diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,20860 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 2974,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.715430917495459,
+      "learning_rate": 2.2222222222222224e-07,
+      "loss": 2.532,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.17694558714131,
+      "learning_rate": 4.444444444444445e-07,
+      "loss": 1.7172,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.852981860601261,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 2.4779,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.049806080940916,
+      "learning_rate": 8.88888888888889e-07,
+      "loss": 1.6707,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.467261416485208,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 1.8766,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.594429177271186,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 2.2758,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.840022857557608,
+      "learning_rate": 1.5555555555555558e-06,
+      "loss": 1.7541,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.783688261843984,
+      "learning_rate": 1.777777777777778e-06,
+      "loss": 1.8476,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.253504483866347,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.7445,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.034782711480892,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 2.4426,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.927538775448608,
+      "learning_rate": 2.4444444444444447e-06,
+      "loss": 1.5423,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.06595124883192,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 1.5731,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.285179935004488,
+      "learning_rate": 2.888888888888889e-06,
+      "loss": 2.3457,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.075524376068787,
+      "learning_rate": 3.1111111111111116e-06,
+      "loss": 1.5826,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.017818166469643,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 2.213,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7318633156311356,
+      "learning_rate": 3.555555555555556e-06,
+      "loss": 1.5819,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.39546853801011,
+      "learning_rate": 3.777777777777778e-06,
+      "loss": 1.5416,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.921650745947644,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 2.0629,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.108978052847991,
+      "learning_rate": 4.222222222222223e-06,
+      "loss": 1.5024,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6181871752374364,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 1.6358,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.7975062951395255,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 1.5625,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.49423992051924,
+      "learning_rate": 4.888888888888889e-06,
+      "loss": 1.3553,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.173827532615298,
+      "learning_rate": 5.1111111111111115e-06,
+      "loss": 1.96,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.944373073326538,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 1.5037,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8608379572360616,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 2.1765,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.570333560531981,
+      "learning_rate": 5.777777777777778e-06,
+      "loss": 1.5346,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2928769024537234,
+      "learning_rate": 6e-06,
+      "loss": 2.1735,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1063165009959404,
+      "learning_rate": 6.222222222222223e-06,
+      "loss": 1.6136,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.699868617625275,
+      "learning_rate": 6.444444444444445e-06,
+      "loss": 1.4974,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.480521181889438,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.8677,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5231580531776756,
+      "learning_rate": 6.88888888888889e-06,
+      "loss": 1.3775,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9141152207867023,
+      "learning_rate": 7.111111111111112e-06,
+      "loss": 1.6464,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1620132177879707,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 1.5225,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.690825499696442,
+      "learning_rate": 7.555555555555556e-06,
+      "loss": 2.0678,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2507799369374044,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 1.3074,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5554741779343515,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.3032,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1059311017537596,
+      "learning_rate": 8.222222222222222e-06,
+      "loss": 2.0574,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.583643022776292,
+      "learning_rate": 8.444444444444446e-06,
+      "loss": 1.3344,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.991514119852815,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 2.0502,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9507675715216084,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 1.4362,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.328937241950173,
+      "learning_rate": 9.111111111111112e-06,
+      "loss": 1.4862,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3059593199880077,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 1.9858,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9542829856880615,
+      "learning_rate": 9.555555555555556e-06,
+      "loss": 1.3439,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1934725094459937,
+      "learning_rate": 9.777777777777779e-06,
+      "loss": 2.306,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8819617356463773,
+      "learning_rate": 1e-05,
+      "loss": 1.338,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1214700144009517,
+      "learning_rate": 1.0222222222222223e-05,
+      "loss": 1.8218,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2989403555588055,
+      "learning_rate": 1.0444444444444445e-05,
+      "loss": 1.432,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.921810174878452,
+      "learning_rate": 1.0666666666666667e-05,
+      "loss": 1.9039,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4092370584286464,
+      "learning_rate": 1.088888888888889e-05,
+      "loss": 2.0572,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1511504427640797,
+      "learning_rate": 1.1111111111111113e-05,
+      "loss": 1.2611,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.506315004642144,
+      "learning_rate": 1.1333333333333334e-05,
+      "loss": 1.9578,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0019113543823839,
+      "learning_rate": 1.1555555555555556e-05,
+      "loss": 1.4237,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4308520514437895,
+      "learning_rate": 1.177777777777778e-05,
+      "loss": 1.7871,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9957137072333171,
+      "learning_rate": 1.2e-05,
+      "loss": 1.4368,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8438417181915796,
+      "learning_rate": 1.2222222222222224e-05,
+      "loss": 1.3727,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1546800134527406,
+      "learning_rate": 1.2444444444444446e-05,
+      "loss": 1.8104,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1628022448782445,
+      "learning_rate": 1.2666666666666667e-05,
+      "loss": 1.3965,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9702494018185307,
+      "learning_rate": 1.288888888888889e-05,
+      "loss": 1.9994,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3463850272322386,
+      "learning_rate": 1.3111111111111113e-05,
+      "loss": 1.3012,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1571121093421481,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.9551,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9703420901162721,
+      "learning_rate": 1.3555555555555557e-05,
+      "loss": 1.1666,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8321792193765819,
+      "learning_rate": 1.377777777777778e-05,
+      "loss": 1.2285,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0996164076366128,
+      "learning_rate": 1.4e-05,
+      "loss": 2.0074,
+      "step": 63
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3381244724400103,
+      "learning_rate": 1.4222222222222224e-05,
+      "loss": 1.42,
+      "step": 64
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4372439924657456,
+      "learning_rate": 1.4444444444444446e-05,
+      "loss": 1.7066,
+      "step": 65
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8208890339820745,
+      "learning_rate": 1.4666666666666666e-05,
+      "loss": 1.3208,
+      "step": 66
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9173967661505593,
+      "learning_rate": 1.488888888888889e-05,
+      "loss": 1.2615,
+      "step": 67
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9560457489927107,
+      "learning_rate": 1.5111111111111112e-05,
+      "loss": 1.7884,
+      "step": 68
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7549335273563856,
+      "learning_rate": 1.5333333333333334e-05,
+      "loss": 1.3065,
+      "step": 69
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3306835388533986,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 1.7792,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9807259208557758,
+      "learning_rate": 1.577777777777778e-05,
+      "loss": 1.3416,
+      "step": 71
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.412405457898274,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.6578,
+      "step": 72
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9777890718408888,
+      "learning_rate": 1.6222222222222223e-05,
+      "loss": 1.3537,
+      "step": 73
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9222233626932654,
+      "learning_rate": 1.6444444444444444e-05,
+      "loss": 1.2838,
+      "step": 74
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0676555532494134,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 1.6579,
+      "step": 75
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7678398262504119,
+      "learning_rate": 1.688888888888889e-05,
+      "loss": 1.295,
+      "step": 76
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8077466088841951,
+      "learning_rate": 1.7111111111111112e-05,
+      "loss": 1.6312,
+      "step": 77
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.179524914848577,
+      "learning_rate": 1.7333333333333336e-05,
+      "loss": 1.1888,
+      "step": 78
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7577401565091403,
+      "learning_rate": 1.7555555555555556e-05,
+      "loss": 1.3214,
+      "step": 79
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6794568745401532,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 1.8492,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6260339034145541,
+      "learning_rate": 1.8e-05,
+      "loss": 1.3796,
+      "step": 81
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9439643222550497,
+      "learning_rate": 1.8222222222222224e-05,
+      "loss": 1.6299,
+      "step": 82
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.293929347664828,
+      "learning_rate": 1.8444444444444448e-05,
+      "loss": 1.1307,
+      "step": 83
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7875928524973604,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 1.2202,
+      "step": 84
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0395910450958372,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 1.6414,
+      "step": 85
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8327153678755055,
+      "learning_rate": 1.9111111111111113e-05,
+      "loss": 1.2989,
+      "step": 86
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1506967589010446,
+      "learning_rate": 1.9333333333333333e-05,
+      "loss": 1.6514,
+      "step": 87
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0620257341923545,
+      "learning_rate": 1.9555555555555557e-05,
+      "loss": 1.1664,
+      "step": 88
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1922367320764897,
+      "learning_rate": 1.977777777777778e-05,
+      "loss": 1.7932,
+      "step": 89
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6939456610018607,
+      "learning_rate": 2e-05,
+      "loss": 1.2367,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5701128233990298,
+      "learning_rate": 1.9999994066933796e-05,
+      "loss": 1.0592,
+      "step": 91
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8166681709474466,
+      "learning_rate": 1.999997626774221e-05,
+      "loss": 1.7546,
+      "step": 92
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8771092138503676,
+      "learning_rate": 1.9999946602446377e-05,
+      "loss": 1.203,
+      "step": 93
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9420254454309991,
+      "learning_rate": 1.9999905071081485e-05,
+      "loss": 1.6937,
+      "step": 94
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7931585429881932,
+      "learning_rate": 1.9999851673696827e-05,
+      "loss": 1.2416,
+      "step": 95
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7766188839145032,
+      "learning_rate": 1.999978641035576e-05,
+      "loss": 1.7619,
+      "step": 96
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7754561507472486,
+      "learning_rate": 1.9999709281135723e-05,
+      "loss": 1.3887,
+      "step": 97
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8540902135988279,
+      "learning_rate": 1.9999620286128243e-05,
+      "loss": 1.1612,
+      "step": 98
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0646677324266913,
+      "learning_rate": 1.9999519425438922e-05,
+      "loss": 1.8008,
+      "step": 99
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8573903328492992,
+      "learning_rate": 1.9999406699187443e-05,
+      "loss": 1.1622,
+      "step": 100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2458112966096304,
+      "learning_rate": 1.9999282107507567e-05,
+      "loss": 1.3055,
+      "step": 101
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0140918422117855,
+      "learning_rate": 1.9999145650547137e-05,
+      "loss": 1.3578,
+      "step": 102
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8391622900713686,
+      "learning_rate": 1.9998997328468075e-05,
+      "loss": 1.4434,
+      "step": 103
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2512902349612376,
+      "learning_rate": 1.9998837141446378e-05,
+      "loss": 1.7464,
+      "step": 104
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7570282709763066,
+      "learning_rate": 1.9998665089672132e-05,
+      "loss": 1.2645,
+      "step": 105
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0475905349102121,
+      "learning_rate": 1.9998481173349497e-05,
+      "loss": 1.4317,
+      "step": 106
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9442530497528,
+      "learning_rate": 1.9998285392696702e-05,
+      "loss": 1.1067,
+      "step": 107
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8558332636691748,
+      "learning_rate": 1.999807774794607e-05,
+      "loss": 1.6658,
+      "step": 108
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8328962112738916,
+      "learning_rate": 1.999785823934399e-05,
+      "loss": 1.2389,
+      "step": 109
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.203854847969168,
+      "learning_rate": 1.999762686715094e-05,
+      "loss": 1.0628,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1888881091101131,
+      "learning_rate": 1.9997383631641464e-05,
+      "loss": 1.7182,
+      "step": 111
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0355012029379682,
+      "learning_rate": 1.9997128533104192e-05,
+      "loss": 1.2197,
+      "step": 112
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9582702511298423,
+      "learning_rate": 1.9996861571841824e-05,
+      "loss": 1.6465,
+      "step": 113
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8347285746888967,
+      "learning_rate": 1.999658274817114e-05,
+      "loss": 1.2388,
+      "step": 114
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7884704257732795,
+      "learning_rate": 1.9996292062423e-05,
+      "loss": 1.5976,
+      "step": 115
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7035119137586714,
+      "learning_rate": 1.999598951494233e-05,
+      "loss": 1.1308,
+      "step": 116
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7492688286033283,
+      "learning_rate": 1.9995675106088143e-05,
+      "loss": 1.1201,
+      "step": 117
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.782178796074627,
+      "learning_rate": 1.9995348836233517e-05,
+      "loss": 1.3695,
+      "step": 118
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7900150046886106,
+      "learning_rate": 1.9995010705765607e-05,
+      "loss": 1.2192,
+      "step": 119
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1730948156448127,
+      "learning_rate": 1.9994660715085645e-05,
+      "loss": 1.602,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7934939841201235,
+      "learning_rate": 1.9994298864608936e-05,
+      "loss": 1.2805,
+      "step": 121
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8776650662512081,
+      "learning_rate": 1.9993925154764856e-05,
+      "loss": 1.2604,
+      "step": 122
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3444489577804166,
+      "learning_rate": 1.999353958599685e-05,
+      "loss": 1.5035,
+      "step": 123
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7185308873205809,
+      "learning_rate": 1.9993142158762443e-05,
+      "loss": 1.2708,
+      "step": 124
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.592023043477027,
+      "learning_rate": 1.9992732873533223e-05,
+      "loss": 1.7342,
+      "step": 125
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8454098063700842,
+      "learning_rate": 1.999231173079486e-05,
+      "loss": 1.2203,
+      "step": 126
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0516519324578995,
+      "learning_rate": 1.9991878731047083e-05,
+      "loss": 1.1577,
+      "step": 127
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6630998805093147,
+      "learning_rate": 1.99914338748037e-05,
+      "loss": 1.2119,
+      "step": 128
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8329166443801544,
+      "learning_rate": 1.9990977162592575e-05,
+      "loss": 1.2433,
+      "step": 129
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3325219834657902,
+      "learning_rate": 1.9990508594955654e-05,
+      "loss": 1.549,
+      "step": 130
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8609800242915061,
+      "learning_rate": 1.9990028172448945e-05,
+      "loss": 1.2546,
+      "step": 131
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7255360813981535,
+      "learning_rate": 1.9989535895642523e-05,
+      "loss": 1.6349,
+      "step": 132
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.210054113102911,
+      "learning_rate": 1.9989031765120534e-05,
+      "loss": 1.2553,
+      "step": 133
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0137292817869574,
+      "learning_rate": 1.998851578148118e-05,
+      "loss": 1.5045,
+      "step": 134
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6718864289198011,
+      "learning_rate": 1.998798794533674e-05,
+      "loss": 1.2788,
+      "step": 135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.253497936648561,
+      "learning_rate": 1.9987448257313545e-05,
+      "loss": 1.0889,
+      "step": 136
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0341488888371062,
+      "learning_rate": 1.9986896718052e-05,
+      "loss": 1.5722,
+      "step": 137
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9529128094115333,
+      "learning_rate": 1.9986333328206567e-05,
+      "loss": 1.1695,
+      "step": 138
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1430571734599142,
+      "learning_rate": 1.9985758088445774e-05,
+      "loss": 1.4587,
+      "step": 139
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2325472126495245,
+      "learning_rate": 1.9985170999452207e-05,
+      "loss": 1.3346,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6462055090021601,
+      "learning_rate": 1.9984572061922515e-05,
+      "loss": 1.2499,
+      "step": 141
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5281465111398673,
+      "learning_rate": 1.99839612765674e-05,
+      "loss": 1.634,
+      "step": 142
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8405905849912727,
+      "learning_rate": 1.9983338644111635e-05,
+      "loss": 1.2297,
+      "step": 143
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8699606125088473,
+      "learning_rate": 1.998270416529404e-05,
+      "loss": 1.4812,
+      "step": 144
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.460993389741857,
+      "learning_rate": 1.9982057840867495e-05,
+      "loss": 1.0798,
+      "step": 145
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1069253834059314,
+      "learning_rate": 1.998139967159894e-05,
+      "loss": 1.4344,
+      "step": 146
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2451818949884668,
+      "learning_rate": 1.9980729658269364e-05,
+      "loss": 1.3808,
+      "step": 147
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7824983384024738,
+      "learning_rate": 1.9980047801673818e-05,
+      "loss": 1.2129,
+      "step": 148
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.398043761467571,
+      "learning_rate": 1.99793541026214e-05,
+      "loss": 1.3733,
+      "step": 149
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8082968843433833,
+      "learning_rate": 1.997864856193526e-05,
+      "loss": 1.28,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7404749537716926,
+      "learning_rate": 1.9977931180452606e-05,
+      "loss": 1.5369,
+      "step": 151
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7969696735345595,
+      "learning_rate": 1.997720195902469e-05,
+      "loss": 1.212,
+      "step": 152
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9712468566318817,
+      "learning_rate": 1.9976460898516817e-05,
+      "loss": 1.4887,
+      "step": 153
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.907806614272928,
+      "learning_rate": 1.9975707999808338e-05,
+      "loss": 0.9463,
+      "step": 154
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.811886254937919,
+      "learning_rate": 1.997494326379265e-05,
+      "loss": 1.2625,
+      "step": 155
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.683065317691926,
+      "learning_rate": 1.997416669137721e-05,
+      "loss": 1.4063,
+      "step": 156
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9342195788602383,
+      "learning_rate": 1.9973378283483497e-05,
+      "loss": 1.2317,
+      "step": 157
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0566508226867115,
+      "learning_rate": 1.997257804104705e-05,
+      "loss": 1.5606,
+      "step": 158
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7466452482653203,
+      "learning_rate": 1.9971765965017446e-05,
+      "loss": 1.1878,
+      "step": 159
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8540026932394673,
+      "learning_rate": 1.997094205635831e-05,
+      "loss": 1.1967,
+      "step": 160
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9430496046968224,
+      "learning_rate": 1.9970106316047295e-05,
+      "loss": 1.4906,
+      "step": 161
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7956958320844714,
+      "learning_rate": 1.996925874507611e-05,
+      "loss": 1.1189,
+      "step": 162
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8262642942936667,
+      "learning_rate": 1.9968399344450492e-05,
+      "loss": 1.2216,
+      "step": 163
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7853586729613815,
+      "learning_rate": 1.996752811519021e-05,
+      "loss": 1.1589,
+      "step": 164
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7708722064999287,
+      "learning_rate": 1.9966645058329087e-05,
+      "loss": 1.1762,
+      "step": 165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1591637642812604,
+      "learning_rate": 1.996575017491496e-05,
+      "loss": 1.4896,
+      "step": 166
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0648397532459584,
+      "learning_rate": 1.9964843466009714e-05,
+      "loss": 1.2251,
+      "step": 167
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8891993029791411,
+      "learning_rate": 1.9963924932689265e-05,
+      "loss": 1.6636,
+      "step": 168
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6155071831324752,
+      "learning_rate": 1.9962994576043555e-05,
+      "loss": 1.2315,
+      "step": 169
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3295362267335242,
+      "learning_rate": 1.9962052397176553e-05,
+      "loss": 1.5969,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9508743905799188,
+      "learning_rate": 1.9961098397206264e-05,
+      "loss": 0.895,
+      "step": 171
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9627719787605238,
+      "learning_rate": 1.996013257726472e-05,
+      "loss": 1.2296,
+      "step": 172
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5434337337348465,
+      "learning_rate": 1.9959154938497968e-05,
+      "loss": 1.3265,
+      "step": 173
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9130175218138912,
+      "learning_rate": 1.9958165482066094e-05,
+      "loss": 1.1398,
+      "step": 174
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9492412558370433,
+      "learning_rate": 1.99571642091432e-05,
+      "loss": 1.5817,
+      "step": 175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7227394388541064,
+      "learning_rate": 1.9956151120917405e-05,
+      "loss": 1.1478,
+      "step": 176
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0849776041006707,
+      "learning_rate": 1.9955126218590858e-05,
+      "loss": 1.5087,
+      "step": 177
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8591745621518452,
+      "learning_rate": 1.9954089503379716e-05,
+      "loss": 1.0655,
+      "step": 178
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6771456588680436,
+      "learning_rate": 1.9953040976514164e-05,
+      "loss": 1.4012,
+      "step": 179
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.030145014116347,
+      "learning_rate": 1.9951980639238398e-05,
+      "loss": 1.1233,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7337635549534972,
+      "learning_rate": 1.9950908492810624e-05,
+      "loss": 1.1237,
+      "step": 181
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1166181511368765,
+      "learning_rate": 1.994982453850307e-05,
+      "loss": 1.6505,
+      "step": 182
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8720923725376568,
+      "learning_rate": 1.9948728777601966e-05,
+      "loss": 1.2515,
+      "step": 183
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9894687550491981,
+      "learning_rate": 1.994762121140756e-05,
+      "loss": 1.3636,
+      "step": 184
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8329580913605024,
+      "learning_rate": 1.99465018412341e-05,
+      "loss": 1.1605,
+      "step": 185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8952137007734138,
+      "learning_rate": 1.994537066840985e-05,
+      "loss": 1.1283,
+      "step": 186
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9831994362570833,
+      "learning_rate": 1.9944227694277076e-05,
+      "loss": 1.3872,
+      "step": 187
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8771942070010402,
+      "learning_rate": 1.9943072920192042e-05,
+      "loss": 1.2486,
+      "step": 188
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0212832064137523,
+      "learning_rate": 1.994190634752502e-05,
+      "loss": 1.6159,
+      "step": 189
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1290199713422984,
+      "learning_rate": 1.9940727977660275e-05,
+      "loss": 0.904,
+      "step": 190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7652997152856574,
+      "learning_rate": 1.993953781199609e-05,
+      "loss": 1.3018,
+      "step": 191
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8364841672737937,
+      "learning_rate": 1.993833585194472e-05,
+      "loss": 1.5265,
+      "step": 192
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7859670513276467,
+      "learning_rate": 1.9937122098932428e-05,
+      "loss": 1.2024,
+      "step": 193
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8845681036747904,
+      "learning_rate": 1.9935896554399472e-05,
+      "loss": 1.3049,
+      "step": 194
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.714011863864167,
+      "learning_rate": 1.99346592198001e-05,
+      "loss": 1.196,
+      "step": 195
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9547307794253863,
+      "learning_rate": 1.9933410096602544e-05,
+      "loss": 1.4956,
+      "step": 196
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7416803252440998,
+      "learning_rate": 1.993214918628904e-05,
+      "loss": 1.1277,
+      "step": 197
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9879778400943108,
+      "learning_rate": 1.993087649035579e-05,
+      "loss": 0.8626,
+      "step": 198
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.100868145566014,
+      "learning_rate": 1.9929592010312997e-05,
+      "loss": 1.5132,
+      "step": 199
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.931056586802545,
+      "learning_rate": 1.9928295747684846e-05,
+      "loss": 1.2023,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7739596186029847,
+      "learning_rate": 1.992698770400949e-05,
+      "loss": 1.6429,
+      "step": 201
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7023077138633494,
+      "learning_rate": 1.992566788083908e-05,
+      "loss": 1.1781,
+      "step": 202
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8055161656325075,
+      "learning_rate": 1.9924336279739732e-05,
+      "loss": 1.2295,
+      "step": 203
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9841642913058554,
+      "learning_rate": 1.992299290229154e-05,
+      "loss": 1.4863,
+      "step": 204
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7124775460785799,
+      "learning_rate": 1.9921637750088577e-05,
+      "loss": 1.2179,
+      "step": 205
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0300672886883806,
+      "learning_rate": 1.9920270824738882e-05,
+      "loss": 1.0534,
+      "step": 206
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1733774201731577,
+      "learning_rate": 1.9918892127864463e-05,
+      "loss": 1.193,
+      "step": 207
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1972514049234146,
+      "learning_rate": 1.9917501661101307e-05,
+      "loss": 1.4427,
+      "step": 208
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0302049178052886,
+      "learning_rate": 1.991609942609936e-05,
+      "loss": 1.1171,
+      "step": 209
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.813663061107973,
+      "learning_rate": 1.9914685424522525e-05,
+      "loss": 1.1538,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2466815000483644,
+      "learning_rate": 1.9913259658048684e-05,
+      "loss": 1.4172,
+      "step": 211
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.719575383179546,
+      "learning_rate": 1.9911822128369668e-05,
+      "loss": 1.3696,
+      "step": 212
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7509348873970256,
+      "learning_rate": 1.991037283719126e-05,
+      "loss": 1.3294,
+      "step": 213
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0384969498814363,
+      "learning_rate": 1.990891178623322e-05,
+      "loss": 1.3021,
+      "step": 214
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.004485336589673,
+      "learning_rate": 1.990743897722925e-05,
+      "loss": 1.4312,
+      "step": 215
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1329726712321508,
+      "learning_rate": 1.9905954411926992e-05,
+      "loss": 0.9156,
+      "step": 216
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6942481586306068,
+      "learning_rate": 1.990445809208806e-05,
+      "loss": 1.07,
+      "step": 217
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6130392823935032,
+      "learning_rate": 1.990295001948801e-05,
+      "loss": 1.3394,
+      "step": 218
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.619756729399947,
+      "learning_rate": 1.9901430195916336e-05,
+      "loss": 1.0542,
+      "step": 219
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8324800692215251,
+      "learning_rate": 1.989989862317648e-05,
+      "loss": 1.5327,
+      "step": 220
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6778970979764347,
+      "learning_rate": 1.9898355303085833e-05,
+      "loss": 1.1163,
+      "step": 221
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7066908781405585,
+      "learning_rate": 1.989680023747571e-05,
+      "loss": 1.1004,
+      "step": 222
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9086253238775837,
+      "learning_rate": 1.989523342819138e-05,
+      "loss": 1.4821,
+      "step": 223
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5197560176610396,
+      "learning_rate": 1.9893654877092032e-05,
+      "loss": 0.8738,
+      "step": 224
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.774368390163999,
+      "learning_rate": 1.9892064586050804e-05,
+      "loss": 1.4269,
+      "step": 225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6890960255818757,
+      "learning_rate": 1.989046255695475e-05,
+      "loss": 1.2427,
+      "step": 226
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9306313416581069,
+      "learning_rate": 1.9888848791704863e-05,
+      "loss": 1.493,
+      "step": 227
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.676905136948933,
+      "learning_rate": 1.9887223292216058e-05,
+      "loss": 1.1269,
+      "step": 228
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7588492322358938,
+      "learning_rate": 1.988558606041717e-05,
+      "loss": 1.1696,
+      "step": 229
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0596065076626853,
+      "learning_rate": 1.9883937098250962e-05,
+      "loss": 1.5369,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6210411822124664,
+      "learning_rate": 1.988227640767412e-05,
+      "loss": 1.1485,
+      "step": 231
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8551370593875253,
+      "learning_rate": 1.988060399065723e-05,
+      "loss": 1.4189,
+      "step": 232
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8813585630703274,
+      "learning_rate": 1.9878919849184813e-05,
+      "loss": 1.1349,
+      "step": 233
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6390791704285876,
+      "learning_rate": 1.9877223985255293e-05,
+      "loss": 0.9899,
+      "step": 234
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.885332378922286,
+      "learning_rate": 1.9875516400881002e-05,
+      "loss": 1.3947,
+      "step": 235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7255839375648175,
+      "learning_rate": 1.9873797098088185e-05,
+      "loss": 1.0421,
+      "step": 236
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8490239575940799,
+      "learning_rate": 1.987206607891699e-05,
+      "loss": 1.4652,
+      "step": 237
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7714050296095142,
+      "learning_rate": 1.9870323345421457e-05,
+      "loss": 1.2254,
+      "step": 238
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8197049203174409,
+      "learning_rate": 1.986856889966955e-05,
+      "loss": 1.4167,
+      "step": 239
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8407503767999983,
+      "learning_rate": 1.986680274374311e-05,
+      "loss": 1.2038,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7280625282466607,
+      "learning_rate": 1.9865024879737885e-05,
+      "loss": 1.253,
+      "step": 241
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.998587961023149,
+      "learning_rate": 1.9863235309763505e-05,
+      "loss": 1.3299,
+      "step": 242
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6735963048006987,
+      "learning_rate": 1.9861434035943507e-05,
+      "loss": 1.0396,
+      "step": 243
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0174666080494899,
+      "learning_rate": 1.9859621060415302e-05,
+      "loss": 1.6098,
+      "step": 244
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8098181621444368,
+      "learning_rate": 1.985779638533019e-05,
+      "loss": 1.154,
+      "step": 245
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8929040711237198,
+      "learning_rate": 1.9855960012853354e-05,
+      "loss": 1.4914,
+      "step": 246
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8145646177884597,
+      "learning_rate": 1.985411194516386e-05,
+      "loss": 1.0293,
+      "step": 247
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7624849389308673,
+      "learning_rate": 1.9852252184454646e-05,
+      "loss": 1.083,
+      "step": 248
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.092748205796077,
+      "learning_rate": 1.9850380732932538e-05,
+      "loss": 1.5549,
+      "step": 249
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6523500669553071,
+      "learning_rate": 1.9848497592818214e-05,
+      "loss": 1.2373,
+      "step": 250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2727781817528012,
+      "learning_rate": 1.9846602766346238e-05,
+      "loss": 1.0709,
+      "step": 251
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7663573388716122,
+      "learning_rate": 1.9844696255765036e-05,
+      "loss": 1.1981,
+      "step": 252
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7245401588018158,
+      "learning_rate": 1.9842778063336898e-05,
+      "loss": 1.1728,
+      "step": 253
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8919422907647013,
+      "learning_rate": 1.9840848191337976e-05,
+      "loss": 1.364,
+      "step": 254
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8096586450460994,
+      "learning_rate": 1.9838906642058283e-05,
+      "loss": 1.2099,
+      "step": 255
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8018839376172252,
+      "learning_rate": 1.9836953417801686e-05,
+      "loss": 1.354,
+      "step": 256
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1486286329269262,
+      "learning_rate": 1.9834988520885912e-05,
+      "loss": 1.2136,
+      "step": 257
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9612876835389118,
+      "learning_rate": 1.9833011953642525e-05,
+      "loss": 1.2728,
+      "step": 258
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8900467694313408,
+      "learning_rate": 1.983102371841695e-05,
+      "loss": 1.1913,
+      "step": 259
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1611162181154129,
+      "learning_rate": 1.9829023817568455e-05,
+      "loss": 0.985,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2033763966392386,
+      "learning_rate": 1.982701225347015e-05,
+      "loss": 1.3902,
+      "step": 261
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7390258519100612,
+      "learning_rate": 1.9824989028508975e-05,
+      "loss": 1.1546,
+      "step": 262
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8981036260864895,
+      "learning_rate": 1.9822954145085723e-05,
+      "loss": 1.3994,
+      "step": 263
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7988980989199648,
+      "learning_rate": 1.9820907605615013e-05,
+      "loss": 1.0695,
+      "step": 264
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6828609112406419,
+      "learning_rate": 1.9818849412525294e-05,
+      "loss": 1.1697,
+      "step": 265
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2862305163713177,
+      "learning_rate": 1.981677956825885e-05,
+      "loss": 1.4442,
+      "step": 266
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7254070446516175,
+      "learning_rate": 1.9814698075271776e-05,
+      "loss": 1.3093,
+      "step": 267
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1547766242211865,
+      "learning_rate": 1.9812604936034004e-05,
+      "loss": 1.3371,
+      "step": 268
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9413236765547888,
+      "learning_rate": 1.9810500153029285e-05,
+      "loss": 0.8613,
+      "step": 269
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9788019631627247,
+      "learning_rate": 1.9808383728755176e-05,
+      "loss": 1.5232,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8632131875740335,
+      "learning_rate": 1.980625566572306e-05,
+      "loss": 1.1041,
+      "step": 271
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7854854715178329,
+      "learning_rate": 1.9804115966458118e-05,
+      "loss": 1.2017,
+      "step": 272
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.062496277095639,
+      "learning_rate": 1.9801964633499352e-05,
+      "loss": 1.4054,
+      "step": 273
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7560548522052498,
+      "learning_rate": 1.9799801669399556e-05,
+      "loss": 1.0545,
+      "step": 274
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3327390590809893,
+      "learning_rate": 1.9797627076725338e-05,
+      "loss": 1.3498,
+      "step": 275
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7977426691763201,
+      "learning_rate": 1.9795440858057094e-05,
+      "loss": 1.2521,
+      "step": 276
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7887563734768773,
+      "learning_rate": 1.979324301598902e-05,
+      "loss": 1.0741,
+      "step": 277
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7080803968602234,
+      "learning_rate": 1.979103355312911e-05,
+      "loss": 1.3167,
+      "step": 278
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7603978502334926,
+      "learning_rate": 1.9788812472099138e-05,
+      "loss": 1.0899,
+      "step": 279
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8402770451353578,
+      "learning_rate": 1.9786579775534663e-05,
+      "loss": 1.4511,
+      "step": 280
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.741982479923261,
+      "learning_rate": 1.978433546608504e-05,
+      "loss": 1.1675,
+      "step": 281
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9463560533756077,
+      "learning_rate": 1.9782079546413396e-05,
+      "loss": 1.4056,
+      "step": 282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7632400251883276,
+      "learning_rate": 1.9779812019196632e-05,
+      "loss": 1.3294,
+      "step": 283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6727880050709257,
+      "learning_rate": 1.9777532887125425e-05,
+      "loss": 1.0752,
+      "step": 284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.928027292088074,
+      "learning_rate": 1.9775242152904228e-05,
+      "loss": 1.5163,
+      "step": 285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.101987857569684,
+      "learning_rate": 1.9772939819251247e-05,
+      "loss": 1.0477,
+      "step": 286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.812540744196496,
+      "learning_rate": 1.9770625888898477e-05,
+      "loss": 0.945,
+      "step": 287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6537976872506521,
+      "learning_rate": 1.9768300364591643e-05,
+      "loss": 1.3631,
+      "step": 288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8975535633190511,
+      "learning_rate": 1.9765963249090254e-05,
+      "loss": 1.4659,
+      "step": 289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7822072351967468,
+      "learning_rate": 1.9763614545167552e-05,
+      "loss": 1.1552,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7855941461706761,
+      "learning_rate": 1.9761254255610553e-05,
+      "loss": 1.1669,
+      "step": 291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6420185913017166,
+      "learning_rate": 1.9758882383219995e-05,
+      "loss": 1.234,
+      "step": 292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6708416907062917,
+      "learning_rate": 1.975649893081038e-05,
+      "loss": 1.0921,
+      "step": 293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1472500838362973,
+      "learning_rate": 1.9754103901209943e-05,
+      "loss": 1.4345,
+      "step": 294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6875030940194743,
+      "learning_rate": 1.975169729726066e-05,
+      "loss": 1.1848,
+      "step": 295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0058098989056334,
+      "learning_rate": 1.9749279121818235e-05,
+      "loss": 1.0437,
+      "step": 296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6528890275546898,
+      "learning_rate": 1.9746849377752113e-05,
+      "loss": 1.1588,
+      "step": 297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7730305783123499,
+      "learning_rate": 1.9744408067945455e-05,
+      "loss": 1.0863,
+      "step": 298
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9015917149040965,
+      "learning_rate": 1.974195519529515e-05,
+      "loss": 1.4946,
+      "step": 299
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8368335356520227,
+      "learning_rate": 1.9739490762711812e-05,
+      "loss": 1.086,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.792642793538534,
+      "learning_rate": 1.9737014773119774e-05,
+      "loss": 1.3063,
+      "step": 301
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6645874825156662,
+      "learning_rate": 1.973452722945707e-05,
+      "loss": 1.0822,
+      "step": 302
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7515654207152128,
+      "learning_rate": 1.9732028134675457e-05,
+      "loss": 1.254,
+      "step": 303
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.218772409924033,
+      "learning_rate": 1.972951749174039e-05,
+      "loss": 1.0565,
+      "step": 304
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6008412117569835,
+      "learning_rate": 1.9726995303631038e-05,
+      "loss": 1.1639,
+      "step": 305
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9793450269403292,
+      "learning_rate": 1.9724461573340258e-05,
+      "loss": 1.3793,
+      "step": 306
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7928901025225913,
+      "learning_rate": 1.9721916303874607e-05,
+      "loss": 1.2444,
+      "step": 307
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.855210201738354,
+      "learning_rate": 1.971935949825434e-05,
+      "loss": 1.2888,
+      "step": 308
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7454384754232674,
+      "learning_rate": 1.971679115951339e-05,
+      "loss": 1.0718,
+      "step": 309
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6773584504749054,
+      "learning_rate": 1.971421129069939e-05,
+      "loss": 1.2326,
+      "step": 310
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.018496728908696,
+      "learning_rate": 1.9711619894873637e-05,
+      "loss": 1.6283,
+      "step": 311
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8369458471416004,
+      "learning_rate": 1.9709016975111123e-05,
+      "loss": 1.0434,
+      "step": 312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6253018762375182,
+      "learning_rate": 1.9706402534500504e-05,
+      "loss": 1.0001,
+      "step": 313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7132096524901521,
+      "learning_rate": 1.9703776576144106e-05,
+      "loss": 1.1822,
+      "step": 314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9520785508745445,
+      "learning_rate": 1.9701139103157937e-05,
+      "loss": 1.3926,
+      "step": 315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6442959669317916,
+      "learning_rate": 1.9698490118671647e-05,
+      "loss": 1.2096,
+      "step": 316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7254304658074563,
+      "learning_rate": 1.9695829625828563e-05,
+      "loss": 1.1583,
+      "step": 317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.201713121582505,
+      "learning_rate": 1.9693157627785654e-05,
+      "loss": 1.3069,
+      "step": 318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7415770323784497,
+      "learning_rate": 1.9690474127713554e-05,
+      "loss": 1.1545,
+      "step": 319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7694229224579856,
+      "learning_rate": 1.9687779128796537e-05,
+      "loss": 1.3122,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6867493830790934,
+      "learning_rate": 1.9685072634232524e-05,
+      "loss": 1.2123,
+      "step": 321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.623453755370031,
+      "learning_rate": 1.9682354647233078e-05,
+      "loss": 0.9277,
+      "step": 322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9151203510190116,
+      "learning_rate": 1.96796251710234e-05,
+      "loss": 1.4697,
+      "step": 323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7196157802516311,
+      "learning_rate": 1.9676884208842323e-05,
+      "loss": 1.1326,
+      "step": 324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7130299232915063,
+      "learning_rate": 1.9674131763942305e-05,
+      "loss": 1.3062,
+      "step": 325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7657361508458136,
+      "learning_rate": 1.9671367839589438e-05,
+      "loss": 1.1602,
+      "step": 326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0513370709681777,
+      "learning_rate": 1.9668592439063426e-05,
+      "loss": 1.2088,
+      "step": 327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9064422551032997,
+      "learning_rate": 1.9665805565657602e-05,
+      "loss": 1.3538,
+      "step": 328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5781338567930846,
+      "learning_rate": 1.9663007222678904e-05,
+      "loss": 1.2085,
+      "step": 329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7547599725124731,
+      "learning_rate": 1.966019741344788e-05,
+      "loss": 1.3987,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8475253878498201,
+      "learning_rate": 1.9657376141298692e-05,
+      "loss": 0.8453,
+      "step": 331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9588673474343349,
+      "learning_rate": 1.96545434095791e-05,
+      "loss": 1.3244,
+      "step": 332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9901473516748829,
+      "learning_rate": 1.9651699221650454e-05,
+      "loss": 1.3275,
+      "step": 333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8640726624209789,
+      "learning_rate": 1.964884358088771e-05,
+      "loss": 1.4516,
+      "step": 334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6713806686020639,
+      "learning_rate": 1.9645976490679402e-05,
+      "loss": 0.9834,
+      "step": 335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7060205964641068,
+      "learning_rate": 1.964309795442767e-05,
+      "loss": 1.1905,
+      "step": 336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.882003110004651,
+      "learning_rate": 1.9640207975548216e-05,
+      "loss": 1.3282,
+      "step": 337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7056877620018042,
+      "learning_rate": 1.9637306557470325e-05,
+      "loss": 1.0693,
+      "step": 338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4359970920448004,
+      "learning_rate": 1.963439370363686e-05,
+      "loss": 1.0379,
+      "step": 339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6918790389959811,
+      "learning_rate": 1.9631469417504253e-05,
+      "loss": 1.114,
+      "step": 340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5659749778015349,
+      "learning_rate": 1.96285337025425e-05,
+      "loss": 1.2222,
+      "step": 341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0102332638165619,
+      "learning_rate": 1.962558656223516e-05,
+      "loss": 1.4535,
+      "step": 342
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7458658416853812,
+      "learning_rate": 1.9622628000079347e-05,
+      "loss": 1.2027,
+      "step": 343
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9089051600078369,
+      "learning_rate": 1.9619658019585736e-05,
+      "loss": 1.4127,
+      "step": 344
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7763010318055139,
+      "learning_rate": 1.9616676624278535e-05,
+      "loss": 1.1267,
+      "step": 345
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0743926059592717,
+      "learning_rate": 1.9613683817695516e-05,
+      "loss": 1.4508,
+      "step": 346
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7031087542938487,
+      "learning_rate": 1.9610679603387977e-05,
+      "loss": 1.2252,
+      "step": 347
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8140503035622717,
+      "learning_rate": 1.960766398492076e-05,
+      "loss": 1.1585,
+      "step": 348
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.234485951843342,
+      "learning_rate": 1.9604636965872243e-05,
+      "loss": 1.0476,
+      "step": 349
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9401530148633382,
+      "learning_rate": 1.9601598549834318e-05,
+      "loss": 1.1011,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7694998281157277,
+      "learning_rate": 1.9598548740412414e-05,
+      "loss": 1.3427,
+      "step": 351
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6271902017252798,
+      "learning_rate": 1.959548754122548e-05,
+      "loss": 1.1057,
+      "step": 352
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9910216640912048,
+      "learning_rate": 1.959241495590597e-05,
+      "loss": 1.1915,
+      "step": 353
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9747004832272841,
+      "learning_rate": 1.958933098809985e-05,
+      "loss": 1.4756,
+      "step": 354
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5278011096088241,
+      "learning_rate": 1.9586235641466607e-05,
+      "loss": 1.177,
+      "step": 355
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8685744668392804,
+      "learning_rate": 1.9583128919679218e-05,
+      "loss": 1.3711,
+      "step": 356
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8524999984371696,
+      "learning_rate": 1.958001082642415e-05,
+      "loss": 0.8924,
+      "step": 357
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2548015752102448,
+      "learning_rate": 1.957688136540139e-05,
+      "loss": 1.3406,
+      "step": 358
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8178901305239519,
+      "learning_rate": 1.9573740540324385e-05,
+      "loss": 0.9532,
+      "step": 359
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6515458235531276,
+      "learning_rate": 1.9570588354920085e-05,
+      "loss": 1.1555,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9233643579229581,
+      "learning_rate": 1.9567424812928918e-05,
+      "loss": 1.4461,
+      "step": 361
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9126404973268626,
+      "learning_rate": 1.9564249918104777e-05,
+      "loss": 1.1104,
+      "step": 362
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7692041412877402,
+      "learning_rate": 1.9561063674215038e-05,
+      "loss": 1.3565,
+      "step": 363
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7565991458998549,
+      "learning_rate": 1.955786608504054e-05,
+      "loss": 1.2892,
+      "step": 364
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9239617593643333,
+      "learning_rate": 1.955465715437559e-05,
+      "loss": 1.144,
+      "step": 365
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8013394695843024,
+      "learning_rate": 1.9551436886027938e-05,
+      "loss": 1.0164,
+      "step": 366
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6566661877030538,
+      "learning_rate": 1.9548205283818803e-05,
+      "loss": 1.1902,
+      "step": 367
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.838998795989866,
+      "learning_rate": 1.9544962351582844e-05,
+      "loss": 1.4504,
+      "step": 368
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.912213305527829,
+      "learning_rate": 1.954170809316817e-05,
+      "loss": 1.169,
+      "step": 369
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8971074576345622,
+      "learning_rate": 1.953844251243633e-05,
+      "loss": 1.3087,
+      "step": 370
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7723983617281817,
+      "learning_rate": 1.9535165613262296e-05,
+      "loss": 0.9865,
+      "step": 371
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7759445052326314,
+      "learning_rate": 1.953187739953449e-05,
+      "loss": 1.0454,
+      "step": 372
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0145831186636523,
+      "learning_rate": 1.952857787515474e-05,
+      "loss": 1.2782,
+      "step": 373
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7502960166033179,
+      "learning_rate": 1.9525267044038318e-05,
+      "loss": 1.265,
+      "step": 374
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8065792375655307,
+      "learning_rate": 1.9521944910113887e-05,
+      "loss": 1.3283,
+      "step": 375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0583370276096757,
+      "learning_rate": 1.951861147732354e-05,
+      "loss": 0.7999,
+      "step": 376
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8379068202262686,
+      "learning_rate": 1.951526674962278e-05,
+      "loss": 1.2792,
+      "step": 377
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.029711807601846,
+      "learning_rate": 1.951191073098049e-05,
+      "loss": 1.3751,
+      "step": 378
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9975696791353594,
+      "learning_rate": 1.9508543425378983e-05,
+      "loss": 1.1439,
+      "step": 379
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6761821898421084,
+      "learning_rate": 1.9505164836813938e-05,
+      "loss": 1.2797,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0485799528423485,
+      "learning_rate": 1.950177496929443e-05,
+      "loss": 1.0962,
+      "step": 381
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.056801960400868,
+      "learning_rate": 1.949837382684293e-05,
+      "loss": 1.3174,
+      "step": 382
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.15406276910975,
+      "learning_rate": 1.9494961413495273e-05,
+      "loss": 0.8314,
+      "step": 383
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6462127765677035,
+      "learning_rate": 1.9491537733300677e-05,
+      "loss": 1.1923,
+      "step": 384
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1787353478778047,
+      "learning_rate": 1.9488102790321726e-05,
+      "loss": 1.332,
+      "step": 385
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7038797678585512,
+      "learning_rate": 1.9484656588634362e-05,
+      "loss": 1.0979,
+      "step": 386
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9006008523500977,
+      "learning_rate": 1.94811991323279e-05,
+      "loss": 1.5793,
+      "step": 387
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6212157292600377,
+      "learning_rate": 1.947773042550501e-05,
+      "loss": 1.1902,
+      "step": 388
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7410880440417607,
+      "learning_rate": 1.947425047228169e-05,
+      "loss": 1.1356,
+      "step": 389
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8038347234132924,
+      "learning_rate": 1.947075927678731e-05,
+      "loss": 1.3641,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5716112009967688,
+      "learning_rate": 1.946725684316456e-05,
+      "loss": 1.1648,
+      "step": 391
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1235529633557455,
+      "learning_rate": 1.946374317556948e-05,
+      "loss": 0.9401,
+      "step": 392
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.771301385827115,
+      "learning_rate": 1.946021827817144e-05,
+      "loss": 1.1305,
+      "step": 393
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0660393999808895,
+      "learning_rate": 1.945668215515312e-05,
+      "loss": 1.3615,
+      "step": 394
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8492572627255481,
+      "learning_rate": 1.9453134810710536e-05,
+      "loss": 0.9957,
+      "step": 395
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9559041671447255,
+      "learning_rate": 1.9449576249053008e-05,
+      "loss": 1.3615,
+      "step": 396
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7810313697203073,
+      "learning_rate": 1.9446006474403177e-05,
+      "loss": 1.1142,
+      "step": 397
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6976034036830887,
+      "learning_rate": 1.9442425490996987e-05,
+      "loss": 1.2947,
+      "step": 398
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8842586041169124,
+      "learning_rate": 1.9438833303083677e-05,
+      "loss": 1.2749,
+      "step": 399
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7093797942471453,
+      "learning_rate": 1.9435229914925785e-05,
+      "loss": 1.2181,
+      "step": 400
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9450255366208501,
+      "learning_rate": 1.943161533079914e-05,
+      "loss": 1.3139,
+      "step": 401
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.93090415684721,
+      "learning_rate": 1.9427989554992854e-05,
+      "loss": 0.8149,
+      "step": 402
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5382348131645471,
+      "learning_rate": 1.9424352591809322e-05,
+      "loss": 1.1556,
+      "step": 403
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9898042440246739,
+      "learning_rate": 1.942070444556421e-05,
+      "loss": 1.2433,
+      "step": 404
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7891219898151527,
+      "learning_rate": 1.941704512058646e-05,
+      "loss": 1.0548,
+      "step": 405
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7398662340039216,
+      "learning_rate": 1.9413374621218275e-05,
+      "loss": 1.3924,
+      "step": 406
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7333489791641546,
+      "learning_rate": 1.9409692951815117e-05,
+      "loss": 1.0441,
+      "step": 407
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9489158977077279,
+      "learning_rate": 1.94060001167457e-05,
+      "loss": 1.4258,
+      "step": 408
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7426697020930576,
+      "learning_rate": 1.9402296120392003e-05,
+      "loss": 1.2232,
+      "step": 409
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2292510163991663,
+      "learning_rate": 1.9398580967149222e-05,
+      "loss": 0.8214,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.821101899301408,
+      "learning_rate": 1.9394854661425816e-05,
+      "loss": 1.413,
+      "step": 411
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7756357045739198,
+      "learning_rate": 1.939111720764347e-05,
+      "loss": 1.1421,
+      "step": 412
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7898121352086022,
+      "learning_rate": 1.938736861023709e-05,
+      "loss": 1.4255,
+      "step": 413
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6797048119982116,
+      "learning_rate": 1.9383608873654817e-05,
+      "loss": 1.0743,
+      "step": 414
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7427140655506289,
+      "learning_rate": 1.9379838002357995e-05,
+      "loss": 1.1509,
+      "step": 415
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0320932718591904,
+      "learning_rate": 1.93760560008212e-05,
+      "loss": 1.4114,
+      "step": 416
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7982201716054931,
+      "learning_rate": 1.9372262873532207e-05,
+      "loss": 1.1277,
+      "step": 417
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7866154392336399,
+      "learning_rate": 1.936845862499198e-05,
+      "loss": 1.2904,
+      "step": 418
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7612314245209197,
+      "learning_rate": 1.9364643259714696e-05,
+      "loss": 1.0697,
+      "step": 419
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0466598191463565,
+      "learning_rate": 1.936081678222772e-05,
+      "loss": 1.1094,
+      "step": 420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9823373169695131,
+      "learning_rate": 1.9356979197071597e-05,
+      "loss": 1.1696,
+      "step": 421
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7543389898567571,
+      "learning_rate": 1.935313050880006e-05,
+      "loss": 0.9725,
+      "step": 422
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8074595526344972,
+      "learning_rate": 1.934927072198001e-05,
+      "loss": 1.3851,
+      "step": 423
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7367737712965701,
+      "learning_rate": 1.9345399841191527e-05,
+      "loss": 1.1308,
+      "step": 424
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8876151742182143,
+      "learning_rate": 1.9341517871027843e-05,
+      "loss": 1.2721,
+      "step": 425
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9013948338339061,
+      "learning_rate": 1.933762481609536e-05,
+      "loss": 1.2725,
+      "step": 426
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9675064918242701,
+      "learning_rate": 1.9333720681013618e-05,
+      "loss": 1.3362,
+      "step": 427
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7672579695456346,
+      "learning_rate": 1.932980547041533e-05,
+      "loss": 1.0462,
+      "step": 428
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6622932738171964,
+      "learning_rate": 1.9325879188946328e-05,
+      "loss": 1.0172,
+      "step": 429
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.007681601538217,
+      "learning_rate": 1.9321941841265596e-05,
+      "loss": 1.4616,
+      "step": 430
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7083915502429695,
+      "learning_rate": 1.9317993432045236e-05,
+      "loss": 1.1801,
+      "step": 431
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7206029071207327,
+      "learning_rate": 1.9314033965970483e-05,
+      "loss": 1.2917,
+      "step": 432
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7033523337663382,
+      "learning_rate": 1.9310063447739697e-05,
+      "loss": 1.05,
+      "step": 433
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7225919094401517,
+      "learning_rate": 1.9306081882064346e-05,
+      "loss": 1.1253,
+      "step": 434
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0701864934697372,
+      "learning_rate": 1.9302089273669002e-05,
+      "loss": 1.4468,
+      "step": 435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5935792071888989,
+      "learning_rate": 1.929808562729136e-05,
+      "loss": 0.9938,
+      "step": 436
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.878672695636262,
+      "learning_rate": 1.9294070947682186e-05,
+      "loss": 1.2997,
+      "step": 437
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6804027726547095,
+      "learning_rate": 1.929004523960536e-05,
+      "loss": 1.2468,
+      "step": 438
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8433024048794079,
+      "learning_rate": 1.9286008507837838e-05,
+      "loss": 1.4001,
+      "step": 439
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6873639768130565,
+      "learning_rate": 1.928196075716966e-05,
+      "loss": 1.0525,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6858757609637378,
+      "learning_rate": 1.9277901992403943e-05,
+      "loss": 1.1598,
+      "step": 441
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8414105783727663,
+      "learning_rate": 1.9273832218356867e-05,
+      "loss": 1.4507,
+      "step": 442
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7085542026227669,
+      "learning_rate": 1.926975143985768e-05,
+      "loss": 0.989,
+      "step": 443
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.778223066341804,
+      "learning_rate": 1.926565966174869e-05,
+      "loss": 1.187,
+      "step": 444
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8887639508000164,
+      "learning_rate": 1.9261556888885257e-05,
+      "loss": 1.1373,
+      "step": 445
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7195991341979856,
+      "learning_rate": 1.9257443126135776e-05,
+      "loss": 1.0519,
+      "step": 446
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6636519236825621,
+      "learning_rate": 1.9253318378381703e-05,
+      "loss": 1.0628,
+      "step": 447
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.677750336351,
+      "learning_rate": 1.924918265051752e-05,
+      "loss": 1.1516,
+      "step": 448
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9492035017746221,
+      "learning_rate": 1.924503594745072e-05,
+      "loss": 1.248,
+      "step": 449
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7968508756227021,
+      "learning_rate": 1.9240878274101854e-05,
+      "loss": 1.0981,
+      "step": 450
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6133219525879076,
+      "learning_rate": 1.9236709635404462e-05,
+      "loss": 1.1659,
+      "step": 451
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7565265788037104,
+      "learning_rate": 1.923253003630511e-05,
+      "loss": 1.3055,
+      "step": 452
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6168960688307307,
+      "learning_rate": 1.9228339481763365e-05,
+      "loss": 1.2585,
+      "step": 453
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8782634132215074,
+      "learning_rate": 1.9224137976751797e-05,
+      "loss": 0.8828,
+      "step": 454
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7152956113344424,
+      "learning_rate": 1.9219925526255956e-05,
+      "loss": 1.1783,
+      "step": 455
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8404195168458753,
+      "learning_rate": 1.921570213527441e-05,
+      "loss": 1.2763,
+      "step": 456
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7921022219418018,
+      "learning_rate": 1.921146780881867e-05,
+      "loss": 1.0831,
+      "step": 457
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8268636953714084,
+      "learning_rate": 1.9207222551913263e-05,
+      "loss": 1.1777,
+      "step": 458
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7997534972242479,
+      "learning_rate": 1.9202966369595653e-05,
+      "loss": 1.3435,
+      "step": 459
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7873678148180306,
+      "learning_rate": 1.919869926691629e-05,
+      "loss": 1.041,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8317078158558062,
+      "learning_rate": 1.9194421248938576e-05,
+      "loss": 1.2488,
+      "step": 461
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6474476858806775,
+      "learning_rate": 1.9190132320738856e-05,
+      "loss": 1.2292,
+      "step": 462
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2035435472536058,
+      "learning_rate": 1.9185832487406433e-05,
+      "loss": 1.0487,
+      "step": 463
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7172962552169628,
+      "learning_rate": 1.918152175404355e-05,
+      "loss": 1.1392,
+      "step": 464
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7776206676728886,
+      "learning_rate": 1.917720012576537e-05,
+      "loss": 1.0809,
+      "step": 465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8961394721339129,
+      "learning_rate": 1.9172867607700004e-05,
+      "loss": 1.2545,
+      "step": 466
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7209949728367009,
+      "learning_rate": 1.9168524204988475e-05,
+      "loss": 1.1254,
+      "step": 467
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9110485887797789,
+      "learning_rate": 1.9164169922784717e-05,
+      "loss": 1.3808,
+      "step": 468
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8856345088673786,
+      "learning_rate": 1.9159804766255583e-05,
+      "loss": 1.1509,
+      "step": 469
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.830476959738851,
+      "learning_rate": 1.915542874058082e-05,
+      "loss": 1.3631,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7010698001530615,
+      "learning_rate": 1.9151041850953086e-05,
+      "loss": 1.0734,
+      "step": 471
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7459072677488164,
+      "learning_rate": 1.9146644102577917e-05,
+      "loss": 1.024,
+      "step": 472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3201480256908966,
+      "learning_rate": 1.914223550067374e-05,
+      "loss": 0.8965,
+      "step": 473
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8494721794095446,
+      "learning_rate": 1.913781605047186e-05,
+      "loss": 1.3481,
+      "step": 474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8843575007277853,
+      "learning_rate": 1.9133385757216458e-05,
+      "loss": 1.3806,
+      "step": 475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7047289879973072,
+      "learning_rate": 1.9128944626164576e-05,
+      "loss": 1.1794,
+      "step": 476
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7095527844965216,
+      "learning_rate": 1.912449266258612e-05,
+      "loss": 1.0268,
+      "step": 477
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6712581487771736,
+      "learning_rate": 1.912002987176385e-05,
+      "loss": 1.2393,
+      "step": 478
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1301264859906577,
+      "learning_rate": 1.9115556258993368e-05,
+      "loss": 1.0611,
+      "step": 479
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2278830678148989,
+      "learning_rate": 1.911107182958313e-05,
+      "loss": 1.3699,
+      "step": 480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7879692240581528,
+      "learning_rate": 1.910657658885441e-05,
+      "loss": 1.1699,
+      "step": 481
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4393915656797696,
+      "learning_rate": 1.910207054214133e-05,
+      "loss": 0.9921,
+      "step": 482
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8356519038651575,
+      "learning_rate": 1.9097553694790816e-05,
+      "loss": 1.1616,
+      "step": 483
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9048596932840296,
+      "learning_rate": 1.9093026052162623e-05,
+      "loss": 1.0326,
+      "step": 484
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8299331427490336,
+      "learning_rate": 1.9088487619629314e-05,
+      "loss": 1.4401,
+      "step": 485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7082737668333174,
+      "learning_rate": 1.9083938402576246e-05,
+      "loss": 1.0571,
+      "step": 486
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1221900767981918,
+      "learning_rate": 1.907937840640159e-05,
+      "loss": 1.3002,
+      "step": 487
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7082112233732729,
+      "learning_rate": 1.9074807636516286e-05,
+      "loss": 1.0462,
+      "step": 488
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.011748084419257,
+      "learning_rate": 1.907022609834408e-05,
+      "loss": 1.4842,
+      "step": 489
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.993792427752119,
+      "learning_rate": 1.9065633797321484e-05,
+      "loss": 0.8652,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7383116642018505,
+      "learning_rate": 1.9061030738897778e-05,
+      "loss": 1.0938,
+      "step": 491
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.969952764604425,
+      "learning_rate": 1.9056416928535016e-05,
+      "loss": 1.2982,
+      "step": 492
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8420184123652142,
+      "learning_rate": 1.905179237170801e-05,
+      "loss": 1.293,
+      "step": 493
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7360869682230602,
+      "learning_rate": 1.904715707390431e-05,
+      "loss": 1.1625,
+      "step": 494
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.022980394319569,
+      "learning_rate": 1.9042511040624235e-05,
+      "loss": 1.1389,
+      "step": 495
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7718143110585082,
+      "learning_rate": 1.903785427738082e-05,
+      "loss": 1.0706,
+      "step": 496
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0219287237777723,
+      "learning_rate": 1.9033186789699846e-05,
+      "loss": 1.5274,
+      "step": 497
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7474576823824255,
+      "learning_rate": 1.902850858311981e-05,
+      "loss": 1.1008,
+      "step": 498
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7064032943313605,
+      "learning_rate": 1.902381966319194e-05,
+      "loss": 0.8876,
+      "step": 499
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7859925693921797,
+      "learning_rate": 1.9019120035480174e-05,
+      "loss": 1.058,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1558847069812817,
+      "learning_rate": 1.9014409705561142e-05,
+      "loss": 1.3944,
+      "step": 501
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.630069196662396,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 1.1739,
+      "step": 502
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6853356558451542,
+      "learning_rate": 1.9004956961471355e-05,
+      "loss": 1.1256,
+      "step": 503
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9353298826550004,
+      "learning_rate": 1.9000214558517343e-05,
+      "loss": 1.2193,
+      "step": 504
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7702955411335701,
+      "learning_rate": 1.899546147578956e-05,
+      "loss": 1.1885,
+      "step": 505
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7060572411419763,
+      "learning_rate": 1.8990697718928084e-05,
+      "loss": 1.3013,
+      "step": 506
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5618180426117285,
+      "learning_rate": 1.8985923293585638e-05,
+      "loss": 0.9268,
+      "step": 507
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8819411243712292,
+      "learning_rate": 1.898113820542763e-05,
+      "loss": 1.36,
+      "step": 508
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8421537842604641,
+      "learning_rate": 1.8976342460132097e-05,
+      "loss": 1.1084,
+      "step": 509
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7485683255978856,
+      "learning_rate": 1.8971536063389745e-05,
+      "loss": 1.0838,
+      "step": 510
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7969832189839798,
+      "learning_rate": 1.89667190209039e-05,
+      "loss": 1.2347,
+      "step": 511
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6784094338540485,
+      "learning_rate": 1.8961891338390534e-05,
+      "loss": 1.209,
+      "step": 512
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9001770097822541,
+      "learning_rate": 1.895705302157824e-05,
+      "loss": 1.3682,
+      "step": 513
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7152275864603884,
+      "learning_rate": 1.8952204076208224e-05,
+      "loss": 1.0689,
+      "step": 514
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5435734165483932,
+      "learning_rate": 1.894734450803431e-05,
+      "loss": 1.1868,
+      "step": 515
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7144335234430003,
+      "learning_rate": 1.8942474322822924e-05,
+      "loss": 1.3171,
+      "step": 516
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6596860586964988,
+      "learning_rate": 1.8937593526353098e-05,
+      "loss": 0.778,
+      "step": 517
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0301139019704448,
+      "learning_rate": 1.8932702124416446e-05,
+      "loss": 1.3205,
+      "step": 518
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7293470756377571,
+      "learning_rate": 1.892780012281717e-05,
+      "loss": 1.2217,
+      "step": 519
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8796016304309106,
+      "learning_rate": 1.8922887527372053e-05,
+      "loss": 1.3595,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7191656896981602,
+      "learning_rate": 1.891796434391044e-05,
+      "loss": 0.9915,
+      "step": 521
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8480546214451106,
+      "learning_rate": 1.8913030578274254e-05,
+      "loss": 1.1641,
+      "step": 522
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9073100368213134,
+      "learning_rate": 1.8908086236317958e-05,
+      "loss": 1.3364,
+      "step": 523
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7646345606501418,
+      "learning_rate": 1.890313132390858e-05,
+      "loss": 1.0784,
+      "step": 524
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7912109635816573,
+      "learning_rate": 1.889816584692568e-05,
+      "loss": 1.0174,
+      "step": 525
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7107030806492641,
+      "learning_rate": 1.8893189811261364e-05,
+      "loss": 0.9809,
+      "step": 526
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8236254312289656,
+      "learning_rate": 1.8888203222820255e-05,
+      "loss": 1.2577,
+      "step": 527
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7942590760923528,
+      "learning_rate": 1.8883206087519512e-05,
+      "loss": 1.1423,
+      "step": 528
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6038964558321538,
+      "learning_rate": 1.88781984112888e-05,
+      "loss": 1.2141,
+      "step": 529
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7749920720820082,
+      "learning_rate": 1.887318020007029e-05,
+      "loss": 1.3479,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6960824536969341,
+      "learning_rate": 1.886815145981866e-05,
+      "loss": 1.0195,
+      "step": 531
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.044378409100927,
+      "learning_rate": 1.886311219650108e-05,
+      "loss": 1.3639,
+      "step": 532
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8965623853431665,
+      "learning_rate": 1.8858062416097208e-05,
+      "loss": 1.0827,
+      "step": 533
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7144618209932374,
+      "learning_rate": 1.885300212459918e-05,
+      "loss": 1.2047,
+      "step": 534
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.120091235192801,
+      "learning_rate": 1.8847931328011602e-05,
+      "loss": 0.9474,
+      "step": 535
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7744643418870258,
+      "learning_rate": 1.884285003235155e-05,
+      "loss": 1.0376,
+      "step": 536
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9220197772461103,
+      "learning_rate": 1.883775824364856e-05,
+      "loss": 1.2663,
+      "step": 537
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.718882161412235,
+      "learning_rate": 1.8832655967944607e-05,
+      "loss": 1.0759,
+      "step": 538
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9490636807941163,
+      "learning_rate": 1.8827543211294134e-05,
+      "loss": 1.1596,
+      "step": 539
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8255578560756088,
+      "learning_rate": 1.882241997976399e-05,
+      "loss": 1.274,
+      "step": 540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7717313650787484,
+      "learning_rate": 1.8817286279433474e-05,
+      "loss": 1.065,
+      "step": 541
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1455579672802718,
+      "learning_rate": 1.8812142116394308e-05,
+      "loss": 0.9848,
+      "step": 542
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7556152085711264,
+      "learning_rate": 1.8806987496750623e-05,
+      "loss": 1.086,
+      "step": 543
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8455746647801307,
+      "learning_rate": 1.8801822426618956e-05,
+      "loss": 1.3059,
+      "step": 544
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8024632577256957,
+      "learning_rate": 1.8796646912128247e-05,
+      "loss": 0.8912,
+      "step": 545
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6134337168257409,
+      "learning_rate": 1.8791460959419832e-05,
+      "loss": 1.1039,
+      "step": 546
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8787931167408898,
+      "learning_rate": 1.8786264574647428e-05,
+      "loss": 1.4546,
+      "step": 547
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.714022996419147,
+      "learning_rate": 1.878105776397714e-05,
+      "loss": 1.0555,
+      "step": 548
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.936775651290531,
+      "learning_rate": 1.8775840533587436e-05,
+      "loss": 1.2748,
+      "step": 549
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5921856385940377,
+      "learning_rate": 1.8770612889669146e-05,
+      "loss": 1.2778,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7272462857147484,
+      "learning_rate": 1.8765374838425467e-05,
+      "loss": 1.0679,
+      "step": 551
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.817661556115021,
+      "learning_rate": 1.8760126386071937e-05,
+      "loss": 0.9836,
+      "step": 552
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7823596117699664,
+      "learning_rate": 1.8754867538836437e-05,
+      "loss": 1.1737,
+      "step": 553
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8793936410295087,
+      "learning_rate": 1.8749598302959195e-05,
+      "loss": 1.2727,
+      "step": 554
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9033007948467923,
+      "learning_rate": 1.8744318684692742e-05,
+      "loss": 1.0006,
+      "step": 555
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9244715396386729,
+      "learning_rate": 1.8739028690301953e-05,
+      "loss": 1.313,
+      "step": 556
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9476369660403333,
+      "learning_rate": 1.8733728326064e-05,
+      "loss": 0.9256,
+      "step": 557
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.761283504723219,
+      "learning_rate": 1.8728417598268375e-05,
+      "loss": 0.9258,
+      "step": 558
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.86767042174229,
+      "learning_rate": 1.8723096513216842e-05,
+      "loss": 1.2512,
+      "step": 559
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7168204098875576,
+      "learning_rate": 1.8717765077223485e-05,
+      "loss": 1.1157,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8049083290558939,
+      "learning_rate": 1.871242329661465e-05,
+      "loss": 0.9735,
+      "step": 561
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7928206860185085,
+      "learning_rate": 1.8707071177728962e-05,
+      "loss": 1.0563,
+      "step": 562
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7213207707059076,
+      "learning_rate": 1.8701708726917325e-05,
+      "loss": 1.3711,
+      "step": 563
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7238326111943297,
+      "learning_rate": 1.8696335950542886e-05,
+      "loss": 1.1441,
+      "step": 564
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6409546964610308,
+      "learning_rate": 1.869095285498106e-05,
+      "loss": 1.1581,
+      "step": 565
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0480950421095994,
+      "learning_rate": 1.868555944661949e-05,
+      "loss": 1.3047,
+      "step": 566
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8413356987939475,
+      "learning_rate": 1.868015573185807e-05,
+      "loss": 1.0798,
+      "step": 567
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7281879327640832,
+      "learning_rate": 1.8674741717108923e-05,
+      "loss": 1.2461,
+      "step": 568
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6351625739992325,
+      "learning_rate": 1.8669317408796387e-05,
+      "loss": 0.7552,
+      "step": 569
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7465424294377462,
+      "learning_rate": 1.866388281335702e-05,
+      "loss": 1.0933,
+      "step": 570
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9690830745773897,
+      "learning_rate": 1.8658437937239586e-05,
+      "loss": 1.3051,
+      "step": 571
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7771806315109613,
+      "learning_rate": 1.8652982786905043e-05,
+      "loss": 1.1531,
+      "step": 572
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8147709244596334,
+      "learning_rate": 1.8647517368826545e-05,
+      "loss": 1.5456,
+      "step": 573
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5731660578681383,
+      "learning_rate": 1.8642041689489435e-05,
+      "loss": 1.158,
+      "step": 574
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8352949944323447,
+      "learning_rate": 1.863655575539122e-05,
+      "loss": 1.3033,
+      "step": 575
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6810530343503973,
+      "learning_rate": 1.8631059573041583e-05,
+      "loss": 1.0565,
+      "step": 576
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7362131487233594,
+      "learning_rate": 1.8625553148962373e-05,
+      "loss": 1.1184,
+      "step": 577
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0257567793795197,
+      "learning_rate": 1.8620036489687578e-05,
+      "loss": 0.8913,
+      "step": 578
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7036561285540086,
+      "learning_rate": 1.8614509601763343e-05,
+      "loss": 1.1527,
+      "step": 579
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8099470440589424,
+      "learning_rate": 1.8608972491747946e-05,
+      "loss": 1.1388,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6668841596328575,
+      "learning_rate": 1.8603425166211795e-05,
+      "loss": 1.0865,
+      "step": 581
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9741400459374031,
+      "learning_rate": 1.859786763173742e-05,
+      "loss": 1.4172,
+      "step": 582
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7208861633273779,
+      "learning_rate": 1.8592299894919462e-05,
+      "loss": 1.1263,
+      "step": 583
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7283302170882795,
+      "learning_rate": 1.8586721962364676e-05,
+      "loss": 1.1961,
+      "step": 584
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0104354147564716,
+      "learning_rate": 1.8581133840691907e-05,
+      "loss": 1.2099,
+      "step": 585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8418144957056612,
+      "learning_rate": 1.8575535536532093e-05,
+      "loss": 1.1891,
+      "step": 586
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6501566186528025,
+      "learning_rate": 1.8569927056528265e-05,
+      "loss": 1.2826,
+      "step": 587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2220326243728103,
+      "learning_rate": 1.8564308407335508e-05,
+      "loss": 0.7679,
+      "step": 588
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5599228717943621,
+      "learning_rate": 1.8558679595620993e-05,
+      "loss": 1.1518,
+      "step": 589
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9499168667797513,
+      "learning_rate": 1.855304062806394e-05,
+      "loss": 1.1537,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.779802327083929,
+      "learning_rate": 1.854739151135562e-05,
+      "loss": 1.0955,
+      "step": 591
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6998020765804661,
+      "learning_rate": 1.8541732252199357e-05,
+      "loss": 1.289,
+      "step": 592
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7084782234823248,
+      "learning_rate": 1.8536062857310496e-05,
+      "loss": 1.037,
+      "step": 593
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9068115081009135,
+      "learning_rate": 1.853038333341642e-05,
+      "loss": 1.2435,
+      "step": 594
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7377468124411005,
+      "learning_rate": 1.852469368725652e-05,
+      "loss": 1.3231,
+      "step": 595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5331377184842,
+      "learning_rate": 1.8518993925582217e-05,
+      "loss": 0.8159,
+      "step": 596
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.783915202133209,
+      "learning_rate": 1.8513284055156916e-05,
+      "loss": 1.2639,
+      "step": 597
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6979368332400752,
+      "learning_rate": 1.8507564082756024e-05,
+      "loss": 1.1289,
+      "step": 598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9142543998284723,
+      "learning_rate": 1.850183401516694e-05,
+      "loss": 1.448,
+      "step": 599
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7434582041145807,
+      "learning_rate": 1.8496093859189037e-05,
+      "loss": 1.0562,
+      "step": 600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0707634010398528,
+      "learning_rate": 1.849034362163366e-05,
+      "loss": 1.0632,
+      "step": 601
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3283682266681323,
+      "learning_rate": 1.848458330932412e-05,
+      "loss": 1.3279,
+      "step": 602
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5036864251259108,
+      "learning_rate": 1.847881292909567e-05,
+      "loss": 1.0793,
+      "step": 603
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.914652327237121,
+      "learning_rate": 1.847303248779553e-05,
+      "loss": 1.0496,
+      "step": 604
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.780625569083374,
+      "learning_rate": 1.8467241992282842e-05,
+      "loss": 1.1436,
+      "step": 605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.195042017662793,
+      "learning_rate": 1.846144144942869e-05,
+      "loss": 1.0829,
+      "step": 606
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6892747506475624,
+      "learning_rate": 1.8455630866116073e-05,
+      "loss": 1.0576,
+      "step": 607
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7372738234577917,
+      "learning_rate": 1.84498102492399e-05,
+      "loss": 1.0048,
+      "step": 608
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8676177569327136,
+      "learning_rate": 1.8443979605707e-05,
+      "loss": 1.357,
+      "step": 609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8303987146844867,
+      "learning_rate": 1.843813894243609e-05,
+      "loss": 1.05,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8684252219774679,
+      "learning_rate": 1.8432288266357778e-05,
+      "loss": 1.278,
+      "step": 611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6614680413502932,
+      "learning_rate": 1.842642758441455e-05,
+      "loss": 1.2298,
+      "step": 612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9226504190501511,
+      "learning_rate": 1.8420556903560773e-05,
+      "loss": 1.2964,
+      "step": 613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7506732412195121,
+      "learning_rate": 1.8414676230762674e-05,
+      "loss": 0.7742,
+      "step": 614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5185868061938737,
+      "learning_rate": 1.8408785572998335e-05,
+      "loss": 1.1064,
+      "step": 615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8065795339973685,
+      "learning_rate": 1.8402884937257693e-05,
+      "loss": 1.4313,
+      "step": 616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7736434998542212,
+      "learning_rate": 1.8396974330542518e-05,
+      "loss": 0.9542,
+      "step": 617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9045069770284873,
+      "learning_rate": 1.8391053759866412e-05,
+      "loss": 1.3584,
+      "step": 618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6301888331739336,
+      "learning_rate": 1.8385123232254805e-05,
+      "loss": 1.0132,
+      "step": 619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7042209037643595,
+      "learning_rate": 1.8379182754744936e-05,
+      "loss": 1.0839,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8469822980709343,
+      "learning_rate": 1.837323233438586e-05,
+      "loss": 1.2594,
+      "step": 621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6913857350983793,
+      "learning_rate": 1.8367271978238422e-05,
+      "loss": 0.9593,
+      "step": 622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0567757682912906,
+      "learning_rate": 1.836130169337526e-05,
+      "loss": 1.2802,
+      "step": 623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.679282518723731,
+      "learning_rate": 1.835532148688079e-05,
+      "loss": 1.1347,
+      "step": 624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8342251485592597,
+      "learning_rate": 1.834933136585121e-05,
+      "loss": 1.2913,
+      "step": 625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7752485411716104,
+      "learning_rate": 1.8343331337394472e-05,
+      "loss": 1.0616,
+      "step": 626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4824065330802627,
+      "learning_rate": 1.8337321408630292e-05,
+      "loss": 1.2096,
+      "step": 627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7711363006951849,
+      "learning_rate": 1.833130158669013e-05,
+      "loss": 1.2805,
+      "step": 628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7403584808578256,
+      "learning_rate": 1.8325271878717187e-05,
+      "loss": 0.9433,
+      "step": 629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6399276903722305,
+      "learning_rate": 1.8319232291866388e-05,
+      "loss": 1.1478,
+      "step": 630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7785356950292148,
+      "learning_rate": 1.8313182833304397e-05,
+      "loss": 1.1756,
+      "step": 631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6202266952177202,
+      "learning_rate": 1.830712351020958e-05,
+      "loss": 1.0516,
+      "step": 632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7601755971434777,
+      "learning_rate": 1.8301054329772e-05,
+      "loss": 0.9948,
+      "step": 633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.907723273969106,
+      "learning_rate": 1.829497529919344e-05,
+      "loss": 1.1131,
+      "step": 634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6995062259542859,
+      "learning_rate": 1.828888642568735e-05,
+      "loss": 1.186,
+      "step": 635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7465197759790763,
+      "learning_rate": 1.828278771647887e-05,
+      "loss": 1.0404,
+      "step": 636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6837401966566865,
+      "learning_rate": 1.827667917880481e-05,
+      "loss": 1.2889,
+      "step": 637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6715182203587148,
+      "learning_rate": 1.8270560819913637e-05,
+      "loss": 1.0699,
+      "step": 638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4719986420413043,
+      "learning_rate": 1.8264432647065485e-05,
+      "loss": 1.2049,
+      "step": 639
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.644368733708318,
+      "learning_rate": 1.825829466753212e-05,
+      "loss": 0.907,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7162100520165472,
+      "learning_rate": 1.8252146888596946e-05,
+      "loss": 1.0522,
+      "step": 641
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7339103721511632,
+      "learning_rate": 1.8245989317555008e-05,
+      "loss": 1.1912,
+      "step": 642
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6194304305490401,
+      "learning_rate": 1.8239821961712953e-05,
+      "loss": 1.1569,
+      "step": 643
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7203828470348336,
+      "learning_rate": 1.8233644828389056e-05,
+      "loss": 1.0856,
+      "step": 644
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7079992054753402,
+      "learning_rate": 1.8227457924913175e-05,
+      "loss": 1.2588,
+      "step": 645
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7256050278069859,
+      "learning_rate": 1.8221261258626777e-05,
+      "loss": 0.9787,
+      "step": 646
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3792976342460939,
+      "learning_rate": 1.821505483688291e-05,
+      "loss": 1.2092,
+      "step": 647
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7045892197680266,
+      "learning_rate": 1.8208838667046193e-05,
+      "loss": 1.2447,
+      "step": 648
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6589363488399195,
+      "learning_rate": 1.8202612756492817e-05,
+      "loss": 0.9908,
+      "step": 649
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6696251647836866,
+      "learning_rate": 1.8196377112610524e-05,
+      "loss": 1.1644,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7310030339729314,
+      "learning_rate": 1.8190131742798623e-05,
+      "loss": 1.0959,
+      "step": 651
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5565353686962363,
+      "learning_rate": 1.8183876654467947e-05,
+      "loss": 1.1937,
+      "step": 652
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7558042295016341,
+      "learning_rate": 1.8177611855040865e-05,
+      "loss": 0.9886,
+      "step": 653
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8360813853974584,
+      "learning_rate": 1.8171337351951268e-05,
+      "loss": 1.3243,
+      "step": 654
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6850775001780893,
+      "learning_rate": 1.816505315264457e-05,
+      "loss": 1.155,
+      "step": 655
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8969065600226596,
+      "learning_rate": 1.8158759264577684e-05,
+      "loss": 1.3058,
+      "step": 656
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.731247416758428,
+      "learning_rate": 1.8152455695219025e-05,
+      "loss": 1.0272,
+      "step": 657
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7487397087099822,
+      "learning_rate": 1.8146142452048484e-05,
+      "loss": 1.0353,
+      "step": 658
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7206413461205357,
+      "learning_rate": 1.813981954255744e-05,
+      "loss": 0.8403,
+      "step": 659
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0117137668272682,
+      "learning_rate": 1.8133486974248748e-05,
+      "loss": 1.4087,
+      "step": 660
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9593658017510689,
+      "learning_rate": 1.812714475463671e-05,
+      "loss": 1.1829,
+      "step": 661
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6468977116464418,
+      "learning_rate": 1.812079289124709e-05,
+      "loss": 1.1367,
+      "step": 662
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.721974120658871,
+      "learning_rate": 1.8114431391617096e-05,
+      "loss": 1.1287,
+      "step": 663
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6861118180169854,
+      "learning_rate": 1.8108060263295364e-05,
+      "loss": 1.2014,
+      "step": 664
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9961217870161995,
+      "learning_rate": 1.810167951384196e-05,
+      "loss": 1.0734,
+      "step": 665
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1537511745174924,
+      "learning_rate": 1.8095289150828368e-05,
+      "loss": 1.2195,
+      "step": 666
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6072668179519647,
+      "learning_rate": 1.8088889181837474e-05,
+      "loss": 0.8899,
+      "step": 667
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1043903401660173,
+      "learning_rate": 1.808247961446357e-05,
+      "loss": 1.2926,
+      "step": 668
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7042767449248901,
+      "learning_rate": 1.807606045631233e-05,
+      "loss": 1.1771,
+      "step": 669
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6363809136178574,
+      "learning_rate": 1.8069631715000815e-05,
+      "loss": 1.0234,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9088665904113629,
+      "learning_rate": 1.8063193398157453e-05,
+      "loss": 1.4307,
+      "step": 671
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8786714119555,
+      "learning_rate": 1.8056745513422034e-05,
+      "loss": 0.9971,
+      "step": 672
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9217413507607104,
+      "learning_rate": 1.8050288068445706e-05,
+      "loss": 1.2779,
+      "step": 673
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8440836026619959,
+      "learning_rate": 1.8043821070890958e-05,
+      "loss": 0.975,
+      "step": 674
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0174295058357292,
+      "learning_rate": 1.8037344528431616e-05,
+      "loss": 1.3753,
+      "step": 675
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.493689764516947,
+      "learning_rate": 1.803085844875283e-05,
+      "loss": 0.798,
+      "step": 676
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6013564385480249,
+      "learning_rate": 1.802436283955107e-05,
+      "loss": 1.0953,
+      "step": 677
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8758675912660442,
+      "learning_rate": 1.8017857708534107e-05,
+      "loss": 1.3275,
+      "step": 678
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.100863870610081,
+      "learning_rate": 1.8011343063421022e-05,
+      "loss": 1.1747,
+      "step": 679
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7673569117420678,
+      "learning_rate": 1.8004818911942173e-05,
+      "loss": 1.222,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.680207164307256,
+      "learning_rate": 1.7998285261839208e-05,
+      "loss": 1.1734,
+      "step": 681
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6955115627982883,
+      "learning_rate": 1.799174212086504e-05,
+      "loss": 0.9245,
+      "step": 682
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9135385495759433,
+      "learning_rate": 1.7985189496783853e-05,
+      "loss": 1.5139,
+      "step": 683
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.863975076381669,
+      "learning_rate": 1.7978627397371072e-05,
+      "loss": 0.9831,
+      "step": 684
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6101048685389358,
+      "learning_rate": 1.7972055830413372e-05,
+      "loss": 0.8668,
+      "step": 685
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7680865708462963,
+      "learning_rate": 1.7965474803708657e-05,
+      "loss": 1.028,
+      "step": 686
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7053134693974854,
+      "learning_rate": 1.7958884325066073e-05,
+      "loss": 1.332,
+      "step": 687
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5777910775648893,
+      "learning_rate": 1.7952284402305956e-05,
+      "loss": 1.1661,
+      "step": 688
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8068155576035376,
+      "learning_rate": 1.7945675043259868e-05,
+      "loss": 1.0012,
+      "step": 689
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.954951092913106,
+      "learning_rate": 1.7939056255770562e-05,
+      "loss": 1.2359,
+      "step": 690
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7839111022361347,
+      "learning_rate": 1.7932428047691978e-05,
+      "loss": 1.1483,
+      "step": 691
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7146920304901228,
+      "learning_rate": 1.7925790426889237e-05,
+      "loss": 1.2987,
+      "step": 692
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5134225072142214,
+      "learning_rate": 1.7919143401238623e-05,
+      "loss": 0.8563,
+      "step": 693
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7618611849012844,
+      "learning_rate": 1.791248697862759e-05,
+      "loss": 1.2969,
+      "step": 694
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7846268186065233,
+      "learning_rate": 1.7905821166954732e-05,
+      "loss": 0.981,
+      "step": 695
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.869724938888259,
+      "learning_rate": 1.7899145974129793e-05,
+      "loss": 1.0213,
+      "step": 696
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8597966867521096,
+      "learning_rate": 1.789246140807365e-05,
+      "loss": 1.2553,
+      "step": 697
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0506091549570147,
+      "learning_rate": 1.7885767476718285e-05,
+      "loss": 1.1748,
+      "step": 698
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9249660042793997,
+      "learning_rate": 1.7879064188006818e-05,
+      "loss": 1.3499,
+      "step": 699
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7533434368094664,
+      "learning_rate": 1.7872351549893452e-05,
+      "loss": 1.0309,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5542502232012351,
+      "learning_rate": 1.7865629570343496e-05,
+      "loss": 1.2774,
+      "step": 701
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5631566318090143,
+      "learning_rate": 1.785889825733334e-05,
+      "loss": 0.8988,
+      "step": 702
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7249243977267447,
+      "learning_rate": 1.7852157618850447e-05,
+      "loss": 0.9867,
+      "step": 703
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9966544939536518,
+      "learning_rate": 1.784540766289335e-05,
+      "loss": 1.2632,
+      "step": 704
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7366973337524998,
+      "learning_rate": 1.7838648397471634e-05,
+      "loss": 1.1628,
+      "step": 705
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9702350724811187,
+      "learning_rate": 1.783187983060594e-05,
+      "loss": 1.26,
+      "step": 706
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6849847126778983,
+      "learning_rate": 1.7825101970327926e-05,
+      "loss": 0.9937,
+      "step": 707
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.664974724617022,
+      "learning_rate": 1.78183148246803e-05,
+      "loss": 1.1187,
+      "step": 708
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.966545453511582,
+      "learning_rate": 1.7811518401716776e-05,
+      "loss": 1.2444,
+      "step": 709
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8508363248848653,
+      "learning_rate": 1.7804712709502083e-05,
+      "loss": 1.068,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6126061281449963,
+      "learning_rate": 1.7797897756111942e-05,
+      "loss": 0.939,
+      "step": 711
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9001607884907395,
+      "learning_rate": 1.7791073549633065e-05,
+      "loss": 1.0679,
+      "step": 712
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.48195145218721386,
+      "learning_rate": 1.7784240098163154e-05,
+      "loss": 1.2519,
+      "step": 713
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7075254753706363,
+      "learning_rate": 1.7777397409810863e-05,
+      "loss": 1.1323,
+      "step": 714
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6323923536167474,
+      "learning_rate": 1.7770545492695825e-05,
+      "loss": 1.1095,
+      "step": 715
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9502418402739828,
+      "learning_rate": 1.7763684354948608e-05,
+      "loss": 1.2294,
+      "step": 716
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7940069540236292,
+      "learning_rate": 1.7756814004710734e-05,
+      "loss": 1.0879,
+      "step": 717
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8182751207922533,
+      "learning_rate": 1.774993445013465e-05,
+      "loss": 1.3178,
+      "step": 718
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6666223977830773,
+      "learning_rate": 1.7743045699383726e-05,
+      "loss": 1.0586,
+      "step": 719
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.676278871509962,
+      "learning_rate": 1.773614776063225e-05,
+      "loss": 1.1234,
+      "step": 720
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7350573483444648,
+      "learning_rate": 1.7729240642065396e-05,
+      "loss": 0.8682,
+      "step": 721
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6772396819142212,
+      "learning_rate": 1.7722324351879252e-05,
+      "loss": 0.9467,
+      "step": 722
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8541401957940518,
+      "learning_rate": 1.771539889828078e-05,
+      "loss": 1.2631,
+      "step": 723
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6865384431455315,
+      "learning_rate": 1.770846428948781e-05,
+      "loss": 1.1065,
+      "step": 724
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6738854491118084,
+      "learning_rate": 1.770152053372904e-05,
+      "loss": 1.1742,
+      "step": 725
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4547106464942515,
+      "learning_rate": 1.769456763924403e-05,
+      "loss": 1.2784,
+      "step": 726
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6929654573615152,
+      "learning_rate": 1.7687605614283167e-05,
+      "loss": 1.0198,
+      "step": 727
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9591167248733101,
+      "learning_rate": 1.768063446710769e-05,
+      "loss": 0.9446,
+      "step": 728
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7808600583138944,
+      "learning_rate": 1.767365420598965e-05,
+      "loss": 1.1164,
+      "step": 729
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8341756470866187,
+      "learning_rate": 1.766666483921192e-05,
+      "loss": 1.1599,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5988624238278876,
+      "learning_rate": 1.765966637506817e-05,
+      "loss": 0.9272,
+      "step": 731
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6844884793719104,
+      "learning_rate": 1.765265882186288e-05,
+      "loss": 1.0823,
+      "step": 732
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9539760509090023,
+      "learning_rate": 1.7645642187911292e-05,
+      "loss": 1.3698,
+      "step": 733
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7315046599293284,
+      "learning_rate": 1.763861648153945e-05,
+      "loss": 1.1049,
+      "step": 734
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8263000690544668,
+      "learning_rate": 1.763158171108414e-05,
+      "loss": 1.2526,
+      "step": 735
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7193966384092704,
+      "learning_rate": 1.7624537884892923e-05,
+      "loss": 1.3074,
+      "step": 736
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8168681110796431,
+      "learning_rate": 1.761748501132409e-05,
+      "loss": 1.3299,
+      "step": 737
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1731436432693851,
+      "learning_rate": 1.761042309874667e-05,
+      "loss": 0.8265,
+      "step": 738
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.692933158804206,
+      "learning_rate": 1.7603352155540433e-05,
+      "loss": 1.0769,
+      "step": 739
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6605763215556315,
+      "learning_rate": 1.759627219009585e-05,
+      "loss": 1.2421,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6456110763413848,
+      "learning_rate": 1.7589183210814094e-05,
+      "loss": 1.0008,
+      "step": 741
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8451271480219952,
+      "learning_rate": 1.7582085226107054e-05,
+      "loss": 1.2759,
+      "step": 742
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0149242484611167,
+      "learning_rate": 1.7574978244397286e-05,
+      "loss": 0.9793,
+      "step": 743
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7793313975500494,
+      "learning_rate": 1.7567862274118027e-05,
+      "loss": 0.9904,
+      "step": 744
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8570668115547272,
+      "learning_rate": 1.7560737323713184e-05,
+      "loss": 1.1931,
+      "step": 745
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7009867726079833,
+      "learning_rate": 1.7553603401637317e-05,
+      "loss": 1.1614,
+      "step": 746
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1846619270410839,
+      "learning_rate": 1.754646051635563e-05,
+      "loss": 0.9597,
+      "step": 747
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8046880409894583,
+      "learning_rate": 1.7539308676343972e-05,
+      "loss": 1.0609,
+      "step": 748
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.026040421718317,
+      "learning_rate": 1.7532147890088807e-05,
+      "loss": 1.3885,
+      "step": 749
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9412151552002544,
+      "learning_rate": 1.752497816608722e-05,
+      "loss": 1.144,
+      "step": 750
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6089395177705107,
+      "learning_rate": 1.7517799512846897e-05,
+      "loss": 1.1438,
+      "step": 751
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7524226335775703,
+      "learning_rate": 1.7510611938886123e-05,
+      "loss": 1.1791,
+      "step": 752
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7737907589982677,
+      "learning_rate": 1.7503415452733774e-05,
+      "loss": 1.0603,
+      "step": 753
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7767425167908045,
+      "learning_rate": 1.7496210062929287e-05,
+      "loss": 1.2535,
+      "step": 754
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8010888278388304,
+      "learning_rate": 1.7488995778022687e-05,
+      "loss": 0.7662,
+      "step": 755
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6787719793919639,
+      "learning_rate": 1.7481772606574526e-05,
+      "loss": 1.2451,
+      "step": 756
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7962488004190779,
+      "learning_rate": 1.7474540557155917e-05,
+      "loss": 1.0378,
+      "step": 757
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.771593985051702,
+      "learning_rate": 1.7467299638348514e-05,
+      "loss": 1.0284,
+      "step": 758
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6893570145177388,
+      "learning_rate": 1.7460049858744482e-05,
+      "loss": 1.5014,
+      "step": 759
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.725488660414114,
+      "learning_rate": 1.7452791226946502e-05,
+      "loss": 1.1325,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7993589098570456,
+      "learning_rate": 1.744552375156777e-05,
+      "loss": 1.2895,
+      "step": 761
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.69956783141036,
+      "learning_rate": 1.743824744123196e-05,
+      "loss": 0.9592,
+      "step": 762
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5387582732951872,
+      "learning_rate": 1.743096230457325e-05,
+      "loss": 1.1057,
+      "step": 763
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8178579247203464,
+      "learning_rate": 1.7423668350236265e-05,
+      "loss": 0.8963,
+      "step": 764
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7102284739162593,
+      "learning_rate": 1.7416365586876118e-05,
+      "loss": 1.0588,
+      "step": 765
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2744971875780038,
+      "learning_rate": 1.7409054023158362e-05,
+      "loss": 1.127,
+      "step": 766
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6710266422282334,
+      "learning_rate": 1.7401733667758993e-05,
+      "loss": 1.0693,
+      "step": 767
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8334620564263383,
+      "learning_rate": 1.7394404529364447e-05,
+      "loss": 1.2173,
+      "step": 768
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7314598648125761,
+      "learning_rate": 1.7387066616671572e-05,
+      "loss": 1.0522,
+      "step": 769
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6005832495578228,
+      "learning_rate": 1.7379719938387634e-05,
+      "loss": 1.1217,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7606652503185934,
+      "learning_rate": 1.73723645032303e-05,
+      "loss": 1.2641,
+      "step": 771
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5418667667121188,
+      "learning_rate": 1.7365000319927624e-05,
+      "loss": 1.2072,
+      "step": 772
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8776613608878291,
+      "learning_rate": 1.735762739721804e-05,
+      "loss": 0.9283,
+      "step": 773
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7257037331084358,
+      "learning_rate": 1.7350245743850367e-05,
+      "loss": 1.039,
+      "step": 774
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5524915443072163,
+      "learning_rate": 1.7342855368583764e-05,
+      "loss": 1.138,
+      "step": 775
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8614184187507274,
+      "learning_rate": 1.7335456280187752e-05,
+      "loss": 1.1764,
+      "step": 776
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6605570661278745,
+      "learning_rate": 1.7328048487442184e-05,
+      "loss": 1.1115,
+      "step": 777
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6798781841613644,
+      "learning_rate": 1.7320631999137245e-05,
+      "loss": 1.2916,
+      "step": 778
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6872632788034485,
+      "learning_rate": 1.7313206824073444e-05,
+      "loss": 0.9647,
+      "step": 779
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6857890212272909,
+      "learning_rate": 1.7305772971061583e-05,
+      "loss": 1.3716,
+      "step": 780
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6211340796840452,
+      "learning_rate": 1.729833044892278e-05,
+      "loss": 0.7657,
+      "step": 781
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6425182331968939,
+      "learning_rate": 1.729087926648842e-05,
+      "loss": 1.0334,
+      "step": 782
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6936731108931592,
+      "learning_rate": 1.7283419432600184e-05,
+      "loss": 1.3067,
+      "step": 783
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6245842220732276,
+      "learning_rate": 1.7275950956110003e-05,
+      "loss": 1.1604,
+      "step": 784
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7676933499344513,
+      "learning_rate": 1.7268473845880076e-05,
+      "loss": 1.2832,
+      "step": 785
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6477642975717636,
+      "learning_rate": 1.7260988110782836e-05,
+      "loss": 1.015,
+      "step": 786
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9519123152633837,
+      "learning_rate": 1.7253493759700958e-05,
+      "loss": 1.2744,
+      "step": 787
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6473569580184851,
+      "learning_rate": 1.724599080152734e-05,
+      "loss": 1.0473,
+      "step": 788
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5060888704257502,
+      "learning_rate": 1.7238479245165084e-05,
+      "loss": 1.0849,
+      "step": 789
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8545399588852657,
+      "learning_rate": 1.7230959099527512e-05,
+      "loss": 1.1003,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6322594140625621,
+      "learning_rate": 1.7223430373538125e-05,
+      "loss": 1.1453,
+      "step": 791
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7300251698618195,
+      "learning_rate": 1.7215893076130603e-05,
+      "loss": 1.0287,
+      "step": 792
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6649391952731999,
+      "learning_rate": 1.720834721624881e-05,
+      "loss": 1.0641,
+      "step": 793
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6714710554694968,
+      "learning_rate": 1.7200792802846763e-05,
+      "loss": 1.0251,
+      "step": 794
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8883612732125409,
+      "learning_rate": 1.7193229844888625e-05,
+      "loss": 1.3063,
+      "step": 795
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7617311184285979,
+      "learning_rate": 1.7185658351348703e-05,
+      "loss": 0.9862,
+      "step": 796
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.671219978860752,
+      "learning_rate": 1.717807833121143e-05,
+      "loss": 1.2957,
+      "step": 797
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6563239043032849,
+      "learning_rate": 1.7170489793471365e-05,
+      "loss": 1.1462,
+      "step": 798
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8209333403151466,
+      "learning_rate": 1.7162892747133163e-05,
+      "loss": 1.1688,
+      "step": 799
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7453589791278252,
+      "learning_rate": 1.7155287201211578e-05,
+      "loss": 0.742,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5395130873056281,
+      "learning_rate": 1.7147673164731453e-05,
+      "loss": 1.113,
+      "step": 801
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7641271934746884,
+      "learning_rate": 1.7140050646727706e-05,
+      "loss": 1.3603,
+      "step": 802
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7735883961081865,
+      "learning_rate": 1.7132419656245316e-05,
+      "loss": 1.1747,
+      "step": 803
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7604006034709704,
+      "learning_rate": 1.712478020233932e-05,
+      "loss": 1.2675,
+      "step": 804
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.625117195470551,
+      "learning_rate": 1.7117132294074793e-05,
+      "loss": 0.9791,
+      "step": 805
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7122815103882818,
+      "learning_rate": 1.710947594052684e-05,
+      "loss": 1.2486,
+      "step": 806
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8475471792267535,
+      "learning_rate": 1.7101811150780602e-05,
+      "loss": 1.1367,
+      "step": 807
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7595938288709491,
+      "learning_rate": 1.709413793393121e-05,
+      "loss": 0.8084,
+      "step": 808
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6686229160735851,
+      "learning_rate": 1.708645629908381e-05,
+      "loss": 1.2858,
+      "step": 809
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6580233760238006,
+      "learning_rate": 1.707876625535353e-05,
+      "loss": 1.1259,
+      "step": 810
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7847620361314064,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 1.3641,
+      "step": 811
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6648284762619924,
+      "learning_rate": 1.7063360977754726e-05,
+      "loss": 1.0232,
+      "step": 812
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4820017888070664,
+      "learning_rate": 1.7055645762166314e-05,
+      "loss": 1.2244,
+      "step": 813
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7141920666637412,
+      "learning_rate": 1.704792217425521e-05,
+      "loss": 1.1159,
+      "step": 814
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6761954323833159,
+      "learning_rate": 1.7040190223186327e-05,
+      "loss": 1.0085,
+      "step": 815
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7086646778782698,
+      "learning_rate": 1.7032449918134506e-05,
+      "loss": 1.2097,
+      "step": 816
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6816317910414367,
+      "learning_rate": 1.7024701268284492e-05,
+      "loss": 1.0178,
+      "step": 817
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5876234920760909,
+      "learning_rate": 1.7016944282830935e-05,
+      "loss": 1.001,
+      "step": 818
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6533800834938732,
+      "learning_rate": 1.7009178970978373e-05,
+      "loss": 0.9974,
+      "step": 819
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6454669869211669,
+      "learning_rate": 1.700140534194124e-05,
+      "loss": 1.0652,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6792546325910553,
+      "learning_rate": 1.699362340494381e-05,
+      "loss": 1.1824,
+      "step": 821
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6300370003081341,
+      "learning_rate": 1.698583316922025e-05,
+      "loss": 1.0202,
+      "step": 822
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6780908528310852,
+      "learning_rate": 1.6978034644014544e-05,
+      "loss": 1.3193,
+      "step": 823
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6776588957763418,
+      "learning_rate": 1.697022783858053e-05,
+      "loss": 1.2156,
+      "step": 824
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.47806018893798496,
+      "learning_rate": 1.6962412762181867e-05,
+      "loss": 1.1371,
+      "step": 825
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6977052181429596,
+      "learning_rate": 1.695458942409203e-05,
+      "loss": 0.854,
+      "step": 826
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6321532114132546,
+      "learning_rate": 1.6946757833594293e-05,
+      "loss": 1.0772,
+      "step": 827
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8154055765155754,
+      "learning_rate": 1.693891799998172e-05,
+      "loss": 1.15,
+      "step": 828
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6499405741116515,
+      "learning_rate": 1.6931069932557173e-05,
+      "loss": 1.1007,
+      "step": 829
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7363805010418552,
+      "learning_rate": 1.692321364063326e-05,
+      "loss": 1.282,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.659585284332615,
+      "learning_rate": 1.691534913353237e-05,
+      "loss": 1.1043,
+      "step": 831
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.565626810988002,
+      "learning_rate": 1.6907476420586632e-05,
+      "loss": 0.885,
+      "step": 832
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7265935877263326,
+      "learning_rate": 1.6899595511137905e-05,
+      "loss": 1.0695,
+      "step": 833
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.49158235849853676,
+      "learning_rate": 1.689170641453778e-05,
+      "loss": 0.8961,
+      "step": 834
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7104796591215876,
+      "learning_rate": 1.6883809140147575e-05,
+      "loss": 1.3267,
+      "step": 835
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7556034617406349,
+      "learning_rate": 1.687590369733829e-05,
+      "loss": 1.0727,
+      "step": 836
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.721924603503175,
+      "learning_rate": 1.6867990095490626e-05,
+      "loss": 1.0822,
+      "step": 837
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6368525862447001,
+      "learning_rate": 1.6860068343994975e-05,
+      "loss": 1.2184,
+      "step": 838
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6828091713274155,
+      "learning_rate": 1.685213845225139e-05,
+      "loss": 0.8821,
+      "step": 839
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.877420469796768,
+      "learning_rate": 1.6844200429669582e-05,
+      "loss": 1.293,
+      "step": 840
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6774624572274394,
+      "learning_rate": 1.683625428566892e-05,
+      "loss": 1.1321,
+      "step": 841
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6423559438911743,
+      "learning_rate": 1.68283000296784e-05,
+      "loss": 1.3354,
+      "step": 842
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8986172560205634,
+      "learning_rate": 1.682033767113665e-05,
+      "loss": 1.0331,
+      "step": 843
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7556654496711835,
+      "learning_rate": 1.6812367219491907e-05,
+      "loss": 0.977,
+      "step": 844
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6219818410723333,
+      "learning_rate": 1.6804388684202012e-05,
+      "loss": 0.8665,
+      "step": 845
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6073284105793156,
+      "learning_rate": 1.6796402074734404e-05,
+      "loss": 1.3075,
+      "step": 846
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7912324816740604,
+      "learning_rate": 1.67884074005661e-05,
+      "loss": 1.3586,
+      "step": 847
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6658089376376938,
+      "learning_rate": 1.678040467118369e-05,
+      "loss": 1.1171,
+      "step": 848
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9043686405104602,
+      "learning_rate": 1.6772393896083314e-05,
+      "loss": 1.0171,
+      "step": 849
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5962028270124824,
+      "learning_rate": 1.6764375084770657e-05,
+      "loss": 1.1963,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6336929674970481,
+      "learning_rate": 1.675634824676096e-05,
+      "loss": 0.9771,
+      "step": 851
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8344181450769514,
+      "learning_rate": 1.6748313391578973e-05,
+      "loss": 1.2522,
+      "step": 852
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5153577703556346,
+      "learning_rate": 1.674027052875895e-05,
+      "loss": 0.8712,
+      "step": 853
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8780630760446395,
+      "learning_rate": 1.673221966784467e-05,
+      "loss": 1.2449,
+      "step": 854
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7087649381507963,
+      "learning_rate": 1.672416081838939e-05,
+      "loss": 1.1647,
+      "step": 855
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6039342792756718,
+      "learning_rate": 1.671609398995584e-05,
+      "loss": 1.1228,
+      "step": 856
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6728488167800359,
+      "learning_rate": 1.6708019192116233e-05,
+      "loss": 1.2936,
+      "step": 857
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6320886343882752,
+      "learning_rate": 1.669993643445223e-05,
+      "loss": 0.9972,
+      "step": 858
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8395701417922717,
+      "learning_rate": 1.6691845726554933e-05,
+      "loss": 1.3342,
+      "step": 859
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6433330138173542,
+      "learning_rate": 1.6683747078024887e-05,
+      "loss": 0.9502,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7354854603726799,
+      "learning_rate": 1.6675640498472055e-05,
+      "loss": 1.3078,
+      "step": 861
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5189728731856527,
+      "learning_rate": 1.6667525997515815e-05,
+      "loss": 0.8503,
+      "step": 862
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.47349551042037746,
+      "learning_rate": 1.6659403584784937e-05,
+      "loss": 1.0975,
+      "step": 863
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7212840627177489,
+      "learning_rate": 1.6651273269917586e-05,
+      "loss": 1.2394,
+      "step": 864
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6251494352739007,
+      "learning_rate": 1.66431350625613e-05,
+      "loss": 1.1907,
+      "step": 865
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6723658860548202,
+      "learning_rate": 1.663498897237298e-05,
+      "loss": 1.1431,
+      "step": 866
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.654418133296627,
+      "learning_rate": 1.6626835009018892e-05,
+      "loss": 1.1458,
+      "step": 867
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6798663023844795,
+      "learning_rate": 1.6618673182174635e-05,
+      "loss": 0.9325,
+      "step": 868
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.73433501742257,
+      "learning_rate": 1.6610503501525136e-05,
+      "loss": 1.4364,
+      "step": 869
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6060138249354762,
+      "learning_rate": 1.660232597676465e-05,
+      "loss": 0.828,
+      "step": 870
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5157503807238196,
+      "learning_rate": 1.6594140617596733e-05,
+      "loss": 0.9213,
+      "step": 871
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.639544785477699,
+      "learning_rate": 1.6585947433734247e-05,
+      "loss": 1.0475,
+      "step": 872
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7235580285980011,
+      "learning_rate": 1.6577746434899324e-05,
+      "loss": 1.3362,
+      "step": 873
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5628535432184121,
+      "learning_rate": 1.6569537630823385e-05,
+      "loss": 1.1609,
+      "step": 874
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7380693920329718,
+      "learning_rate": 1.65613210312471e-05,
+      "loss": 0.9262,
+      "step": 875
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7473231699490854,
+      "learning_rate": 1.65530966459204e-05,
+      "loss": 1.324,
+      "step": 876
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7587877303921416,
+      "learning_rate": 1.6544864484602446e-05,
+      "loss": 1.0711,
+      "step": 877
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7499727953956332,
+      "learning_rate": 1.6536624557061632e-05,
+      "loss": 1.2359,
+      "step": 878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.46880561212167293,
+      "learning_rate": 1.652837687307556e-05,
+      "loss": 0.8655,
+      "step": 879
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7232636851154305,
+      "learning_rate": 1.6520121442431053e-05,
+      "loss": 1.3281,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7064469365791719,
+      "learning_rate": 1.6511858274924098e-05,
+      "loss": 0.9664,
+      "step": 881
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7121006642233513,
+      "learning_rate": 1.6503587380359894e-05,
+      "loss": 1.0693,
+      "step": 882
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.78918471423169,
+      "learning_rate": 1.6495308768552783e-05,
+      "loss": 1.2354,
+      "step": 883
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6899777340335218,
+      "learning_rate": 1.6487022449326282e-05,
+      "loss": 1.0913,
+      "step": 884
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7767137832734547,
+      "learning_rate": 1.6478728432513043e-05,
+      "loss": 1.2934,
+      "step": 885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5435269188528136,
+      "learning_rate": 1.647042672795486e-05,
+      "loss": 1.1159,
+      "step": 886
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8311308610675473,
+      "learning_rate": 1.6462117345502638e-05,
+      "loss": 1.3321,
+      "step": 887
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5485440875986949,
+      "learning_rate": 1.645380029501641e-05,
+      "loss": 0.8951,
+      "step": 888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6914876847575447,
+      "learning_rate": 1.6445475586365295e-05,
+      "loss": 0.9513,
+      "step": 889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6742780895556274,
+      "learning_rate": 1.64371432294275e-05,
+      "loss": 1.2241,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8277910491113315,
+      "learning_rate": 1.642880323409031e-05,
+      "loss": 1.222,
+      "step": 891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.73504542825639,
+      "learning_rate": 1.6420455610250077e-05,
+      "loss": 1.2052,
+      "step": 892
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6061353290332375,
+      "learning_rate": 1.6412100367812204e-05,
+      "loss": 0.9811,
+      "step": 893
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5968799121734012,
+      "learning_rate": 1.6403737516691122e-05,
+      "loss": 1.0432,
+      "step": 894
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7070431257892189,
+      "learning_rate": 1.6395367066810312e-05,
+      "loss": 1.267,
+      "step": 895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7002102830014076,
+      "learning_rate": 1.6386989028102257e-05,
+      "loss": 1.0262,
+      "step": 896
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6682234711830171,
+      "learning_rate": 1.6378603410508444e-05,
+      "loss": 0.9694,
+      "step": 897
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7630320771556174,
+      "learning_rate": 1.6370210223979366e-05,
+      "loss": 1.0648,
+      "step": 898
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.602734044745189,
+      "learning_rate": 1.6361809478474483e-05,
+      "loss": 1.2225,
+      "step": 899
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7001488353321577,
+      "learning_rate": 1.6353401183962235e-05,
+      "loss": 1.118,
+      "step": 900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7004687381407505,
+      "learning_rate": 1.634498535042001e-05,
+      "loss": 1.0462,
+      "step": 901
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8231169483200436,
+      "learning_rate": 1.6336561987834155e-05,
+      "loss": 1.2397,
+      "step": 902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6554212825228936,
+      "learning_rate": 1.6328131106199936e-05,
+      "loss": 1.0778,
+      "step": 903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8101657765799328,
+      "learning_rate": 1.6319692715521552e-05,
+      "loss": 1.2446,
+      "step": 904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6961161919181069,
+      "learning_rate": 1.6311246825812115e-05,
+      "loss": 1.0911,
+      "step": 905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6891637178912318,
+      "learning_rate": 1.630279344709362e-05,
+      "loss": 1.0698,
+      "step": 906
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6783987591168238,
+      "learning_rate": 1.6294332589396965e-05,
+      "loss": 0.8499,
+      "step": 907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5875832341958067,
+      "learning_rate": 1.628586426276191e-05,
+      "loss": 0.936,
+      "step": 908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7772309305649111,
+      "learning_rate": 1.6277388477237087e-05,
+      "loss": 1.3489,
+      "step": 909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6085994597995661,
+      "learning_rate": 1.6268905242879975e-05,
+      "loss": 1.041,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6202038135358384,
+      "learning_rate": 1.6260414569756894e-05,
+      "loss": 1.2974,
+      "step": 911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.45957340141962044,
+      "learning_rate": 1.625191646794299e-05,
+      "loss": 1.1874,
+      "step": 912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7395666970757436,
+      "learning_rate": 1.6243410947522214e-05,
+      "loss": 0.959,
+      "step": 913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7346520070427289,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 0.8635,
+      "step": 914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8906795703982472,
+      "learning_rate": 1.6226377691239913e-05,
+      "loss": 1.1127,
+      "step": 915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.744056409109231,
+      "learning_rate": 1.6217849975590275e-05,
+      "loss": 1.1863,
+      "step": 916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6010953352756473,
+      "learning_rate": 1.620931488175752e-05,
+      "loss": 0.9132,
+      "step": 917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7660501034762701,
+      "learning_rate": 1.6200772419869508e-05,
+      "loss": 1.2852,
+      "step": 918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7141654857663119,
+      "learning_rate": 1.6192222600062833e-05,
+      "loss": 1.1192,
+      "step": 919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6543597158181249,
+      "learning_rate": 1.618366543248283e-05,
+      "loss": 1.059,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8621555324411371,
+      "learning_rate": 1.617510092728354e-05,
+      "loss": 1.1527,
+      "step": 921
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.597032251395634,
+      "learning_rate": 1.6166529094627727e-05,
+      "loss": 1.2858,
+      "step": 922
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6905204872237268,
+      "learning_rate": 1.6157949944686827e-05,
+      "loss": 1.2491,
+      "step": 923
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.674605164285396,
+      "learning_rate": 1.614936348764099e-05,
+      "loss": 0.8027,
+      "step": 924
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6980110006135604,
+      "learning_rate": 1.614076973367901e-05,
+      "loss": 1.135,
+      "step": 925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6451352245888328,
+      "learning_rate": 1.6132168692998353e-05,
+      "loss": 1.2072,
+      "step": 926
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6325326295737436,
+      "learning_rate": 1.6123560375805123e-05,
+      "loss": 0.9306,
+      "step": 927
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9937395580126966,
+      "learning_rate": 1.6114944792314067e-05,
+      "loss": 1.2513,
+      "step": 928
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7079641018041453,
+      "learning_rate": 1.610632195274855e-05,
+      "loss": 0.9403,
+      "step": 929
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7166198681349214,
+      "learning_rate": 1.6097691867340547e-05,
+      "loss": 0.8496,
+      "step": 930
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.329064596837447,
+      "learning_rate": 1.608905454633063e-05,
+      "loss": 1.244,
+      "step": 931
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6758966883313452,
+      "learning_rate": 1.6080409999967962e-05,
+      "loss": 1.1579,
+      "step": 932
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7479244067699421,
+      "learning_rate": 1.6071758238510267e-05,
+      "loss": 0.9327,
+      "step": 933
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.699116185243805,
+      "learning_rate": 1.6063099272223856e-05,
+      "loss": 0.9953,
+      "step": 934
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8775163973438572,
+      "learning_rate": 1.6054433111383562e-05,
+      "loss": 1.3885,
+      "step": 935
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7175167874707578,
+      "learning_rate": 1.604575976627277e-05,
+      "loss": 0.9883,
+      "step": 936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5057848397447293,
+      "learning_rate": 1.603707924718338e-05,
+      "loss": 1.1841,
+      "step": 937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6780493343122458,
+      "learning_rate": 1.602839156441582e-05,
+      "loss": 1.0453,
+      "step": 938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7105099764817611,
+      "learning_rate": 1.6019696728279e-05,
+      "loss": 1.0201,
+      "step": 939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9891722116706507,
+      "learning_rate": 1.6010994749090338e-05,
+      "loss": 1.1391,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6294250323441212,
+      "learning_rate": 1.6002285637175706e-05,
+      "loss": 0.7224,
+      "step": 941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7458806305289697,
+      "learning_rate": 1.599356940286946e-05,
+      "loss": 1.2236,
+      "step": 942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7009111830398861,
+      "learning_rate": 1.59848460565144e-05,
+      "loss": 0.9759,
+      "step": 943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8473206113576789,
+      "learning_rate": 1.5976115608461756e-05,
+      "loss": 1.1572,
+      "step": 944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1227449090489299,
+      "learning_rate": 1.59673780690712e-05,
+      "loss": 1.5153,
+      "step": 945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7043970976447826,
+      "learning_rate": 1.595863344871081e-05,
+      "loss": 1.0985,
+      "step": 946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7920127202522171,
+      "learning_rate": 1.5949881757757073e-05,
+      "loss": 1.2301,
+      "step": 947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6995539435958329,
+      "learning_rate": 1.594112300659485e-05,
+      "loss": 0.9868,
+      "step": 948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.48331928580166006,
+      "learning_rate": 1.5932357205617406e-05,
+      "loss": 1.1297,
+      "step": 949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.105286507789448,
+      "learning_rate": 1.592358436522634e-05,
+      "loss": 0.8796,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6774379539615926,
+      "learning_rate": 1.5914804495831633e-05,
+      "loss": 1.055,
+      "step": 951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7171856392143265,
+      "learning_rate": 1.5906017607851595e-05,
+      "loss": 1.1458,
+      "step": 952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6210020189993924,
+      "learning_rate": 1.589722371171285e-05,
+      "loss": 0.9865,
+      "step": 953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.055581918783621,
+      "learning_rate": 1.5888422817850367e-05,
+      "loss": 1.2501,
+      "step": 954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6732994468147585,
+      "learning_rate": 1.5879614936707395e-05,
+      "loss": 1.1619,
+      "step": 955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7179456233783037,
+      "learning_rate": 1.5870800078735484e-05,
+      "loss": 0.9963,
+      "step": 956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7440520603836308,
+      "learning_rate": 1.5861978254394465e-05,
+      "loss": 1.2664,
+      "step": 957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6071718271028421,
+      "learning_rate": 1.585314947415242e-05,
+      "loss": 1.1701,
+      "step": 958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8954241222134898,
+      "learning_rate": 1.5844313748485714e-05,
+      "loss": 0.9065,
+      "step": 959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6517104630412198,
+      "learning_rate": 1.583547108787892e-05,
+      "loss": 0.952,
+      "step": 960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6185936389068427,
+      "learning_rate": 1.5826621502824862e-05,
+      "loss": 1.2291,
+      "step": 961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8750684331431058,
+      "learning_rate": 1.5817765003824578e-05,
+      "loss": 0.9614,
+      "step": 962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6864802934237031,
+      "learning_rate": 1.5808901601387304e-05,
+      "loss": 1.0866,
+      "step": 963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7057859958868131,
+      "learning_rate": 1.5800031306030476e-05,
+      "loss": 1.2897,
+      "step": 964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7459146961397404,
+      "learning_rate": 1.5791154128279694e-05,
+      "loss": 0.9684,
+      "step": 965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7588116444813651,
+      "learning_rate": 1.578227007866874e-05,
+      "loss": 1.2665,
+      "step": 966
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.777001986132632,
+      "learning_rate": 1.5773379167739547e-05,
+      "loss": 0.7672,
+      "step": 967
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7011787302332867,
+      "learning_rate": 1.5764481406042183e-05,
+      "loss": 0.991,
+      "step": 968
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5856276525850926,
+      "learning_rate": 1.5755576804134855e-05,
+      "loss": 1.2862,
+      "step": 969
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7107172741974491,
+      "learning_rate": 1.5746665372583878e-05,
+      "loss": 1.136,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8561638947986256,
+      "learning_rate": 1.5737747121963673e-05,
+      "loss": 1.2969,
+      "step": 971
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6920586114528242,
+      "learning_rate": 1.5728822062856758e-05,
+      "loss": 1.0067,
+      "step": 972
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8109380350334501,
+      "learning_rate": 1.5719890205853724e-05,
+      "loss": 1.296,
+      "step": 973
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7321327864946914,
+      "learning_rate": 1.5710951561553237e-05,
+      "loss": 1.0083,
+      "step": 974
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4722714473004538,
+      "learning_rate": 1.5702006140562e-05,
+      "loss": 1.132,
+      "step": 975
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.647902208461065,
+      "learning_rate": 1.5693053953494776e-05,
+      "loss": 0.8966,
+      "step": 976
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6876490448661633,
+      "learning_rate": 1.5684095010974346e-05,
+      "loss": 0.8279,
+      "step": 977
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7935269321880268,
+      "learning_rate": 1.567512932363151e-05,
+      "loss": 1.391,
+      "step": 978
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.585979362067601,
+      "learning_rate": 1.5666156902105073e-05,
+      "loss": 1.0327,
+      "step": 979
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3865265627835714,
+      "learning_rate": 1.5657177757041823e-05,
+      "loss": 1.2098,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6730233613098047,
+      "learning_rate": 1.5648191899096538e-05,
+      "loss": 1.1493,
+      "step": 981
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7807131196795682,
+      "learning_rate": 1.5639199338931958e-05,
+      "loss": 1.0804,
+      "step": 982
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7042395975633827,
+      "learning_rate": 1.563020008721877e-05,
+      "loss": 1.2995,
+      "step": 983
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6314377030255415,
+      "learning_rate": 1.5621194154635604e-05,
+      "loss": 1.1321,
+      "step": 984
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7321216936654898,
+      "learning_rate": 1.5612181551869027e-05,
+      "loss": 1.1741,
+      "step": 985
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5184405070329727,
+      "learning_rate": 1.5603162289613503e-05,
+      "loss": 0.9646,
+      "step": 986
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6620612469277165,
+      "learning_rate": 1.559413637857141e-05,
+      "loss": 1.0786,
+      "step": 987
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8390437523405418,
+      "learning_rate": 1.5585103829453018e-05,
+      "loss": 1.2044,
+      "step": 988
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6780287984007061,
+      "learning_rate": 1.5576064652976465e-05,
+      "loss": 1.125,
+      "step": 989
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6709278045674042,
+      "learning_rate": 1.556701885986776e-05,
+      "loss": 1.1971,
+      "step": 990
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6477983990636407,
+      "learning_rate": 1.5557966460860763e-05,
+      "loss": 1.0409,
+      "step": 991
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7918138961482594,
+      "learning_rate": 1.554890746669716e-05,
+      "loss": 1.2881,
+      "step": 992
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6295888122750699,
+      "learning_rate": 1.553984188812649e-05,
+      "loss": 1.1612,
+      "step": 993
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5211942484206588,
+      "learning_rate": 1.5530769735906076e-05,
+      "loss": 0.7864,
+      "step": 994
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7155918640179518,
+      "learning_rate": 1.552169102080106e-05,
+      "loss": 1.172,
+      "step": 995
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6780481083899393,
+      "learning_rate": 1.551260575358436e-05,
+      "loss": 1.1501,
+      "step": 996
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7937204479515761,
+      "learning_rate": 1.5503513945036682e-05,
+      "loss": 1.2822,
+      "step": 997
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6379429254885414,
+      "learning_rate": 1.549441560594648e-05,
+      "loss": 0.9804,
+      "step": 998
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5640498136008442,
+      "learning_rate": 1.5485310747109964e-05,
+      "loss": 1.1956,
+      "step": 999
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7529975621247231,
+      "learning_rate": 1.547619937933108e-05,
+      "loss": 1.16,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7162347855085158,
+      "learning_rate": 1.5467081513421504e-05,
+      "loss": 1.0016,
+      "step": 1001
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8252659399291628,
+      "learning_rate": 1.5457957160200608e-05,
+      "loss": 1.1596,
+      "step": 1002
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7097519789896262,
+      "learning_rate": 1.544882633049548e-05,
+      "loss": 0.7591,
+      "step": 1003
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6396406289579913,
+      "learning_rate": 1.5439689035140873e-05,
+      "loss": 1.2415,
+      "step": 1004
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6478474368449425,
+      "learning_rate": 1.5430545284979225e-05,
+      "loss": 1.0239,
+      "step": 1005
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6666191423609759,
+      "learning_rate": 1.542139509086064e-05,
+      "loss": 1.1467,
+      "step": 1006
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7876656618243175,
+      "learning_rate": 1.5412238463642845e-05,
+      "loss": 1.0672,
+      "step": 1007
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6577200885296666,
+      "learning_rate": 1.540307541419123e-05,
+      "loss": 1.0088,
+      "step": 1008
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6961134263305242,
+      "learning_rate": 1.539390595337878e-05,
+      "loss": 1.303,
+      "step": 1009
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4801355698581469,
+      "learning_rate": 1.5384730092086106e-05,
+      "loss": 1.1807,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8434221959481305,
+      "learning_rate": 1.5375547841201397e-05,
+      "loss": 1.2499,
+      "step": 1011
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9143035538273653,
+      "learning_rate": 1.5366359211620442e-05,
+      "loss": 0.6358,
+      "step": 1012
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.587906645748543,
+      "learning_rate": 1.535716421424659e-05,
+      "loss": 1.0922,
+      "step": 1013
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6956017324254078,
+      "learning_rate": 1.5347962859990744e-05,
+      "loss": 1.1538,
+      "step": 1014
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6428792432778797,
+      "learning_rate": 1.533875515977135e-05,
+      "loss": 1.0725,
+      "step": 1015
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6346564142911305,
+      "learning_rate": 1.5329541124514394e-05,
+      "loss": 1.2599,
+      "step": 1016
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6437386410954381,
+      "learning_rate": 1.5320320765153367e-05,
+      "loss": 1.0818,
+      "step": 1017
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7217324284256089,
+      "learning_rate": 1.531109409262927e-05,
+      "loss": 0.8951,
+      "step": 1018
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6931592792423259,
+      "learning_rate": 1.5301861117890598e-05,
+      "loss": 1.0467,
+      "step": 1019
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0866253317368506,
+      "learning_rate": 1.5292621851893318e-05,
+      "loss": 0.9337,
+      "step": 1020
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7051400287884385,
+      "learning_rate": 1.5283376305600866e-05,
+      "loss": 1.3524,
+      "step": 1021
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.636036437448482,
+      "learning_rate": 1.527412448998413e-05,
+      "loss": 1.0893,
+      "step": 1022
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7717621348718937,
+      "learning_rate": 1.526486641602144e-05,
+      "loss": 1.2114,
+      "step": 1023
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.47668092412634866,
+      "learning_rate": 1.525560209469854e-05,
+      "loss": 1.0989,
+      "step": 1024
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6799529585341583,
+      "learning_rate": 1.5246331537008607e-05,
+      "loss": 0.8581,
+      "step": 1025
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6649525121673968,
+      "learning_rate": 1.5237054753952203e-05,
+      "loss": 1.2608,
+      "step": 1026
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.612325659149975,
+      "learning_rate": 1.522777175653728e-05,
+      "loss": 1.045,
+      "step": 1027
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6962232474522888,
+      "learning_rate": 1.5218482555779167e-05,
+      "loss": 1.3344,
+      "step": 1028
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6070212180704007,
+      "learning_rate": 1.5209187162700551e-05,
+      "loss": 1.0237,
+      "step": 1029
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7388112410716953,
+      "learning_rate": 1.5199885588331476e-05,
+      "loss": 1.2187,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.19146457216263,
+      "learning_rate": 1.5190577843709302e-05,
+      "loss": 0.7137,
+      "step": 1031
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6027572002959518,
+      "learning_rate": 1.5181263939878728e-05,
+      "loss": 1.1883,
+      "step": 1032
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7133277869259731,
+      "learning_rate": 1.517194388789176e-05,
+      "loss": 1.325,
+      "step": 1033
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.706711644518095,
+      "learning_rate": 1.5162617698807686e-05,
+      "loss": 1.0217,
+      "step": 1034
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7449900048482037,
+      "learning_rate": 1.5153285383693091e-05,
+      "loss": 1.2109,
+      "step": 1035
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.48153289785679476,
+      "learning_rate": 1.5143946953621825e-05,
+      "loss": 1.1036,
+      "step": 1036
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7225129361045097,
+      "learning_rate": 1.5134602419674986e-05,
+      "loss": 1.0614,
+      "step": 1037
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8721511102526129,
+      "learning_rate": 1.5125251792940927e-05,
+      "loss": 1.2281,
+      "step": 1038
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7993084192852423,
+      "learning_rate": 1.5115895084515226e-05,
+      "loss": 0.8622,
+      "step": 1039
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8624562486995264,
+      "learning_rate": 1.5106532305500677e-05,
+      "loss": 1.207,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7560668013948787,
+      "learning_rate": 1.5097163467007272e-05,
+      "loss": 1.1911,
+      "step": 1041
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6712024937583518,
+      "learning_rate": 1.5087788580152207e-05,
+      "loss": 1.3439,
+      "step": 1042
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6515267165464833,
+      "learning_rate": 1.5078407656059839e-05,
+      "loss": 1.1092,
+      "step": 1043
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.658920154104335,
+      "learning_rate": 1.5069020705861705e-05,
+      "loss": 0.971,
+      "step": 1044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7855259128243599,
+      "learning_rate": 1.5059627740696476e-05,
+      "loss": 1.2997,
+      "step": 1045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.602245334148906,
+      "learning_rate": 1.5050228771709975e-05,
+      "loss": 0.8929,
+      "step": 1046
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7224282840603793,
+      "learning_rate": 1.5040823810055143e-05,
+      "loss": 1.3729,
+      "step": 1047
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.39993761135169537,
+      "learning_rate": 1.5031412866892025e-05,
+      "loss": 0.9801,
+      "step": 1048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7701917616489522,
+      "learning_rate": 1.502199595338778e-05,
+      "loss": 1.1818,
+      "step": 1049
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6355415317197834,
+      "learning_rate": 1.5012573080716636e-05,
+      "loss": 0.9862,
+      "step": 1050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5792450113060941,
+      "learning_rate": 1.50031442600599e-05,
+      "loss": 1.1358,
+      "step": 1051
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6830869558738272,
+      "learning_rate": 1.4993709502605937e-05,
+      "loss": 1.0982,
+      "step": 1052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5908835678775852,
+      "learning_rate": 1.4984268819550154e-05,
+      "loss": 1.0785,
+      "step": 1053
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8030971995812586,
+      "learning_rate": 1.4974822222094991e-05,
+      "loss": 1.2376,
+      "step": 1054
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6272738616314707,
+      "learning_rate": 1.4965369721449902e-05,
+      "loss": 1.221,
+      "step": 1055
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.46936909964032497,
+      "learning_rate": 1.4955911328831357e-05,
+      "loss": 0.6715,
+      "step": 1056
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6876224941483501,
+      "learning_rate": 1.4946447055462803e-05,
+      "loss": 1.1829,
+      "step": 1057
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6735008861512458,
+      "learning_rate": 1.4936976912574675e-05,
+      "loss": 1.0265,
+      "step": 1058
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7422620517803281,
+      "learning_rate": 1.492750091140437e-05,
+      "loss": 1.2429,
+      "step": 1059
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4500513276184683,
+      "learning_rate": 1.4918019063196236e-05,
+      "loss": 1.1141,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7082524412240172,
+      "learning_rate": 1.4908531379201561e-05,
+      "loss": 0.9524,
+      "step": 1061
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7412128262314986,
+      "learning_rate": 1.4899037870678552e-05,
+      "loss": 1.2989,
+      "step": 1062
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6347626995745161,
+      "learning_rate": 1.4889538548892336e-05,
+      "loss": 0.9873,
+      "step": 1063
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5788761937545864,
+      "learning_rate": 1.4880033425114937e-05,
+      "loss": 1.1518,
+      "step": 1064
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5122748237549893,
+      "learning_rate": 1.4870522510625252e-05,
+      "loss": 0.8153,
+      "step": 1065
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8020860036417907,
+      "learning_rate": 1.4861005816709064e-05,
+      "loss": 1.2507,
+      "step": 1066
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.775264798031918,
+      "learning_rate": 1.4851483354659009e-05,
+      "loss": 0.9499,
+      "step": 1067
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7817590176049422,
+      "learning_rate": 1.4841955135774562e-05,
+      "loss": 1.0166,
+      "step": 1068
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7583846479187806,
+      "learning_rate": 1.4832421171362037e-05,
+      "loss": 1.3088,
+      "step": 1069
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7295504980526636,
+      "learning_rate": 1.4822881472734563e-05,
+      "loss": 1.0614,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6979214510821162,
+      "learning_rate": 1.481333605121207e-05,
+      "loss": 1.2093,
+      "step": 1071
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5770329921444182,
+      "learning_rate": 1.4803784918121282e-05,
+      "loss": 1.0629,
+      "step": 1072
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6819826965910243,
+      "learning_rate": 1.4794228084795704e-05,
+      "loss": 1.3009,
+      "step": 1073
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.47491163507424794,
+      "learning_rate": 1.4784665562575596e-05,
+      "loss": 0.9439,
+      "step": 1074
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8239498777548107,
+      "learning_rate": 1.4775097362807973e-05,
+      "loss": 0.874,
+      "step": 1075
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7217471676141414,
+      "learning_rate": 1.4765523496846592e-05,
+      "loss": 1.2335,
+      "step": 1076
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6612588804110043,
+      "learning_rate": 1.4755943976051928e-05,
+      "loss": 1.3106,
+      "step": 1077
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6738538500523209,
+      "learning_rate": 1.4746358811791164e-05,
+      "loss": 1.0705,
+      "step": 1078
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5945300174045554,
+      "learning_rate": 1.4736768015438185e-05,
+      "loss": 1.0409,
+      "step": 1079
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.611293433732696,
+      "learning_rate": 1.472717159837356e-05,
+      "loss": 1.0974,
+      "step": 1080
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6909511773432386,
+      "learning_rate": 1.471756957198452e-05,
+      "loss": 1.2043,
+      "step": 1081
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7435097300825215,
+      "learning_rate": 1.4707961947664958e-05,
+      "loss": 0.982,
+      "step": 1082
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5101789246542716,
+      "learning_rate": 1.4698348736815407e-05,
+      "loss": 0.9373,
+      "step": 1083
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7114958596494187,
+      "learning_rate": 1.4688729950843035e-05,
+      "loss": 1.0359,
+      "step": 1084
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5347697769461668,
+      "learning_rate": 1.4679105601161614e-05,
+      "loss": 1.2211,
+      "step": 1085
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6899915563998417,
+      "learning_rate": 1.4669475699191531e-05,
+      "loss": 1.0319,
+      "step": 1086
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6287834088080008,
+      "learning_rate": 1.465984025635975e-05,
+      "loss": 1.1006,
+      "step": 1087
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8644829038616366,
+      "learning_rate": 1.4650199284099817e-05,
+      "loss": 1.276,
+      "step": 1088
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.738916712799064,
+      "learning_rate": 1.4640552793851842e-05,
+      "loss": 1.0655,
+      "step": 1089
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6995724741959414,
+      "learning_rate": 1.4630900797062473e-05,
+      "loss": 1.2073,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6297976088616022,
+      "learning_rate": 1.4621243305184898e-05,
+      "loss": 1.1002,
+      "step": 1091
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8214462775359825,
+      "learning_rate": 1.4611580329678822e-05,
+      "loss": 1.2626,
+      "step": 1092
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5232971862655486,
+      "learning_rate": 1.4601911882010465e-05,
+      "loss": 0.6571,
+      "step": 1093
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6434008493487356,
+      "learning_rate": 1.4592237973652532e-05,
+      "loss": 0.9998,
+      "step": 1094
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7897134121026835,
+      "learning_rate": 1.458255861608421e-05,
+      "loss": 1.2089,
+      "step": 1095
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6343007490120041,
+      "learning_rate": 1.4572873820791159e-05,
+      "loss": 1.067,
+      "step": 1096
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7910339513066708,
+      "learning_rate": 1.4563183599265476e-05,
+      "loss": 1.287,
+      "step": 1097
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5049237794452476,
+      "learning_rate": 1.4553487963005712e-05,
+      "loss": 1.1761,
+      "step": 1098
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7220646306919767,
+      "learning_rate": 1.4543786923516835e-05,
+      "loss": 1.1656,
+      "step": 1099
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5724722515544172,
+      "learning_rate": 1.453408049231023e-05,
+      "loss": 0.7261,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8911573879490616,
+      "learning_rate": 1.4524368680903676e-05,
+      "loss": 1.0981,
+      "step": 1101
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7186384141909375,
+      "learning_rate": 1.4514651500821333e-05,
+      "loss": 1.1091,
+      "step": 1102
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6692130344035534,
+      "learning_rate": 1.4504928963593738e-05,
+      "loss": 0.979,
+      "step": 1103
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6799476167156224,
+      "learning_rate": 1.4495201080757786e-05,
+      "loss": 1.3113,
+      "step": 1104
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.698090118993706,
+      "learning_rate": 1.4485467863856704e-05,
+      "loss": 1.0545,
+      "step": 1105
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7197444378594662,
+      "learning_rate": 1.4475729324440063e-05,
+      "loss": 1.0124,
+      "step": 1106
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7471491922524424,
+      "learning_rate": 1.4465985474063741e-05,
+      "loss": 1.1631,
+      "step": 1107
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5914533359942418,
+      "learning_rate": 1.4456236324289919e-05,
+      "loss": 1.2899,
+      "step": 1108
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.653415889602487,
+      "learning_rate": 1.4446481886687067e-05,
+      "loss": 0.8839,
+      "step": 1109
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6906402872018644,
+      "learning_rate": 1.4436722172829931e-05,
+      "loss": 0.9949,
+      "step": 1110
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7892861962578261,
+      "learning_rate": 1.4426957194299516e-05,
+      "loss": 1.3287,
+      "step": 1111
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.615711376190678,
+      "learning_rate": 1.4417186962683072e-05,
+      "loss": 1.0965,
+      "step": 1112
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6035024196640562,
+      "learning_rate": 1.4407411489574088e-05,
+      "loss": 0.9307,
+      "step": 1113
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7999037488786508,
+      "learning_rate": 1.4397630786572272e-05,
+      "loss": 1.3351,
+      "step": 1114
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6726531710745075,
+      "learning_rate": 1.4387844865283534e-05,
+      "loss": 0.9041,
+      "step": 1115
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7442045114552803,
+      "learning_rate": 1.4378053737319975e-05,
+      "loss": 1.106,
+      "step": 1116
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5945185566175396,
+      "learning_rate": 1.436825741429988e-05,
+      "loss": 1.007,
+      "step": 1117
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6214768924864732,
+      "learning_rate": 1.4358455907847695e-05,
+      "loss": 1.1621,
+      "step": 1118
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.593729012431377,
+      "learning_rate": 1.4348649229594019e-05,
+      "loss": 0.8465,
+      "step": 1119
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6450448358627279,
+      "learning_rate": 1.4338837391175582e-05,
+      "loss": 1.1463,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6653766316873773,
+      "learning_rate": 1.4329020404235245e-05,
+      "loss": 1.3137,
+      "step": 1121
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.745870043535574,
+      "learning_rate": 1.4319198280421973e-05,
+      "loss": 0.9931,
+      "step": 1122
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.57231374737015,
+      "learning_rate": 1.430937103139083e-05,
+      "loss": 1.1594,
+      "step": 1123
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6161356160438434,
+      "learning_rate": 1.429953866880296e-05,
+      "loss": 1.0893,
+      "step": 1124
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7180972118373872,
+      "learning_rate": 1.428970120432557e-05,
+      "loss": 0.924,
+      "step": 1125
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6861300359407946,
+      "learning_rate": 1.427985864963193e-05,
+      "loss": 1.1485,
+      "step": 1126
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6578147650322156,
+      "learning_rate": 1.4270011016401345e-05,
+      "loss": 0.7149,
+      "step": 1127
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6355944048193626,
+      "learning_rate": 1.4260158316319146e-05,
+      "loss": 1.2222,
+      "step": 1128
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7734608914605209,
+      "learning_rate": 1.425030056107668e-05,
+      "loss": 0.9954,
+      "step": 1129
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.724927688988428,
+      "learning_rate": 1.4240437762371287e-05,
+      "loss": 1.3179,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5863441950979631,
+      "learning_rate": 1.4230569931906295e-05,
+      "loss": 1.2413,
+      "step": 1131
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5255189387197852,
+      "learning_rate": 1.4220697081391002e-05,
+      "loss": 1.0527,
+      "step": 1132
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7222828758513119,
+      "learning_rate": 1.4210819222540665e-05,
+      "loss": 1.2249,
+      "step": 1133
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9215845931528397,
+      "learning_rate": 1.4200936367076481e-05,
+      "loss": 0.9569,
+      "step": 1134
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.46383190734922214,
+      "learning_rate": 1.4191048526725575e-05,
+      "loss": 1.1704,
+      "step": 1135
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8192916575829242,
+      "learning_rate": 1.4181155713220994e-05,
+      "loss": 0.6418,
+      "step": 1136
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6922878681149328,
+      "learning_rate": 1.4171257938301679e-05,
+      "loss": 1.0695,
+      "step": 1137
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8613287253574752,
+      "learning_rate": 1.416135521371246e-05,
+      "loss": 1.1059,
+      "step": 1138
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6370130307787831,
+      "learning_rate": 1.415144755120404e-05,
+      "loss": 0.9272,
+      "step": 1139
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.740994046996833,
+      "learning_rate": 1.4141534962532986e-05,
+      "loss": 1.3288,
+      "step": 1140
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6560758058569891,
+      "learning_rate": 1.4131617459461706e-05,
+      "loss": 1.1499,
+      "step": 1141
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7526151079184843,
+      "learning_rate": 1.4121695053758437e-05,
+      "loss": 1.138,
+      "step": 1142
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7497338111240673,
+      "learning_rate": 1.4111767757197243e-05,
+      "loss": 1.0639,
+      "step": 1143
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6462315006753913,
+      "learning_rate": 1.4101835581557983e-05,
+      "loss": 1.1825,
+      "step": 1144
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9235682001170321,
+      "learning_rate": 1.4091898538626306e-05,
+      "loss": 0.798,
+      "step": 1145
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6758130386239178,
+      "learning_rate": 1.4081956640193639e-05,
+      "loss": 1.0135,
+      "step": 1146
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6898944359547435,
+      "learning_rate": 1.4072009898057173e-05,
+      "loss": 1.1811,
+      "step": 1147
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8238754226854186,
+      "learning_rate": 1.4062058324019842e-05,
+      "loss": 0.9752,
+      "step": 1148
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6411864705265709,
+      "learning_rate": 1.4052101929890317e-05,
+      "loss": 1.1363,
+      "step": 1149
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6628858543493821,
+      "learning_rate": 1.4042140727482985e-05,
+      "loss": 1.2062,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6514747304730117,
+      "learning_rate": 1.4032174728617943e-05,
+      "loss": 0.9949,
+      "step": 1151
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.626336610795871,
+      "learning_rate": 1.4022203945120971e-05,
+      "loss": 1.2652,
+      "step": 1152
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7434942016320347,
+      "learning_rate": 1.4012228388823543e-05,
+      "loss": 0.7672,
+      "step": 1153
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7192909769356365,
+      "learning_rate": 1.400224807156278e-05,
+      "loss": 1.2708,
+      "step": 1154
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5397975399956528,
+      "learning_rate": 1.3992263005181458e-05,
+      "loss": 1.1361,
+      "step": 1155
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.571282309232512,
+      "learning_rate": 1.3982273201527994e-05,
+      "loss": 1.1446,
+      "step": 1156
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8163047527051556,
+      "learning_rate": 1.3972278672456416e-05,
+      "loss": 1.2567,
+      "step": 1157
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6622137950829772,
+      "learning_rate": 1.3962279429826367e-05,
+      "loss": 1.0625,
+      "step": 1158
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7619666704306116,
+      "learning_rate": 1.3952275485503077e-05,
+      "loss": 1.2403,
+      "step": 1159
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5379848477288309,
+      "learning_rate": 1.3942266851357365e-05,
+      "loss": 1.1213,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6988283531325076,
+      "learning_rate": 1.3932253539265603e-05,
+      "loss": 1.1578,
+      "step": 1161
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5329491663862629,
+      "learning_rate": 1.3922235561109725e-05,
+      "loss": 0.9169,
+      "step": 1162
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6591060430710877,
+      "learning_rate": 1.3912212928777192e-05,
+      "loss": 0.736,
+      "step": 1163
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6591208976028379,
+      "learning_rate": 1.3902185654160995e-05,
+      "loss": 1.3174,
+      "step": 1164
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6055201826322275,
+      "learning_rate": 1.3892153749159629e-05,
+      "loss": 1.0157,
+      "step": 1165
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7086762754672035,
+      "learning_rate": 1.3882117225677087e-05,
+      "loss": 1.2232,
+      "step": 1166
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.63965492928474,
+      "learning_rate": 1.3872076095622842e-05,
+      "loss": 1.1302,
+      "step": 1167
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7044210935698713,
+      "learning_rate": 1.386203037091183e-05,
+      "loss": 1.0831,
+      "step": 1168
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6827192671999947,
+      "learning_rate": 1.3851980063464436e-05,
+      "loss": 1.2516,
+      "step": 1169
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6327821128357803,
+      "learning_rate": 1.3841925185206497e-05,
+      "loss": 1.1218,
+      "step": 1170
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.610397585329582,
+      "learning_rate": 1.383186574806926e-05,
+      "loss": 1.1946,
+      "step": 1171
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.45077151393324866,
+      "learning_rate": 1.3821801763989385e-05,
+      "loss": 0.9458,
+      "step": 1172
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6733410943353803,
+      "learning_rate": 1.3811733244908932e-05,
+      "loss": 1.3127,
+      "step": 1173
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8994244975838934,
+      "learning_rate": 1.3801660202775335e-05,
+      "loss": 0.9263,
+      "step": 1174
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6602209986636345,
+      "learning_rate": 1.3791582649541404e-05,
+      "loss": 1.1239,
+      "step": 1175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5915712154970959,
+      "learning_rate": 1.3781500597165294e-05,
+      "loss": 1.1383,
+      "step": 1176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6668871327519896,
+      "learning_rate": 1.3771414057610505e-05,
+      "loss": 1.0634,
+      "step": 1177
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6645267000958958,
+      "learning_rate": 1.3761323042845855e-05,
+      "loss": 1.1867,
+      "step": 1178
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7907247442759607,
+      "learning_rate": 1.3751227564845475e-05,
+      "loss": 1.1547,
+      "step": 1179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.48346241913913707,
+      "learning_rate": 1.3741127635588795e-05,
+      "loss": 0.8575,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6303566990821807,
+      "learning_rate": 1.3731023267060526e-05,
+      "loss": 1.0786,
+      "step": 1181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.813846694589603,
+      "learning_rate": 1.3720914471250644e-05,
+      "loss": 1.0542,
+      "step": 1182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7448261794139879,
+      "learning_rate": 1.3710801260154378e-05,
+      "loss": 1.2264,
+      "step": 1183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6307249956275212,
+      "learning_rate": 1.37006836457722e-05,
+      "loss": 0.9388,
+      "step": 1184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6416591140205378,
+      "learning_rate": 1.3690561640109807e-05,
+      "loss": 1.2049,
+      "step": 1185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7111872125359983,
+      "learning_rate": 1.3680435255178104e-05,
+      "loss": 1.0563,
+      "step": 1186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6677593103822954,
+      "learning_rate": 1.3670304502993192e-05,
+      "loss": 0.8652,
+      "step": 1187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7358131809838315,
+      "learning_rate": 1.3660169395576357e-05,
+      "loss": 1.2853,
+      "step": 1188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.573828471658806,
+      "learning_rate": 1.3650029944954048e-05,
+      "loss": 0.6497,
+      "step": 1189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5774588670068339,
+      "learning_rate": 1.3639886163157878e-05,
+      "loss": 1.1925,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7259310084626268,
+      "learning_rate": 1.3629738062224589e-05,
+      "loss": 1.0288,
+      "step": 1191
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6702191018541447,
+      "learning_rate": 1.361958565419605e-05,
+      "loss": 1.0353,
+      "step": 1192
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6588222558871119,
+      "learning_rate": 1.3609428951119248e-05,
+      "loss": 1.0888,
+      "step": 1193
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6546310778372426,
+      "learning_rate": 1.3599267965046257e-05,
+      "loss": 1.1413,
+      "step": 1194
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6189868703007219,
+      "learning_rate": 1.3589102708034236e-05,
+      "loss": 1.239,
+      "step": 1195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5639035248465848,
+      "learning_rate": 1.3578933192145421e-05,
+      "loss": 1.1544,
+      "step": 1196
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6486504246539861,
+      "learning_rate": 1.3568759429447087e-05,
+      "loss": 1.2177,
+      "step": 1197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6692240779339084,
+      "learning_rate": 1.3558581432011559e-05,
+      "loss": 0.702,
+      "step": 1198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.603135075628886,
+      "learning_rate": 1.3548399211916183e-05,
+      "loss": 1.0488,
+      "step": 1199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6714308323179666,
+      "learning_rate": 1.3538212781243317e-05,
+      "loss": 1.0787,
+      "step": 1200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6273497334363648,
+      "learning_rate": 1.3528022152080314e-05,
+      "loss": 1.1826,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6123122325311134,
+      "learning_rate": 1.3517827336519505e-05,
+      "loss": 1.215,
+      "step": 1202
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6259902873969785,
+      "learning_rate": 1.35076283466582e-05,
+      "loss": 1.0728,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6965169620962551,
+      "learning_rate": 1.3497425194598654e-05,
+      "loss": 1.0599,
+      "step": 1204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6299187898920778,
+      "learning_rate": 1.3487217892448059e-05,
+      "loss": 0.8533,
+      "step": 1205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6033150578526557,
+      "learning_rate": 1.3477006452318538e-05,
+      "loss": 0.8499,
+      "step": 1206
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6465035492734666,
+      "learning_rate": 1.346679088632712e-05,
+      "loss": 1.2672,
+      "step": 1207
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6237198421160911,
+      "learning_rate": 1.345657120659573e-05,
+      "loss": 1.1325,
+      "step": 1208
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7890529633388019,
+      "learning_rate": 1.3446347425251176e-05,
+      "loss": 1.2081,
+      "step": 1209
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4963395003483872,
+      "learning_rate": 1.3436119554425133e-05,
+      "loss": 1.0624,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6218222501497052,
+      "learning_rate": 1.3425887606254126e-05,
+      "loss": 0.8169,
+      "step": 1211
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8219439227354162,
+      "learning_rate": 1.3415651592879522e-05,
+      "loss": 1.2959,
+      "step": 1212
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6488751708037896,
+      "learning_rate": 1.3405411526447511e-05,
+      "loss": 1.0283,
+      "step": 1213
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7166281735819375,
+      "learning_rate": 1.339516741910909e-05,
+      "loss": 1.3604,
+      "step": 1214
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7235201072105357,
+      "learning_rate": 1.338491928302005e-05,
+      "loss": 0.9865,
+      "step": 1215
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6021665919152875,
+      "learning_rate": 1.3374667130340968e-05,
+      "loss": 0.8442,
+      "step": 1216
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.820177210906781,
+      "learning_rate": 1.3364410973237185e-05,
+      "loss": 0.8467,
+      "step": 1217
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8347197566488663,
+      "learning_rate": 1.3354150823878792e-05,
+      "loss": 1.2461,
+      "step": 1218
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.76443706746482,
+      "learning_rate": 1.3343886694440615e-05,
+      "loss": 1.2156,
+      "step": 1219
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6743550579822668,
+      "learning_rate": 1.333361859710221e-05,
+      "loss": 1.0952,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8020873669210526,
+      "learning_rate": 1.3323346544047835e-05,
+      "loss": 1.1894,
+      "step": 1221
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5074662067428154,
+      "learning_rate": 1.3313070547466445e-05,
+      "loss": 1.0746,
+      "step": 1222
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7504338066404367,
+      "learning_rate": 1.3302790619551673e-05,
+      "loss": 1.206,
+      "step": 1223
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6915718869691989,
+      "learning_rate": 1.3292506772501819e-05,
+      "loss": 1.012,
+      "step": 1224
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5571392942574636,
+      "learning_rate": 1.3282219018519832e-05,
+      "loss": 0.828,
+      "step": 1225
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8157619375856772,
+      "learning_rate": 1.3271927369813292e-05,
+      "loss": 1.2492,
+      "step": 1226
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7060949987447763,
+      "learning_rate": 1.326163183859441e-05,
+      "loss": 1.0724,
+      "step": 1227
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5839847222213234,
+      "learning_rate": 1.3251332437080003e-05,
+      "loss": 1.3384,
+      "step": 1228
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6306606529520874,
+      "learning_rate": 1.3241029177491472e-05,
+      "loss": 1.0642,
+      "step": 1229
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6475351425362795,
+      "learning_rate": 1.3230722072054803e-05,
+      "loss": 1.0108,
+      "step": 1230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.832505246546327,
+      "learning_rate": 1.3220411133000544e-05,
+      "loss": 1.1898,
+      "step": 1231
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6379160643975103,
+      "learning_rate": 1.321009637256379e-05,
+      "loss": 0.9636,
+      "step": 1232
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5494123250633818,
+      "learning_rate": 1.3199777802984177e-05,
+      "loss": 0.961,
+      "step": 1233
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.47711979347549993,
+      "learning_rate": 1.318945543650585e-05,
+      "loss": 1.0835,
+      "step": 1234
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7551317426820015,
+      "learning_rate": 1.3179129285377472e-05,
+      "loss": 1.2442,
+      "step": 1235
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6574001403368904,
+      "learning_rate": 1.3168799361852184e-05,
+      "loss": 0.9906,
+      "step": 1236
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6265098996697052,
+      "learning_rate": 1.3158465678187616e-05,
+      "loss": 1.1078,
+      "step": 1237
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7206843432328163,
+      "learning_rate": 1.314812824664585e-05,
+      "loss": 1.1014,
+      "step": 1238
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7078269361995053,
+      "learning_rate": 1.313778707949342e-05,
+      "loss": 1.0892,
+      "step": 1239
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8063952708858396,
+      "learning_rate": 1.3127442189001296e-05,
+      "loss": 1.2227,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6645277410013207,
+      "learning_rate": 1.3117093587444858e-05,
+      "loss": 1.2421,
+      "step": 1241
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.48447285630592624,
+      "learning_rate": 1.3106741287103891e-05,
+      "loss": 0.7889,
+      "step": 1242
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7475246024444787,
+      "learning_rate": 1.3096385300262578e-05,
+      "loss": 0.9565,
+      "step": 1243
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.641780369047058,
+      "learning_rate": 1.3086025639209465e-05,
+      "loss": 1.0514,
+      "step": 1244
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8618652146767543,
+      "learning_rate": 1.3075662316237466e-05,
+      "loss": 1.2224,
+      "step": 1245
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.43634638470412346,
+      "learning_rate": 1.3065295343643833e-05,
+      "loss": 1.1228,
+      "step": 1246
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7769822249109041,
+      "learning_rate": 1.3054924733730161e-05,
+      "loss": 1.0887,
+      "step": 1247
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6230198016909922,
+      "learning_rate": 1.3044550498802346e-05,
+      "loss": 1.0893,
+      "step": 1248
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8717987267891015,
+      "learning_rate": 1.3034172651170591e-05,
+      "loss": 0.9403,
+      "step": 1249
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7158672214009872,
+      "learning_rate": 1.3023791203149396e-05,
+      "loss": 1.2021,
+      "step": 1250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.47110398687234095,
+      "learning_rate": 1.3013406167057521e-05,
+      "loss": 0.7917,
+      "step": 1251
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6583910270894365,
+      "learning_rate": 1.300301755521798e-05,
+      "loss": 1.2582,
+      "step": 1252
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7018983838487509,
+      "learning_rate": 1.2992625379958048e-05,
+      "loss": 0.9765,
+      "step": 1253
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6613238497581861,
+      "learning_rate": 1.2982229653609215e-05,
+      "loss": 1.144,
+      "step": 1254
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7232971319385698,
+      "learning_rate": 1.2971830388507183e-05,
+      "loss": 1.0821,
+      "step": 1255
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0638509345575333,
+      "learning_rate": 1.296142759699186e-05,
+      "loss": 0.9874,
+      "step": 1256
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7290664593769742,
+      "learning_rate": 1.2951021291407338e-05,
+      "loss": 1.2418,
+      "step": 1257
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4782407862911107,
+      "learning_rate": 1.2940611484101875e-05,
+      "loss": 1.0856,
+      "step": 1258
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.809019437119441,
+      "learning_rate": 1.2930198187427888e-05,
+      "loss": 1.4137,
+      "step": 1259
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4623585112749311,
+      "learning_rate": 1.291978141374193e-05,
+      "loss": 0.8933,
+      "step": 1260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7404831350466217,
+      "learning_rate": 1.2909361175404684e-05,
+      "loss": 1.0975,
+      "step": 1261
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6255749018493926,
+      "learning_rate": 1.2898937484780944e-05,
+      "loss": 1.1495,
+      "step": 1262
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7878970866968047,
+      "learning_rate": 1.2888510354239598e-05,
+      "loss": 0.9818,
+      "step": 1263
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6257630635596145,
+      "learning_rate": 1.287807979615362e-05,
+      "loss": 0.9946,
+      "step": 1264
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6158618636541043,
+      "learning_rate": 1.2867645822900043e-05,
+      "loss": 1.0802,
+      "step": 1265
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6713499420154545,
+      "learning_rate": 1.285720844685996e-05,
+      "loss": 1.2176,
+      "step": 1266
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6599875672031015,
+      "learning_rate": 1.2846767680418502e-05,
+      "loss": 0.9939,
+      "step": 1267
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6943709978020111,
+      "learning_rate": 1.283632353596482e-05,
+      "loss": 1.067,
+      "step": 1268
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5267326663518901,
+      "learning_rate": 1.282587602589207e-05,
+      "loss": 0.8582,
+      "step": 1269
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6965719080323455,
+      "learning_rate": 1.281542516259741e-05,
+      "loss": 1.0059,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.45034454711850436,
+      "learning_rate": 1.2804970958481973e-05,
+      "loss": 1.2239,
+      "step": 1271
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6797072297951187,
+      "learning_rate": 1.2794513425950851e-05,
+      "loss": 0.9219,
+      "step": 1272
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6379334976991707,
+      "learning_rate": 1.2784052577413097e-05,
+      "loss": 1.2856,
+      "step": 1273
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6424119773223786,
+      "learning_rate": 1.2773588425281692e-05,
+      "loss": 0.9798,
+      "step": 1274
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7800550116247131,
+      "learning_rate": 1.2763120981973532e-05,
+      "loss": 1.0529,
+      "step": 1275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7223098347753881,
+      "learning_rate": 1.2752650259909425e-05,
+      "loss": 1.2077,
+      "step": 1276
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6516966846987423,
+      "learning_rate": 1.2742176271514076e-05,
+      "loss": 0.8988,
+      "step": 1277
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5108080717625647,
+      "learning_rate": 1.273169902921605e-05,
+      "loss": 0.9018,
+      "step": 1278
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6366683187952542,
+      "learning_rate": 1.272121854544778e-05,
+      "loss": 0.917,
+      "step": 1279
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6814327959648425,
+      "learning_rate": 1.2710734832645557e-05,
+      "loss": 0.9456,
+      "step": 1280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6287641465138618,
+      "learning_rate": 1.2700247903249485e-05,
+      "loss": 1.2051,
+      "step": 1281
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9479039858314011,
+      "learning_rate": 1.2689757769703495e-05,
+      "loss": 1.0309,
+      "step": 1282
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.617293878561528,
+      "learning_rate": 1.2679264444455322e-05,
+      "loss": 1.2917,
+      "step": 1283
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.45285892648456083,
+      "learning_rate": 1.266876793995648e-05,
+      "loss": 1.1711,
+      "step": 1284
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7690006031707403,
+      "learning_rate": 1.2658268268662259e-05,
+      "loss": 1.1603,
+      "step": 1285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4989780339046183,
+      "learning_rate": 1.2647765443031715e-05,
+      "loss": 0.7338,
+      "step": 1286
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6990094054023813,
+      "learning_rate": 1.2637259475527635e-05,
+      "loss": 1.0651,
+      "step": 1287
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6646537673481084,
+      "learning_rate": 1.2626750378616539e-05,
+      "loss": 1.056,
+      "step": 1288
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6310082999310661,
+      "learning_rate": 1.2616238164768659e-05,
+      "loss": 1.0174,
+      "step": 1289
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.67038223869757,
+      "learning_rate": 1.2605722846457934e-05,
+      "loss": 1.2606,
+      "step": 1290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5952716181224449,
+      "learning_rate": 1.2595204436161976e-05,
+      "loss": 0.9781,
+      "step": 1291
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8075353319716565,
+      "learning_rate": 1.2584682946362067e-05,
+      "loss": 1.0391,
+      "step": 1292
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6956646785107894,
+      "learning_rate": 1.2574158389543152e-05,
+      "loss": 1.1154,
+      "step": 1293
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5818174725793924,
+      "learning_rate": 1.2563630778193805e-05,
+      "loss": 1.2466,
+      "step": 1294
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5459963580450579,
+      "learning_rate": 1.255310012480623e-05,
+      "loss": 0.8438,
+      "step": 1295
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.717520394117697,
+      "learning_rate": 1.2542566441876237e-05,
+      "loss": 0.9854,
+      "step": 1296
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6564156435387603,
+      "learning_rate": 1.253202974190324e-05,
+      "loss": 1.2607,
+      "step": 1297
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4789138733501013,
+      "learning_rate": 1.2521490037390224e-05,
+      "loss": 1.1255,
+      "step": 1298
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5912570860721738,
+      "learning_rate": 1.251094734084374e-05,
+      "loss": 0.9253,
+      "step": 1299
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7753057988708718,
+      "learning_rate": 1.2500401664773894e-05,
+      "loss": 1.2575,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6756285581194484,
+      "learning_rate": 1.2489853021694323e-05,
+      "loss": 0.8751,
+      "step": 1301
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6719584893910867,
+      "learning_rate": 1.2479301424122186e-05,
+      "loss": 1.1527,
+      "step": 1302
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6137888205882567,
+      "learning_rate": 1.2468746884578149e-05,
+      "loss": 1.1152,
+      "step": 1303
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.785072195484246,
+      "learning_rate": 1.245818941558637e-05,
+      "loss": 1.0575,
+      "step": 1304
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6520146582951981,
+      "learning_rate": 1.2447629029674482e-05,
+      "loss": 0.8703,
+      "step": 1305
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8471297399346396,
+      "learning_rate": 1.2437065739373573e-05,
+      "loss": 1.1419,
+      "step": 1306
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6226998614059672,
+      "learning_rate": 1.2426499557218191e-05,
+      "loss": 1.2821,
+      "step": 1307
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7182943852465148,
+      "learning_rate": 1.2415930495746301e-05,
+      "loss": 1.0276,
+      "step": 1308
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4776669371370162,
+      "learning_rate": 1.2405358567499295e-05,
+      "loss": 1.1606,
+      "step": 1309
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6023514205154299,
+      "learning_rate": 1.2394783785021964e-05,
+      "loss": 1.1125,
+      "step": 1310
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6720788246361755,
+      "learning_rate": 1.2384206160862484e-05,
+      "loss": 0.9136,
+      "step": 1311
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6707238450529656,
+      "learning_rate": 1.2373625707572401e-05,
+      "loss": 1.1451,
+      "step": 1312
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.61867318353924,
+      "learning_rate": 1.2363042437706623e-05,
+      "loss": 0.7123,
+      "step": 1313
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7034528129039623,
+      "learning_rate": 1.2352456363823397e-05,
+      "loss": 1.2631,
+      "step": 1314
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7865610449248472,
+      "learning_rate": 1.2341867498484303e-05,
+      "loss": 1.0117,
+      "step": 1315
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6881958355400845,
+      "learning_rate": 1.2331275854254222e-05,
+      "loss": 1.3358,
+      "step": 1316
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5774627741125089,
+      "learning_rate": 1.2320681443701345e-05,
+      "loss": 1.1735,
+      "step": 1317
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6394198485817109,
+      "learning_rate": 1.2310084279397137e-05,
+      "loss": 1.1439,
+      "step": 1318
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6487861803659727,
+      "learning_rate": 1.2299484373916333e-05,
+      "loss": 1.2004,
+      "step": 1319
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6954554563693431,
+      "learning_rate": 1.2288881739836925e-05,
+      "loss": 0.9423,
+      "step": 1320
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4675752285908601,
+      "learning_rate": 1.2278276389740134e-05,
+      "loss": 1.1494,
+      "step": 1321
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6194398741680138,
+      "learning_rate": 1.2267668336210411e-05,
+      "loss": 0.7636,
+      "step": 1322
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6936954416083271,
+      "learning_rate": 1.2257057591835416e-05,
+      "loss": 1.0078,
+      "step": 1323
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6015564162816882,
+      "learning_rate": 1.2246444169205994e-05,
+      "loss": 1.0524,
+      "step": 1324
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6440290254959613,
+      "learning_rate": 1.2235828080916174e-05,
+      "loss": 1.0289,
+      "step": 1325
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6506006745083538,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 1.1797,
+      "step": 1326
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6661413977301752,
+      "learning_rate": 1.2214587957747253e-05,
+      "loss": 1.1879,
+      "step": 1327
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7245745572639448,
+      "learning_rate": 1.2203963948071964e-05,
+      "loss": 1.1156,
+      "step": 1328
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6321754658924643,
+      "learning_rate": 1.2193337323143867e-05,
+      "loss": 1.0907,
+      "step": 1329
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7036855652742022,
+      "learning_rate": 1.2182708095572667e-05,
+      "loss": 1.1348,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7212855227770205,
+      "learning_rate": 1.2172076277971132e-05,
+      "loss": 0.8277,
+      "step": 1331
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7531732696623763,
+      "learning_rate": 1.2161441882955127e-05,
+      "loss": 0.9978,
+      "step": 1332
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4769216096734017,
+      "learning_rate": 1.2150804923143562e-05,
+      "loss": 1.1611,
+      "step": 1333
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9088090420579229,
+      "learning_rate": 1.2140165411158395e-05,
+      "loss": 0.8924,
+      "step": 1334
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6641048169261382,
+      "learning_rate": 1.2129523359624615e-05,
+      "loss": 1.3098,
+      "step": 1335
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5718855878741727,
+      "learning_rate": 1.2118878781170213e-05,
+      "loss": 0.9722,
+      "step": 1336
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6886074520757323,
+      "learning_rate": 1.2108231688426197e-05,
+      "loss": 0.983,
+      "step": 1337
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6575725519347373,
+      "learning_rate": 1.2097582094026544e-05,
+      "loss": 1.2673,
+      "step": 1338
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6815492051212265,
+      "learning_rate": 1.2086930010608198e-05,
+      "loss": 0.7608,
+      "step": 1339
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7044494045174514,
+      "learning_rate": 1.207627545081107e-05,
+      "loss": 1.2791,
+      "step": 1340
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5666503022953212,
+      "learning_rate": 1.2065618427278e-05,
+      "loss": 1.107,
+      "step": 1341
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6099820000393962,
+      "learning_rate": 1.2054958952654751e-05,
+      "loss": 1.1336,
+      "step": 1342
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6976316401731115,
+      "learning_rate": 1.2044297039589999e-05,
+      "loss": 1.1842,
+      "step": 1343
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6770835047679008,
+      "learning_rate": 1.203363270073531e-05,
+      "loss": 0.9342,
+      "step": 1344
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7575018481293777,
+      "learning_rate": 1.2022965948745129e-05,
+      "loss": 1.2424,
+      "step": 1345
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4260001012090418,
+      "learning_rate": 1.2012296796276765e-05,
+      "loss": 1.0799,
+      "step": 1346
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6584084491470842,
+      "learning_rate": 1.200162525599038e-05,
+      "loss": 1.1923,
+      "step": 1347
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5969272512606041,
+      "learning_rate": 1.1990951340548961e-05,
+      "loss": 0.9296,
+      "step": 1348
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5648087266528528,
+      "learning_rate": 1.1980275062618316e-05,
+      "loss": 0.7211,
+      "step": 1349
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6495217316433443,
+      "learning_rate": 1.1969596434867063e-05,
+      "loss": 1.2465,
+      "step": 1350
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5920365518094078,
+      "learning_rate": 1.19589154699666e-05,
+      "loss": 1.0109,
+      "step": 1351
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8030407170706492,
+      "learning_rate": 1.1948232180591097e-05,
+      "loss": 1.2732,
+      "step": 1352
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5780497523248272,
+      "learning_rate": 1.1937546579417492e-05,
+      "loss": 1.1123,
+      "step": 1353
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6485410735237879,
+      "learning_rate": 1.1926858679125461e-05,
+      "loss": 1.0053,
+      "step": 1354
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7224122526563094,
+      "learning_rate": 1.1916168492397408e-05,
+      "loss": 1.2558,
+      "step": 1355
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6333499665927491,
+      "learning_rate": 1.190547603191845e-05,
+      "loss": 1.0619,
+      "step": 1356
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6014225238952144,
+      "learning_rate": 1.1894781310376397e-05,
+      "loss": 1.1116,
+      "step": 1357
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3865010917243142,
+      "learning_rate": 1.1884084340461757e-05,
+      "loss": 0.9347,
+      "step": 1358
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6809244126830575,
+      "learning_rate": 1.1873385134867687e-05,
+      "loss": 1.2668,
+      "step": 1359
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7367284373783829,
+      "learning_rate": 1.186268370629001e-05,
+      "loss": 1.019,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6075355682873748,
+      "learning_rate": 1.1851980067427182e-05,
+      "loss": 1.2079,
+      "step": 1361
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6744350841980605,
+      "learning_rate": 1.1841274230980284e-05,
+      "loss": 1.0873,
+      "step": 1362
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5891318945217233,
+      "learning_rate": 1.1830566209653001e-05,
+      "loss": 1.0551,
+      "step": 1363
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7042383692945317,
+      "learning_rate": 1.1819856016151617e-05,
+      "loss": 1.1944,
+      "step": 1364
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6395259681195232,
+      "learning_rate": 1.1809143663184983e-05,
+      "loss": 1.1022,
+      "step": 1365
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.48472994603821323,
+      "learning_rate": 1.1798429163464524e-05,
+      "loss": 0.8677,
+      "step": 1366
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6413923598283056,
+      "learning_rate": 1.1787712529704205e-05,
+      "loss": 1.0982,
+      "step": 1367
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7054892921541949,
+      "learning_rate": 1.177699377462053e-05,
+      "loss": 0.988,
+      "step": 1368
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6495528212487423,
+      "learning_rate": 1.1766272910932508e-05,
+      "loss": 1.1451,
+      "step": 1369
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6427524342765549,
+      "learning_rate": 1.1755549951361662e-05,
+      "loss": 0.9542,
+      "step": 1370
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3944618846178395,
+      "learning_rate": 1.1744824908631997e-05,
+      "loss": 1.1781,
+      "step": 1371
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6920579881052477,
+      "learning_rate": 1.1734097795469995e-05,
+      "loss": 1.0023,
+      "step": 1372
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5759510449383966,
+      "learning_rate": 1.1723368624604585e-05,
+      "loss": 1.0354,
+      "step": 1373
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6569385658360568,
+      "learning_rate": 1.1712637408767145e-05,
+      "loss": 1.1229,
+      "step": 1374
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4789762528433423,
+      "learning_rate": 1.1701904160691477e-05,
+      "loss": 0.6445,
+      "step": 1375
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.632509485591356,
+      "learning_rate": 1.1691168893113798e-05,
+      "loss": 1.245,
+      "step": 1376
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5926407191209843,
+      "learning_rate": 1.1680431618772714e-05,
+      "loss": 0.951,
+      "step": 1377
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6430631864816272,
+      "learning_rate": 1.1669692350409223e-05,
+      "loss": 1.0506,
+      "step": 1378
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6131865064130512,
+      "learning_rate": 1.1658951100766677e-05,
+      "loss": 1.0818,
+      "step": 1379
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7301533360846921,
+      "learning_rate": 1.1648207882590788e-05,
+      "loss": 1.0967,
+      "step": 1380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7023014069699557,
+      "learning_rate": 1.1637462708629603e-05,
+      "loss": 1.2491,
+      "step": 1381
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3953970708673346,
+      "learning_rate": 1.1626715591633484e-05,
+      "loss": 1.1793,
+      "step": 1382
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6947268675372436,
+      "learning_rate": 1.1615966544355106e-05,
+      "loss": 1.1722,
+      "step": 1383
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5692672978056039,
+      "learning_rate": 1.1605215579549426e-05,
+      "loss": 0.6924,
+      "step": 1384
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7539740882216881,
+      "learning_rate": 1.1594462709973684e-05,
+      "loss": 1.0573,
+      "step": 1385
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6218369158829165,
+      "learning_rate": 1.1583707948387376e-05,
+      "loss": 1.1171,
+      "step": 1386
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5456678854805951,
+      "learning_rate": 1.1572951307552249e-05,
+      "loss": 1.1098,
+      "step": 1387
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6605873397144274,
+      "learning_rate": 1.156219280023227e-05,
+      "loss": 1.2634,
+      "step": 1388
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6498032708707829,
+      "learning_rate": 1.155143243919363e-05,
+      "loss": 0.9416,
+      "step": 1389
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6363843223631512,
+      "learning_rate": 1.1540670237204715e-05,
+      "loss": 1.151,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6001642477401777,
+      "learning_rate": 1.1529906207036095e-05,
+      "loss": 0.8752,
+      "step": 1391
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.599361284070199,
+      "learning_rate": 1.151914036146051e-05,
+      "loss": 0.8257,
+      "step": 1392
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6362011046689622,
+      "learning_rate": 1.1508372713252858e-05,
+      "loss": 1.3058,
+      "step": 1393
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6726828670311317,
+      "learning_rate": 1.149760327519017e-05,
+      "loss": 1.0095,
+      "step": 1394
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8033635504315053,
+      "learning_rate": 1.1486832060051608e-05,
+      "loss": 1.2018,
+      "step": 1395
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4394344004424297,
+      "learning_rate": 1.147605908061843e-05,
+      "loss": 1.0766,
+      "step": 1396
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6466230412693648,
+      "learning_rate": 1.1465284349674006e-05,
+      "loss": 1.0403,
+      "step": 1397
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6357249284806146,
+      "learning_rate": 1.1454507880003768e-05,
+      "loss": 1.23,
+      "step": 1398
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5818959334721676,
+      "learning_rate": 1.1443729684395225e-05,
+      "loss": 1.0029,
+      "step": 1399
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8098859788460234,
+      "learning_rate": 1.1432949775637913e-05,
+      "loss": 1.375,
+      "step": 1400
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7476011235169718,
+      "learning_rate": 1.1422168166523431e-05,
+      "loss": 0.9491,
+      "step": 1401
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5475673053295717,
+      "learning_rate": 1.1411384869845367e-05,
+      "loss": 0.7806,
+      "step": 1402
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6830177321333211,
+      "learning_rate": 1.1400599898399329e-05,
+      "loss": 1.0158,
+      "step": 1403
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7548211769335414,
+      "learning_rate": 1.1389813264982903e-05,
+      "loss": 1.1939,
+      "step": 1404
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8960708701147407,
+      "learning_rate": 1.1379024982395654e-05,
+      "loss": 1.2861,
+      "step": 1405
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5815536718118312,
+      "learning_rate": 1.1368235063439103e-05,
+      "loss": 1.0571,
+      "step": 1406
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.753286473311435,
+      "learning_rate": 1.1357443520916704e-05,
+      "loss": 1.2508,
+      "step": 1407
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.46366496082196085,
+      "learning_rate": 1.1346650367633854e-05,
+      "loss": 1.0564,
+      "step": 1408
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7711212543531224,
+      "learning_rate": 1.1335855616397844e-05,
+      "loss": 1.2127,
+      "step": 1409
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6305561770370115,
+      "learning_rate": 1.1325059280017866e-05,
+      "loss": 0.9931,
+      "step": 1410
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4783199160697449,
+      "learning_rate": 1.1314261371305006e-05,
+      "loss": 0.8495,
+      "step": 1411
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7983045319772163,
+      "learning_rate": 1.1303461903072201e-05,
+      "loss": 1.2486,
+      "step": 1412
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7626413142716281,
+      "learning_rate": 1.1292660888134242e-05,
+      "loss": 1.0577,
+      "step": 1413
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6083212191638633,
+      "learning_rate": 1.1281858339307753e-05,
+      "loss": 1.288,
+      "step": 1414
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6296904974810957,
+      "learning_rate": 1.1271054269411186e-05,
+      "loss": 1.0457,
+      "step": 1415
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5969712320208101,
+      "learning_rate": 1.1260248691264792e-05,
+      "loss": 0.8859,
+      "step": 1416
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8288826436367692,
+      "learning_rate": 1.1249441617690613e-05,
+      "loss": 1.2438,
+      "step": 1417
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7891391542059927,
+      "learning_rate": 1.1238633061512467e-05,
+      "loss": 0.9591,
+      "step": 1418
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5017066504545008,
+      "learning_rate": 1.122782303555593e-05,
+      "loss": 0.9214,
+      "step": 1419
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.45944991229620286,
+      "learning_rate": 1.1217011552648316e-05,
+      "loss": 1.0833,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8457597945526325,
+      "learning_rate": 1.1206198625618683e-05,
+      "loss": 1.2258,
+      "step": 1421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5979138542835742,
+      "learning_rate": 1.1195384267297788e-05,
+      "loss": 0.9653,
+      "step": 1422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6078935775967749,
+      "learning_rate": 1.1184568490518095e-05,
+      "loss": 1.0942,
+      "step": 1423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6816369606017822,
+      "learning_rate": 1.1173751308113741e-05,
+      "loss": 1.132,
+      "step": 1424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5359686977943143,
+      "learning_rate": 1.1162932732920547e-05,
+      "loss": 1.0512,
+      "step": 1425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8090517711069645,
+      "learning_rate": 1.115211277777597e-05,
+      "loss": 1.3722,
+      "step": 1426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.75870962078287,
+      "learning_rate": 1.1141291455519116e-05,
+      "loss": 1.0127,
+      "step": 1427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.45096708599013285,
+      "learning_rate": 1.1130468778990708e-05,
+      "loss": 0.7439,
+      "step": 1428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.584755344909728,
+      "learning_rate": 1.1119644761033079e-05,
+      "loss": 1.0026,
+      "step": 1429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6267585220719076,
+      "learning_rate": 1.1108819414490151e-05,
+      "loss": 1.0189,
+      "step": 1430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6297826538131319,
+      "learning_rate": 1.1097992752207423e-05,
+      "loss": 1.1887,
+      "step": 1431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.48731497212887454,
+      "learning_rate": 1.1087164787031957e-05,
+      "loss": 1.1262,
+      "step": 1432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6917375546818298,
+      "learning_rate": 1.1076335531812357e-05,
+      "loss": 1.081,
+      "step": 1433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5599253984396574,
+      "learning_rate": 1.1065504999398762e-05,
+      "loss": 1.107,
+      "step": 1434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7158063833634571,
+      "learning_rate": 1.1054673202642828e-05,
+      "loss": 1.1507,
+      "step": 1435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5679949624799373,
+      "learning_rate": 1.1043840154397706e-05,
+      "loss": 1.1654,
+      "step": 1436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6489712225214253,
+      "learning_rate": 1.1033005867518034e-05,
+      "loss": 0.6687,
+      "step": 1437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6483283853521938,
+      "learning_rate": 1.1022170354859922e-05,
+      "loss": 1.2584,
+      "step": 1438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6876989686690028,
+      "learning_rate": 1.1011333629280932e-05,
+      "loss": 0.9923,
+      "step": 1439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6468595981278983,
+      "learning_rate": 1.1000495703640063e-05,
+      "loss": 1.1149,
+      "step": 1440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7035146443062912,
+      "learning_rate": 1.0989656590797748e-05,
+      "loss": 1.098,
+      "step": 1441
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6487733646099713,
+      "learning_rate": 1.0978816303615815e-05,
+      "loss": 0.9138,
+      "step": 1442
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6765727774825113,
+      "learning_rate": 1.0967974854957497e-05,
+      "loss": 1.2047,
+      "step": 1443
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4439102505367503,
+      "learning_rate": 1.0957132257687397e-05,
+      "loss": 1.1236,
+      "step": 1444
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6732364279186194,
+      "learning_rate": 1.0946288524671489e-05,
+      "loss": 1.1833,
+      "step": 1445
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4966976699333215,
+      "learning_rate": 1.0935443668777086e-05,
+      "loss": 0.8484,
+      "step": 1446
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6951781005255787,
+      "learning_rate": 1.0924597702872836e-05,
+      "loss": 1.1223,
+      "step": 1447
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5524074477780084,
+      "learning_rate": 1.091375063982871e-05,
+      "loss": 1.1234,
+      "step": 1448
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6202630721097784,
+      "learning_rate": 1.0902902492515978e-05,
+      "loss": 0.915,
+      "step": 1449
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6422049549354176,
+      "learning_rate": 1.0892053273807193e-05,
+      "loss": 1.064,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6757137852391039,
+      "learning_rate": 1.0881202996576176e-05,
+      "loss": 1.1164,
+      "step": 1451
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6840488129269842,
+      "learning_rate": 1.087035167369802e-05,
+      "loss": 1.2362,
+      "step": 1452
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5893132175508196,
+      "learning_rate": 1.0859499318049043e-05,
+      "loss": 1.0004,
+      "step": 1453
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5714078218262268,
+      "learning_rate": 1.084864594250679e-05,
+      "loss": 0.9955,
+      "step": 1454
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5389751074655719,
+      "learning_rate": 1.0837791559950029e-05,
+      "loss": 0.8587,
+      "step": 1455
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6477438579318805,
+      "learning_rate": 1.0826936183258705e-05,
+      "loss": 0.9636,
+      "step": 1456
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.36860615867413044,
+      "learning_rate": 1.081607982531396e-05,
+      "loss": 1.2276,
+      "step": 1457
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6321141658797487,
+      "learning_rate": 1.0805222498998085e-05,
+      "loss": 0.8775,
+      "step": 1458
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8100482559912244,
+      "learning_rate": 1.0794364217194533e-05,
+      "loss": 1.3301,
+      "step": 1459
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6334908953686155,
+      "learning_rate": 1.0783504992787882e-05,
+      "loss": 0.9893,
+      "step": 1460
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6588736767636358,
+      "learning_rate": 1.0772644838663828e-05,
+      "loss": 1.0316,
+      "step": 1461
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6628639468247045,
+      "learning_rate": 1.0761783767709182e-05,
+      "loss": 1.2141,
+      "step": 1462
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7855973321981303,
+      "learning_rate": 1.0750921792811829e-05,
+      "loss": 0.9376,
+      "step": 1463
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5462386978636936,
+      "learning_rate": 1.074005892686073e-05,
+      "loss": 0.9383,
+      "step": 1464
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6849851356081986,
+      "learning_rate": 1.0729195182745914e-05,
+      "loss": 0.907,
+      "step": 1465
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5857138160720231,
+      "learning_rate": 1.0718330573358436e-05,
+      "loss": 0.943,
+      "step": 1466
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6669296819374146,
+      "learning_rate": 1.0707465111590385e-05,
+      "loss": 1.1802,
+      "step": 1467
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5486556955242292,
+      "learning_rate": 1.0696598810334868e-05,
+      "loss": 1.1204,
+      "step": 1468
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.617527749880873,
+      "learning_rate": 1.068573168248598e-05,
+      "loss": 1.3687,
+      "step": 1469
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3887448142346477,
+      "learning_rate": 1.0674863740938797e-05,
+      "loss": 1.1112,
+      "step": 1470
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7262247492856034,
+      "learning_rate": 1.066399499858936e-05,
+      "loss": 1.1781,
+      "step": 1471
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5750278275711763,
+      "learning_rate": 1.0653125468334667e-05,
+      "loss": 0.6939,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6117763166220475,
+      "learning_rate": 1.0642255163072646e-05,
+      "loss": 1.0363,
+      "step": 1473
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6833865877393956,
+      "learning_rate": 1.063138409570214e-05,
+      "loss": 1.0615,
+      "step": 1474
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6001916933432839,
+      "learning_rate": 1.062051227912291e-05,
+      "loss": 0.9306,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6533338833542807,
+      "learning_rate": 1.0609639726235592e-05,
+      "loss": 1.3031,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6721535087511247,
+      "learning_rate": 1.05987664499417e-05,
+      "loss": 0.9985,
+      "step": 1477
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6167111415727138,
+      "learning_rate": 1.0587892463143611e-05,
+      "loss": 1.059,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.752583482818967,
+      "learning_rate": 1.0577017778744545e-05,
+      "loss": 1.1117,
+      "step": 1479
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5938151774779902,
+      "learning_rate": 1.0566142409648539e-05,
+      "loss": 1.259,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5454925293950738,
+      "learning_rate": 1.0555266368760451e-05,
+      "loss": 0.9014,
+      "step": 1481
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5509177597091167,
+      "learning_rate": 1.054438966898594e-05,
+      "loss": 0.9165,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7650049866617863,
+      "learning_rate": 1.0533512323231438e-05,
+      "loss": 1.3791,
+      "step": 1483
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.400311620040117,
+      "learning_rate": 1.0522634344404148e-05,
+      "loss": 1.1047,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6652821536380081,
+      "learning_rate": 1.0511755745412026e-05,
+      "loss": 0.8961,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7382979720848317,
+      "learning_rate": 1.050087653916376e-05,
+      "loss": 1.2133,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7210206677228688,
+      "learning_rate": 1.0489996738568757e-05,
+      "loss": 0.7421,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7447859004782797,
+      "learning_rate": 1.047911635653714e-05,
+      "loss": 1.2086,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5917977805972752,
+      "learning_rate": 1.046823540597971e-05,
+      "loss": 1.0612,
+      "step": 1489
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6497810217640843,
+      "learning_rate": 1.0457353899807947e-05,
+      "loss": 1.216,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.49386836211133484,
+      "learning_rate": 1.0446471850933988e-05,
+      "loss": 0.7352,
+      "step": 1491
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7183228829617457,
+      "learning_rate": 1.0435589272270622e-05,
+      "loss": 1.2031,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.704508480612499,
+      "learning_rate": 1.0424706176731254e-05,
+      "loss": 1.276,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6191131949570136,
+      "learning_rate": 1.0413822577229913e-05,
+      "loss": 0.9835,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4911544336070786,
+      "learning_rate": 1.0402938486681224e-05,
+      "loss": 1.1207,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5981500677866467,
+      "learning_rate": 1.0392053918000393e-05,
+      "loss": 1.0514,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6533104259676071,
+      "learning_rate": 1.0381168884103187e-05,
+      "loss": 1.1739,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6070547841626751,
+      "learning_rate": 1.0370283397905941e-05,
+      "loss": 0.983,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.495734418193103,
+      "learning_rate": 1.035939747232551e-05,
+      "loss": 0.6677,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7825446702840669,
+      "learning_rate": 1.034851112027928e-05,
+      "loss": 1.2238,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7263848734535964,
+      "learning_rate": 1.0337624354685136e-05,
+      "loss": 1.0264,
+      "step": 1501
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6778998172126479,
+      "learning_rate": 1.0326737188461465e-05,
+      "loss": 1.251,
+      "step": 1502
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.570654770426243,
+      "learning_rate": 1.031584963452712e-05,
+      "loss": 1.124,
+      "step": 1503
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5660925085716312,
+      "learning_rate": 1.0304961705801415e-05,
+      "loss": 1.104,
+      "step": 1504
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7204618416044699,
+      "learning_rate": 1.0294073415204114e-05,
+      "loss": 1.1524,
+      "step": 1505
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6501236804314708,
+      "learning_rate": 1.0283184775655401e-05,
+      "loss": 0.9315,
+      "step": 1506
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3686136918086963,
+      "learning_rate": 1.0272295800075885e-05,
+      "loss": 0.9825,
+      "step": 1507
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6208701280607554,
+      "learning_rate": 1.0261406501386568e-05,
+      "loss": 1.1322,
+      "step": 1508
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6407580100541301,
+      "learning_rate": 1.0250516892508835e-05,
+      "loss": 0.9173,
+      "step": 1509
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.553171658564712,
+      "learning_rate": 1.0239626986364441e-05,
+      "loss": 1.0601,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6628137350517861,
+      "learning_rate": 1.0228736795875489e-05,
+      "loss": 0.998,
+      "step": 1511
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6776384465929601,
+      "learning_rate": 1.021784633396443e-05,
+      "loss": 1.2109,
+      "step": 1512
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.579038033897784,
+      "learning_rate": 1.0206955613554024e-05,
+      "loss": 1.1261,
+      "step": 1513
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6412800853468694,
+      "learning_rate": 1.0196064647567346e-05,
+      "loss": 1.092,
+      "step": 1514
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.617106055802382,
+      "learning_rate": 1.0185173448927762e-05,
+      "loss": 1.1233,
+      "step": 1515
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7650757071326688,
+      "learning_rate": 1.017428203055891e-05,
+      "loss": 1.1161,
+      "step": 1516
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5013231571067144,
+      "learning_rate": 1.0163390405384694e-05,
+      "loss": 0.8277,
+      "step": 1517
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5991588332086645,
+      "learning_rate": 1.015249858632926e-05,
+      "loss": 0.9807,
+      "step": 1518
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4898199315906657,
+      "learning_rate": 1.0141606586316985e-05,
+      "loss": 1.1754,
+      "step": 1519
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6706549925491424,
+      "learning_rate": 1.013071441827246e-05,
+      "loss": 0.8949,
+      "step": 1520
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.63846443484541,
+      "learning_rate": 1.0119822095120471e-05,
+      "loss": 1.2409,
+      "step": 1521
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5926730786497912,
+      "learning_rate": 1.0108929629786001e-05,
+      "loss": 0.9696,
+      "step": 1522
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6831866353328723,
+      "learning_rate": 1.009803703519419e-05,
+      "loss": 0.9278,
+      "step": 1523
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.850113042314646,
+      "learning_rate": 1.0087144324270332e-05,
+      "loss": 1.2132,
+      "step": 1524
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5256159756612994,
+      "learning_rate": 1.0076251509939867e-05,
+      "loss": 0.7384,
+      "step": 1525
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.78543256385224,
+      "learning_rate": 1.0065358605128352e-05,
+      "loss": 1.2094,
+      "step": 1526
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5725646651991668,
+      "learning_rate": 1.0054465622761446e-05,
+      "loss": 1.1119,
+      "step": 1527
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6322419226661333,
+      "learning_rate": 1.0043572575764916e-05,
+      "loss": 1.1344,
+      "step": 1528
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7348819992328303,
+      "learning_rate": 1.0032679477064591e-05,
+      "loss": 1.1186,
+      "step": 1529
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7127670639639816,
+      "learning_rate": 1.0021786339586364e-05,
+      "loss": 0.9649,
+      "step": 1530
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7516202872069921,
+      "learning_rate": 1.0010893176256176e-05,
+      "loss": 1.2716,
+      "step": 1531
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3745608966281977,
+      "learning_rate": 1e-05,
+      "loss": 1.0746,
+      "step": 1532
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6064340521412436,
+      "learning_rate": 9.989106823743828e-06,
+      "loss": 1.1304,
+      "step": 1533
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5504023490969719,
+      "learning_rate": 9.97821366041364e-06,
+      "loss": 0.9747,
+      "step": 1534
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4803409147596974,
+      "learning_rate": 9.967320522935412e-06,
+      "loss": 0.6639,
+      "step": 1535
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.670955149408394,
+      "learning_rate": 9.956427424235087e-06,
+      "loss": 1.2388,
+      "step": 1536
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5997629088448405,
+      "learning_rate": 9.945534377238555e-06,
+      "loss": 0.9728,
+      "step": 1537
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0982530565674133,
+      "learning_rate": 9.934641394871651e-06,
+      "loss": 1.2521,
+      "step": 1538
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5718523892253395,
+      "learning_rate": 9.923748490060135e-06,
+      "loss": 1.1009,
+      "step": 1539
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6317277333877067,
+      "learning_rate": 9.91285567572967e-06,
+      "loss": 1.1298,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6458220401667738,
+      "learning_rate": 9.901962964805813e-06,
+      "loss": 1.0611,
+      "step": 1541
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5849153002888605,
+      "learning_rate": 9.891070370214e-06,
+      "loss": 1.0465,
+      "step": 1542
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6686555311493239,
+      "learning_rate": 9.88017790487953e-06,
+      "loss": 1.1478,
+      "step": 1543
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.37849085901027574,
+      "learning_rate": 9.869285581727543e-06,
+      "loss": 0.9018,
+      "step": 1544
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6638751317263084,
+      "learning_rate": 9.858393413683017e-06,
+      "loss": 1.3366,
+      "step": 1545
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7251100236875296,
+      "learning_rate": 9.847501413670742e-06,
+      "loss": 0.9208,
+      "step": 1546
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.664018551414427,
+      "learning_rate": 9.836609594615309e-06,
+      "loss": 1.3303,
+      "step": 1547
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5791613383393673,
+      "learning_rate": 9.825717969441092e-06,
+      "loss": 0.9627,
+      "step": 1548
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5805977672394328,
+      "learning_rate": 9.81482655107224e-06,
+      "loss": 0.9165,
+      "step": 1549
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7979532400713067,
+      "learning_rate": 9.803935352432657e-06,
+      "loss": 1.3406,
+      "step": 1550
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6564744362521798,
+      "learning_rate": 9.79304438644598e-06,
+      "loss": 1.1469,
+      "step": 1551
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.49154768127987275,
+      "learning_rate": 9.782153666035575e-06,
+      "loss": 0.8715,
+      "step": 1552
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6381981448184686,
+      "learning_rate": 9.771263204124513e-06,
+      "loss": 1.0441,
+      "step": 1553
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7356994877779364,
+      "learning_rate": 9.760373013635562e-06,
+      "loss": 1.0056,
+      "step": 1554
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6865781827169765,
+      "learning_rate": 9.749483107491167e-06,
+      "loss": 1.1826,
+      "step": 1555
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4486820112619568,
+      "learning_rate": 9.738593498613435e-06,
+      "loss": 1.0956,
+      "step": 1556
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5694789327698413,
+      "learning_rate": 9.727704199924117e-06,
+      "loss": 1.2512,
+      "step": 1557
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6902294694257273,
+      "learning_rate": 9.7168152243446e-06,
+      "loss": 1.0583,
+      "step": 1558
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5998732479926534,
+      "learning_rate": 9.705926584795891e-06,
+      "loss": 1.0113,
+      "step": 1559
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6748790851621241,
+      "learning_rate": 9.695038294198588e-06,
+      "loss": 1.1319,
+      "step": 1560
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.47435159022145673,
+      "learning_rate": 9.684150365472883e-06,
+      "loss": 0.8204,
+      "step": 1561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6367446646134197,
+      "learning_rate": 9.673262811538537e-06,
+      "loss": 1.1117,
+      "step": 1562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5804235869796706,
+      "learning_rate": 9.662375645314867e-06,
+      "loss": 1.0282,
+      "step": 1563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6061202544773693,
+      "learning_rate": 9.651488879720724e-06,
+      "loss": 1.1213,
+      "step": 1564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6484052441549107,
+      "learning_rate": 9.640602527674495e-06,
+      "loss": 0.9659,
+      "step": 1565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6026109607969359,
+      "learning_rate": 9.629716602094062e-06,
+      "loss": 1.0815,
+      "step": 1566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7606941820843298,
+      "learning_rate": 9.618831115896814e-06,
+      "loss": 1.2237,
+      "step": 1567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4347594185763318,
+      "learning_rate": 9.607946081999609e-06,
+      "loss": 1.1753,
+      "step": 1568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8089007738864563,
+      "learning_rate": 9.597061513318778e-06,
+      "loss": 1.1331,
+      "step": 1569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5608761702740999,
+      "learning_rate": 9.586177422770089e-06,
+      "loss": 0.7189,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6905933048939615,
+      "learning_rate": 9.575293823268749e-06,
+      "loss": 1.0027,
+      "step": 1571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5750539382327454,
+      "learning_rate": 9.564410727729383e-06,
+      "loss": 1.13,
+      "step": 1572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5665095693041945,
+      "learning_rate": 9.553528149066016e-06,
+      "loss": 1.0736,
+      "step": 1573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.658889552514107,
+      "learning_rate": 9.542646100192056e-06,
+      "loss": 1.212,
+      "step": 1574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6263517126604736,
+      "learning_rate": 9.531764594020293e-06,
+      "loss": 0.919,
+      "step": 1575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6149598172121,
+      "learning_rate": 9.520883643462862e-06,
+      "loss": 1.1313,
+      "step": 1576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5201003382670517,
+      "learning_rate": 9.510003261431244e-06,
+      "loss": 1.0339,
+      "step": 1577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5344554914393138,
+      "learning_rate": 9.499123460836242e-06,
+      "loss": 0.6715,
+      "step": 1578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6921966469253894,
+      "learning_rate": 9.488244254587977e-06,
+      "loss": 1.2932,
+      "step": 1579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.586239757156232,
+      "learning_rate": 9.477365655595855e-06,
+      "loss": 0.9604,
+      "step": 1580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7645452502717458,
+      "learning_rate": 9.466487676768563e-06,
+      "loss": 1.1818,
+      "step": 1581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3957126906860972,
+      "learning_rate": 9.455610331014064e-06,
+      "loss": 1.0725,
+      "step": 1582
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7768410288890621,
+      "learning_rate": 9.444733631239552e-06,
+      "loss": 1.1873,
+      "step": 1583
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5210059820209807,
+      "learning_rate": 9.433857590351465e-06,
+      "loss": 1.1692,
+      "step": 1584
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5641755621757018,
+      "learning_rate": 9.422982221255458e-06,
+      "loss": 0.9518,
+      "step": 1585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7188134828118146,
+      "learning_rate": 9.41210753685639e-06,
+      "loss": 1.2917,
+      "step": 1586
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6281415438920909,
+      "learning_rate": 9.401233550058303e-06,
+      "loss": 0.9296,
+      "step": 1587
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5030972613882833,
+      "learning_rate": 9.390360273764411e-06,
+      "loss": 0.7339,
+      "step": 1588
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7554949852640267,
+      "learning_rate": 9.379487720877093e-06,
+      "loss": 1.0917,
+      "step": 1589
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5951884875112878,
+      "learning_rate": 9.368615904297861e-06,
+      "loss": 1.1753,
+      "step": 1590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7798674866296921,
+      "learning_rate": 9.357744836927359e-06,
+      "loss": 1.2562,
+      "step": 1591
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6739059063020579,
+      "learning_rate": 9.346874531665336e-06,
+      "loss": 1.064,
+      "step": 1592
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4774278677914438,
+      "learning_rate": 9.336005001410644e-06,
+      "loss": 1.1623,
+      "step": 1593
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6963196946604675,
+      "learning_rate": 9.325136259061207e-06,
+      "loss": 0.8849,
+      "step": 1594
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6725277242598445,
+      "learning_rate": 9.314268317514023e-06,
+      "loss": 1.1809,
+      "step": 1595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5097766303055257,
+      "learning_rate": 9.303401189665135e-06,
+      "loss": 1.0652,
+      "step": 1596
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4941432367260272,
+      "learning_rate": 9.292534888409618e-06,
+      "loss": 0.6932,
+      "step": 1597
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8339039919324277,
+      "learning_rate": 9.28166942664157e-06,
+      "loss": 1.2746,
+      "step": 1598
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7297171716099446,
+      "learning_rate": 9.270804817254091e-06,
+      "loss": 0.9677,
+      "step": 1599
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.600020595722802,
+      "learning_rate": 9.259941073139272e-06,
+      "loss": 1.2322,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6155877721707952,
+      "learning_rate": 9.249078207188175e-06,
+      "loss": 1.0126,
+      "step": 1601
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6205831431276233,
+      "learning_rate": 9.238216232290821e-06,
+      "loss": 1.1362,
+      "step": 1602
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6392441850579318,
+      "learning_rate": 9.227355161336174e-06,
+      "loss": 0.9987,
+      "step": 1603
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6777564617975294,
+      "learning_rate": 9.216495007212122e-06,
+      "loss": 0.9544,
+      "step": 1604
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5067413407869783,
+      "learning_rate": 9.20563578280547e-06,
+      "loss": 0.9717,
+      "step": 1605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.42757077142096456,
+      "learning_rate": 9.194777501001919e-06,
+      "loss": 1.0591,
+      "step": 1606
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6805756199925198,
+      "learning_rate": 9.183920174686046e-06,
+      "loss": 1.188,
+      "step": 1607
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6920306555867014,
+      "learning_rate": 9.173063816741296e-06,
+      "loss": 1.0308,
+      "step": 1608
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5039650145135657,
+      "learning_rate": 9.162208440049976e-06,
+      "loss": 1.0827,
+      "step": 1609
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6209798884499514,
+      "learning_rate": 9.151354057493213e-06,
+      "loss": 1.0908,
+      "step": 1610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6016326044203245,
+      "learning_rate": 9.14050068195096e-06,
+      "loss": 1.0819,
+      "step": 1611
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6190831687570866,
+      "learning_rate": 9.129648326301983e-06,
+      "loss": 1.3734,
+      "step": 1612
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6390059951133684,
+      "learning_rate": 9.118797003423826e-06,
+      "loss": 0.9877,
+      "step": 1613
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4631017604384936,
+      "learning_rate": 9.10794672619281e-06,
+      "loss": 0.7371,
+      "step": 1614
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5630574184974934,
+      "learning_rate": 9.097097507484023e-06,
+      "loss": 1.0231,
+      "step": 1615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6172470665784344,
+      "learning_rate": 9.086249360171291e-06,
+      "loss": 1.0546,
+      "step": 1616
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7343452413307036,
+      "learning_rate": 9.075402297127167e-06,
+      "loss": 1.1482,
+      "step": 1617
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3732726767684167,
+      "learning_rate": 9.064556331222918e-06,
+      "loss": 1.1158,
+      "step": 1618
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6934665966667566,
+      "learning_rate": 9.053711475328515e-06,
+      "loss": 1.1179,
+      "step": 1619
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5508347913433024,
+      "learning_rate": 9.042867742312605e-06,
+      "loss": 1.0887,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.621268529823606,
+      "learning_rate": 9.032025145042504e-06,
+      "loss": 1.1608,
+      "step": 1621
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.565933393987109,
+      "learning_rate": 9.021183696384186e-06,
+      "loss": 1.1599,
+      "step": 1622
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4844722738649361,
+      "learning_rate": 9.010343409202256e-06,
+      "loss": 0.6979,
+      "step": 1623
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6223099898129011,
+      "learning_rate": 8.99950429635994e-06,
+      "loss": 1.2111,
+      "step": 1624
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.651475042707302,
+      "learning_rate": 8.988666370719071e-06,
+      "loss": 0.9909,
+      "step": 1625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7105287061193927,
+      "learning_rate": 8.977829645140081e-06,
+      "loss": 1.1366,
+      "step": 1626
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6906091404729082,
+      "learning_rate": 8.96699413248197e-06,
+      "loss": 1.0353,
+      "step": 1627
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6892956067516528,
+      "learning_rate": 8.956159845602297e-06,
+      "loss": 0.9539,
+      "step": 1628
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6901089379893459,
+      "learning_rate": 8.945326797357174e-06,
+      "loss": 1.2235,
+      "step": 1629
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3887198634058874,
+      "learning_rate": 8.934495000601241e-06,
+      "loss": 1.1099,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5898010888877552,
+      "learning_rate": 8.923664468187647e-06,
+      "loss": 1.2355,
+      "step": 1631
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5032435510514436,
+      "learning_rate": 8.912835212968047e-06,
+      "loss": 0.7329,
+      "step": 1632
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6549743068597293,
+      "learning_rate": 8.902007247792582e-06,
+      "loss": 1.0957,
+      "step": 1633
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6665629117576923,
+      "learning_rate": 8.891180585509854e-06,
+      "loss": 1.125,
+      "step": 1634
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6385888376081512,
+      "learning_rate": 8.880355238966923e-06,
+      "loss": 1.028,
+      "step": 1635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5995615427115175,
+      "learning_rate": 8.869531221009294e-06,
+      "loss": 0.9723,
+      "step": 1636
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5630080114476215,
+      "learning_rate": 8.858708544480887e-06,
+      "loss": 1.043,
+      "step": 1637
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6169567769901627,
+      "learning_rate": 8.847887222224032e-06,
+      "loss": 1.18,
+      "step": 1638
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.621036839457768,
+      "learning_rate": 8.837067267079456e-06,
+      "loss": 1.0623,
+      "step": 1639
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6310832851776162,
+      "learning_rate": 8.82624869188626e-06,
+      "loss": 1.1557,
+      "step": 1640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5083113425793323,
+      "learning_rate": 8.815431509481909e-06,
+      "loss": 0.6705,
+      "step": 1641
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6243826625979744,
+      "learning_rate": 8.804615732702213e-06,
+      "loss": 1.0869,
+      "step": 1642
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4876484425142108,
+      "learning_rate": 8.793801374381319e-06,
+      "loss": 1.1979,
+      "step": 1643
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5908803539436984,
+      "learning_rate": 8.782988447351685e-06,
+      "loss": 1.1015,
+      "step": 1644
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7589153630745795,
+      "learning_rate": 8.772176964444073e-06,
+      "loss": 1.1464,
+      "step": 1645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6338716354820473,
+      "learning_rate": 8.761366938487536e-06,
+      "loss": 1.0298,
+      "step": 1646
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6981393063478256,
+      "learning_rate": 8.75055838230939e-06,
+      "loss": 1.0658,
+      "step": 1647
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6470728828387629,
+      "learning_rate": 8.73975130873521e-06,
+      "loss": 1.1702,
+      "step": 1648
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7042679138970357,
+      "learning_rate": 8.728945730588818e-06,
+      "loss": 0.9164,
+      "step": 1649
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5200239563316101,
+      "learning_rate": 8.718141660692252e-06,
+      "loss": 0.8627,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7456646605507268,
+      "learning_rate": 8.707339111865763e-06,
+      "loss": 0.9215,
+      "step": 1651
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5489542568882777,
+      "learning_rate": 8.6965380969278e-06,
+      "loss": 0.9699,
+      "step": 1652
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6406967632167619,
+      "learning_rate": 8.685738628694995e-06,
+      "loss": 1.1583,
+      "step": 1653
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.562506868144836,
+      "learning_rate": 8.674940719982135e-06,
+      "loss": 1.0677,
+      "step": 1654
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6207614917155918,
+      "learning_rate": 8.66414438360216e-06,
+      "loss": 1.3554,
+      "step": 1655
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.40320478975387314,
+      "learning_rate": 8.65334963236615e-06,
+      "loss": 1.0921,
+      "step": 1656
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7384760874473211,
+      "learning_rate": 8.642556479083297e-06,
+      "loss": 1.2036,
+      "step": 1657
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5126181192439799,
+      "learning_rate": 8.6317649365609e-06,
+      "loss": 0.6727,
+      "step": 1658
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5874804114771308,
+      "learning_rate": 8.620975017604347e-06,
+      "loss": 1.0663,
+      "step": 1659
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5832789711561721,
+      "learning_rate": 8.610186735017102e-06,
+      "loss": 1.0114,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6174094986531017,
+      "learning_rate": 8.599400101600675e-06,
+      "loss": 0.9928,
+      "step": 1661
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7042851126360115,
+      "learning_rate": 8.588615130154636e-06,
+      "loss": 1.3288,
+      "step": 1662
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6612900723562849,
+      "learning_rate": 8.577831833476574e-06,
+      "loss": 0.9991,
+      "step": 1663
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.609988764617547,
+      "learning_rate": 8.567050224362088e-06,
+      "loss": 1.1791,
+      "step": 1664
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6358504977852428,
+      "learning_rate": 8.556270315604779e-06,
+      "loss": 0.9328,
+      "step": 1665
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5457722166739702,
+      "learning_rate": 8.545492119996233e-06,
+      "loss": 1.272,
+      "step": 1666
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5441606512852258,
+      "learning_rate": 8.534715650325997e-06,
+      "loss": 0.8496,
+      "step": 1667
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6350157703609416,
+      "learning_rate": 8.523940919381571e-06,
+      "loss": 0.9606,
+      "step": 1668
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6537673422362633,
+      "learning_rate": 8.513167939948397e-06,
+      "loss": 1.2407,
+      "step": 1669
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8175399332223765,
+      "learning_rate": 8.502396724809834e-06,
+      "loss": 1.1118,
+      "step": 1670
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5927183463909232,
+      "learning_rate": 8.491627286747145e-06,
+      "loss": 1.1139,
+      "step": 1671
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6897165404596587,
+      "learning_rate": 8.480859638539492e-06,
+      "loss": 0.9619,
+      "step": 1672
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6171240959745348,
+      "learning_rate": 8.47009379296391e-06,
+      "loss": 0.8203,
+      "step": 1673
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6753166854723239,
+      "learning_rate": 8.45932976279529e-06,
+      "loss": 1.0958,
+      "step": 1674
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5801901458828851,
+      "learning_rate": 8.448567560806371e-06,
+      "loss": 1.0489,
+      "step": 1675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6548134158014519,
+      "learning_rate": 8.437807199767731e-06,
+      "loss": 1.3044,
+      "step": 1676
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4521411530775893,
+      "learning_rate": 8.427048692447754e-06,
+      "loss": 0.7106,
+      "step": 1677
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7040209803280151,
+      "learning_rate": 8.416292051612625e-06,
+      "loss": 1.1888,
+      "step": 1678
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.560551909765743,
+      "learning_rate": 8.40553729002632e-06,
+      "loss": 1.2459,
+      "step": 1679
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6085811356477794,
+      "learning_rate": 8.394784420450578e-06,
+      "loss": 1.0136,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4164277856072426,
+      "learning_rate": 8.384033455644897e-06,
+      "loss": 1.1158,
+      "step": 1681
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5887380709010454,
+      "learning_rate": 8.373284408366518e-06,
+      "loss": 1.0851,
+      "step": 1682
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.701157395301471,
+      "learning_rate": 8.3625372913704e-06,
+      "loss": 1.2496,
+      "step": 1683
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5938237132587976,
+      "learning_rate": 8.351792117409214e-06,
+      "loss": 1.0163,
+      "step": 1684
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.486164047402809,
+      "learning_rate": 8.341048899233324e-06,
+      "loss": 0.6901,
+      "step": 1685
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.785039612004658,
+      "learning_rate": 8.330307649590782e-06,
+      "loss": 1.2043,
+      "step": 1686
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.624683394094322,
+      "learning_rate": 8.319568381227288e-06,
+      "loss": 0.9663,
+      "step": 1687
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6401434100977058,
+      "learning_rate": 8.308831106886206e-06,
+      "loss": 1.405,
+      "step": 1688
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.63128616720811,
+      "learning_rate": 8.298095839308526e-06,
+      "loss": 1.0294,
+      "step": 1689
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5109126249209484,
+      "learning_rate": 8.287362591232859e-06,
+      "loss": 1.0911,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7013104987666103,
+      "learning_rate": 8.276631375395417e-06,
+      "loss": 1.1752,
+      "step": 1691
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.399532178946096,
+      "learning_rate": 8.265902204530007e-06,
+      "loss": 1.0999,
+      "step": 1692
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5052196145321185,
+      "learning_rate": 8.255175091368004e-06,
+      "loss": 0.8154,
+      "step": 1693
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6568338109501972,
+      "learning_rate": 8.244450048638341e-06,
+      "loss": 1.0757,
+      "step": 1694
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6567904566815637,
+      "learning_rate": 8.233727089067494e-06,
+      "loss": 1.1303,
+      "step": 1695
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5135337424419454,
+      "learning_rate": 8.223006225379474e-06,
+      "loss": 0.8795,
+      "step": 1696
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6280991954129682,
+      "learning_rate": 8.212287470295796e-06,
+      "loss": 0.9885,
+      "step": 1697
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6662575807396421,
+      "learning_rate": 8.201570836535477e-06,
+      "loss": 1.1892,
+      "step": 1698
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6396899548758538,
+      "learning_rate": 8.190856336815019e-06,
+      "loss": 1.2024,
+      "step": 1699
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7400287697193875,
+      "learning_rate": 8.180143983848388e-06,
+      "loss": 1.0856,
+      "step": 1700
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5899737279640858,
+      "learning_rate": 8.169433790347002e-06,
+      "loss": 1.126,
+      "step": 1701
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5772955247089394,
+      "learning_rate": 8.158725769019719e-06,
+      "loss": 1.1562,
+      "step": 1702
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4825034100564359,
+      "learning_rate": 8.14801993257282e-06,
+      "loss": 0.6914,
+      "step": 1703
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6104445777729363,
+      "learning_rate": 8.137316293709994e-06,
+      "loss": 0.9426,
+      "step": 1704
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4107907850916833,
+      "learning_rate": 8.126614865132314e-06,
+      "loss": 1.1816,
+      "step": 1705
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.675313178487816,
+      "learning_rate": 8.115915659538246e-06,
+      "loss": 0.8656,
+      "step": 1706
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6407586301814115,
+      "learning_rate": 8.105218689623605e-06,
+      "loss": 1.2643,
+      "step": 1707
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6026918518819275,
+      "learning_rate": 8.094523968081553e-06,
+      "loss": 0.962,
+      "step": 1708
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7116864923211448,
+      "learning_rate": 8.083831507602593e-06,
+      "loss": 0.9343,
+      "step": 1709
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9027270246514997,
+      "learning_rate": 8.07314132087454e-06,
+      "loss": 1.2531,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.49626607217975993,
+      "learning_rate": 8.06245342058251e-06,
+      "loss": 0.722,
+      "step": 1711
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5995732580809142,
+      "learning_rate": 8.051767819408907e-06,
+      "loss": 1.2049,
+      "step": 1712
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5198639344460568,
+      "learning_rate": 8.041084530033406e-06,
+      "loss": 1.1242,
+      "step": 1713
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6549681814981623,
+      "learning_rate": 8.030403565132942e-06,
+      "loss": 1.2326,
+      "step": 1714
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5936433042566616,
+      "learning_rate": 8.019724937381687e-06,
+      "loss": 0.958,
+      "step": 1715
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6275394953295488,
+      "learning_rate": 8.009048659451044e-06,
+      "loss": 1.0153,
+      "step": 1716
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6625033579272616,
+      "learning_rate": 7.998374744009624e-06,
+      "loss": 1.2785,
+      "step": 1717
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4295393832760123,
+      "learning_rate": 7.987703203723236e-06,
+      "loss": 1.0322,
+      "step": 1718
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5760587122516111,
+      "learning_rate": 7.977034051254875e-06,
+      "loss": 1.0977,
+      "step": 1719
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5339107992868738,
+      "learning_rate": 7.966367299264696e-06,
+      "loss": 0.9617,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4565091652073978,
+      "learning_rate": 7.955702960410006e-06,
+      "loss": 0.7832,
+      "step": 1721
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.624818327905017,
+      "learning_rate": 7.945041047345254e-06,
+      "loss": 1.2492,
+      "step": 1722
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6154556681383374,
+      "learning_rate": 7.934381572722006e-06,
+      "loss": 0.9148,
+      "step": 1723
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7613283334642198,
+      "learning_rate": 7.923724549188935e-06,
+      "loss": 1.2158,
+      "step": 1724
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5514093628998498,
+      "learning_rate": 7.913069989391805e-06,
+      "loss": 1.1026,
+      "step": 1725
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6248813684699196,
+      "learning_rate": 7.902417905973463e-06,
+      "loss": 1.181,
+      "step": 1726
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6925841263504048,
+      "learning_rate": 7.89176831157381e-06,
+      "loss": 1.1205,
+      "step": 1727
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6024152812171941,
+      "learning_rate": 7.881121218829789e-06,
+      "loss": 1.0799,
+      "step": 1728
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6573792131577955,
+      "learning_rate": 7.87047664037539e-06,
+      "loss": 1.1476,
+      "step": 1729
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3423120745039066,
+      "learning_rate": 7.859834588841608e-06,
+      "loss": 0.9063,
+      "step": 1730
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6617894393736447,
+      "learning_rate": 7.849195076856443e-06,
+      "loss": 1.3174,
+      "step": 1731
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6834436514222975,
+      "learning_rate": 7.838558117044876e-06,
+      "loss": 0.9833,
+      "step": 1732
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5920713491201179,
+      "learning_rate": 7.827923722028871e-06,
+      "loss": 1.2798,
+      "step": 1733
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5606859125660747,
+      "learning_rate": 7.81729190442734e-06,
+      "loss": 0.8888,
+      "step": 1734
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6193807123731256,
+      "learning_rate": 7.806662676856134e-06,
+      "loss": 0.9773,
+      "step": 1735
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6892060159929576,
+      "learning_rate": 7.79603605192804e-06,
+      "loss": 1.2337,
+      "step": 1736
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6021291223978197,
+      "learning_rate": 7.785412042252752e-06,
+      "loss": 1.1025,
+      "step": 1737
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.510658449965839,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.827,
+      "step": 1738
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7664083239556623,
+      "learning_rate": 7.764171919083831e-06,
+      "loss": 1.0673,
+      "step": 1739
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5820692378866911,
+      "learning_rate": 7.753555830794013e-06,
+      "loss": 0.9369,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6697569572453655,
+      "learning_rate": 7.742942408164591e-06,
+      "loss": 1.2008,
+      "step": 1741
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.38561082587922957,
+      "learning_rate": 7.732331663789592e-06,
+      "loss": 1.1069,
+      "step": 1742
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5606704945563207,
+      "learning_rate": 7.72172361025987e-06,
+      "loss": 1.2404,
+      "step": 1743
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6178805601386427,
+      "learning_rate": 7.71111826016308e-06,
+      "loss": 1.0004,
+      "step": 1744
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.584155618632914,
+      "learning_rate": 7.700515626083669e-06,
+      "loss": 1.0096,
+      "step": 1745
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6234125837289846,
+      "learning_rate": 7.689915720602868e-06,
+      "loss": 0.9415,
+      "step": 1746
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.44644279563409095,
+      "learning_rate": 7.67931855629866e-06,
+      "loss": 0.8198,
+      "step": 1747
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7455057456173436,
+      "learning_rate": 7.66872414574578e-06,
+      "loss": 1.1128,
+      "step": 1748
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6103105133286771,
+      "learning_rate": 7.658132501515702e-06,
+      "loss": 0.9903,
+      "step": 1749
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6606199426120424,
+      "learning_rate": 7.647543636176606e-06,
+      "loss": 1.187,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6106548290889563,
+      "learning_rate": 7.636957562293383e-06,
+      "loss": 0.9588,
+      "step": 1751
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5676042042125817,
+      "learning_rate": 7.626374292427604e-06,
+      "loss": 1.054,
+      "step": 1752
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7447869240519877,
+      "learning_rate": 7.615793839137522e-06,
+      "loss": 1.1758,
+      "step": 1753
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.37508433112072775,
+      "learning_rate": 7.60521621497804e-06,
+      "loss": 1.1724,
+      "step": 1754
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7331781099897409,
+      "learning_rate": 7.594641432500707e-06,
+      "loss": 1.0458,
+      "step": 1755
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5155828174420033,
+      "learning_rate": 7.584069504253703e-06,
+      "loss": 0.7293,
+      "step": 1756
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7146553563362854,
+      "learning_rate": 7.573500442781815e-06,
+      "loss": 1.2881,
+      "step": 1757
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5394691201590462,
+      "learning_rate": 7.56293426062643e-06,
+      "loss": 0.9959,
+      "step": 1758
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5116163691145993,
+      "learning_rate": 7.552370970325524e-06,
+      "loss": 1.0563,
+      "step": 1759
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5848714964903177,
+      "learning_rate": 7.541810584413635e-06,
+      "loss": 1.2147,
+      "step": 1760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5946138066583494,
+      "learning_rate": 7.531253115421856e-06,
+      "loss": 0.8947,
+      "step": 1761
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7403244681130777,
+      "learning_rate": 7.520698575877819e-06,
+      "loss": 1.0968,
+      "step": 1762
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5620102123722084,
+      "learning_rate": 7.510146978305683e-06,
+      "loss": 1.0943,
+      "step": 1763
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.49452934629039713,
+      "learning_rate": 7.499598335226113e-06,
+      "loss": 0.841,
+      "step": 1764
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6229944259585142,
+      "learning_rate": 7.489052659156264e-06,
+      "loss": 1.1905,
+      "step": 1765
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6521515841907193,
+      "learning_rate": 7.47850996260978e-06,
+      "loss": 0.9237,
+      "step": 1766
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6582218185621798,
+      "learning_rate": 7.467970258096764e-06,
+      "loss": 1.1784,
+      "step": 1767
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.36383394725776924,
+      "learning_rate": 7.457433558123764e-06,
+      "loss": 1.0441,
+      "step": 1768
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7505908972168122,
+      "learning_rate": 7.446899875193775e-06,
+      "loss": 1.2031,
+      "step": 1769
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5712943619486661,
+      "learning_rate": 7.436369221806201e-06,
+      "loss": 1.079,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6959826065152196,
+      "learning_rate": 7.425841610456854e-06,
+      "loss": 0.8983,
+      "step": 1771
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6325561925546603,
+      "learning_rate": 7.415317053637936e-06,
+      "loss": 1.2341,
+      "step": 1772
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6389652999603104,
+      "learning_rate": 7.404795563838027e-06,
+      "loss": 0.8974,
+      "step": 1773
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4685762443910463,
+      "learning_rate": 7.394277153542069e-06,
+      "loss": 0.75,
+      "step": 1774
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6102465761428463,
+      "learning_rate": 7.3837618352313426e-06,
+      "loss": 1.1888,
+      "step": 1775
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6607164065877673,
+      "learning_rate": 7.373249621383466e-06,
+      "loss": 1.2281,
+      "step": 1776
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6820314608322438,
+      "learning_rate": 7.3627405244723715e-06,
+      "loss": 1.0743,
+      "step": 1777
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6083270238090359,
+      "learning_rate": 7.352234556968289e-06,
+      "loss": 1.0148,
+      "step": 1778
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4123110547330525,
+      "learning_rate": 7.341731731337745e-06,
+      "loss": 1.1331,
+      "step": 1779
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7065953281495146,
+      "learning_rate": 7.331232060043527e-06,
+      "loss": 1.0183,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6741664914073735,
+      "learning_rate": 7.320735555544685e-06,
+      "loss": 1.2354,
+      "step": 1781
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7460631012014125,
+      "learning_rate": 7.310242230296508e-06,
+      "loss": 1.0266,
+      "step": 1782
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49835808845851554,
+      "learning_rate": 7.29975209675052e-06,
+      "loss": 0.8396,
+      "step": 1783
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6781804474281321,
+      "learning_rate": 7.289265167354449e-06,
+      "loss": 1.1249,
+      "step": 1784
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6427510922311185,
+      "learning_rate": 7.278781454552222e-06,
+      "loss": 0.9692,
+      "step": 1785
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5688975349026931,
+      "learning_rate": 7.2683009707839555e-06,
+      "loss": 1.2797,
+      "step": 1786
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6095494904542564,
+      "learning_rate": 7.257823728485931e-06,
+      "loss": 0.9611,
+      "step": 1787
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6790663188059377,
+      "learning_rate": 7.247349740090579e-06,
+      "loss": 1.1201,
+      "step": 1788
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6165301824111209,
+      "learning_rate": 7.236879018026472e-06,
+      "loss": 0.9555,
+      "step": 1789
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7030340483795291,
+      "learning_rate": 7.226411574718314e-06,
+      "loss": 1.1119,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5125415478421086,
+      "learning_rate": 7.215947422586907e-06,
+      "loss": 0.8519,
+      "step": 1791
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4199735601822432,
+      "learning_rate": 7.205486574049152e-06,
+      "loss": 1.0425,
+      "step": 1792
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6994511393039545,
+      "learning_rate": 7.195029041518032e-06,
+      "loss": 1.1085,
+      "step": 1793
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7307488588149369,
+      "learning_rate": 7.1845748374025955e-06,
+      "loss": 1.1446,
+      "step": 1794
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.530173726543926,
+      "learning_rate": 7.174123974107934e-06,
+      "loss": 1.1413,
+      "step": 1795
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6812230077324409,
+      "learning_rate": 7.163676464035186e-06,
+      "loss": 0.9604,
+      "step": 1796
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.553994365605793,
+      "learning_rate": 7.153232319581502e-06,
+      "loss": 1.1082,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6806693146836076,
+      "learning_rate": 7.142791553140045e-06,
+      "loss": 1.3383,
+      "step": 1798
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6366856846270691,
+      "learning_rate": 7.1323541770999615e-06,
+      "loss": 0.9356,
+      "step": 1799
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.44771730631818246,
+      "learning_rate": 7.121920203846387e-06,
+      "loss": 0.7506,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.600561335300782,
+      "learning_rate": 7.111489645760405e-06,
+      "loss": 1.0482,
+      "step": 1801
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5799265560089045,
+      "learning_rate": 7.101062515219059e-06,
+      "loss": 1.2092,
+      "step": 1802
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.41866350602582736,
+      "learning_rate": 7.09063882459532e-06,
+      "loss": 1.1006,
+      "step": 1803
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5978197040295232,
+      "learning_rate": 7.0802185862580756e-06,
+      "loss": 1.025,
+      "step": 1804
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7252632094128787,
+      "learning_rate": 7.069801812572117e-06,
+      "loss": 1.1114,
+      "step": 1805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6229730671079425,
+      "learning_rate": 7.059388515898129e-06,
+      "loss": 1.0376,
+      "step": 1806
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6165895293778959,
+      "learning_rate": 7.0489787085926664e-06,
+      "loss": 1.1858,
+      "step": 1807
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6092282553500928,
+      "learning_rate": 7.038572403008146e-06,
+      "loss": 1.1167,
+      "step": 1808
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4305814027652885,
+      "learning_rate": 7.0281696114928214e-06,
+      "loss": 0.7999,
+      "step": 1809
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7137204476331173,
+      "learning_rate": 7.017770346390791e-06,
+      "loss": 1.2054,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6831441505897714,
+      "learning_rate": 7.007374620041956e-06,
+      "loss": 0.9856,
+      "step": 1811
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5533003915883605,
+      "learning_rate": 6.996982444782021e-06,
+      "loss": 1.0685,
+      "step": 1812
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6757381651843989,
+      "learning_rate": 6.986593832942486e-06,
+      "loss": 1.0157,
+      "step": 1813
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6426685316617226,
+      "learning_rate": 6.976208796850608e-06,
+      "loss": 1.0125,
+      "step": 1814
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8083718181421006,
+      "learning_rate": 6.96582734882941e-06,
+      "loss": 1.173,
+      "step": 1815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3765245541304035,
+      "learning_rate": 6.95544950119766e-06,
+      "loss": 1.1168,
+      "step": 1816
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.523930915445097,
+      "learning_rate": 6.945075266269845e-06,
+      "loss": 1.21,
+      "step": 1817
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4351766577061068,
+      "learning_rate": 6.934704656356171e-06,
+      "loss": 0.7076,
+      "step": 1818
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5572329320351831,
+      "learning_rate": 6.92433768376254e-06,
+      "loss": 1.0795,
+      "step": 1819
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.691142593396509,
+      "learning_rate": 6.9139743607905405e-06,
+      "loss": 1.1411,
+      "step": 1820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7245924202298865,
+      "learning_rate": 6.903614699737429e-06,
+      "loss": 1.2277,
+      "step": 1821
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5232419446214315,
+      "learning_rate": 6.8932587128961124e-06,
+      "loss": 0.8261,
+      "step": 1822
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5736150202157103,
+      "learning_rate": 6.8829064125551485e-06,
+      "loss": 1.1206,
+      "step": 1823
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9855940277979123,
+      "learning_rate": 6.872557810998709e-06,
+      "loss": 1.1469,
+      "step": 1824
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6165550446702174,
+      "learning_rate": 6.862212920506581e-06,
+      "loss": 1.003,
+      "step": 1825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4552234459484642,
+      "learning_rate": 6.851871753354154e-06,
+      "loss": 0.8328,
+      "step": 1826
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.691485249724544,
+      "learning_rate": 6.841534321812388e-06,
+      "loss": 1.0019,
+      "step": 1827
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3615553226056506,
+      "learning_rate": 6.8312006381478216e-06,
+      "loss": 1.1198,
+      "step": 1828
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6524645449439104,
+      "learning_rate": 6.8208707146225324e-06,
+      "loss": 1.2836,
+      "step": 1829
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5490809553946286,
+      "learning_rate": 6.810544563494153e-06,
+      "loss": 1.0638,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6164704789183157,
+      "learning_rate": 6.800222197015829e-06,
+      "loss": 1.1757,
+      "step": 1831
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6022129167679178,
+      "learning_rate": 6.7899036274362115e-06,
+      "loss": 0.9884,
+      "step": 1832
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7055933621246638,
+      "learning_rate": 6.77958886699946e-06,
+      "loss": 1.146,
+      "step": 1833
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5333914698382922,
+      "learning_rate": 6.769277927945201e-06,
+      "loss": 1.0629,
+      "step": 1834
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6402355108337455,
+      "learning_rate": 6.758970822508532e-06,
+      "loss": 0.9508,
+      "step": 1835
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4558508925066594,
+      "learning_rate": 6.748667562920002e-06,
+      "loss": 0.8039,
+      "step": 1836
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6477265845253874,
+      "learning_rate": 6.738368161405593e-06,
+      "loss": 0.9559,
+      "step": 1837
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6485587660312412,
+      "learning_rate": 6.728072630186715e-06,
+      "loss": 1.0982,
+      "step": 1838
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5431495698420269,
+      "learning_rate": 6.717780981480175e-06,
+      "loss": 0.9881,
+      "step": 1839
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5417573810602639,
+      "learning_rate": 6.707493227498187e-06,
+      "loss": 1.0885,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6039948368098219,
+      "learning_rate": 6.697209380448333e-06,
+      "loss": 1.3001,
+      "step": 1841
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5422391844375475,
+      "learning_rate": 6.68692945253356e-06,
+      "loss": 1.0928,
+      "step": 1842
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.768316527689756,
+      "learning_rate": 6.6766534559521686e-06,
+      "loss": 1.1916,
+      "step": 1843
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4341847690135904,
+      "learning_rate": 6.666381402897794e-06,
+      "loss": 0.7143,
+      "step": 1844
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6398325205535565,
+      "learning_rate": 6.656113305559388e-06,
+      "loss": 1.1798,
+      "step": 1845
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5580582498639356,
+      "learning_rate": 6.645849176121213e-06,
+      "loss": 0.7411,
+      "step": 1846
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.582930656920711,
+      "learning_rate": 6.635589026762819e-06,
+      "loss": 1.0686,
+      "step": 1847
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6585640799717891,
+      "learning_rate": 6.625332869659036e-06,
+      "loss": 1.2552,
+      "step": 1848
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6655463600050409,
+      "learning_rate": 6.615080716979954e-06,
+      "loss": 0.9994,
+      "step": 1849
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5920297343892135,
+      "learning_rate": 6.604832580890917e-06,
+      "loss": 1.1457,
+      "step": 1850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5932249216712535,
+      "learning_rate": 6.5945884735524945e-06,
+      "loss": 1.1528,
+      "step": 1851
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5545904232396878,
+      "learning_rate": 6.584348407120481e-06,
+      "loss": 1.0991,
+      "step": 1852
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.47494195646299,
+      "learning_rate": 6.5741123937458776e-06,
+      "loss": 0.8488,
+      "step": 1853
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.561492940334939,
+      "learning_rate": 6.563880445574873e-06,
+      "loss": 1.034,
+      "step": 1854
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6520527691971301,
+      "learning_rate": 6.553652574748828e-06,
+      "loss": 1.2129,
+      "step": 1855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4006096654076564,
+      "learning_rate": 6.543428793404274e-06,
+      "loss": 1.0657,
+      "step": 1856
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6879427668472303,
+      "learning_rate": 6.5332091136728846e-06,
+      "loss": 1.0769,
+      "step": 1857
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5203023365764469,
+      "learning_rate": 6.5229935476814664e-06,
+      "loss": 0.7916,
+      "step": 1858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6376196768501075,
+      "learning_rate": 6.512782107551945e-06,
+      "loss": 0.9114,
+      "step": 1859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5913769404996492,
+      "learning_rate": 6.502574805401351e-06,
+      "loss": 1.096,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5990908451057825,
+      "learning_rate": 6.4923716533418035e-06,
+      "loss": 1.1482,
+      "step": 1861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6404297995211009,
+      "learning_rate": 6.482172663480497e-06,
+      "loss": 1.1503,
+      "step": 1862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5959375247449396,
+      "learning_rate": 6.471977847919691e-06,
+      "loss": 0.651,
+      "step": 1863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5494746355727408,
+      "learning_rate": 6.461787218756687e-06,
+      "loss": 1.1553,
+      "step": 1864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.565467698607377,
+      "learning_rate": 6.451600788083822e-06,
+      "loss": 1.199,
+      "step": 1865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5747446460608361,
+      "learning_rate": 6.441418567988444e-06,
+      "loss": 1.0146,
+      "step": 1866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.36932112246363435,
+      "learning_rate": 6.431240570552917e-06,
+      "loss": 1.1362,
+      "step": 1867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6172894159845662,
+      "learning_rate": 6.421066807854585e-06,
+      "loss": 1.0006,
+      "step": 1868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7158308925513351,
+      "learning_rate": 6.410897291965766e-06,
+      "loss": 1.1034,
+      "step": 1869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.46601072913766733,
+      "learning_rate": 6.400732034953749e-06,
+      "loss": 0.6832,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.578321429981474,
+      "learning_rate": 6.390571048880758e-06,
+      "loss": 1.051,
+      "step": 1871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7082141409557645,
+      "learning_rate": 6.3804143458039535e-06,
+      "loss": 1.0849,
+      "step": 1872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7478465841697114,
+      "learning_rate": 6.370261937775415e-06,
+      "loss": 0.9772,
+      "step": 1873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.550723619145891,
+      "learning_rate": 6.360113836842125e-06,
+      "loss": 1.3814,
+      "step": 1874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5019853317206143,
+      "learning_rate": 6.3499700550459554e-06,
+      "loss": 1.0911,
+      "step": 1875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5472012514243573,
+      "learning_rate": 6.339830604423647e-06,
+      "loss": 1.0033,
+      "step": 1876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6955079102675771,
+      "learning_rate": 6.3296954970068125e-06,
+      "loss": 1.2289,
+      "step": 1877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.35677003120967643,
+      "learning_rate": 6.3195647448219e-06,
+      "loss": 1.0864,
+      "step": 1878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49311713501757737,
+      "learning_rate": 6.309438359890195e-06,
+      "loss": 0.7621,
+      "step": 1879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5848992016811067,
+      "learning_rate": 6.2993163542278025e-06,
+      "loss": 1.0201,
+      "step": 1880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.734484184051005,
+      "learning_rate": 6.2891987398456265e-06,
+      "loss": 1.2242,
+      "step": 1881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5320868908964363,
+      "learning_rate": 6.2790855287493605e-06,
+      "loss": 0.863,
+      "step": 1882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.644011620935649,
+      "learning_rate": 6.268976732939478e-06,
+      "loss": 1.2134,
+      "step": 1883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.562435455145442,
+      "learning_rate": 6.2588723644112095e-06,
+      "loss": 1.0208,
+      "step": 1884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6418139899584338,
+      "learning_rate": 6.248772435154531e-06,
+      "loss": 1.1159,
+      "step": 1885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6263313155940169,
+      "learning_rate": 6.238676957154151e-06,
+      "loss": 1.0646,
+      "step": 1886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5838134537165897,
+      "learning_rate": 6.2285859423895005e-06,
+      "loss": 1.1262,
+      "step": 1887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5722968935912606,
+      "learning_rate": 6.21849940283471e-06,
+      "loss": 1.1327,
+      "step": 1888
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5916235159772455,
+      "learning_rate": 6.208417350458599e-06,
+      "loss": 0.6579,
+      "step": 1889
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.39789755046584213,
+      "learning_rate": 6.1983397972246686e-06,
+      "loss": 1.101,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6062668569306259,
+      "learning_rate": 6.188266755091073e-06,
+      "loss": 1.0691,
+      "step": 1891
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6542719613111978,
+      "learning_rate": 6.178198236010619e-06,
+      "loss": 0.9178,
+      "step": 1892
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5871934044332487,
+      "learning_rate": 6.168134251930743e-06,
+      "loss": 1.2494,
+      "step": 1893
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6051224129510447,
+      "learning_rate": 6.1580748147935046e-06,
+      "loss": 0.9103,
+      "step": 1894
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6759738578022096,
+      "learning_rate": 6.148019936535564e-06,
+      "loss": 1.268,
+      "step": 1895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5468267458193188,
+      "learning_rate": 6.137969629088174e-06,
+      "loss": 1.072,
+      "step": 1896
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5813370233708846,
+      "learning_rate": 6.127923904377159e-06,
+      "loss": 0.6512,
+      "step": 1897
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5760376412212763,
+      "learning_rate": 6.117882774322912e-06,
+      "loss": 1.2434,
+      "step": 1898
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5843029650540363,
+      "learning_rate": 6.107846250840371e-06,
+      "loss": 1.0098,
+      "step": 1899
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6885224618484364,
+      "learning_rate": 6.097814345839005e-06,
+      "loss": 1.2344,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.596568474318507,
+      "learning_rate": 6.087787071222808e-06,
+      "loss": 0.9622,
+      "step": 1901
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.759254685470349,
+      "learning_rate": 6.077764438890277e-06,
+      "loss": 0.9914,
+      "step": 1902
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6564929006860758,
+      "learning_rate": 6.067746460734399e-06,
+      "loss": 1.2182,
+      "step": 1903
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3864626967990134,
+      "learning_rate": 6.057733148642637e-06,
+      "loss": 1.0494,
+      "step": 1904
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6132202400516928,
+      "learning_rate": 6.047724514496924e-06,
+      "loss": 1.1095,
+      "step": 1905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5174704497134439,
+      "learning_rate": 6.037720570173637e-06,
+      "loss": 0.913,
+      "step": 1906
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.49341283733274405,
+      "learning_rate": 6.027721327543585e-06,
+      "loss": 0.9159,
+      "step": 1907
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5853424556840318,
+      "learning_rate": 6.0177267984720074e-06,
+      "loss": 1.0788,
+      "step": 1908
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5936728953749388,
+      "learning_rate": 6.007736994818542e-06,
+      "loss": 0.8775,
+      "step": 1909
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6150194531479851,
+      "learning_rate": 5.99775192843722e-06,
+      "loss": 1.2315,
+      "step": 1910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6466279319207325,
+      "learning_rate": 5.987771611176457e-06,
+      "loss": 1.0199,
+      "step": 1911
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6216925092907963,
+      "learning_rate": 5.977796054879028e-06,
+      "loss": 1.1213,
+      "step": 1912
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6264904894984901,
+      "learning_rate": 5.9678252713820616e-06,
+      "loss": 1.1838,
+      "step": 1913
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6328724385509992,
+      "learning_rate": 5.957859272517017e-06,
+      "loss": 1.2004,
+      "step": 1914
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5752107607058548,
+      "learning_rate": 5.947898070109686e-06,
+      "loss": 0.965,
+      "step": 1915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.31595393057909704,
+      "learning_rate": 5.93794167598016e-06,
+      "loss": 0.9037,
+      "step": 1916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6744603375755757,
+      "learning_rate": 5.9279901019428275e-06,
+      "loss": 1.3342,
+      "step": 1917
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.607069400911977,
+      "learning_rate": 5.918043359806363e-06,
+      "loss": 1.0295,
+      "step": 1918
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5532148708531329,
+      "learning_rate": 5.908101461373696e-06,
+      "loss": 1.2387,
+      "step": 1919
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5823945781599829,
+      "learning_rate": 5.898164418442018e-06,
+      "loss": 0.9173,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5823489753117057,
+      "learning_rate": 5.888232242802757e-06,
+      "loss": 1.0176,
+      "step": 1921
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8978431561638388,
+      "learning_rate": 5.878304946241562e-06,
+      "loss": 1.244,
+      "step": 1922
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.45304859663146047,
+      "learning_rate": 5.868382540538296e-06,
+      "loss": 0.87,
+      "step": 1923
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8312939113189861,
+      "learning_rate": 5.858465037467014e-06,
+      "loss": 1.1405,
+      "step": 1924
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6057647702873029,
+      "learning_rate": 5.848552448795961e-06,
+      "loss": 1.0862,
+      "step": 1925
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6483315886602936,
+      "learning_rate": 5.838644786287544e-06,
+      "loss": 1.1658,
+      "step": 1926
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6478898812330568,
+      "learning_rate": 5.828742061698325e-06,
+      "loss": 0.9093,
+      "step": 1927
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3880752178138668,
+      "learning_rate": 5.818844286779008e-06,
+      "loss": 1.0749,
+      "step": 1928
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5962121997826215,
+      "learning_rate": 5.8089514732744245e-06,
+      "loss": 1.2783,
+      "step": 1929
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7898889625517892,
+      "learning_rate": 5.799063632923523e-06,
+      "loss": 0.8978,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5800926787859898,
+      "learning_rate": 5.789180777459337e-06,
+      "loss": 1.0358,
+      "step": 1931
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6315046665199877,
+      "learning_rate": 5.779302918608998e-06,
+      "loss": 0.9912,
+      "step": 1932
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.41165088630108226,
+      "learning_rate": 5.769430068093708e-06,
+      "loss": 0.8998,
+      "step": 1933
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6297028922062043,
+      "learning_rate": 5.7595622376287144e-06,
+      "loss": 0.9184,
+      "step": 1934
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5624390008712857,
+      "learning_rate": 5.749699438923324e-06,
+      "loss": 1.0117,
+      "step": 1935
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5742240684395015,
+      "learning_rate": 5.739841683680855e-06,
+      "loss": 1.0623,
+      "step": 1936
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.864924534632004,
+      "learning_rate": 5.7299889835986555e-06,
+      "loss": 1.0055,
+      "step": 1937
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5537577424750562,
+      "learning_rate": 5.720141350368072e-06,
+      "loss": 1.0662,
+      "step": 1938
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6823517632462021,
+      "learning_rate": 5.7102987956744315e-06,
+      "loss": 1.1137,
+      "step": 1939
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3681652879928452,
+      "learning_rate": 5.700461331197041e-06,
+      "loss": 1.1703,
+      "step": 1940
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4591989640845307,
+      "learning_rate": 5.690628968609172e-06,
+      "loss": 0.7423,
+      "step": 1941
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6099659171847991,
+      "learning_rate": 5.680801719578026e-06,
+      "loss": 1.0423,
+      "step": 1942
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6579828058301525,
+      "learning_rate": 5.670979595764754e-06,
+      "loss": 1.1547,
+      "step": 1943
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5599436270612447,
+      "learning_rate": 5.66116260882442e-06,
+      "loss": 0.9824,
+      "step": 1944
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5421369823154413,
+      "learning_rate": 5.6513507704059835e-06,
+      "loss": 1.0408,
+      "step": 1945
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5747752112744381,
+      "learning_rate": 5.6415440921523045e-06,
+      "loss": 1.1856,
+      "step": 1946
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6788289436816123,
+      "learning_rate": 5.6317425857001214e-06,
+      "loss": 0.9303,
+      "step": 1947
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6456606812734857,
+      "learning_rate": 5.621946262680026e-06,
+      "loss": 1.0374,
+      "step": 1948
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5863169310017314,
+      "learning_rate": 5.612155134716466e-06,
+      "loss": 1.1222,
+      "step": 1949
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49030905505766764,
+      "learning_rate": 5.6023692134277296e-06,
+      "loss": 0.8825,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5995471442934329,
+      "learning_rate": 5.5925885104259115e-06,
+      "loss": 1.0385,
+      "step": 1951
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6206529069980593,
+      "learning_rate": 5.5828130373169274e-06,
+      "loss": 0.9773,
+      "step": 1952
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5888976283456936,
+      "learning_rate": 5.573042805700488e-06,
+      "loss": 1.1061,
+      "step": 1953
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.38617388974206196,
+      "learning_rate": 5.5632778271700705e-06,
+      "loss": 1.0381,
+      "step": 1954
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8018352243716472,
+      "learning_rate": 5.553518113312936e-06,
+      "loss": 1.2318,
+      "step": 1955
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5028138908163671,
+      "learning_rate": 5.543763675710082e-06,
+      "loss": 1.0665,
+      "step": 1956
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6425704398552563,
+      "learning_rate": 5.534014525936258e-06,
+      "loss": 1.2438,
+      "step": 1957
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6102297310833057,
+      "learning_rate": 5.524270675559938e-06,
+      "loss": 0.9717,
+      "step": 1958
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5582163966759608,
+      "learning_rate": 5.514532136143296e-06,
+      "loss": 0.8912,
+      "step": 1959
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.41568131587820617,
+      "learning_rate": 5.504798919242216e-06,
+      "loss": 0.7129,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8704830295837914,
+      "learning_rate": 5.495071036406263e-06,
+      "loss": 1.2388,
+      "step": 1961
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6712658947832851,
+      "learning_rate": 5.4853484991786686e-06,
+      "loss": 1.2137,
+      "step": 1962
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6842459250497742,
+      "learning_rate": 5.475631319096325e-06,
+      "loss": 1.0989,
+      "step": 1963
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6190538228805453,
+      "learning_rate": 5.465919507689772e-06,
+      "loss": 0.9502,
+      "step": 1964
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3904716170432197,
+      "learning_rate": 5.4562130764831645e-06,
+      "loss": 1.1277,
+      "step": 1965
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5513703509998014,
+      "learning_rate": 5.446512036994287e-06,
+      "loss": 0.9478,
+      "step": 1966
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6517066922824777,
+      "learning_rate": 5.436816400734526e-06,
+      "loss": 1.1987,
+      "step": 1967
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4921573492227528,
+      "learning_rate": 5.427126179208844e-06,
+      "loss": 1.08,
+      "step": 1968
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4914255149284396,
+      "learning_rate": 5.4174413839157865e-06,
+      "loss": 0.8285,
+      "step": 1969
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6179070788444037,
+      "learning_rate": 5.407762026347469e-06,
+      "loss": 1.0981,
+      "step": 1970
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5822048468721,
+      "learning_rate": 5.398088117989535e-06,
+      "loss": 0.9591,
+      "step": 1971
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5913844944266359,
+      "learning_rate": 5.388419670321177e-06,
+      "loss": 1.3395,
+      "step": 1972
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6677073446185213,
+      "learning_rate": 5.378756694815106e-06,
+      "loss": 0.9332,
+      "step": 1973
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6342663340409184,
+      "learning_rate": 5.369099202937528e-06,
+      "loss": 1.1252,
+      "step": 1974
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5708136044859343,
+      "learning_rate": 5.359447206148161e-06,
+      "loss": 0.9318,
+      "step": 1975
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6995793138570234,
+      "learning_rate": 5.349800715900182e-06,
+      "loss": 1.2878,
+      "step": 1976
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.42477354758561575,
+      "learning_rate": 5.34015974364025e-06,
+      "loss": 0.7296,
+      "step": 1977
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3854680118371307,
+      "learning_rate": 5.330524300808473e-06,
+      "loss": 1.0148,
+      "step": 1978
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6147807897958151,
+      "learning_rate": 5.320894398838388e-06,
+      "loss": 1.1604,
+      "step": 1979
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5314112647208283,
+      "learning_rate": 5.311270049156967e-06,
+      "loss": 1.1693,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5598148394181668,
+      "learning_rate": 5.301651263184595e-06,
+      "loss": 1.0752,
+      "step": 1981
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6301469865838644,
+      "learning_rate": 5.292038052335045e-06,
+      "loss": 1.0743,
+      "step": 1982
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6325983625458872,
+      "learning_rate": 5.2824304280154815e-06,
+      "loss": 1.0322,
+      "step": 1983
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6973631037511987,
+      "learning_rate": 5.272828401626443e-06,
+      "loss": 1.3125,
+      "step": 1984
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7099342712388711,
+      "learning_rate": 5.263231984561814e-06,
+      "loss": 1.0457,
+      "step": 1985
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4051026318528006,
+      "learning_rate": 5.253641188208836e-06,
+      "loss": 0.7615,
+      "step": 1986
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5519287435197433,
+      "learning_rate": 5.244056023948075e-06,
+      "loss": 0.9552,
+      "step": 1987
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6361492257668449,
+      "learning_rate": 5.234476503153409e-06,
+      "loss": 1.2248,
+      "step": 1988
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3904727115211689,
+      "learning_rate": 5.2249026371920265e-06,
+      "loss": 1.0813,
+      "step": 1989
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5910408826099562,
+      "learning_rate": 5.2153344374244086e-06,
+      "loss": 1.0353,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6040426189120501,
+      "learning_rate": 5.2057719152042985e-06,
+      "loss": 0.9839,
+      "step": 1991
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5426711717507292,
+      "learning_rate": 5.196215081878717e-06,
+      "loss": 1.1144,
+      "step": 1992
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6281038827305634,
+      "learning_rate": 5.186663948787932e-06,
+      "loss": 1.1709,
+      "step": 1993
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4202148792986245,
+      "learning_rate": 5.177118527265438e-06,
+      "loss": 0.7749,
+      "step": 1994
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5681903237898593,
+      "learning_rate": 5.167578828637965e-06,
+      "loss": 1.1903,
+      "step": 1995
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7322144114618212,
+      "learning_rate": 5.15804486422544e-06,
+      "loss": 1.0453,
+      "step": 1996
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5896327140343831,
+      "learning_rate": 5.148516645340992e-06,
+      "loss": 0.9842,
+      "step": 1997
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6020550583734402,
+      "learning_rate": 5.138994183290937e-06,
+      "loss": 1.0798,
+      "step": 1998
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6524346561203568,
+      "learning_rate": 5.129477489374749e-06,
+      "loss": 1.056,
+      "step": 1999
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.767546334578665,
+      "learning_rate": 5.119966574885065e-06,
+      "loss": 1.2612,
+      "step": 2000
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6030549695654147,
+      "learning_rate": 5.110461451107664e-06,
+      "loss": 0.9554,
+      "step": 2001
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.36678351242254237,
+      "learning_rate": 5.100962129321448e-06,
+      "loss": 1.0983,
+      "step": 2002
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.48699000570646334,
+      "learning_rate": 5.09146862079844e-06,
+      "loss": 1.1952,
+      "step": 2003
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.45724472087564466,
+      "learning_rate": 5.081980936803766e-06,
+      "loss": 0.6538,
+      "step": 2004
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6186976312310234,
+      "learning_rate": 5.07249908859563e-06,
+      "loss": 1.1511,
+      "step": 2005
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6160539997823697,
+      "learning_rate": 5.063023087425324e-06,
+      "loss": 1.1609,
+      "step": 2006
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7582946736063471,
+      "learning_rate": 5.053552944537199e-06,
+      "loss": 1.1867,
+      "step": 2007
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5904381752867235,
+      "learning_rate": 5.044088671168644e-06,
+      "loss": 0.8994,
+      "step": 2008
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6077808888081304,
+      "learning_rate": 5.034630278550095e-06,
+      "loss": 1.066,
+      "step": 2009
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6542632961283458,
+      "learning_rate": 5.0251777779050114e-06,
+      "loss": 1.1447,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5584155187315218,
+      "learning_rate": 5.015731180449847e-06,
+      "loss": 0.9433,
+      "step": 2011
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.46795159634460654,
+      "learning_rate": 5.006290497394066e-06,
+      "loss": 0.8431,
+      "step": 2012
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6341594928600376,
+      "learning_rate": 4.996855739940101e-06,
+      "loss": 0.925,
+      "step": 2013
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3553434479672026,
+      "learning_rate": 4.9874269192833645e-06,
+      "loss": 1.2091,
+      "step": 2014
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6293803152358128,
+      "learning_rate": 4.978004046612224e-06,
+      "loss": 1.0439,
+      "step": 2015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5238745722723024,
+      "learning_rate": 4.968587133107975e-06,
+      "loss": 1.0653,
+      "step": 2016
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6994622830529673,
+      "learning_rate": 4.959176189944859e-06,
+      "loss": 1.176,
+      "step": 2017
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5260549163404403,
+      "learning_rate": 4.9497712282900265e-06,
+      "loss": 0.8956,
+      "step": 2018
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8030197541857261,
+      "learning_rate": 4.9403722593035245e-06,
+      "loss": 1.252,
+      "step": 2019
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5885803689520657,
+      "learning_rate": 4.930979294138296e-06,
+      "loss": 1.0192,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6470750182054976,
+      "learning_rate": 4.921592343940161e-06,
+      "loss": 1.0619,
+      "step": 2021
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4406901396338242,
+      "learning_rate": 4.912211419847795e-06,
+      "loss": 0.786,
+      "step": 2022
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6512892903898472,
+      "learning_rate": 4.902836532992727e-06,
+      "loss": 0.9219,
+      "step": 2023
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5447935912263774,
+      "learning_rate": 4.8934676944993265e-06,
+      "loss": 1.0721,
+      "step": 2024
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5533329048509985,
+      "learning_rate": 4.8841049154847744e-06,
+      "loss": 0.9922,
+      "step": 2025
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5578664236655886,
+      "learning_rate": 4.874748207059071e-06,
+      "loss": 1.0874,
+      "step": 2026
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.34947155540785196,
+      "learning_rate": 4.865397580325016e-06,
+      "loss": 1.2117,
+      "step": 2027
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6100272155349724,
+      "learning_rate": 4.8560530463781775e-06,
+      "loss": 0.9608,
+      "step": 2028
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5097259363686736,
+      "learning_rate": 4.846714616306908e-06,
+      "loss": 0.8619,
+      "step": 2029
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7041439828560394,
+      "learning_rate": 4.837382301192316e-06,
+      "loss": 0.9545,
+      "step": 2030
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5309183875011766,
+      "learning_rate": 4.828056112108243e-06,
+      "loss": 1.1505,
+      "step": 2031
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6391795388051007,
+      "learning_rate": 4.818736060121274e-06,
+      "loss": 0.7969,
+      "step": 2032
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5293104610445706,
+      "learning_rate": 4.8094221562906996e-06,
+      "loss": 0.9766,
+      "step": 2033
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6557894370098634,
+      "learning_rate": 4.800114411668527e-06,
+      "loss": 1.2636,
+      "step": 2034
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6109733953886896,
+      "learning_rate": 4.790812837299451e-06,
+      "loss": 0.9629,
+      "step": 2035
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5573196076692472,
+      "learning_rate": 4.781517444220836e-06,
+      "loss": 1.0721,
+      "step": 2036
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5358610114697743,
+      "learning_rate": 4.772228243462722e-06,
+      "loss": 1.2456,
+      "step": 2037
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7020837237946431,
+      "learning_rate": 4.762945246047801e-06,
+      "loss": 0.9925,
+      "step": 2038
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.48591534271018477,
+      "learning_rate": 4.7536684629913945e-06,
+      "loss": 0.8391,
+      "step": 2039
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6031833352119006,
+      "learning_rate": 4.744397905301459e-06,
+      "loss": 1.052,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3834690548623736,
+      "learning_rate": 4.7351335839785655e-06,
+      "loss": 1.1447,
+      "step": 2041
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5679275783123425,
+      "learning_rate": 4.72587551001587e-06,
+      "loss": 0.8772,
+      "step": 2042
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6599737438755665,
+      "learning_rate": 4.716623694399134e-06,
+      "loss": 1.1835,
+      "step": 2043
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6111927803105563,
+      "learning_rate": 4.707378148106684e-06,
+      "loss": 0.8194,
+      "step": 2044
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5949404019049055,
+      "learning_rate": 4.698138882109403e-06,
+      "loss": 0.817,
+      "step": 2045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6095355706867437,
+      "learning_rate": 4.688905907370729e-06,
+      "loss": 1.1303,
+      "step": 2046
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5900117235314186,
+      "learning_rate": 4.679679234846636e-06,
+      "loss": 1.0929,
+      "step": 2047
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4408673800312333,
+      "learning_rate": 4.670458875485607e-06,
+      "loss": 0.8349,
+      "step": 2048
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6532045305972171,
+      "learning_rate": 4.6612448402286485e-06,
+      "loss": 0.9924,
+      "step": 2049
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6271829612031671,
+      "learning_rate": 4.652037140009259e-06,
+      "loss": 1.2825,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.579819542052795,
+      "learning_rate": 4.642835785753412e-06,
+      "loss": 1.0675,
+      "step": 2051
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4560386042039919,
+      "learning_rate": 4.63364078837956e-06,
+      "loss": 1.1232,
+      "step": 2052
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6861313197348113,
+      "learning_rate": 4.624452158798604e-06,
+      "loss": 1.1248,
+      "step": 2053
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5346538579709584,
+      "learning_rate": 4.615269907913896e-06,
+      "loss": 0.9832,
+      "step": 2054
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5941720781706972,
+      "learning_rate": 4.606094046621222e-06,
+      "loss": 1.1132,
+      "step": 2055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4724981927283896,
+      "learning_rate": 4.596924585808772e-06,
+      "loss": 0.672,
+      "step": 2056
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5761767658897454,
+      "learning_rate": 4.587761536357153e-06,
+      "loss": 1.0284,
+      "step": 2057
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7040438665674074,
+      "learning_rate": 4.5786049091393645e-06,
+      "loss": 1.1578,
+      "step": 2058
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6778823926330952,
+      "learning_rate": 4.5694547150207745e-06,
+      "loss": 1.0278,
+      "step": 2059
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6202822058151938,
+      "learning_rate": 4.560310964859129e-06,
+      "loss": 1.4384,
+      "step": 2060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5187139230768162,
+      "learning_rate": 4.551173669504525e-06,
+      "loss": 1.0691,
+      "step": 2061
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5544624194882747,
+      "learning_rate": 4.542042839799392e-06,
+      "loss": 1.1994,
+      "step": 2062
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6831268382842568,
+      "learning_rate": 4.532918486578497e-06,
+      "loss": 0.9665,
+      "step": 2063
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3658137073493532,
+      "learning_rate": 4.523800620668921e-06,
+      "loss": 1.0504,
+      "step": 2064
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4719243802934354,
+      "learning_rate": 4.514689252890039e-06,
+      "loss": 0.7557,
+      "step": 2065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6372281873588925,
+      "learning_rate": 4.505584394053523e-06,
+      "loss": 1.0837,
+      "step": 2066
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.60040327319363,
+      "learning_rate": 4.496486054963323e-06,
+      "loss": 0.9978,
+      "step": 2067
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6084236096907124,
+      "learning_rate": 4.487394246415641e-06,
+      "loss": 1.008,
+      "step": 2068
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8051755998001795,
+      "learning_rate": 4.47830897919894e-06,
+      "loss": 1.2581,
+      "step": 2069
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6291927345856505,
+      "learning_rate": 4.469230264093924e-06,
+      "loss": 0.9762,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6166505574532433,
+      "learning_rate": 4.460158111873511e-06,
+      "loss": 1.0858,
+      "step": 2071
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6006902765163586,
+      "learning_rate": 4.45109253330284e-06,
+      "loss": 1.076,
+      "step": 2072
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.76158522985408,
+      "learning_rate": 4.442033539139241e-06,
+      "loss": 1.0867,
+      "step": 2073
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5261134315934477,
+      "learning_rate": 4.43298114013224e-06,
+      "loss": 1.1411,
+      "step": 2074
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.43413919637250886,
+      "learning_rate": 4.423935347023538e-06,
+      "loss": 0.6721,
+      "step": 2075
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4158273706577782,
+      "learning_rate": 4.414896170546985e-06,
+      "loss": 1.0866,
+      "step": 2076
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6005943336625283,
+      "learning_rate": 4.405863621428591e-06,
+      "loss": 1.0227,
+      "step": 2077
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7756026008296718,
+      "learning_rate": 4.396837710386503e-06,
+      "loss": 0.9646,
+      "step": 2078
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.569975609816249,
+      "learning_rate": 4.387818448130977e-06,
+      "loss": 1.2214,
+      "step": 2079
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.629642003018112,
+      "learning_rate": 4.378805845364395e-06,
+      "loss": 0.9262,
+      "step": 2080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6558145468300554,
+      "learning_rate": 4.369799912781233e-06,
+      "loss": 1.127,
+      "step": 2081
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.657359050040428,
+      "learning_rate": 4.360800661068043e-06,
+      "loss": 1.1981,
+      "step": 2082
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4257248572478554,
+      "learning_rate": 4.35180810090346e-06,
+      "loss": 0.6817,
+      "step": 2083
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5522694254371675,
+      "learning_rate": 4.34282224295818e-06,
+      "loss": 1.1908,
+      "step": 2084
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.561247628150593,
+      "learning_rate": 4.333843097894932e-06,
+      "loss": 1.0202,
+      "step": 2085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6571882355357918,
+      "learning_rate": 4.324870676368491e-06,
+      "loss": 1.2988,
+      "step": 2086
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5951414503311463,
+      "learning_rate": 4.315904989025657e-06,
+      "loss": 0.9714,
+      "step": 2087
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6389459831051107,
+      "learning_rate": 4.306946046505226e-06,
+      "loss": 0.9625,
+      "step": 2088
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6579589492704967,
+      "learning_rate": 4.297993859438003e-06,
+      "loss": 1.1948,
+      "step": 2089
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.39351965230718783,
+      "learning_rate": 4.289048438446766e-06,
+      "loss": 1.0456,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5571336455453639,
+      "learning_rate": 4.2801097941462755e-06,
+      "loss": 0.9286,
+      "step": 2091
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5816223973638044,
+      "learning_rate": 4.271177937143245e-06,
+      "loss": 1.0402,
+      "step": 2092
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.444759466811791,
+      "learning_rate": 4.2622528780363295e-06,
+      "loss": 0.9305,
+      "step": 2093
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5793860545670172,
+      "learning_rate": 4.253334627416124e-06,
+      "loss": 0.9754,
+      "step": 2094
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6488475518125807,
+      "learning_rate": 4.244423195865149e-06,
+      "loss": 0.8807,
+      "step": 2095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6006921074428845,
+      "learning_rate": 4.2355185939578184e-06,
+      "loss": 1.2494,
+      "step": 2096
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6397092388442902,
+      "learning_rate": 4.226620832260454e-06,
+      "loss": 0.9993,
+      "step": 2097
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6083014836399611,
+      "learning_rate": 4.217729921331263e-06,
+      "loss": 1.1644,
+      "step": 2098
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5937978598299868,
+      "learning_rate": 4.208845871720309e-06,
+      "loss": 1.1945,
+      "step": 2099
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7301152113221664,
+      "learning_rate": 4.199968693969526e-06,
+      "loss": 1.13,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4167360843702326,
+      "learning_rate": 4.191098398612696e-06,
+      "loss": 0.6759,
+      "step": 2101
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3757496554262989,
+      "learning_rate": 4.182234996175421e-06,
+      "loss": 1.0323,
+      "step": 2102
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6218405293592891,
+      "learning_rate": 4.173378497175137e-06,
+      "loss": 1.3337,
+      "step": 2103
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6332221463178977,
+      "learning_rate": 4.164528912121084e-06,
+      "loss": 0.8603,
+      "step": 2104
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6038743494579316,
+      "learning_rate": 4.1556862515142905e-06,
+      "loss": 1.2598,
+      "step": 2105
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5522596437573986,
+      "learning_rate": 4.1468505258475785e-06,
+      "loss": 0.8929,
+      "step": 2106
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5777824087963662,
+      "learning_rate": 4.13802174560554e-06,
+      "loss": 0.9849,
+      "step": 2107
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6856857386515897,
+      "learning_rate": 4.129199921264518e-06,
+      "loss": 1.2148,
+      "step": 2108
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3979842306510994,
+      "learning_rate": 4.120385063292609e-06,
+      "loss": 0.8453,
+      "step": 2109
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8103118814154855,
+      "learning_rate": 4.111577182149635e-06,
+      "loss": 1.1314,
+      "step": 2110
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1976611102597547,
+      "learning_rate": 4.1027762882871496e-06,
+      "loss": 1.0986,
+      "step": 2111
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6676411615234094,
+      "learning_rate": 4.093982392148411e-06,
+      "loss": 1.1516,
+      "step": 2112
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5597823366828354,
+      "learning_rate": 4.085195504168367e-06,
+      "loss": 0.9543,
+      "step": 2113
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3436960271389281,
+      "learning_rate": 4.07641563477366e-06,
+      "loss": 1.1402,
+      "step": 2114
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5988735500293194,
+      "learning_rate": 4.0676427943826e-06,
+      "loss": 1.1564,
+      "step": 2115
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6287241868022766,
+      "learning_rate": 4.0588769934051495e-06,
+      "loss": 0.8514,
+      "step": 2116
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5434832954955672,
+      "learning_rate": 4.050118242242929e-06,
+      "loss": 1.0651,
+      "step": 2117
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6712068568409073,
+      "learning_rate": 4.041366551289191e-06,
+      "loss": 1.0325,
+      "step": 2118
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.41437108843645926,
+      "learning_rate": 4.0326219309288e-06,
+      "loss": 0.951,
+      "step": 2119
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5848408432689475,
+      "learning_rate": 4.023884391538244e-06,
+      "loss": 0.9211,
+      "step": 2120
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9729848139857786,
+      "learning_rate": 4.015153943485606e-06,
+      "loss": 0.9814,
+      "step": 2121
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5848335140050243,
+      "learning_rate": 4.0064305971305416e-06,
+      "loss": 1.0708,
+      "step": 2122
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6837083188178108,
+      "learning_rate": 3.997714362824295e-06,
+      "loss": 1.0067,
+      "step": 2123
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7146265711912283,
+      "learning_rate": 3.989005250909668e-06,
+      "loss": 1.1479,
+      "step": 2124
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6152153588246168,
+      "learning_rate": 3.9803032717210024e-06,
+      "loss": 0.9754,
+      "step": 2125
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.34439721637689613,
+      "learning_rate": 3.971608435584183e-06,
+      "loss": 1.1562,
+      "step": 2126
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.49326896740799403,
+      "learning_rate": 3.9629207528166224e-06,
+      "loss": 0.7491,
+      "step": 2127
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5505341700807806,
+      "learning_rate": 3.954240233727234e-06,
+      "loss": 0.992,
+      "step": 2128
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6236640171788971,
+      "learning_rate": 3.945566888616442e-06,
+      "loss": 1.0858,
+      "step": 2129
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5670098487825718,
+      "learning_rate": 3.936900727776146e-06,
+      "loss": 1.0846,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5176765993567863,
+      "learning_rate": 3.928241761489732e-06,
+      "loss": 1.0356,
+      "step": 2131
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5695447534737139,
+      "learning_rate": 3.919590000032046e-06,
+      "loss": 1.1788,
+      "step": 2132
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5528950306381658,
+      "learning_rate": 3.910945453669375e-06,
+      "loss": 0.8666,
+      "step": 2133
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6369710580824364,
+      "learning_rate": 3.902308132659457e-06,
+      "loss": 1.0745,
+      "step": 2134
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6964158566159449,
+      "learning_rate": 3.8936780472514555e-06,
+      "loss": 1.0756,
+      "step": 2135
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4700892424916087,
+      "learning_rate": 3.885055207685937e-06,
+      "loss": 0.8928,
+      "step": 2136
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8368943348359603,
+      "learning_rate": 3.876439624194879e-06,
+      "loss": 1.0584,
+      "step": 2137
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6608905452095184,
+      "learning_rate": 3.867831307001652e-06,
+      "loss": 0.9667,
+      "step": 2138
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.39337531392017444,
+      "learning_rate": 3.859230266320992e-06,
+      "loss": 1.1383,
+      "step": 2139
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6184462840075517,
+      "learning_rate": 3.85063651235901e-06,
+      "loss": 0.86,
+      "step": 2140
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7248776587890676,
+      "learning_rate": 3.842050055313174e-06,
+      "loss": 1.2021,
+      "step": 2141
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.494658974971039,
+      "learning_rate": 3.833470905372278e-06,
+      "loss": 1.075,
+      "step": 2142
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5723104936581375,
+      "learning_rate": 3.8248990727164616e-06,
+      "loss": 1.2011,
+      "step": 2143
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5791768594629465,
+      "learning_rate": 3.8163345675171745e-06,
+      "loss": 0.961,
+      "step": 2144
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6114813813303155,
+      "learning_rate": 3.807777399937168e-06,
+      "loss": 0.9817,
+      "step": 2145
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4137957145445829,
+      "learning_rate": 3.7992275801304935e-06,
+      "loss": 0.7339,
+      "step": 2146
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6600542624969092,
+      "learning_rate": 3.790685118242482e-06,
+      "loss": 1.2576,
+      "step": 2147
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7597708615018006,
+      "learning_rate": 3.7821500244097276e-06,
+      "loss": 1.1493,
+      "step": 2148
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6301432291221892,
+      "learning_rate": 3.773622308760091e-06,
+      "loss": 1.1057,
+      "step": 2149
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5983064295795961,
+      "learning_rate": 3.7651019814126656e-06,
+      "loss": 1.0082,
+      "step": 2150
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.37759545890850604,
+      "learning_rate": 3.7565890524777883e-06,
+      "loss": 1.1485,
+      "step": 2151
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.512825391078499,
+      "learning_rate": 3.7480835320570173e-06,
+      "loss": 0.9976,
+      "step": 2152
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6536153869800704,
+      "learning_rate": 3.739585430243109e-06,
+      "loss": 1.1217,
+      "step": 2153
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.383787205432439,
+      "learning_rate": 3.731094757120025e-06,
+      "loss": 0.833,
+      "step": 2154
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6628488425531903,
+      "learning_rate": 3.722611522762917e-06,
+      "loss": 1.109,
+      "step": 2155
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6214726550901478,
+      "learning_rate": 3.7141357372380926e-06,
+      "loss": 1.0778,
+      "step": 2156
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5422503749212811,
+      "learning_rate": 3.7056674106030376e-06,
+      "loss": 0.9129,
+      "step": 2157
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.597511358518464,
+      "learning_rate": 3.6972065529063837e-06,
+      "loss": 1.3263,
+      "step": 2158
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6045008167346052,
+      "learning_rate": 3.688753174187888e-06,
+      "loss": 0.9009,
+      "step": 2159
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6183417391963012,
+      "learning_rate": 3.6803072844784473e-06,
+      "loss": 1.133,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5810866600488653,
+      "learning_rate": 3.6718688938000667e-06,
+      "loss": 0.8933,
+      "step": 2161
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6232555709669914,
+      "learning_rate": 3.6634380121658484e-06,
+      "loss": 1.2297,
+      "step": 2162
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4287546138841184,
+      "learning_rate": 3.65501464957999e-06,
+      "loss": 0.7095,
+      "step": 2163
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3762774559237429,
+      "learning_rate": 3.6465988160377695e-06,
+      "loss": 1.0363,
+      "step": 2164
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5967454440857434,
+      "learning_rate": 3.6381905215255187e-06,
+      "loss": 1.1381,
+      "step": 2165
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.48983775100814386,
+      "learning_rate": 3.6297897760206347e-06,
+      "loss": 1.1272,
+      "step": 2166
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6273285074596875,
+      "learning_rate": 3.621396589491557e-06,
+      "loss": 1.1069,
+      "step": 2167
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.597827456176715,
+      "learning_rate": 3.6130109718977457e-06,
+      "loss": 1.0958,
+      "step": 2168
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5423833194006255,
+      "learning_rate": 3.604632933189691e-06,
+      "loss": 0.8963,
+      "step": 2169
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5943062548276571,
+      "learning_rate": 3.5962624833088788e-06,
+      "loss": 1.3475,
+      "step": 2170
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6873275130148255,
+      "learning_rate": 3.587899632187799e-06,
+      "loss": 0.9323,
+      "step": 2171
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.37525623190986135,
+      "learning_rate": 3.579544389749925e-06,
+      "loss": 0.7512,
+      "step": 2172
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6363545050722027,
+      "learning_rate": 3.571196765909691e-06,
+      "loss": 0.9412,
+      "step": 2173
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.619408904150126,
+      "learning_rate": 3.5628567705725014e-06,
+      "loss": 1.2687,
+      "step": 2174
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3639327927799437,
+      "learning_rate": 3.554524413634709e-06,
+      "loss": 1.0762,
+      "step": 2175
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.638561437232273,
+      "learning_rate": 3.546199704983592e-06,
+      "loss": 0.9495,
+      "step": 2176
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5936495725132503,
+      "learning_rate": 3.537882654497362e-06,
+      "loss": 1.084,
+      "step": 2177
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5603228695291737,
+      "learning_rate": 3.529573272045146e-06,
+      "loss": 1.0462,
+      "step": 2178
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6538485809006394,
+      "learning_rate": 3.52127156748696e-06,
+      "loss": 1.1668,
+      "step": 2179
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.38445846228374325,
+      "learning_rate": 3.512977550673721e-06,
+      "loss": 0.7816,
+      "step": 2180
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5674883958772726,
+      "learning_rate": 3.5046912314472204e-06,
+      "loss": 1.2143,
+      "step": 2181
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.65544129309934,
+      "learning_rate": 3.4964126196401094e-06,
+      "loss": 0.9694,
+      "step": 2182
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6530036473371638,
+      "learning_rate": 3.4881417250759007e-06,
+      "loss": 0.9502,
+      "step": 2183
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6064096292819138,
+      "learning_rate": 3.479878557568952e-06,
+      "loss": 1.1621,
+      "step": 2184
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6227462018521666,
+      "learning_rate": 3.4716231269244384e-06,
+      "loss": 1.0899,
+      "step": 2185
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7109067759789365,
+      "learning_rate": 3.4633754429383726e-06,
+      "loss": 1.1607,
+      "step": 2186
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5864873697155012,
+      "learning_rate": 3.455135515397556e-06,
+      "loss": 0.9778,
+      "step": 2187
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.40314126483907387,
+      "learning_rate": 3.4469033540796017e-06,
+      "loss": 1.1627,
+      "step": 2188
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.35615840521629044,
+      "learning_rate": 3.438678968752903e-06,
+      "loss": 0.8041,
+      "step": 2189
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5974142110178924,
+      "learning_rate": 3.4304623691766193e-06,
+      "loss": 0.9147,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6427948498582853,
+      "learning_rate": 3.422253565100678e-06,
+      "loss": 1.2113,
+      "step": 2191
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5907872764716388,
+      "learning_rate": 3.4140525662657597e-06,
+      "loss": 1.1641,
+      "step": 2192
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6467490805990703,
+      "learning_rate": 3.405859382403269e-06,
+      "loss": 1.1302,
+      "step": 2193
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5175418747538464,
+      "learning_rate": 3.397674023235353e-06,
+      "loss": 0.9552,
+      "step": 2194
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5570448460711517,
+      "learning_rate": 3.3894964984748692e-06,
+      "loss": 1.0161,
+      "step": 2195
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6604634196633761,
+      "learning_rate": 3.3813268178253687e-06,
+      "loss": 1.1654,
+      "step": 2196
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5685013421902334,
+      "learning_rate": 3.373164990981108e-06,
+      "loss": 1.0322,
+      "step": 2197
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.43168267628729606,
+      "learning_rate": 3.365011027627021e-06,
+      "loss": 0.8562,
+      "step": 2198
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6749062289019431,
+      "learning_rate": 3.356864937438703e-06,
+      "loss": 0.9263,
+      "step": 2199
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.34026297069132977,
+      "learning_rate": 3.3487267300824155e-06,
+      "loss": 1.179,
+      "step": 2200
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5990748283322206,
+      "learning_rate": 3.340596415215065e-06,
+      "loss": 1.0292,
+      "step": 2201
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5204063238913748,
+      "learning_rate": 3.3324740024841863e-06,
+      "loss": 0.9962,
+      "step": 2202
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6658188006762306,
+      "learning_rate": 3.324359501527944e-06,
+      "loss": 1.1295,
+      "step": 2203
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5982530527649813,
+      "learning_rate": 3.316252921975116e-06,
+      "loss": 0.9763,
+      "step": 2204
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7054174185688421,
+      "learning_rate": 3.30815427344507e-06,
+      "loss": 1.1883,
+      "step": 2205
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5691888976551835,
+      "learning_rate": 3.3000635655477773e-06,
+      "loss": 0.9744,
+      "step": 2206
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6608350750359396,
+      "learning_rate": 3.2919808078837712e-06,
+      "loss": 1.0863,
+      "step": 2207
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4332118561381112,
+      "learning_rate": 3.283906010044162e-06,
+      "loss": 0.7866,
+      "step": 2208
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6466937071800672,
+      "learning_rate": 3.275839181610616e-06,
+      "loss": 0.8947,
+      "step": 2209
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6263838916889417,
+      "learning_rate": 3.2677803321553324e-06,
+      "loss": 1.1651,
+      "step": 2210
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6086334281939717,
+      "learning_rate": 3.259729471241051e-06,
+      "loss": 1.0086,
+      "step": 2211
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5420626127650936,
+      "learning_rate": 3.251686608421033e-06,
+      "loss": 1.0584,
+      "step": 2212
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.36507685796948186,
+      "learning_rate": 3.24365175323904e-06,
+      "loss": 1.221,
+      "step": 2213
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6189911461468462,
+      "learning_rate": 3.2356249152293416e-06,
+      "loss": 0.9147,
+      "step": 2214
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.47080969872919815,
+      "learning_rate": 3.2276061039166917e-06,
+      "loss": 0.8283,
+      "step": 2215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6116332622863169,
+      "learning_rate": 3.219595328816313e-06,
+      "loss": 0.9812,
+      "step": 2216
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6186078192338866,
+      "learning_rate": 3.211592599433898e-06,
+      "loss": 1.136,
+      "step": 2217
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5498416419680126,
+      "learning_rate": 3.203597925265598e-06,
+      "loss": 0.8584,
+      "step": 2218
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.612025710644167,
+      "learning_rate": 3.1956113157979906e-06,
+      "loss": 0.9929,
+      "step": 2219
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6575001330741064,
+      "learning_rate": 3.1876327805080966e-06,
+      "loss": 1.2728,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5612645195725545,
+      "learning_rate": 3.1796623288633544e-06,
+      "loss": 1.0431,
+      "step": 2221
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.591727711267572,
+      "learning_rate": 3.171699970321601e-06,
+      "loss": 1.0729,
+      "step": 2222
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5218651461566923,
+      "learning_rate": 3.163745714331079e-06,
+      "loss": 1.2237,
+      "step": 2223
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6430123243867981,
+      "learning_rate": 3.155799570330419e-06,
+      "loss": 1.1627,
+      "step": 2224
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4203869534440781,
+      "learning_rate": 3.147861547748612e-06,
+      "loss": 0.7044,
+      "step": 2225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5282170620562382,
+      "learning_rate": 3.1399316560050287e-06,
+      "loss": 1.0047,
+      "step": 2226
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.385316224656729,
+      "learning_rate": 3.1320099045093766e-06,
+      "loss": 1.1705,
+      "step": 2227
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6084024052040216,
+      "learning_rate": 3.1240963026617142e-06,
+      "loss": 0.9343,
+      "step": 2228
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6247211165212978,
+      "learning_rate": 3.1161908598524293e-06,
+      "loss": 1.1164,
+      "step": 2229
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6084870440990048,
+      "learning_rate": 3.1082935854622186e-06,
+      "loss": 0.8585,
+      "step": 2230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5900784146736906,
+      "learning_rate": 3.1004044888620963e-06,
+      "loss": 0.8522,
+      "step": 2231
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5940383104748846,
+      "learning_rate": 3.0925235794133725e-06,
+      "loss": 1.1306,
+      "step": 2232
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9709744661555897,
+      "learning_rate": 3.08465086646763e-06,
+      "loss": 1.0943,
+      "step": 2233
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4203937407426138,
+      "learning_rate": 3.0767863593667404e-06,
+      "loss": 0.8325,
+      "step": 2234
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.625744172624661,
+      "learning_rate": 3.0689300674428325e-06,
+      "loss": 0.9978,
+      "step": 2235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5497840216548314,
+      "learning_rate": 3.0610820000182817e-06,
+      "loss": 1.2801,
+      "step": 2236
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.808075911386779,
+      "learning_rate": 3.053242166405711e-06,
+      "loss": 1.0725,
+      "step": 2237
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.34522385798194555,
+      "learning_rate": 3.0454105759079745e-06,
+      "loss": 1.1073,
+      "step": 2238
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5688404069582282,
+      "learning_rate": 3.0375872378181336e-06,
+      "loss": 1.0945,
+      "step": 2239
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.608140739260357,
+      "learning_rate": 3.0297721614194697e-06,
+      "loss": 0.9383,
+      "step": 2240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5658186808440911,
+      "learning_rate": 3.021965355985459e-06,
+      "loss": 1.1441,
+      "step": 2241
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.444413064177629,
+      "learning_rate": 3.014166830779753e-06,
+      "loss": 0.675,
+      "step": 2242
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.637585752343436,
+      "learning_rate": 3.0063765950561884e-06,
+      "loss": 1.1463,
+      "step": 2243
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6598205302803303,
+      "learning_rate": 2.9985946580587653e-06,
+      "loss": 0.9721,
+      "step": 2244
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6769867942058577,
+      "learning_rate": 2.9908210290216266e-06,
+      "loss": 0.9819,
+      "step": 2245
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6135338850182529,
+      "learning_rate": 2.98305571716907e-06,
+      "loss": 1.4317,
+      "step": 2246
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4893644386793855,
+      "learning_rate": 2.9752987317155115e-06,
+      "loss": 1.0689,
+      "step": 2247
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5428946762080946,
+      "learning_rate": 2.9675500818654956e-06,
+      "loss": 1.1783,
+      "step": 2248
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.786783739239104,
+      "learning_rate": 2.959809776813676e-06,
+      "loss": 0.8874,
+      "step": 2249
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3416348546942729,
+      "learning_rate": 2.9520778257447935e-06,
+      "loss": 1.0572,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5024674732531793,
+      "learning_rate": 2.944354237833689e-06,
+      "loss": 0.7728,
+      "step": 2251
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5946783936878172,
+      "learning_rate": 2.9366390222452756e-06,
+      "loss": 1.0128,
+      "step": 2252
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5973832457872528,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 1.0262,
+      "step": 2253
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5492449743875691,
+      "learning_rate": 2.921233744646472e-06,
+      "loss": 0.9842,
+      "step": 2254
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6627898975325832,
+      "learning_rate": 2.913543700916194e-06,
+      "loss": 1.1858,
+      "step": 2255
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6466280519555071,
+      "learning_rate": 2.905862066068793e-06,
+      "loss": 0.9826,
+      "step": 2256
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.859333112361138,
+      "learning_rate": 2.8981888492194e-06,
+      "loss": 1.0789,
+      "step": 2257
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6442918004458631,
+      "learning_rate": 2.890524059473161e-06,
+      "loss": 1.1434,
+      "step": 2258
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4956728101196375,
+      "learning_rate": 2.8828677059252107e-06,
+      "loss": 1.1424,
+      "step": 2259
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.44252454228731486,
+      "learning_rate": 2.875219797660681e-06,
+      "loss": 0.7842,
+      "step": 2260
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.557919947766656,
+      "learning_rate": 2.8675803437546854e-06,
+      "loss": 0.9575,
+      "step": 2261
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3709972928593833,
+      "learning_rate": 2.8599493532722954e-06,
+      "loss": 1.0794,
+      "step": 2262
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6134908956668577,
+      "learning_rate": 2.8523268352685495e-06,
+      "loss": 1.102,
+      "step": 2263
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5630675086843554,
+      "learning_rate": 2.8447127987884247e-06,
+      "loss": 1.0257,
+      "step": 2264
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.564155616834009,
+      "learning_rate": 2.83710725286684e-06,
+      "loss": 1.1765,
+      "step": 2265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6200010784994556,
+      "learning_rate": 2.8295102065286385e-06,
+      "loss": 0.8871,
+      "step": 2266
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6016301660594479,
+      "learning_rate": 2.821921668788571e-06,
+      "loss": 1.2544,
+      "step": 2267
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4886408075807357,
+      "learning_rate": 2.8143416486513e-06,
+      "loss": 0.701,
+      "step": 2268
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5678805351409051,
+      "learning_rate": 2.80677015511138e-06,
+      "loss": 0.9122,
+      "step": 2269
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4812764293964824,
+      "learning_rate": 2.799207197153241e-06,
+      "loss": 1.2202,
+      "step": 2270
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5856166351298977,
+      "learning_rate": 2.791652783751191e-06,
+      "loss": 1.1105,
+      "step": 2271
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6861386657026765,
+      "learning_rate": 2.7841069238693997e-06,
+      "loss": 1.1596,
+      "step": 2272
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6269089740106483,
+      "learning_rate": 2.776569626461879e-06,
+      "loss": 0.9717,
+      "step": 2273
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7607729757735184,
+      "learning_rate": 2.7690409004724883e-06,
+      "loss": 1.1722,
+      "step": 2274
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.611034623590975,
+      "learning_rate": 2.7615207548349164e-06,
+      "loss": 0.9969,
+      "step": 2275
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.40757777254612887,
+      "learning_rate": 2.7540091984726625e-06,
+      "loss": 1.0388,
+      "step": 2276
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5959757588334864,
+      "learning_rate": 2.7465062402990417e-06,
+      "loss": 1.0442,
+      "step": 2277
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5957219870237999,
+      "learning_rate": 2.739011889217166e-06,
+      "loss": 1.1091,
+      "step": 2278
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4258081487929069,
+      "learning_rate": 2.7315261541199266e-06,
+      "loss": 0.9218,
+      "step": 2279
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5610676280553291,
+      "learning_rate": 2.724049043889997e-06,
+      "loss": 1.0055,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6479963615501372,
+      "learning_rate": 2.71658056739982e-06,
+      "loss": 0.9516,
+      "step": 2281
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5406121666737639,
+      "learning_rate": 2.7091207335115822e-06,
+      "loss": 1.1872,
+      "step": 2282
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6806849909637263,
+      "learning_rate": 2.701669551077226e-06,
+      "loss": 1.0111,
+      "step": 2283
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5862329640647629,
+      "learning_rate": 2.69422702893842e-06,
+      "loss": 1.1911,
+      "step": 2284
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5776338126240509,
+      "learning_rate": 2.6867931759265598e-06,
+      "loss": 1.1127,
+      "step": 2285
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9547249298342974,
+      "learning_rate": 2.6793680008627578e-06,
+      "loss": 1.0798,
+      "step": 2286
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.427089169709061,
+      "learning_rate": 2.6719515125578198e-06,
+      "loss": 0.659,
+      "step": 2287
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3773891209850748,
+      "learning_rate": 2.6645437198122503e-06,
+      "loss": 1.0684,
+      "step": 2288
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.613347235578605,
+      "learning_rate": 2.6571446314162385e-06,
+      "loss": 1.3306,
+      "step": 2289
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.696089650663291,
+      "learning_rate": 2.649754256149635e-06,
+      "loss": 1.0388,
+      "step": 2290
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5583714062587771,
+      "learning_rate": 2.6423726027819595e-06,
+      "loss": 1.1532,
+      "step": 2291
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5628365213004118,
+      "learning_rate": 2.6349996800723822e-06,
+      "loss": 0.9037,
+      "step": 2292
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5735638571236613,
+      "learning_rate": 2.627635496769705e-06,
+      "loss": 1.1583,
+      "step": 2293
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5083089945789436,
+      "learning_rate": 2.620280061612368e-06,
+      "loss": 1.0753,
+      "step": 2294
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.449167219588012,
+      "learning_rate": 2.612933383328432e-06,
+      "loss": 0.697,
+      "step": 2295
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5804310008144933,
+      "learning_rate": 2.605595470635556e-06,
+      "loss": 1.1818,
+      "step": 2296
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6266340610412856,
+      "learning_rate": 2.598266332241007e-06,
+      "loss": 1.039,
+      "step": 2297
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7204832016837601,
+      "learning_rate": 2.5909459768416424e-06,
+      "loss": 1.2054,
+      "step": 2298
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.595260809438306,
+      "learning_rate": 2.5836344131238844e-06,
+      "loss": 0.967,
+      "step": 2299
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.366022701715051,
+      "learning_rate": 2.576331649763736e-06,
+      "loss": 1.1389,
+      "step": 2300
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5901295554360567,
+      "learning_rate": 2.569037695426756e-06,
+      "loss": 1.0281,
+      "step": 2301
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6168639473364829,
+      "learning_rate": 2.56175255876804e-06,
+      "loss": 0.9025,
+      "step": 2302
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5845556946004383,
+      "learning_rate": 2.554476248432234e-06,
+      "loss": 1.0881,
+      "step": 2303
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6898659699347192,
+      "learning_rate": 2.547208773053499e-06,
+      "loss": 0.9278,
+      "step": 2304
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4034380671903585,
+      "learning_rate": 2.5399501412555207e-06,
+      "loss": 0.9087,
+      "step": 2305
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.603627644237883,
+      "learning_rate": 2.532700361651489e-06,
+      "loss": 0.934,
+      "step": 2306
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5822655582606127,
+      "learning_rate": 2.5254594428440826e-06,
+      "loss": 0.9931,
+      "step": 2307
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5580811940067331,
+      "learning_rate": 2.5182273934254765e-06,
+      "loss": 1.1008,
+      "step": 2308
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5925631737896657,
+      "learning_rate": 2.5110042219773178e-06,
+      "loss": 0.9319,
+      "step": 2309
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6579000401850993,
+      "learning_rate": 2.503789937070712e-06,
+      "loss": 1.2523,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6153542956614589,
+      "learning_rate": 2.4965845472662287e-06,
+      "loss": 1.1248,
+      "step": 2311
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4035960022683258,
+      "learning_rate": 2.48938806111388e-06,
+      "loss": 1.1079,
+      "step": 2312
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4685581017738537,
+      "learning_rate": 2.482200487153107e-06,
+      "loss": 0.7795,
+      "step": 2313
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5877160326972619,
+      "learning_rate": 2.475021833912783e-06,
+      "loss": 1.0022,
+      "step": 2314
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5598037362077729,
+      "learning_rate": 2.467852109911195e-06,
+      "loss": 1.1099,
+      "step": 2315
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5852840574811058,
+      "learning_rate": 2.4606913236560283e-06,
+      "loss": 1.0182,
+      "step": 2316
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5246775372511999,
+      "learning_rate": 2.453539483644368e-06,
+      "loss": 1.1876,
+      "step": 2317
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.557438315760145,
+      "learning_rate": 2.4463965983626857e-06,
+      "loss": 1.0636,
+      "step": 2318
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5947893156503268,
+      "learning_rate": 2.4392626762868187e-06,
+      "loss": 0.8782,
+      "step": 2319
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6269253207235415,
+      "learning_rate": 2.4321377258819747e-06,
+      "loss": 1.0398,
+      "step": 2320
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3985200106047152,
+      "learning_rate": 2.4250217556027176e-06,
+      "loss": 0.8171,
+      "step": 2321
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6577151074546569,
+      "learning_rate": 2.4179147738929475e-06,
+      "loss": 1.2414,
+      "step": 2322
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5272938571141003,
+      "learning_rate": 2.410816789185907e-06,
+      "loss": 1.0202,
+      "step": 2323
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.651850561317289,
+      "learning_rate": 2.403727809904154e-06,
+      "loss": 0.94,
+      "step": 2324
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4242814757458441,
+      "learning_rate": 2.3966478444595674e-06,
+      "loss": 1.1553,
+      "step": 2325
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6818579422991817,
+      "learning_rate": 2.389576901253331e-06,
+      "loss": 0.8194,
+      "step": 2326
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7583934366450799,
+      "learning_rate": 2.3825149886759147e-06,
+      "loss": 1.223,
+      "step": 2327
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.49454608984550724,
+      "learning_rate": 2.375462115107079e-06,
+      "loss": 1.0116,
+      "step": 2328
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6412251060799331,
+      "learning_rate": 2.3684182889158615e-06,
+      "loss": 1.2673,
+      "step": 2329
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5932746537292899,
+      "learning_rate": 2.3613835184605527e-06,
+      "loss": 0.9326,
+      "step": 2330
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6224210646982978,
+      "learning_rate": 2.354357812088707e-06,
+      "loss": 0.9581,
+      "step": 2331
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3955783671635789,
+      "learning_rate": 2.3473411781371245e-06,
+      "loss": 0.7545,
+      "step": 2332
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5640524127478806,
+      "learning_rate": 2.3403336249318297e-06,
+      "loss": 1.154,
+      "step": 2333
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6492039671717227,
+      "learning_rate": 2.333335160788082e-06,
+      "loss": 1.2595,
+      "step": 2334
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6051480243282906,
+      "learning_rate": 2.326345794010352e-06,
+      "loss": 1.0329,
+      "step": 2335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5898341816520989,
+      "learning_rate": 2.3193655328923116e-06,
+      "loss": 0.9765,
+      "step": 2336
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3745887908364113,
+      "learning_rate": 2.3123943857168316e-06,
+      "loss": 1.1227,
+      "step": 2337
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.56865432800337,
+      "learning_rate": 2.3054323607559736e-06,
+      "loss": 0.9593,
+      "step": 2338
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6174234481371316,
+      "learning_rate": 2.29847946627096e-06,
+      "loss": 1.2132,
+      "step": 2339
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3823271388219208,
+      "learning_rate": 2.2915357105121938e-06,
+      "loss": 0.7999,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6590989722706944,
+      "learning_rate": 2.284601101719224e-06,
+      "loss": 1.1232,
+      "step": 2341
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.692128629325722,
+      "learning_rate": 2.2776756481207483e-06,
+      "loss": 1.1081,
+      "step": 2342
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5384900259009543,
+      "learning_rate": 2.2707593579346075e-06,
+      "loss": 1.0617,
+      "step": 2343
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6059883766149475,
+      "learning_rate": 2.2638522393677564e-06,
+      "loss": 1.245,
+      "step": 2344
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6368763818834191,
+      "learning_rate": 2.2569543006162753e-06,
+      "loss": 0.9787,
+      "step": 2345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7637037396658732,
+      "learning_rate": 2.2500655498653544e-06,
+      "loss": 1.2489,
+      "step": 2346
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6135998691745124,
+      "learning_rate": 2.24318599528927e-06,
+      "loss": 0.909,
+      "step": 2347
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6135998186533691,
+      "learning_rate": 2.2363156450513946e-06,
+      "loss": 1.2269,
+      "step": 2348
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4345705187687559,
+      "learning_rate": 2.229454507304181e-06,
+      "loss": 0.7962,
+      "step": 2349
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.38247656432113175,
+      "learning_rate": 2.222602590189139e-06,
+      "loss": 1.0384,
+      "step": 2350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.606503686953892,
+      "learning_rate": 2.2157599018368492e-06,
+      "loss": 1.1659,
+      "step": 2351
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5045561608041567,
+      "learning_rate": 2.2089264503669374e-06,
+      "loss": 1.1027,
+      "step": 2352
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6144524454192053,
+      "learning_rate": 2.2021022438880614e-06,
+      "loss": 1.0655,
+      "step": 2353
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5576291431920251,
+      "learning_rate": 2.1952872904979184e-06,
+      "loss": 1.0658,
+      "step": 2354
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6298754066074913,
+      "learning_rate": 2.1884815982832265e-06,
+      "loss": 0.9038,
+      "step": 2355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6327649140805596,
+      "learning_rate": 2.1816851753197023e-06,
+      "loss": 1.3259,
+      "step": 2356
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5516383630940972,
+      "learning_rate": 2.1748980296720767e-06,
+      "loss": 0.7439,
+      "step": 2357
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4035673260261629,
+      "learning_rate": 2.1681201693940667e-06,
+      "loss": 0.8294,
+      "step": 2358
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5869890839493557,
+      "learning_rate": 2.1613516025283666e-06,
+      "loss": 0.9866,
+      "step": 2359
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.597291042671418,
+      "learning_rate": 2.1545923371066534e-06,
+      "loss": 1.2047,
+      "step": 2360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.34312741160953,
+      "learning_rate": 2.1478423811495554e-06,
+      "loss": 1.1044,
+      "step": 2361
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6322599533848013,
+      "learning_rate": 2.1411017426666627e-06,
+      "loss": 0.8744,
+      "step": 2362
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.527372558788041,
+      "learning_rate": 2.134370429656507e-06,
+      "loss": 1.2137,
+      "step": 2363
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9899452140476295,
+      "learning_rate": 2.1276484501065508e-06,
+      "loss": 0.9414,
+      "step": 2364
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6102906289073282,
+      "learning_rate": 2.1209358119931845e-06,
+      "loss": 1.1326,
+      "step": 2365
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4077015322965913,
+      "learning_rate": 2.114232523281717e-06,
+      "loss": 0.8004,
+      "step": 2366
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6022821118793912,
+      "learning_rate": 2.1075385919263547e-06,
+      "loss": 1.2233,
+      "step": 2367
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6065790719217187,
+      "learning_rate": 2.100854025870207e-06,
+      "loss": 0.8913,
+      "step": 2368
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.64100162448229,
+      "learning_rate": 2.0941788330452718e-06,
+      "loss": 0.9702,
+      "step": 2369
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6878389747164028,
+      "learning_rate": 2.087513021372415e-06,
+      "loss": 1.1399,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5788088900840865,
+      "learning_rate": 2.0808565987613794e-06,
+      "loss": 0.9952,
+      "step": 2371
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6855709031607685,
+      "learning_rate": 2.0742095731107693e-06,
+      "loss": 1.2124,
+      "step": 2372
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.35669937868280194,
+      "learning_rate": 2.067571952308025e-06,
+      "loss": 1.0771,
+      "step": 2373
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6671855433505333,
+      "learning_rate": 2.06094374422944e-06,
+      "loss": 1.2894,
+      "step": 2374
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.38120170351383376,
+      "learning_rate": 2.0543249567401356e-06,
+      "loss": 0.8882,
+      "step": 2375
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6299945825423293,
+      "learning_rate": 2.0477155976940466e-06,
+      "loss": 0.8597,
+      "step": 2376
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5938373307379275,
+      "learning_rate": 2.041115674933931e-06,
+      "loss": 1.1371,
+      "step": 2377
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.642732122678309,
+      "learning_rate": 2.0345251962913447e-06,
+      "loss": 1.0983,
+      "step": 2378
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6507803478561082,
+      "learning_rate": 2.027944169586633e-06,
+      "loss": 1.1754,
+      "step": 2379
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5427451251607734,
+      "learning_rate": 2.0213726026289336e-06,
+      "loss": 0.8728,
+      "step": 2380
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5152210931852016,
+      "learning_rate": 2.0148105032161512e-06,
+      "loss": 0.992,
+      "step": 2381
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5801403138681049,
+      "learning_rate": 2.0082578791349604e-06,
+      "loss": 1.154,
+      "step": 2382
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6512854472131884,
+      "learning_rate": 2.001714738160796e-06,
+      "loss": 0.9951,
+      "step": 2383
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.44518345427986983,
+      "learning_rate": 1.9951810880578303e-06,
+      "loss": 0.8549,
+      "step": 2384
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.658297780011658,
+      "learning_rate": 1.988656936578981e-06,
+      "loss": 0.9341,
+      "step": 2385
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.33144081034020084,
+      "learning_rate": 1.982142291465896e-06,
+      "loss": 1.1734,
+      "step": 2386
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6333702495399343,
+      "learning_rate": 1.9756371604489333e-06,
+      "loss": 1.0632,
+      "step": 2387
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5129280604737195,
+      "learning_rate": 1.96914155124717e-06,
+      "loss": 1.0415,
+      "step": 2388
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6361787628470469,
+      "learning_rate": 1.9626554715683856e-06,
+      "loss": 1.1538,
+      "step": 2389
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6334625741798205,
+      "learning_rate": 1.956178929109043e-06,
+      "loss": 1.0151,
+      "step": 2390
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6642806002569595,
+      "learning_rate": 1.9497119315542945e-06,
+      "loss": 1.1322,
+      "step": 2391
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5747008487336475,
+      "learning_rate": 1.943254486577969e-06,
+      "loss": 1.0074,
+      "step": 2392
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6870663822405759,
+      "learning_rate": 1.9368066018425504e-06,
+      "loss": 1.0174,
+      "step": 2393
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.43909262354626044,
+      "learning_rate": 1.9303682849991867e-06,
+      "loss": 0.7756,
+      "step": 2394
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5239876126833043,
+      "learning_rate": 1.9239395436876716e-06,
+      "loss": 0.8369,
+      "step": 2395
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7388920687643276,
+      "learning_rate": 1.9175203855364323e-06,
+      "loss": 1.2535,
+      "step": 2396
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5567120853833185,
+      "learning_rate": 1.911110818162526e-06,
+      "loss": 1.0138,
+      "step": 2397
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.537657714606848,
+      "learning_rate": 1.904710849171636e-06,
+      "loss": 1.1811,
+      "step": 2398
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3576500774293892,
+      "learning_rate": 1.8983204861580418e-06,
+      "loss": 1.1555,
+      "step": 2399
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7424983140274568,
+      "learning_rate": 1.891939736704641e-06,
+      "loss": 0.8947,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.44104379485382805,
+      "learning_rate": 1.8855686083829083e-06,
+      "loss": 0.8066,
+      "step": 2401
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6391769657456815,
+      "learning_rate": 1.8792071087529118e-06,
+      "loss": 1.0767,
+      "step": 2402
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5841057888706607,
+      "learning_rate": 1.8728552453632952e-06,
+      "loss": 1.1305,
+      "step": 2403
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4669530268894037,
+      "learning_rate": 1.8665130257512566e-06,
+      "loss": 0.8368,
+      "step": 2404
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6250538514246601,
+      "learning_rate": 1.8601804574425609e-06,
+      "loss": 1.1899,
+      "step": 2405
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6743908518676187,
+      "learning_rate": 1.8538575479515208e-06,
+      "loss": 1.0756,
+      "step": 2406
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.591897137549232,
+      "learning_rate": 1.8475443047809782e-06,
+      "loss": 0.9898,
+      "step": 2407
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6166831898537256,
+      "learning_rate": 1.8412407354223149e-06,
+      "loss": 1.069,
+      "step": 2408
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5184589892769288,
+      "learning_rate": 1.8349468473554311e-06,
+      "loss": 1.2714,
+      "step": 2409
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6116817108414461,
+      "learning_rate": 1.8286626480487336e-06,
+      "loss": 1.1801,
+      "step": 2410
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4175459546440712,
+      "learning_rate": 1.8223881449591385e-06,
+      "loss": 0.7354,
+      "step": 2411
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6706903720654206,
+      "learning_rate": 1.8161233455320571e-06,
+      "loss": 1.0489,
+      "step": 2412
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34259212096906677,
+      "learning_rate": 1.8098682572013782e-06,
+      "loss": 1.1391,
+      "step": 2413
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6751648321263958,
+      "learning_rate": 1.8036228873894745e-06,
+      "loss": 0.8541,
+      "step": 2414
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7103601210541732,
+      "learning_rate": 1.7973872435071882e-06,
+      "loss": 1.2206,
+      "step": 2415
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5938805455451017,
+      "learning_rate": 1.7911613329538102e-06,
+      "loss": 0.9046,
+      "step": 2416
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6310647753236,
+      "learning_rate": 1.7849451631170922e-06,
+      "loss": 0.8299,
+      "step": 2417
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6261860436239886,
+      "learning_rate": 1.7787387413732248e-06,
+      "loss": 1.1982,
+      "step": 2418
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5730788808802759,
+      "learning_rate": 1.7725420750868273e-06,
+      "loss": 1.1078,
+      "step": 2419
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.41773789176162274,
+      "learning_rate": 1.7663551716109495e-06,
+      "loss": 0.8201,
+      "step": 2420
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6534932181164625,
+      "learning_rate": 1.760178038287048e-06,
+      "loss": 0.9542,
+      "step": 2421
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4884448896830803,
+      "learning_rate": 1.7540106824449943e-06,
+      "loss": 1.3426,
+      "step": 2422
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6549322542767119,
+      "learning_rate": 1.7478531114030573e-06,
+      "loss": 0.9528,
+      "step": 2423
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.35602169430230257,
+      "learning_rate": 1.7417053324678856e-06,
+      "loss": 1.0992,
+      "step": 2424
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5633658091515056,
+      "learning_rate": 1.7355673529345184e-06,
+      "loss": 1.0563,
+      "step": 2425
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6067953667854861,
+      "learning_rate": 1.7294391800863664e-06,
+      "loss": 0.9647,
+      "step": 2426
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6170048108558881,
+      "learning_rate": 1.7233208211951946e-06,
+      "loss": 1.1087,
+      "step": 2427
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.43926505913037356,
+      "learning_rate": 1.7172122835211337e-06,
+      "loss": 0.6708,
+      "step": 2428
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5551149221157199,
+      "learning_rate": 1.711113574312655e-06,
+      "loss": 1.1105,
+      "step": 2429
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6169476054414381,
+      "learning_rate": 1.7050247008065634e-06,
+      "loss": 0.9235,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6594242321303965,
+      "learning_rate": 1.6989456702280005e-06,
+      "loss": 1.0583,
+      "step": 2431
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6233820856281812,
+      "learning_rate": 1.692876489790426e-06,
+      "loss": 1.4309,
+      "step": 2432
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.48010732136680145,
+      "learning_rate": 1.6868171666956045e-06,
+      "loss": 1.0601,
+      "step": 2433
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5327201490504982,
+      "learning_rate": 1.6807677081336116e-06,
+      "loss": 1.1004,
+      "step": 2434
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6895271260329561,
+      "learning_rate": 1.6747281212828193e-06,
+      "loss": 0.9167,
+      "step": 2435
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.33128837065939165,
+      "learning_rate": 1.6686984133098737e-06,
+      "loss": 1.069,
+      "step": 2436
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5664383223025716,
+      "learning_rate": 1.6626785913697118e-06,
+      "loss": 0.7755,
+      "step": 2437
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.60468783978114,
+      "learning_rate": 1.6566686626055307e-06,
+      "loss": 1.009,
+      "step": 2438
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6055747525562829,
+      "learning_rate": 1.650668634148792e-06,
+      "loss": 1.0652,
+      "step": 2439
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5724638042571207,
+      "learning_rate": 1.6446785131192123e-06,
+      "loss": 0.9456,
+      "step": 2440
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6900600044090341,
+      "learning_rate": 1.638698306624743e-06,
+      "loss": 1.1938,
+      "step": 2441
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5663475487268133,
+      "learning_rate": 1.6327280217615793e-06,
+      "loss": 1.1404,
+      "step": 2442
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5679579922891559,
+      "learning_rate": 1.6267676656141429e-06,
+      "loss": 0.9041,
+      "step": 2443
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6621713743035647,
+      "learning_rate": 1.6208172452550664e-06,
+      "loss": 1.2297,
+      "step": 2444
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4981586194762279,
+      "learning_rate": 1.6148767677451994e-06,
+      "loss": 1.1248,
+      "step": 2445
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.460819473119904,
+      "learning_rate": 1.6089462401335932e-06,
+      "loss": 0.8141,
+      "step": 2446
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5513179251261788,
+      "learning_rate": 1.6030256694574876e-06,
+      "loss": 0.9083,
+      "step": 2447
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3694159728230469,
+      "learning_rate": 1.5971150627423093e-06,
+      "loss": 1.1472,
+      "step": 2448
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8561255380364157,
+      "learning_rate": 1.5912144270016672e-06,
+      "loss": 0.8721,
+      "step": 2449
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6544750017977187,
+      "learning_rate": 1.5853237692373291e-06,
+      "loss": 1.0088,
+      "step": 2450
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5685144271169036,
+      "learning_rate": 1.5794430964392293e-06,
+      "loss": 1.1966,
+      "step": 2451
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.634395638244246,
+      "learning_rate": 1.5735724155854538e-06,
+      "loss": 0.8818,
+      "step": 2452
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5532775572224042,
+      "learning_rate": 1.5677117336422265e-06,
+      "loss": 1.2264,
+      "step": 2453
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.44790325429134437,
+      "learning_rate": 1.5618610575639115e-06,
+      "loss": 0.6376,
+      "step": 2454
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5681274255931023,
+      "learning_rate": 1.556020394293003e-06,
+      "loss": 0.9158,
+      "step": 2455
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.47050623447795137,
+      "learning_rate": 1.5501897507601016e-06,
+      "loss": 1.1955,
+      "step": 2456
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5951728447130993,
+      "learning_rate": 1.5443691338839328e-06,
+      "loss": 1.095,
+      "step": 2457
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7005563614479381,
+      "learning_rate": 1.5385585505713129e-06,
+      "loss": 1.2032,
+      "step": 2458
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5955597567904634,
+      "learning_rate": 1.5327580077171589e-06,
+      "loss": 0.9564,
+      "step": 2459
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6490748292825601,
+      "learning_rate": 1.5269675122044746e-06,
+      "loss": 1.1695,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6276166425360743,
+      "learning_rate": 1.521187070904333e-06,
+      "loss": 0.9893,
+      "step": 2461
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.37548369453677793,
+      "learning_rate": 1.515416690675885e-06,
+      "loss": 1.1019,
+      "step": 2462
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5753379301113504,
+      "learning_rate": 1.509656378366343e-06,
+      "loss": 0.8718,
+      "step": 2463
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.41611489609210667,
+      "learning_rate": 1.5039061408109656e-06,
+      "loss": 0.7634,
+      "step": 2464
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6026411336270459,
+      "learning_rate": 1.4981659848330622e-06,
+      "loss": 1.2863,
+      "step": 2465
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6186145437996736,
+      "learning_rate": 1.4924359172439794e-06,
+      "loss": 1.0117,
+      "step": 2466
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6667602957957186,
+      "learning_rate": 1.486715944843089e-06,
+      "loss": 1.1176,
+      "step": 2467
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5244436576272918,
+      "learning_rate": 1.4810060744177868e-06,
+      "loss": 1.126,
+      "step": 2468
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.659283312597,
+      "learning_rate": 1.475306312743483e-06,
+      "loss": 1.0035,
+      "step": 2469
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5730890677395521,
+      "learning_rate": 1.4696166665835853e-06,
+      "loss": 1.1737,
+      "step": 2470
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5988995676542108,
+      "learning_rate": 1.4639371426895066e-06,
+      "loss": 1.1322,
+      "step": 2471
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6079886283026333,
+      "learning_rate": 1.4582677478006468e-06,
+      "loss": 1.0904,
+      "step": 2472
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2913468698094277,
+      "learning_rate": 1.4526084886443813e-06,
+      "loss": 0.9217,
+      "step": 2473
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6149802008549778,
+      "learning_rate": 1.446959371936062e-06,
+      "loss": 1.0748,
+      "step": 2474
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6482940775258998,
+      "learning_rate": 1.4413204043790108e-06,
+      "loss": 1.1808,
+      "step": 2475
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6402305210274707,
+      "learning_rate": 1.4356915926644942e-06,
+      "loss": 1.1034,
+      "step": 2476
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5910931884812808,
+      "learning_rate": 1.4300729434717398e-06,
+      "loss": 1.1769,
+      "step": 2477
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5948875431795184,
+      "learning_rate": 1.4244644634679084e-06,
+      "loss": 0.9745,
+      "step": 2478
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6223796635512776,
+      "learning_rate": 1.4188661593080966e-06,
+      "loss": 1.1761,
+      "step": 2479
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5335514992674516,
+      "learning_rate": 1.4132780376353295e-06,
+      "loss": 1.1118,
+      "step": 2480
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4435075646213538,
+      "learning_rate": 1.4077001050805418e-06,
+      "loss": 0.7179,
+      "step": 2481
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6461868584997412,
+      "learning_rate": 1.4021323682625842e-06,
+      "loss": 1.1515,
+      "step": 2482
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6355596004546757,
+      "learning_rate": 1.39657483378821e-06,
+      "loss": 1.1111,
+      "step": 2483
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6599098291462944,
+      "learning_rate": 1.3910275082520575e-06,
+      "loss": 1.1077,
+      "step": 2484
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6209743448655691,
+      "learning_rate": 1.3854903982366585e-06,
+      "loss": 0.9342,
+      "step": 2485
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.31525002124832074,
+      "learning_rate": 1.3799635103124255e-06,
+      "loss": 1.1372,
+      "step": 2486
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5862711059153249,
+      "learning_rate": 1.3744468510376309e-06,
+      "loss": 1.0569,
+      "step": 2487
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6067830971287618,
+      "learning_rate": 1.3689404269584172e-06,
+      "loss": 0.936,
+      "step": 2488
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5725664729249996,
+      "learning_rate": 1.3634442446087837e-06,
+      "loss": 1.0439,
+      "step": 2489
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4136402351655331,
+      "learning_rate": 1.3579583105105677e-06,
+      "loss": 0.698,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49879762557125973,
+      "learning_rate": 1.3524826311734552e-06,
+      "loss": 1.1273,
+      "step": 2491
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5912672673339251,
+      "learning_rate": 1.3470172130949599e-06,
+      "loss": 0.9645,
+      "step": 2492
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6349635185117632,
+      "learning_rate": 1.3415620627604154e-06,
+      "loss": 1.0243,
+      "step": 2493
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5484655255573552,
+      "learning_rate": 1.336117186642979e-06,
+      "loss": 1.0314,
+      "step": 2494
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6868595935560582,
+      "learning_rate": 1.3306825912036137e-06,
+      "loss": 1.0171,
+      "step": 2495
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.600489389719359,
+      "learning_rate": 1.3252582828910775e-06,
+      "loss": 1.2421,
+      "step": 2496
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3526787672739039,
+      "learning_rate": 1.3198442681419321e-06,
+      "loss": 1.1445,
+      "step": 2497
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6127535794835948,
+      "learning_rate": 1.3144405533805138e-06,
+      "loss": 1.2176,
+      "step": 2498
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.45012633183741246,
+      "learning_rate": 1.3090471450189446e-06,
+      "loss": 0.5892,
+      "step": 2499
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5533926940344494,
+      "learning_rate": 1.3036640494571162e-06,
+      "loss": 1.027,
+      "step": 2500
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5954607813530202,
+      "learning_rate": 1.298291273082678e-06,
+      "loss": 1.0737,
+      "step": 2501
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5891511006442205,
+      "learning_rate": 1.292928822271038e-06,
+      "loss": 1.0163,
+      "step": 2502
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4671643914219495,
+      "learning_rate": 1.2875767033853547e-06,
+      "loss": 1.2225,
+      "step": 2503
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6635434349754313,
+      "learning_rate": 1.2822349227765185e-06,
+      "loss": 1.0087,
+      "step": 2504
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6311018507987487,
+      "learning_rate": 1.2769034867831588e-06,
+      "loss": 0.8405,
+      "step": 2505
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5694007771760176,
+      "learning_rate": 1.2715824017316302e-06,
+      "loss": 0.9881,
+      "step": 2506
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.44204461429170816,
+      "learning_rate": 1.2662716739360003e-06,
+      "loss": 0.8324,
+      "step": 2507
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6129348297772778,
+      "learning_rate": 1.260971309698048e-06,
+      "loss": 1.2651,
+      "step": 2508
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5626539463654208,
+      "learning_rate": 1.2556813153072612e-06,
+      "loss": 1.0548,
+      "step": 2509
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6483167902038263,
+      "learning_rate": 1.2504016970408105e-06,
+      "loss": 1.1723,
+      "step": 2510
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.35018038582370237,
+      "learning_rate": 1.245132461163564e-06,
+      "loss": 1.0763,
+      "step": 2511
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5928523179426151,
+      "learning_rate": 1.2398736139280687e-06,
+      "loss": 0.8483,
+      "step": 2512
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5466100951833754,
+      "learning_rate": 1.2346251615745375e-06,
+      "loss": 1.2154,
+      "step": 2513
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5427662276007422,
+      "learning_rate": 1.2293871103308596e-06,
+      "loss": 0.9625,
+      "step": 2514
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6364091142476384,
+      "learning_rate": 1.22415946641257e-06,
+      "loss": 1.2928,
+      "step": 2515
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5808470275780007,
+      "learning_rate": 1.218942236022862e-06,
+      "loss": 0.9667,
+      "step": 2516
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6689872468919316,
+      "learning_rate": 1.2137354253525734e-06,
+      "loss": 1.1488,
+      "step": 2517
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3805753438602651,
+      "learning_rate": 1.208539040580171e-06,
+      "loss": 0.6171,
+      "step": 2518
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6802220751325196,
+      "learning_rate": 1.2033530878717548e-06,
+      "loss": 1.1861,
+      "step": 2519
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6401280258831019,
+      "learning_rate": 1.1981775733810474e-06,
+      "loss": 1.2418,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6400158358440874,
+      "learning_rate": 1.193012503249379e-06,
+      "loss": 1.0137,
+      "step": 2521
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6168902781766576,
+      "learning_rate": 1.187857883605691e-06,
+      "loss": 1.1641,
+      "step": 2522
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.337194059633294,
+      "learning_rate": 1.1827137205665273e-06,
+      "loss": 1.0497,
+      "step": 2523
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5456666024510607,
+      "learning_rate": 1.1775800202360133e-06,
+      "loss": 1.011,
+      "step": 2524
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6870262963190622,
+      "learning_rate": 1.1724567887058702e-06,
+      "loss": 1.1923,
+      "step": 2525
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3902021061575714,
+      "learning_rate": 1.1673440320553941e-06,
+      "loss": 0.7772,
+      "step": 2526
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6598957882067263,
+      "learning_rate": 1.1622417563514443e-06,
+      "loss": 1.1976,
+      "step": 2527
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6046499399665036,
+      "learning_rate": 1.1571499676484522e-06,
+      "loss": 1.099,
+      "step": 2528
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.558714376163596,
+      "learning_rate": 1.1520686719884022e-06,
+      "loss": 1.2257,
+      "step": 2529
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5419699506356975,
+      "learning_rate": 1.1469978754008238e-06,
+      "loss": 1.0647,
+      "step": 2530
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5794227969660144,
+      "learning_rate": 1.1419375839027947e-06,
+      "loss": 0.9703,
+      "step": 2531
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6347225950533961,
+      "learning_rate": 1.1368878034989227e-06,
+      "loss": 1.1058,
+      "step": 2532
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9701491348406989,
+      "learning_rate": 1.1318485401813438e-06,
+      "loss": 0.8549,
+      "step": 2533
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6608191252268751,
+      "learning_rate": 1.126819799929716e-06,
+      "loss": 1.2997,
+      "step": 2534
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30628639831854393,
+      "learning_rate": 1.1218015887112055e-06,
+      "loss": 0.9234,
+      "step": 2535
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6942627975753114,
+      "learning_rate": 1.1167939124804894e-06,
+      "loss": 1.0664,
+      "step": 2536
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5991536178218384,
+      "learning_rate": 1.1117967771797477e-06,
+      "loss": 0.9386,
+      "step": 2537
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.47576159704696686,
+      "learning_rate": 1.1068101887386395e-06,
+      "loss": 1.0873,
+      "step": 2538
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6311452669900304,
+      "learning_rate": 1.1018341530743215e-06,
+      "loss": 1.0431,
+      "step": 2539
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5208898253641899,
+      "learning_rate": 1.0968686760914248e-06,
+      "loss": 1.0731,
+      "step": 2540
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6728349005103083,
+      "learning_rate": 1.0919137636820442e-06,
+      "loss": 1.1571,
+      "step": 2541
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5714058629374836,
+      "learning_rate": 1.0869694217257498e-06,
+      "loss": 1.1121,
+      "step": 2542
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4197629089923067,
+      "learning_rate": 1.0820356560895629e-06,
+      "loss": 0.612,
+      "step": 2543
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5518504542471828,
+      "learning_rate": 1.0771124726279514e-06,
+      "loss": 1.1345,
+      "step": 2544
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5948181854982192,
+      "learning_rate": 1.0721998771828312e-06,
+      "loss": 0.989,
+      "step": 2545
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6384751208434685,
+      "learning_rate": 1.0672978755835572e-06,
+      "loss": 1.164,
+      "step": 2546
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.34438381946310953,
+      "learning_rate": 1.0624064736469054e-06,
+      "loss": 1.0993,
+      "step": 2547
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6896761393076286,
+      "learning_rate": 1.057525677177077e-06,
+      "loss": 0.8911,
+      "step": 2548
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5482877987618444,
+      "learning_rate": 1.052655491965695e-06,
+      "loss": 1.2423,
+      "step": 2549
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.656556579351074,
+      "learning_rate": 1.04779592379178e-06,
+      "loss": 0.9801,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.514659112998211,
+      "learning_rate": 1.0429469784217626e-06,
+      "loss": 1.0939,
+      "step": 2551
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.44646838820226126,
+      "learning_rate": 1.0381086616094671e-06,
+      "loss": 0.7841,
+      "step": 2552
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5579006429285338,
+      "learning_rate": 1.0332809790961019e-06,
+      "loss": 1.1843,
+      "step": 2553
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6069479756223043,
+      "learning_rate": 1.02846393661026e-06,
+      "loss": 0.9334,
+      "step": 2554
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5395620193516197,
+      "learning_rate": 1.023657539867906e-06,
+      "loss": 0.9421,
+      "step": 2555
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.668218088959416,
+      "learning_rate": 1.0188617945723755e-06,
+      "loss": 1.21,
+      "step": 2556
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7092546866072872,
+      "learning_rate": 1.0140767064143653e-06,
+      "loss": 0.9549,
+      "step": 2557
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7783969407837017,
+      "learning_rate": 1.00930228107192e-06,
+      "loss": 1.1489,
+      "step": 2558
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3229011042064861,
+      "learning_rate": 1.0045385242104378e-06,
+      "loss": 1.0345,
+      "step": 2559
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6531579039586352,
+      "learning_rate": 9.99785441482659e-07,
+      "loss": 1.2827,
+      "step": 2560
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.36701685395999706,
+      "learning_rate": 9.950430385286492e-07,
+      "loss": 0.8724,
+      "step": 2561
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6784878711147015,
+      "learning_rate": 9.903113209758098e-07,
+      "loss": 0.8583,
+      "step": 2562
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.563533691039236,
+      "learning_rate": 9.855902944388595e-07,
+      "loss": 1.0985,
+      "step": 2563
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.632440093385196,
+      "learning_rate": 9.808799645198285e-07,
+      "loss": 1.1424,
+      "step": 2564
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6579059759398108,
+      "learning_rate": 9.76180336808058e-07,
+      "loss": 0.9846,
+      "step": 2565
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6104348329429727,
+      "learning_rate": 9.714914168801915e-07,
+      "loss": 1.0217,
+      "step": 2566
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5583978712923524,
+      "learning_rate": 9.66813210300157e-07,
+      "loss": 1.0203,
+      "step": 2567
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5515635243572715,
+      "learning_rate": 9.62145722619182e-07,
+      "loss": 1.079,
+      "step": 2568
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6590848553069691,
+      "learning_rate": 9.574889593757686e-07,
+      "loss": 0.959,
+      "step": 2569
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4178507966169768,
+      "learning_rate": 9.528429260956895e-07,
+      "loss": 0.8396,
+      "step": 2570
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6751102476903273,
+      "learning_rate": 9.482076282919927e-07,
+      "loss": 0.9825,
+      "step": 2571
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.316618440057052,
+      "learning_rate": 9.435830714649851e-07,
+      "loss": 1.1865,
+      "step": 2572
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5966538324511182,
+      "learning_rate": 9.389692611022239e-07,
+      "loss": 0.8909,
+      "step": 2573
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5935389799063877,
+      "learning_rate": 9.343662026785217e-07,
+      "loss": 1.0538,
+      "step": 2574
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6647576265120833,
+      "learning_rate": 9.297739016559226e-07,
+      "loss": 1.1645,
+      "step": 2575
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.587345451284134,
+      "learning_rate": 9.251923634837156e-07,
+      "loss": 0.9922,
+      "step": 2576
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6398184523837471,
+      "learning_rate": 9.206215935984164e-07,
+      "loss": 1.1328,
+      "step": 2577
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5604898444775521,
+      "learning_rate": 9.160615974237563e-07,
+      "loss": 1.0231,
+      "step": 2578
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.609849577416939,
+      "learning_rate": 9.115123803706904e-07,
+      "loss": 1.2143,
+      "step": 2579
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3958143135290962,
+      "learning_rate": 9.069739478373807e-07,
+      "loss": 0.6158,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5631999733862879,
+      "learning_rate": 9.024463052091869e-07,
+      "loss": 0.879,
+      "step": 2581
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5387302333953191,
+      "learning_rate": 8.979294578586739e-07,
+      "loss": 1.1522,
+      "step": 2582
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6335232712225267,
+      "learning_rate": 8.934234111455919e-07,
+      "loss": 1.0675,
+      "step": 2583
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5338932817447397,
+      "learning_rate": 8.889281704168739e-07,
+      "loss": 1.1753,
+      "step": 2584
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3238725702341666,
+      "learning_rate": 8.844437410066331e-07,
+      "loss": 1.1582,
+      "step": 2585
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6292848308321076,
+      "learning_rate": 8.79970128236154e-07,
+      "loss": 1.0599,
+      "step": 2586
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.41490562846402657,
+      "learning_rate": 8.755073374138823e-07,
+      "loss": 0.6473,
+      "step": 2587
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6208432044325296,
+      "learning_rate": 8.710553738354254e-07,
+      "loss": 1.0553,
+      "step": 2588
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.576948129174551,
+      "learning_rate": 8.666142427835444e-07,
+      "loss": 1.0508,
+      "step": 2589
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5071357835062615,
+      "learning_rate": 8.621839495281414e-07,
+      "loss": 0.8724,
+      "step": 2590
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5787423900172943,
+      "learning_rate": 8.577644993262646e-07,
+      "loss": 1.2767,
+      "step": 2591
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6683495003226363,
+      "learning_rate": 8.533558974220868e-07,
+      "loss": 0.9591,
+      "step": 2592
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6019658306865548,
+      "learning_rate": 8.489581490469168e-07,
+      "loss": 1.0067,
+      "step": 2593
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6223859515137519,
+      "learning_rate": 8.445712594191824e-07,
+      "loss": 1.1014,
+      "step": 2594
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.49447710255170846,
+      "learning_rate": 8.401952337444208e-07,
+      "loss": 1.2189,
+      "step": 2595
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4402146907053026,
+      "learning_rate": 8.35830077215285e-07,
+      "loss": 0.7982,
+      "step": 2596
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5692547101841601,
+      "learning_rate": 8.314757950115294e-07,
+      "loss": 0.9595,
+      "step": 2597
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6829589052194625,
+      "learning_rate": 8.271323922999974e-07,
+      "loss": 1.2789,
+      "step": 2598
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.32752243828156913,
+      "learning_rate": 8.22799874234631e-07,
+      "loss": 1.0772,
+      "step": 2599
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5467955203594544,
+      "learning_rate": 8.184782459564556e-07,
+      "loss": 0.8854,
+      "step": 2600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6569382385265886,
+      "learning_rate": 8.1416751259357e-07,
+      "loss": 1.1838,
+      "step": 2601
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.573718308521015,
+      "learning_rate": 8.098676792611471e-07,
+      "loss": 0.8619,
+      "step": 2602
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6664694415241098,
+      "learning_rate": 8.055787510614288e-07,
+      "loss": 1.0548,
+      "step": 2603
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5963769442404825,
+      "learning_rate": 8.013007330837109e-07,
+      "loss": 1.0178,
+      "step": 2604
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5817055726659197,
+      "learning_rate": 7.970336304043469e-07,
+      "loss": 1.1084,
+      "step": 2605
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4154815332909861,
+      "learning_rate": 7.927774480867411e-07,
+      "loss": 0.7906,
+      "step": 2606
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5809804062078907,
+      "learning_rate": 7.885321911813304e-07,
+      "loss": 1.1389,
+      "step": 2607
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5309226066281271,
+      "learning_rate": 7.84297864725595e-07,
+      "loss": 1.2763,
+      "step": 2608
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6515347565876803,
+      "learning_rate": 7.800744737440446e-07,
+      "loss": 0.9206,
+      "step": 2609
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3296083322315197,
+      "learning_rate": 7.758620232482083e-07,
+      "loss": 1.1187,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5429235262524593,
+      "learning_rate": 7.716605182366377e-07,
+      "loss": 1.0384,
+      "step": 2611
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6138104261472467,
+      "learning_rate": 7.674699636948913e-07,
+      "loss": 0.9037,
+      "step": 2612
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6327437495468021,
+      "learning_rate": 7.632903645955392e-07,
+      "loss": 1.063,
+      "step": 2613
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4330474393015752,
+      "learning_rate": 7.591217258981476e-07,
+      "loss": 0.6955,
+      "step": 2614
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5494795520560288,
+      "learning_rate": 7.549640525492796e-07,
+      "loss": 1.1649,
+      "step": 2615
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6973319430859883,
+      "learning_rate": 7.508173494824844e-07,
+      "loss": 0.9314,
+      "step": 2616
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6643269470455986,
+      "learning_rate": 7.46681621618297e-07,
+      "loss": 1.2635,
+      "step": 2617
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6142631592360346,
+      "learning_rate": 7.425568738642242e-07,
+      "loss": 1.2044,
+      "step": 2618
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.459029486689326,
+      "learning_rate": 7.384431111147483e-07,
+      "loss": 1.0626,
+      "step": 2619
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5407153957566548,
+      "learning_rate": 7.343403382513115e-07,
+      "loss": 1.1641,
+      "step": 2620
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6979333401436747,
+      "learning_rate": 7.302485601423215e-07,
+      "loss": 0.929,
+      "step": 2621
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.36374353616630933,
+      "learning_rate": 7.261677816431367e-07,
+      "loss": 1.1223,
+      "step": 2622
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.43857937874405123,
+      "learning_rate": 7.220980075960593e-07,
+      "loss": 0.6125,
+      "step": 2623
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5341956690240642,
+      "learning_rate": 7.180392428303395e-07,
+      "loss": 1.0176,
+      "step": 2624
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7486218262824258,
+      "learning_rate": 7.139914921621638e-07,
+      "loss": 1.0687,
+      "step": 2625
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5820229403369112,
+      "learning_rate": 7.099547603946422e-07,
+      "loss": 0.8871,
+      "step": 2626
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6209277239568368,
+      "learning_rate": 7.059290523178153e-07,
+      "loss": 1.2311,
+      "step": 2627
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5771965148966812,
+      "learning_rate": 7.019143727086441e-07,
+      "loss": 1.1373,
+      "step": 2628
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5966320067032248,
+      "learning_rate": 6.979107263309969e-07,
+      "loss": 1.1272,
+      "step": 2629
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5465885443526891,
+      "learning_rate": 6.939181179356568e-07,
+      "loss": 1.0182,
+      "step": 2630
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5266146225882635,
+      "learning_rate": 6.899365522603041e-07,
+      "loss": 1.1206,
+      "step": 2631
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.41218346118584864,
+      "learning_rate": 6.859660340295171e-07,
+      "loss": 0.7254,
+      "step": 2632
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6769422603527846,
+      "learning_rate": 6.82006567954766e-07,
+      "loss": 0.9479,
+      "step": 2633
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3297169994648152,
+      "learning_rate": 6.780581587344071e-07,
+      "loss": 1.1564,
+      "step": 2634
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6538513815366065,
+      "learning_rate": 6.741208110536712e-07,
+      "loss": 0.9388,
+      "step": 2635
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.623792782542539,
+      "learning_rate": 6.701945295846701e-07,
+      "loss": 1.0822,
+      "step": 2636
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5667323218272413,
+      "learning_rate": 6.662793189863825e-07,
+      "loss": 1.1307,
+      "step": 2637
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.616474469928555,
+      "learning_rate": 6.623751839046455e-07,
+      "loss": 0.956,
+      "step": 2638
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5310550994386763,
+      "learning_rate": 6.584821289721599e-07,
+      "loss": 1.1924,
+      "step": 2639
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4214237347281163,
+      "learning_rate": 6.54600158808476e-07,
+      "loss": 0.7058,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6610037163494487,
+      "learning_rate": 6.507292780199892e-07,
+      "loss": 1.2426,
+      "step": 2641
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4726182747103725,
+      "learning_rate": 6.468694911999418e-07,
+      "loss": 1.0971,
+      "step": 2642
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.557039258758837,
+      "learning_rate": 6.430208029284035e-07,
+      "loss": 1.0626,
+      "step": 2643
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6029909688126425,
+      "learning_rate": 6.391832177722812e-07,
+      "loss": 1.1077,
+      "step": 2644
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5770294160094728,
+      "learning_rate": 6.353567402853056e-07,
+      "loss": 0.9702,
+      "step": 2645
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6668617192111583,
+      "learning_rate": 6.315413750080224e-07,
+      "loss": 1.1494,
+      "step": 2646
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33375712653839523,
+      "learning_rate": 6.277371264677956e-07,
+      "loss": 1.0779,
+      "step": 2647
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.585541360778433,
+      "learning_rate": 6.239439991787988e-07,
+      "loss": 1.0987,
+      "step": 2648
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.524265027241065,
+      "learning_rate": 6.201619976420037e-07,
+      "loss": 0.9036,
+      "step": 2649
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.39696274403289405,
+      "learning_rate": 6.163911263451861e-07,
+      "loss": 0.7289,
+      "step": 2650
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5880761966528067,
+      "learning_rate": 6.126313897629121e-07,
+      "loss": 1.2325,
+      "step": 2651
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5208963189579281,
+      "learning_rate": 6.088827923565321e-07,
+      "loss": 0.9515,
+      "step": 2652
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6795436505773136,
+      "learning_rate": 6.051453385741835e-07,
+      "loss": 1.2074,
+      "step": 2653
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5192641629069749,
+      "learning_rate": 6.014190328507785e-07,
+      "loss": 1.0253,
+      "step": 2654
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6718398670793712,
+      "learning_rate": 5.977038796080004e-07,
+      "loss": 1.019,
+      "step": 2655
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8026798238064052,
+      "learning_rate": 5.939998832542992e-07,
+      "loss": 1.2243,
+      "step": 2656
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6005057008731853,
+      "learning_rate": 5.90307048184886e-07,
+      "loss": 1.0507,
+      "step": 2657
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5770946473555505,
+      "learning_rate": 5.866253787817267e-07,
+      "loss": 1.1496,
+      "step": 2658
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28571635073143087,
+      "learning_rate": 5.82954879413542e-07,
+      "loss": 0.8962,
+      "step": 2659
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6218939378362358,
+      "learning_rate": 5.792955544357914e-07,
+      "loss": 1.2469,
+      "step": 2660
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6745373906843892,
+      "learning_rate": 5.756474081906804e-07,
+      "loss": 0.9218,
+      "step": 2661
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5826081074024696,
+      "learning_rate": 5.720104450071484e-07,
+      "loss": 1.0791,
+      "step": 2662
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5597136001207512,
+      "learning_rate": 5.683846692008621e-07,
+      "loss": 1.0714,
+      "step": 2663
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.638612133974622,
+      "learning_rate": 5.647700850742166e-07,
+      "loss": 1.0153,
+      "step": 2664
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5533622497915317,
+      "learning_rate": 5.611666969163243e-07,
+      "loss": 1.119,
+      "step": 2665
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5272502855674812,
+      "learning_rate": 5.575745090030138e-07,
+      "loss": 1.0914,
+      "step": 2666
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4175085334895554,
+      "learning_rate": 5.539935255968221e-07,
+      "loss": 0.8158,
+      "step": 2667
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5891638330416858,
+      "learning_rate": 5.504237509469934e-07,
+      "loss": 1.0047,
+      "step": 2668
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6429234810946971,
+      "learning_rate": 5.468651892894672e-07,
+      "loss": 1.0692,
+      "step": 2669
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6698686524630421,
+      "learning_rate": 5.433178448468812e-07,
+      "loss": 1.1631,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5861645624670939,
+      "learning_rate": 5.397817218285628e-07,
+      "loss": 0.9131,
+      "step": 2671
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3476914036685893,
+      "learning_rate": 5.362568244305188e-07,
+      "loss": 1.1507,
+      "step": 2672
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6214907294471493,
+      "learning_rate": 5.327431568354402e-07,
+      "loss": 1.0103,
+      "step": 2673
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5078428121849178,
+      "learning_rate": 5.29240723212695e-07,
+      "loss": 0.7842,
+      "step": 2674
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7166272906459107,
+      "learning_rate": 5.257495277183122e-07,
+      "loss": 1.2536,
+      "step": 2675
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.41557812832222685,
+      "learning_rate": 5.222695744949957e-07,
+      "loss": 0.6021,
+      "step": 2676
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5147020948707235,
+      "learning_rate": 5.18800867672099e-07,
+      "loss": 1.1757,
+      "step": 2677
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5608650429866177,
+      "learning_rate": 5.153434113656386e-07,
+      "loss": 0.9258,
+      "step": 2678
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5699652442457541,
+      "learning_rate": 5.118972096782782e-07,
+      "loss": 0.965,
+      "step": 2679
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6084226242043131,
+      "learning_rate": 5.084622666993244e-07,
+      "loss": 1.0321,
+      "step": 2680
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5984969461608697,
+      "learning_rate": 5.05038586504727e-07,
+      "loss": 1.0185,
+      "step": 2681
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5457601451125329,
+      "learning_rate": 5.016261731570715e-07,
+      "loss": 1.1358,
+      "step": 2682
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3392388056629793,
+      "learning_rate": 4.982250307055703e-07,
+      "loss": 1.1244,
+      "step": 2683
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6066345716264735,
+      "learning_rate": 4.948351631860648e-07,
+      "loss": 1.1323,
+      "step": 2684
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4727764702219753,
+      "learning_rate": 4.914565746210198e-07,
+      "loss": 0.6375,
+      "step": 2685
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.550666931692341,
+      "learning_rate": 4.880892690195083e-07,
+      "loss": 0.9671,
+      "step": 2686
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5259229283170924,
+      "learning_rate": 4.847332503772229e-07,
+      "loss": 1.0522,
+      "step": 2687
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5174676280812263,
+      "learning_rate": 4.813885226764592e-07,
+      "loss": 1.1599,
+      "step": 2688
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5225245732466497,
+      "learning_rate": 4.780550898861147e-07,
+      "loss": 1.1726,
+      "step": 2689
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6166779298444285,
+      "learning_rate": 4.7473295596168536e-07,
+      "loss": 1.0165,
+      "step": 2690
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6361971111112008,
+      "learning_rate": 4.7142212484525995e-07,
+      "loss": 1.1022,
+      "step": 2691
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5875260323865119,
+      "learning_rate": 4.681226004655126e-07,
+      "loss": 0.8245,
+      "step": 2692
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4598382271610722,
+      "learning_rate": 4.6483438673770364e-07,
+      "loss": 0.7829,
+      "step": 2693
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5876880063854101,
+      "learning_rate": 4.61557487563673e-07,
+      "loss": 1.2484,
+      "step": 2694
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5263927879088985,
+      "learning_rate": 4.5829190683183034e-07,
+      "loss": 1.0533,
+      "step": 2695
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7202735597823287,
+      "learning_rate": 4.55037648417157e-07,
+      "loss": 1.1195,
+      "step": 2696
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3260902811083263,
+      "learning_rate": 4.5179471618119973e-07,
+      "loss": 1.0343,
+      "step": 2697
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5942112667714785,
+      "learning_rate": 4.4856311397206364e-07,
+      "loss": 0.8391,
+      "step": 2698
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6624601117702901,
+      "learning_rate": 4.4534284562441356e-07,
+      "loss": 1.2315,
+      "step": 2699
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5582199464777541,
+      "learning_rate": 4.4213391495946054e-07,
+      "loss": 0.9733,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6175374421218277,
+      "learning_rate": 4.3893632578496327e-07,
+      "loss": 1.2718,
+      "step": 2701
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5394706278596847,
+      "learning_rate": 4.3575008189522536e-07,
+      "loss": 0.9297,
+      "step": 2702
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4414219919424287,
+      "learning_rate": 4.325751870710848e-07,
+      "loss": 0.7781,
+      "step": 2703
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5691423759601713,
+      "learning_rate": 4.294116450799146e-07,
+      "loss": 0.8646,
+      "step": 2704
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5604969540518228,
+      "learning_rate": 4.2625945967561646e-07,
+      "loss": 1.155,
+      "step": 2705
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6150967683489559,
+      "learning_rate": 4.231186345986116e-07,
+      "loss": 1.1815,
+      "step": 2706
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6549575079064966,
+      "learning_rate": 4.199891735758477e-07,
+      "loss": 1.1037,
+      "step": 2707
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6702275249730077,
+      "learning_rate": 4.168710803207865e-07,
+      "loss": 1.1415,
+      "step": 2708
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.34469925579918603,
+      "learning_rate": 4.137643585333939e-07,
+      "loss": 1.0518,
+      "step": 2709
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5503783853841134,
+      "learning_rate": 4.1066901190015e-07,
+      "loss": 1.153,
+      "step": 2710
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5883490521949297,
+      "learning_rate": 4.075850440940343e-07,
+      "loss": 0.9417,
+      "step": 2711
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.38245058027097223,
+      "learning_rate": 4.0451245877452193e-07,
+      "loss": 0.814,
+      "step": 2712
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6716281464907146,
+      "learning_rate": 4.014512595875852e-07,
+      "loss": 1.1553,
+      "step": 2713
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6829234664588952,
+      "learning_rate": 3.984014501656841e-07,
+      "loss": 1.0116,
+      "step": 2714
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5244586973062355,
+      "learning_rate": 3.953630341277603e-07,
+      "loss": 1.2444,
+      "step": 2715
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8010581985519528,
+      "learning_rate": 3.92336015079241e-07,
+      "loss": 1.0081,
+      "step": 2716
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5289550440663165,
+      "learning_rate": 3.8932039661202514e-07,
+      "loss": 0.9609,
+      "step": 2717
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6373211731745531,
+      "learning_rate": 3.8631618230448696e-07,
+      "loss": 1.12,
+      "step": 2718
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5340136224716474,
+      "learning_rate": 3.8332337572146606e-07,
+      "loss": 0.9393,
+      "step": 2719
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4573366687104803,
+      "learning_rate": 3.8034198041426715e-07,
+      "loss": 0.9003,
+      "step": 2720
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.32587908081428707,
+      "learning_rate": 3.7737199992065154e-07,
+      "loss": 1.0463,
+      "step": 2721
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6347462586558509,
+      "learning_rate": 3.7441343776484116e-07,
+      "loss": 1.1738,
+      "step": 2722
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6069315272931041,
+      "learning_rate": 3.714662974575012e-07,
+      "loss": 0.9686,
+      "step": 2723
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.49199491703820836,
+      "learning_rate": 3.6853058249574745e-07,
+      "loss": 1.1048,
+      "step": 2724
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5961212685060715,
+      "learning_rate": 3.6560629636314236e-07,
+      "loss": 1.0541,
+      "step": 2725
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.49232036938039525,
+      "learning_rate": 3.62693442529678e-07,
+      "loss": 1.0382,
+      "step": 2726
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6275529933780228,
+      "learning_rate": 3.5979202445178716e-07,
+      "loss": 1.144,
+      "step": 2727
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6667496863278709,
+      "learning_rate": 3.5690204557233157e-07,
+      "loss": 1.2223,
+      "step": 2728
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.39390652859511655,
+      "learning_rate": 3.540235093205979e-07,
+      "loss": 0.7331,
+      "step": 2729
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5237188242885945,
+      "learning_rate": 3.51156419112294e-07,
+      "loss": 0.9503,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5590675061220469,
+      "learning_rate": 3.4830077834954956e-07,
+      "loss": 0.9563,
+      "step": 2731
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5911970279798421,
+      "learning_rate": 3.4545659042090284e-07,
+      "loss": 1.1929,
+      "step": 2732
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3391513750694738,
+      "learning_rate": 3.4262385870130866e-07,
+      "loss": 1.1159,
+      "step": 2733
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5744289002152775,
+      "learning_rate": 3.398025865521204e-07,
+      "loss": 1.0566,
+      "step": 2734
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5438086505383488,
+      "learning_rate": 3.369927773210979e-07,
+      "loss": 1.0709,
+      "step": 2735
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6256468530255168,
+      "learning_rate": 3.3419443434240083e-07,
+      "loss": 0.9106,
+      "step": 2736
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5573477671431672,
+      "learning_rate": 3.314075609365752e-07,
+      "loss": 1.1229,
+      "step": 2737
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4112082567775563,
+      "learning_rate": 3.2863216041056465e-07,
+      "loss": 0.7536,
+      "step": 2738
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5486950480125719,
+      "learning_rate": 3.25868236057697e-07,
+      "loss": 1.1981,
+      "step": 2739
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6414852631976968,
+      "learning_rate": 3.2311579115767987e-07,
+      "loss": 0.9435,
+      "step": 2740
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6182856777092339,
+      "learning_rate": 3.2037482897660066e-07,
+      "loss": 1.1004,
+      "step": 2741
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6073418806896529,
+      "learning_rate": 3.1764535276692435e-07,
+      "loss": 1.0703,
+      "step": 2742
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6799413930387912,
+      "learning_rate": 3.14927365767479e-07,
+      "loss": 0.9242,
+      "step": 2743
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7181398509989748,
+      "learning_rate": 3.122208712034669e-07,
+      "loss": 1.1946,
+      "step": 2744
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33124141774516885,
+      "learning_rate": 3.0952587228645024e-07,
+      "loss": 1.0486,
+      "step": 2745
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6398251623393346,
+      "learning_rate": 3.0684237221434877e-07,
+      "loss": 1.3217,
+      "step": 2746
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3641302085150502,
+      "learning_rate": 3.041703741714408e-07,
+      "loss": 0.8513,
+      "step": 2747
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7351106691930778,
+      "learning_rate": 3.015098813283546e-07,
+      "loss": 1.0166,
+      "step": 2748
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5254136687292851,
+      "learning_rate": 2.9886089684206477e-07,
+      "loss": 1.1073,
+      "step": 2749
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7186359444732682,
+      "learning_rate": 2.9622342385589256e-07,
+      "loss": 0.9249,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5135165575105186,
+      "learning_rate": 2.935974654995e-07,
+      "loss": 0.9377,
+      "step": 2751
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.561590048942733,
+      "learning_rate": 2.9098302488888007e-07,
+      "loss": 1.0381,
+      "step": 2752
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.583524481461157,
+      "learning_rate": 2.883801051263657e-07,
+      "loss": 1.1539,
+      "step": 2753
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6274371415367397,
+      "learning_rate": 2.85788709300614e-07,
+      "loss": 0.9703,
+      "step": 2754
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.60126507618939,
+      "learning_rate": 2.8320884048661067e-07,
+      "loss": 0.9917,
+      "step": 2755
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.43929825859633015,
+      "learning_rate": 2.806405017456626e-07,
+      "loss": 0.8225,
+      "step": 2756
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6457728385775139,
+      "learning_rate": 2.780836961253941e-07,
+      "loss": 0.9451,
+      "step": 2757
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3052620474633285,
+      "learning_rate": 2.7553842665974385e-07,
+      "loss": 1.1782,
+      "step": 2758
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6453948689755384,
+      "learning_rate": 2.7300469636896365e-07,
+      "loss": 0.8855,
+      "step": 2759
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5722211354294777,
+      "learning_rate": 2.7048250825960963e-07,
+      "loss": 1.2738,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5477015939229972,
+      "learning_rate": 2.6797186532454446e-07,
+      "loss": 0.9814,
+      "step": 2761
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6703622171746818,
+      "learning_rate": 2.6547277054293276e-07,
+      "loss": 0.9944,
+      "step": 2762
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5773116558236284,
+      "learning_rate": 2.629852268802291e-07,
+      "loss": 1.188,
+      "step": 2763
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6364059148028924,
+      "learning_rate": 2.605092372881879e-07,
+      "loss": 0.8793,
+      "step": 2764
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4225078906836869,
+      "learning_rate": 2.580448047048523e-07,
+      "loss": 0.8675,
+      "step": 2765
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6032019112943826,
+      "learning_rate": 2.555919320545486e-07,
+      "loss": 0.8824,
+      "step": 2766
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5789766359230805,
+      "learning_rate": 2.531506222478886e-07,
+      "loss": 0.885,
+      "step": 2767
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5103134894924822,
+      "learning_rate": 2.507208781817638e-07,
+      "loss": 1.1311,
+      "step": 2768
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6348363341005125,
+      "learning_rate": 2.483027027393403e-07,
+      "loss": 1.0071,
+      "step": 2769
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5209329744111996,
+      "learning_rate": 2.458960987900549e-07,
+      "loss": 1.2204,
+      "step": 2770
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.35423556305891907,
+      "learning_rate": 2.4350106918962e-07,
+      "loss": 1.1487,
+      "step": 2771
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6643041058077058,
+      "learning_rate": 2.4111761678000555e-07,
+      "loss": 1.1347,
+      "step": 2772
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.41389167200967397,
+      "learning_rate": 2.387457443894503e-07,
+      "loss": 0.6772,
+      "step": 2773
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5329184086516637,
+      "learning_rate": 2.3638545483244735e-07,
+      "loss": 1.001,
+      "step": 2774
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5736116394587887,
+      "learning_rate": 2.3403675090974852e-07,
+      "loss": 1.0091,
+      "step": 2775
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.54288390719785,
+      "learning_rate": 2.3169963540835783e-07,
+      "loss": 0.9708,
+      "step": 2776
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5767028369287261,
+      "learning_rate": 2.2937411110152574e-07,
+      "loss": 1.246,
+      "step": 2777
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5792156238444103,
+      "learning_rate": 2.2706018074875046e-07,
+      "loss": 0.9569,
+      "step": 2778
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6124038945879582,
+      "learning_rate": 2.2475784709577453e-07,
+      "loss": 1.0069,
+      "step": 2779
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6711527843620623,
+      "learning_rate": 2.224671128745759e-07,
+      "loss": 1.0925,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4874408678172013,
+      "learning_rate": 2.201879808033691e-07,
+      "loss": 1.2273,
+      "step": 2781
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.42245343943158764,
+      "learning_rate": 2.179204535866053e-07,
+      "loss": 0.8334,
+      "step": 2782
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5431493134425016,
+      "learning_rate": 2.156645339149588e-07,
+      "loss": 0.9588,
+      "step": 2783
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5718219250623688,
+      "learning_rate": 2.1342022446533606e-07,
+      "loss": 1.1778,
+      "step": 2784
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3303582540233042,
+      "learning_rate": 2.111875279008657e-07,
+      "loss": 1.0991,
+      "step": 2785
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5256692611861373,
+      "learning_rate": 2.0896644687089074e-07,
+      "loss": 0.892,
+      "step": 2786
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.798165665859673,
+      "learning_rate": 2.0675698401097956e-07,
+      "loss": 1.2055,
+      "step": 2787
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6463156310018408,
+      "learning_rate": 2.0455914194290827e-07,
+      "loss": 0.7975,
+      "step": 2788
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5899852434947159,
+      "learning_rate": 2.0237292327466406e-07,
+      "loss": 1.0933,
+      "step": 2789
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5736615363502883,
+      "learning_rate": 2.0019833060044512e-07,
+      "loss": 1.0194,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6390750256419516,
+      "learning_rate": 1.980353665006507e-07,
+      "loss": 0.9513,
+      "step": 2791
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4142869058593482,
+      "learning_rate": 1.9588403354188324e-07,
+      "loss": 0.8324,
+      "step": 2792
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6151688438534623,
+      "learning_rate": 1.9374433427694405e-07,
+      "loss": 1.0915,
+      "step": 2793
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5451639297424066,
+      "learning_rate": 1.9161627124482552e-07,
+      "loss": 1.2392,
+      "step": 2794
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6949371472754857,
+      "learning_rate": 1.8949984697071767e-07,
+      "loss": 0.9823,
+      "step": 2795
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3293741220412471,
+      "learning_rate": 1.8739506396599717e-07,
+      "loss": 1.106,
+      "step": 2796
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5417978984767167,
+      "learning_rate": 1.8530192472822728e-07,
+      "loss": 1.028,
+      "step": 2797
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5925616154248103,
+      "learning_rate": 1.8322043174115457e-07,
+      "loss": 0.9238,
+      "step": 2798
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5795038145634107,
+      "learning_rate": 1.811505874747066e-07,
+      "loss": 1.1085,
+      "step": 2799
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4189096260698769,
+      "learning_rate": 1.790923943849876e-07,
+      "loss": 0.712,
+      "step": 2800
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7477564393870062,
+      "learning_rate": 1.770458549142773e-07,
+      "loss": 1.2101,
+      "step": 2801
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6715827820251866,
+      "learning_rate": 1.750109714910264e-07,
+      "loss": 0.9701,
+      "step": 2802
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9009519554856652,
+      "learning_rate": 1.729877465298535e-07,
+      "loss": 1.2842,
+      "step": 2803
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5686505756742919,
+      "learning_rate": 1.709761824315448e-07,
+      "loss": 1.18,
+      "step": 2804
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5304230567223936,
+      "learning_rate": 1.6897628158304979e-07,
+      "loss": 1.1039,
+      "step": 2805
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5151714031451273,
+      "learning_rate": 1.669880463574758e-07,
+      "loss": 1.1766,
+      "step": 2806
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6836767290977754,
+      "learning_rate": 1.6501147911409e-07,
+      "loss": 0.9006,
+      "step": 2807
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3345619210641026,
+      "learning_rate": 1.6304658219831292e-07,
+      "loss": 1.0988,
+      "step": 2808
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4138304558065702,
+      "learning_rate": 1.6109335794171733e-07,
+      "loss": 0.7507,
+      "step": 2809
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5660931352842604,
+      "learning_rate": 1.5915180866202585e-07,
+      "loss": 0.9166,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5340524699482041,
+      "learning_rate": 1.572219366631045e-07,
+      "loss": 0.9896,
+      "step": 2811
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5581800371685302,
+      "learning_rate": 1.5530374423496696e-07,
+      "loss": 0.9159,
+      "step": 2812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5406748520382112,
+      "learning_rate": 1.533972336537648e-07,
+      "loss": 1.1933,
+      "step": 2813
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5906204313564906,
+      "learning_rate": 1.5150240718178833e-07,
+      "loss": 1.1418,
+      "step": 2814
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6886246446257751,
+      "learning_rate": 1.4961926706746453e-07,
+      "loss": 1.0646,
+      "step": 2815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5665916240850327,
+      "learning_rate": 1.4774781554535377e-07,
+      "loss": 1.0367,
+      "step": 2816
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6930597672908867,
+      "learning_rate": 1.4588805483614188e-07,
+      "loss": 1.0772,
+      "step": 2817
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4379481064162324,
+      "learning_rate": 1.440399871466469e-07,
+      "loss": 0.7865,
+      "step": 2818
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5908068090980874,
+      "learning_rate": 1.4220361466981248e-07,
+      "loss": 0.9374,
+      "step": 2819
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3343852386322352,
+      "learning_rate": 1.4037893958469994e-07,
+      "loss": 1.1473,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6093880683910542,
+      "learning_rate": 1.3856596405649292e-07,
+      "loss": 0.8264,
+      "step": 2821
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6047019789548064,
+      "learning_rate": 1.3676469023649497e-07,
+      "loss": 1.2295,
+      "step": 2822
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5871230691862822,
+      "learning_rate": 1.3497512026211857e-07,
+      "loss": 0.9725,
+      "step": 2823
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5643707924506447,
+      "learning_rate": 1.3319725625689173e-07,
+      "loss": 0.8812,
+      "step": 2824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7991218146583534,
+      "learning_rate": 1.314311003304525e-07,
+      "loss": 1.212,
+      "step": 2825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.42225718024570896,
+      "learning_rate": 1.2967665457854438e-07,
+      "loss": 0.7312,
+      "step": 2826
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6603937836190398,
+      "learning_rate": 1.2793392108301438e-07,
+      "loss": 1.2215,
+      "step": 2827
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4527981321888755,
+      "learning_rate": 1.2620290191181606e-07,
+      "loss": 1.0758,
+      "step": 2828
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6011881005843123,
+      "learning_rate": 1.2448359911899743e-07,
+      "loss": 1.1229,
+      "step": 2829
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6282737045981703,
+      "learning_rate": 1.2277601474470767e-07,
+      "loss": 1.1942,
+      "step": 2830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6424926099297722,
+      "learning_rate": 1.2108015081518598e-07,
+      "loss": 0.9402,
+      "step": 2831
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7497394143782933,
+      "learning_rate": 1.193960093427704e-07,
+      "loss": 1.2387,
+      "step": 2832
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.35013017396090496,
+      "learning_rate": 1.1772359232588459e-07,
+      "loss": 1.0571,
+      "step": 2833
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5568552564759318,
+      "learning_rate": 1.160629017490389e-07,
+      "loss": 1.118,
+      "step": 2834
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5839719993558248,
+      "learning_rate": 1.1441393958283253e-07,
+      "loss": 0.8839,
+      "step": 2835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.38960070162077304,
+      "learning_rate": 1.127767077839459e-07,
+      "loss": 0.674,
+      "step": 2836
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5504607509930985,
+      "learning_rate": 1.1115120829513825e-07,
+      "loss": 1.2469,
+      "step": 2837
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5308181552892823,
+      "learning_rate": 1.0953744304525116e-07,
+      "loss": 1.0215,
+      "step": 2838
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7115984991007881,
+      "learning_rate": 1.0793541394919838e-07,
+      "loss": 1.1617,
+      "step": 2839
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4962538474842914,
+      "learning_rate": 1.0634512290796816e-07,
+      "loss": 1.087,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7711069335612667,
+      "learning_rate": 1.0476657180862327e-07,
+      "loss": 1.0464,
+      "step": 2841
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5787192020643743,
+      "learning_rate": 1.0319976252429086e-07,
+      "loss": 1.2247,
+      "step": 2842
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5138817405777263,
+      "learning_rate": 1.0164469691416934e-07,
+      "loss": 1.0037,
+      "step": 2843
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5712820737665514,
+      "learning_rate": 1.0010137682351928e-07,
+      "loss": 1.1161,
+      "step": 2844
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.28455327110529127,
+      "learning_rate": 9.856980408366468e-08,
+      "loss": 0.927,
+      "step": 2845
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.638677167256228,
+      "learning_rate": 9.704998051199066e-08,
+      "loss": 1.1985,
+      "step": 2846
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6352498890287589,
+      "learning_rate": 9.554190791193907e-08,
+      "loss": 0.88,
+      "step": 2847
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5882073445707625,
+      "learning_rate": 9.404558807301067e-08,
+      "loss": 1.1212,
+      "step": 2848
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5615684343460592,
+      "learning_rate": 9.256102277075519e-08,
+      "loss": 1.014,
+      "step": 2849
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8263330071599387,
+      "learning_rate": 9.108821376678123e-08,
+      "loss": 1.0531,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.609928028981056,
+      "learning_rate": 8.962716280874084e-08,
+      "loss": 1.16,
+      "step": 2851
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5339502339472671,
+      "learning_rate": 8.817787163033609e-08,
+      "loss": 1.0982,
+      "step": 2852
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.41957490265534303,
+      "learning_rate": 8.674034195131798e-08,
+      "loss": 0.8269,
+      "step": 2853
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6360744977241545,
+      "learning_rate": 8.531457547747534e-08,
+      "loss": 1.0544,
+      "step": 2854
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1067458342867462,
+      "learning_rate": 8.390057390064266e-08,
+      "loss": 0.9777,
+      "step": 2855
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5769221297876588,
+      "learning_rate": 8.24983388986933e-08,
+      "loss": 1.1053,
+      "step": 2856
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5865146008789448,
+      "learning_rate": 8.11078721355374e-08,
+      "loss": 0.9059,
+      "step": 2857
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.32082270769636806,
+      "learning_rate": 7.972917526112178e-08,
+      "loss": 1.1608,
+      "step": 2858
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.806272534987201,
+      "learning_rate": 7.83622499114256e-08,
+      "loss": 1.002,
+      "step": 2859
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.532796528775421,
+      "learning_rate": 7.700709770846138e-08,
+      "loss": 1.0048,
+      "step": 2860
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6244882711020103,
+      "learning_rate": 7.566372026026947e-08,
+      "loss": 1.0814,
+      "step": 2861
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.40131808856423556,
+      "learning_rate": 7.433211916092143e-08,
+      "loss": 0.5976,
+      "step": 2862
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.47405375837124863,
+      "learning_rate": 7.301229599050997e-08,
+      "loss": 1.1724,
+      "step": 2863
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.540503715904292,
+      "learning_rate": 7.170425231515676e-08,
+      "loss": 0.9317,
+      "step": 2864
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7748169061806772,
+      "learning_rate": 7.040798968700357e-08,
+      "loss": 0.9912,
+      "step": 2865
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5542588824577496,
+      "learning_rate": 6.912350964421221e-08,
+      "loss": 1.0167,
+      "step": 2866
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5999818545984814,
+      "learning_rate": 6.785081371096458e-08,
+      "loss": 1.0922,
+      "step": 2867
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6642210845383534,
+      "learning_rate": 6.658990339745708e-08,
+      "loss": 1.185,
+      "step": 2868
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.32756212309268373,
+      "learning_rate": 6.534078019990397e-08,
+      "loss": 1.1469,
+      "step": 2869
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5438266027572908,
+      "learning_rate": 6.410344560053072e-08,
+      "loss": 1.078,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4037197964765387,
+      "learning_rate": 6.287790106757396e-08,
+      "loss": 0.6102,
+      "step": 2871
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6442689737731137,
+      "learning_rate": 6.166414805528265e-08,
+      "loss": 1.0193,
+      "step": 2872
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.51822517885872,
+      "learning_rate": 6.046218800391245e-08,
+      "loss": 1.0964,
+      "step": 2873
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5342760609763295,
+      "learning_rate": 5.9272022339723624e-08,
+      "loss": 1.0478,
+      "step": 2874
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5762472249602643,
+      "learning_rate": 5.809365247498311e-08,
+      "loss": 1.2242,
+      "step": 2875
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6623557222714956,
+      "learning_rate": 5.69270798079613e-08,
+      "loss": 0.84,
+      "step": 2876
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.61364506703365,
+      "learning_rate": 5.5772305722926467e-08,
+      "loss": 1.1045,
+      "step": 2877
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6622357701453077,
+      "learning_rate": 5.462933159015027e-08,
+      "loss": 0.799,
+      "step": 2878
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4018796477546526,
+      "learning_rate": 5.3498158765901145e-08,
+      "loss": 0.8081,
+      "step": 2879
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5688338196778191,
+      "learning_rate": 5.2378788592443167e-08,
+      "loss": 1.3185,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.559868131735317,
+      "learning_rate": 5.1271222398036056e-08,
+      "loss": 0.9658,
+      "step": 2881
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6622524586310968,
+      "learning_rate": 5.017546149693298e-08,
+      "loss": 1.1219,
+      "step": 2882
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3343347681974828,
+      "learning_rate": 4.9091507189377163e-08,
+      "loss": 1.041,
+      "step": 2883
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6282163906296964,
+      "learning_rate": 4.8019360761604186e-08,
+      "loss": 0.995,
+      "step": 2884
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5205283555981339,
+      "learning_rate": 4.695902348583636e-08,
+      "loss": 1.2236,
+      "step": 2885
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5109315430873833,
+      "learning_rate": 4.591049662028502e-08,
+      "loss": 1.0028,
+      "step": 2886
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.665429524896136,
+      "learning_rate": 4.487378140914489e-08,
+      "loss": 1.3286,
+      "step": 2887
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5772924789384919,
+      "learning_rate": 4.384887908259638e-08,
+      "loss": 0.9246,
+      "step": 2888
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.39011840228881584,
+      "learning_rate": 4.283579085680223e-08,
+      "loss": 0.7217,
+      "step": 2889
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6256243713091748,
+      "learning_rate": 4.183451793390747e-08,
+      "loss": 1.0365,
+      "step": 2890
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6180146078067673,
+      "learning_rate": 4.0845061502033935e-08,
+      "loss": 1.1772,
+      "step": 2891
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.788496067555884,
+      "learning_rate": 3.9867422735283544e-08,
+      "loss": 1.2071,
+      "step": 2892
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5640676558810162,
+      "learning_rate": 3.8901602793738334e-08,
+      "loss": 1.0447,
+      "step": 2893
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.68325690932463,
+      "learning_rate": 3.7947602823449335e-08,
+      "loss": 1.23,
+      "step": 2894
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33399480841416,
+      "learning_rate": 3.7005423956447686e-08,
+      "loss": 1.0237,
+      "step": 2895
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5849772371694506,
+      "learning_rate": 3.607506731073573e-08,
+      "loss": 1.1653,
+      "step": 2896
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5566238040967867,
+      "learning_rate": 3.515653399028596e-08,
+      "loss": 0.9413,
+      "step": 2897
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3803135141041098,
+      "learning_rate": 3.4249825085040936e-08,
+      "loss": 0.7939,
+      "step": 2898
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6581142310055779,
+      "learning_rate": 3.3354941670916687e-08,
+      "loss": 1.2523,
+      "step": 2899
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7167251660773277,
+      "learning_rate": 3.247188480979158e-08,
+      "loss": 1.0364,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5566656467443095,
+      "learning_rate": 3.160065554951186e-08,
+      "loss": 1.1961,
+      "step": 2901
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6296830400711719,
+      "learning_rate": 3.074125492389168e-08,
+      "loss": 1.0478,
+      "step": 2902
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6131813891778514,
+      "learning_rate": 2.989368395270531e-08,
+      "loss": 0.8777,
+      "step": 2903
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7175463075948563,
+      "learning_rate": 2.9057943641693787e-08,
+      "loss": 1.2091,
+      "step": 2904
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5850613403888548,
+      "learning_rate": 2.823403498255606e-08,
+      "loss": 0.9254,
+      "step": 2905
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.43618065161256026,
+      "learning_rate": 2.742195895295341e-08,
+      "loss": 0.8731,
+      "step": 2906
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.33460557178889166,
+      "learning_rate": 2.6621716516506137e-08,
+      "loss": 1.0557,
+      "step": 2907
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6694272708205963,
+      "learning_rate": 2.583330862279243e-08,
+      "loss": 1.193,
+      "step": 2908
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5959727357149021,
+      "learning_rate": 2.5056736207348387e-08,
+      "loss": 0.9572,
+      "step": 2909
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.47571633819165393,
+      "learning_rate": 2.4292000191663555e-08,
+      "loss": 1.1041,
+      "step": 2910
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6266239937902833,
+      "learning_rate": 2.353910148318428e-08,
+      "loss": 1.072,
+      "step": 2911
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5432107442931395,
+      "learning_rate": 2.2798040975310353e-08,
+      "loss": 1.0419,
+      "step": 2912
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6122469748941513,
+      "learning_rate": 2.2068819547395038e-08,
+      "loss": 1.3142,
+      "step": 2913
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6458511076654208,
+      "learning_rate": 2.1351438064740605e-08,
+      "loss": 0.9441,
+      "step": 2914
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.39569853310005276,
+      "learning_rate": 2.0645897378601677e-08,
+      "loss": 0.7255,
+      "step": 2915
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5425712119777671,
+      "learning_rate": 1.9952198326183004e-08,
+      "loss": 0.9375,
+      "step": 2916
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5771755054295563,
+      "learning_rate": 1.9270341730636134e-08,
+      "loss": 1.0393,
+      "step": 2917
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.589996368813655,
+      "learning_rate": 1.860032840106163e-08,
+      "loss": 1.1512,
+      "step": 2918
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.33717360210358893,
+      "learning_rate": 1.7942159132505742e-08,
+      "loss": 1.1224,
+      "step": 2919
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6510250735462159,
+      "learning_rate": 1.7295834705961524e-08,
+      "loss": 1.0033,
+      "step": 2920
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5574642135621758,
+      "learning_rate": 1.6661355888366593e-08,
+      "loss": 1.0997,
+      "step": 2921
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6430264472744239,
+      "learning_rate": 1.6038723432599823e-08,
+      "loss": 1.0915,
+      "step": 2922
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5151954562760477,
+      "learning_rate": 1.5427938077486883e-08,
+      "loss": 1.1281,
+      "step": 2923
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4037455702420488,
+      "learning_rate": 1.4829000547793571e-08,
+      "loss": 0.6844,
+      "step": 2924
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5212496009614612,
+      "learning_rate": 1.424191155422583e-08,
+      "loss": 1.1883,
+      "step": 2925
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6093797097660874,
+      "learning_rate": 1.3666671793433062e-08,
+      "loss": 0.9434,
+      "step": 2926
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5784764945771071,
+      "learning_rate": 1.3103281948001478e-08,
+      "loss": 1.1156,
+      "step": 2927
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5827375927854387,
+      "learning_rate": 1.2551742686456313e-08,
+      "loss": 1.0179,
+      "step": 2928
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6748984011483465,
+      "learning_rate": 1.201205466326294e-08,
+      "loss": 0.9461,
+      "step": 2929
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6186807653726479,
+      "learning_rate": 1.1484218518820201e-08,
+      "loss": 1.2057,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3252474680582571,
+      "learning_rate": 1.0968234879468187e-08,
+      "loss": 1.0925,
+      "step": 2931
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5968226264286355,
+      "learning_rate": 1.0464104357477133e-08,
+      "loss": 1.1555,
+      "step": 2932
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3951976220224852,
+      "learning_rate": 9.971827551056302e-09,
+      "loss": 0.8289,
+      "step": 2933
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6114102824603712,
+      "learning_rate": 9.491405044348422e-09,
+      "loss": 1.0587,
+      "step": 2934
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.572436959860283,
+      "learning_rate": 9.022837407427487e-09,
+      "loss": 1.144,
+      "step": 2935
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6127980819695596,
+      "learning_rate": 8.566125196304287e-09,
+      "loss": 0.8956,
+      "step": 2936
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5660565583166278,
+      "learning_rate": 8.121268952917538e-09,
+      "loss": 1.0232,
+      "step": 2937
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6161107645732048,
+      "learning_rate": 7.688269205141653e-09,
+      "loss": 1.0634,
+      "step": 2938
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6017405847953226,
+      "learning_rate": 7.267126466777852e-09,
+      "loss": 1.1831,
+      "step": 2939
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.549582462849232,
+      "learning_rate": 6.857841237560836e-09,
+      "loss": 0.9937,
+      "step": 2940
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5350817445558836,
+      "learning_rate": 6.460414003153226e-09,
+      "loss": 0.9481,
+      "step": 2941
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4216271867395014,
+      "learning_rate": 6.074845235147786e-09,
+      "loss": 0.7901,
+      "step": 2942
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7765162453822767,
+      "learning_rate": 5.701135391065205e-09,
+      "loss": 0.9229,
+      "step": 2943
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3077617488507821,
+      "learning_rate": 5.339284914355203e-09,
+      "loss": 1.1925,
+      "step": 2944
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6188539895499965,
+      "learning_rate": 4.989294234393205e-09,
+      "loss": 0.8663,
+      "step": 2945
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5505529017520336,
+      "learning_rate": 4.651163766484779e-09,
+      "loss": 1.2602,
+      "step": 2946
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5742576889840544,
+      "learning_rate": 4.324893911857864e-09,
+      "loss": 0.9295,
+      "step": 2947
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.616904389134475,
+      "learning_rate": 4.010485057669433e-09,
+      "loss": 1.0227,
+      "step": 2948
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5490668949746352,
+      "learning_rate": 3.707937577002163e-09,
+      "loss": 1.1281,
+      "step": 2949
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6461576516203833,
+      "learning_rate": 3.4172518288611014e-09,
+      "loss": 0.8893,
+      "step": 2950
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.43326365062975325,
+      "learning_rate": 3.1384281581792187e-09,
+      "loss": 0.9112,
+      "step": 2951
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5716256633615495,
+      "learning_rate": 2.8714668958118585e-09,
+      "loss": 0.8462,
+      "step": 2952
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5324384026988067,
+      "learning_rate": 2.6163683585389565e-09,
+      "loss": 0.9344,
+      "step": 2953
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.559754953335191,
+      "learning_rate": 2.3731328490628202e-09,
+      "loss": 1.1434,
+      "step": 2954
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5029834011393842,
+      "learning_rate": 2.1417606560114603e-09,
+      "loss": 1.0803,
+      "step": 2955
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6380124044870235,
+      "learning_rate": 1.922252053933038e-09,
+      "loss": 1.3189,
+      "step": 2956
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32311431372451754,
+      "learning_rate": 1.7146073033003085e-09,
+      "loss": 1.0883,
+      "step": 2957
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6426855729727804,
+      "learning_rate": 1.5188266505061776e-09,
+      "loss": 1.1152,
+      "step": 2958
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4263897417175533,
+      "learning_rate": 1.3349103278681441e-09,
+      "loss": 0.6663,
+      "step": 2959
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5404964927993937,
+      "learning_rate": 1.1628585536216374e-09,
+      "loss": 1.0105,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.543833875405751,
+      "learning_rate": 1.0026715319277901e-09,
+      "loss": 1.0294,
+      "step": 2961
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5535235269534569,
+      "learning_rate": 8.543494528645557e-10,
+      "loss": 0.8936,
+      "step": 2962
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5899107210318071,
+      "learning_rate": 7.178924924344799e-10,
+      "loss": 1.2687,
+      "step": 2963
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5588270776702076,
+      "learning_rate": 5.933008125580397e-10,
+      "loss": 0.979,
+      "step": 2964
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5903805779843307,
+      "learning_rate": 4.805745610791945e-10,
+      "loss": 1.0208,
+      "step": 2965
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6649241015686055,
+      "learning_rate": 3.797138717576143e-10,
+      "loss": 1.0796,
+      "step": 2966
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.509150284599488,
+      "learning_rate": 2.907188642786718e-10,
+      "loss": 1.2436,
+      "step": 2967
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4231241910762904,
+      "learning_rate": 2.135896442434504e-10,
+      "loss": 0.8598,
+      "step": 2968
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5350695813467743,
+      "learning_rate": 1.4832630317429542e-10,
+      "loss": 0.8925,
+      "step": 2969
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.670302681862561,
+      "learning_rate": 9.49289185148139e-11,
+      "loss": 1.218,
+      "step": 2970
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.32792394105432693,
+      "learning_rate": 5.339755362543386e-11,
+      "loss": 1.0753,
+      "step": 2971
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6287039568408325,
+      "learning_rate": 2.3732257790065606e-11,
+      "loss": 0.8871,
+      "step": 2972
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6539404245360348,
+      "learning_rate": 5.933066207219895e-12,
+      "loss": 1.1709,
+      "step": 2973
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5492518571128793,
+      "learning_rate": 0.0,
+      "loss": 0.6842,
+      "step": 2974
+    },
+    {
+      "epoch": 1.0,
+      "step": 2974,
+      "total_flos": 0.0,
+      "train_loss": 0.0688135369800447,
+      "train_runtime": 2470.1389,
+      "train_samples_per_second": 2468.981,
+      "train_steps_per_second": 1.204
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2974,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}