diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10008 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9994739610731194,
+  "eval_steps": 500,
+  "global_step": 1425,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0007013852358407856,
+      "grad_norm": NaN,
+      "learning_rate": 0.0,
+      "loss": 0.9295,
+      "step": 1
+    },
+    {
+      "epoch": 0.0014027704716815712,
+      "grad_norm": NaN,
+      "learning_rate": 0.0,
+      "loss": 1.432,
+      "step": 2
+    },
+    {
+      "epoch": 0.0021041557075223566,
+      "grad_norm": NaN,
+      "learning_rate": 0.0,
+      "loss": 0.3822,
+      "step": 3
+    },
+    {
+      "epoch": 0.0028055409433631424,
+      "grad_norm": NaN,
+      "learning_rate": 0.0,
+      "loss": 0.6542,
+      "step": 4
+    },
+    {
+      "epoch": 0.003506926179203928,
+      "grad_norm": 29.30962562561035,
+      "learning_rate": 1.7543859649122808e-07,
+      "loss": 0.3371,
+      "step": 5
+    },
+    {
+      "epoch": 0.004208311415044713,
+      "grad_norm": 34.5477294921875,
+      "learning_rate": 3.5087719298245616e-07,
+      "loss": 0.1946,
+      "step": 6
+    },
+    {
+      "epoch": 0.004909696650885499,
+      "grad_norm": 55.77326965332031,
+      "learning_rate": 5.263157894736843e-07,
+      "loss": 0.5682,
+      "step": 7
+    },
+    {
+      "epoch": 0.005611081886726285,
+      "grad_norm": 52.84772491455078,
+      "learning_rate": 7.017543859649123e-07,
+      "loss": 0.3959,
+      "step": 8
+    },
+    {
+      "epoch": 0.00631246712256707,
+      "grad_norm": 44.6428337097168,
+      "learning_rate": 8.771929824561404e-07,
+      "loss": 0.2899,
+      "step": 9
+    },
+    {
+      "epoch": 0.007013852358407856,
+      "grad_norm": 69.89168548583984,
+      "learning_rate": 1.0526315789473685e-06,
+      "loss": 0.6255,
+      "step": 10
+    },
+    {
+      "epoch": 0.007715237594248641,
+      "grad_norm": NaN,
+      "learning_rate": 1.0526315789473685e-06,
+      "loss": 0.4706,
+      "step": 11
+    },
+    {
+      "epoch": 0.008416622830089426,
+      "grad_norm": 65.1708755493164,
+      "learning_rate": 1.2280701754385965e-06,
+      "loss": 0.6875,
+      "step": 12
+    },
+    {
+      "epoch": 0.009118008065930211,
+      "grad_norm": 62.44850158691406,
+      "learning_rate": 1.4035087719298246e-06,
+      "loss": 0.8691,
+      "step": 13
+    },
+    {
+      "epoch": 0.009819393301770998,
+      "grad_norm": 50.52750778198242,
+      "learning_rate": 1.5789473684210528e-06,
+      "loss": 0.5682,
+      "step": 14
+    },
+    {
+      "epoch": 0.010520778537611783,
+      "grad_norm": 38.50554656982422,
+      "learning_rate": 1.7543859649122807e-06,
+      "loss": 0.4118,
+      "step": 15
+    },
+    {
+      "epoch": 0.01122216377345257,
+      "grad_norm": Infinity,
+      "learning_rate": 1.7543859649122807e-06,
+      "loss": 0.4622,
+      "step": 16
+    },
+    {
+      "epoch": 0.011923549009293355,
+      "grad_norm": 41.367347717285156,
+      "learning_rate": 1.929824561403509e-06,
+      "loss": 0.2331,
+      "step": 17
+    },
+    {
+      "epoch": 0.01262493424513414,
+      "grad_norm": 47.666603088378906,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 0.489,
+      "step": 18
+    },
+    {
+      "epoch": 0.013326319480974926,
+      "grad_norm": NaN,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 6.2172,
+      "step": 19
+    },
+    {
+      "epoch": 0.014027704716815711,
+      "grad_norm": 161.6815643310547,
+      "learning_rate": 2.2807017543859652e-06,
+      "loss": 1.9947,
+      "step": 20
+    },
+    {
+      "epoch": 0.014729089952656496,
+      "grad_norm": 14.031229019165039,
+      "learning_rate": 2.456140350877193e-06,
+      "loss": 0.0529,
+      "step": 21
+    },
+    {
+      "epoch": 0.015430475188497283,
+      "grad_norm": 51.389862060546875,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 0.3272,
+      "step": 22
+    },
+    {
+      "epoch": 0.016131860424338066,
+      "grad_norm": 16.31728744506836,
+      "learning_rate": 2.8070175438596493e-06,
+      "loss": 0.0668,
+      "step": 23
+    },
+    {
+      "epoch": 0.016833245660178853,
+      "grad_norm": 32.19120407104492,
+      "learning_rate": 2.9824561403508774e-06,
+      "loss": 0.1298,
+      "step": 24
+    },
+    {
+      "epoch": 0.01753463089601964,
+      "grad_norm": 8.127516746520996,
+      "learning_rate": 3.1578947368421056e-06,
+      "loss": 0.0245,
+      "step": 25
+    },
+    {
+      "epoch": 0.018236016131860423,
+      "grad_norm": 31.48802375793457,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.1808,
+      "step": 26
+    },
+    {
+      "epoch": 0.01893740136770121,
+      "grad_norm": 174.7755584716797,
+      "learning_rate": 3.5087719298245615e-06,
+      "loss": 1.5976,
+      "step": 27
+    },
+    {
+      "epoch": 0.019638786603541996,
+      "grad_norm": 173.22021484375,
+      "learning_rate": 3.6842105263157892e-06,
+      "loss": 2.0731,
+      "step": 28
+    },
+    {
+      "epoch": 0.020340171839382783,
+      "grad_norm": 16.176433563232422,
+      "learning_rate": 3.859649122807018e-06,
+      "loss": 0.0533,
+      "step": 29
+    },
+    {
+      "epoch": 0.021041557075223566,
+      "grad_norm": 25.456438064575195,
+      "learning_rate": 4.035087719298246e-06,
+      "loss": 0.2239,
+      "step": 30
+    },
+    {
+      "epoch": 0.021742942311064353,
+      "grad_norm": 2.50178861618042,
+      "learning_rate": 4.210526315789474e-06,
+      "loss": 0.007,
+      "step": 31
+    },
+    {
+      "epoch": 0.02244432754690514,
+      "grad_norm": 8.369728088378906,
+      "learning_rate": 4.3859649122807014e-06,
+      "loss": 0.0312,
+      "step": 32
+    },
+    {
+      "epoch": 0.023145712782745922,
+      "grad_norm": 1.419299602508545,
+      "learning_rate": 4.5614035087719304e-06,
+      "loss": 0.0039,
+      "step": 33
+    },
+    {
+      "epoch": 0.02384709801858671,
+      "grad_norm": 21.602149963378906,
+      "learning_rate": 4.736842105263159e-06,
+      "loss": 0.1907,
+      "step": 34
+    },
+    {
+      "epoch": 0.024548483254427496,
+      "grad_norm": 96.2072525024414,
+      "learning_rate": 4.912280701754386e-06,
+      "loss": 0.4621,
+      "step": 35
+    },
+    {
+      "epoch": 0.02524986849026828,
+      "grad_norm": 186.85726928710938,
+      "learning_rate": 5.087719298245614e-06,
+      "loss": 3.6381,
+      "step": 36
+    },
+    {
+      "epoch": 0.025951253726109066,
+      "grad_norm": 222.0509033203125,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 2.1687,
+      "step": 37
+    },
+    {
+      "epoch": 0.026652638961949852,
+      "grad_norm": 395.810791015625,
+      "learning_rate": 5.43859649122807e-06,
+      "loss": 4.1469,
+      "step": 38
+    },
+    {
+      "epoch": 0.027354024197790636,
+      "grad_norm": 24.080060958862305,
+      "learning_rate": 5.6140350877192985e-06,
+      "loss": 0.0944,
+      "step": 39
+    },
+    {
+      "epoch": 0.028055409433631422,
+      "grad_norm": 23.396860122680664,
+      "learning_rate": 5.789473684210527e-06,
+      "loss": 0.2287,
+      "step": 40
+    },
+    {
+      "epoch": 0.02875679466947221,
+      "grad_norm": 26.664474487304688,
+      "learning_rate": 5.964912280701755e-06,
+      "loss": 0.2467,
+      "step": 41
+    },
+    {
+      "epoch": 0.029458179905312992,
+      "grad_norm": 0.3562469482421875,
+      "learning_rate": 6.140350877192982e-06,
+      "loss": 0.0005,
+      "step": 42
+    },
+    {
+      "epoch": 0.03015956514115378,
+      "grad_norm": Infinity,
+      "learning_rate": 6.140350877192982e-06,
+      "loss": 1.6967,
+      "step": 43
+    },
+    {
+      "epoch": 0.030860950376994566,
+      "grad_norm": 0.1253829002380371,
+      "learning_rate": 6.315789473684211e-06,
+      "loss": 0.0003,
+      "step": 44
+    },
+    {
+      "epoch": 0.03156233561283535,
+      "grad_norm": 18.994827270507812,
+      "learning_rate": 6.4912280701754385e-06,
+      "loss": 0.233,
+      "step": 45
+    },
+    {
+      "epoch": 0.03226372084867613,
+      "grad_norm": 34.921852111816406,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.3528,
+      "step": 46
+    },
+    {
+      "epoch": 0.03296510608451692,
+      "grad_norm": 202.69903564453125,
+      "learning_rate": 6.842105263157896e-06,
+      "loss": 2.7552,
+      "step": 47
+    },
+    {
+      "epoch": 0.033666491320357705,
+      "grad_norm": 359.0797424316406,
+      "learning_rate": 7.017543859649123e-06,
+      "loss": 2.112,
+      "step": 48
+    },
+    {
+      "epoch": 0.03436787655619849,
+      "grad_norm": 3.0262975692749023,
+      "learning_rate": 7.192982456140351e-06,
+      "loss": 0.0094,
+      "step": 49
+    },
+    {
+      "epoch": 0.03506926179203928,
+      "grad_norm": 504.8307189941406,
+      "learning_rate": 7.3684210526315784e-06,
+      "loss": 6.4525,
+      "step": 50
+    },
+    {
+      "epoch": 0.03577064702788006,
+      "grad_norm": 7.283820152282715,
+      "learning_rate": 7.5438596491228074e-06,
+      "loss": 0.0419,
+      "step": 51
+    },
+    {
+      "epoch": 0.036472032263720845,
+      "grad_norm": 10.193113327026367,
+      "learning_rate": 7.719298245614036e-06,
+      "loss": 0.0805,
+      "step": 52
+    },
+    {
+      "epoch": 0.037173417499561635,
+      "grad_norm": 21.247915267944336,
+      "learning_rate": 7.894736842105263e-06,
+      "loss": 0.1385,
+      "step": 53
+    },
+    {
+      "epoch": 0.03787480273540242,
+      "grad_norm": 12.750287055969238,
+      "learning_rate": 8.070175438596492e-06,
+      "loss": 0.1786,
+      "step": 54
+    },
+    {
+      "epoch": 0.03857618797124321,
+      "grad_norm": 186.3012237548828,
+      "learning_rate": 8.245614035087721e-06,
+      "loss": 2.5431,
+      "step": 55
+    },
+    {
+      "epoch": 0.03927757320708399,
+      "grad_norm": 0.6861621141433716,
+      "learning_rate": 8.421052631578948e-06,
+      "loss": 0.0019,
+      "step": 56
+    },
+    {
+      "epoch": 0.039978958442924775,
+      "grad_norm": 0.36288923025131226,
+      "learning_rate": 8.596491228070176e-06,
+      "loss": 0.0008,
+      "step": 57
+    },
+    {
+      "epoch": 0.040680343678765565,
+      "grad_norm": 0.08618122339248657,
+      "learning_rate": 8.771929824561403e-06,
+      "loss": 0.0001,
+      "step": 58
+    },
+    {
+      "epoch": 0.04138172891460635,
+      "grad_norm": 341.2626037597656,
+      "learning_rate": 8.947368421052632e-06,
+      "loss": 3.6742,
+      "step": 59
+    },
+    {
+      "epoch": 0.04208311415044713,
+      "grad_norm": 5.118091583251953,
+      "learning_rate": 9.122807017543861e-06,
+      "loss": 0.0128,
+      "step": 60
+    },
+    {
+      "epoch": 0.04278449938628792,
+      "grad_norm": 3.345242977142334,
+      "learning_rate": 9.298245614035088e-06,
+      "loss": 0.0095,
+      "step": 61
+    },
+    {
+      "epoch": 0.043485884622128705,
+      "grad_norm": 319.7175598144531,
+      "learning_rate": 9.473684210526317e-06,
+      "loss": 4.3047,
+      "step": 62
+    },
+    {
+      "epoch": 0.04418726985796949,
+      "grad_norm": 279.8736267089844,
+      "learning_rate": 9.649122807017545e-06,
+      "loss": 1.8524,
+      "step": 63
+    },
+    {
+      "epoch": 0.04488865509381028,
+      "grad_norm": 187.50588989257812,
+      "learning_rate": 9.824561403508772e-06,
+      "loss": 2.2166,
+      "step": 64
+    },
+    {
+      "epoch": 0.04559004032965106,
+      "grad_norm": 3.1667685508728027,
+      "learning_rate": 1e-05,
+      "loss": 0.0051,
+      "step": 65
+    },
+    {
+      "epoch": 0.046291425565491845,
+      "grad_norm": 347.2233581542969,
+      "learning_rate": 1.0175438596491228e-05,
+      "loss": 6.7317,
+      "step": 66
+    },
+    {
+      "epoch": 0.046992810801332635,
+      "grad_norm": 0.21984274685382843,
+      "learning_rate": 1.0350877192982457e-05,
+      "loss": 0.0006,
+      "step": 67
+    },
+    {
+      "epoch": 0.04769419603717342,
+      "grad_norm": 657.2495727539062,
+      "learning_rate": 1.0526315789473684e-05,
+      "loss": 12.4963,
+      "step": 68
+    },
+    {
+      "epoch": 0.0483955812730142,
+      "grad_norm": 180.615234375,
+      "learning_rate": 1.0701754385964913e-05,
+      "loss": 3.2259,
+      "step": 69
+    },
+    {
+      "epoch": 0.04909696650885499,
+      "grad_norm": 0.4523675739765167,
+      "learning_rate": 1.087719298245614e-05,
+      "loss": 0.0011,
+      "step": 70
+    },
+    {
+      "epoch": 0.049798351744695775,
+      "grad_norm": 241.3996124267578,
+      "learning_rate": 1.1052631578947368e-05,
+      "loss": 2.9636,
+      "step": 71
+    },
+    {
+      "epoch": 0.05049973698053656,
+      "grad_norm": 0.21977363526821136,
+      "learning_rate": 1.1228070175438597e-05,
+      "loss": 0.0005,
+      "step": 72
+    },
+    {
+      "epoch": 0.05120112221637735,
+      "grad_norm": 102.25936889648438,
+      "learning_rate": 1.1403508771929824e-05,
+      "loss": 0.4978,
+      "step": 73
+    },
+    {
+      "epoch": 0.05190250745221813,
+      "grad_norm": 24.031539916992188,
+      "learning_rate": 1.1578947368421053e-05,
+      "loss": 0.119,
+      "step": 74
+    },
+    {
+      "epoch": 0.052603892688058915,
+      "grad_norm": 187.41812133789062,
+      "learning_rate": 1.1754385964912282e-05,
+      "loss": 1.0223,
+      "step": 75
+    },
+    {
+      "epoch": 0.053305277923899705,
+      "grad_norm": 45.128421783447266,
+      "learning_rate": 1.192982456140351e-05,
+      "loss": 0.4026,
+      "step": 76
+    },
+    {
+      "epoch": 0.05400666315974049,
+      "grad_norm": 12.320211410522461,
+      "learning_rate": 1.2105263157894737e-05,
+      "loss": 0.1124,
+      "step": 77
+    },
+    {
+      "epoch": 0.05470804839558127,
+      "grad_norm": 26.743837356567383,
+      "learning_rate": 1.2280701754385964e-05,
+      "loss": 0.1488,
+      "step": 78
+    },
+    {
+      "epoch": 0.05540943363142206,
+      "grad_norm": 30.013471603393555,
+      "learning_rate": 1.2456140350877193e-05,
+      "loss": 0.3796,
+      "step": 79
+    },
+    {
+      "epoch": 0.056110818867262845,
+      "grad_norm": 38.910221099853516,
+      "learning_rate": 1.2631578947368422e-05,
+      "loss": 0.7097,
+      "step": 80
+    },
+    {
+      "epoch": 0.05681220410310363,
+      "grad_norm": 26.001049041748047,
+      "learning_rate": 1.2807017543859651e-05,
+      "loss": 0.1451,
+      "step": 81
+    },
+    {
+      "epoch": 0.05751358933894442,
+      "grad_norm": 12.577742576599121,
+      "learning_rate": 1.2982456140350877e-05,
+      "loss": 0.0548,
+      "step": 82
+    },
+    {
+      "epoch": 0.0582149745747852,
+      "grad_norm": 334.09478759765625,
+      "learning_rate": 1.3157894736842106e-05,
+      "loss": 8.8326,
+      "step": 83
+    },
+    {
+      "epoch": 0.058916359810625984,
+      "grad_norm": 33.031837463378906,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.2555,
+      "step": 84
+    },
+    {
+      "epoch": 0.059617745046466775,
+      "grad_norm": 47.51133346557617,
+      "learning_rate": 1.3508771929824562e-05,
+      "loss": 0.4288,
+      "step": 85
+    },
+    {
+      "epoch": 0.06031913028230756,
+      "grad_norm": 41.160430908203125,
+      "learning_rate": 1.3684210526315791e-05,
+      "loss": 0.3075,
+      "step": 86
+    },
+    {
+      "epoch": 0.06102051551814834,
+      "grad_norm": 9.073790550231934,
+      "learning_rate": 1.3859649122807017e-05,
+      "loss": 0.045,
+      "step": 87
+    },
+    {
+      "epoch": 0.06172190075398913,
+      "grad_norm": 352.0498352050781,
+      "learning_rate": 1.4035087719298246e-05,
+      "loss": 1.3396,
+      "step": 88
+    },
+    {
+      "epoch": 0.062423285989829914,
+      "grad_norm": 36.5557746887207,
+      "learning_rate": 1.4210526315789475e-05,
+      "loss": 0.1926,
+      "step": 89
+    },
+    {
+      "epoch": 0.0631246712256707,
+      "grad_norm": 6.991324424743652,
+      "learning_rate": 1.4385964912280702e-05,
+      "loss": 0.0341,
+      "step": 90
+    },
+    {
+      "epoch": 0.06382605646151149,
+      "grad_norm": 6.0483078956604,
+      "learning_rate": 1.4561403508771931e-05,
+      "loss": 0.0219,
+      "step": 91
+    },
+    {
+      "epoch": 0.06452744169735226,
+      "grad_norm": 4.510119438171387,
+      "learning_rate": 1.4736842105263157e-05,
+      "loss": 0.0161,
+      "step": 92
+    },
+    {
+      "epoch": 0.06522882693319305,
+      "grad_norm": 0.8209487199783325,
+      "learning_rate": 1.4912280701754386e-05,
+      "loss": 0.0022,
+      "step": 93
+    },
+    {
+      "epoch": 0.06593021216903384,
+      "grad_norm": 0.0308829378336668,
+      "learning_rate": 1.5087719298245615e-05,
+      "loss": 0.0001,
+      "step": 94
+    },
+    {
+      "epoch": 0.06663159740487462,
+      "grad_norm": 171.68765258789062,
+      "learning_rate": 1.5263157894736842e-05,
+      "loss": 4.8668,
+      "step": 95
+    },
+    {
+      "epoch": 0.06733298264071541,
+      "grad_norm": 0.0008907412411645055,
+      "learning_rate": 1.543859649122807e-05,
+      "loss": 0.0,
+      "step": 96
+    },
+    {
+      "epoch": 0.0680343678765562,
+      "grad_norm": 0.0062671443447470665,
+      "learning_rate": 1.56140350877193e-05,
+      "loss": 0.0,
+      "step": 97
+    },
+    {
+      "epoch": 0.06873575311239698,
+      "grad_norm": 0.02987469919025898,
+      "learning_rate": 1.5789473684210526e-05,
+      "loss": 0.0001,
+      "step": 98
+    },
+    {
+      "epoch": 0.06943713834823777,
+      "grad_norm": 0.00033108692150563,
+      "learning_rate": 1.5964912280701755e-05,
+      "loss": 0.0,
+      "step": 99
+    },
+    {
+      "epoch": 0.07013852358407856,
+      "grad_norm": 0.21173974871635437,
+      "learning_rate": 1.6140350877192984e-05,
+      "loss": 0.0003,
+      "step": 100
+    },
+    {
+      "epoch": 0.07083990881991933,
+      "grad_norm": 0.0003154289734084159,
+      "learning_rate": 1.6315789473684213e-05,
+      "loss": 0.0,
+      "step": 101
+    },
+    {
+      "epoch": 0.07154129405576012,
+      "grad_norm": 0.0001488685520598665,
+      "learning_rate": 1.6491228070175442e-05,
+      "loss": 0.0,
+      "step": 102
+    },
+    {
+      "epoch": 0.07224267929160091,
+      "grad_norm": 1.3430032595351804e-05,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.0,
+      "step": 103
+    },
+    {
+      "epoch": 0.07294406452744169,
+      "grad_norm": 0.011110284365713596,
+      "learning_rate": 1.6842105263157896e-05,
+      "loss": 0.0,
+      "step": 104
+    },
+    {
+      "epoch": 0.07364544976328248,
+      "grad_norm": 1.7930891772266477e-05,
+      "learning_rate": 1.7017543859649125e-05,
+      "loss": 0.0,
+      "step": 105
+    },
+    {
+      "epoch": 0.07434683499912327,
+      "grad_norm": 5.653702146446449e-07,
+      "learning_rate": 1.719298245614035e-05,
+      "loss": 0.0,
+      "step": 106
+    },
+    {
+      "epoch": 0.07504822023496406,
+      "grad_norm": 135.4245147705078,
+      "learning_rate": 1.736842105263158e-05,
+      "loss": 6.4792,
+      "step": 107
+    },
+    {
+      "epoch": 0.07574960547080484,
+      "grad_norm": 153.99264526367188,
+      "learning_rate": 1.7543859649122806e-05,
+      "loss": 6.485,
+      "step": 108
+    },
+    {
+      "epoch": 0.07645099070664563,
+      "grad_norm": 159.35337829589844,
+      "learning_rate": 1.7719298245614035e-05,
+      "loss": 2.4936,
+      "step": 109
+    },
+    {
+      "epoch": 0.07715237594248642,
+      "grad_norm": 0.00017455461784265935,
+      "learning_rate": 1.7894736842105264e-05,
+      "loss": 0.0,
+      "step": 110
+    },
+    {
+      "epoch": 0.0778537611783272,
+      "grad_norm": 146.1802978515625,
+      "learning_rate": 1.8070175438596493e-05,
+      "loss": 4.3088,
+      "step": 111
+    },
+    {
+      "epoch": 0.07855514641416798,
+      "grad_norm": 0.022132212296128273,
+      "learning_rate": 1.8245614035087722e-05,
+      "loss": 0.0,
+      "step": 112
+    },
+    {
+      "epoch": 0.07925653165000877,
+      "grad_norm": 160.43975830078125,
+      "learning_rate": 1.8421052631578947e-05,
+      "loss": 4.5829,
+      "step": 113
+    },
+    {
+      "epoch": 0.07995791688584955,
+      "grad_norm": 164.45138549804688,
+      "learning_rate": 1.8596491228070176e-05,
+      "loss": 4.1681,
+      "step": 114
+    },
+    {
+      "epoch": 0.08065930212169034,
+      "grad_norm": 0.02766432799398899,
+      "learning_rate": 1.8771929824561405e-05,
+      "loss": 0.0001,
+      "step": 115
+    },
+    {
+      "epoch": 0.08136068735753113,
+      "grad_norm": 163.59400939941406,
+      "learning_rate": 1.8947368421052634e-05,
+      "loss": 3.0252,
+      "step": 116
+    },
+    {
+      "epoch": 0.08206207259337191,
+      "grad_norm": 7.11077356338501,
+      "learning_rate": 1.9122807017543863e-05,
+      "loss": 0.0309,
+      "step": 117
+    },
+    {
+      "epoch": 0.0827634578292127,
+      "grad_norm": 15.318146705627441,
+      "learning_rate": 1.929824561403509e-05,
+      "loss": 0.0856,
+      "step": 118
+    },
+    {
+      "epoch": 0.08346484306505349,
+      "grad_norm": 9.1824951171875,
+      "learning_rate": 1.9473684210526315e-05,
+      "loss": 0.0416,
+      "step": 119
+    },
+    {
+      "epoch": 0.08416622830089426,
+      "grad_norm": 22.720653533935547,
+      "learning_rate": 1.9649122807017544e-05,
+      "loss": 0.1365,
+      "step": 120
+    },
+    {
+      "epoch": 0.08486761353673505,
+      "grad_norm": 26.200096130371094,
+      "learning_rate": 1.9824561403508773e-05,
+      "loss": 0.2579,
+      "step": 121
+    },
+    {
+      "epoch": 0.08556899877257584,
+      "grad_norm": 15.874689102172852,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 122
+    },
+    {
+      "epoch": 0.08627038400841662,
+      "grad_norm": 11.818103790283203,
+      "learning_rate": 2.0175438596491227e-05,
+      "loss": 0.0541,
+      "step": 123
+    },
+    {
+      "epoch": 0.08697176924425741,
+      "grad_norm": 116.5608139038086,
+      "learning_rate": 2.0350877192982456e-05,
+      "loss": 0.9623,
+      "step": 124
+    },
+    {
+      "epoch": 0.0876731544800982,
+      "grad_norm": 10.346467018127441,
+      "learning_rate": 2.0526315789473685e-05,
+      "loss": 0.0562,
+      "step": 125
+    },
+    {
+      "epoch": 0.08837453971593898,
+      "grad_norm": 137.41136169433594,
+      "learning_rate": 2.0701754385964914e-05,
+      "loss": 1.1566,
+      "step": 126
+    },
+    {
+      "epoch": 0.08907592495177977,
+      "grad_norm": 2.9712822437286377,
+      "learning_rate": 2.0877192982456143e-05,
+      "loss": 0.0104,
+      "step": 127
+    },
+    {
+      "epoch": 0.08977731018762056,
+      "grad_norm": 18.723491668701172,
+      "learning_rate": 2.105263157894737e-05,
+      "loss": 0.0667,
+      "step": 128
+    },
+    {
+      "epoch": 0.09047869542346133,
+      "grad_norm": 156.25631713867188,
+      "learning_rate": 2.1228070175438598e-05,
+      "loss": 2.4023,
+      "step": 129
+    },
+    {
+      "epoch": 0.09118008065930212,
+      "grad_norm": 298.97552490234375,
+      "learning_rate": 2.1403508771929827e-05,
+      "loss": 6.4928,
+      "step": 130
+    },
+    {
+      "epoch": 0.09188146589514291,
+      "grad_norm": 87.54296112060547,
+      "learning_rate": 2.1578947368421053e-05,
+      "loss": 0.5526,
+      "step": 131
+    },
+    {
+      "epoch": 0.09258285113098369,
+      "grad_norm": 0.25955039262771606,
+      "learning_rate": 2.175438596491228e-05,
+      "loss": 0.0007,
+      "step": 132
+    },
+    {
+      "epoch": 0.09328423636682448,
+      "grad_norm": 123.57398986816406,
+      "learning_rate": 2.1929824561403507e-05,
+      "loss": 1.0402,
+      "step": 133
+    },
+    {
+      "epoch": 0.09398562160266527,
+      "grad_norm": 238.33229064941406,
+      "learning_rate": 2.2105263157894736e-05,
+      "loss": 2.4997,
+      "step": 134
+    },
+    {
+      "epoch": 0.09468700683850605,
+      "grad_norm": 13.356133460998535,
+      "learning_rate": 2.2280701754385965e-05,
+      "loss": 0.0549,
+      "step": 135
+    },
+    {
+      "epoch": 0.09538839207434684,
+      "grad_norm": 44.56560134887695,
+      "learning_rate": 2.2456140350877194e-05,
+      "loss": 0.4078,
+      "step": 136
+    },
+    {
+      "epoch": 0.09608977731018763,
+      "grad_norm": 25.394866943359375,
+      "learning_rate": 2.2631578947368423e-05,
+      "loss": 0.1681,
+      "step": 137
+    },
+    {
+      "epoch": 0.0967911625460284,
+      "grad_norm": 28.068824768066406,
+      "learning_rate": 2.280701754385965e-05,
+      "loss": 0.174,
+      "step": 138
+    },
+    {
+      "epoch": 0.0974925477818692,
+      "grad_norm": 27.76630401611328,
+      "learning_rate": 2.2982456140350878e-05,
+      "loss": 0.1515,
+      "step": 139
+    },
+    {
+      "epoch": 0.09819393301770998,
+      "grad_norm": 21.082311630249023,
+      "learning_rate": 2.3157894736842107e-05,
+      "loss": 0.1379,
+      "step": 140
+    },
+    {
+      "epoch": 0.09889531825355076,
+      "grad_norm": 16.697647094726562,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 0.1066,
+      "step": 141
+    },
+    {
+      "epoch": 0.09959670348939155,
+      "grad_norm": 127.46953582763672,
+      "learning_rate": 2.3508771929824565e-05,
+      "loss": 1.3053,
+      "step": 142
+    },
+    {
+      "epoch": 0.10029808872523234,
+      "grad_norm": 6.2219648361206055,
+      "learning_rate": 2.368421052631579e-05,
+      "loss": 0.0323,
+      "step": 143
+    },
+    {
+      "epoch": 0.10099947396107312,
+      "grad_norm": 0.1309046596288681,
+      "learning_rate": 2.385964912280702e-05,
+      "loss": 0.0004,
+      "step": 144
+    },
+    {
+      "epoch": 0.1017008591969139,
+      "grad_norm": 149.50645446777344,
+      "learning_rate": 2.4035087719298245e-05,
+      "loss": 1.8353,
+      "step": 145
+    },
+    {
+      "epoch": 0.1024022444327547,
+      "grad_norm": 159.4170684814453,
+      "learning_rate": 2.4210526315789474e-05,
+      "loss": 3.2714,
+      "step": 146
+    },
+    {
+      "epoch": 0.10310362966859547,
+      "grad_norm": 0.038805797696113586,
+      "learning_rate": 2.4385964912280703e-05,
+      "loss": 0.0001,
+      "step": 147
+    },
+    {
+      "epoch": 0.10380501490443626,
+      "grad_norm": 301.7401123046875,
+      "learning_rate": 2.456140350877193e-05,
+      "loss": 5.5114,
+      "step": 148
+    },
+    {
+      "epoch": 0.10450640014027705,
+      "grad_norm": 160.01580810546875,
+      "learning_rate": 2.4736842105263158e-05,
+      "loss": 3.0398,
+      "step": 149
+    },
+    {
+      "epoch": 0.10520778537611783,
+      "grad_norm": 1.0593466758728027,
+      "learning_rate": 2.4912280701754387e-05,
+      "loss": 0.0026,
+      "step": 150
+    },
+    {
+      "epoch": 0.10590917061195862,
+      "grad_norm": 0.1588684320449829,
+      "learning_rate": 2.5087719298245616e-05,
+      "loss": 0.0004,
+      "step": 151
+    },
+    {
+      "epoch": 0.10661055584779941,
+      "grad_norm": 0.12831349670886993,
+      "learning_rate": 2.5263157894736845e-05,
+      "loss": 0.0004,
+      "step": 152
+    },
+    {
+      "epoch": 0.10731194108364019,
+      "grad_norm": 1.908389925956726,
+      "learning_rate": 2.5438596491228074e-05,
+      "loss": 0.0062,
+      "step": 153
+    },
+    {
+      "epoch": 0.10801332631948098,
+      "grad_norm": 0.9305810928344727,
+      "learning_rate": 2.5614035087719303e-05,
+      "loss": 0.0029,
+      "step": 154
+    },
+    {
+      "epoch": 0.10871471155532177,
+      "grad_norm": 149.55746459960938,
+      "learning_rate": 2.578947368421053e-05,
+      "loss": 2.1251,
+      "step": 155
+    },
+    {
+      "epoch": 0.10941609679116254,
+      "grad_norm": 83.3541030883789,
+      "learning_rate": 2.5964912280701754e-05,
+      "loss": 0.4841,
+      "step": 156
+    },
+    {
+      "epoch": 0.11011748202700333,
+      "grad_norm": 0.37118402123451233,
+      "learning_rate": 2.6140350877192983e-05,
+      "loss": 0.0009,
+      "step": 157
+    },
+    {
+      "epoch": 0.11081886726284412,
+      "grad_norm": 212.9443359375,
+      "learning_rate": 2.6315789473684212e-05,
+      "loss": 1.6639,
+      "step": 158
+    },
+    {
+      "epoch": 0.1115202524986849,
+      "grad_norm": 1.386982798576355,
+      "learning_rate": 2.6491228070175438e-05,
+      "loss": 0.0049,
+      "step": 159
+    },
+    {
+      "epoch": 0.11222163773452569,
+      "grad_norm": 151.23684692382812,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 1.913,
+      "step": 160
+    },
+    {
+      "epoch": 0.11292302297036648,
+      "grad_norm": 276.9901123046875,
+      "learning_rate": 2.6842105263157896e-05,
+      "loss": 5.0838,
+      "step": 161
+    },
+    {
+      "epoch": 0.11362440820620726,
+      "grad_norm": 7.047855377197266,
+      "learning_rate": 2.7017543859649125e-05,
+      "loss": 0.0422,
+      "step": 162
+    },
+    {
+      "epoch": 0.11432579344204805,
+      "grad_norm": 5.565479755401611,
+      "learning_rate": 2.7192982456140354e-05,
+      "loss": 0.0252,
+      "step": 163
+    },
+    {
+      "epoch": 0.11502717867788884,
+      "grad_norm": 11.038246154785156,
+      "learning_rate": 2.7368421052631583e-05,
+      "loss": 0.064,
+      "step": 164
+    },
+    {
+      "epoch": 0.11572856391372961,
+      "grad_norm": 1.4180843830108643,
+      "learning_rate": 2.754385964912281e-05,
+      "loss": 0.0048,
+      "step": 165
+    },
+    {
+      "epoch": 0.1164299491495704,
+      "grad_norm": 131.0841064453125,
+      "learning_rate": 2.7719298245614034e-05,
+      "loss": 1.1797,
+      "step": 166
+    },
+    {
+      "epoch": 0.11713133438541119,
+      "grad_norm": 286.4391784667969,
+      "learning_rate": 2.7894736842105263e-05,
+      "loss": 4.2848,
+      "step": 167
+    },
+    {
+      "epoch": 0.11783271962125197,
+      "grad_norm": 0.354282945394516,
+      "learning_rate": 2.8070175438596492e-05,
+      "loss": 0.001,
+      "step": 168
+    },
+    {
+      "epoch": 0.11853410485709276,
+      "grad_norm": 148.43812561035156,
+      "learning_rate": 2.824561403508772e-05,
+      "loss": 2.3637,
+      "step": 169
+    },
+    {
+      "epoch": 0.11923549009293355,
+      "grad_norm": 3.4153621196746826,
+      "learning_rate": 2.842105263157895e-05,
+      "loss": 0.0121,
+      "step": 170
+    },
+    {
+      "epoch": 0.11993687532877433,
+      "grad_norm": 6.812062740325928,
+      "learning_rate": 2.8596491228070175e-05,
+      "loss": 0.038,
+      "step": 171
+    },
+    {
+      "epoch": 0.12063826056461512,
+      "grad_norm": 0.6381294131278992,
+      "learning_rate": 2.8771929824561404e-05,
+      "loss": 0.002,
+      "step": 172
+    },
+    {
+      "epoch": 0.1213396458004559,
+      "grad_norm": 0.05846899002790451,
+      "learning_rate": 2.8947368421052634e-05,
+      "loss": 0.0002,
+      "step": 173
+    },
+    {
+      "epoch": 0.12204103103629668,
+      "grad_norm": 140.1736297607422,
+      "learning_rate": 2.9122807017543863e-05,
+      "loss": 4.0868,
+      "step": 174
+    },
+    {
+      "epoch": 0.12274241627213747,
+      "grad_norm": 1.7416075468063354,
+      "learning_rate": 2.929824561403509e-05,
+      "loss": 0.0057,
+      "step": 175
+    },
+    {
+      "epoch": 0.12344380150797826,
+      "grad_norm": 0.35068169236183167,
+      "learning_rate": 2.9473684210526314e-05,
+      "loss": 0.0009,
+      "step": 176
+    },
+    {
+      "epoch": 0.12414518674381904,
+      "grad_norm": 135.71665954589844,
+      "learning_rate": 2.9649122807017543e-05,
+      "loss": 2.1025,
+      "step": 177
+    },
+    {
+      "epoch": 0.12484657197965983,
+      "grad_norm": 134.73190307617188,
+      "learning_rate": 2.9824561403508772e-05,
+      "loss": 1.8685,
+      "step": 178
+    },
+    {
+      "epoch": 0.1255479572155006,
+      "grad_norm": 266.5934143066406,
+      "learning_rate": 3e-05,
+      "loss": 5.1122,
+      "step": 179
+    },
+    {
+      "epoch": 0.1262493424513414,
+      "grad_norm": 383.29931640625,
+      "learning_rate": 3.017543859649123e-05,
+      "loss": 10.0944,
+      "step": 180
+    },
+    {
+      "epoch": 0.12695072768718219,
+      "grad_norm": 3.7992992401123047,
+      "learning_rate": 3.035087719298246e-05,
+      "loss": 0.0161,
+      "step": 181
+    },
+    {
+      "epoch": 0.12765211292302298,
+      "grad_norm": 206.0655975341797,
+      "learning_rate": 3.0526315789473684e-05,
+      "loss": 1.5303,
+      "step": 182
+    },
+    {
+      "epoch": 0.12835349815886377,
+      "grad_norm": 13.431243896484375,
+      "learning_rate": 3.0701754385964913e-05,
+      "loss": 0.0912,
+      "step": 183
+    },
+    {
+      "epoch": 0.12905488339470453,
+      "grad_norm": 16.69830894470215,
+      "learning_rate": 3.087719298245614e-05,
+      "loss": 0.0824,
+      "step": 184
+    },
+    {
+      "epoch": 0.12975626863054532,
+      "grad_norm": 13.488637924194336,
+      "learning_rate": 3.105263157894737e-05,
+      "loss": 0.0746,
+      "step": 185
+    },
+    {
+      "epoch": 0.1304576538663861,
+      "grad_norm": 10.065983772277832,
+      "learning_rate": 3.12280701754386e-05,
+      "loss": 0.0503,
+      "step": 186
+    },
+    {
+      "epoch": 0.1311590391022269,
+      "grad_norm": 123.50023651123047,
+      "learning_rate": 3.140350877192982e-05,
+      "loss": 1.2874,
+      "step": 187
+    },
+    {
+      "epoch": 0.1318604243380677,
+      "grad_norm": 1.7797540426254272,
+      "learning_rate": 3.157894736842105e-05,
+      "loss": 0.0067,
+      "step": 188
+    },
+    {
+      "epoch": 0.13256180957390848,
+      "grad_norm": 0.33675551414489746,
+      "learning_rate": 3.175438596491228e-05,
+      "loss": 0.0011,
+      "step": 189
+    },
+    {
+      "epoch": 0.13326319480974924,
+      "grad_norm": 118.4156723022461,
+      "learning_rate": 3.192982456140351e-05,
+      "loss": 1.2059,
+      "step": 190
+    },
+    {
+      "epoch": 0.13396458004559003,
+      "grad_norm": 136.65228271484375,
+      "learning_rate": 3.210526315789474e-05,
+      "loss": 1.7856,
+      "step": 191
+    },
+    {
+      "epoch": 0.13466596528143082,
+      "grad_norm": 145.7415313720703,
+      "learning_rate": 3.228070175438597e-05,
+      "loss": 2.8991,
+      "step": 192
+    },
+    {
+      "epoch": 0.1353673505172716,
+      "grad_norm": 262.7060241699219,
+      "learning_rate": 3.24561403508772e-05,
+      "loss": 3.4817,
+      "step": 193
+    },
+    {
+      "epoch": 0.1360687357531124,
+      "grad_norm": 0.25727707147598267,
+      "learning_rate": 3.2631578947368426e-05,
+      "loss": 0.0008,
+      "step": 194
+    },
+    {
+      "epoch": 0.1367701209889532,
+      "grad_norm": 1.0702918767929077,
+      "learning_rate": 3.2807017543859655e-05,
+      "loss": 0.0037,
+      "step": 195
+    },
+    {
+      "epoch": 0.13747150622479395,
+      "grad_norm": 1.5917607545852661,
+      "learning_rate": 3.2982456140350884e-05,
+      "loss": 0.0055,
+      "step": 196
+    },
+    {
+      "epoch": 0.13817289146063474,
+      "grad_norm": 355.18511962890625,
+      "learning_rate": 3.3157894736842106e-05,
+      "loss": 11.4677,
+      "step": 197
+    },
+    {
+      "epoch": 0.13887427669647553,
+      "grad_norm": 1.7947548627853394,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.0074,
+      "step": 198
+    },
+    {
+      "epoch": 0.13957566193231633,
+      "grad_norm": 1.5046889781951904,
+      "learning_rate": 3.3508771929824564e-05,
+      "loss": 0.0058,
+      "step": 199
+    },
+    {
+      "epoch": 0.14027704716815712,
+      "grad_norm": 0.2553917467594147,
+      "learning_rate": 3.368421052631579e-05,
+      "loss": 0.0008,
+      "step": 200
+    },
+    {
+      "epoch": 0.1409784324039979,
+      "grad_norm": 0.773563027381897,
+      "learning_rate": 3.385964912280702e-05,
+      "loss": 0.0027,
+      "step": 201
+    },
+    {
+      "epoch": 0.14167981763983867,
+      "grad_norm": 166.71820068359375,
+      "learning_rate": 3.403508771929825e-05,
+      "loss": 6.1496,
+      "step": 202
+    },
+    {
+      "epoch": 0.14238120287567946,
+      "grad_norm": 0.017217637971043587,
+      "learning_rate": 3.421052631578947e-05,
+      "loss": 0.0001,
+      "step": 203
+    },
+    {
+      "epoch": 0.14308258811152025,
+      "grad_norm": 0.0036761912051588297,
+      "learning_rate": 3.43859649122807e-05,
+      "loss": 0.0,
+      "step": 204
+    },
+    {
+      "epoch": 0.14378397334736104,
+      "grad_norm": 0.000679672637488693,
+      "learning_rate": 3.456140350877193e-05,
+      "loss": 0.0,
+      "step": 205
+    },
+    {
+      "epoch": 0.14448535858320183,
+      "grad_norm": 0.00011312704009469599,
+      "learning_rate": 3.473684210526316e-05,
+      "loss": 0.0,
+      "step": 206
+    },
+    {
+      "epoch": 0.14518674381904262,
+      "grad_norm": 0.00012077714200131595,
+      "learning_rate": 3.491228070175438e-05,
+      "loss": 0.0,
+      "step": 207
+    },
+    {
+      "epoch": 0.14588812905488338,
+      "grad_norm": 0.00012628763215616345,
+      "learning_rate": 3.508771929824561e-05,
+      "loss": 0.0,
+      "step": 208
+    },
+    {
+      "epoch": 0.14658951429072417,
+      "grad_norm": 1.7496255168225616e-05,
+      "learning_rate": 3.526315789473684e-05,
+      "loss": 0.0,
+      "step": 209
+    },
+    {
+      "epoch": 0.14729089952656496,
+      "grad_norm": 1.5739196896902286e-05,
+      "learning_rate": 3.543859649122807e-05,
+      "loss": 0.0,
+      "step": 210
+    },
+    {
+      "epoch": 0.14799228476240575,
+      "grad_norm": 0.08704076707363129,
+      "learning_rate": 3.56140350877193e-05,
+      "loss": 0.0001,
+      "step": 211
+    },
+    {
+      "epoch": 0.14869366999824654,
+      "grad_norm": 0.0016435593133792281,
+      "learning_rate": 3.578947368421053e-05,
+      "loss": 0.0,
+      "step": 212
+    },
+    {
+      "epoch": 0.14939505523408733,
+      "grad_norm": 1.5830031633377075,
+      "learning_rate": 3.5964912280701756e-05,
+      "loss": 0.005,
+      "step": 213
+    },
+    {
+      "epoch": 0.15009644046992812,
+      "grad_norm": 0.00016859408060554415,
+      "learning_rate": 3.6140350877192985e-05,
+      "loss": 0.0,
+      "step": 214
+    },
+    {
+      "epoch": 0.15079782570576888,
+      "grad_norm": 1.2645250535570085e-06,
+      "learning_rate": 3.6315789473684214e-05,
+      "loss": 0.0,
+      "step": 215
+    },
+    {
+      "epoch": 0.15149921094160967,
+      "grad_norm": 336.5249328613281,
+      "learning_rate": 3.6491228070175443e-05,
+      "loss": 4.1939,
+      "step": 216
+    },
+    {
+      "epoch": 0.15220059617745046,
+      "grad_norm": 9.010884127746976e-07,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 0.0,
+      "step": 217
+    },
+    {
+      "epoch": 0.15290198141329125,
+      "grad_norm": 5.530741304937692e-07,
+      "learning_rate": 3.6842105263157895e-05,
+      "loss": 0.0,
+      "step": 218
+    },
+    {
+      "epoch": 0.15360336664913204,
+      "grad_norm": 222.7361297607422,
+      "learning_rate": 3.7017543859649124e-05,
+      "loss": 0.7712,
+      "step": 219
+    },
+    {
+      "epoch": 0.15430475188497284,
+      "grad_norm": 123.51704406738281,
+      "learning_rate": 3.719298245614035e-05,
+      "loss": 7.8703,
+      "step": 220
+    },
+    {
+      "epoch": 0.1550061371208136,
+      "grad_norm": 119.08999633789062,
+      "learning_rate": 3.736842105263158e-05,
+      "loss": 6.6427,
+      "step": 221
+    },
+    {
+      "epoch": 0.1557075223566544,
+      "grad_norm": 0.005513960495591164,
+      "learning_rate": 3.754385964912281e-05,
+      "loss": 0.0,
+      "step": 222
+    },
+    {
+      "epoch": 0.15640890759249518,
+      "grad_norm": 0.021047895774245262,
+      "learning_rate": 3.771929824561404e-05,
+      "loss": 0.0,
+      "step": 223
+    },
+    {
+      "epoch": 0.15711029282833597,
+      "grad_norm": 0.001549599110148847,
+      "learning_rate": 3.789473684210527e-05,
+      "loss": 0.0,
+      "step": 224
+    },
+    {
+      "epoch": 0.15781167806417676,
+      "grad_norm": 0.008736948482692242,
+      "learning_rate": 3.80701754385965e-05,
+      "loss": 0.0,
+      "step": 225
+    },
+    {
+      "epoch": 0.15851306330001755,
+      "grad_norm": 0.03641926124691963,
+      "learning_rate": 3.824561403508773e-05,
+      "loss": 0.0001,
+      "step": 226
+    },
+    {
+      "epoch": 0.1592144485358583,
+      "grad_norm": 0.026789208874106407,
+      "learning_rate": 3.842105263157895e-05,
+      "loss": 0.0001,
+      "step": 227
+    },
+    {
+      "epoch": 0.1599158337716991,
+      "grad_norm": 130.79087829589844,
+      "learning_rate": 3.859649122807018e-05,
+      "loss": 3.4204,
+      "step": 228
+    },
+    {
+      "epoch": 0.1606172190075399,
+      "grad_norm": 404.3518981933594,
+      "learning_rate": 3.877192982456141e-05,
+      "loss": 5.6741,
+      "step": 229
+    },
+    {
+      "epoch": 0.16131860424338068,
+      "grad_norm": 5.39788818359375,
+      "learning_rate": 3.894736842105263e-05,
+      "loss": 0.0251,
+      "step": 230
+    },
+    {
+      "epoch": 0.16201998947922147,
+      "grad_norm": 18.25702667236328,
+      "learning_rate": 3.912280701754386e-05,
+      "loss": 0.1059,
+      "step": 231
+    },
+    {
+      "epoch": 0.16272137471506226,
+      "grad_norm": 2.5649735927581787,
+      "learning_rate": 3.929824561403509e-05,
+      "loss": 0.0108,
+      "step": 232
+    },
+    {
+      "epoch": 0.16342275995090302,
+      "grad_norm": 389.410400390625,
+      "learning_rate": 3.9473684210526316e-05,
+      "loss": 4.1361,
+      "step": 233
+    },
+    {
+      "epoch": 0.16412414518674381,
+      "grad_norm": 97.98786926269531,
+      "learning_rate": 3.9649122807017545e-05,
+      "loss": 1.0207,
+      "step": 234
+    },
+    {
+      "epoch": 0.1648255304225846,
+      "grad_norm": 446.8147277832031,
+      "learning_rate": 3.9824561403508774e-05,
+      "loss": 2.3678,
+      "step": 235
+    },
+    {
+      "epoch": 0.1655269156584254,
+      "grad_norm": 88.88761901855469,
+      "learning_rate": 4e-05,
+      "loss": 0.814,
+      "step": 236
+    },
+    {
+      "epoch": 0.16622830089426618,
+      "grad_norm": 217.3586883544922,
+      "learning_rate": 4.017543859649123e-05,
+      "loss": 9.7291,
+      "step": 237
+    },
+    {
+      "epoch": 0.16692968613010697,
+      "grad_norm": 106.2865219116211,
+      "learning_rate": 4.0350877192982455e-05,
+      "loss": 1.1716,
+      "step": 238
+    },
+    {
+      "epoch": 0.16763107136594774,
+      "grad_norm": 154.2283477783203,
+      "learning_rate": 4.0526315789473684e-05,
+      "loss": 0.6351,
+      "step": 239
+    },
+    {
+      "epoch": 0.16833245660178853,
+      "grad_norm": 37.964447021484375,
+      "learning_rate": 4.070175438596491e-05,
+      "loss": 0.3484,
+      "step": 240
+    },
+    {
+      "epoch": 0.16903384183762932,
+      "grad_norm": 24.696714401245117,
+      "learning_rate": 4.087719298245614e-05,
+      "loss": 0.2762,
+      "step": 241
+    },
+    {
+      "epoch": 0.1697352270734701,
+      "grad_norm": 82.12918853759766,
+      "learning_rate": 4.105263157894737e-05,
+      "loss": 0.9978,
+      "step": 242
+    },
+    {
+      "epoch": 0.1704366123093109,
+      "grad_norm": 41.84835433959961,
+      "learning_rate": 4.12280701754386e-05,
+      "loss": 0.4799,
+      "step": 243
+    },
+    {
+      "epoch": 0.1711379975451517,
+      "grad_norm": 45.34370803833008,
+      "learning_rate": 4.140350877192983e-05,
+      "loss": 0.4255,
+      "step": 244
+    },
+    {
+      "epoch": 0.17183938278099245,
+      "grad_norm": 41.85076904296875,
+      "learning_rate": 4.157894736842106e-05,
+      "loss": 0.616,
+      "step": 245
+    },
+    {
+      "epoch": 0.17254076801683324,
+      "grad_norm": 31.718475341796875,
+      "learning_rate": 4.1754385964912287e-05,
+      "loss": 0.2034,
+      "step": 246
+    },
+    {
+      "epoch": 0.17324215325267403,
+      "grad_norm": 149.61940002441406,
+      "learning_rate": 4.1929824561403516e-05,
+      "loss": 1.5976,
+      "step": 247
+    },
+    {
+      "epoch": 0.17394353848851482,
+      "grad_norm": 127.43466186523438,
+      "learning_rate": 4.210526315789474e-05,
+      "loss": 1.4465,
+      "step": 248
+    },
+    {
+      "epoch": 0.1746449237243556,
+      "grad_norm": 36.885379791259766,
+      "learning_rate": 4.228070175438597e-05,
+      "loss": 0.5706,
+      "step": 249
+    },
+    {
+      "epoch": 0.1753463089601964,
+      "grad_norm": 343.46649169921875,
+      "learning_rate": 4.2456140350877196e-05,
+      "loss": 5.0778,
+      "step": 250
+    },
+    {
+      "epoch": 0.17604769419603716,
+      "grad_norm": 23.604751586914062,
+      "learning_rate": 4.2631578947368425e-05,
+      "loss": 0.1635,
+      "step": 251
+    },
+    {
+      "epoch": 0.17674907943187795,
+      "grad_norm": 37.630043029785156,
+      "learning_rate": 4.2807017543859654e-05,
+      "loss": 0.5644,
+      "step": 252
+    },
+    {
+      "epoch": 0.17745046466771874,
+      "grad_norm": 41.07538986206055,
+      "learning_rate": 4.298245614035088e-05,
+      "loss": 0.5337,
+      "step": 253
+    },
+    {
+      "epoch": 0.17815184990355953,
+      "grad_norm": 66.12769317626953,
+      "learning_rate": 4.3157894736842105e-05,
+      "loss": 0.8491,
+      "step": 254
+    },
+    {
+      "epoch": 0.17885323513940032,
+      "grad_norm": 38.53345489501953,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 0.4,
+      "step": 255
+    },
+    {
+      "epoch": 0.17955462037524111,
+      "grad_norm": 37.57467269897461,
+      "learning_rate": 4.350877192982456e-05,
+      "loss": 0.6272,
+      "step": 256
+    },
+    {
+      "epoch": 0.18025600561108188,
+      "grad_norm": 27.241796493530273,
+      "learning_rate": 4.368421052631579e-05,
+      "loss": 0.2307,
+      "step": 257
+    },
+    {
+      "epoch": 0.18095739084692267,
+      "grad_norm": 24.353778839111328,
+      "learning_rate": 4.3859649122807014e-05,
+      "loss": 0.1238,
+      "step": 258
+    },
+    {
+      "epoch": 0.18165877608276346,
+      "grad_norm": 5.489215850830078,
+      "learning_rate": 4.403508771929824e-05,
+      "loss": 0.031,
+      "step": 259
+    },
+    {
+      "epoch": 0.18236016131860425,
+      "grad_norm": 16.305925369262695,
+      "learning_rate": 4.421052631578947e-05,
+      "loss": 0.0398,
+      "step": 260
+    },
+    {
+      "epoch": 0.18306154655444504,
+      "grad_norm": 6.403785228729248,
+      "learning_rate": 4.43859649122807e-05,
+      "loss": 0.0304,
+      "step": 261
+    },
+    {
+      "epoch": 0.18376293179028583,
+      "grad_norm": 4.471805572509766,
+      "learning_rate": 4.456140350877193e-05,
+      "loss": 0.0077,
+      "step": 262
+    },
+    {
+      "epoch": 0.1844643170261266,
+      "grad_norm": 0.9698334336280823,
+      "learning_rate": 4.473684210526316e-05,
+      "loss": 0.0031,
+      "step": 263
+    },
+    {
+      "epoch": 0.18516570226196738,
+      "grad_norm": 0.014244407415390015,
+      "learning_rate": 4.491228070175439e-05,
+      "loss": 0.0,
+      "step": 264
+    },
+    {
+      "epoch": 0.18586708749780817,
+      "grad_norm": 121.71415710449219,
+      "learning_rate": 4.508771929824562e-05,
+      "loss": 4.2289,
+      "step": 265
+    },
+    {
+      "epoch": 0.18656847273364896,
+      "grad_norm": 0.02348833531141281,
+      "learning_rate": 4.5263157894736846e-05,
+      "loss": 0.0001,
+      "step": 266
+    },
+    {
+      "epoch": 0.18726985796948975,
+      "grad_norm": 0.0016702099237591028,
+      "learning_rate": 4.5438596491228075e-05,
+      "loss": 0.0,
+      "step": 267
+    },
+    {
+      "epoch": 0.18797124320533054,
+      "grad_norm": 0.0027486609760671854,
+      "learning_rate": 4.56140350877193e-05,
+      "loss": 0.0,
+      "step": 268
+    },
+    {
+      "epoch": 0.1886726284411713,
+      "grad_norm": 0.0012574723223224282,
+      "learning_rate": 4.5789473684210527e-05,
+      "loss": 0.0,
+      "step": 269
+    },
+    {
+      "epoch": 0.1893740136770121,
+      "grad_norm": 0.002944325562566519,
+      "learning_rate": 4.5964912280701756e-05,
+      "loss": 0.0,
+      "step": 270
+    },
+    {
+      "epoch": 0.19007539891285288,
+      "grad_norm": 0.0020754581782966852,
+      "learning_rate": 4.6140350877192985e-05,
+      "loss": 0.0,
+      "step": 271
+    },
+    {
+      "epoch": 0.19077678414869367,
+      "grad_norm": 0.008111892268061638,
+      "learning_rate": 4.6315789473684214e-05,
+      "loss": 0.0,
+      "step": 272
+    },
+    {
+      "epoch": 0.19147816938453446,
+      "grad_norm": 118.25870513916016,
+      "learning_rate": 4.649122807017544e-05,
+      "loss": 3.5798,
+      "step": 273
+    },
+    {
+      "epoch": 0.19217955462037525,
+      "grad_norm": 118.91422271728516,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 2.0933,
+      "step": 274
+    },
+    {
+      "epoch": 0.19288093985621602,
+      "grad_norm": 0.1281914860010147,
+      "learning_rate": 4.68421052631579e-05,
+      "loss": 0.0004,
+      "step": 275
+    },
+    {
+      "epoch": 0.1935823250920568,
+      "grad_norm": 0.9166204929351807,
+      "learning_rate": 4.701754385964913e-05,
+      "loss": 0.0034,
+      "step": 276
+    },
+    {
+      "epoch": 0.1942837103278976,
+      "grad_norm": 3.7037837505340576,
+      "learning_rate": 4.719298245614036e-05,
+      "loss": 0.0179,
+      "step": 277
+    },
+    {
+      "epoch": 0.1949850955637384,
+      "grad_norm": 0.716585099697113,
+      "learning_rate": 4.736842105263158e-05,
+      "loss": 0.0028,
+      "step": 278
+    },
+    {
+      "epoch": 0.19568648079957918,
+      "grad_norm": 119.46430969238281,
+      "learning_rate": 4.754385964912281e-05,
+      "loss": 1.7075,
+      "step": 279
+    },
+    {
+      "epoch": 0.19638786603541997,
+      "grad_norm": 0.07794260233640671,
+      "learning_rate": 4.771929824561404e-05,
+      "loss": 0.0003,
+      "step": 280
+    },
+    {
+      "epoch": 0.19708925127126073,
+      "grad_norm": 119.65841674804688,
+      "learning_rate": 4.789473684210526e-05,
+      "loss": 3.3725,
+      "step": 281
+    },
+    {
+      "epoch": 0.19779063650710152,
+      "grad_norm": 0.06007947400212288,
+      "learning_rate": 4.807017543859649e-05,
+      "loss": 0.0002,
+      "step": 282
+    },
+    {
+      "epoch": 0.1984920217429423,
+      "grad_norm": 0.33029523491859436,
+      "learning_rate": 4.824561403508772e-05,
+      "loss": 0.0013,
+      "step": 283
+    },
+    {
+      "epoch": 0.1991934069787831,
+      "grad_norm": 0.19337859749794006,
+      "learning_rate": 4.842105263157895e-05,
+      "loss": 0.0007,
+      "step": 284
+    },
+    {
+      "epoch": 0.1998947922146239,
+      "grad_norm": 0.4252503514289856,
+      "learning_rate": 4.859649122807018e-05,
+      "loss": 0.0017,
+      "step": 285
+    },
+    {
+      "epoch": 0.20059617745046468,
+      "grad_norm": 0.07317744940519333,
+      "learning_rate": 4.8771929824561406e-05,
+      "loss": 0.0003,
+      "step": 286
+    },
+    {
+      "epoch": 0.20129756268630544,
+      "grad_norm": 0.01806594245135784,
+      "learning_rate": 4.8947368421052635e-05,
+      "loss": 0.0001,
+      "step": 287
+    },
+    {
+      "epoch": 0.20199894792214623,
+      "grad_norm": 106.39412689208984,
+      "learning_rate": 4.912280701754386e-05,
+      "loss": 3.9844,
+      "step": 288
+    },
+    {
+      "epoch": 0.20270033315798702,
+      "grad_norm": 0.008989217691123486,
+      "learning_rate": 4.9298245614035086e-05,
+      "loss": 0.0,
+      "step": 289
+    },
+    {
+      "epoch": 0.2034017183938278,
+      "grad_norm": 0.05597059056162834,
+      "learning_rate": 4.9473684210526315e-05,
+      "loss": 0.0002,
+      "step": 290
+    },
+    {
+      "epoch": 0.2041031036296686,
+      "grad_norm": 113.05389404296875,
+      "learning_rate": 4.9649122807017544e-05,
+      "loss": 2.5036,
+      "step": 291
+    },
+    {
+      "epoch": 0.2048044888655094,
+      "grad_norm": 0.39913010597229004,
+      "learning_rate": 4.9824561403508773e-05,
+      "loss": 0.0015,
+      "step": 292
+    },
+    {
+      "epoch": 0.20550587410135016,
+      "grad_norm": 0.944106936454773,
+      "learning_rate": 5e-05,
+      "loss": 0.0038,
+      "step": 293
+    },
+    {
+      "epoch": 0.20620725933719095,
+      "grad_norm": 108.76079559326172,
+      "learning_rate": 4.9999995795715716e-05,
+      "loss": 1.4773,
+      "step": 294
+    },
+    {
+      "epoch": 0.20690864457303174,
+      "grad_norm": 1.2869099378585815,
+      "learning_rate": 4.999998318286425e-05,
+      "loss": 0.0054,
+      "step": 295
+    },
+    {
+      "epoch": 0.20761002980887253,
+      "grad_norm": 0.7701173424720764,
+      "learning_rate": 4.999996216144987e-05,
+      "loss": 0.0029,
+      "step": 296
+    },
+    {
+      "epoch": 0.20831141504471332,
+      "grad_norm": 104.06690216064453,
+      "learning_rate": 4.9999932731479625e-05,
+      "loss": 1.2931,
+      "step": 297
+    },
+    {
+      "epoch": 0.2090128002805541,
+      "grad_norm": 234.58531188964844,
+      "learning_rate": 4.999989489296344e-05,
+      "loss": 4.5688,
+      "step": 298
+    },
+    {
+      "epoch": 0.20971418551639487,
+      "grad_norm": 108.04949188232422,
+      "learning_rate": 4.999984864591401e-05,
+      "loss": 1.5544,
+      "step": 299
+    },
+    {
+      "epoch": 0.21041557075223566,
+      "grad_norm": 1.8640260696411133,
+      "learning_rate": 4.999979399034691e-05,
+      "loss": 0.0078,
+      "step": 300
+    },
+    {
+      "epoch": 0.21111695598807645,
+      "grad_norm": 3.089526414871216,
+      "learning_rate": 4.999973092628052e-05,
+      "loss": 0.0138,
+      "step": 301
+    },
+    {
+      "epoch": 0.21181834122391724,
+      "grad_norm": 1.4183728694915771,
+      "learning_rate": 4.999965945373605e-05,
+      "loss": 0.006,
+      "step": 302
+    },
+    {
+      "epoch": 0.21251972645975803,
+      "grad_norm": 0.9733534455299377,
+      "learning_rate": 4.9999579572737533e-05,
+      "loss": 0.0038,
+      "step": 303
+    },
+    {
+      "epoch": 0.21322111169559882,
+      "grad_norm": 121.11236572265625,
+      "learning_rate": 4.9999491283311836e-05,
+      "loss": 2.2543,
+      "step": 304
+    },
+    {
+      "epoch": 0.21392249693143958,
+      "grad_norm": 117.718017578125,
+      "learning_rate": 4.999939458548868e-05,
+      "loss": 2.5007,
+      "step": 305
+    },
+    {
+      "epoch": 0.21462388216728037,
+      "grad_norm": 0.013796065002679825,
+      "learning_rate": 4.9999289479300557e-05,
+      "loss": 0.0,
+      "step": 306
+    },
+    {
+      "epoch": 0.21532526740312116,
+      "grad_norm": 116.46581268310547,
+      "learning_rate": 4.999917596478283e-05,
+      "loss": 2.6005,
+      "step": 307
+    },
+    {
+      "epoch": 0.21602665263896195,
+      "grad_norm": 0.3713426887989044,
+      "learning_rate": 4.999905404197368e-05,
+      "loss": 0.0014,
+      "step": 308
+    },
+    {
+      "epoch": 0.21672803787480274,
+      "grad_norm": 0.44403278827667236,
+      "learning_rate": 4.999892371091411e-05,
+      "loss": 0.0017,
+      "step": 309
+    },
+    {
+      "epoch": 0.21742942311064353,
+      "grad_norm": 139.35816955566406,
+      "learning_rate": 4.999878497164797e-05,
+      "loss": 11.1347,
+      "step": 310
+    },
+    {
+      "epoch": 0.2181308083464843,
+      "grad_norm": 0.20512554049491882,
+      "learning_rate": 4.999863782422191e-05,
+      "loss": 0.0007,
+      "step": 311
+    },
+    {
+      "epoch": 0.21883219358232509,
+      "grad_norm": 88.1954345703125,
+      "learning_rate": 4.9998482268685434e-05,
+      "loss": 0.843,
+      "step": 312
+    },
+    {
+      "epoch": 0.21953357881816588,
+      "grad_norm": 93.83118438720703,
+      "learning_rate": 4.999831830509084e-05,
+      "loss": 1.0804,
+      "step": 313
+    },
+    {
+      "epoch": 0.22023496405400667,
+      "grad_norm": 5.042333602905273,
+      "learning_rate": 4.99981459334933e-05,
+      "loss": 0.0256,
+      "step": 314
+    },
+    {
+      "epoch": 0.22093634928984746,
+      "grad_norm": 7.477559566497803,
+      "learning_rate": 4.9997965153950775e-05,
+      "loss": 0.0402,
+      "step": 315
+    },
+    {
+      "epoch": 0.22163773452568825,
+      "grad_norm": 156.78961181640625,
+      "learning_rate": 4.999777596652408e-05,
+      "loss": 6.0824,
+      "step": 316
+    },
+    {
+      "epoch": 0.222339119761529,
+      "grad_norm": 9.518619537353516,
+      "learning_rate": 4.999757837127683e-05,
+      "loss": 0.0577,
+      "step": 317
+    },
+    {
+      "epoch": 0.2230405049973698,
+      "grad_norm": 2.860729455947876,
+      "learning_rate": 4.9997372368275506e-05,
+      "loss": 0.0135,
+      "step": 318
+    },
+    {
+      "epoch": 0.2237418902332106,
+      "grad_norm": 1.1225600242614746,
+      "learning_rate": 4.999715795758938e-05,
+      "loss": 0.0049,
+      "step": 319
+    },
+    {
+      "epoch": 0.22444327546905138,
+      "grad_norm": 0.12374955415725708,
+      "learning_rate": 4.9996935139290575e-05,
+      "loss": 0.0004,
+      "step": 320
+    },
+    {
+      "epoch": 0.22514466070489217,
+      "grad_norm": 112.18521881103516,
+      "learning_rate": 4.999670391345403e-05,
+      "loss": 3.4372,
+      "step": 321
+    },
+    {
+      "epoch": 0.22584604594073296,
+      "grad_norm": 0.0004930093418806791,
+      "learning_rate": 4.999646428015752e-05,
+      "loss": 0.0,
+      "step": 322
+    },
+    {
+      "epoch": 0.22654743117657372,
+      "grad_norm": 0.00027333476464264095,
+      "learning_rate": 4.9996216239481643e-05,
+      "loss": 0.0,
+      "step": 323
+    },
+    {
+      "epoch": 0.2272488164124145,
+      "grad_norm": 0.0001735862751957029,
+      "learning_rate": 4.999595979150982e-05,
+      "loss": 0.0,
+      "step": 324
+    },
+    {
+      "epoch": 0.2279502016482553,
+      "grad_norm": 2.878743180190213e-05,
+      "learning_rate": 4.9995694936328316e-05,
+      "loss": 0.0,
+      "step": 325
+    },
+    {
+      "epoch": 0.2286515868840961,
+      "grad_norm": 109.58158111572266,
+      "learning_rate": 4.99954216740262e-05,
+      "loss": 4.991,
+      "step": 326
+    },
+    {
+      "epoch": 0.22935297211993688,
+      "grad_norm": 113.2159194946289,
+      "learning_rate": 4.9995140004695396e-05,
+      "loss": 4.7235,
+      "step": 327
+    },
+    {
+      "epoch": 0.23005435735577767,
+      "grad_norm": 0.00015405479643959552,
+      "learning_rate": 4.999484992843063e-05,
+      "loss": 0.0,
+      "step": 328
+    },
+    {
+      "epoch": 0.23075574259161843,
+      "grad_norm": 0.011504331603646278,
+      "learning_rate": 4.999455144532947e-05,
+      "loss": 0.0,
+      "step": 329
+    },
+    {
+      "epoch": 0.23145712782745922,
+      "grad_norm": 0.001587590086273849,
+      "learning_rate": 4.9994244555492315e-05,
+      "loss": 0.0,
+      "step": 330
+    },
+    {
+      "epoch": 0.23215851306330001,
+      "grad_norm": 0.0038678112905472517,
+      "learning_rate": 4.999392925902238e-05,
+      "loss": 0.0,
+      "step": 331
+    },
+    {
+      "epoch": 0.2328598982991408,
+      "grad_norm": 0.044445931911468506,
+      "learning_rate": 4.999360555602571e-05,
+      "loss": 0.0001,
+      "step": 332
+    },
+    {
+      "epoch": 0.2335612835349816,
+      "grad_norm": 108.20228576660156,
+      "learning_rate": 4.999327344661118e-05,
+      "loss": 2.0004,
+      "step": 333
+    },
+    {
+      "epoch": 0.23426266877082239,
+      "grad_norm": 0.17047792673110962,
+      "learning_rate": 4.9992932930890505e-05,
+      "loss": 0.0005,
+      "step": 334
+    },
+    {
+      "epoch": 0.23496405400666315,
+      "grad_norm": 0.39230290055274963,
+      "learning_rate": 4.999258400897819e-05,
+      "loss": 0.0016,
+      "step": 335
+    },
+    {
+      "epoch": 0.23566543924250394,
+      "grad_norm": 191.8765106201172,
+      "learning_rate": 4.999222668099163e-05,
+      "loss": 2.4042,
+      "step": 336
+    },
+    {
+      "epoch": 0.23636682447834473,
+      "grad_norm": 87.04983520507812,
+      "learning_rate": 4.999186094705097e-05,
+      "loss": 0.9342,
+      "step": 337
+    },
+    {
+      "epoch": 0.23706820971418552,
+      "grad_norm": 14.748551368713379,
+      "learning_rate": 4.999148680727925e-05,
+      "loss": 0.1165,
+      "step": 338
+    },
+    {
+      "epoch": 0.2377695949500263,
+      "grad_norm": 16.103792190551758,
+      "learning_rate": 4.999110426180229e-05,
+      "loss": 0.11,
+      "step": 339
+    },
+    {
+      "epoch": 0.2384709801858671,
+      "grad_norm": 4.507053375244141,
+      "learning_rate": 4.999071331074877e-05,
+      "loss": 0.1421,
+      "step": 340
+    },
+    {
+      "epoch": 0.23917236542170786,
+      "grad_norm": 8.822351455688477,
+      "learning_rate": 4.999031395425018e-05,
+      "loss": 0.2372,
+      "step": 341
+    },
+    {
+      "epoch": 0.23987375065754865,
+      "grad_norm": 19.92313575744629,
+      "learning_rate": 4.998990619244084e-05,
+      "loss": 0.1537,
+      "step": 342
+    },
+    {
+      "epoch": 0.24057513589338944,
+      "grad_norm": 12.00966739654541,
+      "learning_rate": 4.998949002545789e-05,
+      "loss": 0.0872,
+      "step": 343
+    },
+    {
+      "epoch": 0.24127652112923023,
+      "grad_norm": 4.180269718170166,
+      "learning_rate": 4.998906545344132e-05,
+      "loss": 0.0218,
+      "step": 344
+    },
+    {
+      "epoch": 0.24197790636507102,
+      "grad_norm": 0.596821129322052,
+      "learning_rate": 4.998863247653392e-05,
+      "loss": 0.0025,
+      "step": 345
+    },
+    {
+      "epoch": 0.2426792916009118,
+      "grad_norm": 0.4522903263568878,
+      "learning_rate": 4.998819109488132e-05,
+      "loss": 0.0013,
+      "step": 346
+    },
+    {
+      "epoch": 0.24338067683675257,
+      "grad_norm": 0.0007598726078867912,
+      "learning_rate": 4.998774130863199e-05,
+      "loss": 0.0,
+      "step": 347
+    },
+    {
+      "epoch": 0.24408206207259336,
+      "grad_norm": 8.103512664092705e-05,
+      "learning_rate": 4.998728311793719e-05,
+      "loss": 0.0,
+      "step": 348
+    },
+    {
+      "epoch": 0.24478344730843415,
+      "grad_norm": 105.80138397216797,
+      "learning_rate": 4.998681652295104e-05,
+      "loss": 4.8223,
+      "step": 349
+    },
+    {
+      "epoch": 0.24548483254427494,
+      "grad_norm": 127.92369079589844,
+      "learning_rate": 4.998634152383047e-05,
+      "loss": 7.0698,
+      "step": 350
+    },
+    {
+      "epoch": 0.24618621778011573,
+      "grad_norm": 2.636051704030251e-06,
+      "learning_rate": 4.998585812073525e-05,
+      "loss": 0.0,
+      "step": 351
+    },
+    {
+      "epoch": 0.24688760301595652,
+      "grad_norm": 2.710710077735712e-06,
+      "learning_rate": 4.9985366313827975e-05,
+      "loss": 0.0,
+      "step": 352
+    },
+    {
+      "epoch": 0.2475889882517973,
+      "grad_norm": 103.96717834472656,
+      "learning_rate": 4.998486610327405e-05,
+      "loss": 6.485,
+      "step": 353
+    },
+    {
+      "epoch": 0.24829037348763808,
+      "grad_norm": 103.35551452636719,
+      "learning_rate": 4.9984357489241715e-05,
+      "loss": 5.7037,
+      "step": 354
+    },
+    {
+      "epoch": 0.24899175872347887,
+      "grad_norm": 100.56452178955078,
+      "learning_rate": 4.998384047190204e-05,
+      "loss": 4.1908,
+      "step": 355
+    },
+    {
+      "epoch": 0.24969314395931966,
+      "grad_norm": 200.88751220703125,
+      "learning_rate": 4.998331505142893e-05,
+      "loss": 4.8984,
+      "step": 356
+    },
+    {
+      "epoch": 0.2503945291951604,
+      "grad_norm": 77.69247436523438,
+      "learning_rate": 4.99827812279991e-05,
+      "loss": 0.9449,
+      "step": 357
+    },
+    {
+      "epoch": 0.2510959144310012,
+      "grad_norm": 26.622453689575195,
+      "learning_rate": 4.9982239001792095e-05,
+      "loss": 0.3613,
+      "step": 358
+    },
+    {
+      "epoch": 0.251797299666842,
+      "grad_norm": 26.442935943603516,
+      "learning_rate": 4.9981688372990284e-05,
+      "loss": 0.5424,
+      "step": 359
+    },
+    {
+      "epoch": 0.2524986849026828,
+      "grad_norm": 39.277618408203125,
+      "learning_rate": 4.9981129341778886e-05,
+      "loss": 0.9456,
+      "step": 360
+    },
+    {
+      "epoch": 0.2532000701385236,
+      "grad_norm": 43.96021270751953,
+      "learning_rate": 4.9980561908345916e-05,
+      "loss": 0.6765,
+      "step": 361
+    },
+    {
+      "epoch": 0.25390145537436437,
+      "grad_norm": 14.254291534423828,
+      "learning_rate": 4.997998607288222e-05,
+      "loss": 0.1276,
+      "step": 362
+    },
+    {
+      "epoch": 0.25460284061020516,
+      "grad_norm": 0.6544491052627563,
+      "learning_rate": 4.9979401835581476e-05,
+      "loss": 0.0032,
+      "step": 363
+    },
+    {
+      "epoch": 0.25530422584604595,
+      "grad_norm": 0.011080354452133179,
+      "learning_rate": 4.99788091966402e-05,
+      "loss": 0.0,
+      "step": 364
+    },
+    {
+      "epoch": 0.25600561108188674,
+      "grad_norm": 284.8055114746094,
+      "learning_rate": 4.997820815625771e-05,
+      "loss": 7.6058,
+      "step": 365
+    },
+    {
+      "epoch": 0.25670699631772753,
+      "grad_norm": 89.02140045166016,
+      "learning_rate": 4.997759871463618e-05,
+      "loss": 5.079,
+      "step": 366
+    },
+    {
+      "epoch": 0.2574083815535683,
+      "grad_norm": 4.7222201828844845e-05,
+      "learning_rate": 4.997698087198056e-05,
+      "loss": 0.0,
+      "step": 367
+    },
+    {
+      "epoch": 0.25810976678940906,
+      "grad_norm": 85.66144561767578,
+      "learning_rate": 4.997635462849869e-05,
+      "loss": 4.8183,
+      "step": 368
+    },
+    {
+      "epoch": 0.25881115202524985,
+      "grad_norm": 0.0015366720035672188,
+      "learning_rate": 4.997571998440118e-05,
+      "loss": 0.0,
+      "step": 369
+    },
+    {
+      "epoch": 0.25951253726109064,
+      "grad_norm": 100.14698028564453,
+      "learning_rate": 4.99750769399015e-05,
+      "loss": 2.2663,
+      "step": 370
+    },
+    {
+      "epoch": 0.2602139224969314,
+      "grad_norm": 0.4829590618610382,
+      "learning_rate": 4.997442549521592e-05,
+      "loss": 0.0021,
+      "step": 371
+    },
+    {
+      "epoch": 0.2609153077327722,
+      "grad_norm": 70.44209289550781,
+      "learning_rate": 4.9973765650563564e-05,
+      "loss": 0.7532,
+      "step": 372
+    },
+    {
+      "epoch": 0.261616692968613,
+      "grad_norm": 272.1163024902344,
+      "learning_rate": 4.9973097406166355e-05,
+      "loss": 3.9935,
+      "step": 373
+    },
+    {
+      "epoch": 0.2623180782044538,
+      "grad_norm": 25.12706184387207,
+      "learning_rate": 4.997242076224906e-05,
+      "loss": 0.3738,
+      "step": 374
+    },
+    {
+      "epoch": 0.2630194634402946,
+      "grad_norm": 35.15785598754883,
+      "learning_rate": 4.997173571903926e-05,
+      "loss": 0.6808,
+      "step": 375
+    },
+    {
+      "epoch": 0.2637208486761354,
+      "grad_norm": 27.46406364440918,
+      "learning_rate": 4.9971042276767366e-05,
+      "loss": 0.2568,
+      "step": 376
+    },
+    {
+      "epoch": 0.26442223391197617,
+      "grad_norm": 6.682714462280273,
+      "learning_rate": 4.997034043566661e-05,
+      "loss": 0.0374,
+      "step": 377
+    },
+    {
+      "epoch": 0.26512361914781696,
+      "grad_norm": 0.057287223637104034,
+      "learning_rate": 4.9969630195973045e-05,
+      "loss": 0.0001,
+      "step": 378
+    },
+    {
+      "epoch": 0.26582500438365775,
+      "grad_norm": 0.006871811114251614,
+      "learning_rate": 4.9968911557925564e-05,
+      "loss": 0.0,
+      "step": 379
+    },
+    {
+      "epoch": 0.2665263896194985,
+      "grad_norm": 0.0037054684944450855,
+      "learning_rate": 4.996818452176587e-05,
+      "loss": 0.0,
+      "step": 380
+    },
+    {
+      "epoch": 0.2672277748553393,
+      "grad_norm": 123.15022277832031,
+      "learning_rate": 4.9967449087738505e-05,
+      "loss": 5.8346,
+      "step": 381
+    },
+    {
+      "epoch": 0.26792916009118006,
+      "grad_norm": 121.81306457519531,
+      "learning_rate": 4.996670525609082e-05,
+      "loss": 4.3778,
+      "step": 382
+    },
+    {
+      "epoch": 0.26863054532702085,
+      "grad_norm": 0.00026572938077151775,
+      "learning_rate": 4.9965953027072996e-05,
+      "loss": 0.0,
+      "step": 383
+    },
+    {
+      "epoch": 0.26933193056286164,
+      "grad_norm": 198.09776306152344,
+      "learning_rate": 4.996519240093806e-05,
+      "loss": 5.2574,
+      "step": 384
+    },
+    {
+      "epoch": 0.27003331579870243,
+      "grad_norm": 130.21214294433594,
+      "learning_rate": 4.996442337794181e-05,
+      "loss": 3.217,
+      "step": 385
+    },
+    {
+      "epoch": 0.2707347010345432,
+      "grad_norm": 133.94178771972656,
+      "learning_rate": 4.996364595834291e-05,
+      "loss": 1.3412,
+      "step": 386
+    },
+    {
+      "epoch": 0.271436086270384,
+      "grad_norm": 12.223856925964355,
+      "learning_rate": 4.9962860142402853e-05,
+      "loss": 0.0924,
+      "step": 387
+    },
+    {
+      "epoch": 0.2721374715062248,
+      "grad_norm": 21.94683074951172,
+      "learning_rate": 4.9962065930385934e-05,
+      "loss": 0.1824,
+      "step": 388
+    },
+    {
+      "epoch": 0.2728388567420656,
+      "grad_norm": 46.802406311035156,
+      "learning_rate": 4.996126332255929e-05,
+      "loss": 0.3585,
+      "step": 389
+    },
+    {
+      "epoch": 0.2735402419779064,
+      "grad_norm": 7.21342134475708,
+      "learning_rate": 4.9960452319192866e-05,
+      "loss": 0.1045,
+      "step": 390
+    },
+    {
+      "epoch": 0.2742416272137472,
+      "grad_norm": 33.5169677734375,
+      "learning_rate": 4.995963292055943e-05,
+      "loss": 0.248,
+      "step": 391
+    },
+    {
+      "epoch": 0.2749430124495879,
+      "grad_norm": 16.911136627197266,
+      "learning_rate": 4.995880512693458e-05,
+      "loss": 0.0876,
+      "step": 392
+    },
+    {
+      "epoch": 0.2756443976854287,
+      "grad_norm": 206.7276153564453,
+      "learning_rate": 4.995796893859675e-05,
+      "loss": 1.6484,
+      "step": 393
+    },
+    {
+      "epoch": 0.2763457829212695,
+      "grad_norm": 94.29922485351562,
+      "learning_rate": 4.9957124355827186e-05,
+      "loss": 0.8745,
+      "step": 394
+    },
+    {
+      "epoch": 0.2770471681571103,
+      "grad_norm": 156.6600799560547,
+      "learning_rate": 4.9956271378909945e-05,
+      "loss": 0.9049,
+      "step": 395
+    },
+    {
+      "epoch": 0.27774855339295107,
+      "grad_norm": 6.1268744468688965,
+      "learning_rate": 4.995541000813193e-05,
+      "loss": 0.0343,
+      "step": 396
+    },
+    {
+      "epoch": 0.27844993862879186,
+      "grad_norm": 1.731136679649353,
+      "learning_rate": 4.995454024378285e-05,
+      "loss": 0.0075,
+      "step": 397
+    },
+    {
+      "epoch": 0.27915132386463265,
+      "grad_norm": 144.5174560546875,
+      "learning_rate": 4.995366208615526e-05,
+      "loss": 2.6072,
+      "step": 398
+    },
+    {
+      "epoch": 0.27985270910047344,
+      "grad_norm": 0.19711287319660187,
+      "learning_rate": 4.9952775535544494e-05,
+      "loss": 0.0007,
+      "step": 399
+    },
+    {
+      "epoch": 0.28055409433631423,
+      "grad_norm": 0.13065628707408905,
+      "learning_rate": 4.9951880592248754e-05,
+      "loss": 0.0004,
+      "step": 400
+    },
+    {
+      "epoch": 0.281255479572155,
+      "grad_norm": 125.28437042236328,
+      "learning_rate": 4.995097725656905e-05,
+      "loss": 1.8792,
+      "step": 401
+    },
+    {
+      "epoch": 0.2819568648079958,
+      "grad_norm": 111.38099670410156,
+      "learning_rate": 4.9950065528809214e-05,
+      "loss": 2.548,
+      "step": 402
+    },
+    {
+      "epoch": 0.2826582500438366,
+      "grad_norm": 1.2473199367523193,
+      "learning_rate": 4.994914540927589e-05,
+      "loss": 0.0042,
+      "step": 403
+    },
+    {
+      "epoch": 0.28335963527967734,
+      "grad_norm": 0.7474699020385742,
+      "learning_rate": 4.994821689827856e-05,
+      "loss": 0.0026,
+      "step": 404
+    },
+    {
+      "epoch": 0.2840610205155181,
+      "grad_norm": 0.14196322858333588,
+      "learning_rate": 4.99472799961295e-05,
+      "loss": 0.0005,
+      "step": 405
+    },
+    {
+      "epoch": 0.2847624057513589,
+      "grad_norm": 0.6420115828514099,
+      "learning_rate": 4.994633470314387e-05,
+      "loss": 0.0026,
+      "step": 406
+    },
+    {
+      "epoch": 0.2854637909871997,
+      "grad_norm": 213.8779296875,
+      "learning_rate": 4.994538101963958e-05,
+      "loss": 2.1989,
+      "step": 407
+    },
+    {
+      "epoch": 0.2861651762230405,
+      "grad_norm": 108.08120727539062,
+      "learning_rate": 4.994441894593741e-05,
+      "loss": 3.9965,
+      "step": 408
+    },
+    {
+      "epoch": 0.2868665614588813,
+      "grad_norm": 0.40015965700149536,
+      "learning_rate": 4.994344848236094e-05,
+      "loss": 0.0016,
+      "step": 409
+    },
+    {
+      "epoch": 0.2875679466947221,
+      "grad_norm": 125.80431365966797,
+      "learning_rate": 4.9942469629236585e-05,
+      "loss": 2.3139,
+      "step": 410
+    },
+    {
+      "epoch": 0.28826933193056287,
+      "grad_norm": 101.27831268310547,
+      "learning_rate": 4.994148238689357e-05,
+      "loss": 2.1065,
+      "step": 411
+    },
+    {
+      "epoch": 0.28897071716640366,
+      "grad_norm": 3.526338577270508,
+      "learning_rate": 4.9940486755663946e-05,
+      "loss": 0.015,
+      "step": 412
+    },
+    {
+      "epoch": 0.28967210240224445,
+      "grad_norm": 2.501554489135742,
+      "learning_rate": 4.993948273588259e-05,
+      "loss": 0.0123,
+      "step": 413
+    },
+    {
+      "epoch": 0.29037348763808524,
+      "grad_norm": 75.39434814453125,
+      "learning_rate": 4.993847032788719e-05,
+      "loss": 0.7107,
+      "step": 414
+    },
+    {
+      "epoch": 0.291074872873926,
+      "grad_norm": 3.751978635787964,
+      "learning_rate": 4.993744953201827e-05,
+      "loss": 0.0183,
+      "step": 415
+    },
+    {
+      "epoch": 0.29177625810976676,
+      "grad_norm": 5.31666374206543,
+      "learning_rate": 4.9936420348619164e-05,
+      "loss": 0.0385,
+      "step": 416
+    },
+    {
+      "epoch": 0.29247764334560755,
+      "grad_norm": 104.77239227294922,
+      "learning_rate": 4.9935382778036036e-05,
+      "loss": 1.7353,
+      "step": 417
+    },
+    {
+      "epoch": 0.29317902858144834,
+      "grad_norm": 92.44829559326172,
+      "learning_rate": 4.993433682061786e-05,
+      "loss": 1.1895,
+      "step": 418
+    },
+    {
+      "epoch": 0.29388041381728913,
+      "grad_norm": 89.00714874267578,
+      "learning_rate": 4.993328247671643e-05,
+      "loss": 1.2323,
+      "step": 419
+    },
+    {
+      "epoch": 0.2945817990531299,
+      "grad_norm": 172.867431640625,
+      "learning_rate": 4.993221974668637e-05,
+      "loss": 1.7303,
+      "step": 420
+    },
+    {
+      "epoch": 0.2952831842889707,
+      "grad_norm": 19.69775390625,
+      "learning_rate": 4.9931148630885127e-05,
+      "loss": 0.1678,
+      "step": 421
+    },
+    {
+      "epoch": 0.2959845695248115,
+      "grad_norm": 22.828763961791992,
+      "learning_rate": 4.9930069129672965e-05,
+      "loss": 0.3184,
+      "step": 422
+    },
+    {
+      "epoch": 0.2966859547606523,
+      "grad_norm": 19.460153579711914,
+      "learning_rate": 4.992898124341296e-05,
+      "loss": 0.2748,
+      "step": 423
+    },
+    {
+      "epoch": 0.2973873399964931,
+      "grad_norm": 2.9695911407470703,
+      "learning_rate": 4.9927884972471015e-05,
+      "loss": 0.1925,
+      "step": 424
+    },
+    {
+      "epoch": 0.2980887252323339,
+      "grad_norm": 12.33647346496582,
+      "learning_rate": 4.9926780317215854e-05,
+      "loss": 0.1908,
+      "step": 425
+    },
+    {
+      "epoch": 0.29879011046817466,
+      "grad_norm": 2.8421826362609863,
+      "learning_rate": 4.992566727801902e-05,
+      "loss": 0.101,
+      "step": 426
+    },
+    {
+      "epoch": 0.29949149570401545,
+      "grad_norm": 171.8658447265625,
+      "learning_rate": 4.9924545855254875e-05,
+      "loss": 3.9311,
+      "step": 427
+    },
+    {
+      "epoch": 0.30019288093985624,
+      "grad_norm": 83.66468048095703,
+      "learning_rate": 4.992341604930061e-05,
+      "loss": 1.6806,
+      "step": 428
+    },
+    {
+      "epoch": 0.300894266175697,
+      "grad_norm": 0.7609736323356628,
+      "learning_rate": 4.9922277860536216e-05,
+      "loss": 0.0041,
+      "step": 429
+    },
+    {
+      "epoch": 0.30159565141153777,
+      "grad_norm": 4.077641487121582,
+      "learning_rate": 4.9921131289344514e-05,
+      "loss": 0.0418,
+      "step": 430
+    },
+    {
+      "epoch": 0.30229703664737856,
+      "grad_norm": 3.2341866493225098,
+      "learning_rate": 4.9919976336111155e-05,
+      "loss": 0.0293,
+      "step": 431
+    },
+    {
+      "epoch": 0.30299842188321935,
+      "grad_norm": 148.74095153808594,
+      "learning_rate": 4.991881300122458e-05,
+      "loss": 3.3951,
+      "step": 432
+    },
+    {
+      "epoch": 0.30369980711906014,
+      "grad_norm": 77.130126953125,
+      "learning_rate": 4.99176412850761e-05,
+      "loss": 2.9407,
+      "step": 433
+    },
+    {
+      "epoch": 0.30440119235490093,
+      "grad_norm": 162.98777770996094,
+      "learning_rate": 4.991646118805978e-05,
+      "loss": 4.3504,
+      "step": 434
+    },
+    {
+      "epoch": 0.3051025775907417,
+      "grad_norm": 84.4205551147461,
+      "learning_rate": 4.991527271057256e-05,
+      "loss": 1.7142,
+      "step": 435
+    },
+    {
+      "epoch": 0.3058039628265825,
+      "grad_norm": 0.8636782765388489,
+      "learning_rate": 4.991407585301416e-05,
+      "loss": 0.0053,
+      "step": 436
+    },
+    {
+      "epoch": 0.3065053480624233,
+      "grad_norm": 1.0418014526367188,
+      "learning_rate": 4.9912870615787144e-05,
+      "loss": 0.0059,
+      "step": 437
+    },
+    {
+      "epoch": 0.3072067332982641,
+      "grad_norm": 0.7050487399101257,
+      "learning_rate": 4.991165699929688e-05,
+      "loss": 0.0041,
+      "step": 438
+    },
+    {
+      "epoch": 0.3079081185341049,
+      "grad_norm": 0.3087022304534912,
+      "learning_rate": 4.991043500395157e-05,
+      "loss": 0.0017,
+      "step": 439
+    },
+    {
+      "epoch": 0.30860950376994567,
+      "grad_norm": 0.09886661171913147,
+      "learning_rate": 4.990920463016221e-05,
+      "loss": 0.0005,
+      "step": 440
+    },
+    {
+      "epoch": 0.3093108890057864,
+      "grad_norm": 0.03563763573765755,
+      "learning_rate": 4.9907965878342634e-05,
+      "loss": 0.0002,
+      "step": 441
+    },
+    {
+      "epoch": 0.3100122742416272,
+      "grad_norm": 79.55757141113281,
+      "learning_rate": 4.990671874890948e-05,
+      "loss": 2.966,
+      "step": 442
+    },
+    {
+      "epoch": 0.310713659477468,
+      "grad_norm": 0.6562897562980652,
+      "learning_rate": 4.990546324228222e-05,
+      "loss": 0.0022,
+      "step": 443
+    },
+    {
+      "epoch": 0.3114150447133088,
+      "grad_norm": 74.84564208984375,
+      "learning_rate": 4.990419935888313e-05,
+      "loss": 3.0154,
+      "step": 444
+    },
+    {
+      "epoch": 0.31211642994914957,
+      "grad_norm": 0.03038967400789261,
+      "learning_rate": 4.9902927099137305e-05,
+      "loss": 0.0002,
+      "step": 445
+    },
+    {
+      "epoch": 0.31281781518499036,
+      "grad_norm": 0.023526914417743683,
+      "learning_rate": 4.9901646463472674e-05,
+      "loss": 0.0001,
+      "step": 446
+    },
+    {
+      "epoch": 0.31351920042083115,
+      "grad_norm": 0.08529358357191086,
+      "learning_rate": 4.990035745231995e-05,
+      "loss": 0.0005,
+      "step": 447
+    },
+    {
+      "epoch": 0.31422058565667194,
+      "grad_norm": 0.11940905451774597,
+      "learning_rate": 4.989906006611269e-05,
+      "loss": 0.0007,
+      "step": 448
+    },
+    {
+      "epoch": 0.3149219708925127,
+      "grad_norm": 99.68978118896484,
+      "learning_rate": 4.989775430528727e-05,
+      "loss": 1.6958,
+      "step": 449
+    },
+    {
+      "epoch": 0.3156233561283535,
+      "grad_norm": 0.3344658315181732,
+      "learning_rate": 4.9896440170282864e-05,
+      "loss": 0.002,
+      "step": 450
+    },
+    {
+      "epoch": 0.3163247413641943,
+      "grad_norm": 0.7184603810310364,
+      "learning_rate": 4.9895117661541476e-05,
+      "loss": 0.0043,
+      "step": 451
+    },
+    {
+      "epoch": 0.3170261266000351,
+      "grad_norm": 0.5113491415977478,
+      "learning_rate": 4.989378677950791e-05,
+      "loss": 0.0031,
+      "step": 452
+    },
+    {
+      "epoch": 0.31772751183587583,
+      "grad_norm": 0.49223193526268005,
+      "learning_rate": 4.989244752462982e-05,
+      "loss": 0.0029,
+      "step": 453
+    },
+    {
+      "epoch": 0.3184288970717166,
+      "grad_norm": 64.00535583496094,
+      "learning_rate": 4.989109989735763e-05,
+      "loss": 1.2629,
+      "step": 454
+    },
+    {
+      "epoch": 0.3191302823075574,
+      "grad_norm": 0.18930155038833618,
+      "learning_rate": 4.988974389814463e-05,
+      "loss": 0.0011,
+      "step": 455
+    },
+    {
+      "epoch": 0.3198316675433982,
+      "grad_norm": 0.3088482916355133,
+      "learning_rate": 4.9888379527446885e-05,
+      "loss": 0.0018,
+      "step": 456
+    },
+    {
+      "epoch": 0.320533052779239,
+      "grad_norm": 71.42636108398438,
+      "learning_rate": 4.988700678572329e-05,
+      "loss": 1.9691,
+      "step": 457
+    },
+    {
+      "epoch": 0.3212344380150798,
+      "grad_norm": 0.20487411320209503,
+      "learning_rate": 4.988562567343557e-05,
+      "loss": 0.0012,
+      "step": 458
+    },
+    {
+      "epoch": 0.32193582325092057,
+      "grad_norm": 73.37615203857422,
+      "learning_rate": 4.988423619104823e-05,
+      "loss": 1.8636,
+      "step": 459
+    },
+    {
+      "epoch": 0.32263720848676136,
+      "grad_norm": 66.4114761352539,
+      "learning_rate": 4.988283833902863e-05,
+      "loss": 1.3073,
+      "step": 460
+    },
+    {
+      "epoch": 0.32333859372260215,
+      "grad_norm": 4.2518744468688965,
+      "learning_rate": 4.988143211784693e-05,
+      "loss": 0.0318,
+      "step": 461
+    },
+    {
+      "epoch": 0.32403997895844294,
+      "grad_norm": 4.782449245452881,
+      "learning_rate": 4.9880017527976083e-05,
+      "loss": 0.0384,
+      "step": 462
+    },
+    {
+      "epoch": 0.32474136419428373,
+      "grad_norm": 22.517131805419922,
+      "learning_rate": 4.9878594569891896e-05,
+      "loss": 0.2973,
+      "step": 463
+    },
+    {
+      "epoch": 0.3254427494301245,
+      "grad_norm": 5.63627290725708,
+      "learning_rate": 4.987716324407296e-05,
+      "loss": 0.0496,
+      "step": 464
+    },
+    {
+      "epoch": 0.32614413466596526,
+      "grad_norm": 6.364257335662842,
+      "learning_rate": 4.9875723551000696e-05,
+      "loss": 0.0578,
+      "step": 465
+    },
+    {
+      "epoch": 0.32684551990180605,
+      "grad_norm": 147.82330322265625,
+      "learning_rate": 4.987427549115933e-05,
+      "loss": 1.8188,
+      "step": 466
+    },
+    {
+      "epoch": 0.32754690513764684,
+      "grad_norm": 3.7355575561523438,
+      "learning_rate": 4.987281906503591e-05,
+      "loss": 0.0274,
+      "step": 467
+    },
+    {
+      "epoch": 0.32824829037348763,
+      "grad_norm": 0.6107962131500244,
+      "learning_rate": 4.987135427312029e-05,
+      "loss": 0.0038,
+      "step": 468
+    },
+    {
+      "epoch": 0.3289496756093284,
+      "grad_norm": 61.26199722290039,
+      "learning_rate": 4.9869881115905146e-05,
+      "loss": 0.8915,
+      "step": 469
+    },
+    {
+      "epoch": 0.3296510608451692,
+      "grad_norm": 0.3030995726585388,
+      "learning_rate": 4.986839959388596e-05,
+      "loss": 0.0018,
+      "step": 470
+    },
+    {
+      "epoch": 0.33035244608101,
+      "grad_norm": 0.12271957844495773,
+      "learning_rate": 4.986690970756104e-05,
+      "loss": 0.0007,
+      "step": 471
+    },
+    {
+      "epoch": 0.3310538313168508,
+      "grad_norm": 71.7486801147461,
+      "learning_rate": 4.9865411457431485e-05,
+      "loss": 1.8673,
+      "step": 472
+    },
+    {
+      "epoch": 0.3317552165526916,
+      "grad_norm": 69.17851257324219,
+      "learning_rate": 4.986390484400124e-05,
+      "loss": 1.6587,
+      "step": 473
+    },
+    {
+      "epoch": 0.33245660178853237,
+      "grad_norm": 0.18208004534244537,
+      "learning_rate": 4.9862389867777023e-05,
+      "loss": 0.001,
+      "step": 474
+    },
+    {
+      "epoch": 0.33315798702437316,
+      "grad_norm": 122.65370178222656,
+      "learning_rate": 4.986086652926839e-05,
+      "loss": 0.9068,
+      "step": 475
+    },
+    {
+      "epoch": 0.33385937226021395,
+      "grad_norm": 0.6119470596313477,
+      "learning_rate": 4.9859334828987715e-05,
+      "loss": 0.0039,
+      "step": 476
+    },
+    {
+      "epoch": 0.3345607574960547,
+      "grad_norm": 4.483019828796387,
+      "learning_rate": 4.985779476745017e-05,
+      "loss": 0.0343,
+      "step": 477
+    },
+    {
+      "epoch": 0.3352621427318955,
+      "grad_norm": 0.7592105269432068,
+      "learning_rate": 4.985624634517374e-05,
+      "loss": 0.005,
+      "step": 478
+    },
+    {
+      "epoch": 0.33596352796773626,
+      "grad_norm": 0.6566614508628845,
+      "learning_rate": 4.985468956267922e-05,
+      "loss": 0.0037,
+      "step": 479
+    },
+    {
+      "epoch": 0.33666491320357705,
+      "grad_norm": 67.39823150634766,
+      "learning_rate": 4.985312442049023e-05,
+      "loss": 1.7049,
+      "step": 480
+    },
+    {
+      "epoch": 0.33736629843941784,
+      "grad_norm": 0.13803933560848236,
+      "learning_rate": 4.98515509191332e-05,
+      "loss": 0.0008,
+      "step": 481
+    },
+    {
+      "epoch": 0.33806768367525863,
+      "grad_norm": 72.15523529052734,
+      "learning_rate": 4.984996905913736e-05,
+      "loss": 1.6634,
+      "step": 482
+    },
+    {
+      "epoch": 0.3387690689110994,
+      "grad_norm": 0.10033612698316574,
+      "learning_rate": 4.984837884103475e-05,
+      "loss": 0.0005,
+      "step": 483
+    },
+    {
+      "epoch": 0.3394704541469402,
+      "grad_norm": 62.519317626953125,
+      "learning_rate": 4.984678026536024e-05,
+      "loss": 1.2577,
+      "step": 484
+    },
+    {
+      "epoch": 0.340171839382781,
+      "grad_norm": 0.36229056119918823,
+      "learning_rate": 4.984517333265148e-05,
+      "loss": 0.0023,
+      "step": 485
+    },
+    {
+      "epoch": 0.3408732246186218,
+      "grad_norm": 0.8776547312736511,
+      "learning_rate": 4.9843558043448985e-05,
+      "loss": 0.0056,
+      "step": 486
+    },
+    {
+      "epoch": 0.3415746098544626,
+      "grad_norm": 108.73171997070312,
+      "learning_rate": 4.984193439829602e-05,
+      "loss": 2.7267,
+      "step": 487
+    },
+    {
+      "epoch": 0.3422759950903034,
+      "grad_norm": 0.6344736814498901,
+      "learning_rate": 4.984030239773868e-05,
+      "loss": 0.0041,
+      "step": 488
+    },
+    {
+      "epoch": 0.3429773803261441,
+      "grad_norm": 113.62140655517578,
+      "learning_rate": 4.98386620423259e-05,
+      "loss": 1.7271,
+      "step": 489
+    },
+    {
+      "epoch": 0.3436787655619849,
+      "grad_norm": 1.618384599685669,
+      "learning_rate": 4.983701333260938e-05,
+      "loss": 0.011,
+      "step": 490
+    },
+    {
+      "epoch": 0.3443801507978257,
+      "grad_norm": 138.78854370117188,
+      "learning_rate": 4.9835356269143676e-05,
+      "loss": 1.9143,
+      "step": 491
+    },
+    {
+      "epoch": 0.3450815360336665,
+      "grad_norm": 49.62142562866211,
+      "learning_rate": 4.98336908524861e-05,
+      "loss": 0.753,
+      "step": 492
+    },
+    {
+      "epoch": 0.34578292126950727,
+      "grad_norm": 3.6228299140930176,
+      "learning_rate": 4.9832017083196816e-05,
+      "loss": 0.0277,
+      "step": 493
+    },
+    {
+      "epoch": 0.34648430650534806,
+      "grad_norm": 24.915790557861328,
+      "learning_rate": 4.98303349618388e-05,
+      "loss": 0.356,
+      "step": 494
+    },
+    {
+      "epoch": 0.34718569174118885,
+      "grad_norm": 7.880053997039795,
+      "learning_rate": 4.9828644488977795e-05,
+      "loss": 0.0727,
+      "step": 495
+    },
+    {
+      "epoch": 0.34788707697702964,
+      "grad_norm": 12.95199203491211,
+      "learning_rate": 4.982694566518239e-05,
+      "loss": 0.25,
+      "step": 496
+    },
+    {
+      "epoch": 0.34858846221287043,
+      "grad_norm": 2.059384822845459,
+      "learning_rate": 4.982523849102397e-05,
+      "loss": 0.2069,
+      "step": 497
+    },
+    {
+      "epoch": 0.3492898474487112,
+      "grad_norm": 17.132720947265625,
+      "learning_rate": 4.982352296707675e-05,
+      "loss": 0.2093,
+      "step": 498
+    },
+    {
+      "epoch": 0.349991232684552,
+      "grad_norm": 12.591998100280762,
+      "learning_rate": 4.98217990939177e-05,
+      "loss": 0.1889,
+      "step": 499
+    },
+    {
+      "epoch": 0.3506926179203928,
+      "grad_norm": 12.747396469116211,
+      "learning_rate": 4.982006687212666e-05,
+      "loss": 0.1755,
+      "step": 500
+    },
+    {
+      "epoch": 0.35139400315623354,
+      "grad_norm": 77.32592010498047,
+      "learning_rate": 4.981832630228624e-05,
+      "loss": 0.9517,
+      "step": 501
+    },
+    {
+      "epoch": 0.3520953883920743,
+      "grad_norm": 8.452153205871582,
+      "learning_rate": 4.9816577384981866e-05,
+      "loss": 0.0913,
+      "step": 502
+    },
+    {
+      "epoch": 0.3527967736279151,
+      "grad_norm": 2.1267127990722656,
+      "learning_rate": 4.9814820120801776e-05,
+      "loss": 0.0178,
+      "step": 503
+    },
+    {
+      "epoch": 0.3534981588637559,
+      "grad_norm": 0.8161488771438599,
+      "learning_rate": 4.981305451033701e-05,
+      "loss": 0.0058,
+      "step": 504
+    },
+    {
+      "epoch": 0.3541995440995967,
+      "grad_norm": 0.10489612072706223,
+      "learning_rate": 4.981128055418143e-05,
+      "loss": 0.0006,
+      "step": 505
+    },
+    {
+      "epoch": 0.3549009293354375,
+      "grad_norm": 69.51275634765625,
+      "learning_rate": 4.9809498252931674e-05,
+      "loss": 2.62,
+      "step": 506
+    },
+    {
+      "epoch": 0.3556023145712783,
+      "grad_norm": 0.004451240878552198,
+      "learning_rate": 4.980770760718721e-05,
+      "loss": 0.0,
+      "step": 507
+    },
+    {
+      "epoch": 0.35630369980711907,
+      "grad_norm": 67.6517105102539,
+      "learning_rate": 4.9805908617550335e-05,
+      "loss": 3.25,
+      "step": 508
+    },
+    {
+      "epoch": 0.35700508504295986,
+      "grad_norm": 0.016789492219686508,
+      "learning_rate": 4.980410128462609e-05,
+      "loss": 0.0001,
+      "step": 509
+    },
+    {
+      "epoch": 0.35770647027880065,
+      "grad_norm": 0.025487884879112244,
+      "learning_rate": 4.980228560902239e-05,
+      "loss": 0.0001,
+      "step": 510
+    },
+    {
+      "epoch": 0.35840785551464144,
+      "grad_norm": 0.042408283799886703,
+      "learning_rate": 4.98004615913499e-05,
+      "loss": 0.0002,
+      "step": 511
+    },
+    {
+      "epoch": 0.35910924075048223,
+      "grad_norm": 0.038653161376714706,
+      "learning_rate": 4.979862923222212e-05,
+      "loss": 0.0002,
+      "step": 512
+    },
+    {
+      "epoch": 0.35981062598632296,
+      "grad_norm": 0.03510050103068352,
+      "learning_rate": 4.9796788532255356e-05,
+      "loss": 0.0002,
+      "step": 513
+    },
+    {
+      "epoch": 0.36051201122216375,
+      "grad_norm": 0.008441315963864326,
+      "learning_rate": 4.979493949206872e-05,
+      "loss": 0.0,
+      "step": 514
+    },
+    {
+      "epoch": 0.36121339645800454,
+      "grad_norm": 0.006373463664203882,
+      "learning_rate": 4.979308211228411e-05,
+      "loss": 0.0,
+      "step": 515
+    },
+    {
+      "epoch": 0.36191478169384533,
+      "grad_norm": 79.59523010253906,
+      "learning_rate": 4.979121639352625e-05,
+      "loss": 2.9487,
+      "step": 516
+    },
+    {
+      "epoch": 0.3626161669296861,
+      "grad_norm": 153.0602264404297,
+      "learning_rate": 4.9789342336422665e-05,
+      "loss": 3.6533,
+      "step": 517
+    },
+    {
+      "epoch": 0.3633175521655269,
+      "grad_norm": 0.7085576057434082,
+      "learning_rate": 4.9787459941603666e-05,
+      "loss": 0.0042,
+      "step": 518
+    },
+    {
+      "epoch": 0.3640189374013677,
+      "grad_norm": 155.89154052734375,
+      "learning_rate": 4.9785569209702395e-05,
+      "loss": 7.9018,
+      "step": 519
+    },
+    {
+      "epoch": 0.3647203226372085,
+      "grad_norm": 5.253132343292236,
+      "learning_rate": 4.9783670141354776e-05,
+      "loss": 0.042,
+      "step": 520
+    },
+    {
+      "epoch": 0.3654217078730493,
+      "grad_norm": 26.310007095336914,
+      "learning_rate": 4.978176273719957e-05,
+      "loss": 0.2762,
+      "step": 521
+    },
+    {
+      "epoch": 0.3661230931088901,
+      "grad_norm": 9.570679664611816,
+      "learning_rate": 4.977984699787829e-05,
+      "loss": 0.0898,
+      "step": 522
+    },
+    {
+      "epoch": 0.36682447834473086,
+      "grad_norm": 214.85060119628906,
+      "learning_rate": 4.977792292403529e-05,
+      "loss": 6.7239,
+      "step": 523
+    },
+    {
+      "epoch": 0.36752586358057165,
+      "grad_norm": 2.7513771057128906,
+      "learning_rate": 4.977599051631773e-05,
+      "loss": 0.0228,
+      "step": 524
+    },
+    {
+      "epoch": 0.3682272488164124,
+      "grad_norm": 33.629356384277344,
+      "learning_rate": 4.977404977537554e-05,
+      "loss": 0.4374,
+      "step": 525
+    },
+    {
+      "epoch": 0.3689286340522532,
+      "grad_norm": 3.8864808082580566,
+      "learning_rate": 4.97721007018615e-05,
+      "loss": 0.0308,
+      "step": 526
+    },
+    {
+      "epoch": 0.36963001928809397,
+      "grad_norm": 1.0654228925704956,
+      "learning_rate": 4.9770143296431144e-05,
+      "loss": 0.0077,
+      "step": 527
+    },
+    {
+      "epoch": 0.37033140452393476,
+      "grad_norm": 131.74990844726562,
+      "learning_rate": 4.9768177559742844e-05,
+      "loss": 4.0998,
+      "step": 528
+    },
+    {
+      "epoch": 0.37103278975977555,
+      "grad_norm": 0.10420957952737808,
+      "learning_rate": 4.976620349245776e-05,
+      "loss": 0.0006,
+      "step": 529
+    },
+    {
+      "epoch": 0.37173417499561634,
+      "grad_norm": 72.78388977050781,
+      "learning_rate": 4.976422109523985e-05,
+      "loss": 3.8242,
+      "step": 530
+    },
+    {
+      "epoch": 0.37243556023145713,
+      "grad_norm": 69.0342025756836,
+      "learning_rate": 4.976223036875588e-05,
+      "loss": 1.9872,
+      "step": 531
+    },
+    {
+      "epoch": 0.3731369454672979,
+      "grad_norm": 0.43323075771331787,
+      "learning_rate": 4.9760231313675426e-05,
+      "loss": 0.0028,
+      "step": 532
+    },
+    {
+      "epoch": 0.3738383307031387,
+      "grad_norm": 1.0747731924057007,
+      "learning_rate": 4.975822393067084e-05,
+      "loss": 0.0061,
+      "step": 533
+    },
+    {
+      "epoch": 0.3745397159389795,
+      "grad_norm": 69.71436309814453,
+      "learning_rate": 4.9756208220417304e-05,
+      "loss": 1.8796,
+      "step": 534
+    },
+    {
+      "epoch": 0.3752411011748203,
+      "grad_norm": 127.05543518066406,
+      "learning_rate": 4.975418418359278e-05,
+      "loss": 2.8441,
+      "step": 535
+    },
+    {
+      "epoch": 0.3759424864106611,
+      "grad_norm": 0.9959923624992371,
+      "learning_rate": 4.975215182087804e-05,
+      "loss": 0.0065,
+      "step": 536
+    },
+    {
+      "epoch": 0.3766438716465018,
+      "grad_norm": 2.1746952533721924,
+      "learning_rate": 4.9750111132956654e-05,
+      "loss": 0.0156,
+      "step": 537
+    },
+    {
+      "epoch": 0.3773452568823426,
+      "grad_norm": 92.57381439208984,
+      "learning_rate": 4.9748062120514996e-05,
+      "loss": 1.4085,
+      "step": 538
+    },
+    {
+      "epoch": 0.3780466421181834,
+      "grad_norm": 75.45511627197266,
+      "learning_rate": 4.974600478424223e-05,
+      "loss": 0.9909,
+      "step": 539
+    },
+    {
+      "epoch": 0.3787480273540242,
+      "grad_norm": 25.877460479736328,
+      "learning_rate": 4.974393912483034e-05,
+      "loss": 0.3578,
+      "step": 540
+    },
+    {
+      "epoch": 0.379449412589865,
+      "grad_norm": 12.262901306152344,
+      "learning_rate": 4.9741865142974076e-05,
+      "loss": 0.1338,
+      "step": 541
+    },
+    {
+      "epoch": 0.38015079782570577,
+      "grad_norm": 10.54548168182373,
+      "learning_rate": 4.973978283937102e-05,
+      "loss": 0.2111,
+      "step": 542
+    },
+    {
+      "epoch": 0.38085218306154656,
+      "grad_norm": 35.909976959228516,
+      "learning_rate": 4.973769221472153e-05,
+      "loss": 0.5549,
+      "step": 543
+    },
+    {
+      "epoch": 0.38155356829738735,
+      "grad_norm": 13.012696266174316,
+      "learning_rate": 4.973559326972878e-05,
+      "loss": 0.1784,
+      "step": 544
+    },
+    {
+      "epoch": 0.38225495353322814,
+      "grad_norm": 16.022974014282227,
+      "learning_rate": 4.973348600509874e-05,
+      "loss": 0.2235,
+      "step": 545
+    },
+    {
+      "epoch": 0.3829563387690689,
+      "grad_norm": 22.86570167541504,
+      "learning_rate": 4.973137042154016e-05,
+      "loss": 0.3561,
+      "step": 546
+    },
+    {
+      "epoch": 0.3836577240049097,
+      "grad_norm": 13.543822288513184,
+      "learning_rate": 4.972924651976461e-05,
+      "loss": 0.3077,
+      "step": 547
+    },
+    {
+      "epoch": 0.3843591092407505,
+      "grad_norm": 17.42881965637207,
+      "learning_rate": 4.972711430048644e-05,
+      "loss": 0.2779,
+      "step": 548
+    },
+    {
+      "epoch": 0.38506049447659124,
+      "grad_norm": 19.7313175201416,
+      "learning_rate": 4.972497376442282e-05,
+      "loss": 0.2582,
+      "step": 549
+    },
+    {
+      "epoch": 0.38576187971243203,
+      "grad_norm": 12.354894638061523,
+      "learning_rate": 4.9722824912293695e-05,
+      "loss": 0.2374,
+      "step": 550
+    },
+    {
+      "epoch": 0.3864632649482728,
+      "grad_norm": 17.16146469116211,
+      "learning_rate": 4.972066774482182e-05,
+      "loss": 0.2085,
+      "step": 551
+    },
+    {
+      "epoch": 0.3871646501841136,
+      "grad_norm": 49.288734436035156,
+      "learning_rate": 4.971850226273274e-05,
+      "loss": 0.6211,
+      "step": 552
+    },
+    {
+      "epoch": 0.3878660354199544,
+      "grad_norm": 4.870689392089844,
+      "learning_rate": 4.97163284667548e-05,
+      "loss": 0.0413,
+      "step": 553
+    },
+    {
+      "epoch": 0.3885674206557952,
+      "grad_norm": 4.268405437469482,
+      "learning_rate": 4.971414635761914e-05,
+      "loss": 0.0354,
+      "step": 554
+    },
+    {
+      "epoch": 0.389268805891636,
+      "grad_norm": 2.800025701522827,
+      "learning_rate": 4.971195593605969e-05,
+      "loss": 0.0215,
+      "step": 555
+    },
+    {
+      "epoch": 0.3899701911274768,
+      "grad_norm": 136.0956573486328,
+      "learning_rate": 4.9709757202813195e-05,
+      "loss": 3.3931,
+      "step": 556
+    },
+    {
+      "epoch": 0.39067157636331756,
+      "grad_norm": 68.56623840332031,
+      "learning_rate": 4.970755015861918e-05,
+      "loss": 1.8678,
+      "step": 557
+    },
+    {
+      "epoch": 0.39137296159915835,
+      "grad_norm": 68.25215911865234,
+      "learning_rate": 4.970533480421996e-05,
+      "loss": 1.9984,
+      "step": 558
+    },
+    {
+      "epoch": 0.39207434683499914,
+      "grad_norm": 69.02233123779297,
+      "learning_rate": 4.9703111140360656e-05,
+      "loss": 2.0662,
+      "step": 559
+    },
+    {
+      "epoch": 0.39277573207083993,
+      "grad_norm": 61.60824203491211,
+      "learning_rate": 4.970087916778918e-05,
+      "loss": 1.5281,
+      "step": 560
+    },
+    {
+      "epoch": 0.39347711730668067,
+      "grad_norm": 48.50178146362305,
+      "learning_rate": 4.9698638887256254e-05,
+      "loss": 0.8115,
+      "step": 561
+    },
+    {
+      "epoch": 0.39417850254252146,
+      "grad_norm": 41.32649612426758,
+      "learning_rate": 4.969639029951537e-05,
+      "loss": 0.6374,
+      "step": 562
+    },
+    {
+      "epoch": 0.39487988777836225,
+      "grad_norm": 6.946910858154297,
+      "learning_rate": 4.9694133405322815e-05,
+      "loss": 0.2402,
+      "step": 563
+    },
+    {
+      "epoch": 0.39558127301420304,
+      "grad_norm": 22.8358097076416,
+      "learning_rate": 4.9691868205437694e-05,
+      "loss": 0.3883,
+      "step": 564
+    },
+    {
+      "epoch": 0.39628265825004383,
+      "grad_norm": 15.402759552001953,
+      "learning_rate": 4.968959470062189e-05,
+      "loss": 0.7708,
+      "step": 565
+    },
+    {
+      "epoch": 0.3969840434858846,
+      "grad_norm": 28.08781623840332,
+      "learning_rate": 4.968731289164006e-05,
+      "loss": 0.7703,
+      "step": 566
+    },
+    {
+      "epoch": 0.3976854287217254,
+      "grad_norm": 26.910886764526367,
+      "learning_rate": 4.9685022779259696e-05,
+      "loss": 0.6299,
+      "step": 567
+    },
+    {
+      "epoch": 0.3983868139575662,
+      "grad_norm": 24.341957092285156,
+      "learning_rate": 4.968272436425105e-05,
+      "loss": 0.5751,
+      "step": 568
+    },
+    {
+      "epoch": 0.399088199193407,
+      "grad_norm": 23.469263076782227,
+      "learning_rate": 4.9680417647387186e-05,
+      "loss": 0.4664,
+      "step": 569
+    },
+    {
+      "epoch": 0.3997895844292478,
+      "grad_norm": 15.347972869873047,
+      "learning_rate": 4.967810262944394e-05,
+      "loss": 0.2252,
+      "step": 570
+    },
+    {
+      "epoch": 0.40049096966508857,
+      "grad_norm": 5.6460280418396,
+      "learning_rate": 4.967577931119996e-05,
+      "loss": 0.2232,
+      "step": 571
+    },
+    {
+      "epoch": 0.40119235490092936,
+      "grad_norm": 4.2383856773376465,
+      "learning_rate": 4.967344769343667e-05,
+      "loss": 0.0396,
+      "step": 572
+    },
+    {
+      "epoch": 0.4018937401367701,
+      "grad_norm": 1.6932095289230347,
+      "learning_rate": 4.96711077769383e-05,
+      "loss": 0.0137,
+      "step": 573
+    },
+    {
+      "epoch": 0.4025951253726109,
+      "grad_norm": 0.41152113676071167,
+      "learning_rate": 4.966875956249186e-05,
+      "loss": 0.0029,
+      "step": 574
+    },
+    {
+      "epoch": 0.4032965106084517,
+      "grad_norm": 0.06895897537469864,
+      "learning_rate": 4.9666403050887155e-05,
+      "loss": 0.0004,
+      "step": 575
+    },
+    {
+      "epoch": 0.40399789584429247,
+      "grad_norm": 0.004564318805932999,
+      "learning_rate": 4.9664038242916786e-05,
+      "loss": 0.0,
+      "step": 576
+    },
+    {
+      "epoch": 0.40469928108013326,
+      "grad_norm": 0.0019707169849425554,
+      "learning_rate": 4.966166513937613e-05,
+      "loss": 0.0,
+      "step": 577
+    },
+    {
+      "epoch": 0.40540066631597405,
+      "grad_norm": 0.00037522497586905956,
+      "learning_rate": 4.965928374106337e-05,
+      "loss": 0.0,
+      "step": 578
+    },
+    {
+      "epoch": 0.40610205155181484,
+      "grad_norm": 5.6252967624459416e-05,
+      "learning_rate": 4.965689404877946e-05,
+      "loss": 0.0,
+      "step": 579
+    },
+    {
+      "epoch": 0.4068034367876556,
+      "grad_norm": 1.1128309779451229e-05,
+      "learning_rate": 4.9654496063328174e-05,
+      "loss": 0.0,
+      "step": 580
+    },
+    {
+      "epoch": 0.4075048220234964,
+      "grad_norm": 119.89021301269531,
+      "learning_rate": 4.965208978551605e-05,
+      "loss": 10.8344,
+      "step": 581
+    },
+    {
+      "epoch": 0.4082062072593372,
+      "grad_norm": 1.3851165931555443e-05,
+      "learning_rate": 4.964967521615241e-05,
+      "loss": 0.0,
+      "step": 582
+    },
+    {
+      "epoch": 0.408907592495178,
+      "grad_norm": 3.7249556044116616e-06,
+      "learning_rate": 4.96472523560494e-05,
+      "loss": 0.0,
+      "step": 583
+    },
+    {
+      "epoch": 0.4096089777310188,
+      "grad_norm": 2.917176971095614e-06,
+      "learning_rate": 4.9644821206021906e-05,
+      "loss": 0.0,
+      "step": 584
+    },
+    {
+      "epoch": 0.4103103629668595,
+      "grad_norm": 3.3871406230900902e-06,
+      "learning_rate": 4.964238176688765e-05,
+      "loss": 0.0,
+      "step": 585
+    },
+    {
+      "epoch": 0.4110117482027003,
+      "grad_norm": 2.260056135128252e-06,
+      "learning_rate": 4.9639934039467106e-05,
+      "loss": 0.0,
+      "step": 586
+    },
+    {
+      "epoch": 0.4117131334385411,
+      "grad_norm": 1.5038928040667088e-06,
+      "learning_rate": 4.963747802458355e-05,
+      "loss": 0.0,
+      "step": 587
+    },
+    {
+      "epoch": 0.4124145186743819,
+      "grad_norm": 59.293521881103516,
+      "learning_rate": 4.963501372306306e-05,
+      "loss": 5.3994,
+      "step": 588
+    },
+    {
+      "epoch": 0.4131159039102227,
+      "grad_norm": 59.763816833496094,
+      "learning_rate": 4.963254113573446e-05,
+      "loss": 5.6505,
+      "step": 589
+    },
+    {
+      "epoch": 0.41381728914606347,
+      "grad_norm": 132.7192840576172,
+      "learning_rate": 4.963006026342941e-05,
+      "loss": 9.1499,
+      "step": 590
+    },
+    {
+      "epoch": 0.41451867438190426,
+      "grad_norm": 0.00024285895051434636,
+      "learning_rate": 4.962757110698232e-05,
+      "loss": 0.0,
+      "step": 591
+    },
+    {
+      "epoch": 0.41522005961774505,
+      "grad_norm": 0.0009618580807000399,
+      "learning_rate": 4.9625073667230404e-05,
+      "loss": 0.0,
+      "step": 592
+    },
+    {
+      "epoch": 0.41592144485358584,
+      "grad_norm": 0.004245178773999214,
+      "learning_rate": 4.962256794501365e-05,
+      "loss": 0.0,
+      "step": 593
+    },
+    {
+      "epoch": 0.41662283008942663,
+      "grad_norm": 0.01505737192928791,
+      "learning_rate": 4.962005394117486e-05,
+      "loss": 0.0001,
+      "step": 594
+    },
+    {
+      "epoch": 0.4173242153252674,
+      "grad_norm": 0.01870640181005001,
+      "learning_rate": 4.961753165655958e-05,
+      "loss": 0.0001,
+      "step": 595
+    },
+    {
+      "epoch": 0.4180256005611082,
+      "grad_norm": 0.047609515488147736,
+      "learning_rate": 4.961500109201618e-05,
+      "loss": 0.0003,
+      "step": 596
+    },
+    {
+      "epoch": 0.41872698579694895,
+      "grad_norm": 0.15332278609275818,
+      "learning_rate": 4.9612462248395775e-05,
+      "loss": 0.001,
+      "step": 597
+    },
+    {
+      "epoch": 0.41942837103278974,
+      "grad_norm": 113.53760528564453,
+      "learning_rate": 4.96099151265523e-05,
+      "loss": 2.3303,
+      "step": 598
+    },
+    {
+      "epoch": 0.4201297562686305,
+      "grad_norm": 0.6282661557197571,
+      "learning_rate": 4.9607359727342465e-05,
+      "loss": 0.0044,
+      "step": 599
+    },
+    {
+      "epoch": 0.4208311415044713,
+      "grad_norm": 43.630615234375,
+      "learning_rate": 4.9604796051625756e-05,
+      "loss": 0.6733,
+      "step": 600
+    },
+    {
+      "epoch": 0.4215325267403121,
+      "grad_norm": 72.0724868774414,
+      "learning_rate": 4.960222410026444e-05,
+      "loss": 1.4145,
+      "step": 601
+    },
+    {
+      "epoch": 0.4222339119761529,
+      "grad_norm": 6.166507244110107,
+      "learning_rate": 4.959964387412358e-05,
+      "loss": 0.0608,
+      "step": 602
+    },
+    {
+      "epoch": 0.4229352972119937,
+      "grad_norm": 80.03913879394531,
+      "learning_rate": 4.9597055374071024e-05,
+      "loss": 1.2555,
+      "step": 603
+    },
+    {
+      "epoch": 0.4236366824478345,
+      "grad_norm": 17.456140518188477,
+      "learning_rate": 4.959445860097738e-05,
+      "loss": 0.238,
+      "step": 604
+    },
+    {
+      "epoch": 0.42433806768367527,
+      "grad_norm": 36.63374328613281,
+      "learning_rate": 4.9591853555716064e-05,
+      "loss": 0.8324,
+      "step": 605
+    },
+    {
+      "epoch": 0.42503945291951606,
+      "grad_norm": 16.27304458618164,
+      "learning_rate": 4.9589240239163255e-05,
+      "loss": 0.2125,
+      "step": 606
+    },
+    {
+      "epoch": 0.42574083815535685,
+      "grad_norm": 19.427776336669922,
+      "learning_rate": 4.9586618652197934e-05,
+      "loss": 0.2701,
+      "step": 607
+    },
+    {
+      "epoch": 0.42644222339119764,
+      "grad_norm": 8.516806602478027,
+      "learning_rate": 4.958398879570185e-05,
+      "loss": 0.3408,
+      "step": 608
+    },
+    {
+      "epoch": 0.4271436086270384,
+      "grad_norm": 1.857074499130249,
+      "learning_rate": 4.9581350670559524e-05,
+      "loss": 0.2039,
+      "step": 609
+    },
+    {
+      "epoch": 0.42784499386287916,
+      "grad_norm": 8.925201416015625,
+      "learning_rate": 4.9578704277658286e-05,
+      "loss": 0.09,
+      "step": 610
+    },
+    {
+      "epoch": 0.42854637909871995,
+      "grad_norm": 4.490631580352783,
+      "learning_rate": 4.957604961788823e-05,
+      "loss": 0.0384,
+      "step": 611
+    },
+    {
+      "epoch": 0.42924776433456074,
+      "grad_norm": 0.866288959980011,
+      "learning_rate": 4.957338669214222e-05,
+      "loss": 0.0063,
+      "step": 612
+    },
+    {
+      "epoch": 0.42994914957040153,
+      "grad_norm": 0.6116040349006653,
+      "learning_rate": 4.9570715501315925e-05,
+      "loss": 0.0035,
+      "step": 613
+    },
+    {
+      "epoch": 0.4306505348062423,
+      "grad_norm": 0.014350208453834057,
+      "learning_rate": 4.956803604630777e-05,
+      "loss": 0.0001,
+      "step": 614
+    },
+    {
+      "epoch": 0.4313519200420831,
+      "grad_norm": 0.004978867247700691,
+      "learning_rate": 4.956534832801897e-05,
+      "loss": 0.0,
+      "step": 615
+    },
+    {
+      "epoch": 0.4320533052779239,
+      "grad_norm": 0.0008859955123625696,
+      "learning_rate": 4.9562652347353535e-05,
+      "loss": 0.0,
+      "step": 616
+    },
+    {
+      "epoch": 0.4327546905137647,
+      "grad_norm": 0.0007402606424875557,
+      "learning_rate": 4.955994810521822e-05,
+      "loss": 0.0,
+      "step": 617
+    },
+    {
+      "epoch": 0.4334560757496055,
+      "grad_norm": 60.98609161376953,
+      "learning_rate": 4.955723560252259e-05,
+      "loss": 4.6234,
+      "step": 618
+    },
+    {
+      "epoch": 0.4341574609854463,
+      "grad_norm": 60.63393783569336,
+      "learning_rate": 4.955451484017896e-05,
+      "loss": 4.669,
+      "step": 619
+    },
+    {
+      "epoch": 0.43485884622128707,
+      "grad_norm": 0.00016736971156205982,
+      "learning_rate": 4.955178581910246e-05,
+      "loss": 0.0,
+      "step": 620
+    },
+    {
+      "epoch": 0.4355602314571278,
+      "grad_norm": 0.00026081904070451856,
+      "learning_rate": 4.9549048540210963e-05,
+      "loss": 0.0,
+      "step": 621
+    },
+    {
+      "epoch": 0.4362616166929686,
+      "grad_norm": 0.0002771130239125341,
+      "learning_rate": 4.954630300442513e-05,
+      "loss": 0.0,
+      "step": 622
+    },
+    {
+      "epoch": 0.4369630019288094,
+      "grad_norm": 0.00024890206987038255,
+      "learning_rate": 4.954354921266841e-05,
+      "loss": 0.0,
+      "step": 623
+    },
+    {
+      "epoch": 0.43766438716465017,
+      "grad_norm": 0.0003540183824952692,
+      "learning_rate": 4.954078716586702e-05,
+      "loss": 0.0,
+      "step": 624
+    },
+    {
+      "epoch": 0.43836577240049096,
+      "grad_norm": 61.1258659362793,
+      "learning_rate": 4.9538016864949953e-05,
+      "loss": 4.1063,
+      "step": 625
+    },
+    {
+      "epoch": 0.43906715763633175,
+      "grad_norm": 61.82120895385742,
+      "learning_rate": 4.953523831084897e-05,
+      "loss": 3.2575,
+      "step": 626
+    },
+    {
+      "epoch": 0.43976854287217254,
+      "grad_norm": 0.009341489523649216,
+      "learning_rate": 4.9532451504498644e-05,
+      "loss": 0.0001,
+      "step": 627
+    },
+    {
+      "epoch": 0.44046992810801333,
+      "grad_norm": 126.25260925292969,
+      "learning_rate": 4.952965644683627e-05,
+      "loss": 4.8894,
+      "step": 628
+    },
+    {
+      "epoch": 0.4411713133438541,
+      "grad_norm": 64.0586166381836,
+      "learning_rate": 4.9526853138801954e-05,
+      "loss": 2.0939,
+      "step": 629
+    },
+    {
+      "epoch": 0.4418726985796949,
+      "grad_norm": 66.46090698242188,
+      "learning_rate": 4.952404158133857e-05,
+      "loss": 2.1672,
+      "step": 630
+    },
+    {
+      "epoch": 0.4425740838155357,
+      "grad_norm": 8.158260345458984,
+      "learning_rate": 4.952122177539177e-05,
+      "loss": 0.089,
+      "step": 631
+    },
+    {
+      "epoch": 0.4432754690513765,
+      "grad_norm": 8.28303337097168,
+      "learning_rate": 4.951839372190996e-05,
+      "loss": 0.1792,
+      "step": 632
+    },
+    {
+      "epoch": 0.4439768542872172,
+      "grad_norm": 6.670780658721924,
+      "learning_rate": 4.951555742184435e-05,
+      "loss": 0.3495,
+      "step": 633
+    },
+    {
+      "epoch": 0.444678239523058,
+      "grad_norm": 18.7882022857666,
+      "learning_rate": 4.95127128761489e-05,
+      "loss": 0.2761,
+      "step": 634
+    },
+    {
+      "epoch": 0.4453796247588988,
+      "grad_norm": 3.4713053703308105,
+      "learning_rate": 4.950986008578037e-05,
+      "loss": 0.2429,
+      "step": 635
+    },
+    {
+      "epoch": 0.4460810099947396,
+      "grad_norm": 8.415278434753418,
+      "learning_rate": 4.950699905169825e-05,
+      "loss": 0.0883,
+      "step": 636
+    },
+    {
+      "epoch": 0.4467823952305804,
+      "grad_norm": 5.512305736541748,
+      "learning_rate": 4.950412977486484e-05,
+      "loss": 0.0515,
+      "step": 637
+    },
+    {
+      "epoch": 0.4474837804664212,
+      "grad_norm": 1.983485460281372,
+      "learning_rate": 4.95012522562452e-05,
+      "loss": 0.0162,
+      "step": 638
+    },
+    {
+      "epoch": 0.44818516570226197,
+      "grad_norm": 0.5145743489265442,
+      "learning_rate": 4.949836649680717e-05,
+      "loss": 0.0038,
+      "step": 639
+    },
+    {
+      "epoch": 0.44888655093810276,
+      "grad_norm": 0.0464518703520298,
+      "learning_rate": 4.949547249752134e-05,
+      "loss": 0.0003,
+      "step": 640
+    },
+    {
+      "epoch": 0.44958793617394355,
+      "grad_norm": 63.47880935668945,
+      "learning_rate": 4.9492570259361094e-05,
+      "loss": 2.606,
+      "step": 641
+    },
+    {
+      "epoch": 0.45028932140978434,
+      "grad_norm": 63.044219970703125,
+      "learning_rate": 4.948965978330258e-05,
+      "loss": 3.3714,
+      "step": 642
+    },
+    {
+      "epoch": 0.45099070664562513,
+      "grad_norm": 61.190128326416016,
+      "learning_rate": 4.948674107032471e-05,
+      "loss": 3.3129,
+      "step": 643
+    },
+    {
+      "epoch": 0.4516920918814659,
+      "grad_norm": 59.03745651245117,
+      "learning_rate": 4.948381412140917e-05,
+      "loss": 2.3587,
+      "step": 644
+    },
+    {
+      "epoch": 0.4523934771173067,
+      "grad_norm": 61.125858306884766,
+      "learning_rate": 4.948087893754043e-05,
+      "loss": 2.7351,
+      "step": 645
+    },
+    {
+      "epoch": 0.45309486235314744,
+      "grad_norm": 60.69954299926758,
+      "learning_rate": 4.947793551970571e-05,
+      "loss": 2.2391,
+      "step": 646
+    },
+    {
+      "epoch": 0.45379624758898823,
+      "grad_norm": 56.91267395019531,
+      "learning_rate": 4.9474983868895e-05,
+      "loss": 1.5786,
+      "step": 647
+    },
+    {
+      "epoch": 0.454497632824829,
+      "grad_norm": 0.8762818574905396,
+      "learning_rate": 4.947202398610108e-05,
+      "loss": 0.0067,
+      "step": 648
+    },
+    {
+      "epoch": 0.4551990180606698,
+      "grad_norm": 2.4974522590637207,
+      "learning_rate": 4.9469055872319483e-05,
+      "loss": 0.0224,
+      "step": 649
+    },
+    {
+      "epoch": 0.4559004032965106,
+      "grad_norm": 22.98019790649414,
+      "learning_rate": 4.9466079528548505e-05,
+      "loss": 0.3665,
+      "step": 650
+    },
+    {
+      "epoch": 0.4566017885323514,
+      "grad_norm": 16.03438949584961,
+      "learning_rate": 4.9463094955789214e-05,
+      "loss": 0.315,
+      "step": 651
+    },
+    {
+      "epoch": 0.4573031737681922,
+      "grad_norm": 8.860062599182129,
+      "learning_rate": 4.946010215504546e-05,
+      "loss": 0.1067,
+      "step": 652
+    },
+    {
+      "epoch": 0.458004559004033,
+      "grad_norm": 5.521644115447998,
+      "learning_rate": 4.945710112732386e-05,
+      "loss": 0.2473,
+      "step": 653
+    },
+    {
+      "epoch": 0.45870594423987376,
+      "grad_norm": 15.16481876373291,
+      "learning_rate": 4.945409187363376e-05,
+      "loss": 0.2257,
+      "step": 654
+    },
+    {
+      "epoch": 0.45940732947571455,
+      "grad_norm": 15.714698791503906,
+      "learning_rate": 4.945107439498732e-05,
+      "loss": 0.2359,
+      "step": 655
+    },
+    {
+      "epoch": 0.46010871471155534,
+      "grad_norm": 6.4135589599609375,
+      "learning_rate": 4.944804869239944e-05,
+      "loss": 0.2336,
+      "step": 656
+    },
+    {
+      "epoch": 0.46081009994739613,
+      "grad_norm": 9.600846290588379,
+      "learning_rate": 4.94450147668878e-05,
+      "loss": 0.1174,
+      "step": 657
+    },
+    {
+      "epoch": 0.46151148518323687,
+      "grad_norm": 7.477170467376709,
+      "learning_rate": 4.944197261947283e-05,
+      "loss": 0.0857,
+      "step": 658
+    },
+    {
+      "epoch": 0.46221287041907766,
+      "grad_norm": 2.0777835845947266,
+      "learning_rate": 4.943892225117774e-05,
+      "loss": 0.0196,
+      "step": 659
+    },
+    {
+      "epoch": 0.46291425565491845,
+      "grad_norm": 41.072837829589844,
+      "learning_rate": 4.9435863663028495e-05,
+      "loss": 0.7781,
+      "step": 660
+    },
+    {
+      "epoch": 0.46361564089075924,
+      "grad_norm": 0.5752171277999878,
+      "learning_rate": 4.943279685605383e-05,
+      "loss": 0.0045,
+      "step": 661
+    },
+    {
+      "epoch": 0.46431702612660003,
+      "grad_norm": 51.6488151550293,
+      "learning_rate": 4.9429721831285246e-05,
+      "loss": 1.1945,
+      "step": 662
+    },
+    {
+      "epoch": 0.4650184113624408,
+      "grad_norm": 0.11998272687196732,
+      "learning_rate": 4.942663858975701e-05,
+      "loss": 0.0009,
+      "step": 663
+    },
+    {
+      "epoch": 0.4657197965982816,
+      "grad_norm": 57.40029525756836,
+      "learning_rate": 4.9423547132506135e-05,
+      "loss": 1.7823,
+      "step": 664
+    },
+    {
+      "epoch": 0.4664211818341224,
+      "grad_norm": 0.07954677194356918,
+      "learning_rate": 4.942044746057242e-05,
+      "loss": 0.0006,
+      "step": 665
+    },
+    {
+      "epoch": 0.4671225670699632,
+      "grad_norm": 77.24893951416016,
+      "learning_rate": 4.941733957499841e-05,
+      "loss": 3.3749,
+      "step": 666
+    },
+    {
+      "epoch": 0.467823952305804,
+      "grad_norm": 57.04423522949219,
+      "learning_rate": 4.9414223476829434e-05,
+      "loss": 1.6771,
+      "step": 667
+    },
+    {
+      "epoch": 0.46852533754164477,
+      "grad_norm": 0.14917835593223572,
+      "learning_rate": 4.941109916711355e-05,
+      "loss": 0.0011,
+      "step": 668
+    },
+    {
+      "epoch": 0.46922672277748556,
+      "grad_norm": 47.98805236816406,
+      "learning_rate": 4.940796664690161e-05,
+      "loss": 1.0673,
+      "step": 669
+    },
+    {
+      "epoch": 0.4699281080133263,
+      "grad_norm": 63.10405349731445,
+      "learning_rate": 4.940482591724721e-05,
+      "loss": 1.7249,
+      "step": 670
+    },
+    {
+      "epoch": 0.4706294932491671,
+      "grad_norm": 47.01327133178711,
+      "learning_rate": 4.940167697920671e-05,
+      "loss": 0.9642,
+      "step": 671
+    },
+    {
+      "epoch": 0.4713308784850079,
+      "grad_norm": 4.804792881011963,
+      "learning_rate": 4.9398519833839234e-05,
+      "loss": 0.0514,
+      "step": 672
+    },
+    {
+      "epoch": 0.47203226372084867,
+      "grad_norm": 46.21573257446289,
+      "learning_rate": 4.939535448220667e-05,
+      "loss": 0.7498,
+      "step": 673
+    },
+    {
+      "epoch": 0.47273364895668946,
+      "grad_norm": 12.326519966125488,
+      "learning_rate": 4.939218092537365e-05,
+      "loss": 0.1734,
+      "step": 674
+    },
+    {
+      "epoch": 0.47343503419253025,
+      "grad_norm": 2.402872085571289,
+      "learning_rate": 4.938899916440759e-05,
+      "loss": 0.2545,
+      "step": 675
+    },
+    {
+      "epoch": 0.47413641942837104,
+      "grad_norm": 4.058804512023926,
+      "learning_rate": 4.938580920037864e-05,
+      "loss": 0.2573,
+      "step": 676
+    },
+    {
+      "epoch": 0.4748378046642118,
+      "grad_norm": 11.151541709899902,
+      "learning_rate": 4.938261103435973e-05,
+      "loss": 0.4285,
+      "step": 677
+    },
+    {
+      "epoch": 0.4755391899000526,
+      "grad_norm": 6.957059860229492,
+      "learning_rate": 4.937940466742653e-05,
+      "loss": 0.2347,
+      "step": 678
+    },
+    {
+      "epoch": 0.4762405751358934,
+      "grad_norm": 5.085255146026611,
+      "learning_rate": 4.9376190100657495e-05,
+      "loss": 0.2671,
+      "step": 679
+    },
+    {
+      "epoch": 0.4769419603717342,
+      "grad_norm": 9.267155647277832,
+      "learning_rate": 4.93729673351338e-05,
+      "loss": 0.1125,
+      "step": 680
+    },
+    {
+      "epoch": 0.477643345607575,
+      "grad_norm": 6.162266731262207,
+      "learning_rate": 4.9369736371939426e-05,
+      "loss": 0.0683,
+      "step": 681
+    },
+    {
+      "epoch": 0.4783447308434157,
+      "grad_norm": 3.6411261558532715,
+      "learning_rate": 4.936649721216105e-05,
+      "loss": 0.0356,
+      "step": 682
+    },
+    {
+      "epoch": 0.4790461160792565,
+      "grad_norm": 48.720794677734375,
+      "learning_rate": 4.9363249856888165e-05,
+      "loss": 1.0375,
+      "step": 683
+    },
+    {
+      "epoch": 0.4797475013150973,
+      "grad_norm": 0.3436586856842041,
+      "learning_rate": 4.9359994307212994e-05,
+      "loss": 0.0027,
+      "step": 684
+    },
+    {
+      "epoch": 0.4804488865509381,
+      "grad_norm": 52.97607421875,
+      "learning_rate": 4.93567305642305e-05,
+      "loss": 1.6276,
+      "step": 685
+    },
+    {
+      "epoch": 0.4811502717867789,
+      "grad_norm": 0.09510497748851776,
+      "learning_rate": 4.935345862903844e-05,
+      "loss": 0.0007,
+      "step": 686
+    },
+    {
+      "epoch": 0.4818516570226197,
+      "grad_norm": 56.231292724609375,
+      "learning_rate": 4.93501785027373e-05,
+      "loss": 2.0342,
+      "step": 687
+    },
+    {
+      "epoch": 0.48255304225846046,
+      "grad_norm": 0.21096964180469513,
+      "learning_rate": 4.934689018643032e-05,
+      "loss": 0.0015,
+      "step": 688
+    },
+    {
+      "epoch": 0.48325442749430125,
+      "grad_norm": 0.19893679022789001,
+      "learning_rate": 4.93435936812235e-05,
+      "loss": 0.0014,
+      "step": 689
+    },
+    {
+      "epoch": 0.48395581273014204,
+      "grad_norm": 51.99702072143555,
+      "learning_rate": 4.9340288988225594e-05,
+      "loss": 1.0089,
+      "step": 690
+    },
+    {
+      "epoch": 0.48465719796598283,
+      "grad_norm": 53.519405364990234,
+      "learning_rate": 4.9336976108548126e-05,
+      "loss": 1.7908,
+      "step": 691
+    },
+    {
+      "epoch": 0.4853585832018236,
+      "grad_norm": 106.84630584716797,
+      "learning_rate": 4.9333655043305354e-05,
+      "loss": 3.6845,
+      "step": 692
+    },
+    {
+      "epoch": 0.4860599684376644,
+      "grad_norm": 0.24259038269519806,
+      "learning_rate": 4.9330325793614285e-05,
+      "loss": 0.0019,
+      "step": 693
+    },
+    {
+      "epoch": 0.48676135367350515,
+      "grad_norm": 0.8419340252876282,
+      "learning_rate": 4.93269883605947e-05,
+      "loss": 0.0072,
+      "step": 694
+    },
+    {
+      "epoch": 0.48746273890934594,
+      "grad_norm": 1.204994559288025,
+      "learning_rate": 4.932364274536912e-05,
+      "loss": 0.0087,
+      "step": 695
+    },
+    {
+      "epoch": 0.48816412414518673,
+      "grad_norm": 1.1827445030212402,
+      "learning_rate": 4.932028894906279e-05,
+      "loss": 0.0104,
+      "step": 696
+    },
+    {
+      "epoch": 0.4888655093810275,
+      "grad_norm": 0.5520339608192444,
+      "learning_rate": 4.931692697280378e-05,
+      "loss": 0.0045,
+      "step": 697
+    },
+    {
+      "epoch": 0.4895668946168683,
+      "grad_norm": 0.9636574983596802,
+      "learning_rate": 4.931355681772283e-05,
+      "loss": 0.0079,
+      "step": 698
+    },
+    {
+      "epoch": 0.4902682798527091,
+      "grad_norm": 0.5019407868385315,
+      "learning_rate": 4.931017848495348e-05,
+      "loss": 0.004,
+      "step": 699
+    },
+    {
+      "epoch": 0.4909696650885499,
+      "grad_norm": 0.0757029727101326,
+      "learning_rate": 4.9306791975632026e-05,
+      "loss": 0.0006,
+      "step": 700
+    },
+    {
+      "epoch": 0.4916710503243907,
+      "grad_norm": 110.13203430175781,
+      "learning_rate": 4.9303397290897465e-05,
+      "loss": 4.3731,
+      "step": 701
+    },
+    {
+      "epoch": 0.49237243556023147,
+      "grad_norm": 0.03152499720454216,
+      "learning_rate": 4.929999443189159e-05,
+      "loss": 0.0002,
+      "step": 702
+    },
+    {
+      "epoch": 0.49307382079607226,
+      "grad_norm": 0.029290443286299706,
+      "learning_rate": 4.929658339975893e-05,
+      "loss": 0.0002,
+      "step": 703
+    },
+    {
+      "epoch": 0.49377520603191305,
+      "grad_norm": 0.015484875068068504,
+      "learning_rate": 4.9293164195646754e-05,
+      "loss": 0.0001,
+      "step": 704
+    },
+    {
+      "epoch": 0.49447659126775384,
+      "grad_norm": 53.61540222167969,
+      "learning_rate": 4.928973682070509e-05,
+      "loss": 2.7716,
+      "step": 705
+    },
+    {
+      "epoch": 0.4951779765035946,
+      "grad_norm": 55.3106575012207,
+      "learning_rate": 4.9286301276086715e-05,
+      "loss": 2.3966,
+      "step": 706
+    },
+    {
+      "epoch": 0.49587936173943536,
+      "grad_norm": 0.014001145958900452,
+      "learning_rate": 4.928285756294714e-05,
+      "loss": 0.0001,
+      "step": 707
+    },
+    {
+      "epoch": 0.49658074697527615,
+      "grad_norm": 0.026434065774083138,
+      "learning_rate": 4.927940568244465e-05,
+      "loss": 0.0002,
+      "step": 708
+    },
+    {
+      "epoch": 0.49728213221111695,
+      "grad_norm": 0.04528193548321724,
+      "learning_rate": 4.927594563574023e-05,
+      "loss": 0.0003,
+      "step": 709
+    },
+    {
+      "epoch": 0.49798351744695774,
+      "grad_norm": 0.0950818806886673,
+      "learning_rate": 4.9272477423997666e-05,
+      "loss": 0.0007,
+      "step": 710
+    },
+    {
+      "epoch": 0.4986849026827985,
+      "grad_norm": 0.1095154732465744,
+      "learning_rate": 4.926900104838346e-05,
+      "loss": 0.0008,
+      "step": 711
+    },
+    {
+      "epoch": 0.4993862879186393,
+      "grad_norm": 0.13107261061668396,
+      "learning_rate": 4.9265516510066864e-05,
+      "loss": 0.001,
+      "step": 712
+    },
+    {
+      "epoch": 0.5000876731544801,
+      "grad_norm": 53.58122634887695,
+      "learning_rate": 4.926202381021988e-05,
+      "loss": 1.8605,
+      "step": 713
+    },
+    {
+      "epoch": 0.5007890583903208,
+      "grad_norm": 48.86884689331055,
+      "learning_rate": 4.925852295001725e-05,
+      "loss": 1.3723,
+      "step": 714
+    },
+    {
+      "epoch": 0.5014904436261617,
+      "grad_norm": 0.40202492475509644,
+      "learning_rate": 4.9255013930636465e-05,
+      "loss": 0.0033,
+      "step": 715
+    },
+    {
+      "epoch": 0.5021918288620024,
+      "grad_norm": 43.420196533203125,
+      "learning_rate": 4.925149675325775e-05,
+      "loss": 0.9394,
+      "step": 716
+    },
+    {
+      "epoch": 0.5028932140978433,
+      "grad_norm": 56.66259002685547,
+      "learning_rate": 4.9247971419064097e-05,
+      "loss": 1.2755,
+      "step": 717
+    },
+    {
+      "epoch": 0.503594599333684,
+      "grad_norm": 3.452730894088745,
+      "learning_rate": 4.9244437929241215e-05,
+      "loss": 0.0353,
+      "step": 718
+    },
+    {
+      "epoch": 0.5042959845695248,
+      "grad_norm": 4.454288005828857,
+      "learning_rate": 4.924089628497757e-05,
+      "loss": 0.0486,
+      "step": 719
+    },
+    {
+      "epoch": 0.5049973698053656,
+      "grad_norm": 18.87127685546875,
+      "learning_rate": 4.923734648746437e-05,
+      "loss": 0.3667,
+      "step": 720
+    },
+    {
+      "epoch": 0.5056987550412064,
+      "grad_norm": 11.85842514038086,
+      "learning_rate": 4.923378853789556e-05,
+      "loss": 0.3075,
+      "step": 721
+    },
+    {
+      "epoch": 0.5064001402770472,
+      "grad_norm": 9.63955307006836,
+      "learning_rate": 4.923022243746783e-05,
+      "loss": 0.1304,
+      "step": 722
+    },
+    {
+      "epoch": 0.507101525512888,
+      "grad_norm": 9.134459495544434,
+      "learning_rate": 4.922664818738063e-05,
+      "loss": 0.1346,
+      "step": 723
+    },
+    {
+      "epoch": 0.5078029107487287,
+      "grad_norm": 2.0840797424316406,
+      "learning_rate": 4.92230657888361e-05,
+      "loss": 0.2203,
+      "step": 724
+    },
+    {
+      "epoch": 0.5085042959845695,
+      "grad_norm": 8.857693672180176,
+      "learning_rate": 4.921947524303919e-05,
+      "loss": 0.1232,
+      "step": 725
+    },
+    {
+      "epoch": 0.5092056812204103,
+      "grad_norm": 8.82702350616455,
+      "learning_rate": 4.921587655119753e-05,
+      "loss": 0.1143,
+      "step": 726
+    },
+    {
+      "epoch": 0.5099070664562511,
+      "grad_norm": 4.017044544219971,
+      "learning_rate": 4.9212269714521505e-05,
+      "loss": 0.0433,
+      "step": 727
+    },
+    {
+      "epoch": 0.5106084516920919,
+      "grad_norm": 1.7636412382125854,
+      "learning_rate": 4.920865473422428e-05,
+      "loss": 0.0171,
+      "step": 728
+    },
+    {
+      "epoch": 0.5113098369279326,
+      "grad_norm": 107.01025390625,
+      "learning_rate": 4.920503161152171e-05,
+      "loss": 2.4773,
+      "step": 729
+    },
+    {
+      "epoch": 0.5120112221637735,
+      "grad_norm": 0.35363465547561646,
+      "learning_rate": 4.9201400347632404e-05,
+      "loss": 0.0028,
+      "step": 730
+    },
+    {
+      "epoch": 0.5127126073996142,
+      "grad_norm": 0.19405458867549896,
+      "learning_rate": 4.919776094377771e-05,
+      "loss": 0.0015,
+      "step": 731
+    },
+    {
+      "epoch": 0.5134139926354551,
+      "grad_norm": 55.00040054321289,
+      "learning_rate": 4.919411340118173e-05,
+      "loss": 1.9537,
+      "step": 732
+    },
+    {
+      "epoch": 0.5141153778712958,
+      "grad_norm": 55.760963439941406,
+      "learning_rate": 4.919045772107127e-05,
+      "loss": 2.1316,
+      "step": 733
+    },
+    {
+      "epoch": 0.5148167631071366,
+      "grad_norm": 55.24003601074219,
+      "learning_rate": 4.91867939046759e-05,
+      "loss": 2.1231,
+      "step": 734
+    },
+    {
+      "epoch": 0.5155181483429774,
+      "grad_norm": 0.03399452567100525,
+      "learning_rate": 4.918312195322792e-05,
+      "loss": 0.0002,
+      "step": 735
+    },
+    {
+      "epoch": 0.5162195335788181,
+      "grad_norm": 204.8887176513672,
+      "learning_rate": 4.917944186796235e-05,
+      "loss": 8.6129,
+      "step": 736
+    },
+    {
+      "epoch": 0.516920918814659,
+      "grad_norm": 124.1363754272461,
+      "learning_rate": 4.917575365011699e-05,
+      "loss": 3.9954,
+      "step": 737
+    },
+    {
+      "epoch": 0.5176223040504997,
+      "grad_norm": 0.14609888195991516,
+      "learning_rate": 4.9172057300932306e-05,
+      "loss": 0.0011,
+      "step": 738
+    },
+    {
+      "epoch": 0.5183236892863405,
+      "grad_norm": 0.33696234226226807,
+      "learning_rate": 4.9168352821651575e-05,
+      "loss": 0.0027,
+      "step": 739
+    },
+    {
+      "epoch": 0.5190250745221813,
+      "grad_norm": 105.28889465332031,
+      "learning_rate": 4.916464021352075e-05,
+      "loss": 3.0367,
+      "step": 740
+    },
+    {
+      "epoch": 0.5197264597580221,
+      "grad_norm": 1.017066240310669,
+      "learning_rate": 4.916091947778855e-05,
+      "loss": 0.0088,
+      "step": 741
+    },
+    {
+      "epoch": 0.5204278449938629,
+      "grad_norm": 1.143968939781189,
+      "learning_rate": 4.91571906157064e-05,
+      "loss": 0.0102,
+      "step": 742
+    },
+    {
+      "epoch": 0.5211292302297037,
+      "grad_norm": 0.7457045912742615,
+      "learning_rate": 4.91534536285285e-05,
+      "loss": 0.0063,
+      "step": 743
+    },
+    {
+      "epoch": 0.5218306154655444,
+      "grad_norm": 0.700978696346283,
+      "learning_rate": 4.9149708517511743e-05,
+      "loss": 0.0058,
+      "step": 744
+    },
+    {
+      "epoch": 0.5225320007013853,
+      "grad_norm": 0.38463106751441956,
+      "learning_rate": 4.914595528391578e-05,
+      "loss": 0.0031,
+      "step": 745
+    },
+    {
+      "epoch": 0.523233385937226,
+      "grad_norm": 0.14777745306491852,
+      "learning_rate": 4.914219392900297e-05,
+      "loss": 0.0011,
+      "step": 746
+    },
+    {
+      "epoch": 0.5239347711730669,
+      "grad_norm": 0.07884679734706879,
+      "learning_rate": 4.913842445403843e-05,
+      "loss": 0.0006,
+      "step": 747
+    },
+    {
+      "epoch": 0.5246361564089076,
+      "grad_norm": 61.55155563354492,
+      "learning_rate": 4.913464686028999e-05,
+      "loss": 1.992,
+      "step": 748
+    },
+    {
+      "epoch": 0.5253375416447483,
+      "grad_norm": 0.027440281584858894,
+      "learning_rate": 4.913086114902822e-05,
+      "loss": 0.0002,
+      "step": 749
+    },
+    {
+      "epoch": 0.5260389268805892,
+      "grad_norm": 55.66230010986328,
+      "learning_rate": 4.912706732152641e-05,
+      "loss": 2.231,
+      "step": 750
+    },
+    {
+      "epoch": 0.5267403121164299,
+      "grad_norm": 54.453102111816406,
+      "learning_rate": 4.912326537906059e-05,
+      "loss": 2.0983,
+      "step": 751
+    },
+    {
+      "epoch": 0.5274416973522708,
+      "grad_norm": 0.0637647733092308,
+      "learning_rate": 4.911945532290951e-05,
+      "loss": 0.0005,
+      "step": 752
+    },
+    {
+      "epoch": 0.5281430825881115,
+      "grad_norm": 0.13056063652038574,
+      "learning_rate": 4.9115637154354674e-05,
+      "loss": 0.001,
+      "step": 753
+    },
+    {
+      "epoch": 0.5288444678239523,
+      "grad_norm": 52.72209167480469,
+      "learning_rate": 4.911181087468027e-05,
+      "loss": 1.5666,
+      "step": 754
+    },
+    {
+      "epoch": 0.5295458530597931,
+      "grad_norm": 48.385467529296875,
+      "learning_rate": 4.9107976485173255e-05,
+      "loss": 1.2941,
+      "step": 755
+    },
+    {
+      "epoch": 0.5302472382956339,
+      "grad_norm": 0.5080452561378479,
+      "learning_rate": 4.91041339871233e-05,
+      "loss": 0.0041,
+      "step": 756
+    },
+    {
+      "epoch": 0.5309486235314747,
+      "grad_norm": 80.44468688964844,
+      "learning_rate": 4.910028338182279e-05,
+      "loss": 1.6558,
+      "step": 757
+    },
+    {
+      "epoch": 0.5316500087673155,
+      "grad_norm": 1.59328031539917,
+      "learning_rate": 4.909642467056685e-05,
+      "loss": 0.0141,
+      "step": 758
+    },
+    {
+      "epoch": 0.5323513940031562,
+      "grad_norm": 2.678886651992798,
+      "learning_rate": 4.9092557854653344e-05,
+      "loss": 0.0255,
+      "step": 759
+    },
+    {
+      "epoch": 0.533052779238997,
+      "grad_norm": 2.742159366607666,
+      "learning_rate": 4.908868293538282e-05,
+      "loss": 0.0271,
+      "step": 760
+    },
+    {
+      "epoch": 0.5337541644748378,
+      "grad_norm": 2.275885820388794,
+      "learning_rate": 4.908479991405861e-05,
+      "loss": 0.0215,
+      "step": 761
+    },
+    {
+      "epoch": 0.5344555497106785,
+      "grad_norm": 1.4059484004974365,
+      "learning_rate": 4.9080908791986724e-05,
+      "loss": 0.0126,
+      "step": 762
+    },
+    {
+      "epoch": 0.5351569349465194,
+      "grad_norm": 0.7012129426002502,
+      "learning_rate": 4.907700957047592e-05,
+      "loss": 0.0058,
+      "step": 763
+    },
+    {
+      "epoch": 0.5358583201823601,
+      "grad_norm": 0.2713419497013092,
+      "learning_rate": 4.9073102250837655e-05,
+      "loss": 0.0021,
+      "step": 764
+    },
+    {
+      "epoch": 0.536559705418201,
+      "grad_norm": 0.13228341937065125,
+      "learning_rate": 4.9069186834386145e-05,
+      "loss": 0.001,
+      "step": 765
+    },
+    {
+      "epoch": 0.5372610906540417,
+      "grad_norm": 0.06449166685342789,
+      "learning_rate": 4.90652633224383e-05,
+      "loss": 0.0005,
+      "step": 766
+    },
+    {
+      "epoch": 0.5379624758898826,
+      "grad_norm": 0.017581632360816002,
+      "learning_rate": 4.9061331716313783e-05,
+      "loss": 0.0001,
+      "step": 767
+    },
+    {
+      "epoch": 0.5386638611257233,
+      "grad_norm": 0.016885504126548767,
+      "learning_rate": 4.905739201733495e-05,
+      "loss": 0.0001,
+      "step": 768
+    },
+    {
+      "epoch": 0.5393652463615641,
+      "grad_norm": 0.005497175268828869,
+      "learning_rate": 4.9053444226826886e-05,
+      "loss": 0.0,
+      "step": 769
+    },
+    {
+      "epoch": 0.5400666315974049,
+      "grad_norm": 58.73657989501953,
+      "learning_rate": 4.90494883461174e-05,
+      "loss": 3.1923,
+      "step": 770
+    },
+    {
+      "epoch": 0.5407680168332457,
+      "grad_norm": 0.0022041036281734705,
+      "learning_rate": 4.9045524376537034e-05,
+      "loss": 0.0,
+      "step": 771
+    },
+    {
+      "epoch": 0.5414694020690864,
+      "grad_norm": 52.654136657714844,
+      "learning_rate": 4.9041552319419035e-05,
+      "loss": 3.097,
+      "step": 772
+    },
+    {
+      "epoch": 0.5421707873049272,
+      "grad_norm": 0.0026611709035933018,
+      "learning_rate": 4.903757217609938e-05,
+      "loss": 0.0,
+      "step": 773
+    },
+    {
+      "epoch": 0.542872172540768,
+      "grad_norm": 0.0028983913362026215,
+      "learning_rate": 4.9033583947916765e-05,
+      "loss": 0.0,
+      "step": 774
+    },
+    {
+      "epoch": 0.5435735577766088,
+      "grad_norm": 105.84739685058594,
+      "learning_rate": 4.902958763621258e-05,
+      "loss": 5.9716,
+      "step": 775
+    },
+    {
+      "epoch": 0.5442749430124496,
+      "grad_norm": 54.44437789916992,
+      "learning_rate": 4.902558324233098e-05,
+      "loss": 2.7511,
+      "step": 776
+    },
+    {
+      "epoch": 0.5449763282482903,
+      "grad_norm": 0.009829580783843994,
+      "learning_rate": 4.9021570767618794e-05,
+      "loss": 0.0001,
+      "step": 777
+    },
+    {
+      "epoch": 0.5456777134841312,
+      "grad_norm": 53.416908264160156,
+      "learning_rate": 4.90175502134256e-05,
+      "loss": 2.337,
+      "step": 778
+    },
+    {
+      "epoch": 0.5463790987199719,
+      "grad_norm": 0.050978660583496094,
+      "learning_rate": 4.9013521581103686e-05,
+      "loss": 0.0004,
+      "step": 779
+    },
+    {
+      "epoch": 0.5470804839558128,
+      "grad_norm": 0.11917990446090698,
+      "learning_rate": 4.900948487200804e-05,
+      "loss": 0.0009,
+      "step": 780
+    },
+    {
+      "epoch": 0.5477818691916535,
+      "grad_norm": 48.541015625,
+      "learning_rate": 4.900544008749639e-05,
+      "loss": 1.2937,
+      "step": 781
+    },
+    {
+      "epoch": 0.5484832544274943,
+      "grad_norm": 0.3613051474094391,
+      "learning_rate": 4.900138722892917e-05,
+      "loss": 0.003,
+      "step": 782
+    },
+    {
+      "epoch": 0.5491846396633351,
+      "grad_norm": 35.976261138916016,
+      "learning_rate": 4.899732629766953e-05,
+      "loss": 0.7424,
+      "step": 783
+    },
+    {
+      "epoch": 0.5498860248991758,
+      "grad_norm": 1.3365657329559326,
+      "learning_rate": 4.899325729508333e-05,
+      "loss": 0.012,
+      "step": 784
+    },
+    {
+      "epoch": 0.5505874101350167,
+      "grad_norm": 2.2534306049346924,
+      "learning_rate": 4.898918022253916e-05,
+      "loss": 0.0228,
+      "step": 785
+    },
+    {
+      "epoch": 0.5512887953708574,
+      "grad_norm": 24.77885627746582,
+      "learning_rate": 4.898509508140829e-05,
+      "loss": 0.4601,
+      "step": 786
+    },
+    {
+      "epoch": 0.5519901806066982,
+      "grad_norm": 4.28820276260376,
+      "learning_rate": 4.898100187306475e-05,
+      "loss": 0.0482,
+      "step": 787
+    },
+    {
+      "epoch": 0.552691565842539,
+      "grad_norm": 4.493249416351318,
+      "learning_rate": 4.897690059888524e-05,
+      "loss": 0.0503,
+      "step": 788
+    },
+    {
+      "epoch": 0.5533929510783798,
+      "grad_norm": 3.269726037979126,
+      "learning_rate": 4.897279126024923e-05,
+      "loss": 0.0349,
+      "step": 789
+    },
+    {
+      "epoch": 0.5540943363142206,
+      "grad_norm": 3.0403542518615723,
+      "learning_rate": 4.896867385853883e-05,
+      "loss": 0.0316,
+      "step": 790
+    },
+    {
+      "epoch": 0.5547957215500614,
+      "grad_norm": 25.840961456298828,
+      "learning_rate": 4.896454839513892e-05,
+      "loss": 0.4405,
+      "step": 791
+    },
+    {
+      "epoch": 0.5554971067859021,
+      "grad_norm": 72.00930786132812,
+      "learning_rate": 4.8960414871437055e-05,
+      "loss": 1.4578,
+      "step": 792
+    },
+    {
+      "epoch": 0.556198492021743,
+      "grad_norm": 0.7836804389953613,
+      "learning_rate": 4.895627328882352e-05,
+      "loss": 0.0068,
+      "step": 793
+    },
+    {
+      "epoch": 0.5568998772575837,
+      "grad_norm": 0.7037481069564819,
+      "learning_rate": 4.895212364869132e-05,
+      "loss": 0.0061,
+      "step": 794
+    },
+    {
+      "epoch": 0.5576012624934246,
+      "grad_norm": 0.4775547981262207,
+      "learning_rate": 4.894796595243614e-05,
+      "loss": 0.004,
+      "step": 795
+    },
+    {
+      "epoch": 0.5583026477292653,
+      "grad_norm": 0.4537871778011322,
+      "learning_rate": 4.894380020145639e-05,
+      "loss": 0.0038,
+      "step": 796
+    },
+    {
+      "epoch": 0.559004032965106,
+      "grad_norm": 0.2530267536640167,
+      "learning_rate": 4.8939626397153196e-05,
+      "loss": 0.0021,
+      "step": 797
+    },
+    {
+      "epoch": 0.5597054182009469,
+      "grad_norm": 0.08755356073379517,
+      "learning_rate": 4.8935444540930385e-05,
+      "loss": 0.0007,
+      "step": 798
+    },
+    {
+      "epoch": 0.5604068034367876,
+      "grad_norm": 0.058386508375406265,
+      "learning_rate": 4.8931254634194504e-05,
+      "loss": 0.0004,
+      "step": 799
+    },
+    {
+      "epoch": 0.5611081886726285,
+      "grad_norm": 0.020555956289172173,
+      "learning_rate": 4.892705667835479e-05,
+      "loss": 0.0002,
+      "step": 800
+    },
+    {
+      "epoch": 0.5618095739084692,
+      "grad_norm": 0.023624928668141365,
+      "learning_rate": 4.892285067482319e-05,
+      "loss": 0.0002,
+      "step": 801
+    },
+    {
+      "epoch": 0.56251095914431,
+      "grad_norm": 54.471675872802734,
+      "learning_rate": 4.891863662501437e-05,
+      "loss": 2.8657,
+      "step": 802
+    },
+    {
+      "epoch": 0.5632123443801508,
+      "grad_norm": 0.005444019101560116,
+      "learning_rate": 4.8914414530345684e-05,
+      "loss": 0.0,
+      "step": 803
+    },
+    {
+      "epoch": 0.5639137296159916,
+      "grad_norm": 0.008992165327072144,
+      "learning_rate": 4.891018439223722e-05,
+      "loss": 0.0001,
+      "step": 804
+    },
+    {
+      "epoch": 0.5646151148518324,
+      "grad_norm": 0.0062455967999994755,
+      "learning_rate": 4.890594621211174e-05,
+      "loss": 0.0,
+      "step": 805
+    },
+    {
+      "epoch": 0.5653165000876732,
+      "grad_norm": 0.006425914354622364,
+      "learning_rate": 4.8901699991394735e-05,
+      "loss": 0.0,
+      "step": 806
+    },
+    {
+      "epoch": 0.5660178853235139,
+      "grad_norm": 0.008212440647184849,
+      "learning_rate": 4.8897445731514384e-05,
+      "loss": 0.0001,
+      "step": 807
+    },
+    {
+      "epoch": 0.5667192705593547,
+      "grad_norm": 0.004950184375047684,
+      "learning_rate": 4.889318343390158e-05,
+      "loss": 0.0,
+      "step": 808
+    },
+    {
+      "epoch": 0.5674206557951955,
+      "grad_norm": 0.0056209079921245575,
+      "learning_rate": 4.8888913099989916e-05,
+      "loss": 0.0,
+      "step": 809
+    },
+    {
+      "epoch": 0.5681220410310363,
+      "grad_norm": 0.0028246166184544563,
+      "learning_rate": 4.888463473121568e-05,
+      "loss": 0.0,
+      "step": 810
+    },
+    {
+      "epoch": 0.5688234262668771,
+      "grad_norm": 53.988868713378906,
+      "learning_rate": 4.888034832901788e-05,
+      "loss": 2.7359,
+      "step": 811
+    },
+    {
+      "epoch": 0.5695248115027178,
+      "grad_norm": 0.004621135536581278,
+      "learning_rate": 4.8876053894838215e-05,
+      "loss": 0.0,
+      "step": 812
+    },
+    {
+      "epoch": 0.5702261967385587,
+      "grad_norm": 0.011877781711518764,
+      "learning_rate": 4.887175143012108e-05,
+      "loss": 0.0001,
+      "step": 813
+    },
+    {
+      "epoch": 0.5709275819743994,
+      "grad_norm": 65.13773345947266,
+      "learning_rate": 4.886744093631358e-05,
+      "loss": 3.5609,
+      "step": 814
+    },
+    {
+      "epoch": 0.5716289672102403,
+      "grad_norm": 161.86300659179688,
+      "learning_rate": 4.886312241486554e-05,
+      "loss": 7.2583,
+      "step": 815
+    },
+    {
+      "epoch": 0.572330352446081,
+      "grad_norm": 0.032078277319669724,
+      "learning_rate": 4.8858795867229425e-05,
+      "loss": 0.0002,
+      "step": 816
+    },
+    {
+      "epoch": 0.5730317376819218,
+      "grad_norm": 113.32162475585938,
+      "learning_rate": 4.8854461294860454e-05,
+      "loss": 4.1658,
+      "step": 817
+    },
+    {
+      "epoch": 0.5737331229177626,
+      "grad_norm": 0.17656919360160828,
+      "learning_rate": 4.885011869921654e-05,
+      "loss": 0.0014,
+      "step": 818
+    },
+    {
+      "epoch": 0.5744345081536034,
+      "grad_norm": 0.2188340574502945,
+      "learning_rate": 4.884576808175828e-05,
+      "loss": 0.0018,
+      "step": 819
+    },
+    {
+      "epoch": 0.5751358933894442,
+      "grad_norm": 0.38712555170059204,
+      "learning_rate": 4.884140944394896e-05,
+      "loss": 0.0033,
+      "step": 820
+    },
+    {
+      "epoch": 0.5758372786252849,
+      "grad_norm": 0.7210093140602112,
+      "learning_rate": 4.883704278725458e-05,
+      "loss": 0.0062,
+      "step": 821
+    },
+    {
+      "epoch": 0.5765386638611257,
+      "grad_norm": 0.7694672346115112,
+      "learning_rate": 4.883266811314385e-05,
+      "loss": 0.0069,
+      "step": 822
+    },
+    {
+      "epoch": 0.5772400490969665,
+      "grad_norm": 0.6353862881660461,
+      "learning_rate": 4.8828285423088146e-05,
+      "loss": 0.0056,
+      "step": 823
+    },
+    {
+      "epoch": 0.5779414343328073,
+      "grad_norm": 87.86915588378906,
+      "learning_rate": 4.8823894718561545e-05,
+      "loss": 2.1381,
+      "step": 824
+    },
+    {
+      "epoch": 0.578642819568648,
+      "grad_norm": 0.42429307103157043,
+      "learning_rate": 4.8819496001040846e-05,
+      "loss": 0.0036,
+      "step": 825
+    },
+    {
+      "epoch": 0.5793442048044889,
+      "grad_norm": 0.40686339139938354,
+      "learning_rate": 4.881508927200552e-05,
+      "loss": 0.0034,
+      "step": 826
+    },
+    {
+      "epoch": 0.5800455900403296,
+      "grad_norm": 0.27934542298316956,
+      "learning_rate": 4.881067453293774e-05,
+      "loss": 0.0022,
+      "step": 827
+    },
+    {
+      "epoch": 0.5807469752761705,
+      "grad_norm": 44.31092834472656,
+      "learning_rate": 4.8806251785322355e-05,
+      "loss": 1.1003,
+      "step": 828
+    },
+    {
+      "epoch": 0.5814483605120112,
+      "grad_norm": 0.17587609589099884,
+      "learning_rate": 4.880182103064695e-05,
+      "loss": 0.0014,
+      "step": 829
+    },
+    {
+      "epoch": 0.582149745747852,
+      "grad_norm": 0.2651273012161255,
+      "learning_rate": 4.879738227040176e-05,
+      "loss": 0.0021,
+      "step": 830
+    },
+    {
+      "epoch": 0.5828511309836928,
+      "grad_norm": 149.13035583496094,
+      "learning_rate": 4.879293550607974e-05,
+      "loss": 2.9684,
+      "step": 831
+    },
+    {
+      "epoch": 0.5835525162195335,
+      "grad_norm": 0.17069180309772491,
+      "learning_rate": 4.878848073917651e-05,
+      "loss": 0.0014,
+      "step": 832
+    },
+    {
+      "epoch": 0.5842539014553744,
+      "grad_norm": 0.24799254536628723,
+      "learning_rate": 4.878401797119042e-05,
+      "loss": 0.002,
+      "step": 833
+    },
+    {
+      "epoch": 0.5849552866912151,
+      "grad_norm": 0.15171018242835999,
+      "learning_rate": 4.8779547203622477e-05,
+      "loss": 0.0012,
+      "step": 834
+    },
+    {
+      "epoch": 0.585656671927056,
+      "grad_norm": 45.8453254699707,
+      "learning_rate": 4.877506843797639e-05,
+      "loss": 1.2116,
+      "step": 835
+    },
+    {
+      "epoch": 0.5863580571628967,
+      "grad_norm": 70.00684356689453,
+      "learning_rate": 4.877058167575857e-05,
+      "loss": 2.1496,
+      "step": 836
+    },
+    {
+      "epoch": 0.5870594423987375,
+      "grad_norm": 43.86625671386719,
+      "learning_rate": 4.876608691847809e-05,
+      "loss": 1.077,
+      "step": 837
+    },
+    {
+      "epoch": 0.5877608276345783,
+      "grad_norm": 39.75180435180664,
+      "learning_rate": 4.876158416764675e-05,
+      "loss": 0.8959,
+      "step": 838
+    },
+    {
+      "epoch": 0.5884622128704191,
+      "grad_norm": 78.85111999511719,
+      "learning_rate": 4.8757073424779e-05,
+      "loss": 1.609,
+      "step": 839
+    },
+    {
+      "epoch": 0.5891635981062598,
+      "grad_norm": 2.562579393386841,
+      "learning_rate": 4.8752554691392004e-05,
+      "loss": 0.0262,
+      "step": 840
+    },
+    {
+      "epoch": 0.5898649833421007,
+      "grad_norm": 3.174607276916504,
+      "learning_rate": 4.8748027969005606e-05,
+      "loss": 0.0331,
+      "step": 841
+    },
+    {
+      "epoch": 0.5905663685779414,
+      "grad_norm": 24.42951202392578,
+      "learning_rate": 4.874349325914232e-05,
+      "loss": 0.494,
+      "step": 842
+    },
+    {
+      "epoch": 0.5912677538137823,
+      "grad_norm": 5.76008415222168,
+      "learning_rate": 4.87389505633274e-05,
+      "loss": 0.069,
+      "step": 843
+    },
+    {
+      "epoch": 0.591969139049623,
+      "grad_norm": 16.484464645385742,
+      "learning_rate": 4.873439988308871e-05,
+      "loss": 0.3677,
+      "step": 844
+    },
+    {
+      "epoch": 0.5926705242854637,
+      "grad_norm": 8.441133499145508,
+      "learning_rate": 4.872984121995686e-05,
+      "loss": 0.1122,
+      "step": 845
+    },
+    {
+      "epoch": 0.5933719095213046,
+      "grad_norm": 7.171937942504883,
+      "learning_rate": 4.8725274575465106e-05,
+      "loss": 0.093,
+      "step": 846
+    },
+    {
+      "epoch": 0.5940732947571453,
+      "grad_norm": 7.630092620849609,
+      "learning_rate": 4.872069995114943e-05,
+      "loss": 0.0997,
+      "step": 847
+    },
+    {
+      "epoch": 0.5947746799929862,
+      "grad_norm": 5.521340370178223,
+      "learning_rate": 4.871611734854845e-05,
+      "loss": 0.065,
+      "step": 848
+    },
+    {
+      "epoch": 0.5954760652288269,
+      "grad_norm": 40.95110321044922,
+      "learning_rate": 4.871152676920351e-05,
+      "loss": 0.6631,
+      "step": 849
+    },
+    {
+      "epoch": 0.5961774504646677,
+      "grad_norm": 2.602055311203003,
+      "learning_rate": 4.8706928214658596e-05,
+      "loss": 0.0269,
+      "step": 850
+    },
+    {
+      "epoch": 0.5968788357005085,
+      "grad_norm": 30.752164840698242,
+      "learning_rate": 4.8702321686460424e-05,
+      "loss": 0.5944,
+      "step": 851
+    },
+    {
+      "epoch": 0.5975802209363493,
+      "grad_norm": 34.97576904296875,
+      "learning_rate": 4.8697707186158345e-05,
+      "loss": 0.7216,
+      "step": 852
+    },
+    {
+      "epoch": 0.5982816061721901,
+      "grad_norm": 1.3377373218536377,
+      "learning_rate": 4.869308471530443e-05,
+      "loss": 0.0126,
+      "step": 853
+    },
+    {
+      "epoch": 0.5989829914080309,
+      "grad_norm": 33.73777389526367,
+      "learning_rate": 4.868845427545341e-05,
+      "loss": 0.6655,
+      "step": 854
+    },
+    {
+      "epoch": 0.5996843766438716,
+      "grad_norm": 27.722604751586914,
+      "learning_rate": 4.868381586816268e-05,
+      "loss": 0.5284,
+      "step": 855
+    },
+    {
+      "epoch": 0.6003857618797125,
+      "grad_norm": 39.0617561340332,
+      "learning_rate": 4.867916949499236e-05,
+      "loss": 0.9141,
+      "step": 856
+    },
+    {
+      "epoch": 0.6010871471155532,
+      "grad_norm": 25.998207092285156,
+      "learning_rate": 4.8674515157505205e-05,
+      "loss": 0.4843,
+      "step": 857
+    },
+    {
+      "epoch": 0.601788532351394,
+      "grad_norm": 70.33052062988281,
+      "learning_rate": 4.8669852857266686e-05,
+      "loss": 1.1847,
+      "step": 858
+    },
+    {
+      "epoch": 0.6024899175872348,
+      "grad_norm": 2.859663724899292,
+      "learning_rate": 4.866518259584492e-05,
+      "loss": 0.1724,
+      "step": 859
+    },
+    {
+      "epoch": 0.6031913028230755,
+      "grad_norm": 7.6815185546875,
+      "learning_rate": 4.866050437481072e-05,
+      "loss": 0.2053,
+      "step": 860
+    },
+    {
+      "epoch": 0.6038926880589164,
+      "grad_norm": 15.574623107910156,
+      "learning_rate": 4.8655818195737576e-05,
+      "loss": 0.2786,
+      "step": 861
+    },
+    {
+      "epoch": 0.6045940732947571,
+      "grad_norm": 17.501440048217773,
+      "learning_rate": 4.865112406020164e-05,
+      "loss": 0.3404,
+      "step": 862
+    },
+    {
+      "epoch": 0.605295458530598,
+      "grad_norm": 18.48247718811035,
+      "learning_rate": 4.8646421969781757e-05,
+      "loss": 0.3676,
+      "step": 863
+    },
+    {
+      "epoch": 0.6059968437664387,
+      "grad_norm": 15.630892753601074,
+      "learning_rate": 4.864171192605944e-05,
+      "loss": 0.2856,
+      "step": 864
+    },
+    {
+      "epoch": 0.6066982290022795,
+      "grad_norm": 15.12441349029541,
+      "learning_rate": 4.863699393061888e-05,
+      "loss": 0.2728,
+      "step": 865
+    },
+    {
+      "epoch": 0.6073996142381203,
+      "grad_norm": 10.427019119262695,
+      "learning_rate": 4.863226798504694e-05,
+      "loss": 0.308,
+      "step": 866
+    },
+    {
+      "epoch": 0.6081009994739611,
+      "grad_norm": 10.065474510192871,
+      "learning_rate": 4.8627534090933156e-05,
+      "loss": 0.1617,
+      "step": 867
+    },
+    {
+      "epoch": 0.6088023847098019,
+      "grad_norm": 11.481346130371094,
+      "learning_rate": 4.862279224986973e-05,
+      "loss": 0.2136,
+      "step": 868
+    },
+    {
+      "epoch": 0.6095037699456426,
+      "grad_norm": 5.78438663482666,
+      "learning_rate": 4.8618042463451566e-05,
+      "loss": 0.0726,
+      "step": 869
+    },
+    {
+      "epoch": 0.6102051551814834,
+      "grad_norm": 39.987937927246094,
+      "learning_rate": 4.8613284733276206e-05,
+      "loss": 0.8901,
+      "step": 870
+    },
+    {
+      "epoch": 0.6109065404173242,
+      "grad_norm": 3.24749755859375,
+      "learning_rate": 4.860851906094388e-05,
+      "loss": 0.0369,
+      "step": 871
+    },
+    {
+      "epoch": 0.611607925653165,
+      "grad_norm": 2.4093921184539795,
+      "learning_rate": 4.860374544805749e-05,
+      "loss": 0.0254,
+      "step": 872
+    },
+    {
+      "epoch": 0.6123093108890058,
+      "grad_norm": 0.970287024974823,
+      "learning_rate": 4.85989638962226e-05,
+      "loss": 0.0092,
+      "step": 873
+    },
+    {
+      "epoch": 0.6130106961248466,
+      "grad_norm": 0.4123772084712982,
+      "learning_rate": 4.859417440704746e-05,
+      "loss": 0.0036,
+      "step": 874
+    },
+    {
+      "epoch": 0.6137120813606873,
+      "grad_norm": 0.14014503359794617,
+      "learning_rate": 4.8589376982142977e-05,
+      "loss": 0.0011,
+      "step": 875
+    },
+    {
+      "epoch": 0.6144134665965282,
+      "grad_norm": 0.05853430926799774,
+      "learning_rate": 4.858457162312272e-05,
+      "loss": 0.0005,
+      "step": 876
+    },
+    {
+      "epoch": 0.6151148518323689,
+      "grad_norm": 104.4763412475586,
+      "learning_rate": 4.857975833160295e-05,
+      "loss": 4.5228,
+      "step": 877
+    },
+    {
+      "epoch": 0.6158162370682098,
+      "grad_norm": 60.22207260131836,
+      "learning_rate": 4.857493710920258e-05,
+      "loss": 3.2127,
+      "step": 878
+    },
+    {
+      "epoch": 0.6165176223040505,
+      "grad_norm": 0.014606939628720284,
+      "learning_rate": 4.8570107957543185e-05,
+      "loss": 0.0001,
+      "step": 879
+    },
+    {
+      "epoch": 0.6172190075398913,
+      "grad_norm": 52.49748229980469,
+      "learning_rate": 4.856527087824902e-05,
+      "loss": 2.5181,
+      "step": 880
+    },
+    {
+      "epoch": 0.6179203927757321,
+      "grad_norm": 60.33863067626953,
+      "learning_rate": 4.8560425872947004e-05,
+      "loss": 3.4192,
+      "step": 881
+    },
+    {
+      "epoch": 0.6186217780115728,
+      "grad_norm": 0.017534319311380386,
+      "learning_rate": 4.8555572943266715e-05,
+      "loss": 0.0001,
+      "step": 882
+    },
+    {
+      "epoch": 0.6193231632474137,
+      "grad_norm": 0.03253171965479851,
+      "learning_rate": 4.85507120908404e-05,
+      "loss": 0.0002,
+      "step": 883
+    },
+    {
+      "epoch": 0.6200245484832544,
+      "grad_norm": 100.64522552490234,
+      "learning_rate": 4.8545843317302983e-05,
+      "loss": 3.6464,
+      "step": 884
+    },
+    {
+      "epoch": 0.6207259337190952,
+      "grad_norm": 52.276100158691406,
+      "learning_rate": 4.854096662429203e-05,
+      "loss": 1.5974,
+      "step": 885
+    },
+    {
+      "epoch": 0.621427318954936,
+      "grad_norm": 0.26515480875968933,
+      "learning_rate": 4.853608201344778e-05,
+      "loss": 0.0022,
+      "step": 886
+    },
+    {
+      "epoch": 0.6221287041907768,
+      "grad_norm": 0.4160686731338501,
+      "learning_rate": 4.853118948641314e-05,
+      "loss": 0.0036,
+      "step": 887
+    },
+    {
+      "epoch": 0.6228300894266176,
+      "grad_norm": 40.08952331542969,
+      "learning_rate": 4.852628904483367e-05,
+      "loss": 0.9215,
+      "step": 888
+    },
+    {
+      "epoch": 0.6235314746624584,
+      "grad_norm": 35.90172576904297,
+      "learning_rate": 4.852138069035761e-05,
+      "loss": 0.7604,
+      "step": 889
+    },
+    {
+      "epoch": 0.6242328598982991,
+      "grad_norm": 2.5494091510772705,
+      "learning_rate": 4.851646442463585e-05,
+      "loss": 0.0269,
+      "step": 890
+    },
+    {
+      "epoch": 0.62493424513414,
+      "grad_norm": 4.3015456199646,
+      "learning_rate": 4.851154024932193e-05,
+      "loss": 0.0495,
+      "step": 891
+    },
+    {
+      "epoch": 0.6256356303699807,
+      "grad_norm": 3.570831537246704,
+      "learning_rate": 4.8506608166072054e-05,
+      "loss": 0.0385,
+      "step": 892
+    },
+    {
+      "epoch": 0.6263370156058214,
+      "grad_norm": 3.575366735458374,
+      "learning_rate": 4.8501668176545115e-05,
+      "loss": 0.0397,
+      "step": 893
+    },
+    {
+      "epoch": 0.6270384008416623,
+      "grad_norm": 2.7419331073760986,
+      "learning_rate": 4.849672028240262e-05,
+      "loss": 0.0288,
+      "step": 894
+    },
+    {
+      "epoch": 0.627739786077503,
+      "grad_norm": 41.37546157836914,
+      "learning_rate": 4.849176448530878e-05,
+      "loss": 0.8202,
+      "step": 895
+    },
+    {
+      "epoch": 0.6284411713133439,
+      "grad_norm": 34.27451705932617,
+      "learning_rate": 4.8486800786930415e-05,
+      "loss": 0.7397,
+      "step": 896
+    },
+    {
+      "epoch": 0.6291425565491846,
+      "grad_norm": 1.0924946069717407,
+      "learning_rate": 4.848182918893705e-05,
+      "loss": 0.0104,
+      "step": 897
+    },
+    {
+      "epoch": 0.6298439417850255,
+      "grad_norm": 0.9149582386016846,
+      "learning_rate": 4.8476849693000836e-05,
+      "loss": 0.0085,
+      "step": 898
+    },
+    {
+      "epoch": 0.6305453270208662,
+      "grad_norm": 0.4877033829689026,
+      "learning_rate": 4.8471862300796594e-05,
+      "loss": 0.0043,
+      "step": 899
+    },
+    {
+      "epoch": 0.631246712256707,
+      "grad_norm": 50.69717788696289,
+      "learning_rate": 4.846686701400179e-05,
+      "loss": 1.3647,
+      "step": 900
+    },
+    {
+      "epoch": 0.6319480974925478,
+      "grad_norm": 45.04077911376953,
+      "learning_rate": 4.846186383429657e-05,
+      "loss": 1.366,
+      "step": 901
+    },
+    {
+      "epoch": 0.6326494827283886,
+      "grad_norm": 47.341819763183594,
+      "learning_rate": 4.8456852763363706e-05,
+      "loss": 1.2603,
+      "step": 902
+    },
+    {
+      "epoch": 0.6333508679642293,
+      "grad_norm": 44.485530853271484,
+      "learning_rate": 4.8451833802888625e-05,
+      "loss": 1.3248,
+      "step": 903
+    },
+    {
+      "epoch": 0.6340522532000702,
+      "grad_norm": 0.46667349338531494,
+      "learning_rate": 4.844680695455943e-05,
+      "loss": 0.0041,
+      "step": 904
+    },
+    {
+      "epoch": 0.6347536384359109,
+      "grad_norm": 0.7628422379493713,
+      "learning_rate": 4.844177222006687e-05,
+      "loss": 0.0067,
+      "step": 905
+    },
+    {
+      "epoch": 0.6354550236717517,
+      "grad_norm": 0.7451947927474976,
+      "learning_rate": 4.8436729601104327e-05,
+      "loss": 0.0069,
+      "step": 906
+    },
+    {
+      "epoch": 0.6361564089075925,
+      "grad_norm": 0.6644980907440186,
+      "learning_rate": 4.843167909936786e-05,
+      "loss": 0.006,
+      "step": 907
+    },
+    {
+      "epoch": 0.6368577941434332,
+      "grad_norm": 41.53312683105469,
+      "learning_rate": 4.8426620716556164e-05,
+      "loss": 1.1098,
+      "step": 908
+    },
+    {
+      "epoch": 0.6375591793792741,
+      "grad_norm": 0.5270825624465942,
+      "learning_rate": 4.8421554454370596e-05,
+      "loss": 0.0046,
+      "step": 909
+    },
+    {
+      "epoch": 0.6382605646151148,
+      "grad_norm": 90.92555236816406,
+      "learning_rate": 4.841648031451514e-05,
+      "loss": 2.5683,
+      "step": 910
+    },
+    {
+      "epoch": 0.6389619498509557,
+      "grad_norm": 46.597808837890625,
+      "learning_rate": 4.841139829869646e-05,
+      "loss": 1.4381,
+      "step": 911
+    },
+    {
+      "epoch": 0.6396633350867964,
+      "grad_norm": 0.9347837567329407,
+      "learning_rate": 4.8406308408623865e-05,
+      "loss": 0.0083,
+      "step": 912
+    },
+    {
+      "epoch": 0.6403647203226372,
+      "grad_norm": 0.8483829498291016,
+      "learning_rate": 4.840121064600928e-05,
+      "loss": 0.0077,
+      "step": 913
+    },
+    {
+      "epoch": 0.641066105558478,
+      "grad_norm": 37.393951416015625,
+      "learning_rate": 4.8396105012567315e-05,
+      "loss": 0.8573,
+      "step": 914
+    },
+    {
+      "epoch": 0.6417674907943188,
+      "grad_norm": 69.15557098388672,
+      "learning_rate": 4.839099151001521e-05,
+      "loss": 1.491,
+      "step": 915
+    },
+    {
+      "epoch": 0.6424688760301596,
+      "grad_norm": 1.3047503232955933,
+      "learning_rate": 4.838587014007284e-05,
+      "loss": 0.0126,
+      "step": 916
+    },
+    {
+      "epoch": 0.6431702612660003,
+      "grad_norm": 1.462365746498108,
+      "learning_rate": 4.838074090446276e-05,
+      "loss": 0.0144,
+      "step": 917
+    },
+    {
+      "epoch": 0.6438716465018411,
+      "grad_norm": 1.4955204725265503,
+      "learning_rate": 4.8375603804910154e-05,
+      "loss": 0.0146,
+      "step": 918
+    },
+    {
+      "epoch": 0.6445730317376819,
+      "grad_norm": 1.0593093633651733,
+      "learning_rate": 4.837045884314283e-05,
+      "loss": 0.0101,
+      "step": 919
+    },
+    {
+      "epoch": 0.6452744169735227,
+      "grad_norm": 0.5433834791183472,
+      "learning_rate": 4.836530602089127e-05,
+      "loss": 0.0049,
+      "step": 920
+    },
+    {
+      "epoch": 0.6459758022093635,
+      "grad_norm": 44.653175354003906,
+      "learning_rate": 4.836014533988857e-05,
+      "loss": 1.2462,
+      "step": 921
+    },
+    {
+      "epoch": 0.6466771874452043,
+      "grad_norm": 0.15624181926250458,
+      "learning_rate": 4.835497680187052e-05,
+      "loss": 0.0013,
+      "step": 922
+    },
+    {
+      "epoch": 0.647378572681045,
+      "grad_norm": 0.13333559036254883,
+      "learning_rate": 4.8349800408575506e-05,
+      "loss": 0.0011,
+      "step": 923
+    },
+    {
+      "epoch": 0.6480799579168859,
+      "grad_norm": 0.08697400987148285,
+      "learning_rate": 4.834461616174456e-05,
+      "loss": 0.0007,
+      "step": 924
+    },
+    {
+      "epoch": 0.6487813431527266,
+      "grad_norm": 49.62416076660156,
+      "learning_rate": 4.8339424063121375e-05,
+      "loss": 2.1264,
+      "step": 925
+    },
+    {
+      "epoch": 0.6494827283885675,
+      "grad_norm": 47.912906646728516,
+      "learning_rate": 4.833422411445228e-05,
+      "loss": 1.8048,
+      "step": 926
+    },
+    {
+      "epoch": 0.6501841136244082,
+      "grad_norm": 0.0816919356584549,
+      "learning_rate": 4.832901631748622e-05,
+      "loss": 0.0007,
+      "step": 927
+    },
+    {
+      "epoch": 0.650885498860249,
+      "grad_norm": 49.09192657470703,
+      "learning_rate": 4.8323800673974826e-05,
+      "loss": 1.7493,
+      "step": 928
+    },
+    {
+      "epoch": 0.6515868840960898,
+      "grad_norm": 98.90199279785156,
+      "learning_rate": 4.8318577185672325e-05,
+      "loss": 3.7461,
+      "step": 929
+    },
+    {
+      "epoch": 0.6522882693319305,
+      "grad_norm": 0.22995713353157043,
+      "learning_rate": 4.831334585433561e-05,
+      "loss": 0.002,
+      "step": 930
+    },
+    {
+      "epoch": 0.6529896545677714,
+      "grad_norm": 0.4566737413406372,
+      "learning_rate": 4.830810668172419e-05,
+      "loss": 0.0041,
+      "step": 931
+    },
+    {
+      "epoch": 0.6536910398036121,
+      "grad_norm": 0.5328819155693054,
+      "learning_rate": 4.830285966960023e-05,
+      "loss": 0.0048,
+      "step": 932
+    },
+    {
+      "epoch": 0.6543924250394529,
+      "grad_norm": 0.6611910462379456,
+      "learning_rate": 4.829760481972852e-05,
+      "loss": 0.0061,
+      "step": 933
+    },
+    {
+      "epoch": 0.6550938102752937,
+      "grad_norm": 76.24475860595703,
+      "learning_rate": 4.82923421338765e-05,
+      "loss": 1.7814,
+      "step": 934
+    },
+    {
+      "epoch": 0.6557951955111345,
+      "grad_norm": 0.7961150407791138,
+      "learning_rate": 4.828707161381423e-05,
+      "loss": 0.0075,
+      "step": 935
+    },
+    {
+      "epoch": 0.6564965807469753,
+      "grad_norm": 33.920326232910156,
+      "learning_rate": 4.8281793261314404e-05,
+      "loss": 0.7362,
+      "step": 936
+    },
+    {
+      "epoch": 0.6571979659828161,
+      "grad_norm": 0.9916706681251526,
+      "learning_rate": 4.827650707815236e-05,
+      "loss": 0.0096,
+      "step": 937
+    },
+    {
+      "epoch": 0.6578993512186568,
+      "grad_norm": 0.7823458909988403,
+      "learning_rate": 4.827121306610608e-05,
+      "loss": 0.0074,
+      "step": 938
+    },
+    {
+      "epoch": 0.6586007364544977,
+      "grad_norm": 0.8576016426086426,
+      "learning_rate": 4.8265911226956154e-05,
+      "loss": 0.008,
+      "step": 939
+    },
+    {
+      "epoch": 0.6593021216903384,
+      "grad_norm": 0.5242276787757874,
+      "learning_rate": 4.826060156248582e-05,
+      "loss": 0.0047,
+      "step": 940
+    },
+    {
+      "epoch": 0.6600035069261792,
+      "grad_norm": 0.31952106952667236,
+      "learning_rate": 4.825528407448096e-05,
+      "loss": 0.0028,
+      "step": 941
+    },
+    {
+      "epoch": 0.66070489216202,
+      "grad_norm": 0.17970554530620575,
+      "learning_rate": 4.824995876473004e-05,
+      "loss": 0.0015,
+      "step": 942
+    },
+    {
+      "epoch": 0.6614062773978607,
+      "grad_norm": 0.10021814703941345,
+      "learning_rate": 4.824462563502422e-05,
+      "loss": 0.0008,
+      "step": 943
+    },
+    {
+      "epoch": 0.6621076626337016,
+      "grad_norm": 146.59963989257812,
+      "learning_rate": 4.823928468715724e-05,
+      "loss": 5.6442,
+      "step": 944
+    },
+    {
+      "epoch": 0.6628090478695423,
+      "grad_norm": 0.060028087347745895,
+      "learning_rate": 4.82339359229255e-05,
+      "loss": 0.0005,
+      "step": 945
+    },
+    {
+      "epoch": 0.6635104331053832,
+      "grad_norm": 0.05350363254547119,
+      "learning_rate": 4.822857934412802e-05,
+      "loss": 0.0004,
+      "step": 946
+    },
+    {
+      "epoch": 0.6642118183412239,
+      "grad_norm": 197.06178283691406,
+      "learning_rate": 4.822321495256644e-05,
+      "loss": 8.2989,
+      "step": 947
+    },
+    {
+      "epoch": 0.6649132035770647,
+      "grad_norm": 0.047274697571992874,
+      "learning_rate": 4.821784275004503e-05,
+      "loss": 0.0004,
+      "step": 948
+    },
+    {
+      "epoch": 0.6656145888129055,
+      "grad_norm": 0.037716954946517944,
+      "learning_rate": 4.82124627383707e-05,
+      "loss": 0.0003,
+      "step": 949
+    },
+    {
+      "epoch": 0.6663159740487463,
+      "grad_norm": 48.574623107910156,
+      "learning_rate": 4.8207074919352976e-05,
+      "loss": 1.7823,
+      "step": 950
+    },
+    {
+      "epoch": 0.667017359284587,
+      "grad_norm": 0.06911084055900574,
+      "learning_rate": 4.8201679294804e-05,
+      "loss": 0.0006,
+      "step": 951
+    },
+    {
+      "epoch": 0.6677187445204279,
+      "grad_norm": 0.07155507057905197,
+      "learning_rate": 4.8196275866538574e-05,
+      "loss": 0.0006,
+      "step": 952
+    },
+    {
+      "epoch": 0.6684201297562686,
+      "grad_norm": 0.10516396909952164,
+      "learning_rate": 4.8190864636374086e-05,
+      "loss": 0.0009,
+      "step": 953
+    },
+    {
+      "epoch": 0.6691215149921094,
+      "grad_norm": 46.50364685058594,
+      "learning_rate": 4.818544560613056e-05,
+      "loss": 1.4989,
+      "step": 954
+    },
+    {
+      "epoch": 0.6698229002279502,
+      "grad_norm": 0.2198321670293808,
+      "learning_rate": 4.8180018777630664e-05,
+      "loss": 0.0019,
+      "step": 955
+    },
+    {
+      "epoch": 0.670524285463791,
+      "grad_norm": 87.00233459472656,
+      "learning_rate": 4.8174584152699665e-05,
+      "loss": 2.3952,
+      "step": 956
+    },
+    {
+      "epoch": 0.6712256706996318,
+      "grad_norm": 39.932960510253906,
+      "learning_rate": 4.816914173316545e-05,
+      "loss": 1.0167,
+      "step": 957
+    },
+    {
+      "epoch": 0.6719270559354725,
+      "grad_norm": 0.8517984747886658,
+      "learning_rate": 4.816369152085855e-05,
+      "loss": 0.0082,
+      "step": 958
+    },
+    {
+      "epoch": 0.6726284411713134,
+      "grad_norm": 1.0399830341339111,
+      "learning_rate": 4.8158233517612096e-05,
+      "loss": 0.0102,
+      "step": 959
+    },
+    {
+      "epoch": 0.6733298264071541,
+      "grad_norm": 1.2987735271453857,
+      "learning_rate": 4.815276772526186e-05,
+      "loss": 0.013,
+      "step": 960
+    },
+    {
+      "epoch": 0.674031211642995,
+      "grad_norm": 1.1722874641418457,
+      "learning_rate": 4.8147294145646207e-05,
+      "loss": 0.0119,
+      "step": 961
+    },
+    {
+      "epoch": 0.6747325968788357,
+      "grad_norm": 30.018028259277344,
+      "learning_rate": 4.8141812780606137e-05,
+      "loss": 0.606,
+      "step": 962
+    },
+    {
+      "epoch": 0.6754339821146765,
+      "grad_norm": 1.0533334016799927,
+      "learning_rate": 4.813632363198527e-05,
+      "loss": 0.0105,
+      "step": 963
+    },
+    {
+      "epoch": 0.6761353673505173,
+      "grad_norm": 52.103050231933594,
+      "learning_rate": 4.813082670162986e-05,
+      "loss": 1.5312,
+      "step": 964
+    },
+    {
+      "epoch": 0.676836752586358,
+      "grad_norm": 36.39682388305664,
+      "learning_rate": 4.812532199138873e-05,
+      "loss": 0.8569,
+      "step": 965
+    },
+    {
+      "epoch": 0.6775381378221988,
+      "grad_norm": 84.3869400024414,
+      "learning_rate": 4.8119809503113363e-05,
+      "loss": 2.1346,
+      "step": 966
+    },
+    {
+      "epoch": 0.6782395230580396,
+      "grad_norm": 0.8813936710357666,
+      "learning_rate": 4.811428923865785e-05,
+      "loss": 0.0085,
+      "step": 967
+    },
+    {
+      "epoch": 0.6789409082938804,
+      "grad_norm": 1.1427128314971924,
+      "learning_rate": 4.810876119987888e-05,
+      "loss": 0.0113,
+      "step": 968
+    },
+    {
+      "epoch": 0.6796422935297212,
+      "grad_norm": 30.400836944580078,
+      "learning_rate": 4.810322538863578e-05,
+      "loss": 0.6341,
+      "step": 969
+    },
+    {
+      "epoch": 0.680343678765562,
+      "grad_norm": 77.0494155883789,
+      "learning_rate": 4.8097681806790466e-05,
+      "loss": 1.7974,
+      "step": 970
+    },
+    {
+      "epoch": 0.6810450640014027,
+      "grad_norm": 29.669004440307617,
+      "learning_rate": 4.8092130456207495e-05,
+      "loss": 0.5951,
+      "step": 971
+    },
+    {
+      "epoch": 0.6817464492372436,
+      "grad_norm": 2.0304183959960938,
+      "learning_rate": 4.808657133875402e-05,
+      "loss": 0.0214,
+      "step": 972
+    },
+    {
+      "epoch": 0.6824478344730843,
+      "grad_norm": 22.323291778564453,
+      "learning_rate": 4.808100445629981e-05,
+      "loss": 0.4383,
+      "step": 973
+    },
+    {
+      "epoch": 0.6831492197089252,
+      "grad_norm": 11.758098602294922,
+      "learning_rate": 4.807542981071724e-05,
+      "loss": 0.245,
+      "step": 974
+    },
+    {
+      "epoch": 0.6838506049447659,
+      "grad_norm": 9.657792091369629,
+      "learning_rate": 4.80698474038813e-05,
+      "loss": 0.2276,
+      "step": 975
+    },
+    {
+      "epoch": 0.6845519901806068,
+      "grad_norm": 0.9010874629020691,
+      "learning_rate": 4.806425723766961e-05,
+      "loss": 0.1934,
+      "step": 976
+    },
+    {
+      "epoch": 0.6852533754164475,
+      "grad_norm": 13.536153793334961,
+      "learning_rate": 4.805865931396236e-05,
+      "loss": 0.2465,
+      "step": 977
+    },
+    {
+      "epoch": 0.6859547606522882,
+      "grad_norm": 12.555607795715332,
+      "learning_rate": 4.8053053634642384e-05,
+      "loss": 0.2135,
+      "step": 978
+    },
+    {
+      "epoch": 0.6866561458881291,
+      "grad_norm": 15.943707466125488,
+      "learning_rate": 4.804744020159511e-05,
+      "loss": 0.3215,
+      "step": 979
+    },
+    {
+      "epoch": 0.6873575311239698,
+      "grad_norm": 9.25063419342041,
+      "learning_rate": 4.804181901670856e-05,
+      "loss": 0.2798,
+      "step": 980
+    },
+    {
+      "epoch": 0.6880589163598106,
+      "grad_norm": 8.65345573425293,
+      "learning_rate": 4.80361900818734e-05,
+      "loss": 0.1354,
+      "step": 981
+    },
+    {
+      "epoch": 0.6887603015956514,
+      "grad_norm": 10.924114227294922,
+      "learning_rate": 4.803055339898288e-05,
+      "loss": 0.1829,
+      "step": 982
+    },
+    {
+      "epoch": 0.6894616868314922,
+      "grad_norm": 7.537076950073242,
+      "learning_rate": 4.8024908969932844e-05,
+      "loss": 0.1064,
+      "step": 983
+    },
+    {
+      "epoch": 0.690163072067333,
+      "grad_norm": 43.258338928222656,
+      "learning_rate": 4.8019256796621764e-05,
+      "loss": 0.957,
+      "step": 984
+    },
+    {
+      "epoch": 0.6908644573031738,
+      "grad_norm": 26.308055877685547,
+      "learning_rate": 4.801359688095071e-05,
+      "loss": 0.541,
+      "step": 985
+    },
+    {
+      "epoch": 0.6915658425390145,
+      "grad_norm": 1.8512808084487915,
+      "learning_rate": 4.800792922482333e-05,
+      "loss": 0.0192,
+      "step": 986
+    },
+    {
+      "epoch": 0.6922672277748554,
+      "grad_norm": 1.0505925416946411,
+      "learning_rate": 4.800225383014594e-05,
+      "loss": 0.0104,
+      "step": 987
+    },
+    {
+      "epoch": 0.6929686130106961,
+      "grad_norm": 40.21651840209961,
+      "learning_rate": 4.799657069882738e-05,
+      "loss": 1.0056,
+      "step": 988
+    },
+    {
+      "epoch": 0.6936699982465369,
+      "grad_norm": 0.17625601589679718,
+      "learning_rate": 4.7990879832779156e-05,
+      "loss": 0.0015,
+      "step": 989
+    },
+    {
+      "epoch": 0.6943713834823777,
+      "grad_norm": 0.15104326605796814,
+      "learning_rate": 4.7985181233915335e-05,
+      "loss": 0.0013,
+      "step": 990
+    },
+    {
+      "epoch": 0.6950727687182184,
+      "grad_norm": 0.09187794476747513,
+      "learning_rate": 4.79794749041526e-05,
+      "loss": 0.0008,
+      "step": 991
+    },
+    {
+      "epoch": 0.6957741539540593,
+      "grad_norm": 48.85366439819336,
+      "learning_rate": 4.797376084541024e-05,
+      "loss": 1.9766,
+      "step": 992
+    },
+    {
+      "epoch": 0.6964755391899,
+      "grad_norm": 0.04326195642352104,
+      "learning_rate": 4.796803905961013e-05,
+      "loss": 0.0004,
+      "step": 993
+    },
+    {
+      "epoch": 0.6971769244257409,
+      "grad_norm": 0.04548042640089989,
+      "learning_rate": 4.796230954867676e-05,
+      "loss": 0.0004,
+      "step": 994
+    },
+    {
+      "epoch": 0.6978783096615816,
+      "grad_norm": 48.174827575683594,
+      "learning_rate": 4.79565723145372e-05,
+      "loss": 1.9204,
+      "step": 995
+    },
+    {
+      "epoch": 0.6985796948974224,
+      "grad_norm": 0.04808010160923004,
+      "learning_rate": 4.7950827359121135e-05,
+      "loss": 0.0004,
+      "step": 996
+    },
+    {
+      "epoch": 0.6992810801332632,
+      "grad_norm": 48.49928665161133,
+      "learning_rate": 4.794507468436084e-05,
+      "loss": 1.8132,
+      "step": 997
+    },
+    {
+      "epoch": 0.699982465369104,
+      "grad_norm": 47.405731201171875,
+      "learning_rate": 4.793931429219117e-05,
+      "loss": 1.6124,
+      "step": 998
+    },
+    {
+      "epoch": 0.7006838506049448,
+      "grad_norm": 0.2506006062030792,
+      "learning_rate": 4.793354618454961e-05,
+      "loss": 0.0022,
+      "step": 999
+    },
+    {
+      "epoch": 0.7013852358407856,
+      "grad_norm": 90.9857177734375,
+      "learning_rate": 4.792777036337621e-05,
+      "loss": 2.7266,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7020866210766263,
+      "grad_norm": 0.9927763938903809,
+      "learning_rate": 4.792198683061363e-05,
+      "loss": 0.0099,
+      "step": 1001
+    },
+    {
+      "epoch": 0.7027880063124671,
+      "grad_norm": 1.5031591653823853,
+      "learning_rate": 4.791619558820712e-05,
+      "loss": 0.0152,
+      "step": 1002
+    },
+    {
+      "epoch": 0.7034893915483079,
+      "grad_norm": 29.284543991088867,
+      "learning_rate": 4.791039663810452e-05,
+      "loss": 0.5838,
+      "step": 1003
+    },
+    {
+      "epoch": 0.7041907767841487,
+      "grad_norm": 1.6648647785186768,
+      "learning_rate": 4.790458998225627e-05,
+      "loss": 0.0173,
+      "step": 1004
+    },
+    {
+      "epoch": 0.7048921620199895,
+      "grad_norm": 2.1368696689605713,
+      "learning_rate": 4.7898775622615384e-05,
+      "loss": 0.023,
+      "step": 1005
+    },
+    {
+      "epoch": 0.7055935472558302,
+      "grad_norm": 1.5531522035598755,
+      "learning_rate": 4.789295356113749e-05,
+      "loss": 0.0161,
+      "step": 1006
+    },
+    {
+      "epoch": 0.7062949324916711,
+      "grad_norm": 1.0346095561981201,
+      "learning_rate": 4.7887123799780795e-05,
+      "loss": 0.0096,
+      "step": 1007
+    },
+    {
+      "epoch": 0.7069963177275118,
+      "grad_norm": 0.5431362986564636,
+      "learning_rate": 4.788128634050609e-05,
+      "loss": 0.005,
+      "step": 1008
+    },
+    {
+      "epoch": 0.7076977029633527,
+      "grad_norm": 0.19359983503818512,
+      "learning_rate": 4.787544118527677e-05,
+      "loss": 0.0017,
+      "step": 1009
+    },
+    {
+      "epoch": 0.7083990881991934,
+      "grad_norm": 49.78437042236328,
+      "learning_rate": 4.786958833605882e-05,
+      "loss": 1.7876,
+      "step": 1010
+    },
+    {
+      "epoch": 0.7091004734350342,
+      "grad_norm": 57.95826721191406,
+      "learning_rate": 4.786372779482078e-05,
+      "loss": 2.9286,
+      "step": 1011
+    },
+    {
+      "epoch": 0.709801858670875,
+      "grad_norm": 48.60955047607422,
+      "learning_rate": 4.7857859563533814e-05,
+      "loss": 1.9456,
+      "step": 1012
+    },
+    {
+      "epoch": 0.7105032439067157,
+      "grad_norm": 48.62706756591797,
+      "learning_rate": 4.785198364417166e-05,
+      "loss": 1.8589,
+      "step": 1013
+    },
+    {
+      "epoch": 0.7112046291425566,
+      "grad_norm": 94.03495788574219,
+      "learning_rate": 4.784610003871063e-05,
+      "loss": 3.2697,
+      "step": 1014
+    },
+    {
+      "epoch": 0.7119060143783973,
+      "grad_norm": 44.21165084838867,
+      "learning_rate": 4.784020874912965e-05,
+      "loss": 1.3436,
+      "step": 1015
+    },
+    {
+      "epoch": 0.7126073996142381,
+      "grad_norm": 0.36148276925086975,
+      "learning_rate": 4.7834309777410195e-05,
+      "loss": 0.0033,
+      "step": 1016
+    },
+    {
+      "epoch": 0.7133087848500789,
+      "grad_norm": 74.5863265991211,
+      "learning_rate": 4.7828403125536354e-05,
+      "loss": 1.8025,
+      "step": 1017
+    },
+    {
+      "epoch": 0.7140101700859197,
+      "grad_norm": 43.957252502441406,
+      "learning_rate": 4.782248879549478e-05,
+      "loss": 1.1635,
+      "step": 1018
+    },
+    {
+      "epoch": 0.7147115553217604,
+      "grad_norm": 69.43833923339844,
+      "learning_rate": 4.781656678927471e-05,
+      "loss": 1.2,
+      "step": 1019
+    },
+    {
+      "epoch": 0.7154129405576013,
+      "grad_norm": 4.946931838989258,
+      "learning_rate": 4.7810637108867986e-05,
+      "loss": 0.0617,
+      "step": 1020
+    },
+    {
+      "epoch": 0.716114325793442,
+      "grad_norm": 2.806675434112549,
+      "learning_rate": 4.7804699756268987e-05,
+      "loss": 0.1936,
+      "step": 1021
+    },
+    {
+      "epoch": 0.7168157110292829,
+      "grad_norm": 10.467972755432129,
+      "learning_rate": 4.7798754733474724e-05,
+      "loss": 0.17,
+      "step": 1022
+    },
+    {
+      "epoch": 0.7175170962651236,
+      "grad_norm": 13.816707611083984,
+      "learning_rate": 4.7792802042484746e-05,
+      "loss": 0.2481,
+      "step": 1023
+    },
+    {
+      "epoch": 0.7182184815009645,
+      "grad_norm": 14.67568588256836,
+      "learning_rate": 4.77868416853012e-05,
+      "loss": 0.2728,
+      "step": 1024
+    },
+    {
+      "epoch": 0.7189198667368052,
+      "grad_norm": 3.8597590923309326,
+      "learning_rate": 4.778087366392881e-05,
+      "loss": 0.3473,
+      "step": 1025
+    },
+    {
+      "epoch": 0.7196212519726459,
+      "grad_norm": 6.207676410675049,
+      "learning_rate": 4.777489798037489e-05,
+      "loss": 0.276,
+      "step": 1026
+    },
+    {
+      "epoch": 0.7203226372084868,
+      "grad_norm": 1.4748679399490356,
+      "learning_rate": 4.77689146366493e-05,
+      "loss": 0.2487,
+      "step": 1027
+    },
+    {
+      "epoch": 0.7210240224443275,
+      "grad_norm": 3.2220802307128906,
+      "learning_rate": 4.77629236347645e-05,
+      "loss": 0.1941,
+      "step": 1028
+    },
+    {
+      "epoch": 0.7217254076801684,
+      "grad_norm": 8.897248268127441,
+      "learning_rate": 4.775692497673552e-05,
+      "loss": 0.135,
+      "step": 1029
+    },
+    {
+      "epoch": 0.7224267929160091,
+      "grad_norm": 6.12791633605957,
+      "learning_rate": 4.7750918664579956e-05,
+      "loss": 0.0784,
+      "step": 1030
+    },
+    {
+      "epoch": 0.7231281781518499,
+      "grad_norm": 20.013917922973633,
+      "learning_rate": 4.7744904700318006e-05,
+      "loss": 0.3768,
+      "step": 1031
+    },
+    {
+      "epoch": 0.7238295633876907,
+      "grad_norm": 2.475820541381836,
+      "learning_rate": 4.7738883085972406e-05,
+      "loss": 0.0274,
+      "step": 1032
+    },
+    {
+      "epoch": 0.7245309486235315,
+      "grad_norm": 1.4534077644348145,
+      "learning_rate": 4.773285382356849e-05,
+      "loss": 0.015,
+      "step": 1033
+    },
+    {
+      "epoch": 0.7252323338593722,
+      "grad_norm": 36.22199630737305,
+      "learning_rate": 4.772681691513417e-05,
+      "loss": 0.9046,
+      "step": 1034
+    },
+    {
+      "epoch": 0.7259337190952131,
+      "grad_norm": 0.2810744643211365,
+      "learning_rate": 4.772077236269989e-05,
+      "loss": 0.0025,
+      "step": 1035
+    },
+    {
+      "epoch": 0.7266351043310538,
+      "grad_norm": 45.44533920288086,
+      "learning_rate": 4.771472016829871e-05,
+      "loss": 1.3375,
+      "step": 1036
+    },
+    {
+      "epoch": 0.7273364895668946,
+      "grad_norm": 0.12535035610198975,
+      "learning_rate": 4.770866033396623e-05,
+      "loss": 0.0011,
+      "step": 1037
+    },
+    {
+      "epoch": 0.7280378748027354,
+      "grad_norm": 45.64175033569336,
+      "learning_rate": 4.770259286174065e-05,
+      "loss": 1.6729,
+      "step": 1038
+    },
+    {
+      "epoch": 0.7287392600385761,
+      "grad_norm": 46.727394104003906,
+      "learning_rate": 4.7696517753662696e-05,
+      "loss": 1.7577,
+      "step": 1039
+    },
+    {
+      "epoch": 0.729440645274417,
+      "grad_norm": 0.17975449562072754,
+      "learning_rate": 4.76904350117757e-05,
+      "loss": 0.0016,
+      "step": 1040
+    },
+    {
+      "epoch": 0.7301420305102577,
+      "grad_norm": 85.70226287841797,
+      "learning_rate": 4.7684344638125546e-05,
+      "loss": 2.662,
+      "step": 1041
+    },
+    {
+      "epoch": 0.7308434157460986,
+      "grad_norm": 39.62570571899414,
+      "learning_rate": 4.767824663476069e-05,
+      "loss": 1.1086,
+      "step": 1042
+    },
+    {
+      "epoch": 0.7315448009819393,
+      "grad_norm": 33.90923309326172,
+      "learning_rate": 4.767214100373214e-05,
+      "loss": 0.8021,
+      "step": 1043
+    },
+    {
+      "epoch": 0.7322461862177801,
+      "grad_norm": 23.501230239868164,
+      "learning_rate": 4.76660277470935e-05,
+      "loss": 0.4899,
+      "step": 1044
+    },
+    {
+      "epoch": 0.7329475714536209,
+      "grad_norm": 10.125417709350586,
+      "learning_rate": 4.76599068669009e-05,
+      "loss": 0.2671,
+      "step": 1045
+    },
+    {
+      "epoch": 0.7336489566894617,
+      "grad_norm": 3.0318775177001953,
+      "learning_rate": 4.7653778365213076e-05,
+      "loss": 0.2053,
+      "step": 1046
+    },
+    {
+      "epoch": 0.7343503419253025,
+      "grad_norm": 7.598801612854004,
+      "learning_rate": 4.7647642244091276e-05,
+      "loss": 0.2937,
+      "step": 1047
+    },
+    {
+      "epoch": 0.7350517271611433,
+      "grad_norm": 11.337316513061523,
+      "learning_rate": 4.7641498505599356e-05,
+      "loss": 0.283,
+      "step": 1048
+    },
+    {
+      "epoch": 0.735753112396984,
+      "grad_norm": 2.909923553466797,
+      "learning_rate": 4.763534715180373e-05,
+      "loss": 0.2887,
+      "step": 1049
+    },
+    {
+      "epoch": 0.7364544976328248,
+      "grad_norm": 12.51762866973877,
+      "learning_rate": 4.762918818477334e-05,
+      "loss": 0.2306,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7371558828686656,
+      "grad_norm": 10.550107955932617,
+      "learning_rate": 4.762302160657971e-05,
+      "loss": 0.1879,
+      "step": 1051
+    },
+    {
+      "epoch": 0.7378572681045064,
+      "grad_norm": 10.876016616821289,
+      "learning_rate": 4.7616847419296945e-05,
+      "loss": 0.1877,
+      "step": 1052
+    },
+    {
+      "epoch": 0.7385586533403472,
+      "grad_norm": 3.0670106410980225,
+      "learning_rate": 4.761066562500167e-05,
+      "loss": 0.2327,
+      "step": 1053
+    },
+    {
+      "epoch": 0.7392600385761879,
+      "grad_norm": 5.0195770263671875,
+      "learning_rate": 4.7604476225773084e-05,
+      "loss": 0.0651,
+      "step": 1054
+    },
+    {
+      "epoch": 0.7399614238120288,
+      "grad_norm": 2.679926872253418,
+      "learning_rate": 4.759827922369296e-05,
+      "loss": 0.0317,
+      "step": 1055
+    },
+    {
+      "epoch": 0.7406628090478695,
+      "grad_norm": 1.752316951751709,
+      "learning_rate": 4.7592074620845616e-05,
+      "loss": 0.0184,
+      "step": 1056
+    },
+    {
+      "epoch": 0.7413641942837104,
+      "grad_norm": 0.6544418334960938,
+      "learning_rate": 4.758586241931791e-05,
+      "loss": 0.0062,
+      "step": 1057
+    },
+    {
+      "epoch": 0.7420655795195511,
+      "grad_norm": 0.19106708467006683,
+      "learning_rate": 4.757964262119928e-05,
+      "loss": 0.0017,
+      "step": 1058
+    },
+    {
+      "epoch": 0.742766964755392,
+      "grad_norm": 0.0741598978638649,
+      "learning_rate": 4.757341522858171e-05,
+      "loss": 0.0006,
+      "step": 1059
+    },
+    {
+      "epoch": 0.7434683499912327,
+      "grad_norm": 0.02510019950568676,
+      "learning_rate": 4.7567180243559736e-05,
+      "loss": 0.0002,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7441697352270735,
+      "grad_norm": 0.00887396652251482,
+      "learning_rate": 4.7560937668230454e-05,
+      "loss": 0.0001,
+      "step": 1061
+    },
+    {
+      "epoch": 0.7448711204629143,
+      "grad_norm": 0.0027078278362751007,
+      "learning_rate": 4.75546875046935e-05,
+      "loss": 0.0,
+      "step": 1062
+    },
+    {
+      "epoch": 0.745572505698755,
+      "grad_norm": 0.001226524356752634,
+      "learning_rate": 4.7548429755051085e-05,
+      "loss": 0.0,
+      "step": 1063
+    },
+    {
+      "epoch": 0.7462738909345958,
+      "grad_norm": 0.0005389907164499164,
+      "learning_rate": 4.754216442140794e-05,
+      "loss": 0.0,
+      "step": 1064
+    },
+    {
+      "epoch": 0.7469752761704366,
+      "grad_norm": 55.77132797241211,
+      "learning_rate": 4.7535891505871375e-05,
+      "loss": 3.9293,
+      "step": 1065
+    },
+    {
+      "epoch": 0.7476766614062774,
+      "grad_norm": 103.8635025024414,
+      "learning_rate": 4.752961101055124e-05,
+      "loss": 7.9354,
+      "step": 1066
+    },
+    {
+      "epoch": 0.7483780466421182,
+      "grad_norm": 103.83565521240234,
+      "learning_rate": 4.7523322937559946e-05,
+      "loss": 7.4002,
+      "step": 1067
+    },
+    {
+      "epoch": 0.749079431877959,
+      "grad_norm": 0.0008341918000951409,
+      "learning_rate": 4.751702728901242e-05,
+      "loss": 0.0,
+      "step": 1068
+    },
+    {
+      "epoch": 0.7497808171137997,
+      "grad_norm": 48.793819427490234,
+      "learning_rate": 4.7510724067026156e-05,
+      "loss": 3.1709,
+      "step": 1069
+    },
+    {
+      "epoch": 0.7504822023496406,
+      "grad_norm": 97.3034439086914,
+      "learning_rate": 4.750441327372122e-05,
+      "loss": 5.7,
+      "step": 1070
+    },
+    {
+      "epoch": 0.7511835875854813,
+      "grad_norm": 0.0208453219383955,
+      "learning_rate": 4.7498094911220185e-05,
+      "loss": 0.0002,
+      "step": 1071
+    },
+    {
+      "epoch": 0.7518849728213222,
+      "grad_norm": 0.05400345101952553,
+      "learning_rate": 4.749176898164818e-05,
+      "loss": 0.0004,
+      "step": 1072
+    },
+    {
+      "epoch": 0.7525863580571629,
+      "grad_norm": 0.0927615538239479,
+      "learning_rate": 4.7485435487132904e-05,
+      "loss": 0.0008,
+      "step": 1073
+    },
+    {
+      "epoch": 0.7532877432930036,
+      "grad_norm": 0.19018051028251648,
+      "learning_rate": 4.747909442980456e-05,
+      "loss": 0.0016,
+      "step": 1074
+    },
+    {
+      "epoch": 0.7539891285288445,
+      "grad_norm": 0.3820917010307312,
+      "learning_rate": 4.7472745811795936e-05,
+      "loss": 0.0035,
+      "step": 1075
+    },
+    {
+      "epoch": 0.7546905137646852,
+      "grad_norm": 0.6652111411094666,
+      "learning_rate": 4.746638963524235e-05,
+      "loss": 0.0063,
+      "step": 1076
+    },
+    {
+      "epoch": 0.7553918990005261,
+      "grad_norm": 91.62269592285156,
+      "learning_rate": 4.7460025902281625e-05,
+      "loss": 2.2848,
+      "step": 1077
+    },
+    {
+      "epoch": 0.7560932842363668,
+      "grad_norm": 0.6401788592338562,
+      "learning_rate": 4.7453654615054174e-05,
+      "loss": 0.006,
+      "step": 1078
+    },
+    {
+      "epoch": 0.7567946694722076,
+      "grad_norm": 0.9835159182548523,
+      "learning_rate": 4.744727577570295e-05,
+      "loss": 0.0096,
+      "step": 1079
+    },
+    {
+      "epoch": 0.7574960547080484,
+      "grad_norm": 0.665366530418396,
+      "learning_rate": 4.74408893863734e-05,
+      "loss": 0.0064,
+      "step": 1080
+    },
+    {
+      "epoch": 0.7581974399438892,
+      "grad_norm": 0.3792603313922882,
+      "learning_rate": 4.743449544921356e-05,
+      "loss": 0.0034,
+      "step": 1081
+    },
+    {
+      "epoch": 0.75889882517973,
+      "grad_norm": 0.18616673350334167,
+      "learning_rate": 4.742809396637397e-05,
+      "loss": 0.0016,
+      "step": 1082
+    },
+    {
+      "epoch": 0.7596002104155708,
+      "grad_norm": 0.12753531336784363,
+      "learning_rate": 4.7421684940007725e-05,
+      "loss": 0.0011,
+      "step": 1083
+    },
+    {
+      "epoch": 0.7603015956514115,
+      "grad_norm": 92.24988555908203,
+      "learning_rate": 4.741526837227047e-05,
+      "loss": 3.521,
+      "step": 1084
+    },
+    {
+      "epoch": 0.7610029808872524,
+      "grad_norm": 48.01845169067383,
+      "learning_rate": 4.740884426532035e-05,
+      "loss": 1.9461,
+      "step": 1085
+    },
+    {
+      "epoch": 0.7617043661230931,
+      "grad_norm": 0.07205016911029816,
+      "learning_rate": 4.740241262131808e-05,
+      "loss": 0.0006,
+      "step": 1086
+    },
+    {
+      "epoch": 0.7624057513589338,
+      "grad_norm": 0.07788128405809402,
+      "learning_rate": 4.7395973442426895e-05,
+      "loss": 0.0007,
+      "step": 1087
+    },
+    {
+      "epoch": 0.7631071365947747,
+      "grad_norm": 94.46582794189453,
+      "learning_rate": 4.738952673081256e-05,
+      "loss": 3.4385,
+      "step": 1088
+    },
+    {
+      "epoch": 0.7638085218306154,
+      "grad_norm": 0.17950232326984406,
+      "learning_rate": 4.738307248864339e-05,
+      "loss": 0.0016,
+      "step": 1089
+    },
+    {
+      "epoch": 0.7645099070664563,
+      "grad_norm": 93.51290130615234,
+      "learning_rate": 4.737661071809021e-05,
+      "loss": 3.3519,
+      "step": 1090
+    },
+    {
+      "epoch": 0.765211292302297,
+      "grad_norm": 92.91374206542969,
+      "learning_rate": 4.73701414213264e-05,
+      "loss": 2.7943,
+      "step": 1091
+    },
+    {
+      "epoch": 0.7659126775381379,
+      "grad_norm": 0.5978343486785889,
+      "learning_rate": 4.736366460052786e-05,
+      "loss": 0.0056,
+      "step": 1092
+    },
+    {
+      "epoch": 0.7666140627739786,
+      "grad_norm": 30.616474151611328,
+      "learning_rate": 4.7357180257873014e-05,
+      "loss": 0.6588,
+      "step": 1093
+    },
+    {
+      "epoch": 0.7673154480098194,
+      "grad_norm": 1.5538511276245117,
+      "learning_rate": 4.7350688395542825e-05,
+      "loss": 0.0162,
+      "step": 1094
+    },
+    {
+      "epoch": 0.7680168332456602,
+      "grad_norm": 75.00251007080078,
+      "learning_rate": 4.734418901572079e-05,
+      "loss": 1.4034,
+      "step": 1095
+    },
+    {
+      "epoch": 0.768718218481501,
+      "grad_norm": 3.7650375366210938,
+      "learning_rate": 4.733768212059293e-05,
+      "loss": 0.0454,
+      "step": 1096
+    },
+    {
+      "epoch": 0.7694196037173417,
+      "grad_norm": 5.952811241149902,
+      "learning_rate": 4.733116771234778e-05,
+      "loss": 0.0766,
+      "step": 1097
+    },
+    {
+      "epoch": 0.7701209889531825,
+      "grad_norm": 15.600150108337402,
+      "learning_rate": 4.7324645793176425e-05,
+      "loss": 0.3215,
+      "step": 1098
+    },
+    {
+      "epoch": 0.7708223741890233,
+      "grad_norm": 9.213759422302246,
+      "learning_rate": 4.731811636527246e-05,
+      "loss": 0.2548,
+      "step": 1099
+    },
+    {
+      "epoch": 0.7715237594248641,
+      "grad_norm": 7.49055814743042,
+      "learning_rate": 4.731157943083202e-05,
+      "loss": 0.1,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7722251446607049,
+      "grad_norm": 7.580368518829346,
+      "learning_rate": 4.730503499205373e-05,
+      "loss": 0.1067,
+      "step": 1101
+    },
+    {
+      "epoch": 0.7729265298965456,
+      "grad_norm": 6.691344261169434,
+      "learning_rate": 4.7298483051138796e-05,
+      "loss": 0.0911,
+      "step": 1102
+    },
+    {
+      "epoch": 0.7736279151323865,
+      "grad_norm": 39.10125732421875,
+      "learning_rate": 4.72919236102909e-05,
+      "loss": 0.7558,
+      "step": 1103
+    },
+    {
+      "epoch": 0.7743293003682272,
+      "grad_norm": 4.285506725311279,
+      "learning_rate": 4.728535667171626e-05,
+      "loss": 0.051,
+      "step": 1104
+    },
+    {
+      "epoch": 0.7750306856040681,
+      "grad_norm": 2.88086199760437,
+      "learning_rate": 4.727878223762363e-05,
+      "loss": 0.0324,
+      "step": 1105
+    },
+    {
+      "epoch": 0.7757320708399088,
+      "grad_norm": 1.6831408739089966,
+      "learning_rate": 4.7272200310224256e-05,
+      "loss": 0.0172,
+      "step": 1106
+    },
+    {
+      "epoch": 0.7764334560757497,
+      "grad_norm": 37.848419189453125,
+      "learning_rate": 4.726561089173194e-05,
+      "loss": 0.934,
+      "step": 1107
+    },
+    {
+      "epoch": 0.7771348413115904,
+      "grad_norm": 38.1249885559082,
+      "learning_rate": 4.725901398436297e-05,
+      "loss": 0.9811,
+      "step": 1108
+    },
+    {
+      "epoch": 0.7778362265474312,
+      "grad_norm": 51.93446731567383,
+      "learning_rate": 4.725240959033618e-05,
+      "loss": 1.5657,
+      "step": 1109
+    },
+    {
+      "epoch": 0.778537611783272,
+      "grad_norm": 0.33258679509162903,
+      "learning_rate": 4.7245797711872895e-05,
+      "loss": 0.003,
+      "step": 1110
+    },
+    {
+      "epoch": 0.7792389970191127,
+      "grad_norm": 61.720924377441406,
+      "learning_rate": 4.723917835119699e-05,
+      "loss": 1.6662,
+      "step": 1111
+    },
+    {
+      "epoch": 0.7799403822549535,
+      "grad_norm": 37.942874908447266,
+      "learning_rate": 4.723255151053483e-05,
+      "loss": 1.0046,
+      "step": 1112
+    },
+    {
+      "epoch": 0.7806417674907943,
+      "grad_norm": 0.8135581016540527,
+      "learning_rate": 4.7225917192115297e-05,
+      "loss": 0.0076,
+      "step": 1113
+    },
+    {
+      "epoch": 0.7813431527266351,
+      "grad_norm": 36.66408920288086,
+      "learning_rate": 4.721927539816981e-05,
+      "loss": 0.9099,
+      "step": 1114
+    },
+    {
+      "epoch": 0.7820445379624759,
+      "grad_norm": 1.1302094459533691,
+      "learning_rate": 4.721262613093228e-05,
+      "loss": 0.0115,
+      "step": 1115
+    },
+    {
+      "epoch": 0.7827459231983167,
+      "grad_norm": 1.4476618766784668,
+      "learning_rate": 4.720596939263914e-05,
+      "loss": 0.0146,
+      "step": 1116
+    },
+    {
+      "epoch": 0.7834473084341574,
+      "grad_norm": 1.0591875314712524,
+      "learning_rate": 4.719930518552934e-05,
+      "loss": 0.0106,
+      "step": 1117
+    },
+    {
+      "epoch": 0.7841486936699983,
+      "grad_norm": 0.7351242899894714,
+      "learning_rate": 4.7192633511844334e-05,
+      "loss": 0.0069,
+      "step": 1118
+    },
+    {
+      "epoch": 0.784850078905839,
+      "grad_norm": 0.43580272793769836,
+      "learning_rate": 4.71859543738281e-05,
+      "loss": 0.004,
+      "step": 1119
+    },
+    {
+      "epoch": 0.7855514641416799,
+      "grad_norm": 0.256906658411026,
+      "learning_rate": 4.71792677737271e-05,
+      "loss": 0.0022,
+      "step": 1120
+    },
+    {
+      "epoch": 0.7862528493775206,
+      "grad_norm": 0.13365890085697174,
+      "learning_rate": 4.717257371379034e-05,
+      "loss": 0.0012,
+      "step": 1121
+    },
+    {
+      "epoch": 0.7869542346133613,
+      "grad_norm": 0.07990420609712601,
+      "learning_rate": 4.71658721962693e-05,
+      "loss": 0.0007,
+      "step": 1122
+    },
+    {
+      "epoch": 0.7876556198492022,
+      "grad_norm": 0.0479491651058197,
+      "learning_rate": 4.715916322341801e-05,
+      "loss": 0.0004,
+      "step": 1123
+    },
+    {
+      "epoch": 0.7883570050850429,
+      "grad_norm": 47.28627014160156,
+      "learning_rate": 4.7152446797492963e-05,
+      "loss": 2.0982,
+      "step": 1124
+    },
+    {
+      "epoch": 0.7890583903208838,
+      "grad_norm": 96.63157653808594,
+      "learning_rate": 4.71457229207532e-05,
+      "loss": 4.5257,
+      "step": 1125
+    },
+    {
+      "epoch": 0.7897597755567245,
+      "grad_norm": 47.81488800048828,
+      "learning_rate": 4.713899159546024e-05,
+      "loss": 2.3268,
+      "step": 1126
+    },
+    {
+      "epoch": 0.7904611607925653,
+      "grad_norm": 47.34000015258789,
+      "learning_rate": 4.713225282387811e-05,
+      "loss": 2.1484,
+      "step": 1127
+    },
+    {
+      "epoch": 0.7911625460284061,
+      "grad_norm": 46.76436996459961,
+      "learning_rate": 4.7125506608273344e-05,
+      "loss": 1.8822,
+      "step": 1128
+    },
+    {
+      "epoch": 0.7918639312642469,
+      "grad_norm": 0.12850354611873627,
+      "learning_rate": 4.711875295091499e-05,
+      "loss": 0.0011,
+      "step": 1129
+    },
+    {
+      "epoch": 0.7925653165000877,
+      "grad_norm": 126.00606536865234,
+      "learning_rate": 4.71119918540746e-05,
+      "loss": 3.7093,
+      "step": 1130
+    },
+    {
+      "epoch": 0.7932667017359285,
+      "grad_norm": 0.4526232182979584,
+      "learning_rate": 4.7105223320026205e-05,
+      "loss": 0.0042,
+      "step": 1131
+    },
+    {
+      "epoch": 0.7939680869717692,
+      "grad_norm": 33.449928283691406,
+      "learning_rate": 4.709844735104636e-05,
+      "loss": 0.7578,
+      "step": 1132
+    },
+    {
+      "epoch": 0.7946694722076101,
+      "grad_norm": 1.6459629535675049,
+      "learning_rate": 4.7091663949414105e-05,
+      "loss": 0.0174,
+      "step": 1133
+    },
+    {
+      "epoch": 0.7953708574434508,
+      "grad_norm": 34.26624298095703,
+      "learning_rate": 4.7084873117411e-05,
+      "loss": 0.7867,
+      "step": 1134
+    },
+    {
+      "epoch": 0.7960722426792916,
+      "grad_norm": 3.238154888153076,
+      "learning_rate": 4.7078074857321085e-05,
+      "loss": 0.0382,
+      "step": 1135
+    },
+    {
+      "epoch": 0.7967736279151324,
+      "grad_norm": 3.573115348815918,
+      "learning_rate": 4.70712691714309e-05,
+      "loss": 0.0431,
+      "step": 1136
+    },
+    {
+      "epoch": 0.7974750131509731,
+      "grad_norm": 4.002458095550537,
+      "learning_rate": 4.7064456062029506e-05,
+      "loss": 0.0488,
+      "step": 1137
+    },
+    {
+      "epoch": 0.798176398386814,
+      "grad_norm": 3.4627814292907715,
+      "learning_rate": 4.7057635531408424e-05,
+      "loss": 0.0394,
+      "step": 1138
+    },
+    {
+      "epoch": 0.7988777836226547,
+      "grad_norm": 18.19850730895996,
+      "learning_rate": 4.70508075818617e-05,
+      "loss": 0.3679,
+      "step": 1139
+    },
+    {
+      "epoch": 0.7995791688584956,
+      "grad_norm": 2.7357804775238037,
+      "learning_rate": 4.704397221568586e-05,
+      "loss": 0.0312,
+      "step": 1140
+    },
+    {
+      "epoch": 0.8002805540943363,
+      "grad_norm": 51.47498321533203,
+      "learning_rate": 4.703712943517993e-05,
+      "loss": 1.0259,
+      "step": 1141
+    },
+    {
+      "epoch": 0.8009819393301771,
+      "grad_norm": 1.7935067415237427,
+      "learning_rate": 4.7030279242645434e-05,
+      "loss": 0.0192,
+      "step": 1142
+    },
+    {
+      "epoch": 0.8016833245660179,
+      "grad_norm": 28.79102325439453,
+      "learning_rate": 4.702342164038639e-05,
+      "loss": 0.6393,
+      "step": 1143
+    },
+    {
+      "epoch": 0.8023847098018587,
+      "grad_norm": 1.230648398399353,
+      "learning_rate": 4.7016556630709284e-05,
+      "loss": 0.0126,
+      "step": 1144
+    },
+    {
+      "epoch": 0.8030860950376995,
+      "grad_norm": 0.9287470579147339,
+      "learning_rate": 4.700968421592313e-05,
+      "loss": 0.0094,
+      "step": 1145
+    },
+    {
+      "epoch": 0.8037874802735402,
+      "grad_norm": 62.944549560546875,
+      "learning_rate": 4.700280439833941e-05,
+      "loss": 1.3975,
+      "step": 1146
+    },
+    {
+      "epoch": 0.804488865509381,
+      "grad_norm": 44.87819290161133,
+      "learning_rate": 4.69959171802721e-05,
+      "loss": 1.0857,
+      "step": 1147
+    },
+    {
+      "epoch": 0.8051902507452218,
+      "grad_norm": 0.5283865928649902,
+      "learning_rate": 4.6989022564037654e-05,
+      "loss": 0.0049,
+      "step": 1148
+    },
+    {
+      "epoch": 0.8058916359810626,
+      "grad_norm": 0.5992218852043152,
+      "learning_rate": 4.698212055195505e-05,
+      "loss": 0.0058,
+      "step": 1149
+    },
+    {
+      "epoch": 0.8065930212169033,
+      "grad_norm": 0.5411283373832703,
+      "learning_rate": 4.697521114634571e-05,
+      "loss": 0.0052,
+      "step": 1150
+    },
+    {
+      "epoch": 0.8072944064527442,
+      "grad_norm": 75.1710433959961,
+      "learning_rate": 4.696829434953357e-05,
+      "loss": 1.9684,
+      "step": 1151
+    },
+    {
+      "epoch": 0.8079957916885849,
+      "grad_norm": 0.3737725615501404,
+      "learning_rate": 4.6961370163845044e-05,
+      "loss": 0.0035,
+      "step": 1152
+    },
+    {
+      "epoch": 0.8086971769244258,
+      "grad_norm": 48.369232177734375,
+      "learning_rate": 4.6954438591609027e-05,
+      "loss": 1.2948,
+      "step": 1153
+    },
+    {
+      "epoch": 0.8093985621602665,
+      "grad_norm": 0.4170537292957306,
+      "learning_rate": 4.694749963515691e-05,
+      "loss": 0.0036,
+      "step": 1154
+    },
+    {
+      "epoch": 0.8100999473961074,
+      "grad_norm": 0.3476961553096771,
+      "learning_rate": 4.6940553296822565e-05,
+      "loss": 0.0031,
+      "step": 1155
+    },
+    {
+      "epoch": 0.8108013326319481,
+      "grad_norm": 40.558204650878906,
+      "learning_rate": 4.693359957894232e-05,
+      "loss": 1.199,
+      "step": 1156
+    },
+    {
+      "epoch": 0.8115027178677889,
+      "grad_norm": 78.08576202392578,
+      "learning_rate": 4.692663848385503e-05,
+      "loss": 2.1697,
+      "step": 1157
+    },
+    {
+      "epoch": 0.8122041031036297,
+      "grad_norm": 0.5332819819450378,
+      "learning_rate": 4.6919670013902015e-05,
+      "loss": 0.0051,
+      "step": 1158
+    },
+    {
+      "epoch": 0.8129054883394704,
+      "grad_norm": 0.6460719108581543,
+      "learning_rate": 4.691269417142704e-05,
+      "loss": 0.0063,
+      "step": 1159
+    },
+    {
+      "epoch": 0.8136068735753113,
+      "grad_norm": 0.6999556422233582,
+      "learning_rate": 4.69057109587764e-05,
+      "loss": 0.0068,
+      "step": 1160
+    },
+    {
+      "epoch": 0.814308258811152,
+      "grad_norm": 0.5912299156188965,
+      "learning_rate": 4.6898720378298846e-05,
+      "loss": 0.0057,
+      "step": 1161
+    },
+    {
+      "epoch": 0.8150096440469928,
+      "grad_norm": 0.4620411992073059,
+      "learning_rate": 4.6891722432345597e-05,
+      "loss": 0.0044,
+      "step": 1162
+    },
+    {
+      "epoch": 0.8157110292828336,
+      "grad_norm": 0.3710283935070038,
+      "learning_rate": 4.6884717123270374e-05,
+      "loss": 0.0034,
+      "step": 1163
+    },
+    {
+      "epoch": 0.8164124145186744,
+      "grad_norm": 0.22289921343326569,
+      "learning_rate": 4.687770445342936e-05,
+      "loss": 0.002,
+      "step": 1164
+    },
+    {
+      "epoch": 0.8171137997545151,
+      "grad_norm": 89.52765655517578,
+      "learning_rate": 4.687068442518121e-05,
+      "loss": 2.8479,
+      "step": 1165
+    },
+    {
+      "epoch": 0.817815184990356,
+      "grad_norm": 0.15177088975906372,
+      "learning_rate": 4.6863657040887065e-05,
+      "loss": 0.0013,
+      "step": 1166
+    },
+    {
+      "epoch": 0.8185165702261967,
+      "grad_norm": 0.13902179896831512,
+      "learning_rate": 4.685662230291054e-05,
+      "loss": 0.0012,
+      "step": 1167
+    },
+    {
+      "epoch": 0.8192179554620376,
+      "grad_norm": 0.10139818489551544,
+      "learning_rate": 4.68495802136177e-05,
+      "loss": 0.0009,
+      "step": 1168
+    },
+    {
+      "epoch": 0.8199193406978783,
+      "grad_norm": 0.08662685006856918,
+      "learning_rate": 4.684253077537712e-05,
+      "loss": 0.0007,
+      "step": 1169
+    },
+    {
+      "epoch": 0.820620725933719,
+      "grad_norm": 0.07555234432220459,
+      "learning_rate": 4.683547399055982e-05,
+      "loss": 0.0006,
+      "step": 1170
+    },
+    {
+      "epoch": 0.8213221111695599,
+      "grad_norm": 0.06445518881082535,
+      "learning_rate": 4.682840986153929e-05,
+      "loss": 0.0006,
+      "step": 1171
+    },
+    {
+      "epoch": 0.8220234964054006,
+      "grad_norm": 46.72000503540039,
+      "learning_rate": 4.68213383906915e-05,
+      "loss": 2.1281,
+      "step": 1172
+    },
+    {
+      "epoch": 0.8227248816412415,
+      "grad_norm": 46.803985595703125,
+      "learning_rate": 4.6814259580394915e-05,
+      "loss": 1.9621,
+      "step": 1173
+    },
+    {
+      "epoch": 0.8234262668770822,
+      "grad_norm": 0.07187401503324509,
+      "learning_rate": 4.68071734330304e-05,
+      "loss": 0.0006,
+      "step": 1174
+    },
+    {
+      "epoch": 0.824127652112923,
+      "grad_norm": 0.09642617404460907,
+      "learning_rate": 4.680007995098136e-05,
+      "loss": 0.0008,
+      "step": 1175
+    },
+    {
+      "epoch": 0.8248290373487638,
+      "grad_norm": 0.11466558277606964,
+      "learning_rate": 4.679297913663363e-05,
+      "loss": 0.001,
+      "step": 1176
+    },
+    {
+      "epoch": 0.8255304225846046,
+      "grad_norm": 0.11334109306335449,
+      "learning_rate": 4.67858709923755e-05,
+      "loss": 0.001,
+      "step": 1177
+    },
+    {
+      "epoch": 0.8262318078204454,
+      "grad_norm": 0.1963043361902237,
+      "learning_rate": 4.677875552059777e-05,
+      "loss": 0.0017,
+      "step": 1178
+    },
+    {
+      "epoch": 0.8269331930562862,
+      "grad_norm": 0.17632529139518738,
+      "learning_rate": 4.677163272369365e-05,
+      "loss": 0.0016,
+      "step": 1179
+    },
+    {
+      "epoch": 0.8276345782921269,
+      "grad_norm": 124.80657958984375,
+      "learning_rate": 4.6764502604058864e-05,
+      "loss": 3.9859,
+      "step": 1180
+    },
+    {
+      "epoch": 0.8283359635279678,
+      "grad_norm": 0.19632282853126526,
+      "learning_rate": 4.675736516409156e-05,
+      "loss": 0.0018,
+      "step": 1181
+    },
+    {
+      "epoch": 0.8290373487638085,
+      "grad_norm": 43.057350158691406,
+      "learning_rate": 4.675022040619238e-05,
+      "loss": 1.2696,
+      "step": 1182
+    },
+    {
+      "epoch": 0.8297387339996493,
+      "grad_norm": 39.46339416503906,
+      "learning_rate": 4.674306833276439e-05,
+      "loss": 1.093,
+      "step": 1183
+    },
+    {
+      "epoch": 0.8304401192354901,
+      "grad_norm": 0.653581976890564,
+      "learning_rate": 4.6735908946213156e-05,
+      "loss": 0.0065,
+      "step": 1184
+    },
+    {
+      "epoch": 0.8311415044713308,
+      "grad_norm": 0.8602274656295776,
+      "learning_rate": 4.6728742248946677e-05,
+      "loss": 0.0088,
+      "step": 1185
+    },
+    {
+      "epoch": 0.8318428897071717,
+      "grad_norm": 0.9501532912254333,
+      "learning_rate": 4.6721568243375425e-05,
+      "loss": 0.0095,
+      "step": 1186
+    },
+    {
+      "epoch": 0.8325442749430124,
+      "grad_norm": 0.8702390193939209,
+      "learning_rate": 4.671438693191232e-05,
+      "loss": 0.0088,
+      "step": 1187
+    },
+    {
+      "epoch": 0.8332456601788533,
+      "grad_norm": 0.6245111227035522,
+      "learning_rate": 4.6707198316972746e-05,
+      "loss": 0.0061,
+      "step": 1188
+    },
+    {
+      "epoch": 0.833947045414694,
+      "grad_norm": 0.4615590572357178,
+      "learning_rate": 4.6700002400974544e-05,
+      "loss": 0.0045,
+      "step": 1189
+    },
+    {
+      "epoch": 0.8346484306505348,
+      "grad_norm": 0.3482006788253784,
+      "learning_rate": 4.6692799186338e-05,
+      "loss": 0.0033,
+      "step": 1190
+    },
+    {
+      "epoch": 0.8353498158863756,
+      "grad_norm": 43.396461486816406,
+      "learning_rate": 4.668558867548587e-05,
+      "loss": 1.3748,
+      "step": 1191
+    },
+    {
+      "epoch": 0.8360512011222164,
+      "grad_norm": 49.66410827636719,
+      "learning_rate": 4.6678370870843357e-05,
+      "loss": 1.7113,
+      "step": 1192
+    },
+    {
+      "epoch": 0.8367525863580572,
+      "grad_norm": 0.1562856137752533,
+      "learning_rate": 4.667114577483812e-05,
+      "loss": 0.0014,
+      "step": 1193
+    },
+    {
+      "epoch": 0.8374539715938979,
+      "grad_norm": 47.60147476196289,
+      "learning_rate": 4.666391338990026e-05,
+      "loss": 1.8113,
+      "step": 1194
+    },
+    {
+      "epoch": 0.8381553568297387,
+      "grad_norm": 0.19547221064567566,
+      "learning_rate": 4.665667371846234e-05,
+      "loss": 0.0018,
+      "step": 1195
+    },
+    {
+      "epoch": 0.8388567420655795,
+      "grad_norm": 0.23496605455875397,
+      "learning_rate": 4.664942676295937e-05,
+      "loss": 0.0022,
+      "step": 1196
+    },
+    {
+      "epoch": 0.8395581273014203,
+      "grad_norm": 0.19859689474105835,
+      "learning_rate": 4.6642172525828806e-05,
+      "loss": 0.0018,
+      "step": 1197
+    },
+    {
+      "epoch": 0.840259512537261,
+      "grad_norm": 0.21561066806316376,
+      "learning_rate": 4.6634911009510574e-05,
+      "loss": 0.002,
+      "step": 1198
+    },
+    {
+      "epoch": 0.8409608977731019,
+      "grad_norm": 0.16780033707618713,
+      "learning_rate": 4.662764221644702e-05,
+      "loss": 0.0016,
+      "step": 1199
+    },
+    {
+      "epoch": 0.8416622830089426,
+      "grad_norm": 0.170645073056221,
+      "learning_rate": 4.6620366149082954e-05,
+      "loss": 0.0015,
+      "step": 1200
+    },
+    {
+      "epoch": 0.8423636682447835,
+      "grad_norm": 0.16811063885688782,
+      "learning_rate": 4.661308280986563e-05,
+      "loss": 0.0015,
+      "step": 1201
+    },
+    {
+      "epoch": 0.8430650534806242,
+      "grad_norm": 42.41083908081055,
+      "learning_rate": 4.6605792201244733e-05,
+      "loss": 1.4838,
+      "step": 1202
+    },
+    {
+      "epoch": 0.8437664387164651,
+      "grad_norm": 43.9367790222168,
+      "learning_rate": 4.6598494325672426e-05,
+      "loss": 1.4149,
+      "step": 1203
+    },
+    {
+      "epoch": 0.8444678239523058,
+      "grad_norm": 0.1912125051021576,
+      "learning_rate": 4.659118918560329e-05,
+      "loss": 0.0018,
+      "step": 1204
+    },
+    {
+      "epoch": 0.8451692091881466,
+      "grad_norm": 0.27999013662338257,
+      "learning_rate": 4.6583876783494343e-05,
+      "loss": 0.0025,
+      "step": 1205
+    },
+    {
+      "epoch": 0.8458705944239874,
+      "grad_norm": 36.06639099121094,
+      "learning_rate": 4.657655712180507e-05,
+      "loss": 0.9947,
+      "step": 1206
+    },
+    {
+      "epoch": 0.8465719796598281,
+      "grad_norm": 0.4413144588470459,
+      "learning_rate": 4.6569230202997385e-05,
+      "loss": 0.0042,
+      "step": 1207
+    },
+    {
+      "epoch": 0.847273364895669,
+      "grad_norm": 0.530229926109314,
+      "learning_rate": 4.656189602953565e-05,
+      "loss": 0.0052,
+      "step": 1208
+    },
+    {
+      "epoch": 0.8479747501315097,
+      "grad_norm": 0.6561914086341858,
+      "learning_rate": 4.655455460388665e-05,
+      "loss": 0.0066,
+      "step": 1209
+    },
+    {
+      "epoch": 0.8486761353673505,
+      "grad_norm": 0.6459203362464905,
+      "learning_rate": 4.654720592851962e-05,
+      "loss": 0.0063,
+      "step": 1210
+    },
+    {
+      "epoch": 0.8493775206031913,
+      "grad_norm": 37.68455505371094,
+      "learning_rate": 4.653985000590624e-05,
+      "loss": 1.0854,
+      "step": 1211
+    },
+    {
+      "epoch": 0.8500789058390321,
+      "grad_norm": 0.4941748380661011,
+      "learning_rate": 4.6532486838520616e-05,
+      "loss": 0.0049,
+      "step": 1212
+    },
+    {
+      "epoch": 0.8507802910748729,
+      "grad_norm": 38.626068115234375,
+      "learning_rate": 4.6525116428839307e-05,
+      "loss": 1.0544,
+      "step": 1213
+    },
+    {
+      "epoch": 0.8514816763107137,
+      "grad_norm": 0.5088845491409302,
+      "learning_rate": 4.651773877934129e-05,
+      "loss": 0.005,
+      "step": 1214
+    },
+    {
+      "epoch": 0.8521830615465544,
+      "grad_norm": 0.5165444016456604,
+      "learning_rate": 4.651035389250798e-05,
+      "loss": 0.005,
+      "step": 1215
+    },
+    {
+      "epoch": 0.8528844467823953,
+      "grad_norm": 37.2906379699707,
+      "learning_rate": 4.650296177082323e-05,
+      "loss": 1.0227,
+      "step": 1216
+    },
+    {
+      "epoch": 0.853585832018236,
+      "grad_norm": 0.5255476832389832,
+      "learning_rate": 4.649556241677333e-05,
+      "loss": 0.0052,
+      "step": 1217
+    },
+    {
+      "epoch": 0.8542872172540767,
+      "grad_norm": 37.65691375732422,
+      "learning_rate": 4.6488155832847006e-05,
+      "loss": 1.1044,
+      "step": 1218
+    },
+    {
+      "epoch": 0.8549886024899176,
+      "grad_norm": 35.80206298828125,
+      "learning_rate": 4.648074202153539e-05,
+      "loss": 0.8649,
+      "step": 1219
+    },
+    {
+      "epoch": 0.8556899877257583,
+      "grad_norm": 0.8690975308418274,
+      "learning_rate": 4.647332098533208e-05,
+      "loss": 0.0089,
+      "step": 1220
+    },
+    {
+      "epoch": 0.8563913729615992,
+      "grad_norm": 1.0327463150024414,
+      "learning_rate": 4.646589272673308e-05,
+      "loss": 0.0109,
+      "step": 1221
+    },
+    {
+      "epoch": 0.8570927581974399,
+      "grad_norm": 32.14468765258789,
+      "learning_rate": 4.645845724823684e-05,
+      "loss": 0.7778,
+      "step": 1222
+    },
+    {
+      "epoch": 0.8577941434332808,
+      "grad_norm": 1.0274231433868408,
+      "learning_rate": 4.645101455234422e-05,
+      "loss": 0.0106,
+      "step": 1223
+    },
+    {
+      "epoch": 0.8584955286691215,
+      "grad_norm": 29.724233627319336,
+      "learning_rate": 4.644356464155852e-05,
+      "loss": 0.6951,
+      "step": 1224
+    },
+    {
+      "epoch": 0.8591969139049623,
+      "grad_norm": 1.3864684104919434,
+      "learning_rate": 4.6436107518385465e-05,
+      "loss": 0.0151,
+      "step": 1225
+    },
+    {
+      "epoch": 0.8598982991408031,
+      "grad_norm": 1.3250905275344849,
+      "learning_rate": 4.6428643185333196e-05,
+      "loss": 0.0144,
+      "step": 1226
+    },
+    {
+      "epoch": 0.8605996843766439,
+      "grad_norm": 29.683332443237305,
+      "learning_rate": 4.6421171644912295e-05,
+      "loss": 0.7007,
+      "step": 1227
+    },
+    {
+      "epoch": 0.8613010696124846,
+      "grad_norm": 29.98341178894043,
+      "learning_rate": 4.6413692899635764e-05,
+      "loss": 0.6981,
+      "step": 1228
+    },
+    {
+      "epoch": 0.8620024548483255,
+      "grad_norm": 30.50311279296875,
+      "learning_rate": 4.6406206952019014e-05,
+      "loss": 0.7341,
+      "step": 1229
+    },
+    {
+      "epoch": 0.8627038400841662,
+      "grad_norm": 1.5321385860443115,
+      "learning_rate": 4.6398713804579896e-05,
+      "loss": 0.0171,
+      "step": 1230
+    },
+    {
+      "epoch": 0.863405225320007,
+      "grad_norm": 25.173595428466797,
+      "learning_rate": 4.6391213459838676e-05,
+      "loss": 0.5393,
+      "step": 1231
+    },
+    {
+      "epoch": 0.8641066105558478,
+      "grad_norm": 1.9056795835494995,
+      "learning_rate": 4.638370592031804e-05,
+      "loss": 0.0216,
+      "step": 1232
+    },
+    {
+      "epoch": 0.8648079957916885,
+      "grad_norm": 54.3871955871582,
+      "learning_rate": 4.63761911885431e-05,
+      "loss": 1.1726,
+      "step": 1233
+    },
+    {
+      "epoch": 0.8655093810275294,
+      "grad_norm": 2.3975331783294678,
+      "learning_rate": 4.6368669267041364e-05,
+      "loss": 0.0284,
+      "step": 1234
+    },
+    {
+      "epoch": 0.8662107662633701,
+      "grad_norm": 2.3505735397338867,
+      "learning_rate": 4.636114015834279e-05,
+      "loss": 0.027,
+      "step": 1235
+    },
+    {
+      "epoch": 0.866912151499211,
+      "grad_norm": 19.353313446044922,
+      "learning_rate": 4.635360386497973e-05,
+      "loss": 0.4174,
+      "step": 1236
+    },
+    {
+      "epoch": 0.8676135367350517,
+      "grad_norm": 2.5313189029693604,
+      "learning_rate": 4.6346060389486976e-05,
+      "loss": 0.0301,
+      "step": 1237
+    },
+    {
+      "epoch": 0.8683149219708926,
+      "grad_norm": 2.249114751815796,
+      "learning_rate": 4.633850973440171e-05,
+      "loss": 0.0262,
+      "step": 1238
+    },
+    {
+      "epoch": 0.8690163072067333,
+      "grad_norm": 1.087750792503357,
+      "learning_rate": 4.633095190226355e-05,
+      "loss": 0.0116,
+      "step": 1239
+    },
+    {
+      "epoch": 0.8697176924425741,
+      "grad_norm": 0.7214330434799194,
+      "learning_rate": 4.63233868956145e-05,
+      "loss": 0.0075,
+      "step": 1240
+    },
+    {
+      "epoch": 0.8704190776784149,
+      "grad_norm": 0.5183669924736023,
+      "learning_rate": 4.6315814716999003e-05,
+      "loss": 0.0051,
+      "step": 1241
+    },
+    {
+      "epoch": 0.8711204629142556,
+      "grad_norm": 0.26819220185279846,
+      "learning_rate": 4.630823536896392e-05,
+      "loss": 0.0026,
+      "step": 1242
+    },
+    {
+      "epoch": 0.8718218481500964,
+      "grad_norm": 42.74302291870117,
+      "learning_rate": 4.6300648854058496e-05,
+      "loss": 1.4745,
+      "step": 1243
+    },
+    {
+      "epoch": 0.8725232333859372,
+      "grad_norm": 0.06916595250368118,
+      "learning_rate": 4.62930551748344e-05,
+      "loss": 0.0006,
+      "step": 1244
+    },
+    {
+      "epoch": 0.873224618621778,
+      "grad_norm": 0.04199717566370964,
+      "learning_rate": 4.6285454333845714e-05,
+      "loss": 0.0004,
+      "step": 1245
+    },
+    {
+      "epoch": 0.8739260038576188,
+      "grad_norm": 46.651554107666016,
+      "learning_rate": 4.6277846333648926e-05,
+      "loss": 2.1377,
+      "step": 1246
+    },
+    {
+      "epoch": 0.8746273890934596,
+      "grad_norm": 0.019042061641812325,
+      "learning_rate": 4.6270231176802936e-05,
+      "loss": 0.0002,
+      "step": 1247
+    },
+    {
+      "epoch": 0.8753287743293003,
+      "grad_norm": 43.70697784423828,
+      "learning_rate": 4.626260886586903e-05,
+      "loss": 1.9754,
+      "step": 1248
+    },
+    {
+      "epoch": 0.8760301595651412,
+      "grad_norm": 0.03968697413802147,
+      "learning_rate": 4.625497940341094e-05,
+      "loss": 0.0003,
+      "step": 1249
+    },
+    {
+      "epoch": 0.8767315448009819,
+      "grad_norm": 0.023654410615563393,
+      "learning_rate": 4.624734279199476e-05,
+      "loss": 0.0002,
+      "step": 1250
+    },
+    {
+      "epoch": 0.8774329300368228,
+      "grad_norm": 0.05060436204075813,
+      "learning_rate": 4.623969903418903e-05,
+      "loss": 0.0004,
+      "step": 1251
+    },
+    {
+      "epoch": 0.8781343152726635,
+      "grad_norm": 0.062442123889923096,
+      "learning_rate": 4.623204813256465e-05,
+      "loss": 0.0006,
+      "step": 1252
+    },
+    {
+      "epoch": 0.8788357005085043,
+      "grad_norm": 0.05655084550380707,
+      "learning_rate": 4.622439008969497e-05,
+      "loss": 0.0005,
+      "step": 1253
+    },
+    {
+      "epoch": 0.8795370857443451,
+      "grad_norm": 0.05976495519280434,
+      "learning_rate": 4.621672490815569e-05,
+      "loss": 0.0005,
+      "step": 1254
+    },
+    {
+      "epoch": 0.8802384709801858,
+      "grad_norm": 0.06660816818475723,
+      "learning_rate": 4.620905259052496e-05,
+      "loss": 0.0006,
+      "step": 1255
+    },
+    {
+      "epoch": 0.8809398562160267,
+      "grad_norm": 45.397789001464844,
+      "learning_rate": 4.6201373139383284e-05,
+      "loss": 1.7976,
+      "step": 1256
+    },
+    {
+      "epoch": 0.8816412414518674,
+      "grad_norm": 44.1479606628418,
+      "learning_rate": 4.619368655731362e-05,
+      "loss": 1.7646,
+      "step": 1257
+    },
+    {
+      "epoch": 0.8823426266877082,
+      "grad_norm": 0.13133244216442108,
+      "learning_rate": 4.618599284690127e-05,
+      "loss": 0.0012,
+      "step": 1258
+    },
+    {
+      "epoch": 0.883044011923549,
+      "grad_norm": 82.86041259765625,
+      "learning_rate": 4.6178292010733984e-05,
+      "loss": 2.727,
+      "step": 1259
+    },
+    {
+      "epoch": 0.8837453971593898,
+      "grad_norm": 36.786067962646484,
+      "learning_rate": 4.6170584051401846e-05,
+      "loss": 1.0486,
+      "step": 1260
+    },
+    {
+      "epoch": 0.8844467823952306,
+      "grad_norm": 43.94260787963867,
+      "learning_rate": 4.6162868971497396e-05,
+      "loss": 1.3111,
+      "step": 1261
+    },
+    {
+      "epoch": 0.8851481676310714,
+      "grad_norm": 85.85136413574219,
+      "learning_rate": 4.615514677361554e-05,
+      "loss": 2.2235,
+      "step": 1262
+    },
+    {
+      "epoch": 0.8858495528669121,
+      "grad_norm": 35.138553619384766,
+      "learning_rate": 4.6147417460353596e-05,
+      "loss": 0.6498,
+      "step": 1263
+    },
+    {
+      "epoch": 0.886550938102753,
+      "grad_norm": 5.97971773147583,
+      "learning_rate": 4.613968103431124e-05,
+      "loss": 0.0857,
+      "step": 1264
+    },
+    {
+      "epoch": 0.8872523233385937,
+      "grad_norm": 7.489769458770752,
+      "learning_rate": 4.6131937498090575e-05,
+      "loss": 0.1154,
+      "step": 1265
+    },
+    {
+      "epoch": 0.8879537085744345,
+      "grad_norm": 9.91996955871582,
+      "learning_rate": 4.612418685429608e-05,
+      "loss": 0.1685,
+      "step": 1266
+    },
+    {
+      "epoch": 0.8886550938102753,
+      "grad_norm": 2.7866861820220947,
+      "learning_rate": 4.611642910553463e-05,
+      "loss": 0.2187,
+      "step": 1267
+    },
+    {
+      "epoch": 0.889356479046116,
+      "grad_norm": 9.682534217834473,
+      "learning_rate": 4.610866425441549e-05,
+      "loss": 0.1669,
+      "step": 1268
+    },
+    {
+      "epoch": 0.8900578642819569,
+      "grad_norm": 8.847615242004395,
+      "learning_rate": 4.6100892303550305e-05,
+      "loss": 0.1458,
+      "step": 1269
+    },
+    {
+      "epoch": 0.8907592495177976,
+      "grad_norm": 6.194153308868408,
+      "learning_rate": 4.609311325555311e-05,
+      "loss": 0.0885,
+      "step": 1270
+    },
+    {
+      "epoch": 0.8914606347536385,
+      "grad_norm": 11.383880615234375,
+      "learning_rate": 4.6085327113040346e-05,
+      "loss": 0.2945,
+      "step": 1271
+    },
+    {
+      "epoch": 0.8921620199894792,
+      "grad_norm": 41.971961975097656,
+      "learning_rate": 4.607753387863082e-05,
+      "loss": 0.8504,
+      "step": 1272
+    },
+    {
+      "epoch": 0.89286340522532,
+      "grad_norm": 15.673029899597168,
+      "learning_rate": 4.6069733554945725e-05,
+      "loss": 0.3576,
+      "step": 1273
+    },
+    {
+      "epoch": 0.8935647904611608,
+      "grad_norm": 2.994936466217041,
+      "learning_rate": 4.606192614460865e-05,
+      "loss": 0.0371,
+      "step": 1274
+    },
+    {
+      "epoch": 0.8942661756970016,
+      "grad_norm": 15.890610694885254,
+      "learning_rate": 4.6054111650245544e-05,
+      "loss": 0.3475,
+      "step": 1275
+    },
+    {
+      "epoch": 0.8949675609328424,
+      "grad_norm": 3.3848869800567627,
+      "learning_rate": 4.604629007448479e-05,
+      "loss": 0.0433,
+      "step": 1276
+    },
+    {
+      "epoch": 0.8956689461686832,
+      "grad_norm": 26.11835289001465,
+      "learning_rate": 4.603846141995708e-05,
+      "loss": 0.5389,
+      "step": 1277
+    },
+    {
+      "epoch": 0.8963703314045239,
+      "grad_norm": 3.1392459869384766,
+      "learning_rate": 4.6030625689295546e-05,
+      "loss": 0.0383,
+      "step": 1278
+    },
+    {
+      "epoch": 0.8970717166403647,
+      "grad_norm": 3.49643611907959,
+      "learning_rate": 4.6022782885135674e-05,
+      "loss": 0.0445,
+      "step": 1279
+    },
+    {
+      "epoch": 0.8977731018762055,
+      "grad_norm": 2.069392204284668,
+      "learning_rate": 4.601493301011534e-05,
+      "loss": 0.0228,
+      "step": 1280
+    },
+    {
+      "epoch": 0.8984744871120462,
+      "grad_norm": 1.4139840602874756,
+      "learning_rate": 4.6007076066874786e-05,
+      "loss": 0.016,
+      "step": 1281
+    },
+    {
+      "epoch": 0.8991758723478871,
+      "grad_norm": 0.8678575158119202,
+      "learning_rate": 4.599921205805663e-05,
+      "loss": 0.0092,
+      "step": 1282
+    },
+    {
+      "epoch": 0.8998772575837278,
+      "grad_norm": 0.3188234567642212,
+      "learning_rate": 4.5991340986305896e-05,
+      "loss": 0.0031,
+      "step": 1283
+    },
+    {
+      "epoch": 0.9005786428195687,
+      "grad_norm": 0.13479161262512207,
+      "learning_rate": 4.598346285426994e-05,
+      "loss": 0.0012,
+      "step": 1284
+    },
+    {
+      "epoch": 0.9012800280554094,
+      "grad_norm": 44.85018539428711,
+      "learning_rate": 4.5975577664598535e-05,
+      "loss": 1.8893,
+      "step": 1285
+    },
+    {
+      "epoch": 0.9019814132912503,
+      "grad_norm": 0.030871881172060966,
+      "learning_rate": 4.596768541994379e-05,
+      "loss": 0.0003,
+      "step": 1286
+    },
+    {
+      "epoch": 0.902682798527091,
+      "grad_norm": 0.025855490937829018,
+      "learning_rate": 4.595978612296021e-05,
+      "loss": 0.0002,
+      "step": 1287
+    },
+    {
+      "epoch": 0.9033841837629318,
+      "grad_norm": 46.28804397583008,
+      "learning_rate": 4.595187977630467e-05,
+      "loss": 2.4086,
+      "step": 1288
+    },
+    {
+      "epoch": 0.9040855689987726,
+      "grad_norm": 46.17903518676758,
+      "learning_rate": 4.5943966382636413e-05,
+      "loss": 2.5441,
+      "step": 1289
+    },
+    {
+      "epoch": 0.9047869542346134,
+      "grad_norm": 48.56827926635742,
+      "learning_rate": 4.593604594461704e-05,
+      "loss": 2.5474,
+      "step": 1290
+    },
+    {
+      "epoch": 0.9054883394704542,
+      "grad_norm": 47.86493682861328,
+      "learning_rate": 4.592811846491055e-05,
+      "loss": 2.1781,
+      "step": 1291
+    },
+    {
+      "epoch": 0.9061897247062949,
+      "grad_norm": 43.444766998291016,
+      "learning_rate": 4.592018394618328e-05,
+      "loss": 1.8371,
+      "step": 1292
+    },
+    {
+      "epoch": 0.9068911099421357,
+      "grad_norm": 46.283565521240234,
+      "learning_rate": 4.5912242391103944e-05,
+      "loss": 1.5648,
+      "step": 1293
+    },
+    {
+      "epoch": 0.9075924951779765,
+      "grad_norm": 0.2855938971042633,
+      "learning_rate": 4.5904293802343646e-05,
+      "loss": 0.0028,
+      "step": 1294
+    },
+    {
+      "epoch": 0.9082938804138173,
+      "grad_norm": 44.066627502441406,
+      "learning_rate": 4.589633818257582e-05,
+      "loss": 1.3621,
+      "step": 1295
+    },
+    {
+      "epoch": 0.908995265649658,
+      "grad_norm": 1.1252837181091309,
+      "learning_rate": 4.588837553447628e-05,
+      "loss": 0.0122,
+      "step": 1296
+    },
+    {
+      "epoch": 0.9096966508854989,
+      "grad_norm": 1.7556012868881226,
+      "learning_rate": 4.588040586072321e-05,
+      "loss": 0.0199,
+      "step": 1297
+    },
+    {
+      "epoch": 0.9103980361213396,
+      "grad_norm": 2.0839264392852783,
+      "learning_rate": 4.5872429163997163e-05,
+      "loss": 0.0248,
+      "step": 1298
+    },
+    {
+      "epoch": 0.9110994213571805,
+      "grad_norm": 2.153773784637451,
+      "learning_rate": 4.5864445446981025e-05,
+      "loss": 0.0254,
+      "step": 1299
+    },
+    {
+      "epoch": 0.9118008065930212,
+      "grad_norm": 2.1063156127929688,
+      "learning_rate": 4.5856454712360067e-05,
+      "loss": 0.0235,
+      "step": 1300
+    },
+    {
+      "epoch": 0.912502191828862,
+      "grad_norm": 1.26010000705719,
+      "learning_rate": 4.584845696282192e-05,
+      "loss": 0.0134,
+      "step": 1301
+    },
+    {
+      "epoch": 0.9132035770647028,
+      "grad_norm": 0.7219848036766052,
+      "learning_rate": 4.584045220105656e-05,
+      "loss": 0.0075,
+      "step": 1302
+    },
+    {
+      "epoch": 0.9139049623005435,
+      "grad_norm": 0.3715592324733734,
+      "learning_rate": 4.583244042975634e-05,
+      "loss": 0.0037,
+      "step": 1303
+    },
+    {
+      "epoch": 0.9146063475363844,
+      "grad_norm": 0.16978369653224945,
+      "learning_rate": 4.582442165161596e-05,
+      "loss": 0.0017,
+      "step": 1304
+    },
+    {
+      "epoch": 0.9153077327722251,
+      "grad_norm": 0.10813824832439423,
+      "learning_rate": 4.5816395869332476e-05,
+      "loss": 0.001,
+      "step": 1305
+    },
+    {
+      "epoch": 0.916009118008066,
+      "grad_norm": 0.05026814341545105,
+      "learning_rate": 4.5808363085605296e-05,
+      "loss": 0.0005,
+      "step": 1306
+    },
+    {
+      "epoch": 0.9167105032439067,
+      "grad_norm": 0.025502556934952736,
+      "learning_rate": 4.58003233031362e-05,
+      "loss": 0.0002,
+      "step": 1307
+    },
+    {
+      "epoch": 0.9174118884797475,
+      "grad_norm": 45.29407501220703,
+      "learning_rate": 4.579227652462928e-05,
+      "loss": 2.3544,
+      "step": 1308
+    },
+    {
+      "epoch": 0.9181132737155883,
+      "grad_norm": 0.012745014391839504,
+      "learning_rate": 4.578422275279105e-05,
+      "loss": 0.0001,
+      "step": 1309
+    },
+    {
+      "epoch": 0.9188146589514291,
+      "grad_norm": 0.009426114149391651,
+      "learning_rate": 4.577616199033033e-05,
+      "loss": 0.0001,
+      "step": 1310
+    },
+    {
+      "epoch": 0.9195160441872698,
+      "grad_norm": 0.008077832870185375,
+      "learning_rate": 4.5768094239958285e-05,
+      "loss": 0.0001,
+      "step": 1311
+    },
+    {
+      "epoch": 0.9202174294231107,
+      "grad_norm": 0.006128115579485893,
+      "learning_rate": 4.576001950438845e-05,
+      "loss": 0.0001,
+      "step": 1312
+    },
+    {
+      "epoch": 0.9209188146589514,
+      "grad_norm": 88.39704895019531,
+      "learning_rate": 4.575193778633671e-05,
+      "loss": 5.5453,
+      "step": 1313
+    },
+    {
+      "epoch": 0.9216201998947923,
+      "grad_norm": 45.73112869262695,
+      "learning_rate": 4.574384908852128e-05,
+      "loss": 2.724,
+      "step": 1314
+    },
+    {
+      "epoch": 0.922321585130633,
+      "grad_norm": 45.08832550048828,
+      "learning_rate": 4.573575341366274e-05,
+      "loss": 2.5745,
+      "step": 1315
+    },
+    {
+      "epoch": 0.9230229703664737,
+      "grad_norm": 0.01642024703323841,
+      "learning_rate": 4.572765076448402e-05,
+      "loss": 0.0001,
+      "step": 1316
+    },
+    {
+      "epoch": 0.9237243556023146,
+      "grad_norm": 0.0225964467972517,
+      "learning_rate": 4.5719541143710375e-05,
+      "loss": 0.0002,
+      "step": 1317
+    },
+    {
+      "epoch": 0.9244257408381553,
+      "grad_norm": 0.03809621185064316,
+      "learning_rate": 4.571142455406943e-05,
+      "loss": 0.0003,
+      "step": 1318
+    },
+    {
+      "epoch": 0.9251271260739962,
+      "grad_norm": 42.60374450683594,
+      "learning_rate": 4.5703300998291124e-05,
+      "loss": 1.805,
+      "step": 1319
+    },
+    {
+      "epoch": 0.9258285113098369,
+      "grad_norm": 0.08730126917362213,
+      "learning_rate": 4.569517047910777e-05,
+      "loss": 0.0008,
+      "step": 1320
+    },
+    {
+      "epoch": 0.9265298965456777,
+      "grad_norm": 119.11152648925781,
+      "learning_rate": 4.5687032999254e-05,
+      "loss": 4.0878,
+      "step": 1321
+    },
+    {
+      "epoch": 0.9272312817815185,
+      "grad_norm": 0.3243617117404938,
+      "learning_rate": 4.567888856146681e-05,
+      "loss": 0.0031,
+      "step": 1322
+    },
+    {
+      "epoch": 0.9279326670173593,
+      "grad_norm": 0.4152386784553528,
+      "learning_rate": 4.567073716848551e-05,
+      "loss": 0.004,
+      "step": 1323
+    },
+    {
+      "epoch": 0.9286340522532001,
+      "grad_norm": 30.812305450439453,
+      "learning_rate": 4.566257882305176e-05,
+      "loss": 0.7747,
+      "step": 1324
+    },
+    {
+      "epoch": 0.9293354374890409,
+      "grad_norm": 30.550561904907227,
+      "learning_rate": 4.565441352790957e-05,
+      "loss": 0.7148,
+      "step": 1325
+    },
+    {
+      "epoch": 0.9300368227248816,
+      "grad_norm": 79.6876220703125,
+      "learning_rate": 4.564624128580528e-05,
+      "loss": 1.728,
+      "step": 1326
+    },
+    {
+      "epoch": 0.9307382079607224,
+      "grad_norm": 3.2627832889556885,
+      "learning_rate": 4.5638062099487545e-05,
+      "loss": 0.0427,
+      "step": 1327
+    },
+    {
+      "epoch": 0.9314395931965632,
+      "grad_norm": 10.727757453918457,
+      "learning_rate": 4.5629875971707395e-05,
+      "loss": 0.2789,
+      "step": 1328
+    },
+    {
+      "epoch": 0.932140978432404,
+      "grad_norm": 5.946654319763184,
+      "learning_rate": 4.562168290521817e-05,
+      "loss": 0.0887,
+      "step": 1329
+    },
+    {
+      "epoch": 0.9328423636682448,
+      "grad_norm": 1.3037784099578857,
+      "learning_rate": 4.561348290277554e-05,
+      "loss": 0.1909,
+      "step": 1330
+    },
+    {
+      "epoch": 0.9335437489040855,
+      "grad_norm": 8.197032928466797,
+      "learning_rate": 4.560527596713752e-05,
+      "loss": 0.1388,
+      "step": 1331
+    },
+    {
+      "epoch": 0.9342451341399264,
+      "grad_norm": 8.003156661987305,
+      "learning_rate": 4.559706210106446e-05,
+      "loss": 0.1299,
+      "step": 1332
+    },
+    {
+      "epoch": 0.9349465193757671,
+      "grad_norm": 13.647798538208008,
+      "learning_rate": 4.558884130731903e-05,
+      "loss": 0.3574,
+      "step": 1333
+    },
+    {
+      "epoch": 0.935647904611608,
+      "grad_norm": 5.755825996398926,
+      "learning_rate": 4.558061358866623e-05,
+      "loss": 0.0832,
+      "step": 1334
+    },
+    {
+      "epoch": 0.9363492898474487,
+      "grad_norm": 21.87740135192871,
+      "learning_rate": 4.55723789478734e-05,
+      "loss": 0.4666,
+      "step": 1335
+    },
+    {
+      "epoch": 0.9370506750832895,
+      "grad_norm": 9.851075172424316,
+      "learning_rate": 4.5564137387710196e-05,
+      "loss": 0.2624,
+      "step": 1336
+    },
+    {
+      "epoch": 0.9377520603191303,
+      "grad_norm": 4.983445644378662,
+      "learning_rate": 4.555588891094862e-05,
+      "loss": 0.0688,
+      "step": 1337
+    },
+    {
+      "epoch": 0.9384534455549711,
+      "grad_norm": 4.145101547241211,
+      "learning_rate": 4.554763352036296e-05,
+      "loss": 0.0537,
+      "step": 1338
+    },
+    {
+      "epoch": 0.9391548307908119,
+      "grad_norm": 4.348505020141602,
+      "learning_rate": 4.5539371218729894e-05,
+      "loss": 0.058,
+      "step": 1339
+    },
+    {
+      "epoch": 0.9398562160266526,
+      "grad_norm": 16.51106834411621,
+      "learning_rate": 4.5531102008828354e-05,
+      "loss": 0.3685,
+      "step": 1340
+    },
+    {
+      "epoch": 0.9405576012624934,
+      "grad_norm": 2.958770513534546,
+      "learning_rate": 4.5522825893439645e-05,
+      "loss": 0.0375,
+      "step": 1341
+    },
+    {
+      "epoch": 0.9412589864983342,
+      "grad_norm": 1.705260157585144,
+      "learning_rate": 4.5514542875347376e-05,
+      "loss": 0.0194,
+      "step": 1342
+    },
+    {
+      "epoch": 0.941960371734175,
+      "grad_norm": 28.18739891052246,
+      "learning_rate": 4.550625295733748e-05,
+      "loss": 0.6777,
+      "step": 1343
+    },
+    {
+      "epoch": 0.9426617569700158,
+      "grad_norm": 39.53862380981445,
+      "learning_rate": 4.5497956142198214e-05,
+      "loss": 1.1208,
+      "step": 1344
+    },
+    {
+      "epoch": 0.9433631422058566,
+      "grad_norm": 1.0771280527114868,
+      "learning_rate": 4.5489652432720145e-05,
+      "loss": 0.0118,
+      "step": 1345
+    },
+    {
+      "epoch": 0.9440645274416973,
+      "grad_norm": 96.40424346923828,
+      "learning_rate": 4.548134183169617e-05,
+      "loss": 2.4673,
+      "step": 1346
+    },
+    {
+      "epoch": 0.9447659126775382,
+      "grad_norm": 0.5970985293388367,
+      "learning_rate": 4.547302434192149e-05,
+      "loss": 0.0061,
+      "step": 1347
+    },
+    {
+      "epoch": 0.9454672979133789,
+      "grad_norm": 36.97736358642578,
+      "learning_rate": 4.546469996619364e-05,
+      "loss": 0.9705,
+      "step": 1348
+    },
+    {
+      "epoch": 0.9461686831492198,
+      "grad_norm": 0.6519697308540344,
+      "learning_rate": 4.545636870731247e-05,
+      "loss": 0.0066,
+      "step": 1349
+    },
+    {
+      "epoch": 0.9468700683850605,
+      "grad_norm": 0.4174421429634094,
+      "learning_rate": 4.544803056808013e-05,
+      "loss": 0.0042,
+      "step": 1350
+    },
+    {
+      "epoch": 0.9475714536209012,
+      "grad_norm": 0.4273722469806671,
+      "learning_rate": 4.5439685551301094e-05,
+      "loss": 0.0043,
+      "step": 1351
+    },
+    {
+      "epoch": 0.9482728388567421,
+      "grad_norm": 0.3083280324935913,
+      "learning_rate": 4.543133365978214e-05,
+      "loss": 0.003,
+      "step": 1352
+    },
+    {
+      "epoch": 0.9489742240925828,
+      "grad_norm": 0.2918597161769867,
+      "learning_rate": 4.542297489633238e-05,
+      "loss": 0.0028,
+      "step": 1353
+    },
+    {
+      "epoch": 0.9496756093284237,
+      "grad_norm": 47.360557556152344,
+      "learning_rate": 4.541460926376322e-05,
+      "loss": 1.7171,
+      "step": 1354
+    },
+    {
+      "epoch": 0.9503769945642644,
+      "grad_norm": 81.7466049194336,
+      "learning_rate": 4.5406236764888374e-05,
+      "loss": 2.8326,
+      "step": 1355
+    },
+    {
+      "epoch": 0.9510783798001052,
+      "grad_norm": 39.265384674072266,
+      "learning_rate": 4.539785740252388e-05,
+      "loss": 1.3337,
+      "step": 1356
+    },
+    {
+      "epoch": 0.951779765035946,
+      "grad_norm": 0.28257325291633606,
+      "learning_rate": 4.5389471179488074e-05,
+      "loss": 0.0027,
+      "step": 1357
+    },
+    {
+      "epoch": 0.9524811502717868,
+      "grad_norm": 41.259849548339844,
+      "learning_rate": 4.538107809860159e-05,
+      "loss": 1.2651,
+      "step": 1358
+    },
+    {
+      "epoch": 0.9531825355076275,
+      "grad_norm": 0.5483337640762329,
+      "learning_rate": 4.537267816268739e-05,
+      "loss": 0.0055,
+      "step": 1359
+    },
+    {
+      "epoch": 0.9538839207434684,
+      "grad_norm": 0.9649966359138489,
+      "learning_rate": 4.536427137457072e-05,
+      "loss": 0.0098,
+      "step": 1360
+    },
+    {
+      "epoch": 0.9545853059793091,
+      "grad_norm": 28.19986915588379,
+      "learning_rate": 4.5355857737079166e-05,
+      "loss": 0.6184,
+      "step": 1361
+    },
+    {
+      "epoch": 0.95528669121515,
+      "grad_norm": 1.310431957244873,
+      "learning_rate": 4.534743725304257e-05,
+      "loss": 0.0142,
+      "step": 1362
+    },
+    {
+      "epoch": 0.9559880764509907,
+      "grad_norm": 1.170506477355957,
+      "learning_rate": 4.5339009925293105e-05,
+      "loss": 0.0125,
+      "step": 1363
+    },
+    {
+      "epoch": 0.9566894616868314,
+      "grad_norm": 1.135798454284668,
+      "learning_rate": 4.533057575666526e-05,
+      "loss": 0.0118,
+      "step": 1364
+    },
+    {
+      "epoch": 0.9573908469226723,
+      "grad_norm": 0.7438721060752869,
+      "learning_rate": 4.532213474999578e-05,
+      "loss": 0.0077,
+      "step": 1365
+    },
+    {
+      "epoch": 0.958092232158513,
+      "grad_norm": 0.46775007247924805,
+      "learning_rate": 4.531368690812376e-05,
+      "loss": 0.0046,
+      "step": 1366
+    },
+    {
+      "epoch": 0.9587936173943539,
+      "grad_norm": 0.2829347848892212,
+      "learning_rate": 4.530523223389056e-05,
+      "loss": 0.0027,
+      "step": 1367
+    },
+    {
+      "epoch": 0.9594950026301946,
+      "grad_norm": 0.12059792876243591,
+      "learning_rate": 4.529677073013984e-05,
+      "loss": 0.0011,
+      "step": 1368
+    },
+    {
+      "epoch": 0.9601963878660355,
+      "grad_norm": 41.76416015625,
+      "learning_rate": 4.5288302399717575e-05,
+      "loss": 1.5052,
+      "step": 1369
+    },
+    {
+      "epoch": 0.9608977731018762,
+      "grad_norm": 0.05551628768444061,
+      "learning_rate": 4.5279827245472025e-05,
+      "loss": 0.0005,
+      "step": 1370
+    },
+    {
+      "epoch": 0.961599158337717,
+      "grad_norm": 44.579280853271484,
+      "learning_rate": 4.527134527025374e-05,
+      "loss": 1.9268,
+      "step": 1371
+    },
+    {
+      "epoch": 0.9623005435735578,
+      "grad_norm": 92.77808380126953,
+      "learning_rate": 4.526285647691559e-05,
+      "loss": 3.8084,
+      "step": 1372
+    },
+    {
+      "epoch": 0.9630019288093986,
+      "grad_norm": 0.06515678018331528,
+      "learning_rate": 4.5254360868312694e-05,
+      "loss": 0.0006,
+      "step": 1373
+    },
+    {
+      "epoch": 0.9637033140452393,
+      "grad_norm": 44.31393814086914,
+      "learning_rate": 4.5245858447302506e-05,
+      "loss": 1.8936,
+      "step": 1374
+    },
+    {
+      "epoch": 0.9644046992810801,
+      "grad_norm": 42.123191833496094,
+      "learning_rate": 4.523734921674475e-05,
+      "loss": 1.5489,
+      "step": 1375
+    },
+    {
+      "epoch": 0.9651060845169209,
+      "grad_norm": 78.27152252197266,
+      "learning_rate": 4.522883317950144e-05,
+      "loss": 1.4444,
+      "step": 1376
+    },
+    {
+      "epoch": 0.9658074697527617,
+      "grad_norm": 0.4181966185569763,
+      "learning_rate": 4.5220310338436885e-05,
+      "loss": 0.0041,
+      "step": 1377
+    },
+    {
+      "epoch": 0.9665088549886025,
+      "grad_norm": 61.79771423339844,
+      "learning_rate": 4.521178069641768e-05,
+      "loss": 1.5033,
+      "step": 1378
+    },
+    {
+      "epoch": 0.9672102402244432,
+      "grad_norm": 1.026915431022644,
+      "learning_rate": 4.520324425631271e-05,
+      "loss": 0.011,
+      "step": 1379
+    },
+    {
+      "epoch": 0.9679116254602841,
+      "grad_norm": 1.1025463342666626,
+      "learning_rate": 4.519470102099315e-05,
+      "loss": 0.0117,
+      "step": 1380
+    },
+    {
+      "epoch": 0.9686130106961248,
+      "grad_norm": 1.5587780475616455,
+      "learning_rate": 4.518615099333243e-05,
+      "loss": 0.018,
+      "step": 1381
+    },
+    {
+      "epoch": 0.9693143959319657,
+      "grad_norm": 1.4015796184539795,
+      "learning_rate": 4.517759417620633e-05,
+      "loss": 0.016,
+      "step": 1382
+    },
+    {
+      "epoch": 0.9700157811678064,
+      "grad_norm": 1.1269611120224,
+      "learning_rate": 4.516903057249284e-05,
+      "loss": 0.0124,
+      "step": 1383
+    },
+    {
+      "epoch": 0.9707171664036472,
+      "grad_norm": 64.7952651977539,
+      "learning_rate": 4.516046018507229e-05,
+      "loss": 1.7053,
+      "step": 1384
+    },
+    {
+      "epoch": 0.971418551639488,
+      "grad_norm": 36.87698745727539,
+      "learning_rate": 4.515188301682725e-05,
+      "loss": 0.8805,
+      "step": 1385
+    },
+    {
+      "epoch": 0.9721199368753288,
+      "grad_norm": 39.0102653503418,
+      "learning_rate": 4.5143299070642584e-05,
+      "loss": 1.2212,
+      "step": 1386
+    },
+    {
+      "epoch": 0.9728213221111696,
+      "grad_norm": 62.35078811645508,
+      "learning_rate": 4.5134708349405457e-05,
+      "loss": 1.6357,
+      "step": 1387
+    },
+    {
+      "epoch": 0.9735227073470103,
+      "grad_norm": 31.161195755004883,
+      "learning_rate": 4.5126110856005285e-05,
+      "loss": 0.7379,
+      "step": 1388
+    },
+    {
+      "epoch": 0.9742240925828511,
+      "grad_norm": 1.4460724592208862,
+      "learning_rate": 4.511750659333379e-05,
+      "loss": 0.0166,
+      "step": 1389
+    },
+    {
+      "epoch": 0.9749254778186919,
+      "grad_norm": 1.6871577501296997,
+      "learning_rate": 4.5108895564284924e-05,
+      "loss": 0.0201,
+      "step": 1390
+    },
+    {
+      "epoch": 0.9756268630545327,
+      "grad_norm": 18.601655960083008,
+      "learning_rate": 4.510027777175497e-05,
+      "loss": 0.3773,
+      "step": 1391
+    },
+    {
+      "epoch": 0.9763282482903735,
+      "grad_norm": 24.591615676879883,
+      "learning_rate": 4.509165321864245e-05,
+      "loss": 0.6012,
+      "step": 1392
+    },
+    {
+      "epoch": 0.9770296335262143,
+      "grad_norm": 2.3374860286712646,
+      "learning_rate": 4.5083021907848164e-05,
+      "loss": 0.0286,
+      "step": 1393
+    },
+    {
+      "epoch": 0.977731018762055,
+      "grad_norm": 3.1211438179016113,
+      "learning_rate": 4.50743838422752e-05,
+      "loss": 0.0388,
+      "step": 1394
+    },
+    {
+      "epoch": 0.9784324039978959,
+      "grad_norm": 15.115840911865234,
+      "learning_rate": 4.50657390248289e-05,
+      "loss": 0.3451,
+      "step": 1395
+    },
+    {
+      "epoch": 0.9791337892337366,
+      "grad_norm": 13.081830978393555,
+      "learning_rate": 4.50570874584169e-05,
+      "loss": 0.3082,
+      "step": 1396
+    },
+    {
+      "epoch": 0.9798351744695775,
+      "grad_norm": 3.7496583461761475,
+      "learning_rate": 4.5048429145949076e-05,
+      "loss": 0.0503,
+      "step": 1397
+    },
+    {
+      "epoch": 0.9805365597054182,
+      "grad_norm": 22.880462646484375,
+      "learning_rate": 4.503976409033759e-05,
+      "loss": 0.4432,
+      "step": 1398
+    },
+    {
+      "epoch": 0.9812379449412589,
+      "grad_norm": 4.944215774536133,
+      "learning_rate": 4.503109229449688e-05,
+      "loss": 0.0706,
+      "step": 1399
+    },
+    {
+      "epoch": 0.9819393301770998,
+      "grad_norm": 4.719092845916748,
+      "learning_rate": 4.502241376134364e-05,
+      "loss": 0.0667,
+      "step": 1400
+    },
+    {
+      "epoch": 0.9826407154129405,
+      "grad_norm": 4.378533363342285,
+      "learning_rate": 4.5013728493796826e-05,
+      "loss": 0.0611,
+      "step": 1401
+    },
+    {
+      "epoch": 0.9833421006487814,
+      "grad_norm": 4.234534740447998,
+      "learning_rate": 4.5005036494777666e-05,
+      "loss": 0.055,
+      "step": 1402
+    },
+    {
+      "epoch": 0.9840434858846221,
+      "grad_norm": 2.942084789276123,
+      "learning_rate": 4.4996337767209656e-05,
+      "loss": 0.0382,
+      "step": 1403
+    },
+    {
+      "epoch": 0.9847448711204629,
+      "grad_norm": 1.5875838994979858,
+      "learning_rate": 4.498763231401855e-05,
+      "loss": 0.018,
+      "step": 1404
+    },
+    {
+      "epoch": 0.9854462563563037,
+      "grad_norm": 53.9989128112793,
+      "learning_rate": 4.497892013813235e-05,
+      "loss": 1.2552,
+      "step": 1405
+    },
+    {
+      "epoch": 0.9861476415921445,
+      "grad_norm": 28.49068832397461,
+      "learning_rate": 4.497020124248135e-05,
+      "loss": 0.7336,
+      "step": 1406
+    },
+    {
+      "epoch": 0.9868490268279853,
+      "grad_norm": 0.4417467713356018,
+      "learning_rate": 4.496147562999807e-05,
+      "loss": 0.0045,
+      "step": 1407
+    },
+    {
+      "epoch": 0.9875504120638261,
+      "grad_norm": 104.99906921386719,
+      "learning_rate": 4.4952743303617325e-05,
+      "loss": 3.0756,
+      "step": 1408
+    },
+    {
+      "epoch": 0.9882517972996668,
+      "grad_norm": 0.3293454647064209,
+      "learning_rate": 4.4944004266276164e-05,
+      "loss": 0.0033,
+      "step": 1409
+    },
+    {
+      "epoch": 0.9889531825355077,
+      "grad_norm": 41.633243560791016,
+      "learning_rate": 4.4935258520913894e-05,
+      "loss": 1.3174,
+      "step": 1410
+    },
+    {
+      "epoch": 0.9896545677713484,
+      "grad_norm": 0.23380160331726074,
+      "learning_rate": 4.4926506070472085e-05,
+      "loss": 0.0023,
+      "step": 1411
+    },
+    {
+      "epoch": 0.9903559530071891,
+      "grad_norm": 0.44888490438461304,
+      "learning_rate": 4.4917746917894556e-05,
+      "loss": 0.0047,
+      "step": 1412
+    },
+    {
+      "epoch": 0.99105733824303,
+      "grad_norm": 33.958248138427734,
+      "learning_rate": 4.490898106612739e-05,
+      "loss": 0.9667,
+      "step": 1413
+    },
+    {
+      "epoch": 0.9917587234788707,
+      "grad_norm": 0.3081722855567932,
+      "learning_rate": 4.490020851811892e-05,
+      "loss": 0.003,
+      "step": 1414
+    },
+    {
+      "epoch": 0.9924601087147116,
+      "grad_norm": 32.04133224487305,
+      "learning_rate": 4.4891429276819726e-05,
+      "loss": 0.9003,
+      "step": 1415
+    },
+    {
+      "epoch": 0.9931614939505523,
+      "grad_norm": 43.6740608215332,
+      "learning_rate": 4.488264334518264e-05,
+      "loss": 0.9288,
+      "step": 1416
+    },
+    {
+      "epoch": 0.9938628791863932,
+      "grad_norm": 29.450580596923828,
+      "learning_rate": 4.487385072616275e-05,
+      "loss": 0.7131,
+      "step": 1417
+    },
+    {
+      "epoch": 0.9945642644222339,
+      "grad_norm": 1.1718950271606445,
+      "learning_rate": 4.486505142271739e-05,
+      "loss": 0.0137,
+      "step": 1418
+    },
+    {
+      "epoch": 0.9952656496580747,
+      "grad_norm": 1.5927923917770386,
+      "learning_rate": 4.485624543780613e-05,
+      "loss": 0.0192,
+      "step": 1419
+    },
+    {
+      "epoch": 0.9959670348939155,
+      "grad_norm": 24.835670471191406,
+      "learning_rate": 4.4847432774390807e-05,
+      "loss": 0.5717,
+      "step": 1420
+    },
+    {
+      "epoch": 0.9966684201297563,
+      "grad_norm": 1.8400554656982422,
+      "learning_rate": 4.4838613435435505e-05,
+      "loss": 0.022,
+      "step": 1421
+    },
+    {
+      "epoch": 0.997369805365597,
+      "grad_norm": 2.3654353618621826,
+      "learning_rate": 4.482978742390654e-05,
+      "loss": 0.0298,
+      "step": 1422
+    },
+    {
+      "epoch": 0.9980711906014378,
+      "grad_norm": 24.44013023376465,
+      "learning_rate": 4.482095474277247e-05,
+      "loss": 0.5608,
+      "step": 1423
+    },
+    {
+      "epoch": 0.9987725758372786,
+      "grad_norm": 2.5648183822631836,
+      "learning_rate": 4.48121153950041e-05,
+      "loss": 0.034,
+      "step": 1424
+    },
+    {
+      "epoch": 0.9994739610731194,
+      "grad_norm": 2.0194385051727295,
+      "learning_rate": 4.4803269383574507e-05,
+      "loss": 0.0249,
+      "step": 1425
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1425,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.863941975950848e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}