diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,122082 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999856620546276,
+  "eval_steps": 500,
+  "global_step": 17436,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.558266639709473,
+      "learning_rate": 3.816793893129771e-08,
+      "loss": 1.6675,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.988743782043457,
+      "learning_rate": 7.633587786259542e-08,
+      "loss": 1.5673,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.858505249023438,
+      "learning_rate": 1.1450381679389314e-07,
+      "loss": 1.0491,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.732449531555176,
+      "learning_rate": 1.5267175572519085e-07,
+      "loss": 1.5693,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.124704360961914,
+      "learning_rate": 1.9083969465648858e-07,
+      "loss": 1.6515,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.981911659240723,
+      "learning_rate": 2.2900763358778629e-07,
+      "loss": 1.6218,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.896223068237305,
+      "learning_rate": 2.67175572519084e-07,
+      "loss": 1.6901,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.718731880187988,
+      "learning_rate": 3.053435114503817e-07,
+      "loss": 1.6153,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.451005935668945,
+      "learning_rate": 3.4351145038167945e-07,
+      "loss": 1.6435,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.86658763885498,
+      "learning_rate": 3.8167938931297716e-07,
+      "loss": 1.6422,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.805266380310059,
+      "learning_rate": 4.1984732824427486e-07,
+      "loss": 1.5509,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.717146873474121,
+      "learning_rate": 4.5801526717557257e-07,
+      "loss": 1.5424,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.827333450317383,
+      "learning_rate": 4.961832061068702e-07,
+      "loss": 1.5796,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.937980651855469,
+      "learning_rate": 5.34351145038168e-07,
+      "loss": 1.5997,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.568488121032715,
+      "learning_rate": 5.725190839694656e-07,
+      "loss": 1.5657,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.67142677307129,
+      "learning_rate": 6.106870229007634e-07,
+      "loss": 0.8693,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.242477893829346,
+      "learning_rate": 6.48854961832061e-07,
+      "loss": 1.4926,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.931870937347412,
+      "learning_rate": 6.870229007633589e-07,
+      "loss": 1.4668,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.434217929840088,
+      "learning_rate": 7.251908396946565e-07,
+      "loss": 1.4932,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.345139980316162,
+      "learning_rate": 7.633587786259543e-07,
+      "loss": 1.5443,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.9464006423950195,
+      "learning_rate": 8.01526717557252e-07,
+      "loss": 1.4134,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.003053665161133,
+      "learning_rate": 8.396946564885497e-07,
+      "loss": 1.431,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.969738721847534,
+      "learning_rate": 8.778625954198474e-07,
+      "loss": 1.431,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.9027457237243652,
+      "learning_rate": 9.160305343511451e-07,
+      "loss": 1.3926,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.465153694152832,
+      "learning_rate": 9.54198473282443e-07,
+      "loss": 1.4641,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.1631879806518555,
+      "learning_rate": 9.923664122137404e-07,
+      "loss": 1.4377,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.10839319229126,
+      "learning_rate": 1.0305343511450382e-06,
+      "loss": 1.4771,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.254977226257324,
+      "learning_rate": 1.068702290076336e-06,
+      "loss": 1.4225,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8101425170898438,
+      "learning_rate": 1.1068702290076337e-06,
+      "loss": 1.44,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.160100936889648,
+      "learning_rate": 1.1450381679389313e-06,
+      "loss": 1.4205,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5531253814697266,
+      "learning_rate": 1.1832061068702292e-06,
+      "loss": 1.5099,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6297051906585693,
+      "learning_rate": 1.2213740458015268e-06,
+      "loss": 1.4695,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.042447566986084,
+      "learning_rate": 1.2595419847328243e-06,
+      "loss": 1.3552,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2630228996276855,
+      "learning_rate": 1.297709923664122e-06,
+      "loss": 1.3628,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.807941436767578,
+      "learning_rate": 1.33587786259542e-06,
+      "loss": 1.4268,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7850215435028076,
+      "learning_rate": 1.3740458015267178e-06,
+      "loss": 1.3714,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.831404447555542,
+      "learning_rate": 1.4122137404580156e-06,
+      "loss": 1.3748,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.020758628845215,
+      "learning_rate": 1.450381679389313e-06,
+      "loss": 1.3267,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6790385246276855,
+      "learning_rate": 1.4885496183206109e-06,
+      "loss": 1.4149,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7830984592437744,
+      "learning_rate": 1.5267175572519086e-06,
+      "loss": 1.3075,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0869674682617188,
+      "learning_rate": 1.5648854961832064e-06,
+      "loss": 1.3319,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6376333236694336,
+      "learning_rate": 1.603053435114504e-06,
+      "loss": 1.4106,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0565459728240967,
+      "learning_rate": 1.6412213740458017e-06,
+      "loss": 1.3215,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8039722442626953,
+      "learning_rate": 1.6793893129770995e-06,
+      "loss": 1.3915,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.494558811187744,
+      "learning_rate": 1.7175572519083972e-06,
+      "loss": 1.251,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8768439292907715,
+      "learning_rate": 1.7557251908396948e-06,
+      "loss": 1.3015,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.027677536010742,
+      "learning_rate": 1.7938931297709925e-06,
+      "loss": 1.2823,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.196527481079102,
+      "learning_rate": 1.8320610687022903e-06,
+      "loss": 0.8777,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.764543056488037,
+      "learning_rate": 1.870229007633588e-06,
+      "loss": 1.2758,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6398603916168213,
+      "learning_rate": 1.908396946564886e-06,
+      "loss": 1.3325,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.84912395477295,
+      "learning_rate": 1.946564885496183e-06,
+      "loss": 0.8353,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.275533437728882,
+      "learning_rate": 1.984732824427481e-06,
+      "loss": 1.2948,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6710312366485596,
+      "learning_rate": 2.0229007633587786e-06,
+      "loss": 1.3265,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5640649795532227,
+      "learning_rate": 2.0610687022900764e-06,
+      "loss": 1.3774,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6674256324768066,
+      "learning_rate": 2.099236641221374e-06,
+      "loss": 1.3859,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7106809616088867,
+      "learning_rate": 2.137404580152672e-06,
+      "loss": 1.3243,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6420164108276367,
+      "learning_rate": 2.1755725190839697e-06,
+      "loss": 1.427,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4347591400146484,
+      "learning_rate": 2.2137404580152674e-06,
+      "loss": 1.3067,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5336878299713135,
+      "learning_rate": 2.2519083969465648e-06,
+      "loss": 1.3438,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2951669692993164,
+      "learning_rate": 2.2900763358778625e-06,
+      "loss": 1.2477,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7502987384796143,
+      "learning_rate": 2.3282442748091603e-06,
+      "loss": 1.2977,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4772419929504395,
+      "learning_rate": 2.3664122137404585e-06,
+      "loss": 1.2858,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3816637992858887,
+      "learning_rate": 2.4045801526717562e-06,
+      "loss": 1.2402,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4924306869506836,
+      "learning_rate": 2.4427480916030536e-06,
+      "loss": 1.2361,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4357974529266357,
+      "learning_rate": 2.4809160305343513e-06,
+      "loss": 1.3761,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.770843029022217,
+      "learning_rate": 2.5190839694656487e-06,
+      "loss": 0.6892,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.287750005722046,
+      "learning_rate": 2.5572519083969464e-06,
+      "loss": 1.359,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.177046298980713,
+      "learning_rate": 2.595419847328244e-06,
+      "loss": 1.2689,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.238996982574463,
+      "learning_rate": 2.633587786259542e-06,
+      "loss": 1.3359,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.560926675796509,
+      "learning_rate": 2.67175572519084e-06,
+      "loss": 1.2537,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5880041122436523,
+      "learning_rate": 2.709923664122138e-06,
+      "loss": 1.3544,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2030255794525146,
+      "learning_rate": 2.7480916030534356e-06,
+      "loss": 1.2867,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4377143383026123,
+      "learning_rate": 2.7862595419847334e-06,
+      "loss": 1.2642,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3077785968780518,
+      "learning_rate": 2.824427480916031e-06,
+      "loss": 1.3151,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2587642669677734,
+      "learning_rate": 2.862595419847328e-06,
+      "loss": 1.2545,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4077560901641846,
+      "learning_rate": 2.900763358778626e-06,
+      "loss": 1.3183,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.194283962249756,
+      "learning_rate": 2.938931297709924e-06,
+      "loss": 1.2414,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.358806848526001,
+      "learning_rate": 2.9770992366412218e-06,
+      "loss": 1.2463,
+      "step": 78
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3932883739471436,
+      "learning_rate": 3.0152671755725195e-06,
+      "loss": 1.1616,
+      "step": 79
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4160308837890625,
+      "learning_rate": 3.0534351145038173e-06,
+      "loss": 1.1614,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3357017040252686,
+      "learning_rate": 3.091603053435115e-06,
+      "loss": 1.2228,
+      "step": 81
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3943848609924316,
+      "learning_rate": 3.129770992366413e-06,
+      "loss": 1.3283,
+      "step": 82
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.553863525390625,
+      "learning_rate": 3.1679389312977097e-06,
+      "loss": 1.2791,
+      "step": 83
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3540492057800293,
+      "learning_rate": 3.206106870229008e-06,
+      "loss": 1.2112,
+      "step": 84
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.340452194213867,
+      "learning_rate": 3.2442748091603056e-06,
+      "loss": 1.2269,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4069883823394775,
+      "learning_rate": 3.2824427480916034e-06,
+      "loss": 1.207,
+      "step": 86
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.083019256591797,
+      "learning_rate": 3.320610687022901e-06,
+      "loss": 1.1862,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1076905727386475,
+      "learning_rate": 3.358778625954199e-06,
+      "loss": 1.267,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.48537540435791,
+      "learning_rate": 3.3969465648854967e-06,
+      "loss": 1.3569,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5323920249938965,
+      "learning_rate": 3.4351145038167944e-06,
+      "loss": 1.2864,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3568899631500244,
+      "learning_rate": 3.473282442748092e-06,
+      "loss": 1.3478,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.498481273651123,
+      "learning_rate": 3.5114503816793895e-06,
+      "loss": 1.2504,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.007481336593628,
+      "learning_rate": 3.5496183206106873e-06,
+      "loss": 1.1569,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.409057378768921,
+      "learning_rate": 3.587786259541985e-06,
+      "loss": 1.2249,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.408141136169434,
+      "learning_rate": 3.625954198473283e-06,
+      "loss": 1.2368,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.244300603866577,
+      "learning_rate": 3.6641221374045806e-06,
+      "loss": 1.3195,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4743666648864746,
+      "learning_rate": 3.7022900763358783e-06,
+      "loss": 1.2646,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.102954149246216,
+      "learning_rate": 3.740458015267176e-06,
+      "loss": 0.5534,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.413912534713745,
+      "learning_rate": 3.778625954198474e-06,
+      "loss": 1.2545,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.450010299682617,
+      "learning_rate": 3.816793893129772e-06,
+      "loss": 1.2882,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2480010986328125,
+      "learning_rate": 3.8549618320610685e-06,
+      "loss": 1.2514,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.745875358581543,
+      "learning_rate": 3.893129770992366e-06,
+      "loss": 1.1647,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5601766109466553,
+      "learning_rate": 3.931297709923664e-06,
+      "loss": 1.2509,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.514998435974121,
+      "learning_rate": 3.969465648854962e-06,
+      "loss": 1.2079,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.424623727798462,
+      "learning_rate": 4.0076335877862595e-06,
+      "loss": 1.2166,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2762906551361084,
+      "learning_rate": 4.045801526717557e-06,
+      "loss": 1.2389,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.150784969329834,
+      "learning_rate": 4.083969465648855e-06,
+      "loss": 1.1847,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3234362602233887,
+      "learning_rate": 4.122137404580153e-06,
+      "loss": 1.2278,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4249842166900635,
+      "learning_rate": 4.1603053435114506e-06,
+      "loss": 1.207,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2878878116607666,
+      "learning_rate": 4.198473282442748e-06,
+      "loss": 1.2264,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.549649953842163,
+      "learning_rate": 4.236641221374046e-06,
+      "loss": 1.159,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4554660320281982,
+      "learning_rate": 4.274809160305344e-06,
+      "loss": 1.319,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0981709957122803,
+      "learning_rate": 4.312977099236642e-06,
+      "loss": 1.2092,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.541743040084839,
+      "learning_rate": 4.351145038167939e-06,
+      "loss": 1.1693,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5217630863189697,
+      "learning_rate": 4.389312977099237e-06,
+      "loss": 1.2889,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.495091199874878,
+      "learning_rate": 4.427480916030535e-06,
+      "loss": 1.1888,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4056596755981445,
+      "learning_rate": 4.465648854961833e-06,
+      "loss": 1.1838,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.335122585296631,
+      "learning_rate": 4.5038167938931296e-06,
+      "loss": 1.2866,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1947944164276123,
+      "learning_rate": 4.541984732824427e-06,
+      "loss": 1.2358,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.29399037361145,
+      "learning_rate": 4.580152671755725e-06,
+      "loss": 1.2221,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5239224433898926,
+      "learning_rate": 4.618320610687023e-06,
+      "loss": 1.2033,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3639538288116455,
+      "learning_rate": 4.656488549618321e-06,
+      "loss": 1.2669,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1837053298950195,
+      "learning_rate": 4.694656488549618e-06,
+      "loss": 1.2633,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.715895175933838,
+      "learning_rate": 4.732824427480917e-06,
+      "loss": 0.6809,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.198807716369629,
+      "learning_rate": 4.770992366412215e-06,
+      "loss": 1.2618,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3895256519317627,
+      "learning_rate": 4.8091603053435125e-06,
+      "loss": 1.1939,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3140602111816406,
+      "learning_rate": 4.847328244274809e-06,
+      "loss": 1.3025,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.333887815475464,
+      "learning_rate": 4.885496183206107e-06,
+      "loss": 1.2884,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.109236478805542,
+      "learning_rate": 4.923664122137405e-06,
+      "loss": 1.2517,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2595057487487793,
+      "learning_rate": 4.961832061068703e-06,
+      "loss": 1.2674,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.338866710662842,
+      "learning_rate": 5e-06,
+      "loss": 1.2765,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3303911685943604,
+      "learning_rate": 5.038167938931297e-06,
+      "loss": 1.1896,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8968067169189453,
+      "learning_rate": 5.076335877862596e-06,
+      "loss": 1.2421,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.240675687789917,
+      "learning_rate": 5.114503816793893e-06,
+      "loss": 1.2187,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4179251194000244,
+      "learning_rate": 5.1526717557251914e-06,
+      "loss": 1.2101,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.381080389022827,
+      "learning_rate": 5.190839694656488e-06,
+      "loss": 1.2158,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3089752197265625,
+      "learning_rate": 5.229007633587787e-06,
+      "loss": 1.2186,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6404616832733154,
+      "learning_rate": 5.267175572519084e-06,
+      "loss": 1.2566,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.435467004776001,
+      "learning_rate": 5.3053435114503825e-06,
+      "loss": 1.1878,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.396341562271118,
+      "learning_rate": 5.34351145038168e-06,
+      "loss": 1.3049,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2711474895477295,
+      "learning_rate": 5.381679389312977e-06,
+      "loss": 1.1792,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1534476280212402,
+      "learning_rate": 5.419847328244276e-06,
+      "loss": 1.2657,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.453995704650879,
+      "learning_rate": 5.458015267175573e-06,
+      "loss": 1.1267,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.391569137573242,
+      "learning_rate": 5.496183206106871e-06,
+      "loss": 1.2367,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3522192239761353,
+      "learning_rate": 5.534351145038168e-06,
+      "loss": 0.6216,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.379929304122925,
+      "learning_rate": 5.572519083969467e-06,
+      "loss": 1.2342,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2897443771362305,
+      "learning_rate": 5.610687022900764e-06,
+      "loss": 1.2379,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2341549396514893,
+      "learning_rate": 5.648854961832062e-06,
+      "loss": 1.2158,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.227285146713257,
+      "learning_rate": 5.687022900763359e-06,
+      "loss": 1.2156,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.068729877471924,
+      "learning_rate": 5.725190839694656e-06,
+      "loss": 1.1695,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.238799810409546,
+      "learning_rate": 5.763358778625955e-06,
+      "loss": 1.1904,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5148062705993652,
+      "learning_rate": 5.801526717557252e-06,
+      "loss": 1.2824,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.326028347015381,
+      "learning_rate": 5.83969465648855e-06,
+      "loss": 1.1364,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5881175994873047,
+      "learning_rate": 5.877862595419848e-06,
+      "loss": 1.2886,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.399954080581665,
+      "learning_rate": 5.916030534351146e-06,
+      "loss": 1.1618,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.169663190841675,
+      "learning_rate": 5.9541984732824435e-06,
+      "loss": 1.114,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.353415012359619,
+      "learning_rate": 5.992366412213741e-06,
+      "loss": 1.2032,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1866724491119385,
+      "learning_rate": 6.030534351145039e-06,
+      "loss": 1.2034,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.327688455581665,
+      "learning_rate": 6.068702290076336e-06,
+      "loss": 1.1703,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3042783737182617,
+      "learning_rate": 6.1068702290076346e-06,
+      "loss": 1.2322,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.503833293914795,
+      "learning_rate": 6.1450381679389315e-06,
+      "loss": 1.1789,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.350928783416748,
+      "learning_rate": 6.18320610687023e-06,
+      "loss": 1.236,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4380338191986084,
+      "learning_rate": 6.221374045801527e-06,
+      "loss": 1.1927,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.268906593322754,
+      "learning_rate": 6.259541984732826e-06,
+      "loss": 1.2072,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.231147527694702,
+      "learning_rate": 6.2977099236641225e-06,
+      "loss": 1.2117,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.418914794921875,
+      "learning_rate": 6.335877862595419e-06,
+      "loss": 1.1606,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3540070056915283,
+      "learning_rate": 6.374045801526718e-06,
+      "loss": 1.2334,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2606024742126465,
+      "learning_rate": 6.412213740458016e-06,
+      "loss": 1.12,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2669661045074463,
+      "learning_rate": 6.4503816793893135e-06,
+      "loss": 1.204,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.357459306716919,
+      "learning_rate": 6.488549618320611e-06,
+      "loss": 1.2345,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.168829917907715,
+      "learning_rate": 6.526717557251909e-06,
+      "loss": 1.2474,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1880061626434326,
+      "learning_rate": 6.564885496183207e-06,
+      "loss": 1.1261,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.262659788131714,
+      "learning_rate": 6.6030534351145046e-06,
+      "loss": 1.1857,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3341357707977295,
+      "learning_rate": 6.641221374045802e-06,
+      "loss": 1.1984,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.26676869392395,
+      "learning_rate": 6.679389312977099e-06,
+      "loss": 1.146,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5866336822509766,
+      "learning_rate": 6.717557251908398e-06,
+      "loss": 1.2126,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.249685525894165,
+      "learning_rate": 6.755725190839695e-06,
+      "loss": 1.2455,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.474029779434204,
+      "learning_rate": 6.793893129770993e-06,
+      "loss": 1.2297,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3840372562408447,
+      "learning_rate": 6.83206106870229e-06,
+      "loss": 1.146,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0976841449737549,
+      "learning_rate": 6.870229007633589e-06,
+      "loss": 0.5018,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.552171468734741,
+      "learning_rate": 6.908396946564886e-06,
+      "loss": 1.2003,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7230982780456543,
+      "learning_rate": 6.946564885496184e-06,
+      "loss": 1.1734,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2825429439544678,
+      "learning_rate": 6.984732824427481e-06,
+      "loss": 1.2047,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2877326011657715,
+      "learning_rate": 7.022900763358779e-06,
+      "loss": 1.1578,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.451024055480957,
+      "learning_rate": 7.061068702290077e-06,
+      "loss": 1.2439,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2070621252059937,
+      "learning_rate": 7.0992366412213746e-06,
+      "loss": 0.5192,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5960514545440674,
+      "learning_rate": 7.137404580152672e-06,
+      "loss": 1.3137,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5490591526031494,
+      "learning_rate": 7.17557251908397e-06,
+      "loss": 1.2176,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.679558038711548,
+      "learning_rate": 7.213740458015268e-06,
+      "loss": 1.1846,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.649095296859741,
+      "learning_rate": 7.251908396946566e-06,
+      "loss": 1.1863,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5023233890533447,
+      "learning_rate": 7.290076335877863e-06,
+      "loss": 1.1572,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1537967920303345,
+      "learning_rate": 7.328244274809161e-06,
+      "loss": 0.5124,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.515532970428467,
+      "learning_rate": 7.366412213740458e-06,
+      "loss": 1.0807,
+      "step": 193
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3911054134368896,
+      "learning_rate": 7.404580152671757e-06,
+      "loss": 1.211,
+      "step": 194
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3482248783111572,
+      "learning_rate": 7.4427480916030536e-06,
+      "loss": 1.1568,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3433539867401123,
+      "learning_rate": 7.480916030534352e-06,
+      "loss": 1.1664,
+      "step": 196
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.395803689956665,
+      "learning_rate": 7.519083969465649e-06,
+      "loss": 1.2107,
+      "step": 197
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.49005126953125,
+      "learning_rate": 7.557251908396948e-06,
+      "loss": 1.2141,
+      "step": 198
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.116427898406982,
+      "learning_rate": 7.595419847328245e-06,
+      "loss": 1.0903,
+      "step": 199
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6343843936920166,
+      "learning_rate": 7.633587786259543e-06,
+      "loss": 1.0758,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.756282329559326,
+      "learning_rate": 7.671755725190841e-06,
+      "loss": 1.1098,
+      "step": 201
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.110990285873413,
+      "learning_rate": 7.709923664122137e-06,
+      "loss": 1.1557,
+      "step": 202
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.480382204055786,
+      "learning_rate": 7.748091603053436e-06,
+      "loss": 1.1931,
+      "step": 203
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.264718532562256,
+      "learning_rate": 7.786259541984733e-06,
+      "loss": 1.1575,
+      "step": 204
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4612679481506348,
+      "learning_rate": 7.824427480916032e-06,
+      "loss": 1.1488,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4971280097961426,
+      "learning_rate": 7.862595419847328e-06,
+      "loss": 1.2338,
+      "step": 206
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.420489549636841,
+      "learning_rate": 7.900763358778627e-06,
+      "loss": 1.187,
+      "step": 207
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2725765705108643,
+      "learning_rate": 7.938931297709924e-06,
+      "loss": 1.169,
+      "step": 208
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.435575485229492,
+      "learning_rate": 7.977099236641223e-06,
+      "loss": 1.2375,
+      "step": 209
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2475221157073975,
+      "learning_rate": 8.015267175572519e-06,
+      "loss": 1.1516,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1970536708831787,
+      "learning_rate": 8.053435114503817e-06,
+      "loss": 1.1942,
+      "step": 211
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1885790824890137,
+      "learning_rate": 8.091603053435115e-06,
+      "loss": 1.2051,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4636831283569336,
+      "learning_rate": 8.129770992366412e-06,
+      "loss": 1.1964,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.704554319381714,
+      "learning_rate": 8.16793893129771e-06,
+      "loss": 1.177,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.538517475128174,
+      "learning_rate": 8.206106870229008e-06,
+      "loss": 1.1312,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2675399780273438,
+      "learning_rate": 8.244274809160306e-06,
+      "loss": 1.1574,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3015332221984863,
+      "learning_rate": 8.282442748091603e-06,
+      "loss": 1.1842,
+      "step": 217
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1983731985092163,
+      "learning_rate": 8.320610687022901e-06,
+      "loss": 0.5498,
+      "step": 218
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3395638465881348,
+      "learning_rate": 8.358778625954199e-06,
+      "loss": 1.1192,
+      "step": 219
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7942235469818115,
+      "learning_rate": 8.396946564885497e-06,
+      "loss": 1.2464,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.191277027130127,
+      "learning_rate": 8.435114503816794e-06,
+      "loss": 1.1368,
+      "step": 221
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.591585397720337,
+      "learning_rate": 8.473282442748092e-06,
+      "loss": 1.0736,
+      "step": 222
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.48899507522583,
+      "learning_rate": 8.51145038167939e-06,
+      "loss": 1.2223,
+      "step": 223
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.433845281600952,
+      "learning_rate": 8.549618320610688e-06,
+      "loss": 1.1786,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.364865303039551,
+      "learning_rate": 8.587786259541985e-06,
+      "loss": 1.1529,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3349416255950928,
+      "learning_rate": 8.625954198473283e-06,
+      "loss": 1.1746,
+      "step": 226
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.141732692718506,
+      "learning_rate": 8.664122137404581e-06,
+      "loss": 1.2164,
+      "step": 227
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3822381496429443,
+      "learning_rate": 8.702290076335879e-06,
+      "loss": 1.1444,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4650022983551025,
+      "learning_rate": 8.740458015267176e-06,
+      "loss": 1.1352,
+      "step": 229
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.655257225036621,
+      "learning_rate": 8.778625954198474e-06,
+      "loss": 1.2208,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4323301315307617,
+      "learning_rate": 8.816793893129772e-06,
+      "loss": 1.1644,
+      "step": 231
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3118544816970825,
+      "learning_rate": 8.85496183206107e-06,
+      "loss": 0.5754,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6206159591674805,
+      "learning_rate": 8.893129770992368e-06,
+      "loss": 1.2652,
+      "step": 233
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5270016193389893,
+      "learning_rate": 8.931297709923665e-06,
+      "loss": 1.0734,
+      "step": 234
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.644437551498413,
+      "learning_rate": 8.969465648854963e-06,
+      "loss": 1.1781,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3275222778320312,
+      "learning_rate": 9.007633587786259e-06,
+      "loss": 1.1713,
+      "step": 236
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2932794094085693,
+      "learning_rate": 9.045801526717559e-06,
+      "loss": 1.1651,
+      "step": 237
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.413956642150879,
+      "learning_rate": 9.083969465648855e-06,
+      "loss": 1.2262,
+      "step": 238
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.231199026107788,
+      "learning_rate": 9.122137404580154e-06,
+      "loss": 1.0769,
+      "step": 239
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3945555686950684,
+      "learning_rate": 9.16030534351145e-06,
+      "loss": 1.1812,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.110468626022339,
+      "learning_rate": 9.19847328244275e-06,
+      "loss": 1.2368,
+      "step": 241
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3194801807403564,
+      "learning_rate": 9.236641221374046e-06,
+      "loss": 1.2372,
+      "step": 242
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.27785325050354,
+      "learning_rate": 9.274809160305345e-06,
+      "loss": 1.1542,
+      "step": 243
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.417180061340332,
+      "learning_rate": 9.312977099236641e-06,
+      "loss": 1.2068,
+      "step": 244
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3117575645446777,
+      "learning_rate": 9.351145038167939e-06,
+      "loss": 1.1361,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.495323896408081,
+      "learning_rate": 9.389312977099237e-06,
+      "loss": 1.2174,
+      "step": 246
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.228602170944214,
+      "learning_rate": 9.427480916030534e-06,
+      "loss": 1.1987,
+      "step": 247
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.187556505203247,
+      "learning_rate": 9.465648854961834e-06,
+      "loss": 1.1911,
+      "step": 248
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5458500385284424,
+      "learning_rate": 9.50381679389313e-06,
+      "loss": 1.132,
+      "step": 249
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.205439805984497,
+      "learning_rate": 9.54198473282443e-06,
+      "loss": 1.0681,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2369577884674072,
+      "learning_rate": 9.580152671755725e-06,
+      "loss": 1.2095,
+      "step": 251
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3744304180145264,
+      "learning_rate": 9.618320610687025e-06,
+      "loss": 1.2271,
+      "step": 252
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2684192657470703,
+      "learning_rate": 9.656488549618321e-06,
+      "loss": 1.0788,
+      "step": 253
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1698288917541504,
+      "learning_rate": 9.694656488549619e-06,
+      "loss": 1.1339,
+      "step": 254
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4651522636413574,
+      "learning_rate": 9.732824427480917e-06,
+      "loss": 1.1854,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.563157081604004,
+      "learning_rate": 9.770992366412214e-06,
+      "loss": 1.1689,
+      "step": 256
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.057300090789795,
+      "learning_rate": 9.809160305343512e-06,
+      "loss": 0.602,
+      "step": 257
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.101321816444397,
+      "learning_rate": 9.84732824427481e-06,
+      "loss": 0.4822,
+      "step": 258
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.476306200027466,
+      "learning_rate": 9.885496183206108e-06,
+      "loss": 1.2241,
+      "step": 259
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.863455057144165,
+      "learning_rate": 9.923664122137405e-06,
+      "loss": 1.1446,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.314697742462158,
+      "learning_rate": 9.961832061068703e-06,
+      "loss": 1.1508,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6635262966156006,
+      "learning_rate": 1e-05,
+      "loss": 1.2458,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.232550859451294,
+      "learning_rate": 1.0038167938931299e-05,
+      "loss": 1.2193,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.306417465209961,
+      "learning_rate": 1.0076335877862595e-05,
+      "loss": 1.1676,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8053040504455566,
+      "learning_rate": 1.0114503816793894e-05,
+      "loss": 1.207,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5533223152160645,
+      "learning_rate": 1.0152671755725192e-05,
+      "loss": 1.1838,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.469116449356079,
+      "learning_rate": 1.019083969465649e-05,
+      "loss": 1.2508,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3705906867980957,
+      "learning_rate": 1.0229007633587786e-05,
+      "loss": 1.1399,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1668001413345337,
+      "learning_rate": 1.0267175572519085e-05,
+      "loss": 0.5029,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4078330993652344,
+      "learning_rate": 1.0305343511450383e-05,
+      "loss": 1.1489,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.458362102508545,
+      "learning_rate": 1.034351145038168e-05,
+      "loss": 1.1049,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5195271968841553,
+      "learning_rate": 1.0381679389312977e-05,
+      "loss": 1.2086,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2613472938537598,
+      "learning_rate": 1.0419847328244274e-05,
+      "loss": 1.0938,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3890202045440674,
+      "learning_rate": 1.0458015267175574e-05,
+      "loss": 1.1137,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.255711317062378,
+      "learning_rate": 1.0496183206106872e-05,
+      "loss": 1.2094,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0289547443389893,
+      "learning_rate": 1.0534351145038168e-05,
+      "loss": 0.5386,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.49617600440979,
+      "learning_rate": 1.0572519083969465e-05,
+      "loss": 1.2289,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.490529775619507,
+      "learning_rate": 1.0610687022900765e-05,
+      "loss": 1.1398,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4953157901763916,
+      "learning_rate": 1.0648854961832063e-05,
+      "loss": 1.2283,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.51371169090271,
+      "learning_rate": 1.068702290076336e-05,
+      "loss": 1.1693,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.34928560256958,
+      "learning_rate": 1.0725190839694657e-05,
+      "loss": 1.0796,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3046045303344727,
+      "learning_rate": 1.0763358778625954e-05,
+      "loss": 1.2126,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4181103706359863,
+      "learning_rate": 1.0801526717557254e-05,
+      "loss": 1.0868,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4339473247528076,
+      "learning_rate": 1.0839694656488552e-05,
+      "loss": 1.1992,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1299593448638916,
+      "learning_rate": 1.0877862595419848e-05,
+      "loss": 1.1235,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2972185611724854,
+      "learning_rate": 1.0916030534351145e-05,
+      "loss": 1.1303,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.604118585586548,
+      "learning_rate": 1.0954198473282445e-05,
+      "loss": 1.1724,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3645050525665283,
+      "learning_rate": 1.0992366412213743e-05,
+      "loss": 1.3014,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4452872276306152,
+      "learning_rate": 1.1030534351145039e-05,
+      "loss": 1.1139,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1248505115509033,
+      "learning_rate": 1.1068702290076336e-05,
+      "loss": 1.1383,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3322041034698486,
+      "learning_rate": 1.1106870229007634e-05,
+      "loss": 1.2256,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.192044734954834,
+      "learning_rate": 1.1145038167938934e-05,
+      "loss": 1.2323,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5046188831329346,
+      "learning_rate": 1.118320610687023e-05,
+      "loss": 1.1631,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.338428258895874,
+      "learning_rate": 1.1221374045801527e-05,
+      "loss": 1.174,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.187605619430542,
+      "learning_rate": 1.1259541984732825e-05,
+      "loss": 1.1683,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1554183959960938,
+      "learning_rate": 1.1297709923664125e-05,
+      "loss": 1.1883,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3376195430755615,
+      "learning_rate": 1.133587786259542e-05,
+      "loss": 1.2237,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1107397079467773,
+      "learning_rate": 1.1374045801526718e-05,
+      "loss": 1.1311,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.434945583343506,
+      "learning_rate": 1.1412213740458016e-05,
+      "loss": 1.2061,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3518757820129395,
+      "learning_rate": 1.1450381679389312e-05,
+      "loss": 1.1361,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.270901679992676,
+      "learning_rate": 1.1488549618320612e-05,
+      "loss": 1.1548,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.242243528366089,
+      "learning_rate": 1.152671755725191e-05,
+      "loss": 1.181,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2376561164855957,
+      "learning_rate": 1.1564885496183207e-05,
+      "loss": 1.1291,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4677982330322266,
+      "learning_rate": 1.1603053435114503e-05,
+      "loss": 1.128,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2829904556274414,
+      "learning_rate": 1.1641221374045803e-05,
+      "loss": 1.1789,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0435593128204346,
+      "learning_rate": 1.16793893129771e-05,
+      "loss": 1.0958,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1729190349578857,
+      "learning_rate": 1.1717557251908398e-05,
+      "loss": 1.1834,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.403015613555908,
+      "learning_rate": 1.1755725190839696e-05,
+      "loss": 1.1334,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.153867244720459,
+      "learning_rate": 1.1793893129770992e-05,
+      "loss": 1.1676,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1921372413635254,
+      "learning_rate": 1.1832061068702292e-05,
+      "loss": 1.1697,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.451202154159546,
+      "learning_rate": 1.187022900763359e-05,
+      "loss": 1.2018,
+      "step": 311
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.538238763809204,
+      "learning_rate": 1.1908396946564887e-05,
+      "loss": 1.1653,
+      "step": 312
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3259646892547607,
+      "learning_rate": 1.1946564885496183e-05,
+      "loss": 1.277,
+      "step": 313
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5378835201263428,
+      "learning_rate": 1.1984732824427483e-05,
+      "loss": 1.167,
+      "step": 314
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.323946714401245,
+      "learning_rate": 1.202290076335878e-05,
+      "loss": 1.1798,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3234386444091797,
+      "learning_rate": 1.2061068702290078e-05,
+      "loss": 1.1706,
+      "step": 316
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.256312847137451,
+      "learning_rate": 1.2099236641221374e-05,
+      "loss": 1.2463,
+      "step": 317
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.189204454421997,
+      "learning_rate": 1.2137404580152672e-05,
+      "loss": 1.1542,
+      "step": 318
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2717251777648926,
+      "learning_rate": 1.2175572519083971e-05,
+      "loss": 1.1562,
+      "step": 319
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.362816572189331,
+      "learning_rate": 1.2213740458015269e-05,
+      "loss": 1.0983,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5264642238616943,
+      "learning_rate": 1.2251908396946565e-05,
+      "loss": 1.1484,
+      "step": 321
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1048452854156494,
+      "learning_rate": 1.2290076335877863e-05,
+      "loss": 0.5417,
+      "step": 322
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.439535140991211,
+      "learning_rate": 1.2328244274809162e-05,
+      "loss": 1.1397,
+      "step": 323
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.338682174682617,
+      "learning_rate": 1.236641221374046e-05,
+      "loss": 1.1522,
+      "step": 324
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.173095703125,
+      "learning_rate": 1.2404580152671756e-05,
+      "loss": 1.2423,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2724859714508057,
+      "learning_rate": 1.2442748091603054e-05,
+      "loss": 1.1686,
+      "step": 326
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3127293586730957,
+      "learning_rate": 1.2480916030534352e-05,
+      "loss": 1.18,
+      "step": 327
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1922414302825928,
+      "learning_rate": 1.2519083969465651e-05,
+      "loss": 1.2393,
+      "step": 328
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.249760866165161,
+      "learning_rate": 1.2557251908396947e-05,
+      "loss": 1.1924,
+      "step": 329
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5361180305480957,
+      "learning_rate": 1.2595419847328245e-05,
+      "loss": 1.2124,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4809746742248535,
+      "learning_rate": 1.2633587786259543e-05,
+      "loss": 1.1432,
+      "step": 331
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.598536252975464,
+      "learning_rate": 1.2671755725190839e-05,
+      "loss": 1.1612,
+      "step": 332
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3069939613342285,
+      "learning_rate": 1.2709923664122138e-05,
+      "loss": 1.0567,
+      "step": 333
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.311523199081421,
+      "learning_rate": 1.2748091603053436e-05,
+      "loss": 1.2515,
+      "step": 334
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0850589275360107,
+      "learning_rate": 1.2786259541984734e-05,
+      "loss": 1.0596,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3131818771362305,
+      "learning_rate": 1.2824427480916032e-05,
+      "loss": 1.1662,
+      "step": 336
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3318216800689697,
+      "learning_rate": 1.2862595419847331e-05,
+      "loss": 1.1282,
+      "step": 337
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2734594345092773,
+      "learning_rate": 1.2900763358778627e-05,
+      "loss": 1.1454,
+      "step": 338
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7358264923095703,
+      "learning_rate": 1.2938931297709925e-05,
+      "loss": 1.0995,
+      "step": 339
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.342430830001831,
+      "learning_rate": 1.2977099236641223e-05,
+      "loss": 1.1517,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.223081350326538,
+      "learning_rate": 1.3015267175572519e-05,
+      "loss": 1.1219,
+      "step": 341
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2671773433685303,
+      "learning_rate": 1.3053435114503818e-05,
+      "loss": 1.2077,
+      "step": 342
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.451024293899536,
+      "learning_rate": 1.3091603053435116e-05,
+      "loss": 1.152,
+      "step": 343
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.571594715118408,
+      "learning_rate": 1.3129770992366414e-05,
+      "loss": 1.2044,
+      "step": 344
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4298548698425293,
+      "learning_rate": 1.316793893129771e-05,
+      "loss": 1.1181,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.064215660095215,
+      "learning_rate": 1.3206106870229009e-05,
+      "loss": 1.1252,
+      "step": 346
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.479543685913086,
+      "learning_rate": 1.3244274809160307e-05,
+      "loss": 1.057,
+      "step": 347
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3734121322631836,
+      "learning_rate": 1.3282442748091605e-05,
+      "loss": 1.2348,
+      "step": 348
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6014323234558105,
+      "learning_rate": 1.33206106870229e-05,
+      "loss": 1.0837,
+      "step": 349
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5107686519622803,
+      "learning_rate": 1.3358778625954198e-05,
+      "loss": 1.1559,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.419914722442627,
+      "learning_rate": 1.3396946564885498e-05,
+      "loss": 1.1412,
+      "step": 351
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.224263906478882,
+      "learning_rate": 1.3435114503816796e-05,
+      "loss": 1.1477,
+      "step": 352
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3645153045654297,
+      "learning_rate": 1.3473282442748092e-05,
+      "loss": 1.1707,
+      "step": 353
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.332211971282959,
+      "learning_rate": 1.351145038167939e-05,
+      "loss": 1.1284,
+      "step": 354
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.29952073097229,
+      "learning_rate": 1.3549618320610689e-05,
+      "loss": 1.1937,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.55173921585083,
+      "learning_rate": 1.3587786259541987e-05,
+      "loss": 1.2429,
+      "step": 356
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.178745746612549,
+      "learning_rate": 1.3625954198473283e-05,
+      "loss": 1.1681,
+      "step": 357
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2342777252197266,
+      "learning_rate": 1.366412213740458e-05,
+      "loss": 1.1756,
+      "step": 358
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4349939823150635,
+      "learning_rate": 1.3702290076335878e-05,
+      "loss": 1.1194,
+      "step": 359
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2043814659118652,
+      "learning_rate": 1.3740458015267178e-05,
+      "loss": 1.0826,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.24560809135437,
+      "learning_rate": 1.3778625954198474e-05,
+      "loss": 1.1475,
+      "step": 361
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.424367666244507,
+      "learning_rate": 1.3816793893129772e-05,
+      "loss": 1.2111,
+      "step": 362
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1970183849334717,
+      "learning_rate": 1.385496183206107e-05,
+      "loss": 1.162,
+      "step": 363
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3800458908081055,
+      "learning_rate": 1.3893129770992369e-05,
+      "loss": 1.1419,
+      "step": 364
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1797505617141724,
+      "learning_rate": 1.3931297709923667e-05,
+      "loss": 0.5251,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.403242349624634,
+      "learning_rate": 1.3969465648854963e-05,
+      "loss": 1.1738,
+      "step": 366
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.359656572341919,
+      "learning_rate": 1.400763358778626e-05,
+      "loss": 1.1734,
+      "step": 367
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3161683082580566,
+      "learning_rate": 1.4045801526717558e-05,
+      "loss": 1.1039,
+      "step": 368
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5263845920562744,
+      "learning_rate": 1.4083969465648858e-05,
+      "loss": 1.1701,
+      "step": 369
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3975117206573486,
+      "learning_rate": 1.4122137404580154e-05,
+      "loss": 1.2048,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2706034183502197,
+      "learning_rate": 1.4160305343511451e-05,
+      "loss": 1.1579,
+      "step": 371
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3223025798797607,
+      "learning_rate": 1.4198473282442749e-05,
+      "loss": 1.151,
+      "step": 372
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.183082103729248,
+      "learning_rate": 1.4236641221374049e-05,
+      "loss": 1.2054,
+      "step": 373
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5176961421966553,
+      "learning_rate": 1.4274809160305345e-05,
+      "loss": 1.2146,
+      "step": 374
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.539336681365967,
+      "learning_rate": 1.4312977099236642e-05,
+      "loss": 1.165,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0185726881027222,
+      "learning_rate": 1.435114503816794e-05,
+      "loss": 0.4935,
+      "step": 376
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5876379013061523,
+      "learning_rate": 1.4389312977099236e-05,
+      "loss": 1.2338,
+      "step": 377
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3072938919067383,
+      "learning_rate": 1.4427480916030536e-05,
+      "loss": 1.0365,
+      "step": 378
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4691827297210693,
+      "learning_rate": 1.4465648854961833e-05,
+      "loss": 1.1449,
+      "step": 379
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.482402801513672,
+      "learning_rate": 1.4503816793893131e-05,
+      "loss": 1.2034,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1270785331726074,
+      "learning_rate": 1.4541984732824427e-05,
+      "loss": 1.0944,
+      "step": 381
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2602219581604004,
+      "learning_rate": 1.4580152671755727e-05,
+      "loss": 1.1551,
+      "step": 382
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.679429292678833,
+      "learning_rate": 1.4618320610687024e-05,
+      "loss": 1.1828,
+      "step": 383
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.602813720703125,
+      "learning_rate": 1.4656488549618322e-05,
+      "loss": 1.2577,
+      "step": 384
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4410369396209717,
+      "learning_rate": 1.4694656488549618e-05,
+      "loss": 1.1418,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3489840030670166,
+      "learning_rate": 1.4732824427480916e-05,
+      "loss": 1.0767,
+      "step": 386
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6419332027435303,
+      "learning_rate": 1.4770992366412216e-05,
+      "loss": 1.143,
+      "step": 387
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2403504848480225,
+      "learning_rate": 1.4809160305343513e-05,
+      "loss": 1.1068,
+      "step": 388
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5824480056762695,
+      "learning_rate": 1.484732824427481e-05,
+      "loss": 1.219,
+      "step": 389
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3028483390808105,
+      "learning_rate": 1.4885496183206107e-05,
+      "loss": 1.2444,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.359236478805542,
+      "learning_rate": 1.4923664122137407e-05,
+      "loss": 1.1451,
+      "step": 391
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.207191228866577,
+      "learning_rate": 1.4961832061068704e-05,
+      "loss": 1.1604,
+      "step": 392
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6509203910827637,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.1415,
+      "step": 393
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4773623943328857,
+      "learning_rate": 1.5038167938931298e-05,
+      "loss": 1.0801,
+      "step": 394
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.52006196975708,
+      "learning_rate": 1.5076335877862596e-05,
+      "loss": 1.198,
+      "step": 395
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1149399280548096,
+      "learning_rate": 1.5114503816793895e-05,
+      "loss": 0.5713,
+      "step": 396
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.519970655441284,
+      "learning_rate": 1.5152671755725193e-05,
+      "loss": 1.1421,
+      "step": 397
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.680042028427124,
+      "learning_rate": 1.519083969465649e-05,
+      "loss": 1.1582,
+      "step": 398
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6061649322509766,
+      "learning_rate": 1.5229007633587787e-05,
+      "loss": 1.2406,
+      "step": 399
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2451324462890625,
+      "learning_rate": 1.5267175572519086e-05,
+      "loss": 0.6114,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6718413829803467,
+      "learning_rate": 1.5305343511450384e-05,
+      "loss": 1.1769,
+      "step": 401
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.228839159011841,
+      "learning_rate": 1.5343511450381682e-05,
+      "loss": 1.144,
+      "step": 402
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2783102989196777,
+      "learning_rate": 1.5381679389312976e-05,
+      "loss": 1.1056,
+      "step": 403
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5894501209259033,
+      "learning_rate": 1.5419847328244274e-05,
+      "loss": 1.1594,
+      "step": 404
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3087751865386963,
+      "learning_rate": 1.5458015267175575e-05,
+      "loss": 1.1738,
+      "step": 405
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3469791412353516,
+      "learning_rate": 1.5496183206106873e-05,
+      "loss": 1.1155,
+      "step": 406
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2417876720428467,
+      "learning_rate": 1.5534351145038167e-05,
+      "loss": 1.0861,
+      "step": 407
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1803297996520996,
+      "learning_rate": 1.5572519083969465e-05,
+      "loss": 1.1145,
+      "step": 408
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.469996929168701,
+      "learning_rate": 1.5610687022900766e-05,
+      "loss": 1.2239,
+      "step": 409
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3944132328033447,
+      "learning_rate": 1.5648854961832064e-05,
+      "loss": 1.1156,
+      "step": 410
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.118701934814453,
+      "learning_rate": 1.5687022900763362e-05,
+      "loss": 1.0973,
+      "step": 411
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.096414804458618,
+      "learning_rate": 1.5725190839694656e-05,
+      "loss": 1.0902,
+      "step": 412
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3658647537231445,
+      "learning_rate": 1.5763358778625954e-05,
+      "loss": 1.1405,
+      "step": 413
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6369361877441406,
+      "learning_rate": 1.5801526717557255e-05,
+      "loss": 1.1106,
+      "step": 414
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9504631757736206,
+      "learning_rate": 1.5839694656488553e-05,
+      "loss": 1.018,
+      "step": 415
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.468120813369751,
+      "learning_rate": 1.5877862595419847e-05,
+      "loss": 1.1755,
+      "step": 416
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.623765707015991,
+      "learning_rate": 1.5916030534351145e-05,
+      "loss": 1.1679,
+      "step": 417
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4666054248809814,
+      "learning_rate": 1.5954198473282446e-05,
+      "loss": 1.1704,
+      "step": 418
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.231243848800659,
+      "learning_rate": 1.5992366412213744e-05,
+      "loss": 1.0936,
+      "step": 419
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.238954782485962,
+      "learning_rate": 1.6030534351145038e-05,
+      "loss": 1.1977,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2670278549194336,
+      "learning_rate": 1.6068702290076336e-05,
+      "loss": 1.125,
+      "step": 421
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.356731414794922,
+      "learning_rate": 1.6106870229007634e-05,
+      "loss": 1.2258,
+      "step": 422
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3338921070098877,
+      "learning_rate": 1.6145038167938935e-05,
+      "loss": 1.1358,
+      "step": 423
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2780981063842773,
+      "learning_rate": 1.618320610687023e-05,
+      "loss": 1.2258,
+      "step": 424
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3498711585998535,
+      "learning_rate": 1.6221374045801527e-05,
+      "loss": 1.1348,
+      "step": 425
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.09081768989563,
+      "learning_rate": 1.6259541984732825e-05,
+      "loss": 1.203,
+      "step": 426
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5033562183380127,
+      "learning_rate": 1.6297709923664126e-05,
+      "loss": 1.1688,
+      "step": 427
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.005417823791504,
+      "learning_rate": 1.633587786259542e-05,
+      "loss": 0.504,
+      "step": 428
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3319952487945557,
+      "learning_rate": 1.6374045801526718e-05,
+      "loss": 1.1287,
+      "step": 429
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2967989444732666,
+      "learning_rate": 1.6412213740458016e-05,
+      "loss": 1.1932,
+      "step": 430
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.175524950027466,
+      "learning_rate": 1.6450381679389314e-05,
+      "loss": 1.1379,
+      "step": 431
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.226332902908325,
+      "learning_rate": 1.648854961832061e-05,
+      "loss": 1.2015,
+      "step": 432
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.348856210708618,
+      "learning_rate": 1.652671755725191e-05,
+      "loss": 1.1079,
+      "step": 433
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.217881679534912,
+      "learning_rate": 1.6564885496183207e-05,
+      "loss": 1.1985,
+      "step": 434
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1928725242614746,
+      "learning_rate": 1.6603053435114505e-05,
+      "loss": 1.1277,
+      "step": 435
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3529856204986572,
+      "learning_rate": 1.6641221374045802e-05,
+      "loss": 1.2094,
+      "step": 436
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.230029344558716,
+      "learning_rate": 1.66793893129771e-05,
+      "loss": 1.1525,
+      "step": 437
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5140163898468018,
+      "learning_rate": 1.6717557251908398e-05,
+      "loss": 1.1015,
+      "step": 438
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.147373914718628,
+      "learning_rate": 1.6755725190839696e-05,
+      "loss": 1.1411,
+      "step": 439
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0082900524139404,
+      "learning_rate": 1.6793893129770993e-05,
+      "loss": 1.1816,
+      "step": 440
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1405038833618164,
+      "learning_rate": 1.683206106870229e-05,
+      "loss": 1.1419,
+      "step": 441
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0462840795516968,
+      "learning_rate": 1.687022900763359e-05,
+      "loss": 0.5146,
+      "step": 442
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9637937545776367,
+      "learning_rate": 1.6908396946564887e-05,
+      "loss": 1.1293,
+      "step": 443
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.505474328994751,
+      "learning_rate": 1.6946564885496184e-05,
+      "loss": 1.1538,
+      "step": 444
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1799428462982178,
+      "learning_rate": 1.6984732824427482e-05,
+      "loss": 1.1244,
+      "step": 445
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3049933910369873,
+      "learning_rate": 1.702290076335878e-05,
+      "loss": 1.1634,
+      "step": 446
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2087562084198,
+      "learning_rate": 1.7061068702290078e-05,
+      "loss": 1.17,
+      "step": 447
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.218954086303711,
+      "learning_rate": 1.7099236641221375e-05,
+      "loss": 1.0968,
+      "step": 448
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.418799877166748,
+      "learning_rate": 1.7137404580152673e-05,
+      "loss": 1.2075,
+      "step": 449
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.171356678009033,
+      "learning_rate": 1.717557251908397e-05,
+      "loss": 1.2747,
+      "step": 450
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5126261711120605,
+      "learning_rate": 1.721374045801527e-05,
+      "loss": 1.2013,
+      "step": 451
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.594771385192871,
+      "learning_rate": 1.7251908396946566e-05,
+      "loss": 1.1434,
+      "step": 452
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4237306118011475,
+      "learning_rate": 1.7290076335877864e-05,
+      "loss": 1.165,
+      "step": 453
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2773149013519287,
+      "learning_rate": 1.7328244274809162e-05,
+      "loss": 1.1997,
+      "step": 454
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.128214120864868,
+      "learning_rate": 1.736641221374046e-05,
+      "loss": 1.1609,
+      "step": 455
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2007956504821777,
+      "learning_rate": 1.7404580152671757e-05,
+      "loss": 1.0923,
+      "step": 456
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4340739250183105,
+      "learning_rate": 1.7442748091603055e-05,
+      "loss": 1.1317,
+      "step": 457
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1774098873138428,
+      "learning_rate": 1.7480916030534353e-05,
+      "loss": 1.1116,
+      "step": 458
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4845714569091797,
+      "learning_rate": 1.751908396946565e-05,
+      "loss": 1.0849,
+      "step": 459
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3281002044677734,
+      "learning_rate": 1.755725190839695e-05,
+      "loss": 1.1561,
+      "step": 460
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0520999431610107,
+      "learning_rate": 1.7595419847328246e-05,
+      "loss": 1.0587,
+      "step": 461
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3974833488464355,
+      "learning_rate": 1.7633587786259544e-05,
+      "loss": 1.0751,
+      "step": 462
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2551488876342773,
+      "learning_rate": 1.767175572519084e-05,
+      "loss": 1.1902,
+      "step": 463
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5492265224456787,
+      "learning_rate": 1.770992366412214e-05,
+      "loss": 1.1645,
+      "step": 464
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.333768129348755,
+      "learning_rate": 1.7748091603053437e-05,
+      "loss": 1.0968,
+      "step": 465
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9941705465316772,
+      "learning_rate": 1.7786259541984735e-05,
+      "loss": 0.4776,
+      "step": 466
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0160634517669678,
+      "learning_rate": 1.7824427480916033e-05,
+      "loss": 1.1978,
+      "step": 467
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2153005599975586,
+      "learning_rate": 1.786259541984733e-05,
+      "loss": 1.1401,
+      "step": 468
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.150308609008789,
+      "learning_rate": 1.790076335877863e-05,
+      "loss": 1.1996,
+      "step": 469
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0668734312057495,
+      "learning_rate": 1.7938931297709926e-05,
+      "loss": 0.5641,
+      "step": 470
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3119425773620605,
+      "learning_rate": 1.7977099236641224e-05,
+      "loss": 1.2263,
+      "step": 471
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3993160724639893,
+      "learning_rate": 1.8015267175572518e-05,
+      "loss": 1.1162,
+      "step": 472
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.261383056640625,
+      "learning_rate": 1.805343511450382e-05,
+      "loss": 1.1778,
+      "step": 473
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.249664783477783,
+      "learning_rate": 1.8091603053435117e-05,
+      "loss": 1.1903,
+      "step": 474
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4010398387908936,
+      "learning_rate": 1.8129770992366415e-05,
+      "loss": 1.0956,
+      "step": 475
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1136553287506104,
+      "learning_rate": 1.816793893129771e-05,
+      "loss": 1.1892,
+      "step": 476
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4139397144317627,
+      "learning_rate": 1.820610687022901e-05,
+      "loss": 1.1943,
+      "step": 477
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.240676164627075,
+      "learning_rate": 1.8244274809160308e-05,
+      "loss": 1.0821,
+      "step": 478
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.408395528793335,
+      "learning_rate": 1.8282442748091606e-05,
+      "loss": 1.1172,
+      "step": 479
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.193232297897339,
+      "learning_rate": 1.83206106870229e-05,
+      "loss": 1.157,
+      "step": 480
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1655828952789307,
+      "learning_rate": 1.8358778625954198e-05,
+      "loss": 1.2059,
+      "step": 481
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1595618724823,
+      "learning_rate": 1.83969465648855e-05,
+      "loss": 1.0686,
+      "step": 482
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5173721313476562,
+      "learning_rate": 1.8435114503816797e-05,
+      "loss": 1.2197,
+      "step": 483
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.280892848968506,
+      "learning_rate": 1.847328244274809e-05,
+      "loss": 1.1005,
+      "step": 484
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3220772743225098,
+      "learning_rate": 1.851145038167939e-05,
+      "loss": 1.1063,
+      "step": 485
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.079601287841797,
+      "learning_rate": 1.854961832061069e-05,
+      "loss": 1.1777,
+      "step": 486
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2345283031463623,
+      "learning_rate": 1.8587786259541988e-05,
+      "loss": 1.1491,
+      "step": 487
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3589625358581543,
+      "learning_rate": 1.8625954198473282e-05,
+      "loss": 1.1647,
+      "step": 488
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4348666667938232,
+      "learning_rate": 1.866412213740458e-05,
+      "loss": 1.1538,
+      "step": 489
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2719566822052,
+      "learning_rate": 1.8702290076335878e-05,
+      "loss": 1.1317,
+      "step": 490
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.241027355194092,
+      "learning_rate": 1.874045801526718e-05,
+      "loss": 1.1695,
+      "step": 491
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3261735439300537,
+      "learning_rate": 1.8778625954198473e-05,
+      "loss": 1.1719,
+      "step": 492
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1342687606811523,
+      "learning_rate": 1.881679389312977e-05,
+      "loss": 1.1512,
+      "step": 493
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1521098613739014,
+      "learning_rate": 1.885496183206107e-05,
+      "loss": 1.2239,
+      "step": 494
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1821930408477783,
+      "learning_rate": 1.889312977099237e-05,
+      "loss": 1.0751,
+      "step": 495
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2680542469024658,
+      "learning_rate": 1.8931297709923668e-05,
+      "loss": 0.5478,
+      "step": 496
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6782658100128174,
+      "learning_rate": 1.8969465648854962e-05,
+      "loss": 1.1517,
+      "step": 497
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.575467109680176,
+      "learning_rate": 1.900763358778626e-05,
+      "loss": 1.1372,
+      "step": 498
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9334840774536133,
+      "learning_rate": 1.9045801526717558e-05,
+      "loss": 1.0998,
+      "step": 499
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.164106845855713,
+      "learning_rate": 1.908396946564886e-05,
+      "loss": 0.5547,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.201937675476074,
+      "learning_rate": 1.9122137404580153e-05,
+      "loss": 1.1048,
+      "step": 501
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4689877033233643,
+      "learning_rate": 1.916030534351145e-05,
+      "loss": 1.1524,
+      "step": 502
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3615005016326904,
+      "learning_rate": 1.919847328244275e-05,
+      "loss": 1.1679,
+      "step": 503
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.327754020690918,
+      "learning_rate": 1.923664122137405e-05,
+      "loss": 1.1288,
+      "step": 504
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1735355854034424,
+      "learning_rate": 1.9274809160305344e-05,
+      "loss": 1.212,
+      "step": 505
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6156322956085205,
+      "learning_rate": 1.9312977099236642e-05,
+      "loss": 1.2327,
+      "step": 506
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.54784893989563,
+      "learning_rate": 1.935114503816794e-05,
+      "loss": 1.1699,
+      "step": 507
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.143899440765381,
+      "learning_rate": 1.9389312977099238e-05,
+      "loss": 1.1415,
+      "step": 508
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2697091102600098,
+      "learning_rate": 1.9427480916030535e-05,
+      "loss": 1.2161,
+      "step": 509
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1283698081970215,
+      "learning_rate": 1.9465648854961833e-05,
+      "loss": 1.1471,
+      "step": 510
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2076292037963867,
+      "learning_rate": 1.950381679389313e-05,
+      "loss": 1.1259,
+      "step": 511
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4508249759674072,
+      "learning_rate": 1.954198473282443e-05,
+      "loss": 1.1896,
+      "step": 512
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.149657964706421,
+      "learning_rate": 1.9580152671755726e-05,
+      "loss": 1.112,
+      "step": 513
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2458279132843018,
+      "learning_rate": 1.9618320610687024e-05,
+      "loss": 1.1661,
+      "step": 514
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.184749126434326,
+      "learning_rate": 1.9656488549618322e-05,
+      "loss": 1.1254,
+      "step": 515
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3220924139022827,
+      "learning_rate": 1.969465648854962e-05,
+      "loss": 0.5261,
+      "step": 516
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3829903602600098,
+      "learning_rate": 1.9732824427480917e-05,
+      "loss": 1.2131,
+      "step": 517
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.315910577774048,
+      "learning_rate": 1.9770992366412215e-05,
+      "loss": 1.2225,
+      "step": 518
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.187544107437134,
+      "learning_rate": 1.9809160305343513e-05,
+      "loss": 1.1266,
+      "step": 519
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.349717617034912,
+      "learning_rate": 1.984732824427481e-05,
+      "loss": 1.1396,
+      "step": 520
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3578975200653076,
+      "learning_rate": 1.988549618320611e-05,
+      "loss": 1.1718,
+      "step": 521
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4661965370178223,
+      "learning_rate": 1.9923664122137406e-05,
+      "loss": 1.1849,
+      "step": 522
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.371136426925659,
+      "learning_rate": 1.9961832061068704e-05,
+      "loss": 1.1481,
+      "step": 523
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2340142726898193,
+      "learning_rate": 2e-05,
+      "loss": 1.2411,
+      "step": 524
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2085105180740356,
+      "learning_rate": 1.9999999827463968e-05,
+      "loss": 0.5797,
+      "step": 525
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4866862297058105,
+      "learning_rate": 1.9999999309855876e-05,
+      "loss": 1.1384,
+      "step": 526
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.42767596244812,
+      "learning_rate": 1.999999844717574e-05,
+      "loss": 1.165,
+      "step": 527
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4111995697021484,
+      "learning_rate": 1.9999997239423593e-05,
+      "loss": 1.0462,
+      "step": 528
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.409727096557617,
+      "learning_rate": 1.999999568659947e-05,
+      "loss": 1.1697,
+      "step": 529
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0967955589294434,
+      "learning_rate": 1.9999993788703435e-05,
+      "loss": 1.1773,
+      "step": 530
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5718483924865723,
+      "learning_rate": 1.999999154573555e-05,
+      "loss": 1.2503,
+      "step": 531
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3113183975219727,
+      "learning_rate": 1.9999988957695886e-05,
+      "loss": 1.1549,
+      "step": 532
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3366270065307617,
+      "learning_rate": 1.999998602458454e-05,
+      "loss": 1.05,
+      "step": 533
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3591933250427246,
+      "learning_rate": 1.9999982746401607e-05,
+      "loss": 1.1218,
+      "step": 534
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3871476650238037,
+      "learning_rate": 1.9999979123147204e-05,
+      "loss": 1.1413,
+      "step": 535
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.11716890335083,
+      "learning_rate": 1.9999975154821454e-05,
+      "loss": 1.2012,
+      "step": 536
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.312731981277466,
+      "learning_rate": 1.99999708414245e-05,
+      "loss": 1.2154,
+      "step": 537
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3737916946411133,
+      "learning_rate": 1.9999966182956486e-05,
+      "loss": 1.0983,
+      "step": 538
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3799524307250977,
+      "learning_rate": 1.999996117941757e-05,
+      "loss": 1.2057,
+      "step": 539
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4459667205810547,
+      "learning_rate": 1.9999955830807925e-05,
+      "loss": 1.1362,
+      "step": 540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.422807216644287,
+      "learning_rate": 1.999995013712774e-05,
+      "loss": 1.0987,
+      "step": 541
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3833820819854736,
+      "learning_rate": 1.9999944098377214e-05,
+      "loss": 1.154,
+      "step": 542
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2140049934387207,
+      "learning_rate": 1.9999937714556546e-05,
+      "loss": 1.1917,
+      "step": 543
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7330381870269775,
+      "learning_rate": 1.999993098566596e-05,
+      "loss": 1.1322,
+      "step": 544
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.265702247619629,
+      "learning_rate": 1.9999923911705693e-05,
+      "loss": 1.2275,
+      "step": 545
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.586068868637085,
+      "learning_rate": 1.9999916492675984e-05,
+      "loss": 1.1429,
+      "step": 546
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4351508617401123,
+      "learning_rate": 1.999990872857709e-05,
+      "loss": 1.2053,
+      "step": 547
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2522010803222656,
+      "learning_rate": 1.999990061940928e-05,
+      "loss": 1.1497,
+      "step": 548
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0282084941864014,
+      "learning_rate": 1.999989216517283e-05,
+      "loss": 1.134,
+      "step": 549
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3976423740386963,
+      "learning_rate": 1.999988336586804e-05,
+      "loss": 1.1103,
+      "step": 550
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2513225078582764,
+      "learning_rate": 1.999987422149521e-05,
+      "loss": 1.16,
+      "step": 551
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.314650297164917,
+      "learning_rate": 1.999986473205465e-05,
+      "loss": 1.1013,
+      "step": 552
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2503671646118164,
+      "learning_rate": 1.999985489754669e-05,
+      "loss": 1.1604,
+      "step": 553
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1580429077148438,
+      "learning_rate": 1.9999844717971674e-05,
+      "loss": 1.1881,
+      "step": 554
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1537997722625732,
+      "learning_rate": 1.9999834193329952e-05,
+      "loss": 1.1533,
+      "step": 555
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0059038400650024,
+      "learning_rate": 1.999982332362188e-05,
+      "loss": 0.4717,
+      "step": 556
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3200104236602783,
+      "learning_rate": 1.9999812108847844e-05,
+      "loss": 1.0886,
+      "step": 557
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1596384048461914,
+      "learning_rate": 1.999980054900822e-05,
+      "loss": 1.2114,
+      "step": 558
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.192187786102295,
+      "learning_rate": 1.9999788644103418e-05,
+      "loss": 1.0871,
+      "step": 559
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4987313747406006,
+      "learning_rate": 1.999977639413384e-05,
+      "loss": 1.1774,
+      "step": 560
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.245305061340332,
+      "learning_rate": 1.9999763799099912e-05,
+      "loss": 1.1757,
+      "step": 561
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0853499174118042,
+      "learning_rate": 1.9999750859002066e-05,
+      "loss": 0.5508,
+      "step": 562
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2679243087768555,
+      "learning_rate": 1.9999737573840755e-05,
+      "loss": 1.1046,
+      "step": 563
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.236968517303467,
+      "learning_rate": 1.9999723943616435e-05,
+      "loss": 1.1015,
+      "step": 564
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2850563526153564,
+      "learning_rate": 1.9999709968329572e-05,
+      "loss": 1.197,
+      "step": 565
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.270822525024414,
+      "learning_rate": 1.999969564798065e-05,
+      "loss": 1.1892,
+      "step": 566
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2477834224700928,
+      "learning_rate": 1.9999680982570165e-05,
+      "loss": 1.1031,
+      "step": 567
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.257143497467041,
+      "learning_rate": 1.9999665972098624e-05,
+      "loss": 1.1298,
+      "step": 568
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.267540216445923,
+      "learning_rate": 1.9999650616566542e-05,
+      "loss": 1.2086,
+      "step": 569
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1472766399383545,
+      "learning_rate": 1.999963491597445e-05,
+      "loss": 1.1297,
+      "step": 570
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4439661502838135,
+      "learning_rate": 1.999961887032289e-05,
+      "loss": 1.1425,
+      "step": 571
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.184567928314209,
+      "learning_rate": 1.9999602479612416e-05,
+      "loss": 1.1636,
+      "step": 572
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2469265460968018,
+      "learning_rate": 1.9999585743843592e-05,
+      "loss": 1.1684,
+      "step": 573
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3082656860351562,
+      "learning_rate": 1.9999568663016998e-05,
+      "loss": 1.1929,
+      "step": 574
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.273785352706909,
+      "learning_rate": 1.999955123713322e-05,
+      "loss": 1.1653,
+      "step": 575
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2690200805664062,
+      "learning_rate": 1.9999533466192864e-05,
+      "loss": 1.0866,
+      "step": 576
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0065724849700928,
+      "learning_rate": 1.9999515350196538e-05,
+      "loss": 1.1763,
+      "step": 577
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.245633363723755,
+      "learning_rate": 1.9999496889144874e-05,
+      "loss": 1.1285,
+      "step": 578
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2888190746307373,
+      "learning_rate": 1.9999478083038503e-05,
+      "loss": 1.2325,
+      "step": 579
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.347853183746338,
+      "learning_rate": 1.999945893187807e-05,
+      "loss": 1.2205,
+      "step": 580
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.051862955093384,
+      "learning_rate": 1.999943943566425e-05,
+      "loss": 1.1805,
+      "step": 581
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1901590824127197,
+      "learning_rate": 1.9999419594397706e-05,
+      "loss": 1.1634,
+      "step": 582
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.16910982131958,
+      "learning_rate": 1.999939940807912e-05,
+      "loss": 1.1977,
+      "step": 583
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0214115381240845,
+      "learning_rate": 1.9999378876709194e-05,
+      "loss": 0.5249,
+      "step": 584
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3862781524658203,
+      "learning_rate": 1.9999358000288637e-05,
+      "loss": 1.1643,
+      "step": 585
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.666750431060791,
+      "learning_rate": 1.9999336778818167e-05,
+      "loss": 1.1773,
+      "step": 586
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.133347988128662,
+      "learning_rate": 1.9999315212298516e-05,
+      "loss": 1.1353,
+      "step": 587
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2714200019836426,
+      "learning_rate": 1.9999293300730426e-05,
+      "loss": 1.179,
+      "step": 588
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1183888912200928,
+      "learning_rate": 1.9999271044114663e-05,
+      "loss": 1.112,
+      "step": 589
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2575178146362305,
+      "learning_rate": 1.9999248442451984e-05,
+      "loss": 1.06,
+      "step": 590
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2987260818481445,
+      "learning_rate": 1.999922549574317e-05,
+      "loss": 1.1754,
+      "step": 591
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.289649724960327,
+      "learning_rate": 1.9999202203989022e-05,
+      "loss": 1.1596,
+      "step": 592
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4353740215301514,
+      "learning_rate": 1.9999178567190334e-05,
+      "loss": 1.1812,
+      "step": 593
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1835286617279053,
+      "learning_rate": 1.9999154585347926e-05,
+      "loss": 1.1053,
+      "step": 594
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4543471336364746,
+      "learning_rate": 1.9999130258462626e-05,
+      "loss": 1.1579,
+      "step": 595
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.156053066253662,
+      "learning_rate": 1.999910558653527e-05,
+      "loss": 1.1879,
+      "step": 596
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1831308603286743,
+      "learning_rate": 1.999908056956671e-05,
+      "loss": 0.5483,
+      "step": 597
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.505997657775879,
+      "learning_rate": 1.9999055207557814e-05,
+      "loss": 1.1579,
+      "step": 598
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2746193408966064,
+      "learning_rate": 1.9999029500509453e-05,
+      "loss": 1.2156,
+      "step": 599
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5150904655456543,
+      "learning_rate": 1.9999003448422516e-05,
+      "loss": 1.1332,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3727567195892334,
+      "learning_rate": 1.99989770512979e-05,
+      "loss": 1.1354,
+      "step": 601
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.439591646194458,
+      "learning_rate": 1.999895030913652e-05,
+      "loss": 1.1366,
+      "step": 602
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.472810745239258,
+      "learning_rate": 1.9998923221939294e-05,
+      "loss": 1.121,
+      "step": 603
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1895086765289307,
+      "learning_rate": 1.9998895789707156e-05,
+      "loss": 1.0808,
+      "step": 604
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1236534118652344,
+      "learning_rate": 1.9998868012441056e-05,
+      "loss": 1.1354,
+      "step": 605
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1101601123809814,
+      "learning_rate": 1.9998839890141953e-05,
+      "loss": 1.1826,
+      "step": 606
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3621020317077637,
+      "learning_rate": 1.9998811422810816e-05,
+      "loss": 1.235,
+      "step": 607
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.186937093734741,
+      "learning_rate": 1.9998782610448625e-05,
+      "loss": 1.1058,
+      "step": 608
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4096362590789795,
+      "learning_rate": 1.999875345305638e-05,
+      "loss": 1.1583,
+      "step": 609
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.364793539047241,
+      "learning_rate": 1.999872395063508e-05,
+      "loss": 1.1775,
+      "step": 610
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.067514181137085,
+      "learning_rate": 1.9998694103185753e-05,
+      "loss": 1.0976,
+      "step": 611
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5292017459869385,
+      "learning_rate": 1.9998663910709416e-05,
+      "loss": 1.1454,
+      "step": 612
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.290858268737793,
+      "learning_rate": 1.999863337320712e-05,
+      "loss": 1.1362,
+      "step": 613
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3363685607910156,
+      "learning_rate": 1.9998602490679916e-05,
+      "loss": 1.1575,
+      "step": 614
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.250690460205078,
+      "learning_rate": 1.9998571263128873e-05,
+      "loss": 1.189,
+      "step": 615
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.323308229446411,
+      "learning_rate": 1.999853969055506e-05,
+      "loss": 1.0895,
+      "step": 616
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.347003936767578,
+      "learning_rate": 1.9998507772959578e-05,
+      "loss": 1.1435,
+      "step": 617
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0986454486846924,
+      "learning_rate": 1.999847551034352e-05,
+      "loss": 1.1879,
+      "step": 618
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2946653366088867,
+      "learning_rate": 1.9998442902708e-05,
+      "loss": 1.1618,
+      "step": 619
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4341859817504883,
+      "learning_rate": 1.999840995005415e-05,
+      "loss": 1.1794,
+      "step": 620
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2120912075042725,
+      "learning_rate": 1.9998376652383095e-05,
+      "loss": 1.076,
+      "step": 621
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1352121829986572,
+      "learning_rate": 1.9998343009695995e-05,
+      "loss": 1.0731,
+      "step": 622
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2443056106567383,
+      "learning_rate": 1.9998309021994006e-05,
+      "loss": 1.2216,
+      "step": 623
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.225339412689209,
+      "learning_rate": 1.9998274689278302e-05,
+      "loss": 1.0892,
+      "step": 624
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3853800296783447,
+      "learning_rate": 1.999824001155007e-05,
+      "loss": 1.1826,
+      "step": 625
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2819066047668457,
+      "learning_rate": 1.99982049888105e-05,
+      "loss": 1.1986,
+      "step": 626
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5107710361480713,
+      "learning_rate": 1.999816962106081e-05,
+      "loss": 1.1054,
+      "step": 627
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2733571529388428,
+      "learning_rate": 1.999813390830221e-05,
+      "loss": 1.1267,
+      "step": 628
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1569218635559082,
+      "learning_rate": 1.999809785053594e-05,
+      "loss": 0.5558,
+      "step": 629
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.067953586578369,
+      "learning_rate": 1.999806144776324e-05,
+      "loss": 1.1731,
+      "step": 630
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3705060482025146,
+      "learning_rate": 1.999802469998537e-05,
+      "loss": 1.2497,
+      "step": 631
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.12733793258667,
+      "learning_rate": 1.9997987607203596e-05,
+      "loss": 1.1613,
+      "step": 632
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.433986186981201,
+      "learning_rate": 1.9997950169419194e-05,
+      "loss": 1.1397,
+      "step": 633
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2078075408935547,
+      "learning_rate": 1.9997912386633464e-05,
+      "loss": 1.12,
+      "step": 634
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.938741683959961,
+      "learning_rate": 1.99978742588477e-05,
+      "loss": 1.0168,
+      "step": 635
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2172467708587646,
+      "learning_rate": 1.999783578606323e-05,
+      "loss": 1.1134,
+      "step": 636
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2737231254577637,
+      "learning_rate": 1.9997796968281373e-05,
+      "loss": 1.1391,
+      "step": 637
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.331183671951294,
+      "learning_rate": 1.999775780550347e-05,
+      "loss": 1.1118,
+      "step": 638
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3759515285491943,
+      "learning_rate": 1.9997718297730874e-05,
+      "loss": 1.182,
+      "step": 639
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.184342622756958,
+      "learning_rate": 1.9997678444964947e-05,
+      "loss": 1.0702,
+      "step": 640
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2764105796813965,
+      "learning_rate": 1.9997638247207057e-05,
+      "loss": 1.1988,
+      "step": 641
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9475120306015015,
+      "learning_rate": 1.9997597704458608e-05,
+      "loss": 1.0982,
+      "step": 642
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0692942142486572,
+      "learning_rate": 1.9997556816720985e-05,
+      "loss": 1.148,
+      "step": 643
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1277546882629395,
+      "learning_rate": 1.9997515583995604e-05,
+      "loss": 1.1454,
+      "step": 644
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.091170072555542,
+      "learning_rate": 1.9997474006283885e-05,
+      "loss": 1.1404,
+      "step": 645
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2826859951019287,
+      "learning_rate": 1.9997432083587268e-05,
+      "loss": 1.1263,
+      "step": 646
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2109992504119873,
+      "learning_rate": 1.9997389815907193e-05,
+      "loss": 1.2241,
+      "step": 647
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.08583927154541,
+      "learning_rate": 1.9997347203245126e-05,
+      "loss": 1.1907,
+      "step": 648
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0009896755218506,
+      "learning_rate": 1.9997304245602533e-05,
+      "loss": 1.1202,
+      "step": 649
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2349495887756348,
+      "learning_rate": 1.9997260942980895e-05,
+      "loss": 1.0894,
+      "step": 650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2654592990875244,
+      "learning_rate": 1.999721729538171e-05,
+      "loss": 1.0681,
+      "step": 651
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2119839191436768,
+      "learning_rate": 1.9997173302806478e-05,
+      "loss": 1.1419,
+      "step": 652
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.471290349960327,
+      "learning_rate": 1.9997128965256726e-05,
+      "loss": 1.1598,
+      "step": 653
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.283198118209839,
+      "learning_rate": 1.9997084282733975e-05,
+      "loss": 1.1286,
+      "step": 654
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6239185333251953,
+      "learning_rate": 1.9997039255239774e-05,
+      "loss": 1.2641,
+      "step": 655
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.125765562057495,
+      "learning_rate": 1.9996993882775674e-05,
+      "loss": 1.1104,
+      "step": 656
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5450098514556885,
+      "learning_rate": 1.9996948165343243e-05,
+      "loss": 1.1035,
+      "step": 657
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2383806705474854,
+      "learning_rate": 1.999690210294405e-05,
+      "loss": 1.1286,
+      "step": 658
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2511579990386963,
+      "learning_rate": 1.9996855695579694e-05,
+      "loss": 1.1192,
+      "step": 659
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0633535385131836,
+      "learning_rate": 1.9996808943251773e-05,
+      "loss": 1.1526,
+      "step": 660
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3751425743103027,
+      "learning_rate": 1.99967618459619e-05,
+      "loss": 1.195,
+      "step": 661
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.061765193939209,
+      "learning_rate": 1.99967144037117e-05,
+      "loss": 1.1203,
+      "step": 662
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1556942462921143,
+      "learning_rate": 1.9996666616502812e-05,
+      "loss": 1.1355,
+      "step": 663
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3233346939086914,
+      "learning_rate": 1.9996618484336885e-05,
+      "loss": 1.1094,
+      "step": 664
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.210195779800415,
+      "learning_rate": 1.9996570007215578e-05,
+      "loss": 1.1181,
+      "step": 665
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1433939933776855,
+      "learning_rate": 1.999652118514056e-05,
+      "loss": 1.1799,
+      "step": 666
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1322736740112305,
+      "learning_rate": 1.9996472018113523e-05,
+      "loss": 1.1177,
+      "step": 667
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.336160659790039,
+      "learning_rate": 1.999642250613616e-05,
+      "loss": 1.1672,
+      "step": 668
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.432091474533081,
+      "learning_rate": 1.9996372649210182e-05,
+      "loss": 1.1951,
+      "step": 669
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.310286521911621,
+      "learning_rate": 1.9996322447337307e-05,
+      "loss": 1.2545,
+      "step": 670
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2603771686553955,
+      "learning_rate": 1.9996271900519267e-05,
+      "loss": 1.2332,
+      "step": 671
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2341551780700684,
+      "learning_rate": 1.9996221008757807e-05,
+      "loss": 1.1276,
+      "step": 672
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2875022888183594,
+      "learning_rate": 1.9996169772054684e-05,
+      "loss": 1.1259,
+      "step": 673
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.266982078552246,
+      "learning_rate": 1.9996118190411664e-05,
+      "loss": 1.1116,
+      "step": 674
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.226486921310425,
+      "learning_rate": 1.9996066263830533e-05,
+      "loss": 1.1122,
+      "step": 675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.379892110824585,
+      "learning_rate": 1.9996013992313072e-05,
+      "loss": 1.1502,
+      "step": 676
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.562361478805542,
+      "learning_rate": 1.9995961375861092e-05,
+      "loss": 1.1756,
+      "step": 677
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3117401599884033,
+      "learning_rate": 1.999590841447641e-05,
+      "loss": 1.1153,
+      "step": 678
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3392574787139893,
+      "learning_rate": 1.9995855108160852e-05,
+      "loss": 1.185,
+      "step": 679
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1996610164642334,
+      "learning_rate": 1.9995801456916252e-05,
+      "loss": 1.1068,
+      "step": 680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.414916753768921,
+      "learning_rate": 1.9995747460744467e-05,
+      "loss": 1.223,
+      "step": 681
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2416093349456787,
+      "learning_rate": 1.999569311964736e-05,
+      "loss": 1.1219,
+      "step": 682
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.197721004486084,
+      "learning_rate": 1.999563843362681e-05,
+      "loss": 1.2113,
+      "step": 683
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0379817485809326,
+      "learning_rate": 1.9995583402684697e-05,
+      "loss": 1.137,
+      "step": 684
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0361454486846924,
+      "learning_rate": 1.9995528026822916e-05,
+      "loss": 1.0917,
+      "step": 685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.349773406982422,
+      "learning_rate": 1.999547230604339e-05,
+      "loss": 1.1195,
+      "step": 686
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.339024782180786,
+      "learning_rate": 1.9995416240348034e-05,
+      "loss": 1.1838,
+      "step": 687
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.271723985671997,
+      "learning_rate": 1.9995359829738784e-05,
+      "loss": 1.1244,
+      "step": 688
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.658097982406616,
+      "learning_rate": 1.999530307421759e-05,
+      "loss": 1.1088,
+      "step": 689
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.427680253982544,
+      "learning_rate": 1.9995245973786404e-05,
+      "loss": 1.1421,
+      "step": 690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1488356590270996,
+      "learning_rate": 1.9995188528447205e-05,
+      "loss": 1.1048,
+      "step": 691
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1456120014190674,
+      "learning_rate": 1.9995130738201966e-05,
+      "loss": 1.1714,
+      "step": 692
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.168026924133301,
+      "learning_rate": 1.9995072603052687e-05,
+      "loss": 1.1326,
+      "step": 693
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3212029933929443,
+      "learning_rate": 1.9995014123001374e-05,
+      "loss": 1.1415,
+      "step": 694
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.588196039199829,
+      "learning_rate": 1.999495529805004e-05,
+      "loss": 1.1638,
+      "step": 695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.14258074760437,
+      "learning_rate": 1.9994896128200723e-05,
+      "loss": 1.1866,
+      "step": 696
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.253497838973999,
+      "learning_rate": 1.9994836613455456e-05,
+      "loss": 1.1213,
+      "step": 697
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9785290956497192,
+      "learning_rate": 1.99947767538163e-05,
+      "loss": 1.1912,
+      "step": 698
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.480708122253418,
+      "learning_rate": 1.9994716549285312e-05,
+      "loss": 1.1657,
+      "step": 699
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3822779655456543,
+      "learning_rate": 1.9994655999864583e-05,
+      "loss": 1.0756,
+      "step": 700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1088130474090576,
+      "learning_rate": 1.999459510555619e-05,
+      "loss": 1.1789,
+      "step": 701
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.321089744567871,
+      "learning_rate": 1.999453386636224e-05,
+      "loss": 1.1927,
+      "step": 702
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4990336894989014,
+      "learning_rate": 1.9994472282284843e-05,
+      "loss": 1.1084,
+      "step": 703
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5052928924560547,
+      "learning_rate": 1.9994410353326126e-05,
+      "loss": 1.0988,
+      "step": 704
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1726348400115967,
+      "learning_rate": 1.9994348079488225e-05,
+      "loss": 1.1309,
+      "step": 705
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.193157911300659,
+      "learning_rate": 1.9994285460773294e-05,
+      "loss": 1.1344,
+      "step": 706
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2607314586639404,
+      "learning_rate": 1.9994222497183487e-05,
+      "loss": 1.1726,
+      "step": 707
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1786909103393555,
+      "learning_rate": 1.999415918872098e-05,
+      "loss": 1.1471,
+      "step": 708
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3930792808532715,
+      "learning_rate": 1.999409553538796e-05,
+      "loss": 1.0704,
+      "step": 709
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2470383644104004,
+      "learning_rate": 1.9994031537186615e-05,
+      "loss": 1.1074,
+      "step": 710
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1238656044006348,
+      "learning_rate": 1.999396719411916e-05,
+      "loss": 1.0823,
+      "step": 711
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.032536029815674,
+      "learning_rate": 1.9993902506187815e-05,
+      "loss": 1.0965,
+      "step": 712
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1003777980804443,
+      "learning_rate": 1.999383747339481e-05,
+      "loss": 1.1183,
+      "step": 713
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.355907678604126,
+      "learning_rate": 1.9993772095742396e-05,
+      "loss": 1.1553,
+      "step": 714
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1636922359466553,
+      "learning_rate": 1.9993706373232818e-05,
+      "loss": 1.1997,
+      "step": 715
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1999868154525757,
+      "learning_rate": 1.999364030586835e-05,
+      "loss": 0.5811,
+      "step": 716
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.041339874267578,
+      "learning_rate": 1.9993573893651273e-05,
+      "loss": 1.1038,
+      "step": 717
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0734360218048096,
+      "learning_rate": 1.9993507136583876e-05,
+      "loss": 1.1119,
+      "step": 718
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2639509439468384,
+      "learning_rate": 1.9993440034668462e-05,
+      "loss": 0.6233,
+      "step": 719
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.218365430831909,
+      "learning_rate": 1.9993372587907348e-05,
+      "loss": 1.1584,
+      "step": 720
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0786736011505127,
+      "learning_rate": 1.9993304796302865e-05,
+      "loss": 0.5584,
+      "step": 721
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1460120677948,
+      "learning_rate": 1.9993236659857347e-05,
+      "loss": 1.153,
+      "step": 722
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3159492015838623,
+      "learning_rate": 1.9993168178573146e-05,
+      "loss": 1.1244,
+      "step": 723
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.178205966949463,
+      "learning_rate": 1.9993099352452626e-05,
+      "loss": 1.0997,
+      "step": 724
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.245976686477661,
+      "learning_rate": 1.9993030181498163e-05,
+      "loss": 1.1348,
+      "step": 725
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0968141555786133,
+      "learning_rate": 1.999296066571214e-05,
+      "loss": 1.2025,
+      "step": 726
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3151137828826904,
+      "learning_rate": 1.999289080509696e-05,
+      "loss": 0.5537,
+      "step": 727
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8331804275512695,
+      "learning_rate": 1.9992820599655034e-05,
+      "loss": 1.0328,
+      "step": 728
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.202378511428833,
+      "learning_rate": 1.9992750049388783e-05,
+      "loss": 1.1926,
+      "step": 729
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.318918228149414,
+      "learning_rate": 1.999267915430064e-05,
+      "loss": 1.1047,
+      "step": 730
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3065872192382812,
+      "learning_rate": 1.999260791439305e-05,
+      "loss": 1.1673,
+      "step": 731
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.137610673904419,
+      "learning_rate": 1.999253632966848e-05,
+      "loss": 1.1542,
+      "step": 732
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.495720863342285,
+      "learning_rate": 1.999246440012939e-05,
+      "loss": 1.1439,
+      "step": 733
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5278029441833496,
+      "learning_rate": 1.9992392125778267e-05,
+      "loss": 1.2329,
+      "step": 734
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2882354259490967,
+      "learning_rate": 1.9992319506617606e-05,
+      "loss": 1.0965,
+      "step": 735
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1578209400177,
+      "learning_rate": 1.999224654264991e-05,
+      "loss": 1.1287,
+      "step": 736
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.230405569076538,
+      "learning_rate": 1.99921732338777e-05,
+      "loss": 1.0937,
+      "step": 737
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.175158739089966,
+      "learning_rate": 1.99920995803035e-05,
+      "loss": 1.179,
+      "step": 738
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.416019916534424,
+      "learning_rate": 1.9992025581929856e-05,
+      "loss": 1.1751,
+      "step": 739
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1297595500946045,
+      "learning_rate": 1.9991951238759323e-05,
+      "loss": 1.0965,
+      "step": 740
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.335035562515259,
+      "learning_rate": 1.9991876550794465e-05,
+      "loss": 1.0905,
+      "step": 741
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9685287475585938,
+      "learning_rate": 1.9991801518037856e-05,
+      "loss": 1.1004,
+      "step": 742
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.143871784210205,
+      "learning_rate": 1.9991726140492088e-05,
+      "loss": 1.094,
+      "step": 743
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1374764442443848,
+      "learning_rate": 1.9991650418159763e-05,
+      "loss": 1.1805,
+      "step": 744
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.418219566345215,
+      "learning_rate": 1.999157435104349e-05,
+      "loss": 1.1657,
+      "step": 745
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1903107166290283,
+      "learning_rate": 1.9991497939145898e-05,
+      "loss": 1.1089,
+      "step": 746
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.141371250152588,
+      "learning_rate": 1.9991421182469624e-05,
+      "loss": 1.1817,
+      "step": 747
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0492637157440186,
+      "learning_rate": 1.9991344081017312e-05,
+      "loss": 1.0642,
+      "step": 748
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3152265548706055,
+      "learning_rate": 1.9991266634791627e-05,
+      "loss": 1.1055,
+      "step": 749
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.195970296859741,
+      "learning_rate": 1.9991188843795238e-05,
+      "loss": 1.187,
+      "step": 750
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2096188068389893,
+      "learning_rate": 1.9991110708030836e-05,
+      "loss": 1.1665,
+      "step": 751
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1097246408462524,
+      "learning_rate": 1.999103222750111e-05,
+      "loss": 0.5414,
+      "step": 752
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.29532527923584,
+      "learning_rate": 1.9990953402208767e-05,
+      "loss": 1.1721,
+      "step": 753
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.391451597213745,
+      "learning_rate": 1.9990874232156533e-05,
+      "loss": 1.2523,
+      "step": 754
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2653884887695312,
+      "learning_rate": 1.999079471734714e-05,
+      "loss": 1.1512,
+      "step": 755
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1708734035491943,
+      "learning_rate": 1.9990714857783327e-05,
+      "loss": 1.1133,
+      "step": 756
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0662405490875244,
+      "learning_rate": 1.9990634653467854e-05,
+      "loss": 0.61,
+      "step": 757
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.258378744125366,
+      "learning_rate": 1.9990554104403484e-05,
+      "loss": 1.1304,
+      "step": 758
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3086485862731934,
+      "learning_rate": 1.9990473210593e-05,
+      "loss": 1.1219,
+      "step": 759
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2723228931427,
+      "learning_rate": 1.9990391972039197e-05,
+      "loss": 1.1812,
+      "step": 760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.930712342262268,
+      "learning_rate": 1.9990310388744868e-05,
+      "loss": 1.054,
+      "step": 761
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0871713161468506,
+      "learning_rate": 1.999022846071284e-05,
+      "loss": 1.0973,
+      "step": 762
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.204150915145874,
+      "learning_rate": 1.9990146187945928e-05,
+      "loss": 1.1459,
+      "step": 763
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.039926767349243,
+      "learning_rate": 1.9990063570446985e-05,
+      "loss": 1.0296,
+      "step": 764
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.004918098449707,
+      "learning_rate": 1.9989980608218847e-05,
+      "loss": 1.1188,
+      "step": 765
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.079618215560913,
+      "learning_rate": 1.998989730126439e-05,
+      "loss": 1.1185,
+      "step": 766
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.041402578353882,
+      "learning_rate": 1.998981364958648e-05,
+      "loss": 1.0871,
+      "step": 767
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1574597358703613,
+      "learning_rate": 1.9989729653188006e-05,
+      "loss": 1.1514,
+      "step": 768
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.337864637374878,
+      "learning_rate": 1.9989645312071867e-05,
+      "loss": 1.1673,
+      "step": 769
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9658972024917603,
+      "learning_rate": 1.9989560626240975e-05,
+      "loss": 1.0791,
+      "step": 770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1030495166778564,
+      "learning_rate": 1.9989475595698245e-05,
+      "loss": 0.5206,
+      "step": 771
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.949268102645874,
+      "learning_rate": 1.9989390220446624e-05,
+      "loss": 1.1177,
+      "step": 772
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1365692615509033,
+      "learning_rate": 1.9989304500489047e-05,
+      "loss": 1.1042,
+      "step": 773
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.040756940841675,
+      "learning_rate": 1.9989218435828478e-05,
+      "loss": 1.1084,
+      "step": 774
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.290026903152466,
+      "learning_rate": 1.998913202646788e-05,
+      "loss": 1.2119,
+      "step": 775
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.152768611907959,
+      "learning_rate": 1.9989045272410242e-05,
+      "loss": 1.1365,
+      "step": 776
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.016942262649536,
+      "learning_rate": 1.9988958173658556e-05,
+      "loss": 1.105,
+      "step": 777
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.157235860824585,
+      "learning_rate": 1.9988870730215827e-05,
+      "loss": 1.014,
+      "step": 778
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.21296763420105,
+      "learning_rate": 1.998878294208507e-05,
+      "loss": 1.0783,
+      "step": 779
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2792489528656006,
+      "learning_rate": 1.9988694809269316e-05,
+      "loss": 1.1374,
+      "step": 780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.087193250656128,
+      "learning_rate": 1.9988606331771608e-05,
+      "loss": 1.203,
+      "step": 781
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.255445957183838,
+      "learning_rate": 1.9988517509594994e-05,
+      "loss": 1.1693,
+      "step": 782
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.696228504180908,
+      "learning_rate": 1.9988428342742544e-05,
+      "loss": 1.1136,
+      "step": 783
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1770858764648438,
+      "learning_rate": 1.9988338831217335e-05,
+      "loss": 1.0596,
+      "step": 784
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.083209991455078,
+      "learning_rate": 1.9988248975022455e-05,
+      "loss": 1.0909,
+      "step": 785
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.146885871887207,
+      "learning_rate": 1.9988158774161003e-05,
+      "loss": 0.5317,
+      "step": 786
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9633382558822632,
+      "learning_rate": 1.9988068228636092e-05,
+      "loss": 1.0857,
+      "step": 787
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.03809654712677,
+      "learning_rate": 1.9987977338450845e-05,
+      "loss": 0.5621,
+      "step": 788
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.098221778869629,
+      "learning_rate": 1.9987886103608403e-05,
+      "loss": 1.1404,
+      "step": 789
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0369505882263184,
+      "learning_rate": 1.998779452411191e-05,
+      "loss": 1.146,
+      "step": 790
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.35223126411438,
+      "learning_rate": 1.998770259996453e-05,
+      "loss": 1.1684,
+      "step": 791
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3191168308258057,
+      "learning_rate": 1.998761033116943e-05,
+      "loss": 1.1308,
+      "step": 792
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.250437021255493,
+      "learning_rate": 1.99875177177298e-05,
+      "loss": 1.1598,
+      "step": 793
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.364640712738037,
+      "learning_rate": 1.9987424759648834e-05,
+      "loss": 1.2441,
+      "step": 794
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.301056385040283,
+      "learning_rate": 1.9987331456929734e-05,
+      "loss": 1.1276,
+      "step": 795
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3249990940093994,
+      "learning_rate": 1.9987237809575722e-05,
+      "loss": 1.0879,
+      "step": 796
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.185084104537964,
+      "learning_rate": 1.998714381759004e-05,
+      "loss": 1.1389,
+      "step": 797
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1008193492889404,
+      "learning_rate": 1.9987049480975913e-05,
+      "loss": 1.0826,
+      "step": 798
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1822409629821777,
+      "learning_rate": 1.998695479973661e-05,
+      "loss": 1.1188,
+      "step": 799
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3375747203826904,
+      "learning_rate": 1.9986859773875397e-05,
+      "loss": 1.099,
+      "step": 800
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.00004506111145,
+      "learning_rate": 1.9986764403395546e-05,
+      "loss": 1.0803,
+      "step": 801
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.527397394180298,
+      "learning_rate": 1.9986668688300354e-05,
+      "loss": 1.1633,
+      "step": 802
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4554026126861572,
+      "learning_rate": 1.9986572628593124e-05,
+      "loss": 1.1651,
+      "step": 803
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0630898475646973,
+      "learning_rate": 1.9986476224277167e-05,
+      "loss": 1.149,
+      "step": 804
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.213650703430176,
+      "learning_rate": 1.9986379475355813e-05,
+      "loss": 1.1072,
+      "step": 805
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1035141944885254,
+      "learning_rate": 1.9986282381832396e-05,
+      "loss": 1.1364,
+      "step": 806
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2496273517608643,
+      "learning_rate": 1.9986184943710274e-05,
+      "loss": 1.1111,
+      "step": 807
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.480142593383789,
+      "learning_rate": 1.99860871609928e-05,
+      "loss": 1.1618,
+      "step": 808
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1323018074035645,
+      "learning_rate": 1.9985989033683357e-05,
+      "loss": 1.1097,
+      "step": 809
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1185803413391113,
+      "learning_rate": 1.9985890561785326e-05,
+      "loss": 1.1726,
+      "step": 810
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2804670333862305,
+      "learning_rate": 1.9985791745302108e-05,
+      "loss": 1.1251,
+      "step": 811
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.079725503921509,
+      "learning_rate": 1.998569258423711e-05,
+      "loss": 1.0719,
+      "step": 812
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9008687734603882,
+      "learning_rate": 1.9985593078593753e-05,
+      "loss": 1.0758,
+      "step": 813
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.283749580383301,
+      "learning_rate": 1.9985493228375473e-05,
+      "loss": 1.1424,
+      "step": 814
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.214712619781494,
+      "learning_rate": 1.9985393033585715e-05,
+      "loss": 1.1616,
+      "step": 815
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1830966472625732,
+      "learning_rate": 1.9985292494227937e-05,
+      "loss": 1.2065,
+      "step": 816
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2901535034179688,
+      "learning_rate": 1.9985191610305607e-05,
+      "loss": 1.0891,
+      "step": 817
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0829639434814453,
+      "learning_rate": 1.998509038182221e-05,
+      "loss": 1.0692,
+      "step": 818
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.216646432876587,
+      "learning_rate": 1.998498880878123e-05,
+      "loss": 1.1806,
+      "step": 819
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.129645586013794,
+      "learning_rate": 1.9984886891186184e-05,
+      "loss": 1.1581,
+      "step": 820
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1460018157958984,
+      "learning_rate": 1.9984784629040584e-05,
+      "loss": 1.1521,
+      "step": 821
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9727392196655273,
+      "learning_rate": 1.998468202234795e-05,
+      "loss": 1.0632,
+      "step": 822
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.717949867248535,
+      "learning_rate": 1.998457907111184e-05,
+      "loss": 1.1785,
+      "step": 823
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.086191415786743,
+      "learning_rate": 1.998447577533579e-05,
+      "loss": 1.0725,
+      "step": 824
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5673916339874268,
+      "learning_rate": 1.9984372135023375e-05,
+      "loss": 1.1785,
+      "step": 825
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.241105556488037,
+      "learning_rate": 1.998426815017817e-05,
+      "loss": 1.1964,
+      "step": 826
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2614150047302246,
+      "learning_rate": 1.9984163820803755e-05,
+      "loss": 1.1448,
+      "step": 827
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.063995599746704,
+      "learning_rate": 1.9984059146903738e-05,
+      "loss": 1.0534,
+      "step": 828
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.260887384414673,
+      "learning_rate": 1.998395412848173e-05,
+      "loss": 1.0778,
+      "step": 829
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.075248956680298,
+      "learning_rate": 1.9983848765541355e-05,
+      "loss": 1.1163,
+      "step": 830
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9940873384475708,
+      "learning_rate": 1.998374305808625e-05,
+      "loss": 1.1226,
+      "step": 831
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.212376356124878,
+      "learning_rate": 1.9983637006120054e-05,
+      "loss": 1.1138,
+      "step": 832
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.664011240005493,
+      "learning_rate": 1.998353060964644e-05,
+      "loss": 1.197,
+      "step": 833
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.432181715965271,
+      "learning_rate": 1.9983423868669068e-05,
+      "loss": 0.5363,
+      "step": 834
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5739877223968506,
+      "learning_rate": 1.9983316783191626e-05,
+      "loss": 1.1845,
+      "step": 835
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4608774185180664,
+      "learning_rate": 1.998320935321781e-05,
+      "loss": 1.1445,
+      "step": 836
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.483450412750244,
+      "learning_rate": 1.9983101578751326e-05,
+      "loss": 1.2228,
+      "step": 837
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2174501419067383,
+      "learning_rate": 1.9982993459795897e-05,
+      "loss": 1.1913,
+      "step": 838
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.273247241973877,
+      "learning_rate": 1.9982884996355248e-05,
+      "loss": 1.128,
+      "step": 839
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.36789608001709,
+      "learning_rate": 1.998277618843312e-05,
+      "loss": 1.1906,
+      "step": 840
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.30841064453125,
+      "learning_rate": 1.9982667036033274e-05,
+      "loss": 1.1506,
+      "step": 841
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3792459964752197,
+      "learning_rate": 1.9982557539159476e-05,
+      "loss": 1.1036,
+      "step": 842
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5239768028259277,
+      "learning_rate": 1.99824476978155e-05,
+      "loss": 1.1803,
+      "step": 843
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2788782119750977,
+      "learning_rate": 1.998233751200514e-05,
+      "loss": 1.0712,
+      "step": 844
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.088679552078247,
+      "learning_rate": 1.9982226981732197e-05,
+      "loss": 1.1522,
+      "step": 845
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.284238338470459,
+      "learning_rate": 1.9982116107000485e-05,
+      "loss": 1.1452,
+      "step": 846
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2757132053375244,
+      "learning_rate": 1.998200488781383e-05,
+      "loss": 1.1883,
+      "step": 847
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2445168495178223,
+      "learning_rate": 1.9981893324176067e-05,
+      "loss": 1.2275,
+      "step": 848
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1674227714538574,
+      "learning_rate": 1.998178141609105e-05,
+      "loss": 1.0809,
+      "step": 849
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.361021041870117,
+      "learning_rate": 1.9981669163562642e-05,
+      "loss": 1.1416,
+      "step": 850
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0345869064331055,
+      "learning_rate": 1.9981556566594712e-05,
+      "loss": 1.1171,
+      "step": 851
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9510729312896729,
+      "learning_rate": 1.9981443625191148e-05,
+      "loss": 1.072,
+      "step": 852
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1771600246429443,
+      "learning_rate": 1.9981330339355846e-05,
+      "loss": 1.0939,
+      "step": 853
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4299519062042236,
+      "learning_rate": 1.9981216709092715e-05,
+      "loss": 0.5579,
+      "step": 854
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.224153518676758,
+      "learning_rate": 1.9981102734405676e-05,
+      "loss": 1.1466,
+      "step": 855
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.350743055343628,
+      "learning_rate": 1.9980988415298666e-05,
+      "loss": 0.59,
+      "step": 856
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.270078659057617,
+      "learning_rate": 1.9980873751775625e-05,
+      "loss": 1.1537,
+      "step": 857
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3195390701293945,
+      "learning_rate": 1.998075874384051e-05,
+      "loss": 1.2126,
+      "step": 858
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1503279209136963,
+      "learning_rate": 1.998064339149729e-05,
+      "loss": 1.05,
+      "step": 859
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3343451023101807,
+      "learning_rate": 1.9980527694749952e-05,
+      "loss": 1.1342,
+      "step": 860
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1604702472686768,
+      "learning_rate": 1.9980411653602477e-05,
+      "loss": 1.1234,
+      "step": 861
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3645260334014893,
+      "learning_rate": 1.998029526805888e-05,
+      "loss": 1.1606,
+      "step": 862
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0832858085632324,
+      "learning_rate": 1.998017853812317e-05,
+      "loss": 1.0962,
+      "step": 863
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1821837425231934,
+      "learning_rate": 1.9980061463799374e-05,
+      "loss": 1.1678,
+      "step": 864
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.194148063659668,
+      "learning_rate": 1.9979944045091537e-05,
+      "loss": 1.1331,
+      "step": 865
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.22012197971344,
+      "learning_rate": 1.9979826282003708e-05,
+      "loss": 0.5133,
+      "step": 866
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0265305042266846,
+      "learning_rate": 1.9979708174539954e-05,
+      "loss": 1.0999,
+      "step": 867
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.016631841659546,
+      "learning_rate": 1.9979589722704348e-05,
+      "loss": 1.0722,
+      "step": 868
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9256972074508667,
+      "learning_rate": 1.9979470926500977e-05,
+      "loss": 1.1134,
+      "step": 869
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1395270824432373,
+      "learning_rate": 1.997935178593394e-05,
+      "loss": 1.1621,
+      "step": 870
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0393176078796387,
+      "learning_rate": 1.9979232301007348e-05,
+      "loss": 1.116,
+      "step": 871
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2395260334014893,
+      "learning_rate": 1.9979112471725326e-05,
+      "loss": 1.1365,
+      "step": 872
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9232733249664307,
+      "learning_rate": 1.997899229809201e-05,
+      "loss": 1.1507,
+      "step": 873
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.900830626487732,
+      "learning_rate": 1.9978871780111544e-05,
+      "loss": 1.0534,
+      "step": 874
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9295040369033813,
+      "learning_rate": 1.997875091778809e-05,
+      "loss": 1.0768,
+      "step": 875
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.216456651687622,
+      "learning_rate": 1.997862971112581e-05,
+      "loss": 1.1466,
+      "step": 876
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.250154972076416,
+      "learning_rate": 1.9978508160128896e-05,
+      "loss": 1.0858,
+      "step": 877
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1935391426086426,
+      "learning_rate": 1.997838626480154e-05,
+      "loss": 1.083,
+      "step": 878
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9666025638580322,
+      "learning_rate": 1.9978264025147947e-05,
+      "loss": 1.1814,
+      "step": 879
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9172236919403076,
+      "learning_rate": 1.997814144117234e-05,
+      "loss": 1.0622,
+      "step": 880
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1978591680526733,
+      "learning_rate": 1.9978018512878938e-05,
+      "loss": 0.5714,
+      "step": 881
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.999265193939209,
+      "learning_rate": 1.9977895240271992e-05,
+      "loss": 1.1923,
+      "step": 882
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.092198133468628,
+      "learning_rate": 1.9977771623355752e-05,
+      "loss": 1.0511,
+      "step": 883
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.153632164001465,
+      "learning_rate": 1.997764766213449e-05,
+      "loss": 1.1513,
+      "step": 884
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2152228355407715,
+      "learning_rate": 1.997752335661247e-05,
+      "loss": 1.0787,
+      "step": 885
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1043803691864014,
+      "learning_rate": 1.9977398706794e-05,
+      "loss": 1.0355,
+      "step": 886
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.202075481414795,
+      "learning_rate": 1.9977273712683366e-05,
+      "loss": 1.112,
+      "step": 887
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.01200008392334,
+      "learning_rate": 1.9977148374284886e-05,
+      "loss": 1.0716,
+      "step": 888
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.144124984741211,
+      "learning_rate": 1.9977022691602888e-05,
+      "loss": 1.1628,
+      "step": 889
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.320427894592285,
+      "learning_rate": 1.9976896664641706e-05,
+      "loss": 1.1079,
+      "step": 890
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.182781219482422,
+      "learning_rate": 1.9976770293405687e-05,
+      "loss": 1.0892,
+      "step": 891
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0892481803894043,
+      "learning_rate": 1.9976643577899194e-05,
+      "loss": 1.1918,
+      "step": 892
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1765358448028564,
+      "learning_rate": 1.9976516518126603e-05,
+      "loss": 1.1628,
+      "step": 893
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0505526065826416,
+      "learning_rate": 1.9976389114092293e-05,
+      "loss": 1.1478,
+      "step": 894
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2590537071228027,
+      "learning_rate": 1.9976261365800666e-05,
+      "loss": 1.1414,
+      "step": 895
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4349069595336914,
+      "learning_rate": 1.9976133273256126e-05,
+      "loss": 1.1164,
+      "step": 896
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8681683540344238,
+      "learning_rate": 1.997600483646309e-05,
+      "loss": 1.1127,
+      "step": 897
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2176530361175537,
+      "learning_rate": 1.9975876055425995e-05,
+      "loss": 1.0954,
+      "step": 898
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.453000068664551,
+      "learning_rate": 1.9975746930149287e-05,
+      "loss": 1.143,
+      "step": 899
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.231546401977539,
+      "learning_rate": 1.9975617460637417e-05,
+      "loss": 1.0971,
+      "step": 900
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0050864219665527,
+      "learning_rate": 1.9975487646894854e-05,
+      "loss": 1.1196,
+      "step": 901
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.118800640106201,
+      "learning_rate": 1.9975357488926077e-05,
+      "loss": 1.1198,
+      "step": 902
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1400129795074463,
+      "learning_rate": 1.9975226986735578e-05,
+      "loss": 1.1429,
+      "step": 903
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1674747467041016,
+      "learning_rate": 1.997509614032786e-05,
+      "loss": 1.0515,
+      "step": 904
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.41999888420105,
+      "learning_rate": 1.997496494970744e-05,
+      "loss": 1.1189,
+      "step": 905
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1876188516616821,
+      "learning_rate": 1.9974833414878846e-05,
+      "loss": 0.5326,
+      "step": 906
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1859302520751953,
+      "learning_rate": 1.997470153584661e-05,
+      "loss": 1.0934,
+      "step": 907
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0946156978607178,
+      "learning_rate": 1.997456931261529e-05,
+      "loss": 1.0923,
+      "step": 908
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.942283272743225,
+      "learning_rate": 1.9974436745189444e-05,
+      "loss": 1.0891,
+      "step": 909
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2902328968048096,
+      "learning_rate": 1.9974303833573646e-05,
+      "loss": 1.1458,
+      "step": 910
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.140209674835205,
+      "learning_rate": 1.997417057777249e-05,
+      "loss": 1.1264,
+      "step": 911
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2432613372802734,
+      "learning_rate": 1.9974036977790566e-05,
+      "loss": 1.0763,
+      "step": 912
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3520562648773193,
+      "learning_rate": 1.997390303363249e-05,
+      "loss": 1.1707,
+      "step": 913
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0676722526550293,
+      "learning_rate": 1.997376874530288e-05,
+      "loss": 1.1722,
+      "step": 914
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.253856897354126,
+      "learning_rate": 1.997363411280637e-05,
+      "loss": 1.1562,
+      "step": 915
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1041007041931152,
+      "learning_rate": 1.997349913614761e-05,
+      "loss": 1.0894,
+      "step": 916
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2459332942962646,
+      "learning_rate": 1.9973363815331248e-05,
+      "loss": 1.108,
+      "step": 917
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2529101371765137,
+      "learning_rate": 1.9973228150361965e-05,
+      "loss": 1.0633,
+      "step": 918
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.222435712814331,
+      "learning_rate": 1.9973092141244436e-05,
+      "loss": 1.1954,
+      "step": 919
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0947751998901367,
+      "learning_rate": 1.997295578798336e-05,
+      "loss": 1.1247,
+      "step": 920
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.228332757949829,
+      "learning_rate": 1.9972819090583433e-05,
+      "loss": 1.1268,
+      "step": 921
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1569807529449463,
+      "learning_rate": 1.9972682049049378e-05,
+      "loss": 1.0752,
+      "step": 922
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2037651538848877,
+      "learning_rate": 1.9972544663385927e-05,
+      "loss": 1.1937,
+      "step": 923
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0998852252960205,
+      "learning_rate": 1.9972406933597812e-05,
+      "loss": 0.996,
+      "step": 924
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0960259437561035,
+      "learning_rate": 1.9972268859689792e-05,
+      "loss": 1.2225,
+      "step": 925
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0508906841278076,
+      "learning_rate": 1.997213044166663e-05,
+      "loss": 1.1079,
+      "step": 926
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.122688055038452,
+      "learning_rate": 1.9971991679533103e-05,
+      "loss": 1.0868,
+      "step": 927
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0843045711517334,
+      "learning_rate": 1.9971852573294003e-05,
+      "loss": 1.087,
+      "step": 928
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.288818836212158,
+      "learning_rate": 1.997171312295412e-05,
+      "loss": 1.1175,
+      "step": 929
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.209801197052002,
+      "learning_rate": 1.9971573328518273e-05,
+      "loss": 1.1453,
+      "step": 930
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.073472738265991,
+      "learning_rate": 1.9971433189991286e-05,
+      "loss": 1.1381,
+      "step": 931
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2973203659057617,
+      "learning_rate": 1.997129270737799e-05,
+      "loss": 1.1153,
+      "step": 932
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0285723209381104,
+      "learning_rate": 1.997115188068324e-05,
+      "loss": 1.1248,
+      "step": 933
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3216347694396973,
+      "learning_rate": 1.9971010709911892e-05,
+      "loss": 1.1412,
+      "step": 934
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.091198205947876,
+      "learning_rate": 1.997086919506882e-05,
+      "loss": 1.1571,
+      "step": 935
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.179415225982666,
+      "learning_rate": 1.9970727336158902e-05,
+      "loss": 1.1332,
+      "step": 936
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1212592124938965,
+      "learning_rate": 1.9970585133187034e-05,
+      "loss": 1.1411,
+      "step": 937
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2156178951263428,
+      "learning_rate": 1.997044258615813e-05,
+      "loss": 1.1483,
+      "step": 938
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3334157466888428,
+      "learning_rate": 1.9970299695077095e-05,
+      "loss": 1.1107,
+      "step": 939
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.22662353515625,
+      "learning_rate": 1.9970156459948872e-05,
+      "loss": 1.1499,
+      "step": 940
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.175327777862549,
+      "learning_rate": 1.9970012880778404e-05,
+      "loss": 1.0863,
+      "step": 941
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9993934631347656,
+      "learning_rate": 1.9969868957570638e-05,
+      "loss": 1.0823,
+      "step": 942
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5160818099975586,
+      "learning_rate": 1.9969724690330543e-05,
+      "loss": 1.1077,
+      "step": 943
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1422719955444336,
+      "learning_rate": 1.9969580079063104e-05,
+      "loss": 0.5942,
+      "step": 944
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0571203231811523,
+      "learning_rate": 1.99694351237733e-05,
+      "loss": 0.5021,
+      "step": 945
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6621596813201904,
+      "learning_rate": 1.9969289824466138e-05,
+      "loss": 1.1798,
+      "step": 946
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3779985904693604,
+      "learning_rate": 1.9969144181146634e-05,
+      "loss": 1.1387,
+      "step": 947
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.196004629135132,
+      "learning_rate": 1.996899819381981e-05,
+      "loss": 1.165,
+      "step": 948
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9988819360733032,
+      "learning_rate": 1.996885186249071e-05,
+      "loss": 1.191,
+      "step": 949
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.956302523612976,
+      "learning_rate": 1.9968705187164375e-05,
+      "loss": 1.0856,
+      "step": 950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3769612312316895,
+      "learning_rate": 1.9968558167845873e-05,
+      "loss": 1.1194,
+      "step": 951
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7141518592834473,
+      "learning_rate": 1.9968410804540273e-05,
+      "loss": 1.162,
+      "step": 952
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9573251008987427,
+      "learning_rate": 1.9968263097252666e-05,
+      "loss": 1.0224,
+      "step": 953
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.112016201019287,
+      "learning_rate": 1.996811504598814e-05,
+      "loss": 1.1112,
+      "step": 954
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.384512424468994,
+      "learning_rate": 1.9967966650751808e-05,
+      "loss": 1.2376,
+      "step": 955
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.255016803741455,
+      "learning_rate": 1.9967817911548796e-05,
+      "loss": 1.1659,
+      "step": 956
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.12328839302063,
+      "learning_rate": 1.9967668828384227e-05,
+      "loss": 1.0954,
+      "step": 957
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.163557767868042,
+      "learning_rate": 1.996751940126325e-05,
+      "loss": 0.5296,
+      "step": 958
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.215634822845459,
+      "learning_rate": 1.996736963019103e-05,
+      "loss": 1.2042,
+      "step": 959
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.477297782897949,
+      "learning_rate": 1.996721951517272e-05,
+      "loss": 1.145,
+      "step": 960
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.229360342025757,
+      "learning_rate": 1.9967069056213503e-05,
+      "loss": 1.0969,
+      "step": 961
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1252036094665527,
+      "learning_rate": 1.996691825331858e-05,
+      "loss": 0.5572,
+      "step": 962
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.113152027130127,
+      "learning_rate": 1.996676710649315e-05,
+      "loss": 1.0193,
+      "step": 963
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.158536434173584,
+      "learning_rate": 1.9966615615742423e-05,
+      "loss": 1.1747,
+      "step": 964
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.04012131690979,
+      "learning_rate": 1.9966463781071637e-05,
+      "loss": 1.2352,
+      "step": 965
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0038750171661377,
+      "learning_rate": 1.9966311602486027e-05,
+      "loss": 1.1696,
+      "step": 966
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1827383041381836,
+      "learning_rate": 1.996615907999084e-05,
+      "loss": 1.0811,
+      "step": 967
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.425194025039673,
+      "learning_rate": 1.9966006213591346e-05,
+      "loss": 1.1728,
+      "step": 968
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1980483531951904,
+      "learning_rate": 1.9965853003292813e-05,
+      "loss": 1.0274,
+      "step": 969
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.334479331970215,
+      "learning_rate": 1.996569944910053e-05,
+      "loss": 1.0978,
+      "step": 970
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.199012517929077,
+      "learning_rate": 1.99655455510198e-05,
+      "loss": 1.0849,
+      "step": 971
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.089942455291748,
+      "learning_rate": 1.996539130905593e-05,
+      "loss": 1.096,
+      "step": 972
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2454452514648438,
+      "learning_rate": 1.9965236723214242e-05,
+      "loss": 0.5598,
+      "step": 973
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.190110206604004,
+      "learning_rate": 1.9965081793500073e-05,
+      "loss": 1.1166,
+      "step": 974
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2716000080108643,
+      "learning_rate": 1.996492651991877e-05,
+      "loss": 1.1074,
+      "step": 975
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.482877731323242,
+      "learning_rate": 1.9964770902475686e-05,
+      "loss": 1.058,
+      "step": 976
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.189729690551758,
+      "learning_rate": 1.9964614941176194e-05,
+      "loss": 1.1188,
+      "step": 977
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3052361011505127,
+      "learning_rate": 1.9964458636025673e-05,
+      "loss": 1.1462,
+      "step": 978
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1463403701782227,
+      "learning_rate": 1.9964301987029523e-05,
+      "loss": 1.0888,
+      "step": 979
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.032259941101074,
+      "learning_rate": 1.9964144994193143e-05,
+      "loss": 1.0542,
+      "step": 980
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0936243534088135,
+      "learning_rate": 1.996398765752195e-05,
+      "loss": 1.1355,
+      "step": 981
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9559952020645142,
+      "learning_rate": 1.9963829977021384e-05,
+      "loss": 1.0634,
+      "step": 982
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9189354181289673,
+      "learning_rate": 1.9963671952696876e-05,
+      "loss": 1.0758,
+      "step": 983
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0477375984191895,
+      "learning_rate": 1.9963513584553878e-05,
+      "loss": 1.0469,
+      "step": 984
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9667751789093018,
+      "learning_rate": 1.996335487259786e-05,
+      "loss": 1.108,
+      "step": 985
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4961416721343994,
+      "learning_rate": 1.9963195816834297e-05,
+      "loss": 1.1275,
+      "step": 986
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1340491771698,
+      "learning_rate": 1.9963036417268674e-05,
+      "loss": 1.0724,
+      "step": 987
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0602165460586548,
+      "learning_rate": 1.99628766739065e-05,
+      "loss": 0.5734,
+      "step": 988
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.125601053237915,
+      "learning_rate": 1.996271658675328e-05,
+      "loss": 1.0747,
+      "step": 989
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.029818296432495,
+      "learning_rate": 1.996255615581454e-05,
+      "loss": 1.1072,
+      "step": 990
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.109917640686035,
+      "learning_rate": 1.9962395381095817e-05,
+      "loss": 1.0952,
+      "step": 991
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0499300956726074,
+      "learning_rate": 1.9962234262602656e-05,
+      "loss": 1.1763,
+      "step": 992
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9918546676635742,
+      "learning_rate": 1.9962072800340625e-05,
+      "loss": 1.0755,
+      "step": 993
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.075986623764038,
+      "learning_rate": 1.9961910994315284e-05,
+      "loss": 1.0751,
+      "step": 994
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.937086582183838,
+      "learning_rate": 1.996174884453222e-05,
+      "loss": 1.0872,
+      "step": 995
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0692360401153564,
+      "learning_rate": 1.9961586350997037e-05,
+      "loss": 1.1102,
+      "step": 996
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.037346363067627,
+      "learning_rate": 1.9961423513715333e-05,
+      "loss": 1.094,
+      "step": 997
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.230212450027466,
+      "learning_rate": 1.9961260332692728e-05,
+      "loss": 1.0867,
+      "step": 998
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1400532722473145,
+      "learning_rate": 1.9961096807934855e-05,
+      "loss": 1.2046,
+      "step": 999
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0511980056762695,
+      "learning_rate": 1.996093293944736e-05,
+      "loss": 1.151,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9836289882659912,
+      "learning_rate": 1.9960768727235888e-05,
+      "loss": 1.1205,
+      "step": 1001
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0572667121887207,
+      "learning_rate": 1.9960604171306116e-05,
+      "loss": 1.082,
+      "step": 1002
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1632437705993652,
+      "learning_rate": 1.9960439271663717e-05,
+      "loss": 1.1114,
+      "step": 1003
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.331650495529175,
+      "learning_rate": 1.996027402831438e-05,
+      "loss": 1.1712,
+      "step": 1004
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2957890033721924,
+      "learning_rate": 1.996010844126381e-05,
+      "loss": 1.1529,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9415627717971802,
+      "learning_rate": 1.995994251051772e-05,
+      "loss": 1.0596,
+      "step": 1006
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.061467409133911,
+      "learning_rate": 1.9959776236081837e-05,
+      "loss": 1.0728,
+      "step": 1007
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1467807292938232,
+      "learning_rate": 1.9959609617961898e-05,
+      "loss": 1.1793,
+      "step": 1008
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.126514196395874,
+      "learning_rate": 1.9959442656163653e-05,
+      "loss": 1.1063,
+      "step": 1009
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0872507095336914,
+      "learning_rate": 1.9959275350692862e-05,
+      "loss": 1.1129,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.078003406524658,
+      "learning_rate": 1.9959107701555295e-05,
+      "loss": 1.0982,
+      "step": 1011
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3776357173919678,
+      "learning_rate": 1.9958939708756746e-05,
+      "loss": 1.1107,
+      "step": 1012
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.223130941390991,
+      "learning_rate": 1.9958771372303e-05,
+      "loss": 1.1322,
+      "step": 1013
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.150520086288452,
+      "learning_rate": 1.9958602692199883e-05,
+      "loss": 1.0494,
+      "step": 1014
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2399572134017944,
+      "learning_rate": 1.99584336684532e-05,
+      "loss": 0.5704,
+      "step": 1015
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.046128273010254,
+      "learning_rate": 1.9958264301068788e-05,
+      "loss": 1.1085,
+      "step": 1016
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2082788944244385,
+      "learning_rate": 1.995809459005249e-05,
+      "loss": 1.1426,
+      "step": 1017
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.005742073059082,
+      "learning_rate": 1.9957924535410166e-05,
+      "loss": 1.1323,
+      "step": 1018
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1082708835601807,
+      "learning_rate": 1.9957754137147683e-05,
+      "loss": 1.172,
+      "step": 1019
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9350619316101074,
+      "learning_rate": 1.9957583395270924e-05,
+      "loss": 1.1907,
+      "step": 1020
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.968170404434204,
+      "learning_rate": 1.9957412309785776e-05,
+      "loss": 1.1801,
+      "step": 1021
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2556710243225098,
+      "learning_rate": 1.995724088069814e-05,
+      "loss": 1.1172,
+      "step": 1022
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.90117347240448,
+      "learning_rate": 1.995706910801394e-05,
+      "loss": 1.1361,
+      "step": 1023
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2194061279296875,
+      "learning_rate": 1.9956896991739095e-05,
+      "loss": 1.1273,
+      "step": 1024
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3678059577941895,
+      "learning_rate": 1.995672453187955e-05,
+      "loss": 1.1841,
+      "step": 1025
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1089444160461426,
+      "learning_rate": 1.995655172844126e-05,
+      "loss": 1.1475,
+      "step": 1026
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.067833662033081,
+      "learning_rate": 1.9956378581430175e-05,
+      "loss": 1.1074,
+      "step": 1027
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.961103916168213,
+      "learning_rate": 1.995620509085228e-05,
+      "loss": 1.1198,
+      "step": 1028
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.339808702468872,
+      "learning_rate": 1.995603125671356e-05,
+      "loss": 1.0691,
+      "step": 1029
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1119163036346436,
+      "learning_rate": 1.995585707902001e-05,
+      "loss": 1.076,
+      "step": 1030
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0598556995391846,
+      "learning_rate": 1.9955682557777644e-05,
+      "loss": 1.1673,
+      "step": 1031
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.155109167098999,
+      "learning_rate": 1.9955507692992482e-05,
+      "loss": 1.104,
+      "step": 1032
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.061760663986206,
+      "learning_rate": 1.995533248467056e-05,
+      "loss": 1.0374,
+      "step": 1033
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8904452323913574,
+      "learning_rate": 1.995515693281792e-05,
+      "loss": 1.0838,
+      "step": 1034
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0027434825897217,
+      "learning_rate": 1.995498103744063e-05,
+      "loss": 1.1246,
+      "step": 1035
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.001688003540039,
+      "learning_rate": 1.9954804798544748e-05,
+      "loss": 1.1625,
+      "step": 1036
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0883662700653076,
+      "learning_rate": 1.995462821613636e-05,
+      "loss": 1.1605,
+      "step": 1037
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1495673656463623,
+      "learning_rate": 1.9954451290221558e-05,
+      "loss": 1.1232,
+      "step": 1038
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0513052940368652,
+      "learning_rate": 1.995427402080645e-05,
+      "loss": 1.032,
+      "step": 1039
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2151739597320557,
+      "learning_rate": 1.9954096407897154e-05,
+      "loss": 1.0669,
+      "step": 1040
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0288562774658203,
+      "learning_rate": 1.9953918451499797e-05,
+      "loss": 1.216,
+      "step": 1041
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1106815338134766,
+      "learning_rate": 1.9953740151620515e-05,
+      "loss": 1.1526,
+      "step": 1042
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.88935124874115,
+      "learning_rate": 1.9953561508265466e-05,
+      "loss": 1.136,
+      "step": 1043
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.146790027618408,
+      "learning_rate": 1.9953382521440816e-05,
+      "loss": 1.1014,
+      "step": 1044
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2105722427368164,
+      "learning_rate": 1.9953203191152736e-05,
+      "loss": 1.1983,
+      "step": 1045
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9890414476394653,
+      "learning_rate": 1.995302351740742e-05,
+      "loss": 1.149,
+      "step": 1046
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.050382137298584,
+      "learning_rate": 1.9952843500211062e-05,
+      "loss": 1.0702,
+      "step": 1047
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1800410747528076,
+      "learning_rate": 1.995266313956988e-05,
+      "loss": 1.133,
+      "step": 1048
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.163120985031128,
+      "learning_rate": 1.9952482435490094e-05,
+      "loss": 1.1368,
+      "step": 1049
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1013380289077759,
+      "learning_rate": 1.9952301387977937e-05,
+      "loss": 0.5949,
+      "step": 1050
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.332686185836792,
+      "learning_rate": 1.9952119997039664e-05,
+      "loss": 1.102,
+      "step": 1051
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0294029712677,
+      "learning_rate": 1.9951938262681527e-05,
+      "loss": 1.0779,
+      "step": 1052
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4254696369171143,
+      "learning_rate": 1.99517561849098e-05,
+      "loss": 1.1108,
+      "step": 1053
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8786166906356812,
+      "learning_rate": 1.995157376373077e-05,
+      "loss": 1.0588,
+      "step": 1054
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1033036708831787,
+      "learning_rate": 1.9951390999150723e-05,
+      "loss": 1.1072,
+      "step": 1055
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1297013759613037,
+      "learning_rate": 1.9951207891175973e-05,
+      "loss": 0.5833,
+      "step": 1056
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.257452964782715,
+      "learning_rate": 1.9951024439812837e-05,
+      "loss": 1.0726,
+      "step": 1057
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0183420181274414,
+      "learning_rate": 1.9950840645067643e-05,
+      "loss": 1.0994,
+      "step": 1058
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3594791889190674,
+      "learning_rate": 1.9950656506946737e-05,
+      "loss": 1.0316,
+      "step": 1059
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9593797326087952,
+      "learning_rate": 1.9950472025456472e-05,
+      "loss": 0.5592,
+      "step": 1060
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1703927516937256,
+      "learning_rate": 1.9950287200603208e-05,
+      "loss": 1.1544,
+      "step": 1061
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1522278785705566,
+      "learning_rate": 1.995010203239333e-05,
+      "loss": 1.0469,
+      "step": 1062
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4177536964416504,
+      "learning_rate": 1.9949916520833228e-05,
+      "loss": 1.1175,
+      "step": 1063
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.304537534713745,
+      "learning_rate": 1.99497306659293e-05,
+      "loss": 1.1106,
+      "step": 1064
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.193523406982422,
+      "learning_rate": 1.9949544467687957e-05,
+      "loss": 1.1737,
+      "step": 1065
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5079402923583984,
+      "learning_rate": 1.994935792611563e-05,
+      "loss": 1.1795,
+      "step": 1066
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.331172466278076,
+      "learning_rate": 1.9949171041218754e-05,
+      "loss": 1.1103,
+      "step": 1067
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.36053729057312,
+      "learning_rate": 1.9948983813003776e-05,
+      "loss": 1.127,
+      "step": 1068
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.178652763366699,
+      "learning_rate": 1.9948796241477157e-05,
+      "loss": 1.1637,
+      "step": 1069
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9500905275344849,
+      "learning_rate": 1.994860832664537e-05,
+      "loss": 1.1083,
+      "step": 1070
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9139505624771118,
+      "learning_rate": 1.9948420068514904e-05,
+      "loss": 1.0694,
+      "step": 1071
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4539225101470947,
+      "learning_rate": 1.9948231467092248e-05,
+      "loss": 1.0925,
+      "step": 1072
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2782180309295654,
+      "learning_rate": 1.9948042522383915e-05,
+      "loss": 1.0413,
+      "step": 1073
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1228864192962646,
+      "learning_rate": 1.9947853234396423e-05,
+      "loss": 0.6092,
+      "step": 1074
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0677382946014404,
+      "learning_rate": 1.99476636031363e-05,
+      "loss": 0.9715,
+      "step": 1075
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9538472890853882,
+      "learning_rate": 1.99474736286101e-05,
+      "loss": 1.0481,
+      "step": 1076
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.227790117263794,
+      "learning_rate": 1.994728331082437e-05,
+      "loss": 1.1765,
+      "step": 1077
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1632280349731445,
+      "learning_rate": 1.9947092649785675e-05,
+      "loss": 1.1241,
+      "step": 1078
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1495938301086426,
+      "learning_rate": 1.9946901645500604e-05,
+      "loss": 1.1532,
+      "step": 1079
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1210474967956543,
+      "learning_rate": 1.9946710297975743e-05,
+      "loss": 1.0604,
+      "step": 1080
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.076061487197876,
+      "learning_rate": 1.994651860721769e-05,
+      "loss": 1.1047,
+      "step": 1081
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0261378288269043,
+      "learning_rate": 1.9946326573233068e-05,
+      "loss": 1.1278,
+      "step": 1082
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0028727054595947,
+      "learning_rate": 1.99461341960285e-05,
+      "loss": 1.1157,
+      "step": 1083
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0081613063812256,
+      "learning_rate": 1.9945941475610623e-05,
+      "loss": 1.0783,
+      "step": 1084
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.984207034111023,
+      "learning_rate": 1.994574841198609e-05,
+      "loss": 1.1184,
+      "step": 1085
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1088995933532715,
+      "learning_rate": 1.9945555005161562e-05,
+      "loss": 1.0614,
+      "step": 1086
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.206160545349121,
+      "learning_rate": 1.9945361255143712e-05,
+      "loss": 1.1837,
+      "step": 1087
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.813061237335205,
+      "learning_rate": 1.9945167161939225e-05,
+      "loss": 1.1681,
+      "step": 1088
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0713086128234863,
+      "learning_rate": 1.99449727255548e-05,
+      "loss": 1.123,
+      "step": 1089
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0151636600494385,
+      "learning_rate": 1.9944777945997146e-05,
+      "loss": 1.184,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.175811290740967,
+      "learning_rate": 1.9944582823272985e-05,
+      "loss": 1.1253,
+      "step": 1091
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.008009672164917,
+      "learning_rate": 1.994438735738905e-05,
+      "loss": 1.1741,
+      "step": 1092
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1865029335021973,
+      "learning_rate": 1.9944191548352088e-05,
+      "loss": 0.513,
+      "step": 1093
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.547581434249878,
+      "learning_rate": 1.994399539616885e-05,
+      "loss": 1.0834,
+      "step": 1094
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.438591241836548,
+      "learning_rate": 1.994379890084611e-05,
+      "loss": 1.1187,
+      "step": 1095
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3844690322875977,
+      "learning_rate": 1.994360206239065e-05,
+      "loss": 1.1299,
+      "step": 1096
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0553101301193237,
+      "learning_rate": 1.9943404880809254e-05,
+      "loss": 0.5425,
+      "step": 1097
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0255768299102783,
+      "learning_rate": 1.9943207356108733e-05,
+      "loss": 1.0835,
+      "step": 1098
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.084327220916748,
+      "learning_rate": 1.9943009488295903e-05,
+      "loss": 0.5322,
+      "step": 1099
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4381301403045654,
+      "learning_rate": 1.994281127737759e-05,
+      "loss": 1.1489,
+      "step": 1100
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0825705528259277,
+      "learning_rate": 1.9942612723360632e-05,
+      "loss": 1.1164,
+      "step": 1101
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.114013910293579,
+      "learning_rate": 1.9942413826251885e-05,
+      "loss": 1.077,
+      "step": 1102
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.288700819015503,
+      "learning_rate": 1.994221458605821e-05,
+      "loss": 1.1001,
+      "step": 1103
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0281717777252197,
+      "learning_rate": 1.994201500278648e-05,
+      "loss": 1.1124,
+      "step": 1104
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1523780822753906,
+      "learning_rate": 1.9941815076443586e-05,
+      "loss": 1.0916,
+      "step": 1105
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1003830432891846,
+      "learning_rate": 1.994161480703642e-05,
+      "loss": 1.0749,
+      "step": 1106
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0702121257781982,
+      "learning_rate": 1.9941414194571905e-05,
+      "loss": 1.0608,
+      "step": 1107
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.113926649093628,
+      "learning_rate": 1.994121323905695e-05,
+      "loss": 0.5346,
+      "step": 1108
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3483593463897705,
+      "learning_rate": 1.9941011940498497e-05,
+      "loss": 1.1051,
+      "step": 1109
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1991989612579346,
+      "learning_rate": 1.9940810298903495e-05,
+      "loss": 1.0794,
+      "step": 1110
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0061705112457275,
+      "learning_rate": 1.9940608314278895e-05,
+      "loss": 1.1055,
+      "step": 1111
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.084113836288452,
+      "learning_rate": 1.994040598663167e-05,
+      "loss": 1.0985,
+      "step": 1112
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2391879558563232,
+      "learning_rate": 1.9940203315968806e-05,
+      "loss": 1.1343,
+      "step": 1113
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9646164178848267,
+      "learning_rate": 1.994000030229729e-05,
+      "loss": 1.1927,
+      "step": 1114
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.478668451309204,
+      "learning_rate": 1.9939796945624127e-05,
+      "loss": 1.1542,
+      "step": 1115
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0290873050689697,
+      "learning_rate": 1.993959324595634e-05,
+      "loss": 1.1116,
+      "step": 1116
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.150301456451416,
+      "learning_rate": 1.9939389203300952e-05,
+      "loss": 1.1534,
+      "step": 1117
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1240397691726685,
+      "learning_rate": 1.993918481766501e-05,
+      "loss": 0.5673,
+      "step": 1118
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1398959159851074,
+      "learning_rate": 1.9938980089055565e-05,
+      "loss": 1.0975,
+      "step": 1119
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2083184719085693,
+      "learning_rate": 1.9938775017479678e-05,
+      "loss": 1.0974,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2078986167907715,
+      "learning_rate": 1.993856960294443e-05,
+      "loss": 0.6306,
+      "step": 1121
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0882461071014404,
+      "learning_rate": 1.9938363845456904e-05,
+      "loss": 1.1713,
+      "step": 1122
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.103302240371704,
+      "learning_rate": 1.9938157745024208e-05,
+      "loss": 1.0596,
+      "step": 1123
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0410983562469482,
+      "learning_rate": 1.9937951301653444e-05,
+      "loss": 1.1503,
+      "step": 1124
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.01668381690979,
+      "learning_rate": 1.9937744515351746e-05,
+      "loss": 1.173,
+      "step": 1125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4383652210235596,
+      "learning_rate": 1.9937537386126242e-05,
+      "loss": 1.1395,
+      "step": 1126
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0189781188964844,
+      "learning_rate": 1.9937329913984084e-05,
+      "loss": 1.036,
+      "step": 1127
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.122084379196167,
+      "learning_rate": 1.9937122098932428e-05,
+      "loss": 1.1449,
+      "step": 1128
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.011646270751953,
+      "learning_rate": 1.9936913940978447e-05,
+      "loss": 1.0708,
+      "step": 1129
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0631135702133179,
+      "learning_rate": 1.9936705440129326e-05,
+      "loss": 0.5822,
+      "step": 1130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0753798484802246,
+      "learning_rate": 1.9936496596392253e-05,
+      "loss": 1.1244,
+      "step": 1131
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1454896926879883,
+      "learning_rate": 1.993628740977444e-05,
+      "loss": 1.1368,
+      "step": 1132
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8620635271072388,
+      "learning_rate": 1.9936077880283108e-05,
+      "loss": 1.1696,
+      "step": 1133
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1776223182678223,
+      "learning_rate": 1.993586800792548e-05,
+      "loss": 1.0594,
+      "step": 1134
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2247000932693481,
+      "learning_rate": 1.9935657792708803e-05,
+      "loss": 0.5753,
+      "step": 1135
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2741165161132812,
+      "learning_rate": 1.9935447234640328e-05,
+      "loss": 1.1022,
+      "step": 1136
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1057348251342773,
+      "learning_rate": 1.9935236333727322e-05,
+      "loss": 1.1318,
+      "step": 1137
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.031235456466675,
+      "learning_rate": 1.9935025089977067e-05,
+      "loss": 1.0913,
+      "step": 1138
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0759148597717285,
+      "learning_rate": 1.9934813503396847e-05,
+      "loss": 1.2036,
+      "step": 1139
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0053555965423584,
+      "learning_rate": 1.993460157399396e-05,
+      "loss": 1.1345,
+      "step": 1140
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.204301357269287,
+      "learning_rate": 1.993438930177573e-05,
+      "loss": 1.0714,
+      "step": 1141
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1923940181732178,
+      "learning_rate": 1.9934176686749476e-05,
+      "loss": 1.1855,
+      "step": 1142
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2842936515808105,
+      "learning_rate": 1.9933963728922532e-05,
+      "loss": 1.1546,
+      "step": 1143
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.091099262237549,
+      "learning_rate": 1.993375042830225e-05,
+      "loss": 1.1156,
+      "step": 1144
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0759966373443604,
+      "learning_rate": 1.993353678489599e-05,
+      "loss": 0.5534,
+      "step": 1145
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2083656787872314,
+      "learning_rate": 1.993332279871112e-05,
+      "loss": 1.1551,
+      "step": 1146
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9549657106399536,
+      "learning_rate": 1.9933108469755032e-05,
+      "loss": 1.0497,
+      "step": 1147
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.377147912979126,
+      "learning_rate": 1.9932893798035118e-05,
+      "loss": 1.1436,
+      "step": 1148
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3849093914031982,
+      "learning_rate": 1.9932678783558785e-05,
+      "loss": 1.1112,
+      "step": 1149
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.092904806137085,
+      "learning_rate": 1.993246342633345e-05,
+      "loss": 0.9871,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9337904453277588,
+      "learning_rate": 1.993224772636655e-05,
+      "loss": 1.0408,
+      "step": 1151
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0744550228118896,
+      "learning_rate": 1.9932031683665523e-05,
+      "loss": 1.1605,
+      "step": 1152
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0451710224151611,
+      "learning_rate": 1.993181529823783e-05,
+      "loss": 0.549,
+      "step": 1153
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3024978637695312,
+      "learning_rate": 1.993159857009093e-05,
+      "loss": 1.1693,
+      "step": 1154
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.349405527114868,
+      "learning_rate": 1.993138149923231e-05,
+      "loss": 1.1245,
+      "step": 1155
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3075320720672607,
+      "learning_rate": 1.9931164085669456e-05,
+      "loss": 1.1042,
+      "step": 1156
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9487121105194092,
+      "learning_rate": 1.993094632940987e-05,
+      "loss": 1.0848,
+      "step": 1157
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.157090187072754,
+      "learning_rate": 1.993072823046107e-05,
+      "loss": 0.539,
+      "step": 1158
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2875044345855713,
+      "learning_rate": 1.9930509788830575e-05,
+      "loss": 1.1062,
+      "step": 1159
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.094050168991089,
+      "learning_rate": 1.993029100452593e-05,
+      "loss": 1.1257,
+      "step": 1160
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2589635848999023,
+      "learning_rate": 1.9930071877554683e-05,
+      "loss": 1.0986,
+      "step": 1161
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.198604106903076,
+      "learning_rate": 1.9929852407924392e-05,
+      "loss": 1.1653,
+      "step": 1162
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9290698766708374,
+      "learning_rate": 1.992963259564263e-05,
+      "loss": 1.1296,
+      "step": 1163
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0985058546066284,
+      "learning_rate": 1.9929412440716988e-05,
+      "loss": 0.6325,
+      "step": 1164
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.008709669113159,
+      "learning_rate": 1.9929191943155057e-05,
+      "loss": 1.0745,
+      "step": 1165
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.088191270828247,
+      "learning_rate": 1.9928971102964447e-05,
+      "loss": 1.1129,
+      "step": 1166
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0278618335723877,
+      "learning_rate": 1.992874992015278e-05,
+      "loss": 1.0919,
+      "step": 1167
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0869240760803223,
+      "learning_rate": 1.992852839472769e-05,
+      "loss": 1.0378,
+      "step": 1168
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.011765241622925,
+      "learning_rate": 1.992830652669682e-05,
+      "loss": 1.12,
+      "step": 1169
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0093207359313965,
+      "learning_rate": 1.9928084316067823e-05,
+      "loss": 1.0846,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.262693405151367,
+      "learning_rate": 1.992786176284837e-05,
+      "loss": 1.0494,
+      "step": 1171
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1503779888153076,
+      "learning_rate": 1.9927638867046143e-05,
+      "loss": 1.0779,
+      "step": 1172
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.014343738555908,
+      "learning_rate": 1.992741562866883e-05,
+      "loss": 1.1392,
+      "step": 1173
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0086891651153564,
+      "learning_rate": 1.992719204772413e-05,
+      "loss": 1.095,
+      "step": 1174
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0909321308135986,
+      "learning_rate": 1.9926968124219767e-05,
+      "loss": 1.2311,
+      "step": 1175
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2445292472839355,
+      "learning_rate": 1.9926743858163463e-05,
+      "loss": 1.2062,
+      "step": 1176
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0266518592834473,
+      "learning_rate": 1.9926519249562955e-05,
+      "loss": 1.1218,
+      "step": 1177
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.060330390930176,
+      "learning_rate": 1.9926294298426e-05,
+      "loss": 1.2147,
+      "step": 1178
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.157761335372925,
+      "learning_rate": 1.9926069004760356e-05,
+      "loss": 1.1117,
+      "step": 1179
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1145176887512207,
+      "learning_rate": 1.9925843368573792e-05,
+      "loss": 1.1634,
+      "step": 1180
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.917935848236084,
+      "learning_rate": 1.9925617389874108e-05,
+      "loss": 1.0402,
+      "step": 1181
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.046093463897705,
+      "learning_rate": 1.9925391068669093e-05,
+      "loss": 1.1241,
+      "step": 1182
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0455591678619385,
+      "learning_rate": 1.9925164404966556e-05,
+      "loss": 1.1234,
+      "step": 1183
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8993330001831055,
+      "learning_rate": 1.992493739877432e-05,
+      "loss": 1.0769,
+      "step": 1184
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1404120922088623,
+      "learning_rate": 1.9924710050100217e-05,
+      "loss": 1.1305,
+      "step": 1185
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.159856081008911,
+      "learning_rate": 1.9924482358952098e-05,
+      "loss": 1.084,
+      "step": 1186
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.113464117050171,
+      "learning_rate": 1.9924254325337812e-05,
+      "loss": 1.091,
+      "step": 1187
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.978493332862854,
+      "learning_rate": 1.9924025949265232e-05,
+      "loss": 1.1047,
+      "step": 1188
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1816928386688232,
+      "learning_rate": 1.992379723074224e-05,
+      "loss": 1.1469,
+      "step": 1189
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.222705602645874,
+      "learning_rate": 1.9923568169776725e-05,
+      "loss": 1.1837,
+      "step": 1190
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8997925519943237,
+      "learning_rate": 1.9923338766376593e-05,
+      "loss": 1.1483,
+      "step": 1191
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.17057204246521,
+      "learning_rate": 1.992310902054976e-05,
+      "loss": 1.147,
+      "step": 1192
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1332062482833862,
+      "learning_rate": 1.9922878932304152e-05,
+      "loss": 0.556,
+      "step": 1193
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1345150470733643,
+      "learning_rate": 1.9922648501647714e-05,
+      "loss": 1.0373,
+      "step": 1194
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1468448638916016,
+      "learning_rate": 1.9922417728588394e-05,
+      "loss": 1.086,
+      "step": 1195
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1684858798980713,
+      "learning_rate": 1.9922186613134152e-05,
+      "loss": 1.1128,
+      "step": 1196
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2784924507141113,
+      "learning_rate": 1.9921955155292968e-05,
+      "loss": 1.1567,
+      "step": 1197
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0885435342788696,
+      "learning_rate": 1.992172335507283e-05,
+      "loss": 0.5883,
+      "step": 1198
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2788376808166504,
+      "learning_rate": 1.992149121248173e-05,
+      "loss": 1.1182,
+      "step": 1199
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1025798320770264,
+      "learning_rate": 1.9921258727527685e-05,
+      "loss": 1.0308,
+      "step": 1200
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1366171836853027,
+      "learning_rate": 1.9921025900218715e-05,
+      "loss": 1.1531,
+      "step": 1201
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.092592477798462,
+      "learning_rate": 1.9920792730562853e-05,
+      "loss": 1.116,
+      "step": 1202
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.01593017578125,
+      "learning_rate": 1.992055921856815e-05,
+      "loss": 1.0914,
+      "step": 1203
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9623132944107056,
+      "learning_rate": 1.9920325364242658e-05,
+      "loss": 1.0876,
+      "step": 1204
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1434998512268066,
+      "learning_rate": 1.992009116759445e-05,
+      "loss": 1.1394,
+      "step": 1205
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0466878414154053,
+      "learning_rate": 1.9919856628631606e-05,
+      "loss": 1.0924,
+      "step": 1206
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0747780799865723,
+      "learning_rate": 1.991962174736222e-05,
+      "loss": 1.1534,
+      "step": 1207
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9507187604904175,
+      "learning_rate": 1.9919386523794396e-05,
+      "loss": 1.1179,
+      "step": 1208
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.163022994995117,
+      "learning_rate": 1.9919150957936252e-05,
+      "loss": 1.1486,
+      "step": 1209
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.39188814163208,
+      "learning_rate": 1.9918915049795916e-05,
+      "loss": 1.097,
+      "step": 1210
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0100901126861572,
+      "learning_rate": 1.991867879938153e-05,
+      "loss": 1.1397,
+      "step": 1211
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9968924522399902,
+      "learning_rate": 1.9918442206701244e-05,
+      "loss": 1.1673,
+      "step": 1212
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9430325031280518,
+      "learning_rate": 1.9918205271763225e-05,
+      "loss": 1.144,
+      "step": 1213
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3547229766845703,
+      "learning_rate": 1.9917967994575646e-05,
+      "loss": 1.1018,
+      "step": 1214
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.100503444671631,
+      "learning_rate": 1.9917730375146697e-05,
+      "loss": 1.0594,
+      "step": 1215
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8590503931045532,
+      "learning_rate": 1.991749241348458e-05,
+      "loss": 1.0954,
+      "step": 1216
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.084115505218506,
+      "learning_rate": 1.9917254109597496e-05,
+      "loss": 1.2,
+      "step": 1217
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0761280059814453,
+      "learning_rate": 1.991701546349368e-05,
+      "loss": 1.1649,
+      "step": 1218
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.069512128829956,
+      "learning_rate": 1.9916776475181363e-05,
+      "loss": 1.1133,
+      "step": 1219
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9715789556503296,
+      "learning_rate": 1.991653714466879e-05,
+      "loss": 1.095,
+      "step": 1220
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4183547496795654,
+      "learning_rate": 1.991629747196422e-05,
+      "loss": 1.14,
+      "step": 1221
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.221222400665283,
+      "learning_rate": 1.9916057457075925e-05,
+      "loss": 1.1237,
+      "step": 1222
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0392346382141113,
+      "learning_rate": 1.991581710001219e-05,
+      "loss": 1.0845,
+      "step": 1223
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1710734367370605,
+      "learning_rate": 1.99155764007813e-05,
+      "loss": 1.105,
+      "step": 1224
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1350514888763428,
+      "learning_rate": 1.991533535939157e-05,
+      "loss": 1.0766,
+      "step": 1225
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0256454944610596,
+      "learning_rate": 1.9915093975851313e-05,
+      "loss": 1.1516,
+      "step": 1226
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1307461261749268,
+      "learning_rate": 1.991485225016886e-05,
+      "loss": 1.1361,
+      "step": 1227
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1100406646728516,
+      "learning_rate": 1.991461018235255e-05,
+      "loss": 1.0608,
+      "step": 1228
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9077250957489014,
+      "learning_rate": 1.991436777241074e-05,
+      "loss": 1.1867,
+      "step": 1229
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2169835567474365,
+      "learning_rate": 1.9914125020351794e-05,
+      "loss": 1.1167,
+      "step": 1230
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.226994276046753,
+      "learning_rate": 1.9913881926184084e-05,
+      "loss": 1.1852,
+      "step": 1231
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.021265983581543,
+      "learning_rate": 1.9913638489916006e-05,
+      "loss": 1.1297,
+      "step": 1232
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1348133087158203,
+      "learning_rate": 1.991339471155595e-05,
+      "loss": 1.065,
+      "step": 1233
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2522923946380615,
+      "learning_rate": 1.991315059111234e-05,
+      "loss": 1.102,
+      "step": 1234
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0552234649658203,
+      "learning_rate": 1.9912906128593595e-05,
+      "loss": 1.1832,
+      "step": 1235
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2361905574798584,
+      "learning_rate": 1.9912661324008147e-05,
+      "loss": 1.1429,
+      "step": 1236
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.079887866973877,
+      "learning_rate": 1.991241617736445e-05,
+      "loss": 0.5342,
+      "step": 1237
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2332348823547363,
+      "learning_rate": 1.9912170688670956e-05,
+      "loss": 1.2202,
+      "step": 1238
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3862452507019043,
+      "learning_rate": 1.9911924857936142e-05,
+      "loss": 1.1217,
+      "step": 1239
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9727429151535034,
+      "learning_rate": 1.9911678685168486e-05,
+      "loss": 1.0248,
+      "step": 1240
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2376153469085693,
+      "learning_rate": 1.991143217037649e-05,
+      "loss": 1.1876,
+      "step": 1241
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.046086311340332,
+      "learning_rate": 1.9911185313568655e-05,
+      "loss": 1.0718,
+      "step": 1242
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0690810680389404,
+      "learning_rate": 1.99109381147535e-05,
+      "loss": 1.091,
+      "step": 1243
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.156420946121216,
+      "learning_rate": 1.991069057393956e-05,
+      "loss": 1.1693,
+      "step": 1244
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3771722316741943,
+      "learning_rate": 1.9910442691135364e-05,
+      "loss": 1.0267,
+      "step": 1245
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.202235460281372,
+      "learning_rate": 1.991019446634948e-05,
+      "loss": 1.1042,
+      "step": 1246
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.227205753326416,
+      "learning_rate": 1.9909945899590468e-05,
+      "loss": 1.1327,
+      "step": 1247
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.118591785430908,
+      "learning_rate": 1.9909696990866903e-05,
+      "loss": 1.1394,
+      "step": 1248
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1210837364196777,
+      "learning_rate": 1.990944774018738e-05,
+      "loss": 1.0889,
+      "step": 1249
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3601319789886475,
+      "learning_rate": 1.9909198147560492e-05,
+      "loss": 1.0905,
+      "step": 1250
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.173114061355591,
+      "learning_rate": 1.990894821299486e-05,
+      "loss": 1.133,
+      "step": 1251
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1293811798095703,
+      "learning_rate": 1.9908697936499105e-05,
+      "loss": 1.0958,
+      "step": 1252
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.268125295639038,
+      "learning_rate": 1.990844731808186e-05,
+      "loss": 1.1005,
+      "step": 1253
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2594525814056396,
+      "learning_rate": 1.9908196357751778e-05,
+      "loss": 1.1161,
+      "step": 1254
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.474522829055786,
+      "learning_rate": 1.9907945055517517e-05,
+      "loss": 1.1654,
+      "step": 1255
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.132887840270996,
+      "learning_rate": 1.990769341138775e-05,
+      "loss": 1.1022,
+      "step": 1256
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.103699207305908,
+      "learning_rate": 1.9907441425371155e-05,
+      "loss": 1.0682,
+      "step": 1257
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9943654537200928,
+      "learning_rate": 1.9907189097476434e-05,
+      "loss": 1.1605,
+      "step": 1258
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.103573799133301,
+      "learning_rate": 1.9906936427712295e-05,
+      "loss": 1.0961,
+      "step": 1259
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9811022281646729,
+      "learning_rate": 1.990668341608745e-05,
+      "loss": 1.1111,
+      "step": 1260
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.086775064468384,
+      "learning_rate": 1.9906430062610634e-05,
+      "loss": 1.0567,
+      "step": 1261
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.085899591445923,
+      "learning_rate": 1.990617636729059e-05,
+      "loss": 1.1824,
+      "step": 1262
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.326277017593384,
+      "learning_rate": 1.990592233013607e-05,
+      "loss": 1.1446,
+      "step": 1263
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0884342193603516,
+      "learning_rate": 1.9905667951155846e-05,
+      "loss": 1.1348,
+      "step": 1264
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2065106630325317,
+      "learning_rate": 1.9905413230358687e-05,
+      "loss": 0.5621,
+      "step": 1265
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.869314432144165,
+      "learning_rate": 1.990515816775339e-05,
+      "loss": 1.173,
+      "step": 1266
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.268204689025879,
+      "learning_rate": 1.990490276334875e-05,
+      "loss": 1.1702,
+      "step": 1267
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2038028240203857,
+      "learning_rate": 1.9904647017153584e-05,
+      "loss": 1.1344,
+      "step": 1268
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0081677436828613,
+      "learning_rate": 1.9904390929176716e-05,
+      "loss": 1.1108,
+      "step": 1269
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0541017055511475,
+      "learning_rate": 1.990413449942699e-05,
+      "loss": 1.0601,
+      "step": 1270
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.270810842514038,
+      "learning_rate": 1.9903877727913245e-05,
+      "loss": 1.0704,
+      "step": 1271
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0008022785186768,
+      "learning_rate": 1.9903620614644344e-05,
+      "loss": 1.106,
+      "step": 1272
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.928165078163147,
+      "learning_rate": 1.990336315962916e-05,
+      "loss": 1.0916,
+      "step": 1273
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9835573434829712,
+      "learning_rate": 1.990310536287658e-05,
+      "loss": 1.0741,
+      "step": 1274
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0650382041931152,
+      "learning_rate": 1.9902847224395495e-05,
+      "loss": 1.1363,
+      "step": 1275
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.894583821296692,
+      "learning_rate": 1.9902588744194815e-05,
+      "loss": 1.1038,
+      "step": 1276
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.855492353439331,
+      "learning_rate": 1.990232992228346e-05,
+      "loss": 1.0987,
+      "step": 1277
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9532442092895508,
+      "learning_rate": 1.9902070758670357e-05,
+      "loss": 1.1315,
+      "step": 1278
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.154186725616455,
+      "learning_rate": 1.9901811253364458e-05,
+      "loss": 1.0325,
+      "step": 1279
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9850854873657227,
+      "learning_rate": 1.9901551406374707e-05,
+      "loss": 1.0386,
+      "step": 1280
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3562185764312744,
+      "learning_rate": 1.990129121771008e-05,
+      "loss": 1.0897,
+      "step": 1281
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9751865863800049,
+      "learning_rate": 1.990103068737955e-05,
+      "loss": 1.0541,
+      "step": 1282
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1249632835388184,
+      "learning_rate": 1.9900769815392106e-05,
+      "loss": 0.5727,
+      "step": 1283
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2207763195037842,
+      "learning_rate": 1.9900508601756755e-05,
+      "loss": 0.6802,
+      "step": 1284
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.27069354057312,
+      "learning_rate": 1.9900247046482507e-05,
+      "loss": 1.1251,
+      "step": 1285
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1679444313049316,
+      "learning_rate": 1.989998514957839e-05,
+      "loss": 1.1064,
+      "step": 1286
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.284734010696411,
+      "learning_rate": 1.9899722911053438e-05,
+      "loss": 1.2097,
+      "step": 1287
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1041972637176514,
+      "learning_rate": 1.9899460330916706e-05,
+      "loss": 0.6174,
+      "step": 1288
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0110692977905273,
+      "learning_rate": 1.9899197409177245e-05,
+      "loss": 1.1213,
+      "step": 1289
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0938215255737305,
+      "learning_rate": 1.989893414584414e-05,
+      "loss": 1.1117,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9446513652801514,
+      "learning_rate": 1.9898670540926463e-05,
+      "loss": 1.1011,
+      "step": 1291
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0141987800598145,
+      "learning_rate": 1.989840659443332e-05,
+      "loss": 1.1537,
+      "step": 1292
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9681075811386108,
+      "learning_rate": 1.9898142306373816e-05,
+      "loss": 1.0889,
+      "step": 1293
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.014925241470337,
+      "learning_rate": 1.989787767675707e-05,
+      "loss": 1.0744,
+      "step": 1294
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2276012897491455,
+      "learning_rate": 1.9897612705592215e-05,
+      "loss": 0.6677,
+      "step": 1295
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2037341594696045,
+      "learning_rate": 1.989734739288839e-05,
+      "loss": 1.1658,
+      "step": 1296
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.159756660461426,
+      "learning_rate": 1.9897081738654754e-05,
+      "loss": 1.0833,
+      "step": 1297
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.310328483581543,
+      "learning_rate": 1.989681574290048e-05,
+      "loss": 1.1384,
+      "step": 1298
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9768158197402954,
+      "learning_rate": 1.989654940563473e-05,
+      "loss": 1.2013,
+      "step": 1299
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9453753232955933,
+      "learning_rate": 1.9896282726866713e-05,
+      "loss": 1.0661,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.080828905105591,
+      "learning_rate": 1.989601570660562e-05,
+      "loss": 1.0526,
+      "step": 1301
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1605697870254517,
+      "learning_rate": 1.9895748344860667e-05,
+      "loss": 0.6008,
+      "step": 1302
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5316574573516846,
+      "learning_rate": 1.9895480641641086e-05,
+      "loss": 1.1312,
+      "step": 1303
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1274194717407227,
+      "learning_rate": 1.9895212596956104e-05,
+      "loss": 1.047,
+      "step": 1304
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2341084480285645,
+      "learning_rate": 1.989494421081498e-05,
+      "loss": 1.08,
+      "step": 1305
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1698412895202637,
+      "learning_rate": 1.989467548322697e-05,
+      "loss": 1.1666,
+      "step": 1306
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1324745416641235,
+      "learning_rate": 1.989440641420135e-05,
+      "loss": 0.5499,
+      "step": 1307
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3786816596984863,
+      "learning_rate": 1.9894137003747404e-05,
+      "loss": 1.1462,
+      "step": 1308
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2287256717681885,
+      "learning_rate": 1.9893867251874428e-05,
+      "loss": 1.1352,
+      "step": 1309
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.190324544906616,
+      "learning_rate": 1.9893597158591728e-05,
+      "loss": 1.0935,
+      "step": 1310
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.148191452026367,
+      "learning_rate": 1.9893326723908634e-05,
+      "loss": 1.1491,
+      "step": 1311
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.108069658279419,
+      "learning_rate": 1.9893055947834464e-05,
+      "loss": 1.1705,
+      "step": 1312
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.154458999633789,
+      "learning_rate": 1.989278483037857e-05,
+      "loss": 1.0735,
+      "step": 1313
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1555721759796143,
+      "learning_rate": 1.9892513371550303e-05,
+      "loss": 1.1025,
+      "step": 1314
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0585250854492188,
+      "learning_rate": 1.9892241571359035e-05,
+      "loss": 1.103,
+      "step": 1315
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9872242212295532,
+      "learning_rate": 1.9891969429814147e-05,
+      "loss": 1.1643,
+      "step": 1316
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.19966459274292,
+      "learning_rate": 1.9891696946925024e-05,
+      "loss": 1.0738,
+      "step": 1317
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0834670066833496,
+      "learning_rate": 1.9891424122701067e-05,
+      "loss": 0.5802,
+      "step": 1318
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0283498764038086,
+      "learning_rate": 1.9891150957151696e-05,
+      "loss": 1.1474,
+      "step": 1319
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.191458225250244,
+      "learning_rate": 1.989087745028634e-05,
+      "loss": 1.143,
+      "step": 1320
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.289520740509033,
+      "learning_rate": 1.9890603602114428e-05,
+      "loss": 1.0767,
+      "step": 1321
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1580259799957275,
+      "learning_rate": 1.9890329412645417e-05,
+      "loss": 1.2069,
+      "step": 1322
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.199875593185425,
+      "learning_rate": 1.9890054881888758e-05,
+      "loss": 1.138,
+      "step": 1323
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0213574171066284,
+      "learning_rate": 1.988978000985394e-05,
+      "loss": 0.5916,
+      "step": 1324
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.097529888153076,
+      "learning_rate": 1.9889504796550435e-05,
+      "loss": 1.0179,
+      "step": 1325
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.137336492538452,
+      "learning_rate": 1.9889229241987747e-05,
+      "loss": 1.1846,
+      "step": 1326
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.064444065093994,
+      "learning_rate": 1.9888953346175383e-05,
+      "loss": 1.1054,
+      "step": 1327
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.009007692337036,
+      "learning_rate": 1.988867710912286e-05,
+      "loss": 1.1348,
+      "step": 1328
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9943597316741943,
+      "learning_rate": 1.9888400530839713e-05,
+      "loss": 1.1011,
+      "step": 1329
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.055872917175293,
+      "learning_rate": 1.988812361133549e-05,
+      "loss": 1.0661,
+      "step": 1330
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2273976802825928,
+      "learning_rate": 1.9887846350619736e-05,
+      "loss": 1.1775,
+      "step": 1331
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1020774841308594,
+      "learning_rate": 1.9887568748702032e-05,
+      "loss": 1.1429,
+      "step": 1332
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1669255495071411,
+      "learning_rate": 1.9887290805591946e-05,
+      "loss": 0.6053,
+      "step": 1333
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.249624013900757,
+      "learning_rate": 1.9887012521299072e-05,
+      "loss": 1.158,
+      "step": 1334
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.042140483856201,
+      "learning_rate": 1.9886733895833017e-05,
+      "loss": 1.1374,
+      "step": 1335
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.120642900466919,
+      "learning_rate": 1.9886454929203394e-05,
+      "loss": 1.0874,
+      "step": 1336
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.055746555328369,
+      "learning_rate": 1.9886175621419824e-05,
+      "loss": 1.0788,
+      "step": 1337
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.217156171798706,
+      "learning_rate": 1.9885895972491952e-05,
+      "loss": 1.0792,
+      "step": 1338
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9077961444854736,
+      "learning_rate": 1.9885615982429425e-05,
+      "loss": 1.0627,
+      "step": 1339
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0082945823669434,
+      "learning_rate": 1.9885335651241905e-05,
+      "loss": 1.1286,
+      "step": 1340
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9954739809036255,
+      "learning_rate": 1.9885054978939062e-05,
+      "loss": 1.0454,
+      "step": 1341
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9854028224945068,
+      "learning_rate": 1.988477396553059e-05,
+      "loss": 1.1135,
+      "step": 1342
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.302316188812256,
+      "learning_rate": 1.9884492611026177e-05,
+      "loss": 1.1539,
+      "step": 1343
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.038797616958618,
+      "learning_rate": 1.9884210915435536e-05,
+      "loss": 1.1108,
+      "step": 1344
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0623676776885986,
+      "learning_rate": 1.9883928878768386e-05,
+      "loss": 1.1101,
+      "step": 1345
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9175207614898682,
+      "learning_rate": 1.988364650103446e-05,
+      "loss": 1.0681,
+      "step": 1346
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.099000930786133,
+      "learning_rate": 1.98833637822435e-05,
+      "loss": 1.1031,
+      "step": 1347
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9874680042266846,
+      "learning_rate": 1.988308072240527e-05,
+      "loss": 1.0718,
+      "step": 1348
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3376691341400146,
+      "learning_rate": 1.9882797321529526e-05,
+      "loss": 1.048,
+      "step": 1349
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.420262336730957,
+      "learning_rate": 1.9882513579626056e-05,
+      "loss": 1.1714,
+      "step": 1350
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8753174543380737,
+      "learning_rate": 1.9882229496704647e-05,
+      "loss": 1.0122,
+      "step": 1351
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2582755088806152,
+      "learning_rate": 1.9881945072775106e-05,
+      "loss": 1.1672,
+      "step": 1352
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.077871561050415,
+      "learning_rate": 1.988166030784724e-05,
+      "loss": 1.1688,
+      "step": 1353
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.053107261657715,
+      "learning_rate": 1.988137520193088e-05,
+      "loss": 1.0456,
+      "step": 1354
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0064969062805176,
+      "learning_rate": 1.9881089755035864e-05,
+      "loss": 1.1126,
+      "step": 1355
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.039945363998413,
+      "learning_rate": 1.9880803967172048e-05,
+      "loss": 1.1581,
+      "step": 1356
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1099343299865723,
+      "learning_rate": 1.988051783834928e-05,
+      "loss": 1.0637,
+      "step": 1357
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2105233669281006,
+      "learning_rate": 1.988023136857745e-05,
+      "loss": 1.1763,
+      "step": 1358
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.122349262237549,
+      "learning_rate": 1.987994455786643e-05,
+      "loss": 1.0911,
+      "step": 1359
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1639063358306885,
+      "learning_rate": 1.9879657406226123e-05,
+      "loss": 1.1139,
+      "step": 1360
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.036627769470215,
+      "learning_rate": 1.9879369913666434e-05,
+      "loss": 1.1323,
+      "step": 1361
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.110201835632324,
+      "learning_rate": 1.9879082080197288e-05,
+      "loss": 1.1668,
+      "step": 1362
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3206636905670166,
+      "learning_rate": 1.987879390582862e-05,
+      "loss": 1.039,
+      "step": 1363
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0441465377807617,
+      "learning_rate": 1.987850539057036e-05,
+      "loss": 1.0845,
+      "step": 1364
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.022278070449829,
+      "learning_rate": 1.987821653443248e-05,
+      "loss": 1.0997,
+      "step": 1365
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1065540313720703,
+      "learning_rate": 1.987792733742494e-05,
+      "loss": 1.0949,
+      "step": 1366
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2054505348205566,
+      "learning_rate": 1.987763779955772e-05,
+      "loss": 1.0267,
+      "step": 1367
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.15751576423645,
+      "learning_rate": 1.9877347920840812e-05,
+      "loss": 1.0617,
+      "step": 1368
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1031882762908936,
+      "learning_rate": 1.9877057701284217e-05,
+      "loss": 1.1658,
+      "step": 1369
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1594465970993042,
+      "learning_rate": 1.9876767140897953e-05,
+      "loss": 0.5538,
+      "step": 1370
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0366275310516357,
+      "learning_rate": 1.9876476239692045e-05,
+      "loss": 1.0723,
+      "step": 1371
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.989363670349121,
+      "learning_rate": 1.987618499767653e-05,
+      "loss": 1.1171,
+      "step": 1372
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2282488346099854,
+      "learning_rate": 1.987589341486146e-05,
+      "loss": 1.1912,
+      "step": 1373
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.123347043991089,
+      "learning_rate": 1.9875601491256893e-05,
+      "loss": 1.117,
+      "step": 1374
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9939744472503662,
+      "learning_rate": 1.9875309226872907e-05,
+      "loss": 1.109,
+      "step": 1375
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9738434553146362,
+      "learning_rate": 1.9875016621719584e-05,
+      "loss": 1.1615,
+      "step": 1376
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1679413318634033,
+      "learning_rate": 1.987472367580702e-05,
+      "loss": 1.151,
+      "step": 1377
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0676372051239014,
+      "learning_rate": 1.987443038914533e-05,
+      "loss": 1.1471,
+      "step": 1378
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3361377716064453,
+      "learning_rate": 1.987413676174463e-05,
+      "loss": 1.0903,
+      "step": 1379
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0200390815734863,
+      "learning_rate": 1.987384279361505e-05,
+      "loss": 1.0676,
+      "step": 1380
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.272378444671631,
+      "learning_rate": 1.9873548484766737e-05,
+      "loss": 1.288,
+      "step": 1381
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0421066284179688,
+      "learning_rate": 1.9873253835209848e-05,
+      "loss": 1.0762,
+      "step": 1382
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2948830127716064,
+      "learning_rate": 1.9872958844954548e-05,
+      "loss": 0.575,
+      "step": 1383
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.772346258163452,
+      "learning_rate": 1.9872663514011016e-05,
+      "loss": 1.1606,
+      "step": 1384
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1205649375915527,
+      "learning_rate": 1.9872367842389448e-05,
+      "loss": 1.1125,
+      "step": 1385
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0386877059936523,
+      "learning_rate": 1.987207183010004e-05,
+      "loss": 1.1183,
+      "step": 1386
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.187908411026001,
+      "learning_rate": 1.9871775477153012e-05,
+      "loss": 1.0917,
+      "step": 1387
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3645827770233154,
+      "learning_rate": 1.9871478783558586e-05,
+      "loss": 1.0845,
+      "step": 1388
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4528961181640625,
+      "learning_rate": 1.9871181749327e-05,
+      "loss": 1.1272,
+      "step": 1389
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.261669397354126,
+      "learning_rate": 1.9870884374468512e-05,
+      "loss": 1.0338,
+      "step": 1390
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2685914039611816,
+      "learning_rate": 1.9870586658993375e-05,
+      "loss": 1.1373,
+      "step": 1391
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9602514505386353,
+      "learning_rate": 1.987028860291186e-05,
+      "loss": 1.1259,
+      "step": 1392
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.986748218536377,
+      "learning_rate": 1.986999020623426e-05,
+      "loss": 1.1957,
+      "step": 1393
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.138457775115967,
+      "learning_rate": 1.9869691468970873e-05,
+      "loss": 1.1411,
+      "step": 1394
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.100825309753418,
+      "learning_rate": 1.9869392391132e-05,
+      "loss": 1.0838,
+      "step": 1395
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.173616409301758,
+      "learning_rate": 1.986909297272796e-05,
+      "loss": 1.03,
+      "step": 1396
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.161248207092285,
+      "learning_rate": 1.9868793213769096e-05,
+      "loss": 1.1236,
+      "step": 1397
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2289650440216064,
+      "learning_rate": 1.9868493114265743e-05,
+      "loss": 1.1663,
+      "step": 1398
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0525355339050293,
+      "learning_rate": 1.986819267422826e-05,
+      "loss": 1.0905,
+      "step": 1399
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.002760648727417,
+      "learning_rate": 1.9867891893667012e-05,
+      "loss": 1.0849,
+      "step": 1400
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.117975950241089,
+      "learning_rate": 1.9867590772592384e-05,
+      "loss": 1.1182,
+      "step": 1401
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1687686443328857,
+      "learning_rate": 1.9867289311014756e-05,
+      "loss": 1.0956,
+      "step": 1402
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.150573253631592,
+      "learning_rate": 1.986698750894454e-05,
+      "loss": 1.1856,
+      "step": 1403
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9842984676361084,
+      "learning_rate": 1.9866685366392148e-05,
+      "loss": 1.0706,
+      "step": 1404
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9404444694519043,
+      "learning_rate": 1.9866382883368005e-05,
+      "loss": 1.2065,
+      "step": 1405
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1583411693573,
+      "learning_rate": 1.9866080059882547e-05,
+      "loss": 1.139,
+      "step": 1406
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3189854621887207,
+      "learning_rate": 1.986577689594623e-05,
+      "loss": 0.6255,
+      "step": 1407
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.076507806777954,
+      "learning_rate": 1.986547339156951e-05,
+      "loss": 1.1835,
+      "step": 1408
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1921162605285645,
+      "learning_rate": 1.986516954676286e-05,
+      "loss": 1.1611,
+      "step": 1409
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1837761402130127,
+      "learning_rate": 1.986486536153677e-05,
+      "loss": 1.0383,
+      "step": 1410
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0888397693634033,
+      "learning_rate": 1.9864560835901728e-05,
+      "loss": 1.0853,
+      "step": 1411
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9503192901611328,
+      "learning_rate": 1.9864255969868248e-05,
+      "loss": 1.0118,
+      "step": 1412
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9684503078460693,
+      "learning_rate": 1.986395076344685e-05,
+      "loss": 1.0846,
+      "step": 1413
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9836734533309937,
+      "learning_rate": 1.9863645216648067e-05,
+      "loss": 1.0429,
+      "step": 1414
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.076437473297119,
+      "learning_rate": 1.9863339329482437e-05,
+      "loss": 1.1452,
+      "step": 1415
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7226722240447998,
+      "learning_rate": 1.986303310196052e-05,
+      "loss": 1.0353,
+      "step": 1416
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0731446743011475,
+      "learning_rate": 1.9862726534092884e-05,
+      "loss": 1.1249,
+      "step": 1417
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.060102105140686,
+      "learning_rate": 1.9862419625890104e-05,
+      "loss": 0.5784,
+      "step": 1418
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0847055912017822,
+      "learning_rate": 1.9862112377362773e-05,
+      "loss": 1.1337,
+      "step": 1419
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1206746101379395,
+      "learning_rate": 1.986180478852149e-05,
+      "loss": 1.2428,
+      "step": 1420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3331515789031982,
+      "learning_rate": 1.9861496859376876e-05,
+      "loss": 1.2532,
+      "step": 1421
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1501359939575195,
+      "learning_rate": 1.9861188589939548e-05,
+      "loss": 1.0665,
+      "step": 1422
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.265876531600952,
+      "learning_rate": 1.9860879980220148e-05,
+      "loss": 1.0789,
+      "step": 1423
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1215591430664062,
+      "learning_rate": 1.9860571030229327e-05,
+      "loss": 1.051,
+      "step": 1424
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.048396348953247,
+      "learning_rate": 1.9860261739977748e-05,
+      "loss": 1.1216,
+      "step": 1425
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0101122856140137,
+      "learning_rate": 1.9859952109476076e-05,
+      "loss": 1.2047,
+      "step": 1426
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0449963808059692,
+      "learning_rate": 1.9859642138734995e-05,
+      "loss": 0.5079,
+      "step": 1427
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.464859962463379,
+      "learning_rate": 1.9859331827765214e-05,
+      "loss": 1.1511,
+      "step": 1428
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9723495244979858,
+      "learning_rate": 1.9859021176577426e-05,
+      "loss": 1.0401,
+      "step": 1429
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.02600359916687,
+      "learning_rate": 1.985871018518236e-05,
+      "loss": 1.1363,
+      "step": 1430
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9769604206085205,
+      "learning_rate": 1.9858398853590745e-05,
+      "loss": 1.1085,
+      "step": 1431
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1852800846099854,
+      "learning_rate": 1.9858087181813325e-05,
+      "loss": 1.1026,
+      "step": 1432
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.135044813156128,
+      "learning_rate": 1.985777516986085e-05,
+      "loss": 1.0805,
+      "step": 1433
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.150658130645752,
+      "learning_rate": 1.9857462817744093e-05,
+      "loss": 0.9837,
+      "step": 1434
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1276448965072632,
+      "learning_rate": 1.9857150125473826e-05,
+      "loss": 0.545,
+      "step": 1435
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1883535385131836,
+      "learning_rate": 1.985683709306085e-05,
+      "loss": 0.633,
+      "step": 1436
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2820518016815186,
+      "learning_rate": 1.9856523720515954e-05,
+      "loss": 1.1596,
+      "step": 1437
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.283904790878296,
+      "learning_rate": 1.9856210007849965e-05,
+      "loss": 1.1307,
+      "step": 1438
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0505027770996094,
+      "learning_rate": 1.985589595507369e-05,
+      "loss": 1.141,
+      "step": 1439
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.033250331878662,
+      "learning_rate": 1.9855581562197985e-05,
+      "loss": 1.0985,
+      "step": 1440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9846283197402954,
+      "learning_rate": 1.9855266829233688e-05,
+      "loss": 1.0485,
+      "step": 1441
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2820816040039062,
+      "learning_rate": 1.9854951756191664e-05,
+      "loss": 1.0767,
+      "step": 1442
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.033313751220703,
+      "learning_rate": 1.9854636343082784e-05,
+      "loss": 1.1953,
+      "step": 1443
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9995808601379395,
+      "learning_rate": 1.9854320589917928e-05,
+      "loss": 1.0346,
+      "step": 1444
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.438791036605835,
+      "learning_rate": 1.9854004496708e-05,
+      "loss": 1.1486,
+      "step": 1445
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9010323286056519,
+      "learning_rate": 1.9853688063463896e-05,
+      "loss": 1.0793,
+      "step": 1446
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.157346248626709,
+      "learning_rate": 1.9853371290196547e-05,
+      "loss": 1.0036,
+      "step": 1447
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.060274600982666,
+      "learning_rate": 1.9853054176916877e-05,
+      "loss": 1.0971,
+      "step": 1448
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.134073495864868,
+      "learning_rate": 1.985273672363583e-05,
+      "loss": 1.0651,
+      "step": 1449
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9882314205169678,
+      "learning_rate": 1.9852418930364366e-05,
+      "loss": 1.0077,
+      "step": 1450
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0374741554260254,
+      "learning_rate": 1.9852100797113443e-05,
+      "loss": 1.0779,
+      "step": 1451
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1281180381774902,
+      "learning_rate": 1.985178232389404e-05,
+      "loss": 1.0282,
+      "step": 1452
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0943431854248047,
+      "learning_rate": 1.9851463510717154e-05,
+      "loss": 1.0647,
+      "step": 1453
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0908193588256836,
+      "learning_rate": 1.985114435759378e-05,
+      "loss": 1.1352,
+      "step": 1454
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9162050485610962,
+      "learning_rate": 1.985082486453493e-05,
+      "loss": 1.0421,
+      "step": 1455
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.172769784927368,
+      "learning_rate": 1.985050503155163e-05,
+      "loss": 1.1203,
+      "step": 1456
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0628693103790283,
+      "learning_rate": 1.985018485865492e-05,
+      "loss": 1.0958,
+      "step": 1457
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1457321643829346,
+      "learning_rate": 1.9849864345855847e-05,
+      "loss": 1.1264,
+      "step": 1458
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.002023696899414,
+      "learning_rate": 1.984954349316547e-05,
+      "loss": 1.0712,
+      "step": 1459
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8922641277313232,
+      "learning_rate": 1.984922230059486e-05,
+      "loss": 1.0871,
+      "step": 1460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9675027132034302,
+      "learning_rate": 1.98489007681551e-05,
+      "loss": 1.1516,
+      "step": 1461
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0231220722198486,
+      "learning_rate": 1.9848578895857286e-05,
+      "loss": 1.0958,
+      "step": 1462
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0484676361083984,
+      "learning_rate": 1.9848256683712524e-05,
+      "loss": 1.1855,
+      "step": 1463
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0738344192504883,
+      "learning_rate": 1.984793413173194e-05,
+      "loss": 1.0251,
+      "step": 1464
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0766384601593018,
+      "learning_rate": 1.984761123992665e-05,
+      "loss": 1.0946,
+      "step": 1465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0351288318634033,
+      "learning_rate": 1.984728800830781e-05,
+      "loss": 1.0554,
+      "step": 1466
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.038163900375366,
+      "learning_rate": 1.9846964436886567e-05,
+      "loss": 1.0246,
+      "step": 1467
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9563249349594116,
+      "learning_rate": 1.9846640525674085e-05,
+      "loss": 1.097,
+      "step": 1468
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8876897096633911,
+      "learning_rate": 1.9846316274681547e-05,
+      "loss": 1.0876,
+      "step": 1469
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9763652086257935,
+      "learning_rate": 1.9845991683920136e-05,
+      "loss": 1.149,
+      "step": 1470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1362626552581787,
+      "learning_rate": 1.9845666753401057e-05,
+      "loss": 1.0093,
+      "step": 1471
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1086034774780273,
+      "learning_rate": 1.984534148313552e-05,
+      "loss": 1.0959,
+      "step": 1472
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.030452251434326,
+      "learning_rate": 1.9845015873134754e-05,
+      "loss": 1.067,
+      "step": 1473
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.197798728942871,
+      "learning_rate": 1.984468992340999e-05,
+      "loss": 1.075,
+      "step": 1474
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.923909068107605,
+      "learning_rate": 1.9844363633972477e-05,
+      "loss": 1.1272,
+      "step": 1475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0921998023986816,
+      "learning_rate": 1.984403700483347e-05,
+      "loss": 1.149,
+      "step": 1476
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9096180200576782,
+      "learning_rate": 1.984371003600425e-05,
+      "loss": 1.1273,
+      "step": 1477
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9657894372940063,
+      "learning_rate": 1.984338272749609e-05,
+      "loss": 1.0852,
+      "step": 1478
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.162951946258545,
+      "learning_rate": 1.9843055079320292e-05,
+      "loss": 1.0699,
+      "step": 1479
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.099306344985962,
+      "learning_rate": 1.9842727091488153e-05,
+      "loss": 1.1107,
+      "step": 1480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.193556547164917,
+      "learning_rate": 1.9842398764011e-05,
+      "loss": 1.1818,
+      "step": 1481
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0320000648498535,
+      "learning_rate": 1.9842070096900158e-05,
+      "loss": 1.0553,
+      "step": 1482
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.025590658187866,
+      "learning_rate": 1.984174109016697e-05,
+      "loss": 1.1743,
+      "step": 1483
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.200209856033325,
+      "learning_rate": 1.9841411743822792e-05,
+      "loss": 1.1698,
+      "step": 1484
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9037857055664062,
+      "learning_rate": 1.984108205787898e-05,
+      "loss": 1.0508,
+      "step": 1485
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9431012868881226,
+      "learning_rate": 1.984075203234692e-05,
+      "loss": 1.0926,
+      "step": 1486
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3005175590515137,
+      "learning_rate": 1.9840421667237998e-05,
+      "loss": 1.0724,
+      "step": 1487
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7720915079116821,
+      "learning_rate": 1.984009096256361e-05,
+      "loss": 0.9589,
+      "step": 1488
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0283432006835938,
+      "learning_rate": 1.9839759918335168e-05,
+      "loss": 0.5142,
+      "step": 1489
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9947141408920288,
+      "learning_rate": 1.98394285345641e-05,
+      "loss": 1.1564,
+      "step": 1490
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.241518974304199,
+      "learning_rate": 1.9839096811261838e-05,
+      "loss": 1.121,
+      "step": 1491
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1695207357406616,
+      "learning_rate": 1.983876474843983e-05,
+      "loss": 0.5971,
+      "step": 1492
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0368034839630127,
+      "learning_rate": 1.983843234610953e-05,
+      "loss": 1.0841,
+      "step": 1493
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2994043827056885,
+      "learning_rate": 1.9838099604282418e-05,
+      "loss": 1.1654,
+      "step": 1494
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.300075054168701,
+      "learning_rate": 1.983776652296997e-05,
+      "loss": 1.1359,
+      "step": 1495
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1157824993133545,
+      "learning_rate": 1.9837433102183677e-05,
+      "loss": 1.0804,
+      "step": 1496
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9550588130950928,
+      "learning_rate": 1.9837099341935052e-05,
+      "loss": 0.9942,
+      "step": 1497
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.383256673812866,
+      "learning_rate": 1.9836765242235604e-05,
+      "loss": 1.1689,
+      "step": 1498
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9425299167633057,
+      "learning_rate": 1.9836430803096865e-05,
+      "loss": 1.0567,
+      "step": 1499
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.98634672164917,
+      "learning_rate": 1.9836096024530373e-05,
+      "loss": 1.0626,
+      "step": 1500
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.000257968902588,
+      "learning_rate": 1.983576090654769e-05,
+      "loss": 1.0638,
+      "step": 1501
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.091320037841797,
+      "learning_rate": 1.9835425449160367e-05,
+      "loss": 1.0527,
+      "step": 1502
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2507221698760986,
+      "learning_rate": 1.983508965237999e-05,
+      "loss": 1.127,
+      "step": 1503
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.130293130874634,
+      "learning_rate": 1.9834753516218138e-05,
+      "loss": 1.0692,
+      "step": 1504
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.126779317855835,
+      "learning_rate": 1.983441704068642e-05,
+      "loss": 1.1485,
+      "step": 1505
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.020146369934082,
+      "learning_rate": 1.9834080225796438e-05,
+      "loss": 1.1033,
+      "step": 1506
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2400009632110596,
+      "learning_rate": 1.983374307155982e-05,
+      "loss": 1.063,
+      "step": 1507
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0688021183013916,
+      "learning_rate": 1.9833405577988198e-05,
+      "loss": 1.0404,
+      "step": 1508
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9649642705917358,
+      "learning_rate": 1.9833067745093214e-05,
+      "loss": 1.0055,
+      "step": 1509
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9329557418823242,
+      "learning_rate": 1.9832729572886533e-05,
+      "loss": 1.0667,
+      "step": 1510
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0614495277404785,
+      "learning_rate": 1.983239106137982e-05,
+      "loss": 1.1431,
+      "step": 1511
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2273495197296143,
+      "learning_rate": 1.983205221058476e-05,
+      "loss": 1.1212,
+      "step": 1512
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9913157224655151,
+      "learning_rate": 1.9831713020513038e-05,
+      "loss": 1.0755,
+      "step": 1513
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.638934373855591,
+      "learning_rate": 1.9831373491176365e-05,
+      "loss": 1.124,
+      "step": 1514
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0530943870544434,
+      "learning_rate": 1.983103362258646e-05,
+      "loss": 1.1661,
+      "step": 1515
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8650323152542114,
+      "learning_rate": 1.983069341475504e-05,
+      "loss": 1.1003,
+      "step": 1516
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0902040004730225,
+      "learning_rate": 1.9830352867693854e-05,
+      "loss": 1.1327,
+      "step": 1517
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.889133334159851,
+      "learning_rate": 1.983001198141465e-05,
+      "loss": 1.1197,
+      "step": 1518
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9659128189086914,
+      "learning_rate": 1.9829670755929196e-05,
+      "loss": 1.1055,
+      "step": 1519
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1931116580963135,
+      "learning_rate": 1.9829329191249254e-05,
+      "loss": 1.086,
+      "step": 1520
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1118412017822266,
+      "learning_rate": 1.9828987287386624e-05,
+      "loss": 1.0795,
+      "step": 1521
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.164911985397339,
+      "learning_rate": 1.98286450443531e-05,
+      "loss": 1.1465,
+      "step": 1522
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.487156391143799,
+      "learning_rate": 1.982830246216049e-05,
+      "loss": 1.0907,
+      "step": 1523
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1853439807891846,
+      "learning_rate": 1.9827959540820615e-05,
+      "loss": 1.1219,
+      "step": 1524
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.068653106689453,
+      "learning_rate": 1.982761628034531e-05,
+      "loss": 1.1896,
+      "step": 1525
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0271356105804443,
+      "learning_rate": 1.982727268074642e-05,
+      "loss": 1.1053,
+      "step": 1526
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1470835208892822,
+      "learning_rate": 1.98269287420358e-05,
+      "loss": 1.1822,
+      "step": 1527
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8804415464401245,
+      "learning_rate": 1.982658446422532e-05,
+      "loss": 1.0792,
+      "step": 1528
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2540791034698486,
+      "learning_rate": 1.982623984732686e-05,
+      "loss": 1.0937,
+      "step": 1529
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3177947998046875,
+      "learning_rate": 1.982589489135231e-05,
+      "loss": 0.9667,
+      "step": 1530
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.124906539916992,
+      "learning_rate": 1.9825549596313576e-05,
+      "loss": 1.1543,
+      "step": 1531
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0264222621917725,
+      "learning_rate": 1.9825203962222573e-05,
+      "loss": 1.0412,
+      "step": 1532
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9088367223739624,
+      "learning_rate": 1.9824857989091228e-05,
+      "loss": 1.1721,
+      "step": 1533
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2159401178359985,
+      "learning_rate": 1.9824511676931472e-05,
+      "loss": 0.5885,
+      "step": 1534
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.108255386352539,
+      "learning_rate": 1.9824165025755267e-05,
+      "loss": 1.0751,
+      "step": 1535
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2997970581054688,
+      "learning_rate": 1.9823818035574568e-05,
+      "loss": 1.1204,
+      "step": 1536
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0071542263031006,
+      "learning_rate": 1.982347070640135e-05,
+      "loss": 1.1042,
+      "step": 1537
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3428642749786377,
+      "learning_rate": 1.98231230382476e-05,
+      "loss": 1.0368,
+      "step": 1538
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1721205711364746,
+      "learning_rate": 1.982277503112531e-05,
+      "loss": 0.593,
+      "step": 1539
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2708261013031006,
+      "learning_rate": 1.9822426685046498e-05,
+      "loss": 1.1309,
+      "step": 1540
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3629820346832275,
+      "learning_rate": 1.9822078000023176e-05,
+      "loss": 1.0052,
+      "step": 1541
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.15863037109375,
+      "learning_rate": 1.9821728976067382e-05,
+      "loss": 1.1463,
+      "step": 1542
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0399534702301025,
+      "learning_rate": 1.9821379613191154e-05,
+      "loss": 1.0613,
+      "step": 1543
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0192134380340576,
+      "learning_rate": 1.982102991140655e-05,
+      "loss": 1.1146,
+      "step": 1544
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.148620843887329,
+      "learning_rate": 1.9820679870725642e-05,
+      "loss": 1.103,
+      "step": 1545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.079495429992676,
+      "learning_rate": 1.9820329491160498e-05,
+      "loss": 1.0989,
+      "step": 1546
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1147868633270264,
+      "learning_rate": 1.981997877272322e-05,
+      "loss": 0.5885,
+      "step": 1547
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.35673451423645,
+      "learning_rate": 1.9819627715425904e-05,
+      "loss": 1.0839,
+      "step": 1548
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1471893787384033,
+      "learning_rate": 1.9819276319280666e-05,
+      "loss": 1.174,
+      "step": 1549
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.152531623840332,
+      "learning_rate": 1.9818924584299634e-05,
+      "loss": 1.1352,
+      "step": 1550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1739799976348877,
+      "learning_rate": 1.9818572510494936e-05,
+      "loss": 0.5836,
+      "step": 1551
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.113603115081787,
+      "learning_rate": 1.981822009787873e-05,
+      "loss": 1.1016,
+      "step": 1552
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0553548336029053,
+      "learning_rate": 1.981786734646318e-05,
+      "loss": 1.1185,
+      "step": 1553
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1447396278381348,
+      "learning_rate": 1.981751425626045e-05,
+      "loss": 1.1113,
+      "step": 1554
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.059035301208496,
+      "learning_rate": 1.9817160827282725e-05,
+      "loss": 1.1281,
+      "step": 1555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9455541372299194,
+      "learning_rate": 1.9816807059542204e-05,
+      "loss": 1.1825,
+      "step": 1556
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2844650745391846,
+      "learning_rate": 1.9816452953051092e-05,
+      "loss": 1.1109,
+      "step": 1557
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.350222110748291,
+      "learning_rate": 1.981609850782161e-05,
+      "loss": 1.1377,
+      "step": 1558
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0737218856811523,
+      "learning_rate": 1.9815743723865993e-05,
+      "loss": 1.0395,
+      "step": 1559
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1310346126556396,
+      "learning_rate": 1.9815388601196475e-05,
+      "loss": 1.0461,
+      "step": 1560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0816190242767334,
+      "learning_rate": 1.9815033139825315e-05,
+      "loss": 1.0525,
+      "step": 1561
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.175285816192627,
+      "learning_rate": 1.981467733976478e-05,
+      "loss": 1.1439,
+      "step": 1562
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.250605583190918,
+      "learning_rate": 1.9814321201027144e-05,
+      "loss": 1.0734,
+      "step": 1563
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2713162899017334,
+      "learning_rate": 1.98139647236247e-05,
+      "loss": 1.1042,
+      "step": 1564
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0425164699554443,
+      "learning_rate": 1.9813607907569747e-05,
+      "loss": 1.1371,
+      "step": 1565
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1377336978912354,
+      "learning_rate": 1.98132507528746e-05,
+      "loss": 1.0622,
+      "step": 1566
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2684459686279297,
+      "learning_rate": 1.9812893259551582e-05,
+      "loss": 0.5724,
+      "step": 1567
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1233832836151123,
+      "learning_rate": 1.981253542761303e-05,
+      "loss": 1.1783,
+      "step": 1568
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1962993144989014,
+      "learning_rate": 1.9812177257071284e-05,
+      "loss": 1.1578,
+      "step": 1569
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1200573444366455,
+      "learning_rate": 1.9811818747938717e-05,
+      "loss": 1.1289,
+      "step": 1570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.270577907562256,
+      "learning_rate": 1.981145990022769e-05,
+      "loss": 1.2382,
+      "step": 1571
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0432496070861816,
+      "learning_rate": 1.9811100713950587e-05,
+      "loss": 0.9483,
+      "step": 1572
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.045334815979004,
+      "learning_rate": 1.981074118911981e-05,
+      "loss": 1.0953,
+      "step": 1573
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1907827854156494,
+      "learning_rate": 1.9810381325747757e-05,
+      "loss": 1.1282,
+      "step": 1574
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.030099630355835,
+      "learning_rate": 1.9810021123846845e-05,
+      "loss": 1.0778,
+      "step": 1575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.286919355392456,
+      "learning_rate": 1.980966058342951e-05,
+      "loss": 1.112,
+      "step": 1576
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.13248348236084,
+      "learning_rate": 1.9809299704508193e-05,
+      "loss": 1.0865,
+      "step": 1577
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1283349990844727,
+      "learning_rate": 1.9808938487095343e-05,
+      "loss": 1.066,
+      "step": 1578
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9209989309310913,
+      "learning_rate": 1.9808576931203424e-05,
+      "loss": 1.0962,
+      "step": 1579
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2582900524139404,
+      "learning_rate": 1.980821503684492e-05,
+      "loss": 1.0814,
+      "step": 1580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1568522453308105,
+      "learning_rate": 1.9807852804032306e-05,
+      "loss": 0.5813,
+      "step": 1581
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.098873496055603,
+      "learning_rate": 1.980749023277809e-05,
+      "loss": 0.5676,
+      "step": 1582
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1907334327697754,
+      "learning_rate": 1.9807127323094784e-05,
+      "loss": 1.0623,
+      "step": 1583
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0981929302215576,
+      "learning_rate": 1.980676407499491e-05,
+      "loss": 1.1046,
+      "step": 1584
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.335172414779663,
+      "learning_rate": 1.9806400488491003e-05,
+      "loss": 1.1292,
+      "step": 1585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1574161052703857,
+      "learning_rate": 1.9806036563595606e-05,
+      "loss": 0.5256,
+      "step": 1586
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1440577507019043,
+      "learning_rate": 1.980567230032128e-05,
+      "loss": 1.1451,
+      "step": 1587
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.04799485206604,
+      "learning_rate": 1.9805307698680592e-05,
+      "loss": 1.0126,
+      "step": 1588
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0676918029785156,
+      "learning_rate": 1.980494275868613e-05,
+      "loss": 1.1077,
+      "step": 1589
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2290658950805664,
+      "learning_rate": 1.9804577480350477e-05,
+      "loss": 1.13,
+      "step": 1590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1077401638031006,
+      "learning_rate": 1.9804211863686244e-05,
+      "loss": 1.0092,
+      "step": 1591
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.29612398147583,
+      "learning_rate": 1.980384590870605e-05,
+      "loss": 1.1052,
+      "step": 1592
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0813913345336914,
+      "learning_rate": 1.9803479615422515e-05,
+      "loss": 1.1089,
+      "step": 1593
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0125625133514404,
+      "learning_rate": 1.9803112983848287e-05,
+      "loss": 1.0348,
+      "step": 1594
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9546971321105957,
+      "learning_rate": 1.9802746013996012e-05,
+      "loss": 1.114,
+      "step": 1595
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0356884002685547,
+      "learning_rate": 1.9802378705878354e-05,
+      "loss": 1.078,
+      "step": 1596
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4589121341705322,
+      "learning_rate": 1.9802011059507993e-05,
+      "loss": 1.1889,
+      "step": 1597
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1145145893096924,
+      "learning_rate": 1.980164307489761e-05,
+      "loss": 1.0445,
+      "step": 1598
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9814850091934204,
+      "learning_rate": 1.98012747520599e-05,
+      "loss": 1.0659,
+      "step": 1599
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.982138752937317,
+      "learning_rate": 1.980090609100758e-05,
+      "loss": 1.0136,
+      "step": 1600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0517921447753906,
+      "learning_rate": 1.980053709175337e-05,
+      "loss": 1.064,
+      "step": 1601
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.168419599533081,
+      "learning_rate": 1.980016775431e-05,
+      "loss": 1.0297,
+      "step": 1602
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1289284229278564,
+      "learning_rate": 1.9799798078690216e-05,
+      "loss": 1.0203,
+      "step": 1603
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.037092447280884,
+      "learning_rate": 1.9799428064906775e-05,
+      "loss": 1.134,
+      "step": 1604
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0955047607421875,
+      "learning_rate": 1.9799057712972444e-05,
+      "loss": 1.1261,
+      "step": 1605
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.04819917678833,
+      "learning_rate": 1.979868702290001e-05,
+      "loss": 0.983,
+      "step": 1606
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.179774045944214,
+      "learning_rate": 1.979831599470225e-05,
+      "loss": 1.0752,
+      "step": 1607
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.177022933959961,
+      "learning_rate": 1.9797944628391978e-05,
+      "loss": 1.0054,
+      "step": 1608
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8800945281982422,
+      "learning_rate": 1.979757292398201e-05,
+      "loss": 1.1204,
+      "step": 1609
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.152066707611084,
+      "learning_rate": 1.9797200881485166e-05,
+      "loss": 1.0746,
+      "step": 1610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0069997310638428,
+      "learning_rate": 1.9796828500914285e-05,
+      "loss": 1.1132,
+      "step": 1611
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3057225942611694,
+      "learning_rate": 1.979645578228222e-05,
+      "loss": 0.517,
+      "step": 1612
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9819554090499878,
+      "learning_rate": 1.9796082725601836e-05,
+      "loss": 1.1542,
+      "step": 1613
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0664961338043213,
+      "learning_rate": 1.9795709330885996e-05,
+      "loss": 1.0918,
+      "step": 1614
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9755898714065552,
+      "learning_rate": 1.9795335598147592e-05,
+      "loss": 1.1102,
+      "step": 1615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.813048005104065,
+      "learning_rate": 1.9794961527399518e-05,
+      "loss": 1.0469,
+      "step": 1616
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.861116886138916,
+      "learning_rate": 1.9794587118654686e-05,
+      "loss": 1.0873,
+      "step": 1617
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9587714672088623,
+      "learning_rate": 1.9794212371926008e-05,
+      "loss": 1.0606,
+      "step": 1618
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.183161735534668,
+      "learning_rate": 1.9793837287226424e-05,
+      "loss": 0.6656,
+      "step": 1619
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1780409812927246,
+      "learning_rate": 1.979346186456887e-05,
+      "loss": 1.1627,
+      "step": 1620
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.119764804840088,
+      "learning_rate": 1.9793086103966305e-05,
+      "loss": 1.1192,
+      "step": 1621
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0270514488220215,
+      "learning_rate": 1.9792710005431695e-05,
+      "loss": 1.0353,
+      "step": 1622
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0057146549224854,
+      "learning_rate": 1.9792333568978018e-05,
+      "loss": 1.1613,
+      "step": 1623
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.018458127975464,
+      "learning_rate": 1.9791956794618263e-05,
+      "loss": 1.0525,
+      "step": 1624
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.075502872467041,
+      "learning_rate": 1.979157968236543e-05,
+      "loss": 1.065,
+      "step": 1625
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9254873991012573,
+      "learning_rate": 1.9791202232232534e-05,
+      "loss": 1.0411,
+      "step": 1626
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2357261180877686,
+      "learning_rate": 1.97908244442326e-05,
+      "loss": 1.0471,
+      "step": 1627
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9962939023971558,
+      "learning_rate": 1.9790446318378667e-05,
+      "loss": 1.1121,
+      "step": 1628
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.823634147644043,
+      "learning_rate": 1.9790067854683778e-05,
+      "loss": 1.1161,
+      "step": 1629
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9571431875228882,
+      "learning_rate": 1.9789689053160995e-05,
+      "loss": 1.0946,
+      "step": 1630
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.890627145767212,
+      "learning_rate": 1.9789309913823387e-05,
+      "loss": 1.0639,
+      "step": 1631
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0019779205322266,
+      "learning_rate": 1.9788930436684043e-05,
+      "loss": 1.1484,
+      "step": 1632
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.835968255996704,
+      "learning_rate": 1.978855062175605e-05,
+      "loss": 1.0711,
+      "step": 1633
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.005263566970825,
+      "learning_rate": 1.978817046905252e-05,
+      "loss": 1.1072,
+      "step": 1634
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9493229389190674,
+      "learning_rate": 1.978778997858657e-05,
+      "loss": 1.1526,
+      "step": 1635
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0082051753997803,
+      "learning_rate": 1.9787409150371327e-05,
+      "loss": 1.0992,
+      "step": 1636
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9586727619171143,
+      "learning_rate": 1.978702798441994e-05,
+      "loss": 1.1867,
+      "step": 1637
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8139171600341797,
+      "learning_rate": 1.9786646480745547e-05,
+      "loss": 1.0407,
+      "step": 1638
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8802461624145508,
+      "learning_rate": 1.978626463936133e-05,
+      "loss": 1.0505,
+      "step": 1639
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0512077808380127,
+      "learning_rate": 1.9785882460280452e-05,
+      "loss": 1.1202,
+      "step": 1640
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8554996252059937,
+      "learning_rate": 1.9785499943516108e-05,
+      "loss": 1.1558,
+      "step": 1641
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.029660701751709,
+      "learning_rate": 1.9785117089081497e-05,
+      "loss": 1.1417,
+      "step": 1642
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0053927898406982,
+      "learning_rate": 1.9784733896989826e-05,
+      "loss": 1.0932,
+      "step": 1643
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0511276721954346,
+      "learning_rate": 1.9784350367254322e-05,
+      "loss": 1.1392,
+      "step": 1644
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8730463981628418,
+      "learning_rate": 1.978396649988822e-05,
+      "loss": 1.0124,
+      "step": 1645
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9827896356582642,
+      "learning_rate": 1.9783582294904762e-05,
+      "loss": 1.0992,
+      "step": 1646
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.067535400390625,
+      "learning_rate": 1.9783197752317207e-05,
+      "loss": 1.0703,
+      "step": 1647
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.218522071838379,
+      "learning_rate": 1.978281287213883e-05,
+      "loss": 1.0789,
+      "step": 1648
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1139981746673584,
+      "learning_rate": 1.9782427654382906e-05,
+      "loss": 1.0125,
+      "step": 1649
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.27189040184021,
+      "learning_rate": 1.978204209906273e-05,
+      "loss": 1.1011,
+      "step": 1650
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0736489295959473,
+      "learning_rate": 1.9781656206191604e-05,
+      "loss": 1.0902,
+      "step": 1651
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0755748748779297,
+      "learning_rate": 1.9781269975782848e-05,
+      "loss": 1.0653,
+      "step": 1652
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.988792061805725,
+      "learning_rate": 1.978088340784979e-05,
+      "loss": 1.0032,
+      "step": 1653
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.036255121231079,
+      "learning_rate": 1.9780496502405765e-05,
+      "loss": 1.1107,
+      "step": 1654
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.14223313331604,
+      "learning_rate": 1.9780109259464127e-05,
+      "loss": 1.073,
+      "step": 1655
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1095829010009766,
+      "learning_rate": 1.9779721679038237e-05,
+      "loss": 1.0566,
+      "step": 1656
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1173906326293945,
+      "learning_rate": 1.977933376114147e-05,
+      "loss": 0.9956,
+      "step": 1657
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.165433168411255,
+      "learning_rate": 1.9778945505787216e-05,
+      "loss": 1.1912,
+      "step": 1658
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1865346431732178,
+      "learning_rate": 1.9778556912988865e-05,
+      "loss": 1.1065,
+      "step": 1659
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.797239065170288,
+      "learning_rate": 1.9778167982759836e-05,
+      "loss": 1.0452,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.170520782470703,
+      "learning_rate": 1.977777871511354e-05,
+      "loss": 1.1055,
+      "step": 1661
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5014305114746094,
+      "learning_rate": 1.977738911006341e-05,
+      "loss": 1.2075,
+      "step": 1662
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1980966329574585,
+      "learning_rate": 1.9776999167622902e-05,
+      "loss": 0.6351,
+      "step": 1663
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0420920848846436,
+      "learning_rate": 1.9776608887805456e-05,
+      "loss": 1.0724,
+      "step": 1664
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9526374340057373,
+      "learning_rate": 1.977621827062455e-05,
+      "loss": 1.0672,
+      "step": 1665
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0691959857940674,
+      "learning_rate": 1.9775827316093663e-05,
+      "loss": 1.032,
+      "step": 1666
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.259793519973755,
+      "learning_rate": 1.9775436024226283e-05,
+      "loss": 1.2199,
+      "step": 1667
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9516514539718628,
+      "learning_rate": 1.977504439503591e-05,
+      "loss": 1.0821,
+      "step": 1668
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0254062414169312,
+      "learning_rate": 1.977465242853606e-05,
+      "loss": 0.5414,
+      "step": 1669
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.119377851486206,
+      "learning_rate": 1.9774260124740257e-05,
+      "loss": 1.1291,
+      "step": 1670
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2170608043670654,
+      "learning_rate": 1.9773867483662044e-05,
+      "loss": 1.132,
+      "step": 1671
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.119483470916748,
+      "learning_rate": 1.9773474505314966e-05,
+      "loss": 1.073,
+      "step": 1672
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8645308017730713,
+      "learning_rate": 1.977308118971258e-05,
+      "loss": 1.1062,
+      "step": 1673
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0861430168151855,
+      "learning_rate": 1.9772687536868468e-05,
+      "loss": 1.143,
+      "step": 1674
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1440775394439697,
+      "learning_rate": 1.9772293546796205e-05,
+      "loss": 1.106,
+      "step": 1675
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.320091724395752,
+      "learning_rate": 1.9771899219509388e-05,
+      "loss": 1.115,
+      "step": 1676
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8981430530548096,
+      "learning_rate": 1.977150455502163e-05,
+      "loss": 1.085,
+      "step": 1677
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4733781814575195,
+      "learning_rate": 1.9771109553346542e-05,
+      "loss": 1.1499,
+      "step": 1678
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.916886806488037,
+      "learning_rate": 1.977071421449776e-05,
+      "loss": 1.1342,
+      "step": 1679
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9651583433151245,
+      "learning_rate": 1.9770318538488923e-05,
+      "loss": 1.1639,
+      "step": 1680
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0920627117156982,
+      "learning_rate": 1.9769922525333688e-05,
+      "loss": 1.129,
+      "step": 1681
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9595839977264404,
+      "learning_rate": 1.9769526175045713e-05,
+      "loss": 1.0276,
+      "step": 1682
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2484540939331055,
+      "learning_rate": 1.976912948763868e-05,
+      "loss": 0.5912,
+      "step": 1683
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2011122703552246,
+      "learning_rate": 1.9768732463126282e-05,
+      "loss": 1.0826,
+      "step": 1684
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9850212335586548,
+      "learning_rate": 1.9768335101522212e-05,
+      "loss": 1.0202,
+      "step": 1685
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0429179668426514,
+      "learning_rate": 1.976793740284018e-05,
+      "loss": 1.1044,
+      "step": 1686
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9772089719772339,
+      "learning_rate": 1.976753936709392e-05,
+      "loss": 1.029,
+      "step": 1687
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9813590049743652,
+      "learning_rate": 1.976714099429716e-05,
+      "loss": 1.1413,
+      "step": 1688
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9641249179840088,
+      "learning_rate": 1.9766742284463645e-05,
+      "loss": 1.1728,
+      "step": 1689
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9910151958465576,
+      "learning_rate": 1.9766343237607136e-05,
+      "loss": 1.0805,
+      "step": 1690
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.102437973022461,
+      "learning_rate": 1.976594385374141e-05,
+      "loss": 0.9996,
+      "step": 1691
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8996132612228394,
+      "learning_rate": 1.976554413288023e-05,
+      "loss": 1.1539,
+      "step": 1692
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.278383255004883,
+      "learning_rate": 1.976514407503741e-05,
+      "loss": 1.0643,
+      "step": 1693
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0727622509002686,
+      "learning_rate": 1.9764743680226744e-05,
+      "loss": 1.0664,
+      "step": 1694
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.003706932067871,
+      "learning_rate": 1.9764342948462047e-05,
+      "loss": 1.078,
+      "step": 1695
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0936602354049683,
+      "learning_rate": 1.9763941879757155e-05,
+      "loss": 0.5485,
+      "step": 1696
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1821393966674805,
+      "learning_rate": 1.97635404741259e-05,
+      "loss": 1.077,
+      "step": 1697
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9630897045135498,
+      "learning_rate": 1.9763138731582138e-05,
+      "loss": 1.1377,
+      "step": 1698
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9337328672409058,
+      "learning_rate": 1.9762736652139727e-05,
+      "loss": 1.0559,
+      "step": 1699
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.851905107498169,
+      "learning_rate": 1.976233423581255e-05,
+      "loss": 1.0427,
+      "step": 1700
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8319166898727417,
+      "learning_rate": 1.976193148261449e-05,
+      "loss": 1.0862,
+      "step": 1701
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0709736347198486,
+      "learning_rate": 1.976152839255944e-05,
+      "loss": 0.9576,
+      "step": 1702
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1122336387634277,
+      "learning_rate": 1.9761124965661313e-05,
+      "loss": 1.1438,
+      "step": 1703
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.115537405014038,
+      "learning_rate": 1.9760721201934026e-05,
+      "loss": 1.0307,
+      "step": 1704
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.004452705383301,
+      "learning_rate": 1.9760317101391525e-05,
+      "loss": 1.0156,
+      "step": 1705
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9772614240646362,
+      "learning_rate": 1.975991266404774e-05,
+      "loss": 1.0899,
+      "step": 1706
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0291634798049927,
+      "learning_rate": 1.975950788991663e-05,
+      "loss": 0.5634,
+      "step": 1707
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1664555072784424,
+      "learning_rate": 1.9759102779012167e-05,
+      "loss": 1.1282,
+      "step": 1708
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.046359062194824,
+      "learning_rate": 1.9758697331348328e-05,
+      "loss": 1.0731,
+      "step": 1709
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0227787494659424,
+      "learning_rate": 1.9758291546939107e-05,
+      "loss": 1.0674,
+      "step": 1710
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.312598705291748,
+      "learning_rate": 1.9757885425798497e-05,
+      "loss": 1.1017,
+      "step": 1711
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.135272741317749,
+      "learning_rate": 1.9757478967940526e-05,
+      "loss": 1.1085,
+      "step": 1712
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.454162120819092,
+      "learning_rate": 1.9757072173379206e-05,
+      "loss": 1.1246,
+      "step": 1713
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0647523403167725,
+      "learning_rate": 1.975666504212858e-05,
+      "loss": 1.0551,
+      "step": 1714
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2257614135742188,
+      "learning_rate": 1.9756257574202705e-05,
+      "loss": 1.0485,
+      "step": 1715
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2690961360931396,
+      "learning_rate": 1.975584976961563e-05,
+      "loss": 1.0749,
+      "step": 1716
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9459435939788818,
+      "learning_rate": 1.975544162838143e-05,
+      "loss": 1.1076,
+      "step": 1717
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9962010383605957,
+      "learning_rate": 1.975503315051419e-05,
+      "loss": 0.5681,
+      "step": 1718
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.205332040786743,
+      "learning_rate": 1.9754624336028007e-05,
+      "loss": 1.1968,
+      "step": 1719
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0800867080688477,
+      "learning_rate": 1.975421518493699e-05,
+      "loss": 1.0004,
+      "step": 1720
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0179686546325684,
+      "learning_rate": 1.9753805697255246e-05,
+      "loss": 1.0951,
+      "step": 1721
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1346912384033203,
+      "learning_rate": 1.975339587299692e-05,
+      "loss": 1.1005,
+      "step": 1722
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0877277851104736,
+      "learning_rate": 1.975298571217614e-05,
+      "loss": 1.0645,
+      "step": 1723
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.123556137084961,
+      "learning_rate": 1.9752575214807077e-05,
+      "loss": 1.0744,
+      "step": 1724
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0165324211120605,
+      "learning_rate": 1.975216438090388e-05,
+      "loss": 1.0771,
+      "step": 1725
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.134835958480835,
+      "learning_rate": 1.9751753210480733e-05,
+      "loss": 1.0927,
+      "step": 1726
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2924907207489014,
+      "learning_rate": 1.9751341703551824e-05,
+      "loss": 1.2181,
+      "step": 1727
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1274254322052,
+      "learning_rate": 1.9750929860131353e-05,
+      "loss": 1.1463,
+      "step": 1728
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1014857292175293,
+      "learning_rate": 1.975051768023353e-05,
+      "loss": 1.0484,
+      "step": 1729
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0884482860565186,
+      "learning_rate": 1.9750105163872577e-05,
+      "loss": 1.1372,
+      "step": 1730
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2393875122070312,
+      "learning_rate": 1.9749692311062733e-05,
+      "loss": 1.1367,
+      "step": 1731
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.917853832244873,
+      "learning_rate": 1.9749279121818235e-05,
+      "loss": 1.1067,
+      "step": 1732
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.194563388824463,
+      "learning_rate": 1.9748865596153356e-05,
+      "loss": 1.1273,
+      "step": 1733
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.126788854598999,
+      "learning_rate": 1.9748451734082356e-05,
+      "loss": 1.091,
+      "step": 1734
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.010206699371338,
+      "learning_rate": 1.9748037535619518e-05,
+      "loss": 1.1165,
+      "step": 1735
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.023131847381592,
+      "learning_rate": 1.974762300077913e-05,
+      "loss": 1.1873,
+      "step": 1736
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9872779846191406,
+      "learning_rate": 1.9747208129575507e-05,
+      "loss": 1.0712,
+      "step": 1737
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1089580059051514,
+      "learning_rate": 1.9746792922022956e-05,
+      "loss": 0.6061,
+      "step": 1738
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8352059125900269,
+      "learning_rate": 1.974637737813581e-05,
+      "loss": 1.0785,
+      "step": 1739
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9942800998687744,
+      "learning_rate": 1.9745961497928406e-05,
+      "loss": 1.1035,
+      "step": 1740
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0443503856658936,
+      "learning_rate": 1.974554528141509e-05,
+      "loss": 1.1267,
+      "step": 1741
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3537492752075195,
+      "learning_rate": 1.9745128728610235e-05,
+      "loss": 1.094,
+      "step": 1742
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9897205829620361,
+      "learning_rate": 1.974471183952821e-05,
+      "loss": 1.0581,
+      "step": 1743
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0357699394226074,
+      "learning_rate": 1.9744294614183397e-05,
+      "loss": 1.1107,
+      "step": 1744
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2005937099456787,
+      "learning_rate": 1.97438770525902e-05,
+      "loss": 1.0821,
+      "step": 1745
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.001901865005493,
+      "learning_rate": 1.974345915476302e-05,
+      "loss": 1.0564,
+      "step": 1746
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.468035936355591,
+      "learning_rate": 1.9743040920716282e-05,
+      "loss": 1.1817,
+      "step": 1747
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9884804487228394,
+      "learning_rate": 1.974262235046442e-05,
+      "loss": 1.1304,
+      "step": 1748
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.144960641860962,
+      "learning_rate": 1.9742203444021878e-05,
+      "loss": 1.0166,
+      "step": 1749
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.011486291885376,
+      "learning_rate": 1.9741784201403104e-05,
+      "loss": 1.0824,
+      "step": 1750
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0152251720428467,
+      "learning_rate": 1.974136462262257e-05,
+      "loss": 1.0871,
+      "step": 1751
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9161790609359741,
+      "learning_rate": 1.9740944707694757e-05,
+      "loss": 1.0788,
+      "step": 1752
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0590739250183105,
+      "learning_rate": 1.974052445663415e-05,
+      "loss": 1.0372,
+      "step": 1753
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0413475036621094,
+      "learning_rate": 1.9740103869455257e-05,
+      "loss": 1.0788,
+      "step": 1754
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.130965232849121,
+      "learning_rate": 1.9739682946172584e-05,
+      "loss": 1.1716,
+      "step": 1755
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.155247449874878,
+      "learning_rate": 1.9739261686800662e-05,
+      "loss": 1.1196,
+      "step": 1756
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9767510890960693,
+      "learning_rate": 1.973884009135402e-05,
+      "loss": 1.1925,
+      "step": 1757
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2008259296417236,
+      "learning_rate": 1.9738418159847216e-05,
+      "loss": 0.597,
+      "step": 1758
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.095219612121582,
+      "learning_rate": 1.9737995892294803e-05,
+      "loss": 1.0242,
+      "step": 1759
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1930441856384277,
+      "learning_rate": 1.9737573288711348e-05,
+      "loss": 1.1751,
+      "step": 1760
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9918029308319092,
+      "learning_rate": 1.9737150349111447e-05,
+      "loss": 1.1299,
+      "step": 1761
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.037109851837158,
+      "learning_rate": 1.9736727073509684e-05,
+      "loss": 1.0889,
+      "step": 1762
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.012993097305298,
+      "learning_rate": 1.9736303461920667e-05,
+      "loss": 1.0657,
+      "step": 1763
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.165102958679199,
+      "learning_rate": 1.9735879514359017e-05,
+      "loss": 1.0382,
+      "step": 1764
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.009129762649536,
+      "learning_rate": 1.9735455230839363e-05,
+      "loss": 1.0391,
+      "step": 1765
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9560577869415283,
+      "learning_rate": 1.973503061137634e-05,
+      "loss": 1.0763,
+      "step": 1766
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7726469039916992,
+      "learning_rate": 1.9734605655984604e-05,
+      "loss": 0.9933,
+      "step": 1767
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.263556718826294,
+      "learning_rate": 1.9734180364678824e-05,
+      "loss": 1.144,
+      "step": 1768
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1548538208007812,
+      "learning_rate": 1.973375473747367e-05,
+      "loss": 1.1339,
+      "step": 1769
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8769031763076782,
+      "learning_rate": 1.9733328774383825e-05,
+      "loss": 1.174,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.128791570663452,
+      "learning_rate": 1.9732902475423995e-05,
+      "loss": 0.9804,
+      "step": 1771
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.067493438720703,
+      "learning_rate": 1.973247584060889e-05,
+      "loss": 1.1482,
+      "step": 1772
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8724035024642944,
+      "learning_rate": 1.973204886995323e-05,
+      "loss": 1.0242,
+      "step": 1773
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9555832147598267,
+      "learning_rate": 1.9731621563471748e-05,
+      "loss": 1.0877,
+      "step": 1774
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.945119857788086,
+      "learning_rate": 1.9731193921179192e-05,
+      "loss": 1.181,
+      "step": 1775
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8445144891738892,
+      "learning_rate": 1.9730765943090314e-05,
+      "loss": 1.081,
+      "step": 1776
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.080310106277466,
+      "learning_rate": 1.9730337629219886e-05,
+      "loss": 1.1375,
+      "step": 1777
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0283448696136475,
+      "learning_rate": 1.972990897958269e-05,
+      "loss": 1.0731,
+      "step": 1778
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3141883611679077,
+      "learning_rate": 1.972947999419351e-05,
+      "loss": 0.6173,
+      "step": 1779
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1313295364379883,
+      "learning_rate": 1.9729050673067156e-05,
+      "loss": 1.0467,
+      "step": 1780
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1361844539642334,
+      "learning_rate": 1.972862101621844e-05,
+      "loss": 1.1688,
+      "step": 1781
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.201749086380005,
+      "learning_rate": 1.9728191023662188e-05,
+      "loss": 1.1626,
+      "step": 1782
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8586986064910889,
+      "learning_rate": 1.972776069541324e-05,
+      "loss": 1.1071,
+      "step": 1783
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.819804072380066,
+      "learning_rate": 1.9727330031486443e-05,
+      "loss": 1.1585,
+      "step": 1784
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1963603496551514,
+      "learning_rate": 1.972689903189666e-05,
+      "loss": 1.0215,
+      "step": 1785
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.010779857635498,
+      "learning_rate": 1.972646769665876e-05,
+      "loss": 1.0796,
+      "step": 1786
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9218958616256714,
+      "learning_rate": 1.972603602578763e-05,
+      "loss": 1.0612,
+      "step": 1787
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.994632601737976,
+      "learning_rate": 1.9725604019298162e-05,
+      "loss": 1.0832,
+      "step": 1788
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.118058919906616,
+      "learning_rate": 1.9725171677205273e-05,
+      "loss": 1.0512,
+      "step": 1789
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8827786445617676,
+      "learning_rate": 1.9724738999523874e-05,
+      "loss": 1.0859,
+      "step": 1790
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1775686740875244,
+      "learning_rate": 1.9724305986268898e-05,
+      "loss": 1.1353,
+      "step": 1791
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.099546432495117,
+      "learning_rate": 1.972387263745528e-05,
+      "loss": 1.0989,
+      "step": 1792
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9460331201553345,
+      "learning_rate": 1.9723438953097985e-05,
+      "loss": 1.0469,
+      "step": 1793
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8811291456222534,
+      "learning_rate": 1.9723004933211975e-05,
+      "loss": 1.0835,
+      "step": 1794
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9230514764785767,
+      "learning_rate": 1.972257057781222e-05,
+      "loss": 1.1082,
+      "step": 1795
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9974192380905151,
+      "learning_rate": 1.9722135886913716e-05,
+      "loss": 1.1541,
+      "step": 1796
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9699289798736572,
+      "learning_rate": 1.972170086053146e-05,
+      "loss": 1.0339,
+      "step": 1797
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9669231176376343,
+      "learning_rate": 1.972126549868046e-05,
+      "loss": 1.1009,
+      "step": 1798
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0928196907043457,
+      "learning_rate": 1.972082980137575e-05,
+      "loss": 1.0161,
+      "step": 1799
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.152636766433716,
+      "learning_rate": 1.972039376863235e-05,
+      "loss": 1.0703,
+      "step": 1800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.007467746734619,
+      "learning_rate": 1.971995740046532e-05,
+      "loss": 1.063,
+      "step": 1801
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.372265338897705,
+      "learning_rate": 1.971952069688971e-05,
+      "loss": 1.1111,
+      "step": 1802
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9397143125534058,
+      "learning_rate": 1.971908365792059e-05,
+      "loss": 1.081,
+      "step": 1803
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1565940380096436,
+      "learning_rate": 1.971864628357304e-05,
+      "loss": 1.1147,
+      "step": 1804
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.154966115951538,
+      "learning_rate": 1.971820857386216e-05,
+      "loss": 1.1186,
+      "step": 1805
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2199151515960693,
+      "learning_rate": 1.9717770528803046e-05,
+      "loss": 1.0974,
+      "step": 1806
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0128839015960693,
+      "learning_rate": 1.9717332148410817e-05,
+      "loss": 1.0827,
+      "step": 1807
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1589548587799072,
+      "learning_rate": 1.97168934327006e-05,
+      "loss": 1.0958,
+      "step": 1808
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9092520475387573,
+      "learning_rate": 1.9716454381687535e-05,
+      "loss": 1.0037,
+      "step": 1809
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.080681324005127,
+      "learning_rate": 1.9716014995386767e-05,
+      "loss": 1.0566,
+      "step": 1810
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.865631341934204,
+      "learning_rate": 1.9715575273813466e-05,
+      "loss": 0.9919,
+      "step": 1811
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0890274047851562,
+      "learning_rate": 1.97151352169828e-05,
+      "loss": 1.1962,
+      "step": 1812
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9518492221832275,
+      "learning_rate": 1.9714694824909954e-05,
+      "loss": 1.0589,
+      "step": 1813
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9185514450073242,
+      "learning_rate": 1.9714254097610128e-05,
+      "loss": 1.076,
+      "step": 1814
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9747978448867798,
+      "learning_rate": 1.971381303509853e-05,
+      "loss": 1.0736,
+      "step": 1815
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0762267112731934,
+      "learning_rate": 1.9713371637390376e-05,
+      "loss": 1.1286,
+      "step": 1816
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.014113664627075,
+      "learning_rate": 1.9712929904500905e-05,
+      "loss": 1.0808,
+      "step": 1817
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9392385482788086,
+      "learning_rate": 1.971248783644535e-05,
+      "loss": 1.0036,
+      "step": 1818
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.078747272491455,
+      "learning_rate": 1.9712045433238972e-05,
+      "loss": 1.1318,
+      "step": 1819
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0274078845977783,
+      "learning_rate": 1.971160269489704e-05,
+      "loss": 1.0583,
+      "step": 1820
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.037526845932007,
+      "learning_rate": 1.9711159621434822e-05,
+      "loss": 1.0786,
+      "step": 1821
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1042792797088623,
+      "learning_rate": 1.971071621286761e-05,
+      "loss": 1.1058,
+      "step": 1822
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1857566833496094,
+      "learning_rate": 1.9710272469210713e-05,
+      "loss": 1.0701,
+      "step": 1823
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9370806217193604,
+      "learning_rate": 1.9709828390479436e-05,
+      "loss": 1.0763,
+      "step": 1824
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6989450454711914,
+      "learning_rate": 1.9709383976689102e-05,
+      "loss": 1.0247,
+      "step": 1825
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.118069648742676,
+      "learning_rate": 1.970893922785505e-05,
+      "loss": 1.2092,
+      "step": 1826
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1400671005249023,
+      "learning_rate": 1.970849414399263e-05,
+      "loss": 1.0178,
+      "step": 1827
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1096243858337402,
+      "learning_rate": 1.9708048725117194e-05,
+      "loss": 1.095,
+      "step": 1828
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9998549222946167,
+      "learning_rate": 1.9707602971244115e-05,
+      "loss": 1.0742,
+      "step": 1829
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9400320053100586,
+      "learning_rate": 1.9707156882388773e-05,
+      "loss": 0.9742,
+      "step": 1830
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1331746578216553,
+      "learning_rate": 1.9706710458566564e-05,
+      "loss": 1.0566,
+      "step": 1831
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9247522354125977,
+      "learning_rate": 1.9706263699792895e-05,
+      "loss": 1.0662,
+      "step": 1832
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9742368459701538,
+      "learning_rate": 1.970581660608317e-05,
+      "loss": 1.0608,
+      "step": 1833
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.023930072784424,
+      "learning_rate": 1.9705369177452835e-05,
+      "loss": 0.9979,
+      "step": 1834
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9788657426834106,
+      "learning_rate": 1.970492141391732e-05,
+      "loss": 1.0738,
+      "step": 1835
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9974464178085327,
+      "learning_rate": 1.9704473315492072e-05,
+      "loss": 1.059,
+      "step": 1836
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.179856777191162,
+      "learning_rate": 1.9704024882192562e-05,
+      "loss": 1.1156,
+      "step": 1837
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.971026062965393,
+      "learning_rate": 1.9703576114034257e-05,
+      "loss": 1.1142,
+      "step": 1838
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9058090448379517,
+      "learning_rate": 1.9703127011032646e-05,
+      "loss": 1.0117,
+      "step": 1839
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9456253051757812,
+      "learning_rate": 1.9702677573203232e-05,
+      "loss": 1.0769,
+      "step": 1840
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.038161039352417,
+      "learning_rate": 1.9702227800561514e-05,
+      "loss": 1.1014,
+      "step": 1841
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.078125476837158,
+      "learning_rate": 1.9701777693123017e-05,
+      "loss": 1.141,
+      "step": 1842
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.029832124710083,
+      "learning_rate": 1.9701327250903273e-05,
+      "loss": 1.1286,
+      "step": 1843
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8729559183120728,
+      "learning_rate": 1.9700876473917825e-05,
+      "loss": 1.1493,
+      "step": 1844
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.325866222381592,
+      "learning_rate": 1.970042536218223e-05,
+      "loss": 1.1554,
+      "step": 1845
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.217384099960327,
+      "learning_rate": 1.9699973915712046e-05,
+      "loss": 1.1087,
+      "step": 1846
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1590614318847656,
+      "learning_rate": 1.9699522134522866e-05,
+      "loss": 1.0392,
+      "step": 1847
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8054181337356567,
+      "learning_rate": 1.969907001863027e-05,
+      "loss": 1.0825,
+      "step": 1848
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.164466619491577,
+      "learning_rate": 1.9698617568049857e-05,
+      "loss": 1.1268,
+      "step": 1849
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7488116025924683,
+      "learning_rate": 1.9698164782797247e-05,
+      "loss": 1.0256,
+      "step": 1850
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.063260078430176,
+      "learning_rate": 1.9697711662888062e-05,
+      "loss": 1.128,
+      "step": 1851
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.043689727783203,
+      "learning_rate": 1.9697258208337935e-05,
+      "loss": 1.1244,
+      "step": 1852
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1509673595428467,
+      "learning_rate": 1.9696804419162513e-05,
+      "loss": 0.5746,
+      "step": 1853
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0850212574005127,
+      "learning_rate": 1.969635029537746e-05,
+      "loss": 0.6038,
+      "step": 1854
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0359199047088623,
+      "learning_rate": 1.9695895836998448e-05,
+      "loss": 1.1049,
+      "step": 1855
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0931057929992676,
+      "learning_rate": 1.969544104404115e-05,
+      "loss": 1.0613,
+      "step": 1856
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.351095676422119,
+      "learning_rate": 1.9694985916521266e-05,
+      "loss": 1.1317,
+      "step": 1857
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.973114252090454,
+      "learning_rate": 1.96945304544545e-05,
+      "loss": 1.1384,
+      "step": 1858
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8562629222869873,
+      "learning_rate": 1.969407465785657e-05,
+      "loss": 1.0101,
+      "step": 1859
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.224827766418457,
+      "learning_rate": 1.96936185267432e-05,
+      "loss": 1.1184,
+      "step": 1860
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1688177585601807,
+      "learning_rate": 1.969316206113013e-05,
+      "loss": 1.0626,
+      "step": 1861
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.143812417984009,
+      "learning_rate": 1.969270526103312e-05,
+      "loss": 1.0885,
+      "step": 1862
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8832557201385498,
+      "learning_rate": 1.9692248126467923e-05,
+      "loss": 1.0971,
+      "step": 1863
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.322359561920166,
+      "learning_rate": 1.9691790657450324e-05,
+      "loss": 1.1657,
+      "step": 1864
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.358800172805786,
+      "learning_rate": 1.9691332853996093e-05,
+      "loss": 1.0904,
+      "step": 1865
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9578733444213867,
+      "learning_rate": 1.969087471612104e-05,
+      "loss": 1.0162,
+      "step": 1866
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.180544137954712,
+      "learning_rate": 1.9690416243840976e-05,
+      "loss": 1.0867,
+      "step": 1867
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.208935022354126,
+      "learning_rate": 1.968995743717171e-05,
+      "loss": 1.1689,
+      "step": 1868
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.13218355178833,
+      "learning_rate": 1.9689498296129084e-05,
+      "loss": 1.0683,
+      "step": 1869
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3065497875213623,
+      "learning_rate": 1.968903882072894e-05,
+      "loss": 1.1383,
+      "step": 1870
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1199758052825928,
+      "learning_rate": 1.968857901098713e-05,
+      "loss": 1.0516,
+      "step": 1871
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2501637935638428,
+      "learning_rate": 1.968811886691952e-05,
+      "loss": 1.052,
+      "step": 1872
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8632465600967407,
+      "learning_rate": 1.968765838854199e-05,
+      "loss": 1.0023,
+      "step": 1873
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9477862119674683,
+      "learning_rate": 1.9687197575870435e-05,
+      "loss": 1.1966,
+      "step": 1874
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.224736452102661,
+      "learning_rate": 1.9686736428920748e-05,
+      "loss": 1.0793,
+      "step": 1875
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.985801100730896,
+      "learning_rate": 1.9686274947708848e-05,
+      "loss": 1.1082,
+      "step": 1876
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.041146755218506,
+      "learning_rate": 1.9685813132250655e-05,
+      "loss": 1.166,
+      "step": 1877
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9730193614959717,
+      "learning_rate": 1.968535098256211e-05,
+      "loss": 1.0668,
+      "step": 1878
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.007770299911499,
+      "learning_rate": 1.9684888498659154e-05,
+      "loss": 1.1746,
+      "step": 1879
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2531676292419434,
+      "learning_rate": 1.968442568055775e-05,
+      "loss": 1.0496,
+      "step": 1880
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1449577808380127,
+      "learning_rate": 1.968396252827387e-05,
+      "loss": 1.1088,
+      "step": 1881
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.030939817428589,
+      "learning_rate": 1.9683499041823495e-05,
+      "loss": 1.1058,
+      "step": 1882
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7645504474639893,
+      "learning_rate": 1.9683035221222617e-05,
+      "loss": 0.601,
+      "step": 1883
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0631253719329834,
+      "learning_rate": 1.9682571066487242e-05,
+      "loss": 1.1194,
+      "step": 1884
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9948899745941162,
+      "learning_rate": 1.9682106577633385e-05,
+      "loss": 1.0699,
+      "step": 1885
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2630064487457275,
+      "learning_rate": 1.9681641754677076e-05,
+      "loss": 1.1082,
+      "step": 1886
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.015381336212158,
+      "learning_rate": 1.9681176597634353e-05,
+      "loss": 1.0819,
+      "step": 1887
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3420841693878174,
+      "learning_rate": 1.9680711106521274e-05,
+      "loss": 0.6015,
+      "step": 1888
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.021106719970703,
+      "learning_rate": 1.9680245281353894e-05,
+      "loss": 1.1318,
+      "step": 1889
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0298774242401123,
+      "learning_rate": 1.967977912214829e-05,
+      "loss": 1.1653,
+      "step": 1890
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0051352977752686,
+      "learning_rate": 1.9679312628920546e-05,
+      "loss": 1.1263,
+      "step": 1891
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2140235900878906,
+      "learning_rate": 1.9678845801686766e-05,
+      "loss": 1.1582,
+      "step": 1892
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0819430351257324,
+      "learning_rate": 1.9678378640463053e-05,
+      "loss": 1.0603,
+      "step": 1893
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0939204692840576,
+      "learning_rate": 1.9677911145265524e-05,
+      "loss": 1.0927,
+      "step": 1894
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9242463111877441,
+      "learning_rate": 1.9677443316110317e-05,
+      "loss": 1.0455,
+      "step": 1895
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8855524063110352,
+      "learning_rate": 1.9676975153013574e-05,
+      "loss": 1.0853,
+      "step": 1896
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9202202558517456,
+      "learning_rate": 1.9676506655991453e-05,
+      "loss": 1.0659,
+      "step": 1897
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2485251426696777,
+      "learning_rate": 1.9676037825060117e-05,
+      "loss": 1.0615,
+      "step": 1898
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.113287091255188,
+      "learning_rate": 1.9675568660235746e-05,
+      "loss": 0.5635,
+      "step": 1899
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0088906288146973,
+      "learning_rate": 1.9675099161534524e-05,
+      "loss": 1.0554,
+      "step": 1900
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8528177738189697,
+      "learning_rate": 1.9674629328972657e-05,
+      "loss": 1.1226,
+      "step": 1901
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.98314368724823,
+      "learning_rate": 1.967415916256636e-05,
+      "loss": 1.0786,
+      "step": 1902
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.992766261100769,
+      "learning_rate": 1.9673688662331848e-05,
+      "loss": 1.0806,
+      "step": 1903
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0704426765441895,
+      "learning_rate": 1.967321782828537e-05,
+      "loss": 1.0969,
+      "step": 1904
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5542867183685303,
+      "learning_rate": 1.967274666044316e-05,
+      "loss": 1.0653,
+      "step": 1905
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.056108236312866,
+      "learning_rate": 1.9672275158821486e-05,
+      "loss": 1.0861,
+      "step": 1906
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.040116548538208,
+      "learning_rate": 1.9671803323436612e-05,
+      "loss": 1.0995,
+      "step": 1907
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.223264694213867,
+      "learning_rate": 1.9671331154304823e-05,
+      "loss": 1.0498,
+      "step": 1908
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.055952548980713,
+      "learning_rate": 1.967085865144241e-05,
+      "loss": 1.1523,
+      "step": 1909
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.131486654281616,
+      "learning_rate": 1.9670385814865685e-05,
+      "loss": 1.1237,
+      "step": 1910
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.133578062057495,
+      "learning_rate": 1.9669912644590954e-05,
+      "loss": 1.0621,
+      "step": 1911
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0090739727020264,
+      "learning_rate": 1.9669439140634552e-05,
+      "loss": 1.0827,
+      "step": 1912
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9790887832641602,
+      "learning_rate": 1.9668965303012815e-05,
+      "loss": 1.119,
+      "step": 1913
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1349194049835205,
+      "learning_rate": 1.9668491131742092e-05,
+      "loss": 1.0418,
+      "step": 1914
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.134464979171753,
+      "learning_rate": 1.9668016626838753e-05,
+      "loss": 1.0664,
+      "step": 1915
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0129239559173584,
+      "learning_rate": 1.966754178831916e-05,
+      "loss": 1.2199,
+      "step": 1916
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9092477560043335,
+      "learning_rate": 1.9667066616199712e-05,
+      "loss": 1.0602,
+      "step": 1917
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0964646339416504,
+      "learning_rate": 1.9666591110496794e-05,
+      "loss": 1.006,
+      "step": 1918
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0655882358551025,
+      "learning_rate": 1.9666115271226823e-05,
+      "loss": 1.1782,
+      "step": 1919
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.125945806503296,
+      "learning_rate": 1.9665639098406215e-05,
+      "loss": 1.0972,
+      "step": 1920
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.007563591003418,
+      "learning_rate": 1.9665162592051397e-05,
+      "loss": 1.048,
+      "step": 1921
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.987390160560608,
+      "learning_rate": 1.9664685752178817e-05,
+      "loss": 1.0839,
+      "step": 1922
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0545575618743896,
+      "learning_rate": 1.9664208578804934e-05,
+      "loss": 1.0552,
+      "step": 1923
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9538164138793945,
+      "learning_rate": 1.9663731071946207e-05,
+      "loss": 1.0589,
+      "step": 1924
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.092909574508667,
+      "learning_rate": 1.9663253231619113e-05,
+      "loss": 1.0941,
+      "step": 1925
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1029818058013916,
+      "learning_rate": 1.9662775057840145e-05,
+      "loss": 1.122,
+      "step": 1926
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0009849071502686,
+      "learning_rate": 1.96622965506258e-05,
+      "loss": 1.08,
+      "step": 1927
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9353209733963013,
+      "learning_rate": 1.9661817709992593e-05,
+      "loss": 1.0749,
+      "step": 1928
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9612815380096436,
+      "learning_rate": 1.9661338535957046e-05,
+      "loss": 1.0463,
+      "step": 1929
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9495552778244019,
+      "learning_rate": 1.9660859028535694e-05,
+      "loss": 1.0123,
+      "step": 1930
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0252058506011963,
+      "learning_rate": 1.966037918774508e-05,
+      "loss": 1.0127,
+      "step": 1931
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.110083818435669,
+      "learning_rate": 1.9659899013601772e-05,
+      "loss": 1.0423,
+      "step": 1932
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.054593801498413,
+      "learning_rate": 1.9659418506122328e-05,
+      "loss": 1.0695,
+      "step": 1933
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9878177642822266,
+      "learning_rate": 1.9658937665323337e-05,
+      "loss": 1.1189,
+      "step": 1934
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.144376516342163,
+      "learning_rate": 1.9658456491221387e-05,
+      "loss": 1.106,
+      "step": 1935
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2118537425994873,
+      "learning_rate": 1.965797498383308e-05,
+      "loss": 1.0582,
+      "step": 1936
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9113328456878662,
+      "learning_rate": 1.965749314317504e-05,
+      "loss": 1.1397,
+      "step": 1937
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8301200866699219,
+      "learning_rate": 1.9657010969263887e-05,
+      "loss": 1.0227,
+      "step": 1938
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0810914039611816,
+      "learning_rate": 1.965652846211626e-05,
+      "loss": 1.0967,
+      "step": 1939
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9043028354644775,
+      "learning_rate": 1.965604562174881e-05,
+      "loss": 1.0748,
+      "step": 1940
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9406917095184326,
+      "learning_rate": 1.96555624481782e-05,
+      "loss": 0.9878,
+      "step": 1941
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1192684173583984,
+      "learning_rate": 1.96550789414211e-05,
+      "loss": 1.0755,
+      "step": 1942
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9011651277542114,
+      "learning_rate": 1.9654595101494198e-05,
+      "loss": 1.0645,
+      "step": 1943
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8229209184646606,
+      "learning_rate": 1.965411092841419e-05,
+      "loss": 1.0192,
+      "step": 1944
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9243736267089844,
+      "learning_rate": 1.9653626422197778e-05,
+      "loss": 1.1633,
+      "step": 1945
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8593775033950806,
+      "learning_rate": 1.9653141582861683e-05,
+      "loss": 1.0684,
+      "step": 1946
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.969380259513855,
+      "learning_rate": 1.965265641042264e-05,
+      "loss": 1.137,
+      "step": 1947
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.909165620803833,
+      "learning_rate": 1.965217090489739e-05,
+      "loss": 1.0478,
+      "step": 1948
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.000683307647705,
+      "learning_rate": 1.965168506630268e-05,
+      "loss": 1.0516,
+      "step": 1949
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9607040882110596,
+      "learning_rate": 1.9651198894655278e-05,
+      "loss": 1.1354,
+      "step": 1950
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.953071117401123,
+      "learning_rate": 1.9650712389971964e-05,
+      "loss": 1.0954,
+      "step": 1951
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.007289171218872,
+      "learning_rate": 1.9650225552269526e-05,
+      "loss": 1.0816,
+      "step": 1952
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.94691801071167,
+      "learning_rate": 1.964973838156476e-05,
+      "loss": 1.0984,
+      "step": 1953
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9262628555297852,
+      "learning_rate": 1.9649250877874476e-05,
+      "loss": 1.0559,
+      "step": 1954
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9184095859527588,
+      "learning_rate": 1.96487630412155e-05,
+      "loss": 1.0098,
+      "step": 1955
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0065128803253174,
+      "learning_rate": 1.9648274871604663e-05,
+      "loss": 1.0948,
+      "step": 1956
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.178722858428955,
+      "learning_rate": 1.964778636905881e-05,
+      "loss": 0.598,
+      "step": 1957
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8920912742614746,
+      "learning_rate": 1.96472975335948e-05,
+      "loss": 1.0701,
+      "step": 1958
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.931671142578125,
+      "learning_rate": 1.9646808365229506e-05,
+      "loss": 1.137,
+      "step": 1959
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.992349624633789,
+      "learning_rate": 1.9646318863979797e-05,
+      "loss": 1.0803,
+      "step": 1960
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.033719778060913,
+      "learning_rate": 1.964582902986257e-05,
+      "loss": 0.5966,
+      "step": 1961
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.057173013687134,
+      "learning_rate": 1.964533886289473e-05,
+      "loss": 1.1329,
+      "step": 1962
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.076580286026001,
+      "learning_rate": 1.964484836309319e-05,
+      "loss": 1.0282,
+      "step": 1963
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.943109393119812,
+      "learning_rate": 1.9644357530474875e-05,
+      "loss": 1.1523,
+      "step": 1964
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.162402629852295,
+      "learning_rate": 1.964386636505672e-05,
+      "loss": 0.6459,
+      "step": 1965
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.960707426071167,
+      "learning_rate": 1.9643374866855674e-05,
+      "loss": 1.07,
+      "step": 1966
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9734525680541992,
+      "learning_rate": 1.96428830358887e-05,
+      "loss": 1.0551,
+      "step": 1967
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0760449171066284,
+      "learning_rate": 1.9642390872172773e-05,
+      "loss": 0.6081,
+      "step": 1968
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9683618545532227,
+      "learning_rate": 1.964189837572487e-05,
+      "loss": 1.1434,
+      "step": 1969
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.862407922744751,
+      "learning_rate": 1.9641405546561984e-05,
+      "loss": 1.0504,
+      "step": 1970
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9058871269226074,
+      "learning_rate": 1.9640912384701124e-05,
+      "loss": 1.0426,
+      "step": 1971
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0674712657928467,
+      "learning_rate": 1.9640418890159313e-05,
+      "loss": 0.9859,
+      "step": 1972
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9644389152526855,
+      "learning_rate": 1.9639925062953576e-05,
+      "loss": 1.0761,
+      "step": 1973
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9584554433822632,
+      "learning_rate": 1.963943090310095e-05,
+      "loss": 1.1416,
+      "step": 1974
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9575107097625732,
+      "learning_rate": 1.963893641061849e-05,
+      "loss": 1.0598,
+      "step": 1975
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8889191150665283,
+      "learning_rate": 1.963844158552326e-05,
+      "loss": 1.108,
+      "step": 1976
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.106147289276123,
+      "learning_rate": 1.9637946427832337e-05,
+      "loss": 1.0572,
+      "step": 1977
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8799062967300415,
+      "learning_rate": 1.9637450937562805e-05,
+      "loss": 1.028,
+      "step": 1978
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9847042560577393,
+      "learning_rate": 1.963695511473176e-05,
+      "loss": 1.0838,
+      "step": 1979
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9777265787124634,
+      "learning_rate": 1.963645895935632e-05,
+      "loss": 1.1227,
+      "step": 1980
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9350816011428833,
+      "learning_rate": 1.963596247145359e-05,
+      "loss": 1.0609,
+      "step": 1981
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.146388292312622,
+      "learning_rate": 1.9635465651040717e-05,
+      "loss": 1.0326,
+      "step": 1982
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9863266944885254,
+      "learning_rate": 1.963496849813484e-05,
+      "loss": 1.1206,
+      "step": 1983
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9964183568954468,
+      "learning_rate": 1.9634471012753115e-05,
+      "loss": 1.1081,
+      "step": 1984
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8813986778259277,
+      "learning_rate": 1.9633973194912708e-05,
+      "loss": 1.0745,
+      "step": 1985
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8746603727340698,
+      "learning_rate": 1.9633475044630795e-05,
+      "loss": 1.024,
+      "step": 1986
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8395682573318481,
+      "learning_rate": 1.9632976561924572e-05,
+      "loss": 1.0373,
+      "step": 1987
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8946608304977417,
+      "learning_rate": 1.9632477746811232e-05,
+      "loss": 0.9967,
+      "step": 1988
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.043721914291382,
+      "learning_rate": 1.9631978599308e-05,
+      "loss": 1.0953,
+      "step": 1989
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9646673202514648,
+      "learning_rate": 1.9631479119432085e-05,
+      "loss": 1.1034,
+      "step": 1990
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1654207706451416,
+      "learning_rate": 1.9630979307200732e-05,
+      "loss": 1.0211,
+      "step": 1991
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.406587600708008,
+      "learning_rate": 1.9630479162631183e-05,
+      "loss": 1.0901,
+      "step": 1992
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0283422470092773,
+      "learning_rate": 1.9629978685740706e-05,
+      "loss": 1.1152,
+      "step": 1993
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0785393714904785,
+      "learning_rate": 1.962947787654656e-05,
+      "loss": 1.1314,
+      "step": 1994
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.983107089996338,
+      "learning_rate": 1.9628976735066036e-05,
+      "loss": 1.0518,
+      "step": 1995
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.92808198928833,
+      "learning_rate": 1.962847526131642e-05,
+      "loss": 1.1066,
+      "step": 1996
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.173936367034912,
+      "learning_rate": 1.9627973455315014e-05,
+      "loss": 1.1249,
+      "step": 1997
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9850523471832275,
+      "learning_rate": 1.9627471317079146e-05,
+      "loss": 1.1744,
+      "step": 1998
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9090620279312134,
+      "learning_rate": 1.9626968846626134e-05,
+      "loss": 1.0545,
+      "step": 1999
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2168614864349365,
+      "learning_rate": 1.962646604397332e-05,
+      "loss": 0.662,
+      "step": 2000
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0397729873657227,
+      "learning_rate": 1.9625962909138048e-05,
+      "loss": 1.1416,
+      "step": 2001
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2100512981414795,
+      "learning_rate": 1.9625459442137688e-05,
+      "loss": 1.1167,
+      "step": 2002
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1123437881469727,
+      "learning_rate": 1.962495564298961e-05,
+      "loss": 1.1972,
+      "step": 2003
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.799432396888733,
+      "learning_rate": 1.96244515117112e-05,
+      "loss": 1.1289,
+      "step": 2004
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.344395160675049,
+      "learning_rate": 1.9623947048319854e-05,
+      "loss": 1.1671,
+      "step": 2005
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9976422786712646,
+      "learning_rate": 1.962344225283298e-05,
+      "loss": 1.0938,
+      "step": 2006
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.09765625,
+      "learning_rate": 1.962293712526799e-05,
+      "loss": 1.0949,
+      "step": 2007
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8186978101730347,
+      "learning_rate": 1.9622431665642324e-05,
+      "loss": 1.0912,
+      "step": 2008
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8941640853881836,
+      "learning_rate": 1.962192587397342e-05,
+      "loss": 1.0673,
+      "step": 2009
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1862854957580566,
+      "learning_rate": 1.9621419750278732e-05,
+      "loss": 1.0662,
+      "step": 2010
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.113650321960449,
+      "learning_rate": 1.9620913294575724e-05,
+      "loss": 1.0753,
+      "step": 2011
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.036569118499756,
+      "learning_rate": 1.9620406506881876e-05,
+      "loss": 1.0115,
+      "step": 2012
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9064944982528687,
+      "learning_rate": 1.961989938721467e-05,
+      "loss": 1.1051,
+      "step": 2013
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9965492486953735,
+      "learning_rate": 1.961939193559161e-05,
+      "loss": 1.0674,
+      "step": 2014
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1026313304901123,
+      "learning_rate": 1.9618884152030206e-05,
+      "loss": 1.0898,
+      "step": 2015
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.073885202407837,
+      "learning_rate": 1.9618376036547974e-05,
+      "loss": 1.0513,
+      "step": 2016
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2151987552642822,
+      "learning_rate": 1.9617867589162457e-05,
+      "loss": 1.0662,
+      "step": 2017
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0801613330841064,
+      "learning_rate": 1.9617358809891198e-05,
+      "loss": 1.0828,
+      "step": 2018
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9782872200012207,
+      "learning_rate": 1.9616849698751748e-05,
+      "loss": 1.1069,
+      "step": 2019
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.11164927482605,
+      "learning_rate": 1.9616340255761676e-05,
+      "loss": 1.1182,
+      "step": 2020
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3223094940185547,
+      "learning_rate": 1.961583048093857e-05,
+      "loss": 1.12,
+      "step": 2021
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.186772346496582,
+      "learning_rate": 1.961532037430001e-05,
+      "loss": 1.0665,
+      "step": 2022
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7606115341186523,
+      "learning_rate": 1.961480993586361e-05,
+      "loss": 1.099,
+      "step": 2023
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.172051191329956,
+      "learning_rate": 1.961429916564697e-05,
+      "loss": 1.0889,
+      "step": 2024
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.078150987625122,
+      "learning_rate": 1.9613788063667722e-05,
+      "loss": 1.0649,
+      "step": 2025
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.193549394607544,
+      "learning_rate": 1.9613276629943504e-05,
+      "loss": 1.1117,
+      "step": 2026
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.124544143676758,
+      "learning_rate": 1.9612764864491968e-05,
+      "loss": 1.0571,
+      "step": 2027
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6797614097595215,
+      "learning_rate": 1.9612252767330763e-05,
+      "loss": 1.0941,
+      "step": 2028
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.235015392303467,
+      "learning_rate": 1.961174033847757e-05,
+      "loss": 1.1047,
+      "step": 2029
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.124994993209839,
+      "learning_rate": 1.9611227577950065e-05,
+      "loss": 1.0651,
+      "step": 2030
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.851344347000122,
+      "learning_rate": 1.961071448576594e-05,
+      "loss": 1.0507,
+      "step": 2031
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0498554706573486,
+      "learning_rate": 1.9610201061942913e-05,
+      "loss": 1.095,
+      "step": 2032
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.255980968475342,
+      "learning_rate": 1.9609687306498686e-05,
+      "loss": 1.0895,
+      "step": 2033
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.180277943611145,
+      "learning_rate": 1.9609173219450998e-05,
+      "loss": 0.6238,
+      "step": 2034
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.249208688735962,
+      "learning_rate": 1.9608658800817582e-05,
+      "loss": 1.1537,
+      "step": 2035
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.799227237701416,
+      "learning_rate": 1.9608144050616192e-05,
+      "loss": 1.0423,
+      "step": 2036
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0315423011779785,
+      "learning_rate": 1.9607628968864588e-05,
+      "loss": 1.0973,
+      "step": 2037
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9563486576080322,
+      "learning_rate": 1.9607113555580548e-05,
+      "loss": 1.1245,
+      "step": 2038
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1117281913757324,
+      "learning_rate": 1.9606597810781856e-05,
+      "loss": 1.0557,
+      "step": 2039
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2255120277404785,
+      "learning_rate": 1.9606081734486307e-05,
+      "loss": 1.248,
+      "step": 2040
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9935283660888672,
+      "learning_rate": 1.9605565326711712e-05,
+      "loss": 1.109,
+      "step": 2041
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.140305995941162,
+      "learning_rate": 1.960504858747589e-05,
+      "loss": 1.1328,
+      "step": 2042
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9693390130996704,
+      "learning_rate": 1.960453151679667e-05,
+      "loss": 1.1152,
+      "step": 2043
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.154205322265625,
+      "learning_rate": 1.96040141146919e-05,
+      "loss": 1.0748,
+      "step": 2044
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9009878635406494,
+      "learning_rate": 1.9603496381179428e-05,
+      "loss": 1.1232,
+      "step": 2045
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2208219766616821,
+      "learning_rate": 1.9602978316277124e-05,
+      "loss": 0.6276,
+      "step": 2046
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9022663831710815,
+      "learning_rate": 1.9602459920002862e-05,
+      "loss": 1.0385,
+      "step": 2047
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.243781566619873,
+      "learning_rate": 1.960194119237453e-05,
+      "loss": 1.1051,
+      "step": 2048
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0564043521881104,
+      "learning_rate": 1.9601422133410032e-05,
+      "loss": 1.0592,
+      "step": 2049
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.875037431716919,
+      "learning_rate": 1.9600902743127276e-05,
+      "loss": 1.0151,
+      "step": 2050
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.07086181640625,
+      "learning_rate": 1.960038302154418e-05,
+      "loss": 1.0879,
+      "step": 2051
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8443243503570557,
+      "learning_rate": 1.9599862968678687e-05,
+      "loss": 1.0805,
+      "step": 2052
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0580105781555176,
+      "learning_rate": 1.9599342584548745e-05,
+      "loss": 1.1073,
+      "step": 2053
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.344372034072876,
+      "learning_rate": 1.9598821869172298e-05,
+      "loss": 1.1261,
+      "step": 2054
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9451229572296143,
+      "learning_rate": 1.9598300822567324e-05,
+      "loss": 1.1401,
+      "step": 2055
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.053720474243164,
+      "learning_rate": 1.95977794447518e-05,
+      "loss": 1.0937,
+      "step": 2056
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9877467155456543,
+      "learning_rate": 1.959725773574372e-05,
+      "loss": 0.986,
+      "step": 2057
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7685565948486328,
+      "learning_rate": 1.9596735695561082e-05,
+      "loss": 1.0411,
+      "step": 2058
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.957061529159546,
+      "learning_rate": 1.95962133242219e-05,
+      "loss": 1.0615,
+      "step": 2059
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9303545951843262,
+      "learning_rate": 1.9595690621744208e-05,
+      "loss": 1.0526,
+      "step": 2060
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1168856620788574,
+      "learning_rate": 1.9595167588146036e-05,
+      "loss": 0.9692,
+      "step": 2061
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.991184949874878,
+      "learning_rate": 1.9594644223445432e-05,
+      "loss": 1.036,
+      "step": 2062
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9844151735305786,
+      "learning_rate": 1.9594120527660453e-05,
+      "loss": 0.9962,
+      "step": 2063
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9535188674926758,
+      "learning_rate": 1.9593596500809183e-05,
+      "loss": 1.0292,
+      "step": 2064
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.180851459503174,
+      "learning_rate": 1.9593072142909692e-05,
+      "loss": 1.1836,
+      "step": 2065
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.012561321258545,
+      "learning_rate": 1.9592547453980076e-05,
+      "loss": 1.0704,
+      "step": 2066
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.056328296661377,
+      "learning_rate": 1.9592022434038447e-05,
+      "loss": 0.5338,
+      "step": 2067
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.183631420135498,
+      "learning_rate": 1.9591497083102916e-05,
+      "loss": 1.1009,
+      "step": 2068
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.96914803981781,
+      "learning_rate": 1.9590971401191616e-05,
+      "loss": 1.0772,
+      "step": 2069
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.887779712677002,
+      "learning_rate": 1.959044538832268e-05,
+      "loss": 1.1447,
+      "step": 2070
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3084380626678467,
+      "learning_rate": 1.9589919044514267e-05,
+      "loss": 1.0451,
+      "step": 2071
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.010209083557129,
+      "learning_rate": 1.9589392369784536e-05,
+      "loss": 1.1424,
+      "step": 2072
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9774534702301025,
+      "learning_rate": 1.958886536415166e-05,
+      "loss": 1.027,
+      "step": 2073
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9513734579086304,
+      "learning_rate": 1.9588338027633824e-05,
+      "loss": 1.0508,
+      "step": 2074
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.01263427734375,
+      "learning_rate": 1.9587810360249228e-05,
+      "loss": 1.06,
+      "step": 2075
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0426251888275146,
+      "learning_rate": 1.9587282362016083e-05,
+      "loss": 1.0911,
+      "step": 2076
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9721602201461792,
+      "learning_rate": 1.9586754032952598e-05,
+      "loss": 1.1562,
+      "step": 2077
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0045042037963867,
+      "learning_rate": 1.9586225373077018e-05,
+      "loss": 1.0035,
+      "step": 2078
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9218757152557373,
+      "learning_rate": 1.9585696382407573e-05,
+      "loss": 1.0527,
+      "step": 2079
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.082410454750061,
+      "learning_rate": 1.9585167060962523e-05,
+      "loss": 0.5503,
+      "step": 2080
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9107838869094849,
+      "learning_rate": 1.9584637408760133e-05,
+      "loss": 1.068,
+      "step": 2081
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1332125663757324,
+      "learning_rate": 1.9584107425818682e-05,
+      "loss": 1.1124,
+      "step": 2082
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8978484869003296,
+      "learning_rate": 1.9583577112156456e-05,
+      "loss": 1.0579,
+      "step": 2083
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9840205907821655,
+      "learning_rate": 1.958304646779175e-05,
+      "loss": 1.0873,
+      "step": 2084
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0169150829315186,
+      "learning_rate": 1.9582515492742883e-05,
+      "loss": 1.1131,
+      "step": 2085
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6841520071029663,
+      "learning_rate": 1.9581984187028174e-05,
+      "loss": 1.0539,
+      "step": 2086
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9439979791641235,
+      "learning_rate": 1.9581452550665956e-05,
+      "loss": 1.1329,
+      "step": 2087
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.186230421066284,
+      "learning_rate": 1.9580920583674573e-05,
+      "loss": 1.121,
+      "step": 2088
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9828510284423828,
+      "learning_rate": 1.9580388286072388e-05,
+      "loss": 1.103,
+      "step": 2089
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8692365884780884,
+      "learning_rate": 1.9579855657877763e-05,
+      "loss": 1.0305,
+      "step": 2090
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.121267318725586,
+      "learning_rate": 1.957932269910908e-05,
+      "loss": 1.0995,
+      "step": 2091
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8128423690795898,
+      "learning_rate": 1.9578789409784727e-05,
+      "loss": 1.1519,
+      "step": 2092
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1580734252929688,
+      "learning_rate": 1.957825578992311e-05,
+      "loss": 1.0303,
+      "step": 2093
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0924642086029053,
+      "learning_rate": 1.9577721839542646e-05,
+      "loss": 1.1129,
+      "step": 2094
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.051870346069336,
+      "learning_rate": 1.957718755866175e-05,
+      "loss": 1.1684,
+      "step": 2095
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.242689847946167,
+      "learning_rate": 1.957665294729887e-05,
+      "loss": 1.0697,
+      "step": 2096
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.096494436264038,
+      "learning_rate": 1.9576118005472442e-05,
+      "loss": 1.0437,
+      "step": 2097
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9058254957199097,
+      "learning_rate": 1.957558273320093e-05,
+      "loss": 1.1211,
+      "step": 2098
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.138946533203125,
+      "learning_rate": 1.9575047130502813e-05,
+      "loss": 1.0596,
+      "step": 2099
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.873805046081543,
+      "learning_rate": 1.9574511197396563e-05,
+      "loss": 1.0888,
+      "step": 2100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9857383966445923,
+      "learning_rate": 1.9573974933900677e-05,
+      "loss": 1.1667,
+      "step": 2101
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.02386736869812,
+      "learning_rate": 1.957343834003366e-05,
+      "loss": 1.1306,
+      "step": 2102
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9653023481369019,
+      "learning_rate": 1.9572901415814027e-05,
+      "loss": 1.0629,
+      "step": 2103
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0490076541900635,
+      "learning_rate": 1.957236416126031e-05,
+      "loss": 1.0303,
+      "step": 2104
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9465950131416321,
+      "learning_rate": 1.9571826576391042e-05,
+      "loss": 0.5271,
+      "step": 2105
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0076417922973633,
+      "learning_rate": 1.957128866122478e-05,
+      "loss": 1.1345,
+      "step": 2106
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0429115295410156,
+      "learning_rate": 1.957075041578008e-05,
+      "loss": 1.0879,
+      "step": 2107
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4451446533203125,
+      "learning_rate": 1.9570211840075518e-05,
+      "loss": 1.1297,
+      "step": 2108
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2899601459503174,
+      "learning_rate": 1.9569672934129676e-05,
+      "loss": 1.0908,
+      "step": 2109
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0632998943328857,
+      "learning_rate": 1.9569133697961158e-05,
+      "loss": 1.0509,
+      "step": 2110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1583948135375977,
+      "learning_rate": 1.9568594131588562e-05,
+      "loss": 1.118,
+      "step": 2111
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9642024040222168,
+      "learning_rate": 1.9568054235030515e-05,
+      "loss": 1.0881,
+      "step": 2112
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.161700487136841,
+      "learning_rate": 1.9567514008305643e-05,
+      "loss": 1.0973,
+      "step": 2113
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8933343887329102,
+      "learning_rate": 1.9566973451432586e-05,
+      "loss": 1.0766,
+      "step": 2114
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3618459701538086,
+      "learning_rate": 1.9566432564430003e-05,
+      "loss": 1.0475,
+      "step": 2115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1570398807525635,
+      "learning_rate": 1.9565891347316553e-05,
+      "loss": 0.5926,
+      "step": 2116
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2885169982910156,
+      "learning_rate": 1.9565349800110915e-05,
+      "loss": 1.1096,
+      "step": 2117
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9491864442825317,
+      "learning_rate": 1.9564807922831773e-05,
+      "loss": 1.1581,
+      "step": 2118
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9246591329574585,
+      "learning_rate": 1.9564265715497827e-05,
+      "loss": 1.0032,
+      "step": 2119
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0605428218841553,
+      "learning_rate": 1.956372317812779e-05,
+      "loss": 1.0935,
+      "step": 2120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9146169424057007,
+      "learning_rate": 1.956318031074038e-05,
+      "loss": 1.0664,
+      "step": 2121
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.99824059009552,
+      "learning_rate": 1.9562637113354332e-05,
+      "loss": 1.1267,
+      "step": 2122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0577216148376465,
+      "learning_rate": 1.9562093585988392e-05,
+      "loss": 1.0702,
+      "step": 2123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8620915412902832,
+      "learning_rate": 1.9561549728661312e-05,
+      "loss": 1.0276,
+      "step": 2124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9248203039169312,
+      "learning_rate": 1.9561005541391857e-05,
+      "loss": 1.0848,
+      "step": 2125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.223719358444214,
+      "learning_rate": 1.956046102419881e-05,
+      "loss": 1.179,
+      "step": 2126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.071540117263794,
+      "learning_rate": 1.9559916177100958e-05,
+      "loss": 1.059,
+      "step": 2127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3402137756347656,
+      "learning_rate": 1.9559371000117106e-05,
+      "loss": 1.062,
+      "step": 2128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.191251516342163,
+      "learning_rate": 1.955882549326606e-05,
+      "loss": 1.1068,
+      "step": 2129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9132071733474731,
+      "learning_rate": 1.955827965656665e-05,
+      "loss": 1.1004,
+      "step": 2130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9674185514450073,
+      "learning_rate": 1.955773349003771e-05,
+      "loss": 1.1286,
+      "step": 2131
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9617855548858643,
+      "learning_rate": 1.9557186993698082e-05,
+      "loss": 1.1046,
+      "step": 2132
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.911357045173645,
+      "learning_rate": 1.9556640167566632e-05,
+      "loss": 1.0262,
+      "step": 2133
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1601715087890625,
+      "learning_rate": 1.9556093011662222e-05,
+      "loss": 0.6082,
+      "step": 2134
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9079619646072388,
+      "learning_rate": 1.955554552600374e-05,
+      "loss": 1.0338,
+      "step": 2135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.042503595352173,
+      "learning_rate": 1.9554997710610068e-05,
+      "loss": 1.0693,
+      "step": 2136
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8659805059432983,
+      "learning_rate": 1.9554449565500122e-05,
+      "loss": 1.0073,
+      "step": 2137
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9416954517364502,
+      "learning_rate": 1.955390109069281e-05,
+      "loss": 1.1641,
+      "step": 2138
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9294977188110352,
+      "learning_rate": 1.9553352286207056e-05,
+      "loss": 1.1296,
+      "step": 2139
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7696038484573364,
+      "learning_rate": 1.9552803152061803e-05,
+      "loss": 1.1345,
+      "step": 2140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.120213747024536,
+      "learning_rate": 1.9552253688276e-05,
+      "loss": 1.1134,
+      "step": 2141
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.024939775466919,
+      "learning_rate": 1.9551703894868597e-05,
+      "loss": 1.1007,
+      "step": 2142
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.229119062423706,
+      "learning_rate": 1.9551153771858578e-05,
+      "loss": 1.0399,
+      "step": 2143
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8080370426177979,
+      "learning_rate": 1.9550603319264926e-05,
+      "loss": 1.028,
+      "step": 2144
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.298851728439331,
+      "learning_rate": 1.955005253710663e-05,
+      "loss": 1.1039,
+      "step": 2145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9364268779754639,
+      "learning_rate": 1.95495014254027e-05,
+      "loss": 0.9986,
+      "step": 2146
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9137351512908936,
+      "learning_rate": 1.9548949984172148e-05,
+      "loss": 1.0187,
+      "step": 2147
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0432533025741577,
+      "learning_rate": 1.954839821343401e-05,
+      "loss": 0.5284,
+      "step": 2148
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.917034387588501,
+      "learning_rate": 1.9547846113207317e-05,
+      "loss": 1.0481,
+      "step": 2149
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.120081663131714,
+      "learning_rate": 1.954729368351113e-05,
+      "loss": 1.129,
+      "step": 2150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.048861026763916,
+      "learning_rate": 1.9546740924364504e-05,
+      "loss": 1.1379,
+      "step": 2151
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0498645305633545,
+      "learning_rate": 1.9546187835786515e-05,
+      "loss": 1.0716,
+      "step": 2152
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0628933906555176,
+      "learning_rate": 1.9545634417796255e-05,
+      "loss": 1.076,
+      "step": 2153
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9649591445922852,
+      "learning_rate": 1.9545080670412814e-05,
+      "loss": 1.1029,
+      "step": 2154
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9629981517791748,
+      "learning_rate": 1.9544526593655296e-05,
+      "loss": 1.0701,
+      "step": 2155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9879732131958008,
+      "learning_rate": 1.9543972187542833e-05,
+      "loss": 1.129,
+      "step": 2156
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.687589168548584,
+      "learning_rate": 1.9543417452094552e-05,
+      "loss": 1.0697,
+      "step": 2157
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9478509426116943,
+      "learning_rate": 1.954286238732959e-05,
+      "loss": 1.0446,
+      "step": 2158
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0066754817962646,
+      "learning_rate": 1.9542306993267105e-05,
+      "loss": 1.0888,
+      "step": 2159
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8370342254638672,
+      "learning_rate": 1.954175126992626e-05,
+      "loss": 1.0222,
+      "step": 2160
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8456997871398926,
+      "learning_rate": 1.9541195217326233e-05,
+      "loss": 1.0897,
+      "step": 2161
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.966375470161438,
+      "learning_rate": 1.954063883548621e-05,
+      "loss": 1.1349,
+      "step": 2162
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8924496173858643,
+      "learning_rate": 1.9540082124425393e-05,
+      "loss": 1.1214,
+      "step": 2163
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0375827550888062,
+      "learning_rate": 1.9539525084162993e-05,
+      "loss": 0.5749,
+      "step": 2164
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9985837936401367,
+      "learning_rate": 1.9538967714718226e-05,
+      "loss": 1.0815,
+      "step": 2165
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8893191814422607,
+      "learning_rate": 1.953841001611033e-05,
+      "loss": 1.058,
+      "step": 2166
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.957874059677124,
+      "learning_rate": 1.953785198835855e-05,
+      "loss": 1.0964,
+      "step": 2167
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0782155990600586,
+      "learning_rate": 1.953729363148214e-05,
+      "loss": 0.9915,
+      "step": 2168
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2153515815734863,
+      "learning_rate": 1.953673494550037e-05,
+      "loss": 1.1288,
+      "step": 2169
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1902804374694824,
+      "learning_rate": 1.9536175930432512e-05,
+      "loss": 1.0764,
+      "step": 2170
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.84712815284729,
+      "learning_rate": 1.9535616586297866e-05,
+      "loss": 1.1065,
+      "step": 2171
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0183358192443848,
+      "learning_rate": 1.9535056913115725e-05,
+      "loss": 1.1351,
+      "step": 2172
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9436144828796387,
+      "learning_rate": 1.9534496910905404e-05,
+      "loss": 1.0999,
+      "step": 2173
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8144478797912598,
+      "learning_rate": 1.9533936579686233e-05,
+      "loss": 1.0999,
+      "step": 2174
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.830190658569336,
+      "learning_rate": 1.953337591947754e-05,
+      "loss": 1.1035,
+      "step": 2175
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9984219074249268,
+      "learning_rate": 1.9532814930298673e-05,
+      "loss": 1.07,
+      "step": 2176
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.066513776779175,
+      "learning_rate": 1.9532253612168994e-05,
+      "loss": 1.0819,
+      "step": 2177
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0593111515045166,
+      "learning_rate": 1.953169196510787e-05,
+      "loss": 1.0248,
+      "step": 2178
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.387378215789795,
+      "learning_rate": 1.953112998913468e-05,
+      "loss": 1.0998,
+      "step": 2179
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.031853675842285,
+      "learning_rate": 1.9530567684268823e-05,
+      "loss": 1.0108,
+      "step": 2180
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.252674102783203,
+      "learning_rate": 1.953000505052969e-05,
+      "loss": 1.1022,
+      "step": 2181
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.006401538848877,
+      "learning_rate": 1.952944208793671e-05,
+      "loss": 1.1395,
+      "step": 2182
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0618743896484375,
+      "learning_rate": 1.95288787965093e-05,
+      "loss": 1.1236,
+      "step": 2183
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0417888164520264,
+      "learning_rate": 1.9528315176266904e-05,
+      "loss": 1.1649,
+      "step": 2184
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.171774387359619,
+      "learning_rate": 1.9527751227228964e-05,
+      "loss": 1.1012,
+      "step": 2185
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8630516529083252,
+      "learning_rate": 1.9527186949414948e-05,
+      "loss": 1.0835,
+      "step": 2186
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0474302768707275,
+      "learning_rate": 1.9526622342844318e-05,
+      "loss": 1.0943,
+      "step": 2187
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8254406452178955,
+      "learning_rate": 1.9526057407536565e-05,
+      "loss": 1.0067,
+      "step": 2188
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0415186882019043,
+      "learning_rate": 1.9525492143511182e-05,
+      "loss": 1.0362,
+      "step": 2189
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.998120665550232,
+      "learning_rate": 1.952492655078767e-05,
+      "loss": 1.0416,
+      "step": 2190
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9627357721328735,
+      "learning_rate": 1.9524360629385554e-05,
+      "loss": 0.9909,
+      "step": 2191
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9530761241912842,
+      "learning_rate": 1.9523794379324354e-05,
+      "loss": 1.1086,
+      "step": 2192
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0925235748291016,
+      "learning_rate": 1.9523227800623616e-05,
+      "loss": 0.9154,
+      "step": 2193
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.854523777961731,
+      "learning_rate": 1.952266089330289e-05,
+      "loss": 1.0055,
+      "step": 2194
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9151043891906738,
+      "learning_rate": 1.9522093657381733e-05,
+      "loss": 1.1637,
+      "step": 2195
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.061143398284912,
+      "learning_rate": 1.9521526092879725e-05,
+      "loss": 1.0934,
+      "step": 2196
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7574697732925415,
+      "learning_rate": 1.9520958199816448e-05,
+      "loss": 1.0863,
+      "step": 2197
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0976905822753906,
+      "learning_rate": 1.95203899782115e-05,
+      "loss": 1.1654,
+      "step": 2198
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3165946006774902,
+      "learning_rate": 1.9519821428084488e-05,
+      "loss": 1.0739,
+      "step": 2199
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0492351055145264,
+      "learning_rate": 1.9519252549455033e-05,
+      "loss": 1.1017,
+      "step": 2200
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1428918838500977,
+      "learning_rate": 1.9518683342342762e-05,
+      "loss": 1.1016,
+      "step": 2201
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2069575786590576,
+      "learning_rate": 1.9518113806767316e-05,
+      "loss": 1.0892,
+      "step": 2202
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1398799419403076,
+      "learning_rate": 1.9517543942748353e-05,
+      "loss": 1.0395,
+      "step": 2203
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9363088607788086,
+      "learning_rate": 1.951697375030553e-05,
+      "loss": 1.067,
+      "step": 2204
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.207046627998352,
+      "learning_rate": 1.9516403229458535e-05,
+      "loss": 0.6198,
+      "step": 2205
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2757289409637451,
+      "learning_rate": 1.9515832380227044e-05,
+      "loss": 0.6216,
+      "step": 2206
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8511155843734741,
+      "learning_rate": 1.9515261202630758e-05,
+      "loss": 1.1289,
+      "step": 2207
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9863609075546265,
+      "learning_rate": 1.9514689696689388e-05,
+      "loss": 1.0329,
+      "step": 2208
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.21726393699646,
+      "learning_rate": 1.9514117862422655e-05,
+      "loss": 1.162,
+      "step": 2209
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9351719617843628,
+      "learning_rate": 1.9513545699850292e-05,
+      "loss": 1.1288,
+      "step": 2210
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9839487075805664,
+      "learning_rate": 1.951297320899204e-05,
+      "loss": 1.054,
+      "step": 2211
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7833126783370972,
+      "learning_rate": 1.951240038986766e-05,
+      "loss": 0.9817,
+      "step": 2212
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2012832164764404,
+      "learning_rate": 1.951182724249691e-05,
+      "loss": 0.5744,
+      "step": 2213
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1642181873321533,
+      "learning_rate": 1.9511253766899574e-05,
+      "loss": 1.1509,
+      "step": 2214
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.121772527694702,
+      "learning_rate": 1.9510679963095437e-05,
+      "loss": 1.1528,
+      "step": 2215
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.904455304145813,
+      "learning_rate": 1.9510105831104305e-05,
+      "loss": 1.052,
+      "step": 2216
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1524360179901123,
+      "learning_rate": 1.9509531370945982e-05,
+      "loss": 1.0748,
+      "step": 2217
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9766215085983276,
+      "learning_rate": 1.95089565826403e-05,
+      "loss": 1.0318,
+      "step": 2218
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9146275520324707,
+      "learning_rate": 1.9508381466207086e-05,
+      "loss": 1.0427,
+      "step": 2219
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9278593063354492,
+      "learning_rate": 1.9507806021666188e-05,
+      "loss": 1.1385,
+      "step": 2220
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.076561689376831,
+      "learning_rate": 1.9507230249037462e-05,
+      "loss": 1.1266,
+      "step": 2221
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0840296745300293,
+      "learning_rate": 1.9506654148340783e-05,
+      "loss": 1.0215,
+      "step": 2222
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1414427757263184,
+      "learning_rate": 1.950607771959602e-05,
+      "loss": 1.1184,
+      "step": 2223
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3353934288024902,
+      "learning_rate": 1.950550096282307e-05,
+      "loss": 1.1756,
+      "step": 2224
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2285561561584473,
+      "learning_rate": 1.9504923878041834e-05,
+      "loss": 1.1113,
+      "step": 2225
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9018921852111816,
+      "learning_rate": 1.9504346465272225e-05,
+      "loss": 1.0906,
+      "step": 2226
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.903281807899475,
+      "learning_rate": 1.9503768724534172e-05,
+      "loss": 1.053,
+      "step": 2227
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2642760276794434,
+      "learning_rate": 1.9503190655847605e-05,
+      "loss": 1.0613,
+      "step": 2228
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1987807750701904,
+      "learning_rate": 1.9502612259232477e-05,
+      "loss": 1.1309,
+      "step": 2229
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0226805210113525,
+      "learning_rate": 1.9502033534708743e-05,
+      "loss": 1.0917,
+      "step": 2230
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9941848516464233,
+      "learning_rate": 1.9501454482296376e-05,
+      "loss": 1.0317,
+      "step": 2231
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9081685543060303,
+      "learning_rate": 1.9500875102015354e-05,
+      "loss": 1.1692,
+      "step": 2232
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.053341865539551,
+      "learning_rate": 1.9500295393885672e-05,
+      "loss": 1.1273,
+      "step": 2233
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.083022117614746,
+      "learning_rate": 1.9499715357927335e-05,
+      "loss": 1.0835,
+      "step": 2234
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.09658145904541,
+      "learning_rate": 1.949913499416036e-05,
+      "loss": 1.1934,
+      "step": 2235
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.007136583328247,
+      "learning_rate": 1.9498554302604768e-05,
+      "loss": 1.1158,
+      "step": 2236
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.944387674331665,
+      "learning_rate": 1.94979732832806e-05,
+      "loss": 1.0922,
+      "step": 2237
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8987300395965576,
+      "learning_rate": 1.9497391936207905e-05,
+      "loss": 1.0937,
+      "step": 2238
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9165974855422974,
+      "learning_rate": 1.949681026140674e-05,
+      "loss": 1.1416,
+      "step": 2239
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8277428150177002,
+      "learning_rate": 1.949622825889719e-05,
+      "loss": 1.1068,
+      "step": 2240
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1218738555908203,
+      "learning_rate": 1.9495645928699324e-05,
+      "loss": 1.0911,
+      "step": 2241
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1599841117858887,
+      "learning_rate": 1.9495063270833247e-05,
+      "loss": 1.0769,
+      "step": 2242
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.067570686340332,
+      "learning_rate": 1.9494480285319057e-05,
+      "loss": 1.063,
+      "step": 2243
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1810226440429688,
+      "learning_rate": 1.949389697217687e-05,
+      "loss": 1.0818,
+      "step": 2244
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.756366491317749,
+      "learning_rate": 1.9493313331426825e-05,
+      "loss": 1.0479,
+      "step": 2245
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0172007083892822,
+      "learning_rate": 1.949272936308905e-05,
+      "loss": 1.051,
+      "step": 2246
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.112607479095459,
+      "learning_rate": 1.9492145067183705e-05,
+      "loss": 1.051,
+      "step": 2247
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.024402379989624,
+      "learning_rate": 1.949156044373095e-05,
+      "loss": 1.1802,
+      "step": 2248
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2420358657836914,
+      "learning_rate": 1.9490975492750953e-05,
+      "loss": 1.0159,
+      "step": 2249
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.102362632751465,
+      "learning_rate": 1.9490390214263908e-05,
+      "loss": 1.1737,
+      "step": 2250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0069520473480225,
+      "learning_rate": 1.9489804608290005e-05,
+      "loss": 1.0418,
+      "step": 2251
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9474588632583618,
+      "learning_rate": 1.9489218674849454e-05,
+      "loss": 1.0917,
+      "step": 2252
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.03008770942688,
+      "learning_rate": 1.9488632413962473e-05,
+      "loss": 1.089,
+      "step": 2253
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8665211200714111,
+      "learning_rate": 1.94880458256493e-05,
+      "loss": 1.0565,
+      "step": 2254
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0243096351623535,
+      "learning_rate": 1.948745890993016e-05,
+      "loss": 1.1318,
+      "step": 2255
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.156773567199707,
+      "learning_rate": 1.9486871666825318e-05,
+      "loss": 1.1509,
+      "step": 2256
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.857231855392456,
+      "learning_rate": 1.9486284096355036e-05,
+      "loss": 0.989,
+      "step": 2257
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8624000549316406,
+      "learning_rate": 1.9485696198539588e-05,
+      "loss": 1.0586,
+      "step": 2258
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1422600746154785,
+      "learning_rate": 1.948510797339926e-05,
+      "loss": 1.0586,
+      "step": 2259
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.047529935836792,
+      "learning_rate": 1.9484519420954356e-05,
+      "loss": 1.1701,
+      "step": 2260
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.060432195663452,
+      "learning_rate": 1.9483930541225177e-05,
+      "loss": 1.0574,
+      "step": 2261
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.491337537765503,
+      "learning_rate": 1.9483341334232048e-05,
+      "loss": 1.0739,
+      "step": 2262
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4662020206451416,
+      "learning_rate": 1.94827517999953e-05,
+      "loss": 1.0691,
+      "step": 2263
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0126819610595703,
+      "learning_rate": 1.9482161938535275e-05,
+      "loss": 1.0706,
+      "step": 2264
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0396909713745117,
+      "learning_rate": 1.948157174987233e-05,
+      "loss": 1.0675,
+      "step": 2265
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8847274780273438,
+      "learning_rate": 1.948098123402683e-05,
+      "loss": 1.0138,
+      "step": 2266
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0830929279327393,
+      "learning_rate": 1.9480390391019153e-05,
+      "loss": 1.1052,
+      "step": 2267
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6002235412597656,
+      "learning_rate": 1.9479799220869686e-05,
+      "loss": 1.156,
+      "step": 2268
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.180441379547119,
+      "learning_rate": 1.9479207723598828e-05,
+      "loss": 0.9706,
+      "step": 2269
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9346991777420044,
+      "learning_rate": 1.947861589922699e-05,
+      "loss": 1.0266,
+      "step": 2270
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0881309509277344,
+      "learning_rate": 1.9478023747774593e-05,
+      "loss": 1.0243,
+      "step": 2271
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8805605173110962,
+      "learning_rate": 1.9477431269262076e-05,
+      "loss": 1.0334,
+      "step": 2272
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.139005661010742,
+      "learning_rate": 1.9476838463709878e-05,
+      "loss": 1.1812,
+      "step": 2273
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8011958599090576,
+      "learning_rate": 1.9476245331138455e-05,
+      "loss": 1.0882,
+      "step": 2274
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1845293045043945,
+      "learning_rate": 1.947565187156828e-05,
+      "loss": 1.1186,
+      "step": 2275
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.024052143096924,
+      "learning_rate": 1.9475058085019825e-05,
+      "loss": 1.106,
+      "step": 2276
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8693737983703613,
+      "learning_rate": 1.9474463971513584e-05,
+      "loss": 1.0715,
+      "step": 2277
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9654878377914429,
+      "learning_rate": 1.947386953107006e-05,
+      "loss": 0.9929,
+      "step": 2278
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8518799543380737,
+      "learning_rate": 1.9473274763709758e-05,
+      "loss": 1.0829,
+      "step": 2279
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.038874626159668,
+      "learning_rate": 1.9472679669453208e-05,
+      "loss": 1.1494,
+      "step": 2280
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0306472778320312,
+      "learning_rate": 1.947208424832095e-05,
+      "loss": 1.1107,
+      "step": 2281
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.212695360183716,
+      "learning_rate": 1.9471488500333518e-05,
+      "loss": 1.1498,
+      "step": 2282
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.839547038078308,
+      "learning_rate": 1.9470892425511475e-05,
+      "loss": 1.0395,
+      "step": 2283
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.993290662765503,
+      "learning_rate": 1.9470296023875387e-05,
+      "loss": 1.14,
+      "step": 2284
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1583175659179688,
+      "learning_rate": 1.9469699295445842e-05,
+      "loss": 1.0084,
+      "step": 2285
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9848265647888184,
+      "learning_rate": 1.9469102240243428e-05,
+      "loss": 1.1052,
+      "step": 2286
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.328465461730957,
+      "learning_rate": 1.9468504858288747e-05,
+      "loss": 1.0701,
+      "step": 2287
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.250830888748169,
+      "learning_rate": 1.9467907149602406e-05,
+      "loss": 1.1655,
+      "step": 2288
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3036826848983765,
+      "learning_rate": 1.9467309114205043e-05,
+      "loss": 0.6335,
+      "step": 2289
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0632299184799194,
+      "learning_rate": 1.9466710752117286e-05,
+      "loss": 0.6031,
+      "step": 2290
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9720453023910522,
+      "learning_rate": 1.9466112063359785e-05,
+      "loss": 1.0932,
+      "step": 2291
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8090431690216064,
+      "learning_rate": 1.9465513047953202e-05,
+      "loss": 1.0898,
+      "step": 2292
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9945141077041626,
+      "learning_rate": 1.94649137059182e-05,
+      "loss": 1.0587,
+      "step": 2293
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9482027292251587,
+      "learning_rate": 1.9464314037275468e-05,
+      "loss": 1.0473,
+      "step": 2294
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1146042346954346,
+      "learning_rate": 1.9463714042045695e-05,
+      "loss": 1.1568,
+      "step": 2295
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.001692295074463,
+      "learning_rate": 1.9463113720249587e-05,
+      "loss": 1.1414,
+      "step": 2296
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.757969617843628,
+      "learning_rate": 1.946251307190786e-05,
+      "loss": 0.9973,
+      "step": 2297
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1525330543518066,
+      "learning_rate": 1.9461912097041238e-05,
+      "loss": 1.0598,
+      "step": 2298
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.615339756011963,
+      "learning_rate": 1.946131079567046e-05,
+      "loss": 1.1176,
+      "step": 2299
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1890640258789062,
+      "learning_rate": 1.9460709167816274e-05,
+      "loss": 1.1411,
+      "step": 2300
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.885799765586853,
+      "learning_rate": 1.9460107213499445e-05,
+      "loss": 1.0585,
+      "step": 2301
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8118035793304443,
+      "learning_rate": 1.945950493274074e-05,
+      "loss": 1.0494,
+      "step": 2302
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.264744281768799,
+      "learning_rate": 1.9458902325560945e-05,
+      "loss": 1.0665,
+      "step": 2303
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1410720348358154,
+      "learning_rate": 1.945829939198085e-05,
+      "loss": 1.1687,
+      "step": 2304
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.058357000350952,
+      "learning_rate": 1.945769613202127e-05,
+      "loss": 1.1369,
+      "step": 2305
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9387421607971191,
+      "learning_rate": 1.9457092545703008e-05,
+      "loss": 1.0832,
+      "step": 2306
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9720033407211304,
+      "learning_rate": 1.9456488633046905e-05,
+      "loss": 1.0818,
+      "step": 2307
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9730970859527588,
+      "learning_rate": 1.9455884394073792e-05,
+      "loss": 1.1148,
+      "step": 2308
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0752041339874268,
+      "learning_rate": 1.9455279828804526e-05,
+      "loss": 1.0324,
+      "step": 2309
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9665589332580566,
+      "learning_rate": 1.945467493725996e-05,
+      "loss": 1.1027,
+      "step": 2310
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.019516706466675,
+      "learning_rate": 1.945406971946098e-05,
+      "loss": 1.1563,
+      "step": 2311
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2828277349472046,
+      "learning_rate": 1.9453464175428456e-05,
+      "loss": 0.6239,
+      "step": 2312
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.040358066558838,
+      "learning_rate": 1.945285830518329e-05,
+      "loss": 1.0476,
+      "step": 2313
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8861701488494873,
+      "learning_rate": 1.9452252108746395e-05,
+      "loss": 1.0633,
+      "step": 2314
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.930243730545044,
+      "learning_rate": 1.945164558613868e-05,
+      "loss": 1.0379,
+      "step": 2315
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9522651433944702,
+      "learning_rate": 1.9451038737381078e-05,
+      "loss": 1.0818,
+      "step": 2316
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.057356119155884,
+      "learning_rate": 1.945043156249453e-05,
+      "loss": 1.0102,
+      "step": 2317
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1683175563812256,
+      "learning_rate": 1.9449824061499986e-05,
+      "loss": 1.1164,
+      "step": 2318
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1286325454711914,
+      "learning_rate": 1.9449216234418412e-05,
+      "loss": 1.0906,
+      "step": 2319
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0584769248962402,
+      "learning_rate": 1.9448608081270782e-05,
+      "loss": 0.5686,
+      "step": 2320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1297574043273926,
+      "learning_rate": 1.944799960207808e-05,
+      "loss": 1.0278,
+      "step": 2321
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.008143186569214,
+      "learning_rate": 1.9447390796861304e-05,
+      "loss": 1.1446,
+      "step": 2322
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.016409397125244,
+      "learning_rate": 1.9446781665641465e-05,
+      "loss": 1.1213,
+      "step": 2323
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8787120580673218,
+      "learning_rate": 1.9446172208439576e-05,
+      "loss": 1.0607,
+      "step": 2324
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8465402126312256,
+      "learning_rate": 1.944556242527667e-05,
+      "loss": 0.992,
+      "step": 2325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9858829975128174,
+      "learning_rate": 1.944495231617379e-05,
+      "loss": 1.1399,
+      "step": 2326
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.149644136428833,
+      "learning_rate": 1.944434188115199e-05,
+      "loss": 0.5856,
+      "step": 2327
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8389830589294434,
+      "learning_rate": 1.9443731120232332e-05,
+      "loss": 1.0137,
+      "step": 2328
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0238380432128906,
+      "learning_rate": 1.9443120033435895e-05,
+      "loss": 1.1014,
+      "step": 2329
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.935685396194458,
+      "learning_rate": 1.9442508620783763e-05,
+      "loss": 1.144,
+      "step": 2330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.134855270385742,
+      "learning_rate": 1.9441896882297033e-05,
+      "loss": 1.1397,
+      "step": 2331
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.136145830154419,
+      "learning_rate": 1.944128481799682e-05,
+      "loss": 1.0695,
+      "step": 2332
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.144909381866455,
+      "learning_rate": 1.9440672427904238e-05,
+      "loss": 1.061,
+      "step": 2333
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9671698808670044,
+      "learning_rate": 1.9440059712040424e-05,
+      "loss": 1.0852,
+      "step": 2334
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.70538330078125,
+      "learning_rate": 1.943944667042652e-05,
+      "loss": 1.0379,
+      "step": 2335
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.898211121559143,
+      "learning_rate": 1.9438833303083677e-05,
+      "loss": 1.0494,
+      "step": 2336
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8669668436050415,
+      "learning_rate": 1.9438219610033066e-05,
+      "loss": 1.094,
+      "step": 2337
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9458999633789062,
+      "learning_rate": 1.9437605591295857e-05,
+      "loss": 1.068,
+      "step": 2338
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.055189847946167,
+      "learning_rate": 1.9436991246893244e-05,
+      "loss": 1.0871,
+      "step": 2339
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8226596117019653,
+      "learning_rate": 1.9436376576846422e-05,
+      "loss": 1.0595,
+      "step": 2340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8832745552062988,
+      "learning_rate": 1.9435761581176608e-05,
+      "loss": 0.9507,
+      "step": 2341
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0675411224365234,
+      "learning_rate": 1.9435146259905018e-05,
+      "loss": 1.1602,
+      "step": 2342
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.885493516921997,
+      "learning_rate": 1.9434530613052883e-05,
+      "loss": 1.0434,
+      "step": 2343
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1025185585021973,
+      "learning_rate": 1.9433914640641456e-05,
+      "loss": 1.11,
+      "step": 2344
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.904587984085083,
+      "learning_rate": 1.9433298342691987e-05,
+      "loss": 1.0662,
+      "step": 2345
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9344552755355835,
+      "learning_rate": 1.9432681719225737e-05,
+      "loss": 1.0696,
+      "step": 2346
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9837193489074707,
+      "learning_rate": 1.9432064770263998e-05,
+      "loss": 1.0466,
+      "step": 2347
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7730023860931396,
+      "learning_rate": 1.9431447495828046e-05,
+      "loss": 1.0446,
+      "step": 2348
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9897903203964233,
+      "learning_rate": 1.943082989593919e-05,
+      "loss": 1.1333,
+      "step": 2349
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.148547887802124,
+      "learning_rate": 1.9430211970618736e-05,
+      "loss": 1.1457,
+      "step": 2350
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9025495052337646,
+      "learning_rate": 1.9429593719888008e-05,
+      "loss": 1.0363,
+      "step": 2351
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9933695793151855,
+      "learning_rate": 1.9428975143768344e-05,
+      "loss": 1.119,
+      "step": 2352
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1240458488464355,
+      "learning_rate": 1.9428356242281084e-05,
+      "loss": 1.1107,
+      "step": 2353
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.945935845375061,
+      "learning_rate": 1.942773701544759e-05,
+      "loss": 1.059,
+      "step": 2354
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8880828619003296,
+      "learning_rate": 1.9427117463289223e-05,
+      "loss": 1.0784,
+      "step": 2355
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0734593868255615,
+      "learning_rate": 1.942649758582737e-05,
+      "loss": 1.0563,
+      "step": 2356
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0611612796783447,
+      "learning_rate": 1.9425877383083414e-05,
+      "loss": 1.1901,
+      "step": 2357
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9515230655670166,
+      "learning_rate": 1.9425256855078762e-05,
+      "loss": 1.0399,
+      "step": 2358
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8852448463439941,
+      "learning_rate": 1.942463600183482e-05,
+      "loss": 1.0744,
+      "step": 2359
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1457905769348145,
+      "learning_rate": 1.9424014823373024e-05,
+      "loss": 1.1605,
+      "step": 2360
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0216314792633057,
+      "learning_rate": 1.9423393319714797e-05,
+      "loss": 1.058,
+      "step": 2361
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9243296384811401,
+      "learning_rate": 1.942277149088159e-05,
+      "loss": 1.0756,
+      "step": 2362
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0187437534332275,
+      "learning_rate": 1.9422149336894858e-05,
+      "loss": 0.9591,
+      "step": 2363
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2650585174560547,
+      "learning_rate": 1.9421526857776074e-05,
+      "loss": 0.6359,
+      "step": 2364
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1112375259399414,
+      "learning_rate": 1.9420904053546716e-05,
+      "loss": 1.0625,
+      "step": 2365
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.089199423789978,
+      "learning_rate": 1.9420280924228277e-05,
+      "loss": 0.5762,
+      "step": 2366
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0326247215270996,
+      "learning_rate": 1.9419657469842256e-05,
+      "loss": 1.0875,
+      "step": 2367
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9220284223556519,
+      "learning_rate": 1.941903369041017e-05,
+      "loss": 1.0838,
+      "step": 2368
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.114212989807129,
+      "learning_rate": 1.941840958595354e-05,
+      "loss": 1.081,
+      "step": 2369
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0175459384918213,
+      "learning_rate": 1.9417785156493906e-05,
+      "loss": 1.0756,
+      "step": 2370
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.993507981300354,
+      "learning_rate": 1.9417160402052813e-05,
+      "loss": 1.1363,
+      "step": 2371
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.014669895172119,
+      "learning_rate": 1.941653532265182e-05,
+      "loss": 1.0987,
+      "step": 2372
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3594884872436523,
+      "learning_rate": 1.9415909918312497e-05,
+      "loss": 1.1388,
+      "step": 2373
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.099210023880005,
+      "learning_rate": 1.9415284189056426e-05,
+      "loss": 1.1534,
+      "step": 2374
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.098428249359131,
+      "learning_rate": 1.94146581349052e-05,
+      "loss": 1.0132,
+      "step": 2375
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3804538249969482,
+      "learning_rate": 1.9414031755880417e-05,
+      "loss": 1.1292,
+      "step": 2376
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.987931251525879,
+      "learning_rate": 1.9413405052003696e-05,
+      "loss": 1.0543,
+      "step": 2377
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8174099922180176,
+      "learning_rate": 1.9412778023296663e-05,
+      "loss": 1.0528,
+      "step": 2378
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8591465950012207,
+      "learning_rate": 1.9412150669780952e-05,
+      "loss": 1.1006,
+      "step": 2379
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.882876992225647,
+      "learning_rate": 1.9411522991478217e-05,
+      "loss": 1.104,
+      "step": 2380
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8860564231872559,
+      "learning_rate": 1.9410894988410113e-05,
+      "loss": 1.0943,
+      "step": 2381
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0436627864837646,
+      "learning_rate": 1.941026666059831e-05,
+      "loss": 1.0568,
+      "step": 2382
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8623932600021362,
+      "learning_rate": 1.9409638008064487e-05,
+      "loss": 1.1268,
+      "step": 2383
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9900166988372803,
+      "learning_rate": 1.9409009030830347e-05,
+      "loss": 1.068,
+      "step": 2384
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8611586093902588,
+      "learning_rate": 1.9408379728917585e-05,
+      "loss": 1.0122,
+      "step": 2385
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.971740961074829,
+      "learning_rate": 1.940775010234792e-05,
+      "loss": 1.0872,
+      "step": 2386
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8328596353530884,
+      "learning_rate": 1.940712015114308e-05,
+      "loss": 1.1006,
+      "step": 2387
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.890426516532898,
+      "learning_rate": 1.9406489875324798e-05,
+      "loss": 1.0538,
+      "step": 2388
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7593294382095337,
+      "learning_rate": 1.940585927491483e-05,
+      "loss": 1.06,
+      "step": 2389
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8916653394699097,
+      "learning_rate": 1.9405228349934933e-05,
+      "loss": 1.0819,
+      "step": 2390
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.983296275138855,
+      "learning_rate": 1.9404597100406878e-05,
+      "loss": 1.0942,
+      "step": 2391
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0611073970794678,
+      "learning_rate": 1.9403965526352447e-05,
+      "loss": 1.1865,
+      "step": 2392
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.013436794281006,
+      "learning_rate": 1.940333362779343e-05,
+      "loss": 1.1293,
+      "step": 2393
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.317640781402588,
+      "learning_rate": 1.9402701404751644e-05,
+      "loss": 0.5838,
+      "step": 2394
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8269084692001343,
+      "learning_rate": 1.9402068857248894e-05,
+      "loss": 1.0508,
+      "step": 2395
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9530673027038574,
+      "learning_rate": 1.940143598530701e-05,
+      "loss": 1.0642,
+      "step": 2396
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.137480616569519,
+      "learning_rate": 1.9400802788947833e-05,
+      "loss": 0.6317,
+      "step": 2397
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.073695182800293,
+      "learning_rate": 1.9400169268193213e-05,
+      "loss": 1.0602,
+      "step": 2398
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.39306378364563,
+      "learning_rate": 1.9399535423065014e-05,
+      "loss": 1.1315,
+      "step": 2399
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.995789647102356,
+      "learning_rate": 1.9398901253585097e-05,
+      "loss": 1.0869,
+      "step": 2400
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9781782627105713,
+      "learning_rate": 1.9398266759775354e-05,
+      "loss": 1.1367,
+      "step": 2401
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9745179414749146,
+      "learning_rate": 1.939763194165768e-05,
+      "loss": 1.1736,
+      "step": 2402
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.975692629814148,
+      "learning_rate": 1.939699679925398e-05,
+      "loss": 1.0529,
+      "step": 2403
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8244720697402954,
+      "learning_rate": 1.9396361332586168e-05,
+      "loss": 1.0436,
+      "step": 2404
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1581244468688965,
+      "learning_rate": 1.9395725541676174e-05,
+      "loss": 1.1372,
+      "step": 2405
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.852168321609497,
+      "learning_rate": 1.9395089426545938e-05,
+      "loss": 1.1173,
+      "step": 2406
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.027801990509033,
+      "learning_rate": 1.939445298721741e-05,
+      "loss": 1.1059,
+      "step": 2407
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3611423969268799,
+      "learning_rate": 1.9393816223712553e-05,
+      "loss": 0.6187,
+      "step": 2408
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0657310485839844,
+      "learning_rate": 1.9393179136053334e-05,
+      "loss": 1.0492,
+      "step": 2409
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9881436824798584,
+      "learning_rate": 1.9392541724261745e-05,
+      "loss": 1.2039,
+      "step": 2410
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.133744478225708,
+      "learning_rate": 1.9391903988359776e-05,
+      "loss": 1.08,
+      "step": 2411
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8247300386428833,
+      "learning_rate": 1.939126592836944e-05,
+      "loss": 1.0087,
+      "step": 2412
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0829696655273438,
+      "learning_rate": 1.9390627544312748e-05,
+      "loss": 1.0425,
+      "step": 2413
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0927734375,
+      "learning_rate": 1.938998883621173e-05,
+      "loss": 1.0497,
+      "step": 2414
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2136967182159424,
+      "learning_rate": 1.938934980408843e-05,
+      "loss": 1.0664,
+      "step": 2415
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1908979415893555,
+      "learning_rate": 1.9388710447964894e-05,
+      "loss": 1.0983,
+      "step": 2416
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8190349340438843,
+      "learning_rate": 1.9388070767863186e-05,
+      "loss": 1.0703,
+      "step": 2417
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9157389402389526,
+      "learning_rate": 1.9387430763805383e-05,
+      "loss": 1.023,
+      "step": 2418
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7293500900268555,
+      "learning_rate": 1.9386790435813564e-05,
+      "loss": 1.0421,
+      "step": 2419
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9776259660720825,
+      "learning_rate": 1.9386149783909827e-05,
+      "loss": 1.073,
+      "step": 2420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.014918565750122,
+      "learning_rate": 1.9385508808116287e-05,
+      "loss": 1.0156,
+      "step": 2421
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.047497272491455,
+      "learning_rate": 1.938486750845505e-05,
+      "loss": 1.034,
+      "step": 2422
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0469491481781006,
+      "learning_rate": 1.938422588494825e-05,
+      "loss": 1.1204,
+      "step": 2423
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7876993417739868,
+      "learning_rate": 1.9383583937618034e-05,
+      "loss": 1.101,
+      "step": 2424
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9604291915893555,
+      "learning_rate": 1.9382941666486542e-05,
+      "loss": 1.1114,
+      "step": 2425
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.338437080383301,
+      "learning_rate": 1.9382299071575947e-05,
+      "loss": 1.0781,
+      "step": 2426
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.771887183189392,
+      "learning_rate": 1.9381656152908418e-05,
+      "loss": 1.0719,
+      "step": 2427
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.921116590499878,
+      "learning_rate": 1.9381012910506146e-05,
+      "loss": 1.0684,
+      "step": 2428
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.204495906829834,
+      "learning_rate": 1.9380369344391318e-05,
+      "loss": 1.0315,
+      "step": 2429
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9952232837677002,
+      "learning_rate": 1.9379725454586145e-05,
+      "loss": 1.1281,
+      "step": 2430
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1511647701263428,
+      "learning_rate": 1.9379081241112855e-05,
+      "loss": 1.1339,
+      "step": 2431
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.198298931121826,
+      "learning_rate": 1.9378436703993666e-05,
+      "loss": 1.0468,
+      "step": 2432
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0758845806121826,
+      "learning_rate": 1.9377791843250825e-05,
+      "loss": 1.134,
+      "step": 2433
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8433973789215088,
+      "learning_rate": 1.937714665890658e-05,
+      "loss": 1.1025,
+      "step": 2434
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9933902025222778,
+      "learning_rate": 1.9376501150983205e-05,
+      "loss": 1.0997,
+      "step": 2435
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.266794204711914,
+      "learning_rate": 1.9375855319502964e-05,
+      "loss": 1.1177,
+      "step": 2436
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.857455849647522,
+      "learning_rate": 1.9375209164488145e-05,
+      "loss": 1.0671,
+      "step": 2437
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9561847448349,
+      "learning_rate": 1.9374562685961045e-05,
+      "loss": 1.028,
+      "step": 2438
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.066419839859009,
+      "learning_rate": 1.9373915883943975e-05,
+      "loss": 1.1097,
+      "step": 2439
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.045851707458496,
+      "learning_rate": 1.9373268758459256e-05,
+      "loss": 0.9915,
+      "step": 2440
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9956412315368652,
+      "learning_rate": 1.9372621309529213e-05,
+      "loss": 1.094,
+      "step": 2441
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8995851278305054,
+      "learning_rate": 1.937197353717619e-05,
+      "loss": 1.049,
+      "step": 2442
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.845262050628662,
+      "learning_rate": 1.9371325441422537e-05,
+      "loss": 1.0271,
+      "step": 2443
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9874464273452759,
+      "learning_rate": 1.9370677022290625e-05,
+      "loss": 1.0712,
+      "step": 2444
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.226375937461853,
+      "learning_rate": 1.9370028279802825e-05,
+      "loss": 0.5934,
+      "step": 2445
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8723481893539429,
+      "learning_rate": 1.936937921398152e-05,
+      "loss": 1.1358,
+      "step": 2446
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8159383535385132,
+      "learning_rate": 1.936872982484911e-05,
+      "loss": 1.0442,
+      "step": 2447
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.230492353439331,
+      "learning_rate": 1.9368080112428008e-05,
+      "loss": 1.1768,
+      "step": 2448
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9398930072784424,
+      "learning_rate": 1.936743007674063e-05,
+      "loss": 1.0932,
+      "step": 2449
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7045669555664062,
+      "learning_rate": 1.9366779717809402e-05,
+      "loss": 0.9882,
+      "step": 2450
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8414820432662964,
+      "learning_rate": 1.9366129035656777e-05,
+      "loss": 1.181,
+      "step": 2451
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8981040716171265,
+      "learning_rate": 1.9365478030305195e-05,
+      "loss": 1.0567,
+      "step": 2452
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1965668201446533,
+      "learning_rate": 1.9364826701777133e-05,
+      "loss": 1.1267,
+      "step": 2453
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2412266731262207,
+      "learning_rate": 1.9364175050095058e-05,
+      "loss": 1.0584,
+      "step": 2454
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8191626071929932,
+      "learning_rate": 1.9363523075281464e-05,
+      "loss": 1.0284,
+      "step": 2455
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2468509674072266,
+      "learning_rate": 1.936287077735884e-05,
+      "loss": 1.0617,
+      "step": 2456
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.997544765472412,
+      "learning_rate": 1.9362218156349707e-05,
+      "loss": 1.0942,
+      "step": 2457
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7855714559555054,
+      "learning_rate": 1.9361565212276572e-05,
+      "loss": 1.0649,
+      "step": 2458
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.017965078353882,
+      "learning_rate": 1.9360911945161975e-05,
+      "loss": 1.0604,
+      "step": 2459
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0366060733795166,
+      "learning_rate": 1.9360258355028452e-05,
+      "loss": 1.1378,
+      "step": 2460
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.044109344482422,
+      "learning_rate": 1.935960444189856e-05,
+      "loss": 1.0719,
+      "step": 2461
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7298020124435425,
+      "learning_rate": 1.9358950205794863e-05,
+      "loss": 1.0558,
+      "step": 2462
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.026771068572998,
+      "learning_rate": 1.935829564673994e-05,
+      "loss": 1.0679,
+      "step": 2463
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.177840232849121,
+      "learning_rate": 1.9357640764756377e-05,
+      "loss": 1.0011,
+      "step": 2464
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.854079246520996,
+      "learning_rate": 1.935698555986677e-05,
+      "loss": 1.0859,
+      "step": 2465
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.986857295036316,
+      "learning_rate": 1.9356330032093728e-05,
+      "loss": 1.0594,
+      "step": 2466
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.891068935394287,
+      "learning_rate": 1.935567418145987e-05,
+      "loss": 1.0968,
+      "step": 2467
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9980063438415527,
+      "learning_rate": 1.9355018007987832e-05,
+      "loss": 1.0502,
+      "step": 2468
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9222664833068848,
+      "learning_rate": 1.9354361511700256e-05,
+      "loss": 0.9977,
+      "step": 2469
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7580130100250244,
+      "learning_rate": 1.935370469261979e-05,
+      "loss": 1.059,
+      "step": 2470
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0984249114990234,
+      "learning_rate": 1.9353047550769108e-05,
+      "loss": 1.1054,
+      "step": 2471
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.885148525238037,
+      "learning_rate": 1.935239008617088e-05,
+      "loss": 1.062,
+      "step": 2472
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.815417766571045,
+      "learning_rate": 1.9351732298847797e-05,
+      "loss": 1.0972,
+      "step": 2473
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1468958854675293,
+      "learning_rate": 1.9351074188822557e-05,
+      "loss": 1.0896,
+      "step": 2474
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.023013114929199,
+      "learning_rate": 1.9350415756117863e-05,
+      "loss": 1.0329,
+      "step": 2475
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9057968854904175,
+      "learning_rate": 1.9349757000756442e-05,
+      "loss": 1.0331,
+      "step": 2476
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9393936395645142,
+      "learning_rate": 1.9349097922761026e-05,
+      "loss": 1.0806,
+      "step": 2477
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.968632698059082,
+      "learning_rate": 1.9348438522154355e-05,
+      "loss": 1.0075,
+      "step": 2478
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.918735146522522,
+      "learning_rate": 1.9347778798959184e-05,
+      "loss": 1.122,
+      "step": 2479
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9503058195114136,
+      "learning_rate": 1.934711875319828e-05,
+      "loss": 1.0386,
+      "step": 2480
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9704668521881104,
+      "learning_rate": 1.9346458384894418e-05,
+      "loss": 1.0505,
+      "step": 2481
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.732666015625,
+      "learning_rate": 1.9345797694070387e-05,
+      "loss": 1.1212,
+      "step": 2482
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.935217261314392,
+      "learning_rate": 1.934513668074898e-05,
+      "loss": 1.0639,
+      "step": 2483
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9657056331634521,
+      "learning_rate": 1.934447534495301e-05,
+      "loss": 1.0973,
+      "step": 2484
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.199621319770813,
+      "learning_rate": 1.9343813686705302e-05,
+      "loss": 0.6188,
+      "step": 2485
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1108856201171875,
+      "learning_rate": 1.9343151706028684e-05,
+      "loss": 1.0553,
+      "step": 2486
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1071860790252686,
+      "learning_rate": 1.9342489402945997e-05,
+      "loss": 1.1577,
+      "step": 2487
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8007738590240479,
+      "learning_rate": 1.9341826777480103e-05,
+      "loss": 1.0552,
+      "step": 2488
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7889597415924072,
+      "learning_rate": 1.934116382965386e-05,
+      "loss": 0.9961,
+      "step": 2489
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8007550239562988,
+      "learning_rate": 1.9340500559490146e-05,
+      "loss": 0.9874,
+      "step": 2490
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8030359745025635,
+      "learning_rate": 1.933983696701185e-05,
+      "loss": 1.004,
+      "step": 2491
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.926586389541626,
+      "learning_rate": 1.933917305224187e-05,
+      "loss": 1.0909,
+      "step": 2492
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9421710968017578,
+      "learning_rate": 1.9338508815203116e-05,
+      "loss": 1.0998,
+      "step": 2493
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.027650833129883,
+      "learning_rate": 1.9337844255918506e-05,
+      "loss": 1.084,
+      "step": 2494
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.977394461631775,
+      "learning_rate": 1.933717937441098e-05,
+      "loss": 1.0822,
+      "step": 2495
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0998809337615967,
+      "learning_rate": 1.933651417070347e-05,
+      "loss": 1.0521,
+      "step": 2496
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0941131114959717,
+      "learning_rate": 1.9335848644818942e-05,
+      "loss": 1.1314,
+      "step": 2497
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0855953693389893,
+      "learning_rate": 1.9335182796780354e-05,
+      "loss": 1.0283,
+      "step": 2498
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9205164909362793,
+      "learning_rate": 1.9334516626610685e-05,
+      "loss": 1.0564,
+      "step": 2499
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0995781421661377,
+      "learning_rate": 1.9333850134332918e-05,
+      "loss": 0.9803,
+      "step": 2500
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1396867036819458,
+      "learning_rate": 1.933318331997006e-05,
+      "loss": 0.6401,
+      "step": 2501
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.987968921661377,
+      "learning_rate": 1.9332516183545116e-05,
+      "loss": 1.1081,
+      "step": 2502
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3172152042388916,
+      "learning_rate": 1.933184872508111e-05,
+      "loss": 1.0898,
+      "step": 2503
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9781608581542969,
+      "learning_rate": 1.9331180944601067e-05,
+      "loss": 0.5482,
+      "step": 2504
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1662814617156982,
+      "learning_rate": 1.933051284212804e-05,
+      "loss": 1.0913,
+      "step": 2505
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9019211530685425,
+      "learning_rate": 1.9329844417685078e-05,
+      "loss": 1.0814,
+      "step": 2506
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.074995517730713,
+      "learning_rate": 1.9329175671295247e-05,
+      "loss": 0.9853,
+      "step": 2507
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8957983255386353,
+      "learning_rate": 1.932850660298162e-05,
+      "loss": 1.0263,
+      "step": 2508
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8568568229675293,
+      "learning_rate": 1.932783721276729e-05,
+      "loss": 0.9991,
+      "step": 2509
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9765918254852295,
+      "learning_rate": 1.9327167500675354e-05,
+      "loss": 1.08,
+      "step": 2510
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1233248710632324,
+      "learning_rate": 1.9326497466728924e-05,
+      "loss": 1.1166,
+      "step": 2511
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.980760097503662,
+      "learning_rate": 1.932582711095112e-05,
+      "loss": 1.0881,
+      "step": 2512
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.837073564529419,
+      "learning_rate": 1.932515643336507e-05,
+      "loss": 1.0646,
+      "step": 2513
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8227471113204956,
+      "learning_rate": 1.932448543399392e-05,
+      "loss": 1.0062,
+      "step": 2514
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.110034704208374,
+      "learning_rate": 1.9323814112860826e-05,
+      "loss": 1.0069,
+      "step": 2515
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9318060874938965,
+      "learning_rate": 1.9323142469988953e-05,
+      "loss": 1.0787,
+      "step": 2516
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.850795030593872,
+      "learning_rate": 1.932247050540147e-05,
+      "loss": 0.9666,
+      "step": 2517
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.138353109359741,
+      "learning_rate": 1.9321798219121575e-05,
+      "loss": 1.0713,
+      "step": 2518
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7956981658935547,
+      "learning_rate": 1.9321125611172468e-05,
+      "loss": 0.9648,
+      "step": 2519
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.281947135925293,
+      "learning_rate": 1.9320452681577348e-05,
+      "loss": 1.0801,
+      "step": 2520
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3190159797668457,
+      "learning_rate": 1.9319779430359443e-05,
+      "loss": 1.1107,
+      "step": 2521
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8956717252731323,
+      "learning_rate": 1.9319105857541983e-05,
+      "loss": 1.0818,
+      "step": 2522
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.035945415496826,
+      "learning_rate": 1.9318431963148214e-05,
+      "loss": 1.1642,
+      "step": 2523
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9879200458526611,
+      "learning_rate": 1.9317757747201386e-05,
+      "loss": 1.1075,
+      "step": 2524
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.971221685409546,
+      "learning_rate": 1.9317083209724767e-05,
+      "loss": 1.0936,
+      "step": 2525
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1511213779449463,
+      "learning_rate": 1.931640835074163e-05,
+      "loss": 1.0769,
+      "step": 2526
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8749141693115234,
+      "learning_rate": 1.9315733170275268e-05,
+      "loss": 1.0545,
+      "step": 2527
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8905028104782104,
+      "learning_rate": 1.931505766834898e-05,
+      "loss": 1.1361,
+      "step": 2528
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7451789379119873,
+      "learning_rate": 1.931438184498607e-05,
+      "loss": 1.0834,
+      "step": 2529
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0431735515594482,
+      "learning_rate": 1.9313705700209853e-05,
+      "loss": 1.0975,
+      "step": 2530
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8834114074707031,
+      "learning_rate": 1.931302923404368e-05,
+      "loss": 1.1329,
+      "step": 2531
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2932240962982178,
+      "learning_rate": 1.9312352446510877e-05,
+      "loss": 1.0912,
+      "step": 2532
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1283633708953857,
+      "learning_rate": 1.931167533763481e-05,
+      "loss": 1.056,
+      "step": 2533
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8803224563598633,
+      "learning_rate": 1.931099790743883e-05,
+      "loss": 0.9982,
+      "step": 2534
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.971746563911438,
+      "learning_rate": 1.9310320155946326e-05,
+      "loss": 1.0834,
+      "step": 2535
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9654436111450195,
+      "learning_rate": 1.9309642083180682e-05,
+      "loss": 1.1105,
+      "step": 2536
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8840471506118774,
+      "learning_rate": 1.930896368916529e-05,
+      "loss": 1.1006,
+      "step": 2537
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.26127290725708,
+      "learning_rate": 1.930828497392357e-05,
+      "loss": 1.0847,
+      "step": 2538
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3020989894866943,
+      "learning_rate": 1.9307605937478937e-05,
+      "loss": 0.6401,
+      "step": 2539
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1201605796813965,
+      "learning_rate": 1.930692657985482e-05,
+      "loss": 1.0584,
+      "step": 2540
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0672385692596436,
+      "learning_rate": 1.9306246901074666e-05,
+      "loss": 1.0254,
+      "step": 2541
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8749616146087646,
+      "learning_rate": 1.9305566901161928e-05,
+      "loss": 1.0872,
+      "step": 2542
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9761897325515747,
+      "learning_rate": 1.930488658014007e-05,
+      "loss": 1.0704,
+      "step": 2543
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8861124515533447,
+      "learning_rate": 1.9304205938032567e-05,
+      "loss": 1.0419,
+      "step": 2544
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8207732439041138,
+      "learning_rate": 1.930352497486291e-05,
+      "loss": 1.1339,
+      "step": 2545
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8331044912338257,
+      "learning_rate": 1.930284369065459e-05,
+      "loss": 1.0997,
+      "step": 2546
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9984328746795654,
+      "learning_rate": 1.9302162085431125e-05,
+      "loss": 1.1047,
+      "step": 2547
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9100768566131592,
+      "learning_rate": 1.9301480159216028e-05,
+      "loss": 1.0297,
+      "step": 2548
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.022785186767578,
+      "learning_rate": 1.9300797912032834e-05,
+      "loss": 1.0806,
+      "step": 2549
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9129924774169922,
+      "learning_rate": 1.9300115343905086e-05,
+      "loss": 1.0462,
+      "step": 2550
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9246619939804077,
+      "learning_rate": 1.9299432454856335e-05,
+      "loss": 1.062,
+      "step": 2551
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0852985382080078,
+      "learning_rate": 1.929874924491015e-05,
+      "loss": 0.5822,
+      "step": 2552
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.53330397605896,
+      "learning_rate": 1.9298065714090098e-05,
+      "loss": 1.1222,
+      "step": 2553
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.130354642868042,
+      "learning_rate": 1.9297381862419776e-05,
+      "loss": 1.0678,
+      "step": 2554
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0234578847885132,
+      "learning_rate": 1.9296697689922775e-05,
+      "loss": 0.6062,
+      "step": 2555
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9899568557739258,
+      "learning_rate": 1.929601319662271e-05,
+      "loss": 1.0975,
+      "step": 2556
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.061023235321045,
+      "learning_rate": 1.929532838254319e-05,
+      "loss": 0.554,
+      "step": 2557
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9634816646575928,
+      "learning_rate": 1.9294643247707858e-05,
+      "loss": 1.0688,
+      "step": 2558
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8687609434127808,
+      "learning_rate": 1.9293957792140348e-05,
+      "loss": 1.0515,
+      "step": 2559
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.528010606765747,
+      "learning_rate": 1.9293272015864318e-05,
+      "loss": 1.0907,
+      "step": 2560
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0524628162384033,
+      "learning_rate": 1.929258591890343e-05,
+      "loss": 1.1302,
+      "step": 2561
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.045297145843506,
+      "learning_rate": 1.929189950128136e-05,
+      "loss": 1.035,
+      "step": 2562
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9476463794708252,
+      "learning_rate": 1.9291212763021792e-05,
+      "loss": 1.0099,
+      "step": 2563
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.022969961166382,
+      "learning_rate": 1.929052570414843e-05,
+      "loss": 1.099,
+      "step": 2564
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.16217303276062,
+      "learning_rate": 1.9289838324684974e-05,
+      "loss": 1.0807,
+      "step": 2565
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1511733531951904,
+      "learning_rate": 1.928915062465515e-05,
+      "loss": 1.0421,
+      "step": 2566
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.124356985092163,
+      "learning_rate": 1.9288462604082684e-05,
+      "loss": 0.9788,
+      "step": 2567
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2225966453552246,
+      "learning_rate": 1.9287774262991324e-05,
+      "loss": 1.0287,
+      "step": 2568
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9184764623641968,
+      "learning_rate": 1.9287085601404813e-05,
+      "loss": 0.9997,
+      "step": 2569
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8319218158721924,
+      "learning_rate": 1.9286396619346925e-05,
+      "loss": 1.06,
+      "step": 2570
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9651515483856201,
+      "learning_rate": 1.9285707316841425e-05,
+      "loss": 1.1011,
+      "step": 2571
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2106685638427734,
+      "learning_rate": 1.9285017693912107e-05,
+      "loss": 1.0117,
+      "step": 2572
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3083195686340332,
+      "learning_rate": 1.9284327750582767e-05,
+      "loss": 0.5883,
+      "step": 2573
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9038532972335815,
+      "learning_rate": 1.928363748687721e-05,
+      "loss": 1.137,
+      "step": 2574
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9197347164154053,
+      "learning_rate": 1.9282946902819253e-05,
+      "loss": 1.1562,
+      "step": 2575
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.170259952545166,
+      "learning_rate": 1.928225599843273e-05,
+      "loss": 1.1273,
+      "step": 2576
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8560339212417603,
+      "learning_rate": 1.9281564773741487e-05,
+      "loss": 1.1398,
+      "step": 2577
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9716919660568237,
+      "learning_rate": 1.9280873228769365e-05,
+      "loss": 1.0671,
+      "step": 2578
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.026346206665039,
+      "learning_rate": 1.9280181363540236e-05,
+      "loss": 1.0435,
+      "step": 2579
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1012442111968994,
+      "learning_rate": 1.9279489178077968e-05,
+      "loss": 1.0075,
+      "step": 2580
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9520851373672485,
+      "learning_rate": 1.927879667240645e-05,
+      "loss": 0.9199,
+      "step": 2581
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9819574356079102,
+      "learning_rate": 1.9278103846549582e-05,
+      "loss": 1.0847,
+      "step": 2582
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.941069483757019,
+      "learning_rate": 1.9277410700531264e-05,
+      "loss": 0.9903,
+      "step": 2583
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.026662588119507,
+      "learning_rate": 1.927671723437542e-05,
+      "loss": 1.06,
+      "step": 2584
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3707916736602783,
+      "learning_rate": 1.927602344810598e-05,
+      "loss": 1.0584,
+      "step": 2585
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.093108892440796,
+      "learning_rate": 1.927532934174688e-05,
+      "loss": 1.0611,
+      "step": 2586
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9259631633758545,
+      "learning_rate": 1.927463491532207e-05,
+      "loss": 1.0199,
+      "step": 2587
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.926685094833374,
+      "learning_rate": 1.9273940168855518e-05,
+      "loss": 1.0613,
+      "step": 2588
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8998945951461792,
+      "learning_rate": 1.92732451023712e-05,
+      "loss": 1.0206,
+      "step": 2589
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0324254035949707,
+      "learning_rate": 1.9272549715893097e-05,
+      "loss": 1.0275,
+      "step": 2590
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1829004287719727,
+      "learning_rate": 1.9271854009445202e-05,
+      "loss": 0.5712,
+      "step": 2591
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3386924266815186,
+      "learning_rate": 1.927115798305153e-05,
+      "loss": 1.0463,
+      "step": 2592
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.066807746887207,
+      "learning_rate": 1.9270461636736087e-05,
+      "loss": 1.0889,
+      "step": 2593
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9856654405593872,
+      "learning_rate": 1.9269764970522915e-05,
+      "loss": 1.0015,
+      "step": 2594
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9622710943222046,
+      "learning_rate": 1.9269067984436045e-05,
+      "loss": 1.094,
+      "step": 2595
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.111218214035034,
+      "learning_rate": 1.926837067849953e-05,
+      "loss": 1.0647,
+      "step": 2596
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9046671390533447,
+      "learning_rate": 1.9267673052737438e-05,
+      "loss": 1.1955,
+      "step": 2597
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.029859781265259,
+      "learning_rate": 1.9266975107173834e-05,
+      "loss": 1.0461,
+      "step": 2598
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.970551609992981,
+      "learning_rate": 1.9266276841832802e-05,
+      "loss": 1.0148,
+      "step": 2599
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9541776180267334,
+      "learning_rate": 1.9265578256738445e-05,
+      "loss": 1.0768,
+      "step": 2600
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.333456039428711,
+      "learning_rate": 1.9264879351914866e-05,
+      "loss": 1.0762,
+      "step": 2601
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1288232803344727,
+      "learning_rate": 1.9264180127386176e-05,
+      "loss": 1.0375,
+      "step": 2602
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1100027561187744,
+      "learning_rate": 1.9263480583176514e-05,
+      "loss": 1.1167,
+      "step": 2603
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8198777437210083,
+      "learning_rate": 1.9262780719310008e-05,
+      "loss": 1.0951,
+      "step": 2604
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8601609468460083,
+      "learning_rate": 1.9262080535810815e-05,
+      "loss": 1.0751,
+      "step": 2605
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8959596157073975,
+      "learning_rate": 1.9261380032703095e-05,
+      "loss": 1.0317,
+      "step": 2606
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9837675094604492,
+      "learning_rate": 1.9260679210011024e-05,
+      "loss": 0.9995,
+      "step": 2607
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0680274963378906,
+      "learning_rate": 1.925997806775878e-05,
+      "loss": 1.0594,
+      "step": 2608
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2864267826080322,
+      "learning_rate": 1.925927660597056e-05,
+      "loss": 1.1585,
+      "step": 2609
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9845783710479736,
+      "learning_rate": 1.9258574824670567e-05,
+      "loss": 1.1241,
+      "step": 2610
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.182053565979004,
+      "learning_rate": 1.925787272388302e-05,
+      "loss": 1.112,
+      "step": 2611
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.167057991027832,
+      "learning_rate": 1.925717030363215e-05,
+      "loss": 1.0497,
+      "step": 2612
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0042035579681396,
+      "learning_rate": 1.925646756394219e-05,
+      "loss": 1.076,
+      "step": 2613
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.099884271621704,
+      "learning_rate": 1.9255764504837387e-05,
+      "loss": 1.0682,
+      "step": 2614
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9429821968078613,
+      "learning_rate": 1.9255061126342013e-05,
+      "loss": 1.0021,
+      "step": 2615
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0103068351745605,
+      "learning_rate": 1.925435742848033e-05,
+      "loss": 1.1184,
+      "step": 2616
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9453660249710083,
+      "learning_rate": 1.925365341127662e-05,
+      "loss": 1.1758,
+      "step": 2617
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9802542924880981,
+      "learning_rate": 1.925294907475518e-05,
+      "loss": 1.1017,
+      "step": 2618
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9817343950271606,
+      "learning_rate": 1.925224441894032e-05,
+      "loss": 1.0786,
+      "step": 2619
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0544378757476807,
+      "learning_rate": 1.9251539443856344e-05,
+      "loss": 0.9645,
+      "step": 2620
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1256073713302612,
+      "learning_rate": 1.925083414952759e-05,
+      "loss": 0.5849,
+      "step": 2621
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9506025314331055,
+      "learning_rate": 1.9250128535978384e-05,
+      "loss": 1.0662,
+      "step": 2622
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0655899047851562,
+      "learning_rate": 1.9249422603233086e-05,
+      "loss": 1.0693,
+      "step": 2623
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9858718514442444,
+      "learning_rate": 1.9248716351316054e-05,
+      "loss": 0.5865,
+      "step": 2624
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1444947719573975,
+      "learning_rate": 1.924800978025165e-05,
+      "loss": 0.6595,
+      "step": 2625
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2073960304260254,
+      "learning_rate": 1.9247302890064264e-05,
+      "loss": 1.0985,
+      "step": 2626
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.026515483856201,
+      "learning_rate": 1.924659568077829e-05,
+      "loss": 1.0698,
+      "step": 2627
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5139331817626953,
+      "learning_rate": 1.9245888152418123e-05,
+      "loss": 1.1387,
+      "step": 2628
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0619704723358154,
+      "learning_rate": 1.9245180305008187e-05,
+      "loss": 1.093,
+      "step": 2629
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9288030862808228,
+      "learning_rate": 1.92444721385729e-05,
+      "loss": 1.0548,
+      "step": 2630
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.969240427017212,
+      "learning_rate": 1.9243763653136707e-05,
+      "loss": 1.1001,
+      "step": 2631
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8991674184799194,
+      "learning_rate": 1.9243054848724048e-05,
+      "loss": 0.9836,
+      "step": 2632
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.947309672832489,
+      "learning_rate": 1.9242345725359392e-05,
+      "loss": 0.5858,
+      "step": 2633
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9963483810424805,
+      "learning_rate": 1.9241636283067197e-05,
+      "loss": 1.0546,
+      "step": 2634
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.097155809402466,
+      "learning_rate": 1.924092652187195e-05,
+      "loss": 1.0627,
+      "step": 2635
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.131049633026123,
+      "learning_rate": 1.9240216441798145e-05,
+      "loss": 1.0536,
+      "step": 2636
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0207908153533936,
+      "learning_rate": 1.9239506042870276e-05,
+      "loss": 1.0597,
+      "step": 2637
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1051024198532104,
+      "learning_rate": 1.9238795325112867e-05,
+      "loss": 0.563,
+      "step": 2638
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8110027313232422,
+      "learning_rate": 1.923808428855044e-05,
+      "loss": 0.9528,
+      "step": 2639
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.171895742416382,
+      "learning_rate": 1.923737293320753e-05,
+      "loss": 1.0618,
+      "step": 2640
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0588314533233643,
+      "learning_rate": 1.923666125910868e-05,
+      "loss": 1.054,
+      "step": 2641
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.107593536376953,
+      "learning_rate": 1.9235949266278455e-05,
+      "loss": 1.1372,
+      "step": 2642
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.891647219657898,
+      "learning_rate": 1.923523695474142e-05,
+      "loss": 1.064,
+      "step": 2643
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8993183374404907,
+      "learning_rate": 1.9234524324522153e-05,
+      "loss": 1.0468,
+      "step": 2644
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9360493421554565,
+      "learning_rate": 1.9233811375645247e-05,
+      "loss": 1.0893,
+      "step": 2645
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8919892311096191,
+      "learning_rate": 1.9233098108135308e-05,
+      "loss": 1.0294,
+      "step": 2646
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8131908178329468,
+      "learning_rate": 1.923238452201694e-05,
+      "loss": 0.9944,
+      "step": 2647
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9440006017684937,
+      "learning_rate": 1.9231670617314778e-05,
+      "loss": 1.0423,
+      "step": 2648
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.010502576828003,
+      "learning_rate": 1.9230956394053445e-05,
+      "loss": 1.0898,
+      "step": 2649
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1989331245422363,
+      "learning_rate": 1.9230241852257595e-05,
+      "loss": 1.0519,
+      "step": 2650
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8532699346542358,
+      "learning_rate": 1.9229526991951883e-05,
+      "loss": 1.1416,
+      "step": 2651
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3194479942321777,
+      "learning_rate": 1.9228811813160972e-05,
+      "loss": 1.1164,
+      "step": 2652
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0304808616638184,
+      "learning_rate": 1.922809631590955e-05,
+      "loss": 1.0666,
+      "step": 2653
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0340416431427,
+      "learning_rate": 1.92273805002223e-05,
+      "loss": 1.0318,
+      "step": 2654
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8742103576660156,
+      "learning_rate": 1.922666436612392e-05,
+      "loss": 1.0353,
+      "step": 2655
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2732443809509277,
+      "learning_rate": 1.9225947913639133e-05,
+      "loss": 1.1026,
+      "step": 2656
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2090322971343994,
+      "learning_rate": 1.9225231142792653e-05,
+      "loss": 1.0508,
+      "step": 2657
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7809934616088867,
+      "learning_rate": 1.9224514053609217e-05,
+      "loss": 1.0544,
+      "step": 2658
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9456626176834106,
+      "learning_rate": 1.9223796646113567e-05,
+      "loss": 1.0268,
+      "step": 2659
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7916826009750366,
+      "learning_rate": 1.922307892033046e-05,
+      "loss": 1.0921,
+      "step": 2660
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7888703346252441,
+      "learning_rate": 1.9222360876284662e-05,
+      "loss": 1.0584,
+      "step": 2661
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9444453716278076,
+      "learning_rate": 1.9221642514000955e-05,
+      "loss": 1.0953,
+      "step": 2662
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1149709224700928,
+      "learning_rate": 1.922092383350412e-05,
+      "loss": 1.0664,
+      "step": 2663
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9952826499938965,
+      "learning_rate": 1.9220204834818966e-05,
+      "loss": 1.0262,
+      "step": 2664
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.842921495437622,
+      "learning_rate": 1.9219485517970297e-05,
+      "loss": 0.9498,
+      "step": 2665
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7758725881576538,
+      "learning_rate": 1.9218765882982937e-05,
+      "loss": 1.0086,
+      "step": 2666
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.022393226623535,
+      "learning_rate": 1.9218045929881715e-05,
+      "loss": 1.0296,
+      "step": 2667
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.178647041320801,
+      "learning_rate": 1.921732565869148e-05,
+      "loss": 1.2327,
+      "step": 2668
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0059146881103516,
+      "learning_rate": 1.9216605069437086e-05,
+      "loss": 1.1601,
+      "step": 2669
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6977680921554565,
+      "learning_rate": 1.9215884162143393e-05,
+      "loss": 1.0294,
+      "step": 2670
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9692310094833374,
+      "learning_rate": 1.9215162936835283e-05,
+      "loss": 1.0727,
+      "step": 2671
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7743926048278809,
+      "learning_rate": 1.9214441393537642e-05,
+      "loss": 1.0896,
+      "step": 2672
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8910232782363892,
+      "learning_rate": 1.921371953227537e-05,
+      "loss": 1.1039,
+      "step": 2673
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.16933012008667,
+      "learning_rate": 1.9212997353073367e-05,
+      "loss": 1.1841,
+      "step": 2674
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.077240467071533,
+      "learning_rate": 1.9212274855956566e-05,
+      "loss": 1.1365,
+      "step": 2675
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9534598588943481,
+      "learning_rate": 1.9211552040949892e-05,
+      "loss": 1.1375,
+      "step": 2676
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8483035564422607,
+      "learning_rate": 1.9210828908078292e-05,
+      "loss": 1.0145,
+      "step": 2677
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8195797204971313,
+      "learning_rate": 1.921010545736671e-05,
+      "loss": 0.9977,
+      "step": 2678
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.084599494934082,
+      "learning_rate": 1.920938168884012e-05,
+      "loss": 1.0422,
+      "step": 2679
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1527047157287598,
+      "learning_rate": 1.9208657602523494e-05,
+      "loss": 1.1163,
+      "step": 2680
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9527807235717773,
+      "learning_rate": 1.920793319844181e-05,
+      "loss": 1.185,
+      "step": 2681
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3483762741088867,
+      "learning_rate": 1.920720847662008e-05,
+      "loss": 1.0799,
+      "step": 2682
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.065059185028076,
+      "learning_rate": 1.9206483437083302e-05,
+      "loss": 1.0359,
+      "step": 2683
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1101040840148926,
+      "learning_rate": 1.92057580798565e-05,
+      "loss": 1.0937,
+      "step": 2684
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8338804244995117,
+      "learning_rate": 1.92050324049647e-05,
+      "loss": 1.1392,
+      "step": 2685
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.127398729324341,
+      "learning_rate": 1.9204306412432947e-05,
+      "loss": 1.1117,
+      "step": 2686
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8999232053756714,
+      "learning_rate": 1.920358010228629e-05,
+      "loss": 1.0613,
+      "step": 2687
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1035425662994385,
+      "learning_rate": 1.9202853474549796e-05,
+      "loss": 1.0639,
+      "step": 2688
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0180768966674805,
+      "learning_rate": 1.9202126529248528e-05,
+      "loss": 1.0571,
+      "step": 2689
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0199365615844727,
+      "learning_rate": 1.9201399266407582e-05,
+      "loss": 1.0469,
+      "step": 2690
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0198137760162354,
+      "learning_rate": 1.9200671686052053e-05,
+      "loss": 1.1409,
+      "step": 2691
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8435190916061401,
+      "learning_rate": 1.9199943788207044e-05,
+      "loss": 1.1006,
+      "step": 2692
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.826055884361267,
+      "learning_rate": 1.9199215572897675e-05,
+      "loss": 1.0922,
+      "step": 2693
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0210392475128174,
+      "learning_rate": 1.919848704014907e-05,
+      "loss": 1.0804,
+      "step": 2694
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8186566829681396,
+      "learning_rate": 1.9197758189986372e-05,
+      "loss": 1.0592,
+      "step": 2695
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0916662216186523,
+      "learning_rate": 1.9197029022434734e-05,
+      "loss": 1.1399,
+      "step": 2696
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3249537944793701,
+      "learning_rate": 1.9196299537519314e-05,
+      "loss": 0.5983,
+      "step": 2697
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8300137519836426,
+      "learning_rate": 1.9195569735265288e-05,
+      "loss": 1.0982,
+      "step": 2698
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.977926254272461,
+      "learning_rate": 1.9194839615697836e-05,
+      "loss": 1.0305,
+      "step": 2699
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.085047960281372,
+      "learning_rate": 1.9194109178842155e-05,
+      "loss": 1.0664,
+      "step": 2700
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0125699043273926,
+      "learning_rate": 1.9193378424723446e-05,
+      "loss": 1.1197,
+      "step": 2701
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1143689155578613,
+      "learning_rate": 1.919264735336693e-05,
+      "loss": 1.1328,
+      "step": 2702
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8425114154815674,
+      "learning_rate": 1.919191596479783e-05,
+      "loss": 1.0824,
+      "step": 2703
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1642632484436035,
+      "learning_rate": 1.919118425904139e-05,
+      "loss": 0.61,
+      "step": 2704
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1337926387786865,
+      "learning_rate": 1.9190452236122856e-05,
+      "loss": 1.0865,
+      "step": 2705
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0472490787506104,
+      "learning_rate": 1.9189719896067487e-05,
+      "loss": 1.0851,
+      "step": 2706
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.060659646987915,
+      "learning_rate": 1.9188987238900554e-05,
+      "loss": 1.1239,
+      "step": 2707
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8582466840744019,
+      "learning_rate": 1.9188254264647338e-05,
+      "loss": 1.137,
+      "step": 2708
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0046050548553467,
+      "learning_rate": 1.9187520973333136e-05,
+      "loss": 1.058,
+      "step": 2709
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8665271997451782,
+      "learning_rate": 1.918678736498325e-05,
+      "loss": 1.0962,
+      "step": 2710
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2069183588027954,
+      "learning_rate": 1.9186053439622995e-05,
+      "loss": 0.6754,
+      "step": 2711
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7997500896453857,
+      "learning_rate": 1.9185319197277693e-05,
+      "loss": 1.1427,
+      "step": 2712
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.048140525817871,
+      "learning_rate": 1.9184584637972685e-05,
+      "loss": 1.0705,
+      "step": 2713
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8470405340194702,
+      "learning_rate": 1.9183849761733316e-05,
+      "loss": 1.0805,
+      "step": 2714
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9241894483566284,
+      "learning_rate": 1.9183114568584948e-05,
+      "loss": 1.0193,
+      "step": 2715
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9826548099517822,
+      "learning_rate": 1.918237905855295e-05,
+      "loss": 1.1926,
+      "step": 2716
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.166938543319702,
+      "learning_rate": 1.9181643231662695e-05,
+      "loss": 1.0218,
+      "step": 2717
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9338405132293701,
+      "learning_rate": 1.9180907087939588e-05,
+      "loss": 1.0929,
+      "step": 2718
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1510963439941406,
+      "learning_rate": 1.9180170627409014e-05,
+      "loss": 1.1066,
+      "step": 2719
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.833640217781067,
+      "learning_rate": 1.91794338500964e-05,
+      "loss": 1.0707,
+      "step": 2720
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0463311672210693,
+      "learning_rate": 1.9178696756027167e-05,
+      "loss": 1.0824,
+      "step": 2721
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8670568466186523,
+      "learning_rate": 1.9177959345226746e-05,
+      "loss": 1.1162,
+      "step": 2722
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8612425327301025,
+      "learning_rate": 1.9177221617720584e-05,
+      "loss": 1.0989,
+      "step": 2723
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9941425323486328,
+      "learning_rate": 1.9176483573534142e-05,
+      "loss": 1.0405,
+      "step": 2724
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9859180450439453,
+      "learning_rate": 1.917574521269289e-05,
+      "loss": 1.1413,
+      "step": 2725
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.81574547290802,
+      "learning_rate": 1.9175006535222293e-05,
+      "loss": 1.0651,
+      "step": 2726
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.118943452835083,
+      "learning_rate": 1.9174267541147856e-05,
+      "loss": 0.5386,
+      "step": 2727
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0453383922576904,
+      "learning_rate": 1.9173528230495072e-05,
+      "loss": 1.0662,
+      "step": 2728
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.864418864250183,
+      "learning_rate": 1.9172788603289453e-05,
+      "loss": 1.0685,
+      "step": 2729
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.033733606338501,
+      "learning_rate": 1.9172048659556523e-05,
+      "loss": 1.0565,
+      "step": 2730
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9367201328277588,
+      "learning_rate": 1.9171308399321817e-05,
+      "loss": 1.1281,
+      "step": 2731
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9522922039031982,
+      "learning_rate": 1.9170567822610872e-05,
+      "loss": 0.9853,
+      "step": 2732
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8821351528167725,
+      "learning_rate": 1.916982692944925e-05,
+      "loss": 1.0346,
+      "step": 2733
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8420169353485107,
+      "learning_rate": 1.9169085719862522e-05,
+      "loss": 1.0688,
+      "step": 2734
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.851286768913269,
+      "learning_rate": 1.916834419387625e-05,
+      "loss": 1.0329,
+      "step": 2735
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9359358549118042,
+      "learning_rate": 1.916760235151604e-05,
+      "loss": 1.0353,
+      "step": 2736
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.072796106338501,
+      "learning_rate": 1.9166860192807472e-05,
+      "loss": 1.0303,
+      "step": 2737
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.11775279045105,
+      "learning_rate": 1.9166117717776166e-05,
+      "loss": 1.0727,
+      "step": 2738
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9249122142791748,
+      "learning_rate": 1.9165374926447748e-05,
+      "loss": 1.1123,
+      "step": 2739
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.963181495666504,
+      "learning_rate": 1.9164631818847842e-05,
+      "loss": 1.1166,
+      "step": 2740
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7474950551986694,
+      "learning_rate": 1.916388839500209e-05,
+      "loss": 1.0789,
+      "step": 2741
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9499820470809937,
+      "learning_rate": 1.9163144654936148e-05,
+      "loss": 1.0622,
+      "step": 2742
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.911020278930664,
+      "learning_rate": 1.9162400598675682e-05,
+      "loss": 1.0566,
+      "step": 2743
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8519973754882812,
+      "learning_rate": 1.9161656226246362e-05,
+      "loss": 1.0631,
+      "step": 2744
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0770349502563477,
+      "learning_rate": 1.9160911537673884e-05,
+      "loss": 1.096,
+      "step": 2745
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9932708740234375,
+      "learning_rate": 1.9160166532983932e-05,
+      "loss": 1.1188,
+      "step": 2746
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9023292064666748,
+      "learning_rate": 1.9159421212202223e-05,
+      "loss": 1.0229,
+      "step": 2747
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0837292671203613,
+      "learning_rate": 1.9158675575354477e-05,
+      "loss": 1.0788,
+      "step": 2748
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.885361671447754,
+      "learning_rate": 1.9157929622466418e-05,
+      "loss": 1.0586,
+      "step": 2749
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.217736005783081,
+      "learning_rate": 1.9157183353563787e-05,
+      "loss": 1.0771,
+      "step": 2750
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.992760419845581,
+      "learning_rate": 1.9156436768672344e-05,
+      "loss": 1.0196,
+      "step": 2751
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0330630540847778,
+      "learning_rate": 1.9155689867817845e-05,
+      "loss": 0.5798,
+      "step": 2752
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1774089336395264,
+      "learning_rate": 1.9154942651026057e-05,
+      "loss": 1.0755,
+      "step": 2753
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9684513807296753,
+      "learning_rate": 1.9154195118322774e-05,
+      "loss": 1.0862,
+      "step": 2754
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.125251293182373,
+      "learning_rate": 1.9153447269733794e-05,
+      "loss": 1.0151,
+      "step": 2755
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9299724102020264,
+      "learning_rate": 1.9152699105284912e-05,
+      "loss": 1.0336,
+      "step": 2756
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.053964138031006,
+      "learning_rate": 1.9151950625001955e-05,
+      "loss": 1.0421,
+      "step": 2757
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9560880661010742,
+      "learning_rate": 1.9151201828910745e-05,
+      "loss": 1.0798,
+      "step": 2758
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.105623245239258,
+      "learning_rate": 1.915045271703712e-05,
+      "loss": 1.0867,
+      "step": 2759
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.883391261100769,
+      "learning_rate": 1.914970328940694e-05,
+      "loss": 1.028,
+      "step": 2760
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9990217685699463,
+      "learning_rate": 1.914895354604605e-05,
+      "loss": 1.0616,
+      "step": 2761
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2924914360046387,
+      "learning_rate": 1.9148203486980335e-05,
+      "loss": 1.1067,
+      "step": 2762
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.080641984939575,
+      "learning_rate": 1.914745311223567e-05,
+      "loss": 1.0492,
+      "step": 2763
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.065049409866333,
+      "learning_rate": 1.9146702421837952e-05,
+      "loss": 1.0923,
+      "step": 2764
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0786685943603516,
+      "learning_rate": 1.9145951415813084e-05,
+      "loss": 1.034,
+      "step": 2765
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9505268335342407,
+      "learning_rate": 1.9145200094186975e-05,
+      "loss": 1.0556,
+      "step": 2766
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1301217079162598,
+      "learning_rate": 1.914444845698556e-05,
+      "loss": 1.1408,
+      "step": 2767
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0864977836608887,
+      "learning_rate": 1.9143696504234777e-05,
+      "loss": 1.0588,
+      "step": 2768
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9977463483810425,
+      "learning_rate": 1.9142944235960566e-05,
+      "loss": 1.0589,
+      "step": 2769
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0642709732055664,
+      "learning_rate": 1.914219165218889e-05,
+      "loss": 1.1203,
+      "step": 2770
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8074480295181274,
+      "learning_rate": 1.9141438752945717e-05,
+      "loss": 1.0095,
+      "step": 2771
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1132543087005615,
+      "learning_rate": 1.9140685538257027e-05,
+      "loss": 1.1107,
+      "step": 2772
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1416449546813965,
+      "learning_rate": 1.9139932008148816e-05,
+      "loss": 1.1041,
+      "step": 2773
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9659407138824463,
+      "learning_rate": 1.913917816264708e-05,
+      "loss": 1.1376,
+      "step": 2774
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9418716430664062,
+      "learning_rate": 1.913842400177784e-05,
+      "loss": 1.0105,
+      "step": 2775
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9128941297531128,
+      "learning_rate": 1.9137669525567108e-05,
+      "loss": 1.097,
+      "step": 2776
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9174975156784058,
+      "learning_rate": 1.913691473404093e-05,
+      "loss": 1.0629,
+      "step": 2777
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.08526349067688,
+      "learning_rate": 1.9136159627225342e-05,
+      "loss": 1.1187,
+      "step": 2778
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.918385624885559,
+      "learning_rate": 1.9135404205146414e-05,
+      "loss": 1.087,
+      "step": 2779
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8826457262039185,
+      "learning_rate": 1.9134648467830198e-05,
+      "loss": 0.9506,
+      "step": 2780
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9759894609451294,
+      "learning_rate": 1.9133892415302783e-05,
+      "loss": 1.0828,
+      "step": 2781
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.948956847190857,
+      "learning_rate": 1.9133136047590258e-05,
+      "loss": 0.9818,
+      "step": 2782
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.021775484085083,
+      "learning_rate": 1.913237936471872e-05,
+      "loss": 1.1329,
+      "step": 2783
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8469105958938599,
+      "learning_rate": 1.9131622366714277e-05,
+      "loss": 1.1223,
+      "step": 2784
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0182039737701416,
+      "learning_rate": 1.9130865053603055e-05,
+      "loss": 1.0881,
+      "step": 2785
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9165613651275635,
+      "learning_rate": 1.9130107425411186e-05,
+      "loss": 1.0766,
+      "step": 2786
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1084506511688232,
+      "learning_rate": 1.9129349482164815e-05,
+      "loss": 1.0226,
+      "step": 2787
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8359850645065308,
+      "learning_rate": 1.9128591223890094e-05,
+      "loss": 0.9454,
+      "step": 2788
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9818309545516968,
+      "learning_rate": 1.912783265061319e-05,
+      "loss": 1.1711,
+      "step": 2789
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.741380214691162,
+      "learning_rate": 1.912707376236028e-05,
+      "loss": 1.0016,
+      "step": 2790
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9307808876037598,
+      "learning_rate": 1.912631455915755e-05,
+      "loss": 1.0517,
+      "step": 2791
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8927243947982788,
+      "learning_rate": 1.9125555041031196e-05,
+      "loss": 1.1103,
+      "step": 2792
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9926817417144775,
+      "learning_rate": 1.912479520800743e-05,
+      "loss": 1.1048,
+      "step": 2793
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8304252624511719,
+      "learning_rate": 1.912403506011247e-05,
+      "loss": 0.9985,
+      "step": 2794
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.867504358291626,
+      "learning_rate": 1.9123274597372547e-05,
+      "loss": 1.1196,
+      "step": 2795
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9199800491333008,
+      "learning_rate": 1.91225138198139e-05,
+      "loss": 1.1098,
+      "step": 2796
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7561461925506592,
+      "learning_rate": 1.9121752727462787e-05,
+      "loss": 0.9618,
+      "step": 2797
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8149597644805908,
+      "learning_rate": 1.912099132034547e-05,
+      "loss": 1.0378,
+      "step": 2798
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0402047634124756,
+      "learning_rate": 1.9120229598488218e-05,
+      "loss": 1.1333,
+      "step": 2799
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7843079566955566,
+      "learning_rate": 1.911946756191732e-05,
+      "loss": 1.1115,
+      "step": 2800
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1256554126739502,
+      "learning_rate": 1.9118705210659067e-05,
+      "loss": 0.618,
+      "step": 2801
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1451330184936523,
+      "learning_rate": 1.911794254473977e-05,
+      "loss": 1.0175,
+      "step": 2802
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.235419750213623,
+      "learning_rate": 1.911717956418575e-05,
+      "loss": 1.0795,
+      "step": 2803
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8935601711273193,
+      "learning_rate": 1.911641626902333e-05,
+      "loss": 0.9933,
+      "step": 2804
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9885404109954834,
+      "learning_rate": 1.911565265927885e-05,
+      "loss": 1.0476,
+      "step": 2805
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0454165935516357,
+      "learning_rate": 1.911488873497866e-05,
+      "loss": 0.9931,
+      "step": 2806
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9443235397338867,
+      "learning_rate": 1.911412449614912e-05,
+      "loss": 1.0239,
+      "step": 2807
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.016200304031372,
+      "learning_rate": 1.9113359942816602e-05,
+      "loss": 1.1213,
+      "step": 2808
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9532594680786133,
+      "learning_rate": 1.9112595075007492e-05,
+      "loss": 1.0586,
+      "step": 2809
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.084554433822632,
+      "learning_rate": 1.911182989274818e-05,
+      "loss": 0.9784,
+      "step": 2810
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.808603048324585,
+      "learning_rate": 1.911106439606507e-05,
+      "loss": 1.1076,
+      "step": 2811
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8961806297302246,
+      "learning_rate": 1.911029858498458e-05,
+      "loss": 1.0701,
+      "step": 2812
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1251559257507324,
+      "learning_rate": 1.9109532459533136e-05,
+      "loss": 1.0397,
+      "step": 2813
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9431532621383667,
+      "learning_rate": 1.9108766019737168e-05,
+      "loss": 1.0586,
+      "step": 2814
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0151922702789307,
+      "learning_rate": 1.9107999265623133e-05,
+      "loss": 1.0769,
+      "step": 2815
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8579295873641968,
+      "learning_rate": 1.9107232197217483e-05,
+      "loss": 1.0889,
+      "step": 2816
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0122504234313965,
+      "learning_rate": 1.9106464814546695e-05,
+      "loss": 1.0166,
+      "step": 2817
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.951866626739502,
+      "learning_rate": 1.910569711763724e-05,
+      "loss": 1.1338,
+      "step": 2818
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7444969415664673,
+      "learning_rate": 1.9104929106515616e-05,
+      "loss": 1.0497,
+      "step": 2819
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7327609062194824,
+      "learning_rate": 1.910416078120832e-05,
+      "loss": 1.0935,
+      "step": 2820
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.132636070251465,
+      "learning_rate": 1.9103392141741865e-05,
+      "loss": 1.074,
+      "step": 2821
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9666560888290405,
+      "learning_rate": 1.910262318814278e-05,
+      "loss": 1.1461,
+      "step": 2822
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8750109672546387,
+      "learning_rate": 1.9101853920437594e-05,
+      "loss": 0.9635,
+      "step": 2823
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0285205841064453,
+      "learning_rate": 1.9101084338652855e-05,
+      "loss": 1.1015,
+      "step": 2824
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9769545793533325,
+      "learning_rate": 1.910031444281512e-05,
+      "loss": 1.1294,
+      "step": 2825
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1257152557373047,
+      "learning_rate": 1.909954423295095e-05,
+      "loss": 1.0726,
+      "step": 2826
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.031496047973633,
+      "learning_rate": 1.909877370908693e-05,
+      "loss": 1.0683,
+      "step": 2827
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.915421962738037,
+      "learning_rate": 1.9098002871249644e-05,
+      "loss": 1.072,
+      "step": 2828
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.938776969909668,
+      "learning_rate": 1.9097231719465695e-05,
+      "loss": 1.0576,
+      "step": 2829
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.993424415588379,
+      "learning_rate": 1.909646025376169e-05,
+      "loss": 1.0303,
+      "step": 2830
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.067265033721924,
+      "learning_rate": 1.9095688474164254e-05,
+      "loss": 1.0114,
+      "step": 2831
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9131224155426025,
+      "learning_rate": 1.9094916380700015e-05,
+      "loss": 1.0502,
+      "step": 2832
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1012070178985596,
+      "learning_rate": 1.9094143973395614e-05,
+      "loss": 1.1246,
+      "step": 2833
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9724221229553223,
+      "learning_rate": 1.909337125227771e-05,
+      "loss": 1.1616,
+      "step": 2834
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8480886220932007,
+      "learning_rate": 1.909259821737297e-05,
+      "loss": 1.0921,
+      "step": 2835
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9709773063659668,
+      "learning_rate": 1.909182486870806e-05,
+      "loss": 0.9865,
+      "step": 2836
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1934001445770264,
+      "learning_rate": 1.9091051206309674e-05,
+      "loss": 0.5629,
+      "step": 2837
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.012815475463867,
+      "learning_rate": 1.9090277230204503e-05,
+      "loss": 1.0509,
+      "step": 2838
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9511667490005493,
+      "learning_rate": 1.9089502940419258e-05,
+      "loss": 1.1037,
+      "step": 2839
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8742932081222534,
+      "learning_rate": 1.9088728336980656e-05,
+      "loss": 1.0173,
+      "step": 2840
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7490102052688599,
+      "learning_rate": 1.9087953419915427e-05,
+      "loss": 0.9781,
+      "step": 2841
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.992408275604248,
+      "learning_rate": 1.9087178189250314e-05,
+      "loss": 1.1066,
+      "step": 2842
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0442935228347778,
+      "learning_rate": 1.9086402645012065e-05,
+      "loss": 0.6306,
+      "step": 2843
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0599634647369385,
+      "learning_rate": 1.9085626787227444e-05,
+      "loss": 1.0428,
+      "step": 2844
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9415106773376465,
+      "learning_rate": 1.9084850615923217e-05,
+      "loss": 1.0039,
+      "step": 2845
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.987179160118103,
+      "learning_rate": 1.908407413112618e-05,
+      "loss": 1.0064,
+      "step": 2846
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0024447441101074,
+      "learning_rate": 1.9083297332863114e-05,
+      "loss": 1.0419,
+      "step": 2847
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8485029935836792,
+      "learning_rate": 1.9082520221160835e-05,
+      "loss": 1.07,
+      "step": 2848
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9870679378509521,
+      "learning_rate": 1.908174279604615e-05,
+      "loss": 0.9968,
+      "step": 2849
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9794743061065674,
+      "learning_rate": 1.9080965057545894e-05,
+      "loss": 1.1163,
+      "step": 2850
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.929158329963684,
+      "learning_rate": 1.9080187005686896e-05,
+      "loss": 1.0327,
+      "step": 2851
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0727713108062744,
+      "learning_rate": 1.9079408640496012e-05,
+      "loss": 1.0528,
+      "step": 2852
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9584839344024658,
+      "learning_rate": 1.90786299620001e-05,
+      "loss": 1.1685,
+      "step": 2853
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9692672491073608,
+      "learning_rate": 1.9077850970226025e-05,
+      "loss": 0.9874,
+      "step": 2854
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.855149745941162,
+      "learning_rate": 1.907707166520067e-05,
+      "loss": 1.1424,
+      "step": 2855
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.854702353477478,
+      "learning_rate": 1.907629204695093e-05,
+      "loss": 1.0632,
+      "step": 2856
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.853702187538147,
+      "learning_rate": 1.9075512115503707e-05,
+      "loss": 1.0292,
+      "step": 2857
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0127134323120117,
+      "learning_rate": 1.9074731870885907e-05,
+      "loss": 1.0929,
+      "step": 2858
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.887139081954956,
+      "learning_rate": 1.9073951313124462e-05,
+      "loss": 0.9609,
+      "step": 2859
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.975973129272461,
+      "learning_rate": 1.9073170442246304e-05,
+      "loss": 1.0774,
+      "step": 2860
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.976212978363037,
+      "learning_rate": 1.9072389258278378e-05,
+      "loss": 1.0252,
+      "step": 2861
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8501232862472534,
+      "learning_rate": 1.9071607761247644e-05,
+      "loss": 1.0484,
+      "step": 2862
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8229377269744873,
+      "learning_rate": 1.9070825951181065e-05,
+      "loss": 1.0586,
+      "step": 2863
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1713542938232422,
+      "learning_rate": 1.9070043828105616e-05,
+      "loss": 0.5867,
+      "step": 2864
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0396790504455566,
+      "learning_rate": 1.90692613920483e-05,
+      "loss": 1.1764,
+      "step": 2865
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.997402548789978,
+      "learning_rate": 1.9068478643036102e-05,
+      "loss": 1.0735,
+      "step": 2866
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0114099979400635,
+      "learning_rate": 1.9067695581096037e-05,
+      "loss": 1.107,
+      "step": 2867
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8693817853927612,
+      "learning_rate": 1.906691220625513e-05,
+      "loss": 1.0884,
+      "step": 2868
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9016669988632202,
+      "learning_rate": 1.9066128518540408e-05,
+      "loss": 1.0759,
+      "step": 2869
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9285945892333984,
+      "learning_rate": 1.906534451797892e-05,
+      "loss": 1.1167,
+      "step": 2870
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.843967318534851,
+      "learning_rate": 1.906456020459771e-05,
+      "loss": 1.0286,
+      "step": 2871
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9659533500671387,
+      "learning_rate": 1.906377557842385e-05,
+      "loss": 1.1093,
+      "step": 2872
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1897130012512207,
+      "learning_rate": 1.9062990639484416e-05,
+      "loss": 1.0701,
+      "step": 2873
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.050063371658325,
+      "learning_rate": 1.906220538780649e-05,
+      "loss": 1.044,
+      "step": 2874
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9924061298370361,
+      "learning_rate": 1.906141982341717e-05,
+      "loss": 1.1113,
+      "step": 2875
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7729718685150146,
+      "learning_rate": 1.906063394634356e-05,
+      "loss": 1.053,
+      "step": 2876
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7325552701950073,
+      "learning_rate": 1.905984775661279e-05,
+      "loss": 1.0026,
+      "step": 2877
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8477683067321777,
+      "learning_rate": 1.9059061254251978e-05,
+      "loss": 1.0763,
+      "step": 2878
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.196218252182007,
+      "learning_rate": 1.9058274439288267e-05,
+      "loss": 1.1098,
+      "step": 2879
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9315710067749023,
+      "learning_rate": 1.905748731174881e-05,
+      "loss": 1.1487,
+      "step": 2880
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.876016616821289,
+      "learning_rate": 1.9056699871660763e-05,
+      "loss": 1.0634,
+      "step": 2881
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.927635908126831,
+      "learning_rate": 1.9055912119051305e-05,
+      "loss": 1.0757,
+      "step": 2882
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7561267614364624,
+      "learning_rate": 1.905512405394762e-05,
+      "loss": 1.0485,
+      "step": 2883
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1301265954971313,
+      "learning_rate": 1.9054335676376893e-05,
+      "loss": 0.5229,
+      "step": 2884
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.156092643737793,
+      "learning_rate": 1.9053546986366332e-05,
+      "loss": 1.0645,
+      "step": 2885
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2195944786071777,
+      "learning_rate": 1.9052757983943162e-05,
+      "loss": 1.0956,
+      "step": 2886
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9298347234725952,
+      "learning_rate": 1.9051968669134597e-05,
+      "loss": 1.1206,
+      "step": 2887
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.022909164428711,
+      "learning_rate": 1.905117904196788e-05,
+      "loss": 0.5451,
+      "step": 2888
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7126446962356567,
+      "learning_rate": 1.9050389102470258e-05,
+      "loss": 1.0747,
+      "step": 2889
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.088958978652954,
+      "learning_rate": 1.9049598850668988e-05,
+      "loss": 1.0442,
+      "step": 2890
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.002758264541626,
+      "learning_rate": 1.9048808286591343e-05,
+      "loss": 1.0963,
+      "step": 2891
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9130730628967285,
+      "learning_rate": 1.90480174102646e-05,
+      "loss": 1.1149,
+      "step": 2892
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9090899229049683,
+      "learning_rate": 1.904722622171605e-05,
+      "loss": 1.1241,
+      "step": 2893
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.970346212387085,
+      "learning_rate": 1.9046434720973e-05,
+      "loss": 1.0268,
+      "step": 2894
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8828133344650269,
+      "learning_rate": 1.904564290806275e-05,
+      "loss": 0.985,
+      "step": 2895
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0196754932403564,
+      "learning_rate": 1.9044850783012636e-05,
+      "loss": 1.0554,
+      "step": 2896
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8040275573730469,
+      "learning_rate": 1.9044058345849988e-05,
+      "loss": 1.0266,
+      "step": 2897
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9745535850524902,
+      "learning_rate": 1.9043265596602146e-05,
+      "loss": 1.0612,
+      "step": 2898
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.021299362182617,
+      "learning_rate": 1.9042472535296474e-05,
+      "loss": 1.0923,
+      "step": 2899
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1055734157562256,
+      "learning_rate": 1.904167916196033e-05,
+      "loss": 1.0128,
+      "step": 2900
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8815861940383911,
+      "learning_rate": 1.9040885476621097e-05,
+      "loss": 1.0756,
+      "step": 2901
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8467283248901367,
+      "learning_rate": 1.9040091479306163e-05,
+      "loss": 1.0524,
+      "step": 2902
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5922529697418213,
+      "learning_rate": 1.9039297170042922e-05,
+      "loss": 1.0293,
+      "step": 2903
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9232429265975952,
+      "learning_rate": 1.9038502548858786e-05,
+      "loss": 1.0476,
+      "step": 2904
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.830834984779358,
+      "learning_rate": 1.9037707615781177e-05,
+      "loss": 1.0375,
+      "step": 2905
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.071190118789673,
+      "learning_rate": 1.9036912370837523e-05,
+      "loss": 1.042,
+      "step": 2906
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2560254335403442,
+      "learning_rate": 1.9036116814055264e-05,
+      "loss": 0.5797,
+      "step": 2907
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7809886932373047,
+      "learning_rate": 1.903532094546186e-05,
+      "loss": 1.0051,
+      "step": 2908
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.940935492515564,
+      "learning_rate": 1.9034524765084764e-05,
+      "loss": 1.0706,
+      "step": 2909
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9670809507369995,
+      "learning_rate": 1.9033728272951458e-05,
+      "loss": 1.078,
+      "step": 2910
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.199079751968384,
+      "learning_rate": 1.9032931469089425e-05,
+      "loss": 1.0578,
+      "step": 2911
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.948774814605713,
+      "learning_rate": 1.9032134353526158e-05,
+      "loss": 0.9922,
+      "step": 2912
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8382983207702637,
+      "learning_rate": 1.9031336926289167e-05,
+      "loss": 1.0867,
+      "step": 2913
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9122934341430664,
+      "learning_rate": 1.9030539187405962e-05,
+      "loss": 1.0126,
+      "step": 2914
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.012712001800537,
+      "learning_rate": 1.902974113690408e-05,
+      "loss": 1.093,
+      "step": 2915
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.950904369354248,
+      "learning_rate": 1.902894277481105e-05,
+      "loss": 1.0707,
+      "step": 2916
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1292970180511475,
+      "learning_rate": 1.902814410115443e-05,
+      "loss": 0.6413,
+      "step": 2917
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3109323978424072,
+      "learning_rate": 1.9027345115961778e-05,
+      "loss": 1.0625,
+      "step": 2918
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.045332670211792,
+      "learning_rate": 1.902654581926066e-05,
+      "loss": 1.0038,
+      "step": 2919
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.892066240310669,
+      "learning_rate": 1.9025746211078658e-05,
+      "loss": 1.0534,
+      "step": 2920
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.176978588104248,
+      "learning_rate": 1.902494629144337e-05,
+      "loss": 1.076,
+      "step": 2921
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.276561975479126,
+      "learning_rate": 1.9024146060382396e-05,
+      "loss": 1.0489,
+      "step": 2922
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8345601558685303,
+      "learning_rate": 1.902334551792335e-05,
+      "loss": 1.0183,
+      "step": 2923
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9069886207580566,
+      "learning_rate": 1.9022544664093854e-05,
+      "loss": 1.1266,
+      "step": 2924
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0337042808532715,
+      "learning_rate": 1.9021743498921544e-05,
+      "loss": 1.0125,
+      "step": 2925
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9993294477462769,
+      "learning_rate": 1.9020942022434072e-05,
+      "loss": 1.0337,
+      "step": 2926
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9678328037261963,
+      "learning_rate": 1.9020140234659084e-05,
+      "loss": 1.071,
+      "step": 2927
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.171757221221924,
+      "learning_rate": 1.9019338135624256e-05,
+      "loss": 1.0099,
+      "step": 2928
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8959630727767944,
+      "learning_rate": 1.901853572535726e-05,
+      "loss": 1.032,
+      "step": 2929
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0473315715789795,
+      "learning_rate": 1.9017733003885794e-05,
+      "loss": 1.1325,
+      "step": 2930
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8315461874008179,
+      "learning_rate": 1.901692997123755e-05,
+      "loss": 1.0791,
+      "step": 2931
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9948164224624634,
+      "learning_rate": 1.901612662744024e-05,
+      "loss": 1.0553,
+      "step": 2932
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9197337627410889,
+      "learning_rate": 1.9015322972521587e-05,
+      "loss": 0.961,
+      "step": 2933
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8900346755981445,
+      "learning_rate": 1.901451900650932e-05,
+      "loss": 1.0264,
+      "step": 2934
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4162036180496216,
+      "learning_rate": 1.9013714729431183e-05,
+      "loss": 0.6802,
+      "step": 2935
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1681153774261475,
+      "learning_rate": 1.9012910141314928e-05,
+      "loss": 1.1044,
+      "step": 2936
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.876365065574646,
+      "learning_rate": 1.9012105242188323e-05,
+      "loss": 1.0254,
+      "step": 2937
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9956624507904053,
+      "learning_rate": 1.901130003207914e-05,
+      "loss": 1.0748,
+      "step": 2938
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0879809856414795,
+      "learning_rate": 1.9010494511015164e-05,
+      "loss": 1.123,
+      "step": 2939
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9485974311828613,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 1.1134,
+      "step": 2940
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7256534099578857,
+      "learning_rate": 1.9008882536134035e-05,
+      "loss": 0.9859,
+      "step": 2941
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.959047794342041,
+      "learning_rate": 1.9008076082372504e-05,
+      "loss": 1.0945,
+      "step": 2942
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3929054737091064,
+      "learning_rate": 1.9007269317767428e-05,
+      "loss": 0.5858,
+      "step": 2943
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.03013014793396,
+      "learning_rate": 1.9006462242346654e-05,
+      "loss": 1.0717,
+      "step": 2944
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8899321556091309,
+      "learning_rate": 1.900565485613802e-05,
+      "loss": 1.0111,
+      "step": 2945
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.825768232345581,
+      "learning_rate": 1.9004847159169397e-05,
+      "loss": 1.0079,
+      "step": 2946
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8760653734207153,
+      "learning_rate": 1.9004039151468654e-05,
+      "loss": 1.0923,
+      "step": 2947
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9713869094848633,
+      "learning_rate": 1.900323083306367e-05,
+      "loss": 0.9865,
+      "step": 2948
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1316449642181396,
+      "learning_rate": 1.9002422203982342e-05,
+      "loss": 1.1449,
+      "step": 2949
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8565106391906738,
+      "learning_rate": 1.9001613264252565e-05,
+      "loss": 1.0821,
+      "step": 2950
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7604048252105713,
+      "learning_rate": 1.9000804013902264e-05,
+      "loss": 1.0402,
+      "step": 2951
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0256030559539795,
+      "learning_rate": 1.8999994452959357e-05,
+      "loss": 1.0772,
+      "step": 2952
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9991750717163086,
+      "learning_rate": 1.8999184581451783e-05,
+      "loss": 1.0487,
+      "step": 2953
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0230250358581543,
+      "learning_rate": 1.8998374399407486e-05,
+      "loss": 1.0922,
+      "step": 2954
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9650601148605347,
+      "learning_rate": 1.8997563906854427e-05,
+      "loss": 1.1754,
+      "step": 2955
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9861899614334106,
+      "learning_rate": 1.899675310382057e-05,
+      "loss": 1.0609,
+      "step": 2956
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8957674503326416,
+      "learning_rate": 1.8995941990333894e-05,
+      "loss": 1.0995,
+      "step": 2957
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7356202602386475,
+      "learning_rate": 1.899513056642239e-05,
+      "loss": 1.0554,
+      "step": 2958
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.076918601989746,
+      "learning_rate": 1.899431883211406e-05,
+      "loss": 1.0885,
+      "step": 2959
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.065945863723755,
+      "learning_rate": 1.8993506787436905e-05,
+      "loss": 1.056,
+      "step": 2960
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.993909239768982,
+      "learning_rate": 1.8992694432418958e-05,
+      "loss": 1.0548,
+      "step": 2961
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0165367126464844,
+      "learning_rate": 1.8991881767088245e-05,
+      "loss": 1.0799,
+      "step": 2962
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0284125804901123,
+      "learning_rate": 1.899106879147281e-05,
+      "loss": 1.0216,
+      "step": 2963
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9928864240646362,
+      "learning_rate": 1.8990255505600706e-05,
+      "loss": 1.0449,
+      "step": 2964
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0046610832214355,
+      "learning_rate": 1.8989441909499998e-05,
+      "loss": 1.1715,
+      "step": 2965
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1145431995391846,
+      "learning_rate": 1.8988628003198762e-05,
+      "loss": 1.0832,
+      "step": 2966
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1668436527252197,
+      "learning_rate": 1.898781378672508e-05,
+      "loss": 1.167,
+      "step": 2967
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8951785564422607,
+      "learning_rate": 1.8986999260107054e-05,
+      "loss": 1.0043,
+      "step": 2968
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2734827995300293,
+      "learning_rate": 1.8986184423372784e-05,
+      "loss": 0.6073,
+      "step": 2969
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.136033296585083,
+      "learning_rate": 1.898536927655039e-05,
+      "loss": 1.0259,
+      "step": 2970
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9802865982055664,
+      "learning_rate": 1.898455381966801e-05,
+      "loss": 1.0753,
+      "step": 2971
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7699776887893677,
+      "learning_rate": 1.8983738052753767e-05,
+      "loss": 0.9932,
+      "step": 2972
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9671918153762817,
+      "learning_rate": 1.898292197583582e-05,
+      "loss": 1.0428,
+      "step": 2973
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.134178876876831,
+      "learning_rate": 1.8982105588942333e-05,
+      "loss": 1.0647,
+      "step": 2974
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9318534135818481,
+      "learning_rate": 1.8981288892101468e-05,
+      "loss": 1.0227,
+      "step": 2975
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9363925457000732,
+      "learning_rate": 1.8980471885341415e-05,
+      "loss": 1.156,
+      "step": 2976
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8565434217453003,
+      "learning_rate": 1.8979654568690363e-05,
+      "loss": 1.0217,
+      "step": 2977
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8469220399856567,
+      "learning_rate": 1.8978836942176513e-05,
+      "loss": 1.1264,
+      "step": 2978
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0001449584960938,
+      "learning_rate": 1.897801900582808e-05,
+      "loss": 1.0021,
+      "step": 2979
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.389589786529541,
+      "learning_rate": 1.8977200759673295e-05,
+      "loss": 1.0617,
+      "step": 2980
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0419857501983643,
+      "learning_rate": 1.8976382203740383e-05,
+      "loss": 1.1454,
+      "step": 2981
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1098897457122803,
+      "learning_rate": 1.8975563338057602e-05,
+      "loss": 0.6071,
+      "step": 2982
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8723129034042358,
+      "learning_rate": 1.89747441626532e-05,
+      "loss": 1.0691,
+      "step": 2983
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0793051719665527,
+      "learning_rate": 1.8973924677555448e-05,
+      "loss": 1.1307,
+      "step": 2984
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0163912773132324,
+      "learning_rate": 1.897310488279262e-05,
+      "loss": 1.023,
+      "step": 2985
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.043722629547119,
+      "learning_rate": 1.897228477839301e-05,
+      "loss": 1.1534,
+      "step": 2986
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0760319232940674,
+      "learning_rate": 1.897146436438491e-05,
+      "loss": 1.0985,
+      "step": 2987
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8473470211029053,
+      "learning_rate": 1.8970643640796642e-05,
+      "loss": 1.0897,
+      "step": 2988
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1732301712036133,
+      "learning_rate": 1.896982260765652e-05,
+      "loss": 1.0501,
+      "step": 2989
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3138396739959717,
+      "learning_rate": 1.8969001264992872e-05,
+      "loss": 1.1972,
+      "step": 2990
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1274964809417725,
+      "learning_rate": 1.8968179612834048e-05,
+      "loss": 1.1422,
+      "step": 2991
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7372972965240479,
+      "learning_rate": 1.8967357651208396e-05,
+      "loss": 1.0065,
+      "step": 2992
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9205377101898193,
+      "learning_rate": 1.8966535380144278e-05,
+      "loss": 1.0313,
+      "step": 2993
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1258206367492676,
+      "learning_rate": 1.8965712799670077e-05,
+      "loss": 0.6516,
+      "step": 2994
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.858955979347229,
+      "learning_rate": 1.8964889909814167e-05,
+      "loss": 0.9919,
+      "step": 2995
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0021536350250244,
+      "learning_rate": 1.8964066710604953e-05,
+      "loss": 1.1322,
+      "step": 2996
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.010690450668335,
+      "learning_rate": 1.8963243202070832e-05,
+      "loss": 0.9942,
+      "step": 2997
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7876743078231812,
+      "learning_rate": 1.896241938424023e-05,
+      "loss": 1.1047,
+      "step": 2998
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7891483306884766,
+      "learning_rate": 1.896159525714157e-05,
+      "loss": 1.0348,
+      "step": 2999
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9011098146438599,
+      "learning_rate": 1.8960770820803286e-05,
+      "loss": 1.0941,
+      "step": 3000
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.029694080352783,
+      "learning_rate": 1.8959946075253833e-05,
+      "loss": 1.0809,
+      "step": 3001
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.061575412750244,
+      "learning_rate": 1.8959121020521674e-05,
+      "loss": 1.1,
+      "step": 3002
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0008037090301514,
+      "learning_rate": 1.8958295656635273e-05,
+      "loss": 1.0101,
+      "step": 3003
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9375660419464111,
+      "learning_rate": 1.8957469983623113e-05,
+      "loss": 1.0455,
+      "step": 3004
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.990166425704956,
+      "learning_rate": 1.8956644001513686e-05,
+      "loss": 1.0524,
+      "step": 3005
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8718637228012085,
+      "learning_rate": 1.8955817710335493e-05,
+      "loss": 1.0437,
+      "step": 3006
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9028778076171875,
+      "learning_rate": 1.895499111011705e-05,
+      "loss": 1.1123,
+      "step": 3007
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9340474605560303,
+      "learning_rate": 1.8954164200886874e-05,
+      "loss": 1.0475,
+      "step": 3008
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.953302264213562,
+      "learning_rate": 1.8953336982673506e-05,
+      "loss": 1.0313,
+      "step": 3009
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.781287670135498,
+      "learning_rate": 1.895250945550549e-05,
+      "loss": 0.9338,
+      "step": 3010
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8158373832702637,
+      "learning_rate": 1.895168161941138e-05,
+      "loss": 1.027,
+      "step": 3011
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.973016381263733,
+      "learning_rate": 1.8950853474419745e-05,
+      "loss": 1.0986,
+      "step": 3012
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8059771060943604,
+      "learning_rate": 1.8950025020559155e-05,
+      "loss": 1.0776,
+      "step": 3013
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9360907077789307,
+      "learning_rate": 1.8949196257858205e-05,
+      "loss": 1.017,
+      "step": 3014
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0138602256774902,
+      "learning_rate": 1.894836718634549e-05,
+      "loss": 1.0888,
+      "step": 3015
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1574063301086426,
+      "learning_rate": 1.894753780604962e-05,
+      "loss": 0.9724,
+      "step": 3016
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8222215175628662,
+      "learning_rate": 1.8946708116999216e-05,
+      "loss": 1.0829,
+      "step": 3017
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.050697088241577,
+      "learning_rate": 1.8945878119222904e-05,
+      "loss": 1.0059,
+      "step": 3018
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9466689825057983,
+      "learning_rate": 1.894504781274933e-05,
+      "loss": 1.1105,
+      "step": 3019
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.949623942375183,
+      "learning_rate": 1.8944217197607142e-05,
+      "loss": 1.1131,
+      "step": 3020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9778258800506592,
+      "learning_rate": 1.8943386273825e-05,
+      "loss": 1.0343,
+      "step": 3021
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0692856311798096,
+      "learning_rate": 1.8942555041431584e-05,
+      "loss": 1.1159,
+      "step": 3022
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9074277877807617,
+      "learning_rate": 1.8941723500455575e-05,
+      "loss": 1.0707,
+      "step": 3023
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.133254289627075,
+      "learning_rate": 1.8940891650925662e-05,
+      "loss": 1.0496,
+      "step": 3024
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9440248012542725,
+      "learning_rate": 1.8940059492870552e-05,
+      "loss": 1.0953,
+      "step": 3025
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.788736343383789,
+      "learning_rate": 1.8939227026318964e-05,
+      "loss": 1.0241,
+      "step": 3026
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8978806734085083,
+      "learning_rate": 1.8938394251299624e-05,
+      "loss": 1.105,
+      "step": 3027
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8875573873519897,
+      "learning_rate": 1.8937561167841262e-05,
+      "loss": 1.0331,
+      "step": 3028
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9286848306655884,
+      "learning_rate": 1.8936727775972638e-05,
+      "loss": 1.1025,
+      "step": 3029
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0652108192443848,
+      "learning_rate": 1.8935894075722495e-05,
+      "loss": 1.121,
+      "step": 3030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8857553005218506,
+      "learning_rate": 1.893506006711961e-05,
+      "loss": 1.0003,
+      "step": 3031
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.961744785308838,
+      "learning_rate": 1.8934225750192762e-05,
+      "loss": 1.0014,
+      "step": 3032
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9467982053756714,
+      "learning_rate": 1.8933391124970742e-05,
+      "loss": 1.1093,
+      "step": 3033
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.111811637878418,
+      "learning_rate": 1.8932556191482347e-05,
+      "loss": 0.5851,
+      "step": 3034
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.873814344406128,
+      "learning_rate": 1.893172094975639e-05,
+      "loss": 1.0432,
+      "step": 3035
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9854656457901,
+      "learning_rate": 1.8930885399821693e-05,
+      "loss": 1.017,
+      "step": 3036
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.216555118560791,
+      "learning_rate": 1.8930049541707088e-05,
+      "loss": 0.6279,
+      "step": 3037
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9797500371932983,
+      "learning_rate": 1.892921337544142e-05,
+      "loss": 1.153,
+      "step": 3038
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0771334171295166,
+      "learning_rate": 1.892837690105354e-05,
+      "loss": 1.1322,
+      "step": 3039
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0623233318328857,
+      "learning_rate": 1.8927540118572314e-05,
+      "loss": 1.1078,
+      "step": 3040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0152978897094727,
+      "learning_rate": 1.8926703028026617e-05,
+      "loss": 1.1086,
+      "step": 3041
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4055519104003906,
+      "learning_rate": 1.892586562944533e-05,
+      "loss": 1.0539,
+      "step": 3042
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.848069667816162,
+      "learning_rate": 1.8925027922857358e-05,
+      "loss": 1.0976,
+      "step": 3043
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.042470932006836,
+      "learning_rate": 1.89241899082916e-05,
+      "loss": 0.9783,
+      "step": 3044
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8244280815124512,
+      "learning_rate": 1.892335158577698e-05,
+      "loss": 1.0713,
+      "step": 3045
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8841606378555298,
+      "learning_rate": 1.8922512955342423e-05,
+      "loss": 0.9789,
+      "step": 3046
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7972135543823242,
+      "learning_rate": 1.892167401701687e-05,
+      "loss": 1.0037,
+      "step": 3047
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9971961975097656,
+      "learning_rate": 1.8920834770829262e-05,
+      "loss": 1.012,
+      "step": 3048
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1417475938796997,
+      "learning_rate": 1.891999521680857e-05,
+      "loss": 0.6142,
+      "step": 3049
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9814648628234863,
+      "learning_rate": 1.891915535498376e-05,
+      "loss": 1.1373,
+      "step": 3050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8645546436309814,
+      "learning_rate": 1.8918315185383812e-05,
+      "loss": 1.0334,
+      "step": 3051
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1152050495147705,
+      "learning_rate": 1.891747470803772e-05,
+      "loss": 1.0105,
+      "step": 3052
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9987908601760864,
+      "learning_rate": 1.8916633922974487e-05,
+      "loss": 1.0848,
+      "step": 3053
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.033486843109131,
+      "learning_rate": 1.8915792830223122e-05,
+      "loss": 1.1923,
+      "step": 3054
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0743985176086426,
+      "learning_rate": 1.8914951429812653e-05,
+      "loss": 1.085,
+      "step": 3055
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7913750410079956,
+      "learning_rate": 1.8914109721772113e-05,
+      "loss": 0.9951,
+      "step": 3056
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9485929012298584,
+      "learning_rate": 1.891326770613055e-05,
+      "loss": 0.9978,
+      "step": 3057
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0151426792144775,
+      "learning_rate": 1.8912425382917013e-05,
+      "loss": 1.0312,
+      "step": 3058
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.888296127319336,
+      "learning_rate": 1.8911582752160572e-05,
+      "loss": 1.0874,
+      "step": 3059
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.062451124191284,
+      "learning_rate": 1.89107398138903e-05,
+      "loss": 1.0763,
+      "step": 3060
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9340568780899048,
+      "learning_rate": 1.8909896568135297e-05,
+      "loss": 0.9979,
+      "step": 3061
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.137028217315674,
+      "learning_rate": 1.8909053014924646e-05,
+      "loss": 1.1171,
+      "step": 3062
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8640222549438477,
+      "learning_rate": 1.8908209154287466e-05,
+      "loss": 1.027,
+      "step": 3063
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0309529304504395,
+      "learning_rate": 1.890736498625287e-05,
+      "loss": 1.0476,
+      "step": 3064
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5965559482574463,
+      "learning_rate": 1.890652051084999e-05,
+      "loss": 0.9815,
+      "step": 3065
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9703634977340698,
+      "learning_rate": 1.8905675728107966e-05,
+      "loss": 1.0342,
+      "step": 3066
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.154611349105835,
+      "learning_rate": 1.8904830638055948e-05,
+      "loss": 1.0083,
+      "step": 3067
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9131062030792236,
+      "learning_rate": 1.8903985240723104e-05,
+      "loss": 1.0713,
+      "step": 3068
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7599763870239258,
+      "learning_rate": 1.89031395361386e-05,
+      "loss": 1.0879,
+      "step": 3069
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0864375829696655,
+      "learning_rate": 1.890229352433162e-05,
+      "loss": 0.6227,
+      "step": 3070
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8956810235977173,
+      "learning_rate": 1.8901447205331354e-05,
+      "loss": 1.0572,
+      "step": 3071
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0365216732025146,
+      "learning_rate": 1.8900600579167014e-05,
+      "loss": 1.076,
+      "step": 3072
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.890666127204895,
+      "learning_rate": 1.8899753645867813e-05,
+      "loss": 1.0794,
+      "step": 3073
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0366103649139404,
+      "learning_rate": 1.8898906405462972e-05,
+      "loss": 1.0762,
+      "step": 3074
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6437525749206543,
+      "learning_rate": 1.889805885798173e-05,
+      "loss": 1.0652,
+      "step": 3075
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9615285396575928,
+      "learning_rate": 1.889721100345333e-05,
+      "loss": 1.0105,
+      "step": 3076
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7844871282577515,
+      "learning_rate": 1.8896362841907033e-05,
+      "loss": 1.0286,
+      "step": 3077
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0820298194885254,
+      "learning_rate": 1.8895514373372107e-05,
+      "loss": 0.617,
+      "step": 3078
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0752005577087402,
+      "learning_rate": 1.8894665597877824e-05,
+      "loss": 1.076,
+      "step": 3079
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0140368938446045,
+      "learning_rate": 1.889381651545348e-05,
+      "loss": 1.1663,
+      "step": 3080
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0944695472717285,
+      "learning_rate": 1.8892967126128373e-05,
+      "loss": 1.0689,
+      "step": 3081
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7865242958068848,
+      "learning_rate": 1.889211742993181e-05,
+      "loss": 1.1034,
+      "step": 3082
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3344666957855225,
+      "learning_rate": 1.8891267426893116e-05,
+      "loss": 1.0753,
+      "step": 3083
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8635098934173584,
+      "learning_rate": 1.8890417117041618e-05,
+      "loss": 1.0148,
+      "step": 3084
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.172523021697998,
+      "learning_rate": 1.8889566500406662e-05,
+      "loss": 0.6651,
+      "step": 3085
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9754880666732788,
+      "learning_rate": 1.8888715577017596e-05,
+      "loss": 1.0089,
+      "step": 3086
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0755326747894287,
+      "learning_rate": 1.8887864346903784e-05,
+      "loss": 1.1142,
+      "step": 3087
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8751474618911743,
+      "learning_rate": 1.8887012810094606e-05,
+      "loss": 1.0768,
+      "step": 3088
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7957533597946167,
+      "learning_rate": 1.8886160966619433e-05,
+      "loss": 1.0156,
+      "step": 3089
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0717203617095947,
+      "learning_rate": 1.8885308816507674e-05,
+      "loss": 1.0328,
+      "step": 3090
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8285044431686401,
+      "learning_rate": 1.8884456359788725e-05,
+      "loss": 1.0114,
+      "step": 3091
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9245933294296265,
+      "learning_rate": 1.8883603596492004e-05,
+      "loss": 1.1163,
+      "step": 3092
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1680736541748047,
+      "learning_rate": 1.888275052664694e-05,
+      "loss": 1.0796,
+      "step": 3093
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9532865285873413,
+      "learning_rate": 1.888189715028297e-05,
+      "loss": 1.059,
+      "step": 3094
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4088635444641113,
+      "learning_rate": 1.8881043467429533e-05,
+      "loss": 1.0819,
+      "step": 3095
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8426268100738525,
+      "learning_rate": 1.88801894781161e-05,
+      "loss": 1.0336,
+      "step": 3096
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.975570797920227,
+      "learning_rate": 1.8879335182372133e-05,
+      "loss": 1.0416,
+      "step": 3097
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0050601959228516,
+      "learning_rate": 1.887848058022711e-05,
+      "loss": 1.066,
+      "step": 3098
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0849416255950928,
+      "learning_rate": 1.887762567171053e-05,
+      "loss": 1.1271,
+      "step": 3099
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7428510189056396,
+      "learning_rate": 1.887677045685188e-05,
+      "loss": 1.1412,
+      "step": 3100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3142433166503906,
+      "learning_rate": 1.887591493568068e-05,
+      "loss": 0.9902,
+      "step": 3101
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7968847751617432,
+      "learning_rate": 1.887505910822645e-05,
+      "loss": 0.9576,
+      "step": 3102
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8358029127120972,
+      "learning_rate": 1.887420297451872e-05,
+      "loss": 1.0698,
+      "step": 3103
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0887221097946167,
+      "learning_rate": 1.8873346534587033e-05,
+      "loss": 0.5851,
+      "step": 3104
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.169832944869995,
+      "learning_rate": 1.8872489788460947e-05,
+      "loss": 1.1116,
+      "step": 3105
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9128682613372803,
+      "learning_rate": 1.8871632736170024e-05,
+      "loss": 1.1064,
+      "step": 3106
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9288883209228516,
+      "learning_rate": 1.887077537774383e-05,
+      "loss": 1.0524,
+      "step": 3107
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.997115969657898,
+      "learning_rate": 1.8869917713211964e-05,
+      "loss": 1.1389,
+      "step": 3108
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7426297664642334,
+      "learning_rate": 1.8869059742604013e-05,
+      "loss": 1.0374,
+      "step": 3109
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9589606523513794,
+      "learning_rate": 1.8868201465949585e-05,
+      "loss": 1.035,
+      "step": 3110
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9675849676132202,
+      "learning_rate": 1.8867342883278293e-05,
+      "loss": 1.0522,
+      "step": 3111
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9334344863891602,
+      "learning_rate": 1.8866483994619775e-05,
+      "loss": 1.0036,
+      "step": 3112
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.07240629196167,
+      "learning_rate": 1.886562480000366e-05,
+      "loss": 0.9803,
+      "step": 3113
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0564050674438477,
+      "learning_rate": 1.8864765299459595e-05,
+      "loss": 1.0422,
+      "step": 3114
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.148075819015503,
+      "learning_rate": 1.8863905493017242e-05,
+      "loss": 0.6243,
+      "step": 3115
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0131330490112305,
+      "learning_rate": 1.8863045380706275e-05,
+      "loss": 1.083,
+      "step": 3116
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8728803396224976,
+      "learning_rate": 1.8862184962556366e-05,
+      "loss": 1.0468,
+      "step": 3117
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8426618576049805,
+      "learning_rate": 1.8861324238597212e-05,
+      "loss": 1.116,
+      "step": 3118
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7704581022262573,
+      "learning_rate": 1.8860463208858513e-05,
+      "loss": 1.046,
+      "step": 3119
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8465875387191772,
+      "learning_rate": 1.8859601873369974e-05,
+      "loss": 1.0698,
+      "step": 3120
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9481022357940674,
+      "learning_rate": 1.8858740232161325e-05,
+      "loss": 1.0968,
+      "step": 3121
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.836522102355957,
+      "learning_rate": 1.88578782852623e-05,
+      "loss": 1.0386,
+      "step": 3122
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9044045209884644,
+      "learning_rate": 1.8857016032702634e-05,
+      "loss": 0.998,
+      "step": 3123
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1791692972183228,
+      "learning_rate": 1.8856153474512088e-05,
+      "loss": 0.66,
+      "step": 3124
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.986464500427246,
+      "learning_rate": 1.8855290610720426e-05,
+      "loss": 1.0451,
+      "step": 3125
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9232033491134644,
+      "learning_rate": 1.8854427441357418e-05,
+      "loss": 0.9975,
+      "step": 3126
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.05118989944458,
+      "learning_rate": 1.8853563966452854e-05,
+      "loss": 1.115,
+      "step": 3127
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7756752967834473,
+      "learning_rate": 1.8852700186036526e-05,
+      "loss": 1.076,
+      "step": 3128
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.897196888923645,
+      "learning_rate": 1.885183610013825e-05,
+      "loss": 1.1529,
+      "step": 3129
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6670805215835571,
+      "learning_rate": 1.8850971708787833e-05,
+      "loss": 1.0476,
+      "step": 3130
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.045642614364624,
+      "learning_rate": 1.8850107012015105e-05,
+      "loss": 0.5897,
+      "step": 3131
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.773659348487854,
+      "learning_rate": 1.884924200984991e-05,
+      "loss": 1.0037,
+      "step": 3132
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7240121364593506,
+      "learning_rate": 1.884837670232209e-05,
+      "loss": 1.0696,
+      "step": 3133
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.855925440788269,
+      "learning_rate": 1.884751108946151e-05,
+      "loss": 1.0964,
+      "step": 3134
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9446394443511963,
+      "learning_rate": 1.884664517129803e-05,
+      "loss": 1.0342,
+      "step": 3135
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8911516666412354,
+      "learning_rate": 1.8845778947861545e-05,
+      "loss": 1.0954,
+      "step": 3136
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9340369701385498,
+      "learning_rate": 1.8844912419181934e-05,
+      "loss": 1.0858,
+      "step": 3137
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9892759323120117,
+      "learning_rate": 1.8844045585289105e-05,
+      "loss": 1.0719,
+      "step": 3138
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8575681447982788,
+      "learning_rate": 1.8843178446212965e-05,
+      "loss": 1.0873,
+      "step": 3139
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7944180965423584,
+      "learning_rate": 1.884231100198344e-05,
+      "loss": 0.9844,
+      "step": 3140
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7894824743270874,
+      "learning_rate": 1.8841443252630463e-05,
+      "loss": 1.0887,
+      "step": 3141
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.845396876335144,
+      "learning_rate": 1.8840575198183977e-05,
+      "loss": 1.0653,
+      "step": 3142
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9280120134353638,
+      "learning_rate": 1.8839706838673933e-05,
+      "loss": 1.0616,
+      "step": 3143
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9053345918655396,
+      "learning_rate": 1.8838838174130303e-05,
+      "loss": 1.0198,
+      "step": 3144
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7476216554641724,
+      "learning_rate": 1.8837969204583055e-05,
+      "loss": 1.003,
+      "step": 3145
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0599719285964966,
+      "learning_rate": 1.883709993006218e-05,
+      "loss": 0.5704,
+      "step": 3146
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.021195411682129,
+      "learning_rate": 1.8836230350597667e-05,
+      "loss": 1.0717,
+      "step": 3147
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9832763671875,
+      "learning_rate": 1.8835360466219534e-05,
+      "loss": 1.0759,
+      "step": 3148
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9397449493408203,
+      "learning_rate": 1.8834490276957788e-05,
+      "loss": 1.1081,
+      "step": 3149
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0987125635147095,
+      "learning_rate": 1.8833619782842464e-05,
+      "loss": 0.608,
+      "step": 3150
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9179284572601318,
+      "learning_rate": 1.8832748983903593e-05,
+      "loss": 1.0453,
+      "step": 3151
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1638340950012207,
+      "learning_rate": 1.8831877880171233e-05,
+      "loss": 1.0745,
+      "step": 3152
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9122291803359985,
+      "learning_rate": 1.8831006471675433e-05,
+      "loss": 1.0547,
+      "step": 3153
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9634008407592773,
+      "learning_rate": 1.883013475844627e-05,
+      "loss": 1.0948,
+      "step": 3154
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9297114610671997,
+      "learning_rate": 1.8829262740513823e-05,
+      "loss": 1.0947,
+      "step": 3155
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.066471815109253,
+      "learning_rate": 1.882839041790818e-05,
+      "loss": 1.0247,
+      "step": 3156
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.388782262802124,
+      "learning_rate": 1.8827517790659447e-05,
+      "loss": 1.0798,
+      "step": 3157
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.885628342628479,
+      "learning_rate": 1.8826644858797734e-05,
+      "loss": 1.1219,
+      "step": 3158
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1876800060272217,
+      "learning_rate": 1.8825771622353164e-05,
+      "loss": 1.0793,
+      "step": 3159
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9918705224990845,
+      "learning_rate": 1.8824898081355866e-05,
+      "loss": 1.1487,
+      "step": 3160
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1957788467407227,
+      "learning_rate": 1.882402423583599e-05,
+      "loss": 0.6636,
+      "step": 3161
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8979698419570923,
+      "learning_rate": 1.8823150085823685e-05,
+      "loss": 1.0492,
+      "step": 3162
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8649994134902954,
+      "learning_rate": 1.8822275631349115e-05,
+      "loss": 1.045,
+      "step": 3163
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.918664813041687,
+      "learning_rate": 1.8821400872442458e-05,
+      "loss": 1.0176,
+      "step": 3164
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9375090599060059,
+      "learning_rate": 1.88205258091339e-05,
+      "loss": 1.0676,
+      "step": 3165
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8347877264022827,
+      "learning_rate": 1.8819650441453635e-05,
+      "loss": 1.1164,
+      "step": 3166
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1002001762390137,
+      "learning_rate": 1.881877476943187e-05,
+      "loss": 1.0222,
+      "step": 3167
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1007864475250244,
+      "learning_rate": 1.881789879309882e-05,
+      "loss": 1.0886,
+      "step": 3168
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9871070384979248,
+      "learning_rate": 1.8817022512484718e-05,
+      "loss": 1.0781,
+      "step": 3169
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9151084423065186,
+      "learning_rate": 1.8816145927619795e-05,
+      "loss": 1.0012,
+      "step": 3170
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.860733151435852,
+      "learning_rate": 1.8815269038534305e-05,
+      "loss": 1.0587,
+      "step": 3171
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.898160457611084,
+      "learning_rate": 1.8814391845258507e-05,
+      "loss": 1.0481,
+      "step": 3172
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9673105478286743,
+      "learning_rate": 1.8813514347822662e-05,
+      "loss": 1.1394,
+      "step": 3173
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.014975070953369,
+      "learning_rate": 1.8812636546257062e-05,
+      "loss": 1.0822,
+      "step": 3174
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9710543155670166,
+      "learning_rate": 1.881175844059199e-05,
+      "loss": 1.0113,
+      "step": 3175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.058241605758667,
+      "learning_rate": 1.881088003085775e-05,
+      "loss": 1.1249,
+      "step": 3176
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9997801780700684,
+      "learning_rate": 1.881000131708465e-05,
+      "loss": 0.9042,
+      "step": 3177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9435893297195435,
+      "learning_rate": 1.8809122299303015e-05,
+      "loss": 1.1371,
+      "step": 3178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9610495567321777,
+      "learning_rate": 1.880824297754318e-05,
+      "loss": 1.0608,
+      "step": 3179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8750948905944824,
+      "learning_rate": 1.880736335183548e-05,
+      "loss": 1.0738,
+      "step": 3180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9201828241348267,
+      "learning_rate": 1.8806483422210275e-05,
+      "loss": 1.1539,
+      "step": 3181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8751220703125,
+      "learning_rate": 1.880560318869793e-05,
+      "loss": 1.0392,
+      "step": 3182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9925799369812012,
+      "learning_rate": 1.8804722651328814e-05,
+      "loss": 1.1547,
+      "step": 3183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9806839227676392,
+      "learning_rate": 1.8803841810133315e-05,
+      "loss": 1.0571,
+      "step": 3184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8010292053222656,
+      "learning_rate": 1.8802960665141824e-05,
+      "loss": 1.036,
+      "step": 3185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0512712001800537,
+      "learning_rate": 1.8802079216384754e-05,
+      "loss": 1.1245,
+      "step": 3186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8635333776474,
+      "learning_rate": 1.8801197463892516e-05,
+      "loss": 1.1249,
+      "step": 3187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.824039340019226,
+      "learning_rate": 1.880031540769554e-05,
+      "loss": 1.1187,
+      "step": 3188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1490957736968994,
+      "learning_rate": 1.879943304782426e-05,
+      "loss": 1.0597,
+      "step": 3189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9700634479522705,
+      "learning_rate": 1.8798550384309128e-05,
+      "loss": 1.022,
+      "step": 3190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1496310234069824,
+      "learning_rate": 1.87976674171806e-05,
+      "loss": 0.6336,
+      "step": 3191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9290885925292969,
+      "learning_rate": 1.8796784146469146e-05,
+      "loss": 1.1359,
+      "step": 3192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.145732879638672,
+      "learning_rate": 1.8795900572205238e-05,
+      "loss": 1.031,
+      "step": 3193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9401350021362305,
+      "learning_rate": 1.8795016694419378e-05,
+      "loss": 1.04,
+      "step": 3194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.002579927444458,
+      "learning_rate": 1.8794132513142057e-05,
+      "loss": 1.0572,
+      "step": 3195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8517241477966309,
+      "learning_rate": 1.879324802840379e-05,
+      "loss": 1.0808,
+      "step": 3196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9809143543243408,
+      "learning_rate": 1.8792363240235097e-05,
+      "loss": 0.9701,
+      "step": 3197
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9905089139938354,
+      "learning_rate": 1.8791478148666504e-05,
+      "loss": 1.1054,
+      "step": 3198
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.018239736557007,
+      "learning_rate": 1.8790592753728562e-05,
+      "loss": 0.978,
+      "step": 3199
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0903102159500122,
+      "learning_rate": 1.878970705545182e-05,
+      "loss": 0.6354,
+      "step": 3200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0063588619232178,
+      "learning_rate": 1.878882105386684e-05,
+      "loss": 1.0131,
+      "step": 3201
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9619927406311035,
+      "learning_rate": 1.8787934749004194e-05,
+      "loss": 1.1232,
+      "step": 3202
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.878528118133545,
+      "learning_rate": 1.878704814089447e-05,
+      "loss": 0.9598,
+      "step": 3203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8036553859710693,
+      "learning_rate": 1.878616122956826e-05,
+      "loss": 1.0327,
+      "step": 3204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0007259845733643,
+      "learning_rate": 1.878527401505617e-05,
+      "loss": 1.0349,
+      "step": 3205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9058406352996826,
+      "learning_rate": 1.8784386497388813e-05,
+      "loss": 1.1136,
+      "step": 3206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7690937519073486,
+      "learning_rate": 1.8783498676596815e-05,
+      "loss": 1.0781,
+      "step": 3207
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8112245798110962,
+      "learning_rate": 1.8782610552710817e-05,
+      "loss": 0.9604,
+      "step": 3208
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7245216369628906,
+      "learning_rate": 1.878172212576146e-05,
+      "loss": 1.0069,
+      "step": 3209
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.879065752029419,
+      "learning_rate": 1.8780833395779402e-05,
+      "loss": 1.0045,
+      "step": 3210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9663444757461548,
+      "learning_rate": 1.8779944362795314e-05,
+      "loss": 1.1072,
+      "step": 3211
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8845505714416504,
+      "learning_rate": 1.877905502683987e-05,
+      "loss": 0.965,
+      "step": 3212
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3410561084747314,
+      "learning_rate": 1.877816538794376e-05,
+      "loss": 0.9502,
+      "step": 3213
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.131164789199829,
+      "learning_rate": 1.8777275446137687e-05,
+      "loss": 1.0873,
+      "step": 3214
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8176981210708618,
+      "learning_rate": 1.877638520145235e-05,
+      "loss": 1.0653,
+      "step": 3215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.87510085105896,
+      "learning_rate": 1.8775494653918482e-05,
+      "loss": 1.0659,
+      "step": 3216
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8689242601394653,
+      "learning_rate": 1.8774603803566804e-05,
+      "loss": 1.0962,
+      "step": 3217
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.958552598953247,
+      "learning_rate": 1.877371265042806e-05,
+      "loss": 1.0491,
+      "step": 3218
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8431156873703003,
+      "learning_rate": 1.8772821194533e-05,
+      "loss": 1.1154,
+      "step": 3219
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7707875967025757,
+      "learning_rate": 1.877192943591239e-05,
+      "loss": 1.0431,
+      "step": 3220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7103362083435059,
+      "learning_rate": 1.8771037374596995e-05,
+      "loss": 0.9648,
+      "step": 3221
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.987158179283142,
+      "learning_rate": 1.8770145010617604e-05,
+      "loss": 1.1035,
+      "step": 3222
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0184786319732666,
+      "learning_rate": 1.8769252344005005e-05,
+      "loss": 1.0932,
+      "step": 3223
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.153855085372925,
+      "learning_rate": 1.876835937479e-05,
+      "loss": 1.1398,
+      "step": 3224
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9682793617248535,
+      "learning_rate": 1.876746610300341e-05,
+      "loss": 1.0241,
+      "step": 3225
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9574861526489258,
+      "learning_rate": 1.8766572528676055e-05,
+      "loss": 1.1308,
+      "step": 3226
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1643271446228027,
+      "learning_rate": 1.8765678651838774e-05,
+      "loss": 1.0126,
+      "step": 3227
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8782780170440674,
+      "learning_rate": 1.8764784472522405e-05,
+      "loss": 1.0943,
+      "step": 3228
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9205485582351685,
+      "learning_rate": 1.8763889990757808e-05,
+      "loss": 1.0141,
+      "step": 3229
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0391829013824463,
+      "learning_rate": 1.8762995206575848e-05,
+      "loss": 1.0574,
+      "step": 3230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8242876529693604,
+      "learning_rate": 1.8762100120007402e-05,
+      "loss": 1.0889,
+      "step": 3231
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.039702296257019,
+      "learning_rate": 1.876120473108336e-05,
+      "loss": 0.6092,
+      "step": 3232
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9650447368621826,
+      "learning_rate": 1.8760309039834613e-05,
+      "loss": 1.008,
+      "step": 3233
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8859398365020752,
+      "learning_rate": 1.875941304629207e-05,
+      "loss": 0.9959,
+      "step": 3234
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.901572823524475,
+      "learning_rate": 1.8758516750486655e-05,
+      "loss": 1.045,
+      "step": 3235
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.124387502670288,
+      "learning_rate": 1.875762015244929e-05,
+      "loss": 0.9704,
+      "step": 3236
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8097633123397827,
+      "learning_rate": 1.8756723252210917e-05,
+      "loss": 1.0535,
+      "step": 3237
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8400793075561523,
+      "learning_rate": 1.8755826049802487e-05,
+      "loss": 0.9506,
+      "step": 3238
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7167398929595947,
+      "learning_rate": 1.875492854525496e-05,
+      "loss": 1.0451,
+      "step": 3239
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3077800273895264,
+      "learning_rate": 1.8754030738599302e-05,
+      "loss": 1.094,
+      "step": 3240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.194966197013855,
+      "learning_rate": 1.87531326298665e-05,
+      "loss": 0.6595,
+      "step": 3241
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0183727741241455,
+      "learning_rate": 1.8752234219087538e-05,
+      "loss": 1.0343,
+      "step": 3242
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2822582721710205,
+      "learning_rate": 1.8751335506293423e-05,
+      "loss": 1.0888,
+      "step": 3243
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9930367469787598,
+      "learning_rate": 1.8750436491515165e-05,
+      "loss": 1.0744,
+      "step": 3244
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1007680892944336,
+      "learning_rate": 1.8749537174783787e-05,
+      "loss": 1.0164,
+      "step": 3245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8790193796157837,
+      "learning_rate": 1.8748637556130323e-05,
+      "loss": 1.0029,
+      "step": 3246
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9194238185882568,
+      "learning_rate": 1.8747737635585817e-05,
+      "loss": 0.979,
+      "step": 3247
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2853827476501465,
+      "learning_rate": 1.874683741318132e-05,
+      "loss": 1.0845,
+      "step": 3248
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9042471647262573,
+      "learning_rate": 1.8745936888947893e-05,
+      "loss": 0.9847,
+      "step": 3249
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9672541618347168,
+      "learning_rate": 1.8745036062916617e-05,
+      "loss": 1.0156,
+      "step": 3250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9093239307403564,
+      "learning_rate": 1.8744134935118575e-05,
+      "loss": 1.0319,
+      "step": 3251
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8691750764846802,
+      "learning_rate": 1.8743233505584863e-05,
+      "loss": 1.0655,
+      "step": 3252
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8544687032699585,
+      "learning_rate": 1.8742331774346587e-05,
+      "loss": 1.0632,
+      "step": 3253
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.965006709098816,
+      "learning_rate": 1.874142974143486e-05,
+      "loss": 1.0937,
+      "step": 3254
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.952591896057129,
+      "learning_rate": 1.8740527406880813e-05,
+      "loss": 1.0957,
+      "step": 3255
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.837761402130127,
+      "learning_rate": 1.873962477071558e-05,
+      "loss": 0.9728,
+      "step": 3256
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9160650968551636,
+      "learning_rate": 1.8738721832970308e-05,
+      "loss": 1.0494,
+      "step": 3257
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1636738777160645,
+      "learning_rate": 1.8737818593676155e-05,
+      "loss": 1.0909,
+      "step": 3258
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8780964612960815,
+      "learning_rate": 1.8736915052864293e-05,
+      "loss": 0.9485,
+      "step": 3259
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1274797916412354,
+      "learning_rate": 1.8736011210565897e-05,
+      "loss": 1.0766,
+      "step": 3260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7647018432617188,
+      "learning_rate": 1.873510706681216e-05,
+      "loss": 0.9751,
+      "step": 3261
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.05830454826355,
+      "learning_rate": 1.8734202621634275e-05,
+      "loss": 1.0551,
+      "step": 3262
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8194825649261475,
+      "learning_rate": 1.8733297875063457e-05,
+      "loss": 1.0364,
+      "step": 3263
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9840232133865356,
+      "learning_rate": 1.8732392827130924e-05,
+      "loss": 1.1237,
+      "step": 3264
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.031609296798706,
+      "learning_rate": 1.8731487477867908e-05,
+      "loss": 1.053,
+      "step": 3265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8754713535308838,
+      "learning_rate": 1.873058182730565e-05,
+      "loss": 0.9853,
+      "step": 3266
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0270116329193115,
+      "learning_rate": 1.8729675875475402e-05,
+      "loss": 1.0821,
+      "step": 3267
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1765875816345215,
+      "learning_rate": 1.8728769622408423e-05,
+      "loss": 1.057,
+      "step": 3268
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0628955364227295,
+      "learning_rate": 1.8727863068135987e-05,
+      "loss": 1.0832,
+      "step": 3269
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7419322729110718,
+      "learning_rate": 1.872695621268938e-05,
+      "loss": 1.0309,
+      "step": 3270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.24467396736145,
+      "learning_rate": 1.8726049056099887e-05,
+      "loss": 1.0639,
+      "step": 3271
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3124465942382812,
+      "learning_rate": 1.8725141598398823e-05,
+      "loss": 0.9719,
+      "step": 3272
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9237641096115112,
+      "learning_rate": 1.872423383961749e-05,
+      "loss": 1.1478,
+      "step": 3273
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1040890216827393,
+      "learning_rate": 1.8723325779787218e-05,
+      "loss": 0.968,
+      "step": 3274
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6902024745941162,
+      "learning_rate": 1.872241741893934e-05,
+      "loss": 1.071,
+      "step": 3275
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.755793809890747,
+      "learning_rate": 1.8721508757105203e-05,
+      "loss": 1.0132,
+      "step": 3276
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1633825302124023,
+      "learning_rate": 1.872059979431616e-05,
+      "loss": 1.0303,
+      "step": 3277
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9733517169952393,
+      "learning_rate": 1.8719690530603582e-05,
+      "loss": 1.0649,
+      "step": 3278
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9661974906921387,
+      "learning_rate": 1.871878096599884e-05,
+      "loss": 1.0751,
+      "step": 3279
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0361452102661133,
+      "learning_rate": 1.8717871100533317e-05,
+      "loss": 1.047,
+      "step": 3280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1370222568511963,
+      "learning_rate": 1.8716960934238422e-05,
+      "loss": 1.0261,
+      "step": 3281
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9634020328521729,
+      "learning_rate": 1.871605046714555e-05,
+      "loss": 1.1747,
+      "step": 3282
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.172603964805603,
+      "learning_rate": 1.8715139699286125e-05,
+      "loss": 0.5776,
+      "step": 3283
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8870103359222412,
+      "learning_rate": 1.8714228630691576e-05,
+      "loss": 0.9769,
+      "step": 3284
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8174453973770142,
+      "learning_rate": 1.8713317261393337e-05,
+      "loss": 1.0209,
+      "step": 3285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8726760149002075,
+      "learning_rate": 1.8712405591422857e-05,
+      "loss": 0.9879,
+      "step": 3286
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9008108377456665,
+      "learning_rate": 1.8711493620811602e-05,
+      "loss": 0.9863,
+      "step": 3287
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9532196521759033,
+      "learning_rate": 1.8710581349591034e-05,
+      "loss": 1.0251,
+      "step": 3288
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.012089729309082,
+      "learning_rate": 1.8709668777792633e-05,
+      "loss": 1.0548,
+      "step": 3289
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8110605478286743,
+      "learning_rate": 1.8708755905447897e-05,
+      "loss": 1.0962,
+      "step": 3290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0551087856292725,
+      "learning_rate": 1.870784273258832e-05,
+      "loss": 1.123,
+      "step": 3291
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.985762357711792,
+      "learning_rate": 1.8706929259245412e-05,
+      "loss": 1.0283,
+      "step": 3292
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9035676717758179,
+      "learning_rate": 1.8706015485450697e-05,
+      "loss": 1.0683,
+      "step": 3293
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.752871036529541,
+      "learning_rate": 1.870510141123571e-05,
+      "loss": 1.0262,
+      "step": 3294
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8591305017471313,
+      "learning_rate": 1.870418703663199e-05,
+      "loss": 1.0719,
+      "step": 3295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.891185998916626,
+      "learning_rate": 1.870327236167109e-05,
+      "loss": 0.9989,
+      "step": 3296
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9542032480239868,
+      "learning_rate": 1.870235738638457e-05,
+      "loss": 1.0994,
+      "step": 3297
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9030170440673828,
+      "learning_rate": 1.8701442110804004e-05,
+      "loss": 1.0069,
+      "step": 3298
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8853739500045776,
+      "learning_rate": 1.870052653496098e-05,
+      "loss": 1.0701,
+      "step": 3299
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7397522926330566,
+      "learning_rate": 1.869961065888709e-05,
+      "loss": 1.0167,
+      "step": 3300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9591866731643677,
+      "learning_rate": 1.8698694482613937e-05,
+      "loss": 1.0855,
+      "step": 3301
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7865917682647705,
+      "learning_rate": 1.8697778006173134e-05,
+      "loss": 1.0791,
+      "step": 3302
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6885157823562622,
+      "learning_rate": 1.8696861229596307e-05,
+      "loss": 1.1335,
+      "step": 3303
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7753127813339233,
+      "learning_rate": 1.86959441529151e-05,
+      "loss": 0.9196,
+      "step": 3304
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.025127649307251,
+      "learning_rate": 1.8695026776161146e-05,
+      "loss": 1.0472,
+      "step": 3305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0208792686462402,
+      "learning_rate": 1.869410909936611e-05,
+      "loss": 1.0222,
+      "step": 3306
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9484285116195679,
+      "learning_rate": 1.869319112256165e-05,
+      "loss": 1.0533,
+      "step": 3307
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7399507761001587,
+      "learning_rate": 1.8692272845779448e-05,
+      "loss": 1.0167,
+      "step": 3308
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2569055557250977,
+      "learning_rate": 1.8691354269051192e-05,
+      "loss": 0.6248,
+      "step": 3309
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.904214859008789,
+      "learning_rate": 1.8690435392408584e-05,
+      "loss": 1.0831,
+      "step": 3310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.956451654434204,
+      "learning_rate": 1.868951621588332e-05,
+      "loss": 1.1825,
+      "step": 3311
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0369679927825928,
+      "learning_rate": 1.8688596739507127e-05,
+      "loss": 0.9904,
+      "step": 3312
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.966463565826416,
+      "learning_rate": 1.868767696331173e-05,
+      "loss": 1.1563,
+      "step": 3313
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8489439487457275,
+      "learning_rate": 1.868675688732887e-05,
+      "loss": 1.0889,
+      "step": 3314
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.895443081855774,
+      "learning_rate": 1.8685836511590297e-05,
+      "loss": 1.0644,
+      "step": 3315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.765350103378296,
+      "learning_rate": 1.8684915836127766e-05,
+      "loss": 0.9662,
+      "step": 3316
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9570988416671753,
+      "learning_rate": 1.8683994860973053e-05,
+      "loss": 1.0477,
+      "step": 3317
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.866875410079956,
+      "learning_rate": 1.8683073586157933e-05,
+      "loss": 1.0396,
+      "step": 3318
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8933188915252686,
+      "learning_rate": 1.86821520117142e-05,
+      "loss": 1.0332,
+      "step": 3319
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9142025709152222,
+      "learning_rate": 1.868123013767365e-05,
+      "loss": 0.9864,
+      "step": 3320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1308761835098267,
+      "learning_rate": 1.86803079640681e-05,
+      "loss": 0.6163,
+      "step": 3321
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.093803882598877,
+      "learning_rate": 1.867938549092937e-05,
+      "loss": 1.1464,
+      "step": 3322
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0949089527130127,
+      "learning_rate": 1.8678462718289293e-05,
+      "loss": 1.0002,
+      "step": 3323
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8703641891479492,
+      "learning_rate": 1.8677539646179706e-05,
+      "loss": 1.1262,
+      "step": 3324
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0862178802490234,
+      "learning_rate": 1.867661627463247e-05,
+      "loss": 1.0913,
+      "step": 3325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9975895881652832,
+      "learning_rate": 1.8675692603679443e-05,
+      "loss": 1.061,
+      "step": 3326
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8370732069015503,
+      "learning_rate": 1.8674768633352497e-05,
+      "loss": 0.9806,
+      "step": 3327
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7796324491500854,
+      "learning_rate": 1.8673844363683517e-05,
+      "loss": 1.1344,
+      "step": 3328
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3345690965652466,
+      "learning_rate": 1.8672919794704398e-05,
+      "loss": 0.649,
+      "step": 3329
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0007505416870117,
+      "learning_rate": 1.8671994926447047e-05,
+      "loss": 1.16,
+      "step": 3330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7431893348693848,
+      "learning_rate": 1.867106975894337e-05,
+      "loss": 1.105,
+      "step": 3331
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9228627681732178,
+      "learning_rate": 1.86701442922253e-05,
+      "loss": 1.0223,
+      "step": 3332
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.188016653060913,
+      "learning_rate": 1.866921852632477e-05,
+      "loss": 1.0894,
+      "step": 3333
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9352843761444092,
+      "learning_rate": 1.8668292461273726e-05,
+      "loss": 1.0402,
+      "step": 3334
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.834652066230774,
+      "learning_rate": 1.866736609710412e-05,
+      "loss": 1.141,
+      "step": 3335
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.803381085395813,
+      "learning_rate": 1.866643943384792e-05,
+      "loss": 1.0879,
+      "step": 3336
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0487242937088013,
+      "learning_rate": 1.8665512471537108e-05,
+      "loss": 0.5524,
+      "step": 3337
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7956326007843018,
+      "learning_rate": 1.8664585210203663e-05,
+      "loss": 1.0008,
+      "step": 3338
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.905713677406311,
+      "learning_rate": 1.866365764987959e-05,
+      "loss": 1.0815,
+      "step": 3339
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.097097873687744,
+      "learning_rate": 1.866272979059689e-05,
+      "loss": 1.0612,
+      "step": 3340
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0210537910461426,
+      "learning_rate": 1.8661801632387586e-05,
+      "loss": 1.2117,
+      "step": 3341
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9372286796569824,
+      "learning_rate": 1.86608731752837e-05,
+      "loss": 1.1173,
+      "step": 3342
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.006133794784546,
+      "learning_rate": 1.8659944419317275e-05,
+      "loss": 0.5819,
+      "step": 3343
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7832244634628296,
+      "learning_rate": 1.8659015364520358e-05,
+      "loss": 1.0938,
+      "step": 3344
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0531816482543945,
+      "learning_rate": 1.8658086010925012e-05,
+      "loss": 1.0564,
+      "step": 3345
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0361785888671875,
+      "learning_rate": 1.8657156358563298e-05,
+      "loss": 1.1287,
+      "step": 3346
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9186803102493286,
+      "learning_rate": 1.8656226407467307e-05,
+      "loss": 1.0378,
+      "step": 3347
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0930813550949097,
+      "learning_rate": 1.8655296157669118e-05,
+      "loss": 0.6092,
+      "step": 3348
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7512125968933105,
+      "learning_rate": 1.8654365609200837e-05,
+      "loss": 1.0061,
+      "step": 3349
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8459097146987915,
+      "learning_rate": 1.8653434762094577e-05,
+      "loss": 1.0309,
+      "step": 3350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8281456232070923,
+      "learning_rate": 1.8652503616382453e-05,
+      "loss": 1.0914,
+      "step": 3351
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8712741136550903,
+      "learning_rate": 1.86515721720966e-05,
+      "loss": 1.12,
+      "step": 3352
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.785926342010498,
+      "learning_rate": 1.865064042926916e-05,
+      "loss": 0.9756,
+      "step": 3353
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8476792573928833,
+      "learning_rate": 1.8649708387932278e-05,
+      "loss": 1.0182,
+      "step": 3354
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1880877017974854,
+      "learning_rate": 1.8648776048118123e-05,
+      "loss": 1.0432,
+      "step": 3355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0624444484710693,
+      "learning_rate": 1.864784340985887e-05,
+      "loss": 1.0405,
+      "step": 3356
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2413864135742188,
+      "learning_rate": 1.8646910473186696e-05,
+      "loss": 1.1001,
+      "step": 3357
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9817475080490112,
+      "learning_rate": 1.8645977238133794e-05,
+      "loss": 1.0978,
+      "step": 3358
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.786462664604187,
+      "learning_rate": 1.8645043704732367e-05,
+      "loss": 1.0761,
+      "step": 3359
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.049424409866333,
+      "learning_rate": 1.8644109873014637e-05,
+      "loss": 1.0944,
+      "step": 3360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.22489595413208,
+      "learning_rate": 1.8643175743012822e-05,
+      "loss": 1.068,
+      "step": 3361
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8220051527023315,
+      "learning_rate": 1.864224131475915e-05,
+      "loss": 1.132,
+      "step": 3362
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9313114881515503,
+      "learning_rate": 1.8641306588285873e-05,
+      "loss": 1.0718,
+      "step": 3363
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8049237728118896,
+      "learning_rate": 1.8640371563625246e-05,
+      "loss": 1.0327,
+      "step": 3364
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8267239332199097,
+      "learning_rate": 1.863943624080953e-05,
+      "loss": 1.0397,
+      "step": 3365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0097880363464355,
+      "learning_rate": 1.8638500619871004e-05,
+      "loss": 1.0868,
+      "step": 3366
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.853818655014038,
+      "learning_rate": 1.8637564700841953e-05,
+      "loss": 1.0749,
+      "step": 3367
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8577972650527954,
+      "learning_rate": 1.8636628483754672e-05,
+      "loss": 1.0314,
+      "step": 3368
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9461338520050049,
+      "learning_rate": 1.8635691968641465e-05,
+      "loss": 1.0079,
+      "step": 3369
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8037561178207397,
+      "learning_rate": 1.8634755155534653e-05,
+      "loss": 1.0111,
+      "step": 3370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9555385112762451,
+      "learning_rate": 1.863381804446656e-05,
+      "loss": 1.038,
+      "step": 3371
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1431832313537598,
+      "learning_rate": 1.8632880635469526e-05,
+      "loss": 1.1583,
+      "step": 3372
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.827445387840271,
+      "learning_rate": 1.8631942928575896e-05,
+      "loss": 1.0799,
+      "step": 3373
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9537386894226074,
+      "learning_rate": 1.8631004923818025e-05,
+      "loss": 1.0936,
+      "step": 3374
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8883380889892578,
+      "learning_rate": 1.8630066621228285e-05,
+      "loss": 1.0691,
+      "step": 3375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.809682011604309,
+      "learning_rate": 1.8629128020839054e-05,
+      "loss": 1.0347,
+      "step": 3376
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0518288612365723,
+      "learning_rate": 1.8628189122682716e-05,
+      "loss": 1.0668,
+      "step": 3377
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9575802087783813,
+      "learning_rate": 1.8627249926791678e-05,
+      "loss": 1.0137,
+      "step": 3378
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9513120651245117,
+      "learning_rate": 1.8626310433198344e-05,
+      "loss": 1.0154,
+      "step": 3379
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9968680143356323,
+      "learning_rate": 1.862537064193513e-05,
+      "loss": 1.1619,
+      "step": 3380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7917680740356445,
+      "learning_rate": 1.8624430553034472e-05,
+      "loss": 1.0194,
+      "step": 3381
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.778450608253479,
+      "learning_rate": 1.8623490166528807e-05,
+      "loss": 1.0188,
+      "step": 3382
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8233938217163086,
+      "learning_rate": 1.8622549482450584e-05,
+      "loss": 0.9837,
+      "step": 3383
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8853827714920044,
+      "learning_rate": 1.8621608500832265e-05,
+      "loss": 1.0428,
+      "step": 3384
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9374910593032837,
+      "learning_rate": 1.862066722170632e-05,
+      "loss": 1.0215,
+      "step": 3385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.876924991607666,
+      "learning_rate": 1.8619725645105228e-05,
+      "loss": 0.9934,
+      "step": 3386
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7990772724151611,
+      "learning_rate": 1.8618783771061487e-05,
+      "loss": 1.0593,
+      "step": 3387
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.000581979751587,
+      "learning_rate": 1.8617841599607588e-05,
+      "loss": 1.093,
+      "step": 3388
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9209024906158447,
+      "learning_rate": 1.861689913077605e-05,
+      "loss": 1.0887,
+      "step": 3389
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.840140700340271,
+      "learning_rate": 1.8615956364599395e-05,
+      "loss": 0.9985,
+      "step": 3390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.940855622291565,
+      "learning_rate": 1.8615013301110153e-05,
+      "loss": 1.08,
+      "step": 3391
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.937943696975708,
+      "learning_rate": 1.8614069940340867e-05,
+      "loss": 1.057,
+      "step": 3392
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9030269384384155,
+      "learning_rate": 1.8613126282324092e-05,
+      "loss": 1.0093,
+      "step": 3393
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1251370906829834,
+      "learning_rate": 1.8612182327092386e-05,
+      "loss": 1.0717,
+      "step": 3394
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0540411472320557,
+      "learning_rate": 1.8611238074678324e-05,
+      "loss": 1.0871,
+      "step": 3395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.980051875114441,
+      "learning_rate": 1.8610293525114492e-05,
+      "loss": 1.1254,
+      "step": 3396
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9048339128494263,
+      "learning_rate": 1.8609348678433485e-05,
+      "loss": 1.0603,
+      "step": 3397
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9517136812210083,
+      "learning_rate": 1.8608403534667902e-05,
+      "loss": 1.0888,
+      "step": 3398
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.987550139427185,
+      "learning_rate": 1.860745809385036e-05,
+      "loss": 1.0881,
+      "step": 3399
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9593169689178467,
+      "learning_rate": 1.8606512356013482e-05,
+      "loss": 0.9967,
+      "step": 3400
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9651020765304565,
+      "learning_rate": 1.8605566321189907e-05,
+      "loss": 1.0839,
+      "step": 3401
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.277466058731079,
+      "learning_rate": 1.860461998941227e-05,
+      "loss": 1.0311,
+      "step": 3402
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9531781673431396,
+      "learning_rate": 1.860367336071324e-05,
+      "loss": 1.0481,
+      "step": 3403
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2110350131988525,
+      "learning_rate": 1.8602726435125472e-05,
+      "loss": 1.0074,
+      "step": 3404
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9735784530639648,
+      "learning_rate": 1.8601779212681646e-05,
+      "loss": 1.0323,
+      "step": 3405
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0155084133148193,
+      "learning_rate": 1.860083169341445e-05,
+      "loss": 1.0625,
+      "step": 3406
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.830021858215332,
+      "learning_rate": 1.8599883877356576e-05,
+      "loss": 1.09,
+      "step": 3407
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.08437442779541,
+      "learning_rate": 1.8598935764540732e-05,
+      "loss": 1.0314,
+      "step": 3408
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.030372381210327,
+      "learning_rate": 1.8597987354999635e-05,
+      "loss": 1.0192,
+      "step": 3409
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.255364418029785,
+      "learning_rate": 1.8597038648766013e-05,
+      "loss": 1.0897,
+      "step": 3410
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0484447479248047,
+      "learning_rate": 1.85960896458726e-05,
+      "loss": 1.0403,
+      "step": 3411
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9664161205291748,
+      "learning_rate": 1.859514034635215e-05,
+      "loss": 1.1266,
+      "step": 3412
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.064326763153076,
+      "learning_rate": 1.8594190750237412e-05,
+      "loss": 1.0785,
+      "step": 3413
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8610503673553467,
+      "learning_rate": 1.859324085756116e-05,
+      "loss": 1.0605,
+      "step": 3414
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1601450443267822,
+      "learning_rate": 1.859229066835617e-05,
+      "loss": 1.0711,
+      "step": 3415
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9858617782592773,
+      "learning_rate": 1.8591340182655227e-05,
+      "loss": 1.0829,
+      "step": 3416
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9081016778945923,
+      "learning_rate": 1.8590389400491137e-05,
+      "loss": 0.9656,
+      "step": 3417
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.811676263809204,
+      "learning_rate": 1.8589438321896707e-05,
+      "loss": 1.0363,
+      "step": 3418
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9044095277786255,
+      "learning_rate": 1.858848694690475e-05,
+      "loss": 0.9673,
+      "step": 3419
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9564266204833984,
+      "learning_rate": 1.8587535275548102e-05,
+      "loss": 0.9847,
+      "step": 3420
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8827168941497803,
+      "learning_rate": 1.85865833078596e-05,
+      "loss": 1.1169,
+      "step": 3421
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0502302646636963,
+      "learning_rate": 1.858563104387209e-05,
+      "loss": 1.0226,
+      "step": 3422
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9666664600372314,
+      "learning_rate": 1.8584678483618437e-05,
+      "loss": 1.0455,
+      "step": 3423
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9862865209579468,
+      "learning_rate": 1.858372562713151e-05,
+      "loss": 1.0348,
+      "step": 3424
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1505987644195557,
+      "learning_rate": 1.8582772474444192e-05,
+      "loss": 1.0554,
+      "step": 3425
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7604612112045288,
+      "learning_rate": 1.858181902558937e-05,
+      "loss": 1.0212,
+      "step": 3426
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2355408668518066,
+      "learning_rate": 1.8580865280599943e-05,
+      "loss": 0.591,
+      "step": 3427
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.901374101638794,
+      "learning_rate": 1.8579911239508827e-05,
+      "loss": 1.0276,
+      "step": 3428
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.189349412918091,
+      "learning_rate": 1.8578956902348945e-05,
+      "loss": 1.0036,
+      "step": 3429
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.777085065841675,
+      "learning_rate": 1.857800226915322e-05,
+      "loss": 1.0639,
+      "step": 3430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7158477306365967,
+      "learning_rate": 1.8577047339954597e-05,
+      "loss": 0.9864,
+      "step": 3431
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9763731956481934,
+      "learning_rate": 1.8576092114786034e-05,
+      "loss": 1.0439,
+      "step": 3432
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9182837009429932,
+      "learning_rate": 1.8575136593680483e-05,
+      "loss": 1.0567,
+      "step": 3433
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.181162118911743,
+      "learning_rate": 1.8574180776670924e-05,
+      "loss": 1.0551,
+      "step": 3434
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9289743900299072,
+      "learning_rate": 1.8573224663790338e-05,
+      "loss": 1.0667,
+      "step": 3435
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7823630571365356,
+      "learning_rate": 1.8572268255071718e-05,
+      "loss": 1.0212,
+      "step": 3436
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8837275505065918,
+      "learning_rate": 1.8571311550548064e-05,
+      "loss": 1.0507,
+      "step": 3437
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8886528015136719,
+      "learning_rate": 1.8570354550252398e-05,
+      "loss": 1.0988,
+      "step": 3438
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7002193927764893,
+      "learning_rate": 1.856939725421773e-05,
+      "loss": 1.0583,
+      "step": 3439
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.813125491142273,
+      "learning_rate": 1.85684396624771e-05,
+      "loss": 0.9866,
+      "step": 3440
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9589003324508667,
+      "learning_rate": 1.8567481775063556e-05,
+      "loss": 1.0748,
+      "step": 3441
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.978636622428894,
+      "learning_rate": 1.8566523592010143e-05,
+      "loss": 1.0436,
+      "step": 3442
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0496766567230225,
+      "learning_rate": 1.8565565113349934e-05,
+      "loss": 1.0731,
+      "step": 3443
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.119873046875,
+      "learning_rate": 1.8564606339116e-05,
+      "loss": 1.0264,
+      "step": 3444
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0026862621307373,
+      "learning_rate": 1.856364726934143e-05,
+      "loss": 1.1001,
+      "step": 3445
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.878288745880127,
+      "learning_rate": 1.8562687904059307e-05,
+      "loss": 1.081,
+      "step": 3446
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7916675806045532,
+      "learning_rate": 1.8561728243302745e-05,
+      "loss": 1.0133,
+      "step": 3447
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8412566184997559,
+      "learning_rate": 1.856076828710486e-05,
+      "loss": 1.1449,
+      "step": 3448
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0950841903686523,
+      "learning_rate": 1.855980803549877e-05,
+      "loss": 1.078,
+      "step": 3449
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8982582092285156,
+      "learning_rate": 1.855884748851762e-05,
+      "loss": 0.9412,
+      "step": 3450
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7044872045516968,
+      "learning_rate": 1.8557886646194553e-05,
+      "loss": 1.0361,
+      "step": 3451
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9325010776519775,
+      "learning_rate": 1.855692550856272e-05,
+      "loss": 1.1081,
+      "step": 3452
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0346431732177734,
+      "learning_rate": 1.855596407565529e-05,
+      "loss": 1.094,
+      "step": 3453
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.117912530899048,
+      "learning_rate": 1.855500234750544e-05,
+      "loss": 1.0912,
+      "step": 3454
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9300957918167114,
+      "learning_rate": 1.855404032414636e-05,
+      "loss": 1.021,
+      "step": 3455
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9170085191726685,
+      "learning_rate": 1.855307800561124e-05,
+      "loss": 1.0564,
+      "step": 3456
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8575444221496582,
+      "learning_rate": 1.855211539193329e-05,
+      "loss": 1.0836,
+      "step": 3457
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9159964323043823,
+      "learning_rate": 1.8551152483145728e-05,
+      "loss": 1.0474,
+      "step": 3458
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6806856393814087,
+      "learning_rate": 1.855018927928178e-05,
+      "loss": 0.9806,
+      "step": 3459
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2402334213256836,
+      "learning_rate": 1.8549225780374683e-05,
+      "loss": 1.1379,
+      "step": 3460
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.904198169708252,
+      "learning_rate": 1.854826198645769e-05,
+      "loss": 1.0081,
+      "step": 3461
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7969858646392822,
+      "learning_rate": 1.8547297897564053e-05,
+      "loss": 1.0711,
+      "step": 3462
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9743412733078003,
+      "learning_rate": 1.8546333513727037e-05,
+      "loss": 1.0907,
+      "step": 3463
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.043072462081909,
+      "learning_rate": 1.8545368834979925e-05,
+      "loss": 1.0075,
+      "step": 3464
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9212602376937866,
+      "learning_rate": 1.854440386135601e-05,
+      "loss": 1.0117,
+      "step": 3465
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7230650186538696,
+      "learning_rate": 1.8543438592888585e-05,
+      "loss": 1.0933,
+      "step": 3466
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9055861234664917,
+      "learning_rate": 1.8542473029610954e-05,
+      "loss": 1.0289,
+      "step": 3467
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9236326217651367,
+      "learning_rate": 1.8541507171556445e-05,
+      "loss": 1.0834,
+      "step": 3468
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.863621473312378,
+      "learning_rate": 1.8540541018758383e-05,
+      "loss": 1.0588,
+      "step": 3469
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8660657405853271,
+      "learning_rate": 1.853957457125011e-05,
+      "loss": 1.0279,
+      "step": 3470
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7194898128509521,
+      "learning_rate": 1.853860782906497e-05,
+      "loss": 1.0292,
+      "step": 3471
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0675582885742188,
+      "learning_rate": 1.8537640792236326e-05,
+      "loss": 1.056,
+      "step": 3472
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9705510139465332,
+      "learning_rate": 1.8536673460797546e-05,
+      "loss": 1.0987,
+      "step": 3473
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6624950170516968,
+      "learning_rate": 1.853570583478201e-05,
+      "loss": 0.9893,
+      "step": 3474
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1362228393554688,
+      "learning_rate": 1.8534737914223113e-05,
+      "loss": 1.1606,
+      "step": 3475
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.08270001411438,
+      "learning_rate": 1.853376969915425e-05,
+      "loss": 1.0718,
+      "step": 3476
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8066281080245972,
+      "learning_rate": 1.8532801189608833e-05,
+      "loss": 1.14,
+      "step": 3477
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8894119262695312,
+      "learning_rate": 1.8531832385620282e-05,
+      "loss": 1.1019,
+      "step": 3478
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0659496784210205,
+      "learning_rate": 1.8530863287222026e-05,
+      "loss": 1.0491,
+      "step": 3479
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9510080814361572,
+      "learning_rate": 1.852989389444751e-05,
+      "loss": 1.1009,
+      "step": 3480
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8240132331848145,
+      "learning_rate": 1.8528924207330183e-05,
+      "loss": 1.0875,
+      "step": 3481
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8344157934188843,
+      "learning_rate": 1.8527954225903508e-05,
+      "loss": 1.0414,
+      "step": 3482
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9397079944610596,
+      "learning_rate": 1.852698395020095e-05,
+      "loss": 1.0563,
+      "step": 3483
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0839195251464844,
+      "learning_rate": 1.8526013380255998e-05,
+      "loss": 0.979,
+      "step": 3484
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7874321937561035,
+      "learning_rate": 1.852504251610214e-05,
+      "loss": 1.1049,
+      "step": 3485
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.026287078857422,
+      "learning_rate": 1.852407135777288e-05,
+      "loss": 1.1652,
+      "step": 3486
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9050946235656738,
+      "learning_rate": 1.8523099905301726e-05,
+      "loss": 1.0989,
+      "step": 3487
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9575291872024536,
+      "learning_rate": 1.8522128158722204e-05,
+      "loss": 0.9963,
+      "step": 3488
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9820384979248047,
+      "learning_rate": 1.8521156118067846e-05,
+      "loss": 1.095,
+      "step": 3489
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7276792526245117,
+      "learning_rate": 1.8520183783372193e-05,
+      "loss": 1.0635,
+      "step": 3490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7215031385421753,
+      "learning_rate": 1.8519211154668796e-05,
+      "loss": 0.9974,
+      "step": 3491
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8326390981674194,
+      "learning_rate": 1.851823823199122e-05,
+      "loss": 1.0732,
+      "step": 3492
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2233939170837402,
+      "learning_rate": 1.8517265015373035e-05,
+      "loss": 0.681,
+      "step": 3493
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9728939533233643,
+      "learning_rate": 1.851629150484783e-05,
+      "loss": 1.0716,
+      "step": 3494
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0011157989501953,
+      "learning_rate": 1.8515317700449195e-05,
+      "loss": 1.043,
+      "step": 3495
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7853963375091553,
+      "learning_rate": 1.851434360221073e-05,
+      "loss": 1.0615,
+      "step": 3496
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.888091802597046,
+      "learning_rate": 1.8513369210166052e-05,
+      "loss": 1.0872,
+      "step": 3497
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.10886287689209,
+      "learning_rate": 1.8512394524348786e-05,
+      "loss": 0.9939,
+      "step": 3498
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1124346256256104,
+      "learning_rate": 1.851141954479256e-05,
+      "loss": 1.1193,
+      "step": 3499
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.958008050918579,
+      "learning_rate": 1.8510444271531023e-05,
+      "loss": 1.138,
+      "step": 3500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8277994394302368,
+      "learning_rate": 1.8509468704597827e-05,
+      "loss": 1.0539,
+      "step": 3501
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9879865646362305,
+      "learning_rate": 1.8508492844026636e-05,
+      "loss": 1.075,
+      "step": 3502
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9964429140090942,
+      "learning_rate": 1.8507516689851126e-05,
+      "loss": 1.035,
+      "step": 3503
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8694556951522827,
+      "learning_rate": 1.8506540242104977e-05,
+      "loss": 1.1122,
+      "step": 3504
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8541712760925293,
+      "learning_rate": 1.8505563500821888e-05,
+      "loss": 1.087,
+      "step": 3505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8021937608718872,
+      "learning_rate": 1.850458646603556e-05,
+      "loss": 1.0213,
+      "step": 3506
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.835868000984192,
+      "learning_rate": 1.8503609137779712e-05,
+      "loss": 1.0971,
+      "step": 3507
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9602806568145752,
+      "learning_rate": 1.8502631516088067e-05,
+      "loss": 1.0101,
+      "step": 3508
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9882681369781494,
+      "learning_rate": 1.850165360099436e-05,
+      "loss": 0.9701,
+      "step": 3509
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0302655696868896,
+      "learning_rate": 1.850067539253233e-05,
+      "loss": 1.0132,
+      "step": 3510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0800695419311523,
+      "learning_rate": 1.8499696890735742e-05,
+      "loss": 1.0684,
+      "step": 3511
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.922810673713684,
+      "learning_rate": 1.8498718095638357e-05,
+      "loss": 0.9766,
+      "step": 3512
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.787361979484558,
+      "learning_rate": 1.849773900727395e-05,
+      "loss": 1.0926,
+      "step": 3513
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.770956039428711,
+      "learning_rate": 1.849675962567631e-05,
+      "loss": 1.0391,
+      "step": 3514
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9045885801315308,
+      "learning_rate": 1.8495779950879226e-05,
+      "loss": 1.0787,
+      "step": 3515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2738239765167236,
+      "learning_rate": 1.8494799982916512e-05,
+      "loss": 0.6159,
+      "step": 3516
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0047459602355957,
+      "learning_rate": 1.8493819721821977e-05,
+      "loss": 1.0209,
+      "step": 3517
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.923448920249939,
+      "learning_rate": 1.849283916762945e-05,
+      "loss": 0.9357,
+      "step": 3518
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0513598918914795,
+      "learning_rate": 1.849185832037277e-05,
+      "loss": 1.1697,
+      "step": 3519
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8110374212265015,
+      "learning_rate": 1.8490877180085774e-05,
+      "loss": 1.1264,
+      "step": 3520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9033101797103882,
+      "learning_rate": 1.8489895746802333e-05,
+      "loss": 1.0743,
+      "step": 3521
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8522331714630127,
+      "learning_rate": 1.84889140205563e-05,
+      "loss": 1.0904,
+      "step": 3522
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.831287145614624,
+      "learning_rate": 1.848793200138156e-05,
+      "loss": 1.074,
+      "step": 3523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.103928565979004,
+      "learning_rate": 1.8486949689311995e-05,
+      "loss": 1.0133,
+      "step": 3524
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.107046604156494,
+      "learning_rate": 1.8485967084381502e-05,
+      "loss": 1.1111,
+      "step": 3525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8897216320037842,
+      "learning_rate": 1.8484984186623994e-05,
+      "loss": 1.0076,
+      "step": 3526
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.059037208557129,
+      "learning_rate": 1.848400099607338e-05,
+      "loss": 1.0314,
+      "step": 3527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.891284227371216,
+      "learning_rate": 1.848301751276359e-05,
+      "loss": 0.9937,
+      "step": 3528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.828878402709961,
+      "learning_rate": 1.8482033736728563e-05,
+      "loss": 1.0455,
+      "step": 3529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8921170234680176,
+      "learning_rate": 1.8481049668002246e-05,
+      "loss": 1.0579,
+      "step": 3530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9101365804672241,
+      "learning_rate": 1.84800653066186e-05,
+      "loss": 0.9849,
+      "step": 3531
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.079787015914917,
+      "learning_rate": 1.8479080652611584e-05,
+      "loss": 1.0394,
+      "step": 3532
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0759100914001465,
+      "learning_rate": 1.8478095706015177e-05,
+      "loss": 1.0842,
+      "step": 3533
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.086484909057617,
+      "learning_rate": 1.847711046686337e-05,
+      "loss": 1.0996,
+      "step": 3534
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9296739101409912,
+      "learning_rate": 1.8476124935190168e-05,
+      "loss": 1.0444,
+      "step": 3535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2011492252349854,
+      "learning_rate": 1.8475139111029565e-05,
+      "loss": 1.0634,
+      "step": 3536
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2124176025390625,
+      "learning_rate": 1.8474152994415588e-05,
+      "loss": 0.5315,
+      "step": 3537
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7310326099395752,
+      "learning_rate": 1.8473166585382266e-05,
+      "loss": 0.9949,
+      "step": 3538
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.79548180103302,
+      "learning_rate": 1.847217988396363e-05,
+      "loss": 1.0466,
+      "step": 3539
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.035682439804077,
+      "learning_rate": 1.847119289019373e-05,
+      "loss": 1.0682,
+      "step": 3540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9045240879058838,
+      "learning_rate": 1.847020560410663e-05,
+      "loss": 1.0211,
+      "step": 3541
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8681426048278809,
+      "learning_rate": 1.8469218025736393e-05,
+      "loss": 1.0566,
+      "step": 3542
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7367184162139893,
+      "learning_rate": 1.8468230155117106e-05,
+      "loss": 1.0097,
+      "step": 3543
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.842137336730957,
+      "learning_rate": 1.8467241992282842e-05,
+      "loss": 1.0695,
+      "step": 3544
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0482707023620605,
+      "learning_rate": 1.8466253537267714e-05,
+      "loss": 1.0629,
+      "step": 3545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8125511407852173,
+      "learning_rate": 1.8465264790105827e-05,
+      "loss": 1.0347,
+      "step": 3546
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.749755620956421,
+      "learning_rate": 1.84642757508313e-05,
+      "loss": 1.0436,
+      "step": 3547
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8040045499801636,
+      "learning_rate": 1.8463286419478256e-05,
+      "loss": 1.0531,
+      "step": 3548
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7575068473815918,
+      "learning_rate": 1.8462296796080843e-05,
+      "loss": 0.9501,
+      "step": 3549
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8651032447814941,
+      "learning_rate": 1.84613068806732e-05,
+      "loss": 1.0134,
+      "step": 3550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.925028920173645,
+      "learning_rate": 1.84603166732895e-05,
+      "loss": 1.1533,
+      "step": 3551
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9582946300506592,
+      "learning_rate": 1.84593261739639e-05,
+      "loss": 1.0809,
+      "step": 3552
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9587719440460205,
+      "learning_rate": 1.8458335382730585e-05,
+      "loss": 1.0741,
+      "step": 3553
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8932819366455078,
+      "learning_rate": 1.8457344299623747e-05,
+      "loss": 1.0782,
+      "step": 3554
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6897850036621094,
+      "learning_rate": 1.8456352924677575e-05,
+      "loss": 1.0466,
+      "step": 3555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9577382802963257,
+      "learning_rate": 1.845536125792629e-05,
+      "loss": 1.1555,
+      "step": 3556
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9770492315292358,
+      "learning_rate": 1.8454369299404106e-05,
+      "loss": 1.0394,
+      "step": 3557
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0685949325561523,
+      "learning_rate": 1.8453377049145254e-05,
+      "loss": 1.1444,
+      "step": 3558
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9234570264816284,
+      "learning_rate": 1.845238450718397e-05,
+      "loss": 1.0338,
+      "step": 3559
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9207348823547363,
+      "learning_rate": 1.8451391673554514e-05,
+      "loss": 1.0503,
+      "step": 3560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0540645122528076,
+      "learning_rate": 1.8450398548291135e-05,
+      "loss": 1.0758,
+      "step": 3561
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8427129983901978,
+      "learning_rate": 1.844940513142811e-05,
+      "loss": 1.0298,
+      "step": 3562
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7799781560897827,
+      "learning_rate": 1.8448411422999714e-05,
+      "loss": 1.0992,
+      "step": 3563
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.090266227722168,
+      "learning_rate": 1.844741742304024e-05,
+      "loss": 1.0868,
+      "step": 3564
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8227992057800293,
+      "learning_rate": 1.844642313158399e-05,
+      "loss": 0.9584,
+      "step": 3565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9015003442764282,
+      "learning_rate": 1.8445428548665268e-05,
+      "loss": 1.0416,
+      "step": 3566
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.138782501220703,
+      "learning_rate": 1.84444336743184e-05,
+      "loss": 1.0707,
+      "step": 3567
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1841679811477661,
+      "learning_rate": 1.8443438508577712e-05,
+      "loss": 0.6803,
+      "step": 3568
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.939046025276184,
+      "learning_rate": 1.844244305147755e-05,
+      "loss": 1.0225,
+      "step": 3569
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.09078848361969,
+      "learning_rate": 1.8441447303052262e-05,
+      "loss": 0.6489,
+      "step": 3570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1279072761535645,
+      "learning_rate": 1.8440451263336204e-05,
+      "loss": 0.9861,
+      "step": 3571
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.095539093017578,
+      "learning_rate": 1.8439454932363757e-05,
+      "loss": 0.9878,
+      "step": 3572
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.234347105026245,
+      "learning_rate": 1.8438458310169287e-05,
+      "loss": 1.0193,
+      "step": 3573
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1745190620422363,
+      "learning_rate": 1.8437461396787198e-05,
+      "loss": 1.0471,
+      "step": 3574
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1308064460754395,
+      "learning_rate": 1.843646419225188e-05,
+      "loss": 1.1014,
+      "step": 3575
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7232040166854858,
+      "learning_rate": 1.8435466696597758e-05,
+      "loss": 1.0707,
+      "step": 3576
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9038829803466797,
+      "learning_rate": 1.8434468909859235e-05,
+      "loss": 1.0573,
+      "step": 3577
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.886281132698059,
+      "learning_rate": 1.8433470832070758e-05,
+      "loss": 1.0526,
+      "step": 3578
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.254626512527466,
+      "learning_rate": 1.8432472463266754e-05,
+      "loss": 1.0314,
+      "step": 3579
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.07153058052063,
+      "learning_rate": 1.8431473803481682e-05,
+      "loss": 1.0698,
+      "step": 3580
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.941596269607544,
+      "learning_rate": 1.843047485275e-05,
+      "loss": 1.03,
+      "step": 3581
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8435593843460083,
+      "learning_rate": 1.8429475611106186e-05,
+      "loss": 1.1292,
+      "step": 3582
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9185932874679565,
+      "learning_rate": 1.842847607858471e-05,
+      "loss": 1.0221,
+      "step": 3583
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8758773803710938,
+      "learning_rate": 1.8427476255220074e-05,
+      "loss": 1.076,
+      "step": 3584
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9200859069824219,
+      "learning_rate": 1.842647614104677e-05,
+      "loss": 1.1302,
+      "step": 3585
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8744639158248901,
+      "learning_rate": 1.8425475736099316e-05,
+      "loss": 1.0557,
+      "step": 3586
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.028812885284424,
+      "learning_rate": 1.8424475040412224e-05,
+      "loss": 1.1477,
+      "step": 3587
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9398387670516968,
+      "learning_rate": 1.8423474054020034e-05,
+      "loss": 0.9862,
+      "step": 3588
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8377379179000854,
+      "learning_rate": 1.8422472776957287e-05,
+      "loss": 1.046,
+      "step": 3589
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.956353783607483,
+      "learning_rate": 1.8421471209258528e-05,
+      "loss": 1.0648,
+      "step": 3590
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8554823398590088,
+      "learning_rate": 1.8420469350958323e-05,
+      "loss": 1.0575,
+      "step": 3591
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1042957305908203,
+      "learning_rate": 1.8419467202091245e-05,
+      "loss": 1.0791,
+      "step": 3592
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.987033486366272,
+      "learning_rate": 1.841846476269187e-05,
+      "loss": 1.0193,
+      "step": 3593
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8211930990219116,
+      "learning_rate": 1.8417462032794792e-05,
+      "loss": 1.053,
+      "step": 3594
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3913238048553467,
+      "learning_rate": 1.8416459012434613e-05,
+      "loss": 1.0419,
+      "step": 3595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7238380908966064,
+      "learning_rate": 1.8415455701645942e-05,
+      "loss": 1.0998,
+      "step": 3596
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.732304334640503,
+      "learning_rate": 1.8414452100463407e-05,
+      "loss": 1.0255,
+      "step": 3597
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8758773803710938,
+      "learning_rate": 1.8413448208921632e-05,
+      "loss": 0.9784,
+      "step": 3598
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9735000133514404,
+      "learning_rate": 1.841244402705526e-05,
+      "loss": 1.0186,
+      "step": 3599
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9037452936172485,
+      "learning_rate": 1.8411439554898946e-05,
+      "loss": 1.0458,
+      "step": 3600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8801592588424683,
+      "learning_rate": 1.841043479248735e-05,
+      "loss": 1.0962,
+      "step": 3601
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.825833797454834,
+      "learning_rate": 1.8409429739855144e-05,
+      "loss": 1.1485,
+      "step": 3602
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.905579924583435,
+      "learning_rate": 1.8408424397037004e-05,
+      "loss": 0.9764,
+      "step": 3603
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1809136867523193,
+      "learning_rate": 1.8407418764067627e-05,
+      "loss": 1.1141,
+      "step": 3604
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0089707374572754,
+      "learning_rate": 1.8406412840981717e-05,
+      "loss": 1.0615,
+      "step": 3605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8194379806518555,
+      "learning_rate": 1.8405406627813978e-05,
+      "loss": 1.0775,
+      "step": 3606
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9196796417236328,
+      "learning_rate": 1.840440012459914e-05,
+      "loss": 1.0786,
+      "step": 3607
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8719857931137085,
+      "learning_rate": 1.8403393331371925e-05,
+      "loss": 1.1196,
+      "step": 3608
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9615795612335205,
+      "learning_rate": 1.8402386248167084e-05,
+      "loss": 0.9984,
+      "step": 3609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.973728060722351,
+      "learning_rate": 1.8401378875019366e-05,
+      "loss": 1.0674,
+      "step": 3610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8056793212890625,
+      "learning_rate": 1.840037121196353e-05,
+      "loss": 0.9996,
+      "step": 3611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.862863540649414,
+      "learning_rate": 1.8399363259034345e-05,
+      "loss": 1.0211,
+      "step": 3612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9475704431533813,
+      "learning_rate": 1.8398355016266604e-05,
+      "loss": 1.041,
+      "step": 3613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9655884504318237,
+      "learning_rate": 1.8397346483695085e-05,
+      "loss": 1.0674,
+      "step": 3614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.957646131515503,
+      "learning_rate": 1.8396337661354597e-05,
+      "loss": 1.1249,
+      "step": 3615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0246176719665527,
+      "learning_rate": 1.8395328549279955e-05,
+      "loss": 1.0956,
+      "step": 3616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9781476259231567,
+      "learning_rate": 1.839431914750597e-05,
+      "loss": 1.0352,
+      "step": 3617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0354580879211426,
+      "learning_rate": 1.8393309456067482e-05,
+      "loss": 0.9871,
+      "step": 3618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9261014461517334,
+      "learning_rate": 1.8392299474999333e-05,
+      "loss": 1.0737,
+      "step": 3619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8048653602600098,
+      "learning_rate": 1.839128920433637e-05,
+      "loss": 1.1286,
+      "step": 3620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.799453854560852,
+      "learning_rate": 1.8390278644113454e-05,
+      "loss": 1.1792,
+      "step": 3621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.790726900100708,
+      "learning_rate": 1.8389267794365465e-05,
+      "loss": 1.0064,
+      "step": 3622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0814177989959717,
+      "learning_rate": 1.8388256655127273e-05,
+      "loss": 1.0578,
+      "step": 3623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.81192147731781,
+      "learning_rate": 1.838724522643378e-05,
+      "loss": 1.0251,
+      "step": 3624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9143368005752563,
+      "learning_rate": 1.838623350831988e-05,
+      "loss": 1.1239,
+      "step": 3625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8702278137207031,
+      "learning_rate": 1.838522150082049e-05,
+      "loss": 1.1061,
+      "step": 3626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6882659196853638,
+      "learning_rate": 1.838420920397053e-05,
+      "loss": 0.9589,
+      "step": 3627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.862864375114441,
+      "learning_rate": 1.838319661780493e-05,
+      "loss": 1.0136,
+      "step": 3628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1542258262634277,
+      "learning_rate": 1.8382183742358627e-05,
+      "loss": 0.7157,
+      "step": 3629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9286577701568604,
+      "learning_rate": 1.8381170577666584e-05,
+      "loss": 1.0604,
+      "step": 3630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.041041135787964,
+      "learning_rate": 1.8380157123763755e-05,
+      "loss": 1.0544,
+      "step": 3631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.990922451019287,
+      "learning_rate": 1.837914338068511e-05,
+      "loss": 1.0884,
+      "step": 3632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0906779766082764,
+      "learning_rate": 1.8378129348465636e-05,
+      "loss": 0.6396,
+      "step": 3633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1289966106414795,
+      "learning_rate": 1.837711502714032e-05,
+      "loss": 0.9812,
+      "step": 3634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.916698694229126,
+      "learning_rate": 1.8376100416744166e-05,
+      "loss": 1.1153,
+      "step": 3635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.053574800491333,
+      "learning_rate": 1.8375085517312185e-05,
+      "loss": 1.0624,
+      "step": 3636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9801465272903442,
+      "learning_rate": 1.8374070328879395e-05,
+      "loss": 1.0463,
+      "step": 3637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.109030246734619,
+      "learning_rate": 1.8373054851480832e-05,
+      "loss": 1.1164,
+      "step": 3638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9094069004058838,
+      "learning_rate": 1.8372039085151537e-05,
+      "loss": 1.1147,
+      "step": 3639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9854366779327393,
+      "learning_rate": 1.8371023029926552e-05,
+      "loss": 1.0585,
+      "step": 3640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7335041761398315,
+      "learning_rate": 1.8370006685840953e-05,
+      "loss": 0.9966,
+      "step": 3641
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0833959579467773,
+      "learning_rate": 1.8368990052929804e-05,
+      "loss": 1.1091,
+      "step": 3642
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.952837347984314,
+      "learning_rate": 1.8367973131228182e-05,
+      "loss": 1.0576,
+      "step": 3643
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7813069820404053,
+      "learning_rate": 1.8366955920771183e-05,
+      "loss": 1.0638,
+      "step": 3644
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0346055030822754,
+      "learning_rate": 1.836593842159391e-05,
+      "loss": 0.9939,
+      "step": 3645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8170734643936157,
+      "learning_rate": 1.836492063373147e-05,
+      "loss": 0.9769,
+      "step": 3646
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9153265953063965,
+      "learning_rate": 1.8363902557218985e-05,
+      "loss": 1.1302,
+      "step": 3647
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9709967374801636,
+      "learning_rate": 1.8362884192091588e-05,
+      "loss": 1.1009,
+      "step": 3648
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7561345100402832,
+      "learning_rate": 1.8361865538384416e-05,
+      "loss": 1.0893,
+      "step": 3649
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8429702520370483,
+      "learning_rate": 1.8360846596132625e-05,
+      "loss": 1.0619,
+      "step": 3650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8159621953964233,
+      "learning_rate": 1.835982736537137e-05,
+      "loss": 1.0749,
+      "step": 3651
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2080602645874023,
+      "learning_rate": 1.8358807846135828e-05,
+      "loss": 1.0278,
+      "step": 3652
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0812442302703857,
+      "learning_rate": 1.8357788038461174e-05,
+      "loss": 1.1144,
+      "step": 3653
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.031829595565796,
+      "learning_rate": 1.83567679423826e-05,
+      "loss": 1.0162,
+      "step": 3654
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7498936653137207,
+      "learning_rate": 1.835574755793531e-05,
+      "loss": 1.0801,
+      "step": 3655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.816951036453247,
+      "learning_rate": 1.8354726885154512e-05,
+      "loss": 0.9975,
+      "step": 3656
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.783962368965149,
+      "learning_rate": 1.835370592407543e-05,
+      "loss": 1.0146,
+      "step": 3657
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.015430212020874,
+      "learning_rate": 1.8352684674733287e-05,
+      "loss": 1.0431,
+      "step": 3658
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9650129079818726,
+      "learning_rate": 1.8351663137163333e-05,
+      "loss": 1.0084,
+      "step": 3659
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8860795497894287,
+      "learning_rate": 1.8350641311400813e-05,
+      "loss": 1.0484,
+      "step": 3660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6969646215438843,
+      "learning_rate": 1.8349619197480985e-05,
+      "loss": 0.9705,
+      "step": 3661
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6013673543930054,
+      "learning_rate": 1.834859679543912e-05,
+      "loss": 1.0254,
+      "step": 3662
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8783289194107056,
+      "learning_rate": 1.8347574105310508e-05,
+      "loss": 1.134,
+      "step": 3663
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8577848672866821,
+      "learning_rate": 1.8346551127130424e-05,
+      "loss": 1.013,
+      "step": 3664
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7345956563949585,
+      "learning_rate": 1.834552786093418e-05,
+      "loss": 0.9427,
+      "step": 3665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9436713457107544,
+      "learning_rate": 1.834450430675708e-05,
+      "loss": 1.0893,
+      "step": 3666
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9044986963272095,
+      "learning_rate": 1.8343480464634448e-05,
+      "loss": 1.0126,
+      "step": 3667
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9315669536590576,
+      "learning_rate": 1.834245633460161e-05,
+      "loss": 1.0972,
+      "step": 3668
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8200615644454956,
+      "learning_rate": 1.8341431916693908e-05,
+      "loss": 0.9836,
+      "step": 3669
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.425615072250366,
+      "learning_rate": 1.8340407210946695e-05,
+      "loss": 1.0994,
+      "step": 3670
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1943169832229614,
+      "learning_rate": 1.833938221739532e-05,
+      "loss": 0.5752,
+      "step": 3671
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7126250267028809,
+      "learning_rate": 1.8338356936075165e-05,
+      "loss": 0.9713,
+      "step": 3672
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8821932077407837,
+      "learning_rate": 1.83373313670216e-05,
+      "loss": 1.1028,
+      "step": 3673
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.068502426147461,
+      "learning_rate": 1.8336305510270025e-05,
+      "loss": 1.0516,
+      "step": 3674
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.74851393699646,
+      "learning_rate": 1.833527936585583e-05,
+      "loss": 0.9962,
+      "step": 3675
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8025509119033813,
+      "learning_rate": 1.833425293381443e-05,
+      "loss": 0.9765,
+      "step": 3676
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7696292400360107,
+      "learning_rate": 1.8333226214181236e-05,
+      "loss": 1.1622,
+      "step": 3677
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8834055662155151,
+      "learning_rate": 1.833219920699169e-05,
+      "loss": 1.0331,
+      "step": 3678
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.986501693725586,
+      "learning_rate": 1.833117191228122e-05,
+      "loss": 1.0145,
+      "step": 3679
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8849607706069946,
+      "learning_rate": 1.8330144330085283e-05,
+      "loss": 1.0548,
+      "step": 3680
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0000438690185547,
+      "learning_rate": 1.8329116460439332e-05,
+      "loss": 1.1064,
+      "step": 3681
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.923536777496338,
+      "learning_rate": 1.832808830337884e-05,
+      "loss": 1.1025,
+      "step": 3682
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9162832498550415,
+      "learning_rate": 1.8327059858939283e-05,
+      "loss": 1.0627,
+      "step": 3683
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.022949457168579,
+      "learning_rate": 1.832603112715615e-05,
+      "loss": 1.0995,
+      "step": 3684
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.844857096672058,
+      "learning_rate": 1.832500210806494e-05,
+      "loss": 1.0216,
+      "step": 3685
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9799689054489136,
+      "learning_rate": 1.8323972801701166e-05,
+      "loss": 1.0612,
+      "step": 3686
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1007819175720215,
+      "learning_rate": 1.832294320810034e-05,
+      "loss": 1.064,
+      "step": 3687
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1841572523117065,
+      "learning_rate": 1.8321913327297997e-05,
+      "loss": 0.6001,
+      "step": 3688
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3432774543762207,
+      "learning_rate": 1.832088315932967e-05,
+      "loss": 1.0285,
+      "step": 3689
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.96401846408844,
+      "learning_rate": 1.8319852704230903e-05,
+      "loss": 1.02,
+      "step": 3690
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9793747663497925,
+      "learning_rate": 1.8318821962037266e-05,
+      "loss": 0.9902,
+      "step": 3691
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8920596837997437,
+      "learning_rate": 1.8317790932784315e-05,
+      "loss": 1.0429,
+      "step": 3692
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.148195505142212,
+      "learning_rate": 1.8316759616507637e-05,
+      "loss": 0.6211,
+      "step": 3693
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.043264865875244,
+      "learning_rate": 1.8315728013242816e-05,
+      "loss": 1.004,
+      "step": 3694
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2051589488983154,
+      "learning_rate": 1.8314696123025456e-05,
+      "loss": 1.0406,
+      "step": 3695
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.099808931350708,
+      "learning_rate": 1.8313663945891155e-05,
+      "loss": 1.0684,
+      "step": 3696
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8362714052200317,
+      "learning_rate": 1.8312631481875532e-05,
+      "loss": 1.0776,
+      "step": 3697
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0712838172912598,
+      "learning_rate": 1.8311598731014218e-05,
+      "loss": 1.0563,
+      "step": 3698
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.089859962463379,
+      "learning_rate": 1.831056569334285e-05,
+      "loss": 1.0467,
+      "step": 3699
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8799108266830444,
+      "learning_rate": 1.830953236889707e-05,
+      "loss": 1.0748,
+      "step": 3700
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9862091541290283,
+      "learning_rate": 1.8308498757712548e-05,
+      "loss": 1.0638,
+      "step": 3701
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.047929525375366,
+      "learning_rate": 1.830746485982494e-05,
+      "loss": 0.9358,
+      "step": 3702
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9723416566848755,
+      "learning_rate": 1.8306430675269922e-05,
+      "loss": 1.0568,
+      "step": 3703
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8958004713058472,
+      "learning_rate": 1.8305396204083185e-05,
+      "loss": 1.0427,
+      "step": 3704
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.230467677116394,
+      "learning_rate": 1.830436144630043e-05,
+      "loss": 0.577,
+      "step": 3705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8862000703811646,
+      "learning_rate": 1.8303326401957357e-05,
+      "loss": 1.0562,
+      "step": 3706
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2858850955963135,
+      "learning_rate": 1.830229107108968e-05,
+      "loss": 1.1023,
+      "step": 3707
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9600496292114258,
+      "learning_rate": 1.8301255453733135e-05,
+      "loss": 1.0856,
+      "step": 3708
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9136651754379272,
+      "learning_rate": 1.830021954992345e-05,
+      "loss": 1.1109,
+      "step": 3709
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7664834260940552,
+      "learning_rate": 1.8299183359696376e-05,
+      "loss": 1.0726,
+      "step": 3710
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7448902130126953,
+      "learning_rate": 1.8298146883087663e-05,
+      "loss": 1.0322,
+      "step": 3711
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8630681037902832,
+      "learning_rate": 1.8297110120133082e-05,
+      "loss": 1.0952,
+      "step": 3712
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9166920185089111,
+      "learning_rate": 1.8296073070868413e-05,
+      "loss": 1.1426,
+      "step": 3713
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.922372817993164,
+      "learning_rate": 1.8295035735329433e-05,
+      "loss": 1.0175,
+      "step": 3714
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.150103807449341,
+      "learning_rate": 1.8293998113551942e-05,
+      "loss": 1.1089,
+      "step": 3715
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8648571968078613,
+      "learning_rate": 1.8292960205571742e-05,
+      "loss": 1.0083,
+      "step": 3716
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9953773021697998,
+      "learning_rate": 1.8291922011424655e-05,
+      "loss": 1.0355,
+      "step": 3717
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.792277216911316,
+      "learning_rate": 1.82908835311465e-05,
+      "loss": 1.058,
+      "step": 3718
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.157439947128296,
+      "learning_rate": 1.828984476477311e-05,
+      "loss": 0.6291,
+      "step": 3719
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8322031497955322,
+      "learning_rate": 1.828880571234034e-05,
+      "loss": 1.0024,
+      "step": 3720
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0621535778045654,
+      "learning_rate": 1.8287766373884034e-05,
+      "loss": 1.0,
+      "step": 3721
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8825392723083496,
+      "learning_rate": 1.828672674944006e-05,
+      "loss": 1.0303,
+      "step": 3722
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9418011903762817,
+      "learning_rate": 1.82856868390443e-05,
+      "loss": 0.9897,
+      "step": 3723
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7633953094482422,
+      "learning_rate": 1.828464664273263e-05,
+      "loss": 1.0081,
+      "step": 3724
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.840912103652954,
+      "learning_rate": 1.8283606160540945e-05,
+      "loss": 1.0405,
+      "step": 3725
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8150430917739868,
+      "learning_rate": 1.8282565392505152e-05,
+      "loss": 1.0015,
+      "step": 3726
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.797620415687561,
+      "learning_rate": 1.8281524338661164e-05,
+      "loss": 0.977,
+      "step": 3727
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8623230457305908,
+      "learning_rate": 1.8280482999044905e-05,
+      "loss": 1.0798,
+      "step": 3728
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8047195672988892,
+      "learning_rate": 1.8279441373692307e-05,
+      "loss": 1.0924,
+      "step": 3729
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.004473924636841,
+      "learning_rate": 1.8278399462639314e-05,
+      "loss": 1.1145,
+      "step": 3730
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.983450174331665,
+      "learning_rate": 1.8277357265921885e-05,
+      "loss": 1.1546,
+      "step": 3731
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.800101637840271,
+      "learning_rate": 1.827631478357597e-05,
+      "loss": 1.0722,
+      "step": 3732
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7286827564239502,
+      "learning_rate": 1.827527201563756e-05,
+      "loss": 1.0384,
+      "step": 3733
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9980391263961792,
+      "learning_rate": 1.8274228962142623e-05,
+      "loss": 1.1244,
+      "step": 3734
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7999382019042969,
+      "learning_rate": 1.8273185623127162e-05,
+      "loss": 1.1046,
+      "step": 3735
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8693015575408936,
+      "learning_rate": 1.8272141998627172e-05,
+      "loss": 1.0314,
+      "step": 3736
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8798842430114746,
+      "learning_rate": 1.8271098088678667e-05,
+      "loss": 1.0771,
+      "step": 3737
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0632619857788086,
+      "learning_rate": 1.8270053893317675e-05,
+      "loss": 1.0244,
+      "step": 3738
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.885330319404602,
+      "learning_rate": 1.8269009412580223e-05,
+      "loss": 1.0393,
+      "step": 3739
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0820708274841309,
+      "learning_rate": 1.8267964646502356e-05,
+      "loss": 0.5998,
+      "step": 3740
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.836755633354187,
+      "learning_rate": 1.8266919595120126e-05,
+      "loss": 1.0356,
+      "step": 3741
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.150808334350586,
+      "learning_rate": 1.8265874258469593e-05,
+      "loss": 0.6624,
+      "step": 3742
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8213987350463867,
+      "learning_rate": 1.8264828636586824e-05,
+      "loss": 1.0488,
+      "step": 3743
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7294355630874634,
+      "learning_rate": 1.8263782729507912e-05,
+      "loss": 1.0708,
+      "step": 3744
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0099728107452393,
+      "learning_rate": 1.826273653726894e-05,
+      "loss": 0.5983,
+      "step": 3745
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9380937814712524,
+      "learning_rate": 1.8261690059906006e-05,
+      "loss": 1.1014,
+      "step": 3746
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1081117391586304,
+      "learning_rate": 1.826064329745523e-05,
+      "loss": 0.5906,
+      "step": 3747
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.816329836845398,
+      "learning_rate": 1.825959624995273e-05,
+      "loss": 1.0881,
+      "step": 3748
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6400295495986938,
+      "learning_rate": 1.825854891743464e-05,
+      "loss": 1.0661,
+      "step": 3749
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.898664951324463,
+      "learning_rate": 1.825750129993709e-05,
+      "loss": 1.0862,
+      "step": 3750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7420895099639893,
+      "learning_rate": 1.8256453397496233e-05,
+      "loss": 1.073,
+      "step": 3751
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8840978145599365,
+      "learning_rate": 1.825540521014824e-05,
+      "loss": 1.0515,
+      "step": 3752
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8699301481246948,
+      "learning_rate": 1.825435673792927e-05,
+      "loss": 1.0868,
+      "step": 3753
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8465110063552856,
+      "learning_rate": 1.825330798087551e-05,
+      "loss": 1.0654,
+      "step": 3754
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9346208572387695,
+      "learning_rate": 1.825225893902314e-05,
+      "loss": 1.1373,
+      "step": 3755
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7104047536849976,
+      "learning_rate": 1.8251209612408375e-05,
+      "loss": 1.0191,
+      "step": 3756
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9211071729660034,
+      "learning_rate": 1.8250160001067408e-05,
+      "loss": 0.9918,
+      "step": 3757
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.994146466255188,
+      "learning_rate": 1.8249110105036468e-05,
+      "loss": 1.0419,
+      "step": 3758
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9202604293823242,
+      "learning_rate": 1.824805992435178e-05,
+      "loss": 1.0157,
+      "step": 3759
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7936993837356567,
+      "learning_rate": 1.8247009459049585e-05,
+      "loss": 1.0232,
+      "step": 3760
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9010344743728638,
+      "learning_rate": 1.824595870916613e-05,
+      "loss": 1.0992,
+      "step": 3761
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8472884893417358,
+      "learning_rate": 1.8244907674737672e-05,
+      "loss": 0.9942,
+      "step": 3762
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7190096378326416,
+      "learning_rate": 1.8243856355800485e-05,
+      "loss": 1.138,
+      "step": 3763
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8474845886230469,
+      "learning_rate": 1.8242804752390844e-05,
+      "loss": 1.1125,
+      "step": 3764
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1347293853759766,
+      "learning_rate": 1.8241752864545032e-05,
+      "loss": 0.6192,
+      "step": 3765
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.795087218284607,
+      "learning_rate": 1.8240700692299357e-05,
+      "loss": 1.0523,
+      "step": 3766
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9675209522247314,
+      "learning_rate": 1.8239648235690115e-05,
+      "loss": 1.0007,
+      "step": 3767
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8827106952667236,
+      "learning_rate": 1.8238595494753633e-05,
+      "loss": 1.1101,
+      "step": 3768
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8106427192687988,
+      "learning_rate": 1.823754246952623e-05,
+      "loss": 1.0869,
+      "step": 3769
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9490290880203247,
+      "learning_rate": 1.8236489160044247e-05,
+      "loss": 0.9808,
+      "step": 3770
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.100027322769165,
+      "learning_rate": 1.8235435566344034e-05,
+      "loss": 1.0246,
+      "step": 3771
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7031404972076416,
+      "learning_rate": 1.8234381688461943e-05,
+      "loss": 1.0067,
+      "step": 3772
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6785268783569336,
+      "learning_rate": 1.8233327526434342e-05,
+      "loss": 1.0026,
+      "step": 3773
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.702819585800171,
+      "learning_rate": 1.8232273080297606e-05,
+      "loss": 1.0284,
+      "step": 3774
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9300413131713867,
+      "learning_rate": 1.8231218350088124e-05,
+      "loss": 1.0865,
+      "step": 3775
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0553581714630127,
+      "learning_rate": 1.8230163335842288e-05,
+      "loss": 1.0674,
+      "step": 3776
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7888224124908447,
+      "learning_rate": 1.822910803759651e-05,
+      "loss": 0.9917,
+      "step": 3777
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0336437225341797,
+      "learning_rate": 1.8228052455387194e-05,
+      "loss": 1.0064,
+      "step": 3778
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.813965082168579,
+      "learning_rate": 1.8226996589250775e-05,
+      "loss": 1.079,
+      "step": 3779
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9376791715621948,
+      "learning_rate": 1.8225940439223684e-05,
+      "loss": 1.0624,
+      "step": 3780
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9597405195236206,
+      "learning_rate": 1.8224884005342367e-05,
+      "loss": 1.1065,
+      "step": 3781
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8007218837738037,
+      "learning_rate": 1.822382728764328e-05,
+      "loss": 0.9965,
+      "step": 3782
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.034085273742676,
+      "learning_rate": 1.8222770286162884e-05,
+      "loss": 1.0951,
+      "step": 3783
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0285332202911377,
+      "learning_rate": 1.8221713000937653e-05,
+      "loss": 0.9982,
+      "step": 3784
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7554453611373901,
+      "learning_rate": 1.8220655432004073e-05,
+      "loss": 1.018,
+      "step": 3785
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8582874536514282,
+      "learning_rate": 1.821959757939864e-05,
+      "loss": 0.9533,
+      "step": 3786
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.01412296295166,
+      "learning_rate": 1.8218539443157855e-05,
+      "loss": 1.0441,
+      "step": 3787
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8150533437728882,
+      "learning_rate": 1.8217481023318232e-05,
+      "loss": 1.0775,
+      "step": 3788
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.001422166824341,
+      "learning_rate": 1.8216422319916288e-05,
+      "loss": 1.0918,
+      "step": 3789
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7037065029144287,
+      "learning_rate": 1.8215363332988568e-05,
+      "loss": 0.9502,
+      "step": 3790
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.753208041191101,
+      "learning_rate": 1.8214304062571605e-05,
+      "loss": 1.0031,
+      "step": 3791
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0934661626815796,
+      "learning_rate": 1.821324450870195e-05,
+      "loss": 0.5973,
+      "step": 3792
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9076343774795532,
+      "learning_rate": 1.821218467141618e-05,
+      "loss": 1.0807,
+      "step": 3793
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.901941180229187,
+      "learning_rate": 1.8211124550750853e-05,
+      "loss": 1.0313,
+      "step": 3794
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.100982666015625,
+      "learning_rate": 1.821006414674255e-05,
+      "loss": 1.1513,
+      "step": 3795
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.02571439743042,
+      "learning_rate": 1.820900345942787e-05,
+      "loss": 0.6006,
+      "step": 3796
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.005312204360962,
+      "learning_rate": 1.8207942488843416e-05,
+      "loss": 1.0021,
+      "step": 3797
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7871723175048828,
+      "learning_rate": 1.8206881235025786e-05,
+      "loss": 1.0317,
+      "step": 3798
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0059335231781006,
+      "learning_rate": 1.8205819698011615e-05,
+      "loss": 1.1389,
+      "step": 3799
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8365846872329712,
+      "learning_rate": 1.8204757877837528e-05,
+      "loss": 1.0041,
+      "step": 3800
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8571128845214844,
+      "learning_rate": 1.8203695774540167e-05,
+      "loss": 1.0772,
+      "step": 3801
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8227801322937012,
+      "learning_rate": 1.8202633388156176e-05,
+      "loss": 1.0904,
+      "step": 3802
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7804046869277954,
+      "learning_rate": 1.8201570718722225e-05,
+      "loss": 1.1316,
+      "step": 3803
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7756474018096924,
+      "learning_rate": 1.8200507766274978e-05,
+      "loss": 1.0126,
+      "step": 3804
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8382691144943237,
+      "learning_rate": 1.819944453085111e-05,
+      "loss": 1.0328,
+      "step": 3805
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8535228967666626,
+      "learning_rate": 1.8198381012487322e-05,
+      "loss": 1.0575,
+      "step": 3806
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9102948904037476,
+      "learning_rate": 1.8197317211220302e-05,
+      "loss": 1.0751,
+      "step": 3807
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9193061590194702,
+      "learning_rate": 1.8196253127086765e-05,
+      "loss": 1.0632,
+      "step": 3808
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9411903619766235,
+      "learning_rate": 1.819518876012343e-05,
+      "loss": 1.0677,
+      "step": 3809
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9638195037841797,
+      "learning_rate": 1.819412411036702e-05,
+      "loss": 1.0493,
+      "step": 3810
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4751646518707275,
+      "learning_rate": 1.8193059177854278e-05,
+      "loss": 0.6146,
+      "step": 3811
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.757002592086792,
+      "learning_rate": 1.819199396262195e-05,
+      "loss": 1.0057,
+      "step": 3812
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2763197422027588,
+      "learning_rate": 1.819092846470679e-05,
+      "loss": 0.6732,
+      "step": 3813
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.876028299331665,
+      "learning_rate": 1.8189862684145577e-05,
+      "loss": 1.0339,
+      "step": 3814
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.818593978881836,
+      "learning_rate": 1.8188796620975073e-05,
+      "loss": 1.0312,
+      "step": 3815
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.893667459487915,
+      "learning_rate": 1.8187730275232075e-05,
+      "loss": 1.0567,
+      "step": 3816
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9479000568389893,
+      "learning_rate": 1.8186663646953376e-05,
+      "loss": 0.9315,
+      "step": 3817
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8655383586883545,
+      "learning_rate": 1.8185596736175782e-05,
+      "loss": 0.9801,
+      "step": 3818
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8937088251113892,
+      "learning_rate": 1.8184529542936113e-05,
+      "loss": 0.9915,
+      "step": 3819
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.815807819366455,
+      "learning_rate": 1.8183462067271193e-05,
+      "loss": 0.9819,
+      "step": 3820
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8278753757476807,
+      "learning_rate": 1.8182394309217852e-05,
+      "loss": 1.0679,
+      "step": 3821
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7043217420578003,
+      "learning_rate": 1.8181326268812946e-05,
+      "loss": 0.9628,
+      "step": 3822
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6850651502609253,
+      "learning_rate": 1.8180257946093317e-05,
+      "loss": 1.0968,
+      "step": 3823
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8540571928024292,
+      "learning_rate": 1.817918934109584e-05,
+      "loss": 1.0846,
+      "step": 3824
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.836405873298645,
+      "learning_rate": 1.817812045385739e-05,
+      "loss": 0.9513,
+      "step": 3825
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0068178176879883,
+      "learning_rate": 1.8177051284414844e-05,
+      "loss": 0.9994,
+      "step": 3826
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.716103434562683,
+      "learning_rate": 1.81759818328051e-05,
+      "loss": 0.5534,
+      "step": 3827
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8814142942428589,
+      "learning_rate": 1.817491209906506e-05,
+      "loss": 1.1194,
+      "step": 3828
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9128639698028564,
+      "learning_rate": 1.8173842083231643e-05,
+      "loss": 1.0408,
+      "step": 3829
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.016286611557007,
+      "learning_rate": 1.8172771785341766e-05,
+      "loss": 1.0341,
+      "step": 3830
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.343295693397522,
+      "learning_rate": 1.8171701205432365e-05,
+      "loss": 0.6534,
+      "step": 3831
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9161622524261475,
+      "learning_rate": 1.8170630343540382e-05,
+      "loss": 1.0144,
+      "step": 3832
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7268056869506836,
+      "learning_rate": 1.816955919970277e-05,
+      "loss": 1.0105,
+      "step": 3833
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9056082963943481,
+      "learning_rate": 1.8168487773956493e-05,
+      "loss": 1.0088,
+      "step": 3834
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8977272510528564,
+      "learning_rate": 1.8167416066338518e-05,
+      "loss": 1.0595,
+      "step": 3835
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8341658115386963,
+      "learning_rate": 1.816634407688583e-05,
+      "loss": 1.053,
+      "step": 3836
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.966023325920105,
+      "learning_rate": 1.816527180563542e-05,
+      "loss": 1.1511,
+      "step": 3837
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.091402292251587,
+      "learning_rate": 1.8164199252624285e-05,
+      "loss": 1.1314,
+      "step": 3838
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8178540468215942,
+      "learning_rate": 1.8163126417889444e-05,
+      "loss": 1.052,
+      "step": 3839
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1357970237731934,
+      "learning_rate": 1.816205330146791e-05,
+      "loss": 1.0246,
+      "step": 3840
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8733441829681396,
+      "learning_rate": 1.816097990339672e-05,
+      "loss": 1.1315,
+      "step": 3841
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9849085807800293,
+      "learning_rate": 1.815990622371291e-05,
+      "loss": 1.0377,
+      "step": 3842
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1089699268341064,
+      "learning_rate": 1.815883226245353e-05,
+      "loss": 1.0951,
+      "step": 3843
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.814235806465149,
+      "learning_rate": 1.8157758019655633e-05,
+      "loss": 1.0507,
+      "step": 3844
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7230116128921509,
+      "learning_rate": 1.81566834953563e-05,
+      "loss": 1.0042,
+      "step": 3845
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7044602632522583,
+      "learning_rate": 1.8155608689592604e-05,
+      "loss": 1.0332,
+      "step": 3846
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.63759183883667,
+      "learning_rate": 1.8154533602401634e-05,
+      "loss": 1.0546,
+      "step": 3847
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8536738157272339,
+      "learning_rate": 1.8153458233820487e-05,
+      "loss": 1.0999,
+      "step": 3848
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7037180662155151,
+      "learning_rate": 1.8152382583886272e-05,
+      "loss": 1.0081,
+      "step": 3849
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.159395933151245,
+      "learning_rate": 1.815130665263611e-05,
+      "loss": 1.0148,
+      "step": 3850
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7641210556030273,
+      "learning_rate": 1.815023044010712e-05,
+      "loss": 1.0557,
+      "step": 3851
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.738863468170166,
+      "learning_rate": 1.8149153946336448e-05,
+      "loss": 0.9173,
+      "step": 3852
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9852241277694702,
+      "learning_rate": 1.8148077171361237e-05,
+      "loss": 1.0803,
+      "step": 3853
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.66620934009552,
+      "learning_rate": 1.814700011521864e-05,
+      "loss": 1.047,
+      "step": 3854
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9558688402175903,
+      "learning_rate": 1.8145922777945832e-05,
+      "loss": 1.0513,
+      "step": 3855
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1312966346740723,
+      "learning_rate": 1.814484515957998e-05,
+      "loss": 1.0644,
+      "step": 3856
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1521785259246826,
+      "learning_rate": 1.8143767260158272e-05,
+      "loss": 1.0731,
+      "step": 3857
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9486254453659058,
+      "learning_rate": 1.8142689079717908e-05,
+      "loss": 1.0388,
+      "step": 3858
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9918071031570435,
+      "learning_rate": 1.8141610618296087e-05,
+      "loss": 0.9962,
+      "step": 3859
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.089367389678955,
+      "learning_rate": 1.814053187593003e-05,
+      "loss": 1.0867,
+      "step": 3860
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.987486481666565,
+      "learning_rate": 1.8139452852656955e-05,
+      "loss": 1.0599,
+      "step": 3861
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.958200216293335,
+      "learning_rate": 1.81383735485141e-05,
+      "loss": 0.9996,
+      "step": 3862
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8884671926498413,
+      "learning_rate": 1.8137293963538705e-05,
+      "loss": 1.0481,
+      "step": 3863
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.927946925163269,
+      "learning_rate": 1.813621409776803e-05,
+      "loss": 1.0794,
+      "step": 3864
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9613637924194336,
+      "learning_rate": 1.8135133951239327e-05,
+      "loss": 1.0209,
+      "step": 3865
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9980244636535645,
+      "learning_rate": 1.8134053523989883e-05,
+      "loss": 1.145,
+      "step": 3866
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8582040071487427,
+      "learning_rate": 1.813297281605697e-05,
+      "loss": 1.0088,
+      "step": 3867
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.085115432739258,
+      "learning_rate": 1.8131891827477884e-05,
+      "loss": 1.1203,
+      "step": 3868
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1039021015167236,
+      "learning_rate": 1.8130810558289925e-05,
+      "loss": 0.9948,
+      "step": 3869
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7838964462280273,
+      "learning_rate": 1.812972900853041e-05,
+      "loss": 1.0531,
+      "step": 3870
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8884449005126953,
+      "learning_rate": 1.8128647178236654e-05,
+      "loss": 0.9906,
+      "step": 3871
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8240293264389038,
+      "learning_rate": 1.812756506744599e-05,
+      "loss": 1.0183,
+      "step": 3872
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8504832983016968,
+      "learning_rate": 1.812648267619576e-05,
+      "loss": 1.059,
+      "step": 3873
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.988913893699646,
+      "learning_rate": 1.8125400004523316e-05,
+      "loss": 1.1195,
+      "step": 3874
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8549951314926147,
+      "learning_rate": 1.812431705246601e-05,
+      "loss": 1.0648,
+      "step": 3875
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9300460815429688,
+      "learning_rate": 1.812323382006122e-05,
+      "loss": 1.1128,
+      "step": 3876
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8917664289474487,
+      "learning_rate": 1.812215030734632e-05,
+      "loss": 1.0421,
+      "step": 3877
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0504982471466064,
+      "learning_rate": 1.8121066514358703e-05,
+      "loss": 0.9862,
+      "step": 3878
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9585716724395752,
+      "learning_rate": 1.8119982441135765e-05,
+      "loss": 1.0213,
+      "step": 3879
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8183696269989014,
+      "learning_rate": 1.8118898087714915e-05,
+      "loss": 0.9669,
+      "step": 3880
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.857771396636963,
+      "learning_rate": 1.8117813454133574e-05,
+      "loss": 0.9849,
+      "step": 3881
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.001291036605835,
+      "learning_rate": 1.8116728540429164e-05,
+      "loss": 1.0847,
+      "step": 3882
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9891024827957153,
+      "learning_rate": 1.8115643346639124e-05,
+      "loss": 1.0465,
+      "step": 3883
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9021621942520142,
+      "learning_rate": 1.8114557872800906e-05,
+      "loss": 1.0694,
+      "step": 3884
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8179564476013184,
+      "learning_rate": 1.811347211895196e-05,
+      "loss": 0.9768,
+      "step": 3885
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8908690214157104,
+      "learning_rate": 1.8112386085129757e-05,
+      "loss": 1.0638,
+      "step": 3886
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0184385776519775,
+      "learning_rate": 1.811129977137177e-05,
+      "loss": 1.0277,
+      "step": 3887
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.994249939918518,
+      "learning_rate": 1.811021317771549e-05,
+      "loss": 1.023,
+      "step": 3888
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9378025531768799,
+      "learning_rate": 1.8109126304198402e-05,
+      "loss": 1.0506,
+      "step": 3889
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.799944281578064,
+      "learning_rate": 1.8108039150858018e-05,
+      "loss": 1.0305,
+      "step": 3890
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7559455633163452,
+      "learning_rate": 1.8106951717731854e-05,
+      "loss": 1.0524,
+      "step": 3891
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.690826654434204,
+      "learning_rate": 1.8105864004857433e-05,
+      "loss": 0.9264,
+      "step": 3892
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9688596725463867,
+      "learning_rate": 1.810477601227229e-05,
+      "loss": 1.0857,
+      "step": 3893
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3253986835479736,
+      "learning_rate": 1.8103687740013958e-05,
+      "loss": 1.0815,
+      "step": 3894
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7403953075408936,
+      "learning_rate": 1.8102599188120006e-05,
+      "loss": 1.009,
+      "step": 3895
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8972994089126587,
+      "learning_rate": 1.810151035662799e-05,
+      "loss": 1.1136,
+      "step": 3896
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8872801065444946,
+      "learning_rate": 1.810042124557548e-05,
+      "loss": 0.964,
+      "step": 3897
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.848447561264038,
+      "learning_rate": 1.809933185500006e-05,
+      "loss": 1.0557,
+      "step": 3898
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.015592575073242,
+      "learning_rate": 1.8098242184939324e-05,
+      "loss": 1.05,
+      "step": 3899
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.927268147468567,
+      "learning_rate": 1.8097152235430872e-05,
+      "loss": 1.0204,
+      "step": 3900
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1515603065490723,
+      "learning_rate": 1.8096062006512308e-05,
+      "loss": 1.0896,
+      "step": 3901
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.051318645477295,
+      "learning_rate": 1.809497149822127e-05,
+      "loss": 1.0001,
+      "step": 3902
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9856501817703247,
+      "learning_rate": 1.8093880710595372e-05,
+      "loss": 1.0594,
+      "step": 3903
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1264731884002686,
+      "learning_rate": 1.809278964367226e-05,
+      "loss": 1.0749,
+      "step": 3904
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.848768711090088,
+      "learning_rate": 1.8091698297489585e-05,
+      "loss": 1.1265,
+      "step": 3905
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8376882076263428,
+      "learning_rate": 1.8090606672085003e-05,
+      "loss": 1.0638,
+      "step": 3906
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8967822790145874,
+      "learning_rate": 1.808951476749619e-05,
+      "loss": 0.9877,
+      "step": 3907
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7814844846725464,
+      "learning_rate": 1.8088422583760814e-05,
+      "loss": 1.0653,
+      "step": 3908
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.190186023712158,
+      "learning_rate": 1.808733012091657e-05,
+      "loss": 0.9881,
+      "step": 3909
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9718918800354004,
+      "learning_rate": 1.8086237379001156e-05,
+      "loss": 1.0011,
+      "step": 3910
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1417131423950195,
+      "learning_rate": 1.8085144358052276e-05,
+      "loss": 1.1141,
+      "step": 3911
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9288969039916992,
+      "learning_rate": 1.808405105810765e-05,
+      "loss": 1.0754,
+      "step": 3912
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8807377815246582,
+      "learning_rate": 1.8082957479205006e-05,
+      "loss": 1.122,
+      "step": 3913
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9672008752822876,
+      "learning_rate": 1.8081863621382075e-05,
+      "loss": 1.0566,
+      "step": 3914
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0628933906555176,
+      "learning_rate": 1.8080769484676613e-05,
+      "loss": 1.0905,
+      "step": 3915
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9678635597229004,
+      "learning_rate": 1.807967506912636e-05,
+      "loss": 1.049,
+      "step": 3916
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9444068670272827,
+      "learning_rate": 1.8078580374769096e-05,
+      "loss": 0.9538,
+      "step": 3917
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2512242794036865,
+      "learning_rate": 1.8077485401642586e-05,
+      "loss": 1.1235,
+      "step": 3918
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9790058135986328,
+      "learning_rate": 1.8076390149784622e-05,
+      "loss": 1.0946,
+      "step": 3919
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9601472616195679,
+      "learning_rate": 1.807529461923299e-05,
+      "loss": 1.0174,
+      "step": 3920
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0217413902282715,
+      "learning_rate": 1.80741988100255e-05,
+      "loss": 1.0095,
+      "step": 3921
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9320117235183716,
+      "learning_rate": 1.8073102722199966e-05,
+      "loss": 0.9607,
+      "step": 3922
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.010375499725342,
+      "learning_rate": 1.8072006355794206e-05,
+      "loss": 1.0236,
+      "step": 3923
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8503814935684204,
+      "learning_rate": 1.8070909710846053e-05,
+      "loss": 1.0506,
+      "step": 3924
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0421550273895264,
+      "learning_rate": 1.8069812787393355e-05,
+      "loss": 1.0711,
+      "step": 3925
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2336549758911133,
+      "learning_rate": 1.8068715585473957e-05,
+      "loss": 0.6175,
+      "step": 3926
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.07673716545105,
+      "learning_rate": 1.8067618105125725e-05,
+      "loss": 1.0859,
+      "step": 3927
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8096039295196533,
+      "learning_rate": 1.8066520346386526e-05,
+      "loss": 1.031,
+      "step": 3928
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8365750312805176,
+      "learning_rate": 1.8065422309294245e-05,
+      "loss": 0.9916,
+      "step": 3929
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8973675966262817,
+      "learning_rate": 1.8064323993886768e-05,
+      "loss": 1.0873,
+      "step": 3930
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7288068532943726,
+      "learning_rate": 1.8063225400202e-05,
+      "loss": 1.088,
+      "step": 3931
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8798779249191284,
+      "learning_rate": 1.8062126528277846e-05,
+      "loss": 1.0496,
+      "step": 3932
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0560450553894043,
+      "learning_rate": 1.8061027378152224e-05,
+      "loss": 1.0716,
+      "step": 3933
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1856849193572998,
+      "learning_rate": 1.8059927949863066e-05,
+      "loss": 0.6826,
+      "step": 3934
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9146655797958374,
+      "learning_rate": 1.8058828243448308e-05,
+      "loss": 1.0853,
+      "step": 3935
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7568093538284302,
+      "learning_rate": 1.8057728258945902e-05,
+      "loss": 0.9345,
+      "step": 3936
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7870181798934937,
+      "learning_rate": 1.8056627996393797e-05,
+      "loss": 1.0161,
+      "step": 3937
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8653303384780884,
+      "learning_rate": 1.8055527455829968e-05,
+      "loss": 1.054,
+      "step": 3938
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.840993881225586,
+      "learning_rate": 1.805442663729239e-05,
+      "loss": 1.069,
+      "step": 3939
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0293118953704834,
+      "learning_rate": 1.8053325540819048e-05,
+      "loss": 1.0657,
+      "step": 3940
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6711387634277344,
+      "learning_rate": 1.8052224166447936e-05,
+      "loss": 0.9387,
+      "step": 3941
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9441683292388916,
+      "learning_rate": 1.805112251421706e-05,
+      "loss": 1.0373,
+      "step": 3942
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8204078674316406,
+      "learning_rate": 1.8050020584164437e-05,
+      "loss": 1.1179,
+      "step": 3943
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8146203756332397,
+      "learning_rate": 1.804891837632809e-05,
+      "loss": 1.0273,
+      "step": 3944
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9963690042495728,
+      "learning_rate": 1.8047815890746056e-05,
+      "loss": 1.1278,
+      "step": 3945
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7636953592300415,
+      "learning_rate": 1.8046713127456375e-05,
+      "loss": 1.1083,
+      "step": 3946
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7529535293579102,
+      "learning_rate": 1.80456100864971e-05,
+      "loss": 1.0615,
+      "step": 3947
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.898754596710205,
+      "learning_rate": 1.8044506767906297e-05,
+      "loss": 1.0569,
+      "step": 3948
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.186434268951416,
+      "learning_rate": 1.8043403171722034e-05,
+      "loss": 0.6396,
+      "step": 3949
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8913542032241821,
+      "learning_rate": 1.80422992979824e-05,
+      "loss": 1.1204,
+      "step": 3950
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8315101861953735,
+      "learning_rate": 1.804119514672548e-05,
+      "loss": 0.9885,
+      "step": 3951
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8892968893051147,
+      "learning_rate": 1.8040090717989378e-05,
+      "loss": 1.0374,
+      "step": 3952
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.880257248878479,
+      "learning_rate": 1.8038986011812203e-05,
+      "loss": 1.0305,
+      "step": 3953
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7758201360702515,
+      "learning_rate": 1.803788102823208e-05,
+      "loss": 1.0407,
+      "step": 3954
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9291921854019165,
+      "learning_rate": 1.8036775767287135e-05,
+      "loss": 0.9676,
+      "step": 3955
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.881773829460144,
+      "learning_rate": 1.803567022901551e-05,
+      "loss": 1.0348,
+      "step": 3956
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.04146146774292,
+      "learning_rate": 1.8034564413455345e-05,
+      "loss": 1.0909,
+      "step": 3957
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9385111331939697,
+      "learning_rate": 1.803345832064481e-05,
+      "loss": 1.0612,
+      "step": 3958
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8683481216430664,
+      "learning_rate": 1.8032351950622068e-05,
+      "loss": 1.0705,
+      "step": 3959
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9467734098434448,
+      "learning_rate": 1.80312453034253e-05,
+      "loss": 1.0596,
+      "step": 3960
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8971284627914429,
+      "learning_rate": 1.8030138379092688e-05,
+      "loss": 1.0351,
+      "step": 3961
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8684958219528198,
+      "learning_rate": 1.8029031177662434e-05,
+      "loss": 1.0458,
+      "step": 3962
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8525906801223755,
+      "learning_rate": 1.802792369917274e-05,
+      "loss": 1.057,
+      "step": 3963
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.958800196647644,
+      "learning_rate": 1.8026815943661828e-05,
+      "loss": 1.056,
+      "step": 3964
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.715982437133789,
+      "learning_rate": 1.802570791116792e-05,
+      "loss": 1.1179,
+      "step": 3965
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.19415283203125,
+      "learning_rate": 1.8024599601729245e-05,
+      "loss": 1.0545,
+      "step": 3966
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9145385026931763,
+      "learning_rate": 1.802349101538406e-05,
+      "loss": 1.0073,
+      "step": 3967
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8941882848739624,
+      "learning_rate": 1.8022382152170607e-05,
+      "loss": 1.0987,
+      "step": 3968
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9278075695037842,
+      "learning_rate": 1.802127301212716e-05,
+      "loss": 1.0493,
+      "step": 3969
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8007793426513672,
+      "learning_rate": 1.8020163595291987e-05,
+      "loss": 0.9988,
+      "step": 3970
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7699190378189087,
+      "learning_rate": 1.801905390170337e-05,
+      "loss": 0.9807,
+      "step": 3971
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8725440502166748,
+      "learning_rate": 1.8017943931399604e-05,
+      "loss": 1.0459,
+      "step": 3972
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.173238754272461,
+      "learning_rate": 1.801683368441899e-05,
+      "loss": 1.0929,
+      "step": 3973
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0531492233276367,
+      "learning_rate": 1.801572316079984e-05,
+      "loss": 1.0759,
+      "step": 3974
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2277255058288574,
+      "learning_rate": 1.8014612360580477e-05,
+      "loss": 1.0659,
+      "step": 3975
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7957488298416138,
+      "learning_rate": 1.8013501283799225e-05,
+      "loss": 1.0445,
+      "step": 3976
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.010195255279541,
+      "learning_rate": 1.801238993049443e-05,
+      "loss": 1.0546,
+      "step": 3977
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.798598051071167,
+      "learning_rate": 1.8011278300704443e-05,
+      "loss": 0.9403,
+      "step": 3978
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0803163051605225,
+      "learning_rate": 1.8010166394467617e-05,
+      "loss": 1.0313,
+      "step": 3979
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1398024559020996,
+      "learning_rate": 1.8009054211822324e-05,
+      "loss": 0.6528,
+      "step": 3980
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8305085897445679,
+      "learning_rate": 1.800794175280695e-05,
+      "loss": 1.0046,
+      "step": 3981
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8218560218811035,
+      "learning_rate": 1.8006829017459868e-05,
+      "loss": 1.0187,
+      "step": 3982
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1577584743499756,
+      "learning_rate": 1.8005716005819482e-05,
+      "loss": 1.1369,
+      "step": 3983
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9718064069747925,
+      "learning_rate": 1.8004602717924204e-05,
+      "loss": 1.0778,
+      "step": 3984
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9038029909133911,
+      "learning_rate": 1.800348915381244e-05,
+      "loss": 1.0723,
+      "step": 3985
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7515453100204468,
+      "learning_rate": 1.800237531352263e-05,
+      "loss": 1.0191,
+      "step": 3986
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9517369270324707,
+      "learning_rate": 1.8001261197093196e-05,
+      "loss": 1.0386,
+      "step": 3987
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7540301084518433,
+      "learning_rate": 1.800014680456259e-05,
+      "loss": 1.1015,
+      "step": 3988
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.024225950241089,
+      "learning_rate": 1.7999032135969265e-05,
+      "loss": 1.0741,
+      "step": 3989
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8054072856903076,
+      "learning_rate": 1.7997917191351688e-05,
+      "loss": 0.9977,
+      "step": 3990
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7811944484710693,
+      "learning_rate": 1.7996801970748326e-05,
+      "loss": 1.0172,
+      "step": 3991
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9467167854309082,
+      "learning_rate": 1.799568647419767e-05,
+      "loss": 1.0672,
+      "step": 3992
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.906733512878418,
+      "learning_rate": 1.7994570701738208e-05,
+      "loss": 0.9953,
+      "step": 3993
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9821357727050781,
+      "learning_rate": 1.7993454653408443e-05,
+      "loss": 1.0176,
+      "step": 3994
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.114565372467041,
+      "learning_rate": 1.7992338329246885e-05,
+      "loss": 1.0517,
+      "step": 3995
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8689619302749634,
+      "learning_rate": 1.799122172929206e-05,
+      "loss": 1.0324,
+      "step": 3996
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0412685871124268,
+      "learning_rate": 1.7990104853582494e-05,
+      "loss": 1.1121,
+      "step": 3997
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.068769931793213,
+      "learning_rate": 1.7988987702156725e-05,
+      "loss": 1.1178,
+      "step": 3998
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9666004180908203,
+      "learning_rate": 1.798787027505331e-05,
+      "loss": 0.9617,
+      "step": 3999
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1355674266815186,
+      "learning_rate": 1.798675257231081e-05,
+      "loss": 1.0383,
+      "step": 4000
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.835336685180664,
+      "learning_rate": 1.7985634593967782e-05,
+      "loss": 1.0279,
+      "step": 4001
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9524002075195312,
+      "learning_rate": 1.7984516340062814e-05,
+      "loss": 1.1141,
+      "step": 4002
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0472798347473145,
+      "learning_rate": 1.7983397810634488e-05,
+      "loss": 1.0881,
+      "step": 4003
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8359564542770386,
+      "learning_rate": 1.7982279005721408e-05,
+      "loss": 1.0657,
+      "step": 4004
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8843165636062622,
+      "learning_rate": 1.7981159925362174e-05,
+      "loss": 0.9319,
+      "step": 4005
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9676355123519897,
+      "learning_rate": 1.798004056959541e-05,
+      "loss": 1.0645,
+      "step": 4006
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.235625147819519,
+      "learning_rate": 1.7978920938459735e-05,
+      "loss": 0.6301,
+      "step": 4007
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0407421588897705,
+      "learning_rate": 1.7977801031993785e-05,
+      "loss": 1.0161,
+      "step": 4008
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.739117980003357,
+      "learning_rate": 1.7976680850236204e-05,
+      "loss": 1.07,
+      "step": 4009
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9061022996902466,
+      "learning_rate": 1.7975560393225656e-05,
+      "loss": 1.0762,
+      "step": 4010
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8710026741027832,
+      "learning_rate": 1.7974439661000794e-05,
+      "loss": 1.1173,
+      "step": 4011
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8702064752578735,
+      "learning_rate": 1.7973318653600294e-05,
+      "loss": 1.0246,
+      "step": 4012
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.949748158454895,
+      "learning_rate": 1.797219737106284e-05,
+      "loss": 1.0695,
+      "step": 4013
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7159366607666016,
+      "learning_rate": 1.7971075813427125e-05,
+      "loss": 1.0074,
+      "step": 4014
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9137449264526367,
+      "learning_rate": 1.796995398073185e-05,
+      "loss": 1.0661,
+      "step": 4015
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.788398027420044,
+      "learning_rate": 1.7968831873015725e-05,
+      "loss": 0.9237,
+      "step": 4016
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.975713849067688,
+      "learning_rate": 1.7967709490317475e-05,
+      "loss": 1.0727,
+      "step": 4017
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.964179277420044,
+      "learning_rate": 1.7966586832675824e-05,
+      "loss": 1.1154,
+      "step": 4018
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7685472965240479,
+      "learning_rate": 1.7965463900129517e-05,
+      "loss": 1.0324,
+      "step": 4019
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0564913749694824,
+      "learning_rate": 1.7964340692717303e-05,
+      "loss": 1.0233,
+      "step": 4020
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9170676469802856,
+      "learning_rate": 1.796321721047794e-05,
+      "loss": 1.1237,
+      "step": 4021
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1880359649658203,
+      "learning_rate": 1.796209345345019e-05,
+      "loss": 1.0362,
+      "step": 4022
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8392949104309082,
+      "learning_rate": 1.7960969421672837e-05,
+      "loss": 1.0199,
+      "step": 4023
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0556066036224365,
+      "learning_rate": 1.795984511518467e-05,
+      "loss": 0.9562,
+      "step": 4024
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8696030378341675,
+      "learning_rate": 1.7958720534024484e-05,
+      "loss": 0.988,
+      "step": 4025
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8294761180877686,
+      "learning_rate": 1.795759567823108e-05,
+      "loss": 1.0314,
+      "step": 4026
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9035627841949463,
+      "learning_rate": 1.7956470547843283e-05,
+      "loss": 1.1086,
+      "step": 4027
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9037648439407349,
+      "learning_rate": 1.795534514289991e-05,
+      "loss": 1.0568,
+      "step": 4028
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7723984718322754,
+      "learning_rate": 1.79542194634398e-05,
+      "loss": 0.9877,
+      "step": 4029
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9419113397598267,
+      "learning_rate": 1.7953093509501794e-05,
+      "loss": 1.0511,
+      "step": 4030
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8378154039382935,
+      "learning_rate": 1.7951967281124746e-05,
+      "loss": 1.0863,
+      "step": 4031
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.854482889175415,
+      "learning_rate": 1.7950840778347524e-05,
+      "loss": 1.0396,
+      "step": 4032
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.858854055404663,
+      "learning_rate": 1.794971400120899e-05,
+      "loss": 1.0536,
+      "step": 4033
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7688071727752686,
+      "learning_rate": 1.7948586949748036e-05,
+      "loss": 0.9924,
+      "step": 4034
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.932106852531433,
+      "learning_rate": 1.7947459624003553e-05,
+      "loss": 1.056,
+      "step": 4035
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9195826053619385,
+      "learning_rate": 1.7946332024014433e-05,
+      "loss": 1.0277,
+      "step": 4036
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9722931385040283,
+      "learning_rate": 1.7945204149819596e-05,
+      "loss": 1.0682,
+      "step": 4037
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9842535257339478,
+      "learning_rate": 1.7944076001457958e-05,
+      "loss": 1.0162,
+      "step": 4038
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8828541040420532,
+      "learning_rate": 1.7942947578968445e-05,
+      "loss": 1.0254,
+      "step": 4039
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9666810035705566,
+      "learning_rate": 1.7941818882390004e-05,
+      "loss": 1.0358,
+      "step": 4040
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7862322330474854,
+      "learning_rate": 1.7940689911761574e-05,
+      "loss": 1.0118,
+      "step": 4041
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.96208918094635,
+      "learning_rate": 1.7939560667122117e-05,
+      "loss": 1.0565,
+      "step": 4042
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.79078209400177,
+      "learning_rate": 1.7938431148510597e-05,
+      "loss": 1.0774,
+      "step": 4043
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8950724601745605,
+      "learning_rate": 1.7937301355965997e-05,
+      "loss": 1.074,
+      "step": 4044
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.86475670337677,
+      "learning_rate": 1.7936171289527296e-05,
+      "loss": 1.0352,
+      "step": 4045
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9543344974517822,
+      "learning_rate": 1.7935040949233496e-05,
+      "loss": 1.041,
+      "step": 4046
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9720805883407593,
+      "learning_rate": 1.79339103351236e-05,
+      "loss": 1.0929,
+      "step": 4047
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8948408365249634,
+      "learning_rate": 1.7932779447236613e-05,
+      "loss": 1.0214,
+      "step": 4048
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.233901023864746,
+      "learning_rate": 1.7931648285611576e-05,
+      "loss": 0.6518,
+      "step": 4049
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7442704439163208,
+      "learning_rate": 1.7930516850287506e-05,
+      "loss": 0.9626,
+      "step": 4050
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.020441770553589,
+      "learning_rate": 1.7929385141303456e-05,
+      "loss": 1.0738,
+      "step": 4051
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0234367847442627,
+      "learning_rate": 1.7928253158698474e-05,
+      "loss": 1.0075,
+      "step": 4052
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.963760495185852,
+      "learning_rate": 1.792712090251162e-05,
+      "loss": 1.1171,
+      "step": 4053
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6826575994491577,
+      "learning_rate": 1.792598837278197e-05,
+      "loss": 0.8942,
+      "step": 4054
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9428327083587646,
+      "learning_rate": 1.79248555695486e-05,
+      "loss": 1.1102,
+      "step": 4055
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7691515684127808,
+      "learning_rate": 1.7923722492850602e-05,
+      "loss": 0.9965,
+      "step": 4056
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8592867851257324,
+      "learning_rate": 1.7922589142727074e-05,
+      "loss": 1.1101,
+      "step": 4057
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9486405849456787,
+      "learning_rate": 1.7921455519217127e-05,
+      "loss": 1.0463,
+      "step": 4058
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7884804010391235,
+      "learning_rate": 1.7920321622359876e-05,
+      "loss": 1.0523,
+      "step": 4059
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8796781301498413,
+      "learning_rate": 1.7919187452194452e-05,
+      "loss": 0.9999,
+      "step": 4060
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8021658658981323,
+      "learning_rate": 1.791805300875999e-05,
+      "loss": 1.0585,
+      "step": 4061
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9235745668411255,
+      "learning_rate": 1.7916918292095636e-05,
+      "loss": 1.0027,
+      "step": 4062
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.882026195526123,
+      "learning_rate": 1.7915783302240548e-05,
+      "loss": 1.0375,
+      "step": 4063
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6535778045654297,
+      "learning_rate": 1.791464803923389e-05,
+      "loss": 1.0345,
+      "step": 4064
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1898133754730225,
+      "learning_rate": 1.791351250311484e-05,
+      "loss": 0.6315,
+      "step": 4065
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9608352184295654,
+      "learning_rate": 1.791237669392257e-05,
+      "loss": 1.071,
+      "step": 4066
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8725255727767944,
+      "learning_rate": 1.791124061169629e-05,
+      "loss": 1.0111,
+      "step": 4067
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2051154375076294,
+      "learning_rate": 1.7910104256475194e-05,
+      "loss": 0.6134,
+      "step": 4068
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8643743991851807,
+      "learning_rate": 1.7908967628298493e-05,
+      "loss": 1.0173,
+      "step": 4069
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9550732374191284,
+      "learning_rate": 1.7907830727205416e-05,
+      "loss": 1.035,
+      "step": 4070
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9909507036209106,
+      "learning_rate": 1.7906693553235192e-05,
+      "loss": 1.0533,
+      "step": 4071
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9447014331817627,
+      "learning_rate": 1.7905556106427054e-05,
+      "loss": 0.9918,
+      "step": 4072
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7819938659667969,
+      "learning_rate": 1.7904418386820262e-05,
+      "loss": 1.0638,
+      "step": 4073
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.059746265411377,
+      "learning_rate": 1.790328039445407e-05,
+      "loss": 1.0943,
+      "step": 4074
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0561931133270264,
+      "learning_rate": 1.790214212936775e-05,
+      "loss": 0.5938,
+      "step": 4075
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9607278108596802,
+      "learning_rate": 1.7901003591600575e-05,
+      "loss": 1.0749,
+      "step": 4076
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0960516929626465,
+      "learning_rate": 1.7899864781191842e-05,
+      "loss": 1.0295,
+      "step": 4077
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9445511102676392,
+      "learning_rate": 1.789872569818084e-05,
+      "loss": 1.0216,
+      "step": 4078
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8433399200439453,
+      "learning_rate": 1.7897586342606875e-05,
+      "loss": 1.0463,
+      "step": 4079
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2118213176727295,
+      "learning_rate": 1.7896446714509272e-05,
+      "loss": 1.0345,
+      "step": 4080
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7436401844024658,
+      "learning_rate": 1.789530681392735e-05,
+      "loss": 1.0291,
+      "step": 4081
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8584575653076172,
+      "learning_rate": 1.789416664090044e-05,
+      "loss": 1.0291,
+      "step": 4082
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7580806016921997,
+      "learning_rate": 1.7893026195467897e-05,
+      "loss": 1.0553,
+      "step": 4083
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.833535075187683,
+      "learning_rate": 1.7891885477669065e-05,
+      "loss": 0.9951,
+      "step": 4084
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.125997543334961,
+      "learning_rate": 1.789074448754331e-05,
+      "loss": 1.0531,
+      "step": 4085
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6811869144439697,
+      "learning_rate": 1.7889603225130004e-05,
+      "loss": 1.0068,
+      "step": 4086
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.740148663520813,
+      "learning_rate": 1.788846169046853e-05,
+      "loss": 1.0026,
+      "step": 4087
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.566307783126831,
+      "learning_rate": 1.7887319883598278e-05,
+      "loss": 0.9683,
+      "step": 4088
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9611505270004272,
+      "learning_rate": 1.788617780455865e-05,
+      "loss": 0.977,
+      "step": 4089
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9645882844924927,
+      "learning_rate": 1.7885035453389057e-05,
+      "loss": 1.0087,
+      "step": 4090
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1712177991867065,
+      "learning_rate": 1.7883892830128915e-05,
+      "loss": 0.6679,
+      "step": 4091
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8754403591156006,
+      "learning_rate": 1.7882749934817654e-05,
+      "loss": 0.9969,
+      "step": 4092
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.131348133087158,
+      "learning_rate": 1.7881606767494712e-05,
+      "loss": 1.0345,
+      "step": 4093
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.49139142036438,
+      "learning_rate": 1.788046332819954e-05,
+      "loss": 1.0746,
+      "step": 4094
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1379097700119019,
+      "learning_rate": 1.7879319616971584e-05,
+      "loss": 0.5899,
+      "step": 4095
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.898483395576477,
+      "learning_rate": 1.7878175633850326e-05,
+      "loss": 1.0469,
+      "step": 4096
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8450311422348022,
+      "learning_rate": 1.787703137887523e-05,
+      "loss": 0.9974,
+      "step": 4097
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8753880262374878,
+      "learning_rate": 1.7875886852085785e-05,
+      "loss": 1.028,
+      "step": 4098
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1251187324523926,
+      "learning_rate": 1.7874742053521483e-05,
+      "loss": 1.0544,
+      "step": 4099
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8532066345214844,
+      "learning_rate": 1.7873596983221832e-05,
+      "loss": 1.0042,
+      "step": 4100
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6483250856399536,
+      "learning_rate": 1.7872451641226345e-05,
+      "loss": 1.0167,
+      "step": 4101
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.075258493423462,
+      "learning_rate": 1.7871306027574544e-05,
+      "loss": 1.0776,
+      "step": 4102
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.879873514175415,
+      "learning_rate": 1.7870160142305954e-05,
+      "loss": 1.1043,
+      "step": 4103
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.802793264389038,
+      "learning_rate": 1.7869013985460123e-05,
+      "loss": 1.0716,
+      "step": 4104
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.625813364982605,
+      "learning_rate": 1.7867867557076604e-05,
+      "loss": 0.9365,
+      "step": 4105
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8565022945404053,
+      "learning_rate": 1.786672085719495e-05,
+      "loss": 1.0765,
+      "step": 4106
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4801981449127197,
+      "learning_rate": 1.7865573885854737e-05,
+      "loss": 0.9989,
+      "step": 4107
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.040534257888794,
+      "learning_rate": 1.7864426643095537e-05,
+      "loss": 1.0585,
+      "step": 4108
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.120488405227661,
+      "learning_rate": 1.7863279128956946e-05,
+      "loss": 1.0781,
+      "step": 4109
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8138446807861328,
+      "learning_rate": 1.7862131343478556e-05,
+      "loss": 1.039,
+      "step": 4110
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.99754798412323,
+      "learning_rate": 1.7860983286699976e-05,
+      "loss": 1.0417,
+      "step": 4111
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8607620000839233,
+      "learning_rate": 1.785983495866082e-05,
+      "loss": 1.0624,
+      "step": 4112
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9502263069152832,
+      "learning_rate": 1.7858686359400715e-05,
+      "loss": 0.9974,
+      "step": 4113
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8580212593078613,
+      "learning_rate": 1.7857537488959297e-05,
+      "loss": 1.1063,
+      "step": 4114
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0122132301330566,
+      "learning_rate": 1.785638834737621e-05,
+      "loss": 1.0533,
+      "step": 4115
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0008320808410645,
+      "learning_rate": 1.785523893469111e-05,
+      "loss": 1.0244,
+      "step": 4116
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8050484657287598,
+      "learning_rate": 1.785408925094365e-05,
+      "loss": 1.0728,
+      "step": 4117
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7926274538040161,
+      "learning_rate": 1.7852939296173516e-05,
+      "loss": 1.0203,
+      "step": 4118
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9154744148254395,
+      "learning_rate": 1.785178907042038e-05,
+      "loss": 0.9789,
+      "step": 4119
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8084694147109985,
+      "learning_rate": 1.7850638573723932e-05,
+      "loss": 1.0608,
+      "step": 4120
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.920157551765442,
+      "learning_rate": 1.7849487806123885e-05,
+      "loss": 1.0513,
+      "step": 4121
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8571562767028809,
+      "learning_rate": 1.7848336767659934e-05,
+      "loss": 1.0582,
+      "step": 4122
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7227673530578613,
+      "learning_rate": 1.7847185458371808e-05,
+      "loss": 0.9684,
+      "step": 4123
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8197344541549683,
+      "learning_rate": 1.7846033878299232e-05,
+      "loss": 1.0166,
+      "step": 4124
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9192757606506348,
+      "learning_rate": 1.7844882027481943e-05,
+      "loss": 1.0545,
+      "step": 4125
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9091987609863281,
+      "learning_rate": 1.7843729905959687e-05,
+      "loss": 1.0706,
+      "step": 4126
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9407949447631836,
+      "learning_rate": 1.7842577513772227e-05,
+      "loss": 1.1138,
+      "step": 4127
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8984144926071167,
+      "learning_rate": 1.784142485095932e-05,
+      "loss": 0.9695,
+      "step": 4128
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.867648959159851,
+      "learning_rate": 1.784027191756075e-05,
+      "loss": 1.081,
+      "step": 4129
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.133528709411621,
+      "learning_rate": 1.7839118713616296e-05,
+      "loss": 1.1055,
+      "step": 4130
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9219939708709717,
+      "learning_rate": 1.7837965239165748e-05,
+      "loss": 1.0505,
+      "step": 4131
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.851528286933899,
+      "learning_rate": 1.7836811494248917e-05,
+      "loss": 1.0243,
+      "step": 4132
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9875849485397339,
+      "learning_rate": 1.7835657478905613e-05,
+      "loss": 0.9717,
+      "step": 4133
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8717445135116577,
+      "learning_rate": 1.783450319317566e-05,
+      "loss": 1.1689,
+      "step": 4134
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8347394466400146,
+      "learning_rate": 1.7833348637098883e-05,
+      "loss": 0.9974,
+      "step": 4135
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0256121158599854,
+      "learning_rate": 1.7832193810715125e-05,
+      "loss": 1.0943,
+      "step": 4136
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8869954347610474,
+      "learning_rate": 1.783103871406424e-05,
+      "loss": 1.1003,
+      "step": 4137
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6921499967575073,
+      "learning_rate": 1.7829883347186086e-05,
+      "loss": 0.9748,
+      "step": 4138
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7227833271026611,
+      "learning_rate": 1.782872771012053e-05,
+      "loss": 1.104,
+      "step": 4139
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9167836904525757,
+      "learning_rate": 1.7827571802907443e-05,
+      "loss": 0.9952,
+      "step": 4140
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9519836902618408,
+      "learning_rate": 1.782641562558672e-05,
+      "loss": 1.0962,
+      "step": 4141
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9016164541244507,
+      "learning_rate": 1.7825259178198258e-05,
+      "loss": 1.0461,
+      "step": 4142
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8856735229492188,
+      "learning_rate": 1.7824102460781962e-05,
+      "loss": 1.0639,
+      "step": 4143
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9405943155288696,
+      "learning_rate": 1.7822945473377744e-05,
+      "loss": 1.0286,
+      "step": 4144
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7457057237625122,
+      "learning_rate": 1.782178821602553e-05,
+      "loss": 1.0439,
+      "step": 4145
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9829154014587402,
+      "learning_rate": 1.7820630688765253e-05,
+      "loss": 1.0167,
+      "step": 4146
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9646515846252441,
+      "learning_rate": 1.7819472891636863e-05,
+      "loss": 1.0232,
+      "step": 4147
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9647910594940186,
+      "learning_rate": 1.78183148246803e-05,
+      "loss": 1.0514,
+      "step": 4148
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.876084327697754,
+      "learning_rate": 1.7817156487935534e-05,
+      "loss": 1.025,
+      "step": 4149
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8375605344772339,
+      "learning_rate": 1.781599788144253e-05,
+      "loss": 1.037,
+      "step": 4150
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8500174283981323,
+      "learning_rate": 1.781483900524128e-05,
+      "loss": 1.1388,
+      "step": 4151
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7848176956176758,
+      "learning_rate": 1.781367985937176e-05,
+      "loss": 1.0623,
+      "step": 4152
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0311336517333984,
+      "learning_rate": 1.7812520443873976e-05,
+      "loss": 1.0831,
+      "step": 4153
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8714094161987305,
+      "learning_rate": 1.7811360758787938e-05,
+      "loss": 0.9857,
+      "step": 4154
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.746452808380127,
+      "learning_rate": 1.7810200804153657e-05,
+      "loss": 1.0028,
+      "step": 4155
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.739323377609253,
+      "learning_rate": 1.780904058001116e-05,
+      "loss": 0.9855,
+      "step": 4156
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8690398931503296,
+      "learning_rate": 1.7807880086400496e-05,
+      "loss": 1.0177,
+      "step": 4157
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8942416906356812,
+      "learning_rate": 1.780671932336169e-05,
+      "loss": 1.0529,
+      "step": 4158
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6740044355392456,
+      "learning_rate": 1.7805558290934814e-05,
+      "loss": 0.9964,
+      "step": 4159
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9355802536010742,
+      "learning_rate": 1.7804396989159923e-05,
+      "loss": 1.0431,
+      "step": 4160
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9288440942764282,
+      "learning_rate": 1.7803235418077094e-05,
+      "loss": 1.0848,
+      "step": 4161
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8997458219528198,
+      "learning_rate": 1.7802073577726407e-05,
+      "loss": 1.0653,
+      "step": 4162
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9431718587875366,
+      "learning_rate": 1.7800911468147955e-05,
+      "loss": 1.0946,
+      "step": 4163
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.852612853050232,
+      "learning_rate": 1.7799749089381843e-05,
+      "loss": 1.0666,
+      "step": 4164
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7472355365753174,
+      "learning_rate": 1.7798586441468172e-05,
+      "loss": 1.0431,
+      "step": 4165
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7824941873550415,
+      "learning_rate": 1.779742352444707e-05,
+      "loss": 1.0488,
+      "step": 4166
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.048429489135742,
+      "learning_rate": 1.7796260338358663e-05,
+      "loss": 1.1056,
+      "step": 4167
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.182837963104248,
+      "learning_rate": 1.7795096883243088e-05,
+      "loss": 0.6198,
+      "step": 4168
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9492744207382202,
+      "learning_rate": 1.7793933159140495e-05,
+      "loss": 1.0678,
+      "step": 4169
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8401384353637695,
+      "learning_rate": 1.779276916609104e-05,
+      "loss": 1.1216,
+      "step": 4170
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8888349533081055,
+      "learning_rate": 1.7791604904134893e-05,
+      "loss": 1.0545,
+      "step": 4171
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8480182886123657,
+      "learning_rate": 1.7790440373312222e-05,
+      "loss": 1.0774,
+      "step": 4172
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7159712314605713,
+      "learning_rate": 1.7789275573663215e-05,
+      "loss": 0.9273,
+      "step": 4173
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9522992372512817,
+      "learning_rate": 1.7788110505228072e-05,
+      "loss": 1.1068,
+      "step": 4174
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8960472345352173,
+      "learning_rate": 1.7786945168046983e-05,
+      "loss": 1.1205,
+      "step": 4175
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9161450862884521,
+      "learning_rate": 1.7785779562160176e-05,
+      "loss": 1.0631,
+      "step": 4176
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8746070861816406,
+      "learning_rate": 1.778461368760786e-05,
+      "loss": 1.1112,
+      "step": 4177
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6447917222976685,
+      "learning_rate": 1.778344754443027e-05,
+      "loss": 1.05,
+      "step": 4178
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0296075344085693,
+      "learning_rate": 1.778228113266765e-05,
+      "loss": 1.1127,
+      "step": 4179
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9357746839523315,
+      "learning_rate": 1.7781114452360246e-05,
+      "loss": 1.0803,
+      "step": 4180
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.743965744972229,
+      "learning_rate": 1.7779947503548318e-05,
+      "loss": 1.0436,
+      "step": 4181
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9591288566589355,
+      "learning_rate": 1.7778780286272134e-05,
+      "loss": 1.0628,
+      "step": 4182
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.173356771469116,
+      "learning_rate": 1.777761280057197e-05,
+      "loss": 1.0876,
+      "step": 4183
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.734431266784668,
+      "learning_rate": 1.7776445046488117e-05,
+      "loss": 1.024,
+      "step": 4184
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8294124603271484,
+      "learning_rate": 1.7775277024060868e-05,
+      "loss": 0.9752,
+      "step": 4185
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8191791772842407,
+      "learning_rate": 1.777410873333053e-05,
+      "loss": 1.0552,
+      "step": 4186
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8683364391326904,
+      "learning_rate": 1.7772940174337412e-05,
+      "loss": 1.0618,
+      "step": 4187
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.780064344406128,
+      "learning_rate": 1.7771771347121842e-05,
+      "loss": 1.0393,
+      "step": 4188
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.722326636314392,
+      "learning_rate": 1.7770602251724153e-05,
+      "loss": 1.003,
+      "step": 4189
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7124961614608765,
+      "learning_rate": 1.7769432888184685e-05,
+      "loss": 1.0586,
+      "step": 4190
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8572055101394653,
+      "learning_rate": 1.7768263256543795e-05,
+      "loss": 1.0152,
+      "step": 4191
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7632818222045898,
+      "learning_rate": 1.7767093356841837e-05,
+      "loss": 1.0479,
+      "step": 4192
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.939591646194458,
+      "learning_rate": 1.7765923189119182e-05,
+      "loss": 1.0995,
+      "step": 4193
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8527276515960693,
+      "learning_rate": 1.776475275341621e-05,
+      "loss": 1.0232,
+      "step": 4194
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7614784240722656,
+      "learning_rate": 1.7763582049773317e-05,
+      "loss": 1.0405,
+      "step": 4195
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7259291410446167,
+      "learning_rate": 1.776241107823089e-05,
+      "loss": 1.0399,
+      "step": 4196
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.925403356552124,
+      "learning_rate": 1.776123983882934e-05,
+      "loss": 1.0503,
+      "step": 4197
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7659294605255127,
+      "learning_rate": 1.7760068331609084e-05,
+      "loss": 1.079,
+      "step": 4198
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9547407627105713,
+      "learning_rate": 1.7758896556610547e-05,
+      "loss": 1.0364,
+      "step": 4199
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1354563236236572,
+      "learning_rate": 1.7757724513874164e-05,
+      "loss": 0.6551,
+      "step": 4200
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8680031299591064,
+      "learning_rate": 1.7756552203440377e-05,
+      "loss": 1.0394,
+      "step": 4201
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.780556082725525,
+      "learning_rate": 1.775537962534964e-05,
+      "loss": 0.9578,
+      "step": 4202
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.08740496635437,
+      "learning_rate": 1.7754206779642416e-05,
+      "loss": 1.015,
+      "step": 4203
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9441726207733154,
+      "learning_rate": 1.7753033666359178e-05,
+      "loss": 1.0404,
+      "step": 4204
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9989814758300781,
+      "learning_rate": 1.7751860285540406e-05,
+      "loss": 0.9956,
+      "step": 4205
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9172426462173462,
+      "learning_rate": 1.7750686637226587e-05,
+      "loss": 1.0918,
+      "step": 4206
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8495709896087646,
+      "learning_rate": 1.7749512721458225e-05,
+      "loss": 0.9887,
+      "step": 4207
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7149168252944946,
+      "learning_rate": 1.7748338538275826e-05,
+      "loss": 1.0178,
+      "step": 4208
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0932137966156006,
+      "learning_rate": 1.7747164087719908e-05,
+      "loss": 1.0206,
+      "step": 4209
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.12624192237854,
+      "learning_rate": 1.7745989369831e-05,
+      "loss": 1.077,
+      "step": 4210
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.929797649383545,
+      "learning_rate": 1.7744814384649633e-05,
+      "loss": 1.0638,
+      "step": 4211
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8133859634399414,
+      "learning_rate": 1.7743639132216355e-05,
+      "loss": 0.9956,
+      "step": 4212
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9237077236175537,
+      "learning_rate": 1.7742463612571724e-05,
+      "loss": 1.0771,
+      "step": 4213
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0847067832946777,
+      "learning_rate": 1.7741287825756303e-05,
+      "loss": 1.1072,
+      "step": 4214
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7332103252410889,
+      "learning_rate": 1.774011177181066e-05,
+      "loss": 0.9776,
+      "step": 4215
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8185864686965942,
+      "learning_rate": 1.773893545077538e-05,
+      "loss": 1.0761,
+      "step": 4216
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8146651983261108,
+      "learning_rate": 1.773775886269106e-05,
+      "loss": 0.9719,
+      "step": 4217
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1219741106033325,
+      "learning_rate": 1.7736582007598295e-05,
+      "loss": 0.6138,
+      "step": 4218
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7386362552642822,
+      "learning_rate": 1.7735404885537693e-05,
+      "loss": 1.055,
+      "step": 4219
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7388397455215454,
+      "learning_rate": 1.773422749654988e-05,
+      "loss": 1.0367,
+      "step": 4220
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9700113534927368,
+      "learning_rate": 1.773304984067548e-05,
+      "loss": 1.0607,
+      "step": 4221
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8276171684265137,
+      "learning_rate": 1.773187191795513e-05,
+      "loss": 1.0387,
+      "step": 4222
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.882305383682251,
+      "learning_rate": 1.773069372842948e-05,
+      "loss": 1.0452,
+      "step": 4223
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8763865232467651,
+      "learning_rate": 1.7729515272139185e-05,
+      "loss": 1.0268,
+      "step": 4224
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6875226497650146,
+      "learning_rate": 1.7728336549124907e-05,
+      "loss": 0.9949,
+      "step": 4225
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9628195762634277,
+      "learning_rate": 1.7727157559427322e-05,
+      "loss": 1.0455,
+      "step": 4226
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7888374328613281,
+      "learning_rate": 1.7725978303087117e-05,
+      "loss": 0.9878,
+      "step": 4227
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9839298725128174,
+      "learning_rate": 1.7724798780144983e-05,
+      "loss": 0.9171,
+      "step": 4228
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.847846269607544,
+      "learning_rate": 1.772361899064162e-05,
+      "loss": 1.0838,
+      "step": 4229
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6741400957107544,
+      "learning_rate": 1.7722438934617742e-05,
+      "loss": 0.9886,
+      "step": 4230
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8983056545257568,
+      "learning_rate": 1.7721258612114066e-05,
+      "loss": 1.0238,
+      "step": 4231
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8268014192581177,
+      "learning_rate": 1.7720078023171325e-05,
+      "loss": 1.0343,
+      "step": 4232
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8222181797027588,
+      "learning_rate": 1.7718897167830257e-05,
+      "loss": 1.0102,
+      "step": 4233
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.109616994857788,
+      "learning_rate": 1.771771604613161e-05,
+      "loss": 1.0471,
+      "step": 4234
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9435971975326538,
+      "learning_rate": 1.7716534658116135e-05,
+      "loss": 1.0223,
+      "step": 4235
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8442895412445068,
+      "learning_rate": 1.7715353003824613e-05,
+      "loss": 1.0654,
+      "step": 4236
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.049138069152832,
+      "learning_rate": 1.7714171083297804e-05,
+      "loss": 0.5889,
+      "step": 4237
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.034022331237793,
+      "learning_rate": 1.7712988896576503e-05,
+      "loss": 0.5975,
+      "step": 4238
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1796789169311523,
+      "learning_rate": 1.77118064437015e-05,
+      "loss": 1.071,
+      "step": 4239
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.094520092010498,
+      "learning_rate": 1.77106237247136e-05,
+      "loss": 1.0393,
+      "step": 4240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.105039358139038,
+      "learning_rate": 1.770944073965361e-05,
+      "loss": 1.0383,
+      "step": 4241
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.888999342918396,
+      "learning_rate": 1.770825748856236e-05,
+      "loss": 1.1102,
+      "step": 4242
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8211272954940796,
+      "learning_rate": 1.7707073971480676e-05,
+      "loss": 1.1087,
+      "step": 4243
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8571031093597412,
+      "learning_rate": 1.7705890188449396e-05,
+      "loss": 0.996,
+      "step": 4244
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8587195873260498,
+      "learning_rate": 1.7704706139509372e-05,
+      "loss": 1.1054,
+      "step": 4245
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9014601707458496,
+      "learning_rate": 1.770352182470146e-05,
+      "loss": 1.0158,
+      "step": 4246
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7314354181289673,
+      "learning_rate": 1.770233724406653e-05,
+      "loss": 1.0906,
+      "step": 4247
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8266500234603882,
+      "learning_rate": 1.770115239764546e-05,
+      "loss": 1.0105,
+      "step": 4248
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9728732109069824,
+      "learning_rate": 1.7699967285479126e-05,
+      "loss": 1.0199,
+      "step": 4249
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8988287448883057,
+      "learning_rate": 1.7698781907608436e-05,
+      "loss": 0.9853,
+      "step": 4250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.057539939880371,
+      "learning_rate": 1.7697596264074285e-05,
+      "loss": 1.0625,
+      "step": 4251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7484532594680786,
+      "learning_rate": 1.769641035491759e-05,
+      "loss": 1.0259,
+      "step": 4252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7445478439331055,
+      "learning_rate": 1.7695224180179275e-05,
+      "loss": 1.0879,
+      "step": 4253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7427477836608887,
+      "learning_rate": 1.7694037739900266e-05,
+      "loss": 1.0591,
+      "step": 4254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.768796443939209,
+      "learning_rate": 1.7692851034121507e-05,
+      "loss": 1.0464,
+      "step": 4255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9365317821502686,
+      "learning_rate": 1.769166406288395e-05,
+      "loss": 1.0303,
+      "step": 4256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9032156467437744,
+      "learning_rate": 1.7690476826228555e-05,
+      "loss": 1.0552,
+      "step": 4257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9427025318145752,
+      "learning_rate": 1.768928932419628e-05,
+      "loss": 1.0931,
+      "step": 4258
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8406506776809692,
+      "learning_rate": 1.7688101556828113e-05,
+      "loss": 1.0801,
+      "step": 4259
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.049480676651001,
+      "learning_rate": 1.7686913524165035e-05,
+      "loss": 1.1148,
+      "step": 4260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7732404470443726,
+      "learning_rate": 1.7685725226248047e-05,
+      "loss": 1.0077,
+      "step": 4261
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9690828323364258,
+      "learning_rate": 1.7684536663118152e-05,
+      "loss": 1.0852,
+      "step": 4262
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9237252473831177,
+      "learning_rate": 1.768334783481636e-05,
+      "loss": 0.989,
+      "step": 4263
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8147753477096558,
+      "learning_rate": 1.7682158741383697e-05,
+      "loss": 0.9964,
+      "step": 4264
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8905417919158936,
+      "learning_rate": 1.7680969382861194e-05,
+      "loss": 1.0317,
+      "step": 4265
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7306811809539795,
+      "learning_rate": 1.7679779759289894e-05,
+      "loss": 1.1156,
+      "step": 4266
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8386883735656738,
+      "learning_rate": 1.767858987071085e-05,
+      "loss": 1.087,
+      "step": 4267
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.775655746459961,
+      "learning_rate": 1.7677399717165116e-05,
+      "loss": 1.0489,
+      "step": 4268
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9242711067199707,
+      "learning_rate": 1.7676209298693765e-05,
+      "loss": 1.0338,
+      "step": 4269
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.301798939704895,
+      "learning_rate": 1.7675018615337874e-05,
+      "loss": 0.6448,
+      "step": 4270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8412302732467651,
+      "learning_rate": 1.767382766713853e-05,
+      "loss": 1.0277,
+      "step": 4271
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9537311792373657,
+      "learning_rate": 1.7672636454136826e-05,
+      "loss": 1.0361,
+      "step": 4272
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0520377159118652,
+      "learning_rate": 1.7671444976373874e-05,
+      "loss": 0.9908,
+      "step": 4273
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8259596824645996,
+      "learning_rate": 1.767025323389078e-05,
+      "loss": 1.0774,
+      "step": 4274
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.840742588043213,
+      "learning_rate": 1.766906122672868e-05,
+      "loss": 0.993,
+      "step": 4275
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.103976249694824,
+      "learning_rate": 1.7667868954928695e-05,
+      "loss": 1.0403,
+      "step": 4276
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9355628490447998,
+      "learning_rate": 1.7666676418531975e-05,
+      "loss": 1.0885,
+      "step": 4277
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.779179334640503,
+      "learning_rate": 1.7665483617579666e-05,
+      "loss": 1.1026,
+      "step": 4278
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8350656032562256,
+      "learning_rate": 1.766429055211293e-05,
+      "loss": 1.0463,
+      "step": 4279
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8098876476287842,
+      "learning_rate": 1.7663097222172936e-05,
+      "loss": 1.1038,
+      "step": 4280
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8046627044677734,
+      "learning_rate": 1.7661903627800864e-05,
+      "loss": 1.0161,
+      "step": 4281
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1524578332901,
+      "learning_rate": 1.76607097690379e-05,
+      "loss": 0.6045,
+      "step": 4282
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9195706844329834,
+      "learning_rate": 1.7659515645925242e-05,
+      "loss": 1.0875,
+      "step": 4283
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8675733804702759,
+      "learning_rate": 1.7658321258504092e-05,
+      "loss": 1.0441,
+      "step": 4284
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.755473017692566,
+      "learning_rate": 1.7657126606815672e-05,
+      "loss": 1.064,
+      "step": 4285
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0828771591186523,
+      "learning_rate": 1.7655931690901197e-05,
+      "loss": 1.0681,
+      "step": 4286
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.719041109085083,
+      "learning_rate": 1.765473651080191e-05,
+      "loss": 1.0533,
+      "step": 4287
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7719743251800537,
+      "learning_rate": 1.7653541066559044e-05,
+      "loss": 0.9605,
+      "step": 4288
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6929140090942383,
+      "learning_rate": 1.765234535821386e-05,
+      "loss": 1.0061,
+      "step": 4289
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.009722948074341,
+      "learning_rate": 1.7651149385807612e-05,
+      "loss": 1.0627,
+      "step": 4290
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9348431825637817,
+      "learning_rate": 1.7649953149381572e-05,
+      "loss": 1.184,
+      "step": 4291
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9220441579818726,
+      "learning_rate": 1.764875664897702e-05,
+      "loss": 1.1498,
+      "step": 4292
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8753741979599,
+      "learning_rate": 1.7647559884635238e-05,
+      "loss": 1.0068,
+      "step": 4293
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0409554243087769,
+      "learning_rate": 1.7646362856397527e-05,
+      "loss": 0.5888,
+      "step": 4294
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7593097686767578,
+      "learning_rate": 1.7645165564305197e-05,
+      "loss": 0.9922,
+      "step": 4295
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7983962297439575,
+      "learning_rate": 1.7643968008399553e-05,
+      "loss": 0.9999,
+      "step": 4296
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7414953708648682,
+      "learning_rate": 1.764277018872193e-05,
+      "loss": 0.9985,
+      "step": 4297
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8488049507141113,
+      "learning_rate": 1.7641572105313657e-05,
+      "loss": 1.0848,
+      "step": 4298
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.142622709274292,
+      "learning_rate": 1.7640373758216075e-05,
+      "loss": 1.0983,
+      "step": 4299
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.771493673324585,
+      "learning_rate": 1.7639175147470537e-05,
+      "loss": 1.0218,
+      "step": 4300
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9123984575271606,
+      "learning_rate": 1.7637976273118405e-05,
+      "loss": 1.0646,
+      "step": 4301
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2467190027236938,
+      "learning_rate": 1.763677713520105e-05,
+      "loss": 0.6547,
+      "step": 4302
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8806278705596924,
+      "learning_rate": 1.7635577733759843e-05,
+      "loss": 1.0897,
+      "step": 4303
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9061013460159302,
+      "learning_rate": 1.763437806883618e-05,
+      "loss": 1.0713,
+      "step": 4304
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0251996517181396,
+      "learning_rate": 1.763317814047146e-05,
+      "loss": 1.0655,
+      "step": 4305
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1427624225616455,
+      "learning_rate": 1.763197794870708e-05,
+      "loss": 1.0701,
+      "step": 4306
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7627131938934326,
+      "learning_rate": 1.763077749358446e-05,
+      "loss": 0.9516,
+      "step": 4307
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0266106128692627,
+      "learning_rate": 1.7629576775145026e-05,
+      "loss": 1.0311,
+      "step": 4308
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.843187689781189,
+      "learning_rate": 1.762837579343021e-05,
+      "loss": 0.9992,
+      "step": 4309
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.825524091720581,
+      "learning_rate": 1.7627174548481455e-05,
+      "loss": 1.0868,
+      "step": 4310
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2021127939224243,
+      "learning_rate": 1.7625973040340208e-05,
+      "loss": 0.6306,
+      "step": 4311
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9457613229751587,
+      "learning_rate": 1.7624771269047935e-05,
+      "loss": 1.1067,
+      "step": 4312
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.35219144821167,
+      "learning_rate": 1.7623569234646108e-05,
+      "loss": 1.1091,
+      "step": 4313
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8737244606018066,
+      "learning_rate": 1.76223669371762e-05,
+      "loss": 1.1332,
+      "step": 4314
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7767356634140015,
+      "learning_rate": 1.7621164376679697e-05,
+      "loss": 1.0771,
+      "step": 4315
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.869818091392517,
+      "learning_rate": 1.761996155319811e-05,
+      "loss": 0.9751,
+      "step": 4316
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7541027069091797,
+      "learning_rate": 1.7618758466772928e-05,
+      "loss": 1.0052,
+      "step": 4317
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.764745831489563,
+      "learning_rate": 1.7617555117445674e-05,
+      "loss": 0.979,
+      "step": 4318
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7584518194198608,
+      "learning_rate": 1.7616351505257873e-05,
+      "loss": 0.9849,
+      "step": 4319
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7625293731689453,
+      "learning_rate": 1.7615147630251055e-05,
+      "loss": 1.0115,
+      "step": 4320
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8104900121688843,
+      "learning_rate": 1.7613943492466767e-05,
+      "loss": 1.0396,
+      "step": 4321
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8959300518035889,
+      "learning_rate": 1.7612739091946556e-05,
+      "loss": 1.0087,
+      "step": 4322
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7578295469284058,
+      "learning_rate": 1.7611534428731986e-05,
+      "loss": 1.0243,
+      "step": 4323
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7342609167099,
+      "learning_rate": 1.7610329502864625e-05,
+      "loss": 1.0253,
+      "step": 4324
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8796322345733643,
+      "learning_rate": 1.7609124314386053e-05,
+      "loss": 1.0681,
+      "step": 4325
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.453695297241211,
+      "learning_rate": 1.760791886333785e-05,
+      "loss": 1.046,
+      "step": 4326
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.800392508506775,
+      "learning_rate": 1.7606713149761626e-05,
+      "loss": 0.965,
+      "step": 4327
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8954888582229614,
+      "learning_rate": 1.760550717369897e-05,
+      "loss": 0.9225,
+      "step": 4328
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8303425312042236,
+      "learning_rate": 1.7604300935191517e-05,
+      "loss": 1.0918,
+      "step": 4329
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7157204151153564,
+      "learning_rate": 1.7603094434280878e-05,
+      "loss": 1.0535,
+      "step": 4330
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8541831970214844,
+      "learning_rate": 1.7601887671008685e-05,
+      "loss": 1.0628,
+      "step": 4331
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8427479267120361,
+      "learning_rate": 1.7600680645416583e-05,
+      "loss": 0.9964,
+      "step": 4332
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1192212104797363,
+      "learning_rate": 1.759947335754623e-05,
+      "loss": 0.9555,
+      "step": 4333
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7982256412506104,
+      "learning_rate": 1.759826580743927e-05,
+      "loss": 1.0827,
+      "step": 4334
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7763760089874268,
+      "learning_rate": 1.759705799513739e-05,
+      "loss": 0.9659,
+      "step": 4335
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8327903747558594,
+      "learning_rate": 1.7595849920682258e-05,
+      "loss": 1.062,
+      "step": 4336
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0384023189544678,
+      "learning_rate": 1.7594641584115565e-05,
+      "loss": 1.0663,
+      "step": 4337
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.988202452659607,
+      "learning_rate": 1.7593432985479003e-05,
+      "loss": 1.0789,
+      "step": 4338
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9371157884597778,
+      "learning_rate": 1.759222412481428e-05,
+      "loss": 1.0223,
+      "step": 4339
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.844335675239563,
+      "learning_rate": 1.759101500216311e-05,
+      "loss": 1.0108,
+      "step": 4340
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1359611749649048,
+      "learning_rate": 1.7589805617567218e-05,
+      "loss": 0.6675,
+      "step": 4341
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9542769193649292,
+      "learning_rate": 1.7588595971068335e-05,
+      "loss": 1.0133,
+      "step": 4342
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9332380294799805,
+      "learning_rate": 1.7587386062708206e-05,
+      "loss": 0.9716,
+      "step": 4343
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8213160037994385,
+      "learning_rate": 1.7586175892528574e-05,
+      "loss": 1.0374,
+      "step": 4344
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9793041944503784,
+      "learning_rate": 1.7584965460571207e-05,
+      "loss": 0.9901,
+      "step": 4345
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0708868503570557,
+      "learning_rate": 1.7583754766877864e-05,
+      "loss": 0.9619,
+      "step": 4346
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.815346598625183,
+      "learning_rate": 1.7582543811490334e-05,
+      "loss": 1.047,
+      "step": 4347
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1594117879867554,
+      "learning_rate": 1.7581332594450394e-05,
+      "loss": 0.6383,
+      "step": 4348
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9160046577453613,
+      "learning_rate": 1.7580121115799845e-05,
+      "loss": 1.0448,
+      "step": 4349
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9320781230926514,
+      "learning_rate": 1.7578909375580494e-05,
+      "loss": 1.1024,
+      "step": 4350
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.172434091567993,
+      "learning_rate": 1.7577697373834147e-05,
+      "loss": 0.981,
+      "step": 4351
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8760137557983398,
+      "learning_rate": 1.7576485110602634e-05,
+      "loss": 1.0946,
+      "step": 4352
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7859572172164917,
+      "learning_rate": 1.757527258592778e-05,
+      "loss": 1.0843,
+      "step": 4353
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8375836610794067,
+      "learning_rate": 1.7574059799851433e-05,
+      "loss": 1.0549,
+      "step": 4354
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.125635862350464,
+      "learning_rate": 1.757284675241544e-05,
+      "loss": 1.0184,
+      "step": 4355
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.682332158088684,
+      "learning_rate": 1.7571633443661658e-05,
+      "loss": 1.0099,
+      "step": 4356
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.990230679512024,
+      "learning_rate": 1.757041987363196e-05,
+      "loss": 1.0209,
+      "step": 4357
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9476579427719116,
+      "learning_rate": 1.7569206042368213e-05,
+      "loss": 1.0016,
+      "step": 4358
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9963138103485107,
+      "learning_rate": 1.7567991949912313e-05,
+      "loss": 1.0659,
+      "step": 4359
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6647831201553345,
+      "learning_rate": 1.756677759630615e-05,
+      "loss": 1.0327,
+      "step": 4360
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8915528059005737,
+      "learning_rate": 1.7565562981591628e-05,
+      "loss": 0.9759,
+      "step": 4361
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.921199917793274,
+      "learning_rate": 1.7564348105810663e-05,
+      "loss": 1.1693,
+      "step": 4362
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7560514211654663,
+      "learning_rate": 1.7563132969005175e-05,
+      "loss": 1.1282,
+      "step": 4363
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7801347970962524,
+      "learning_rate": 1.7561917571217094e-05,
+      "loss": 1.0528,
+      "step": 4364
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5496139526367188,
+      "learning_rate": 1.7560701912488362e-05,
+      "loss": 0.9204,
+      "step": 4365
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9893279075622559,
+      "learning_rate": 1.7559485992860924e-05,
+      "loss": 0.9709,
+      "step": 4366
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.764581322669983,
+      "learning_rate": 1.7558269812376746e-05,
+      "loss": 1.0323,
+      "step": 4367
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.669557809829712,
+      "learning_rate": 1.7557053371077785e-05,
+      "loss": 0.9808,
+      "step": 4368
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8658924102783203,
+      "learning_rate": 1.7555836669006026e-05,
+      "loss": 1.0525,
+      "step": 4369
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9508601427078247,
+      "learning_rate": 1.755461970620345e-05,
+      "loss": 1.0479,
+      "step": 4370
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8994783163070679,
+      "learning_rate": 1.7553402482712048e-05,
+      "loss": 1.0795,
+      "step": 4371
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6652858257293701,
+      "learning_rate": 1.7552184998573827e-05,
+      "loss": 0.9697,
+      "step": 4372
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.014883518218994,
+      "learning_rate": 1.7550967253830795e-05,
+      "loss": 1.049,
+      "step": 4373
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7456210851669312,
+      "learning_rate": 1.7549749248524982e-05,
+      "loss": 1.036,
+      "step": 4374
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8527990579605103,
+      "learning_rate": 1.7548530982698405e-05,
+      "loss": 1.1032,
+      "step": 4375
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1514252424240112,
+      "learning_rate": 1.7547312456393116e-05,
+      "loss": 0.5883,
+      "step": 4376
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9269376993179321,
+      "learning_rate": 1.7546093669651155e-05,
+      "loss": 1.0394,
+      "step": 4377
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9862536191940308,
+      "learning_rate": 1.7544874622514577e-05,
+      "loss": 1.0888,
+      "step": 4378
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.036888599395752,
+      "learning_rate": 1.7543655315025458e-05,
+      "loss": 1.0216,
+      "step": 4379
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.854652762413025,
+      "learning_rate": 1.7542435747225858e-05,
+      "loss": 1.0913,
+      "step": 4380
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.871500849723816,
+      "learning_rate": 1.7541215919157876e-05,
+      "loss": 1.0197,
+      "step": 4381
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7790777683258057,
+      "learning_rate": 1.7539995830863598e-05,
+      "loss": 1.0294,
+      "step": 4382
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8373584747314453,
+      "learning_rate": 1.753877548238512e-05,
+      "loss": 1.0527,
+      "step": 4383
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.931557536125183,
+      "learning_rate": 1.7537554873764566e-05,
+      "loss": 0.979,
+      "step": 4384
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8186033964157104,
+      "learning_rate": 1.7536334005044044e-05,
+      "loss": 1.0621,
+      "step": 4385
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7134222984313965,
+      "learning_rate": 1.753511287626569e-05,
+      "loss": 0.9737,
+      "step": 4386
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.976021409034729,
+      "learning_rate": 1.7533891487471636e-05,
+      "loss": 1.0808,
+      "step": 4387
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.96632719039917,
+      "learning_rate": 1.7532669838704036e-05,
+      "loss": 1.0336,
+      "step": 4388
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7563728094100952,
+      "learning_rate": 1.753144793000504e-05,
+      "loss": 1.023,
+      "step": 4389
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.106621026992798,
+      "learning_rate": 1.7530225761416814e-05,
+      "loss": 1.0728,
+      "step": 4390
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1241793632507324,
+      "learning_rate": 1.752900333298153e-05,
+      "loss": 1.1316,
+      "step": 4391
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9995064735412598,
+      "learning_rate": 1.752778064474137e-05,
+      "loss": 0.9829,
+      "step": 4392
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7408829927444458,
+      "learning_rate": 1.7526557696738536e-05,
+      "loss": 1.0173,
+      "step": 4393
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.857115387916565,
+      "learning_rate": 1.7525334489015217e-05,
+      "loss": 0.9446,
+      "step": 4394
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8432306051254272,
+      "learning_rate": 1.7524111021613625e-05,
+      "loss": 1.0162,
+      "step": 4395
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7006149291992188,
+      "learning_rate": 1.7522887294575978e-05,
+      "loss": 1.0203,
+      "step": 4396
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9943126440048218,
+      "learning_rate": 1.7521663307944504e-05,
+      "loss": 1.0712,
+      "step": 4397
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8605749607086182,
+      "learning_rate": 1.7520439061761444e-05,
+      "loss": 1.051,
+      "step": 4398
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9962127208709717,
+      "learning_rate": 1.7519214556069037e-05,
+      "loss": 1.076,
+      "step": 4399
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9175949096679688,
+      "learning_rate": 1.751798979090954e-05,
+      "loss": 1.0559,
+      "step": 4400
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8520915508270264,
+      "learning_rate": 1.751676476632522e-05,
+      "loss": 0.9975,
+      "step": 4401
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9299004077911377,
+      "learning_rate": 1.751553948235834e-05,
+      "loss": 0.9731,
+      "step": 4402
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8639037609100342,
+      "learning_rate": 1.751431393905119e-05,
+      "loss": 1.1013,
+      "step": 4403
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8737397193908691,
+      "learning_rate": 1.7513088136446055e-05,
+      "loss": 1.0056,
+      "step": 4404
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7145204544067383,
+      "learning_rate": 1.7511862074585232e-05,
+      "loss": 1.0403,
+      "step": 4405
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.861163854598999,
+      "learning_rate": 1.7510635753511037e-05,
+      "loss": 0.9912,
+      "step": 4406
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.879435420036316,
+      "learning_rate": 1.750940917326578e-05,
+      "loss": 1.0508,
+      "step": 4407
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9297387599945068,
+      "learning_rate": 1.750818233389179e-05,
+      "loss": 1.0349,
+      "step": 4408
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.729915738105774,
+      "learning_rate": 1.75069552354314e-05,
+      "loss": 0.9792,
+      "step": 4409
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.050724506378174,
+      "learning_rate": 1.7505727877926957e-05,
+      "loss": 1.003,
+      "step": 4410
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9200385808944702,
+      "learning_rate": 1.750450026142081e-05,
+      "loss": 1.0401,
+      "step": 4411
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7514429092407227,
+      "learning_rate": 1.750327238595532e-05,
+      "loss": 1.0196,
+      "step": 4412
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0090529918670654,
+      "learning_rate": 1.750204425157286e-05,
+      "loss": 1.0068,
+      "step": 4413
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.790141224861145,
+      "learning_rate": 1.750081585831581e-05,
+      "loss": 0.9904,
+      "step": 4414
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7692090272903442,
+      "learning_rate": 1.7499587206226558e-05,
+      "loss": 1.0581,
+      "step": 4415
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8720014095306396,
+      "learning_rate": 1.74983582953475e-05,
+      "loss": 0.9857,
+      "step": 4416
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.079662322998047,
+      "learning_rate": 1.7497129125721043e-05,
+      "loss": 1.0777,
+      "step": 4417
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6887396574020386,
+      "learning_rate": 1.7495899697389602e-05,
+      "loss": 1.1083,
+      "step": 4418
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8303890228271484,
+      "learning_rate": 1.7494670010395603e-05,
+      "loss": 1.0502,
+      "step": 4419
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8686891794204712,
+      "learning_rate": 1.7493440064781477e-05,
+      "loss": 1.019,
+      "step": 4420
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9353564977645874,
+      "learning_rate": 1.7492209860589665e-05,
+      "loss": 1.0761,
+      "step": 4421
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7239558696746826,
+      "learning_rate": 1.749097939786262e-05,
+      "loss": 0.9892,
+      "step": 4422
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9947906732559204,
+      "learning_rate": 1.74897486766428e-05,
+      "loss": 1.0948,
+      "step": 4423
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.148057460784912,
+      "learning_rate": 1.7488517696972675e-05,
+      "loss": 1.072,
+      "step": 4424
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.994001865386963,
+      "learning_rate": 1.7487286458894725e-05,
+      "loss": 0.9893,
+      "step": 4425
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.927014708518982,
+      "learning_rate": 1.7486054962451435e-05,
+      "loss": 1.0698,
+      "step": 4426
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7765557765960693,
+      "learning_rate": 1.7484823207685298e-05,
+      "loss": 1.0266,
+      "step": 4427
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9424831867218018,
+      "learning_rate": 1.748359119463882e-05,
+      "loss": 1.0077,
+      "step": 4428
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.102088689804077,
+      "learning_rate": 1.748235892335451e-05,
+      "loss": 1.0057,
+      "step": 4429
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9213680028915405,
+      "learning_rate": 1.74811263938749e-05,
+      "loss": 1.0632,
+      "step": 4430
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8730134963989258,
+      "learning_rate": 1.7479893606242517e-05,
+      "loss": 1.0234,
+      "step": 4431
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8309963941574097,
+      "learning_rate": 1.74786605604999e-05,
+      "loss": 0.9764,
+      "step": 4432
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8340407609939575,
+      "learning_rate": 1.747742725668959e-05,
+      "loss": 1.0787,
+      "step": 4433
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0280392169952393,
+      "learning_rate": 1.747619369485416e-05,
+      "loss": 0.5216,
+      "step": 4434
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7461820840835571,
+      "learning_rate": 1.7474959875036167e-05,
+      "loss": 1.0842,
+      "step": 4435
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0159056186676025,
+      "learning_rate": 1.747372579727819e-05,
+      "loss": 0.9791,
+      "step": 4436
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0121371746063232,
+      "learning_rate": 1.7472491461622813e-05,
+      "loss": 1.0148,
+      "step": 4437
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8012595176696777,
+      "learning_rate": 1.7471256868112632e-05,
+      "loss": 1.0073,
+      "step": 4438
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9510135650634766,
+      "learning_rate": 1.747002201679024e-05,
+      "loss": 0.9764,
+      "step": 4439
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7875256538391113,
+      "learning_rate": 1.746878690769826e-05,
+      "loss": 1.0573,
+      "step": 4440
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7916224002838135,
+      "learning_rate": 1.7467551540879303e-05,
+      "loss": 0.9429,
+      "step": 4441
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.997593641281128,
+      "learning_rate": 1.7466315916376008e-05,
+      "loss": 1.1146,
+      "step": 4442
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.287393569946289,
+      "learning_rate": 1.7465080034231002e-05,
+      "loss": 1.0024,
+      "step": 4443
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.137901782989502,
+      "learning_rate": 1.746384389448694e-05,
+      "loss": 1.0588,
+      "step": 4444
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8191730976104736,
+      "learning_rate": 1.7462607497186473e-05,
+      "loss": 1.0947,
+      "step": 4445
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.983291745185852,
+      "learning_rate": 1.7461370842372263e-05,
+      "loss": 1.0369,
+      "step": 4446
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1294503211975098,
+      "learning_rate": 1.7460133930086997e-05,
+      "loss": 1.038,
+      "step": 4447
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8829761743545532,
+      "learning_rate": 1.7458896760373337e-05,
+      "loss": 1.0484,
+      "step": 4448
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7346512079238892,
+      "learning_rate": 1.7457659333273995e-05,
+      "loss": 1.0111,
+      "step": 4449
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8734426498413086,
+      "learning_rate": 1.7456421648831658e-05,
+      "loss": 1.0173,
+      "step": 4450
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8326090574264526,
+      "learning_rate": 1.745518370708904e-05,
+      "loss": 1.0303,
+      "step": 4451
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7571946382522583,
+      "learning_rate": 1.7453945508088853e-05,
+      "loss": 1.0763,
+      "step": 4452
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0093625783920288,
+      "learning_rate": 1.7452707051873835e-05,
+      "loss": 0.5638,
+      "step": 4453
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1001510620117188,
+      "learning_rate": 1.745146833848671e-05,
+      "loss": 1.037,
+      "step": 4454
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7653131484985352,
+      "learning_rate": 1.745022936797023e-05,
+      "loss": 1.1007,
+      "step": 4455
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9968433380126953,
+      "learning_rate": 1.7448990140367143e-05,
+      "loss": 1.0205,
+      "step": 4456
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.86422860622406,
+      "learning_rate": 1.7447750655720215e-05,
+      "loss": 1.0656,
+      "step": 4457
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1134369373321533,
+      "learning_rate": 1.7446510914072217e-05,
+      "loss": 0.6609,
+      "step": 4458
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7147307395935059,
+      "learning_rate": 1.7445270915465927e-05,
+      "loss": 1.0442,
+      "step": 4459
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.755669116973877,
+      "learning_rate": 1.7444030659944138e-05,
+      "loss": 0.9659,
+      "step": 4460
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8571805953979492,
+      "learning_rate": 1.7442790147549644e-05,
+      "loss": 0.9453,
+      "step": 4461
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6257529258728027,
+      "learning_rate": 1.744154937832525e-05,
+      "loss": 1.0123,
+      "step": 4462
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9559122323989868,
+      "learning_rate": 1.744030835231378e-05,
+      "loss": 1.0895,
+      "step": 4463
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.782806634902954,
+      "learning_rate": 1.7439067069558048e-05,
+      "loss": 1.0253,
+      "step": 4464
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1015076637268066,
+      "learning_rate": 1.7437825530100892e-05,
+      "loss": 1.0328,
+      "step": 4465
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8133772611618042,
+      "learning_rate": 1.7436583733985154e-05,
+      "loss": 1.029,
+      "step": 4466
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7965253591537476,
+      "learning_rate": 1.7435341681253683e-05,
+      "loss": 1.0085,
+      "step": 4467
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7258120775222778,
+      "learning_rate": 1.7434099371949345e-05,
+      "loss": 0.9902,
+      "step": 4468
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8406016826629639,
+      "learning_rate": 1.7432856806114998e-05,
+      "loss": 1.0623,
+      "step": 4469
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8869602680206299,
+      "learning_rate": 1.7431613983793528e-05,
+      "loss": 0.9949,
+      "step": 4470
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.639609932899475,
+      "learning_rate": 1.743037090502782e-05,
+      "loss": 0.9081,
+      "step": 4471
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7113995552062988,
+      "learning_rate": 1.7429127569860768e-05,
+      "loss": 1.0229,
+      "step": 4472
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7718567848205566,
+      "learning_rate": 1.7427883978335275e-05,
+      "loss": 1.0297,
+      "step": 4473
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9718457460403442,
+      "learning_rate": 1.7426640130494258e-05,
+      "loss": 0.9871,
+      "step": 4474
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7503424882888794,
+      "learning_rate": 1.742539602638063e-05,
+      "loss": 0.9506,
+      "step": 4475
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.717276692390442,
+      "learning_rate": 1.742415166603733e-05,
+      "loss": 1.081,
+      "step": 4476
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9617878198623657,
+      "learning_rate": 1.7422907049507295e-05,
+      "loss": 0.9606,
+      "step": 4477
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8451484441757202,
+      "learning_rate": 1.742166217683347e-05,
+      "loss": 0.9326,
+      "step": 4478
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.778012990951538,
+      "learning_rate": 1.7420417048058816e-05,
+      "loss": 1.0367,
+      "step": 4479
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.862652063369751,
+      "learning_rate": 1.74191716632263e-05,
+      "loss": 0.9743,
+      "step": 4480
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6514818668365479,
+      "learning_rate": 1.7417926022378894e-05,
+      "loss": 1.0858,
+      "step": 4481
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0770589113235474,
+      "learning_rate": 1.741668012555958e-05,
+      "loss": 0.6152,
+      "step": 4482
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7427926063537598,
+      "learning_rate": 1.7415433972811356e-05,
+      "loss": 0.993,
+      "step": 4483
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0804619789123535,
+      "learning_rate": 1.741418756417722e-05,
+      "loss": 1.0858,
+      "step": 4484
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9790469408035278,
+      "learning_rate": 1.741294089970018e-05,
+      "loss": 0.9608,
+      "step": 4485
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8288018703460693,
+      "learning_rate": 1.741169397942325e-05,
+      "loss": 0.9527,
+      "step": 4486
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.710623860359192,
+      "learning_rate": 1.7410446803389477e-05,
+      "loss": 0.9809,
+      "step": 4487
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9912240505218506,
+      "learning_rate": 1.7409199371641875e-05,
+      "loss": 1.0205,
+      "step": 4488
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7927414178848267,
+      "learning_rate": 1.7407951684223504e-05,
+      "loss": 1.0715,
+      "step": 4489
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6931726932525635,
+      "learning_rate": 1.7406703741177416e-05,
+      "loss": 1.0477,
+      "step": 4490
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.932991862297058,
+      "learning_rate": 1.7405455542546668e-05,
+      "loss": 1.0067,
+      "step": 4491
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8996227979660034,
+      "learning_rate": 1.7404207088374333e-05,
+      "loss": 1.0551,
+      "step": 4492
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7182364463806152,
+      "learning_rate": 1.74029583787035e-05,
+      "loss": 0.9863,
+      "step": 4493
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.776246428489685,
+      "learning_rate": 1.7401709413577248e-05,
+      "loss": 1.0253,
+      "step": 4494
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.094330072402954,
+      "learning_rate": 1.7400460193038684e-05,
+      "loss": 1.107,
+      "step": 4495
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8532600402832031,
+      "learning_rate": 1.739921071713091e-05,
+      "loss": 1.1167,
+      "step": 4496
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9286553859710693,
+      "learning_rate": 1.739796098589704e-05,
+      "loss": 1.0665,
+      "step": 4497
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.848468542098999,
+      "learning_rate": 1.7396710999380205e-05,
+      "loss": 0.9989,
+      "step": 4498
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.123755931854248,
+      "learning_rate": 1.739546075762353e-05,
+      "loss": 1.1226,
+      "step": 4499
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9550516605377197,
+      "learning_rate": 1.739421026067017e-05,
+      "loss": 1.085,
+      "step": 4500
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7448196411132812,
+      "learning_rate": 1.7392959508563266e-05,
+      "loss": 1.068,
+      "step": 4501
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8628637790679932,
+      "learning_rate": 1.739170850134598e-05,
+      "loss": 1.0651,
+      "step": 4502
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7664508819580078,
+      "learning_rate": 1.739045723906148e-05,
+      "loss": 1.1084,
+      "step": 4503
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8520745038986206,
+      "learning_rate": 1.7389205721752952e-05,
+      "loss": 0.9457,
+      "step": 4504
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.015570878982544,
+      "learning_rate": 1.7387953949463566e-05,
+      "loss": 0.9678,
+      "step": 4505
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.813125729560852,
+      "learning_rate": 1.7386701922236534e-05,
+      "loss": 1.0742,
+      "step": 4506
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7884405851364136,
+      "learning_rate": 1.738544964011505e-05,
+      "loss": 1.0867,
+      "step": 4507
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.817070484161377,
+      "learning_rate": 1.738419710314233e-05,
+      "loss": 0.9926,
+      "step": 4508
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8422431945800781,
+      "learning_rate": 1.7382944311361593e-05,
+      "loss": 1.0771,
+      "step": 4509
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.764309048652649,
+      "learning_rate": 1.738169126481607e-05,
+      "loss": 0.9801,
+      "step": 4510
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.784175992012024,
+      "learning_rate": 1.7380437963549005e-05,
+      "loss": 1.0063,
+      "step": 4511
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7415882349014282,
+      "learning_rate": 1.737918440760364e-05,
+      "loss": 1.0233,
+      "step": 4512
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.759074330329895,
+      "learning_rate": 1.7377930597023235e-05,
+      "loss": 1.0348,
+      "step": 4513
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9451662302017212,
+      "learning_rate": 1.7376676531851053e-05,
+      "loss": 0.9744,
+      "step": 4514
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8722350597381592,
+      "learning_rate": 1.7375422212130372e-05,
+      "loss": 1.0651,
+      "step": 4515
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.734866976737976,
+      "learning_rate": 1.737416763790447e-05,
+      "loss": 0.9875,
+      "step": 4516
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1796067953109741,
+      "learning_rate": 1.7372912809216647e-05,
+      "loss": 0.6717,
+      "step": 4517
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8314779996871948,
+      "learning_rate": 1.7371657726110192e-05,
+      "loss": 1.0172,
+      "step": 4518
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7801262140274048,
+      "learning_rate": 1.7370402388628424e-05,
+      "loss": 0.9716,
+      "step": 4519
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9206441640853882,
+      "learning_rate": 1.736914679681466e-05,
+      "loss": 1.0691,
+      "step": 4520
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8735765218734741,
+      "learning_rate": 1.7367890950712222e-05,
+      "loss": 1.0628,
+      "step": 4521
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5241646766662598,
+      "learning_rate": 1.736663485036445e-05,
+      "loss": 0.9981,
+      "step": 4522
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.891777515411377,
+      "learning_rate": 1.7365378495814688e-05,
+      "loss": 1.0586,
+      "step": 4523
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.002377986907959,
+      "learning_rate": 1.7364121887106285e-05,
+      "loss": 1.0557,
+      "step": 4524
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0251810550689697,
+      "learning_rate": 1.736286502428261e-05,
+      "loss": 1.0239,
+      "step": 4525
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.738773226737976,
+      "learning_rate": 1.736160790738703e-05,
+      "loss": 0.9556,
+      "step": 4526
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6899263858795166,
+      "learning_rate": 1.7360350536462924e-05,
+      "loss": 1.0476,
+      "step": 4527
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.873380422592163,
+      "learning_rate": 1.735909291155368e-05,
+      "loss": 0.9954,
+      "step": 4528
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7814782857894897,
+      "learning_rate": 1.7357835032702696e-05,
+      "loss": 1.0363,
+      "step": 4529
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8374149799346924,
+      "learning_rate": 1.735657689995338e-05,
+      "loss": 1.0731,
+      "step": 4530
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6497817039489746,
+      "learning_rate": 1.7355318513349148e-05,
+      "loss": 0.9492,
+      "step": 4531
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8671873807907104,
+      "learning_rate": 1.7354059872933414e-05,
+      "loss": 0.982,
+      "step": 4532
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.743689775466919,
+      "learning_rate": 1.735280097874962e-05,
+      "loss": 1.0225,
+      "step": 4533
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.767421841621399,
+      "learning_rate": 1.7351541830841205e-05,
+      "loss": 0.9663,
+      "step": 4534
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8889718055725098,
+      "learning_rate": 1.735028242925161e-05,
+      "loss": 1.1091,
+      "step": 4535
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7472302913665771,
+      "learning_rate": 1.7349022774024307e-05,
+      "loss": 0.93,
+      "step": 4536
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9255951642990112,
+      "learning_rate": 1.734776286520276e-05,
+      "loss": 1.0153,
+      "step": 4537
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8141483068466187,
+      "learning_rate": 1.7346502702830436e-05,
+      "loss": 0.9766,
+      "step": 4538
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8694795370101929,
+      "learning_rate": 1.7345242286950825e-05,
+      "loss": 1.0406,
+      "step": 4539
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8059786558151245,
+      "learning_rate": 1.7343981617607423e-05,
+      "loss": 0.9805,
+      "step": 4540
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.116225004196167,
+      "learning_rate": 1.734272069484373e-05,
+      "loss": 1.0873,
+      "step": 4541
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0251054763793945,
+      "learning_rate": 1.7341459518703256e-05,
+      "loss": 1.066,
+      "step": 4542
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8019906282424927,
+      "learning_rate": 1.7340198089229522e-05,
+      "loss": 1.0596,
+      "step": 4543
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6431965827941895,
+      "learning_rate": 1.7338936406466052e-05,
+      "loss": 1.0467,
+      "step": 4544
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8506948947906494,
+      "learning_rate": 1.7337674470456394e-05,
+      "loss": 1.052,
+      "step": 4545
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7110610008239746,
+      "learning_rate": 1.7336412281244085e-05,
+      "loss": 1.0088,
+      "step": 4546
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7550017833709717,
+      "learning_rate": 1.733514983887268e-05,
+      "loss": 1.0249,
+      "step": 4547
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.937481164932251,
+      "learning_rate": 1.7333887143385742e-05,
+      "loss": 0.9878,
+      "step": 4548
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8417145013809204,
+      "learning_rate": 1.7332624194826847e-05,
+      "loss": 1.0479,
+      "step": 4549
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8994860649108887,
+      "learning_rate": 1.7331360993239577e-05,
+      "loss": 1.0669,
+      "step": 4550
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.829964280128479,
+      "learning_rate": 1.7330097538667513e-05,
+      "loss": 1.0346,
+      "step": 4551
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7882614135742188,
+      "learning_rate": 1.7328833831154264e-05,
+      "loss": 1.0568,
+      "step": 4552
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.729219913482666,
+      "learning_rate": 1.7327569870743427e-05,
+      "loss": 1.0643,
+      "step": 4553
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9621493816375732,
+      "learning_rate": 1.7326305657478626e-05,
+      "loss": 1.0187,
+      "step": 4554
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9608837962150574,
+      "learning_rate": 1.732504119140348e-05,
+      "loss": 0.5806,
+      "step": 4555
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.086506962776184,
+      "learning_rate": 1.7323776472561625e-05,
+      "loss": 0.6169,
+      "step": 4556
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2444100379943848,
+      "learning_rate": 1.7322511500996704e-05,
+      "loss": 1.0632,
+      "step": 4557
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7177653312683105,
+      "learning_rate": 1.732124627675236e-05,
+      "loss": 0.9777,
+      "step": 4558
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7906980514526367,
+      "learning_rate": 1.7319980799872266e-05,
+      "loss": 0.9654,
+      "step": 4559
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.863255262374878,
+      "learning_rate": 1.7318715070400075e-05,
+      "loss": 1.0643,
+      "step": 4560
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7572040557861328,
+      "learning_rate": 1.7317449088379477e-05,
+      "loss": 1.0482,
+      "step": 4561
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8347163200378418,
+      "learning_rate": 1.7316182853854147e-05,
+      "loss": 1.0712,
+      "step": 4562
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1304402351379395,
+      "learning_rate": 1.7314916366867784e-05,
+      "loss": 0.9882,
+      "step": 4563
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.151083469390869,
+      "learning_rate": 1.731364962746409e-05,
+      "loss": 1.0339,
+      "step": 4564
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1241965293884277,
+      "learning_rate": 1.731238263568678e-05,
+      "loss": 1.0243,
+      "step": 4565
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1738831996917725,
+      "learning_rate": 1.731111539157957e-05,
+      "loss": 1.0225,
+      "step": 4566
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8628838062286377,
+      "learning_rate": 1.730984789518619e-05,
+      "loss": 1.0596,
+      "step": 4567
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.861246109008789,
+      "learning_rate": 1.7308580146550382e-05,
+      "loss": 1.038,
+      "step": 4568
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.05255126953125,
+      "learning_rate": 1.7307312145715887e-05,
+      "loss": 0.9758,
+      "step": 4569
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.001434564590454,
+      "learning_rate": 1.730604389272646e-05,
+      "loss": 1.0073,
+      "step": 4570
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8627930879592896,
+      "learning_rate": 1.7304775387625867e-05,
+      "loss": 1.0492,
+      "step": 4571
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.665010690689087,
+      "learning_rate": 1.7303506630457882e-05,
+      "loss": 1.1345,
+      "step": 4572
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3181103467941284,
+      "learning_rate": 1.7302237621266283e-05,
+      "loss": 0.6996,
+      "step": 4573
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8638582229614258,
+      "learning_rate": 1.7300968360094863e-05,
+      "loss": 1.1246,
+      "step": 4574
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7731599807739258,
+      "learning_rate": 1.7299698846987422e-05,
+      "loss": 1.0939,
+      "step": 4575
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.907395362854004,
+      "learning_rate": 1.729842908198776e-05,
+      "loss": 1.0166,
+      "step": 4576
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8753777742385864,
+      "learning_rate": 1.72971590651397e-05,
+      "loss": 1.104,
+      "step": 4577
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8011924028396606,
+      "learning_rate": 1.729588879648706e-05,
+      "loss": 1.0512,
+      "step": 4578
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0923328399658203,
+      "learning_rate": 1.7294618276073684e-05,
+      "loss": 1.1141,
+      "step": 4579
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.905290126800537,
+      "learning_rate": 1.729334750394341e-05,
+      "loss": 1.0355,
+      "step": 4580
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7892175912857056,
+      "learning_rate": 1.7292076480140078e-05,
+      "loss": 1.0007,
+      "step": 4581
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9723823070526123,
+      "learning_rate": 1.7290805204707563e-05,
+      "loss": 1.0848,
+      "step": 4582
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8557202816009521,
+      "learning_rate": 1.7289533677689724e-05,
+      "loss": 1.0225,
+      "step": 4583
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7672028541564941,
+      "learning_rate": 1.728826189913044e-05,
+      "loss": 1.0743,
+      "step": 4584
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8175417184829712,
+      "learning_rate": 1.7286989869073597e-05,
+      "loss": 1.0305,
+      "step": 4585
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7689151763916016,
+      "learning_rate": 1.728571758756309e-05,
+      "loss": 0.9986,
+      "step": 4586
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8418059349060059,
+      "learning_rate": 1.728444505464282e-05,
+      "loss": 1.0289,
+      "step": 4587
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8006080389022827,
+      "learning_rate": 1.7283172270356702e-05,
+      "loss": 0.9365,
+      "step": 4588
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.159023404121399,
+      "learning_rate": 1.7281899234748648e-05,
+      "loss": 0.6608,
+      "step": 4589
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9467965364456177,
+      "learning_rate": 1.7280625947862594e-05,
+      "loss": 1.0871,
+      "step": 4590
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8746761083602905,
+      "learning_rate": 1.727935240974248e-05,
+      "loss": 1.0334,
+      "step": 4591
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9868035316467285,
+      "learning_rate": 1.7278078620432247e-05,
+      "loss": 1.1233,
+      "step": 4592
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8632564544677734,
+      "learning_rate": 1.727680457997585e-05,
+      "loss": 1.0035,
+      "step": 4593
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6722391843795776,
+      "learning_rate": 1.7275530288417256e-05,
+      "loss": 1.0509,
+      "step": 4594
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.812365174293518,
+      "learning_rate": 1.7274255745800433e-05,
+      "loss": 0.9502,
+      "step": 4595
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8872088193893433,
+      "learning_rate": 1.7272980952169364e-05,
+      "loss": 1.0553,
+      "step": 4596
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7986476421356201,
+      "learning_rate": 1.727170590756804e-05,
+      "loss": 1.0079,
+      "step": 4597
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8733775615692139,
+      "learning_rate": 1.7270430612040456e-05,
+      "loss": 1.0904,
+      "step": 4598
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9841808080673218,
+      "learning_rate": 1.7269155065630624e-05,
+      "loss": 1.0962,
+      "step": 4599
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9636262655258179,
+      "learning_rate": 1.7267879268382556e-05,
+      "loss": 1.0425,
+      "step": 4600
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.87576162815094,
+      "learning_rate": 1.7266603220340273e-05,
+      "loss": 0.9587,
+      "step": 4601
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8474199771881104,
+      "learning_rate": 1.7265326921547815e-05,
+      "loss": 0.9974,
+      "step": 4602
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.010124921798706,
+      "learning_rate": 1.7264050372049216e-05,
+      "loss": 1.054,
+      "step": 4603
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1257333755493164,
+      "learning_rate": 1.726277357188853e-05,
+      "loss": 1.0429,
+      "step": 4604
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0573055744171143,
+      "learning_rate": 1.7261496521109817e-05,
+      "loss": 1.062,
+      "step": 4605
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0497677326202393,
+      "learning_rate": 1.7260219219757145e-05,
+      "loss": 1.0711,
+      "step": 4606
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.906580924987793,
+      "learning_rate": 1.7258941667874587e-05,
+      "loss": 1.0824,
+      "step": 4607
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9586677551269531,
+      "learning_rate": 1.725766386550623e-05,
+      "loss": 1.0637,
+      "step": 4608
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9318772554397583,
+      "learning_rate": 1.7256385812696165e-05,
+      "loss": 1.0484,
+      "step": 4609
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.085712194442749,
+      "learning_rate": 1.7255107509488494e-05,
+      "loss": 1.0499,
+      "step": 4610
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8690803050994873,
+      "learning_rate": 1.725382895592733e-05,
+      "loss": 1.0317,
+      "step": 4611
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1537272930145264,
+      "learning_rate": 1.7252550152056795e-05,
+      "loss": 0.6377,
+      "step": 4612
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1774359941482544,
+      "learning_rate": 1.725127109792101e-05,
+      "loss": 0.6328,
+      "step": 4613
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.14308762550354,
+      "learning_rate": 1.7249991793564116e-05,
+      "loss": 1.065,
+      "step": 4614
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4846396446228027,
+      "learning_rate": 1.7248712239030257e-05,
+      "loss": 1.0505,
+      "step": 4615
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9136874675750732,
+      "learning_rate": 1.724743243436359e-05,
+      "loss": 0.9939,
+      "step": 4616
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1476850509643555,
+      "learning_rate": 1.724615237960827e-05,
+      "loss": 1.067,
+      "step": 4617
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1069937944412231,
+      "learning_rate": 1.7244872074808478e-05,
+      "loss": 0.6178,
+      "step": 4618
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8680189847946167,
+      "learning_rate": 1.7243591520008384e-05,
+      "loss": 0.9612,
+      "step": 4619
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7486422061920166,
+      "learning_rate": 1.724231071525218e-05,
+      "loss": 1.1352,
+      "step": 4620
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7785797119140625,
+      "learning_rate": 1.7241029660584068e-05,
+      "loss": 0.9916,
+      "step": 4621
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8769954442977905,
+      "learning_rate": 1.7239748356048248e-05,
+      "loss": 0.9827,
+      "step": 4622
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.692273736000061,
+      "learning_rate": 1.7238466801688934e-05,
+      "loss": 0.9932,
+      "step": 4623
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6582568883895874,
+      "learning_rate": 1.723718499755035e-05,
+      "loss": 1.0232,
+      "step": 4624
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.938091516494751,
+      "learning_rate": 1.723590294367673e-05,
+      "loss": 0.9657,
+      "step": 4625
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0084519386291504,
+      "learning_rate": 1.7234620640112313e-05,
+      "loss": 1.003,
+      "step": 4626
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7339975833892822,
+      "learning_rate": 1.7233338086901342e-05,
+      "loss": 1.0643,
+      "step": 4627
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8406918048858643,
+      "learning_rate": 1.7232055284088085e-05,
+      "loss": 0.9802,
+      "step": 4628
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.019582509994507,
+      "learning_rate": 1.72307722317168e-05,
+      "loss": 1.147,
+      "step": 4629
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.017970085144043,
+      "learning_rate": 1.7229488929831762e-05,
+      "loss": 1.0419,
+      "step": 4630
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8190929889678955,
+      "learning_rate": 1.7228205378477258e-05,
+      "loss": 0.9979,
+      "step": 4631
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1298515796661377,
+      "learning_rate": 1.7226921577697575e-05,
+      "loss": 1.0833,
+      "step": 4632
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0303947925567627,
+      "learning_rate": 1.7225637527537017e-05,
+      "loss": 1.0403,
+      "step": 4633
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7437316179275513,
+      "learning_rate": 1.7224353228039893e-05,
+      "loss": 0.9475,
+      "step": 4634
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7206364870071411,
+      "learning_rate": 1.722306867925052e-05,
+      "loss": 1.0109,
+      "step": 4635
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8375470638275146,
+      "learning_rate": 1.7221783881213222e-05,
+      "loss": 1.0014,
+      "step": 4636
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2050999402999878,
+      "learning_rate": 1.7220498833972333e-05,
+      "loss": 0.5861,
+      "step": 4637
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9113013744354248,
+      "learning_rate": 1.7219213537572203e-05,
+      "loss": 1.0175,
+      "step": 4638
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2827603816986084,
+      "learning_rate": 1.721792799205718e-05,
+      "loss": 1.0758,
+      "step": 4639
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.275219440460205,
+      "learning_rate": 1.7216642197471626e-05,
+      "loss": 1.11,
+      "step": 4640
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7857520580291748,
+      "learning_rate": 1.7215356153859904e-05,
+      "loss": 1.124,
+      "step": 4641
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6672146320343018,
+      "learning_rate": 1.7214069861266398e-05,
+      "loss": 1.0573,
+      "step": 4642
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7119561433792114,
+      "learning_rate": 1.7212783319735492e-05,
+      "loss": 0.9232,
+      "step": 4643
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6102291345596313,
+      "learning_rate": 1.7211496529311582e-05,
+      "loss": 1.022,
+      "step": 4644
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.908078908920288,
+      "learning_rate": 1.7210209490039075e-05,
+      "loss": 1.0198,
+      "step": 4645
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9277675151824951,
+      "learning_rate": 1.7208922201962376e-05,
+      "loss": 1.0722,
+      "step": 4646
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9552894830703735,
+      "learning_rate": 1.7207634665125907e-05,
+      "loss": 1.0283,
+      "step": 4647
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.826055645942688,
+      "learning_rate": 1.7206346879574104e-05,
+      "loss": 1.0395,
+      "step": 4648
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8464971780776978,
+      "learning_rate": 1.7205058845351398e-05,
+      "loss": 0.9822,
+      "step": 4649
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8195899724960327,
+      "learning_rate": 1.7203770562502238e-05,
+      "loss": 1.0106,
+      "step": 4650
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8553869724273682,
+      "learning_rate": 1.720248203107108e-05,
+      "loss": 1.0796,
+      "step": 4651
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1967116594314575,
+      "learning_rate": 1.7201193251102383e-05,
+      "loss": 0.6502,
+      "step": 4652
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8553239107131958,
+      "learning_rate": 1.7199904222640627e-05,
+      "loss": 1.0031,
+      "step": 4653
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.784236192703247,
+      "learning_rate": 1.7198614945730287e-05,
+      "loss": 1.0453,
+      "step": 4654
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9137392044067383,
+      "learning_rate": 1.7197325420415852e-05,
+      "loss": 1.0384,
+      "step": 4655
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.961923599243164,
+      "learning_rate": 1.7196035646741824e-05,
+      "loss": 1.1053,
+      "step": 4656
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.788831114768982,
+      "learning_rate": 1.7194745624752704e-05,
+      "loss": 1.0171,
+      "step": 4657
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0162508487701416,
+      "learning_rate": 1.7193455354493013e-05,
+      "loss": 1.0585,
+      "step": 4658
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7559564113616943,
+      "learning_rate": 1.719216483600727e-05,
+      "loss": 1.032,
+      "step": 4659
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7875580787658691,
+      "learning_rate": 1.7190874069340015e-05,
+      "loss": 0.9777,
+      "step": 4660
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8427859544754028,
+      "learning_rate": 1.7189583054535775e-05,
+      "loss": 1.0401,
+      "step": 4661
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8399351835250854,
+      "learning_rate": 1.7188291791639108e-05,
+      "loss": 0.9954,
+      "step": 4662
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.886370062828064,
+      "learning_rate": 1.7187000280694572e-05,
+      "loss": 1.1176,
+      "step": 4663
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9610744714736938,
+      "learning_rate": 1.7185708521746734e-05,
+      "loss": 1.0222,
+      "step": 4664
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8299083709716797,
+      "learning_rate": 1.718441651484017e-05,
+      "loss": 1.0333,
+      "step": 4665
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9010825157165527,
+      "learning_rate": 1.7183124260019454e-05,
+      "loss": 0.9701,
+      "step": 4666
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6604851484298706,
+      "learning_rate": 1.7181831757329187e-05,
+      "loss": 1.0368,
+      "step": 4667
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9394150972366333,
+      "learning_rate": 1.7180539006813973e-05,
+      "loss": 0.971,
+      "step": 4668
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8684107065200806,
+      "learning_rate": 1.717924600851841e-05,
+      "loss": 0.9659,
+      "step": 4669
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7895206212997437,
+      "learning_rate": 1.7177952762487125e-05,
+      "loss": 0.9975,
+      "step": 4670
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.070091962814331,
+      "learning_rate": 1.717665926876474e-05,
+      "loss": 1.039,
+      "step": 4671
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7868837118148804,
+      "learning_rate": 1.717536552739589e-05,
+      "loss": 0.9915,
+      "step": 4672
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7172470092773438,
+      "learning_rate": 1.717407153842522e-05,
+      "loss": 1.0347,
+      "step": 4673
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7209681272506714,
+      "learning_rate": 1.7172777301897382e-05,
+      "loss": 1.016,
+      "step": 4674
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8555841445922852,
+      "learning_rate": 1.7171482817857034e-05,
+      "loss": 0.9958,
+      "step": 4675
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7856978178024292,
+      "learning_rate": 1.7170188086348847e-05,
+      "loss": 1.0656,
+      "step": 4676
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6724331378936768,
+      "learning_rate": 1.7168893107417498e-05,
+      "loss": 1.0039,
+      "step": 4677
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0794757604599,
+      "learning_rate": 1.7167597881107673e-05,
+      "loss": 0.6129,
+      "step": 4678
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8575519323349,
+      "learning_rate": 1.716630240746407e-05,
+      "loss": 1.0745,
+      "step": 4679
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1793243885040283,
+      "learning_rate": 1.7165006686531387e-05,
+      "loss": 0.9818,
+      "step": 4680
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.378910779953003,
+      "learning_rate": 1.7163710718354338e-05,
+      "loss": 1.0682,
+      "step": 4681
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.91934072971344,
+      "learning_rate": 1.7162414502977643e-05,
+      "loss": 1.009,
+      "step": 4682
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.720177173614502,
+      "learning_rate": 1.716111804044603e-05,
+      "loss": 1.0029,
+      "step": 4683
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7971895933151245,
+      "learning_rate": 1.7159821330804237e-05,
+      "loss": 1.0726,
+      "step": 4684
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.093899965286255,
+      "learning_rate": 1.715852437409701e-05,
+      "loss": 1.0698,
+      "step": 4685
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9821538925170898,
+      "learning_rate": 1.715722717036911e-05,
+      "loss": 1.1076,
+      "step": 4686
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.133533239364624,
+      "learning_rate": 1.7155929719665286e-05,
+      "loss": 1.0199,
+      "step": 4687
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7891730070114136,
+      "learning_rate": 1.7154632022030318e-05,
+      "loss": 1.0255,
+      "step": 4688
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8691664934158325,
+      "learning_rate": 1.7153334077508983e-05,
+      "loss": 0.9721,
+      "step": 4689
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7908364534378052,
+      "learning_rate": 1.7152035886146076e-05,
+      "loss": 1.0807,
+      "step": 4690
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7693136930465698,
+      "learning_rate": 1.7150737447986384e-05,
+      "loss": 1.0969,
+      "step": 4691
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6246225833892822,
+      "learning_rate": 1.714943876307472e-05,
+      "loss": 1.0339,
+      "step": 4692
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.029156446456909,
+      "learning_rate": 1.7148139831455896e-05,
+      "loss": 0.967,
+      "step": 4693
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7252434492111206,
+      "learning_rate": 1.7146840653174732e-05,
+      "loss": 0.9975,
+      "step": 4694
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.956209421157837,
+      "learning_rate": 1.7145541228276063e-05,
+      "loss": 1.0548,
+      "step": 4695
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.685001015663147,
+      "learning_rate": 1.7144241556804724e-05,
+      "loss": 0.9349,
+      "step": 4696
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8997920751571655,
+      "learning_rate": 1.7142941638805564e-05,
+      "loss": 1.029,
+      "step": 4697
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.901615858078003,
+      "learning_rate": 1.714164147432345e-05,
+      "loss": 1.077,
+      "step": 4698
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8288229703903198,
+      "learning_rate": 1.7140341063403226e-05,
+      "loss": 1.0495,
+      "step": 4699
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5870082378387451,
+      "learning_rate": 1.7139040406089786e-05,
+      "loss": 1.0468,
+      "step": 4700
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9229363203048706,
+      "learning_rate": 1.7137739502428005e-05,
+      "loss": 1.0361,
+      "step": 4701
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7878434658050537,
+      "learning_rate": 1.7136438352462764e-05,
+      "loss": 1.0346,
+      "step": 4702
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8636884689331055,
+      "learning_rate": 1.7135136956238977e-05,
+      "loss": 1.1064,
+      "step": 4703
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6639491319656372,
+      "learning_rate": 1.713383531380154e-05,
+      "loss": 1.0152,
+      "step": 4704
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8125758171081543,
+      "learning_rate": 1.713253342519538e-05,
+      "loss": 1.0325,
+      "step": 4705
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9059473276138306,
+      "learning_rate": 1.713123129046541e-05,
+      "loss": 0.9953,
+      "step": 4706
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7418614625930786,
+      "learning_rate": 1.7129928909656573e-05,
+      "loss": 1.0012,
+      "step": 4707
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7379263639450073,
+      "learning_rate": 1.7128626282813803e-05,
+      "loss": 0.9819,
+      "step": 4708
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8412891626358032,
+      "learning_rate": 1.7127323409982053e-05,
+      "loss": 1.0359,
+      "step": 4709
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7066762447357178,
+      "learning_rate": 1.712602029120628e-05,
+      "loss": 1.0111,
+      "step": 4710
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9225943088531494,
+      "learning_rate": 1.7124716926531454e-05,
+      "loss": 1.0073,
+      "step": 4711
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9674371480941772,
+      "learning_rate": 1.712341331600255e-05,
+      "loss": 1.0675,
+      "step": 4712
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8190809488296509,
+      "learning_rate": 1.712210945966455e-05,
+      "loss": 1.1216,
+      "step": 4713
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9054967164993286,
+      "learning_rate": 1.712080535756245e-05,
+      "loss": 0.9991,
+      "step": 4714
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8096667528152466,
+      "learning_rate": 1.7119501009741245e-05,
+      "loss": 1.0812,
+      "step": 4715
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6954387426376343,
+      "learning_rate": 1.7118196416245947e-05,
+      "loss": 0.9051,
+      "step": 4716
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1597070693969727,
+      "learning_rate": 1.7116891577121576e-05,
+      "loss": 1.0382,
+      "step": 4717
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1272761821746826,
+      "learning_rate": 1.711558649241316e-05,
+      "loss": 1.0742,
+      "step": 4718
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9281593561172485,
+      "learning_rate": 1.7114281162165726e-05,
+      "loss": 1.0587,
+      "step": 4719
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7473527193069458,
+      "learning_rate": 1.7112975586424327e-05,
+      "loss": 1.0668,
+      "step": 4720
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8329812288284302,
+      "learning_rate": 1.7111669765234006e-05,
+      "loss": 0.9759,
+      "step": 4721
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7359579801559448,
+      "learning_rate": 1.7110363698639826e-05,
+      "loss": 1.0285,
+      "step": 4722
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8638657331466675,
+      "learning_rate": 1.710905738668686e-05,
+      "loss": 1.1208,
+      "step": 4723
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8270941972732544,
+      "learning_rate": 1.7107750829420177e-05,
+      "loss": 1.0529,
+      "step": 4724
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8194390535354614,
+      "learning_rate": 1.7106444026884873e-05,
+      "loss": 1.0184,
+      "step": 4725
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8500094413757324,
+      "learning_rate": 1.7105136979126036e-05,
+      "loss": 1.034,
+      "step": 4726
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8671813011169434,
+      "learning_rate": 1.7103829686188766e-05,
+      "loss": 0.9974,
+      "step": 4727
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7800992727279663,
+      "learning_rate": 1.710252214811818e-05,
+      "loss": 1.0502,
+      "step": 4728
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.939017415046692,
+      "learning_rate": 1.7101214364959392e-05,
+      "loss": 0.9988,
+      "step": 4729
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0655815601348877,
+      "learning_rate": 1.7099906336757533e-05,
+      "loss": 1.1468,
+      "step": 4730
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6860171556472778,
+      "learning_rate": 1.7098598063557744e-05,
+      "loss": 1.0033,
+      "step": 4731
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7304410934448242,
+      "learning_rate": 1.709728954540516e-05,
+      "loss": 0.9844,
+      "step": 4732
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9833543300628662,
+      "learning_rate": 1.7095980782344942e-05,
+      "loss": 1.1093,
+      "step": 4733
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8146415948867798,
+      "learning_rate": 1.7094671774422245e-05,
+      "loss": 0.9643,
+      "step": 4734
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.025599956512451,
+      "learning_rate": 1.709336252168225e-05,
+      "loss": 1.0958,
+      "step": 4735
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.826157569885254,
+      "learning_rate": 1.7092053024170122e-05,
+      "loss": 0.9876,
+      "step": 4736
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.041050910949707,
+      "learning_rate": 1.709074328193106e-05,
+      "loss": 1.033,
+      "step": 4737
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7011879682540894,
+      "learning_rate": 1.7089433295010252e-05,
+      "loss": 1.0384,
+      "step": 4738
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9061089754104614,
+      "learning_rate": 1.7088123063452905e-05,
+      "loss": 0.9956,
+      "step": 4739
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.837290644645691,
+      "learning_rate": 1.7086812587304233e-05,
+      "loss": 1.0254,
+      "step": 4740
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9794304370880127,
+      "learning_rate": 1.7085501866609455e-05,
+      "loss": 1.1418,
+      "step": 4741
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.885650634765625,
+      "learning_rate": 1.7084190901413795e-05,
+      "loss": 1.0039,
+      "step": 4742
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.779986023902893,
+      "learning_rate": 1.70828796917625e-05,
+      "loss": 1.0113,
+      "step": 4743
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.76728355884552,
+      "learning_rate": 1.708156823770081e-05,
+      "loss": 1.0148,
+      "step": 4744
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7319813966751099,
+      "learning_rate": 1.7080256539273984e-05,
+      "loss": 0.9863,
+      "step": 4745
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9736579656600952,
+      "learning_rate": 1.707894459652728e-05,
+      "loss": 1.0475,
+      "step": 4746
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.776868224143982,
+      "learning_rate": 1.7077632409505974e-05,
+      "loss": 0.9779,
+      "step": 4747
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.718104600906372,
+      "learning_rate": 1.7076319978255345e-05,
+      "loss": 1.0393,
+      "step": 4748
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0262227058410645,
+      "learning_rate": 1.707500730282068e-05,
+      "loss": 0.9863,
+      "step": 4749
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9351303577423096,
+      "learning_rate": 1.7073694383247273e-05,
+      "loss": 1.0606,
+      "step": 4750
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0892128944396973,
+      "learning_rate": 1.7072381219580438e-05,
+      "loss": 1.0418,
+      "step": 4751
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8069987297058105,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 1.0253,
+      "step": 4752
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8135216236114502,
+      "learning_rate": 1.706975416014772e-05,
+      "loss": 1.0444,
+      "step": 4753
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8808784484863281,
+      "learning_rate": 1.7068440264472496e-05,
+      "loss": 1.0551,
+      "step": 4754
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0072484016418457,
+      "learning_rate": 1.7067126124885144e-05,
+      "loss": 1.0224,
+      "step": 4755
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8094242811203003,
+      "learning_rate": 1.706581174143101e-05,
+      "loss": 0.9853,
+      "step": 4756
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.072641611099243,
+      "learning_rate": 1.706449711415545e-05,
+      "loss": 1.0902,
+      "step": 4757
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8186986446380615,
+      "learning_rate": 1.706318224310383e-05,
+      "loss": 1.0231,
+      "step": 4758
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7402868270874023,
+      "learning_rate": 1.7061867128321524e-05,
+      "loss": 0.9819,
+      "step": 4759
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.684104561805725,
+      "learning_rate": 1.7060551769853904e-05,
+      "loss": 1.039,
+      "step": 4760
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9414888620376587,
+      "learning_rate": 1.7059236167746367e-05,
+      "loss": 1.0714,
+      "step": 4761
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.885617971420288,
+      "learning_rate": 1.705792032204431e-05,
+      "loss": 1.0369,
+      "step": 4762
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7878416776657104,
+      "learning_rate": 1.7056604232793137e-05,
+      "loss": 1.0263,
+      "step": 4763
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8834227323532104,
+      "learning_rate": 1.7055287900038264e-05,
+      "loss": 0.9933,
+      "step": 4764
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9338070154190063,
+      "learning_rate": 1.7053971323825114e-05,
+      "loss": 1.0804,
+      "step": 4765
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8026280403137207,
+      "learning_rate": 1.705265450419912e-05,
+      "loss": 0.9992,
+      "step": 4766
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1608020067214966,
+      "learning_rate": 1.705133744120572e-05,
+      "loss": 0.5857,
+      "step": 4767
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1691300868988037,
+      "learning_rate": 1.705002013489036e-05,
+      "loss": 1.0393,
+      "step": 4768
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1441478729248047,
+      "learning_rate": 1.7048702585298493e-05,
+      "loss": 0.5956,
+      "step": 4769
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8856031894683838,
+      "learning_rate": 1.7047384792475597e-05,
+      "loss": 0.9926,
+      "step": 4770
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7053017616271973,
+      "learning_rate": 1.7046066756467134e-05,
+      "loss": 1.0177,
+      "step": 4771
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7317636013031006,
+      "learning_rate": 1.7044748477318595e-05,
+      "loss": 0.9415,
+      "step": 4772
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.906692624092102,
+      "learning_rate": 1.704342995507546e-05,
+      "loss": 1.0166,
+      "step": 4773
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8884797096252441,
+      "learning_rate": 1.704211118978323e-05,
+      "loss": 1.0386,
+      "step": 4774
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.837873101234436,
+      "learning_rate": 1.7040792181487423e-05,
+      "loss": 1.0969,
+      "step": 4775
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0613327026367188,
+      "learning_rate": 1.703947293023354e-05,
+      "loss": 1.0391,
+      "step": 4776
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1687113046646118,
+      "learning_rate": 1.703815343606711e-05,
+      "loss": 0.6493,
+      "step": 4777
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.845878005027771,
+      "learning_rate": 1.7036833699033665e-05,
+      "loss": 1.0088,
+      "step": 4778
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6320486068725586,
+      "learning_rate": 1.7035513719178747e-05,
+      "loss": 0.9654,
+      "step": 4779
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8643770217895508,
+      "learning_rate": 1.7034193496547903e-05,
+      "loss": 1.0328,
+      "step": 4780
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.97599196434021,
+      "learning_rate": 1.703287303118669e-05,
+      "loss": 1.0173,
+      "step": 4781
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8334367275238037,
+      "learning_rate": 1.7031552323140674e-05,
+      "loss": 1.0995,
+      "step": 4782
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1179115772247314,
+      "learning_rate": 1.703023137245543e-05,
+      "loss": 1.1057,
+      "step": 4783
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9273347854614258,
+      "learning_rate": 1.702891017917654e-05,
+      "loss": 1.0102,
+      "step": 4784
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.70667564868927,
+      "learning_rate": 1.7027588743349596e-05,
+      "loss": 0.9865,
+      "step": 4785
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6817822456359863,
+      "learning_rate": 1.7026267065020193e-05,
+      "loss": 0.9957,
+      "step": 4786
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9020365476608276,
+      "learning_rate": 1.702494514423394e-05,
+      "loss": 1.044,
+      "step": 4787
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8188600540161133,
+      "learning_rate": 1.7023622981036454e-05,
+      "loss": 1.0867,
+      "step": 4788
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.997410774230957,
+      "learning_rate": 1.7022300575473356e-05,
+      "loss": 1.0236,
+      "step": 4789
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1429548263549805,
+      "learning_rate": 1.7020977927590286e-05,
+      "loss": 1.075,
+      "step": 4790
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8220083713531494,
+      "learning_rate": 1.7019655037432875e-05,
+      "loss": 1.1165,
+      "step": 4791
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.569589376449585,
+      "learning_rate": 1.701833190504678e-05,
+      "loss": 1.0598,
+      "step": 4792
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8345558643341064,
+      "learning_rate": 1.7017008530477658e-05,
+      "loss": 1.0093,
+      "step": 4793
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2149317264556885,
+      "learning_rate": 1.701568491377117e-05,
+      "loss": 0.6669,
+      "step": 4794
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7364712953567505,
+      "learning_rate": 1.7014361054972995e-05,
+      "loss": 1.081,
+      "step": 4795
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9525762796401978,
+      "learning_rate": 1.701303695412881e-05,
+      "loss": 1.1081,
+      "step": 4796
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8129020929336548,
+      "learning_rate": 1.7011712611284316e-05,
+      "loss": 1.0315,
+      "step": 4797
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9507430791854858,
+      "learning_rate": 1.7010388026485202e-05,
+      "loss": 1.0581,
+      "step": 4798
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.078479290008545,
+      "learning_rate": 1.700906319977718e-05,
+      "loss": 1.0721,
+      "step": 4799
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0087268352508545,
+      "learning_rate": 1.7007738131205966e-05,
+      "loss": 1.0073,
+      "step": 4800
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.795243740081787,
+      "learning_rate": 1.7006412820817288e-05,
+      "loss": 1.0518,
+      "step": 4801
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9656835794448853,
+      "learning_rate": 1.700508726865687e-05,
+      "loss": 1.0878,
+      "step": 4802
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.929245948791504,
+      "learning_rate": 1.7003761474770462e-05,
+      "loss": 0.9508,
+      "step": 4803
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7213767766952515,
+      "learning_rate": 1.700243543920381e-05,
+      "loss": 0.9865,
+      "step": 4804
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7619291543960571,
+      "learning_rate": 1.7001109162002668e-05,
+      "loss": 0.9814,
+      "step": 4805
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5582529306411743,
+      "learning_rate": 1.699978264321281e-05,
+      "loss": 1.093,
+      "step": 4806
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6498557329177856,
+      "learning_rate": 1.6998455882880002e-05,
+      "loss": 0.9467,
+      "step": 4807
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9310967922210693,
+      "learning_rate": 1.6997128881050028e-05,
+      "loss": 1.049,
+      "step": 4808
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8834171295166016,
+      "learning_rate": 1.6995801637768687e-05,
+      "loss": 1.0561,
+      "step": 4809
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9829078912734985,
+      "learning_rate": 1.6994474153081774e-05,
+      "loss": 0.9531,
+      "step": 4810
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7405204772949219,
+      "learning_rate": 1.6993146427035093e-05,
+      "loss": 0.9849,
+      "step": 4811
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.348369598388672,
+      "learning_rate": 1.699181845967447e-05,
+      "loss": 1.1075,
+      "step": 4812
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9127343893051147,
+      "learning_rate": 1.6990490251045717e-05,
+      "loss": 1.0327,
+      "step": 4813
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8989533185958862,
+      "learning_rate": 1.6989161801194675e-05,
+      "loss": 1.0079,
+      "step": 4814
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.642884612083435,
+      "learning_rate": 1.698783311016718e-05,
+      "loss": 0.9756,
+      "step": 4815
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.902197003364563,
+      "learning_rate": 1.6986504178009085e-05,
+      "loss": 0.9689,
+      "step": 4816
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7135981321334839,
+      "learning_rate": 1.698517500476625e-05,
+      "loss": 1.0756,
+      "step": 4817
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9768487215042114,
+      "learning_rate": 1.6983845590484535e-05,
+      "loss": 0.6036,
+      "step": 4818
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9502159357070923,
+      "learning_rate": 1.698251593520982e-05,
+      "loss": 0.9728,
+      "step": 4819
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.913528561592102,
+      "learning_rate": 1.698118603898798e-05,
+      "loss": 1.0414,
+      "step": 4820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8836307525634766,
+      "learning_rate": 1.6979855901864914e-05,
+      "loss": 0.9903,
+      "step": 4821
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7608457803726196,
+      "learning_rate": 1.6978525523886515e-05,
+      "loss": 1.0579,
+      "step": 4822
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8165781497955322,
+      "learning_rate": 1.69771949050987e-05,
+      "loss": 0.9679,
+      "step": 4823
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.764281153678894,
+      "learning_rate": 1.6975864045547373e-05,
+      "loss": 1.0234,
+      "step": 4824
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.777406930923462,
+      "learning_rate": 1.6974532945278468e-05,
+      "loss": 1.0061,
+      "step": 4825
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.918281078338623,
+      "learning_rate": 1.697320160433791e-05,
+      "loss": 1.0516,
+      "step": 4826
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0181868076324463,
+      "learning_rate": 1.6971870022771648e-05,
+      "loss": 1.0691,
+      "step": 4827
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.971143364906311,
+      "learning_rate": 1.6970538200625622e-05,
+      "loss": 1.1327,
+      "step": 4828
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7758426666259766,
+      "learning_rate": 1.6969206137945797e-05,
+      "loss": 1.0397,
+      "step": 4829
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9450143575668335,
+      "learning_rate": 1.6967873834778136e-05,
+      "loss": 1.0463,
+      "step": 4830
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8333852291107178,
+      "learning_rate": 1.6966541291168616e-05,
+      "loss": 1.0376,
+      "step": 4831
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.133472442626953,
+      "learning_rate": 1.696520850716321e-05,
+      "loss": 1.0861,
+      "step": 4832
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8106203079223633,
+      "learning_rate": 1.6963875482807916e-05,
+      "loss": 1.0421,
+      "step": 4833
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8422704935073853,
+      "learning_rate": 1.6962542218148735e-05,
+      "loss": 1.0832,
+      "step": 4834
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8666435480117798,
+      "learning_rate": 1.696120871323167e-05,
+      "loss": 1.1086,
+      "step": 4835
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.793318510055542,
+      "learning_rate": 1.6959874968102736e-05,
+      "loss": 0.9876,
+      "step": 4836
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8047312498092651,
+      "learning_rate": 1.6958540982807958e-05,
+      "loss": 1.0864,
+      "step": 4837
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8125795125961304,
+      "learning_rate": 1.6957206757393372e-05,
+      "loss": 0.9864,
+      "step": 4838
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.914389967918396,
+      "learning_rate": 1.6955872291905014e-05,
+      "loss": 1.0606,
+      "step": 4839
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7422670125961304,
+      "learning_rate": 1.6954537586388932e-05,
+      "loss": 1.0353,
+      "step": 4840
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8671469688415527,
+      "learning_rate": 1.6953202640891187e-05,
+      "loss": 1.0014,
+      "step": 4841
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7410883903503418,
+      "learning_rate": 1.695186745545784e-05,
+      "loss": 1.0261,
+      "step": 4842
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7394968271255493,
+      "learning_rate": 1.6950532030134966e-05,
+      "loss": 1.002,
+      "step": 4843
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7510324716567993,
+      "learning_rate": 1.6949196364968648e-05,
+      "loss": 0.9897,
+      "step": 4844
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.797247052192688,
+      "learning_rate": 1.6947860460004974e-05,
+      "loss": 1.051,
+      "step": 4845
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8486158847808838,
+      "learning_rate": 1.6946524315290047e-05,
+      "loss": 1.0578,
+      "step": 4846
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8467079401016235,
+      "learning_rate": 1.6945187930869967e-05,
+      "loss": 1.0301,
+      "step": 4847
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8644989728927612,
+      "learning_rate": 1.6943851306790852e-05,
+      "loss": 1.0498,
+      "step": 4848
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7966654300689697,
+      "learning_rate": 1.6942514443098826e-05,
+      "loss": 1.0643,
+      "step": 4849
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6879764795303345,
+      "learning_rate": 1.694117733984002e-05,
+      "loss": 0.9921,
+      "step": 4850
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.915128469467163,
+      "learning_rate": 1.6939839997060575e-05,
+      "loss": 1.0155,
+      "step": 4851
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8353031873703003,
+      "learning_rate": 1.6938502414806633e-05,
+      "loss": 1.0629,
+      "step": 4852
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9062198400497437,
+      "learning_rate": 1.693716459312436e-05,
+      "loss": 1.0046,
+      "step": 4853
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0432519912719727,
+      "learning_rate": 1.6935826532059913e-05,
+      "loss": 1.0641,
+      "step": 4854
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7728166580200195,
+      "learning_rate": 1.6934488231659465e-05,
+      "loss": 0.9788,
+      "step": 4855
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6331583261489868,
+      "learning_rate": 1.69331496919692e-05,
+      "loss": 1.0028,
+      "step": 4856
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9410744905471802,
+      "learning_rate": 1.6931810913035306e-05,
+      "loss": 0.9283,
+      "step": 4857
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8948951959609985,
+      "learning_rate": 1.693047189490398e-05,
+      "loss": 1.0785,
+      "step": 4858
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0742077827453613,
+      "learning_rate": 1.6929132637621433e-05,
+      "loss": 1.0692,
+      "step": 4859
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7253693342208862,
+      "learning_rate": 1.692779314123387e-05,
+      "loss": 1.0015,
+      "step": 4860
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7998379468917847,
+      "learning_rate": 1.6926453405787518e-05,
+      "loss": 0.9652,
+      "step": 4861
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9471490383148193,
+      "learning_rate": 1.6925113431328605e-05,
+      "loss": 1.0556,
+      "step": 4862
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0721354484558105,
+      "learning_rate": 1.6923773217903378e-05,
+      "loss": 1.0313,
+      "step": 4863
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9251173734664917,
+      "learning_rate": 1.692243276555807e-05,
+      "loss": 1.0408,
+      "step": 4864
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7999091148376465,
+      "learning_rate": 1.6921092074338953e-05,
+      "loss": 0.9218,
+      "step": 4865
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8684536218643188,
+      "learning_rate": 1.691975114429228e-05,
+      "loss": 1.0716,
+      "step": 4866
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7986088991165161,
+      "learning_rate": 1.691840997546432e-05,
+      "loss": 1.0215,
+      "step": 4867
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8519375324249268,
+      "learning_rate": 1.6917068567901358e-05,
+      "loss": 1.0603,
+      "step": 4868
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.848416805267334,
+      "learning_rate": 1.6915726921649685e-05,
+      "loss": 1.0526,
+      "step": 4869
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2385339736938477,
+      "learning_rate": 1.691438503675559e-05,
+      "loss": 0.7072,
+      "step": 4870
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.118136405944824,
+      "learning_rate": 1.6913042913265388e-05,
+      "loss": 1.058,
+      "step": 4871
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0882911682128906,
+      "learning_rate": 1.6911700551225382e-05,
+      "loss": 0.6453,
+      "step": 4872
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.727582573890686,
+      "learning_rate": 1.6910357950681898e-05,
+      "loss": 1.0236,
+      "step": 4873
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8469053506851196,
+      "learning_rate": 1.6909015111681265e-05,
+      "loss": 0.9702,
+      "step": 4874
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9877110719680786,
+      "learning_rate": 1.690767203426982e-05,
+      "loss": 1.1292,
+      "step": 4875
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9954696893692017,
+      "learning_rate": 1.6906328718493906e-05,
+      "loss": 1.1002,
+      "step": 4876
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.833004355430603,
+      "learning_rate": 1.690498516439988e-05,
+      "loss": 1.0518,
+      "step": 4877
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9706647396087646,
+      "learning_rate": 1.6903641372034107e-05,
+      "loss": 1.0977,
+      "step": 4878
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.052793264389038,
+      "learning_rate": 1.6902297341442956e-05,
+      "loss": 1.0011,
+      "step": 4879
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9225096702575684,
+      "learning_rate": 1.6900953072672802e-05,
+      "loss": 0.9972,
+      "step": 4880
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9569402933120728,
+      "learning_rate": 1.6899608565770035e-05,
+      "loss": 1.0629,
+      "step": 4881
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0029420852661133,
+      "learning_rate": 1.6898263820781047e-05,
+      "loss": 1.1177,
+      "step": 4882
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7940871715545654,
+      "learning_rate": 1.689691883775225e-05,
+      "loss": 1.0343,
+      "step": 4883
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9395021200180054,
+      "learning_rate": 1.6895573616730046e-05,
+      "loss": 1.0731,
+      "step": 4884
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.554253339767456,
+      "learning_rate": 1.689422815776086e-05,
+      "loss": 1.0751,
+      "step": 4885
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8766225576400757,
+      "learning_rate": 1.6892882460891118e-05,
+      "loss": 1.0117,
+      "step": 4886
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.883284330368042,
+      "learning_rate": 1.6891536526167252e-05,
+      "loss": 1.0988,
+      "step": 4887
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.821362853050232,
+      "learning_rate": 1.689019035363572e-05,
+      "loss": 1.0654,
+      "step": 4888
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2597479820251465,
+      "learning_rate": 1.688884394334296e-05,
+      "loss": 1.0652,
+      "step": 4889
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7758064270019531,
+      "learning_rate": 1.688749729533544e-05,
+      "loss": 1.0065,
+      "step": 4890
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6758724451065063,
+      "learning_rate": 1.688615040965963e-05,
+      "loss": 1.0092,
+      "step": 4891
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0197389125823975,
+      "learning_rate": 1.6884803286362e-05,
+      "loss": 1.1238,
+      "step": 4892
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1135830879211426,
+      "learning_rate": 1.6883455925489044e-05,
+      "loss": 1.0507,
+      "step": 4893
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1909617185592651,
+      "learning_rate": 1.6882108327087252e-05,
+      "loss": 0.6107,
+      "step": 4894
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.937817931175232,
+      "learning_rate": 1.6880760491203124e-05,
+      "loss": 1.0924,
+      "step": 4895
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8181272745132446,
+      "learning_rate": 1.6879412417883175e-05,
+      "loss": 0.9535,
+      "step": 4896
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9131006002426147,
+      "learning_rate": 1.687806410717392e-05,
+      "loss": 0.9977,
+      "step": 4897
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6708403825759888,
+      "learning_rate": 1.6876715559121883e-05,
+      "loss": 0.9352,
+      "step": 4898
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8211203813552856,
+      "learning_rate": 1.6875366773773604e-05,
+      "loss": 0.9778,
+      "step": 4899
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0896263122558594,
+      "learning_rate": 1.6874017751175622e-05,
+      "loss": 1.0195,
+      "step": 4900
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.047825336456299,
+      "learning_rate": 1.687266849137449e-05,
+      "loss": 1.035,
+      "step": 4901
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.026298761367798,
+      "learning_rate": 1.6871318994416766e-05,
+      "loss": 1.0087,
+      "step": 4902
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9729666709899902,
+      "learning_rate": 1.686996926034902e-05,
+      "loss": 1.1111,
+      "step": 4903
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1248681545257568,
+      "learning_rate": 1.686861928921782e-05,
+      "loss": 0.6334,
+      "step": 4904
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1286895275115967,
+      "learning_rate": 1.6867269081069756e-05,
+      "loss": 1.0766,
+      "step": 4905
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.057871103286743,
+      "learning_rate": 1.6865918635951425e-05,
+      "loss": 1.0375,
+      "step": 4906
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.907536268234253,
+      "learning_rate": 1.6864567953909417e-05,
+      "loss": 1.0141,
+      "step": 4907
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9625297784805298,
+      "learning_rate": 1.6863217034990343e-05,
+      "loss": 0.9866,
+      "step": 4908
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8968021869659424,
+      "learning_rate": 1.6861865879240822e-05,
+      "loss": 0.9839,
+      "step": 4909
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8067208528518677,
+      "learning_rate": 1.6860514486707477e-05,
+      "loss": 0.9844,
+      "step": 4910
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.87395441532135,
+      "learning_rate": 1.6859162857436943e-05,
+      "loss": 1.0789,
+      "step": 4911
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.927905559539795,
+      "learning_rate": 1.6857810991475857e-05,
+      "loss": 0.9781,
+      "step": 4912
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0463930368423462,
+      "learning_rate": 1.685645888887087e-05,
+      "loss": 0.6187,
+      "step": 4913
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7887893915176392,
+      "learning_rate": 1.685510654966864e-05,
+      "loss": 1.0948,
+      "step": 4914
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7676926851272583,
+      "learning_rate": 1.6853753973915832e-05,
+      "loss": 1.0488,
+      "step": 4915
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7640408277511597,
+      "learning_rate": 1.6852401161659122e-05,
+      "loss": 0.9792,
+      "step": 4916
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7754625082015991,
+      "learning_rate": 1.685104811294519e-05,
+      "loss": 1.0557,
+      "step": 4917
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.764940619468689,
+      "learning_rate": 1.6849694827820718e-05,
+      "loss": 1.0284,
+      "step": 4918
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7750033140182495,
+      "learning_rate": 1.6848341306332418e-05,
+      "loss": 1.0486,
+      "step": 4919
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8623255491256714,
+      "learning_rate": 1.6846987548526987e-05,
+      "loss": 0.9688,
+      "step": 4920
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7640178203582764,
+      "learning_rate": 1.6845633554451142e-05,
+      "loss": 1.0442,
+      "step": 4921
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8016480207443237,
+      "learning_rate": 1.6844279324151608e-05,
+      "loss": 0.9812,
+      "step": 4922
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9299976825714111,
+      "learning_rate": 1.684292485767511e-05,
+      "loss": 0.9732,
+      "step": 4923
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8280783891677856,
+      "learning_rate": 1.684157015506839e-05,
+      "loss": 0.9375,
+      "step": 4924
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8758163452148438,
+      "learning_rate": 1.6840215216378198e-05,
+      "loss": 0.9807,
+      "step": 4925
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.743076205253601,
+      "learning_rate": 1.6838860041651286e-05,
+      "loss": 0.9689,
+      "step": 4926
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9167157411575317,
+      "learning_rate": 1.6837504630934412e-05,
+      "loss": 0.9753,
+      "step": 4927
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8990880250930786,
+      "learning_rate": 1.683614898427436e-05,
+      "loss": 0.9734,
+      "step": 4928
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7996269464492798,
+      "learning_rate": 1.6834793101717897e-05,
+      "loss": 1.001,
+      "step": 4929
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.061493396759033,
+      "learning_rate": 1.6833436983311823e-05,
+      "loss": 1.0002,
+      "step": 4930
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6283007860183716,
+      "learning_rate": 1.6832080629102923e-05,
+      "loss": 1.0448,
+      "step": 4931
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7978817224502563,
+      "learning_rate": 1.6830724039138005e-05,
+      "loss": 1.0868,
+      "step": 4932
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8094483613967896,
+      "learning_rate": 1.682936721346388e-05,
+      "loss": 0.9893,
+      "step": 4933
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.836949348449707,
+      "learning_rate": 1.682801015212737e-05,
+      "loss": 0.9779,
+      "step": 4934
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1489976644515991,
+      "learning_rate": 1.6826652855175304e-05,
+      "loss": 0.5886,
+      "step": 4935
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0390729904174805,
+      "learning_rate": 1.6825295322654517e-05,
+      "loss": 1.0726,
+      "step": 4936
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9677833318710327,
+      "learning_rate": 1.6823937554611856e-05,
+      "loss": 1.1067,
+      "step": 4937
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.645825743675232,
+      "learning_rate": 1.682257955109417e-05,
+      "loss": 1.0226,
+      "step": 4938
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.138099193572998,
+      "learning_rate": 1.6821221312148322e-05,
+      "loss": 0.6241,
+      "step": 4939
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7123996019363403,
+      "learning_rate": 1.681986283782118e-05,
+      "loss": 1.056,
+      "step": 4940
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8262661695480347,
+      "learning_rate": 1.6818504128159628e-05,
+      "loss": 1.1014,
+      "step": 4941
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.103005051612854,
+      "learning_rate": 1.6817145183210538e-05,
+      "loss": 0.6455,
+      "step": 4942
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9934080243110657,
+      "learning_rate": 1.6815786003020812e-05,
+      "loss": 0.5659,
+      "step": 4943
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.700119137763977,
+      "learning_rate": 1.6814426587637354e-05,
+      "loss": 1.0563,
+      "step": 4944
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7847425937652588,
+      "learning_rate": 1.6813066937107065e-05,
+      "loss": 1.0749,
+      "step": 4945
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.731642723083496,
+      "learning_rate": 1.6811707051476868e-05,
+      "loss": 1.0615,
+      "step": 4946
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9088008403778076,
+      "learning_rate": 1.681034693079369e-05,
+      "loss": 0.9486,
+      "step": 4947
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9439805746078491,
+      "learning_rate": 1.6808986575104464e-05,
+      "loss": 1.0261,
+      "step": 4948
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8109700679779053,
+      "learning_rate": 1.680762598445613e-05,
+      "loss": 0.9845,
+      "step": 4949
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.028010845184326,
+      "learning_rate": 1.6806265158895642e-05,
+      "loss": 1.0151,
+      "step": 4950
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1376056671142578,
+      "learning_rate": 1.6804904098469955e-05,
+      "loss": 0.6496,
+      "step": 4951
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9380031824111938,
+      "learning_rate": 1.6803542803226034e-05,
+      "loss": 1.0977,
+      "step": 4952
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9276891946792603,
+      "learning_rate": 1.6802181273210858e-05,
+      "loss": 1.0836,
+      "step": 4953
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8618677854537964,
+      "learning_rate": 1.6800819508471407e-05,
+      "loss": 1.0399,
+      "step": 4954
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7445107698440552,
+      "learning_rate": 1.679945750905467e-05,
+      "loss": 1.1631,
+      "step": 4955
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7674551010131836,
+      "learning_rate": 1.679809527500765e-05,
+      "loss": 1.0302,
+      "step": 4956
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.70111083984375,
+      "learning_rate": 1.679673280637735e-05,
+      "loss": 0.9669,
+      "step": 4957
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9820137023925781,
+      "learning_rate": 1.679537010321079e-05,
+      "loss": 1.0471,
+      "step": 4958
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8192939758300781,
+      "learning_rate": 1.679400716555499e-05,
+      "loss": 1.0805,
+      "step": 4959
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.147865653038025,
+      "learning_rate": 1.6792643993456978e-05,
+      "loss": 0.646,
+      "step": 4960
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6857572793960571,
+      "learning_rate": 1.6791280586963798e-05,
+      "loss": 1.0064,
+      "step": 4961
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.774776577949524,
+      "learning_rate": 1.6789916946122494e-05,
+      "loss": 1.0044,
+      "step": 4962
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9410014152526855,
+      "learning_rate": 1.6788553070980126e-05,
+      "loss": 0.9876,
+      "step": 4963
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7348839044570923,
+      "learning_rate": 1.678718896158375e-05,
+      "loss": 0.9517,
+      "step": 4964
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6816319227218628,
+      "learning_rate": 1.6785824617980446e-05,
+      "loss": 1.0669,
+      "step": 4965
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7540005445480347,
+      "learning_rate": 1.6784460040217286e-05,
+      "loss": 1.0939,
+      "step": 4966
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8014140129089355,
+      "learning_rate": 1.6783095228341365e-05,
+      "loss": 0.9779,
+      "step": 4967
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8831634521484375,
+      "learning_rate": 1.6781730182399774e-05,
+      "loss": 1.0254,
+      "step": 4968
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8566021919250488,
+      "learning_rate": 1.678036490243962e-05,
+      "loss": 1.0221,
+      "step": 4969
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9947357177734375,
+      "learning_rate": 1.6778999388508013e-05,
+      "loss": 1.0193,
+      "step": 4970
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.705660104751587,
+      "learning_rate": 1.6777633640652072e-05,
+      "loss": 1.0313,
+      "step": 4971
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9963876008987427,
+      "learning_rate": 1.6776267658918927e-05,
+      "loss": 1.0854,
+      "step": 4972
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.702654242515564,
+      "learning_rate": 1.6774901443355717e-05,
+      "loss": 1.0393,
+      "step": 4973
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.784767746925354,
+      "learning_rate": 1.677353499400958e-05,
+      "loss": 1.0797,
+      "step": 4974
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0459593534469604,
+      "learning_rate": 1.6772168310927673e-05,
+      "loss": 0.6739,
+      "step": 4975
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.039383053779602,
+      "learning_rate": 1.677080139415715e-05,
+      "loss": 0.6218,
+      "step": 4976
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7655045986175537,
+      "learning_rate": 1.676943424374519e-05,
+      "loss": 1.0503,
+      "step": 4977
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.754356861114502,
+      "learning_rate": 1.6768066859738963e-05,
+      "loss": 0.9762,
+      "step": 4978
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.992211937904358,
+      "learning_rate": 1.6766699242185653e-05,
+      "loss": 1.0499,
+      "step": 4979
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7898869514465332,
+      "learning_rate": 1.6765331391132454e-05,
+      "loss": 0.9944,
+      "step": 4980
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0913188457489014,
+      "learning_rate": 1.676396330662657e-05,
+      "loss": 1.1137,
+      "step": 4981
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8172208070755005,
+      "learning_rate": 1.6762594988715204e-05,
+      "loss": 1.0002,
+      "step": 4982
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8199137449264526,
+      "learning_rate": 1.6761226437445577e-05,
+      "loss": 1.0955,
+      "step": 4983
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.869014859199524,
+      "learning_rate": 1.675985765286491e-05,
+      "loss": 1.0695,
+      "step": 4984
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9903357028961182,
+      "learning_rate": 1.675848863502044e-05,
+      "loss": 1.1009,
+      "step": 4985
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8338927030563354,
+      "learning_rate": 1.6757119383959406e-05,
+      "loss": 1.0384,
+      "step": 4986
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7249780893325806,
+      "learning_rate": 1.6755749899729056e-05,
+      "loss": 0.9809,
+      "step": 4987
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.76215398311615,
+      "learning_rate": 1.675438018237665e-05,
+      "loss": 0.9627,
+      "step": 4988
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9811723232269287,
+      "learning_rate": 1.675301023194945e-05,
+      "loss": 1.0637,
+      "step": 4989
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.747698187828064,
+      "learning_rate": 1.6751640048494734e-05,
+      "loss": 0.9672,
+      "step": 4990
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1844778060913086,
+      "learning_rate": 1.6750269632059776e-05,
+      "loss": 1.117,
+      "step": 4991
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.948578953742981,
+      "learning_rate": 1.674889898269187e-05,
+      "loss": 0.9905,
+      "step": 4992
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6798242330551147,
+      "learning_rate": 1.6747528100438316e-05,
+      "loss": 1.026,
+      "step": 4993
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8882676362991333,
+      "learning_rate": 1.6746156985346413e-05,
+      "loss": 1.1141,
+      "step": 4994
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.584547758102417,
+      "learning_rate": 1.6744785637463476e-05,
+      "loss": 0.9406,
+      "step": 4995
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.051281452178955,
+      "learning_rate": 1.6743414056836827e-05,
+      "loss": 0.9816,
+      "step": 4996
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.90603768825531,
+      "learning_rate": 1.67420422435138e-05,
+      "loss": 0.9714,
+      "step": 4997
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7681667804718018,
+      "learning_rate": 1.6740670197541722e-05,
+      "loss": 1.1069,
+      "step": 4998
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.075596570968628,
+      "learning_rate": 1.6739297918967948e-05,
+      "loss": 0.9976,
+      "step": 4999
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8993091583251953,
+      "learning_rate": 1.6737925407839828e-05,
+      "loss": 1.0662,
+      "step": 5000
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6743183135986328,
+      "learning_rate": 1.6736552664204725e-05,
+      "loss": 0.9614,
+      "step": 5001
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8447792530059814,
+      "learning_rate": 1.6735179688110004e-05,
+      "loss": 1.0693,
+      "step": 5002
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6768145561218262,
+      "learning_rate": 1.673380647960305e-05,
+      "loss": 0.9933,
+      "step": 5003
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7443569898605347,
+      "learning_rate": 1.6732433038731245e-05,
+      "loss": 0.9433,
+      "step": 5004
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8673535585403442,
+      "learning_rate": 1.673105936554198e-05,
+      "loss": 1.0735,
+      "step": 5005
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9199897050857544,
+      "learning_rate": 1.6729685460082658e-05,
+      "loss": 1.0182,
+      "step": 5006
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6692949533462524,
+      "learning_rate": 1.6728311322400693e-05,
+      "loss": 0.943,
+      "step": 5007
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.002382516860962,
+      "learning_rate": 1.6726936952543494e-05,
+      "loss": 1.0232,
+      "step": 5008
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.011814832687378,
+      "learning_rate": 1.6725562350558494e-05,
+      "loss": 1.0256,
+      "step": 5009
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.85011625289917,
+      "learning_rate": 1.6724187516493125e-05,
+      "loss": 0.9765,
+      "step": 5010
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9433430433273315,
+      "learning_rate": 1.6722812450394826e-05,
+      "loss": 1.0563,
+      "step": 5011
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2836203575134277,
+      "learning_rate": 1.6721437152311052e-05,
+      "loss": 0.6912,
+      "step": 5012
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.831094741821289,
+      "learning_rate": 1.6720061622289258e-05,
+      "loss": 1.0179,
+      "step": 5013
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.651019811630249,
+      "learning_rate": 1.6718685860376903e-05,
+      "loss": 1.0214,
+      "step": 5014
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7902356386184692,
+      "learning_rate": 1.6717309866621473e-05,
+      "loss": 1.0352,
+      "step": 5015
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7322484254837036,
+      "learning_rate": 1.6715933641070443e-05,
+      "loss": 1.0983,
+      "step": 5016
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7539037466049194,
+      "learning_rate": 1.67145571837713e-05,
+      "loss": 1.0511,
+      "step": 5017
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.787750005722046,
+      "learning_rate": 1.6713180494771545e-05,
+      "loss": 0.9072,
+      "step": 5018
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8441230058670044,
+      "learning_rate": 1.6711803574118687e-05,
+      "loss": 1.0385,
+      "step": 5019
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9119000434875488,
+      "learning_rate": 1.6710426421860236e-05,
+      "loss": 0.9924,
+      "step": 5020
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8587931394577026,
+      "learning_rate": 1.670904903804371e-05,
+      "loss": 0.9626,
+      "step": 5021
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7519148588180542,
+      "learning_rate": 1.6707671422716644e-05,
+      "loss": 1.0367,
+      "step": 5022
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.089341402053833,
+      "learning_rate": 1.670629357592658e-05,
+      "loss": 1.0558,
+      "step": 5023
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2060391902923584,
+      "learning_rate": 1.670491549772105e-05,
+      "loss": 0.6475,
+      "step": 5024
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7893842458724976,
+      "learning_rate": 1.6703537188147622e-05,
+      "loss": 0.9636,
+      "step": 5025
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8013216257095337,
+      "learning_rate": 1.6702158647253846e-05,
+      "loss": 1.0143,
+      "step": 5026
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8165167570114136,
+      "learning_rate": 1.6700779875087302e-05,
+      "loss": 1.0265,
+      "step": 5027
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7995188236236572,
+      "learning_rate": 1.6699400871695556e-05,
+      "loss": 1.043,
+      "step": 5028
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6800528764724731,
+      "learning_rate": 1.66980216371262e-05,
+      "loss": 0.9996,
+      "step": 5029
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8691574335098267,
+      "learning_rate": 1.6696642171426834e-05,
+      "loss": 0.9599,
+      "step": 5030
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9034206867218018,
+      "learning_rate": 1.669526247464505e-05,
+      "loss": 1.0361,
+      "step": 5031
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7886236906051636,
+      "learning_rate": 1.6693882546828462e-05,
+      "loss": 0.9895,
+      "step": 5032
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8249578475952148,
+      "learning_rate": 1.6692502388024684e-05,
+      "loss": 1.0423,
+      "step": 5033
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8827708959579468,
+      "learning_rate": 1.6691121998281343e-05,
+      "loss": 1.0048,
+      "step": 5034
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9637030363082886,
+      "learning_rate": 1.6689741377646075e-05,
+      "loss": 1.1184,
+      "step": 5035
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.108392357826233,
+      "learning_rate": 1.6688360526166514e-05,
+      "loss": 0.5615,
+      "step": 5036
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9205940961837769,
+      "learning_rate": 1.668697944389032e-05,
+      "loss": 1.0902,
+      "step": 5037
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9189637899398804,
+      "learning_rate": 1.6685598130865143e-05,
+      "loss": 0.9799,
+      "step": 5038
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.967287302017212,
+      "learning_rate": 1.6684216587138647e-05,
+      "loss": 1.0766,
+      "step": 5039
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9516175985336304,
+      "learning_rate": 1.668283481275851e-05,
+      "loss": 0.9908,
+      "step": 5040
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8366619348526,
+      "learning_rate": 1.6681452807772413e-05,
+      "loss": 1.0563,
+      "step": 5041
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0738985538482666,
+      "learning_rate": 1.6680070572228043e-05,
+      "loss": 0.6312,
+      "step": 5042
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.040541410446167,
+      "learning_rate": 1.6678688106173097e-05,
+      "loss": 1.0372,
+      "step": 5043
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9387565851211548,
+      "learning_rate": 1.667730540965528e-05,
+      "loss": 1.0402,
+      "step": 5044
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.158446788787842,
+      "learning_rate": 1.667592248272231e-05,
+      "loss": 1.0442,
+      "step": 5045
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1286211013793945,
+      "learning_rate": 1.6674539325421897e-05,
+      "loss": 0.6194,
+      "step": 5046
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0675065517425537,
+      "learning_rate": 1.667315593780178e-05,
+      "loss": 1.079,
+      "step": 5047
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8281991481781006,
+      "learning_rate": 1.6671772319909692e-05,
+      "loss": 0.9893,
+      "step": 5048
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7437891960144043,
+      "learning_rate": 1.6670388471793377e-05,
+      "loss": 0.9611,
+      "step": 5049
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8292734622955322,
+      "learning_rate": 1.666900439350059e-05,
+      "loss": 1.0034,
+      "step": 5050
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0750789642333984,
+      "learning_rate": 1.666762008507909e-05,
+      "loss": 1.0279,
+      "step": 5051
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1034513711929321,
+      "learning_rate": 1.6666235546576648e-05,
+      "loss": 0.6067,
+      "step": 5052
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9065146446228027,
+      "learning_rate": 1.6664850778041036e-05,
+      "loss": 1.0109,
+      "step": 5053
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.070749044418335,
+      "learning_rate": 1.6663465779520042e-05,
+      "loss": 0.948,
+      "step": 5054
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7005939483642578,
+      "learning_rate": 1.6662080551061458e-05,
+      "loss": 1.0125,
+      "step": 5055
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9674650430679321,
+      "learning_rate": 1.6660695092713083e-05,
+      "loss": 0.9087,
+      "step": 5056
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8961045742034912,
+      "learning_rate": 1.6659309404522725e-05,
+      "loss": 1.0591,
+      "step": 5057
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9492416381835938,
+      "learning_rate": 1.6657923486538203e-05,
+      "loss": 1.1025,
+      "step": 5058
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9426721334457397,
+      "learning_rate": 1.665653733880734e-05,
+      "loss": 1.059,
+      "step": 5059
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9625918865203857,
+      "learning_rate": 1.665515096137797e-05,
+      "loss": 1.0059,
+      "step": 5060
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9408493041992188,
+      "learning_rate": 1.665376435429793e-05,
+      "loss": 1.0875,
+      "step": 5061
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.980059266090393,
+      "learning_rate": 1.6652377517615065e-05,
+      "loss": 1.1142,
+      "step": 5062
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8643765449523926,
+      "learning_rate": 1.6650990451377237e-05,
+      "loss": 1.0869,
+      "step": 5063
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7100104093551636,
+      "learning_rate": 1.6649603155632305e-05,
+      "loss": 1.0375,
+      "step": 5064
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2048871517181396,
+      "learning_rate": 1.6648215630428146e-05,
+      "loss": 0.9923,
+      "step": 5065
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9793179035186768,
+      "learning_rate": 1.6646827875812635e-05,
+      "loss": 1.0599,
+      "step": 5066
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.914188027381897,
+      "learning_rate": 1.664543989183366e-05,
+      "loss": 1.0172,
+      "step": 5067
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9987032413482666,
+      "learning_rate": 1.6644051678539122e-05,
+      "loss": 0.963,
+      "step": 5068
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7849905490875244,
+      "learning_rate": 1.6642663235976916e-05,
+      "loss": 1.0345,
+      "step": 5069
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9672876596450806,
+      "learning_rate": 1.6641274564194956e-05,
+      "loss": 1.0449,
+      "step": 5070
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8944051265716553,
+      "learning_rate": 1.663988566324117e-05,
+      "loss": 0.9923,
+      "step": 5071
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7426607608795166,
+      "learning_rate": 1.663849653316347e-05,
+      "loss": 0.9722,
+      "step": 5072
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6451948881149292,
+      "learning_rate": 1.66371071740098e-05,
+      "loss": 0.9605,
+      "step": 5073
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1996660232543945,
+      "learning_rate": 1.6635717585828102e-05,
+      "loss": 0.9767,
+      "step": 5074
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9899603128433228,
+      "learning_rate": 1.6634327768666328e-05,
+      "loss": 1.0257,
+      "step": 5075
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.007981538772583,
+      "learning_rate": 1.6632937722572436e-05,
+      "loss": 1.0709,
+      "step": 5076
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.003831386566162,
+      "learning_rate": 1.6631547447594388e-05,
+      "loss": 0.9635,
+      "step": 5077
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8250536918640137,
+      "learning_rate": 1.663015694378016e-05,
+      "loss": 1.0368,
+      "step": 5078
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9970078468322754,
+      "learning_rate": 1.6628766211177744e-05,
+      "loss": 0.988,
+      "step": 5079
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8606919050216675,
+      "learning_rate": 1.6627375249835117e-05,
+      "loss": 1.0887,
+      "step": 5080
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.050845146179199,
+      "learning_rate": 1.6625984059800285e-05,
+      "loss": 1.0635,
+      "step": 5081
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6655102968215942,
+      "learning_rate": 1.6624592641121252e-05,
+      "loss": 1.055,
+      "step": 5082
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8206238746643066,
+      "learning_rate": 1.662320099384603e-05,
+      "loss": 1.1186,
+      "step": 5083
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9102427959442139,
+      "learning_rate": 1.6621809118022646e-05,
+      "loss": 1.0264,
+      "step": 5084
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8041369915008545,
+      "learning_rate": 1.6620417013699122e-05,
+      "loss": 1.0616,
+      "step": 5085
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8059991598129272,
+      "learning_rate": 1.6619024680923505e-05,
+      "loss": 0.9807,
+      "step": 5086
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7358617782592773,
+      "learning_rate": 1.6617632119743837e-05,
+      "loss": 1.021,
+      "step": 5087
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8036466836929321,
+      "learning_rate": 1.6616239330208163e-05,
+      "loss": 0.9799,
+      "step": 5088
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.068758487701416,
+      "learning_rate": 1.661484631236456e-05,
+      "loss": 1.1062,
+      "step": 5089
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.035245656967163,
+      "learning_rate": 1.661345306626108e-05,
+      "loss": 1.0171,
+      "step": 5090
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0842854976654053,
+      "learning_rate": 1.6612059591945815e-05,
+      "loss": 1.0239,
+      "step": 5091
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8434911966323853,
+      "learning_rate": 1.661066588946684e-05,
+      "loss": 1.0085,
+      "step": 5092
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8452248573303223,
+      "learning_rate": 1.660927195887225e-05,
+      "loss": 0.9346,
+      "step": 5093
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8949992656707764,
+      "learning_rate": 1.6607877800210156e-05,
+      "loss": 1.0137,
+      "step": 5094
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9097779989242554,
+      "learning_rate": 1.660648341352865e-05,
+      "loss": 1.083,
+      "step": 5095
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6986191272735596,
+      "learning_rate": 1.6605088798875856e-05,
+      "loss": 1.0189,
+      "step": 5096
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8721935749053955,
+      "learning_rate": 1.66036939562999e-05,
+      "loss": 1.0203,
+      "step": 5097
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.862821102142334,
+      "learning_rate": 1.660229888584891e-05,
+      "loss": 1.0707,
+      "step": 5098
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.097388744354248,
+      "learning_rate": 1.6600903587571028e-05,
+      "loss": 1.0541,
+      "step": 5099
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9460285902023315,
+      "learning_rate": 1.6599508061514404e-05,
+      "loss": 0.9935,
+      "step": 5100
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9337083101272583,
+      "learning_rate": 1.659811230772719e-05,
+      "loss": 1.0208,
+      "step": 5101
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7883825302124023,
+      "learning_rate": 1.6596716326257552e-05,
+      "loss": 1.001,
+      "step": 5102
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8423614501953125,
+      "learning_rate": 1.6595320117153664e-05,
+      "loss": 1.0452,
+      "step": 5103
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9980061054229736,
+      "learning_rate": 1.6593923680463698e-05,
+      "loss": 1.0483,
+      "step": 5104
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8504562377929688,
+      "learning_rate": 1.6592527016235848e-05,
+      "loss": 1.0736,
+      "step": 5105
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8085793256759644,
+      "learning_rate": 1.6591130124518305e-05,
+      "loss": 0.9559,
+      "step": 5106
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8882025480270386,
+      "learning_rate": 1.6589733005359274e-05,
+      "loss": 0.9677,
+      "step": 5107
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.173730492591858,
+      "learning_rate": 1.6588335658806964e-05,
+      "loss": 0.6819,
+      "step": 5108
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.008368730545044,
+      "learning_rate": 1.658693808490959e-05,
+      "loss": 0.9706,
+      "step": 5109
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2888143062591553,
+      "learning_rate": 1.658554028371539e-05,
+      "loss": 1.0056,
+      "step": 5110
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8251432180404663,
+      "learning_rate": 1.6584142255272587e-05,
+      "loss": 0.9935,
+      "step": 5111
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8009058237075806,
+      "learning_rate": 1.6582743999629426e-05,
+      "loss": 0.9475,
+      "step": 5112
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.013921022415161,
+      "learning_rate": 1.6581345516834158e-05,
+      "loss": 1.057,
+      "step": 5113
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7876862287521362,
+      "learning_rate": 1.657994680693504e-05,
+      "loss": 1.0784,
+      "step": 5114
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7720376253128052,
+      "learning_rate": 1.657854786998034e-05,
+      "loss": 1.0035,
+      "step": 5115
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6997485160827637,
+      "learning_rate": 1.657714870601833e-05,
+      "loss": 1.0185,
+      "step": 5116
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1311860084533691,
+      "learning_rate": 1.657574931509729e-05,
+      "loss": 0.6178,
+      "step": 5117
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0236198902130127,
+      "learning_rate": 1.6574349697265507e-05,
+      "loss": 0.6048,
+      "step": 5118
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0134334564208984,
+      "learning_rate": 1.657294985257128e-05,
+      "loss": 1.0408,
+      "step": 5119
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.226640224456787,
+      "learning_rate": 1.6571549781062917e-05,
+      "loss": 0.6493,
+      "step": 5120
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.914505958557129,
+      "learning_rate": 1.6570149482788732e-05,
+      "loss": 0.9974,
+      "step": 5121
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8020858764648438,
+      "learning_rate": 1.6568748957797038e-05,
+      "loss": 1.009,
+      "step": 5122
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7695931196212769,
+      "learning_rate": 1.6567348206136165e-05,
+      "loss": 0.9871,
+      "step": 5123
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8216640949249268,
+      "learning_rate": 1.656594722785445e-05,
+      "loss": 1.0726,
+      "step": 5124
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.829126238822937,
+      "learning_rate": 1.6564546023000237e-05,
+      "loss": 1.0043,
+      "step": 5125
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7772107124328613,
+      "learning_rate": 1.656314459162188e-05,
+      "loss": 0.9755,
+      "step": 5126
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9471821784973145,
+      "learning_rate": 1.6561742933767738e-05,
+      "loss": 1.1311,
+      "step": 5127
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7489556074142456,
+      "learning_rate": 1.6560341049486176e-05,
+      "loss": 1.0658,
+      "step": 5128
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7661609649658203,
+      "learning_rate": 1.6558938938825568e-05,
+      "loss": 0.9817,
+      "step": 5129
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6405469179153442,
+      "learning_rate": 1.65575366018343e-05,
+      "loss": 1.0084,
+      "step": 5130
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.707180380821228,
+      "learning_rate": 1.655613403856076e-05,
+      "loss": 1.0275,
+      "step": 5131
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7378404140472412,
+      "learning_rate": 1.6554731249053352e-05,
+      "loss": 1.0347,
+      "step": 5132
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.713561773300171,
+      "learning_rate": 1.6553328233360477e-05,
+      "loss": 0.9501,
+      "step": 5133
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6333541870117188,
+      "learning_rate": 1.655192499153055e-05,
+      "loss": 1.0285,
+      "step": 5134
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7564549446105957,
+      "learning_rate": 1.655052152361199e-05,
+      "loss": 1.0502,
+      "step": 5135
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.903883695602417,
+      "learning_rate": 1.6549117829653238e-05,
+      "loss": 0.9278,
+      "step": 5136
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6957582235336304,
+      "learning_rate": 1.6547713909702716e-05,
+      "loss": 0.9768,
+      "step": 5137
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.824670672416687,
+      "learning_rate": 1.6546309763808883e-05,
+      "loss": 0.9913,
+      "step": 5138
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.724562406539917,
+      "learning_rate": 1.6544905392020182e-05,
+      "loss": 1.0095,
+      "step": 5139
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7832239866256714,
+      "learning_rate": 1.6543500794385084e-05,
+      "loss": 1.0758,
+      "step": 5140
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7708734273910522,
+      "learning_rate": 1.6542095970952046e-05,
+      "loss": 0.9918,
+      "step": 5141
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0592641830444336,
+      "learning_rate": 1.6540690921769556e-05,
+      "loss": 0.9831,
+      "step": 5142
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9066232442855835,
+      "learning_rate": 1.653928564688609e-05,
+      "loss": 1.0807,
+      "step": 5143
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8743990659713745,
+      "learning_rate": 1.6537880146350144e-05,
+      "loss": 1.053,
+      "step": 5144
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8483422994613647,
+      "learning_rate": 1.6536474420210215e-05,
+      "loss": 1.0806,
+      "step": 5145
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8380682468414307,
+      "learning_rate": 1.6535068468514817e-05,
+      "loss": 1.0396,
+      "step": 5146
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8999302387237549,
+      "learning_rate": 1.653366229131246e-05,
+      "loss": 1.0137,
+      "step": 5147
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.827161192893982,
+      "learning_rate": 1.6532255888651665e-05,
+      "loss": 0.9832,
+      "step": 5148
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8261295557022095,
+      "learning_rate": 1.6530849260580967e-05,
+      "loss": 1.0127,
+      "step": 5149
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6067414283752441,
+      "learning_rate": 1.6529442407148907e-05,
+      "loss": 0.9998,
+      "step": 5150
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8082988262176514,
+      "learning_rate": 1.6528035328404026e-05,
+      "loss": 1.064,
+      "step": 5151
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8491404056549072,
+      "learning_rate": 1.6526628024394883e-05,
+      "loss": 1.0527,
+      "step": 5152
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.740983247756958,
+      "learning_rate": 1.6525220495170037e-05,
+      "loss": 1.1291,
+      "step": 5153
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7249277830123901,
+      "learning_rate": 1.652381274077806e-05,
+      "loss": 1.0145,
+      "step": 5154
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8582860231399536,
+      "learning_rate": 1.652240476126753e-05,
+      "loss": 0.9817,
+      "step": 5155
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7936407327651978,
+      "learning_rate": 1.6520996556687026e-05,
+      "loss": 0.9869,
+      "step": 5156
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7154349088668823,
+      "learning_rate": 1.6519588127085155e-05,
+      "loss": 1.0545,
+      "step": 5157
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.817953109741211,
+      "learning_rate": 1.6518179472510506e-05,
+      "loss": 1.0644,
+      "step": 5158
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0917179584503174,
+      "learning_rate": 1.651677059301169e-05,
+      "loss": 1.0515,
+      "step": 5159
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9930673837661743,
+      "learning_rate": 1.6515361488637323e-05,
+      "loss": 0.9829,
+      "step": 5160
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.724372148513794,
+      "learning_rate": 1.6513952159436033e-05,
+      "loss": 0.982,
+      "step": 5161
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8292368650436401,
+      "learning_rate": 1.651254260545645e-05,
+      "loss": 1.1008,
+      "step": 5162
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8004170656204224,
+      "learning_rate": 1.6511132826747212e-05,
+      "loss": 1.0003,
+      "step": 5163
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8540232181549072,
+      "learning_rate": 1.650972282335697e-05,
+      "loss": 1.0383,
+      "step": 5164
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.018101215362549,
+      "learning_rate": 1.6508312595334378e-05,
+      "loss": 1.0866,
+      "step": 5165
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.314603090286255,
+      "learning_rate": 1.65069021427281e-05,
+      "loss": 1.0029,
+      "step": 5166
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.185894250869751,
+      "learning_rate": 1.65054914655868e-05,
+      "loss": 0.5862,
+      "step": 5167
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7470319271087646,
+      "learning_rate": 1.6504080563959167e-05,
+      "loss": 1.0095,
+      "step": 5168
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7473456859588623,
+      "learning_rate": 1.6502669437893878e-05,
+      "loss": 1.1083,
+      "step": 5169
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9529571533203125,
+      "learning_rate": 1.6501258087439637e-05,
+      "loss": 0.9961,
+      "step": 5170
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7644704580307007,
+      "learning_rate": 1.6499846512645136e-05,
+      "loss": 1.0215,
+      "step": 5171
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8650532960891724,
+      "learning_rate": 1.649843471355909e-05,
+      "loss": 1.0283,
+      "step": 5172
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7721192836761475,
+      "learning_rate": 1.649702269023021e-05,
+      "loss": 1.0563,
+      "step": 5173
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7326829433441162,
+      "learning_rate": 1.649561044270723e-05,
+      "loss": 1.0945,
+      "step": 5174
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8108159303665161,
+      "learning_rate": 1.6494197971038876e-05,
+      "loss": 0.9759,
+      "step": 5175
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7008742094039917,
+      "learning_rate": 1.649278527527389e-05,
+      "loss": 1.0745,
+      "step": 5176
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5801929235458374,
+      "learning_rate": 1.6491372355461028e-05,
+      "loss": 0.9273,
+      "step": 5177
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6170562505722046,
+      "learning_rate": 1.6489959211649035e-05,
+      "loss": 1.0075,
+      "step": 5178
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0312623977661133,
+      "learning_rate": 1.6488545843886677e-05,
+      "loss": 0.9267,
+      "step": 5179
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9025565385818481,
+      "learning_rate": 1.648713225222273e-05,
+      "loss": 0.9545,
+      "step": 5180
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9409915208816528,
+      "learning_rate": 1.6485718436705965e-05,
+      "loss": 1.0998,
+      "step": 5181
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8895087242126465,
+      "learning_rate": 1.6484304397385177e-05,
+      "loss": 1.0841,
+      "step": 5182
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9464365243911743,
+      "learning_rate": 1.6482890134309156e-05,
+      "loss": 1.0579,
+      "step": 5183
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8552864789962769,
+      "learning_rate": 1.648147564752671e-05,
+      "loss": 1.0387,
+      "step": 5184
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9646834135055542,
+      "learning_rate": 1.648006093708664e-05,
+      "loss": 0.9989,
+      "step": 5185
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.746027946472168,
+      "learning_rate": 1.647864600303777e-05,
+      "loss": 1.0073,
+      "step": 5186
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8757826089859009,
+      "learning_rate": 1.6477230845428925e-05,
+      "loss": 0.9697,
+      "step": 5187
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1826835870742798,
+      "learning_rate": 1.6475815464308933e-05,
+      "loss": 0.6374,
+      "step": 5188
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7431520223617554,
+      "learning_rate": 1.6474399859726644e-05,
+      "loss": 1.0647,
+      "step": 5189
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8256975412368774,
+      "learning_rate": 1.64729840317309e-05,
+      "loss": 1.0547,
+      "step": 5190
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0221009254455566,
+      "learning_rate": 1.6471567980370556e-05,
+      "loss": 1.1086,
+      "step": 5191
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.819825291633606,
+      "learning_rate": 1.6470151705694478e-05,
+      "loss": 0.9987,
+      "step": 5192
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.950721025466919,
+      "learning_rate": 1.646873520775154e-05,
+      "loss": 1.0497,
+      "step": 5193
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.791812777519226,
+      "learning_rate": 1.6467318486590623e-05,
+      "loss": 1.0024,
+      "step": 5194
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7883446216583252,
+      "learning_rate": 1.6465901542260607e-05,
+      "loss": 0.9573,
+      "step": 5195
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8049689531326294,
+      "learning_rate": 1.646448437481039e-05,
+      "loss": 1.0273,
+      "step": 5196
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9341366291046143,
+      "learning_rate": 1.646306698428888e-05,
+      "loss": 1.0573,
+      "step": 5197
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6553109884262085,
+      "learning_rate": 1.6461649370744975e-05,
+      "loss": 0.9983,
+      "step": 5198
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.001251459121704,
+      "learning_rate": 1.6460231534227603e-05,
+      "loss": 1.0859,
+      "step": 5199
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8395774364471436,
+      "learning_rate": 1.6458813474785685e-05,
+      "loss": 0.9716,
+      "step": 5200
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6906391382217407,
+      "learning_rate": 1.6457395192468158e-05,
+      "loss": 0.9144,
+      "step": 5201
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.063892364501953,
+      "learning_rate": 1.645597668732396e-05,
+      "loss": 1.0668,
+      "step": 5202
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8227604627609253,
+      "learning_rate": 1.6454557959402048e-05,
+      "loss": 1.0833,
+      "step": 5203
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9198225736618042,
+      "learning_rate": 1.645313900875136e-05,
+      "loss": 0.9916,
+      "step": 5204
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8252094984054565,
+      "learning_rate": 1.645171983542088e-05,
+      "loss": 1.0474,
+      "step": 5205
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7386993169784546,
+      "learning_rate": 1.6450300439459562e-05,
+      "loss": 1.0368,
+      "step": 5206
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.783159613609314,
+      "learning_rate": 1.64488808209164e-05,
+      "loss": 1.0037,
+      "step": 5207
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6322764158248901,
+      "learning_rate": 1.6447460979840373e-05,
+      "loss": 0.9427,
+      "step": 5208
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.839501142501831,
+      "learning_rate": 1.644604091628048e-05,
+      "loss": 1.0476,
+      "step": 5209
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9075543880462646,
+      "learning_rate": 1.6444620630285717e-05,
+      "loss": 1.084,
+      "step": 5210
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8340729475021362,
+      "learning_rate": 1.64432001219051e-05,
+      "loss": 1.0892,
+      "step": 5211
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9423549175262451,
+      "learning_rate": 1.6441779391187647e-05,
+      "loss": 1.0034,
+      "step": 5212
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7347936630249023,
+      "learning_rate": 1.6440358438182383e-05,
+      "loss": 0.9915,
+      "step": 5213
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8718029260635376,
+      "learning_rate": 1.6438937262938336e-05,
+      "loss": 1.0649,
+      "step": 5214
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8925142288208008,
+      "learning_rate": 1.643751586550455e-05,
+      "loss": 0.9537,
+      "step": 5215
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.822401523590088,
+      "learning_rate": 1.6436094245930077e-05,
+      "loss": 1.0798,
+      "step": 5216
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9212912321090698,
+      "learning_rate": 1.643467240426397e-05,
+      "loss": 0.9887,
+      "step": 5217
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.145285725593567,
+      "learning_rate": 1.6433250340555292e-05,
+      "loss": 0.6171,
+      "step": 5218
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8513644933700562,
+      "learning_rate": 1.6431828054853112e-05,
+      "loss": 1.028,
+      "step": 5219
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7444010972976685,
+      "learning_rate": 1.6430405547206518e-05,
+      "loss": 1.1106,
+      "step": 5220
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7501215934753418,
+      "learning_rate": 1.6428982817664586e-05,
+      "loss": 0.9859,
+      "step": 5221
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.007359266281128,
+      "learning_rate": 1.642755986627642e-05,
+      "loss": 1.0562,
+      "step": 5222
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.095323324203491,
+      "learning_rate": 1.6426136693091116e-05,
+      "loss": 1.101,
+      "step": 5223
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8750954866409302,
+      "learning_rate": 1.6424713298157784e-05,
+      "loss": 0.9934,
+      "step": 5224
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8303214311599731,
+      "learning_rate": 1.6423289681525544e-05,
+      "loss": 1.0742,
+      "step": 5225
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.825567603111267,
+      "learning_rate": 1.6421865843243522e-05,
+      "loss": 0.9778,
+      "step": 5226
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5514206886291504,
+      "learning_rate": 1.6420441783360845e-05,
+      "loss": 1.051,
+      "step": 5227
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6774661540985107,
+      "learning_rate": 1.641901750192666e-05,
+      "loss": 0.9834,
+      "step": 5228
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8114384412765503,
+      "learning_rate": 1.6417592998990107e-05,
+      "loss": 1.029,
+      "step": 5229
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9738600254058838,
+      "learning_rate": 1.641616827460035e-05,
+      "loss": 1.0089,
+      "step": 5230
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.774478793144226,
+      "learning_rate": 1.6414743328806547e-05,
+      "loss": 1.2079,
+      "step": 5231
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8085359334945679,
+      "learning_rate": 1.641331816165787e-05,
+      "loss": 1.0086,
+      "step": 5232
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6217718124389648,
+      "learning_rate": 1.64118927732035e-05,
+      "loss": 1.0355,
+      "step": 5233
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6912823915481567,
+      "learning_rate": 1.6410467163492624e-05,
+      "loss": 1.0614,
+      "step": 5234
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.951919436454773,
+      "learning_rate": 1.640904133257443e-05,
+      "loss": 1.1116,
+      "step": 5235
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8196876049041748,
+      "learning_rate": 1.6407615280498125e-05,
+      "loss": 1.0041,
+      "step": 5236
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6498008966445923,
+      "learning_rate": 1.640618900731291e-05,
+      "loss": 1.0425,
+      "step": 5237
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7506357431411743,
+      "learning_rate": 1.6404762513068014e-05,
+      "loss": 0.9513,
+      "step": 5238
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8732815980911255,
+      "learning_rate": 1.640333579781265e-05,
+      "loss": 0.9872,
+      "step": 5239
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7445695400238037,
+      "learning_rate": 1.6401908861596054e-05,
+      "loss": 1.1108,
+      "step": 5240
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7546354532241821,
+      "learning_rate": 1.6400481704467468e-05,
+      "loss": 1.0896,
+      "step": 5241
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8240140676498413,
+      "learning_rate": 1.6399054326476142e-05,
+      "loss": 0.9225,
+      "step": 5242
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7127807140350342,
+      "learning_rate": 1.639762672767132e-05,
+      "loss": 1.0578,
+      "step": 5243
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.932915449142456,
+      "learning_rate": 1.6396198908102273e-05,
+      "loss": 1.0383,
+      "step": 5244
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8805278539657593,
+      "learning_rate": 1.6394770867818267e-05,
+      "loss": 1.0237,
+      "step": 5245
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.798864722251892,
+      "learning_rate": 1.6393342606868582e-05,
+      "loss": 0.9055,
+      "step": 5246
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7802367210388184,
+      "learning_rate": 1.6391914125302505e-05,
+      "loss": 1.0358,
+      "step": 5247
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6692166328430176,
+      "learning_rate": 1.6390485423169323e-05,
+      "loss": 1.002,
+      "step": 5248
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8615047931671143,
+      "learning_rate": 1.6389056500518343e-05,
+      "loss": 1.0055,
+      "step": 5249
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7731149196624756,
+      "learning_rate": 1.638762735739887e-05,
+      "loss": 1.0703,
+      "step": 5250
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8083915710449219,
+      "learning_rate": 1.6386197993860218e-05,
+      "loss": 0.9796,
+      "step": 5251
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.074035167694092,
+      "learning_rate": 1.6384768409951714e-05,
+      "loss": 1.0288,
+      "step": 5252
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7210679054260254,
+      "learning_rate": 1.6383338605722686e-05,
+      "loss": 0.9803,
+      "step": 5253
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8645868301391602,
+      "learning_rate": 1.6381908581222477e-05,
+      "loss": 1.0703,
+      "step": 5254
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0244362354278564,
+      "learning_rate": 1.6380478336500427e-05,
+      "loss": 1.0403,
+      "step": 5255
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9616894721984863,
+      "learning_rate": 1.6379047871605897e-05,
+      "loss": 0.9535,
+      "step": 5256
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.73922598361969,
+      "learning_rate": 1.6377617186588236e-05,
+      "loss": 0.9678,
+      "step": 5257
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7541407346725464,
+      "learning_rate": 1.637618628149683e-05,
+      "loss": 1.0348,
+      "step": 5258
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8831788301467896,
+      "learning_rate": 1.637475515638104e-05,
+      "loss": 1.073,
+      "step": 5259
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8657195568084717,
+      "learning_rate": 1.6373323811290262e-05,
+      "loss": 0.9982,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7818225622177124,
+      "learning_rate": 1.6371892246273877e-05,
+      "loss": 1.0534,
+      "step": 5261
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.843577265739441,
+      "learning_rate": 1.6370460461381292e-05,
+      "loss": 1.0181,
+      "step": 5262
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.842872142791748,
+      "learning_rate": 1.636902845666191e-05,
+      "loss": 0.9868,
+      "step": 5263
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.831213355064392,
+      "learning_rate": 1.636759623216515e-05,
+      "loss": 1.05,
+      "step": 5264
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6557680368423462,
+      "learning_rate": 1.636616378794043e-05,
+      "loss": 0.9932,
+      "step": 5265
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8199869394302368,
+      "learning_rate": 1.636473112403718e-05,
+      "loss": 0.9797,
+      "step": 5266
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.40049409866333,
+      "learning_rate": 1.6363298240504842e-05,
+      "loss": 0.665,
+      "step": 5267
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7990132570266724,
+      "learning_rate": 1.6361865137392855e-05,
+      "loss": 1.0224,
+      "step": 5268
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.637906551361084,
+      "learning_rate": 1.636043181475067e-05,
+      "loss": 0.9486,
+      "step": 5269
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7352979183197021,
+      "learning_rate": 1.6358998272627754e-05,
+      "loss": 0.9786,
+      "step": 5270
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7840981483459473,
+      "learning_rate": 1.6357564511073567e-05,
+      "loss": 1.0571,
+      "step": 5271
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6980482339859009,
+      "learning_rate": 1.635613053013759e-05,
+      "loss": 1.0016,
+      "step": 5272
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8260037899017334,
+      "learning_rate": 1.6354696329869307e-05,
+      "loss": 0.959,
+      "step": 5273
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8289821147918701,
+      "learning_rate": 1.63532619103182e-05,
+      "loss": 1.0565,
+      "step": 5274
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.04422926902771,
+      "learning_rate": 1.6351827271533774e-05,
+      "loss": 1.0866,
+      "step": 5275
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.857997179031372,
+      "learning_rate": 1.635039241356553e-05,
+      "loss": 1.0391,
+      "step": 5276
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9671493768692017,
+      "learning_rate": 1.6348957336462982e-05,
+      "loss": 1.0893,
+      "step": 5277
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1050386428833008,
+      "learning_rate": 1.6347522040275653e-05,
+      "loss": 0.6131,
+      "step": 5278
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7517123222351074,
+      "learning_rate": 1.6346086525053072e-05,
+      "loss": 1.1215,
+      "step": 5279
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0481855869293213,
+      "learning_rate": 1.634465079084477e-05,
+      "loss": 0.9805,
+      "step": 5280
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6988558769226074,
+      "learning_rate": 1.6343214837700296e-05,
+      "loss": 1.0837,
+      "step": 5281
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8934874534606934,
+      "learning_rate": 1.634177866566919e-05,
+      "loss": 1.0736,
+      "step": 5282
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8343819379806519,
+      "learning_rate": 1.6340342274801024e-05,
+      "loss": 1.0241,
+      "step": 5283
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8228952884674072,
+      "learning_rate": 1.6338905665145352e-05,
+      "loss": 0.9884,
+      "step": 5284
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8130691051483154,
+      "learning_rate": 1.6337468836751753e-05,
+      "loss": 0.9255,
+      "step": 5285
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6233998537063599,
+      "learning_rate": 1.6336031789669808e-05,
+      "loss": 0.9477,
+      "step": 5286
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.87525475025177,
+      "learning_rate": 1.6334594523949107e-05,
+      "loss": 1.0055,
+      "step": 5287
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.672725796699524,
+      "learning_rate": 1.633315703963924e-05,
+      "loss": 1.0006,
+      "step": 5288
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.157294511795044,
+      "learning_rate": 1.6331719336789817e-05,
+      "loss": 0.9921,
+      "step": 5289
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8449910879135132,
+      "learning_rate": 1.6330281415450446e-05,
+      "loss": 1.077,
+      "step": 5290
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9979575872421265,
+      "learning_rate": 1.6328843275670748e-05,
+      "loss": 0.9675,
+      "step": 5291
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9981364011764526,
+      "learning_rate": 1.6327404917500345e-05,
+      "loss": 1.0143,
+      "step": 5292
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9902487993240356,
+      "learning_rate": 1.6325966340988877e-05,
+      "loss": 1.0263,
+      "step": 5293
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7841180562973022,
+      "learning_rate": 1.6324527546185977e-05,
+      "loss": 0.9924,
+      "step": 5294
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.185438871383667,
+      "learning_rate": 1.6323088533141304e-05,
+      "loss": 0.9932,
+      "step": 5295
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8306366205215454,
+      "learning_rate": 1.6321649301904505e-05,
+      "loss": 0.9876,
+      "step": 5296
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.998640775680542,
+      "learning_rate": 1.6320209852525242e-05,
+      "loss": 1.0704,
+      "step": 5297
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.028083086013794,
+      "learning_rate": 1.6318770185053197e-05,
+      "loss": 0.9842,
+      "step": 5298
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9142224788665771,
+      "learning_rate": 1.6317330299538046e-05,
+      "loss": 0.9059,
+      "step": 5299
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9877183437347412,
+      "learning_rate": 1.631589019602947e-05,
+      "loss": 1.0,
+      "step": 5300
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8000242710113525,
+      "learning_rate": 1.6314449874577166e-05,
+      "loss": 1.0743,
+      "step": 5301
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8279976844787598,
+      "learning_rate": 1.631300933523084e-05,
+      "loss": 0.9637,
+      "step": 5302
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7045879364013672,
+      "learning_rate": 1.631156857804019e-05,
+      "loss": 0.9535,
+      "step": 5303
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7607024908065796,
+      "learning_rate": 1.6310127603054945e-05,
+      "loss": 1.0108,
+      "step": 5304
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.92744779586792,
+      "learning_rate": 1.630868641032482e-05,
+      "loss": 0.991,
+      "step": 5305
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0643630027770996,
+      "learning_rate": 1.6307244999899547e-05,
+      "loss": 1.0553,
+      "step": 5306
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.817264437675476,
+      "learning_rate": 1.6305803371828874e-05,
+      "loss": 1.0755,
+      "step": 5307
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9802658557891846,
+      "learning_rate": 1.6304361526162534e-05,
+      "loss": 0.9962,
+      "step": 5308
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.743973731994629,
+      "learning_rate": 1.6302919462950294e-05,
+      "loss": 1.056,
+      "step": 5309
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9193590879440308,
+      "learning_rate": 1.6301477182241903e-05,
+      "loss": 1.0127,
+      "step": 5310
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1546295881271362,
+      "learning_rate": 1.6300034684087145e-05,
+      "loss": 0.568,
+      "step": 5311
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8126161098480225,
+      "learning_rate": 1.6298591968535784e-05,
+      "loss": 1.0318,
+      "step": 5312
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7339115142822266,
+      "learning_rate": 1.6297149035637608e-05,
+      "loss": 0.9933,
+      "step": 5313
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9690638780593872,
+      "learning_rate": 1.6295705885442413e-05,
+      "loss": 1.0665,
+      "step": 5314
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7355304956436157,
+      "learning_rate": 1.6294262517999994e-05,
+      "loss": 0.9647,
+      "step": 5315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6641390323638916,
+      "learning_rate": 1.6292818933360153e-05,
+      "loss": 0.9512,
+      "step": 5316
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1275404691696167,
+      "learning_rate": 1.629137513157271e-05,
+      "loss": 0.6151,
+      "step": 5317
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9910144805908203,
+      "learning_rate": 1.628993111268749e-05,
+      "loss": 0.9756,
+      "step": 5318
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8117080926895142,
+      "learning_rate": 1.6288486876754314e-05,
+      "loss": 1.0279,
+      "step": 5319
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9032024145126343,
+      "learning_rate": 1.628704242382302e-05,
+      "loss": 1.0483,
+      "step": 5320
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7163499593734741,
+      "learning_rate": 1.6285597753943458e-05,
+      "loss": 0.9742,
+      "step": 5321
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8539586067199707,
+      "learning_rate": 1.6284152867165475e-05,
+      "loss": 1.0372,
+      "step": 5322
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7598193883895874,
+      "learning_rate": 1.628270776353893e-05,
+      "loss": 0.9659,
+      "step": 5323
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8572660684585571,
+      "learning_rate": 1.628126244311369e-05,
+      "loss": 1.0197,
+      "step": 5324
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7211501598358154,
+      "learning_rate": 1.6279816905939627e-05,
+      "loss": 1.1184,
+      "step": 5325
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6807236671447754,
+      "learning_rate": 1.627837115206663e-05,
+      "loss": 0.9651,
+      "step": 5326
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1544266939163208,
+      "learning_rate": 1.6276925181544577e-05,
+      "loss": 0.5849,
+      "step": 5327
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0740216970443726,
+      "learning_rate": 1.6275478994423372e-05,
+      "loss": 0.7002,
+      "step": 5328
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.021015167236328,
+      "learning_rate": 1.627403259075292e-05,
+      "loss": 1.0665,
+      "step": 5329
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7597944736480713,
+      "learning_rate": 1.6272585970583124e-05,
+      "loss": 1.0188,
+      "step": 5330
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6504372358322144,
+      "learning_rate": 1.6271139133963906e-05,
+      "loss": 0.9756,
+      "step": 5331
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8334649801254272,
+      "learning_rate": 1.62696920809452e-05,
+      "loss": 1.0298,
+      "step": 5332
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.993582010269165,
+      "learning_rate": 1.6268244811576932e-05,
+      "loss": 1.0806,
+      "step": 5333
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0547441244125366,
+      "learning_rate": 1.6266797325909045e-05,
+      "loss": 0.6393,
+      "step": 5334
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.827067494392395,
+      "learning_rate": 1.626534962399149e-05,
+      "loss": 1.0917,
+      "step": 5335
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8336405754089355,
+      "learning_rate": 1.626390170587422e-05,
+      "loss": 0.9961,
+      "step": 5336
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6455762386322021,
+      "learning_rate": 1.6262453571607198e-05,
+      "loss": 0.9522,
+      "step": 5337
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6795203685760498,
+      "learning_rate": 1.6261005221240394e-05,
+      "loss": 1.0395,
+      "step": 5338
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8535797595977783,
+      "learning_rate": 1.6259556654823793e-05,
+      "loss": 0.9948,
+      "step": 5339
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8329442739486694,
+      "learning_rate": 1.6258107872407376e-05,
+      "loss": 1.013,
+      "step": 5340
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0877106189727783,
+      "learning_rate": 1.625665887404114e-05,
+      "loss": 0.6125,
+      "step": 5341
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8665584325790405,
+      "learning_rate": 1.6255209659775082e-05,
+      "loss": 1.0,
+      "step": 5342
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8733494281768799,
+      "learning_rate": 1.625376022965921e-05,
+      "loss": 0.9807,
+      "step": 5343
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.891006588935852,
+      "learning_rate": 1.6252310583743544e-05,
+      "loss": 0.9436,
+      "step": 5344
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.889841079711914,
+      "learning_rate": 1.6250860722078106e-05,
+      "loss": 1.0553,
+      "step": 5345
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9409970045089722,
+      "learning_rate": 1.6249410644712925e-05,
+      "loss": 1.1267,
+      "step": 5346
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9099448919296265,
+      "learning_rate": 1.624796035169804e-05,
+      "loss": 1.0782,
+      "step": 5347
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9502898454666138,
+      "learning_rate": 1.6246509843083492e-05,
+      "loss": 1.0599,
+      "step": 5348
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7492371797561646,
+      "learning_rate": 1.6245059118919342e-05,
+      "loss": 1.041,
+      "step": 5349
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9055966138839722,
+      "learning_rate": 1.6243608179255645e-05,
+      "loss": 0.9941,
+      "step": 5350
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9000831842422485,
+      "learning_rate": 1.624215702414247e-05,
+      "loss": 0.9467,
+      "step": 5351
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8919672966003418,
+      "learning_rate": 1.6240705653629896e-05,
+      "loss": 1.0168,
+      "step": 5352
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.907387614250183,
+      "learning_rate": 1.6239254067768002e-05,
+      "loss": 1.0353,
+      "step": 5353
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7466566562652588,
+      "learning_rate": 1.6237802266606877e-05,
+      "loss": 0.9752,
+      "step": 5354
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6579678058624268,
+      "learning_rate": 1.623635025019662e-05,
+      "loss": 0.951,
+      "step": 5355
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9517204761505127,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 0.9835,
+      "step": 5356
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.91655695438385,
+      "learning_rate": 1.623344557182914e-05,
+      "loss": 0.9505,
+      "step": 5357
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.124234437942505,
+      "learning_rate": 1.623199290997215e-05,
+      "loss": 1.0431,
+      "step": 5358
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.070302724838257,
+      "learning_rate": 1.6230540033066492e-05,
+      "loss": 0.9741,
+      "step": 5359
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.90458083152771,
+      "learning_rate": 1.62290869411623e-05,
+      "loss": 1.0118,
+      "step": 5360
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8572077751159668,
+      "learning_rate": 1.622763363430972e-05,
+      "loss": 0.9646,
+      "step": 5361
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7768325805664062,
+      "learning_rate": 1.6226180112558897e-05,
+      "loss": 1.0824,
+      "step": 5362
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8336414098739624,
+      "learning_rate": 1.6224726375959994e-05,
+      "loss": 1.0104,
+      "step": 5363
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1761006116867065,
+      "learning_rate": 1.6223272424563174e-05,
+      "loss": 0.6472,
+      "step": 5364
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2218785285949707,
+      "learning_rate": 1.62218182584186e-05,
+      "loss": 0.9537,
+      "step": 5365
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.465400218963623,
+      "learning_rate": 1.622036387757646e-05,
+      "loss": 1.0643,
+      "step": 5366
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8587623834609985,
+      "learning_rate": 1.621890928208694e-05,
+      "loss": 1.1061,
+      "step": 5367
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9155011177062988,
+      "learning_rate": 1.6217454472000232e-05,
+      "loss": 1.0137,
+      "step": 5368
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8350976705551147,
+      "learning_rate": 1.621599944736654e-05,
+      "loss": 1.1068,
+      "step": 5369
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9839016795158386,
+      "learning_rate": 1.6214544208236066e-05,
+      "loss": 0.5773,
+      "step": 5370
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5843640565872192,
+      "learning_rate": 1.6213088754659033e-05,
+      "loss": 1.0513,
+      "step": 5371
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8846186399459839,
+      "learning_rate": 1.6211633086685666e-05,
+      "loss": 1.0879,
+      "step": 5372
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.819551944732666,
+      "learning_rate": 1.6210177204366187e-05,
+      "loss": 1.0244,
+      "step": 5373
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8161970376968384,
+      "learning_rate": 1.6208721107750845e-05,
+      "loss": 0.9999,
+      "step": 5374
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9262241125106812,
+      "learning_rate": 1.6207264796889875e-05,
+      "loss": 1.0385,
+      "step": 5375
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0153942108154297,
+      "learning_rate": 1.6205808271833542e-05,
+      "loss": 1.0683,
+      "step": 5376
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7681382894515991,
+      "learning_rate": 1.6204351532632098e-05,
+      "loss": 1.1153,
+      "step": 5377
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.666382908821106,
+      "learning_rate": 1.6202894579335815e-05,
+      "loss": 1.0607,
+      "step": 5378
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8089468479156494,
+      "learning_rate": 1.6201437411994967e-05,
+      "loss": 0.9768,
+      "step": 5379
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7946926355361938,
+      "learning_rate": 1.6199980030659837e-05,
+      "loss": 0.9943,
+      "step": 5380
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8243076801300049,
+      "learning_rate": 1.6198522435380716e-05,
+      "loss": 1.049,
+      "step": 5381
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8830714225769043,
+      "learning_rate": 1.61970646262079e-05,
+      "loss": 1.0789,
+      "step": 5382
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8030109405517578,
+      "learning_rate": 1.6195606603191692e-05,
+      "loss": 1.0082,
+      "step": 5383
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8486520051956177,
+      "learning_rate": 1.619414836638241e-05,
+      "loss": 1.0603,
+      "step": 5384
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0993711948394775,
+      "learning_rate": 1.619268991583037e-05,
+      "loss": 1.0644,
+      "step": 5385
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1216024160385132,
+      "learning_rate": 1.61912312515859e-05,
+      "loss": 0.6071,
+      "step": 5386
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8520299196243286,
+      "learning_rate": 1.6189772373699334e-05,
+      "loss": 0.9751,
+      "step": 5387
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.023817300796509,
+      "learning_rate": 1.6188313282221008e-05,
+      "loss": 1.0642,
+      "step": 5388
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0483896732330322,
+      "learning_rate": 1.618685397720128e-05,
+      "loss": 1.0392,
+      "step": 5389
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0124213695526123,
+      "learning_rate": 1.618539445869051e-05,
+      "loss": 1.0367,
+      "step": 5390
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9030776023864746,
+      "learning_rate": 1.618393472673905e-05,
+      "loss": 1.0302,
+      "step": 5391
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8303340673446655,
+      "learning_rate": 1.6182474781397277e-05,
+      "loss": 0.9987,
+      "step": 5392
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.844299077987671,
+      "learning_rate": 1.6181014622715568e-05,
+      "loss": 1.0248,
+      "step": 5393
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.733989953994751,
+      "learning_rate": 1.6179554250744315e-05,
+      "loss": 1.0522,
+      "step": 5394
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9461452960968018,
+      "learning_rate": 1.6178093665533903e-05,
+      "loss": 0.9966,
+      "step": 5395
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7193036079406738,
+      "learning_rate": 1.6176632867134738e-05,
+      "loss": 1.0738,
+      "step": 5396
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0400829315185547,
+      "learning_rate": 1.6175171855597224e-05,
+      "loss": 1.0613,
+      "step": 5397
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.936630129814148,
+      "learning_rate": 1.617371063097178e-05,
+      "loss": 1.0325,
+      "step": 5398
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8583730459213257,
+      "learning_rate": 1.6172249193308827e-05,
+      "loss": 0.976,
+      "step": 5399
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.042346239089966,
+      "learning_rate": 1.61707875426588e-05,
+      "loss": 1.0281,
+      "step": 5400
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2337827682495117,
+      "learning_rate": 1.6169325679072127e-05,
+      "loss": 1.1197,
+      "step": 5401
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8800380229949951,
+      "learning_rate": 1.6167863602599263e-05,
+      "loss": 1.0653,
+      "step": 5402
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8144276142120361,
+      "learning_rate": 1.616640131329065e-05,
+      "loss": 1.0001,
+      "step": 5403
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9197808504104614,
+      "learning_rate": 1.6164938811196758e-05,
+      "loss": 0.9966,
+      "step": 5404
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.933760166168213,
+      "learning_rate": 1.6163476096368046e-05,
+      "loss": 0.9963,
+      "step": 5405
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8016561269760132,
+      "learning_rate": 1.6162013168854992e-05,
+      "loss": 1.0201,
+      "step": 5406
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9271571636199951,
+      "learning_rate": 1.6160550028708077e-05,
+      "loss": 1.0579,
+      "step": 5407
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.022021532058716,
+      "learning_rate": 1.6159086675977785e-05,
+      "loss": 1.0895,
+      "step": 5408
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9065383672714233,
+      "learning_rate": 1.6157623110714618e-05,
+      "loss": 0.9586,
+      "step": 5409
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9301396608352661,
+      "learning_rate": 1.615615933296908e-05,
+      "loss": 1.0325,
+      "step": 5410
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8570078611373901,
+      "learning_rate": 1.6154695342791682e-05,
+      "loss": 0.9274,
+      "step": 5411
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.914941668510437,
+      "learning_rate": 1.6153231140232936e-05,
+      "loss": 1.042,
+      "step": 5412
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.126347303390503,
+      "learning_rate": 1.6151766725343373e-05,
+      "loss": 1.0154,
+      "step": 5413
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1510863304138184,
+      "learning_rate": 1.6150302098173523e-05,
+      "loss": 1.0231,
+      "step": 5414
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.855362892150879,
+      "learning_rate": 1.6148837258773934e-05,
+      "loss": 1.0199,
+      "step": 5415
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.887709617614746,
+      "learning_rate": 1.6147372207195142e-05,
+      "loss": 1.0571,
+      "step": 5416
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8217957019805908,
+      "learning_rate": 1.6145906943487706e-05,
+      "loss": 1.0676,
+      "step": 5417
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1708956956863403,
+      "learning_rate": 1.6144441467702194e-05,
+      "loss": 0.6114,
+      "step": 5418
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7735998630523682,
+      "learning_rate": 1.6142975779889167e-05,
+      "loss": 1.0294,
+      "step": 5419
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9344372749328613,
+      "learning_rate": 1.6141509880099205e-05,
+      "loss": 0.9958,
+      "step": 5420
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8749792575836182,
+      "learning_rate": 1.6140043768382894e-05,
+      "loss": 1.0335,
+      "step": 5421
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9425380229949951,
+      "learning_rate": 1.6138577444790826e-05,
+      "loss": 0.9642,
+      "step": 5422
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8311879634857178,
+      "learning_rate": 1.6137110909373595e-05,
+      "loss": 0.9392,
+      "step": 5423
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1880133152008057,
+      "learning_rate": 1.6135644162181814e-05,
+      "loss": 1.025,
+      "step": 5424
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9264405965805054,
+      "learning_rate": 1.6134177203266087e-05,
+      "loss": 1.0532,
+      "step": 5425
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8340034484863281,
+      "learning_rate": 1.6132710032677043e-05,
+      "loss": 1.0363,
+      "step": 5426
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6725407838821411,
+      "learning_rate": 1.6131242650465305e-05,
+      "loss": 1.0238,
+      "step": 5427
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.705064058303833,
+      "learning_rate": 1.6129775056681515e-05,
+      "loss": 0.9729,
+      "step": 5428
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9954407215118408,
+      "learning_rate": 1.6128307251376304e-05,
+      "loss": 0.966,
+      "step": 5429
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7256083488464355,
+      "learning_rate": 1.612683923460033e-05,
+      "loss": 0.9855,
+      "step": 5430
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9120780229568481,
+      "learning_rate": 1.612537100640425e-05,
+      "loss": 1.0178,
+      "step": 5431
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.009671688079834,
+      "learning_rate": 1.6123902566838726e-05,
+      "loss": 0.9795,
+      "step": 5432
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6778771877288818,
+      "learning_rate": 1.6122433915954433e-05,
+      "loss": 0.9789,
+      "step": 5433
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7159498929977417,
+      "learning_rate": 1.6120965053802047e-05,
+      "loss": 1.0189,
+      "step": 5434
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1774489879608154,
+      "learning_rate": 1.6119495980432254e-05,
+      "loss": 0.6219,
+      "step": 5435
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8734923601150513,
+      "learning_rate": 1.611802669589575e-05,
+      "loss": 1.0445,
+      "step": 5436
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0146141052246094,
+      "learning_rate": 1.6116557200243234e-05,
+      "loss": 1.0338,
+      "step": 5437
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8196088075637817,
+      "learning_rate": 1.6115087493525416e-05,
+      "loss": 0.9966,
+      "step": 5438
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7152007818222046,
+      "learning_rate": 1.611361757579301e-05,
+      "loss": 1.0772,
+      "step": 5439
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9442161321640015,
+      "learning_rate": 1.611214744709674e-05,
+      "loss": 1.0263,
+      "step": 5440
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8195992708206177,
+      "learning_rate": 1.611067710748733e-05,
+      "loss": 1.0616,
+      "step": 5441
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8805011510849,
+      "learning_rate": 1.610920655701553e-05,
+      "loss": 1.0128,
+      "step": 5442
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9722251892089844,
+      "learning_rate": 1.6107735795732072e-05,
+      "loss": 1.03,
+      "step": 5443
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9770876169204712,
+      "learning_rate": 1.6106264823687716e-05,
+      "loss": 1.017,
+      "step": 5444
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.781185507774353,
+      "learning_rate": 1.6104793640933215e-05,
+      "loss": 1.0172,
+      "step": 5445
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7444640398025513,
+      "learning_rate": 1.6103322247519343e-05,
+      "loss": 1.0289,
+      "step": 5446
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0717390775680542,
+      "learning_rate": 1.6101850643496865e-05,
+      "loss": 0.6202,
+      "step": 5447
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1352232694625854,
+      "learning_rate": 1.610037882891657e-05,
+      "loss": 0.6291,
+      "step": 5448
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6499055624008179,
+      "learning_rate": 1.6098906803829238e-05,
+      "loss": 0.9389,
+      "step": 5449
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.085078716278076,
+      "learning_rate": 1.609743456828567e-05,
+      "loss": 0.9547,
+      "step": 5450
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8660414218902588,
+      "learning_rate": 1.609596212233667e-05,
+      "loss": 1.01,
+      "step": 5451
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7397100925445557,
+      "learning_rate": 1.609448946603304e-05,
+      "loss": 1.0309,
+      "step": 5452
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8257865905761719,
+      "learning_rate": 1.609301659942561e-05,
+      "loss": 0.9469,
+      "step": 5453
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9190455675125122,
+      "learning_rate": 1.6091543522565194e-05,
+      "loss": 1.0128,
+      "step": 5454
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0373733043670654,
+      "learning_rate": 1.6090070235502625e-05,
+      "loss": 1.0213,
+      "step": 5455
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.737097144126892,
+      "learning_rate": 1.608859673828875e-05,
+      "loss": 1.0284,
+      "step": 5456
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8663862943649292,
+      "learning_rate": 1.6087123030974403e-05,
+      "loss": 1.029,
+      "step": 5457
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8301271200180054,
+      "learning_rate": 1.6085649113610447e-05,
+      "loss": 0.9813,
+      "step": 5458
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8740274906158447,
+      "learning_rate": 1.6084174986247738e-05,
+      "loss": 1.0954,
+      "step": 5459
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8680083751678467,
+      "learning_rate": 1.6082700648937146e-05,
+      "loss": 0.9862,
+      "step": 5460
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0268795490264893,
+      "learning_rate": 1.6081226101729547e-05,
+      "loss": 1.0101,
+      "step": 5461
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7622219324111938,
+      "learning_rate": 1.6079751344675823e-05,
+      "loss": 1.0289,
+      "step": 5462
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8151675462722778,
+      "learning_rate": 1.6078276377826862e-05,
+      "loss": 1.0342,
+      "step": 5463
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8670762777328491,
+      "learning_rate": 1.6076801201233562e-05,
+      "loss": 0.9738,
+      "step": 5464
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8911508321762085,
+      "learning_rate": 1.6075325814946828e-05,
+      "loss": 1.0476,
+      "step": 5465
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7749561071395874,
+      "learning_rate": 1.6073850219017572e-05,
+      "loss": 1.0473,
+      "step": 5466
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8277431726455688,
+      "learning_rate": 1.607237441349671e-05,
+      "loss": 1.0334,
+      "step": 5467
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8503179550170898,
+      "learning_rate": 1.6070898398435167e-05,
+      "loss": 1.0163,
+      "step": 5468
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9117094278335571,
+      "learning_rate": 1.6069422173883883e-05,
+      "loss": 1.0444,
+      "step": 5469
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.813779592514038,
+      "learning_rate": 1.606794573989379e-05,
+      "loss": 1.0513,
+      "step": 5470
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8578535318374634,
+      "learning_rate": 1.6066469096515845e-05,
+      "loss": 1.0688,
+      "step": 5471
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.020658016204834,
+      "learning_rate": 1.6064992243800993e-05,
+      "loss": 1.0417,
+      "step": 5472
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.888777732849121,
+      "learning_rate": 1.6063515181800203e-05,
+      "loss": 1.0996,
+      "step": 5473
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8295797109603882,
+      "learning_rate": 1.606203791056444e-05,
+      "loss": 1.022,
+      "step": 5474
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8093209266662598,
+      "learning_rate": 1.6060560430144683e-05,
+      "loss": 1.0136,
+      "step": 5475
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.977956771850586,
+      "learning_rate": 1.6059082740591915e-05,
+      "loss": 1.0341,
+      "step": 5476
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9586621522903442,
+      "learning_rate": 1.6057604841957127e-05,
+      "loss": 1.054,
+      "step": 5477
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8799439668655396,
+      "learning_rate": 1.605612673429132e-05,
+      "loss": 1.0302,
+      "step": 5478
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8172889947891235,
+      "learning_rate": 1.6054648417645493e-05,
+      "loss": 1.0173,
+      "step": 5479
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8336008787155151,
+      "learning_rate": 1.6053169892070664e-05,
+      "loss": 1.0112,
+      "step": 5480
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9452742338180542,
+      "learning_rate": 1.605169115761785e-05,
+      "loss": 0.9467,
+      "step": 5481
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9023585319519043,
+      "learning_rate": 1.6050212214338076e-05,
+      "loss": 1.0445,
+      "step": 5482
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.932289481163025,
+      "learning_rate": 1.6048733062282385e-05,
+      "loss": 1.07,
+      "step": 5483
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6552865505218506,
+      "learning_rate": 1.6047253701501807e-05,
+      "loss": 1.0699,
+      "step": 5484
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9912241697311401,
+      "learning_rate": 1.60457741320474e-05,
+      "loss": 1.0184,
+      "step": 5485
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9628201723098755,
+      "learning_rate": 1.6044294353970212e-05,
+      "loss": 0.9072,
+      "step": 5486
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9564534425735474,
+      "learning_rate": 1.6042814367321313e-05,
+      "loss": 1.0166,
+      "step": 5487
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8115447759628296,
+      "learning_rate": 1.604133417215177e-05,
+      "loss": 0.9885,
+      "step": 5488
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9882843494415283,
+      "learning_rate": 1.603985376851266e-05,
+      "loss": 1.0018,
+      "step": 5489
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8080862760543823,
+      "learning_rate": 1.6038373156455068e-05,
+      "loss": 0.9718,
+      "step": 5490
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4758456945419312,
+      "learning_rate": 1.6036892336030086e-05,
+      "loss": 0.7086,
+      "step": 5491
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8091721534729004,
+      "learning_rate": 1.6035411307288814e-05,
+      "loss": 1.008,
+      "step": 5492
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1586380004882812,
+      "learning_rate": 1.6033930070282357e-05,
+      "loss": 0.9716,
+      "step": 5493
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7824774980545044,
+      "learning_rate": 1.603244862506182e-05,
+      "loss": 0.9699,
+      "step": 5494
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.866288185119629,
+      "learning_rate": 1.603096697167834e-05,
+      "loss": 0.9373,
+      "step": 5495
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.987775444984436,
+      "learning_rate": 1.6029485110183037e-05,
+      "loss": 1.0997,
+      "step": 5496
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8538475036621094,
+      "learning_rate": 1.6028003040627042e-05,
+      "loss": 1.0204,
+      "step": 5497
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.847609043121338,
+      "learning_rate": 1.6026520763061504e-05,
+      "loss": 1.0592,
+      "step": 5498
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7513768672943115,
+      "learning_rate": 1.602503827753757e-05,
+      "loss": 0.9833,
+      "step": 5499
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0010900497436523,
+      "learning_rate": 1.6023555584106392e-05,
+      "loss": 1.0425,
+      "step": 5500
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.662872076034546,
+      "learning_rate": 1.6022072682819138e-05,
+      "loss": 1.1165,
+      "step": 5501
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8837188482284546,
+      "learning_rate": 1.6020589573726976e-05,
+      "loss": 0.9901,
+      "step": 5502
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7075861692428589,
+      "learning_rate": 1.6019106256881088e-05,
+      "loss": 0.9575,
+      "step": 5503
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8860602378845215,
+      "learning_rate": 1.6017622732332656e-05,
+      "loss": 1.0477,
+      "step": 5504
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9120173454284668,
+      "learning_rate": 1.6016139000132873e-05,
+      "loss": 0.9778,
+      "step": 5505
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9226152896881104,
+      "learning_rate": 1.601465506033294e-05,
+      "loss": 1.0247,
+      "step": 5506
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8007512092590332,
+      "learning_rate": 1.601317091298406e-05,
+      "loss": 0.9058,
+      "step": 5507
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2160261869430542,
+      "learning_rate": 1.601168655813745e-05,
+      "loss": 0.6049,
+      "step": 5508
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8075487613677979,
+      "learning_rate": 1.6010201995844328e-05,
+      "loss": 1.0097,
+      "step": 5509
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7775397300720215,
+      "learning_rate": 1.6008717226155925e-05,
+      "loss": 0.9358,
+      "step": 5510
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8586055040359497,
+      "learning_rate": 1.6007232249123478e-05,
+      "loss": 1.0363,
+      "step": 5511
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7582409381866455,
+      "learning_rate": 1.6005747064798224e-05,
+      "loss": 0.9495,
+      "step": 5512
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.885279655456543,
+      "learning_rate": 1.6004261673231414e-05,
+      "loss": 0.9902,
+      "step": 5513
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9768409729003906,
+      "learning_rate": 1.6002776074474308e-05,
+      "loss": 1.0861,
+      "step": 5514
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7169522047042847,
+      "learning_rate": 1.6001290268578164e-05,
+      "loss": 0.9648,
+      "step": 5515
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8215198516845703,
+      "learning_rate": 1.5999804255594262e-05,
+      "loss": 1.019,
+      "step": 5516
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8581769466400146,
+      "learning_rate": 1.599831803557387e-05,
+      "loss": 1.0159,
+      "step": 5517
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4309744834899902,
+      "learning_rate": 1.599683160856828e-05,
+      "loss": 1.0321,
+      "step": 5518
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9362409114837646,
+      "learning_rate": 1.5995344974628787e-05,
+      "loss": 1.0545,
+      "step": 5519
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8471745252609253,
+      "learning_rate": 1.5993858133806684e-05,
+      "loss": 1.0043,
+      "step": 5520
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6273977756500244,
+      "learning_rate": 1.5992371086153276e-05,
+      "loss": 0.9699,
+      "step": 5521
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.956954002380371,
+      "learning_rate": 1.5990883831719886e-05,
+      "loss": 1.0649,
+      "step": 5522
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.799927830696106,
+      "learning_rate": 1.5989396370557824e-05,
+      "loss": 0.9226,
+      "step": 5523
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8341554403305054,
+      "learning_rate": 1.598790870271843e-05,
+      "loss": 1.0271,
+      "step": 5524
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3146204948425293,
+      "learning_rate": 1.5986420828253032e-05,
+      "loss": 0.9936,
+      "step": 5525
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9320725202560425,
+      "learning_rate": 1.598493274721297e-05,
+      "loss": 1.1027,
+      "step": 5526
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7183300256729126,
+      "learning_rate": 1.59834444596496e-05,
+      "loss": 1.1023,
+      "step": 5527
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8065663576126099,
+      "learning_rate": 1.5981955965614274e-05,
+      "loss": 1.0101,
+      "step": 5528
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.793352723121643,
+      "learning_rate": 1.598046726515836e-05,
+      "loss": 0.9984,
+      "step": 5529
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7683485746383667,
+      "learning_rate": 1.5978978358333223e-05,
+      "loss": 0.9855,
+      "step": 5530
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7994049787521362,
+      "learning_rate": 1.597748924519025e-05,
+      "loss": 0.912,
+      "step": 5531
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7815009355545044,
+      "learning_rate": 1.5975999925780812e-05,
+      "loss": 0.9812,
+      "step": 5532
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9630521535873413,
+      "learning_rate": 1.5974510400156316e-05,
+      "loss": 1.0515,
+      "step": 5533
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8291627168655396,
+      "learning_rate": 1.5973020668368155e-05,
+      "loss": 1.0555,
+      "step": 5534
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7765790224075317,
+      "learning_rate": 1.5971530730467736e-05,
+      "loss": 0.9725,
+      "step": 5535
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.124052047729492,
+      "learning_rate": 1.597004058650647e-05,
+      "loss": 1.0564,
+      "step": 5536
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6187347173690796,
+      "learning_rate": 1.596855023653578e-05,
+      "loss": 1.0079,
+      "step": 5537
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7103477716445923,
+      "learning_rate": 1.5967059680607097e-05,
+      "loss": 1.019,
+      "step": 5538
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8535102605819702,
+      "learning_rate": 1.596556891877185e-05,
+      "loss": 0.9583,
+      "step": 5539
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9701313972473145,
+      "learning_rate": 1.5964077951081484e-05,
+      "loss": 1.0428,
+      "step": 5540
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.68707275390625,
+      "learning_rate": 1.596258677758745e-05,
+      "loss": 0.9534,
+      "step": 5541
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7589253187179565,
+      "learning_rate": 1.59610953983412e-05,
+      "loss": 1.0687,
+      "step": 5542
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7668746709823608,
+      "learning_rate": 1.59596038133942e-05,
+      "loss": 1.0145,
+      "step": 5543
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.00268292427063,
+      "learning_rate": 1.5958112022797917e-05,
+      "loss": 0.9909,
+      "step": 5544
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.097862958908081,
+      "learning_rate": 1.5956620026603835e-05,
+      "loss": 0.5782,
+      "step": 5545
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.908186674118042,
+      "learning_rate": 1.595512782486344e-05,
+      "loss": 1.0802,
+      "step": 5546
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9989211559295654,
+      "learning_rate": 1.5953635417628207e-05,
+      "loss": 1.0182,
+      "step": 5547
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.927767038345337,
+      "learning_rate": 1.5952142804949654e-05,
+      "loss": 1.0438,
+      "step": 5548
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8537846803665161,
+      "learning_rate": 1.595064998687928e-05,
+      "loss": 1.0221,
+      "step": 5549
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7231436967849731,
+      "learning_rate": 1.5949156963468593e-05,
+      "loss": 1.08,
+      "step": 5550
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7040066719055176,
+      "learning_rate": 1.594766373476912e-05,
+      "loss": 1.037,
+      "step": 5551
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6797269582748413,
+      "learning_rate": 1.5946170300832385e-05,
+      "loss": 0.9258,
+      "step": 5552
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7724076509475708,
+      "learning_rate": 1.5944676661709922e-05,
+      "loss": 0.9654,
+      "step": 5553
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7064090967178345,
+      "learning_rate": 1.5943182817453277e-05,
+      "loss": 1.0494,
+      "step": 5554
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.843024730682373,
+      "learning_rate": 1.594168876811399e-05,
+      "loss": 1.0331,
+      "step": 5555
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7479760646820068,
+      "learning_rate": 1.5940194513743623e-05,
+      "loss": 0.9791,
+      "step": 5556
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6669811010360718,
+      "learning_rate": 1.593870005439374e-05,
+      "loss": 1.0223,
+      "step": 5557
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.663368582725525,
+      "learning_rate": 1.5937205390115902e-05,
+      "loss": 0.9502,
+      "step": 5558
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7865049839019775,
+      "learning_rate": 1.5935710520961693e-05,
+      "loss": 0.9874,
+      "step": 5559
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.81930673122406,
+      "learning_rate": 1.5934215446982696e-05,
+      "loss": 1.1008,
+      "step": 5560
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0533955097198486,
+      "learning_rate": 1.5932720168230497e-05,
+      "loss": 0.9945,
+      "step": 5561
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8189257383346558,
+      "learning_rate": 1.5931224684756698e-05,
+      "loss": 0.9611,
+      "step": 5562
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8144233226776123,
+      "learning_rate": 1.5929728996612905e-05,
+      "loss": 1.0139,
+      "step": 5563
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8488507270812988,
+      "learning_rate": 1.592823310385073e-05,
+      "loss": 0.9654,
+      "step": 5564
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9606572389602661,
+      "learning_rate": 1.5926737006521787e-05,
+      "loss": 1.1029,
+      "step": 5565
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7714370489120483,
+      "learning_rate": 1.5925240704677708e-05,
+      "loss": 1.0483,
+      "step": 5566
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8103426694869995,
+      "learning_rate": 1.5923744198370124e-05,
+      "loss": 1.052,
+      "step": 5567
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7000718116760254,
+      "learning_rate": 1.5922247487650674e-05,
+      "loss": 0.9843,
+      "step": 5568
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1136692762374878,
+      "learning_rate": 1.5920750572571004e-05,
+      "loss": 0.619,
+      "step": 5569
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1499528884887695,
+      "learning_rate": 1.5919253453182776e-05,
+      "loss": 0.9689,
+      "step": 5570
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9264980554580688,
+      "learning_rate": 1.591775612953764e-05,
+      "loss": 1.0648,
+      "step": 5571
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8386764526367188,
+      "learning_rate": 1.5916258601687276e-05,
+      "loss": 0.9729,
+      "step": 5572
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8946834802627563,
+      "learning_rate": 1.591476086968335e-05,
+      "loss": 0.9932,
+      "step": 5573
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7496845722198486,
+      "learning_rate": 1.591326293357755e-05,
+      "loss": 1.0286,
+      "step": 5574
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.745600700378418,
+      "learning_rate": 1.5911764793421563e-05,
+      "loss": 0.9319,
+      "step": 5575
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.769769310951233,
+      "learning_rate": 1.5910266449267088e-05,
+      "loss": 0.9636,
+      "step": 5576
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.09451425075531,
+      "learning_rate": 1.590876790116583e-05,
+      "loss": 0.6322,
+      "step": 5577
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9613984823226929,
+      "learning_rate": 1.5907269149169496e-05,
+      "loss": 0.9264,
+      "step": 5578
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8515127897262573,
+      "learning_rate": 1.5905770193329802e-05,
+      "loss": 0.9714,
+      "step": 5579
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9208461046218872,
+      "learning_rate": 1.590427103369848e-05,
+      "loss": 1.1082,
+      "step": 5580
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0351321697235107,
+      "learning_rate": 1.590277167032725e-05,
+      "loss": 0.9883,
+      "step": 5581
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9064292907714844,
+      "learning_rate": 1.5901272103267865e-05,
+      "loss": 0.9593,
+      "step": 5582
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.047064781188965,
+      "learning_rate": 1.5899772332572064e-05,
+      "loss": 1.0164,
+      "step": 5583
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8514686822891235,
+      "learning_rate": 1.58982723582916e-05,
+      "loss": 0.9972,
+      "step": 5584
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8108243942260742,
+      "learning_rate": 1.5896772180478232e-05,
+      "loss": 0.9882,
+      "step": 5585
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.162264347076416,
+      "learning_rate": 1.5895271799183728e-05,
+      "loss": 0.9096,
+      "step": 5586
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1043074131011963,
+      "learning_rate": 1.589377121445986e-05,
+      "loss": 0.9905,
+      "step": 5587
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.933682918548584,
+      "learning_rate": 1.5892270426358413e-05,
+      "loss": 1.0075,
+      "step": 5588
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.959614872932434,
+      "learning_rate": 1.5890769434931173e-05,
+      "loss": 1.0735,
+      "step": 5589
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.91002357006073,
+      "learning_rate": 1.5889268240229938e-05,
+      "loss": 1.0017,
+      "step": 5590
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.061537265777588,
+      "learning_rate": 1.58877668423065e-05,
+      "loss": 1.016,
+      "step": 5591
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.711720585823059,
+      "learning_rate": 1.5886265241212684e-05,
+      "loss": 0.9937,
+      "step": 5592
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6773245334625244,
+      "learning_rate": 1.588476343700029e-05,
+      "loss": 1.0657,
+      "step": 5593
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9883131980895996,
+      "learning_rate": 1.588326142972115e-05,
+      "loss": 1.0662,
+      "step": 5594
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8492990732192993,
+      "learning_rate": 1.5881759219427092e-05,
+      "loss": 1.0012,
+      "step": 5595
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.649088978767395,
+      "learning_rate": 1.5880256806169954e-05,
+      "loss": 0.9264,
+      "step": 5596
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0660146474838257,
+      "learning_rate": 1.587875419000158e-05,
+      "loss": 0.6128,
+      "step": 5597
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.83261239528656,
+      "learning_rate": 1.587725137097382e-05,
+      "loss": 1.0174,
+      "step": 5598
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8204060792922974,
+      "learning_rate": 1.5875748349138533e-05,
+      "loss": 0.9952,
+      "step": 5599
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7471671104431152,
+      "learning_rate": 1.5874245124547583e-05,
+      "loss": 0.967,
+      "step": 5600
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7384865283966064,
+      "learning_rate": 1.5872741697252843e-05,
+      "loss": 1.0371,
+      "step": 5601
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.831261396408081,
+      "learning_rate": 1.5871238067306196e-05,
+      "loss": 1.0413,
+      "step": 5602
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9228148460388184,
+      "learning_rate": 1.5869734234759516e-05,
+      "loss": 1.0485,
+      "step": 5603
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6715623140335083,
+      "learning_rate": 1.586823019966471e-05,
+      "loss": 0.9621,
+      "step": 5604
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.996870517730713,
+      "learning_rate": 1.586672596207367e-05,
+      "loss": 1.0174,
+      "step": 5605
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8512481451034546,
+      "learning_rate": 1.5865221522038304e-05,
+      "loss": 0.9463,
+      "step": 5606
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.84958016872406,
+      "learning_rate": 1.5863716879610528e-05,
+      "loss": 0.982,
+      "step": 5607
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.824703335762024,
+      "learning_rate": 1.5862212034842265e-05,
+      "loss": 0.9914,
+      "step": 5608
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1317521333694458,
+      "learning_rate": 1.5860706987785437e-05,
+      "loss": 0.6052,
+      "step": 5609
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2002053260803223,
+      "learning_rate": 1.5859201738491982e-05,
+      "loss": 1.0076,
+      "step": 5610
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8917196989059448,
+      "learning_rate": 1.5857696287013843e-05,
+      "loss": 0.9436,
+      "step": 5611
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8004605770111084,
+      "learning_rate": 1.585619063340297e-05,
+      "loss": 0.9872,
+      "step": 5612
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8486406803131104,
+      "learning_rate": 1.5854684777711312e-05,
+      "loss": 1.0223,
+      "step": 5613
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.403108835220337,
+      "learning_rate": 1.5853178719990842e-05,
+      "loss": 0.9588,
+      "step": 5614
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8655530214309692,
+      "learning_rate": 1.585167246029352e-05,
+      "loss": 1.0219,
+      "step": 5615
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7775142192840576,
+      "learning_rate": 1.585016599867133e-05,
+      "loss": 0.9926,
+      "step": 5616
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8702837228775024,
+      "learning_rate": 1.584865933517625e-05,
+      "loss": 1.0295,
+      "step": 5617
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.155032157897949,
+      "learning_rate": 1.5847152469860277e-05,
+      "loss": 1.0145,
+      "step": 5618
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1903616189956665,
+      "learning_rate": 1.5845645402775404e-05,
+      "loss": 0.5977,
+      "step": 5619
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.033189535140991,
+      "learning_rate": 1.584413813397364e-05,
+      "loss": 1.0135,
+      "step": 5620
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.007765054702759,
+      "learning_rate": 1.584263066350699e-05,
+      "loss": 1.105,
+      "step": 5621
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7189234495162964,
+      "learning_rate": 1.584112299142748e-05,
+      "loss": 0.9226,
+      "step": 5622
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0227088928222656,
+      "learning_rate": 1.5839615117787132e-05,
+      "loss": 1.0538,
+      "step": 5623
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1617214679718018,
+      "learning_rate": 1.5838107042637974e-05,
+      "loss": 1.0109,
+      "step": 5624
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9102849960327148,
+      "learning_rate": 1.5836598766032055e-05,
+      "loss": 1.0907,
+      "step": 5625
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0651750564575195,
+      "learning_rate": 1.5835090288021414e-05,
+      "loss": 1.097,
+      "step": 5626
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8451707363128662,
+      "learning_rate": 1.5833581608658108e-05,
+      "loss": 1.0849,
+      "step": 5627
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7253296375274658,
+      "learning_rate": 1.5832072727994193e-05,
+      "loss": 1.0893,
+      "step": 5628
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7815706729888916,
+      "learning_rate": 1.5830563646081746e-05,
+      "loss": 1.0086,
+      "step": 5629
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.816599726676941,
+      "learning_rate": 1.582905436297283e-05,
+      "loss": 1.0314,
+      "step": 5630
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7699759006500244,
+      "learning_rate": 1.5827544878719532e-05,
+      "loss": 1.0011,
+      "step": 5631
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9050698280334473,
+      "learning_rate": 1.5826035193373935e-05,
+      "loss": 1.0166,
+      "step": 5632
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8246219158172607,
+      "learning_rate": 1.5824525306988144e-05,
+      "loss": 0.9555,
+      "step": 5633
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8009560108184814,
+      "learning_rate": 1.582301521961425e-05,
+      "loss": 1.1471,
+      "step": 5634
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0377798080444336,
+      "learning_rate": 1.582150493130437e-05,
+      "loss": 1.066,
+      "step": 5635
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.820115327835083,
+      "learning_rate": 1.5819994442110617e-05,
+      "loss": 1.0461,
+      "step": 5636
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.898526668548584,
+      "learning_rate": 1.581848375208511e-05,
+      "loss": 0.9808,
+      "step": 5637
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8746848106384277,
+      "learning_rate": 1.5816972861279985e-05,
+      "loss": 1.0011,
+      "step": 5638
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7512683868408203,
+      "learning_rate": 1.5815461769747372e-05,
+      "loss": 1.0096,
+      "step": 5639
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8255621194839478,
+      "learning_rate": 1.581395047753942e-05,
+      "loss": 1.0531,
+      "step": 5640
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.998863697052002,
+      "learning_rate": 1.581243898470828e-05,
+      "loss": 1.0372,
+      "step": 5641
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.783408761024475,
+      "learning_rate": 1.58109272913061e-05,
+      "loss": 1.0707,
+      "step": 5642
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.94863760471344,
+      "learning_rate": 1.580941539738506e-05,
+      "loss": 1.0407,
+      "step": 5643
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.849313735961914,
+      "learning_rate": 1.580790330299732e-05,
+      "loss": 1.0509,
+      "step": 5644
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3469746112823486,
+      "learning_rate": 1.5806391008195058e-05,
+      "loss": 1.0892,
+      "step": 5645
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.767053246498108,
+      "learning_rate": 1.5804878513030463e-05,
+      "loss": 0.9959,
+      "step": 5646
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2576100826263428,
+      "learning_rate": 1.5803365817555726e-05,
+      "loss": 0.6746,
+      "step": 5647
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7269951105117798,
+      "learning_rate": 1.5801852921823047e-05,
+      "loss": 0.9814,
+      "step": 5648
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8020143508911133,
+      "learning_rate": 1.580033982588463e-05,
+      "loss": 1.0421,
+      "step": 5649
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7122759819030762,
+      "learning_rate": 1.5798826529792684e-05,
+      "loss": 1.0338,
+      "step": 5650
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.742413878440857,
+      "learning_rate": 1.579731303359944e-05,
+      "loss": 1.0062,
+      "step": 5651
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7944848537445068,
+      "learning_rate": 1.5795799337357115e-05,
+      "loss": 0.9278,
+      "step": 5652
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.805260419845581,
+      "learning_rate": 1.579428544111794e-05,
+      "loss": 1.0102,
+      "step": 5653
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5833542346954346,
+      "learning_rate": 1.5792771344934167e-05,
+      "loss": 1.0237,
+      "step": 5654
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7975929975509644,
+      "learning_rate": 1.579125704885803e-05,
+      "loss": 0.9802,
+      "step": 5655
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6812245845794678,
+      "learning_rate": 1.5789742552941794e-05,
+      "loss": 0.91,
+      "step": 5656
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7998672723770142,
+      "learning_rate": 1.5788227857237715e-05,
+      "loss": 1.0114,
+      "step": 5657
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8616605997085571,
+      "learning_rate": 1.578671296179806e-05,
+      "loss": 1.007,
+      "step": 5658
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7945549488067627,
+      "learning_rate": 1.5785197866675107e-05,
+      "loss": 1.0494,
+      "step": 5659
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8168004751205444,
+      "learning_rate": 1.5783682571921132e-05,
+      "loss": 0.9598,
+      "step": 5660
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.83732008934021,
+      "learning_rate": 1.578216707758843e-05,
+      "loss": 0.9933,
+      "step": 5661
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0592806339263916,
+      "learning_rate": 1.5780651383729292e-05,
+      "loss": 1.0096,
+      "step": 5662
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3441640138626099,
+      "learning_rate": 1.5779135490396025e-05,
+      "loss": 0.6417,
+      "step": 5663
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0483028888702393,
+      "learning_rate": 1.5777619397640937e-05,
+      "loss": 1.0512,
+      "step": 5664
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7721612453460693,
+      "learning_rate": 1.577610310551634e-05,
+      "loss": 0.9649,
+      "step": 5665
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7482903003692627,
+      "learning_rate": 1.577458661407456e-05,
+      "loss": 0.9648,
+      "step": 5666
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9614330530166626,
+      "learning_rate": 1.5773069923367927e-05,
+      "loss": 1.0459,
+      "step": 5667
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7487666606903076,
+      "learning_rate": 1.5771553033448777e-05,
+      "loss": 1.0002,
+      "step": 5668
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8344194889068604,
+      "learning_rate": 1.5770035944369456e-05,
+      "loss": 1.037,
+      "step": 5669
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8469412326812744,
+      "learning_rate": 1.576851865618231e-05,
+      "loss": 0.9667,
+      "step": 5670
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6935985088348389,
+      "learning_rate": 1.57670011689397e-05,
+      "loss": 0.9824,
+      "step": 5671
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.026711106300354,
+      "learning_rate": 1.5765483482693987e-05,
+      "loss": 0.5761,
+      "step": 5672
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9349530935287476,
+      "learning_rate": 1.5763965597497547e-05,
+      "loss": 1.0748,
+      "step": 5673
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8719233274459839,
+      "learning_rate": 1.5762447513402755e-05,
+      "loss": 0.9511,
+      "step": 5674
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9921724796295166,
+      "learning_rate": 1.5760929230461994e-05,
+      "loss": 1.0241,
+      "step": 5675
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8035528659820557,
+      "learning_rate": 1.5759410748727663e-05,
+      "loss": 1.0403,
+      "step": 5676
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9983935356140137,
+      "learning_rate": 1.5757892068252148e-05,
+      "loss": 0.6209,
+      "step": 5677
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3461449146270752,
+      "learning_rate": 1.5756373189087864e-05,
+      "loss": 0.6136,
+      "step": 5678
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7944227457046509,
+      "learning_rate": 1.5754854111287222e-05,
+      "loss": 1.0384,
+      "step": 5679
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3363876342773438,
+      "learning_rate": 1.5753334834902643e-05,
+      "loss": 1.1349,
+      "step": 5680
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8927228450775146,
+      "learning_rate": 1.5751815359986548e-05,
+      "loss": 1.0747,
+      "step": 5681
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9742066860198975,
+      "learning_rate": 1.5750295686591372e-05,
+      "loss": 1.0283,
+      "step": 5682
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8273694515228271,
+      "learning_rate": 1.5748775814769553e-05,
+      "loss": 0.955,
+      "step": 5683
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.903059482574463,
+      "learning_rate": 1.5747255744573542e-05,
+      "loss": 1.0214,
+      "step": 5684
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3763656616210938,
+      "learning_rate": 1.574573547605579e-05,
+      "loss": 1.0977,
+      "step": 5685
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1012463569641113,
+      "learning_rate": 1.574421500926875e-05,
+      "loss": 1.0923,
+      "step": 5686
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0620412826538086,
+      "learning_rate": 1.57426943442649e-05,
+      "loss": 0.9615,
+      "step": 5687
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.822073221206665,
+      "learning_rate": 1.5741173481096713e-05,
+      "loss": 1.0159,
+      "step": 5688
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6255770921707153,
+      "learning_rate": 1.573965241981666e-05,
+      "loss": 0.9443,
+      "step": 5689
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1343984603881836,
+      "learning_rate": 1.5738131160477242e-05,
+      "loss": 0.6628,
+      "step": 5690
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7767374515533447,
+      "learning_rate": 1.5736609703130942e-05,
+      "loss": 1.013,
+      "step": 5691
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6196361780166626,
+      "learning_rate": 1.573508804783027e-05,
+      "loss": 0.9825,
+      "step": 5692
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9072165489196777,
+      "learning_rate": 1.5733566194627722e-05,
+      "loss": 0.9395,
+      "step": 5693
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7007228136062622,
+      "learning_rate": 1.5732044143575827e-05,
+      "loss": 0.9975,
+      "step": 5694
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7965205907821655,
+      "learning_rate": 1.5730521894727098e-05,
+      "loss": 1.0064,
+      "step": 5695
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8211930990219116,
+      "learning_rate": 1.572899944813407e-05,
+      "loss": 0.9962,
+      "step": 5696
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8731894493103027,
+      "learning_rate": 1.572747680384927e-05,
+      "loss": 1.0479,
+      "step": 5697
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6269744634628296,
+      "learning_rate": 1.5725953961925245e-05,
+      "loss": 1.0081,
+      "step": 5698
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.949883222579956,
+      "learning_rate": 1.5724430922414543e-05,
+      "loss": 1.074,
+      "step": 5699
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6587811708450317,
+      "learning_rate": 1.5722907685369724e-05,
+      "loss": 1.0096,
+      "step": 5700
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0263965129852295,
+      "learning_rate": 1.5721384250843343e-05,
+      "loss": 1.0261,
+      "step": 5701
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9395532608032227,
+      "learning_rate": 1.5719860618887976e-05,
+      "loss": 1.063,
+      "step": 5702
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8952118158340454,
+      "learning_rate": 1.5718336789556195e-05,
+      "loss": 0.9465,
+      "step": 5703
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6189630031585693,
+      "learning_rate": 1.5716812762900588e-05,
+      "loss": 1.0284,
+      "step": 5704
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.797673225402832,
+      "learning_rate": 1.571528853897374e-05,
+      "loss": 0.9073,
+      "step": 5705
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8497487306594849,
+      "learning_rate": 1.5713764117828253e-05,
+      "loss": 0.9876,
+      "step": 5706
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8215322494506836,
+      "learning_rate": 1.571223949951672e-05,
+      "loss": 0.9146,
+      "step": 5707
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7548412084579468,
+      "learning_rate": 1.5710714684091764e-05,
+      "loss": 1.0178,
+      "step": 5708
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8014018535614014,
+      "learning_rate": 1.5709189671605992e-05,
+      "loss": 1.0698,
+      "step": 5709
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.776074767112732,
+      "learning_rate": 1.5707664462112035e-05,
+      "loss": 1.0065,
+      "step": 5710
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6989144086837769,
+      "learning_rate": 1.570613905566252e-05,
+      "loss": 0.9914,
+      "step": 5711
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8957029581069946,
+      "learning_rate": 1.570461345231009e-05,
+      "loss": 1.0267,
+      "step": 5712
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8715628385543823,
+      "learning_rate": 1.570308765210738e-05,
+      "loss": 1.0299,
+      "step": 5713
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.807898998260498,
+      "learning_rate": 1.5701561655107047e-05,
+      "loss": 1.0139,
+      "step": 5714
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6347683668136597,
+      "learning_rate": 1.5700035461361748e-05,
+      "loss": 0.9268,
+      "step": 5715
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.837496042251587,
+      "learning_rate": 1.569850907092415e-05,
+      "loss": 1.0004,
+      "step": 5716
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.831099033355713,
+      "learning_rate": 1.569698248384692e-05,
+      "loss": 0.9972,
+      "step": 5717
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9904072284698486,
+      "learning_rate": 1.569545570018274e-05,
+      "loss": 1.0602,
+      "step": 5718
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6455543041229248,
+      "learning_rate": 1.5693928719984292e-05,
+      "loss": 0.9568,
+      "step": 5719
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7781951427459717,
+      "learning_rate": 1.569240154330427e-05,
+      "loss": 1.0427,
+      "step": 5720
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8802788257598877,
+      "learning_rate": 1.5690874170195368e-05,
+      "loss": 1.0397,
+      "step": 5721
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8736258745193481,
+      "learning_rate": 1.56893466007103e-05,
+      "loss": 1.0137,
+      "step": 5722
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.831794261932373,
+      "learning_rate": 1.568781883490177e-05,
+      "loss": 0.9285,
+      "step": 5723
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.732837200164795,
+      "learning_rate": 1.5686290872822504e-05,
+      "loss": 0.9913,
+      "step": 5724
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8350539207458496,
+      "learning_rate": 1.5684762714525222e-05,
+      "loss": 1.0,
+      "step": 5725
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.713833212852478,
+      "learning_rate": 1.568323436006266e-05,
+      "loss": 1.032,
+      "step": 5726
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7734463214874268,
+      "learning_rate": 1.5681705809487554e-05,
+      "loss": 0.9892,
+      "step": 5727
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8690985441207886,
+      "learning_rate": 1.568017706285265e-05,
+      "loss": 1.0635,
+      "step": 5728
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6964085102081299,
+      "learning_rate": 1.5678648120210703e-05,
+      "loss": 0.9062,
+      "step": 5729
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8863768577575684,
+      "learning_rate": 1.5677118981614477e-05,
+      "loss": 1.0252,
+      "step": 5730
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8796589374542236,
+      "learning_rate": 1.567558964711673e-05,
+      "loss": 0.9903,
+      "step": 5731
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.273630142211914,
+      "learning_rate": 1.5674060116770234e-05,
+      "loss": 1.1655,
+      "step": 5732
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.728611707687378,
+      "learning_rate": 1.567253039062778e-05,
+      "loss": 1.0736,
+      "step": 5733
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8526363372802734,
+      "learning_rate": 1.5671000468742144e-05,
+      "loss": 1.0292,
+      "step": 5734
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.833158016204834,
+      "learning_rate": 1.5669470351166125e-05,
+      "loss": 1.0352,
+      "step": 5735
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8526487350463867,
+      "learning_rate": 1.566794003795252e-05,
+      "loss": 1.0781,
+      "step": 5736
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.829857349395752,
+      "learning_rate": 1.5666409529154138e-05,
+      "loss": 1.0385,
+      "step": 5737
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7702175378799438,
+      "learning_rate": 1.5664878824823794e-05,
+      "loss": 1.1065,
+      "step": 5738
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8521746397018433,
+      "learning_rate": 1.5663347925014302e-05,
+      "loss": 0.9984,
+      "step": 5739
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.819317102432251,
+      "learning_rate": 1.5661816829778493e-05,
+      "loss": 0.9998,
+      "step": 5740
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8256032466888428,
+      "learning_rate": 1.5660285539169202e-05,
+      "loss": 1.0779,
+      "step": 5741
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.978366494178772,
+      "learning_rate": 1.5658754053239267e-05,
+      "loss": 1.0641,
+      "step": 5742
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0026516914367676,
+      "learning_rate": 1.565722237204154e-05,
+      "loss": 1.0453,
+      "step": 5743
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9630029201507568,
+      "learning_rate": 1.5655690495628867e-05,
+      "loss": 1.0341,
+      "step": 5744
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0062742233276367,
+      "learning_rate": 1.565415842405412e-05,
+      "loss": 1.0353,
+      "step": 5745
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7329479455947876,
+      "learning_rate": 1.5652626157370154e-05,
+      "loss": 1.0637,
+      "step": 5746
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.720922827720642,
+      "learning_rate": 1.5651093695629854e-05,
+      "loss": 1.009,
+      "step": 5747
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8163689374923706,
+      "learning_rate": 1.5649561038886093e-05,
+      "loss": 1.0117,
+      "step": 5748
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.862666368484497,
+      "learning_rate": 1.5648028187191764e-05,
+      "loss": 1.0208,
+      "step": 5749
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8742403984069824,
+      "learning_rate": 1.5646495140599758e-05,
+      "loss": 1.0665,
+      "step": 5750
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.866719365119934,
+      "learning_rate": 1.5644961899162977e-05,
+      "loss": 1.0787,
+      "step": 5751
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8796030282974243,
+      "learning_rate": 1.564342846293433e-05,
+      "loss": 0.959,
+      "step": 5752
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8757416009902954,
+      "learning_rate": 1.5641894831966732e-05,
+      "loss": 1.0584,
+      "step": 5753
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7255322933197021,
+      "learning_rate": 1.5640361006313103e-05,
+      "loss": 1.079,
+      "step": 5754
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7396575212478638,
+      "learning_rate": 1.5638826986026373e-05,
+      "loss": 1.0705,
+      "step": 5755
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8356553316116333,
+      "learning_rate": 1.563729277115947e-05,
+      "loss": 1.0386,
+      "step": 5756
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.698919653892517,
+      "learning_rate": 1.5635758361765345e-05,
+      "loss": 0.989,
+      "step": 5757
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8141980171203613,
+      "learning_rate": 1.5634223757896943e-05,
+      "loss": 0.9852,
+      "step": 5758
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9051424264907837,
+      "learning_rate": 1.563268895960721e-05,
+      "loss": 0.9707,
+      "step": 5759
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9231992959976196,
+      "learning_rate": 1.5631153966949125e-05,
+      "loss": 1.0113,
+      "step": 5760
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7726978063583374,
+      "learning_rate": 1.562961877997564e-05,
+      "loss": 0.969,
+      "step": 5761
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.743573546409607,
+      "learning_rate": 1.562808339873974e-05,
+      "loss": 0.9293,
+      "step": 5762
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8523484468460083,
+      "learning_rate": 1.56265478232944e-05,
+      "loss": 1.0762,
+      "step": 5763
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9541845321655273,
+      "learning_rate": 1.5625012053692615e-05,
+      "loss": 0.9897,
+      "step": 5764
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.692139744758606,
+      "learning_rate": 1.5623476089987376e-05,
+      "loss": 0.9145,
+      "step": 5765
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.643104910850525,
+      "learning_rate": 1.5621939932231685e-05,
+      "loss": 0.8562,
+      "step": 5766
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.732215404510498,
+      "learning_rate": 1.5620403580478552e-05,
+      "loss": 0.9352,
+      "step": 5767
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6711032390594482,
+      "learning_rate": 1.561886703478099e-05,
+      "loss": 1.0885,
+      "step": 5768
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7550568580627441,
+      "learning_rate": 1.5617330295192025e-05,
+      "loss": 1.0372,
+      "step": 5769
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9764037132263184,
+      "learning_rate": 1.561579336176468e-05,
+      "loss": 1.011,
+      "step": 5770
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9520660638809204,
+      "learning_rate": 1.5614256234551995e-05,
+      "loss": 0.9581,
+      "step": 5771
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8582338094711304,
+      "learning_rate": 1.561271891360701e-05,
+      "loss": 0.9808,
+      "step": 5772
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9764502048492432,
+      "learning_rate": 1.561118139898277e-05,
+      "loss": 1.0704,
+      "step": 5773
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.891813039779663,
+      "learning_rate": 1.5609643690732337e-05,
+      "loss": 0.9736,
+      "step": 5774
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.73569917678833,
+      "learning_rate": 1.560810578890877e-05,
+      "loss": 1.0572,
+      "step": 5775
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7014878988265991,
+      "learning_rate": 1.5606567693565143e-05,
+      "loss": 1.0153,
+      "step": 5776
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.748089075088501,
+      "learning_rate": 1.560502940475452e-05,
+      "loss": 1.0455,
+      "step": 5777
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8214471340179443,
+      "learning_rate": 1.560349092252999e-05,
+      "loss": 1.0017,
+      "step": 5778
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6725749969482422,
+      "learning_rate": 1.5601952246944642e-05,
+      "loss": 0.9291,
+      "step": 5779
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0799410343170166,
+      "learning_rate": 1.560041337805157e-05,
+      "loss": 1.0551,
+      "step": 5780
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1333547830581665,
+      "learning_rate": 1.5598874315903878e-05,
+      "loss": 0.6249,
+      "step": 5781
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.76744544506073,
+      "learning_rate": 1.5597335060554673e-05,
+      "loss": 1.0528,
+      "step": 5782
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.71015465259552,
+      "learning_rate": 1.5595795612057067e-05,
+      "loss": 1.0575,
+      "step": 5783
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7831852436065674,
+      "learning_rate": 1.5594255970464192e-05,
+      "loss": 1.021,
+      "step": 5784
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.828924536705017,
+      "learning_rate": 1.5592716135829164e-05,
+      "loss": 0.9835,
+      "step": 5785
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7111735343933105,
+      "learning_rate": 1.5591176108205127e-05,
+      "loss": 1.0607,
+      "step": 5786
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7572084665298462,
+      "learning_rate": 1.558963588764522e-05,
+      "loss": 0.9941,
+      "step": 5787
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7194637060165405,
+      "learning_rate": 1.5588095474202597e-05,
+      "loss": 0.9593,
+      "step": 5788
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8021361827850342,
+      "learning_rate": 1.5586554867930404e-05,
+      "loss": 1.0019,
+      "step": 5789
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8215588331222534,
+      "learning_rate": 1.558501406888181e-05,
+      "loss": 1.0669,
+      "step": 5790
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8258445262908936,
+      "learning_rate": 1.558347307710998e-05,
+      "loss": 1.0771,
+      "step": 5791
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9110820293426514,
+      "learning_rate": 1.5581931892668093e-05,
+      "loss": 1.0837,
+      "step": 5792
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0173819065093994,
+      "learning_rate": 1.5580390515609325e-05,
+      "loss": 1.0265,
+      "step": 5793
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7407664060592651,
+      "learning_rate": 1.5578848945986872e-05,
+      "loss": 1.0047,
+      "step": 5794
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7931828498840332,
+      "learning_rate": 1.5577307183853925e-05,
+      "loss": 0.9203,
+      "step": 5795
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8555265665054321,
+      "learning_rate": 1.5575765229263686e-05,
+      "loss": 1.0104,
+      "step": 5796
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7339000701904297,
+      "learning_rate": 1.5574223082269366e-05,
+      "loss": 1.0278,
+      "step": 5797
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.973382592201233,
+      "learning_rate": 1.5572680742924178e-05,
+      "loss": 1.0666,
+      "step": 5798
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7116643190383911,
+      "learning_rate": 1.557113821128134e-05,
+      "loss": 0.995,
+      "step": 5799
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3167226314544678,
+      "learning_rate": 1.556959548739409e-05,
+      "loss": 1.0367,
+      "step": 5800
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.818730115890503,
+      "learning_rate": 1.556805257131566e-05,
+      "loss": 0.9967,
+      "step": 5801
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8402304649353027,
+      "learning_rate": 1.556650946309928e-05,
+      "loss": 0.9614,
+      "step": 5802
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8830138444900513,
+      "learning_rate": 1.5564966162798216e-05,
+      "loss": 0.9712,
+      "step": 5803
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8251796960830688,
+      "learning_rate": 1.556342267046571e-05,
+      "loss": 1.0023,
+      "step": 5804
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0738155841827393,
+      "learning_rate": 1.5561878986155033e-05,
+      "loss": 1.1168,
+      "step": 5805
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.943463087081909,
+      "learning_rate": 1.5560335109919445e-05,
+      "loss": 0.9982,
+      "step": 5806
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7479894161224365,
+      "learning_rate": 1.5558791041812226e-05,
+      "loss": 0.9737,
+      "step": 5807
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7748967409133911,
+      "learning_rate": 1.5557246781886657e-05,
+      "loss": 0.9924,
+      "step": 5808
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8681551218032837,
+      "learning_rate": 1.5555702330196024e-05,
+      "loss": 1.0555,
+      "step": 5809
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7283658981323242,
+      "learning_rate": 1.5554157686793623e-05,
+      "loss": 1.0104,
+      "step": 5810
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9318522214889526,
+      "learning_rate": 1.5552612851732757e-05,
+      "loss": 1.0521,
+      "step": 5811
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8806594610214233,
+      "learning_rate": 1.5551067825066727e-05,
+      "loss": 0.9614,
+      "step": 5812
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.769594669342041,
+      "learning_rate": 1.5549522606848855e-05,
+      "loss": 1.057,
+      "step": 5813
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7724529504776,
+      "learning_rate": 1.5547977197132463e-05,
+      "loss": 0.9933,
+      "step": 5814
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2300639152526855,
+      "learning_rate": 1.5546431595970873e-05,
+      "loss": 1.0198,
+      "step": 5815
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9044530391693115,
+      "learning_rate": 1.554488580341742e-05,
+      "loss": 1.1298,
+      "step": 5816
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7599481344223022,
+      "learning_rate": 1.5543339819525448e-05,
+      "loss": 1.0533,
+      "step": 5817
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8673388957977295,
+      "learning_rate": 1.5541793644348305e-05,
+      "loss": 1.0476,
+      "step": 5818
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.710741400718689,
+      "learning_rate": 1.5540247277939343e-05,
+      "loss": 1.0187,
+      "step": 5819
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7325483560562134,
+      "learning_rate": 1.5538700720351924e-05,
+      "loss": 1.0555,
+      "step": 5820
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7160881757736206,
+      "learning_rate": 1.5537153971639414e-05,
+      "loss": 0.9907,
+      "step": 5821
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1652578115463257,
+      "learning_rate": 1.5535607031855188e-05,
+      "loss": 0.646,
+      "step": 5822
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6653498411178589,
+      "learning_rate": 1.5534059901052628e-05,
+      "loss": 0.8851,
+      "step": 5823
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.701088547706604,
+      "learning_rate": 1.5532512579285118e-05,
+      "loss": 1.014,
+      "step": 5824
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0857338905334473,
+      "learning_rate": 1.5530965066606055e-05,
+      "loss": 0.5671,
+      "step": 5825
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.871946096420288,
+      "learning_rate": 1.5529417363068832e-05,
+      "loss": 0.9477,
+      "step": 5826
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.908758282661438,
+      "learning_rate": 1.5527869468726867e-05,
+      "loss": 0.9356,
+      "step": 5827
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8757386207580566,
+      "learning_rate": 1.552632138363357e-05,
+      "loss": 1.0541,
+      "step": 5828
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8711673021316528,
+      "learning_rate": 1.5524773107842355e-05,
+      "loss": 1.007,
+      "step": 5829
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8330109119415283,
+      "learning_rate": 1.5523224641406653e-05,
+      "loss": 0.9845,
+      "step": 5830
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7495567798614502,
+      "learning_rate": 1.5521675984379898e-05,
+      "loss": 1.0381,
+      "step": 5831
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7670490741729736,
+      "learning_rate": 1.552012713681553e-05,
+      "loss": 1.0639,
+      "step": 5832
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.957664132118225,
+      "learning_rate": 1.5518578098766993e-05,
+      "loss": 0.9583,
+      "step": 5833
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9304425716400146,
+      "learning_rate": 1.5517028870287743e-05,
+      "loss": 1.019,
+      "step": 5834
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7708600759506226,
+      "learning_rate": 1.5515479451431237e-05,
+      "loss": 0.9029,
+      "step": 5835
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8742115497589111,
+      "learning_rate": 1.551392984225094e-05,
+      "loss": 1.0259,
+      "step": 5836
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6818162202835083,
+      "learning_rate": 1.551238004280033e-05,
+      "loss": 1.0079,
+      "step": 5837
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6914740800857544,
+      "learning_rate": 1.5510830053132882e-05,
+      "loss": 1.0007,
+      "step": 5838
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7664865255355835,
+      "learning_rate": 1.550927987330208e-05,
+      "loss": 0.9841,
+      "step": 5839
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6322523355484009,
+      "learning_rate": 1.550772950336142e-05,
+      "loss": 1.0121,
+      "step": 5840
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6135891675949097,
+      "learning_rate": 1.5506178943364406e-05,
+      "loss": 1.02,
+      "step": 5841
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9580302238464355,
+      "learning_rate": 1.550462819336453e-05,
+      "loss": 1.0152,
+      "step": 5842
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9442603588104248,
+      "learning_rate": 1.5503077253415315e-05,
+      "loss": 1.0498,
+      "step": 5843
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7674839496612549,
+      "learning_rate": 1.5501526123570277e-05,
+      "loss": 1.0136,
+      "step": 5844
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8188444375991821,
+      "learning_rate": 1.549997480388294e-05,
+      "loss": 0.9813,
+      "step": 5845
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9018720388412476,
+      "learning_rate": 1.5498423294406833e-05,
+      "loss": 1.0025,
+      "step": 5846
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9595156908035278,
+      "learning_rate": 1.54968715951955e-05,
+      "loss": 1.0294,
+      "step": 5847
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9530587196350098,
+      "learning_rate": 1.5495319706302485e-05,
+      "loss": 0.9885,
+      "step": 5848
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.96371591091156,
+      "learning_rate": 1.5493767627781332e-05,
+      "loss": 1.0414,
+      "step": 5849
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.839888334274292,
+      "learning_rate": 1.549221535968561e-05,
+      "loss": 1.0287,
+      "step": 5850
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7183386087417603,
+      "learning_rate": 1.5490662902068872e-05,
+      "loss": 1.0059,
+      "step": 5851
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8995938301086426,
+      "learning_rate": 1.54891102549847e-05,
+      "loss": 1.0452,
+      "step": 5852
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6958467960357666,
+      "learning_rate": 1.5487557418486666e-05,
+      "loss": 1.03,
+      "step": 5853
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.855422854423523,
+      "learning_rate": 1.548600439262835e-05,
+      "loss": 0.9593,
+      "step": 5854
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.736997127532959,
+      "learning_rate": 1.548445117746335e-05,
+      "loss": 0.9598,
+      "step": 5855
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8337535858154297,
+      "learning_rate": 1.5482897773045262e-05,
+      "loss": 1.0499,
+      "step": 5856
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7215017080307007,
+      "learning_rate": 1.5481344179427688e-05,
+      "loss": 0.974,
+      "step": 5857
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7477277517318726,
+      "learning_rate": 1.5479790396664235e-05,
+      "loss": 1.0548,
+      "step": 5858
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9066221714019775,
+      "learning_rate": 1.547823642480852e-05,
+      "loss": 0.9922,
+      "step": 5859
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.794390320777893,
+      "learning_rate": 1.547668226391417e-05,
+      "loss": 1.0412,
+      "step": 5860
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.836539387702942,
+      "learning_rate": 1.5475127914034816e-05,
+      "loss": 0.9699,
+      "step": 5861
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7617093324661255,
+      "learning_rate": 1.5473573375224093e-05,
+      "loss": 1.0219,
+      "step": 5862
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9734299182891846,
+      "learning_rate": 1.5472018647535637e-05,
+      "loss": 1.0659,
+      "step": 5863
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8893712759017944,
+      "learning_rate": 1.5470463731023107e-05,
+      "loss": 0.9184,
+      "step": 5864
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3721625804901123,
+      "learning_rate": 1.5468908625740157e-05,
+      "loss": 1.0514,
+      "step": 5865
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.836548924446106,
+      "learning_rate": 1.5467353331740445e-05,
+      "loss": 0.9108,
+      "step": 5866
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7330526113510132,
+      "learning_rate": 1.5465797849077643e-05,
+      "loss": 1.0146,
+      "step": 5867
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.785692572593689,
+      "learning_rate": 1.546424217780542e-05,
+      "loss": 0.9858,
+      "step": 5868
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9129208326339722,
+      "learning_rate": 1.546268631797747e-05,
+      "loss": 0.9787,
+      "step": 5869
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7087873220443726,
+      "learning_rate": 1.546113026964747e-05,
+      "loss": 0.9836,
+      "step": 5870
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8789410591125488,
+      "learning_rate": 1.5459574032869126e-05,
+      "loss": 1.0437,
+      "step": 5871
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2202550172805786,
+      "learning_rate": 1.5458017607696124e-05,
+      "loss": 0.6633,
+      "step": 5872
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.895402431488037,
+      "learning_rate": 1.5456460994182185e-05,
+      "loss": 1.0153,
+      "step": 5873
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1224522590637207,
+      "learning_rate": 1.545490419238102e-05,
+      "loss": 1.0592,
+      "step": 5874
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.957767128944397,
+      "learning_rate": 1.5453347202346347e-05,
+      "loss": 0.978,
+      "step": 5875
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7957892417907715,
+      "learning_rate": 1.5451790024131897e-05,
+      "loss": 1.0856,
+      "step": 5876
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7588075399398804,
+      "learning_rate": 1.54502326577914e-05,
+      "loss": 0.9343,
+      "step": 5877
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.670821189880371,
+      "learning_rate": 1.54486751033786e-05,
+      "loss": 0.9639,
+      "step": 5878
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8318747282028198,
+      "learning_rate": 1.5447117360947244e-05,
+      "loss": 0.949,
+      "step": 5879
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3086538314819336,
+      "learning_rate": 1.5445559430551083e-05,
+      "loss": 1.0727,
+      "step": 5880
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1196787357330322,
+      "learning_rate": 1.5444001312243876e-05,
+      "loss": 1.0706,
+      "step": 5881
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7794506549835205,
+      "learning_rate": 1.544244300607939e-05,
+      "loss": 0.9646,
+      "step": 5882
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8270851373672485,
+      "learning_rate": 1.54408845121114e-05,
+      "loss": 1.0972,
+      "step": 5883
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6430906057357788,
+      "learning_rate": 1.5439325830393688e-05,
+      "loss": 1.0369,
+      "step": 5884
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8952572345733643,
+      "learning_rate": 1.543776696098003e-05,
+      "loss": 0.933,
+      "step": 5885
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7971152067184448,
+      "learning_rate": 1.5436207903924226e-05,
+      "loss": 1.0009,
+      "step": 5886
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.692742109298706,
+      "learning_rate": 1.5434648659280072e-05,
+      "loss": 0.9657,
+      "step": 5887
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7461479902267456,
+      "learning_rate": 1.5433089227101374e-05,
+      "loss": 1.0204,
+      "step": 5888
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0098862648010254,
+      "learning_rate": 1.5431529607441945e-05,
+      "loss": 1.0327,
+      "step": 5889
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.040712356567383,
+      "learning_rate": 1.5429969800355602e-05,
+      "loss": 1.1232,
+      "step": 5890
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7322931289672852,
+      "learning_rate": 1.5428409805896166e-05,
+      "loss": 0.9897,
+      "step": 5891
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6626914739608765,
+      "learning_rate": 1.5426849624117474e-05,
+      "loss": 0.9986,
+      "step": 5892
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6629664897918701,
+      "learning_rate": 1.542528925507336e-05,
+      "loss": 0.9222,
+      "step": 5893
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.017216205596924,
+      "learning_rate": 1.5423728698817665e-05,
+      "loss": 1.0044,
+      "step": 5894
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.89260733127594,
+      "learning_rate": 1.542216795540425e-05,
+      "loss": 1.0251,
+      "step": 5895
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8202911615371704,
+      "learning_rate": 1.542060702488696e-05,
+      "loss": 1.0319,
+      "step": 5896
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8159570693969727,
+      "learning_rate": 1.5419045907319666e-05,
+      "loss": 1.0769,
+      "step": 5897
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8324774503707886,
+      "learning_rate": 1.5417484602756237e-05,
+      "loss": 1.0931,
+      "step": 5898
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7690989971160889,
+      "learning_rate": 1.5415923111250543e-05,
+      "loss": 0.9444,
+      "step": 5899
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9186985492706299,
+      "learning_rate": 1.5414361432856475e-05,
+      "loss": 1.0817,
+      "step": 5900
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8851598501205444,
+      "learning_rate": 1.5412799567627915e-05,
+      "loss": 1.058,
+      "step": 5901
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.834693431854248,
+      "learning_rate": 1.5411237515618764e-05,
+      "loss": 0.9521,
+      "step": 5902
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7473396062850952,
+      "learning_rate": 1.540967527688292e-05,
+      "loss": 1.0651,
+      "step": 5903
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8309718370437622,
+      "learning_rate": 1.54081128514743e-05,
+      "loss": 1.0007,
+      "step": 5904
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0772550106048584,
+      "learning_rate": 1.5406550239446808e-05,
+      "loss": 1.0377,
+      "step": 5905
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7661163806915283,
+      "learning_rate": 1.5404987440854367e-05,
+      "loss": 1.0147,
+      "step": 5906
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7555384635925293,
+      "learning_rate": 1.540342445575091e-05,
+      "loss": 1.0931,
+      "step": 5907
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.64422607421875,
+      "learning_rate": 1.5401861284190368e-05,
+      "loss": 0.9206,
+      "step": 5908
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7226132154464722,
+      "learning_rate": 1.5400297926226683e-05,
+      "loss": 0.9446,
+      "step": 5909
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9372684955596924,
+      "learning_rate": 1.5398734381913802e-05,
+      "loss": 0.9841,
+      "step": 5910
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8741693496704102,
+      "learning_rate": 1.539717065130568e-05,
+      "loss": 0.9853,
+      "step": 5911
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.889244794845581,
+      "learning_rate": 1.5395606734456273e-05,
+      "loss": 1.0806,
+      "step": 5912
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6520521640777588,
+      "learning_rate": 1.539404263141955e-05,
+      "loss": 0.9568,
+      "step": 5913
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7411062717437744,
+      "learning_rate": 1.5392478342249485e-05,
+      "loss": 0.9141,
+      "step": 5914
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.041492462158203,
+      "learning_rate": 1.5390913867000056e-05,
+      "loss": 0.9167,
+      "step": 5915
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7665635347366333,
+      "learning_rate": 1.5389349205725244e-05,
+      "loss": 0.9472,
+      "step": 5916
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.006621837615967,
+      "learning_rate": 1.5387784358479046e-05,
+      "loss": 0.9455,
+      "step": 5917
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9990004301071167,
+      "learning_rate": 1.5386219325315465e-05,
+      "loss": 1.054,
+      "step": 5918
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6952786445617676,
+      "learning_rate": 1.53846541062885e-05,
+      "loss": 1.007,
+      "step": 5919
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.677554726600647,
+      "learning_rate": 1.538308870145216e-05,
+      "loss": 1.0355,
+      "step": 5920
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9471287727355957,
+      "learning_rate": 1.5381523110860466e-05,
+      "loss": 0.9643,
+      "step": 5921
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.646493911743164,
+      "learning_rate": 1.5379957334567444e-05,
+      "loss": 1.1153,
+      "step": 5922
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8899216651916504,
+      "learning_rate": 1.537839137262712e-05,
+      "loss": 0.9747,
+      "step": 5923
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7572290897369385,
+      "learning_rate": 1.537682522509354e-05,
+      "loss": 1.1002,
+      "step": 5924
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8194000720977783,
+      "learning_rate": 1.5375258892020734e-05,
+      "loss": 1.0451,
+      "step": 5925
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7214579582214355,
+      "learning_rate": 1.5373692373462762e-05,
+      "loss": 1.0656,
+      "step": 5926
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.785620927810669,
+      "learning_rate": 1.5372125669473676e-05,
+      "loss": 0.9336,
+      "step": 5927
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7861748933792114,
+      "learning_rate": 1.537055878010754e-05,
+      "loss": 1.0356,
+      "step": 5928
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9255220890045166,
+      "learning_rate": 1.536899170541842e-05,
+      "loss": 1.0626,
+      "step": 5929
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8526057004928589,
+      "learning_rate": 1.53674244454604e-05,
+      "loss": 1.0259,
+      "step": 5930
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6921361684799194,
+      "learning_rate": 1.536585700028755e-05,
+      "loss": 1.0125,
+      "step": 5931
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1956487894058228,
+      "learning_rate": 1.5364289369953967e-05,
+      "loss": 0.6069,
+      "step": 5932
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7433998584747314,
+      "learning_rate": 1.5362721554513743e-05,
+      "loss": 0.9658,
+      "step": 5933
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7707922458648682,
+      "learning_rate": 1.5361153554020977e-05,
+      "loss": 1.1592,
+      "step": 5934
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.015620708465576,
+      "learning_rate": 1.5359585368529778e-05,
+      "loss": 1.0309,
+      "step": 5935
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8554457426071167,
+      "learning_rate": 1.5358016998094255e-05,
+      "loss": 1.0816,
+      "step": 5936
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7883185148239136,
+      "learning_rate": 1.5356448442768535e-05,
+      "loss": 0.9611,
+      "step": 5937
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8557204008102417,
+      "learning_rate": 1.5354879702606745e-05,
+      "loss": 1.061,
+      "step": 5938
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.801736831665039,
+      "learning_rate": 1.5353310777663014e-05,
+      "loss": 1.0063,
+      "step": 5939
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6246137619018555,
+      "learning_rate": 1.535174166799148e-05,
+      "loss": 1.058,
+      "step": 5940
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.909960389137268,
+      "learning_rate": 1.5350172373646292e-05,
+      "loss": 0.9802,
+      "step": 5941
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1383817195892334,
+      "learning_rate": 1.53486028946816e-05,
+      "loss": 0.6415,
+      "step": 5942
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8180601596832275,
+      "learning_rate": 1.5347033231151562e-05,
+      "loss": 1.0727,
+      "step": 5943
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9397879838943481,
+      "learning_rate": 1.5345463383110345e-05,
+      "loss": 0.9781,
+      "step": 5944
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8618950843811035,
+      "learning_rate": 1.534389335061212e-05,
+      "loss": 1.0772,
+      "step": 5945
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8925445079803467,
+      "learning_rate": 1.534232313371106e-05,
+      "loss": 1.0837,
+      "step": 5946
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.876947045326233,
+      "learning_rate": 1.5340752732461352e-05,
+      "loss": 1.0523,
+      "step": 5947
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7619332075119019,
+      "learning_rate": 1.5339182146917185e-05,
+      "loss": 1.0467,
+      "step": 5948
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6400269269943237,
+      "learning_rate": 1.5337611377132757e-05,
+      "loss": 1.0656,
+      "step": 5949
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7686179876327515,
+      "learning_rate": 1.533604042316227e-05,
+      "loss": 1.101,
+      "step": 5950
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8215978145599365,
+      "learning_rate": 1.5334469285059935e-05,
+      "loss": 1.0201,
+      "step": 5951
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0492578744888306,
+      "learning_rate": 1.533289796287997e-05,
+      "loss": 0.6133,
+      "step": 5952
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6992031335830688,
+      "learning_rate": 1.5331326456676588e-05,
+      "loss": 1.0827,
+      "step": 5953
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8196260929107666,
+      "learning_rate": 1.5329754766504025e-05,
+      "loss": 0.9568,
+      "step": 5954
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8492751121520996,
+      "learning_rate": 1.532818289241651e-05,
+      "loss": 0.9926,
+      "step": 5955
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8144549131393433,
+      "learning_rate": 1.532661083446829e-05,
+      "loss": 1.0395,
+      "step": 5956
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6317473649978638,
+      "learning_rate": 1.532503859271361e-05,
+      "loss": 0.8731,
+      "step": 5957
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9456760883331299,
+      "learning_rate": 1.532346616720672e-05,
+      "loss": 1.0129,
+      "step": 5958
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.693047285079956,
+      "learning_rate": 1.5321893558001884e-05,
+      "loss": 1.0422,
+      "step": 5959
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9552749395370483,
+      "learning_rate": 1.5320320765153367e-05,
+      "loss": 0.5617,
+      "step": 5960
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.862405776977539,
+      "learning_rate": 1.5318747788715445e-05,
+      "loss": 1.0217,
+      "step": 5961
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.856614112854004,
+      "learning_rate": 1.5317174628742387e-05,
+      "loss": 0.9888,
+      "step": 5962
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8407907485961914,
+      "learning_rate": 1.531560128528849e-05,
+      "loss": 1.0287,
+      "step": 5963
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.756582260131836,
+      "learning_rate": 1.5314027758408046e-05,
+      "loss": 1.0105,
+      "step": 5964
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8108060359954834,
+      "learning_rate": 1.531245404815534e-05,
+      "loss": 0.986,
+      "step": 5965
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9277286529541016,
+      "learning_rate": 1.5310880154584693e-05,
+      "loss": 0.97,
+      "step": 5966
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8403834104537964,
+      "learning_rate": 1.5309306077750403e-05,
+      "loss": 0.9523,
+      "step": 5967
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9529494047164917,
+      "learning_rate": 1.530773181770679e-05,
+      "loss": 1.0312,
+      "step": 5968
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7818015813827515,
+      "learning_rate": 1.530615737450818e-05,
+      "loss": 1.0062,
+      "step": 5969
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.015958786010742,
+      "learning_rate": 1.53045827482089e-05,
+      "loss": 1.0113,
+      "step": 5970
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.821475625038147,
+      "learning_rate": 1.5303007938863287e-05,
+      "loss": 1.0062,
+      "step": 5971
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8706809282302856,
+      "learning_rate": 1.5301432946525684e-05,
+      "loss": 1.0421,
+      "step": 5972
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6613487005233765,
+      "learning_rate": 1.5299857771250442e-05,
+      "loss": 0.9531,
+      "step": 5973
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.906445860862732,
+      "learning_rate": 1.529828241309191e-05,
+      "loss": 0.9952,
+      "step": 5974
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9234700202941895,
+      "learning_rate": 1.529670687210445e-05,
+      "loss": 1.0448,
+      "step": 5975
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.779085397720337,
+      "learning_rate": 1.5295131148342432e-05,
+      "loss": 0.9388,
+      "step": 5976
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7602256536483765,
+      "learning_rate": 1.5293555241860235e-05,
+      "loss": 0.9842,
+      "step": 5977
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.092806100845337,
+      "learning_rate": 1.529197915271223e-05,
+      "loss": 1.0015,
+      "step": 5978
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7246185541152954,
+      "learning_rate": 1.5290402880952802e-05,
+      "loss": 0.9688,
+      "step": 5979
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6515874862670898,
+      "learning_rate": 1.5288826426636356e-05,
+      "loss": 1.0068,
+      "step": 5980
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.674303412437439,
+      "learning_rate": 1.5287249789817283e-05,
+      "loss": 1.0054,
+      "step": 5981
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8856505155563354,
+      "learning_rate": 1.5285672970549987e-05,
+      "loss": 0.9598,
+      "step": 5982
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8281220197677612,
+      "learning_rate": 1.528409596888888e-05,
+      "loss": 0.9682,
+      "step": 5983
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.896794080734253,
+      "learning_rate": 1.5282518784888384e-05,
+      "loss": 1.043,
+      "step": 5984
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9264347553253174,
+      "learning_rate": 1.528094141860292e-05,
+      "loss": 1.0371,
+      "step": 5985
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7788622379302979,
+      "learning_rate": 1.527936387008692e-05,
+      "loss": 1.0592,
+      "step": 5986
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0962557792663574,
+      "learning_rate": 1.527778613939482e-05,
+      "loss": 1.1051,
+      "step": 5987
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7764137983322144,
+      "learning_rate": 1.5276208226581062e-05,
+      "loss": 1.0425,
+      "step": 5988
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8246161937713623,
+      "learning_rate": 1.5274630131700098e-05,
+      "loss": 0.9956,
+      "step": 5989
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7876452207565308,
+      "learning_rate": 1.5273051854806383e-05,
+      "loss": 0.9528,
+      "step": 5990
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9007985591888428,
+      "learning_rate": 1.5271473395954374e-05,
+      "loss": 1.0857,
+      "step": 5991
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.715610146522522,
+      "learning_rate": 1.5269894755198548e-05,
+      "loss": 0.9945,
+      "step": 5992
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.752297282218933,
+      "learning_rate": 1.5268315932593373e-05,
+      "loss": 0.9845,
+      "step": 5993
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.76646089553833,
+      "learning_rate": 1.5266736928193333e-05,
+      "loss": 0.9602,
+      "step": 5994
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8575209379196167,
+      "learning_rate": 1.5265157742052914e-05,
+      "loss": 0.8897,
+      "step": 5995
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.895287275314331,
+      "learning_rate": 1.5263578374226607e-05,
+      "loss": 1.0185,
+      "step": 5996
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8003556728363037,
+      "learning_rate": 1.526199882476891e-05,
+      "loss": 0.8937,
+      "step": 5997
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9263912439346313,
+      "learning_rate": 1.526041909373434e-05,
+      "loss": 1.0516,
+      "step": 5998
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8928062915802002,
+      "learning_rate": 1.5258839181177397e-05,
+      "loss": 1.0488,
+      "step": 5999
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8394184112548828,
+      "learning_rate": 1.5257259087152606e-05,
+      "loss": 1.0482,
+      "step": 6000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8424128293991089,
+      "learning_rate": 1.5255678811714489e-05,
+      "loss": 1.0242,
+      "step": 6001
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7534763813018799,
+      "learning_rate": 1.5254098354917575e-05,
+      "loss": 1.0441,
+      "step": 6002
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9388951063156128,
+      "learning_rate": 1.5252517716816404e-05,
+      "loss": 0.9905,
+      "step": 6003
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8896952867507935,
+      "learning_rate": 1.5250936897465521e-05,
+      "loss": 1.0085,
+      "step": 6004
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.785737156867981,
+      "learning_rate": 1.5249355896919473e-05,
+      "loss": 1.0191,
+      "step": 6005
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6935712099075317,
+      "learning_rate": 1.5247774715232817e-05,
+      "loss": 0.9441,
+      "step": 6006
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8514976501464844,
+      "learning_rate": 1.5246193352460112e-05,
+      "loss": 0.9502,
+      "step": 6007
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9389610290527344,
+      "learning_rate": 1.524461180865593e-05,
+      "loss": 0.9978,
+      "step": 6008
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8652697801589966,
+      "learning_rate": 1.5243030083874847e-05,
+      "loss": 1.0548,
+      "step": 6009
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7653182744979858,
+      "learning_rate": 1.5241448178171442e-05,
+      "loss": 0.9906,
+      "step": 6010
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.811454176902771,
+      "learning_rate": 1.52398660916003e-05,
+      "loss": 1.006,
+      "step": 6011
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.865007758140564,
+      "learning_rate": 1.5238283824216015e-05,
+      "loss": 1.0213,
+      "step": 6012
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.127016544342041,
+      "learning_rate": 1.5236701376073188e-05,
+      "loss": 0.6384,
+      "step": 6013
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.676629662513733,
+      "learning_rate": 1.5235118747226425e-05,
+      "loss": 0.9842,
+      "step": 6014
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1022276878356934,
+      "learning_rate": 1.5233535937730337e-05,
+      "loss": 1.082,
+      "step": 6015
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8264867067337036,
+      "learning_rate": 1.5231952947639546e-05,
+      "loss": 1.0443,
+      "step": 6016
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7683968544006348,
+      "learning_rate": 1.5230369777008672e-05,
+      "loss": 1.0689,
+      "step": 6017
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.632810354232788,
+      "learning_rate": 1.5228786425892348e-05,
+      "loss": 0.948,
+      "step": 6018
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8630067110061646,
+      "learning_rate": 1.522720289434521e-05,
+      "loss": 1.0269,
+      "step": 6019
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8933451175689697,
+      "learning_rate": 1.52256191824219e-05,
+      "loss": 0.9956,
+      "step": 6020
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.631516695022583,
+      "learning_rate": 1.5224035290177073e-05,
+      "loss": 0.9119,
+      "step": 6021
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8633639812469482,
+      "learning_rate": 1.5222451217665376e-05,
+      "loss": 1.0632,
+      "step": 6022
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6615869998931885,
+      "learning_rate": 1.522086696494148e-05,
+      "loss": 1.0187,
+      "step": 6023
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8441771268844604,
+      "learning_rate": 1.5219282532060047e-05,
+      "loss": 1.0207,
+      "step": 6024
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7313274145126343,
+      "learning_rate": 1.521769791907575e-05,
+      "loss": 1.095,
+      "step": 6025
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.977745532989502,
+      "learning_rate": 1.5216113126043279e-05,
+      "loss": 1.0391,
+      "step": 6026
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0305464267730713,
+      "learning_rate": 1.5214528153017311e-05,
+      "loss": 1.0074,
+      "step": 6027
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7988184690475464,
+      "learning_rate": 1.5212943000052547e-05,
+      "loss": 1.0014,
+      "step": 6028
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.647784948348999,
+      "learning_rate": 1.5211357667203674e-05,
+      "loss": 1.0294,
+      "step": 6029
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7534931898117065,
+      "learning_rate": 1.5209772154525411e-05,
+      "loss": 0.9793,
+      "step": 6030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.926552176475525,
+      "learning_rate": 1.5208186462072463e-05,
+      "loss": 1.0769,
+      "step": 6031
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8396424055099487,
+      "learning_rate": 1.520660058989955e-05,
+      "loss": 1.0784,
+      "step": 6032
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7301712036132812,
+      "learning_rate": 1.5205014538061394e-05,
+      "loss": 0.9417,
+      "step": 6033
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8157248497009277,
+      "learning_rate": 1.5203428306612722e-05,
+      "loss": 0.9629,
+      "step": 6034
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7172398567199707,
+      "learning_rate": 1.520184189560828e-05,
+      "loss": 1.0255,
+      "step": 6035
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0596046447753906,
+      "learning_rate": 1.5200255305102802e-05,
+      "loss": 0.6104,
+      "step": 6036
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.848392367362976,
+      "learning_rate": 1.519866853515104e-05,
+      "loss": 1.0094,
+      "step": 6037
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6890674829483032,
+      "learning_rate": 1.519708158580775e-05,
+      "loss": 0.9585,
+      "step": 6038
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7839975357055664,
+      "learning_rate": 1.519549445712769e-05,
+      "loss": 0.9934,
+      "step": 6039
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.98940110206604,
+      "learning_rate": 1.519390714916563e-05,
+      "loss": 1.0531,
+      "step": 6040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9757086038589478,
+      "learning_rate": 1.519231966197634e-05,
+      "loss": 1.0638,
+      "step": 6041
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8217767477035522,
+      "learning_rate": 1.5190731995614606e-05,
+      "loss": 0.9674,
+      "step": 6042
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0176331996917725,
+      "learning_rate": 1.5189144150135211e-05,
+      "loss": 0.9583,
+      "step": 6043
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9761642217636108,
+      "learning_rate": 1.5187556125592946e-05,
+      "loss": 0.9362,
+      "step": 6044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6965320110321045,
+      "learning_rate": 1.518596792204261e-05,
+      "loss": 1.0941,
+      "step": 6045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5800977945327759,
+      "learning_rate": 1.5184379539539007e-05,
+      "loss": 0.9586,
+      "step": 6046
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.757814645767212,
+      "learning_rate": 1.5182790978136948e-05,
+      "loss": 1.0211,
+      "step": 6047
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.762105941772461,
+      "learning_rate": 1.518120223789125e-05,
+      "loss": 0.9928,
+      "step": 6048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.616483211517334,
+      "learning_rate": 1.5179613318856739e-05,
+      "loss": 0.9047,
+      "step": 6049
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.975540280342102,
+      "learning_rate": 1.5178024221088237e-05,
+      "loss": 0.9895,
+      "step": 6050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.921586513519287,
+      "learning_rate": 1.5176434944640583e-05,
+      "loss": 0.9914,
+      "step": 6051
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0907998085021973,
+      "learning_rate": 1.5174845489568622e-05,
+      "loss": 0.5833,
+      "step": 6052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7525326013565063,
+      "learning_rate": 1.5173255855927194e-05,
+      "loss": 0.9782,
+      "step": 6053
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8479973077774048,
+      "learning_rate": 1.517166604377116e-05,
+      "loss": 1.0146,
+      "step": 6054
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6547385454177856,
+      "learning_rate": 1.5170076053155378e-05,
+      "loss": 0.9334,
+      "step": 6055
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1483852863311768,
+      "learning_rate": 1.5168485884134714e-05,
+      "loss": 0.613,
+      "step": 6056
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7013059854507446,
+      "learning_rate": 1.5166895536764035e-05,
+      "loss": 1.0071,
+      "step": 6057
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9270102977752686,
+      "learning_rate": 1.5165305011098228e-05,
+      "loss": 1.0449,
+      "step": 6058
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.813938856124878,
+      "learning_rate": 1.5163714307192174e-05,
+      "loss": 0.9821,
+      "step": 6059
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.597678542137146,
+      "learning_rate": 1.5162123425100764e-05,
+      "loss": 0.9407,
+      "step": 6060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9122754335403442,
+      "learning_rate": 1.5160532364878892e-05,
+      "loss": 1.094,
+      "step": 6061
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.734326720237732,
+      "learning_rate": 1.5158941126581466e-05,
+      "loss": 1.0234,
+      "step": 6062
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8525351285934448,
+      "learning_rate": 1.5157349710263391e-05,
+      "loss": 0.9608,
+      "step": 6063
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7561818361282349,
+      "learning_rate": 1.5155758115979585e-05,
+      "loss": 0.9605,
+      "step": 6064
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6374677419662476,
+      "learning_rate": 1.515416634378497e-05,
+      "loss": 0.9448,
+      "step": 6065
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7455236911773682,
+      "learning_rate": 1.5152574393734467e-05,
+      "loss": 1.0512,
+      "step": 6066
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7492396831512451,
+      "learning_rate": 1.5150982265883019e-05,
+      "loss": 0.995,
+      "step": 6067
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9079885482788086,
+      "learning_rate": 1.514938996028556e-05,
+      "loss": 1.0407,
+      "step": 6068
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7905523777008057,
+      "learning_rate": 1.5147797476997037e-05,
+      "loss": 0.9975,
+      "step": 6069
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.74009108543396,
+      "learning_rate": 1.5146204816072402e-05,
+      "loss": 1.0052,
+      "step": 6070
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.683832049369812,
+      "learning_rate": 1.5144611977566619e-05,
+      "loss": 0.8887,
+      "step": 6071
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9672870635986328,
+      "learning_rate": 1.5143018961534646e-05,
+      "loss": 1.0812,
+      "step": 6072
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9116899967193604,
+      "learning_rate": 1.5141425768031452e-05,
+      "loss": 0.9548,
+      "step": 6073
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.64646315574646,
+      "learning_rate": 1.5139832397112018e-05,
+      "loss": 0.9556,
+      "step": 6074
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7728618383407593,
+      "learning_rate": 1.5138238848831326e-05,
+      "loss": 0.9591,
+      "step": 6075
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9289491176605225,
+      "learning_rate": 1.5136645123244366e-05,
+      "loss": 1.0289,
+      "step": 6076
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.836245059967041,
+      "learning_rate": 1.513505122040613e-05,
+      "loss": 1.0431,
+      "step": 6077
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.042660713195801,
+      "learning_rate": 1.513345714037162e-05,
+      "loss": 0.9803,
+      "step": 6078
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7818243503570557,
+      "learning_rate": 1.5131862883195844e-05,
+      "loss": 1.0601,
+      "step": 6079
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8721898794174194,
+      "learning_rate": 1.5130268448933815e-05,
+      "loss": 1.0128,
+      "step": 6080
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1211024522781372,
+      "learning_rate": 1.5128673837640552e-05,
+      "loss": 0.6676,
+      "step": 6081
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8897124528884888,
+      "learning_rate": 1.5127079049371083e-05,
+      "loss": 1.0237,
+      "step": 6082
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8602781295776367,
+      "learning_rate": 1.5125484084180437e-05,
+      "loss": 0.9586,
+      "step": 6083
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0244290828704834,
+      "learning_rate": 1.5123888942123652e-05,
+      "loss": 1.0342,
+      "step": 6084
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.880847454071045,
+      "learning_rate": 1.5122293623255777e-05,
+      "loss": 0.9993,
+      "step": 6085
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6572504043579102,
+      "learning_rate": 1.5120698127631851e-05,
+      "loss": 1.0889,
+      "step": 6086
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8460131883621216,
+      "learning_rate": 1.5119102455306943e-05,
+      "loss": 1.0564,
+      "step": 6087
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0404727458953857,
+      "learning_rate": 1.5117506606336105e-05,
+      "loss": 0.9933,
+      "step": 6088
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6624268293380737,
+      "learning_rate": 1.511591058077441e-05,
+      "loss": 0.9102,
+      "step": 6089
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.831697702407837,
+      "learning_rate": 1.5114314378676928e-05,
+      "loss": 0.9825,
+      "step": 6090
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6961880922317505,
+      "learning_rate": 1.5112718000098746e-05,
+      "loss": 1.1015,
+      "step": 6091
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8591513633728027,
+      "learning_rate": 1.5111121445094952e-05,
+      "loss": 1.139,
+      "step": 6092
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7808781862258911,
+      "learning_rate": 1.510952471372063e-05,
+      "loss": 1.0073,
+      "step": 6093
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6450496912002563,
+      "learning_rate": 1.5107927806030885e-05,
+      "loss": 0.9958,
+      "step": 6094
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.808458924293518,
+      "learning_rate": 1.5106330722080815e-05,
+      "loss": 1.0669,
+      "step": 6095
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9477546215057373,
+      "learning_rate": 1.510473346192554e-05,
+      "loss": 1.0831,
+      "step": 6096
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7996560335159302,
+      "learning_rate": 1.5103136025620173e-05,
+      "loss": 0.9667,
+      "step": 6097
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7238742113113403,
+      "learning_rate": 1.5101538413219834e-05,
+      "loss": 0.9596,
+      "step": 6098
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7662135362625122,
+      "learning_rate": 1.5099940624779659e-05,
+      "loss": 1.0272,
+      "step": 6099
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8242950439453125,
+      "learning_rate": 1.5098342660354774e-05,
+      "loss": 1.0443,
+      "step": 6100
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.747917890548706,
+      "learning_rate": 1.509674452000033e-05,
+      "loss": 0.9921,
+      "step": 6101
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7819490432739258,
+      "learning_rate": 1.5095146203771466e-05,
+      "loss": 1.0778,
+      "step": 6102
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.910750389099121,
+      "learning_rate": 1.5093547711723343e-05,
+      "loss": 0.9859,
+      "step": 6103
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8931645154953003,
+      "learning_rate": 1.5091949043911114e-05,
+      "loss": 0.915,
+      "step": 6104
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8700298070907593,
+      "learning_rate": 1.5090350200389949e-05,
+      "loss": 0.9615,
+      "step": 6105
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8552329540252686,
+      "learning_rate": 1.5088751181215018e-05,
+      "loss": 1.0613,
+      "step": 6106
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0202159881591797,
+      "learning_rate": 1.5087151986441495e-05,
+      "loss": 1.0778,
+      "step": 6107
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7837328910827637,
+      "learning_rate": 1.508555261612457e-05,
+      "loss": 0.9643,
+      "step": 6108
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7941837310791016,
+      "learning_rate": 1.508395307031943e-05,
+      "loss": 1.0355,
+      "step": 6109
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1752427816390991,
+      "learning_rate": 1.5082353349081271e-05,
+      "loss": 0.5669,
+      "step": 6110
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.304492235183716,
+      "learning_rate": 1.5080753452465296e-05,
+      "loss": 1.0545,
+      "step": 6111
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0493496656417847,
+      "learning_rate": 1.507915338052671e-05,
+      "loss": 0.5654,
+      "step": 6112
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.802672266960144,
+      "learning_rate": 1.5077553133320732e-05,
+      "loss": 0.9976,
+      "step": 6113
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9464176893234253,
+      "learning_rate": 1.5075952710902577e-05,
+      "loss": 1.0095,
+      "step": 6114
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8905532360076904,
+      "learning_rate": 1.507435211332747e-05,
+      "loss": 1.0249,
+      "step": 6115
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.676032543182373,
+      "learning_rate": 1.5072751340650651e-05,
+      "loss": 0.9244,
+      "step": 6116
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.775314450263977,
+      "learning_rate": 1.5071150392927351e-05,
+      "loss": 0.9537,
+      "step": 6117
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8706495761871338,
+      "learning_rate": 1.5069549270212818e-05,
+      "loss": 1.0171,
+      "step": 6118
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7748523950576782,
+      "learning_rate": 1.5067947972562299e-05,
+      "loss": 0.9707,
+      "step": 6119
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.915094256401062,
+      "learning_rate": 1.5066346500031053e-05,
+      "loss": 0.9948,
+      "step": 6120
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7447072267532349,
+      "learning_rate": 1.5064744852674343e-05,
+      "loss": 0.9228,
+      "step": 6121
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.743086814880371,
+      "learning_rate": 1.5063143030547434e-05,
+      "loss": 0.9608,
+      "step": 6122
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7502856254577637,
+      "learning_rate": 1.50615410337056e-05,
+      "loss": 0.9502,
+      "step": 6123
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9303454160690308,
+      "learning_rate": 1.5059938862204126e-05,
+      "loss": 1.063,
+      "step": 6124
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9362964630126953,
+      "learning_rate": 1.5058336516098298e-05,
+      "loss": 0.9969,
+      "step": 6125
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9840667247772217,
+      "learning_rate": 1.5056733995443407e-05,
+      "loss": 1.0826,
+      "step": 6126
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.814815640449524,
+      "learning_rate": 1.505513130029475e-05,
+      "loss": 1.016,
+      "step": 6127
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8090100288391113,
+      "learning_rate": 1.5053528430707632e-05,
+      "loss": 1.0587,
+      "step": 6128
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.698258399963379,
+      "learning_rate": 1.5051925386737365e-05,
+      "loss": 0.9588,
+      "step": 6129
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8670573234558105,
+      "learning_rate": 1.5050322168439265e-05,
+      "loss": 0.9804,
+      "step": 6130
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8872298002243042,
+      "learning_rate": 1.5048718775868654e-05,
+      "loss": 1.0135,
+      "step": 6131
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.180208683013916,
+      "learning_rate": 1.504711520908086e-05,
+      "loss": 0.6059,
+      "step": 6132
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8020473718643188,
+      "learning_rate": 1.5045511468131222e-05,
+      "loss": 1.0464,
+      "step": 6133
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9412001371383667,
+      "learning_rate": 1.5043907553075072e-05,
+      "loss": 0.9998,
+      "step": 6134
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8525619506835938,
+      "learning_rate": 1.5042303463967767e-05,
+      "loss": 0.9515,
+      "step": 6135
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8055158853530884,
+      "learning_rate": 1.5040699200864653e-05,
+      "loss": 1.0043,
+      "step": 6136
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8440173864364624,
+      "learning_rate": 1.5039094763821091e-05,
+      "loss": 1.0779,
+      "step": 6137
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8687825202941895,
+      "learning_rate": 1.5037490152892443e-05,
+      "loss": 0.9529,
+      "step": 6138
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8497471809387207,
+      "learning_rate": 1.5035885368134082e-05,
+      "loss": 1.0351,
+      "step": 6139
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.84963858127594,
+      "learning_rate": 1.5034280409601386e-05,
+      "loss": 0.9766,
+      "step": 6140
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.051640748977661,
+      "learning_rate": 1.5032675277349733e-05,
+      "loss": 0.9926,
+      "step": 6141
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9018441438674927,
+      "learning_rate": 1.5031069971434517e-05,
+      "loss": 1.0332,
+      "step": 6142
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1138722896575928,
+      "learning_rate": 1.502946449191113e-05,
+      "loss": 1.0232,
+      "step": 6143
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9888110160827637,
+      "learning_rate": 1.502785883883497e-05,
+      "loss": 1.0505,
+      "step": 6144
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9782544374465942,
+      "learning_rate": 1.5026253012261448e-05,
+      "loss": 0.9862,
+      "step": 6145
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8075544834136963,
+      "learning_rate": 1.5024647012245972e-05,
+      "loss": 0.9944,
+      "step": 6146
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7304812669754028,
+      "learning_rate": 1.5023040838843966e-05,
+      "loss": 0.9817,
+      "step": 6147
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0388548374176025,
+      "learning_rate": 1.5021434492110851e-05,
+      "loss": 1.0171,
+      "step": 6148
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.899377703666687,
+      "learning_rate": 1.501982797210206e-05,
+      "loss": 0.9672,
+      "step": 6149
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7409586906433105,
+      "learning_rate": 1.5018221278873028e-05,
+      "loss": 1.0767,
+      "step": 6150
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.828242301940918,
+      "learning_rate": 1.5016614412479195e-05,
+      "loss": 0.9295,
+      "step": 6151
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.278900146484375,
+      "learning_rate": 1.5015007372976013e-05,
+      "loss": 0.6004,
+      "step": 6152
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1635106801986694,
+      "learning_rate": 1.5013400160418939e-05,
+      "loss": 0.635,
+      "step": 6153
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8994755744934082,
+      "learning_rate": 1.5011792774863425e-05,
+      "loss": 0.9922,
+      "step": 6154
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0562361478805542,
+      "learning_rate": 1.5010185216364947e-05,
+      "loss": 0.5855,
+      "step": 6155
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.134864568710327,
+      "learning_rate": 1.5008577484978966e-05,
+      "loss": 1.0352,
+      "step": 6156
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7977651357650757,
+      "learning_rate": 1.5006969580760973e-05,
+      "loss": 1.0137,
+      "step": 6157
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.972727656364441,
+      "learning_rate": 1.5005361503766442e-05,
+      "loss": 1.0332,
+      "step": 6158
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.76993989944458,
+      "learning_rate": 1.500375325405087e-05,
+      "loss": 0.9736,
+      "step": 6159
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9606564044952393,
+      "learning_rate": 1.5002144831669752e-05,
+      "loss": 1.0053,
+      "step": 6160
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.914209008216858,
+      "learning_rate": 1.5000536236678583e-05,
+      "loss": 0.9853,
+      "step": 6161
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9595354795455933,
+      "learning_rate": 1.4998927469132881e-05,
+      "loss": 1.031,
+      "step": 6162
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8119326829910278,
+      "learning_rate": 1.4997318529088153e-05,
+      "loss": 0.9371,
+      "step": 6163
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0008444786071777,
+      "learning_rate": 1.4995709416599927e-05,
+      "loss": 0.9842,
+      "step": 6164
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7979505062103271,
+      "learning_rate": 1.4994100131723721e-05,
+      "loss": 0.9776,
+      "step": 6165
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8173017501831055,
+      "learning_rate": 1.499249067451507e-05,
+      "loss": 1.0238,
+      "step": 6166
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8522803783416748,
+      "learning_rate": 1.4990881045029512e-05,
+      "loss": 1.0515,
+      "step": 6167
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8449554443359375,
+      "learning_rate": 1.4989271243322592e-05,
+      "loss": 1.0575,
+      "step": 6168
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9337587356567383,
+      "learning_rate": 1.4987661269449858e-05,
+      "loss": 1.0309,
+      "step": 6169
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5684670209884644,
+      "learning_rate": 1.4986051123466864e-05,
+      "loss": 0.9466,
+      "step": 6170
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8391071557998657,
+      "learning_rate": 1.4984440805429175e-05,
+      "loss": 1.0791,
+      "step": 6171
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8247380256652832,
+      "learning_rate": 1.4982830315392357e-05,
+      "loss": 0.928,
+      "step": 6172
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9018083810806274,
+      "learning_rate": 1.4981219653411983e-05,
+      "loss": 0.9236,
+      "step": 6173
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9272807836532593,
+      "learning_rate": 1.4979608819543635e-05,
+      "loss": 0.9611,
+      "step": 6174
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9058659076690674,
+      "learning_rate": 1.4977997813842894e-05,
+      "loss": 1.0078,
+      "step": 6175
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9096859693527222,
+      "learning_rate": 1.4976386636365358e-05,
+      "loss": 0.9601,
+      "step": 6176
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6966136693954468,
+      "learning_rate": 1.4974775287166616e-05,
+      "loss": 1.004,
+      "step": 6177
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8732268810272217,
+      "learning_rate": 1.4973163766302279e-05,
+      "loss": 1.0294,
+      "step": 6178
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6418588161468506,
+      "learning_rate": 1.497155207382795e-05,
+      "loss": 0.9561,
+      "step": 6179
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9602961540222168,
+      "learning_rate": 1.4969940209799248e-05,
+      "loss": 1.0145,
+      "step": 6180
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7430047988891602,
+      "learning_rate": 1.4968328174271791e-05,
+      "loss": 0.8779,
+      "step": 6181
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8561376333236694,
+      "learning_rate": 1.4966715967301209e-05,
+      "loss": 1.0128,
+      "step": 6182
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0214855670928955,
+      "learning_rate": 1.4965103588943131e-05,
+      "loss": 0.9816,
+      "step": 6183
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7786353826522827,
+      "learning_rate": 1.4963491039253198e-05,
+      "loss": 1.0303,
+      "step": 6184
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.715368628501892,
+      "learning_rate": 1.4961878318287051e-05,
+      "loss": 1.0221,
+      "step": 6185
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.076876401901245,
+      "learning_rate": 1.4960265426100348e-05,
+      "loss": 1.0527,
+      "step": 6186
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3664613962173462,
+      "learning_rate": 1.4958652362748741e-05,
+      "loss": 0.658,
+      "step": 6187
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1998190879821777,
+      "learning_rate": 1.4957039128287891e-05,
+      "loss": 0.6562,
+      "step": 6188
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.003871440887451,
+      "learning_rate": 1.4955425722773467e-05,
+      "loss": 1.0703,
+      "step": 6189
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8492906093597412,
+      "learning_rate": 1.4953812146261143e-05,
+      "loss": 0.8865,
+      "step": 6190
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8066935539245605,
+      "learning_rate": 1.4952198398806603e-05,
+      "loss": 1.1185,
+      "step": 6191
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.806857943534851,
+      "learning_rate": 1.4950584480465526e-05,
+      "loss": 1.0256,
+      "step": 6192
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.044191598892212,
+      "learning_rate": 1.494897039129361e-05,
+      "loss": 0.947,
+      "step": 6193
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9217848777770996,
+      "learning_rate": 1.4947356131346551e-05,
+      "loss": 0.9707,
+      "step": 6194
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7288991212844849,
+      "learning_rate": 1.4945741700680052e-05,
+      "loss": 0.9765,
+      "step": 6195
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0345983505249023,
+      "learning_rate": 1.4944127099349821e-05,
+      "loss": 1.0394,
+      "step": 6196
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8446462154388428,
+      "learning_rate": 1.4942512327411574e-05,
+      "loss": 1.0694,
+      "step": 6197
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.843661904335022,
+      "learning_rate": 1.4940897384921034e-05,
+      "loss": 0.9276,
+      "step": 6198
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9998055696487427,
+      "learning_rate": 1.4939282271933926e-05,
+      "loss": 0.9923,
+      "step": 6199
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9066799879074097,
+      "learning_rate": 1.4937666988505984e-05,
+      "loss": 1.0521,
+      "step": 6200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9892860651016235,
+      "learning_rate": 1.4936051534692948e-05,
+      "loss": 1.0712,
+      "step": 6201
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.926393985748291,
+      "learning_rate": 1.4934435910550562e-05,
+      "loss": 0.9949,
+      "step": 6202
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.122161388397217,
+      "learning_rate": 1.4932820116134575e-05,
+      "loss": 0.9733,
+      "step": 6203
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6975136995315552,
+      "learning_rate": 1.4931204151500746e-05,
+      "loss": 0.9931,
+      "step": 6204
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7935056686401367,
+      "learning_rate": 1.4929588016704837e-05,
+      "loss": 0.9855,
+      "step": 6205
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.783319354057312,
+      "learning_rate": 1.4927971711802615e-05,
+      "loss": 1.0147,
+      "step": 6206
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7814267873764038,
+      "learning_rate": 1.4926355236849857e-05,
+      "loss": 1.1237,
+      "step": 6207
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.939418077468872,
+      "learning_rate": 1.492473859190234e-05,
+      "loss": 1.0465,
+      "step": 6208
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6745635271072388,
+      "learning_rate": 1.492312177701585e-05,
+      "loss": 0.9628,
+      "step": 6209
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6615424156188965,
+      "learning_rate": 1.492150479224618e-05,
+      "loss": 0.9644,
+      "step": 6210
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9777902364730835,
+      "learning_rate": 1.4919887637649127e-05,
+      "loss": 1.0234,
+      "step": 6211
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6302640438079834,
+      "learning_rate": 1.4918270313280494e-05,
+      "loss": 1.0398,
+      "step": 6212
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3890011310577393,
+      "learning_rate": 1.4916652819196091e-05,
+      "loss": 0.6534,
+      "step": 6213
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8851747512817383,
+      "learning_rate": 1.4915035155451736e-05,
+      "loss": 0.9988,
+      "step": 6214
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9260203838348389,
+      "learning_rate": 1.4913417322103245e-05,
+      "loss": 1.0126,
+      "step": 6215
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9152077436447144,
+      "learning_rate": 1.4911799319206449e-05,
+      "loss": 1.0496,
+      "step": 6216
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8295962810516357,
+      "learning_rate": 1.4910181146817178e-05,
+      "loss": 1.0143,
+      "step": 6217
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.916728138923645,
+      "learning_rate": 1.490856280499127e-05,
+      "loss": 1.0762,
+      "step": 6218
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7609875202178955,
+      "learning_rate": 1.4906944293784574e-05,
+      "loss": 1.046,
+      "step": 6219
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7399111986160278,
+      "learning_rate": 1.4905325613252937e-05,
+      "loss": 1.0126,
+      "step": 6220
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7363574504852295,
+      "learning_rate": 1.4903706763452214e-05,
+      "loss": 0.9556,
+      "step": 6221
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1364474296569824,
+      "learning_rate": 1.4902087744438269e-05,
+      "loss": 1.0122,
+      "step": 6222
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.840543270111084,
+      "learning_rate": 1.490046855626697e-05,
+      "loss": 0.9901,
+      "step": 6223
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0427427291870117,
+      "learning_rate": 1.489884919899419e-05,
+      "loss": 1.0179,
+      "step": 6224
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7339650392532349,
+      "learning_rate": 1.4897229672675807e-05,
+      "loss": 0.9775,
+      "step": 6225
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6437357664108276,
+      "learning_rate": 1.489560997736771e-05,
+      "loss": 0.9673,
+      "step": 6226
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.935800313949585,
+      "learning_rate": 1.4893990113125786e-05,
+      "loss": 1.1411,
+      "step": 6227
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9649231433868408,
+      "learning_rate": 1.4892370080005936e-05,
+      "loss": 1.037,
+      "step": 6228
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8260817527770996,
+      "learning_rate": 1.489074987806406e-05,
+      "loss": 1.0888,
+      "step": 6229
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7246772050857544,
+      "learning_rate": 1.4889129507356068e-05,
+      "loss": 1.0277,
+      "step": 6230
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9139410257339478,
+      "learning_rate": 1.4887508967937874e-05,
+      "loss": 0.9334,
+      "step": 6231
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7284680604934692,
+      "learning_rate": 1.4885888259865398e-05,
+      "loss": 0.9824,
+      "step": 6232
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6977698802947998,
+      "learning_rate": 1.4884267383194567e-05,
+      "loss": 1.0428,
+      "step": 6233
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9950957298278809,
+      "learning_rate": 1.488264633798131e-05,
+      "loss": 0.593,
+      "step": 6234
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7299916744232178,
+      "learning_rate": 1.488102512428157e-05,
+      "loss": 1.019,
+      "step": 6235
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.940701961517334,
+      "learning_rate": 1.4879403742151283e-05,
+      "loss": 0.9799,
+      "step": 6236
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0809202194213867,
+      "learning_rate": 1.4877782191646408e-05,
+      "loss": 0.612,
+      "step": 6237
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9608681201934814,
+      "learning_rate": 1.4876160472822894e-05,
+      "loss": 0.9512,
+      "step": 6238
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.684124231338501,
+      "learning_rate": 1.48745385857367e-05,
+      "loss": 1.0196,
+      "step": 6239
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9151595830917358,
+      "learning_rate": 1.4872916530443797e-05,
+      "loss": 0.96,
+      "step": 6240
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6824043989181519,
+      "learning_rate": 1.4871294307000158e-05,
+      "loss": 0.9408,
+      "step": 6241
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8919451236724854,
+      "learning_rate": 1.486967191546176e-05,
+      "loss": 1.0664,
+      "step": 6242
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.03244948387146,
+      "learning_rate": 1.4868049355884586e-05,
+      "loss": 1.1093,
+      "step": 6243
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6975504159927368,
+      "learning_rate": 1.4866426628324625e-05,
+      "loss": 0.9314,
+      "step": 6244
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.744381070137024,
+      "learning_rate": 1.4864803732837878e-05,
+      "loss": 1.0092,
+      "step": 6245
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.679764986038208,
+      "learning_rate": 1.4863180669480344e-05,
+      "loss": 1.0801,
+      "step": 6246
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6461364030838013,
+      "learning_rate": 1.486155743830803e-05,
+      "loss": 0.9785,
+      "step": 6247
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8205231428146362,
+      "learning_rate": 1.4859934039376947e-05,
+      "loss": 1.0272,
+      "step": 6248
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1533944606781006,
+      "learning_rate": 1.4858310472743117e-05,
+      "loss": 0.9756,
+      "step": 6249
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6899497509002686,
+      "learning_rate": 1.4856686738462563e-05,
+      "loss": 0.9829,
+      "step": 6250
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9039406776428223,
+      "learning_rate": 1.4855062836591313e-05,
+      "loss": 1.0296,
+      "step": 6251
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7875765562057495,
+      "learning_rate": 1.4853438767185411e-05,
+      "loss": 0.8543,
+      "step": 6252
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8219566345214844,
+      "learning_rate": 1.4851814530300895e-05,
+      "loss": 0.9051,
+      "step": 6253
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7640273571014404,
+      "learning_rate": 1.4850190125993811e-05,
+      "loss": 1.0048,
+      "step": 6254
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7041285037994385,
+      "learning_rate": 1.484856555432021e-05,
+      "loss": 0.982,
+      "step": 6255
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8866806030273438,
+      "learning_rate": 1.4846940815336162e-05,
+      "loss": 1.0454,
+      "step": 6256
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8135604858398438,
+      "learning_rate": 1.4845315909097724e-05,
+      "loss": 0.9845,
+      "step": 6257
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6515693664550781,
+      "learning_rate": 1.4843690835660968e-05,
+      "loss": 0.9832,
+      "step": 6258
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.726922631263733,
+      "learning_rate": 1.4842065595081973e-05,
+      "loss": 0.9974,
+      "step": 6259
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7245620489120483,
+      "learning_rate": 1.484044018741682e-05,
+      "loss": 0.9304,
+      "step": 6260
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7319328784942627,
+      "learning_rate": 1.4838814612721599e-05,
+      "loss": 1.0461,
+      "step": 6261
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7668333053588867,
+      "learning_rate": 1.4837188871052399e-05,
+      "loss": 1.0281,
+      "step": 6262
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.839624047279358,
+      "learning_rate": 1.4835562962465323e-05,
+      "loss": 0.9605,
+      "step": 6263
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6296578645706177,
+      "learning_rate": 1.483393688701648e-05,
+      "loss": 1.0166,
+      "step": 6264
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7272576093673706,
+      "learning_rate": 1.4832310644761978e-05,
+      "loss": 1.0149,
+      "step": 6265
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.641893982887268,
+      "learning_rate": 1.483068423575793e-05,
+      "loss": 0.996,
+      "step": 6266
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7102335691452026,
+      "learning_rate": 1.4829057660060464e-05,
+      "loss": 0.9954,
+      "step": 6267
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.698488712310791,
+      "learning_rate": 1.482743091772571e-05,
+      "loss": 0.9074,
+      "step": 6268
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7537356615066528,
+      "learning_rate": 1.4825804008809799e-05,
+      "loss": 0.9874,
+      "step": 6269
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8343656063079834,
+      "learning_rate": 1.4824176933368873e-05,
+      "loss": 1.013,
+      "step": 6270
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8277825117111206,
+      "learning_rate": 1.4822549691459077e-05,
+      "loss": 0.9601,
+      "step": 6271
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9550387859344482,
+      "learning_rate": 1.482092228313656e-05,
+      "loss": 1.0608,
+      "step": 6272
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8219711780548096,
+      "learning_rate": 1.4819294708457484e-05,
+      "loss": 1.0604,
+      "step": 6273
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8420875072479248,
+      "learning_rate": 1.4817666967478008e-05,
+      "loss": 0.9718,
+      "step": 6274
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0286264419555664,
+      "learning_rate": 1.4816039060254304e-05,
+      "loss": 1.0603,
+      "step": 6275
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6736133098602295,
+      "learning_rate": 1.4814410986842544e-05,
+      "loss": 0.99,
+      "step": 6276
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9076026678085327,
+      "learning_rate": 1.4812782747298911e-05,
+      "loss": 1.0321,
+      "step": 6277
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1877388954162598,
+      "learning_rate": 1.4811154341679585e-05,
+      "loss": 1.0694,
+      "step": 6278
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7865681648254395,
+      "learning_rate": 1.4809525770040764e-05,
+      "loss": 1.0604,
+      "step": 6279
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7880630493164062,
+      "learning_rate": 1.4807897032438646e-05,
+      "loss": 0.9887,
+      "step": 6280
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5941312313079834,
+      "learning_rate": 1.4806268128929431e-05,
+      "loss": 0.9557,
+      "step": 6281
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7703405618667603,
+      "learning_rate": 1.4804639059569327e-05,
+      "loss": 0.9952,
+      "step": 6282
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8247032165527344,
+      "learning_rate": 1.4803009824414552e-05,
+      "loss": 1.0362,
+      "step": 6283
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6750893592834473,
+      "learning_rate": 1.4801380423521323e-05,
+      "loss": 1.014,
+      "step": 6284
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8853751420974731,
+      "learning_rate": 1.4799750856945869e-05,
+      "loss": 1.0335,
+      "step": 6285
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.031217336654663,
+      "learning_rate": 1.4798121124744421e-05,
+      "loss": 0.9928,
+      "step": 6286
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7678987979888916,
+      "learning_rate": 1.4796491226973215e-05,
+      "loss": 0.8818,
+      "step": 6287
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7393865585327148,
+      "learning_rate": 1.4794861163688495e-05,
+      "loss": 1.0452,
+      "step": 6288
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8612635135650635,
+      "learning_rate": 1.479323093494651e-05,
+      "loss": 1.0073,
+      "step": 6289
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8555412292480469,
+      "learning_rate": 1.4791600540803514e-05,
+      "loss": 0.9929,
+      "step": 6290
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.912619709968567,
+      "learning_rate": 1.478996998131577e-05,
+      "loss": 1.0378,
+      "step": 6291
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6263010501861572,
+      "learning_rate": 1.4788339256539543e-05,
+      "loss": 1.0443,
+      "step": 6292
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1084511280059814,
+      "learning_rate": 1.47867083665311e-05,
+      "loss": 0.9555,
+      "step": 6293
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8931496143341064,
+      "learning_rate": 1.4785077311346725e-05,
+      "loss": 1.0787,
+      "step": 6294
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0524985790252686,
+      "learning_rate": 1.4783446091042698e-05,
+      "loss": 0.6046,
+      "step": 6295
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8386329412460327,
+      "learning_rate": 1.478181470567531e-05,
+      "loss": 1.0903,
+      "step": 6296
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8875433206558228,
+      "learning_rate": 1.4780183155300853e-05,
+      "loss": 1.027,
+      "step": 6297
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.003554344177246,
+      "learning_rate": 1.4778551439975629e-05,
+      "loss": 0.5215,
+      "step": 6298
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7032344341278076,
+      "learning_rate": 1.477691955975594e-05,
+      "loss": 1.079,
+      "step": 6299
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8045324087142944,
+      "learning_rate": 1.4775287514698105e-05,
+      "loss": 0.971,
+      "step": 6300
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.139028787612915,
+      "learning_rate": 1.4773655304858434e-05,
+      "loss": 0.6167,
+      "step": 6301
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7120414972305298,
+      "learning_rate": 1.4772022930293256e-05,
+      "loss": 1.0214,
+      "step": 6302
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8945528268814087,
+      "learning_rate": 1.4770390391058894e-05,
+      "loss": 1.0712,
+      "step": 6303
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6438384056091309,
+      "learning_rate": 1.4768757687211685e-05,
+      "loss": 0.9501,
+      "step": 6304
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6727845668792725,
+      "learning_rate": 1.476712481880797e-05,
+      "loss": 0.9553,
+      "step": 6305
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8367928266525269,
+      "learning_rate": 1.4765491785904094e-05,
+      "loss": 0.9926,
+      "step": 6306
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7160555124282837,
+      "learning_rate": 1.476385858855641e-05,
+      "loss": 1.0785,
+      "step": 6307
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7079044580459595,
+      "learning_rate": 1.4762225226821272e-05,
+      "loss": 1.105,
+      "step": 6308
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6129440069198608,
+      "learning_rate": 1.4760591700755042e-05,
+      "loss": 0.9922,
+      "step": 6309
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.846893310546875,
+      "learning_rate": 1.4758958010414094e-05,
+      "loss": 0.9897,
+      "step": 6310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9567654132843018,
+      "learning_rate": 1.4757324155854798e-05,
+      "loss": 1.0259,
+      "step": 6311
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7855849266052246,
+      "learning_rate": 1.4755690137133534e-05,
+      "loss": 0.9321,
+      "step": 6312
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7195664644241333,
+      "learning_rate": 1.4754055954306687e-05,
+      "loss": 0.9601,
+      "step": 6313
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9063020944595337,
+      "learning_rate": 1.4752421607430649e-05,
+      "loss": 0.9526,
+      "step": 6314
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7135204076766968,
+      "learning_rate": 1.4750787096561818e-05,
+      "loss": 0.9973,
+      "step": 6315
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8679938316345215,
+      "learning_rate": 1.4749152421756596e-05,
+      "loss": 1.0395,
+      "step": 6316
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8654956817626953,
+      "learning_rate": 1.4747517583071386e-05,
+      "loss": 1.0344,
+      "step": 6317
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.706494927406311,
+      "learning_rate": 1.4745882580562609e-05,
+      "loss": 0.9955,
+      "step": 6318
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.710939645767212,
+      "learning_rate": 1.4744247414286681e-05,
+      "loss": 1.0618,
+      "step": 6319
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.870609998703003,
+      "learning_rate": 1.4742612084300025e-05,
+      "loss": 0.9628,
+      "step": 6320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.637471079826355,
+      "learning_rate": 1.4740976590659075e-05,
+      "loss": 1.0338,
+      "step": 6321
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7514467239379883,
+      "learning_rate": 1.4739340933420268e-05,
+      "loss": 0.9147,
+      "step": 6322
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.685638666152954,
+      "learning_rate": 1.4737705112640044e-05,
+      "loss": 1.1065,
+      "step": 6323
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6635195016860962,
+      "learning_rate": 1.4736069128374851e-05,
+      "loss": 0.9881,
+      "step": 6324
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1888723373413086,
+      "learning_rate": 1.473443298068114e-05,
+      "loss": 0.6156,
+      "step": 6325
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7495137453079224,
+      "learning_rate": 1.4732796669615372e-05,
+      "loss": 0.8819,
+      "step": 6326
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8505098819732666,
+      "learning_rate": 1.4731160195234013e-05,
+      "loss": 1.0184,
+      "step": 6327
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7478581666946411,
+      "learning_rate": 1.4729523557593532e-05,
+      "loss": 1.0741,
+      "step": 6328
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7161797285079956,
+      "learning_rate": 1.4727886756750404e-05,
+      "loss": 0.9647,
+      "step": 6329
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9601303339004517,
+      "learning_rate": 1.472624979276111e-05,
+      "loss": 1.0916,
+      "step": 6330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6027662754058838,
+      "learning_rate": 1.4724612665682139e-05,
+      "loss": 0.9793,
+      "step": 6331
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.978075385093689,
+      "learning_rate": 1.4722975375569978e-05,
+      "loss": 0.9719,
+      "step": 6332
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8380613327026367,
+      "learning_rate": 1.4721337922481135e-05,
+      "loss": 0.9598,
+      "step": 6333
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1491286754608154,
+      "learning_rate": 1.4719700306472108e-05,
+      "loss": 0.6358,
+      "step": 6334
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7171493768692017,
+      "learning_rate": 1.4718062527599408e-05,
+      "loss": 0.9946,
+      "step": 6335
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.051431894302368,
+      "learning_rate": 1.4716424585919548e-05,
+      "loss": 1.1288,
+      "step": 6336
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7927138805389404,
+      "learning_rate": 1.4714786481489052e-05,
+      "loss": 0.9239,
+      "step": 6337
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7960854768753052,
+      "learning_rate": 1.4713148214364443e-05,
+      "loss": 0.9576,
+      "step": 6338
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7240616083145142,
+      "learning_rate": 1.4711509784602256e-05,
+      "loss": 1.0141,
+      "step": 6339
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.897529125213623,
+      "learning_rate": 1.4709871192259027e-05,
+      "loss": 1.0801,
+      "step": 6340
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0229244232177734,
+      "learning_rate": 1.4708232437391299e-05,
+      "loss": 0.6452,
+      "step": 6341
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8683282136917114,
+      "learning_rate": 1.4706593520055624e-05,
+      "loss": 1.0475,
+      "step": 6342
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7752506732940674,
+      "learning_rate": 1.470495444030855e-05,
+      "loss": 0.9797,
+      "step": 6343
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7672481536865234,
+      "learning_rate": 1.4703315198206643e-05,
+      "loss": 1.0958,
+      "step": 6344
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7327262163162231,
+      "learning_rate": 1.4701675793806464e-05,
+      "loss": 1.0391,
+      "step": 6345
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6891984939575195,
+      "learning_rate": 1.4700036227164592e-05,
+      "loss": 0.9485,
+      "step": 6346
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9141579866409302,
+      "learning_rate": 1.4698396498337595e-05,
+      "loss": 1.1333,
+      "step": 6347
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8592197895050049,
+      "learning_rate": 1.469675660738206e-05,
+      "loss": 0.9578,
+      "step": 6348
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.065803289413452,
+      "learning_rate": 1.4695116554354576e-05,
+      "loss": 1.0292,
+      "step": 6349
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8045185804367065,
+      "learning_rate": 1.4693476339311734e-05,
+      "loss": 0.9475,
+      "step": 6350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7326642274856567,
+      "learning_rate": 1.4691835962310135e-05,
+      "loss": 1.0681,
+      "step": 6351
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9410340785980225,
+      "learning_rate": 1.4690195423406381e-05,
+      "loss": 1.0174,
+      "step": 6352
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0535422563552856,
+      "learning_rate": 1.4688554722657087e-05,
+      "loss": 0.6152,
+      "step": 6353
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.753043532371521,
+      "learning_rate": 1.4686913860118865e-05,
+      "loss": 1.0644,
+      "step": 6354
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.030503511428833,
+      "learning_rate": 1.4685272835848336e-05,
+      "loss": 0.948,
+      "step": 6355
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0237739086151123,
+      "learning_rate": 1.4683631649902132e-05,
+      "loss": 1.0836,
+      "step": 6356
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.687549114227295,
+      "learning_rate": 1.4681990302336884e-05,
+      "loss": 0.9991,
+      "step": 6357
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7775124311447144,
+      "learning_rate": 1.4680348793209227e-05,
+      "loss": 0.9783,
+      "step": 6358
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7740108966827393,
+      "learning_rate": 1.4678707122575806e-05,
+      "loss": 0.9924,
+      "step": 6359
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9009690284729004,
+      "learning_rate": 1.4677065290493273e-05,
+      "loss": 0.9882,
+      "step": 6360
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6432981491088867,
+      "learning_rate": 1.4675423297018283e-05,
+      "loss": 0.9709,
+      "step": 6361
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.709234356880188,
+      "learning_rate": 1.4673781142207496e-05,
+      "loss": 0.9513,
+      "step": 6362
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6530418395996094,
+      "learning_rate": 1.4672138826117576e-05,
+      "loss": 0.9322,
+      "step": 6363
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7418795824050903,
+      "learning_rate": 1.4670496348805197e-05,
+      "loss": 0.9852,
+      "step": 6364
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8017892837524414,
+      "learning_rate": 1.4668853710327033e-05,
+      "loss": 1.0185,
+      "step": 6365
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0400454998016357,
+      "learning_rate": 1.466721091073977e-05,
+      "loss": 0.9944,
+      "step": 6366
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8270628452301025,
+      "learning_rate": 1.46655679501001e-05,
+      "loss": 1.0031,
+      "step": 6367
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9862697720527649,
+      "learning_rate": 1.4663924828464709e-05,
+      "loss": 0.5768,
+      "step": 6368
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9824302196502686,
+      "learning_rate": 1.46622815458903e-05,
+      "loss": 1.108,
+      "step": 6369
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.803748607635498,
+      "learning_rate": 1.466063810243358e-05,
+      "loss": 1.0048,
+      "step": 6370
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.897152066230774,
+      "learning_rate": 1.4658994498151255e-05,
+      "loss": 1.081,
+      "step": 6371
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8599333763122559,
+      "learning_rate": 1.465735073310005e-05,
+      "loss": 1.0439,
+      "step": 6372
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7976322174072266,
+      "learning_rate": 1.4655706807336676e-05,
+      "loss": 0.9446,
+      "step": 6373
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.88971745967865,
+      "learning_rate": 1.4654062720917868e-05,
+      "loss": 1.0468,
+      "step": 6374
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8467110395431519,
+      "learning_rate": 1.4652418473900355e-05,
+      "loss": 1.0239,
+      "step": 6375
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7547907829284668,
+      "learning_rate": 1.4650774066340877e-05,
+      "loss": 0.9928,
+      "step": 6376
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7801387310028076,
+      "learning_rate": 1.4649129498296175e-05,
+      "loss": 0.9929,
+      "step": 6377
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1547889709472656,
+      "learning_rate": 1.4647484769823004e-05,
+      "loss": 1.089,
+      "step": 6378
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9286860227584839,
+      "learning_rate": 1.4645839880978114e-05,
+      "loss": 0.9628,
+      "step": 6379
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6162844896316528,
+      "learning_rate": 1.4644194831818268e-05,
+      "loss": 0.9821,
+      "step": 6380
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7161037921905518,
+      "learning_rate": 1.4642549622400233e-05,
+      "loss": 0.9686,
+      "step": 6381
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.887052059173584,
+      "learning_rate": 1.4640904252780776e-05,
+      "loss": 0.9602,
+      "step": 6382
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7911127805709839,
+      "learning_rate": 1.4639258723016676e-05,
+      "loss": 1.0207,
+      "step": 6383
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.834877610206604,
+      "learning_rate": 1.4637613033164719e-05,
+      "loss": 1.0345,
+      "step": 6384
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.651520848274231,
+      "learning_rate": 1.4635967183281692e-05,
+      "loss": 0.9821,
+      "step": 6385
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8115218877792358,
+      "learning_rate": 1.4634321173424386e-05,
+      "loss": 1.0714,
+      "step": 6386
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.091355562210083,
+      "learning_rate": 1.4632675003649604e-05,
+      "loss": 1.0827,
+      "step": 6387
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0397772789001465,
+      "learning_rate": 1.4631028674014143e-05,
+      "loss": 1.1114,
+      "step": 6388
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6521655321121216,
+      "learning_rate": 1.4629382184574823e-05,
+      "loss": 0.9815,
+      "step": 6389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8779476881027222,
+      "learning_rate": 1.4627735535388455e-05,
+      "loss": 1.0096,
+      "step": 6390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9257357120513916,
+      "learning_rate": 1.462608872651186e-05,
+      "loss": 1.0815,
+      "step": 6391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8574851751327515,
+      "learning_rate": 1.4624441758001865e-05,
+      "loss": 0.9644,
+      "step": 6392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8194639682769775,
+      "learning_rate": 1.4622794629915306e-05,
+      "loss": 0.9431,
+      "step": 6393
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8634132146835327,
+      "learning_rate": 1.4621147342309016e-05,
+      "loss": 1.0495,
+      "step": 6394
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9924507141113281,
+      "learning_rate": 1.4619499895239839e-05,
+      "loss": 0.9619,
+      "step": 6395
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1242406368255615,
+      "learning_rate": 1.4617852288764624e-05,
+      "loss": 0.6205,
+      "step": 6396
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.799599051475525,
+      "learning_rate": 1.4616204522940227e-05,
+      "loss": 1.0294,
+      "step": 6397
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.718880534172058,
+      "learning_rate": 1.4614556597823506e-05,
+      "loss": 0.9918,
+      "step": 6398
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.833927869796753,
+      "learning_rate": 1.4612908513471329e-05,
+      "loss": 0.9039,
+      "step": 6399
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8740451335906982,
+      "learning_rate": 1.4611260269940563e-05,
+      "loss": 1.0074,
+      "step": 6400
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.761946678161621,
+      "learning_rate": 1.4609611867288087e-05,
+      "loss": 0.9872,
+      "step": 6401
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6574726104736328,
+      "learning_rate": 1.4607963305570783e-05,
+      "loss": 1.0523,
+      "step": 6402
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8972047567367554,
+      "learning_rate": 1.4606314584845536e-05,
+      "loss": 0.952,
+      "step": 6403
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1345824003219604,
+      "learning_rate": 1.4604665705169239e-05,
+      "loss": 0.64,
+      "step": 6404
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8390077352523804,
+      "learning_rate": 1.4603016666598793e-05,
+      "loss": 1.0317,
+      "step": 6405
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8301200866699219,
+      "learning_rate": 1.4601367469191098e-05,
+      "loss": 1.07,
+      "step": 6406
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0173637866973877,
+      "learning_rate": 1.4599718113003065e-05,
+      "loss": 1.0378,
+      "step": 6407
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7590110301971436,
+      "learning_rate": 1.459806859809161e-05,
+      "loss": 1.004,
+      "step": 6408
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4076590538024902,
+      "learning_rate": 1.4596418924513652e-05,
+      "loss": 0.608,
+      "step": 6409
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7717429399490356,
+      "learning_rate": 1.4594769092326113e-05,
+      "loss": 1.0398,
+      "step": 6410
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9422847032546997,
+      "learning_rate": 1.4593119101585931e-05,
+      "loss": 0.9805,
+      "step": 6411
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9014897346496582,
+      "learning_rate": 1.4591468952350039e-05,
+      "loss": 1.0414,
+      "step": 6412
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.776681661605835,
+      "learning_rate": 1.4589818644675378e-05,
+      "loss": 1.0561,
+      "step": 6413
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7230682373046875,
+      "learning_rate": 1.4588168178618897e-05,
+      "loss": 1.0137,
+      "step": 6414
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0297831296920776,
+      "learning_rate": 1.4586517554237549e-05,
+      "loss": 0.6216,
+      "step": 6415
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0024008750915527,
+      "learning_rate": 1.4584866771588294e-05,
+      "loss": 0.9889,
+      "step": 6416
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6077396869659424,
+      "learning_rate": 1.4583215830728092e-05,
+      "loss": 1.0507,
+      "step": 6417
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6634632349014282,
+      "learning_rate": 1.4581564731713915e-05,
+      "loss": 1.0837,
+      "step": 6418
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8909369707107544,
+      "learning_rate": 1.4579913474602738e-05,
+      "loss": 1.0191,
+      "step": 6419
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.847243070602417,
+      "learning_rate": 1.4578262059451538e-05,
+      "loss": 1.0072,
+      "step": 6420
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7284141778945923,
+      "learning_rate": 1.4576610486317302e-05,
+      "loss": 1.0019,
+      "step": 6421
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6591328382492065,
+      "learning_rate": 1.4574958755257024e-05,
+      "loss": 0.9397,
+      "step": 6422
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8538094758987427,
+      "learning_rate": 1.4573306866327702e-05,
+      "loss": 1.0108,
+      "step": 6423
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6468374729156494,
+      "learning_rate": 1.4571654819586334e-05,
+      "loss": 1.0046,
+      "step": 6424
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8968156576156616,
+      "learning_rate": 1.4570002615089924e-05,
+      "loss": 1.0159,
+      "step": 6425
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8063596487045288,
+      "learning_rate": 1.4568350252895494e-05,
+      "loss": 1.0217,
+      "step": 6426
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.019664764404297,
+      "learning_rate": 1.4566697733060057e-05,
+      "loss": 0.9998,
+      "step": 6427
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7261220216751099,
+      "learning_rate": 1.4565045055640639e-05,
+      "loss": 0.9801,
+      "step": 6428
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1193735599517822,
+      "learning_rate": 1.4563392220694265e-05,
+      "loss": 0.5985,
+      "step": 6429
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9723857641220093,
+      "learning_rate": 1.4561739228277976e-05,
+      "loss": 0.9833,
+      "step": 6430
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8732644319534302,
+      "learning_rate": 1.4560086078448807e-05,
+      "loss": 1.0046,
+      "step": 6431
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.885548710823059,
+      "learning_rate": 1.4558432771263806e-05,
+      "loss": 1.0395,
+      "step": 6432
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9934170246124268,
+      "learning_rate": 1.4556779306780024e-05,
+      "loss": 1.0837,
+      "step": 6433
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.682637333869934,
+      "learning_rate": 1.4555125685054519e-05,
+      "loss": 0.9274,
+      "step": 6434
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8780604600906372,
+      "learning_rate": 1.4553471906144347e-05,
+      "loss": 1.0211,
+      "step": 6435
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.812703251838684,
+      "learning_rate": 1.455181797010658e-05,
+      "loss": 0.9943,
+      "step": 6436
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7835032939910889,
+      "learning_rate": 1.4550163876998288e-05,
+      "loss": 1.0734,
+      "step": 6437
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8049976825714111,
+      "learning_rate": 1.4548509626876554e-05,
+      "loss": 1.0012,
+      "step": 6438
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6568845510482788,
+      "learning_rate": 1.454685521979846e-05,
+      "loss": 1.0279,
+      "step": 6439
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.812314510345459,
+      "learning_rate": 1.454520065582109e-05,
+      "loss": 0.9825,
+      "step": 6440
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9423649311065674,
+      "learning_rate": 1.4543545935001544e-05,
+      "loss": 1.0251,
+      "step": 6441
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.859851360321045,
+      "learning_rate": 1.4541891057396917e-05,
+      "loss": 1.012,
+      "step": 6442
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8464343547821045,
+      "learning_rate": 1.454023602306432e-05,
+      "loss": 0.9928,
+      "step": 6443
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9029483795166016,
+      "learning_rate": 1.4538580832060861e-05,
+      "loss": 1.025,
+      "step": 6444
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5566385984420776,
+      "learning_rate": 1.4536925484443653e-05,
+      "loss": 0.9894,
+      "step": 6445
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5982820987701416,
+      "learning_rate": 1.4535269980269822e-05,
+      "loss": 0.9552,
+      "step": 6446
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.888586401939392,
+      "learning_rate": 1.4533614319596489e-05,
+      "loss": 1.0156,
+      "step": 6447
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7846451997756958,
+      "learning_rate": 1.4531958502480794e-05,
+      "loss": 0.8955,
+      "step": 6448
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7921074628829956,
+      "learning_rate": 1.4530302528979868e-05,
+      "loss": 1.0495,
+      "step": 6449
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.711888313293457,
+      "learning_rate": 1.4528646399150857e-05,
+      "loss": 0.9183,
+      "step": 6450
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.621599793434143,
+      "learning_rate": 1.452699011305091e-05,
+      "loss": 1.0458,
+      "step": 6451
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.770071029663086,
+      "learning_rate": 1.4525333670737181e-05,
+      "loss": 1.0048,
+      "step": 6452
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7624177932739258,
+      "learning_rate": 1.4523677072266825e-05,
+      "loss": 0.956,
+      "step": 6453
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8492014408111572,
+      "learning_rate": 1.452202031769701e-05,
+      "loss": 1.0619,
+      "step": 6454
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7370258569717407,
+      "learning_rate": 1.4520363407084905e-05,
+      "loss": 1.0078,
+      "step": 6455
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7034357786178589,
+      "learning_rate": 1.4518706340487689e-05,
+      "loss": 0.9863,
+      "step": 6456
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.729231834411621,
+      "learning_rate": 1.4517049117962539e-05,
+      "loss": 1.039,
+      "step": 6457
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7120835781097412,
+      "learning_rate": 1.4515391739566642e-05,
+      "loss": 0.9876,
+      "step": 6458
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7186849117279053,
+      "learning_rate": 1.4513734205357186e-05,
+      "loss": 0.9244,
+      "step": 6459
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9158945083618164,
+      "learning_rate": 1.4512076515391375e-05,
+      "loss": 0.9648,
+      "step": 6460
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0482406616210938,
+      "learning_rate": 1.4510418669726407e-05,
+      "loss": 0.946,
+      "step": 6461
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9669641256332397,
+      "learning_rate": 1.4508760668419489e-05,
+      "loss": 0.9763,
+      "step": 6462
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8329441547393799,
+      "learning_rate": 1.4507102511527834e-05,
+      "loss": 1.0406,
+      "step": 6463
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0751601457595825,
+      "learning_rate": 1.4505444199108662e-05,
+      "loss": 0.6183,
+      "step": 6464
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8225510120391846,
+      "learning_rate": 1.4503785731219195e-05,
+      "loss": 0.9144,
+      "step": 6465
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6742932796478271,
+      "learning_rate": 1.4502127107916666e-05,
+      "loss": 0.973,
+      "step": 6466
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7082477807998657,
+      "learning_rate": 1.4500468329258305e-05,
+      "loss": 1.0539,
+      "step": 6467
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8570054769515991,
+      "learning_rate": 1.4498809395301356e-05,
+      "loss": 1.0599,
+      "step": 6468
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7183752059936523,
+      "learning_rate": 1.4497150306103061e-05,
+      "loss": 0.9962,
+      "step": 6469
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8262039422988892,
+      "learning_rate": 1.4495491061720671e-05,
+      "loss": 0.9805,
+      "step": 6470
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7520779371261597,
+      "learning_rate": 1.4493831662211439e-05,
+      "loss": 0.9918,
+      "step": 6471
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.983081579208374,
+      "learning_rate": 1.449217210763263e-05,
+      "loss": 1.0585,
+      "step": 6472
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.738904356956482,
+      "learning_rate": 1.4490512398041515e-05,
+      "loss": 0.9685,
+      "step": 6473
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8170249462127686,
+      "learning_rate": 1.4488852533495357e-05,
+      "loss": 0.9681,
+      "step": 6474
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8998768329620361,
+      "learning_rate": 1.4487192514051437e-05,
+      "loss": 1.015,
+      "step": 6475
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7171269655227661,
+      "learning_rate": 1.4485532339767036e-05,
+      "loss": 0.9578,
+      "step": 6476
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7854276895523071,
+      "learning_rate": 1.4483872010699446e-05,
+      "loss": 1.0138,
+      "step": 6477
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6501131057739258,
+      "learning_rate": 1.448221152690596e-05,
+      "loss": 1.0863,
+      "step": 6478
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7809715270996094,
+      "learning_rate": 1.4480550888443871e-05,
+      "loss": 0.9552,
+      "step": 6479
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7217483520507812,
+      "learning_rate": 1.4478890095370491e-05,
+      "loss": 0.9768,
+      "step": 6480
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8227529525756836,
+      "learning_rate": 1.447722914774312e-05,
+      "loss": 0.9444,
+      "step": 6481
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9200953245162964,
+      "learning_rate": 1.4475568045619084e-05,
+      "loss": 1.1285,
+      "step": 6482
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7601783275604248,
+      "learning_rate": 1.4473906789055692e-05,
+      "loss": 0.9523,
+      "step": 6483
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.842558741569519,
+      "learning_rate": 1.4472245378110276e-05,
+      "loss": 0.942,
+      "step": 6484
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9267534017562866,
+      "learning_rate": 1.4470583812840164e-05,
+      "loss": 0.9987,
+      "step": 6485
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.05393123626709,
+      "learning_rate": 1.4468922093302693e-05,
+      "loss": 1.0468,
+      "step": 6486
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6043117046356201,
+      "learning_rate": 1.44672602195552e-05,
+      "loss": 0.954,
+      "step": 6487
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1881109476089478,
+      "learning_rate": 1.4465598191655042e-05,
+      "loss": 0.6342,
+      "step": 6488
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7878910303115845,
+      "learning_rate": 1.4463936009659563e-05,
+      "loss": 1.0544,
+      "step": 6489
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8502981662750244,
+      "learning_rate": 1.446227367362612e-05,
+      "loss": 0.9376,
+      "step": 6490
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0396090745925903,
+      "learning_rate": 1.4460611183612074e-05,
+      "loss": 0.526,
+      "step": 6491
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8148415088653564,
+      "learning_rate": 1.4458948539674802e-05,
+      "loss": 0.9569,
+      "step": 6492
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.768983244895935,
+      "learning_rate": 1.445728574187167e-05,
+      "loss": 0.9423,
+      "step": 6493
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.862540364265442,
+      "learning_rate": 1.4455622790260057e-05,
+      "loss": 1.0599,
+      "step": 6494
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8864184617996216,
+      "learning_rate": 1.445395968489735e-05,
+      "loss": 1.0239,
+      "step": 6495
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8307377099990845,
+      "learning_rate": 1.4452296425840935e-05,
+      "loss": 1.0407,
+      "step": 6496
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.954124093055725,
+      "learning_rate": 1.4450633013148205e-05,
+      "loss": 0.963,
+      "step": 6497
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7054812908172607,
+      "learning_rate": 1.4448969446876567e-05,
+      "loss": 0.9572,
+      "step": 6498
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7556042671203613,
+      "learning_rate": 1.4447305727083416e-05,
+      "loss": 1.0189,
+      "step": 6499
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8345447778701782,
+      "learning_rate": 1.4445641853826172e-05,
+      "loss": 0.9704,
+      "step": 6500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.005666732788086,
+      "learning_rate": 1.4443977827162242e-05,
+      "loss": 1.0405,
+      "step": 6501
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.002800941467285,
+      "learning_rate": 1.4442313647149053e-05,
+      "loss": 0.952,
+      "step": 6502
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8509241342544556,
+      "learning_rate": 1.4440649313844026e-05,
+      "loss": 1.0527,
+      "step": 6503
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7495306730270386,
+      "learning_rate": 1.44389848273046e-05,
+      "loss": 0.9529,
+      "step": 6504
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.722464919090271,
+      "learning_rate": 1.4437320187588204e-05,
+      "loss": 0.9982,
+      "step": 6505
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8324607610702515,
+      "learning_rate": 1.4435655394752287e-05,
+      "loss": 0.9718,
+      "step": 6506
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6986242532730103,
+      "learning_rate": 1.4433990448854289e-05,
+      "loss": 0.9991,
+      "step": 6507
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8668314218521118,
+      "learning_rate": 1.4432325349951668e-05,
+      "loss": 0.9255,
+      "step": 6508
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.831986665725708,
+      "learning_rate": 1.443066009810188e-05,
+      "loss": 1.0221,
+      "step": 6509
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.221524715423584,
+      "learning_rate": 1.442899469336239e-05,
+      "loss": 1.0437,
+      "step": 6510
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8463190793991089,
+      "learning_rate": 1.4427329135790667e-05,
+      "loss": 1.0553,
+      "step": 6511
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7398473024368286,
+      "learning_rate": 1.4425663425444179e-05,
+      "loss": 1.0165,
+      "step": 6512
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7463241815567017,
+      "learning_rate": 1.442399756238041e-05,
+      "loss": 1.0556,
+      "step": 6513
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7856740951538086,
+      "learning_rate": 1.4422331546656842e-05,
+      "loss": 1.0234,
+      "step": 6514
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.804658055305481,
+      "learning_rate": 1.4420665378330964e-05,
+      "loss": 0.9906,
+      "step": 6515
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9732418060302734,
+      "learning_rate": 1.4418999057460277e-05,
+      "loss": 1.0376,
+      "step": 6516
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8557568788528442,
+      "learning_rate": 1.4417332584102273e-05,
+      "loss": 1.0359,
+      "step": 6517
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0801892280578613,
+      "learning_rate": 1.4415665958314465e-05,
+      "loss": 1.0006,
+      "step": 6518
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.638874888420105,
+      "learning_rate": 1.4413999180154355e-05,
+      "loss": 1.0863,
+      "step": 6519
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7250996828079224,
+      "learning_rate": 1.4412332249679463e-05,
+      "loss": 1.0019,
+      "step": 6520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8014070987701416,
+      "learning_rate": 1.4410665166947312e-05,
+      "loss": 1.0091,
+      "step": 6521
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0862135887145996,
+      "learning_rate": 1.4408997932015428e-05,
+      "loss": 1.0418,
+      "step": 6522
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7028090953826904,
+      "learning_rate": 1.440733054494134e-05,
+      "loss": 0.98,
+      "step": 6523
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.893875002861023,
+      "learning_rate": 1.440566300578259e-05,
+      "loss": 1.0148,
+      "step": 6524
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.834351658821106,
+      "learning_rate": 1.4403995314596711e-05,
+      "loss": 1.0467,
+      "step": 6525
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9340988397598267,
+      "learning_rate": 1.4402327471441257e-05,
+      "loss": 1.0258,
+      "step": 6526
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6812983751296997,
+      "learning_rate": 1.4400659476373781e-05,
+      "loss": 0.9937,
+      "step": 6527
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.060377359390259,
+      "learning_rate": 1.439899132945184e-05,
+      "loss": 1.0227,
+      "step": 6528
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7195736169815063,
+      "learning_rate": 1.4397323030732994e-05,
+      "loss": 1.0075,
+      "step": 6529
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.198493242263794,
+      "learning_rate": 1.4395654580274814e-05,
+      "loss": 0.5961,
+      "step": 6530
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8344513177871704,
+      "learning_rate": 1.4393985978134874e-05,
+      "loss": 1.1127,
+      "step": 6531
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5952489376068115,
+      "learning_rate": 1.4392317224370751e-05,
+      "loss": 0.9578,
+      "step": 6532
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8457087278366089,
+      "learning_rate": 1.4390648319040032e-05,
+      "loss": 1.0494,
+      "step": 6533
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8811098337173462,
+      "learning_rate": 1.4388979262200302e-05,
+      "loss": 1.0539,
+      "step": 6534
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7097623348236084,
+      "learning_rate": 1.438731005390916e-05,
+      "loss": 0.9877,
+      "step": 6535
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8394737243652344,
+      "learning_rate": 1.4385640694224203e-05,
+      "loss": 1.0304,
+      "step": 6536
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7272309064865112,
+      "learning_rate": 1.4383971183203036e-05,
+      "loss": 1.0012,
+      "step": 6537
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0144572257995605,
+      "learning_rate": 1.4382301520903267e-05,
+      "loss": 0.9956,
+      "step": 6538
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7602953910827637,
+      "learning_rate": 1.4380631707382517e-05,
+      "loss": 0.9671,
+      "step": 6539
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8607141971588135,
+      "learning_rate": 1.4378961742698403e-05,
+      "loss": 1.0021,
+      "step": 6540
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8680787086486816,
+      "learning_rate": 1.4377291626908552e-05,
+      "loss": 0.9664,
+      "step": 6541
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6423228979110718,
+      "learning_rate": 1.4375621360070593e-05,
+      "loss": 0.9743,
+      "step": 6542
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8883764743804932,
+      "learning_rate": 1.4373950942242166e-05,
+      "loss": 0.9784,
+      "step": 6543
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9475061893463135,
+      "learning_rate": 1.4372280373480907e-05,
+      "loss": 1.0052,
+      "step": 6544
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9167377948760986,
+      "learning_rate": 1.4370609653844467e-05,
+      "loss": 0.9683,
+      "step": 6545
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.782257318496704,
+      "learning_rate": 1.4368938783390498e-05,
+      "loss": 0.9838,
+      "step": 6546
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8462469577789307,
+      "learning_rate": 1.4367267762176655e-05,
+      "loss": 1.0275,
+      "step": 6547
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7571134567260742,
+      "learning_rate": 1.43655965902606e-05,
+      "loss": 0.9948,
+      "step": 6548
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.936461091041565,
+      "learning_rate": 1.4363925267700003e-05,
+      "loss": 1.046,
+      "step": 6549
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8618601560592651,
+      "learning_rate": 1.4362253794552534e-05,
+      "loss": 1.0277,
+      "step": 6550
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6052945852279663,
+      "learning_rate": 1.4360582170875872e-05,
+      "loss": 0.9506,
+      "step": 6551
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8968398571014404,
+      "learning_rate": 1.4358910396727701e-05,
+      "loss": 0.9646,
+      "step": 6552
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9234265089035034,
+      "learning_rate": 1.4357238472165707e-05,
+      "loss": 0.9882,
+      "step": 6553
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7836437225341797,
+      "learning_rate": 1.4355566397247584e-05,
+      "loss": 0.9707,
+      "step": 6554
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2310750484466553,
+      "learning_rate": 1.4353894172031033e-05,
+      "loss": 1.0473,
+      "step": 6555
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0779621601104736,
+      "learning_rate": 1.4352221796573758e-05,
+      "loss": 0.9836,
+      "step": 6556
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8310410976409912,
+      "learning_rate": 1.4350549270933463e-05,
+      "loss": 1.0107,
+      "step": 6557
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7569674253463745,
+      "learning_rate": 1.4348876595167865e-05,
+      "loss": 0.9478,
+      "step": 6558
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3535518646240234,
+      "learning_rate": 1.4347203769334685e-05,
+      "loss": 0.643,
+      "step": 6559
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6123547554016113,
+      "learning_rate": 1.434553079349165e-05,
+      "loss": 1.0544,
+      "step": 6560
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4126694202423096,
+      "learning_rate": 1.4343857667696481e-05,
+      "loss": 0.9932,
+      "step": 6561
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9189002513885498,
+      "learning_rate": 1.4342184392006922e-05,
+      "loss": 0.9629,
+      "step": 6562
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8644987344741821,
+      "learning_rate": 1.4340510966480707e-05,
+      "loss": 1.0123,
+      "step": 6563
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.769840121269226,
+      "learning_rate": 1.4338837391175582e-05,
+      "loss": 0.975,
+      "step": 6564
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8416407108306885,
+      "learning_rate": 1.4337163666149301e-05,
+      "loss": 0.9625,
+      "step": 6565
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9921787977218628,
+      "learning_rate": 1.4335489791459615e-05,
+      "loss": 0.6578,
+      "step": 6566
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8825032711029053,
+      "learning_rate": 1.433381576716429e-05,
+      "loss": 1.0717,
+      "step": 6567
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7608774900436401,
+      "learning_rate": 1.4332141593321087e-05,
+      "loss": 0.9692,
+      "step": 6568
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8526568412780762,
+      "learning_rate": 1.4330467269987778e-05,
+      "loss": 1.0264,
+      "step": 6569
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2145869731903076,
+      "learning_rate": 1.4328792797222142e-05,
+      "loss": 1.0822,
+      "step": 6570
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8882297277450562,
+      "learning_rate": 1.4327118175081957e-05,
+      "loss": 1.0456,
+      "step": 6571
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6885595321655273,
+      "learning_rate": 1.4325443403625012e-05,
+      "loss": 1.0059,
+      "step": 6572
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6838698387145996,
+      "learning_rate": 1.4323768482909097e-05,
+      "loss": 0.9376,
+      "step": 6573
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.997868299484253,
+      "learning_rate": 1.432209341299201e-05,
+      "loss": 0.9534,
+      "step": 6574
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8583295345306396,
+      "learning_rate": 1.4320418193931556e-05,
+      "loss": 1.1367,
+      "step": 6575
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7825112342834473,
+      "learning_rate": 1.4318742825785535e-05,
+      "loss": 1.0238,
+      "step": 6576
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.830778956413269,
+      "learning_rate": 1.4317067308611762e-05,
+      "loss": 1.0095,
+      "step": 6577
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7713563442230225,
+      "learning_rate": 1.4315391642468059e-05,
+      "loss": 1.0108,
+      "step": 6578
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8108627796173096,
+      "learning_rate": 1.4313715827412243e-05,
+      "loss": 0.9999,
+      "step": 6579
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9374496936798096,
+      "learning_rate": 1.4312039863502145e-05,
+      "loss": 0.9924,
+      "step": 6580
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8295040130615234,
+      "learning_rate": 1.4310363750795593e-05,
+      "loss": 1.0622,
+      "step": 6581
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.069732666015625,
+      "learning_rate": 1.430868748935043e-05,
+      "loss": 0.596,
+      "step": 6582
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.973830223083496,
+      "learning_rate": 1.4307011079224498e-05,
+      "loss": 1.0789,
+      "step": 6583
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6908007860183716,
+      "learning_rate": 1.4305334520475647e-05,
+      "loss": 0.8879,
+      "step": 6584
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.850256085395813,
+      "learning_rate": 1.4303657813161725e-05,
+      "loss": 1.0637,
+      "step": 6585
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7185535430908203,
+      "learning_rate": 1.4301980957340593e-05,
+      "loss": 0.9622,
+      "step": 6586
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8960317373275757,
+      "learning_rate": 1.4300303953070118e-05,
+      "loss": 0.9339,
+      "step": 6587
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9350983500480652,
+      "learning_rate": 1.4298626800408166e-05,
+      "loss": 0.5962,
+      "step": 6588
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9939955472946167,
+      "learning_rate": 1.4296949499412609e-05,
+      "loss": 1.0499,
+      "step": 6589
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7241488695144653,
+      "learning_rate": 1.4295272050141329e-05,
+      "loss": 0.9086,
+      "step": 6590
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9936388731002808,
+      "learning_rate": 1.4293594452652212e-05,
+      "loss": 1.0211,
+      "step": 6591
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8468677997589111,
+      "learning_rate": 1.429191670700314e-05,
+      "loss": 1.0188,
+      "step": 6592
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.759409785270691,
+      "learning_rate": 1.4290238813252011e-05,
+      "loss": 1.0064,
+      "step": 6593
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.863101840019226,
+      "learning_rate": 1.4288560771456727e-05,
+      "loss": 1.005,
+      "step": 6594
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7478598356246948,
+      "learning_rate": 1.428688258167519e-05,
+      "loss": 0.9722,
+      "step": 6595
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8459408283233643,
+      "learning_rate": 1.4285204243965307e-05,
+      "loss": 1.0568,
+      "step": 6596
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.86139714717865,
+      "learning_rate": 1.4283525758385e-05,
+      "loss": 0.9188,
+      "step": 6597
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7600229978561401,
+      "learning_rate": 1.4281847124992181e-05,
+      "loss": 0.9026,
+      "step": 6598
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7581877708435059,
+      "learning_rate": 1.428016834384478e-05,
+      "loss": 0.9924,
+      "step": 6599
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7071934938430786,
+      "learning_rate": 1.4278489415000727e-05,
+      "loss": 1.0168,
+      "step": 6600
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8092888593673706,
+      "learning_rate": 1.4276810338517955e-05,
+      "loss": 1.0107,
+      "step": 6601
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.741211175918579,
+      "learning_rate": 1.4275131114454405e-05,
+      "loss": 1.008,
+      "step": 6602
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9269744157791138,
+      "learning_rate": 1.4273451742868023e-05,
+      "loss": 1.0116,
+      "step": 6603
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9591504335403442,
+      "learning_rate": 1.4271772223816758e-05,
+      "loss": 1.0416,
+      "step": 6604
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.297698736190796,
+      "learning_rate": 1.4270092557358566e-05,
+      "loss": 1.0133,
+      "step": 6605
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0246243476867676,
+      "learning_rate": 1.426841274355141e-05,
+      "loss": 1.0371,
+      "step": 6606
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7749741077423096,
+      "learning_rate": 1.4266732782453252e-05,
+      "loss": 1.0152,
+      "step": 6607
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7851496934890747,
+      "learning_rate": 1.4265052674122063e-05,
+      "loss": 0.9268,
+      "step": 6608
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0950188636779785,
+      "learning_rate": 1.426337241861582e-05,
+      "loss": 0.9803,
+      "step": 6609
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8657783269882202,
+      "learning_rate": 1.4261692015992505e-05,
+      "loss": 1.0346,
+      "step": 6610
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8111631870269775,
+      "learning_rate": 1.4260011466310104e-05,
+      "loss": 0.9339,
+      "step": 6611
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.068686008453369,
+      "learning_rate": 1.4258330769626607e-05,
+      "loss": 1.0543,
+      "step": 6612
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8311845064163208,
+      "learning_rate": 1.425664992600001e-05,
+      "loss": 0.9377,
+      "step": 6613
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7941935062408447,
+      "learning_rate": 1.4254968935488314e-05,
+      "loss": 0.9926,
+      "step": 6614
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.934528112411499,
+      "learning_rate": 1.4253287798149526e-05,
+      "loss": 0.9832,
+      "step": 6615
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.912278175354004,
+      "learning_rate": 1.4251606514041659e-05,
+      "loss": 1.0557,
+      "step": 6616
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6515284776687622,
+      "learning_rate": 1.4249925083222724e-05,
+      "loss": 0.9598,
+      "step": 6617
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.774816870689392,
+      "learning_rate": 1.424824350575075e-05,
+      "loss": 1.0422,
+      "step": 6618
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.816463828086853,
+      "learning_rate": 1.4246561781683754e-05,
+      "loss": 1.0368,
+      "step": 6619
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.824295163154602,
+      "learning_rate": 1.4244879911079779e-05,
+      "loss": 0.9043,
+      "step": 6620
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8551970720291138,
+      "learning_rate": 1.4243197893996855e-05,
+      "loss": 0.9994,
+      "step": 6621
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8014181852340698,
+      "learning_rate": 1.4241515730493021e-05,
+      "loss": 1.0175,
+      "step": 6622
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9140993356704712,
+      "learning_rate": 1.4239833420626328e-05,
+      "loss": 0.9517,
+      "step": 6623
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9484046697616577,
+      "learning_rate": 1.423815096445483e-05,
+      "loss": 1.0351,
+      "step": 6624
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.500910758972168,
+      "learning_rate": 1.423646836203658e-05,
+      "loss": 0.953,
+      "step": 6625
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5796787738800049,
+      "learning_rate": 1.423478561342964e-05,
+      "loss": 0.939,
+      "step": 6626
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.881302833557129,
+      "learning_rate": 1.4233102718692078e-05,
+      "loss": 1.0325,
+      "step": 6627
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8480826616287231,
+      "learning_rate": 1.4231419677881966e-05,
+      "loss": 0.9877,
+      "step": 6628
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8555617332458496,
+      "learning_rate": 1.4229736491057382e-05,
+      "loss": 1.027,
+      "step": 6629
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8782488107681274,
+      "learning_rate": 1.4228053158276407e-05,
+      "loss": 0.9052,
+      "step": 6630
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.79643714427948,
+      "learning_rate": 1.4226369679597127e-05,
+      "loss": 0.9605,
+      "step": 6631
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7273995876312256,
+      "learning_rate": 1.4224686055077636e-05,
+      "loss": 0.9487,
+      "step": 6632
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8837653398513794,
+      "learning_rate": 1.422300228477603e-05,
+      "loss": 1.0666,
+      "step": 6633
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7214385271072388,
+      "learning_rate": 1.4221318368750411e-05,
+      "loss": 0.9393,
+      "step": 6634
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8205844163894653,
+      "learning_rate": 1.4219634307058888e-05,
+      "loss": 0.973,
+      "step": 6635
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.814313292503357,
+      "learning_rate": 1.421795009975957e-05,
+      "loss": 0.9621,
+      "step": 6636
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6666721105575562,
+      "learning_rate": 1.4216265746910579e-05,
+      "loss": 0.9338,
+      "step": 6637
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8279637098312378,
+      "learning_rate": 1.4214581248570034e-05,
+      "loss": 0.9916,
+      "step": 6638
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8509894609451294,
+      "learning_rate": 1.4212896604796064e-05,
+      "loss": 1.0023,
+      "step": 6639
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7838656902313232,
+      "learning_rate": 1.42112118156468e-05,
+      "loss": 0.9304,
+      "step": 6640
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7575708627700806,
+      "learning_rate": 1.4209526881180377e-05,
+      "loss": 0.9876,
+      "step": 6641
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1256905794143677,
+      "learning_rate": 1.4207841801454945e-05,
+      "loss": 0.6455,
+      "step": 6642
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.654044508934021,
+      "learning_rate": 1.4206156576528643e-05,
+      "loss": 0.9592,
+      "step": 6643
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.731591820716858,
+      "learning_rate": 1.4204471206459629e-05,
+      "loss": 0.9452,
+      "step": 6644
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7719027996063232,
+      "learning_rate": 1.4202785691306056e-05,
+      "loss": 1.0676,
+      "step": 6645
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.768100619316101,
+      "learning_rate": 1.4201100031126091e-05,
+      "loss": 0.9503,
+      "step": 6646
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.136711597442627,
+      "learning_rate": 1.4199414225977897e-05,
+      "loss": 1.0227,
+      "step": 6647
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.775230884552002,
+      "learning_rate": 1.4197728275919649e-05,
+      "loss": 1.0366,
+      "step": 6648
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9589234590530396,
+      "learning_rate": 1.4196042181009525e-05,
+      "loss": 1.0704,
+      "step": 6649
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8622713088989258,
+      "learning_rate": 1.4194355941305706e-05,
+      "loss": 1.0001,
+      "step": 6650
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9297192096710205,
+      "learning_rate": 1.419266955686638e-05,
+      "loss": 1.0856,
+      "step": 6651
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7351832389831543,
+      "learning_rate": 1.419098302774974e-05,
+      "loss": 0.9397,
+      "step": 6652
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7967897653579712,
+      "learning_rate": 1.4189296354013982e-05,
+      "loss": 0.9777,
+      "step": 6653
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6718227863311768,
+      "learning_rate": 1.418760953571731e-05,
+      "loss": 1.0362,
+      "step": 6654
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8721386194229126,
+      "learning_rate": 1.4185922572917931e-05,
+      "loss": 0.8875,
+      "step": 6655
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6840529441833496,
+      "learning_rate": 1.4184235465674055e-05,
+      "loss": 0.9631,
+      "step": 6656
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6847478151321411,
+      "learning_rate": 1.4182548214043902e-05,
+      "loss": 1.0208,
+      "step": 6657
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7295044660568237,
+      "learning_rate": 1.4180860818085695e-05,
+      "loss": 0.867,
+      "step": 6658
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.013784408569336,
+      "learning_rate": 1.4179173277857658e-05,
+      "loss": 0.9899,
+      "step": 6659
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6859959363937378,
+      "learning_rate": 1.4177485593418028e-05,
+      "loss": 0.9711,
+      "step": 6660
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.724387288093567,
+      "learning_rate": 1.4175797764825036e-05,
+      "loss": 1.0283,
+      "step": 6661
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8912831544876099,
+      "learning_rate": 1.417410979213693e-05,
+      "loss": 1.0973,
+      "step": 6662
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7722456455230713,
+      "learning_rate": 1.4172421675411954e-05,
+      "loss": 1.0363,
+      "step": 6663
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0262619256973267,
+      "learning_rate": 1.4170733414708363e-05,
+      "loss": 0.6007,
+      "step": 6664
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8084995746612549,
+      "learning_rate": 1.4169045010084411e-05,
+      "loss": 0.9921,
+      "step": 6665
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0457794666290283,
+      "learning_rate": 1.416735646159836e-05,
+      "loss": 1.0544,
+      "step": 6666
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8310672044754028,
+      "learning_rate": 1.416566776930848e-05,
+      "loss": 1.0378,
+      "step": 6667
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8013845682144165,
+      "learning_rate": 1.4163978933273041e-05,
+      "loss": 0.9975,
+      "step": 6668
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7868287563323975,
+      "learning_rate": 1.4162289953550322e-05,
+      "loss": 0.9903,
+      "step": 6669
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.899398922920227,
+      "learning_rate": 1.4160600830198602e-05,
+      "loss": 0.9949,
+      "step": 6670
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8585964441299438,
+      "learning_rate": 1.4158911563276172e-05,
+      "loss": 0.9943,
+      "step": 6671
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8636623620986938,
+      "learning_rate": 1.415722215284132e-05,
+      "loss": 1.0294,
+      "step": 6672
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.705580472946167,
+      "learning_rate": 1.4155532598952345e-05,
+      "loss": 0.9942,
+      "step": 6673
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.713517665863037,
+      "learning_rate": 1.4153842901667543e-05,
+      "loss": 1.0279,
+      "step": 6674
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.743633508682251,
+      "learning_rate": 1.415215306104523e-05,
+      "loss": 0.9224,
+      "step": 6675
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8368037939071655,
+      "learning_rate": 1.4150463077143712e-05,
+      "loss": 0.995,
+      "step": 6676
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9264042377471924,
+      "learning_rate": 1.414877295002131e-05,
+      "loss": 0.9746,
+      "step": 6677
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.871768832206726,
+      "learning_rate": 1.4147082679736342e-05,
+      "loss": 0.9933,
+      "step": 6678
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6611647605895996,
+      "learning_rate": 1.4145392266347133e-05,
+      "loss": 1.0094,
+      "step": 6679
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.935665249824524,
+      "learning_rate": 1.4143701709912017e-05,
+      "loss": 1.0416,
+      "step": 6680
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9268814325332642,
+      "learning_rate": 1.4142011010489332e-05,
+      "loss": 0.9669,
+      "step": 6681
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.011317014694214,
+      "learning_rate": 1.4140320168137414e-05,
+      "loss": 1.0625,
+      "step": 6682
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7357138395309448,
+      "learning_rate": 1.4138629182914617e-05,
+      "loss": 1.0074,
+      "step": 6683
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8733261823654175,
+      "learning_rate": 1.4136938054879284e-05,
+      "loss": 1.0088,
+      "step": 6684
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8755565881729126,
+      "learning_rate": 1.4135246784089774e-05,
+      "loss": 0.957,
+      "step": 6685
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7794195413589478,
+      "learning_rate": 1.413355537060445e-05,
+      "loss": 0.9682,
+      "step": 6686
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5756385326385498,
+      "learning_rate": 1.413186381448168e-05,
+      "loss": 0.9283,
+      "step": 6687
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9130840301513672,
+      "learning_rate": 1.4130172115779828e-05,
+      "loss": 1.1402,
+      "step": 6688
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7617560625076294,
+      "learning_rate": 1.412848027455727e-05,
+      "loss": 0.9422,
+      "step": 6689
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7759631872177124,
+      "learning_rate": 1.4126788290872395e-05,
+      "loss": 0.9418,
+      "step": 6690
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7937480211257935,
+      "learning_rate": 1.4125096164783586e-05,
+      "loss": 0.9844,
+      "step": 6691
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6398866176605225,
+      "learning_rate": 1.4123403896349227e-05,
+      "loss": 0.9924,
+      "step": 6692
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7397969961166382,
+      "learning_rate": 1.412171148562772e-05,
+      "loss": 0.9804,
+      "step": 6693
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4397199153900146,
+      "learning_rate": 1.4120018932677461e-05,
+      "loss": 1.0353,
+      "step": 6694
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7143027782440186,
+      "learning_rate": 1.411832623755686e-05,
+      "loss": 0.9389,
+      "step": 6695
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8062744140625,
+      "learning_rate": 1.4116633400324325e-05,
+      "loss": 0.9498,
+      "step": 6696
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8488270044326782,
+      "learning_rate": 1.411494042103827e-05,
+      "loss": 0.9168,
+      "step": 6697
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7461563348770142,
+      "learning_rate": 1.4113247299757116e-05,
+      "loss": 0.864,
+      "step": 6698
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8032582998275757,
+      "learning_rate": 1.4111554036539285e-05,
+      "loss": 1.0144,
+      "step": 6699
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9199655055999756,
+      "learning_rate": 1.4109860631443214e-05,
+      "loss": 1.0324,
+      "step": 6700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8765710592269897,
+      "learning_rate": 1.4108167084527327e-05,
+      "loss": 1.0288,
+      "step": 6701
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9236133098602295,
+      "learning_rate": 1.4106473395850073e-05,
+      "loss": 1.003,
+      "step": 6702
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9396934509277344,
+      "learning_rate": 1.4104779565469892e-05,
+      "loss": 1.0093,
+      "step": 6703
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7690138816833496,
+      "learning_rate": 1.4103085593445236e-05,
+      "loss": 0.9209,
+      "step": 6704
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8163502216339111,
+      "learning_rate": 1.4101391479834558e-05,
+      "loss": 1.0551,
+      "step": 6705
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6216959953308105,
+      "learning_rate": 1.4099697224696316e-05,
+      "loss": 1.0234,
+      "step": 6706
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9014475345611572,
+      "learning_rate": 1.4098002828088974e-05,
+      "loss": 1.0392,
+      "step": 6707
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1175824403762817,
+      "learning_rate": 1.4096308290071003e-05,
+      "loss": 0.603,
+      "step": 6708
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0823798179626465,
+      "learning_rate": 1.4094613610700876e-05,
+      "loss": 0.6117,
+      "step": 6709
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9367735385894775,
+      "learning_rate": 1.4092918790037069e-05,
+      "loss": 1.0192,
+      "step": 6710
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0108258724212646,
+      "learning_rate": 1.4091223828138068e-05,
+      "loss": 1.0669,
+      "step": 6711
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.77130126953125,
+      "learning_rate": 1.4089528725062362e-05,
+      "loss": 1.011,
+      "step": 6712
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7122178077697754,
+      "learning_rate": 1.4087833480868442e-05,
+      "loss": 0.9043,
+      "step": 6713
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.718496322631836,
+      "learning_rate": 1.4086138095614808e-05,
+      "loss": 0.9639,
+      "step": 6714
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8527882099151611,
+      "learning_rate": 1.4084442569359964e-05,
+      "loss": 0.926,
+      "step": 6715
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9653902053833008,
+      "learning_rate": 1.4082746902162414e-05,
+      "loss": 1.0,
+      "step": 6716
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9845950603485107,
+      "learning_rate": 1.4081051094080675e-05,
+      "loss": 0.9188,
+      "step": 6717
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8779118061065674,
+      "learning_rate": 1.407935514517326e-05,
+      "loss": 1.0235,
+      "step": 6718
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.11811363697052,
+      "learning_rate": 1.4077659055498695e-05,
+      "loss": 0.6848,
+      "step": 6719
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.69560706615448,
+      "learning_rate": 1.407596282511551e-05,
+      "loss": 1.006,
+      "step": 6720
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7870867252349854,
+      "learning_rate": 1.407426645408223e-05,
+      "loss": 1.003,
+      "step": 6721
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.717315673828125,
+      "learning_rate": 1.4072569942457399e-05,
+      "loss": 0.9686,
+      "step": 6722
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.03489351272583,
+      "learning_rate": 1.4070873290299554e-05,
+      "loss": 0.9558,
+      "step": 6723
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8468509912490845,
+      "learning_rate": 1.4069176497667242e-05,
+      "loss": 1.016,
+      "step": 6724
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8713451623916626,
+      "learning_rate": 1.4067479564619018e-05,
+      "loss": 0.9966,
+      "step": 6725
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.603393793106079,
+      "learning_rate": 1.4065782491213433e-05,
+      "loss": 0.9698,
+      "step": 6726
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.21260929107666,
+      "learning_rate": 1.4064085277509055e-05,
+      "loss": 1.0467,
+      "step": 6727
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6719169616699219,
+      "learning_rate": 1.4062387923564443e-05,
+      "loss": 0.9636,
+      "step": 6728
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8306374549865723,
+      "learning_rate": 1.4060690429438176e-05,
+      "loss": 0.9977,
+      "step": 6729
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0939902067184448,
+      "learning_rate": 1.4058992795188822e-05,
+      "loss": 0.6028,
+      "step": 6730
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9729236364364624,
+      "learning_rate": 1.4057295020874966e-05,
+      "loss": 0.5598,
+      "step": 6731
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0358269214630127,
+      "learning_rate": 1.4055597106555193e-05,
+      "loss": 0.9921,
+      "step": 6732
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9271126985549927,
+      "learning_rate": 1.4053899052288091e-05,
+      "loss": 1.0187,
+      "step": 6733
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7927757501602173,
+      "learning_rate": 1.405220085813226e-05,
+      "loss": 1.0521,
+      "step": 6734
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0081684589385986,
+      "learning_rate": 1.4050502524146294e-05,
+      "loss": 0.9678,
+      "step": 6735
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.77253258228302,
+      "learning_rate": 1.4048804050388802e-05,
+      "loss": 0.9605,
+      "step": 6736
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.882011890411377,
+      "learning_rate": 1.4047105436918392e-05,
+      "loss": 0.9201,
+      "step": 6737
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9428033828735352,
+      "learning_rate": 1.4045406683793677e-05,
+      "loss": 0.9867,
+      "step": 6738
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6767597198486328,
+      "learning_rate": 1.4043707791073278e-05,
+      "loss": 1.0221,
+      "step": 6739
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.70387601852417,
+      "learning_rate": 1.404200875881582e-05,
+      "loss": 1.0827,
+      "step": 6740
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.03206205368042,
+      "learning_rate": 1.4040309587079928e-05,
+      "loss": 1.0983,
+      "step": 6741
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.788863182067871,
+      "learning_rate": 1.403861027592424e-05,
+      "loss": 0.9548,
+      "step": 6742
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.174391269683838,
+      "learning_rate": 1.4036910825407395e-05,
+      "loss": 0.6276,
+      "step": 6743
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7766765356063843,
+      "learning_rate": 1.4035211235588032e-05,
+      "loss": 0.9986,
+      "step": 6744
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.102076292037964,
+      "learning_rate": 1.4033511506524802e-05,
+      "loss": 0.9219,
+      "step": 6745
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8500827550888062,
+      "learning_rate": 1.4031811638276356e-05,
+      "loss": 0.9636,
+      "step": 6746
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9082595109939575,
+      "learning_rate": 1.4030111630901354e-05,
+      "loss": 0.9861,
+      "step": 6747
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.641597867012024,
+      "learning_rate": 1.4028411484458456e-05,
+      "loss": 0.9329,
+      "step": 6748
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1459851264953613,
+      "learning_rate": 1.4026711199006331e-05,
+      "loss": 1.0106,
+      "step": 6749
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8246526718139648,
+      "learning_rate": 1.402501077460365e-05,
+      "loss": 0.9948,
+      "step": 6750
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7775335311889648,
+      "learning_rate": 1.4023310211309092e-05,
+      "loss": 1.0089,
+      "step": 6751
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8576596975326538,
+      "learning_rate": 1.4021609509181335e-05,
+      "loss": 1.0497,
+      "step": 6752
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9051514863967896,
+      "learning_rate": 1.401990866827907e-05,
+      "loss": 1.0876,
+      "step": 6753
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.922931432723999,
+      "learning_rate": 1.4018207688660985e-05,
+      "loss": 0.979,
+      "step": 6754
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7118853330612183,
+      "learning_rate": 1.4016506570385775e-05,
+      "loss": 0.9974,
+      "step": 6755
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8786288499832153,
+      "learning_rate": 1.4014805313512146e-05,
+      "loss": 1.0175,
+      "step": 6756
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7783482074737549,
+      "learning_rate": 1.4013103918098801e-05,
+      "loss": 0.9528,
+      "step": 6757
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9152061939239502,
+      "learning_rate": 1.401140238420445e-05,
+      "loss": 1.0323,
+      "step": 6758
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6809674501419067,
+      "learning_rate": 1.4009700711887806e-05,
+      "loss": 0.9861,
+      "step": 6759
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7512511014938354,
+      "learning_rate": 1.400799890120759e-05,
+      "loss": 1.0672,
+      "step": 6760
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1490086317062378,
+      "learning_rate": 1.400629695222253e-05,
+      "loss": 0.6663,
+      "step": 6761
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.817386507987976,
+      "learning_rate": 1.4004594864991354e-05,
+      "loss": 0.9611,
+      "step": 6762
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.295001268386841,
+      "learning_rate": 1.4002892639572795e-05,
+      "loss": 1.0126,
+      "step": 6763
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.11795711517334,
+      "learning_rate": 1.4001190276025593e-05,
+      "loss": 1.0335,
+      "step": 6764
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0941977500915527,
+      "learning_rate": 1.3999487774408491e-05,
+      "loss": 0.9809,
+      "step": 6765
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8022984266281128,
+      "learning_rate": 1.3997785134780239e-05,
+      "loss": 0.9806,
+      "step": 6766
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7007839679718018,
+      "learning_rate": 1.3996082357199585e-05,
+      "loss": 1.0436,
+      "step": 6767
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8461931943893433,
+      "learning_rate": 1.3994379441725297e-05,
+      "loss": 0.9859,
+      "step": 6768
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7574536800384521,
+      "learning_rate": 1.3992676388416128e-05,
+      "loss": 0.996,
+      "step": 6769
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9120289087295532,
+      "learning_rate": 1.3990973197330853e-05,
+      "loss": 0.9505,
+      "step": 6770
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7500176429748535,
+      "learning_rate": 1.3989269868528242e-05,
+      "loss": 0.9464,
+      "step": 6771
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.78048837184906,
+      "learning_rate": 1.398756640206707e-05,
+      "loss": 0.9325,
+      "step": 6772
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.755582571029663,
+      "learning_rate": 1.398586279800612e-05,
+      "loss": 0.9235,
+      "step": 6773
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1130605936050415,
+      "learning_rate": 1.3984159056404178e-05,
+      "loss": 0.583,
+      "step": 6774
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7887859344482422,
+      "learning_rate": 1.3982455177320038e-05,
+      "loss": 0.9839,
+      "step": 6775
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7818692922592163,
+      "learning_rate": 1.3980751160812491e-05,
+      "loss": 0.9966,
+      "step": 6776
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.048404335975647,
+      "learning_rate": 1.3979047006940345e-05,
+      "loss": 0.5909,
+      "step": 6777
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.016364812850952,
+      "learning_rate": 1.39773427157624e-05,
+      "loss": 1.0495,
+      "step": 6778
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6175923347473145,
+      "learning_rate": 1.397563828733747e-05,
+      "loss": 0.9709,
+      "step": 6779
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.005697727203369,
+      "learning_rate": 1.3973933721724364e-05,
+      "loss": 0.9522,
+      "step": 6780
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1397303342819214,
+      "learning_rate": 1.3972229018981911e-05,
+      "loss": 0.6316,
+      "step": 6781
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.811553955078125,
+      "learning_rate": 1.3970524179168927e-05,
+      "loss": 1.0376,
+      "step": 6782
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7347733974456787,
+      "learning_rate": 1.3968819202344246e-05,
+      "loss": 0.9579,
+      "step": 6783
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.750199556350708,
+      "learning_rate": 1.39671140885667e-05,
+      "loss": 0.9878,
+      "step": 6784
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7455170154571533,
+      "learning_rate": 1.3965408837895129e-05,
+      "loss": 1.022,
+      "step": 6785
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7671785354614258,
+      "learning_rate": 1.3963703450388377e-05,
+      "loss": 1.0123,
+      "step": 6786
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8536680936813354,
+      "learning_rate": 1.396199792610529e-05,
+      "loss": 1.0056,
+      "step": 6787
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8019664287567139,
+      "learning_rate": 1.3960292265104723e-05,
+      "loss": 0.9702,
+      "step": 6788
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0354859828948975,
+      "learning_rate": 1.3958586467445532e-05,
+      "loss": 1.0861,
+      "step": 6789
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8166725635528564,
+      "learning_rate": 1.3956880533186582e-05,
+      "loss": 1.005,
+      "step": 6790
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6663466691970825,
+      "learning_rate": 1.3955174462386733e-05,
+      "loss": 1.0318,
+      "step": 6791
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.814225435256958,
+      "learning_rate": 1.3953468255104865e-05,
+      "loss": 1.0157,
+      "step": 6792
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.781333565711975,
+      "learning_rate": 1.395176191139985e-05,
+      "loss": 0.9414,
+      "step": 6793
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7532334327697754,
+      "learning_rate": 1.3950055431330568e-05,
+      "loss": 1.0687,
+      "step": 6794
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8517855405807495,
+      "learning_rate": 1.3948348814955912e-05,
+      "loss": 0.9925,
+      "step": 6795
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7015671730041504,
+      "learning_rate": 1.3946642062334765e-05,
+      "loss": 1.0437,
+      "step": 6796
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6779364347457886,
+      "learning_rate": 1.3944935173526026e-05,
+      "loss": 1.0208,
+      "step": 6797
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.703795075416565,
+      "learning_rate": 1.3943228148588595e-05,
+      "loss": 0.9557,
+      "step": 6798
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.821058750152588,
+      "learning_rate": 1.3941520987581371e-05,
+      "loss": 0.9087,
+      "step": 6799
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.049903154373169,
+      "learning_rate": 1.3939813690563274e-05,
+      "loss": 1.0012,
+      "step": 6800
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5722832679748535,
+      "learning_rate": 1.3938106257593207e-05,
+      "loss": 0.9447,
+      "step": 6801
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8461346626281738,
+      "learning_rate": 1.3936398688730095e-05,
+      "loss": 0.9467,
+      "step": 6802
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6874440908432007,
+      "learning_rate": 1.393469098403286e-05,
+      "loss": 0.9443,
+      "step": 6803
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.957535982131958,
+      "learning_rate": 1.3932983143560433e-05,
+      "loss": 0.9429,
+      "step": 6804
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8040337562561035,
+      "learning_rate": 1.3931275167371743e-05,
+      "loss": 0.927,
+      "step": 6805
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7531129121780396,
+      "learning_rate": 1.3929567055525726e-05,
+      "loss": 1.073,
+      "step": 6806
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7696815729141235,
+      "learning_rate": 1.3927858808081327e-05,
+      "loss": 0.9081,
+      "step": 6807
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8680943250656128,
+      "learning_rate": 1.3926150425097493e-05,
+      "loss": 0.8907,
+      "step": 6808
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.846307396888733,
+      "learning_rate": 1.3924441906633174e-05,
+      "loss": 1.0195,
+      "step": 6809
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6509976387023926,
+      "learning_rate": 1.3922733252747332e-05,
+      "loss": 1.0058,
+      "step": 6810
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6506422758102417,
+      "learning_rate": 1.392102446349892e-05,
+      "loss": 0.8888,
+      "step": 6811
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5828776359558105,
+      "learning_rate": 1.3919315538946907e-05,
+      "loss": 0.9122,
+      "step": 6812
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8559921979904175,
+      "learning_rate": 1.3917606479150263e-05,
+      "loss": 1.0352,
+      "step": 6813
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7429535388946533,
+      "learning_rate": 1.3915897284167963e-05,
+      "loss": 1.0226,
+      "step": 6814
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1186811923980713,
+      "learning_rate": 1.3914187954058986e-05,
+      "loss": 0.6241,
+      "step": 6815
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.86961829662323,
+      "learning_rate": 1.3912478488882317e-05,
+      "loss": 1.0792,
+      "step": 6816
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.73786461353302,
+      "learning_rate": 1.3910768888696941e-05,
+      "loss": 0.9467,
+      "step": 6817
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9818168878555298,
+      "learning_rate": 1.3909059153561859e-05,
+      "loss": 0.9886,
+      "step": 6818
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6181373596191406,
+      "learning_rate": 1.3907349283536065e-05,
+      "loss": 0.9501,
+      "step": 6819
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.706716775894165,
+      "learning_rate": 1.390563927867856e-05,
+      "loss": 0.9533,
+      "step": 6820
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0678229331970215,
+      "learning_rate": 1.3903929139048353e-05,
+      "loss": 1.0081,
+      "step": 6821
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0080697536468506,
+      "learning_rate": 1.3902218864704458e-05,
+      "loss": 1.0186,
+      "step": 6822
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6773377656936646,
+      "learning_rate": 1.390050845570589e-05,
+      "loss": 1.0428,
+      "step": 6823
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8540889024734497,
+      "learning_rate": 1.3898797912111673e-05,
+      "loss": 0.9548,
+      "step": 6824
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7127068042755127,
+      "learning_rate": 1.3897087233980828e-05,
+      "loss": 1.0449,
+      "step": 6825
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7764031887054443,
+      "learning_rate": 1.3895376421372391e-05,
+      "loss": 0.9341,
+      "step": 6826
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7259466648101807,
+      "learning_rate": 1.3893665474345392e-05,
+      "loss": 0.9454,
+      "step": 6827
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.684853196144104,
+      "learning_rate": 1.3891954392958877e-05,
+      "loss": 0.9527,
+      "step": 6828
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7955219745635986,
+      "learning_rate": 1.3890243177271887e-05,
+      "loss": 1.0108,
+      "step": 6829
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.741733193397522,
+      "learning_rate": 1.388853182734347e-05,
+      "loss": 1.0072,
+      "step": 6830
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7035363912582397,
+      "learning_rate": 1.3886820343232685e-05,
+      "loss": 0.9151,
+      "step": 6831
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9676433801651,
+      "learning_rate": 1.3885108724998583e-05,
+      "loss": 1.0395,
+      "step": 6832
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7374156713485718,
+      "learning_rate": 1.3883396972700233e-05,
+      "loss": 1.0525,
+      "step": 6833
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7043983936309814,
+      "learning_rate": 1.3881685086396704e-05,
+      "loss": 1.0299,
+      "step": 6834
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.656599998474121,
+      "learning_rate": 1.3879973066147063e-05,
+      "loss": 0.9625,
+      "step": 6835
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7608959674835205,
+      "learning_rate": 1.3878260912010393e-05,
+      "loss": 0.9978,
+      "step": 6836
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8383724689483643,
+      "learning_rate": 1.3876548624045769e-05,
+      "loss": 1.0015,
+      "step": 6837
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7328393459320068,
+      "learning_rate": 1.3874836202312283e-05,
+      "loss": 1.0019,
+      "step": 6838
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0246083736419678,
+      "learning_rate": 1.3873123646869022e-05,
+      "loss": 1.0334,
+      "step": 6839
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.849651575088501,
+      "learning_rate": 1.3871410957775085e-05,
+      "loss": 1.0286,
+      "step": 6840
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8208962678909302,
+      "learning_rate": 1.3869698135089566e-05,
+      "loss": 0.9502,
+      "step": 6841
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6560593843460083,
+      "learning_rate": 1.3867985178871579e-05,
+      "loss": 1.0554,
+      "step": 6842
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6062977313995361,
+      "learning_rate": 1.3866272089180224e-05,
+      "loss": 0.9123,
+      "step": 6843
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8182982206344604,
+      "learning_rate": 1.3864558866074622e-05,
+      "loss": 1.0288,
+      "step": 6844
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6653316020965576,
+      "learning_rate": 1.386284550961389e-05,
+      "loss": 1.029,
+      "step": 6845
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.941753625869751,
+      "learning_rate": 1.3861132019857143e-05,
+      "loss": 1.0552,
+      "step": 6846
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.833056092262268,
+      "learning_rate": 1.3859418396863522e-05,
+      "loss": 0.9556,
+      "step": 6847
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.657699704170227,
+      "learning_rate": 1.385770464069215e-05,
+      "loss": 0.9047,
+      "step": 6848
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9108439683914185,
+      "learning_rate": 1.3855990751402169e-05,
+      "loss": 1.0864,
+      "step": 6849
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7507915496826172,
+      "learning_rate": 1.3854276729052716e-05,
+      "loss": 0.8967,
+      "step": 6850
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.08841073513031,
+      "learning_rate": 1.3852562573702941e-05,
+      "loss": 0.6313,
+      "step": 6851
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0323333740234375,
+      "learning_rate": 1.3850848285411994e-05,
+      "loss": 0.5743,
+      "step": 6852
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9940459728240967,
+      "learning_rate": 1.384913386423903e-05,
+      "loss": 0.5656,
+      "step": 6853
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0336462259292603,
+      "learning_rate": 1.3847419310243209e-05,
+      "loss": 0.5885,
+      "step": 6854
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7249271869659424,
+      "learning_rate": 1.3845704623483691e-05,
+      "loss": 0.9527,
+      "step": 6855
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7896692752838135,
+      "learning_rate": 1.3843989804019653e-05,
+      "loss": 1.0075,
+      "step": 6856
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8599122762680054,
+      "learning_rate": 1.3842274851910263e-05,
+      "loss": 0.9814,
+      "step": 6857
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9365019798278809,
+      "learning_rate": 1.38405597672147e-05,
+      "loss": 0.9887,
+      "step": 6858
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6805779933929443,
+      "learning_rate": 1.383884454999215e-05,
+      "loss": 1.0246,
+      "step": 6859
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.837536096572876,
+      "learning_rate": 1.3837129200301794e-05,
+      "loss": 1.0547,
+      "step": 6860
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8172812461853027,
+      "learning_rate": 1.3835413718202831e-05,
+      "loss": 0.9238,
+      "step": 6861
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.101880431175232,
+      "learning_rate": 1.3833698103754454e-05,
+      "loss": 0.5785,
+      "step": 6862
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6543262004852295,
+      "learning_rate": 1.3831982357015866e-05,
+      "loss": 1.0067,
+      "step": 6863
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8261693716049194,
+      "learning_rate": 1.3830266478046268e-05,
+      "loss": 1.0631,
+      "step": 6864
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.795305609703064,
+      "learning_rate": 1.3828550466904876e-05,
+      "loss": 0.883,
+      "step": 6865
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6449183225631714,
+      "learning_rate": 1.3826834323650899e-05,
+      "loss": 0.9758,
+      "step": 6866
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8044474124908447,
+      "learning_rate": 1.3825118048343562e-05,
+      "loss": 1.0641,
+      "step": 6867
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.758386492729187,
+      "learning_rate": 1.3823401641042085e-05,
+      "loss": 0.9946,
+      "step": 6868
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7067718505859375,
+      "learning_rate": 1.3821685101805697e-05,
+      "loss": 0.9277,
+      "step": 6869
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7688261270523071,
+      "learning_rate": 1.3819968430693633e-05,
+      "loss": 1.1033,
+      "step": 6870
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5806578397750854,
+      "learning_rate": 1.3818251627765129e-05,
+      "loss": 1.0465,
+      "step": 6871
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.844901204109192,
+      "learning_rate": 1.3816534693079426e-05,
+      "loss": 0.9981,
+      "step": 6872
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6712594032287598,
+      "learning_rate": 1.3814817626695771e-05,
+      "loss": 0.9352,
+      "step": 6873
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9066039323806763,
+      "learning_rate": 1.3813100428673419e-05,
+      "loss": 0.9937,
+      "step": 6874
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7405380010604858,
+      "learning_rate": 1.3811383099071618e-05,
+      "loss": 0.9602,
+      "step": 6875
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8275471925735474,
+      "learning_rate": 1.3809665637949636e-05,
+      "loss": 0.9518,
+      "step": 6876
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7132142782211304,
+      "learning_rate": 1.3807948045366737e-05,
+      "loss": 0.966,
+      "step": 6877
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9485523700714111,
+      "learning_rate": 1.3806230321382183e-05,
+      "loss": 1.0468,
+      "step": 6878
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7423691749572754,
+      "learning_rate": 1.3804512466055257e-05,
+      "loss": 0.934,
+      "step": 6879
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8203736543655396,
+      "learning_rate": 1.3802794479445232e-05,
+      "loss": 1.032,
+      "step": 6880
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8252575397491455,
+      "learning_rate": 1.3801076361611393e-05,
+      "loss": 1.0588,
+      "step": 6881
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.797673225402832,
+      "learning_rate": 1.3799358112613026e-05,
+      "loss": 1.0269,
+      "step": 6882
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7945324182510376,
+      "learning_rate": 1.379763973250942e-05,
+      "loss": 1.0038,
+      "step": 6883
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.857056975364685,
+      "learning_rate": 1.3795921221359877e-05,
+      "loss": 1.0242,
+      "step": 6884
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6810377836227417,
+      "learning_rate": 1.3794202579223699e-05,
+      "loss": 0.9972,
+      "step": 6885
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.825913429260254,
+      "learning_rate": 1.3792483806160188e-05,
+      "loss": 1.0659,
+      "step": 6886
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.655322551727295,
+      "learning_rate": 1.3790764902228653e-05,
+      "loss": 1.0265,
+      "step": 6887
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.776117205619812,
+      "learning_rate": 1.3789045867488411e-05,
+      "loss": 1.0244,
+      "step": 6888
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7154232263565063,
+      "learning_rate": 1.378732670199878e-05,
+      "loss": 0.9343,
+      "step": 6889
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.900773048400879,
+      "learning_rate": 1.3785607405819085e-05,
+      "loss": 0.8913,
+      "step": 6890
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7355782985687256,
+      "learning_rate": 1.3783887979008652e-05,
+      "loss": 1.0004,
+      "step": 6891
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7621214389801025,
+      "learning_rate": 1.3782168421626817e-05,
+      "loss": 1.0453,
+      "step": 6892
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6618645191192627,
+      "learning_rate": 1.3780448733732911e-05,
+      "loss": 0.9573,
+      "step": 6893
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9396284818649292,
+      "learning_rate": 1.3778728915386284e-05,
+      "loss": 0.9723,
+      "step": 6894
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8881758451461792,
+      "learning_rate": 1.3777008966646275e-05,
+      "loss": 0.995,
+      "step": 6895
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7376147508621216,
+      "learning_rate": 1.3775288887572238e-05,
+      "loss": 1.0451,
+      "step": 6896
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6251953840255737,
+      "learning_rate": 1.3773568678223525e-05,
+      "loss": 1.0418,
+      "step": 6897
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9007586240768433,
+      "learning_rate": 1.3771848338659502e-05,
+      "loss": 1.0188,
+      "step": 6898
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.880305528640747,
+      "learning_rate": 1.3770127868939528e-05,
+      "loss": 0.906,
+      "step": 6899
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8768893480300903,
+      "learning_rate": 1.3768407269122968e-05,
+      "loss": 0.9587,
+      "step": 6900
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7819699048995972,
+      "learning_rate": 1.3766686539269203e-05,
+      "loss": 0.9231,
+      "step": 6901
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6596421003341675,
+      "learning_rate": 1.3764965679437608e-05,
+      "loss": 1.0304,
+      "step": 6902
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8156112432479858,
+      "learning_rate": 1.3763244689687562e-05,
+      "loss": 0.9472,
+      "step": 6903
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.604722261428833,
+      "learning_rate": 1.3761523570078459e-05,
+      "loss": 0.989,
+      "step": 6904
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7487188577651978,
+      "learning_rate": 1.3759802320669681e-05,
+      "loss": 0.9988,
+      "step": 6905
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8543919324874878,
+      "learning_rate": 1.3758080941520628e-05,
+      "loss": 0.9908,
+      "step": 6906
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.846685767173767,
+      "learning_rate": 1.37563594326907e-05,
+      "loss": 1.0371,
+      "step": 6907
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8190960884094238,
+      "learning_rate": 1.3754637794239303e-05,
+      "loss": 1.0285,
+      "step": 6908
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7319422960281372,
+      "learning_rate": 1.375291602622584e-05,
+      "loss": 0.9807,
+      "step": 6909
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9516987800598145,
+      "learning_rate": 1.3751194128709731e-05,
+      "loss": 0.9238,
+      "step": 6910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.894558072090149,
+      "learning_rate": 1.374947210175039e-05,
+      "loss": 1.0307,
+      "step": 6911
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7581887245178223,
+      "learning_rate": 1.3747749945407238e-05,
+      "loss": 0.9743,
+      "step": 6912
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0241750478744507,
+      "learning_rate": 1.374602765973971e-05,
+      "loss": 0.6493,
+      "step": 6913
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7180957794189453,
+      "learning_rate": 1.3744305244807228e-05,
+      "loss": 0.966,
+      "step": 6914
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8838133811950684,
+      "learning_rate": 1.3742582700669229e-05,
+      "loss": 1.0685,
+      "step": 6915
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8051873445510864,
+      "learning_rate": 1.374086002738516e-05,
+      "loss": 0.9628,
+      "step": 6916
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8534681797027588,
+      "learning_rate": 1.3739137225014458e-05,
+      "loss": 0.9043,
+      "step": 6917
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6830724477767944,
+      "learning_rate": 1.3737414293616575e-05,
+      "loss": 0.9057,
+      "step": 6918
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7066988945007324,
+      "learning_rate": 1.3735691233250968e-05,
+      "loss": 0.9748,
+      "step": 6919
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8414219617843628,
+      "learning_rate": 1.373396804397709e-05,
+      "loss": 1.0403,
+      "step": 6920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9916975498199463,
+      "learning_rate": 1.3732244725854407e-05,
+      "loss": 0.5958,
+      "step": 6921
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7530094385147095,
+      "learning_rate": 1.3730521278942382e-05,
+      "loss": 0.9549,
+      "step": 6922
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7721494436264038,
+      "learning_rate": 1.3728797703300489e-05,
+      "loss": 1.0594,
+      "step": 6923
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7973453998565674,
+      "learning_rate": 1.3727073998988202e-05,
+      "loss": 0.9808,
+      "step": 6924
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.945422887802124,
+      "learning_rate": 1.3725350166065006e-05,
+      "loss": 0.9587,
+      "step": 6925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1564247608184814,
+      "learning_rate": 1.3723626204590376e-05,
+      "loss": 0.6194,
+      "step": 6926
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.943947434425354,
+      "learning_rate": 1.3721902114623812e-05,
+      "loss": 0.9891,
+      "step": 6927
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7373510599136353,
+      "learning_rate": 1.3720177896224802e-05,
+      "loss": 1.0493,
+      "step": 6928
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6268665790557861,
+      "learning_rate": 1.3718453549452843e-05,
+      "loss": 0.9699,
+      "step": 6929
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.950081706047058,
+      "learning_rate": 1.3716729074367443e-05,
+      "loss": 0.9893,
+      "step": 6930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9322916269302368,
+      "learning_rate": 1.3715004471028101e-05,
+      "loss": 0.9537,
+      "step": 6931
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6880568265914917,
+      "learning_rate": 1.3713279739494334e-05,
+      "loss": 0.9861,
+      "step": 6932
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6641559600830078,
+      "learning_rate": 1.3711554879825655e-05,
+      "loss": 0.9366,
+      "step": 6933
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.087035059928894,
+      "learning_rate": 1.3709829892081588e-05,
+      "loss": 0.5984,
+      "step": 6934
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8527957201004028,
+      "learning_rate": 1.3708104776321652e-05,
+      "loss": 0.9584,
+      "step": 6935
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5877742767333984,
+      "learning_rate": 1.3706379532605377e-05,
+      "loss": 0.999,
+      "step": 6936
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.744698405265808,
+      "learning_rate": 1.3704654160992298e-05,
+      "loss": 0.977,
+      "step": 6937
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8975149393081665,
+      "learning_rate": 1.3702928661541955e-05,
+      "loss": 1.0295,
+      "step": 6938
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8641890287399292,
+      "learning_rate": 1.3701203034313884e-05,
+      "loss": 1.0936,
+      "step": 6939
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7193855047225952,
+      "learning_rate": 1.3699477279367636e-05,
+      "loss": 0.9429,
+      "step": 6940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9784059524536133,
+      "learning_rate": 1.3697751396762762e-05,
+      "loss": 0.9404,
+      "step": 6941
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6951725482940674,
+      "learning_rate": 1.3696025386558817e-05,
+      "loss": 0.9774,
+      "step": 6942
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7396291494369507,
+      "learning_rate": 1.3694299248815362e-05,
+      "loss": 0.9084,
+      "step": 6943
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7987840175628662,
+      "learning_rate": 1.3692572983591957e-05,
+      "loss": 1.0254,
+      "step": 6944
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7492117881774902,
+      "learning_rate": 1.3690846590948172e-05,
+      "loss": 1.0077,
+      "step": 6945
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8775699138641357,
+      "learning_rate": 1.3689120070943584e-05,
+      "loss": 1.074,
+      "step": 6946
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.823535680770874,
+      "learning_rate": 1.3687393423637767e-05,
+      "loss": 0.9436,
+      "step": 6947
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7458266019821167,
+      "learning_rate": 1.3685666649090301e-05,
+      "loss": 0.9932,
+      "step": 6948
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8429923057556152,
+      "learning_rate": 1.3683939747360776e-05,
+      "loss": 0.938,
+      "step": 6949
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.896311640739441,
+      "learning_rate": 1.3682212718508781e-05,
+      "loss": 1.007,
+      "step": 6950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.826287031173706,
+      "learning_rate": 1.3680485562593911e-05,
+      "loss": 0.934,
+      "step": 6951
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.824033498764038,
+      "learning_rate": 1.3678758279675766e-05,
+      "loss": 0.9872,
+      "step": 6952
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7327228784561157,
+      "learning_rate": 1.3677030869813946e-05,
+      "loss": 0.969,
+      "step": 6953
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8105082511901855,
+      "learning_rate": 1.3675303333068062e-05,
+      "loss": 1.0242,
+      "step": 6954
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9966410398483276,
+      "learning_rate": 1.3673575669497729e-05,
+      "loss": 0.9841,
+      "step": 6955
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1057027578353882,
+      "learning_rate": 1.3671847879162562e-05,
+      "loss": 0.6237,
+      "step": 6956
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9029927253723145,
+      "learning_rate": 1.3670119962122182e-05,
+      "loss": 0.9535,
+      "step": 6957
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3604249954223633,
+      "learning_rate": 1.3668391918436212e-05,
+      "loss": 0.9971,
+      "step": 6958
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9180368185043335,
+      "learning_rate": 1.3666663748164286e-05,
+      "loss": 1.1134,
+      "step": 6959
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.919784426689148,
+      "learning_rate": 1.3664935451366035e-05,
+      "loss": 0.9921,
+      "step": 6960
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7360732555389404,
+      "learning_rate": 1.36632070281011e-05,
+      "loss": 1.0155,
+      "step": 6961
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.751959204673767,
+      "learning_rate": 1.3661478478429123e-05,
+      "loss": 1.0331,
+      "step": 6962
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7400237321853638,
+      "learning_rate": 1.3659749802409752e-05,
+      "loss": 0.9868,
+      "step": 6963
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7216694355010986,
+      "learning_rate": 1.3658021000102638e-05,
+      "loss": 0.9858,
+      "step": 6964
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7637081146240234,
+      "learning_rate": 1.3656292071567436e-05,
+      "loss": 1.0183,
+      "step": 6965
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9082213640213013,
+      "learning_rate": 1.365456301686381e-05,
+      "loss": 0.9342,
+      "step": 6966
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.762948989868164,
+      "learning_rate": 1.365283383605142e-05,
+      "loss": 0.9563,
+      "step": 6967
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9059149026870728,
+      "learning_rate": 1.365110452918994e-05,
+      "loss": 1.012,
+      "step": 6968
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7151412963867188,
+      "learning_rate": 1.3649375096339044e-05,
+      "loss": 0.9545,
+      "step": 6969
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6515100002288818,
+      "learning_rate": 1.3647645537558406e-05,
+      "loss": 0.955,
+      "step": 6970
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8940740823745728,
+      "learning_rate": 1.3645915852907709e-05,
+      "loss": 0.9831,
+      "step": 6971
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8581762313842773,
+      "learning_rate": 1.3644186042446641e-05,
+      "loss": 0.9713,
+      "step": 6972
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0670125484466553,
+      "learning_rate": 1.364245610623489e-05,
+      "loss": 0.5949,
+      "step": 6973
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8109681606292725,
+      "learning_rate": 1.3640726044332157e-05,
+      "loss": 1.1458,
+      "step": 6974
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7546714544296265,
+      "learning_rate": 1.3638995856798138e-05,
+      "loss": 0.9952,
+      "step": 6975
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6644359827041626,
+      "learning_rate": 1.3637265543692536e-05,
+      "loss": 0.9806,
+      "step": 6976
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7586390972137451,
+      "learning_rate": 1.363553510507506e-05,
+      "loss": 0.9561,
+      "step": 6977
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7128090858459473,
+      "learning_rate": 1.3633804541005423e-05,
+      "loss": 0.9075,
+      "step": 6978
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7180993556976318,
+      "learning_rate": 1.363207385154334e-05,
+      "loss": 1.0022,
+      "step": 6979
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7044589519500732,
+      "learning_rate": 1.3630343036748536e-05,
+      "loss": 1.0041,
+      "step": 6980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9900785684585571,
+      "learning_rate": 1.3628612096680738e-05,
+      "loss": 1.0296,
+      "step": 6981
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6514198780059814,
+      "learning_rate": 1.3626881031399669e-05,
+      "loss": 1.0069,
+      "step": 6982
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9177031517028809,
+      "learning_rate": 1.3625149840965066e-05,
+      "loss": 0.9796,
+      "step": 6983
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8580693006515503,
+      "learning_rate": 1.3623418525436668e-05,
+      "loss": 0.9831,
+      "step": 6984
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6237668991088867,
+      "learning_rate": 1.3621687084874222e-05,
+      "loss": 0.9995,
+      "step": 6985
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6746234893798828,
+      "learning_rate": 1.361995551933747e-05,
+      "loss": 0.978,
+      "step": 6986
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6842522621154785,
+      "learning_rate": 1.3618223828886165e-05,
+      "loss": 1.0357,
+      "step": 6987
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8021342754364014,
+      "learning_rate": 1.3616492013580063e-05,
+      "loss": 0.9742,
+      "step": 6988
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9917511940002441,
+      "learning_rate": 1.3614760073478923e-05,
+      "loss": 0.9304,
+      "step": 6989
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7161378860473633,
+      "learning_rate": 1.3613028008642512e-05,
+      "loss": 0.9298,
+      "step": 6990
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8116116523742676,
+      "learning_rate": 1.3611295819130597e-05,
+      "loss": 1.0215,
+      "step": 6991
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8013050556182861,
+      "learning_rate": 1.3609563505002949e-05,
+      "loss": 0.9854,
+      "step": 6992
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9005085229873657,
+      "learning_rate": 1.3607831066319346e-05,
+      "loss": 1.029,
+      "step": 6993
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7874274253845215,
+      "learning_rate": 1.3606098503139573e-05,
+      "loss": 0.9384,
+      "step": 6994
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.910658836364746,
+      "learning_rate": 1.3604365815523415e-05,
+      "loss": 1.0032,
+      "step": 6995
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.856735110282898,
+      "learning_rate": 1.3602633003530658e-05,
+      "loss": 0.9484,
+      "step": 6996
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8812077045440674,
+      "learning_rate": 1.3600900067221103e-05,
+      "loss": 1.0087,
+      "step": 6997
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8383809328079224,
+      "learning_rate": 1.3599167006654545e-05,
+      "loss": 0.9521,
+      "step": 6998
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7604742050170898,
+      "learning_rate": 1.3597433821890787e-05,
+      "loss": 0.9567,
+      "step": 6999
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7840937376022339,
+      "learning_rate": 1.3595700512989635e-05,
+      "loss": 1.0047,
+      "step": 7000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9289796352386475,
+      "learning_rate": 1.3593967080010905e-05,
+      "loss": 0.9753,
+      "step": 7001
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7640876770019531,
+      "learning_rate": 1.359223352301441e-05,
+      "loss": 1.0175,
+      "step": 7002
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.800082802772522,
+      "learning_rate": 1.359049984205997e-05,
+      "loss": 0.9792,
+      "step": 7003
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.99664306640625,
+      "learning_rate": 1.3588766037207411e-05,
+      "loss": 1.1687,
+      "step": 7004
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6999276876449585,
+      "learning_rate": 1.3587032108516555e-05,
+      "loss": 1.0019,
+      "step": 7005
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9114924669265747,
+      "learning_rate": 1.3585298056047247e-05,
+      "loss": 0.9636,
+      "step": 7006
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9935848712921143,
+      "learning_rate": 1.3583563879859318e-05,
+      "loss": 0.9371,
+      "step": 7007
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8532017469406128,
+      "learning_rate": 1.358182958001261e-05,
+      "loss": 1.0441,
+      "step": 7008
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7209014892578125,
+      "learning_rate": 1.3580095156566966e-05,
+      "loss": 1.0167,
+      "step": 7009
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6647123098373413,
+      "learning_rate": 1.3578360609582242e-05,
+      "loss": 0.949,
+      "step": 7010
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8390610218048096,
+      "learning_rate": 1.3576625939118286e-05,
+      "loss": 1.0074,
+      "step": 7011
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6940075159072876,
+      "learning_rate": 1.3574891145234962e-05,
+      "loss": 0.9493,
+      "step": 7012
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7103420495986938,
+      "learning_rate": 1.357315622799213e-05,
+      "loss": 0.9497,
+      "step": 7013
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.854063630104065,
+      "learning_rate": 1.3571421187449656e-05,
+      "loss": 1.0326,
+      "step": 7014
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7468311786651611,
+      "learning_rate": 1.3569686023667415e-05,
+      "loss": 1.0708,
+      "step": 7015
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.913588523864746,
+      "learning_rate": 1.356795073670528e-05,
+      "loss": 1.0091,
+      "step": 7016
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.942193865776062,
+      "learning_rate": 1.3566215326623131e-05,
+      "loss": 0.9449,
+      "step": 7017
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7962298393249512,
+      "learning_rate": 1.3564479793480856e-05,
+      "loss": 1.0253,
+      "step": 7018
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.058428168296814,
+      "learning_rate": 1.3562744137338336e-05,
+      "loss": 0.602,
+      "step": 7019
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.857191801071167,
+      "learning_rate": 1.356100835825547e-05,
+      "loss": 0.9692,
+      "step": 7020
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7414098978042603,
+      "learning_rate": 1.3559272456292153e-05,
+      "loss": 0.9871,
+      "step": 7021
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7420209646224976,
+      "learning_rate": 1.3557536431508287e-05,
+      "loss": 0.9514,
+      "step": 7022
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7479898929595947,
+      "learning_rate": 1.3555800283963775e-05,
+      "loss": 0.9887,
+      "step": 7023
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7601523399353027,
+      "learning_rate": 1.3554064013718528e-05,
+      "loss": 1.0652,
+      "step": 7024
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8693417310714722,
+      "learning_rate": 1.3552327620832461e-05,
+      "loss": 0.9381,
+      "step": 7025
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7442224025726318,
+      "learning_rate": 1.3550591105365492e-05,
+      "loss": 1.0031,
+      "step": 7026
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6652826070785522,
+      "learning_rate": 1.354885446737754e-05,
+      "loss": 0.9465,
+      "step": 7027
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6905291080474854,
+      "learning_rate": 1.3547117706928532e-05,
+      "loss": 0.9896,
+      "step": 7028
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7153137922286987,
+      "learning_rate": 1.3545380824078403e-05,
+      "loss": 1.0494,
+      "step": 7029
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7378743886947632,
+      "learning_rate": 1.3543643818887084e-05,
+      "loss": 0.9866,
+      "step": 7030
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6543859243392944,
+      "learning_rate": 1.3541906691414517e-05,
+      "loss": 0.9887,
+      "step": 7031
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9482614994049072,
+      "learning_rate": 1.3540169441720641e-05,
+      "loss": 1.0719,
+      "step": 7032
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8409395217895508,
+      "learning_rate": 1.3538432069865408e-05,
+      "loss": 0.925,
+      "step": 7033
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8252531290054321,
+      "learning_rate": 1.353669457590877e-05,
+      "loss": 1.0035,
+      "step": 7034
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8132827281951904,
+      "learning_rate": 1.3534956959910682e-05,
+      "loss": 0.9581,
+      "step": 7035
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6536005735397339,
+      "learning_rate": 1.3533219221931102e-05,
+      "loss": 0.9243,
+      "step": 7036
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0901060104370117,
+      "learning_rate": 1.3531481362029997e-05,
+      "loss": 0.6558,
+      "step": 7037
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8233550786972046,
+      "learning_rate": 1.3529743380267335e-05,
+      "loss": 0.9779,
+      "step": 7038
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0900726318359375,
+      "learning_rate": 1.3528005276703089e-05,
+      "loss": 0.5776,
+      "step": 7039
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9141634702682495,
+      "learning_rate": 1.3526267051397235e-05,
+      "loss": 1.0117,
+      "step": 7040
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.820834994316101,
+      "learning_rate": 1.3524528704409759e-05,
+      "loss": 1.0554,
+      "step": 7041
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6915125846862793,
+      "learning_rate": 1.3522790235800638e-05,
+      "loss": 0.966,
+      "step": 7042
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0108846426010132,
+      "learning_rate": 1.3521051645629867e-05,
+      "loss": 0.6209,
+      "step": 7043
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7411608695983887,
+      "learning_rate": 1.351931293395744e-05,
+      "loss": 0.966,
+      "step": 7044
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.961833119392395,
+      "learning_rate": 1.3517574100843356e-05,
+      "loss": 0.4893,
+      "step": 7045
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1222561597824097,
+      "learning_rate": 1.3515835146347616e-05,
+      "loss": 0.6678,
+      "step": 7046
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9756367206573486,
+      "learning_rate": 1.3514096070530225e-05,
+      "loss": 0.9883,
+      "step": 7047
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9438972473144531,
+      "learning_rate": 1.3512356873451191e-05,
+      "loss": 1.0327,
+      "step": 7048
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.832377552986145,
+      "learning_rate": 1.3510617555170538e-05,
+      "loss": 1.0228,
+      "step": 7049
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6580421924591064,
+      "learning_rate": 1.3508878115748279e-05,
+      "loss": 0.9652,
+      "step": 7050
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6248708963394165,
+      "learning_rate": 1.3507138555244436e-05,
+      "loss": 0.9789,
+      "step": 7051
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8159586191177368,
+      "learning_rate": 1.350539887371904e-05,
+      "loss": 0.965,
+      "step": 7052
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7084332704544067,
+      "learning_rate": 1.350365907123212e-05,
+      "loss": 1.0746,
+      "step": 7053
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7992106676101685,
+      "learning_rate": 1.3501919147843715e-05,
+      "loss": 0.9973,
+      "step": 7054
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.803873896598816,
+      "learning_rate": 1.350017910361386e-05,
+      "loss": 0.926,
+      "step": 7055
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.045374631881714,
+      "learning_rate": 1.3498438938602601e-05,
+      "loss": 1.0142,
+      "step": 7056
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6891943216323853,
+      "learning_rate": 1.3496698652869985e-05,
+      "loss": 0.9741,
+      "step": 7057
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6306461095809937,
+      "learning_rate": 1.3494958246476071e-05,
+      "loss": 1.007,
+      "step": 7058
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6677608489990234,
+      "learning_rate": 1.3493217719480907e-05,
+      "loss": 1.0304,
+      "step": 7059
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9547065496444702,
+      "learning_rate": 1.349147707194456e-05,
+      "loss": 0.9932,
+      "step": 7060
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8835636377334595,
+      "learning_rate": 1.3489736303927088e-05,
+      "loss": 1.0529,
+      "step": 7061
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9034584760665894,
+      "learning_rate": 1.3487995415488568e-05,
+      "loss": 0.9915,
+      "step": 7062
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7705930471420288,
+      "learning_rate": 1.348625440668907e-05,
+      "loss": 0.9218,
+      "step": 7063
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8528423309326172,
+      "learning_rate": 1.3484513277588668e-05,
+      "loss": 0.9631,
+      "step": 7064
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9423719644546509,
+      "learning_rate": 1.3482772028247448e-05,
+      "loss": 0.9791,
+      "step": 7065
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.749715805053711,
+      "learning_rate": 1.3481030658725496e-05,
+      "loss": 0.9833,
+      "step": 7066
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7337987422943115,
+      "learning_rate": 1.3479289169082899e-05,
+      "loss": 0.9834,
+      "step": 7067
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6208419799804688,
+      "learning_rate": 1.3477547559379748e-05,
+      "loss": 0.926,
+      "step": 7068
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.850780725479126,
+      "learning_rate": 1.3475805829676149e-05,
+      "loss": 0.9694,
+      "step": 7069
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.962349534034729,
+      "learning_rate": 1.34740639800322e-05,
+      "loss": 1.0264,
+      "step": 7070
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.781683087348938,
+      "learning_rate": 1.3472322010508003e-05,
+      "loss": 0.9417,
+      "step": 7071
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6838470697402954,
+      "learning_rate": 1.3470579921163675e-05,
+      "loss": 1.0583,
+      "step": 7072
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8170595169067383,
+      "learning_rate": 1.3468837712059331e-05,
+      "loss": 1.0239,
+      "step": 7073
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7725492715835571,
+      "learning_rate": 1.3467095383255087e-05,
+      "loss": 1.0163,
+      "step": 7074
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7359145879745483,
+      "learning_rate": 1.3465352934811065e-05,
+      "loss": 0.9067,
+      "step": 7075
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.535945177078247,
+      "learning_rate": 1.3463610366787392e-05,
+      "loss": 1.078,
+      "step": 7076
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7793300151824951,
+      "learning_rate": 1.3461867679244203e-05,
+      "loss": 0.9522,
+      "step": 7077
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8639154434204102,
+      "learning_rate": 1.346012487224163e-05,
+      "loss": 1.0918,
+      "step": 7078
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.718047857284546,
+      "learning_rate": 1.3458381945839814e-05,
+      "loss": 1.0124,
+      "step": 7079
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.898523211479187,
+      "learning_rate": 1.3456638900098895e-05,
+      "loss": 0.9628,
+      "step": 7080
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6874685287475586,
+      "learning_rate": 1.3454895735079024e-05,
+      "loss": 0.9438,
+      "step": 7081
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7769930362701416,
+      "learning_rate": 1.3453152450840353e-05,
+      "loss": 1.0077,
+      "step": 7082
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7403358221054077,
+      "learning_rate": 1.3451409047443036e-05,
+      "loss": 0.9738,
+      "step": 7083
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5668665170669556,
+      "learning_rate": 1.3449665524947234e-05,
+      "loss": 0.9835,
+      "step": 7084
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.709720492362976,
+      "learning_rate": 1.3447921883413114e-05,
+      "loss": 1.0087,
+      "step": 7085
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7522914409637451,
+      "learning_rate": 1.3446178122900837e-05,
+      "loss": 1.0208,
+      "step": 7086
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8360804319381714,
+      "learning_rate": 1.344443424347058e-05,
+      "loss": 1.0475,
+      "step": 7087
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9595894813537598,
+      "learning_rate": 1.3442690245182521e-05,
+      "loss": 1.0712,
+      "step": 7088
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.747092604637146,
+      "learning_rate": 1.3440946128096836e-05,
+      "loss": 0.9465,
+      "step": 7089
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1992155313491821,
+      "learning_rate": 1.3439201892273715e-05,
+      "loss": 0.6542,
+      "step": 7090
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.942929744720459,
+      "learning_rate": 1.3437457537773341e-05,
+      "loss": 1.0208,
+      "step": 7091
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0027027130126953,
+      "learning_rate": 1.3435713064655913e-05,
+      "loss": 1.0236,
+      "step": 7092
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7496685981750488,
+      "learning_rate": 1.3433968472981622e-05,
+      "loss": 0.9923,
+      "step": 7093
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9361763000488281,
+      "learning_rate": 1.3432223762810672e-05,
+      "loss": 1.0381,
+      "step": 7094
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.713435173034668,
+      "learning_rate": 1.3430478934203265e-05,
+      "loss": 1.0205,
+      "step": 7095
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0750579833984375,
+      "learning_rate": 1.3428733987219618e-05,
+      "loss": 1.022,
+      "step": 7096
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9263370037078857,
+      "learning_rate": 1.3426988921919934e-05,
+      "loss": 0.9751,
+      "step": 7097
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.691001296043396,
+      "learning_rate": 1.3425243738364435e-05,
+      "loss": 0.9429,
+      "step": 7098
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7930668592453003,
+      "learning_rate": 1.3423498436613347e-05,
+      "loss": 0.938,
+      "step": 7099
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.732703447341919,
+      "learning_rate": 1.3421753016726889e-05,
+      "loss": 0.9909,
+      "step": 7100
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6890366077423096,
+      "learning_rate": 1.3420007478765291e-05,
+      "loss": 0.9478,
+      "step": 7101
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.833478569984436,
+      "learning_rate": 1.3418261822788789e-05,
+      "loss": 0.9757,
+      "step": 7102
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6822532415390015,
+      "learning_rate": 1.3416516048857623e-05,
+      "loss": 0.8885,
+      "step": 7103
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7534940242767334,
+      "learning_rate": 1.3414770157032026e-05,
+      "loss": 0.9465,
+      "step": 7104
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.797637701034546,
+      "learning_rate": 1.3413024147372256e-05,
+      "loss": 0.9854,
+      "step": 7105
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.754135251045227,
+      "learning_rate": 1.3411278019938552e-05,
+      "loss": 1.0972,
+      "step": 7106
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7949976921081543,
+      "learning_rate": 1.3409531774791175e-05,
+      "loss": 0.9662,
+      "step": 7107
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7370597124099731,
+      "learning_rate": 1.340778541199038e-05,
+      "loss": 0.9612,
+      "step": 7108
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7453358173370361,
+      "learning_rate": 1.340603893159643e-05,
+      "loss": 0.9048,
+      "step": 7109
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.835410475730896,
+      "learning_rate": 1.3404292333669588e-05,
+      "loss": 0.8921,
+      "step": 7110
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8565512895584106,
+      "learning_rate": 1.3402545618270128e-05,
+      "loss": 1.0153,
+      "step": 7111
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8340508937835693,
+      "learning_rate": 1.3400798785458326e-05,
+      "loss": 1.0048,
+      "step": 7112
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.602967619895935,
+      "learning_rate": 1.339905183529446e-05,
+      "loss": 0.9358,
+      "step": 7113
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.698876142501831,
+      "learning_rate": 1.3397304767838801e-05,
+      "loss": 1.0939,
+      "step": 7114
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7387785911560059,
+      "learning_rate": 1.339555758315165e-05,
+      "loss": 0.999,
+      "step": 7115
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1303051710128784,
+      "learning_rate": 1.3393810281293294e-05,
+      "loss": 0.6442,
+      "step": 7116
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8230528831481934,
+      "learning_rate": 1.3392062862324023e-05,
+      "loss": 0.9495,
+      "step": 7117
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.907477855682373,
+      "learning_rate": 1.3390315326304138e-05,
+      "loss": 1.0355,
+      "step": 7118
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5617702007293701,
+      "learning_rate": 1.3388567673293942e-05,
+      "loss": 0.9178,
+      "step": 7119
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7005559206008911,
+      "learning_rate": 1.338681990335374e-05,
+      "loss": 1.031,
+      "step": 7120
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0625462532043457,
+      "learning_rate": 1.3385072016543846e-05,
+      "loss": 0.9813,
+      "step": 7121
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.018613338470459,
+      "learning_rate": 1.3383324012924571e-05,
+      "loss": 1.0617,
+      "step": 7122
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6336359977722168,
+      "learning_rate": 1.3381575892556236e-05,
+      "loss": 1.0361,
+      "step": 7123
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5899081230163574,
+      "learning_rate": 1.3379827655499163e-05,
+      "loss": 1.0524,
+      "step": 7124
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9728938341140747,
+      "learning_rate": 1.3378079301813676e-05,
+      "loss": 1.0791,
+      "step": 7125
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9653327465057373,
+      "learning_rate": 1.3376330831560111e-05,
+      "loss": 1.0768,
+      "step": 7126
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6958613395690918,
+      "learning_rate": 1.33745822447988e-05,
+      "loss": 0.9449,
+      "step": 7127
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6509110927581787,
+      "learning_rate": 1.3372833541590082e-05,
+      "loss": 1.0184,
+      "step": 7128
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7942200899124146,
+      "learning_rate": 1.33710847219943e-05,
+      "loss": 1.0238,
+      "step": 7129
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6847031116485596,
+      "learning_rate": 1.3369335786071805e-05,
+      "loss": 0.9873,
+      "step": 7130
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8261550664901733,
+      "learning_rate": 1.3367586733882941e-05,
+      "loss": 1.008,
+      "step": 7131
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.85873544216156,
+      "learning_rate": 1.3365837565488065e-05,
+      "loss": 1.0458,
+      "step": 7132
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6591531038284302,
+      "learning_rate": 1.3364088280947535e-05,
+      "loss": 0.9591,
+      "step": 7133
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7034783363342285,
+      "learning_rate": 1.336233888032172e-05,
+      "loss": 0.9304,
+      "step": 7134
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0323245525360107,
+      "learning_rate": 1.3360589363670979e-05,
+      "loss": 1.0623,
+      "step": 7135
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8291207551956177,
+      "learning_rate": 1.3358839731055688e-05,
+      "loss": 0.9202,
+      "step": 7136
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7845864295959473,
+      "learning_rate": 1.3357089982536217e-05,
+      "loss": 0.9608,
+      "step": 7137
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5888057947158813,
+      "learning_rate": 1.3355340118172953e-05,
+      "loss": 0.964,
+      "step": 7138
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9319008588790894,
+      "learning_rate": 1.3353590138026273e-05,
+      "loss": 1.0629,
+      "step": 7139
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.931626558303833,
+      "learning_rate": 1.3351840042156565e-05,
+      "loss": 0.9517,
+      "step": 7140
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8275007009506226,
+      "learning_rate": 1.335008983062422e-05,
+      "loss": 0.9566,
+      "step": 7141
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.721826195716858,
+      "learning_rate": 1.3348339503489634e-05,
+      "loss": 0.9645,
+      "step": 7142
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6183193922042847,
+      "learning_rate": 1.3346589060813205e-05,
+      "loss": 0.9714,
+      "step": 7143
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7025712728500366,
+      "learning_rate": 1.3344838502655333e-05,
+      "loss": 1.0471,
+      "step": 7144
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.911240577697754,
+      "learning_rate": 1.334308782907643e-05,
+      "loss": 1.0072,
+      "step": 7145
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7872273921966553,
+      "learning_rate": 1.3341337040136905e-05,
+      "loss": 0.9913,
+      "step": 7146
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.749800443649292,
+      "learning_rate": 1.3339586135897168e-05,
+      "loss": 1.0408,
+      "step": 7147
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.014974594116211,
+      "learning_rate": 1.3337835116417649e-05,
+      "loss": 0.9006,
+      "step": 7148
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8732120990753174,
+      "learning_rate": 1.3336083981758758e-05,
+      "loss": 0.9283,
+      "step": 7149
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8451341390609741,
+      "learning_rate": 1.3334332731980933e-05,
+      "loss": 1.004,
+      "step": 7150
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6996300220489502,
+      "learning_rate": 1.3332581367144598e-05,
+      "loss": 0.991,
+      "step": 7151
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7470327615737915,
+      "learning_rate": 1.3330829887310186e-05,
+      "loss": 1.0038,
+      "step": 7152
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8499749898910522,
+      "learning_rate": 1.332907829253814e-05,
+      "loss": 0.9096,
+      "step": 7153
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6973888874053955,
+      "learning_rate": 1.3327326582888902e-05,
+      "loss": 1.0011,
+      "step": 7154
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8498083353042603,
+      "learning_rate": 1.3325574758422919e-05,
+      "loss": 1.0434,
+      "step": 7155
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9180599451065063,
+      "learning_rate": 1.3323822819200642e-05,
+      "loss": 0.9956,
+      "step": 7156
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7870728969573975,
+      "learning_rate": 1.3322070765282522e-05,
+      "loss": 0.9784,
+      "step": 7157
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5605841875076294,
+      "learning_rate": 1.332031859672902e-05,
+      "loss": 0.9699,
+      "step": 7158
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1336910724639893,
+      "learning_rate": 1.33185663136006e-05,
+      "loss": 0.6108,
+      "step": 7159
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9131513833999634,
+      "learning_rate": 1.3316813915957724e-05,
+      "loss": 0.9672,
+      "step": 7160
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8049116134643555,
+      "learning_rate": 1.3315061403860868e-05,
+      "loss": 0.9577,
+      "step": 7161
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9845693111419678,
+      "learning_rate": 1.3313308777370502e-05,
+      "loss": 0.9679,
+      "step": 7162
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.634359359741211,
+      "learning_rate": 1.3311556036547104e-05,
+      "loss": 0.9336,
+      "step": 7163
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7526347637176514,
+      "learning_rate": 1.3309803181451155e-05,
+      "loss": 1.0303,
+      "step": 7164
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.046959161758423,
+      "learning_rate": 1.330805021214315e-05,
+      "loss": 1.0261,
+      "step": 7165
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.863598108291626,
+      "learning_rate": 1.330629712868357e-05,
+      "loss": 1.083,
+      "step": 7166
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0913314819335938,
+      "learning_rate": 1.330454393113291e-05,
+      "loss": 0.9598,
+      "step": 7167
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0542597770690918,
+      "learning_rate": 1.3302790619551673e-05,
+      "loss": 0.5728,
+      "step": 7168
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8831356763839722,
+      "learning_rate": 1.3301037194000355e-05,
+      "loss": 1.0135,
+      "step": 7169
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7314691543579102,
+      "learning_rate": 1.3299283654539467e-05,
+      "loss": 1.0009,
+      "step": 7170
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7369725704193115,
+      "learning_rate": 1.3297530001229515e-05,
+      "loss": 0.9467,
+      "step": 7171
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.62538480758667,
+      "learning_rate": 1.3295776234131015e-05,
+      "loss": 0.9317,
+      "step": 7172
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.219205141067505,
+      "learning_rate": 1.3294022353304481e-05,
+      "loss": 1.016,
+      "step": 7173
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.77105712890625,
+      "learning_rate": 1.329226835881044e-05,
+      "loss": 1.0153,
+      "step": 7174
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.790759563446045,
+      "learning_rate": 1.3290514250709414e-05,
+      "loss": 0.9287,
+      "step": 7175
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7438914775848389,
+      "learning_rate": 1.3288760029061929e-05,
+      "loss": 0.9405,
+      "step": 7176
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7851142883300781,
+      "learning_rate": 1.3287005693928525e-05,
+      "loss": 1.0582,
+      "step": 7177
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8819966316223145,
+      "learning_rate": 1.3285251245369736e-05,
+      "loss": 0.9538,
+      "step": 7178
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8589634895324707,
+      "learning_rate": 1.3283496683446106e-05,
+      "loss": 0.984,
+      "step": 7179
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6994068622589111,
+      "learning_rate": 1.3281742008218173e-05,
+      "loss": 0.9692,
+      "step": 7180
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7802623510360718,
+      "learning_rate": 1.3279987219746495e-05,
+      "loss": 1.0179,
+      "step": 7181
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.731086254119873,
+      "learning_rate": 1.3278232318091618e-05,
+      "loss": 1.0565,
+      "step": 7182
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.723964810371399,
+      "learning_rate": 1.3276477303314102e-05,
+      "loss": 0.9658,
+      "step": 7183
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7675085067749023,
+      "learning_rate": 1.3274722175474505e-05,
+      "loss": 1.0576,
+      "step": 7184
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9074996709823608,
+      "learning_rate": 1.3272966934633396e-05,
+      "loss": 1.117,
+      "step": 7185
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7390044927597046,
+      "learning_rate": 1.327121158085134e-05,
+      "loss": 0.9902,
+      "step": 7186
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.738469123840332,
+      "learning_rate": 1.3269456114188908e-05,
+      "loss": 0.9622,
+      "step": 7187
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7782527208328247,
+      "learning_rate": 1.326770053470668e-05,
+      "loss": 1.0451,
+      "step": 7188
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7204747200012207,
+      "learning_rate": 1.3265944842465236e-05,
+      "loss": 1.0044,
+      "step": 7189
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.970015287399292,
+      "learning_rate": 1.3264189037525154e-05,
+      "loss": 1.0453,
+      "step": 7190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6674587726593018,
+      "learning_rate": 1.3262433119947028e-05,
+      "loss": 1.0404,
+      "step": 7191
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6568924188613892,
+      "learning_rate": 1.3260677089791449e-05,
+      "loss": 0.9335,
+      "step": 7192
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9165407419204712,
+      "learning_rate": 1.3258920947119013e-05,
+      "loss": 0.944,
+      "step": 7193
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6607844829559326,
+      "learning_rate": 1.3257164691990321e-05,
+      "loss": 1.0182,
+      "step": 7194
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7086931467056274,
+      "learning_rate": 1.3255408324465971e-05,
+      "loss": 1.005,
+      "step": 7195
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8508728742599487,
+      "learning_rate": 1.3253651844606571e-05,
+      "loss": 1.021,
+      "step": 7196
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7938482761383057,
+      "learning_rate": 1.3251895252472738e-05,
+      "loss": 0.9506,
+      "step": 7197
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7508562803268433,
+      "learning_rate": 1.3250138548125082e-05,
+      "loss": 1.0347,
+      "step": 7198
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7429879903793335,
+      "learning_rate": 1.3248381731624225e-05,
+      "loss": 0.9703,
+      "step": 7199
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.802351713180542,
+      "learning_rate": 1.3246624803030787e-05,
+      "loss": 1.0595,
+      "step": 7200
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.117609739303589,
+      "learning_rate": 1.3244867762405398e-05,
+      "loss": 1.0867,
+      "step": 7201
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.744667649269104,
+      "learning_rate": 1.3243110609808685e-05,
+      "loss": 0.8805,
+      "step": 7202
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7889891862869263,
+      "learning_rate": 1.3241353345301282e-05,
+      "loss": 1.0372,
+      "step": 7203
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8246865272521973,
+      "learning_rate": 1.3239595968943832e-05,
+      "loss": 1.0495,
+      "step": 7204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7911354303359985,
+      "learning_rate": 1.3237838480796976e-05,
+      "loss": 0.9979,
+      "step": 7205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9357727766036987,
+      "learning_rate": 1.3236080880921355e-05,
+      "loss": 0.9865,
+      "step": 7206
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9034751653671265,
+      "learning_rate": 1.3234323169377627e-05,
+      "loss": 0.9316,
+      "step": 7207
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8116472959518433,
+      "learning_rate": 1.3232565346226439e-05,
+      "loss": 0.9508,
+      "step": 7208
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5838007926940918,
+      "learning_rate": 1.323080741152845e-05,
+      "loss": 0.9717,
+      "step": 7209
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7649577856063843,
+      "learning_rate": 1.3229049365344322e-05,
+      "loss": 0.9972,
+      "step": 7210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8748691082000732,
+      "learning_rate": 1.322729120773472e-05,
+      "loss": 1.024,
+      "step": 7211
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8980518579483032,
+      "learning_rate": 1.3225532938760317e-05,
+      "loss": 0.9375,
+      "step": 7212
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8921834230422974,
+      "learning_rate": 1.3223774558481776e-05,
+      "loss": 0.9723,
+      "step": 7213
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.560067057609558,
+      "learning_rate": 1.3222016066959786e-05,
+      "loss": 0.8823,
+      "step": 7214
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.986478328704834,
+      "learning_rate": 1.322025746425502e-05,
+      "loss": 0.9961,
+      "step": 7215
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8217487335205078,
+      "learning_rate": 1.3218498750428164e-05,
+      "loss": 1.0161,
+      "step": 7216
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6017341613769531,
+      "learning_rate": 1.3216739925539908e-05,
+      "loss": 0.9567,
+      "step": 7217
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0322140455245972,
+      "learning_rate": 1.3214980989650939e-05,
+      "loss": 0.5558,
+      "step": 7218
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8201757669448853,
+      "learning_rate": 1.3213221942821958e-05,
+      "loss": 0.9655,
+      "step": 7219
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.087665557861328,
+      "learning_rate": 1.3211462785113666e-05,
+      "loss": 0.9508,
+      "step": 7220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.868225336074829,
+      "learning_rate": 1.3209703516586763e-05,
+      "loss": 1.0054,
+      "step": 7221
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1085224151611328,
+      "learning_rate": 1.3207944137301958e-05,
+      "loss": 0.6108,
+      "step": 7222
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.904455542564392,
+      "learning_rate": 1.3206184647319961e-05,
+      "loss": 1.0375,
+      "step": 7223
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.720018744468689,
+      "learning_rate": 1.3204425046701487e-05,
+      "loss": 1.0041,
+      "step": 7224
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8209823369979858,
+      "learning_rate": 1.3202665335507261e-05,
+      "loss": 1.0151,
+      "step": 7225
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7863492965698242,
+      "learning_rate": 1.3200905513797997e-05,
+      "loss": 1.0613,
+      "step": 7226
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.901233434677124,
+      "learning_rate": 1.3199145581634425e-05,
+      "loss": 1.0055,
+      "step": 7227
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7353391647338867,
+      "learning_rate": 1.3197385539077274e-05,
+      "loss": 1.0249,
+      "step": 7228
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7799358367919922,
+      "learning_rate": 1.319562538618728e-05,
+      "loss": 0.9793,
+      "step": 7229
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7576563358306885,
+      "learning_rate": 1.319386512302518e-05,
+      "loss": 1.0232,
+      "step": 7230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.70844566822052,
+      "learning_rate": 1.3192104749651717e-05,
+      "loss": 0.9748,
+      "step": 7231
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7767360210418701,
+      "learning_rate": 1.3190344266127639e-05,
+      "loss": 0.9327,
+      "step": 7232
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8299583196640015,
+      "learning_rate": 1.318858367251369e-05,
+      "loss": 1.0248,
+      "step": 7233
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6567928791046143,
+      "learning_rate": 1.3186822968870624e-05,
+      "loss": 0.9787,
+      "step": 7234
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0927796363830566,
+      "learning_rate": 1.31850621552592e-05,
+      "loss": 0.5708,
+      "step": 7235
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6882855892181396,
+      "learning_rate": 1.3183301231740182e-05,
+      "loss": 1.0278,
+      "step": 7236
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6057868003845215,
+      "learning_rate": 1.3181540198374325e-05,
+      "loss": 0.9118,
+      "step": 7237
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7376574277877808,
+      "learning_rate": 1.3179779055222407e-05,
+      "loss": 1.0209,
+      "step": 7238
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5406112670898438,
+      "learning_rate": 1.3178017802345196e-05,
+      "loss": 0.9771,
+      "step": 7239
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.91118586063385,
+      "learning_rate": 1.3176256439803465e-05,
+      "loss": 1.0388,
+      "step": 7240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.890557050704956,
+      "learning_rate": 1.3174494967658e-05,
+      "loss": 0.8887,
+      "step": 7241
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.65822172164917,
+      "learning_rate": 1.3172733385969579e-05,
+      "loss": 1.0199,
+      "step": 7242
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7154620885849,
+      "learning_rate": 1.317097169479899e-05,
+      "loss": 1.0009,
+      "step": 7243
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7485640048980713,
+      "learning_rate": 1.316920989420703e-05,
+      "loss": 1.0079,
+      "step": 7244
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6414271593093872,
+      "learning_rate": 1.3167447984254486e-05,
+      "loss": 1.0301,
+      "step": 7245
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7783876657485962,
+      "learning_rate": 1.3165685965002159e-05,
+      "loss": 1.0128,
+      "step": 7246
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0908946990966797,
+      "learning_rate": 1.3163923836510854e-05,
+      "loss": 1.0684,
+      "step": 7247
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7661899328231812,
+      "learning_rate": 1.3162161598841378e-05,
+      "loss": 1.0287,
+      "step": 7248
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7683403491973877,
+      "learning_rate": 1.3160399252054536e-05,
+      "loss": 0.8959,
+      "step": 7249
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.848463535308838,
+      "learning_rate": 1.3158636796211143e-05,
+      "loss": 0.95,
+      "step": 7250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.726299524307251,
+      "learning_rate": 1.3156874231372022e-05,
+      "loss": 0.9156,
+      "step": 7251
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8885399103164673,
+      "learning_rate": 1.3155111557597987e-05,
+      "loss": 0.9677,
+      "step": 7252
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6882926225662231,
+      "learning_rate": 1.3153348774949864e-05,
+      "loss": 1.0393,
+      "step": 7253
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8314157724380493,
+      "learning_rate": 1.3151585883488485e-05,
+      "loss": 1.042,
+      "step": 7254
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6638668775558472,
+      "learning_rate": 1.314982288327468e-05,
+      "loss": 0.9388,
+      "step": 7255
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8940447568893433,
+      "learning_rate": 1.3148059774369286e-05,
+      "loss": 1.0152,
+      "step": 7256
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8125624656677246,
+      "learning_rate": 1.314629655683314e-05,
+      "loss": 0.9426,
+      "step": 7257
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8944802284240723,
+      "learning_rate": 1.3144533230727092e-05,
+      "loss": 1.0074,
+      "step": 7258
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7745938301086426,
+      "learning_rate": 1.3142769796111987e-05,
+      "loss": 0.9879,
+      "step": 7259
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7447702884674072,
+      "learning_rate": 1.3141006253048674e-05,
+      "loss": 1.0788,
+      "step": 7260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.86833918094635,
+      "learning_rate": 1.313924260159801e-05,
+      "loss": 1.045,
+      "step": 7261
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8839757442474365,
+      "learning_rate": 1.3137478841820853e-05,
+      "loss": 0.9568,
+      "step": 7262
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.644152283668518,
+      "learning_rate": 1.3135714973778064e-05,
+      "loss": 0.9814,
+      "step": 7263
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5099202394485474,
+      "learning_rate": 1.3133950997530512e-05,
+      "loss": 0.9721,
+      "step": 7264
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7904218435287476,
+      "learning_rate": 1.3132186913139064e-05,
+      "loss": 0.9817,
+      "step": 7265
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7721269130706787,
+      "learning_rate": 1.3130422720664596e-05,
+      "loss": 0.9944,
+      "step": 7266
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7509828805923462,
+      "learning_rate": 1.3128658420167985e-05,
+      "loss": 0.9935,
+      "step": 7267
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.729807734489441,
+      "learning_rate": 1.312689401171011e-05,
+      "loss": 0.9767,
+      "step": 7268
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6518383026123047,
+      "learning_rate": 1.3125129495351856e-05,
+      "loss": 1.0219,
+      "step": 7269
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7479612827301025,
+      "learning_rate": 1.3123364871154113e-05,
+      "loss": 1.0241,
+      "step": 7270
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7142798900604248,
+      "learning_rate": 1.3121600139177777e-05,
+      "loss": 1.0556,
+      "step": 7271
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8249727487564087,
+      "learning_rate": 1.3119835299483738e-05,
+      "loss": 0.9209,
+      "step": 7272
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7366821765899658,
+      "learning_rate": 1.3118070352132896e-05,
+      "loss": 1.0192,
+      "step": 7273
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.795507788658142,
+      "learning_rate": 1.3116305297186159e-05,
+      "loss": 0.9037,
+      "step": 7274
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6106244325637817,
+      "learning_rate": 1.311454013470443e-05,
+      "loss": 0.9856,
+      "step": 7275
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8018512725830078,
+      "learning_rate": 1.311277486474862e-05,
+      "loss": 0.9717,
+      "step": 7276
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7001457214355469,
+      "learning_rate": 1.3111009487379647e-05,
+      "loss": 0.9706,
+      "step": 7277
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.211688756942749,
+      "learning_rate": 1.3109244002658425e-05,
+      "loss": 1.0413,
+      "step": 7278
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6731610298156738,
+      "learning_rate": 1.3107478410645875e-05,
+      "loss": 0.9965,
+      "step": 7279
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6553643941879272,
+      "learning_rate": 1.310571271140293e-05,
+      "loss": 0.947,
+      "step": 7280
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8816719055175781,
+      "learning_rate": 1.3103946904990515e-05,
+      "loss": 0.9851,
+      "step": 7281
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8195414543151855,
+      "learning_rate": 1.310218099146956e-05,
+      "loss": 1.0232,
+      "step": 7282
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7219852209091187,
+      "learning_rate": 1.3100414970901008e-05,
+      "loss": 1.0163,
+      "step": 7283
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.802555799484253,
+      "learning_rate": 1.3098648843345789e-05,
+      "loss": 1.0537,
+      "step": 7284
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5903615951538086,
+      "learning_rate": 1.309688260886486e-05,
+      "loss": 1.0201,
+      "step": 7285
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8213194608688354,
+      "learning_rate": 1.3095116267519163e-05,
+      "loss": 0.9884,
+      "step": 7286
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7187060117721558,
+      "learning_rate": 1.3093349819369647e-05,
+      "loss": 1.0325,
+      "step": 7287
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.052876353263855,
+      "learning_rate": 1.3091583264477273e-05,
+      "loss": 0.5511,
+      "step": 7288
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8216989040374756,
+      "learning_rate": 1.3089816602902993e-05,
+      "loss": 0.9528,
+      "step": 7289
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8386342525482178,
+      "learning_rate": 1.3088049834707777e-05,
+      "loss": 1.0003,
+      "step": 7290
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0155565738677979,
+      "learning_rate": 1.3086282959952583e-05,
+      "loss": 0.6111,
+      "step": 7291
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6320468187332153,
+      "learning_rate": 1.3084515978698389e-05,
+      "loss": 0.9852,
+      "step": 7292
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.714259147644043,
+      "learning_rate": 1.3082748891006164e-05,
+      "loss": 1.0311,
+      "step": 7293
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7873072624206543,
+      "learning_rate": 1.3080981696936883e-05,
+      "loss": 0.9709,
+      "step": 7294
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9880605340003967,
+      "learning_rate": 1.3079214396551532e-05,
+      "loss": 0.5795,
+      "step": 7295
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7565131187438965,
+      "learning_rate": 1.3077446989911092e-05,
+      "loss": 1.033,
+      "step": 7296
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7618820667266846,
+      "learning_rate": 1.3075679477076556e-05,
+      "loss": 0.894,
+      "step": 7297
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6277841329574585,
+      "learning_rate": 1.3073911858108911e-05,
+      "loss": 0.9919,
+      "step": 7298
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8146573305130005,
+      "learning_rate": 1.3072144133069156e-05,
+      "loss": 0.9427,
+      "step": 7299
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.079033374786377,
+      "learning_rate": 1.3070376302018287e-05,
+      "loss": 0.9939,
+      "step": 7300
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6928759813308716,
+      "learning_rate": 1.3068608365017308e-05,
+      "loss": 1.0517,
+      "step": 7301
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5805751085281372,
+      "learning_rate": 1.3066840322127227e-05,
+      "loss": 0.9607,
+      "step": 7302
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7294808626174927,
+      "learning_rate": 1.3065072173409055e-05,
+      "loss": 1.0378,
+      "step": 7303
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8863801956176758,
+      "learning_rate": 1.3063303918923802e-05,
+      "loss": 0.9914,
+      "step": 7304
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.181049346923828,
+      "learning_rate": 1.306153555873249e-05,
+      "loss": 1.0169,
+      "step": 7305
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8362523317337036,
+      "learning_rate": 1.3059767092896136e-05,
+      "loss": 0.9995,
+      "step": 7306
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7565770149230957,
+      "learning_rate": 1.3057998521475768e-05,
+      "loss": 1.0351,
+      "step": 7307
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8455919027328491,
+      "learning_rate": 1.305622984453241e-05,
+      "loss": 1.0054,
+      "step": 7308
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5867325067520142,
+      "learning_rate": 1.3054461062127099e-05,
+      "loss": 0.902,
+      "step": 7309
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.780295491218567,
+      "learning_rate": 1.305269217432087e-05,
+      "loss": 0.9707,
+      "step": 7310
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6085394620895386,
+      "learning_rate": 1.3050923181174762e-05,
+      "loss": 0.937,
+      "step": 7311
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6629457473754883,
+      "learning_rate": 1.3049154082749813e-05,
+      "loss": 1.0596,
+      "step": 7312
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.814444661140442,
+      "learning_rate": 1.3047384879107079e-05,
+      "loss": 0.9792,
+      "step": 7313
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8534595966339111,
+      "learning_rate": 1.3045615570307604e-05,
+      "loss": 0.9752,
+      "step": 7314
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9800876379013062,
+      "learning_rate": 1.3043846156412443e-05,
+      "loss": 0.9695,
+      "step": 7315
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8575453758239746,
+      "learning_rate": 1.3042076637482655e-05,
+      "loss": 0.9578,
+      "step": 7316
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.806773066520691,
+      "learning_rate": 1.3040307013579299e-05,
+      "loss": 1.05,
+      "step": 7317
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8021187782287598,
+      "learning_rate": 1.3038537284763443e-05,
+      "loss": 0.9848,
+      "step": 7318
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9050631523132324,
+      "learning_rate": 1.3036767451096148e-05,
+      "loss": 0.9586,
+      "step": 7319
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.667320728302002,
+      "learning_rate": 1.3034997512638493e-05,
+      "loss": 0.9902,
+      "step": 7320
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8604519367218018,
+      "learning_rate": 1.3033227469451555e-05,
+      "loss": 0.9761,
+      "step": 7321
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6019165515899658,
+      "learning_rate": 1.3031457321596409e-05,
+      "loss": 0.9924,
+      "step": 7322
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.70156729221344,
+      "learning_rate": 1.3029687069134134e-05,
+      "loss": 0.9556,
+      "step": 7323
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.903174638748169,
+      "learning_rate": 1.3027916712125825e-05,
+      "loss": 0.966,
+      "step": 7324
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7214933633804321,
+      "learning_rate": 1.302614625063257e-05,
+      "loss": 0.961,
+      "step": 7325
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.663572072982788,
+      "learning_rate": 1.3024375684715458e-05,
+      "loss": 0.8844,
+      "step": 7326
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.064779281616211,
+      "learning_rate": 1.3022605014435591e-05,
+      "loss": 0.5537,
+      "step": 7327
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7520649433135986,
+      "learning_rate": 1.3020834239854068e-05,
+      "loss": 0.9155,
+      "step": 7328
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7561284303665161,
+      "learning_rate": 1.3019063361031994e-05,
+      "loss": 1.0436,
+      "step": 7329
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6289538145065308,
+      "learning_rate": 1.3017292378030477e-05,
+      "loss": 0.9728,
+      "step": 7330
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8337544202804565,
+      "learning_rate": 1.3015521290910628e-05,
+      "loss": 1.0454,
+      "step": 7331
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6629467010498047,
+      "learning_rate": 1.3013750099733561e-05,
+      "loss": 0.9906,
+      "step": 7332
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8491228818893433,
+      "learning_rate": 1.3011978804560401e-05,
+      "loss": 0.8925,
+      "step": 7333
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7411162853240967,
+      "learning_rate": 1.3010207405452265e-05,
+      "loss": 1.0043,
+      "step": 7334
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8061507940292358,
+      "learning_rate": 1.3008435902470276e-05,
+      "loss": 1.0701,
+      "step": 7335
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9804774522781372,
+      "learning_rate": 1.300666429567557e-05,
+      "loss": 0.9755,
+      "step": 7336
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7164946794509888,
+      "learning_rate": 1.3004892585129279e-05,
+      "loss": 0.9141,
+      "step": 7337
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6265473365783691,
+      "learning_rate": 1.3003120770892536e-05,
+      "loss": 0.9445,
+      "step": 7338
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.803780436515808,
+      "learning_rate": 1.3001348853026488e-05,
+      "loss": 0.9294,
+      "step": 7339
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7544447183609009,
+      "learning_rate": 1.2999576831592273e-05,
+      "loss": 1.0138,
+      "step": 7340
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.751644492149353,
+      "learning_rate": 1.299780470665104e-05,
+      "loss": 1.002,
+      "step": 7341
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6455060243606567,
+      "learning_rate": 1.2996032478263943e-05,
+      "loss": 0.9731,
+      "step": 7342
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7202250957489014,
+      "learning_rate": 1.2994260146492133e-05,
+      "loss": 1.0097,
+      "step": 7343
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7249274253845215,
+      "learning_rate": 1.2992487711396768e-05,
+      "loss": 0.959,
+      "step": 7344
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7013154029846191,
+      "learning_rate": 1.299071517303901e-05,
+      "loss": 1.0827,
+      "step": 7345
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.958591341972351,
+      "learning_rate": 1.2988942531480028e-05,
+      "loss": 1.1027,
+      "step": 7346
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5968042612075806,
+      "learning_rate": 1.2987169786780988e-05,
+      "loss": 0.9098,
+      "step": 7347
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6995233297348022,
+      "learning_rate": 1.2985396939003065e-05,
+      "loss": 0.9614,
+      "step": 7348
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8575081825256348,
+      "learning_rate": 1.2983623988207432e-05,
+      "loss": 1.0174,
+      "step": 7349
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8969746828079224,
+      "learning_rate": 1.2981850934455267e-05,
+      "loss": 1.0359,
+      "step": 7350
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7084649801254272,
+      "learning_rate": 1.2980077777807755e-05,
+      "loss": 0.9774,
+      "step": 7351
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8542094230651855,
+      "learning_rate": 1.2978304518326088e-05,
+      "loss": 0.939,
+      "step": 7352
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9423540830612183,
+      "learning_rate": 1.297653115607145e-05,
+      "loss": 1.062,
+      "step": 7353
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.897090196609497,
+      "learning_rate": 1.2974757691105038e-05,
+      "loss": 0.9799,
+      "step": 7354
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8937289714813232,
+      "learning_rate": 1.2972984123488045e-05,
+      "loss": 1.013,
+      "step": 7355
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8164328336715698,
+      "learning_rate": 1.2971210453281675e-05,
+      "loss": 0.9474,
+      "step": 7356
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8939083814620972,
+      "learning_rate": 1.2969436680547132e-05,
+      "loss": 0.9777,
+      "step": 7357
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.850019931793213,
+      "learning_rate": 1.2967662805345625e-05,
+      "loss": 1.0043,
+      "step": 7358
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.152945876121521,
+      "learning_rate": 1.2965888827738365e-05,
+      "loss": 0.5927,
+      "step": 7359
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8057949542999268,
+      "learning_rate": 1.2964114747786564e-05,
+      "loss": 0.995,
+      "step": 7360
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6508755683898926,
+      "learning_rate": 1.2962340565551443e-05,
+      "loss": 0.9802,
+      "step": 7361
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7664345502853394,
+      "learning_rate": 1.2960566281094224e-05,
+      "loss": 0.9584,
+      "step": 7362
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.007944107055664,
+      "learning_rate": 1.2958791894476134e-05,
+      "loss": 0.8857,
+      "step": 7363
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7279704809188843,
+      "learning_rate": 1.29570174057584e-05,
+      "loss": 0.8938,
+      "step": 7364
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7731008529663086,
+      "learning_rate": 1.2955242815002258e-05,
+      "loss": 0.9324,
+      "step": 7365
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8896539211273193,
+      "learning_rate": 1.295346812226894e-05,
+      "loss": 0.9683,
+      "step": 7366
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9950226545333862,
+      "learning_rate": 1.2951693327619689e-05,
+      "loss": 0.9765,
+      "step": 7367
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7307549715042114,
+      "learning_rate": 1.2949918431115742e-05,
+      "loss": 0.9967,
+      "step": 7368
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.878368616104126,
+      "learning_rate": 1.2948143432818352e-05,
+      "loss": 0.978,
+      "step": 7369
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8204724788665771,
+      "learning_rate": 1.294636833278877e-05,
+      "loss": 0.9809,
+      "step": 7370
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.819919466972351,
+      "learning_rate": 1.2944593131088246e-05,
+      "loss": 1.0129,
+      "step": 7371
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6566475629806519,
+      "learning_rate": 1.2942817827778037e-05,
+      "loss": 0.9865,
+      "step": 7372
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9487743377685547,
+      "learning_rate": 1.2941042422919405e-05,
+      "loss": 0.999,
+      "step": 7373
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7331446409225464,
+      "learning_rate": 1.2939266916573614e-05,
+      "loss": 0.9563,
+      "step": 7374
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8267320394515991,
+      "learning_rate": 1.2937491308801936e-05,
+      "loss": 1.0745,
+      "step": 7375
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1133317947387695,
+      "learning_rate": 1.2935715599665635e-05,
+      "loss": 0.612,
+      "step": 7376
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6918549537658691,
+      "learning_rate": 1.293393978922599e-05,
+      "loss": 0.8991,
+      "step": 7377
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.829156756401062,
+      "learning_rate": 1.2932163877544277e-05,
+      "loss": 0.9734,
+      "step": 7378
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5627474784851074,
+      "learning_rate": 1.293038786468178e-05,
+      "loss": 0.9992,
+      "step": 7379
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8583108186721802,
+      "learning_rate": 1.2928611750699784e-05,
+      "loss": 0.9215,
+      "step": 7380
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8099446296691895,
+      "learning_rate": 1.2926835535659579e-05,
+      "loss": 0.9328,
+      "step": 7381
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0165079832077026,
+      "learning_rate": 1.2925059219622455e-05,
+      "loss": 0.6059,
+      "step": 7382
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6953763961791992,
+      "learning_rate": 1.2923282802649708e-05,
+      "loss": 0.9706,
+      "step": 7383
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8917162418365479,
+      "learning_rate": 1.2921506284802636e-05,
+      "loss": 1.0104,
+      "step": 7384
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0757641792297363,
+      "learning_rate": 1.2919729666142545e-05,
+      "loss": 0.6491,
+      "step": 7385
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9731879234313965,
+      "learning_rate": 1.2917952946730737e-05,
+      "loss": 1.0203,
+      "step": 7386
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8856728076934814,
+      "learning_rate": 1.2916176126628527e-05,
+      "loss": 0.9763,
+      "step": 7387
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8110123872756958,
+      "learning_rate": 1.2914399205897221e-05,
+      "loss": 1.0556,
+      "step": 7388
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7422555685043335,
+      "learning_rate": 1.2912622184598138e-05,
+      "loss": 0.9648,
+      "step": 7389
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8223832845687866,
+      "learning_rate": 1.2910845062792604e-05,
+      "loss": 1.0258,
+      "step": 7390
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8142021894454956,
+      "learning_rate": 1.2909067840541935e-05,
+      "loss": 1.0817,
+      "step": 7391
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.776807188987732,
+      "learning_rate": 1.2907290517907462e-05,
+      "loss": 1.0026,
+      "step": 7392
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6979384422302246,
+      "learning_rate": 1.2905513094950517e-05,
+      "loss": 0.9178,
+      "step": 7393
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8462640047073364,
+      "learning_rate": 1.290373557173243e-05,
+      "loss": 1.0767,
+      "step": 7394
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8140606880187988,
+      "learning_rate": 1.2901957948314539e-05,
+      "loss": 1.0402,
+      "step": 7395
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7951174974441528,
+      "learning_rate": 1.2900180224758186e-05,
+      "loss": 0.9011,
+      "step": 7396
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9115961790084839,
+      "learning_rate": 1.2898402401124713e-05,
+      "loss": 1.0766,
+      "step": 7397
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0440149307250977,
+      "learning_rate": 1.289662447747547e-05,
+      "loss": 1.0204,
+      "step": 7398
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7560451030731201,
+      "learning_rate": 1.289484645387181e-05,
+      "loss": 0.9514,
+      "step": 7399
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8194767236709595,
+      "learning_rate": 1.2893068330375082e-05,
+      "loss": 1.047,
+      "step": 7400
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9434022903442383,
+      "learning_rate": 1.2891290107046647e-05,
+      "loss": 0.9533,
+      "step": 7401
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7496813535690308,
+      "learning_rate": 1.288951178394787e-05,
+      "loss": 0.9679,
+      "step": 7402
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.859490156173706,
+      "learning_rate": 1.288773336114011e-05,
+      "loss": 1.0121,
+      "step": 7403
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.778328776359558,
+      "learning_rate": 1.2885954838684742e-05,
+      "loss": 0.9174,
+      "step": 7404
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.059141755104065,
+      "learning_rate": 1.2884176216643132e-05,
+      "loss": 0.5498,
+      "step": 7405
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.801032304763794,
+      "learning_rate": 1.2882397495076657e-05,
+      "loss": 1.0558,
+      "step": 7406
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6838574409484863,
+      "learning_rate": 1.28806186740467e-05,
+      "loss": 0.9721,
+      "step": 7407
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7295992374420166,
+      "learning_rate": 1.2878839753614633e-05,
+      "loss": 0.9707,
+      "step": 7408
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8102807998657227,
+      "learning_rate": 1.287706073384185e-05,
+      "loss": 1.0825,
+      "step": 7409
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6353942155838013,
+      "learning_rate": 1.287528161478974e-05,
+      "loss": 0.9245,
+      "step": 7410
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7287774085998535,
+      "learning_rate": 1.2873502396519692e-05,
+      "loss": 1.027,
+      "step": 7411
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.825989007949829,
+      "learning_rate": 1.2871723079093101e-05,
+      "loss": 0.9144,
+      "step": 7412
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.811667799949646,
+      "learning_rate": 1.2869943662571372e-05,
+      "loss": 0.994,
+      "step": 7413
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.970324158668518,
+      "learning_rate": 1.28681641470159e-05,
+      "loss": 1.0173,
+      "step": 7414
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9308664798736572,
+      "learning_rate": 1.2866384532488098e-05,
+      "loss": 1.0567,
+      "step": 7415
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6111783981323242,
+      "learning_rate": 1.286460481904937e-05,
+      "loss": 0.9629,
+      "step": 7416
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7394113540649414,
+      "learning_rate": 1.2862825006761136e-05,
+      "loss": 0.9924,
+      "step": 7417
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8474863767623901,
+      "learning_rate": 1.2861045095684805e-05,
+      "loss": 0.9743,
+      "step": 7418
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7442917823791504,
+      "learning_rate": 1.28592650858818e-05,
+      "loss": 1.0255,
+      "step": 7419
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9154739379882812,
+      "learning_rate": 1.2857484977413545e-05,
+      "loss": 1.0643,
+      "step": 7420
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.830653429031372,
+      "learning_rate": 1.2855704770341463e-05,
+      "loss": 1.0451,
+      "step": 7421
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7709776163101196,
+      "learning_rate": 1.285392446472699e-05,
+      "loss": 0.9858,
+      "step": 7422
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8861711025238037,
+      "learning_rate": 1.2852144060631556e-05,
+      "loss": 0.8848,
+      "step": 7423
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8036357164382935,
+      "learning_rate": 1.2850363558116596e-05,
+      "loss": 1.0782,
+      "step": 7424
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.729099988937378,
+      "learning_rate": 1.2848582957243552e-05,
+      "loss": 0.919,
+      "step": 7425
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6754510402679443,
+      "learning_rate": 1.2846802258073867e-05,
+      "loss": 0.9521,
+      "step": 7426
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7148399353027344,
+      "learning_rate": 1.2845021460668988e-05,
+      "loss": 0.6629,
+      "step": 7427
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8017168045043945,
+      "learning_rate": 1.2843240565090365e-05,
+      "loss": 0.9795,
+      "step": 7428
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7993075847625732,
+      "learning_rate": 1.2841459571399453e-05,
+      "loss": 0.9253,
+      "step": 7429
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8552762269973755,
+      "learning_rate": 1.2839678479657709e-05,
+      "loss": 0.9809,
+      "step": 7430
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.77656888961792,
+      "learning_rate": 1.2837897289926592e-05,
+      "loss": 0.9731,
+      "step": 7431
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5843607187271118,
+      "learning_rate": 1.283611600226757e-05,
+      "loss": 0.8335,
+      "step": 7432
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7424556016921997,
+      "learning_rate": 1.2834334616742104e-05,
+      "loss": 0.9747,
+      "step": 7433
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.724855661392212,
+      "learning_rate": 1.2832553133411666e-05,
+      "loss": 1.0031,
+      "step": 7434
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.924340844154358,
+      "learning_rate": 1.2830771552337735e-05,
+      "loss": 1.1026,
+      "step": 7435
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8070130348205566,
+      "learning_rate": 1.2828989873581786e-05,
+      "loss": 1.0251,
+      "step": 7436
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.75368070602417,
+      "learning_rate": 1.2827208097205298e-05,
+      "loss": 1.009,
+      "step": 7437
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8113374710083008,
+      "learning_rate": 1.2825426223269755e-05,
+      "loss": 1.0823,
+      "step": 7438
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.751863956451416,
+      "learning_rate": 1.2823644251836647e-05,
+      "loss": 0.9411,
+      "step": 7439
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8712741136550903,
+      "learning_rate": 1.282186218296746e-05,
+      "loss": 1.0149,
+      "step": 7440
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.980992078781128,
+      "learning_rate": 1.2820080016723695e-05,
+      "loss": 1.0026,
+      "step": 7441
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9852229356765747,
+      "learning_rate": 1.2818297753166844e-05,
+      "loss": 1.0551,
+      "step": 7442
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7283494472503662,
+      "learning_rate": 1.2816515392358413e-05,
+      "loss": 1.039,
+      "step": 7443
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9550997018814087,
+      "learning_rate": 1.2814732934359901e-05,
+      "loss": 1.0094,
+      "step": 7444
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0406551361083984,
+      "learning_rate": 1.2812950379232816e-05,
+      "loss": 1.0285,
+      "step": 7445
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9118375778198242,
+      "learning_rate": 1.2811167727038675e-05,
+      "loss": 0.9471,
+      "step": 7446
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7764334678649902,
+      "learning_rate": 1.2809384977838988e-05,
+      "loss": 1.0103,
+      "step": 7447
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.877103567123413,
+      "learning_rate": 1.2807602131695274e-05,
+      "loss": 0.9244,
+      "step": 7448
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8031103610992432,
+      "learning_rate": 1.2805819188669051e-05,
+      "loss": 1.0056,
+      "step": 7449
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6693018674850464,
+      "learning_rate": 1.2804036148821846e-05,
+      "loss": 0.9546,
+      "step": 7450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9337722063064575,
+      "learning_rate": 1.2802253012215187e-05,
+      "loss": 0.9952,
+      "step": 7451
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.594313144683838,
+      "learning_rate": 1.2800469778910603e-05,
+      "loss": 0.9433,
+      "step": 7452
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5926663875579834,
+      "learning_rate": 1.279868644896963e-05,
+      "loss": 0.9507,
+      "step": 7453
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8125909566879272,
+      "learning_rate": 1.2796903022453808e-05,
+      "loss": 0.9863,
+      "step": 7454
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7210795879364014,
+      "learning_rate": 1.279511949942467e-05,
+      "loss": 0.9471,
+      "step": 7455
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2077887058258057,
+      "learning_rate": 1.2793335879943771e-05,
+      "loss": 0.6219,
+      "step": 7456
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1810319423675537,
+      "learning_rate": 1.2791552164072652e-05,
+      "loss": 0.5648,
+      "step": 7457
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7424389123916626,
+      "learning_rate": 1.2789768351872867e-05,
+      "loss": 0.9128,
+      "step": 7458
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.820887565612793,
+      "learning_rate": 1.2787984443405966e-05,
+      "loss": 1.0088,
+      "step": 7459
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6214430332183838,
+      "learning_rate": 1.2786200438733512e-05,
+      "loss": 0.8781,
+      "step": 7460
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0633918046951294,
+      "learning_rate": 1.2784416337917063e-05,
+      "loss": 0.5948,
+      "step": 7461
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6853479146957397,
+      "learning_rate": 1.2782632141018185e-05,
+      "loss": 1.0068,
+      "step": 7462
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7700046300888062,
+      "learning_rate": 1.2780847848098445e-05,
+      "loss": 1.0108,
+      "step": 7463
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.879514217376709,
+      "learning_rate": 1.2779063459219414e-05,
+      "loss": 0.984,
+      "step": 7464
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.722773790359497,
+      "learning_rate": 1.2777278974442664e-05,
+      "loss": 1.0109,
+      "step": 7465
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8056957721710205,
+      "learning_rate": 1.2775494393829777e-05,
+      "loss": 1.0415,
+      "step": 7466
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9671504497528076,
+      "learning_rate": 1.2773709717442326e-05,
+      "loss": 1.0492,
+      "step": 7467
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.986178994178772,
+      "learning_rate": 1.2771924945341906e-05,
+      "loss": 0.9528,
+      "step": 7468
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8241297006607056,
+      "learning_rate": 1.2770140077590098e-05,
+      "loss": 1.0811,
+      "step": 7469
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9104567766189575,
+      "learning_rate": 1.2768355114248493e-05,
+      "loss": 1.0288,
+      "step": 7470
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.701006293296814,
+      "learning_rate": 1.276657005537869e-05,
+      "loss": 0.9418,
+      "step": 7471
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7929165363311768,
+      "learning_rate": 1.276478490104228e-05,
+      "loss": 0.9289,
+      "step": 7472
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7173610925674438,
+      "learning_rate": 1.2762999651300865e-05,
+      "loss": 0.9649,
+      "step": 7473
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8171846866607666,
+      "learning_rate": 1.2761214306216052e-05,
+      "loss": 1.0279,
+      "step": 7474
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.109318256378174,
+      "learning_rate": 1.2759428865849445e-05,
+      "loss": 0.9936,
+      "step": 7475
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.917927622795105,
+      "learning_rate": 1.2757643330262656e-05,
+      "loss": 0.925,
+      "step": 7476
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9283018112182617,
+      "learning_rate": 1.27558576995173e-05,
+      "loss": 0.9698,
+      "step": 7477
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7373062372207642,
+      "learning_rate": 1.275407197367499e-05,
+      "loss": 0.9673,
+      "step": 7478
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8121575117111206,
+      "learning_rate": 1.2752286152797352e-05,
+      "loss": 1.0262,
+      "step": 7479
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.773172378540039,
+      "learning_rate": 1.2750500236946008e-05,
+      "loss": 0.941,
+      "step": 7480
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9105595350265503,
+      "learning_rate": 1.2748714226182583e-05,
+      "loss": 1.1366,
+      "step": 7481
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6814420223236084,
+      "learning_rate": 1.2746928120568707e-05,
+      "loss": 1.0099,
+      "step": 7482
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.756831169128418,
+      "learning_rate": 1.2745141920166016e-05,
+      "loss": 1.0091,
+      "step": 7483
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.653228998184204,
+      "learning_rate": 1.2743355625036145e-05,
+      "loss": 1.0106,
+      "step": 7484
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8476029634475708,
+      "learning_rate": 1.2741569235240733e-05,
+      "loss": 0.995,
+      "step": 7485
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7879499197006226,
+      "learning_rate": 1.2739782750841428e-05,
+      "loss": 0.9372,
+      "step": 7486
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9753953218460083,
+      "learning_rate": 1.2737996171899873e-05,
+      "loss": 1.0142,
+      "step": 7487
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8585056066513062,
+      "learning_rate": 1.2736209498477719e-05,
+      "loss": 0.9522,
+      "step": 7488
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7398544549942017,
+      "learning_rate": 1.2734422730636617e-05,
+      "loss": 0.9327,
+      "step": 7489
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7137055397033691,
+      "learning_rate": 1.2732635868438225e-05,
+      "loss": 0.9418,
+      "step": 7490
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5938587188720703,
+      "learning_rate": 1.2730848911944204e-05,
+      "loss": 1.0401,
+      "step": 7491
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7662118673324585,
+      "learning_rate": 1.2729061861216214e-05,
+      "loss": 1.039,
+      "step": 7492
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.983756184577942,
+      "learning_rate": 1.2727274716315922e-05,
+      "loss": 0.9517,
+      "step": 7493
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3117972612380981,
+      "learning_rate": 1.2725487477304999e-05,
+      "loss": 0.6674,
+      "step": 7494
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7978792190551758,
+      "learning_rate": 1.2723700144245115e-05,
+      "loss": 0.9926,
+      "step": 7495
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7574079036712646,
+      "learning_rate": 1.2721912717197949e-05,
+      "loss": 0.9103,
+      "step": 7496
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6949598789215088,
+      "learning_rate": 1.2720125196225178e-05,
+      "loss": 1.0089,
+      "step": 7497
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9450141191482544,
+      "learning_rate": 1.2718337581388485e-05,
+      "loss": 1.0455,
+      "step": 7498
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7450401782989502,
+      "learning_rate": 1.2716549872749555e-05,
+      "loss": 0.9784,
+      "step": 7499
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7866034507751465,
+      "learning_rate": 1.2714762070370078e-05,
+      "loss": 0.978,
+      "step": 7500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8792423009872437,
+      "learning_rate": 1.2712974174311743e-05,
+      "loss": 0.9371,
+      "step": 7501
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7580718994140625,
+      "learning_rate": 1.271118618463625e-05,
+      "loss": 1.059,
+      "step": 7502
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.822844386100769,
+      "learning_rate": 1.2709398101405296e-05,
+      "loss": 0.9845,
+      "step": 7503
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7986738681793213,
+      "learning_rate": 1.270760992468058e-05,
+      "loss": 0.8986,
+      "step": 7504
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7601100206375122,
+      "learning_rate": 1.2705821654523809e-05,
+      "loss": 1.0599,
+      "step": 7505
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.813385009765625,
+      "learning_rate": 1.270403329099669e-05,
+      "loss": 0.9624,
+      "step": 7506
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6762651205062866,
+      "learning_rate": 1.2702244834160937e-05,
+      "loss": 1.043,
+      "step": 7507
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7727586030960083,
+      "learning_rate": 1.2700456284078263e-05,
+      "loss": 1.0071,
+      "step": 7508
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.211315155029297,
+      "learning_rate": 1.2698667640810386e-05,
+      "loss": 0.993,
+      "step": 7509
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9319093227386475,
+      "learning_rate": 1.2696878904419028e-05,
+      "loss": 0.9659,
+      "step": 7510
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8859407901763916,
+      "learning_rate": 1.269509007496591e-05,
+      "loss": 0.9505,
+      "step": 7511
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8175667524337769,
+      "learning_rate": 1.2693301152512765e-05,
+      "loss": 0.9882,
+      "step": 7512
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6579301357269287,
+      "learning_rate": 1.269151213712132e-05,
+      "loss": 0.8946,
+      "step": 7513
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8423120975494385,
+      "learning_rate": 1.268972302885331e-05,
+      "loss": 0.9619,
+      "step": 7514
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7640901803970337,
+      "learning_rate": 1.2687933827770469e-05,
+      "loss": 1.0234,
+      "step": 7515
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6898865699768066,
+      "learning_rate": 1.268614453393454e-05,
+      "loss": 0.8837,
+      "step": 7516
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.79181969165802,
+      "learning_rate": 1.268435514740727e-05,
+      "loss": 0.9206,
+      "step": 7517
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.774214506149292,
+      "learning_rate": 1.2682565668250401e-05,
+      "loss": 1.0339,
+      "step": 7518
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1134421825408936,
+      "learning_rate": 1.2680776096525684e-05,
+      "loss": 0.6506,
+      "step": 7519
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8094152212142944,
+      "learning_rate": 1.2678986432294872e-05,
+      "loss": 0.9961,
+      "step": 7520
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1506853103637695,
+      "learning_rate": 1.267719667561972e-05,
+      "loss": 0.9644,
+      "step": 7521
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.753096342086792,
+      "learning_rate": 1.2675406826561991e-05,
+      "loss": 1.0185,
+      "step": 7522
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7113707065582275,
+      "learning_rate": 1.2673616885183449e-05,
+      "loss": 0.983,
+      "step": 7523
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7953295707702637,
+      "learning_rate": 1.2671826851545851e-05,
+      "loss": 1.0852,
+      "step": 7524
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8164705038070679,
+      "learning_rate": 1.2670036725710974e-05,
+      "loss": 0.9757,
+      "step": 7525
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.865089774131775,
+      "learning_rate": 1.266824650774059e-05,
+      "loss": 0.9623,
+      "step": 7526
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0285768508911133,
+      "learning_rate": 1.2666456197696473e-05,
+      "loss": 0.9957,
+      "step": 7527
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6920452117919922,
+      "learning_rate": 1.2664665795640399e-05,
+      "loss": 1.0866,
+      "step": 7528
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9247956275939941,
+      "learning_rate": 1.2662875301634152e-05,
+      "loss": 1.0301,
+      "step": 7529
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8386216163635254,
+      "learning_rate": 1.2661084715739516e-05,
+      "loss": 1.0137,
+      "step": 7530
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1674944162368774,
+      "learning_rate": 1.2659294038018279e-05,
+      "loss": 0.6569,
+      "step": 7531
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7139341831207275,
+      "learning_rate": 1.2657503268532236e-05,
+      "loss": 0.9706,
+      "step": 7532
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9618979692459106,
+      "learning_rate": 1.2655712407343175e-05,
+      "loss": 1.0096,
+      "step": 7533
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7249456644058228,
+      "learning_rate": 1.26539214545129e-05,
+      "loss": 0.8991,
+      "step": 7534
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8893961906433105,
+      "learning_rate": 1.265213041010321e-05,
+      "loss": 1.0104,
+      "step": 7535
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.101256012916565,
+      "learning_rate": 1.2650339274175906e-05,
+      "loss": 0.6349,
+      "step": 7536
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0688300132751465,
+      "learning_rate": 1.26485480467928e-05,
+      "loss": 0.5672,
+      "step": 7537
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8091368675231934,
+      "learning_rate": 1.2646756728015696e-05,
+      "loss": 0.98,
+      "step": 7538
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.759499192237854,
+      "learning_rate": 1.2644965317906413e-05,
+      "loss": 0.9828,
+      "step": 7539
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.136631965637207,
+      "learning_rate": 1.2643173816526763e-05,
+      "loss": 0.5935,
+      "step": 7540
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0222177505493164,
+      "learning_rate": 1.264138222393857e-05,
+      "loss": 0.9537,
+      "step": 7541
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7345727682113647,
+      "learning_rate": 1.2639590540203654e-05,
+      "loss": 0.9587,
+      "step": 7542
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.874272346496582,
+      "learning_rate": 1.2637798765383842e-05,
+      "loss": 0.94,
+      "step": 7543
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6331371068954468,
+      "learning_rate": 1.263600689954096e-05,
+      "loss": 1.004,
+      "step": 7544
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7173949480056763,
+      "learning_rate": 1.2634214942736847e-05,
+      "loss": 0.9903,
+      "step": 7545
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.018784523010254,
+      "learning_rate": 1.2632422895033333e-05,
+      "loss": 1.0095,
+      "step": 7546
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0410802364349365,
+      "learning_rate": 1.263063075649226e-05,
+      "loss": 1.0811,
+      "step": 7547
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8046157360076904,
+      "learning_rate": 1.2628838527175464e-05,
+      "loss": 1.0396,
+      "step": 7548
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1469578742980957,
+      "learning_rate": 1.2627046207144798e-05,
+      "loss": 0.6852,
+      "step": 7549
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.041348934173584,
+      "learning_rate": 1.2625253796462104e-05,
+      "loss": 0.9877,
+      "step": 7550
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8755077123641968,
+      "learning_rate": 1.2623461295189236e-05,
+      "loss": 1.0647,
+      "step": 7551
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6887931823730469,
+      "learning_rate": 1.2621668703388046e-05,
+      "loss": 0.9837,
+      "step": 7552
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7082523107528687,
+      "learning_rate": 1.2619876021120394e-05,
+      "loss": 1.0328,
+      "step": 7553
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6883245706558228,
+      "learning_rate": 1.2618083248448137e-05,
+      "loss": 0.9997,
+      "step": 7554
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8061022758483887,
+      "learning_rate": 1.2616290385433141e-05,
+      "loss": 0.9801,
+      "step": 7555
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6123614311218262,
+      "learning_rate": 1.2614497432137274e-05,
+      "loss": 0.9141,
+      "step": 7556
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7140716314315796,
+      "learning_rate": 1.26127043886224e-05,
+      "loss": 0.9544,
+      "step": 7557
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6678967475891113,
+      "learning_rate": 1.26109112549504e-05,
+      "loss": 0.9468,
+      "step": 7558
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.795954942703247,
+      "learning_rate": 1.2609118031183144e-05,
+      "loss": 0.9323,
+      "step": 7559
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6206252574920654,
+      "learning_rate": 1.260732471738251e-05,
+      "loss": 0.9129,
+      "step": 7560
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8071534633636475,
+      "learning_rate": 1.2605531313610386e-05,
+      "loss": 1.0176,
+      "step": 7561
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7835968732833862,
+      "learning_rate": 1.2603737819928656e-05,
+      "loss": 1.0631,
+      "step": 7562
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8254762887954712,
+      "learning_rate": 1.2601944236399207e-05,
+      "loss": 1.0146,
+      "step": 7563
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.772290825843811,
+      "learning_rate": 1.2600150563083929e-05,
+      "loss": 0.9648,
+      "step": 7564
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8369179964065552,
+      "learning_rate": 1.2598356800044717e-05,
+      "loss": 0.8646,
+      "step": 7565
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7670364379882812,
+      "learning_rate": 1.2596562947343473e-05,
+      "loss": 0.8997,
+      "step": 7566
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8656688928604126,
+      "learning_rate": 1.2594769005042093e-05,
+      "loss": 1.0005,
+      "step": 7567
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8663579225540161,
+      "learning_rate": 1.2592974973202486e-05,
+      "loss": 0.9596,
+      "step": 7568
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6148930788040161,
+      "learning_rate": 1.2591180851886554e-05,
+      "loss": 0.9745,
+      "step": 7569
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8621864318847656,
+      "learning_rate": 1.2589386641156208e-05,
+      "loss": 0.9315,
+      "step": 7570
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6840916872024536,
+      "learning_rate": 1.2587592341073362e-05,
+      "loss": 0.9844,
+      "step": 7571
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6677579879760742,
+      "learning_rate": 1.2585797951699932e-05,
+      "loss": 1.0499,
+      "step": 7572
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8730182647705078,
+      "learning_rate": 1.2584003473097837e-05,
+      "loss": 0.9146,
+      "step": 7573
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.676588773727417,
+      "learning_rate": 1.2582208905329004e-05,
+      "loss": 0.9929,
+      "step": 7574
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.028895378112793,
+      "learning_rate": 1.2580414248455352e-05,
+      "loss": 1.0946,
+      "step": 7575
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8122638463974,
+      "learning_rate": 1.2578619502538814e-05,
+      "loss": 1.0032,
+      "step": 7576
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5693435668945312,
+      "learning_rate": 1.2576824667641317e-05,
+      "loss": 0.9244,
+      "step": 7577
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7245745658874512,
+      "learning_rate": 1.2575029743824803e-05,
+      "loss": 0.933,
+      "step": 7578
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.736441731452942,
+      "learning_rate": 1.2573234731151203e-05,
+      "loss": 0.9737,
+      "step": 7579
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.622036337852478,
+      "learning_rate": 1.257143962968246e-05,
+      "loss": 0.929,
+      "step": 7580
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7343754768371582,
+      "learning_rate": 1.256964443948052e-05,
+      "loss": 1.0073,
+      "step": 7581
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7654688358306885,
+      "learning_rate": 1.2567849160607327e-05,
+      "loss": 0.9599,
+      "step": 7582
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6042214632034302,
+      "learning_rate": 1.2566053793124834e-05,
+      "loss": 0.9153,
+      "step": 7583
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6926013231277466,
+      "learning_rate": 1.2564258337094994e-05,
+      "loss": 0.9967,
+      "step": 7584
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.724165439605713,
+      "learning_rate": 1.2562462792579759e-05,
+      "loss": 0.9516,
+      "step": 7585
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7788724899291992,
+      "learning_rate": 1.2560667159641092e-05,
+      "loss": 0.9947,
+      "step": 7586
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7585062980651855,
+      "learning_rate": 1.2558871438340951e-05,
+      "loss": 0.9913,
+      "step": 7587
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8138631582260132,
+      "learning_rate": 1.2557075628741309e-05,
+      "loss": 1.032,
+      "step": 7588
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6059385538101196,
+      "learning_rate": 1.2555279730904128e-05,
+      "loss": 1.036,
+      "step": 7589
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7083841562271118,
+      "learning_rate": 1.2553483744891382e-05,
+      "loss": 1.0165,
+      "step": 7590
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8931975364685059,
+      "learning_rate": 1.2551687670765045e-05,
+      "loss": 0.9667,
+      "step": 7591
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1866353750228882,
+      "learning_rate": 1.2549891508587095e-05,
+      "loss": 0.6447,
+      "step": 7592
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8239480257034302,
+      "learning_rate": 1.254809525841951e-05,
+      "loss": 1.0698,
+      "step": 7593
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6755266189575195,
+      "learning_rate": 1.2546298920324277e-05,
+      "loss": 1.0232,
+      "step": 7594
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7988396883010864,
+      "learning_rate": 1.2544502494363382e-05,
+      "loss": 1.0231,
+      "step": 7595
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7721139192581177,
+      "learning_rate": 1.2542705980598813e-05,
+      "loss": 1.022,
+      "step": 7596
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8031858205795288,
+      "learning_rate": 1.254090937909256e-05,
+      "loss": 0.9576,
+      "step": 7597
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7613509893417358,
+      "learning_rate": 1.2539112689906627e-05,
+      "loss": 0.9114,
+      "step": 7598
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8318668603897095,
+      "learning_rate": 1.2537315913103003e-05,
+      "loss": 1.0007,
+      "step": 7599
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8220456838607788,
+      "learning_rate": 1.2535519048743696e-05,
+      "loss": 1.0284,
+      "step": 7600
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7607215642929077,
+      "learning_rate": 1.2533722096890713e-05,
+      "loss": 1.0121,
+      "step": 7601
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7073352336883545,
+      "learning_rate": 1.2531925057606053e-05,
+      "loss": 0.9583,
+      "step": 7602
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0248639583587646,
+      "learning_rate": 1.2530127930951736e-05,
+      "loss": 1.0115,
+      "step": 7603
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9395157098770142,
+      "learning_rate": 1.252833071698977e-05,
+      "loss": 1.0344,
+      "step": 7604
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.098695158958435,
+      "learning_rate": 1.2526533415782173e-05,
+      "loss": 0.6311,
+      "step": 7605
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.638378620147705,
+      "learning_rate": 1.2524736027390968e-05,
+      "loss": 0.9969,
+      "step": 7606
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9952657222747803,
+      "learning_rate": 1.2522938551878171e-05,
+      "loss": 0.9236,
+      "step": 7607
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6767350435256958,
+      "learning_rate": 1.2521140989305816e-05,
+      "loss": 0.9621,
+      "step": 7608
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.055700421333313,
+      "learning_rate": 1.2519343339735925e-05,
+      "loss": 0.6334,
+      "step": 7609
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6364747285842896,
+      "learning_rate": 1.2517545603230534e-05,
+      "loss": 0.9466,
+      "step": 7610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7909554243087769,
+      "learning_rate": 1.2515747779851677e-05,
+      "loss": 1.0274,
+      "step": 7611
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9285619258880615,
+      "learning_rate": 1.251394986966139e-05,
+      "loss": 0.9975,
+      "step": 7612
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7856172323226929,
+      "learning_rate": 1.2512151872721718e-05,
+      "loss": 1.0448,
+      "step": 7613
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8012738227844238,
+      "learning_rate": 1.25103537890947e-05,
+      "loss": 0.973,
+      "step": 7614
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.797629475593567,
+      "learning_rate": 1.2508555618842386e-05,
+      "loss": 0.9628,
+      "step": 7615
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8345519304275513,
+      "learning_rate": 1.2506757362026824e-05,
+      "loss": 0.9593,
+      "step": 7616
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7342952489852905,
+      "learning_rate": 1.250495901871007e-05,
+      "loss": 0.9834,
+      "step": 7617
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8109760284423828,
+      "learning_rate": 1.2503160588954178e-05,
+      "loss": 0.9337,
+      "step": 7618
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.911319613456726,
+      "learning_rate": 1.2501362072821204e-05,
+      "loss": 1.0144,
+      "step": 7619
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0442311763763428,
+      "learning_rate": 1.2499563470373213e-05,
+      "loss": 1.0142,
+      "step": 7620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9256324768066406,
+      "learning_rate": 1.249776478167227e-05,
+      "loss": 1.0058,
+      "step": 7621
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0462584495544434,
+      "learning_rate": 1.249596600678044e-05,
+      "loss": 0.5621,
+      "step": 7622
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8107529878616333,
+      "learning_rate": 1.2494167145759797e-05,
+      "loss": 0.9807,
+      "step": 7623
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7807387113571167,
+      "learning_rate": 1.2492368198672411e-05,
+      "loss": 0.9759,
+      "step": 7624
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.803370475769043,
+      "learning_rate": 1.2490569165580363e-05,
+      "loss": 1.0265,
+      "step": 7625
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0468337535858154,
+      "learning_rate": 1.2488770046545727e-05,
+      "loss": 0.9558,
+      "step": 7626
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0263333320617676,
+      "learning_rate": 1.248697084163059e-05,
+      "loss": 1.0306,
+      "step": 7627
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8878659009933472,
+      "learning_rate": 1.2485171550897037e-05,
+      "loss": 0.9994,
+      "step": 7628
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.81484055519104,
+      "learning_rate": 1.2483372174407155e-05,
+      "loss": 0.9838,
+      "step": 7629
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8727859258651733,
+      "learning_rate": 1.2481572712223038e-05,
+      "loss": 1.0808,
+      "step": 7630
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.78900945186615,
+      "learning_rate": 1.2479773164406779e-05,
+      "loss": 0.9959,
+      "step": 7631
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9696929454803467,
+      "learning_rate": 1.2477973531020473e-05,
+      "loss": 0.5828,
+      "step": 7632
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8541227579116821,
+      "learning_rate": 1.2476173812126224e-05,
+      "loss": 1.0434,
+      "step": 7633
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.861672043800354,
+      "learning_rate": 1.2474374007786133e-05,
+      "loss": 1.0057,
+      "step": 7634
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8727742433547974,
+      "learning_rate": 1.2472574118062305e-05,
+      "loss": 0.8561,
+      "step": 7635
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8147246837615967,
+      "learning_rate": 1.2470774143016854e-05,
+      "loss": 0.9538,
+      "step": 7636
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.119937539100647,
+      "learning_rate": 1.2468974082711889e-05,
+      "loss": 0.6745,
+      "step": 7637
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9202128648757935,
+      "learning_rate": 1.2467173937209523e-05,
+      "loss": 1.0411,
+      "step": 7638
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7603487968444824,
+      "learning_rate": 1.2465373706571878e-05,
+      "loss": 1.011,
+      "step": 7639
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7497239112854004,
+      "learning_rate": 1.246357339086107e-05,
+      "loss": 1.0062,
+      "step": 7640
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.776878833770752,
+      "learning_rate": 1.246177299013923e-05,
+      "loss": 0.985,
+      "step": 7641
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7239664793014526,
+      "learning_rate": 1.2459972504468479e-05,
+      "loss": 0.9534,
+      "step": 7642
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7702629566192627,
+      "learning_rate": 1.2458171933910946e-05,
+      "loss": 0.932,
+      "step": 7643
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8115798234939575,
+      "learning_rate": 1.2456371278528769e-05,
+      "loss": 1.0158,
+      "step": 7644
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7451653480529785,
+      "learning_rate": 1.2454570538384081e-05,
+      "loss": 1.0677,
+      "step": 7645
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.817245364189148,
+      "learning_rate": 1.245276971353902e-05,
+      "loss": 0.934,
+      "step": 7646
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8699016571044922,
+      "learning_rate": 1.2450968804055728e-05,
+      "loss": 1.0077,
+      "step": 7647
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6603144407272339,
+      "learning_rate": 1.244916780999635e-05,
+      "loss": 0.9548,
+      "step": 7648
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8910303115844727,
+      "learning_rate": 1.2447366731423029e-05,
+      "loss": 1.0763,
+      "step": 7649
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8313363790512085,
+      "learning_rate": 1.2445565568397921e-05,
+      "loss": 0.9859,
+      "step": 7650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9035089015960693,
+      "learning_rate": 1.2443764320983179e-05,
+      "loss": 1.0293,
+      "step": 7651
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6988542079925537,
+      "learning_rate": 1.2441962989240953e-05,
+      "loss": 1.0218,
+      "step": 7652
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8387292623519897,
+      "learning_rate": 1.2440161573233404e-05,
+      "loss": 1.1222,
+      "step": 7653
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8995107412338257,
+      "learning_rate": 1.24383600730227e-05,
+      "loss": 1.0028,
+      "step": 7654
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0339196920394897,
+      "learning_rate": 1.2436558488670997e-05,
+      "loss": 0.6084,
+      "step": 7655
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8126020431518555,
+      "learning_rate": 1.243475682024047e-05,
+      "loss": 0.9405,
+      "step": 7656
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0082863569259644,
+      "learning_rate": 1.2432955067793286e-05,
+      "loss": 0.5183,
+      "step": 7657
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.778639316558838,
+      "learning_rate": 1.2431153231391617e-05,
+      "loss": 0.9502,
+      "step": 7658
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1354997158050537,
+      "learning_rate": 1.2429351311097643e-05,
+      "loss": 0.5436,
+      "step": 7659
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6467963457107544,
+      "learning_rate": 1.242754930697354e-05,
+      "loss": 1.0063,
+      "step": 7660
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8586868047714233,
+      "learning_rate": 1.2425747219081494e-05,
+      "loss": 1.0034,
+      "step": 7661
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7889364957809448,
+      "learning_rate": 1.2423945047483686e-05,
+      "loss": 1.0121,
+      "step": 7662
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.883612871170044,
+      "learning_rate": 1.2422142792242305e-05,
+      "loss": 1.0411,
+      "step": 7663
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7482819557189941,
+      "learning_rate": 1.2420340453419542e-05,
+      "loss": 0.9479,
+      "step": 7664
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.66978120803833,
+      "learning_rate": 1.241853803107759e-05,
+      "loss": 0.93,
+      "step": 7665
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9105827808380127,
+      "learning_rate": 1.2416735525278648e-05,
+      "loss": 1.0029,
+      "step": 7666
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.589983344078064,
+      "learning_rate": 1.2414932936084914e-05,
+      "loss": 0.9611,
+      "step": 7667
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.971234679222107,
+      "learning_rate": 1.2413130263558588e-05,
+      "loss": 0.9823,
+      "step": 7668
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9066078662872314,
+      "learning_rate": 1.2411327507761879e-05,
+      "loss": 0.9584,
+      "step": 7669
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8142876625061035,
+      "learning_rate": 1.2409524668756996e-05,
+      "loss": 0.9721,
+      "step": 7670
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6707004308700562,
+      "learning_rate": 1.2407721746606145e-05,
+      "loss": 0.9261,
+      "step": 7671
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.970712661743164,
+      "learning_rate": 1.240591874137154e-05,
+      "loss": 1.0006,
+      "step": 7672
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7521528005599976,
+      "learning_rate": 1.2404115653115403e-05,
+      "loss": 0.9639,
+      "step": 7673
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7101728916168213,
+      "learning_rate": 1.240231248189995e-05,
+      "loss": 0.9361,
+      "step": 7674
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7056487798690796,
+      "learning_rate": 1.2400509227787406e-05,
+      "loss": 0.9641,
+      "step": 7675
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8381524085998535,
+      "learning_rate": 1.2398705890839988e-05,
+      "loss": 0.9816,
+      "step": 7676
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0057384967803955,
+      "learning_rate": 1.2396902471119934e-05,
+      "loss": 0.9668,
+      "step": 7677
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6697067022323608,
+      "learning_rate": 1.2395098968689471e-05,
+      "loss": 0.8417,
+      "step": 7678
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0697829723358154,
+      "learning_rate": 1.2393295383610832e-05,
+      "loss": 1.1301,
+      "step": 7679
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8120660781860352,
+      "learning_rate": 1.2391491715946256e-05,
+      "loss": 1.0325,
+      "step": 7680
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.709254264831543,
+      "learning_rate": 1.238968796575798e-05,
+      "loss": 0.8657,
+      "step": 7681
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9467140436172485,
+      "learning_rate": 1.238788413310825e-05,
+      "loss": 0.9921,
+      "step": 7682
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.815127968788147,
+      "learning_rate": 1.2386080218059307e-05,
+      "loss": 0.9492,
+      "step": 7683
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.869968295097351,
+      "learning_rate": 1.2384276220673401e-05,
+      "loss": 1.0001,
+      "step": 7684
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6904853582382202,
+      "learning_rate": 1.2382472141012785e-05,
+      "loss": 0.9237,
+      "step": 7685
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.066488265991211,
+      "learning_rate": 1.2380667979139709e-05,
+      "loss": 1.0309,
+      "step": 7686
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.782029390335083,
+      "learning_rate": 1.237886373511643e-05,
+      "loss": 0.9428,
+      "step": 7687
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8218485116958618,
+      "learning_rate": 1.237705940900521e-05,
+      "loss": 0.9441,
+      "step": 7688
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8111777305603027,
+      "learning_rate": 1.2375255000868309e-05,
+      "loss": 0.9998,
+      "step": 7689
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9606832265853882,
+      "learning_rate": 1.2373450510767993e-05,
+      "loss": 0.9992,
+      "step": 7690
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.948456883430481,
+      "learning_rate": 1.2371645938766532e-05,
+      "loss": 0.9433,
+      "step": 7691
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0028083324432373,
+      "learning_rate": 1.236984128492619e-05,
+      "loss": 0.9702,
+      "step": 7692
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.822094440460205,
+      "learning_rate": 1.2368036549309248e-05,
+      "loss": 0.976,
+      "step": 7693
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6483851671218872,
+      "learning_rate": 1.236623173197798e-05,
+      "loss": 1.0515,
+      "step": 7694
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.798875093460083,
+      "learning_rate": 1.2364426832994663e-05,
+      "loss": 0.9366,
+      "step": 7695
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9957859516143799,
+      "learning_rate": 1.2362621852421583e-05,
+      "loss": 1.0341,
+      "step": 7696
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.893798589706421,
+      "learning_rate": 1.2360816790321023e-05,
+      "loss": 0.9339,
+      "step": 7697
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.888737678527832,
+      "learning_rate": 1.2359011646755268e-05,
+      "loss": 0.9602,
+      "step": 7698
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9217383861541748,
+      "learning_rate": 1.2357206421786611e-05,
+      "loss": 1.0519,
+      "step": 7699
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7134835720062256,
+      "learning_rate": 1.2355401115477347e-05,
+      "loss": 0.9607,
+      "step": 7700
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7297958135604858,
+      "learning_rate": 1.2353595727889767e-05,
+      "loss": 0.971,
+      "step": 7701
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7277300357818604,
+      "learning_rate": 1.2351790259086174e-05,
+      "loss": 0.9512,
+      "step": 7702
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9131211042404175,
+      "learning_rate": 1.234998470912887e-05,
+      "loss": 0.9907,
+      "step": 7703
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.914372444152832,
+      "learning_rate": 1.2348179078080155e-05,
+      "loss": 1.0042,
+      "step": 7704
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7643848657608032,
+      "learning_rate": 1.2346373366002342e-05,
+      "loss": 1.0128,
+      "step": 7705
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8489370346069336,
+      "learning_rate": 1.2344567572957738e-05,
+      "loss": 0.9128,
+      "step": 7706
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8097583055496216,
+      "learning_rate": 1.2342761699008656e-05,
+      "loss": 1.0348,
+      "step": 7707
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8438140153884888,
+      "learning_rate": 1.2340955744217413e-05,
+      "loss": 1.0085,
+      "step": 7708
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8783791065216064,
+      "learning_rate": 1.2339149708646323e-05,
+      "loss": 0.9383,
+      "step": 7709
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.824316143989563,
+      "learning_rate": 1.2337343592357713e-05,
+      "loss": 0.8786,
+      "step": 7710
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.642930269241333,
+      "learning_rate": 1.2335537395413906e-05,
+      "loss": 0.9145,
+      "step": 7711
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7573450803756714,
+      "learning_rate": 1.2333731117877228e-05,
+      "loss": 1.0059,
+      "step": 7712
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7118175029754639,
+      "learning_rate": 1.2331924759810008e-05,
+      "loss": 0.9653,
+      "step": 7713
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0859718322753906,
+      "learning_rate": 1.2330118321274576e-05,
+      "loss": 0.9895,
+      "step": 7714
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8060108423233032,
+      "learning_rate": 1.232831180233327e-05,
+      "loss": 0.9569,
+      "step": 7715
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9823864698410034,
+      "learning_rate": 1.232650520304843e-05,
+      "loss": 0.9702,
+      "step": 7716
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8027502298355103,
+      "learning_rate": 1.2324698523482393e-05,
+      "loss": 0.9291,
+      "step": 7717
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7098044157028198,
+      "learning_rate": 1.2322891763697505e-05,
+      "loss": 0.9706,
+      "step": 7718
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.840698003768921,
+      "learning_rate": 1.2321084923756108e-05,
+      "loss": 1.0376,
+      "step": 7719
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.799296498298645,
+      "learning_rate": 1.2319278003720554e-05,
+      "loss": 0.9955,
+      "step": 7720
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6471272706985474,
+      "learning_rate": 1.2317471003653196e-05,
+      "loss": 0.9732,
+      "step": 7721
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9313384294509888,
+      "learning_rate": 1.2315663923616388e-05,
+      "loss": 1.0424,
+      "step": 7722
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9198538064956665,
+      "learning_rate": 1.2313856763672486e-05,
+      "loss": 0.9698,
+      "step": 7723
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7545133829116821,
+      "learning_rate": 1.2312049523883851e-05,
+      "loss": 1.0192,
+      "step": 7724
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8588281869888306,
+      "learning_rate": 1.2310242204312845e-05,
+      "loss": 1.0653,
+      "step": 7725
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8679583072662354,
+      "learning_rate": 1.2308434805021836e-05,
+      "loss": 0.9915,
+      "step": 7726
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.875706672668457,
+      "learning_rate": 1.2306627326073189e-05,
+      "loss": 0.9695,
+      "step": 7727
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8187111616134644,
+      "learning_rate": 1.2304819767529274e-05,
+      "loss": 0.9801,
+      "step": 7728
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9930998086929321,
+      "learning_rate": 1.2303012129452469e-05,
+      "loss": 0.9449,
+      "step": 7729
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8013794422149658,
+      "learning_rate": 1.230120441190515e-05,
+      "loss": 1.008,
+      "step": 7730
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8797240257263184,
+      "learning_rate": 1.2299396614949691e-05,
+      "loss": 0.9384,
+      "step": 7731
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7570873498916626,
+      "learning_rate": 1.2297588738648481e-05,
+      "loss": 0.983,
+      "step": 7732
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6738688945770264,
+      "learning_rate": 1.22957807830639e-05,
+      "loss": 0.9441,
+      "step": 7733
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7166593074798584,
+      "learning_rate": 1.2293972748258338e-05,
+      "loss": 1.0917,
+      "step": 7734
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.752254843711853,
+      "learning_rate": 1.2292164634294184e-05,
+      "loss": 0.9098,
+      "step": 7735
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5763689279556274,
+      "learning_rate": 1.2290356441233833e-05,
+      "loss": 0.946,
+      "step": 7736
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8233758211135864,
+      "learning_rate": 1.2288548169139676e-05,
+      "loss": 0.9967,
+      "step": 7737
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7821087837219238,
+      "learning_rate": 1.2286739818074116e-05,
+      "loss": 0.8766,
+      "step": 7738
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9034427404403687,
+      "learning_rate": 1.2284931388099554e-05,
+      "loss": 1.0466,
+      "step": 7739
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8669012784957886,
+      "learning_rate": 1.2283122879278393e-05,
+      "loss": 1.0304,
+      "step": 7740
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7810660600662231,
+      "learning_rate": 1.2281314291673036e-05,
+      "loss": 0.9702,
+      "step": 7741
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9883533716201782,
+      "learning_rate": 1.2279505625345896e-05,
+      "loss": 0.8985,
+      "step": 7742
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9422756433486938,
+      "learning_rate": 1.2277696880359384e-05,
+      "loss": 1.0119,
+      "step": 7743
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8473113775253296,
+      "learning_rate": 1.2275888056775918e-05,
+      "loss": 0.9856,
+      "step": 7744
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7119449377059937,
+      "learning_rate": 1.227407915465791e-05,
+      "loss": 0.9976,
+      "step": 7745
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1289796829223633,
+      "learning_rate": 1.2272270174067782e-05,
+      "loss": 1.008,
+      "step": 7746
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9426058530807495,
+      "learning_rate": 1.227046111506796e-05,
+      "loss": 0.9196,
+      "step": 7747
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9121708869934082,
+      "learning_rate": 1.2268651977720867e-05,
+      "loss": 0.9814,
+      "step": 7748
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7229068279266357,
+      "learning_rate": 1.2266842762088932e-05,
+      "loss": 0.9712,
+      "step": 7749
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9257116317749023,
+      "learning_rate": 1.2265033468234584e-05,
+      "loss": 0.9506,
+      "step": 7750
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7360285520553589,
+      "learning_rate": 1.2263224096220258e-05,
+      "loss": 1.006,
+      "step": 7751
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.664992094039917,
+      "learning_rate": 1.2261414646108391e-05,
+      "loss": 0.976,
+      "step": 7752
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.885969638824463,
+      "learning_rate": 1.2259605117961422e-05,
+      "loss": 1.0596,
+      "step": 7753
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6871249675750732,
+      "learning_rate": 1.2257795511841792e-05,
+      "loss": 1.0256,
+      "step": 7754
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1521257162094116,
+      "learning_rate": 1.2255985827811947e-05,
+      "loss": 0.5943,
+      "step": 7755
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6556180715560913,
+      "learning_rate": 1.2254176065934332e-05,
+      "loss": 0.9489,
+      "step": 7756
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7253577709197998,
+      "learning_rate": 1.2252366226271398e-05,
+      "loss": 1.0308,
+      "step": 7757
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6542346477508545,
+      "learning_rate": 1.2250556308885595e-05,
+      "loss": 1.0086,
+      "step": 7758
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.724663257598877,
+      "learning_rate": 1.224874631383938e-05,
+      "loss": 0.9801,
+      "step": 7759
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1577228307724,
+      "learning_rate": 1.2246936241195215e-05,
+      "loss": 0.6109,
+      "step": 7760
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.045353651046753,
+      "learning_rate": 1.2245126091015556e-05,
+      "loss": 1.0538,
+      "step": 7761
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7128798961639404,
+      "learning_rate": 1.2243315863362866e-05,
+      "loss": 0.9856,
+      "step": 7762
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0578300952911377,
+      "learning_rate": 1.2241505558299614e-05,
+      "loss": 0.9877,
+      "step": 7763
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7511426210403442,
+      "learning_rate": 1.2239695175888264e-05,
+      "loss": 1.0442,
+      "step": 7764
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7828105688095093,
+      "learning_rate": 1.223788471619129e-05,
+      "loss": 0.9138,
+      "step": 7765
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2071444988250732,
+      "learning_rate": 1.223607417927117e-05,
+      "loss": 0.6095,
+      "step": 7766
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0276472568511963,
+      "learning_rate": 1.2234263565190372e-05,
+      "loss": 0.9679,
+      "step": 7767
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.854109287261963,
+      "learning_rate": 1.2232452874011381e-05,
+      "loss": 1.0191,
+      "step": 7768
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8650341033935547,
+      "learning_rate": 1.2230642105796674e-05,
+      "loss": 1.0227,
+      "step": 7769
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9766743183135986,
+      "learning_rate": 1.2228831260608745e-05,
+      "loss": 0.5656,
+      "step": 7770
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8758357763290405,
+      "learning_rate": 1.222702033851007e-05,
+      "loss": 0.9577,
+      "step": 7771
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7652212381362915,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 0.9378,
+      "step": 7772
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9341726303100586,
+      "learning_rate": 1.2223398263830463e-05,
+      "loss": 0.9483,
+      "step": 7773
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6811860799789429,
+      "learning_rate": 1.2221587111374519e-05,
+      "loss": 0.9361,
+      "step": 7774
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9038540124893188,
+      "learning_rate": 1.2219775882257804e-05,
+      "loss": 0.9776,
+      "step": 7775
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8591549396514893,
+      "learning_rate": 1.2217964576542829e-05,
+      "loss": 0.983,
+      "step": 7776
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8306639194488525,
+      "learning_rate": 1.221615319429209e-05,
+      "loss": 1.0137,
+      "step": 7777
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8915072679519653,
+      "learning_rate": 1.2214341735568099e-05,
+      "loss": 1.0339,
+      "step": 7778
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.753797173500061,
+      "learning_rate": 1.2212530200433355e-05,
+      "loss": 0.9676,
+      "step": 7779
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.785038948059082,
+      "learning_rate": 1.2210718588950376e-05,
+      "loss": 0.9451,
+      "step": 7780
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7709153890609741,
+      "learning_rate": 1.2208906901181675e-05,
+      "loss": 0.9271,
+      "step": 7781
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7793200016021729,
+      "learning_rate": 1.2207095137189766e-05,
+      "loss": 0.8301,
+      "step": 7782
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.197963237762451,
+      "learning_rate": 1.2205283297037172e-05,
+      "loss": 1.0018,
+      "step": 7783
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7792103290557861,
+      "learning_rate": 1.2203471380786407e-05,
+      "loss": 1.0014,
+      "step": 7784
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6629672050476074,
+      "learning_rate": 1.22016593885e-05,
+      "loss": 0.9557,
+      "step": 7785
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7825459241867065,
+      "learning_rate": 1.219984732024048e-05,
+      "loss": 1.009,
+      "step": 7786
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8265187740325928,
+      "learning_rate": 1.2198035176070375e-05,
+      "loss": 1.0344,
+      "step": 7787
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7456082105636597,
+      "learning_rate": 1.2196222956052215e-05,
+      "loss": 0.9857,
+      "step": 7788
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.899399757385254,
+      "learning_rate": 1.2194410660248535e-05,
+      "loss": 1.0053,
+      "step": 7789
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8614953756332397,
+      "learning_rate": 1.2192598288721873e-05,
+      "loss": 0.9883,
+      "step": 7790
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7467591762542725,
+      "learning_rate": 1.219078584153477e-05,
+      "loss": 1.0712,
+      "step": 7791
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6163705587387085,
+      "learning_rate": 1.2188973318749766e-05,
+      "loss": 0.9939,
+      "step": 7792
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9156336784362793,
+      "learning_rate": 1.2187160720429407e-05,
+      "loss": 0.9976,
+      "step": 7793
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8359031677246094,
+      "learning_rate": 1.2185348046636243e-05,
+      "loss": 1.0469,
+      "step": 7794
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8725281953811646,
+      "learning_rate": 1.2183535297432821e-05,
+      "loss": 0.9965,
+      "step": 7795
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.815598726272583,
+      "learning_rate": 1.2181722472881697e-05,
+      "loss": 1.0652,
+      "step": 7796
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.960411787033081,
+      "learning_rate": 1.217990957304542e-05,
+      "loss": 0.9459,
+      "step": 7797
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.717895746231079,
+      "learning_rate": 1.2178096597986557e-05,
+      "loss": 0.958,
+      "step": 7798
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7529269456863403,
+      "learning_rate": 1.2176283547767665e-05,
+      "loss": 0.9301,
+      "step": 7799
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8150732517242432,
+      "learning_rate": 1.2174470422451306e-05,
+      "loss": 0.9927,
+      "step": 7800
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0498924255371094,
+      "learning_rate": 1.2172657222100047e-05,
+      "loss": 0.6094,
+      "step": 7801
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8583680391311646,
+      "learning_rate": 1.2170843946776457e-05,
+      "loss": 0.987,
+      "step": 7802
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7384819984436035,
+      "learning_rate": 1.2169030596543106e-05,
+      "loss": 0.9862,
+      "step": 7803
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7967960834503174,
+      "learning_rate": 1.2167217171462566e-05,
+      "loss": 1.0252,
+      "step": 7804
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8554235696792603,
+      "learning_rate": 1.2165403671597418e-05,
+      "loss": 0.9671,
+      "step": 7805
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.789666771888733,
+      "learning_rate": 1.2163590097010239e-05,
+      "loss": 0.966,
+      "step": 7806
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9626660346984863,
+      "learning_rate": 1.2161776447763607e-05,
+      "loss": 0.9962,
+      "step": 7807
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7834597826004028,
+      "learning_rate": 1.2159962723920107e-05,
+      "loss": 0.983,
+      "step": 7808
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6984755992889404,
+      "learning_rate": 1.2158148925542328e-05,
+      "loss": 0.8972,
+      "step": 7809
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6644375324249268,
+      "learning_rate": 1.215633505269286e-05,
+      "loss": 1.0004,
+      "step": 7810
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7690021991729736,
+      "learning_rate": 1.2154521105434292e-05,
+      "loss": 0.9507,
+      "step": 7811
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0975890159606934,
+      "learning_rate": 1.2152707083829218e-05,
+      "loss": 0.9303,
+      "step": 7812
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6571881771087646,
+      "learning_rate": 1.2150892987940236e-05,
+      "loss": 0.8909,
+      "step": 7813
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0248130559921265,
+      "learning_rate": 1.2149078817829947e-05,
+      "loss": 0.5726,
+      "step": 7814
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7024688720703125,
+      "learning_rate": 1.214726457356095e-05,
+      "loss": 0.9333,
+      "step": 7815
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6930779218673706,
+      "learning_rate": 1.2145450255195852e-05,
+      "loss": 0.933,
+      "step": 7816
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7007426023483276,
+      "learning_rate": 1.2143635862797258e-05,
+      "loss": 0.95,
+      "step": 7817
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0356431007385254,
+      "learning_rate": 1.214182139642778e-05,
+      "loss": 0.9791,
+      "step": 7818
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6855437755584717,
+      "learning_rate": 1.2140006856150026e-05,
+      "loss": 0.9086,
+      "step": 7819
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6805263757705688,
+      "learning_rate": 1.2138192242026613e-05,
+      "loss": 0.9183,
+      "step": 7820
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8973623514175415,
+      "learning_rate": 1.213637755412016e-05,
+      "loss": 0.9967,
+      "step": 7821
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6158584356307983,
+      "learning_rate": 1.2134562792493285e-05,
+      "loss": 1.0595,
+      "step": 7822
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.979712724685669,
+      "learning_rate": 1.2132747957208613e-05,
+      "loss": 0.9056,
+      "step": 7823
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.799809455871582,
+      "learning_rate": 1.2130933048328762e-05,
+      "loss": 0.9442,
+      "step": 7824
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.094165563583374,
+      "learning_rate": 1.2129118065916366e-05,
+      "loss": 1.0416,
+      "step": 7825
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7371410131454468,
+      "learning_rate": 1.2127303010034052e-05,
+      "loss": 0.9937,
+      "step": 7826
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6301283836364746,
+      "learning_rate": 1.2125487880744456e-05,
+      "loss": 0.9775,
+      "step": 7827
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8274223804473877,
+      "learning_rate": 1.212367267811021e-05,
+      "loss": 0.9874,
+      "step": 7828
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8016401529312134,
+      "learning_rate": 1.2121857402193951e-05,
+      "loss": 0.9359,
+      "step": 7829
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8613020181655884,
+      "learning_rate": 1.212004205305832e-05,
+      "loss": 0.9893,
+      "step": 7830
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8588120937347412,
+      "learning_rate": 1.211822663076596e-05,
+      "loss": 0.9684,
+      "step": 7831
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.768686056137085,
+      "learning_rate": 1.2116411135379517e-05,
+      "loss": 0.9467,
+      "step": 7832
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6390390396118164,
+      "learning_rate": 1.211459556696164e-05,
+      "loss": 0.9352,
+      "step": 7833
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9199225902557373,
+      "learning_rate": 1.2112779925574973e-05,
+      "loss": 1.0228,
+      "step": 7834
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7528667449951172,
+      "learning_rate": 1.2110964211282175e-05,
+      "loss": 1.0292,
+      "step": 7835
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9582585096359253,
+      "learning_rate": 1.2109148424145897e-05,
+      "loss": 0.9846,
+      "step": 7836
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.739980936050415,
+      "learning_rate": 1.2107332564228798e-05,
+      "loss": 0.997,
+      "step": 7837
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6895437240600586,
+      "learning_rate": 1.2105516631593539e-05,
+      "loss": 0.992,
+      "step": 7838
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7470614910125732,
+      "learning_rate": 1.2103700626302784e-05,
+      "loss": 0.9675,
+      "step": 7839
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7087090015411377,
+      "learning_rate": 1.2101884548419196e-05,
+      "loss": 0.9401,
+      "step": 7840
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5694940090179443,
+      "learning_rate": 1.2100068398005443e-05,
+      "loss": 0.9315,
+      "step": 7841
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8077785968780518,
+      "learning_rate": 1.2098252175124197e-05,
+      "loss": 0.9982,
+      "step": 7842
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1268465518951416,
+      "learning_rate": 1.2096435879838129e-05,
+      "loss": 0.5888,
+      "step": 7843
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7400039434432983,
+      "learning_rate": 1.2094619512209915e-05,
+      "loss": 0.9264,
+      "step": 7844
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7248669862747192,
+      "learning_rate": 1.2092803072302233e-05,
+      "loss": 0.9711,
+      "step": 7845
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.725252389907837,
+      "learning_rate": 1.2090986560177764e-05,
+      "loss": 0.9837,
+      "step": 7846
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8736921548843384,
+      "learning_rate": 1.208916997589919e-05,
+      "loss": 0.9879,
+      "step": 7847
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5278563499450684,
+      "learning_rate": 1.2087353319529193e-05,
+      "loss": 0.9096,
+      "step": 7848
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.782125473022461,
+      "learning_rate": 1.2085536591130467e-05,
+      "loss": 0.929,
+      "step": 7849
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8642730712890625,
+      "learning_rate": 1.2083719790765698e-05,
+      "loss": 1.0413,
+      "step": 7850
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.118824005126953,
+      "learning_rate": 1.2081902918497577e-05,
+      "loss": 0.9199,
+      "step": 7851
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9222227334976196,
+      "learning_rate": 1.2080085974388802e-05,
+      "loss": 1.0483,
+      "step": 7852
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1640421152114868,
+      "learning_rate": 1.2078268958502073e-05,
+      "loss": 0.6379,
+      "step": 7853
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7426316738128662,
+      "learning_rate": 1.2076451870900087e-05,
+      "loss": 1.0212,
+      "step": 7854
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9776769876480103,
+      "learning_rate": 1.2074634711645548e-05,
+      "loss": 1.0042,
+      "step": 7855
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3835716247558594,
+      "learning_rate": 1.207281748080116e-05,
+      "loss": 0.9722,
+      "step": 7856
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0638364553451538,
+      "learning_rate": 1.207100017842963e-05,
+      "loss": 0.5733,
+      "step": 7857
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.876054286956787,
+      "learning_rate": 1.2069182804593671e-05,
+      "loss": 0.9346,
+      "step": 7858
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6998441219329834,
+      "learning_rate": 1.2067365359355991e-05,
+      "loss": 0.9379,
+      "step": 7859
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9469122886657715,
+      "learning_rate": 1.206554784277931e-05,
+      "loss": 0.9101,
+      "step": 7860
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.681156873703003,
+      "learning_rate": 1.206373025492634e-05,
+      "loss": 1.0031,
+      "step": 7861
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9619028568267822,
+      "learning_rate": 1.2061912595859806e-05,
+      "loss": 1.0183,
+      "step": 7862
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8572429418563843,
+      "learning_rate": 1.2060094865642427e-05,
+      "loss": 0.9742,
+      "step": 7863
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.036673665046692,
+      "learning_rate": 1.2058277064336928e-05,
+      "loss": 0.6323,
+      "step": 7864
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.765555739402771,
+      "learning_rate": 1.2056459192006038e-05,
+      "loss": 0.9486,
+      "step": 7865
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9937422275543213,
+      "learning_rate": 1.2054641248712487e-05,
+      "loss": 0.6318,
+      "step": 7866
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9794669151306152,
+      "learning_rate": 1.2052823234519004e-05,
+      "loss": 0.5837,
+      "step": 7867
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6570795774459839,
+      "learning_rate": 1.2051005149488326e-05,
+      "loss": 0.8826,
+      "step": 7868
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8475841283798218,
+      "learning_rate": 1.204918699368319e-05,
+      "loss": 1.0037,
+      "step": 7869
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0002281665802,
+      "learning_rate": 1.2047368767166334e-05,
+      "loss": 0.9915,
+      "step": 7870
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.937207818031311,
+      "learning_rate": 1.2045550470000502e-05,
+      "loss": 1.0088,
+      "step": 7871
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6322695016860962,
+      "learning_rate": 1.2043732102248437e-05,
+      "loss": 0.9787,
+      "step": 7872
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7556989192962646,
+      "learning_rate": 1.2041913663972886e-05,
+      "loss": 1.0925,
+      "step": 7873
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6921501159667969,
+      "learning_rate": 1.2040095155236597e-05,
+      "loss": 0.9283,
+      "step": 7874
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.780142068862915,
+      "learning_rate": 1.2038276576102324e-05,
+      "loss": 0.8611,
+      "step": 7875
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6809784173965454,
+      "learning_rate": 1.203645792663282e-05,
+      "loss": 0.9607,
+      "step": 7876
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7665553092956543,
+      "learning_rate": 1.2034639206890843e-05,
+      "loss": 0.9632,
+      "step": 7877
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9764895439147949,
+      "learning_rate": 1.2032820416939148e-05,
+      "loss": 0.5531,
+      "step": 7878
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.643336296081543,
+      "learning_rate": 1.20310015568405e-05,
+      "loss": 0.9801,
+      "step": 7879
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.797683835029602,
+      "learning_rate": 1.2029182626657662e-05,
+      "loss": 1.0349,
+      "step": 7880
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8186728954315186,
+      "learning_rate": 1.20273636264534e-05,
+      "loss": 0.9759,
+      "step": 7881
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.934433937072754,
+      "learning_rate": 1.2025544556290483e-05,
+      "loss": 1.0516,
+      "step": 7882
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9675723314285278,
+      "learning_rate": 1.202372541623168e-05,
+      "loss": 1.0712,
+      "step": 7883
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7322179079055786,
+      "learning_rate": 1.2021906206339766e-05,
+      "loss": 0.9636,
+      "step": 7884
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.852795958518982,
+      "learning_rate": 1.202008692667752e-05,
+      "loss": 1.0067,
+      "step": 7885
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8027853965759277,
+      "learning_rate": 1.2018267577307714e-05,
+      "loss": 0.9376,
+      "step": 7886
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6885671615600586,
+      "learning_rate": 1.2016448158293133e-05,
+      "loss": 0.9805,
+      "step": 7887
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6978824138641357,
+      "learning_rate": 1.2014628669696557e-05,
+      "loss": 0.9744,
+      "step": 7888
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6917245388031006,
+      "learning_rate": 1.2012809111580774e-05,
+      "loss": 0.9792,
+      "step": 7889
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9685620069503784,
+      "learning_rate": 1.201098948400857e-05,
+      "loss": 0.9795,
+      "step": 7890
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0712857246398926,
+      "learning_rate": 1.2009169787042739e-05,
+      "loss": 0.9485,
+      "step": 7891
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0722960233688354,
+      "learning_rate": 1.2007350020746069e-05,
+      "loss": 0.5886,
+      "step": 7892
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6066468954086304,
+      "learning_rate": 1.2005530185181358e-05,
+      "loss": 1.0109,
+      "step": 7893
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.867124080657959,
+      "learning_rate": 1.2003710280411403e-05,
+      "loss": 0.9757,
+      "step": 7894
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.778659462928772,
+      "learning_rate": 1.2001890306499003e-05,
+      "loss": 0.9148,
+      "step": 7895
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7782255411148071,
+      "learning_rate": 1.200007026350696e-05,
+      "loss": 0.907,
+      "step": 7896
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.765851616859436,
+      "learning_rate": 1.1998250151498078e-05,
+      "loss": 0.9826,
+      "step": 7897
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8597368001937866,
+      "learning_rate": 1.1996429970535169e-05,
+      "loss": 0.9972,
+      "step": 7898
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.056516408920288,
+      "learning_rate": 1.1994609720681036e-05,
+      "loss": 0.6222,
+      "step": 7899
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5763027667999268,
+      "learning_rate": 1.1992789401998492e-05,
+      "loss": 0.9862,
+      "step": 7900
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8898308277130127,
+      "learning_rate": 1.1990969014550355e-05,
+      "loss": 1.0403,
+      "step": 7901
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6723557710647583,
+      "learning_rate": 1.1989148558399436e-05,
+      "loss": 0.9295,
+      "step": 7902
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9528573751449585,
+      "learning_rate": 1.198732803360856e-05,
+      "loss": 0.9839,
+      "step": 7903
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.119760274887085,
+      "learning_rate": 1.1985507440240543e-05,
+      "loss": 0.5821,
+      "step": 7904
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7641072273254395,
+      "learning_rate": 1.198368677835821e-05,
+      "loss": 0.9693,
+      "step": 7905
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8989454507827759,
+      "learning_rate": 1.1981866048024388e-05,
+      "loss": 0.9342,
+      "step": 7906
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7934963703155518,
+      "learning_rate": 1.1980045249301904e-05,
+      "loss": 1.0204,
+      "step": 7907
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7942888736724854,
+      "learning_rate": 1.197822438225359e-05,
+      "loss": 0.9444,
+      "step": 7908
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.915575623512268,
+      "learning_rate": 1.197640344694228e-05,
+      "loss": 1.032,
+      "step": 7909
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8801555633544922,
+      "learning_rate": 1.1974582443430807e-05,
+      "loss": 0.9634,
+      "step": 7910
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.098273515701294,
+      "learning_rate": 1.1972761371782008e-05,
+      "loss": 1.0093,
+      "step": 7911
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.729427456855774,
+      "learning_rate": 1.1970940232058727e-05,
+      "loss": 1.0106,
+      "step": 7912
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6927303075790405,
+      "learning_rate": 1.1969119024323805e-05,
+      "loss": 0.9869,
+      "step": 7913
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7167794704437256,
+      "learning_rate": 1.1967297748640085e-05,
+      "loss": 0.9666,
+      "step": 7914
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.813736081123352,
+      "learning_rate": 1.1965476405070415e-05,
+      "loss": 1.0,
+      "step": 7915
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1631368398666382,
+      "learning_rate": 1.1963654993677645e-05,
+      "loss": 0.6381,
+      "step": 7916
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1261005401611328,
+      "learning_rate": 1.1961833514524624e-05,
+      "loss": 0.605,
+      "step": 7917
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8264689445495605,
+      "learning_rate": 1.196001196767421e-05,
+      "loss": 0.9705,
+      "step": 7918
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.828500509262085,
+      "learning_rate": 1.1958190353189259e-05,
+      "loss": 0.9949,
+      "step": 7919
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8491536378860474,
+      "learning_rate": 1.1956368671132628e-05,
+      "loss": 0.9823,
+      "step": 7920
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.795048475265503,
+      "learning_rate": 1.1954546921567179e-05,
+      "loss": 1.0092,
+      "step": 7921
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0442445278167725,
+      "learning_rate": 1.1952725104555775e-05,
+      "loss": 1.0333,
+      "step": 7922
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9495928287506104,
+      "learning_rate": 1.1950903220161286e-05,
+      "loss": 1.0267,
+      "step": 7923
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8137692213058472,
+      "learning_rate": 1.1949081268446573e-05,
+      "loss": 1.0312,
+      "step": 7924
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.158942461013794,
+      "learning_rate": 1.194725924947451e-05,
+      "loss": 0.5532,
+      "step": 7925
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.695214033126831,
+      "learning_rate": 1.1945437163307971e-05,
+      "loss": 0.8582,
+      "step": 7926
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.821165919303894,
+      "learning_rate": 1.1943615010009828e-05,
+      "loss": 0.9679,
+      "step": 7927
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7312781810760498,
+      "learning_rate": 1.1941792789642963e-05,
+      "loss": 0.9957,
+      "step": 7928
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.896626591682434,
+      "learning_rate": 1.1939970502270253e-05,
+      "loss": 1.0303,
+      "step": 7929
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9577184915542603,
+      "learning_rate": 1.1938148147954575e-05,
+      "loss": 0.9517,
+      "step": 7930
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6960945129394531,
+      "learning_rate": 1.1936325726758822e-05,
+      "loss": 0.9566,
+      "step": 7931
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7003276348114014,
+      "learning_rate": 1.1934503238745878e-05,
+      "loss": 0.9321,
+      "step": 7932
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6604646444320679,
+      "learning_rate": 1.1932680683978631e-05,
+      "loss": 1.0728,
+      "step": 7933
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7263479232788086,
+      "learning_rate": 1.193085806251997e-05,
+      "loss": 0.9398,
+      "step": 7934
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.780462622642517,
+      "learning_rate": 1.1929035374432794e-05,
+      "loss": 1.0487,
+      "step": 7935
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9222217798233032,
+      "learning_rate": 1.1927212619779994e-05,
+      "loss": 0.9582,
+      "step": 7936
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9137370586395264,
+      "learning_rate": 1.192538979862447e-05,
+      "loss": 0.9669,
+      "step": 7937
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7538650035858154,
+      "learning_rate": 1.1923566911029123e-05,
+      "loss": 0.9555,
+      "step": 7938
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.064792275428772,
+      "learning_rate": 1.1921743957056854e-05,
+      "loss": 0.5748,
+      "step": 7939
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6400911808013916,
+      "learning_rate": 1.1919920936770568e-05,
+      "loss": 0.9217,
+      "step": 7940
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7615841627120972,
+      "learning_rate": 1.1918097850233177e-05,
+      "loss": 0.9309,
+      "step": 7941
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0309568643569946,
+      "learning_rate": 1.1916274697507583e-05,
+      "loss": 0.6075,
+      "step": 7942
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8571319580078125,
+      "learning_rate": 1.1914451478656708e-05,
+      "loss": 0.9384,
+      "step": 7943
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8746356964111328,
+      "learning_rate": 1.1912628193743454e-05,
+      "loss": 0.991,
+      "step": 7944
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9811104536056519,
+      "learning_rate": 1.1910804842830746e-05,
+      "loss": 0.9576,
+      "step": 7945
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.821722149848938,
+      "learning_rate": 1.1908981425981502e-05,
+      "loss": 0.9125,
+      "step": 7946
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9698566794395447,
+      "learning_rate": 1.190715794325864e-05,
+      "loss": 0.5708,
+      "step": 7947
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.666179895401001,
+      "learning_rate": 1.1905334394725086e-05,
+      "loss": 1.0343,
+      "step": 7948
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7561556100845337,
+      "learning_rate": 1.1903510780443765e-05,
+      "loss": 1.0325,
+      "step": 7949
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8011491298675537,
+      "learning_rate": 1.1901687100477604e-05,
+      "loss": 1.0059,
+      "step": 7950
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.019629955291748,
+      "learning_rate": 1.1899863354889532e-05,
+      "loss": 1.0272,
+      "step": 7951
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8673694133758545,
+      "learning_rate": 1.1898039543742484e-05,
+      "loss": 0.9911,
+      "step": 7952
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8260685205459595,
+      "learning_rate": 1.189621566709939e-05,
+      "loss": 0.9564,
+      "step": 7953
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7339130640029907,
+      "learning_rate": 1.1894391725023194e-05,
+      "loss": 0.9645,
+      "step": 7954
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8669044971466064,
+      "learning_rate": 1.1892567717576831e-05,
+      "loss": 1.0324,
+      "step": 7955
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8711845874786377,
+      "learning_rate": 1.1890743644823242e-05,
+      "loss": 1.0058,
+      "step": 7956
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1033128499984741,
+      "learning_rate": 1.188891950682537e-05,
+      "loss": 0.5578,
+      "step": 7957
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7915196418762207,
+      "learning_rate": 1.1887095303646161e-05,
+      "loss": 0.8906,
+      "step": 7958
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.736484408378601,
+      "learning_rate": 1.1885271035348569e-05,
+      "loss": 1.0089,
+      "step": 7959
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8839147090911865,
+      "learning_rate": 1.1883446701995536e-05,
+      "loss": 0.9998,
+      "step": 7960
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.742846131324768,
+      "learning_rate": 1.1881622303650022e-05,
+      "loss": 0.9501,
+      "step": 7961
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9567749500274658,
+      "learning_rate": 1.1879797840374976e-05,
+      "loss": 0.9694,
+      "step": 7962
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.821663498878479,
+      "learning_rate": 1.1877973312233358e-05,
+      "loss": 1.0121,
+      "step": 7963
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.778533697128296,
+      "learning_rate": 1.1876148719288128e-05,
+      "loss": 1.0397,
+      "step": 7964
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8641010522842407,
+      "learning_rate": 1.1874324061602245e-05,
+      "loss": 0.8945,
+      "step": 7965
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8051846027374268,
+      "learning_rate": 1.1872499339238677e-05,
+      "loss": 1.0041,
+      "step": 7966
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6744557619094849,
+      "learning_rate": 1.1870674552260384e-05,
+      "loss": 0.9544,
+      "step": 7967
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8837062120437622,
+      "learning_rate": 1.186884970073034e-05,
+      "loss": 1.0154,
+      "step": 7968
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.777496576309204,
+      "learning_rate": 1.186702478471151e-05,
+      "loss": 1.0355,
+      "step": 7969
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7890945672988892,
+      "learning_rate": 1.1865199804266872e-05,
+      "loss": 0.9163,
+      "step": 7970
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7591489553451538,
+      "learning_rate": 1.1863374759459401e-05,
+      "loss": 0.9771,
+      "step": 7971
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0123162269592285,
+      "learning_rate": 1.1861549650352069e-05,
+      "loss": 0.957,
+      "step": 7972
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7793066501617432,
+      "learning_rate": 1.185972447700786e-05,
+      "loss": 0.9913,
+      "step": 7973
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8957061767578125,
+      "learning_rate": 1.1857899239489753e-05,
+      "loss": 0.9469,
+      "step": 7974
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1403964757919312,
+      "learning_rate": 1.1856073937860735e-05,
+      "loss": 0.6193,
+      "step": 7975
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6930732727050781,
+      "learning_rate": 1.1854248572183789e-05,
+      "loss": 1.0251,
+      "step": 7976
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6233232021331787,
+      "learning_rate": 1.1852423142521904e-05,
+      "loss": 0.8828,
+      "step": 7977
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8118256330490112,
+      "learning_rate": 1.1850597648938073e-05,
+      "loss": 0.9743,
+      "step": 7978
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.783571481704712,
+      "learning_rate": 1.1848772091495287e-05,
+      "loss": 0.9379,
+      "step": 7979
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7771198749542236,
+      "learning_rate": 1.1846946470256537e-05,
+      "loss": 0.9804,
+      "step": 7980
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9490060806274414,
+      "learning_rate": 1.1845120785284827e-05,
+      "loss": 0.9742,
+      "step": 7981
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6073194742202759,
+      "learning_rate": 1.184329503664315e-05,
+      "loss": 1.0533,
+      "step": 7982
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8105151653289795,
+      "learning_rate": 1.184146922439451e-05,
+      "loss": 0.9468,
+      "step": 7983
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0121195316314697,
+      "learning_rate": 1.1839643348601912e-05,
+      "loss": 0.5665,
+      "step": 7984
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.763900876045227,
+      "learning_rate": 1.1837817409328362e-05,
+      "loss": 1.0206,
+      "step": 7985
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8150074481964111,
+      "learning_rate": 1.1835991406636866e-05,
+      "loss": 0.8469,
+      "step": 7986
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.016606330871582,
+      "learning_rate": 1.1834165340590436e-05,
+      "loss": 0.9691,
+      "step": 7987
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6387733221054077,
+      "learning_rate": 1.1832339211252084e-05,
+      "loss": 0.9324,
+      "step": 7988
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8975329399108887,
+      "learning_rate": 1.1830513018684824e-05,
+      "loss": 0.9369,
+      "step": 7989
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7825971841812134,
+      "learning_rate": 1.1828686762951674e-05,
+      "loss": 0.9607,
+      "step": 7990
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6111706495285034,
+      "learning_rate": 1.1826860444115648e-05,
+      "loss": 0.9725,
+      "step": 7991
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7063044309616089,
+      "learning_rate": 1.1825034062239775e-05,
+      "loss": 1.0124,
+      "step": 7992
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6753889322280884,
+      "learning_rate": 1.1823207617387073e-05,
+      "loss": 0.9982,
+      "step": 7993
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6649279594421387,
+      "learning_rate": 1.1821381109620571e-05,
+      "loss": 1.0296,
+      "step": 7994
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8102540969848633,
+      "learning_rate": 1.1819554539003292e-05,
+      "loss": 0.927,
+      "step": 7995
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8892713785171509,
+      "learning_rate": 1.1817727905598268e-05,
+      "loss": 1.0201,
+      "step": 7996
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9476550817489624,
+      "learning_rate": 1.1815901209468535e-05,
+      "loss": 0.9812,
+      "step": 7997
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8439606428146362,
+      "learning_rate": 1.181407445067712e-05,
+      "loss": 1.0306,
+      "step": 7998
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.011625051498413,
+      "learning_rate": 1.1812247629287065e-05,
+      "loss": 0.9561,
+      "step": 7999
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7481447458267212,
+      "learning_rate": 1.1810420745361408e-05,
+      "loss": 1.0133,
+      "step": 8000
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8607566356658936,
+      "learning_rate": 1.1808593798963185e-05,
+      "loss": 0.9616,
+      "step": 8001
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7358875274658203,
+      "learning_rate": 1.1806766790155446e-05,
+      "loss": 0.9752,
+      "step": 8002
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7777528762817383,
+      "learning_rate": 1.180493971900123e-05,
+      "loss": 0.9703,
+      "step": 8003
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.813475489616394,
+      "learning_rate": 1.1803112585563587e-05,
+      "loss": 0.919,
+      "step": 8004
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6517568826675415,
+      "learning_rate": 1.1801285389905567e-05,
+      "loss": 0.9347,
+      "step": 8005
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7512457370758057,
+      "learning_rate": 1.1799458132090214e-05,
+      "loss": 1.0562,
+      "step": 8006
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1174851655960083,
+      "learning_rate": 1.1797630812180592e-05,
+      "loss": 0.6142,
+      "step": 8007
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.728916883468628,
+      "learning_rate": 1.1795803430239752e-05,
+      "loss": 1.0521,
+      "step": 8008
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.625531554222107,
+      "learning_rate": 1.179397598633075e-05,
+      "loss": 0.9305,
+      "step": 8009
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7653635740280151,
+      "learning_rate": 1.1792148480516648e-05,
+      "loss": 0.9921,
+      "step": 8010
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6762639284133911,
+      "learning_rate": 1.1790320912860508e-05,
+      "loss": 0.9083,
+      "step": 8011
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7834948301315308,
+      "learning_rate": 1.1788493283425398e-05,
+      "loss": 1.0187,
+      "step": 8012
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.823069453239441,
+      "learning_rate": 1.1786665592274378e-05,
+      "loss": 0.9573,
+      "step": 8013
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6952208280563354,
+      "learning_rate": 1.1784837839470519e-05,
+      "loss": 0.9276,
+      "step": 8014
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7396788597106934,
+      "learning_rate": 1.1783010025076893e-05,
+      "loss": 0.903,
+      "step": 8015
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7886943817138672,
+      "learning_rate": 1.1781182149156572e-05,
+      "loss": 0.9444,
+      "step": 8016
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8170719146728516,
+      "learning_rate": 1.1779354211772632e-05,
+      "loss": 1.0505,
+      "step": 8017
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.860854983329773,
+      "learning_rate": 1.1777526212988144e-05,
+      "loss": 0.9851,
+      "step": 8018
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.619338870048523,
+      "learning_rate": 1.1775698152866195e-05,
+      "loss": 0.9354,
+      "step": 8019
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.883421778678894,
+      "learning_rate": 1.1773870031469863e-05,
+      "loss": 0.9667,
+      "step": 8020
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8250725269317627,
+      "learning_rate": 1.177204184886223e-05,
+      "loss": 0.9406,
+      "step": 8021
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.766154408454895,
+      "learning_rate": 1.1770213605106384e-05,
+      "loss": 0.9322,
+      "step": 8022
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8425568342208862,
+      "learning_rate": 1.1768385300265409e-05,
+      "loss": 1.1135,
+      "step": 8023
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6613640785217285,
+      "learning_rate": 1.1766556934402398e-05,
+      "loss": 0.9558,
+      "step": 8024
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.895185947418213,
+      "learning_rate": 1.1764728507580442e-05,
+      "loss": 0.8801,
+      "step": 8025
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5727230310440063,
+      "learning_rate": 1.1762900019862635e-05,
+      "loss": 0.8939,
+      "step": 8026
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8296129703521729,
+      "learning_rate": 1.1761071471312075e-05,
+      "loss": 0.9284,
+      "step": 8027
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7631586790084839,
+      "learning_rate": 1.1759242861991855e-05,
+      "loss": 0.9583,
+      "step": 8028
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9332904815673828,
+      "learning_rate": 1.175741419196508e-05,
+      "loss": 0.9527,
+      "step": 8029
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8091992139816284,
+      "learning_rate": 1.175558546129485e-05,
+      "loss": 0.9601,
+      "step": 8030
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7562631368637085,
+      "learning_rate": 1.175375667004427e-05,
+      "loss": 0.9975,
+      "step": 8031
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8765333890914917,
+      "learning_rate": 1.1751927818276445e-05,
+      "loss": 1.0765,
+      "step": 8032
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.712401270866394,
+      "learning_rate": 1.1750098906054485e-05,
+      "loss": 1.0067,
+      "step": 8033
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7910075187683105,
+      "learning_rate": 1.1748269933441501e-05,
+      "loss": 0.9828,
+      "step": 8034
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8870741128921509,
+      "learning_rate": 1.1746440900500604e-05,
+      "loss": 0.9482,
+      "step": 8035
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8458998203277588,
+      "learning_rate": 1.174461180729491e-05,
+      "loss": 0.9521,
+      "step": 8036
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7518543004989624,
+      "learning_rate": 1.1742782653887537e-05,
+      "loss": 0.9304,
+      "step": 8037
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9056423902511597,
+      "learning_rate": 1.1740953440341602e-05,
+      "loss": 0.9669,
+      "step": 8038
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0244139432907104,
+      "learning_rate": 1.1739124166720228e-05,
+      "loss": 0.5684,
+      "step": 8039
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7574100494384766,
+      "learning_rate": 1.1737294833086537e-05,
+      "loss": 0.9641,
+      "step": 8040
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6881746053695679,
+      "learning_rate": 1.1735465439503652e-05,
+      "loss": 0.9955,
+      "step": 8041
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9910071492195129,
+      "learning_rate": 1.1733635986034706e-05,
+      "loss": 0.5522,
+      "step": 8042
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.742068886756897,
+      "learning_rate": 1.1731806472742823e-05,
+      "loss": 0.9728,
+      "step": 8043
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.839636206626892,
+      "learning_rate": 1.1729976899691138e-05,
+      "loss": 1.0572,
+      "step": 8044
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7851154804229736,
+      "learning_rate": 1.172814726694278e-05,
+      "loss": 0.9615,
+      "step": 8045
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.874877691268921,
+      "learning_rate": 1.1726317574560888e-05,
+      "loss": 0.9762,
+      "step": 8046
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7921478748321533,
+      "learning_rate": 1.1724487822608602e-05,
+      "loss": 0.9702,
+      "step": 8047
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8193144798278809,
+      "learning_rate": 1.1722658011149055e-05,
+      "loss": 0.9823,
+      "step": 8048
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.918825626373291,
+      "learning_rate": 1.1720828140245393e-05,
+      "loss": 0.9573,
+      "step": 8049
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8105534315109253,
+      "learning_rate": 1.1718998209960755e-05,
+      "loss": 0.9805,
+      "step": 8050
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.630724549293518,
+      "learning_rate": 1.1717168220358296e-05,
+      "loss": 0.9217,
+      "step": 8051
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7872074842453003,
+      "learning_rate": 1.1715338171501156e-05,
+      "loss": 0.9542,
+      "step": 8052
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8341808319091797,
+      "learning_rate": 1.1713508063452487e-05,
+      "loss": 0.9985,
+      "step": 8053
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8159358501434326,
+      "learning_rate": 1.1711677896275444e-05,
+      "loss": 0.969,
+      "step": 8054
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7973802089691162,
+      "learning_rate": 1.1709847670033176e-05,
+      "loss": 1.0004,
+      "step": 8055
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7079880237579346,
+      "learning_rate": 1.1708017384788842e-05,
+      "loss": 0.9966,
+      "step": 8056
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6823166608810425,
+      "learning_rate": 1.1706187040605598e-05,
+      "loss": 1.0462,
+      "step": 8057
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7455824613571167,
+      "learning_rate": 1.1704356637546606e-05,
+      "loss": 0.8762,
+      "step": 8058
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7439374923706055,
+      "learning_rate": 1.1702526175675026e-05,
+      "loss": 0.9992,
+      "step": 8059
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8739033937454224,
+      "learning_rate": 1.1700695655054027e-05,
+      "loss": 1.0688,
+      "step": 8060
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8617148399353027,
+      "learning_rate": 1.1698865075746768e-05,
+      "loss": 1.0507,
+      "step": 8061
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8059519529342651,
+      "learning_rate": 1.169703443781642e-05,
+      "loss": 1.0325,
+      "step": 8062
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6627306938171387,
+      "learning_rate": 1.1695203741326157e-05,
+      "loss": 0.9753,
+      "step": 8063
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.745413064956665,
+      "learning_rate": 1.169337298633915e-05,
+      "loss": 0.954,
+      "step": 8064
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9172461032867432,
+      "learning_rate": 1.1691542172918566e-05,
+      "loss": 0.9741,
+      "step": 8065
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9184553623199463,
+      "learning_rate": 1.1689711301127591e-05,
+      "loss": 0.9493,
+      "step": 8066
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7978546619415283,
+      "learning_rate": 1.1687880371029398e-05,
+      "loss": 1.0013,
+      "step": 8067
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8455827236175537,
+      "learning_rate": 1.1686049382687168e-05,
+      "loss": 0.9538,
+      "step": 8068
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8107343912124634,
+      "learning_rate": 1.1684218336164083e-05,
+      "loss": 0.9853,
+      "step": 8069
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9131147861480713,
+      "learning_rate": 1.1682387231523328e-05,
+      "loss": 1.0223,
+      "step": 8070
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8731969594955444,
+      "learning_rate": 1.1680556068828092e-05,
+      "loss": 0.9415,
+      "step": 8071
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6293609142303467,
+      "learning_rate": 1.1678724848141555e-05,
+      "loss": 0.9388,
+      "step": 8072
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7343569993972778,
+      "learning_rate": 1.1676893569526918e-05,
+      "loss": 0.9575,
+      "step": 8073
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7709144353866577,
+      "learning_rate": 1.1675062233047365e-05,
+      "loss": 0.9331,
+      "step": 8074
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6687047481536865,
+      "learning_rate": 1.1673230838766094e-05,
+      "loss": 1.0134,
+      "step": 8075
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.072532892227173,
+      "learning_rate": 1.1671399386746301e-05,
+      "loss": 0.9763,
+      "step": 8076
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6952625513076782,
+      "learning_rate": 1.1669567877051184e-05,
+      "loss": 0.9338,
+      "step": 8077
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8068554401397705,
+      "learning_rate": 1.1667736309743945e-05,
+      "loss": 0.9963,
+      "step": 8078
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6265000104904175,
+      "learning_rate": 1.1665904684887784e-05,
+      "loss": 1.0063,
+      "step": 8079
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6523091793060303,
+      "learning_rate": 1.1664073002545903e-05,
+      "loss": 0.9118,
+      "step": 8080
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6876542568206787,
+      "learning_rate": 1.1662241262781515e-05,
+      "loss": 0.9032,
+      "step": 8081
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1174182891845703,
+      "learning_rate": 1.1660409465657822e-05,
+      "loss": 1.0409,
+      "step": 8082
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0477702617645264,
+      "learning_rate": 1.1658577611238037e-05,
+      "loss": 1.0491,
+      "step": 8083
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9220936298370361,
+      "learning_rate": 1.1656745699585373e-05,
+      "loss": 0.949,
+      "step": 8084
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7343533039093018,
+      "learning_rate": 1.165491373076304e-05,
+      "loss": 1.018,
+      "step": 8085
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8112566471099854,
+      "learning_rate": 1.1653081704834259e-05,
+      "loss": 0.979,
+      "step": 8086
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9316182136535645,
+      "learning_rate": 1.1651249621862245e-05,
+      "loss": 0.9953,
+      "step": 8087
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8631106615066528,
+      "learning_rate": 1.164941748191022e-05,
+      "loss": 1.0681,
+      "step": 8088
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7775073051452637,
+      "learning_rate": 1.1647585285041405e-05,
+      "loss": 1.0616,
+      "step": 8089
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7869524955749512,
+      "learning_rate": 1.1645753031319022e-05,
+      "loss": 0.9905,
+      "step": 8090
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8374818563461304,
+      "learning_rate": 1.1643920720806304e-05,
+      "loss": 0.9769,
+      "step": 8091
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7218824625015259,
+      "learning_rate": 1.164208835356647e-05,
+      "loss": 0.9342,
+      "step": 8092
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7000293731689453,
+      "learning_rate": 1.1640255929662756e-05,
+      "loss": 1.0114,
+      "step": 8093
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.101402759552002,
+      "learning_rate": 1.1638423449158388e-05,
+      "loss": 1.0469,
+      "step": 8094
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9071333408355713,
+      "learning_rate": 1.163659091211661e-05,
+      "loss": 1.003,
+      "step": 8095
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7965080738067627,
+      "learning_rate": 1.1634758318600648e-05,
+      "loss": 1.038,
+      "step": 8096
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.691931128501892,
+      "learning_rate": 1.1632925668673743e-05,
+      "loss": 1.0022,
+      "step": 8097
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7160139083862305,
+      "learning_rate": 1.1631092962399134e-05,
+      "loss": 0.937,
+      "step": 8098
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8190345764160156,
+      "learning_rate": 1.1629260199840063e-05,
+      "loss": 0.9406,
+      "step": 8099
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7773182392120361,
+      "learning_rate": 1.1627427381059773e-05,
+      "loss": 0.9615,
+      "step": 8100
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7664669752120972,
+      "learning_rate": 1.162559450612151e-05,
+      "loss": 1.0337,
+      "step": 8101
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8930587768554688,
+      "learning_rate": 1.162376157508852e-05,
+      "loss": 0.9714,
+      "step": 8102
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6648093461990356,
+      "learning_rate": 1.1621928588024058e-05,
+      "loss": 0.9915,
+      "step": 8103
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9049309492111206,
+      "learning_rate": 1.162009554499137e-05,
+      "loss": 1.0744,
+      "step": 8104
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.859045386314392,
+      "learning_rate": 1.1618262446053708e-05,
+      "loss": 0.8913,
+      "step": 8105
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.735291600227356,
+      "learning_rate": 1.1616429291274331e-05,
+      "loss": 0.9519,
+      "step": 8106
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9270039796829224,
+      "learning_rate": 1.1614596080716493e-05,
+      "loss": 0.9057,
+      "step": 8107
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7491557598114014,
+      "learning_rate": 1.1612762814443459e-05,
+      "loss": 0.9619,
+      "step": 8108
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7298047542572021,
+      "learning_rate": 1.1610929492518481e-05,
+      "loss": 1.0075,
+      "step": 8109
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8623096942901611,
+      "learning_rate": 1.1609096115004827e-05,
+      "loss": 1.0101,
+      "step": 8110
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8681354522705078,
+      "learning_rate": 1.1607262681965763e-05,
+      "loss": 0.9269,
+      "step": 8111
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.691609501838684,
+      "learning_rate": 1.1605429193464553e-05,
+      "loss": 0.9408,
+      "step": 8112
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0658972263336182,
+      "learning_rate": 1.1603595649564466e-05,
+      "loss": 0.5628,
+      "step": 8113
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9071061611175537,
+      "learning_rate": 1.160176205032877e-05,
+      "loss": 0.9006,
+      "step": 8114
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7213611602783203,
+      "learning_rate": 1.1599928395820743e-05,
+      "loss": 0.9455,
+      "step": 8115
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8260375261306763,
+      "learning_rate": 1.1598094686103654e-05,
+      "loss": 0.9476,
+      "step": 8116
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9121737480163574,
+      "learning_rate": 1.159626092124078e-05,
+      "loss": 1.0094,
+      "step": 8117
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0367008447647095,
+      "learning_rate": 1.1594427101295404e-05,
+      "loss": 0.5985,
+      "step": 8118
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7241759300231934,
+      "learning_rate": 1.1592593226330802e-05,
+      "loss": 0.9197,
+      "step": 8119
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6419388055801392,
+      "learning_rate": 1.1590759296410256e-05,
+      "loss": 1.0405,
+      "step": 8120
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7511953115463257,
+      "learning_rate": 1.1588925311597052e-05,
+      "loss": 0.9967,
+      "step": 8121
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6367125511169434,
+      "learning_rate": 1.1587091271954471e-05,
+      "loss": 0.9708,
+      "step": 8122
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.884300708770752,
+      "learning_rate": 1.1585257177545805e-05,
+      "loss": 0.9662,
+      "step": 8123
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.030393600463867,
+      "learning_rate": 1.1583423028434343e-05,
+      "loss": 0.9684,
+      "step": 8124
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8546231985092163,
+      "learning_rate": 1.1581588824683375e-05,
+      "loss": 0.9397,
+      "step": 8125
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9013009071350098,
+      "learning_rate": 1.1579754566356195e-05,
+      "loss": 0.9464,
+      "step": 8126
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8146413564682007,
+      "learning_rate": 1.1577920253516097e-05,
+      "loss": 0.9242,
+      "step": 8127
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.724095106124878,
+      "learning_rate": 1.1576085886226376e-05,
+      "loss": 0.9851,
+      "step": 8128
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7956159114837646,
+      "learning_rate": 1.1574251464550337e-05,
+      "loss": 1.0046,
+      "step": 8129
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9521915912628174,
+      "learning_rate": 1.1572416988551277e-05,
+      "loss": 0.9898,
+      "step": 8130
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0285500288009644,
+      "learning_rate": 1.1570582458292499e-05,
+      "loss": 0.5782,
+      "step": 8131
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8780637979507446,
+      "learning_rate": 1.1568747873837307e-05,
+      "loss": 0.9056,
+      "step": 8132
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7956031560897827,
+      "learning_rate": 1.1566913235249008e-05,
+      "loss": 1.0004,
+      "step": 8133
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7687028646469116,
+      "learning_rate": 1.1565078542590912e-05,
+      "loss": 0.9725,
+      "step": 8134
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.656584620475769,
+      "learning_rate": 1.1563243795926327e-05,
+      "loss": 1.0376,
+      "step": 8135
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8720474243164062,
+      "learning_rate": 1.1561408995318565e-05,
+      "loss": 1.0016,
+      "step": 8136
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7733625173568726,
+      "learning_rate": 1.1559574140830938e-05,
+      "loss": 0.9423,
+      "step": 8137
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8585197925567627,
+      "learning_rate": 1.1557739232526766e-05,
+      "loss": 0.9559,
+      "step": 8138
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8304883241653442,
+      "learning_rate": 1.1555904270469363e-05,
+      "loss": 0.9238,
+      "step": 8139
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6804660558700562,
+      "learning_rate": 1.155406925472205e-05,
+      "loss": 0.9287,
+      "step": 8140
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6479648351669312,
+      "learning_rate": 1.155223418534815e-05,
+      "loss": 0.9242,
+      "step": 8141
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7876489162445068,
+      "learning_rate": 1.1550399062410984e-05,
+      "loss": 1.0236,
+      "step": 8142
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9216400384902954,
+      "learning_rate": 1.1548563885973873e-05,
+      "loss": 0.9277,
+      "step": 8143
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0850942134857178,
+      "learning_rate": 1.1546728656100153e-05,
+      "loss": 1.1015,
+      "step": 8144
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9480878114700317,
+      "learning_rate": 1.1544893372853145e-05,
+      "loss": 1.0199,
+      "step": 8145
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8592286109924316,
+      "learning_rate": 1.1543058036296185e-05,
+      "loss": 0.9735,
+      "step": 8146
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7742880582809448,
+      "learning_rate": 1.15412226464926e-05,
+      "loss": 0.9424,
+      "step": 8147
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8639676570892334,
+      "learning_rate": 1.1539387203505728e-05,
+      "loss": 0.9639,
+      "step": 8148
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6871354579925537,
+      "learning_rate": 1.1537551707398904e-05,
+      "loss": 0.8919,
+      "step": 8149
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.184121012687683,
+      "learning_rate": 1.1535716158235466e-05,
+      "loss": 0.5423,
+      "step": 8150
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.103196382522583,
+      "learning_rate": 1.1533880556078751e-05,
+      "loss": 0.933,
+      "step": 8151
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9446314573287964,
+      "learning_rate": 1.1532044900992105e-05,
+      "loss": 0.897,
+      "step": 8152
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8579375743865967,
+      "learning_rate": 1.1530209193038868e-05,
+      "loss": 0.9252,
+      "step": 8153
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8165243864059448,
+      "learning_rate": 1.1528373432282388e-05,
+      "loss": 0.9217,
+      "step": 8154
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8631258010864258,
+      "learning_rate": 1.1526537618786005e-05,
+      "loss": 0.9518,
+      "step": 8155
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6529138088226318,
+      "learning_rate": 1.1524701752613074e-05,
+      "loss": 1.0056,
+      "step": 8156
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9313029050827026,
+      "learning_rate": 1.1522865833826948e-05,
+      "loss": 0.9577,
+      "step": 8157
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.844384789466858,
+      "learning_rate": 1.1521029862490976e-05,
+      "loss": 0.9835,
+      "step": 8158
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1799063682556152,
+      "learning_rate": 1.1519193838668513e-05,
+      "loss": 1.0568,
+      "step": 8159
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.042477607727051,
+      "learning_rate": 1.151735776242291e-05,
+      "loss": 1.042,
+      "step": 8160
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.056021213531494,
+      "learning_rate": 1.1515521633817532e-05,
+      "loss": 1.023,
+      "step": 8161
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7760075330734253,
+      "learning_rate": 1.1513685452915737e-05,
+      "loss": 1.0423,
+      "step": 8162
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7028388977050781,
+      "learning_rate": 1.1511849219780883e-05,
+      "loss": 0.9746,
+      "step": 8163
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.80674409866333,
+      "learning_rate": 1.1510012934476337e-05,
+      "loss": 1.0195,
+      "step": 8164
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8487589359283447,
+      "learning_rate": 1.1508176597065463e-05,
+      "loss": 1.0557,
+      "step": 8165
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7171343564987183,
+      "learning_rate": 1.1506340207611626e-05,
+      "loss": 1.0017,
+      "step": 8166
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.734755039215088,
+      "learning_rate": 1.1504503766178197e-05,
+      "loss": 0.9395,
+      "step": 8167
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.762942910194397,
+      "learning_rate": 1.1502667272828545e-05,
+      "loss": 1.0116,
+      "step": 8168
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7260783910751343,
+      "learning_rate": 1.1500830727626044e-05,
+      "loss": 0.9647,
+      "step": 8169
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.949357032775879,
+      "learning_rate": 1.1498994130634068e-05,
+      "loss": 1.0262,
+      "step": 8170
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8129600286483765,
+      "learning_rate": 1.1497157481915987e-05,
+      "loss": 0.9656,
+      "step": 8171
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7713143825531006,
+      "learning_rate": 1.1495320781535186e-05,
+      "loss": 0.9101,
+      "step": 8172
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7725486755371094,
+      "learning_rate": 1.1493484029555043e-05,
+      "loss": 0.9887,
+      "step": 8173
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.740011215209961,
+      "learning_rate": 1.1491647226038938e-05,
+      "loss": 0.9305,
+      "step": 8174
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7816308736801147,
+      "learning_rate": 1.1489810371050254e-05,
+      "loss": 0.9035,
+      "step": 8175
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7888948917388916,
+      "learning_rate": 1.1487973464652375e-05,
+      "loss": 1.0225,
+      "step": 8176
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7568554878234863,
+      "learning_rate": 1.1486136506908689e-05,
+      "loss": 1.09,
+      "step": 8177
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6500009298324585,
+      "learning_rate": 1.1484299497882585e-05,
+      "loss": 0.9712,
+      "step": 8178
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6973532438278198,
+      "learning_rate": 1.148246243763745e-05,
+      "loss": 0.906,
+      "step": 8179
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7078568935394287,
+      "learning_rate": 1.1480625326236676e-05,
+      "loss": 0.9455,
+      "step": 8180
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6860296726226807,
+      "learning_rate": 1.1478788163743659e-05,
+      "loss": 0.9972,
+      "step": 8181
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8339258432388306,
+      "learning_rate": 1.1476950950221793e-05,
+      "loss": 0.922,
+      "step": 8182
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6800051927566528,
+      "learning_rate": 1.1475113685734476e-05,
+      "loss": 1.0284,
+      "step": 8183
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6942602396011353,
+      "learning_rate": 1.1473276370345105e-05,
+      "loss": 0.9844,
+      "step": 8184
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.689346432685852,
+      "learning_rate": 1.1471439004117082e-05,
+      "loss": 1.1338,
+      "step": 8185
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7736237049102783,
+      "learning_rate": 1.146960158711381e-05,
+      "loss": 0.9605,
+      "step": 8186
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5671947002410889,
+      "learning_rate": 1.1467764119398695e-05,
+      "loss": 0.9928,
+      "step": 8187
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5330764055252075,
+      "learning_rate": 1.1465926601035137e-05,
+      "loss": 0.9783,
+      "step": 8188
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7810726165771484,
+      "learning_rate": 1.1464089032086547e-05,
+      "loss": 0.9475,
+      "step": 8189
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6199408769607544,
+      "learning_rate": 1.1462251412616337e-05,
+      "loss": 0.9272,
+      "step": 8190
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9038352966308594,
+      "learning_rate": 1.1460413742687912e-05,
+      "loss": 1.0067,
+      "step": 8191
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7979843616485596,
+      "learning_rate": 1.1458576022364692e-05,
+      "loss": 0.9196,
+      "step": 8192
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.077195167541504,
+      "learning_rate": 1.1456738251710085e-05,
+      "loss": 0.5643,
+      "step": 8193
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7150534391403198,
+      "learning_rate": 1.1454900430787507e-05,
+      "loss": 1.0892,
+      "step": 8194
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6887699365615845,
+      "learning_rate": 1.1453062559660384e-05,
+      "loss": 1.0092,
+      "step": 8195
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7780054807662964,
+      "learning_rate": 1.145122463839213e-05,
+      "loss": 0.9276,
+      "step": 8196
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7952513694763184,
+      "learning_rate": 1.1449386667046167e-05,
+      "loss": 0.9888,
+      "step": 8197
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9794857501983643,
+      "learning_rate": 1.1447548645685919e-05,
+      "loss": 1.032,
+      "step": 8198
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6215165853500366,
+      "learning_rate": 1.144571057437481e-05,
+      "loss": 0.9448,
+      "step": 8199
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6376278400421143,
+      "learning_rate": 1.144387245317627e-05,
+      "loss": 0.9151,
+      "step": 8200
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8030284643173218,
+      "learning_rate": 1.1442034282153725e-05,
+      "loss": 0.9387,
+      "step": 8201
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0381600856781006,
+      "learning_rate": 1.1440196061370603e-05,
+      "loss": 0.5239,
+      "step": 8202
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7246286869049072,
+      "learning_rate": 1.143835779089034e-05,
+      "loss": 0.889,
+      "step": 8203
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6741750240325928,
+      "learning_rate": 1.1436519470776362e-05,
+      "loss": 0.9248,
+      "step": 8204
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6802411079406738,
+      "learning_rate": 1.1434681101092116e-05,
+      "loss": 1.0066,
+      "step": 8205
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6129857301712036,
+      "learning_rate": 1.1432842681901031e-05,
+      "loss": 0.9193,
+      "step": 8206
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.931418538093567,
+      "learning_rate": 1.1431004213266545e-05,
+      "loss": 0.8692,
+      "step": 8207
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7123961448669434,
+      "learning_rate": 1.1429165695252105e-05,
+      "loss": 1.0199,
+      "step": 8208
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6577389240264893,
+      "learning_rate": 1.1427327127921144e-05,
+      "loss": 1.008,
+      "step": 8209
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6065069437026978,
+      "learning_rate": 1.1425488511337115e-05,
+      "loss": 1.0165,
+      "step": 8210
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.624403715133667,
+      "learning_rate": 1.1423649845563458e-05,
+      "loss": 0.9704,
+      "step": 8211
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7185124158859253,
+      "learning_rate": 1.1421811130663622e-05,
+      "loss": 0.9004,
+      "step": 8212
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7191646099090576,
+      "learning_rate": 1.1419972366701057e-05,
+      "loss": 0.9725,
+      "step": 8213
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9825719594955444,
+      "learning_rate": 1.141813355373921e-05,
+      "loss": 0.5585,
+      "step": 8214
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.939244270324707,
+      "learning_rate": 1.1416294691841539e-05,
+      "loss": 1.0116,
+      "step": 8215
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7359364032745361,
+      "learning_rate": 1.1414455781071489e-05,
+      "loss": 0.9356,
+      "step": 8216
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.899379849433899,
+      "learning_rate": 1.1412616821492526e-05,
+      "loss": 0.9651,
+      "step": 8217
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6946544647216797,
+      "learning_rate": 1.1410777813168102e-05,
+      "loss": 0.9667,
+      "step": 8218
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7538397312164307,
+      "learning_rate": 1.1408938756161675e-05,
+      "loss": 1.01,
+      "step": 8219
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6150877475738525,
+      "learning_rate": 1.1407099650536706e-05,
+      "loss": 0.9478,
+      "step": 8220
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6541019678115845,
+      "learning_rate": 1.1405260496356658e-05,
+      "loss": 0.8805,
+      "step": 8221
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9111180305480957,
+      "learning_rate": 1.1403421293684997e-05,
+      "loss": 0.9774,
+      "step": 8222
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6048967838287354,
+      "learning_rate": 1.1401582042585188e-05,
+      "loss": 1.0198,
+      "step": 8223
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.737343430519104,
+      "learning_rate": 1.1399742743120699e-05,
+      "loss": 0.9027,
+      "step": 8224
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6954782009124756,
+      "learning_rate": 1.1397903395354996e-05,
+      "loss": 0.9416,
+      "step": 8225
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8143391609191895,
+      "learning_rate": 1.1396063999351551e-05,
+      "loss": 0.9159,
+      "step": 8226
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.825358271598816,
+      "learning_rate": 1.1394224555173841e-05,
+      "loss": 0.9694,
+      "step": 8227
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9426848888397217,
+      "learning_rate": 1.1392385062885334e-05,
+      "loss": 0.9851,
+      "step": 8228
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6380014419555664,
+      "learning_rate": 1.1390545522549508e-05,
+      "loss": 0.9515,
+      "step": 8229
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8274394273757935,
+      "learning_rate": 1.138870593422984e-05,
+      "loss": 1.0284,
+      "step": 8230
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7856947183609009,
+      "learning_rate": 1.1386866297989809e-05,
+      "loss": 0.8886,
+      "step": 8231
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.764729619026184,
+      "learning_rate": 1.1385026613892898e-05,
+      "loss": 0.8709,
+      "step": 8232
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9077820777893066,
+      "learning_rate": 1.1383186882002584e-05,
+      "loss": 0.978,
+      "step": 8233
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7588090896606445,
+      "learning_rate": 1.1381347102382356e-05,
+      "loss": 0.9351,
+      "step": 8234
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8538366556167603,
+      "learning_rate": 1.13795072750957e-05,
+      "loss": 1.0181,
+      "step": 8235
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7108629941940308,
+      "learning_rate": 1.13776674002061e-05,
+      "loss": 0.9766,
+      "step": 8236
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7506940364837646,
+      "learning_rate": 1.1375827477777044e-05,
+      "loss": 0.9131,
+      "step": 8237
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7736252546310425,
+      "learning_rate": 1.1373987507872028e-05,
+      "loss": 0.9013,
+      "step": 8238
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.69211745262146,
+      "learning_rate": 1.1372147490554541e-05,
+      "loss": 1.0076,
+      "step": 8239
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0225110054016113,
+      "learning_rate": 1.1370307425888077e-05,
+      "loss": 0.9391,
+      "step": 8240
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.788693904876709,
+      "learning_rate": 1.136846731393613e-05,
+      "loss": 1.0511,
+      "step": 8241
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6437410116195679,
+      "learning_rate": 1.1366627154762202e-05,
+      "loss": 0.9467,
+      "step": 8242
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.671946406364441,
+      "learning_rate": 1.1364786948429788e-05,
+      "loss": 0.9052,
+      "step": 8243
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.76242995262146,
+      "learning_rate": 1.1362946695002383e-05,
+      "loss": 1.0641,
+      "step": 8244
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6688573360443115,
+      "learning_rate": 1.1361106394543502e-05,
+      "loss": 0.8956,
+      "step": 8245
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8608148097991943,
+      "learning_rate": 1.1359266047116636e-05,
+      "loss": 0.9705,
+      "step": 8246
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8750923871994019,
+      "learning_rate": 1.13574256527853e-05,
+      "loss": 1.006,
+      "step": 8247
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8330316543579102,
+      "learning_rate": 1.1355585211612992e-05,
+      "loss": 1.048,
+      "step": 8248
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8987964391708374,
+      "learning_rate": 1.1353744723663227e-05,
+      "loss": 1.017,
+      "step": 8249
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.010899305343628,
+      "learning_rate": 1.1351904188999513e-05,
+      "loss": 1.018,
+      "step": 8250
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.716174840927124,
+      "learning_rate": 1.1350063607685364e-05,
+      "loss": 0.9265,
+      "step": 8251
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7473540306091309,
+      "learning_rate": 1.1348222979784289e-05,
+      "loss": 0.9861,
+      "step": 8252
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9039522409439087,
+      "learning_rate": 1.1346382305359807e-05,
+      "loss": 1.0257,
+      "step": 8253
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9177871942520142,
+      "learning_rate": 1.1344541584475432e-05,
+      "loss": 0.9438,
+      "step": 8254
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8129955530166626,
+      "learning_rate": 1.1342700817194681e-05,
+      "loss": 1.0386,
+      "step": 8255
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7839810848236084,
+      "learning_rate": 1.1340860003581078e-05,
+      "loss": 0.9758,
+      "step": 8256
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7627233266830444,
+      "learning_rate": 1.133901914369814e-05,
+      "loss": 0.9774,
+      "step": 8257
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.714215636253357,
+      "learning_rate": 1.1337178237609391e-05,
+      "loss": 0.9081,
+      "step": 8258
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6672221422195435,
+      "learning_rate": 1.1335337285378359e-05,
+      "loss": 0.9405,
+      "step": 8259
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7233778238296509,
+      "learning_rate": 1.1333496287068565e-05,
+      "loss": 0.8735,
+      "step": 8260
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.244443655014038,
+      "learning_rate": 1.1331655242743538e-05,
+      "loss": 0.9356,
+      "step": 8261
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7163015604019165,
+      "learning_rate": 1.1329814152466811e-05,
+      "loss": 0.9486,
+      "step": 8262
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0177924633026123,
+      "learning_rate": 1.1327973016301912e-05,
+      "loss": 0.9723,
+      "step": 8263
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7468641996383667,
+      "learning_rate": 1.1326131834312372e-05,
+      "loss": 0.9985,
+      "step": 8264
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8751897811889648,
+      "learning_rate": 1.132429060656173e-05,
+      "loss": 0.8387,
+      "step": 8265
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6528464555740356,
+      "learning_rate": 1.1322449333113517e-05,
+      "loss": 0.8574,
+      "step": 8266
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9983199834823608,
+      "learning_rate": 1.1320608014031272e-05,
+      "loss": 0.9916,
+      "step": 8267
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9700661897659302,
+      "learning_rate": 1.1318766649378532e-05,
+      "loss": 1.0186,
+      "step": 8268
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9696274995803833,
+      "learning_rate": 1.1316925239218838e-05,
+      "loss": 1.0042,
+      "step": 8269
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.747190237045288,
+      "learning_rate": 1.1315083783615734e-05,
+      "loss": 0.9391,
+      "step": 8270
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7099398374557495,
+      "learning_rate": 1.1313242282632762e-05,
+      "loss": 0.9736,
+      "step": 8271
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8568205833435059,
+      "learning_rate": 1.1311400736333466e-05,
+      "loss": 0.9857,
+      "step": 8272
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8906890153884888,
+      "learning_rate": 1.1309559144781397e-05,
+      "loss": 0.9708,
+      "step": 8273
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0576772689819336,
+      "learning_rate": 1.1307717508040099e-05,
+      "loss": 0.9543,
+      "step": 8274
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.752837896347046,
+      "learning_rate": 1.1305875826173119e-05,
+      "loss": 1.0138,
+      "step": 8275
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6970767974853516,
+      "learning_rate": 1.1304034099244015e-05,
+      "loss": 1.0017,
+      "step": 8276
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7145463228225708,
+      "learning_rate": 1.1302192327316338e-05,
+      "loss": 0.9518,
+      "step": 8277
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8148404359817505,
+      "learning_rate": 1.130035051045364e-05,
+      "loss": 0.8566,
+      "step": 8278
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7286200523376465,
+      "learning_rate": 1.129850864871948e-05,
+      "loss": 1.036,
+      "step": 8279
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6262943744659424,
+      "learning_rate": 1.129666674217741e-05,
+      "loss": 0.9612,
+      "step": 8280
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.700950264930725,
+      "learning_rate": 1.1294824790890997e-05,
+      "loss": 0.816,
+      "step": 8281
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7229434251785278,
+      "learning_rate": 1.1292982794923795e-05,
+      "loss": 0.8762,
+      "step": 8282
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9246916770935059,
+      "learning_rate": 1.129114075433937e-05,
+      "loss": 0.8443,
+      "step": 8283
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.68325674533844,
+      "learning_rate": 1.1289298669201284e-05,
+      "loss": 0.8917,
+      "step": 8284
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2466845512390137,
+      "learning_rate": 1.12874565395731e-05,
+      "loss": 0.9945,
+      "step": 8285
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8770309686660767,
+      "learning_rate": 1.1285614365518392e-05,
+      "loss": 0.9111,
+      "step": 8286
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8043227195739746,
+      "learning_rate": 1.1283772147100717e-05,
+      "loss": 0.9361,
+      "step": 8287
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7761905193328857,
+      "learning_rate": 1.1281929884383655e-05,
+      "loss": 1.0958,
+      "step": 8288
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.702733039855957,
+      "learning_rate": 1.1280087577430775e-05,
+      "loss": 0.9483,
+      "step": 8289
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0956785678863525,
+      "learning_rate": 1.1278245226305646e-05,
+      "loss": 1.0398,
+      "step": 8290
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.981734275817871,
+      "learning_rate": 1.1276402831071844e-05,
+      "loss": 0.9186,
+      "step": 8291
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.828492522239685,
+      "learning_rate": 1.1274560391792948e-05,
+      "loss": 0.9848,
+      "step": 8292
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8461949825286865,
+      "learning_rate": 1.1272717908532533e-05,
+      "loss": 0.9095,
+      "step": 8293
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9354281425476074,
+      "learning_rate": 1.1270875381354178e-05,
+      "loss": 0.9926,
+      "step": 8294
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7985162734985352,
+      "learning_rate": 1.1269032810321464e-05,
+      "loss": 1.0163,
+      "step": 8295
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.726866602897644,
+      "learning_rate": 1.1267190195497973e-05,
+      "loss": 0.9998,
+      "step": 8296
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.859765887260437,
+      "learning_rate": 1.1265347536947286e-05,
+      "loss": 0.9978,
+      "step": 8297
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7985138893127441,
+      "learning_rate": 1.1263504834732993e-05,
+      "loss": 1.0298,
+      "step": 8298
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.705773949623108,
+      "learning_rate": 1.1261662088918675e-05,
+      "loss": 0.8854,
+      "step": 8299
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.581881046295166,
+      "learning_rate": 1.1259819299567922e-05,
+      "loss": 0.9663,
+      "step": 8300
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8307021856307983,
+      "learning_rate": 1.1257976466744326e-05,
+      "loss": 1.0094,
+      "step": 8301
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6894458532333374,
+      "learning_rate": 1.1256133590511475e-05,
+      "loss": 0.9881,
+      "step": 8302
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8815500736236572,
+      "learning_rate": 1.1254290670932964e-05,
+      "loss": 0.9345,
+      "step": 8303
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7606112957000732,
+      "learning_rate": 1.1252447708072386e-05,
+      "loss": 0.9682,
+      "step": 8304
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.950756549835205,
+      "learning_rate": 1.1250604701993334e-05,
+      "loss": 0.9814,
+      "step": 8305
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8115973472595215,
+      "learning_rate": 1.1248761652759408e-05,
+      "loss": 0.964,
+      "step": 8306
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9059438705444336,
+      "learning_rate": 1.1246918560434206e-05,
+      "loss": 0.966,
+      "step": 8307
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.974716067314148,
+      "learning_rate": 1.1245075425081328e-05,
+      "loss": 0.9919,
+      "step": 8308
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6902050971984863,
+      "learning_rate": 1.1243232246764376e-05,
+      "loss": 0.9353,
+      "step": 8309
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8598041534423828,
+      "learning_rate": 1.124138902554695e-05,
+      "loss": 0.9788,
+      "step": 8310
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9292845726013184,
+      "learning_rate": 1.1239545761492658e-05,
+      "loss": 1.0386,
+      "step": 8311
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.803775668144226,
+      "learning_rate": 1.1237702454665102e-05,
+      "loss": 0.9999,
+      "step": 8312
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9135504961013794,
+      "learning_rate": 1.1235859105127895e-05,
+      "loss": 0.9917,
+      "step": 8313
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8946150541305542,
+      "learning_rate": 1.1234015712944639e-05,
+      "loss": 0.9743,
+      "step": 8314
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1521081924438477,
+      "learning_rate": 1.123217227817895e-05,
+      "loss": 0.5679,
+      "step": 8315
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7653307914733887,
+      "learning_rate": 1.1230328800894437e-05,
+      "loss": 0.9603,
+      "step": 8316
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.795918583869934,
+      "learning_rate": 1.1228485281154713e-05,
+      "loss": 0.9187,
+      "step": 8317
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9153954982757568,
+      "learning_rate": 1.1226641719023395e-05,
+      "loss": 0.9938,
+      "step": 8318
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.774968147277832,
+      "learning_rate": 1.1224798114564097e-05,
+      "loss": 0.9587,
+      "step": 8319
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1230523586273193,
+      "learning_rate": 1.1222954467840439e-05,
+      "loss": 1.0183,
+      "step": 8320
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.866174340248108,
+      "learning_rate": 1.1221110778916037e-05,
+      "loss": 0.9619,
+      "step": 8321
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9503875970840454,
+      "learning_rate": 1.1219267047854515e-05,
+      "loss": 1.0731,
+      "step": 8322
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8368667364120483,
+      "learning_rate": 1.121742327471949e-05,
+      "loss": 0.9705,
+      "step": 8323
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6803312301635742,
+      "learning_rate": 1.121557945957459e-05,
+      "loss": 0.9222,
+      "step": 8324
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8311927318572998,
+      "learning_rate": 1.1213735602483439e-05,
+      "loss": 0.9675,
+      "step": 8325
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8211675882339478,
+      "learning_rate": 1.121189170350966e-05,
+      "loss": 0.9592,
+      "step": 8326
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7949206829071045,
+      "learning_rate": 1.1210047762716885e-05,
+      "loss": 1.0078,
+      "step": 8327
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6693055629730225,
+      "learning_rate": 1.1208203780168743e-05,
+      "loss": 0.8943,
+      "step": 8328
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8428863286972046,
+      "learning_rate": 1.1206359755928865e-05,
+      "loss": 0.9754,
+      "step": 8329
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.856624722480774,
+      "learning_rate": 1.1204515690060878e-05,
+      "loss": 0.987,
+      "step": 8330
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9839272499084473,
+      "learning_rate": 1.1202671582628422e-05,
+      "loss": 0.9798,
+      "step": 8331
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8181593418121338,
+      "learning_rate": 1.1200827433695128e-05,
+      "loss": 0.9482,
+      "step": 8332
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7232624292373657,
+      "learning_rate": 1.1198983243324635e-05,
+      "loss": 0.9369,
+      "step": 8333
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7491872310638428,
+      "learning_rate": 1.1197139011580578e-05,
+      "loss": 0.997,
+      "step": 8334
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.760096549987793,
+      "learning_rate": 1.1195294738526598e-05,
+      "loss": 0.9388,
+      "step": 8335
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8956553936004639,
+      "learning_rate": 1.1193450424226333e-05,
+      "loss": 1.008,
+      "step": 8336
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9834154844284058,
+      "learning_rate": 1.119160606874343e-05,
+      "loss": 0.9404,
+      "step": 8337
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.741002082824707,
+      "learning_rate": 1.118976167214153e-05,
+      "loss": 1.007,
+      "step": 8338
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7230005264282227,
+      "learning_rate": 1.118791723448428e-05,
+      "loss": 0.9516,
+      "step": 8339
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6550581455230713,
+      "learning_rate": 1.1186072755835322e-05,
+      "loss": 0.9014,
+      "step": 8340
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8168171644210815,
+      "learning_rate": 1.1184228236258306e-05,
+      "loss": 1.0582,
+      "step": 8341
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7454965114593506,
+      "learning_rate": 1.1182383675816884e-05,
+      "loss": 0.99,
+      "step": 8342
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8558862209320068,
+      "learning_rate": 1.1180539074574703e-05,
+      "loss": 1.0687,
+      "step": 8343
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.795082926750183,
+      "learning_rate": 1.1178694432595415e-05,
+      "loss": 0.9355,
+      "step": 8344
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.940369963645935,
+      "learning_rate": 1.1176849749942677e-05,
+      "loss": 1.0404,
+      "step": 8345
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7173689603805542,
+      "learning_rate": 1.1175005026680141e-05,
+      "loss": 0.9502,
+      "step": 8346
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7479103803634644,
+      "learning_rate": 1.1173160262871465e-05,
+      "loss": 0.9302,
+      "step": 8347
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9141708612442017,
+      "learning_rate": 1.1171315458580302e-05,
+      "loss": 0.9349,
+      "step": 8348
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8113453388214111,
+      "learning_rate": 1.1169470613870318e-05,
+      "loss": 0.9286,
+      "step": 8349
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6964112520217896,
+      "learning_rate": 1.1167625728805169e-05,
+      "loss": 0.9412,
+      "step": 8350
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0398805141448975,
+      "learning_rate": 1.1165780803448516e-05,
+      "loss": 0.9488,
+      "step": 8351
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.011366605758667,
+      "learning_rate": 1.1163935837864026e-05,
+      "loss": 1.017,
+      "step": 8352
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.75300931930542,
+      "learning_rate": 1.1162090832115357e-05,
+      "loss": 0.946,
+      "step": 8353
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8279283046722412,
+      "learning_rate": 1.1160245786266184e-05,
+      "loss": 0.9376,
+      "step": 8354
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1573725938796997,
+      "learning_rate": 1.115840070038017e-05,
+      "loss": 0.579,
+      "step": 8355
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9647483825683594,
+      "learning_rate": 1.1156555574520982e-05,
+      "loss": 0.9478,
+      "step": 8356
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0390822887420654,
+      "learning_rate": 1.1154710408752294e-05,
+      "loss": 0.9201,
+      "step": 8357
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9553459882736206,
+      "learning_rate": 1.1152865203137773e-05,
+      "loss": 0.9865,
+      "step": 8358
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8672056198120117,
+      "learning_rate": 1.1151019957741096e-05,
+      "loss": 1.0424,
+      "step": 8359
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.752069354057312,
+      "learning_rate": 1.1149174672625933e-05,
+      "loss": 0.9823,
+      "step": 8360
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8578617572784424,
+      "learning_rate": 1.1147329347855964e-05,
+      "loss": 1.0587,
+      "step": 8361
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7429560422897339,
+      "learning_rate": 1.1145483983494865e-05,
+      "loss": 0.9378,
+      "step": 8362
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9389619827270508,
+      "learning_rate": 1.1143638579606313e-05,
+      "loss": 1.0593,
+      "step": 8363
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7277865409851074,
+      "learning_rate": 1.1141793136253987e-05,
+      "loss": 0.9136,
+      "step": 8364
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.920994758605957,
+      "learning_rate": 1.1139947653501569e-05,
+      "loss": 1.0137,
+      "step": 8365
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.847996711730957,
+      "learning_rate": 1.1138102131412742e-05,
+      "loss": 0.9533,
+      "step": 8366
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7338600158691406,
+      "learning_rate": 1.1136256570051192e-05,
+      "loss": 0.8973,
+      "step": 8367
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.866511344909668,
+      "learning_rate": 1.1134410969480598e-05,
+      "loss": 0.8825,
+      "step": 8368
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8479801416397095,
+      "learning_rate": 1.1132565329764651e-05,
+      "loss": 0.9843,
+      "step": 8369
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8212615251541138,
+      "learning_rate": 1.113071965096704e-05,
+      "loss": 1.0527,
+      "step": 8370
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8662315607070923,
+      "learning_rate": 1.1128873933151452e-05,
+      "loss": 0.9717,
+      "step": 8371
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.891660451889038,
+      "learning_rate": 1.1127028176381577e-05,
+      "loss": 0.9898,
+      "step": 8372
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.72786545753479,
+      "learning_rate": 1.1125182380721109e-05,
+      "loss": 0.8789,
+      "step": 8373
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6998101472854614,
+      "learning_rate": 1.112333654623374e-05,
+      "loss": 1.0874,
+      "step": 8374
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7416257858276367,
+      "learning_rate": 1.1121490672983167e-05,
+      "loss": 0.9948,
+      "step": 8375
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8130559921264648,
+      "learning_rate": 1.1119644761033079e-05,
+      "loss": 1.0171,
+      "step": 8376
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.724853515625,
+      "learning_rate": 1.1117798810447182e-05,
+      "loss": 0.981,
+      "step": 8377
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.957699179649353,
+      "learning_rate": 1.1115952821289168e-05,
+      "loss": 1.0227,
+      "step": 8378
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.802990436553955,
+      "learning_rate": 1.1114106793622742e-05,
+      "loss": 0.9424,
+      "step": 8379
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7563058137893677,
+      "learning_rate": 1.1112260727511597e-05,
+      "loss": 0.945,
+      "step": 8380
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8949158191680908,
+      "learning_rate": 1.1110414623019446e-05,
+      "loss": 0.9147,
+      "step": 8381
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5763353109359741,
+      "learning_rate": 1.1108568480209986e-05,
+      "loss": 0.8286,
+      "step": 8382
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0792378187179565,
+      "learning_rate": 1.1106722299146926e-05,
+      "loss": 0.6325,
+      "step": 8383
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9433655738830566,
+      "learning_rate": 1.110487607989397e-05,
+      "loss": 1.0396,
+      "step": 8384
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9687572717666626,
+      "learning_rate": 1.1103029822514828e-05,
+      "loss": 1.0082,
+      "step": 8385
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8394485712051392,
+      "learning_rate": 1.1101183527073207e-05,
+      "loss": 1.0391,
+      "step": 8386
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8573013544082642,
+      "learning_rate": 1.109933719363282e-05,
+      "loss": 0.9547,
+      "step": 8387
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.713350534439087,
+      "learning_rate": 1.1097490822257377e-05,
+      "loss": 1.0192,
+      "step": 8388
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8887578248977661,
+      "learning_rate": 1.1095644413010591e-05,
+      "loss": 1.0308,
+      "step": 8389
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.717733383178711,
+      "learning_rate": 1.1093797965956177e-05,
+      "loss": 1.0069,
+      "step": 8390
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9532909393310547,
+      "learning_rate": 1.109195148115785e-05,
+      "loss": 0.6081,
+      "step": 8391
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7393585443496704,
+      "learning_rate": 1.1090104958679323e-05,
+      "loss": 0.9115,
+      "step": 8392
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6626917123794556,
+      "learning_rate": 1.1088258398584327e-05,
+      "loss": 1.0603,
+      "step": 8393
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9202543497085571,
+      "learning_rate": 1.108641180093657e-05,
+      "loss": 0.9076,
+      "step": 8394
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6771575212478638,
+      "learning_rate": 1.1084565165799777e-05,
+      "loss": 0.979,
+      "step": 8395
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.724423885345459,
+      "learning_rate": 1.108271849323767e-05,
+      "loss": 0.9327,
+      "step": 8396
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.928701400756836,
+      "learning_rate": 1.108087178331397e-05,
+      "loss": 0.9942,
+      "step": 8397
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.933451771736145,
+      "learning_rate": 1.1079025036092408e-05,
+      "loss": 0.9508,
+      "step": 8398
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6016159057617188,
+      "learning_rate": 1.1077178251636702e-05,
+      "loss": 1.053,
+      "step": 8399
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8350372314453125,
+      "learning_rate": 1.1075331430010587e-05,
+      "loss": 0.9047,
+      "step": 8400
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2960610389709473,
+      "learning_rate": 1.1073484571277786e-05,
+      "loss": 0.5847,
+      "step": 8401
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6901441812515259,
+      "learning_rate": 1.107163767550203e-05,
+      "loss": 0.9589,
+      "step": 8402
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.028442859649658,
+      "learning_rate": 1.1069790742747053e-05,
+      "loss": 0.9041,
+      "step": 8403
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9091041088104248,
+      "learning_rate": 1.1067943773076585e-05,
+      "loss": 1.0307,
+      "step": 8404
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0231430530548096,
+      "learning_rate": 1.1066096766554365e-05,
+      "loss": 0.9905,
+      "step": 8405
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8749932050704956,
+      "learning_rate": 1.1064249723244117e-05,
+      "loss": 1.0841,
+      "step": 8406
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3131743669509888,
+      "learning_rate": 1.1062402643209586e-05,
+      "loss": 0.6178,
+      "step": 8407
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8546671867370605,
+      "learning_rate": 1.1060555526514508e-05,
+      "loss": 1.0173,
+      "step": 8408
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8187686204910278,
+      "learning_rate": 1.1058708373222622e-05,
+      "loss": 0.9755,
+      "step": 8409
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7127286195755005,
+      "learning_rate": 1.1056861183397669e-05,
+      "loss": 0.9295,
+      "step": 8410
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6071020364761353,
+      "learning_rate": 1.1055013957103387e-05,
+      "loss": 0.9157,
+      "step": 8411
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5687869787216187,
+      "learning_rate": 1.105316669440352e-05,
+      "loss": 0.8657,
+      "step": 8412
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7713377475738525,
+      "learning_rate": 1.1051319395361812e-05,
+      "loss": 1.0018,
+      "step": 8413
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8914685249328613,
+      "learning_rate": 1.104947206004201e-05,
+      "loss": 0.974,
+      "step": 8414
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1528468132019043,
+      "learning_rate": 1.104762468850786e-05,
+      "loss": 0.6008,
+      "step": 8415
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.817579746246338,
+      "learning_rate": 1.1045777280823105e-05,
+      "loss": 0.9958,
+      "step": 8416
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.805662989616394,
+      "learning_rate": 1.10439298370515e-05,
+      "loss": 0.9923,
+      "step": 8417
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.104353427886963,
+      "learning_rate": 1.1042082357256789e-05,
+      "loss": 1.0412,
+      "step": 8418
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6318070888519287,
+      "learning_rate": 1.1040234841502728e-05,
+      "loss": 0.922,
+      "step": 8419
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6960625648498535,
+      "learning_rate": 1.1038387289853069e-05,
+      "loss": 0.9692,
+      "step": 8420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8021087646484375,
+      "learning_rate": 1.1036539702371565e-05,
+      "loss": 0.9113,
+      "step": 8421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6216925382614136,
+      "learning_rate": 1.1034692079121972e-05,
+      "loss": 0.9411,
+      "step": 8422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7137560844421387,
+      "learning_rate": 1.1032844420168045e-05,
+      "loss": 1.0105,
+      "step": 8423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7940731048583984,
+      "learning_rate": 1.1030996725573544e-05,
+      "loss": 0.9095,
+      "step": 8424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6126354932785034,
+      "learning_rate": 1.1029148995402224e-05,
+      "loss": 0.8851,
+      "step": 8425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9155354499816895,
+      "learning_rate": 1.1027301229717849e-05,
+      "loss": 1.0097,
+      "step": 8426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7939229011535645,
+      "learning_rate": 1.1025453428584176e-05,
+      "loss": 1.014,
+      "step": 8427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.815770149230957,
+      "learning_rate": 1.102360559206497e-05,
+      "loss": 1.0026,
+      "step": 8428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6997798681259155,
+      "learning_rate": 1.1021757720223996e-05,
+      "loss": 0.9044,
+      "step": 8429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8816899061203003,
+      "learning_rate": 1.1019909813125016e-05,
+      "loss": 0.9436,
+      "step": 8430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8161057233810425,
+      "learning_rate": 1.1018061870831795e-05,
+      "loss": 0.9965,
+      "step": 8431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7540086507797241,
+      "learning_rate": 1.1016213893408105e-05,
+      "loss": 1.0504,
+      "step": 8432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8399978876113892,
+      "learning_rate": 1.1014365880917713e-05,
+      "loss": 1.0044,
+      "step": 8433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6707216501235962,
+      "learning_rate": 1.1012517833424387e-05,
+      "loss": 1.0404,
+      "step": 8434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6970584392547607,
+      "learning_rate": 1.1010669750991898e-05,
+      "loss": 0.9086,
+      "step": 8435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7782795429229736,
+      "learning_rate": 1.100882163368402e-05,
+      "loss": 0.9556,
+      "step": 8436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.909121036529541,
+      "learning_rate": 1.1006973481564527e-05,
+      "loss": 0.9839,
+      "step": 8437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.827270746231079,
+      "learning_rate": 1.100512529469719e-05,
+      "loss": 0.9375,
+      "step": 8438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.916403889656067,
+      "learning_rate": 1.1003277073145788e-05,
+      "loss": 0.9614,
+      "step": 8439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8410837650299072,
+      "learning_rate": 1.1001428816974095e-05,
+      "loss": 0.894,
+      "step": 8440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8386131525039673,
+      "learning_rate": 1.0999580526245894e-05,
+      "loss": 1.0033,
+      "step": 8441
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0226311683654785,
+      "learning_rate": 1.099773220102496e-05,
+      "loss": 0.9071,
+      "step": 8442
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.64186692237854,
+      "learning_rate": 1.0995883841375073e-05,
+      "loss": 0.9223,
+      "step": 8443
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.689921498298645,
+      "learning_rate": 1.0994035447360018e-05,
+      "loss": 1.0059,
+      "step": 8444
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7537623643875122,
+      "learning_rate": 1.0992187019043576e-05,
+      "loss": 0.9516,
+      "step": 8445
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6835432052612305,
+      "learning_rate": 1.0990338556489531e-05,
+      "loss": 0.9587,
+      "step": 8446
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7420859336853027,
+      "learning_rate": 1.098849005976167e-05,
+      "loss": 1.077,
+      "step": 8447
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9376620054244995,
+      "learning_rate": 1.0986641528923776e-05,
+      "loss": 1.0043,
+      "step": 8448
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.762676477432251,
+      "learning_rate": 1.0984792964039641e-05,
+      "loss": 0.9822,
+      "step": 8449
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7219209671020508,
+      "learning_rate": 1.0982944365173052e-05,
+      "loss": 0.9196,
+      "step": 8450
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6521238088607788,
+      "learning_rate": 1.0981095732387799e-05,
+      "loss": 0.9653,
+      "step": 8451
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9159728288650513,
+      "learning_rate": 1.0979247065747672e-05,
+      "loss": 1.0285,
+      "step": 8452
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0150840282440186,
+      "learning_rate": 1.0977398365316464e-05,
+      "loss": 1.046,
+      "step": 8453
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6001360416412354,
+      "learning_rate": 1.0975549631157969e-05,
+      "loss": 0.9198,
+      "step": 8454
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.113195776939392,
+      "learning_rate": 1.0973700863335981e-05,
+      "loss": 0.5876,
+      "step": 8455
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7141731977462769,
+      "learning_rate": 1.0971852061914296e-05,
+      "loss": 0.9513,
+      "step": 8456
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6028388738632202,
+      "learning_rate": 1.0970003226956713e-05,
+      "loss": 1.0388,
+      "step": 8457
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9380491971969604,
+      "learning_rate": 1.0968154358527024e-05,
+      "loss": 0.994,
+      "step": 8458
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7320345640182495,
+      "learning_rate": 1.0966305456689034e-05,
+      "loss": 0.8716,
+      "step": 8459
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.802369475364685,
+      "learning_rate": 1.0964456521506545e-05,
+      "loss": 0.9675,
+      "step": 8460
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.738146185874939,
+      "learning_rate": 1.0962607553043354e-05,
+      "loss": 0.9809,
+      "step": 8461
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6915006637573242,
+      "learning_rate": 1.0960758551363265e-05,
+      "loss": 0.9239,
+      "step": 8462
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7923160791397095,
+      "learning_rate": 1.0958909516530082e-05,
+      "loss": 0.9883,
+      "step": 8463
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8369743824005127,
+      "learning_rate": 1.095706044860761e-05,
+      "loss": 1.0698,
+      "step": 8464
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6356401443481445,
+      "learning_rate": 1.0955211347659655e-05,
+      "loss": 0.9155,
+      "step": 8465
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8413978815078735,
+      "learning_rate": 1.0953362213750027e-05,
+      "loss": 0.9745,
+      "step": 8466
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6407099962234497,
+      "learning_rate": 1.095151304694253e-05,
+      "loss": 1.02,
+      "step": 8467
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0899373292922974,
+      "learning_rate": 1.0949663847300976e-05,
+      "loss": 0.6003,
+      "step": 8468
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.731991171836853,
+      "learning_rate": 1.0947814614889174e-05,
+      "loss": 0.9317,
+      "step": 8469
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7802760601043701,
+      "learning_rate": 1.094596534977094e-05,
+      "loss": 0.97,
+      "step": 8470
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6222083568572998,
+      "learning_rate": 1.094411605201008e-05,
+      "loss": 1.0078,
+      "step": 8471
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7317465543746948,
+      "learning_rate": 1.0942266721670418e-05,
+      "loss": 0.9625,
+      "step": 8472
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.783359169960022,
+      "learning_rate": 1.0940417358815758e-05,
+      "loss": 0.8993,
+      "step": 8473
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7969245910644531,
+      "learning_rate": 1.0938567963509925e-05,
+      "loss": 0.9271,
+      "step": 8474
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8140931129455566,
+      "learning_rate": 1.0936718535816733e-05,
+      "loss": 0.9919,
+      "step": 8475
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5985990762710571,
+      "learning_rate": 1.09348690758e-05,
+      "loss": 0.9681,
+      "step": 8476
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8745919466018677,
+      "learning_rate": 1.0933019583523549e-05,
+      "loss": 1.0543,
+      "step": 8477
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9741621017456055,
+      "learning_rate": 1.0931170059051198e-05,
+      "loss": 0.5679,
+      "step": 8478
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.719829797744751,
+      "learning_rate": 1.0929320502446768e-05,
+      "loss": 0.9621,
+      "step": 8479
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8366910219192505,
+      "learning_rate": 1.0927470913774085e-05,
+      "loss": 1.014,
+      "step": 8480
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.743626594543457,
+      "learning_rate": 1.0925621293096971e-05,
+      "loss": 0.9749,
+      "step": 8481
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0183520317077637,
+      "learning_rate": 1.0923771640479251e-05,
+      "loss": 0.5842,
+      "step": 8482
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.734700083732605,
+      "learning_rate": 1.0921921955984753e-05,
+      "loss": 1.0663,
+      "step": 8483
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6550425291061401,
+      "learning_rate": 1.0920072239677302e-05,
+      "loss": 0.9093,
+      "step": 8484
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.676248550415039,
+      "learning_rate": 1.0918222491620726e-05,
+      "loss": 1.0079,
+      "step": 8485
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7471917867660522,
+      "learning_rate": 1.0916372711878862e-05,
+      "loss": 0.9783,
+      "step": 8486
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7320146560668945,
+      "learning_rate": 1.0914522900515535e-05,
+      "loss": 0.9686,
+      "step": 8487
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6843491792678833,
+      "learning_rate": 1.0912673057594574e-05,
+      "loss": 0.9572,
+      "step": 8488
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8031753301620483,
+      "learning_rate": 1.0910823183179818e-05,
+      "loss": 0.9828,
+      "step": 8489
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.805920124053955,
+      "learning_rate": 1.0908973277335097e-05,
+      "loss": 1.0482,
+      "step": 8490
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8119913339614868,
+      "learning_rate": 1.0907123340124248e-05,
+      "loss": 0.9473,
+      "step": 8491
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8013060092926025,
+      "learning_rate": 1.0905273371611104e-05,
+      "loss": 1.0078,
+      "step": 8492
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6332967281341553,
+      "learning_rate": 1.090342337185951e-05,
+      "loss": 0.9601,
+      "step": 8493
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.704505205154419,
+      "learning_rate": 1.0901573340933296e-05,
+      "loss": 0.9334,
+      "step": 8494
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7328894138336182,
+      "learning_rate": 1.0899723278896304e-05,
+      "loss": 0.9862,
+      "step": 8495
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8468217849731445,
+      "learning_rate": 1.0897873185812375e-05,
+      "loss": 0.9961,
+      "step": 8496
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6798930168151855,
+      "learning_rate": 1.089602306174535e-05,
+      "loss": 1.0261,
+      "step": 8497
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6860201358795166,
+      "learning_rate": 1.089417290675907e-05,
+      "loss": 0.9836,
+      "step": 8498
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8676429986953735,
+      "learning_rate": 1.0892322720917386e-05,
+      "loss": 1.0077,
+      "step": 8499
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8589872121810913,
+      "learning_rate": 1.0890472504284133e-05,
+      "loss": 1.0459,
+      "step": 8500
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7926000356674194,
+      "learning_rate": 1.0888622256923165e-05,
+      "loss": 1.01,
+      "step": 8501
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8592532873153687,
+      "learning_rate": 1.0886771978898321e-05,
+      "loss": 1.0103,
+      "step": 8502
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9396649599075317,
+      "learning_rate": 1.0884921670273453e-05,
+      "loss": 0.532,
+      "step": 8503
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9368165731430054,
+      "learning_rate": 1.0883071331112416e-05,
+      "loss": 0.9072,
+      "step": 8504
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.020904064178467,
+      "learning_rate": 1.0881220961479052e-05,
+      "loss": 0.9232,
+      "step": 8505
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7988852262496948,
+      "learning_rate": 1.0879370561437211e-05,
+      "loss": 0.9263,
+      "step": 8506
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6805570125579834,
+      "learning_rate": 1.0877520131050749e-05,
+      "loss": 0.9777,
+      "step": 8507
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5848116874694824,
+      "learning_rate": 1.0875669670383521e-05,
+      "loss": 0.8678,
+      "step": 8508
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7621204853057861,
+      "learning_rate": 1.0873819179499378e-05,
+      "loss": 0.8752,
+      "step": 8509
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.67289400100708,
+      "learning_rate": 1.0871968658462176e-05,
+      "loss": 0.9045,
+      "step": 8510
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8348678350448608,
+      "learning_rate": 1.0870118107335772e-05,
+      "loss": 0.9645,
+      "step": 8511
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.053372383117676,
+      "learning_rate": 1.086826752618402e-05,
+      "loss": 1.0494,
+      "step": 8512
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7624744176864624,
+      "learning_rate": 1.0866416915070781e-05,
+      "loss": 1.0051,
+      "step": 8513
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7851784229278564,
+      "learning_rate": 1.0864566274059919e-05,
+      "loss": 0.9694,
+      "step": 8514
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9425764083862305,
+      "learning_rate": 1.0862715603215285e-05,
+      "loss": 0.9749,
+      "step": 8515
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8974881172180176,
+      "learning_rate": 1.0860864902600748e-05,
+      "loss": 1.0146,
+      "step": 8516
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.948978066444397,
+      "learning_rate": 1.085901417228017e-05,
+      "loss": 0.9767,
+      "step": 8517
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.036729097366333,
+      "learning_rate": 1.085716341231741e-05,
+      "loss": 0.5687,
+      "step": 8518
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.75920569896698,
+      "learning_rate": 1.0855312622776333e-05,
+      "loss": 0.9728,
+      "step": 8519
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.692769169807434,
+      "learning_rate": 1.0853461803720809e-05,
+      "loss": 0.9564,
+      "step": 8520
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9231067895889282,
+      "learning_rate": 1.0851610955214701e-05,
+      "loss": 0.9771,
+      "step": 8521
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7166630029678345,
+      "learning_rate": 1.084976007732188e-05,
+      "loss": 0.8731,
+      "step": 8522
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.053574562072754,
+      "learning_rate": 1.0847909170106213e-05,
+      "loss": 0.6437,
+      "step": 8523
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8035495281219482,
+      "learning_rate": 1.0846058233631565e-05,
+      "loss": 1.0652,
+      "step": 8524
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7434369325637817,
+      "learning_rate": 1.0844207267961813e-05,
+      "loss": 1.0059,
+      "step": 8525
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9722028374671936,
+      "learning_rate": 1.0842356273160825e-05,
+      "loss": 0.5816,
+      "step": 8526
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.784459114074707,
+      "learning_rate": 1.0840505249292477e-05,
+      "loss": 0.9321,
+      "step": 8527
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.665932536125183,
+      "learning_rate": 1.083865419642064e-05,
+      "loss": 1.0443,
+      "step": 8528
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8101881742477417,
+      "learning_rate": 1.083680311460919e-05,
+      "loss": 0.9172,
+      "step": 8529
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.843503713607788,
+      "learning_rate": 1.0834952003922e-05,
+      "loss": 0.9846,
+      "step": 8530
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0145201683044434,
+      "learning_rate": 1.0833100864422952e-05,
+      "loss": 0.9449,
+      "step": 8531
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7863091230392456,
+      "learning_rate": 1.0831249696175918e-05,
+      "loss": 0.9597,
+      "step": 8532
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0934700965881348,
+      "learning_rate": 1.0829398499244781e-05,
+      "loss": 0.6095,
+      "step": 8533
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7591910362243652,
+      "learning_rate": 1.0827547273693418e-05,
+      "loss": 0.906,
+      "step": 8534
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6965126991271973,
+      "learning_rate": 1.0825696019585705e-05,
+      "loss": 0.9477,
+      "step": 8535
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7844785451889038,
+      "learning_rate": 1.0823844736985534e-05,
+      "loss": 0.9171,
+      "step": 8536
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.69741952419281,
+      "learning_rate": 1.0821993425956782e-05,
+      "loss": 1.0158,
+      "step": 8537
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.942112922668457,
+      "learning_rate": 1.0820142086563331e-05,
+      "loss": 0.9425,
+      "step": 8538
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5140103101730347,
+      "learning_rate": 1.0818290718869068e-05,
+      "loss": 0.89,
+      "step": 8539
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6266496181488037,
+      "learning_rate": 1.081643932293788e-05,
+      "loss": 0.9271,
+      "step": 8540
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7963783740997314,
+      "learning_rate": 1.0814587898833651e-05,
+      "loss": 0.9418,
+      "step": 8541
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9350452423095703,
+      "learning_rate": 1.0812736446620269e-05,
+      "loss": 0.933,
+      "step": 8542
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8321863412857056,
+      "learning_rate": 1.0810884966361624e-05,
+      "loss": 0.9576,
+      "step": 8543
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7120146751403809,
+      "learning_rate": 1.0809033458121603e-05,
+      "loss": 0.9496,
+      "step": 8544
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9368250370025635,
+      "learning_rate": 1.0807181921964096e-05,
+      "loss": 0.9965,
+      "step": 8545
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7208921909332275,
+      "learning_rate": 1.0805330357952996e-05,
+      "loss": 0.937,
+      "step": 8546
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5980980396270752,
+      "learning_rate": 1.0803478766152196e-05,
+      "loss": 0.9018,
+      "step": 8547
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7228633165359497,
+      "learning_rate": 1.0801627146625588e-05,
+      "loss": 0.9324,
+      "step": 8548
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8946014642715454,
+      "learning_rate": 1.0799775499437066e-05,
+      "loss": 1.0265,
+      "step": 8549
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8867732286453247,
+      "learning_rate": 1.0797923824650525e-05,
+      "loss": 0.9614,
+      "step": 8550
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8067049980163574,
+      "learning_rate": 1.0796072122329862e-05,
+      "loss": 1.0661,
+      "step": 8551
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8514765501022339,
+      "learning_rate": 1.0794220392538972e-05,
+      "loss": 0.9821,
+      "step": 8552
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7458326816558838,
+      "learning_rate": 1.079236863534176e-05,
+      "loss": 0.9568,
+      "step": 8553
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7729965448379517,
+      "learning_rate": 1.0790516850802115e-05,
+      "loss": 1.0427,
+      "step": 8554
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8417407274246216,
+      "learning_rate": 1.0788665038983942e-05,
+      "loss": 1.0374,
+      "step": 8555
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7935092449188232,
+      "learning_rate": 1.0786813199951145e-05,
+      "loss": 0.8952,
+      "step": 8556
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9512757062911987,
+      "learning_rate": 1.0784961333767621e-05,
+      "loss": 0.9476,
+      "step": 8557
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6605255603790283,
+      "learning_rate": 1.0783109440497273e-05,
+      "loss": 1.0046,
+      "step": 8558
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7901972532272339,
+      "learning_rate": 1.0781257520204004e-05,
+      "loss": 1.0459,
+      "step": 8559
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9524850845336914,
+      "learning_rate": 1.0779405572951724e-05,
+      "loss": 1.0617,
+      "step": 8560
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7421863079071045,
+      "learning_rate": 1.0777553598804333e-05,
+      "loss": 1.0059,
+      "step": 8561
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7468451261520386,
+      "learning_rate": 1.0775701597825741e-05,
+      "loss": 0.9311,
+      "step": 8562
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9935927391052246,
+      "learning_rate": 1.077384957007985e-05,
+      "loss": 0.5852,
+      "step": 8563
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6447386741638184,
+      "learning_rate": 1.0771997515630574e-05,
+      "loss": 0.9573,
+      "step": 8564
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7000644207000732,
+      "learning_rate": 1.077014543454182e-05,
+      "loss": 0.9478,
+      "step": 8565
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5797619819641113,
+      "learning_rate": 1.07682933268775e-05,
+      "loss": 1.0009,
+      "step": 8566
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6906930208206177,
+      "learning_rate": 1.0766441192701521e-05,
+      "loss": 0.9448,
+      "step": 8567
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6575133800506592,
+      "learning_rate": 1.0764589032077799e-05,
+      "loss": 0.9115,
+      "step": 8568
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.831507921218872,
+      "learning_rate": 1.0762736845070244e-05,
+      "loss": 0.9995,
+      "step": 8569
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8337693214416504,
+      "learning_rate": 1.0760884631742771e-05,
+      "loss": 0.9545,
+      "step": 8570
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.821090817451477,
+      "learning_rate": 1.0759032392159296e-05,
+      "loss": 0.9029,
+      "step": 8571
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9774340391159058,
+      "learning_rate": 1.0757180126383736e-05,
+      "loss": 1.0274,
+      "step": 8572
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7982439994812012,
+      "learning_rate": 1.0755327834480001e-05,
+      "loss": 0.9105,
+      "step": 8573
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7567111253738403,
+      "learning_rate": 1.0753475516512015e-05,
+      "loss": 0.9358,
+      "step": 8574
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.771444320678711,
+      "learning_rate": 1.0751623172543693e-05,
+      "loss": 1.0675,
+      "step": 8575
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5541614294052124,
+      "learning_rate": 1.0749770802638952e-05,
+      "loss": 1.0017,
+      "step": 8576
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.624999761581421,
+      "learning_rate": 1.0747918406861719e-05,
+      "loss": 0.893,
+      "step": 8577
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8379334211349487,
+      "learning_rate": 1.0746065985275907e-05,
+      "loss": 1.0735,
+      "step": 8578
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7358770370483398,
+      "learning_rate": 1.0744213537945444e-05,
+      "loss": 0.9421,
+      "step": 8579
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5639175176620483,
+      "learning_rate": 1.074236106493425e-05,
+      "loss": 0.9337,
+      "step": 8580
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.010817050933838,
+      "learning_rate": 1.074050856630625e-05,
+      "loss": 0.5791,
+      "step": 8581
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.046964406967163,
+      "learning_rate": 1.0738656042125368e-05,
+      "loss": 0.8862,
+      "step": 8582
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8380920886993408,
+      "learning_rate": 1.073680349245553e-05,
+      "loss": 1.0147,
+      "step": 8583
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7033231258392334,
+      "learning_rate": 1.0734950917360663e-05,
+      "loss": 0.9233,
+      "step": 8584
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7934552431106567,
+      "learning_rate": 1.073309831690469e-05,
+      "loss": 1.034,
+      "step": 8585
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7326147556304932,
+      "learning_rate": 1.0731245691151544e-05,
+      "loss": 0.9815,
+      "step": 8586
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6948521137237549,
+      "learning_rate": 1.072939304016515e-05,
+      "loss": 0.9377,
+      "step": 8587
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6737289428710938,
+      "learning_rate": 1.072754036400944e-05,
+      "loss": 0.9791,
+      "step": 8588
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.917029857635498,
+      "learning_rate": 1.0725687662748345e-05,
+      "loss": 1.0345,
+      "step": 8589
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5554600954055786,
+      "learning_rate": 1.0723834936445795e-05,
+      "loss": 0.946,
+      "step": 8590
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8516888618469238,
+      "learning_rate": 1.0721982185165723e-05,
+      "loss": 0.9756,
+      "step": 8591
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6908961534500122,
+      "learning_rate": 1.0720129408972063e-05,
+      "loss": 0.9606,
+      "step": 8592
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7380443811416626,
+      "learning_rate": 1.0718276607928751e-05,
+      "loss": 0.8969,
+      "step": 8593
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7556151151657104,
+      "learning_rate": 1.0716423782099716e-05,
+      "loss": 0.9836,
+      "step": 8594
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.11409330368042,
+      "learning_rate": 1.07145709315489e-05,
+      "loss": 0.6102,
+      "step": 8595
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0003728866577148,
+      "learning_rate": 1.0712718056340236e-05,
+      "loss": 0.5407,
+      "step": 8596
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0980660915374756,
+      "learning_rate": 1.0710865156537664e-05,
+      "loss": 0.9516,
+      "step": 8597
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6600666046142578,
+      "learning_rate": 1.070901223220512e-05,
+      "loss": 0.985,
+      "step": 8598
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.784712314605713,
+      "learning_rate": 1.0707159283406546e-05,
+      "loss": 0.9074,
+      "step": 8599
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8646495342254639,
+      "learning_rate": 1.0705306310205878e-05,
+      "loss": 1.0179,
+      "step": 8600
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7804954051971436,
+      "learning_rate": 1.070345331266706e-05,
+      "loss": 0.9179,
+      "step": 8601
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.810348629951477,
+      "learning_rate": 1.0701600290854032e-05,
+      "loss": 0.9391,
+      "step": 8602
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7820229530334473,
+      "learning_rate": 1.0699747244830742e-05,
+      "loss": 0.9815,
+      "step": 8603
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.737383246421814,
+      "learning_rate": 1.0697894174661128e-05,
+      "loss": 1.0452,
+      "step": 8604
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.803390383720398,
+      "learning_rate": 1.0696041080409132e-05,
+      "loss": 1.0465,
+      "step": 8605
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7518434524536133,
+      "learning_rate": 1.0694187962138705e-05,
+      "loss": 0.9799,
+      "step": 8606
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9417475461959839,
+      "learning_rate": 1.069233481991379e-05,
+      "loss": 0.9524,
+      "step": 8607
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.729644775390625,
+      "learning_rate": 1.0690481653798337e-05,
+      "loss": 0.9223,
+      "step": 8608
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8701280355453491,
+      "learning_rate": 1.0688628463856287e-05,
+      "loss": 1.0167,
+      "step": 8609
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5511900186538696,
+      "learning_rate": 1.0686775250151595e-05,
+      "loss": 0.9707,
+      "step": 8610
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8549238443374634,
+      "learning_rate": 1.0684922012748207e-05,
+      "loss": 1.0297,
+      "step": 8611
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6823989152908325,
+      "learning_rate": 1.0683068751710075e-05,
+      "loss": 0.9275,
+      "step": 8612
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.819419264793396,
+      "learning_rate": 1.0681215467101147e-05,
+      "loss": 1.0072,
+      "step": 8613
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7211225032806396,
+      "learning_rate": 1.0679362158985376e-05,
+      "loss": 0.9739,
+      "step": 8614
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9419435262680054,
+      "learning_rate": 1.0677508827426715e-05,
+      "loss": 1.0042,
+      "step": 8615
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0536625385284424,
+      "learning_rate": 1.0675655472489117e-05,
+      "loss": 0.6108,
+      "step": 8616
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.92514967918396,
+      "learning_rate": 1.0673802094236532e-05,
+      "loss": 0.9804,
+      "step": 8617
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9182807207107544,
+      "learning_rate": 1.0671948692732923e-05,
+      "loss": 0.9956,
+      "step": 8618
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9738763570785522,
+      "learning_rate": 1.0670095268042242e-05,
+      "loss": 0.6096,
+      "step": 8619
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7380576133728027,
+      "learning_rate": 1.0668241820228445e-05,
+      "loss": 0.9563,
+      "step": 8620
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.790464997291565,
+      "learning_rate": 1.0666388349355487e-05,
+      "loss": 1.0334,
+      "step": 8621
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8091809749603271,
+      "learning_rate": 1.0664534855487331e-05,
+      "loss": 1.034,
+      "step": 8622
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6292237043380737,
+      "learning_rate": 1.0662681338687932e-05,
+      "loss": 1.0229,
+      "step": 8623
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8130451440811157,
+      "learning_rate": 1.0660827799021253e-05,
+      "loss": 0.9698,
+      "step": 8624
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8048807382583618,
+      "learning_rate": 1.0658974236551252e-05,
+      "loss": 1.0049,
+      "step": 8625
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9106788635253906,
+      "learning_rate": 1.065712065134189e-05,
+      "loss": 0.9699,
+      "step": 8626
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6256356239318848,
+      "learning_rate": 1.065526704345713e-05,
+      "loss": 0.9665,
+      "step": 8627
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7158743143081665,
+      "learning_rate": 1.0653413412960936e-05,
+      "loss": 0.9371,
+      "step": 8628
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.830836296081543,
+      "learning_rate": 1.0651559759917266e-05,
+      "loss": 0.998,
+      "step": 8629
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.944166660308838,
+      "learning_rate": 1.0649706084390093e-05,
+      "loss": 0.9869,
+      "step": 8630
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.752475380897522,
+      "learning_rate": 1.0647852386443375e-05,
+      "loss": 0.9472,
+      "step": 8631
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.762967586517334,
+      "learning_rate": 1.0645998666141085e-05,
+      "loss": 0.9819,
+      "step": 8632
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8161859512329102,
+      "learning_rate": 1.0644144923547184e-05,
+      "loss": 0.8903,
+      "step": 8633
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8336495161056519,
+      "learning_rate": 1.0642291158725638e-05,
+      "loss": 0.9357,
+      "step": 8634
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7623893022537231,
+      "learning_rate": 1.064043737174042e-05,
+      "loss": 0.9326,
+      "step": 8635
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7667280435562134,
+      "learning_rate": 1.0638583562655498e-05,
+      "loss": 1.0291,
+      "step": 8636
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6274546384811401,
+      "learning_rate": 1.063672973153484e-05,
+      "loss": 0.9588,
+      "step": 8637
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7482718229293823,
+      "learning_rate": 1.0634875878442422e-05,
+      "loss": 1.0384,
+      "step": 8638
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7935895919799805,
+      "learning_rate": 1.0633022003442206e-05,
+      "loss": 0.9079,
+      "step": 8639
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.688215732574463,
+      "learning_rate": 1.0631168106598171e-05,
+      "loss": 1.0355,
+      "step": 8640
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7240748405456543,
+      "learning_rate": 1.0629314187974287e-05,
+      "loss": 1.0004,
+      "step": 8641
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7775052785873413,
+      "learning_rate": 1.0627460247634529e-05,
+      "loss": 1.0444,
+      "step": 8642
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7331091165542603,
+      "learning_rate": 1.062560628564287e-05,
+      "loss": 1.0176,
+      "step": 8643
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6832427978515625,
+      "learning_rate": 1.0623752302063284e-05,
+      "loss": 0.9782,
+      "step": 8644
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.744986653327942,
+      "learning_rate": 1.062189829695975e-05,
+      "loss": 0.9356,
+      "step": 8645
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8521672487258911,
+      "learning_rate": 1.0620044270396244e-05,
+      "loss": 0.8572,
+      "step": 8646
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6382694244384766,
+      "learning_rate": 1.0618190222436741e-05,
+      "loss": 0.9336,
+      "step": 8647
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0543371438980103,
+      "learning_rate": 1.0616336153145221e-05,
+      "loss": 0.6145,
+      "step": 8648
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0356029272079468,
+      "learning_rate": 1.061448206258566e-05,
+      "loss": 0.5438,
+      "step": 8649
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.735732078552246,
+      "learning_rate": 1.0612627950822044e-05,
+      "loss": 0.9816,
+      "step": 8650
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0616755485534668,
+      "learning_rate": 1.0610773817918346e-05,
+      "loss": 0.6212,
+      "step": 8651
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.567923665046692,
+      "learning_rate": 1.0608919663938549e-05,
+      "loss": 0.9941,
+      "step": 8652
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8313571214675903,
+      "learning_rate": 1.0607065488946635e-05,
+      "loss": 0.8924,
+      "step": 8653
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8070112466812134,
+      "learning_rate": 1.0605211293006587e-05,
+      "loss": 1.0449,
+      "step": 8654
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7310975790023804,
+      "learning_rate": 1.060335707618239e-05,
+      "loss": 0.9425,
+      "step": 8655
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.725441813468933,
+      "learning_rate": 1.0601502838538022e-05,
+      "loss": 0.8793,
+      "step": 8656
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7714563608169556,
+      "learning_rate": 1.0599648580137474e-05,
+      "loss": 0.9703,
+      "step": 8657
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7605868577957153,
+      "learning_rate": 1.0597794301044728e-05,
+      "loss": 1.0535,
+      "step": 8658
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8259702920913696,
+      "learning_rate": 1.0595940001323771e-05,
+      "loss": 1.0009,
+      "step": 8659
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6600556373596191,
+      "learning_rate": 1.0594085681038589e-05,
+      "loss": 0.941,
+      "step": 8660
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7088278532028198,
+      "learning_rate": 1.059223134025317e-05,
+      "loss": 0.9976,
+      "step": 8661
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7935503721237183,
+      "learning_rate": 1.0590376979031501e-05,
+      "loss": 0.9267,
+      "step": 8662
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.655744194984436,
+      "learning_rate": 1.0588522597437571e-05,
+      "loss": 1.0564,
+      "step": 8663
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.650330901145935,
+      "learning_rate": 1.0586668195535373e-05,
+      "loss": 0.9354,
+      "step": 8664
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7604529857635498,
+      "learning_rate": 1.058481377338889e-05,
+      "loss": 0.9867,
+      "step": 8665
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0636042356491089,
+      "learning_rate": 1.058295933106212e-05,
+      "loss": 0.5901,
+      "step": 8666
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.598160743713379,
+      "learning_rate": 1.058110486861905e-05,
+      "loss": 0.9588,
+      "step": 8667
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8154085874557495,
+      "learning_rate": 1.0579250386123676e-05,
+      "loss": 0.9962,
+      "step": 8668
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0499058961868286,
+      "learning_rate": 1.057739588363999e-05,
+      "loss": 0.5989,
+      "step": 8669
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7893593311309814,
+      "learning_rate": 1.0575541361231984e-05,
+      "loss": 0.9857,
+      "step": 8670
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9737614989280701,
+      "learning_rate": 1.0573686818963651e-05,
+      "loss": 0.5495,
+      "step": 8671
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6470822095870972,
+      "learning_rate": 1.0571832256898991e-05,
+      "loss": 0.8776,
+      "step": 8672
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8030527830123901,
+      "learning_rate": 1.0569977675101997e-05,
+      "loss": 1.0023,
+      "step": 8673
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.705561637878418,
+      "learning_rate": 1.0568123073636666e-05,
+      "loss": 0.9378,
+      "step": 8674
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7909302711486816,
+      "learning_rate": 1.0566268452566995e-05,
+      "loss": 0.9221,
+      "step": 8675
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9130544662475586,
+      "learning_rate": 1.056441381195698e-05,
+      "loss": 0.9481,
+      "step": 8676
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7799227237701416,
+      "learning_rate": 1.0562559151870621e-05,
+      "loss": 0.9503,
+      "step": 8677
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.95915687084198,
+      "learning_rate": 1.0560704472371919e-05,
+      "loss": 0.9859,
+      "step": 8678
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8908278942108154,
+      "learning_rate": 1.055884977352487e-05,
+      "loss": 0.9141,
+      "step": 8679
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6821506023406982,
+      "learning_rate": 1.055699505539348e-05,
+      "loss": 0.9237,
+      "step": 8680
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8918871879577637,
+      "learning_rate": 1.0555140318041743e-05,
+      "loss": 0.9715,
+      "step": 8681
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9311010837554932,
+      "learning_rate": 1.0553285561533664e-05,
+      "loss": 0.9606,
+      "step": 8682
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8428478240966797,
+      "learning_rate": 1.0551430785933246e-05,
+      "loss": 0.9243,
+      "step": 8683
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9872162342071533,
+      "learning_rate": 1.0549575991304493e-05,
+      "loss": 1.0045,
+      "step": 8684
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.168229341506958,
+      "learning_rate": 1.0547721177711407e-05,
+      "loss": 0.5894,
+      "step": 8685
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6577624082565308,
+      "learning_rate": 1.0545866345217994e-05,
+      "loss": 0.9368,
+      "step": 8686
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8130950927734375,
+      "learning_rate": 1.0544011493888258e-05,
+      "loss": 0.9924,
+      "step": 8687
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.862364411354065,
+      "learning_rate": 1.0542156623786206e-05,
+      "loss": 1.0334,
+      "step": 8688
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1388707160949707,
+      "learning_rate": 1.0540301734975842e-05,
+      "loss": 0.607,
+      "step": 8689
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9884395599365234,
+      "learning_rate": 1.0538446827521174e-05,
+      "loss": 0.5887,
+      "step": 8690
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.850339412689209,
+      "learning_rate": 1.053659190148621e-05,
+      "loss": 1.0648,
+      "step": 8691
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8652937412261963,
+      "learning_rate": 1.0534736956934962e-05,
+      "loss": 0.9676,
+      "step": 8692
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.992677927017212,
+      "learning_rate": 1.0532881993931432e-05,
+      "loss": 1.02,
+      "step": 8693
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7602018117904663,
+      "learning_rate": 1.0531027012539632e-05,
+      "loss": 0.9478,
+      "step": 8694
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.908563494682312,
+      "learning_rate": 1.0529172012823575e-05,
+      "loss": 0.9808,
+      "step": 8695
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.672378420829773,
+      "learning_rate": 1.0527316994847268e-05,
+      "loss": 0.9865,
+      "step": 8696
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.615802526473999,
+      "learning_rate": 1.0525461958674725e-05,
+      "loss": 1.0088,
+      "step": 8697
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0042924880981445,
+      "learning_rate": 1.0523606904369961e-05,
+      "loss": 1.0518,
+      "step": 8698
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7680364847183228,
+      "learning_rate": 1.0521751831996983e-05,
+      "loss": 0.9802,
+      "step": 8699
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9430030584335327,
+      "learning_rate": 1.0519896741619803e-05,
+      "loss": 0.9559,
+      "step": 8700
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7214672565460205,
+      "learning_rate": 1.0518041633302442e-05,
+      "loss": 1.0449,
+      "step": 8701
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.128002405166626,
+      "learning_rate": 1.0516186507108915e-05,
+      "loss": 0.6241,
+      "step": 8702
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.013005018234253,
+      "learning_rate": 1.051433136310323e-05,
+      "loss": 0.9865,
+      "step": 8703
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8287417888641357,
+      "learning_rate": 1.0512476201349407e-05,
+      "loss": 0.9167,
+      "step": 8704
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8643275499343872,
+      "learning_rate": 1.051062102191146e-05,
+      "loss": 0.9495,
+      "step": 8705
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7194923162460327,
+      "learning_rate": 1.0508765824853411e-05,
+      "loss": 1.0212,
+      "step": 8706
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0926965475082397,
+      "learning_rate": 1.0506910610239274e-05,
+      "loss": 0.5821,
+      "step": 8707
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8850185871124268,
+      "learning_rate": 1.0505055378133067e-05,
+      "loss": 1.1141,
+      "step": 8708
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7096799612045288,
+      "learning_rate": 1.050320012859881e-05,
+      "loss": 0.9563,
+      "step": 8709
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.082404613494873,
+      "learning_rate": 1.0501344861700518e-05,
+      "loss": 0.9406,
+      "step": 8710
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7201992273330688,
+      "learning_rate": 1.049948957750222e-05,
+      "loss": 0.9683,
+      "step": 8711
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7442508935928345,
+      "learning_rate": 1.0497634276067932e-05,
+      "loss": 0.9851,
+      "step": 8712
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8243485689163208,
+      "learning_rate": 1.0495778957461673e-05,
+      "loss": 0.9937,
+      "step": 8713
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6692544221878052,
+      "learning_rate": 1.0493923621747468e-05,
+      "loss": 0.9115,
+      "step": 8714
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.888960361480713,
+      "learning_rate": 1.0492068268989339e-05,
+      "loss": 1.0181,
+      "step": 8715
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6490732431411743,
+      "learning_rate": 1.0490212899251308e-05,
+      "loss": 1.0199,
+      "step": 8716
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7625213861465454,
+      "learning_rate": 1.0488357512597402e-05,
+      "loss": 1.017,
+      "step": 8717
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6850634813308716,
+      "learning_rate": 1.0486502109091639e-05,
+      "loss": 0.9494,
+      "step": 8718
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9551935195922852,
+      "learning_rate": 1.0484646688798049e-05,
+      "loss": 0.9527,
+      "step": 8719
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.760522484779358,
+      "learning_rate": 1.0482791251780655e-05,
+      "loss": 0.9768,
+      "step": 8720
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7895591259002686,
+      "learning_rate": 1.0480935798103485e-05,
+      "loss": 1.0364,
+      "step": 8721
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7450265884399414,
+      "learning_rate": 1.047908032783056e-05,
+      "loss": 0.9664,
+      "step": 8722
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7582743167877197,
+      "learning_rate": 1.0477224841025912e-05,
+      "loss": 0.9432,
+      "step": 8723
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8529181480407715,
+      "learning_rate": 1.047536933775357e-05,
+      "loss": 0.9086,
+      "step": 8724
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6644318103790283,
+      "learning_rate": 1.047351381807756e-05,
+      "loss": 0.9339,
+      "step": 8725
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.651681900024414,
+      "learning_rate": 1.0471658282061909e-05,
+      "loss": 1.0104,
+      "step": 8726
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7770278453826904,
+      "learning_rate": 1.0469802729770651e-05,
+      "loss": 0.9328,
+      "step": 8727
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6897521018981934,
+      "learning_rate": 1.0467947161267811e-05,
+      "loss": 0.9768,
+      "step": 8728
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.978057861328125,
+      "learning_rate": 1.0466091576617423e-05,
+      "loss": 0.9959,
+      "step": 8729
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6739064455032349,
+      "learning_rate": 1.0464235975883516e-05,
+      "loss": 0.9397,
+      "step": 8730
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8200616836547852,
+      "learning_rate": 1.0462380359130122e-05,
+      "loss": 0.9548,
+      "step": 8731
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8289320468902588,
+      "learning_rate": 1.0460524726421275e-05,
+      "loss": 1.0325,
+      "step": 8732
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6130201816558838,
+      "learning_rate": 1.0458669077821002e-05,
+      "loss": 0.9099,
+      "step": 8733
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.781553864479065,
+      "learning_rate": 1.0456813413393342e-05,
+      "loss": 0.9231,
+      "step": 8734
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9842435121536255,
+      "learning_rate": 1.0454957733202329e-05,
+      "loss": 0.9439,
+      "step": 8735
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.614741563796997,
+      "learning_rate": 1.0453102037311995e-05,
+      "loss": 0.9931,
+      "step": 8736
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9577792882919312,
+      "learning_rate": 1.0451246325786373e-05,
+      "loss": 0.9155,
+      "step": 8737
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.122384786605835,
+      "learning_rate": 1.0449390598689504e-05,
+      "loss": 0.6086,
+      "step": 8738
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.666574239730835,
+      "learning_rate": 1.044753485608542e-05,
+      "loss": 0.9682,
+      "step": 8739
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9014045000076294,
+      "learning_rate": 1.0445679098038158e-05,
+      "loss": 0.9088,
+      "step": 8740
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6517398357391357,
+      "learning_rate": 1.0443823324611754e-05,
+      "loss": 0.9677,
+      "step": 8741
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8512718677520752,
+      "learning_rate": 1.0441967535870248e-05,
+      "loss": 0.9181,
+      "step": 8742
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9636318683624268,
+      "learning_rate": 1.0440111731877678e-05,
+      "loss": 0.951,
+      "step": 8743
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8172403573989868,
+      "learning_rate": 1.043825591269808e-05,
+      "loss": 0.9628,
+      "step": 8744
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.697481632232666,
+      "learning_rate": 1.0436400078395497e-05,
+      "loss": 0.9462,
+      "step": 8745
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.811568021774292,
+      "learning_rate": 1.0434544229033964e-05,
+      "loss": 1.0213,
+      "step": 8746
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9566123485565186,
+      "learning_rate": 1.0432688364677523e-05,
+      "loss": 1.0229,
+      "step": 8747
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0337415933609009,
+      "learning_rate": 1.0430832485390217e-05,
+      "loss": 0.5674,
+      "step": 8748
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7874358892440796,
+      "learning_rate": 1.0428976591236082e-05,
+      "loss": 0.9048,
+      "step": 8749
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9547349214553833,
+      "learning_rate": 1.0427120682279166e-05,
+      "loss": 0.9517,
+      "step": 8750
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9897099733352661,
+      "learning_rate": 1.0425264758583509e-05,
+      "loss": 1.064,
+      "step": 8751
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0182952880859375,
+      "learning_rate": 1.0423408820213153e-05,
+      "loss": 0.9427,
+      "step": 8752
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8943119049072266,
+      "learning_rate": 1.0421552867232141e-05,
+      "loss": 0.9241,
+      "step": 8753
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8008861541748047,
+      "learning_rate": 1.0419696899704517e-05,
+      "loss": 0.9365,
+      "step": 8754
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5976773500442505,
+      "learning_rate": 1.0417840917694324e-05,
+      "loss": 0.9511,
+      "step": 8755
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7022970914840698,
+      "learning_rate": 1.041598492126561e-05,
+      "loss": 1.0267,
+      "step": 8756
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.725212812423706,
+      "learning_rate": 1.0414128910482417e-05,
+      "loss": 0.957,
+      "step": 8757
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7276993989944458,
+      "learning_rate": 1.0412272885408793e-05,
+      "loss": 0.9447,
+      "step": 8758
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.633353352546692,
+      "learning_rate": 1.0410416846108783e-05,
+      "loss": 0.9868,
+      "step": 8759
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7715717554092407,
+      "learning_rate": 1.0408560792646433e-05,
+      "loss": 0.9975,
+      "step": 8760
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9322177171707153,
+      "learning_rate": 1.0406704725085792e-05,
+      "loss": 0.9184,
+      "step": 8761
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9761765003204346,
+      "learning_rate": 1.0404848643490908e-05,
+      "loss": 1.0402,
+      "step": 8762
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7888035774230957,
+      "learning_rate": 1.0402992547925827e-05,
+      "loss": 0.9688,
+      "step": 8763
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7115849256515503,
+      "learning_rate": 1.04011364384546e-05,
+      "loss": 0.9447,
+      "step": 8764
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8671318292617798,
+      "learning_rate": 1.0399280315141275e-05,
+      "loss": 0.9366,
+      "step": 8765
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0708019733428955,
+      "learning_rate": 1.03974241780499e-05,
+      "loss": 0.6696,
+      "step": 8766
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.826255440711975,
+      "learning_rate": 1.0395568027244527e-05,
+      "loss": 0.9416,
+      "step": 8767
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5463672876358032,
+      "learning_rate": 1.0393711862789209e-05,
+      "loss": 0.9357,
+      "step": 8768
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9841461181640625,
+      "learning_rate": 1.039185568474799e-05,
+      "loss": 1.0082,
+      "step": 8769
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7973681688308716,
+      "learning_rate": 1.038999949318493e-05,
+      "loss": 0.9111,
+      "step": 8770
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6945699453353882,
+      "learning_rate": 1.0388143288164077e-05,
+      "loss": 0.9624,
+      "step": 8771
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7679191827774048,
+      "learning_rate": 1.038628706974948e-05,
+      "loss": 0.8842,
+      "step": 8772
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7038698196411133,
+      "learning_rate": 1.03844308380052e-05,
+      "loss": 0.9593,
+      "step": 8773
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.725569248199463,
+      "learning_rate": 1.0382574592995283e-05,
+      "loss": 1.021,
+      "step": 8774
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8513617515563965,
+      "learning_rate": 1.0380718334783785e-05,
+      "loss": 1.0249,
+      "step": 8775
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7253975868225098,
+      "learning_rate": 1.0378862063434757e-05,
+      "loss": 0.982,
+      "step": 8776
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7462536096572876,
+      "learning_rate": 1.0377005779012264e-05,
+      "loss": 0.976,
+      "step": 8777
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8043029308319092,
+      "learning_rate": 1.0375149481580352e-05,
+      "loss": 0.9873,
+      "step": 8778
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7627674341201782,
+      "learning_rate": 1.037329317120308e-05,
+      "loss": 1.0279,
+      "step": 8779
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6667206287384033,
+      "learning_rate": 1.0371436847944503e-05,
+      "loss": 0.8473,
+      "step": 8780
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0618886947631836,
+      "learning_rate": 1.0369580511868678e-05,
+      "loss": 1.0151,
+      "step": 8781
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0162452459335327,
+      "learning_rate": 1.0367724163039663e-05,
+      "loss": 0.5428,
+      "step": 8782
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.601648211479187,
+      "learning_rate": 1.0365867801521515e-05,
+      "loss": 0.9997,
+      "step": 8783
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6492953300476074,
+      "learning_rate": 1.036401142737829e-05,
+      "loss": 0.9617,
+      "step": 8784
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7361164093017578,
+      "learning_rate": 1.0362155040674045e-05,
+      "loss": 1.0498,
+      "step": 8785
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9997910261154175,
+      "learning_rate": 1.0360298641472843e-05,
+      "loss": 0.9897,
+      "step": 8786
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9723209142684937,
+      "learning_rate": 1.0358442229838742e-05,
+      "loss": 0.9892,
+      "step": 8787
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7785959243774414,
+      "learning_rate": 1.0356585805835796e-05,
+      "loss": 0.9454,
+      "step": 8788
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.896394968032837,
+      "learning_rate": 1.0354729369528076e-05,
+      "loss": 0.9076,
+      "step": 8789
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7003003358840942,
+      "learning_rate": 1.0352872920979636e-05,
+      "loss": 0.9885,
+      "step": 8790
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.789709210395813,
+      "learning_rate": 1.0351016460254536e-05,
+      "loss": 0.9294,
+      "step": 8791
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.708404541015625,
+      "learning_rate": 1.0349159987416837e-05,
+      "loss": 0.9278,
+      "step": 8792
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8927345275878906,
+      "learning_rate": 1.0347303502530605e-05,
+      "loss": 1.0119,
+      "step": 8793
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6412158012390137,
+      "learning_rate": 1.0345447005659897e-05,
+      "loss": 0.8636,
+      "step": 8794
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8670347929000854,
+      "learning_rate": 1.0343590496868778e-05,
+      "loss": 0.9322,
+      "step": 8795
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6055206060409546,
+      "learning_rate": 1.0341733976221313e-05,
+      "loss": 0.8791,
+      "step": 8796
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8486522436141968,
+      "learning_rate": 1.033987744378156e-05,
+      "loss": 0.9328,
+      "step": 8797
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7057690620422363,
+      "learning_rate": 1.0338020899613588e-05,
+      "loss": 1.021,
+      "step": 8798
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9186482429504395,
+      "learning_rate": 1.0336164343781457e-05,
+      "loss": 1.0249,
+      "step": 8799
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8854731321334839,
+      "learning_rate": 1.0334307776349235e-05,
+      "loss": 0.9576,
+      "step": 8800
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.795735478401184,
+      "learning_rate": 1.0332451197380987e-05,
+      "loss": 0.9984,
+      "step": 8801
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6469866037368774,
+      "learning_rate": 1.0330594606940773e-05,
+      "loss": 0.9222,
+      "step": 8802
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6200100183486938,
+      "learning_rate": 1.0328738005092662e-05,
+      "loss": 0.9542,
+      "step": 8803
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.627124547958374,
+      "learning_rate": 1.0326881391900726e-05,
+      "loss": 0.9727,
+      "step": 8804
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6503161191940308,
+      "learning_rate": 1.0325024767429022e-05,
+      "loss": 0.9748,
+      "step": 8805
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7582732439041138,
+      "learning_rate": 1.0323168131741623e-05,
+      "loss": 0.9398,
+      "step": 8806
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.88052499294281,
+      "learning_rate": 1.0321311484902594e-05,
+      "loss": 1.0136,
+      "step": 8807
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.01294207572937,
+      "learning_rate": 1.0319454826976006e-05,
+      "loss": 1.0337,
+      "step": 8808
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7034651041030884,
+      "learning_rate": 1.0317598158025921e-05,
+      "loss": 0.9511,
+      "step": 8809
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7480237483978271,
+      "learning_rate": 1.0315741478116413e-05,
+      "loss": 0.9574,
+      "step": 8810
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6877936124801636,
+      "learning_rate": 1.0313884787311545e-05,
+      "loss": 0.9429,
+      "step": 8811
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7521706819534302,
+      "learning_rate": 1.0312028085675393e-05,
+      "loss": 0.99,
+      "step": 8812
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7182331085205078,
+      "learning_rate": 1.0310171373272021e-05,
+      "loss": 0.9597,
+      "step": 8813
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8290330171585083,
+      "learning_rate": 1.0308314650165505e-05,
+      "loss": 0.9407,
+      "step": 8814
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7604749202728271,
+      "learning_rate": 1.0306457916419907e-05,
+      "loss": 0.9294,
+      "step": 8815
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5807100534439087,
+      "learning_rate": 1.0304601172099304e-05,
+      "loss": 1.0314,
+      "step": 8816
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7661210298538208,
+      "learning_rate": 1.0302744417267767e-05,
+      "loss": 0.9776,
+      "step": 8817
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.787975549697876,
+      "learning_rate": 1.0300887651989363e-05,
+      "loss": 1.0192,
+      "step": 8818
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6769720315933228,
+      "learning_rate": 1.029903087632817e-05,
+      "loss": 0.9379,
+      "step": 8819
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9327137470245361,
+      "learning_rate": 1.0297174090348257e-05,
+      "loss": 0.8904,
+      "step": 8820
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7335028648376465,
+      "learning_rate": 1.0295317294113694e-05,
+      "loss": 0.9949,
+      "step": 8821
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7678446769714355,
+      "learning_rate": 1.0293460487688557e-05,
+      "loss": 1.0138,
+      "step": 8822
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6729885339736938,
+      "learning_rate": 1.0291603671136918e-05,
+      "loss": 0.881,
+      "step": 8823
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8116122484207153,
+      "learning_rate": 1.0289746844522851e-05,
+      "loss": 0.9003,
+      "step": 8824
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.689509630203247,
+      "learning_rate": 1.028789000791043e-05,
+      "loss": 0.9123,
+      "step": 8825
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.691733479499817,
+      "learning_rate": 1.0286033161363728e-05,
+      "loss": 0.9698,
+      "step": 8826
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8427883386611938,
+      "learning_rate": 1.0284176304946823e-05,
+      "loss": 0.9147,
+      "step": 8827
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6397954225540161,
+      "learning_rate": 1.0282319438723783e-05,
+      "loss": 1.0396,
+      "step": 8828
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9507653713226318,
+      "learning_rate": 1.028046256275869e-05,
+      "loss": 0.8731,
+      "step": 8829
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4024715423583984,
+      "learning_rate": 1.0278605677115618e-05,
+      "loss": 0.9398,
+      "step": 8830
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.860568881034851,
+      "learning_rate": 1.0276748781858643e-05,
+      "loss": 1.0037,
+      "step": 8831
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1306332349777222,
+      "learning_rate": 1.0274891877051838e-05,
+      "loss": 0.619,
+      "step": 8832
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7977855205535889,
+      "learning_rate": 1.0273034962759286e-05,
+      "loss": 0.9487,
+      "step": 8833
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.559626817703247,
+      "learning_rate": 1.0271178039045058e-05,
+      "loss": 0.9163,
+      "step": 8834
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.644394874572754,
+      "learning_rate": 1.0269321105973233e-05,
+      "loss": 0.9621,
+      "step": 8835
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7126811742782593,
+      "learning_rate": 1.026746416360789e-05,
+      "loss": 0.9586,
+      "step": 8836
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6614716053009033,
+      "learning_rate": 1.0265607212013107e-05,
+      "loss": 0.9819,
+      "step": 8837
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6473829746246338,
+      "learning_rate": 1.026375025125296e-05,
+      "loss": 0.9725,
+      "step": 8838
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7341830730438232,
+      "learning_rate": 1.0261893281391526e-05,
+      "loss": 0.959,
+      "step": 8839
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8899998664855957,
+      "learning_rate": 1.026003630249289e-05,
+      "loss": 1.114,
+      "step": 8840
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.821725845336914,
+      "learning_rate": 1.0258179314621125e-05,
+      "loss": 0.955,
+      "step": 8841
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7690879106521606,
+      "learning_rate": 1.0256322317840313e-05,
+      "loss": 0.9138,
+      "step": 8842
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8985460996627808,
+      "learning_rate": 1.0254465312214534e-05,
+      "loss": 0.9087,
+      "step": 8843
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.691991925239563,
+      "learning_rate": 1.0252608297807871e-05,
+      "loss": 1.0119,
+      "step": 8844
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9804728031158447,
+      "learning_rate": 1.0250751274684399e-05,
+      "loss": 0.976,
+      "step": 8845
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.195158004760742,
+      "learning_rate": 1.02488942429082e-05,
+      "loss": 1.0353,
+      "step": 8846
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6387323141098022,
+      "learning_rate": 1.0247037202543357e-05,
+      "loss": 0.9196,
+      "step": 8847
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6925315856933594,
+      "learning_rate": 1.024518015365395e-05,
+      "loss": 0.9938,
+      "step": 8848
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6406574249267578,
+      "learning_rate": 1.0243323096304063e-05,
+      "loss": 0.9232,
+      "step": 8849
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7291712760925293,
+      "learning_rate": 1.0241466030557775e-05,
+      "loss": 0.9431,
+      "step": 8850
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7111749649047852,
+      "learning_rate": 1.0239608956479165e-05,
+      "loss": 1.0127,
+      "step": 8851
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.795896291732788,
+      "learning_rate": 1.0237751874132323e-05,
+      "loss": 0.9374,
+      "step": 8852
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9535866975784302,
+      "learning_rate": 1.0235894783581328e-05,
+      "loss": 1.0129,
+      "step": 8853
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9126861095428467,
+      "learning_rate": 1.0234037684890258e-05,
+      "loss": 0.9117,
+      "step": 8854
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7033377885818481,
+      "learning_rate": 1.0232180578123206e-05,
+      "loss": 0.8874,
+      "step": 8855
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0191949605941772,
+      "learning_rate": 1.023032346334425e-05,
+      "loss": 0.602,
+      "step": 8856
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.818963646888733,
+      "learning_rate": 1.0228466340617473e-05,
+      "loss": 0.9796,
+      "step": 8857
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.814379334449768,
+      "learning_rate": 1.0226609210006963e-05,
+      "loss": 1.0384,
+      "step": 8858
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7110366821289062,
+      "learning_rate": 1.0224752071576803e-05,
+      "loss": 0.949,
+      "step": 8859
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.020367383956909,
+      "learning_rate": 1.0222894925391074e-05,
+      "loss": 0.9004,
+      "step": 8860
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9149428606033325,
+      "learning_rate": 1.0221037771513867e-05,
+      "loss": 1.0647,
+      "step": 8861
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6878336668014526,
+      "learning_rate": 1.021918061000926e-05,
+      "loss": 0.9423,
+      "step": 8862
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0334285497665405,
+      "learning_rate": 1.0217323440941345e-05,
+      "loss": 0.5866,
+      "step": 8863
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7308261394500732,
+      "learning_rate": 1.0215466264374205e-05,
+      "loss": 0.9525,
+      "step": 8864
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5370523929595947,
+      "learning_rate": 1.0213609080371922e-05,
+      "loss": 0.9416,
+      "step": 8865
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9030011892318726,
+      "learning_rate": 1.0211751888998592e-05,
+      "loss": 1.0226,
+      "step": 8866
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6708587408065796,
+      "learning_rate": 1.0209894690318293e-05,
+      "loss": 0.9322,
+      "step": 8867
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.767116665840149,
+      "learning_rate": 1.0208037484395114e-05,
+      "loss": 0.9616,
+      "step": 8868
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7106602191925049,
+      "learning_rate": 1.0206180271293143e-05,
+      "loss": 0.927,
+      "step": 8869
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7157697677612305,
+      "learning_rate": 1.0204323051076467e-05,
+      "loss": 1.0525,
+      "step": 8870
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6959165334701538,
+      "learning_rate": 1.0202465823809176e-05,
+      "loss": 0.967,
+      "step": 8871
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.992066502571106,
+      "learning_rate": 1.0200608589555352e-05,
+      "loss": 0.9559,
+      "step": 8872
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.719935655593872,
+      "learning_rate": 1.0198751348379085e-05,
+      "loss": 0.9949,
+      "step": 8873
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9515131711959839,
+      "learning_rate": 1.0196894100344467e-05,
+      "loss": 0.9402,
+      "step": 8874
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7132283449172974,
+      "learning_rate": 1.0195036845515583e-05,
+      "loss": 0.9366,
+      "step": 8875
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.781367301940918,
+      "learning_rate": 1.0193179583956523e-05,
+      "loss": 0.9035,
+      "step": 8876
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.778655767440796,
+      "learning_rate": 1.0191322315731374e-05,
+      "loss": 0.9026,
+      "step": 8877
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7946217060089111,
+      "learning_rate": 1.0189465040904224e-05,
+      "loss": 1.0595,
+      "step": 8878
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7899078130722046,
+      "learning_rate": 1.0187607759539168e-05,
+      "loss": 0.9926,
+      "step": 8879
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6764757633209229,
+      "learning_rate": 1.0185750471700293e-05,
+      "loss": 0.9681,
+      "step": 8880
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.797339916229248,
+      "learning_rate": 1.0183893177451683e-05,
+      "loss": 0.9043,
+      "step": 8881
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9926712512969971,
+      "learning_rate": 1.0182035876857437e-05,
+      "loss": 0.5918,
+      "step": 8882
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9849005937576294,
+      "learning_rate": 1.0180178569981641e-05,
+      "loss": 1.024,
+      "step": 8883
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.123988389968872,
+      "learning_rate": 1.0178321256888386e-05,
+      "loss": 0.9618,
+      "step": 8884
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7594813108444214,
+      "learning_rate": 1.0176463937641763e-05,
+      "loss": 1.0231,
+      "step": 8885
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1616874933242798,
+      "learning_rate": 1.0174606612305863e-05,
+      "loss": 0.6022,
+      "step": 8886
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6073048114776611,
+      "learning_rate": 1.0172749280944773e-05,
+      "loss": 0.9811,
+      "step": 8887
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6909401416778564,
+      "learning_rate": 1.0170891943622593e-05,
+      "loss": 0.9614,
+      "step": 8888
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6995933055877686,
+      "learning_rate": 1.0169034600403404e-05,
+      "loss": 1.0046,
+      "step": 8889
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7703731060028076,
+      "learning_rate": 1.0167177251351305e-05,
+      "loss": 0.9076,
+      "step": 8890
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0440077781677246,
+      "learning_rate": 1.0165319896530388e-05,
+      "loss": 0.9302,
+      "step": 8891
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8103872537612915,
+      "learning_rate": 1.0163462536004742e-05,
+      "loss": 0.8734,
+      "step": 8892
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7994543313980103,
+      "learning_rate": 1.0161605169838459e-05,
+      "loss": 0.9685,
+      "step": 8893
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6386631727218628,
+      "learning_rate": 1.0159747798095635e-05,
+      "loss": 0.9588,
+      "step": 8894
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.087648391723633,
+      "learning_rate": 1.015789042084036e-05,
+      "loss": 1.0591,
+      "step": 8895
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7218666076660156,
+      "learning_rate": 1.0156033038136728e-05,
+      "loss": 1.0145,
+      "step": 8896
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7784907817840576,
+      "learning_rate": 1.015417565004883e-05,
+      "loss": 0.9369,
+      "step": 8897
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7313958406448364,
+      "learning_rate": 1.0152318256640761e-05,
+      "loss": 0.922,
+      "step": 8898
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7791234254837036,
+      "learning_rate": 1.0150460857976616e-05,
+      "loss": 1.0624,
+      "step": 8899
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7113159894943237,
+      "learning_rate": 1.0148603454120487e-05,
+      "loss": 0.9925,
+      "step": 8900
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7801169157028198,
+      "learning_rate": 1.0146746045136468e-05,
+      "loss": 0.9928,
+      "step": 8901
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9388582706451416,
+      "learning_rate": 1.0144888631088652e-05,
+      "loss": 0.9619,
+      "step": 8902
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.879156231880188,
+      "learning_rate": 1.0143031212041136e-05,
+      "loss": 0.9727,
+      "step": 8903
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6208269596099854,
+      "learning_rate": 1.0141173788058012e-05,
+      "loss": 0.9769,
+      "step": 8904
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.333559989929199,
+      "learning_rate": 1.0139316359203373e-05,
+      "loss": 1.0546,
+      "step": 8905
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6685092449188232,
+      "learning_rate": 1.0137458925541317e-05,
+      "loss": 0.9691,
+      "step": 8906
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7328840494155884,
+      "learning_rate": 1.0135601487135937e-05,
+      "loss": 0.9246,
+      "step": 8907
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7420519590377808,
+      "learning_rate": 1.0133744044051329e-05,
+      "loss": 0.9754,
+      "step": 8908
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6979494094848633,
+      "learning_rate": 1.0131886596351585e-05,
+      "loss": 1.0703,
+      "step": 8909
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7054520845413208,
+      "learning_rate": 1.0130029144100806e-05,
+      "loss": 0.915,
+      "step": 8910
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7403396368026733,
+      "learning_rate": 1.0128171687363084e-05,
+      "loss": 0.9798,
+      "step": 8911
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8815827369689941,
+      "learning_rate": 1.0126314226202517e-05,
+      "loss": 0.9809,
+      "step": 8912
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6624183654785156,
+      "learning_rate": 1.0124456760683194e-05,
+      "loss": 0.9823,
+      "step": 8913
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.774604082107544,
+      "learning_rate": 1.012259929086922e-05,
+      "loss": 0.9673,
+      "step": 8914
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7927039861679077,
+      "learning_rate": 1.0120741816824686e-05,
+      "loss": 0.9381,
+      "step": 8915
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8010519742965698,
+      "learning_rate": 1.0118884338613688e-05,
+      "loss": 0.9378,
+      "step": 8916
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8775206804275513,
+      "learning_rate": 1.0117026856300326e-05,
+      "loss": 0.9342,
+      "step": 8917
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7072855234146118,
+      "learning_rate": 1.0115169369948692e-05,
+      "loss": 1.0085,
+      "step": 8918
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9570808410644531,
+      "learning_rate": 1.0113311879622884e-05,
+      "loss": 0.955,
+      "step": 8919
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.76759934425354,
+      "learning_rate": 1.0111454385387001e-05,
+      "loss": 0.9547,
+      "step": 8920
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7058250904083252,
+      "learning_rate": 1.0109596887305137e-05,
+      "loss": 1.056,
+      "step": 8921
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7628761529922485,
+      "learning_rate": 1.010773938544139e-05,
+      "loss": 0.9218,
+      "step": 8922
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9185079336166382,
+      "learning_rate": 1.0105881879859862e-05,
+      "loss": 0.9718,
+      "step": 8923
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6816920042037964,
+      "learning_rate": 1.0104024370624644e-05,
+      "loss": 0.8992,
+      "step": 8924
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9165916442871094,
+      "learning_rate": 1.0102166857799835e-05,
+      "loss": 1.0513,
+      "step": 8925
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7411118745803833,
+      "learning_rate": 1.0100309341449532e-05,
+      "loss": 0.9157,
+      "step": 8926
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8978520631790161,
+      "learning_rate": 1.0098451821637837e-05,
+      "loss": 0.9807,
+      "step": 8927
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7074702978134155,
+      "learning_rate": 1.0096594298428841e-05,
+      "loss": 0.9944,
+      "step": 8928
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.689185619354248,
+      "learning_rate": 1.009473677188665e-05,
+      "loss": 0.9556,
+      "step": 8929
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8243353366851807,
+      "learning_rate": 1.0092879242075352e-05,
+      "loss": 0.9741,
+      "step": 8930
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7011597156524658,
+      "learning_rate": 1.0091021709059054e-05,
+      "loss": 0.9364,
+      "step": 8931
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8377373218536377,
+      "learning_rate": 1.008916417290185e-05,
+      "loss": 0.9692,
+      "step": 8932
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6834347248077393,
+      "learning_rate": 1.0087306633667842e-05,
+      "loss": 0.9618,
+      "step": 8933
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.777701735496521,
+      "learning_rate": 1.0085449091421124e-05,
+      "loss": 0.9922,
+      "step": 8934
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.694658637046814,
+      "learning_rate": 1.0083591546225794e-05,
+      "loss": 0.8588,
+      "step": 8935
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8077694177627563,
+      "learning_rate": 1.0081733998145957e-05,
+      "loss": 0.8945,
+      "step": 8936
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8565845489501953,
+      "learning_rate": 1.0079876447245706e-05,
+      "loss": 0.9706,
+      "step": 8937
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8217909336090088,
+      "learning_rate": 1.0078018893589142e-05,
+      "loss": 1.0318,
+      "step": 8938
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1791296005249023,
+      "learning_rate": 1.0076161337240366e-05,
+      "loss": 0.9221,
+      "step": 8939
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.892891764640808,
+      "learning_rate": 1.0074303778263475e-05,
+      "loss": 1.0535,
+      "step": 8940
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0799294710159302,
+      "learning_rate": 1.0072446216722566e-05,
+      "loss": 0.5249,
+      "step": 8941
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8259260654449463,
+      "learning_rate": 1.0070588652681743e-05,
+      "loss": 0.9855,
+      "step": 8942
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9901882410049438,
+      "learning_rate": 1.0068731086205102e-05,
+      "loss": 0.963,
+      "step": 8943
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8897918462753296,
+      "learning_rate": 1.0066873517356743e-05,
+      "loss": 0.9397,
+      "step": 8944
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9522923231124878,
+      "learning_rate": 1.0065015946200765e-05,
+      "loss": 1.0194,
+      "step": 8945
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6233583688735962,
+      "learning_rate": 1.006315837280127e-05,
+      "loss": 0.975,
+      "step": 8946
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6960440874099731,
+      "learning_rate": 1.0061300797222351e-05,
+      "loss": 0.9465,
+      "step": 8947
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8390296697616577,
+      "learning_rate": 1.0059443219528117e-05,
+      "loss": 0.909,
+      "step": 8948
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7989524602890015,
+      "learning_rate": 1.0057585639782663e-05,
+      "loss": 0.9638,
+      "step": 8949
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7077968120574951,
+      "learning_rate": 1.005572805805009e-05,
+      "loss": 0.9626,
+      "step": 8950
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.977864146232605,
+      "learning_rate": 1.0053870474394495e-05,
+      "loss": 0.9194,
+      "step": 8951
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7608526945114136,
+      "learning_rate": 1.0052012888879982e-05,
+      "loss": 0.9273,
+      "step": 8952
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.054431438446045,
+      "learning_rate": 1.0050155301570652e-05,
+      "loss": 0.5972,
+      "step": 8953
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8544378280639648,
+      "learning_rate": 1.0048297712530599e-05,
+      "loss": 0.9707,
+      "step": 8954
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8765212297439575,
+      "learning_rate": 1.0046440121823928e-05,
+      "loss": 0.9882,
+      "step": 8955
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.73099946975708,
+      "learning_rate": 1.0044582529514739e-05,
+      "loss": 0.9418,
+      "step": 8956
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.791661262512207,
+      "learning_rate": 1.0042724935667132e-05,
+      "loss": 0.9792,
+      "step": 8957
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7827130556106567,
+      "learning_rate": 1.0040867340345204e-05,
+      "loss": 0.8902,
+      "step": 8958
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8416775465011597,
+      "learning_rate": 1.003900974361306e-05,
+      "loss": 0.9386,
+      "step": 8959
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8485275506973267,
+      "learning_rate": 1.0037152145534797e-05,
+      "loss": 0.9101,
+      "step": 8960
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.318046808242798,
+      "learning_rate": 1.0035294546174519e-05,
+      "loss": 1.0374,
+      "step": 8961
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6909818649291992,
+      "learning_rate": 1.0033436945596325e-05,
+      "loss": 0.9476,
+      "step": 8962
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8507784605026245,
+      "learning_rate": 1.0031579343864316e-05,
+      "loss": 0.9463,
+      "step": 8963
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.888702630996704,
+      "learning_rate": 1.0029721741042587e-05,
+      "loss": 0.9336,
+      "step": 8964
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8669872283935547,
+      "learning_rate": 1.0027864137195247e-05,
+      "loss": 0.958,
+      "step": 8965
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6972261667251587,
+      "learning_rate": 1.0026006532386394e-05,
+      "loss": 0.8913,
+      "step": 8966
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8669794797897339,
+      "learning_rate": 1.002414892668013e-05,
+      "loss": 0.9802,
+      "step": 8967
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6567277908325195,
+      "learning_rate": 1.0022291320140552e-05,
+      "loss": 0.9698,
+      "step": 8968
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8291791677474976,
+      "learning_rate": 1.0020433712831763e-05,
+      "loss": 1.0483,
+      "step": 8969
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7178864479064941,
+      "learning_rate": 1.0018576104817866e-05,
+      "loss": 1.0496,
+      "step": 8970
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6637693643569946,
+      "learning_rate": 1.0016718496162957e-05,
+      "loss": 0.9756,
+      "step": 8971
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7813897132873535,
+      "learning_rate": 1.0014860886931138e-05,
+      "loss": 0.941,
+      "step": 8972
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6179001331329346,
+      "learning_rate": 1.0013003277186513e-05,
+      "loss": 0.9509,
+      "step": 8973
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.680076003074646,
+      "learning_rate": 1.001114566699318e-05,
+      "loss": 0.9895,
+      "step": 8974
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8443573713302612,
+      "learning_rate": 1.0009288056415243e-05,
+      "loss": 1.0379,
+      "step": 8975
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7588536739349365,
+      "learning_rate": 1.00074304455168e-05,
+      "loss": 0.9279,
+      "step": 8976
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6640065908432007,
+      "learning_rate": 1.0005572834361954e-05,
+      "loss": 0.8448,
+      "step": 8977
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8051307201385498,
+      "learning_rate": 1.0003715223014805e-05,
+      "loss": 0.9805,
+      "step": 8978
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9108141660690308,
+      "learning_rate": 1.0001857611539454e-05,
+      "loss": 0.9759,
+      "step": 8979
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.000741958618164,
+      "learning_rate": 1e-05,
+      "loss": 0.9706,
+      "step": 8980
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8142032623291016,
+      "learning_rate": 9.99814238846055e-06,
+      "loss": 1.072,
+      "step": 8981
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7238868474960327,
+      "learning_rate": 9.996284776985199e-06,
+      "loss": 0.9755,
+      "step": 8982
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7324669361114502,
+      "learning_rate": 9.99442716563805e-06,
+      "loss": 0.9509,
+      "step": 8983
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5926146507263184,
+      "learning_rate": 9.992569554483202e-06,
+      "loss": 0.9385,
+      "step": 8984
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5944912433624268,
+      "learning_rate": 9.99071194358476e-06,
+      "loss": 0.9183,
+      "step": 8985
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9157428741455078,
+      "learning_rate": 9.988854333006823e-06,
+      "loss": 0.9524,
+      "step": 8986
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0344617366790771,
+      "learning_rate": 9.986996722813489e-06,
+      "loss": 0.5842,
+      "step": 8987
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.112147569656372,
+      "learning_rate": 9.985139113068865e-06,
+      "loss": 0.9307,
+      "step": 8988
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.874718427658081,
+      "learning_rate": 9.983281503837047e-06,
+      "loss": 0.9547,
+      "step": 8989
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.783829689025879,
+      "learning_rate": 9.981423895182139e-06,
+      "loss": 1.0163,
+      "step": 8990
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0992252826690674,
+      "learning_rate": 9.97956628716824e-06,
+      "loss": 0.996,
+      "step": 8991
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7792052030563354,
+      "learning_rate": 9.977708679859451e-06,
+      "loss": 1.0038,
+      "step": 8992
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7909373044967651,
+      "learning_rate": 9.975851073319873e-06,
+      "loss": 0.9234,
+      "step": 8993
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.96729576587677,
+      "learning_rate": 9.973993467613607e-06,
+      "loss": 1.0248,
+      "step": 8994
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6747527122497559,
+      "learning_rate": 9.972135862804755e-06,
+      "loss": 0.8754,
+      "step": 8995
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1096645593643188,
+      "learning_rate": 9.970278258957415e-06,
+      "loss": 0.62,
+      "step": 8996
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9389084577560425,
+      "learning_rate": 9.968420656135691e-06,
+      "loss": 0.9249,
+      "step": 8997
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.842962384223938,
+      "learning_rate": 9.96656305440368e-06,
+      "loss": 0.9776,
+      "step": 8998
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7134102582931519,
+      "learning_rate": 9.964705453825484e-06,
+      "loss": 0.8636,
+      "step": 8999
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9879229068756104,
+      "learning_rate": 9.962847854465206e-06,
+      "loss": 0.9611,
+      "step": 9000
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.693435549736023,
+      "learning_rate": 9.960990256386944e-06,
+      "loss": 0.8929,
+      "step": 9001
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.752920389175415,
+      "learning_rate": 9.9591326596548e-06,
+      "loss": 0.8911,
+      "step": 9002
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.840528130531311,
+      "learning_rate": 9.957275064332874e-06,
+      "loss": 0.9188,
+      "step": 9003
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6822412014007568,
+      "learning_rate": 9.955417470485267e-06,
+      "loss": 1.0313,
+      "step": 9004
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6589536666870117,
+      "learning_rate": 9.953559878176077e-06,
+      "loss": 0.9199,
+      "step": 9005
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7368932962417603,
+      "learning_rate": 9.951702287469406e-06,
+      "loss": 0.9307,
+      "step": 9006
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5662494897842407,
+      "learning_rate": 9.949844698429354e-06,
+      "loss": 0.9266,
+      "step": 9007
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9565941095352173,
+      "learning_rate": 9.947987111120023e-06,
+      "loss": 1.0666,
+      "step": 9008
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7266814708709717,
+      "learning_rate": 9.94612952560551e-06,
+      "loss": 0.9671,
+      "step": 9009
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.889132022857666,
+      "learning_rate": 9.944271941949916e-06,
+      "loss": 0.9849,
+      "step": 9010
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6573015451431274,
+      "learning_rate": 9.942414360217339e-06,
+      "loss": 0.9979,
+      "step": 9011
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8467018604278564,
+      "learning_rate": 9.940556780471886e-06,
+      "loss": 1.0034,
+      "step": 9012
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8084139823913574,
+      "learning_rate": 9.938699202777652e-06,
+      "loss": 0.9201,
+      "step": 9013
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7212934494018555,
+      "learning_rate": 9.936841627198734e-06,
+      "loss": 1.0466,
+      "step": 9014
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7020750045776367,
+      "learning_rate": 9.934984053799239e-06,
+      "loss": 0.9605,
+      "step": 9015
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8461289405822754,
+      "learning_rate": 9.933126482643259e-06,
+      "loss": 1.0082,
+      "step": 9016
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9211375713348389,
+      "learning_rate": 9.931268913794898e-06,
+      "loss": 1.0497,
+      "step": 9017
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7913408279418945,
+      "learning_rate": 9.929411347318257e-06,
+      "loss": 0.9537,
+      "step": 9018
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.671887993812561,
+      "learning_rate": 9.927553783277432e-06,
+      "loss": 0.9663,
+      "step": 9019
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7826820611953735,
+      "learning_rate": 9.925696221736525e-06,
+      "loss": 1.0174,
+      "step": 9020
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9593912363052368,
+      "learning_rate": 9.923838662759632e-06,
+      "loss": 0.9936,
+      "step": 9021
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9109236001968384,
+      "learning_rate": 9.921981106410856e-06,
+      "loss": 0.8992,
+      "step": 9022
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.737614393234253,
+      "learning_rate": 9.920123552754294e-06,
+      "loss": 0.9765,
+      "step": 9023
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8170056343078613,
+      "learning_rate": 9.918266001854045e-06,
+      "loss": 0.8944,
+      "step": 9024
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7842293977737427,
+      "learning_rate": 9.916408453774207e-06,
+      "loss": 0.9621,
+      "step": 9025
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8636560440063477,
+      "learning_rate": 9.91455090857888e-06,
+      "loss": 0.8785,
+      "step": 9026
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0510454177856445,
+      "learning_rate": 9.912693366332161e-06,
+      "loss": 0.886,
+      "step": 9027
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9517890214920044,
+      "learning_rate": 9.91083582709815e-06,
+      "loss": 0.8668,
+      "step": 9028
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7905247211456299,
+      "learning_rate": 9.908978290940948e-06,
+      "loss": 0.9611,
+      "step": 9029
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6684703826904297,
+      "learning_rate": 9.90712075792465e-06,
+      "loss": 0.9932,
+      "step": 9030
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6955595016479492,
+      "learning_rate": 9.905263228113354e-06,
+      "loss": 0.9546,
+      "step": 9031
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7910109758377075,
+      "learning_rate": 9.90340570157116e-06,
+      "loss": 0.9908,
+      "step": 9032
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.597535252571106,
+      "learning_rate": 9.901548178362167e-06,
+      "loss": 0.8774,
+      "step": 9033
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9226889610290527,
+      "learning_rate": 9.89969065855047e-06,
+      "loss": 0.9957,
+      "step": 9034
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8893277645111084,
+      "learning_rate": 9.897833142200168e-06,
+      "loss": 0.9714,
+      "step": 9035
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7162786722183228,
+      "learning_rate": 9.89597562937536e-06,
+      "loss": 0.9643,
+      "step": 9036
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.919396996498108,
+      "learning_rate": 9.894118120140141e-06,
+      "loss": 0.9581,
+      "step": 9037
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0325645208358765,
+      "learning_rate": 9.892260614558611e-06,
+      "loss": 0.5513,
+      "step": 9038
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.783811330795288,
+      "learning_rate": 9.890403112694867e-06,
+      "loss": 1.0061,
+      "step": 9039
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9469741582870483,
+      "learning_rate": 9.888545614613e-06,
+      "loss": 0.9227,
+      "step": 9040
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6788498163223267,
+      "learning_rate": 9.886688120377118e-06,
+      "loss": 0.9669,
+      "step": 9041
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9167431592941284,
+      "learning_rate": 9.88483063005131e-06,
+      "loss": 0.9351,
+      "step": 9042
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9411070346832275,
+      "learning_rate": 9.882973143699678e-06,
+      "loss": 0.9683,
+      "step": 9043
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7986396551132202,
+      "learning_rate": 9.881115661386314e-06,
+      "loss": 0.9064,
+      "step": 9044
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6843129396438599,
+      "learning_rate": 9.879258183175317e-06,
+      "loss": 0.9591,
+      "step": 9045
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.621228814125061,
+      "learning_rate": 9.877400709130784e-06,
+      "loss": 0.904,
+      "step": 9046
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8656718730926514,
+      "learning_rate": 9.875543239316808e-06,
+      "loss": 0.9345,
+      "step": 9047
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7246835231781006,
+      "learning_rate": 9.873685773797488e-06,
+      "loss": 0.9726,
+      "step": 9048
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9007236957550049,
+      "learning_rate": 9.871828312636919e-06,
+      "loss": 1.0377,
+      "step": 9049
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.052535057067871,
+      "learning_rate": 9.869970855899197e-06,
+      "loss": 1.0168,
+      "step": 9050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7416030168533325,
+      "learning_rate": 9.868113403648416e-06,
+      "loss": 0.9112,
+      "step": 9051
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8203192949295044,
+      "learning_rate": 9.866255955948676e-06,
+      "loss": 1.0144,
+      "step": 9052
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7618767023086548,
+      "learning_rate": 9.864398512864065e-06,
+      "loss": 0.9815,
+      "step": 9053
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8067446947097778,
+      "learning_rate": 9.862541074458685e-06,
+      "loss": 0.9816,
+      "step": 9054
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6581450700759888,
+      "learning_rate": 9.860683640796629e-06,
+      "loss": 0.9614,
+      "step": 9055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6084994077682495,
+      "learning_rate": 9.858826211941993e-06,
+      "loss": 0.9857,
+      "step": 9056
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.766340970993042,
+      "learning_rate": 9.856968787958867e-06,
+      "loss": 0.9352,
+      "step": 9057
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.782696008682251,
+      "learning_rate": 9.85511136891135e-06,
+      "loss": 0.9287,
+      "step": 9058
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.718967080116272,
+      "learning_rate": 9.853253954863535e-06,
+      "loss": 1.016,
+      "step": 9059
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7111783027648926,
+      "learning_rate": 9.851396545879517e-06,
+      "loss": 1.0233,
+      "step": 9060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5144329071044922,
+      "learning_rate": 9.849539142023386e-06,
+      "loss": 0.9756,
+      "step": 9061
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5721980333328247,
+      "learning_rate": 9.847681743359242e-06,
+      "loss": 1.036,
+      "step": 9062
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6753511428833008,
+      "learning_rate": 9.845824349951175e-06,
+      "loss": 0.9393,
+      "step": 9063
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.693739652633667,
+      "learning_rate": 9.843966961863279e-06,
+      "loss": 1.024,
+      "step": 9064
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6939172744750977,
+      "learning_rate": 9.842109579159645e-06,
+      "loss": 0.9415,
+      "step": 9065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7573553323745728,
+      "learning_rate": 9.840252201904369e-06,
+      "loss": 0.9214,
+      "step": 9066
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.852636694908142,
+      "learning_rate": 9.838394830161546e-06,
+      "loss": 0.9495,
+      "step": 9067
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9712855815887451,
+      "learning_rate": 9.836537463995263e-06,
+      "loss": 0.8734,
+      "step": 9068
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7290847301483154,
+      "learning_rate": 9.834680103469617e-06,
+      "loss": 0.9727,
+      "step": 9069
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8492132425308228,
+      "learning_rate": 9.832822748648699e-06,
+      "loss": 0.9864,
+      "step": 9070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.850852131843567,
+      "learning_rate": 9.8309653995966e-06,
+      "loss": 0.9453,
+      "step": 9071
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6377599239349365,
+      "learning_rate": 9.829108056377414e-06,
+      "loss": 0.9797,
+      "step": 9072
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8012235164642334,
+      "learning_rate": 9.827250719055232e-06,
+      "loss": 0.8401,
+      "step": 9073
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.924228549003601,
+      "learning_rate": 9.825393387694144e-06,
+      "loss": 0.9195,
+      "step": 9074
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.701206922531128,
+      "learning_rate": 9.823536062358244e-06,
+      "loss": 0.9048,
+      "step": 9075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7253954410552979,
+      "learning_rate": 9.82167874311162e-06,
+      "loss": 0.979,
+      "step": 9076
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8000190258026123,
+      "learning_rate": 9.81982143001836e-06,
+      "loss": 0.9002,
+      "step": 9077
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1878587007522583,
+      "learning_rate": 9.817964123142566e-06,
+      "loss": 0.6406,
+      "step": 9078
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7436158657073975,
+      "learning_rate": 9.816106822548319e-06,
+      "loss": 1.0117,
+      "step": 9079
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6903648376464844,
+      "learning_rate": 9.81424952829971e-06,
+      "loss": 0.9145,
+      "step": 9080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1014875173568726,
+      "learning_rate": 9.812392240460833e-06,
+      "loss": 0.6022,
+      "step": 9081
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.596754550933838,
+      "learning_rate": 9.810534959095775e-06,
+      "loss": 0.9253,
+      "step": 9082
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.84114408493042,
+      "learning_rate": 9.808677684268628e-06,
+      "loss": 1.0053,
+      "step": 9083
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8379429578781128,
+      "learning_rate": 9.806820416043478e-06,
+      "loss": 0.9541,
+      "step": 9084
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8191065788269043,
+      "learning_rate": 9.804963154484417e-06,
+      "loss": 1.0128,
+      "step": 9085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8007383346557617,
+      "learning_rate": 9.803105899655533e-06,
+      "loss": 0.9645,
+      "step": 9086
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.796562910079956,
+      "learning_rate": 9.801248651620913e-06,
+      "loss": 1.0077,
+      "step": 9087
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6809942722320557,
+      "learning_rate": 9.799391410444648e-06,
+      "loss": 1.0116,
+      "step": 9088
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0320732593536377,
+      "learning_rate": 9.797534176190826e-06,
+      "loss": 0.989,
+      "step": 9089
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7093507051467896,
+      "learning_rate": 9.795676948923533e-06,
+      "loss": 0.9236,
+      "step": 9090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7944904565811157,
+      "learning_rate": 9.793819728706859e-06,
+      "loss": 1.089,
+      "step": 9091
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.809480905532837,
+      "learning_rate": 9.791962515604887e-06,
+      "loss": 0.9345,
+      "step": 9092
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.785309076309204,
+      "learning_rate": 9.79010530968171e-06,
+      "loss": 0.9445,
+      "step": 9093
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6351906061172485,
+      "learning_rate": 9.78824811100141e-06,
+      "loss": 0.9112,
+      "step": 9094
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8967688083648682,
+      "learning_rate": 9.786390919628076e-06,
+      "loss": 0.9493,
+      "step": 9095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7735562324523926,
+      "learning_rate": 9.784533735625798e-06,
+      "loss": 0.92,
+      "step": 9096
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.552429437637329,
+      "learning_rate": 9.782676559058658e-06,
+      "loss": 0.9485,
+      "step": 9097
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8040074110031128,
+      "learning_rate": 9.780819389990742e-06,
+      "loss": 1.0107,
+      "step": 9098
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7570675611495972,
+      "learning_rate": 9.778962228486138e-06,
+      "loss": 0.9501,
+      "step": 9099
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.812116026878357,
+      "learning_rate": 9.777105074608927e-06,
+      "loss": 1.0212,
+      "step": 9100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8255603313446045,
+      "learning_rate": 9.7752479284232e-06,
+      "loss": 0.9661,
+      "step": 9101
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.671401858329773,
+      "learning_rate": 9.773390789993038e-06,
+      "loss": 0.8711,
+      "step": 9102
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9694647789001465,
+      "learning_rate": 9.771533659382528e-06,
+      "loss": 0.9044,
+      "step": 9103
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.122879147529602,
+      "learning_rate": 9.769676536655754e-06,
+      "loss": 0.6526,
+      "step": 9104
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8121559619903564,
+      "learning_rate": 9.767819421876798e-06,
+      "loss": 0.9984,
+      "step": 9105
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6679710149765015,
+      "learning_rate": 9.765962315109743e-06,
+      "loss": 0.9182,
+      "step": 9106
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7525192499160767,
+      "learning_rate": 9.764105216418675e-06,
+      "loss": 0.9105,
+      "step": 9107
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7969474792480469,
+      "learning_rate": 9.762248125867679e-06,
+      "loss": 0.9791,
+      "step": 9108
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.665952205657959,
+      "learning_rate": 9.760391043520837e-06,
+      "loss": 0.8695,
+      "step": 9109
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9379795789718628,
+      "learning_rate": 9.758533969442229e-06,
+      "loss": 0.9559,
+      "step": 9110
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8320332765579224,
+      "learning_rate": 9.75667690369594e-06,
+      "loss": 0.9557,
+      "step": 9111
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6530497074127197,
+      "learning_rate": 9.754819846346051e-06,
+      "loss": 0.9094,
+      "step": 9112
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6904035806655884,
+      "learning_rate": 9.752962797456645e-06,
+      "loss": 0.8795,
+      "step": 9113
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5185389518737793,
+      "learning_rate": 9.751105757091802e-06,
+      "loss": 0.902,
+      "step": 9114
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8129384517669678,
+      "learning_rate": 9.749248725315605e-06,
+      "loss": 0.937,
+      "step": 9115
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.819229245185852,
+      "learning_rate": 9.747391702192132e-06,
+      "loss": 0.9767,
+      "step": 9116
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6527334451675415,
+      "learning_rate": 9.745534687785467e-06,
+      "loss": 0.931,
+      "step": 9117
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9471772313117981,
+      "learning_rate": 9.74367768215969e-06,
+      "loss": 0.5261,
+      "step": 9118
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8168249130249023,
+      "learning_rate": 9.741820685378876e-06,
+      "loss": 0.9973,
+      "step": 9119
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9497843980789185,
+      "learning_rate": 9.739963697507113e-06,
+      "loss": 0.9345,
+      "step": 9120
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8046777248382568,
+      "learning_rate": 9.738106718608475e-06,
+      "loss": 0.974,
+      "step": 9121
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.705181360244751,
+      "learning_rate": 9.736249748747045e-06,
+      "loss": 0.9757,
+      "step": 9122
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7673641443252563,
+      "learning_rate": 9.734392787986896e-06,
+      "loss": 0.8812,
+      "step": 9123
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8405790328979492,
+      "learning_rate": 9.732535836392112e-06,
+      "loss": 0.924,
+      "step": 9124
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.811075210571289,
+      "learning_rate": 9.730678894026769e-06,
+      "loss": 0.9168,
+      "step": 9125
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8107473850250244,
+      "learning_rate": 9.728821960954945e-06,
+      "loss": 0.9082,
+      "step": 9126
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6464771032333374,
+      "learning_rate": 9.726965037240718e-06,
+      "loss": 0.9954,
+      "step": 9127
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6136360168457031,
+      "learning_rate": 9.725108122948163e-06,
+      "loss": 0.8873,
+      "step": 9128
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.04267156124115,
+      "learning_rate": 9.72325121814136e-06,
+      "loss": 0.5504,
+      "step": 9129
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7047576904296875,
+      "learning_rate": 9.721394322884386e-06,
+      "loss": 0.9303,
+      "step": 9130
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.710576057434082,
+      "learning_rate": 9.719537437241311e-06,
+      "loss": 0.8982,
+      "step": 9131
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.874295711517334,
+      "learning_rate": 9.717680561276219e-06,
+      "loss": 0.9556,
+      "step": 9132
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.632091999053955,
+      "learning_rate": 9.715823695053182e-06,
+      "loss": 0.9892,
+      "step": 9133
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6833040714263916,
+      "learning_rate": 9.713966838636277e-06,
+      "loss": 0.9379,
+      "step": 9134
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7168601751327515,
+      "learning_rate": 9.712109992089575e-06,
+      "loss": 0.9189,
+      "step": 9135
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6759908199310303,
+      "learning_rate": 9.710253155477154e-06,
+      "loss": 0.9862,
+      "step": 9136
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9267933368682861,
+      "learning_rate": 9.708396328863087e-06,
+      "loss": 1.0107,
+      "step": 9137
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8829721212387085,
+      "learning_rate": 9.706539512311448e-06,
+      "loss": 0.9729,
+      "step": 9138
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8684232234954834,
+      "learning_rate": 9.70468270588631e-06,
+      "loss": 0.952,
+      "step": 9139
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7286484241485596,
+      "learning_rate": 9.702825909651748e-06,
+      "loss": 1.0044,
+      "step": 9140
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.702661395072937,
+      "learning_rate": 9.700969123671834e-06,
+      "loss": 1.0631,
+      "step": 9141
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8028011322021484,
+      "learning_rate": 9.69911234801064e-06,
+      "loss": 0.9915,
+      "step": 9142
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9446077346801758,
+      "learning_rate": 9.697255582732235e-06,
+      "loss": 1.0217,
+      "step": 9143
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7118337154388428,
+      "learning_rate": 9.695398827900699e-06,
+      "loss": 0.9008,
+      "step": 9144
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6389646530151367,
+      "learning_rate": 9.693542083580096e-06,
+      "loss": 0.9112,
+      "step": 9145
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8176825046539307,
+      "learning_rate": 9.691685349834499e-06,
+      "loss": 0.9228,
+      "step": 9146
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8787474632263184,
+      "learning_rate": 9.68982862672798e-06,
+      "loss": 0.9737,
+      "step": 9147
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.841701865196228,
+      "learning_rate": 9.687971914324607e-06,
+      "loss": 0.9943,
+      "step": 9148
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8081611394882202,
+      "learning_rate": 9.686115212688455e-06,
+      "loss": 0.9862,
+      "step": 9149
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6243294477462769,
+      "learning_rate": 9.684258521883589e-06,
+      "loss": 1.0066,
+      "step": 9150
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7476487159729004,
+      "learning_rate": 9.68240184197408e-06,
+      "loss": 0.948,
+      "step": 9151
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.863942265510559,
+      "learning_rate": 9.680545173023996e-06,
+      "loss": 0.9946,
+      "step": 9152
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7008764743804932,
+      "learning_rate": 9.678688515097405e-06,
+      "loss": 1.0009,
+      "step": 9153
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.732004165649414,
+      "learning_rate": 9.676831868258377e-06,
+      "loss": 0.9444,
+      "step": 9154
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.787692904472351,
+      "learning_rate": 9.674975232570978e-06,
+      "loss": 0.8751,
+      "step": 9155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7370622158050537,
+      "learning_rate": 9.673118608099276e-06,
+      "loss": 0.9743,
+      "step": 9156
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7364182472229004,
+      "learning_rate": 9.671261994907337e-06,
+      "loss": 0.9604,
+      "step": 9157
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.56843101978302,
+      "learning_rate": 9.669405393059228e-06,
+      "loss": 0.8853,
+      "step": 9158
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.878078818321228,
+      "learning_rate": 9.667548802619018e-06,
+      "loss": 1.0337,
+      "step": 9159
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0043394565582275,
+      "learning_rate": 9.665692223650765e-06,
+      "loss": 0.5717,
+      "step": 9160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8545775413513184,
+      "learning_rate": 9.663835656218545e-06,
+      "loss": 1.0082,
+      "step": 9161
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8088265657424927,
+      "learning_rate": 9.661979100386414e-06,
+      "loss": 0.9772,
+      "step": 9162
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8426729440689087,
+      "learning_rate": 9.660122556218441e-06,
+      "loss": 0.9882,
+      "step": 9163
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5606706142425537,
+      "learning_rate": 9.658266023778689e-06,
+      "loss": 0.9362,
+      "step": 9164
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7727643251419067,
+      "learning_rate": 9.656409503131224e-06,
+      "loss": 0.9656,
+      "step": 9165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7324577569961548,
+      "learning_rate": 9.654552994340104e-06,
+      "loss": 0.9218,
+      "step": 9166
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7717936038970947,
+      "learning_rate": 9.652696497469398e-06,
+      "loss": 0.9775,
+      "step": 9167
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.023068428039551,
+      "learning_rate": 9.650840012583164e-06,
+      "loss": 0.9217,
+      "step": 9168
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6869587898254395,
+      "learning_rate": 9.648983539745468e-06,
+      "loss": 0.9897,
+      "step": 9169
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8586260080337524,
+      "learning_rate": 9.647127079020368e-06,
+      "loss": 0.977,
+      "step": 9170
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7261173725128174,
+      "learning_rate": 9.645270630471927e-06,
+      "loss": 1.013,
+      "step": 9171
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6706613302230835,
+      "learning_rate": 9.643414194164205e-06,
+      "loss": 0.9789,
+      "step": 9172
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6840660572052002,
+      "learning_rate": 9.64155777016126e-06,
+      "loss": 0.8984,
+      "step": 9173
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7468554973602295,
+      "learning_rate": 9.639701358527159e-06,
+      "loss": 1.0471,
+      "step": 9174
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.770385980606079,
+      "learning_rate": 9.637844959325958e-06,
+      "loss": 0.9114,
+      "step": 9175
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6317611932754517,
+      "learning_rate": 9.635988572621716e-06,
+      "loss": 0.9143,
+      "step": 9176
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7013816833496094,
+      "learning_rate": 9.63413219847849e-06,
+      "loss": 0.9388,
+      "step": 9177
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7648004293441772,
+      "learning_rate": 9.632275836960339e-06,
+      "loss": 0.9616,
+      "step": 9178
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6768865585327148,
+      "learning_rate": 9.630419488131324e-06,
+      "loss": 0.8414,
+      "step": 9179
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8724957704544067,
+      "learning_rate": 9.628563152055499e-06,
+      "loss": 0.9046,
+      "step": 9180
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.067582368850708,
+      "learning_rate": 9.626706828796923e-06,
+      "loss": 0.5963,
+      "step": 9181
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7324275970458984,
+      "learning_rate": 9.624850518419651e-06,
+      "loss": 1.0006,
+      "step": 9182
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6639113426208496,
+      "learning_rate": 9.62299422098774e-06,
+      "loss": 0.96,
+      "step": 9183
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.666750431060791,
+      "learning_rate": 9.621137936565244e-06,
+      "loss": 0.8709,
+      "step": 9184
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.788880705833435,
+      "learning_rate": 9.619281665216218e-06,
+      "loss": 0.9644,
+      "step": 9185
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9265905618667603,
+      "learning_rate": 9.61742540700472e-06,
+      "loss": 0.9225,
+      "step": 9186
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6256499290466309,
+      "learning_rate": 9.615569161994804e-06,
+      "loss": 1.0594,
+      "step": 9187
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.769329309463501,
+      "learning_rate": 9.613712930250521e-06,
+      "loss": 0.8919,
+      "step": 9188
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7993172407150269,
+      "learning_rate": 9.611856711835926e-06,
+      "loss": 0.9707,
+      "step": 9189
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7590276002883911,
+      "learning_rate": 9.610000506815072e-06,
+      "loss": 0.9534,
+      "step": 9190
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7609660625457764,
+      "learning_rate": 9.608144315252011e-06,
+      "loss": 0.9863,
+      "step": 9191
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6544277667999268,
+      "learning_rate": 9.606288137210795e-06,
+      "loss": 0.9516,
+      "step": 9192
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1204427480697632,
+      "learning_rate": 9.604431972755477e-06,
+      "loss": 0.6416,
+      "step": 9193
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7792524099349976,
+      "learning_rate": 9.602575821950105e-06,
+      "loss": 0.9991,
+      "step": 9194
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.770972490310669,
+      "learning_rate": 9.60071968485873e-06,
+      "loss": 0.9462,
+      "step": 9195
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6536386013031006,
+      "learning_rate": 9.598863561545404e-06,
+      "loss": 0.9621,
+      "step": 9196
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7517553567886353,
+      "learning_rate": 9.597007452074175e-06,
+      "loss": 0.9583,
+      "step": 9197
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7854695320129395,
+      "learning_rate": 9.595151356509095e-06,
+      "loss": 0.8949,
+      "step": 9198
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.052421808242798,
+      "learning_rate": 9.59329527491421e-06,
+      "loss": 1.0236,
+      "step": 9199
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8106080293655396,
+      "learning_rate": 9.59143920735357e-06,
+      "loss": 0.9872,
+      "step": 9200
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.841724157333374,
+      "learning_rate": 9.589583153891222e-06,
+      "loss": 1.0363,
+      "step": 9201
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.870548963546753,
+      "learning_rate": 9.587727114591212e-06,
+      "loss": 1.0104,
+      "step": 9202
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8253833055496216,
+      "learning_rate": 9.585871089517588e-06,
+      "loss": 0.9651,
+      "step": 9203
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6802380084991455,
+      "learning_rate": 9.584015078734395e-06,
+      "loss": 1.0274,
+      "step": 9204
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7654331922531128,
+      "learning_rate": 9.58215908230568e-06,
+      "loss": 1.0008,
+      "step": 9205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0187466144561768,
+      "learning_rate": 9.58030310029549e-06,
+      "loss": 0.915,
+      "step": 9206
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7666550874710083,
+      "learning_rate": 9.578447132767866e-06,
+      "loss": 0.9878,
+      "step": 9207
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8300156593322754,
+      "learning_rate": 9.576591179786852e-06,
+      "loss": 1.0434,
+      "step": 9208
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0231473445892334,
+      "learning_rate": 9.574735241416495e-06,
+      "loss": 1.0109,
+      "step": 9209
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1398117542266846,
+      "learning_rate": 9.572879317720836e-06,
+      "loss": 0.9557,
+      "step": 9210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8345195055007935,
+      "learning_rate": 9.57102340876392e-06,
+      "loss": 0.8996,
+      "step": 9211
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0773444175720215,
+      "learning_rate": 9.569167514609786e-06,
+      "loss": 0.6073,
+      "step": 9212
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5026386976242065,
+      "learning_rate": 9.567311635322479e-06,
+      "loss": 0.851,
+      "step": 9213
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7157379388809204,
+      "learning_rate": 9.565455770966036e-06,
+      "loss": 0.9678,
+      "step": 9214
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8854520320892334,
+      "learning_rate": 9.563599921604505e-06,
+      "loss": 0.9765,
+      "step": 9215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.846970558166504,
+      "learning_rate": 9.56174408730192e-06,
+      "loss": 0.9221,
+      "step": 9216
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7136763334274292,
+      "learning_rate": 9.559888268122323e-06,
+      "loss": 0.853,
+      "step": 9217
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8666157722473145,
+      "learning_rate": 9.558032464129752e-06,
+      "loss": 0.9567,
+      "step": 9218
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.773330569267273,
+      "learning_rate": 9.556176675388245e-06,
+      "loss": 1.0328,
+      "step": 9219
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7969763278961182,
+      "learning_rate": 9.554320901961844e-06,
+      "loss": 0.9848,
+      "step": 9220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9880284667015076,
+      "learning_rate": 9.55246514391458e-06,
+      "loss": 0.5913,
+      "step": 9221
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7320494651794434,
+      "learning_rate": 9.550609401310498e-06,
+      "loss": 0.954,
+      "step": 9222
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.727244257926941,
+      "learning_rate": 9.548753674213627e-06,
+      "loss": 0.9452,
+      "step": 9223
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.803238034248352,
+      "learning_rate": 9.546897962688007e-06,
+      "loss": 1.0112,
+      "step": 9224
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8354918956756592,
+      "learning_rate": 9.545042266797675e-06,
+      "loss": 0.9314,
+      "step": 9225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7105672359466553,
+      "learning_rate": 9.543186586606657e-06,
+      "loss": 0.9461,
+      "step": 9226
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7394497394561768,
+      "learning_rate": 9.541330922178998e-06,
+      "loss": 0.9825,
+      "step": 9227
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9199997186660767,
+      "learning_rate": 9.539475273578729e-06,
+      "loss": 0.9236,
+      "step": 9228
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.910712480545044,
+      "learning_rate": 9.53761964086988e-06,
+      "loss": 1.0679,
+      "step": 9229
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.696234107017517,
+      "learning_rate": 9.535764024116488e-06,
+      "loss": 0.9596,
+      "step": 9230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.876043677330017,
+      "learning_rate": 9.53390842338258e-06,
+      "loss": 0.9485,
+      "step": 9231
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.908756136894226,
+      "learning_rate": 9.53205283873219e-06,
+      "loss": 0.9829,
+      "step": 9232
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6808531284332275,
+      "learning_rate": 9.530197270229352e-06,
+      "loss": 0.9752,
+      "step": 9233
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7408236265182495,
+      "learning_rate": 9.528341717938093e-06,
+      "loss": 0.9499,
+      "step": 9234
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.89076566696167,
+      "learning_rate": 9.526486181922443e-06,
+      "loss": 0.9591,
+      "step": 9235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.696572184562683,
+      "learning_rate": 9.524630662246432e-06,
+      "loss": 0.939,
+      "step": 9236
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7089451551437378,
+      "learning_rate": 9.522775158974091e-06,
+      "loss": 0.9423,
+      "step": 9237
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6499236822128296,
+      "learning_rate": 9.520919672169444e-06,
+      "loss": 0.8797,
+      "step": 9238
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7998429536819458,
+      "learning_rate": 9.51906420189652e-06,
+      "loss": 0.9286,
+      "step": 9239
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6627498865127563,
+      "learning_rate": 9.517208748219347e-06,
+      "loss": 0.9495,
+      "step": 9240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7416205406188965,
+      "learning_rate": 9.515353311201953e-06,
+      "loss": 0.9525,
+      "step": 9241
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7147578001022339,
+      "learning_rate": 9.513497890908363e-06,
+      "loss": 0.9136,
+      "step": 9242
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7469813823699951,
+      "learning_rate": 9.511642487402601e-06,
+      "loss": 0.9336,
+      "step": 9243
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8452692031860352,
+      "learning_rate": 9.509787100748693e-06,
+      "loss": 0.9226,
+      "step": 9244
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7484073638916016,
+      "learning_rate": 9.507931731010663e-06,
+      "loss": 0.979,
+      "step": 9245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7768722772598267,
+      "learning_rate": 9.506076378252535e-06,
+      "loss": 0.9941,
+      "step": 9246
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7821413278579712,
+      "learning_rate": 9.50422104253833e-06,
+      "loss": 0.9402,
+      "step": 9247
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7205201387405396,
+      "learning_rate": 9.502365723932072e-06,
+      "loss": 0.9296,
+      "step": 9248
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6679078340530396,
+      "learning_rate": 9.500510422497783e-06,
+      "loss": 0.9417,
+      "step": 9249
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.790339469909668,
+      "learning_rate": 9.498655138299484e-06,
+      "loss": 0.9388,
+      "step": 9250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8243279457092285,
+      "learning_rate": 9.496799871401195e-06,
+      "loss": 0.9771,
+      "step": 9251
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.116796851158142,
+      "learning_rate": 9.494944621866938e-06,
+      "loss": 0.6068,
+      "step": 9252
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.671281337738037,
+      "learning_rate": 9.49308938976073e-06,
+      "loss": 0.9475,
+      "step": 9253
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5734124183654785,
+      "learning_rate": 9.491234175146592e-06,
+      "loss": 0.9178,
+      "step": 9254
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.646193265914917,
+      "learning_rate": 9.489378978088542e-06,
+      "loss": 1.0565,
+      "step": 9255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7536954879760742,
+      "learning_rate": 9.487523798650596e-06,
+      "loss": 1.0017,
+      "step": 9256
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7661257982254028,
+      "learning_rate": 9.485668636896774e-06,
+      "loss": 0.9342,
+      "step": 9257
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7019902467727661,
+      "learning_rate": 9.48381349289109e-06,
+      "loss": 1.0087,
+      "step": 9258
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.738187551498413,
+      "learning_rate": 9.48195836669756e-06,
+      "loss": 0.961,
+      "step": 9259
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7771666049957275,
+      "learning_rate": 9.480103258380198e-06,
+      "loss": 0.986,
+      "step": 9260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7559185028076172,
+      "learning_rate": 9.478248168003022e-06,
+      "loss": 1.0406,
+      "step": 9261
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7301557064056396,
+      "learning_rate": 9.476393095630046e-06,
+      "loss": 0.9277,
+      "step": 9262
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8342472314834595,
+      "learning_rate": 9.474538041325277e-06,
+      "loss": 0.9542,
+      "step": 9263
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.74834406375885,
+      "learning_rate": 9.472683005152735e-06,
+      "loss": 0.8808,
+      "step": 9264
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6324621438980103,
+      "learning_rate": 9.47082798717643e-06,
+      "loss": 0.9081,
+      "step": 9265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.949791431427002,
+      "learning_rate": 9.468972987460372e-06,
+      "loss": 0.9579,
+      "step": 9266
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8759971857070923,
+      "learning_rate": 9.467118006068575e-06,
+      "loss": 1.0282,
+      "step": 9267
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6768170595169067,
+      "learning_rate": 9.465263043065045e-06,
+      "loss": 0.9918,
+      "step": 9268
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8442323207855225,
+      "learning_rate": 9.463408098513794e-06,
+      "loss": 0.9271,
+      "step": 9269
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8389896154403687,
+      "learning_rate": 9.461553172478831e-06,
+      "loss": 1.0064,
+      "step": 9270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6300525665283203,
+      "learning_rate": 9.459698265024164e-06,
+      "loss": 0.9981,
+      "step": 9271
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8235679864883423,
+      "learning_rate": 9.4578433762138e-06,
+      "loss": 0.9891,
+      "step": 9272
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0574673414230347,
+      "learning_rate": 9.455988506111747e-06,
+      "loss": 0.6068,
+      "step": 9273
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6184523105621338,
+      "learning_rate": 9.454133654782011e-06,
+      "loss": 0.9987,
+      "step": 9274
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1272081136703491,
+      "learning_rate": 9.452278822288597e-06,
+      "loss": 0.6166,
+      "step": 9275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8617305755615234,
+      "learning_rate": 9.45042400869551e-06,
+      "loss": 1.0194,
+      "step": 9276
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6669964790344238,
+      "learning_rate": 9.448569214066757e-06,
+      "loss": 0.9676,
+      "step": 9277
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9764981865882874,
+      "learning_rate": 9.446714438466338e-06,
+      "loss": 0.5387,
+      "step": 9278
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.910941481590271,
+      "learning_rate": 9.44485968195826e-06,
+      "loss": 0.9878,
+      "step": 9279
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9592466354370117,
+      "learning_rate": 9.443004944606522e-06,
+      "loss": 0.9693,
+      "step": 9280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9266902208328247,
+      "learning_rate": 9.44115022647513e-06,
+      "loss": 1.0737,
+      "step": 9281
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9121421575546265,
+      "learning_rate": 9.439295527628083e-06,
+      "loss": 0.9761,
+      "step": 9282
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0599453449249268,
+      "learning_rate": 9.437440848129377e-06,
+      "loss": 0.6047,
+      "step": 9283
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.684749960899353,
+      "learning_rate": 9.43558618804302e-06,
+      "loss": 0.9396,
+      "step": 9284
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7958449125289917,
+      "learning_rate": 9.433731547433007e-06,
+      "loss": 0.9947,
+      "step": 9285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.284531593322754,
+      "learning_rate": 9.431876926363335e-06,
+      "loss": 0.9178,
+      "step": 9286
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.710998296737671,
+      "learning_rate": 9.430022324898003e-06,
+      "loss": 0.9176,
+      "step": 9287
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6571248769760132,
+      "learning_rate": 9.428167743101009e-06,
+      "loss": 0.9121,
+      "step": 9288
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.716275691986084,
+      "learning_rate": 9.426313181036349e-06,
+      "loss": 0.9716,
+      "step": 9289
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6823360919952393,
+      "learning_rate": 9.424458638768018e-06,
+      "loss": 0.8511,
+      "step": 9290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5675513744354248,
+      "learning_rate": 9.422604116360012e-06,
+      "loss": 0.9822,
+      "step": 9291
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7278447151184082,
+      "learning_rate": 9.420749613876326e-06,
+      "loss": 0.8702,
+      "step": 9292
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7466652393341064,
+      "learning_rate": 9.41889513138095e-06,
+      "loss": 0.9253,
+      "step": 9293
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6461753845214844,
+      "learning_rate": 9.417040668937881e-06,
+      "loss": 0.9027,
+      "step": 9294
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.678087592124939,
+      "learning_rate": 9.415186226611111e-06,
+      "loss": 0.8641,
+      "step": 9295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0825157165527344,
+      "learning_rate": 9.41333180446463e-06,
+      "loss": 0.5885,
+      "step": 9296
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.754571557044983,
+      "learning_rate": 9.411477402562432e-06,
+      "loss": 0.9902,
+      "step": 9297
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7567294836044312,
+      "learning_rate": 9.409623020968502e-06,
+      "loss": 0.9742,
+      "step": 9298
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7099295854568481,
+      "learning_rate": 9.407768659746833e-06,
+      "loss": 0.9486,
+      "step": 9299
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8363503217697144,
+      "learning_rate": 9.405914318961414e-06,
+      "loss": 0.9886,
+      "step": 9300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8244668245315552,
+      "learning_rate": 9.404059998676232e-06,
+      "loss": 0.9367,
+      "step": 9301
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.678518533706665,
+      "learning_rate": 9.402205698955274e-06,
+      "loss": 0.9928,
+      "step": 9302
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.814306616783142,
+      "learning_rate": 9.40035141986253e-06,
+      "loss": 0.9596,
+      "step": 9303
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6924738883972168,
+      "learning_rate": 9.398497161461981e-06,
+      "loss": 0.867,
+      "step": 9304
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6738865375518799,
+      "learning_rate": 9.396642923817613e-06,
+      "loss": 1.0128,
+      "step": 9305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6957411766052246,
+      "learning_rate": 9.394788706993414e-06,
+      "loss": 0.8222,
+      "step": 9306
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9628829956054688,
+      "learning_rate": 9.392934511053367e-06,
+      "loss": 0.9796,
+      "step": 9307
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9452887773513794,
+      "learning_rate": 9.391080336061454e-06,
+      "loss": 1.0289,
+      "step": 9308
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.852339744567871,
+      "learning_rate": 9.38922618208166e-06,
+      "loss": 1.0154,
+      "step": 9309
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5806230306625366,
+      "learning_rate": 9.387372049177961e-06,
+      "loss": 0.9014,
+      "step": 9310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6444921493530273,
+      "learning_rate": 9.385517937414341e-06,
+      "loss": 1.0151,
+      "step": 9311
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7565041780471802,
+      "learning_rate": 9.383663846854782e-06,
+      "loss": 0.9015,
+      "step": 9312
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6916500329971313,
+      "learning_rate": 9.381809777563262e-06,
+      "loss": 0.956,
+      "step": 9313
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6737329959869385,
+      "learning_rate": 9.37995572960376e-06,
+      "loss": 1.0323,
+      "step": 9314
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6440950632095337,
+      "learning_rate": 9.378101703040254e-06,
+      "loss": 0.9892,
+      "step": 9315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7503169775009155,
+      "learning_rate": 9.376247697936719e-06,
+      "loss": 0.9663,
+      "step": 9316
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6544078588485718,
+      "learning_rate": 9.374393714357132e-06,
+      "loss": 0.9875,
+      "step": 9317
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6965941190719604,
+      "learning_rate": 9.372539752365474e-06,
+      "loss": 0.9113,
+      "step": 9318
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6590759754180908,
+      "learning_rate": 9.370685812025716e-06,
+      "loss": 0.8929,
+      "step": 9319
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8665412664413452,
+      "learning_rate": 9.368831893401832e-06,
+      "loss": 0.9715,
+      "step": 9320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6794768571853638,
+      "learning_rate": 9.366977996557797e-06,
+      "loss": 0.9558,
+      "step": 9321
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7764272689819336,
+      "learning_rate": 9.365124121557583e-06,
+      "loss": 0.9906,
+      "step": 9322
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.009660243988037,
+      "learning_rate": 9.363270268465162e-06,
+      "loss": 0.9514,
+      "step": 9323
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7077337503433228,
+      "learning_rate": 9.361416437344504e-06,
+      "loss": 0.9117,
+      "step": 9324
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8543405532836914,
+      "learning_rate": 9.359562628259582e-06,
+      "loss": 0.9906,
+      "step": 9325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6126928329467773,
+      "learning_rate": 9.357708841274365e-06,
+      "loss": 0.9519,
+      "step": 9326
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1099298000335693,
+      "learning_rate": 9.355855076452823e-06,
+      "loss": 0.9479,
+      "step": 9327
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7311211824417114,
+      "learning_rate": 9.35400133385892e-06,
+      "loss": 0.9436,
+      "step": 9328
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7908272743225098,
+      "learning_rate": 9.352147613556626e-06,
+      "loss": 0.9988,
+      "step": 9329
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7593470811843872,
+      "learning_rate": 9.35029391560991e-06,
+      "loss": 0.9288,
+      "step": 9330
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6753427982330322,
+      "learning_rate": 9.348440240082737e-06,
+      "loss": 0.9635,
+      "step": 9331
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.61228609085083,
+      "learning_rate": 9.34658658703907e-06,
+      "loss": 0.9349,
+      "step": 9332
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.681319236755371,
+      "learning_rate": 9.344732956542874e-06,
+      "loss": 0.9856,
+      "step": 9333
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7990999221801758,
+      "learning_rate": 9.342879348658115e-06,
+      "loss": 1.0457,
+      "step": 9334
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.905035138130188,
+      "learning_rate": 9.341025763448753e-06,
+      "loss": 0.9639,
+      "step": 9335
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.023582935333252,
+      "learning_rate": 9.339172200978752e-06,
+      "loss": 1.0008,
+      "step": 9336
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.716108798980713,
+      "learning_rate": 9.337318661312072e-06,
+      "loss": 0.9706,
+      "step": 9337
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.643920660018921,
+      "learning_rate": 9.335465144512674e-06,
+      "loss": 0.9526,
+      "step": 9338
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7576969861984253,
+      "learning_rate": 9.333611650644518e-06,
+      "loss": 0.9388,
+      "step": 9339
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0834410190582275,
+      "learning_rate": 9.331758179771562e-06,
+      "loss": 0.593,
+      "step": 9340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.848089337348938,
+      "learning_rate": 9.329904731957761e-06,
+      "loss": 1.021,
+      "step": 9341
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.681815266609192,
+      "learning_rate": 9.328051307267079e-06,
+      "loss": 0.897,
+      "step": 9342
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6038150787353516,
+      "learning_rate": 9.32619790576347e-06,
+      "loss": 0.965,
+      "step": 9343
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6405913829803467,
+      "learning_rate": 9.324344527510886e-06,
+      "loss": 0.9329,
+      "step": 9344
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6053853034973145,
+      "learning_rate": 9.32249117257329e-06,
+      "loss": 0.968,
+      "step": 9345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1093873977661133,
+      "learning_rate": 9.320637841014625e-06,
+      "loss": 0.9681,
+      "step": 9346
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.633998990058899,
+      "learning_rate": 9.318784532898855e-06,
+      "loss": 0.9681,
+      "step": 9347
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8004143238067627,
+      "learning_rate": 9.316931248289926e-06,
+      "loss": 0.9307,
+      "step": 9348
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7232537269592285,
+      "learning_rate": 9.315077987251793e-06,
+      "loss": 0.9118,
+      "step": 9349
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8785570859909058,
+      "learning_rate": 9.313224749848405e-06,
+      "loss": 0.9349,
+      "step": 9350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.798021674156189,
+      "learning_rate": 9.311371536143713e-06,
+      "loss": 0.949,
+      "step": 9351
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9915432929992676,
+      "learning_rate": 9.309518346201665e-06,
+      "loss": 0.9075,
+      "step": 9352
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.695725440979004,
+      "learning_rate": 9.30766518008621e-06,
+      "loss": 0.9895,
+      "step": 9353
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0562973022460938,
+      "learning_rate": 9.305812037861296e-06,
+      "loss": 0.9156,
+      "step": 9354
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8085405826568604,
+      "learning_rate": 9.30395891959087e-06,
+      "loss": 0.9798,
+      "step": 9355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8736507892608643,
+      "learning_rate": 9.302105825338876e-06,
+      "loss": 0.9638,
+      "step": 9356
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.645461082458496,
+      "learning_rate": 9.300252755169261e-06,
+      "loss": 1.0142,
+      "step": 9357
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8870527744293213,
+      "learning_rate": 9.29839970914597e-06,
+      "loss": 0.9952,
+      "step": 9358
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7292526960372925,
+      "learning_rate": 9.296546687332941e-06,
+      "loss": 0.8701,
+      "step": 9359
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8257315158843994,
+      "learning_rate": 9.294693689794123e-06,
+      "loss": 0.988,
+      "step": 9360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8182116746902466,
+      "learning_rate": 9.292840716593458e-06,
+      "loss": 0.9103,
+      "step": 9361
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8039677143096924,
+      "learning_rate": 9.290987767794883e-06,
+      "loss": 0.9649,
+      "step": 9362
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6090418100357056,
+      "learning_rate": 9.28913484346234e-06,
+      "loss": 0.9163,
+      "step": 9363
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.694377064704895,
+      "learning_rate": 9.287281943659767e-06,
+      "loss": 0.9383,
+      "step": 9364
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7505521774291992,
+      "learning_rate": 9.285429068451103e-06,
+      "loss": 1.0301,
+      "step": 9365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6647368669509888,
+      "learning_rate": 9.283576217900286e-06,
+      "loss": 0.9736,
+      "step": 9366
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.736624002456665,
+      "learning_rate": 9.281723392071254e-06,
+      "loss": 0.9525,
+      "step": 9367
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7116786241531372,
+      "learning_rate": 9.279870591027939e-06,
+      "loss": 0.9935,
+      "step": 9368
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7518374919891357,
+      "learning_rate": 9.27801781483428e-06,
+      "loss": 0.9534,
+      "step": 9369
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7609783411026,
+      "learning_rate": 9.27616506355421e-06,
+      "loss": 0.9716,
+      "step": 9370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8115838766098022,
+      "learning_rate": 9.274312337251658e-06,
+      "loss": 0.9897,
+      "step": 9371
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8622715473175049,
+      "learning_rate": 9.272459635990563e-06,
+      "loss": 0.9372,
+      "step": 9372
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8954293727874756,
+      "learning_rate": 9.270606959834853e-06,
+      "loss": 0.9923,
+      "step": 9373
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6513603925704956,
+      "learning_rate": 9.26875430884846e-06,
+      "loss": 0.8766,
+      "step": 9374
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7509721517562866,
+      "learning_rate": 9.266901683095313e-06,
+      "loss": 0.9584,
+      "step": 9375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7485777139663696,
+      "learning_rate": 9.26504908263934e-06,
+      "loss": 0.967,
+      "step": 9376
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8830461502075195,
+      "learning_rate": 9.263196507544472e-06,
+      "loss": 0.9521,
+      "step": 9377
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.940224051475525,
+      "learning_rate": 9.261343957874633e-06,
+      "loss": 0.9947,
+      "step": 9378
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9655793905258179,
+      "learning_rate": 9.259491433693751e-06,
+      "loss": 1.0147,
+      "step": 9379
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.828822135925293,
+      "learning_rate": 9.257638935065752e-06,
+      "loss": 1.0006,
+      "step": 9380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.077743649482727,
+      "learning_rate": 9.255786462054559e-06,
+      "loss": 0.6011,
+      "step": 9381
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8199617862701416,
+      "learning_rate": 9.253934014724097e-06,
+      "loss": 0.9852,
+      "step": 9382
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.846769094467163,
+      "learning_rate": 9.252081593138284e-06,
+      "loss": 0.926,
+      "step": 9383
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7593907117843628,
+      "learning_rate": 9.25022919736105e-06,
+      "loss": 0.8686,
+      "step": 9384
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9020034074783325,
+      "learning_rate": 9.248376827456312e-06,
+      "loss": 0.9573,
+      "step": 9385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.859734296798706,
+      "learning_rate": 9.246524483487988e-06,
+      "loss": 0.9477,
+      "step": 9386
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.704759120941162,
+      "learning_rate": 9.24467216552e-06,
+      "loss": 0.9924,
+      "step": 9387
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8651659488677979,
+      "learning_rate": 9.242819873616268e-06,
+      "loss": 1.0056,
+      "step": 9388
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.623112678527832,
+      "learning_rate": 9.240967607840706e-06,
+      "loss": 0.9713,
+      "step": 9389
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8262887001037598,
+      "learning_rate": 9.23911536825723e-06,
+      "loss": 1.0327,
+      "step": 9390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5998939275741577,
+      "learning_rate": 9.237263154929759e-06,
+      "loss": 0.9604,
+      "step": 9391
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9035791158676147,
+      "learning_rate": 9.235410967922205e-06,
+      "loss": 0.9565,
+      "step": 9392
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.736545205116272,
+      "learning_rate": 9.233558807298484e-06,
+      "loss": 1.0426,
+      "step": 9393
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8222191333770752,
+      "learning_rate": 9.231706673122504e-06,
+      "loss": 0.913,
+      "step": 9394
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8094574213027954,
+      "learning_rate": 9.229854565458181e-06,
+      "loss": 1.0163,
+      "step": 9395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.698534607887268,
+      "learning_rate": 9.228002484369429e-06,
+      "loss": 0.9728,
+      "step": 9396
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6703506708145142,
+      "learning_rate": 9.226150429920153e-06,
+      "loss": 0.8821,
+      "step": 9397
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7057803869247437,
+      "learning_rate": 9.224298402174264e-06,
+      "loss": 0.9655,
+      "step": 9398
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8563436269760132,
+      "learning_rate": 9.222446401195672e-06,
+      "loss": 0.9388,
+      "step": 9399
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.850363850593567,
+      "learning_rate": 9.22059442704828e-06,
+      "loss": 0.9431,
+      "step": 9400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6936535835266113,
+      "learning_rate": 9.218742479796e-06,
+      "loss": 0.9031,
+      "step": 9401
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9065234661102295,
+      "learning_rate": 9.216890559502732e-06,
+      "loss": 0.9411,
+      "step": 9402
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7509254217147827,
+      "learning_rate": 9.215038666232385e-06,
+      "loss": 1.0015,
+      "step": 9403
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7968378067016602,
+      "learning_rate": 9.213186800048862e-06,
+      "loss": 0.9325,
+      "step": 9404
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8421032428741455,
+      "learning_rate": 9.211334961016063e-06,
+      "loss": 0.9295,
+      "step": 9405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9317408800125122,
+      "learning_rate": 9.20948314919789e-06,
+      "loss": 1.0458,
+      "step": 9406
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9427756071090698,
+      "learning_rate": 9.207631364658244e-06,
+      "loss": 0.9244,
+      "step": 9407
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1277761459350586,
+      "learning_rate": 9.20577960746103e-06,
+      "loss": 0.6026,
+      "step": 9408
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7018071413040161,
+      "learning_rate": 9.203927877670143e-06,
+      "loss": 0.9916,
+      "step": 9409
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9066749811172485,
+      "learning_rate": 9.202076175349477e-06,
+      "loss": 0.9313,
+      "step": 9410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7801448106765747,
+      "learning_rate": 9.200224500562937e-06,
+      "loss": 0.908,
+      "step": 9411
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7181084156036377,
+      "learning_rate": 9.198372853374415e-06,
+      "loss": 0.9213,
+      "step": 9412
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7172428369522095,
+      "learning_rate": 9.196521233847806e-06,
+      "loss": 0.892,
+      "step": 9413
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.706556797027588,
+      "learning_rate": 9.194669642047004e-06,
+      "loss": 0.9235,
+      "step": 9414
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6751763820648193,
+      "learning_rate": 9.192818078035904e-06,
+      "loss": 0.9211,
+      "step": 9415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.077202558517456,
+      "learning_rate": 9.190966541878399e-06,
+      "loss": 0.8807,
+      "step": 9416
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7662979364395142,
+      "learning_rate": 9.189115033638378e-06,
+      "loss": 0.9822,
+      "step": 9417
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.911409854888916,
+      "learning_rate": 9.187263553379731e-06,
+      "loss": 0.9367,
+      "step": 9418
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8815838098526,
+      "learning_rate": 9.185412101166349e-06,
+      "loss": 1.0048,
+      "step": 9419
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6018904447555542,
+      "learning_rate": 9.18356067706212e-06,
+      "loss": 0.943,
+      "step": 9420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7018402814865112,
+      "learning_rate": 9.181709281130932e-06,
+      "loss": 0.9695,
+      "step": 9421
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0471857786178589,
+      "learning_rate": 9.17985791343667e-06,
+      "loss": 0.621,
+      "step": 9422
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5961575508117676,
+      "learning_rate": 9.178006574043221e-06,
+      "loss": 0.9629,
+      "step": 9423
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9312336444854736,
+      "learning_rate": 9.17615526301447e-06,
+      "loss": 0.9628,
+      "step": 9424
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6334311962127686,
+      "learning_rate": 9.174303980414295e-06,
+      "loss": 0.9408,
+      "step": 9425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6464003324508667,
+      "learning_rate": 9.172452726306586e-06,
+      "loss": 0.9167,
+      "step": 9426
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7353788614273071,
+      "learning_rate": 9.170601500755224e-06,
+      "loss": 1.0614,
+      "step": 9427
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.622787594795227,
+      "learning_rate": 9.168750303824085e-06,
+      "loss": 0.9679,
+      "step": 9428
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8990715742111206,
+      "learning_rate": 9.166899135577052e-06,
+      "loss": 1.0443,
+      "step": 9429
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7969086170196533,
+      "learning_rate": 9.165047996078001e-06,
+      "loss": 0.9661,
+      "step": 9430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7733854055404663,
+      "learning_rate": 9.163196885390812e-06,
+      "loss": 0.9886,
+      "step": 9431
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5882633924484253,
+      "learning_rate": 9.161345803579362e-06,
+      "loss": 0.9392,
+      "step": 9432
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7996106147766113,
+      "learning_rate": 9.159494750707527e-06,
+      "loss": 0.9823,
+      "step": 9433
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6894115209579468,
+      "learning_rate": 9.157643726839177e-06,
+      "loss": 0.9926,
+      "step": 9434
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8034989833831787,
+      "learning_rate": 9.155792732038192e-06,
+      "loss": 0.9992,
+      "step": 9435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6855391263961792,
+      "learning_rate": 9.153941766368439e-06,
+      "loss": 0.9735,
+      "step": 9436
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.645350694656372,
+      "learning_rate": 9.152090829893792e-06,
+      "loss": 0.9966,
+      "step": 9437
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6695024967193604,
+      "learning_rate": 9.150239922678122e-06,
+      "loss": 0.9534,
+      "step": 9438
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7206846475601196,
+      "learning_rate": 9.1483890447853e-06,
+      "loss": 0.9488,
+      "step": 9439
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.875939130783081,
+      "learning_rate": 9.146538196279193e-06,
+      "loss": 0.9287,
+      "step": 9440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6875778436660767,
+      "learning_rate": 9.144687377223669e-06,
+      "loss": 0.9785,
+      "step": 9441
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7431803941726685,
+      "learning_rate": 9.142836587682594e-06,
+      "loss": 0.9611,
+      "step": 9442
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6714915037155151,
+      "learning_rate": 9.140985827719835e-06,
+      "loss": 0.8847,
+      "step": 9443
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6376293897628784,
+      "learning_rate": 9.139135097399254e-06,
+      "loss": 0.9481,
+      "step": 9444
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.180797815322876,
+      "learning_rate": 9.137284396784716e-06,
+      "loss": 0.6542,
+      "step": 9445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.837324619293213,
+      "learning_rate": 9.135433725940086e-06,
+      "loss": 0.9308,
+      "step": 9446
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7160722017288208,
+      "learning_rate": 9.13358308492922e-06,
+      "loss": 0.9645,
+      "step": 9447
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8531137704849243,
+      "learning_rate": 9.131732473815984e-06,
+      "loss": 0.9529,
+      "step": 9448
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7481144666671753,
+      "learning_rate": 9.129881892664232e-06,
+      "loss": 0.9322,
+      "step": 9449
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6420543193817139,
+      "learning_rate": 9.128031341537826e-06,
+      "loss": 0.9179,
+      "step": 9450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7773021459579468,
+      "learning_rate": 9.126180820500624e-06,
+      "loss": 1.0342,
+      "step": 9451
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8343827724456787,
+      "learning_rate": 9.124330329616482e-06,
+      "loss": 0.9579,
+      "step": 9452
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7513643503189087,
+      "learning_rate": 9.122479868949253e-06,
+      "loss": 0.9919,
+      "step": 9453
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.779710292816162,
+      "learning_rate": 9.12062943856279e-06,
+      "loss": 0.8962,
+      "step": 9454
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8562480211257935,
+      "learning_rate": 9.118779038520953e-06,
+      "loss": 0.9752,
+      "step": 9455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8598366975784302,
+      "learning_rate": 9.116928668887587e-06,
+      "loss": 0.9388,
+      "step": 9456
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6017711162567139,
+      "learning_rate": 9.115078329726548e-06,
+      "loss": 0.8695,
+      "step": 9457
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.767075538635254,
+      "learning_rate": 9.113228021101682e-06,
+      "loss": 0.9458,
+      "step": 9458
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7146070003509521,
+      "learning_rate": 9.111377743076842e-06,
+      "loss": 0.9633,
+      "step": 9459
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0590147972106934,
+      "learning_rate": 9.109527495715872e-06,
+      "loss": 1.0209,
+      "step": 9460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7618181705474854,
+      "learning_rate": 9.107677279082619e-06,
+      "loss": 0.8306,
+      "step": 9461
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7003214359283447,
+      "learning_rate": 9.105827093240932e-06,
+      "loss": 0.9516,
+      "step": 9462
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6746435165405273,
+      "learning_rate": 9.103976938254656e-06,
+      "loss": 0.9773,
+      "step": 9463
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7741769552230835,
+      "learning_rate": 9.10212681418763e-06,
+      "loss": 0.9066,
+      "step": 9464
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8328062295913696,
+      "learning_rate": 9.100276721103703e-06,
+      "loss": 0.98,
+      "step": 9465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7269577980041504,
+      "learning_rate": 9.098426659066711e-06,
+      "loss": 0.9496,
+      "step": 9466
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7527824640274048,
+      "learning_rate": 9.096576628140497e-06,
+      "loss": 0.9444,
+      "step": 9467
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8261911869049072,
+      "learning_rate": 9.094726628388899e-06,
+      "loss": 1.012,
+      "step": 9468
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.655664324760437,
+      "learning_rate": 9.092876659875757e-06,
+      "loss": 1.013,
+      "step": 9469
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7257881164550781,
+      "learning_rate": 9.091026722664908e-06,
+      "loss": 0.8881,
+      "step": 9470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8430830240249634,
+      "learning_rate": 9.089176816820187e-06,
+      "loss": 1.0273,
+      "step": 9471
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2765910625457764,
+      "learning_rate": 9.08732694240543e-06,
+      "loss": 1.0921,
+      "step": 9472
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.716312289237976,
+      "learning_rate": 9.08547709948447e-06,
+      "loss": 0.9452,
+      "step": 9473
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.932315707206726,
+      "learning_rate": 9.083627288121141e-06,
+      "loss": 1.0633,
+      "step": 9474
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8637512922286987,
+      "learning_rate": 9.081777508379275e-06,
+      "loss": 1.0597,
+      "step": 9475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9153664112091064,
+      "learning_rate": 9.0799277603227e-06,
+      "loss": 0.8944,
+      "step": 9476
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8246362209320068,
+      "learning_rate": 9.07807804401525e-06,
+      "loss": 0.9529,
+      "step": 9477
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9576621055603027,
+      "learning_rate": 9.076228359520752e-06,
+      "loss": 0.972,
+      "step": 9478
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.001964807510376,
+      "learning_rate": 9.074378706903029e-06,
+      "loss": 0.9575,
+      "step": 9479
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8684083223342896,
+      "learning_rate": 9.072529086225917e-06,
+      "loss": 0.8825,
+      "step": 9480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8071510791778564,
+      "learning_rate": 9.070679497553232e-06,
+      "loss": 1.0095,
+      "step": 9481
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7188029289245605,
+      "learning_rate": 9.068829940948802e-06,
+      "loss": 0.9754,
+      "step": 9482
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7659193277359009,
+      "learning_rate": 9.06698041647645e-06,
+      "loss": 0.9619,
+      "step": 9483
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6424344778060913,
+      "learning_rate": 9.065130924199998e-06,
+      "loss": 0.9424,
+      "step": 9484
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.927290678024292,
+      "learning_rate": 9.063281464183267e-06,
+      "loss": 0.9697,
+      "step": 9485
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.84980046749115,
+      "learning_rate": 9.061432036490076e-06,
+      "loss": 0.939,
+      "step": 9486
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.702980637550354,
+      "learning_rate": 9.059582641184242e-06,
+      "loss": 1.0235,
+      "step": 9487
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7925997972488403,
+      "learning_rate": 9.057733278329585e-06,
+      "loss": 0.9668,
+      "step": 9488
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7005348205566406,
+      "learning_rate": 9.055883947989921e-06,
+      "loss": 0.9953,
+      "step": 9489
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7919210195541382,
+      "learning_rate": 9.054034650229065e-06,
+      "loss": 0.9414,
+      "step": 9490
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9344539642333984,
+      "learning_rate": 9.052185385110826e-06,
+      "loss": 0.9659,
+      "step": 9491
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8405370712280273,
+      "learning_rate": 9.050336152699026e-06,
+      "loss": 0.9985,
+      "step": 9492
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7546119689941406,
+      "learning_rate": 9.048486953057472e-06,
+      "loss": 0.9996,
+      "step": 9493
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9716124534606934,
+      "learning_rate": 9.046637786249977e-06,
+      "loss": 0.9922,
+      "step": 9494
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7934556007385254,
+      "learning_rate": 9.044788652340346e-06,
+      "loss": 0.9561,
+      "step": 9495
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8514511585235596,
+      "learning_rate": 9.042939551392392e-06,
+      "loss": 0.9761,
+      "step": 9496
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8618247509002686,
+      "learning_rate": 9.041090483469921e-06,
+      "loss": 0.9264,
+      "step": 9497
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.636122465133667,
+      "learning_rate": 9.039241448636739e-06,
+      "loss": 0.9431,
+      "step": 9498
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9527537822723389,
+      "learning_rate": 9.03739244695665e-06,
+      "loss": 0.5931,
+      "step": 9499
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0867558717727661,
+      "learning_rate": 9.035543478493458e-06,
+      "loss": 0.6989,
+      "step": 9500
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7727773189544678,
+      "learning_rate": 9.033694543310968e-06,
+      "loss": 1.0118,
+      "step": 9501
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7828236818313599,
+      "learning_rate": 9.031845641472978e-06,
+      "loss": 0.9518,
+      "step": 9502
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7448800802230835,
+      "learning_rate": 9.02999677304329e-06,
+      "loss": 1.0723,
+      "step": 9503
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7413157224655151,
+      "learning_rate": 9.028147938085705e-06,
+      "loss": 0.9257,
+      "step": 9504
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9028218984603882,
+      "learning_rate": 9.02629913666402e-06,
+      "loss": 0.9451,
+      "step": 9505
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.788261890411377,
+      "learning_rate": 9.024450368842033e-06,
+      "loss": 0.9795,
+      "step": 9506
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7059335708618164,
+      "learning_rate": 9.022601634683539e-06,
+      "loss": 0.9401,
+      "step": 9507
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.094498634338379,
+      "learning_rate": 9.02075293425233e-06,
+      "loss": 1.0031,
+      "step": 9508
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8183048963546753,
+      "learning_rate": 9.018904267612205e-06,
+      "loss": 0.9388,
+      "step": 9509
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0388621091842651,
+      "learning_rate": 9.01705563482695e-06,
+      "loss": 0.5598,
+      "step": 9510
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7996293306350708,
+      "learning_rate": 9.01520703596036e-06,
+      "loss": 0.9998,
+      "step": 9511
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6867812871932983,
+      "learning_rate": 9.013358471076226e-06,
+      "loss": 1.0428,
+      "step": 9512
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6633232831954956,
+      "learning_rate": 9.011509940238335e-06,
+      "loss": 0.9994,
+      "step": 9513
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6506688594818115,
+      "learning_rate": 9.009661443510472e-06,
+      "loss": 0.9509,
+      "step": 9514
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8423984050750732,
+      "learning_rate": 9.007812980956427e-06,
+      "loss": 0.9095,
+      "step": 9515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9332531690597534,
+      "learning_rate": 9.005964552639983e-06,
+      "loss": 0.9607,
+      "step": 9516
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8392046689987183,
+      "learning_rate": 9.004116158624928e-06,
+      "loss": 0.9435,
+      "step": 9517
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7351148128509521,
+      "learning_rate": 9.002267798975044e-06,
+      "loss": 0.8928,
+      "step": 9518
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.748577356338501,
+      "learning_rate": 9.00041947375411e-06,
+      "loss": 0.973,
+      "step": 9519
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.611171841621399,
+      "learning_rate": 8.998571183025906e-06,
+      "loss": 0.9533,
+      "step": 9520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6906951665878296,
+      "learning_rate": 8.996722926854215e-06,
+      "loss": 0.9997,
+      "step": 9521
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8013076782226562,
+      "learning_rate": 8.994874705302814e-06,
+      "loss": 0.8942,
+      "step": 9522
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.580102801322937,
+      "learning_rate": 8.993026518435477e-06,
+      "loss": 1.0187,
+      "step": 9523
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.818587303161621,
+      "learning_rate": 8.991178366315982e-06,
+      "loss": 0.9452,
+      "step": 9524
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6145761013031006,
+      "learning_rate": 8.989330249008106e-06,
+      "loss": 0.8558,
+      "step": 9525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8406963348388672,
+      "learning_rate": 8.987482166575618e-06,
+      "loss": 0.9711,
+      "step": 9526
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7605501413345337,
+      "learning_rate": 8.985634119082289e-06,
+      "loss": 0.9203,
+      "step": 9527
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6978834867477417,
+      "learning_rate": 8.983786106591897e-06,
+      "loss": 1.0072,
+      "step": 9528
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6757129430770874,
+      "learning_rate": 8.981938129168208e-06,
+      "loss": 0.8962,
+      "step": 9529
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9993942975997925,
+      "learning_rate": 8.980090186874989e-06,
+      "loss": 0.9645,
+      "step": 9530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.742680311203003,
+      "learning_rate": 8.978242279776009e-06,
+      "loss": 0.8854,
+      "step": 9531
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7286977767944336,
+      "learning_rate": 8.976394407935034e-06,
+      "loss": 0.9304,
+      "step": 9532
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7718924283981323,
+      "learning_rate": 8.974546571415829e-06,
+      "loss": 0.9673,
+      "step": 9533
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8872159719467163,
+      "learning_rate": 8.972698770282156e-06,
+      "loss": 0.8919,
+      "step": 9534
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7218289375305176,
+      "learning_rate": 8.97085100459778e-06,
+      "loss": 0.9493,
+      "step": 9535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7465647459030151,
+      "learning_rate": 8.96900327442646e-06,
+      "loss": 0.9487,
+      "step": 9536
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8240687847137451,
+      "learning_rate": 8.967155579831959e-06,
+      "loss": 0.9612,
+      "step": 9537
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7849608659744263,
+      "learning_rate": 8.965307920878033e-06,
+      "loss": 0.9452,
+      "step": 9538
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7726281881332397,
+      "learning_rate": 8.963460297628437e-06,
+      "loss": 0.9387,
+      "step": 9539
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.056203842163086,
+      "learning_rate": 8.961612710146934e-06,
+      "loss": 0.6264,
+      "step": 9540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0727276802062988,
+      "learning_rate": 8.959765158497275e-06,
+      "loss": 0.6305,
+      "step": 9541
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7222744226455688,
+      "learning_rate": 8.957917642743214e-06,
+      "loss": 0.9104,
+      "step": 9542
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7554842233657837,
+      "learning_rate": 8.956070162948505e-06,
+      "loss": 0.9713,
+      "step": 9543
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.740689754486084,
+      "learning_rate": 8.954222719176898e-06,
+      "loss": 0.9424,
+      "step": 9544
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7328921556472778,
+      "learning_rate": 8.952375311492142e-06,
+      "loss": 1.0324,
+      "step": 9545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.779259204864502,
+      "learning_rate": 8.95052793995799e-06,
+      "loss": 0.9891,
+      "step": 9546
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7910897731781006,
+      "learning_rate": 8.948680604638188e-06,
+      "loss": 0.8713,
+      "step": 9547
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.660340428352356,
+      "learning_rate": 8.946833305596481e-06,
+      "loss": 0.8751,
+      "step": 9548
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8062788248062134,
+      "learning_rate": 8.944986042896615e-06,
+      "loss": 0.9456,
+      "step": 9549
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8999450206756592,
+      "learning_rate": 8.943138816602333e-06,
+      "loss": 0.977,
+      "step": 9550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8570343255996704,
+      "learning_rate": 8.941291626777378e-06,
+      "loss": 0.9803,
+      "step": 9551
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8765219449996948,
+      "learning_rate": 8.939444473485492e-06,
+      "loss": 1.0303,
+      "step": 9552
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.908486008644104,
+      "learning_rate": 8.937597356790414e-06,
+      "loss": 0.8866,
+      "step": 9553
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.767295241355896,
+      "learning_rate": 8.935750276755884e-06,
+      "loss": 1.0025,
+      "step": 9554
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.754777193069458,
+      "learning_rate": 8.93390323344564e-06,
+      "loss": 0.909,
+      "step": 9555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.950538992881775,
+      "learning_rate": 8.932056226923416e-06,
+      "loss": 0.9791,
+      "step": 9556
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0358662605285645,
+      "learning_rate": 8.930209257252948e-06,
+      "loss": 0.6559,
+      "step": 9557
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9117839336395264,
+      "learning_rate": 8.92836232449797e-06,
+      "loss": 0.9522,
+      "step": 9558
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0651201009750366,
+      "learning_rate": 8.926515428722217e-06,
+      "loss": 0.5732,
+      "step": 9559
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6351854801177979,
+      "learning_rate": 8.924668569989416e-06,
+      "loss": 0.9464,
+      "step": 9560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.02516508102417,
+      "learning_rate": 8.9228217483633e-06,
+      "loss": 0.5854,
+      "step": 9561
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8900099992752075,
+      "learning_rate": 8.920974963907596e-06,
+      "loss": 1.0156,
+      "step": 9562
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7059491872787476,
+      "learning_rate": 8.919128216686033e-06,
+      "loss": 0.8825,
+      "step": 9563
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8137855529785156,
+      "learning_rate": 8.917281506762335e-06,
+      "loss": 1.0537,
+      "step": 9564
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7275257110595703,
+      "learning_rate": 8.915434834200228e-06,
+      "loss": 0.9196,
+      "step": 9565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9669073820114136,
+      "learning_rate": 8.913588199063435e-06,
+      "loss": 0.8261,
+      "step": 9566
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7783315181732178,
+      "learning_rate": 8.911741601415678e-06,
+      "loss": 0.9535,
+      "step": 9567
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.639341115951538,
+      "learning_rate": 8.909895041320678e-06,
+      "loss": 0.8925,
+      "step": 9568
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6668494939804077,
+      "learning_rate": 8.908048518842154e-06,
+      "loss": 0.9068,
+      "step": 9569
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7862534523010254,
+      "learning_rate": 8.906202034043828e-06,
+      "loss": 0.952,
+      "step": 9570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.742661952972412,
+      "learning_rate": 8.904355586989414e-06,
+      "loss": 0.8981,
+      "step": 9571
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7427294254302979,
+      "learning_rate": 8.902509177742626e-06,
+      "loss": 0.9565,
+      "step": 9572
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6680155992507935,
+      "learning_rate": 8.900662806367182e-06,
+      "loss": 1.0055,
+      "step": 9573
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8078573942184448,
+      "learning_rate": 8.898816472926795e-06,
+      "loss": 0.8967,
+      "step": 9574
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1516464948654175,
+      "learning_rate": 8.896970177485174e-06,
+      "loss": 0.5923,
+      "step": 9575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8302619457244873,
+      "learning_rate": 8.895123920106033e-06,
+      "loss": 0.9869,
+      "step": 9576
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7211086750030518,
+      "learning_rate": 8.893277700853077e-06,
+      "loss": 0.9434,
+      "step": 9577
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9084489345550537,
+      "learning_rate": 8.891431519790017e-06,
+      "loss": 0.9876,
+      "step": 9578
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7240585088729858,
+      "learning_rate": 8.889585376980557e-06,
+      "loss": 1.0356,
+      "step": 9579
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8124111890792847,
+      "learning_rate": 8.887739272488407e-06,
+      "loss": 1.017,
+      "step": 9580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6830075979232788,
+      "learning_rate": 8.885893206377263e-06,
+      "loss": 0.9794,
+      "step": 9581
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.749099612236023,
+      "learning_rate": 8.884047178710835e-06,
+      "loss": 1.0162,
+      "step": 9582
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7007075548171997,
+      "learning_rate": 8.882201189552821e-06,
+      "loss": 0.93,
+      "step": 9583
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.797532558441162,
+      "learning_rate": 8.880355238966923e-06,
+      "loss": 0.9382,
+      "step": 9584
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7762314081192017,
+      "learning_rate": 8.878509327016838e-06,
+      "loss": 0.9286,
+      "step": 9585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6819589138031006,
+      "learning_rate": 8.876663453766263e-06,
+      "loss": 0.8806,
+      "step": 9586
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7350964546203613,
+      "learning_rate": 8.874817619278893e-06,
+      "loss": 0.991,
+      "step": 9587
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0609499216079712,
+      "learning_rate": 8.872971823618424e-06,
+      "loss": 0.6276,
+      "step": 9588
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.787153959274292,
+      "learning_rate": 8.871126066848552e-06,
+      "loss": 0.9548,
+      "step": 9589
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0259146690368652,
+      "learning_rate": 8.869280349032962e-06,
+      "loss": 0.9922,
+      "step": 9590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8086462020874023,
+      "learning_rate": 8.867434670235352e-06,
+      "loss": 0.9891,
+      "step": 9591
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8335354328155518,
+      "learning_rate": 8.865589030519405e-06,
+      "loss": 1.0128,
+      "step": 9592
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.828158974647522,
+      "learning_rate": 8.863743429948812e-06,
+      "loss": 0.9048,
+      "step": 9593
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8213618993759155,
+      "learning_rate": 8.861897868587262e-06,
+      "loss": 1.0193,
+      "step": 9594
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7731385231018066,
+      "learning_rate": 8.860052346498435e-06,
+      "loss": 0.9123,
+      "step": 9595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6486048698425293,
+      "learning_rate": 8.858206863746018e-06,
+      "loss": 0.907,
+      "step": 9596
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0992177724838257,
+      "learning_rate": 8.856361420393694e-06,
+      "loss": 0.7051,
+      "step": 9597
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6694741249084473,
+      "learning_rate": 8.85451601650514e-06,
+      "loss": 0.8699,
+      "step": 9598
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7299197912216187,
+      "learning_rate": 8.85267065214404e-06,
+      "loss": 0.9088,
+      "step": 9599
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8383482694625854,
+      "learning_rate": 8.85082532737407e-06,
+      "loss": 0.9431,
+      "step": 9600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8750901222229004,
+      "learning_rate": 8.84898004225891e-06,
+      "loss": 1.0146,
+      "step": 9601
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8442833423614502,
+      "learning_rate": 8.847134796862232e-06,
+      "loss": 0.9215,
+      "step": 9602
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9772388935089111,
+      "learning_rate": 8.845289591247713e-06,
+      "loss": 0.9364,
+      "step": 9603
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9751039743423462,
+      "learning_rate": 8.843444425479023e-06,
+      "loss": 0.4999,
+      "step": 9604
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7800713777542114,
+      "learning_rate": 8.841599299619834e-06,
+      "loss": 0.9667,
+      "step": 9605
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5958842039108276,
+      "learning_rate": 8.839754213733818e-06,
+      "loss": 0.9445,
+      "step": 9606
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7308917045593262,
+      "learning_rate": 8.837909167884646e-06,
+      "loss": 1.0054,
+      "step": 9607
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9478790760040283,
+      "learning_rate": 8.836064162135977e-06,
+      "loss": 1.0303,
+      "step": 9608
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7335511445999146,
+      "learning_rate": 8.834219196551486e-06,
+      "loss": 0.9714,
+      "step": 9609
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6473861932754517,
+      "learning_rate": 8.832374271194834e-06,
+      "loss": 0.9467,
+      "step": 9610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9433598518371582,
+      "learning_rate": 8.830529386129683e-06,
+      "loss": 1.0636,
+      "step": 9611
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9085807800292969,
+      "learning_rate": 8.828684541419696e-06,
+      "loss": 0.8839,
+      "step": 9612
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8517625331878662,
+      "learning_rate": 8.826839737128537e-06,
+      "loss": 0.9621,
+      "step": 9613
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8316911458969116,
+      "learning_rate": 8.824994973319859e-06,
+      "loss": 0.9305,
+      "step": 9614
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7378853559494019,
+      "learning_rate": 8.823150250057323e-06,
+      "loss": 0.9346,
+      "step": 9615
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7135299444198608,
+      "learning_rate": 8.821305567404583e-06,
+      "loss": 1.0465,
+      "step": 9616
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.796515941619873,
+      "learning_rate": 8.819460925425297e-06,
+      "loss": 0.996,
+      "step": 9617
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7214481830596924,
+      "learning_rate": 8.817616324183116e-06,
+      "loss": 0.89,
+      "step": 9618
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6071430444717407,
+      "learning_rate": 8.815771763741694e-06,
+      "loss": 0.9728,
+      "step": 9619
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9085267782211304,
+      "learning_rate": 8.81392724416468e-06,
+      "loss": 0.9682,
+      "step": 9620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8057479858398438,
+      "learning_rate": 8.812082765515722e-06,
+      "loss": 0.947,
+      "step": 9621
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8132497072219849,
+      "learning_rate": 8.810238327858471e-06,
+      "loss": 0.9344,
+      "step": 9622
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8121356964111328,
+      "learning_rate": 8.80839393125657e-06,
+      "loss": 1.0095,
+      "step": 9623
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6763031482696533,
+      "learning_rate": 8.806549575773667e-06,
+      "loss": 0.9563,
+      "step": 9624
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7980499267578125,
+      "learning_rate": 8.804705261473405e-06,
+      "loss": 0.9171,
+      "step": 9625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6353436708450317,
+      "learning_rate": 8.802860988419427e-06,
+      "loss": 1.0482,
+      "step": 9626
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6753578186035156,
+      "learning_rate": 8.801016756675368e-06,
+      "loss": 0.9315,
+      "step": 9627
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7558215856552124,
+      "learning_rate": 8.799172566304874e-06,
+      "loss": 0.9647,
+      "step": 9628
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0095503330230713,
+      "learning_rate": 8.797328417371581e-06,
+      "loss": 1.0234,
+      "step": 9629
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7538678646087646,
+      "learning_rate": 8.795484309939124e-06,
+      "loss": 1.0021,
+      "step": 9630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9550825357437134,
+      "learning_rate": 8.793640244071139e-06,
+      "loss": 0.9885,
+      "step": 9631
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7170133590698242,
+      "learning_rate": 8.791796219831259e-06,
+      "loss": 0.9583,
+      "step": 9632
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7709131240844727,
+      "learning_rate": 8.789952237283117e-06,
+      "loss": 0.9078,
+      "step": 9633
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7496020793914795,
+      "learning_rate": 8.788108296490343e-06,
+      "loss": 1.0123,
+      "step": 9634
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.831620454788208,
+      "learning_rate": 8.786264397516564e-06,
+      "loss": 0.9095,
+      "step": 9635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6673592329025269,
+      "learning_rate": 8.784420540425413e-06,
+      "loss": 0.9155,
+      "step": 9636
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6360702514648438,
+      "learning_rate": 8.782576725280513e-06,
+      "loss": 0.9543,
+      "step": 9637
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.646201729774475,
+      "learning_rate": 8.78073295214549e-06,
+      "loss": 0.8538,
+      "step": 9638
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7288867235183716,
+      "learning_rate": 8.778889221083966e-06,
+      "loss": 0.9037,
+      "step": 9639
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.034565806388855,
+      "learning_rate": 8.777045532159564e-06,
+      "loss": 0.6406,
+      "step": 9640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.004770278930664,
+      "learning_rate": 8.775201885435906e-06,
+      "loss": 1.0462,
+      "step": 9641
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7726222276687622,
+      "learning_rate": 8.773358280976607e-06,
+      "loss": 1.0457,
+      "step": 9642
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6694080829620361,
+      "learning_rate": 8.77151471884529e-06,
+      "loss": 0.9629,
+      "step": 9643
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.769039273262024,
+      "learning_rate": 8.769671199105566e-06,
+      "loss": 0.9704,
+      "step": 9644
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.746877908706665,
+      "learning_rate": 8.767827721821054e-06,
+      "loss": 0.9845,
+      "step": 9645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6474612951278687,
+      "learning_rate": 8.765984287055364e-06,
+      "loss": 0.9397,
+      "step": 9646
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8083648681640625,
+      "learning_rate": 8.764140894872108e-06,
+      "loss": 1.0045,
+      "step": 9647
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8130719661712646,
+      "learning_rate": 8.7622975453349e-06,
+      "loss": 0.9449,
+      "step": 9648
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.750686526298523,
+      "learning_rate": 8.760454238507345e-06,
+      "loss": 0.9287,
+      "step": 9649
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8263676166534424,
+      "learning_rate": 8.758610974453052e-06,
+      "loss": 0.9459,
+      "step": 9650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8185365200042725,
+      "learning_rate": 8.756767753235628e-06,
+      "loss": 0.9605,
+      "step": 9651
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7819435596466064,
+      "learning_rate": 8.754924574918675e-06,
+      "loss": 0.9143,
+      "step": 9652
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.625867486000061,
+      "learning_rate": 8.753081439565795e-06,
+      "loss": 0.96,
+      "step": 9653
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9162887334823608,
+      "learning_rate": 8.751238347240595e-06,
+      "loss": 0.9645,
+      "step": 9654
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.726010799407959,
+      "learning_rate": 8.749395298006668e-06,
+      "loss": 0.929,
+      "step": 9655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6468002796173096,
+      "learning_rate": 8.74755229192762e-06,
+      "loss": 0.8925,
+      "step": 9656
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6392147541046143,
+      "learning_rate": 8.74570932906704e-06,
+      "loss": 0.9376,
+      "step": 9657
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7753158807754517,
+      "learning_rate": 8.743866409488529e-06,
+      "loss": 1.0555,
+      "step": 9658
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8286948204040527,
+      "learning_rate": 8.742023533255677e-06,
+      "loss": 0.8981,
+      "step": 9659
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6443310976028442,
+      "learning_rate": 8.74018070043208e-06,
+      "loss": 0.9228,
+      "step": 9660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.683529257774353,
+      "learning_rate": 8.738337911081329e-06,
+      "loss": 0.9325,
+      "step": 9661
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0657858848571777,
+      "learning_rate": 8.736495165267012e-06,
+      "loss": 0.5665,
+      "step": 9662
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7507518529891968,
+      "learning_rate": 8.734652463052717e-06,
+      "loss": 1.0048,
+      "step": 9663
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0817203521728516,
+      "learning_rate": 8.732809804502032e-06,
+      "loss": 1.0015,
+      "step": 9664
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.75890052318573,
+      "learning_rate": 8.73096718967854e-06,
+      "loss": 0.9469,
+      "step": 9665
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9110064506530762,
+      "learning_rate": 8.729124618645827e-06,
+      "loss": 0.9162,
+      "step": 9666
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8386422395706177,
+      "learning_rate": 8.727282091467472e-06,
+      "loss": 0.9129,
+      "step": 9667
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1576179265975952,
+      "learning_rate": 8.725439608207056e-06,
+      "loss": 0.6172,
+      "step": 9668
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7457475662231445,
+      "learning_rate": 8.723597168928159e-06,
+      "loss": 0.9466,
+      "step": 9669
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8802131414413452,
+      "learning_rate": 8.72175477369436e-06,
+      "loss": 0.9078,
+      "step": 9670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.935111165046692,
+      "learning_rate": 8.719912422569232e-06,
+      "loss": 1.0129,
+      "step": 9671
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6842056512832642,
+      "learning_rate": 8.718070115616348e-06,
+      "loss": 0.9663,
+      "step": 9672
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.790739893913269,
+      "learning_rate": 8.716227852899286e-06,
+      "loss": 0.8655,
+      "step": 9673
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7785013914108276,
+      "learning_rate": 8.714385634481613e-06,
+      "loss": 0.9837,
+      "step": 9674
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8138951063156128,
+      "learning_rate": 8.712543460426901e-06,
+      "loss": 0.9868,
+      "step": 9675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9091839790344238,
+      "learning_rate": 8.71070133079872e-06,
+      "loss": 0.9166,
+      "step": 9676
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2488070726394653,
+      "learning_rate": 8.70885924566063e-06,
+      "loss": 0.6598,
+      "step": 9677
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8477082252502441,
+      "learning_rate": 8.707017205076205e-06,
+      "loss": 0.9076,
+      "step": 9678
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7101072072982788,
+      "learning_rate": 8.705175209109003e-06,
+      "loss": 1.0161,
+      "step": 9679
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.840926170349121,
+      "learning_rate": 8.70333325782259e-06,
+      "loss": 0.9264,
+      "step": 9680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7042949199676514,
+      "learning_rate": 8.701491351280521e-06,
+      "loss": 0.924,
+      "step": 9681
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8257490396499634,
+      "learning_rate": 8.69964948954636e-06,
+      "loss": 0.9654,
+      "step": 9682
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7320343255996704,
+      "learning_rate": 8.697807672683662e-06,
+      "loss": 0.9265,
+      "step": 9683
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6063027381896973,
+      "learning_rate": 8.695965900755985e-06,
+      "loss": 0.9934,
+      "step": 9684
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.703420877456665,
+      "learning_rate": 8.694124173826881e-06,
+      "loss": 0.899,
+      "step": 9685
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5832927227020264,
+      "learning_rate": 8.692282491959904e-06,
+      "loss": 0.9626,
+      "step": 9686
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7099549770355225,
+      "learning_rate": 8.690440855218606e-06,
+      "loss": 0.9516,
+      "step": 9687
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8984639644622803,
+      "learning_rate": 8.688599263666536e-06,
+      "loss": 0.9517,
+      "step": 9688
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.898252248764038,
+      "learning_rate": 8.68675771736724e-06,
+      "loss": 0.934,
+      "step": 9689
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.888367772102356,
+      "learning_rate": 8.684916216384268e-06,
+      "loss": 0.9491,
+      "step": 9690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8562384843826294,
+      "learning_rate": 8.683074760781163e-06,
+      "loss": 1.0361,
+      "step": 9691
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.880313754081726,
+      "learning_rate": 8.681233350621472e-06,
+      "loss": 0.9818,
+      "step": 9692
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7483007907867432,
+      "learning_rate": 8.679391985968732e-06,
+      "loss": 1.0765,
+      "step": 9693
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6772500276565552,
+      "learning_rate": 8.677550666886486e-06,
+      "loss": 0.9398,
+      "step": 9694
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8419322967529297,
+      "learning_rate": 8.675709393438273e-06,
+      "loss": 0.953,
+      "step": 9695
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7664134502410889,
+      "learning_rate": 8.67386816568763e-06,
+      "loss": 0.968,
+      "step": 9696
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7157539129257202,
+      "learning_rate": 8.67202698369809e-06,
+      "loss": 0.9796,
+      "step": 9697
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7701255083084106,
+      "learning_rate": 8.67018584753319e-06,
+      "loss": 0.9877,
+      "step": 9698
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.855797290802002,
+      "learning_rate": 8.668344757256464e-06,
+      "loss": 0.9418,
+      "step": 9699
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8157871961593628,
+      "learning_rate": 8.666503712931439e-06,
+      "loss": 0.9642,
+      "step": 9700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7075386047363281,
+      "learning_rate": 8.664662714621643e-06,
+      "loss": 0.9357,
+      "step": 9701
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.830863118171692,
+      "learning_rate": 8.66282176239061e-06,
+      "loss": 0.8789,
+      "step": 9702
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7107125520706177,
+      "learning_rate": 8.660980856301862e-06,
+      "loss": 0.9236,
+      "step": 9703
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0277130603790283,
+      "learning_rate": 8.659139996418925e-06,
+      "loss": 0.5594,
+      "step": 9704
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6675851345062256,
+      "learning_rate": 8.65729918280532e-06,
+      "loss": 0.9527,
+      "step": 9705
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.654502511024475,
+      "learning_rate": 8.655458415524571e-06,
+      "loss": 0.9099,
+      "step": 9706
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9035966396331787,
+      "learning_rate": 8.653617694640196e-06,
+      "loss": 0.9928,
+      "step": 9707
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6585253477096558,
+      "learning_rate": 8.651777020215713e-06,
+      "loss": 0.9427,
+      "step": 9708
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7313233613967896,
+      "learning_rate": 8.649936392314638e-06,
+      "loss": 0.9722,
+      "step": 9709
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5806974172592163,
+      "learning_rate": 8.648095811000488e-06,
+      "loss": 0.8853,
+      "step": 9710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.775068998336792,
+      "learning_rate": 8.646255276336775e-06,
+      "loss": 0.9058,
+      "step": 9711
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.642682671546936,
+      "learning_rate": 8.644414788387012e-06,
+      "loss": 0.9213,
+      "step": 9712
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8018274307250977,
+      "learning_rate": 8.642574347214702e-06,
+      "loss": 0.9542,
+      "step": 9713
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6730376482009888,
+      "learning_rate": 8.640733952883365e-06,
+      "loss": 0.9341,
+      "step": 9714
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6986688375473022,
+      "learning_rate": 8.638893605456502e-06,
+      "loss": 0.9835,
+      "step": 9715
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.712127447128296,
+      "learning_rate": 8.637053304997618e-06,
+      "loss": 1.0183,
+      "step": 9716
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7850602865219116,
+      "learning_rate": 8.635213051570217e-06,
+      "loss": 1.0927,
+      "step": 9717
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6969475746154785,
+      "learning_rate": 8.633372845237803e-06,
+      "loss": 0.921,
+      "step": 9718
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6289851665496826,
+      "learning_rate": 8.631532686063871e-06,
+      "loss": 0.9846,
+      "step": 9719
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.767514705657959,
+      "learning_rate": 8.629692574111926e-06,
+      "loss": 1.0,
+      "step": 9720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7698767185211182,
+      "learning_rate": 8.627852509445462e-06,
+      "loss": 0.9593,
+      "step": 9721
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8133702278137207,
+      "learning_rate": 8.626012492127975e-06,
+      "loss": 0.9646,
+      "step": 9722
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.889578104019165,
+      "learning_rate": 8.624172522222959e-06,
+      "loss": 0.9346,
+      "step": 9723
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0171505212783813,
+      "learning_rate": 8.622332599793906e-06,
+      "loss": 0.5231,
+      "step": 9724
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6942979097366333,
+      "learning_rate": 8.620492724904304e-06,
+      "loss": 0.9819,
+      "step": 9725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6233772039413452,
+      "learning_rate": 8.618652897617646e-06,
+      "loss": 0.9597,
+      "step": 9726
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7178518772125244,
+      "learning_rate": 8.61681311799742e-06,
+      "loss": 1.038,
+      "step": 9727
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.79873788356781,
+      "learning_rate": 8.614973386107107e-06,
+      "loss": 1.0083,
+      "step": 9728
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.760603666305542,
+      "learning_rate": 8.613133702010196e-06,
+      "loss": 1.0058,
+      "step": 9729
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.648322343826294,
+      "learning_rate": 8.611294065770166e-06,
+      "loss": 0.9731,
+      "step": 9730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6967090368270874,
+      "learning_rate": 8.609454477450497e-06,
+      "loss": 0.971,
+      "step": 9731
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.753484845161438,
+      "learning_rate": 8.607614937114671e-06,
+      "loss": 1.0044,
+      "step": 9732
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8227661848068237,
+      "learning_rate": 8.605775444826164e-06,
+      "loss": 0.9016,
+      "step": 9733
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7095043659210205,
+      "learning_rate": 8.603936000648452e-06,
+      "loss": 0.8888,
+      "step": 9734
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6874574422836304,
+      "learning_rate": 8.602096604645009e-06,
+      "loss": 0.9798,
+      "step": 9735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.68003249168396,
+      "learning_rate": 8.600257256879306e-06,
+      "loss": 0.9337,
+      "step": 9736
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9060606956481934,
+      "learning_rate": 8.598417957414817e-06,
+      "loss": 0.962,
+      "step": 9737
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7026256322860718,
+      "learning_rate": 8.596578706315006e-06,
+      "loss": 0.9992,
+      "step": 9738
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8191571235656738,
+      "learning_rate": 8.594739503643345e-06,
+      "loss": 0.9129,
+      "step": 9739
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7354254722595215,
+      "learning_rate": 8.592900349463297e-06,
+      "loss": 0.8972,
+      "step": 9740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7470617294311523,
+      "learning_rate": 8.59106124383833e-06,
+      "loss": 0.9997,
+      "step": 9741
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7821121215820312,
+      "learning_rate": 8.589222186831903e-06,
+      "loss": 1.0658,
+      "step": 9742
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.844748854637146,
+      "learning_rate": 8.587383178507474e-06,
+      "loss": 0.9313,
+      "step": 9743
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.834816336631775,
+      "learning_rate": 8.58554421892851e-06,
+      "loss": 0.8901,
+      "step": 9744
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7154957056045532,
+      "learning_rate": 8.583705308158463e-06,
+      "loss": 0.9643,
+      "step": 9745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7639740705490112,
+      "learning_rate": 8.581866446260789e-06,
+      "loss": 0.9867,
+      "step": 9746
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6491105556488037,
+      "learning_rate": 8.580027633298945e-06,
+      "loss": 0.9272,
+      "step": 9747
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.973211646080017,
+      "learning_rate": 8.578188869336378e-06,
+      "loss": 1.006,
+      "step": 9748
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7114208936691284,
+      "learning_rate": 8.576350154436542e-06,
+      "loss": 0.909,
+      "step": 9749
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7240196466445923,
+      "learning_rate": 8.574511488662886e-06,
+      "loss": 0.9678,
+      "step": 9750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8778692483901978,
+      "learning_rate": 8.572672872078856e-06,
+      "loss": 0.9763,
+      "step": 9751
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8747127056121826,
+      "learning_rate": 8.570834304747898e-06,
+      "loss": 1.0946,
+      "step": 9752
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.891045331954956,
+      "learning_rate": 8.568995786733456e-06,
+      "loss": 0.9068,
+      "step": 9753
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7993545532226562,
+      "learning_rate": 8.567157318098974e-06,
+      "loss": 0.9032,
+      "step": 9754
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6323741674423218,
+      "learning_rate": 8.565318898907886e-06,
+      "loss": 0.9372,
+      "step": 9755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9460618495941162,
+      "learning_rate": 8.563480529223638e-06,
+      "loss": 1.0282,
+      "step": 9756
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7004189491271973,
+      "learning_rate": 8.561642209109664e-06,
+      "loss": 0.9513,
+      "step": 9757
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8786660432815552,
+      "learning_rate": 8.5598039386294e-06,
+      "loss": 1.0091,
+      "step": 9758
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1535083055496216,
+      "learning_rate": 8.557965717846278e-06,
+      "loss": 0.6237,
+      "step": 9759
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7741538286209106,
+      "learning_rate": 8.556127546823732e-06,
+      "loss": 0.9286,
+      "step": 9760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7362631559371948,
+      "learning_rate": 8.554289425625191e-06,
+      "loss": 1.0297,
+      "step": 9761
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7978144884109497,
+      "learning_rate": 8.552451354314083e-06,
+      "loss": 0.9479,
+      "step": 9762
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6696869134902954,
+      "learning_rate": 8.550613332953835e-06,
+      "loss": 0.9023,
+      "step": 9763
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9857523441314697,
+      "learning_rate": 8.548775361607872e-06,
+      "loss": 1.014,
+      "step": 9764
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7713627815246582,
+      "learning_rate": 8.54693744033962e-06,
+      "loss": 0.8942,
+      "step": 9765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7986689805984497,
+      "learning_rate": 8.545099569212496e-06,
+      "loss": 0.9259,
+      "step": 9766
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7333914041519165,
+      "learning_rate": 8.543261748289919e-06,
+      "loss": 0.8948,
+      "step": 9767
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8085017204284668,
+      "learning_rate": 8.541423977635313e-06,
+      "loss": 0.9572,
+      "step": 9768
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.556166410446167,
+      "learning_rate": 8.539586257312091e-06,
+      "loss": 0.945,
+      "step": 9769
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.851192831993103,
+      "learning_rate": 8.537748587383667e-06,
+      "loss": 0.9588,
+      "step": 9770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.839781641960144,
+      "learning_rate": 8.535910967913454e-06,
+      "loss": 0.9402,
+      "step": 9771
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8548117876052856,
+      "learning_rate": 8.534073398964866e-06,
+      "loss": 0.9723,
+      "step": 9772
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.003147602081299,
+      "learning_rate": 8.532235880601309e-06,
+      "loss": 1.0372,
+      "step": 9773
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7658926248550415,
+      "learning_rate": 8.530398412886192e-06,
+      "loss": 0.8876,
+      "step": 9774
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7920811176300049,
+      "learning_rate": 8.52856099588292e-06,
+      "loss": 0.8905,
+      "step": 9775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8868311643600464,
+      "learning_rate": 8.526723629654898e-06,
+      "loss": 0.9874,
+      "step": 9776
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6354767084121704,
+      "learning_rate": 8.524886314265527e-06,
+      "loss": 1.015,
+      "step": 9777
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.606435775756836,
+      "learning_rate": 8.523049049778212e-06,
+      "loss": 0.8059,
+      "step": 9778
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6480079889297485,
+      "learning_rate": 8.521211836256343e-06,
+      "loss": 0.9127,
+      "step": 9779
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6213204860687256,
+      "learning_rate": 8.519374673763326e-06,
+      "loss": 0.9295,
+      "step": 9780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8984014987945557,
+      "learning_rate": 8.517537562362554e-06,
+      "loss": 0.944,
+      "step": 9781
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.905485987663269,
+      "learning_rate": 8.515700502117418e-06,
+      "loss": 0.9928,
+      "step": 9782
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5779186487197876,
+      "learning_rate": 8.513863493091313e-06,
+      "loss": 0.9735,
+      "step": 9783
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8514306545257568,
+      "learning_rate": 8.512026535347627e-06,
+      "loss": 0.9884,
+      "step": 9784
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.719305396080017,
+      "learning_rate": 8.51018962894975e-06,
+      "loss": 0.9532,
+      "step": 9785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6308788061141968,
+      "learning_rate": 8.508352773961063e-06,
+      "loss": 0.9655,
+      "step": 9786
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6522870063781738,
+      "learning_rate": 8.50651597044496e-06,
+      "loss": 0.8783,
+      "step": 9787
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7631512880325317,
+      "learning_rate": 8.504679218464816e-06,
+      "loss": 0.9546,
+      "step": 9788
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6883199214935303,
+      "learning_rate": 8.502842518084015e-06,
+      "loss": 0.9983,
+      "step": 9789
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8403781652450562,
+      "learning_rate": 8.501005869365939e-06,
+      "loss": 0.9601,
+      "step": 9790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1367347240447998,
+      "learning_rate": 8.499169272373961e-06,
+      "loss": 0.6541,
+      "step": 9791
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7039514780044556,
+      "learning_rate": 8.497332727171458e-06,
+      "loss": 0.8586,
+      "step": 9792
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.794058918952942,
+      "learning_rate": 8.495496233821808e-06,
+      "loss": 1.0559,
+      "step": 9793
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6120576858520508,
+      "learning_rate": 8.493659792388378e-06,
+      "loss": 1.0517,
+      "step": 9794
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5673608779907227,
+      "learning_rate": 8.491823402934542e-06,
+      "loss": 0.8646,
+      "step": 9795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6555083990097046,
+      "learning_rate": 8.489987065523668e-06,
+      "loss": 0.9611,
+      "step": 9796
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9197218418121338,
+      "learning_rate": 8.488150780219122e-06,
+      "loss": 0.9224,
+      "step": 9797
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3184118270874023,
+      "learning_rate": 8.48631454708427e-06,
+      "loss": 1.0038,
+      "step": 9798
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0872104167938232,
+      "learning_rate": 8.484478366182472e-06,
+      "loss": 0.5469,
+      "step": 9799
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7852264642715454,
+      "learning_rate": 8.482642237577094e-06,
+      "loss": 0.9636,
+      "step": 9800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.748695731163025,
+      "learning_rate": 8.480806161331494e-06,
+      "loss": 0.9493,
+      "step": 9801
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.024226427078247,
+      "learning_rate": 8.478970137509029e-06,
+      "loss": 0.6269,
+      "step": 9802
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.771871566772461,
+      "learning_rate": 8.477134166173057e-06,
+      "loss": 0.9976,
+      "step": 9803
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6688435077667236,
+      "learning_rate": 8.475298247386927e-06,
+      "loss": 1.072,
+      "step": 9804
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.708177924156189,
+      "learning_rate": 8.473462381213999e-06,
+      "loss": 1.0109,
+      "step": 9805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9090077877044678,
+      "learning_rate": 8.471626567717617e-06,
+      "loss": 0.9891,
+      "step": 9806
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7106572389602661,
+      "learning_rate": 8.469790806961136e-06,
+      "loss": 0.9655,
+      "step": 9807
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7648698091506958,
+      "learning_rate": 8.467955099007899e-06,
+      "loss": 0.9127,
+      "step": 9808
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8491175174713135,
+      "learning_rate": 8.466119443921249e-06,
+      "loss": 0.9318,
+      "step": 9809
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.848143458366394,
+      "learning_rate": 8.464283841764536e-06,
+      "loss": 0.9889,
+      "step": 9810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8523794412612915,
+      "learning_rate": 8.462448292601096e-06,
+      "loss": 0.981,
+      "step": 9811
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6393706798553467,
+      "learning_rate": 8.460612796494272e-06,
+      "loss": 0.8938,
+      "step": 9812
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6241719722747803,
+      "learning_rate": 8.4587773535074e-06,
+      "loss": 0.9579,
+      "step": 9813
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8510385751724243,
+      "learning_rate": 8.456941963703817e-06,
+      "loss": 1.04,
+      "step": 9814
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6346877813339233,
+      "learning_rate": 8.455106627146855e-06,
+      "loss": 0.868,
+      "step": 9815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8974014520645142,
+      "learning_rate": 8.453271343899849e-06,
+      "loss": 0.9612,
+      "step": 9816
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6034828424453735,
+      "learning_rate": 8.451436114026127e-06,
+      "loss": 0.9871,
+      "step": 9817
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7454249858856201,
+      "learning_rate": 8.449600937589019e-06,
+      "loss": 0.9252,
+      "step": 9818
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6624367237091064,
+      "learning_rate": 8.447765814651853e-06,
+      "loss": 0.9011,
+      "step": 9819
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7782756090164185,
+      "learning_rate": 8.445930745277953e-06,
+      "loss": 1.0924,
+      "step": 9820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7412757873535156,
+      "learning_rate": 8.444095729530638e-06,
+      "loss": 0.9848,
+      "step": 9821
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0630688667297363,
+      "learning_rate": 8.442260767473236e-06,
+      "loss": 0.5285,
+      "step": 9822
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6591832637786865,
+      "learning_rate": 8.440425859169064e-06,
+      "loss": 0.9191,
+      "step": 9823
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.804245948791504,
+      "learning_rate": 8.438591004681439e-06,
+      "loss": 0.9802,
+      "step": 9824
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0639244318008423,
+      "learning_rate": 8.436756204073676e-06,
+      "loss": 0.7007,
+      "step": 9825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6975889205932617,
+      "learning_rate": 8.434921457409091e-06,
+      "loss": 0.9165,
+      "step": 9826
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0042953491210938,
+      "learning_rate": 8.433086764750993e-06,
+      "loss": 0.5333,
+      "step": 9827
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.662895917892456,
+      "learning_rate": 8.431252126162695e-06,
+      "loss": 0.9927,
+      "step": 9828
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.726589322090149,
+      "learning_rate": 8.429417541707505e-06,
+      "loss": 0.9874,
+      "step": 9829
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6898778676986694,
+      "learning_rate": 8.427583011448725e-06,
+      "loss": 0.8727,
+      "step": 9830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6102042198181152,
+      "learning_rate": 8.425748535449666e-06,
+      "loss": 0.9873,
+      "step": 9831
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4961017370224,
+      "learning_rate": 8.423914113773627e-06,
+      "loss": 1.0199,
+      "step": 9832
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7345589399337769,
+      "learning_rate": 8.422079746483907e-06,
+      "loss": 0.8976,
+      "step": 9833
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8311835527420044,
+      "learning_rate": 8.420245433643807e-06,
+      "loss": 1.0025,
+      "step": 9834
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7003817558288574,
+      "learning_rate": 8.418411175316627e-06,
+      "loss": 1.0129,
+      "step": 9835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6672784090042114,
+      "learning_rate": 8.41657697156566e-06,
+      "loss": 0.938,
+      "step": 9836
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7642420530319214,
+      "learning_rate": 8.414742822454197e-06,
+      "loss": 0.9119,
+      "step": 9837
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7299810647964478,
+      "learning_rate": 8.41290872804553e-06,
+      "loss": 0.9908,
+      "step": 9838
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5857008695602417,
+      "learning_rate": 8.411074688402952e-06,
+      "loss": 0.9169,
+      "step": 9839
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5904239416122437,
+      "learning_rate": 8.409240703589746e-06,
+      "loss": 0.9951,
+      "step": 9840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7446154356002808,
+      "learning_rate": 8.4074067736692e-06,
+      "loss": 0.9103,
+      "step": 9841
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9121451377868652,
+      "learning_rate": 8.405572898704598e-06,
+      "loss": 0.9484,
+      "step": 9842
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6011571884155273,
+      "learning_rate": 8.403739078759221e-06,
+      "loss": 0.8354,
+      "step": 9843
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7281208038330078,
+      "learning_rate": 8.40190531389635e-06,
+      "loss": 0.9697,
+      "step": 9844
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7923554182052612,
+      "learning_rate": 8.40007160417926e-06,
+      "loss": 1.0145,
+      "step": 9845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9364678859710693,
+      "learning_rate": 8.398237949671231e-06,
+      "loss": 1.012,
+      "step": 9846
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9073313474655151,
+      "learning_rate": 8.396404350435539e-06,
+      "loss": 0.9566,
+      "step": 9847
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9555952548980713,
+      "learning_rate": 8.39457080653545e-06,
+      "loss": 0.9763,
+      "step": 9848
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8375389575958252,
+      "learning_rate": 8.392737318034239e-06,
+      "loss": 1.0071,
+      "step": 9849
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7012312412261963,
+      "learning_rate": 8.390903884995174e-06,
+      "loss": 0.9439,
+      "step": 9850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7312678098678589,
+      "learning_rate": 8.389070507481522e-06,
+      "loss": 0.8852,
+      "step": 9851
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7921741008758545,
+      "learning_rate": 8.387237185556544e-06,
+      "loss": 1.0518,
+      "step": 9852
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.680347204208374,
+      "learning_rate": 8.385403919283508e-06,
+      "loss": 0.9043,
+      "step": 9853
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8480693101882935,
+      "learning_rate": 8.383570708725672e-06,
+      "loss": 0.9366,
+      "step": 9854
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5710793733596802,
+      "learning_rate": 8.381737553946296e-06,
+      "loss": 0.9025,
+      "step": 9855
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7552309036254883,
+      "learning_rate": 8.379904455008635e-06,
+      "loss": 0.9226,
+      "step": 9856
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6855357885360718,
+      "learning_rate": 8.378071411975947e-06,
+      "loss": 0.9774,
+      "step": 9857
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.665492296218872,
+      "learning_rate": 8.376238424911481e-06,
+      "loss": 0.9258,
+      "step": 9858
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7533986568450928,
+      "learning_rate": 8.374405493878494e-06,
+      "loss": 0.9493,
+      "step": 9859
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9120346307754517,
+      "learning_rate": 8.372572618940232e-06,
+      "loss": 0.9401,
+      "step": 9860
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6676182746887207,
+      "learning_rate": 8.370739800159944e-06,
+      "loss": 1.0289,
+      "step": 9861
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.775700330734253,
+      "learning_rate": 8.368907037600873e-06,
+      "loss": 0.9437,
+      "step": 9862
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6866471767425537,
+      "learning_rate": 8.367074331326264e-06,
+      "loss": 0.9376,
+      "step": 9863
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5305966138839722,
+      "learning_rate": 8.365241681399359e-06,
+      "loss": 0.9331,
+      "step": 9864
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.730548620223999,
+      "learning_rate": 8.363409087883397e-06,
+      "loss": 0.8898,
+      "step": 9865
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7145718336105347,
+      "learning_rate": 8.361576550841615e-06,
+      "loss": 0.8388,
+      "step": 9866
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.746213436126709,
+      "learning_rate": 8.35974407033725e-06,
+      "loss": 0.9281,
+      "step": 9867
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6540430784225464,
+      "learning_rate": 8.357911646433534e-06,
+      "loss": 0.9637,
+      "step": 9868
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9239869117736816,
+      "learning_rate": 8.356079279193703e-06,
+      "loss": 0.9915,
+      "step": 9869
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7954456806182861,
+      "learning_rate": 8.35424696868098e-06,
+      "loss": 0.9518,
+      "step": 9870
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0025910139083862,
+      "learning_rate": 8.3524147149586e-06,
+      "loss": 0.6069,
+      "step": 9871
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7625956535339355,
+      "learning_rate": 8.350582518089781e-06,
+      "loss": 0.9031,
+      "step": 9872
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.857361912727356,
+      "learning_rate": 8.348750378137756e-06,
+      "loss": 0.9891,
+      "step": 9873
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7598695755004883,
+      "learning_rate": 8.346918295165743e-06,
+      "loss": 0.894,
+      "step": 9874
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6191054582595825,
+      "learning_rate": 8.34508626923696e-06,
+      "loss": 0.901,
+      "step": 9875
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8950486183166504,
+      "learning_rate": 8.343254300414629e-06,
+      "loss": 0.9912,
+      "step": 9876
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.622578740119934,
+      "learning_rate": 8.341422388761964e-06,
+      "loss": 0.948,
+      "step": 9877
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1127142906188965,
+      "learning_rate": 8.33959053434218e-06,
+      "loss": 0.588,
+      "step": 9878
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8276445865631104,
+      "learning_rate": 8.337758737218487e-06,
+      "loss": 0.9216,
+      "step": 9879
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5599991083145142,
+      "learning_rate": 8.335926997454097e-06,
+      "loss": 0.9535,
+      "step": 9880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8944109678268433,
+      "learning_rate": 8.334095315112218e-06,
+      "loss": 0.9727,
+      "step": 9881
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.729617714881897,
+      "learning_rate": 8.332263690256056e-06,
+      "loss": 0.9828,
+      "step": 9882
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.883064866065979,
+      "learning_rate": 8.330432122948816e-06,
+      "loss": 0.9903,
+      "step": 9883
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6110795736312866,
+      "learning_rate": 8.3286006132537e-06,
+      "loss": 0.9622,
+      "step": 9884
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9348753690719604,
+      "learning_rate": 8.326769161233907e-06,
+      "loss": 0.9983,
+      "step": 9885
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8125224113464355,
+      "learning_rate": 8.324937766952638e-06,
+      "loss": 0.9462,
+      "step": 9886
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.894012928009033,
+      "learning_rate": 8.323106430473084e-06,
+      "loss": 0.9146,
+      "step": 9887
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.705432653427124,
+      "learning_rate": 8.321275151858445e-06,
+      "loss": 0.9496,
+      "step": 9888
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2724661827087402,
+      "learning_rate": 8.319443931171911e-06,
+      "loss": 0.984,
+      "step": 9889
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.600510597229004,
+      "learning_rate": 8.317612768476673e-06,
+      "loss": 0.9081,
+      "step": 9890
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8133643865585327,
+      "learning_rate": 8.315781663835918e-06,
+      "loss": 0.9857,
+      "step": 9891
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8171542882919312,
+      "learning_rate": 8.313950617312835e-06,
+      "loss": 0.9706,
+      "step": 9892
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.671372413635254,
+      "learning_rate": 8.312119628970605e-06,
+      "loss": 0.9608,
+      "step": 9893
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6408926248550415,
+      "learning_rate": 8.310288698872412e-06,
+      "loss": 0.8968,
+      "step": 9894
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8124191761016846,
+      "learning_rate": 8.308457827081436e-06,
+      "loss": 0.9211,
+      "step": 9895
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.907081127166748,
+      "learning_rate": 8.306627013660856e-06,
+      "loss": 0.9142,
+      "step": 9896
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7713135480880737,
+      "learning_rate": 8.304796258673845e-06,
+      "loss": 0.9305,
+      "step": 9897
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.721226692199707,
+      "learning_rate": 8.302965562183583e-06,
+      "loss": 0.8517,
+      "step": 9898
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7855409383773804,
+      "learning_rate": 8.301134924253233e-06,
+      "loss": 0.9,
+      "step": 9899
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7462619543075562,
+      "learning_rate": 8.299304344945977e-06,
+      "loss": 0.9325,
+      "step": 9900
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9759937524795532,
+      "learning_rate": 8.297473824324976e-06,
+      "loss": 0.9968,
+      "step": 9901
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7004410028457642,
+      "learning_rate": 8.295643362453397e-06,
+      "loss": 0.9483,
+      "step": 9902
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7353190183639526,
+      "learning_rate": 8.293812959394405e-06,
+      "loss": 0.9268,
+      "step": 9903
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8573025465011597,
+      "learning_rate": 8.291982615211163e-06,
+      "loss": 0.9673,
+      "step": 9904
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6955972909927368,
+      "learning_rate": 8.290152329966827e-06,
+      "loss": 0.958,
+      "step": 9905
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.865373969078064,
+      "learning_rate": 8.28832210372456e-06,
+      "loss": 0.9732,
+      "step": 9906
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7247673273086548,
+      "learning_rate": 8.286491936547514e-06,
+      "loss": 0.9136,
+      "step": 9907
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.706994891166687,
+      "learning_rate": 8.284661828498847e-06,
+      "loss": 0.9503,
+      "step": 9908
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6628011465072632,
+      "learning_rate": 8.282831779641708e-06,
+      "loss": 0.9114,
+      "step": 9909
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8326587677001953,
+      "learning_rate": 8.281001790039246e-06,
+      "loss": 0.9767,
+      "step": 9910
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6019359827041626,
+      "learning_rate": 8.27917185975461e-06,
+      "loss": 0.8618,
+      "step": 9911
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.680530309677124,
+      "learning_rate": 8.277341988850949e-06,
+      "loss": 0.9544,
+      "step": 9912
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7658640146255493,
+      "learning_rate": 8.275512177391403e-06,
+      "loss": 0.929,
+      "step": 9913
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.848766565322876,
+      "learning_rate": 8.273682425439114e-06,
+      "loss": 0.9147,
+      "step": 9914
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8004672527313232,
+      "learning_rate": 8.271852733057222e-06,
+      "loss": 0.9916,
+      "step": 9915
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.755155086517334,
+      "learning_rate": 8.270023100308865e-06,
+      "loss": 0.9459,
+      "step": 9916
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9252840280532837,
+      "learning_rate": 8.26819352725718e-06,
+      "loss": 0.9535,
+      "step": 9917
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.923312783241272,
+      "learning_rate": 8.266364013965297e-06,
+      "loss": 0.5343,
+      "step": 9918
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6745131015777588,
+      "learning_rate": 8.26453456049635e-06,
+      "loss": 0.9126,
+      "step": 9919
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7967758178710938,
+      "learning_rate": 8.262705166913467e-06,
+      "loss": 0.8683,
+      "step": 9920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6511890888214111,
+      "learning_rate": 8.260875833279776e-06,
+      "loss": 0.8783,
+      "step": 9921
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5698761940002441,
+      "learning_rate": 8.259046559658401e-06,
+      "loss": 0.8749,
+      "step": 9922
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8676340579986572,
+      "learning_rate": 8.257217346112468e-06,
+      "loss": 1.0066,
+      "step": 9923
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8737683296203613,
+      "learning_rate": 8.255388192705092e-06,
+      "loss": 0.9175,
+      "step": 9924
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8011157512664795,
+      "learning_rate": 8.2535590994994e-06,
+      "loss": 0.9462,
+      "step": 9925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.796507716178894,
+      "learning_rate": 8.251730066558504e-06,
+      "loss": 0.9056,
+      "step": 9926
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9044804573059082,
+      "learning_rate": 8.24990109394552e-06,
+      "loss": 0.9592,
+      "step": 9927
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7152119874954224,
+      "learning_rate": 8.24807218172356e-06,
+      "loss": 0.85,
+      "step": 9928
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0453922748565674,
+      "learning_rate": 8.246243329955735e-06,
+      "loss": 0.5773,
+      "step": 9929
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6330358982086182,
+      "learning_rate": 8.244414538705155e-06,
+      "loss": 0.9323,
+      "step": 9930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.782004714012146,
+      "learning_rate": 8.242585808034924e-06,
+      "loss": 0.908,
+      "step": 9931
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7026286125183105,
+      "learning_rate": 8.240757138008149e-06,
+      "loss": 0.9192,
+      "step": 9932
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8290990591049194,
+      "learning_rate": 8.23892852868793e-06,
+      "loss": 1.0071,
+      "step": 9933
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7360644340515137,
+      "learning_rate": 8.237099980137368e-06,
+      "loss": 0.9037,
+      "step": 9934
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7625229358673096,
+      "learning_rate": 8.235271492419563e-06,
+      "loss": 0.9103,
+      "step": 9935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.077831983566284,
+      "learning_rate": 8.233443065597605e-06,
+      "loss": 0.9098,
+      "step": 9936
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.783826231956482,
+      "learning_rate": 8.231614699734595e-06,
+      "loss": 0.9188,
+      "step": 9937
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.857846736907959,
+      "learning_rate": 8.22978639489362e-06,
+      "loss": 0.984,
+      "step": 9938
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8796751499176025,
+      "learning_rate": 8.227958151137773e-06,
+      "loss": 0.9687,
+      "step": 9939
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7597626447677612,
+      "learning_rate": 8.22612996853014e-06,
+      "loss": 0.9487,
+      "step": 9940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6525269746780396,
+      "learning_rate": 8.224301847133805e-06,
+      "loss": 0.899,
+      "step": 9941
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6960855722427368,
+      "learning_rate": 8.222473787011855e-06,
+      "loss": 0.9641,
+      "step": 9942
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9196513891220093,
+      "learning_rate": 8.22064578822737e-06,
+      "loss": 0.9329,
+      "step": 9943
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6569267511367798,
+      "learning_rate": 8.218817850843428e-06,
+      "loss": 0.9485,
+      "step": 9944
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8788106441497803,
+      "learning_rate": 8.216989974923107e-06,
+      "loss": 0.9351,
+      "step": 9945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7779889106750488,
+      "learning_rate": 8.21516216052948e-06,
+      "loss": 0.9016,
+      "step": 9946
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7726404666900635,
+      "learning_rate": 8.213334407725622e-06,
+      "loss": 0.9259,
+      "step": 9947
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.700067162513733,
+      "learning_rate": 8.211506716574604e-06,
+      "loss": 0.9279,
+      "step": 9948
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6555423736572266,
+      "learning_rate": 8.209679087139491e-06,
+      "loss": 0.9613,
+      "step": 9949
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6706092357635498,
+      "learning_rate": 8.207851519483352e-06,
+      "loss": 0.9813,
+      "step": 9950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.842773199081421,
+      "learning_rate": 8.206024013669253e-06,
+      "loss": 0.9573,
+      "step": 9951
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7060863971710205,
+      "learning_rate": 8.204196569760252e-06,
+      "loss": 0.895,
+      "step": 9952
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.686597228050232,
+      "learning_rate": 8.20236918781941e-06,
+      "loss": 0.9311,
+      "step": 9953
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.580003261566162,
+      "learning_rate": 8.200541867909786e-06,
+      "loss": 0.9929,
+      "step": 9954
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1099209785461426,
+      "learning_rate": 8.198714610094438e-06,
+      "loss": 0.9683,
+      "step": 9955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8280285596847534,
+      "learning_rate": 8.196887414436416e-06,
+      "loss": 0.9237,
+      "step": 9956
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.651206374168396,
+      "learning_rate": 8.195060280998772e-06,
+      "loss": 0.8837,
+      "step": 9957
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8081951141357422,
+      "learning_rate": 8.193233209844557e-06,
+      "loss": 0.9543,
+      "step": 9958
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7081942558288574,
+      "learning_rate": 8.191406201036816e-06,
+      "loss": 0.9403,
+      "step": 9959
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6036102771759033,
+      "learning_rate": 8.189579254638595e-06,
+      "loss": 0.9286,
+      "step": 9960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5856201648712158,
+      "learning_rate": 8.187752370712936e-06,
+      "loss": 0.8996,
+      "step": 9961
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8253306150436401,
+      "learning_rate": 8.185925549322883e-06,
+      "loss": 1.033,
+      "step": 9962
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6382275819778442,
+      "learning_rate": 8.18409879053147e-06,
+      "loss": 0.9808,
+      "step": 9963
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7461841106414795,
+      "learning_rate": 8.182272094401735e-06,
+      "loss": 0.9837,
+      "step": 9964
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.797199010848999,
+      "learning_rate": 8.180445460996711e-06,
+      "loss": 0.9297,
+      "step": 9965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.779309868812561,
+      "learning_rate": 8.178618890379432e-06,
+      "loss": 0.9404,
+      "step": 9966
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7046159505844116,
+      "learning_rate": 8.17679238261293e-06,
+      "loss": 0.8942,
+      "step": 9967
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6449289321899414,
+      "learning_rate": 8.174965937760228e-06,
+      "loss": 1.0307,
+      "step": 9968
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8228262662887573,
+      "learning_rate": 8.173139555884353e-06,
+      "loss": 0.9957,
+      "step": 9969
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7197256088256836,
+      "learning_rate": 8.171313237048331e-06,
+      "loss": 0.9318,
+      "step": 9970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8620012998580933,
+      "learning_rate": 8.16948698131518e-06,
+      "loss": 0.9437,
+      "step": 9971
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6902499198913574,
+      "learning_rate": 8.16766078874792e-06,
+      "loss": 0.8836,
+      "step": 9972
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9670344591140747,
+      "learning_rate": 8.165834659409566e-06,
+      "loss": 0.9847,
+      "step": 9973
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9216474294662476,
+      "learning_rate": 8.164008593363136e-06,
+      "loss": 0.9186,
+      "step": 9974
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.835475206375122,
+      "learning_rate": 8.16218259067164e-06,
+      "loss": 1.0679,
+      "step": 9975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7477036714553833,
+      "learning_rate": 8.16035665139809e-06,
+      "loss": 0.8871,
+      "step": 9976
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9078035354614258,
+      "learning_rate": 8.158530775605493e-06,
+      "loss": 0.9443,
+      "step": 9977
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0806448459625244,
+      "learning_rate": 8.156704963356851e-06,
+      "loss": 0.9899,
+      "step": 9978
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7123740911483765,
+      "learning_rate": 8.154879214715176e-06,
+      "loss": 0.9562,
+      "step": 9979
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0956957340240479,
+      "learning_rate": 8.153053529743465e-06,
+      "loss": 0.598,
+      "step": 9980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9688538312911987,
+      "learning_rate": 8.151227908504718e-06,
+      "loss": 0.544,
+      "step": 9981
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7711527347564697,
+      "learning_rate": 8.14940235106193e-06,
+      "loss": 0.9526,
+      "step": 9982
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7858059406280518,
+      "learning_rate": 8.147576857478098e-06,
+      "loss": 0.9197,
+      "step": 9983
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8245915174484253,
+      "learning_rate": 8.145751427816215e-06,
+      "loss": 0.9208,
+      "step": 9984
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8225047588348389,
+      "learning_rate": 8.143926062139268e-06,
+      "loss": 0.8478,
+      "step": 9985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.947296380996704,
+      "learning_rate": 8.142100760510249e-06,
+      "loss": 0.9458,
+      "step": 9986
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6904819011688232,
+      "learning_rate": 8.140275522992146e-06,
+      "loss": 0.9517,
+      "step": 9987
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6851528882980347,
+      "learning_rate": 8.138450349647936e-06,
+      "loss": 0.9918,
+      "step": 9988
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8243478536605835,
+      "learning_rate": 8.136625240540605e-06,
+      "loss": 1.0168,
+      "step": 9989
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.660402536392212,
+      "learning_rate": 8.13480019573313e-06,
+      "loss": 0.9552,
+      "step": 9990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6985617876052856,
+      "learning_rate": 8.132975215288494e-06,
+      "loss": 0.936,
+      "step": 9991
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9087896347045898,
+      "learning_rate": 8.131150299269665e-06,
+      "loss": 0.9941,
+      "step": 9992
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7848143577575684,
+      "learning_rate": 8.12932544773962e-06,
+      "loss": 1.0171,
+      "step": 9993
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4789185523986816,
+      "learning_rate": 8.12750066076133e-06,
+      "loss": 0.925,
+      "step": 9994
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.866515874862671,
+      "learning_rate": 8.125675938397759e-06,
+      "loss": 0.8677,
+      "step": 9995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5574911832809448,
+      "learning_rate": 8.123851280711877e-06,
+      "loss": 0.9505,
+      "step": 9996
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8167750835418701,
+      "learning_rate": 8.122026687766647e-06,
+      "loss": 0.9371,
+      "step": 9997
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9497883319854736,
+      "learning_rate": 8.120202159625029e-06,
+      "loss": 0.5246,
+      "step": 9998
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7061885595321655,
+      "learning_rate": 8.118377696349984e-06,
+      "loss": 0.9318,
+      "step": 9999
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7085875272750854,
+      "learning_rate": 8.116553298004467e-06,
+      "loss": 0.9913,
+      "step": 10000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0322234630584717,
+      "learning_rate": 8.114728964651438e-06,
+      "loss": 0.9057,
+      "step": 10001
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.708474040031433,
+      "learning_rate": 8.11290469635384e-06,
+      "loss": 0.9604,
+      "step": 10002
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6908295154571533,
+      "learning_rate": 8.111080493174635e-06,
+      "loss": 0.931,
+      "step": 10003
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.865363597869873,
+      "learning_rate": 8.109256355176761e-06,
+      "loss": 0.9613,
+      "step": 10004
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6232551336288452,
+      "learning_rate": 8.107432282423172e-06,
+      "loss": 0.8894,
+      "step": 10005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6600151062011719,
+      "learning_rate": 8.105608274976808e-06,
+      "loss": 1.0031,
+      "step": 10006
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7602753639221191,
+      "learning_rate": 8.10378433290061e-06,
+      "loss": 0.9707,
+      "step": 10007
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9686120748519897,
+      "learning_rate": 8.101960456257518e-06,
+      "loss": 0.9289,
+      "step": 10008
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7205287218093872,
+      "learning_rate": 8.10013664511047e-06,
+      "loss": 0.9626,
+      "step": 10009
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6464554071426392,
+      "learning_rate": 8.098312899522398e-06,
+      "loss": 0.8431,
+      "step": 10010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.621382236480713,
+      "learning_rate": 8.096489219556237e-06,
+      "loss": 0.9092,
+      "step": 10011
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6266937255859375,
+      "learning_rate": 8.094665605274914e-06,
+      "loss": 0.952,
+      "step": 10012
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.698685884475708,
+      "learning_rate": 8.09284205674136e-06,
+      "loss": 0.9421,
+      "step": 10013
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6341575384140015,
+      "learning_rate": 8.091018574018499e-06,
+      "loss": 0.8872,
+      "step": 10014
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7809375524520874,
+      "learning_rate": 8.089195157169254e-06,
+      "loss": 0.8652,
+      "step": 10015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8846242427825928,
+      "learning_rate": 8.087371806256548e-06,
+      "loss": 0.8921,
+      "step": 10016
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.857582688331604,
+      "learning_rate": 8.085548521343296e-06,
+      "loss": 0.936,
+      "step": 10017
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7509684562683105,
+      "learning_rate": 8.083725302492418e-06,
+      "loss": 0.9823,
+      "step": 10018
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.935383677482605,
+      "learning_rate": 8.081902149766825e-06,
+      "loss": 0.9181,
+      "step": 10019
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7712290287017822,
+      "learning_rate": 8.080079063229432e-06,
+      "loss": 0.9885,
+      "step": 10020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.823823094367981,
+      "learning_rate": 8.078256042943149e-06,
+      "loss": 0.9332,
+      "step": 10021
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6421974897384644,
+      "learning_rate": 8.07643308897088e-06,
+      "loss": 0.8633,
+      "step": 10022
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8431400060653687,
+      "learning_rate": 8.074610201375532e-06,
+      "loss": 0.9373,
+      "step": 10023
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8571799993515015,
+      "learning_rate": 8.07278738022001e-06,
+      "loss": 0.9569,
+      "step": 10024
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.780059576034546,
+      "learning_rate": 8.070964625567209e-06,
+      "loss": 0.9423,
+      "step": 10025
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.074397087097168,
+      "learning_rate": 8.069141937480031e-06,
+      "loss": 0.6162,
+      "step": 10026
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8878297805786133,
+      "learning_rate": 8.067319316021372e-06,
+      "loss": 0.9086,
+      "step": 10027
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7411837577819824,
+      "learning_rate": 8.065496761254126e-06,
+      "loss": 0.9747,
+      "step": 10028
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7861181497573853,
+      "learning_rate": 8.06367427324118e-06,
+      "loss": 0.9468,
+      "step": 10029
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9160903692245483,
+      "learning_rate": 8.061851852045428e-06,
+      "loss": 0.9071,
+      "step": 10030
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8891775608062744,
+      "learning_rate": 8.060029497729752e-06,
+      "loss": 0.937,
+      "step": 10031
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7994674444198608,
+      "learning_rate": 8.05820721035704e-06,
+      "loss": 0.9338,
+      "step": 10032
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7510249614715576,
+      "learning_rate": 8.056384989990173e-06,
+      "loss": 0.8641,
+      "step": 10033
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7696845531463623,
+      "learning_rate": 8.054562836692032e-06,
+      "loss": 0.9854,
+      "step": 10034
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7008767127990723,
+      "learning_rate": 8.052740750525492e-06,
+      "loss": 0.9997,
+      "step": 10035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.846603274345398,
+      "learning_rate": 8.05091873155343e-06,
+      "loss": 0.8249,
+      "step": 10036
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8892041444778442,
+      "learning_rate": 8.04909677983872e-06,
+      "loss": 0.9787,
+      "step": 10037
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.979797601699829,
+      "learning_rate": 8.047274895444227e-06,
+      "loss": 0.9783,
+      "step": 10038
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7236649990081787,
+      "learning_rate": 8.045453078432824e-06,
+      "loss": 0.9483,
+      "step": 10039
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.763467788696289,
+      "learning_rate": 8.043631328867376e-06,
+      "loss": 0.9365,
+      "step": 10040
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.706706166267395,
+      "learning_rate": 8.041809646810745e-06,
+      "loss": 0.9042,
+      "step": 10041
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8034722805023193,
+      "learning_rate": 8.039988032325794e-06,
+      "loss": 0.9613,
+      "step": 10042
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8166176080703735,
+      "learning_rate": 8.038166485475381e-06,
+      "loss": 0.9866,
+      "step": 10043
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7812343835830688,
+      "learning_rate": 8.036345006322358e-06,
+      "loss": 0.9458,
+      "step": 10044
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5142858028411865,
+      "learning_rate": 8.034523594929588e-06,
+      "loss": 0.8928,
+      "step": 10045
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9605401158332825,
+      "learning_rate": 8.032702251359918e-06,
+      "loss": 0.5909,
+      "step": 10046
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8225682973861694,
+      "learning_rate": 8.030880975676198e-06,
+      "loss": 0.8881,
+      "step": 10047
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.013690233230591,
+      "learning_rate": 8.029059767941275e-06,
+      "loss": 1.1004,
+      "step": 10048
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7007765769958496,
+      "learning_rate": 8.027238628217993e-06,
+      "loss": 0.9715,
+      "step": 10049
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8556448221206665,
+      "learning_rate": 8.025417556569196e-06,
+      "loss": 0.9892,
+      "step": 10050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7061582803726196,
+      "learning_rate": 8.023596553057723e-06,
+      "loss": 0.9882,
+      "step": 10051
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.878564715385437,
+      "learning_rate": 8.021775617746412e-06,
+      "loss": 1.0291,
+      "step": 10052
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.933752417564392,
+      "learning_rate": 8.0199547506981e-06,
+      "loss": 0.9178,
+      "step": 10053
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8793507814407349,
+      "learning_rate": 8.018133951975617e-06,
+      "loss": 0.9142,
+      "step": 10054
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7183701992034912,
+      "learning_rate": 8.016313221641795e-06,
+      "loss": 0.9182,
+      "step": 10055
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3512232303619385,
+      "learning_rate": 8.01449255975946e-06,
+      "loss": 1.0324,
+      "step": 10056
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.810836672782898,
+      "learning_rate": 8.012671966391444e-06,
+      "loss": 0.8897,
+      "step": 10057
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7067466974258423,
+      "learning_rate": 8.010851441600567e-06,
+      "loss": 0.8918,
+      "step": 10058
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.859833002090454,
+      "learning_rate": 8.00903098544965e-06,
+      "loss": 0.9134,
+      "step": 10059
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0736366510391235,
+      "learning_rate": 8.007210598001511e-06,
+      "loss": 0.5943,
+      "step": 10060
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6433991193771362,
+      "learning_rate": 8.00539027931897e-06,
+      "loss": 0.9065,
+      "step": 10061
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.650827169418335,
+      "learning_rate": 8.003570029464836e-06,
+      "loss": 0.9097,
+      "step": 10062
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7230761051177979,
+      "learning_rate": 8.001749848501925e-06,
+      "loss": 0.9745,
+      "step": 10063
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7422531843185425,
+      "learning_rate": 7.999929736493046e-06,
+      "loss": 1.0152,
+      "step": 10064
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9442970752716064,
+      "learning_rate": 7.998109693501002e-06,
+      "loss": 0.927,
+      "step": 10065
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6801656484603882,
+      "learning_rate": 7.996289719588604e-06,
+      "loss": 0.9136,
+      "step": 10066
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7054427862167358,
+      "learning_rate": 7.994469814818647e-06,
+      "loss": 0.994,
+      "step": 10067
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7884531021118164,
+      "learning_rate": 7.992649979253934e-06,
+      "loss": 1.0502,
+      "step": 10068
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7164208889007568,
+      "learning_rate": 7.990830212957266e-06,
+      "loss": 0.9148,
+      "step": 10069
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9825434684753418,
+      "learning_rate": 7.989010515991433e-06,
+      "loss": 0.9415,
+      "step": 10070
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7256280183792114,
+      "learning_rate": 7.987190888419229e-06,
+      "loss": 0.9046,
+      "step": 10071
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1246126890182495,
+      "learning_rate": 7.985371330303446e-06,
+      "loss": 0.5958,
+      "step": 10072
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8341796398162842,
+      "learning_rate": 7.983551841706869e-06,
+      "loss": 1.0017,
+      "step": 10073
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.876781940460205,
+      "learning_rate": 7.981732422692288e-06,
+      "loss": 0.9257,
+      "step": 10074
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.823730707168579,
+      "learning_rate": 7.979913073322482e-06,
+      "loss": 1.043,
+      "step": 10075
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6162852048873901,
+      "learning_rate": 7.978093793660234e-06,
+      "loss": 0.9329,
+      "step": 10076
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8555052280426025,
+      "learning_rate": 7.976274583768322e-06,
+      "loss": 0.9239,
+      "step": 10077
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7418303489685059,
+      "learning_rate": 7.974455443709519e-06,
+      "loss": 0.9221,
+      "step": 10078
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0510542392730713,
+      "learning_rate": 7.9726363735466e-06,
+      "loss": 0.5679,
+      "step": 10079
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6530357599258423,
+      "learning_rate": 7.97081737334234e-06,
+      "loss": 0.9626,
+      "step": 10080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9245362281799316,
+      "learning_rate": 7.968998443159502e-06,
+      "loss": 1.0087,
+      "step": 10081
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8035534620285034,
+      "learning_rate": 7.967179583060853e-06,
+      "loss": 1.0413,
+      "step": 10082
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6676276922225952,
+      "learning_rate": 7.96536079310916e-06,
+      "loss": 1.0009,
+      "step": 10083
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9292770624160767,
+      "learning_rate": 7.963542073367183e-06,
+      "loss": 0.8833,
+      "step": 10084
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8010612726211548,
+      "learning_rate": 7.961723423897676e-06,
+      "loss": 1.0337,
+      "step": 10085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8244218826293945,
+      "learning_rate": 7.959904844763405e-06,
+      "loss": 0.9809,
+      "step": 10086
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7090595960617065,
+      "learning_rate": 7.958086336027116e-06,
+      "loss": 1.0106,
+      "step": 10087
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.880307674407959,
+      "learning_rate": 7.956267897751566e-06,
+      "loss": 0.964,
+      "step": 10088
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.741143822669983,
+      "learning_rate": 7.954449529999501e-06,
+      "loss": 0.9734,
+      "step": 10089
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.65348219871521,
+      "learning_rate": 7.952631232833669e-06,
+      "loss": 0.8861,
+      "step": 10090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9793020486831665,
+      "learning_rate": 7.950813006316813e-06,
+      "loss": 1.0093,
+      "step": 10091
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5826938152313232,
+      "learning_rate": 7.948994850511678e-06,
+      "loss": 0.932,
+      "step": 10092
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7913683652877808,
+      "learning_rate": 7.947176765481e-06,
+      "loss": 0.8892,
+      "step": 10093
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6705474853515625,
+      "learning_rate": 7.945358751287518e-06,
+      "loss": 0.9189,
+      "step": 10094
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7742568254470825,
+      "learning_rate": 7.943540807993965e-06,
+      "loss": 0.9934,
+      "step": 10095
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6560449600219727,
+      "learning_rate": 7.941722935663076e-06,
+      "loss": 0.933,
+      "step": 10096
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8294296264648438,
+      "learning_rate": 7.939905134357574e-06,
+      "loss": 0.9435,
+      "step": 10097
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8178365230560303,
+      "learning_rate": 7.938087404140196e-06,
+      "loss": 0.9668,
+      "step": 10098
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7293378114700317,
+      "learning_rate": 7.936269745073661e-06,
+      "loss": 0.8747,
+      "step": 10099
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6465486288070679,
+      "learning_rate": 7.934452157220693e-06,
+      "loss": 0.8598,
+      "step": 10100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.574589729309082,
+      "learning_rate": 7.93263464064401e-06,
+      "loss": 1.0007,
+      "step": 10101
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5112907886505127,
+      "learning_rate": 7.930817195406332e-06,
+      "loss": 0.8726,
+      "step": 10102
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0750426054000854,
+      "learning_rate": 7.928999821570372e-06,
+      "loss": 0.576,
+      "step": 10103
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7453200817108154,
+      "learning_rate": 7.927182519198843e-06,
+      "loss": 0.9616,
+      "step": 10104
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7644621133804321,
+      "learning_rate": 7.925365288354453e-06,
+      "loss": 1.017,
+      "step": 10105
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6096513271331787,
+      "learning_rate": 7.923548129099914e-06,
+      "loss": 0.9792,
+      "step": 10106
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7948856353759766,
+      "learning_rate": 7.921731041497928e-06,
+      "loss": 1.0133,
+      "step": 10107
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5832029581069946,
+      "learning_rate": 7.9199140256112e-06,
+      "loss": 0.9992,
+      "step": 10108
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7737842798233032,
+      "learning_rate": 7.918097081502426e-06,
+      "loss": 0.9635,
+      "step": 10109
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.668912410736084,
+      "learning_rate": 7.916280209234307e-06,
+      "loss": 0.9632,
+      "step": 10110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6767700910568237,
+      "learning_rate": 7.914463408869537e-06,
+      "loss": 0.8997,
+      "step": 10111
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7865148782730103,
+      "learning_rate": 7.91264668047081e-06,
+      "loss": 0.9156,
+      "step": 10112
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7553844451904297,
+      "learning_rate": 7.910830024100816e-06,
+      "loss": 0.9804,
+      "step": 10113
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.698000192642212,
+      "learning_rate": 7.90901343982224e-06,
+      "loss": 1.0,
+      "step": 10114
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5750150680541992,
+      "learning_rate": 7.90719692769777e-06,
+      "loss": 0.9558,
+      "step": 10115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8725422620773315,
+      "learning_rate": 7.905380487790088e-06,
+      "loss": 0.9985,
+      "step": 10116
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.736026406288147,
+      "learning_rate": 7.903564120161876e-06,
+      "loss": 1.009,
+      "step": 10117
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7956511974334717,
+      "learning_rate": 7.901747824875807e-06,
+      "loss": 0.9785,
+      "step": 10118
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9132428169250488,
+      "learning_rate": 7.89993160199456e-06,
+      "loss": 1.0134,
+      "step": 10119
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8169267177581787,
+      "learning_rate": 7.898115451580809e-06,
+      "loss": 0.9388,
+      "step": 10120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7750738859176636,
+      "learning_rate": 7.896299373697221e-06,
+      "loss": 0.998,
+      "step": 10121
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8177005052566528,
+      "learning_rate": 7.894483368406464e-06,
+      "loss": 0.8983,
+      "step": 10122
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6939191818237305,
+      "learning_rate": 7.892667435771207e-06,
+      "loss": 0.9132,
+      "step": 10123
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.83375883102417,
+      "learning_rate": 7.890851575854108e-06,
+      "loss": 0.9686,
+      "step": 10124
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.798264503479004,
+      "learning_rate": 7.88903578871783e-06,
+      "loss": 0.9657,
+      "step": 10125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6605677604675293,
+      "learning_rate": 7.887220074425032e-06,
+      "loss": 0.9687,
+      "step": 10126
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6503233909606934,
+      "learning_rate": 7.885404433038366e-06,
+      "loss": 0.9301,
+      "step": 10127
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7085896730422974,
+      "learning_rate": 7.883588864620486e-06,
+      "loss": 0.9524,
+      "step": 10128
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6330724954605103,
+      "learning_rate": 7.881773369234043e-06,
+      "loss": 0.9499,
+      "step": 10129
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8343591690063477,
+      "learning_rate": 7.879957946941683e-06,
+      "loss": 1.0077,
+      "step": 10130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7827332019805908,
+      "learning_rate": 7.878142597806054e-06,
+      "loss": 0.927,
+      "step": 10131
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.814094066619873,
+      "learning_rate": 7.876327321889794e-06,
+      "loss": 0.9986,
+      "step": 10132
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7959904670715332,
+      "learning_rate": 7.87451211925555e-06,
+      "loss": 0.9828,
+      "step": 10133
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.809969425201416,
+      "learning_rate": 7.87269698996595e-06,
+      "loss": 0.8209,
+      "step": 10134
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.641908884048462,
+      "learning_rate": 7.870881934083637e-06,
+      "loss": 0.9187,
+      "step": 10135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.74166738986969,
+      "learning_rate": 7.869066951671241e-06,
+      "loss": 0.8763,
+      "step": 10136
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7387738227844238,
+      "learning_rate": 7.867252042791392e-06,
+      "loss": 0.8779,
+      "step": 10137
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7424275875091553,
+      "learning_rate": 7.865437207506716e-06,
+      "loss": 0.9328,
+      "step": 10138
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7437899112701416,
+      "learning_rate": 7.86362244587984e-06,
+      "loss": 0.8659,
+      "step": 10139
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9675076007843018,
+      "learning_rate": 7.861807757973386e-06,
+      "loss": 0.9703,
+      "step": 10140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6945221424102783,
+      "learning_rate": 7.859993143849976e-06,
+      "loss": 0.9291,
+      "step": 10141
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7178763151168823,
+      "learning_rate": 7.858178603572222e-06,
+      "loss": 0.943,
+      "step": 10142
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7484416961669922,
+      "learning_rate": 7.856364137202742e-06,
+      "loss": 0.9301,
+      "step": 10143
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7419393062591553,
+      "learning_rate": 7.85454974480415e-06,
+      "loss": 0.8959,
+      "step": 10144
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6358963251113892,
+      "learning_rate": 7.85273542643905e-06,
+      "loss": 0.9806,
+      "step": 10145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0152032375335693,
+      "learning_rate": 7.850921182170053e-06,
+      "loss": 0.5685,
+      "step": 10146
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6213083267211914,
+      "learning_rate": 7.849107012059765e-06,
+      "loss": 0.8766,
+      "step": 10147
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.02239727973938,
+      "learning_rate": 7.847292916170783e-06,
+      "loss": 0.9914,
+      "step": 10148
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9953396320343018,
+      "learning_rate": 7.84547889456571e-06,
+      "loss": 1.0276,
+      "step": 10149
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.780761957168579,
+      "learning_rate": 7.843664947307143e-06,
+      "loss": 0.9116,
+      "step": 10150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7339656352996826,
+      "learning_rate": 7.841851074457672e-06,
+      "loss": 0.8628,
+      "step": 10151
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8337950706481934,
+      "learning_rate": 7.840037276079895e-06,
+      "loss": 1.0312,
+      "step": 10152
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6772739887237549,
+      "learning_rate": 7.838223552236396e-06,
+      "loss": 0.9894,
+      "step": 10153
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0506101846694946,
+      "learning_rate": 7.836409902989766e-06,
+      "loss": 0.6064,
+      "step": 10154
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6640840768814087,
+      "learning_rate": 7.834596328402585e-06,
+      "loss": 0.9619,
+      "step": 10155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6776657104492188,
+      "learning_rate": 7.832782828537437e-06,
+      "loss": 0.8892,
+      "step": 10156
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7512624263763428,
+      "learning_rate": 7.830969403456899e-06,
+      "loss": 0.8953,
+      "step": 10157
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7777304649353027,
+      "learning_rate": 7.829156053223546e-06,
+      "loss": 1.0109,
+      "step": 10158
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8017364740371704,
+      "learning_rate": 7.827342777899956e-06,
+      "loss": 1.0039,
+      "step": 10159
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9505505561828613,
+      "learning_rate": 7.825529577548698e-06,
+      "loss": 0.9583,
+      "step": 10160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7678524255752563,
+      "learning_rate": 7.823716452232339e-06,
+      "loss": 0.863,
+      "step": 10161
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.674464225769043,
+      "learning_rate": 7.821903402013447e-06,
+      "loss": 0.9751,
+      "step": 10162
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6768087148666382,
+      "learning_rate": 7.820090426954583e-06,
+      "loss": 0.8927,
+      "step": 10163
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.72789466381073,
+      "learning_rate": 7.818277527118308e-06,
+      "loss": 0.9972,
+      "step": 10164
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6604609489440918,
+      "learning_rate": 7.816464702567182e-06,
+      "loss": 0.9636,
+      "step": 10165
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9407632350921631,
+      "learning_rate": 7.81465195336376e-06,
+      "loss": 0.5517,
+      "step": 10166
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.732412338256836,
+      "learning_rate": 7.812839279570596e-06,
+      "loss": 0.8963,
+      "step": 10167
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7932050228118896,
+      "learning_rate": 7.811026681250237e-06,
+      "loss": 0.9524,
+      "step": 10168
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8364158868789673,
+      "learning_rate": 7.809214158465234e-06,
+      "loss": 0.9195,
+      "step": 10169
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.795823574066162,
+      "learning_rate": 7.807401711278132e-06,
+      "loss": 0.9191,
+      "step": 10170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.636948585510254,
+      "learning_rate": 7.80558933975147e-06,
+      "loss": 1.0192,
+      "step": 10171
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7334684133529663,
+      "learning_rate": 7.80377704394779e-06,
+      "loss": 0.9352,
+      "step": 10172
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6473677158355713,
+      "learning_rate": 7.801964823929628e-06,
+      "loss": 0.984,
+      "step": 10173
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5900925397872925,
+      "learning_rate": 7.800152679759523e-06,
+      "loss": 1.0016,
+      "step": 10174
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7851715087890625,
+      "learning_rate": 7.798340611500002e-06,
+      "loss": 0.9217,
+      "step": 10175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6785316467285156,
+      "learning_rate": 7.796528619213594e-06,
+      "loss": 0.8743,
+      "step": 10176
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8954955339431763,
+      "learning_rate": 7.794716702962832e-06,
+      "loss": 0.9065,
+      "step": 10177
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8364949226379395,
+      "learning_rate": 7.792904862810236e-06,
+      "loss": 0.9438,
+      "step": 10178
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.845595359802246,
+      "learning_rate": 7.791093098818328e-06,
+      "loss": 0.8886,
+      "step": 10179
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7551746368408203,
+      "learning_rate": 7.789281411049626e-06,
+      "loss": 0.9246,
+      "step": 10180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.775098204612732,
+      "learning_rate": 7.787469799566647e-06,
+      "loss": 1.0187,
+      "step": 10181
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5932780504226685,
+      "learning_rate": 7.785658264431906e-06,
+      "loss": 0.9131,
+      "step": 10182
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.742674708366394,
+      "learning_rate": 7.783846805707911e-06,
+      "loss": 0.882,
+      "step": 10183
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6689084768295288,
+      "learning_rate": 7.782035423457173e-06,
+      "loss": 0.959,
+      "step": 10184
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8702350854873657,
+      "learning_rate": 7.780224117742197e-06,
+      "loss": 1.0103,
+      "step": 10185
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9909065961837769,
+      "learning_rate": 7.778412888625486e-06,
+      "loss": 1.028,
+      "step": 10186
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.704916000366211,
+      "learning_rate": 7.776601736169542e-06,
+      "loss": 0.9748,
+      "step": 10187
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9401311874389648,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.9973,
+      "step": 10188
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.815200686454773,
+      "learning_rate": 7.772979661489934e-06,
+      "loss": 0.9165,
+      "step": 10189
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7874748706817627,
+      "learning_rate": 7.77116873939126e-06,
+      "loss": 0.9812,
+      "step": 10190
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7537847757339478,
+      "learning_rate": 7.769357894203329e-06,
+      "loss": 0.9927,
+      "step": 10191
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6697522401809692,
+      "learning_rate": 7.767547125988624e-06,
+      "loss": 0.9763,
+      "step": 10192
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7794636487960815,
+      "learning_rate": 7.765736434809633e-06,
+      "loss": 0.9326,
+      "step": 10193
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.832753300666809,
+      "learning_rate": 7.763925820728838e-06,
+      "loss": 0.9674,
+      "step": 10194
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7453020811080933,
+      "learning_rate": 7.762115283808713e-06,
+      "loss": 0.9168,
+      "step": 10195
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8311960697174072,
+      "learning_rate": 7.760304824111741e-06,
+      "loss": 0.8921,
+      "step": 10196
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.859073519706726,
+      "learning_rate": 7.758494441700391e-06,
+      "loss": 0.9462,
+      "step": 10197
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7316911220550537,
+      "learning_rate": 7.756684136637139e-06,
+      "loss": 0.9028,
+      "step": 10198
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.588106632232666,
+      "learning_rate": 7.75487390898445e-06,
+      "loss": 0.973,
+      "step": 10199
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7877694368362427,
+      "learning_rate": 7.753063758804787e-06,
+      "loss": 0.9899,
+      "step": 10200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2713561058044434,
+      "learning_rate": 7.751253686160621e-06,
+      "loss": 0.907,
+      "step": 10201
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5533236265182495,
+      "learning_rate": 7.749443691114409e-06,
+      "loss": 0.9074,
+      "step": 10202
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7497491836547852,
+      "learning_rate": 7.747633773728606e-06,
+      "loss": 0.9786,
+      "step": 10203
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9755339622497559,
+      "learning_rate": 7.745823934065672e-06,
+      "loss": 0.9897,
+      "step": 10204
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7884821891784668,
+      "learning_rate": 7.744014172188055e-06,
+      "loss": 0.9738,
+      "step": 10205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7109590768814087,
+      "learning_rate": 7.742204488158207e-06,
+      "loss": 0.8953,
+      "step": 10206
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.713253378868103,
+      "learning_rate": 7.740394882038578e-06,
+      "loss": 0.9875,
+      "step": 10207
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8571285009384155,
+      "learning_rate": 7.738585353891609e-06,
+      "loss": 0.9634,
+      "step": 10208
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8915430307388306,
+      "learning_rate": 7.736775903779744e-06,
+      "loss": 0.8547,
+      "step": 10209
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8626590967178345,
+      "learning_rate": 7.734966531765417e-06,
+      "loss": 0.9033,
+      "step": 10210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.676995873451233,
+      "learning_rate": 7.73315723791107e-06,
+      "loss": 0.9495,
+      "step": 10211
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7670599222183228,
+      "learning_rate": 7.731348022279135e-06,
+      "loss": 0.9953,
+      "step": 10212
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7109105587005615,
+      "learning_rate": 7.72953888493204e-06,
+      "loss": 0.8834,
+      "step": 10213
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0694222450256348,
+      "learning_rate": 7.727729825932218e-06,
+      "loss": 0.6004,
+      "step": 10214
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8354840278625488,
+      "learning_rate": 7.725920845342091e-06,
+      "loss": 0.9872,
+      "step": 10215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.745019793510437,
+      "learning_rate": 7.724111943224085e-06,
+      "loss": 0.8719,
+      "step": 10216
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8354718685150146,
+      "learning_rate": 7.722303119640616e-06,
+      "loss": 0.8687,
+      "step": 10217
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8223716020584106,
+      "learning_rate": 7.720494374654104e-06,
+      "loss": 0.9698,
+      "step": 10218
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8163396120071411,
+      "learning_rate": 7.718685708326965e-06,
+      "loss": 0.9147,
+      "step": 10219
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.718186616897583,
+      "learning_rate": 7.716877120721612e-06,
+      "loss": 1.0772,
+      "step": 10220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.687212347984314,
+      "learning_rate": 7.71506861190045e-06,
+      "loss": 0.9258,
+      "step": 10221
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6812412738800049,
+      "learning_rate": 7.713260181925886e-06,
+      "loss": 0.9733,
+      "step": 10222
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6774266958236694,
+      "learning_rate": 7.711451830860325e-06,
+      "loss": 0.9446,
+      "step": 10223
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5561604499816895,
+      "learning_rate": 7.70964355876617e-06,
+      "loss": 0.8825,
+      "step": 10224
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8358871936798096,
+      "learning_rate": 7.70783536570582e-06,
+      "loss": 0.9304,
+      "step": 10225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.658037781715393,
+      "learning_rate": 7.706027251741666e-06,
+      "loss": 0.9222,
+      "step": 10226
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6748814582824707,
+      "learning_rate": 7.704219216936104e-06,
+      "loss": 0.9108,
+      "step": 10227
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3357272148132324,
+      "learning_rate": 7.702411261351524e-06,
+      "loss": 0.8942,
+      "step": 10228
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8117284774780273,
+      "learning_rate": 7.700603385050312e-06,
+      "loss": 0.9566,
+      "step": 10229
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9078834056854248,
+      "learning_rate": 7.698795588094855e-06,
+      "loss": 0.9866,
+      "step": 10230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9172916412353516,
+      "learning_rate": 7.696987870547533e-06,
+      "loss": 1.0027,
+      "step": 10231
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7475751638412476,
+      "learning_rate": 7.695180232470727e-06,
+      "loss": 0.9186,
+      "step": 10232
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.149897575378418,
+      "learning_rate": 7.693372673926814e-06,
+      "loss": 0.9182,
+      "step": 10233
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6616199016571045,
+      "learning_rate": 7.691565194978167e-06,
+      "loss": 1.0003,
+      "step": 10234
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7549474239349365,
+      "learning_rate": 7.689757795687156e-06,
+      "loss": 0.9944,
+      "step": 10235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0900800228118896,
+      "learning_rate": 7.68795047611615e-06,
+      "loss": 0.6007,
+      "step": 10236
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0690689086914062,
+      "learning_rate": 7.686143236327515e-06,
+      "loss": 0.5851,
+      "step": 10237
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.01665461063385,
+      "learning_rate": 7.684336076383614e-06,
+      "loss": 0.5323,
+      "step": 10238
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.686279535293579,
+      "learning_rate": 7.682528996346805e-06,
+      "loss": 0.9305,
+      "step": 10239
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7071090936660767,
+      "learning_rate": 7.680721996279448e-06,
+      "loss": 0.9139,
+      "step": 10240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8833144903182983,
+      "learning_rate": 7.678915076243895e-06,
+      "loss": 0.9461,
+      "step": 10241
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.81498384475708,
+      "learning_rate": 7.677108236302499e-06,
+      "loss": 0.9373,
+      "step": 10242
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7244994640350342,
+      "learning_rate": 7.675301476517609e-06,
+      "loss": 0.8794,
+      "step": 10243
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7051539421081543,
+      "learning_rate": 7.673494796951573e-06,
+      "loss": 0.9557,
+      "step": 10244
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7417227029800415,
+      "learning_rate": 7.671688197666731e-06,
+      "loss": 1.0163,
+      "step": 10245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7964205741882324,
+      "learning_rate": 7.669881678725426e-06,
+      "loss": 0.9567,
+      "step": 10246
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6077373027801514,
+      "learning_rate": 7.668075240189996e-06,
+      "loss": 0.8828,
+      "step": 10247
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.931624174118042,
+      "learning_rate": 7.666268882122775e-06,
+      "loss": 0.9854,
+      "step": 10248
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9547218084335327,
+      "learning_rate": 7.664462604586095e-06,
+      "loss": 1.0109,
+      "step": 10249
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7365643978118896,
+      "learning_rate": 7.662656407642288e-06,
+      "loss": 0.9014,
+      "step": 10250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.807495355606079,
+      "learning_rate": 7.660850291353679e-06,
+      "loss": 0.9813,
+      "step": 10251
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.787226915359497,
+      "learning_rate": 7.659044255782592e-06,
+      "loss": 0.9302,
+      "step": 10252
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7445251941680908,
+      "learning_rate": 7.65723830099135e-06,
+      "loss": 0.9756,
+      "step": 10253
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7434368133544922,
+      "learning_rate": 7.655432427042266e-06,
+      "loss": 0.9532,
+      "step": 10254
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8607277870178223,
+      "learning_rate": 7.653626633997661e-06,
+      "loss": 0.924,
+      "step": 10255
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.94423246383667,
+      "learning_rate": 7.651820921919848e-06,
+      "loss": 0.9836,
+      "step": 10256
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7074990272521973,
+      "learning_rate": 7.650015290871135e-06,
+      "loss": 0.9772,
+      "step": 10257
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7803990840911865,
+      "learning_rate": 7.648209740913831e-06,
+      "loss": 0.986,
+      "step": 10258
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8392260074615479,
+      "learning_rate": 7.646404272110238e-06,
+      "loss": 0.9683,
+      "step": 10259
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.81463623046875,
+      "learning_rate": 7.644598884522659e-06,
+      "loss": 0.9455,
+      "step": 10260
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7497608661651611,
+      "learning_rate": 7.642793578213394e-06,
+      "loss": 0.9348,
+      "step": 10261
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.675525188446045,
+      "learning_rate": 7.640988353244739e-06,
+      "loss": 0.9311,
+      "step": 10262
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8487236499786377,
+      "learning_rate": 7.639183209678984e-06,
+      "loss": 0.97,
+      "step": 10263
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.765504240989685,
+      "learning_rate": 7.637378147578422e-06,
+      "loss": 0.942,
+      "step": 10264
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.772786259651184,
+      "learning_rate": 7.63557316700534e-06,
+      "loss": 0.9335,
+      "step": 10265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7995020151138306,
+      "learning_rate": 7.633768268022023e-06,
+      "loss": 0.9533,
+      "step": 10266
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7972110509872437,
+      "learning_rate": 7.631963450690755e-06,
+      "loss": 0.9128,
+      "step": 10267
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8026013374328613,
+      "learning_rate": 7.630158715073813e-06,
+      "loss": 0.9882,
+      "step": 10268
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8926401138305664,
+      "learning_rate": 7.628354061233472e-06,
+      "loss": 0.8931,
+      "step": 10269
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.671213150024414,
+      "learning_rate": 7.626549489232009e-06,
+      "loss": 0.9555,
+      "step": 10270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1038405895233154,
+      "learning_rate": 7.624744999131691e-06,
+      "loss": 0.6132,
+      "step": 10271
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7962102890014648,
+      "learning_rate": 7.6229405909947915e-06,
+      "loss": 0.9754,
+      "step": 10272
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.042463779449463,
+      "learning_rate": 7.621136264883571e-06,
+      "loss": 0.5654,
+      "step": 10273
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.761847972869873,
+      "learning_rate": 7.619332020860293e-06,
+      "loss": 1.0196,
+      "step": 10274
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.72986900806427,
+      "learning_rate": 7.617527858987217e-06,
+      "loss": 1.0209,
+      "step": 10275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6604474782943726,
+      "learning_rate": 7.6157237793265996e-06,
+      "loss": 0.9374,
+      "step": 10276
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5567196607589722,
+      "learning_rate": 7.613919781940694e-06,
+      "loss": 1.004,
+      "step": 10277
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.776759147644043,
+      "learning_rate": 7.612115866891751e-06,
+      "loss": 0.9036,
+      "step": 10278
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7751779556274414,
+      "learning_rate": 7.61031203424202e-06,
+      "loss": 0.9085,
+      "step": 10279
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7614938020706177,
+      "learning_rate": 7.608508284053746e-06,
+      "loss": 0.9481,
+      "step": 10280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.625900387763977,
+      "learning_rate": 7.606704616389169e-06,
+      "loss": 0.8383,
+      "step": 10281
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.795332431793213,
+      "learning_rate": 7.604901031310532e-06,
+      "loss": 0.9474,
+      "step": 10282
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7623851299285889,
+      "learning_rate": 7.603097528880067e-06,
+      "loss": 0.9397,
+      "step": 10283
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5453157424926758,
+      "learning_rate": 7.601294109160012e-06,
+      "loss": 0.9392,
+      "step": 10284
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.791609287261963,
+      "learning_rate": 7.599490772212599e-06,
+      "loss": 0.9385,
+      "step": 10285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7089269161224365,
+      "learning_rate": 7.597687518100052e-06,
+      "loss": 0.9342,
+      "step": 10286
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1518092155456543,
+      "learning_rate": 7.595884346884599e-06,
+      "loss": 0.9135,
+      "step": 10287
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.794111728668213,
+      "learning_rate": 7.594081258628461e-06,
+      "loss": 0.918,
+      "step": 10288
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7815576791763306,
+      "learning_rate": 7.592278253393859e-06,
+      "loss": 0.9136,
+      "step": 10289
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6652026176452637,
+      "learning_rate": 7.590475331243008e-06,
+      "loss": 0.9964,
+      "step": 10290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9116684794425964,
+      "learning_rate": 7.588672492238123e-06,
+      "loss": 0.526,
+      "step": 10291
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0427523851394653,
+      "learning_rate": 7.586869736441413e-06,
+      "loss": 0.5481,
+      "step": 10292
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7694807052612305,
+      "learning_rate": 7.5850670639150904e-06,
+      "loss": 0.8908,
+      "step": 10293
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.902209997177124,
+      "learning_rate": 7.583264474721356e-06,
+      "loss": 0.9052,
+      "step": 10294
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9816991090774536,
+      "learning_rate": 7.581461968922413e-06,
+      "loss": 0.9873,
+      "step": 10295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9553471803665161,
+      "learning_rate": 7.5796595465804616e-06,
+      "loss": 0.9211,
+      "step": 10296
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6685162782669067,
+      "learning_rate": 7.577857207757698e-06,
+      "loss": 0.924,
+      "step": 10297
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6475584506988525,
+      "learning_rate": 7.576054952516318e-06,
+      "loss": 0.8465,
+      "step": 10298
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8763052225112915,
+      "learning_rate": 7.57425278091851e-06,
+      "loss": 0.9843,
+      "step": 10299
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7583242654800415,
+      "learning_rate": 7.572450693026462e-06,
+      "loss": 0.9446,
+      "step": 10300
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6700351238250732,
+      "learning_rate": 7.57064868890236e-06,
+      "loss": 0.9628,
+      "step": 10301
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8223263025283813,
+      "learning_rate": 7.5688467686083845e-06,
+      "loss": 0.9452,
+      "step": 10302
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9018670320510864,
+      "learning_rate": 7.567044932206717e-06,
+      "loss": 0.9898,
+      "step": 10303
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9095746278762817,
+      "learning_rate": 7.565243179759533e-06,
+      "loss": 0.9818,
+      "step": 10304
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5821911096572876,
+      "learning_rate": 7.563441511329005e-06,
+      "loss": 0.843,
+      "step": 10305
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8519865274429321,
+      "learning_rate": 7.561639926977304e-06,
+      "loss": 1.0019,
+      "step": 10306
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8512591123580933,
+      "learning_rate": 7.559838426766598e-06,
+      "loss": 0.9718,
+      "step": 10307
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7307250499725342,
+      "learning_rate": 7.55803701075905e-06,
+      "loss": 0.9418,
+      "step": 10308
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7194267511367798,
+      "learning_rate": 7.5562356790168256e-06,
+      "loss": 0.9845,
+      "step": 10309
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.724992036819458,
+      "learning_rate": 7.5544344316020804e-06,
+      "loss": 0.8839,
+      "step": 10310
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.851946234703064,
+      "learning_rate": 7.552633268576972e-06,
+      "loss": 0.969,
+      "step": 10311
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7737661600112915,
+      "learning_rate": 7.550832190003654e-06,
+      "loss": 0.9666,
+      "step": 10312
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7596232891082764,
+      "learning_rate": 7.549031195944274e-06,
+      "loss": 0.9188,
+      "step": 10313
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.709008812904358,
+      "learning_rate": 7.547230286460983e-06,
+      "loss": 0.8701,
+      "step": 10314
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0186187028884888,
+      "learning_rate": 7.5454294616159215e-06,
+      "loss": 0.5712,
+      "step": 10315
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8753888607025146,
+      "learning_rate": 7.543628721471234e-06,
+      "loss": 0.9725,
+      "step": 10316
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.641258716583252,
+      "learning_rate": 7.5418280660890565e-06,
+      "loss": 0.9877,
+      "step": 10317
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8166675567626953,
+      "learning_rate": 7.540027495531527e-06,
+      "loss": 0.9519,
+      "step": 10318
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8743252754211426,
+      "learning_rate": 7.538227009860775e-06,
+      "loss": 1.0089,
+      "step": 10319
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.784569501876831,
+      "learning_rate": 7.536426609138933e-06,
+      "loss": 0.8955,
+      "step": 10320
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7723467350006104,
+      "learning_rate": 7.534626293428127e-06,
+      "loss": 1.041,
+      "step": 10321
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9346673488616943,
+      "learning_rate": 7.532826062790482e-06,
+      "loss": 0.9564,
+      "step": 10322
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6984739303588867,
+      "learning_rate": 7.531025917288116e-06,
+      "loss": 0.9992,
+      "step": 10323
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8053065538406372,
+      "learning_rate": 7.529225856983151e-06,
+      "loss": 0.9708,
+      "step": 10324
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8159064054489136,
+      "learning_rate": 7.527425881937699e-06,
+      "loss": 0.9246,
+      "step": 10325
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5712523460388184,
+      "learning_rate": 7.525625992213872e-06,
+      "loss": 0.9322,
+      "step": 10326
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7395873069763184,
+      "learning_rate": 7.5238261878737815e-06,
+      "loss": 1.0413,
+      "step": 10327
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6705348491668701,
+      "learning_rate": 7.522026468979532e-06,
+      "loss": 0.9279,
+      "step": 10328
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8817106485366821,
+      "learning_rate": 7.520226835593226e-06,
+      "loss": 1.0056,
+      "step": 10329
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8142273426055908,
+      "learning_rate": 7.518427287776966e-06,
+      "loss": 0.9287,
+      "step": 10330
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9435560703277588,
+      "learning_rate": 7.516627825592848e-06,
+      "loss": 0.8942,
+      "step": 10331
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6227105855941772,
+      "learning_rate": 7.514828449102965e-06,
+      "loss": 0.9093,
+      "step": 10332
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5616214275360107,
+      "learning_rate": 7.513029158369412e-06,
+      "loss": 0.9424,
+      "step": 10333
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.938692569732666,
+      "learning_rate": 7.511229953454276e-06,
+      "loss": 0.9645,
+      "step": 10334
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7903739213943481,
+      "learning_rate": 7.50943083441964e-06,
+      "loss": 0.909,
+      "step": 10335
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7871263027191162,
+      "learning_rate": 7.50763180132759e-06,
+      "loss": 0.8656,
+      "step": 10336
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8659932613372803,
+      "learning_rate": 7.5058328542402035e-06,
+      "loss": 0.9399,
+      "step": 10337
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9322506189346313,
+      "learning_rate": 7.504033993219559e-06,
+      "loss": 0.9597,
+      "step": 10338
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6774568557739258,
+      "learning_rate": 7.50223521832773e-06,
+      "loss": 1.0672,
+      "step": 10339
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6441165208816528,
+      "learning_rate": 7.500436529626787e-06,
+      "loss": 0.9803,
+      "step": 10340
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.534072995185852,
+      "learning_rate": 7.498637927178796e-06,
+      "loss": 0.8899,
+      "step": 10341
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7665263414382935,
+      "learning_rate": 7.496839411045824e-06,
+      "loss": 0.9251,
+      "step": 10342
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7371271848678589,
+      "learning_rate": 7.495040981289931e-06,
+      "loss": 0.9546,
+      "step": 10343
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6054104566574097,
+      "learning_rate": 7.493242637973175e-06,
+      "loss": 0.9281,
+      "step": 10344
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.692980408668518,
+      "learning_rate": 7.491444381157616e-06,
+      "loss": 0.9072,
+      "step": 10345
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8429760932922363,
+      "learning_rate": 7.489646210905301e-06,
+      "loss": 0.9423,
+      "step": 10346
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8078008890151978,
+      "learning_rate": 7.487848127278285e-06,
+      "loss": 0.9181,
+      "step": 10347
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8376256227493286,
+      "learning_rate": 7.486050130338611e-06,
+      "loss": 0.9126,
+      "step": 10348
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.14129376411438,
+      "learning_rate": 7.484252220148327e-06,
+      "loss": 0.9624,
+      "step": 10349
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7582612037658691,
+      "learning_rate": 7.482454396769468e-06,
+      "loss": 1.0191,
+      "step": 10350
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6357886791229248,
+      "learning_rate": 7.480656660264076e-06,
+      "loss": 0.9612,
+      "step": 10351
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7550610303878784,
+      "learning_rate": 7.478859010694187e-06,
+      "loss": 0.9674,
+      "step": 10352
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.855189323425293,
+      "learning_rate": 7.477061448121832e-06,
+      "loss": 0.9267,
+      "step": 10353
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7542481422424316,
+      "learning_rate": 7.4752639726090374e-06,
+      "loss": 0.953,
+      "step": 10354
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6325795650482178,
+      "learning_rate": 7.47346658421783e-06,
+      "loss": 0.9645,
+      "step": 10355
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7339776754379272,
+      "learning_rate": 7.4716692830102335e-06,
+      "loss": 0.8925,
+      "step": 10356
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6789088249206543,
+      "learning_rate": 7.469872069048267e-06,
+      "loss": 0.967,
+      "step": 10357
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7512871026992798,
+      "learning_rate": 7.468074942393949e-06,
+      "loss": 0.9476,
+      "step": 10358
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.722955584526062,
+      "learning_rate": 7.466277903109291e-06,
+      "loss": 0.9465,
+      "step": 10359
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7761191129684448,
+      "learning_rate": 7.464480951256306e-06,
+      "loss": 0.9638,
+      "step": 10360
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7436084747314453,
+      "learning_rate": 7.462684086897001e-06,
+      "loss": 0.9487,
+      "step": 10361
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6918977499008179,
+      "learning_rate": 7.460887310093377e-06,
+      "loss": 0.9617,
+      "step": 10362
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.652294874191284,
+      "learning_rate": 7.459090620907441e-06,
+      "loss": 0.9451,
+      "step": 10363
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7144298553466797,
+      "learning_rate": 7.457294019401191e-06,
+      "loss": 0.9358,
+      "step": 10364
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8511468172073364,
+      "learning_rate": 7.455497505636622e-06,
+      "loss": 1.0208,
+      "step": 10365
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8542863130569458,
+      "learning_rate": 7.4537010796757244e-06,
+      "loss": 0.9762,
+      "step": 10366
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6627072095870972,
+      "learning_rate": 7.451904741580491e-06,
+      "loss": 0.9919,
+      "step": 10367
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6468545198440552,
+      "learning_rate": 7.450108491412909e-06,
+      "loss": 0.9141,
+      "step": 10368
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.729580044746399,
+      "learning_rate": 7.448312329234957e-06,
+      "loss": 0.9566,
+      "step": 10369
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9567667245864868,
+      "learning_rate": 7.44651625510862e-06,
+      "loss": 0.9225,
+      "step": 10370
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0888370275497437,
+      "learning_rate": 7.444720269095875e-06,
+      "loss": 0.5679,
+      "step": 10371
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6800328493118286,
+      "learning_rate": 7.442924371258694e-06,
+      "loss": 0.8681,
+      "step": 10372
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8042755126953125,
+      "learning_rate": 7.4411285616590505e-06,
+      "loss": 0.9504,
+      "step": 10373
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6325427293777466,
+      "learning_rate": 7.4393328403589105e-06,
+      "loss": 0.8894,
+      "step": 10374
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6852833032608032,
+      "learning_rate": 7.437537207420243e-06,
+      "loss": 0.9377,
+      "step": 10375
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6984894275665283,
+      "learning_rate": 7.435741662905009e-06,
+      "loss": 0.9038,
+      "step": 10376
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5487910509109497,
+      "learning_rate": 7.433946206875167e-06,
+      "loss": 0.9157,
+      "step": 10377
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7332934141159058,
+      "learning_rate": 7.432150839392674e-06,
+      "loss": 0.9208,
+      "step": 10378
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8254334926605225,
+      "learning_rate": 7.4303555605194825e-06,
+      "loss": 0.9517,
+      "step": 10379
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7641140222549438,
+      "learning_rate": 7.428560370317542e-06,
+      "loss": 0.9257,
+      "step": 10380
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8517814874649048,
+      "learning_rate": 7.426765268848801e-06,
+      "loss": 0.9889,
+      "step": 10381
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.831356406211853,
+      "learning_rate": 7.424970256175201e-06,
+      "loss": 0.9247,
+      "step": 10382
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7396743297576904,
+      "learning_rate": 7.423175332358686e-06,
+      "loss": 1.0379,
+      "step": 10383
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6547646522521973,
+      "learning_rate": 7.421380497461191e-06,
+      "loss": 0.9259,
+      "step": 10384
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8959203958511353,
+      "learning_rate": 7.419585751544654e-06,
+      "loss": 0.9567,
+      "step": 10385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6283398866653442,
+      "learning_rate": 7.417791094671e-06,
+      "loss": 0.984,
+      "step": 10386
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9545018672943115,
+      "learning_rate": 7.415996526902165e-06,
+      "loss": 0.9362,
+      "step": 10387
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.826254963874817,
+      "learning_rate": 7.414202048300072e-06,
+      "loss": 0.9179,
+      "step": 10388
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9001611471176147,
+      "learning_rate": 7.412407658926644e-06,
+      "loss": 0.9545,
+      "step": 10389
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6383544206619263,
+      "learning_rate": 7.4106133588437975e-06,
+      "loss": 0.9945,
+      "step": 10390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.791081428527832,
+      "learning_rate": 7.408819148113453e-06,
+      "loss": 0.9062,
+      "step": 10391
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7783887386322021,
+      "learning_rate": 7.407025026797521e-06,
+      "loss": 0.9097,
+      "step": 10392
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.778610348701477,
+      "learning_rate": 7.405230994957911e-06,
+      "loss": 0.9445,
+      "step": 10393
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7498493194580078,
+      "learning_rate": 7.403437052656531e-06,
+      "loss": 0.9177,
+      "step": 10394
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6312103271484375,
+      "learning_rate": 7.401643199955286e-06,
+      "loss": 0.8852,
+      "step": 10395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6708095073699951,
+      "learning_rate": 7.399849436916076e-06,
+      "loss": 0.8823,
+      "step": 10396
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8595303297042847,
+      "learning_rate": 7.3980557636008e-06,
+      "loss": 1.0069,
+      "step": 10397
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.739510178565979,
+      "learning_rate": 7.3962621800713475e-06,
+      "loss": 0.9061,
+      "step": 10398
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7827552556991577,
+      "learning_rate": 7.394468686389615e-06,
+      "loss": 0.9607,
+      "step": 10399
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6982142925262451,
+      "learning_rate": 7.3926752826174916e-06,
+      "loss": 0.9281,
+      "step": 10400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8110274076461792,
+      "learning_rate": 7.390881968816859e-06,
+      "loss": 0.8984,
+      "step": 10401
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.720268726348877,
+      "learning_rate": 7.389088745049604e-06,
+      "loss": 1.0069,
+      "step": 10402
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7948675155639648,
+      "learning_rate": 7.387295611377599e-06,
+      "loss": 1.0496,
+      "step": 10403
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.751927375793457,
+      "learning_rate": 7.385502567862728e-06,
+      "loss": 0.9373,
+      "step": 10404
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6295148134231567,
+      "learning_rate": 7.383709614566859e-06,
+      "loss": 0.9132,
+      "step": 10405
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6818876266479492,
+      "learning_rate": 7.381916751551863e-06,
+      "loss": 1.029,
+      "step": 10406
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.627815842628479,
+      "learning_rate": 7.3801239788796075e-06,
+      "loss": 0.9505,
+      "step": 10407
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1063754558563232,
+      "learning_rate": 7.3783312966119535e-06,
+      "loss": 0.5933,
+      "step": 10408
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7942789793014526,
+      "learning_rate": 7.376538704810765e-06,
+      "loss": 1.0138,
+      "step": 10409
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8698800802230835,
+      "learning_rate": 7.374746203537897e-06,
+      "loss": 0.9455,
+      "step": 10410
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7118797302246094,
+      "learning_rate": 7.372953792855203e-06,
+      "loss": 0.9469,
+      "step": 10411
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.63349449634552,
+      "learning_rate": 7.3711614728245364e-06,
+      "loss": 0.9426,
+      "step": 10412
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7269929647445679,
+      "learning_rate": 7.3693692435077425e-06,
+      "loss": 0.9628,
+      "step": 10413
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8332700729370117,
+      "learning_rate": 7.36757710496667e-06,
+      "loss": 0.8703,
+      "step": 10414
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7432461977005005,
+      "learning_rate": 7.365785057263156e-06,
+      "loss": 0.9429,
+      "step": 10415
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6571274995803833,
+      "learning_rate": 7.36399310045904e-06,
+      "loss": 0.9949,
+      "step": 10416
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7190614938735962,
+      "learning_rate": 7.362201234616162e-06,
+      "loss": 1.0038,
+      "step": 10417
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7729727029800415,
+      "learning_rate": 7.3604094597963494e-06,
+      "loss": 0.9989,
+      "step": 10418
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.938710331916809,
+      "learning_rate": 7.358617776061434e-06,
+      "loss": 0.8579,
+      "step": 10419
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6284576654434204,
+      "learning_rate": 7.35682618347324e-06,
+      "loss": 0.9837,
+      "step": 10420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7897228002548218,
+      "learning_rate": 7.355034682093591e-06,
+      "loss": 0.9078,
+      "step": 10421
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.685447096824646,
+      "learning_rate": 7.3532432719843075e-06,
+      "loss": 0.9719,
+      "step": 10422
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5817539691925049,
+      "learning_rate": 7.351451953207205e-06,
+      "loss": 0.8721,
+      "step": 10423
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7040961980819702,
+      "learning_rate": 7.349660725824097e-06,
+      "loss": 0.9954,
+      "step": 10424
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8188546895980835,
+      "learning_rate": 7.347869589896794e-06,
+      "loss": 0.9368,
+      "step": 10425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.659582257270813,
+      "learning_rate": 7.346078545487102e-06,
+      "loss": 0.947,
+      "step": 10426
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7515660524368286,
+      "learning_rate": 7.344287592656827e-06,
+      "loss": 0.9685,
+      "step": 10427
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.727772831916809,
+      "learning_rate": 7.342496731467766e-06,
+      "loss": 0.9036,
+      "step": 10428
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9014296531677246,
+      "learning_rate": 7.340705961981722e-06,
+      "loss": 1.0327,
+      "step": 10429
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9315305948257446,
+      "learning_rate": 7.338915284260487e-06,
+      "loss": 0.8965,
+      "step": 10430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7743418216705322,
+      "learning_rate": 7.337124698365851e-06,
+      "loss": 0.9831,
+      "step": 10431
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7623370885849,
+      "learning_rate": 7.335334204359605e-06,
+      "loss": 1.061,
+      "step": 10432
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.877609133720398,
+      "learning_rate": 7.333543802303531e-06,
+      "loss": 1.0535,
+      "step": 10433
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8454853296279907,
+      "learning_rate": 7.331753492259412e-06,
+      "loss": 0.9605,
+      "step": 10434
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6689057350158691,
+      "learning_rate": 7.329963274289027e-06,
+      "loss": 0.9097,
+      "step": 10435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8124525547027588,
+      "learning_rate": 7.328173148454151e-06,
+      "loss": 0.9679,
+      "step": 10436
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8154628276824951,
+      "learning_rate": 7.326383114816555e-06,
+      "loss": 1.0283,
+      "step": 10437
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8383556604385376,
+      "learning_rate": 7.324593173438011e-06,
+      "loss": 0.8785,
+      "step": 10438
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8645848035812378,
+      "learning_rate": 7.322803324380282e-06,
+      "loss": 0.8606,
+      "step": 10439
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.590856671333313,
+      "learning_rate": 7.321013567705131e-06,
+      "loss": 0.9486,
+      "step": 10440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7762575149536133,
+      "learning_rate": 7.319223903474318e-06,
+      "loss": 0.9583,
+      "step": 10441
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6695590019226074,
+      "learning_rate": 7.317434331749602e-06,
+      "loss": 0.9919,
+      "step": 10442
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7676153182983398,
+      "learning_rate": 7.315644852592733e-06,
+      "loss": 0.9094,
+      "step": 10443
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.873572826385498,
+      "learning_rate": 7.31385546606546e-06,
+      "loss": 1.0042,
+      "step": 10444
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0585534572601318,
+      "learning_rate": 7.312066172229534e-06,
+      "loss": 0.65,
+      "step": 10445
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7065507173538208,
+      "learning_rate": 7.310276971146695e-06,
+      "loss": 0.9397,
+      "step": 10446
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8405537605285645,
+      "learning_rate": 7.308487862878684e-06,
+      "loss": 0.934,
+      "step": 10447
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7185841798782349,
+      "learning_rate": 7.306698847487239e-06,
+      "loss": 0.8914,
+      "step": 10448
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8095675706863403,
+      "learning_rate": 7.304909925034093e-06,
+      "loss": 0.9432,
+      "step": 10449
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1023943424224854,
+      "learning_rate": 7.303121095580976e-06,
+      "loss": 0.6007,
+      "step": 10450
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.740768313407898,
+      "learning_rate": 7.301332359189618e-06,
+      "loss": 0.9509,
+      "step": 10451
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.058027505874634,
+      "learning_rate": 7.29954371592174e-06,
+      "loss": 0.9772,
+      "step": 10452
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7246119976043701,
+      "learning_rate": 7.297755165839066e-06,
+      "loss": 0.9351,
+      "step": 10453
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.024247169494629,
+      "learning_rate": 7.295966709003312e-06,
+      "loss": 0.9063,
+      "step": 10454
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9063416719436646,
+      "learning_rate": 7.294178345476195e-06,
+      "loss": 1.0634,
+      "step": 10455
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8643913269042969,
+      "learning_rate": 7.292390075319426e-06,
+      "loss": 0.919,
+      "step": 10456
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9154936075210571,
+      "learning_rate": 7.2906018985947095e-06,
+      "loss": 1.0358,
+      "step": 10457
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7114940881729126,
+      "learning_rate": 7.288813815363754e-06,
+      "loss": 1.0048,
+      "step": 10458
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8005897998809814,
+      "learning_rate": 7.287025825688261e-06,
+      "loss": 0.9033,
+      "step": 10459
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8557074069976807,
+      "learning_rate": 7.285237929629928e-06,
+      "loss": 1.0136,
+      "step": 10460
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6670433282852173,
+      "learning_rate": 7.283450127250451e-06,
+      "loss": 0.9328,
+      "step": 10461
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7676392793655396,
+      "learning_rate": 7.281662418611521e-06,
+      "loss": 0.9511,
+      "step": 10462
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0404930114746094,
+      "learning_rate": 7.279874803774828e-06,
+      "loss": 0.9822,
+      "step": 10463
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8812988996505737,
+      "learning_rate": 7.2780872828020556e-06,
+      "loss": 0.8572,
+      "step": 10464
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6840405464172363,
+      "learning_rate": 7.2762998557548894e-06,
+      "loss": 0.9332,
+      "step": 10465
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7049657106399536,
+      "learning_rate": 7.274512522695006e-06,
+      "loss": 0.8929,
+      "step": 10466
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7308307886123657,
+      "learning_rate": 7.27272528368408e-06,
+      "loss": 0.8621,
+      "step": 10467
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9550319910049438,
+      "learning_rate": 7.2709381387837894e-06,
+      "loss": 0.8522,
+      "step": 10468
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7165430784225464,
+      "learning_rate": 7.269151088055799e-06,
+      "loss": 0.9454,
+      "step": 10469
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5933412313461304,
+      "learning_rate": 7.267364131561775e-06,
+      "loss": 0.9117,
+      "step": 10470
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8131766319274902,
+      "learning_rate": 7.265577269363384e-06,
+      "loss": 0.9682,
+      "step": 10471
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0049012899398804,
+      "learning_rate": 7.263790501522282e-06,
+      "loss": 0.5249,
+      "step": 10472
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.990783929824829,
+      "learning_rate": 7.262003828100127e-06,
+      "loss": 0.8583,
+      "step": 10473
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7652831077575684,
+      "learning_rate": 7.260217249158572e-06,
+      "loss": 0.8636,
+      "step": 10474
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.781686544418335,
+      "learning_rate": 7.258430764759266e-06,
+      "loss": 0.8908,
+      "step": 10475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7794626951217651,
+      "learning_rate": 7.256644374963857e-06,
+      "loss": 0.9199,
+      "step": 10476
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.717576503753662,
+      "learning_rate": 7.254858079833986e-06,
+      "loss": 0.9136,
+      "step": 10477
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.561353087425232,
+      "learning_rate": 7.253071879431295e-06,
+      "loss": 0.956,
+      "step": 10478
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6979806423187256,
+      "learning_rate": 7.25128577381742e-06,
+      "loss": 0.9309,
+      "step": 10479
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.685271978378296,
+      "learning_rate": 7.249499763053996e-06,
+      "loss": 0.9505,
+      "step": 10480
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6365466117858887,
+      "learning_rate": 7.24771384720265e-06,
+      "loss": 0.9827,
+      "step": 10481
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0076375007629395,
+      "learning_rate": 7.24592802632501e-06,
+      "loss": 0.5142,
+      "step": 10482
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7766143083572388,
+      "learning_rate": 7.2441423004827016e-06,
+      "loss": 0.9976,
+      "step": 10483
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.996250569820404,
+      "learning_rate": 7.2423566697373445e-06,
+      "loss": 0.5652,
+      "step": 10484
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6844693422317505,
+      "learning_rate": 7.240571134150558e-06,
+      "loss": 0.8965,
+      "step": 10485
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5797817707061768,
+      "learning_rate": 7.238785693783951e-06,
+      "loss": 0.9805,
+      "step": 10486
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8870216608047485,
+      "learning_rate": 7.237000348699137e-06,
+      "loss": 0.8402,
+      "step": 10487
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.679835319519043,
+      "learning_rate": 7.235215098957723e-06,
+      "loss": 0.9737,
+      "step": 10488
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6840558052062988,
+      "learning_rate": 7.233429944621313e-06,
+      "loss": 0.9124,
+      "step": 10489
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4455772638320923,
+      "learning_rate": 7.2316448857515076e-06,
+      "loss": 0.542,
+      "step": 10490
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.719997525215149,
+      "learning_rate": 7.229859922409903e-06,
+      "loss": 0.9594,
+      "step": 10491
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.718862771987915,
+      "learning_rate": 7.228075054658096e-06,
+      "loss": 0.8893,
+      "step": 10492
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6912163496017456,
+      "learning_rate": 7.226290282557675e-06,
+      "loss": 0.8767,
+      "step": 10493
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9682996273040771,
+      "learning_rate": 7.224505606170227e-06,
+      "loss": 0.9355,
+      "step": 10494
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0403172969818115,
+      "learning_rate": 7.222721025557337e-06,
+      "loss": 0.4955,
+      "step": 10495
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7146739959716797,
+      "learning_rate": 7.22093654078059e-06,
+      "loss": 0.8837,
+      "step": 10496
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.006544351577759,
+      "learning_rate": 7.219152151901558e-06,
+      "loss": 0.8625,
+      "step": 10497
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.672311782836914,
+      "learning_rate": 7.217367858981818e-06,
+      "loss": 0.8896,
+      "step": 10498
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8284478187561035,
+      "learning_rate": 7.215583662082939e-06,
+      "loss": 0.9508,
+      "step": 10499
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8492563962936401,
+      "learning_rate": 7.21379956126649e-06,
+      "loss": 0.9396,
+      "step": 10500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.769406795501709,
+      "learning_rate": 7.212015556594037e-06,
+      "loss": 0.9696,
+      "step": 10501
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7644988298416138,
+      "learning_rate": 7.2102316481271376e-06,
+      "loss": 0.8542,
+      "step": 10502
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5977485179901123,
+      "learning_rate": 7.2084478359273514e-06,
+      "loss": 0.8127,
+      "step": 10503
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9385881423950195,
+      "learning_rate": 7.206664120056232e-06,
+      "loss": 1.0041,
+      "step": 10504
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7143248319625854,
+      "learning_rate": 7.204880500575333e-06,
+      "loss": 0.914,
+      "step": 10505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7707990407943726,
+      "learning_rate": 7.203096977546196e-06,
+      "loss": 0.874,
+      "step": 10506
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6297681331634521,
+      "learning_rate": 7.201313551030373e-06,
+      "loss": 0.944,
+      "step": 10507
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7276585102081299,
+      "learning_rate": 7.199530221089399e-06,
+      "loss": 0.9402,
+      "step": 10508
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7412209510803223,
+      "learning_rate": 7.1977469877848175e-06,
+      "loss": 0.965,
+      "step": 10509
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6354182958602905,
+      "learning_rate": 7.195963851178157e-06,
+      "loss": 0.9107,
+      "step": 10510
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6675649881362915,
+      "learning_rate": 7.194180811330953e-06,
+      "loss": 0.9716,
+      "step": 10511
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.645922303199768,
+      "learning_rate": 7.1923978683047305e-06,
+      "loss": 0.9131,
+      "step": 10512
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6072906255722046,
+      "learning_rate": 7.190615022161015e-06,
+      "loss": 0.9417,
+      "step": 10513
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8283355236053467,
+      "learning_rate": 7.188832272961328e-06,
+      "loss": 0.9146,
+      "step": 10514
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.786779522895813,
+      "learning_rate": 7.187049620767186e-06,
+      "loss": 0.9453,
+      "step": 10515
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7378132343292236,
+      "learning_rate": 7.1852670656401036e-06,
+      "loss": 0.9004,
+      "step": 10516
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7554913759231567,
+      "learning_rate": 7.183484607641593e-06,
+      "loss": 0.95,
+      "step": 10517
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.108429431915283,
+      "learning_rate": 7.181702246833158e-06,
+      "loss": 0.9026,
+      "step": 10518
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.702378511428833,
+      "learning_rate": 7.179919983276309e-06,
+      "loss": 0.9425,
+      "step": 10519
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7587751150131226,
+      "learning_rate": 7.178137817032542e-06,
+      "loss": 1.0535,
+      "step": 10520
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0380104780197144,
+      "learning_rate": 7.176355748163358e-06,
+      "loss": 0.6182,
+      "step": 10521
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6112908124923706,
+      "learning_rate": 7.17457377673025e-06,
+      "loss": 0.8,
+      "step": 10522
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8520015478134155,
+      "learning_rate": 7.1727919027947064e-06,
+      "loss": 0.9574,
+      "step": 10523
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7646820545196533,
+      "learning_rate": 7.171010126418218e-06,
+      "loss": 1.0021,
+      "step": 10524
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8538589477539062,
+      "learning_rate": 7.169228447662269e-06,
+      "loss": 0.9343,
+      "step": 10525
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6463186740875244,
+      "learning_rate": 7.167446866588337e-06,
+      "loss": 0.9269,
+      "step": 10526
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7062673568725586,
+      "learning_rate": 7.165665383257902e-06,
+      "loss": 0.9435,
+      "step": 10527
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6561214923858643,
+      "learning_rate": 7.1638839977324374e-06,
+      "loss": 0.8998,
+      "step": 10528
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.766083836555481,
+      "learning_rate": 7.162102710073413e-06,
+      "loss": 0.9265,
+      "step": 10529
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.144442319869995,
+      "learning_rate": 7.1603215203422945e-06,
+      "loss": 0.9792,
+      "step": 10530
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.583807110786438,
+      "learning_rate": 7.158540428600551e-06,
+      "loss": 0.9106,
+      "step": 10531
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0937477350234985,
+      "learning_rate": 7.1567594349096395e-06,
+      "loss": 0.5903,
+      "step": 10532
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7128753662109375,
+      "learning_rate": 7.154978539331015e-06,
+      "loss": 0.9245,
+      "step": 10533
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5351214408874512,
+      "learning_rate": 7.153197741926137e-06,
+      "loss": 0.8612,
+      "step": 10534
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7781466245651245,
+      "learning_rate": 7.1514170427564525e-06,
+      "loss": 0.9275,
+      "step": 10535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.757417917251587,
+      "learning_rate": 7.149636441883405e-06,
+      "loss": 0.8927,
+      "step": 10536
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8818066120147705,
+      "learning_rate": 7.147855939368445e-06,
+      "loss": 0.9756,
+      "step": 10537
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6473575830459595,
+      "learning_rate": 7.14607553527301e-06,
+      "loss": 0.9219,
+      "step": 10538
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.020093321800232,
+      "learning_rate": 7.144295229658536e-06,
+      "loss": 0.5469,
+      "step": 10539
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7127479314804077,
+      "learning_rate": 7.142515022586456e-06,
+      "loss": 0.8661,
+      "step": 10540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7459585666656494,
+      "learning_rate": 7.1407349141182e-06,
+      "loss": 0.9061,
+      "step": 10541
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7850323915481567,
+      "learning_rate": 7.138954904315196e-06,
+      "loss": 0.9114,
+      "step": 10542
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9461034536361694,
+      "learning_rate": 7.137174993238865e-06,
+      "loss": 1.0089,
+      "step": 10543
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.715523362159729,
+      "learning_rate": 7.13539518095063e-06,
+      "loss": 0.9641,
+      "step": 10544
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6246423721313477,
+      "learning_rate": 7.1336154675119044e-06,
+      "loss": 0.936,
+      "step": 10545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7631171941757202,
+      "learning_rate": 7.131835852984102e-06,
+      "loss": 0.9249,
+      "step": 10546
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6301947832107544,
+      "learning_rate": 7.130056337428633e-06,
+      "loss": 0.9562,
+      "step": 10547
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.62639319896698,
+      "learning_rate": 7.1282769209069005e-06,
+      "loss": 0.9001,
+      "step": 10548
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9143786430358887,
+      "learning_rate": 7.126497603480311e-06,
+      "loss": 0.8719,
+      "step": 10549
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0039377212524414,
+      "learning_rate": 7.124718385210263e-06,
+      "loss": 0.9298,
+      "step": 10550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.73677396774292,
+      "learning_rate": 7.122939266158151e-06,
+      "loss": 0.8826,
+      "step": 10551
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.623577356338501,
+      "learning_rate": 7.121160246385369e-06,
+      "loss": 0.9616,
+      "step": 10552
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.02581524848938,
+      "learning_rate": 7.119381325953305e-06,
+      "loss": 0.915,
+      "step": 10553
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5755760669708252,
+      "learning_rate": 7.117602504923345e-06,
+      "loss": 0.9115,
+      "step": 10554
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8273218870162964,
+      "learning_rate": 7.11582378335687e-06,
+      "loss": 0.8953,
+      "step": 10555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.772381067276001,
+      "learning_rate": 7.11404516131526e-06,
+      "loss": 0.9046,
+      "step": 10556
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6744062900543213,
+      "learning_rate": 7.11226663885989e-06,
+      "loss": 0.8696,
+      "step": 10557
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6818684339523315,
+      "learning_rate": 7.110488216052133e-06,
+      "loss": 0.9822,
+      "step": 10558
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8032218217849731,
+      "learning_rate": 7.108709892953355e-06,
+      "loss": 0.9339,
+      "step": 10559
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.751014232635498,
+      "learning_rate": 7.106931669624919e-06,
+      "loss": 0.8842,
+      "step": 10560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7469888925552368,
+      "learning_rate": 7.105153546128194e-06,
+      "loss": 0.9295,
+      "step": 10561
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7836576700210571,
+      "learning_rate": 7.1033755225245315e-06,
+      "loss": 0.889,
+      "step": 10562
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7243170738220215,
+      "learning_rate": 7.10159759887529e-06,
+      "loss": 0.9444,
+      "step": 10563
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9207457304000854,
+      "learning_rate": 7.099819775241818e-06,
+      "loss": 0.9771,
+      "step": 10564
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.876839280128479,
+      "learning_rate": 7.0980420516854655e-06,
+      "loss": 0.9412,
+      "step": 10565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0860122442245483,
+      "learning_rate": 7.096264428267574e-06,
+      "loss": 0.5333,
+      "step": 10566
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6661583185195923,
+      "learning_rate": 7.094486905049487e-06,
+      "loss": 1.0143,
+      "step": 10567
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8619788885116577,
+      "learning_rate": 7.092709482092539e-06,
+      "loss": 0.9895,
+      "step": 10568
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5746023654937744,
+      "learning_rate": 7.090932159458067e-06,
+      "loss": 0.9401,
+      "step": 10569
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7725682258605957,
+      "learning_rate": 7.0891549372073996e-06,
+      "loss": 0.99,
+      "step": 10570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.750618815422058,
+      "learning_rate": 7.0873778154018636e-06,
+      "loss": 0.8965,
+      "step": 10571
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7544050216674805,
+      "learning_rate": 7.085600794102783e-06,
+      "loss": 0.9502,
+      "step": 10572
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8539271354675293,
+      "learning_rate": 7.0838238733714785e-06,
+      "loss": 0.9348,
+      "step": 10573
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5890002250671387,
+      "learning_rate": 7.0820470532692654e-06,
+      "loss": 0.8919,
+      "step": 10574
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7070835828781128,
+      "learning_rate": 7.080270333857459e-06,
+      "loss": 0.8832,
+      "step": 10575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6903101205825806,
+      "learning_rate": 7.0784937151973666e-06,
+      "loss": 0.9332,
+      "step": 10576
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.779494047164917,
+      "learning_rate": 7.0767171973502955e-06,
+      "loss": 0.9111,
+      "step": 10577
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.853913426399231,
+      "learning_rate": 7.074940780377548e-06,
+      "loss": 0.9215,
+      "step": 10578
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.840926170349121,
+      "learning_rate": 7.073164464340423e-06,
+      "loss": 0.8718,
+      "step": 10579
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.883320689201355,
+      "learning_rate": 7.071388249300217e-06,
+      "loss": 0.949,
+      "step": 10580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.708052158355713,
+      "learning_rate": 7.069612135318222e-06,
+      "loss": 0.9655,
+      "step": 10581
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7948048114776611,
+      "learning_rate": 7.0678361224557265e-06,
+      "loss": 0.8808,
+      "step": 10582
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.682698130607605,
+      "learning_rate": 7.066060210774015e-06,
+      "loss": 0.9309,
+      "step": 10583
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7539793252944946,
+      "learning_rate": 7.064284400334369e-06,
+      "loss": 0.9424,
+      "step": 10584
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.015618085861206,
+      "learning_rate": 7.0625086911980685e-06,
+      "loss": 0.6048,
+      "step": 10585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7441778182983398,
+      "learning_rate": 7.060733083426389e-06,
+      "loss": 0.894,
+      "step": 10586
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7858424186706543,
+      "learning_rate": 7.058957577080599e-06,
+      "loss": 0.8866,
+      "step": 10587
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6261181831359863,
+      "learning_rate": 7.057182172221968e-06,
+      "loss": 0.8644,
+      "step": 10588
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6882188320159912,
+      "learning_rate": 7.055406868911761e-06,
+      "loss": 0.9653,
+      "step": 10589
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.735335350036621,
+      "learning_rate": 7.053631667211236e-06,
+      "loss": 0.8811,
+      "step": 10590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9600070714950562,
+      "learning_rate": 7.051856567181652e-06,
+      "loss": 0.9558,
+      "step": 10591
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7807068824768066,
+      "learning_rate": 7.0500815688842614e-06,
+      "loss": 0.9736,
+      "step": 10592
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8206084966659546,
+      "learning_rate": 7.048306672380318e-06,
+      "loss": 0.946,
+      "step": 10593
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8796347379684448,
+      "learning_rate": 7.046531877731065e-06,
+      "loss": 1.0012,
+      "step": 10594
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.684680461883545,
+      "learning_rate": 7.044757184997747e-06,
+      "loss": 0.8911,
+      "step": 10595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6220518350601196,
+      "learning_rate": 7.0429825942416e-06,
+      "loss": 0.9426,
+      "step": 10596
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8553264141082764,
+      "learning_rate": 7.0412081055238675e-06,
+      "loss": 0.9212,
+      "step": 10597
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7969778776168823,
+      "learning_rate": 7.039433718905777e-06,
+      "loss": 0.8835,
+      "step": 10598
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9352818727493286,
+      "learning_rate": 7.0376594344485586e-06,
+      "loss": 0.9461,
+      "step": 10599
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.75458824634552,
+      "learning_rate": 7.035885252213439e-06,
+      "loss": 1.0264,
+      "step": 10600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6329597234725952,
+      "learning_rate": 7.03411117226164e-06,
+      "loss": 0.9338,
+      "step": 10601
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.737112283706665,
+      "learning_rate": 7.032337194654375e-06,
+      "loss": 0.9172,
+      "step": 10602
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7277860641479492,
+      "learning_rate": 7.0305633194528675e-06,
+      "loss": 0.9527,
+      "step": 10603
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8295855522155762,
+      "learning_rate": 7.028789546718327e-06,
+      "loss": 1.051,
+      "step": 10604
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8651982545852661,
+      "learning_rate": 7.027015876511955e-06,
+      "loss": 0.9216,
+      "step": 10605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8102905750274658,
+      "learning_rate": 7.025242308894964e-06,
+      "loss": 0.9481,
+      "step": 10606
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8126261234283447,
+      "learning_rate": 7.02346884392855e-06,
+      "loss": 0.8496,
+      "step": 10607
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8048622608184814,
+      "learning_rate": 7.021695481673912e-06,
+      "loss": 0.9313,
+      "step": 10608
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9112160205841064,
+      "learning_rate": 7.019922222192243e-06,
+      "loss": 0.9956,
+      "step": 10609
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7338054180145264,
+      "learning_rate": 7.018149065544735e-06,
+      "loss": 1.0393,
+      "step": 10610
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0718231201171875,
+      "learning_rate": 7.016376011792572e-06,
+      "loss": 0.8995,
+      "step": 10611
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6575974225997925,
+      "learning_rate": 7.0146030609969385e-06,
+      "loss": 0.8795,
+      "step": 10612
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.819928526878357,
+      "learning_rate": 7.012830213219013e-06,
+      "loss": 0.8589,
+      "step": 10613
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.666174054145813,
+      "learning_rate": 7.011057468519973e-06,
+      "loss": 0.9066,
+      "step": 10614
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8539707660675049,
+      "learning_rate": 7.009284826960989e-06,
+      "loss": 0.8985,
+      "step": 10615
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9939125776290894,
+      "learning_rate": 7.007512288603234e-06,
+      "loss": 0.9897,
+      "step": 10616
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7586653232574463,
+      "learning_rate": 7.005739853507871e-06,
+      "loss": 1.0008,
+      "step": 10617
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0530898571014404,
+      "learning_rate": 7.00396752173606e-06,
+      "loss": 0.9621,
+      "step": 10618
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9518625736236572,
+      "learning_rate": 7.002195293348961e-06,
+      "loss": 0.9535,
+      "step": 10619
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.890276551246643,
+      "learning_rate": 7.00042316840773e-06,
+      "loss": 0.9704,
+      "step": 10620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9246728420257568,
+      "learning_rate": 6.9986511469735145e-06,
+      "loss": 0.8925,
+      "step": 10621
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.808403491973877,
+      "learning_rate": 6.9968792291074646e-06,
+      "loss": 0.9285,
+      "step": 10622
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9037803411483765,
+      "learning_rate": 6.995107414870725e-06,
+      "loss": 0.9468,
+      "step": 10623
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7801018953323364,
+      "learning_rate": 6.9933357043244335e-06,
+      "loss": 0.8925,
+      "step": 10624
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5667699575424194,
+      "learning_rate": 6.991564097529727e-06,
+      "loss": 0.9389,
+      "step": 10625
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8979694843292236,
+      "learning_rate": 6.989792594547739e-06,
+      "loss": 1.0365,
+      "step": 10626
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.711273193359375,
+      "learning_rate": 6.988021195439603e-06,
+      "loss": 0.8901,
+      "step": 10627
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8914388418197632,
+      "learning_rate": 6.98624990026644e-06,
+      "loss": 0.8564,
+      "step": 10628
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.753403663635254,
+      "learning_rate": 6.984478709089375e-06,
+      "loss": 0.9645,
+      "step": 10629
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5992937088012695,
+      "learning_rate": 6.9827076219695254e-06,
+      "loss": 0.9076,
+      "step": 10630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0181591510772705,
+      "learning_rate": 6.9809366389680075e-06,
+      "loss": 0.9307,
+      "step": 10631
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8274506330490112,
+      "learning_rate": 6.979165760145934e-06,
+      "loss": 1.029,
+      "step": 10632
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6901062726974487,
+      "learning_rate": 6.977394985564412e-06,
+      "loss": 0.9327,
+      "step": 10633
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8407485485076904,
+      "learning_rate": 6.975624315284544e-06,
+      "loss": 0.9415,
+      "step": 10634
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6040914058685303,
+      "learning_rate": 6.973853749367434e-06,
+      "loss": 0.9332,
+      "step": 10635
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1123594045639038,
+      "learning_rate": 6.9720832878741776e-06,
+      "loss": 0.5664,
+      "step": 10636
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6647958755493164,
+      "learning_rate": 6.970312930865868e-06,
+      "loss": 0.9076,
+      "step": 10637
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.694380283355713,
+      "learning_rate": 6.968542678403596e-06,
+      "loss": 0.8781,
+      "step": 10638
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7860279083251953,
+      "learning_rate": 6.966772530548448e-06,
+      "loss": 1.0284,
+      "step": 10639
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7016037702560425,
+      "learning_rate": 6.965002487361507e-06,
+      "loss": 0.9512,
+      "step": 10640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9999542236328125,
+      "learning_rate": 6.963232548903853e-06,
+      "loss": 0.9682,
+      "step": 10641
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8749598264694214,
+      "learning_rate": 6.9614627152365625e-06,
+      "loss": 0.9298,
+      "step": 10642
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0003541707992554,
+      "learning_rate": 6.959692986420703e-06,
+      "loss": 0.5844,
+      "step": 10643
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9773035049438477,
+      "learning_rate": 6.957923362517348e-06,
+      "loss": 0.5531,
+      "step": 10644
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7271779775619507,
+      "learning_rate": 6.956153843587559e-06,
+      "loss": 0.9272,
+      "step": 10645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.640600562095642,
+      "learning_rate": 6.954384429692398e-06,
+      "loss": 0.929,
+      "step": 10646
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7854920625686646,
+      "learning_rate": 6.9526151208929234e-06,
+      "loss": 0.9038,
+      "step": 10647
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6040844917297363,
+      "learning_rate": 6.950845917250188e-06,
+      "loss": 0.9255,
+      "step": 10648
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7694549560546875,
+      "learning_rate": 6.9490768188252435e-06,
+      "loss": 1.0054,
+      "step": 10649
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6183125972747803,
+      "learning_rate": 6.947307825679133e-06,
+      "loss": 0.9455,
+      "step": 10650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8158743381500244,
+      "learning_rate": 6.945538937872903e-06,
+      "loss": 1.0143,
+      "step": 10651
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8781062364578247,
+      "learning_rate": 6.943770155467593e-06,
+      "loss": 0.9514,
+      "step": 10652
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7817882299423218,
+      "learning_rate": 6.9420014785242374e-06,
+      "loss": 0.9948,
+      "step": 10653
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7521567344665527,
+      "learning_rate": 6.940232907103868e-06,
+      "loss": 0.9492,
+      "step": 10654
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6803890466690063,
+      "learning_rate": 6.9384644412675165e-06,
+      "loss": 0.9569,
+      "step": 10655
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9495131969451904,
+      "learning_rate": 6.936696081076202e-06,
+      "loss": 0.5421,
+      "step": 10656
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7499480247497559,
+      "learning_rate": 6.9349278265909506e-06,
+      "loss": 0.9578,
+      "step": 10657
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.877681016921997,
+      "learning_rate": 6.933159677872776e-06,
+      "loss": 0.9675,
+      "step": 10658
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7409342527389526,
+      "learning_rate": 6.931391634982696e-06,
+      "loss": 0.9898,
+      "step": 10659
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9767569899559021,
+      "learning_rate": 6.9296236979817175e-06,
+      "loss": 0.5647,
+      "step": 10660
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8217902183532715,
+      "learning_rate": 6.92785586693085e-06,
+      "loss": 1.0466,
+      "step": 10661
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5689538717269897,
+      "learning_rate": 6.926088141891092e-06,
+      "loss": 0.9663,
+      "step": 10662
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7143189907073975,
+      "learning_rate": 6.924320522923448e-06,
+      "loss": 0.8771,
+      "step": 10663
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7888163328170776,
+      "learning_rate": 6.9225530100889105e-06,
+      "loss": 0.8921,
+      "step": 10664
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8386070728302002,
+      "learning_rate": 6.92078560344847e-06,
+      "loss": 0.9951,
+      "step": 10665
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7091959714889526,
+      "learning_rate": 6.9190183030631185e-06,
+      "loss": 1.0207,
+      "step": 10666
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7623509168624878,
+      "learning_rate": 6.917251108993841e-06,
+      "loss": 0.9466,
+      "step": 10667
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.561977505683899,
+      "learning_rate": 6.915484021301613e-06,
+      "loss": 0.91,
+      "step": 10668
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9120452404022217,
+      "learning_rate": 6.9137170400474164e-06,
+      "loss": 0.8985,
+      "step": 10669
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.798666000366211,
+      "learning_rate": 6.911950165292225e-06,
+      "loss": 0.9788,
+      "step": 10670
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.042483925819397,
+      "learning_rate": 6.9101833970970074e-06,
+      "loss": 0.5321,
+      "step": 10671
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6670053005218506,
+      "learning_rate": 6.9084167355227295e-06,
+      "loss": 1.0227,
+      "step": 10672
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7947410345077515,
+      "learning_rate": 6.906650180630353e-06,
+      "loss": 0.9832,
+      "step": 10673
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.788172721862793,
+      "learning_rate": 6.904883732480838e-06,
+      "loss": 1.0008,
+      "step": 10674
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8236624002456665,
+      "learning_rate": 6.903117391135141e-06,
+      "loss": 0.953,
+      "step": 10675
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8077296018600464,
+      "learning_rate": 6.90135115665421e-06,
+      "loss": 0.9215,
+      "step": 10676
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.737565517425537,
+      "learning_rate": 6.899585029098996e-06,
+      "loss": 0.8719,
+      "step": 10677
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7289538383483887,
+      "learning_rate": 6.897819008530442e-06,
+      "loss": 0.9999,
+      "step": 10678
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9687201976776123,
+      "learning_rate": 6.89605309500949e-06,
+      "loss": 0.511,
+      "step": 10679
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8477734327316284,
+      "learning_rate": 6.89428728859707e-06,
+      "loss": 0.9496,
+      "step": 10680
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0349655151367188,
+      "learning_rate": 6.892521589354124e-06,
+      "loss": 0.9415,
+      "step": 10681
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8025058507919312,
+      "learning_rate": 6.8907559973415776e-06,
+      "loss": 0.9244,
+      "step": 10682
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6989761590957642,
+      "learning_rate": 6.888990512620356e-06,
+      "loss": 0.9642,
+      "step": 10683
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.666062831878662,
+      "learning_rate": 6.887225135251381e-06,
+      "loss": 0.9645,
+      "step": 10684
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7569764852523804,
+      "learning_rate": 6.885459865295573e-06,
+      "loss": 0.9722,
+      "step": 10685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8222192525863647,
+      "learning_rate": 6.883694702813843e-06,
+      "loss": 0.9852,
+      "step": 10686
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7695648670196533,
+      "learning_rate": 6.881929647867105e-06,
+      "loss": 1.0292,
+      "step": 10687
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5964521169662476,
+      "learning_rate": 6.880164700516265e-06,
+      "loss": 0.8765,
+      "step": 10688
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6771517992019653,
+      "learning_rate": 6.878399860822226e-06,
+      "loss": 0.8959,
+      "step": 10689
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8062862157821655,
+      "learning_rate": 6.876635128845888e-06,
+      "loss": 0.9944,
+      "step": 10690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.052683115005493,
+      "learning_rate": 6.874870504648147e-06,
+      "loss": 0.9878,
+      "step": 10691
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.04588782787323,
+      "learning_rate": 6.873105988289892e-06,
+      "loss": 0.5483,
+      "step": 10692
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.980678915977478,
+      "learning_rate": 6.871341579832018e-06,
+      "loss": 0.9433,
+      "step": 10693
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6742863655090332,
+      "learning_rate": 6.869577279335407e-06,
+      "loss": 0.998,
+      "step": 10694
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8063756227493286,
+      "learning_rate": 6.867813086860939e-06,
+      "loss": 0.9092,
+      "step": 10695
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6749621629714966,
+      "learning_rate": 6.8660490024694905e-06,
+      "loss": 0.8654,
+      "step": 10696
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7051033973693848,
+      "learning_rate": 6.864285026221939e-06,
+      "loss": 0.8724,
+      "step": 10697
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8245407342910767,
+      "learning_rate": 6.862521158179151e-06,
+      "loss": 0.9671,
+      "step": 10698
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9305305480957031,
+      "learning_rate": 6.860757398401994e-06,
+      "loss": 0.9709,
+      "step": 10699
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.756739616394043,
+      "learning_rate": 6.858993746951328e-06,
+      "loss": 0.9593,
+      "step": 10700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8252607583999634,
+      "learning_rate": 6.8572302038880155e-06,
+      "loss": 0.9419,
+      "step": 10701
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.632275104522705,
+      "learning_rate": 6.85546676927291e-06,
+      "loss": 0.855,
+      "step": 10702
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.02900767326355,
+      "learning_rate": 6.853703443166861e-06,
+      "loss": 0.9852,
+      "step": 10703
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0382273197174072,
+      "learning_rate": 6.851940225630718e-06,
+      "loss": 0.5807,
+      "step": 10704
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.691987156867981,
+      "learning_rate": 6.8501771167253224e-06,
+      "loss": 0.8632,
+      "step": 10705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7489248514175415,
+      "learning_rate": 6.848414116511519e-06,
+      "loss": 0.9458,
+      "step": 10706
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7351505756378174,
+      "learning_rate": 6.84665122505014e-06,
+      "loss": 0.8572,
+      "step": 10707
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7645676136016846,
+      "learning_rate": 6.844888442402018e-06,
+      "loss": 1.0089,
+      "step": 10708
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7964791059494019,
+      "learning_rate": 6.843125768627983e-06,
+      "loss": 0.9314,
+      "step": 10709
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9650331735610962,
+      "learning_rate": 6.841363203788858e-06,
+      "loss": 1.0,
+      "step": 10710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6995660066604614,
+      "learning_rate": 6.8396007479454675e-06,
+      "loss": 0.8429,
+      "step": 10711
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8449565172195435,
+      "learning_rate": 6.837838401158625e-06,
+      "loss": 0.9166,
+      "step": 10712
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5481001138687134,
+      "learning_rate": 6.836076163489147e-06,
+      "loss": 0.9429,
+      "step": 10713
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0663679838180542,
+      "learning_rate": 6.834314034997844e-06,
+      "loss": 0.5822,
+      "step": 10714
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8308523893356323,
+      "learning_rate": 6.832552015745519e-06,
+      "loss": 0.9478,
+      "step": 10715
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7743624448776245,
+      "learning_rate": 6.8307901057929735e-06,
+      "loss": 1.0199,
+      "step": 10716
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8001407384872437,
+      "learning_rate": 6.829028305201012e-06,
+      "loss": 0.8889,
+      "step": 10717
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7936240434646606,
+      "learning_rate": 6.8272666140304255e-06,
+      "loss": 0.9346,
+      "step": 10718
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0354218482971191,
+      "learning_rate": 6.825505032342005e-06,
+      "loss": 0.487,
+      "step": 10719
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6789531707763672,
+      "learning_rate": 6.823743560196539e-06,
+      "loss": 0.9096,
+      "step": 10720
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9396724700927734,
+      "learning_rate": 6.8219821976548104e-06,
+      "loss": 0.9735,
+      "step": 10721
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6277079582214355,
+      "learning_rate": 6.820220944777598e-06,
+      "loss": 0.9284,
+      "step": 10722
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6848450899124146,
+      "learning_rate": 6.818459801625679e-06,
+      "loss": 0.903,
+      "step": 10723
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8770413398742676,
+      "learning_rate": 6.816698768259824e-06,
+      "loss": 0.9984,
+      "step": 10724
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7362871170043945,
+      "learning_rate": 6.814937844740803e-06,
+      "loss": 0.8919,
+      "step": 10725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7478768825531006,
+      "learning_rate": 6.81317703112938e-06,
+      "loss": 0.9723,
+      "step": 10726
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7467293739318848,
+      "learning_rate": 6.811416327486316e-06,
+      "loss": 0.9714,
+      "step": 10727
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8072682619094849,
+      "learning_rate": 6.8096557338723665e-06,
+      "loss": 0.9422,
+      "step": 10728
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.956426978111267,
+      "learning_rate": 6.807895250348284e-06,
+      "loss": 0.9892,
+      "step": 10729
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7539492845535278,
+      "learning_rate": 6.806134876974821e-06,
+      "loss": 0.9577,
+      "step": 10730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7264825105667114,
+      "learning_rate": 6.804374613812721e-06,
+      "loss": 0.9687,
+      "step": 10731
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7004607915878296,
+      "learning_rate": 6.802614460922728e-06,
+      "loss": 1.0253,
+      "step": 10732
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.651534080505371,
+      "learning_rate": 6.800854418365579e-06,
+      "loss": 0.8433,
+      "step": 10733
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8990579843521118,
+      "learning_rate": 6.799094486202005e-06,
+      "loss": 0.9038,
+      "step": 10734
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7547316551208496,
+      "learning_rate": 6.797334664492741e-06,
+      "loss": 0.8984,
+      "step": 10735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.787498950958252,
+      "learning_rate": 6.795574953298511e-06,
+      "loss": 0.9147,
+      "step": 10736
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7551413774490356,
+      "learning_rate": 6.7938153526800386e-06,
+      "loss": 0.9337,
+      "step": 10737
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6865466833114624,
+      "learning_rate": 6.792055862698042e-06,
+      "loss": 0.9851,
+      "step": 10738
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6120564937591553,
+      "learning_rate": 6.790296483413237e-06,
+      "loss": 0.9142,
+      "step": 10739
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8226311206817627,
+      "learning_rate": 6.788537214886335e-06,
+      "loss": 1.0253,
+      "step": 10740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7005010843276978,
+      "learning_rate": 6.7867780571780416e-06,
+      "loss": 0.9272,
+      "step": 10741
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7235755920410156,
+      "learning_rate": 6.785019010349062e-06,
+      "loss": 0.9282,
+      "step": 10742
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7701646089553833,
+      "learning_rate": 6.783260074460096e-06,
+      "loss": 0.8534,
+      "step": 10743
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.613637089729309,
+      "learning_rate": 6.781501249571839e-06,
+      "loss": 0.9348,
+      "step": 10744
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.94866144657135,
+      "learning_rate": 6.7797425357449844e-06,
+      "loss": 0.8959,
+      "step": 10745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9007580280303955,
+      "learning_rate": 6.777983933040216e-06,
+      "loss": 0.9693,
+      "step": 10746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9180207252502441,
+      "learning_rate": 6.776225441518224e-06,
+      "loss": 1.0191,
+      "step": 10747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7450957298278809,
+      "learning_rate": 6.7744670612396866e-06,
+      "loss": 0.9105,
+      "step": 10748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7718079090118408,
+      "learning_rate": 6.7727087922652815e-06,
+      "loss": 1.0198,
+      "step": 10749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7065348625183105,
+      "learning_rate": 6.77095063465568e-06,
+      "loss": 0.9569,
+      "step": 10750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6835771799087524,
+      "learning_rate": 6.769192588471553e-06,
+      "loss": 0.8948,
+      "step": 10751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.660536766052246,
+      "learning_rate": 6.767434653773564e-06,
+      "loss": 0.9514,
+      "step": 10752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7290070056915283,
+      "learning_rate": 6.765676830622376e-06,
+      "loss": 0.9366,
+      "step": 10753
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7143102884292603,
+      "learning_rate": 6.7639191190786455e-06,
+      "loss": 0.8621,
+      "step": 10754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8951539993286133,
+      "learning_rate": 6.762161519203028e-06,
+      "loss": 0.938,
+      "step": 10755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5635262727737427,
+      "learning_rate": 6.760404031056169e-06,
+      "loss": 0.878,
+      "step": 10756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.835033655166626,
+      "learning_rate": 6.758646654698719e-06,
+      "loss": 0.9498,
+      "step": 10757
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6972578763961792,
+      "learning_rate": 6.756889390191317e-06,
+      "loss": 0.9662,
+      "step": 10758
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6809446811676025,
+      "learning_rate": 6.755132237594605e-06,
+      "loss": 0.8907,
+      "step": 10759
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8069835901260376,
+      "learning_rate": 6.753375196969214e-06,
+      "loss": 0.9287,
+      "step": 10760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5308797359466553,
+      "learning_rate": 6.751618268375777e-06,
+      "loss": 0.9434,
+      "step": 10761
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7994133234024048,
+      "learning_rate": 6.749861451874919e-06,
+      "loss": 0.9521,
+      "step": 10762
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5352572202682495,
+      "learning_rate": 6.748104747527265e-06,
+      "loss": 0.8142,
+      "step": 10763
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.821385383605957,
+      "learning_rate": 6.74634815539343e-06,
+      "loss": 0.9621,
+      "step": 10764
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.73631751537323,
+      "learning_rate": 6.744591675534033e-06,
+      "loss": 0.9163,
+      "step": 10765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7903461456298828,
+      "learning_rate": 6.742835308009683e-06,
+      "loss": 0.9126,
+      "step": 10766
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125685453414917,
+      "learning_rate": 6.7410790528809875e-06,
+      "loss": 1.1167,
+      "step": 10767
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6026840209960938,
+      "learning_rate": 6.7393229102085525e-06,
+      "loss": 0.8783,
+      "step": 10768
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8227370977401733,
+      "learning_rate": 6.737566880052973e-06,
+      "loss": 0.9831,
+      "step": 10769
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8788026571273804,
+      "learning_rate": 6.735810962474847e-06,
+      "loss": 1.0038,
+      "step": 10770
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6938592195510864,
+      "learning_rate": 6.734055157534768e-06,
+      "loss": 0.9911,
+      "step": 10771
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7798479795455933,
+      "learning_rate": 6.732299465293322e-06,
+      "loss": 0.8596,
+      "step": 10772
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1224095821380615,
+      "learning_rate": 6.730543885811094e-06,
+      "loss": 0.9381,
+      "step": 10773
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8793865442276,
+      "learning_rate": 6.728788419148664e-06,
+      "loss": 0.9947,
+      "step": 10774
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6216096878051758,
+      "learning_rate": 6.727033065366609e-06,
+      "loss": 0.9507,
+      "step": 10775
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7437671422958374,
+      "learning_rate": 6.725277824525498e-06,
+      "loss": 0.8766,
+      "step": 10776
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6781764030456543,
+      "learning_rate": 6.723522696685902e-06,
+      "loss": 0.8865,
+      "step": 10777
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.764045238494873,
+      "learning_rate": 6.721767681908386e-06,
+      "loss": 0.9362,
+      "step": 10778
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.993725299835205,
+      "learning_rate": 6.720012780253509e-06,
+      "loss": 1.0574,
+      "step": 10779
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5924664735794067,
+      "learning_rate": 6.7182579917818295e-06,
+      "loss": 0.9438,
+      "step": 10780
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4769564867019653,
+      "learning_rate": 6.716503316553899e-06,
+      "loss": 0.8964,
+      "step": 10781
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6932201385498047,
+      "learning_rate": 6.714748754630264e-06,
+      "loss": 0.9264,
+      "step": 10782
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9355340003967285,
+      "learning_rate": 6.712994306071476e-06,
+      "loss": 0.8863,
+      "step": 10783
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7098079919815063,
+      "learning_rate": 6.711239970938073e-06,
+      "loss": 0.9729,
+      "step": 10784
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6829544305801392,
+      "learning_rate": 6.709485749290592e-06,
+      "loss": 0.9553,
+      "step": 10785
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8375035524368286,
+      "learning_rate": 6.707731641189565e-06,
+      "loss": 0.9197,
+      "step": 10786
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7233394384384155,
+      "learning_rate": 6.705977646695523e-06,
+      "loss": 0.9421,
+      "step": 10787
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0112662315368652,
+      "learning_rate": 6.704223765868991e-06,
+      "loss": 0.5897,
+      "step": 10788
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1351630687713623,
+      "learning_rate": 6.70246999877049e-06,
+      "loss": 0.589,
+      "step": 10789
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6162692308425903,
+      "learning_rate": 6.700716345460538e-06,
+      "loss": 0.9482,
+      "step": 10790
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.758133888244629,
+      "learning_rate": 6.698962805999649e-06,
+      "loss": 0.9673,
+      "step": 10791
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8369076251983643,
+      "learning_rate": 6.697209380448333e-06,
+      "loss": 0.9569,
+      "step": 10792
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.950299620628357,
+      "learning_rate": 6.695456068867094e-06,
+      "loss": 0.8982,
+      "step": 10793
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7521032094955444,
+      "learning_rate": 6.693702871316436e-06,
+      "loss": 0.9578,
+      "step": 10794
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.783825397491455,
+      "learning_rate": 6.691949787856855e-06,
+      "loss": 0.8967,
+      "step": 10795
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6299785375595093,
+      "learning_rate": 6.690196818548846e-06,
+      "loss": 0.9325,
+      "step": 10796
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6558128595352173,
+      "learning_rate": 6.6884439634529e-06,
+      "loss": 0.9219,
+      "step": 10797
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7902382612228394,
+      "learning_rate": 6.686691222629503e-06,
+      "loss": 0.9059,
+      "step": 10798
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6823365688323975,
+      "learning_rate": 6.684938596139135e-06,
+      "loss": 0.9002,
+      "step": 10799
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7720131874084473,
+      "learning_rate": 6.683186084042276e-06,
+      "loss": 0.9064,
+      "step": 10800
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.748056411743164,
+      "learning_rate": 6.681433686399401e-06,
+      "loss": 0.9167,
+      "step": 10801
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7118057012557983,
+      "learning_rate": 6.67968140327098e-06,
+      "loss": 0.9403,
+      "step": 10802
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6603078842163086,
+      "learning_rate": 6.677929234717478e-06,
+      "loss": 0.9045,
+      "step": 10803
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.636451244354248,
+      "learning_rate": 6.676177180799359e-06,
+      "loss": 0.8681,
+      "step": 10804
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8545950651168823,
+      "learning_rate": 6.6744252415770806e-06,
+      "loss": 0.8167,
+      "step": 10805
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7148292064666748,
+      "learning_rate": 6.672673417111098e-06,
+      "loss": 0.947,
+      "step": 10806
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7503153085708618,
+      "learning_rate": 6.670921707461862e-06,
+      "loss": 0.8889,
+      "step": 10807
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7412528991699219,
+      "learning_rate": 6.669170112689816e-06,
+      "loss": 0.9614,
+      "step": 10808
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7296286821365356,
+      "learning_rate": 6.667418632855407e-06,
+      "loss": 0.9172,
+      "step": 10809
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9325380325317383,
+      "learning_rate": 6.665667268019071e-06,
+      "loss": 0.8946,
+      "step": 10810
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9250518083572388,
+      "learning_rate": 6.663916018241244e-06,
+      "loss": 0.9679,
+      "step": 10811
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0168627500534058,
+      "learning_rate": 6.662164883582354e-06,
+      "loss": 0.5764,
+      "step": 10812
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9645469188690186,
+      "learning_rate": 6.660413864102831e-06,
+      "loss": 1.0383,
+      "step": 10813
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8175413608551025,
+      "learning_rate": 6.658662959863098e-06,
+      "loss": 0.9694,
+      "step": 10814
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8003426790237427,
+      "learning_rate": 6.656912170923573e-06,
+      "loss": 0.8811,
+      "step": 10815
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5706884860992432,
+      "learning_rate": 6.6551614973446685e-06,
+      "loss": 0.9244,
+      "step": 10816
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.019797444343567,
+      "learning_rate": 6.653410939186799e-06,
+      "loss": 0.6027,
+      "step": 10817
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7114776372909546,
+      "learning_rate": 6.65166049651037e-06,
+      "loss": 0.9755,
+      "step": 10818
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6805094480514526,
+      "learning_rate": 6.6499101693757815e-06,
+      "loss": 0.9309,
+      "step": 10819
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6730667352676392,
+      "learning_rate": 6.648159957843438e-06,
+      "loss": 0.9252,
+      "step": 10820
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.744978904724121,
+      "learning_rate": 6.64640986197373e-06,
+      "loss": 0.9374,
+      "step": 10821
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6462736129760742,
+      "learning_rate": 6.6446598818270495e-06,
+      "loss": 0.8708,
+      "step": 10822
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8557629585266113,
+      "learning_rate": 6.642910017463784e-06,
+      "loss": 1.0394,
+      "step": 10823
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7058827877044678,
+      "learning_rate": 6.641160268944314e-06,
+      "loss": 0.9515,
+      "step": 10824
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7320085763931274,
+      "learning_rate": 6.6394106363290235e-06,
+      "loss": 0.9082,
+      "step": 10825
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.74837327003479,
+      "learning_rate": 6.637661119678284e-06,
+      "loss": 0.9166,
+      "step": 10826
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7110360860824585,
+      "learning_rate": 6.635911719052466e-06,
+      "loss": 0.9835,
+      "step": 10827
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9405287504196167,
+      "learning_rate": 6.634162434511939e-06,
+      "loss": 0.9671,
+      "step": 10828
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8028181791305542,
+      "learning_rate": 6.632413266117064e-06,
+      "loss": 0.8949,
+      "step": 10829
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.68267822265625,
+      "learning_rate": 6.6306642139281994e-06,
+      "loss": 0.8916,
+      "step": 10830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6243553161621094,
+      "learning_rate": 6.628915278005701e-06,
+      "loss": 0.9176,
+      "step": 10831
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7617520093917847,
+      "learning_rate": 6.627166458409919e-06,
+      "loss": 0.9344,
+      "step": 10832
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0125434398651123,
+      "learning_rate": 6.625417755201202e-06,
+      "loss": 1.0592,
+      "step": 10833
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7374382019042969,
+      "learning_rate": 6.623669168439893e-06,
+      "loss": 0.9634,
+      "step": 10834
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7962983846664429,
+      "learning_rate": 6.621920698186326e-06,
+      "loss": 0.9429,
+      "step": 10835
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.618694543838501,
+      "learning_rate": 6.620172344500841e-06,
+      "loss": 0.9134,
+      "step": 10836
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7785835266113281,
+      "learning_rate": 6.618424107443766e-06,
+      "loss": 0.9473,
+      "step": 10837
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7171711921691895,
+      "learning_rate": 6.616675987075432e-06,
+      "loss": 0.9392,
+      "step": 10838
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.709456443786621,
+      "learning_rate": 6.614927983456156e-06,
+      "loss": 1.0263,
+      "step": 10839
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7950149774551392,
+      "learning_rate": 6.613180096646261e-06,
+      "loss": 0.9222,
+      "step": 10840
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7731181383132935,
+      "learning_rate": 6.611432326706061e-06,
+      "loss": 0.8782,
+      "step": 10841
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9063584804534912,
+      "learning_rate": 6.609684673695864e-06,
+      "loss": 1.0037,
+      "step": 10842
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1089892387390137,
+      "learning_rate": 6.607937137675981e-06,
+      "loss": 0.9239,
+      "step": 10843
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9470454454421997,
+      "learning_rate": 6.606189718706711e-06,
+      "loss": 0.9464,
+      "step": 10844
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0197046995162964,
+      "learning_rate": 6.604442416848351e-06,
+      "loss": 0.5566,
+      "step": 10845
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7876802682876587,
+      "learning_rate": 6.6026952321612005e-06,
+      "loss": 0.9183,
+      "step": 10846
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8760367631912231,
+      "learning_rate": 6.6009481647055475e-06,
+      "loss": 0.8508,
+      "step": 10847
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8140838146209717,
+      "learning_rate": 6.599201214541677e-06,
+      "loss": 1.0108,
+      "step": 10848
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8265868425369263,
+      "learning_rate": 6.597454381729873e-06,
+      "loss": 0.8781,
+      "step": 10849
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8629993200302124,
+      "learning_rate": 6.595707666330414e-06,
+      "loss": 0.9431,
+      "step": 10850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7512820959091187,
+      "learning_rate": 6.5939610684035745e-06,
+      "loss": 0.9436,
+      "step": 10851
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6978334188461304,
+      "learning_rate": 6.592214588009624e-06,
+      "loss": 0.9382,
+      "step": 10852
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7078100442886353,
+      "learning_rate": 6.59046822520883e-06,
+      "loss": 0.9595,
+      "step": 10853
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0426899194717407,
+      "learning_rate": 6.588721980061452e-06,
+      "loss": 0.5675,
+      "step": 10854
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7512562274932861,
+      "learning_rate": 6.58697585262775e-06,
+      "loss": 0.9969,
+      "step": 10855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7518759965896606,
+      "learning_rate": 6.585229842967977e-06,
+      "loss": 0.8917,
+      "step": 10856
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8223766088485718,
+      "learning_rate": 6.583483951142384e-06,
+      "loss": 0.9971,
+      "step": 10857
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8145039081573486,
+      "learning_rate": 6.581738177211215e-06,
+      "loss": 0.9559,
+      "step": 10858
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8618780374526978,
+      "learning_rate": 6.5799925212347145e-06,
+      "loss": 0.9522,
+      "step": 10859
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7096915245056152,
+      "learning_rate": 6.578246983273118e-06,
+      "loss": 0.9688,
+      "step": 10860
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6276496648788452,
+      "learning_rate": 6.576501563386657e-06,
+      "loss": 0.8663,
+      "step": 10861
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5654850006103516,
+      "learning_rate": 6.574756261635567e-06,
+      "loss": 0.8401,
+      "step": 10862
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8588556051254272,
+      "learning_rate": 6.573011078080067e-06,
+      "loss": 0.8103,
+      "step": 10863
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8329823017120361,
+      "learning_rate": 6.571266012780386e-06,
+      "loss": 0.9849,
+      "step": 10864
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5993702411651611,
+      "learning_rate": 6.569521065796735e-06,
+      "loss": 0.9921,
+      "step": 10865
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.525145173072815,
+      "learning_rate": 6.56777623718933e-06,
+      "loss": 0.9066,
+      "step": 10866
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8859401941299438,
+      "learning_rate": 6.56603152701838e-06,
+      "loss": 0.8992,
+      "step": 10867
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6509398221969604,
+      "learning_rate": 6.564286935344088e-06,
+      "loss": 0.9792,
+      "step": 10868
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.602835774421692,
+      "learning_rate": 6.562542462226658e-06,
+      "loss": 0.8529,
+      "step": 10869
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7459063529968262,
+      "learning_rate": 6.560798107726285e-06,
+      "loss": 0.9177,
+      "step": 10870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.983681082725525,
+      "learning_rate": 6.559053871903163e-06,
+      "loss": 0.9336,
+      "step": 10871
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7293874025344849,
+      "learning_rate": 6.55730975481748e-06,
+      "loss": 0.9035,
+      "step": 10872
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7049914598464966,
+      "learning_rate": 6.55556575652942e-06,
+      "loss": 0.9272,
+      "step": 10873
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6534440517425537,
+      "learning_rate": 6.553821877099165e-06,
+      "loss": 0.9224,
+      "step": 10874
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7726149559020996,
+      "learning_rate": 6.55207811658689e-06,
+      "loss": 0.888,
+      "step": 10875
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0222344398498535,
+      "learning_rate": 6.550334475052767e-06,
+      "loss": 0.5555,
+      "step": 10876
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.712051272392273,
+      "learning_rate": 6.548590952556966e-06,
+      "loss": 0.9751,
+      "step": 10877
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9366486072540283,
+      "learning_rate": 6.546847549159648e-06,
+      "loss": 1.012,
+      "step": 10878
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9394659996032715,
+      "learning_rate": 6.545104264920978e-06,
+      "loss": 0.9222,
+      "step": 10879
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7507514953613281,
+      "learning_rate": 6.543361099901106e-06,
+      "loss": 0.867,
+      "step": 10880
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7260664701461792,
+      "learning_rate": 6.541618054160191e-06,
+      "loss": 0.8269,
+      "step": 10881
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7792997360229492,
+      "learning_rate": 6.539875127758373e-06,
+      "loss": 0.9738,
+      "step": 10882
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8152871131896973,
+      "learning_rate": 6.538132320755799e-06,
+      "loss": 0.9572,
+      "step": 10883
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.764359951019287,
+      "learning_rate": 6.53638963321261e-06,
+      "loss": 0.9687,
+      "step": 10884
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6781080961227417,
+      "learning_rate": 6.534647065188939e-06,
+      "loss": 0.993,
+      "step": 10885
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.753397822380066,
+      "learning_rate": 6.532904616744918e-06,
+      "loss": 0.9365,
+      "step": 10886
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9462578296661377,
+      "learning_rate": 6.531162287940672e-06,
+      "loss": 0.9443,
+      "step": 10887
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6521697044372559,
+      "learning_rate": 6.529420078836327e-06,
+      "loss": 0.9058,
+      "step": 10888
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6961244344711304,
+      "learning_rate": 6.527677989492001e-06,
+      "loss": 0.8956,
+      "step": 10889
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.67652428150177,
+      "learning_rate": 6.5259360199678046e-06,
+      "loss": 1.0197,
+      "step": 10890
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6533143520355225,
+      "learning_rate": 6.5241941703238545e-06,
+      "loss": 0.9823,
+      "step": 10891
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6874151229858398,
+      "learning_rate": 6.5224524406202535e-06,
+      "loss": 1.0069,
+      "step": 10892
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8553768396377563,
+      "learning_rate": 6.520710830917105e-06,
+      "loss": 0.9092,
+      "step": 10893
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6393564939498901,
+      "learning_rate": 6.518969341274508e-06,
+      "loss": 0.8662,
+      "step": 10894
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0096694231033325,
+      "learning_rate": 6.517227971752553e-06,
+      "loss": 0.5339,
+      "step": 10895
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6071406602859497,
+      "learning_rate": 6.515486722411334e-06,
+      "loss": 0.8773,
+      "step": 10896
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.675369143486023,
+      "learning_rate": 6.513745593310934e-06,
+      "loss": 0.966,
+      "step": 10897
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8798292875289917,
+      "learning_rate": 6.5120045845114344e-06,
+      "loss": 0.9706,
+      "step": 10898
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.766384243965149,
+      "learning_rate": 6.510263696072914e-06,
+      "loss": 0.9265,
+      "step": 10899
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.820424199104309,
+      "learning_rate": 6.508522928055445e-06,
+      "loss": 1.0018,
+      "step": 10900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5814709663391113,
+      "learning_rate": 6.5067822805190976e-06,
+      "loss": 0.9253,
+      "step": 10901
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6871063709259033,
+      "learning_rate": 6.505041753523932e-06,
+      "loss": 0.9669,
+      "step": 10902
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6259853839874268,
+      "learning_rate": 6.503301347130015e-06,
+      "loss": 1.0228,
+      "step": 10903
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.807197093963623,
+      "learning_rate": 6.501561061397402e-06,
+      "loss": 0.9348,
+      "step": 10904
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.557220220565796,
+      "learning_rate": 6.499820896386144e-06,
+      "loss": 0.911,
+      "step": 10905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5868818759918213,
+      "learning_rate": 6.4980808521562895e-06,
+      "loss": 0.8803,
+      "step": 10906
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6475913524627686,
+      "learning_rate": 6.496340928767881e-06,
+      "loss": 0.8941,
+      "step": 10907
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7701905965805054,
+      "learning_rate": 6.494601126280963e-06,
+      "loss": 0.9548,
+      "step": 10908
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8128606081008911,
+      "learning_rate": 6.492861444755566e-06,
+      "loss": 0.9861,
+      "step": 10909
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6365362405776978,
+      "learning_rate": 6.491121884251724e-06,
+      "loss": 0.9098,
+      "step": 10910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7748537063598633,
+      "learning_rate": 6.489382444829464e-06,
+      "loss": 1.024,
+      "step": 10911
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6241058111190796,
+      "learning_rate": 6.487643126548811e-06,
+      "loss": 0.8813,
+      "step": 10912
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8209609985351562,
+      "learning_rate": 6.485903929469782e-06,
+      "loss": 0.8999,
+      "step": 10913
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6941410303115845,
+      "learning_rate": 6.484164853652391e-06,
+      "loss": 1.0171,
+      "step": 10914
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8765860795974731,
+      "learning_rate": 6.482425899156647e-06,
+      "loss": 0.8233,
+      "step": 10915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7330796718597412,
+      "learning_rate": 6.480687066042562e-06,
+      "loss": 0.8856,
+      "step": 10916
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7512115240097046,
+      "learning_rate": 6.478948354370136e-06,
+      "loss": 0.9015,
+      "step": 10917
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5984995365142822,
+      "learning_rate": 6.477209764199366e-06,
+      "loss": 0.9206,
+      "step": 10918
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.857762336730957,
+      "learning_rate": 6.475471295590248e-06,
+      "loss": 0.9192,
+      "step": 10919
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8268460035324097,
+      "learning_rate": 6.473732948602769e-06,
+      "loss": 1.0099,
+      "step": 10920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6284695863723755,
+      "learning_rate": 6.471994723296915e-06,
+      "loss": 0.8808,
+      "step": 10921
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6751948595046997,
+      "learning_rate": 6.470256619732669e-06,
+      "loss": 0.9119,
+      "step": 10922
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7086914777755737,
+      "learning_rate": 6.4685186379700075e-06,
+      "loss": 0.8154,
+      "step": 10923
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6296610832214355,
+      "learning_rate": 6.466780778068903e-06,
+      "loss": 0.9486,
+      "step": 10924
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.778385043144226,
+      "learning_rate": 6.465043040089322e-06,
+      "loss": 0.9376,
+      "step": 10925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3001983165740967,
+      "learning_rate": 6.463305424091235e-06,
+      "loss": 0.9382,
+      "step": 10926
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6163650751113892,
+      "learning_rate": 6.461567930134593e-06,
+      "loss": 0.8851,
+      "step": 10927
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7785840034484863,
+      "learning_rate": 6.459830558279362e-06,
+      "loss": 0.9041,
+      "step": 10928
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8891711235046387,
+      "learning_rate": 6.458093308585486e-06,
+      "loss": 0.9212,
+      "step": 10929
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7698668241500854,
+      "learning_rate": 6.456356181112919e-06,
+      "loss": 0.9251,
+      "step": 10930
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8802356719970703,
+      "learning_rate": 6.4546191759216e-06,
+      "loss": 0.9468,
+      "step": 10931
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5903085470199585,
+      "learning_rate": 6.452882293071467e-06,
+      "loss": 1.0076,
+      "step": 10932
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7616667747497559,
+      "learning_rate": 6.451145532622463e-06,
+      "loss": 0.9344,
+      "step": 10933
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7080965042114258,
+      "learning_rate": 6.44940889463451e-06,
+      "loss": 0.9825,
+      "step": 10934
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6940077543258667,
+      "learning_rate": 6.44767237916754e-06,
+      "loss": 0.9852,
+      "step": 10935
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5831682682037354,
+      "learning_rate": 6.445935986281472e-06,
+      "loss": 0.8523,
+      "step": 10936
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8369783163070679,
+      "learning_rate": 6.444199716036225e-06,
+      "loss": 0.9479,
+      "step": 10937
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7305128574371338,
+      "learning_rate": 6.442463568491715e-06,
+      "loss": 0.9108,
+      "step": 10938
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8118075132369995,
+      "learning_rate": 6.440727543707847e-06,
+      "loss": 0.8433,
+      "step": 10939
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.747452974319458,
+      "learning_rate": 6.438991641744531e-06,
+      "loss": 0.966,
+      "step": 10940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0185604095458984,
+      "learning_rate": 6.437255862661664e-06,
+      "loss": 0.6005,
+      "step": 10941
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8497415781021118,
+      "learning_rate": 6.435520206519148e-06,
+      "loss": 1.0114,
+      "step": 10942
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7420477867126465,
+      "learning_rate": 6.43378467337687e-06,
+      "loss": 0.8999,
+      "step": 10943
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.77449369430542,
+      "learning_rate": 6.432049263294722e-06,
+      "loss": 0.9797,
+      "step": 10944
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6068028211593628,
+      "learning_rate": 6.4303139763325874e-06,
+      "loss": 0.8382,
+      "step": 10945
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8904341459274292,
+      "learning_rate": 6.428578812550346e-06,
+      "loss": 0.8673,
+      "step": 10946
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0405447483062744,
+      "learning_rate": 6.426843772007873e-06,
+      "loss": 0.5558,
+      "step": 10947
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9702852964401245,
+      "learning_rate": 6.425108854765041e-06,
+      "loss": 0.822,
+      "step": 10948
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7094571590423584,
+      "learning_rate": 6.423374060881716e-06,
+      "loss": 0.8728,
+      "step": 10949
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0238184928894043,
+      "learning_rate": 6.421639390417762e-06,
+      "loss": 0.8342,
+      "step": 10950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7719672918319702,
+      "learning_rate": 6.4199048434330355e-06,
+      "loss": 0.8763,
+      "step": 10951
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7524116039276123,
+      "learning_rate": 6.418170419987393e-06,
+      "loss": 0.9056,
+      "step": 10952
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7774206399917603,
+      "learning_rate": 6.416436120140684e-06,
+      "loss": 0.9202,
+      "step": 10953
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8391846418380737,
+      "learning_rate": 6.414701943952755e-06,
+      "loss": 0.9902,
+      "step": 10954
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7043883800506592,
+      "learning_rate": 6.412967891483446e-06,
+      "loss": 0.8446,
+      "step": 10955
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6859205961227417,
+      "learning_rate": 6.411233962792593e-06,
+      "loss": 0.8399,
+      "step": 10956
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.781709909439087,
+      "learning_rate": 6.409500157940033e-06,
+      "loss": 0.9385,
+      "step": 10957
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7395527362823486,
+      "learning_rate": 6.407766476985593e-06,
+      "loss": 0.924,
+      "step": 10958
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7877651453018188,
+      "learning_rate": 6.406032919989098e-06,
+      "loss": 0.9883,
+      "step": 10959
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6655069589614868,
+      "learning_rate": 6.404299487010366e-06,
+      "loss": 0.9283,
+      "step": 10960
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6002076864242554,
+      "learning_rate": 6.402566178109217e-06,
+      "loss": 0.912,
+      "step": 10961
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7050714492797852,
+      "learning_rate": 6.4008329933454585e-06,
+      "loss": 0.9527,
+      "step": 10962
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8286172151565552,
+      "learning_rate": 6.399099932778898e-06,
+      "loss": 0.9709,
+      "step": 10963
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.755947232246399,
+      "learning_rate": 6.397366996469343e-06,
+      "loss": 0.9697,
+      "step": 10964
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0392966270446777,
+      "learning_rate": 6.395634184476589e-06,
+      "loss": 0.523,
+      "step": 10965
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6759611368179321,
+      "learning_rate": 6.39390149686043e-06,
+      "loss": 0.966,
+      "step": 10966
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6869038343429565,
+      "learning_rate": 6.392168933680657e-06,
+      "loss": 0.8614,
+      "step": 10967
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.754591941833496,
+      "learning_rate": 6.390436494997055e-06,
+      "loss": 0.8291,
+      "step": 10968
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8557289838790894,
+      "learning_rate": 6.388704180869407e-06,
+      "loss": 0.8698,
+      "step": 10969
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7339762449264526,
+      "learning_rate": 6.386971991357491e-06,
+      "loss": 0.966,
+      "step": 10970
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7179497480392456,
+      "learning_rate": 6.385239926521078e-06,
+      "loss": 1.0051,
+      "step": 10971
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.726916790008545,
+      "learning_rate": 6.383507986419939e-06,
+      "loss": 0.9143,
+      "step": 10972
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.942336916923523,
+      "learning_rate": 6.381776171113837e-06,
+      "loss": 0.9306,
+      "step": 10973
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8387527465820312,
+      "learning_rate": 6.3800444806625325e-06,
+      "loss": 1.0069,
+      "step": 10974
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8318450450897217,
+      "learning_rate": 6.378312915125781e-06,
+      "loss": 0.8225,
+      "step": 10975
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.052971839904785,
+      "learning_rate": 6.376581474563332e-06,
+      "loss": 0.8635,
+      "step": 10976
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8037197589874268,
+      "learning_rate": 6.3748501590349374e-06,
+      "loss": 0.9644,
+      "step": 10977
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8193291425704956,
+      "learning_rate": 6.373118968600336e-06,
+      "loss": 0.8925,
+      "step": 10978
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7070635557174683,
+      "learning_rate": 6.371387903319268e-06,
+      "loss": 0.9196,
+      "step": 10979
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.756906509399414,
+      "learning_rate": 6.369656963251467e-06,
+      "loss": 1.0098,
+      "step": 10980
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.69398832321167,
+      "learning_rate": 6.367926148456663e-06,
+      "loss": 0.9388,
+      "step": 10981
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7069170475006104,
+      "learning_rate": 6.366195458994581e-06,
+      "loss": 1.0062,
+      "step": 10982
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7974534034729004,
+      "learning_rate": 6.3644648949249444e-06,
+      "loss": 0.9866,
+      "step": 10983
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.713573932647705,
+      "learning_rate": 6.362734456307469e-06,
+      "loss": 1.0169,
+      "step": 10984
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.866715431213379,
+      "learning_rate": 6.3610041432018675e-06,
+      "loss": 0.8631,
+      "step": 10985
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7902175188064575,
+      "learning_rate": 6.359273955667847e-06,
+      "loss": 0.8962,
+      "step": 10986
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.731526255607605,
+      "learning_rate": 6.3575438937651126e-06,
+      "loss": 0.9179,
+      "step": 10987
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8109718561172485,
+      "learning_rate": 6.355813957553364e-06,
+      "loss": 0.9641,
+      "step": 10988
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0731029510498047,
+      "learning_rate": 6.354084147092296e-06,
+      "loss": 0.6011,
+      "step": 10989
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8114094734191895,
+      "learning_rate": 6.352354462441599e-06,
+      "loss": 0.943,
+      "step": 10990
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0237623453140259,
+      "learning_rate": 6.350624903660961e-06,
+      "loss": 0.5365,
+      "step": 10991
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.690920352935791,
+      "learning_rate": 6.3488954708100635e-06,
+      "loss": 0.9045,
+      "step": 10992
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8189454078674316,
+      "learning_rate": 6.347166163948581e-06,
+      "loss": 0.896,
+      "step": 10993
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7014495134353638,
+      "learning_rate": 6.345436983136195e-06,
+      "loss": 0.954,
+      "step": 10994
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.827431321144104,
+      "learning_rate": 6.343707928432566e-06,
+      "loss": 0.8499,
+      "step": 10995
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.02494215965271,
+      "learning_rate": 6.3419789998973655e-06,
+      "loss": 0.9032,
+      "step": 10996
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9038704633712769,
+      "learning_rate": 6.340250197590252e-06,
+      "loss": 0.9574,
+      "step": 10997
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9325326681137085,
+      "learning_rate": 6.338521521570878e-06,
+      "loss": 0.9384,
+      "step": 10998
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8828848600387573,
+      "learning_rate": 6.336792971898902e-06,
+      "loss": 0.9209,
+      "step": 10999
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1130417585372925,
+      "learning_rate": 6.335064548633967e-06,
+      "loss": 0.5687,
+      "step": 11000
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.816176414489746,
+      "learning_rate": 6.333336251835715e-06,
+      "loss": 0.9401,
+      "step": 11001
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.798572063446045,
+      "learning_rate": 6.331608081563789e-06,
+      "loss": 0.9238,
+      "step": 11002
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5433313846588135,
+      "learning_rate": 6.3298800378778205e-06,
+      "loss": 0.9521,
+      "step": 11003
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7677944898605347,
+      "learning_rate": 6.328152120837438e-06,
+      "loss": 0.8213,
+      "step": 11004
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8854432106018066,
+      "learning_rate": 6.326424330502271e-06,
+      "loss": 0.9288,
+      "step": 11005
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8492082357406616,
+      "learning_rate": 6.324696666931938e-06,
+      "loss": 1.0141,
+      "step": 11006
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8113558292388916,
+      "learning_rate": 6.322969130186057e-06,
+      "loss": 0.9787,
+      "step": 11007
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7282055616378784,
+      "learning_rate": 6.3212417203242386e-06,
+      "loss": 0.9875,
+      "step": 11008
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9558500051498413,
+      "learning_rate": 6.319514437406092e-06,
+      "loss": 0.9387,
+      "step": 11009
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7584469318389893,
+      "learning_rate": 6.317787281491221e-06,
+      "loss": 0.9657,
+      "step": 11010
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7344539165496826,
+      "learning_rate": 6.316060252639226e-06,
+      "loss": 0.9068,
+      "step": 11011
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7521001100540161,
+      "learning_rate": 6.314333350909701e-06,
+      "loss": 0.9043,
+      "step": 11012
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7542681694030762,
+      "learning_rate": 6.312606576362237e-06,
+      "loss": 0.9057,
+      "step": 11013
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.663023829460144,
+      "learning_rate": 6.3108799290564195e-06,
+      "loss": 0.9462,
+      "step": 11014
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7715116739273071,
+      "learning_rate": 6.30915340905183e-06,
+      "loss": 0.8449,
+      "step": 11015
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9177113771438599,
+      "learning_rate": 6.307427016408048e-06,
+      "loss": 0.9331,
+      "step": 11016
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8698233366012573,
+      "learning_rate": 6.3057007511846425e-06,
+      "loss": 0.9832,
+      "step": 11017
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.887035608291626,
+      "learning_rate": 6.303974613441186e-06,
+      "loss": 0.9607,
+      "step": 11018
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9032979011535645,
+      "learning_rate": 6.30224860323724e-06,
+      "loss": 0.8288,
+      "step": 11019
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7999794483184814,
+      "learning_rate": 6.300522720632367e-06,
+      "loss": 0.9558,
+      "step": 11020
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.743431568145752,
+      "learning_rate": 6.29879696568612e-06,
+      "loss": 0.9257,
+      "step": 11021
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.869232177734375,
+      "learning_rate": 6.297071338458049e-06,
+      "loss": 0.9035,
+      "step": 11022
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8955055475234985,
+      "learning_rate": 6.295345839007705e-06,
+      "loss": 0.966,
+      "step": 11023
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8623287677764893,
+      "learning_rate": 6.293620467394626e-06,
+      "loss": 0.9151,
+      "step": 11024
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7950202226638794,
+      "learning_rate": 6.291895223678352e-06,
+      "loss": 1.0403,
+      "step": 11025
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7138208150863647,
+      "learning_rate": 6.290170107918416e-06,
+      "loss": 0.9177,
+      "step": 11026
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7432208061218262,
+      "learning_rate": 6.2884451201743465e-06,
+      "loss": 0.9432,
+      "step": 11027
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6791489124298096,
+      "learning_rate": 6.286720260505667e-06,
+      "loss": 0.9159,
+      "step": 11028
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7297539710998535,
+      "learning_rate": 6.2849955289719015e-06,
+      "loss": 0.9047,
+      "step": 11029
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9915313720703125,
+      "learning_rate": 6.283270925632561e-06,
+      "loss": 0.9114,
+      "step": 11030
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.877548336982727,
+      "learning_rate": 6.281546450547158e-06,
+      "loss": 1.0277,
+      "step": 11031
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6026054620742798,
+      "learning_rate": 6.279822103775202e-06,
+      "loss": 0.8753,
+      "step": 11032
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8172279596328735,
+      "learning_rate": 6.278097885376191e-06,
+      "loss": 0.941,
+      "step": 11033
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.755049467086792,
+      "learning_rate": 6.276373795409626e-06,
+      "loss": 0.9646,
+      "step": 11034
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6424330472946167,
+      "learning_rate": 6.274649833934998e-06,
+      "loss": 1.0022,
+      "step": 11035
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8425216674804688,
+      "learning_rate": 6.2729260010117995e-06,
+      "loss": 0.9213,
+      "step": 11036
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7436614036560059,
+      "learning_rate": 6.271202296699515e-06,
+      "loss": 0.8983,
+      "step": 11037
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8171215057373047,
+      "learning_rate": 6.269478721057621e-06,
+      "loss": 0.9871,
+      "step": 11038
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7208151817321777,
+      "learning_rate": 6.267755274145597e-06,
+      "loss": 0.8681,
+      "step": 11039
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.785952091217041,
+      "learning_rate": 6.266031956022913e-06,
+      "loss": 0.9223,
+      "step": 11040
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1018598079681396,
+      "learning_rate": 6.264308766749034e-06,
+      "loss": 0.9146,
+      "step": 11041
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9064139127731323,
+      "learning_rate": 6.262585706383426e-06,
+      "loss": 0.9593,
+      "step": 11042
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6764744520187378,
+      "learning_rate": 6.260862774985545e-06,
+      "loss": 0.9406,
+      "step": 11043
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9635815620422363,
+      "learning_rate": 6.259139972614845e-06,
+      "loss": 0.5402,
+      "step": 11044
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.744985580444336,
+      "learning_rate": 6.257417299330775e-06,
+      "loss": 0.9671,
+      "step": 11045
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.869409441947937,
+      "learning_rate": 6.2556947551927786e-06,
+      "loss": 0.935,
+      "step": 11046
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.692609429359436,
+      "learning_rate": 6.253972340260295e-06,
+      "loss": 0.9935,
+      "step": 11047
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7915390729904175,
+      "learning_rate": 6.2522500545927635e-06,
+      "loss": 0.9857,
+      "step": 11048
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8140815496444702,
+      "learning_rate": 6.2505278982496146e-06,
+      "loss": 0.9292,
+      "step": 11049
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.841794729232788,
+      "learning_rate": 6.248805871290274e-06,
+      "loss": 0.8641,
+      "step": 11050
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6097465753555298,
+      "learning_rate": 6.247083973774164e-06,
+      "loss": 0.9056,
+      "step": 11051
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6902005672454834,
+      "learning_rate": 6.245362205760703e-06,
+      "loss": 0.9938,
+      "step": 11052
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6900423765182495,
+      "learning_rate": 6.2436405673093035e-06,
+      "loss": 0.995,
+      "step": 11053
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7228493690490723,
+      "learning_rate": 6.2419190584793755e-06,
+      "loss": 0.9032,
+      "step": 11054
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.713371753692627,
+      "learning_rate": 6.240197679330324e-06,
+      "loss": 1.0024,
+      "step": 11055
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.67291259765625,
+      "learning_rate": 6.238476429921547e-06,
+      "loss": 0.9291,
+      "step": 11056
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9409734606742859,
+      "learning_rate": 6.236755310312441e-06,
+      "loss": 0.5372,
+      "step": 11057
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.72934889793396,
+      "learning_rate": 6.235034320562396e-06,
+      "loss": 0.9588,
+      "step": 11058
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.783906102180481,
+      "learning_rate": 6.2333134607308e-06,
+      "loss": 0.9355,
+      "step": 11059
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0282846689224243,
+      "learning_rate": 6.231592730877035e-06,
+      "loss": 0.5902,
+      "step": 11060
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7783414125442505,
+      "learning_rate": 6.229872131060477e-06,
+      "loss": 0.8947,
+      "step": 11061
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.770006537437439,
+      "learning_rate": 6.228151661340503e-06,
+      "loss": 0.9197,
+      "step": 11062
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6996607780456543,
+      "learning_rate": 6.226431321776476e-06,
+      "loss": 0.8681,
+      "step": 11063
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5794992446899414,
+      "learning_rate": 6.224711112427764e-06,
+      "loss": 0.9598,
+      "step": 11064
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8020678758621216,
+      "learning_rate": 6.2229910333537256e-06,
+      "loss": 0.9207,
+      "step": 11065
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8647137880325317,
+      "learning_rate": 6.221271084613718e-06,
+      "loss": 1.0036,
+      "step": 11066
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6966720819473267,
+      "learning_rate": 6.219551266267088e-06,
+      "loss": 0.9952,
+      "step": 11067
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.490164875984192,
+      "learning_rate": 6.217831578373185e-06,
+      "loss": 0.8968,
+      "step": 11068
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6140415668487549,
+      "learning_rate": 6.2161120209913475e-06,
+      "loss": 0.9231,
+      "step": 11069
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9712891578674316,
+      "learning_rate": 6.214392594180915e-06,
+      "loss": 0.9862,
+      "step": 11070
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9715535640716553,
+      "learning_rate": 6.212673298001221e-06,
+      "loss": 0.937,
+      "step": 11071
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.711273431777954,
+      "learning_rate": 6.2109541325115905e-06,
+      "loss": 0.9432,
+      "step": 11072
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7534865140914917,
+      "learning_rate": 6.209235097771349e-06,
+      "loss": 0.9281,
+      "step": 11073
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7710973024368286,
+      "learning_rate": 6.207516193839815e-06,
+      "loss": 0.8858,
+      "step": 11074
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2353761196136475,
+      "learning_rate": 6.205797420776303e-06,
+      "loss": 0.5681,
+      "step": 11075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7199963331222534,
+      "learning_rate": 6.204078778640121e-06,
+      "loss": 0.9494,
+      "step": 11076
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8881731033325195,
+      "learning_rate": 6.2023602674905795e-06,
+      "loss": 0.8742,
+      "step": 11077
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.774097204208374,
+      "learning_rate": 6.2006418873869776e-06,
+      "loss": 0.936,
+      "step": 11078
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7970454692840576,
+      "learning_rate": 6.19892363838861e-06,
+      "loss": 0.8906,
+      "step": 11079
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.851100206375122,
+      "learning_rate": 6.1972055205547696e-06,
+      "loss": 0.9278,
+      "step": 11080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7701952457427979,
+      "learning_rate": 6.195487533944745e-06,
+      "loss": 0.8708,
+      "step": 11081
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9664692878723145,
+      "learning_rate": 6.1937696786178184e-06,
+      "loss": 0.9535,
+      "step": 11082
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7868226766586304,
+      "learning_rate": 6.192051954633267e-06,
+      "loss": 0.9718,
+      "step": 11083
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.709617257118225,
+      "learning_rate": 6.190334362050365e-06,
+      "loss": 0.8835,
+      "step": 11084
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8134889602661133,
+      "learning_rate": 6.188616900928384e-06,
+      "loss": 0.9123,
+      "step": 11085
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1057851314544678,
+      "learning_rate": 6.186899571326586e-06,
+      "loss": 1.0112,
+      "step": 11086
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7677581310272217,
+      "learning_rate": 6.185182373304233e-06,
+      "loss": 0.9094,
+      "step": 11087
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.849670648574829,
+      "learning_rate": 6.183465306920578e-06,
+      "loss": 0.9055,
+      "step": 11088
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5702879428863525,
+      "learning_rate": 6.181748372234875e-06,
+      "loss": 0.8569,
+      "step": 11089
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9833319187164307,
+      "learning_rate": 6.180031569306371e-06,
+      "loss": 1.0073,
+      "step": 11090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8712981939315796,
+      "learning_rate": 6.178314898194305e-06,
+      "loss": 0.9958,
+      "step": 11091
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6491199731826782,
+      "learning_rate": 6.1765983589579185e-06,
+      "loss": 0.8409,
+      "step": 11092
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8637315034866333,
+      "learning_rate": 6.1748819516564414e-06,
+      "loss": 0.9626,
+      "step": 11093
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.911102056503296,
+      "learning_rate": 6.173165676349103e-06,
+      "loss": 0.8994,
+      "step": 11094
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8287254571914673,
+      "learning_rate": 6.1714495330951285e-06,
+      "loss": 0.934,
+      "step": 11095
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.856987714767456,
+      "learning_rate": 6.169733521953735e-06,
+      "loss": 1.0073,
+      "step": 11096
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.696885585784912,
+      "learning_rate": 6.168017642984139e-06,
+      "loss": 0.9138,
+      "step": 11097
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.845579743385315,
+      "learning_rate": 6.166301896245549e-06,
+      "loss": 0.8685,
+      "step": 11098
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6699053049087524,
+      "learning_rate": 6.164586281797171e-06,
+      "loss": 0.945,
+      "step": 11099
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6427404880523682,
+      "learning_rate": 6.162870799698209e-06,
+      "loss": 0.881,
+      "step": 11100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8081845045089722,
+      "learning_rate": 6.161155450007853e-06,
+      "loss": 0.9336,
+      "step": 11101
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.63100004196167,
+      "learning_rate": 6.159440232785301e-06,
+      "loss": 0.8434,
+      "step": 11102
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.61729896068573,
+      "learning_rate": 6.1577251480897394e-06,
+      "loss": 1.0031,
+      "step": 11103
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6938660144805908,
+      "learning_rate": 6.15601019598035e-06,
+      "loss": 0.8987,
+      "step": 11104
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8394631147384644,
+      "learning_rate": 6.1542953765163105e-06,
+      "loss": 0.897,
+      "step": 11105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7753878831863403,
+      "learning_rate": 6.152580689756795e-06,
+      "loss": 0.9828,
+      "step": 11106
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.822212815284729,
+      "learning_rate": 6.150866135760973e-06,
+      "loss": 0.9443,
+      "step": 11107
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7607523202896118,
+      "learning_rate": 6.149151714588009e-06,
+      "loss": 0.9835,
+      "step": 11108
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9261929988861084,
+      "learning_rate": 6.14743742629706e-06,
+      "loss": 0.9983,
+      "step": 11109
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.727971076965332,
+      "learning_rate": 6.1457232709472854e-06,
+      "loss": 0.889,
+      "step": 11110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9376394748687744,
+      "learning_rate": 6.1440092485978355e-06,
+      "loss": 0.9913,
+      "step": 11111
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7738347053527832,
+      "learning_rate": 6.1422953593078535e-06,
+      "loss": 0.9228,
+      "step": 11112
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7786391973495483,
+      "learning_rate": 6.140581603136482e-06,
+      "loss": 0.943,
+      "step": 11113
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.676505446434021,
+      "learning_rate": 6.138867980142859e-06,
+      "loss": 0.9668,
+      "step": 11114
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.002361536026001,
+      "learning_rate": 6.137154490386117e-06,
+      "loss": 0.9506,
+      "step": 11115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7543835639953613,
+      "learning_rate": 6.135441133925382e-06,
+      "loss": 0.9386,
+      "step": 11116
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8200215101242065,
+      "learning_rate": 6.13372791081978e-06,
+      "loss": 0.9455,
+      "step": 11117
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7570029497146606,
+      "learning_rate": 6.132014821128427e-06,
+      "loss": 1.0108,
+      "step": 11118
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.802232027053833,
+      "learning_rate": 6.130301864910437e-06,
+      "loss": 0.9386,
+      "step": 11119
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6663905382156372,
+      "learning_rate": 6.128589042224922e-06,
+      "loss": 0.9714,
+      "step": 11120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7160100936889648,
+      "learning_rate": 6.126876353130984e-06,
+      "loss": 0.8989,
+      "step": 11121
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9042409658432007,
+      "learning_rate": 6.125163797687723e-06,
+      "loss": 0.9002,
+      "step": 11122
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8753575086593628,
+      "learning_rate": 6.123451375954235e-06,
+      "loss": 0.9573,
+      "step": 11123
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7204365730285645,
+      "learning_rate": 6.121739087989613e-06,
+      "loss": 0.931,
+      "step": 11124
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9006472826004028,
+      "learning_rate": 6.120026933852939e-06,
+      "loss": 0.9898,
+      "step": 11125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.850385069847107,
+      "learning_rate": 6.118314913603299e-06,
+      "loss": 1.0244,
+      "step": 11126
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7779486179351807,
+      "learning_rate": 6.116603027299769e-06,
+      "loss": 0.8601,
+      "step": 11127
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.093890428543091,
+      "learning_rate": 6.114891275001417e-06,
+      "loss": 0.9856,
+      "step": 11128
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6598483324050903,
+      "learning_rate": 6.113179656767319e-06,
+      "loss": 0.8929,
+      "step": 11129
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.772676706314087,
+      "learning_rate": 6.111468172656529e-06,
+      "loss": 0.8801,
+      "step": 11130
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7685588598251343,
+      "learning_rate": 6.109756822728114e-06,
+      "loss": 1.0091,
+      "step": 11131
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6190587282180786,
+      "learning_rate": 6.108045607041124e-06,
+      "loss": 0.8415,
+      "step": 11132
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8606104850769043,
+      "learning_rate": 6.106334525654608e-06,
+      "loss": 0.9118,
+      "step": 11133
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.846361517906189,
+      "learning_rate": 6.1046235786276105e-06,
+      "loss": 1.0228,
+      "step": 11134
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0187033414840698,
+      "learning_rate": 6.102912766019173e-06,
+      "loss": 0.5528,
+      "step": 11135
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6623220443725586,
+      "learning_rate": 6.101202087888329e-06,
+      "loss": 0.9492,
+      "step": 11136
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8402243852615356,
+      "learning_rate": 6.099491544294111e-06,
+      "loss": 0.9341,
+      "step": 11137
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6684446334838867,
+      "learning_rate": 6.097781135295543e-06,
+      "loss": 0.9616,
+      "step": 11138
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.869888424873352,
+      "learning_rate": 6.096070860951648e-06,
+      "loss": 0.8574,
+      "step": 11139
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.659745454788208,
+      "learning_rate": 6.094360721321443e-06,
+      "loss": 1.0251,
+      "step": 11140
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5677752494812012,
+      "learning_rate": 6.092650716463939e-06,
+      "loss": 0.8079,
+      "step": 11141
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7860305309295654,
+      "learning_rate": 6.090940846438143e-06,
+      "loss": 0.9112,
+      "step": 11142
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7747595310211182,
+      "learning_rate": 6.08923111130306e-06,
+      "loss": 0.8739,
+      "step": 11143
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6891639232635498,
+      "learning_rate": 6.087521511117686e-06,
+      "loss": 0.8964,
+      "step": 11144
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9224579334259033,
+      "learning_rate": 6.085812045941018e-06,
+      "loss": 0.9193,
+      "step": 11145
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7775453329086304,
+      "learning_rate": 6.084102715832041e-06,
+      "loss": 0.9303,
+      "step": 11146
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6949049234390259,
+      "learning_rate": 6.08239352084974e-06,
+      "loss": 0.8649,
+      "step": 11147
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6505948305130005,
+      "learning_rate": 6.080684461053096e-06,
+      "loss": 1.0045,
+      "step": 11148
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5866869688034058,
+      "learning_rate": 6.0789755365010834e-06,
+      "loss": 0.8772,
+      "step": 11149
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8161391019821167,
+      "learning_rate": 6.077266747252672e-06,
+      "loss": 0.9599,
+      "step": 11150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.048730731010437,
+      "learning_rate": 6.0755580933668265e-06,
+      "loss": 0.5741,
+      "step": 11151
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7437841892242432,
+      "learning_rate": 6.073849574902509e-06,
+      "loss": 0.9133,
+      "step": 11152
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.804383635520935,
+      "learning_rate": 6.0721411919186766e-06,
+      "loss": 0.918,
+      "step": 11153
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7442551851272583,
+      "learning_rate": 6.070432944474276e-06,
+      "loss": 0.9426,
+      "step": 11154
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6498570442199707,
+      "learning_rate": 6.068724832628261e-06,
+      "loss": 0.8492,
+      "step": 11155
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7418148517608643,
+      "learning_rate": 6.06701685643957e-06,
+      "loss": 0.871,
+      "step": 11156
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6323784589767456,
+      "learning_rate": 6.065309015967141e-06,
+      "loss": 0.9127,
+      "step": 11157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6228750944137573,
+      "learning_rate": 6.063601311269906e-06,
+      "loss": 0.9608,
+      "step": 11158
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.959595799446106,
+      "learning_rate": 6.061893742406795e-06,
+      "loss": 0.911,
+      "step": 11159
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.590744972229004,
+      "learning_rate": 6.06018630943673e-06,
+      "loss": 0.965,
+      "step": 11160
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7135217189788818,
+      "learning_rate": 6.05847901241863e-06,
+      "loss": 0.8576,
+      "step": 11161
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6816157102584839,
+      "learning_rate": 6.05677185141141e-06,
+      "loss": 0.8943,
+      "step": 11162
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.673656702041626,
+      "learning_rate": 6.0550648264739776e-06,
+      "loss": 0.9347,
+      "step": 11163
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.862610936164856,
+      "learning_rate": 6.053357937665237e-06,
+      "loss": 0.9805,
+      "step": 11164
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6045889854431152,
+      "learning_rate": 6.051651185044091e-06,
+      "loss": 0.8789,
+      "step": 11165
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.777668833732605,
+      "learning_rate": 6.049944568669432e-06,
+      "loss": 0.9633,
+      "step": 11166
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8737283945083618,
+      "learning_rate": 6.048238088600151e-06,
+      "loss": 0.9442,
+      "step": 11167
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9231356382369995,
+      "learning_rate": 6.046531744895136e-06,
+      "loss": 0.9518,
+      "step": 11168
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.009995698928833,
+      "learning_rate": 6.044825537613268e-06,
+      "loss": 0.9372,
+      "step": 11169
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8889410495758057,
+      "learning_rate": 6.0431194668134226e-06,
+      "loss": 0.9116,
+      "step": 11170
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0155773162841797,
+      "learning_rate": 6.04141353255447e-06,
+      "loss": 0.9155,
+      "step": 11171
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9353569746017456,
+      "learning_rate": 6.039707734895279e-06,
+      "loss": 0.9612,
+      "step": 11172
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7433983087539673,
+      "learning_rate": 6.038002073894712e-06,
+      "loss": 0.9868,
+      "step": 11173
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7999604940414429,
+      "learning_rate": 6.036296549611627e-06,
+      "loss": 0.9284,
+      "step": 11174
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7205383777618408,
+      "learning_rate": 6.034591162104873e-06,
+      "loss": 0.9225,
+      "step": 11175
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6665748357772827,
+      "learning_rate": 6.032885911433303e-06,
+      "loss": 0.9616,
+      "step": 11176
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7283823490142822,
+      "learning_rate": 6.031180797655758e-06,
+      "loss": 0.94,
+      "step": 11177
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.857397198677063,
+      "learning_rate": 6.029475820831077e-06,
+      "loss": 0.8904,
+      "step": 11178
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7711527347564697,
+      "learning_rate": 6.027770981018093e-06,
+      "loss": 0.9774,
+      "step": 11179
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0171586275100708,
+      "learning_rate": 6.026066278275638e-06,
+      "loss": 0.5762,
+      "step": 11180
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5514590740203857,
+      "learning_rate": 6.024361712662534e-06,
+      "loss": 0.8537,
+      "step": 11181
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6581367254257202,
+      "learning_rate": 6.022657284237603e-06,
+      "loss": 0.872,
+      "step": 11182
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8713977336883545,
+      "learning_rate": 6.020952993059659e-06,
+      "loss": 0.8437,
+      "step": 11183
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0557243824005127,
+      "learning_rate": 6.0192488391875125e-06,
+      "loss": 0.9067,
+      "step": 11184
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.955741047859192,
+      "learning_rate": 6.017544822679968e-06,
+      "loss": 0.9168,
+      "step": 11185
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8321826457977295,
+      "learning_rate": 6.015840943595828e-06,
+      "loss": 0.9073,
+      "step": 11186
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8450034856796265,
+      "learning_rate": 6.014137201993886e-06,
+      "loss": 0.9441,
+      "step": 11187
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.584324598312378,
+      "learning_rate": 6.0124335979329365e-06,
+      "loss": 0.8551,
+      "step": 11188
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6888806819915771,
+      "learning_rate": 6.0107301314717635e-06,
+      "loss": 0.9328,
+      "step": 11189
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.745956540107727,
+      "learning_rate": 6.009026802669151e-06,
+      "loss": 0.8763,
+      "step": 11190
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7399253845214844,
+      "learning_rate": 6.007323611583873e-06,
+      "loss": 0.9309,
+      "step": 11191
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.072075128555298,
+      "learning_rate": 6.005620558274707e-06,
+      "loss": 0.8344,
+      "step": 11192
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.795037865638733,
+      "learning_rate": 6.003917642800416e-06,
+      "loss": 0.8578,
+      "step": 11193
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6517858505249023,
+      "learning_rate": 6.002214865219764e-06,
+      "loss": 0.9073,
+      "step": 11194
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7928987741470337,
+      "learning_rate": 6.00051222559151e-06,
+      "loss": 0.9194,
+      "step": 11195
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.718496561050415,
+      "learning_rate": 5.998809723974407e-06,
+      "loss": 0.9122,
+      "step": 11196
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8300230503082275,
+      "learning_rate": 5.997107360427205e-06,
+      "loss": 0.9925,
+      "step": 11197
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6812857389450073,
+      "learning_rate": 5.995405135008645e-06,
+      "loss": 0.8515,
+      "step": 11198
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.677461862564087,
+      "learning_rate": 5.993703047777468e-06,
+      "loss": 0.8564,
+      "step": 11199
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7871259450912476,
+      "learning_rate": 5.9920010987924086e-06,
+      "loss": 0.9006,
+      "step": 11200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9717799425125122,
+      "learning_rate": 5.9902992881121955e-06,
+      "loss": 1.0088,
+      "step": 11201
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0505146980285645,
+      "learning_rate": 5.988597615795553e-06,
+      "loss": 0.9313,
+      "step": 11202
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7951393127441406,
+      "learning_rate": 5.9868960819012e-06,
+      "loss": 1.0173,
+      "step": 11203
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9893182516098022,
+      "learning_rate": 5.985194686487854e-06,
+      "loss": 0.5617,
+      "step": 11204
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9469119310379028,
+      "learning_rate": 5.983493429614224e-06,
+      "loss": 0.978,
+      "step": 11205
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8285976648330688,
+      "learning_rate": 5.981792311339017e-06,
+      "loss": 0.9528,
+      "step": 11206
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7404613494873047,
+      "learning_rate": 5.980091331720933e-06,
+      "loss": 0.9591,
+      "step": 11207
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.718595266342163,
+      "learning_rate": 5.978390490818665e-06,
+      "loss": 0.9273,
+      "step": 11208
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6811379194259644,
+      "learning_rate": 5.97668978869091e-06,
+      "loss": 0.9081,
+      "step": 11209
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9186893701553345,
+      "learning_rate": 5.974989225396352e-06,
+      "loss": 0.9442,
+      "step": 11210
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7086790800094604,
+      "learning_rate": 5.973288800993672e-06,
+      "loss": 0.8348,
+      "step": 11211
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8945207595825195,
+      "learning_rate": 5.971588515541547e-06,
+      "loss": 0.9188,
+      "step": 11212
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.758755087852478,
+      "learning_rate": 5.969888369098649e-06,
+      "loss": 0.9017,
+      "step": 11213
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7454659938812256,
+      "learning_rate": 5.968188361723647e-06,
+      "loss": 0.9104,
+      "step": 11214
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8722522258758545,
+      "learning_rate": 5.9664884934752025e-06,
+      "loss": 0.9755,
+      "step": 11215
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.843132734298706,
+      "learning_rate": 5.964788764411971e-06,
+      "loss": 0.9215,
+      "step": 11216
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9044537544250488,
+      "learning_rate": 5.963089174592609e-06,
+      "loss": 0.9293,
+      "step": 11217
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5448360443115234,
+      "learning_rate": 5.961389724075761e-06,
+      "loss": 0.9017,
+      "step": 11218
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8351337909698486,
+      "learning_rate": 5.959690412920074e-06,
+      "loss": 0.9585,
+      "step": 11219
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7059670686721802,
+      "learning_rate": 5.957991241184184e-06,
+      "loss": 0.9345,
+      "step": 11220
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7572273015975952,
+      "learning_rate": 5.956292208926724e-06,
+      "loss": 0.9058,
+      "step": 11221
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.060110092163086,
+      "learning_rate": 5.954593316206325e-06,
+      "loss": 0.5581,
+      "step": 11222
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.689358115196228,
+      "learning_rate": 5.952894563081612e-06,
+      "loss": 0.9642,
+      "step": 11223
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.02638578414917,
+      "learning_rate": 5.9511959496112015e-06,
+      "loss": 0.9481,
+      "step": 11224
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7780299186706543,
+      "learning_rate": 5.949497475853709e-06,
+      "loss": 0.9089,
+      "step": 11225
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.793791651725769,
+      "learning_rate": 5.947799141867744e-06,
+      "loss": 0.9441,
+      "step": 11226
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.588212013244629,
+      "learning_rate": 5.94610094771191e-06,
+      "loss": 0.8989,
+      "step": 11227
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7925142049789429,
+      "learning_rate": 5.9444028934448105e-06,
+      "loss": 0.985,
+      "step": 11228
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7412967681884766,
+      "learning_rate": 5.942704979125037e-06,
+      "loss": 0.9236,
+      "step": 11229
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7654411792755127,
+      "learning_rate": 5.941007204811181e-06,
+      "loss": 0.9671,
+      "step": 11230
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.689968228340149,
+      "learning_rate": 5.939309570561828e-06,
+      "loss": 0.8924,
+      "step": 11231
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7111839056015015,
+      "learning_rate": 5.9376120764355595e-06,
+      "loss": 0.9579,
+      "step": 11232
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9688478708267212,
+      "learning_rate": 5.935914722490947e-06,
+      "loss": 0.9566,
+      "step": 11233
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7676457166671753,
+      "learning_rate": 5.934217508786569e-06,
+      "loss": 0.9576,
+      "step": 11234
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.965472936630249,
+      "learning_rate": 5.932520435380986e-06,
+      "loss": 1.0277,
+      "step": 11235
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7442251443862915,
+      "learning_rate": 5.930823502332761e-06,
+      "loss": 0.9822,
+      "step": 11236
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8437994718551636,
+      "learning_rate": 5.92912670970045e-06,
+      "loss": 0.9122,
+      "step": 11237
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6816394329071045,
+      "learning_rate": 5.9274300575426045e-06,
+      "loss": 0.9191,
+      "step": 11238
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7000749111175537,
+      "learning_rate": 5.925733545917771e-06,
+      "loss": 0.842,
+      "step": 11239
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6760202646255493,
+      "learning_rate": 5.924037174884494e-06,
+      "loss": 0.878,
+      "step": 11240
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.844265341758728,
+      "learning_rate": 5.922340944501306e-06,
+      "loss": 0.9125,
+      "step": 11241
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7644356489181519,
+      "learning_rate": 5.920644854826742e-06,
+      "loss": 0.8731,
+      "step": 11242
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7012150287628174,
+      "learning_rate": 5.918948905919331e-06,
+      "loss": 0.9895,
+      "step": 11243
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0596972703933716,
+      "learning_rate": 5.91725309783759e-06,
+      "loss": 0.6,
+      "step": 11244
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8861899375915527,
+      "learning_rate": 5.91555743064004e-06,
+      "loss": 0.9187,
+      "step": 11245
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0559172630310059,
+      "learning_rate": 5.913861904385194e-06,
+      "loss": 0.5721,
+      "step": 11246
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.706436038017273,
+      "learning_rate": 5.912166519131561e-06,
+      "loss": 0.8855,
+      "step": 11247
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.769168734550476,
+      "learning_rate": 5.910471274937643e-06,
+      "loss": 1.0049,
+      "step": 11248
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6045399904251099,
+      "learning_rate": 5.908776171861937e-06,
+      "loss": 0.9566,
+      "step": 11249
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6776833534240723,
+      "learning_rate": 5.907081209962937e-06,
+      "loss": 0.9508,
+      "step": 11250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7636966705322266,
+      "learning_rate": 5.9053863892991304e-06,
+      "loss": 0.931,
+      "step": 11251
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9879350066184998,
+      "learning_rate": 5.903691709929002e-06,
+      "loss": 0.5479,
+      "step": 11252
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6759867668151855,
+      "learning_rate": 5.901997171911032e-06,
+      "loss": 0.9567,
+      "step": 11253
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9727058410644531,
+      "learning_rate": 5.90030277530369e-06,
+      "loss": 0.8964,
+      "step": 11254
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6429152488708496,
+      "learning_rate": 5.898608520165448e-06,
+      "loss": 0.9155,
+      "step": 11255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7586252689361572,
+      "learning_rate": 5.896914406554768e-06,
+      "loss": 1.0172,
+      "step": 11256
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.696723222732544,
+      "learning_rate": 5.89522043453011e-06,
+      "loss": 0.9038,
+      "step": 11257
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.055777668952942,
+      "learning_rate": 5.893526604149931e-06,
+      "loss": 0.5995,
+      "step": 11258
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6775919198989868,
+      "learning_rate": 5.891832915472676e-06,
+      "loss": 0.9389,
+      "step": 11259
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6754018068313599,
+      "learning_rate": 5.8901393685567906e-06,
+      "loss": 0.8429,
+      "step": 11260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7547240257263184,
+      "learning_rate": 5.888445963460716e-06,
+      "loss": 1.0018,
+      "step": 11261
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7485063076019287,
+      "learning_rate": 5.886752700242886e-06,
+      "loss": 0.9531,
+      "step": 11262
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6518371105194092,
+      "learning_rate": 5.885059578961732e-06,
+      "loss": 0.9516,
+      "step": 11263
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7253373861312866,
+      "learning_rate": 5.883366599675675e-06,
+      "loss": 0.9161,
+      "step": 11264
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7492817640304565,
+      "learning_rate": 5.881673762443138e-06,
+      "loss": 0.8991,
+      "step": 11265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7205696105957031,
+      "learning_rate": 5.879981067322538e-06,
+      "loss": 1.0574,
+      "step": 11266
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.75700044631958,
+      "learning_rate": 5.878288514372281e-06,
+      "loss": 0.9142,
+      "step": 11267
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5620605945587158,
+      "learning_rate": 5.8765961036507734e-06,
+      "loss": 0.8509,
+      "step": 11268
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6875847578048706,
+      "learning_rate": 5.874903835216417e-06,
+      "loss": 0.9441,
+      "step": 11269
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.839728593826294,
+      "learning_rate": 5.873211709127604e-06,
+      "loss": 0.9261,
+      "step": 11270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8331587314605713,
+      "learning_rate": 5.871519725442729e-06,
+      "loss": 0.9048,
+      "step": 11271
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7244642972946167,
+      "learning_rate": 5.869827884220176e-06,
+      "loss": 0.9746,
+      "step": 11272
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7994509935379028,
+      "learning_rate": 5.868136185518325e-06,
+      "loss": 0.9162,
+      "step": 11273
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6378540992736816,
+      "learning_rate": 5.866444629395551e-06,
+      "loss": 0.8494,
+      "step": 11274
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6356014013290405,
+      "learning_rate": 5.864753215910227e-06,
+      "loss": 0.9277,
+      "step": 11275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8897268772125244,
+      "learning_rate": 5.863061945120719e-06,
+      "loss": 0.9266,
+      "step": 11276
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7455523014068604,
+      "learning_rate": 5.8613708170853875e-06,
+      "loss": 0.8983,
+      "step": 11277
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8324429988861084,
+      "learning_rate": 5.859679831862588e-06,
+      "loss": 0.9782,
+      "step": 11278
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7036374807357788,
+      "learning_rate": 5.857988989510672e-06,
+      "loss": 0.9329,
+      "step": 11279
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8639886379241943,
+      "learning_rate": 5.856298290087985e-06,
+      "loss": 0.9633,
+      "step": 11280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6167266368865967,
+      "learning_rate": 5.854607733652871e-06,
+      "loss": 0.9092,
+      "step": 11281
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9042736291885376,
+      "learning_rate": 5.852917320263662e-06,
+      "loss": 0.9026,
+      "step": 11282
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7621804475784302,
+      "learning_rate": 5.8512270499786925e-06,
+      "loss": 0.9477,
+      "step": 11283
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0519490242004395,
+      "learning_rate": 5.849536922856289e-06,
+      "loss": 0.5061,
+      "step": 11284
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7745466232299805,
+      "learning_rate": 5.847846938954773e-06,
+      "loss": 0.9707,
+      "step": 11285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7042334079742432,
+      "learning_rate": 5.846157098332459e-06,
+      "loss": 1.0151,
+      "step": 11286
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7382415533065796,
+      "learning_rate": 5.8444674010476595e-06,
+      "loss": 0.9163,
+      "step": 11287
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9140769243240356,
+      "learning_rate": 5.842777847158682e-06,
+      "loss": 0.9885,
+      "step": 11288
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7565152645111084,
+      "learning_rate": 5.841088436723832e-06,
+      "loss": 0.9545,
+      "step": 11289
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0483756065368652,
+      "learning_rate": 5.839399169801399e-06,
+      "loss": 0.5986,
+      "step": 11290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8989804983139038,
+      "learning_rate": 5.837710046449681e-06,
+      "loss": 0.8871,
+      "step": 11291
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8919768333435059,
+      "learning_rate": 5.836021066726962e-06,
+      "loss": 0.9383,
+      "step": 11292
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6916006803512573,
+      "learning_rate": 5.8343322306915215e-06,
+      "loss": 0.9034,
+      "step": 11293
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6258602142333984,
+      "learning_rate": 5.832643538401641e-06,
+      "loss": 0.8699,
+      "step": 11294
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8095002174377441,
+      "learning_rate": 5.830954989915593e-06,
+      "loss": 0.9066,
+      "step": 11295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7301563024520874,
+      "learning_rate": 5.82926658529164e-06,
+      "loss": 0.9352,
+      "step": 11296
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8340225219726562,
+      "learning_rate": 5.827578324588049e-06,
+      "loss": 0.8658,
+      "step": 11297
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8023529052734375,
+      "learning_rate": 5.825890207863072e-06,
+      "loss": 0.946,
+      "step": 11298
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1031925678253174,
+      "learning_rate": 5.824202235174967e-06,
+      "loss": 0.5817,
+      "step": 11299
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8104435205459595,
+      "learning_rate": 5.822514406581975e-06,
+      "loss": 0.9293,
+      "step": 11300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8191719055175781,
+      "learning_rate": 5.820826722142345e-06,
+      "loss": 0.8759,
+      "step": 11301
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9445056915283203,
+      "learning_rate": 5.819139181914307e-06,
+      "loss": 0.9718,
+      "step": 11302
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6088212728500366,
+      "learning_rate": 5.817451785956101e-06,
+      "loss": 0.8821,
+      "step": 11303
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.774735689163208,
+      "learning_rate": 5.815764534325947e-06,
+      "loss": 0.8943,
+      "step": 11304
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.787113070487976,
+      "learning_rate": 5.81407742708207e-06,
+      "loss": 0.9508,
+      "step": 11305
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6502835750579834,
+      "learning_rate": 5.812390464282694e-06,
+      "loss": 0.9513,
+      "step": 11306
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7536396980285645,
+      "learning_rate": 5.810703645986018e-06,
+      "loss": 0.9259,
+      "step": 11307
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.833302617073059,
+      "learning_rate": 5.8090169722502634e-06,
+      "loss": 0.9006,
+      "step": 11308
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7453045845031738,
+      "learning_rate": 5.807330443133621e-06,
+      "loss": 0.9877,
+      "step": 11309
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8242830038070679,
+      "learning_rate": 5.805644058694297e-06,
+      "loss": 0.9277,
+      "step": 11310
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7830101251602173,
+      "learning_rate": 5.803957818990478e-06,
+      "loss": 0.9388,
+      "step": 11311
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7013764381408691,
+      "learning_rate": 5.802271724080355e-06,
+      "loss": 1.0082,
+      "step": 11312
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6380224227905273,
+      "learning_rate": 5.800585774022107e-06,
+      "loss": 0.8738,
+      "step": 11313
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.618857979774475,
+      "learning_rate": 5.7988999688739165e-06,
+      "loss": 0.9265,
+      "step": 11314
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.052559733390808,
+      "learning_rate": 5.797214308693948e-06,
+      "loss": 0.5528,
+      "step": 11315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8657804727554321,
+      "learning_rate": 5.7955287935403795e-06,
+      "loss": 0.9717,
+      "step": 11316
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9440536499023438,
+      "learning_rate": 5.793843423471361e-06,
+      "loss": 0.9188,
+      "step": 11317
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7049248218536377,
+      "learning_rate": 5.792158198545059e-06,
+      "loss": 0.8873,
+      "step": 11318
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.775940179824829,
+      "learning_rate": 5.790473118819626e-06,
+      "loss": 0.8851,
+      "step": 11319
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7453688383102417,
+      "learning_rate": 5.788788184353203e-06,
+      "loss": 0.9521,
+      "step": 11320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.721278429031372,
+      "learning_rate": 5.7871033952039416e-06,
+      "loss": 1.0033,
+      "step": 11321
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8560482263565063,
+      "learning_rate": 5.785418751429968e-06,
+      "loss": 0.9488,
+      "step": 11322
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8506866693496704,
+      "learning_rate": 5.783734253089426e-06,
+      "loss": 1.0243,
+      "step": 11323
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7493692636489868,
+      "learning_rate": 5.782049900240432e-06,
+      "loss": 0.8892,
+      "step": 11324
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.66787588596344,
+      "learning_rate": 5.780365692941118e-06,
+      "loss": 0.8545,
+      "step": 11325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.833383321762085,
+      "learning_rate": 5.778681631249588e-06,
+      "loss": 0.8914,
+      "step": 11326
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8372529745101929,
+      "learning_rate": 5.776997715223972e-06,
+      "loss": 0.9183,
+      "step": 11327
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8468822240829468,
+      "learning_rate": 5.775313944922365e-06,
+      "loss": 1.0248,
+      "step": 11328
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1548287868499756,
+      "learning_rate": 5.773630320402875e-06,
+      "loss": 0.9246,
+      "step": 11329
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.88059663772583,
+      "learning_rate": 5.771946841723594e-06,
+      "loss": 0.9688,
+      "step": 11330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.030954599380493,
+      "learning_rate": 5.77026350894262e-06,
+      "loss": 0.9839,
+      "step": 11331
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9893274903297424,
+      "learning_rate": 5.768580322118034e-06,
+      "loss": 0.5852,
+      "step": 11332
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6069986820220947,
+      "learning_rate": 5.766897281307924e-06,
+      "loss": 1.0112,
+      "step": 11333
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8937959671020508,
+      "learning_rate": 5.765214386570361e-06,
+      "loss": 0.9403,
+      "step": 11334
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.613581895828247,
+      "learning_rate": 5.76353163796342e-06,
+      "loss": 0.8648,
+      "step": 11335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9452673196792603,
+      "learning_rate": 5.761849035545171e-06,
+      "loss": 0.5137,
+      "step": 11336
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7590184211730957,
+      "learning_rate": 5.760166579373671e-06,
+      "loss": 0.8906,
+      "step": 11337
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6628491878509521,
+      "learning_rate": 5.758484269506981e-06,
+      "loss": 0.9672,
+      "step": 11338
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7405368089675903,
+      "learning_rate": 5.756802106003148e-06,
+      "loss": 0.938,
+      "step": 11339
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6419308185577393,
+      "learning_rate": 5.755120088920225e-06,
+      "loss": 0.9541,
+      "step": 11340
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.029478073120117,
+      "learning_rate": 5.753438218316245e-06,
+      "loss": 0.9222,
+      "step": 11341
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9166191816329956,
+      "learning_rate": 5.751756494249255e-06,
+      "loss": 0.8825,
+      "step": 11342
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.964041829109192,
+      "learning_rate": 5.7500749167772775e-06,
+      "loss": 0.954,
+      "step": 11343
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.697587013244629,
+      "learning_rate": 5.7483934859583465e-06,
+      "loss": 0.8927,
+      "step": 11344
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.761136531829834,
+      "learning_rate": 5.746712201850476e-06,
+      "loss": 0.9865,
+      "step": 11345
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7736387252807617,
+      "learning_rate": 5.74503106451169e-06,
+      "loss": 0.9655,
+      "step": 11346
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6465517282485962,
+      "learning_rate": 5.743350073999994e-06,
+      "loss": 0.8737,
+      "step": 11347
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7512249946594238,
+      "learning_rate": 5.741669230373394e-06,
+      "loss": 0.94,
+      "step": 11348
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8223719596862793,
+      "learning_rate": 5.739988533689899e-06,
+      "loss": 0.9007,
+      "step": 11349
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6956031322479248,
+      "learning_rate": 5.738307984007495e-06,
+      "loss": 0.9315,
+      "step": 11350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8070241212844849,
+      "learning_rate": 5.736627581384182e-06,
+      "loss": 0.9475,
+      "step": 11351
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7960282564163208,
+      "learning_rate": 5.73494732587794e-06,
+      "loss": 0.9441,
+      "step": 11352
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7883763313293457,
+      "learning_rate": 5.7332672175467545e-06,
+      "loss": 0.9634,
+      "step": 11353
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6098722219467163,
+      "learning_rate": 5.731587256448594e-06,
+      "loss": 0.9032,
+      "step": 11354
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7845321893692017,
+      "learning_rate": 5.729907442641438e-06,
+      "loss": 0.8938,
+      "step": 11355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8283565044403076,
+      "learning_rate": 5.728227776183244e-06,
+      "loss": 1.0146,
+      "step": 11356
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.741353154182434,
+      "learning_rate": 5.726548257131981e-06,
+      "loss": 0.9553,
+      "step": 11357
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.688572883605957,
+      "learning_rate": 5.724868885545597e-06,
+      "loss": 0.939,
+      "step": 11358
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.776901125907898,
+      "learning_rate": 5.723189661482045e-06,
+      "loss": 0.9235,
+      "step": 11359
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8392395973205566,
+      "learning_rate": 5.721510584999275e-06,
+      "loss": 0.9456,
+      "step": 11360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6833420991897583,
+      "learning_rate": 5.719831656155219e-06,
+      "loss": 0.9358,
+      "step": 11361
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7693593502044678,
+      "learning_rate": 5.718152875007821e-06,
+      "loss": 0.8412,
+      "step": 11362
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7292180061340332,
+      "learning_rate": 5.716474241615002e-06,
+      "loss": 0.967,
+      "step": 11363
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0668320655822754,
+      "learning_rate": 5.7147957560346955e-06,
+      "loss": 0.86,
+      "step": 11364
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7753098011016846,
+      "learning_rate": 5.713117418324814e-06,
+      "loss": 0.9726,
+      "step": 11365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8005441427230835,
+      "learning_rate": 5.711439228543278e-06,
+      "loss": 0.9117,
+      "step": 11366
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7748939990997314,
+      "learning_rate": 5.7097611867479915e-06,
+      "loss": 1.0715,
+      "step": 11367
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6896727085113525,
+      "learning_rate": 5.708083292996867e-06,
+      "loss": 0.8891,
+      "step": 11368
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5829453468322754,
+      "learning_rate": 5.7064055473477934e-06,
+      "loss": 0.875,
+      "step": 11369
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6841682195663452,
+      "learning_rate": 5.704727949858675e-06,
+      "loss": 0.8664,
+      "step": 11370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.752638578414917,
+      "learning_rate": 5.703050500587393e-06,
+      "loss": 0.8691,
+      "step": 11371
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6753618717193604,
+      "learning_rate": 5.7013731995918355e-06,
+      "loss": 0.9747,
+      "step": 11372
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8384405374526978,
+      "learning_rate": 5.699696046929885e-06,
+      "loss": 0.943,
+      "step": 11373
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9088551998138428,
+      "learning_rate": 5.698019042659407e-06,
+      "loss": 0.9491,
+      "step": 11374
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9232667684555054,
+      "learning_rate": 5.696342186838279e-06,
+      "loss": 0.9104,
+      "step": 11375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7725820541381836,
+      "learning_rate": 5.694665479524357e-06,
+      "loss": 0.9058,
+      "step": 11376
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.647047996520996,
+      "learning_rate": 5.692988920775506e-06,
+      "loss": 0.8647,
+      "step": 11377
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7829203605651855,
+      "learning_rate": 5.6913125106495725e-06,
+      "loss": 0.952,
+      "step": 11378
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.854789137840271,
+      "learning_rate": 5.689636249204412e-06,
+      "loss": 0.9181,
+      "step": 11379
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8685446977615356,
+      "learning_rate": 5.687960136497861e-06,
+      "loss": 0.9711,
+      "step": 11380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8763785362243652,
+      "learning_rate": 5.686284172587764e-06,
+      "loss": 0.9196,
+      "step": 11381
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6788727045059204,
+      "learning_rate": 5.684608357531946e-06,
+      "loss": 0.9355,
+      "step": 11382
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7597674131393433,
+      "learning_rate": 5.682932691388239e-06,
+      "loss": 0.8707,
+      "step": 11383
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.816584825515747,
+      "learning_rate": 5.68125717421447e-06,
+      "loss": 0.957,
+      "step": 11384
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0189027786254883,
+      "learning_rate": 5.679581806068448e-06,
+      "loss": 0.5032,
+      "step": 11385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7265815734863281,
+      "learning_rate": 5.677906587007993e-06,
+      "loss": 0.932,
+      "step": 11386
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7486203908920288,
+      "learning_rate": 5.676231517090904e-06,
+      "loss": 0.9954,
+      "step": 11387
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6739150285720825,
+      "learning_rate": 5.674556596374993e-06,
+      "loss": 0.8794,
+      "step": 11388
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6435792446136475,
+      "learning_rate": 5.672881824918046e-06,
+      "loss": 0.952,
+      "step": 11389
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8992124795913696,
+      "learning_rate": 5.671207202777864e-06,
+      "loss": 0.9311,
+      "step": 11390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.759995698928833,
+      "learning_rate": 5.669532730012226e-06,
+      "loss": 0.9304,
+      "step": 11391
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8289308547973633,
+      "learning_rate": 5.667858406678915e-06,
+      "loss": 0.8461,
+      "step": 11392
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7310559749603271,
+      "learning_rate": 5.666184232835711e-06,
+      "loss": 1.0165,
+      "step": 11393
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6345224380493164,
+      "learning_rate": 5.664510208540386e-06,
+      "loss": 0.8365,
+      "step": 11394
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.841046690940857,
+      "learning_rate": 5.6628363338506995e-06,
+      "loss": 0.9026,
+      "step": 11395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.867016315460205,
+      "learning_rate": 5.66116260882442e-06,
+      "loss": 0.9462,
+      "step": 11396
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.76585054397583,
+      "learning_rate": 5.659489033519294e-06,
+      "loss": 1.0184,
+      "step": 11397
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5102393627166748,
+      "learning_rate": 5.6578156079930824e-06,
+      "loss": 0.8637,
+      "step": 11398
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6155517101287842,
+      "learning_rate": 5.656142332303518e-06,
+      "loss": 0.8823,
+      "step": 11399
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.677160382270813,
+      "learning_rate": 5.65446920650835e-06,
+      "loss": 0.9642,
+      "step": 11400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8145761489868164,
+      "learning_rate": 5.652796230665314e-06,
+      "loss": 0.9221,
+      "step": 11401
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5694677829742432,
+      "learning_rate": 5.6511234048321325e-06,
+      "loss": 0.8843,
+      "step": 11402
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7132309675216675,
+      "learning_rate": 5.649450729066539e-06,
+      "loss": 0.9187,
+      "step": 11403
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7981895208358765,
+      "learning_rate": 5.647778203426244e-06,
+      "loss": 0.9496,
+      "step": 11404
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7436301708221436,
+      "learning_rate": 5.6461058279689685e-06,
+      "loss": 0.9728,
+      "step": 11405
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0067760944366455,
+      "learning_rate": 5.644433602752416e-06,
+      "loss": 0.5576,
+      "step": 11406
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7562768459320068,
+      "learning_rate": 5.642761527834297e-06,
+      "loss": 0.996,
+      "step": 11407
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8708735704421997,
+      "learning_rate": 5.641089603272301e-06,
+      "loss": 1.0431,
+      "step": 11408
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6377204656600952,
+      "learning_rate": 5.639417829124132e-06,
+      "loss": 0.8273,
+      "step": 11409
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7471342086791992,
+      "learning_rate": 5.637746205447469e-06,
+      "loss": 0.915,
+      "step": 11410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9942532777786255,
+      "learning_rate": 5.636074732300002e-06,
+      "loss": 0.9953,
+      "step": 11411
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6542710065841675,
+      "learning_rate": 5.634403409739402e-06,
+      "loss": 0.9299,
+      "step": 11412
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.836714744567871,
+      "learning_rate": 5.632732237823346e-06,
+      "loss": 0.966,
+      "step": 11413
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7244693040847778,
+      "learning_rate": 5.6310612166095055e-06,
+      "loss": 0.9559,
+      "step": 11414
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7735915184020996,
+      "learning_rate": 5.629390346155533e-06,
+      "loss": 0.9378,
+      "step": 11415
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6782585382461548,
+      "learning_rate": 5.627719626519096e-06,
+      "loss": 0.9366,
+      "step": 11416
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5590078830718994,
+      "learning_rate": 5.6260490577578365e-06,
+      "loss": 0.857,
+      "step": 11417
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.706998586654663,
+      "learning_rate": 5.624378639929411e-06,
+      "loss": 0.9245,
+      "step": 11418
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7812827825546265,
+      "learning_rate": 5.62270837309145e-06,
+      "loss": 0.831,
+      "step": 11419
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6757922172546387,
+      "learning_rate": 5.621038257301601e-06,
+      "loss": 0.9411,
+      "step": 11420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7457726001739502,
+      "learning_rate": 5.619368292617484e-06,
+      "loss": 0.8925,
+      "step": 11421
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8011186122894287,
+      "learning_rate": 5.617698479096736e-06,
+      "loss": 0.9744,
+      "step": 11422
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.837487816810608,
+      "learning_rate": 5.616028816796968e-06,
+      "loss": 0.9744,
+      "step": 11423
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7239410877227783,
+      "learning_rate": 5.614359305775803e-06,
+      "loss": 0.9241,
+      "step": 11424
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.007815957069397,
+      "learning_rate": 5.612689946090844e-06,
+      "loss": 0.5286,
+      "step": 11425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1158007383346558,
+      "learning_rate": 5.6110207377996985e-06,
+      "loss": 0.5159,
+      "step": 11426
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7071516513824463,
+      "learning_rate": 5.609351680959971e-06,
+      "loss": 0.9342,
+      "step": 11427
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6868281364440918,
+      "learning_rate": 5.607682775629249e-06,
+      "loss": 0.9794,
+      "step": 11428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.980263352394104,
+      "learning_rate": 5.606014021865129e-06,
+      "loss": 0.5045,
+      "step": 11429
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6178532838821411,
+      "learning_rate": 5.604345419725188e-06,
+      "loss": 0.8148,
+      "step": 11430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5486879348754883,
+      "learning_rate": 5.6026769692670106e-06,
+      "loss": 0.8594,
+      "step": 11431
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9661365747451782,
+      "learning_rate": 5.601008670548162e-06,
+      "loss": 0.4986,
+      "step": 11432
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6334351301193237,
+      "learning_rate": 5.599340523626222e-06,
+      "loss": 0.9142,
+      "step": 11433
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9416334629058838,
+      "learning_rate": 5.5976725285587445e-06,
+      "loss": 0.9804,
+      "step": 11434
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8198814392089844,
+      "learning_rate": 5.596004685403294e-06,
+      "loss": 0.9481,
+      "step": 11435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9834403991699219,
+      "learning_rate": 5.594336994217416e-06,
+      "loss": 1.0434,
+      "step": 11436
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6068751811981201,
+      "learning_rate": 5.59266945505866e-06,
+      "loss": 0.8837,
+      "step": 11437
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6591851711273193,
+      "learning_rate": 5.5910020679845745e-06,
+      "loss": 0.9259,
+      "step": 11438
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.711958408355713,
+      "learning_rate": 5.5893348330526885e-06,
+      "loss": 0.8653,
+      "step": 11439
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7710493803024292,
+      "learning_rate": 5.58766775032054e-06,
+      "loss": 0.9518,
+      "step": 11440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7233673334121704,
+      "learning_rate": 5.586000819845647e-06,
+      "loss": 0.9111,
+      "step": 11441
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6358797550201416,
+      "learning_rate": 5.584334041685542e-06,
+      "loss": 0.8505,
+      "step": 11442
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6612132787704468,
+      "learning_rate": 5.582667415897729e-06,
+      "loss": 0.92,
+      "step": 11443
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7748699188232422,
+      "learning_rate": 5.581000942539729e-06,
+      "loss": 0.9163,
+      "step": 11444
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8957267999649048,
+      "learning_rate": 5.579334621669038e-06,
+      "loss": 0.8912,
+      "step": 11445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6913210153579712,
+      "learning_rate": 5.577668453343165e-06,
+      "loss": 0.8982,
+      "step": 11446
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7763898372650146,
+      "learning_rate": 5.576002437619595e-06,
+      "loss": 0.8938,
+      "step": 11447
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6640903949737549,
+      "learning_rate": 5.574336574555829e-06,
+      "loss": 0.924,
+      "step": 11448
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8456172943115234,
+      "learning_rate": 5.572670864209339e-06,
+      "loss": 0.9085,
+      "step": 11449
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6747535467147827,
+      "learning_rate": 5.571005306637611e-06,
+      "loss": 0.9439,
+      "step": 11450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6667211055755615,
+      "learning_rate": 5.569339901898123e-06,
+      "loss": 0.9794,
+      "step": 11451
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5908657312393188,
+      "learning_rate": 5.567674650048334e-06,
+      "loss": 0.9393,
+      "step": 11452
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8179203271865845,
+      "learning_rate": 5.566009551145716e-06,
+      "loss": 0.9556,
+      "step": 11453
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.90180242061615,
+      "learning_rate": 5.564344605247718e-06,
+      "loss": 1.0459,
+      "step": 11454
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.785592794418335,
+      "learning_rate": 5.5626798124118005e-06,
+      "loss": 0.919,
+      "step": 11455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7312959432601929,
+      "learning_rate": 5.561015172695406e-06,
+      "loss": 0.9251,
+      "step": 11456
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9106122255325317,
+      "learning_rate": 5.559350686155979e-06,
+      "loss": 0.9373,
+      "step": 11457
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6820158958435059,
+      "learning_rate": 5.5576863528509486e-06,
+      "loss": 1.0181,
+      "step": 11458
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.640638828277588,
+      "learning_rate": 5.556022172837761e-06,
+      "loss": 0.8923,
+      "step": 11459
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6908283233642578,
+      "learning_rate": 5.554358146173831e-06,
+      "loss": 0.8973,
+      "step": 11460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8675917387008667,
+      "learning_rate": 5.552694272916586e-06,
+      "loss": 0.9577,
+      "step": 11461
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.731400728225708,
+      "learning_rate": 5.551030553123436e-06,
+      "loss": 0.9217,
+      "step": 11462
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7646234035491943,
+      "learning_rate": 5.549366986851797e-06,
+      "loss": 0.8923,
+      "step": 11463
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7584969997406006,
+      "learning_rate": 5.547703574159067e-06,
+      "loss": 0.9994,
+      "step": 11464
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6685765981674194,
+      "learning_rate": 5.546040315102653e-06,
+      "loss": 0.8648,
+      "step": 11465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7235735654830933,
+      "learning_rate": 5.544377209739943e-06,
+      "loss": 0.8839,
+      "step": 11466
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8909136056900024,
+      "learning_rate": 5.542714258128329e-06,
+      "loss": 0.9504,
+      "step": 11467
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8063448667526245,
+      "learning_rate": 5.541051460325199e-06,
+      "loss": 0.9321,
+      "step": 11468
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7116748094558716,
+      "learning_rate": 5.539388816387922e-06,
+      "loss": 0.9564,
+      "step": 11469
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.92942214012146,
+      "learning_rate": 5.537726326373883e-06,
+      "loss": 0.9518,
+      "step": 11470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8322480916976929,
+      "learning_rate": 5.536063990340439e-06,
+      "loss": 0.975,
+      "step": 11471
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.627907395362854,
+      "learning_rate": 5.5344018083449615e-06,
+      "loss": 0.881,
+      "step": 11472
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.822126030921936,
+      "learning_rate": 5.532739780444799e-06,
+      "loss": 1.0355,
+      "step": 11473
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.627274751663208,
+      "learning_rate": 5.531077906697312e-06,
+      "loss": 0.9008,
+      "step": 11474
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7201422452926636,
+      "learning_rate": 5.52941618715984e-06,
+      "loss": 0.9682,
+      "step": 11475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8258881568908691,
+      "learning_rate": 5.52775462188973e-06,
+      "loss": 0.9198,
+      "step": 11476
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0668574571609497,
+      "learning_rate": 5.52609321094431e-06,
+      "loss": 0.5941,
+      "step": 11477
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7159940004348755,
+      "learning_rate": 5.524431954380922e-06,
+      "loss": 0.9072,
+      "step": 11478
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7637144327163696,
+      "learning_rate": 5.52277085225688e-06,
+      "loss": 0.9198,
+      "step": 11479
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8645601272583008,
+      "learning_rate": 5.521109904629511e-06,
+      "loss": 0.9771,
+      "step": 11480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0816428661346436,
+      "learning_rate": 5.51944911155613e-06,
+      "loss": 0.625,
+      "step": 11481
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7623848915100098,
+      "learning_rate": 5.517788473094041e-06,
+      "loss": 0.9269,
+      "step": 11482
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9972187280654907,
+      "learning_rate": 5.516127989300556e-06,
+      "loss": 0.5633,
+      "step": 11483
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7359601259231567,
+      "learning_rate": 5.514467660232965e-06,
+      "loss": 0.9553,
+      "step": 11484
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8252882957458496,
+      "learning_rate": 5.512807485948568e-06,
+      "loss": 0.8859,
+      "step": 11485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6612387895584106,
+      "learning_rate": 5.5111474665046475e-06,
+      "loss": 0.9042,
+      "step": 11486
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.821286916732788,
+      "learning_rate": 5.509487601958491e-06,
+      "loss": 0.8859,
+      "step": 11487
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7292828559875488,
+      "learning_rate": 5.50782789236737e-06,
+      "loss": 0.9862,
+      "step": 11488
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5921818017959595,
+      "learning_rate": 5.5061683377885645e-06,
+      "loss": 0.9683,
+      "step": 11489
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9562250375747681,
+      "learning_rate": 5.504508938279334e-06,
+      "loss": 0.5405,
+      "step": 11490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8294230699539185,
+      "learning_rate": 5.502849693896941e-06,
+      "loss": 0.8617,
+      "step": 11491
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.721171498298645,
+      "learning_rate": 5.501190604698647e-06,
+      "loss": 0.9466,
+      "step": 11492
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7587980031967163,
+      "learning_rate": 5.499531670741694e-06,
+      "loss": 0.953,
+      "step": 11493
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9093011617660522,
+      "learning_rate": 5.497872892083336e-06,
+      "loss": 0.9648,
+      "step": 11494
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.916135549545288,
+      "learning_rate": 5.496214268780804e-06,
+      "loss": 1.0285,
+      "step": 11495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6544861793518066,
+      "learning_rate": 5.494555800891342e-06,
+      "loss": 0.9849,
+      "step": 11496
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1265995502471924,
+      "learning_rate": 5.492897488472167e-06,
+      "loss": 0.5242,
+      "step": 11497
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8691014051437378,
+      "learning_rate": 5.491239331580515e-06,
+      "loss": 0.9291,
+      "step": 11498
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7157424688339233,
+      "learning_rate": 5.4895813302735965e-06,
+      "loss": 0.8802,
+      "step": 11499
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6429020166397095,
+      "learning_rate": 5.487923484608629e-06,
+      "loss": 0.8892,
+      "step": 11500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8457252979278564,
+      "learning_rate": 5.4862657946428155e-06,
+      "loss": 0.878,
+      "step": 11501
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9185314178466797,
+      "learning_rate": 5.484608260433364e-06,
+      "loss": 0.9211,
+      "step": 11502
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7743207216262817,
+      "learning_rate": 5.4829508820374645e-06,
+      "loss": 0.852,
+      "step": 11503
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6909737586975098,
+      "learning_rate": 5.481293659512312e-06,
+      "loss": 0.9529,
+      "step": 11504
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8193588256835938,
+      "learning_rate": 5.479636592915096e-06,
+      "loss": 0.9207,
+      "step": 11505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.639646053314209,
+      "learning_rate": 5.477979682302992e-06,
+      "loss": 0.9547,
+      "step": 11506
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7361444234848022,
+      "learning_rate": 5.47632292773318e-06,
+      "loss": 0.9694,
+      "step": 11507
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9035722017288208,
+      "learning_rate": 5.474666329262823e-06,
+      "loss": 0.9077,
+      "step": 11508
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6163007020950317,
+      "learning_rate": 5.473009886949094e-06,
+      "loss": 0.8152,
+      "step": 11509
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6628053188323975,
+      "learning_rate": 5.4713536008491455e-06,
+      "loss": 0.9493,
+      "step": 11510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7400606870651245,
+      "learning_rate": 5.4696974710201375e-06,
+      "loss": 0.9133,
+      "step": 11511
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8280506134033203,
+      "learning_rate": 5.468041497519211e-06,
+      "loss": 0.9639,
+      "step": 11512
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.655489206314087,
+      "learning_rate": 5.466385680403517e-06,
+      "loss": 0.9196,
+      "step": 11513
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.751820683479309,
+      "learning_rate": 5.464730019730185e-06,
+      "loss": 0.9209,
+      "step": 11514
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.812620997428894,
+      "learning_rate": 5.46307451555635e-06,
+      "loss": 0.9288,
+      "step": 11515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8815985918045044,
+      "learning_rate": 5.461419167939145e-06,
+      "loss": 0.9429,
+      "step": 11516
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7389006614685059,
+      "learning_rate": 5.459763976935681e-06,
+      "loss": 0.8534,
+      "step": 11517
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.872350811958313,
+      "learning_rate": 5.4581089426030865e-06,
+      "loss": 0.946,
+      "step": 11518
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.04293954372406,
+      "learning_rate": 5.45645406499846e-06,
+      "loss": 0.5744,
+      "step": 11519
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.738649606704712,
+      "learning_rate": 5.454799344178914e-06,
+      "loss": 0.881,
+      "step": 11520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8397550582885742,
+      "learning_rate": 5.4531447802015445e-06,
+      "loss": 0.9773,
+      "step": 11521
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.888254165649414,
+      "learning_rate": 5.45149037312345e-06,
+      "loss": 1.0335,
+      "step": 11522
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6881126165390015,
+      "learning_rate": 5.449836123001714e-06,
+      "loss": 0.9937,
+      "step": 11523
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7345950603485107,
+      "learning_rate": 5.448182029893423e-06,
+      "loss": 0.8285,
+      "step": 11524
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.787186622619629,
+      "learning_rate": 5.4465280938556545e-06,
+      "loss": 0.9602,
+      "step": 11525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6280945539474487,
+      "learning_rate": 5.444874314945485e-06,
+      "loss": 0.9719,
+      "step": 11526
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9787378311157227,
+      "learning_rate": 5.443220693219976e-06,
+      "loss": 0.9074,
+      "step": 11527
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9109443426132202,
+      "learning_rate": 5.441567228736195e-06,
+      "loss": 0.9853,
+      "step": 11528
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9916813969612122,
+      "learning_rate": 5.439913921551193e-06,
+      "loss": 0.569,
+      "step": 11529
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7193677425384521,
+      "learning_rate": 5.438260771722027e-06,
+      "loss": 0.9586,
+      "step": 11530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.85320246219635,
+      "learning_rate": 5.436607779305735e-06,
+      "loss": 0.9215,
+      "step": 11531
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6162906885147095,
+      "learning_rate": 5.434954944359365e-06,
+      "loss": 0.9022,
+      "step": 11532
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.770362138748169,
+      "learning_rate": 5.433302266939944e-06,
+      "loss": 0.905,
+      "step": 11533
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0092549324035645,
+      "learning_rate": 5.431649747104505e-06,
+      "loss": 0.926,
+      "step": 11534
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7938748598098755,
+      "learning_rate": 5.429997384910075e-06,
+      "loss": 0.9024,
+      "step": 11535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.872809886932373,
+      "learning_rate": 5.428345180413667e-06,
+      "loss": 0.9583,
+      "step": 11536
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.754981517791748,
+      "learning_rate": 5.426693133672301e-06,
+      "loss": 0.9708,
+      "step": 11537
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6853177547454834,
+      "learning_rate": 5.425041244742975e-06,
+      "loss": 0.9505,
+      "step": 11538
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9788057804107666,
+      "learning_rate": 5.4233895136827e-06,
+      "loss": 0.9462,
+      "step": 11539
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.873182773590088,
+      "learning_rate": 5.421737940548464e-06,
+      "loss": 0.9124,
+      "step": 11540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7483211755752563,
+      "learning_rate": 5.420086525397268e-06,
+      "loss": 0.9743,
+      "step": 11541
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7042179107666016,
+      "learning_rate": 5.418435268286089e-06,
+      "loss": 1.0,
+      "step": 11542
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.109450340270996,
+      "learning_rate": 5.416784169271913e-06,
+      "loss": 0.9745,
+      "step": 11543
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7788567543029785,
+      "learning_rate": 5.415133228411709e-06,
+      "loss": 0.8561,
+      "step": 11544
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6479825973510742,
+      "learning_rate": 5.4134824457624504e-06,
+      "loss": 0.8752,
+      "step": 11545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5855357646942139,
+      "learning_rate": 5.411831821381105e-06,
+      "loss": 0.8289,
+      "step": 11546
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.67849600315094,
+      "learning_rate": 5.410181355324622e-06,
+      "loss": 0.9589,
+      "step": 11547
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7494593858718872,
+      "learning_rate": 5.408531047649964e-06,
+      "loss": 0.9427,
+      "step": 11548
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7740103006362915,
+      "learning_rate": 5.406880898414069e-06,
+      "loss": 0.9479,
+      "step": 11549
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8178991079330444,
+      "learning_rate": 5.405230907673889e-06,
+      "loss": 1.0246,
+      "step": 11550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8297910690307617,
+      "learning_rate": 5.403581075486351e-06,
+      "loss": 0.8878,
+      "step": 11551
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9474565982818604,
+      "learning_rate": 5.401931401908394e-06,
+      "loss": 0.9151,
+      "step": 11552
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0433881282806396,
+      "learning_rate": 5.400281886996938e-06,
+      "loss": 0.9608,
+      "step": 11553
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0580730438232422,
+      "learning_rate": 5.3986325308089075e-06,
+      "loss": 0.5894,
+      "step": 11554
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5779671669006348,
+      "learning_rate": 5.396983333401211e-06,
+      "loss": 0.9373,
+      "step": 11555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.670257329940796,
+      "learning_rate": 5.395334294830766e-06,
+      "loss": 0.8942,
+      "step": 11556
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8482142686843872,
+      "learning_rate": 5.393685415154468e-06,
+      "loss": 0.9514,
+      "step": 11557
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7733861207962036,
+      "learning_rate": 5.392036694429219e-06,
+      "loss": 0.942,
+      "step": 11558
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0109962224960327,
+      "learning_rate": 5.390388132711916e-06,
+      "loss": 0.5292,
+      "step": 11559
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7582389116287231,
+      "learning_rate": 5.388739730059438e-06,
+      "loss": 0.9298,
+      "step": 11560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.804871678352356,
+      "learning_rate": 5.387091486528675e-06,
+      "loss": 0.9533,
+      "step": 11561
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0298603773117065,
+      "learning_rate": 5.385443402176494e-06,
+      "loss": 0.5349,
+      "step": 11562
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0929852724075317,
+      "learning_rate": 5.383795477059776e-06,
+      "loss": 0.6014,
+      "step": 11563
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8419651985168457,
+      "learning_rate": 5.382147711235377e-06,
+      "loss": 0.8793,
+      "step": 11564
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7546218633651733,
+      "learning_rate": 5.380500104760165e-06,
+      "loss": 0.9293,
+      "step": 11565
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.869836688041687,
+      "learning_rate": 5.3788526576909874e-06,
+      "loss": 0.8758,
+      "step": 11566
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7971305847167969,
+      "learning_rate": 5.3772053700847e-06,
+      "loss": 0.9042,
+      "step": 11567
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9503904581069946,
+      "learning_rate": 5.3755582419981354e-06,
+      "loss": 1.0142,
+      "step": 11568
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9284682273864746,
+      "learning_rate": 5.373911273488139e-06,
+      "loss": 0.9625,
+      "step": 11569
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.793882131576538,
+      "learning_rate": 5.372264464611548e-06,
+      "loss": 0.924,
+      "step": 11570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7408554553985596,
+      "learning_rate": 5.370617815425177e-06,
+      "loss": 0.9551,
+      "step": 11571
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0461440086364746,
+      "learning_rate": 5.368971325985859e-06,
+      "loss": 0.5158,
+      "step": 11572
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.737757682800293,
+      "learning_rate": 5.3673249963504005e-06,
+      "loss": 0.9009,
+      "step": 11573
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6210299730300903,
+      "learning_rate": 5.3656788265756175e-06,
+      "loss": 0.9465,
+      "step": 11574
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8598971366882324,
+      "learning_rate": 5.364032816718311e-06,
+      "loss": 0.863,
+      "step": 11575
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8453575372695923,
+      "learning_rate": 5.362386966835285e-06,
+      "loss": 0.92,
+      "step": 11576
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9715495705604553,
+      "learning_rate": 5.360741276983325e-06,
+      "loss": 0.5064,
+      "step": 11577
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.72489595413208,
+      "learning_rate": 5.359095747219231e-06,
+      "loss": 0.903,
+      "step": 11578
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7449196577072144,
+      "learning_rate": 5.357450377599773e-06,
+      "loss": 0.9767,
+      "step": 11579
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7011772394180298,
+      "learning_rate": 5.355805168181738e-06,
+      "loss": 1.0102,
+      "step": 11580
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8990588188171387,
+      "learning_rate": 5.354160119021891e-06,
+      "loss": 0.9222,
+      "step": 11581
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6559391021728516,
+      "learning_rate": 5.352515230177e-06,
+      "loss": 0.9839,
+      "step": 11582
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7523138523101807,
+      "learning_rate": 5.350870501703829e-06,
+      "loss": 0.9601,
+      "step": 11583
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7118678092956543,
+      "learning_rate": 5.3492259336591275e-06,
+      "loss": 0.8945,
+      "step": 11584
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8525192737579346,
+      "learning_rate": 5.347581526099651e-06,
+      "loss": 0.9597,
+      "step": 11585
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8023669719696045,
+      "learning_rate": 5.345937279082136e-06,
+      "loss": 0.897,
+      "step": 11586
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7100729942321777,
+      "learning_rate": 5.344293192663329e-06,
+      "loss": 0.8538,
+      "step": 11587
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.794411063194275,
+      "learning_rate": 5.342649266899955e-06,
+      "loss": 1.0126,
+      "step": 11588
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6426360607147217,
+      "learning_rate": 5.341005501848749e-06,
+      "loss": 0.9104,
+      "step": 11589
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.933236837387085,
+      "learning_rate": 5.33936189756642e-06,
+      "loss": 0.9685,
+      "step": 11590
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.662968635559082,
+      "learning_rate": 5.337718454109702e-06,
+      "loss": 0.9731,
+      "step": 11591
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6944464445114136,
+      "learning_rate": 5.336075171535292e-06,
+      "loss": 0.8069,
+      "step": 11592
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.620992660522461,
+      "learning_rate": 5.334432049899904e-06,
+      "loss": 0.9808,
+      "step": 11593
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.782029151916504,
+      "learning_rate": 5.3327890892602286e-06,
+      "loss": 0.8903,
+      "step": 11594
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8381882905960083,
+      "learning_rate": 5.331146289672968e-06,
+      "loss": 0.95,
+      "step": 11595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.635562777519226,
+      "learning_rate": 5.329503651194805e-06,
+      "loss": 0.8921,
+      "step": 11596
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8567148447036743,
+      "learning_rate": 5.327861173882427e-06,
+      "loss": 0.8928,
+      "step": 11597
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6727715730667114,
+      "learning_rate": 5.326218857792505e-06,
+      "loss": 0.9496,
+      "step": 11598
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6957740783691406,
+      "learning_rate": 5.324576702981716e-06,
+      "loss": 0.9577,
+      "step": 11599
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5796741247177124,
+      "learning_rate": 5.322934709506726e-06,
+      "loss": 0.9309,
+      "step": 11600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8734960556030273,
+      "learning_rate": 5.321292877424192e-06,
+      "loss": 0.9592,
+      "step": 11601
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8251913785934448,
+      "learning_rate": 5.319651206790775e-06,
+      "loss": 0.8968,
+      "step": 11602
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6572778224945068,
+      "learning_rate": 5.318009697663118e-06,
+      "loss": 0.9261,
+      "step": 11603
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.157059669494629,
+      "learning_rate": 5.316368350097869e-06,
+      "loss": 0.9059,
+      "step": 11604
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8795872926712036,
+      "learning_rate": 5.314727164151663e-06,
+      "loss": 0.9272,
+      "step": 11605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8853521347045898,
+      "learning_rate": 5.3130861398811385e-06,
+      "loss": 0.9651,
+      "step": 11606
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7526181936264038,
+      "learning_rate": 5.311445277342915e-06,
+      "loss": 0.9949,
+      "step": 11607
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7402299642562866,
+      "learning_rate": 5.309804576593623e-06,
+      "loss": 1.0122,
+      "step": 11608
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.888508915901184,
+      "learning_rate": 5.308164037689867e-06,
+      "loss": 0.9748,
+      "step": 11609
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8047722578048706,
+      "learning_rate": 5.30652366068827e-06,
+      "loss": 0.9391,
+      "step": 11610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8021529912948608,
+      "learning_rate": 5.304883445645425e-06,
+      "loss": 0.9236,
+      "step": 11611
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.682714581489563,
+      "learning_rate": 5.3032433926179395e-06,
+      "loss": 0.8347,
+      "step": 11612
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.193568706512451,
+      "learning_rate": 5.301603501662407e-06,
+      "loss": 1.0248,
+      "step": 11613
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9750372171401978,
+      "learning_rate": 5.29996377283541e-06,
+      "loss": 0.9867,
+      "step": 11614
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8413037061691284,
+      "learning_rate": 5.2983242061935365e-06,
+      "loss": 0.9245,
+      "step": 11615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.947635531425476,
+      "learning_rate": 5.296684801793359e-06,
+      "loss": 0.9093,
+      "step": 11616
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6264203786849976,
+      "learning_rate": 5.295045559691454e-06,
+      "loss": 0.9554,
+      "step": 11617
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0501279830932617,
+      "learning_rate": 5.293406479944381e-06,
+      "loss": 0.5505,
+      "step": 11618
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7028065919876099,
+      "learning_rate": 5.291767562608705e-06,
+      "loss": 1.0395,
+      "step": 11619
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.570185661315918,
+      "learning_rate": 5.290128807740976e-06,
+      "loss": 0.9477,
+      "step": 11620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7570148706436157,
+      "learning_rate": 5.288490215397749e-06,
+      "loss": 0.8676,
+      "step": 11621
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7937171459197998,
+      "learning_rate": 5.286851785635559e-06,
+      "loss": 0.8738,
+      "step": 11622
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8685243129730225,
+      "learning_rate": 5.28521351851095e-06,
+      "loss": 0.9151,
+      "step": 11623
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8277575969696045,
+      "learning_rate": 5.283575414080455e-06,
+      "loss": 0.9532,
+      "step": 11624
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6919366121292114,
+      "learning_rate": 5.281937472400594e-06,
+      "loss": 0.9393,
+      "step": 11625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.760114073753357,
+      "learning_rate": 5.280299693527895e-06,
+      "loss": 0.9267,
+      "step": 11626
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6749792098999023,
+      "learning_rate": 5.278662077518866e-06,
+      "loss": 0.9551,
+      "step": 11627
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.653491497039795,
+      "learning_rate": 5.2770246244300225e-06,
+      "loss": 0.9102,
+      "step": 11628
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9866293668746948,
+      "learning_rate": 5.275387334317864e-06,
+      "loss": 0.5441,
+      "step": 11629
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.599632740020752,
+      "learning_rate": 5.273750207238894e-06,
+      "loss": 0.9997,
+      "step": 11630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7295571565628052,
+      "learning_rate": 5.272113243249599e-06,
+      "loss": 0.9848,
+      "step": 11631
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7558050155639648,
+      "learning_rate": 5.270476442406472e-06,
+      "loss": 0.8713,
+      "step": 11632
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0101670026779175,
+      "learning_rate": 5.268839804765988e-06,
+      "loss": 0.5259,
+      "step": 11633
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7342720031738281,
+      "learning_rate": 5.267203330384632e-06,
+      "loss": 0.939,
+      "step": 11634
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9709718227386475,
+      "learning_rate": 5.265567019318862e-06,
+      "loss": 0.9353,
+      "step": 11635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9546594619750977,
+      "learning_rate": 5.263930871625151e-06,
+      "loss": 0.9891,
+      "step": 11636
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8432539701461792,
+      "learning_rate": 5.2622948873599595e-06,
+      "loss": 0.9238,
+      "step": 11637
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7171056270599365,
+      "learning_rate": 5.260659066579733e-06,
+      "loss": 0.9101,
+      "step": 11638
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.618330955505371,
+      "learning_rate": 5.259023409340926e-06,
+      "loss": 0.9904,
+      "step": 11639
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6643435955047607,
+      "learning_rate": 5.257387915699976e-06,
+      "loss": 0.8719,
+      "step": 11640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6756975650787354,
+      "learning_rate": 5.255752585713324e-06,
+      "loss": 0.7993,
+      "step": 11641
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7428895235061646,
+      "learning_rate": 5.254117419437394e-06,
+      "loss": 0.9586,
+      "step": 11642
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.834367275238037,
+      "learning_rate": 5.252482416928619e-06,
+      "loss": 0.9845,
+      "step": 11643
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.76563560962677,
+      "learning_rate": 5.2508475782434095e-06,
+      "loss": 0.9248,
+      "step": 11644
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7358566522598267,
+      "learning_rate": 5.2492129034381875e-06,
+      "loss": 1.0042,
+      "step": 11645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7760509252548218,
+      "learning_rate": 5.247578392569354e-06,
+      "loss": 1.0442,
+      "step": 11646
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6131383180618286,
+      "learning_rate": 5.2459440456933156e-06,
+      "loss": 0.888,
+      "step": 11647
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7445423603057861,
+      "learning_rate": 5.24430986286647e-06,
+      "loss": 0.8852,
+      "step": 11648
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8820563554763794,
+      "learning_rate": 5.242675844145204e-06,
+      "loss": 0.9549,
+      "step": 11649
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5935759544372559,
+      "learning_rate": 5.241041989585911e-06,
+      "loss": 0.9394,
+      "step": 11650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.603927731513977,
+      "learning_rate": 5.239408299244959e-06,
+      "loss": 0.8361,
+      "step": 11651
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.622622013092041,
+      "learning_rate": 5.237774773178734e-06,
+      "loss": 0.8539,
+      "step": 11652
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5777705907821655,
+      "learning_rate": 5.236141411443594e-06,
+      "loss": 0.8871,
+      "step": 11653
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.781461477279663,
+      "learning_rate": 5.23450821409591e-06,
+      "loss": 1.0093,
+      "step": 11654
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6174107789993286,
+      "learning_rate": 5.232875181192033e-06,
+      "loss": 0.9867,
+      "step": 11655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0862786769866943,
+      "learning_rate": 5.231242312788316e-06,
+      "loss": 0.55,
+      "step": 11656
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7979928255081177,
+      "learning_rate": 5.229609608941106e-06,
+      "loss": 0.8687,
+      "step": 11657
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9316529035568237,
+      "learning_rate": 5.227977069706748e-06,
+      "loss": 0.9461,
+      "step": 11658
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7856119871139526,
+      "learning_rate": 5.226344695141567e-06,
+      "loss": 0.9508,
+      "step": 11659
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6399339437484741,
+      "learning_rate": 5.224712485301898e-06,
+      "loss": 0.9082,
+      "step": 11660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.898725152015686,
+      "learning_rate": 5.223080440244059e-06,
+      "loss": 0.9459,
+      "step": 11661
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7254668474197388,
+      "learning_rate": 5.2214485600243756e-06,
+      "loss": 0.9364,
+      "step": 11662
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9469966888427734,
+      "learning_rate": 5.219816844699148e-06,
+      "loss": 1.025,
+      "step": 11663
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5856157541275024,
+      "learning_rate": 5.218185294324694e-06,
+      "loss": 0.8914,
+      "step": 11664
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9075679779052734,
+      "learning_rate": 5.2165539089573025e-06,
+      "loss": 1.0112,
+      "step": 11665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7695640325546265,
+      "learning_rate": 5.214922688653274e-06,
+      "loss": 0.8524,
+      "step": 11666
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0467355251312256,
+      "learning_rate": 5.213291633468901e-06,
+      "loss": 1.0058,
+      "step": 11667
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.226175546646118,
+      "learning_rate": 5.211660743460458e-06,
+      "loss": 1.026,
+      "step": 11668
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.812477946281433,
+      "learning_rate": 5.210030018684233e-06,
+      "loss": 0.9607,
+      "step": 11669
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1455178260803223,
+      "learning_rate": 5.208399459196486e-06,
+      "loss": 0.8835,
+      "step": 11670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9750603437423706,
+      "learning_rate": 5.206769065053494e-06,
+      "loss": 0.9793,
+      "step": 11671
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6402770280838013,
+      "learning_rate": 5.205138836311508e-06,
+      "loss": 0.8243,
+      "step": 11672
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8187425136566162,
+      "learning_rate": 5.20350877302679e-06,
+      "loss": 0.9573,
+      "step": 11673
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6716643571853638,
+      "learning_rate": 5.201878875255582e-06,
+      "loss": 0.9245,
+      "step": 11674
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.706623911857605,
+      "learning_rate": 5.2002491430541346e-06,
+      "loss": 0.9697,
+      "step": 11675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6241567134857178,
+      "learning_rate": 5.198619576478678e-06,
+      "loss": 0.9486,
+      "step": 11676
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8172380924224854,
+      "learning_rate": 5.196990175585449e-06,
+      "loss": 0.9072,
+      "step": 11677
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.908457636833191,
+      "learning_rate": 5.195360940430676e-06,
+      "loss": 0.8448,
+      "step": 11678
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6959284543991089,
+      "learning_rate": 5.1937318710705706e-06,
+      "loss": 0.949,
+      "step": 11679
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7938159704208374,
+      "learning_rate": 5.192102967561357e-06,
+      "loss": 0.8996,
+      "step": 11680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.590562105178833,
+      "learning_rate": 5.190474229959236e-06,
+      "loss": 0.9546,
+      "step": 11681
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7886768579483032,
+      "learning_rate": 5.188845658320419e-06,
+      "loss": 0.9238,
+      "step": 11682
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8281091451644897,
+      "learning_rate": 5.187217252701093e-06,
+      "loss": 0.8072,
+      "step": 11683
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8246984481811523,
+      "learning_rate": 5.1855890131574615e-06,
+      "loss": 0.9469,
+      "step": 11684
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7635284662246704,
+      "learning_rate": 5.1839609397457e-06,
+      "loss": 0.9573,
+      "step": 11685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.762611746788025,
+      "learning_rate": 5.182333032521997e-06,
+      "loss": 0.9996,
+      "step": 11686
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7961598634719849,
+      "learning_rate": 5.18070529154252e-06,
+      "loss": 0.955,
+      "step": 11687
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.871989130973816,
+      "learning_rate": 5.179077716863445e-06,
+      "loss": 0.9361,
+      "step": 11688
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0644457340240479,
+      "learning_rate": 5.177450308540928e-06,
+      "loss": 0.5879,
+      "step": 11689
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8947426080703735,
+      "learning_rate": 5.1758230666311286e-06,
+      "loss": 0.9638,
+      "step": 11690
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7619361877441406,
+      "learning_rate": 5.174195991190203e-06,
+      "loss": 0.884,
+      "step": 11691
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.802830457687378,
+      "learning_rate": 5.17256908227429e-06,
+      "loss": 0.9009,
+      "step": 11692
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.696345329284668,
+      "learning_rate": 5.170942339939538e-06,
+      "loss": 0.9669,
+      "step": 11693
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7990154027938843,
+      "learning_rate": 5.169315764242071e-06,
+      "loss": 1.0317,
+      "step": 11694
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9006434679031372,
+      "learning_rate": 5.167689355238028e-06,
+      "loss": 0.912,
+      "step": 11695
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.524186134338379,
+      "learning_rate": 5.166063112983522e-06,
+      "loss": 0.9333,
+      "step": 11696
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8372793197631836,
+      "learning_rate": 5.16443703753468e-06,
+      "loss": 0.8728,
+      "step": 11697
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6507500410079956,
+      "learning_rate": 5.1628111289476025e-06,
+      "loss": 0.9579,
+      "step": 11698
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7946605682373047,
+      "learning_rate": 5.1611853872784065e-06,
+      "loss": 0.9745,
+      "step": 11699
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.781618356704712,
+      "learning_rate": 5.159559812583181e-06,
+      "loss": 0.9549,
+      "step": 11700
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.867897391319275,
+      "learning_rate": 5.157934404918025e-06,
+      "loss": 0.8843,
+      "step": 11701
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9105182886123657,
+      "learning_rate": 5.1563091643390324e-06,
+      "loss": 0.9926,
+      "step": 11702
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.707181692123413,
+      "learning_rate": 5.154684090902275e-06,
+      "loss": 0.8772,
+      "step": 11703
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5971965789794922,
+      "learning_rate": 5.15305918466384e-06,
+      "loss": 0.8584,
+      "step": 11704
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7153371572494507,
+      "learning_rate": 5.151434445679788e-06,
+      "loss": 1.0003,
+      "step": 11705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7008590698242188,
+      "learning_rate": 5.149809874006194e-06,
+      "loss": 0.9284,
+      "step": 11706
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5508931875228882,
+      "learning_rate": 5.148185469699109e-06,
+      "loss": 0.9212,
+      "step": 11707
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6856895685195923,
+      "learning_rate": 5.146561232814593e-06,
+      "loss": 0.9562,
+      "step": 11708
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6909445524215698,
+      "learning_rate": 5.144937163408689e-06,
+      "loss": 0.9435,
+      "step": 11709
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8079134225845337,
+      "learning_rate": 5.143313261537443e-06,
+      "loss": 0.9669,
+      "step": 11710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7138488292694092,
+      "learning_rate": 5.141689527256889e-06,
+      "loss": 0.9888,
+      "step": 11711
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.628553867340088,
+      "learning_rate": 5.140065960623061e-06,
+      "loss": 0.8738,
+      "step": 11712
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8324013948440552,
+      "learning_rate": 5.138442561691976e-06,
+      "loss": 0.9254,
+      "step": 11713
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7348748445510864,
+      "learning_rate": 5.136819330519659e-06,
+      "loss": 0.8754,
+      "step": 11714
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7780094146728516,
+      "learning_rate": 5.135196267162126e-06,
+      "loss": 0.9213,
+      "step": 11715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.78496515750885,
+      "learning_rate": 5.133573371675375e-06,
+      "loss": 0.8827,
+      "step": 11716
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6222641468048096,
+      "learning_rate": 5.1319506441154195e-06,
+      "loss": 1.0103,
+      "step": 11717
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8690211772918701,
+      "learning_rate": 5.130328084538244e-06,
+      "loss": 0.9449,
+      "step": 11718
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.813813328742981,
+      "learning_rate": 5.128705692999847e-06,
+      "loss": 0.9515,
+      "step": 11719
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6950210332870483,
+      "learning_rate": 5.127083469556206e-06,
+      "loss": 0.8646,
+      "step": 11720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6613376140594482,
+      "learning_rate": 5.1254614142633064e-06,
+      "loss": 0.9313,
+      "step": 11721
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7989271879196167,
+      "learning_rate": 5.123839527177108e-06,
+      "loss": 0.8764,
+      "step": 11722
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6891721487045288,
+      "learning_rate": 5.122217808353596e-06,
+      "loss": 0.9692,
+      "step": 11723
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5907065868377686,
+      "learning_rate": 5.120596257848716e-06,
+      "loss": 0.9337,
+      "step": 11724
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8290561437606812,
+      "learning_rate": 5.118974875718434e-06,
+      "loss": 0.9892,
+      "step": 11725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8799984455108643,
+      "learning_rate": 5.117353662018692e-06,
+      "loss": 0.9835,
+      "step": 11726
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8811370134353638,
+      "learning_rate": 5.1157326168054374e-06,
+      "loss": 0.9903,
+      "step": 11727
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7325929403305054,
+      "learning_rate": 5.114111740134604e-06,
+      "loss": 0.8456,
+      "step": 11728
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7307196855545044,
+      "learning_rate": 5.112491032062129e-06,
+      "loss": 0.949,
+      "step": 11729
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9088267087936401,
+      "learning_rate": 5.110870492643934e-06,
+      "loss": 1.0396,
+      "step": 11730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.07021427154541,
+      "learning_rate": 5.109250121935938e-06,
+      "loss": 1.0772,
+      "step": 11731
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7922099828720093,
+      "learning_rate": 5.107629919994065e-06,
+      "loss": 0.9785,
+      "step": 11732
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8231827020645142,
+      "learning_rate": 5.106009886874212e-06,
+      "loss": 0.9594,
+      "step": 11733
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6988617181777954,
+      "learning_rate": 5.104390022632292e-06,
+      "loss": 0.8619,
+      "step": 11734
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7075978517532349,
+      "learning_rate": 5.102770327324193e-06,
+      "loss": 0.7801,
+      "step": 11735
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.885631799697876,
+      "learning_rate": 5.101150801005813e-06,
+      "loss": 0.9459,
+      "step": 11736
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.802761435508728,
+      "learning_rate": 5.0995314437330315e-06,
+      "loss": 0.901,
+      "step": 11737
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.91435968875885,
+      "learning_rate": 5.0979122555617345e-06,
+      "loss": 0.9863,
+      "step": 11738
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0569753646850586,
+      "learning_rate": 5.096293236547787e-06,
+      "loss": 0.5421,
+      "step": 11739
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7725480794906616,
+      "learning_rate": 5.0946743867470675e-06,
+      "loss": 0.9424,
+      "step": 11740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9501739740371704,
+      "learning_rate": 5.093055706215428e-06,
+      "loss": 0.9644,
+      "step": 11741
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.853328824043274,
+      "learning_rate": 5.0914371950087325e-06,
+      "loss": 0.8084,
+      "step": 11742
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8748635053634644,
+      "learning_rate": 5.089818853182825e-06,
+      "loss": 0.9379,
+      "step": 11743
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6371344327926636,
+      "learning_rate": 5.088200680793553e-06,
+      "loss": 0.9422,
+      "step": 11744
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9204833507537842,
+      "learning_rate": 5.086582677896758e-06,
+      "loss": 0.9899,
+      "step": 11745
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8673197031021118,
+      "learning_rate": 5.084964844548266e-06,
+      "loss": 1.0478,
+      "step": 11746
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8951619863510132,
+      "learning_rate": 5.083347180803911e-06,
+      "loss": 0.8845,
+      "step": 11747
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8026243448257446,
+      "learning_rate": 5.081729686719507e-06,
+      "loss": 0.8968,
+      "step": 11748
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7013325691223145,
+      "learning_rate": 5.080112362350877e-06,
+      "loss": 0.896,
+      "step": 11749
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6203408241271973,
+      "learning_rate": 5.078495207753824e-06,
+      "loss": 0.9532,
+      "step": 11750
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.763732671737671,
+      "learning_rate": 5.076878222984156e-06,
+      "loss": 0.9439,
+      "step": 11751
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9145323038101196,
+      "learning_rate": 5.075261408097665e-06,
+      "loss": 0.9206,
+      "step": 11752
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8830153942108154,
+      "learning_rate": 5.073644763150148e-06,
+      "loss": 0.8938,
+      "step": 11753
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9114141464233398,
+      "learning_rate": 5.072028288197387e-06,
+      "loss": 0.9421,
+      "step": 11754
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8215630054473877,
+      "learning_rate": 5.070411983295164e-06,
+      "loss": 0.8783,
+      "step": 11755
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6587389707565308,
+      "learning_rate": 5.068795848499257e-06,
+      "loss": 0.8897,
+      "step": 11756
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.73112154006958,
+      "learning_rate": 5.067179883865425e-06,
+      "loss": 0.8657,
+      "step": 11757
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9408777952194214,
+      "learning_rate": 5.0655640894494415e-06,
+      "loss": 0.9885,
+      "step": 11758
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7125475406646729,
+      "learning_rate": 5.0639484653070535e-06,
+      "loss": 0.8915,
+      "step": 11759
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7384669780731201,
+      "learning_rate": 5.0623330114940195e-06,
+      "loss": 0.9165,
+      "step": 11760
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8590176105499268,
+      "learning_rate": 5.060717728066076e-06,
+      "loss": 0.9095,
+      "step": 11761
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7750022411346436,
+      "learning_rate": 5.059102615078972e-06,
+      "loss": 0.9658,
+      "step": 11762
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.625806212425232,
+      "learning_rate": 5.057487672588428e-06,
+      "loss": 0.989,
+      "step": 11763
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6001558303833008,
+      "learning_rate": 5.055872900650185e-06,
+      "loss": 0.9151,
+      "step": 11764
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8214077949523926,
+      "learning_rate": 5.054258299319952e-06,
+      "loss": 0.8698,
+      "step": 11765
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6866214275360107,
+      "learning_rate": 5.052643868653453e-06,
+      "loss": 0.8927,
+      "step": 11766
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.838830828666687,
+      "learning_rate": 5.05102960870639e-06,
+      "loss": 0.9075,
+      "step": 11767
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7893980741500854,
+      "learning_rate": 5.049415519534473e-06,
+      "loss": 1.0027,
+      "step": 11768
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7837446928024292,
+      "learning_rate": 5.047801601193401e-06,
+      "loss": 0.9065,
+      "step": 11769
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7983431816101074,
+      "learning_rate": 5.0461878537388575e-06,
+      "loss": 0.8972,
+      "step": 11770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0220587253570557,
+      "learning_rate": 5.044574277226537e-06,
+      "loss": 0.5722,
+      "step": 11771
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0957280397415161,
+      "learning_rate": 5.042960871712112e-06,
+      "loss": 0.5698,
+      "step": 11772
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.776816487312317,
+      "learning_rate": 5.041347637251264e-06,
+      "loss": 0.9815,
+      "step": 11773
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5999451875686646,
+      "learning_rate": 5.039734573899655e-06,
+      "loss": 0.9808,
+      "step": 11774
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.758184790611267,
+      "learning_rate": 5.038121681712953e-06,
+      "loss": 0.9628,
+      "step": 11775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7792437076568604,
+      "learning_rate": 5.036508960746806e-06,
+      "loss": 0.8652,
+      "step": 11776
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.678625464439392,
+      "learning_rate": 5.034896411056875e-06,
+      "loss": 0.8422,
+      "step": 11777
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.768420934677124,
+      "learning_rate": 5.033284032698797e-06,
+      "loss": 0.8544,
+      "step": 11778
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6039161682128906,
+      "learning_rate": 5.031671825728211e-06,
+      "loss": 0.8034,
+      "step": 11779
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.097721576690674,
+      "learning_rate": 5.0300597902007565e-06,
+      "loss": 0.9364,
+      "step": 11780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9498729705810547,
+      "learning_rate": 5.028447926172052e-06,
+      "loss": 0.8816,
+      "step": 11781
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.829207420349121,
+      "learning_rate": 5.026836233697725e-06,
+      "loss": 0.9438,
+      "step": 11782
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7565690279006958,
+      "learning_rate": 5.025224712833385e-06,
+      "loss": 0.937,
+      "step": 11783
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7777293920516968,
+      "learning_rate": 5.023613363634647e-06,
+      "loss": 0.9255,
+      "step": 11784
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8801199197769165,
+      "learning_rate": 5.0220021861571064e-06,
+      "loss": 0.9024,
+      "step": 11785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8420298099517822,
+      "learning_rate": 5.0203911804563695e-06,
+      "loss": 0.9135,
+      "step": 11786
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6504629850387573,
+      "learning_rate": 5.018780346588019e-06,
+      "loss": 0.8379,
+      "step": 11787
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.661375880241394,
+      "learning_rate": 5.017169684607644e-06,
+      "loss": 0.95,
+      "step": 11788
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8354562520980835,
+      "learning_rate": 5.015559194570825e-06,
+      "loss": 0.9028,
+      "step": 11789
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.717236876487732,
+      "learning_rate": 5.013948876533138e-06,
+      "loss": 0.898,
+      "step": 11790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7261273860931396,
+      "learning_rate": 5.012338730550144e-06,
+      "loss": 0.9238,
+      "step": 11791
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.855106234550476,
+      "learning_rate": 5.01072875667741e-06,
+      "loss": 0.9425,
+      "step": 11792
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8412870168685913,
+      "learning_rate": 5.009118954970488e-06,
+      "loss": 0.9069,
+      "step": 11793
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8281900882720947,
+      "learning_rate": 5.007509325484932e-06,
+      "loss": 0.8786,
+      "step": 11794
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8697614669799805,
+      "learning_rate": 5.005899868276279e-06,
+      "loss": 0.8988,
+      "step": 11795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8560070991516113,
+      "learning_rate": 5.004290583400076e-06,
+      "loss": 0.9358,
+      "step": 11796
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9895504117012024,
+      "learning_rate": 5.002681470911846e-06,
+      "loss": 0.5185,
+      "step": 11797
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.906506061553955,
+      "learning_rate": 5.001072530867119e-06,
+      "loss": 0.9455,
+      "step": 11798
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9184057712554932,
+      "learning_rate": 4.999463763321419e-06,
+      "loss": 0.9266,
+      "step": 11799
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2055280208587646,
+      "learning_rate": 4.997855168330251e-06,
+      "loss": 0.9154,
+      "step": 11800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.797656774520874,
+      "learning_rate": 4.996246745949133e-06,
+      "loss": 0.9651,
+      "step": 11801
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7430778741836548,
+      "learning_rate": 4.994638496233558e-06,
+      "loss": 0.8839,
+      "step": 11802
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.739729642868042,
+      "learning_rate": 4.993030419239031e-06,
+      "loss": 0.8473,
+      "step": 11803
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0470771789550781,
+      "learning_rate": 4.991422515021034e-06,
+      "loss": 0.5337,
+      "step": 11804
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.750593662261963,
+      "learning_rate": 4.989814783635059e-06,
+      "loss": 0.8905,
+      "step": 11805
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7154282331466675,
+      "learning_rate": 4.988207225136577e-06,
+      "loss": 0.8077,
+      "step": 11806
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7973541021347046,
+      "learning_rate": 4.986599839581065e-06,
+      "loss": 0.8583,
+      "step": 11807
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7965013980865479,
+      "learning_rate": 4.9849926270239865e-06,
+      "loss": 0.9084,
+      "step": 11808
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6194044351577759,
+      "learning_rate": 4.983385587520804e-06,
+      "loss": 1.0169,
+      "step": 11809
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6442198753356934,
+      "learning_rate": 4.981778721126975e-06,
+      "loss": 0.9304,
+      "step": 11810
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8897299766540527,
+      "learning_rate": 4.98017202789794e-06,
+      "loss": 0.955,
+      "step": 11811
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8244643211364746,
+      "learning_rate": 4.97856550788915e-06,
+      "loss": 0.8707,
+      "step": 11812
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6431162357330322,
+      "learning_rate": 4.976959161156034e-06,
+      "loss": 0.9616,
+      "step": 11813
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5817203521728516,
+      "learning_rate": 4.975352987754031e-06,
+      "loss": 0.9714,
+      "step": 11814
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6769627332687378,
+      "learning_rate": 4.973746987738555e-06,
+      "loss": 0.8568,
+      "step": 11815
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6666783094406128,
+      "learning_rate": 4.972141161165035e-06,
+      "loss": 0.878,
+      "step": 11816
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6715227365493774,
+      "learning_rate": 4.970535508088874e-06,
+      "loss": 0.9001,
+      "step": 11817
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8833880424499512,
+      "learning_rate": 4.9689300285654886e-06,
+      "loss": 0.9476,
+      "step": 11818
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8556007146835327,
+      "learning_rate": 4.9673247226502684e-06,
+      "loss": 0.9316,
+      "step": 11819
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6133644580841064,
+      "learning_rate": 4.965719590398619e-06,
+      "loss": 0.9271,
+      "step": 11820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8462836742401123,
+      "learning_rate": 4.964114631865919e-06,
+      "loss": 0.8824,
+      "step": 11821
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7447646856307983,
+      "learning_rate": 4.962509847107557e-06,
+      "loss": 0.9709,
+      "step": 11822
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5922834873199463,
+      "learning_rate": 4.960905236178912e-06,
+      "loss": 0.9975,
+      "step": 11823
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7459512948989868,
+      "learning_rate": 4.959300799135348e-06,
+      "loss": 0.8725,
+      "step": 11824
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6210883855819702,
+      "learning_rate": 4.957696536032236e-06,
+      "loss": 0.8809,
+      "step": 11825
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6789504289627075,
+      "learning_rate": 4.9560924469249276e-06,
+      "loss": 0.8933,
+      "step": 11826
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7001127004623413,
+      "learning_rate": 4.954488531868783e-06,
+      "loss": 0.8888,
+      "step": 11827
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8419580459594727,
+      "learning_rate": 4.9528847909191414e-06,
+      "loss": 0.9265,
+      "step": 11828
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6779720783233643,
+      "learning_rate": 4.95128122413135e-06,
+      "loss": 0.9137,
+      "step": 11829
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0809009075164795,
+      "learning_rate": 4.949677831560738e-06,
+      "loss": 0.5939,
+      "step": 11830
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9956091046333313,
+      "learning_rate": 4.94807461326264e-06,
+      "loss": 0.5585,
+      "step": 11831
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6134732961654663,
+      "learning_rate": 4.946471569292372e-06,
+      "loss": 0.8912,
+      "step": 11832
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7498338222503662,
+      "learning_rate": 4.944868699705252e-06,
+      "loss": 0.8553,
+      "step": 11833
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0120846033096313,
+      "learning_rate": 4.943266004556597e-06,
+      "loss": 0.6111,
+      "step": 11834
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8234264850616455,
+      "learning_rate": 4.941663483901703e-06,
+      "loss": 0.9485,
+      "step": 11835
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8544602394104004,
+      "learning_rate": 4.940061137795876e-06,
+      "loss": 0.871,
+      "step": 11836
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8312692642211914,
+      "learning_rate": 4.9384589662944005e-06,
+      "loss": 0.8945,
+      "step": 11837
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.806687593460083,
+      "learning_rate": 4.936856969452572e-06,
+      "loss": 0.9047,
+      "step": 11838
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8743573427200317,
+      "learning_rate": 4.935255147325661e-06,
+      "loss": 0.8792,
+      "step": 11839
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6780905723571777,
+      "learning_rate": 4.933653499968952e-06,
+      "loss": 0.878,
+      "step": 11840
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8195533752441406,
+      "learning_rate": 4.932052027437705e-06,
+      "loss": 0.9269,
+      "step": 11841
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.683165431022644,
+      "learning_rate": 4.930450729787188e-06,
+      "loss": 0.9617,
+      "step": 11842
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.706084966659546,
+      "learning_rate": 4.928849607072654e-06,
+      "loss": 0.8759,
+      "step": 11843
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.837035059928894,
+      "learning_rate": 4.927248659349355e-06,
+      "loss": 1.0567,
+      "step": 11844
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6546556949615479,
+      "learning_rate": 4.9256478866725325e-06,
+      "loss": 0.8836,
+      "step": 11845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7245676517486572,
+      "learning_rate": 4.924047289097426e-06,
+      "loss": 0.9767,
+      "step": 11846
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8457307815551758,
+      "learning_rate": 4.922446866679274e-06,
+      "loss": 0.8356,
+      "step": 11847
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.672777771949768,
+      "learning_rate": 4.920846619473292e-06,
+      "loss": 0.9419,
+      "step": 11848
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7533535957336426,
+      "learning_rate": 4.919246547534709e-06,
+      "loss": 0.8678,
+      "step": 11849
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8325262069702148,
+      "learning_rate": 4.917646650918731e-06,
+      "loss": 0.9241,
+      "step": 11850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8409937620162964,
+      "learning_rate": 4.9160469296805735e-06,
+      "loss": 0.9361,
+      "step": 11851
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6650915145874023,
+      "learning_rate": 4.914447383875433e-06,
+      "loss": 0.932,
+      "step": 11852
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0347445011138916,
+      "learning_rate": 4.912848013558509e-06,
+      "loss": 0.9281,
+      "step": 11853
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.161604881286621,
+      "learning_rate": 4.911248818784984e-06,
+      "loss": 0.9575,
+      "step": 11854
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7821305990219116,
+      "learning_rate": 4.909649799610054e-06,
+      "loss": 0.9008,
+      "step": 11855
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8175138235092163,
+      "learning_rate": 4.908050956088886e-06,
+      "loss": 0.9265,
+      "step": 11856
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9369957447052002,
+      "learning_rate": 4.90645228827666e-06,
+      "loss": 0.926,
+      "step": 11857
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6925907135009766,
+      "learning_rate": 4.904853796228534e-06,
+      "loss": 0.9258,
+      "step": 11858
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7461246252059937,
+      "learning_rate": 4.9032554799996735e-06,
+      "loss": 0.8678,
+      "step": 11859
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.863371729850769,
+      "learning_rate": 4.901657339645226e-06,
+      "loss": 0.9126,
+      "step": 11860
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7510188817977905,
+      "learning_rate": 4.900059375220345e-06,
+      "loss": 0.8835,
+      "step": 11861
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6668410301208496,
+      "learning_rate": 4.8984615867801664e-06,
+      "loss": 0.9261,
+      "step": 11862
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7928352355957031,
+      "learning_rate": 4.896863974379828e-06,
+      "loss": 0.9124,
+      "step": 11863
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6541231870651245,
+      "learning_rate": 4.895266538074461e-06,
+      "loss": 0.8353,
+      "step": 11864
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0209603309631348,
+      "learning_rate": 4.893669277919184e-06,
+      "loss": 0.9066,
+      "step": 11865
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8958548307418823,
+      "learning_rate": 4.892072193969119e-06,
+      "loss": 0.9218,
+      "step": 11866
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8142685890197754,
+      "learning_rate": 4.8904752862793705e-06,
+      "loss": 0.9999,
+      "step": 11867
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7322319746017456,
+      "learning_rate": 4.888878554905051e-06,
+      "loss": 0.9048,
+      "step": 11868
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.686623215675354,
+      "learning_rate": 4.887281999901253e-06,
+      "loss": 0.9522,
+      "step": 11869
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7443772554397583,
+      "learning_rate": 4.885685621323073e-06,
+      "loss": 0.9411,
+      "step": 11870
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0071076154708862,
+      "learning_rate": 4.884089419225593e-06,
+      "loss": 0.5505,
+      "step": 11871
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9278382062911987,
+      "learning_rate": 4.8824933936639e-06,
+      "loss": 0.9589,
+      "step": 11872
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.830845832824707,
+      "learning_rate": 4.880897544693061e-06,
+      "loss": 0.9267,
+      "step": 11873
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7285884618759155,
+      "learning_rate": 4.879301872368152e-06,
+      "loss": 0.9682,
+      "step": 11874
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5627127885818481,
+      "learning_rate": 4.877706376744227e-06,
+      "loss": 0.9164,
+      "step": 11875
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8125560283660889,
+      "learning_rate": 4.8761110578763475e-06,
+      "loss": 0.9179,
+      "step": 11876
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0074329376220703,
+      "learning_rate": 4.874515915819565e-06,
+      "loss": 0.5481,
+      "step": 11877
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.965358018875122,
+      "learning_rate": 4.872920950628918e-06,
+      "loss": 0.9204,
+      "step": 11878
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6264110803604126,
+      "learning_rate": 4.8713261623594495e-06,
+      "loss": 0.895,
+      "step": 11879
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6801484823226929,
+      "learning_rate": 4.869731551066185e-06,
+      "loss": 0.9491,
+      "step": 11880
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9870801568031311,
+      "learning_rate": 4.86813711680416e-06,
+      "loss": 0.5671,
+      "step": 11881
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9008806943893433,
+      "learning_rate": 4.866542859628383e-06,
+      "loss": 0.9284,
+      "step": 11882
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8084572553634644,
+      "learning_rate": 4.864948779593874e-06,
+      "loss": 0.9221,
+      "step": 11883
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7177457809448242,
+      "learning_rate": 4.863354876755637e-06,
+      "loss": 0.9533,
+      "step": 11884
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9433872699737549,
+      "learning_rate": 4.861761151168678e-06,
+      "loss": 0.9925,
+      "step": 11885
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7670843601226807,
+      "learning_rate": 4.860167602887984e-06,
+      "loss": 0.942,
+      "step": 11886
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7366704940795898,
+      "learning_rate": 4.858574231968548e-06,
+      "loss": 0.9313,
+      "step": 11887
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.809315800666809,
+      "learning_rate": 4.8569810384653585e-06,
+      "loss": 0.8818,
+      "step": 11888
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8434861898422241,
+      "learning_rate": 4.855388022433383e-06,
+      "loss": 0.9349,
+      "step": 11889
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8136955499649048,
+      "learning_rate": 4.8537951839275985e-06,
+      "loss": 0.8954,
+      "step": 11890
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6073514223098755,
+      "learning_rate": 4.852202523002964e-06,
+      "loss": 0.8456,
+      "step": 11891
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6485495567321777,
+      "learning_rate": 4.850610039714444e-06,
+      "loss": 0.8092,
+      "step": 11892
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8869320154190063,
+      "learning_rate": 4.849017734116984e-06,
+      "loss": 1.0245,
+      "step": 11893
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7046549320220947,
+      "learning_rate": 4.847425606265537e-06,
+      "loss": 0.9873,
+      "step": 11894
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7491726875305176,
+      "learning_rate": 4.845833656215034e-06,
+      "loss": 0.8055,
+      "step": 11895
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.614396572113037,
+      "learning_rate": 4.8442418840204195e-06,
+      "loss": 0.9741,
+      "step": 11896
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.599818229675293,
+      "learning_rate": 4.842650289736611e-06,
+      "loss": 0.9123,
+      "step": 11897
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7163891792297363,
+      "learning_rate": 4.84105887341854e-06,
+      "loss": 0.9908,
+      "step": 11898
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6395829916000366,
+      "learning_rate": 4.83946763512111e-06,
+      "loss": 0.9483,
+      "step": 11899
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7141355276107788,
+      "learning_rate": 4.837876574899237e-06,
+      "loss": 0.9483,
+      "step": 11900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8705010414123535,
+      "learning_rate": 4.836285692807828e-06,
+      "loss": 0.9262,
+      "step": 11901
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6290096044540405,
+      "learning_rate": 4.834694988901772e-06,
+      "loss": 0.943,
+      "step": 11902
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8338147401809692,
+      "learning_rate": 4.833104463235967e-06,
+      "loss": 0.9686,
+      "step": 11903
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.906256914138794,
+      "learning_rate": 4.83151411586529e-06,
+      "loss": 0.9605,
+      "step": 11904
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6055206060409546,
+      "learning_rate": 4.829923946844627e-06,
+      "loss": 0.91,
+      "step": 11905
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6356450319290161,
+      "learning_rate": 4.828333956228842e-06,
+      "loss": 0.9496,
+      "step": 11906
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.76207435131073,
+      "learning_rate": 4.82674414407281e-06,
+      "loss": 0.9235,
+      "step": 11907
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0050086975097656,
+      "learning_rate": 4.825154510431383e-06,
+      "loss": 1.0014,
+      "step": 11908
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8124254941940308,
+      "learning_rate": 4.823565055359423e-06,
+      "loss": 0.9586,
+      "step": 11909
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6688618659973145,
+      "learning_rate": 4.821975778911768e-06,
+      "loss": 0.9545,
+      "step": 11910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.654943823814392,
+      "learning_rate": 4.82038668114327e-06,
+      "loss": 0.9573,
+      "step": 11911
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8712985515594482,
+      "learning_rate": 4.818797762108754e-06,
+      "loss": 0.9806,
+      "step": 11912
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8634648323059082,
+      "learning_rate": 4.817209021863054e-06,
+      "loss": 0.9103,
+      "step": 11913
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8785457611083984,
+      "learning_rate": 4.815620460460997e-06,
+      "loss": 0.8856,
+      "step": 11914
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6534260511398315,
+      "learning_rate": 4.814032077957392e-06,
+      "loss": 0.9501,
+      "step": 11915
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.777207612991333,
+      "learning_rate": 4.812443874407059e-06,
+      "loss": 0.883,
+      "step": 11916
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2961912155151367,
+      "learning_rate": 4.810855849864792e-06,
+      "loss": 0.8819,
+      "step": 11917
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.073600172996521,
+      "learning_rate": 4.809268004385398e-06,
+      "loss": 0.5544,
+      "step": 11918
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.741358757019043,
+      "learning_rate": 4.807680338023661e-06,
+      "loss": 0.9771,
+      "step": 11919
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7747488021850586,
+      "learning_rate": 4.806092850834373e-06,
+      "loss": 0.9182,
+      "step": 11920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7467007637023926,
+      "learning_rate": 4.804505542872311e-06,
+      "loss": 0.8902,
+      "step": 11921
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.837347149848938,
+      "learning_rate": 4.802918414192254e-06,
+      "loss": 0.9105,
+      "step": 11922
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7731757164001465,
+      "learning_rate": 4.801331464848961e-06,
+      "loss": 0.9517,
+      "step": 11923
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.208500862121582,
+      "learning_rate": 4.7997446948972015e-06,
+      "loss": 0.9847,
+      "step": 11924
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6921570301055908,
+      "learning_rate": 4.798158104391721e-06,
+      "loss": 0.8937,
+      "step": 11925
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8473275899887085,
+      "learning_rate": 4.796571693387278e-06,
+      "loss": 1.0081,
+      "step": 11926
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.70350980758667,
+      "learning_rate": 4.7949854619386086e-06,
+      "loss": 0.9556,
+      "step": 11927
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8426882028579712,
+      "learning_rate": 4.793399410100453e-06,
+      "loss": 0.8594,
+      "step": 11928
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7434173822402954,
+      "learning_rate": 4.791813537927537e-06,
+      "loss": 0.9984,
+      "step": 11929
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.97019362449646,
+      "learning_rate": 4.790227845474588e-06,
+      "loss": 0.9194,
+      "step": 11930
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6433249711990356,
+      "learning_rate": 4.788642332796325e-06,
+      "loss": 0.9674,
+      "step": 11931
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9288480281829834,
+      "learning_rate": 4.7870569999474545e-06,
+      "loss": 0.9144,
+      "step": 11932
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6517329216003418,
+      "learning_rate": 4.78547184698269e-06,
+      "loss": 0.853,
+      "step": 11933
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.876889705657959,
+      "learning_rate": 4.783886873956721e-06,
+      "loss": 0.939,
+      "step": 11934
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5641628503799438,
+      "learning_rate": 4.78230208092425e-06,
+      "loss": 0.9281,
+      "step": 11935
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9833928346633911,
+      "learning_rate": 4.780717467939955e-06,
+      "loss": 0.9072,
+      "step": 11936
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7088240385055542,
+      "learning_rate": 4.779133035058524e-06,
+      "loss": 0.9,
+      "step": 11937
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7135251760482788,
+      "learning_rate": 4.777548782334626e-06,
+      "loss": 0.8758,
+      "step": 11938
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8125405311584473,
+      "learning_rate": 4.7759647098229335e-06,
+      "loss": 0.9986,
+      "step": 11939
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6916090250015259,
+      "learning_rate": 4.774380817578101e-06,
+      "loss": 0.9156,
+      "step": 11940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.758058786392212,
+      "learning_rate": 4.7727971056547915e-06,
+      "loss": 0.958,
+      "step": 11941
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.90669584274292,
+      "learning_rate": 4.771213574107656e-06,
+      "loss": 0.8799,
+      "step": 11942
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7429393529891968,
+      "learning_rate": 4.769630222991329e-06,
+      "loss": 0.8746,
+      "step": 11943
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8275842666625977,
+      "learning_rate": 4.768047052360457e-06,
+      "loss": 0.9892,
+      "step": 11944
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6740633249282837,
+      "learning_rate": 4.7664640622696626e-06,
+      "loss": 0.9438,
+      "step": 11945
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.126016139984131,
+      "learning_rate": 4.7648812527735775e-06,
+      "loss": 0.9547,
+      "step": 11946
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7013310194015503,
+      "learning_rate": 4.7632986239268145e-06,
+      "loss": 0.8546,
+      "step": 11947
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.70201575756073,
+      "learning_rate": 4.7617161757839895e-06,
+      "loss": 0.9254,
+      "step": 11948
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6803851127624512,
+      "learning_rate": 4.760133908399705e-06,
+      "loss": 0.8593,
+      "step": 11949
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8167961835861206,
+      "learning_rate": 4.758551821828564e-06,
+      "loss": 0.9036,
+      "step": 11950
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7445735931396484,
+      "learning_rate": 4.756969916125155e-06,
+      "loss": 0.9293,
+      "step": 11951
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.948959231376648,
+      "learning_rate": 4.755388191344073e-06,
+      "loss": 0.9645,
+      "step": 11952
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.778679609298706,
+      "learning_rate": 4.7538066475398905e-06,
+      "loss": 0.8778,
+      "step": 11953
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7203950881958008,
+      "learning_rate": 4.752225284767185e-06,
+      "loss": 0.8853,
+      "step": 11954
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8872040510177612,
+      "learning_rate": 4.750644103080529e-06,
+      "loss": 0.9498,
+      "step": 11955
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8887821435928345,
+      "learning_rate": 4.7490631025344805e-06,
+      "loss": 0.9093,
+      "step": 11956
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7111982107162476,
+      "learning_rate": 4.747482283183598e-06,
+      "loss": 0.8984,
+      "step": 11957
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.802860975265503,
+      "learning_rate": 4.745901645082426e-06,
+      "loss": 0.952,
+      "step": 11958
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.728653907775879,
+      "learning_rate": 4.744321188285516e-06,
+      "loss": 0.8727,
+      "step": 11959
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8139851093292236,
+      "learning_rate": 4.742740912847397e-06,
+      "loss": 0.9184,
+      "step": 11960
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8228660821914673,
+      "learning_rate": 4.741160818822607e-06,
+      "loss": 0.9789,
+      "step": 11961
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.754248023033142,
+      "learning_rate": 4.739580906265663e-06,
+      "loss": 0.8698,
+      "step": 11962
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7690085172653198,
+      "learning_rate": 4.738001175231091e-06,
+      "loss": 0.9068,
+      "step": 11963
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0565402507781982,
+      "learning_rate": 4.736421625773396e-06,
+      "loss": 0.5449,
+      "step": 11964
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7604234218597412,
+      "learning_rate": 4.734842257947089e-06,
+      "loss": 0.8692,
+      "step": 11965
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6503328084945679,
+      "learning_rate": 4.73326307180667e-06,
+      "loss": 0.9616,
+      "step": 11966
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.69640052318573,
+      "learning_rate": 4.731684067406628e-06,
+      "loss": 0.9819,
+      "step": 11967
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7732524871826172,
+      "learning_rate": 4.730105244801455e-06,
+      "loss": 0.931,
+      "step": 11968
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6727626323699951,
+      "learning_rate": 4.7285266040456255e-06,
+      "loss": 0.8658,
+      "step": 11969
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7181884050369263,
+      "learning_rate": 4.726948145193622e-06,
+      "loss": 0.931,
+      "step": 11970
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.002288818359375,
+      "learning_rate": 4.725369868299904e-06,
+      "loss": 0.8838,
+      "step": 11971
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8661658763885498,
+      "learning_rate": 4.723791773418942e-06,
+      "loss": 0.9205,
+      "step": 11972
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8935105800628662,
+      "learning_rate": 4.722213860605184e-06,
+      "loss": 1.0292,
+      "step": 11973
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.77283775806427,
+      "learning_rate": 4.720636129913086e-06,
+      "loss": 0.8934,
+      "step": 11974
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0479719638824463,
+      "learning_rate": 4.719058581397084e-06,
+      "loss": 0.6187,
+      "step": 11975
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8851581811904907,
+      "learning_rate": 4.717481215111622e-06,
+      "loss": 0.9473,
+      "step": 11976
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7985254526138306,
+      "learning_rate": 4.715904031111124e-06,
+      "loss": 0.9009,
+      "step": 11977
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1421436071395874,
+      "learning_rate": 4.714327029450016e-06,
+      "loss": 0.5794,
+      "step": 11978
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9085428714752197,
+      "learning_rate": 4.712750210182724e-06,
+      "loss": 0.8492,
+      "step": 11979
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.741111159324646,
+      "learning_rate": 4.711173573363647e-06,
+      "loss": 0.8585,
+      "step": 11980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8865671157836914,
+      "learning_rate": 4.7095971190472e-06,
+      "loss": 0.939,
+      "step": 11981
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9194602966308594,
+      "learning_rate": 4.708020847287776e-06,
+      "loss": 0.8701,
+      "step": 11982
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7661586999893188,
+      "learning_rate": 4.706444758139772e-06,
+      "loss": 0.9776,
+      "step": 11983
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8294540643692017,
+      "learning_rate": 4.704868851657569e-06,
+      "loss": 0.8541,
+      "step": 11984
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6922866106033325,
+      "learning_rate": 4.703293127895555e-06,
+      "loss": 0.9931,
+      "step": 11985
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7151439189910889,
+      "learning_rate": 4.701717586908091e-06,
+      "loss": 0.9293,
+      "step": 11986
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8376166820526123,
+      "learning_rate": 4.700142228749561e-06,
+      "loss": 0.9037,
+      "step": 11987
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.656103491783142,
+      "learning_rate": 4.698567053474316e-06,
+      "loss": 0.8858,
+      "step": 11988
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6496952772140503,
+      "learning_rate": 4.6969920611367145e-06,
+      "loss": 0.8841,
+      "step": 11989
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8571563959121704,
+      "learning_rate": 4.6954172517911e-06,
+      "loss": 0.8724,
+      "step": 11990
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7825052738189697,
+      "learning_rate": 4.6938426254918235e-06,
+      "loss": 0.9618,
+      "step": 11991
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9085901975631714,
+      "learning_rate": 4.692268182293211e-06,
+      "loss": 0.9337,
+      "step": 11992
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.733119010925293,
+      "learning_rate": 4.690693922249601e-06,
+      "loss": 0.9527,
+      "step": 11993
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0514566898345947,
+      "learning_rate": 4.689119845415308e-06,
+      "loss": 0.5433,
+      "step": 11994
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.064592123031616,
+      "learning_rate": 4.687545951844656e-06,
+      "loss": 0.9177,
+      "step": 11995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7091907262802124,
+      "learning_rate": 4.685972241591956e-06,
+      "loss": 0.9259,
+      "step": 11996
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.891315221786499,
+      "learning_rate": 4.684398714711507e-06,
+      "loss": 0.907,
+      "step": 11997
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.800525426864624,
+      "learning_rate": 4.6828253712576125e-06,
+      "loss": 0.8894,
+      "step": 11998
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.640904426574707,
+      "learning_rate": 4.681252211284557e-06,
+      "loss": 0.8976,
+      "step": 11999
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7103461027145386,
+      "learning_rate": 4.679679234846636e-06,
+      "loss": 0.8161,
+      "step": 12000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7877203226089478,
+      "learning_rate": 4.678106441998118e-06,
+      "loss": 0.9406,
+      "step": 12001
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7582924365997314,
+      "learning_rate": 4.676533832793284e-06,
+      "loss": 0.9277,
+      "step": 12002
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.628161072731018,
+      "learning_rate": 4.674961407286393e-06,
+      "loss": 0.9072,
+      "step": 12003
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6190695762634277,
+      "learning_rate": 4.673389165531714e-06,
+      "loss": 0.8787,
+      "step": 12004
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.80056893825531,
+      "learning_rate": 4.6718171075834916e-06,
+      "loss": 0.9068,
+      "step": 12005
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0112545490264893,
+      "learning_rate": 4.67024523349598e-06,
+      "loss": 0.9393,
+      "step": 12006
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7300282716751099,
+      "learning_rate": 4.668673543323414e-06,
+      "loss": 0.948,
+      "step": 12007
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6679729223251343,
+      "learning_rate": 4.6671020371200324e-06,
+      "loss": 0.9344,
+      "step": 12008
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.790862798690796,
+      "learning_rate": 4.665530714940067e-06,
+      "loss": 0.9331,
+      "step": 12009
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7741063833236694,
+      "learning_rate": 4.663959576837729e-06,
+      "loss": 0.9388,
+      "step": 12010
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9999828934669495,
+      "learning_rate": 4.662388622867246e-06,
+      "loss": 0.582,
+      "step": 12011
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6702301502227783,
+      "learning_rate": 4.6608178530828176e-06,
+      "loss": 1.0275,
+      "step": 12012
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.947562575340271,
+      "learning_rate": 4.6592472675386535e-06,
+      "loss": 0.9604,
+      "step": 12013
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6805510520935059,
+      "learning_rate": 4.657676866288945e-06,
+      "loss": 0.9143,
+      "step": 12014
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.771214246749878,
+      "learning_rate": 4.656106649387887e-06,
+      "loss": 0.8037,
+      "step": 12015
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7808375358581543,
+      "learning_rate": 4.654536616889658e-06,
+      "loss": 0.9087,
+      "step": 12016
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.641385555267334,
+      "learning_rate": 4.652966768848442e-06,
+      "loss": 0.7922,
+      "step": 12017
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.814820408821106,
+      "learning_rate": 4.651397105318402e-06,
+      "loss": 0.8584,
+      "step": 12018
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7147387266159058,
+      "learning_rate": 4.649827626353709e-06,
+      "loss": 0.9931,
+      "step": 12019
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6887844800949097,
+      "learning_rate": 4.648258332008523e-06,
+      "loss": 0.9519,
+      "step": 12020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7287869453430176,
+      "learning_rate": 4.646689222336988e-06,
+      "loss": 0.8996,
+      "step": 12021
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.738333821296692,
+      "learning_rate": 4.645120297393257e-06,
+      "loss": 0.8911,
+      "step": 12022
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7822545766830444,
+      "learning_rate": 4.643551557231464e-06,
+      "loss": 0.9667,
+      "step": 12023
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0141637325286865,
+      "learning_rate": 4.641983001905747e-06,
+      "loss": 1.0143,
+      "step": 12024
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9001177549362183,
+      "learning_rate": 4.640414631470226e-06,
+      "loss": 1.0146,
+      "step": 12025
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7878799438476562,
+      "learning_rate": 4.638846445979028e-06,
+      "loss": 0.9018,
+      "step": 12026
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6512539386749268,
+      "learning_rate": 4.6372784454862605e-06,
+      "loss": 0.9063,
+      "step": 12027
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.748706340789795,
+      "learning_rate": 4.635710630046037e-06,
+      "loss": 0.903,
+      "step": 12028
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6269668340682983,
+      "learning_rate": 4.634142999712451e-06,
+      "loss": 0.9733,
+      "step": 12029
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6721644401550293,
+      "learning_rate": 4.632575554539605e-06,
+      "loss": 0.9289,
+      "step": 12030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.604628086090088,
+      "learning_rate": 4.6310082945815805e-06,
+      "loss": 0.8758,
+      "step": 12031
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8042831420898438,
+      "learning_rate": 4.6294412198924625e-06,
+      "loss": 0.8787,
+      "step": 12032
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7345798015594482,
+      "learning_rate": 4.627874330526328e-06,
+      "loss": 0.8784,
+      "step": 12033
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7797752618789673,
+      "learning_rate": 4.626307626537241e-06,
+      "loss": 0.9824,
+      "step": 12034
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7025492191314697,
+      "learning_rate": 4.62474110797927e-06,
+      "loss": 0.9857,
+      "step": 12035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6464751958847046,
+      "learning_rate": 4.623174774906464e-06,
+      "loss": 0.8602,
+      "step": 12036
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8642346858978271,
+      "learning_rate": 4.621608627372883e-06,
+      "loss": 0.9015,
+      "step": 12037
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6643532514572144,
+      "learning_rate": 4.620042665432559e-06,
+      "loss": 0.8842,
+      "step": 12038
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.704205870628357,
+      "learning_rate": 4.618476889139538e-06,
+      "loss": 0.8622,
+      "step": 12039
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8148225545883179,
+      "learning_rate": 4.616911298547845e-06,
+      "loss": 0.9606,
+      "step": 12040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9357045888900757,
+      "learning_rate": 4.615345893711508e-06,
+      "loss": 0.9485,
+      "step": 12041
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.62363862991333,
+      "learning_rate": 4.61378067468454e-06,
+      "loss": 0.9628,
+      "step": 12042
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8985458612442017,
+      "learning_rate": 4.612215641520957e-06,
+      "loss": 0.9177,
+      "step": 12043
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9085721969604492,
+      "learning_rate": 4.6106507942747595e-06,
+      "loss": 0.9103,
+      "step": 12044
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8194293975830078,
+      "learning_rate": 4.609086132999949e-06,
+      "loss": 0.966,
+      "step": 12045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7097723484039307,
+      "learning_rate": 4.60752165775052e-06,
+      "loss": 0.9256,
+      "step": 12046
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.81070876121521,
+      "learning_rate": 4.605957368580453e-06,
+      "loss": 0.8806,
+      "step": 12047
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.984039306640625,
+      "learning_rate": 4.6043932655437316e-06,
+      "loss": 0.9777,
+      "step": 12048
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7470107078552246,
+      "learning_rate": 4.6028293486943234e-06,
+      "loss": 0.8602,
+      "step": 12049
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6861248016357422,
+      "learning_rate": 4.6012656180862024e-06,
+      "loss": 0.9372,
+      "step": 12050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8057801723480225,
+      "learning_rate": 4.59970207377332e-06,
+      "loss": 0.9022,
+      "step": 12051
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.058417558670044,
+      "learning_rate": 4.598138715809634e-06,
+      "loss": 0.5787,
+      "step": 12052
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7225613594055176,
+      "learning_rate": 4.596575544249091e-06,
+      "loss": 0.9108,
+      "step": 12053
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.819343090057373,
+      "learning_rate": 4.595012559145636e-06,
+      "loss": 0.8622,
+      "step": 12054
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8904342651367188,
+      "learning_rate": 4.5934497605531955e-06,
+      "loss": 0.9406,
+      "step": 12055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7186291217803955,
+      "learning_rate": 4.5918871485257055e-06,
+      "loss": 0.9006,
+      "step": 12056
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7375426292419434,
+      "learning_rate": 4.5903247231170785e-06,
+      "loss": 0.9171,
+      "step": 12057
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8099197149276733,
+      "learning_rate": 4.588762484381238e-06,
+      "loss": 0.9935,
+      "step": 12058
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7822587490081787,
+      "learning_rate": 4.587200432372085e-06,
+      "loss": 0.8809,
+      "step": 12059
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7568249702453613,
+      "learning_rate": 4.5856385671435285e-06,
+      "loss": 0.9707,
+      "step": 12060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.298645257949829,
+      "learning_rate": 4.584076888749458e-06,
+      "loss": 0.9125,
+      "step": 12061
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9083985090255737,
+      "learning_rate": 4.582515397243764e-06,
+      "loss": 0.9895,
+      "step": 12062
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7404065132141113,
+      "learning_rate": 4.580954092680334e-06,
+      "loss": 0.9085,
+      "step": 12063
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6660234928131104,
+      "learning_rate": 4.5793929751130384e-06,
+      "loss": 0.8512,
+      "step": 12064
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8452751636505127,
+      "learning_rate": 4.577832044595752e-06,
+      "loss": 0.8905,
+      "step": 12065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0838899612426758,
+      "learning_rate": 4.576271301182332e-06,
+      "loss": 0.5105,
+      "step": 12066
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6553512811660767,
+      "learning_rate": 4.574710744926643e-06,
+      "loss": 0.914,
+      "step": 12067
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6799341440200806,
+      "learning_rate": 4.573150375882527e-06,
+      "loss": 0.8506,
+      "step": 12068
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.718163251876831,
+      "learning_rate": 4.571590194103836e-06,
+      "loss": 1.0113,
+      "step": 12069
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6394116878509521,
+      "learning_rate": 4.570030199644401e-06,
+      "loss": 0.9512,
+      "step": 12070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8188966512680054,
+      "learning_rate": 4.568470392558059e-06,
+      "loss": 1.0151,
+      "step": 12071
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1250067949295044,
+      "learning_rate": 4.566910772898627e-06,
+      "loss": 0.5999,
+      "step": 12072
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6914042234420776,
+      "learning_rate": 4.565351340719928e-06,
+      "loss": 0.8688,
+      "step": 12073
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8462719917297363,
+      "learning_rate": 4.563792096075777e-06,
+      "loss": 0.9154,
+      "step": 12074
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6726685762405396,
+      "learning_rate": 4.562233039019971e-06,
+      "loss": 0.9769,
+      "step": 12075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6645078659057617,
+      "learning_rate": 4.560674169606317e-06,
+      "loss": 0.9177,
+      "step": 12076
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.757951021194458,
+      "learning_rate": 4.5591154878886e-06,
+      "loss": 0.9182,
+      "step": 12077
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.641312837600708,
+      "learning_rate": 4.5575569939206125e-06,
+      "loss": 0.8785,
+      "step": 12078
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.945239782333374,
+      "learning_rate": 4.555998687756127e-06,
+      "loss": 0.8914,
+      "step": 12079
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5375421047210693,
+      "learning_rate": 4.5544405694489224e-06,
+      "loss": 0.8649,
+      "step": 12080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.819831371307373,
+      "learning_rate": 4.5528826390527594e-06,
+      "loss": 0.9037,
+      "step": 12081
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.689393401145935,
+      "learning_rate": 4.551324896621403e-06,
+      "loss": 0.8749,
+      "step": 12082
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8362040519714355,
+      "learning_rate": 4.549767342208602e-06,
+      "loss": 0.9851,
+      "step": 12083
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8308439254760742,
+      "learning_rate": 4.548209975868109e-06,
+      "loss": 0.9242,
+      "step": 12084
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8101271390914917,
+      "learning_rate": 4.546652797653656e-06,
+      "loss": 0.8712,
+      "step": 12085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9170933961868286,
+      "learning_rate": 4.5450958076189825e-06,
+      "loss": 0.8969,
+      "step": 12086
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8410924673080444,
+      "learning_rate": 4.543539005817818e-06,
+      "loss": 0.9218,
+      "step": 12087
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8555163145065308,
+      "learning_rate": 4.541982392303876e-06,
+      "loss": 1.0267,
+      "step": 12088
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7057043313980103,
+      "learning_rate": 4.540425967130881e-06,
+      "loss": 0.9537,
+      "step": 12089
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.855972409248352,
+      "learning_rate": 4.53886973035253e-06,
+      "loss": 0.8994,
+      "step": 12090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5824469327926636,
+      "learning_rate": 4.537313682022534e-06,
+      "loss": 0.8573,
+      "step": 12091
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7325880527496338,
+      "learning_rate": 4.53575782219458e-06,
+      "loss": 0.8998,
+      "step": 12092
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.878062129020691,
+      "learning_rate": 4.534202150922362e-06,
+      "loss": 0.9307,
+      "step": 12093
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8051809072494507,
+      "learning_rate": 4.532646668259557e-06,
+      "loss": 0.9662,
+      "step": 12094
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7402708530426025,
+      "learning_rate": 4.531091374259848e-06,
+      "loss": 0.9049,
+      "step": 12095
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7610024213790894,
+      "learning_rate": 4.529536268976893e-06,
+      "loss": 0.8916,
+      "step": 12096
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.026312828063965,
+      "learning_rate": 4.5279813524643644e-06,
+      "loss": 0.8098,
+      "step": 12097
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8060368299484253,
+      "learning_rate": 4.526426624775911e-06,
+      "loss": 0.9179,
+      "step": 12098
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9569159746170044,
+      "learning_rate": 4.524872085965184e-06,
+      "loss": 0.9601,
+      "step": 12099
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.655694603919983,
+      "learning_rate": 4.523317736085832e-06,
+      "loss": 0.9731,
+      "step": 12100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8205318450927734,
+      "learning_rate": 4.521763575191482e-06,
+      "loss": 0.9043,
+      "step": 12101
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8622970581054688,
+      "learning_rate": 4.520209603335772e-06,
+      "loss": 0.8329,
+      "step": 12102
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6752930879592896,
+      "learning_rate": 4.518655820572317e-06,
+      "loss": 0.9472,
+      "step": 12103
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7246054410934448,
+      "learning_rate": 4.517102226954744e-06,
+      "loss": 0.9573,
+      "step": 12104
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8553173542022705,
+      "learning_rate": 4.515548822536652e-06,
+      "loss": 0.9446,
+      "step": 12105
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.625590205192566,
+      "learning_rate": 4.513995607371654e-06,
+      "loss": 0.8991,
+      "step": 12106
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8012144565582275,
+      "learning_rate": 4.512442581513339e-06,
+      "loss": 0.9701,
+      "step": 12107
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.044556975364685,
+      "learning_rate": 4.510889745015306e-06,
+      "loss": 0.5489,
+      "step": 12108
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9414899349212646,
+      "learning_rate": 4.50933709793113e-06,
+      "loss": 0.8543,
+      "step": 12109
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8772963285446167,
+      "learning_rate": 4.507784640314393e-06,
+      "loss": 0.9221,
+      "step": 12110
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8219832181930542,
+      "learning_rate": 4.50623237221867e-06,
+      "loss": 0.9939,
+      "step": 12111
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8231613636016846,
+      "learning_rate": 4.50468029369752e-06,
+      "loss": 0.9314,
+      "step": 12112
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.864905834197998,
+      "learning_rate": 4.5031284048045045e-06,
+      "loss": 1.0244,
+      "step": 12113
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6909316778182983,
+      "learning_rate": 4.5015767055931695e-06,
+      "loss": 0.9142,
+      "step": 12114
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4885127544403076,
+      "learning_rate": 4.500025196117066e-06,
+      "loss": 0.8463,
+      "step": 12115
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.721638798713684,
+      "learning_rate": 4.498473876429727e-06,
+      "loss": 0.9042,
+      "step": 12116
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8825305700302124,
+      "learning_rate": 4.49692274658469e-06,
+      "loss": 0.9038,
+      "step": 12117
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5840251445770264,
+      "learning_rate": 4.495371806635469e-06,
+      "loss": 0.9152,
+      "step": 12118
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9720197916030884,
+      "learning_rate": 4.493821056635598e-06,
+      "loss": 0.9172,
+      "step": 12119
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.777827501296997,
+      "learning_rate": 4.492270496638578e-06,
+      "loss": 0.8906,
+      "step": 12120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.783775806427002,
+      "learning_rate": 4.490720126697921e-06,
+      "loss": 0.9353,
+      "step": 12121
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.855164647102356,
+      "learning_rate": 4.489169946867119e-06,
+      "loss": 0.8433,
+      "step": 12122
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7153263092041016,
+      "learning_rate": 4.487619957199672e-06,
+      "loss": 0.8493,
+      "step": 12123
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6919550895690918,
+      "learning_rate": 4.486070157749059e-06,
+      "loss": 0.8684,
+      "step": 12124
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6777186393737793,
+      "learning_rate": 4.484520548568766e-06,
+      "loss": 0.9316,
+      "step": 12125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.784044623374939,
+      "learning_rate": 4.482971129712258e-06,
+      "loss": 0.9159,
+      "step": 12126
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8387353420257568,
+      "learning_rate": 4.481421901233005e-06,
+      "loss": 0.9539,
+      "step": 12127
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.764176845550537,
+      "learning_rate": 4.4798728631844715e-06,
+      "loss": 0.9344,
+      "step": 12128
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6732069253921509,
+      "learning_rate": 4.478324015620101e-06,
+      "loss": 0.9919,
+      "step": 12129
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.015463948249817,
+      "learning_rate": 4.476775358593348e-06,
+      "loss": 0.5797,
+      "step": 12130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6946240663528442,
+      "learning_rate": 4.475226892157646e-06,
+      "loss": 0.9111,
+      "step": 12131
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6743507385253906,
+      "learning_rate": 4.473678616366434e-06,
+      "loss": 0.9501,
+      "step": 12132
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6993016004562378,
+      "learning_rate": 4.472130531273132e-06,
+      "loss": 0.9076,
+      "step": 12133
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6751279830932617,
+      "learning_rate": 4.470582636931168e-06,
+      "loss": 0.8764,
+      "step": 12134
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8724472522735596,
+      "learning_rate": 4.469034933393948e-06,
+      "loss": 0.9181,
+      "step": 12135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0837968587875366,
+      "learning_rate": 4.467487420714885e-06,
+      "loss": 0.5223,
+      "step": 12136
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6412482261657715,
+      "learning_rate": 4.4659400989473744e-06,
+      "loss": 0.9073,
+      "step": 12137
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7140542268753052,
+      "learning_rate": 4.464392968144815e-06,
+      "loss": 0.9707,
+      "step": 12138
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7308902740478516,
+      "learning_rate": 4.462846028360588e-06,
+      "loss": 1.0517,
+      "step": 12139
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6907967329025269,
+      "learning_rate": 4.461299279648077e-06,
+      "loss": 0.9189,
+      "step": 12140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8545448780059814,
+      "learning_rate": 4.45975272206066e-06,
+      "loss": 0.9468,
+      "step": 12141
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7968237400054932,
+      "learning_rate": 4.4582063556516955e-06,
+      "loss": 0.923,
+      "step": 12142
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.546665072441101,
+      "learning_rate": 4.456660180474554e-06,
+      "loss": 0.8514,
+      "step": 12143
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7471485137939453,
+      "learning_rate": 4.455114196582582e-06,
+      "loss": 0.9956,
+      "step": 12144
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7082730531692505,
+      "learning_rate": 4.453568404029131e-06,
+      "loss": 0.9179,
+      "step": 12145
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9506896734237671,
+      "learning_rate": 4.452022802867541e-06,
+      "loss": 0.5619,
+      "step": 12146
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.96599543094635,
+      "learning_rate": 4.450477393151148e-06,
+      "loss": 0.9262,
+      "step": 12147
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.548830270767212,
+      "learning_rate": 4.448932174933274e-06,
+      "loss": 1.0198,
+      "step": 12148
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0511407852172852,
+      "learning_rate": 4.447387148267249e-06,
+      "loss": 0.5041,
+      "step": 12149
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7605023384094238,
+      "learning_rate": 4.44584231320638e-06,
+      "loss": 0.9343,
+      "step": 12150
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.782924771308899,
+      "learning_rate": 4.444297669803981e-06,
+      "loss": 0.9455,
+      "step": 12151
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3206787109375,
+      "learning_rate": 4.442753218113346e-06,
+      "loss": 0.8799,
+      "step": 12152
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7880959510803223,
+      "learning_rate": 4.441208958187774e-06,
+      "loss": 0.9132,
+      "step": 12153
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.594754695892334,
+      "learning_rate": 4.4396648900805574e-06,
+      "loss": 0.8757,
+      "step": 12154
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7561460733413696,
+      "learning_rate": 4.4381210138449685e-06,
+      "loss": 0.9267,
+      "step": 12155
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.897371768951416,
+      "learning_rate": 4.436577329534291e-06,
+      "loss": 0.9138,
+      "step": 12156
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6250697374343872,
+      "learning_rate": 4.435033837201785e-06,
+      "loss": 0.9545,
+      "step": 12157
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0469470024108887,
+      "learning_rate": 4.433490536900721e-06,
+      "loss": 0.553,
+      "step": 12158
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6664888858795166,
+      "learning_rate": 4.4319474286843446e-06,
+      "loss": 0.9255,
+      "step": 12159
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5987811088562012,
+      "learning_rate": 4.4304045126059126e-06,
+      "loss": 0.8882,
+      "step": 12160
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.665635585784912,
+      "learning_rate": 4.428861788718659e-06,
+      "loss": 0.8743,
+      "step": 12161
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5753731727600098,
+      "learning_rate": 4.427319257075827e-06,
+      "loss": 0.8321,
+      "step": 12162
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6694729328155518,
+      "learning_rate": 4.425776917730636e-06,
+      "loss": 0.8196,
+      "step": 12163
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8965317010879517,
+      "learning_rate": 4.424234770736314e-06,
+      "loss": 0.9573,
+      "step": 12164
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8113521337509155,
+      "learning_rate": 4.422692816146078e-06,
+      "loss": 0.9978,
+      "step": 12165
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0293681621551514,
+      "learning_rate": 4.42115105401313e-06,
+      "loss": 0.9883,
+      "step": 12166
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8602954149246216,
+      "learning_rate": 4.419609484390678e-06,
+      "loss": 0.8724,
+      "step": 12167
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7014737129211426,
+      "learning_rate": 4.418068107331911e-06,
+      "loss": 0.8917,
+      "step": 12168
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.682297945022583,
+      "learning_rate": 4.416526922890024e-06,
+      "loss": 0.9549,
+      "step": 12169
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8611048460006714,
+      "learning_rate": 4.414985931118194e-06,
+      "loss": 0.9257,
+      "step": 12170
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8226057291030884,
+      "learning_rate": 4.413445132069601e-06,
+      "loss": 0.8677,
+      "step": 12171
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.738637089729309,
+      "learning_rate": 4.411904525797408e-06,
+      "loss": 0.911,
+      "step": 12172
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7301138639450073,
+      "learning_rate": 4.4103641123547836e-06,
+      "loss": 0.9653,
+      "step": 12173
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8996912240982056,
+      "learning_rate": 4.4088238917948765e-06,
+      "loss": 0.9327,
+      "step": 12174
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7330734729766846,
+      "learning_rate": 4.4072838641708415e-06,
+      "loss": 0.9284,
+      "step": 12175
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5642669200897217,
+      "learning_rate": 4.405744029535815e-06,
+      "loss": 0.8801,
+      "step": 12176
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9499053955078125,
+      "learning_rate": 4.404204387942934e-06,
+      "loss": 0.9138,
+      "step": 12177
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.825777292251587,
+      "learning_rate": 4.402664939445333e-06,
+      "loss": 0.9849,
+      "step": 12178
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0183806419372559,
+      "learning_rate": 4.401125684096124e-06,
+      "loss": 0.5204,
+      "step": 12179
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7793660163879395,
+      "learning_rate": 4.399586621948433e-06,
+      "loss": 0.9181,
+      "step": 12180
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6689112186431885,
+      "learning_rate": 4.39804775305536e-06,
+      "loss": 0.8612,
+      "step": 12181
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.710805892944336,
+      "learning_rate": 4.3965090774700135e-06,
+      "loss": 0.9253,
+      "step": 12182
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7184252738952637,
+      "learning_rate": 4.394970595245483e-06,
+      "loss": 0.9507,
+      "step": 12183
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8182810544967651,
+      "learning_rate": 4.3934323064348636e-06,
+      "loss": 0.9591,
+      "step": 12184
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6814923286437988,
+      "learning_rate": 4.391894211091227e-06,
+      "loss": 0.9022,
+      "step": 12185
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8337434530258179,
+      "learning_rate": 4.3903563092676626e-06,
+      "loss": 0.9207,
+      "step": 12186
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8550876379013062,
+      "learning_rate": 4.388818601017228e-06,
+      "loss": 0.8593,
+      "step": 12187
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.964364230632782,
+      "learning_rate": 4.387281086392994e-06,
+      "loss": 0.5373,
+      "step": 12188
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5979536771774292,
+      "learning_rate": 4.385743765448006e-06,
+      "loss": 0.8781,
+      "step": 12189
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.090733766555786,
+      "learning_rate": 4.384206638235322e-06,
+      "loss": 0.8764,
+      "step": 12190
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.638323426246643,
+      "learning_rate": 4.382669704807977e-06,
+      "loss": 0.9201,
+      "step": 12191
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8556439876556396,
+      "learning_rate": 4.3811329652190126e-06,
+      "loss": 0.9341,
+      "step": 12192
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7892338037490845,
+      "learning_rate": 4.37959641952145e-06,
+      "loss": 0.9075,
+      "step": 12193
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.609830379486084,
+      "learning_rate": 4.3780600677683145e-06,
+      "loss": 0.8846,
+      "step": 12194
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7032337188720703,
+      "learning_rate": 4.376523910012627e-06,
+      "loss": 0.9057,
+      "step": 12195
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7046489715576172,
+      "learning_rate": 4.3749879463073854e-06,
+      "loss": 0.9698,
+      "step": 12196
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6963069438934326,
+      "learning_rate": 4.373452176705601e-06,
+      "loss": 0.9395,
+      "step": 12197
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8119169473648071,
+      "learning_rate": 4.371916601260262e-06,
+      "loss": 0.9465,
+      "step": 12198
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6900408267974854,
+      "learning_rate": 4.370381220024362e-06,
+      "loss": 1.0242,
+      "step": 12199
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9985848665237427,
+      "learning_rate": 4.368846033050879e-06,
+      "loss": 0.8655,
+      "step": 12200
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.646001935005188,
+      "learning_rate": 4.367311040392791e-06,
+      "loss": 0.9745,
+      "step": 12201
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6562753915786743,
+      "learning_rate": 4.365776242103062e-06,
+      "loss": 0.9819,
+      "step": 12202
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6765224933624268,
+      "learning_rate": 4.364241638234659e-06,
+      "loss": 0.9175,
+      "step": 12203
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.613690733909607,
+      "learning_rate": 4.362707228840531e-06,
+      "loss": 0.8811,
+      "step": 12204
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6638638973236084,
+      "learning_rate": 4.36117301397363e-06,
+      "loss": 0.8766,
+      "step": 12205
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0590672492980957,
+      "learning_rate": 4.3596389936869e-06,
+      "loss": 0.5949,
+      "step": 12206
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7519869804382324,
+      "learning_rate": 4.358105168033269e-06,
+      "loss": 0.9128,
+      "step": 12207
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.703728437423706,
+      "learning_rate": 4.3565715370656725e-06,
+      "loss": 0.9564,
+      "step": 12208
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8182377815246582,
+      "learning_rate": 4.355038100837023e-06,
+      "loss": 0.9266,
+      "step": 12209
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6696603298187256,
+      "learning_rate": 4.353504859400246e-06,
+      "loss": 0.86,
+      "step": 12210
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9939159154891968,
+      "learning_rate": 4.351971812808239e-06,
+      "loss": 0.513,
+      "step": 12211
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8564430475234985,
+      "learning_rate": 4.350438961113911e-06,
+      "loss": 1.0003,
+      "step": 12212
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.786815881729126,
+      "learning_rate": 4.348906304370148e-06,
+      "loss": 0.8335,
+      "step": 12213
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6922428607940674,
+      "learning_rate": 4.3473738426298485e-06,
+      "loss": 0.9523,
+      "step": 12214
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7936313152313232,
+      "learning_rate": 4.345841575945884e-06,
+      "loss": 0.9573,
+      "step": 12215
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7848342657089233,
+      "learning_rate": 4.344309504371135e-06,
+      "loss": 0.8772,
+      "step": 12216
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7737482786178589,
+      "learning_rate": 4.342777627958463e-06,
+      "loss": 1.0202,
+      "step": 12217
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8887823820114136,
+      "learning_rate": 4.341245946760733e-06,
+      "loss": 0.9995,
+      "step": 12218
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6565766334533691,
+      "learning_rate": 4.339714460830802e-06,
+      "loss": 0.904,
+      "step": 12219
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8045833110809326,
+      "learning_rate": 4.338183170221508e-06,
+      "loss": 0.939,
+      "step": 12220
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.729711651802063,
+      "learning_rate": 4.336652074985703e-06,
+      "loss": 0.91,
+      "step": 12221
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0387065410614014,
+      "learning_rate": 4.3351211751762104e-06,
+      "loss": 0.5337,
+      "step": 12222
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7121163606643677,
+      "learning_rate": 4.333590470845866e-06,
+      "loss": 0.872,
+      "step": 12223
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7705730199813843,
+      "learning_rate": 4.332059962047481e-06,
+      "loss": 0.8938,
+      "step": 12224
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8683247566223145,
+      "learning_rate": 4.330529648833879e-06,
+      "loss": 0.9195,
+      "step": 12225
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6935882568359375,
+      "learning_rate": 4.3289995312578585e-06,
+      "loss": 0.942,
+      "step": 12226
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.761215090751648,
+      "learning_rate": 4.327469609372224e-06,
+      "loss": 0.9786,
+      "step": 12227
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6463446617126465,
+      "learning_rate": 4.3259398832297665e-06,
+      "loss": 0.8471,
+      "step": 12228
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6273789405822754,
+      "learning_rate": 4.324410352883277e-06,
+      "loss": 0.9037,
+      "step": 12229
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8586559295654297,
+      "learning_rate": 4.322881018385527e-06,
+      "loss": 0.9008,
+      "step": 12230
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6294889450073242,
+      "learning_rate": 4.321351879789296e-06,
+      "loss": 0.8492,
+      "step": 12231
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7973731756210327,
+      "learning_rate": 4.3198229371473535e-06,
+      "loss": 0.9872,
+      "step": 12232
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8881688117980957,
+      "learning_rate": 4.31829419051245e-06,
+      "loss": 0.9911,
+      "step": 12233
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6258580684661865,
+      "learning_rate": 4.316765639937346e-06,
+      "loss": 0.9493,
+      "step": 12234
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.773603081703186,
+      "learning_rate": 4.31523728547478e-06,
+      "loss": 0.9609,
+      "step": 12235
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9875374436378479,
+      "learning_rate": 4.3137091271775e-06,
+      "loss": 0.5411,
+      "step": 12236
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6680865287780762,
+      "learning_rate": 4.3121811650982306e-06,
+      "loss": 0.859,
+      "step": 12237
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8523112535476685,
+      "learning_rate": 4.310653399289705e-06,
+      "loss": 0.9491,
+      "step": 12238
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8099464178085327,
+      "learning_rate": 4.309125829804633e-06,
+      "loss": 0.8759,
+      "step": 12239
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7375110387802124,
+      "learning_rate": 4.307598456695736e-06,
+      "loss": 0.8817,
+      "step": 12240
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2342143058776855,
+      "learning_rate": 4.306071280015713e-06,
+      "loss": 0.9653,
+      "step": 12241
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0205042362213135,
+      "learning_rate": 4.304544299817263e-06,
+      "loss": 0.93,
+      "step": 12242
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7261152267456055,
+      "learning_rate": 4.303017516153083e-06,
+      "loss": 0.9328,
+      "step": 12243
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8399591445922852,
+      "learning_rate": 4.3014909290758525e-06,
+      "loss": 0.891,
+      "step": 12244
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.704454779624939,
+      "learning_rate": 4.299964538638255e-06,
+      "loss": 0.8718,
+      "step": 12245
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7072902917861938,
+      "learning_rate": 4.298438344892954e-06,
+      "loss": 0.9065,
+      "step": 12246
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7964297533035278,
+      "learning_rate": 4.296912347892625e-06,
+      "loss": 0.8728,
+      "step": 12247
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6738181114196777,
+      "learning_rate": 4.295386547689913e-06,
+      "loss": 0.8801,
+      "step": 12248
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.88579523563385,
+      "learning_rate": 4.293860944337482e-06,
+      "loss": 0.9585,
+      "step": 12249
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.784139633178711,
+      "learning_rate": 4.2923355378879675e-06,
+      "loss": 0.9878,
+      "step": 12250
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8959250450134277,
+      "learning_rate": 4.290810328394008e-06,
+      "loss": 0.8566,
+      "step": 12251
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6999598741531372,
+      "learning_rate": 4.289285315908237e-06,
+      "loss": 0.8978,
+      "step": 12252
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2608280181884766,
+      "learning_rate": 4.2877605004832816e-06,
+      "loss": 0.9765,
+      "step": 12253
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7862107753753662,
+      "learning_rate": 4.2862358821717496e-06,
+      "loss": 0.8848,
+      "step": 12254
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8780516386032104,
+      "learning_rate": 4.284711461026262e-06,
+      "loss": 0.9402,
+      "step": 12255
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.730836272239685,
+      "learning_rate": 4.283187237099412e-06,
+      "loss": 0.9146,
+      "step": 12256
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8147203922271729,
+      "learning_rate": 4.281663210443805e-06,
+      "loss": 0.9374,
+      "step": 12257
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9632763862609863,
+      "learning_rate": 4.280139381112024e-06,
+      "loss": 0.9564,
+      "step": 12258
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7735322713851929,
+      "learning_rate": 4.278615749156655e-06,
+      "loss": 0.901,
+      "step": 12259
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5340155363082886,
+      "learning_rate": 4.277092314630278e-06,
+      "loss": 0.9071,
+      "step": 12260
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9996885657310486,
+      "learning_rate": 4.275569077585455e-06,
+      "loss": 0.4977,
+      "step": 12261
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8660857677459717,
+      "learning_rate": 4.274046038074756e-06,
+      "loss": 0.9849,
+      "step": 12262
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7176084518432617,
+      "learning_rate": 4.27252319615073e-06,
+      "loss": 0.9249,
+      "step": 12263
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6369363069534302,
+      "learning_rate": 4.271000551865934e-06,
+      "loss": 0.8533,
+      "step": 12264
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8407223224639893,
+      "learning_rate": 4.269478105272901e-06,
+      "loss": 0.9376,
+      "step": 12265
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8782856464385986,
+      "learning_rate": 4.267955856424175e-06,
+      "loss": 0.9669,
+      "step": 12266
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7713334560394287,
+      "learning_rate": 4.266433805372278e-06,
+      "loss": 0.929,
+      "step": 12267
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8101853132247925,
+      "learning_rate": 4.264911952169736e-06,
+      "loss": 0.966,
+      "step": 12268
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7709208726882935,
+      "learning_rate": 4.26339029686906e-06,
+      "loss": 0.8807,
+      "step": 12269
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7185662984848022,
+      "learning_rate": 4.2618688395227624e-06,
+      "loss": 0.968,
+      "step": 12270
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.855184555053711,
+      "learning_rate": 4.26034758018334e-06,
+      "loss": 0.8851,
+      "step": 12271
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8644254207611084,
+      "learning_rate": 4.25882651890329e-06,
+      "loss": 1.0355,
+      "step": 12272
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6256617307662964,
+      "learning_rate": 4.2573056557351015e-06,
+      "loss": 0.8821,
+      "step": 12273
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7973995208740234,
+      "learning_rate": 4.2557849907312494e-06,
+      "loss": 0.9267,
+      "step": 12274
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.726332664489746,
+      "learning_rate": 4.254264523944217e-06,
+      "loss": 0.8717,
+      "step": 12275
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6894577741622925,
+      "learning_rate": 4.252744255426461e-06,
+      "loss": 0.8925,
+      "step": 12276
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8342831134796143,
+      "learning_rate": 4.2512241852304506e-06,
+      "loss": 0.9195,
+      "step": 12277
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6363437175750732,
+      "learning_rate": 4.249704313408632e-06,
+      "loss": 0.9135,
+      "step": 12278
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8010531663894653,
+      "learning_rate": 4.248184640013456e-06,
+      "loss": 0.9287,
+      "step": 12279
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7862954139709473,
+      "learning_rate": 4.24666516509736e-06,
+      "loss": 0.921,
+      "step": 12280
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6051653623580933,
+      "learning_rate": 4.24514588871278e-06,
+      "loss": 0.8599,
+      "step": 12281
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.816988468170166,
+      "learning_rate": 4.243626810912137e-06,
+      "loss": 0.9146,
+      "step": 12282
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6443513631820679,
+      "learning_rate": 4.242107931747855e-06,
+      "loss": 0.9604,
+      "step": 12283
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6637365818023682,
+      "learning_rate": 4.240589251272342e-06,
+      "loss": 0.8863,
+      "step": 12284
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7601675987243652,
+      "learning_rate": 4.2390707695380065e-06,
+      "loss": 0.9393,
+      "step": 12285
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8307349681854248,
+      "learning_rate": 4.2375524865972485e-06,
+      "loss": 0.9495,
+      "step": 12286
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7934890985488892,
+      "learning_rate": 4.236034402502454e-06,
+      "loss": 0.9336,
+      "step": 12287
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.90567147731781,
+      "learning_rate": 4.234516517306016e-06,
+      "loss": 0.9258,
+      "step": 12288
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8131781816482544,
+      "learning_rate": 4.2329988310603025e-06,
+      "loss": 0.9204,
+      "step": 12289
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0351872444152832,
+      "learning_rate": 4.231481343817694e-06,
+      "loss": 0.5383,
+      "step": 12290
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5632492303848267,
+      "learning_rate": 4.229964055630547e-06,
+      "loss": 0.8826,
+      "step": 12291
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6366617679595947,
+      "learning_rate": 4.2284469665512265e-06,
+      "loss": 0.8572,
+      "step": 12292
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8293219804763794,
+      "learning_rate": 4.226930076632075e-06,
+      "loss": 0.9581,
+      "step": 12293
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7509188652038574,
+      "learning_rate": 4.2254133859254445e-06,
+      "loss": 0.8889,
+      "step": 12294
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7620365619659424,
+      "learning_rate": 4.223896894483664e-06,
+      "loss": 0.9245,
+      "step": 12295
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0621201992034912,
+      "learning_rate": 4.222380602359065e-06,
+      "loss": 0.4952,
+      "step": 12296
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8416523933410645,
+      "learning_rate": 4.220864509603977e-06,
+      "loss": 0.9421,
+      "step": 12297
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8055000305175781,
+      "learning_rate": 4.219348616270707e-06,
+      "loss": 0.9643,
+      "step": 12298
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.926814079284668,
+      "learning_rate": 4.217832922411574e-06,
+      "loss": 0.937,
+      "step": 12299
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6999422311782837,
+      "learning_rate": 4.21631742807887e-06,
+      "loss": 0.9289,
+      "step": 12300
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5582475662231445,
+      "learning_rate": 4.2148021333249e-06,
+      "loss": 0.7853,
+      "step": 12301
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8325923681259155,
+      "learning_rate": 4.213287038201943e-06,
+      "loss": 0.9646,
+      "step": 12302
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7302120923995972,
+      "learning_rate": 4.2117721427622916e-06,
+      "loss": 0.9201,
+      "step": 12303
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.658416748046875,
+      "learning_rate": 4.2102574470582094e-06,
+      "loss": 0.7855,
+      "step": 12304
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.869757056236267,
+      "learning_rate": 4.208742951141974e-06,
+      "loss": 0.8731,
+      "step": 12305
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5227854251861572,
+      "learning_rate": 4.207228655065838e-06,
+      "loss": 0.922,
+      "step": 12306
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7483675479888916,
+      "learning_rate": 4.205714558882064e-06,
+      "loss": 0.9423,
+      "step": 12307
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9386212825775146,
+      "learning_rate": 4.204200662642891e-06,
+      "loss": 0.9667,
+      "step": 12308
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5964329242706299,
+      "learning_rate": 4.2026869664005635e-06,
+      "loss": 0.9097,
+      "step": 12309
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6607552766799927,
+      "learning_rate": 4.201173470207317e-06,
+      "loss": 0.9291,
+      "step": 12310
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.658761739730835,
+      "learning_rate": 4.199660174115373e-06,
+      "loss": 0.9318,
+      "step": 12311
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7204349040985107,
+      "learning_rate": 4.1981470781769574e-06,
+      "loss": 0.8651,
+      "step": 12312
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.692347526550293,
+      "learning_rate": 4.196634182444276e-06,
+      "loss": 0.9394,
+      "step": 12313
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8105977773666382,
+      "learning_rate": 4.195121486969541e-06,
+      "loss": 0.984,
+      "step": 12314
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9936108589172363,
+      "learning_rate": 4.193608991804945e-06,
+      "loss": 0.9871,
+      "step": 12315
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7505136728286743,
+      "learning_rate": 4.192096697002686e-06,
+      "loss": 0.9003,
+      "step": 12316
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6852811574935913,
+      "learning_rate": 4.19058460261494e-06,
+      "loss": 0.8776,
+      "step": 12317
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7839736938476562,
+      "learning_rate": 4.189072708693899e-06,
+      "loss": 0.9221,
+      "step": 12318
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.588187336921692,
+      "learning_rate": 4.1875610152917225e-06,
+      "loss": 0.8894,
+      "step": 12319
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.654762625694275,
+      "learning_rate": 4.186049522460581e-06,
+      "loss": 0.835,
+      "step": 12320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8535449504852295,
+      "learning_rate": 4.184538230252628e-06,
+      "loss": 0.9031,
+      "step": 12321
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5364501476287842,
+      "learning_rate": 4.183027138720019e-06,
+      "loss": 0.8726,
+      "step": 12322
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8365474939346313,
+      "learning_rate": 4.181516247914892e-06,
+      "loss": 0.9154,
+      "step": 12323
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5075379610061646,
+      "learning_rate": 4.180005557889388e-06,
+      "loss": 0.7941,
+      "step": 12324
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7151494026184082,
+      "learning_rate": 4.178495068695632e-06,
+      "loss": 0.976,
+      "step": 12325
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7232338190078735,
+      "learning_rate": 4.176984780385749e-06,
+      "loss": 1.0321,
+      "step": 12326
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9207673072814941,
+      "learning_rate": 4.175474693011858e-06,
+      "loss": 0.967,
+      "step": 12327
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7690528631210327,
+      "learning_rate": 4.173964806626063e-06,
+      "loss": 0.9015,
+      "step": 12328
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6922712326049805,
+      "learning_rate": 4.172455121280471e-06,
+      "loss": 0.9439,
+      "step": 12329
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8027880191802979,
+      "learning_rate": 4.1709456370271716e-06,
+      "loss": 0.9474,
+      "step": 12330
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.946854591369629,
+      "learning_rate": 4.169436353918258e-06,
+      "loss": 0.9256,
+      "step": 12331
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.715364933013916,
+      "learning_rate": 4.167927272005805e-06,
+      "loss": 0.8614,
+      "step": 12332
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6620863676071167,
+      "learning_rate": 4.1664183913418955e-06,
+      "loss": 0.8832,
+      "step": 12333
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7056200504302979,
+      "learning_rate": 4.164909711978587e-06,
+      "loss": 0.8893,
+      "step": 12334
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.67353093624115,
+      "learning_rate": 4.163401233967949e-06,
+      "loss": 0.9014,
+      "step": 12335
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7750691175460815,
+      "learning_rate": 4.161892957362027e-06,
+      "loss": 0.896,
+      "step": 12336
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7011590003967285,
+      "learning_rate": 4.160384882212875e-06,
+      "loss": 0.8289,
+      "step": 12337
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7766766548156738,
+      "learning_rate": 4.158877008572523e-06,
+      "loss": 0.8957,
+      "step": 12338
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6974159479141235,
+      "learning_rate": 4.15736933649301e-06,
+      "loss": 0.8842,
+      "step": 12339
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.032731056213379,
+      "learning_rate": 4.155861866026364e-06,
+      "loss": 0.9313,
+      "step": 12340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.81290602684021,
+      "learning_rate": 4.154354597224597e-06,
+      "loss": 0.9109,
+      "step": 12341
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8402010202407837,
+      "learning_rate": 4.152847530139726e-06,
+      "loss": 0.9116,
+      "step": 12342
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7119414806365967,
+      "learning_rate": 4.151340664823751e-06,
+      "loss": 0.9269,
+      "step": 12343
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7739782333374023,
+      "learning_rate": 4.1498340013286755e-06,
+      "loss": 0.8528,
+      "step": 12344
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.058659315109253,
+      "learning_rate": 4.148327539706483e-06,
+      "loss": 0.5841,
+      "step": 12345
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7927967309951782,
+      "learning_rate": 4.146821280009165e-06,
+      "loss": 0.8347,
+      "step": 12346
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8593764305114746,
+      "learning_rate": 4.14531522228869e-06,
+      "loss": 0.9265,
+      "step": 12347
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7242666482925415,
+      "learning_rate": 4.143809366597037e-06,
+      "loss": 0.9712,
+      "step": 12348
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7882723808288574,
+      "learning_rate": 4.14230371298616e-06,
+      "loss": 0.9409,
+      "step": 12349
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.588816523551941,
+      "learning_rate": 4.140798261508019e-06,
+      "loss": 0.9311,
+      "step": 12350
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5768983364105225,
+      "learning_rate": 4.139293012214566e-06,
+      "loss": 0.8495,
+      "step": 12351
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.850172519683838,
+      "learning_rate": 4.137787965157737e-06,
+      "loss": 0.9402,
+      "step": 12352
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6511579751968384,
+      "learning_rate": 4.136283120389474e-06,
+      "loss": 0.8052,
+      "step": 12353
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7065151929855347,
+      "learning_rate": 4.134778477961696e-06,
+      "loss": 0.9554,
+      "step": 12354
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.922829270362854,
+      "learning_rate": 4.1332740379263335e-06,
+      "loss": 0.9357,
+      "step": 12355
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.006742000579834,
+      "learning_rate": 4.131769800335293e-06,
+      "loss": 0.9558,
+      "step": 12356
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7688363790512085,
+      "learning_rate": 4.1302657652404865e-06,
+      "loss": 0.9481,
+      "step": 12357
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6695479154586792,
+      "learning_rate": 4.128761932693809e-06,
+      "loss": 0.9734,
+      "step": 12358
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.860970377922058,
+      "learning_rate": 4.127258302747159e-06,
+      "loss": 0.939,
+      "step": 12359
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7460209131240845,
+      "learning_rate": 4.1257548754524175e-06,
+      "loss": 0.9254,
+      "step": 12360
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9051164388656616,
+      "learning_rate": 4.124251650861471e-06,
+      "loss": 0.9823,
+      "step": 12361
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7319355010986328,
+      "learning_rate": 4.122748629026182e-06,
+      "loss": 0.9391,
+      "step": 12362
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7394003868103027,
+      "learning_rate": 4.12124580999842e-06,
+      "loss": 0.871,
+      "step": 12363
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6891359090805054,
+      "learning_rate": 4.119743193830048e-06,
+      "loss": 0.9179,
+      "step": 12364
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8783525228500366,
+      "learning_rate": 4.1182407805729084e-06,
+      "loss": 0.9281,
+      "step": 12365
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6569445133209229,
+      "learning_rate": 4.116738570278853e-06,
+      "loss": 0.8388,
+      "step": 12366
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0325391292572021,
+      "learning_rate": 4.115236562999713e-06,
+      "loss": 0.5731,
+      "step": 12367
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8691225051879883,
+      "learning_rate": 4.113734758787322e-06,
+      "loss": 0.9262,
+      "step": 12368
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6074047088623047,
+      "learning_rate": 4.112233157693501e-06,
+      "loss": 0.8511,
+      "step": 12369
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8227767944335938,
+      "learning_rate": 4.110731759770068e-06,
+      "loss": 0.9323,
+      "step": 12370
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.777860403060913,
+      "learning_rate": 4.109230565068828e-06,
+      "loss": 0.9444,
+      "step": 12371
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.888817310333252,
+      "learning_rate": 4.10772957364159e-06,
+      "loss": 0.907,
+      "step": 12372
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.888682246208191,
+      "learning_rate": 4.106228785540141e-06,
+      "loss": 0.8766,
+      "step": 12373
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8184161186218262,
+      "learning_rate": 4.1047282008162734e-06,
+      "loss": 0.9125,
+      "step": 12374
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7346991300582886,
+      "learning_rate": 4.1032278195217725e-06,
+      "loss": 0.9246,
+      "step": 12375
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8360766172409058,
+      "learning_rate": 4.101727641708403e-06,
+      "loss": 0.985,
+      "step": 12376
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.703584909439087,
+      "learning_rate": 4.1002276674279395e-06,
+      "loss": 0.8077,
+      "step": 12377
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.140605092048645,
+      "learning_rate": 4.098727896732135e-06,
+      "loss": 0.6147,
+      "step": 12378
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.724928855895996,
+      "learning_rate": 4.097228329672751e-06,
+      "loss": 0.8605,
+      "step": 12379
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5698235034942627,
+      "learning_rate": 4.095728966301526e-06,
+      "loss": 0.8455,
+      "step": 12380
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.629245400428772,
+      "learning_rate": 4.0942298066702026e-06,
+      "loss": 0.9636,
+      "step": 12381
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8827617168426514,
+      "learning_rate": 4.092730850830509e-06,
+      "loss": 0.9127,
+      "step": 12382
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7989907264709473,
+      "learning_rate": 4.0912320988341725e-06,
+      "loss": 1.0247,
+      "step": 12383
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9501909613609314,
+      "learning_rate": 4.0897335507329104e-06,
+      "loss": 0.4938,
+      "step": 12384
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7616958618164062,
+      "learning_rate": 4.088235206578438e-06,
+      "loss": 0.8138,
+      "step": 12385
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6607574224472046,
+      "learning_rate": 4.086737066422451e-06,
+      "loss": 0.9306,
+      "step": 12386
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7319327592849731,
+      "learning_rate": 4.085239130316653e-06,
+      "loss": 0.9814,
+      "step": 12387
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.812543272972107,
+      "learning_rate": 4.083741398312727e-06,
+      "loss": 0.8766,
+      "step": 12388
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7206640243530273,
+      "learning_rate": 4.082243870462362e-06,
+      "loss": 0.9932,
+      "step": 12389
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7464792728424072,
+      "learning_rate": 4.080746546817228e-06,
+      "loss": 0.9365,
+      "step": 12390
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7988104820251465,
+      "learning_rate": 4.079249427428995e-06,
+      "loss": 0.8834,
+      "step": 12391
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.004652261734009,
+      "learning_rate": 4.077752512349329e-06,
+      "loss": 0.8995,
+      "step": 12392
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.943991780281067,
+      "learning_rate": 4.076255801629877e-06,
+      "loss": 0.9497,
+      "step": 12393
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8857355117797852,
+      "learning_rate": 4.074759295322295e-06,
+      "loss": 0.8529,
+      "step": 12394
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.818435788154602,
+      "learning_rate": 4.073262993478213e-06,
+      "loss": 0.9409,
+      "step": 12395
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7896023988723755,
+      "learning_rate": 4.0717668961492725e-06,
+      "loss": 0.9767,
+      "step": 12396
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6305948495864868,
+      "learning_rate": 4.0702710033870955e-06,
+      "loss": 0.9032,
+      "step": 12397
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8447567224502563,
+      "learning_rate": 4.068775315243303e-06,
+      "loss": 0.8821,
+      "step": 12398
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6577025651931763,
+      "learning_rate": 4.067279831769504e-06,
+      "loss": 0.941,
+      "step": 12399
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6408772468566895,
+      "learning_rate": 4.065784553017309e-06,
+      "loss": 0.9069,
+      "step": 12400
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7687045335769653,
+      "learning_rate": 4.0642894790383094e-06,
+      "loss": 0.959,
+      "step": 12401
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1135081052780151,
+      "learning_rate": 4.062794609884102e-06,
+      "loss": 0.5891,
+      "step": 12402
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.851867914199829,
+      "learning_rate": 4.061299945606264e-06,
+      "loss": 0.9897,
+      "step": 12403
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8630967140197754,
+      "learning_rate": 4.059805486256376e-06,
+      "loss": 0.8719,
+      "step": 12404
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7124943733215332,
+      "learning_rate": 4.058311231886012e-06,
+      "loss": 0.9115,
+      "step": 12405
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8081378936767578,
+      "learning_rate": 4.056817182546725e-06,
+      "loss": 0.8665,
+      "step": 12406
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7391353845596313,
+      "learning_rate": 4.055323338290079e-06,
+      "loss": 0.9366,
+      "step": 12407
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8912296295166016,
+      "learning_rate": 4.053829699167616e-06,
+      "loss": 0.8939,
+      "step": 12408
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6939836740493774,
+      "learning_rate": 4.052336265230884e-06,
+      "loss": 0.844,
+      "step": 12409
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7051935195922852,
+      "learning_rate": 4.050843036531409e-06,
+      "loss": 0.8683,
+      "step": 12410
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7583786249160767,
+      "learning_rate": 4.049350013120726e-06,
+      "loss": 0.8762,
+      "step": 12411
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.638037919998169,
+      "learning_rate": 4.047857195050349e-06,
+      "loss": 0.8813,
+      "step": 12412
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7822891473770142,
+      "learning_rate": 4.046364582371795e-06,
+      "loss": 0.9287,
+      "step": 12413
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7306482791900635,
+      "learning_rate": 4.0448721751365675e-06,
+      "loss": 0.9232,
+      "step": 12414
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1819629669189453,
+      "learning_rate": 4.0433799733961685e-06,
+      "loss": 0.9664,
+      "step": 12415
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7338840961456299,
+      "learning_rate": 4.0418879772020835e-06,
+      "loss": 1.0119,
+      "step": 12416
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8047226667404175,
+      "learning_rate": 4.040396186605803e-06,
+      "loss": 0.8692,
+      "step": 12417
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8259817361831665,
+      "learning_rate": 4.038904601658804e-06,
+      "loss": 0.9406,
+      "step": 12418
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7484159469604492,
+      "learning_rate": 4.037413222412553e-06,
+      "loss": 0.8792,
+      "step": 12419
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5469627380371094,
+      "learning_rate": 4.035922048918519e-06,
+      "loss": 0.8614,
+      "step": 12420
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8891232013702393,
+      "learning_rate": 4.034431081228152e-06,
+      "loss": 0.9197,
+      "step": 12421
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7687270641326904,
+      "learning_rate": 4.0329403193929075e-06,
+      "loss": 0.9467,
+      "step": 12422
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0497314929962158,
+      "learning_rate": 4.031449763464222e-06,
+      "loss": 0.5542,
+      "step": 12423
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7079578638076782,
+      "learning_rate": 4.0299594134935335e-06,
+      "loss": 0.9123,
+      "step": 12424
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.802924633026123,
+      "learning_rate": 4.028469269532268e-06,
+      "loss": 0.9238,
+      "step": 12425
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.74862539768219,
+      "learning_rate": 4.0269793316318496e-06,
+      "loss": 0.9317,
+      "step": 12426
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.230196475982666,
+      "learning_rate": 4.025489599843686e-06,
+      "loss": 0.8103,
+      "step": 12427
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1775816679000854,
+      "learning_rate": 4.0240000742191875e-06,
+      "loss": 0.5971,
+      "step": 12428
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.729634404182434,
+      "learning_rate": 4.022510754809757e-06,
+      "loss": 0.8588,
+      "step": 12429
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8385225534439087,
+      "learning_rate": 4.021021641666778e-06,
+      "loss": 0.9568,
+      "step": 12430
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9453097581863403,
+      "learning_rate": 4.019532734841645e-06,
+      "loss": 0.8932,
+      "step": 12431
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0154201984405518,
+      "learning_rate": 4.018044034385728e-06,
+      "loss": 0.9893,
+      "step": 12432
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0026189088821411,
+      "learning_rate": 4.0165555403504055e-06,
+      "loss": 0.5369,
+      "step": 12433
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6947625875473022,
+      "learning_rate": 4.015067252787033e-06,
+      "loss": 0.9101,
+      "step": 12434
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.704961895942688,
+      "learning_rate": 4.013579171746975e-06,
+      "loss": 0.8898,
+      "step": 12435
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.846923828125,
+      "learning_rate": 4.012091297281574e-06,
+      "loss": 0.9626,
+      "step": 12436
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9454129934310913,
+      "learning_rate": 4.010603629442179e-06,
+      "loss": 0.9867,
+      "step": 12437
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5986747741699219,
+      "learning_rate": 4.009116168280119e-06,
+      "loss": 0.891,
+      "step": 12438
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6478374004364014,
+      "learning_rate": 4.0076289138467286e-06,
+      "loss": 0.8941,
+      "step": 12439
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9266692399978638,
+      "learning_rate": 4.006141866193321e-06,
+      "loss": 0.9009,
+      "step": 12440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8235458135604858,
+      "learning_rate": 4.004655025371215e-06,
+      "loss": 0.9583,
+      "step": 12441
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7677868604660034,
+      "learning_rate": 4.003168391431721e-06,
+      "loss": 0.8511,
+      "step": 12442
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0000358819961548,
+      "learning_rate": 4.001681964426131e-06,
+      "loss": 0.5646,
+      "step": 12443
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7779160737991333,
+      "learning_rate": 4.000195744405742e-06,
+      "loss": 1.0032,
+      "step": 12444
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6591830253601074,
+      "learning_rate": 3.998709731421837e-06,
+      "loss": 0.8998,
+      "step": 12445
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0530970096588135,
+      "learning_rate": 3.997223925525698e-06,
+      "loss": 0.5643,
+      "step": 12446
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9968852400779724,
+      "learning_rate": 3.995738326768589e-06,
+      "loss": 0.5334,
+      "step": 12447
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.000629425048828,
+      "learning_rate": 3.994252935201782e-06,
+      "loss": 0.949,
+      "step": 12448
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8586641550064087,
+      "learning_rate": 3.9927677508765235e-06,
+      "loss": 0.9192,
+      "step": 12449
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0317986011505127,
+      "learning_rate": 3.991282773844076e-06,
+      "loss": 0.9873,
+      "step": 12450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.721819519996643,
+      "learning_rate": 3.989798004155671e-06,
+      "loss": 0.8754,
+      "step": 12451
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7990550994873047,
+      "learning_rate": 3.9883134418625535e-06,
+      "loss": 0.9468,
+      "step": 12452
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5679570436477661,
+      "learning_rate": 3.986829087015941e-06,
+      "loss": 0.9563,
+      "step": 12453
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8023875951766968,
+      "learning_rate": 3.985344939667064e-06,
+      "loss": 0.894,
+      "step": 12454
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7562695741653442,
+      "learning_rate": 3.983860999867128e-06,
+      "loss": 0.939,
+      "step": 12455
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7863434553146362,
+      "learning_rate": 3.982377267667347e-06,
+      "loss": 0.8763,
+      "step": 12456
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6930670738220215,
+      "learning_rate": 3.980893743118913e-06,
+      "loss": 0.901,
+      "step": 12457
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6185472011566162,
+      "learning_rate": 3.979410426273022e-06,
+      "loss": 0.8511,
+      "step": 12458
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8135086297988892,
+      "learning_rate": 3.977927317180864e-06,
+      "loss": 0.8511,
+      "step": 12459
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6804667711257935,
+      "learning_rate": 3.9764444158936075e-06,
+      "loss": 1.0038,
+      "step": 12460
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.563314437866211,
+      "learning_rate": 3.9749617224624325e-06,
+      "loss": 0.8486,
+      "step": 12461
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8163152933120728,
+      "learning_rate": 3.9734792369384945e-06,
+      "loss": 1.011,
+      "step": 12462
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.882642388343811,
+      "learning_rate": 3.971996959372958e-06,
+      "loss": 0.934,
+      "step": 12463
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6761990785598755,
+      "learning_rate": 3.970514889816963e-06,
+      "loss": 0.8904,
+      "step": 12464
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7790278196334839,
+      "learning_rate": 3.96903302832166e-06,
+      "loss": 0.9655,
+      "step": 12465
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7485060691833496,
+      "learning_rate": 3.967551374938178e-06,
+      "loss": 0.974,
+      "step": 12466
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7336326837539673,
+      "learning_rate": 3.96606992971765e-06,
+      "loss": 0.9246,
+      "step": 12467
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7338225841522217,
+      "learning_rate": 3.96458869271119e-06,
+      "loss": 0.9192,
+      "step": 12468
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.676155924797058,
+      "learning_rate": 3.9631076639699185e-06,
+      "loss": 0.9595,
+      "step": 12469
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6723685264587402,
+      "learning_rate": 3.961626843544935e-06,
+      "loss": 0.9599,
+      "step": 12470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.919965147972107,
+      "learning_rate": 3.9601462314873405e-06,
+      "loss": 0.9325,
+      "step": 12471
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9680243730545044,
+      "learning_rate": 3.958665827848233e-06,
+      "loss": 0.8786,
+      "step": 12472
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.614292025566101,
+      "learning_rate": 3.957185632678687e-06,
+      "loss": 1.0014,
+      "step": 12473
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.982468605041504,
+      "learning_rate": 3.95570564602979e-06,
+      "loss": 0.9419,
+      "step": 12474
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0683231353759766,
+      "learning_rate": 3.954225867952602e-06,
+      "loss": 0.5804,
+      "step": 12475
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6746337413787842,
+      "learning_rate": 3.9527462984981954e-06,
+      "loss": 0.9248,
+      "step": 12476
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7570018768310547,
+      "learning_rate": 3.951266937717619e-06,
+      "loss": 0.9298,
+      "step": 12477
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7199463844299316,
+      "learning_rate": 3.949787785661926e-06,
+      "loss": 0.9107,
+      "step": 12478
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9739207029342651,
+      "learning_rate": 3.948308842382154e-06,
+      "loss": 1.0158,
+      "step": 12479
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.806829810142517,
+      "learning_rate": 3.946830107929342e-06,
+      "loss": 0.9366,
+      "step": 12480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7213636636734009,
+      "learning_rate": 3.94535158235451e-06,
+      "loss": 1.0085,
+      "step": 12481
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.725403070449829,
+      "learning_rate": 3.943873265708682e-06,
+      "loss": 0.9565,
+      "step": 12482
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7160180807113647,
+      "learning_rate": 3.9423951580428744e-06,
+      "loss": 0.9792,
+      "step": 12483
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6791510581970215,
+      "learning_rate": 3.940917259408085e-06,
+      "loss": 0.9852,
+      "step": 12484
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8221672773361206,
+      "learning_rate": 3.939439569855319e-06,
+      "loss": 0.9131,
+      "step": 12485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7410014867782593,
+      "learning_rate": 3.937962089435561e-06,
+      "loss": 0.9524,
+      "step": 12486
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6324843168258667,
+      "learning_rate": 3.936484818199801e-06,
+      "loss": 0.8046,
+      "step": 12487
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7280384302139282,
+      "learning_rate": 3.935007756199009e-06,
+      "loss": 0.8583,
+      "step": 12488
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6598620414733887,
+      "learning_rate": 3.9335309034841595e-06,
+      "loss": 0.9518,
+      "step": 12489
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9313338994979858,
+      "learning_rate": 3.932054260106209e-06,
+      "loss": 0.9691,
+      "step": 12490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8046241998672485,
+      "learning_rate": 3.9305778261161205e-06,
+      "loss": 0.9595,
+      "step": 12491
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7612308263778687,
+      "learning_rate": 3.929101601564834e-06,
+      "loss": 0.9195,
+      "step": 12492
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9295299053192139,
+      "learning_rate": 3.9276255865032965e-06,
+      "loss": 0.9285,
+      "step": 12493
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8986449241638184,
+      "learning_rate": 3.926149780982432e-06,
+      "loss": 1.0166,
+      "step": 12494
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7338577508926392,
+      "learning_rate": 3.924674185053173e-06,
+      "loss": 0.9299,
+      "step": 12495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.605047345161438,
+      "learning_rate": 3.923198798766441e-06,
+      "loss": 0.8553,
+      "step": 12496
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.671620488166809,
+      "learning_rate": 3.92172362217314e-06,
+      "loss": 0.9145,
+      "step": 12497
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8116626739501953,
+      "learning_rate": 3.920248655324182e-06,
+      "loss": 0.9558,
+      "step": 12498
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6567705869674683,
+      "learning_rate": 3.918773898270455e-06,
+      "loss": 0.9004,
+      "step": 12499
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.235962390899658,
+      "learning_rate": 3.917299351062858e-06,
+      "loss": 0.9996,
+      "step": 12500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.701453685760498,
+      "learning_rate": 3.915825013752265e-06,
+      "loss": 0.9198,
+      "step": 12501
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7635984420776367,
+      "learning_rate": 3.914350886389558e-06,
+      "loss": 0.9939,
+      "step": 12502
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.929249882698059,
+      "learning_rate": 3.912876969025601e-06,
+      "loss": 0.8475,
+      "step": 12503
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7297722101211548,
+      "learning_rate": 3.911403261711257e-06,
+      "loss": 0.8454,
+      "step": 12504
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6612706184387207,
+      "learning_rate": 3.909929764497377e-06,
+      "loss": 0.8418,
+      "step": 12505
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6837877035140991,
+      "learning_rate": 3.908456477434809e-06,
+      "loss": 0.8637,
+      "step": 12506
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9422123432159424,
+      "learning_rate": 3.906983400574394e-06,
+      "loss": 0.9981,
+      "step": 12507
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8037015199661255,
+      "learning_rate": 3.905510533966959e-06,
+      "loss": 0.9109,
+      "step": 12508
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.581459879875183,
+      "learning_rate": 3.9040378776633355e-06,
+      "loss": 0.9029,
+      "step": 12509
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.871157169342041,
+      "learning_rate": 3.902565431714333e-06,
+      "loss": 0.9423,
+      "step": 12510
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.815946340560913,
+      "learning_rate": 3.901093196170766e-06,
+      "loss": 0.9139,
+      "step": 12511
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.022512674331665,
+      "learning_rate": 3.899621171083435e-06,
+      "loss": 0.9518,
+      "step": 12512
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.766019582748413,
+      "learning_rate": 3.898149356503139e-06,
+      "loss": 0.9181,
+      "step": 12513
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7870184183120728,
+      "learning_rate": 3.896677752480662e-06,
+      "loss": 0.9607,
+      "step": 12514
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7253893613815308,
+      "learning_rate": 3.8952063590667855e-06,
+      "loss": 0.8236,
+      "step": 12515
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9726859331130981,
+      "learning_rate": 3.893735176312284e-06,
+      "loss": 0.5207,
+      "step": 12516
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.633440375328064,
+      "learning_rate": 3.892264204267929e-06,
+      "loss": 0.7954,
+      "step": 12517
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7915269136428833,
+      "learning_rate": 3.890793442984471e-06,
+      "loss": 0.9201,
+      "step": 12518
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6709668636322021,
+      "learning_rate": 3.889322892512669e-06,
+      "loss": 0.9227,
+      "step": 12519
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7345893383026123,
+      "learning_rate": 3.887852552903262e-06,
+      "loss": 0.9785,
+      "step": 12520
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8026156425476074,
+      "learning_rate": 3.886382424206992e-06,
+      "loss": 0.9719,
+      "step": 12521
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7650612592697144,
+      "learning_rate": 3.884912506474585e-06,
+      "loss": 0.9712,
+      "step": 12522
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6858350038528442,
+      "learning_rate": 3.883442799756768e-06,
+      "loss": 0.9125,
+      "step": 12523
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6993342638015747,
+      "learning_rate": 3.881973304104252e-06,
+      "loss": 0.8961,
+      "step": 12524
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.997534990310669,
+      "learning_rate": 3.880504019567746e-06,
+      "loss": 0.9865,
+      "step": 12525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7246320247650146,
+      "learning_rate": 3.879034946197955e-06,
+      "loss": 0.9523,
+      "step": 12526
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5362396240234375,
+      "learning_rate": 3.877566084045567e-06,
+      "loss": 0.8882,
+      "step": 12527
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.66085684299469,
+      "learning_rate": 3.876097433161275e-06,
+      "loss": 0.974,
+      "step": 12528
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7776702642440796,
+      "learning_rate": 3.87462899359575e-06,
+      "loss": 0.9397,
+      "step": 12529
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.78606379032135,
+      "learning_rate": 3.873160765399672e-06,
+      "loss": 0.8514,
+      "step": 12530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9417780637741089,
+      "learning_rate": 3.8716927486236975e-06,
+      "loss": 0.8324,
+      "step": 12531
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6343858242034912,
+      "learning_rate": 3.870224943318491e-06,
+      "loss": 0.8932,
+      "step": 12532
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9871156811714172,
+      "learning_rate": 3.868757349534695e-06,
+      "loss": 0.5515,
+      "step": 12533
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8294332027435303,
+      "learning_rate": 3.86728996732296e-06,
+      "loss": 0.9094,
+      "step": 12534
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7224035263061523,
+      "learning_rate": 3.865822796733914e-06,
+      "loss": 0.9669,
+      "step": 12535
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7133840322494507,
+      "learning_rate": 3.864355837818188e-06,
+      "loss": 0.8429,
+      "step": 12536
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7271113395690918,
+      "learning_rate": 3.862889090626406e-06,
+      "loss": 0.8986,
+      "step": 12537
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7311434745788574,
+      "learning_rate": 3.8614225552091745e-06,
+      "loss": 0.9566,
+      "step": 12538
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6086724996566772,
+      "learning_rate": 3.859956231617107e-06,
+      "loss": 0.8714,
+      "step": 12539
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.857214331626892,
+      "learning_rate": 3.858490119900794e-06,
+      "loss": 0.9976,
+      "step": 12540
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6624752283096313,
+      "learning_rate": 3.857024220110837e-06,
+      "loss": 0.8879,
+      "step": 12541
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7396091222763062,
+      "learning_rate": 3.855558532297808e-06,
+      "loss": 0.8906,
+      "step": 12542
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8602701425552368,
+      "learning_rate": 3.854093056512296e-06,
+      "loss": 0.9584,
+      "step": 12543
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6431396007537842,
+      "learning_rate": 3.85262779280486e-06,
+      "loss": 0.834,
+      "step": 12544
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7836977243423462,
+      "learning_rate": 3.851162741226071e-06,
+      "loss": 0.8997,
+      "step": 12545
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6669334173202515,
+      "learning_rate": 3.849697901826477e-06,
+      "loss": 0.9244,
+      "step": 12546
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7953133583068848,
+      "learning_rate": 3.848233274656631e-06,
+      "loss": 0.9685,
+      "step": 12547
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6984896659851074,
+      "learning_rate": 3.846768859767066e-06,
+      "loss": 0.9369,
+      "step": 12548
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7227824926376343,
+      "learning_rate": 3.845304657208321e-06,
+      "loss": 0.917,
+      "step": 12549
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.754768967628479,
+      "learning_rate": 3.8438406670309215e-06,
+      "loss": 0.9208,
+      "step": 12550
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7911325693130493,
+      "learning_rate": 3.842376889285382e-06,
+      "loss": 0.8674,
+      "step": 12551
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7789198160171509,
+      "learning_rate": 3.840913324022218e-06,
+      "loss": 0.9308,
+      "step": 12552
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8998196125030518,
+      "learning_rate": 3.8394499712919275e-06,
+      "loss": 0.9987,
+      "step": 12553
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9097747802734375,
+      "learning_rate": 3.8379868311450134e-06,
+      "loss": 0.8982,
+      "step": 12554
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6499348878860474,
+      "learning_rate": 3.8365239036319565e-06,
+      "loss": 0.9653,
+      "step": 12555
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8675527572631836,
+      "learning_rate": 3.8350611888032474e-06,
+      "loss": 0.9251,
+      "step": 12556
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8499308824539185,
+      "learning_rate": 3.833598686709351e-06,
+      "loss": 0.954,
+      "step": 12557
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.83310067653656,
+      "learning_rate": 3.832136397400743e-06,
+      "loss": 0.9913,
+      "step": 12558
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1436314582824707,
+      "learning_rate": 3.830674320927875e-06,
+      "loss": 0.9536,
+      "step": 12559
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9855970144271851,
+      "learning_rate": 3.829212457341203e-06,
+      "loss": 0.5498,
+      "step": 12560
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.843736171722412,
+      "learning_rate": 3.827750806691175e-06,
+      "loss": 0.9276,
+      "step": 12561
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0240211486816406,
+      "learning_rate": 3.8262893690282214e-06,
+      "loss": 0.9149,
+      "step": 12562
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6047064065933228,
+      "learning_rate": 3.82482814440278e-06,
+      "loss": 0.8451,
+      "step": 12563
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.83649742603302,
+      "learning_rate": 3.823367132865266e-06,
+      "loss": 0.9819,
+      "step": 12564
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6730918884277344,
+      "learning_rate": 3.821906334466102e-06,
+      "loss": 0.9815,
+      "step": 12565
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7875293493270874,
+      "learning_rate": 3.820445749255689e-06,
+      "loss": 0.9596,
+      "step": 12566
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6322795152664185,
+      "learning_rate": 3.818985377284435e-06,
+      "loss": 0.9155,
+      "step": 12567
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9037878513336182,
+      "learning_rate": 3.817525218602727e-06,
+      "loss": 0.9464,
+      "step": 12568
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9884686470031738,
+      "learning_rate": 3.816065273260956e-06,
+      "loss": 0.9943,
+      "step": 12569
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.983309030532837,
+      "learning_rate": 3.814605541309495e-06,
+      "loss": 0.952,
+      "step": 12570
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9923788905143738,
+      "learning_rate": 3.8131460227987214e-06,
+      "loss": 0.571,
+      "step": 12571
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7488611936569214,
+      "learning_rate": 3.811686717778994e-06,
+      "loss": 0.9517,
+      "step": 12572
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7324405908584595,
+      "learning_rate": 3.810227626300671e-06,
+      "loss": 0.8915,
+      "step": 12573
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.570957064628601,
+      "learning_rate": 3.8087687484141055e-06,
+      "loss": 0.9517,
+      "step": 12574
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9039630889892578,
+      "learning_rate": 3.8073100841696333e-06,
+      "loss": 0.9445,
+      "step": 12575
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8011610507965088,
+      "learning_rate": 3.8058516336175942e-06,
+      "loss": 0.9286,
+      "step": 12576
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.805119276046753,
+      "learning_rate": 3.80439339680831e-06,
+      "loss": 0.8667,
+      "step": 12577
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8024210929870605,
+      "learning_rate": 3.802935373792106e-06,
+      "loss": 0.9213,
+      "step": 12578
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6766784191131592,
+      "learning_rate": 3.801477564619287e-06,
+      "loss": 0.948,
+      "step": 12579
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6544581651687622,
+      "learning_rate": 3.8000199693401675e-06,
+      "loss": 0.9568,
+      "step": 12580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9259475469589233,
+      "learning_rate": 3.7985625880050315e-06,
+      "loss": 0.9144,
+      "step": 12581
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6724088191986084,
+      "learning_rate": 3.7971054206641854e-06,
+      "loss": 0.9235,
+      "step": 12582
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8785679340362549,
+      "learning_rate": 3.7956484673679006e-06,
+      "loss": 0.9227,
+      "step": 12583
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6434869766235352,
+      "learning_rate": 3.7941917281664586e-06,
+      "loss": 0.9661,
+      "step": 12584
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.777374029159546,
+      "learning_rate": 3.7927352031101228e-06,
+      "loss": 0.9159,
+      "step": 12585
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6906895637512207,
+      "learning_rate": 3.7912788922491582e-06,
+      "loss": 0.8463,
+      "step": 12586
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6666299104690552,
+      "learning_rate": 3.789822795633813e-06,
+      "loss": 0.9892,
+      "step": 12587
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6618847846984863,
+      "learning_rate": 3.7883669133143388e-06,
+      "loss": 0.8577,
+      "step": 12588
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.839779257774353,
+      "learning_rate": 3.7869112453409673e-06,
+      "loss": 0.9683,
+      "step": 12589
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.899091362953186,
+      "learning_rate": 3.7854557917639333e-06,
+      "loss": 0.8988,
+      "step": 12590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.85112726688385,
+      "learning_rate": 3.7840005526334633e-06,
+      "loss": 0.9093,
+      "step": 12591
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7250407934188843,
+      "learning_rate": 3.782545527999768e-06,
+      "loss": 0.8787,
+      "step": 12592
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8389370441436768,
+      "learning_rate": 3.781090717913062e-06,
+      "loss": 0.8964,
+      "step": 12593
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7947970628738403,
+      "learning_rate": 3.77963612242354e-06,
+      "loss": 0.8455,
+      "step": 12594
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7338473796844482,
+      "learning_rate": 3.778181741581403e-06,
+      "loss": 0.9537,
+      "step": 12595
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.795505404472351,
+      "learning_rate": 3.7767275754368292e-06,
+      "loss": 0.8984,
+      "step": 12596
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0807818174362183,
+      "learning_rate": 3.775273624040008e-06,
+      "loss": 0.5417,
+      "step": 12597
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.979111909866333,
+      "learning_rate": 3.7738198874411026e-06,
+      "loss": 0.9746,
+      "step": 12598
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.93534255027771,
+      "learning_rate": 3.772366365690283e-06,
+      "loss": 0.9461,
+      "step": 12599
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7157135009765625,
+      "learning_rate": 3.7709130588377007e-06,
+      "loss": 0.8516,
+      "step": 12600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.702778935432434,
+      "learning_rate": 3.7694599669335132e-06,
+      "loss": 0.9412,
+      "step": 12601
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7491122484207153,
+      "learning_rate": 3.7680070900278533e-06,
+      "loss": 1.0122,
+      "step": 12602
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9815839529037476,
+      "learning_rate": 3.766554428170861e-06,
+      "loss": 0.9377,
+      "step": 12603
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0403605699539185,
+      "learning_rate": 3.7651019814126656e-06,
+      "loss": 0.551,
+      "step": 12604
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6842703819274902,
+      "learning_rate": 3.7636497498033817e-06,
+      "loss": 0.9202,
+      "step": 12605
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7488585710525513,
+      "learning_rate": 3.762197733393127e-06,
+      "loss": 0.8912,
+      "step": 12606
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.752842903137207,
+      "learning_rate": 3.7607459322320015e-06,
+      "loss": 0.9385,
+      "step": 12607
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7367526292800903,
+      "learning_rate": 3.7592943463701083e-06,
+      "loss": 0.9174,
+      "step": 12608
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.698089838027954,
+      "learning_rate": 3.7578429758575306e-06,
+      "loss": 0.936,
+      "step": 12609
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7510919570922852,
+      "learning_rate": 3.7563918207443583e-06,
+      "loss": 0.8927,
+      "step": 12610
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.940737009048462,
+      "learning_rate": 3.754940881080661e-06,
+      "loss": 0.8255,
+      "step": 12611
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8052895069122314,
+      "learning_rate": 3.7534901569165117e-06,
+      "loss": 0.9187,
+      "step": 12612
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8674594163894653,
+      "learning_rate": 3.7520396483019648e-06,
+      "loss": 0.9784,
+      "step": 12613
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7453211545944214,
+      "learning_rate": 3.7505893552870774e-06,
+      "loss": 0.8841,
+      "step": 12614
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6215099096298218,
+      "learning_rate": 3.749139277921897e-06,
+      "loss": 0.9096,
+      "step": 12615
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7010674476623535,
+      "learning_rate": 3.747689416256456e-06,
+      "loss": 0.9248,
+      "step": 12616
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6624914407730103,
+      "learning_rate": 3.7462397703407917e-06,
+      "loss": 0.8962,
+      "step": 12617
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7322510480880737,
+      "learning_rate": 3.7447903402249197e-06,
+      "loss": 0.8598,
+      "step": 12618
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.720736026763916,
+      "learning_rate": 3.7433411259588635e-06,
+      "loss": 0.9601,
+      "step": 12619
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7048335075378418,
+      "learning_rate": 3.7418921275926245e-06,
+      "loss": 0.9032,
+      "step": 12620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5797029733657837,
+      "learning_rate": 3.74044334517621e-06,
+      "loss": 0.8538,
+      "step": 12621
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8111974000930786,
+      "learning_rate": 3.738994778759607e-06,
+      "loss": 1.0421,
+      "step": 12622
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5807260274887085,
+      "learning_rate": 3.7375464283928086e-06,
+      "loss": 0.9306,
+      "step": 12623
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6567331552505493,
+      "learning_rate": 3.736098294125785e-06,
+      "loss": 0.9437,
+      "step": 12624
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7192879915237427,
+      "learning_rate": 3.734650376008516e-06,
+      "loss": 0.9957,
+      "step": 12625
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.410449504852295,
+      "learning_rate": 3.7332026740909576e-06,
+      "loss": 0.9806,
+      "step": 12626
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.782019019126892,
+      "learning_rate": 3.7317551884230697e-06,
+      "loss": 0.9629,
+      "step": 12627
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7913216352462769,
+      "learning_rate": 3.730307919054803e-06,
+      "loss": 0.9445,
+      "step": 12628
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.861438274383545,
+      "learning_rate": 3.7288608660360935e-06,
+      "loss": 0.9476,
+      "step": 12629
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6599637269973755,
+      "learning_rate": 3.7274140294168813e-06,
+      "loss": 0.9118,
+      "step": 12630
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7982391119003296,
+      "learning_rate": 3.7259674092470853e-06,
+      "loss": 0.9777,
+      "step": 12631
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7976106405258179,
+      "learning_rate": 3.7245210055766324e-06,
+      "loss": 0.9944,
+      "step": 12632
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8489376306533813,
+      "learning_rate": 3.7230748184554254e-06,
+      "loss": 0.9221,
+      "step": 12633
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7034870386123657,
+      "learning_rate": 3.7216288479333763e-06,
+      "loss": 0.8935,
+      "step": 12634
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.614827275276184,
+      "learning_rate": 3.7201830940603747e-06,
+      "loss": 0.8533,
+      "step": 12635
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.673074245452881,
+      "learning_rate": 3.718737556886316e-06,
+      "loss": 0.9384,
+      "step": 12636
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.713295340538025,
+      "learning_rate": 3.717292236461074e-06,
+      "loss": 0.9422,
+      "step": 12637
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6888242959976196,
+      "learning_rate": 3.715847132834528e-06,
+      "loss": 0.9148,
+      "step": 12638
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.663413166999817,
+      "learning_rate": 3.7144022460565452e-06,
+      "loss": 0.9475,
+      "step": 12639
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6894317865371704,
+      "learning_rate": 3.712957576176981e-06,
+      "loss": 0.9171,
+      "step": 12640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6570804119110107,
+      "learning_rate": 3.7115131232456915e-06,
+      "loss": 0.9264,
+      "step": 12641
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6393131017684937,
+      "learning_rate": 3.7100688873125147e-06,
+      "loss": 0.9128,
+      "step": 12642
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.886132001876831,
+      "learning_rate": 3.708624868427293e-06,
+      "loss": 0.8699,
+      "step": 12643
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.738251805305481,
+      "learning_rate": 3.70718106663985e-06,
+      "loss": 0.9156,
+      "step": 12644
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7354727983474731,
+      "learning_rate": 3.7057374820000137e-06,
+      "loss": 0.8879,
+      "step": 12645
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0547871589660645,
+      "learning_rate": 3.7042941145575915e-06,
+      "loss": 0.5625,
+      "step": 12646
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.902708649635315,
+      "learning_rate": 3.702850964362392e-06,
+      "loss": 0.8941,
+      "step": 12647
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.593868374824524,
+      "learning_rate": 3.7014080314642163e-06,
+      "loss": 0.8353,
+      "step": 12648
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6904938220977783,
+      "learning_rate": 3.699965315912858e-06,
+      "loss": 0.9223,
+      "step": 12649
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9393669366836548,
+      "learning_rate": 3.6985228177580944e-06,
+      "loss": 0.9283,
+      "step": 12650
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7096010446548462,
+      "learning_rate": 3.697080537049711e-06,
+      "loss": 0.9653,
+      "step": 12651
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7063405513763428,
+      "learning_rate": 3.695638473837466e-06,
+      "loss": 0.9289,
+      "step": 12652
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8004151582717896,
+      "learning_rate": 3.6941966281711318e-06,
+      "loss": 1.0359,
+      "step": 12653
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6368452310562134,
+      "learning_rate": 3.692755000100453e-06,
+      "loss": 0.9187,
+      "step": 12654
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.017239570617676,
+      "learning_rate": 3.691313589675185e-06,
+      "loss": 0.9316,
+      "step": 12655
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7595504522323608,
+      "learning_rate": 3.689872396945059e-06,
+      "loss": 0.9018,
+      "step": 12656
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7971736192703247,
+      "learning_rate": 3.6884314219598095e-06,
+      "loss": 1.0437,
+      "step": 12657
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7783476114273071,
+      "learning_rate": 3.6869906647691635e-06,
+      "loss": 0.9333,
+      "step": 12658
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6664040088653564,
+      "learning_rate": 3.6855501254228322e-06,
+      "loss": 0.9443,
+      "step": 12659
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.905690312385559,
+      "learning_rate": 3.6841098039705313e-06,
+      "loss": 0.8788,
+      "step": 12660
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8161495923995972,
+      "learning_rate": 3.682669700461955e-06,
+      "loss": 0.913,
+      "step": 12661
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.781017780303955,
+      "learning_rate": 3.681229814946803e-06,
+      "loss": 0.9054,
+      "step": 12662
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7194368839263916,
+      "learning_rate": 3.6797901474747567e-06,
+      "loss": 0.8599,
+      "step": 12663
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7641019821166992,
+      "learning_rate": 3.6783506980955007e-06,
+      "loss": 0.9318,
+      "step": 12664
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7111732959747314,
+      "learning_rate": 3.6769114668587e-06,
+      "loss": 0.9093,
+      "step": 12665
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7966961860656738,
+      "learning_rate": 3.675472453814025e-06,
+      "loss": 0.8629,
+      "step": 12666
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.786097526550293,
+      "learning_rate": 3.674033659011126e-06,
+      "loss": 0.9613,
+      "step": 12667
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7995284795761108,
+      "learning_rate": 3.6725950824996537e-06,
+      "loss": 0.9167,
+      "step": 12668
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8264379501342773,
+      "learning_rate": 3.6711567243292547e-06,
+      "loss": 0.9118,
+      "step": 12669
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.905930995941162,
+      "learning_rate": 3.669718584549553e-06,
+      "loss": 0.9728,
+      "step": 12670
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7546838521957397,
+      "learning_rate": 3.6682806632101852e-06,
+      "loss": 0.8725,
+      "step": 12671
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.646043062210083,
+      "learning_rate": 3.6668429603607604e-06,
+      "loss": 0.8587,
+      "step": 12672
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7822115421295166,
+      "learning_rate": 3.6654054760508983e-06,
+      "loss": 0.9413,
+      "step": 12673
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.821334958076477,
+      "learning_rate": 3.6639682103301943e-06,
+      "loss": 1.004,
+      "step": 12674
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7244815826416016,
+      "learning_rate": 3.662531163248252e-06,
+      "loss": 0.914,
+      "step": 12675
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.846912145614624,
+      "learning_rate": 3.6610943348546524e-06,
+      "loss": 0.9597,
+      "step": 12676
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6697893142700195,
+      "learning_rate": 3.659657725198984e-06,
+      "loss": 0.9108,
+      "step": 12677
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2002352476119995,
+      "learning_rate": 3.6582213343308126e-06,
+      "loss": 0.6341,
+      "step": 12678
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.776036262512207,
+      "learning_rate": 3.656785162299712e-06,
+      "loss": 0.9095,
+      "step": 12679
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7355620861053467,
+      "learning_rate": 3.6553492091552324e-06,
+      "loss": 0.9476,
+      "step": 12680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7779364585876465,
+      "learning_rate": 3.6539134749469284e-06,
+      "loss": 0.9138,
+      "step": 12681
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6813569068908691,
+      "learning_rate": 3.652477959724348e-06,
+      "loss": 0.929,
+      "step": 12682
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9758507013320923,
+      "learning_rate": 3.6510426635370178e-06,
+      "loss": 0.522,
+      "step": 12683
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7634559869766235,
+      "learning_rate": 3.6496075864344736e-06,
+      "loss": 0.9375,
+      "step": 12684
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7748652696609497,
+      "learning_rate": 3.6481727284662284e-06,
+      "loss": 0.8694,
+      "step": 12685
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6467344760894775,
+      "learning_rate": 3.6467380896818037e-06,
+      "loss": 0.8919,
+      "step": 12686
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.902411699295044,
+      "learning_rate": 3.6453036701306964e-06,
+      "loss": 0.8775,
+      "step": 12687
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.739104151725769,
+      "learning_rate": 3.643869469862412e-06,
+      "loss": 0.9339,
+      "step": 12688
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7552645206451416,
+      "learning_rate": 3.6424354889264334e-06,
+      "loss": 0.8617,
+      "step": 12689
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9040225744247437,
+      "learning_rate": 3.641001727372251e-06,
+      "loss": 0.9645,
+      "step": 12690
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7261549234390259,
+      "learning_rate": 3.6395681852493326e-06,
+      "loss": 0.9759,
+      "step": 12691
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8241993188858032,
+      "learning_rate": 3.6381348626071477e-06,
+      "loss": 0.9326,
+      "step": 12692
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8203476667404175,
+      "learning_rate": 3.6367017594951615e-06,
+      "loss": 0.8533,
+      "step": 12693
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.871741533279419,
+      "learning_rate": 3.63526887596282e-06,
+      "loss": 0.9062,
+      "step": 12694
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.611861228942871,
+      "learning_rate": 3.6338362120595726e-06,
+      "loss": 0.9197,
+      "step": 12695
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6157146692276,
+      "learning_rate": 3.6324037678348513e-06,
+      "loss": 0.8956,
+      "step": 12696
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.637628436088562,
+      "learning_rate": 3.630971543338092e-06,
+      "loss": 0.9061,
+      "step": 12697
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0322898626327515,
+      "learning_rate": 3.6295395386187103e-06,
+      "loss": 0.5595,
+      "step": 12698
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8189359903335571,
+      "learning_rate": 3.6281077537261276e-06,
+      "loss": 0.9409,
+      "step": 12699
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7638901472091675,
+      "learning_rate": 3.6266761887097433e-06,
+      "loss": 0.9291,
+      "step": 12700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0128118991851807,
+      "learning_rate": 3.625244843618965e-06,
+      "loss": 0.896,
+      "step": 12701
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7139759063720703,
+      "learning_rate": 3.6238137185031765e-06,
+      "loss": 0.9513,
+      "step": 12702
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7586177587509155,
+      "learning_rate": 3.6223828134117678e-06,
+      "loss": 0.9365,
+      "step": 12703
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7549299001693726,
+      "learning_rate": 3.6209521283941097e-06,
+      "loss": 0.8844,
+      "step": 12704
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8163676261901855,
+      "learning_rate": 3.6195216634995743e-06,
+      "loss": 0.8534,
+      "step": 12705
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9223564863204956,
+      "learning_rate": 3.6180914187775273e-06,
+      "loss": 0.875,
+      "step": 12706
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8886995315551758,
+      "learning_rate": 3.6166613942773156e-06,
+      "loss": 0.9143,
+      "step": 12707
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5925272703170776,
+      "learning_rate": 3.6152315900482904e-06,
+      "loss": 0.9418,
+      "step": 12708
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7665399312973022,
+      "learning_rate": 3.613802006139785e-06,
+      "loss": 0.9588,
+      "step": 12709
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0779120922088623,
+      "learning_rate": 3.6123726426011363e-06,
+      "loss": 0.5234,
+      "step": 12710
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6971087455749512,
+      "learning_rate": 3.6109434994816606e-06,
+      "loss": 0.8811,
+      "step": 12711
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.682674765586853,
+      "learning_rate": 3.6095145768306817e-06,
+      "loss": 0.9488,
+      "step": 12712
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0810747146606445,
+      "learning_rate": 3.6080858746974965e-06,
+      "loss": 0.5753,
+      "step": 12713
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.746949315071106,
+      "learning_rate": 3.6066573931314198e-06,
+      "loss": 0.9033,
+      "step": 12714
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6196736097335815,
+      "learning_rate": 3.6052291321817343e-06,
+      "loss": 0.8162,
+      "step": 12715
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6244525909423828,
+      "learning_rate": 3.6038010918977308e-06,
+      "loss": 0.9264,
+      "step": 12716
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6909512281417847,
+      "learning_rate": 3.602373272328682e-06,
+      "loss": 0.8933,
+      "step": 12717
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6652984619140625,
+      "learning_rate": 3.6009456735238633e-06,
+      "loss": 0.9474,
+      "step": 12718
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8534907102584839,
+      "learning_rate": 3.5995182955325313e-06,
+      "loss": 0.9021,
+      "step": 12719
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7665417194366455,
+      "learning_rate": 3.598091138403947e-06,
+      "loss": 0.9186,
+      "step": 12720
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7442777156829834,
+      "learning_rate": 3.596664202187352e-06,
+      "loss": 0.9859,
+      "step": 12721
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6241544485092163,
+      "learning_rate": 3.5952374869319884e-06,
+      "loss": 0.8833,
+      "step": 12722
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8073933124542236,
+      "learning_rate": 3.5938109926870914e-06,
+      "loss": 0.9094,
+      "step": 12723
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7068259716033936,
+      "learning_rate": 3.592384719501878e-06,
+      "loss": 0.8896,
+      "step": 12724
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6717661619186401,
+      "learning_rate": 3.5909586674255723e-06,
+      "loss": 0.9558,
+      "step": 12725
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.688665747642517,
+      "learning_rate": 3.5895328365073768e-06,
+      "loss": 0.9212,
+      "step": 12726
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8968111276626587,
+      "learning_rate": 3.5881072267965e-06,
+      "loss": 0.8936,
+      "step": 12727
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6330198049545288,
+      "learning_rate": 3.5866818383421288e-06,
+      "loss": 0.954,
+      "step": 12728
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8906724452972412,
+      "learning_rate": 3.5852566711934545e-06,
+      "loss": 0.8737,
+      "step": 12729
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8431918621063232,
+      "learning_rate": 3.5838317253996514e-06,
+      "loss": 0.9466,
+      "step": 12730
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8814911842346191,
+      "learning_rate": 3.5824070010098956e-06,
+      "loss": 0.9088,
+      "step": 12731
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1299934387207031,
+      "learning_rate": 3.5809824980733445e-06,
+      "loss": 0.5222,
+      "step": 12732
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9598451256752014,
+      "learning_rate": 3.5795582166391597e-06,
+      "loss": 0.5171,
+      "step": 12733
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8097270727157593,
+      "learning_rate": 3.578134156756482e-06,
+      "loss": 0.9194,
+      "step": 12734
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5717129707336426,
+      "learning_rate": 3.5767103184744566e-06,
+      "loss": 0.9148,
+      "step": 12735
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8004220724105835,
+      "learning_rate": 3.575286701842218e-06,
+      "loss": 0.968,
+      "step": 12736
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6115024089813232,
+      "learning_rate": 3.5738633069088857e-06,
+      "loss": 0.9142,
+      "step": 12737
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9924904108047485,
+      "learning_rate": 3.5724401337235835e-06,
+      "loss": 0.8534,
+      "step": 12738
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7243096828460693,
+      "learning_rate": 3.5710171823354145e-06,
+      "loss": 0.8912,
+      "step": 12739
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0127779245376587,
+      "learning_rate": 3.5695944527934868e-06,
+      "loss": 0.478,
+      "step": 12740
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7696268558502197,
+      "learning_rate": 3.568171945146889e-06,
+      "loss": 0.9013,
+      "step": 12741
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9781690239906311,
+      "learning_rate": 3.566749659444714e-06,
+      "loss": 0.497,
+      "step": 12742
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7435624599456787,
+      "learning_rate": 3.5653275957360333e-06,
+      "loss": 0.9278,
+      "step": 12743
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.006174325942993,
+      "learning_rate": 3.5639057540699274e-06,
+      "loss": 0.9099,
+      "step": 12744
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8570996522903442,
+      "learning_rate": 3.5624841344954508e-06,
+      "loss": 0.91,
+      "step": 12745
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8896645307540894,
+      "learning_rate": 3.5610627370616656e-06,
+      "loss": 0.8788,
+      "step": 12746
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7396416664123535,
+      "learning_rate": 3.5596415618176215e-06,
+      "loss": 0.8968,
+      "step": 12747
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8568254709243774,
+      "learning_rate": 3.558220608812354e-06,
+      "loss": 0.8122,
+      "step": 12748
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1276328563690186,
+      "learning_rate": 3.556799878094901e-06,
+      "loss": 0.5746,
+      "step": 12749
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8160136938095093,
+      "learning_rate": 3.5553793697142837e-06,
+      "loss": 0.9087,
+      "step": 12750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.771316647529602,
+      "learning_rate": 3.553959083719525e-06,
+      "loss": 0.9419,
+      "step": 12751
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8243526220321655,
+      "learning_rate": 3.552539020159629e-06,
+      "loss": 0.8467,
+      "step": 12752
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.841591477394104,
+      "learning_rate": 3.551119179083603e-06,
+      "loss": 0.955,
+      "step": 12753
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9740155339241028,
+      "learning_rate": 3.549699560540438e-06,
+      "loss": 0.5323,
+      "step": 12754
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8601003885269165,
+      "learning_rate": 3.5482801645791266e-06,
+      "loss": 0.9313,
+      "step": 12755
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7400918006896973,
+      "learning_rate": 3.5468609912486405e-06,
+      "loss": 0.9171,
+      "step": 12756
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8677492141723633,
+      "learning_rate": 3.5454420405979583e-06,
+      "loss": 1.0072,
+      "step": 12757
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.836872935295105,
+      "learning_rate": 3.544023312676039e-06,
+      "loss": 0.8738,
+      "step": 12758
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.095333218574524,
+      "learning_rate": 3.542604807531841e-06,
+      "loss": 0.5914,
+      "step": 12759
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0798290967941284,
+      "learning_rate": 3.541186525214316e-06,
+      "loss": 0.505,
+      "step": 12760
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.845782995223999,
+      "learning_rate": 3.5397684657723986e-06,
+      "loss": 0.8721,
+      "step": 12761
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8541338443756104,
+      "learning_rate": 3.5383506292550296e-06,
+      "loss": 0.9548,
+      "step": 12762
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7853888273239136,
+      "learning_rate": 3.536933015711126e-06,
+      "loss": 0.8882,
+      "step": 12763
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6120736598968506,
+      "learning_rate": 3.535515625189614e-06,
+      "loss": 0.8486,
+      "step": 12764
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8972859382629395,
+      "learning_rate": 3.5340984577393966e-06,
+      "loss": 0.939,
+      "step": 12765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.79887056350708,
+      "learning_rate": 3.532681513409384e-06,
+      "loss": 0.8357,
+      "step": 12766
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.855502963066101,
+      "learning_rate": 3.531264792248462e-06,
+      "loss": 0.8781,
+      "step": 12767
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6616597175598145,
+      "learning_rate": 3.5298482943055266e-06,
+      "loss": 0.9153,
+      "step": 12768
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.823286533355713,
+      "learning_rate": 3.5284320196294486e-06,
+      "loss": 0.8766,
+      "step": 12769
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7013206481933594,
+      "learning_rate": 3.527015968269105e-06,
+      "loss": 0.989,
+      "step": 12770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8120239973068237,
+      "learning_rate": 3.5256001402733607e-06,
+      "loss": 0.9408,
+      "step": 12771
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8070712089538574,
+      "learning_rate": 3.5241845356910688e-06,
+      "loss": 0.8923,
+      "step": 12772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9496750831604004,
+      "learning_rate": 3.5227691545710807e-06,
+      "loss": 0.9271,
+      "step": 12773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9979255795478821,
+      "learning_rate": 3.5213539969622335e-06,
+      "loss": 0.5467,
+      "step": 12774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7153676748275757,
+      "learning_rate": 3.5199390629133645e-06,
+      "loss": 0.9151,
+      "step": 12775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8439545631408691,
+      "learning_rate": 3.518524352473295e-06,
+      "loss": 0.9153,
+      "step": 12776
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5927504301071167,
+      "learning_rate": 3.5171098656908475e-06,
+      "loss": 0.8571,
+      "step": 12777
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1567935943603516,
+      "learning_rate": 3.515695602614826e-06,
+      "loss": 0.8705,
+      "step": 12778
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6935616731643677,
+      "learning_rate": 3.514281563294036e-06,
+      "loss": 0.9514,
+      "step": 12779
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.590765357017517,
+      "learning_rate": 3.5128677477772733e-06,
+      "loss": 0.8496,
+      "step": 12780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.817599892616272,
+      "learning_rate": 3.5114541561133253e-06,
+      "loss": 0.9306,
+      "step": 12781
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6007765531539917,
+      "learning_rate": 3.510040788350967e-06,
+      "loss": 0.9193,
+      "step": 12782
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6822649240493774,
+      "learning_rate": 3.5086276445389756e-06,
+      "loss": 0.9429,
+      "step": 12783
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9569311141967773,
+      "learning_rate": 3.507214724726107e-06,
+      "loss": 0.5317,
+      "step": 12784
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.951790690422058,
+      "learning_rate": 3.505802028961125e-06,
+      "loss": 1.0424,
+      "step": 12785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8282763957977295,
+      "learning_rate": 3.504389557292771e-06,
+      "loss": 0.8751,
+      "step": 12786
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7141307592391968,
+      "learning_rate": 3.5029773097697928e-06,
+      "loss": 0.8671,
+      "step": 12787
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8596477508544922,
+      "learning_rate": 3.5015652864409142e-06,
+      "loss": 0.95,
+      "step": 12788
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8929619789123535,
+      "learning_rate": 3.500153487354866e-06,
+      "loss": 0.8451,
+      "step": 12789
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6982077360153198,
+      "learning_rate": 3.4987419125603674e-06,
+      "loss": 0.9502,
+      "step": 12790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.958797574043274,
+      "learning_rate": 3.4973305621061214e-06,
+      "loss": 0.8893,
+      "step": 12791
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5980812311172485,
+      "learning_rate": 3.4959194360408368e-06,
+      "loss": 0.8945,
+      "step": 12792
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9950487017631531,
+      "learning_rate": 3.4945085344132e-06,
+      "loss": 0.5542,
+      "step": 12793
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6351646184921265,
+      "learning_rate": 3.4930978572719054e-06,
+      "loss": 0.925,
+      "step": 12794
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7312103509902954,
+      "learning_rate": 3.4916874046656235e-06,
+      "loss": 0.9859,
+      "step": 12795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7557244300842285,
+      "learning_rate": 3.490277176643033e-06,
+      "loss": 0.8722,
+      "step": 12796
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6782313585281372,
+      "learning_rate": 3.488867173252789e-06,
+      "loss": 0.8342,
+      "step": 12797
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9762368202209473,
+      "learning_rate": 3.487457394543554e-06,
+      "loss": 0.514,
+      "step": 12798
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7493423223495483,
+      "learning_rate": 3.48604784056397e-06,
+      "loss": 0.838,
+      "step": 12799
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.642878770828247,
+      "learning_rate": 3.484638511362678e-06,
+      "loss": 0.9094,
+      "step": 12800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7214213609695435,
+      "learning_rate": 3.4832294069883143e-06,
+      "loss": 0.9423,
+      "step": 12801
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6931219100952148,
+      "learning_rate": 3.4818205274894977e-06,
+      "loss": 0.932,
+      "step": 12802
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.876623272895813,
+      "learning_rate": 3.4804118729148494e-06,
+      "loss": 0.9454,
+      "step": 12803
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7917015552520752,
+      "learning_rate": 3.4790034433129727e-06,
+      "loss": 0.8338,
+      "step": 12804
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8057525157928467,
+      "learning_rate": 3.477595238732474e-06,
+      "loss": 0.8977,
+      "step": 12805
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.622389554977417,
+      "learning_rate": 3.4761872592219416e-06,
+      "loss": 0.8829,
+      "step": 12806
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0450387001037598,
+      "learning_rate": 3.474779504829966e-06,
+      "loss": 0.922,
+      "step": 12807
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8976327180862427,
+      "learning_rate": 3.473371975605119e-06,
+      "loss": 0.8312,
+      "step": 12808
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.992012619972229,
+      "learning_rate": 3.4719646715959777e-06,
+      "loss": 0.9721,
+      "step": 12809
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0269834995269775,
+      "learning_rate": 3.470557592851096e-06,
+      "loss": 0.5479,
+      "step": 12810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8272058963775635,
+      "learning_rate": 3.469150739419036e-06,
+      "loss": 1.0019,
+      "step": 12811
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7720597982406616,
+      "learning_rate": 3.467744111348338e-06,
+      "loss": 0.8357,
+      "step": 12812
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8477842807769775,
+      "learning_rate": 3.466337708687544e-06,
+      "loss": 0.929,
+      "step": 12813
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7937548160552979,
+      "learning_rate": 3.4649315314851874e-06,
+      "loss": 0.936,
+      "step": 12814
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9334877729415894,
+      "learning_rate": 3.463525579789785e-06,
+      "loss": 0.9758,
+      "step": 12815
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.843354344367981,
+      "learning_rate": 3.462119853649859e-06,
+      "loss": 0.9759,
+      "step": 12816
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7610127925872803,
+      "learning_rate": 3.460714353113912e-06,
+      "loss": 0.8868,
+      "step": 12817
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8768399953842163,
+      "learning_rate": 3.459309078230448e-06,
+      "loss": 0.9059,
+      "step": 12818
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7744067907333374,
+      "learning_rate": 3.4579040290479536e-06,
+      "loss": 0.8708,
+      "step": 12819
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6929404735565186,
+      "learning_rate": 3.4564992056149216e-06,
+      "loss": 0.9353,
+      "step": 12820
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6844855546951294,
+      "learning_rate": 3.4550946079798187e-06,
+      "loss": 0.9305,
+      "step": 12821
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7147749662399292,
+      "learning_rate": 3.4536902361911218e-06,
+      "loss": 0.9118,
+      "step": 12822
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.916273832321167,
+      "learning_rate": 3.4522860902972854e-06,
+      "loss": 0.9458,
+      "step": 12823
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.822318434715271,
+      "learning_rate": 3.4508821703467653e-06,
+      "loss": 1.0188,
+      "step": 12824
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.675005316734314,
+      "learning_rate": 3.449478476388012e-06,
+      "loss": 0.9206,
+      "step": 12825
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8033742904663086,
+      "learning_rate": 3.4480750084694537e-06,
+      "loss": 0.9396,
+      "step": 12826
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7420192956924438,
+      "learning_rate": 3.446671766639528e-06,
+      "loss": 0.9026,
+      "step": 12827
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.564064621925354,
+      "learning_rate": 3.445268750946651e-06,
+      "loss": 0.8778,
+      "step": 12828
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7758115530014038,
+      "learning_rate": 3.4438659614392423e-06,
+      "loss": 0.8397,
+      "step": 12829
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8669469356536865,
+      "learning_rate": 3.442463398165703e-06,
+      "loss": 0.8843,
+      "step": 12830
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.731524109840393,
+      "learning_rate": 3.4410610611744368e-06,
+      "loss": 0.992,
+      "step": 12831
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8694032430648804,
+      "learning_rate": 3.439658950513828e-06,
+      "loss": 0.9731,
+      "step": 12832
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9750151634216309,
+      "learning_rate": 3.4382570662322667e-06,
+      "loss": 0.9662,
+      "step": 12833
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6615134477615356,
+      "learning_rate": 3.4368554083781224e-06,
+      "loss": 0.9445,
+      "step": 12834
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.627312183380127,
+      "learning_rate": 3.4354539769997664e-06,
+      "loss": 0.9308,
+      "step": 12835
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6794954538345337,
+      "learning_rate": 3.4340527721455542e-06,
+      "loss": 0.8634,
+      "step": 12836
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7179930210113525,
+      "learning_rate": 3.432651793863838e-06,
+      "loss": 0.9342,
+      "step": 12837
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7547636032104492,
+      "learning_rate": 3.4312510422029687e-06,
+      "loss": 0.9639,
+      "step": 12838
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7991423606872559,
+      "learning_rate": 3.4298505172112716e-06,
+      "loss": 0.9634,
+      "step": 12839
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7628287076950073,
+      "learning_rate": 3.428450218937085e-06,
+      "loss": 0.9223,
+      "step": 12840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6760315895080566,
+      "learning_rate": 3.42705014742872e-06,
+      "loss": 0.8576,
+      "step": 12841
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9345563650131226,
+      "learning_rate": 3.425650302734498e-06,
+      "loss": 0.9181,
+      "step": 12842
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0413291454315186,
+      "learning_rate": 3.4242506849027146e-06,
+      "loss": 0.9273,
+      "step": 12843
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7162268161773682,
+      "learning_rate": 3.422851293981676e-06,
+      "loss": 0.8944,
+      "step": 12844
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8405197858810425,
+      "learning_rate": 3.42145213001966e-06,
+      "loss": 0.8331,
+      "step": 12845
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.737568736076355,
+      "learning_rate": 3.4200531930649607e-06,
+      "loss": 0.9446,
+      "step": 12846
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8552114963531494,
+      "learning_rate": 3.418654483165842e-06,
+      "loss": 0.8681,
+      "step": 12847
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.609439730644226,
+      "learning_rate": 3.417256000370577e-06,
+      "loss": 0.866,
+      "step": 12848
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6716490983963013,
+      "learning_rate": 3.4158577447274156e-06,
+      "loss": 0.9368,
+      "step": 12849
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6060247421264648,
+      "learning_rate": 3.4144597162846137e-06,
+      "loss": 0.9161,
+      "step": 12850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8354800939559937,
+      "learning_rate": 3.413061915090409e-06,
+      "loss": 0.9907,
+      "step": 12851
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.675417184829712,
+      "learning_rate": 3.4116643411930405e-06,
+      "loss": 0.9179,
+      "step": 12852
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6660715341567993,
+      "learning_rate": 3.4102669946407284e-06,
+      "loss": 0.8612,
+      "step": 12853
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7027852535247803,
+      "learning_rate": 3.408869875481695e-06,
+      "loss": 0.9306,
+      "step": 12854
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0397684574127197,
+      "learning_rate": 3.407472983764153e-06,
+      "loss": 1.0035,
+      "step": 12855
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8944162130355835,
+      "learning_rate": 3.406076319536301e-06,
+      "loss": 1.0013,
+      "step": 12856
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.761699914932251,
+      "learning_rate": 3.404679882846338e-06,
+      "loss": 0.8935,
+      "step": 12857
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4569480419158936,
+      "learning_rate": 3.4032836737424456e-06,
+      "loss": 0.8644,
+      "step": 12858
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7800631523132324,
+      "learning_rate": 3.4018876922728105e-06,
+      "loss": 0.8816,
+      "step": 12859
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7472339868545532,
+      "learning_rate": 3.400491938485596e-06,
+      "loss": 0.9345,
+      "step": 12860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7179112434387207,
+      "learning_rate": 3.399096412428974e-06,
+      "loss": 0.9177,
+      "step": 12861
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0143840312957764,
+      "learning_rate": 3.3977011141510917e-06,
+      "loss": 0.5033,
+      "step": 12862
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7133642435073853,
+      "learning_rate": 3.396306043700105e-06,
+      "loss": 0.9369,
+      "step": 12863
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7558554410934448,
+      "learning_rate": 3.394911201124147e-06,
+      "loss": 0.9346,
+      "step": 12864
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7255535125732422,
+      "learning_rate": 3.393516586471356e-06,
+      "loss": 0.8891,
+      "step": 12865
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8447771072387695,
+      "learning_rate": 3.392122199789849e-06,
+      "loss": 0.9164,
+      "step": 12866
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7963228225708008,
+      "learning_rate": 3.3907280411277478e-06,
+      "loss": 0.8766,
+      "step": 12867
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7364304065704346,
+      "learning_rate": 3.3893341105331612e-06,
+      "loss": 0.8031,
+      "step": 12868
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.608424425125122,
+      "learning_rate": 3.3879404080541866e-06,
+      "loss": 0.8474,
+      "step": 12869
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7260738611221313,
+      "learning_rate": 3.386546933738921e-06,
+      "loss": 0.8671,
+      "step": 12870
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8440991640090942,
+      "learning_rate": 3.385153687635444e-06,
+      "loss": 1.015,
+      "step": 12871
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9038454294204712,
+      "learning_rate": 3.383760669791838e-06,
+      "loss": 0.9003,
+      "step": 12872
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6984950304031372,
+      "learning_rate": 3.3823678802561677e-06,
+      "loss": 0.9703,
+      "step": 12873
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7837727069854736,
+      "learning_rate": 3.3809753190764983e-06,
+      "loss": 0.9246,
+      "step": 12874
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.101508378982544,
+      "learning_rate": 3.3795829863008777e-06,
+      "loss": 0.5568,
+      "step": 12875
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.751475214958191,
+      "learning_rate": 3.378190881977359e-06,
+      "loss": 0.8941,
+      "step": 12876
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8393185138702393,
+      "learning_rate": 3.376799006153971e-06,
+      "loss": 0.979,
+      "step": 12877
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.878891110420227,
+      "learning_rate": 3.3754073588787494e-06,
+      "loss": 0.9126,
+      "step": 12878
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8797088861465454,
+      "learning_rate": 3.3740159401997173e-06,
+      "loss": 0.8809,
+      "step": 12879
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0130059719085693,
+      "learning_rate": 3.3726247501648846e-06,
+      "loss": 0.5417,
+      "step": 12880
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6608471870422363,
+      "learning_rate": 3.37123378882226e-06,
+      "loss": 0.8555,
+      "step": 12881
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7912064790725708,
+      "learning_rate": 3.369843056219839e-06,
+      "loss": 0.9811,
+      "step": 12882
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8159829378128052,
+      "learning_rate": 3.3684525524056156e-06,
+      "loss": 0.828,
+      "step": 12883
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7412052154541016,
+      "learning_rate": 3.3670622774275676e-06,
+      "loss": 0.8353,
+      "step": 12884
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7044848203659058,
+      "learning_rate": 3.3656722313336755e-06,
+      "loss": 0.806,
+      "step": 12885
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.548651933670044,
+      "learning_rate": 3.3642824141718986e-06,
+      "loss": 0.9308,
+      "step": 12886
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6368460655212402,
+      "learning_rate": 3.362892825990203e-06,
+      "loss": 1.0051,
+      "step": 12887
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8173948526382446,
+      "learning_rate": 3.361503466836532e-06,
+      "loss": 0.9137,
+      "step": 12888
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1178308725357056,
+      "learning_rate": 3.3601143367588362e-06,
+      "loss": 0.5839,
+      "step": 12889
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6334292888641357,
+      "learning_rate": 3.358725435805045e-06,
+      "loss": 0.9938,
+      "step": 12890
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.720916509628296,
+      "learning_rate": 3.3573367640230846e-06,
+      "loss": 0.9691,
+      "step": 12891
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7262259721755981,
+      "learning_rate": 3.3559483214608822e-06,
+      "loss": 0.9261,
+      "step": 12892
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.726747989654541,
+      "learning_rate": 3.3545601081663405e-06,
+      "loss": 0.9819,
+      "step": 12893
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6461094617843628,
+      "learning_rate": 3.3531721241873684e-06,
+      "loss": 0.9791,
+      "step": 12894
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6887973546981812,
+      "learning_rate": 3.3517843695718567e-06,
+      "loss": 0.8968,
+      "step": 12895
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.913801670074463,
+      "learning_rate": 3.350396844367698e-06,
+      "loss": 0.8532,
+      "step": 12896
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8553576469421387,
+      "learning_rate": 3.349009548622767e-06,
+      "loss": 0.9709,
+      "step": 12897
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8029811382293701,
+      "learning_rate": 3.34762248238494e-06,
+      "loss": 1.0315,
+      "step": 12898
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7179360389709473,
+      "learning_rate": 3.3462356457020762e-06,
+      "loss": 0.8922,
+      "step": 12899
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7548192739486694,
+      "learning_rate": 3.3448490386220355e-06,
+      "loss": 0.8201,
+      "step": 12900
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.611159086227417,
+      "learning_rate": 3.3434626611926625e-06,
+      "loss": 0.9366,
+      "step": 12901
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0283126831054688,
+      "learning_rate": 3.3420765134618006e-06,
+      "loss": 1.0067,
+      "step": 12902
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7182762622833252,
+      "learning_rate": 3.340690595477277e-06,
+      "loss": 0.9121,
+      "step": 12903
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.576112985610962,
+      "learning_rate": 3.3393049072869198e-06,
+      "loss": 0.8513,
+      "step": 12904
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.735619068145752,
+      "learning_rate": 3.337919448938547e-06,
+      "loss": 0.8703,
+      "step": 12905
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8262648582458496,
+      "learning_rate": 3.3365342204799613e-06,
+      "loss": 0.9832,
+      "step": 12906
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.909550666809082,
+      "learning_rate": 3.33514922195897e-06,
+      "loss": 0.9173,
+      "step": 12907
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7568025588989258,
+      "learning_rate": 3.333764453423357e-06,
+      "loss": 0.9044,
+      "step": 12908
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8140579462051392,
+      "learning_rate": 3.332379914920915e-06,
+      "loss": 0.907,
+      "step": 12909
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7118470668792725,
+      "learning_rate": 3.330995606499413e-06,
+      "loss": 0.8968,
+      "step": 12910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.729712963104248,
+      "learning_rate": 3.3296115282066245e-06,
+      "loss": 0.9335,
+      "step": 12911
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.75454580783844,
+      "learning_rate": 3.328227680090309e-06,
+      "loss": 0.8658,
+      "step": 12912
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0110607147216797,
+      "learning_rate": 3.3268440621982222e-06,
+      "loss": 0.9002,
+      "step": 12913
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8920130729675293,
+      "learning_rate": 3.3254606745781026e-06,
+      "loss": 0.8568,
+      "step": 12914
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8047552108764648,
+      "learning_rate": 3.3240775172776952e-06,
+      "loss": 0.9476,
+      "step": 12915
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.833772897720337,
+      "learning_rate": 3.3226945903447196e-06,
+      "loss": 0.916,
+      "step": 12916
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.830235481262207,
+      "learning_rate": 3.321311893826905e-06,
+      "loss": 0.8187,
+      "step": 12917
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.154555082321167,
+      "learning_rate": 3.3199294277719573e-06,
+      "loss": 0.5193,
+      "step": 12918
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.790439248085022,
+      "learning_rate": 3.318547192227589e-06,
+      "loss": 0.8954,
+      "step": 12919
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.786900281906128,
+      "learning_rate": 3.31716518724149e-06,
+      "loss": 0.9008,
+      "step": 12920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0621657371520996,
+      "learning_rate": 3.315783412861352e-06,
+      "loss": 0.5529,
+      "step": 12921
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0171537399291992,
+      "learning_rate": 3.3144018691348602e-06,
+      "loss": 0.5673,
+      "step": 12922
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.809259295463562,
+      "learning_rate": 3.3130205561096818e-06,
+      "loss": 0.948,
+      "step": 12923
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.755825161933899,
+      "learning_rate": 3.311639473833487e-06,
+      "loss": 0.9291,
+      "step": 12924
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.730514645576477,
+      "learning_rate": 3.310258622353928e-06,
+      "loss": 0.9534,
+      "step": 12925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7440922260284424,
+      "learning_rate": 3.3088780017186608e-06,
+      "loss": 0.8313,
+      "step": 12926
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7911334037780762,
+      "learning_rate": 3.3074976119753178e-06,
+      "loss": 0.8897,
+      "step": 12927
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6797412633895874,
+      "learning_rate": 3.3061174531715425e-06,
+      "loss": 0.8669,
+      "step": 12928
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.809531569480896,
+      "learning_rate": 3.304737525354951e-06,
+      "loss": 0.9346,
+      "step": 12929
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5755363702774048,
+      "learning_rate": 3.3033578285731693e-06,
+      "loss": 0.8273,
+      "step": 12930
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.689409613609314,
+      "learning_rate": 3.301978362873798e-06,
+      "loss": 0.8505,
+      "step": 12931
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9128230810165405,
+      "learning_rate": 3.3005991283044436e-06,
+      "loss": 0.9701,
+      "step": 12932
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6728066205978394,
+      "learning_rate": 3.2992201249127033e-06,
+      "loss": 0.9033,
+      "step": 12933
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9323869943618774,
+      "learning_rate": 3.2978413527461552e-06,
+      "loss": 0.8976,
+      "step": 12934
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7050065994262695,
+      "learning_rate": 3.2964628118523832e-06,
+      "loss": 0.8384,
+      "step": 12935
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7353442907333374,
+      "learning_rate": 3.295084502278951e-06,
+      "loss": 0.9225,
+      "step": 12936
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8594201803207397,
+      "learning_rate": 3.2937064240734262e-06,
+      "loss": 0.924,
+      "step": 12937
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7685085535049438,
+      "learning_rate": 3.292328577283356e-06,
+      "loss": 0.8707,
+      "step": 12938
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7772799730300903,
+      "learning_rate": 3.290950961956293e-06,
+      "loss": 0.8769,
+      "step": 12939
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.679996132850647,
+      "learning_rate": 3.289573578139769e-06,
+      "loss": 0.9618,
+      "step": 12940
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7996190786361694,
+      "learning_rate": 3.2881964258813172e-06,
+      "loss": 0.9067,
+      "step": 12941
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6842286586761475,
+      "learning_rate": 3.2868195052284557e-06,
+      "loss": 0.8927,
+      "step": 12942
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8408339023590088,
+      "learning_rate": 3.2854428162287046e-06,
+      "loss": 0.9548,
+      "step": 12943
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7755848169326782,
+      "learning_rate": 3.2840663589295617e-06,
+      "loss": 0.9062,
+      "step": 12944
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5631972551345825,
+      "learning_rate": 3.282690133378529e-06,
+      "loss": 0.7774,
+      "step": 12945
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7434961795806885,
+      "learning_rate": 3.2813141396230986e-06,
+      "loss": 0.9717,
+      "step": 12946
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7239450216293335,
+      "learning_rate": 3.2799383777107453e-06,
+      "loss": 0.9651,
+      "step": 12947
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7930244207382202,
+      "learning_rate": 3.278562847688951e-06,
+      "loss": 0.9002,
+      "step": 12948
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7142078876495361,
+      "learning_rate": 3.2771875496051743e-06,
+      "loss": 0.9856,
+      "step": 12949
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6232517957687378,
+      "learning_rate": 3.275812483506878e-06,
+      "loss": 0.8529,
+      "step": 12950
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8481017351150513,
+      "learning_rate": 3.2744376494415075e-06,
+      "loss": 0.9415,
+      "step": 12951
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7313182353973389,
+      "learning_rate": 3.2730630474565096e-06,
+      "loss": 0.9001,
+      "step": 12952
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7166863679885864,
+      "learning_rate": 3.2716886775993117e-06,
+      "loss": 0.9696,
+      "step": 12953
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6979272365570068,
+      "learning_rate": 3.2703145399173453e-06,
+      "loss": 0.9094,
+      "step": 12954
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.65584135055542,
+      "learning_rate": 3.2689406344580233e-06,
+      "loss": 0.8902,
+      "step": 12955
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7679846286773682,
+      "learning_rate": 3.2675669612687565e-06,
+      "loss": 0.9014,
+      "step": 12956
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5945175886154175,
+      "learning_rate": 3.2661935203969518e-06,
+      "loss": 0.8953,
+      "step": 12957
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7289154529571533,
+      "learning_rate": 3.264820311889996e-06,
+      "loss": 0.8785,
+      "step": 12958
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2390129566192627,
+      "learning_rate": 3.263447335795279e-06,
+      "loss": 0.8568,
+      "step": 12959
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6853278875350952,
+      "learning_rate": 3.2620745921601737e-06,
+      "loss": 0.9589,
+      "step": 12960
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7010923624038696,
+      "learning_rate": 3.2607020810320558e-06,
+      "loss": 0.9058,
+      "step": 12961
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.695469617843628,
+      "learning_rate": 3.259329802458281e-06,
+      "loss": 0.917,
+      "step": 12962
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7931337356567383,
+      "learning_rate": 3.2579577564862076e-06,
+      "loss": 0.8935,
+      "step": 12963
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7772189378738403,
+      "learning_rate": 3.2565859431631765e-06,
+      "loss": 0.8834,
+      "step": 12964
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9803166389465332,
+      "learning_rate": 3.2552143625365306e-06,
+      "loss": 0.9446,
+      "step": 12965
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8056056499481201,
+      "learning_rate": 3.2538430146535927e-06,
+      "loss": 0.8732,
+      "step": 12966
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7885921001434326,
+      "learning_rate": 3.2524718995616913e-06,
+      "loss": 0.8989,
+      "step": 12967
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7689409255981445,
+      "learning_rate": 3.2511010173081327e-06,
+      "loss": 0.9502,
+      "step": 12968
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.860883355140686,
+      "learning_rate": 3.2497303679402258e-06,
+      "loss": 0.93,
+      "step": 12969
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7048285007476807,
+      "learning_rate": 3.2483599515052723e-06,
+      "loss": 0.8876,
+      "step": 12970
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.713229775428772,
+      "learning_rate": 3.2469897680505515e-06,
+      "loss": 0.8596,
+      "step": 12971
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6767103672027588,
+      "learning_rate": 3.2456198176233545e-06,
+      "loss": 0.9543,
+      "step": 12972
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7357815504074097,
+      "learning_rate": 3.244250100270947e-06,
+      "loss": 0.8644,
+      "step": 12973
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.882466435432434,
+      "learning_rate": 3.242880616040599e-06,
+      "loss": 0.8935,
+      "step": 12974
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.663617730140686,
+      "learning_rate": 3.241511364979564e-06,
+      "loss": 0.8514,
+      "step": 12975
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.702240228652954,
+      "learning_rate": 3.2401423471350955e-06,
+      "loss": 0.9161,
+      "step": 12976
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.813599705696106,
+      "learning_rate": 3.238773562554425e-06,
+      "loss": 0.8906,
+      "step": 12977
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5721182823181152,
+      "learning_rate": 3.237405011284799e-06,
+      "loss": 0.9431,
+      "step": 12978
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8185781240463257,
+      "learning_rate": 3.236036693373431e-06,
+      "loss": 0.9359,
+      "step": 12979
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7295923233032227,
+      "learning_rate": 3.234668608867547e-06,
+      "loss": 0.853,
+      "step": 12980
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7073886394500732,
+      "learning_rate": 3.2333007578143473e-06,
+      "loss": 0.8379,
+      "step": 12981
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.779390573501587,
+      "learning_rate": 3.2319331402610397e-06,
+      "loss": 0.889,
+      "step": 12982
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8085041046142578,
+      "learning_rate": 3.23056575625481e-06,
+      "loss": 0.8762,
+      "step": 12983
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6807066202163696,
+      "learning_rate": 3.2291986058428506e-06,
+      "loss": 0.9476,
+      "step": 12984
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6979188919067383,
+      "learning_rate": 3.2278316890723293e-06,
+      "loss": 0.8902,
+      "step": 12985
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.868937611579895,
+      "learning_rate": 3.2264650059904203e-06,
+      "loss": 0.9113,
+      "step": 12986
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.90029776096344,
+      "learning_rate": 3.225098556644286e-06,
+      "loss": 0.8997,
+      "step": 12987
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7030693292617798,
+      "learning_rate": 3.2237323410810717e-06,
+      "loss": 0.921,
+      "step": 12988
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0223934650421143,
+      "learning_rate": 3.2223663593479293e-06,
+      "loss": 0.8821,
+      "step": 12989
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7602988481521606,
+      "learning_rate": 3.221000611491988e-06,
+      "loss": 0.8874,
+      "step": 12990
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.736777424812317,
+      "learning_rate": 3.219635097560382e-06,
+      "loss": 0.9198,
+      "step": 12991
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8135632276535034,
+      "learning_rate": 3.218269817600226e-06,
+      "loss": 1.0182,
+      "step": 12992
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9046096801757812,
+      "learning_rate": 3.2169047716586364e-06,
+      "loss": 0.9709,
+      "step": 12993
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.081299901008606,
+      "learning_rate": 3.215539959782714e-06,
+      "loss": 0.5018,
+      "step": 12994
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8905421495437622,
+      "learning_rate": 3.2141753820195588e-06,
+      "loss": 0.8946,
+      "step": 12995
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0486741065979004,
+      "learning_rate": 3.2128110384162515e-06,
+      "loss": 0.8801,
+      "step": 12996
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7285798788070679,
+      "learning_rate": 3.21144692901988e-06,
+      "loss": 0.9606,
+      "step": 12997
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8632614612579346,
+      "learning_rate": 3.2100830538775086e-06,
+      "loss": 0.8923,
+      "step": 12998
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7425857782363892,
+      "learning_rate": 3.2087194130362033e-06,
+      "loss": 0.8199,
+      "step": 12999
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4580556154251099,
+      "learning_rate": 3.207356006543024e-06,
+      "loss": 0.8738,
+      "step": 13000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.835984706878662,
+      "learning_rate": 3.205992834445012e-06,
+      "loss": 0.896,
+      "step": 13001
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8194921016693115,
+      "learning_rate": 3.204629896789212e-06,
+      "loss": 0.8516,
+      "step": 13002
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9852017164230347,
+      "learning_rate": 3.203267193622649e-06,
+      "loss": 0.5772,
+      "step": 13003
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8939801454544067,
+      "learning_rate": 3.201904724992352e-06,
+      "loss": 0.9859,
+      "step": 13004
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9345558881759644,
+      "learning_rate": 3.2005424909453297e-06,
+      "loss": 0.9309,
+      "step": 13005
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8502147197723389,
+      "learning_rate": 3.199180491528597e-06,
+      "loss": 0.868,
+      "step": 13006
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.725514531135559,
+      "learning_rate": 3.197818726789144e-06,
+      "loss": 0.9148,
+      "step": 13007
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6633045673370361,
+      "learning_rate": 3.1964571967739687e-06,
+      "loss": 0.8883,
+      "step": 13008
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.844893217086792,
+      "learning_rate": 3.1950959015300486e-06,
+      "loss": 0.8763,
+      "step": 13009
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9353128671646118,
+      "learning_rate": 3.1937348411043588e-06,
+      "loss": 0.935,
+      "step": 13010
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7447443008422852,
+      "learning_rate": 3.192374015543871e-06,
+      "loss": 0.8482,
+      "step": 13011
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7648658752441406,
+      "learning_rate": 3.191013424895536e-06,
+      "loss": 0.9077,
+      "step": 13012
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7657307386398315,
+      "learning_rate": 3.189653069206311e-06,
+      "loss": 0.9166,
+      "step": 13013
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6768075227737427,
+      "learning_rate": 3.1882929485231316e-06,
+      "loss": 0.8811,
+      "step": 13014
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8069775104522705,
+      "learning_rate": 3.1869330628929385e-06,
+      "loss": 0.9,
+      "step": 13015
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7847970724105835,
+      "learning_rate": 3.1855734123626493e-06,
+      "loss": 0.9421,
+      "step": 13016
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7322032451629639,
+      "learning_rate": 3.1842139969791907e-06,
+      "loss": 0.9391,
+      "step": 13017
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6978659629821777,
+      "learning_rate": 3.182854816789465e-06,
+      "loss": 0.9857,
+      "step": 13018
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6952332258224487,
+      "learning_rate": 3.181495871840379e-06,
+      "loss": 0.8781,
+      "step": 13019
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6562113761901855,
+      "learning_rate": 3.1801371621788203e-06,
+      "loss": 0.9255,
+      "step": 13020
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7698861360549927,
+      "learning_rate": 3.1787786878516813e-06,
+      "loss": 1.0115,
+      "step": 13021
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.94884192943573,
+      "learning_rate": 3.1774204489058313e-06,
+      "loss": 0.9073,
+      "step": 13022
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6540274620056152,
+      "learning_rate": 3.176062445388145e-06,
+      "loss": 0.8815,
+      "step": 13023
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8833669424057007,
+      "learning_rate": 3.1747046773454838e-06,
+      "loss": 0.9069,
+      "step": 13024
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9806326627731323,
+      "learning_rate": 3.1733471448246968e-06,
+      "loss": 0.5096,
+      "step": 13025
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.677437424659729,
+      "learning_rate": 3.171989847872632e-06,
+      "loss": 0.9004,
+      "step": 13026
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7745624780654907,
+      "learning_rate": 3.1706327865361218e-06,
+      "loss": 0.9371,
+      "step": 13027
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6608716249465942,
+      "learning_rate": 3.1692759608620004e-06,
+      "loss": 0.962,
+      "step": 13028
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6462064981460571,
+      "learning_rate": 3.167919370897081e-06,
+      "loss": 0.8764,
+      "step": 13029
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8811894655227661,
+      "learning_rate": 3.1665630166881833e-06,
+      "loss": 0.9224,
+      "step": 13030
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.739756464958191,
+      "learning_rate": 3.165206898282104e-06,
+      "loss": 0.8838,
+      "step": 13031
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.033450961112976,
+      "learning_rate": 3.1638510157256453e-06,
+      "loss": 0.5078,
+      "step": 13032
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6108418703079224,
+      "learning_rate": 3.162495369065589e-06,
+      "loss": 0.8754,
+      "step": 13033
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6332170963287354,
+      "learning_rate": 3.1611399583487213e-06,
+      "loss": 0.8445,
+      "step": 13034
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6060373783111572,
+      "learning_rate": 3.1597847836218054e-06,
+      "loss": 0.8578,
+      "step": 13035
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8608494997024536,
+      "learning_rate": 3.158429844931611e-06,
+      "loss": 0.9151,
+      "step": 13036
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9385156631469727,
+      "learning_rate": 3.1570751423248935e-06,
+      "loss": 0.9643,
+      "step": 13037
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.728148102760315,
+      "learning_rate": 3.155720675848396e-06,
+      "loss": 0.9174,
+      "step": 13038
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7340481281280518,
+      "learning_rate": 3.154366445548861e-06,
+      "loss": 0.9254,
+      "step": 13039
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9110102653503418,
+      "learning_rate": 3.1530124514730155e-06,
+      "loss": 0.9293,
+      "step": 13040
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9378286600112915,
+      "learning_rate": 3.1516586936675863e-06,
+      "loss": 0.9301,
+      "step": 13041
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9784626960754395,
+      "learning_rate": 3.1503051721792833e-06,
+      "loss": 0.9802,
+      "step": 13042
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7403974533081055,
+      "learning_rate": 3.148951887054814e-06,
+      "loss": 0.8985,
+      "step": 13043
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2266108989715576,
+      "learning_rate": 3.1475988383408774e-06,
+      "loss": 0.8071,
+      "step": 13044
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7259621620178223,
+      "learning_rate": 3.1462460260841675e-06,
+      "loss": 0.9289,
+      "step": 13045
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8038628101348877,
+      "learning_rate": 3.1448934503313588e-06,
+      "loss": 0.864,
+      "step": 13046
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7620233297348022,
+      "learning_rate": 3.1435411111291304e-06,
+      "loss": 0.952,
+      "step": 13047
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9062740802764893,
+      "learning_rate": 3.1421890085241437e-06,
+      "loss": 0.9023,
+      "step": 13048
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7724922895431519,
+      "learning_rate": 3.14083714256306e-06,
+      "loss": 0.8166,
+      "step": 13049
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7863596677780151,
+      "learning_rate": 3.139485513292523e-06,
+      "loss": 0.9269,
+      "step": 13050
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7511223554611206,
+      "learning_rate": 3.1381341207591797e-06,
+      "loss": 0.8416,
+      "step": 13051
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6451618671417236,
+      "learning_rate": 3.136782965009658e-06,
+      "loss": 0.8982,
+      "step": 13052
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.807515263557434,
+      "learning_rate": 3.135432046090584e-06,
+      "loss": 0.8506,
+      "step": 13053
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9561550617218018,
+      "learning_rate": 3.1340813640485777e-06,
+      "loss": 0.9234,
+      "step": 13054
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7279635667800903,
+      "learning_rate": 3.1327309189302415e-06,
+      "loss": 0.9816,
+      "step": 13055
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6455328464508057,
+      "learning_rate": 3.1313807107821815e-06,
+      "loss": 0.8733,
+      "step": 13056
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6129493713378906,
+      "learning_rate": 3.1300307396509833e-06,
+      "loss": 0.7935,
+      "step": 13057
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6675885915756226,
+      "learning_rate": 3.128681005583236e-06,
+      "loss": 0.9505,
+      "step": 13058
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9253281354904175,
+      "learning_rate": 3.1273315086255106e-06,
+      "loss": 0.9616,
+      "step": 13059
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.643847942352295,
+      "learning_rate": 3.1259822488243805e-06,
+      "loss": 0.8076,
+      "step": 13060
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8107280731201172,
+      "learning_rate": 3.1246332262263977e-06,
+      "loss": 0.9682,
+      "step": 13061
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8633383512496948,
+      "learning_rate": 3.123284440878119e-06,
+      "loss": 0.8906,
+      "step": 13062
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.893514633178711,
+      "learning_rate": 3.1219358928260823e-06,
+      "loss": 0.9793,
+      "step": 13063
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8098845481872559,
+      "learning_rate": 3.120587582116825e-06,
+      "loss": 0.8851,
+      "step": 13064
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9624648094177246,
+      "learning_rate": 3.1192395087968775e-06,
+      "loss": 0.9503,
+      "step": 13065
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.692080020904541,
+      "learning_rate": 3.1178916729127497e-06,
+      "loss": 0.8306,
+      "step": 13066
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.879489541053772,
+      "learning_rate": 3.116544074510959e-06,
+      "loss": 0.8684,
+      "step": 13067
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7086669206619263,
+      "learning_rate": 3.115196713638e-06,
+      "loss": 0.9053,
+      "step": 13068
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8532218933105469,
+      "learning_rate": 3.1138495903403754e-06,
+      "loss": 0.8797,
+      "step": 13069
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6436641216278076,
+      "learning_rate": 3.1125027046645616e-06,
+      "loss": 0.9979,
+      "step": 13070
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8048409223556519,
+      "learning_rate": 3.111156056657044e-06,
+      "loss": 0.8947,
+      "step": 13071
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8001627922058105,
+      "learning_rate": 3.1098096463642834e-06,
+      "loss": 0.8906,
+      "step": 13072
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.830783724784851,
+      "learning_rate": 3.108463473832749e-06,
+      "loss": 0.8569,
+      "step": 13073
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7894561290740967,
+      "learning_rate": 3.1071175391088857e-06,
+      "loss": 0.8195,
+      "step": 13074
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7809464931488037,
+      "learning_rate": 3.105771842239146e-06,
+      "loss": 0.8661,
+      "step": 13075
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8369849920272827,
+      "learning_rate": 3.1044263832699574e-06,
+      "loss": 0.9988,
+      "step": 13076
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9826326370239258,
+      "learning_rate": 3.103081162247752e-06,
+      "loss": 0.5467,
+      "step": 13077
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.825492024421692,
+      "learning_rate": 3.1017361792189537e-06,
+      "loss": 0.9279,
+      "step": 13078
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6958202123641968,
+      "learning_rate": 3.100391434229967e-06,
+      "loss": 0.8574,
+      "step": 13079
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9676889181137085,
+      "learning_rate": 3.0990469273272016e-06,
+      "loss": 0.8827,
+      "step": 13080
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9030475616455078,
+      "learning_rate": 3.0977026585570467e-06,
+      "loss": 0.9926,
+      "step": 13081
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.893059492111206,
+      "learning_rate": 3.0963586279658963e-06,
+      "loss": 0.9696,
+      "step": 13082
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7339398860931396,
+      "learning_rate": 3.095014835600121e-06,
+      "loss": 0.9039,
+      "step": 13083
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6585297584533691,
+      "learning_rate": 3.093671281506099e-06,
+      "loss": 0.9193,
+      "step": 13084
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7605347633361816,
+      "learning_rate": 3.0923279657301853e-06,
+      "loss": 0.868,
+      "step": 13085
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6108390092849731,
+      "learning_rate": 3.090984888318741e-06,
+      "loss": 0.8874,
+      "step": 13086
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7320451736450195,
+      "learning_rate": 3.0896420493181058e-06,
+      "loss": 0.9975,
+      "step": 13087
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9116482734680176,
+      "learning_rate": 3.0882994487746233e-06,
+      "loss": 0.9738,
+      "step": 13088
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7406176328659058,
+      "learning_rate": 3.0869570867346167e-06,
+      "loss": 0.9516,
+      "step": 13089
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8883994817733765,
+      "learning_rate": 3.08561496324441e-06,
+      "loss": 0.9032,
+      "step": 13090
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8087818622589111,
+      "learning_rate": 3.0842730783503195e-06,
+      "loss": 0.8918,
+      "step": 13091
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7573754787445068,
+      "learning_rate": 3.0829314320986436e-06,
+      "loss": 0.8815,
+      "step": 13092
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8649489879608154,
+      "learning_rate": 3.0815900245356857e-06,
+      "loss": 0.9297,
+      "step": 13093
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9662258625030518,
+      "learning_rate": 3.0802488557077257e-06,
+      "loss": 0.5492,
+      "step": 13094
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8766885995864868,
+      "learning_rate": 3.078907925661052e-06,
+      "loss": 0.9706,
+      "step": 13095
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7519360780715942,
+      "learning_rate": 3.0775672344419305e-06,
+      "loss": 0.9203,
+      "step": 13096
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7163918018341064,
+      "learning_rate": 3.0762267820966285e-06,
+      "loss": 0.8335,
+      "step": 13097
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8738588094711304,
+      "learning_rate": 3.074886568671397e-06,
+      "loss": 0.9792,
+      "step": 13098
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6981703042984009,
+      "learning_rate": 3.0735465942124877e-06,
+      "loss": 0.9329,
+      "step": 13099
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8276050090789795,
+      "learning_rate": 3.0722068587661346e-06,
+      "loss": 0.8767,
+      "step": 13100
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7266253232955933,
+      "learning_rate": 3.0708673623785713e-06,
+      "loss": 0.9123,
+      "step": 13101
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7737120389938354,
+      "learning_rate": 3.0695281050960224e-06,
+      "loss": 0.8927,
+      "step": 13102
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.770328164100647,
+      "learning_rate": 3.0681890869646957e-06,
+      "loss": 0.9232,
+      "step": 13103
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7303805351257324,
+      "learning_rate": 3.066850308030803e-06,
+      "loss": 0.8801,
+      "step": 13104
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.819445013999939,
+      "learning_rate": 3.0655117683405378e-06,
+      "loss": 0.9607,
+      "step": 13105
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.722355842590332,
+      "learning_rate": 3.0641734679400925e-06,
+      "loss": 0.9052,
+      "step": 13106
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8268797397613525,
+      "learning_rate": 3.062835406875643e-06,
+      "loss": 0.9597,
+      "step": 13107
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6671241521835327,
+      "learning_rate": 3.0614975851933694e-06,
+      "loss": 0.8755,
+      "step": 13108
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6749218702316284,
+      "learning_rate": 3.060160002939425e-06,
+      "loss": 0.9283,
+      "step": 13109
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7422959804534912,
+      "learning_rate": 3.0588226601599803e-06,
+      "loss": 0.9027,
+      "step": 13110
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7745940685272217,
+      "learning_rate": 3.057485556901173e-06,
+      "loss": 0.8485,
+      "step": 13111
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1069972515106201,
+      "learning_rate": 3.0561486932091487e-06,
+      "loss": 0.5914,
+      "step": 13112
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6952247619628906,
+      "learning_rate": 3.0548120691300344e-06,
+      "loss": 0.9315,
+      "step": 13113
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8256460428237915,
+      "learning_rate": 3.0534756847099567e-06,
+      "loss": 0.852,
+      "step": 13114
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9599213600158691,
+      "learning_rate": 3.052139539995026e-06,
+      "loss": 1.0084,
+      "step": 13115
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.946385145187378,
+      "learning_rate": 3.0508036350313553e-06,
+      "loss": 0.8959,
+      "step": 13116
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.651458501815796,
+      "learning_rate": 3.0494679698650353e-06,
+      "loss": 0.8646,
+      "step": 13117
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7008143663406372,
+      "learning_rate": 3.0481325445421604e-06,
+      "loss": 0.9033,
+      "step": 13118
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6350711584091187,
+      "learning_rate": 3.0467973591088163e-06,
+      "loss": 0.8603,
+      "step": 13119
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.908288598060608,
+      "learning_rate": 3.0454624136110676e-06,
+      "loss": 0.9841,
+      "step": 13120
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8394027948379517,
+      "learning_rate": 3.0441277080949883e-06,
+      "loss": 0.8164,
+      "step": 13121
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.020101308822632,
+      "learning_rate": 3.0427932426066286e-06,
+      "loss": 0.9294,
+      "step": 13122
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7723796367645264,
+      "learning_rate": 3.041459017192042e-06,
+      "loss": 0.8761,
+      "step": 13123
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7550098896026611,
+      "learning_rate": 3.0401250318972643e-06,
+      "loss": 0.8431,
+      "step": 13124
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9901959300041199,
+      "learning_rate": 3.0387912867683334e-06,
+      "loss": 0.5301,
+      "step": 13125
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7972277402877808,
+      "learning_rate": 3.037457781851266e-06,
+      "loss": 0.9158,
+      "step": 13126
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.725225806236267,
+      "learning_rate": 3.0361245171920862e-06,
+      "loss": 0.8734,
+      "step": 13127
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5979008674621582,
+      "learning_rate": 3.0347914928367917e-06,
+      "loss": 0.8932,
+      "step": 13128
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6955655813217163,
+      "learning_rate": 3.0334587088313903e-06,
+      "loss": 0.9194,
+      "step": 13129
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.872408151626587,
+      "learning_rate": 3.0321261652218647e-06,
+      "loss": 0.8756,
+      "step": 13130
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6361650228500366,
+      "learning_rate": 3.0307938620542023e-06,
+      "loss": 0.9277,
+      "step": 13131
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7348065376281738,
+      "learning_rate": 3.029461799374378e-06,
+      "loss": 0.9003,
+      "step": 13132
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9422408938407898,
+      "learning_rate": 3.0281299772283534e-06,
+      "loss": 0.5573,
+      "step": 13133
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7520490884780884,
+      "learning_rate": 3.0267983956620907e-06,
+      "loss": 0.9474,
+      "step": 13134
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8453209400177002,
+      "learning_rate": 3.025467054721534e-06,
+      "loss": 0.9664,
+      "step": 13135
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.849521279335022,
+      "learning_rate": 3.0241359544526296e-06,
+      "loss": 0.9295,
+      "step": 13136
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7623339891433716,
+      "learning_rate": 3.0228050949013033e-06,
+      "loss": 0.8474,
+      "step": 13137
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6508996486663818,
+      "learning_rate": 3.0214744761134863e-06,
+      "loss": 0.8804,
+      "step": 13138
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7103842496871948,
+      "learning_rate": 3.0201440981350892e-06,
+      "loss": 0.9105,
+      "step": 13139
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.468301773071289,
+      "learning_rate": 3.018813961012025e-06,
+      "loss": 0.8877,
+      "step": 13140
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0021722316741943,
+      "learning_rate": 3.017484064790186e-06,
+      "loss": 0.4864,
+      "step": 13141
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6509027481079102,
+      "learning_rate": 3.016154409515467e-06,
+      "loss": 0.8585,
+      "step": 13142
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9704585075378418,
+      "learning_rate": 3.0148249952337536e-06,
+      "loss": 0.9119,
+      "step": 13143
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6537797451019287,
+      "learning_rate": 3.013495821990915e-06,
+      "loss": 0.9343,
+      "step": 13144
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8629426956176758,
+      "learning_rate": 3.0121668898328225e-06,
+      "loss": 0.9428,
+      "step": 13145
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9515079259872437,
+      "learning_rate": 3.0108381988053283e-06,
+      "loss": 0.9222,
+      "step": 13146
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6965047121047974,
+      "learning_rate": 3.0095097489542867e-06,
+      "loss": 0.88,
+      "step": 13147
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.667617917060852,
+      "learning_rate": 3.008181540325533e-06,
+      "loss": 0.8564,
+      "step": 13148
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7944458723068237,
+      "learning_rate": 3.0068535729649074e-06,
+      "loss": 0.9791,
+      "step": 13149
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6819769144058228,
+      "learning_rate": 3.0055258469182267e-06,
+      "loss": 0.8316,
+      "step": 13150
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7586705684661865,
+      "learning_rate": 3.004198362231315e-06,
+      "loss": 0.9353,
+      "step": 13151
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6299564838409424,
+      "learning_rate": 3.0028711189499717e-06,
+      "loss": 0.8682,
+      "step": 13152
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.744849443435669,
+      "learning_rate": 3.0015441171200045e-06,
+      "loss": 1.0035,
+      "step": 13153
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9347338676452637,
+      "learning_rate": 3.0002173567871964e-06,
+      "loss": 0.9465,
+      "step": 13154
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6929874420166016,
+      "learning_rate": 2.9988908379973346e-06,
+      "loss": 0.9204,
+      "step": 13155
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6636193990707397,
+      "learning_rate": 2.997564560796196e-06,
+      "loss": 0.8773,
+      "step": 13156
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7956461906433105,
+      "learning_rate": 2.9962385252295414e-06,
+      "loss": 0.8391,
+      "step": 13157
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6897242069244385,
+      "learning_rate": 2.9949127313431335e-06,
+      "loss": 0.9538,
+      "step": 13158
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8598498106002808,
+      "learning_rate": 2.9935871791827166e-06,
+      "loss": 0.942,
+      "step": 13159
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9431281089782715,
+      "learning_rate": 2.9922618687940374e-06,
+      "loss": 0.9107,
+      "step": 13160
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6600600481033325,
+      "learning_rate": 2.9909368002228223e-06,
+      "loss": 0.8868,
+      "step": 13161
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6209518909454346,
+      "learning_rate": 2.989611973514803e-06,
+      "loss": 0.9199,
+      "step": 13162
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7206671237945557,
+      "learning_rate": 2.9882873887156885e-06,
+      "loss": 0.9183,
+      "step": 13163
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.82438325881958,
+      "learning_rate": 2.986963045871193e-06,
+      "loss": 0.9486,
+      "step": 13164
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9009443521499634,
+      "learning_rate": 2.9856389450270085e-06,
+      "loss": 0.9075,
+      "step": 13165
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6566468477249146,
+      "learning_rate": 2.984315086228834e-06,
+      "loss": 0.9397,
+      "step": 13166
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7568695545196533,
+      "learning_rate": 2.982991469522346e-06,
+      "loss": 0.8926,
+      "step": 13167
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7237212657928467,
+      "learning_rate": 2.9816680949532207e-06,
+      "loss": 0.8874,
+      "step": 13168
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.806574821472168,
+      "learning_rate": 2.9803449625671266e-06,
+      "loss": 0.883,
+      "step": 13169
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.788259744644165,
+      "learning_rate": 2.9790220724097173e-06,
+      "loss": 0.8566,
+      "step": 13170
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6851210594177246,
+      "learning_rate": 2.9776994245266465e-06,
+      "loss": 0.888,
+      "step": 13171
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7277971506118774,
+      "learning_rate": 2.97637701896355e-06,
+      "loss": 0.8776,
+      "step": 13172
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7069425582885742,
+      "learning_rate": 2.9750548557660663e-06,
+      "loss": 0.8802,
+      "step": 13173
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6721842288970947,
+      "learning_rate": 2.9737329349798115e-06,
+      "loss": 0.8595,
+      "step": 13174
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.721893548965454,
+      "learning_rate": 2.9724112566504072e-06,
+      "loss": 0.9999,
+      "step": 13175
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.011451244354248,
+      "learning_rate": 2.9710898208234593e-06,
+      "loss": 0.8833,
+      "step": 13176
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6708012819290161,
+      "learning_rate": 2.9697686275445703e-06,
+      "loss": 0.8679,
+      "step": 13177
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.729573369026184,
+      "learning_rate": 2.968447676859325e-06,
+      "loss": 0.9192,
+      "step": 13178
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8127408027648926,
+      "learning_rate": 2.967126968813312e-06,
+      "loss": 0.9578,
+      "step": 13179
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6849807500839233,
+      "learning_rate": 2.965806503452098e-06,
+      "loss": 1.0057,
+      "step": 13180
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6070747375488281,
+      "learning_rate": 2.964486280821256e-06,
+      "loss": 0.9915,
+      "step": 13181
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9959151744842529,
+      "learning_rate": 2.963166300966336e-06,
+      "loss": 0.5262,
+      "step": 13182
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8654508590698242,
+      "learning_rate": 2.961846563932893e-06,
+      "loss": 0.9353,
+      "step": 13183
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.695183277130127,
+      "learning_rate": 2.9605270697664624e-06,
+      "loss": 0.8821,
+      "step": 13184
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6865360736846924,
+      "learning_rate": 2.9592078185125783e-06,
+      "loss": 0.9424,
+      "step": 13185
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8784617185592651,
+      "learning_rate": 2.957888810216768e-06,
+      "loss": 0.9317,
+      "step": 13186
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8124345541000366,
+      "learning_rate": 2.9565700449245407e-06,
+      "loss": 0.8859,
+      "step": 13187
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7845858335494995,
+      "learning_rate": 2.9552515226814084e-06,
+      "loss": 0.9281,
+      "step": 13188
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.875267744064331,
+      "learning_rate": 2.953933243532865e-06,
+      "loss": 0.9778,
+      "step": 13189
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.817422866821289,
+      "learning_rate": 2.9526152075244054e-06,
+      "loss": 0.9433,
+      "step": 13190
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8604117631912231,
+      "learning_rate": 2.951297414701506e-06,
+      "loss": 0.9248,
+      "step": 13191
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7330478429794312,
+      "learning_rate": 2.9499798651096466e-06,
+      "loss": 0.9422,
+      "step": 13192
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9802653789520264,
+      "learning_rate": 2.9486625587942854e-06,
+      "loss": 0.9167,
+      "step": 13193
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7999430894851685,
+      "learning_rate": 2.947345495800885e-06,
+      "loss": 0.8833,
+      "step": 13194
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.666460633277893,
+      "learning_rate": 2.946028676174888e-06,
+      "loss": 0.8997,
+      "step": 13195
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7075555324554443,
+      "learning_rate": 2.9447120999617363e-06,
+      "loss": 0.8959,
+      "step": 13196
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7928537130355835,
+      "learning_rate": 2.943395767206866e-06,
+      "loss": 0.8873,
+      "step": 13197
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7476816177368164,
+      "learning_rate": 2.9420796779556916e-06,
+      "loss": 0.9267,
+      "step": 13198
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8047982454299927,
+      "learning_rate": 2.940763832253636e-06,
+      "loss": 0.8948,
+      "step": 13199
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8126059770584106,
+      "learning_rate": 2.939448230146098e-06,
+      "loss": 0.9162,
+      "step": 13200
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.937567114830017,
+      "learning_rate": 2.9381328716784816e-06,
+      "loss": 0.8727,
+      "step": 13201
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6910878419876099,
+      "learning_rate": 2.936817756896171e-06,
+      "loss": 0.9754,
+      "step": 13202
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8811239004135132,
+      "learning_rate": 2.935502885844551e-06,
+      "loss": 0.918,
+      "step": 13203
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7432224750518799,
+      "learning_rate": 2.9341882585689908e-06,
+      "loss": 0.9228,
+      "step": 13204
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.743201494216919,
+      "learning_rate": 2.932873875114859e-06,
+      "loss": 0.8639,
+      "step": 13205
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8480312824249268,
+      "learning_rate": 2.9315597355275048e-06,
+      "loss": 0.9222,
+      "step": 13206
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7654367685317993,
+      "learning_rate": 2.9302458398522836e-06,
+      "loss": 0.8815,
+      "step": 13207
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.85563063621521,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.9126,
+      "step": 13208
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.093630075454712,
+      "learning_rate": 2.9276187804195664e-06,
+      "loss": 0.5255,
+      "step": 13209
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7472140789031982,
+      "learning_rate": 2.9263056167527293e-06,
+      "loss": 0.8998,
+      "step": 13210
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8472427129745483,
+      "learning_rate": 2.924992697179324e-06,
+      "loss": 0.9155,
+      "step": 13211
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.938778281211853,
+      "learning_rate": 2.923680021744659e-06,
+      "loss": 0.8698,
+      "step": 13212
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8026741743087769,
+      "learning_rate": 2.9223675904940274e-06,
+      "loss": 0.8661,
+      "step": 13213
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9690545797348022,
+      "learning_rate": 2.9210554034727236e-06,
+      "loss": 1.0288,
+      "step": 13214
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7886452674865723,
+      "learning_rate": 2.919743460726019e-06,
+      "loss": 0.9428,
+      "step": 13215
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6505091190338135,
+      "learning_rate": 2.9184317622991933e-06,
+      "loss": 0.8912,
+      "step": 13216
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8436967134475708,
+      "learning_rate": 2.9171203082375033e-06,
+      "loss": 0.8631,
+      "step": 13217
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7760852575302124,
+      "learning_rate": 2.9158090985862085e-06,
+      "loss": 0.9646,
+      "step": 13218
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.789567470550537,
+      "learning_rate": 2.914498133390551e-06,
+      "loss": 0.8927,
+      "step": 13219
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7957115173339844,
+      "learning_rate": 2.9131874126957728e-06,
+      "loss": 0.9065,
+      "step": 13220
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6785744428634644,
+      "learning_rate": 2.9118769365470967e-06,
+      "loss": 0.9948,
+      "step": 13221
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6264938116073608,
+      "learning_rate": 2.910566704989749e-06,
+      "loss": 0.9087,
+      "step": 13222
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6964391469955444,
+      "learning_rate": 2.9092567180689436e-06,
+      "loss": 0.8738,
+      "step": 13223
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7114852666854858,
+      "learning_rate": 2.907946975829877e-06,
+      "loss": 0.9582,
+      "step": 13224
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8612910509109497,
+      "learning_rate": 2.9066374783177543e-06,
+      "loss": 0.9365,
+      "step": 13225
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7882800102233887,
+      "learning_rate": 2.905328225577755e-06,
+      "loss": 0.9113,
+      "step": 13226
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8253496885299683,
+      "learning_rate": 2.904019217655062e-06,
+      "loss": 0.9329,
+      "step": 13227
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.815693736076355,
+      "learning_rate": 2.9027104545948414e-06,
+      "loss": 0.8785,
+      "step": 13228
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9561164379119873,
+      "learning_rate": 2.9014019364422606e-06,
+      "loss": 0.9548,
+      "step": 13229
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7446643114089966,
+      "learning_rate": 2.9000936632424682e-06,
+      "loss": 0.9702,
+      "step": 13230
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8888019323349,
+      "learning_rate": 2.898785635040612e-06,
+      "loss": 0.8617,
+      "step": 13231
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8743476867675781,
+      "learning_rate": 2.897477851881825e-06,
+      "loss": 0.8172,
+      "step": 13232
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8921974897384644,
+      "learning_rate": 2.896170313811236e-06,
+      "loss": 0.8582,
+      "step": 13233
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8024595975875854,
+      "learning_rate": 2.8948630208739704e-06,
+      "loss": 0.8963,
+      "step": 13234
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9709048271179199,
+      "learning_rate": 2.8935559731151295e-06,
+      "loss": 0.497,
+      "step": 13235
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7381986379623413,
+      "learning_rate": 2.892249170579826e-06,
+      "loss": 0.919,
+      "step": 13236
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6300166845321655,
+      "learning_rate": 2.8909426133131447e-06,
+      "loss": 0.8549,
+      "step": 13237
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6343140602111816,
+      "learning_rate": 2.8896363013601793e-06,
+      "loss": 0.8764,
+      "step": 13238
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8076062202453613,
+      "learning_rate": 2.888330234765999e-06,
+      "loss": 1.0231,
+      "step": 13239
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.988242506980896,
+      "learning_rate": 2.887024413575681e-06,
+      "loss": 0.9234,
+      "step": 13240
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8430019617080688,
+      "learning_rate": 2.8857188378342773e-06,
+      "loss": 0.9245,
+      "step": 13241
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6071193218231201,
+      "learning_rate": 2.884413507586844e-06,
+      "loss": 0.8541,
+      "step": 13242
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1695038080215454,
+      "learning_rate": 2.8831084228784234e-06,
+      "loss": 0.6118,
+      "step": 13243
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8383134603500366,
+      "learning_rate": 2.8818035837540538e-06,
+      "loss": 0.9274,
+      "step": 13244
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1496589183807373,
+      "learning_rate": 2.8804989902587564e-06,
+      "loss": 0.9226,
+      "step": 13245
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0077054500579834,
+      "learning_rate": 2.8791946424375537e-06,
+      "loss": 0.8503,
+      "step": 13246
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.76277494430542,
+      "learning_rate": 2.877890540335451e-06,
+      "loss": 0.9148,
+      "step": 13247
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.645403504371643,
+      "learning_rate": 2.8765866839974522e-06,
+      "loss": 0.8756,
+      "step": 13248
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.716269850730896,
+      "learning_rate": 2.8752830734685466e-06,
+      "loss": 0.9304,
+      "step": 13249
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7370630502700806,
+      "learning_rate": 2.8739797087937194e-06,
+      "loss": 0.9447,
+      "step": 13250
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.80650794506073,
+      "learning_rate": 2.87267659001795e-06,
+      "loss": 0.9511,
+      "step": 13251
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7192002534866333,
+      "learning_rate": 2.871373717186199e-06,
+      "loss": 0.8388,
+      "step": 13252
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.081408739089966,
+      "learning_rate": 2.8700710903434314e-06,
+      "loss": 0.946,
+      "step": 13253
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.868427038192749,
+      "learning_rate": 2.8687687095345894e-06,
+      "loss": 0.9108,
+      "step": 13254
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7838658094406128,
+      "learning_rate": 2.867466574804624e-06,
+      "loss": 0.9335,
+      "step": 13255
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7105185985565186,
+      "learning_rate": 2.866164686198459e-06,
+      "loss": 0.9367,
+      "step": 13256
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7979981899261475,
+      "learning_rate": 2.864863043761026e-06,
+      "loss": 0.9041,
+      "step": 13257
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.832834005355835,
+      "learning_rate": 2.8635616475372365e-06,
+      "loss": 0.9572,
+      "step": 13258
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6792157888412476,
+      "learning_rate": 2.8622604975720016e-06,
+      "loss": 0.9173,
+      "step": 13259
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8305110931396484,
+      "learning_rate": 2.8609595939102153e-06,
+      "loss": 1.0185,
+      "step": 13260
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.931822657585144,
+      "learning_rate": 2.859658936596774e-06,
+      "loss": 0.9481,
+      "step": 13261
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1546175479888916,
+      "learning_rate": 2.8583585256765547e-06,
+      "loss": 0.87,
+      "step": 13262
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8563448190689087,
+      "learning_rate": 2.8570583611944336e-06,
+      "loss": 0.7861,
+      "step": 13263
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6987948417663574,
+      "learning_rate": 2.855758443195278e-06,
+      "loss": 0.8936,
+      "step": 13264
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9658501148223877,
+      "learning_rate": 2.854458771723939e-06,
+      "loss": 0.5109,
+      "step": 13265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6588844060897827,
+      "learning_rate": 2.8531593468252703e-06,
+      "loss": 0.968,
+      "step": 13266
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9360096454620361,
+      "learning_rate": 2.851860168544106e-06,
+      "loss": 0.9487,
+      "step": 13267
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7261602878570557,
+      "learning_rate": 2.8505612369252834e-06,
+      "loss": 0.9179,
+      "step": 13268
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0890910625457764,
+      "learning_rate": 2.8492625520136174e-06,
+      "loss": 0.9482,
+      "step": 13269
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7408963441848755,
+      "learning_rate": 2.847964113853928e-06,
+      "loss": 0.9575,
+      "step": 13270
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9424909353256226,
+      "learning_rate": 2.8466659224910174e-06,
+      "loss": 0.9589,
+      "step": 13271
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7734640836715698,
+      "learning_rate": 2.8453679779696864e-06,
+      "loss": 0.8728,
+      "step": 13272
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.985802173614502,
+      "learning_rate": 2.8440702803347175e-06,
+      "loss": 0.9759,
+      "step": 13273
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9950181841850281,
+      "learning_rate": 2.8427728296308965e-06,
+      "loss": 0.4843,
+      "step": 13274
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7475999593734741,
+      "learning_rate": 2.8414756259029907e-06,
+      "loss": 0.8522,
+      "step": 13275
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8192265033721924,
+      "learning_rate": 2.8401786691957632e-06,
+      "loss": 0.895,
+      "step": 13276
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.819940447807312,
+      "learning_rate": 2.838881959553973e-06,
+      "loss": 0.8825,
+      "step": 13277
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5944887399673462,
+      "learning_rate": 2.8375854970223595e-06,
+      "loss": 0.8606,
+      "step": 13278
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6561224460601807,
+      "learning_rate": 2.8362892816456668e-06,
+      "loss": 0.8669,
+      "step": 13279
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8115332126617432,
+      "learning_rate": 2.8349933134686156e-06,
+      "loss": 0.9033,
+      "step": 13280
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.861738920211792,
+      "learning_rate": 2.8336975925359345e-06,
+      "loss": 0.9444,
+      "step": 13281
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7704592943191528,
+      "learning_rate": 2.8324021188923276e-06,
+      "loss": 0.8754,
+      "step": 13282
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9867587089538574,
+      "learning_rate": 2.8311068925825057e-06,
+      "loss": 0.9557,
+      "step": 13283
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8192524909973145,
+      "learning_rate": 2.829811913651156e-06,
+      "loss": 0.9777,
+      "step": 13284
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7871979475021362,
+      "learning_rate": 2.8285171821429715e-06,
+      "loss": 0.9353,
+      "step": 13285
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7441893815994263,
+      "learning_rate": 2.827222698102622e-06,
+      "loss": 0.8872,
+      "step": 13286
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8543709516525269,
+      "learning_rate": 2.825928461574782e-06,
+      "loss": 0.8975,
+      "step": 13287
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7354766130447388,
+      "learning_rate": 2.824634472604113e-06,
+      "loss": 0.989,
+      "step": 13288
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.805140733718872,
+      "learning_rate": 2.8233407312352623e-06,
+      "loss": 0.8908,
+      "step": 13289
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6988160610198975,
+      "learning_rate": 2.8220472375128793e-06,
+      "loss": 0.9229,
+      "step": 13290
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7211908102035522,
+      "learning_rate": 2.820753991481592e-06,
+      "loss": 0.9281,
+      "step": 13291
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7471613883972168,
+      "learning_rate": 2.819460993186032e-06,
+      "loss": 0.9751,
+      "step": 13292
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.793211817741394,
+      "learning_rate": 2.8181682426708134e-06,
+      "loss": 0.8538,
+      "step": 13293
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8046174049377441,
+      "learning_rate": 2.816875739980549e-06,
+      "loss": 0.8877,
+      "step": 13294
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.811018705368042,
+      "learning_rate": 2.815583485159835e-06,
+      "loss": 0.9231,
+      "step": 13295
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0825761556625366,
+      "learning_rate": 2.8142914782532693e-06,
+      "loss": 0.5365,
+      "step": 13296
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.767999529838562,
+      "learning_rate": 2.8129997193054294e-06,
+      "loss": 0.9104,
+      "step": 13297
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8627296686172485,
+      "learning_rate": 2.811708208360896e-06,
+      "loss": 0.904,
+      "step": 13298
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7052541971206665,
+      "learning_rate": 2.8104169454642293e-06,
+      "loss": 0.8966,
+      "step": 13299
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.656807541847229,
+      "learning_rate": 2.8091259306599905e-06,
+      "loss": 0.9033,
+      "step": 13300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7286090850830078,
+      "learning_rate": 2.8078351639927326e-06,
+      "loss": 0.8502,
+      "step": 13301
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9246201515197754,
+      "learning_rate": 2.806544645506989e-06,
+      "loss": 0.9052,
+      "step": 13302
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8742958307266235,
+      "learning_rate": 2.8052543752472996e-06,
+      "loss": 0.8318,
+      "step": 13303
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6651182174682617,
+      "learning_rate": 2.8039643532581794e-06,
+      "loss": 0.8753,
+      "step": 13304
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.794137716293335,
+      "learning_rate": 2.8026745795841525e-06,
+      "loss": 0.8762,
+      "step": 13305
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.833147406578064,
+      "learning_rate": 2.8013850542697162e-06,
+      "loss": 0.9718,
+      "step": 13306
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8844655752182007,
+      "learning_rate": 2.8000957773593786e-06,
+      "loss": 0.8835,
+      "step": 13307
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6990755796432495,
+      "learning_rate": 2.7988067488976158e-06,
+      "loss": 0.9939,
+      "step": 13308
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6252760887145996,
+      "learning_rate": 2.7975179689289223e-06,
+      "loss": 0.8682,
+      "step": 13309
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8397388458251953,
+      "learning_rate": 2.796229437497762e-06,
+      "loss": 0.9186,
+      "step": 13310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.85454261302948,
+      "learning_rate": 2.7949411546486037e-06,
+      "loss": 0.9231,
+      "step": 13311
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.76252019405365,
+      "learning_rate": 2.7936531204258964e-06,
+      "loss": 0.8301,
+      "step": 13312
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7701213359832764,
+      "learning_rate": 2.7923653348740944e-06,
+      "loss": 0.8594,
+      "step": 13313
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8002903461456299,
+      "learning_rate": 2.7910777980376256e-06,
+      "loss": 0.8304,
+      "step": 13314
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.742992877960205,
+      "learning_rate": 2.789790509960929e-06,
+      "loss": 0.8994,
+      "step": 13315
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7432808876037598,
+      "learning_rate": 2.7885034706884186e-06,
+      "loss": 0.905,
+      "step": 13316
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.689575433731079,
+      "learning_rate": 2.7872166802645073e-06,
+      "loss": 0.8046,
+      "step": 13317
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8014971017837524,
+      "learning_rate": 2.785930138733605e-06,
+      "loss": 0.9158,
+      "step": 13318
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7904638051986694,
+      "learning_rate": 2.784643846140097e-06,
+      "loss": 0.888,
+      "step": 13319
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.850538969039917,
+      "learning_rate": 2.783357802528379e-06,
+      "loss": 0.9077,
+      "step": 13320
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.639488935470581,
+      "learning_rate": 2.782072007942821e-06,
+      "loss": 0.8709,
+      "step": 13321
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6707806587219238,
+      "learning_rate": 2.780786462427798e-06,
+      "loss": 0.8406,
+      "step": 13322
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0482465028762817,
+      "learning_rate": 2.7795011660276662e-06,
+      "loss": 0.6377,
+      "step": 13323
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9531552195549011,
+      "learning_rate": 2.778216118786782e-06,
+      "loss": 0.5236,
+      "step": 13324
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7198026180267334,
+      "learning_rate": 2.776931320749483e-06,
+      "loss": 0.8663,
+      "step": 13325
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7642197608947754,
+      "learning_rate": 2.775646771960111e-06,
+      "loss": 0.9263,
+      "step": 13326
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7483997344970703,
+      "learning_rate": 2.7743624724629847e-06,
+      "loss": 0.9107,
+      "step": 13327
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8054081201553345,
+      "learning_rate": 2.7730784223024255e-06,
+      "loss": 0.8776,
+      "step": 13328
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8185434341430664,
+      "learning_rate": 2.7717946215227453e-06,
+      "loss": 0.9155,
+      "step": 13329
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6876800060272217,
+      "learning_rate": 2.770511070168239e-06,
+      "loss": 1.0118,
+      "step": 13330
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8999838829040527,
+      "learning_rate": 2.769227768283204e-06,
+      "loss": 0.9727,
+      "step": 13331
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9106507301330566,
+      "learning_rate": 2.7679447159119164e-06,
+      "loss": 0.9897,
+      "step": 13332
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8607107400894165,
+      "learning_rate": 2.7666619130986594e-06,
+      "loss": 0.969,
+      "step": 13333
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.718234896659851,
+      "learning_rate": 2.76537935988769e-06,
+      "loss": 0.8793,
+      "step": 13334
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8154743909835815,
+      "learning_rate": 2.764097056323273e-06,
+      "loss": 0.8537,
+      "step": 13335
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8593318462371826,
+      "learning_rate": 2.7628150024496513e-06,
+      "loss": 0.8895,
+      "step": 13336
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0154054164886475,
+      "learning_rate": 2.7615331983110704e-06,
+      "loss": 0.8898,
+      "step": 13337
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.656383991241455,
+      "learning_rate": 2.7602516439517555e-06,
+      "loss": 0.8681,
+      "step": 13338
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6162132024765015,
+      "learning_rate": 2.7589703394159362e-06,
+      "loss": 0.9056,
+      "step": 13339
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7413274049758911,
+      "learning_rate": 2.7576892847478208e-06,
+      "loss": 0.988,
+      "step": 13340
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8440016508102417,
+      "learning_rate": 2.756408479991618e-06,
+      "loss": 0.9361,
+      "step": 13341
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.767145037651062,
+      "learning_rate": 2.7551279251915265e-06,
+      "loss": 0.9143,
+      "step": 13342
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7157105207443237,
+      "learning_rate": 2.7538476203917296e-06,
+      "loss": 0.83,
+      "step": 13343
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.839754581451416,
+      "learning_rate": 2.7525675656364136e-06,
+      "loss": 0.8337,
+      "step": 13344
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8468281030654907,
+      "learning_rate": 2.751287760969743e-06,
+      "loss": 0.8741,
+      "step": 13345
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.713848352432251,
+      "learning_rate": 2.7500082064358855e-06,
+      "loss": 0.8525,
+      "step": 13346
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7702603340148926,
+      "learning_rate": 2.748728902078991e-06,
+      "loss": 0.9346,
+      "step": 13347
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6287257671356201,
+      "learning_rate": 2.7474498479432087e-06,
+      "loss": 0.8488,
+      "step": 13348
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7959941625595093,
+      "learning_rate": 2.7461710440726696e-06,
+      "loss": 0.934,
+      "step": 13349
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8187071084976196,
+      "learning_rate": 2.7448924905115095e-06,
+      "loss": 0.8862,
+      "step": 13350
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6120656728744507,
+      "learning_rate": 2.743614187303838e-06,
+      "loss": 0.9344,
+      "step": 13351
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7234573364257812,
+      "learning_rate": 2.742336134493776e-06,
+      "loss": 0.9042,
+      "step": 13352
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.81466805934906,
+      "learning_rate": 2.741058332125417e-06,
+      "loss": 1.0001,
+      "step": 13353
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6738122701644897,
+      "learning_rate": 2.739780780242857e-06,
+      "loss": 0.8824,
+      "step": 13354
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.862013578414917,
+      "learning_rate": 2.7385034788901853e-06,
+      "loss": 0.8605,
+      "step": 13355
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7717914581298828,
+      "learning_rate": 2.737226428111471e-06,
+      "loss": 0.9429,
+      "step": 13356
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7951974868774414,
+      "learning_rate": 2.735949627950789e-06,
+      "loss": 0.9648,
+      "step": 13357
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8635969161987305,
+      "learning_rate": 2.73467307845219e-06,
+      "loss": 0.8941,
+      "step": 13358
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5822746753692627,
+      "learning_rate": 2.7333967796597317e-06,
+      "loss": 0.9383,
+      "step": 13359
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9676087498664856,
+      "learning_rate": 2.7321207316174493e-06,
+      "loss": 0.5084,
+      "step": 13360
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.760954737663269,
+      "learning_rate": 2.7308449343693812e-06,
+      "loss": 0.9147,
+      "step": 13361
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6818454265594482,
+      "learning_rate": 2.7295693879595453e-06,
+      "loss": 0.8711,
+      "step": 13362
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1502350568771362,
+      "learning_rate": 2.7282940924319647e-06,
+      "loss": 0.6337,
+      "step": 13363
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0412083864212036,
+      "learning_rate": 2.727019047830638e-06,
+      "loss": 0.5221,
+      "step": 13364
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6387964487075806,
+      "learning_rate": 2.7257442541995692e-06,
+      "loss": 0.8635,
+      "step": 13365
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7158979177474976,
+      "learning_rate": 2.724469711582748e-06,
+      "loss": 0.9463,
+      "step": 13366
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6583956480026245,
+      "learning_rate": 2.723195420024152e-06,
+      "loss": 0.9242,
+      "step": 13367
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.728717565536499,
+      "learning_rate": 2.7219213795677567e-06,
+      "loss": 0.9293,
+      "step": 13368
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6908937692642212,
+      "learning_rate": 2.7206475902575225e-06,
+      "loss": 0.8562,
+      "step": 13369
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9748352766036987,
+      "learning_rate": 2.719374052137408e-06,
+      "loss": 0.9284,
+      "step": 13370
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8594862222671509,
+      "learning_rate": 2.718100765251355e-06,
+      "loss": 0.8801,
+      "step": 13371
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7083646059036255,
+      "learning_rate": 2.7168277296433055e-06,
+      "loss": 0.8948,
+      "step": 13372
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.985163688659668,
+      "learning_rate": 2.715554945357184e-06,
+      "loss": 0.8844,
+      "step": 13373
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7561290264129639,
+      "learning_rate": 2.714282412436913e-06,
+      "loss": 0.8,
+      "step": 13374
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.74337637424469,
+      "learning_rate": 2.7130101309264035e-06,
+      "loss": 0.9497,
+      "step": 13375
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.882129430770874,
+      "learning_rate": 2.711738100869563e-06,
+      "loss": 0.8478,
+      "step": 13376
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6719154119491577,
+      "learning_rate": 2.7104663223102776e-06,
+      "loss": 0.8639,
+      "step": 13377
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.651335597038269,
+      "learning_rate": 2.709194795292441e-06,
+      "loss": 0.898,
+      "step": 13378
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8392298221588135,
+      "learning_rate": 2.707923519859922e-06,
+      "loss": 0.9249,
+      "step": 13379
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7724997997283936,
+      "learning_rate": 2.7066524960565965e-06,
+      "loss": 0.8261,
+      "step": 13380
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.262718915939331,
+      "learning_rate": 2.7053817239263168e-06,
+      "loss": 0.9593,
+      "step": 13381
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8630397319793701,
+      "learning_rate": 2.704111203512938e-06,
+      "loss": 0.8823,
+      "step": 13382
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1210665702819824,
+      "learning_rate": 2.7028409348603037e-06,
+      "loss": 0.5562,
+      "step": 13383
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8422825336456299,
+      "learning_rate": 2.7015709180122416e-06,
+      "loss": 0.8691,
+      "step": 13384
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7199374437332153,
+      "learning_rate": 2.7003011530125823e-06,
+      "loss": 0.9395,
+      "step": 13385
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7889477014541626,
+      "learning_rate": 2.6990316399051373e-06,
+      "loss": 0.917,
+      "step": 13386
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.647608995437622,
+      "learning_rate": 2.6977623787337193e-06,
+      "loss": 0.852,
+      "step": 13387
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.870060682296753,
+      "learning_rate": 2.696493369542119e-06,
+      "loss": 0.8578,
+      "step": 13388
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7119452953338623,
+      "learning_rate": 2.6952246123741353e-06,
+      "loss": 0.9926,
+      "step": 13389
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.614616870880127,
+      "learning_rate": 2.693956107273542e-06,
+      "loss": 0.8661,
+      "step": 13390
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.674981951713562,
+      "learning_rate": 2.6926878542841184e-06,
+      "loss": 0.8625,
+      "step": 13391
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8029931783676147,
+      "learning_rate": 2.6914198534496204e-06,
+      "loss": 0.9375,
+      "step": 13392
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7679470777511597,
+      "learning_rate": 2.6901521048138115e-06,
+      "loss": 0.8508,
+      "step": 13393
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.039330005645752,
+      "learning_rate": 2.688884608420431e-06,
+      "loss": 1.0253,
+      "step": 13394
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7551875114440918,
+      "learning_rate": 2.68761736431322e-06,
+      "loss": 0.9618,
+      "step": 13395
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.813559889793396,
+      "learning_rate": 2.6863503725359107e-06,
+      "loss": 0.9196,
+      "step": 13396
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8430988788604736,
+      "learning_rate": 2.685083633132216e-06,
+      "loss": 0.9033,
+      "step": 13397
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.765631914138794,
+      "learning_rate": 2.6838171461458563e-06,
+      "loss": 0.9833,
+      "step": 13398
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.743289589881897,
+      "learning_rate": 2.682550911620526e-06,
+      "loss": 0.8334,
+      "step": 13399
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6356432437896729,
+      "learning_rate": 2.6812849295999267e-06,
+      "loss": 0.991,
+      "step": 13400
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7690151929855347,
+      "learning_rate": 2.680019200127737e-06,
+      "loss": 0.9193,
+      "step": 13401
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7529290914535522,
+      "learning_rate": 2.6787537232476403e-06,
+      "loss": 0.8587,
+      "step": 13402
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8339829444885254,
+      "learning_rate": 2.677488499003299e-06,
+      "loss": 0.9759,
+      "step": 13403
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.732663631439209,
+      "learning_rate": 2.6762235274383775e-06,
+      "loss": 0.8637,
+      "step": 13404
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7830681800842285,
+      "learning_rate": 2.6749588085965216e-06,
+      "loss": 0.9172,
+      "step": 13405
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7096357345581055,
+      "learning_rate": 2.673694342521378e-06,
+      "loss": 0.9225,
+      "step": 13406
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.708238959312439,
+      "learning_rate": 2.6724301292565747e-06,
+      "loss": 0.9508,
+      "step": 13407
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6304820775985718,
+      "learning_rate": 2.671166168845738e-06,
+      "loss": 0.9219,
+      "step": 13408
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7312514781951904,
+      "learning_rate": 2.6699024613324888e-06,
+      "loss": 0.9579,
+      "step": 13409
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1243536472320557,
+      "learning_rate": 2.6686390067604264e-06,
+      "loss": 0.5645,
+      "step": 13410
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.811184287071228,
+      "learning_rate": 2.6673758051731546e-06,
+      "loss": 0.8466,
+      "step": 13411
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8927572965621948,
+      "learning_rate": 2.6661128566142592e-06,
+      "loss": 0.9854,
+      "step": 13412
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5789134502410889,
+      "learning_rate": 2.6648501611273248e-06,
+      "loss": 0.8828,
+      "step": 13413
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8304094076156616,
+      "learning_rate": 2.663587718755919e-06,
+      "loss": 0.9944,
+      "step": 13414
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7803597450256348,
+      "learning_rate": 2.66232552954361e-06,
+      "loss": 0.9091,
+      "step": 13415
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.892556071281433,
+      "learning_rate": 2.6610635935339477e-06,
+      "loss": 0.9384,
+      "step": 13416
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6220402717590332,
+      "learning_rate": 2.659801910770483e-06,
+      "loss": 0.8797,
+      "step": 13417
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7127230167388916,
+      "learning_rate": 2.6585404812967476e-06,
+      "loss": 0.8351,
+      "step": 13418
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6150742769241333,
+      "learning_rate": 2.6572793051562727e-06,
+      "loss": 0.9387,
+      "step": 13419
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7272599935531616,
+      "learning_rate": 2.65601838239258e-06,
+      "loss": 0.8457,
+      "step": 13420
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8239541053771973,
+      "learning_rate": 2.6547577130491764e-06,
+      "loss": 0.9089,
+      "step": 13421
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6838151216506958,
+      "learning_rate": 2.6534972971695683e-06,
+      "loss": 0.9267,
+      "step": 13422
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9428657293319702,
+      "learning_rate": 2.6522371347972444e-06,
+      "loss": 0.9237,
+      "step": 13423
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8919479846954346,
+      "learning_rate": 2.650977225975695e-06,
+      "loss": 0.8336,
+      "step": 13424
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.679810881614685,
+      "learning_rate": 2.649717570748389e-06,
+      "loss": 0.8542,
+      "step": 13425
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.66526460647583,
+      "learning_rate": 2.648458169158801e-06,
+      "loss": 0.942,
+      "step": 13426
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7164236307144165,
+      "learning_rate": 2.647199021250383e-06,
+      "loss": 0.7926,
+      "step": 13427
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6227229833602905,
+      "learning_rate": 2.64594012706659e-06,
+      "loss": 0.8714,
+      "step": 13428
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7704640626907349,
+      "learning_rate": 2.6446814866508587e-06,
+      "loss": 0.9682,
+      "step": 13429
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7707606554031372,
+      "learning_rate": 2.643423100046625e-06,
+      "loss": 0.9444,
+      "step": 13430
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.769217848777771,
+      "learning_rate": 2.6421649672973072e-06,
+      "loss": 0.8163,
+      "step": 13431
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6367143392562866,
+      "learning_rate": 2.6409070884463227e-06,
+      "loss": 0.9359,
+      "step": 13432
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7883613109588623,
+      "learning_rate": 2.6396494635370816e-06,
+      "loss": 0.8401,
+      "step": 13433
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8270783424377441,
+      "learning_rate": 2.6383920926129746e-06,
+      "loss": 0.8557,
+      "step": 13434
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7381024360656738,
+      "learning_rate": 2.6371349757173946e-06,
+      "loss": 0.879,
+      "step": 13435
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7255743741989136,
+      "learning_rate": 2.635878112893717e-06,
+      "loss": 0.8718,
+      "step": 13436
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6766544580459595,
+      "learning_rate": 2.6346215041853183e-06,
+      "loss": 0.8918,
+      "step": 13437
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7601232528686523,
+      "learning_rate": 2.6333651496355527e-06,
+      "loss": 0.8911,
+      "step": 13438
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8920098543167114,
+      "learning_rate": 2.6321090492877823e-06,
+      "loss": 1.0118,
+      "step": 13439
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7545466423034668,
+      "learning_rate": 2.630853203185341e-06,
+      "loss": 0.9712,
+      "step": 13440
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7252217531204224,
+      "learning_rate": 2.629597611371576e-06,
+      "loss": 0.9643,
+      "step": 13441
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9029046297073364,
+      "learning_rate": 2.6283422738898067e-06,
+      "loss": 0.9217,
+      "step": 13442
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.743169903755188,
+      "learning_rate": 2.627087190783356e-06,
+      "loss": 0.8654,
+      "step": 13443
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7741267681121826,
+      "learning_rate": 2.6258323620955286e-06,
+      "loss": 0.8985,
+      "step": 13444
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8389676809310913,
+      "learning_rate": 2.62457778786963e-06,
+      "loss": 0.8943,
+      "step": 13445
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6463677883148193,
+      "learning_rate": 2.6233234681489473e-06,
+      "loss": 0.9001,
+      "step": 13446
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.912057638168335,
+      "learning_rate": 2.622069402976768e-06,
+      "loss": 0.9304,
+      "step": 13447
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8358478546142578,
+      "learning_rate": 2.620815592396362e-06,
+      "loss": 0.9083,
+      "step": 13448
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7130416631698608,
+      "learning_rate": 2.6195620364509966e-06,
+      "loss": 0.8868,
+      "step": 13449
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6826281547546387,
+      "learning_rate": 2.618308735183931e-06,
+      "loss": 0.9089,
+      "step": 13450
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8808422088623047,
+      "learning_rate": 2.6170556886384092e-06,
+      "loss": 0.9799,
+      "step": 13451
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7785825729370117,
+      "learning_rate": 2.615802896857674e-06,
+      "loss": 0.9884,
+      "step": 13452
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8639971017837524,
+      "learning_rate": 2.614550359884952e-06,
+      "loss": 0.9035,
+      "step": 13453
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1102173328399658,
+      "learning_rate": 2.61329807776347e-06,
+      "loss": 0.5912,
+      "step": 13454
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6840548515319824,
+      "learning_rate": 2.6120460505364333e-06,
+      "loss": 0.9342,
+      "step": 13455
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7257391214370728,
+      "learning_rate": 2.610794278247053e-06,
+      "loss": 0.8567,
+      "step": 13456
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7970428466796875,
+      "learning_rate": 2.609542760938519e-06,
+      "loss": 0.819,
+      "step": 13457
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7008837461471558,
+      "learning_rate": 2.608291498654023e-06,
+      "loss": 0.8991,
+      "step": 13458
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8798167705535889,
+      "learning_rate": 2.6070404914367355e-06,
+      "loss": 1.0265,
+      "step": 13459
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9884977340698242,
+      "learning_rate": 2.6057897393298328e-06,
+      "loss": 0.5522,
+      "step": 13460
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8128204345703125,
+      "learning_rate": 2.604539242376468e-06,
+      "loss": 1.0019,
+      "step": 13461
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8131340742111206,
+      "learning_rate": 2.6032890006197965e-06,
+      "loss": 0.8568,
+      "step": 13462
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.576253652572632,
+      "learning_rate": 2.6020390141029616e-06,
+      "loss": 0.9143,
+      "step": 13463
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.562553882598877,
+      "learning_rate": 2.6007892828690927e-06,
+      "loss": 0.9111,
+      "step": 13464
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8374035358428955,
+      "learning_rate": 2.5995398069613197e-06,
+      "loss": 0.9413,
+      "step": 13465
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.671708345413208,
+      "learning_rate": 2.5982905864227526e-06,
+      "loss": 0.8831,
+      "step": 13466
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1745402812957764,
+      "learning_rate": 2.5970416212965043e-06,
+      "loss": 0.8832,
+      "step": 13467
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7788275480270386,
+      "learning_rate": 2.5957929116256677e-06,
+      "loss": 0.9464,
+      "step": 13468
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9412552118301392,
+      "learning_rate": 2.5945444574533372e-06,
+      "loss": 0.8723,
+      "step": 13469
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8016103506088257,
+      "learning_rate": 2.5932962588225884e-06,
+      "loss": 0.9362,
+      "step": 13470
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8947603702545166,
+      "learning_rate": 2.5920483157764988e-06,
+      "loss": 0.8644,
+      "step": 13471
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8442952632904053,
+      "learning_rate": 2.5908006283581255e-06,
+      "loss": 0.884,
+      "step": 13472
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0008127689361572,
+      "learning_rate": 2.589553196610527e-06,
+      "loss": 0.9186,
+      "step": 13473
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8190070390701294,
+      "learning_rate": 2.5883060205767495e-06,
+      "loss": 0.8356,
+      "step": 13474
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8109745979309082,
+      "learning_rate": 2.5870591002998235e-06,
+      "loss": 0.8993,
+      "step": 13475
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9837809801101685,
+      "learning_rate": 2.5858124358227856e-06,
+      "loss": 0.9195,
+      "step": 13476
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7921805381774902,
+      "learning_rate": 2.584566027188645e-06,
+      "loss": 0.9471,
+      "step": 13477
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0525033473968506,
+      "learning_rate": 2.583319874440421e-06,
+      "loss": 0.5807,
+      "step": 13478
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6983859539031982,
+      "learning_rate": 2.582073977621107e-06,
+      "loss": 0.8853,
+      "step": 13479
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6854795217514038,
+      "learning_rate": 2.580828336773702e-06,
+      "loss": 0.9147,
+      "step": 13480
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8637653589248657,
+      "learning_rate": 2.579582951941184e-06,
+      "loss": 0.8891,
+      "step": 13481
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9753445386886597,
+      "learning_rate": 2.5783378231665322e-06,
+      "loss": 0.9432,
+      "step": 13482
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.770796537399292,
+      "learning_rate": 2.577092950492708e-06,
+      "loss": 0.8924,
+      "step": 13483
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9099305868148804,
+      "learning_rate": 2.575848333962674e-06,
+      "loss": 0.8746,
+      "step": 13484
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8360024690628052,
+      "learning_rate": 2.5746039736193727e-06,
+      "loss": 0.9524,
+      "step": 13485
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7529373168945312,
+      "learning_rate": 2.573359869505746e-06,
+      "loss": 0.8807,
+      "step": 13486
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.687541127204895,
+      "learning_rate": 2.572116021664728e-06,
+      "loss": 0.8415,
+      "step": 13487
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.944042682647705,
+      "learning_rate": 2.570872430139234e-06,
+      "loss": 0.8743,
+      "step": 13488
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9257749319076538,
+      "learning_rate": 2.5696290949721823e-06,
+      "loss": 0.913,
+      "step": 13489
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6089198589324951,
+      "learning_rate": 2.5683860162064723e-06,
+      "loss": 0.8882,
+      "step": 13490
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8171964883804321,
+      "learning_rate": 2.5671431938850044e-06,
+      "loss": 0.9666,
+      "step": 13491
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7848421335220337,
+      "learning_rate": 2.5659006280506594e-06,
+      "loss": 0.8899,
+      "step": 13492
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.68595290184021,
+      "learning_rate": 2.5646583187463203e-06,
+      "loss": 0.8754,
+      "step": 13493
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6042120456695557,
+      "learning_rate": 2.56341626601485e-06,
+      "loss": 0.8252,
+      "step": 13494
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6256245374679565,
+      "learning_rate": 2.5621744698991134e-06,
+      "loss": 0.9226,
+      "step": 13495
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9064655303955078,
+      "learning_rate": 2.560932930441956e-06,
+      "loss": 0.9921,
+      "step": 13496
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5552937984466553,
+      "learning_rate": 2.5596916476862234e-06,
+      "loss": 0.8624,
+      "step": 13497
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7282001972198486,
+      "learning_rate": 2.5584506216747516e-06,
+      "loss": 0.8774,
+      "step": 13498
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.682398796081543,
+      "learning_rate": 2.5572098524503585e-06,
+      "loss": 0.9536,
+      "step": 13499
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.698731780052185,
+      "learning_rate": 2.555969340055866e-06,
+      "loss": 0.9224,
+      "step": 13500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7543385028839111,
+      "learning_rate": 2.5547290845340745e-06,
+      "loss": 0.8983,
+      "step": 13501
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7062714099884033,
+      "learning_rate": 2.5534890859277873e-06,
+      "loss": 0.8766,
+      "step": 13502
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8012104034423828,
+      "learning_rate": 2.552249344279788e-06,
+      "loss": 0.8833,
+      "step": 13503
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7164487838745117,
+      "learning_rate": 2.5510098596328625e-06,
+      "loss": 0.8712,
+      "step": 13504
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0002520084381104,
+      "learning_rate": 2.5497706320297757e-06,
+      "loss": 0.954,
+      "step": 13505
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6533023118972778,
+      "learning_rate": 2.548531661513293e-06,
+      "loss": 0.8439,
+      "step": 13506
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8816951513290405,
+      "learning_rate": 2.5472929481261677e-06,
+      "loss": 0.9019,
+      "step": 13507
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8175066709518433,
+      "learning_rate": 2.5460544919111473e-06,
+      "loss": 0.8599,
+      "step": 13508
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6753649711608887,
+      "learning_rate": 2.544816292910962e-06,
+      "loss": 0.9854,
+      "step": 13509
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8827247619628906,
+      "learning_rate": 2.5435783511683444e-06,
+      "loss": 0.9681,
+      "step": 13510
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7025583982467651,
+      "learning_rate": 2.5423406667260065e-06,
+      "loss": 0.9546,
+      "step": 13511
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6776599884033203,
+      "learning_rate": 2.541103239626662e-06,
+      "loss": 0.8516,
+      "step": 13512
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.160560131072998,
+      "learning_rate": 2.539866069913007e-06,
+      "loss": 0.5227,
+      "step": 13513
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8358908891677856,
+      "learning_rate": 2.5386291576277343e-06,
+      "loss": 0.9137,
+      "step": 13514
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8254345655441284,
+      "learning_rate": 2.5373925028135304e-06,
+      "loss": 0.9122,
+      "step": 13515
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8013652563095093,
+      "learning_rate": 2.5361561055130625e-06,
+      "loss": 0.8143,
+      "step": 13516
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8198809623718262,
+      "learning_rate": 2.5349199657690004e-06,
+      "loss": 0.9426,
+      "step": 13517
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.804755449295044,
+      "learning_rate": 2.533684083623994e-06,
+      "loss": 0.9327,
+      "step": 13518
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6862956285476685,
+      "learning_rate": 2.5324484591206978e-06,
+      "loss": 0.8778,
+      "step": 13519
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7597792148590088,
+      "learning_rate": 2.531213092301742e-06,
+      "loss": 0.8879,
+      "step": 13520
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7173988819122314,
+      "learning_rate": 2.5299779832097616e-06,
+      "loss": 0.8587,
+      "step": 13521
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.000418186187744,
+      "learning_rate": 2.528743131887371e-06,
+      "loss": 0.952,
+      "step": 13522
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7533934116363525,
+      "learning_rate": 2.527508538377189e-06,
+      "loss": 0.8514,
+      "step": 13523
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8522275686264038,
+      "learning_rate": 2.5262742027218102e-06,
+      "loss": 0.8705,
+      "step": 13524
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8937947750091553,
+      "learning_rate": 2.5250401249638344e-06,
+      "loss": 0.8968,
+      "step": 13525
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0209144353866577,
+      "learning_rate": 2.5238063051458415e-06,
+      "loss": 0.5407,
+      "step": 13526
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8867063522338867,
+      "learning_rate": 2.5225727433104085e-06,
+      "loss": 0.4912,
+      "step": 13527
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.712629795074463,
+      "learning_rate": 2.5213394395001055e-06,
+      "loss": 0.8222,
+      "step": 13528
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7257368564605713,
+      "learning_rate": 2.5201063937574842e-06,
+      "loss": 0.8699,
+      "step": 13529
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.822894811630249,
+      "learning_rate": 2.5188736061251016e-06,
+      "loss": 0.8894,
+      "step": 13530
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7532531023025513,
+      "learning_rate": 2.5176410766454884e-06,
+      "loss": 0.9672,
+      "step": 13531
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8516680002212524,
+      "learning_rate": 2.5164088053611844e-06,
+      "loss": 0.9776,
+      "step": 13532
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6414748430252075,
+      "learning_rate": 2.515176792314705e-06,
+      "loss": 0.8894,
+      "step": 13533
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7627924680709839,
+      "learning_rate": 2.51394503754857e-06,
+      "loss": 0.8332,
+      "step": 13534
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8156533241271973,
+      "learning_rate": 2.512713541105276e-06,
+      "loss": 0.8449,
+      "step": 13535
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7378655672073364,
+      "learning_rate": 2.5114823030273273e-06,
+      "loss": 0.8479,
+      "step": 13536
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8265857696533203,
+      "learning_rate": 2.510251323357201e-06,
+      "loss": 0.9534,
+      "step": 13537
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8134419918060303,
+      "learning_rate": 2.509020602137384e-06,
+      "loss": 0.8817,
+      "step": 13538
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.701204776763916,
+      "learning_rate": 2.5077901394103386e-06,
+      "loss": 0.8588,
+      "step": 13539
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8620517253875732,
+      "learning_rate": 2.5065599352185255e-06,
+      "loss": 0.9378,
+      "step": 13540
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7137367725372314,
+      "learning_rate": 2.5053299896044e-06,
+      "loss": 0.8855,
+      "step": 13541
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8661954402923584,
+      "learning_rate": 2.5041003026103994e-06,
+      "loss": 1.0024,
+      "step": 13542
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8329602479934692,
+      "learning_rate": 2.50287087427896e-06,
+      "loss": 0.9197,
+      "step": 13543
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0175042152404785,
+      "learning_rate": 2.501641704652502e-06,
+      "loss": 0.844,
+      "step": 13544
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8630995750427246,
+      "learning_rate": 2.5004127937734456e-06,
+      "loss": 0.9515,
+      "step": 13545
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8460482358932495,
+      "learning_rate": 2.4991841416841922e-06,
+      "loss": 0.9665,
+      "step": 13546
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9038846492767334,
+      "learning_rate": 2.497955748427143e-06,
+      "loss": 0.9269,
+      "step": 13547
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6748512983322144,
+      "learning_rate": 2.496727614044683e-06,
+      "loss": 0.9063,
+      "step": 13548
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8474611043930054,
+      "learning_rate": 2.4954997385791967e-06,
+      "loss": 0.8088,
+      "step": 13549
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8525071144104004,
+      "learning_rate": 2.494272122073047e-06,
+      "loss": 0.8748,
+      "step": 13550
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1093207597732544,
+      "learning_rate": 2.4930447645686016e-06,
+      "loss": 0.5935,
+      "step": 13551
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7060428857803345,
+      "learning_rate": 2.491817666108214e-06,
+      "loss": 0.8993,
+      "step": 13552
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7593170404434204,
+      "learning_rate": 2.4905908267342216e-06,
+      "loss": 1.0324,
+      "step": 13553
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8856810331344604,
+      "learning_rate": 2.4893642464889667e-06,
+      "loss": 0.8973,
+      "step": 13554
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8496859073638916,
+      "learning_rate": 2.4881379254147685e-06,
+      "loss": 0.9747,
+      "step": 13555
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7298897504806519,
+      "learning_rate": 2.48691186355395e-06,
+      "loss": 0.9879,
+      "step": 13556
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7991303205490112,
+      "learning_rate": 2.4856860609488133e-06,
+      "loss": 0.8397,
+      "step": 13557
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9909332990646362,
+      "learning_rate": 2.484460517641664e-06,
+      "loss": 0.9396,
+      "step": 13558
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7192751169204712,
+      "learning_rate": 2.4832352336747833e-06,
+      "loss": 0.8254,
+      "step": 13559
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7192391157150269,
+      "learning_rate": 2.482010209090462e-06,
+      "loss": 0.9126,
+      "step": 13560
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6755099296569824,
+      "learning_rate": 2.480785443930964e-06,
+      "loss": 0.9509,
+      "step": 13561
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.794945240020752,
+      "learning_rate": 2.47956093823856e-06,
+      "loss": 0.9382,
+      "step": 13562
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9940567016601562,
+      "learning_rate": 2.4783366920554973e-06,
+      "loss": 0.9632,
+      "step": 13563
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7845189571380615,
+      "learning_rate": 2.477112705424024e-06,
+      "loss": 0.9667,
+      "step": 13564
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5141937732696533,
+      "learning_rate": 2.4758889783863803e-06,
+      "loss": 0.9175,
+      "step": 13565
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.864034652709961,
+      "learning_rate": 2.4746655109847874e-06,
+      "loss": 0.9436,
+      "step": 13566
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6060407161712646,
+      "learning_rate": 2.4734423032614695e-06,
+      "loss": 0.8613,
+      "step": 13567
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8270012140274048,
+      "learning_rate": 2.4722193552586295e-06,
+      "loss": 0.9514,
+      "step": 13568
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6867666244506836,
+      "learning_rate": 2.4709966670184747e-06,
+      "loss": 0.9665,
+      "step": 13569
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9030689001083374,
+      "learning_rate": 2.4697742385831915e-06,
+      "loss": 0.8916,
+      "step": 13570
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.793871521949768,
+      "learning_rate": 2.468552069994966e-06,
+      "loss": 0.9407,
+      "step": 13571
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6520037651062012,
+      "learning_rate": 2.4673301612959653e-06,
+      "loss": 0.8373,
+      "step": 13572
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8157391548156738,
+      "learning_rate": 2.4661085125283647e-06,
+      "loss": 0.8917,
+      "step": 13573
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.650205135345459,
+      "learning_rate": 2.464887123734312e-06,
+      "loss": 0.9705,
+      "step": 13574
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.680430293083191,
+      "learning_rate": 2.4636659949559583e-06,
+      "loss": 0.8629,
+      "step": 13575
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6670992374420166,
+      "learning_rate": 2.4624451262354365e-06,
+      "loss": 0.9356,
+      "step": 13576
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6412323713302612,
+      "learning_rate": 2.461224517614881e-06,
+      "loss": 0.911,
+      "step": 13577
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6828844547271729,
+      "learning_rate": 2.4600041691364053e-06,
+      "loss": 0.8202,
+      "step": 13578
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8257712125778198,
+      "learning_rate": 2.458784080842127e-06,
+      "loss": 0.8662,
+      "step": 13579
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8983169794082642,
+      "learning_rate": 2.457564252774142e-06,
+      "loss": 0.9116,
+      "step": 13580
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7737958431243896,
+      "learning_rate": 2.4563446849745453e-06,
+      "loss": 0.9236,
+      "step": 13581
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9499841928482056,
+      "learning_rate": 2.455125377485423e-06,
+      "loss": 0.8549,
+      "step": 13582
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6789606809616089,
+      "learning_rate": 2.4539063303488474e-06,
+      "loss": 0.8902,
+      "step": 13583
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7524129152297974,
+      "learning_rate": 2.4526875436068865e-06,
+      "loss": 0.9344,
+      "step": 13584
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7539042234420776,
+      "learning_rate": 2.4514690173015944e-06,
+      "loss": 0.9338,
+      "step": 13585
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7185512781143188,
+      "learning_rate": 2.450250751475022e-06,
+      "loss": 0.9036,
+      "step": 13586
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0133153200149536,
+      "learning_rate": 2.4490327461692043e-06,
+      "loss": 0.5656,
+      "step": 13587
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9754283428192139,
+      "learning_rate": 2.447815001426177e-06,
+      "loss": 0.9252,
+      "step": 13588
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.76350736618042,
+      "learning_rate": 2.446597517287954e-06,
+      "loss": 0.9511,
+      "step": 13589
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8235632181167603,
+      "learning_rate": 2.445380293796555e-06,
+      "loss": 0.8567,
+      "step": 13590
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.664875864982605,
+      "learning_rate": 2.4441633309939762e-06,
+      "loss": 0.8897,
+      "step": 13591
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6128277778625488,
+      "learning_rate": 2.442946628922217e-06,
+      "loss": 0.8974,
+      "step": 13592
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7497379779815674,
+      "learning_rate": 2.4417301876232568e-06,
+      "loss": 0.936,
+      "step": 13593
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7339715957641602,
+      "learning_rate": 2.4405140071390755e-06,
+      "loss": 0.8352,
+      "step": 13594
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0066795349121094,
+      "learning_rate": 2.4392980875116414e-06,
+      "loss": 0.8483,
+      "step": 13595
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.13130784034729,
+      "learning_rate": 2.4380824287829073e-06,
+      "loss": 0.922,
+      "step": 13596
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8409433364868164,
+      "learning_rate": 2.4368670309948283e-06,
+      "loss": 0.8528,
+      "step": 13597
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9464130401611328,
+      "learning_rate": 2.435651894189338e-06,
+      "loss": 0.9887,
+      "step": 13598
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7921556234359741,
+      "learning_rate": 2.4344370184083742e-06,
+      "loss": 0.8776,
+      "step": 13599
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.861660361289978,
+      "learning_rate": 2.4332224036938524e-06,
+      "loss": 0.8166,
+      "step": 13600
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6708663702011108,
+      "learning_rate": 2.432008050087692e-06,
+      "loss": 1.0309,
+      "step": 13601
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9906009435653687,
+      "learning_rate": 2.4307939576317897e-06,
+      "loss": 0.8884,
+      "step": 13602
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7988314628601074,
+      "learning_rate": 2.429580126368046e-06,
+      "loss": 0.9101,
+      "step": 13603
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9699030518531799,
+      "learning_rate": 2.428366556338344e-06,
+      "loss": 0.5127,
+      "step": 13604
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9651240110397339,
+      "learning_rate": 2.4271532475845617e-06,
+      "loss": 0.9634,
+      "step": 13605
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.623925805091858,
+      "learning_rate": 2.425940200148569e-06,
+      "loss": 0.8451,
+      "step": 13606
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8109289407730103,
+      "learning_rate": 2.4247274140722197e-06,
+      "loss": 0.9301,
+      "step": 13607
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0365108251571655,
+      "learning_rate": 2.4235148893973693e-06,
+      "loss": 0.4759,
+      "step": 13608
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9081228971481323,
+      "learning_rate": 2.4223026261658546e-06,
+      "loss": 0.9117,
+      "step": 13609
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8595786094665527,
+      "learning_rate": 2.42109062441951e-06,
+      "loss": 0.8851,
+      "step": 13610
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.882069706916809,
+      "learning_rate": 2.419878884200155e-06,
+      "loss": 0.8586,
+      "step": 13611
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7658051252365112,
+      "learning_rate": 2.4186674055496084e-06,
+      "loss": 0.9557,
+      "step": 13612
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9237983226776123,
+      "learning_rate": 2.417456188509669e-06,
+      "loss": 0.9422,
+      "step": 13613
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.093208074569702,
+      "learning_rate": 2.4162452331221387e-06,
+      "loss": 0.9901,
+      "step": 13614
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6953976154327393,
+      "learning_rate": 2.415034539428798e-06,
+      "loss": 0.9097,
+      "step": 13615
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6519988775253296,
+      "learning_rate": 2.413824107471431e-06,
+      "loss": 0.8395,
+      "step": 13616
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7033830881118774,
+      "learning_rate": 2.412613937291799e-06,
+      "loss": 0.8968,
+      "step": 13617
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.73103928565979,
+      "learning_rate": 2.4114040289316665e-06,
+      "loss": 0.9502,
+      "step": 13618
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8082640171051025,
+      "learning_rate": 2.4101943824327855e-06,
+      "loss": 0.9526,
+      "step": 13619
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8229402303695679,
+      "learning_rate": 2.4089849978368917e-06,
+      "loss": 0.9362,
+      "step": 13620
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9007649421691895,
+      "learning_rate": 2.407775875185725e-06,
+      "loss": 0.8726,
+      "step": 13621
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5575422048568726,
+      "learning_rate": 2.4065670145210006e-06,
+      "loss": 0.9016,
+      "step": 13622
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.653222918510437,
+      "learning_rate": 2.4053584158844412e-06,
+      "loss": 0.8523,
+      "step": 13623
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9007463455200195,
+      "learning_rate": 2.4041500793177454e-06,
+      "loss": 0.8686,
+      "step": 13624
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7739113569259644,
+      "learning_rate": 2.402942004862614e-06,
+      "loss": 0.8704,
+      "step": 13625
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7059307098388672,
+      "learning_rate": 2.4017341925607296e-06,
+      "loss": 0.9077,
+      "step": 13626
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8686466217041016,
+      "learning_rate": 2.4005266424537767e-06,
+      "loss": 0.9074,
+      "step": 13627
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6895369291305542,
+      "learning_rate": 2.3993193545834182e-06,
+      "loss": 0.8743,
+      "step": 13628
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6835358142852783,
+      "learning_rate": 2.3981123289913176e-06,
+      "loss": 0.8506,
+      "step": 13629
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.864768385887146,
+      "learning_rate": 2.3969055657191276e-06,
+      "loss": 0.8837,
+      "step": 13630
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8432718515396118,
+      "learning_rate": 2.3956990648084855e-06,
+      "loss": 0.888,
+      "step": 13631
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7115341424942017,
+      "learning_rate": 2.39449282630103e-06,
+      "loss": 0.8813,
+      "step": 13632
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7658716440200806,
+      "learning_rate": 2.3932868502383788e-06,
+      "loss": 0.8436,
+      "step": 13633
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7238898277282715,
+      "learning_rate": 2.3920811366621533e-06,
+      "loss": 0.9932,
+      "step": 13634
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0151317119598389,
+      "learning_rate": 2.3908756856139524e-06,
+      "loss": 0.5165,
+      "step": 13635
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7716412544250488,
+      "learning_rate": 2.389670497135379e-06,
+      "loss": 0.9194,
+      "step": 13636
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8116930723190308,
+      "learning_rate": 2.388465571268016e-06,
+      "loss": 0.9041,
+      "step": 13637
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8976644277572632,
+      "learning_rate": 2.3872609080534436e-06,
+      "loss": 0.9432,
+      "step": 13638
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.69411039352417,
+      "learning_rate": 2.386056507533232e-06,
+      "loss": 0.8547,
+      "step": 13639
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.659548282623291,
+      "learning_rate": 2.384852369748946e-06,
+      "loss": 0.8915,
+      "step": 13640
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.754137396812439,
+      "learning_rate": 2.3836484947421278e-06,
+      "loss": 0.9148,
+      "step": 13641
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7559254169464111,
+      "learning_rate": 2.382444882554328e-06,
+      "loss": 0.9491,
+      "step": 13642
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8168963193893433,
+      "learning_rate": 2.3812415332270742e-06,
+      "loss": 0.8787,
+      "step": 13643
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7588047981262207,
+      "learning_rate": 2.3800384468018954e-06,
+      "loss": 0.9218,
+      "step": 13644
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8390454053878784,
+      "learning_rate": 2.3788356233203014e-06,
+      "loss": 0.8287,
+      "step": 13645
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5857840776443481,
+      "learning_rate": 2.377633062823804e-06,
+      "loss": 0.7898,
+      "step": 13646
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.965441107749939,
+      "learning_rate": 2.3764307653538954e-06,
+      "loss": 0.8902,
+      "step": 13647
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.928942084312439,
+      "learning_rate": 2.3752287309520637e-06,
+      "loss": 0.9285,
+      "step": 13648
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8728222846984863,
+      "learning_rate": 2.3740269596597943e-06,
+      "loss": 0.827,
+      "step": 13649
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9338548183441162,
+      "learning_rate": 2.372825451518549e-06,
+      "loss": 1.0039,
+      "step": 13650
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8214329481124878,
+      "learning_rate": 2.3716242065697938e-06,
+      "loss": 0.9199,
+      "step": 13651
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.948134183883667,
+      "learning_rate": 2.3704232248549753e-06,
+      "loss": 0.8203,
+      "step": 13652
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8129678964614868,
+      "learning_rate": 2.3692225064155427e-06,
+      "loss": 0.9602,
+      "step": 13653
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8240019083023071,
+      "learning_rate": 2.368022051292922e-06,
+      "loss": 0.8448,
+      "step": 13654
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8858380317687988,
+      "learning_rate": 2.366821859528544e-06,
+      "loss": 0.9295,
+      "step": 13655
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7104382514953613,
+      "learning_rate": 2.3656219311638194e-06,
+      "loss": 0.956,
+      "step": 13656
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8154139518737793,
+      "learning_rate": 2.3644222662401583e-06,
+      "loss": 0.9469,
+      "step": 13657
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7367316484451294,
+      "learning_rate": 2.363222864798953e-06,
+      "loss": 0.8136,
+      "step": 13658
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7884933948516846,
+      "learning_rate": 2.362023726881594e-06,
+      "loss": 0.9232,
+      "step": 13659
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8147273063659668,
+      "learning_rate": 2.360824852529463e-06,
+      "loss": 0.9362,
+      "step": 13660
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.663731336593628,
+      "learning_rate": 2.3596262417839256e-06,
+      "loss": 0.8924,
+      "step": 13661
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.760013461112976,
+      "learning_rate": 2.358427894686346e-06,
+      "loss": 0.8986,
+      "step": 13662
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8135193586349487,
+      "learning_rate": 2.3572298112780702e-06,
+      "loss": 0.9135,
+      "step": 13663
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6551241874694824,
+      "learning_rate": 2.356031991600448e-06,
+      "loss": 0.8573,
+      "step": 13664
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7942925691604614,
+      "learning_rate": 2.3548344356948063e-06,
+      "loss": 0.882,
+      "step": 13665
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8022816181182861,
+      "learning_rate": 2.353637143602475e-06,
+      "loss": 0.888,
+      "step": 13666
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7360374927520752,
+      "learning_rate": 2.3524401153647646e-06,
+      "loss": 0.8702,
+      "step": 13667
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7473527193069458,
+      "learning_rate": 2.3512433510229858e-06,
+      "loss": 0.9101,
+      "step": 13668
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.761274814605713,
+      "learning_rate": 2.350046850618429e-06,
+      "loss": 0.9056,
+      "step": 13669
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.79006028175354,
+      "learning_rate": 2.3488506141923907e-06,
+      "loss": 0.8481,
+      "step": 13670
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8377692699432373,
+      "learning_rate": 2.347654641786141e-06,
+      "loss": 1.0105,
+      "step": 13671
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.837234377861023,
+      "learning_rate": 2.346458933440954e-06,
+      "loss": 0.8925,
+      "step": 13672
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.252354621887207,
+      "learning_rate": 2.345263489198093e-06,
+      "loss": 1.0055,
+      "step": 13673
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8179010152816772,
+      "learning_rate": 2.3440683090988024e-06,
+      "loss": 0.8645,
+      "step": 13674
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6748462915420532,
+      "learning_rate": 2.342873393184333e-06,
+      "loss": 0.864,
+      "step": 13675
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6932932138442993,
+      "learning_rate": 2.3416787414959097e-06,
+      "loss": 0.8868,
+      "step": 13676
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9458096027374268,
+      "learning_rate": 2.3404843540747634e-06,
+      "loss": 0.9213,
+      "step": 13677
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8058936595916748,
+      "learning_rate": 2.3392902309621025e-06,
+      "loss": 0.8676,
+      "step": 13678
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7884318828582764,
+      "learning_rate": 2.33809637219914e-06,
+      "loss": 0.8709,
+      "step": 13679
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.700257658958435,
+      "learning_rate": 2.3369027778270657e-06,
+      "loss": 0.9089,
+      "step": 13680
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8179317712783813,
+      "learning_rate": 2.3357094478870747e-06,
+      "loss": 0.8374,
+      "step": 13681
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7130640745162964,
+      "learning_rate": 2.3345163824203377e-06,
+      "loss": 0.884,
+      "step": 13682
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7481637001037598,
+      "learning_rate": 2.3333235814680264e-06,
+      "loss": 0.8403,
+      "step": 13683
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6055896282196045,
+      "learning_rate": 2.3321310450713066e-06,
+      "loss": 0.8416,
+      "step": 13684
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.894039511680603,
+      "learning_rate": 2.330938773271322e-06,
+      "loss": 0.9605,
+      "step": 13685
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6541424989700317,
+      "learning_rate": 2.329746766109221e-06,
+      "loss": 0.9665,
+      "step": 13686
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7017853260040283,
+      "learning_rate": 2.328555023626129e-06,
+      "loss": 0.8191,
+      "step": 13687
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7103816270828247,
+      "learning_rate": 2.327363545863177e-06,
+      "loss": 0.8927,
+      "step": 13688
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7522919178009033,
+      "learning_rate": 2.3261723328614747e-06,
+      "loss": 0.9455,
+      "step": 13689
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9626909494400024,
+      "learning_rate": 2.3249813846621307e-06,
+      "loss": 0.9702,
+      "step": 13690
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9076628684997559,
+      "learning_rate": 2.3237907013062377e-06,
+      "loss": 0.8757,
+      "step": 13691
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0335936546325684,
+      "learning_rate": 2.322600282834888e-06,
+      "loss": 0.9886,
+      "step": 13692
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8790507316589355,
+      "learning_rate": 2.3214101292891535e-06,
+      "loss": 0.8645,
+      "step": 13693
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8087753057479858,
+      "learning_rate": 2.3202202407101084e-06,
+      "loss": 0.8683,
+      "step": 13694
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6997405290603638,
+      "learning_rate": 2.3190306171388077e-06,
+      "loss": 0.9287,
+      "step": 13695
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8791561126708984,
+      "learning_rate": 2.3178412586163046e-06,
+      "loss": 0.9458,
+      "step": 13696
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8820852041244507,
+      "learning_rate": 2.3166521651836437e-06,
+      "loss": 0.9908,
+      "step": 13697
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7594972848892212,
+      "learning_rate": 2.315463336881851e-06,
+      "loss": 0.9198,
+      "step": 13698
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0052285194396973,
+      "learning_rate": 2.3142747737519555e-06,
+      "loss": 0.9397,
+      "step": 13699
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7913938760757446,
+      "learning_rate": 2.3130864758349645e-06,
+      "loss": 0.8404,
+      "step": 13700
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.629380464553833,
+      "learning_rate": 2.3118984431718903e-06,
+      "loss": 0.8831,
+      "step": 13701
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.844380497932434,
+      "learning_rate": 2.3107106758037225e-06,
+      "loss": 0.8203,
+      "step": 13702
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7498242855072021,
+      "learning_rate": 2.309523173771453e-06,
+      "loss": 0.8491,
+      "step": 13703
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.773834466934204,
+      "learning_rate": 2.3083359371160497e-06,
+      "loss": 0.886,
+      "step": 13704
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6575134992599487,
+      "learning_rate": 2.307148965878494e-06,
+      "loss": 0.8911,
+      "step": 13705
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6364333629608154,
+      "learning_rate": 2.3059622600997355e-06,
+      "loss": 0.9379,
+      "step": 13706
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0666403770446777,
+      "learning_rate": 2.304775819820729e-06,
+      "loss": 0.8677,
+      "step": 13707
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8420294523239136,
+      "learning_rate": 2.3035896450824115e-06,
+      "loss": 0.8903,
+      "step": 13708
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9709292054176331,
+      "learning_rate": 2.302403735925718e-06,
+      "loss": 0.479,
+      "step": 13709
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9285016059875488,
+      "learning_rate": 2.3012180923915673e-06,
+      "loss": 0.8433,
+      "step": 13710
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8556067943572998,
+      "learning_rate": 2.300032714520877e-06,
+      "loss": 0.9879,
+      "step": 13711
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8201324939727783,
+      "learning_rate": 2.2988476023545447e-06,
+      "loss": 0.9486,
+      "step": 13712
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.862945795059204,
+      "learning_rate": 2.2976627559334707e-06,
+      "loss": 0.8615,
+      "step": 13713
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8331682682037354,
+      "learning_rate": 2.296478175298542e-06,
+      "loss": 0.9561,
+      "step": 13714
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.592003583908081,
+      "learning_rate": 2.2952938604906303e-06,
+      "loss": 0.9171,
+      "step": 13715
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8083757162094116,
+      "learning_rate": 2.2941098115506065e-06,
+      "loss": 0.9086,
+      "step": 13716
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7005000114440918,
+      "learning_rate": 2.2929260285193266e-06,
+      "loss": 1.0155,
+      "step": 13717
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9146922826766968,
+      "learning_rate": 2.291742511437642e-06,
+      "loss": 0.9498,
+      "step": 13718
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6851366758346558,
+      "learning_rate": 2.2905592603463888e-06,
+      "loss": 0.9733,
+      "step": 13719
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6600128412246704,
+      "learning_rate": 2.2893762752864035e-06,
+      "loss": 0.8193,
+      "step": 13720
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.816359519958496,
+      "learning_rate": 2.2881935562985015e-06,
+      "loss": 0.9327,
+      "step": 13721
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7869069576263428,
+      "learning_rate": 2.2870111034235e-06,
+      "loss": 0.988,
+      "step": 13722
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8428736925125122,
+      "learning_rate": 2.2858289167021963e-06,
+      "loss": 0.9121,
+      "step": 13723
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8796955347061157,
+      "learning_rate": 2.2846469961753916e-06,
+      "loss": 0.8068,
+      "step": 13724
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8612310886383057,
+      "learning_rate": 2.2834653418838647e-06,
+      "loss": 0.894,
+      "step": 13725
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6684895753860474,
+      "learning_rate": 2.282283953868393e-06,
+      "loss": 0.8273,
+      "step": 13726
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7755604982376099,
+      "learning_rate": 2.281102832169747e-06,
+      "loss": 0.9481,
+      "step": 13727
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.819770336151123,
+      "learning_rate": 2.2799219768286774e-06,
+      "loss": 0.8916,
+      "step": 13728
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9963771104812622,
+      "learning_rate": 2.278741387885938e-06,
+      "loss": 0.5464,
+      "step": 13729
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.715556025505066,
+      "learning_rate": 2.277561065382261e-06,
+      "loss": 0.956,
+      "step": 13730
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7296901941299438,
+      "learning_rate": 2.276381009358384e-06,
+      "loss": 0.9237,
+      "step": 13731
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8876839876174927,
+      "learning_rate": 2.27520121985502e-06,
+      "loss": 0.9787,
+      "step": 13732
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9505195617675781,
+      "learning_rate": 2.274021696912886e-06,
+      "loss": 0.497,
+      "step": 13733
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7306569814682007,
+      "learning_rate": 2.272842440572679e-06,
+      "loss": 0.921,
+      "step": 13734
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.859546184539795,
+      "learning_rate": 2.271663450875097e-06,
+      "loss": 0.9257,
+      "step": 13735
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.832834005355835,
+      "learning_rate": 2.2704847278608187e-06,
+      "loss": 0.8853,
+      "step": 13736
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7615388631820679,
+      "learning_rate": 2.2693062715705203e-06,
+      "loss": 0.874,
+      "step": 13737
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.811547040939331,
+      "learning_rate": 2.2681280820448715e-06,
+      "loss": 0.9464,
+      "step": 13738
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.780483603477478,
+      "learning_rate": 2.2669501593245214e-06,
+      "loss": 0.979,
+      "step": 13739
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7343541383743286,
+      "learning_rate": 2.265772503450122e-06,
+      "loss": 0.881,
+      "step": 13740
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9140340089797974,
+      "learning_rate": 2.264595114462307e-06,
+      "loss": 0.9782,
+      "step": 13741
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7881622314453125,
+      "learning_rate": 2.2634179924017086e-06,
+      "loss": 0.7928,
+      "step": 13742
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8407269716262817,
+      "learning_rate": 2.2622411373089415e-06,
+      "loss": 0.9731,
+      "step": 13743
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7623176574707031,
+      "learning_rate": 2.2610645492246207e-06,
+      "loss": 0.9707,
+      "step": 13744
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6502431631088257,
+      "learning_rate": 2.2598882281893417e-06,
+      "loss": 0.7793,
+      "step": 13745
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.78104829788208,
+      "learning_rate": 2.2587121742437024e-06,
+      "loss": 0.8254,
+      "step": 13746
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.828385829925537,
+      "learning_rate": 2.2575363874282784e-06,
+      "loss": 0.9187,
+      "step": 13747
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6948174238204956,
+      "learning_rate": 2.256360867783648e-06,
+      "loss": 0.966,
+      "step": 13748
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7490049600601196,
+      "learning_rate": 2.2551856153503714e-06,
+      "loss": 0.8107,
+      "step": 13749
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.704943060874939,
+      "learning_rate": 2.2540106301690044e-06,
+      "loss": 0.9108,
+      "step": 13750
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8893815279006958,
+      "learning_rate": 2.2528359122800957e-06,
+      "loss": 0.891,
+      "step": 13751
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7886872291564941,
+      "learning_rate": 2.251661461724176e-06,
+      "loss": 0.9073,
+      "step": 13752
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7814863920211792,
+      "learning_rate": 2.2504872785417776e-06,
+      "loss": 0.9209,
+      "step": 13753
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8895375728607178,
+      "learning_rate": 2.249313362773414e-06,
+      "loss": 0.9159,
+      "step": 13754
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7596838474273682,
+      "learning_rate": 2.2481397144595975e-06,
+      "loss": 0.9126,
+      "step": 13755
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8434736728668213,
+      "learning_rate": 2.246966333640823e-06,
+      "loss": 0.9373,
+      "step": 13756
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8783890008926392,
+      "learning_rate": 2.245793220357586e-06,
+      "loss": 0.9616,
+      "step": 13757
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6653900146484375,
+      "learning_rate": 2.2446203746503626e-06,
+      "loss": 0.9444,
+      "step": 13758
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6814123392105103,
+      "learning_rate": 2.243447796559628e-06,
+      "loss": 0.9646,
+      "step": 13759
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7036863565444946,
+      "learning_rate": 2.2422754861258402e-06,
+      "loss": 0.8768,
+      "step": 13760
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6109156608581543,
+      "learning_rate": 2.241103443389455e-06,
+      "loss": 0.9263,
+      "step": 13761
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7284893989562988,
+      "learning_rate": 2.239931668390919e-06,
+      "loss": 0.9258,
+      "step": 13762
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7164733409881592,
+      "learning_rate": 2.238760161170662e-06,
+      "loss": 0.9379,
+      "step": 13763
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7378226518630981,
+      "learning_rate": 2.237588921769114e-06,
+      "loss": 0.9341,
+      "step": 13764
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7952555418014526,
+      "learning_rate": 2.236417950226686e-06,
+      "loss": 0.9198,
+      "step": 13765
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6079233884811401,
+      "learning_rate": 2.2352472465837915e-06,
+      "loss": 0.8893,
+      "step": 13766
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.808095097541809,
+      "learning_rate": 2.234076810880821e-06,
+      "loss": 0.9644,
+      "step": 13767
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7293614149093628,
+      "learning_rate": 2.2329066431581693e-06,
+      "loss": 0.9086,
+      "step": 13768
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.137747049331665,
+      "learning_rate": 2.23173674345621e-06,
+      "loss": 0.9378,
+      "step": 13769
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8463839292526245,
+      "learning_rate": 2.230567111815316e-06,
+      "loss": 0.8991,
+      "step": 13770
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8261408805847168,
+      "learning_rate": 2.229397748275849e-06,
+      "loss": 0.886,
+      "step": 13771
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6731531620025635,
+      "learning_rate": 2.2282286528781604e-06,
+      "loss": 0.9302,
+      "step": 13772
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8516348600387573,
+      "learning_rate": 2.22705982566259e-06,
+      "loss": 0.9173,
+      "step": 13773
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6551576852798462,
+      "learning_rate": 2.225891266669474e-06,
+      "loss": 0.9373,
+      "step": 13774
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.026163935661316,
+      "learning_rate": 2.224722975939133e-06,
+      "loss": 0.6001,
+      "step": 13775
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5885018110275269,
+      "learning_rate": 2.2235549535118838e-06,
+      "loss": 0.9574,
+      "step": 13776
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.783301830291748,
+      "learning_rate": 2.222387199428029e-06,
+      "loss": 0.8903,
+      "step": 13777
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7355778217315674,
+      "learning_rate": 2.221219713727868e-06,
+      "loss": 0.9144,
+      "step": 13778
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7261461019515991,
+      "learning_rate": 2.2200524964516835e-06,
+      "loss": 0.8655,
+      "step": 13779
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.708254337310791,
+      "learning_rate": 2.218885547639754e-06,
+      "loss": 0.8532,
+      "step": 13780
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8537296056747437,
+      "learning_rate": 2.2177188673323523e-06,
+      "loss": 0.9126,
+      "step": 13781
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7386155128479004,
+      "learning_rate": 2.2165524555697306e-06,
+      "loss": 0.8627,
+      "step": 13782
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.623005747795105,
+      "learning_rate": 2.2153863123921435e-06,
+      "loss": 0.8558,
+      "step": 13783
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5860615968704224,
+      "learning_rate": 2.214220437839827e-06,
+      "loss": 0.7943,
+      "step": 13784
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8126397132873535,
+      "learning_rate": 2.2130548319530177e-06,
+      "loss": 0.8235,
+      "step": 13785
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6117854118347168,
+      "learning_rate": 2.2118894947719305e-06,
+      "loss": 0.8962,
+      "step": 13786
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9143046140670776,
+      "learning_rate": 2.2107244263367855e-06,
+      "loss": 0.8604,
+      "step": 13787
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6688520908355713,
+      "learning_rate": 2.2095596266877783e-06,
+      "loss": 0.9114,
+      "step": 13788
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7845020294189453,
+      "learning_rate": 2.2083950958651103e-06,
+      "loss": 0.9671,
+      "step": 13789
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8241589069366455,
+      "learning_rate": 2.2072308339089597e-06,
+      "loss": 0.8848,
+      "step": 13790
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8034124374389648,
+      "learning_rate": 2.206066840859504e-06,
+      "loss": 0.9238,
+      "step": 13791
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7303788661956787,
+      "learning_rate": 2.2049031167569134e-06,
+      "loss": 0.9272,
+      "step": 13792
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6216984987258911,
+      "learning_rate": 2.2037396616413386e-06,
+      "loss": 0.9558,
+      "step": 13793
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8044848442077637,
+      "learning_rate": 2.202576475552933e-06,
+      "loss": 0.8795,
+      "step": 13794
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7683902978897095,
+      "learning_rate": 2.2014135585318296e-06,
+      "loss": 0.9747,
+      "step": 13795
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8716888427734375,
+      "learning_rate": 2.2002509106181625e-06,
+      "loss": 0.9392,
+      "step": 13796
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8777893781661987,
+      "learning_rate": 2.199088531852046e-06,
+      "loss": 1.0105,
+      "step": 13797
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8201268911361694,
+      "learning_rate": 2.197926422273595e-06,
+      "loss": 0.8881,
+      "step": 13798
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8586938381195068,
+      "learning_rate": 2.1967645819229077e-06,
+      "loss": 1.0509,
+      "step": 13799
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5527417659759521,
+      "learning_rate": 2.1956030108400796e-06,
+      "loss": 0.8054,
+      "step": 13800
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7363436222076416,
+      "learning_rate": 2.194441709065187e-06,
+      "loss": 0.9149,
+      "step": 13801
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8199156522750854,
+      "learning_rate": 2.193280676638311e-06,
+      "loss": 0.896,
+      "step": 13802
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.851406455039978,
+      "learning_rate": 2.1921199135995086e-06,
+      "loss": 0.9581,
+      "step": 13803
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7970741987228394,
+      "learning_rate": 2.1909594199888374e-06,
+      "loss": 0.9468,
+      "step": 13804
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7619885206222534,
+      "learning_rate": 2.189799195846346e-06,
+      "loss": 0.9835,
+      "step": 13805
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7285693883895874,
+      "learning_rate": 2.188639241212065e-06,
+      "loss": 0.9125,
+      "step": 13806
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0640184879302979,
+      "learning_rate": 2.1874795561260256e-06,
+      "loss": 0.516,
+      "step": 13807
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7181600332260132,
+      "learning_rate": 2.186320140628241e-06,
+      "loss": 0.8774,
+      "step": 13808
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.740336537361145,
+      "learning_rate": 2.185160994758724e-06,
+      "loss": 0.9246,
+      "step": 13809
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8006306886672974,
+      "learning_rate": 2.184002118557469e-06,
+      "loss": 0.8765,
+      "step": 13810
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.747934341430664,
+      "learning_rate": 2.1828435120644698e-06,
+      "loss": 0.9317,
+      "step": 13811
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9009658098220825,
+      "learning_rate": 2.1816851753197023e-06,
+      "loss": 0.9235,
+      "step": 13812
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7725614309310913,
+      "learning_rate": 2.180527108363143e-06,
+      "loss": 0.902,
+      "step": 13813
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.901047945022583,
+      "learning_rate": 2.179369311234747e-06,
+      "loss": 0.8909,
+      "step": 13814
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7487728595733643,
+      "learning_rate": 2.178211783974471e-06,
+      "loss": 0.8616,
+      "step": 13815
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6686267852783203,
+      "learning_rate": 2.1770545266222587e-06,
+      "loss": 0.9567,
+      "step": 13816
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7387621402740479,
+      "learning_rate": 2.1758975392180405e-06,
+      "loss": 0.9245,
+      "step": 13817
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6347938776016235,
+      "learning_rate": 2.174740821801744e-06,
+      "loss": 0.8989,
+      "step": 13818
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7126402854919434,
+      "learning_rate": 2.17358437441328e-06,
+      "loss": 0.9098,
+      "step": 13819
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9313092827796936,
+      "learning_rate": 2.172428197092561e-06,
+      "loss": 0.4543,
+      "step": 13820
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7997554540634155,
+      "learning_rate": 2.1712722898794756e-06,
+      "loss": 0.8912,
+      "step": 13821
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8088159561157227,
+      "learning_rate": 2.1701166528139182e-06,
+      "loss": 0.9595,
+      "step": 13822
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8501399755477905,
+      "learning_rate": 2.16896128593576e-06,
+      "loss": 0.9482,
+      "step": 13823
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0146775245666504,
+      "learning_rate": 2.167806189284877e-06,
+      "loss": 0.5353,
+      "step": 13824
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5741924047470093,
+      "learning_rate": 2.166651362901119e-06,
+      "loss": 0.9345,
+      "step": 13825
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9845336079597473,
+      "learning_rate": 2.1654968068243455e-06,
+      "loss": 0.543,
+      "step": 13826
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6851930618286133,
+      "learning_rate": 2.164342521094388e-06,
+      "loss": 0.8325,
+      "step": 13827
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.082017421722412,
+      "learning_rate": 2.1631885057510836e-06,
+      "loss": 0.5781,
+      "step": 13828
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.819605827331543,
+      "learning_rate": 2.162034760834254e-06,
+      "loss": 0.9403,
+      "step": 13829
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9804980754852295,
+      "learning_rate": 2.160881286383708e-06,
+      "loss": 0.5368,
+      "step": 13830
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7803066968917847,
+      "learning_rate": 2.159728082439255e-06,
+      "loss": 0.8666,
+      "step": 13831
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7022088766098022,
+      "learning_rate": 2.1585751490406816e-06,
+      "loss": 0.9315,
+      "step": 13832
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9040809869766235,
+      "learning_rate": 2.157422486227778e-06,
+      "loss": 0.8583,
+      "step": 13833
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7470202445983887,
+      "learning_rate": 2.1562700940403134e-06,
+      "loss": 0.9411,
+      "step": 13834
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8454785346984863,
+      "learning_rate": 2.1551179725180613e-06,
+      "loss": 0.9635,
+      "step": 13835
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8282545804977417,
+      "learning_rate": 2.153966121700769e-06,
+      "loss": 0.9495,
+      "step": 13836
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.636711597442627,
+      "learning_rate": 2.152814541628193e-06,
+      "loss": 0.9873,
+      "step": 13837
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.758360743522644,
+      "learning_rate": 2.1516632323400656e-06,
+      "loss": 0.8658,
+      "step": 13838
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8387466669082642,
+      "learning_rate": 2.1505121938761184e-06,
+      "loss": 0.8796,
+      "step": 13839
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.75312340259552,
+      "learning_rate": 2.1493614262760664e-06,
+      "loss": 0.97,
+      "step": 13840
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9700262546539307,
+      "learning_rate": 2.148210929579625e-06,
+      "loss": 0.875,
+      "step": 13841
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6707663536071777,
+      "learning_rate": 2.1470607038264878e-06,
+      "loss": 0.8742,
+      "step": 13842
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.769095540046692,
+      "learning_rate": 2.1459107490563522e-06,
+      "loss": 0.9435,
+      "step": 13843
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7184206247329712,
+      "learning_rate": 2.1447610653088946e-06,
+      "loss": 0.9229,
+      "step": 13844
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0802392959594727,
+      "learning_rate": 2.1436116526237894e-06,
+      "loss": 0.9026,
+      "step": 13845
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9855695962905884,
+      "learning_rate": 2.1424625110407036e-06,
+      "loss": 0.5231,
+      "step": 13846
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7364611625671387,
+      "learning_rate": 2.141313640599284e-06,
+      "loss": 0.9147,
+      "step": 13847
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.954888939857483,
+      "learning_rate": 2.1401650413391816e-06,
+      "loss": 0.8847,
+      "step": 13848
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7651033401489258,
+      "learning_rate": 2.139016713300025e-06,
+      "loss": 0.9216,
+      "step": 13849
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6987709999084473,
+      "learning_rate": 2.137868656521446e-06,
+      "loss": 0.8798,
+      "step": 13850
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0484561920166016,
+      "learning_rate": 2.136720871043054e-06,
+      "loss": 0.9179,
+      "step": 13851
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7574348449707031,
+      "learning_rate": 2.1355733569044633e-06,
+      "loss": 0.8653,
+      "step": 13852
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9542453289031982,
+      "learning_rate": 2.134426114145265e-06,
+      "loss": 0.9537,
+      "step": 13853
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7410023212432861,
+      "learning_rate": 2.1332791428050526e-06,
+      "loss": 0.8958,
+      "step": 13854
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8932121992111206,
+      "learning_rate": 2.132132442923398e-06,
+      "loss": 0.8597,
+      "step": 13855
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7822086811065674,
+      "learning_rate": 2.1309860145398788e-06,
+      "loss": 0.8556,
+      "step": 13856
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7748568058013916,
+      "learning_rate": 2.129839857694048e-06,
+      "loss": 0.8818,
+      "step": 13857
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.751509189605713,
+      "learning_rate": 2.1286939724254598e-06,
+      "loss": 0.9335,
+      "step": 13858
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7816684246063232,
+      "learning_rate": 2.1275483587736577e-06,
+      "loss": 0.9292,
+      "step": 13859
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6210846900939941,
+      "learning_rate": 2.126403016778168e-06,
+      "loss": 0.8625,
+      "step": 13860
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6334526538848877,
+      "learning_rate": 2.1252579464785185e-06,
+      "loss": 0.9132,
+      "step": 13861
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.672934651374817,
+      "learning_rate": 2.1241131479142175e-06,
+      "loss": 0.978,
+      "step": 13862
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6934659481048584,
+      "learning_rate": 2.1229686211247737e-06,
+      "loss": 0.8937,
+      "step": 13863
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.836025357246399,
+      "learning_rate": 2.1218243661496773e-06,
+      "loss": 0.9629,
+      "step": 13864
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7841970920562744,
+      "learning_rate": 2.120680383028417e-06,
+      "loss": 0.8626,
+      "step": 13865
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6765799522399902,
+      "learning_rate": 2.119536671800465e-06,
+      "loss": 0.8911,
+      "step": 13866
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7843064069747925,
+      "learning_rate": 2.1183932325052915e-06,
+      "loss": 0.9016,
+      "step": 13867
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8544692993164062,
+      "learning_rate": 2.117250065182349e-06,
+      "loss": 0.9317,
+      "step": 13868
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6293355226516724,
+      "learning_rate": 2.1161071698710866e-06,
+      "loss": 0.9438,
+      "step": 13869
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.718941569328308,
+      "learning_rate": 2.114964546610946e-06,
+      "loss": 0.8709,
+      "step": 13870
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6374696493148804,
+      "learning_rate": 2.1138221954413496e-06,
+      "loss": 0.9263,
+      "step": 13871
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8791415691375732,
+      "learning_rate": 2.112680116401723e-06,
+      "loss": 0.9,
+      "step": 13872
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9028964042663574,
+      "learning_rate": 2.1115383095314712e-06,
+      "loss": 0.9502,
+      "step": 13873
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6713298559188843,
+      "learning_rate": 2.1103967748699995e-06,
+      "loss": 0.9257,
+      "step": 13874
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9702057242393494,
+      "learning_rate": 2.1092555124566925e-06,
+      "loss": 0.5332,
+      "step": 13875
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6842060089111328,
+      "learning_rate": 2.1081145223309397e-06,
+      "loss": 1.0034,
+      "step": 13876
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8811030387878418,
+      "learning_rate": 2.1069738045321063e-06,
+      "loss": 0.9115,
+      "step": 13877
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6667803525924683,
+      "learning_rate": 2.1058333590995617e-06,
+      "loss": 0.9203,
+      "step": 13878
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5625908374786377,
+      "learning_rate": 2.104693186072654e-06,
+      "loss": 0.9246,
+      "step": 13879
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.691623330116272,
+      "learning_rate": 2.1035532854907314e-06,
+      "loss": 0.9152,
+      "step": 13880
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8293339014053345,
+      "learning_rate": 2.1024136573931252e-06,
+      "loss": 0.9322,
+      "step": 13881
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6824750900268555,
+      "learning_rate": 2.101274301819163e-06,
+      "loss": 0.9162,
+      "step": 13882
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0978751182556152,
+      "learning_rate": 2.1001352188081627e-06,
+      "loss": 0.5641,
+      "step": 13883
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9204086065292358,
+      "learning_rate": 2.0989964083994254e-06,
+      "loss": 1.0214,
+      "step": 13884
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8783453702926636,
+      "learning_rate": 2.097857870632255e-06,
+      "loss": 0.8456,
+      "step": 13885
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.049328565597534,
+      "learning_rate": 2.0967196055459327e-06,
+      "loss": 0.9211,
+      "step": 13886
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8419142961502075,
+      "learning_rate": 2.0955816131797425e-06,
+      "loss": 0.9897,
+      "step": 13887
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9320902824401855,
+      "learning_rate": 2.0944438935729484e-06,
+      "loss": 0.9657,
+      "step": 13888
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9545713663101196,
+      "learning_rate": 2.0933064467648147e-06,
+      "loss": 0.8988,
+      "step": 13889
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8225746154785156,
+      "learning_rate": 2.0921692727945863e-06,
+      "loss": 1.001,
+      "step": 13890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.784244179725647,
+      "learning_rate": 2.091032371701509e-06,
+      "loss": 0.7933,
+      "step": 13891
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.695263385772705,
+      "learning_rate": 2.08989574352481e-06,
+      "loss": 0.8289,
+      "step": 13892
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7006781101226807,
+      "learning_rate": 2.0887593883037116e-06,
+      "loss": 0.9493,
+      "step": 13893
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7475663423538208,
+      "learning_rate": 2.087623306077431e-06,
+      "loss": 0.8839,
+      "step": 13894
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7519831657409668,
+      "learning_rate": 2.0864874968851657e-06,
+      "loss": 0.8756,
+      "step": 13895
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.724940538406372,
+      "learning_rate": 2.085351960766113e-06,
+      "loss": 0.8132,
+      "step": 13896
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9793107509613037,
+      "learning_rate": 2.0842166977594538e-06,
+      "loss": 0.891,
+      "step": 13897
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.643418312072754,
+      "learning_rate": 2.0830817079043663e-06,
+      "loss": 1.0207,
+      "step": 13898
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8368695974349976,
+      "learning_rate": 2.0819469912400113e-06,
+      "loss": 0.9198,
+      "step": 13899
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6786694526672363,
+      "learning_rate": 2.0808125478055507e-06,
+      "loss": 0.9134,
+      "step": 13900
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7283631563186646,
+      "learning_rate": 2.0796783776401252e-06,
+      "loss": 0.9707,
+      "step": 13901
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.717724084854126,
+      "learning_rate": 2.0785444807828737e-06,
+      "loss": 0.902,
+      "step": 13902
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.774915099143982,
+      "learning_rate": 2.0774108572729256e-06,
+      "loss": 0.8784,
+      "step": 13903
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7804466485977173,
+      "learning_rate": 2.076277507149399e-06,
+      "loss": 0.89,
+      "step": 13904
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5639712810516357,
+      "learning_rate": 2.0751444304514002e-06,
+      "loss": 0.9027,
+      "step": 13905
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.874142050743103,
+      "learning_rate": 2.074011627218032e-06,
+      "loss": 0.9437,
+      "step": 13906
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5796946287155151,
+      "learning_rate": 2.0728790974883793e-06,
+      "loss": 0.8641,
+      "step": 13907
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9472652673721313,
+      "learning_rate": 2.0717468413015285e-06,
+      "loss": 0.8275,
+      "step": 13908
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7220498323440552,
+      "learning_rate": 2.0706148586965457e-06,
+      "loss": 0.8823,
+      "step": 13909
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7351999282836914,
+      "learning_rate": 2.0694831497124958e-06,
+      "loss": 0.8439,
+      "step": 13910
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6093543767929077,
+      "learning_rate": 2.068351714388427e-06,
+      "loss": 0.8857,
+      "step": 13911
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8661140203475952,
+      "learning_rate": 2.0672205527633837e-06,
+      "loss": 0.8783,
+      "step": 13912
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6557588577270508,
+      "learning_rate": 2.066089664876404e-06,
+      "loss": 0.8129,
+      "step": 13913
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6607632637023926,
+      "learning_rate": 2.064959050766504e-06,
+      "loss": 0.85,
+      "step": 13914
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7957963943481445,
+      "learning_rate": 2.063828710472704e-06,
+      "loss": 0.9353,
+      "step": 13915
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7717972993850708,
+      "learning_rate": 2.0626986440340036e-06,
+      "loss": 0.9331,
+      "step": 13916
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.737707495689392,
+      "learning_rate": 2.061568851489404e-06,
+      "loss": 0.8456,
+      "step": 13917
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.75895094871521,
+      "learning_rate": 2.060439332877886e-06,
+      "loss": 0.9747,
+      "step": 13918
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6888923645019531,
+      "learning_rate": 2.0593100882384297e-06,
+      "loss": 0.9101,
+      "step": 13919
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7585532665252686,
+      "learning_rate": 2.0581811176099997e-06,
+      "loss": 0.9234,
+      "step": 13920
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.872012734413147,
+      "learning_rate": 2.057052421031557e-06,
+      "loss": 0.9339,
+      "step": 13921
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7240371704101562,
+      "learning_rate": 2.0559239985420444e-06,
+      "loss": 0.8901,
+      "step": 13922
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0078750848770142,
+      "learning_rate": 2.0547958501804034e-06,
+      "loss": 0.5132,
+      "step": 13923
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1504322290420532,
+      "learning_rate": 2.053667975985567e-06,
+      "loss": 0.5625,
+      "step": 13924
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8820008039474487,
+      "learning_rate": 2.052540375996449e-06,
+      "loss": 0.9026,
+      "step": 13925
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7437798976898193,
+      "learning_rate": 2.051413050251965e-06,
+      "loss": 0.9219,
+      "step": 13926
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7602156400680542,
+      "learning_rate": 2.0502859987910097e-06,
+      "loss": 0.9531,
+      "step": 13927
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6916152238845825,
+      "learning_rate": 2.0491592216524813e-06,
+      "loss": 0.9219,
+      "step": 13928
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9786920547485352,
+      "learning_rate": 2.048032718875255e-06,
+      "loss": 0.914,
+      "step": 13929
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8582561016082764,
+      "learning_rate": 2.0469064904982094e-06,
+      "loss": 0.9101,
+      "step": 13930
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.024723529815674,
+      "learning_rate": 2.0457805365602023e-06,
+      "loss": 0.8898,
+      "step": 13931
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7189216613769531,
+      "learning_rate": 2.0446548571000936e-06,
+      "loss": 0.8826,
+      "step": 13932
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8184411525726318,
+      "learning_rate": 2.0435294521567194e-06,
+      "loss": 0.8386,
+      "step": 13933
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6967767477035522,
+      "learning_rate": 2.0424043217689204e-06,
+      "loss": 0.9319,
+      "step": 13934
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6388914585113525,
+      "learning_rate": 2.0412794659755187e-06,
+      "loss": 1.0292,
+      "step": 13935
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6549962759017944,
+      "learning_rate": 2.0401548848153296e-06,
+      "loss": 0.9041,
+      "step": 13936
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5967315435409546,
+      "learning_rate": 2.0390305783271636e-06,
+      "loss": 0.8298,
+      "step": 13937
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5790693759918213,
+      "learning_rate": 2.0379065465498114e-06,
+      "loss": 0.9771,
+      "step": 13938
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5990328788757324,
+      "learning_rate": 2.036782789522066e-06,
+      "loss": 0.8833,
+      "step": 13939
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6504746675491333,
+      "learning_rate": 2.035659307282699e-06,
+      "loss": 0.8712,
+      "step": 13940
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7570942640304565,
+      "learning_rate": 2.0345360998704843e-06,
+      "loss": 0.9264,
+      "step": 13941
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0953291654586792,
+      "learning_rate": 2.0334131673241763e-06,
+      "loss": 0.5458,
+      "step": 13942
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.848275899887085,
+      "learning_rate": 2.0322905096825283e-06,
+      "loss": 0.8963,
+      "step": 13943
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.793375015258789,
+      "learning_rate": 2.0311681269842755e-06,
+      "loss": 0.8808,
+      "step": 13944
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6269214153289795,
+      "learning_rate": 2.030046019268154e-06,
+      "loss": 0.9117,
+      "step": 13945
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8680565357208252,
+      "learning_rate": 2.028924186572877e-06,
+      "loss": 1.0309,
+      "step": 13946
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8572474718093872,
+      "learning_rate": 2.027802628937161e-06,
+      "loss": 0.8297,
+      "step": 13947
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.782065510749817,
+      "learning_rate": 2.0266813463997092e-06,
+      "loss": 0.8893,
+      "step": 13948
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.405534029006958,
+      "learning_rate": 2.0255603389992084e-06,
+      "loss": 0.4688,
+      "step": 13949
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8521493673324585,
+      "learning_rate": 2.024439606774349e-06,
+      "loss": 0.9972,
+      "step": 13950
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5672870874404907,
+      "learning_rate": 2.023319149763796e-06,
+      "loss": 0.9191,
+      "step": 13951
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.757883071899414,
+      "learning_rate": 2.0221989680062193e-06,
+      "loss": 0.8822,
+      "step": 13952
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7798980474472046,
+      "learning_rate": 2.02107906154027e-06,
+      "loss": 0.9268,
+      "step": 13953
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.721358299255371,
+      "learning_rate": 2.0199594304045956e-06,
+      "loss": 0.9446,
+      "step": 13954
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7716293334960938,
+      "learning_rate": 2.0188400746378268e-06,
+      "loss": 0.8869,
+      "step": 13955
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8402817249298096,
+      "learning_rate": 2.017720994278596e-06,
+      "loss": 1.0162,
+      "step": 13956
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.780182123184204,
+      "learning_rate": 2.0166021893655143e-06,
+      "loss": 0.9375,
+      "step": 13957
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8040544986724854,
+      "learning_rate": 2.0154836599371917e-06,
+      "loss": 0.8863,
+      "step": 13958
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.145618200302124,
+      "learning_rate": 2.0143654060322214e-06,
+      "loss": 0.928,
+      "step": 13959
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.738170862197876,
+      "learning_rate": 2.0132474276891945e-06,
+      "loss": 0.9122,
+      "step": 13960
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.910254955291748,
+      "learning_rate": 2.012129724946692e-06,
+      "loss": 0.922,
+      "step": 13961
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7241897583007812,
+      "learning_rate": 2.0110122978432754e-06,
+      "loss": 0.8984,
+      "step": 13962
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1332030296325684,
+      "learning_rate": 2.009895146417512e-06,
+      "loss": 0.5619,
+      "step": 13963
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7244709730148315,
+      "learning_rate": 2.008778270707944e-06,
+      "loss": 0.8796,
+      "step": 13964
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.96401047706604,
+      "learning_rate": 2.007661670753118e-06,
+      "loss": 0.5084,
+      "step": 13965
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0133216381073,
+      "learning_rate": 2.0065453465915608e-06,
+      "loss": 0.956,
+      "step": 13966
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7604944705963135,
+      "learning_rate": 2.0054292982617964e-06,
+      "loss": 0.8845,
+      "step": 13967
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7436622381210327,
+      "learning_rate": 2.0043135258023294e-06,
+      "loss": 0.8628,
+      "step": 13968
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8912303447723389,
+      "learning_rate": 2.003198029251674e-06,
+      "loss": 0.8703,
+      "step": 13969
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7419434785842896,
+      "learning_rate": 2.0020828086483124e-06,
+      "loss": 0.9019,
+      "step": 13970
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7680869102478027,
+      "learning_rate": 2.000967864030735e-06,
+      "loss": 0.867,
+      "step": 13971
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8261722326278687,
+      "learning_rate": 1.99985319543741e-06,
+      "loss": 0.9777,
+      "step": 13972
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7912230491638184,
+      "learning_rate": 1.9987388029068068e-06,
+      "loss": 0.9246,
+      "step": 13973
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9042878150939941,
+      "learning_rate": 1.997624686477373e-06,
+      "loss": 0.9507,
+      "step": 13974
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7716833353042603,
+      "learning_rate": 1.9965108461875602e-06,
+      "loss": 0.8996,
+      "step": 13975
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7671964168548584,
+      "learning_rate": 1.9953972820757994e-06,
+      "loss": 0.888,
+      "step": 13976
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7693727016448975,
+      "learning_rate": 1.9942839941805183e-06,
+      "loss": 0.8743,
+      "step": 13977
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.862221360206604,
+      "learning_rate": 1.9931709825401358e-06,
+      "loss": 0.9559,
+      "step": 13978
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8267016410827637,
+      "learning_rate": 1.992058247193054e-06,
+      "loss": 0.9201,
+      "step": 13979
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7337477207183838,
+      "learning_rate": 1.990945788177676e-06,
+      "loss": 0.8135,
+      "step": 13980
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6489856243133545,
+      "learning_rate": 1.989833605532383e-06,
+      "loss": 0.881,
+      "step": 13981
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5279343128204346,
+      "learning_rate": 1.9887216992955605e-06,
+      "loss": 0.8935,
+      "step": 13982
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5997909307479858,
+      "learning_rate": 1.98761006950557e-06,
+      "loss": 0.9414,
+      "step": 13983
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7981884479522705,
+      "learning_rate": 1.9864987162007764e-06,
+      "loss": 0.8647,
+      "step": 13984
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8458565473556519,
+      "learning_rate": 1.985387639419526e-06,
+      "loss": 0.8627,
+      "step": 13985
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6984550952911377,
+      "learning_rate": 1.984276839200162e-06,
+      "loss": 0.9655,
+      "step": 13986
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7024176120758057,
+      "learning_rate": 1.983166315581011e-06,
+      "loss": 0.9092,
+      "step": 13987
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8675802946090698,
+      "learning_rate": 1.9820560686003985e-06,
+      "loss": 0.9218,
+      "step": 13988
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.82809579372406,
+      "learning_rate": 1.9809460982966323e-06,
+      "loss": 0.9861,
+      "step": 13989
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7346214056015015,
+      "learning_rate": 1.979836404708014e-06,
+      "loss": 0.9485,
+      "step": 13990
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8084725141525269,
+      "learning_rate": 1.978726987872842e-06,
+      "loss": 0.8094,
+      "step": 13991
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7948328256607056,
+      "learning_rate": 1.9776178478293926e-06,
+      "loss": 0.9049,
+      "step": 13992
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7461326122283936,
+      "learning_rate": 1.9765089846159433e-06,
+      "loss": 0.8277,
+      "step": 13993
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7405855655670166,
+      "learning_rate": 1.9754003982707546e-06,
+      "loss": 0.9028,
+      "step": 13994
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9237440824508667,
+      "learning_rate": 1.974292088832085e-06,
+      "loss": 0.9199,
+      "step": 13995
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2398149967193604,
+      "learning_rate": 1.973184056338173e-06,
+      "loss": 0.9631,
+      "step": 13996
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7465142011642456,
+      "learning_rate": 1.9720763008272604e-06,
+      "loss": 0.824,
+      "step": 13997
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7676247358322144,
+      "learning_rate": 1.970968822337567e-06,
+      "loss": 0.9268,
+      "step": 13998
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8269869089126587,
+      "learning_rate": 1.969861620907314e-06,
+      "loss": 0.8781,
+      "step": 13999
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7873669862747192,
+      "learning_rate": 1.9687546965747018e-06,
+      "loss": 0.8698,
+      "step": 14000
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7634211778640747,
+      "learning_rate": 1.9676480493779314e-06,
+      "loss": 0.9295,
+      "step": 14001
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7665032148361206,
+      "learning_rate": 1.9665416793551917e-06,
+      "loss": 0.9265,
+      "step": 14002
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9484899044036865,
+      "learning_rate": 1.965435586544656e-06,
+      "loss": 0.9425,
+      "step": 14003
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7135900259017944,
+      "learning_rate": 1.9643297709844964e-06,
+      "loss": 0.9304,
+      "step": 14004
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8288850784301758,
+      "learning_rate": 1.963224232712868e-06,
+      "loss": 0.8891,
+      "step": 14005
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6682873964309692,
+      "learning_rate": 1.9621189717679236e-06,
+      "loss": 0.9036,
+      "step": 14006
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8813554048538208,
+      "learning_rate": 1.9610139881877977e-06,
+      "loss": 0.9836,
+      "step": 14007
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.85567045211792,
+      "learning_rate": 1.9599092820106257e-06,
+      "loss": 0.8597,
+      "step": 14008
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7214148044586182,
+      "learning_rate": 1.958804853274523e-06,
+      "loss": 0.8757,
+      "step": 14009
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8692643642425537,
+      "learning_rate": 1.957700702017604e-06,
+      "loss": 0.9252,
+      "step": 14010
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.819067358970642,
+      "learning_rate": 1.956596828277968e-06,
+      "loss": 0.8756,
+      "step": 14011
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8559890985488892,
+      "learning_rate": 1.9554932320937083e-06,
+      "loss": 0.9353,
+      "step": 14012
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7926456928253174,
+      "learning_rate": 1.9543899135029034e-06,
+      "loss": 0.9687,
+      "step": 14013
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7329295873641968,
+      "learning_rate": 1.953286872543628e-06,
+      "loss": 0.8625,
+      "step": 14014
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5481125116348267,
+      "learning_rate": 1.9521841092539485e-06,
+      "loss": 0.9414,
+      "step": 14015
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8968678712844849,
+      "learning_rate": 1.951081623671911e-06,
+      "loss": 0.9576,
+      "step": 14016
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6864975690841675,
+      "learning_rate": 1.9499794158355658e-06,
+      "loss": 0.9485,
+      "step": 14017
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7293339967727661,
+      "learning_rate": 1.948877485782942e-06,
+      "loss": 0.9023,
+      "step": 14018
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.652435541152954,
+      "learning_rate": 1.947775833552069e-06,
+      "loss": 0.8574,
+      "step": 14019
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0346343517303467,
+      "learning_rate": 1.946674459180955e-06,
+      "loss": 0.5031,
+      "step": 14020
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6752530336380005,
+      "learning_rate": 1.9455733627076136e-06,
+      "loss": 0.9135,
+      "step": 14021
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.02181077003479,
+      "learning_rate": 1.944472544170033e-06,
+      "loss": 0.9037,
+      "step": 14022
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.712759256362915,
+      "learning_rate": 1.9433720036062055e-06,
+      "loss": 0.9107,
+      "step": 14023
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5891083478927612,
+      "learning_rate": 1.9422717410541016e-06,
+      "loss": 0.8673,
+      "step": 14024
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7310612201690674,
+      "learning_rate": 1.941171756551695e-06,
+      "loss": 0.8613,
+      "step": 14025
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8287407159805298,
+      "learning_rate": 1.9400720501369363e-06,
+      "loss": 0.8781,
+      "step": 14026
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7315726280212402,
+      "learning_rate": 1.938972621847778e-06,
+      "loss": 0.8755,
+      "step": 14027
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8258755207061768,
+      "learning_rate": 1.937873471722158e-06,
+      "loss": 0.8767,
+      "step": 14028
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8173223733901978,
+      "learning_rate": 1.9367745997980026e-06,
+      "loss": 0.9064,
+      "step": 14029
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7682009935379028,
+      "learning_rate": 1.935676006113234e-06,
+      "loss": 0.8903,
+      "step": 14030
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8015422821044922,
+      "learning_rate": 1.9345776907057566e-06,
+      "loss": 0.9974,
+      "step": 14031
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5687092542648315,
+      "learning_rate": 1.933479653613476e-06,
+      "loss": 0.8866,
+      "step": 14032
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.937098503112793,
+      "learning_rate": 1.932381894874278e-06,
+      "loss": 0.9031,
+      "step": 14033
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8051369190216064,
+      "learning_rate": 1.9312844145260435e-06,
+      "loss": 0.9186,
+      "step": 14034
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.686700463294983,
+      "learning_rate": 1.930187212606646e-06,
+      "loss": 0.9152,
+      "step": 14035
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7500659227371216,
+      "learning_rate": 1.9290902891539475e-06,
+      "loss": 0.8877,
+      "step": 14036
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8407310247421265,
+      "learning_rate": 1.927993644205796e-06,
+      "loss": 0.9643,
+      "step": 14037
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7866275310516357,
+      "learning_rate": 1.9268972778000373e-06,
+      "loss": 0.9623,
+      "step": 14038
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7575585842132568,
+      "learning_rate": 1.9258011899744998e-06,
+      "loss": 0.9314,
+      "step": 14039
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8576430082321167,
+      "learning_rate": 1.924705380767011e-06,
+      "loss": 0.9341,
+      "step": 14040
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7025312185287476,
+      "learning_rate": 1.923609850215381e-06,
+      "loss": 0.9,
+      "step": 14041
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8535430431365967,
+      "learning_rate": 1.9225145983574166e-06,
+      "loss": 0.9393,
+      "step": 14042
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1030341386795044,
+      "learning_rate": 1.921419625230907e-06,
+      "loss": 0.5638,
+      "step": 14043
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7576714754104614,
+      "learning_rate": 1.920324930873639e-06,
+      "loss": 0.921,
+      "step": 14044
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.692887544631958,
+      "learning_rate": 1.9192305153233913e-06,
+      "loss": 0.9261,
+      "step": 14045
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8226488828659058,
+      "learning_rate": 1.918136378617923e-06,
+      "loss": 0.9028,
+      "step": 14046
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9208482503890991,
+      "learning_rate": 1.917042520794995e-06,
+      "loss": 0.8834,
+      "step": 14047
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.788041114807129,
+      "learning_rate": 1.9159489418923493e-06,
+      "loss": 0.8817,
+      "step": 14048
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.809095025062561,
+      "learning_rate": 1.914855641947725e-06,
+      "loss": 0.8783,
+      "step": 14049
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9852890372276306,
+      "learning_rate": 1.913762620998846e-06,
+      "loss": 0.5117,
+      "step": 14050
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8035486936569214,
+      "learning_rate": 1.912669879083432e-06,
+      "loss": 0.8581,
+      "step": 14051
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.917277455329895,
+      "learning_rate": 1.9115774162391876e-06,
+      "loss": 0.876,
+      "step": 14052
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.809247374534607,
+      "learning_rate": 1.910485232503816e-06,
+      "loss": 0.952,
+      "step": 14053
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9863134622573853,
+      "learning_rate": 1.909393327914998e-06,
+      "loss": 0.9518,
+      "step": 14054
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6907293796539307,
+      "learning_rate": 1.9083017025104166e-06,
+      "loss": 0.8399,
+      "step": 14055
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7695196866989136,
+      "learning_rate": 1.9072103563277423e-06,
+      "loss": 0.9217,
+      "step": 14056
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7960482835769653,
+      "learning_rate": 1.906119289404631e-06,
+      "loss": 0.9491,
+      "step": 14057
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8133835792541504,
+      "learning_rate": 1.9050285017787351e-06,
+      "loss": 0.9302,
+      "step": 14058
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8223031759262085,
+      "learning_rate": 1.9039379934876912e-06,
+      "loss": 0.8936,
+      "step": 14059
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7321873903274536,
+      "learning_rate": 1.9028477645691334e-06,
+      "loss": 0.9289,
+      "step": 14060
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7210311889648438,
+      "learning_rate": 1.9017578150606786e-06,
+      "loss": 0.9112,
+      "step": 14061
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.041524887084961,
+      "learning_rate": 1.900668144999943e-06,
+      "loss": 0.8705,
+      "step": 14062
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0023767948150635,
+      "learning_rate": 1.8995787544245225e-06,
+      "loss": 0.9344,
+      "step": 14063
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8140009641647339,
+      "learning_rate": 1.8984896433720147e-06,
+      "loss": 0.9044,
+      "step": 14064
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.752633810043335,
+      "learning_rate": 1.8974008118799947e-06,
+      "loss": 0.8332,
+      "step": 14065
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.741737723350525,
+      "learning_rate": 1.8963122599860428e-06,
+      "loss": 0.9709,
+      "step": 14066
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1367217302322388,
+      "learning_rate": 1.8952239877277145e-06,
+      "loss": 0.5821,
+      "step": 14067
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7521618604660034,
+      "learning_rate": 1.8941359951425675e-06,
+      "loss": 0.9497,
+      "step": 14068
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8025600910186768,
+      "learning_rate": 1.8930482822681473e-06,
+      "loss": 0.9244,
+      "step": 14069
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7236820459365845,
+      "learning_rate": 1.8919608491419816e-06,
+      "loss": 0.8974,
+      "step": 14070
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.762858271598816,
+      "learning_rate": 1.8908736958016006e-06,
+      "loss": 0.9244,
+      "step": 14071
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8352046012878418,
+      "learning_rate": 1.8897868222845139e-06,
+      "loss": 0.9184,
+      "step": 14072
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9757412075996399,
+      "learning_rate": 1.8887002286282318e-06,
+      "loss": 0.5177,
+      "step": 14073
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8258934020996094,
+      "learning_rate": 1.8876139148702444e-06,
+      "loss": 0.9007,
+      "step": 14074
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.670674443244934,
+      "learning_rate": 1.8865278810480425e-06,
+      "loss": 0.927,
+      "step": 14075
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6569859981536865,
+      "learning_rate": 1.8854421271990964e-06,
+      "loss": 0.84,
+      "step": 14076
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7142783403396606,
+      "learning_rate": 1.884356653360878e-06,
+      "loss": 0.954,
+      "step": 14077
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6480296850204468,
+      "learning_rate": 1.883271459570839e-06,
+      "loss": 0.8044,
+      "step": 14078
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.731122374534607,
+      "learning_rate": 1.8821865458664291e-06,
+      "loss": 0.928,
+      "step": 14079
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.768945574760437,
+      "learning_rate": 1.8811019122850872e-06,
+      "loss": 0.917,
+      "step": 14080
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8149460554122925,
+      "learning_rate": 1.8800175588642366e-06,
+      "loss": 0.9792,
+      "step": 14081
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.107393741607666,
+      "learning_rate": 1.8789334856413e-06,
+      "loss": 0.9472,
+      "step": 14082
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9868326783180237,
+      "learning_rate": 1.8778496926536815e-06,
+      "loss": 0.5967,
+      "step": 14083
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7033144235610962,
+      "learning_rate": 1.8767661799387848e-06,
+      "loss": 0.8628,
+      "step": 14084
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6941184997558594,
+      "learning_rate": 1.8756829475339922e-06,
+      "loss": 0.8656,
+      "step": 14085
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8487378358840942,
+      "learning_rate": 1.8745999954766903e-06,
+      "loss": 1.0283,
+      "step": 14086
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7418134212493896,
+      "learning_rate": 1.8735173238042415e-06,
+      "loss": 0.8322,
+      "step": 14087
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5997120141983032,
+      "learning_rate": 1.8724349325540137e-06,
+      "loss": 0.8643,
+      "step": 14088
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7328933477401733,
+      "learning_rate": 1.8713528217633491e-06,
+      "loss": 0.9392,
+      "step": 14089
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7457698583602905,
+      "learning_rate": 1.8702709914695949e-06,
+      "loss": 0.8546,
+      "step": 14090
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.752020001411438,
+      "learning_rate": 1.8691894417100764e-06,
+      "loss": 0.8275,
+      "step": 14091
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8976435661315918,
+      "learning_rate": 1.8681081725221185e-06,
+      "loss": 0.8771,
+      "step": 14092
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7588292360305786,
+      "learning_rate": 1.8670271839430343e-06,
+      "loss": 0.8791,
+      "step": 14093
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7215102910995483,
+      "learning_rate": 1.865946476010121e-06,
+      "loss": 0.8491,
+      "step": 14094
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.014331340789795,
+      "learning_rate": 1.8648660487606752e-06,
+      "loss": 0.5498,
+      "step": 14095
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8594752550125122,
+      "learning_rate": 1.863785902231976e-06,
+      "loss": 0.8232,
+      "step": 14096
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8511710166931152,
+      "learning_rate": 1.8627060364612993e-06,
+      "loss": 0.8858,
+      "step": 14097
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0701453685760498,
+      "learning_rate": 1.8616264514859051e-06,
+      "loss": 0.5091,
+      "step": 14098
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9040848016738892,
+      "learning_rate": 1.8605471473430503e-06,
+      "loss": 0.8619,
+      "step": 14099
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7075958251953125,
+      "learning_rate": 1.8594681240699708e-06,
+      "loss": 0.8836,
+      "step": 14100
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.697514295578003,
+      "learning_rate": 1.8583893817039134e-06,
+      "loss": 0.8698,
+      "step": 14101
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.950579047203064,
+      "learning_rate": 1.8573109202820927e-06,
+      "loss": 0.9734,
+      "step": 14102
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.675250768661499,
+      "learning_rate": 1.856232739841729e-06,
+      "loss": 0.8716,
+      "step": 14103
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7621372938156128,
+      "learning_rate": 1.8551548404200215e-06,
+      "loss": 0.9217,
+      "step": 14104
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8805630207061768,
+      "learning_rate": 1.8540772220541725e-06,
+      "loss": 0.9084,
+      "step": 14105
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8633058071136475,
+      "learning_rate": 1.8529998847813602e-06,
+      "loss": 0.9215,
+      "step": 14106
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.76390540599823,
+      "learning_rate": 1.8519228286387668e-06,
+      "loss": 0.9464,
+      "step": 14107
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8566062450408936,
+      "learning_rate": 1.8508460536635542e-06,
+      "loss": 0.9695,
+      "step": 14108
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6882922649383545,
+      "learning_rate": 1.84976955989288e-06,
+      "loss": 0.8501,
+      "step": 14109
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6523946523666382,
+      "learning_rate": 1.8486933473638945e-06,
+      "loss": 0.8227,
+      "step": 14110
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6837847232818604,
+      "learning_rate": 1.8476174161137283e-06,
+      "loss": 0.8144,
+      "step": 14111
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.713645577430725,
+      "learning_rate": 1.846541766179516e-06,
+      "loss": 0.922,
+      "step": 14112
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6728363037109375,
+      "learning_rate": 1.8454663975983677e-06,
+      "loss": 0.9482,
+      "step": 14113
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7649805545806885,
+      "learning_rate": 1.8443913104073984e-06,
+      "loss": 0.8473,
+      "step": 14114
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.617882490158081,
+      "learning_rate": 1.8433165046437018e-06,
+      "loss": 0.8963,
+      "step": 14115
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.833274006843567,
+      "learning_rate": 1.8422419803443692e-06,
+      "loss": 0.938,
+      "step": 14116
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1394277811050415,
+      "learning_rate": 1.8411677375464754e-06,
+      "loss": 0.5762,
+      "step": 14117
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7469112873077393,
+      "learning_rate": 1.840093776287095e-06,
+      "loss": 0.9194,
+      "step": 14118
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9957841634750366,
+      "learning_rate": 1.8390200966032822e-06,
+      "loss": 0.5626,
+      "step": 14119
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7444697618484497,
+      "learning_rate": 1.8379466985320915e-06,
+      "loss": 0.8521,
+      "step": 14120
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8393453359603882,
+      "learning_rate": 1.8368735821105588e-06,
+      "loss": 0.9045,
+      "step": 14121
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8354759216308594,
+      "learning_rate": 1.8358007473757145e-06,
+      "loss": 0.9408,
+      "step": 14122
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.677212119102478,
+      "learning_rate": 1.8347281943645846e-06,
+      "loss": 0.8847,
+      "step": 14123
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7355605363845825,
+      "learning_rate": 1.8336559231141726e-06,
+      "loss": 0.8906,
+      "step": 14124
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0447041988372803,
+      "learning_rate": 1.8325839336614858e-06,
+      "loss": 0.5264,
+      "step": 14125
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7561938762664795,
+      "learning_rate": 1.8315122260435092e-06,
+      "loss": 0.8997,
+      "step": 14126
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7475374937057495,
+      "learning_rate": 1.8304408002972318e-06,
+      "loss": 0.9372,
+      "step": 14127
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8813424110412598,
+      "learning_rate": 1.8293696564596186e-06,
+      "loss": 0.9757,
+      "step": 14128
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6131538152694702,
+      "learning_rate": 1.8282987945676368e-06,
+      "loss": 0.876,
+      "step": 14129
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7284905910491943,
+      "learning_rate": 1.8272282146582354e-06,
+      "loss": 0.8574,
+      "step": 14130
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8039865493774414,
+      "learning_rate": 1.8261579167683597e-06,
+      "loss": 0.9642,
+      "step": 14131
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7227176427841187,
+      "learning_rate": 1.8250879009349398e-06,
+      "loss": 0.9131,
+      "step": 14132
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.846163034439087,
+      "learning_rate": 1.824018167194901e-06,
+      "loss": 0.9538,
+      "step": 14133
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.778264045715332,
+      "learning_rate": 1.8229487155851589e-06,
+      "loss": 0.9239,
+      "step": 14134
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0351076126098633,
+      "learning_rate": 1.821879546142613e-06,
+      "loss": 0.5484,
+      "step": 14135
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.67878258228302,
+      "learning_rate": 1.8208106589041608e-06,
+      "loss": 0.8323,
+      "step": 14136
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.779456377029419,
+      "learning_rate": 1.8197420539066834e-06,
+      "loss": 0.8745,
+      "step": 14137
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8889758586883545,
+      "learning_rate": 1.8186737311870596e-06,
+      "loss": 0.9385,
+      "step": 14138
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7490098476409912,
+      "learning_rate": 1.8176056907821482e-06,
+      "loss": 0.8726,
+      "step": 14139
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9873440265655518,
+      "learning_rate": 1.8165379327288113e-06,
+      "loss": 0.5469,
+      "step": 14140
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7836222648620605,
+      "learning_rate": 1.8154704570638882e-06,
+      "loss": 0.9273,
+      "step": 14141
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.867698073387146,
+      "learning_rate": 1.8144032638242192e-06,
+      "loss": 0.8756,
+      "step": 14142
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6066627502441406,
+      "learning_rate": 1.8133363530466253e-06,
+      "loss": 0.9629,
+      "step": 14143
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7265434265136719,
+      "learning_rate": 1.8122697247679288e-06,
+      "loss": 0.9542,
+      "step": 14144
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7514406442642212,
+      "learning_rate": 1.8112033790249294e-06,
+      "loss": 0.9315,
+      "step": 14145
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7310880422592163,
+      "learning_rate": 1.8101373158544267e-06,
+      "loss": 0.8511,
+      "step": 14146
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8452088832855225,
+      "learning_rate": 1.809071535293211e-06,
+      "loss": 0.8755,
+      "step": 14147
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7872775793075562,
+      "learning_rate": 1.808006037378053e-06,
+      "loss": 0.8991,
+      "step": 14148
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7407293319702148,
+      "learning_rate": 1.8069408221457264e-06,
+      "loss": 0.898,
+      "step": 14149
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7654609680175781,
+      "learning_rate": 1.805875889632982e-06,
+      "loss": 0.8645,
+      "step": 14150
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.651309847831726,
+      "learning_rate": 1.804811239876575e-06,
+      "loss": 0.9322,
+      "step": 14151
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.718447208404541,
+      "learning_rate": 1.8037468729132368e-06,
+      "loss": 0.8514,
+      "step": 14152
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7021558284759521,
+      "learning_rate": 1.8026827887797016e-06,
+      "loss": 0.9079,
+      "step": 14153
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7367775440216064,
+      "learning_rate": 1.8016189875126821e-06,
+      "loss": 0.9014,
+      "step": 14154
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7291998863220215,
+      "learning_rate": 1.8005554691488924e-06,
+      "loss": 0.892,
+      "step": 14155
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8266780376434326,
+      "learning_rate": 1.7994922337250276e-06,
+      "loss": 1.017,
+      "step": 14156
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6764843463897705,
+      "learning_rate": 1.7984292812777805e-06,
+      "loss": 0.8096,
+      "step": 14157
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7339614629745483,
+      "learning_rate": 1.797366611843826e-06,
+      "loss": 0.8607,
+      "step": 14158
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0795891284942627,
+      "learning_rate": 1.7963042254598362e-06,
+      "loss": 0.5529,
+      "step": 14159
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7569904327392578,
+      "learning_rate": 1.795242122162475e-06,
+      "loss": 0.9161,
+      "step": 14160
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1390608549118042,
+      "learning_rate": 1.7941803019883864e-06,
+      "loss": 0.6185,
+      "step": 14161
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9225541353225708,
+      "learning_rate": 1.7931187649742155e-06,
+      "loss": 0.8756,
+      "step": 14162
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7774367332458496,
+      "learning_rate": 1.7920575111565896e-06,
+      "loss": 0.8882,
+      "step": 14163
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7486282587051392,
+      "learning_rate": 1.790996540572133e-06,
+      "loss": 0.8801,
+      "step": 14164
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7437165975570679,
+      "learning_rate": 1.7899358532574518e-06,
+      "loss": 0.914,
+      "step": 14165
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7274329662322998,
+      "learning_rate": 1.788875449249151e-06,
+      "loss": 0.898,
+      "step": 14166
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7874555587768555,
+      "learning_rate": 1.7878153285838206e-06,
+      "loss": 0.8526,
+      "step": 14167
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7938071489334106,
+      "learning_rate": 1.7867554912980478e-06,
+      "loss": 0.8716,
+      "step": 14168
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.744019627571106,
+      "learning_rate": 1.7856959374283967e-06,
+      "loss": 0.8601,
+      "step": 14169
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8557969331741333,
+      "learning_rate": 1.7846366670114345e-06,
+      "loss": 0.9045,
+      "step": 14170
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9055627584457397,
+      "learning_rate": 1.7835776800837113e-06,
+      "loss": 0.9082,
+      "step": 14171
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5974041223526,
+      "learning_rate": 1.782518976681773e-06,
+      "loss": 0.9352,
+      "step": 14172
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7303529977798462,
+      "learning_rate": 1.7814605568421473e-06,
+      "loss": 0.8609,
+      "step": 14173
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9245283603668213,
+      "learning_rate": 1.7804024206013625e-06,
+      "loss": 0.927,
+      "step": 14174
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6998205184936523,
+      "learning_rate": 1.7793445679959276e-06,
+      "loss": 0.9598,
+      "step": 14175
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7247503995895386,
+      "learning_rate": 1.7782869990623475e-06,
+      "loss": 0.8842,
+      "step": 14176
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0022276639938354,
+      "learning_rate": 1.7772297138371197e-06,
+      "loss": 0.4701,
+      "step": 14177
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.750462293624878,
+      "learning_rate": 1.776172712356723e-06,
+      "loss": 0.8759,
+      "step": 14178
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.079314708709717,
+      "learning_rate": 1.7751159946576357e-06,
+      "loss": 0.9231,
+      "step": 14179
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8474836349487305,
+      "learning_rate": 1.7740595607763177e-06,
+      "loss": 0.902,
+      "step": 14180
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7702025175094604,
+      "learning_rate": 1.7730034107492278e-06,
+      "loss": 0.8724,
+      "step": 14181
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7252347469329834,
+      "learning_rate": 1.7719475446128076e-06,
+      "loss": 0.8878,
+      "step": 14182
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8300334215164185,
+      "learning_rate": 1.770891962403496e-06,
+      "loss": 0.9086,
+      "step": 14183
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7947404384613037,
+      "learning_rate": 1.7698366641577124e-06,
+      "loss": 0.8064,
+      "step": 14184
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6846437454223633,
+      "learning_rate": 1.7687816499118781e-06,
+      "loss": 0.8231,
+      "step": 14185
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7311357259750366,
+      "learning_rate": 1.7677269197023938e-06,
+      "loss": 0.9341,
+      "step": 14186
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7213209867477417,
+      "learning_rate": 1.7666724735656583e-06,
+      "loss": 0.8893,
+      "step": 14187
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.854055643081665,
+      "learning_rate": 1.7656183115380577e-06,
+      "loss": 0.9033,
+      "step": 14188
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7185779809951782,
+      "learning_rate": 1.7645644336559665e-06,
+      "loss": 0.8737,
+      "step": 14189
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.839493751525879,
+      "learning_rate": 1.7635108399557532e-06,
+      "loss": 0.9342,
+      "step": 14190
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0113407373428345,
+      "learning_rate": 1.7624575304737713e-06,
+      "loss": 0.5621,
+      "step": 14191
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6885428428649902,
+      "learning_rate": 1.7614045052463724e-06,
+      "loss": 0.8431,
+      "step": 14192
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.880582332611084,
+      "learning_rate": 1.7603517643098866e-06,
+      "loss": 1.0317,
+      "step": 14193
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6084463596343994,
+      "learning_rate": 1.7592993077006482e-06,
+      "loss": 0.9631,
+      "step": 14194
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.681396484375,
+      "learning_rate": 1.758247135454969e-06,
+      "loss": 0.8958,
+      "step": 14195
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.828576922416687,
+      "learning_rate": 1.7571952476091604e-06,
+      "loss": 0.9367,
+      "step": 14196
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6702998876571655,
+      "learning_rate": 1.756143644199516e-06,
+      "loss": 0.9021,
+      "step": 14197
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7651079893112183,
+      "learning_rate": 1.7550923252623299e-06,
+      "loss": 0.9085,
+      "step": 14198
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7151782512664795,
+      "learning_rate": 1.7540412908338723e-06,
+      "loss": 0.9235,
+      "step": 14199
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7548067569732666,
+      "learning_rate": 1.7529905409504167e-06,
+      "loss": 0.8431,
+      "step": 14200
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7655396461486816,
+      "learning_rate": 1.751940075648223e-06,
+      "loss": 0.8821,
+      "step": 14201
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8697059154510498,
+      "learning_rate": 1.7508898949635345e-06,
+      "loss": 0.8703,
+      "step": 14202
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8848613500595093,
+      "learning_rate": 1.7498399989325943e-06,
+      "loss": 0.9447,
+      "step": 14203
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8009474277496338,
+      "learning_rate": 1.748790387591629e-06,
+      "loss": 0.8516,
+      "step": 14204
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5711179971694946,
+      "learning_rate": 1.7477410609768597e-06,
+      "loss": 0.9435,
+      "step": 14205
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.739617109298706,
+      "learning_rate": 1.746692019124493e-06,
+      "loss": 0.9154,
+      "step": 14206
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8294473886489868,
+      "learning_rate": 1.745643262070732e-06,
+      "loss": 0.9235,
+      "step": 14207
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6991270780563354,
+      "learning_rate": 1.7445947898517624e-06,
+      "loss": 0.8487,
+      "step": 14208
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8120815753936768,
+      "learning_rate": 1.7435466025037684e-06,
+      "loss": 0.8244,
+      "step": 14209
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.776281476020813,
+      "learning_rate": 1.7424987000629146e-06,
+      "loss": 0.8771,
+      "step": 14210
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6794159412384033,
+      "learning_rate": 1.7414510825653674e-06,
+      "loss": 0.915,
+      "step": 14211
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6037272214889526,
+      "learning_rate": 1.7404037500472714e-06,
+      "loss": 0.8254,
+      "step": 14212
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8331823348999023,
+      "learning_rate": 1.73935670254477e-06,
+      "loss": 0.9082,
+      "step": 14213
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7944010496139526,
+      "learning_rate": 1.7383099400939963e-06,
+      "loss": 0.8992,
+      "step": 14214
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8769935369491577,
+      "learning_rate": 1.7372634627310647e-06,
+      "loss": 0.8835,
+      "step": 14215
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7185628414154053,
+      "learning_rate": 1.7362172704920933e-06,
+      "loss": 0.921,
+      "step": 14216
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.995455801486969,
+      "learning_rate": 1.7351713634131773e-06,
+      "loss": 0.4698,
+      "step": 14217
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8464833498001099,
+      "learning_rate": 1.7341257415304137e-06,
+      "loss": 0.8171,
+      "step": 14218
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6105918884277344,
+      "learning_rate": 1.7330804048798777e-06,
+      "loss": 0.8858,
+      "step": 14219
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.681216835975647,
+      "learning_rate": 1.7320353534976474e-06,
+      "loss": 0.9911,
+      "step": 14220
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7502812147140503,
+      "learning_rate": 1.7309905874197786e-06,
+      "loss": 0.9489,
+      "step": 14221
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6263118982315063,
+      "learning_rate": 1.7299461066823286e-06,
+      "loss": 0.9049,
+      "step": 14222
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7487337589263916,
+      "learning_rate": 1.7289019113213346e-06,
+      "loss": 0.8828,
+      "step": 14223
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6987881660461426,
+      "learning_rate": 1.7278580013728307e-06,
+      "loss": 0.9132,
+      "step": 14224
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.771954894065857,
+      "learning_rate": 1.7268143768728429e-06,
+      "loss": 0.8872,
+      "step": 14225
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6657689809799194,
+      "learning_rate": 1.725771037857379e-06,
+      "loss": 0.9551,
+      "step": 14226
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.898425817489624,
+      "learning_rate": 1.7247279843624455e-06,
+      "loss": 0.8439,
+      "step": 14227
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7319918870925903,
+      "learning_rate": 1.7236852164240292e-06,
+      "loss": 0.9171,
+      "step": 14228
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.794089436531067,
+      "learning_rate": 1.7226427340781215e-06,
+      "loss": 0.8083,
+      "step": 14229
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8060024976730347,
+      "learning_rate": 1.7216005373606881e-06,
+      "loss": 0.9427,
+      "step": 14230
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9346424341201782,
+      "learning_rate": 1.7205586263076978e-06,
+      "loss": 0.8922,
+      "step": 14231
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7874208688735962,
+      "learning_rate": 1.7195170009550966e-06,
+      "loss": 0.9424,
+      "step": 14232
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9846954345703125,
+      "learning_rate": 1.7184756613388376e-06,
+      "loss": 0.525,
+      "step": 14233
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7438921928405762,
+      "learning_rate": 1.7174346074948478e-06,
+      "loss": 0.924,
+      "step": 14234
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.921628475189209,
+      "learning_rate": 1.7163938394590563e-06,
+      "loss": 0.9272,
+      "step": 14235
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.915191411972046,
+      "learning_rate": 1.7153533572673708e-06,
+      "loss": 0.9535,
+      "step": 14236
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7039462327957153,
+      "learning_rate": 1.7143131609557017e-06,
+      "loss": 0.92,
+      "step": 14237
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.993430733680725,
+      "learning_rate": 1.713273250559938e-06,
+      "loss": 0.8601,
+      "step": 14238
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6365638971328735,
+      "learning_rate": 1.7122336261159689e-06,
+      "loss": 0.935,
+      "step": 14239
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8261125087738037,
+      "learning_rate": 1.7111942876596633e-06,
+      "loss": 0.9117,
+      "step": 14240
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9686861038208008,
+      "learning_rate": 1.7101552352268901e-06,
+      "loss": 0.9045,
+      "step": 14241
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7271994352340698,
+      "learning_rate": 1.7091164688535044e-06,
+      "loss": 0.9066,
+      "step": 14242
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8059329986572266,
+      "learning_rate": 1.7080779885753473e-06,
+      "loss": 0.8867,
+      "step": 14243
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.871031641960144,
+      "learning_rate": 1.707039794428259e-06,
+      "loss": 0.7928,
+      "step": 14244
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6465849876403809,
+      "learning_rate": 1.7060018864480598e-06,
+      "loss": 0.9053,
+      "step": 14245
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7444630861282349,
+      "learning_rate": 1.7049642646705688e-06,
+      "loss": 0.8862,
+      "step": 14246
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.780835747718811,
+      "learning_rate": 1.7039269291315885e-06,
+      "loss": 0.8361,
+      "step": 14247
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8187954425811768,
+      "learning_rate": 1.702889879866917e-06,
+      "loss": 0.8388,
+      "step": 14248
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.773315668106079,
+      "learning_rate": 1.7018531169123364e-06,
+      "loss": 0.8386,
+      "step": 14249
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8218978643417358,
+      "learning_rate": 1.7008166403036286e-06,
+      "loss": 0.9175,
+      "step": 14250
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4901678562164307,
+      "learning_rate": 1.6997804500765513e-06,
+      "loss": 0.7915,
+      "step": 14251
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6536728143692017,
+      "learning_rate": 1.6987445462668695e-06,
+      "loss": 0.8286,
+      "step": 14252
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7746291160583496,
+      "learning_rate": 1.697708928910321e-06,
+      "loss": 0.9312,
+      "step": 14253
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7446489334106445,
+      "learning_rate": 1.6966735980426453e-06,
+      "loss": 0.7996,
+      "step": 14254
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7436262369155884,
+      "learning_rate": 1.6956385536995735e-06,
+      "loss": 0.8566,
+      "step": 14255
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8321533203125,
+      "learning_rate": 1.6946037959168138e-06,
+      "loss": 0.861,
+      "step": 14256
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6946762800216675,
+      "learning_rate": 1.69356932473008e-06,
+      "loss": 0.9216,
+      "step": 14257
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8879473209381104,
+      "learning_rate": 1.6925351401750634e-06,
+      "loss": 0.8533,
+      "step": 14258
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8471488952636719,
+      "learning_rate": 1.6915012422874555e-06,
+      "loss": 0.9009,
+      "step": 14259
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0366759300231934,
+      "learning_rate": 1.6904676311029289e-06,
+      "loss": 0.9227,
+      "step": 14260
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.857603907585144,
+      "learning_rate": 1.689434306657154e-06,
+      "loss": 0.848,
+      "step": 14261
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6743396520614624,
+      "learning_rate": 1.6884012689857854e-06,
+      "loss": 0.8402,
+      "step": 14262
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8774621486663818,
+      "learning_rate": 1.6873685181244726e-06,
+      "loss": 0.944,
+      "step": 14263
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6760950088500977,
+      "learning_rate": 1.6863360541088503e-06,
+      "loss": 0.8466,
+      "step": 14264
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7994827032089233,
+      "learning_rate": 1.6853038769745466e-06,
+      "loss": 0.9446,
+      "step": 14265
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.735919713973999,
+      "learning_rate": 1.6842719867571832e-06,
+      "loss": 0.8534,
+      "step": 14266
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7978187799453735,
+      "learning_rate": 1.6832403834923617e-06,
+      "loss": 0.9337,
+      "step": 14267
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7791756391525269,
+      "learning_rate": 1.6822090672156854e-06,
+      "loss": 0.872,
+      "step": 14268
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5988049507141113,
+      "learning_rate": 1.6811780379627374e-06,
+      "loss": 0.8608,
+      "step": 14269
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.950165033340454,
+      "learning_rate": 1.6801472957690989e-06,
+      "loss": 0.8776,
+      "step": 14270
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.726991891860962,
+      "learning_rate": 1.6791168406703351e-06,
+      "loss": 0.9663,
+      "step": 14271
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9134963750839233,
+      "learning_rate": 1.6780866727020074e-06,
+      "loss": 0.9385,
+      "step": 14272
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8603650331497192,
+      "learning_rate": 1.6770567918996604e-06,
+      "loss": 0.965,
+      "step": 14273
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9431560039520264,
+      "learning_rate": 1.6760271982988363e-06,
+      "loss": 0.9282,
+      "step": 14274
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7547402381896973,
+      "learning_rate": 1.6749978919350595e-06,
+      "loss": 0.8584,
+      "step": 14275
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8730500936508179,
+      "learning_rate": 1.673968872843853e-06,
+      "loss": 0.9098,
+      "step": 14276
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7788445949554443,
+      "learning_rate": 1.6729401410607205e-06,
+      "loss": 0.8579,
+      "step": 14277
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0393218994140625,
+      "learning_rate": 1.6719116966211624e-06,
+      "loss": 0.9371,
+      "step": 14278
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7952711582183838,
+      "learning_rate": 1.6708835395606704e-06,
+      "loss": 0.9065,
+      "step": 14279
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6670825481414795,
+      "learning_rate": 1.6698556699147195e-06,
+      "loss": 0.9141,
+      "step": 14280
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9021333456039429,
+      "learning_rate": 1.6688280877187824e-06,
+      "loss": 0.9704,
+      "step": 14281
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9523084163665771,
+      "learning_rate": 1.667800793008313e-06,
+      "loss": 0.9238,
+      "step": 14282
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7435333728790283,
+      "learning_rate": 1.6667737858187649e-06,
+      "loss": 0.9254,
+      "step": 14283
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7668544054031372,
+      "learning_rate": 1.6657470661855746e-06,
+      "loss": 0.8475,
+      "step": 14284
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7622661590576172,
+      "learning_rate": 1.6647206341441735e-06,
+      "loss": 0.9824,
+      "step": 14285
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7271367311477661,
+      "learning_rate": 1.6636944897299777e-06,
+      "loss": 0.9462,
+      "step": 14286
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7134380340576172,
+      "learning_rate": 1.6626686329784003e-06,
+      "loss": 0.8616,
+      "step": 14287
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1179444789886475,
+      "learning_rate": 1.661643063924837e-06,
+      "loss": 0.5259,
+      "step": 14288
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0577850341796875,
+      "learning_rate": 1.6606177826046822e-06,
+      "loss": 0.5911,
+      "step": 14289
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6966627836227417,
+      "learning_rate": 1.6595927890533103e-06,
+      "loss": 0.9122,
+      "step": 14290
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8448669910430908,
+      "learning_rate": 1.6585680833060923e-06,
+      "loss": 0.9101,
+      "step": 14291
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6884130239486694,
+      "learning_rate": 1.6575436653983923e-06,
+      "loss": 1.0163,
+      "step": 14292
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8530460596084595,
+      "learning_rate": 1.656519535365554e-06,
+      "loss": 0.8685,
+      "step": 14293
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7355821132659912,
+      "learning_rate": 1.6554956932429223e-06,
+      "loss": 0.8415,
+      "step": 14294
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6747983694076538,
+      "learning_rate": 1.6544721390658213e-06,
+      "loss": 0.8036,
+      "step": 14295
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7244642972946167,
+      "learning_rate": 1.6534488728695786e-06,
+      "loss": 0.8406,
+      "step": 14296
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7265794277191162,
+      "learning_rate": 1.6524258946894966e-06,
+      "loss": 0.8608,
+      "step": 14297
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7757856845855713,
+      "learning_rate": 1.6514032045608819e-06,
+      "loss": 1.0044,
+      "step": 14298
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.68044114112854,
+      "learning_rate": 1.650380802519017e-06,
+      "loss": 0.8888,
+      "step": 14299
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7359893321990967,
+      "learning_rate": 1.6493586885991908e-06,
+      "loss": 0.976,
+      "step": 14300
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6907399892807007,
+      "learning_rate": 1.648336862836668e-06,
+      "loss": 0.9145,
+      "step": 14301
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8975483179092407,
+      "learning_rate": 1.647315325266714e-06,
+      "loss": 0.8715,
+      "step": 14302
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.63186514377594,
+      "learning_rate": 1.6462940759245716e-06,
+      "loss": 0.8925,
+      "step": 14303
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7615134716033936,
+      "learning_rate": 1.6452731148454893e-06,
+      "loss": 0.9618,
+      "step": 14304
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7152031660079956,
+      "learning_rate": 1.644252442064691e-06,
+      "loss": 1.0029,
+      "step": 14305
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.65349543094635,
+      "learning_rate": 1.643232057617402e-06,
+      "loss": 0.9047,
+      "step": 14306
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9353394508361816,
+      "learning_rate": 1.6422119615388288e-06,
+      "loss": 0.8764,
+      "step": 14307
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9299299716949463,
+      "learning_rate": 1.641192153864175e-06,
+      "loss": 0.4796,
+      "step": 14308
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.756333827972412,
+      "learning_rate": 1.6401726346286317e-06,
+      "loss": 0.8628,
+      "step": 14309
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.831635594367981,
+      "learning_rate": 1.6391534038673774e-06,
+      "loss": 0.8596,
+      "step": 14310
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5277256965637207,
+      "learning_rate": 1.6381344616155859e-06,
+      "loss": 0.8635,
+      "step": 14311
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7213791608810425,
+      "learning_rate": 1.6371158079084136e-06,
+      "loss": 0.8277,
+      "step": 14312
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.756650686264038,
+      "learning_rate": 1.6360974427810172e-06,
+      "loss": 0.8685,
+      "step": 14313
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7933175563812256,
+      "learning_rate": 1.6350793662685305e-06,
+      "loss": 0.9306,
+      "step": 14314
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8213019371032715,
+      "learning_rate": 1.634061578406092e-06,
+      "loss": 0.8476,
+      "step": 14315
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6642775535583496,
+      "learning_rate": 1.633044079228817e-06,
+      "loss": 0.8428,
+      "step": 14316
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.776218056678772,
+      "learning_rate": 1.6320268687718199e-06,
+      "loss": 0.8844,
+      "step": 14317
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7835209369659424,
+      "learning_rate": 1.631009947070199e-06,
+      "loss": 0.8475,
+      "step": 14318
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.727022409439087,
+      "learning_rate": 1.6299933141590473e-06,
+      "loss": 0.8876,
+      "step": 14319
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9957889318466187,
+      "learning_rate": 1.628976970073447e-06,
+      "loss": 0.8827,
+      "step": 14320
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7999436855316162,
+      "learning_rate": 1.6279609148484666e-06,
+      "loss": 0.9053,
+      "step": 14321
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8848061561584473,
+      "learning_rate": 1.6269451485191701e-06,
+      "loss": 0.8885,
+      "step": 14322
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8123704195022583,
+      "learning_rate": 1.6259296711206051e-06,
+      "loss": 0.8348,
+      "step": 14323
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7478015422821045,
+      "learning_rate": 1.6249144826878182e-06,
+      "loss": 0.9418,
+      "step": 14324
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.67262601852417,
+      "learning_rate": 1.6238995832558358e-06,
+      "loss": 0.8111,
+      "step": 14325
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5477443933486938,
+      "learning_rate": 1.6228849728596818e-06,
+      "loss": 0.8651,
+      "step": 14326
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8882970809936523,
+      "learning_rate": 1.6218706515343652e-06,
+      "loss": 0.8542,
+      "step": 14327
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8037967681884766,
+      "learning_rate": 1.6208566193148922e-06,
+      "loss": 0.9123,
+      "step": 14328
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7822704315185547,
+      "learning_rate": 1.6198428762362473e-06,
+      "loss": 0.9266,
+      "step": 14329
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.728736400604248,
+      "learning_rate": 1.618829422333419e-06,
+      "loss": 0.9156,
+      "step": 14330
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1977746486663818,
+      "learning_rate": 1.6178162576413736e-06,
+      "loss": 0.5566,
+      "step": 14331
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8046597242355347,
+      "learning_rate": 1.6168033821950735e-06,
+      "loss": 0.9103,
+      "step": 14332
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7510817050933838,
+      "learning_rate": 1.615790796029474e-06,
+      "loss": 0.8537,
+      "step": 14333
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7535678148269653,
+      "learning_rate": 1.6147784991795113e-06,
+      "loss": 0.8846,
+      "step": 14334
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6615991592407227,
+      "learning_rate": 1.613766491680121e-06,
+      "loss": 0.8082,
+      "step": 14335
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8091779947280884,
+      "learning_rate": 1.6127547735662218e-06,
+      "loss": 0.9701,
+      "step": 14336
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9102187156677246,
+      "learning_rate": 1.6117433448727282e-06,
+      "loss": 0.9518,
+      "step": 14337
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7487438917160034,
+      "learning_rate": 1.6107322056345388e-06,
+      "loss": 0.948,
+      "step": 14338
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.793394684791565,
+      "learning_rate": 1.6097213558865478e-06,
+      "loss": 0.9151,
+      "step": 14339
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6446669101715088,
+      "learning_rate": 1.6087107956636338e-06,
+      "loss": 0.898,
+      "step": 14340
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6988474130630493,
+      "learning_rate": 1.6077005250006717e-06,
+      "loss": 0.8762,
+      "step": 14341
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8521496057510376,
+      "learning_rate": 1.6066905439325199e-06,
+      "loss": 0.7738,
+      "step": 14342
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7256203889846802,
+      "learning_rate": 1.6056808524940338e-06,
+      "loss": 0.956,
+      "step": 14343
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7402937412261963,
+      "learning_rate": 1.60467145072005e-06,
+      "loss": 0.8862,
+      "step": 14344
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7349143028259277,
+      "learning_rate": 1.6036623386454041e-06,
+      "loss": 0.9085,
+      "step": 14345
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7306697368621826,
+      "learning_rate": 1.6026535163049184e-06,
+      "loss": 0.8174,
+      "step": 14346
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.723577618598938,
+      "learning_rate": 1.6016449837334004e-06,
+      "loss": 0.9039,
+      "step": 14347
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6758160591125488,
+      "learning_rate": 1.6006367409656564e-06,
+      "loss": 0.8787,
+      "step": 14348
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.863316297531128,
+      "learning_rate": 1.5996287880364736e-06,
+      "loss": 0.9003,
+      "step": 14349
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8070380687713623,
+      "learning_rate": 1.5986211249806382e-06,
+      "loss": 0.8482,
+      "step": 14350
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0333231687545776,
+      "learning_rate": 1.5976137518329182e-06,
+      "loss": 0.5935,
+      "step": 14351
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6390674114227295,
+      "learning_rate": 1.5966066686280778e-06,
+      "loss": 0.8684,
+      "step": 14352
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8405100107192993,
+      "learning_rate": 1.595599875400865e-06,
+      "loss": 0.9065,
+      "step": 14353
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9881863594055176,
+      "learning_rate": 1.5945933721860263e-06,
+      "loss": 0.4933,
+      "step": 14354
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0013110637664795,
+      "learning_rate": 1.5935871590182883e-06,
+      "loss": 0.4961,
+      "step": 14355
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.699596881866455,
+      "learning_rate": 1.5925812359323745e-06,
+      "loss": 0.8972,
+      "step": 14356
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0379343032836914,
+      "learning_rate": 1.5915756029630004e-06,
+      "loss": 0.5512,
+      "step": 14357
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.777631163597107,
+      "learning_rate": 1.5905702601448615e-06,
+      "loss": 0.8946,
+      "step": 14358
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8313721418380737,
+      "learning_rate": 1.5895652075126545e-06,
+      "loss": 0.9193,
+      "step": 14359
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9245115518569946,
+      "learning_rate": 1.588560445101056e-06,
+      "loss": 0.9203,
+      "step": 14360
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8366649150848389,
+      "learning_rate": 1.587555972944742e-06,
+      "loss": 0.8777,
+      "step": 14361
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.961328387260437,
+      "learning_rate": 1.5865517910783712e-06,
+      "loss": 0.962,
+      "step": 14362
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8175925016403198,
+      "learning_rate": 1.585547899536598e-06,
+      "loss": 0.9832,
+      "step": 14363
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7593896389007568,
+      "learning_rate": 1.5845442983540593e-06,
+      "loss": 0.8848,
+      "step": 14364
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9625341892242432,
+      "learning_rate": 1.5835409875653884e-06,
+      "loss": 0.5052,
+      "step": 14365
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.770939588546753,
+      "learning_rate": 1.5825379672052088e-06,
+      "loss": 0.8532,
+      "step": 14366
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7892173528671265,
+      "learning_rate": 1.5815352373081328e-06,
+      "loss": 0.8539,
+      "step": 14367
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8567909002304077,
+      "learning_rate": 1.580532797908757e-06,
+      "loss": 0.9543,
+      "step": 14368
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.013876438140869,
+      "learning_rate": 1.5795306490416784e-06,
+      "loss": 0.9796,
+      "step": 14369
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7063997983932495,
+      "learning_rate": 1.5785287907414726e-06,
+      "loss": 0.9636,
+      "step": 14370
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.025414228439331,
+      "learning_rate": 1.5775272230427164e-06,
+      "loss": 0.5462,
+      "step": 14371
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9535168409347534,
+      "learning_rate": 1.5765259459799664e-06,
+      "loss": 0.923,
+      "step": 14372
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6779085397720337,
+      "learning_rate": 1.5755249595877752e-06,
+      "loss": 1.0168,
+      "step": 14373
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.902982473373413,
+      "learning_rate": 1.5745242639006886e-06,
+      "loss": 0.8881,
+      "step": 14374
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5876433849334717,
+      "learning_rate": 1.573523858953231e-06,
+      "loss": 0.946,
+      "step": 14375
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.75193190574646,
+      "learning_rate": 1.572523744779928e-06,
+      "loss": 0.9079,
+      "step": 14376
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9131932258605957,
+      "learning_rate": 1.5715239214152877e-06,
+      "loss": 0.8516,
+      "step": 14377
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.594518780708313,
+      "learning_rate": 1.570524388893816e-06,
+      "loss": 0.9242,
+      "step": 14378
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7981634140014648,
+      "learning_rate": 1.5695251472499974e-06,
+      "loss": 1.0,
+      "step": 14379
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6218276023864746,
+      "learning_rate": 1.5685261965183196e-06,
+      "loss": 0.9352,
+      "step": 14380
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.071269154548645,
+      "learning_rate": 1.5675275367332476e-06,
+      "loss": 0.5487,
+      "step": 14381
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6671252250671387,
+      "learning_rate": 1.5665291679292472e-06,
+      "loss": 0.8756,
+      "step": 14382
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8355664014816284,
+      "learning_rate": 1.565531090140765e-06,
+      "loss": 0.8914,
+      "step": 14383
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8472294807434082,
+      "learning_rate": 1.564533303402247e-06,
+      "loss": 0.8645,
+      "step": 14384
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7986302375793457,
+      "learning_rate": 1.563535807748119e-06,
+      "loss": 0.9585,
+      "step": 14385
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6741151809692383,
+      "learning_rate": 1.562538603212803e-06,
+      "loss": 0.9099,
+      "step": 14386
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.724277377128601,
+      "learning_rate": 1.5615416898307135e-06,
+      "loss": 0.9718,
+      "step": 14387
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8047869205474854,
+      "learning_rate": 1.5605450676362465e-06,
+      "loss": 0.9433,
+      "step": 14388
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.727473258972168,
+      "learning_rate": 1.5595487366637962e-06,
+      "loss": 0.8724,
+      "step": 14389
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7540818452835083,
+      "learning_rate": 1.5585526969477394e-06,
+      "loss": 0.8661,
+      "step": 14390
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8403637409210205,
+      "learning_rate": 1.5575569485224519e-06,
+      "loss": 0.8312,
+      "step": 14391
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7430156469345093,
+      "learning_rate": 1.556561491422287e-06,
+      "loss": 0.8698,
+      "step": 14392
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8151819705963135,
+      "learning_rate": 1.5555663256816033e-06,
+      "loss": 0.9164,
+      "step": 14393
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6449781656265259,
+      "learning_rate": 1.5545714513347343e-06,
+      "loss": 0.8366,
+      "step": 14394
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0620030164718628,
+      "learning_rate": 1.5535768684160158e-06,
+      "loss": 0.5468,
+      "step": 14395
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.733941674232483,
+      "learning_rate": 1.5525825769597625e-06,
+      "loss": 0.938,
+      "step": 14396
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.75095796585083,
+      "learning_rate": 1.5515885770002891e-06,
+      "loss": 0.9465,
+      "step": 14397
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7997056245803833,
+      "learning_rate": 1.550594868571893e-06,
+      "loss": 0.9154,
+      "step": 14398
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7493253946304321,
+      "learning_rate": 1.5496014517088654e-06,
+      "loss": 0.8431,
+      "step": 14399
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7949256896972656,
+      "learning_rate": 1.5486083264454887e-06,
+      "loss": 0.9434,
+      "step": 14400
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8511708974838257,
+      "learning_rate": 1.547615492816029e-06,
+      "loss": 0.9624,
+      "step": 14401
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9339182376861572,
+      "learning_rate": 1.5466229508547492e-06,
+      "loss": 0.8446,
+      "step": 14402
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5918800830841064,
+      "learning_rate": 1.545630700595896e-06,
+      "loss": 0.8253,
+      "step": 14403
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9751707315444946,
+      "learning_rate": 1.544638742073713e-06,
+      "loss": 0.9335,
+      "step": 14404
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7182022333145142,
+      "learning_rate": 1.5436470753224264e-06,
+      "loss": 0.8629,
+      "step": 14405
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8121334314346313,
+      "learning_rate": 1.5426557003762587e-06,
+      "loss": 0.8312,
+      "step": 14406
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8647698163986206,
+      "learning_rate": 1.541664617269416e-06,
+      "loss": 0.9006,
+      "step": 14407
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6378693580627441,
+      "learning_rate": 1.5406738260361031e-06,
+      "loss": 0.9883,
+      "step": 14408
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6593563556671143,
+      "learning_rate": 1.5396833267105026e-06,
+      "loss": 0.7769,
+      "step": 14409
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6766915321350098,
+      "learning_rate": 1.5386931193267983e-06,
+      "loss": 0.8609,
+      "step": 14410
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9499013423919678,
+      "learning_rate": 1.5377032039191608e-06,
+      "loss": 0.9692,
+      "step": 14411
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6247973442077637,
+      "learning_rate": 1.536713580521746e-06,
+      "loss": 0.9148,
+      "step": 14412
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.914451003074646,
+      "learning_rate": 1.5357242491687052e-06,
+      "loss": 0.9583,
+      "step": 14413
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9622336626052856,
+      "learning_rate": 1.5347352098941748e-06,
+      "loss": 0.9131,
+      "step": 14414
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8761560916900635,
+      "learning_rate": 1.5337464627322884e-06,
+      "loss": 0.8787,
+      "step": 14415
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7535345554351807,
+      "learning_rate": 1.5327580077171589e-06,
+      "loss": 0.9009,
+      "step": 14416
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8927239179611206,
+      "learning_rate": 1.531769844882901e-06,
+      "loss": 0.8654,
+      "step": 14417
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7496799230575562,
+      "learning_rate": 1.5307819742636088e-06,
+      "loss": 0.926,
+      "step": 14418
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7160295248031616,
+      "learning_rate": 1.5297943958933748e-06,
+      "loss": 0.8637,
+      "step": 14419
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0564165115356445,
+      "learning_rate": 1.5288071098062728e-06,
+      "loss": 0.9452,
+      "step": 14420
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.796385407447815,
+      "learning_rate": 1.527820116036377e-06,
+      "loss": 0.8583,
+      "step": 14421
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1175018548965454,
+      "learning_rate": 1.5268334146177399e-06,
+      "loss": 0.6195,
+      "step": 14422
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.62832510471344,
+      "learning_rate": 1.5258470055844131e-06,
+      "loss": 0.8988,
+      "step": 14423
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7216987609863281,
+      "learning_rate": 1.5248608889704374e-06,
+      "loss": 0.9537,
+      "step": 14424
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9396353960037231,
+      "learning_rate": 1.5238750648098354e-06,
+      "loss": 0.9557,
+      "step": 14425
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7440906763076782,
+      "learning_rate": 1.5228895331366301e-06,
+      "loss": 0.8261,
+      "step": 14426
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7885385751724243,
+      "learning_rate": 1.5219042939848249e-06,
+      "loss": 0.8973,
+      "step": 14427
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0642056465148926,
+      "learning_rate": 1.5209193473884232e-06,
+      "loss": 0.5111,
+      "step": 14428
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7156176567077637,
+      "learning_rate": 1.5199346933814052e-06,
+      "loss": 0.8487,
+      "step": 14429
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7713056802749634,
+      "learning_rate": 1.5189503319977573e-06,
+      "loss": 0.8843,
+      "step": 14430
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7010061740875244,
+      "learning_rate": 1.5179662632714364e-06,
+      "loss": 0.8402,
+      "step": 14431
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.752675175666809,
+      "learning_rate": 1.5169824872364115e-06,
+      "loss": 0.855,
+      "step": 14432
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8046355247497559,
+      "learning_rate": 1.5159990039266215e-06,
+      "loss": 0.9296,
+      "step": 14433
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7720823287963867,
+      "learning_rate": 1.5150158133760095e-06,
+      "loss": 0.8991,
+      "step": 14434
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7415226697921753,
+      "learning_rate": 1.5140329156184974e-06,
+      "loss": 0.9119,
+      "step": 14435
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8239010572433472,
+      "learning_rate": 1.513050310688008e-06,
+      "loss": 0.9358,
+      "step": 14436
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.730281949043274,
+      "learning_rate": 1.5120679986184417e-06,
+      "loss": 0.8895,
+      "step": 14437
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6903257369995117,
+      "learning_rate": 1.5110859794437016e-06,
+      "loss": 0.8775,
+      "step": 14438
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.767596960067749,
+      "learning_rate": 1.5101042531976696e-06,
+      "loss": 0.8957,
+      "step": 14439
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7511062622070312,
+      "learning_rate": 1.5091228199142238e-06,
+      "loss": 0.8839,
+      "step": 14440
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6825076341629028,
+      "learning_rate": 1.508141679627233e-06,
+      "loss": 0.8404,
+      "step": 14441
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6293699741363525,
+      "learning_rate": 1.50716083237055e-06,
+      "loss": 0.8414,
+      "step": 14442
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.715467095375061,
+      "learning_rate": 1.5061802781780244e-06,
+      "loss": 0.9687,
+      "step": 14443
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5926454067230225,
+      "learning_rate": 1.50520001708349e-06,
+      "loss": 0.8927,
+      "step": 14444
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7566200494766235,
+      "learning_rate": 1.5042200491207747e-06,
+      "loss": 0.8873,
+      "step": 14445
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6798533201217651,
+      "learning_rate": 1.5032403743236924e-06,
+      "loss": 0.8558,
+      "step": 14446
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.879533052444458,
+      "learning_rate": 1.5022609927260512e-06,
+      "loss": 0.9455,
+      "step": 14447
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0210976600646973,
+      "learning_rate": 1.5012819043616445e-06,
+      "loss": 0.9137,
+      "step": 14448
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.763628602027893,
+      "learning_rate": 1.5003031092642605e-06,
+      "loss": 0.9455,
+      "step": 14449
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7218728065490723,
+      "learning_rate": 1.4993246074676714e-06,
+      "loss": 0.9265,
+      "step": 14450
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5834935903549194,
+      "learning_rate": 1.4983463990056467e-06,
+      "loss": 0.8959,
+      "step": 14451
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6832629442214966,
+      "learning_rate": 1.4973684839119362e-06,
+      "loss": 0.8288,
+      "step": 14452
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8081368207931519,
+      "learning_rate": 1.4963908622202894e-06,
+      "loss": 0.9487,
+      "step": 14453
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7063283920288086,
+      "learning_rate": 1.4954135339644416e-06,
+      "loss": 0.9616,
+      "step": 14454
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.804421067237854,
+      "learning_rate": 1.4944364991781147e-06,
+      "loss": 0.9016,
+      "step": 14455
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.031509518623352,
+      "learning_rate": 1.493459757895026e-06,
+      "loss": 0.558,
+      "step": 14456
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6179637908935547,
+      "learning_rate": 1.4924833101488768e-06,
+      "loss": 0.9058,
+      "step": 14457
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6956065893173218,
+      "learning_rate": 1.4915071559733673e-06,
+      "loss": 0.9841,
+      "step": 14458
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6311208009719849,
+      "learning_rate": 1.4905312954021745e-06,
+      "loss": 0.9052,
+      "step": 14459
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0202057361602783,
+      "learning_rate": 1.48955572846898e-06,
+      "loss": 0.9269,
+      "step": 14460
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7691367864608765,
+      "learning_rate": 1.4885804552074413e-06,
+      "loss": 0.8756,
+      "step": 14461
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7620258331298828,
+      "learning_rate": 1.4876054756512182e-06,
+      "loss": 0.8745,
+      "step": 14462
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.795127034187317,
+      "learning_rate": 1.4866307898339493e-06,
+      "loss": 0.9139,
+      "step": 14463
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9033477306365967,
+      "learning_rate": 1.48565639778927e-06,
+      "loss": 0.8792,
+      "step": 14464
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7459214925765991,
+      "learning_rate": 1.4846822995508082e-06,
+      "loss": 0.9281,
+      "step": 14465
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8313684463500977,
+      "learning_rate": 1.4837084951521708e-06,
+      "loss": 0.953,
+      "step": 14466
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8939236402511597,
+      "learning_rate": 1.4827349846269656e-06,
+      "loss": 0.9173,
+      "step": 14467
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8759583234786987,
+      "learning_rate": 1.4817617680087826e-06,
+      "loss": 0.9299,
+      "step": 14468
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1508959531784058,
+      "learning_rate": 1.480788845331208e-06,
+      "loss": 0.5976,
+      "step": 14469
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7790790796279907,
+      "learning_rate": 1.4798162166278108e-06,
+      "loss": 0.8131,
+      "step": 14470
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7428867816925049,
+      "learning_rate": 1.4788438819321582e-06,
+      "loss": 0.8461,
+      "step": 14471
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.825386643409729,
+      "learning_rate": 1.477871841277797e-06,
+      "loss": 0.8968,
+      "step": 14472
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0135072469711304,
+      "learning_rate": 1.476900094698277e-06,
+      "loss": 0.5121,
+      "step": 14473
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0950936079025269,
+      "learning_rate": 1.4759286422271224e-06,
+      "loss": 0.5238,
+      "step": 14474
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8173545598983765,
+      "learning_rate": 1.474957483897863e-06,
+      "loss": 0.8334,
+      "step": 14475
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7673684358596802,
+      "learning_rate": 1.4739866197440046e-06,
+      "loss": 0.8571,
+      "step": 14476
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0635735988616943,
+      "learning_rate": 1.4730160497990509e-06,
+      "loss": 0.5199,
+      "step": 14477
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8114100694656372,
+      "learning_rate": 1.4720457740964966e-06,
+      "loss": 0.9029,
+      "step": 14478
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8555278778076172,
+      "learning_rate": 1.4710757926698182e-06,
+      "loss": 0.8888,
+      "step": 14479
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7651171684265137,
+      "learning_rate": 1.4701061055524924e-06,
+      "loss": 0.8153,
+      "step": 14480
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6576436758041382,
+      "learning_rate": 1.4691367127779754e-06,
+      "loss": 0.867,
+      "step": 14481
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7941083908081055,
+      "learning_rate": 1.468167614379723e-06,
+      "loss": 0.9149,
+      "step": 14482
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.943681001663208,
+      "learning_rate": 1.4671988103911704e-06,
+      "loss": 0.9049,
+      "step": 14483
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8133623600006104,
+      "learning_rate": 1.4662303008457536e-06,
+      "loss": 0.8955,
+      "step": 14484
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7643803358078003,
+      "learning_rate": 1.4652620857768895e-06,
+      "loss": 0.9603,
+      "step": 14485
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8418092727661133,
+      "learning_rate": 1.464294165217992e-06,
+      "loss": 0.8432,
+      "step": 14486
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7747565507888794,
+      "learning_rate": 1.4633265392024564e-06,
+      "loss": 0.9211,
+      "step": 14487
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8276382684707642,
+      "learning_rate": 1.4623592077636772e-06,
+      "loss": 0.8965,
+      "step": 14488
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5917805433273315,
+      "learning_rate": 1.4613921709350342e-06,
+      "loss": 0.8688,
+      "step": 14489
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7457362413406372,
+      "learning_rate": 1.460425428749893e-06,
+      "loss": 0.9517,
+      "step": 14490
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.878147840499878,
+      "learning_rate": 1.4594589812416182e-06,
+      "loss": 0.9404,
+      "step": 14491
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8353899717330933,
+      "learning_rate": 1.458492828443555e-06,
+      "loss": 0.8644,
+      "step": 14492
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8390419483184814,
+      "learning_rate": 1.4575269703890471e-06,
+      "loss": 0.7979,
+      "step": 14493
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8015797138214111,
+      "learning_rate": 1.4565614071114187e-06,
+      "loss": 0.9113,
+      "step": 14494
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.731790542602539,
+      "learning_rate": 1.4555961386439933e-06,
+      "loss": 0.8135,
+      "step": 14495
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9548718333244324,
+      "learning_rate": 1.454631165020075e-06,
+      "loss": 0.5364,
+      "step": 14496
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.733827829360962,
+      "learning_rate": 1.4536664862729643e-06,
+      "loss": 0.9148,
+      "step": 14497
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8217209577560425,
+      "learning_rate": 1.45270210243595e-06,
+      "loss": 0.8724,
+      "step": 14498
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8738255500793457,
+      "learning_rate": 1.4517380135423132e-06,
+      "loss": 0.8954,
+      "step": 14499
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6562740802764893,
+      "learning_rate": 1.450774219625316e-06,
+      "loss": 0.906,
+      "step": 14500
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7145521640777588,
+      "learning_rate": 1.449810720718221e-06,
+      "loss": 0.9558,
+      "step": 14501
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6991722583770752,
+      "learning_rate": 1.4488475168542725e-06,
+      "loss": 0.9062,
+      "step": 14502
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7761893272399902,
+      "learning_rate": 1.447884608066712e-06,
+      "loss": 0.8824,
+      "step": 14503
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8761807680130005,
+      "learning_rate": 1.4469219943887613e-06,
+      "loss": 0.8981,
+      "step": 14504
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8143270015716553,
+      "learning_rate": 1.445959675853641e-06,
+      "loss": 0.8536,
+      "step": 14505
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.098664402961731,
+      "learning_rate": 1.4449976524945598e-06,
+      "loss": 0.5375,
+      "step": 14506
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8778107166290283,
+      "learning_rate": 1.44403592434471e-06,
+      "loss": 0.8831,
+      "step": 14507
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8655641078948975,
+      "learning_rate": 1.443074491437283e-06,
+      "loss": 0.9147,
+      "step": 14508
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.838841199874878,
+      "learning_rate": 1.442113353805449e-06,
+      "loss": 0.8532,
+      "step": 14509
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6403008699417114,
+      "learning_rate": 1.441152511482381e-06,
+      "loss": 0.8071,
+      "step": 14510
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9378477334976196,
+      "learning_rate": 1.4401919645012286e-06,
+      "loss": 0.9109,
+      "step": 14511
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.655186653137207,
+      "learning_rate": 1.4392317128951438e-06,
+      "loss": 0.8912,
+      "step": 14512
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8928735256195068,
+      "learning_rate": 1.438271756697256e-06,
+      "loss": 0.956,
+      "step": 14513
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.75174880027771,
+      "learning_rate": 1.437312095940696e-06,
+      "loss": 0.9112,
+      "step": 14514
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7490835189819336,
+      "learning_rate": 1.4363527306585744e-06,
+      "loss": 0.9233,
+      "step": 14515
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8052923679351807,
+      "learning_rate": 1.4353936608840014e-06,
+      "loss": 0.8525,
+      "step": 14516
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.828856110572815,
+      "learning_rate": 1.4344348866500657e-06,
+      "loss": 0.945,
+      "step": 14517
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8553471565246582,
+      "learning_rate": 1.4334764079898556e-06,
+      "loss": 0.9025,
+      "step": 14518
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7000377178192139,
+      "learning_rate": 1.4325182249364477e-06,
+      "loss": 0.9362,
+      "step": 14519
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8315136432647705,
+      "learning_rate": 1.4315603375229003e-06,
+      "loss": 0.9235,
+      "step": 14520
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8198537826538086,
+      "learning_rate": 1.4306027457822735e-06,
+      "loss": 0.8198,
+      "step": 14521
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8960990905761719,
+      "learning_rate": 1.4296454497476064e-06,
+      "loss": 0.8894,
+      "step": 14522
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.934336543083191,
+      "learning_rate": 1.428688449451937e-06,
+      "loss": 0.8653,
+      "step": 14523
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8146134614944458,
+      "learning_rate": 1.4277317449282834e-06,
+      "loss": 0.9836,
+      "step": 14524
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7093007564544678,
+      "learning_rate": 1.4267753362096637e-06,
+      "loss": 0.9747,
+      "step": 14525
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.744572639465332,
+      "learning_rate": 1.4258192233290769e-06,
+      "loss": 0.8567,
+      "step": 14526
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0023268461227417,
+      "learning_rate": 1.4248634063195198e-06,
+      "loss": 0.5738,
+      "step": 14527
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5838576555252075,
+      "learning_rate": 1.4239078852139698e-06,
+      "loss": 0.8924,
+      "step": 14528
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9270954132080078,
+      "learning_rate": 1.4229526600454058e-06,
+      "loss": 0.869,
+      "step": 14529
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6948373317718506,
+      "learning_rate": 1.4219977308467836e-06,
+      "loss": 0.884,
+      "step": 14530
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.615168809890747,
+      "learning_rate": 1.421043097651058e-06,
+      "loss": 0.9307,
+      "step": 14531
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.896254062652588,
+      "learning_rate": 1.420088760491174e-06,
+      "loss": 0.8447,
+      "step": 14532
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7818957567214966,
+      "learning_rate": 1.419134719400057e-06,
+      "loss": 0.942,
+      "step": 14533
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0576672554016113,
+      "learning_rate": 1.4181809744106334e-06,
+      "loss": 0.5439,
+      "step": 14534
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7533237934112549,
+      "learning_rate": 1.4172275255558088e-06,
+      "loss": 0.7857,
+      "step": 14535
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.831253170967102,
+      "learning_rate": 1.4162743728684914e-06,
+      "loss": 0.8599,
+      "step": 14536
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9453730583190918,
+      "learning_rate": 1.4153215163815637e-06,
+      "loss": 0.8341,
+      "step": 14537
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7503539323806763,
+      "learning_rate": 1.4143689561279138e-06,
+      "loss": 0.8818,
+      "step": 14538
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9909722208976746,
+      "learning_rate": 1.4134166921404047e-06,
+      "loss": 0.5849,
+      "step": 14539
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6401209831237793,
+      "learning_rate": 1.412464724451903e-06,
+      "loss": 0.9332,
+      "step": 14540
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8063476085662842,
+      "learning_rate": 1.4115130530952526e-06,
+      "loss": 0.9075,
+      "step": 14541
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6439341306686401,
+      "learning_rate": 1.410561678103296e-06,
+      "loss": 0.8542,
+      "step": 14542
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8168108463287354,
+      "learning_rate": 1.4096105995088648e-06,
+      "loss": 0.8325,
+      "step": 14543
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7581793069839478,
+      "learning_rate": 1.4086598173447729e-06,
+      "loss": 0.8185,
+      "step": 14544
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7775856256484985,
+      "learning_rate": 1.407709331643834e-06,
+      "loss": 0.8453,
+      "step": 14545
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8567034006118774,
+      "learning_rate": 1.4067591424388427e-06,
+      "loss": 0.9075,
+      "step": 14546
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8719321489334106,
+      "learning_rate": 1.405809249762591e-06,
+      "loss": 0.9418,
+      "step": 14547
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7581191062927246,
+      "learning_rate": 1.404859653647853e-06,
+      "loss": 0.9579,
+      "step": 14548
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.715213656425476,
+      "learning_rate": 1.4039103541274013e-06,
+      "loss": 0.8069,
+      "step": 14549
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7716665267944336,
+      "learning_rate": 1.402961351233989e-06,
+      "loss": 0.9809,
+      "step": 14550
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7329015731811523,
+      "learning_rate": 1.4020126450003669e-06,
+      "loss": 0.8811,
+      "step": 14551
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8000420331954956,
+      "learning_rate": 1.4010642354592697e-06,
+      "loss": 0.8852,
+      "step": 14552
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.110851764678955,
+      "learning_rate": 1.4001161226434267e-06,
+      "loss": 0.9364,
+      "step": 14553
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8728327751159668,
+      "learning_rate": 1.399168306585552e-06,
+      "loss": 0.9032,
+      "step": 14554
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.024374008178711,
+      "learning_rate": 1.398220787318354e-06,
+      "loss": 0.5773,
+      "step": 14555
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0640523433685303,
+      "learning_rate": 1.3972735648745295e-06,
+      "loss": 0.8925,
+      "step": 14556
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6798175573349,
+      "learning_rate": 1.3963266392867624e-06,
+      "loss": 0.8254,
+      "step": 14557
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8340222835540771,
+      "learning_rate": 1.3953800105877313e-06,
+      "loss": 0.897,
+      "step": 14558
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.649770975112915,
+      "learning_rate": 1.3944336788100976e-06,
+      "loss": 0.9169,
+      "step": 14559
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7816296815872192,
+      "learning_rate": 1.393487643986522e-06,
+      "loss": 0.9888,
+      "step": 14560
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1148887872695923,
+      "learning_rate": 1.3925419061496436e-06,
+      "loss": 0.5355,
+      "step": 14561
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7746391296386719,
+      "learning_rate": 1.3915964653321023e-06,
+      "loss": 0.9097,
+      "step": 14562
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.766114592552185,
+      "learning_rate": 1.3906513215665162e-06,
+      "loss": 0.9787,
+      "step": 14563
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7979731559753418,
+      "learning_rate": 1.3897064748855083e-06,
+      "loss": 0.8803,
+      "step": 14564
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9663290977478027,
+      "learning_rate": 1.3887619253216756e-06,
+      "loss": 0.5337,
+      "step": 14565
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8370264768600464,
+      "learning_rate": 1.3878176729076166e-06,
+      "loss": 0.8683,
+      "step": 14566
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8813267946243286,
+      "learning_rate": 1.3868737176759105e-06,
+      "loss": 0.9833,
+      "step": 14567
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.843646764755249,
+      "learning_rate": 1.3859300596591342e-06,
+      "loss": 0.9492,
+      "step": 14568
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8345117568969727,
+      "learning_rate": 1.3849866988898474e-06,
+      "loss": 0.9316,
+      "step": 14569
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7999660968780518,
+      "learning_rate": 1.384043635400607e-06,
+      "loss": 0.904,
+      "step": 14570
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9210731983184814,
+      "learning_rate": 1.38310086922395e-06,
+      "loss": 0.8121,
+      "step": 14571
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8331910371780396,
+      "learning_rate": 1.3821584003924127e-06,
+      "loss": 0.8792,
+      "step": 14572
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7950514554977417,
+      "learning_rate": 1.3812162289385178e-06,
+      "loss": 0.9092,
+      "step": 14573
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8383709192276,
+      "learning_rate": 1.3802743548947729e-06,
+      "loss": 0.933,
+      "step": 14574
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0912328958511353,
+      "learning_rate": 1.3793327782936839e-06,
+      "loss": 0.5875,
+      "step": 14575
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8132764101028442,
+      "learning_rate": 1.3783914991677373e-06,
+      "loss": 0.8852,
+      "step": 14576
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8958594799041748,
+      "learning_rate": 1.3774505175494191e-06,
+      "loss": 0.9726,
+      "step": 14577
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.754158616065979,
+      "learning_rate": 1.3765098334711958e-06,
+      "loss": 0.8495,
+      "step": 14578
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8186291456222534,
+      "learning_rate": 1.375569446965531e-06,
+      "loss": 0.901,
+      "step": 14579
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7348216772079468,
+      "learning_rate": 1.3746293580648718e-06,
+      "loss": 0.8185,
+      "step": 14580
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9503923654556274,
+      "learning_rate": 1.3736895668016603e-06,
+      "loss": 0.9087,
+      "step": 14581
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6520909070968628,
+      "learning_rate": 1.3727500732083242e-06,
+      "loss": 0.9836,
+      "step": 14582
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6311596632003784,
+      "learning_rate": 1.3718108773172855e-06,
+      "loss": 0.8208,
+      "step": 14583
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9474674463272095,
+      "learning_rate": 1.3708719791609494e-06,
+      "loss": 0.8268,
+      "step": 14584
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6714344024658203,
+      "learning_rate": 1.3699333787717173e-06,
+      "loss": 0.7629,
+      "step": 14585
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0072879791259766,
+      "learning_rate": 1.3689950761819781e-06,
+      "loss": 0.5606,
+      "step": 14586
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6865061521530151,
+      "learning_rate": 1.3680570714241082e-06,
+      "loss": 0.8563,
+      "step": 14587
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7102704048156738,
+      "learning_rate": 1.367119364530478e-06,
+      "loss": 0.8933,
+      "step": 14588
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7438291311264038,
+      "learning_rate": 1.366181955533441e-06,
+      "loss": 0.9024,
+      "step": 14589
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6683228015899658,
+      "learning_rate": 1.3652448444653499e-06,
+      "loss": 0.8125,
+      "step": 14590
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7147173881530762,
+      "learning_rate": 1.3643080313585366e-06,
+      "loss": 0.8906,
+      "step": 14591
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8748008012771606,
+      "learning_rate": 1.363371516245333e-06,
+      "loss": 0.9295,
+      "step": 14592
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.777127981185913,
+      "learning_rate": 1.3624352991580503e-06,
+      "loss": 0.8788,
+      "step": 14593
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0049889087677002,
+      "learning_rate": 1.361499380129e-06,
+      "loss": 0.5152,
+      "step": 14594
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.757861614227295,
+      "learning_rate": 1.360563759190473e-06,
+      "loss": 0.9012,
+      "step": 14595
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7264031171798706,
+      "learning_rate": 1.359628436374757e-06,
+      "loss": 0.8299,
+      "step": 14596
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7234222888946533,
+      "learning_rate": 1.3586934117141304e-06,
+      "loss": 0.8478,
+      "step": 14597
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.686714768409729,
+      "learning_rate": 1.357758685240853e-06,
+      "loss": 0.9318,
+      "step": 14598
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7773762941360474,
+      "learning_rate": 1.3568242569871847e-06,
+      "loss": 0.914,
+      "step": 14599
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6407865285873413,
+      "learning_rate": 1.3558901269853653e-06,
+      "loss": 0.9006,
+      "step": 14600
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9167946577072144,
+      "learning_rate": 1.354956295267633e-06,
+      "loss": 0.499,
+      "step": 14601
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7722874879837036,
+      "learning_rate": 1.3540227618662082e-06,
+      "loss": 0.8829,
+      "step": 14602
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5769129991531372,
+      "learning_rate": 1.3530895268133083e-06,
+      "loss": 0.9115,
+      "step": 14603
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7653549909591675,
+      "learning_rate": 1.3521565901411327e-06,
+      "loss": 0.8497,
+      "step": 14604
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7872886657714844,
+      "learning_rate": 1.3512239518818793e-06,
+      "loss": 0.9064,
+      "step": 14605
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.924564242362976,
+      "learning_rate": 1.3502916120677246e-06,
+      "loss": 0.8283,
+      "step": 14606
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.663714051246643,
+      "learning_rate": 1.3493595707308472e-06,
+      "loss": 0.8743,
+      "step": 14607
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.709092378616333,
+      "learning_rate": 1.3484278279034046e-06,
+      "loss": 0.8754,
+      "step": 14608
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8538317680358887,
+      "learning_rate": 1.3474963836175492e-06,
+      "loss": 0.9097,
+      "step": 14609
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7197909355163574,
+      "learning_rate": 1.3465652379054273e-06,
+      "loss": 0.9511,
+      "step": 14610
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.766197919845581,
+      "learning_rate": 1.3456343907991632e-06,
+      "loss": 0.9089,
+      "step": 14611
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7603367567062378,
+      "learning_rate": 1.3447038423308845e-06,
+      "loss": 0.9528,
+      "step": 14612
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7487378120422363,
+      "learning_rate": 1.3437735925326968e-06,
+      "loss": 0.9684,
+      "step": 14613
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0952643156051636,
+      "learning_rate": 1.342843641436703e-06,
+      "loss": 0.5713,
+      "step": 14614
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.793286681175232,
+      "learning_rate": 1.341913989074991e-06,
+      "loss": 0.8595,
+      "step": 14615
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.833357810974121,
+      "learning_rate": 1.340984635479644e-06,
+      "loss": 0.8689,
+      "step": 14616
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.680869221687317,
+      "learning_rate": 1.3400555806827265e-06,
+      "loss": 0.7951,
+      "step": 14617
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8264920711517334,
+      "learning_rate": 1.3391268247163037e-06,
+      "loss": 0.9169,
+      "step": 14618
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.704236626625061,
+      "learning_rate": 1.3381983676124178e-06,
+      "loss": 0.7916,
+      "step": 14619
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.979659914970398,
+      "learning_rate": 1.337270209403111e-06,
+      "loss": 0.8328,
+      "step": 14620
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7868047952651978,
+      "learning_rate": 1.336342350120413e-06,
+      "loss": 0.9431,
+      "step": 14621
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8141860961914062,
+      "learning_rate": 1.3354147897963365e-06,
+      "loss": 0.9386,
+      "step": 14622
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7215797901153564,
+      "learning_rate": 1.3344875284628956e-06,
+      "loss": 0.8701,
+      "step": 14623
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6068801879882812,
+      "learning_rate": 1.333560566152081e-06,
+      "loss": 0.8972,
+      "step": 14624
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7842477560043335,
+      "learning_rate": 1.3326339028958846e-06,
+      "loss": 0.8862,
+      "step": 14625
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0018666982650757,
+      "learning_rate": 1.331707538726279e-06,
+      "loss": 0.5578,
+      "step": 14626
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.108668088912964,
+      "learning_rate": 1.3307814736752344e-06,
+      "loss": 0.9717,
+      "step": 14627
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7983465194702148,
+      "learning_rate": 1.3298557077747032e-06,
+      "loss": 0.8959,
+      "step": 14628
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.971538245677948,
+      "learning_rate": 1.3289302410566318e-06,
+      "loss": 0.5085,
+      "step": 14629
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8674952983856201,
+      "learning_rate": 1.328005073552956e-06,
+      "loss": 0.9247,
+      "step": 14630
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7466880083084106,
+      "learning_rate": 1.3270802052956033e-06,
+      "loss": 0.8768,
+      "step": 14631
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8704296350479126,
+      "learning_rate": 1.326155636316483e-06,
+      "loss": 0.8788,
+      "step": 14632
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.635634422302246,
+      "learning_rate": 1.3252313666475058e-06,
+      "loss": 0.8601,
+      "step": 14633
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7505536079406738,
+      "learning_rate": 1.3243073963205589e-06,
+      "loss": 0.8722,
+      "step": 14634
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7863550186157227,
+      "learning_rate": 1.3233837253675319e-06,
+      "loss": 0.8488,
+      "step": 14635
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7028264999389648,
+      "learning_rate": 1.3224603538202929e-06,
+      "loss": 0.9937,
+      "step": 14636
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8244420289993286,
+      "learning_rate": 1.3215372817107098e-06,
+      "loss": 0.896,
+      "step": 14637
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9127250909805298,
+      "learning_rate": 1.3206145090706302e-06,
+      "loss": 0.9085,
+      "step": 14638
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7490317821502686,
+      "learning_rate": 1.3196920359318998e-06,
+      "loss": 0.8647,
+      "step": 14639
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7005019187927246,
+      "learning_rate": 1.3187698623263511e-06,
+      "loss": 0.8674,
+      "step": 14640
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5824023485183716,
+      "learning_rate": 1.317847988285803e-06,
+      "loss": 0.8092,
+      "step": 14641
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8160111904144287,
+      "learning_rate": 1.31692641384207e-06,
+      "loss": 0.8675,
+      "step": 14642
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7849373817443848,
+      "learning_rate": 1.316005139026949e-06,
+      "loss": 0.8471,
+      "step": 14643
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7585822343826294,
+      "learning_rate": 1.3150841638722355e-06,
+      "loss": 0.8718,
+      "step": 14644
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8520047664642334,
+      "learning_rate": 1.3141634884097043e-06,
+      "loss": 0.9042,
+      "step": 14645
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.841574788093567,
+      "learning_rate": 1.313243112671131e-06,
+      "loss": 0.8655,
+      "step": 14646
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7779295444488525,
+      "learning_rate": 1.31232303668827e-06,
+      "loss": 0.9831,
+      "step": 14647
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9942909479141235,
+      "learning_rate": 1.311403260492875e-06,
+      "loss": 0.5375,
+      "step": 14648
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8605635166168213,
+      "learning_rate": 1.3104837841166807e-06,
+      "loss": 0.8539,
+      "step": 14649
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7608391046524048,
+      "learning_rate": 1.309564607591418e-06,
+      "loss": 0.8003,
+      "step": 14650
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7137819528579712,
+      "learning_rate": 1.3086457309488066e-06,
+      "loss": 0.8902,
+      "step": 14651
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6845453977584839,
+      "learning_rate": 1.3077271542205517e-06,
+      "loss": 0.8504,
+      "step": 14652
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8603036403656006,
+      "learning_rate": 1.3068088774383525e-06,
+      "loss": 0.8432,
+      "step": 14653
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8364906311035156,
+      "learning_rate": 1.305890900633895e-06,
+      "loss": 0.9811,
+      "step": 14654
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7778388261795044,
+      "learning_rate": 1.304973223838857e-06,
+      "loss": 0.9553,
+      "step": 14655
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8376907110214233,
+      "learning_rate": 1.304055847084903e-06,
+      "loss": 0.9702,
+      "step": 14656
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8152296543121338,
+      "learning_rate": 1.3031387704036935e-06,
+      "loss": 0.8531,
+      "step": 14657
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9518057703971863,
+      "learning_rate": 1.3022219938268677e-06,
+      "loss": 0.4785,
+      "step": 14658
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8720519542694092,
+      "learning_rate": 1.3013055173860678e-06,
+      "loss": 0.9051,
+      "step": 14659
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7312129735946655,
+      "learning_rate": 1.3003893411129131e-06,
+      "loss": 0.9132,
+      "step": 14660
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6815332174301147,
+      "learning_rate": 1.2994734650390239e-06,
+      "loss": 0.8798,
+      "step": 14661
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8160277605056763,
+      "learning_rate": 1.2985578891959983e-06,
+      "loss": 0.8936,
+      "step": 14662
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.795061707496643,
+      "learning_rate": 1.297642613615434e-06,
+      "loss": 0.8924,
+      "step": 14663
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8633986711502075,
+      "learning_rate": 1.296727638328915e-06,
+      "loss": 0.8444,
+      "step": 14664
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6985819339752197,
+      "learning_rate": 1.2958129633680128e-06,
+      "loss": 0.8705,
+      "step": 14665
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6409038305282593,
+      "learning_rate": 1.294898588764293e-06,
+      "loss": 0.8598,
+      "step": 14666
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8164708614349365,
+      "learning_rate": 1.2939845145493036e-06,
+      "loss": 0.8959,
+      "step": 14667
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5908163785934448,
+      "learning_rate": 1.2930707407545917e-06,
+      "loss": 0.9183,
+      "step": 14668
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9382691383361816,
+      "learning_rate": 1.2921572674116845e-06,
+      "loss": 0.9225,
+      "step": 14669
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8026989698410034,
+      "learning_rate": 1.2912440945521087e-06,
+      "loss": 0.9021,
+      "step": 14670
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8587267398834229,
+      "learning_rate": 1.2903312222073695e-06,
+      "loss": 0.9341,
+      "step": 14671
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7153987884521484,
+      "learning_rate": 1.2894186504089712e-06,
+      "loss": 0.8362,
+      "step": 14672
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8463294506072998,
+      "learning_rate": 1.2885063791884023e-06,
+      "loss": 0.956,
+      "step": 14673
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7521140575408936,
+      "learning_rate": 1.2875944085771441e-06,
+      "loss": 0.8768,
+      "step": 14674
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0366640090942383,
+      "learning_rate": 1.2866827386066672e-06,
+      "loss": 0.5232,
+      "step": 14675
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7474634647369385,
+      "learning_rate": 1.2857713693084272e-06,
+      "loss": 0.9478,
+      "step": 14676
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0166056156158447,
+      "learning_rate": 1.2848603007138772e-06,
+      "loss": 0.9165,
+      "step": 14677
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6873531341552734,
+      "learning_rate": 1.2839495328544515e-06,
+      "loss": 0.9043,
+      "step": 14678
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9208621978759766,
+      "learning_rate": 1.2830390657615821e-06,
+      "loss": 0.8333,
+      "step": 14679
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8275736570358276,
+      "learning_rate": 1.2821288994666824e-06,
+      "loss": 0.8807,
+      "step": 14680
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.805232286453247,
+      "learning_rate": 1.2812190340011654e-06,
+      "loss": 0.938,
+      "step": 14681
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9696357250213623,
+      "learning_rate": 1.2803094693964214e-06,
+      "loss": 0.9082,
+      "step": 14682
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.716220736503601,
+      "learning_rate": 1.2794002056838417e-06,
+      "loss": 0.8086,
+      "step": 14683
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7015950679779053,
+      "learning_rate": 1.2784912428947994e-06,
+      "loss": 0.9517,
+      "step": 14684
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.683508276939392,
+      "learning_rate": 1.2775825810606635e-06,
+      "loss": 0.8513,
+      "step": 14685
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.800771951675415,
+      "learning_rate": 1.2766742202127858e-06,
+      "loss": 0.9126,
+      "step": 14686
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7635225057601929,
+      "learning_rate": 1.2757661603825133e-06,
+      "loss": 0.907,
+      "step": 14687
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7570717334747314,
+      "learning_rate": 1.2748584016011834e-06,
+      "loss": 0.9581,
+      "step": 14688
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8386635780334473,
+      "learning_rate": 1.273950943900114e-06,
+      "loss": 0.8998,
+      "step": 14689
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9459898471832275,
+      "learning_rate": 1.273043787310625e-06,
+      "loss": 0.8974,
+      "step": 14690
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7218345403671265,
+      "learning_rate": 1.2721369318640142e-06,
+      "loss": 0.9154,
+      "step": 14691
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.892529010772705,
+      "learning_rate": 1.2712303775915803e-06,
+      "loss": 0.8184,
+      "step": 14692
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7933043241500854,
+      "learning_rate": 1.2703241245246012e-06,
+      "loss": 0.886,
+      "step": 14693
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7797266244888306,
+      "learning_rate": 1.2694181726943533e-06,
+      "loss": 0.9099,
+      "step": 14694
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0341979265213013,
+      "learning_rate": 1.2685125221320915e-06,
+      "loss": 0.4811,
+      "step": 14695
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7987921237945557,
+      "learning_rate": 1.2676071728690765e-06,
+      "loss": 0.8194,
+      "step": 14696
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7758455276489258,
+      "learning_rate": 1.2667021249365442e-06,
+      "loss": 0.9164,
+      "step": 14697
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9008455276489258,
+      "learning_rate": 1.2657973783657262e-06,
+      "loss": 0.9224,
+      "step": 14698
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9163676500320435,
+      "learning_rate": 1.2648929331878423e-06,
+      "loss": 0.9543,
+      "step": 14699
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7035263776779175,
+      "learning_rate": 1.2639887894341042e-06,
+      "loss": 0.9427,
+      "step": 14700
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7910116910934448,
+      "learning_rate": 1.2630849471357075e-06,
+      "loss": 0.9357,
+      "step": 14701
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7919225692749023,
+      "learning_rate": 1.2621814063238457e-06,
+      "loss": 0.8516,
+      "step": 14702
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8103469610214233,
+      "learning_rate": 1.2612781670296936e-06,
+      "loss": 0.9023,
+      "step": 14703
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7206581830978394,
+      "learning_rate": 1.2603752292844219e-06,
+      "loss": 0.7913,
+      "step": 14704
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0068676471710205,
+      "learning_rate": 1.2594725931191898e-06,
+      "loss": 0.5049,
+      "step": 14705
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8440759181976318,
+      "learning_rate": 1.2585702585651404e-06,
+      "loss": 0.8608,
+      "step": 14706
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9100841283798218,
+      "learning_rate": 1.2576682256534144e-06,
+      "loss": 0.8481,
+      "step": 14707
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.646077275276184,
+      "learning_rate": 1.256766494415137e-06,
+      "loss": 0.9455,
+      "step": 14708
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7600090503692627,
+      "learning_rate": 1.2558650648814253e-06,
+      "loss": 0.9275,
+      "step": 14709
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6959959268569946,
+      "learning_rate": 1.2549639370833832e-06,
+      "loss": 0.9367,
+      "step": 14710
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7412863969802856,
+      "learning_rate": 1.2540631110521085e-06,
+      "loss": 0.9544,
+      "step": 14711
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.886684536933899,
+      "learning_rate": 1.2531625868186835e-06,
+      "loss": 0.9061,
+      "step": 14712
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8368618488311768,
+      "learning_rate": 1.2522623644141863e-06,
+      "loss": 0.8694,
+      "step": 14713
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7116056680679321,
+      "learning_rate": 1.2513624438696782e-06,
+      "loss": 0.9193,
+      "step": 14714
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7453137636184692,
+      "learning_rate": 1.2504628252162143e-06,
+      "loss": 0.7886,
+      "step": 14715
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8006501197814941,
+      "learning_rate": 1.2495635084848356e-06,
+      "loss": 0.9402,
+      "step": 14716
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0127891302108765,
+      "learning_rate": 1.2486644937065774e-06,
+      "loss": 0.5225,
+      "step": 14717
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7752655744552612,
+      "learning_rate": 1.2477657809124632e-06,
+      "loss": 0.923,
+      "step": 14718
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9243180751800537,
+      "learning_rate": 1.2468673701335022e-06,
+      "loss": 0.9176,
+      "step": 14719
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6877802610397339,
+      "learning_rate": 1.245969261400699e-06,
+      "loss": 0.8868,
+      "step": 14720
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7470992803573608,
+      "learning_rate": 1.2450714547450414e-06,
+      "loss": 0.8554,
+      "step": 14721
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6644604206085205,
+      "learning_rate": 1.2441739501975137e-06,
+      "loss": 0.8545,
+      "step": 14722
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6703284978866577,
+      "learning_rate": 1.2432767477890828e-06,
+      "loss": 0.8596,
+      "step": 14723
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8213419914245605,
+      "learning_rate": 1.242379847550712e-06,
+      "loss": 0.9732,
+      "step": 14724
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9877550601959229,
+      "learning_rate": 1.2414832495133477e-06,
+      "loss": 0.9412,
+      "step": 14725
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0882115364074707,
+      "learning_rate": 1.2405869537079317e-06,
+      "loss": 0.8787,
+      "step": 14726
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7571953535079956,
+      "learning_rate": 1.2396909601653906e-06,
+      "loss": 0.7984,
+      "step": 14727
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8994741439819336,
+      "learning_rate": 1.2387952689166426e-06,
+      "loss": 0.9049,
+      "step": 14728
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7688002586364746,
+      "learning_rate": 1.237899879992599e-06,
+      "loss": 0.8589,
+      "step": 14729
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.685086965560913,
+      "learning_rate": 1.2370047934241525e-06,
+      "loss": 0.9536,
+      "step": 14730
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7471513748168945,
+      "learning_rate": 1.2361100092421941e-06,
+      "loss": 0.9025,
+      "step": 14731
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8116681575775146,
+      "learning_rate": 1.2352155274775967e-06,
+      "loss": 0.9751,
+      "step": 14732
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9792311787605286,
+      "learning_rate": 1.2343213481612293e-06,
+      "loss": 0.472,
+      "step": 14733
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9582418203353882,
+      "learning_rate": 1.2334274713239447e-06,
+      "loss": 0.4785,
+      "step": 14734
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8422155380249023,
+      "learning_rate": 1.2325338969965916e-06,
+      "loss": 0.8384,
+      "step": 14735
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8625606298446655,
+      "learning_rate": 1.2316406252100011e-06,
+      "loss": 0.8972,
+      "step": 14736
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8275922536849976,
+      "learning_rate": 1.2307476559950004e-06,
+      "loss": 0.894,
+      "step": 14737
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7536301612854004,
+      "learning_rate": 1.2298549893824008e-06,
+      "loss": 0.8512,
+      "step": 14738
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0371437072753906,
+      "learning_rate": 1.2289626254030084e-06,
+      "loss": 0.9631,
+      "step": 14739
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9717128276824951,
+      "learning_rate": 1.2280705640876134e-06,
+      "loss": 0.8285,
+      "step": 14740
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7757149934768677,
+      "learning_rate": 1.2271788054669997e-06,
+      "loss": 0.864,
+      "step": 14741
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7596187591552734,
+      "learning_rate": 1.2262873495719418e-06,
+      "loss": 0.9644,
+      "step": 14742
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7524757385253906,
+      "learning_rate": 1.2253961964331973e-06,
+      "loss": 0.8478,
+      "step": 14743
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.780735969543457,
+      "learning_rate": 1.2245053460815204e-06,
+      "loss": 0.9573,
+      "step": 14744
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8444665670394897,
+      "learning_rate": 1.22361479854765e-06,
+      "loss": 0.8974,
+      "step": 14745
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7797430753707886,
+      "learning_rate": 1.2227245538623178e-06,
+      "loss": 0.8409,
+      "step": 14746
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7705284357070923,
+      "learning_rate": 1.2218346120562407e-06,
+      "loss": 0.8893,
+      "step": 14747
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8666045665740967,
+      "learning_rate": 1.220944973160133e-06,
+      "loss": 0.8328,
+      "step": 14748
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.661313772201538,
+      "learning_rate": 1.220055637204689e-06,
+      "loss": 0.8439,
+      "step": 14749
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7973288297653198,
+      "learning_rate": 1.2191666042206007e-06,
+      "loss": 0.9281,
+      "step": 14750
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7284934520721436,
+      "learning_rate": 1.2182778742385438e-06,
+      "loss": 0.9466,
+      "step": 14751
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.858979344367981,
+      "learning_rate": 1.2173894472891857e-06,
+      "loss": 0.9179,
+      "step": 14752
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7412071228027344,
+      "learning_rate": 1.2165013234031864e-06,
+      "loss": 0.9106,
+      "step": 14753
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.856900930404663,
+      "learning_rate": 1.2156135026111892e-06,
+      "loss": 0.9255,
+      "step": 14754
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7105193138122559,
+      "learning_rate": 1.2147259849438342e-06,
+      "loss": 0.9548,
+      "step": 14755
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.759567141532898,
+      "learning_rate": 1.2138387704317422e-06,
+      "loss": 0.8662,
+      "step": 14756
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9352344274520874,
+      "learning_rate": 1.2129518591055323e-06,
+      "loss": 0.8995,
+      "step": 14757
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7994494438171387,
+      "learning_rate": 1.2120652509958075e-06,
+      "loss": 0.9484,
+      "step": 14758
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7424341440200806,
+      "learning_rate": 1.2111789461331646e-06,
+      "loss": 0.9011,
+      "step": 14759
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7624995708465576,
+      "learning_rate": 1.2102929445481827e-06,
+      "loss": 0.9092,
+      "step": 14760
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8309754133224487,
+      "learning_rate": 1.209407246271439e-06,
+      "loss": 0.9086,
+      "step": 14761
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7244170904159546,
+      "learning_rate": 1.208521851333495e-06,
+      "loss": 0.9141,
+      "step": 14762
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6323603391647339,
+      "learning_rate": 1.2076367597649075e-06,
+      "loss": 0.8591,
+      "step": 14763
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7281302213668823,
+      "learning_rate": 1.2067519715962116e-06,
+      "loss": 0.8696,
+      "step": 14764
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9572097063064575,
+      "learning_rate": 1.2058674868579446e-06,
+      "loss": 0.8585,
+      "step": 14765
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9160422086715698,
+      "learning_rate": 1.2049833055806227e-06,
+      "loss": 0.9354,
+      "step": 14766
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7224925756454468,
+      "learning_rate": 1.2040994277947615e-06,
+      "loss": 0.8782,
+      "step": 14767
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.69132661819458,
+      "learning_rate": 1.203215853530857e-06,
+      "loss": 0.8399,
+      "step": 14768
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8126194477081299,
+      "learning_rate": 1.202332582819402e-06,
+      "loss": 0.9485,
+      "step": 14769
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7348047494888306,
+      "learning_rate": 1.2014496156908728e-06,
+      "loss": 0.9025,
+      "step": 14770
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8124852180480957,
+      "learning_rate": 1.200566952175739e-06,
+      "loss": 0.9332,
+      "step": 14771
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7152577638626099,
+      "learning_rate": 1.199684592304462e-06,
+      "loss": 0.8171,
+      "step": 14772
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6856400966644287,
+      "learning_rate": 1.198802536107484e-06,
+      "loss": 0.9956,
+      "step": 14773
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9402037858963013,
+      "learning_rate": 1.1979207836152484e-06,
+      "loss": 0.8895,
+      "step": 14774
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.668998122215271,
+      "learning_rate": 1.1970393348581766e-06,
+      "loss": 0.8255,
+      "step": 14775
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7074435949325562,
+      "learning_rate": 1.1961581898666895e-06,
+      "loss": 0.9107,
+      "step": 14776
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6808173656463623,
+      "learning_rate": 1.195277348671189e-06,
+      "loss": 0.7387,
+      "step": 14777
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7139525413513184,
+      "learning_rate": 1.1943968113020733e-06,
+      "loss": 0.8979,
+      "step": 14778
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7274811267852783,
+      "learning_rate": 1.193516577789725e-06,
+      "loss": 0.903,
+      "step": 14779
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6515867710113525,
+      "learning_rate": 1.1926366481645213e-06,
+      "loss": 0.9149,
+      "step": 14780
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8225899934768677,
+      "learning_rate": 1.191757022456822e-06,
+      "loss": 0.91,
+      "step": 14781
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.883935570716858,
+      "learning_rate": 1.1908777006969841e-06,
+      "loss": 0.8197,
+      "step": 14782
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7968236207962036,
+      "learning_rate": 1.189998682915351e-06,
+      "loss": 0.8987,
+      "step": 14783
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7596200704574585,
+      "learning_rate": 1.1891199691422517e-06,
+      "loss": 0.9379,
+      "step": 14784
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8464593887329102,
+      "learning_rate": 1.1882415594080111e-06,
+      "loss": 0.8107,
+      "step": 14785
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6986284255981445,
+      "learning_rate": 1.187363453742939e-06,
+      "loss": 0.8749,
+      "step": 14786
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7600467205047607,
+      "learning_rate": 1.1864856521773382e-06,
+      "loss": 0.8687,
+      "step": 14787
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7135143280029297,
+      "learning_rate": 1.1856081547414965e-06,
+      "loss": 0.9506,
+      "step": 14788
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9230210781097412,
+      "learning_rate": 1.1847309614656966e-06,
+      "loss": 0.8943,
+      "step": 14789
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.797799825668335,
+      "learning_rate": 1.183854072380205e-06,
+      "loss": 0.8831,
+      "step": 14790
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7772151231765747,
+      "learning_rate": 1.1829774875152854e-06,
+      "loss": 0.857,
+      "step": 14791
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6105971336364746,
+      "learning_rate": 1.1821012069011806e-06,
+      "loss": 0.9017,
+      "step": 14792
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.744730830192566,
+      "learning_rate": 1.1812252305681326e-06,
+      "loss": 0.8856,
+      "step": 14793
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8852442502975464,
+      "learning_rate": 1.1803495585463665e-06,
+      "loss": 0.9199,
+      "step": 14794
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.849757432937622,
+      "learning_rate": 1.1794741908661012e-06,
+      "loss": 0.8482,
+      "step": 14795
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7568823099136353,
+      "learning_rate": 1.1785991275575426e-06,
+      "loss": 0.8506,
+      "step": 14796
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8058956861495972,
+      "learning_rate": 1.1777243686508854e-06,
+      "loss": 0.8276,
+      "step": 14797
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7867754697799683,
+      "learning_rate": 1.176849914176319e-06,
+      "loss": 0.8471,
+      "step": 14798
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7113209962844849,
+      "learning_rate": 1.1759757641640125e-06,
+      "loss": 0.876,
+      "step": 14799
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6810534000396729,
+      "learning_rate": 1.175101918644136e-06,
+      "loss": 0.9932,
+      "step": 14800
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7239222526550293,
+      "learning_rate": 1.1742283776468389e-06,
+      "loss": 0.8098,
+      "step": 14801
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8601396083831787,
+      "learning_rate": 1.1733551412022682e-06,
+      "loss": 0.8095,
+      "step": 14802
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.791898488998413,
+      "learning_rate": 1.1724822093405542e-06,
+      "loss": 0.859,
+      "step": 14803
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9741474390029907,
+      "learning_rate": 1.1716095820918217e-06,
+      "loss": 0.5525,
+      "step": 14804
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8405451774597168,
+      "learning_rate": 1.17073725948618e-06,
+      "loss": 0.9578,
+      "step": 14805
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.761673092842102,
+      "learning_rate": 1.1698652415537315e-06,
+      "loss": 1.0171,
+      "step": 14806
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7625985145568848,
+      "learning_rate": 1.16899352832457e-06,
+      "loss": 0.8912,
+      "step": 14807
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9864790439605713,
+      "learning_rate": 1.1681221198287707e-06,
+      "loss": 0.8616,
+      "step": 14808
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.670388102531433,
+      "learning_rate": 1.167251016096409e-06,
+      "loss": 0.8667,
+      "step": 14809
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5771379470825195,
+      "learning_rate": 1.166380217157539e-06,
+      "loss": 0.9119,
+      "step": 14810
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.735488772392273,
+      "learning_rate": 1.1655097230422141e-06,
+      "loss": 0.8951,
+      "step": 14811
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0001025199890137,
+      "learning_rate": 1.1646395337804684e-06,
+      "loss": 0.4815,
+      "step": 14812
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9586142301559448,
+      "learning_rate": 1.1637696494023331e-06,
+      "loss": 0.8887,
+      "step": 14813
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7583764791488647,
+      "learning_rate": 1.1629000699378235e-06,
+      "loss": 0.8856,
+      "step": 14814
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8681720495224,
+      "learning_rate": 1.1620307954169484e-06,
+      "loss": 0.8413,
+      "step": 14815
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7538697719573975,
+      "learning_rate": 1.1611618258696999e-06,
+      "loss": 0.9364,
+      "step": 14816
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7741611003875732,
+      "learning_rate": 1.1602931613260694e-06,
+      "loss": 0.9007,
+      "step": 14817
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1062372922897339,
+      "learning_rate": 1.159424801816027e-06,
+      "loss": 0.5678,
+      "step": 14818
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7400236129760742,
+      "learning_rate": 1.1585567473695403e-06,
+      "loss": 0.8709,
+      "step": 14819
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7755361795425415,
+      "learning_rate": 1.157688998016564e-06,
+      "loss": 0.8898,
+      "step": 14820
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6871265172958374,
+      "learning_rate": 1.1568215537870376e-06,
+      "loss": 0.9804,
+      "step": 14821
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7692261934280396,
+      "learning_rate": 1.1559544147109004e-06,
+      "loss": 0.8133,
+      "step": 14822
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7094820737838745,
+      "learning_rate": 1.1550875808180685e-06,
+      "loss": 0.8723,
+      "step": 14823
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6173863410949707,
+      "learning_rate": 1.154221052138459e-06,
+      "loss": 0.8899,
+      "step": 14824
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7471132278442383,
+      "learning_rate": 1.1533548287019702e-06,
+      "loss": 0.8852,
+      "step": 14825
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7408438920974731,
+      "learning_rate": 1.152488910538495e-06,
+      "loss": 0.8076,
+      "step": 14826
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9347789287567139,
+      "learning_rate": 1.1516232976779095e-06,
+      "loss": 0.8943,
+      "step": 14827
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7290682792663574,
+      "learning_rate": 1.150757990150091e-06,
+      "loss": 0.8012,
+      "step": 14828
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7400461435317993,
+      "learning_rate": 1.149892987984893e-06,
+      "loss": 0.8784,
+      "step": 14829
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8327698707580566,
+      "learning_rate": 1.1490282912121686e-06,
+      "loss": 0.9392,
+      "step": 14830
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7748452425003052,
+      "learning_rate": 1.1481638998617507e-06,
+      "loss": 0.865,
+      "step": 14831
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8569293022155762,
+      "learning_rate": 1.1472998139634727e-06,
+      "loss": 0.853,
+      "step": 14832
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7124027013778687,
+      "learning_rate": 1.146436033547147e-06,
+      "loss": 0.8749,
+      "step": 14833
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6331645250320435,
+      "learning_rate": 1.1455725586425847e-06,
+      "loss": 0.8735,
+      "step": 14834
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0667500495910645,
+      "learning_rate": 1.1447093892795769e-06,
+      "loss": 0.9016,
+      "step": 14835
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7744994163513184,
+      "learning_rate": 1.1438465254879116e-06,
+      "loss": 0.9426,
+      "step": 14836
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7284291982650757,
+      "learning_rate": 1.1429839672973665e-06,
+      "loss": 0.8813,
+      "step": 14837
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9361088275909424,
+      "learning_rate": 1.1421217147377018e-06,
+      "loss": 0.8769,
+      "step": 14838
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8605183362960815,
+      "learning_rate": 1.141259767838675e-06,
+      "loss": 0.8413,
+      "step": 14839
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9905667304992676,
+      "learning_rate": 1.1403981266300258e-06,
+      "loss": 0.5687,
+      "step": 14840
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8911683559417725,
+      "learning_rate": 1.1395367911414911e-06,
+      "loss": 0.9432,
+      "step": 14841
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.788665771484375,
+      "learning_rate": 1.1386757614027888e-06,
+      "loss": 0.8168,
+      "step": 14842
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8488212823867798,
+      "learning_rate": 1.1378150374436347e-06,
+      "loss": 0.9767,
+      "step": 14843
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6769204139709473,
+      "learning_rate": 1.1369546192937264e-06,
+      "loss": 0.8694,
+      "step": 14844
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.801372766494751,
+      "learning_rate": 1.136094506982759e-06,
+      "loss": 0.9206,
+      "step": 14845
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9626538157463074,
+      "learning_rate": 1.1352347005404062e-06,
+      "loss": 0.5129,
+      "step": 14846
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9301203489303589,
+      "learning_rate": 1.1343751999963448e-06,
+      "loss": 0.9109,
+      "step": 14847
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8862485885620117,
+      "learning_rate": 1.1335160053802273e-06,
+      "loss": 0.959,
+      "step": 14848
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8061212301254272,
+      "learning_rate": 1.132657116721705e-06,
+      "loss": 0.9016,
+      "step": 14849
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.064295530319214,
+      "learning_rate": 1.1317985340504178e-06,
+      "loss": 0.9604,
+      "step": 14850
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9635558128356934,
+      "learning_rate": 1.1309402573959882e-06,
+      "loss": 0.8797,
+      "step": 14851
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7758803367614746,
+      "learning_rate": 1.1300822867880378e-06,
+      "loss": 0.8734,
+      "step": 14852
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.030975341796875,
+      "learning_rate": 1.1292246222561697e-06,
+      "loss": 0.9173,
+      "step": 14853
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7642524242401123,
+      "learning_rate": 1.1283672638299813e-06,
+      "loss": 0.8545,
+      "step": 14854
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7300866842269897,
+      "learning_rate": 1.1275102115390546e-06,
+      "loss": 1.0123,
+      "step": 14855
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7304534912109375,
+      "learning_rate": 1.126653465412969e-06,
+      "loss": 0.9208,
+      "step": 14856
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6619575023651123,
+      "learning_rate": 1.1257970254812833e-06,
+      "loss": 0.8994,
+      "step": 14857
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7392650842666626,
+      "learning_rate": 1.124940891773555e-06,
+      "loss": 0.8464,
+      "step": 14858
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7891305685043335,
+      "learning_rate": 1.1240850643193236e-06,
+      "loss": 0.8295,
+      "step": 14859
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7653645277023315,
+      "learning_rate": 1.1232295431481222e-06,
+      "loss": 0.9218,
+      "step": 14860
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6638380289077759,
+      "learning_rate": 1.122374328289475e-06,
+      "loss": 0.9276,
+      "step": 14861
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6876366138458252,
+      "learning_rate": 1.1215194197728886e-06,
+      "loss": 0.8602,
+      "step": 14862
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6116549968719482,
+      "learning_rate": 1.120664817627869e-06,
+      "loss": 0.8931,
+      "step": 14863
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.932853102684021,
+      "learning_rate": 1.1198105218839007e-06,
+      "loss": 0.9732,
+      "step": 14864
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8430944681167603,
+      "learning_rate": 1.1189565325704677e-06,
+      "loss": 0.8903,
+      "step": 14865
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8514463901519775,
+      "learning_rate": 1.1181028497170344e-06,
+      "loss": 0.8543,
+      "step": 14866
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6708588600158691,
+      "learning_rate": 1.1172494733530625e-06,
+      "loss": 0.8563,
+      "step": 14867
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7908377647399902,
+      "learning_rate": 1.1163964035079976e-06,
+      "loss": 0.9798,
+      "step": 14868
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8075577020645142,
+      "learning_rate": 1.1155436402112785e-06,
+      "loss": 0.9929,
+      "step": 14869
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6757240295410156,
+      "learning_rate": 1.114691183492329e-06,
+      "loss": 0.8865,
+      "step": 14870
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.799923062324524,
+      "learning_rate": 1.1138390333805682e-06,
+      "loss": 0.959,
+      "step": 14871
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6274442672729492,
+      "learning_rate": 1.112987189905399e-06,
+      "loss": 0.8843,
+      "step": 14872
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0177228450775146,
+      "learning_rate": 1.1121356530962157e-06,
+      "loss": 0.9739,
+      "step": 14873
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6361910104751587,
+      "learning_rate": 1.1112844229824071e-06,
+      "loss": 0.868,
+      "step": 14874
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9618886709213257,
+      "learning_rate": 1.1104334995933407e-06,
+      "loss": 0.8465,
+      "step": 14875
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7846596240997314,
+      "learning_rate": 1.1095828829583844e-06,
+      "loss": 0.8327,
+      "step": 14876
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6273856163024902,
+      "learning_rate": 1.1087325731068854e-06,
+      "loss": 0.8913,
+      "step": 14877
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8435790538787842,
+      "learning_rate": 1.1078825700681918e-06,
+      "loss": 0.8831,
+      "step": 14878
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0711643695831299,
+      "learning_rate": 1.1070328738716285e-06,
+      "loss": 0.5418,
+      "step": 14879
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8297356367111206,
+      "learning_rate": 1.1061834845465225e-06,
+      "loss": 0.9101,
+      "step": 14880
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8431520462036133,
+      "learning_rate": 1.1053344021221778e-06,
+      "loss": 0.855,
+      "step": 14881
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.582067847251892,
+      "learning_rate": 1.104485626627899e-06,
+      "loss": 0.8298,
+      "step": 14882
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.15044641494751,
+      "learning_rate": 1.1036371580929706e-06,
+      "loss": 0.8888,
+      "step": 14883
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5642304420471191,
+      "learning_rate": 1.102788996546672e-06,
+      "loss": 0.9006,
+      "step": 14884
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7663689851760864,
+      "learning_rate": 1.1019411420182747e-06,
+      "loss": 0.8598,
+      "step": 14885
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6893409490585327,
+      "learning_rate": 1.1010935945370305e-06,
+      "loss": 0.9682,
+      "step": 14886
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.782151222229004,
+      "learning_rate": 1.1002463541321906e-06,
+      "loss": 0.9291,
+      "step": 14887
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.722662091255188,
+      "learning_rate": 1.0993994208329862e-06,
+      "loss": 0.9829,
+      "step": 14888
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9793977737426758,
+      "learning_rate": 1.098552794668648e-06,
+      "loss": 0.8827,
+      "step": 14889
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7547744512557983,
+      "learning_rate": 1.0977064756683841e-06,
+      "loss": 0.8589,
+      "step": 14890
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7955750226974487,
+      "learning_rate": 1.0968604638614055e-06,
+      "loss": 0.8553,
+      "step": 14891
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8794209957122803,
+      "learning_rate": 1.096014759276899e-06,
+      "loss": 0.9655,
+      "step": 14892
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.872417688369751,
+      "learning_rate": 1.0951693619440517e-06,
+      "loss": 0.9323,
+      "step": 14893
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8672363758087158,
+      "learning_rate": 1.0943242718920355e-06,
+      "loss": 0.8697,
+      "step": 14894
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.697163462638855,
+      "learning_rate": 1.0934794891500134e-06,
+      "loss": 0.8536,
+      "step": 14895
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7430111169815063,
+      "learning_rate": 1.092635013747132e-06,
+      "loss": 0.8852,
+      "step": 14896
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8425328731536865,
+      "learning_rate": 1.091790845712537e-06,
+      "loss": 0.9685,
+      "step": 14897
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8222815990447998,
+      "learning_rate": 1.090946985075354e-06,
+      "loss": 0.9451,
+      "step": 14898
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.637719750404358,
+      "learning_rate": 1.0901034318647063e-06,
+      "loss": 0.8964,
+      "step": 14899
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6801607608795166,
+      "learning_rate": 1.0892601861096985e-06,
+      "loss": 0.9068,
+      "step": 14900
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8916633129119873,
+      "learning_rate": 1.0884172478394317e-06,
+      "loss": 1.0096,
+      "step": 14901
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7380305528640747,
+      "learning_rate": 1.0875746170829903e-06,
+      "loss": 0.9066,
+      "step": 14902
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8806736469268799,
+      "learning_rate": 1.0867322938694535e-06,
+      "loss": 0.953,
+      "step": 14903
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7062561511993408,
+      "learning_rate": 1.085890278227889e-06,
+      "loss": 0.7576,
+      "step": 14904
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8777328729629517,
+      "learning_rate": 1.085048570187348e-06,
+      "loss": 0.8432,
+      "step": 14905
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.79771089553833,
+      "learning_rate": 1.0842071697768808e-06,
+      "loss": 0.8779,
+      "step": 14906
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8964142799377441,
+      "learning_rate": 1.0833660770255162e-06,
+      "loss": 0.9333,
+      "step": 14907
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.013614535331726,
+      "learning_rate": 1.082525291962283e-06,
+      "loss": 0.5683,
+      "step": 14908
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8889468908309937,
+      "learning_rate": 1.0816848146161895e-06,
+      "loss": 0.9218,
+      "step": 14909
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.679005742073059,
+      "learning_rate": 1.0808446450162435e-06,
+      "loss": 0.8734,
+      "step": 14910
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.699961543083191,
+      "learning_rate": 1.0800047831914317e-06,
+      "loss": 0.9117,
+      "step": 14911
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7158069610595703,
+      "learning_rate": 1.07916522917074e-06,
+      "loss": 0.9132,
+      "step": 14912
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8884469270706177,
+      "learning_rate": 1.078325982983134e-06,
+      "loss": 0.8976,
+      "step": 14913
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9490313529968262,
+      "learning_rate": 1.077487044657578e-06,
+      "loss": 0.8698,
+      "step": 14914
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6767425537109375,
+      "learning_rate": 1.0766484142230215e-06,
+      "loss": 0.8781,
+      "step": 14915
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.943113923072815,
+      "learning_rate": 1.075810091708399e-06,
+      "loss": 0.9006,
+      "step": 14916
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.725237250328064,
+      "learning_rate": 1.0749720771426443e-06,
+      "loss": 0.9273,
+      "step": 14917
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7463523149490356,
+      "learning_rate": 1.0741343705546704e-06,
+      "loss": 0.883,
+      "step": 14918
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6340445280075073,
+      "learning_rate": 1.0732969719733877e-06,
+      "loss": 0.8899,
+      "step": 14919
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7823466062545776,
+      "learning_rate": 1.0724598814276887e-06,
+      "loss": 0.9085,
+      "step": 14920
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7317627668380737,
+      "learning_rate": 1.0716230989464638e-06,
+      "loss": 0.9062,
+      "step": 14921
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9632351398468018,
+      "learning_rate": 1.070786624558583e-06,
+      "loss": 0.8874,
+      "step": 14922
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8129502534866333,
+      "learning_rate": 1.0699504582929144e-06,
+      "loss": 0.9309,
+      "step": 14923
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6596673727035522,
+      "learning_rate": 1.0691146001783081e-06,
+      "loss": 0.8475,
+      "step": 14924
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.796264886856079,
+      "learning_rate": 1.0682790502436124e-06,
+      "loss": 0.8789,
+      "step": 14925
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8723270893096924,
+      "learning_rate": 1.0674438085176553e-06,
+      "loss": 0.9483,
+      "step": 14926
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5680409669876099,
+      "learning_rate": 1.066608875029259e-06,
+      "loss": 0.8487,
+      "step": 14927
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8118727207183838,
+      "learning_rate": 1.0657742498072388e-06,
+      "loss": 0.9055,
+      "step": 14928
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.996239423751831,
+      "learning_rate": 1.0649399328803912e-06,
+      "loss": 0.8977,
+      "step": 14929
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8458224534988403,
+      "learning_rate": 1.0641059242775087e-06,
+      "loss": 1.0045,
+      "step": 14930
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6864839792251587,
+      "learning_rate": 1.0632722240273662e-06,
+      "loss": 0.8899,
+      "step": 14931
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6400095224380493,
+      "learning_rate": 1.0624388321587387e-06,
+      "loss": 0.908,
+      "step": 14932
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7352604866027832,
+      "learning_rate": 1.0616057487003794e-06,
+      "loss": 0.8605,
+      "step": 14933
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.819554328918457,
+      "learning_rate": 1.060772973681039e-06,
+      "loss": 0.9217,
+      "step": 14934
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8149404525756836,
+      "learning_rate": 1.05994050712945e-06,
+      "loss": 0.8856,
+      "step": 14935
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.877528190612793,
+      "learning_rate": 1.0591083490743437e-06,
+      "loss": 0.9106,
+      "step": 14936
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8921500444412231,
+      "learning_rate": 1.0582764995444305e-06,
+      "loss": 0.937,
+      "step": 14937
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1025562286376953,
+      "learning_rate": 1.0574449585684176e-06,
+      "loss": 0.5401,
+      "step": 14938
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8973225355148315,
+      "learning_rate": 1.056613726175002e-06,
+      "loss": 0.9134,
+      "step": 14939
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8091366291046143,
+      "learning_rate": 1.0557828023928607e-06,
+      "loss": 0.8404,
+      "step": 14940
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8649598360061646,
+      "learning_rate": 1.054952187250674e-06,
+      "loss": 0.8943,
+      "step": 14941
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7879281044006348,
+      "learning_rate": 1.054121880777097e-06,
+      "loss": 0.8626,
+      "step": 14942
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7447692155838013,
+      "learning_rate": 1.0532918830007876e-06,
+      "loss": 0.9264,
+      "step": 14943
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7139854431152344,
+      "learning_rate": 1.052462193950381e-06,
+      "loss": 0.9175,
+      "step": 14944
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8053346872329712,
+      "learning_rate": 1.0516328136545129e-06,
+      "loss": 0.9155,
+      "step": 14945
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7508553266525269,
+      "learning_rate": 1.0508037421417971e-06,
+      "loss": 0.8043,
+      "step": 14946
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.010089159011841,
+      "learning_rate": 1.0499749794408475e-06,
+      "loss": 0.9632,
+      "step": 14947
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6463533639907837,
+      "learning_rate": 1.0491465255802603e-06,
+      "loss": 0.9299,
+      "step": 14948
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7881395816802979,
+      "learning_rate": 1.0483183805886233e-06,
+      "loss": 1.0043,
+      "step": 14949
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.694732666015625,
+      "learning_rate": 1.0474905444945128e-06,
+      "loss": 0.8883,
+      "step": 14950
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6043202877044678,
+      "learning_rate": 1.0466630173264946e-06,
+      "loss": 0.9465,
+      "step": 14951
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.818852424621582,
+      "learning_rate": 1.0458357991131284e-06,
+      "loss": 0.9115,
+      "step": 14952
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7274229526519775,
+      "learning_rate": 1.0450088898829547e-06,
+      "loss": 0.9153,
+      "step": 14953
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7107564210891724,
+      "learning_rate": 1.0441822896645104e-06,
+      "loss": 0.8062,
+      "step": 14954
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7387783527374268,
+      "learning_rate": 1.0433559984863162e-06,
+      "loss": 0.8776,
+      "step": 14955
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5985060930252075,
+      "learning_rate": 1.0425300163768903e-06,
+      "loss": 0.9396,
+      "step": 14956
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0059682130813599,
+      "learning_rate": 1.0417043433647289e-06,
+      "loss": 0.4969,
+      "step": 14957
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9177589416503906,
+      "learning_rate": 1.0408789794783292e-06,
+      "loss": 0.9313,
+      "step": 14958
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8948943614959717,
+      "learning_rate": 1.040053924746165e-06,
+      "loss": 0.8487,
+      "step": 14959
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4074482917785645,
+      "learning_rate": 1.0392291791967158e-06,
+      "loss": 0.8925,
+      "step": 14960
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.882422685623169,
+      "learning_rate": 1.0384047428584344e-06,
+      "loss": 0.9122,
+      "step": 14961
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7333823442459106,
+      "learning_rate": 1.0375806157597734e-06,
+      "loss": 0.9573,
+      "step": 14962
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8123652935028076,
+      "learning_rate": 1.0367567979291694e-06,
+      "loss": 0.8534,
+      "step": 14963
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8629822731018066,
+      "learning_rate": 1.0359332893950514e-06,
+      "loss": 0.9177,
+      "step": 14964
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5997167825698853,
+      "learning_rate": 1.0351100901858335e-06,
+      "loss": 0.8768,
+      "step": 14965
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8188073635101318,
+      "learning_rate": 1.0342872003299265e-06,
+      "loss": 0.967,
+      "step": 14966
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.846309781074524,
+      "learning_rate": 1.0334646198557208e-06,
+      "loss": 0.95,
+      "step": 14967
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.982743501663208,
+      "learning_rate": 1.0326423487916048e-06,
+      "loss": 0.8022,
+      "step": 14968
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7773852348327637,
+      "learning_rate": 1.0318203871659538e-06,
+      "loss": 0.8311,
+      "step": 14969
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1019771099090576,
+      "learning_rate": 1.0309987350071281e-06,
+      "loss": 0.6092,
+      "step": 14970
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.590195655822754,
+      "learning_rate": 1.0301773923434833e-06,
+      "loss": 0.904,
+      "step": 14971
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.666215419769287,
+      "learning_rate": 1.0293563592033595e-06,
+      "loss": 0.922,
+      "step": 14972
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8345155715942383,
+      "learning_rate": 1.0285356356150899e-06,
+      "loss": 0.8458,
+      "step": 14973
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.841991901397705,
+      "learning_rate": 1.0277152216069942e-06,
+      "loss": 0.923,
+      "step": 14974
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9013420343399048,
+      "learning_rate": 1.0268951172073838e-06,
+      "loss": 0.9206,
+      "step": 14975
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.766484260559082,
+      "learning_rate": 1.0260753224445564e-06,
+      "loss": 0.8579,
+      "step": 14976
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6873103380203247,
+      "learning_rate": 1.0252558373468036e-06,
+      "loss": 0.8281,
+      "step": 14977
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8229858875274658,
+      "learning_rate": 1.0244366619424006e-06,
+      "loss": 0.9006,
+      "step": 14978
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6585370302200317,
+      "learning_rate": 1.0236177962596173e-06,
+      "loss": 0.8202,
+      "step": 14979
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9619442224502563,
+      "learning_rate": 1.0227992403267074e-06,
+      "loss": 0.529,
+      "step": 14980
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.014090061187744,
+      "learning_rate": 1.0219809941719195e-06,
+      "loss": 0.9412,
+      "step": 14981
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0536394119262695,
+      "learning_rate": 1.0211630578234899e-06,
+      "loss": 0.8079,
+      "step": 14982
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6927961111068726,
+      "learning_rate": 1.0203454313096407e-06,
+      "loss": 0.8818,
+      "step": 14983
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.124111294746399,
+      "learning_rate": 1.0195281146585879e-06,
+      "loss": 0.6468,
+      "step": 14984
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8876206874847412,
+      "learning_rate": 1.0187111078985324e-06,
+      "loss": 1.0668,
+      "step": 14985
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.752833604812622,
+      "learning_rate": 1.0178944110576704e-06,
+      "loss": 0.8663,
+      "step": 14986
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6628209352493286,
+      "learning_rate": 1.0170780241641798e-06,
+      "loss": 0.8837,
+      "step": 14987
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7642520666122437,
+      "learning_rate": 1.0162619472462355e-06,
+      "loss": 0.8182,
+      "step": 14988
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.793492078781128,
+      "learning_rate": 1.0154461803319938e-06,
+      "loss": 0.8995,
+      "step": 14989
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7705007791519165,
+      "learning_rate": 1.01463072344961e-06,
+      "loss": 0.8643,
+      "step": 14990
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9232949018478394,
+      "learning_rate": 1.0138155766272185e-06,
+      "loss": 0.9302,
+      "step": 14991
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9237415790557861,
+      "learning_rate": 1.0130007398929486e-06,
+      "loss": 0.9553,
+      "step": 14992
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7248718738555908,
+      "learning_rate": 1.0121862132749216e-06,
+      "loss": 0.8873,
+      "step": 14993
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6651725769042969,
+      "learning_rate": 1.0113719968012403e-06,
+      "loss": 0.889,
+      "step": 14994
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.003671169281006,
+      "learning_rate": 1.0105580905000045e-06,
+      "loss": 0.8715,
+      "step": 14995
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5913702249526978,
+      "learning_rate": 1.009744494399295e-06,
+      "loss": 0.8059,
+      "step": 14996
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9233933687210083,
+      "learning_rate": 1.008931208527193e-06,
+      "loss": 0.9002,
+      "step": 14997
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8158527612686157,
+      "learning_rate": 1.0081182329117566e-06,
+      "loss": 0.8847,
+      "step": 14998
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7748826742172241,
+      "learning_rate": 1.007305567581045e-06,
+      "loss": 0.8688,
+      "step": 14999
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8594638109207153,
+      "learning_rate": 1.0064932125630956e-06,
+      "loss": 0.8767,
+      "step": 15000
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7900792360305786,
+      "learning_rate": 1.0056811678859458e-06,
+      "loss": 0.8046,
+      "step": 15001
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8237502574920654,
+      "learning_rate": 1.0048694335776111e-06,
+      "loss": 0.8702,
+      "step": 15002
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8914153575897217,
+      "learning_rate": 1.0040580096661079e-06,
+      "loss": 0.9477,
+      "step": 15003
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7254283428192139,
+      "learning_rate": 1.0032468961794317e-06,
+      "loss": 0.8188,
+      "step": 15004
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8993984460830688,
+      "learning_rate": 1.0024360931455735e-06,
+      "loss": 0.8996,
+      "step": 15005
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8457854986190796,
+      "learning_rate": 1.0016256005925152e-06,
+      "loss": 0.9287,
+      "step": 15006
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9410896301269531,
+      "learning_rate": 1.0008154185482178e-06,
+      "loss": 0.9062,
+      "step": 15007
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0482933521270752,
+      "learning_rate": 1.0000055470406445e-06,
+      "loss": 0.4759,
+      "step": 15008
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9075489044189453,
+      "learning_rate": 9.991959860977384e-07,
+      "loss": 0.8627,
+      "step": 15009
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7341303825378418,
+      "learning_rate": 9.983867357474374e-07,
+      "loss": 0.955,
+      "step": 15010
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.688016414642334,
+      "learning_rate": 9.975777960176625e-07,
+      "loss": 0.835,
+      "step": 15011
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.80156409740448,
+      "learning_rate": 9.967691669363334e-07,
+      "loss": 0.8995,
+      "step": 15012
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6473296880722046,
+      "learning_rate": 9.959608485313488e-07,
+      "loss": 0.9735,
+      "step": 15013
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7560105323791504,
+      "learning_rate": 9.951528408306054e-07,
+      "loss": 0.9356,
+      "step": 15014
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7326314449310303,
+      "learning_rate": 9.94345143861981e-07,
+      "loss": 0.9622,
+      "step": 15015
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6863282918930054,
+      "learning_rate": 9.935377576533523e-07,
+      "loss": 0.864,
+      "step": 15016
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7707799673080444,
+      "learning_rate": 9.927306822325745e-07,
+      "loss": 0.9043,
+      "step": 15017
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7395671606063843,
+      "learning_rate": 9.919239176274998e-07,
+      "loss": 0.9411,
+      "step": 15018
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.670168399810791,
+      "learning_rate": 9.911174638659703e-07,
+      "loss": 0.9057,
+      "step": 15019
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8225750923156738,
+      "learning_rate": 9.903113209758098e-07,
+      "loss": 0.9408,
+      "step": 15020
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8781756162643433,
+      "learning_rate": 9.895054889848389e-07,
+      "loss": 0.9022,
+      "step": 15021
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8368732929229736,
+      "learning_rate": 9.88699967920863e-07,
+      "loss": 0.8923,
+      "step": 15022
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8686180114746094,
+      "learning_rate": 9.878947578116804e-07,
+      "loss": 0.9404,
+      "step": 15023
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8079357147216797,
+      "learning_rate": 9.870898586850742e-07,
+      "loss": 0.9248,
+      "step": 15024
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8096425533294678,
+      "learning_rate": 9.862852705688198e-07,
+      "loss": 0.8643,
+      "step": 15025
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7631109952926636,
+      "learning_rate": 9.85480993490683e-07,
+      "loss": 0.9202,
+      "step": 15026
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.845440149307251,
+      "learning_rate": 9.846770274784168e-07,
+      "loss": 0.9432,
+      "step": 15027
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7919163703918457,
+      "learning_rate": 9.838733725597615e-07,
+      "loss": 0.9022,
+      "step": 15028
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8412429094314575,
+      "learning_rate": 9.830700287624528e-07,
+      "loss": 0.8272,
+      "step": 15029
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7344106435775757,
+      "learning_rate": 9.822669961142074e-07,
+      "loss": 0.8765,
+      "step": 15030
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6600439548492432,
+      "learning_rate": 9.814642746427394e-07,
+      "loss": 0.906,
+      "step": 15031
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6861951351165771,
+      "learning_rate": 9.806618643757459e-07,
+      "loss": 0.9848,
+      "step": 15032
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.814062476158142,
+      "learning_rate": 9.79859765340918e-07,
+      "loss": 0.8422,
+      "step": 15033
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.857266902923584,
+      "learning_rate": 9.790579775659326e-07,
+      "loss": 0.9027,
+      "step": 15034
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6585924625396729,
+      "learning_rate": 9.78256501078456e-07,
+      "loss": 0.8907,
+      "step": 15035
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.982703447341919,
+      "learning_rate": 9.77455335906149e-07,
+      "loss": 0.8893,
+      "step": 15036
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7560428380966187,
+      "learning_rate": 9.766544820766522e-07,
+      "loss": 0.9072,
+      "step": 15037
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7224732637405396,
+      "learning_rate": 9.75853939617606e-07,
+      "loss": 0.9246,
+      "step": 15038
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7476567029953003,
+      "learning_rate": 9.750537085566302e-07,
+      "loss": 0.9563,
+      "step": 15039
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7501264810562134,
+      "learning_rate": 9.74253788921342e-07,
+      "loss": 0.9091,
+      "step": 15040
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9733738899230957,
+      "learning_rate": 9.734541807393428e-07,
+      "loss": 0.8785,
+      "step": 15041
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7972707748413086,
+      "learning_rate": 9.726548840382256e-07,
+      "loss": 0.94,
+      "step": 15042
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8973783254623413,
+      "learning_rate": 9.718558988455706e-07,
+      "loss": 0.8888,
+      "step": 15043
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8386483192443848,
+      "learning_rate": 9.710572251889505e-07,
+      "loss": 0.8885,
+      "step": 15044
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8734389543533325,
+      "learning_rate": 9.70258863095923e-07,
+      "loss": 0.9636,
+      "step": 15045
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9930318593978882,
+      "learning_rate": 9.694608125940385e-07,
+      "loss": 0.5489,
+      "step": 15046
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.852248191833496,
+      "learning_rate": 9.686630737108372e-07,
+      "loss": 0.8996,
+      "step": 15047
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8760344982147217,
+      "learning_rate": 9.678656464738433e-07,
+      "loss": 0.9484,
+      "step": 15048
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8614131212234497,
+      "learning_rate": 9.670685309105786e-07,
+      "loss": 0.8531,
+      "step": 15049
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6949517726898193,
+      "learning_rate": 9.662717270485432e-07,
+      "loss": 0.802,
+      "step": 15050
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.816502571105957,
+      "learning_rate": 9.654752349152384e-07,
+      "loss": 0.893,
+      "step": 15051
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8275442123413086,
+      "learning_rate": 9.646790545381447e-07,
+      "loss": 0.9257,
+      "step": 15052
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7770590782165527,
+      "learning_rate": 9.638831859447385e-07,
+      "loss": 0.8659,
+      "step": 15053
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8152029514312744,
+      "learning_rate": 9.630876291624802e-07,
+      "loss": 0.8958,
+      "step": 15054
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7617536783218384,
+      "learning_rate": 9.62292384218827e-07,
+      "loss": 0.8945,
+      "step": 15055
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1205124855041504,
+      "learning_rate": 9.614974511412156e-07,
+      "loss": 0.8867,
+      "step": 15056
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.732620120048523,
+      "learning_rate": 9.60702829957082e-07,
+      "loss": 0.8581,
+      "step": 15057
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7106659412384033,
+      "learning_rate": 9.599085206938397e-07,
+      "loss": 0.8498,
+      "step": 15058
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.76889967918396,
+      "learning_rate": 9.591145233789034e-07,
+      "loss": 0.9676,
+      "step": 15059
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.727650761604309,
+      "learning_rate": 9.583208380396714e-07,
+      "loss": 0.7985,
+      "step": 15060
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8442251682281494,
+      "learning_rate": 9.575274647035282e-07,
+      "loss": 0.9606,
+      "step": 15061
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.798671007156372,
+      "learning_rate": 9.567344033978555e-07,
+      "loss": 0.8341,
+      "step": 15062
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0089762210845947,
+      "learning_rate": 9.559416541500154e-07,
+      "loss": 0.4918,
+      "step": 15063
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8915917873382568,
+      "learning_rate": 9.551492169873666e-07,
+      "loss": 0.9208,
+      "step": 15064
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8015803098678589,
+      "learning_rate": 9.543570919372513e-07,
+      "loss": 0.904,
+      "step": 15065
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8134909868240356,
+      "learning_rate": 9.535652790270067e-07,
+      "loss": 0.9379,
+      "step": 15066
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7277216911315918,
+      "learning_rate": 9.527737782839519e-07,
+      "loss": 0.9128,
+      "step": 15067
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.757851481437683,
+      "learning_rate": 9.519825897354029e-07,
+      "loss": 0.856,
+      "step": 15068
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8581829071044922,
+      "learning_rate": 9.51191713408659e-07,
+      "loss": 0.9758,
+      "step": 15069
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8184823989868164,
+      "learning_rate": 9.504011493310128e-07,
+      "loss": 0.8702,
+      "step": 15070
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7623789310455322,
+      "learning_rate": 9.496108975297447e-07,
+      "loss": 0.8647,
+      "step": 15071
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.661779522895813,
+      "learning_rate": 9.488209580321217e-07,
+      "loss": 0.8028,
+      "step": 15072
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7767212390899658,
+      "learning_rate": 9.480313308654054e-07,
+      "loss": 0.9565,
+      "step": 15073
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0178263187408447,
+      "learning_rate": 9.472420160568407e-07,
+      "loss": 0.5721,
+      "step": 15074
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6802396774291992,
+      "learning_rate": 9.464530136336691e-07,
+      "loss": 0.7907,
+      "step": 15075
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7329049110412598,
+      "learning_rate": 9.456643236231111e-07,
+      "loss": 0.8673,
+      "step": 15076
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6688580513000488,
+      "learning_rate": 9.448759460523871e-07,
+      "loss": 0.9539,
+      "step": 15077
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9005753993988037,
+      "learning_rate": 9.440878809486975e-07,
+      "loss": 0.9538,
+      "step": 15078
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6351451873779297,
+      "learning_rate": 9.433001283392407e-07,
+      "loss": 0.8418,
+      "step": 15079
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8428897857666016,
+      "learning_rate": 9.425126882511948e-07,
+      "loss": 0.8815,
+      "step": 15080
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.722648024559021,
+      "learning_rate": 9.417255607117382e-07,
+      "loss": 0.9011,
+      "step": 15081
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.618863821029663,
+      "learning_rate": 9.409387457480268e-07,
+      "loss": 0.8018,
+      "step": 15082
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7407630681991577,
+      "learning_rate": 9.401522433872135e-07,
+      "loss": 0.8597,
+      "step": 15083
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6356356143951416,
+      "learning_rate": 9.393660536564408e-07,
+      "loss": 0.9266,
+      "step": 15084
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8879754543304443,
+      "learning_rate": 9.385801765828339e-07,
+      "loss": 0.9055,
+      "step": 15085
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6995222568511963,
+      "learning_rate": 9.377946121935144e-07,
+      "loss": 0.8397,
+      "step": 15086
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8619149923324585,
+      "learning_rate": 9.370093605155872e-07,
+      "loss": 0.93,
+      "step": 15087
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.723923683166504,
+      "learning_rate": 9.362244215761529e-07,
+      "loss": 0.8559,
+      "step": 15088
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7686301469802856,
+      "learning_rate": 9.354397954022931e-07,
+      "loss": 0.8448,
+      "step": 15089
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7810715436935425,
+      "learning_rate": 9.346554820210863e-07,
+      "loss": 0.8413,
+      "step": 15090
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7437270879745483,
+      "learning_rate": 9.338714814595928e-07,
+      "loss": 0.9076,
+      "step": 15091
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8411232233047485,
+      "learning_rate": 9.330877937448724e-07,
+      "loss": 0.939,
+      "step": 15092
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7314757108688354,
+      "learning_rate": 9.323044189039632e-07,
+      "loss": 0.9164,
+      "step": 15093
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8791905641555786,
+      "learning_rate": 9.315213569639004e-07,
+      "loss": 0.9976,
+      "step": 15094
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7264031171798706,
+      "learning_rate": 9.307386079517022e-07,
+      "loss": 0.8594,
+      "step": 15095
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7713074684143066,
+      "learning_rate": 9.299561718943829e-07,
+      "loss": 0.8778,
+      "step": 15096
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7188791036605835,
+      "learning_rate": 9.291740488189383e-07,
+      "loss": 0.7986,
+      "step": 15097
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7765638828277588,
+      "learning_rate": 9.283922387523603e-07,
+      "loss": 0.8795,
+      "step": 15098
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8715845346450806,
+      "learning_rate": 9.27610741721624e-07,
+      "loss": 0.9844,
+      "step": 15099
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.764789342880249,
+      "learning_rate": 9.268295577536979e-07,
+      "loss": 0.8705,
+      "step": 15100
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7246286869049072,
+      "learning_rate": 9.260486868755414e-07,
+      "loss": 0.8275,
+      "step": 15101
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.606115460395813,
+      "learning_rate": 9.252681291140953e-07,
+      "loss": 0.8048,
+      "step": 15102
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.71206533908844,
+      "learning_rate": 9.24487884496299e-07,
+      "loss": 0.8757,
+      "step": 15103
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7506659030914307,
+      "learning_rate": 9.237079530490722e-07,
+      "loss": 0.8756,
+      "step": 15104
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6884684562683105,
+      "learning_rate": 9.229283347993324e-07,
+      "loss": 0.9052,
+      "step": 15105
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.730265498161316,
+      "learning_rate": 9.221490297739777e-07,
+      "loss": 0.8346,
+      "step": 15106
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0522516965866089,
+      "learning_rate": 9.213700379999036e-07,
+      "loss": 0.5865,
+      "step": 15107
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9417134523391724,
+      "learning_rate": 9.205913595039883e-07,
+      "loss": 0.5228,
+      "step": 15108
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.880674123764038,
+      "learning_rate": 9.198129943131051e-07,
+      "loss": 0.8776,
+      "step": 15109
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8207496404647827,
+      "learning_rate": 9.190349424541078e-07,
+      "loss": 0.8326,
+      "step": 15110
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8347065448760986,
+      "learning_rate": 9.182572039538506e-07,
+      "loss": 0.8712,
+      "step": 15111
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8834484815597534,
+      "learning_rate": 9.174797788391676e-07,
+      "loss": 0.8954,
+      "step": 15112
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8277864456176758,
+      "learning_rate": 9.167026671368851e-07,
+      "loss": 0.8122,
+      "step": 15113
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6601061820983887,
+      "learning_rate": 9.159258688738226e-07,
+      "loss": 0.7879,
+      "step": 15114
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8744237422943115,
+      "learning_rate": 9.151493840767811e-07,
+      "loss": 0.8523,
+      "step": 15115
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.925638198852539,
+      "learning_rate": 9.143732127725591e-07,
+      "loss": 0.8906,
+      "step": 15116
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5890482664108276,
+      "learning_rate": 9.135973549879351e-07,
+      "loss": 0.8341,
+      "step": 15117
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9622678756713867,
+      "learning_rate": 9.128218107496878e-07,
+      "loss": 0.8564,
+      "step": 15118
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7663103342056274,
+      "learning_rate": 9.120465800845723e-07,
+      "loss": 0.9508,
+      "step": 15119
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0400755405426025,
+      "learning_rate": 9.112716630193463e-07,
+      "loss": 0.487,
+      "step": 15120
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.796094536781311,
+      "learning_rate": 9.10497059580745e-07,
+      "loss": 0.9509,
+      "step": 15121
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8268377780914307,
+      "learning_rate": 9.097227697955003e-07,
+      "loss": 0.9927,
+      "step": 15122
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.602809190750122,
+      "learning_rate": 9.089487936903296e-07,
+      "loss": 0.8955,
+      "step": 15123
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.638709306716919,
+      "learning_rate": 9.081751312919406e-07,
+      "loss": 0.9518,
+      "step": 15124
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8332468271255493,
+      "learning_rate": 9.074017826270332e-07,
+      "loss": 0.9191,
+      "step": 15125
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.694878339767456,
+      "learning_rate": 9.066287477222879e-07,
+      "loss": 0.9124,
+      "step": 15126
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8779165744781494,
+      "learning_rate": 9.058560266043869e-07,
+      "loss": 0.9178,
+      "step": 15127
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6200839281082153,
+      "learning_rate": 9.050836192999879e-07,
+      "loss": 0.8678,
+      "step": 15128
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5960053205490112,
+      "learning_rate": 9.043115258357494e-07,
+      "loss": 0.8739,
+      "step": 15129
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6541297435760498,
+      "learning_rate": 9.035397462383111e-07,
+      "loss": 0.9939,
+      "step": 15130
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6889852285385132,
+      "learning_rate": 9.027682805343074e-07,
+      "loss": 0.927,
+      "step": 15131
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6429500579833984,
+      "learning_rate": 9.01997128750357e-07,
+      "loss": 0.8253,
+      "step": 15132
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7418010234832764,
+      "learning_rate": 9.012262909130732e-07,
+      "loss": 0.8779,
+      "step": 15133
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7570445537567139,
+      "learning_rate": 9.004557670490522e-07,
+      "loss": 0.8924,
+      "step": 15134
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7874356508255005,
+      "learning_rate": 8.99685557184885e-07,
+      "loss": 0.9848,
+      "step": 15135
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.73018479347229,
+      "learning_rate": 8.989156613471473e-07,
+      "loss": 0.8521,
+      "step": 15136
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5975689888000488,
+      "learning_rate": 8.981460795624075e-07,
+      "loss": 0.863,
+      "step": 15137
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7245250940322876,
+      "learning_rate": 8.973768118572234e-07,
+      "loss": 0.882,
+      "step": 15138
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7696150541305542,
+      "learning_rate": 8.966078582581361e-07,
+      "loss": 0.8726,
+      "step": 15139
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6990641355514526,
+      "learning_rate": 8.958392187916842e-07,
+      "loss": 0.8661,
+      "step": 15140
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.181406259536743,
+      "learning_rate": 8.950708934843876e-07,
+      "loss": 0.9239,
+      "step": 15141
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7135200500488281,
+      "learning_rate": 8.94302882362762e-07,
+      "loss": 0.9212,
+      "step": 15142
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.819069743156433,
+      "learning_rate": 8.93535185453308e-07,
+      "loss": 0.8345,
+      "step": 15143
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9195365905761719,
+      "learning_rate": 8.92767802782517e-07,
+      "loss": 0.9481,
+      "step": 15144
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.87208092212677,
+      "learning_rate": 8.920007343768689e-07,
+      "loss": 0.9822,
+      "step": 15145
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.758432388305664,
+      "learning_rate": 8.912339802628333e-07,
+      "loss": 0.8813,
+      "step": 15146
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6271404027938843,
+      "learning_rate": 8.904675404668683e-07,
+      "loss": 0.8822,
+      "step": 15147
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6618837118148804,
+      "learning_rate": 8.897014150154237e-07,
+      "loss": 0.8927,
+      "step": 15148
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6085976362228394,
+      "learning_rate": 8.889356039349317e-07,
+      "loss": 0.8504,
+      "step": 15149
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.740433931350708,
+      "learning_rate": 8.881701072518223e-07,
+      "loss": 0.871,
+      "step": 15150
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8165339231491089,
+      "learning_rate": 8.874049249925121e-07,
+      "loss": 0.8814,
+      "step": 15151
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7171921730041504,
+      "learning_rate": 8.866400571833999e-07,
+      "loss": 0.8754,
+      "step": 15152
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7877271175384521,
+      "learning_rate": 8.858755038508849e-07,
+      "loss": 0.8934,
+      "step": 15153
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6933064460754395,
+      "learning_rate": 8.851112650213445e-07,
+      "loss": 0.9077,
+      "step": 15154
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8075060844421387,
+      "learning_rate": 8.843473407211545e-07,
+      "loss": 0.8545,
+      "step": 15155
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7054356336593628,
+      "learning_rate": 8.835837309766726e-07,
+      "loss": 0.8384,
+      "step": 15156
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9078304767608643,
+      "learning_rate": 8.828204358142511e-07,
+      "loss": 0.8444,
+      "step": 15157
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.974923849105835,
+      "learning_rate": 8.820574552602279e-07,
+      "loss": 0.9792,
+      "step": 15158
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.695038080215454,
+      "learning_rate": 8.81294789340934e-07,
+      "loss": 0.9137,
+      "step": 15159
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7332298755645752,
+      "learning_rate": 8.805324380826829e-07,
+      "loss": 0.9554,
+      "step": 15160
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6782861948013306,
+      "learning_rate": 8.797704015117847e-07,
+      "loss": 0.9144,
+      "step": 15161
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7296689748764038,
+      "learning_rate": 8.790086796545328e-07,
+      "loss": 0.9229,
+      "step": 15162
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.762547492980957,
+      "learning_rate": 8.782472725372138e-07,
+      "loss": 0.9497,
+      "step": 15163
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6892166137695312,
+      "learning_rate": 8.774861801861001e-07,
+      "loss": 0.9997,
+      "step": 15164
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9292035102844238,
+      "learning_rate": 8.767254026274563e-07,
+      "loss": 0.9049,
+      "step": 15165
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7236320972442627,
+      "learning_rate": 8.759649398875325e-07,
+      "loss": 0.8938,
+      "step": 15166
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6645889282226562,
+      "learning_rate": 8.752047919925722e-07,
+      "loss": 0.9131,
+      "step": 15167
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8525749444961548,
+      "learning_rate": 8.744449589688064e-07,
+      "loss": 0.8895,
+      "step": 15168
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.763445496559143,
+      "learning_rate": 8.736854408424522e-07,
+      "loss": 0.8549,
+      "step": 15169
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6880347728729248,
+      "learning_rate": 8.729262376397219e-07,
+      "loss": 0.9102,
+      "step": 15170
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0608956813812256,
+      "learning_rate": 8.721673493868111e-07,
+      "loss": 0.8427,
+      "step": 15171
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8918505907058716,
+      "learning_rate": 8.714087761099077e-07,
+      "loss": 0.8568,
+      "step": 15172
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.645162582397461,
+      "learning_rate": 8.706505178351865e-07,
+      "loss": 0.8695,
+      "step": 15173
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6317764520645142,
+      "learning_rate": 8.698925745888165e-07,
+      "loss": 0.8892,
+      "step": 15174
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7027106285095215,
+      "learning_rate": 8.691349463969467e-07,
+      "loss": 0.8413,
+      "step": 15175
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9443846940994263,
+      "learning_rate": 8.683776332857274e-07,
+      "loss": 0.9003,
+      "step": 15176
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9066749811172485,
+      "learning_rate": 8.676206352812844e-07,
+      "loss": 0.9099,
+      "step": 15177
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6307892799377441,
+      "learning_rate": 8.668639524097444e-07,
+      "loss": 0.9106,
+      "step": 15178
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8742786645889282,
+      "learning_rate": 8.661075846972177e-07,
+      "loss": 0.8782,
+      "step": 15179
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.99409818649292,
+      "learning_rate": 8.653515321698025e-07,
+      "loss": 0.8777,
+      "step": 15180
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.911147117614746,
+      "learning_rate": 8.64595794853591e-07,
+      "loss": 0.9458,
+      "step": 15181
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8192955255508423,
+      "learning_rate": 8.63840372774658e-07,
+      "loss": 0.931,
+      "step": 15182
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.830049753189087,
+      "learning_rate": 8.630852659590749e-07,
+      "loss": 0.8523,
+      "step": 15183
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.703125238418579,
+      "learning_rate": 8.623304744328942e-07,
+      "loss": 0.894,
+      "step": 15184
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9576330184936523,
+      "learning_rate": 8.61575998222166e-07,
+      "loss": 0.5642,
+      "step": 15185
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.644372582435608,
+      "learning_rate": 8.608218373529209e-07,
+      "loss": 0.8432,
+      "step": 15186
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6589711904525757,
+      "learning_rate": 8.600679918511868e-07,
+      "loss": 0.88,
+      "step": 15187
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9155924320220947,
+      "learning_rate": 8.593144617429727e-07,
+      "loss": 0.8935,
+      "step": 15188
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.742495059967041,
+      "learning_rate": 8.585612470542859e-07,
+      "loss": 0.8478,
+      "step": 15189
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8457953929901123,
+      "learning_rate": 8.578083478111121e-07,
+      "loss": 0.8878,
+      "step": 15190
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.741173505783081,
+      "learning_rate": 8.570557640394351e-07,
+      "loss": 0.8439,
+      "step": 15191
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.852959394454956,
+      "learning_rate": 8.56303495765225e-07,
+      "loss": 0.9241,
+      "step": 15192
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8397605419158936,
+      "learning_rate": 8.555515430144379e-07,
+      "loss": 0.8774,
+      "step": 15193
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.769058346748352,
+      "learning_rate": 8.547999058130252e-07,
+      "loss": 0.8968,
+      "step": 15194
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7914997339248657,
+      "learning_rate": 8.540485841869195e-07,
+      "loss": 0.8567,
+      "step": 15195
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8630309104919434,
+      "learning_rate": 8.532975781620511e-07,
+      "loss": 0.9072,
+      "step": 15196
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7516695261001587,
+      "learning_rate": 8.525468877643316e-07,
+      "loss": 1.0045,
+      "step": 15197
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6907166242599487,
+      "learning_rate": 8.517965130196681e-07,
+      "loss": 0.871,
+      "step": 15198
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8259077072143555,
+      "learning_rate": 8.510464539539498e-07,
+      "loss": 0.8816,
+      "step": 15199
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5879443883895874,
+      "learning_rate": 8.502967105930648e-07,
+      "loss": 0.9223,
+      "step": 15200
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8313791751861572,
+      "learning_rate": 8.495472829628793e-07,
+      "loss": 0.9014,
+      "step": 15201
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8462830781936646,
+      "learning_rate": 8.487981710892579e-07,
+      "loss": 0.853,
+      "step": 15202
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8247673511505127,
+      "learning_rate": 8.480493749980468e-07,
+      "loss": 0.9014,
+      "step": 15203
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5998510122299194,
+      "learning_rate": 8.473008947150873e-07,
+      "loss": 0.824,
+      "step": 15204
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.997332751750946,
+      "learning_rate": 8.465527302662091e-07,
+      "loss": 0.5273,
+      "step": 15205
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5914613008499146,
+      "learning_rate": 8.458048816772246e-07,
+      "loss": 0.7913,
+      "step": 15206
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6336804628372192,
+      "learning_rate": 8.450573489739445e-07,
+      "loss": 0.7912,
+      "step": 15207
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8204389810562134,
+      "learning_rate": 8.443101321821601e-07,
+      "loss": 0.8699,
+      "step": 15208
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7686796188354492,
+      "learning_rate": 8.435632313276587e-07,
+      "loss": 0.9126,
+      "step": 15209
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9280214309692383,
+      "learning_rate": 8.428166464362119e-07,
+      "loss": 0.9493,
+      "step": 15210
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7088472843170166,
+      "learning_rate": 8.420703775335848e-07,
+      "loss": 0.8734,
+      "step": 15211
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6044533252716064,
+      "learning_rate": 8.413244246455255e-07,
+      "loss": 0.9046,
+      "step": 15212
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8642398118972778,
+      "learning_rate": 8.405787877977778e-07,
+      "loss": 0.865,
+      "step": 15213
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.808875560760498,
+      "learning_rate": 8.39833467016069e-07,
+      "loss": 0.9862,
+      "step": 15214
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.900651216506958,
+      "learning_rate": 8.390884623261198e-07,
+      "loss": 0.9192,
+      "step": 15215
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8250494003295898,
+      "learning_rate": 8.383437737536382e-07,
+      "loss": 0.8297,
+      "step": 15216
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6764636039733887,
+      "learning_rate": 8.375994013243205e-07,
+      "loss": 0.9021,
+      "step": 15217
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7157115936279297,
+      "learning_rate": 8.368553450638539e-07,
+      "loss": 0.7863,
+      "step": 15218
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7138679027557373,
+      "learning_rate": 8.361116049979124e-07,
+      "loss": 0.8739,
+      "step": 15219
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0029304027557373,
+      "learning_rate": 8.35368181152163e-07,
+      "loss": 0.8971,
+      "step": 15220
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6574784517288208,
+      "learning_rate": 8.346250735522543e-07,
+      "loss": 0.8902,
+      "step": 15221
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8043324947357178,
+      "learning_rate": 8.338822822238346e-07,
+      "loss": 0.9497,
+      "step": 15222
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6096408367156982,
+      "learning_rate": 8.33139807192529e-07,
+      "loss": 0.8967,
+      "step": 15223
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7971632480621338,
+      "learning_rate": 8.323976484839657e-07,
+      "loss": 0.8081,
+      "step": 15224
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8918967247009277,
+      "learning_rate": 8.31655806123749e-07,
+      "loss": 0.9406,
+      "step": 15225
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0474677085876465,
+      "learning_rate": 8.309142801374825e-07,
+      "loss": 0.9861,
+      "step": 15226
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6858329772949219,
+      "learning_rate": 8.301730705507483e-07,
+      "loss": 0.8984,
+      "step": 15227
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6777770519256592,
+      "learning_rate": 8.294321773891289e-07,
+      "loss": 0.8557,
+      "step": 15228
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7256790399551392,
+      "learning_rate": 8.286916006781865e-07,
+      "loss": 0.8595,
+      "step": 15229
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7659342288970947,
+      "learning_rate": 8.279513404434792e-07,
+      "loss": 0.8736,
+      "step": 15230
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8282809257507324,
+      "learning_rate": 8.272113967105477e-07,
+      "loss": 0.8682,
+      "step": 15231
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0442488193511963,
+      "learning_rate": 8.264717695049284e-07,
+      "loss": 0.5211,
+      "step": 15232
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.88129460811615,
+      "learning_rate": 8.257324588521454e-07,
+      "loss": 0.9203,
+      "step": 15233
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0429025888442993,
+      "learning_rate": 8.249934647777058e-07,
+      "loss": 0.5053,
+      "step": 15234
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.761221170425415,
+      "learning_rate": 8.24254787307115e-07,
+      "loss": 0.9238,
+      "step": 15235
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5914571285247803,
+      "learning_rate": 8.235164264658568e-07,
+      "loss": 0.7877,
+      "step": 15236
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6997902393341064,
+      "learning_rate": 8.227783822794155e-07,
+      "loss": 0.9062,
+      "step": 15237
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7110528945922852,
+      "learning_rate": 8.220406547732551e-07,
+      "loss": 0.8518,
+      "step": 15238
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6429951190948486,
+      "learning_rate": 8.213032439728364e-07,
+      "loss": 0.8685,
+      "step": 15239
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8685827255249023,
+      "learning_rate": 8.20566149903601e-07,
+      "loss": 0.9036,
+      "step": 15240
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7199651002883911,
+      "learning_rate": 8.198293725909867e-07,
+      "loss": 0.8735,
+      "step": 15241
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9008992910385132,
+      "learning_rate": 8.190929120604163e-07,
+      "loss": 0.9161,
+      "step": 15242
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.880269169807434,
+      "learning_rate": 8.183567683373062e-07,
+      "loss": 0.9481,
+      "step": 15243
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0379102230072021,
+      "learning_rate": 8.176209414470526e-07,
+      "loss": 0.5368,
+      "step": 15244
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7062644958496094,
+      "learning_rate": 8.16885431415052e-07,
+      "loss": 0.8623,
+      "step": 15245
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8536968231201172,
+      "learning_rate": 8.161502382666841e-07,
+      "loss": 0.8754,
+      "step": 15246
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8501344919204712,
+      "learning_rate": 8.154153620273153e-07,
+      "loss": 0.897,
+      "step": 15247
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8082139492034912,
+      "learning_rate": 8.146808027223085e-07,
+      "loss": 0.9115,
+      "step": 15248
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7853705883026123,
+      "learning_rate": 8.139465603770069e-07,
+      "loss": 0.8958,
+      "step": 15249
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.682013988494873,
+      "learning_rate": 8.132126350167513e-07,
+      "loss": 0.8813,
+      "step": 15250
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.837856411933899,
+      "learning_rate": 8.124790266668647e-07,
+      "loss": 0.9178,
+      "step": 15251
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8251163959503174,
+      "learning_rate": 8.117457353526626e-07,
+      "loss": 0.9806,
+      "step": 15252
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8140475749969482,
+      "learning_rate": 8.110127610994478e-07,
+      "loss": 0.8961,
+      "step": 15253
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8378630876541138,
+      "learning_rate": 8.10280103932517e-07,
+      "loss": 0.9992,
+      "step": 15254
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7668818235397339,
+      "learning_rate": 8.095477638771465e-07,
+      "loss": 0.8274,
+      "step": 15255
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.733489751815796,
+      "learning_rate": 8.088157409586094e-07,
+      "loss": 0.8541,
+      "step": 15256
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7906293869018555,
+      "learning_rate": 8.080840352021702e-07,
+      "loss": 0.8856,
+      "step": 15257
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6362950801849365,
+      "learning_rate": 8.073526466330716e-07,
+      "loss": 0.9162,
+      "step": 15258
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8058305978775024,
+      "learning_rate": 8.06621575276556e-07,
+      "loss": 0.9176,
+      "step": 15259
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8573954105377197,
+      "learning_rate": 8.058908211578476e-07,
+      "loss": 0.9214,
+      "step": 15260
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6094926595687866,
+      "learning_rate": 8.05160384302166e-07,
+      "loss": 0.9377,
+      "step": 15261
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8726974725723267,
+      "learning_rate": 8.044302647347135e-07,
+      "loss": 0.8235,
+      "step": 15262
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7805413007736206,
+      "learning_rate": 8.037004624806866e-07,
+      "loss": 0.9444,
+      "step": 15263
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.680458664894104,
+      "learning_rate": 8.029709775652672e-07,
+      "loss": 0.8491,
+      "step": 15264
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.614868402481079,
+      "learning_rate": 8.022418100136298e-07,
+      "loss": 0.8931,
+      "step": 15265
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6888545751571655,
+      "learning_rate": 8.01512959850933e-07,
+      "loss": 0.8708,
+      "step": 15266
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7408480644226074,
+      "learning_rate": 8.007844271023301e-07,
+      "loss": 0.8794,
+      "step": 15267
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7664724588394165,
+      "learning_rate": 8.000562117929589e-07,
+      "loss": 0.8884,
+      "step": 15268
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9032137393951416,
+      "learning_rate": 7.993283139479479e-07,
+      "loss": 0.8686,
+      "step": 15269
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7440332174301147,
+      "learning_rate": 7.986007335924184e-07,
+      "loss": 0.8824,
+      "step": 15270
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8190518617630005,
+      "learning_rate": 7.978734707514724e-07,
+      "loss": 0.9718,
+      "step": 15271
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6874903440475464,
+      "learning_rate": 7.971465254502097e-07,
+      "loss": 0.9005,
+      "step": 15272
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0436069965362549,
+      "learning_rate": 7.964198977137116e-07,
+      "loss": 0.513,
+      "step": 15273
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8280973434448242,
+      "learning_rate": 7.956935875670547e-07,
+      "loss": 0.8382,
+      "step": 15274
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0568522214889526,
+      "learning_rate": 7.949675950352998e-07,
+      "loss": 0.5608,
+      "step": 15275
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.754268765449524,
+      "learning_rate": 7.942419201435014e-07,
+      "loss": 0.8751,
+      "step": 15276
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7185472249984741,
+      "learning_rate": 7.935165629166974e-07,
+      "loss": 0.8699,
+      "step": 15277
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7946736812591553,
+      "learning_rate": 7.927915233799221e-07,
+      "loss": 0.8884,
+      "step": 15278
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.730396032333374,
+      "learning_rate": 7.92066801558189e-07,
+      "loss": 0.8469,
+      "step": 15279
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8024253845214844,
+      "learning_rate": 7.913423974765111e-07,
+      "loss": 0.9022,
+      "step": 15280
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9227482080459595,
+      "learning_rate": 7.906183111598831e-07,
+      "loss": 0.8795,
+      "step": 15281
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9201732873916626,
+      "learning_rate": 7.898945426332905e-07,
+      "loss": 0.8677,
+      "step": 15282
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.964017391204834,
+      "learning_rate": 7.891710919217133e-07,
+      "loss": 0.5305,
+      "step": 15283
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7004503011703491,
+      "learning_rate": 7.884479590501093e-07,
+      "loss": 0.9149,
+      "step": 15284
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8003718852996826,
+      "learning_rate": 7.877251440434363e-07,
+      "loss": 0.9443,
+      "step": 15285
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7350023984909058,
+      "learning_rate": 7.870026469266334e-07,
+      "loss": 0.8812,
+      "step": 15286
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.770645260810852,
+      "learning_rate": 7.86280467724636e-07,
+      "loss": 0.9295,
+      "step": 15287
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8011971712112427,
+      "learning_rate": 7.85558606462361e-07,
+      "loss": 0.8509,
+      "step": 15288
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7753578424453735,
+      "learning_rate": 7.848370631647184e-07,
+      "loss": 0.8769,
+      "step": 15289
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6029423475265503,
+      "learning_rate": 7.84115837856606e-07,
+      "loss": 0.8389,
+      "step": 15290
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6923832893371582,
+      "learning_rate": 7.833949305629163e-07,
+      "loss": 0.9069,
+      "step": 15291
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0122928619384766,
+      "learning_rate": 7.826743413085192e-07,
+      "loss": 0.8846,
+      "step": 15292
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7595362663269043,
+      "learning_rate": 7.819540701182848e-07,
+      "loss": 0.8581,
+      "step": 15293
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6417657136917114,
+      "learning_rate": 7.812341170170646e-07,
+      "loss": 0.9159,
+      "step": 15294
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5641648769378662,
+      "learning_rate": 7.80514482029704e-07,
+      "loss": 0.8807,
+      "step": 15295
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8082481622695923,
+      "learning_rate": 7.797951651810343e-07,
+      "loss": 0.9057,
+      "step": 15296
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.857820749282837,
+      "learning_rate": 7.790761664958791e-07,
+      "loss": 0.8962,
+      "step": 15297
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7280539274215698,
+      "learning_rate": 7.783574859990461e-07,
+      "loss": 0.9306,
+      "step": 15298
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8621299266815186,
+      "learning_rate": 7.776391237153369e-07,
+      "loss": 0.894,
+      "step": 15299
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.017876148223877,
+      "learning_rate": 7.769210796695415e-07,
+      "loss": 0.9237,
+      "step": 15300
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8639944791793823,
+      "learning_rate": 7.762033538864344e-07,
+      "loss": 0.783,
+      "step": 15301
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.844602108001709,
+      "learning_rate": 7.75485946390786e-07,
+      "loss": 0.8771,
+      "step": 15302
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8975077867507935,
+      "learning_rate": 7.747688572073475e-07,
+      "loss": 0.9442,
+      "step": 15303
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7101982831954956,
+      "learning_rate": 7.740520863608681e-07,
+      "loss": 0.8695,
+      "step": 15304
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7556662559509277,
+      "learning_rate": 7.733356338760778e-07,
+      "loss": 0.8965,
+      "step": 15305
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7240653038024902,
+      "learning_rate": 7.726194997777036e-07,
+      "loss": 0.8471,
+      "step": 15306
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7658778429031372,
+      "learning_rate": 7.719036840904525e-07,
+      "loss": 0.9852,
+      "step": 15307
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7187371253967285,
+      "learning_rate": 7.711881868390292e-07,
+      "loss": 0.8245,
+      "step": 15308
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.834443211555481,
+      "learning_rate": 7.704730080481205e-07,
+      "loss": 0.8566,
+      "step": 15309
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.628485083580017,
+      "learning_rate": 7.697581477424055e-07,
+      "loss": 1.0261,
+      "step": 15310
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8690414428710938,
+      "learning_rate": 7.690436059465567e-07,
+      "loss": 0.9066,
+      "step": 15311
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.736380934715271,
+      "learning_rate": 7.683293826852245e-07,
+      "loss": 0.859,
+      "step": 15312
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8093090057373047,
+      "learning_rate": 7.676154779830591e-07,
+      "loss": 0.8795,
+      "step": 15313
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7221871614456177,
+      "learning_rate": 7.669018918646932e-07,
+      "loss": 0.8856,
+      "step": 15314
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.830049991607666,
+      "learning_rate": 7.661886243547534e-07,
+      "loss": 0.9,
+      "step": 15315
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6073338985443115,
+      "learning_rate": 7.654756754778481e-07,
+      "loss": 0.8534,
+      "step": 15316
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.694115161895752,
+      "learning_rate": 7.64763045258583e-07,
+      "loss": 0.9407,
+      "step": 15317
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.664668321609497,
+      "learning_rate": 7.640507337215463e-07,
+      "loss": 0.8868,
+      "step": 15318
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7007136344909668,
+      "learning_rate": 7.633387408913207e-07,
+      "loss": 0.9312,
+      "step": 15319
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8537704944610596,
+      "learning_rate": 7.626270667924728e-07,
+      "loss": 0.9057,
+      "step": 15320
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0101597309112549,
+      "learning_rate": 7.619157114495623e-07,
+      "loss": 0.4836,
+      "step": 15321
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8031200170516968,
+      "learning_rate": 7.612046748871327e-07,
+      "loss": 0.8791,
+      "step": 15322
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7335458993911743,
+      "learning_rate": 7.604939571297232e-07,
+      "loss": 0.8428,
+      "step": 15323
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6624542474746704,
+      "learning_rate": 7.597835582018586e-07,
+      "loss": 0.8616,
+      "step": 15324
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.754930019378662,
+      "learning_rate": 7.590734781280506e-07,
+      "loss": 0.9545,
+      "step": 15325
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8274152278900146,
+      "learning_rate": 7.583637169328062e-07,
+      "loss": 0.9444,
+      "step": 15326
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8776918649673462,
+      "learning_rate": 7.576542746406112e-07,
+      "loss": 0.8457,
+      "step": 15327
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7671899795532227,
+      "learning_rate": 7.569451512759518e-07,
+      "loss": 0.8489,
+      "step": 15328
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9043998718261719,
+      "learning_rate": 7.562363468632949e-07,
+      "loss": 0.9008,
+      "step": 15329
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.903128743171692,
+      "learning_rate": 7.555278614271011e-07,
+      "loss": 0.8482,
+      "step": 15330
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8257449865341187,
+      "learning_rate": 7.548196949918152e-07,
+      "loss": 0.8914,
+      "step": 15331
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7120627164840698,
+      "learning_rate": 7.541118475818787e-07,
+      "loss": 0.8215,
+      "step": 15332
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6204255819320679,
+      "learning_rate": 7.534043192217133e-07,
+      "loss": 0.9274,
+      "step": 15333
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7735533714294434,
+      "learning_rate": 7.526971099357372e-07,
+      "loss": 0.8458,
+      "step": 15334
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.703644037246704,
+      "learning_rate": 7.519902197483508e-07,
+      "loss": 0.8486,
+      "step": 15335
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8146334886550903,
+      "learning_rate": 7.512836486839492e-07,
+      "loss": 0.8954,
+      "step": 15336
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.004289150238037,
+      "learning_rate": 7.50577396766915e-07,
+      "loss": 0.8612,
+      "step": 15337
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9826062917709351,
+      "learning_rate": 7.498714640216154e-07,
+      "loss": 0.5011,
+      "step": 15338
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.919113278388977,
+      "learning_rate": 7.491658504724142e-07,
+      "loss": 0.9769,
+      "step": 15339
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7631869316101074,
+      "learning_rate": 7.484605561436575e-07,
+      "loss": 0.8657,
+      "step": 15340
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0938211679458618,
+      "learning_rate": 7.477555810596848e-07,
+      "loss": 0.4978,
+      "step": 15341
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8437411785125732,
+      "learning_rate": 7.470509252448199e-07,
+      "loss": 0.9147,
+      "step": 15342
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.836773157119751,
+      "learning_rate": 7.463465887233834e-07,
+      "loss": 0.8811,
+      "step": 15343
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7091957330703735,
+      "learning_rate": 7.456425715196747e-07,
+      "loss": 0.8978,
+      "step": 15344
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8344340324401855,
+      "learning_rate": 7.44938873657991e-07,
+      "loss": 0.8608,
+      "step": 15345
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8668969869613647,
+      "learning_rate": 7.44235495162613e-07,
+      "loss": 0.8088,
+      "step": 15346
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6947628259658813,
+      "learning_rate": 7.435324360578122e-07,
+      "loss": 0.9283,
+      "step": 15347
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8333630561828613,
+      "learning_rate": 7.428296963678527e-07,
+      "loss": 0.888,
+      "step": 15348
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9681718349456787,
+      "learning_rate": 7.421272761169795e-07,
+      "loss": 0.9329,
+      "step": 15349
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6789928674697876,
+      "learning_rate": 7.414251753294344e-07,
+      "loss": 0.847,
+      "step": 15350
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9197742938995361,
+      "learning_rate": 7.407233940294422e-07,
+      "loss": 0.8764,
+      "step": 15351
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6077734231948853,
+      "learning_rate": 7.400219322412239e-07,
+      "loss": 0.9421,
+      "step": 15352
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6395260095596313,
+      "learning_rate": 7.393207899889787e-07,
+      "loss": 0.822,
+      "step": 15353
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7577667236328125,
+      "learning_rate": 7.386199672969063e-07,
+      "loss": 0.9119,
+      "step": 15354
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.912465214729309,
+      "learning_rate": 7.379194641891874e-07,
+      "loss": 0.8642,
+      "step": 15355
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8100286722183228,
+      "learning_rate": 7.372192806899947e-07,
+      "loss": 0.8424,
+      "step": 15356
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0536580085754395,
+      "learning_rate": 7.365194168234902e-07,
+      "loss": 0.8748,
+      "step": 15357
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.845513939857483,
+      "learning_rate": 7.358198726138255e-07,
+      "loss": 1.0209,
+      "step": 15358
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7771759033203125,
+      "learning_rate": 7.35120648085137e-07,
+      "loss": 0.7919,
+      "step": 15359
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8251824378967285,
+      "learning_rate": 7.344217432615564e-07,
+      "loss": 0.8886,
+      "step": 15360
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6980317831039429,
+      "learning_rate": 7.337231581671977e-07,
+      "loss": 0.8138,
+      "step": 15361
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7436755895614624,
+      "learning_rate": 7.330248928261697e-07,
+      "loss": 1.0136,
+      "step": 15362
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0328305959701538,
+      "learning_rate": 7.32326947262565e-07,
+      "loss": 0.5768,
+      "step": 15363
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8750241994857788,
+      "learning_rate": 7.316293215004689e-07,
+      "loss": 0.9608,
+      "step": 15364
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.871216893196106,
+      "learning_rate": 7.309320155639565e-07,
+      "loss": 0.9215,
+      "step": 15365
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.061514377593994,
+      "learning_rate": 7.302350294770866e-07,
+      "loss": 0.8723,
+      "step": 15366
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7541924715042114,
+      "learning_rate": 7.29538363263913e-07,
+      "loss": 0.8933,
+      "step": 15367
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8216084241867065,
+      "learning_rate": 7.288420169484734e-07,
+      "loss": 0.8812,
+      "step": 15368
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.685610294342041,
+      "learning_rate": 7.281459905547994e-07,
+      "loss": 0.8764,
+      "step": 15369
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8366074562072754,
+      "learning_rate": 7.274502841069053e-07,
+      "loss": 0.8876,
+      "step": 15370
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7516920566558838,
+      "learning_rate": 7.267548976288019e-07,
+      "loss": 0.8532,
+      "step": 15371
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.838331937789917,
+      "learning_rate": 7.260598311444822e-07,
+      "loss": 0.9454,
+      "step": 15372
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7801814079284668,
+      "learning_rate": 7.253650846779325e-07,
+      "loss": 0.8898,
+      "step": 15373
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8323547840118408,
+      "learning_rate": 7.246706582531249e-07,
+      "loss": 0.9131,
+      "step": 15374
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8539369106292725,
+      "learning_rate": 7.239765518940256e-07,
+      "loss": 0.9305,
+      "step": 15375
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.696033000946045,
+      "learning_rate": 7.232827656245823e-07,
+      "loss": 0.9876,
+      "step": 15376
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7499042749404907,
+      "learning_rate": 7.225892994687367e-07,
+      "loss": 0.9156,
+      "step": 15377
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9029293060302734,
+      "learning_rate": 7.218961534504209e-07,
+      "loss": 0.8594,
+      "step": 15378
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0207233428955078,
+      "learning_rate": 7.212033275935493e-07,
+      "loss": 0.5412,
+      "step": 15379
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8558810949325562,
+      "learning_rate": 7.205108219220336e-07,
+      "loss": 0.9017,
+      "step": 15380
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7028923034667969,
+      "learning_rate": 7.19818636459767e-07,
+      "loss": 0.891,
+      "step": 15381
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8055140972137451,
+      "learning_rate": 7.191267712306372e-07,
+      "loss": 0.7845,
+      "step": 15382
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0377800464630127,
+      "learning_rate": 7.184352262585159e-07,
+      "loss": 0.5333,
+      "step": 15383
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8510552644729614,
+      "learning_rate": 7.177440015672699e-07,
+      "loss": 0.8386,
+      "step": 15384
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8055551052093506,
+      "learning_rate": 7.170530971807477e-07,
+      "loss": 0.9669,
+      "step": 15385
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8047500848770142,
+      "learning_rate": 7.163625131227936e-07,
+      "loss": 0.9222,
+      "step": 15386
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.619687557220459,
+      "learning_rate": 7.156722494172352e-07,
+      "loss": 0.8661,
+      "step": 15387
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8669445514678955,
+      "learning_rate": 7.149823060878946e-07,
+      "loss": 0.9242,
+      "step": 15388
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6497832536697388,
+      "learning_rate": 7.142926831585761e-07,
+      "loss": 0.9129,
+      "step": 15389
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6180696487426758,
+      "learning_rate": 7.136033806530784e-07,
+      "loss": 0.8539,
+      "step": 15390
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9061260223388672,
+      "learning_rate": 7.129143985951892e-07,
+      "loss": 0.9253,
+      "step": 15391
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6800200939178467,
+      "learning_rate": 7.122257370086793e-07,
+      "loss": 0.8556,
+      "step": 15392
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.700907826423645,
+      "learning_rate": 7.115373959173177e-07,
+      "loss": 0.8428,
+      "step": 15393
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8817559480667114,
+      "learning_rate": 7.10849375344852e-07,
+      "loss": 0.838,
+      "step": 15394
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.793295979499817,
+      "learning_rate": 7.101616753150275e-07,
+      "loss": 0.8888,
+      "step": 15395
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8626763820648193,
+      "learning_rate": 7.094742958515722e-07,
+      "loss": 0.8945,
+      "step": 15396
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.722426176071167,
+      "learning_rate": 7.08787236978209e-07,
+      "loss": 0.8826,
+      "step": 15397
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.987001895904541,
+      "learning_rate": 7.081004987186424e-07,
+      "loss": 0.8784,
+      "step": 15398
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7001419067382812,
+      "learning_rate": 7.074140810965724e-07,
+      "loss": 0.9045,
+      "step": 15399
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9331961870193481,
+      "learning_rate": 7.067279841356844e-07,
+      "loss": 0.8307,
+      "step": 15400
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7904248237609863,
+      "learning_rate": 7.060422078596529e-07,
+      "loss": 0.9443,
+      "step": 15401
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7708702087402344,
+      "learning_rate": 7.053567522921457e-07,
+      "loss": 0.8315,
+      "step": 15402
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8293017148971558,
+      "learning_rate": 7.046716174568114e-07,
+      "loss": 0.8291,
+      "step": 15403
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6780261993408203,
+      "learning_rate": 7.039868033772956e-07,
+      "loss": 0.8253,
+      "step": 15404
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8134949207305908,
+      "learning_rate": 7.033023100772262e-07,
+      "loss": 0.8511,
+      "step": 15405
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6957945823669434,
+      "learning_rate": 7.026181375802266e-07,
+      "loss": 0.8699,
+      "step": 15406
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9956048727035522,
+      "learning_rate": 7.019342859099032e-07,
+      "loss": 0.8516,
+      "step": 15407
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7115371227264404,
+      "learning_rate": 7.012507550898551e-07,
+      "loss": 0.9003,
+      "step": 15408
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6581414937973022,
+      "learning_rate": 7.005675451436667e-07,
+      "loss": 0.8825,
+      "step": 15409
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.609484314918518,
+      "learning_rate": 6.998846560949168e-07,
+      "loss": 0.8471,
+      "step": 15410
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.700181007385254,
+      "learning_rate": 6.992020879671679e-07,
+      "loss": 0.9138,
+      "step": 15411
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7253769636154175,
+      "learning_rate": 6.985198407839755e-07,
+      "loss": 0.9419,
+      "step": 15412
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7705073356628418,
+      "learning_rate": 6.978379145688785e-07,
+      "loss": 0.9007,
+      "step": 15413
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8615552186965942,
+      "learning_rate": 6.971563093454114e-07,
+      "loss": 0.8839,
+      "step": 15414
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7783173322677612,
+      "learning_rate": 6.964750251370945e-07,
+      "loss": 0.8425,
+      "step": 15415
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7722424268722534,
+      "learning_rate": 6.957940619674352e-07,
+      "loss": 0.8919,
+      "step": 15416
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8861054182052612,
+      "learning_rate": 6.951134198599341e-07,
+      "loss": 0.9458,
+      "step": 15417
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8971647024154663,
+      "learning_rate": 6.944330988380743e-07,
+      "loss": 0.8908,
+      "step": 15418
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8293219804763794,
+      "learning_rate": 6.93753098925336e-07,
+      "loss": 0.857,
+      "step": 15419
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.776326060295105,
+      "learning_rate": 6.930734201451817e-07,
+      "loss": 0.8946,
+      "step": 15420
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0064682960510254,
+      "learning_rate": 6.923940625210668e-07,
+      "loss": 0.9297,
+      "step": 15421
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7492276430130005,
+      "learning_rate": 6.917150260764293e-07,
+      "loss": 0.8724,
+      "step": 15422
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7210890054702759,
+      "learning_rate": 6.910363108347084e-07,
+      "loss": 0.9525,
+      "step": 15423
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9671790599822998,
+      "learning_rate": 6.903579168193197e-07,
+      "loss": 0.8913,
+      "step": 15424
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7515835762023926,
+      "learning_rate": 6.896798440536744e-07,
+      "loss": 0.9254,
+      "step": 15425
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7188644409179688,
+      "learning_rate": 6.890020925611696e-07,
+      "loss": 0.9043,
+      "step": 15426
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7102930545806885,
+      "learning_rate": 6.883246623651951e-07,
+      "loss": 0.8841,
+      "step": 15427
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7879194021224976,
+      "learning_rate": 6.876475534891236e-07,
+      "loss": 0.9297,
+      "step": 15428
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7796441316604614,
+      "learning_rate": 6.86970765956323e-07,
+      "loss": 0.8203,
+      "step": 15429
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8255034685134888,
+      "learning_rate": 6.862942997901456e-07,
+      "loss": 0.8743,
+      "step": 15430
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7486450672149658,
+      "learning_rate": 6.856181550139341e-07,
+      "loss": 0.8166,
+      "step": 15431
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8995726108551025,
+      "learning_rate": 6.849423316510239e-07,
+      "loss": 0.9474,
+      "step": 15432
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7616682052612305,
+      "learning_rate": 6.842668297247312e-07,
+      "loss": 0.8631,
+      "step": 15433
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1272461414337158,
+      "learning_rate": 6.835916492583694e-07,
+      "loss": 0.5383,
+      "step": 15434
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7339755296707153,
+      "learning_rate": 6.829167902752342e-07,
+      "loss": 0.9032,
+      "step": 15435
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1219072341918945,
+      "learning_rate": 6.822422527986161e-07,
+      "loss": 0.9946,
+      "step": 15436
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.712729811668396,
+      "learning_rate": 6.815680368517874e-07,
+      "loss": 0.8669,
+      "step": 15437
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9047094583511353,
+      "learning_rate": 6.808941424580184e-07,
+      "loss": 0.9317,
+      "step": 15438
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7539697885513306,
+      "learning_rate": 6.802205696405584e-07,
+      "loss": 0.8974,
+      "step": 15439
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8198806047439575,
+      "learning_rate": 6.795473184226542e-07,
+      "loss": 0.9215,
+      "step": 15440
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8968397378921509,
+      "learning_rate": 6.788743888275351e-07,
+      "loss": 0.927,
+      "step": 15441
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8130605220794678,
+      "learning_rate": 6.782017808784236e-07,
+      "loss": 0.8549,
+      "step": 15442
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.68769371509552,
+      "learning_rate": 6.7752949459853e-07,
+      "loss": 0.8329,
+      "step": 15443
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7522064447402954,
+      "learning_rate": 6.768575300110514e-07,
+      "loss": 0.8102,
+      "step": 15444
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.762490153312683,
+      "learning_rate": 6.76185887139178e-07,
+      "loss": 0.9177,
+      "step": 15445
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8453413248062134,
+      "learning_rate": 6.755145660060825e-07,
+      "loss": 0.8769,
+      "step": 15446
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6166960000991821,
+      "learning_rate": 6.74843566634934e-07,
+      "loss": 0.9805,
+      "step": 15447
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.856553316116333,
+      "learning_rate": 6.741728890488841e-07,
+      "loss": 0.8987,
+      "step": 15448
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9598171710968018,
+      "learning_rate": 6.735025332710776e-07,
+      "loss": 0.9108,
+      "step": 15449
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7338714599609375,
+      "learning_rate": 6.72832499324646e-07,
+      "loss": 0.9798,
+      "step": 15450
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8061928749084473,
+      "learning_rate": 6.721627872327119e-07,
+      "loss": 0.9121,
+      "step": 15451
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6309852600097656,
+      "learning_rate": 6.714933970183813e-07,
+      "loss": 0.9267,
+      "step": 15452
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7296868562698364,
+      "learning_rate": 6.708243287047578e-07,
+      "loss": 0.8577,
+      "step": 15453
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6793081760406494,
+      "learning_rate": 6.701555823149242e-07,
+      "loss": 0.9314,
+      "step": 15454
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7906148433685303,
+      "learning_rate": 6.694871578719608e-07,
+      "loss": 0.8813,
+      "step": 15455
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7418558597564697,
+      "learning_rate": 6.688190553989327e-07,
+      "loss": 0.9142,
+      "step": 15456
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7237902879714966,
+      "learning_rate": 6.681512749188923e-07,
+      "loss": 0.8722,
+      "step": 15457
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6942803859710693,
+      "learning_rate": 6.674838164548847e-07,
+      "loss": 0.9119,
+      "step": 15458
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9457521438598633,
+      "learning_rate": 6.668166800299402e-07,
+      "loss": 0.8638,
+      "step": 15459
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7421132326126099,
+      "learning_rate": 6.661498656670828e-07,
+      "loss": 0.8351,
+      "step": 15460
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6366709470748901,
+      "learning_rate": 6.654833733893184e-07,
+      "loss": 0.8619,
+      "step": 15461
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0346012115478516,
+      "learning_rate": 6.648172032196487e-07,
+      "loss": 0.4982,
+      "step": 15462
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.909983515739441,
+      "learning_rate": 6.641513551810608e-07,
+      "loss": 0.9558,
+      "step": 15463
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6495444774627686,
+      "learning_rate": 6.634858292965307e-07,
+      "loss": 0.7992,
+      "step": 15464
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8287988901138306,
+      "learning_rate": 6.628206255890235e-07,
+      "loss": 0.9156,
+      "step": 15465
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6711454391479492,
+      "learning_rate": 6.621557440814963e-07,
+      "loss": 0.896,
+      "step": 15466
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7442529201507568,
+      "learning_rate": 6.614911847968875e-07,
+      "loss": 0.907,
+      "step": 15467
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6668674945831299,
+      "learning_rate": 6.60826947758132e-07,
+      "loss": 0.8805,
+      "step": 15468
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.871113657951355,
+      "learning_rate": 6.601630329881525e-07,
+      "loss": 0.8464,
+      "step": 15469
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1420254707336426,
+      "learning_rate": 6.594994405098554e-07,
+      "loss": 0.6099,
+      "step": 15470
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9286190271377563,
+      "learning_rate": 6.588361703461433e-07,
+      "loss": 0.9271,
+      "step": 15471
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7409298419952393,
+      "learning_rate": 6.581732225198989e-07,
+      "loss": 0.9413,
+      "step": 15472
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8043605089187622,
+      "learning_rate": 6.57510597054003e-07,
+      "loss": 0.89,
+      "step": 15473
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6918832063674927,
+      "learning_rate": 6.568482939713172e-07,
+      "loss": 0.8051,
+      "step": 15474
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8590641021728516,
+      "learning_rate": 6.561863132947e-07,
+      "loss": 0.9181,
+      "step": 15475
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6623128652572632,
+      "learning_rate": 6.555246550469907e-07,
+      "loss": 0.9372,
+      "step": 15476
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7385882139205933,
+      "learning_rate": 6.548633192510234e-07,
+      "loss": 0.9223,
+      "step": 15477
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0345485210418701,
+      "learning_rate": 6.542023059296176e-07,
+      "loss": 0.5543,
+      "step": 15478
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7842947244644165,
+      "learning_rate": 6.53541615105584e-07,
+      "loss": 0.8839,
+      "step": 15479
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7564575672149658,
+      "learning_rate": 6.528812468017221e-07,
+      "loss": 0.9038,
+      "step": 15480
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7557716369628906,
+      "learning_rate": 6.522212010408168e-07,
+      "loss": 0.934,
+      "step": 15481
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8131500482559204,
+      "learning_rate": 6.51561477845648e-07,
+      "loss": 0.8901,
+      "step": 15482
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0179673433303833,
+      "learning_rate": 6.509020772389763e-07,
+      "loss": 0.5193,
+      "step": 15483
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9805660247802734,
+      "learning_rate": 6.502429992435599e-07,
+      "loss": 0.9208,
+      "step": 15484
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8453505039215088,
+      "learning_rate": 6.495842438821387e-07,
+      "loss": 0.871,
+      "step": 15485
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7040224075317383,
+      "learning_rate": 6.489258111774477e-07,
+      "loss": 0.8238,
+      "step": 15486
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7427481412887573,
+      "learning_rate": 6.482677011522042e-07,
+      "loss": 0.9231,
+      "step": 15487
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.814207673072815,
+      "learning_rate": 6.476099138291192e-07,
+      "loss": 0.9063,
+      "step": 15488
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8178634643554688,
+      "learning_rate": 6.469524492308921e-07,
+      "loss": 0.8599,
+      "step": 15489
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7496860027313232,
+      "learning_rate": 6.462953073802103e-07,
+      "loss": 0.8705,
+      "step": 15490
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7155566215515137,
+      "learning_rate": 6.456384882997468e-07,
+      "loss": 0.9304,
+      "step": 15491
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6795132160186768,
+      "learning_rate": 6.449819920121702e-07,
+      "loss": 0.9265,
+      "step": 15492
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7608604431152344,
+      "learning_rate": 6.443258185401324e-07,
+      "loss": 0.8753,
+      "step": 15493
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.008249282836914,
+      "learning_rate": 6.436699679062775e-07,
+      "loss": 0.9294,
+      "step": 15494
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1016560792922974,
+      "learning_rate": 6.430144401332338e-07,
+      "loss": 0.524,
+      "step": 15495
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6687204837799072,
+      "learning_rate": 6.423592352436248e-07,
+      "loss": 0.8754,
+      "step": 15496
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9280526638031006,
+      "learning_rate": 6.417043532600609e-07,
+      "loss": 0.9332,
+      "step": 15497
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.694471836090088,
+      "learning_rate": 6.410497942051363e-07,
+      "loss": 0.9066,
+      "step": 15498
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8368369340896606,
+      "learning_rate": 6.403955581014421e-07,
+      "loss": 0.8805,
+      "step": 15499
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7158629894256592,
+      "learning_rate": 6.39741644971551e-07,
+      "loss": 0.869,
+      "step": 15500
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7269920110702515,
+      "learning_rate": 6.390880548380296e-07,
+      "loss": 0.9283,
+      "step": 15501
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9694809913635254,
+      "learning_rate": 6.384347877234299e-07,
+      "loss": 0.829,
+      "step": 15502
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.016885280609131,
+      "learning_rate": 6.377818436502969e-07,
+      "loss": 0.9519,
+      "step": 15503
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8114701509475708,
+      "learning_rate": 6.371292226411574e-07,
+      "loss": 0.8159,
+      "step": 15504
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6478521823883057,
+      "learning_rate": 6.364769247185376e-07,
+      "loss": 0.8758,
+      "step": 15505
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.737593650817871,
+      "learning_rate": 6.358249499049407e-07,
+      "loss": 0.852,
+      "step": 15506
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7148728370666504,
+      "learning_rate": 6.351732982228687e-07,
+      "loss": 0.8892,
+      "step": 15507
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7233200073242188,
+      "learning_rate": 6.345219696948046e-07,
+      "loss": 0.913,
+      "step": 15508
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8590095043182373,
+      "learning_rate": 6.338709643432261e-07,
+      "loss": 0.9252,
+      "step": 15509
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.750510811805725,
+      "learning_rate": 6.332202821905986e-07,
+      "loss": 0.8538,
+      "step": 15510
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.114027500152588,
+      "learning_rate": 6.32569923259373e-07,
+      "loss": 0.5378,
+      "step": 15511
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7650784254074097,
+      "learning_rate": 6.319198875719945e-07,
+      "loss": 0.8798,
+      "step": 15512
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8563258647918701,
+      "learning_rate": 6.312701751508898e-07,
+      "loss": 0.8992,
+      "step": 15513
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9289206266403198,
+      "learning_rate": 6.306207860184832e-07,
+      "loss": 0.9787,
+      "step": 15514
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.861488938331604,
+      "learning_rate": 6.299717201971789e-07,
+      "loss": 0.8757,
+      "step": 15515
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8455007076263428,
+      "learning_rate": 6.293229777093779e-07,
+      "loss": 0.97,
+      "step": 15516
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9264119863510132,
+      "learning_rate": 6.286745585774634e-07,
+      "loss": 0.9504,
+      "step": 15517
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9496403336524963,
+      "learning_rate": 6.28026462823813e-07,
+      "loss": 0.4688,
+      "step": 15518
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7758005857467651,
+      "learning_rate": 6.273786904707901e-07,
+      "loss": 0.9176,
+      "step": 15519
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9176400899887085,
+      "learning_rate": 6.267312415407478e-07,
+      "loss": 0.905,
+      "step": 15520
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.701231837272644,
+      "learning_rate": 6.260841160560249e-07,
+      "loss": 0.882,
+      "step": 15521
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.010457158088684,
+      "learning_rate": 6.254373140389546e-07,
+      "loss": 0.4875,
+      "step": 15522
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8519154787063599,
+      "learning_rate": 6.247908355118581e-07,
+      "loss": 0.8354,
+      "step": 15523
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8883432149887085,
+      "learning_rate": 6.241446804970397e-07,
+      "loss": 0.9274,
+      "step": 15524
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.86014986038208,
+      "learning_rate": 6.234988490167981e-07,
+      "loss": 0.9595,
+      "step": 15525
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6856034994125366,
+      "learning_rate": 6.22853341093419e-07,
+      "loss": 0.8962,
+      "step": 15526
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8290272951126099,
+      "learning_rate": 6.222081567491778e-07,
+      "loss": 0.8819,
+      "step": 15527
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7978019714355469,
+      "learning_rate": 6.215632960063367e-07,
+      "loss": 0.8613,
+      "step": 15528
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7104436159133911,
+      "learning_rate": 6.20918758887149e-07,
+      "loss": 0.9095,
+      "step": 15529
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8589082956314087,
+      "learning_rate": 6.202745454138548e-07,
+      "loss": 0.9629,
+      "step": 15530
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.744292974472046,
+      "learning_rate": 6.196306556086862e-07,
+      "loss": 0.9372,
+      "step": 15531
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8318431377410889,
+      "learning_rate": 6.189870894938587e-07,
+      "loss": 0.8565,
+      "step": 15532
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.932905912399292,
+      "learning_rate": 6.183438470915826e-07,
+      "loss": 0.8697,
+      "step": 15533
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9996007084846497,
+      "learning_rate": 6.177009284240542e-07,
+      "loss": 0.5517,
+      "step": 15534
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.324336528778076,
+      "learning_rate": 6.170583335134584e-07,
+      "loss": 0.9225,
+      "step": 15535
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6341477632522583,
+      "learning_rate": 6.164160623819693e-07,
+      "loss": 0.8925,
+      "step": 15536
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.696211814880371,
+      "learning_rate": 6.157741150517494e-07,
+      "loss": 0.8662,
+      "step": 15537
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6755071878433228,
+      "learning_rate": 6.15132491544952e-07,
+      "loss": 0.9652,
+      "step": 15538
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8382385969161987,
+      "learning_rate": 6.14491191883716e-07,
+      "loss": 0.9468,
+      "step": 15539
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6141189336776733,
+      "learning_rate": 6.138502160901727e-07,
+      "loss": 0.849,
+      "step": 15540
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.722029447555542,
+      "learning_rate": 6.132095641864378e-07,
+      "loss": 0.8759,
+      "step": 15541
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6465224027633667,
+      "learning_rate": 6.125692361946211e-07,
+      "loss": 0.8796,
+      "step": 15542
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.679895281791687,
+      "learning_rate": 6.119292321368153e-07,
+      "loss": 0.8483,
+      "step": 15543
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8032406568527222,
+      "learning_rate": 6.112895520351103e-07,
+      "loss": 0.9443,
+      "step": 15544
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6419126987457275,
+      "learning_rate": 6.10650195911574e-07,
+      "loss": 0.8833,
+      "step": 15545
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5212006568908691,
+      "learning_rate": 6.100111637882711e-07,
+      "loss": 0.765,
+      "step": 15546
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.759412407875061,
+      "learning_rate": 6.093724556872549e-07,
+      "loss": 0.9028,
+      "step": 15547
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0550742149353027,
+      "learning_rate": 6.087340716305623e-07,
+      "loss": 0.5482,
+      "step": 15548
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.059455156326294,
+      "learning_rate": 6.080960116402245e-07,
+      "loss": 0.9912,
+      "step": 15549
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.888584017753601,
+      "learning_rate": 6.074582757382575e-07,
+      "loss": 0.8927,
+      "step": 15550
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.664782166481018,
+      "learning_rate": 6.068208639466688e-07,
+      "loss": 0.9334,
+      "step": 15551
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.723462700843811,
+      "learning_rate": 6.061837762874523e-07,
+      "loss": 0.8874,
+      "step": 15552
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8481340408325195,
+      "learning_rate": 6.055470127825946e-07,
+      "loss": 0.88,
+      "step": 15553
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0063462257385254,
+      "learning_rate": 6.049105734540639e-07,
+      "loss": 0.4903,
+      "step": 15554
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6952629089355469,
+      "learning_rate": 6.042744583238291e-07,
+      "loss": 0.8773,
+      "step": 15555
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.717790961265564,
+      "learning_rate": 6.036386674138339e-07,
+      "loss": 0.9307,
+      "step": 15556
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.775532841682434,
+      "learning_rate": 6.030032007460229e-07,
+      "loss": 0.8684,
+      "step": 15557
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.68381667137146,
+      "learning_rate": 6.023680583423209e-07,
+      "loss": 0.8578,
+      "step": 15558
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6865746974945068,
+      "learning_rate": 6.017332402246468e-07,
+      "loss": 0.8708,
+      "step": 15559
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9333467483520508,
+      "learning_rate": 6.010987464149043e-07,
+      "loss": 0.8167,
+      "step": 15560
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.737566590309143,
+      "learning_rate": 6.004645769349915e-07,
+      "loss": 0.9618,
+      "step": 15561
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7951821088790894,
+      "learning_rate": 5.998307318067875e-07,
+      "loss": 0.9258,
+      "step": 15562
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.781555414199829,
+      "learning_rate": 5.99197211052166e-07,
+      "loss": 0.8705,
+      "step": 15563
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.73175048828125,
+      "learning_rate": 5.985640146929906e-07,
+      "loss": 0.8543,
+      "step": 15564
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5270600318908691,
+      "learning_rate": 5.979311427511081e-07,
+      "loss": 0.8526,
+      "step": 15565
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8287838697433472,
+      "learning_rate": 5.972985952483601e-07,
+      "loss": 0.9412,
+      "step": 15566
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6838566064834595,
+      "learning_rate": 5.966663722065691e-07,
+      "loss": 0.8423,
+      "step": 15567
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7391963005065918,
+      "learning_rate": 5.960344736475576e-07,
+      "loss": 0.8679,
+      "step": 15568
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6428356170654297,
+      "learning_rate": 5.95402899593125e-07,
+      "loss": 0.8845,
+      "step": 15569
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8367490768432617,
+      "learning_rate": 5.947716500650702e-07,
+      "loss": 0.9521,
+      "step": 15570
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7687405347824097,
+      "learning_rate": 5.941407250851705e-07,
+      "loss": 0.9225,
+      "step": 15571
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8504630327224731,
+      "learning_rate": 5.935101246752029e-07,
+      "loss": 0.8998,
+      "step": 15572
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8147363662719727,
+      "learning_rate": 5.928798488569221e-07,
+      "loss": 0.9059,
+      "step": 15573
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8694202899932861,
+      "learning_rate": 5.922498976520818e-07,
+      "loss": 0.9067,
+      "step": 15574
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6132831573486328,
+      "learning_rate": 5.916202710824171e-07,
+      "loss": 0.9781,
+      "step": 15575
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8362089395523071,
+      "learning_rate": 5.909909691696558e-07,
+      "loss": 0.8844,
+      "step": 15576
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.705001711845398,
+      "learning_rate": 5.903619919355141e-07,
+      "loss": 0.8366,
+      "step": 15577
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8806865215301514,
+      "learning_rate": 5.897333394016935e-07,
+      "loss": 0.8806,
+      "step": 15578
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7190494537353516,
+      "learning_rate": 5.891050115898911e-07,
+      "loss": 0.8013,
+      "step": 15579
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.841902494430542,
+      "learning_rate": 5.88477008521785e-07,
+      "loss": 0.9714,
+      "step": 15580
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7925434112548828,
+      "learning_rate": 5.87849330219048e-07,
+      "loss": 0.8926,
+      "step": 15581
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7538598775863647,
+      "learning_rate": 5.872219767033382e-07,
+      "loss": 0.8815,
+      "step": 15582
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8988202810287476,
+      "learning_rate": 5.865949479963052e-07,
+      "loss": 0.9301,
+      "step": 15583
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6966590881347656,
+      "learning_rate": 5.859682441195846e-07,
+      "loss": 0.9732,
+      "step": 15584
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8185322284698486,
+      "learning_rate": 5.853418650948039e-07,
+      "loss": 0.8626,
+      "step": 15585
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0373144149780273,
+      "learning_rate": 5.847158109435746e-07,
+      "loss": 0.5108,
+      "step": 15586
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8377195596694946,
+      "learning_rate": 5.840900816875028e-07,
+      "loss": 0.9463,
+      "step": 15587
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6262853145599365,
+      "learning_rate": 5.834646773481811e-07,
+      "loss": 0.9424,
+      "step": 15588
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.780945062637329,
+      "learning_rate": 5.82839597947189e-07,
+      "loss": 0.916,
+      "step": 15589
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8211147785186768,
+      "learning_rate": 5.822148435060971e-07,
+      "loss": 0.8665,
+      "step": 15590
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5948418378829956,
+      "learning_rate": 5.815904140464623e-07,
+      "loss": 0.8741,
+      "step": 15591
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6296415328979492,
+      "learning_rate": 5.809663095898332e-07,
+      "loss": 0.8931,
+      "step": 15592
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8155616521835327,
+      "learning_rate": 5.803425301577459e-07,
+      "loss": 0.887,
+      "step": 15593
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7767034769058228,
+      "learning_rate": 5.797190757717264e-07,
+      "loss": 0.8596,
+      "step": 15594
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8372561931610107,
+      "learning_rate": 5.790959464532852e-07,
+      "loss": 0.9633,
+      "step": 15595
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.804306149482727,
+      "learning_rate": 5.784731422239276e-07,
+      "loss": 0.9006,
+      "step": 15596
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8901151418685913,
+      "learning_rate": 5.77850663105144e-07,
+      "loss": 0.9367,
+      "step": 15597
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6477718353271484,
+      "learning_rate": 5.772285091184138e-07,
+      "loss": 0.8859,
+      "step": 15598
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7240761518478394,
+      "learning_rate": 5.766066802852066e-07,
+      "loss": 0.9518,
+      "step": 15599
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9692060947418213,
+      "learning_rate": 5.759851766269786e-07,
+      "loss": 0.9452,
+      "step": 15600
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8080519437789917,
+      "learning_rate": 5.753639981651792e-07,
+      "loss": 0.9087,
+      "step": 15601
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6249909400939941,
+      "learning_rate": 5.747431449212393e-07,
+      "loss": 0.8644,
+      "step": 15602
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7710704803466797,
+      "learning_rate": 5.74122616916587e-07,
+      "loss": 0.8791,
+      "step": 15603
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1158372163772583,
+      "learning_rate": 5.73502414172632e-07,
+      "loss": 0.5356,
+      "step": 15604
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8361241817474365,
+      "learning_rate": 5.728825367107782e-07,
+      "loss": 0.9456,
+      "step": 15605
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.802695393562317,
+      "learning_rate": 5.722629845524131e-07,
+      "loss": 1.0117,
+      "step": 15606
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7006168365478516,
+      "learning_rate": 5.716437577189182e-07,
+      "loss": 0.8232,
+      "step": 15607
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8938943147659302,
+      "learning_rate": 5.710248562316589e-07,
+      "loss": 0.9121,
+      "step": 15608
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7587521076202393,
+      "learning_rate": 5.704062801119947e-07,
+      "loss": 0.8805,
+      "step": 15609
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.756843090057373,
+      "learning_rate": 5.697880293812674e-07,
+      "loss": 0.9176,
+      "step": 15610
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7367825508117676,
+      "learning_rate": 5.691701040608133e-07,
+      "loss": 0.8907,
+      "step": 15611
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7477002143859863,
+      "learning_rate": 5.685525041719553e-07,
+      "loss": 0.8997,
+      "step": 15612
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8525936603546143,
+      "learning_rate": 5.679352297360041e-07,
+      "loss": 0.9056,
+      "step": 15613
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9039686918258667,
+      "learning_rate": 5.673182807742627e-07,
+      "loss": 0.9175,
+      "step": 15614
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.80568265914917,
+      "learning_rate": 5.667016573080164e-07,
+      "loss": 0.9697,
+      "step": 15615
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8497459888458252,
+      "learning_rate": 5.660853593585458e-07,
+      "loss": 0.8802,
+      "step": 15616
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7925468683242798,
+      "learning_rate": 5.654693869471162e-07,
+      "loss": 0.8679,
+      "step": 15617
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8453634977340698,
+      "learning_rate": 5.64853740094985e-07,
+      "loss": 0.9208,
+      "step": 15618
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7078968286514282,
+      "learning_rate": 5.64238418823394e-07,
+      "loss": 0.8676,
+      "step": 15619
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.749570608139038,
+      "learning_rate": 5.636234231535775e-07,
+      "loss": 0.8422,
+      "step": 15620
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.775765299797058,
+      "learning_rate": 5.630087531067574e-07,
+      "loss": 0.7934,
+      "step": 15621
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7809996604919434,
+      "learning_rate": 5.623944087041444e-07,
+      "loss": 0.8311,
+      "step": 15622
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8752776384353638,
+      "learning_rate": 5.617803899669372e-07,
+      "loss": 0.9098,
+      "step": 15623
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9358185529708862,
+      "learning_rate": 5.611666969163243e-07,
+      "loss": 0.5014,
+      "step": 15624
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7852346897125244,
+      "learning_rate": 5.605533295734822e-07,
+      "loss": 0.957,
+      "step": 15625
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7740525007247925,
+      "learning_rate": 5.599402879595772e-07,
+      "loss": 0.8256,
+      "step": 15626
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8986852169036865,
+      "learning_rate": 5.593275720957625e-07,
+      "loss": 0.9134,
+      "step": 15627
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7122132778167725,
+      "learning_rate": 5.587151820031811e-07,
+      "loss": 0.8797,
+      "step": 15628
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8016579151153564,
+      "learning_rate": 5.581031177029672e-07,
+      "loss": 0.8761,
+      "step": 15629
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7391622066497803,
+      "learning_rate": 5.574913792162395e-07,
+      "loss": 0.9061,
+      "step": 15630
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8038105964660645,
+      "learning_rate": 5.568799665641078e-07,
+      "loss": 0.9224,
+      "step": 15631
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9782445430755615,
+      "learning_rate": 5.562688797676696e-07,
+      "loss": 0.9664,
+      "step": 15632
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.905548334121704,
+      "learning_rate": 5.556581188480126e-07,
+      "loss": 0.8787,
+      "step": 15633
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6789883375167847,
+      "learning_rate": 5.55047683826212e-07,
+      "loss": 0.8991,
+      "step": 15634
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.811604380607605,
+      "learning_rate": 5.544375747233333e-07,
+      "loss": 0.9344,
+      "step": 15635
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6902881860733032,
+      "learning_rate": 5.538277915604273e-07,
+      "loss": 0.8849,
+      "step": 15636
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.715160846710205,
+      "learning_rate": 5.532183343585396e-07,
+      "loss": 0.9737,
+      "step": 15637
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0077650547027588,
+      "learning_rate": 5.526092031386965e-07,
+      "loss": 0.5393,
+      "step": 15638
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7462610006332397,
+      "learning_rate": 5.520003979219202e-07,
+      "loss": 0.9459,
+      "step": 15639
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.730873703956604,
+      "learning_rate": 5.513919187292182e-07,
+      "loss": 0.8546,
+      "step": 15640
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6832468509674072,
+      "learning_rate": 5.507837655815873e-07,
+      "loss": 0.9624,
+      "step": 15641
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0448713302612305,
+      "learning_rate": 5.501759385000138e-07,
+      "loss": 0.9672,
+      "step": 15642
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6207120418548584,
+      "learning_rate": 5.495684375054711e-07,
+      "loss": 0.9207,
+      "step": 15643
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.773476004600525,
+      "learning_rate": 5.489612626189245e-07,
+      "loss": 0.8333,
+      "step": 15644
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7257273197174072,
+      "learning_rate": 5.483544138613217e-07,
+      "loss": 0.8128,
+      "step": 15645
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.684701919555664,
+      "learning_rate": 5.477478912536083e-07,
+      "loss": 0.8832,
+      "step": 15646
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9602268934249878,
+      "learning_rate": 5.471416948167107e-07,
+      "loss": 0.9757,
+      "step": 15647
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7607377767562866,
+      "learning_rate": 5.46535824571548e-07,
+      "loss": 0.9789,
+      "step": 15648
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.098794460296631,
+      "learning_rate": 5.459302805390254e-07,
+      "loss": 0.9081,
+      "step": 15649
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7969435453414917,
+      "learning_rate": 5.453250627400419e-07,
+      "loss": 0.9311,
+      "step": 15650
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7865899801254272,
+      "learning_rate": 5.447201711954775e-07,
+      "loss": 0.9158,
+      "step": 15651
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7946916818618774,
+      "learning_rate": 5.441156059262109e-07,
+      "loss": 0.8306,
+      "step": 15652
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.000335693359375,
+      "learning_rate": 5.435113669530978e-07,
+      "loss": 0.9067,
+      "step": 15653
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7875988483428955,
+      "learning_rate": 5.429074542969926e-07,
+      "loss": 0.8493,
+      "step": 15654
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.774566411972046,
+      "learning_rate": 5.423038679787352e-07,
+      "loss": 0.8609,
+      "step": 15655
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.794792890548706,
+      "learning_rate": 5.417006080191501e-07,
+      "loss": 0.9016,
+      "step": 15656
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8023427724838257,
+      "learning_rate": 5.410976744390584e-07,
+      "loss": 0.931,
+      "step": 15657
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.76619291305542,
+      "learning_rate": 5.404950672592623e-07,
+      "loss": 0.9598,
+      "step": 15658
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7884960174560547,
+      "learning_rate": 5.398927865005588e-07,
+      "loss": 1.0358,
+      "step": 15659
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6713100671768188,
+      "learning_rate": 5.392908321837275e-07,
+      "loss": 0.8963,
+      "step": 15660
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.721921682357788,
+      "learning_rate": 5.386892043295433e-07,
+      "loss": 0.8999,
+      "step": 15661
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7572952508926392,
+      "learning_rate": 5.380879029587649e-07,
+      "loss": 0.8587,
+      "step": 15662
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9742989540100098,
+      "learning_rate": 5.374869280921436e-07,
+      "loss": 0.8605,
+      "step": 15663
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6886658668518066,
+      "learning_rate": 5.368862797504149e-07,
+      "loss": 0.8876,
+      "step": 15664
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6170834302902222,
+      "learning_rate": 5.362859579543056e-07,
+      "loss": 0.8683,
+      "step": 15665
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8819648027420044,
+      "learning_rate": 5.356859627245337e-07,
+      "loss": 0.8384,
+      "step": 15666
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7028852701187134,
+      "learning_rate": 5.350862940818014e-07,
+      "loss": 0.9006,
+      "step": 15667
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5484930276870728,
+      "learning_rate": 5.344869520468021e-07,
+      "loss": 0.9625,
+      "step": 15668
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7978967428207397,
+      "learning_rate": 5.338879366402161e-07,
+      "loss": 0.9439,
+      "step": 15669
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8412784337997437,
+      "learning_rate": 5.332892478827168e-07,
+      "loss": 0.8419,
+      "step": 15670
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6158326864242554,
+      "learning_rate": 5.326908857949586e-07,
+      "loss": 0.9335,
+      "step": 15671
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9648857116699219,
+      "learning_rate": 5.320928503975953e-07,
+      "loss": 0.9231,
+      "step": 15672
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6637235879898071,
+      "learning_rate": 5.31495141711258e-07,
+      "loss": 0.8077,
+      "step": 15673
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.794452428817749,
+      "learning_rate": 5.308977597565756e-07,
+      "loss": 0.8253,
+      "step": 15674
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7804847955703735,
+      "learning_rate": 5.303007045541586e-07,
+      "loss": 0.8895,
+      "step": 15675
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9212371110916138,
+      "learning_rate": 5.297039761246137e-07,
+      "loss": 0.8571,
+      "step": 15676
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0678497552871704,
+      "learning_rate": 5.291075744885288e-07,
+      "loss": 0.531,
+      "step": 15677
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7733186483383179,
+      "learning_rate": 5.285114996664864e-07,
+      "loss": 0.9076,
+      "step": 15678
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7166513204574585,
+      "learning_rate": 5.279157516790545e-07,
+      "loss": 0.8949,
+      "step": 15679
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7937605381011963,
+      "learning_rate": 5.273203305467911e-07,
+      "loss": 0.8907,
+      "step": 15680
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9153492450714111,
+      "learning_rate": 5.267252362902431e-07,
+      "loss": 0.936,
+      "step": 15681
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.749113917350769,
+      "learning_rate": 5.261304689299429e-07,
+      "loss": 0.9013,
+      "step": 15682
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0417803525924683,
+      "learning_rate": 5.255360284864175e-07,
+      "loss": 0.5309,
+      "step": 15683
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7653416395187378,
+      "learning_rate": 5.24941914980176e-07,
+      "loss": 0.8583,
+      "step": 15684
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0027296543121338,
+      "learning_rate": 5.243481284317232e-07,
+      "loss": 0.5005,
+      "step": 15685
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.86319100856781,
+      "learning_rate": 5.237546688615447e-07,
+      "loss": 0.9205,
+      "step": 15686
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.840283989906311,
+      "learning_rate": 5.231615362901255e-07,
+      "loss": 0.9122,
+      "step": 15687
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8656138181686401,
+      "learning_rate": 5.225687307379268e-07,
+      "loss": 0.8741,
+      "step": 15688
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9472781419754028,
+      "learning_rate": 5.219762522254079e-07,
+      "loss": 0.9382,
+      "step": 15689
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.863309621810913,
+      "learning_rate": 5.213841007730125e-07,
+      "loss": 0.9011,
+      "step": 15690
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8095654249191284,
+      "learning_rate": 5.207922764011752e-07,
+      "loss": 0.8874,
+      "step": 15691
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7858742475509644,
+      "learning_rate": 5.202007791303165e-07,
+      "loss": 0.9017,
+      "step": 15692
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7495330572128296,
+      "learning_rate": 5.196096089808489e-07,
+      "loss": 0.8601,
+      "step": 15693
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9624909162521362,
+      "learning_rate": 5.190187659731705e-07,
+      "loss": 0.8931,
+      "step": 15694
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7143641710281372,
+      "learning_rate": 5.184282501276694e-07,
+      "loss": 0.8443,
+      "step": 15695
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.853432059288025,
+      "learning_rate": 5.17838061464726e-07,
+      "loss": 0.9205,
+      "step": 15696
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8027478456497192,
+      "learning_rate": 5.172482000047019e-07,
+      "loss": 0.9873,
+      "step": 15697
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9007970094680786,
+      "learning_rate": 5.166586657679551e-07,
+      "loss": 0.8908,
+      "step": 15698
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1118667125701904,
+      "learning_rate": 5.16069458774825e-07,
+      "loss": 1.0109,
+      "step": 15699
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.816728949546814,
+      "learning_rate": 5.154805790456486e-07,
+      "loss": 0.9295,
+      "step": 15700
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8203498125076294,
+      "learning_rate": 5.148920266007407e-07,
+      "loss": 0.9413,
+      "step": 15701
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8068729639053345,
+      "learning_rate": 5.143038014604152e-07,
+      "loss": 0.8825,
+      "step": 15702
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.86770498752594,
+      "learning_rate": 5.137159036449668e-07,
+      "loss": 0.9791,
+      "step": 15703
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8436784744262695,
+      "learning_rate": 5.131283331746851e-07,
+      "loss": 0.9117,
+      "step": 15704
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8742127418518066,
+      "learning_rate": 5.125410900698425e-07,
+      "loss": 0.9951,
+      "step": 15705
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.696946144104004,
+      "learning_rate": 5.119541743507062e-07,
+      "loss": 0.9304,
+      "step": 15706
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7339062690734863,
+      "learning_rate": 5.113675860375267e-07,
+      "loss": 0.8848,
+      "step": 15707
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8559589385986328,
+      "learning_rate": 5.107813251505456e-07,
+      "loss": 0.9666,
+      "step": 15708
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7508392333984375,
+      "learning_rate": 5.101953917099955e-07,
+      "loss": 0.8681,
+      "step": 15709
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6933382749557495,
+      "learning_rate": 5.096097857360927e-07,
+      "loss": 0.8133,
+      "step": 15710
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6814154386520386,
+      "learning_rate": 5.090245072490474e-07,
+      "loss": 0.9545,
+      "step": 15711
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.718481183052063,
+      "learning_rate": 5.084395562690525e-07,
+      "loss": 0.8492,
+      "step": 15712
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8031295537948608,
+      "learning_rate": 5.078549328162963e-07,
+      "loss": 0.882,
+      "step": 15713
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9920603036880493,
+      "learning_rate": 5.072706369109504e-07,
+      "loss": 0.8628,
+      "step": 15714
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0501608848571777,
+      "learning_rate": 5.066866685731786e-07,
+      "loss": 0.8813,
+      "step": 15715
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6383816003799438,
+      "learning_rate": 5.061030278231305e-07,
+      "loss": 0.8585,
+      "step": 15716
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.768262267112732,
+      "learning_rate": 5.055197146809476e-07,
+      "loss": 0.9008,
+      "step": 15717
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8149127960205078,
+      "learning_rate": 5.049367291667572e-07,
+      "loss": 0.8396,
+      "step": 15718
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.70863938331604,
+      "learning_rate": 5.043540713006756e-07,
+      "loss": 0.9565,
+      "step": 15719
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7773997783660889,
+      "learning_rate": 5.03771741102812e-07,
+      "loss": 0.927,
+      "step": 15720
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8717988729476929,
+      "learning_rate": 5.031897385932582e-07,
+      "loss": 0.8581,
+      "step": 15721
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8253116607666016,
+      "learning_rate": 5.026080637920983e-07,
+      "loss": 0.8837,
+      "step": 15722
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9104567766189575,
+      "learning_rate": 5.020267167194038e-07,
+      "loss": 0.9371,
+      "step": 15723
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9085501432418823,
+      "learning_rate": 5.014456973952375e-07,
+      "loss": 0.8923,
+      "step": 15724
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8560725450515747,
+      "learning_rate": 5.008650058396448e-07,
+      "loss": 0.8573,
+      "step": 15725
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8208476305007935,
+      "learning_rate": 5.00284642072667e-07,
+      "loss": 0.8889,
+      "step": 15726
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.888777494430542,
+      "learning_rate": 4.997046061143296e-07,
+      "loss": 0.9014,
+      "step": 15727
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.78938889503479,
+      "learning_rate": 4.991248979846486e-07,
+      "loss": 0.8624,
+      "step": 15728
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7296125888824463,
+      "learning_rate": 4.985455177036269e-07,
+      "loss": 0.8509,
+      "step": 15729
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6556065082550049,
+      "learning_rate": 4.979664652912597e-07,
+      "loss": 0.9032,
+      "step": 15730
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6765328645706177,
+      "learning_rate": 4.973877407675253e-07,
+      "loss": 0.9402,
+      "step": 15731
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7139443159103394,
+      "learning_rate": 4.968093441523958e-07,
+      "loss": 0.8296,
+      "step": 15732
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7192199230194092,
+      "learning_rate": 4.962312754658305e-07,
+      "loss": 0.8702,
+      "step": 15733
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6450542211532593,
+      "learning_rate": 4.956535347277758e-07,
+      "loss": 0.9321,
+      "step": 15734
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8065743446350098,
+      "learning_rate": 4.95076121958169e-07,
+      "loss": 0.9276,
+      "step": 15735
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7082480192184448,
+      "learning_rate": 4.944990371769331e-07,
+      "loss": 0.8433,
+      "step": 15736
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7138797044754028,
+      "learning_rate": 4.939222804039833e-07,
+      "loss": 0.859,
+      "step": 15737
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.725624680519104,
+      "learning_rate": 4.933458516592216e-07,
+      "loss": 0.9419,
+      "step": 15738
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.776052713394165,
+      "learning_rate": 4.927697509625396e-07,
+      "loss": 0.8684,
+      "step": 15739
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7903958559036255,
+      "learning_rate": 4.921939783338137e-07,
+      "loss": 1.014,
+      "step": 15740
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7034764289855957,
+      "learning_rate": 4.916185337929169e-07,
+      "loss": 0.8972,
+      "step": 15741
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.716565728187561,
+      "learning_rate": 4.910434173597023e-07,
+      "loss": 0.9402,
+      "step": 15742
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8492597341537476,
+      "learning_rate": 4.904686290540184e-07,
+      "loss": 0.9752,
+      "step": 15743
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7823865413665771,
+      "learning_rate": 4.898941688956981e-07,
+      "loss": 0.8729,
+      "step": 15744
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.696645975112915,
+      "learning_rate": 4.893200369045636e-07,
+      "loss": 0.9579,
+      "step": 15745
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.059925079345703,
+      "learning_rate": 4.8874623310043e-07,
+      "loss": 0.9037,
+      "step": 15746
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6956720352172852,
+      "learning_rate": 4.881727575030926e-07,
+      "loss": 0.8757,
+      "step": 15747
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6883294582366943,
+      "learning_rate": 4.875996101323455e-07,
+      "loss": 0.8769,
+      "step": 15748
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7607452869415283,
+      "learning_rate": 4.870267910079618e-07,
+      "loss": 0.9245,
+      "step": 15749
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8052239418029785,
+      "learning_rate": 4.864543001497113e-07,
+      "loss": 0.8465,
+      "step": 15750
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9014909267425537,
+      "learning_rate": 4.858821375773471e-07,
+      "loss": 0.8528,
+      "step": 15751
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7933992147445679,
+      "learning_rate": 4.853103033106143e-07,
+      "loss": 0.8965,
+      "step": 15752
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7145254611968994,
+      "learning_rate": 4.847387973692441e-07,
+      "loss": 0.9,
+      "step": 15753
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8281176090240479,
+      "learning_rate": 4.841676197729594e-07,
+      "loss": 0.7973,
+      "step": 15754
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7307642698287964,
+      "learning_rate": 4.835967705414679e-07,
+      "loss": 0.956,
+      "step": 15755
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9950353503227234,
+      "learning_rate": 4.830262496944693e-07,
+      "loss": 0.5799,
+      "step": 15756
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8266314268112183,
+      "learning_rate": 4.824560572516501e-07,
+      "loss": 0.9225,
+      "step": 15757
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0209341049194336,
+      "learning_rate": 4.818861932326868e-07,
+      "loss": 0.9647,
+      "step": 15758
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9397742748260498,
+      "learning_rate": 4.813166576572415e-07,
+      "loss": 0.8734,
+      "step": 15759
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6650595664978027,
+      "learning_rate": 4.807474505449705e-07,
+      "loss": 0.8131,
+      "step": 15760
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.929392695426941,
+      "learning_rate": 4.801785719155128e-07,
+      "loss": 0.9177,
+      "step": 15761
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7827253341674805,
+      "learning_rate": 4.796100217885003e-07,
+      "loss": 0.8413,
+      "step": 15762
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8335567712783813,
+      "learning_rate": 4.790418001835529e-07,
+      "loss": 0.8995,
+      "step": 15763
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9441468715667725,
+      "learning_rate": 4.78473907120276e-07,
+      "loss": 0.5192,
+      "step": 15764
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7356173992156982,
+      "learning_rate": 4.779063426182684e-07,
+      "loss": 0.8767,
+      "step": 15765
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8008344173431396,
+      "learning_rate": 4.773391066971134e-07,
+      "loss": 0.9068,
+      "step": 15766
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.128982424736023,
+      "learning_rate": 4.767721993763863e-07,
+      "loss": 0.5993,
+      "step": 15767
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8010064363479614,
+      "learning_rate": 4.7620562067564715e-07,
+      "loss": 0.8734,
+      "step": 15768
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.833927869796753,
+      "learning_rate": 4.756393706144491e-07,
+      "loss": 0.9064,
+      "step": 15769
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9079430103302002,
+      "learning_rate": 4.75073449212331e-07,
+      "loss": 0.9044,
+      "step": 15770
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.869832158088684,
+      "learning_rate": 4.745078564888217e-07,
+      "loss": 0.8867,
+      "step": 15771
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8267983198165894,
+      "learning_rate": 4.7394259246343666e-07,
+      "loss": 0.8493,
+      "step": 15772
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6709568500518799,
+      "learning_rate": 4.7337765715568364e-07,
+      "loss": 0.8659,
+      "step": 15773
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7146183252334595,
+      "learning_rate": 4.728130505850559e-07,
+      "loss": 0.9038,
+      "step": 15774
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7966071367263794,
+      "learning_rate": 4.7224877277103673e-07,
+      "loss": 0.8676,
+      "step": 15775
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.967586636543274,
+      "learning_rate": 4.716848237330984e-07,
+      "loss": 0.9224,
+      "step": 15776
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7624552249908447,
+      "learning_rate": 4.7112120349069976e-07,
+      "loss": 0.9134,
+      "step": 15777
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.844422459602356,
+      "learning_rate": 4.7055791206329194e-07,
+      "loss": 0.914,
+      "step": 15778
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7868367433547974,
+      "learning_rate": 4.699949494703093e-07,
+      "loss": 0.8833,
+      "step": 15779
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7061166763305664,
+      "learning_rate": 4.694323157311809e-07,
+      "loss": 0.9134,
+      "step": 15780
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8165570497512817,
+      "learning_rate": 4.6887001086531994e-07,
+      "loss": 0.9239,
+      "step": 15781
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.007720470428467,
+      "learning_rate": 4.6830803489213206e-07,
+      "loss": 0.8854,
+      "step": 15782
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8768231868743896,
+      "learning_rate": 4.6774638783100625e-07,
+      "loss": 0.9213,
+      "step": 15783
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.754347801208496,
+      "learning_rate": 4.6718506970132803e-07,
+      "loss": 0.8921,
+      "step": 15784
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.823531985282898,
+      "learning_rate": 4.66624080522462e-07,
+      "loss": 0.9453,
+      "step": 15785
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.76090407371521,
+      "learning_rate": 4.6606342031376816e-07,
+      "loss": 0.8468,
+      "step": 15786
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.830092430114746,
+      "learning_rate": 4.6550308909459554e-07,
+      "loss": 0.8878,
+      "step": 15787
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9652304649353027,
+      "learning_rate": 4.6494308688427635e-07,
+      "loss": 0.5434,
+      "step": 15788
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7110916376113892,
+      "learning_rate": 4.6438341370213745e-07,
+      "loss": 0.8519,
+      "step": 15789
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9821134805679321,
+      "learning_rate": 4.638240695674889e-07,
+      "loss": 0.558,
+      "step": 15790
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8891531229019165,
+      "learning_rate": 4.6326505449963535e-07,
+      "loss": 0.8531,
+      "step": 15791
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8857909440994263,
+      "learning_rate": 4.6270636851786234e-07,
+      "loss": 0.9731,
+      "step": 15792
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8339778184890747,
+      "learning_rate": 4.621480116414534e-07,
+      "loss": 0.8119,
+      "step": 15793
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.009507656097412,
+      "learning_rate": 4.6158998388967315e-07,
+      "loss": 0.5773,
+      "step": 15794
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.904154658317566,
+      "learning_rate": 4.6103228528177834e-07,
+      "loss": 0.9033,
+      "step": 15795
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9575434923171997,
+      "learning_rate": 4.6047491583701253e-07,
+      "loss": 0.8975,
+      "step": 15796
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7517492771148682,
+      "learning_rate": 4.5991787557460923e-07,
+      "loss": 0.8387,
+      "step": 15797
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9346876740455627,
+      "learning_rate": 4.5936116451379186e-07,
+      "loss": 0.4791,
+      "step": 15798
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6114987134933472,
+      "learning_rate": 4.588047826737696e-07,
+      "loss": 0.8701,
+      "step": 15799
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7082452774047852,
+      "learning_rate": 4.582487300737437e-07,
+      "loss": 0.8997,
+      "step": 15800
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.760166049003601,
+      "learning_rate": 4.5769300673289776e-07,
+      "loss": 0.9715,
+      "step": 15801
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8668776750564575,
+      "learning_rate": 4.57137612670413e-07,
+      "loss": 0.9953,
+      "step": 15802
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0007483959198,
+      "learning_rate": 4.5658254790545085e-07,
+      "loss": 0.9302,
+      "step": 15803
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6467692852020264,
+      "learning_rate": 4.5602781245716707e-07,
+      "loss": 0.7793,
+      "step": 15804
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8213754892349243,
+      "learning_rate": 4.5547340634470303e-07,
+      "loss": 0.8517,
+      "step": 15805
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7014737129211426,
+      "learning_rate": 4.549193295871912e-07,
+      "loss": 0.8727,
+      "step": 15806
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9187073707580566,
+      "learning_rate": 4.543655822037496e-07,
+      "loss": 0.8901,
+      "step": 15807
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7303494215011597,
+      "learning_rate": 4.538121642134874e-07,
+      "loss": 0.894,
+      "step": 15808
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6147066354751587,
+      "learning_rate": 4.5325907563550045e-07,
+      "loss": 0.7855,
+      "step": 15809
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8874821662902832,
+      "learning_rate": 4.5270631648887455e-07,
+      "loss": 0.9575,
+      "step": 15810
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0284878015518188,
+      "learning_rate": 4.5215388679268556e-07,
+      "loss": 0.535,
+      "step": 15811
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.803733229637146,
+      "learning_rate": 4.5160178656599495e-07,
+      "loss": 0.8653,
+      "step": 15812
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8952652215957642,
+      "learning_rate": 4.510500158278541e-07,
+      "loss": 0.8971,
+      "step": 15813
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7323887348175049,
+      "learning_rate": 4.504985745973034e-07,
+      "loss": 0.8429,
+      "step": 15814
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8402196168899536,
+      "learning_rate": 4.49947462893372e-07,
+      "loss": 0.8667,
+      "step": 15815
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.849502682685852,
+      "learning_rate": 4.493966807350758e-07,
+      "loss": 0.8452,
+      "step": 15816
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7073493003845215,
+      "learning_rate": 4.4884622814142187e-07,
+      "loss": 0.8369,
+      "step": 15817
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6291992664337158,
+      "learning_rate": 4.482961051314028e-07,
+      "loss": 0.8464,
+      "step": 15818
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0040384531021118,
+      "learning_rate": 4.4774631172400663e-07,
+      "loss": 0.4712,
+      "step": 15819
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9278392791748047,
+      "learning_rate": 4.471968479381994e-07,
+      "loss": 0.8817,
+      "step": 15820
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.801424503326416,
+      "learning_rate": 4.4664771379294704e-07,
+      "loss": 0.8682,
+      "step": 15821
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7981007099151611,
+      "learning_rate": 4.4609890930719324e-07,
+      "loss": 0.8557,
+      "step": 15822
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8545074462890625,
+      "learning_rate": 4.455504344998807e-07,
+      "loss": 0.9202,
+      "step": 15823
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7270314693450928,
+      "learning_rate": 4.4500228938993195e-07,
+      "loss": 0.9089,
+      "step": 15824
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9364813566207886,
+      "learning_rate": 4.444544739962642e-07,
+      "loss": 0.4811,
+      "step": 15825
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7959704399108887,
+      "learning_rate": 4.439069883377789e-07,
+      "loss": 0.8656,
+      "step": 15826
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.761427402496338,
+      "learning_rate": 4.433598324333699e-07,
+      "loss": 0.9237,
+      "step": 15827
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7590771913528442,
+      "learning_rate": 4.428130063019187e-07,
+      "loss": 0.7513,
+      "step": 15828
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0862759351730347,
+      "learning_rate": 4.4226650996229247e-07,
+      "loss": 0.5485,
+      "step": 15829
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6861910820007324,
+      "learning_rate": 4.417203434333517e-07,
+      "loss": 0.9668,
+      "step": 15830
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7583849430084229,
+      "learning_rate": 4.4117450673394125e-07,
+      "loss": 0.9017,
+      "step": 15831
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9644854068756104,
+      "learning_rate": 4.406289998828972e-07,
+      "loss": 0.8992,
+      "step": 15832
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.851729154586792,
+      "learning_rate": 4.4008382289904337e-07,
+      "loss": 0.9028,
+      "step": 15833
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8522791862487793,
+      "learning_rate": 4.3953897580119255e-07,
+      "loss": 0.7929,
+      "step": 15834
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.827263593673706,
+      "learning_rate": 4.3899445860814516e-07,
+      "loss": 0.997,
+      "step": 15835
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.778059959411621,
+      "learning_rate": 4.384502713386918e-07,
+      "loss": 0.8277,
+      "step": 15836
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.747892141342163,
+      "learning_rate": 4.379064140116096e-07,
+      "loss": 0.91,
+      "step": 15837
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7599698305130005,
+      "learning_rate": 4.3736288664566805e-07,
+      "loss": 0.9697,
+      "step": 15838
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8990840911865234,
+      "learning_rate": 4.3681968925961994e-07,
+      "loss": 0.9091,
+      "step": 15839
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.856643795967102,
+      "learning_rate": 4.362768218722102e-07,
+      "loss": 0.8383,
+      "step": 15840
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8139196634292603,
+      "learning_rate": 4.3573428450217394e-07,
+      "loss": 0.9175,
+      "step": 15841
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7533811330795288,
+      "learning_rate": 4.351920771682283e-07,
+      "loss": 0.8531,
+      "step": 15842
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5952285528182983,
+      "learning_rate": 4.346501998890884e-07,
+      "loss": 0.8685,
+      "step": 15843
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.837520956993103,
+      "learning_rate": 4.341086526834493e-07,
+      "loss": 0.9518,
+      "step": 15844
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9442716836929321,
+      "learning_rate": 4.335674355699992e-07,
+      "loss": 0.4729,
+      "step": 15845
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7076964378356934,
+      "learning_rate": 4.3302654856741343e-07,
+      "loss": 0.8808,
+      "step": 15846
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8478697538375854,
+      "learning_rate": 4.324859916943591e-07,
+      "loss": 0.8889,
+      "step": 15847
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7950849533081055,
+      "learning_rate": 4.3194576496948584e-07,
+      "loss": 0.7923,
+      "step": 15848
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.90106999874115,
+      "learning_rate": 4.314058684114386e-07,
+      "loss": 0.9557,
+      "step": 15849
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8818327188491821,
+      "learning_rate": 4.3086630203884374e-07,
+      "loss": 0.7985,
+      "step": 15850
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.635120153427124,
+      "learning_rate": 4.3032706587032403e-07,
+      "loss": 0.8484,
+      "step": 15851
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0191941261291504,
+      "learning_rate": 4.297881599244857e-07,
+      "loss": 0.516,
+      "step": 15852
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6411631107330322,
+      "learning_rate": 4.2924958421992293e-07,
+      "loss": 0.8478,
+      "step": 15853
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7756330966949463,
+      "learning_rate": 4.28711338775224e-07,
+      "loss": 0.8757,
+      "step": 15854
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7151927947998047,
+      "learning_rate": 4.2817342360895965e-07,
+      "loss": 0.7637,
+      "step": 15855
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9657961130142212,
+      "learning_rate": 4.2763583873969394e-07,
+      "loss": 0.8647,
+      "step": 15856
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9946190118789673,
+      "learning_rate": 4.270985841859743e-07,
+      "loss": 0.893,
+      "step": 15857
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.854508638381958,
+      "learning_rate": 4.2656165996634247e-07,
+      "loss": 0.9895,
+      "step": 15858
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8718602657318115,
+      "learning_rate": 4.2602506609932484e-07,
+      "loss": 0.9856,
+      "step": 15859
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6180856227874756,
+      "learning_rate": 4.2548880260343985e-07,
+      "loss": 0.8654,
+      "step": 15860
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9282160997390747,
+      "learning_rate": 4.2495286949718937e-07,
+      "loss": 0.9833,
+      "step": 15861
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8110114336013794,
+      "learning_rate": 4.244172667990698e-07,
+      "loss": 0.8345,
+      "step": 15862
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.855578899383545,
+      "learning_rate": 4.2388199452756075e-07,
+      "loss": 0.928,
+      "step": 15863
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8390268087387085,
+      "learning_rate": 4.2334705270113405e-07,
+      "loss": 0.8307,
+      "step": 15864
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7110477685928345,
+      "learning_rate": 4.2281244133825173e-07,
+      "loss": 0.8727,
+      "step": 15865
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7367172241210938,
+      "learning_rate": 4.222781604573567e-07,
+      "loss": 0.7846,
+      "step": 15866
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9426498413085938,
+      "learning_rate": 4.2174421007688983e-07,
+      "loss": 0.9402,
+      "step": 15867
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7359305620193481,
+      "learning_rate": 4.21210590215273e-07,
+      "loss": 0.8705,
+      "step": 15868
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7773189544677734,
+      "learning_rate": 4.206773008909226e-07,
+      "loss": 0.9183,
+      "step": 15869
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9586496353149414,
+      "learning_rate": 4.201443421222384e-07,
+      "loss": 0.8646,
+      "step": 15870
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.523102045059204,
+      "learning_rate": 4.196117139276146e-07,
+      "loss": 0.9036,
+      "step": 15871
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7972440719604492,
+      "learning_rate": 4.190794163254275e-07,
+      "loss": 0.9298,
+      "step": 15872
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8890737295150757,
+      "learning_rate": 4.18547449334048e-07,
+      "loss": 0.8857,
+      "step": 15873
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9448047876358032,
+      "learning_rate": 4.1801581297182926e-07,
+      "loss": 0.9867,
+      "step": 15874
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7496426105499268,
+      "learning_rate": 4.1748450725711875e-07,
+      "loss": 1.0168,
+      "step": 15875
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9961042404174805,
+      "learning_rate": 4.169535322082519e-07,
+      "loss": 0.9143,
+      "step": 15876
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7489299774169922,
+      "learning_rate": 4.164228878435483e-07,
+      "loss": 1.0009,
+      "step": 15877
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8291044235229492,
+      "learning_rate": 4.1589257418132135e-07,
+      "loss": 0.9289,
+      "step": 15878
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7482285499572754,
+      "learning_rate": 4.1536259123986735e-07,
+      "loss": 0.9399,
+      "step": 15879
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.015943765640259,
+      "learning_rate": 4.1483293903747944e-07,
+      "loss": 0.8717,
+      "step": 15880
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.93864107131958,
+      "learning_rate": 4.1430361759242976e-07,
+      "loss": 0.942,
+      "step": 15881
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.913555383682251,
+      "learning_rate": 4.1377462692298695e-07,
+      "loss": 0.8557,
+      "step": 15882
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8105446100234985,
+      "learning_rate": 4.1324596704740203e-07,
+      "loss": 0.8742,
+      "step": 15883
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0868711471557617,
+      "learning_rate": 4.127176379839193e-07,
+      "loss": 0.5959,
+      "step": 15884
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9166744947433472,
+      "learning_rate": 4.121896397507708e-07,
+      "loss": 0.9479,
+      "step": 15885
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0831048488616943,
+      "learning_rate": 4.1166197236617634e-07,
+      "loss": 0.541,
+      "step": 15886
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7250031232833862,
+      "learning_rate": 4.1113463584834144e-07,
+      "loss": 0.8544,
+      "step": 15887
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8878393173217773,
+      "learning_rate": 4.106076302154671e-07,
+      "loss": 0.8975,
+      "step": 15888
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.673466682434082,
+      "learning_rate": 4.100809554857343e-07,
+      "loss": 0.9465,
+      "step": 15889
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6400001049041748,
+      "learning_rate": 4.095546116773208e-07,
+      "loss": 0.8602,
+      "step": 15890
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9134602546691895,
+      "learning_rate": 4.0902859880838643e-07,
+      "loss": 0.8467,
+      "step": 15891
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0014296770095825,
+      "learning_rate": 4.0850291689708553e-07,
+      "loss": 0.5115,
+      "step": 15892
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.702964425086975,
+      "learning_rate": 4.079775659615548e-07,
+      "loss": 0.926,
+      "step": 15893
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9945602416992188,
+      "learning_rate": 4.074525460199241e-07,
+      "loss": 0.5498,
+      "step": 15894
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.774322509765625,
+      "learning_rate": 4.0692785709031125e-07,
+      "loss": 0.9044,
+      "step": 15895
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.690029501914978,
+      "learning_rate": 4.0640349919082056e-07,
+      "loss": 0.9713,
+      "step": 15896
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5604135990142822,
+      "learning_rate": 4.058794723395465e-07,
+      "loss": 0.8533,
+      "step": 15897
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8438425064086914,
+      "learning_rate": 4.0535577655457127e-07,
+      "loss": 0.8902,
+      "step": 15898
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.811793327331543,
+      "learning_rate": 4.0483241185396815e-07,
+      "loss": 0.9056,
+      "step": 15899
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7385178804397583,
+      "learning_rate": 4.0430937825579386e-07,
+      "loss": 0.9308,
+      "step": 15900
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0665388107299805,
+      "learning_rate": 4.037866757780995e-07,
+      "loss": 0.5866,
+      "step": 15901
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.056377649307251,
+      "learning_rate": 4.032643044389195e-07,
+      "loss": 0.9549,
+      "step": 15902
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7054698467254639,
+      "learning_rate": 4.027422642562828e-07,
+      "loss": 0.8626,
+      "step": 15903
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0850942134857178,
+      "learning_rate": 4.0222055524820057e-07,
+      "loss": 0.8922,
+      "step": 15904
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7183587551116943,
+      "learning_rate": 4.0169917743267616e-07,
+      "loss": 0.8545,
+      "step": 15905
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8127549886703491,
+      "learning_rate": 4.0117813082770296e-07,
+      "loss": 0.8847,
+      "step": 15906
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7577332258224487,
+      "learning_rate": 4.006574154512577e-07,
+      "loss": 0.8403,
+      "step": 15907
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8667739629745483,
+      "learning_rate": 4.0013703132131153e-07,
+      "loss": 0.8531,
+      "step": 15908
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7228500843048096,
+      "learning_rate": 3.9961697845581905e-07,
+      "loss": 0.7923,
+      "step": 15909
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6586765050888062,
+      "learning_rate": 3.9909725687272914e-07,
+      "loss": 0.7987,
+      "step": 15910
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8758174180984497,
+      "learning_rate": 3.98577866589972e-07,
+      "loss": 0.9429,
+      "step": 15911
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.810036063194275,
+      "learning_rate": 3.9805880762547323e-07,
+      "loss": 0.8145,
+      "step": 15912
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6515954732894897,
+      "learning_rate": 3.975400799971418e-07,
+      "loss": 0.8672,
+      "step": 15913
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.821568250656128,
+      "learning_rate": 3.970216837228802e-07,
+      "loss": 0.9317,
+      "step": 15914
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8133752346038818,
+      "learning_rate": 3.96503618820574e-07,
+      "loss": 0.9489,
+      "step": 15915
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.899345874786377,
+      "learning_rate": 3.9598588530810335e-07,
+      "loss": 0.8526,
+      "step": 15916
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6724337339401245,
+      "learning_rate": 3.9546848320333067e-07,
+      "loss": 0.7796,
+      "step": 15917
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8203648328781128,
+      "learning_rate": 3.949514125241116e-07,
+      "loss": 0.8627,
+      "step": 15918
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.801805019378662,
+      "learning_rate": 3.9443467328828976e-07,
+      "loss": 0.8723,
+      "step": 15919
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.72738778591156,
+      "learning_rate": 3.9391826551369417e-07,
+      "loss": 0.9546,
+      "step": 15920
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7988413572311401,
+      "learning_rate": 3.934021892181461e-07,
+      "loss": 0.8887,
+      "step": 15921
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.786757230758667,
+      "learning_rate": 3.9288644441945356e-07,
+      "loss": 0.8449,
+      "step": 15922
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2779321670532227,
+      "learning_rate": 3.923710311354134e-07,
+      "loss": 0.9014,
+      "step": 15923
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7580676078796387,
+      "learning_rate": 3.918559493838114e-07,
+      "loss": 0.9546,
+      "step": 15924
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6649545431137085,
+      "learning_rate": 3.913411991824212e-07,
+      "loss": 0.8636,
+      "step": 15925
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.76620614528656,
+      "learning_rate": 3.908267805490051e-07,
+      "loss": 0.871,
+      "step": 15926
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8247005939483643,
+      "learning_rate": 3.9031269350131574e-07,
+      "loss": 0.9103,
+      "step": 15927
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7880696058273315,
+      "learning_rate": 3.897989380570899e-07,
+      "loss": 0.8994,
+      "step": 15928
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7409021854400635,
+      "learning_rate": 3.892855142340579e-07,
+      "loss": 0.9078,
+      "step": 15929
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.879544734954834,
+      "learning_rate": 3.8877242204993784e-07,
+      "loss": 0.9704,
+      "step": 15930
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7329720258712769,
+      "learning_rate": 3.882596615224332e-07,
+      "loss": 0.8924,
+      "step": 15931
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8035959005355835,
+      "learning_rate": 3.8774723266923886e-07,
+      "loss": 0.9232,
+      "step": 15932
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7096164226531982,
+      "learning_rate": 3.8723513550803506e-07,
+      "loss": 0.9052,
+      "step": 15933
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8174573183059692,
+      "learning_rate": 3.867233700564965e-07,
+      "loss": 0.8994,
+      "step": 15934
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7200968265533447,
+      "learning_rate": 3.8621193633227916e-07,
+      "loss": 0.8657,
+      "step": 15935
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.842258095741272,
+      "learning_rate": 3.857008343530344e-07,
+      "loss": 0.9125,
+      "step": 15936
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8510268926620483,
+      "learning_rate": 3.851900641363959e-07,
+      "loss": 0.9554,
+      "step": 15937
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0929793119430542,
+      "learning_rate": 3.8467962569999183e-07,
+      "loss": 0.5181,
+      "step": 15938
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6862468719482422,
+      "learning_rate": 3.8416951906143253e-07,
+      "loss": 0.8226,
+      "step": 15939
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.798294186592102,
+      "learning_rate": 3.8365974423832496e-07,
+      "loss": 0.8832,
+      "step": 15940
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7634085416793823,
+      "learning_rate": 3.8315030124825516e-07,
+      "loss": 0.8715,
+      "step": 15941
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.689590573310852,
+      "learning_rate": 3.8264119010880564e-07,
+      "loss": 0.9394,
+      "step": 15942
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9379509687423706,
+      "learning_rate": 3.821324108375446e-07,
+      "loss": 0.9528,
+      "step": 15943
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.763135552406311,
+      "learning_rate": 3.816239634520258e-07,
+      "loss": 0.9559,
+      "step": 15944
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3083832263946533,
+      "learning_rate": 3.811158479697985e-07,
+      "loss": 0.9314,
+      "step": 15945
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6003978252410889,
+      "learning_rate": 3.8060806440839205e-07,
+      "loss": 0.8946,
+      "step": 15946
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.835839867591858,
+      "learning_rate": 3.801006127853313e-07,
+      "loss": 0.9251,
+      "step": 15947
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.002423048019409,
+      "learning_rate": 3.795934931181267e-07,
+      "loss": 0.9791,
+      "step": 15948
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7746679782867432,
+      "learning_rate": 3.7908670542427637e-07,
+      "loss": 0.9778,
+      "step": 15949
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7696954011917114,
+      "learning_rate": 3.785802497212676e-07,
+      "loss": 0.8341,
+      "step": 15950
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8632299900054932,
+      "learning_rate": 3.780741260265808e-07,
+      "loss": 0.9339,
+      "step": 15951
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8545212745666504,
+      "learning_rate": 3.775683343576764e-07,
+      "loss": 0.8383,
+      "step": 15952
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8567663431167603,
+      "learning_rate": 3.7706287473201155e-07,
+      "loss": 0.9113,
+      "step": 15953
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.720462441444397,
+      "learning_rate": 3.7655774716702453e-07,
+      "loss": 0.8549,
+      "step": 15954
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7620278596878052,
+      "learning_rate": 3.7605295168014813e-07,
+      "loss": 0.9328,
+      "step": 15955
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7949262857437134,
+      "learning_rate": 3.755484882888005e-07,
+      "loss": 0.8645,
+      "step": 15956
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8074122667312622,
+      "learning_rate": 3.750443570103912e-07,
+      "loss": 0.9107,
+      "step": 15957
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.808741807937622,
+      "learning_rate": 3.745405578623129e-07,
+      "loss": 0.9098,
+      "step": 15958
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8140826225280762,
+      "learning_rate": 3.740370908619528e-07,
+      "loss": 0.897,
+      "step": 15959
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8312243223190308,
+      "learning_rate": 3.7353395602668486e-07,
+      "loss": 0.9867,
+      "step": 15960
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.059182643890381,
+      "learning_rate": 3.7303115337386843e-07,
+      "loss": 0.8797,
+      "step": 15961
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9641932249069214,
+      "learning_rate": 3.725286829208563e-07,
+      "loss": 0.8713,
+      "step": 15962
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0012917518615723,
+      "learning_rate": 3.720265446849858e-07,
+      "loss": 0.8169,
+      "step": 15963
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7301795482635498,
+      "learning_rate": 3.715247386835841e-07,
+      "loss": 0.902,
+      "step": 15964
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7078012228012085,
+      "learning_rate": 3.7102326493396736e-07,
+      "loss": 0.8431,
+      "step": 15965
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7636345624923706,
+      "learning_rate": 3.7052212345344176e-07,
+      "loss": 0.896,
+      "step": 15966
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8149771690368652,
+      "learning_rate": 3.7002131425929677e-07,
+      "loss": 0.8703,
+      "step": 15967
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6916556358337402,
+      "learning_rate": 3.695208373688175e-07,
+      "loss": 0.8962,
+      "step": 15968
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9374797344207764,
+      "learning_rate": 3.690206927992712e-07,
+      "loss": 0.8997,
+      "step": 15969
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5948518514633179,
+      "learning_rate": 3.685208805679186e-07,
+      "loss": 0.9563,
+      "step": 15970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.761960744857788,
+      "learning_rate": 3.680214006920058e-07,
+      "loss": 0.8667,
+      "step": 15971
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0561821460723877,
+      "learning_rate": 3.675222531887679e-07,
+      "loss": 0.5755,
+      "step": 15972
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7220972776412964,
+      "learning_rate": 3.670234380754312e-07,
+      "loss": 0.8981,
+      "step": 15973
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7780568599700928,
+      "learning_rate": 3.665249553692052e-07,
+      "loss": 0.8575,
+      "step": 15974
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8450536727905273,
+      "learning_rate": 3.6602680508729506e-07,
+      "loss": 0.8799,
+      "step": 15975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8215750455856323,
+      "learning_rate": 3.65528987246887e-07,
+      "loss": 0.9125,
+      "step": 15976
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6484339237213135,
+      "learning_rate": 3.650315018651618e-07,
+      "loss": 0.8873,
+      "step": 15977
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6994446516036987,
+      "learning_rate": 3.645343489592834e-07,
+      "loss": 0.8745,
+      "step": 15978
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7784489393234253,
+      "learning_rate": 3.640375285464115e-07,
+      "loss": 0.8452,
+      "step": 15979
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6236860752105713,
+      "learning_rate": 3.635410406436857e-07,
+      "loss": 0.9537,
+      "step": 15980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.762093424797058,
+      "learning_rate": 3.6304488526824113e-07,
+      "loss": 0.8743,
+      "step": 15981
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8169455528259277,
+      "learning_rate": 3.6254906243719744e-07,
+      "loss": 0.8901,
+      "step": 15982
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9161959886550903,
+      "learning_rate": 3.620535721676643e-07,
+      "loss": 0.8909,
+      "step": 15983
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8205138444900513,
+      "learning_rate": 3.6155841447674035e-07,
+      "loss": 1.0179,
+      "step": 15984
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7061212062835693,
+      "learning_rate": 3.6106358938151065e-07,
+      "loss": 0.8711,
+      "step": 15985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7559456825256348,
+      "learning_rate": 3.6056909689905274e-07,
+      "loss": 0.8928,
+      "step": 15986
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8215025663375854,
+      "learning_rate": 3.600749370464274e-07,
+      "loss": 0.8903,
+      "step": 15987
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8701554536819458,
+      "learning_rate": 3.595811098406887e-07,
+      "loss": 0.897,
+      "step": 15988
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.980040431022644,
+      "learning_rate": 3.5908761529887536e-07,
+      "loss": 0.8842,
+      "step": 15989
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8623727560043335,
+      "learning_rate": 3.5859445343801926e-07,
+      "loss": 0.8437,
+      "step": 15990
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7643805742263794,
+      "learning_rate": 3.581016242751356e-07,
+      "loss": 0.8693,
+      "step": 15991
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9034552574157715,
+      "learning_rate": 3.57609127827232e-07,
+      "loss": 0.9462,
+      "step": 15992
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9294816255569458,
+      "learning_rate": 3.5711696411130035e-07,
+      "loss": 0.9132,
+      "step": 15993
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7443877458572388,
+      "learning_rate": 3.5662513314432823e-07,
+      "loss": 0.8514,
+      "step": 15994
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7230364084243774,
+      "learning_rate": 3.561336349432842e-07,
+      "loss": 0.9014,
+      "step": 15995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8636956214904785,
+      "learning_rate": 3.5564246952512817e-07,
+      "loss": 0.9478,
+      "step": 15996
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7242804765701294,
+      "learning_rate": 3.55151636906812e-07,
+      "loss": 0.9772,
+      "step": 15997
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7394336462020874,
+      "learning_rate": 3.5466113710526997e-07,
+      "loss": 0.9589,
+      "step": 15998
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7706475257873535,
+      "learning_rate": 3.5417097013743075e-07,
+      "loss": 0.8727,
+      "step": 15999
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.981211245059967,
+      "learning_rate": 3.5368113602020414e-07,
+      "loss": 0.5746,
+      "step": 16000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6546509265899658,
+      "learning_rate": 3.531916347704978e-07,
+      "loss": 0.796,
+      "step": 16001
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.804200530052185,
+      "learning_rate": 3.5270246640519925e-07,
+      "loss": 0.898,
+      "step": 16002
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9052832126617432,
+      "learning_rate": 3.5221363094119166e-07,
+      "loss": 0.866,
+      "step": 16003
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.696407437324524,
+      "learning_rate": 3.5172512839533934e-07,
+      "loss": 0.8817,
+      "step": 16004
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8722658157348633,
+      "learning_rate": 3.5123695878450327e-07,
+      "loss": 0.8628,
+      "step": 16005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7976242303848267,
+      "learning_rate": 3.507491221255266e-07,
+      "loss": 0.9414,
+      "step": 16006
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7640900611877441,
+      "learning_rate": 3.5026161843524254e-07,
+      "loss": 0.9239,
+      "step": 16007
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.771964192390442,
+      "learning_rate": 3.4977444773047653e-07,
+      "loss": 0.8632,
+      "step": 16008
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6969386339187622,
+      "learning_rate": 3.4928761002803625e-07,
+      "loss": 0.9189,
+      "step": 16009
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8852514028549194,
+      "learning_rate": 3.4880110534472265e-07,
+      "loss": 0.8797,
+      "step": 16010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1289188861846924,
+      "learning_rate": 3.483149336973235e-07,
+      "loss": 0.9061,
+      "step": 16011
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7376195192337036,
+      "learning_rate": 3.478290951026153e-07,
+      "loss": 0.9327,
+      "step": 16012
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7414181232452393,
+      "learning_rate": 3.4734358957736247e-07,
+      "loss": 0.8543,
+      "step": 16013
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8411325216293335,
+      "learning_rate": 3.4685841713831937e-07,
+      "loss": 0.871,
+      "step": 16014
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9469294548034668,
+      "learning_rate": 3.463735778022259e-07,
+      "loss": 0.9475,
+      "step": 16015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6670212745666504,
+      "learning_rate": 3.458890715858143e-07,
+      "loss": 0.8557,
+      "step": 16016
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0476583242416382,
+      "learning_rate": 3.454048985058034e-07,
+      "loss": 0.5805,
+      "step": 16017
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0331312417984009,
+      "learning_rate": 3.449210585789009e-07,
+      "loss": 0.5023,
+      "step": 16018
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.045556664466858,
+      "learning_rate": 3.4443755182180125e-07,
+      "loss": 0.5286,
+      "step": 16019
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6600151062011719,
+      "learning_rate": 3.4395437825119117e-07,
+      "loss": 0.8311,
+      "step": 16020
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6164562702178955,
+      "learning_rate": 3.4347153788374167e-07,
+      "loss": 0.8996,
+      "step": 16021
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.805677890777588,
+      "learning_rate": 3.429890307361161e-07,
+      "loss": 0.9129,
+      "step": 16022
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8209412097930908,
+      "learning_rate": 3.4250685682496233e-07,
+      "loss": 0.8245,
+      "step": 16023
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7682663202285767,
+      "learning_rate": 3.420250161669203e-07,
+      "loss": 0.8792,
+      "step": 16024
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8460445404052734,
+      "learning_rate": 3.4154350877861565e-07,
+      "loss": 0.8908,
+      "step": 16025
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8630725145339966,
+      "learning_rate": 3.4106233467666504e-07,
+      "loss": 0.8475,
+      "step": 16026
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7359724044799805,
+      "learning_rate": 3.4058149387767305e-07,
+      "loss": 0.931,
+      "step": 16027
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7361557483673096,
+      "learning_rate": 3.4010098639822964e-07,
+      "loss": 0.8737,
+      "step": 16028
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1061549186706543,
+      "learning_rate": 3.396208122549194e-07,
+      "loss": 0.868,
+      "step": 16029
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.253780484199524,
+      "learning_rate": 3.391409714643079e-07,
+      "loss": 0.5529,
+      "step": 16030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1058951616287231,
+      "learning_rate": 3.386614640429553e-07,
+      "loss": 0.5813,
+      "step": 16031
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7368977069854736,
+      "learning_rate": 3.3818229000740833e-07,
+      "loss": 0.8316,
+      "step": 16032
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8827120065689087,
+      "learning_rate": 3.377034493742015e-07,
+      "loss": 0.8448,
+      "step": 16033
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.536138892173767,
+      "learning_rate": 3.372249421598572e-07,
+      "loss": 0.8511,
+      "step": 16034
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8005306720733643,
+      "learning_rate": 3.3674676838088893e-07,
+      "loss": 0.9824,
+      "step": 16035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8767318725585938,
+      "learning_rate": 3.3626892805379565e-07,
+      "loss": 0.8387,
+      "step": 16036
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9783217310905457,
+      "learning_rate": 3.357914211950675e-07,
+      "loss": 0.5722,
+      "step": 16037
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6194586753845215,
+      "learning_rate": 3.353142478211824e-07,
+      "loss": 0.9219,
+      "step": 16038
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7989591360092163,
+      "learning_rate": 3.3483740794860386e-07,
+      "loss": 0.799,
+      "step": 16039
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9079878330230713,
+      "learning_rate": 3.3436090159378987e-07,
+      "loss": 0.9051,
+      "step": 16040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8095263242721558,
+      "learning_rate": 3.338847287731795e-07,
+      "loss": 0.8747,
+      "step": 16041
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8701649904251099,
+      "learning_rate": 3.3340888950320725e-07,
+      "loss": 0.8645,
+      "step": 16042
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7751802206039429,
+      "learning_rate": 3.3293338380029017e-07,
+      "loss": 0.7998,
+      "step": 16043
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7383900880813599,
+      "learning_rate": 3.324582116808395e-07,
+      "loss": 0.9046,
+      "step": 16044
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.743598222732544,
+      "learning_rate": 3.3198337316124987e-07,
+      "loss": 0.9032,
+      "step": 16045
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7166293859481812,
+      "learning_rate": 3.315088682579082e-07,
+      "loss": 0.8994,
+      "step": 16046
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7939342260360718,
+      "learning_rate": 3.3103469698718694e-07,
+      "loss": 0.9371,
+      "step": 16047
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7819585800170898,
+      "learning_rate": 3.305608593654508e-07,
+      "loss": 0.8843,
+      "step": 16048
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7243385314941406,
+      "learning_rate": 3.3008735540904666e-07,
+      "loss": 0.9563,
+      "step": 16049
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7253384590148926,
+      "learning_rate": 3.2961418513431596e-07,
+      "loss": 0.8331,
+      "step": 16050
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7176647186279297,
+      "learning_rate": 3.2914134855758894e-07,
+      "loss": 0.8971,
+      "step": 16051
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9830152988433838,
+      "learning_rate": 3.286688456951781e-07,
+      "loss": 0.8723,
+      "step": 16052
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8986012935638428,
+      "learning_rate": 3.2819667656339036e-07,
+      "loss": 0.8976,
+      "step": 16053
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8135855197906494,
+      "learning_rate": 3.2772484117851723e-07,
+      "loss": 0.9289,
+      "step": 16054
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6852836608886719,
+      "learning_rate": 3.272533395568422e-07,
+      "loss": 0.8648,
+      "step": 16055
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.769431233406067,
+      "learning_rate": 3.2678217171463355e-07,
+      "loss": 0.8704,
+      "step": 16056
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6774264574050903,
+      "learning_rate": 3.263113376681526e-07,
+      "loss": 0.9187,
+      "step": 16057
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7942821979522705,
+      "learning_rate": 3.2584083743364416e-07,
+      "loss": 0.9553,
+      "step": 16058
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7275360822677612,
+      "learning_rate": 3.253706710273452e-07,
+      "loss": 0.8616,
+      "step": 16059
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7733031511306763,
+      "learning_rate": 3.2490083846547835e-07,
+      "loss": 0.996,
+      "step": 16060
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9808598160743713,
+      "learning_rate": 3.2443133976425733e-07,
+      "loss": 0.5259,
+      "step": 16061
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8284358978271484,
+      "learning_rate": 3.239621749398847e-07,
+      "loss": 0.9432,
+      "step": 16062
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7595874071121216,
+      "learning_rate": 3.2349334400854746e-07,
+      "loss": 0.8562,
+      "step": 16063
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0034810304641724,
+      "learning_rate": 3.23024846986425e-07,
+      "loss": 0.5445,
+      "step": 16064
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.833794355392456,
+      "learning_rate": 3.225566838896832e-07,
+      "loss": 0.9258,
+      "step": 16065
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6624916791915894,
+      "learning_rate": 3.2208885473447806e-07,
+      "loss": 0.9563,
+      "step": 16066
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7818577289581299,
+      "learning_rate": 3.216213595369522e-07,
+      "loss": 0.9376,
+      "step": 16067
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7648690938949585,
+      "learning_rate": 3.2115419831323715e-07,
+      "loss": 0.9309,
+      "step": 16068
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7156198024749756,
+      "learning_rate": 3.206873710794545e-07,
+      "loss": 0.9243,
+      "step": 16069
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6340789794921875,
+      "learning_rate": 3.2022087785171243e-07,
+      "loss": 0.9183,
+      "step": 16070
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7318247556686401,
+      "learning_rate": 3.1975471864610805e-07,
+      "loss": 0.8846,
+      "step": 16071
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.781927227973938,
+      "learning_rate": 3.192888934787286e-07,
+      "loss": 0.9002,
+      "step": 16072
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6848148107528687,
+      "learning_rate": 3.188234023656467e-07,
+      "loss": 0.8772,
+      "step": 16073
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1297677755355835,
+      "learning_rate": 3.1835824532292616e-07,
+      "loss": 0.5209,
+      "step": 16074
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8265327215194702,
+      "learning_rate": 3.178934223666186e-07,
+      "loss": 0.9265,
+      "step": 16075
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7990341186523438,
+      "learning_rate": 3.174289335127612e-07,
+      "loss": 0.906,
+      "step": 16076
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9134149551391602,
+      "learning_rate": 3.1696477877738664e-07,
+      "loss": 0.8379,
+      "step": 16077
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0070412158966064,
+      "learning_rate": 3.1650095817650773e-07,
+      "loss": 0.9389,
+      "step": 16078
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.621281385421753,
+      "learning_rate": 3.1603747172613165e-07,
+      "loss": 0.8458,
+      "step": 16079
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0595463514328003,
+      "learning_rate": 3.1557431944225005e-07,
+      "loss": 0.5352,
+      "step": 16080
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9107943773269653,
+      "learning_rate": 3.1511150134084787e-07,
+      "loss": 0.8584,
+      "step": 16081
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7533669471740723,
+      "learning_rate": 3.146490174378913e-07,
+      "loss": 0.8822,
+      "step": 16082
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9546745419502258,
+      "learning_rate": 3.141868677493454e-07,
+      "loss": 0.5225,
+      "step": 16083
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7685881853103638,
+      "learning_rate": 3.137250522911528e-07,
+      "loss": 0.8396,
+      "step": 16084
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8378522396087646,
+      "learning_rate": 3.132635710792531e-07,
+      "loss": 0.9302,
+      "step": 16085
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8987963199615479,
+      "learning_rate": 3.1280242412956687e-07,
+      "loss": 0.9008,
+      "step": 16086
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6857706308364868,
+      "learning_rate": 3.123416114580091e-07,
+      "loss": 0.9239,
+      "step": 16087
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8007545471191406,
+      "learning_rate": 3.118811330804816e-07,
+      "loss": 0.857,
+      "step": 16088
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9181270599365234,
+      "learning_rate": 3.114209890128739e-07,
+      "loss": 0.8288,
+      "step": 16089
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8000327348709106,
+      "learning_rate": 3.1096117927106205e-07,
+      "loss": 0.8175,
+      "step": 16090
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8781778812408447,
+      "learning_rate": 3.105017038709157e-07,
+      "loss": 0.8544,
+      "step": 16091
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9210681915283203,
+      "learning_rate": 3.100425628282899e-07,
+      "loss": 0.8478,
+      "step": 16092
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7471415996551514,
+      "learning_rate": 3.095837561590265e-07,
+      "loss": 0.8533,
+      "step": 16093
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9061459302902222,
+      "learning_rate": 3.0912528387895937e-07,
+      "loss": 0.9306,
+      "step": 16094
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8363149166107178,
+      "learning_rate": 3.0866714600390704e-07,
+      "loss": 1.0026,
+      "step": 16095
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7845964431762695,
+      "learning_rate": 3.0820934254968126e-07,
+      "loss": 0.9465,
+      "step": 16096
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.948388695716858,
+      "learning_rate": 3.0775187353207614e-07,
+      "loss": 0.874,
+      "step": 16097
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.740517497062683,
+      "learning_rate": 3.072947389668823e-07,
+      "loss": 0.8771,
+      "step": 16098
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3168516159057617,
+      "learning_rate": 3.0683793886986943e-07,
+      "loss": 0.879,
+      "step": 16099
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9168965816497803,
+      "learning_rate": 3.063814732568038e-07,
+      "loss": 0.9173,
+      "step": 16100
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.865510106086731,
+      "learning_rate": 3.0592534214343495e-07,
+      "loss": 0.8821,
+      "step": 16101
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.641528606414795,
+      "learning_rate": 3.0546954554550366e-07,
+      "loss": 0.8791,
+      "step": 16102
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6778312921524048,
+      "learning_rate": 3.050140834787374e-07,
+      "loss": 0.909,
+      "step": 16103
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6749008893966675,
+      "learning_rate": 3.0455895595885246e-07,
+      "loss": 0.9597,
+      "step": 16104
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.851684808731079,
+      "learning_rate": 3.041041630015562e-07,
+      "loss": 0.8651,
+      "step": 16105
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7146273851394653,
+      "learning_rate": 3.036497046225395e-07,
+      "loss": 0.8744,
+      "step": 16106
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7540463209152222,
+      "learning_rate": 3.0319558083748754e-07,
+      "loss": 0.9171,
+      "step": 16107
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.798244595527649,
+      "learning_rate": 3.0274179166206785e-07,
+      "loss": 0.9312,
+      "step": 16108
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.01286780834198,
+      "learning_rate": 3.022883371119423e-07,
+      "loss": 0.5771,
+      "step": 16109
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6508347988128662,
+      "learning_rate": 3.01835217202755e-07,
+      "loss": 0.8731,
+      "step": 16110
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8392674922943115,
+      "learning_rate": 3.013824319501446e-07,
+      "loss": 0.8802,
+      "step": 16111
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.656704068183899,
+      "learning_rate": 3.009299813697331e-07,
+      "loss": 0.8668,
+      "step": 16112
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8883048295974731,
+      "learning_rate": 3.0047786547713677e-07,
+      "loss": 0.8809,
+      "step": 16113
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.662778377532959,
+      "learning_rate": 3.000260842879532e-07,
+      "loss": 0.8824,
+      "step": 16114
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6404680013656616,
+      "learning_rate": 2.9957463781777443e-07,
+      "loss": 0.7754,
+      "step": 16115
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8990052938461304,
+      "learning_rate": 2.991235260821779e-07,
+      "loss": 0.9018,
+      "step": 16116
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8069746494293213,
+      "learning_rate": 2.986727490967289e-07,
+      "loss": 0.7892,
+      "step": 16117
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6222139596939087,
+      "learning_rate": 2.982223068769863e-07,
+      "loss": 0.9439,
+      "step": 16118
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5850636959075928,
+      "learning_rate": 2.9777219943848856e-07,
+      "loss": 0.8239,
+      "step": 16119
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8803397417068481,
+      "learning_rate": 2.9732242679677227e-07,
+      "loss": 0.8058,
+      "step": 16120
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6956595182418823,
+      "learning_rate": 2.9687298896735384e-07,
+      "loss": 0.9009,
+      "step": 16121
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7697738409042358,
+      "learning_rate": 2.964238859657453e-07,
+      "loss": 0.941,
+      "step": 16122
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8401726484298706,
+      "learning_rate": 2.9597511780744104e-07,
+      "loss": 0.8962,
+      "step": 16123
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8243794441223145,
+      "learning_rate": 2.9552668450792965e-07,
+      "loss": 0.8883,
+      "step": 16124
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6852229833602905,
+      "learning_rate": 2.9507858608268325e-07,
+      "loss": 0.9278,
+      "step": 16125
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8400501012802124,
+      "learning_rate": 2.9463082254716725e-07,
+      "loss": 0.9365,
+      "step": 16126
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9815301299095154,
+      "learning_rate": 2.941833939168282e-07,
+      "loss": 0.5137,
+      "step": 16127
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8687422275543213,
+      "learning_rate": 2.937363002071081e-07,
+      "loss": 0.877,
+      "step": 16128
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7938392162322998,
+      "learning_rate": 2.932895414334369e-07,
+      "loss": 0.8558,
+      "step": 16129
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.891542911529541,
+      "learning_rate": 2.9284311761122787e-07,
+      "loss": 0.8359,
+      "step": 16130
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8307249546051025,
+      "learning_rate": 2.923970287558875e-07,
+      "loss": 0.8985,
+      "step": 16131
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7825486660003662,
+      "learning_rate": 2.9195127488280795e-07,
+      "loss": 0.9025,
+      "step": 16132
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8482294082641602,
+      "learning_rate": 2.9150585600737247e-07,
+      "loss": 0.9454,
+      "step": 16133
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8095290660858154,
+      "learning_rate": 2.910607721449488e-07,
+      "loss": 0.9362,
+      "step": 16134
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6416980028152466,
+      "learning_rate": 2.90616023310899e-07,
+      "loss": 0.891,
+      "step": 16135
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6791030168533325,
+      "learning_rate": 2.9017160952056646e-07,
+      "loss": 0.9153,
+      "step": 16136
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.949495255947113,
+      "learning_rate": 2.8972753078928994e-07,
+      "loss": 0.5017,
+      "step": 16137
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7111949920654297,
+      "learning_rate": 2.892837871323906e-07,
+      "loss": 0.931,
+      "step": 16138
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4629085063934326,
+      "learning_rate": 2.8884037856518277e-07,
+      "loss": 0.8134,
+      "step": 16139
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8181954622268677,
+      "learning_rate": 2.8839730510296536e-07,
+      "loss": 0.9739,
+      "step": 16140
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7331819534301758,
+      "learning_rate": 2.8795456676102837e-07,
+      "loss": 0.89,
+      "step": 16141
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8123481273651123,
+      "learning_rate": 2.8751216355465075e-07,
+      "loss": 0.809,
+      "step": 16142
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7138338088989258,
+      "learning_rate": 2.8707009549909793e-07,
+      "loss": 0.8243,
+      "step": 16143
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9347686767578125,
+      "learning_rate": 2.8662836260962444e-07,
+      "loss": 0.8708,
+      "step": 16144
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.954247236251831,
+      "learning_rate": 2.861869649014715e-07,
+      "loss": 0.871,
+      "step": 16145
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6859766244888306,
+      "learning_rate": 2.857459023898734e-07,
+      "loss": 0.9496,
+      "step": 16146
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7164487838745117,
+      "learning_rate": 2.853051750900471e-07,
+      "loss": 0.8696,
+      "step": 16147
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8968310356140137,
+      "learning_rate": 2.8486478301720246e-07,
+      "loss": 0.9632,
+      "step": 16148
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5730295181274414,
+      "learning_rate": 2.844247261865363e-07,
+      "loss": 0.8537,
+      "step": 16149
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7843034267425537,
+      "learning_rate": 2.839850046132342e-07,
+      "loss": 0.8481,
+      "step": 16150
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9352452754974365,
+      "learning_rate": 2.835456183124685e-07,
+      "loss": 0.9387,
+      "step": 16151
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6448343992233276,
+      "learning_rate": 2.8310656729940157e-07,
+      "loss": 0.843,
+      "step": 16152
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0222015380859375,
+      "learning_rate": 2.8266785158918454e-07,
+      "loss": 0.9801,
+      "step": 16153
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8364253044128418,
+      "learning_rate": 2.822294711969553e-07,
+      "loss": 0.8992,
+      "step": 16154
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9215871095657349,
+      "learning_rate": 2.817914261378407e-07,
+      "loss": 0.9338,
+      "step": 16155
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8381654024124146,
+      "learning_rate": 2.8135371642695865e-07,
+      "loss": 0.9219,
+      "step": 16156
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.836000680923462,
+      "learning_rate": 2.809163420794114e-07,
+      "loss": 0.9067,
+      "step": 16157
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0308854579925537,
+      "learning_rate": 2.8047930311029147e-07,
+      "loss": 0.5192,
+      "step": 16158
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.734331727027893,
+      "learning_rate": 2.8004259953468115e-07,
+      "loss": 0.8925,
+      "step": 16159
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6828244924545288,
+      "learning_rate": 2.7960623136764844e-07,
+      "loss": 0.7999,
+      "step": 16160
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.671135663986206,
+      "learning_rate": 2.7917019862425344e-07,
+      "loss": 0.8644,
+      "step": 16161
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6954333782196045,
+      "learning_rate": 2.787345013195386e-07,
+      "loss": 0.9731,
+      "step": 16162
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7403264045715332,
+      "learning_rate": 2.7829913946854305e-07,
+      "loss": 0.9298,
+      "step": 16163
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8312069177627563,
+      "learning_rate": 2.778641130862858e-07,
+      "loss": 0.8975,
+      "step": 16164
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8255512714385986,
+      "learning_rate": 2.774294221877816e-07,
+      "loss": 0.8557,
+      "step": 16165
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7426769733428955,
+      "learning_rate": 2.7699506678802837e-07,
+      "loss": 0.9236,
+      "step": 16166
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.88303804397583,
+      "learning_rate": 2.7656104690201636e-07,
+      "loss": 0.9352,
+      "step": 16167
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7245970964431763,
+      "learning_rate": 2.7612736254472026e-07,
+      "loss": 0.878,
+      "step": 16168
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8192698955535889,
+      "learning_rate": 2.7569401373110595e-07,
+      "loss": 0.8589,
+      "step": 16169
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8533731698989868,
+      "learning_rate": 2.7526100047612804e-07,
+      "loss": 0.9104,
+      "step": 16170
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6068453788757324,
+      "learning_rate": 2.7482832279472796e-07,
+      "loss": 0.852,
+      "step": 16171
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0124881267547607,
+      "learning_rate": 2.7439598070183705e-07,
+      "loss": 0.8949,
+      "step": 16172
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9334485530853271,
+      "learning_rate": 2.739639742123723e-07,
+      "loss": 0.866,
+      "step": 16173
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7058124542236328,
+      "learning_rate": 2.735323033412429e-07,
+      "loss": 0.9091,
+      "step": 16174
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7397733926773071,
+      "learning_rate": 2.731009681033436e-07,
+      "loss": 0.936,
+      "step": 16175
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.765415906906128,
+      "learning_rate": 2.726699685135603e-07,
+      "loss": 0.8738,
+      "step": 16176
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8463367223739624,
+      "learning_rate": 2.722393045867622e-07,
+      "loss": 0.8949,
+      "step": 16177
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8276586532592773,
+      "learning_rate": 2.718089763378129e-07,
+      "loss": 0.9126,
+      "step": 16178
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7087610960006714,
+      "learning_rate": 2.713789837815617e-07,
+      "loss": 0.88,
+      "step": 16179
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7833597660064697,
+      "learning_rate": 2.709493269328456e-07,
+      "loss": 0.7971,
+      "step": 16180
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7066795825958252,
+      "learning_rate": 2.705200058064916e-07,
+      "loss": 0.9262,
+      "step": 16181
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6766881942749023,
+      "learning_rate": 2.700910204173124e-07,
+      "loss": 0.8816,
+      "step": 16182
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8699969053268433,
+      "learning_rate": 2.696623707801149e-07,
+      "loss": 0.9221,
+      "step": 16183
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7617461681365967,
+      "learning_rate": 2.692340569096874e-07,
+      "loss": 0.9278,
+      "step": 16184
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7738142013549805,
+      "learning_rate": 2.6880607882081135e-07,
+      "loss": 0.9778,
+      "step": 16185
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.884142279624939,
+      "learning_rate": 2.683784365282527e-07,
+      "loss": 0.8555,
+      "step": 16186
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7024891376495361,
+      "learning_rate": 2.6795113004677187e-07,
+      "loss": 0.9031,
+      "step": 16187
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7708393335342407,
+      "learning_rate": 2.6752415939111154e-07,
+      "loss": 0.9549,
+      "step": 16188
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9742568731307983,
+      "learning_rate": 2.6709752457600657e-07,
+      "loss": 0.482,
+      "step": 16189
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9434130191802979,
+      "learning_rate": 2.6667122561617744e-07,
+      "loss": 0.8558,
+      "step": 16190
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8814247846603394,
+      "learning_rate": 2.6624526252633564e-07,
+      "loss": 0.9564,
+      "step": 16191
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8548383712768555,
+      "learning_rate": 2.6581963532117947e-07,
+      "loss": 0.9453,
+      "step": 16192
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7945811748504639,
+      "learning_rate": 2.653943440153961e-07,
+      "loss": 0.8389,
+      "step": 16193
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7402881383895874,
+      "learning_rate": 2.649693886236626e-07,
+      "loss": 0.8446,
+      "step": 16194
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6852415800094604,
+      "learning_rate": 2.6454476916063953e-07,
+      "loss": 0.8128,
+      "step": 16195
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8853914737701416,
+      "learning_rate": 2.64120485640984e-07,
+      "loss": 0.9464,
+      "step": 16196
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.714269995689392,
+      "learning_rate": 2.6369653807933327e-07,
+      "loss": 0.8983,
+      "step": 16197
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8072229623794556,
+      "learning_rate": 2.6327292649031775e-07,
+      "loss": 0.9442,
+      "step": 16198
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.723638653755188,
+      "learning_rate": 2.6284965088855583e-07,
+      "loss": 0.9605,
+      "step": 16199
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6997871398925781,
+      "learning_rate": 2.624267112886525e-07,
+      "loss": 0.9037,
+      "step": 16200
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8215404748916626,
+      "learning_rate": 2.620041077052016e-07,
+      "loss": 0.8945,
+      "step": 16201
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7320815324783325,
+      "learning_rate": 2.615818401527881e-07,
+      "loss": 0.8379,
+      "step": 16202
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7023767232894897,
+      "learning_rate": 2.611599086459815e-07,
+      "loss": 0.9388,
+      "step": 16203
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7626385688781738,
+      "learning_rate": 2.607383131993424e-07,
+      "loss": 0.8888,
+      "step": 16204
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9082413911819458,
+      "learning_rate": 2.60317053827418e-07,
+      "loss": 0.9471,
+      "step": 16205
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6569337844848633,
+      "learning_rate": 2.598961305447456e-07,
+      "loss": 0.887,
+      "step": 16206
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6226907968521118,
+      "learning_rate": 2.5947554336585134e-07,
+      "loss": 0.8509,
+      "step": 16207
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0655992031097412,
+      "learning_rate": 2.5905529230524475e-07,
+      "loss": 0.5428,
+      "step": 16208
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7991751432418823,
+      "learning_rate": 2.5863537737743196e-07,
+      "loss": 0.8528,
+      "step": 16209
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5879261493682861,
+      "learning_rate": 2.5821579859689914e-07,
+      "loss": 0.8846,
+      "step": 16210
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7403408288955688,
+      "learning_rate": 2.57796555978127e-07,
+      "loss": 0.9023,
+      "step": 16211
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7279956340789795,
+      "learning_rate": 2.573776495355818e-07,
+      "loss": 0.8935,
+      "step": 16212
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7943546772003174,
+      "learning_rate": 2.5695907928371955e-07,
+      "loss": 0.9984,
+      "step": 16213
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8842931985855103,
+      "learning_rate": 2.565408452369822e-07,
+      "loss": 0.9385,
+      "step": 16214
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0422879457473755,
+      "learning_rate": 2.561229474098048e-07,
+      "loss": 0.5418,
+      "step": 16215
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6120043992996216,
+      "learning_rate": 2.5570538581660476e-07,
+      "loss": 0.8638,
+      "step": 16216
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7499767541885376,
+      "learning_rate": 2.5528816047179275e-07,
+      "loss": 0.9719,
+      "step": 16217
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8366278409957886,
+      "learning_rate": 2.5487127138976497e-07,
+      "loss": 0.8156,
+      "step": 16218
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.563151240348816,
+      "learning_rate": 2.544547185849089e-07,
+      "loss": 0.8917,
+      "step": 16219
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.81069815158844,
+      "learning_rate": 2.540385020715963e-07,
+      "loss": 0.912,
+      "step": 16220
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9182506799697876,
+      "learning_rate": 2.536226218641924e-07,
+      "loss": 0.9845,
+      "step": 16221
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7533453702926636,
+      "learning_rate": 2.532070779770446e-07,
+      "loss": 1.0208,
+      "step": 16222
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6919492483139038,
+      "learning_rate": 2.527918704244936e-07,
+      "loss": 0.8924,
+      "step": 16223
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9972357749938965,
+      "learning_rate": 2.523769992208691e-07,
+      "loss": 0.937,
+      "step": 16224
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6524839401245117,
+      "learning_rate": 2.519624643804852e-07,
+      "loss": 0.9195,
+      "step": 16225
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.898565649986267,
+      "learning_rate": 2.51548265917646e-07,
+      "loss": 0.9371,
+      "step": 16226
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.650154709815979,
+      "learning_rate": 2.511344038466457e-07,
+      "loss": 0.9478,
+      "step": 16227
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7800287008285522,
+      "learning_rate": 2.507208781817638e-07,
+      "loss": 0.9161,
+      "step": 16228
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.674424171447754,
+      "learning_rate": 2.503076889372713e-07,
+      "loss": 0.9083,
+      "step": 16229
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8096734285354614,
+      "learning_rate": 2.498948361274267e-07,
+      "loss": 0.8262,
+      "step": 16230
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7749285697937012,
+      "learning_rate": 2.494823197664742e-07,
+      "loss": 0.7951,
+      "step": 16231
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7971230745315552,
+      "learning_rate": 2.4907013986865015e-07,
+      "loss": 0.8337,
+      "step": 16232
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.022780418395996,
+      "learning_rate": 2.4865829644817764e-07,
+      "loss": 0.8521,
+      "step": 16233
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7016489505767822,
+      "learning_rate": 2.4824678951926864e-07,
+      "loss": 0.8447,
+      "step": 16234
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6620415449142456,
+      "learning_rate": 2.4783561909612063e-07,
+      "loss": 0.9028,
+      "step": 16235
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7939611673355103,
+      "learning_rate": 2.474247851929246e-07,
+      "loss": 0.7635,
+      "step": 16236
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7018694877624512,
+      "learning_rate": 2.4701428782385794e-07,
+      "loss": 0.8986,
+      "step": 16237
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8474808931350708,
+      "learning_rate": 2.4660412700308276e-07,
+      "loss": 0.8886,
+      "step": 16238
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.665451169013977,
+      "learning_rate": 2.461943027447555e-07,
+      "loss": 0.8525,
+      "step": 16239
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.752081274986267,
+      "learning_rate": 2.457848150630149e-07,
+      "loss": 0.8582,
+      "step": 16240
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7202935218811035,
+      "learning_rate": 2.4537566397199506e-07,
+      "loss": 0.8369,
+      "step": 16241
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8139984607696533,
+      "learning_rate": 2.449668494858115e-07,
+      "loss": 0.8824,
+      "step": 16242
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9071663618087769,
+      "learning_rate": 2.445583716185729e-07,
+      "loss": 0.9147,
+      "step": 16243
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.819251298904419,
+      "learning_rate": 2.4415023038437345e-07,
+      "loss": 0.8814,
+      "step": 16244
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9118560552597046,
+      "learning_rate": 2.4374242579729866e-07,
+      "loss": 0.8625,
+      "step": 16245
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0151519775390625,
+      "learning_rate": 2.4333495787141837e-07,
+      "loss": 0.5282,
+      "step": 16246
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.732537865638733,
+      "learning_rate": 2.429278266207946e-07,
+      "loss": 0.8882,
+      "step": 16247
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6146734952926636,
+      "learning_rate": 2.425210320594773e-07,
+      "loss": 0.8437,
+      "step": 16248
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.115344524383545,
+      "learning_rate": 2.4211457420150184e-07,
+      "loss": 0.5762,
+      "step": 16249
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.707578420639038,
+      "learning_rate": 2.4170845306089596e-07,
+      "loss": 0.8836,
+      "step": 16250
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8316994905471802,
+      "learning_rate": 2.4130266865167175e-07,
+      "loss": 0.879,
+      "step": 16251
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.923264980316162,
+      "learning_rate": 2.408972209878335e-07,
+      "loss": 0.8821,
+      "step": 16252
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8193572759628296,
+      "learning_rate": 2.4049211008337127e-07,
+      "loss": 0.9427,
+      "step": 16253
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6619873046875,
+      "learning_rate": 2.4008733595226376e-07,
+      "loss": 0.8937,
+      "step": 16254
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6813629865646362,
+      "learning_rate": 2.3968289860847873e-07,
+      "loss": 0.8757,
+      "step": 16255
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7966097593307495,
+      "learning_rate": 2.3927879806597274e-07,
+      "loss": 1.0082,
+      "step": 16256
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7698556184768677,
+      "learning_rate": 2.388750343386903e-07,
+      "loss": 0.8974,
+      "step": 16257
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9665919542312622,
+      "learning_rate": 2.3847160744056354e-07,
+      "loss": 0.932,
+      "step": 16258
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6363818645477295,
+      "learning_rate": 2.380685173855135e-07,
+      "loss": 0.8569,
+      "step": 16259
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6899081468582153,
+      "learning_rate": 2.3766576418745024e-07,
+      "loss": 0.9221,
+      "step": 16260
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9260165691375732,
+      "learning_rate": 2.3726334786027261e-07,
+      "loss": 0.8682,
+      "step": 16261
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7516756057739258,
+      "learning_rate": 2.3686126841786394e-07,
+      "loss": 0.8974,
+      "step": 16262
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.018200159072876,
+      "learning_rate": 2.3645952587410204e-07,
+      "loss": 0.9216,
+      "step": 16263
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8899859189987183,
+      "learning_rate": 2.3605812024284802e-07,
+      "loss": 0.7772,
+      "step": 16264
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.814500331878662,
+      "learning_rate": 2.3565705153795415e-07,
+      "loss": 0.883,
+      "step": 16265
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.633141279220581,
+      "learning_rate": 2.3525631977325825e-07,
+      "loss": 0.9266,
+      "step": 16266
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.062927007675171,
+      "learning_rate": 2.3485592496259258e-07,
+      "loss": 0.5335,
+      "step": 16267
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.648722529411316,
+      "learning_rate": 2.344558671197694e-07,
+      "loss": 0.8153,
+      "step": 16268
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7400288581848145,
+      "learning_rate": 2.3405614625859552e-07,
+      "loss": 0.9327,
+      "step": 16269
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6756274700164795,
+      "learning_rate": 2.3365676239286428e-07,
+      "loss": 0.8973,
+      "step": 16270
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7360039949417114,
+      "learning_rate": 2.332577155363569e-07,
+      "loss": 0.8616,
+      "step": 16271
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.774493932723999,
+      "learning_rate": 2.3285900570284348e-07,
+      "loss": 0.8567,
+      "step": 16272
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8044387102127075,
+      "learning_rate": 2.3246063290608189e-07,
+      "loss": 0.9097,
+      "step": 16273
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.749638557434082,
+      "learning_rate": 2.3206259715982005e-07,
+      "loss": 0.9074,
+      "step": 16274
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6550493240356445,
+      "learning_rate": 2.316648984777925e-07,
+      "loss": 0.9375,
+      "step": 16275
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8353968858718872,
+      "learning_rate": 2.3126753687372273e-07,
+      "loss": 0.9189,
+      "step": 16276
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8457140922546387,
+      "learning_rate": 2.3087051236132086e-07,
+      "loss": 0.9426,
+      "step": 16277
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9276691675186157,
+      "learning_rate": 2.3047382495429037e-07,
+      "loss": 0.9491,
+      "step": 16278
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8845618963241577,
+      "learning_rate": 2.3007747466631701e-07,
+      "loss": 0.935,
+      "step": 16279
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.841586947441101,
+      "learning_rate": 2.2968146151107872e-07,
+      "loss": 0.872,
+      "step": 16280
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.984134316444397,
+      "learning_rate": 2.2928578550224124e-07,
+      "loss": 0.5051,
+      "step": 16281
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.753153681755066,
+      "learning_rate": 2.2889044665345806e-07,
+      "loss": 0.8885,
+      "step": 16282
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0393688678741455,
+      "learning_rate": 2.2849544497837052e-07,
+      "loss": 0.5473,
+      "step": 16283
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7542282342910767,
+      "learning_rate": 2.2810078049061102e-07,
+      "loss": 0.8775,
+      "step": 16284
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7539726495742798,
+      "learning_rate": 2.2770645320379538e-07,
+      "loss": 0.8507,
+      "step": 16285
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7383103370666504,
+      "learning_rate": 2.2731246313153376e-07,
+      "loss": 0.8126,
+      "step": 16286
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8871041536331177,
+      "learning_rate": 2.2691881028741868e-07,
+      "loss": 0.8636,
+      "step": 16287
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.952820897102356,
+      "learning_rate": 2.2652549468503593e-07,
+      "loss": 0.9593,
+      "step": 16288
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8409380912780762,
+      "learning_rate": 2.2613251633795685e-07,
+      "loss": 0.963,
+      "step": 16289
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.829919457435608,
+      "learning_rate": 2.2573987525974284e-07,
+      "loss": 0.8811,
+      "step": 16290
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.87303626537323,
+      "learning_rate": 2.2534757146394305e-07,
+      "loss": 0.9335,
+      "step": 16291
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.613935112953186,
+      "learning_rate": 2.249556049640933e-07,
+      "loss": 0.8567,
+      "step": 16292
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8837584257125854,
+      "learning_rate": 2.2456397577372057e-07,
+      "loss": 0.94,
+      "step": 16293
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8737622499465942,
+      "learning_rate": 2.241726839063385e-07,
+      "loss": 0.9638,
+      "step": 16294
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6680828332901,
+      "learning_rate": 2.237817293754496e-07,
+      "loss": 0.8731,
+      "step": 16295
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6804893016815186,
+      "learning_rate": 2.2339111219454311e-07,
+      "loss": 0.8865,
+      "step": 16296
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7263448238372803,
+      "learning_rate": 2.2300083237710158e-07,
+      "loss": 0.8425,
+      "step": 16297
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9436613321304321,
+      "learning_rate": 2.2261088993658863e-07,
+      "loss": 0.5149,
+      "step": 16298
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8394290208816528,
+      "learning_rate": 2.2222128488646356e-07,
+      "loss": 0.9258,
+      "step": 16299
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6884090900421143,
+      "learning_rate": 2.2183201724016667e-07,
+      "loss": 0.8742,
+      "step": 16300
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8040472269058228,
+      "learning_rate": 2.2144308701113393e-07,
+      "loss": 0.9279,
+      "step": 16301
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7878984212875366,
+      "learning_rate": 2.210544942127857e-07,
+      "loss": 0.9037,
+      "step": 16302
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.868920087814331,
+      "learning_rate": 2.20666238858529e-07,
+      "loss": 0.9256,
+      "step": 16303
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.658509612083435,
+      "learning_rate": 2.2027832096176428e-07,
+      "loss": 0.9527,
+      "step": 16304
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.728318452835083,
+      "learning_rate": 2.1989074053587413e-07,
+      "loss": 0.8891,
+      "step": 16305
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7210944890975952,
+      "learning_rate": 2.1950349759423674e-07,
+      "loss": 0.7879,
+      "step": 16306
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.936753511428833,
+      "learning_rate": 2.1911659215021252e-07,
+      "loss": 0.4847,
+      "step": 16307
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8152029514312744,
+      "learning_rate": 2.1873002421715305e-07,
+      "loss": 0.8708,
+      "step": 16308
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8049331903457642,
+      "learning_rate": 2.1834379380839655e-07,
+      "loss": 0.9684,
+      "step": 16309
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7111965417861938,
+      "learning_rate": 2.1795790093727344e-07,
+      "loss": 0.8844,
+      "step": 16310
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.756780743598938,
+      "learning_rate": 2.175723456170964e-07,
+      "loss": 0.839,
+      "step": 16311
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.860119104385376,
+      "learning_rate": 2.1718712786117258e-07,
+      "loss": 0.8265,
+      "step": 16312
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7923767566680908,
+      "learning_rate": 2.1680224768279356e-07,
+      "loss": 0.9125,
+      "step": 16313
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8300269842147827,
+      "learning_rate": 2.1641770509524095e-07,
+      "loss": 0.8816,
+      "step": 16314
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6666057109832764,
+      "learning_rate": 2.1603350011178416e-07,
+      "loss": 0.9806,
+      "step": 16315
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7954477071762085,
+      "learning_rate": 2.1564963274568028e-07,
+      "loss": 0.8959,
+      "step": 16316
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8551993370056152,
+      "learning_rate": 2.152661030101766e-07,
+      "loss": 0.8808,
+      "step": 16317
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.796030044555664,
+      "learning_rate": 2.1488291091850577e-07,
+      "loss": 0.9032,
+      "step": 16318
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.741132140159607,
+      "learning_rate": 2.1450005648389395e-07,
+      "loss": 0.9399,
+      "step": 16319
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7671053409576416,
+      "learning_rate": 2.1411753971954941e-07,
+      "loss": 0.9029,
+      "step": 16320
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.961742103099823,
+      "learning_rate": 2.1373536063867384e-07,
+      "loss": 0.5443,
+      "step": 16321
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7057327032089233,
+      "learning_rate": 2.1335351925445335e-07,
+      "loss": 0.856,
+      "step": 16322
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6808359622955322,
+      "learning_rate": 2.129720155800652e-07,
+      "loss": 0.9321,
+      "step": 16323
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8065845966339111,
+      "learning_rate": 2.1259084962867326e-07,
+      "loss": 0.8852,
+      "step": 16324
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7093238830566406,
+      "learning_rate": 2.1221002141343261e-07,
+      "loss": 0.9935,
+      "step": 16325
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9793174266815186,
+      "learning_rate": 2.118295309474816e-07,
+      "loss": 0.8969,
+      "step": 16326
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9372689723968506,
+      "learning_rate": 2.1144937824395083e-07,
+      "loss": 0.9584,
+      "step": 16327
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7589476108551025,
+      "learning_rate": 2.1106956331595986e-07,
+      "loss": 0.9013,
+      "step": 16328
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7151142358779907,
+      "learning_rate": 2.1069008617661369e-07,
+      "loss": 0.913,
+      "step": 16329
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0841842889785767,
+      "learning_rate": 2.1031094683900855e-07,
+      "loss": 0.5228,
+      "step": 16330
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9114629030227661,
+      "learning_rate": 2.0993214531622397e-07,
+      "loss": 0.848,
+      "step": 16331
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.725567102432251,
+      "learning_rate": 2.0955368162133504e-07,
+      "loss": 0.8853,
+      "step": 16332
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6830344200134277,
+      "learning_rate": 2.0917555576740022e-07,
+      "loss": 0.7919,
+      "step": 16333
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8170216083526611,
+      "learning_rate": 2.0879776776746684e-07,
+      "loss": 0.9222,
+      "step": 16334
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7129426002502441,
+      "learning_rate": 2.0842031763457228e-07,
+      "loss": 0.8525,
+      "step": 16335
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7161333560943604,
+      "learning_rate": 2.080432053817405e-07,
+      "loss": 0.9027,
+      "step": 16336
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9157575368881226,
+      "learning_rate": 2.0766643102198448e-07,
+      "loss": 0.9308,
+      "step": 16337
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.728074073791504,
+      "learning_rate": 2.0728999456830712e-07,
+      "loss": 0.8365,
+      "step": 16338
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8389438390731812,
+      "learning_rate": 2.0691389603369695e-07,
+      "loss": 0.8551,
+      "step": 16339
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.844509482383728,
+      "learning_rate": 2.0653813543113133e-07,
+      "loss": 0.8367,
+      "step": 16340
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7989506721496582,
+      "learning_rate": 2.061627127735799e-07,
+      "loss": 0.991,
+      "step": 16341
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7589361667633057,
+      "learning_rate": 2.0578762807399343e-07,
+      "loss": 0.846,
+      "step": 16342
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.569637417793274,
+      "learning_rate": 2.0541288134531824e-07,
+      "loss": 0.8799,
+      "step": 16343
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.63938570022583,
+      "learning_rate": 2.050384726004828e-07,
+      "loss": 0.9558,
+      "step": 16344
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7267253398895264,
+      "learning_rate": 2.0466440185241021e-07,
+      "loss": 0.8851,
+      "step": 16345
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8333081007003784,
+      "learning_rate": 2.0429066911400452e-07,
+      "loss": 0.8738,
+      "step": 16346
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.885105013847351,
+      "learning_rate": 2.0391727439816655e-07,
+      "loss": 0.9303,
+      "step": 16347
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9842063188552856,
+      "learning_rate": 2.0354421771777821e-07,
+      "loss": 0.8577,
+      "step": 16348
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8866575956344604,
+      "learning_rate": 2.0317149908571475e-07,
+      "loss": 0.8653,
+      "step": 16349
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6680948734283447,
+      "learning_rate": 2.027991185148359e-07,
+      "loss": 0.8853,
+      "step": 16350
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.658965826034546,
+      "learning_rate": 2.024270760179936e-07,
+      "loss": 0.7819,
+      "step": 16351
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.595818042755127,
+      "learning_rate": 2.0205537160802202e-07,
+      "loss": 0.9195,
+      "step": 16352
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7592500448226929,
+      "learning_rate": 2.01684005297752e-07,
+      "loss": 0.9689,
+      "step": 16353
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.786474585533142,
+      "learning_rate": 2.0131297709999554e-07,
+      "loss": 0.8934,
+      "step": 16354
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7921550273895264,
+      "learning_rate": 2.0094228702755568e-07,
+      "loss": 0.9029,
+      "step": 16355
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6838581562042236,
+      "learning_rate": 2.005719350932267e-07,
+      "loss": 0.8403,
+      "step": 16356
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.04280424118042,
+      "learning_rate": 2.0020192130978611e-07,
+      "loss": 0.534,
+      "step": 16357
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.028695821762085,
+      "learning_rate": 1.998322456900026e-07,
+      "loss": 0.5503,
+      "step": 16358
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8095715045928955,
+      "learning_rate": 1.9946290824663262e-07,
+      "loss": 0.9712,
+      "step": 16359
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.777685523033142,
+      "learning_rate": 1.9909390899242153e-07,
+      "loss": 0.8318,
+      "step": 16360
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6515487432479858,
+      "learning_rate": 1.987252479401014e-07,
+      "loss": 0.8473,
+      "step": 16361
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7955330610275269,
+      "learning_rate": 1.9835692510239424e-07,
+      "loss": 0.8583,
+      "step": 16362
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7963247299194336,
+      "learning_rate": 1.9798894049200878e-07,
+      "loss": 0.8396,
+      "step": 16363
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.807252287864685,
+      "learning_rate": 1.97621294121646e-07,
+      "loss": 0.8203,
+      "step": 16364
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7473043203353882,
+      "learning_rate": 1.9725398600398905e-07,
+      "loss": 0.9325,
+      "step": 16365
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8721873760223389,
+      "learning_rate": 1.9688701615171558e-07,
+      "loss": 0.9282,
+      "step": 16366
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.081289052963257,
+      "learning_rate": 1.9652038457748547e-07,
+      "loss": 0.9149,
+      "step": 16367
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6495956182479858,
+      "learning_rate": 1.9615409129395303e-07,
+      "loss": 0.9168,
+      "step": 16368
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8121864795684814,
+      "learning_rate": 1.9578813631375704e-07,
+      "loss": 0.9037,
+      "step": 16369
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7582117319107056,
+      "learning_rate": 1.9542251964952518e-07,
+      "loss": 0.8961,
+      "step": 16370
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8538521528244019,
+      "learning_rate": 1.9505724131387515e-07,
+      "loss": 0.9293,
+      "step": 16371
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.753697395324707,
+      "learning_rate": 1.9469230131940907e-07,
+      "loss": 0.8655,
+      "step": 16372
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8507347106933594,
+      "learning_rate": 1.9432769967872357e-07,
+      "loss": 0.8575,
+      "step": 16373
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8493958711624146,
+      "learning_rate": 1.939634364043963e-07,
+      "loss": 0.9094,
+      "step": 16374
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6765323877334595,
+      "learning_rate": 1.9359951150900059e-07,
+      "loss": 0.841,
+      "step": 16375
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8059134483337402,
+      "learning_rate": 1.9323592500509082e-07,
+      "loss": 0.8376,
+      "step": 16376
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9414695501327515,
+      "learning_rate": 1.9287267690521582e-07,
+      "loss": 0.932,
+      "step": 16377
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8842369318008423,
+      "learning_rate": 1.9250976722191006e-07,
+      "loss": 0.8829,
+      "step": 16378
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8446831703186035,
+      "learning_rate": 1.921471959676957e-07,
+      "loss": 0.9176,
+      "step": 16379
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.861767292022705,
+      "learning_rate": 1.91784963155085e-07,
+      "loss": 0.9212,
+      "step": 16380
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.675605058670044,
+      "learning_rate": 1.9142306879657569e-07,
+      "loss": 0.9243,
+      "step": 16381
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.812416911125183,
+      "learning_rate": 1.910615129046589e-07,
+      "loss": 0.8695,
+      "step": 16382
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0805389881134033,
+      "learning_rate": 1.9070029549180802e-07,
+      "loss": 0.5867,
+      "step": 16383
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7600376605987549,
+      "learning_rate": 1.903394165704897e-07,
+      "loss": 0.9126,
+      "step": 16384
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9993528723716736,
+      "learning_rate": 1.8997887615315513e-07,
+      "loss": 0.5213,
+      "step": 16385
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7143893241882324,
+      "learning_rate": 1.8961867425224655e-07,
+      "loss": 0.8144,
+      "step": 16386
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6862927675247192,
+      "learning_rate": 1.8925881088019292e-07,
+      "loss": 0.9937,
+      "step": 16387
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7340646982192993,
+      "learning_rate": 1.8889928604941432e-07,
+      "loss": 0.9101,
+      "step": 16388
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7498443126678467,
+      "learning_rate": 1.8854009977231303e-07,
+      "loss": 0.8884,
+      "step": 16389
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8975346088409424,
+      "learning_rate": 1.881812520612869e-07,
+      "loss": 0.8528,
+      "step": 16390
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7859089374542236,
+      "learning_rate": 1.8782274292871717e-07,
+      "loss": 0.9032,
+      "step": 16391
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7822070121765137,
+      "learning_rate": 1.874645723869739e-07,
+      "loss": 0.9685,
+      "step": 16392
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7800108194351196,
+      "learning_rate": 1.8710674044842058e-07,
+      "loss": 0.9029,
+      "step": 16393
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.650575041770935,
+      "learning_rate": 1.8674924712540066e-07,
+      "loss": 0.8885,
+      "step": 16394
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7646385431289673,
+      "learning_rate": 1.8639209243025315e-07,
+      "loss": 0.9304,
+      "step": 16395
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.785351037979126,
+      "learning_rate": 1.860352763753004e-07,
+      "loss": 0.8679,
+      "step": 16396
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7265164852142334,
+      "learning_rate": 1.85678798972857e-07,
+      "loss": 0.9392,
+      "step": 16397
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.689312219619751,
+      "learning_rate": 1.8532266023522206e-07,
+      "loss": 0.8759,
+      "step": 16398
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7069814205169678,
+      "learning_rate": 1.849668601746868e-07,
+      "loss": 0.8919,
+      "step": 16399
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6813130378723145,
+      "learning_rate": 1.8461139880352695e-07,
+      "loss": 0.9555,
+      "step": 16400
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9528535604476929,
+      "learning_rate": 1.8425627613401165e-07,
+      "loss": 0.8984,
+      "step": 16401
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0073697566986084,
+      "learning_rate": 1.8390149217839103e-07,
+      "loss": 0.8432,
+      "step": 16402
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6456698179244995,
+      "learning_rate": 1.835470469489109e-07,
+      "loss": 0.8569,
+      "step": 16403
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7223634719848633,
+      "learning_rate": 1.8319294045779922e-07,
+      "loss": 0.8948,
+      "step": 16404
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.747360110282898,
+      "learning_rate": 1.8283917271727846e-07,
+      "loss": 0.9302,
+      "step": 16405
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7781614065170288,
+      "learning_rate": 1.8248574373955442e-07,
+      "loss": 0.9435,
+      "step": 16406
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0016318559646606,
+      "learning_rate": 1.8213265353682396e-07,
+      "loss": 0.535,
+      "step": 16407
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8278100490570068,
+      "learning_rate": 1.8177990212126962e-07,
+      "loss": 0.8483,
+      "step": 16408
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8017596006393433,
+      "learning_rate": 1.8142748950506494e-07,
+      "loss": 0.8634,
+      "step": 16409
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9471365213394165,
+      "learning_rate": 1.8107541570037136e-07,
+      "loss": 0.8854,
+      "step": 16410
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.760509729385376,
+      "learning_rate": 1.8072368071933577e-07,
+      "loss": 0.88,
+      "step": 16411
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8247904777526855,
+      "learning_rate": 1.803722845740985e-07,
+      "loss": 0.8408,
+      "step": 16412
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8561985492706299,
+      "learning_rate": 1.8002122727678096e-07,
+      "loss": 0.9246,
+      "step": 16413
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6995141506195068,
+      "learning_rate": 1.7967050883950233e-07,
+      "loss": 0.8736,
+      "step": 16414
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.039583444595337,
+      "learning_rate": 1.7932012927436182e-07,
+      "loss": 0.8991,
+      "step": 16415
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7881120443344116,
+      "learning_rate": 1.789700885934509e-07,
+      "loss": 0.9653,
+      "step": 16416
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7039881944656372,
+      "learning_rate": 1.7862038680884763e-07,
+      "loss": 0.8661,
+      "step": 16417
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7617466449737549,
+      "learning_rate": 1.782710239326202e-07,
+      "loss": 0.9392,
+      "step": 16418
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8066859245300293,
+      "learning_rate": 1.7792199997682335e-07,
+      "loss": 0.8956,
+      "step": 16419
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7197020053863525,
+      "learning_rate": 1.7757331495350306e-07,
+      "loss": 0.9227,
+      "step": 16420
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.004526376724243,
+      "learning_rate": 1.7722496887468854e-07,
+      "loss": 0.9252,
+      "step": 16421
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8045226335525513,
+      "learning_rate": 1.768769617524013e-07,
+      "loss": 0.8541,
+      "step": 16422
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7096847295761108,
+      "learning_rate": 1.7652929359865068e-07,
+      "loss": 0.9347,
+      "step": 16423
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8708813190460205,
+      "learning_rate": 1.7618196442543366e-07,
+      "loss": 0.8864,
+      "step": 16424
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.799077033996582,
+      "learning_rate": 1.7583497424473516e-07,
+      "loss": 0.9929,
+      "step": 16425
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9251153469085693,
+      "learning_rate": 1.7548832306852893e-07,
+      "loss": 0.8797,
+      "step": 16426
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7964842319488525,
+      "learning_rate": 1.7514201090877758e-07,
+      "loss": 0.8982,
+      "step": 16427
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.82236909866333,
+      "learning_rate": 1.7479603777742937e-07,
+      "loss": 0.8563,
+      "step": 16428
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5631624460220337,
+      "learning_rate": 1.7445040368642585e-07,
+      "loss": 0.9045,
+      "step": 16429
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8029755353927612,
+      "learning_rate": 1.7410510864769081e-07,
+      "loss": 0.9283,
+      "step": 16430
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9529526233673096,
+      "learning_rate": 1.737601526731425e-07,
+      "loss": 0.8742,
+      "step": 16431
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.764808177947998,
+      "learning_rate": 1.7341553577468252e-07,
+      "loss": 0.8237,
+      "step": 16432
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8694605827331543,
+      "learning_rate": 1.7307125796420133e-07,
+      "loss": 1.0193,
+      "step": 16433
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9062522649765015,
+      "learning_rate": 1.7272731925358277e-07,
+      "loss": 0.9472,
+      "step": 16434
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9673916101455688,
+      "learning_rate": 1.723837196546918e-07,
+      "loss": 0.8847,
+      "step": 16435
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.846174955368042,
+      "learning_rate": 1.7204045917938672e-07,
+      "loss": 0.9184,
+      "step": 16436
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.642804503440857,
+      "learning_rate": 1.7169753783951137e-07,
+      "loss": 0.8761,
+      "step": 16437
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.61570143699646,
+      "learning_rate": 1.7135495564690075e-07,
+      "loss": 0.8591,
+      "step": 16438
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7706712484359741,
+      "learning_rate": 1.7101271261337537e-07,
+      "loss": 0.8105,
+      "step": 16439
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.879559874534607,
+      "learning_rate": 1.706708087507447e-07,
+      "loss": 0.9938,
+      "step": 16440
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7266027927398682,
+      "learning_rate": 1.7032924407080709e-07,
+      "loss": 0.8774,
+      "step": 16441
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.918654441833496,
+      "learning_rate": 1.6998801858534975e-07,
+      "loss": 0.9498,
+      "step": 16442
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8942171335220337,
+      "learning_rate": 1.696471323061466e-07,
+      "loss": 0.9064,
+      "step": 16443
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7254973649978638,
+      "learning_rate": 1.6930658524496158e-07,
+      "loss": 0.9206,
+      "step": 16444
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9922207593917847,
+      "learning_rate": 1.6896637741354417e-07,
+      "loss": 0.912,
+      "step": 16445
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7686071395874023,
+      "learning_rate": 1.6862650882363497e-07,
+      "loss": 0.8724,
+      "step": 16446
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8459292650222778,
+      "learning_rate": 1.6828697948696348e-07,
+      "loss": 0.9444,
+      "step": 16447
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7722623348236084,
+      "learning_rate": 1.679477894152437e-07,
+      "loss": 0.9586,
+      "step": 16448
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5576446056365967,
+      "learning_rate": 1.676089386201818e-07,
+      "loss": 0.9035,
+      "step": 16449
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6673877239227295,
+      "learning_rate": 1.672704271134684e-07,
+      "loss": 0.9382,
+      "step": 16450
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7748799324035645,
+      "learning_rate": 1.6693225490678755e-07,
+      "loss": 0.8937,
+      "step": 16451
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7299423217773438,
+      "learning_rate": 1.6659442201180543e-07,
+      "loss": 0.933,
+      "step": 16452
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8745492696762085,
+      "learning_rate": 1.662569284401827e-07,
+      "loss": 0.8875,
+      "step": 16453
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.934075951576233,
+      "learning_rate": 1.659197742035623e-07,
+      "loss": 0.9163,
+      "step": 16454
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8632644414901733,
+      "learning_rate": 1.655829593135816e-07,
+      "loss": 0.8921,
+      "step": 16455
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5784367322921753,
+      "learning_rate": 1.6524648378186125e-07,
+      "loss": 0.863,
+      "step": 16456
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8589651584625244,
+      "learning_rate": 1.649103476200131e-07,
+      "loss": 0.897,
+      "step": 16457
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.707099199295044,
+      "learning_rate": 1.6457455083963346e-07,
+      "loss": 0.8793,
+      "step": 16458
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6957958936691284,
+      "learning_rate": 1.6423909345231304e-07,
+      "loss": 0.8291,
+      "step": 16459
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.764585256576538,
+      "learning_rate": 1.639039754696281e-07,
+      "loss": 0.9135,
+      "step": 16460
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7115422487258911,
+      "learning_rate": 1.6356919690313832e-07,
+      "loss": 0.8875,
+      "step": 16461
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8564659357070923,
+      "learning_rate": 1.632347577644e-07,
+      "loss": 0.8248,
+      "step": 16462
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8585751056671143,
+      "learning_rate": 1.629006580649528e-07,
+      "loss": 0.8287,
+      "step": 16463
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.029334545135498,
+      "learning_rate": 1.6256689781632416e-07,
+      "loss": 0.5804,
+      "step": 16464
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0663323402404785,
+      "learning_rate": 1.6223347703003156e-07,
+      "loss": 0.9443,
+      "step": 16465
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7953803539276123,
+      "learning_rate": 1.6190039571758243e-07,
+      "loss": 0.9376,
+      "step": 16466
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6454825401306152,
+      "learning_rate": 1.6156765389046868e-07,
+      "loss": 0.9291,
+      "step": 16467
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9765831232070923,
+      "learning_rate": 1.6123525156017228e-07,
+      "loss": 0.935,
+      "step": 16468
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6322416067123413,
+      "learning_rate": 1.6090318873816292e-07,
+      "loss": 0.8417,
+      "step": 16469
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8629791736602783,
+      "learning_rate": 1.605714654359014e-07,
+      "loss": 0.8786,
+      "step": 16470
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.706268310546875,
+      "learning_rate": 1.6024008166483308e-07,
+      "loss": 0.8945,
+      "step": 16471
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7280964851379395,
+      "learning_rate": 1.5990903743639318e-07,
+      "loss": 0.8963,
+      "step": 16472
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7247207164764404,
+      "learning_rate": 1.5957833276200486e-07,
+      "loss": 0.9038,
+      "step": 16473
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.740746259689331,
+      "learning_rate": 1.5924796765308004e-07,
+      "loss": 0.9098,
+      "step": 16474
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7239288091659546,
+      "learning_rate": 1.5891794212102073e-07,
+      "loss": 0.8278,
+      "step": 16475
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7507599592208862,
+      "learning_rate": 1.585882561772112e-07,
+      "loss": 0.9142,
+      "step": 16476
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9095275402069092,
+      "learning_rate": 1.5825890983303115e-07,
+      "loss": 0.9631,
+      "step": 16477
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.691354751586914,
+      "learning_rate": 1.579299030998427e-07,
+      "loss": 0.8627,
+      "step": 16478
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7601280212402344,
+      "learning_rate": 1.5760123598900222e-07,
+      "loss": 0.8822,
+      "step": 16479
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9213297367095947,
+      "learning_rate": 1.572729085118485e-07,
+      "loss": 0.7579,
+      "step": 16480
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9090956449508667,
+      "learning_rate": 1.5694492067971245e-07,
+      "loss": 0.9249,
+      "step": 16481
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7180875539779663,
+      "learning_rate": 1.5661727250391167e-07,
+      "loss": 0.8416,
+      "step": 16482
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7013002634048462,
+      "learning_rate": 1.562899639957538e-07,
+      "loss": 0.8544,
+      "step": 16483
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7529064416885376,
+      "learning_rate": 1.559629951665298e-07,
+      "loss": 0.8571,
+      "step": 16484
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.568910002708435,
+      "learning_rate": 1.556363660275262e-07,
+      "loss": 0.7811,
+      "step": 16485
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8314090967178345,
+      "learning_rate": 1.5531007659001175e-07,
+      "loss": 0.9341,
+      "step": 16486
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8949534893035889,
+      "learning_rate": 1.5498412686524634e-07,
+      "loss": 0.8667,
+      "step": 16487
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0249289274215698,
+      "learning_rate": 1.5465851686447876e-07,
+      "loss": 0.5592,
+      "step": 16488
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.738390564918518,
+      "learning_rate": 1.5433324659894333e-07,
+      "loss": 0.9648,
+      "step": 16489
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6103278398513794,
+      "learning_rate": 1.5400831607986443e-07,
+      "loss": 0.8457,
+      "step": 16490
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7184944152832031,
+      "learning_rate": 1.5368372531845532e-07,
+      "loss": 0.9577,
+      "step": 16491
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7703791856765747,
+      "learning_rate": 1.53359474325917e-07,
+      "loss": 0.8672,
+      "step": 16492
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9693399667739868,
+      "learning_rate": 1.5303556311343616e-07,
+      "loss": 0.8843,
+      "step": 16493
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8252665996551514,
+      "learning_rate": 1.5271199169219264e-07,
+      "loss": 0.8591,
+      "step": 16494
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.89171302318573,
+      "learning_rate": 1.5238876007335091e-07,
+      "loss": 0.8467,
+      "step": 16495
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.709255576133728,
+      "learning_rate": 1.5206586826806536e-07,
+      "loss": 0.8933,
+      "step": 16496
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.865009069442749,
+      "learning_rate": 1.5174331628747596e-07,
+      "loss": 0.8926,
+      "step": 16497
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.86410391330719,
+      "learning_rate": 1.5142110414271716e-07,
+      "loss": 0.9913,
+      "step": 16498
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7553921937942505,
+      "learning_rate": 1.510992318449034e-07,
+      "loss": 0.9054,
+      "step": 16499
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7745091915130615,
+      "learning_rate": 1.5077769940514242e-07,
+      "loss": 0.9481,
+      "step": 16500
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.043612241744995,
+      "learning_rate": 1.5045650683453205e-07,
+      "loss": 0.8476,
+      "step": 16501
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9572433233261108,
+      "learning_rate": 1.5013565414415342e-07,
+      "loss": 0.9025,
+      "step": 16502
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.794930338859558,
+      "learning_rate": 1.498151413450799e-07,
+      "loss": 0.8519,
+      "step": 16503
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.702752709388733,
+      "learning_rate": 1.494949684483693e-07,
+      "loss": 0.9376,
+      "step": 16504
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.575696587562561,
+      "learning_rate": 1.4917513546507167e-07,
+      "loss": 0.8486,
+      "step": 16505
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7192968130111694,
+      "learning_rate": 1.4885564240622265e-07,
+      "loss": 0.9022,
+      "step": 16506
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6231508255004883,
+      "learning_rate": 1.485364892828478e-07,
+      "loss": 0.8916,
+      "step": 16507
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7822118997573853,
+      "learning_rate": 1.482176761059584e-07,
+      "loss": 0.9637,
+      "step": 16508
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7868828773498535,
+      "learning_rate": 1.478992028865589e-07,
+      "loss": 0.9137,
+      "step": 16509
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8035506010055542,
+      "learning_rate": 1.4758106963563612e-07,
+      "loss": 0.9549,
+      "step": 16510
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8400654792785645,
+      "learning_rate": 1.4726327636417015e-07,
+      "loss": 0.8627,
+      "step": 16511
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8744951486587524,
+      "learning_rate": 1.4694582308312444e-07,
+      "loss": 0.866,
+      "step": 16512
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7350282669067383,
+      "learning_rate": 1.466287098034558e-07,
+      "loss": 1.0024,
+      "step": 16513
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.10121750831604,
+      "learning_rate": 1.4631193653610542e-07,
+      "loss": 0.5384,
+      "step": 16514
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0601329803466797,
+      "learning_rate": 1.4599550329200574e-07,
+      "loss": 0.928,
+      "step": 16515
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8255534172058105,
+      "learning_rate": 1.4567941008207466e-07,
+      "loss": 1.0136,
+      "step": 16516
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7194275856018066,
+      "learning_rate": 1.4536365691722122e-07,
+      "loss": 0.8817,
+      "step": 16517
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8004549741744995,
+      "learning_rate": 1.4504824380833892e-07,
+      "loss": 0.9848,
+      "step": 16518
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.864165186882019,
+      "learning_rate": 1.4473317076631355e-07,
+      "loss": 0.9286,
+      "step": 16519
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7690070867538452,
+      "learning_rate": 1.4441843780201747e-07,
+      "loss": 0.9085,
+      "step": 16520
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7431583404541016,
+      "learning_rate": 1.441040449263098e-07,
+      "loss": 0.9473,
+      "step": 16521
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7609999179840088,
+      "learning_rate": 1.4378999215004076e-07,
+      "loss": 0.9518,
+      "step": 16522
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7276779413223267,
+      "learning_rate": 1.4347627948404607e-07,
+      "loss": 0.927,
+      "step": 16523
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0218112468719482,
+      "learning_rate": 1.431629069391516e-07,
+      "loss": 0.927,
+      "step": 16524
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.705073356628418,
+      "learning_rate": 1.4284987452617306e-07,
+      "loss": 0.9891,
+      "step": 16525
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9493002891540527,
+      "learning_rate": 1.425371822559085e-07,
+      "loss": 0.898,
+      "step": 16526
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7836133241653442,
+      "learning_rate": 1.4222483013915156e-07,
+      "loss": 0.9047,
+      "step": 16527
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9222790002822876,
+      "learning_rate": 1.4191281818667914e-07,
+      "loss": 0.9553,
+      "step": 16528
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.626082181930542,
+      "learning_rate": 1.4160114640925704e-07,
+      "loss": 0.8374,
+      "step": 16529
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8756816387176514,
+      "learning_rate": 1.4128981481764115e-07,
+      "loss": 0.8733,
+      "step": 16530
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7788617610931396,
+      "learning_rate": 1.4097882342257508e-07,
+      "loss": 0.9497,
+      "step": 16531
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.744699478149414,
+      "learning_rate": 1.4066817223478913e-07,
+      "loss": 0.8517,
+      "step": 16532
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8373404741287231,
+      "learning_rate": 1.4035786126500473e-07,
+      "loss": 0.935,
+      "step": 16533
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7422071695327759,
+      "learning_rate": 1.4004789052392777e-07,
+      "loss": 0.8591,
+      "step": 16534
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9151790142059326,
+      "learning_rate": 1.3973826002225631e-07,
+      "loss": 0.8738,
+      "step": 16535
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6773675680160522,
+      "learning_rate": 1.39428969770673e-07,
+      "loss": 0.8623,
+      "step": 16536
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6714239120483398,
+      "learning_rate": 1.3912001977985146e-07,
+      "loss": 0.8179,
+      "step": 16537
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.038982629776001,
+      "learning_rate": 1.3881141006045318e-07,
+      "loss": 0.9548,
+      "step": 16538
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7970119714736938,
+      "learning_rate": 1.3850314062312742e-07,
+      "loss": 0.9082,
+      "step": 16539
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.786732792854309,
+      "learning_rate": 1.3819521147851122e-07,
+      "loss": 0.8333,
+      "step": 16540
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8275790214538574,
+      "learning_rate": 1.3788762263722943e-07,
+      "loss": 1.0284,
+      "step": 16541
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7548185586929321,
+      "learning_rate": 1.37580374109898e-07,
+      "loss": 0.9106,
+      "step": 16542
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8583842515945435,
+      "learning_rate": 1.3727346590711843e-07,
+      "loss": 0.8834,
+      "step": 16543
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0873218774795532,
+      "learning_rate": 1.3696689803948114e-07,
+      "loss": 0.5724,
+      "step": 16544
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8137444257736206,
+      "learning_rate": 1.3666067051756326e-07,
+      "loss": 0.8949,
+      "step": 16545
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7303413152694702,
+      "learning_rate": 1.363547833519352e-07,
+      "loss": 0.8784,
+      "step": 16546
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6444858312606812,
+      "learning_rate": 1.360492365531496e-07,
+      "loss": 0.8531,
+      "step": 16547
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7975293397903442,
+      "learning_rate": 1.357440301317525e-07,
+      "loss": 0.8912,
+      "step": 16548
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8063907623291016,
+      "learning_rate": 1.3543916409827328e-07,
+      "loss": 0.8986,
+      "step": 16549
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8259648084640503,
+      "learning_rate": 1.351346384632335e-07,
+      "loss": 0.9029,
+      "step": 16550
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8852148056030273,
+      "learning_rate": 1.348304532371403e-07,
+      "loss": 0.884,
+      "step": 16551
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6538900136947632,
+      "learning_rate": 1.3452660843049082e-07,
+      "loss": 0.8997,
+      "step": 16552
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6715694665908813,
+      "learning_rate": 1.3422310405377003e-07,
+      "loss": 0.9102,
+      "step": 16553
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8713077306747437,
+      "learning_rate": 1.339199401174507e-07,
+      "loss": 0.885,
+      "step": 16554
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7181580066680908,
+      "learning_rate": 1.336171166319955e-07,
+      "loss": 0.9071,
+      "step": 16555
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.732968807220459,
+      "learning_rate": 1.333146336078528e-07,
+      "loss": 0.8967,
+      "step": 16556
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7870614528656006,
+      "learning_rate": 1.3301249105546087e-07,
+      "loss": 0.8683,
+      "step": 16557
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8052188158035278,
+      "learning_rate": 1.3271068898524475e-07,
+      "loss": 0.8922,
+      "step": 16558
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.726914405822754,
+      "learning_rate": 1.3240922740761942e-07,
+      "loss": 0.8456,
+      "step": 16559
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9734898209571838,
+      "learning_rate": 1.3210810633298765e-07,
+      "loss": 0.5558,
+      "step": 16560
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9070172309875488,
+      "learning_rate": 1.3180732577174117e-07,
+      "loss": 0.8776,
+      "step": 16561
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9358142614364624,
+      "learning_rate": 1.3150688573425829e-07,
+      "loss": 0.9322,
+      "step": 16562
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6691879034042358,
+      "learning_rate": 1.3120678623090521e-07,
+      "loss": 0.9626,
+      "step": 16563
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8167921304702759,
+      "learning_rate": 1.3090702727203918e-07,
+      "loss": 0.8886,
+      "step": 16564
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7664705514907837,
+      "learning_rate": 1.3060760886800417e-07,
+      "loss": 0.9267,
+      "step": 16565
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7893266677856445,
+      "learning_rate": 1.3030853102912965e-07,
+      "loss": 0.8985,
+      "step": 16566
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8319549560546875,
+      "learning_rate": 1.300097937657385e-07,
+      "loss": 0.8323,
+      "step": 16567
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7281087636947632,
+      "learning_rate": 1.297113970881403e-07,
+      "loss": 0.8651,
+      "step": 16568
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7932853698730469,
+      "learning_rate": 1.2941334100662784e-07,
+      "loss": 0.971,
+      "step": 16569
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9154644012451172,
+      "learning_rate": 1.291156255314907e-07,
+      "loss": 0.4918,
+      "step": 16570
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6723839044570923,
+      "learning_rate": 1.2881825067299848e-07,
+      "loss": 0.8406,
+      "step": 16571
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8287107944488525,
+      "learning_rate": 1.2852121644141624e-07,
+      "loss": 0.8858,
+      "step": 16572
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8438234329223633,
+      "learning_rate": 1.2822452284699026e-07,
+      "loss": 0.9524,
+      "step": 16573
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7729114294052124,
+      "learning_rate": 1.2792816989996127e-07,
+      "loss": 0.924,
+      "step": 16574
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7965925931930542,
+      "learning_rate": 1.2763215761055437e-07,
+      "loss": 0.8552,
+      "step": 16575
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7768440246582031,
+      "learning_rate": 1.2733648598898475e-07,
+      "loss": 0.8628,
+      "step": 16576
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.625126600265503,
+      "learning_rate": 1.2704115504545312e-07,
+      "loss": 0.8198,
+      "step": 16577
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.651566982269287,
+      "learning_rate": 1.2674616479015355e-07,
+      "loss": 0.9041,
+      "step": 16578
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6268478631973267,
+      "learning_rate": 1.264515152332646e-07,
+      "loss": 0.9429,
+      "step": 16579
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.606135368347168,
+      "learning_rate": 1.2615720638495142e-07,
+      "loss": 0.8608,
+      "step": 16580
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6384949684143066,
+      "learning_rate": 1.2586323825537372e-07,
+      "loss": 0.8969,
+      "step": 16581
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7512062788009644,
+      "learning_rate": 1.2556961085467223e-07,
+      "loss": 0.8414,
+      "step": 16582
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6920546293258667,
+      "learning_rate": 1.2527632419297997e-07,
+      "loss": 0.9094,
+      "step": 16583
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5868085622787476,
+      "learning_rate": 1.2498337828041886e-07,
+      "loss": 0.7516,
+      "step": 16584
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8788325786590576,
+      "learning_rate": 1.2469077312709633e-07,
+      "loss": 0.8882,
+      "step": 16585
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8532047271728516,
+      "learning_rate": 1.2439850874310877e-07,
+      "loss": 0.8798,
+      "step": 16586
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.985978603363037,
+      "learning_rate": 1.241065851385437e-07,
+      "loss": 0.8682,
+      "step": 16587
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7965304851531982,
+      "learning_rate": 1.2381500232347188e-07,
+      "loss": 0.8789,
+      "step": 16588
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6530011892318726,
+      "learning_rate": 1.2352376030795753e-07,
+      "loss": 0.89,
+      "step": 16589
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7447978258132935,
+      "learning_rate": 1.232328591020482e-07,
+      "loss": 0.8827,
+      "step": 16590
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.773024559020996,
+      "learning_rate": 1.2294229871578356e-07,
+      "loss": 0.8629,
+      "step": 16591
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.684670090675354,
+      "learning_rate": 1.2265207915919008e-07,
+      "loss": 0.9349,
+      "step": 16592
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7269409894943237,
+      "learning_rate": 1.2236220044228196e-07,
+      "loss": 0.9854,
+      "step": 16593
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6446491479873657,
+      "learning_rate": 1.2207266257506234e-07,
+      "loss": 0.92,
+      "step": 16594
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7803030014038086,
+      "learning_rate": 1.2178346556752207e-07,
+      "loss": 0.8812,
+      "step": 16595
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9164819717407227,
+      "learning_rate": 1.2149460942964097e-07,
+      "loss": 0.9484,
+      "step": 16596
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.79301917552948,
+      "learning_rate": 1.212060941713855e-07,
+      "loss": 0.9546,
+      "step": 16597
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.036015033721924,
+      "learning_rate": 1.2091791980271438e-07,
+      "loss": 0.9546,
+      "step": 16598
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8012720346450806,
+      "learning_rate": 1.2063008633356743e-07,
+      "loss": 0.8533,
+      "step": 16599
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7246770858764648,
+      "learning_rate": 1.2034259377388113e-07,
+      "loss": 0.9456,
+      "step": 16600
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6861293315887451,
+      "learning_rate": 1.200554421335731e-07,
+      "loss": 0.9462,
+      "step": 16601
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8792359828948975,
+      "learning_rate": 1.1976863142255324e-07,
+      "loss": 0.929,
+      "step": 16602
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8180543184280396,
+      "learning_rate": 1.1948216165071912e-07,
+      "loss": 0.9709,
+      "step": 16603
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9028102159500122,
+      "learning_rate": 1.191960328279551e-07,
+      "loss": 0.8795,
+      "step": 16604
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5768588781356812,
+      "learning_rate": 1.1891024496413661e-07,
+      "loss": 0.8467,
+      "step": 16605
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8773727416992188,
+      "learning_rate": 1.186247980691213e-07,
+      "loss": 0.881,
+      "step": 16606
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9762147665023804,
+      "learning_rate": 1.1833969215276352e-07,
+      "loss": 0.8532,
+      "step": 16607
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.891180396080017,
+      "learning_rate": 1.1805492722489876e-07,
+      "loss": 0.9476,
+      "step": 16608
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7515558004379272,
+      "learning_rate": 1.1777050329535577e-07,
+      "loss": 0.8814,
+      "step": 16609
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.96222984790802,
+      "learning_rate": 1.1748642037394564e-07,
+      "loss": 0.9238,
+      "step": 16610
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8126362562179565,
+      "learning_rate": 1.1720267847047495e-07,
+      "loss": 0.9303,
+      "step": 16611
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5835965871810913,
+      "learning_rate": 1.1691927759473254e-07,
+      "loss": 0.8836,
+      "step": 16612
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0533149242401123,
+      "learning_rate": 1.1663621775649946e-07,
+      "loss": 0.9243,
+      "step": 16613
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7795202732086182,
+      "learning_rate": 1.1635349896554126e-07,
+      "loss": 0.9172,
+      "step": 16614
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.817523717880249,
+      "learning_rate": 1.1607112123161679e-07,
+      "loss": 0.885,
+      "step": 16615
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8167636394500732,
+      "learning_rate": 1.1578908456446713e-07,
+      "loss": 0.8674,
+      "step": 16616
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2300267219543457,
+      "learning_rate": 1.1550738897382563e-07,
+      "loss": 0.8808,
+      "step": 16617
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9196256399154663,
+      "learning_rate": 1.1522603446941338e-07,
+      "loss": 0.892,
+      "step": 16618
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8228071928024292,
+      "learning_rate": 1.1494502106093708e-07,
+      "loss": 0.9416,
+      "step": 16619
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7819263935089111,
+      "learning_rate": 1.1466434875809784e-07,
+      "loss": 0.9334,
+      "step": 16620
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.844464898109436,
+      "learning_rate": 1.1438401757057682e-07,
+      "loss": 0.9009,
+      "step": 16621
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8060771226882935,
+      "learning_rate": 1.1410402750804961e-07,
+      "loss": 0.906,
+      "step": 16622
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8951303958892822,
+      "learning_rate": 1.1382437858017626e-07,
+      "loss": 0.8879,
+      "step": 16623
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7181979417800903,
+      "learning_rate": 1.1354507079660904e-07,
+      "loss": 0.7693,
+      "step": 16624
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.841902494430542,
+      "learning_rate": 1.1326610416698358e-07,
+      "loss": 1.0096,
+      "step": 16625
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.808828353881836,
+      "learning_rate": 1.1298747870092886e-07,
+      "loss": 0.8968,
+      "step": 16626
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9094401597976685,
+      "learning_rate": 1.1270919440805605e-07,
+      "loss": 0.8973,
+      "step": 16627
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7344419956207275,
+      "learning_rate": 1.1243125129797194e-07,
+      "loss": 0.8725,
+      "step": 16628
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7809631824493408,
+      "learning_rate": 1.121536493802644e-07,
+      "loss": 0.9343,
+      "step": 16629
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8784180879592896,
+      "learning_rate": 1.1187638866451355e-07,
+      "loss": 0.8436,
+      "step": 16630
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8009629249572754,
+      "learning_rate": 1.115994691602873e-07,
+      "loss": 0.8634,
+      "step": 16631
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7326936721801758,
+      "learning_rate": 1.1132289087714132e-07,
+      "loss": 0.8608,
+      "step": 16632
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8138861656188965,
+      "learning_rate": 1.110466538246202e-07,
+      "loss": 0.8992,
+      "step": 16633
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.681660532951355,
+      "learning_rate": 1.1077075801225412e-07,
+      "loss": 0.924,
+      "step": 16634
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7696179151535034,
+      "learning_rate": 1.1049520344956654e-07,
+      "loss": 0.9029,
+      "step": 16635
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.738041877746582,
+      "learning_rate": 1.1021999014606322e-07,
+      "loss": 0.9149,
+      "step": 16636
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8930988311767578,
+      "learning_rate": 1.099451181112421e-07,
+      "loss": 0.8898,
+      "step": 16637
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8724775314331055,
+      "learning_rate": 1.096705873545878e-07,
+      "loss": 0.8356,
+      "step": 16638
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8593138456344604,
+      "learning_rate": 1.0939639788557499e-07,
+      "loss": 0.895,
+      "step": 16639
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7877471446990967,
+      "learning_rate": 1.0912254971366276e-07,
+      "loss": 0.895,
+      "step": 16640
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6651346683502197,
+      "learning_rate": 1.0884904284830355e-07,
+      "loss": 0.8845,
+      "step": 16641
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0390714406967163,
+      "learning_rate": 1.0857587729893316e-07,
+      "loss": 0.5337,
+      "step": 16642
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6564645767211914,
+      "learning_rate": 1.0830305307497958e-07,
+      "loss": 0.8884,
+      "step": 16643
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1324530839920044,
+      "learning_rate": 1.0803057018585528e-07,
+      "loss": 0.5359,
+      "step": 16644
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7464479207992554,
+      "learning_rate": 1.0775842864096387e-07,
+      "loss": 0.8361,
+      "step": 16645
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7493220567703247,
+      "learning_rate": 1.0748662844969781e-07,
+      "loss": 0.8915,
+      "step": 16646
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.842637538909912,
+      "learning_rate": 1.0721516962143296e-07,
+      "loss": 0.9855,
+      "step": 16647
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.865507960319519,
+      "learning_rate": 1.069440521655396e-07,
+      "loss": 0.8383,
+      "step": 16648
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5383331775665283,
+      "learning_rate": 1.0667327609137024e-07,
+      "loss": 0.8725,
+      "step": 16649
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7621910572052002,
+      "learning_rate": 1.0640284140827183e-07,
+      "loss": 0.8831,
+      "step": 16650
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.097637891769409,
+      "learning_rate": 1.0613274812557361e-07,
+      "loss": 0.9048,
+      "step": 16651
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6516413688659668,
+      "learning_rate": 1.0586299625259699e-07,
+      "loss": 0.8574,
+      "step": 16652
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7174369096755981,
+      "learning_rate": 1.055935857986512e-07,
+      "loss": 0.8556,
+      "step": 16653
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.951137125492096,
+      "learning_rate": 1.0532451677303102e-07,
+      "loss": 0.5239,
+      "step": 16654
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8946439027786255,
+      "learning_rate": 1.0505578918502124e-07,
+      "loss": 0.7897,
+      "step": 16655
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.584452748298645,
+      "learning_rate": 1.0478740304389668e-07,
+      "loss": 0.7435,
+      "step": 16656
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8971904516220093,
+      "learning_rate": 1.045193583589188e-07,
+      "loss": 0.9921,
+      "step": 16657
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.718258261680603,
+      "learning_rate": 1.0425165513933355e-07,
+      "loss": 0.9481,
+      "step": 16658
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7395762205123901,
+      "learning_rate": 1.0398429339438353e-07,
+      "loss": 0.9145,
+      "step": 16659
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6841444969177246,
+      "learning_rate": 1.0371727313329027e-07,
+      "loss": 1.0403,
+      "step": 16660
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7818424701690674,
+      "learning_rate": 1.0345059436527082e-07,
+      "loss": 0.8786,
+      "step": 16661
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.848825216293335,
+      "learning_rate": 1.0318425709952562e-07,
+      "loss": 0.8428,
+      "step": 16662
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8447976112365723,
+      "learning_rate": 1.029182613452473e-07,
+      "loss": 0.8852,
+      "step": 16663
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6873865127563477,
+      "learning_rate": 1.0265260711161184e-07,
+      "loss": 0.975,
+      "step": 16664
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6499338150024414,
+      "learning_rate": 1.023872944077886e-07,
+      "loss": 0.9421,
+      "step": 16665
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0688316822052002,
+      "learning_rate": 1.0212232324293248e-07,
+      "loss": 0.5359,
+      "step": 16666
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.56336510181427,
+      "learning_rate": 1.0185769362618614e-07,
+      "loss": 0.9357,
+      "step": 16667
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8793442249298096,
+      "learning_rate": 1.0159340556668007e-07,
+      "loss": 0.8075,
+      "step": 16668
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7353289127349854,
+      "learning_rate": 1.0132945907353697e-07,
+      "loss": 0.9387,
+      "step": 16669
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.71719229221344,
+      "learning_rate": 1.010658541558629e-07,
+      "loss": 0.8756,
+      "step": 16670
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7425702810287476,
+      "learning_rate": 1.0080259082275501e-07,
+      "loss": 0.9315,
+      "step": 16671
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.883992314338684,
+      "learning_rate": 1.0053966908329716e-07,
+      "loss": 0.897,
+      "step": 16672
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.644033432006836,
+      "learning_rate": 1.0027708894656208e-07,
+      "loss": 0.8737,
+      "step": 16673
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.781711459159851,
+      "learning_rate": 1.0001485042161141e-07,
+      "loss": 0.843,
+      "step": 16674
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9693657159805298,
+      "learning_rate": 9.975295351749348e-08,
+      "loss": 0.4812,
+      "step": 16675
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.780993938446045,
+      "learning_rate": 9.949139824324661e-08,
+      "loss": 0.8793,
+      "step": 16676
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7134166955947876,
+      "learning_rate": 9.923018460789358e-08,
+      "loss": 0.8925,
+      "step": 16677
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8056004047393799,
+      "learning_rate": 9.896931262045162e-08,
+      "loss": 0.8949,
+      "step": 16678
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7053399085998535,
+      "learning_rate": 9.870878228992132e-08,
+      "loss": 0.8244,
+      "step": 16679
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9220213890075684,
+      "learning_rate": 9.844859362529324e-08,
+      "loss": 0.8981,
+      "step": 16680
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7998857498168945,
+      "learning_rate": 9.818874663554356e-08,
+      "loss": 0.8405,
+      "step": 16681
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8952929973602295,
+      "learning_rate": 9.792924132964287e-08,
+      "loss": 0.9372,
+      "step": 16682
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7541149854660034,
+      "learning_rate": 9.76700777165418e-08,
+      "loss": 0.8867,
+      "step": 16683
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.762635350227356,
+      "learning_rate": 9.741125580518651e-08,
+      "loss": 0.9163,
+      "step": 16684
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7036212682724,
+      "learning_rate": 9.715277560450653e-08,
+      "loss": 0.8809,
+      "step": 16685
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6715259552001953,
+      "learning_rate": 9.689463712342251e-08,
+      "loss": 0.8324,
+      "step": 16686
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7778007984161377,
+      "learning_rate": 9.663684037084064e-08,
+      "loss": 0.8009,
+      "step": 16687
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7729668617248535,
+      "learning_rate": 9.637938535565716e-08,
+      "loss": 0.847,
+      "step": 16688
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8433982133865356,
+      "learning_rate": 9.612227208675718e-08,
+      "loss": 0.8444,
+      "step": 16689
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8206764459609985,
+      "learning_rate": 9.586550057301247e-08,
+      "loss": 0.9588,
+      "step": 16690
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8578920364379883,
+      "learning_rate": 9.560907082328263e-08,
+      "loss": 0.9852,
+      "step": 16691
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6715120077133179,
+      "learning_rate": 9.535298284641725e-08,
+      "loss": 0.8814,
+      "step": 16692
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7341216802597046,
+      "learning_rate": 9.509723665125259e-08,
+      "loss": 0.9187,
+      "step": 16693
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7308123111724854,
+      "learning_rate": 9.484183224661381e-08,
+      "loss": 0.8959,
+      "step": 16694
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8514095544815063,
+      "learning_rate": 9.458676964131496e-08,
+      "loss": 0.962,
+      "step": 16695
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9360171556472778,
+      "learning_rate": 9.433204884415681e-08,
+      "loss": 0.8415,
+      "step": 16696
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7884798049926758,
+      "learning_rate": 9.407766986393007e-08,
+      "loss": 0.8342,
+      "step": 16697
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8000905513763428,
+      "learning_rate": 9.382363270941108e-08,
+      "loss": 0.8586,
+      "step": 16698
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.616525650024414,
+      "learning_rate": 9.356993738936615e-08,
+      "loss": 0.8999,
+      "step": 16699
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9182251691818237,
+      "learning_rate": 9.33165839125516e-08,
+      "loss": 0.8527,
+      "step": 16700
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6902246475219727,
+      "learning_rate": 9.306357228770713e-08,
+      "loss": 0.8372,
+      "step": 16701
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8984614014625549,
+      "learning_rate": 9.281090252356684e-08,
+      "loss": 0.4854,
+      "step": 16702
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1479718685150146,
+      "learning_rate": 9.255857462884599e-08,
+      "loss": 0.8815,
+      "step": 16703
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8117250204086304,
+      "learning_rate": 9.230658861225428e-08,
+      "loss": 0.8504,
+      "step": 16704
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.733026146888733,
+      "learning_rate": 9.205494448248476e-08,
+      "loss": 0.8416,
+      "step": 16705
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.902415156364441,
+      "learning_rate": 9.18036422482238e-08,
+      "loss": 0.9298,
+      "step": 16706
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6994664669036865,
+      "learning_rate": 9.155268191814114e-08,
+      "loss": 0.8525,
+      "step": 16707
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9498339295387268,
+      "learning_rate": 9.130206350089765e-08,
+      "loss": 0.4791,
+      "step": 16708
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7519595623016357,
+      "learning_rate": 9.105178700514084e-08,
+      "loss": 0.8754,
+      "step": 16709
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7514469623565674,
+      "learning_rate": 9.080185243950712e-08,
+      "loss": 0.9439,
+      "step": 16710
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9952343106269836,
+      "learning_rate": 9.055225981262184e-08,
+      "loss": 0.5427,
+      "step": 16711
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.599165916442871,
+      "learning_rate": 9.030300913309698e-08,
+      "loss": 0.8042,
+      "step": 16712
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5810967683792114,
+      "learning_rate": 9.005410040953344e-08,
+      "loss": 0.9077,
+      "step": 16713
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9196925163269043,
+      "learning_rate": 8.980553365051992e-08,
+      "loss": 0.8921,
+      "step": 16714
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9037365913391113,
+      "learning_rate": 8.95573088646362e-08,
+      "loss": 0.8476,
+      "step": 16715
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7106841802597046,
+      "learning_rate": 8.930942606044434e-08,
+      "loss": 0.9139,
+      "step": 16716
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.833024501800537,
+      "learning_rate": 8.906188524650083e-08,
+      "loss": 0.8717,
+      "step": 16717
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6605820655822754,
+      "learning_rate": 8.881468643134661e-08,
+      "loss": 0.9245,
+      "step": 16718
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6086612939834595,
+      "learning_rate": 8.856782962351152e-08,
+      "loss": 0.8964,
+      "step": 16719
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8210073709487915,
+      "learning_rate": 8.832131483151319e-08,
+      "loss": 0.8985,
+      "step": 16720
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0180044174194336,
+      "learning_rate": 8.807514206386037e-08,
+      "loss": 0.5604,
+      "step": 16721
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7961909770965576,
+      "learning_rate": 8.782931132904627e-08,
+      "loss": 1.0014,
+      "step": 16722
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.779679298400879,
+      "learning_rate": 8.758382263555299e-08,
+      "loss": 0.8758,
+      "step": 16723
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.999839425086975,
+      "learning_rate": 8.733867599185487e-08,
+      "loss": 0.889,
+      "step": 16724
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7902342081069946,
+      "learning_rate": 8.709387140640736e-08,
+      "loss": 0.9615,
+      "step": 16725
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9254785776138306,
+      "learning_rate": 8.684940888766036e-08,
+      "loss": 0.9201,
+      "step": 16726
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8096891641616821,
+      "learning_rate": 8.660528844404936e-08,
+      "loss": 0.913,
+      "step": 16727
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7044142484664917,
+      "learning_rate": 8.636151008399762e-08,
+      "loss": 0.9629,
+      "step": 16728
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5496708154678345,
+      "learning_rate": 8.61180738159173e-08,
+      "loss": 0.871,
+      "step": 16729
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0486814975738525,
+      "learning_rate": 8.587497964820946e-08,
+      "loss": 0.8495,
+      "step": 16730
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9881958961486816,
+      "learning_rate": 8.563222758926181e-08,
+      "loss": 0.8516,
+      "step": 16731
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.599428653717041,
+      "learning_rate": 8.538981764745102e-08,
+      "loss": 0.8877,
+      "step": 16732
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0189226865768433,
+      "learning_rate": 8.51477498311426e-08,
+      "loss": 0.5455,
+      "step": 16733
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7823212146759033,
+      "learning_rate": 8.490602414868876e-08,
+      "loss": 0.9574,
+      "step": 16734
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8173750638961792,
+      "learning_rate": 8.466464060843282e-08,
+      "loss": 0.8808,
+      "step": 16735
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5656057596206665,
+      "learning_rate": 8.442359921870148e-08,
+      "loss": 0.9062,
+      "step": 16736
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.764848232269287,
+      "learning_rate": 8.418289998781359e-08,
+      "loss": 0.8444,
+      "step": 16737
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.699743628501892,
+      "learning_rate": 8.394254292407589e-08,
+      "loss": 0.8085,
+      "step": 16738
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8418811559677124,
+      "learning_rate": 8.37025280357806e-08,
+      "loss": 0.9323,
+      "step": 16739
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.830573320388794,
+      "learning_rate": 8.346285533121224e-08,
+      "loss": 0.9605,
+      "step": 16740
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1512264013290405,
+      "learning_rate": 8.322352481863971e-08,
+      "loss": 0.5256,
+      "step": 16741
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.803523302078247,
+      "learning_rate": 8.298453650632088e-08,
+      "loss": 0.8608,
+      "step": 16742
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7614006996154785,
+      "learning_rate": 8.274589040250469e-08,
+      "loss": 0.8161,
+      "step": 16743
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7893513441085815,
+      "learning_rate": 8.250758651542456e-08,
+      "loss": 0.812,
+      "step": 16744
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.780419111251831,
+      "learning_rate": 8.226962485330392e-08,
+      "loss": 0.8996,
+      "step": 16745
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.713121771812439,
+      "learning_rate": 8.203200542435507e-08,
+      "loss": 0.9211,
+      "step": 16746
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6605677604675293,
+      "learning_rate": 8.179472823677703e-08,
+      "loss": 0.848,
+      "step": 16747
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.686124324798584,
+      "learning_rate": 8.155779329875768e-08,
+      "loss": 0.9004,
+      "step": 16748
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7454346418380737,
+      "learning_rate": 8.13212006184727e-08,
+      "loss": 0.8211,
+      "step": 16749
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7194617986679077,
+      "learning_rate": 8.108495020408558e-08,
+      "loss": 0.8278,
+      "step": 16750
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7110462188720703,
+      "learning_rate": 8.084904206375088e-08,
+      "loss": 0.9624,
+      "step": 16751
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7371922731399536,
+      "learning_rate": 8.061347620560656e-08,
+      "loss": 0.854,
+      "step": 16752
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9698975086212158,
+      "learning_rate": 8.037825263778276e-08,
+      "loss": 0.862,
+      "step": 16753
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8095905780792236,
+      "learning_rate": 8.014337136839633e-08,
+      "loss": 0.9402,
+      "step": 16754
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7126554250717163,
+      "learning_rate": 7.990883240555191e-08,
+      "loss": 0.9407,
+      "step": 16755
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7946914434432983,
+      "learning_rate": 7.967463575734413e-08,
+      "loss": 0.8918,
+      "step": 16756
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9343855381011963,
+      "learning_rate": 7.944078143185207e-08,
+      "loss": 0.8248,
+      "step": 16757
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.84098482131958,
+      "learning_rate": 7.920726943714707e-08,
+      "loss": 0.8504,
+      "step": 16758
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9783902168273926,
+      "learning_rate": 7.8974099781286e-08,
+      "loss": 0.9248,
+      "step": 16759
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.650917410850525,
+      "learning_rate": 7.874127247231688e-08,
+      "loss": 0.9578,
+      "step": 16760
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7563252449035645,
+      "learning_rate": 7.850878751827107e-08,
+      "loss": 0.8776,
+      "step": 16761
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.762316107749939,
+      "learning_rate": 7.827664492717323e-08,
+      "loss": 0.8346,
+      "step": 16762
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.75888991355896,
+      "learning_rate": 7.804484470703255e-08,
+      "loss": 0.8616,
+      "step": 16763
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7647573947906494,
+      "learning_rate": 7.781338686584928e-08,
+      "loss": 0.8701,
+      "step": 16764
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.64426589012146,
+      "learning_rate": 7.758227141160923e-08,
+      "loss": 0.8792,
+      "step": 16765
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8904253244400024,
+      "learning_rate": 7.735149835228717e-08,
+      "loss": 0.9248,
+      "step": 16766
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.821258544921875,
+      "learning_rate": 7.712106769584782e-08,
+      "loss": 0.9874,
+      "step": 16767
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7079551219940186,
+      "learning_rate": 7.689097945024149e-08,
+      "loss": 0.9205,
+      "step": 16768
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7915880680084229,
+      "learning_rate": 7.66612336234096e-08,
+      "loss": 0.8896,
+      "step": 16769
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6094939708709717,
+      "learning_rate": 7.643183022327694e-08,
+      "loss": 0.8724,
+      "step": 16770
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6901040077209473,
+      "learning_rate": 7.620276925776271e-08,
+      "loss": 0.8766,
+      "step": 16771
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8767811059951782,
+      "learning_rate": 7.597405073476949e-08,
+      "loss": 0.8464,
+      "step": 16772
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8378677368164062,
+      "learning_rate": 7.574567466219096e-08,
+      "loss": 0.935,
+      "step": 16773
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9892394542694092,
+      "learning_rate": 7.551764104790527e-08,
+      "loss": 0.5458,
+      "step": 16774
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7656549215316772,
+      "learning_rate": 7.528994989978389e-08,
+      "loss": 0.9198,
+      "step": 16775
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6966168880462646,
+      "learning_rate": 7.506260122568277e-08,
+      "loss": 0.8504,
+      "step": 16776
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6978414058685303,
+      "learning_rate": 7.483559503344673e-08,
+      "loss": 0.8553,
+      "step": 16777
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6987836360931396,
+      "learning_rate": 7.460893133090952e-08,
+      "loss": 0.8564,
+      "step": 16778
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.874830961227417,
+      "learning_rate": 7.438261012589265e-08,
+      "loss": 0.923,
+      "step": 16779
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7200143337249756,
+      "learning_rate": 7.415663142620655e-08,
+      "loss": 0.9686,
+      "step": 16780
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.067589282989502,
+      "learning_rate": 7.393099523964719e-08,
+      "loss": 0.5695,
+      "step": 16781
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8484282493591309,
+      "learning_rate": 7.370570157400281e-08,
+      "loss": 0.8738,
+      "step": 16782
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6711485385894775,
+      "learning_rate": 7.348075043704605e-08,
+      "loss": 0.8844,
+      "step": 16783
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7487764358520508,
+      "learning_rate": 7.325614183654072e-08,
+      "loss": 0.8899,
+      "step": 16784
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0184872150421143,
+      "learning_rate": 7.303187578023618e-08,
+      "loss": 0.5378,
+      "step": 16785
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9166746139526367,
+      "learning_rate": 7.280795227587179e-08,
+      "loss": 0.9465,
+      "step": 16786
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0317682027816772,
+      "learning_rate": 7.258437133117468e-08,
+      "loss": 0.5222,
+      "step": 16787
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6860440969467163,
+      "learning_rate": 7.236113295385983e-08,
+      "loss": 0.8689,
+      "step": 16788
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8362212181091309,
+      "learning_rate": 7.213823715162993e-08,
+      "loss": 0.8992,
+      "step": 16789
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9014166593551636,
+      "learning_rate": 7.191568393217774e-08,
+      "loss": 0.8837,
+      "step": 16790
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8883302211761475,
+      "learning_rate": 7.169347330318155e-08,
+      "loss": 0.914,
+      "step": 16791
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8568795919418335,
+      "learning_rate": 7.147160527231079e-08,
+      "loss": 0.8594,
+      "step": 16792
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8146347999572754,
+      "learning_rate": 7.125007984722043e-08,
+      "loss": 0.9158,
+      "step": 16793
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6529438495635986,
+      "learning_rate": 7.102889703555548e-08,
+      "loss": 0.9224,
+      "step": 16794
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5970295667648315,
+      "learning_rate": 7.080805684494652e-08,
+      "loss": 0.8878,
+      "step": 16795
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7198632955551147,
+      "learning_rate": 7.058755928301631e-08,
+      "loss": 0.8553,
+      "step": 16796
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7496482133865356,
+      "learning_rate": 7.036740435737321e-08,
+      "loss": 0.8263,
+      "step": 16797
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6216895580291748,
+      "learning_rate": 7.014759207561339e-08,
+      "loss": 0.9321,
+      "step": 16798
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.854547381401062,
+      "learning_rate": 6.992812244532188e-08,
+      "loss": 0.8848,
+      "step": 16799
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8922069072723389,
+      "learning_rate": 6.97089954740715e-08,
+      "loss": 0.9181,
+      "step": 16800
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6890883445739746,
+      "learning_rate": 6.949021116942622e-08,
+      "loss": 0.8787,
+      "step": 16801
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8593213558197021,
+      "learning_rate": 6.927176953893334e-08,
+      "loss": 0.9281,
+      "step": 16802
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7434062957763672,
+      "learning_rate": 6.905367059013013e-08,
+      "loss": 0.8783,
+      "step": 16803
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8309246301651,
+      "learning_rate": 6.883591433054615e-08,
+      "loss": 0.8093,
+      "step": 16804
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.945576548576355,
+      "learning_rate": 6.861850076769095e-08,
+      "loss": 0.511,
+      "step": 16805
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8680192232131958,
+      "learning_rate": 6.840142990907072e-08,
+      "loss": 0.8819,
+      "step": 16806
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9060392379760742,
+      "learning_rate": 6.818470176217284e-08,
+      "loss": 0.9007,
+      "step": 16807
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8418883085250854,
+      "learning_rate": 6.79683163344791e-08,
+      "loss": 0.9405,
+      "step": 16808
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6961679458618164,
+      "learning_rate": 6.775227363345349e-08,
+      "loss": 0.8422,
+      "step": 16809
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9760373830795288,
+      "learning_rate": 6.75365736665523e-08,
+      "loss": 0.9317,
+      "step": 16810
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9569554328918457,
+      "learning_rate": 6.732121644121958e-08,
+      "loss": 0.9044,
+      "step": 16811
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8706907033920288,
+      "learning_rate": 6.710620196488605e-08,
+      "loss": 0.9505,
+      "step": 16812
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7905306816101074,
+      "learning_rate": 6.689153024497019e-08,
+      "loss": 0.9122,
+      "step": 16813
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8795119524002075,
+      "learning_rate": 6.667720128888056e-08,
+      "loss": 0.8574,
+      "step": 16814
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8214315176010132,
+      "learning_rate": 6.646321510401344e-08,
+      "loss": 0.8992,
+      "step": 16815
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8094909191131592,
+      "learning_rate": 6.624957169775293e-08,
+      "loss": 0.9237,
+      "step": 16816
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7679553031921387,
+      "learning_rate": 6.603627107746979e-08,
+      "loss": 0.937,
+      "step": 16817
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8337820768356323,
+      "learning_rate": 6.582331325052704e-08,
+      "loss": 0.8813,
+      "step": 16818
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0413517951965332,
+      "learning_rate": 6.561069822427103e-08,
+      "loss": 0.5715,
+      "step": 16819
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8462769985198975,
+      "learning_rate": 6.539842600603918e-08,
+      "loss": 0.8639,
+      "step": 16820
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.80979585647583,
+      "learning_rate": 6.518649660315568e-08,
+      "loss": 0.9913,
+      "step": 16821
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.728592038154602,
+      "learning_rate": 6.497491002293576e-08,
+      "loss": 0.9206,
+      "step": 16822
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9578533172607422,
+      "learning_rate": 6.476366627267917e-08,
+      "loss": 0.934,
+      "step": 16823
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8968058824539185,
+      "learning_rate": 6.455276535967448e-08,
+      "loss": 0.897,
+      "step": 16824
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6480591297149658,
+      "learning_rate": 6.434220729120145e-08,
+      "loss": 0.8769,
+      "step": 16825
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.821168065071106,
+      "learning_rate": 6.413199207452314e-08,
+      "loss": 0.9271,
+      "step": 16826
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8033379316329956,
+      "learning_rate": 6.39221197168971e-08,
+      "loss": 0.8754,
+      "step": 16827
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7585662603378296,
+      "learning_rate": 6.371259022556198e-08,
+      "loss": 0.8326,
+      "step": 16828
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6405359506607056,
+      "learning_rate": 6.350340360774976e-08,
+      "loss": 0.8723,
+      "step": 16829
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6803693771362305,
+      "learning_rate": 6.329455987067912e-08,
+      "loss": 0.8518,
+      "step": 16830
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.963234543800354,
+      "learning_rate": 6.30860590215554e-08,
+      "loss": 0.4813,
+      "step": 16831
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7980083227157593,
+      "learning_rate": 6.287790106757396e-08,
+      "loss": 0.8532,
+      "step": 16832
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7277086973190308,
+      "learning_rate": 6.267008601591906e-08,
+      "loss": 0.942,
+      "step": 16833
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7424904108047485,
+      "learning_rate": 6.24626138737594e-08,
+      "loss": 0.8639,
+      "step": 16834
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.754137396812439,
+      "learning_rate": 6.225548464825592e-08,
+      "loss": 0.8617,
+      "step": 16835
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0378326177597046,
+      "learning_rate": 6.204869834655624e-08,
+      "loss": 0.6097,
+      "step": 16836
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7844176292419434,
+      "learning_rate": 6.184225497579577e-08,
+      "loss": 0.9808,
+      "step": 16837
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.645247220993042,
+      "learning_rate": 6.163615454309769e-08,
+      "loss": 0.9077,
+      "step": 16838
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7880001068115234,
+      "learning_rate": 6.143039705557297e-08,
+      "loss": 0.9124,
+      "step": 16839
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.861774206161499,
+      "learning_rate": 6.122498252032483e-08,
+      "loss": 0.9008,
+      "step": 16840
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9493675827980042,
+      "learning_rate": 6.10199109444376e-08,
+      "loss": 0.5769,
+      "step": 16841
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.164233446121216,
+      "learning_rate": 6.081518233499117e-08,
+      "loss": 0.9231,
+      "step": 16842
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9515260457992554,
+      "learning_rate": 6.061079669904879e-08,
+      "loss": 0.8373,
+      "step": 16843
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0363835096359253,
+      "learning_rate": 6.040675404366259e-08,
+      "loss": 0.5166,
+      "step": 16844
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9002394676208496,
+      "learning_rate": 6.02030543758747e-08,
+      "loss": 0.8562,
+      "step": 16845
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.804194450378418,
+      "learning_rate": 5.999969770271397e-08,
+      "loss": 0.8752,
+      "step": 16846
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.719359040260315,
+      "learning_rate": 5.979668403119699e-08,
+      "loss": 0.8818,
+      "step": 16847
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6977514028549194,
+      "learning_rate": 5.959401336833037e-08,
+      "loss": 0.9692,
+      "step": 16848
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8752810955047607,
+      "learning_rate": 5.9391685721106315e-08,
+      "loss": 0.8759,
+      "step": 16849
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6643203496932983,
+      "learning_rate": 5.918970109650701e-08,
+      "loss": 0.8666,
+      "step": 16850
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8345988988876343,
+      "learning_rate": 5.8988059501502434e-08,
+      "loss": 0.8879,
+      "step": 16851
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8764899969100952,
+      "learning_rate": 5.878676094305147e-08,
+      "loss": 0.8845,
+      "step": 16852
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.644471287727356,
+      "learning_rate": 5.858580542809966e-08,
+      "loss": 0.8837,
+      "step": 16853
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7819035053253174,
+      "learning_rate": 5.8385192963580364e-08,
+      "loss": 0.905,
+      "step": 16854
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6999619007110596,
+      "learning_rate": 5.818492355641803e-08,
+      "loss": 0.941,
+      "step": 16855
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8117866516113281,
+      "learning_rate": 5.79849972135238e-08,
+      "loss": 0.9184,
+      "step": 16856
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6418768167495728,
+      "learning_rate": 5.778541394179327e-08,
+      "loss": 0.8844,
+      "step": 16857
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6514885425567627,
+      "learning_rate": 5.7586173748117594e-08,
+      "loss": 0.8537,
+      "step": 16858
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8319910764694214,
+      "learning_rate": 5.738727663936905e-08,
+      "loss": 0.9233,
+      "step": 16859
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9716206789016724,
+      "learning_rate": 5.718872262241215e-08,
+      "loss": 0.4998,
+      "step": 16860
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.785876989364624,
+      "learning_rate": 5.6990511704098086e-08,
+      "loss": 0.8748,
+      "step": 16861
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.663503885269165,
+      "learning_rate": 5.6792643891266927e-08,
+      "loss": 0.9801,
+      "step": 16862
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0122292041778564,
+      "learning_rate": 5.659511919074656e-08,
+      "loss": 0.8861,
+      "step": 16863
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0279662609100342,
+      "learning_rate": 5.6397937609353745e-08,
+      "loss": 0.5392,
+      "step": 16864
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.028484582901001,
+      "learning_rate": 5.620109915388972e-08,
+      "loss": 0.9338,
+      "step": 16865
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.669344425201416,
+      "learning_rate": 5.600460383115014e-08,
+      "loss": 0.9576,
+      "step": 16866
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8382127285003662,
+      "learning_rate": 5.5808451647914044e-08,
+      "loss": 0.9117,
+      "step": 16867
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.753487229347229,
+      "learning_rate": 5.5612642610950454e-08,
+      "loss": 0.9321,
+      "step": 16868
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7272156476974487,
+      "learning_rate": 5.541717672701619e-08,
+      "loss": 0.8381,
+      "step": 16869
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0990896224975586,
+      "learning_rate": 5.522205400285585e-08,
+      "loss": 0.905,
+      "step": 16870
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7755687236785889,
+      "learning_rate": 5.502727444520295e-08,
+      "loss": 0.8522,
+      "step": 16871
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8048200607299805,
+      "learning_rate": 5.483283806077877e-08,
+      "loss": 0.9164,
+      "step": 16872
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7603212594985962,
+      "learning_rate": 5.463874485629239e-08,
+      "loss": 1.0092,
+      "step": 16873
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7477320432662964,
+      "learning_rate": 5.444499483844179e-08,
+      "loss": 0.9092,
+      "step": 16874
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.979417324066162,
+      "learning_rate": 5.425158801391162e-08,
+      "loss": 0.8517,
+      "step": 16875
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7972522974014282,
+      "learning_rate": 5.405852438937764e-08,
+      "loss": 0.9267,
+      "step": 16876
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9225883483886719,
+      "learning_rate": 5.386580397150121e-08,
+      "loss": 0.9067,
+      "step": 16877
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7164772748947144,
+      "learning_rate": 5.3673426766932546e-08,
+      "loss": 0.8054,
+      "step": 16878
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.678081750869751,
+      "learning_rate": 5.3481392782309684e-08,
+      "loss": 0.795,
+      "step": 16879
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.756480097770691,
+      "learning_rate": 5.328970202425954e-08,
+      "loss": 0.9606,
+      "step": 16880
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.753977656364441,
+      "learning_rate": 5.309835449939793e-08,
+      "loss": 0.8448,
+      "step": 16881
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8222942352294922,
+      "learning_rate": 5.2907350214325135e-08,
+      "loss": 0.8967,
+      "step": 16882
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.868146300315857,
+      "learning_rate": 5.271668917563366e-08,
+      "loss": 0.8656,
+      "step": 16883
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6813275814056396,
+      "learning_rate": 5.2526371389902685e-08,
+      "loss": 0.7733,
+      "step": 16884
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9261870384216309,
+      "learning_rate": 5.233639686370029e-08,
+      "loss": 0.8399,
+      "step": 16885
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9739835262298584,
+      "learning_rate": 5.214676560358123e-08,
+      "loss": 0.9294,
+      "step": 16886
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7306627035140991,
+      "learning_rate": 5.195747761608805e-08,
+      "loss": 0.8975,
+      "step": 16887
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8155744075775146,
+      "learning_rate": 5.176853290775441e-08,
+      "loss": 0.8849,
+      "step": 16888
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8863637447357178,
+      "learning_rate": 5.157993148509843e-08,
+      "loss": 0.8946,
+      "step": 16889
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2636477947235107,
+      "learning_rate": 5.1391673354630467e-08,
+      "loss": 0.9218,
+      "step": 16890
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2207958698272705,
+      "learning_rate": 5.1203758522844205e-08,
+      "loss": 0.8633,
+      "step": 16891
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8707972764968872,
+      "learning_rate": 5.101618699622668e-08,
+      "loss": 0.8653,
+      "step": 16892
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.85116708278656,
+      "learning_rate": 5.082895878124827e-08,
+      "loss": 0.9898,
+      "step": 16893
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6779738664627075,
+      "learning_rate": 5.064207388437159e-08,
+      "loss": 0.8801,
+      "step": 16894
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7194525003433228,
+      "learning_rate": 5.04555323120437e-08,
+      "loss": 0.9582,
+      "step": 16895
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7836042642593384,
+      "learning_rate": 5.026933407070167e-08,
+      "loss": 0.8723,
+      "step": 16896
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7618259191513062,
+      "learning_rate": 5.0083479166773696e-08,
+      "loss": 0.8302,
+      "step": 16897
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7172775268554688,
+      "learning_rate": 4.98979676066691e-08,
+      "loss": 0.9439,
+      "step": 16898
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9904892444610596,
+      "learning_rate": 4.971279939679163e-08,
+      "loss": 0.5308,
+      "step": 16899
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9795794486999512,
+      "learning_rate": 4.952797454353064e-08,
+      "loss": 0.8302,
+      "step": 16900
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7926768064498901,
+      "learning_rate": 4.934349305326325e-08,
+      "loss": 0.9091,
+      "step": 16901
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7847084999084473,
+      "learning_rate": 4.915935493235657e-08,
+      "loss": 0.9821,
+      "step": 16902
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.694834589958191,
+      "learning_rate": 4.897556018716443e-08,
+      "loss": 0.9604,
+      "step": 16903
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6887502670288086,
+      "learning_rate": 4.879210882402729e-08,
+      "loss": 0.9233,
+      "step": 16904
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.618674397468567,
+      "learning_rate": 4.8609000849277884e-08,
+      "loss": 0.8824,
+      "step": 16905
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8002557754516602,
+      "learning_rate": 4.8426236269233375e-08,
+      "loss": 0.9415,
+      "step": 16906
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7163259983062744,
+      "learning_rate": 4.824381509020093e-08,
+      "loss": 0.8355,
+      "step": 16907
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.139887571334839,
+      "learning_rate": 4.806173731847441e-08,
+      "loss": 0.8743,
+      "step": 16908
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.826765775680542,
+      "learning_rate": 4.788000296033879e-08,
+      "loss": 0.9019,
+      "step": 16909
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.677130937576294,
+      "learning_rate": 4.769861202206349e-08,
+      "loss": 0.9111,
+      "step": 16910
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.74350106716156,
+      "learning_rate": 4.751756450990908e-08,
+      "loss": 0.8496,
+      "step": 16911
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7927353382110596,
+      "learning_rate": 4.733686043012275e-08,
+      "loss": 0.9233,
+      "step": 16912
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.712623953819275,
+      "learning_rate": 4.715649978893844e-08,
+      "loss": 0.8325,
+      "step": 16913
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7005751132965088,
+      "learning_rate": 4.697648259258225e-08,
+      "loss": 0.8882,
+      "step": 16914
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.08018159866333,
+      "learning_rate": 4.6796808847264787e-08,
+      "loss": 0.9487,
+      "step": 16915
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9266852140426636,
+      "learning_rate": 4.6617478559186634e-08,
+      "loss": 0.9103,
+      "step": 16916
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7743932008743286,
+      "learning_rate": 4.6438491734535077e-08,
+      "loss": 0.9236,
+      "step": 16917
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.790971279144287,
+      "learning_rate": 4.6259848379486275e-08,
+      "loss": 0.9723,
+      "step": 16918
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8152374029159546,
+      "learning_rate": 4.608154850020641e-08,
+      "loss": 0.9049,
+      "step": 16919
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7499529123306274,
+      "learning_rate": 4.5903592102847224e-08,
+      "loss": 0.8602,
+      "step": 16920
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8109657764434814,
+      "learning_rate": 4.5725979193549376e-08,
+      "loss": 0.896,
+      "step": 16921
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8365999460220337,
+      "learning_rate": 4.554870977844128e-08,
+      "loss": 0.8922,
+      "step": 16922
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6344904899597168,
+      "learning_rate": 4.537178386364138e-08,
+      "loss": 0.9225,
+      "step": 16923
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8438054323196411,
+      "learning_rate": 4.519520145525369e-08,
+      "loss": 0.919,
+      "step": 16924
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.926115870475769,
+      "learning_rate": 4.5018962559372216e-08,
+      "loss": 0.9209,
+      "step": 16925
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9236541986465454,
+      "learning_rate": 4.484306718207765e-08,
+      "loss": 0.9418,
+      "step": 16926
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9134057760238647,
+      "learning_rate": 4.4667515329440694e-08,
+      "loss": 0.9301,
+      "step": 16927
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7136832475662231,
+      "learning_rate": 4.449230700751872e-08,
+      "loss": 0.9425,
+      "step": 16928
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6230510473251343,
+      "learning_rate": 4.4317442222356896e-08,
+      "loss": 0.8967,
+      "step": 16929
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7166478633880615,
+      "learning_rate": 4.41429209799904e-08,
+      "loss": 0.9116,
+      "step": 16930
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8351988792419434,
+      "learning_rate": 4.3968743286442186e-08,
+      "loss": 0.9177,
+      "step": 16931
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.733890175819397,
+      "learning_rate": 4.3794909147720776e-08,
+      "loss": 0.9268,
+      "step": 16932
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6140780448913574,
+      "learning_rate": 4.362141856982471e-08,
+      "loss": 0.9459,
+      "step": 16933
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.906572699546814,
+      "learning_rate": 4.344827155874254e-08,
+      "loss": 0.9225,
+      "step": 16934
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7609244585037231,
+      "learning_rate": 4.327546812044836e-08,
+      "loss": 0.8341,
+      "step": 16935
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8001834154129028,
+      "learning_rate": 4.3103008260904076e-08,
+      "loss": 0.9166,
+      "step": 16936
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7748472690582275,
+      "learning_rate": 4.29308919860616e-08,
+      "loss": 0.9032,
+      "step": 16937
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9414796829223633,
+      "learning_rate": 4.2759119301859495e-08,
+      "loss": 0.938,
+      "step": 16938
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7094727754592896,
+      "learning_rate": 4.2587690214226376e-08,
+      "loss": 0.8305,
+      "step": 16939
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.812376618385315,
+      "learning_rate": 4.2416604729077496e-08,
+      "loss": 0.8929,
+      "step": 16940
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.905362606048584,
+      "learning_rate": 4.2245862852315914e-08,
+      "loss": 0.9097,
+      "step": 16941
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.858384370803833,
+      "learning_rate": 4.2075464589833583e-08,
+      "loss": 0.9154,
+      "step": 16942
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7335377931594849,
+      "learning_rate": 4.1905409947510244e-08,
+      "loss": 0.8718,
+      "step": 16943
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9073083400726318,
+      "learning_rate": 4.1735698931215655e-08,
+      "loss": 0.9305,
+      "step": 16944
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.79978609085083,
+      "learning_rate": 4.1566331546804004e-08,
+      "loss": 0.9137,
+      "step": 16945
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5958067178726196,
+      "learning_rate": 4.1397307800120635e-08,
+      "loss": 0.9178,
+      "step": 16946
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.730715036392212,
+      "learning_rate": 4.122862769699753e-08,
+      "loss": 0.8727,
+      "step": 16947
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8058571815490723,
+      "learning_rate": 4.1060291243255615e-08,
+      "loss": 0.8657,
+      "step": 16948
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7328755855560303,
+      "learning_rate": 4.089229844470466e-08,
+      "loss": 0.9298,
+      "step": 16949
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7758475542068481,
+      "learning_rate": 4.072464930714004e-08,
+      "loss": 0.9433,
+      "step": 16950
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.676727294921875,
+      "learning_rate": 4.055734383634824e-08,
+      "loss": 0.9177,
+      "step": 16951
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8348947763442993,
+      "learning_rate": 4.0390382038102416e-08,
+      "loss": 0.8619,
+      "step": 16952
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7150864601135254,
+      "learning_rate": 4.0223763918162404e-08,
+      "loss": 0.8381,
+      "step": 16953
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.750909447669983,
+      "learning_rate": 4.0057489482279166e-08,
+      "loss": 0.849,
+      "step": 16954
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.809551477432251,
+      "learning_rate": 3.989155873619033e-08,
+      "loss": 0.8733,
+      "step": 16955
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.783695936203003,
+      "learning_rate": 3.972597168562131e-08,
+      "loss": 0.8814,
+      "step": 16956
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7003097534179688,
+      "learning_rate": 3.956072833628533e-08,
+      "loss": 0.8092,
+      "step": 16957
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7009940147399902,
+      "learning_rate": 3.939582869388559e-08,
+      "loss": 0.887,
+      "step": 16958
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8587396144866943,
+      "learning_rate": 3.923127276411309e-08,
+      "loss": 0.8572,
+      "step": 16959
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7167260646820068,
+      "learning_rate": 3.906706055264331e-08,
+      "loss": 0.9159,
+      "step": 16960
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8860445022583008,
+      "learning_rate": 3.890319206514615e-08,
+      "loss": 0.9051,
+      "step": 16961
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8863625526428223,
+      "learning_rate": 3.873966730727374e-08,
+      "loss": 0.9757,
+      "step": 16962
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8373031616210938,
+      "learning_rate": 3.857648628466937e-08,
+      "loss": 0.8576,
+      "step": 16963
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6365312337875366,
+      "learning_rate": 3.8413649002965184e-08,
+      "loss": 0.9164,
+      "step": 16964
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8804470300674438,
+      "learning_rate": 3.825115546777891e-08,
+      "loss": 0.8722,
+      "step": 16965
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8623290061950684,
+      "learning_rate": 3.808900568471941e-08,
+      "loss": 0.8616,
+      "step": 16966
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.488790512084961,
+      "learning_rate": 3.792719965937885e-08,
+      "loss": 0.9075,
+      "step": 16967
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8464670181274414,
+      "learning_rate": 3.7765737397343904e-08,
+      "loss": 0.8954,
+      "step": 16968
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6899328231811523,
+      "learning_rate": 3.7604618904184545e-08,
+      "loss": 0.9634,
+      "step": 16969
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7534446716308594,
+      "learning_rate": 3.744384418546188e-08,
+      "loss": 0.8734,
+      "step": 16970
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8187503814697266,
+      "learning_rate": 3.728341324672147e-08,
+      "loss": 0.9168,
+      "step": 16971
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.840823769569397,
+      "learning_rate": 3.712332609350222e-08,
+      "loss": 0.9104,
+      "step": 16972
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7387958765029907,
+      "learning_rate": 3.6963582731326386e-08,
+      "loss": 0.9156,
+      "step": 16973
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6884331703186035,
+      "learning_rate": 3.680418316570622e-08,
+      "loss": 0.8942,
+      "step": 16974
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.890742301940918,
+      "learning_rate": 3.664512740214288e-08,
+      "loss": 0.8811,
+      "step": 16975
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8318909406661987,
+      "learning_rate": 3.648641544612419e-08,
+      "loss": 0.8581,
+      "step": 16976
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6680599451065063,
+      "learning_rate": 3.6328047303128e-08,
+      "loss": 0.8599,
+      "step": 16977
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.918280839920044,
+      "learning_rate": 3.6170022978617715e-08,
+      "loss": 0.9324,
+      "step": 16978
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7631083726882935,
+      "learning_rate": 3.6012342478047854e-08,
+      "loss": 0.9336,
+      "step": 16979
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8177746534347534,
+      "learning_rate": 3.585500580685852e-08,
+      "loss": 0.9714,
+      "step": 16980
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7385541200637817,
+      "learning_rate": 3.569801297047981e-08,
+      "loss": 0.8697,
+      "step": 16981
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7808159589767456,
+      "learning_rate": 3.5541363974327395e-08,
+      "loss": 0.8847,
+      "step": 16982
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7447491884231567,
+      "learning_rate": 3.538505882380916e-08,
+      "loss": 0.9111,
+      "step": 16983
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.592615008354187,
+      "learning_rate": 3.522909752431636e-08,
+      "loss": 0.9204,
+      "step": 16984
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9057852029800415,
+      "learning_rate": 3.507348008123246e-08,
+      "loss": 0.8696,
+      "step": 16985
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6352488994598389,
+      "learning_rate": 3.49182064999265e-08,
+      "loss": 0.9257,
+      "step": 16986
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6955666542053223,
+      "learning_rate": 3.4763276785757525e-08,
+      "loss": 0.8545,
+      "step": 16987
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7541472911834717,
+      "learning_rate": 3.460869094407127e-08,
+      "loss": 0.8645,
+      "step": 16988
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.832167387008667,
+      "learning_rate": 3.4454448980201226e-08,
+      "loss": 0.8578,
+      "step": 16989
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7183431386947632,
+      "learning_rate": 3.430055089947093e-08,
+      "loss": 0.8699,
+      "step": 16990
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7163522243499756,
+      "learning_rate": 3.414699670718946e-08,
+      "loss": 0.9084,
+      "step": 16991
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7016727924346924,
+      "learning_rate": 3.399378640865814e-08,
+      "loss": 0.9657,
+      "step": 16992
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8987613916397095,
+      "learning_rate": 3.3840920009161614e-08,
+      "loss": 0.8524,
+      "step": 16993
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7800244092941284,
+      "learning_rate": 3.3688397513975676e-08,
+      "loss": 0.8475,
+      "step": 16994
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.732125997543335,
+      "learning_rate": 3.353621892836389e-08,
+      "loss": 0.897,
+      "step": 16995
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7798832654953003,
+      "learning_rate": 3.3384384257576505e-08,
+      "loss": 0.9155,
+      "step": 16996
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6733369827270508,
+      "learning_rate": 3.3232893506852657e-08,
+      "loss": 0.9714,
+      "step": 16997
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6715399026870728,
+      "learning_rate": 3.30817466814215e-08,
+      "loss": 0.9048,
+      "step": 16998
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7094799280166626,
+      "learning_rate": 3.293094378649775e-08,
+      "loss": 0.8183,
+      "step": 16999
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.740619421005249,
+      "learning_rate": 3.278048482728502e-08,
+      "loss": 0.8976,
+      "step": 17000
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6426995992660522,
+      "learning_rate": 3.2630369808975827e-08,
+      "loss": 0.8611,
+      "step": 17001
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9045161008834839,
+      "learning_rate": 3.248059873675047e-08,
+      "loss": 0.8481,
+      "step": 17002
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8640708923339844,
+      "learning_rate": 3.233117161577481e-08,
+      "loss": 0.8529,
+      "step": 17003
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7647262811660767,
+      "learning_rate": 3.218208845120807e-08,
+      "loss": 0.875,
+      "step": 17004
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8356826305389404,
+      "learning_rate": 3.2033349248193903e-08,
+      "loss": 0.9043,
+      "step": 17005
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8375816345214844,
+      "learning_rate": 3.1884954011862646e-08,
+      "loss": 0.9352,
+      "step": 17006
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0403692722320557,
+      "learning_rate": 3.173690274733798e-08,
+      "loss": 0.8908,
+      "step": 17007
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.753091812133789,
+      "learning_rate": 3.158919545972694e-08,
+      "loss": 0.889,
+      "step": 17008
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7871108055114746,
+      "learning_rate": 3.144183215412877e-08,
+      "loss": 0.8781,
+      "step": 17009
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.767849087715149,
+      "learning_rate": 3.1294812835624965e-08,
+      "loss": 0.9194,
+      "step": 17010
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8572269678115845,
+      "learning_rate": 3.114813750929258e-08,
+      "loss": 0.867,
+      "step": 17011
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6849069595336914,
+      "learning_rate": 3.1001806180189775e-08,
+      "loss": 0.9161,
+      "step": 17012
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7427431344985962,
+      "learning_rate": 3.085581885336808e-08,
+      "loss": 0.8043,
+      "step": 17013
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.810930848121643,
+      "learning_rate": 3.071017553386346e-08,
+      "loss": 0.8816,
+      "step": 17014
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6672090291976929,
+      "learning_rate": 3.056487622670301e-08,
+      "loss": 0.8447,
+      "step": 17015
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0917577743530273,
+      "learning_rate": 3.0419920936900494e-08,
+      "loss": 0.5408,
+      "step": 17016
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8725041151046753,
+      "learning_rate": 3.027530966945747e-08,
+      "loss": 0.8854,
+      "step": 17017
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9404852390289307,
+      "learning_rate": 3.0131042429364376e-08,
+      "loss": 0.9232,
+      "step": 17018
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9725698828697205,
+      "learning_rate": 2.998711922159836e-08,
+      "loss": 0.4745,
+      "step": 17019
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7312403917312622,
+      "learning_rate": 2.984354005112766e-08,
+      "loss": 0.7786,
+      "step": 17020
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7155433893203735,
+      "learning_rate": 2.9700304922906098e-08,
+      "loss": 0.9062,
+      "step": 17021
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7692140340805054,
+      "learning_rate": 2.955741384187527e-08,
+      "loss": 0.8677,
+      "step": 17022
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8503379821777344,
+      "learning_rate": 2.9414866812967902e-08,
+      "loss": 0.916,
+      "step": 17023
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9333522319793701,
+      "learning_rate": 2.9272663841101168e-08,
+      "loss": 0.9184,
+      "step": 17024
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.113245964050293,
+      "learning_rate": 2.9130804931182254e-08,
+      "loss": 0.573,
+      "step": 17025
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6996816396713257,
+      "learning_rate": 2.8989290088107248e-08,
+      "loss": 0.8168,
+      "step": 17026
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7576831579208374,
+      "learning_rate": 2.8848119316758905e-08,
+      "loss": 0.9034,
+      "step": 17027
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.783080816268921,
+      "learning_rate": 2.8707292622008888e-08,
+      "loss": 0.8541,
+      "step": 17028
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6134307384490967,
+      "learning_rate": 2.856681000871664e-08,
+      "loss": 0.8485,
+      "step": 17029
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7848316431045532,
+      "learning_rate": 2.8426671481728285e-08,
+      "loss": 0.9054,
+      "step": 17030
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7033088207244873,
+      "learning_rate": 2.8286877045882178e-08,
+      "loss": 0.8713,
+      "step": 17031
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7085144519805908,
+      "learning_rate": 2.8147426706001125e-08,
+      "loss": 0.8415,
+      "step": 17032
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7783198356628418,
+      "learning_rate": 2.8008320466896834e-08,
+      "loss": 0.8603,
+      "step": 17033
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7809889316558838,
+      "learning_rate": 2.7869558333369906e-08,
+      "loss": 0.8824,
+      "step": 17034
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7643349170684814,
+      "learning_rate": 2.7731140310208738e-08,
+      "loss": 0.9208,
+      "step": 17035
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8451462984085083,
+      "learning_rate": 2.7593066402189506e-08,
+      "loss": 0.9005,
+      "step": 17036
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7428935766220093,
+      "learning_rate": 2.745533661407729e-08,
+      "loss": 0.9347,
+      "step": 17037
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8069937229156494,
+      "learning_rate": 2.7317950950622728e-08,
+      "loss": 0.8862,
+      "step": 17038
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8138737678527832,
+      "learning_rate": 2.71809094165687e-08,
+      "loss": 0.92,
+      "step": 17039
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8255681991577148,
+      "learning_rate": 2.7044212016643643e-08,
+      "loss": 0.7977,
+      "step": 17040
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8366420269012451,
+      "learning_rate": 2.6907858755564896e-08,
+      "loss": 0.8944,
+      "step": 17041
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9001367092132568,
+      "learning_rate": 2.6771849638036473e-08,
+      "loss": 0.8731,
+      "step": 17042
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8538137674331665,
+      "learning_rate": 2.6636184668753506e-08,
+      "loss": 0.8772,
+      "step": 17043
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7869664430618286,
+      "learning_rate": 2.6500863852395585e-08,
+      "loss": 0.8862,
+      "step": 17044
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6638556718826294,
+      "learning_rate": 2.6365887193633422e-08,
+      "loss": 0.8804,
+      "step": 17045
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8795177936553955,
+      "learning_rate": 2.6231254697123286e-08,
+      "loss": 0.8973,
+      "step": 17046
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7164945602416992,
+      "learning_rate": 2.609696636751258e-08,
+      "loss": 0.8814,
+      "step": 17047
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7197479009628296,
+      "learning_rate": 2.596302220943536e-08,
+      "loss": 0.9137,
+      "step": 17048
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7311943769454956,
+      "learning_rate": 2.582942222751239e-08,
+      "loss": 0.897,
+      "step": 17049
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7729771137237549,
+      "learning_rate": 2.569616642635331e-08,
+      "loss": 0.886,
+      "step": 17050
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8523064851760864,
+      "learning_rate": 2.5563254810558878e-08,
+      "loss": 0.96,
+      "step": 17051
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9374133348464966,
+      "learning_rate": 2.5430687384713217e-08,
+      "loss": 0.862,
+      "step": 17052
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.754805326461792,
+      "learning_rate": 2.5298464153391544e-08,
+      "loss": 0.9155,
+      "step": 17053
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8759046792984009,
+      "learning_rate": 2.5166585121156883e-08,
+      "loss": 0.9156,
+      "step": 17054
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7295581102371216,
+      "learning_rate": 2.5035050292560036e-08,
+      "loss": 0.8677,
+      "step": 17055
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7976747751235962,
+      "learning_rate": 2.4903859672139597e-08,
+      "loss": 0.979,
+      "step": 17056
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8166133165359497,
+      "learning_rate": 2.477301326442305e-08,
+      "loss": 0.8452,
+      "step": 17057
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8200608491897583,
+      "learning_rate": 2.464251107392457e-08,
+      "loss": 0.8777,
+      "step": 17058
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.654309868812561,
+      "learning_rate": 2.4512353105148322e-08,
+      "loss": 0.7942,
+      "step": 17059
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7202975749969482,
+      "learning_rate": 2.4382539362585168e-08,
+      "loss": 0.9671,
+      "step": 17060
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6502807140350342,
+      "learning_rate": 2.4253069850714848e-08,
+      "loss": 0.8972,
+      "step": 17061
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7878464460372925,
+      "learning_rate": 2.412394457400491e-08,
+      "loss": 0.9122,
+      "step": 17062
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6978769302368164,
+      "learning_rate": 2.3995163536911782e-08,
+      "loss": 0.7872,
+      "step": 17063
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6716961860656738,
+      "learning_rate": 2.3866726743877467e-08,
+      "loss": 0.8586,
+      "step": 17064
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0738970041275024,
+      "learning_rate": 2.37386341993362e-08,
+      "loss": 0.5886,
+      "step": 17065
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7975754737854004,
+      "learning_rate": 2.361088590770666e-08,
+      "loss": 0.7864,
+      "step": 17066
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8742730617523193,
+      "learning_rate": 2.348348187339755e-08,
+      "loss": 0.8731,
+      "step": 17067
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6703159809112549,
+      "learning_rate": 2.3356422100805354e-08,
+      "loss": 0.9394,
+      "step": 17068
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0650947093963623,
+      "learning_rate": 2.3229706594314337e-08,
+      "loss": 0.9534,
+      "step": 17069
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.850548267364502,
+      "learning_rate": 2.3103335358297672e-08,
+      "loss": 0.9163,
+      "step": 17070
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9106305837631226,
+      "learning_rate": 2.2977308397115205e-08,
+      "loss": 0.9422,
+      "step": 17071
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8150694370269775,
+      "learning_rate": 2.2851625715115677e-08,
+      "loss": 0.854,
+      "step": 17072
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1853013038635254,
+      "learning_rate": 2.272628731663673e-08,
+      "loss": 0.8768,
+      "step": 17073
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7792284488677979,
+      "learning_rate": 2.2601293206003795e-08,
+      "loss": 0.9033,
+      "step": 17074
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5997140407562256,
+      "learning_rate": 2.2476643387528974e-08,
+      "loss": 0.948,
+      "step": 17075
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8332079648971558,
+      "learning_rate": 2.2352337865514384e-08,
+      "loss": 0.9094,
+      "step": 17076
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7646775245666504,
+      "learning_rate": 2.2228376644248816e-08,
+      "loss": 0.8828,
+      "step": 17077
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.842896580696106,
+      "learning_rate": 2.210475972800996e-08,
+      "loss": 0.859,
+      "step": 17078
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8573428392410278,
+      "learning_rate": 2.1981487121064404e-08,
+      "loss": 0.8593,
+      "step": 17079
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9214030504226685,
+      "learning_rate": 2.18585588276643e-08,
+      "loss": 0.8884,
+      "step": 17080
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.898058533668518,
+      "learning_rate": 2.173597485205403e-08,
+      "loss": 0.9032,
+      "step": 17081
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7773573398590088,
+      "learning_rate": 2.1613735198460216e-08,
+      "loss": 0.8774,
+      "step": 17082
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8364392518997192,
+      "learning_rate": 2.1491839871105034e-08,
+      "loss": 0.8957,
+      "step": 17083
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8194957971572876,
+      "learning_rate": 2.137028887419068e-08,
+      "loss": 0.9094,
+      "step": 17084
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.69001042842865,
+      "learning_rate": 2.1249082211914905e-08,
+      "loss": 0.9411,
+      "step": 17085
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8862391710281372,
+      "learning_rate": 2.1128219888457702e-08,
+      "loss": 0.8236,
+      "step": 17086
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.803242564201355,
+      "learning_rate": 2.1007701907991284e-08,
+      "loss": 0.8839,
+      "step": 17087
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7918617725372314,
+      "learning_rate": 2.088752827467455e-08,
+      "loss": 0.8937,
+      "step": 17088
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8471431732177734,
+      "learning_rate": 2.0767698992653072e-08,
+      "loss": 0.8761,
+      "step": 17089
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.606785774230957,
+      "learning_rate": 2.064821406606243e-08,
+      "loss": 0.8914,
+      "step": 17090
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7698849439620972,
+      "learning_rate": 2.0529073499024886e-08,
+      "loss": 0.9565,
+      "step": 17091
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8024587631225586,
+      "learning_rate": 2.0410277295653814e-08,
+      "loss": 0.9026,
+      "step": 17092
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7660603523254395,
+      "learning_rate": 2.0291825460047044e-08,
+      "loss": 0.8885,
+      "step": 17093
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6082065105438232,
+      "learning_rate": 2.0173717996291308e-08,
+      "loss": 0.9053,
+      "step": 17094
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.676315426826477,
+      "learning_rate": 2.005595490846335e-08,
+      "loss": 0.9445,
+      "step": 17095
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7303417921066284,
+      "learning_rate": 1.993853620062769e-08,
+      "loss": 0.8629,
+      "step": 17096
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5428791046142578,
+      "learning_rate": 1.9821461876833314e-08,
+      "loss": 0.9133,
+      "step": 17097
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6239371299743652,
+      "learning_rate": 1.9704731941122545e-08,
+      "loss": 0.7725,
+      "step": 17098
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7116451263427734,
+      "learning_rate": 1.9588346397523273e-08,
+      "loss": 0.9502,
+      "step": 17099
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0114779472351074,
+      "learning_rate": 1.947230525005006e-08,
+      "loss": 0.5726,
+      "step": 17100
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.896443486213684,
+      "learning_rate": 1.9356608502708595e-08,
+      "loss": 0.8814,
+      "step": 17101
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.716213583946228,
+      "learning_rate": 1.924125615949013e-08,
+      "loss": 0.8606,
+      "step": 17102
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8351266384124756,
+      "learning_rate": 1.912624822437592e-08,
+      "loss": 0.9155,
+      "step": 17103
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8435132503509521,
+      "learning_rate": 1.9011584701335017e-08,
+      "loss": 0.9169,
+      "step": 17104
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8010510206222534,
+      "learning_rate": 1.8897265594323145e-08,
+      "loss": 0.9259,
+      "step": 17105
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8487257957458496,
+      "learning_rate": 1.8783290907286034e-08,
+      "loss": 0.8832,
+      "step": 17106
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7463833093643188,
+      "learning_rate": 1.8669660644156094e-08,
+      "loss": 0.9773,
+      "step": 17107
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8983663320541382,
+      "learning_rate": 1.8556374808853527e-08,
+      "loss": 0.8695,
+      "step": 17108
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.585308313369751,
+      "learning_rate": 1.8443433405289646e-08,
+      "loss": 0.8921,
+      "step": 17109
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7126920223236084,
+      "learning_rate": 1.8330836437360222e-08,
+      "loss": 0.8569,
+      "step": 17110
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.848032832145691,
+      "learning_rate": 1.8218583908949926e-08,
+      "loss": 0.8996,
+      "step": 17111
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6332186460494995,
+      "learning_rate": 1.810667582393455e-08,
+      "loss": 0.8163,
+      "step": 17112
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7423447370529175,
+      "learning_rate": 1.7995112186174334e-08,
+      "loss": 0.8045,
+      "step": 17113
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0589439868927,
+      "learning_rate": 1.7883892999518428e-08,
+      "loss": 0.8867,
+      "step": 17114
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6956969499588013,
+      "learning_rate": 1.7773018267805976e-08,
+      "loss": 0.9109,
+      "step": 17115
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7583951950073242,
+      "learning_rate": 1.7662487994862808e-08,
+      "loss": 0.8571,
+      "step": 17116
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.729231357574463,
+      "learning_rate": 1.7552302184502544e-08,
+      "loss": 0.7545,
+      "step": 17117
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.80427885055542,
+      "learning_rate": 1.744246084052659e-08,
+      "loss": 0.9136,
+      "step": 17118
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8782159090042114,
+      "learning_rate": 1.7332963966726348e-08,
+      "loss": 0.8671,
+      "step": 17119
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.652601957321167,
+      "learning_rate": 1.722381156688102e-08,
+      "loss": 0.9287,
+      "step": 17120
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7854180335998535,
+      "learning_rate": 1.711500364475538e-08,
+      "loss": 0.9002,
+      "step": 17121
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.712899923324585,
+      "learning_rate": 1.7006540204105304e-08,
+      "loss": 0.9656,
+      "step": 17122
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.819166898727417,
+      "learning_rate": 1.6898421248673357e-08,
+      "loss": 0.8584,
+      "step": 17123
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9423781633377075,
+      "learning_rate": 1.679064678218989e-08,
+      "loss": 0.9003,
+      "step": 17124
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9733960628509521,
+      "learning_rate": 1.668321680837415e-08,
+      "loss": 0.8727,
+      "step": 17125
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8883227109909058,
+      "learning_rate": 1.6576131330933166e-08,
+      "loss": 0.957,
+      "step": 17126
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7456860542297363,
+      "learning_rate": 1.6469390353562877e-08,
+      "loss": 0.8811,
+      "step": 17127
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9007147550582886,
+      "learning_rate": 1.6362993879945888e-08,
+      "loss": 0.9507,
+      "step": 17128
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.090998649597168,
+      "learning_rate": 1.6256941913753708e-08,
+      "loss": 0.9503,
+      "step": 17129
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1105899810791016,
+      "learning_rate": 1.615123445864564e-08,
+      "loss": 0.5609,
+      "step": 17130
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.83977210521698,
+      "learning_rate": 1.6045871518269863e-08,
+      "loss": 0.8709,
+      "step": 17131
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.682094931602478,
+      "learning_rate": 1.5940853096262366e-08,
+      "loss": 0.9542,
+      "step": 17132
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.648603916168213,
+      "learning_rate": 1.5836179196246914e-08,
+      "loss": 0.8645,
+      "step": 17133
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9157813787460327,
+      "learning_rate": 1.5731849821833955e-08,
+      "loss": 0.8548,
+      "step": 17134
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7474228143692017,
+      "learning_rate": 1.562786497662616e-08,
+      "loss": 0.8565,
+      "step": 17135
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7426587343215942,
+      "learning_rate": 1.5524224664210662e-08,
+      "loss": 0.9368,
+      "step": 17136
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7092564105987549,
+      "learning_rate": 1.5420928888163487e-08,
+      "loss": 0.8422,
+      "step": 17137
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0438761711120605,
+      "learning_rate": 1.5317977652048455e-08,
+      "loss": 0.8994,
+      "step": 17138
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7591665983200073,
+      "learning_rate": 1.5215370959419383e-08,
+      "loss": 0.9301,
+      "step": 17139
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7031729221343994,
+      "learning_rate": 1.5113108813816778e-08,
+      "loss": 0.8402,
+      "step": 17140
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6974835395812988,
+      "learning_rate": 1.501119121876893e-08,
+      "loss": 0.8846,
+      "step": 17141
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.09260892868042,
+      "learning_rate": 1.4909618177793016e-08,
+      "loss": 0.9626,
+      "step": 17142
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9540131092071533,
+      "learning_rate": 1.4808389694394021e-08,
+      "loss": 0.8754,
+      "step": 17143
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8488633632659912,
+      "learning_rate": 1.4707505772064701e-08,
+      "loss": 0.9224,
+      "step": 17144
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8815505504608154,
+      "learning_rate": 1.4606966414286717e-08,
+      "loss": 0.9979,
+      "step": 17145
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.037135124206543,
+      "learning_rate": 1.4506771624529514e-08,
+      "loss": 0.9023,
+      "step": 17146
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.694784164428711,
+      "learning_rate": 1.440692140624922e-08,
+      "loss": 0.8502,
+      "step": 17147
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7239001989364624,
+      "learning_rate": 1.4307415762893074e-08,
+      "loss": 0.8968,
+      "step": 17148
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7371407747268677,
+      "learning_rate": 1.4208254697894996e-08,
+      "loss": 0.8871,
+      "step": 17149
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7894316911697388,
+      "learning_rate": 1.4109438214674475e-08,
+      "loss": 0.8576,
+      "step": 17150
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.828436255455017,
+      "learning_rate": 1.4010966316643226e-08,
+      "loss": 0.8415,
+      "step": 17151
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6152567863464355,
+      "learning_rate": 1.3912839007199641e-08,
+      "loss": 0.8744,
+      "step": 17152
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.729296326637268,
+      "learning_rate": 1.381505628972879e-08,
+      "loss": 0.875,
+      "step": 17153
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7571072578430176,
+      "learning_rate": 1.371761816760464e-08,
+      "loss": 0.8943,
+      "step": 17154
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.803439974784851,
+      "learning_rate": 1.3620524644188949e-08,
+      "loss": 0.9129,
+      "step": 17155
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8950973749160767,
+      "learning_rate": 1.3523775722834586e-08,
+      "loss": 0.9404,
+      "step": 17156
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7568846940994263,
+      "learning_rate": 1.3427371406877776e-08,
+      "loss": 0.8778,
+      "step": 17157
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6658481359481812,
+      "learning_rate": 1.3331311699646965e-08,
+      "loss": 0.8916,
+      "step": 17158
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9940531253814697,
+      "learning_rate": 1.3235596604455058e-08,
+      "loss": 0.9097,
+      "step": 17159
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8355598449707031,
+      "learning_rate": 1.3140226124606082e-08,
+      "loss": 0.8967,
+      "step": 17160
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6796530485153198,
+      "learning_rate": 1.3045200263390734e-08,
+      "loss": 0.7783,
+      "step": 17161
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6376535892486572,
+      "learning_rate": 1.2950519024088615e-08,
+      "loss": 0.8064,
+      "step": 17162
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9003660678863525,
+      "learning_rate": 1.2856182409966e-08,
+      "loss": 0.8903,
+      "step": 17163
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8660941123962402,
+      "learning_rate": 1.2762190424278065e-08,
+      "loss": 0.9566,
+      "step": 17164
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6976171731948853,
+      "learning_rate": 1.266854307026888e-08,
+      "loss": 0.86,
+      "step": 17165
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8287798166275024,
+      "learning_rate": 1.2575240351170303e-08,
+      "loss": 0.9357,
+      "step": 17166
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7769935131072998,
+      "learning_rate": 1.2482282270200874e-08,
+      "loss": 0.9407,
+      "step": 17167
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1053519248962402,
+      "learning_rate": 1.2389668830569135e-08,
+      "loss": 0.8631,
+      "step": 17168
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8490961790084839,
+      "learning_rate": 1.2297400035471418e-08,
+      "loss": 0.8591,
+      "step": 17169
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7782566547393799,
+      "learning_rate": 1.2205475888089623e-08,
+      "loss": 0.8729,
+      "step": 17170
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9943373203277588,
+      "learning_rate": 1.2113896391597878e-08,
+      "loss": 0.9313,
+      "step": 17171
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9776736497879028,
+      "learning_rate": 1.2022661549154769e-08,
+      "loss": 0.8921,
+      "step": 17172
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.773078203201294,
+      "learning_rate": 1.1931771363909995e-08,
+      "loss": 0.8297,
+      "step": 17173
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7074309587478638,
+      "learning_rate": 1.1841225838998827e-08,
+      "loss": 0.9223,
+      "step": 17174
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8111144304275513,
+      "learning_rate": 1.1751024977546543e-08,
+      "loss": 0.885,
+      "step": 17175
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7480214834213257,
+      "learning_rate": 1.16611687826651e-08,
+      "loss": 0.9442,
+      "step": 17176
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.824697732925415,
+      "learning_rate": 1.1571657257455349e-08,
+      "loss": 0.8634,
+      "step": 17177
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7207027673721313,
+      "learning_rate": 1.148249040500704e-08,
+      "loss": 0.8902,
+      "step": 17178
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7518699169158936,
+      "learning_rate": 1.1393668228395494e-08,
+      "loss": 0.8756,
+      "step": 17179
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.871133804321289,
+      "learning_rate": 1.1305190730686034e-08,
+      "loss": 0.8652,
+      "step": 17180
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7560901641845703,
+      "learning_rate": 1.1217057914932882e-08,
+      "loss": 0.9319,
+      "step": 17181
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8585501909255981,
+      "learning_rate": 1.1129269784175833e-08,
+      "loss": 0.9514,
+      "step": 17182
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7219775915145874,
+      "learning_rate": 1.1041826341445793e-08,
+      "loss": 0.8165,
+      "step": 17183
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6961294412612915,
+      "learning_rate": 1.095472758975813e-08,
+      "loss": 0.8941,
+      "step": 17184
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8639365434646606,
+      "learning_rate": 1.0867973532120436e-08,
+      "loss": 0.9362,
+      "step": 17185
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8855323791503906,
+      "learning_rate": 1.0781564171524761e-08,
+      "loss": 0.8431,
+      "step": 17186
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7122552394866943,
+      "learning_rate": 1.0695499510954276e-08,
+      "loss": 0.8716,
+      "step": 17187
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8580491542816162,
+      "learning_rate": 1.060977955337772e-08,
+      "loss": 0.8543,
+      "step": 17188
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8506311178207397,
+      "learning_rate": 1.0524404301753832e-08,
+      "loss": 0.8649,
+      "step": 17189
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7915905714035034,
+      "learning_rate": 1.0439373759028037e-08,
+      "loss": 0.8932,
+      "step": 17190
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7270772457122803,
+      "learning_rate": 1.0354687928134655e-08,
+      "loss": 0.7863,
+      "step": 17191
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7215646505355835,
+      "learning_rate": 1.027034681199579e-08,
+      "loss": 0.892,
+      "step": 17192
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7147690057754517,
+      "learning_rate": 1.018635041352245e-08,
+      "loss": 0.9039,
+      "step": 17193
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8432629108428955,
+      "learning_rate": 1.0102698735612315e-08,
+      "loss": 0.8989,
+      "step": 17194
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7604790925979614,
+      "learning_rate": 1.0019391781153076e-08,
+      "loss": 0.8715,
+      "step": 17195
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8725320100784302,
+      "learning_rate": 9.93642955301799e-09,
+      "loss": 0.8321,
+      "step": 17196
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7183960676193237,
+      "learning_rate": 9.853812054071432e-09,
+      "loss": 0.9397,
+      "step": 17197
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.798171043395996,
+      "learning_rate": 9.771539287163345e-09,
+      "loss": 0.8991,
+      "step": 17198
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7045540809631348,
+      "learning_rate": 9.68961125513257e-09,
+      "loss": 0.8625,
+      "step": 17199
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.686225175857544,
+      "learning_rate": 9.608027960805732e-09,
+      "loss": 0.9079,
+      "step": 17200
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7695108652114868,
+      "learning_rate": 9.52678940699947e-09,
+      "loss": 0.9661,
+      "step": 17201
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8676689863204956,
+      "learning_rate": 9.445895596517097e-09,
+      "loss": 0.8829,
+      "step": 17202
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7154688835144043,
+      "learning_rate": 9.3653465321486e-09,
+      "loss": 0.8037,
+      "step": 17203
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7137490510940552,
+      "learning_rate": 9.285142216675092e-09,
+      "loss": 0.8875,
+      "step": 17204
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7711985111236572,
+      "learning_rate": 9.205282652862135e-09,
+      "loss": 0.8764,
+      "step": 17205
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8837743997573853,
+      "learning_rate": 9.125767843467526e-09,
+      "loss": 0.8796,
+      "step": 17206
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1066405773162842,
+      "learning_rate": 9.046597791234624e-09,
+      "loss": 0.5435,
+      "step": 17207
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7785933017730713,
+      "learning_rate": 8.967772498894577e-09,
+      "loss": 0.8934,
+      "step": 17208
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7028650045394897,
+      "learning_rate": 8.889291969167435e-09,
+      "loss": 0.92,
+      "step": 17209
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9259285926818848,
+      "learning_rate": 8.811156204762139e-09,
+      "loss": 0.9289,
+      "step": 17210
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9579336643218994,
+      "learning_rate": 8.733365208374312e-09,
+      "loss": 0.9149,
+      "step": 17211
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8345398902893066,
+      "learning_rate": 8.655918982689582e-09,
+      "loss": 0.8363,
+      "step": 17212
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.728969931602478,
+      "learning_rate": 8.578817530378036e-09,
+      "loss": 0.9045,
+      "step": 17213
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7376890182495117,
+      "learning_rate": 8.5020608541031e-09,
+      "loss": 0.7658,
+      "step": 17214
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.830384373664856,
+      "learning_rate": 8.425648956510434e-09,
+      "loss": 1.0337,
+      "step": 17215
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7596899271011353,
+      "learning_rate": 8.34958184023904e-09,
+      "loss": 0.8931,
+      "step": 17216
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8630993366241455,
+      "learning_rate": 8.273859507913485e-09,
+      "loss": 0.9106,
+      "step": 17217
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.742850422859192,
+      "learning_rate": 8.19848196214501e-09,
+      "loss": 0.9151,
+      "step": 17218
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7061578035354614,
+      "learning_rate": 8.123449205537093e-09,
+      "loss": 0.8896,
+      "step": 17219
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7565820217132568,
+      "learning_rate": 8.04876124067766e-09,
+      "loss": 0.8627,
+      "step": 17220
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7931569814682007,
+      "learning_rate": 7.974418070143541e-09,
+      "loss": 0.9215,
+      "step": 17221
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.763895869255066,
+      "learning_rate": 7.900419696500461e-09,
+      "loss": 0.9329,
+      "step": 17222
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.82549250125885,
+      "learning_rate": 7.826766122303042e-09,
+      "loss": 0.8678,
+      "step": 17223
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.936253309249878,
+      "learning_rate": 7.753457350091476e-09,
+      "loss": 0.8659,
+      "step": 17224
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8933451175689697,
+      "learning_rate": 7.680493382395959e-09,
+      "loss": 1.0177,
+      "step": 17225
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8975743055343628,
+      "learning_rate": 7.607874221733369e-09,
+      "loss": 0.8707,
+      "step": 17226
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8239986896514893,
+      "learning_rate": 7.5355998706117e-09,
+      "loss": 0.9135,
+      "step": 17227
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7205934524536133,
+      "learning_rate": 7.463670331523399e-09,
+      "loss": 0.8547,
+      "step": 17228
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.576512336730957,
+      "learning_rate": 7.392085606949817e-09,
+      "loss": 0.8946,
+      "step": 17229
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7212175130844116,
+      "learning_rate": 7.320845699363421e-09,
+      "loss": 0.8746,
+      "step": 17230
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6207765340805054,
+      "learning_rate": 7.249950611220025e-09,
+      "loss": 0.8747,
+      "step": 17231
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6379351615905762,
+      "learning_rate": 7.1794003449676684e-09,
+      "loss": 0.8458,
+      "step": 17232
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.757816195487976,
+      "learning_rate": 7.109194903041073e-09,
+      "loss": 0.9361,
+      "step": 17233
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1014716625213623,
+      "learning_rate": 7.0393342878616325e-09,
+      "loss": 0.5234,
+      "step": 17234
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.861232876777649,
+      "learning_rate": 6.969818501839643e-09,
+      "loss": 0.8701,
+      "step": 17235
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7594435214996338,
+      "learning_rate": 6.900647547376516e-09,
+      "loss": 0.8218,
+      "step": 17236
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.704912543296814,
+      "learning_rate": 6.831821426855901e-09,
+      "loss": 0.8592,
+      "step": 17237
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8003016710281372,
+      "learning_rate": 6.763340142654784e-09,
+      "loss": 0.9959,
+      "step": 17238
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7108994722366333,
+      "learning_rate": 6.695203697136832e-09,
+      "loss": 0.8467,
+      "step": 17239
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.994081735610962,
+      "learning_rate": 6.6274120926512755e-09,
+      "loss": 0.8885,
+      "step": 17240
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7196333408355713,
+      "learning_rate": 6.559965331538465e-09,
+      "loss": 0.8546,
+      "step": 17241
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8456125259399414,
+      "learning_rate": 6.492863416125428e-09,
+      "loss": 0.9658,
+      "step": 17242
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.581201434135437,
+      "learning_rate": 6.4261063487292e-09,
+      "loss": 0.8655,
+      "step": 17243
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8731552362442017,
+      "learning_rate": 6.3596941316501624e-09,
+      "loss": 0.8043,
+      "step": 17244
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7131097316741943,
+      "learning_rate": 6.293626767183148e-09,
+      "loss": 0.9556,
+      "step": 17245
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.699301838874817,
+      "learning_rate": 6.2279042576074425e-09,
+      "loss": 0.9659,
+      "step": 17246
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8515162467956543,
+      "learning_rate": 6.162526605189012e-09,
+      "loss": 0.9216,
+      "step": 17247
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.698272705078125,
+      "learning_rate": 6.097493812186051e-09,
+      "loss": 0.9046,
+      "step": 17248
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0736775398254395,
+      "learning_rate": 6.032805880841209e-09,
+      "loss": 0.8813,
+      "step": 17249
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8179266452789307,
+      "learning_rate": 5.9684628133871435e-09,
+      "loss": 0.8279,
+      "step": 17250
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.759022831916809,
+      "learning_rate": 5.904464612044303e-09,
+      "loss": 0.8102,
+      "step": 17251
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.745347261428833,
+      "learning_rate": 5.840811279020919e-09,
+      "loss": 0.8959,
+      "step": 17252
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.802207589149475,
+      "learning_rate": 5.777502816514124e-09,
+      "loss": 0.8725,
+      "step": 17253
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6186901330947876,
+      "learning_rate": 5.714539226707727e-09,
+      "loss": 0.906,
+      "step": 17254
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7447986602783203,
+      "learning_rate": 5.651920511774433e-09,
+      "loss": 0.8402,
+      "step": 17255
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7617939710617065,
+      "learning_rate": 5.5896466738758485e-09,
+      "loss": 0.8709,
+      "step": 17256
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8744162321090698,
+      "learning_rate": 5.527717715159142e-09,
+      "loss": 0.8727,
+      "step": 17257
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0165075063705444,
+      "learning_rate": 5.466133637763715e-09,
+      "loss": 0.5246,
+      "step": 17258
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9690394401550293,
+      "learning_rate": 5.404894443812314e-09,
+      "loss": 0.9476,
+      "step": 17259
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9371414184570312,
+      "learning_rate": 5.344000135419913e-09,
+      "loss": 0.863,
+      "step": 17260
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7504358291625977,
+      "learning_rate": 5.283450714687055e-09,
+      "loss": 0.9737,
+      "step": 17261
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7477630376815796,
+      "learning_rate": 5.223246183703179e-09,
+      "loss": 0.881,
+      "step": 17262
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6652309894561768,
+      "learning_rate": 5.163386544545512e-09,
+      "loss": 0.9266,
+      "step": 17263
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9076701402664185,
+      "learning_rate": 5.10387179928018e-09,
+      "loss": 0.9125,
+      "step": 17264
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5830901861190796,
+      "learning_rate": 5.044701949961095e-09,
+      "loss": 0.851,
+      "step": 17265
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8062587976455688,
+      "learning_rate": 4.985876998628847e-09,
+      "loss": 0.9222,
+      "step": 17266
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8255336284637451,
+      "learning_rate": 4.927396947315144e-09,
+      "loss": 0.8615,
+      "step": 17267
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.905593752861023,
+      "learning_rate": 4.869261798035041e-09,
+      "loss": 0.8791,
+      "step": 17268
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7742645740509033,
+      "learning_rate": 4.811471552798042e-09,
+      "loss": 0.93,
+      "step": 17269
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9172112941741943,
+      "learning_rate": 4.754026213595886e-09,
+      "loss": 0.8941,
+      "step": 17270
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6594692468643188,
+      "learning_rate": 4.696925782411432e-09,
+      "loss": 0.875,
+      "step": 17271
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7498916387557983,
+      "learning_rate": 4.640170261216437e-09,
+      "loss": 0.9888,
+      "step": 17272
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7695399522781372,
+      "learning_rate": 4.583759651967113e-09,
+      "loss": 0.9271,
+      "step": 17273
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7547574043273926,
+      "learning_rate": 4.5276939566119004e-09,
+      "loss": 0.9253,
+      "step": 17274
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.035147190093994,
+      "learning_rate": 4.47197317708481e-09,
+      "loss": 0.8427,
+      "step": 17275
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.783646583557129,
+      "learning_rate": 4.4165973153076355e-09,
+      "loss": 0.9137,
+      "step": 17276
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8231968879699707,
+      "learning_rate": 4.3615663731932936e-09,
+      "loss": 0.9847,
+      "step": 17277
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.018293857574463,
+      "learning_rate": 4.306880352639154e-09,
+      "loss": 0.8412,
+      "step": 17278
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7986867427825928,
+      "learning_rate": 4.252539255532595e-09,
+      "loss": 0.9551,
+      "step": 17279
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.831531286239624,
+      "learning_rate": 4.198543083748785e-09,
+      "loss": 0.962,
+      "step": 17280
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8445018529891968,
+      "learning_rate": 4.144891839150678e-09,
+      "loss": 0.9136,
+      "step": 17281
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7745819091796875,
+      "learning_rate": 4.091585523591235e-09,
+      "loss": 0.976,
+      "step": 17282
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8485870361328125,
+      "learning_rate": 4.038624138907876e-09,
+      "loss": 0.9329,
+      "step": 17283
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.834294319152832,
+      "learning_rate": 3.9860076869291385e-09,
+      "loss": 0.9365,
+      "step": 17284
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7621880769729614,
+      "learning_rate": 3.933736169471347e-09,
+      "loss": 0.9385,
+      "step": 17285
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.733346700668335,
+      "learning_rate": 3.881809588336394e-09,
+      "loss": 0.9959,
+      "step": 17286
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7475131750106812,
+      "learning_rate": 3.8302279453172885e-09,
+      "loss": 0.9017,
+      "step": 17287
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9727720022201538,
+      "learning_rate": 3.77899124219483e-09,
+      "loss": 0.5222,
+      "step": 17288
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8779473304748535,
+      "learning_rate": 3.728099480735381e-09,
+      "loss": 0.8489,
+      "step": 17289
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8390697240829468,
+      "learning_rate": 3.677552662695316e-09,
+      "loss": 0.9777,
+      "step": 17290
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7679295539855957,
+      "learning_rate": 3.6273507898199057e-09,
+      "loss": 0.8226,
+      "step": 17291
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.679634690284729,
+      "learning_rate": 3.577493863841097e-09,
+      "loss": 0.9521,
+      "step": 17292
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8505855798721313,
+      "learning_rate": 3.5279818864786264e-09,
+      "loss": 0.9067,
+      "step": 17293
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7914808988571167,
+      "learning_rate": 3.4788148594411262e-09,
+      "loss": 0.8174,
+      "step": 17294
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.918171763420105,
+      "learning_rate": 3.429992784426128e-09,
+      "loss": 0.9265,
+      "step": 17295
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.758527159690857,
+      "learning_rate": 3.3815156631178404e-09,
+      "loss": 0.9308,
+      "step": 17296
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7671966552734375,
+      "learning_rate": 3.3333834971882582e-09,
+      "loss": 0.8315,
+      "step": 17297
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8305507898330688,
+      "learning_rate": 3.2855962883004968e-09,
+      "loss": 0.9141,
+      "step": 17298
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8487319946289062,
+      "learning_rate": 3.2381540381010157e-09,
+      "loss": 0.8533,
+      "step": 17299
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8038307428359985,
+      "learning_rate": 3.1910567482285048e-09,
+      "loss": 0.8889,
+      "step": 17300
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8449538946151733,
+      "learning_rate": 3.1443044203072205e-09,
+      "loss": 0.9464,
+      "step": 17301
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7164820432662964,
+      "learning_rate": 3.0978970559503164e-09,
+      "loss": 0.8779,
+      "step": 17302
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8161195516586304,
+      "learning_rate": 3.0518346567609546e-09,
+      "loss": 0.864,
+      "step": 17303
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8048014640808105,
+      "learning_rate": 3.0061172243267543e-09,
+      "loss": 0.925,
+      "step": 17304
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6757572889328003,
+      "learning_rate": 2.9607447602264526e-09,
+      "loss": 0.8084,
+      "step": 17305
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8629672527313232,
+      "learning_rate": 2.915717266025464e-09,
+      "loss": 0.9228,
+      "step": 17306
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.797763466835022,
+      "learning_rate": 2.8710347432769903e-09,
+      "loss": 0.8502,
+      "step": 17307
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6682815551757812,
+      "learning_rate": 2.8266971935231312e-09,
+      "loss": 0.9116,
+      "step": 17308
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.71448814868927,
+      "learning_rate": 2.7827046182937744e-09,
+      "loss": 0.8901,
+      "step": 17309
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8518085479736328,
+      "learning_rate": 2.7390570191077048e-09,
+      "loss": 0.9126,
+      "step": 17310
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6976794004440308,
+      "learning_rate": 2.695754397470385e-09,
+      "loss": 0.9096,
+      "step": 17311
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6713060140609741,
+      "learning_rate": 2.6527967548761747e-09,
+      "loss": 0.9112,
+      "step": 17312
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8100568056106567,
+      "learning_rate": 2.610184092807222e-09,
+      "loss": 0.9074,
+      "step": 17313
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8403061628341675,
+      "learning_rate": 2.5679164127345726e-09,
+      "loss": 0.9149,
+      "step": 17314
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8030495643615723,
+      "learning_rate": 2.5259937161159485e-09,
+      "loss": 0.8348,
+      "step": 17315
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9048752784729004,
+      "learning_rate": 2.4844160043990817e-09,
+      "loss": 0.8158,
+      "step": 17316
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5969892740249634,
+      "learning_rate": 2.4431832790172693e-09,
+      "loss": 0.8478,
+      "step": 17317
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7424389123916626,
+      "learning_rate": 2.402295541394928e-09,
+      "loss": 0.9026,
+      "step": 17318
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7886524200439453,
+      "learning_rate": 2.36175279294204e-09,
+      "loss": 0.844,
+      "step": 17319
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6108866930007935,
+      "learning_rate": 2.3215550350574877e-09,
+      "loss": 0.868,
+      "step": 17320
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8164024353027344,
+      "learning_rate": 2.281702269129049e-09,
+      "loss": 0.9111,
+      "step": 17321
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7763748168945312,
+      "learning_rate": 2.2421944965311803e-09,
+      "loss": 0.9436,
+      "step": 17322
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7680420875549316,
+      "learning_rate": 2.2030317186283457e-09,
+      "loss": 0.9507,
+      "step": 17323
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6039468050003052,
+      "learning_rate": 2.164213936770576e-09,
+      "loss": 0.846,
+      "step": 17324
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.952896237373352,
+      "learning_rate": 2.125741152297911e-09,
+      "loss": 0.9217,
+      "step": 17325
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.07783579826355,
+      "learning_rate": 2.087613366538177e-09,
+      "loss": 0.8864,
+      "step": 17326
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6380054950714111,
+      "learning_rate": 2.0498305808069886e-09,
+      "loss": 0.8736,
+      "step": 17327
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9388900995254517,
+      "learning_rate": 2.0123927964077473e-09,
+      "loss": 0.8911,
+      "step": 17328
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9390250444412231,
+      "learning_rate": 1.9753000146327527e-09,
+      "loss": 0.8948,
+      "step": 17329
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7530548572540283,
+      "learning_rate": 1.9385522367620922e-09,
+      "loss": 0.925,
+      "step": 17330
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6160852909088135,
+      "learning_rate": 1.9021494640625302e-09,
+      "loss": 0.7857,
+      "step": 17331
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7491726875305176,
+      "learning_rate": 1.8660916977919495e-09,
+      "loss": 0.8592,
+      "step": 17332
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7526048421859741,
+      "learning_rate": 1.8303789391937999e-09,
+      "loss": 0.8532,
+      "step": 17333
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6760650873184204,
+      "learning_rate": 1.7950111895004285e-09,
+      "loss": 0.8776,
+      "step": 17334
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8032807111740112,
+      "learning_rate": 1.7599884499319708e-09,
+      "loss": 0.8643,
+      "step": 17335
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9885274171829224,
+      "learning_rate": 1.72531072169857e-09,
+      "loss": 0.8011,
+      "step": 17336
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9964971542358398,
+      "learning_rate": 1.6909780059948255e-09,
+      "loss": 0.8789,
+      "step": 17337
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8368868827819824,
+      "learning_rate": 1.656990304006456e-09,
+      "loss": 0.8164,
+      "step": 17338
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9991557598114014,
+      "learning_rate": 1.623347616905857e-09,
+      "loss": 0.5255,
+      "step": 17339
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8376662731170654,
+      "learning_rate": 1.5900499458543216e-09,
+      "loss": 0.921,
+      "step": 17340
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7481489181518555,
+      "learning_rate": 1.557097292000931e-09,
+      "loss": 0.8681,
+      "step": 17341
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8019217252731323,
+      "learning_rate": 1.524489656482553e-09,
+      "loss": 0.898,
+      "step": 17342
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7518693208694458,
+      "learning_rate": 1.4922270404238436e-09,
+      "loss": 0.904,
+      "step": 17343
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.019370198249817,
+      "learning_rate": 1.4603094449394673e-09,
+      "loss": 0.551,
+      "step": 17344
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8118783235549927,
+      "learning_rate": 1.4287368711296546e-09,
+      "loss": 0.9583,
+      "step": 17345
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.884177565574646,
+      "learning_rate": 1.3975093200835344e-09,
+      "loss": 0.847,
+      "step": 17346
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8861427307128906,
+      "learning_rate": 1.3666267928802435e-09,
+      "loss": 0.9287,
+      "step": 17347
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8120076656341553,
+      "learning_rate": 1.3360892905844858e-09,
+      "loss": 0.9,
+      "step": 17348
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8624743223190308,
+      "learning_rate": 1.305896814249863e-09,
+      "loss": 0.8985,
+      "step": 17349
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8163299560546875,
+      "learning_rate": 1.2760493649188744e-09,
+      "loss": 0.923,
+      "step": 17350
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7511534690856934,
+      "learning_rate": 1.246546943621807e-09,
+      "loss": 0.8586,
+      "step": 17351
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.720291018486023,
+      "learning_rate": 1.2173895513745148e-09,
+      "loss": 0.8695,
+      "step": 17352
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0976109504699707,
+      "learning_rate": 1.188577189186191e-09,
+      "loss": 0.9732,
+      "step": 17353
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7866488695144653,
+      "learning_rate": 1.1601098580482639e-09,
+      "loss": 0.9215,
+      "step": 17354
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6885319948196411,
+      "learning_rate": 1.1319875589443918e-09,
+      "loss": 0.8194,
+      "step": 17355
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9684853553771973,
+      "learning_rate": 1.1042102928460196e-09,
+      "loss": 0.4855,
+      "step": 17356
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9425514936447144,
+      "learning_rate": 1.0767780607090494e-09,
+      "loss": 0.8542,
+      "step": 17357
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6254544258117676,
+      "learning_rate": 1.0496908634827218e-09,
+      "loss": 0.9099,
+      "step": 17358
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.193384051322937,
+      "learning_rate": 1.0229487020996243e-09,
+      "loss": 0.4737,
+      "step": 17359
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7835793495178223,
+      "learning_rate": 9.965515774845723e-10,
+      "loss": 0.8662,
+      "step": 17360
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7936466932296753,
+      "learning_rate": 9.70499490546839e-10,
+      "loss": 0.9262,
+      "step": 17361
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8469680547714233,
+      "learning_rate": 9.447924421868148e-10,
+      "loss": 0.9731,
+      "step": 17362
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8499726057052612,
+      "learning_rate": 9.19430433290458e-10,
+      "loss": 0.9678,
+      "step": 17363
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8740984201431274,
+      "learning_rate": 8.944134647326241e-10,
+      "loss": 0.9736,
+      "step": 17364
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.848207712173462,
+      "learning_rate": 8.697415373770668e-10,
+      "loss": 0.9566,
+      "step": 17365
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7580366134643555,
+      "learning_rate": 8.454146520764373e-10,
+      "loss": 0.9085,
+      "step": 17366
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8541340827941895,
+      "learning_rate": 8.214328096678437e-10,
+      "loss": 0.857,
+      "step": 17367
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8808695077896118,
+      "learning_rate": 7.977960109806226e-10,
+      "loss": 0.8506,
+      "step": 17368
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6626999378204346,
+      "learning_rate": 7.745042568296779e-10,
+      "loss": 0.863,
+      "step": 17369
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.829413652420044,
+      "learning_rate": 7.515575480188109e-10,
+      "loss": 0.9216,
+      "step": 17370
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7048414945602417,
+      "learning_rate": 7.289558853407208e-10,
+      "loss": 0.8215,
+      "step": 17371
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8696891069412231,
+      "learning_rate": 7.066992695736741e-10,
+      "loss": 0.9779,
+      "step": 17372
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8192112445831299,
+      "learning_rate": 6.847877014870552e-10,
+      "loss": 0.9278,
+      "step": 17373
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.810707688331604,
+      "learning_rate": 6.632211818358159e-10,
+      "loss": 0.8488,
+      "step": 17374
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.922389030456543,
+      "learning_rate": 6.419997113649156e-10,
+      "loss": 0.9139,
+      "step": 17375
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8385673761367798,
+      "learning_rate": 6.211232908071018e-10,
+      "loss": 0.8446,
+      "step": 17376
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6024255752563477,
+      "learning_rate": 6.00591920881799e-10,
+      "loss": 0.9176,
+      "step": 17377
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.808584213256836,
+      "learning_rate": 5.804056022973292e-10,
+      "loss": 0.935,
+      "step": 17378
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0808799266815186,
+      "learning_rate": 5.605643357520229e-10,
+      "loss": 0.9457,
+      "step": 17379
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9911454916000366,
+      "learning_rate": 5.410681219286673e-10,
+      "loss": 0.5222,
+      "step": 17380
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.739420771598816,
+      "learning_rate": 5.219169615000574e-10,
+      "loss": 0.9829,
+      "step": 17381
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7970499992370605,
+      "learning_rate": 5.031108551289965e-10,
+      "loss": 0.8973,
+      "step": 17382
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.724595308303833,
+      "learning_rate": 4.846498034616342e-10,
+      "loss": 0.9242,
+      "step": 17383
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6912459135055542,
+      "learning_rate": 4.665338071374592e-10,
+      "loss": 0.9215,
+      "step": 17384
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6548596620559692,
+      "learning_rate": 4.487628667804167e-10,
+      "loss": 0.8796,
+      "step": 17385
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.742425799369812,
+      "learning_rate": 4.3133698300446e-10,
+      "loss": 0.7722,
+      "step": 17386
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6803466081619263,
+      "learning_rate": 4.1425615641021986e-10,
+      "loss": 0.8334,
+      "step": 17387
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7410744428634644,
+      "learning_rate": 3.975203875861144e-10,
+      "loss": 0.8724,
+      "step": 17388
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7138065099716187,
+      "learning_rate": 3.8112967711168013e-10,
+      "loss": 0.9301,
+      "step": 17389
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.581705093383789,
+      "learning_rate": 3.650840255520205e-10,
+      "loss": 0.8427,
+      "step": 17390
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.812705397605896,
+      "learning_rate": 3.4938343346002657e-10,
+      "loss": 0.9257,
+      "step": 17391
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7295976877212524,
+      "learning_rate": 3.3402790137859744e-10,
+      "loss": 0.9071,
+      "step": 17392
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7323287725448608,
+      "learning_rate": 3.190174298361992e-10,
+      "loss": 0.9106,
+      "step": 17393
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8838635683059692,
+      "learning_rate": 3.043520193513061e-10,
+      "loss": 0.8679,
+      "step": 17394
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7170557975769043,
+      "learning_rate": 2.9003167043017975e-10,
+      "loss": 0.7926,
+      "step": 17395
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.724204182624817,
+      "learning_rate": 2.760563835679797e-10,
+      "loss": 0.8397,
+      "step": 17396
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.85691237449646,
+      "learning_rate": 2.6242615924543245e-10,
+      "loss": 0.9537,
+      "step": 17397
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.871584415435791,
+      "learning_rate": 2.4914099793327263e-10,
+      "loss": 0.9361,
+      "step": 17398
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8563990592956543,
+      "learning_rate": 2.3620090009002227e-10,
+      "loss": 0.843,
+      "step": 17399
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7285387516021729,
+      "learning_rate": 2.236058661619911e-10,
+      "loss": 0.8923,
+      "step": 17400
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.770787000656128,
+      "learning_rate": 2.113558965843865e-10,
+      "loss": 0.8035,
+      "step": 17401
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6742857694625854,
+      "learning_rate": 1.9945099178020345e-10,
+      "loss": 0.8655,
+      "step": 17402
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7817342281341553,
+      "learning_rate": 1.8789115215911424e-10,
+      "loss": 0.8919,
+      "step": 17403
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7028727531433105,
+      "learning_rate": 1.7667637812079918e-10,
+      "loss": 0.8223,
+      "step": 17404
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6603620052337646,
+      "learning_rate": 1.6580667005161589e-10,
+      "loss": 0.8239,
+      "step": 17405
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.749579668045044,
+      "learning_rate": 1.5528202832681972e-10,
+      "loss": 0.8359,
+      "step": 17406
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.78286874294281,
+      "learning_rate": 1.4510245331056383e-10,
+      "loss": 0.8856,
+      "step": 17407
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8148283958435059,
+      "learning_rate": 1.3526794535256848e-10,
+      "loss": 0.9131,
+      "step": 17408
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.525618076324463,
+      "learning_rate": 1.2577850479367214e-10,
+      "loss": 0.8512,
+      "step": 17409
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7967215776443481,
+      "learning_rate": 1.1663413196028039e-10,
+      "loss": 0.9035,
+      "step": 17410
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.115170955657959,
+      "learning_rate": 1.0783482716880677e-10,
+      "loss": 0.5575,
+      "step": 17411
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8722461462020874,
+      "learning_rate": 9.938059072123196e-11,
+      "loss": 0.9315,
+      "step": 17412
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7386447191238403,
+      "learning_rate": 9.127142291176506e-11,
+      "loss": 0.8546,
+      "step": 17413
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6093006134033203,
+      "learning_rate": 8.35073240179618e-11,
+      "loss": 0.9021,
+      "step": 17414
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8057047128677368,
+      "learning_rate": 7.60882943096064e-11,
+      "loss": 0.887,
+      "step": 17415
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.009599208831787,
+      "learning_rate": 6.901433404093993e-11,
+      "loss": 0.933,
+      "step": 17416
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.808866024017334,
+      "learning_rate": 6.228544345732168e-11,
+      "loss": 0.8222,
+      "step": 17417
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.743621587753296,
+      "learning_rate": 5.5901622789678035e-11,
+      "loss": 0.9033,
+      "step": 17418
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8964340686798096,
+      "learning_rate": 4.98628722600536e-11,
+      "loss": 0.8462,
+      "step": 17419
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7574759721755981,
+      "learning_rate": 4.416919207606007e-11,
+      "loss": 0.9225,
+      "step": 17420
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6010372638702393,
+      "learning_rate": 3.8820582434206944e-11,
+      "loss": 0.7561,
+      "step": 17421
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.743086576461792,
+      "learning_rate": 3.3817043517681e-11,
+      "loss": 0.8319,
+      "step": 17422
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7555707693099976,
+      "learning_rate": 2.915857550078727e-11,
+      "loss": 0.8599,
+      "step": 17423
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6431801319122314,
+      "learning_rate": 2.4845178544508074e-11,
+      "loss": 0.914,
+      "step": 17424
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.865052342414856,
+      "learning_rate": 2.087685279650309e-11,
+      "loss": 0.8689,
+      "step": 17425
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6582229137420654,
+      "learning_rate": 1.7253598394439965e-11,
+      "loss": 0.8424,
+      "step": 17426
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0003459453582764,
+      "learning_rate": 1.3975415462663678e-11,
+      "loss": 0.8006,
+      "step": 17427
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7288761138916016,
+      "learning_rate": 1.1042304115527203e-11,
+      "loss": 0.8445,
+      "step": 17428
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8551204204559326,
+      "learning_rate": 8.454264452950612e-12,
+      "loss": 0.9486,
+      "step": 17429
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8679002523422241,
+      "learning_rate": 6.211296564861969e-12,
+      "loss": 0.8834,
+      "step": 17430
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8377928733825684,
+      "learning_rate": 4.313400528976886e-12,
+      "loss": 0.9727,
+      "step": 17431
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.751551628112793,
+      "learning_rate": 2.7605764096882983e-12,
+      "loss": 0.8406,
+      "step": 17432
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9908881187438965,
+      "learning_rate": 1.5528242613971345e-12,
+      "loss": 0.9007,
+      "step": 17433
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6909549236297607,
+      "learning_rate": 6.901441262918696e-13,
+      "loss": 0.9673,
+      "step": 17434
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9202371835708618,
+      "learning_rate": 1.7253603323830192e-13,
+      "loss": 0.9763,
+      "step": 17435
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.094381332397461,
+      "learning_rate": 0.0,
+      "loss": 0.7425,
+      "step": 17436
+    },
+    {
+      "epoch": 1.0,
+      "step": 17436,
+      "total_flos": 6.360551734822673e+19,
+      "train_loss": 0.9591581591121454,
+      "train_runtime": 243882.3408,
+      "train_samples_per_second": 9.151,
+      "train_steps_per_second": 0.071
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 17436,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 6.360551734822673e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}