diff --git "a/checkpoint-1676/trainer_state.json" "b/checkpoint-1676/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1676/trainer_state.json"
@@ -0,0 +1,11753 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 1676,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0434608134535555,
+      "learning_rate": 2e-05,
+      "loss": 0.7234,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9514763639201144,
+      "learning_rate": 4e-05,
+      "loss": 0.5916,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8903336235741379,
+      "learning_rate": 6e-05,
+      "loss": 0.5827,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.764603046688215,
+      "learning_rate": 8e-05,
+      "loss": 0.5555,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8956967835024867,
+      "learning_rate": 0.0001,
+      "loss": 0.6073,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.378526599131456,
+      "learning_rate": 0.00012,
+      "loss": 0.6204,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1442512069955262,
+      "learning_rate": 0.00014,
+      "loss": 0.5452,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9100913748328603,
+      "learning_rate": 0.00016,
+      "loss": 0.5858,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6859557442210851,
+      "learning_rate": 0.00018,
+      "loss": 0.5182,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8288606055568941,
+      "learning_rate": 0.0002,
+      "loss": 0.6083,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0417686973414615,
+      "learning_rate": 0.0001999999558168346,
+      "loss": 0.65,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8756017006672853,
+      "learning_rate": 0.00019999982326737747,
+      "loss": 0.5837,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7284541589078422,
+      "learning_rate": 0.0001999996023517457,
+      "loss": 0.5738,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8602799120413903,
+      "learning_rate": 0.0001999992930701345,
+      "loss": 0.595,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7938840633060059,
+      "learning_rate": 0.00019999889542281728,
+      "loss": 0.5907,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7022632853348306,
+      "learning_rate": 0.00019999840941014525,
+      "loss": 0.5513,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6531781263691616,
+      "learning_rate": 0.00019999783503254803,
+      "loss": 0.5475,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5808528104992969,
+      "learning_rate": 0.0001999971722905331,
+      "loss": 0.519,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5452643486331965,
+      "learning_rate": 0.00019999642118468614,
+      "loss": 0.5421,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5893567274117093,
+      "learning_rate": 0.00019999558171567082,
+      "loss": 0.6016,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5039905113559068,
+      "learning_rate": 0.000199994653884229,
+      "loss": 0.6096,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5236847425188783,
+      "learning_rate": 0.00019999363769118055,
+      "loss": 0.5845,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4403777461558745,
+      "learning_rate": 0.00019999253313742344,
+      "loss": 0.5657,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.411935443568472,
+      "learning_rate": 0.00019999134022393375,
+      "loss": 0.5619,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3669921362459581,
+      "learning_rate": 0.0001999900589517656,
+      "loss": 0.6115,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3613613842516578,
+      "learning_rate": 0.0001999886893220512,
+      "loss": 0.5286,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.378560233146017,
+      "learning_rate": 0.0001999872313360008,
+      "loss": 0.5887,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3562687135057843,
+      "learning_rate": 0.00019998568499490283,
+      "loss": 0.5598,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.34581851507208355,
+      "learning_rate": 0.00019998405030012371,
+      "loss": 0.5772,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3812400800902662,
+      "learning_rate": 0.00019998232725310796,
+      "loss": 0.6154,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2876811822631032,
+      "learning_rate": 0.00019998051585537818,
+      "loss": 0.4949,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3291934945038139,
+      "learning_rate": 0.00019997861610853503,
+      "loss": 0.5388,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.35220229516562385,
+      "learning_rate": 0.00019997662801425725,
+      "loss": 0.5801,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3908550087374589,
+      "learning_rate": 0.00019997455157430165,
+      "loss": 0.5783,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.38564043955621646,
+      "learning_rate": 0.00019997238679050308,
+      "loss": 0.5628,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3248792879576579,
+      "learning_rate": 0.00019997013366477453,
+      "loss": 0.5896,
+      "step": 36
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.33243708146621687,
+      "learning_rate": 0.00019996779219910696,
+      "loss": 0.5618,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3589452470436555,
+      "learning_rate": 0.00019996536239556942,
+      "loss": 0.5387,
+      "step": 38
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3635189263065437,
+      "learning_rate": 0.0001999628442563091,
+      "loss": 0.629,
+      "step": 39
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.36761442942017947,
+      "learning_rate": 0.00019996023778355113,
+      "loss": 0.6133,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.30331424568033827,
+      "learning_rate": 0.00019995754297959882,
+      "loss": 0.5377,
+      "step": 41
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3157701200247212,
+      "learning_rate": 0.0001999547598468334,
+      "loss": 0.6249,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3112819252913729,
+      "learning_rate": 0.00019995188838771425,
+      "loss": 0.5424,
+      "step": 43
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.33384944369487113,
+      "learning_rate": 0.0001999489286047788,
+      "loss": 0.6014,
+      "step": 44
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3330564524921197,
+      "learning_rate": 0.00019994588050064243,
+      "loss": 0.5469,
+      "step": 45
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2889450580843479,
+      "learning_rate": 0.00019994274407799872,
+      "loss": 0.512,
+      "step": 46
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3273617072745067,
+      "learning_rate": 0.00019993951933961913,
+      "loss": 0.5456,
+      "step": 47
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3228418979730564,
+      "learning_rate": 0.00019993620628835332,
+      "loss": 0.5716,
+      "step": 48
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3439278828333003,
+      "learning_rate": 0.0001999328049271289,
+      "loss": 0.5177,
+      "step": 49
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3186271172935729,
+      "learning_rate": 0.0001999293152589515,
+      "loss": 0.5502,
+      "step": 50
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.38357922086488366,
+      "learning_rate": 0.0001999257372869048,
+      "loss": 0.6178,
+      "step": 51
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.36013497860303273,
+      "learning_rate": 0.00019992207101415053,
+      "loss": 0.6278,
+      "step": 52
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3146595251755829,
+      "learning_rate": 0.00019991831644392848,
+      "loss": 0.5348,
+      "step": 53
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3986948949803995,
+      "learning_rate": 0.00019991447357955639,
+      "loss": 0.6331,
+      "step": 54
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3194176204715625,
+      "learning_rate": 0.00019991054242443008,
+      "loss": 0.5817,
+      "step": 55
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.29564051537116465,
+      "learning_rate": 0.00019990652298202335,
+      "loss": 0.545,
+      "step": 56
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2908246398716361,
+      "learning_rate": 0.00019990241525588804,
+      "loss": 0.5294,
+      "step": 57
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3480952622658696,
+      "learning_rate": 0.000199898219249654,
+      "loss": 0.6282,
+      "step": 58
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4278607132919695,
+      "learning_rate": 0.00019989393496702907,
+      "loss": 0.7008,
+      "step": 59
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3088760407735635,
+      "learning_rate": 0.00019988956241179912,
+      "loss": 0.5747,
+      "step": 60
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3549589308890128,
+      "learning_rate": 0.00019988510158782804,
+      "loss": 0.615,
+      "step": 61
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.28349959678885256,
+      "learning_rate": 0.00019988055249905767,
+      "loss": 0.577,
+      "step": 62
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.35001480138074803,
+      "learning_rate": 0.00019987591514950787,
+      "loss": 0.5551,
+      "step": 63
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.32895401860950285,
+      "learning_rate": 0.00019987118954327654,
+      "loss": 0.5617,
+      "step": 64
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.347007326906862,
+      "learning_rate": 0.00019986637568453945,
+      "loss": 0.5935,
+      "step": 65
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.30223641676037666,
+      "learning_rate": 0.00019986147357755048,
+      "loss": 0.5355,
+      "step": 66
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.302279046184156,
+      "learning_rate": 0.00019985648322664145,
+      "loss": 0.5571,
+      "step": 67
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3181910281320864,
+      "learning_rate": 0.00019985140463622215,
+      "loss": 0.5198,
+      "step": 68
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.32334719229096776,
+      "learning_rate": 0.0001998462378107803,
+      "loss": 0.5063,
+      "step": 69
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.31038499461943353,
+      "learning_rate": 0.0001998409827548817,
+      "loss": 0.5805,
+      "step": 70
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3032049786542595,
+      "learning_rate": 0.00019983563947316996,
+      "loss": 0.564,
+      "step": 71
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3345643555445713,
+      "learning_rate": 0.00019983020797036683,
+      "loss": 0.5442,
+      "step": 72
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.32583720357675877,
+      "learning_rate": 0.00019982468825127187,
+      "loss": 0.5674,
+      "step": 73
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.30278224625704836,
+      "learning_rate": 0.0001998190803207627,
+      "loss": 0.569,
+      "step": 74
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.29996902392483177,
+      "learning_rate": 0.0001998133841837948,
+      "loss": 0.6142,
+      "step": 75
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2947151895973628,
+      "learning_rate": 0.00019980759984540168,
+      "loss": 0.5084,
+      "step": 76
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.33309005605837944,
+      "learning_rate": 0.0001998017273106947,
+      "loss": 0.5807,
+      "step": 77
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.31281343399912853,
+      "learning_rate": 0.00019979576658486325,
+      "loss": 0.6299,
+      "step": 78
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.30980839190781245,
+      "learning_rate": 0.00019978971767317457,
+      "loss": 0.5521,
+      "step": 79
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2988115356408324,
+      "learning_rate": 0.00019978358058097388,
+      "loss": 0.5645,
+      "step": 80
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2864799366004751,
+      "learning_rate": 0.0001997773553136843,
+      "loss": 0.5604,
+      "step": 81
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.28284272149262185,
+      "learning_rate": 0.00019977104187680688,
+      "loss": 0.5964,
+      "step": 82
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2807639116477172,
+      "learning_rate": 0.00019976464027592053,
+      "loss": 0.5441,
+      "step": 83
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3169919212395633,
+      "learning_rate": 0.00019975815051668217,
+      "loss": 0.5672,
+      "step": 84
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2827524094344841,
+      "learning_rate": 0.0001997515726048265,
+      "loss": 0.5631,
+      "step": 85
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.29538492598974014,
+      "learning_rate": 0.00019974490654616625,
+      "loss": 0.609,
+      "step": 86
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.29397994414783907,
+      "learning_rate": 0.0001997381523465919,
+      "loss": 0.5723,
+      "step": 87
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2824005824065347,
+      "learning_rate": 0.00019973131001207195,
+      "loss": 0.5209,
+      "step": 88
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2847018611508931,
+      "learning_rate": 0.00019972437954865265,
+      "loss": 0.5617,
+      "step": 89
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2908040007926844,
+      "learning_rate": 0.00019971736096245825,
+      "loss": 0.5624,
+      "step": 90
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.27754759063410545,
+      "learning_rate": 0.00019971025425969083,
+      "loss": 0.5353,
+      "step": 91
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2885964599083646,
+      "learning_rate": 0.0001997030594466303,
+      "loss": 0.5181,
+      "step": 92
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.29372725474160666,
+      "learning_rate": 0.00019969577652963444,
+      "loss": 0.5757,
+      "step": 93
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.32149915053639194,
+      "learning_rate": 0.0001996884055151389,
+      "loss": 0.552,
+      "step": 94
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2816717380191918,
+      "learning_rate": 0.00019968094640965717,
+      "loss": 0.4968,
+      "step": 95
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2719020140724135,
+      "learning_rate": 0.00019967339921978062,
+      "loss": 0.5503,
+      "step": 96
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.28166729851780475,
+      "learning_rate": 0.00019966576395217837,
+      "loss": 0.5546,
+      "step": 97
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.27817598279558775,
+      "learning_rate": 0.0001996580406135975,
+      "loss": 0.6145,
+      "step": 98
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3471492732103861,
+      "learning_rate": 0.00019965022921086275,
+      "loss": 0.6464,
+      "step": 99
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2838977359279957,
+      "learning_rate": 0.00019964232975087687,
+      "loss": 0.5576,
+      "step": 100
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2989119777268752,
+      "learning_rate": 0.00019963434224062025,
+      "loss": 0.5747,
+      "step": 101
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.26079447242968457,
+      "learning_rate": 0.0001996262666871512,
+      "loss": 0.5144,
+      "step": 102
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2904578974664885,
+      "learning_rate": 0.00019961810309760577,
+      "loss": 0.5623,
+      "step": 103
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2682051539483259,
+      "learning_rate": 0.00019960985147919778,
+      "loss": 0.5722,
+      "step": 104
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3004061592870477,
+      "learning_rate": 0.00019960151183921897,
+      "loss": 0.5526,
+      "step": 105
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.27675115608209533,
+      "learning_rate": 0.00019959308418503877,
+      "loss": 0.5859,
+      "step": 106
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.26526760651496173,
+      "learning_rate": 0.00019958456852410433,
+      "loss": 0.5395,
+      "step": 107
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.29513224606753785,
+      "learning_rate": 0.0001995759648639406,
+      "loss": 0.59,
+      "step": 108
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.27848396362609984,
+      "learning_rate": 0.00019956727321215044,
+      "loss": 0.6076,
+      "step": 109
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.29804931563512865,
+      "learning_rate": 0.00019955849357641424,
+      "loss": 0.5555,
+      "step": 110
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2756038079003531,
+      "learning_rate": 0.00019954962596449024,
+      "loss": 0.5542,
+      "step": 111
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.28433942136507906,
+      "learning_rate": 0.0001995406703842145,
+      "loss": 0.5527,
+      "step": 112
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.646736414676863,
+      "learning_rate": 0.0001995316268435007,
+      "loss": 0.7024,
+      "step": 113
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.25030459489112267,
+      "learning_rate": 0.00019952249535034025,
+      "loss": 0.4928,
+      "step": 114
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30977646038996587,
+      "learning_rate": 0.00019951327591280236,
+      "loss": 0.5883,
+      "step": 115
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2854791945432696,
+      "learning_rate": 0.0001995039685390339,
+      "loss": 0.6318,
+      "step": 116
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3264119003161445,
+      "learning_rate": 0.00019949457323725946,
+      "loss": 0.5654,
+      "step": 117
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.266512900378873,
+      "learning_rate": 0.0001994850900157813,
+      "loss": 0.5457,
+      "step": 118
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30259231663288877,
+      "learning_rate": 0.0001994755188829794,
+      "loss": 0.5828,
+      "step": 119
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.33798187117798184,
+      "learning_rate": 0.00019946585984731142,
+      "loss": 0.5669,
+      "step": 120
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2769767197030659,
+      "learning_rate": 0.00019945611291731274,
+      "loss": 0.5619,
+      "step": 121
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2657775119925365,
+      "learning_rate": 0.00019944627810159632,
+      "loss": 0.59,
+      "step": 122
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2918370192930011,
+      "learning_rate": 0.00019943635540885279,
+      "loss": 0.5816,
+      "step": 123
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.33417238245851544,
+      "learning_rate": 0.00019942634484785052,
+      "loss": 0.5921,
+      "step": 124
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2482914745160091,
+      "learning_rate": 0.00019941624642743548,
+      "loss": 0.5113,
+      "step": 125
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.28272991564412037,
+      "learning_rate": 0.0001994060601565313,
+      "loss": 0.5543,
+      "step": 126
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.35389083143384653,
+      "learning_rate": 0.00019939578604413912,
+      "loss": 0.5921,
+      "step": 127
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3426787529883331,
+      "learning_rate": 0.00019938542409933787,
+      "loss": 0.6073,
+      "step": 128
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.25621145606980844,
+      "learning_rate": 0.000199374974331284,
+      "loss": 0.5639,
+      "step": 129
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2996022840977615,
+      "learning_rate": 0.00019936443674921158,
+      "loss": 0.5874,
+      "step": 130
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3283726508140091,
+      "learning_rate": 0.0001993538113624323,
+      "loss": 0.6295,
+      "step": 131
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2878523809947236,
+      "learning_rate": 0.00019934309818033544,
+      "loss": 0.5565,
+      "step": 132
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2764010556696928,
+      "learning_rate": 0.0001993322972123878,
+      "loss": 0.554,
+      "step": 133
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.28827215996506156,
+      "learning_rate": 0.0001993214084681338,
+      "loss": 0.5788,
+      "step": 134
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2945105689397871,
+      "learning_rate": 0.00019931043195719548,
+      "loss": 0.5197,
+      "step": 135
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.25439488455073433,
+      "learning_rate": 0.00019929936768927232,
+      "loss": 0.509,
+      "step": 136
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2746742447613063,
+      "learning_rate": 0.00019928821567414144,
+      "loss": 0.5479,
+      "step": 137
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3033960913538536,
+      "learning_rate": 0.00019927697592165747,
+      "loss": 0.5859,
+      "step": 138
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.28486244663010424,
+      "learning_rate": 0.00019926564844175256,
+      "loss": 0.5951,
+      "step": 139
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3208016816012168,
+      "learning_rate": 0.00019925423324443638,
+      "loss": 0.5823,
+      "step": 140
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3005362808247367,
+      "learning_rate": 0.00019924273033979613,
+      "loss": 0.5652,
+      "step": 141
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.26910338931514155,
+      "learning_rate": 0.0001992311397379965,
+      "loss": 0.5463,
+      "step": 142
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28718562659091934,
+      "learning_rate": 0.00019921946144927966,
+      "loss": 0.5245,
+      "step": 143
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2616953117781411,
+      "learning_rate": 0.0001992076954839653,
+      "loss": 0.5358,
+      "step": 144
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3107257337368987,
+      "learning_rate": 0.00019919584185245062,
+      "loss": 0.5536,
+      "step": 145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28427245738065265,
+      "learning_rate": 0.00019918390056521018,
+      "loss": 0.6126,
+      "step": 146
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.25398310452304734,
+      "learning_rate": 0.00019917187163279605,
+      "loss": 0.5068,
+      "step": 147
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2556256469730818,
+      "learning_rate": 0.00019915975506583778,
+      "loss": 0.5416,
+      "step": 148
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2611702329742577,
+      "learning_rate": 0.00019914755087504236,
+      "loss": 0.5276,
+      "step": 149
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2789045987166575,
+      "learning_rate": 0.00019913525907119418,
+      "loss": 0.5591,
+      "step": 150
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.26837658503581957,
+      "learning_rate": 0.000199122879665155,
+      "loss": 0.6581,
+      "step": 151
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.34601396912277804,
+      "learning_rate": 0.0001991104126678641,
+      "loss": 0.5394,
+      "step": 152
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.25684957257052443,
+      "learning_rate": 0.00019909785809033806,
+      "loss": 0.5392,
+      "step": 153
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2906797315813485,
+      "learning_rate": 0.00019908521594367098,
+      "loss": 0.5185,
+      "step": 154
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2852843546202924,
+      "learning_rate": 0.0001990724862390342,
+      "loss": 0.5436,
+      "step": 155
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2875355300862882,
+      "learning_rate": 0.0001990596689876765,
+      "loss": 0.6009,
+      "step": 156
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.32052910212305513,
+      "learning_rate": 0.00019904676420092404,
+      "loss": 0.5831,
+      "step": 157
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.266884162852661,
+      "learning_rate": 0.00019903377189018024,
+      "loss": 0.5459,
+      "step": 158
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2957365744895018,
+      "learning_rate": 0.000199020692066926,
+      "loss": 0.5211,
+      "step": 159
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.24951992931808137,
+      "learning_rate": 0.00019900752474271945,
+      "loss": 0.497,
+      "step": 160
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.30509964150122953,
+      "learning_rate": 0.0001989942699291961,
+      "loss": 0.5812,
+      "step": 161
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2790293776337124,
+      "learning_rate": 0.0001989809276380687,
+      "loss": 0.5856,
+      "step": 162
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.24940387850774506,
+      "learning_rate": 0.00019896749788112737,
+      "loss": 0.5281,
+      "step": 163
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2664890107453781,
+      "learning_rate": 0.0001989539806702395,
+      "loss": 0.524,
+      "step": 164
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2896608423493073,
+      "learning_rate": 0.0001989403760173497,
+      "loss": 0.5171,
+      "step": 165
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2544937836162412,
+      "learning_rate": 0.00019892668393447997,
+      "loss": 0.5546,
+      "step": 166
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2626626371027326,
+      "learning_rate": 0.00019891290443372944,
+      "loss": 0.5498,
+      "step": 167
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.281410490858952,
+      "learning_rate": 0.0001988990375272746,
+      "loss": 0.5377,
+      "step": 168
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3376943176164128,
+      "learning_rate": 0.0001988850832273691,
+      "loss": 0.5469,
+      "step": 169
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2507691377758427,
+      "learning_rate": 0.0001988710415463439,
+      "loss": 0.549,
+      "step": 170
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.27178996752570117,
+      "learning_rate": 0.00019885691249660702,
+      "loss": 0.5636,
+      "step": 171
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3359421766962587,
+      "learning_rate": 0.00019884269609064386,
+      "loss": 0.5957,
+      "step": 172
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2638709645045905,
+      "learning_rate": 0.0001988283923410169,
+      "loss": 0.5793,
+      "step": 173
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.25585919726912226,
+      "learning_rate": 0.00019881400126036582,
+      "loss": 0.5817,
+      "step": 174
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2905067973645414,
+      "learning_rate": 0.00019879952286140754,
+      "loss": 0.5585,
+      "step": 175
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.24197399766587002,
+      "learning_rate": 0.0001987849571569361,
+      "loss": 0.507,
+      "step": 176
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28898034252731664,
+      "learning_rate": 0.0001987703041598226,
+      "loss": 0.5981,
+      "step": 177
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.26516349701479863,
+      "learning_rate": 0.00019875556388301543,
+      "loss": 0.56,
+      "step": 178
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.27235027968517367,
+      "learning_rate": 0.00019874073633953997,
+      "loss": 0.5872,
+      "step": 179
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2692241567318253,
+      "learning_rate": 0.00019872582154249884,
+      "loss": 0.5397,
+      "step": 180
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2560507155942398,
+      "learning_rate": 0.00019871081950507163,
+      "loss": 0.5431,
+      "step": 181
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.26691224099103567,
+      "learning_rate": 0.00019869573024051517,
+      "loss": 0.5608,
+      "step": 182
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2961375506924155,
+      "learning_rate": 0.00019868055376216323,
+      "loss": 0.5784,
+      "step": 183
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.26055755072015874,
+      "learning_rate": 0.00019866529008342673,
+      "loss": 0.5369,
+      "step": 184
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2525359310079611,
+      "learning_rate": 0.00019864993921779361,
+      "loss": 0.5438,
+      "step": 185
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.249327141855566,
+      "learning_rate": 0.0001986345011788289,
+      "loss": 0.5668,
+      "step": 186
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2983950007732028,
+      "learning_rate": 0.00019861897598017457,
+      "loss": 0.5271,
+      "step": 187
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.25610455444964525,
+      "learning_rate": 0.00019860336363554973,
+      "loss": 0.6012,
+      "step": 188
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.24760444410184018,
+      "learning_rate": 0.0001985876641587504,
+      "loss": 0.5066,
+      "step": 189
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2614264060863463,
+      "learning_rate": 0.00019857187756364958,
+      "loss": 0.5792,
+      "step": 190
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.27219045408444215,
+      "learning_rate": 0.00019855600386419744,
+      "loss": 0.543,
+      "step": 191
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.24606131498871828,
+      "learning_rate": 0.00019854004307442088,
+      "loss": 0.5676,
+      "step": 192
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.28394763236035964,
+      "learning_rate": 0.0001985239952084239,
+      "loss": 0.6032,
+      "step": 193
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.28350170917034406,
+      "learning_rate": 0.0001985078602803874,
+      "loss": 0.6264,
+      "step": 194
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24011552907338696,
+      "learning_rate": 0.00019849163830456922,
+      "loss": 0.4793,
+      "step": 195
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2561209086280576,
+      "learning_rate": 0.00019847532929530415,
+      "loss": 0.6198,
+      "step": 196
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.23712886628255178,
+      "learning_rate": 0.00019845893326700384,
+      "loss": 0.4989,
+      "step": 197
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26720592489417233,
+      "learning_rate": 0.00019844245023415685,
+      "loss": 0.4934,
+      "step": 198
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2753251397417421,
+      "learning_rate": 0.0001984258802113287,
+      "loss": 0.5544,
+      "step": 199
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2557869713293877,
+      "learning_rate": 0.0001984092232131616,
+      "loss": 0.5643,
+      "step": 200
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2669651919314609,
+      "learning_rate": 0.0001983924792543748,
+      "loss": 0.5879,
+      "step": 201
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.25579187132615644,
+      "learning_rate": 0.00019837564834976432,
+      "loss": 0.5742,
+      "step": 202
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2550207949421237,
+      "learning_rate": 0.000198358730514203,
+      "loss": 0.574,
+      "step": 203
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.23565090455665763,
+      "learning_rate": 0.0001983417257626405,
+      "loss": 0.5299,
+      "step": 204
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.236980034600526,
+      "learning_rate": 0.00019832463411010331,
+      "loss": 0.5199,
+      "step": 205
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2434029841189093,
+      "learning_rate": 0.0001983074555716947,
+      "loss": 0.5477,
+      "step": 206
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24771511082227154,
+      "learning_rate": 0.00019829019016259468,
+      "loss": 0.5697,
+      "step": 207
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.23705880771864213,
+      "learning_rate": 0.00019827283789806011,
+      "loss": 0.521,
+      "step": 208
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24167295291353477,
+      "learning_rate": 0.0001982553987934245,
+      "loss": 0.558,
+      "step": 209
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2535406245356529,
+      "learning_rate": 0.0001982378728640982,
+      "loss": 0.5693,
+      "step": 210
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24865334136075806,
+      "learning_rate": 0.00019822026012556818,
+      "loss": 0.5499,
+      "step": 211
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2544751551481819,
+      "learning_rate": 0.0001982025605933982,
+      "loss": 0.5449,
+      "step": 212
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2599391794330939,
+      "learning_rate": 0.0001981847742832287,
+      "loss": 0.6222,
+      "step": 213
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.23171934920449544,
+      "learning_rate": 0.00019816690121077674,
+      "loss": 0.5448,
+      "step": 214
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24380268930715565,
+      "learning_rate": 0.00019814894139183614,
+      "loss": 0.5773,
+      "step": 215
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2518755786374484,
+      "learning_rate": 0.00019813089484227732,
+      "loss": 0.5479,
+      "step": 216
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.23133984467720642,
+      "learning_rate": 0.00019811276157804733,
+      "loss": 0.471,
+      "step": 217
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.250968947574734,
+      "learning_rate": 0.00019809454161516993,
+      "loss": 0.5738,
+      "step": 218
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.25976015596485974,
+      "learning_rate": 0.00019807623496974537,
+      "loss": 0.5592,
+      "step": 219
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24400199531999783,
+      "learning_rate": 0.0001980578416579506,
+      "loss": 0.5266,
+      "step": 220
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24001572180370875,
+      "learning_rate": 0.00019803936169603912,
+      "loss": 0.5843,
+      "step": 221
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.22867208326764507,
+      "learning_rate": 0.00019802079510034096,
+      "loss": 0.518,
+      "step": 222
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2381724022911579,
+      "learning_rate": 0.00019800214188726276,
+      "loss": 0.5175,
+      "step": 223
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2700455397530704,
+      "learning_rate": 0.00019798340207328766,
+      "loss": 0.5804,
+      "step": 224
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24320219539003604,
+      "learning_rate": 0.00019796457567497537,
+      "loss": 0.5304,
+      "step": 225
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2370472610839002,
+      "learning_rate": 0.0001979456627089621,
+      "loss": 0.5671,
+      "step": 226
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.26756212643991917,
+      "learning_rate": 0.0001979266631919605,
+      "loss": 0.5528,
+      "step": 227
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.24929490389202372,
+      "learning_rate": 0.00019790757714075979,
+      "loss": 0.5407,
+      "step": 228
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.23090595152280974,
+      "learning_rate": 0.00019788840457222556,
+      "loss": 0.5258,
+      "step": 229
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2979564627406142,
+      "learning_rate": 0.0001978691455033,
+      "loss": 0.5367,
+      "step": 230
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.24228845587479894,
+      "learning_rate": 0.0001978497999510015,
+      "loss": 0.5344,
+      "step": 231
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.25363482164729867,
+      "learning_rate": 0.00019783036793242516,
+      "loss": 0.5669,
+      "step": 232
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.23622712417060854,
+      "learning_rate": 0.00019781084946474226,
+      "loss": 0.5797,
+      "step": 233
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.21594682302559634,
+      "learning_rate": 0.00019779124456520056,
+      "loss": 0.5011,
+      "step": 234
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.24211833950801223,
+      "learning_rate": 0.0001977715532511242,
+      "loss": 0.5164,
+      "step": 235
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2693820157715391,
+      "learning_rate": 0.0001977517755399137,
+      "loss": 0.5806,
+      "step": 236
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.24734981937576542,
+      "learning_rate": 0.00019773191144904586,
+      "loss": 0.5233,
+      "step": 237
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2741663737268136,
+      "learning_rate": 0.00019771196099607386,
+      "loss": 0.5402,
+      "step": 238
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2521489075033339,
+      "learning_rate": 0.00019769192419862716,
+      "loss": 0.5862,
+      "step": 239
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2633255671236209,
+      "learning_rate": 0.0001976718010744116,
+      "loss": 0.548,
+      "step": 240
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2691399721238696,
+      "learning_rate": 0.00019765159164120916,
+      "loss": 0.5648,
+      "step": 241
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.25501545746966797,
+      "learning_rate": 0.00019763129591687827,
+      "loss": 0.5602,
+      "step": 242
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3049976908839563,
+      "learning_rate": 0.00019761091391935347,
+      "loss": 0.5508,
+      "step": 243
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24467331400916031,
+      "learning_rate": 0.00019759044566664558,
+      "loss": 0.5229,
+      "step": 244
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.27011009612786374,
+      "learning_rate": 0.00019756989117684164,
+      "loss": 0.5448,
+      "step": 245
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24427143044387528,
+      "learning_rate": 0.00019754925046810493,
+      "loss": 0.5435,
+      "step": 246
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.22753961311031143,
+      "learning_rate": 0.00019752852355867486,
+      "loss": 0.5369,
+      "step": 247
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.23865046003559778,
+      "learning_rate": 0.00019750771046686704,
+      "loss": 0.5354,
+      "step": 248
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2736283930569903,
+      "learning_rate": 0.00019748681121107325,
+      "loss": 0.5588,
+      "step": 249
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24727127426749082,
+      "learning_rate": 0.00019746582580976136,
+      "loss": 0.5753,
+      "step": 250
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2828829340227291,
+      "learning_rate": 0.00019744475428147546,
+      "loss": 0.6793,
+      "step": 251
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.21818184663896711,
+      "learning_rate": 0.00019742359664483563,
+      "loss": 0.5248,
+      "step": 252
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2320708306833192,
+      "learning_rate": 0.00019740235291853812,
+      "loss": 0.5461,
+      "step": 253
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.25703347930088793,
+      "learning_rate": 0.00019738102312135523,
+      "loss": 0.5713,
+      "step": 254
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24399588128874033,
+      "learning_rate": 0.0001973596072721353,
+      "loss": 0.5178,
+      "step": 255
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2229881452119291,
+      "learning_rate": 0.00019733810538980281,
+      "loss": 0.5144,
+      "step": 256
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.23889465035265364,
+      "learning_rate": 0.0001973165174933581,
+      "loss": 0.5727,
+      "step": 257
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.25790569964877214,
+      "learning_rate": 0.0001972948436018776,
+      "loss": 0.5659,
+      "step": 258
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.22511338701135042,
+      "learning_rate": 0.00019727308373451377,
+      "loss": 0.5292,
+      "step": 259
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23111498863739158,
+      "learning_rate": 0.000197251237910495,
+      "loss": 0.5267,
+      "step": 260
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23740021982137896,
+      "learning_rate": 0.00019722930614912563,
+      "loss": 0.5499,
+      "step": 261
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.24020258332985106,
+      "learning_rate": 0.00019720728846978598,
+      "loss": 0.5604,
+      "step": 262
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23947573439011133,
+      "learning_rate": 0.00019718518489193225,
+      "loss": 0.5638,
+      "step": 263
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23526187217481284,
+      "learning_rate": 0.00019716299543509654,
+      "loss": 0.5436,
+      "step": 264
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2505831003191156,
+      "learning_rate": 0.00019714072011888686,
+      "loss": 0.5039,
+      "step": 265
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.21592487431784965,
+      "learning_rate": 0.00019711835896298713,
+      "loss": 0.484,
+      "step": 266
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26528309878122613,
+      "learning_rate": 0.00019709591198715707,
+      "loss": 0.539,
+      "step": 267
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.21635789850987178,
+      "learning_rate": 0.00019707337921123221,
+      "loss": 0.5553,
+      "step": 268
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23671151623321054,
+      "learning_rate": 0.00019705076065512398,
+      "loss": 0.4968,
+      "step": 269
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.25400871793456326,
+      "learning_rate": 0.00019702805633881957,
+      "loss": 0.5982,
+      "step": 270
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2622810971314154,
+      "learning_rate": 0.0001970052662823819,
+      "loss": 0.5879,
+      "step": 271
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.22931814830456296,
+      "learning_rate": 0.00019698239050594977,
+      "loss": 0.5611,
+      "step": 272
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.213695866193263,
+      "learning_rate": 0.0001969594290297376,
+      "loss": 0.5386,
+      "step": 273
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2431252328609808,
+      "learning_rate": 0.00019693638187403563,
+      "loss": 0.6039,
+      "step": 274
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23108667253454973,
+      "learning_rate": 0.00019691324905920984,
+      "loss": 0.5579,
+      "step": 275
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.22718831415064272,
+      "learning_rate": 0.0001968900306057018,
+      "loss": 0.5196,
+      "step": 276
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.23632362967796033,
+      "learning_rate": 0.0001968667265340288,
+      "loss": 0.5336,
+      "step": 277
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2421878973201691,
+      "learning_rate": 0.00019684333686478383,
+      "loss": 0.5928,
+      "step": 278
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.225775487602821,
+      "learning_rate": 0.00019681986161863542,
+      "loss": 0.552,
+      "step": 279
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.23037375759338816,
+      "learning_rate": 0.00019679630081632782,
+      "loss": 0.4983,
+      "step": 280
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.24684136612832333,
+      "learning_rate": 0.00019677265447868086,
+      "loss": 0.5655,
+      "step": 281
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2412756534364674,
+      "learning_rate": 0.0001967489226265899,
+      "loss": 0.5063,
+      "step": 282
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.22005765622474396,
+      "learning_rate": 0.00019672510528102597,
+      "loss": 0.5188,
+      "step": 283
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.25071531725514384,
+      "learning_rate": 0.0001967012024630355,
+      "loss": 0.5938,
+      "step": 284
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.22139592405512468,
+      "learning_rate": 0.00019667721419374065,
+      "loss": 0.4917,
+      "step": 285
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.23067244251762076,
+      "learning_rate": 0.00019665314049433888,
+      "loss": 0.5584,
+      "step": 286
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.23829875535152545,
+      "learning_rate": 0.00019662898138610323,
+      "loss": 0.5264,
+      "step": 287
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2641034663020514,
+      "learning_rate": 0.00019660473689038228,
+      "loss": 0.5805,
+      "step": 288
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.22321690487140503,
+      "learning_rate": 0.00019658040702859997,
+      "loss": 0.5529,
+      "step": 289
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2502632163555589,
+      "learning_rate": 0.00019655599182225565,
+      "loss": 0.5347,
+      "step": 290
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2392608020883604,
+      "learning_rate": 0.00019653149129292426,
+      "loss": 0.5263,
+      "step": 291
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2539237490519494,
+      "learning_rate": 0.00019650690546225592,
+      "loss": 0.5156,
+      "step": 292
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.21964099103511592,
+      "learning_rate": 0.00019648223435197627,
+      "loss": 0.5101,
+      "step": 293
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.24992985700157416,
+      "learning_rate": 0.00019645747798388628,
+      "loss": 0.5621,
+      "step": 294
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2559439615345381,
+      "learning_rate": 0.0001964326363798622,
+      "loss": 0.5753,
+      "step": 295
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2504368010690795,
+      "learning_rate": 0.00019640770956185567,
+      "loss": 0.5558,
+      "step": 296
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.21022187251218089,
+      "learning_rate": 0.0001963826975518936,
+      "loss": 0.5322,
+      "step": 297
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2422143856532352,
+      "learning_rate": 0.00019635760037207817,
+      "loss": 0.538,
+      "step": 298
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.23174554470584352,
+      "learning_rate": 0.00019633241804458687,
+      "loss": 0.5545,
+      "step": 299
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.267070426953347,
+      "learning_rate": 0.00019630715059167238,
+      "loss": 0.5632,
+      "step": 300
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.22256523603127878,
+      "learning_rate": 0.0001962817980356626,
+      "loss": 0.545,
+      "step": 301
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.24403300497950306,
+      "learning_rate": 0.0001962563603989607,
+      "loss": 0.5448,
+      "step": 302
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2193103621019292,
+      "learning_rate": 0.00019623083770404492,
+      "loss": 0.5064,
+      "step": 303
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.23299992325798072,
+      "learning_rate": 0.0001962052299734688,
+      "loss": 0.5192,
+      "step": 304
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2371054150083945,
+      "learning_rate": 0.00019617953722986096,
+      "loss": 0.5157,
+      "step": 305
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2436064901189273,
+      "learning_rate": 0.00019615375949592504,
+      "loss": 0.5672,
+      "step": 306
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.25098365347678436,
+      "learning_rate": 0.00019612789679443997,
+      "loss": 0.5548,
+      "step": 307
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2319425382974216,
+      "learning_rate": 0.00019610194914825962,
+      "loss": 0.5293,
+      "step": 308
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.24156576209403272,
+      "learning_rate": 0.000196075916580313,
+      "loss": 0.5672,
+      "step": 309
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2337383575844323,
+      "learning_rate": 0.0001960497991136041,
+      "loss": 0.5509,
+      "step": 310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.23799692988502053,
+      "learning_rate": 0.00019602359677121199,
+      "loss": 0.5604,
+      "step": 311
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2296728275122706,
+      "learning_rate": 0.0001959973095762907,
+      "loss": 0.5371,
+      "step": 312
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.22381626870035518,
+      "learning_rate": 0.00019597093755206936,
+      "loss": 0.5465,
+      "step": 313
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.23335681761234933,
+      "learning_rate": 0.00019594448072185182,
+      "loss": 0.5386,
+      "step": 314
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.22582265649304345,
+      "learning_rate": 0.00019591793910901707,
+      "loss": 0.543,
+      "step": 315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2439330072441743,
+      "learning_rate": 0.00019589131273701894,
+      "loss": 0.5177,
+      "step": 316
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2138593422237162,
+      "learning_rate": 0.00019586460162938622,
+      "loss": 0.5157,
+      "step": 317
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.24003613679646058,
+      "learning_rate": 0.00019583780580972253,
+      "loss": 0.5611,
+      "step": 318
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2552582800734971,
+      "learning_rate": 0.00019581092530170633,
+      "loss": 0.5922,
+      "step": 319
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.21898423827197905,
+      "learning_rate": 0.00019578396012909092,
+      "loss": 0.5272,
+      "step": 320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.22013525107478477,
+      "learning_rate": 0.00019575691031570446,
+      "loss": 0.5184,
+      "step": 321
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.21113180640163418,
+      "learning_rate": 0.00019572977588544986,
+      "loss": 0.5134,
+      "step": 322
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.22335860079943387,
+      "learning_rate": 0.00019570255686230485,
+      "loss": 0.5227,
+      "step": 323
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.23006684721287293,
+      "learning_rate": 0.00019567525327032187,
+      "loss": 0.5885,
+      "step": 324
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.21933564641390155,
+      "learning_rate": 0.0001956478651336281,
+      "loss": 0.5598,
+      "step": 325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.21770749652400337,
+      "learning_rate": 0.00019562039247642546,
+      "loss": 0.5082,
+      "step": 326
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.22800344133653658,
+      "learning_rate": 0.00019559283532299043,
+      "loss": 0.5539,
+      "step": 327
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2385574924163353,
+      "learning_rate": 0.00019556519369767438,
+      "loss": 0.5497,
+      "step": 328
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.23099538598172079,
+      "learning_rate": 0.0001955374676249031,
+      "loss": 0.5138,
+      "step": 329
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.21517217725478144,
+      "learning_rate": 0.0001955096571291772,
+      "loss": 0.5051,
+      "step": 330
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.21535771106277588,
+      "learning_rate": 0.0001954817622350717,
+      "loss": 0.524,
+      "step": 331
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.20361747402971658,
+      "learning_rate": 0.00019545378296723635,
+      "loss": 0.4989,
+      "step": 332
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.24644921068325687,
+      "learning_rate": 0.0001954257193503954,
+      "loss": 0.5927,
+      "step": 333
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.24765268362172385,
+      "learning_rate": 0.0001953975714093476,
+      "loss": 0.5451,
+      "step": 334
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.20846277824915477,
+      "learning_rate": 0.00019536933916896633,
+      "loss": 0.5259,
+      "step": 335
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2457371199220107,
+      "learning_rate": 0.00019534102265419932,
+      "loss": 0.5784,
+      "step": 336
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.23029745387228598,
+      "learning_rate": 0.00019531262189006882,
+      "loss": 0.5918,
+      "step": 337
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2387820151516941,
+      "learning_rate": 0.0001952841369016716,
+      "loss": 0.5576,
+      "step": 338
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.226451643448924,
+      "learning_rate": 0.00019525556771417875,
+      "loss": 0.5241,
+      "step": 339
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.22086691724075064,
+      "learning_rate": 0.00019522691435283585,
+      "loss": 0.5392,
+      "step": 340
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2259720671796772,
+      "learning_rate": 0.00019519817684296285,
+      "loss": 0.516,
+      "step": 341
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2244741513315317,
+      "learning_rate": 0.00019516935520995393,
+      "loss": 0.569,
+      "step": 342
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.23890602213836412,
+      "learning_rate": 0.0001951404494792778,
+      "loss": 0.524,
+      "step": 343
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.22136745892767679,
+      "learning_rate": 0.00019511145967647737,
+      "loss": 0.5472,
+      "step": 344
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.22275740066078306,
+      "learning_rate": 0.00019508238582716984,
+      "loss": 0.5553,
+      "step": 345
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.21225155652808625,
+      "learning_rate": 0.00019505322795704676,
+      "loss": 0.5302,
+      "step": 346
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.22704101844750724,
+      "learning_rate": 0.0001950239860918738,
+      "loss": 0.5485,
+      "step": 347
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2135110250199134,
+      "learning_rate": 0.00019499466025749097,
+      "loss": 0.5343,
+      "step": 348
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.22772242632973722,
+      "learning_rate": 0.00019496525047981242,
+      "loss": 0.5159,
+      "step": 349
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4444297049160113,
+      "learning_rate": 0.00019493575678482649,
+      "loss": 0.5121,
+      "step": 350
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.226632712040011,
+      "learning_rate": 0.0001949061791985957,
+      "loss": 0.5304,
+      "step": 351
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.22132303156586286,
+      "learning_rate": 0.00019487651774725663,
+      "loss": 0.4817,
+      "step": 352
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.23206581340772667,
+      "learning_rate": 0.00019484677245702004,
+      "loss": 0.5258,
+      "step": 353
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2374903834541946,
+      "learning_rate": 0.0001948169433541708,
+      "loss": 0.5318,
+      "step": 354
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.22896458770920267,
+      "learning_rate": 0.00019478703046506773,
+      "loss": 0.4806,
+      "step": 355
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.21040301706147688,
+      "learning_rate": 0.00019475703381614375,
+      "loss": 0.5144,
+      "step": 356
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.21179618454444762,
+      "learning_rate": 0.00019472695343390585,
+      "loss": 0.524,
+      "step": 357
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.20436614333218908,
+      "learning_rate": 0.00019469678934493488,
+      "loss": 0.501,
+      "step": 358
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2478256130980173,
+      "learning_rate": 0.0001946665415758858,
+      "loss": 0.5386,
+      "step": 359
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.226116084636948,
+      "learning_rate": 0.00019463621015348748,
+      "loss": 0.5101,
+      "step": 360
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.21838947264457534,
+      "learning_rate": 0.00019460579510454263,
+      "loss": 0.5296,
+      "step": 361
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2152879498375444,
+      "learning_rate": 0.00019457529645592792,
+      "loss": 0.512,
+      "step": 362
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.22514971802642378,
+      "learning_rate": 0.00019454471423459389,
+      "loss": 0.5593,
+      "step": 363
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.23402386101532432,
+      "learning_rate": 0.00019451404846756494,
+      "loss": 0.555,
+      "step": 364
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2244514913016572,
+      "learning_rate": 0.00019448329918193927,
+      "loss": 0.5689,
+      "step": 365
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.22260707231596893,
+      "learning_rate": 0.00019445246640488893,
+      "loss": 0.6062,
+      "step": 366
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.21791090145253736,
+      "learning_rate": 0.00019442155016365965,
+      "loss": 0.531,
+      "step": 367
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.21895889257567258,
+      "learning_rate": 0.00019439055048557101,
+      "loss": 0.5538,
+      "step": 368
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.21306696799872818,
+      "learning_rate": 0.00019435946739801633,
+      "loss": 0.5673,
+      "step": 369
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2294607768810707,
+      "learning_rate": 0.00019432830092846253,
+      "loss": 0.5855,
+      "step": 370
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.22758409665267085,
+      "learning_rate": 0.0001942970511044503,
+      "loss": 0.5783,
+      "step": 371
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.21334100614200935,
+      "learning_rate": 0.00019426571795359398,
+      "loss": 0.5056,
+      "step": 372
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.24187624093875965,
+      "learning_rate": 0.0001942343015035815,
+      "loss": 0.543,
+      "step": 373
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2275714845035408,
+      "learning_rate": 0.00019420280178217443,
+      "loss": 0.5329,
+      "step": 374
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.23237641477505608,
+      "learning_rate": 0.00019417121881720793,
+      "loss": 0.5134,
+      "step": 375
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25196886008416386,
+      "learning_rate": 0.0001941395526365907,
+      "loss": 0.6023,
+      "step": 376
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.22418514390796682,
+      "learning_rate": 0.00019410780326830498,
+      "loss": 0.5529,
+      "step": 377
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.21438856736265666,
+      "learning_rate": 0.0001940759707404065,
+      "loss": 0.5134,
+      "step": 378
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2331754234870151,
+      "learning_rate": 0.00019404405508102455,
+      "loss": 0.5406,
+      "step": 379
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24908239322819828,
+      "learning_rate": 0.00019401205631836178,
+      "loss": 0.5377,
+      "step": 380
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.21332745391417657,
+      "learning_rate": 0.00019397997448069435,
+      "loss": 0.5025,
+      "step": 381
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.20749658696001225,
+      "learning_rate": 0.00019394780959637177,
+      "loss": 0.5257,
+      "step": 382
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2237716482529178,
+      "learning_rate": 0.000193915561693817,
+      "loss": 0.5,
+      "step": 383
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25234282015654147,
+      "learning_rate": 0.00019388323080152633,
+      "loss": 0.5753,
+      "step": 384
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2252939326339829,
+      "learning_rate": 0.00019385081694806936,
+      "loss": 0.5662,
+      "step": 385
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.21979629294660186,
+      "learning_rate": 0.00019381832016208904,
+      "loss": 0.5141,
+      "step": 386
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24762535901866153,
+      "learning_rate": 0.0001937857404723016,
+      "loss": 0.6193,
+      "step": 387
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25032044234085526,
+      "learning_rate": 0.00019375307790749647,
+      "loss": 0.5024,
+      "step": 388
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.22892425302508923,
+      "learning_rate": 0.0001937203324965364,
+      "loss": 0.5401,
+      "step": 389
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2461599771002527,
+      "learning_rate": 0.0001936875042683573,
+      "loss": 0.5301,
+      "step": 390
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.22363255721865732,
+      "learning_rate": 0.00019365459325196825,
+      "loss": 0.5538,
+      "step": 391
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.22482667580972365,
+      "learning_rate": 0.00019362159947645152,
+      "loss": 0.4928,
+      "step": 392
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.22869596173751142,
+      "learning_rate": 0.00019358852297096253,
+      "loss": 0.5546,
+      "step": 393
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2274546469780496,
+      "learning_rate": 0.00019355536376472972,
+      "loss": 0.5763,
+      "step": 394
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.21284874650406885,
+      "learning_rate": 0.0001935221218870547,
+      "loss": 0.5778,
+      "step": 395
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.23158847478661296,
+      "learning_rate": 0.0001934887973673121,
+      "loss": 0.5654,
+      "step": 396
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.24510006704514478,
+      "learning_rate": 0.0001934553902349496,
+      "loss": 0.5053,
+      "step": 397
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.20330878586204656,
+      "learning_rate": 0.00019342190051948777,
+      "loss": 0.5171,
+      "step": 398
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2131804710318274,
+      "learning_rate": 0.0001933883282505203,
+      "loss": 0.5286,
+      "step": 399
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.23297933515492006,
+      "learning_rate": 0.00019335467345771377,
+      "loss": 0.5593,
+      "step": 400
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.24611434220143105,
+      "learning_rate": 0.0001933209361708077,
+      "loss": 0.604,
+      "step": 401
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.23281321736249425,
+      "learning_rate": 0.00019328711641961445,
+      "loss": 0.5579,
+      "step": 402
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.21399822113981087,
+      "learning_rate": 0.00019325321423401933,
+      "loss": 0.5661,
+      "step": 403
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.22113107520723113,
+      "learning_rate": 0.00019321922964398046,
+      "loss": 0.5789,
+      "step": 404
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.23262880002546846,
+      "learning_rate": 0.00019318516267952874,
+      "loss": 0.5447,
+      "step": 405
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.24962941770082592,
+      "learning_rate": 0.00019315101337076792,
+      "loss": 0.5512,
+      "step": 406
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.22210049422713798,
+      "learning_rate": 0.0001931167817478745,
+      "loss": 0.5427,
+      "step": 407
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.22647809883332484,
+      "learning_rate": 0.0001930824678410977,
+      "loss": 0.4888,
+      "step": 408
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.23660763255678552,
+      "learning_rate": 0.00019304807168075944,
+      "loss": 0.5755,
+      "step": 409
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2354103448271752,
+      "learning_rate": 0.00019301359329725436,
+      "loss": 0.5265,
+      "step": 410
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.24322261128085423,
+      "learning_rate": 0.00019297903272104977,
+      "loss": 0.5291,
+      "step": 411
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.20525199182278092,
+      "learning_rate": 0.00019294438998268554,
+      "loss": 0.4996,
+      "step": 412
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.24678535182755174,
+      "learning_rate": 0.00019290966511277422,
+      "loss": 0.567,
+      "step": 413
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.22165331172413838,
+      "learning_rate": 0.00019287485814200087,
+      "loss": 0.5348,
+      "step": 414
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.24541020782476444,
+      "learning_rate": 0.00019283996910112318,
+      "loss": 0.5432,
+      "step": 415
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2255959168063083,
+      "learning_rate": 0.00019280499802097126,
+      "loss": 0.5891,
+      "step": 416
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.21159018099714821,
+      "learning_rate": 0.0001927699449324478,
+      "loss": 0.5003,
+      "step": 417
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.21379995902020923,
+      "learning_rate": 0.00019273480986652794,
+      "loss": 0.5314,
+      "step": 418
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2853169518220406,
+      "learning_rate": 0.0001926995928542592,
+      "loss": 0.6108,
+      "step": 419
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.22738285867292138,
+      "learning_rate": 0.00019266429392676164,
+      "loss": 0.5217,
+      "step": 420
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.23835369502554374,
+      "learning_rate": 0.00019262891311522755,
+      "loss": 0.5318,
+      "step": 421
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.20671557324330114,
+      "learning_rate": 0.0001925934504509217,
+      "loss": 0.5234,
+      "step": 422
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.205212164360302,
+      "learning_rate": 0.00019255790596518112,
+      "loss": 0.5023,
+      "step": 423
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.21664090577036152,
+      "learning_rate": 0.00019252227968941522,
+      "loss": 0.5452,
+      "step": 424
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.22146041084684798,
+      "learning_rate": 0.00019248657165510556,
+      "loss": 0.5474,
+      "step": 425
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2338997589574809,
+      "learning_rate": 0.00019245078189380604,
+      "loss": 0.5516,
+      "step": 426
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2313978280927526,
+      "learning_rate": 0.0001924149104371428,
+      "loss": 0.5831,
+      "step": 427
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2098577112814155,
+      "learning_rate": 0.00019237895731681408,
+      "loss": 0.5452,
+      "step": 428
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.26497439164374026,
+      "learning_rate": 0.0001923429225645904,
+      "loss": 0.5666,
+      "step": 429
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.21859970576834997,
+      "learning_rate": 0.00019230680621231425,
+      "loss": 0.5069,
+      "step": 430
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.20509380886351694,
+      "learning_rate": 0.0001922706082919004,
+      "loss": 0.4573,
+      "step": 431
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2182328366507935,
+      "learning_rate": 0.0001922343288353356,
+      "loss": 0.6133,
+      "step": 432
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2822350271273954,
+      "learning_rate": 0.00019219796787467867,
+      "loss": 0.5709,
+      "step": 433
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.24487543268473794,
+      "learning_rate": 0.00019216152544206049,
+      "loss": 0.546,
+      "step": 434
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.24221176090281485,
+      "learning_rate": 0.00019212500156968383,
+      "loss": 0.5507,
+      "step": 435
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.22053929296251015,
+      "learning_rate": 0.00019208839628982358,
+      "loss": 0.5473,
+      "step": 436
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.22975415570737245,
+      "learning_rate": 0.00019205170963482643,
+      "loss": 0.5181,
+      "step": 437
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.22969105575505203,
+      "learning_rate": 0.00019201494163711104,
+      "loss": 0.5463,
+      "step": 438
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.23764087103158363,
+      "learning_rate": 0.00019197809232916795,
+      "loss": 0.55,
+      "step": 439
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.21997498488474826,
+      "learning_rate": 0.00019194116174355954,
+      "loss": 0.5421,
+      "step": 440
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.22225824990596896,
+      "learning_rate": 0.00019190414991291998,
+      "loss": 0.5439,
+      "step": 441
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.243391488050543,
+      "learning_rate": 0.00019186705686995533,
+      "loss": 0.6289,
+      "step": 442
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.222494273038652,
+      "learning_rate": 0.0001918298826474433,
+      "loss": 0.5088,
+      "step": 443
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.22114450997419682,
+      "learning_rate": 0.0001917926272782334,
+      "loss": 0.5624,
+      "step": 444
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.21964760504534894,
+      "learning_rate": 0.00019175529079524687,
+      "loss": 0.5289,
+      "step": 445
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3042847973140014,
+      "learning_rate": 0.00019171787323147654,
+      "loss": 0.5328,
+      "step": 446
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.22425571202210934,
+      "learning_rate": 0.00019168037461998695,
+      "loss": 0.5699,
+      "step": 447
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.23406959191320909,
+      "learning_rate": 0.00019164279499391427,
+      "loss": 0.5147,
+      "step": 448
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3604500123158513,
+      "learning_rate": 0.00019160513438646617,
+      "loss": 0.5697,
+      "step": 449
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2501436029131694,
+      "learning_rate": 0.00019156739283092205,
+      "loss": 0.6015,
+      "step": 450
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.21928141490521824,
+      "learning_rate": 0.00019152957036063265,
+      "loss": 0.5111,
+      "step": 451
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.257908225365161,
+      "learning_rate": 0.00019149166700902032,
+      "loss": 0.5132,
+      "step": 452
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2713678867101362,
+      "learning_rate": 0.0001914536828095789,
+      "loss": 0.5995,
+      "step": 453
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2398794022246256,
+      "learning_rate": 0.0001914156177958736,
+      "loss": 0.4993,
+      "step": 454
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2373981477389832,
+      "learning_rate": 0.0001913774720015411,
+      "loss": 0.5064,
+      "step": 455
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2188011093608266,
+      "learning_rate": 0.00019133924546028942,
+      "loss": 0.5606,
+      "step": 456
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.24077263566935142,
+      "learning_rate": 0.00019130093820589791,
+      "loss": 0.5606,
+      "step": 457
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.23519919814487683,
+      "learning_rate": 0.00019126255027221735,
+      "loss": 0.5307,
+      "step": 458
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.21480730775028578,
+      "learning_rate": 0.00019122408169316976,
+      "loss": 0.526,
+      "step": 459
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2161668548042441,
+      "learning_rate": 0.00019118553250274832,
+      "loss": 0.5657,
+      "step": 460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.22318400428439122,
+      "learning_rate": 0.00019114690273501765,
+      "loss": 0.513,
+      "step": 461
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.22252447744680176,
+      "learning_rate": 0.00019110819242411337,
+      "loss": 0.5247,
+      "step": 462
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.21358818358042153,
+      "learning_rate": 0.00019106940160424244,
+      "loss": 0.556,
+      "step": 463
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2121229259271081,
+      "learning_rate": 0.0001910305303096828,
+      "loss": 0.5138,
+      "step": 464
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.22636146624511622,
+      "learning_rate": 0.0001909915785747836,
+      "loss": 0.5111,
+      "step": 465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.20571954917028099,
+      "learning_rate": 0.00019095254643396512,
+      "loss": 0.5125,
+      "step": 466
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.21968966730793454,
+      "learning_rate": 0.0001909134339217186,
+      "loss": 0.5358,
+      "step": 467
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.21910723327372644,
+      "learning_rate": 0.00019087424107260627,
+      "loss": 0.5382,
+      "step": 468
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2153841373499183,
+      "learning_rate": 0.00019083496792126153,
+      "loss": 0.5375,
+      "step": 469
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.23479205084160673,
+      "learning_rate": 0.00019079561450238854,
+      "loss": 0.5984,
+      "step": 470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.21595571362737268,
+      "learning_rate": 0.00019075618085076247,
+      "loss": 0.5417,
+      "step": 471
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.24550770571804625,
+      "learning_rate": 0.00019071666700122946,
+      "loss": 0.5306,
+      "step": 472
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.21802243564456578,
+      "learning_rate": 0.00019067707298870638,
+      "loss": 0.5157,
+      "step": 473
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2068796190094572,
+      "learning_rate": 0.00019063739884818103,
+      "loss": 0.5254,
+      "step": 474
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.24034732867281272,
+      "learning_rate": 0.000190597644614712,
+      "loss": 0.6204,
+      "step": 475
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2260836607650634,
+      "learning_rate": 0.00019055781032342864,
+      "loss": 0.5492,
+      "step": 476
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2476351525598878,
+      "learning_rate": 0.00019051789600953102,
+      "loss": 0.5157,
+      "step": 477
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2280151093681579,
+      "learning_rate": 0.00019047790170829003,
+      "loss": 0.4984,
+      "step": 478
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2217333524292061,
+      "learning_rate": 0.00019043782745504711,
+      "loss": 0.5149,
+      "step": 479
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2356369467654302,
+      "learning_rate": 0.00019039767328521442,
+      "loss": 0.5724,
+      "step": 480
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.21541809863677616,
+      "learning_rate": 0.0001903574392342747,
+      "loss": 0.5138,
+      "step": 481
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.21722431891543054,
+      "learning_rate": 0.00019031712533778137,
+      "loss": 0.5536,
+      "step": 482
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2370708268417489,
+      "learning_rate": 0.00019027673163135827,
+      "loss": 0.5038,
+      "step": 483
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.22809310323516838,
+      "learning_rate": 0.00019023625815069989,
+      "loss": 0.5713,
+      "step": 484
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.22374988575329294,
+      "learning_rate": 0.00019019570493157114,
+      "loss": 0.5549,
+      "step": 485
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.20510711707245072,
+      "learning_rate": 0.0001901550720098074,
+      "loss": 0.46,
+      "step": 486
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2621551195786783,
+      "learning_rate": 0.00019011435942131448,
+      "loss": 0.5546,
+      "step": 487
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.20503054358781417,
+      "learning_rate": 0.00019007356720206865,
+      "loss": 0.5547,
+      "step": 488
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.23586140447856616,
+      "learning_rate": 0.00019003269538811647,
+      "loss": 0.6075,
+      "step": 489
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2828040872125889,
+      "learning_rate": 0.00018999174401557488,
+      "loss": 0.602,
+      "step": 490
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2023429982220119,
+      "learning_rate": 0.00018995071312063105,
+      "loss": 0.4975,
+      "step": 491
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2054777673202953,
+      "learning_rate": 0.00018990960273954254,
+      "loss": 0.5295,
+      "step": 492
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1982185225446849,
+      "learning_rate": 0.00018986841290863704,
+      "loss": 0.5461,
+      "step": 493
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.23248022218099268,
+      "learning_rate": 0.0001898271436643125,
+      "loss": 0.5924,
+      "step": 494
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2235279893303581,
+      "learning_rate": 0.00018978579504303706,
+      "loss": 0.5598,
+      "step": 495
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.21675084465821123,
+      "learning_rate": 0.000189744367081349,
+      "loss": 0.5012,
+      "step": 496
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2041881848681654,
+      "learning_rate": 0.00018970285981585662,
+      "loss": 0.526,
+      "step": 497
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.23258761727278376,
+      "learning_rate": 0.00018966127328323842,
+      "loss": 0.553,
+      "step": 498
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.23066266735191,
+      "learning_rate": 0.00018961960752024288,
+      "loss": 0.5506,
+      "step": 499
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.20634958879584178,
+      "learning_rate": 0.0001895778625636885,
+      "loss": 0.5006,
+      "step": 500
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.21082421656186934,
+      "learning_rate": 0.00018953603845046378,
+      "loss": 0.5279,
+      "step": 501
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2057560041730304,
+      "learning_rate": 0.00018949413521752713,
+      "loss": 0.5598,
+      "step": 502
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2096114347066206,
+      "learning_rate": 0.00018945215290190693,
+      "loss": 0.5113,
+      "step": 503
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.23218477255443984,
+      "learning_rate": 0.00018941009154070136,
+      "loss": 0.5169,
+      "step": 504
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.20857717653678057,
+      "learning_rate": 0.00018936795117107855,
+      "loss": 0.5149,
+      "step": 505
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.24006448825761761,
+      "learning_rate": 0.0001893257318302764,
+      "loss": 0.5228,
+      "step": 506
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2146671098435255,
+      "learning_rate": 0.00018928343355560258,
+      "loss": 0.5257,
+      "step": 507
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.20608859556559073,
+      "learning_rate": 0.00018924105638443452,
+      "loss": 0.527,
+      "step": 508
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2336814919363686,
+      "learning_rate": 0.0001891986003542194,
+      "loss": 0.5461,
+      "step": 509
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2409130946928026,
+      "learning_rate": 0.00018915606550247397,
+      "loss": 0.5493,
+      "step": 510
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.21371348825911873,
+      "learning_rate": 0.0001891134518667848,
+      "loss": 0.572,
+      "step": 511
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2014364828041311,
+      "learning_rate": 0.000189070759484808,
+      "loss": 0.5109,
+      "step": 512
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2290945612613713,
+      "learning_rate": 0.0001890279883942692,
+      "loss": 0.5493,
+      "step": 513
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.22127732127756986,
+      "learning_rate": 0.0001889851386329637,
+      "loss": 0.5387,
+      "step": 514
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.20564079598559082,
+      "learning_rate": 0.00018894221023875622,
+      "loss": 0.5192,
+      "step": 515
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.213993086214796,
+      "learning_rate": 0.00018889920324958106,
+      "loss": 0.5044,
+      "step": 516
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.21506249939577854,
+      "learning_rate": 0.00018885611770344185,
+      "loss": 0.4969,
+      "step": 517
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.22792164808811663,
+      "learning_rate": 0.00018881295363841174,
+      "loss": 0.5564,
+      "step": 518
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1978731923118128,
+      "learning_rate": 0.00018876971109263324,
+      "loss": 0.4898,
+      "step": 519
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.22394451521984352,
+      "learning_rate": 0.00018872639010431822,
+      "loss": 0.5586,
+      "step": 520
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.20009625678598073,
+      "learning_rate": 0.0001886829907117478,
+      "loss": 0.5399,
+      "step": 521
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.20448355507434923,
+      "learning_rate": 0.00018863951295327244,
+      "loss": 0.5263,
+      "step": 522
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1967777231547204,
+      "learning_rate": 0.00018859595686731187,
+      "loss": 0.4904,
+      "step": 523
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2052388343929957,
+      "learning_rate": 0.00018855232249235498,
+      "loss": 0.4951,
+      "step": 524
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1970956590240829,
+      "learning_rate": 0.00018850860986695985,
+      "loss": 0.5112,
+      "step": 525
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2102143499682878,
+      "learning_rate": 0.00018846481902975377,
+      "loss": 0.5234,
+      "step": 526
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.23384214794287286,
+      "learning_rate": 0.00018842095001943306,
+      "loss": 0.5387,
+      "step": 527
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.20133953340775343,
+      "learning_rate": 0.00018837700287476316,
+      "loss": 0.4995,
+      "step": 528
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2238467486071384,
+      "learning_rate": 0.00018833297763457858,
+      "loss": 0.5709,
+      "step": 529
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.26170161234282546,
+      "learning_rate": 0.00018828887433778278,
+      "loss": 0.6314,
+      "step": 530
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2317819906199683,
+      "learning_rate": 0.00018824469302334822,
+      "loss": 0.5333,
+      "step": 531
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.21538390925414544,
+      "learning_rate": 0.0001882004337303163,
+      "loss": 0.5603,
+      "step": 532
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.23053571801246284,
+      "learning_rate": 0.0001881560964977974,
+      "loss": 0.593,
+      "step": 533
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.21173642276584706,
+      "learning_rate": 0.0001881116813649706,
+      "loss": 0.5539,
+      "step": 534
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.24587290888576793,
+      "learning_rate": 0.00018806718837108402,
+      "loss": 0.5408,
+      "step": 535
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.22324082101473863,
+      "learning_rate": 0.00018802261755545443,
+      "loss": 0.5857,
+      "step": 536
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.21827653101692504,
+      "learning_rate": 0.0001879779689574674,
+      "loss": 0.5451,
+      "step": 537
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2146222856243753,
+      "learning_rate": 0.00018793324261657737,
+      "loss": 0.5007,
+      "step": 538
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.20994383183759666,
+      "learning_rate": 0.00018788843857230726,
+      "loss": 0.5039,
+      "step": 539
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.23384168426304514,
+      "learning_rate": 0.00018784355686424876,
+      "loss": 0.5329,
+      "step": 540
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.20284382518697272,
+      "learning_rate": 0.00018779859753206225,
+      "loss": 0.5383,
+      "step": 541
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.22307014132513725,
+      "learning_rate": 0.00018775356061547662,
+      "loss": 0.5766,
+      "step": 542
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.21675879523474215,
+      "learning_rate": 0.00018770844615428932,
+      "loss": 0.4994,
+      "step": 543
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2200785983728407,
+      "learning_rate": 0.00018766325418836637,
+      "loss": 0.5615,
+      "step": 544
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.20895654400479502,
+      "learning_rate": 0.00018761798475764224,
+      "loss": 0.4993,
+      "step": 545
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.22152937631276676,
+      "learning_rate": 0.00018757263790211988,
+      "loss": 0.5275,
+      "step": 546
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.209333487906431,
+      "learning_rate": 0.0001875272136618706,
+      "loss": 0.4911,
+      "step": 547
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2123519912763275,
+      "learning_rate": 0.00018748171207703417,
+      "loss": 0.5662,
+      "step": 548
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2147346642469028,
+      "learning_rate": 0.00018743613318781868,
+      "loss": 0.5651,
+      "step": 549
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2017789732342509,
+      "learning_rate": 0.00018739047703450048,
+      "loss": 0.5573,
+      "step": 550
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2084087089737107,
+      "learning_rate": 0.00018734474365742428,
+      "loss": 0.562,
+      "step": 551
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.22130968599178,
+      "learning_rate": 0.00018729893309700295,
+      "loss": 0.5729,
+      "step": 552
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.22736172090948445,
+      "learning_rate": 0.0001872530453937176,
+      "loss": 0.5548,
+      "step": 553
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.21738577850339916,
+      "learning_rate": 0.0001872070805881176,
+      "loss": 0.5191,
+      "step": 554
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.20994273135857797,
+      "learning_rate": 0.00018716103872082026,
+      "loss": 0.5153,
+      "step": 555
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.25944295362906805,
+      "learning_rate": 0.00018711491983251113,
+      "loss": 0.5471,
+      "step": 556
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2138519097360962,
+      "learning_rate": 0.00018706872396394376,
+      "loss": 0.4875,
+      "step": 557
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.23586915663527888,
+      "learning_rate": 0.00018702245115593974,
+      "loss": 0.5224,
+      "step": 558
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.20477148046499385,
+      "learning_rate": 0.0001869761014493887,
+      "loss": 0.5466,
+      "step": 559
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.21783175505387284,
+      "learning_rate": 0.00018692967488524812,
+      "loss": 0.5557,
+      "step": 560
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.20442177589984145,
+      "learning_rate": 0.0001868831715045435,
+      "loss": 0.507,
+      "step": 561
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.21291324212369495,
+      "learning_rate": 0.00018683659134836813,
+      "loss": 0.5779,
+      "step": 562
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.22670486875141618,
+      "learning_rate": 0.00018678993445788323,
+      "loss": 0.5831,
+      "step": 563
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2431493116309222,
+      "learning_rate": 0.00018674320087431768,
+      "loss": 0.5389,
+      "step": 564
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.22102091260855142,
+      "learning_rate": 0.00018669639063896836,
+      "loss": 0.5569,
+      "step": 565
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.20001951850669827,
+      "learning_rate": 0.0001866495037931997,
+      "loss": 0.486,
+      "step": 566
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.22781103196427857,
+      "learning_rate": 0.00018660254037844388,
+      "loss": 0.4973,
+      "step": 567
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.21129685691062433,
+      "learning_rate": 0.00018655550043620073,
+      "loss": 0.5459,
+      "step": 568
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.20363805081315986,
+      "learning_rate": 0.0001865083840080378,
+      "loss": 0.4997,
+      "step": 569
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.22269838654252982,
+      "learning_rate": 0.00018646119113559006,
+      "loss": 0.5406,
+      "step": 570
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.20307002281681275,
+      "learning_rate": 0.00018641392186056016,
+      "loss": 0.4861,
+      "step": 571
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.20146261628709675,
+      "learning_rate": 0.0001863665762247182,
+      "loss": 0.561,
+      "step": 572
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.21049257054009352,
+      "learning_rate": 0.00018631915426990184,
+      "loss": 0.5257,
+      "step": 573
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2245482792823418,
+      "learning_rate": 0.00018627165603801605,
+      "loss": 0.5441,
+      "step": 574
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2106578436256788,
+      "learning_rate": 0.0001862240815710333,
+      "loss": 0.5125,
+      "step": 575
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2091435884054145,
+      "learning_rate": 0.0001861764309109934,
+      "loss": 0.523,
+      "step": 576
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.21256854318600532,
+      "learning_rate": 0.00018612870410000354,
+      "loss": 0.4851,
+      "step": 577
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.24387962798982954,
+      "learning_rate": 0.00018608090118023808,
+      "loss": 0.5423,
+      "step": 578
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2357478920855788,
+      "learning_rate": 0.00018603302219393874,
+      "loss": 0.5386,
+      "step": 579
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.21267780857117077,
+      "learning_rate": 0.0001859850671834144,
+      "loss": 0.5545,
+      "step": 580
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.25049614581715324,
+      "learning_rate": 0.0001859370361910412,
+      "loss": 0.5241,
+      "step": 581
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1937807494598699,
+      "learning_rate": 0.00018588892925926228,
+      "loss": 0.5533,
+      "step": 582
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.21209972240968475,
+      "learning_rate": 0.00018584074643058807,
+      "loss": 0.538,
+      "step": 583
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.22281277082523665,
+      "learning_rate": 0.00018579248774759586,
+      "loss": 0.5456,
+      "step": 584
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.22156542955128883,
+      "learning_rate": 0.00018574415325293018,
+      "loss": 0.5622,
+      "step": 585
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.20068342929250654,
+      "learning_rate": 0.00018569574298930237,
+      "loss": 0.5372,
+      "step": 586
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.21693418845369525,
+      "learning_rate": 0.00018564725699949083,
+      "loss": 0.4874,
+      "step": 587
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2060622909003744,
+      "learning_rate": 0.0001855986953263409,
+      "loss": 0.5331,
+      "step": 588
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.20007419545283933,
+      "learning_rate": 0.00018555005801276463,
+      "loss": 0.5131,
+      "step": 589
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.21905328017125653,
+      "learning_rate": 0.00018550134510174115,
+      "loss": 0.5572,
+      "step": 590
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.21213287568506015,
+      "learning_rate": 0.0001854525566363162,
+      "loss": 0.5359,
+      "step": 591
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.20066093050756748,
+      "learning_rate": 0.00018540369265960242,
+      "loss": 0.5334,
+      "step": 592
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2068811720002483,
+      "learning_rate": 0.00018535475321477906,
+      "loss": 0.5558,
+      "step": 593
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2025287668887073,
+      "learning_rate": 0.00018530573834509215,
+      "loss": 0.5098,
+      "step": 594
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.20807380346718593,
+      "learning_rate": 0.0001852566480938543,
+      "loss": 0.5211,
+      "step": 595
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2049943719782544,
+      "learning_rate": 0.00018520748250444474,
+      "loss": 0.5379,
+      "step": 596
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8558508208219735,
+      "learning_rate": 0.00018515824162030934,
+      "loss": 0.5403,
+      "step": 597
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25414317775682305,
+      "learning_rate": 0.00018510892548496047,
+      "loss": 0.5804,
+      "step": 598
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.20806597400748386,
+      "learning_rate": 0.00018505953414197696,
+      "loss": 0.5419,
+      "step": 599
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1950528976937739,
+      "learning_rate": 0.00018501006763500414,
+      "loss": 0.4956,
+      "step": 600
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.20652545713558523,
+      "learning_rate": 0.00018496052600775376,
+      "loss": 0.4942,
+      "step": 601
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.20955886781649663,
+      "learning_rate": 0.0001849109093040039,
+      "loss": 0.5177,
+      "step": 602
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.21093362015684414,
+      "learning_rate": 0.00018486121756759906,
+      "loss": 0.5672,
+      "step": 603
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.22033088091533184,
+      "learning_rate": 0.00018481145084245002,
+      "loss": 0.5691,
+      "step": 604
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.20322111965044637,
+      "learning_rate": 0.00018476160917253373,
+      "loss": 0.5425,
+      "step": 605
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2028788101278272,
+      "learning_rate": 0.0001847116926018935,
+      "loss": 0.5176,
+      "step": 606
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.19551140156538951,
+      "learning_rate": 0.0001846617011746388,
+      "loss": 0.5115,
+      "step": 607
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.21944694996534547,
+      "learning_rate": 0.00018461163493494517,
+      "loss": 0.5496,
+      "step": 608
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.21506814147924705,
+      "learning_rate": 0.0001845614939270543,
+      "loss": 0.5823,
+      "step": 609
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2220938137588105,
+      "learning_rate": 0.00018451127819527402,
+      "loss": 0.5731,
+      "step": 610
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.21590208362786933,
+      "learning_rate": 0.00018446098778397807,
+      "loss": 0.6063,
+      "step": 611
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.20084594317065918,
+      "learning_rate": 0.00018441062273760628,
+      "loss": 0.5286,
+      "step": 612
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.21847304705653886,
+      "learning_rate": 0.00018436018310066435,
+      "loss": 0.5721,
+      "step": 613
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2467936487351411,
+      "learning_rate": 0.000184309668917724,
+      "loss": 0.571,
+      "step": 614
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.21666156526926003,
+      "learning_rate": 0.0001842590802334227,
+      "loss": 0.5244,
+      "step": 615
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.21336859433357677,
+      "learning_rate": 0.00018420841709246383,
+      "loss": 0.5724,
+      "step": 616
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1933070755110986,
+      "learning_rate": 0.0001841576795396166,
+      "loss": 0.5347,
+      "step": 617
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2332186369470874,
+      "learning_rate": 0.00018410686761971586,
+      "loss": 0.5474,
+      "step": 618
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1996293438855639,
+      "learning_rate": 0.00018405598137766224,
+      "loss": 0.5421,
+      "step": 619
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2012759807756364,
+      "learning_rate": 0.00018400502085842208,
+      "loss": 0.519,
+      "step": 620
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24355300568180752,
+      "learning_rate": 0.00018395398610702733,
+      "loss": 0.597,
+      "step": 621
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2136711983483761,
+      "learning_rate": 0.00018390287716857546,
+      "loss": 0.5398,
+      "step": 622
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.22275088525970024,
+      "learning_rate": 0.00018385169408822964,
+      "loss": 0.5597,
+      "step": 623
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.20011931485707388,
+      "learning_rate": 0.0001838004369112184,
+      "loss": 0.4901,
+      "step": 624
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.19544716159187206,
+      "learning_rate": 0.00018374910568283594,
+      "loss": 0.4726,
+      "step": 625
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2176067620544374,
+      "learning_rate": 0.00018369770044844168,
+      "loss": 0.5369,
+      "step": 626
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2005629047810257,
+      "learning_rate": 0.00018364622125346055,
+      "loss": 0.4914,
+      "step": 627
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.21497281608823432,
+      "learning_rate": 0.0001835946681433829,
+      "loss": 0.5559,
+      "step": 628
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.20354723273049724,
+      "learning_rate": 0.00018354304116376425,
+      "loss": 0.5083,
+      "step": 629
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.23536026550959782,
+      "learning_rate": 0.0001834913403602255,
+      "loss": 0.5449,
+      "step": 630
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.20887211237530257,
+      "learning_rate": 0.00018343956577845276,
+      "loss": 0.5131,
+      "step": 631
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.21728763678777088,
+      "learning_rate": 0.00018338771746419726,
+      "loss": 0.5484,
+      "step": 632
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.21910570476522437,
+      "learning_rate": 0.00018333579546327556,
+      "loss": 0.5452,
+      "step": 633
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.21247350127543838,
+      "learning_rate": 0.00018328379982156915,
+      "loss": 0.5232,
+      "step": 634
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.21706686115456897,
+      "learning_rate": 0.00018323173058502472,
+      "loss": 0.5353,
+      "step": 635
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.19529494853666482,
+      "learning_rate": 0.00018317958779965387,
+      "loss": 0.4611,
+      "step": 636
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2194890381897013,
+      "learning_rate": 0.00018312737151153334,
+      "loss": 0.4884,
+      "step": 637
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24336065627870296,
+      "learning_rate": 0.00018307508176680472,
+      "loss": 0.5708,
+      "step": 638
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.22638828434923797,
+      "learning_rate": 0.00018302271861167456,
+      "loss": 0.5795,
+      "step": 639
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.20501380703607638,
+      "learning_rate": 0.0001829702820924142,
+      "loss": 0.5645,
+      "step": 640
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.22705979847255006,
+      "learning_rate": 0.00018291777225535994,
+      "loss": 0.4974,
+      "step": 641
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.22629645320684777,
+      "learning_rate": 0.00018286518914691272,
+      "loss": 0.5587,
+      "step": 642
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.21772563640763765,
+      "learning_rate": 0.00018281253281353838,
+      "loss": 0.5219,
+      "step": 643
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.20447194133414195,
+      "learning_rate": 0.00018275980330176737,
+      "loss": 0.5425,
+      "step": 644
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24126870503035064,
+      "learning_rate": 0.00018270700065819477,
+      "loss": 0.5119,
+      "step": 645
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.23269297218381896,
+      "learning_rate": 0.00018265412492948042,
+      "loss": 0.5507,
+      "step": 646
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.23416570398912578,
+      "learning_rate": 0.0001826011761623486,
+      "loss": 0.5947,
+      "step": 647
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2186560086983282,
+      "learning_rate": 0.0001825481544035882,
+      "loss": 0.5204,
+      "step": 648
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.20624707271501935,
+      "learning_rate": 0.00018249505970005262,
+      "loss": 0.4785,
+      "step": 649
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.23418189558532218,
+      "learning_rate": 0.00018244189209865974,
+      "loss": 0.4976,
+      "step": 650
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.21372290734059424,
+      "learning_rate": 0.00018238865164639173,
+      "loss": 0.5237,
+      "step": 651
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1986689651795865,
+      "learning_rate": 0.0001823353383902953,
+      "loss": 0.5354,
+      "step": 652
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.21154599437074698,
+      "learning_rate": 0.0001822819523774814,
+      "loss": 0.5292,
+      "step": 653
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.21348268586605149,
+      "learning_rate": 0.00018222849365512523,
+      "loss": 0.5249,
+      "step": 654
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.22296243039072478,
+      "learning_rate": 0.0001821749622704664,
+      "loss": 0.5458,
+      "step": 655
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.22596567506529938,
+      "learning_rate": 0.00018212135827080857,
+      "loss": 0.5085,
+      "step": 656
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.19012132806019622,
+      "learning_rate": 0.00018206768170351962,
+      "loss": 0.4977,
+      "step": 657
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2125366600531234,
+      "learning_rate": 0.0001820139326160316,
+      "loss": 0.5051,
+      "step": 658
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.23677835317412968,
+      "learning_rate": 0.00018196011105584058,
+      "loss": 0.575,
+      "step": 659
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2262210065848097,
+      "learning_rate": 0.00018190621707050671,
+      "loss": 0.5744,
+      "step": 660
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.21618545867420894,
+      "learning_rate": 0.0001818522507076541,
+      "loss": 0.5715,
+      "step": 661
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2050215711297079,
+      "learning_rate": 0.00018179821201497092,
+      "loss": 0.5201,
+      "step": 662
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.20218467055707082,
+      "learning_rate": 0.0001817441010402091,
+      "loss": 0.5058,
+      "step": 663
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.20940987275867923,
+      "learning_rate": 0.00018168991783118452,
+      "loss": 0.5095,
+      "step": 664
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.21341822518403558,
+      "learning_rate": 0.00018163566243577697,
+      "loss": 0.5599,
+      "step": 665
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.20028205017927186,
+      "learning_rate": 0.0001815813349019299,
+      "loss": 0.5318,
+      "step": 666
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.20184912350066175,
+      "learning_rate": 0.00018152693527765057,
+      "loss": 0.5643,
+      "step": 667
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.20882160405967118,
+      "learning_rate": 0.0001814724636110099,
+      "loss": 0.542,
+      "step": 668
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.20252144356881077,
+      "learning_rate": 0.00018141791995014255,
+      "loss": 0.4496,
+      "step": 669
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1956328371434174,
+      "learning_rate": 0.00018136330434324674,
+      "loss": 0.56,
+      "step": 670
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.20691128111503362,
+      "learning_rate": 0.00018130861683858426,
+      "loss": 0.5726,
+      "step": 671
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2258004454621585,
+      "learning_rate": 0.00018125385748448048,
+      "loss": 0.583,
+      "step": 672
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.22330059205477634,
+      "learning_rate": 0.00018119902632932416,
+      "loss": 0.5288,
+      "step": 673
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.20473079466150892,
+      "learning_rate": 0.0001811441234215677,
+      "loss": 0.5085,
+      "step": 674
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.19439333859223318,
+      "learning_rate": 0.0001810891488097267,
+      "loss": 0.5147,
+      "step": 675
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2037181989313857,
+      "learning_rate": 0.00018103410254238021,
+      "loss": 0.5228,
+      "step": 676
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.21580635559566858,
+      "learning_rate": 0.0001809789846681706,
+      "loss": 0.5034,
+      "step": 677
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.21490060304667385,
+      "learning_rate": 0.00018092379523580357,
+      "loss": 0.5347,
+      "step": 678
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.20927738857723482,
+      "learning_rate": 0.00018086853429404793,
+      "loss": 0.5039,
+      "step": 679
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.21391199422702836,
+      "learning_rate": 0.00018081320189173577,
+      "loss": 0.5404,
+      "step": 680
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.22355130583819918,
+      "learning_rate": 0.0001807577980777623,
+      "loss": 0.5147,
+      "step": 681
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.21899190720848985,
+      "learning_rate": 0.00018070232290108584,
+      "loss": 0.5195,
+      "step": 682
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.20636096645560792,
+      "learning_rate": 0.00018064677641072775,
+      "loss": 0.5158,
+      "step": 683
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.20462410706105155,
+      "learning_rate": 0.00018059115865577249,
+      "loss": 0.5194,
+      "step": 684
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.21978634315593423,
+      "learning_rate": 0.00018053546968536735,
+      "loss": 0.4986,
+      "step": 685
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2203882917140438,
+      "learning_rate": 0.00018047970954872264,
+      "loss": 0.5855,
+      "step": 686
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.20144829000454462,
+      "learning_rate": 0.0001804238782951116,
+      "loss": 0.5212,
+      "step": 687
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.21142991796239274,
+      "learning_rate": 0.00018036797597387023,
+      "loss": 0.495,
+      "step": 688
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.21275432668758548,
+      "learning_rate": 0.00018031200263439736,
+      "loss": 0.5694,
+      "step": 689
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2035189446424034,
+      "learning_rate": 0.00018025595832615459,
+      "loss": 0.55,
+      "step": 690
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.20030837247360464,
+      "learning_rate": 0.00018019984309866619,
+      "loss": 0.4748,
+      "step": 691
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.20366715425572,
+      "learning_rate": 0.00018014365700151912,
+      "loss": 0.5792,
+      "step": 692
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2082468197583491,
+      "learning_rate": 0.000180087400084363,
+      "loss": 0.4973,
+      "step": 693
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.21820027454676755,
+      "learning_rate": 0.00018003107239691004,
+      "loss": 0.5512,
+      "step": 694
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2085678250499903,
+      "learning_rate": 0.00017997467398893488,
+      "loss": 0.5148,
+      "step": 695
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.20422653056781329,
+      "learning_rate": 0.00017991820491027472,
+      "loss": 0.4968,
+      "step": 696
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1875899162050169,
+      "learning_rate": 0.0001798616652108293,
+      "loss": 0.5061,
+      "step": 697
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.20869663705218836,
+      "learning_rate": 0.00017980505494056062,
+      "loss": 0.5182,
+      "step": 698
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.19250179476147736,
+      "learning_rate": 0.00017974837414949307,
+      "loss": 0.5184,
+      "step": 699
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.21732108838463451,
+      "learning_rate": 0.00017969162288771347,
+      "loss": 0.5524,
+      "step": 700
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.20200315361578813,
+      "learning_rate": 0.0001796348012053707,
+      "loss": 0.5386,
+      "step": 701
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.20242537832049035,
+      "learning_rate": 0.00017957790915267615,
+      "loss": 0.5656,
+      "step": 702
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1889172192023988,
+      "learning_rate": 0.0001795209467799031,
+      "loss": 0.5115,
+      "step": 703
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.19623435201373893,
+      "learning_rate": 0.0001794639141373872,
+      "loss": 0.497,
+      "step": 704
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.22372809637554478,
+      "learning_rate": 0.00017940681127552604,
+      "loss": 0.5579,
+      "step": 705
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1968536923376666,
+      "learning_rate": 0.0001793496382447794,
+      "loss": 0.4891,
+      "step": 706
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1990723573146364,
+      "learning_rate": 0.00017929239509566894,
+      "loss": 0.5921,
+      "step": 707
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.20388703819339077,
+      "learning_rate": 0.00017923508187877834,
+      "loss": 0.5414,
+      "step": 708
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.23657852979478725,
+      "learning_rate": 0.00017917769864475314,
+      "loss": 0.5672,
+      "step": 709
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.22888252332289927,
+      "learning_rate": 0.00017912024544430088,
+      "loss": 0.5459,
+      "step": 710
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.19383907969249117,
+      "learning_rate": 0.0001790627223281908,
+      "loss": 0.5509,
+      "step": 711
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2154263629956836,
+      "learning_rate": 0.00017900512934725397,
+      "loss": 0.5629,
+      "step": 712
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.19802419635693494,
+      "learning_rate": 0.0001789474665523832,
+      "loss": 0.5128,
+      "step": 713
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.19783321602266912,
+      "learning_rate": 0.00017888973399453296,
+      "loss": 0.5064,
+      "step": 714
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.19864882371614528,
+      "learning_rate": 0.00017883193172471944,
+      "loss": 0.5458,
+      "step": 715
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.23609512585527,
+      "learning_rate": 0.00017877405979402038,
+      "loss": 0.5069,
+      "step": 716
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.19894144678524353,
+      "learning_rate": 0.00017871611825357502,
+      "loss": 0.5812,
+      "step": 717
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.19598819977852033,
+      "learning_rate": 0.00017865810715458427,
+      "loss": 0.5223,
+      "step": 718
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.23274847505011953,
+      "learning_rate": 0.00017860002654831032,
+      "loss": 0.5703,
+      "step": 719
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.19794477486450376,
+      "learning_rate": 0.00017854187648607694,
+      "loss": 0.5538,
+      "step": 720
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2091737019131215,
+      "learning_rate": 0.00017848365701926913,
+      "loss": 0.4962,
+      "step": 721
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.21890749511490995,
+      "learning_rate": 0.00017842536819933337,
+      "loss": 0.5074,
+      "step": 722
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.22746821737803938,
+      "learning_rate": 0.0001783670100777773,
+      "loss": 0.5849,
+      "step": 723
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.20967916540656184,
+      "learning_rate": 0.0001783085827061699,
+      "loss": 0.5246,
+      "step": 724
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.19798059353181535,
+      "learning_rate": 0.00017825008613614127,
+      "loss": 0.4667,
+      "step": 725
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1992664047298655,
+      "learning_rate": 0.00017819152041938265,
+      "loss": 0.5247,
+      "step": 726
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.22025628624147217,
+      "learning_rate": 0.00017813288560764647,
+      "loss": 0.5291,
+      "step": 727
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.20405038516624363,
+      "learning_rate": 0.00017807418175274612,
+      "loss": 0.5235,
+      "step": 728
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.20626127985692586,
+      "learning_rate": 0.00017801540890655609,
+      "loss": 0.5103,
+      "step": 729
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2187527308725265,
+      "learning_rate": 0.00017795656712101172,
+      "loss": 0.5515,
+      "step": 730
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.20386714530070776,
+      "learning_rate": 0.00017789765644810935,
+      "loss": 0.5109,
+      "step": 731
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1990293686392052,
+      "learning_rate": 0.00017783867693990624,
+      "loss": 0.5208,
+      "step": 732
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.19601721442767256,
+      "learning_rate": 0.0001777796286485204,
+      "loss": 0.5318,
+      "step": 733
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.20542580410660244,
+      "learning_rate": 0.0001777205116261306,
+      "loss": 0.5198,
+      "step": 734
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.20998518101289002,
+      "learning_rate": 0.0001776613259249764,
+      "loss": 0.5384,
+      "step": 735
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.20134476803418952,
+      "learning_rate": 0.00017760207159735805,
+      "loss": 0.5448,
+      "step": 736
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.22396912180134018,
+      "learning_rate": 0.00017754274869563637,
+      "loss": 0.59,
+      "step": 737
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2044555533666512,
+      "learning_rate": 0.00017748335727223294,
+      "loss": 0.5152,
+      "step": 738
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2106748606361736,
+      "learning_rate": 0.00017742389737962966,
+      "loss": 0.5233,
+      "step": 739
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.19348093577281505,
+      "learning_rate": 0.0001773643690703691,
+      "loss": 0.5181,
+      "step": 740
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.20393569458294794,
+      "learning_rate": 0.00017730477239705428,
+      "loss": 0.5671,
+      "step": 741
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.19728761757057783,
+      "learning_rate": 0.00017724510741234858,
+      "loss": 0.4919,
+      "step": 742
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2025575313201386,
+      "learning_rate": 0.0001771853741689757,
+      "loss": 0.5452,
+      "step": 743
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.19153867099886435,
+      "learning_rate": 0.0001771255727197198,
+      "loss": 0.4951,
+      "step": 744
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2220125331576081,
+      "learning_rate": 0.00017706570311742516,
+      "loss": 0.5521,
+      "step": 745
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.22704666961693065,
+      "learning_rate": 0.0001770057654149964,
+      "loss": 0.5184,
+      "step": 746
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.20871880228168335,
+      "learning_rate": 0.00017694575966539823,
+      "loss": 0.5205,
+      "step": 747
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2105924919088961,
+      "learning_rate": 0.00017688568592165552,
+      "loss": 0.5448,
+      "step": 748
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.19780662201378688,
+      "learning_rate": 0.00017682554423685329,
+      "loss": 0.6037,
+      "step": 749
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.23105576261963792,
+      "learning_rate": 0.0001767653346641365,
+      "loss": 0.7225,
+      "step": 750
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.21997563912032173,
+      "learning_rate": 0.00017670505725671013,
+      "loss": 0.552,
+      "step": 751
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2033859052649398,
+      "learning_rate": 0.00017664471206783915,
+      "loss": 0.5315,
+      "step": 752
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.19979214467824102,
+      "learning_rate": 0.00017658429915084835,
+      "loss": 0.5697,
+      "step": 753
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.20567412732571028,
+      "learning_rate": 0.00017652381855912247,
+      "loss": 0.5051,
+      "step": 754
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.20563597140752976,
+      "learning_rate": 0.0001764632703461059,
+      "loss": 0.5141,
+      "step": 755
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1979658869623221,
+      "learning_rate": 0.00017640265456530293,
+      "loss": 0.5257,
+      "step": 756
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2241077787834463,
+      "learning_rate": 0.0001763419712702775,
+      "loss": 0.5203,
+      "step": 757
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2197932846972142,
+      "learning_rate": 0.00017628122051465322,
+      "loss": 0.5847,
+      "step": 758
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1990944255813207,
+      "learning_rate": 0.00017622040235211326,
+      "loss": 0.4962,
+      "step": 759
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.22179309744704687,
+      "learning_rate": 0.00017615951683640045,
+      "loss": 0.5635,
+      "step": 760
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.20505896786344424,
+      "learning_rate": 0.00017609856402131703,
+      "loss": 0.4968,
+      "step": 761
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.21771157401053975,
+      "learning_rate": 0.00017603754396072483,
+      "loss": 0.4858,
+      "step": 762
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.23357401076131715,
+      "learning_rate": 0.000175976456708545,
+      "loss": 0.5766,
+      "step": 763
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.21488993425737504,
+      "learning_rate": 0.0001759153023187581,
+      "loss": 0.5419,
+      "step": 764
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2035555999534868,
+      "learning_rate": 0.00017585408084540405,
+      "loss": 0.5272,
+      "step": 765
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.20066829451010718,
+      "learning_rate": 0.00017579279234258198,
+      "loss": 0.5013,
+      "step": 766
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2052255359730049,
+      "learning_rate": 0.00017573143686445034,
+      "loss": 0.5383,
+      "step": 767
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.19180058672325329,
+      "learning_rate": 0.00017567001446522665,
+      "loss": 0.5108,
+      "step": 768
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.22862029943228582,
+      "learning_rate": 0.0001756085251991877,
+      "loss": 0.531,
+      "step": 769
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2180888066741993,
+      "learning_rate": 0.00017554696912066924,
+      "loss": 0.5938,
+      "step": 770
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.19823263656993223,
+      "learning_rate": 0.00017548534628406616,
+      "loss": 0.5158,
+      "step": 771
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18700255356016454,
+      "learning_rate": 0.00017542365674383227,
+      "loss": 0.517,
+      "step": 772
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.22948411236460914,
+      "learning_rate": 0.00017536190055448037,
+      "loss": 0.5464,
+      "step": 773
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.21370070213829387,
+      "learning_rate": 0.00017530007777058213,
+      "loss": 0.5158,
+      "step": 774
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.19174674116457566,
+      "learning_rate": 0.0001752381884467681,
+      "loss": 0.5035,
+      "step": 775
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.19069115110218368,
+      "learning_rate": 0.00017517623263772758,
+      "loss": 0.5341,
+      "step": 776
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2401612943495333,
+      "learning_rate": 0.00017511421039820863,
+      "loss": 0.578,
+      "step": 777
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.20371659209716964,
+      "learning_rate": 0.00017505212178301805,
+      "loss": 0.5103,
+      "step": 778
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2029847143168681,
+      "learning_rate": 0.00017498996684702132,
+      "loss": 0.537,
+      "step": 779
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1904915669257304,
+      "learning_rate": 0.00017492774564514235,
+      "loss": 0.5129,
+      "step": 780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.20640027525482552,
+      "learning_rate": 0.00017486545823236385,
+      "loss": 0.5585,
+      "step": 781
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.23582084058854208,
+      "learning_rate": 0.00017480310466372686,
+      "loss": 0.5648,
+      "step": 782
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2219618762642625,
+      "learning_rate": 0.00017474068499433098,
+      "loss": 0.5365,
+      "step": 783
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2021980496149104,
+      "learning_rate": 0.00017467819927933416,
+      "loss": 0.5232,
+      "step": 784
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.22350007413119136,
+      "learning_rate": 0.00017461564757395272,
+      "loss": 0.571,
+      "step": 785
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1982267515659923,
+      "learning_rate": 0.00017455302993346134,
+      "loss": 0.5228,
+      "step": 786
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1981641437638338,
+      "learning_rate": 0.00017449034641319288,
+      "loss": 0.5233,
+      "step": 787
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.21019781303279997,
+      "learning_rate": 0.00017442759706853855,
+      "loss": 0.5207,
+      "step": 788
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2060312831458839,
+      "learning_rate": 0.00017436478195494756,
+      "loss": 0.5262,
+      "step": 789
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.21829001169718656,
+      "learning_rate": 0.00017430190112792737,
+      "loss": 0.563,
+      "step": 790
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18511782073951058,
+      "learning_rate": 0.00017423895464304342,
+      "loss": 0.5017,
+      "step": 791
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1883852134929889,
+      "learning_rate": 0.00017417594255591927,
+      "loss": 0.4598,
+      "step": 792
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18093236424530848,
+      "learning_rate": 0.00017411286492223632,
+      "loss": 0.4834,
+      "step": 793
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18428120434597678,
+      "learning_rate": 0.000174049721797734,
+      "loss": 0.5032,
+      "step": 794
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.20829275110131446,
+      "learning_rate": 0.00017398651323820958,
+      "loss": 0.5844,
+      "step": 795
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.20484798677763622,
+      "learning_rate": 0.00017392323929951812,
+      "loss": 0.5674,
+      "step": 796
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.24390628538267795,
+      "learning_rate": 0.0001738599000375725,
+      "loss": 0.5415,
+      "step": 797
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7159631217821198,
+      "learning_rate": 0.00017379649550834327,
+      "loss": 0.5248,
+      "step": 798
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2153929799459398,
+      "learning_rate": 0.00017373302576785874,
+      "loss": 0.5362,
+      "step": 799
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18434502268826083,
+      "learning_rate": 0.00017366949087220472,
+      "loss": 0.5179,
+      "step": 800
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1993458339623336,
+      "learning_rate": 0.0001736058908775247,
+      "loss": 0.5378,
+      "step": 801
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.20482117487035287,
+      "learning_rate": 0.0001735422258400197,
+      "loss": 0.5066,
+      "step": 802
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.22481951556352617,
+      "learning_rate": 0.0001734784958159481,
+      "loss": 0.5504,
+      "step": 803
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.20893222575857784,
+      "learning_rate": 0.00017341470086162586,
+      "loss": 0.5558,
+      "step": 804
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.21011978723049574,
+      "learning_rate": 0.0001733508410334262,
+      "loss": 0.5164,
+      "step": 805
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.19493427746334713,
+      "learning_rate": 0.0001732869163877797,
+      "loss": 0.4928,
+      "step": 806
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.21183476672026114,
+      "learning_rate": 0.00017322292698117425,
+      "loss": 0.539,
+      "step": 807
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.19833380077404217,
+      "learning_rate": 0.00017315887287015492,
+      "loss": 0.5271,
+      "step": 808
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1914374518219283,
+      "learning_rate": 0.000173094754111324,
+      "loss": 0.5408,
+      "step": 809
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2086814492670768,
+      "learning_rate": 0.00017303057076134085,
+      "loss": 0.5289,
+      "step": 810
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.20957903788826676,
+      "learning_rate": 0.000172966322876922,
+      "loss": 0.4998,
+      "step": 811
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.20998255172298386,
+      "learning_rate": 0.00017290201051484085,
+      "loss": 0.5481,
+      "step": 812
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2071975609134585,
+      "learning_rate": 0.00017283763373192798,
+      "loss": 0.5183,
+      "step": 813
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.21738328519054306,
+      "learning_rate": 0.00017277319258507073,
+      "loss": 0.539,
+      "step": 814
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.20518040499899,
+      "learning_rate": 0.0001727086871312134,
+      "loss": 0.5109,
+      "step": 815
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.19341379491491822,
+      "learning_rate": 0.00017264411742735707,
+      "loss": 0.4882,
+      "step": 816
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.23128359760674316,
+      "learning_rate": 0.00017257948353055963,
+      "loss": 0.547,
+      "step": 817
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1960131633047162,
+      "learning_rate": 0.0001725147854979357,
+      "loss": 0.5467,
+      "step": 818
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.21053602560644855,
+      "learning_rate": 0.00017245002338665656,
+      "loss": 0.5644,
+      "step": 819
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.19602457133539752,
+      "learning_rate": 0.00017238519725395007,
+      "loss": 0.5121,
+      "step": 820
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1923459024283483,
+      "learning_rate": 0.00017232030715710076,
+      "loss": 0.5335,
+      "step": 821
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.19919783133579333,
+      "learning_rate": 0.00017225535315344955,
+      "loss": 0.5076,
+      "step": 822
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.23727428892467575,
+      "learning_rate": 0.00017219033530039397,
+      "loss": 0.5396,
+      "step": 823
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18048505778792392,
+      "learning_rate": 0.00017212525365538792,
+      "loss": 0.467,
+      "step": 824
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.20071702267002645,
+      "learning_rate": 0.00017206010827594163,
+      "loss": 0.5217,
+      "step": 825
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.20315216612025339,
+      "learning_rate": 0.0001719948992196217,
+      "loss": 0.4975,
+      "step": 826
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2142259292765235,
+      "learning_rate": 0.00017192962654405096,
+      "loss": 0.5148,
+      "step": 827
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.19450012283752555,
+      "learning_rate": 0.00017186429030690848,
+      "loss": 0.5297,
+      "step": 828
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.19853162923467543,
+      "learning_rate": 0.00017179889056592954,
+      "loss": 0.547,
+      "step": 829
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.24873174470750906,
+      "learning_rate": 0.00017173342737890544,
+      "loss": 0.563,
+      "step": 830
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18593730182623175,
+      "learning_rate": 0.00017166790080368357,
+      "loss": 0.4647,
+      "step": 831
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.19387710879340586,
+      "learning_rate": 0.00017160231089816748,
+      "loss": 0.5313,
+      "step": 832
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.20818447206363588,
+      "learning_rate": 0.00017153665772031643,
+      "loss": 0.5333,
+      "step": 833
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17584822143732362,
+      "learning_rate": 0.0001714709413281458,
+      "loss": 0.4467,
+      "step": 834
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.19622166504995672,
+      "learning_rate": 0.00017140516177972676,
+      "loss": 0.5129,
+      "step": 835
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18822988249157332,
+      "learning_rate": 0.00017133931913318625,
+      "loss": 0.5186,
+      "step": 836
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2021164051164271,
+      "learning_rate": 0.00017127341344670696,
+      "loss": 0.551,
+      "step": 837
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.19354327476685654,
+      "learning_rate": 0.00017120744477852745,
+      "loss": 0.5001,
+      "step": 838
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2202448214811264,
+      "learning_rate": 0.00017114141318694167,
+      "loss": 0.5516,
+      "step": 839
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2152071248011757,
+      "learning_rate": 0.00017107531873029942,
+      "loss": 0.603,
+      "step": 840
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.21182530477312786,
+      "learning_rate": 0.0001710091614670059,
+      "loss": 0.522,
+      "step": 841
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.19970970678249778,
+      "learning_rate": 0.00017094294145552188,
+      "loss": 0.547,
+      "step": 842
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.19827173723610417,
+      "learning_rate": 0.00017087665875436354,
+      "loss": 0.5238,
+      "step": 843
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.20841052076060687,
+      "learning_rate": 0.00017081031342210245,
+      "loss": 0.5438,
+      "step": 844
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.20276205913776804,
+      "learning_rate": 0.0001707439055173656,
+      "loss": 0.541,
+      "step": 845
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2235538813636529,
+      "learning_rate": 0.00017067743509883515,
+      "loss": 0.5496,
+      "step": 846
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.19092318633528002,
+      "learning_rate": 0.00017061090222524863,
+      "loss": 0.47,
+      "step": 847
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.23200775329397214,
+      "learning_rate": 0.00017054430695539864,
+      "loss": 0.5268,
+      "step": 848
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.22029427294529383,
+      "learning_rate": 0.00017047764934813303,
+      "loss": 0.499,
+      "step": 849
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.22601440602605696,
+      "learning_rate": 0.00017041092946235467,
+      "loss": 0.5593,
+      "step": 850
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18461060416824568,
+      "learning_rate": 0.00017034414735702145,
+      "loss": 0.495,
+      "step": 851
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1901253135554411,
+      "learning_rate": 0.0001702773030911463,
+      "loss": 0.5333,
+      "step": 852
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.21957939715433938,
+      "learning_rate": 0.00017021039672379703,
+      "loss": 0.5019,
+      "step": 853
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18779803577301077,
+      "learning_rate": 0.00017014342831409634,
+      "loss": 0.5079,
+      "step": 854
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.22887531917427628,
+      "learning_rate": 0.00017007639792122173,
+      "loss": 0.537,
+      "step": 855
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.20978489469439648,
+      "learning_rate": 0.00017000930560440554,
+      "loss": 0.5488,
+      "step": 856
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.21114765352664955,
+      "learning_rate": 0.0001699421514229348,
+      "loss": 0.55,
+      "step": 857
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1963314578424483,
+      "learning_rate": 0.00016987493543615115,
+      "loss": 0.487,
+      "step": 858
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.21946023399145406,
+      "learning_rate": 0.0001698076577034509,
+      "loss": 0.5174,
+      "step": 859
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.20119091495408187,
+      "learning_rate": 0.00016974031828428495,
+      "loss": 0.5127,
+      "step": 860
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.19098558699810822,
+      "learning_rate": 0.00016967291723815863,
+      "loss": 0.5776,
+      "step": 861
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.20834795709316445,
+      "learning_rate": 0.00016960545462463183,
+      "loss": 0.4853,
+      "step": 862
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.20553975175824055,
+      "learning_rate": 0.0001695379305033187,
+      "loss": 0.5244,
+      "step": 863
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.38517636125953164,
+      "learning_rate": 0.00016947034493388786,
+      "loss": 0.5314,
+      "step": 864
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.19821744800803398,
+      "learning_rate": 0.00016940269797606228,
+      "loss": 0.5104,
+      "step": 865
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.21102895742419708,
+      "learning_rate": 0.00016933498968961898,
+      "loss": 0.5355,
+      "step": 866
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.202288584940297,
+      "learning_rate": 0.00016926722013438936,
+      "loss": 0.5497,
+      "step": 867
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.20339292404553427,
+      "learning_rate": 0.00016919938937025886,
+      "loss": 0.5655,
+      "step": 868
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.23789079718932474,
+      "learning_rate": 0.00016913149745716703,
+      "loss": 0.5312,
+      "step": 869
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1930501027423073,
+      "learning_rate": 0.00016906354445510747,
+      "loss": 0.5119,
+      "step": 870
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.20670905101801587,
+      "learning_rate": 0.0001689955304241278,
+      "loss": 0.5297,
+      "step": 871
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18514362407781898,
+      "learning_rate": 0.0001689274554243294,
+      "loss": 0.4913,
+      "step": 872
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.22536281121632565,
+      "learning_rate": 0.00016885931951586783,
+      "loss": 0.5236,
+      "step": 873
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.22244899593514914,
+      "learning_rate": 0.00016879112275895215,
+      "loss": 0.5305,
+      "step": 874
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.19941629299670782,
+      "learning_rate": 0.00016872286521384537,
+      "loss": 0.5809,
+      "step": 875
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.20169129247248377,
+      "learning_rate": 0.0001686545469408642,
+      "loss": 0.5081,
+      "step": 876
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.21516300195343227,
+      "learning_rate": 0.000168586168000379,
+      "loss": 0.5084,
+      "step": 877
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.21422988118226038,
+      "learning_rate": 0.00016851772845281367,
+      "loss": 0.5246,
+      "step": 878
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.19390871621362155,
+      "learning_rate": 0.00016844922835864575,
+      "loss": 0.5528,
+      "step": 879
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.21294605765564745,
+      "learning_rate": 0.00016838066777840629,
+      "loss": 0.5971,
+      "step": 880
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.17876424837335442,
+      "learning_rate": 0.00016831204677267975,
+      "loss": 0.4906,
+      "step": 881
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.21551555099302896,
+      "learning_rate": 0.00016824336540210402,
+      "loss": 0.5501,
+      "step": 882
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2046934200181805,
+      "learning_rate": 0.0001681746237273702,
+      "loss": 0.5372,
+      "step": 883
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.19709464178914807,
+      "learning_rate": 0.00016810582180922293,
+      "loss": 0.5479,
+      "step": 884
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.19373994384361878,
+      "learning_rate": 0.00016803695970845985,
+      "loss": 0.5038,
+      "step": 885
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.19009129266864652,
+      "learning_rate": 0.0001679680374859319,
+      "loss": 0.5047,
+      "step": 886
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.19438125438775203,
+      "learning_rate": 0.0001678990552025431,
+      "loss": 0.5241,
+      "step": 887
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.23198116415021436,
+      "learning_rate": 0.00016783001291925055,
+      "loss": 0.5384,
+      "step": 888
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.19609980449580128,
+      "learning_rate": 0.00016776091069706442,
+      "loss": 0.5096,
+      "step": 889
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.19406158834918122,
+      "learning_rate": 0.00016769174859704783,
+      "loss": 0.494,
+      "step": 890
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.21696546174997547,
+      "learning_rate": 0.00016762252668031674,
+      "loss": 0.5064,
+      "step": 891
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18219835689033384,
+      "learning_rate": 0.00016755324500804,
+      "loss": 0.5172,
+      "step": 892
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.20040255213813227,
+      "learning_rate": 0.00016748390364143938,
+      "loss": 0.5691,
+      "step": 893
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.231096201209158,
+      "learning_rate": 0.00016741450264178917,
+      "loss": 0.5921,
+      "step": 894
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1949281480708862,
+      "learning_rate": 0.00016734504207041663,
+      "loss": 0.5268,
+      "step": 895
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.20701715584416944,
+      "learning_rate": 0.00016727552198870135,
+      "loss": 0.5415,
+      "step": 896
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19117861330981445,
+      "learning_rate": 0.00016720594245807582,
+      "loss": 0.5208,
+      "step": 897
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.203448739160119,
+      "learning_rate": 0.00016713630354002484,
+      "loss": 0.4925,
+      "step": 898
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1961421908595443,
+      "learning_rate": 0.00016706660529608583,
+      "loss": 0.5173,
+      "step": 899
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.21571781486847058,
+      "learning_rate": 0.0001669968477878485,
+      "loss": 0.5288,
+      "step": 900
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1951469824130785,
+      "learning_rate": 0.00016692703107695507,
+      "loss": 0.5454,
+      "step": 901
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.20181422096222762,
+      "learning_rate": 0.00016685715522509994,
+      "loss": 0.5266,
+      "step": 902
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.20297647148502806,
+      "learning_rate": 0.0001667872202940299,
+      "loss": 0.5621,
+      "step": 903
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.21597435470412332,
+      "learning_rate": 0.0001667172263455438,
+      "loss": 0.5458,
+      "step": 904
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18789940455022036,
+      "learning_rate": 0.00016664717344149277,
+      "loss": 0.5198,
+      "step": 905
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.20826993025429438,
+      "learning_rate": 0.00016657706164378,
+      "loss": 0.5513,
+      "step": 906
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.20437016308518594,
+      "learning_rate": 0.00016650689101436073,
+      "loss": 0.549,
+      "step": 907
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19686798551292672,
+      "learning_rate": 0.00016643666161524217,
+      "loss": 0.529,
+      "step": 908
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19123982383563307,
+      "learning_rate": 0.00016636637350848338,
+      "loss": 0.4862,
+      "step": 909
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1949343721096402,
+      "learning_rate": 0.00016629602675619548,
+      "loss": 0.5535,
+      "step": 910
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19482047517370968,
+      "learning_rate": 0.0001662256214205413,
+      "loss": 0.5149,
+      "step": 911
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19782922305356188,
+      "learning_rate": 0.00016615515756373533,
+      "loss": 0.5132,
+      "step": 912
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19929182594748354,
+      "learning_rate": 0.00016608463524804407,
+      "loss": 0.4978,
+      "step": 913
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.20329183437505496,
+      "learning_rate": 0.0001660140545357854,
+      "loss": 0.554,
+      "step": 914
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.20759199356585792,
+      "learning_rate": 0.00016594341548932894,
+      "loss": 0.5478,
+      "step": 915
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2141758006074044,
+      "learning_rate": 0.0001658727181710958,
+      "loss": 0.5097,
+      "step": 916
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.19823795581184833,
+      "learning_rate": 0.0001658019626435586,
+      "loss": 0.4997,
+      "step": 917
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.19545826935570532,
+      "learning_rate": 0.00016573114896924147,
+      "loss": 0.4973,
+      "step": 918
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2011307790801751,
+      "learning_rate": 0.0001656602772107198,
+      "loss": 0.5683,
+      "step": 919
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.220529866848964,
+      "learning_rate": 0.00016558934743062035,
+      "loss": 0.574,
+      "step": 920
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.19374991809265163,
+      "learning_rate": 0.00016551835969162118,
+      "loss": 0.4881,
+      "step": 921
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.188760632867396,
+      "learning_rate": 0.00016544731405645154,
+      "loss": 0.5035,
+      "step": 922
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.21165086312587134,
+      "learning_rate": 0.00016537621058789194,
+      "loss": 0.5457,
+      "step": 923
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.22858469252110178,
+      "learning_rate": 0.00016530504934877377,
+      "loss": 0.5361,
+      "step": 924
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.22860822822332005,
+      "learning_rate": 0.0001652338304019797,
+      "loss": 0.5317,
+      "step": 925
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18446055255941704,
+      "learning_rate": 0.00016516255381044323,
+      "loss": 0.5384,
+      "step": 926
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.19147093978335786,
+      "learning_rate": 0.00016509121963714896,
+      "loss": 0.5462,
+      "step": 927
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2010805530061722,
+      "learning_rate": 0.00016501982794513219,
+      "loss": 0.5224,
+      "step": 928
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2194018595901758,
+      "learning_rate": 0.00016494837879747916,
+      "loss": 0.5403,
+      "step": 929
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18748965563201492,
+      "learning_rate": 0.00016487687225732694,
+      "loss": 0.5024,
+      "step": 930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18976038729649813,
+      "learning_rate": 0.00016480530838786312,
+      "loss": 0.5373,
+      "step": 931
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.20751165573259855,
+      "learning_rate": 0.00016473368725232614,
+      "loss": 0.5524,
+      "step": 932
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.23600220568039398,
+      "learning_rate": 0.0001646620089140049,
+      "loss": 0.5315,
+      "step": 933
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17133056460736396,
+      "learning_rate": 0.00016459027343623906,
+      "loss": 0.4753,
+      "step": 934
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1864042334593607,
+      "learning_rate": 0.00016451848088241847,
+      "loss": 0.5617,
+      "step": 935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.20853995853524807,
+      "learning_rate": 0.00016444663131598365,
+      "loss": 0.5149,
+      "step": 936
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.20631366863701492,
+      "learning_rate": 0.00016437472480042544,
+      "loss": 0.5712,
+      "step": 937
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18606219350180553,
+      "learning_rate": 0.00016430276139928494,
+      "loss": 0.5147,
+      "step": 938
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18702291336595328,
+      "learning_rate": 0.00016423074117615362,
+      "loss": 0.4982,
+      "step": 939
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.19313224950151772,
+      "learning_rate": 0.00016415866419467308,
+      "loss": 0.5283,
+      "step": 940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.19468213657571698,
+      "learning_rate": 0.00016408653051853505,
+      "loss": 0.5222,
+      "step": 941
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1900318796388756,
+      "learning_rate": 0.00016401434021148155,
+      "loss": 0.5215,
+      "step": 942
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.20563975975379326,
+      "learning_rate": 0.00016394209333730437,
+      "loss": 0.5332,
+      "step": 943
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.19499178353885577,
+      "learning_rate": 0.0001638697899598455,
+      "loss": 0.5089,
+      "step": 944
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.193828603254631,
+      "learning_rate": 0.00016379743014299675,
+      "loss": 0.4814,
+      "step": 945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2077835011524252,
+      "learning_rate": 0.00016372501395069984,
+      "loss": 0.5874,
+      "step": 946
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.19612398295653072,
+      "learning_rate": 0.0001636525414469463,
+      "loss": 0.5526,
+      "step": 947
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.20309531438683182,
+      "learning_rate": 0.00016358001269577743,
+      "loss": 0.5197,
+      "step": 948
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1915526531241072,
+      "learning_rate": 0.00016350742776128423,
+      "loss": 0.4595,
+      "step": 949
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18881194099828352,
+      "learning_rate": 0.00016343478670760732,
+      "loss": 0.5367,
+      "step": 950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18139152767925906,
+      "learning_rate": 0.00016336208959893698,
+      "loss": 0.4768,
+      "step": 951
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.17769254483980393,
+      "learning_rate": 0.00016328933649951293,
+      "loss": 0.4676,
+      "step": 952
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.20965333812663062,
+      "learning_rate": 0.00016321652747362445,
+      "loss": 0.5555,
+      "step": 953
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.19321230034521153,
+      "learning_rate": 0.00016314366258561016,
+      "loss": 0.5295,
+      "step": 954
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18597567849314112,
+      "learning_rate": 0.00016307074189985814,
+      "loss": 0.4884,
+      "step": 955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18434368773024823,
+      "learning_rate": 0.0001629977654808057,
+      "loss": 0.4824,
+      "step": 956
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.20328216484120545,
+      "learning_rate": 0.0001629247333929394,
+      "loss": 0.58,
+      "step": 957
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.21563942573107894,
+      "learning_rate": 0.00016285164570079504,
+      "loss": 0.5822,
+      "step": 958
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2001040967836008,
+      "learning_rate": 0.00016277850246895753,
+      "loss": 0.5302,
+      "step": 959
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1937847274208307,
+      "learning_rate": 0.0001627053037620609,
+      "loss": 0.4953,
+      "step": 960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.19489799779815298,
+      "learning_rate": 0.00016263204964478807,
+      "loss": 0.5693,
+      "step": 961
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1746038558868276,
+      "learning_rate": 0.00016255874018187113,
+      "loss": 0.4843,
+      "step": 962
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.19886109498653962,
+      "learning_rate": 0.00016248537543809085,
+      "loss": 0.5251,
+      "step": 963
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.23968663323524653,
+      "learning_rate": 0.00016241195547827704,
+      "loss": 0.5842,
+      "step": 964
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17879861073069336,
+      "learning_rate": 0.00016233848036730818,
+      "loss": 0.5271,
+      "step": 965
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18682851702979986,
+      "learning_rate": 0.00016226495017011155,
+      "loss": 0.5117,
+      "step": 966
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17994919875574256,
+      "learning_rate": 0.0001621913649516631,
+      "loss": 0.5277,
+      "step": 967
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18619266507890767,
+      "learning_rate": 0.00016211772477698737,
+      "loss": 0.5564,
+      "step": 968
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18791500150949197,
+      "learning_rate": 0.0001620440297111575,
+      "loss": 0.5138,
+      "step": 969
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.20203542918576534,
+      "learning_rate": 0.00016197027981929506,
+      "loss": 0.5027,
+      "step": 970
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2028531086326741,
+      "learning_rate": 0.00016189647516657018,
+      "loss": 0.5271,
+      "step": 971
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.20588712233829473,
+      "learning_rate": 0.0001618226158182013,
+      "loss": 0.5453,
+      "step": 972
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17483034796275576,
+      "learning_rate": 0.00016174870183945523,
+      "loss": 0.5213,
+      "step": 973
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1869321547342615,
+      "learning_rate": 0.00016167473329564705,
+      "loss": 0.5445,
+      "step": 974
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.20261074022832953,
+      "learning_rate": 0.00016160071025213998,
+      "loss": 0.5607,
+      "step": 975
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.20067938165729438,
+      "learning_rate": 0.00016152663277434556,
+      "loss": 0.5145,
+      "step": 976
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1886528405831426,
+      "learning_rate": 0.0001614525009277233,
+      "loss": 0.5212,
+      "step": 977
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.20396953786191716,
+      "learning_rate": 0.00016137831477778077,
+      "loss": 0.5546,
+      "step": 978
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.22783734025618152,
+      "learning_rate": 0.00016130407439007355,
+      "loss": 0.5541,
+      "step": 979
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2062351395059516,
+      "learning_rate": 0.0001612297798302052,
+      "loss": 0.5276,
+      "step": 980
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1971656346477824,
+      "learning_rate": 0.00016115543116382707,
+      "loss": 0.5232,
+      "step": 981
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.19528017198031397,
+      "learning_rate": 0.00016108102845663832,
+      "loss": 0.5468,
+      "step": 982
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.213721647209808,
+      "learning_rate": 0.00016100657177438592,
+      "loss": 0.5617,
+      "step": 983
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.21222822398944205,
+      "learning_rate": 0.0001609320611828645,
+      "loss": 0.5014,
+      "step": 984
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.19722416264931983,
+      "learning_rate": 0.0001608574967479163,
+      "loss": 0.5811,
+      "step": 985
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.19590127100369606,
+      "learning_rate": 0.00016078287853543125,
+      "loss": 0.518,
+      "step": 986
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.19341001514547207,
+      "learning_rate": 0.00016070820661134668,
+      "loss": 0.4885,
+      "step": 987
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.20898635250980999,
+      "learning_rate": 0.00016063348104164744,
+      "loss": 0.4751,
+      "step": 988
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.21215717136848186,
+      "learning_rate": 0.00016055870189236578,
+      "loss": 0.5211,
+      "step": 989
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.24258818484100222,
+      "learning_rate": 0.00016048386922958127,
+      "loss": 0.4836,
+      "step": 990
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1798892664861584,
+      "learning_rate": 0.00016040898311942082,
+      "loss": 0.4984,
+      "step": 991
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18450252398162806,
+      "learning_rate": 0.0001603340436280585,
+      "loss": 0.5306,
+      "step": 992
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18837615304247687,
+      "learning_rate": 0.00016025905082171562,
+      "loss": 0.4754,
+      "step": 993
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2214407132100871,
+      "learning_rate": 0.00016018400476666055,
+      "loss": 0.6057,
+      "step": 994
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.19318387417313984,
+      "learning_rate": 0.00016010890552920875,
+      "loss": 0.5205,
+      "step": 995
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18092243312526385,
+      "learning_rate": 0.00016003375317572263,
+      "loss": 0.5283,
+      "step": 996
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18814705771101395,
+      "learning_rate": 0.00015995854777261161,
+      "loss": 0.5231,
+      "step": 997
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2411971952093798,
+      "learning_rate": 0.00015988328938633191,
+      "loss": 0.5382,
+      "step": 998
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19563910470947196,
+      "learning_rate": 0.00015980797808338664,
+      "loss": 0.5228,
+      "step": 999
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1859383484794571,
+      "learning_rate": 0.00015973261393032563,
+      "loss": 0.5166,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19354728242000344,
+      "learning_rate": 0.0001596571969937454,
+      "loss": 0.5317,
+      "step": 1001
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19477100537075334,
+      "learning_rate": 0.0001595817273402891,
+      "loss": 0.5389,
+      "step": 1002
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19658938430385506,
+      "learning_rate": 0.00015950620503664658,
+      "loss": 0.5576,
+      "step": 1003
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18705811284094567,
+      "learning_rate": 0.00015943063014955402,
+      "loss": 0.5339,
+      "step": 1004
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2238775349637154,
+      "learning_rate": 0.00015935500274579426,
+      "loss": 0.5348,
+      "step": 1005
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.20521877935386715,
+      "learning_rate": 0.00015927932289219642,
+      "loss": 0.5817,
+      "step": 1006
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1748831609387364,
+      "learning_rate": 0.00015920359065563604,
+      "loss": 0.5003,
+      "step": 1007
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19535482397472856,
+      "learning_rate": 0.0001591278061030349,
+      "loss": 0.5075,
+      "step": 1008
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2091479313897058,
+      "learning_rate": 0.00015905196930136097,
+      "loss": 0.5555,
+      "step": 1009
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19083157236229129,
+      "learning_rate": 0.0001589760803176286,
+      "loss": 0.5166,
+      "step": 1010
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19553974503051083,
+      "learning_rate": 0.00015890013921889795,
+      "loss": 0.5978,
+      "step": 1011
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18073049964870547,
+      "learning_rate": 0.00015882414607227546,
+      "loss": 0.5136,
+      "step": 1012
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18595442399132076,
+      "learning_rate": 0.00015874810094491343,
+      "loss": 0.494,
+      "step": 1013
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18565792372097342,
+      "learning_rate": 0.00015867200390401023,
+      "loss": 0.5211,
+      "step": 1014
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.19074958534189332,
+      "learning_rate": 0.00015859585501681,
+      "loss": 0.5371,
+      "step": 1015
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.22738878543819066,
+      "learning_rate": 0.00015851965435060262,
+      "loss": 0.5055,
+      "step": 1016
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2316623260697804,
+      "learning_rate": 0.00015844340197272393,
+      "loss": 0.5467,
+      "step": 1017
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18501859711102792,
+      "learning_rate": 0.00015836709795055532,
+      "loss": 0.5152,
+      "step": 1018
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2000854019680551,
+      "learning_rate": 0.0001582907423515239,
+      "loss": 0.4975,
+      "step": 1019
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2164338400016087,
+      "learning_rate": 0.00015821433524310224,
+      "loss": 0.5108,
+      "step": 1020
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1794884681634538,
+      "learning_rate": 0.00015813787669280855,
+      "loss": 0.4852,
+      "step": 1021
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18483080731616555,
+      "learning_rate": 0.00015806136676820639,
+      "loss": 0.5126,
+      "step": 1022
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18580223692725362,
+      "learning_rate": 0.00015798480553690482,
+      "loss": 0.5068,
+      "step": 1023
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2061029141413908,
+      "learning_rate": 0.0001579081930665582,
+      "loss": 0.5358,
+      "step": 1024
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.20179672861743628,
+      "learning_rate": 0.00015783152942486613,
+      "loss": 0.4904,
+      "step": 1025
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1901456318733625,
+      "learning_rate": 0.0001577548146795735,
+      "loss": 0.5189,
+      "step": 1026
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18700717565555658,
+      "learning_rate": 0.00015767804889847025,
+      "loss": 0.5053,
+      "step": 1027
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.20306080744009194,
+      "learning_rate": 0.00015760123214939148,
+      "loss": 0.5648,
+      "step": 1028
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.19427308531076118,
+      "learning_rate": 0.00015752436450021742,
+      "loss": 0.5317,
+      "step": 1029
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.17730364387799943,
+      "learning_rate": 0.0001574474460188731,
+      "loss": 0.4865,
+      "step": 1030
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18919444977285255,
+      "learning_rate": 0.00015737047677332863,
+      "loss": 0.4482,
+      "step": 1031
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2186748845065149,
+      "learning_rate": 0.0001572934568315989,
+      "loss": 0.5166,
+      "step": 1032
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2157209010754937,
+      "learning_rate": 0.00015721638626174354,
+      "loss": 0.5,
+      "step": 1033
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.19854498695084713,
+      "learning_rate": 0.00015713926513186702,
+      "loss": 0.5833,
+      "step": 1034
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.19845689442164263,
+      "learning_rate": 0.00015706209351011848,
+      "loss": 0.5678,
+      "step": 1035
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1816372371442818,
+      "learning_rate": 0.00015698487146469163,
+      "loss": 0.5403,
+      "step": 1036
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1770141417061582,
+      "learning_rate": 0.0001569075990638248,
+      "loss": 0.4799,
+      "step": 1037
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1828187317317013,
+      "learning_rate": 0.00015683027637580066,
+      "loss": 0.4879,
+      "step": 1038
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.20434187555872277,
+      "learning_rate": 0.00015675290346894657,
+      "loss": 0.5189,
+      "step": 1039
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1906338108685848,
+      "learning_rate": 0.00015667548041163406,
+      "loss": 0.5162,
+      "step": 1040
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.19218887365666604,
+      "learning_rate": 0.00015659800727227903,
+      "loss": 0.5571,
+      "step": 1041
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.19183422180053736,
+      "learning_rate": 0.00015652048411934167,
+      "loss": 0.5144,
+      "step": 1042
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.19849906503502973,
+      "learning_rate": 0.00015644291102132635,
+      "loss": 0.5271,
+      "step": 1043
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18585204906477085,
+      "learning_rate": 0.0001563652880467816,
+      "loss": 0.4906,
+      "step": 1044
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.20344554390359765,
+      "learning_rate": 0.00015628761526429992,
+      "loss": 0.5268,
+      "step": 1045
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.22108658658095226,
+      "learning_rate": 0.00015620989274251797,
+      "loss": 0.5253,
+      "step": 1046
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1991875825405089,
+      "learning_rate": 0.00015613212055011624,
+      "loss": 0.5311,
+      "step": 1047
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.20172996858736694,
+      "learning_rate": 0.0001560542987558192,
+      "loss": 0.5008,
+      "step": 1048
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1879727067503482,
+      "learning_rate": 0.00015597642742839506,
+      "loss": 0.539,
+      "step": 1049
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2016771106491948,
+      "learning_rate": 0.00015589850663665593,
+      "loss": 0.5426,
+      "step": 1050
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.20217576692011496,
+      "learning_rate": 0.0001558205364494575,
+      "loss": 0.5315,
+      "step": 1051
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.19326432360974946,
+      "learning_rate": 0.0001557425169356992,
+      "loss": 0.4779,
+      "step": 1052
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.20958769456715673,
+      "learning_rate": 0.000155664448164324,
+      "loss": 0.5368,
+      "step": 1053
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1944842076067339,
+      "learning_rate": 0.00015558633020431835,
+      "loss": 0.4948,
+      "step": 1054
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.19103650755521584,
+      "learning_rate": 0.00015550816312471234,
+      "loss": 0.4687,
+      "step": 1055
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.19130294148586216,
+      "learning_rate": 0.00015542994699457925,
+      "loss": 0.5241,
+      "step": 1056
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.19011766175405304,
+      "learning_rate": 0.00015535168188303585,
+      "loss": 0.5157,
+      "step": 1057
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.19308237510118917,
+      "learning_rate": 0.00015527336785924213,
+      "loss": 0.5158,
+      "step": 1058
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.18949297343958862,
+      "learning_rate": 0.00015519500499240133,
+      "loss": 0.5525,
+      "step": 1059
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.18554479264301466,
+      "learning_rate": 0.00015511659335175985,
+      "loss": 0.516,
+      "step": 1060
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.20363425340107272,
+      "learning_rate": 0.00015503813300660717,
+      "loss": 0.509,
+      "step": 1061
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.18613987636546375,
+      "learning_rate": 0.0001549596240262758,
+      "loss": 0.5097,
+      "step": 1062
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.20481511432281724,
+      "learning_rate": 0.00015488106648014127,
+      "loss": 0.5263,
+      "step": 1063
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.18895386399049763,
+      "learning_rate": 0.00015480246043762198,
+      "loss": 0.528,
+      "step": 1064
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19514309309247793,
+      "learning_rate": 0.00015472380596817922,
+      "loss": 0.5444,
+      "step": 1065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2013716596665657,
+      "learning_rate": 0.0001546451031413171,
+      "loss": 0.535,
+      "step": 1066
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2003122019096167,
+      "learning_rate": 0.0001545663520265823,
+      "loss": 0.5294,
+      "step": 1067
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18908243290011195,
+      "learning_rate": 0.00015448755269356442,
+      "loss": 0.5366,
+      "step": 1068
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1893813159340979,
+      "learning_rate": 0.00015440870521189547,
+      "loss": 0.5397,
+      "step": 1069
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.17856528748444775,
+      "learning_rate": 0.00015432980965125008,
+      "loss": 0.4996,
+      "step": 1070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.20198848206655348,
+      "learning_rate": 0.0001542508660813454,
+      "loss": 0.5084,
+      "step": 1071
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2027016976348092,
+      "learning_rate": 0.00015417187457194092,
+      "loss": 0.5835,
+      "step": 1072
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19225289448462915,
+      "learning_rate": 0.00015409283519283857,
+      "loss": 0.5333,
+      "step": 1073
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19832197684261557,
+      "learning_rate": 0.00015401374801388254,
+      "loss": 0.5183,
+      "step": 1074
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18727329905053802,
+      "learning_rate": 0.00015393461310495926,
+      "loss": 0.5104,
+      "step": 1075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19371954596455335,
+      "learning_rate": 0.0001538554305359974,
+      "loss": 0.5482,
+      "step": 1076
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.17653026485031678,
+      "learning_rate": 0.00015377620037696757,
+      "loss": 0.4785,
+      "step": 1077
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18746235874046693,
+      "learning_rate": 0.00015369692269788266,
+      "loss": 0.5432,
+      "step": 1078
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.20320517317727052,
+      "learning_rate": 0.0001536175975687974,
+      "loss": 0.5358,
+      "step": 1079
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18239445216598227,
+      "learning_rate": 0.00015353822505980854,
+      "loss": 0.5133,
+      "step": 1080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2001726032915372,
+      "learning_rate": 0.00015345880524105462,
+      "loss": 0.5601,
+      "step": 1081
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.17441851789965093,
+      "learning_rate": 0.00015337933818271597,
+      "loss": 0.4992,
+      "step": 1082
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19460177893040895,
+      "learning_rate": 0.00015329982395501478,
+      "loss": 0.5038,
+      "step": 1083
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18954115748132863,
+      "learning_rate": 0.00015322026262821488,
+      "loss": 0.5691,
+      "step": 1084
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18958791658091825,
+      "learning_rate": 0.00015314065427262166,
+      "loss": 0.513,
+      "step": 1085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18280595622163878,
+      "learning_rate": 0.00015306099895858206,
+      "loss": 0.5147,
+      "step": 1086
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.17678570607217403,
+      "learning_rate": 0.00015298129675648462,
+      "loss": 0.4952,
+      "step": 1087
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.17214270706367016,
+      "learning_rate": 0.00015290154773675923,
+      "loss": 0.4614,
+      "step": 1088
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1874186691157256,
+      "learning_rate": 0.00015282175196987721,
+      "loss": 0.5354,
+      "step": 1089
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18779920562332963,
+      "learning_rate": 0.00015274190952635106,
+      "loss": 0.4892,
+      "step": 1090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.17670235761935024,
+      "learning_rate": 0.00015266202047673467,
+      "loss": 0.5265,
+      "step": 1091
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1744750957023612,
+      "learning_rate": 0.00015258208489162312,
+      "loss": 0.547,
+      "step": 1092
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.17422295122430556,
+      "learning_rate": 0.00015250210284165246,
+      "loss": 0.5005,
+      "step": 1093
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.17453851271743742,
+      "learning_rate": 0.00015242207439749992,
+      "loss": 0.4727,
+      "step": 1094
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2020628888829431,
+      "learning_rate": 0.0001523419996298837,
+      "loss": 0.5164,
+      "step": 1095
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18036982996630943,
+      "learning_rate": 0.00015226187860956295,
+      "loss": 0.4998,
+      "step": 1096
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19207762457147357,
+      "learning_rate": 0.00015218171140733773,
+      "loss": 0.4962,
+      "step": 1097
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17728752409159226,
+      "learning_rate": 0.00015210149809404875,
+      "loss": 0.486,
+      "step": 1098
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18658006486346662,
+      "learning_rate": 0.00015202123874057761,
+      "loss": 0.5701,
+      "step": 1099
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1799589924022257,
+      "learning_rate": 0.00015194093341784655,
+      "loss": 0.508,
+      "step": 1100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18488881231943724,
+      "learning_rate": 0.00015186058219681848,
+      "loss": 0.5276,
+      "step": 1101
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.19172132491153843,
+      "learning_rate": 0.00015178018514849678,
+      "loss": 0.5103,
+      "step": 1102
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18192159347628917,
+      "learning_rate": 0.00015169974234392538,
+      "loss": 0.5168,
+      "step": 1103
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17390383147461494,
+      "learning_rate": 0.00015161925385418867,
+      "loss": 0.5117,
+      "step": 1104
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18438067858194318,
+      "learning_rate": 0.00015153871975041131,
+      "loss": 0.5362,
+      "step": 1105
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18794738239211156,
+      "learning_rate": 0.00015145814010375841,
+      "loss": 0.5284,
+      "step": 1106
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18207793056557062,
+      "learning_rate": 0.00015137751498543517,
+      "loss": 0.5157,
+      "step": 1107
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1820180090632629,
+      "learning_rate": 0.00015129684446668713,
+      "loss": 0.4754,
+      "step": 1108
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17889600747067783,
+      "learning_rate": 0.00015121612861879974,
+      "loss": 0.4987,
+      "step": 1109
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17971937316609232,
+      "learning_rate": 0.00015113536751309878,
+      "loss": 0.4911,
+      "step": 1110
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18140328667930378,
+      "learning_rate": 0.00015105456122094983,
+      "loss": 0.5305,
+      "step": 1111
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2025119429324591,
+      "learning_rate": 0.00015097370981375838,
+      "loss": 0.5386,
+      "step": 1112
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.19491035376315885,
+      "learning_rate": 0.0001508928133629699,
+      "loss": 0.4907,
+      "step": 1113
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.21078693446799357,
+      "learning_rate": 0.00015081187194006962,
+      "loss": 0.5365,
+      "step": 1114
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18843997731657822,
+      "learning_rate": 0.0001507308856165825,
+      "loss": 0.5183,
+      "step": 1115
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18301017548445006,
+      "learning_rate": 0.00015064985446407321,
+      "loss": 0.5365,
+      "step": 1116
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.20526106918611095,
+      "learning_rate": 0.00015056877855414594,
+      "loss": 0.5424,
+      "step": 1117
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19687691349301994,
+      "learning_rate": 0.00015048765795844457,
+      "loss": 0.4767,
+      "step": 1118
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.24595956331043234,
+      "learning_rate": 0.00015040649274865238,
+      "loss": 0.5276,
+      "step": 1119
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19028199168907478,
+      "learning_rate": 0.000150325282996492,
+      "loss": 0.5154,
+      "step": 1120
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1869717933744596,
+      "learning_rate": 0.00015024402877372562,
+      "loss": 0.4994,
+      "step": 1121
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19760879222012737,
+      "learning_rate": 0.00015016273015215455,
+      "loss": 0.5181,
+      "step": 1122
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19770025526423807,
+      "learning_rate": 0.00015008138720361942,
+      "loss": 0.5224,
+      "step": 1123
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.21756294254565256,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.5578,
+      "step": 1124
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.20716812622877365,
+      "learning_rate": 0.0001499185686132152,
+      "loss": 0.5812,
+      "step": 1125
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.20335171698711324,
+      "learning_rate": 0.00014983709311522297,
+      "loss": 0.5249,
+      "step": 1126
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2176962106234944,
+      "learning_rate": 0.0001497555735780201,
+      "loss": 0.5856,
+      "step": 1127
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1783334911039492,
+      "learning_rate": 0.00014967401007364255,
+      "loss": 0.4926,
+      "step": 1128
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.21009643426167063,
+      "learning_rate": 0.0001495924026741649,
+      "loss": 0.4825,
+      "step": 1129
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.182586366890212,
+      "learning_rate": 0.0001495107514517007,
+      "loss": 0.4694,
+      "step": 1130
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19955812947692972,
+      "learning_rate": 0.00014942905647840206,
+      "loss": 0.5385,
+      "step": 1131
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18833362414930832,
+      "learning_rate": 0.0001493473178264599,
+      "loss": 0.5331,
+      "step": 1132
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19652826299925613,
+      "learning_rate": 0.0001492655355681036,
+      "loss": 0.5833,
+      "step": 1133
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19744208734697974,
+      "learning_rate": 0.00014918370977560122,
+      "loss": 0.5147,
+      "step": 1134
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19287794254421994,
+      "learning_rate": 0.0001491018405212591,
+      "loss": 0.4678,
+      "step": 1135
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.20401004694813352,
+      "learning_rate": 0.00014901992787742219,
+      "loss": 0.5388,
+      "step": 1136
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1859539975227642,
+      "learning_rate": 0.00014893797191647368,
+      "loss": 0.5121,
+      "step": 1137
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1777739643065349,
+      "learning_rate": 0.00014885597271083499,
+      "loss": 0.4917,
+      "step": 1138
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.24667049739226657,
+      "learning_rate": 0.00014877393033296585,
+      "loss": 0.5631,
+      "step": 1139
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19660914123915638,
+      "learning_rate": 0.00014869184485536408,
+      "loss": 0.4955,
+      "step": 1140
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1849639582602034,
+      "learning_rate": 0.00014860971635056563,
+      "loss": 0.5611,
+      "step": 1141
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18618478737494773,
+      "learning_rate": 0.00014852754489114444,
+      "loss": 0.5668,
+      "step": 1142
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18738098465743458,
+      "learning_rate": 0.0001484453305497124,
+      "loss": 0.491,
+      "step": 1143
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2001668377337724,
+      "learning_rate": 0.00014836307339891934,
+      "loss": 0.5395,
+      "step": 1144
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18840397893983923,
+      "learning_rate": 0.00014828077351145282,
+      "loss": 0.5466,
+      "step": 1145
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18348054201307173,
+      "learning_rate": 0.00014819843096003824,
+      "loss": 0.551,
+      "step": 1146
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.17819691596218545,
+      "learning_rate": 0.0001481160458174388,
+      "loss": 0.4773,
+      "step": 1147
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.17697981019548276,
+      "learning_rate": 0.0001480336181564551,
+      "loss": 0.4586,
+      "step": 1148
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2051551821180113,
+      "learning_rate": 0.0001479511480499255,
+      "loss": 0.5141,
+      "step": 1149
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.19457252340787579,
+      "learning_rate": 0.00014786863557072582,
+      "loss": 0.4769,
+      "step": 1150
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.20436510440947783,
+      "learning_rate": 0.00014778608079176923,
+      "loss": 0.5556,
+      "step": 1151
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18295670635083391,
+      "learning_rate": 0.00014770348378600646,
+      "loss": 0.5063,
+      "step": 1152
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18628367387903996,
+      "learning_rate": 0.00014762084462642539,
+      "loss": 0.5427,
+      "step": 1153
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2161510404034079,
+      "learning_rate": 0.00014753816338605123,
+      "loss": 0.4854,
+      "step": 1154
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.17721958276959857,
+      "learning_rate": 0.00014745544013794636,
+      "loss": 0.5226,
+      "step": 1155
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18343920037243416,
+      "learning_rate": 0.0001473726749552103,
+      "loss": 0.5187,
+      "step": 1156
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.19071533215976352,
+      "learning_rate": 0.00014728986791097957,
+      "loss": 0.5307,
+      "step": 1157
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.17869129856056046,
+      "learning_rate": 0.00014720701907842772,
+      "loss": 0.4744,
+      "step": 1158
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18707001538233387,
+      "learning_rate": 0.00014712412853076524,
+      "loss": 0.4709,
+      "step": 1159
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18697017631013504,
+      "learning_rate": 0.00014704119634123948,
+      "loss": 0.5077,
+      "step": 1160
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18814502472746178,
+      "learning_rate": 0.00014695822258313455,
+      "loss": 0.5233,
+      "step": 1161
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2368823091300486,
+      "learning_rate": 0.00014687520732977128,
+      "loss": 0.4966,
+      "step": 1162
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.21080889905859476,
+      "learning_rate": 0.00014679215065450726,
+      "loss": 0.5262,
+      "step": 1163
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.20202781405176504,
+      "learning_rate": 0.0001467090526307366,
+      "loss": 0.547,
+      "step": 1164
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22096110211580308,
+      "learning_rate": 0.00014662591333189,
+      "loss": 0.5188,
+      "step": 1165
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.20568086982318412,
+      "learning_rate": 0.0001465427328314346,
+      "loss": 0.4716,
+      "step": 1166
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2001156528680437,
+      "learning_rate": 0.0001464595112028739,
+      "loss": 0.4726,
+      "step": 1167
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22271237142339692,
+      "learning_rate": 0.0001463762485197479,
+      "loss": 0.5685,
+      "step": 1168
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.20249978518037207,
+      "learning_rate": 0.00014629294485563271,
+      "loss": 0.4992,
+      "step": 1169
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.21051317014195287,
+      "learning_rate": 0.00014620960028414074,
+      "loss": 0.4941,
+      "step": 1170
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22015178328592389,
+      "learning_rate": 0.0001461262148789205,
+      "loss": 0.5554,
+      "step": 1171
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1718994350084717,
+      "learning_rate": 0.00014604278871365662,
+      "loss": 0.4964,
+      "step": 1172
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.17185816501519585,
+      "learning_rate": 0.0001459593218620698,
+      "loss": 0.5104,
+      "step": 1173
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.21401829521307225,
+      "learning_rate": 0.0001458758143979166,
+      "loss": 0.499,
+      "step": 1174
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19776025857516183,
+      "learning_rate": 0.00014579226639498946,
+      "loss": 0.496,
+      "step": 1175
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1961032370488782,
+      "learning_rate": 0.00014570867792711674,
+      "loss": 0.5107,
+      "step": 1176
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18716882427415388,
+      "learning_rate": 0.0001456250490681625,
+      "loss": 0.4876,
+      "step": 1177
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1956152915992498,
+      "learning_rate": 0.00014554137989202643,
+      "loss": 0.5732,
+      "step": 1178
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19786724370350303,
+      "learning_rate": 0.000145457670472644,
+      "loss": 0.5551,
+      "step": 1179
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18947724734010707,
+      "learning_rate": 0.00014537392088398608,
+      "loss": 0.4944,
+      "step": 1180
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2097627269461171,
+      "learning_rate": 0.00014529013120005916,
+      "loss": 0.5383,
+      "step": 1181
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.17430571658619776,
+      "learning_rate": 0.0001452063014949051,
+      "loss": 0.4849,
+      "step": 1182
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1709299956294579,
+      "learning_rate": 0.0001451224318426011,
+      "loss": 0.4627,
+      "step": 1183
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.17998153878412265,
+      "learning_rate": 0.0001450385223172597,
+      "loss": 0.5102,
+      "step": 1184
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1973625392843972,
+      "learning_rate": 0.0001449545729930287,
+      "loss": 0.5508,
+      "step": 1185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1871194681672037,
+      "learning_rate": 0.00014487058394409104,
+      "loss": 0.5061,
+      "step": 1186
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1869428596068349,
+      "learning_rate": 0.00014478655524466475,
+      "loss": 0.4707,
+      "step": 1187
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1847860359839481,
+      "learning_rate": 0.00014470248696900285,
+      "loss": 0.5141,
+      "step": 1188
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18708736509686738,
+      "learning_rate": 0.00014461837919139348,
+      "loss": 0.5192,
+      "step": 1189
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.22393402471111507,
+      "learning_rate": 0.00014453423198615957,
+      "loss": 0.5192,
+      "step": 1190
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19155699706033086,
+      "learning_rate": 0.00014445004542765888,
+      "loss": 0.5054,
+      "step": 1191
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18623804761748072,
+      "learning_rate": 0.00014436581959028405,
+      "loss": 0.5212,
+      "step": 1192
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18614307922016232,
+      "learning_rate": 0.00014428155454846225,
+      "loss": 0.4667,
+      "step": 1193
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.20760229991014245,
+      "learning_rate": 0.0001441972503766555,
+      "loss": 0.4897,
+      "step": 1194
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1713367358079391,
+      "learning_rate": 0.00014411290714936033,
+      "loss": 0.5075,
+      "step": 1195
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18925060252185533,
+      "learning_rate": 0.00014402852494110768,
+      "loss": 0.5185,
+      "step": 1196
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1845531822650512,
+      "learning_rate": 0.00014394410382646304,
+      "loss": 0.4799,
+      "step": 1197
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.17403525502879638,
+      "learning_rate": 0.00014385964388002623,
+      "loss": 0.4935,
+      "step": 1198
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1991754972420626,
+      "learning_rate": 0.00014377514517643144,
+      "loss": 0.5419,
+      "step": 1199
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.17030413694508353,
+      "learning_rate": 0.00014369060779034708,
+      "loss": 0.4907,
+      "step": 1200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.17984445026847887,
+      "learning_rate": 0.00014360603179647567,
+      "loss": 0.5024,
+      "step": 1201
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1782253350761795,
+      "learning_rate": 0.000143521417269554,
+      "loss": 0.4629,
+      "step": 1202
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18078126508901013,
+      "learning_rate": 0.00014343676428435275,
+      "loss": 0.4884,
+      "step": 1203
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.17808414211423282,
+      "learning_rate": 0.0001433520729156767,
+      "loss": 0.4884,
+      "step": 1204
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.19634028134840262,
+      "learning_rate": 0.0001432673432383645,
+      "loss": 0.5253,
+      "step": 1205
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.17516039333071384,
+      "learning_rate": 0.00014318257532728866,
+      "loss": 0.5234,
+      "step": 1206
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18938507108619154,
+      "learning_rate": 0.0001430977692573554,
+      "loss": 0.4569,
+      "step": 1207
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18529885163701465,
+      "learning_rate": 0.00014301292510350485,
+      "loss": 0.5588,
+      "step": 1208
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1980418349421889,
+      "learning_rate": 0.0001429280429407106,
+      "loss": 0.5567,
+      "step": 1209
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18102359715715138,
+      "learning_rate": 0.00014284312284397994,
+      "loss": 0.5346,
+      "step": 1210
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1682936887439461,
+      "learning_rate": 0.00014275816488835364,
+      "loss": 0.4636,
+      "step": 1211
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1800173543128752,
+      "learning_rate": 0.00014267316914890583,
+      "loss": 0.5273,
+      "step": 1212
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18733177044546032,
+      "learning_rate": 0.00014258813570074429,
+      "loss": 0.5375,
+      "step": 1213
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1748926227325175,
+      "learning_rate": 0.00014250306461900984,
+      "loss": 0.5167,
+      "step": 1214
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18731958978952354,
+      "learning_rate": 0.00014241795597887675,
+      "loss": 0.4931,
+      "step": 1215
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1826531388805938,
+      "learning_rate": 0.00014233280985555234,
+      "loss": 0.5561,
+      "step": 1216
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.17687603086969267,
+      "learning_rate": 0.00014224762632427713,
+      "loss": 0.4916,
+      "step": 1217
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18307772952489884,
+      "learning_rate": 0.0001421624054603247,
+      "loss": 0.4885,
+      "step": 1218
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.17499886290677696,
+      "learning_rate": 0.00014207714733900162,
+      "loss": 0.472,
+      "step": 1219
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.21149209308036526,
+      "learning_rate": 0.00014199185203564728,
+      "loss": 0.5162,
+      "step": 1220
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18757068290035836,
+      "learning_rate": 0.00014190651962563407,
+      "loss": 0.5057,
+      "step": 1221
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1968053254662056,
+      "learning_rate": 0.00014182115018436715,
+      "loss": 0.4965,
+      "step": 1222
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1832324100206531,
+      "learning_rate": 0.0001417357437872843,
+      "loss": 0.5297,
+      "step": 1223
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1947511964403348,
+      "learning_rate": 0.00014165030050985604,
+      "loss": 0.5108,
+      "step": 1224
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1887394891277499,
+      "learning_rate": 0.00014156482042758544,
+      "loss": 0.5378,
+      "step": 1225
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1948370350677339,
+      "learning_rate": 0.0001414793036160081,
+      "loss": 0.5031,
+      "step": 1226
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1876481087251801,
+      "learning_rate": 0.00014139375015069215,
+      "loss": 0.5245,
+      "step": 1227
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.17100568796101634,
+      "learning_rate": 0.00014130816010723805,
+      "loss": 0.474,
+      "step": 1228
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18845060671638836,
+      "learning_rate": 0.0001412225335612785,
+      "loss": 0.4944,
+      "step": 1229
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19905303803856061,
+      "learning_rate": 0.00014113687058847857,
+      "loss": 0.5389,
+      "step": 1230
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18736361174530702,
+      "learning_rate": 0.00014105117126453554,
+      "loss": 0.5128,
+      "step": 1231
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18113672252747182,
+      "learning_rate": 0.00014096543566517871,
+      "loss": 0.5077,
+      "step": 1232
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18697696198984598,
+      "learning_rate": 0.00014087966386616945,
+      "loss": 0.5083,
+      "step": 1233
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18321266285619817,
+      "learning_rate": 0.00014079385594330121,
+      "loss": 0.501,
+      "step": 1234
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2085760786366843,
+      "learning_rate": 0.00014070801197239928,
+      "loss": 0.5337,
+      "step": 1235
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.20073471496417797,
+      "learning_rate": 0.00014062213202932085,
+      "loss": 0.5354,
+      "step": 1236
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.17820357534513603,
+      "learning_rate": 0.00014053621618995488,
+      "loss": 0.4728,
+      "step": 1237
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18692997172911976,
+      "learning_rate": 0.00014045026453022197,
+      "loss": 0.5461,
+      "step": 1238
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2216376697939887,
+      "learning_rate": 0.00014036427712607453,
+      "loss": 0.5099,
+      "step": 1239
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19522613324619895,
+      "learning_rate": 0.00014027825405349642,
+      "loss": 0.5257,
+      "step": 1240
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.23644520056181112,
+      "learning_rate": 0.0001401921953885031,
+      "loss": 0.5637,
+      "step": 1241
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1850220175837294,
+      "learning_rate": 0.00014010610120714147,
+      "loss": 0.5299,
+      "step": 1242
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.17697459178123112,
+      "learning_rate": 0.00014001997158548973,
+      "loss": 0.4723,
+      "step": 1243
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.20359077418663543,
+      "learning_rate": 0.00013993380659965755,
+      "loss": 0.5546,
+      "step": 1244
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1875137496084646,
+      "learning_rate": 0.00013984760632578577,
+      "loss": 0.5747,
+      "step": 1245
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.21313947649725407,
+      "learning_rate": 0.00013976137084004633,
+      "loss": 0.5321,
+      "step": 1246
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.17217418667779574,
+      "learning_rate": 0.0001396751002186424,
+      "loss": 0.4719,
+      "step": 1247
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18863205998380614,
+      "learning_rate": 0.00013958879453780817,
+      "loss": 0.5627,
+      "step": 1248
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.17087031621768445,
+      "learning_rate": 0.00013950245387380882,
+      "loss": 0.4626,
+      "step": 1249
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.17677171504664238,
+      "learning_rate": 0.00013941607830294042,
+      "loss": 0.4593,
+      "step": 1250
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19402672015245978,
+      "learning_rate": 0.00013932966790152987,
+      "loss": 0.5231,
+      "step": 1251
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2008980487940075,
+      "learning_rate": 0.00013924322274593486,
+      "loss": 0.5223,
+      "step": 1252
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19654503986426297,
+      "learning_rate": 0.00013915674291254383,
+      "loss": 0.5489,
+      "step": 1253
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18755885764625013,
+      "learning_rate": 0.00013907022847777585,
+      "loss": 0.5124,
+      "step": 1254
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.16715913531165327,
+      "learning_rate": 0.00013898367951808052,
+      "loss": 0.4741,
+      "step": 1255
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18224361618494386,
+      "learning_rate": 0.000138897096109938,
+      "loss": 0.4852,
+      "step": 1256
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.17309322546124714,
+      "learning_rate": 0.00013881047832985886,
+      "loss": 0.4782,
+      "step": 1257
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.17208927952188605,
+      "learning_rate": 0.00013872382625438405,
+      "loss": 0.5383,
+      "step": 1258
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18195195367386732,
+      "learning_rate": 0.00013863713996008483,
+      "loss": 0.4793,
+      "step": 1259
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1869015589060158,
+      "learning_rate": 0.00013855041952356273,
+      "loss": 0.4918,
+      "step": 1260
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.22970064407630242,
+      "learning_rate": 0.00013846366502144936,
+      "loss": 0.5468,
+      "step": 1261
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18223297061392882,
+      "learning_rate": 0.00013837687653040653,
+      "loss": 0.4836,
+      "step": 1262
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1811453560251677,
+      "learning_rate": 0.00013829005412712607,
+      "loss": 0.4799,
+      "step": 1263
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19531829020393465,
+      "learning_rate": 0.00013820319788832968,
+      "loss": 0.4886,
+      "step": 1264
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1873613837354565,
+      "learning_rate": 0.0001381163078907691,
+      "loss": 0.5082,
+      "step": 1265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2725120336639838,
+      "learning_rate": 0.0001380293842112258,
+      "loss": 0.5434,
+      "step": 1266
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1844679328183984,
+      "learning_rate": 0.00013794242692651102,
+      "loss": 0.51,
+      "step": 1267
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.188306631118848,
+      "learning_rate": 0.00013785543611346578,
+      "loss": 0.4763,
+      "step": 1268
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19531189056502454,
+      "learning_rate": 0.00013776841184896064,
+      "loss": 0.5165,
+      "step": 1269
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20264440596591932,
+      "learning_rate": 0.00013768135420989577,
+      "loss": 0.5387,
+      "step": 1270
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2084557864213027,
+      "learning_rate": 0.00013759426327320074,
+      "loss": 0.5763,
+      "step": 1271
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20500727562381132,
+      "learning_rate": 0.0001375071391158347,
+      "loss": 0.5333,
+      "step": 1272
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18316050604278142,
+      "learning_rate": 0.00013741998181478603,
+      "loss": 0.4955,
+      "step": 1273
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20590646162784712,
+      "learning_rate": 0.00013733279144707245,
+      "loss": 0.519,
+      "step": 1274
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.17998508710498043,
+      "learning_rate": 0.00013724556808974086,
+      "loss": 0.4866,
+      "step": 1275
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20092486892395545,
+      "learning_rate": 0.0001371583118198674,
+      "loss": 0.5918,
+      "step": 1276
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1910150922839534,
+      "learning_rate": 0.0001370710227145572,
+      "loss": 0.5362,
+      "step": 1277
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18781648781714727,
+      "learning_rate": 0.00013698370085094442,
+      "loss": 0.5094,
+      "step": 1278
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.16979981901477645,
+      "learning_rate": 0.0001368963463061922,
+      "loss": 0.4665,
+      "step": 1279
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.17477774487542888,
+      "learning_rate": 0.0001368089591574926,
+      "loss": 0.5096,
+      "step": 1280
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18366095693364085,
+      "learning_rate": 0.00013672153948206635,
+      "loss": 0.4843,
+      "step": 1281
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.17419423552171015,
+      "learning_rate": 0.00013663408735716307,
+      "loss": 0.4886,
+      "step": 1282
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18711956168861,
+      "learning_rate": 0.00013654660286006095,
+      "loss": 0.5093,
+      "step": 1283
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.17966945439077842,
+      "learning_rate": 0.0001364590860680669,
+      "loss": 0.5146,
+      "step": 1284
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1762512874090438,
+      "learning_rate": 0.00013637153705851616,
+      "loss": 0.4863,
+      "step": 1285
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1851970347695891,
+      "learning_rate": 0.00013628395590877277,
+      "loss": 0.5562,
+      "step": 1286
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19299530099729093,
+      "learning_rate": 0.00013619634269622884,
+      "loss": 0.4686,
+      "step": 1287
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19014321879289434,
+      "learning_rate": 0.00013610869749830498,
+      "loss": 0.5187,
+      "step": 1288
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1851421397455001,
+      "learning_rate": 0.00013602102039245002,
+      "loss": 0.5738,
+      "step": 1289
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18412094053224104,
+      "learning_rate": 0.00013593331145614104,
+      "loss": 0.5022,
+      "step": 1290
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.17961356701720557,
+      "learning_rate": 0.00013584557076688322,
+      "loss": 0.5351,
+      "step": 1291
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1705319278777615,
+      "learning_rate": 0.00013575779840220976,
+      "loss": 0.5084,
+      "step": 1292
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18019495479729933,
+      "learning_rate": 0.00013566999443968185,
+      "loss": 0.5236,
+      "step": 1293
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.17974898003188056,
+      "learning_rate": 0.00013558215895688867,
+      "loss": 0.542,
+      "step": 1294
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1797247167748882,
+      "learning_rate": 0.00013549429203144723,
+      "loss": 0.5077,
+      "step": 1295
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1850550416693565,
+      "learning_rate": 0.00013540639374100226,
+      "loss": 0.5396,
+      "step": 1296
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1733382663357764,
+      "learning_rate": 0.00013531846416322627,
+      "loss": 0.5239,
+      "step": 1297
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.17977034098323824,
+      "learning_rate": 0.00013523050337581943,
+      "loss": 0.4717,
+      "step": 1298
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18483046559711522,
+      "learning_rate": 0.0001351425114565094,
+      "loss": 0.5378,
+      "step": 1299
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.17834941323623535,
+      "learning_rate": 0.0001350544884830515,
+      "loss": 0.5662,
+      "step": 1300
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18309245782790126,
+      "learning_rate": 0.00013496643453322828,
+      "loss": 0.5812,
+      "step": 1301
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19808755077064702,
+      "learning_rate": 0.0001348783496848499,
+      "loss": 0.5721,
+      "step": 1302
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18923943372482865,
+      "learning_rate": 0.00013479023401575366,
+      "loss": 0.5383,
+      "step": 1303
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.20010184755781324,
+      "learning_rate": 0.00013470208760380412,
+      "loss": 0.4803,
+      "step": 1304
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2759865843375521,
+      "learning_rate": 0.0001346139105268931,
+      "loss": 0.5583,
+      "step": 1305
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.17317696637576427,
+      "learning_rate": 0.00013452570286293938,
+      "loss": 0.4797,
+      "step": 1306
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1915090235119688,
+      "learning_rate": 0.00013443746468988884,
+      "loss": 0.467,
+      "step": 1307
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.20256119138747514,
+      "learning_rate": 0.00013434919608571437,
+      "loss": 0.5311,
+      "step": 1308
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.17993484843579866,
+      "learning_rate": 0.00013426089712841564,
+      "loss": 0.4832,
+      "step": 1309
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18038829097640488,
+      "learning_rate": 0.00013417256789601925,
+      "loss": 0.5688,
+      "step": 1310
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19541363045484325,
+      "learning_rate": 0.00013408420846657844,
+      "loss": 0.5244,
+      "step": 1311
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19899174638442974,
+      "learning_rate": 0.00013399581891817324,
+      "loss": 0.5588,
+      "step": 1312
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1882127221313746,
+      "learning_rate": 0.00013390739932891022,
+      "loss": 0.4932,
+      "step": 1313
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1823628172344837,
+      "learning_rate": 0.00013381894977692257,
+      "loss": 0.5594,
+      "step": 1314
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.17992657977066912,
+      "learning_rate": 0.00013373047034036988,
+      "loss": 0.5509,
+      "step": 1315
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18381185146899134,
+      "learning_rate": 0.0001336419610974382,
+      "loss": 0.5209,
+      "step": 1316
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.16713876415743062,
+      "learning_rate": 0.00013355342212633986,
+      "loss": 0.4726,
+      "step": 1317
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1945012456185686,
+      "learning_rate": 0.0001334648535053136,
+      "loss": 0.5733,
+      "step": 1318
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1832463509617787,
+      "learning_rate": 0.00013337625531262414,
+      "loss": 0.5257,
+      "step": 1319
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.17703976530840718,
+      "learning_rate": 0.0001332876276265625,
+      "loss": 0.4854,
+      "step": 1320
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1857229688186767,
+      "learning_rate": 0.00013319897052544577,
+      "loss": 0.4803,
+      "step": 1321
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18149715939246025,
+      "learning_rate": 0.00013311028408761688,
+      "loss": 0.4886,
+      "step": 1322
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18600803718548015,
+      "learning_rate": 0.00013302156839144484,
+      "loss": 0.5098,
+      "step": 1323
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18184709807791666,
+      "learning_rate": 0.00013293282351532442,
+      "loss": 0.5253,
+      "step": 1324
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.17926319053458772,
+      "learning_rate": 0.00013284404953767625,
+      "loss": 0.5112,
+      "step": 1325
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18473321145802576,
+      "learning_rate": 0.00013275524653694665,
+      "loss": 0.5257,
+      "step": 1326
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.17432084959871605,
+      "learning_rate": 0.00013266641459160753,
+      "loss": 0.4819,
+      "step": 1327
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2136616093736039,
+      "learning_rate": 0.0001325775537801564,
+      "loss": 0.5289,
+      "step": 1328
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18909582251032753,
+      "learning_rate": 0.00013248866418111635,
+      "loss": 0.5369,
+      "step": 1329
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1915972979886006,
+      "learning_rate": 0.00013239974587303584,
+      "loss": 0.5926,
+      "step": 1330
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18549724952330918,
+      "learning_rate": 0.00013231079893448873,
+      "loss": 0.4909,
+      "step": 1331
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1986357464763957,
+      "learning_rate": 0.00013222182344407415,
+      "loss": 0.5257,
+      "step": 1332
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1649013077802235,
+      "learning_rate": 0.00013213281948041647,
+      "loss": 0.4578,
+      "step": 1333
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.24324520525042556,
+      "learning_rate": 0.0001320437871221652,
+      "loss": 0.5119,
+      "step": 1334
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19388615550793745,
+      "learning_rate": 0.00013195472644799504,
+      "loss": 0.4655,
+      "step": 1335
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.16279182912251683,
+      "learning_rate": 0.00013186563753660562,
+      "loss": 0.4868,
+      "step": 1336
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.17463318249809762,
+      "learning_rate": 0.0001317765204667215,
+      "loss": 0.5026,
+      "step": 1337
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.17843994655748693,
+      "learning_rate": 0.0001316873753170922,
+      "loss": 0.4725,
+      "step": 1338
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.17705143987562852,
+      "learning_rate": 0.00013159820216649198,
+      "loss": 0.4835,
+      "step": 1339
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2071900675528722,
+      "learning_rate": 0.00013150900109371998,
+      "loss": 0.5452,
+      "step": 1340
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1855602116667224,
+      "learning_rate": 0.00013141977217759977,
+      "loss": 0.5166,
+      "step": 1341
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19368584941752956,
+      "learning_rate": 0.00013133051549697977,
+      "loss": 0.5258,
+      "step": 1342
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19453794357182327,
+      "learning_rate": 0.00013124123113073278,
+      "loss": 0.5039,
+      "step": 1343
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.16449765544951725,
+      "learning_rate": 0.0001311519191577562,
+      "loss": 0.4935,
+      "step": 1344
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1817628748335402,
+      "learning_rate": 0.0001310625796569717,
+      "loss": 0.4937,
+      "step": 1345
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2142696862792987,
+      "learning_rate": 0.00013097321270732524,
+      "loss": 0.5004,
+      "step": 1346
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2076553608561363,
+      "learning_rate": 0.0001308838183877872,
+      "loss": 0.499,
+      "step": 1347
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1694419193937227,
+      "learning_rate": 0.00013079439677735207,
+      "loss": 0.477,
+      "step": 1348
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.20906856602739568,
+      "learning_rate": 0.0001307049479550384,
+      "loss": 0.5227,
+      "step": 1349
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.17644464533212287,
+      "learning_rate": 0.00013061547199988885,
+      "loss": 0.5636,
+      "step": 1350
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18582243028910475,
+      "learning_rate": 0.00013052596899097005,
+      "loss": 0.4588,
+      "step": 1351
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19864208660134122,
+      "learning_rate": 0.0001304364390073725,
+      "loss": 0.5367,
+      "step": 1352
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2025705912016794,
+      "learning_rate": 0.00013034688212821058,
+      "loss": 0.4989,
+      "step": 1353
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1850384411891724,
+      "learning_rate": 0.00013025729843262241,
+      "loss": 0.4829,
+      "step": 1354
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18112685382941437,
+      "learning_rate": 0.00013016768799976983,
+      "loss": 0.501,
+      "step": 1355
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19083770143909534,
+      "learning_rate": 0.00013007805090883826,
+      "loss": 0.4693,
+      "step": 1356
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1741168713346763,
+      "learning_rate": 0.00012998838723903675,
+      "loss": 0.5051,
+      "step": 1357
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.17998449629698884,
+      "learning_rate": 0.00012989869706959777,
+      "loss": 0.5192,
+      "step": 1358
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.22733876158298189,
+      "learning_rate": 0.0001298089804797772,
+      "loss": 0.5699,
+      "step": 1359
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18706763134147958,
+      "learning_rate": 0.00012971923754885438,
+      "loss": 0.5352,
+      "step": 1360
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.17137720274694657,
+      "learning_rate": 0.0001296294683561318,
+      "loss": 0.5028,
+      "step": 1361
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1706735720997184,
+      "learning_rate": 0.00012953967298093513,
+      "loss": 0.5152,
+      "step": 1362
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1877012462021262,
+      "learning_rate": 0.00012944985150261341,
+      "loss": 0.4895,
+      "step": 1363
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18521525521633062,
+      "learning_rate": 0.00012936000400053845,
+      "loss": 0.5483,
+      "step": 1364
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.175077365429483,
+      "learning_rate": 0.00012927013055410522,
+      "loss": 0.5405,
+      "step": 1365
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2453533595215046,
+      "learning_rate": 0.00012918023124273165,
+      "loss": 0.5118,
+      "step": 1366
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.20592576443274452,
+      "learning_rate": 0.00012909030614585836,
+      "loss": 0.5008,
+      "step": 1367
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19466804271674717,
+      "learning_rate": 0.00012900035534294893,
+      "loss": 0.4973,
+      "step": 1368
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18407694211273226,
+      "learning_rate": 0.00012891037891348957,
+      "loss": 0.4928,
+      "step": 1369
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1944897009288739,
+      "learning_rate": 0.00012882037693698917,
+      "loss": 0.5042,
+      "step": 1370
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1876729080378583,
+      "learning_rate": 0.00012873034949297912,
+      "loss": 0.5565,
+      "step": 1371
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18740458209713104,
+      "learning_rate": 0.0001286402966610134,
+      "loss": 0.4912,
+      "step": 1372
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.17895655124360504,
+      "learning_rate": 0.00012855021852066842,
+      "loss": 0.4735,
+      "step": 1373
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.21851593513203868,
+      "learning_rate": 0.00012846011515154287,
+      "loss": 0.5548,
+      "step": 1374
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19220669747584535,
+      "learning_rate": 0.00012836998663325782,
+      "loss": 0.5339,
+      "step": 1375
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1877766868834106,
+      "learning_rate": 0.00012827983304545656,
+      "loss": 0.5379,
+      "step": 1376
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1781170900201972,
+      "learning_rate": 0.00012818965446780448,
+      "loss": 0.4614,
+      "step": 1377
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.17271334480103068,
+      "learning_rate": 0.00012809945097998907,
+      "loss": 0.4667,
+      "step": 1378
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19192709550266393,
+      "learning_rate": 0.00012800922266171987,
+      "loss": 0.539,
+      "step": 1379
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18347756297917528,
+      "learning_rate": 0.0001279189695927283,
+      "loss": 0.4869,
+      "step": 1380
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18435293599254082,
+      "learning_rate": 0.0001278286918527677,
+      "loss": 0.5142,
+      "step": 1381
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.17568604254392264,
+      "learning_rate": 0.00012773838952161322,
+      "loss": 0.4944,
+      "step": 1382
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19863759493863561,
+      "learning_rate": 0.0001276480626790617,
+      "loss": 0.5071,
+      "step": 1383
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18656826295630696,
+      "learning_rate": 0.00012755771140493167,
+      "loss": 0.4743,
+      "step": 1384
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.17629250804517016,
+      "learning_rate": 0.0001274673357790632,
+      "loss": 0.5198,
+      "step": 1385
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.17537444688807757,
+      "learning_rate": 0.00012737693588131793,
+      "loss": 0.49,
+      "step": 1386
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1739583842649344,
+      "learning_rate": 0.00012728651179157895,
+      "loss": 0.4826,
+      "step": 1387
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19774251629485748,
+      "learning_rate": 0.00012719606358975073,
+      "loss": 0.5571,
+      "step": 1388
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.17736202390835257,
+      "learning_rate": 0.00012710559135575895,
+      "loss": 0.4928,
+      "step": 1389
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18721704864956284,
+      "learning_rate": 0.00012701509516955067,
+      "loss": 0.4798,
+      "step": 1390
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19216069335454633,
+      "learning_rate": 0.00012692457511109402,
+      "loss": 0.5268,
+      "step": 1391
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.20170703615044985,
+      "learning_rate": 0.00012683403126037825,
+      "loss": 0.5556,
+      "step": 1392
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.17163471940645972,
+      "learning_rate": 0.00012674346369741365,
+      "loss": 0.5224,
+      "step": 1393
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.168318318516553,
+      "learning_rate": 0.0001266528725022315,
+      "loss": 0.5071,
+      "step": 1394
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.17605642876202626,
+      "learning_rate": 0.00012656225775488383,
+      "loss": 0.5356,
+      "step": 1395
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.17094228936642658,
+      "learning_rate": 0.0001264716195354436,
+      "loss": 0.4705,
+      "step": 1396
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.17472714805803075,
+      "learning_rate": 0.00012638095792400452,
+      "loss": 0.4903,
+      "step": 1397
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.17866538442934543,
+      "learning_rate": 0.00012629027300068088,
+      "loss": 0.5151,
+      "step": 1398
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.16752002793148427,
+      "learning_rate": 0.0001261995648456076,
+      "loss": 0.5094,
+      "step": 1399
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18374616312572198,
+      "learning_rate": 0.00012610883353894026,
+      "loss": 0.4547,
+      "step": 1400
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.17615421249336274,
+      "learning_rate": 0.00012601807916085461,
+      "loss": 0.5194,
+      "step": 1401
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.17640888040569758,
+      "learning_rate": 0.00012592730179154712,
+      "loss": 0.4926,
+      "step": 1402
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.17818443729463354,
+      "learning_rate": 0.0001258365015112344,
+      "loss": 0.5138,
+      "step": 1403
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18271989198154837,
+      "learning_rate": 0.00012574567840015324,
+      "loss": 0.5012,
+      "step": 1404
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1828993894744449,
+      "learning_rate": 0.00012565483253856071,
+      "loss": 0.4917,
+      "step": 1405
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.17537537909673637,
+      "learning_rate": 0.00012556396400673403,
+      "loss": 0.4856,
+      "step": 1406
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1791770204684942,
+      "learning_rate": 0.00012547307288497035,
+      "loss": 0.4993,
+      "step": 1407
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1926761899524845,
+      "learning_rate": 0.00012538215925358688,
+      "loss": 0.532,
+      "step": 1408
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1641415902243254,
+      "learning_rate": 0.00012529122319292053,
+      "loss": 0.4994,
+      "step": 1409
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18035208365167332,
+      "learning_rate": 0.00012520026478332822,
+      "loss": 0.457,
+      "step": 1410
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1767179210201008,
+      "learning_rate": 0.00012510928410518663,
+      "loss": 0.515,
+      "step": 1411
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.16556221078273253,
+      "learning_rate": 0.00012501828123889194,
+      "loss": 0.4578,
+      "step": 1412
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.17248580048933712,
+      "learning_rate": 0.00012492725626486013,
+      "loss": 0.4818,
+      "step": 1413
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.17049351620509895,
+      "learning_rate": 0.00012483620926352656,
+      "loss": 0.4889,
+      "step": 1414
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2445103047102675,
+      "learning_rate": 0.00012474514031534617,
+      "loss": 0.543,
+      "step": 1415
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1832392790663422,
+      "learning_rate": 0.00012465404950079325,
+      "loss": 0.5609,
+      "step": 1416
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18408303780573687,
+      "learning_rate": 0.00012456293690036135,
+      "loss": 0.489,
+      "step": 1417
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18185382443521964,
+      "learning_rate": 0.00012447180259456342,
+      "loss": 0.5156,
+      "step": 1418
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19471472250974153,
+      "learning_rate": 0.00012438064666393144,
+      "loss": 0.493,
+      "step": 1419
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19996647014153662,
+      "learning_rate": 0.00012428946918901655,
+      "loss": 0.5355,
+      "step": 1420
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.17841905649814713,
+      "learning_rate": 0.00012419827025038905,
+      "loss": 0.5193,
+      "step": 1421
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.23618023782431,
+      "learning_rate": 0.00012410704992863792,
+      "loss": 0.5052,
+      "step": 1422
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19206431381415717,
+      "learning_rate": 0.00012401580830437135,
+      "loss": 0.5061,
+      "step": 1423
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18232605343034466,
+      "learning_rate": 0.0001239245454582162,
+      "loss": 0.4726,
+      "step": 1424
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1667177719369294,
+      "learning_rate": 0.000123833261470818,
+      "loss": 0.4584,
+      "step": 1425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18664991662366204,
+      "learning_rate": 0.0001237419564228412,
+      "loss": 0.4984,
+      "step": 1426
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18842903468293132,
+      "learning_rate": 0.00012365063039496862,
+      "loss": 0.5021,
+      "step": 1427
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.17356657187313135,
+      "learning_rate": 0.00012355928346790174,
+      "loss": 0.4677,
+      "step": 1428
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19172179721922822,
+      "learning_rate": 0.0001234679157223605,
+      "loss": 0.5298,
+      "step": 1429
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.17844753388051934,
+      "learning_rate": 0.00012337652723908325,
+      "loss": 0.5179,
+      "step": 1430
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.17488823189990194,
+      "learning_rate": 0.0001232851180988266,
+      "loss": 0.504,
+      "step": 1431
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19923950835484613,
+      "learning_rate": 0.00012319368838236547,
+      "loss": 0.4801,
+      "step": 1432
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18421010540531865,
+      "learning_rate": 0.00012310223817049292,
+      "loss": 0.4874,
+      "step": 1433
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18280627148010795,
+      "learning_rate": 0.00012301076754402018,
+      "loss": 0.4807,
+      "step": 1434
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18255969688709006,
+      "learning_rate": 0.00012291927658377648,
+      "loss": 0.5507,
+      "step": 1435
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19369212033571756,
+      "learning_rate": 0.00012282776537060903,
+      "loss": 0.4807,
+      "step": 1436
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.17045351156650784,
+      "learning_rate": 0.0001227362339853829,
+      "loss": 0.5108,
+      "step": 1437
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18805439573471683,
+      "learning_rate": 0.000122644682508981,
+      "loss": 0.5024,
+      "step": 1438
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18557193779276274,
+      "learning_rate": 0.0001225531110223041,
+      "loss": 0.4968,
+      "step": 1439
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18446256740657319,
+      "learning_rate": 0.00012246151960627053,
+      "loss": 0.5296,
+      "step": 1440
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1860920392864418,
+      "learning_rate": 0.0001223699083418162,
+      "loss": 0.5143,
+      "step": 1441
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2044775732039818,
+      "learning_rate": 0.00012227827730989466,
+      "loss": 0.5057,
+      "step": 1442
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18607693244293935,
+      "learning_rate": 0.00012218662659147693,
+      "loss": 0.5174,
+      "step": 1443
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18611453439901565,
+      "learning_rate": 0.00012209495626755134,
+      "loss": 0.4989,
+      "step": 1444
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19583337708028287,
+      "learning_rate": 0.00012200326641912361,
+      "loss": 0.5225,
+      "step": 1445
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19901404620525662,
+      "learning_rate": 0.00012191155712721667,
+      "loss": 0.5108,
+      "step": 1446
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2034918139139227,
+      "learning_rate": 0.0001218198284728707,
+      "loss": 0.53,
+      "step": 1447
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19938382509011504,
+      "learning_rate": 0.00012172808053714292,
+      "loss": 0.4907,
+      "step": 1448
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1979156919673814,
+      "learning_rate": 0.00012163631340110764,
+      "loss": 0.4977,
+      "step": 1449
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1773307028514164,
+      "learning_rate": 0.00012154452714585605,
+      "loss": 0.5305,
+      "step": 1450
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1822927258357729,
+      "learning_rate": 0.00012145272185249634,
+      "loss": 0.4921,
+      "step": 1451
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19519341132650003,
+      "learning_rate": 0.0001213608976021535,
+      "loss": 0.4735,
+      "step": 1452
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18840961162083253,
+      "learning_rate": 0.00012126905447596921,
+      "loss": 0.5396,
+      "step": 1453
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18484188418670194,
+      "learning_rate": 0.00012117719255510188,
+      "loss": 0.4916,
+      "step": 1454
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1787508136445565,
+      "learning_rate": 0.00012108531192072652,
+      "loss": 0.4667,
+      "step": 1455
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.17143949715387075,
+      "learning_rate": 0.0001209934126540347,
+      "loss": 0.4869,
+      "step": 1456
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19700823060671513,
+      "learning_rate": 0.00012090149483623438,
+      "loss": 0.4673,
+      "step": 1457
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18571871176182977,
+      "learning_rate": 0.00012080955854855002,
+      "loss": 0.5284,
+      "step": 1458
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19439561697487548,
+      "learning_rate": 0.00012071760387222229,
+      "loss": 0.5232,
+      "step": 1459
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19667065535729555,
+      "learning_rate": 0.0001206256308885082,
+      "loss": 0.5436,
+      "step": 1460
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18778267851212818,
+      "learning_rate": 0.00012053363967868092,
+      "loss": 0.514,
+      "step": 1461
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18611375927652626,
+      "learning_rate": 0.00012044163032402965,
+      "loss": 0.4708,
+      "step": 1462
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18484579524440076,
+      "learning_rate": 0.0001203496029058597,
+      "loss": 0.464,
+      "step": 1463
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1826656462774754,
+      "learning_rate": 0.00012025755750549233,
+      "loss": 0.4919,
+      "step": 1464
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18150683810049317,
+      "learning_rate": 0.00012016549420426471,
+      "loss": 0.4936,
+      "step": 1465
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18124559007552254,
+      "learning_rate": 0.00012007341308352977,
+      "loss": 0.4972,
+      "step": 1466
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19182333070322963,
+      "learning_rate": 0.00011998131422465621,
+      "loss": 0.4856,
+      "step": 1467
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.178004127767968,
+      "learning_rate": 0.00011988919770902845,
+      "loss": 0.4818,
+      "step": 1468
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.17292199373403655,
+      "learning_rate": 0.00011979706361804644,
+      "loss": 0.4829,
+      "step": 1469
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18196556074734327,
+      "learning_rate": 0.00011970491203312568,
+      "loss": 0.5003,
+      "step": 1470
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18754874881279202,
+      "learning_rate": 0.0001196127430356972,
+      "loss": 0.5743,
+      "step": 1471
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1694899960861821,
+      "learning_rate": 0.00011952055670720732,
+      "loss": 0.4812,
+      "step": 1472
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.17685121597710485,
+      "learning_rate": 0.00011942835312911773,
+      "loss": 0.5106,
+      "step": 1473
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18468147975405122,
+      "learning_rate": 0.00011933613238290535,
+      "loss": 0.4881,
+      "step": 1474
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19502651305818872,
+      "learning_rate": 0.00011924389455006226,
+      "loss": 0.5436,
+      "step": 1475
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1912752770383613,
+      "learning_rate": 0.00011915163971209566,
+      "loss": 0.4975,
+      "step": 1476
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18045614062303392,
+      "learning_rate": 0.00011905936795052774,
+      "loss": 0.4946,
+      "step": 1477
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19188311700454608,
+      "learning_rate": 0.0001189670793468957,
+      "loss": 0.5096,
+      "step": 1478
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.17844071972998474,
+      "learning_rate": 0.00011887477398275162,
+      "loss": 0.4932,
+      "step": 1479
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1825369869650523,
+      "learning_rate": 0.00011878245193966229,
+      "loss": 0.5087,
+      "step": 1480
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.187073417853317,
+      "learning_rate": 0.00011869011329920936,
+      "loss": 0.4916,
+      "step": 1481
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18205562507913237,
+      "learning_rate": 0.00011859775814298905,
+      "loss": 0.5233,
+      "step": 1482
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18850193307348725,
+      "learning_rate": 0.0001185053865526123,
+      "loss": 0.509,
+      "step": 1483
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.16598501445752575,
+      "learning_rate": 0.00011841299860970445,
+      "loss": 0.4606,
+      "step": 1484
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18164697924930495,
+      "learning_rate": 0.00011832059439590533,
+      "loss": 0.5326,
+      "step": 1485
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1744542009553421,
+      "learning_rate": 0.00011822817399286916,
+      "loss": 0.4806,
+      "step": 1486
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1866112771984086,
+      "learning_rate": 0.00011813573748226447,
+      "loss": 0.5142,
+      "step": 1487
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1804334025313157,
+      "learning_rate": 0.00011804328494577402,
+      "loss": 0.5055,
+      "step": 1488
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1827107380774223,
+      "learning_rate": 0.00011795081646509469,
+      "loss": 0.4705,
+      "step": 1489
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.17671386861052463,
+      "learning_rate": 0.00011785833212193749,
+      "loss": 0.5103,
+      "step": 1490
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18983228560014775,
+      "learning_rate": 0.00011776583199802746,
+      "loss": 0.5105,
+      "step": 1491
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.17998266735444157,
+      "learning_rate": 0.00011767331617510358,
+      "loss": 0.5207,
+      "step": 1492
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.17711101362820791,
+      "learning_rate": 0.00011758078473491864,
+      "loss": 0.5266,
+      "step": 1493
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18914416644450918,
+      "learning_rate": 0.00011748823775923934,
+      "loss": 0.5338,
+      "step": 1494
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18057892909261303,
+      "learning_rate": 0.00011739567532984598,
+      "loss": 0.5549,
+      "step": 1495
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.16973466919637561,
+      "learning_rate": 0.00011730309752853261,
+      "loss": 0.4838,
+      "step": 1496
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18240256763305998,
+      "learning_rate": 0.00011721050443710688,
+      "loss": 0.4946,
+      "step": 1497
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18071282928562282,
+      "learning_rate": 0.00011711789613738986,
+      "loss": 0.5128,
+      "step": 1498
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1764697230658466,
+      "learning_rate": 0.00011702527271121609,
+      "loss": 0.5081,
+      "step": 1499
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18036733171638797,
+      "learning_rate": 0.00011693263424043353,
+      "loss": 0.492,
+      "step": 1500
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18783837144683072,
+      "learning_rate": 0.00011683998080690334,
+      "loss": 0.5066,
+      "step": 1501
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1748875971730362,
+      "learning_rate": 0.00011674731249250008,
+      "loss": 0.5045,
+      "step": 1502
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18835075966322473,
+      "learning_rate": 0.00011665462937911124,
+      "loss": 0.5153,
+      "step": 1503
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.16522307101295688,
+      "learning_rate": 0.00011656193154863749,
+      "loss": 0.4485,
+      "step": 1504
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1743041745662128,
+      "learning_rate": 0.00011646921908299254,
+      "loss": 0.4668,
+      "step": 1505
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18915730869999403,
+      "learning_rate": 0.00011637649206410298,
+      "loss": 0.5169,
+      "step": 1506
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1787635275807655,
+      "learning_rate": 0.00011628375057390824,
+      "loss": 0.5218,
+      "step": 1507
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18297477342275548,
+      "learning_rate": 0.00011619099469436061,
+      "loss": 0.497,
+      "step": 1508
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1780890941415051,
+      "learning_rate": 0.00011609822450742507,
+      "loss": 0.4961,
+      "step": 1509
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.17078744735386345,
+      "learning_rate": 0.0001160054400950792,
+      "loss": 0.4832,
+      "step": 1510
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1784315880273548,
+      "learning_rate": 0.00011591264153931321,
+      "loss": 0.4745,
+      "step": 1511
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1871274533549485,
+      "learning_rate": 0.00011581982892212975,
+      "loss": 0.5072,
+      "step": 1512
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18545283275745875,
+      "learning_rate": 0.0001157270023255439,
+      "loss": 0.4982,
+      "step": 1513
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18159317342397205,
+      "learning_rate": 0.00011563416183158318,
+      "loss": 0.5188,
+      "step": 1514
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20002147439421913,
+      "learning_rate": 0.00011554130752228731,
+      "loss": 0.5472,
+      "step": 1515
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18673799334263,
+      "learning_rate": 0.00011544843947970822,
+      "loss": 0.469,
+      "step": 1516
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2099300853781104,
+      "learning_rate": 0.00011535555778590999,
+      "loss": 0.5273,
+      "step": 1517
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18505904288666042,
+      "learning_rate": 0.00011526266252296876,
+      "loss": 0.5222,
+      "step": 1518
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2022588519204151,
+      "learning_rate": 0.0001151697537729727,
+      "loss": 0.5154,
+      "step": 1519
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.17874437237649884,
+      "learning_rate": 0.00011507683161802184,
+      "loss": 0.5493,
+      "step": 1520
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.16382256794398076,
+      "learning_rate": 0.00011498389614022807,
+      "loss": 0.4838,
+      "step": 1521
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1788771786525409,
+      "learning_rate": 0.00011489094742171502,
+      "loss": 0.5068,
+      "step": 1522
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19844148701897005,
+      "learning_rate": 0.00011479798554461818,
+      "loss": 0.499,
+      "step": 1523
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18029997805524606,
+      "learning_rate": 0.0001147050105910845,
+      "loss": 0.5218,
+      "step": 1524
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2070583771119555,
+      "learning_rate": 0.00011461202264327246,
+      "loss": 0.5527,
+      "step": 1525
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.17683689253976204,
+      "learning_rate": 0.00011451902178335219,
+      "loss": 0.475,
+      "step": 1526
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1746582188063431,
+      "learning_rate": 0.0001144260080935051,
+      "loss": 0.5002,
+      "step": 1527
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19411077398249144,
+      "learning_rate": 0.00011433298165592396,
+      "loss": 0.4862,
+      "step": 1528
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1873910047084602,
+      "learning_rate": 0.00011423994255281285,
+      "loss": 0.4882,
+      "step": 1529
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18794249715310352,
+      "learning_rate": 0.000114146890866387,
+      "loss": 0.543,
+      "step": 1530
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1795844721124014,
+      "learning_rate": 0.00011405382667887276,
+      "loss": 0.5305,
+      "step": 1531
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.17955259790256417,
+      "learning_rate": 0.00011396075007250758,
+      "loss": 0.4973,
+      "step": 1532
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.21366812260453688,
+      "learning_rate": 0.00011386766112953977,
+      "loss": 0.5278,
+      "step": 1533
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18167687811794028,
+      "learning_rate": 0.00011377455993222867,
+      "loss": 0.5249,
+      "step": 1534
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.17645370527975052,
+      "learning_rate": 0.00011368144656284436,
+      "loss": 0.4947,
+      "step": 1535
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.16934323070380375,
+      "learning_rate": 0.00011358832110366775,
+      "loss": 0.4616,
+      "step": 1536
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.17196203109279234,
+      "learning_rate": 0.00011349518363699036,
+      "loss": 0.4638,
+      "step": 1537
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1943534034058594,
+      "learning_rate": 0.00011340203424511434,
+      "loss": 0.5094,
+      "step": 1538
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.17128940028769005,
+      "learning_rate": 0.00011330887301035242,
+      "loss": 0.4624,
+      "step": 1539
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.179017474107053,
+      "learning_rate": 0.00011321570001502775,
+      "loss": 0.4872,
+      "step": 1540
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.17288738113525648,
+      "learning_rate": 0.00011312251534147387,
+      "loss": 0.4754,
+      "step": 1541
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.17577830411270592,
+      "learning_rate": 0.0001130293190720347,
+      "loss": 0.4729,
+      "step": 1542
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.16957226974611797,
+      "learning_rate": 0.00011293611128906431,
+      "loss": 0.4814,
+      "step": 1543
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1760426058506451,
+      "learning_rate": 0.00011284289207492706,
+      "loss": 0.4619,
+      "step": 1544
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18329235131533697,
+      "learning_rate": 0.00011274966151199731,
+      "loss": 0.5044,
+      "step": 1545
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18307223453280677,
+      "learning_rate": 0.00011265641968265945,
+      "loss": 0.5241,
+      "step": 1546
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.17773989766788875,
+      "learning_rate": 0.00011256316666930798,
+      "loss": 0.4925,
+      "step": 1547
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1769674765097833,
+      "learning_rate": 0.00011246990255434704,
+      "loss": 0.5262,
+      "step": 1548
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1849871149599951,
+      "learning_rate": 0.00011237662742019075,
+      "loss": 0.4594,
+      "step": 1549
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18867364566508427,
+      "learning_rate": 0.00011228334134926297,
+      "loss": 0.476,
+      "step": 1550
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.17417769451068804,
+      "learning_rate": 0.00011219004442399712,
+      "loss": 0.4972,
+      "step": 1551
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.16859318717354108,
+      "learning_rate": 0.00011209673672683632,
+      "loss": 0.4802,
+      "step": 1552
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1644307458626678,
+      "learning_rate": 0.00011200341834023309,
+      "loss": 0.4829,
+      "step": 1553
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1989496601102084,
+      "learning_rate": 0.00011191008934664951,
+      "loss": 0.6644,
+      "step": 1554
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1844631029567801,
+      "learning_rate": 0.000111816749828557,
+      "loss": 0.4901,
+      "step": 1555
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19816902332331113,
+      "learning_rate": 0.00011172339986843626,
+      "loss": 0.5496,
+      "step": 1556
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19691336622769823,
+      "learning_rate": 0.00011163003954877718,
+      "loss": 0.5586,
+      "step": 1557
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18166390903396584,
+      "learning_rate": 0.00011153666895207885,
+      "loss": 0.5509,
+      "step": 1558
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19059509070126388,
+      "learning_rate": 0.00011144328816084952,
+      "loss": 0.5301,
+      "step": 1559
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1907238818250723,
+      "learning_rate": 0.00011134989725760632,
+      "loss": 0.5015,
+      "step": 1560
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18292558413429197,
+      "learning_rate": 0.00011125649632487538,
+      "loss": 0.5246,
+      "step": 1561
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.17463088614239725,
+      "learning_rate": 0.00011116308544519163,
+      "loss": 0.4912,
+      "step": 1562
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19109769957786926,
+      "learning_rate": 0.00011106966470109888,
+      "loss": 0.5016,
+      "step": 1563
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.20891602551163976,
+      "learning_rate": 0.00011097623417514957,
+      "loss": 0.5061,
+      "step": 1564
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1912861078378425,
+      "learning_rate": 0.00011088279394990491,
+      "loss": 0.5087,
+      "step": 1565
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.20860613345682275,
+      "learning_rate": 0.00011078934410793453,
+      "loss": 0.5475,
+      "step": 1566
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2016276928245179,
+      "learning_rate": 0.00011069588473181663,
+      "loss": 0.4917,
+      "step": 1567
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18622143936681837,
+      "learning_rate": 0.00011060241590413787,
+      "loss": 0.4691,
+      "step": 1568
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1862650789358927,
+      "learning_rate": 0.0001105089377074932,
+      "loss": 0.4783,
+      "step": 1569
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.17314711363865778,
+      "learning_rate": 0.00011041545022448585,
+      "loss": 0.4636,
+      "step": 1570
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18882699212061288,
+      "learning_rate": 0.00011032195353772732,
+      "loss": 0.5071,
+      "step": 1571
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18313968659351176,
+      "learning_rate": 0.00011022844772983716,
+      "loss": 0.4925,
+      "step": 1572
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18989183467798568,
+      "learning_rate": 0.00011013493288344307,
+      "loss": 0.5138,
+      "step": 1573
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.20009639832032847,
+      "learning_rate": 0.00011004140908118069,
+      "loss": 0.5033,
+      "step": 1574
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1864822830628491,
+      "learning_rate": 0.00010994787640569348,
+      "loss": 0.4566,
+      "step": 1575
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18537930991361964,
+      "learning_rate": 0.00010985433493963294,
+      "loss": 0.4931,
+      "step": 1576
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1860930876562612,
+      "learning_rate": 0.00010976078476565818,
+      "loss": 0.4843,
+      "step": 1577
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2386895987379907,
+      "learning_rate": 0.00010966722596643607,
+      "loss": 0.5246,
+      "step": 1578
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.17624730923913154,
+      "learning_rate": 0.00010957365862464106,
+      "loss": 0.4471,
+      "step": 1579
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.17947142111167033,
+      "learning_rate": 0.00010948008282295523,
+      "loss": 0.5057,
+      "step": 1580
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.17182151942408264,
+      "learning_rate": 0.00010938649864406803,
+      "loss": 0.4452,
+      "step": 1581
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.17298006001721172,
+      "learning_rate": 0.0001092929061706764,
+      "loss": 0.4578,
+      "step": 1582
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2070838001590237,
+      "learning_rate": 0.00010919930548548456,
+      "loss": 0.4887,
+      "step": 1583
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18919686254137597,
+      "learning_rate": 0.00010910569667120402,
+      "loss": 0.5132,
+      "step": 1584
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18877811317785967,
+      "learning_rate": 0.0001090120798105534,
+      "loss": 0.4522,
+      "step": 1585
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.17183583532409152,
+      "learning_rate": 0.00010891845498625857,
+      "loss": 0.4938,
+      "step": 1586
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19365576655760885,
+      "learning_rate": 0.00010882482228105229,
+      "loss": 0.5192,
+      "step": 1587
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.17490010992510652,
+      "learning_rate": 0.00010873118177767433,
+      "loss": 0.4711,
+      "step": 1588
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.17342191737901563,
+      "learning_rate": 0.00010863753355887143,
+      "loss": 0.4775,
+      "step": 1589
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.16379055188844005,
+      "learning_rate": 0.00010854387770739707,
+      "loss": 0.4708,
+      "step": 1590
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1842692603288873,
+      "learning_rate": 0.00010845021430601143,
+      "loss": 0.4949,
+      "step": 1591
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18033303940018516,
+      "learning_rate": 0.00010835654343748149,
+      "loss": 0.4991,
+      "step": 1592
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.16891948671510196,
+      "learning_rate": 0.00010826286518458073,
+      "loss": 0.4001,
+      "step": 1593
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.17260620046518338,
+      "learning_rate": 0.00010816917963008916,
+      "loss": 0.4984,
+      "step": 1594
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18883680555663057,
+      "learning_rate": 0.00010807548685679334,
+      "loss": 0.5055,
+      "step": 1595
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.17805386706526016,
+      "learning_rate": 0.00010798178694748607,
+      "loss": 0.4544,
+      "step": 1596
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.16667712713064364,
+      "learning_rate": 0.00010788807998496655,
+      "loss": 0.4746,
+      "step": 1597
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18366350719934063,
+      "learning_rate": 0.00010779436605204017,
+      "loss": 0.5247,
+      "step": 1598
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.22274345582510527,
+      "learning_rate": 0.0001077006452315185,
+      "loss": 0.5095,
+      "step": 1599
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18493326300947324,
+      "learning_rate": 0.00010760691760621921,
+      "loss": 0.5327,
+      "step": 1600
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.16600394767674562,
+      "learning_rate": 0.00010751318325896592,
+      "loss": 0.498,
+      "step": 1601
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1788563275400356,
+      "learning_rate": 0.00010741944227258827,
+      "loss": 0.5264,
+      "step": 1602
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1757402482395372,
+      "learning_rate": 0.00010732569472992171,
+      "loss": 0.5285,
+      "step": 1603
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18663007232179554,
+      "learning_rate": 0.00010723194071380751,
+      "loss": 0.5189,
+      "step": 1604
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19049017159093576,
+      "learning_rate": 0.00010713818030709268,
+      "loss": 0.5003,
+      "step": 1605
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1788772885287754,
+      "learning_rate": 0.00010704441359262982,
+      "loss": 0.465,
+      "step": 1606
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2158972916651406,
+      "learning_rate": 0.00010695064065327712,
+      "loss": 0.5272,
+      "step": 1607
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1736784845280322,
+      "learning_rate": 0.00010685686157189832,
+      "loss": 0.5141,
+      "step": 1608
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19533897028956482,
+      "learning_rate": 0.00010676307643136254,
+      "loss": 0.5258,
+      "step": 1609
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19307284822375173,
+      "learning_rate": 0.00010666928531454428,
+      "loss": 0.5016,
+      "step": 1610
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.17903971380850872,
+      "learning_rate": 0.00010657548830432329,
+      "loss": 0.4497,
+      "step": 1611
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19164639321131477,
+      "learning_rate": 0.00010648168548358455,
+      "loss": 0.5137,
+      "step": 1612
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.17618034640166205,
+      "learning_rate": 0.00010638787693521819,
+      "loss": 0.5087,
+      "step": 1613
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18247646676718576,
+      "learning_rate": 0.00010629406274211934,
+      "loss": 0.5074,
+      "step": 1614
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.17276068231312514,
+      "learning_rate": 0.00010620024298718822,
+      "loss": 0.4787,
+      "step": 1615
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1915641904837803,
+      "learning_rate": 0.00010610641775332983,
+      "loss": 0.5306,
+      "step": 1616
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.17501613856715206,
+      "learning_rate": 0.00010601258712345414,
+      "loss": 0.5037,
+      "step": 1617
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18815087782526124,
+      "learning_rate": 0.00010591875118047588,
+      "loss": 0.5116,
+      "step": 1618
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.186783359821177,
+      "learning_rate": 0.00010582491000731432,
+      "loss": 0.5044,
+      "step": 1619
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.22993928050303203,
+      "learning_rate": 0.00010573106368689352,
+      "loss": 0.4821,
+      "step": 1620
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19722071181591044,
+      "learning_rate": 0.00010563721230214203,
+      "loss": 0.5189,
+      "step": 1621
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1809577680466395,
+      "learning_rate": 0.00010554335593599285,
+      "loss": 0.5428,
+      "step": 1622
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1773652413715408,
+      "learning_rate": 0.00010544949467138346,
+      "loss": 0.4978,
+      "step": 1623
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.16603963794042106,
+      "learning_rate": 0.00010535562859125558,
+      "loss": 0.4497,
+      "step": 1624
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19264000969327572,
+      "learning_rate": 0.0001052617577785552,
+      "loss": 0.5067,
+      "step": 1625
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19997471766094152,
+      "learning_rate": 0.00010516788231623253,
+      "loss": 0.4959,
+      "step": 1626
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1853233178226942,
+      "learning_rate": 0.00010507400228724192,
+      "loss": 0.5259,
+      "step": 1627
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.20841560813457816,
+      "learning_rate": 0.00010498011777454163,
+      "loss": 0.6294,
+      "step": 1628
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.17236196464793524,
+      "learning_rate": 0.000104886228861094,
+      "loss": 0.503,
+      "step": 1629
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.17688449011527507,
+      "learning_rate": 0.00010479233562986519,
+      "loss": 0.4859,
+      "step": 1630
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2609892130797167,
+      "learning_rate": 0.00010469843816382526,
+      "loss": 0.548,
+      "step": 1631
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19046903787621935,
+      "learning_rate": 0.0001046045365459479,
+      "loss": 0.5337,
+      "step": 1632
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18314320161449926,
+      "learning_rate": 0.00010451063085921056,
+      "loss": 0.4978,
+      "step": 1633
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19018122473676155,
+      "learning_rate": 0.00010441672118659422,
+      "loss": 0.4981,
+      "step": 1634
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1879539993234289,
+      "learning_rate": 0.00010432280761108342,
+      "loss": 0.502,
+      "step": 1635
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1957235753033609,
+      "learning_rate": 0.00010422889021566618,
+      "loss": 0.5232,
+      "step": 1636
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1830846528779905,
+      "learning_rate": 0.0001041349690833338,
+      "loss": 0.5185,
+      "step": 1637
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2110613700559306,
+      "learning_rate": 0.00010404104429708097,
+      "loss": 0.5631,
+      "step": 1638
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19406820066555858,
+      "learning_rate": 0.00010394711593990554,
+      "loss": 0.487,
+      "step": 1639
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17641222102986034,
+      "learning_rate": 0.00010385318409480862,
+      "loss": 0.4597,
+      "step": 1640
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1729485551615291,
+      "learning_rate": 0.00010375924884479427,
+      "loss": 0.4869,
+      "step": 1641
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17035672775207952,
+      "learning_rate": 0.00010366531027286967,
+      "loss": 0.4805,
+      "step": 1642
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1718925201317743,
+      "learning_rate": 0.00010357136846204487,
+      "loss": 0.4773,
+      "step": 1643
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17917233153986767,
+      "learning_rate": 0.00010347742349533278,
+      "loss": 0.4535,
+      "step": 1644
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19544887303797653,
+      "learning_rate": 0.00010338347545574916,
+      "loss": 0.5009,
+      "step": 1645
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1759404996666079,
+      "learning_rate": 0.00010328952442631241,
+      "loss": 0.462,
+      "step": 1646
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17490417817834802,
+      "learning_rate": 0.00010319557049004365,
+      "loss": 0.4648,
+      "step": 1647
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18605625193133354,
+      "learning_rate": 0.00010310161372996648,
+      "loss": 0.535,
+      "step": 1648
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18211716389824942,
+      "learning_rate": 0.00010300765422910706,
+      "loss": 0.482,
+      "step": 1649
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.195298449859917,
+      "learning_rate": 0.00010291369207049397,
+      "loss": 0.4966,
+      "step": 1650
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.20359810497996356,
+      "learning_rate": 0.00010281972733715808,
+      "loss": 0.4735,
+      "step": 1651
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18071892768830874,
+      "learning_rate": 0.00010272576011213262,
+      "loss": 0.5107,
+      "step": 1652
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19488059200805252,
+      "learning_rate": 0.00010263179047845297,
+      "loss": 0.5581,
+      "step": 1653
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1827251419622096,
+      "learning_rate": 0.00010253781851915663,
+      "loss": 0.5429,
+      "step": 1654
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1815977598840352,
+      "learning_rate": 0.0001024438443172832,
+      "loss": 0.4719,
+      "step": 1655
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1750166365728634,
+      "learning_rate": 0.00010234986795587418,
+      "loss": 0.4972,
+      "step": 1656
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.16860820894878417,
+      "learning_rate": 0.00010225588951797309,
+      "loss": 0.4881,
+      "step": 1657
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1653555565773814,
+      "learning_rate": 0.00010216190908662522,
+      "loss": 0.4759,
+      "step": 1658
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1973447838806209,
+      "learning_rate": 0.0001020679267448776,
+      "loss": 0.5361,
+      "step": 1659
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.17166094417441266,
+      "learning_rate": 0.00010197394257577902,
+      "loss": 0.4587,
+      "step": 1660
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.17920046555821298,
+      "learning_rate": 0.00010187995666237977,
+      "loss": 0.5237,
+      "step": 1661
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1746051081494759,
+      "learning_rate": 0.00010178596908773179,
+      "loss": 0.5352,
+      "step": 1662
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1904814752787734,
+      "learning_rate": 0.00010169197993488851,
+      "loss": 0.4947,
+      "step": 1663
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1682118449533182,
+      "learning_rate": 0.0001015979892869046,
+      "loss": 0.5075,
+      "step": 1664
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1787167363650672,
+      "learning_rate": 0.00010150399722683623,
+      "loss": 0.503,
+      "step": 1665
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.17433015556420575,
+      "learning_rate": 0.00010141000383774067,
+      "loss": 0.4736,
+      "step": 1666
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18890832234956914,
+      "learning_rate": 0.00010131600920267645,
+      "loss": 0.482,
+      "step": 1667
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.16995768612256246,
+      "learning_rate": 0.00010122201340470321,
+      "loss": 0.4764,
+      "step": 1668
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19189171962262738,
+      "learning_rate": 0.00010112801652688155,
+      "loss": 0.5234,
+      "step": 1669
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1887634223189705,
+      "learning_rate": 0.00010103401865227304,
+      "loss": 0.4763,
+      "step": 1670
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1842920226325373,
+      "learning_rate": 0.00010094001986394024,
+      "loss": 0.5156,
+      "step": 1671
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.17969781851880492,
+      "learning_rate": 0.00010084602024494633,
+      "loss": 0.4826,
+      "step": 1672
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19854544410385247,
+      "learning_rate": 0.0001007520198783554,
+      "loss": 0.5498,
+      "step": 1673
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.16752089299272865,
+      "learning_rate": 0.0001006580188472321,
+      "loss": 0.4453,
+      "step": 1674
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1740348252936511,
+      "learning_rate": 0.00010056401723464166,
+      "loss": 0.5058,
+      "step": 1675
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1734770189678822,
+      "learning_rate": 0.00010047001512364992,
+      "loss": 0.4988,
+      "step": 1676
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3352,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 838,
+  "total_flos": 1659369280241664.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}