diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,20552 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 2930,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0017064846416382253,
+      "grad_norm": 5.5382735829047816,
+      "learning_rate": 1.3651877133105803e-07,
+      "loss": 0.9275,
+      "step": 1
+    },
+    {
+      "epoch": 0.0034129692832764505,
+      "grad_norm": 5.560911534512389,
+      "learning_rate": 2.7303754266211607e-07,
+      "loss": 0.8735,
+      "step": 2
+    },
+    {
+      "epoch": 0.005119453924914676,
+      "grad_norm": 5.571629079857404,
+      "learning_rate": 4.0955631399317407e-07,
+      "loss": 0.8948,
+      "step": 3
+    },
+    {
+      "epoch": 0.006825938566552901,
+      "grad_norm": 5.399904062260569,
+      "learning_rate": 5.460750853242321e-07,
+      "loss": 0.8989,
+      "step": 4
+    },
+    {
+      "epoch": 0.008532423208191127,
+      "grad_norm": 5.520885319552546,
+      "learning_rate": 6.825938566552902e-07,
+      "loss": 0.8903,
+      "step": 5
+    },
+    {
+      "epoch": 0.010238907849829351,
+      "grad_norm": 5.516912402763377,
+      "learning_rate": 8.191126279863481e-07,
+      "loss": 0.8867,
+      "step": 6
+    },
+    {
+      "epoch": 0.011945392491467578,
+      "grad_norm": 5.557317289645866,
+      "learning_rate": 9.556313993174062e-07,
+      "loss": 0.8591,
+      "step": 7
+    },
+    {
+      "epoch": 0.013651877133105802,
+      "grad_norm": 5.045508456607984,
+      "learning_rate": 1.0921501706484643e-06,
+      "loss": 0.8438,
+      "step": 8
+    },
+    {
+      "epoch": 0.015358361774744027,
+      "grad_norm": 4.867775053276732,
+      "learning_rate": 1.2286689419795223e-06,
+      "loss": 0.8324,
+      "step": 9
+    },
+    {
+      "epoch": 0.017064846416382253,
+      "grad_norm": 4.17269521078131,
+      "learning_rate": 1.3651877133105804e-06,
+      "loss": 0.7919,
+      "step": 10
+    },
+    {
+      "epoch": 0.01877133105802048,
+      "grad_norm": 3.8684320532409604,
+      "learning_rate": 1.5017064846416384e-06,
+      "loss": 0.7654,
+      "step": 11
+    },
+    {
+      "epoch": 0.020477815699658702,
+      "grad_norm": 4.0237028768787875,
+      "learning_rate": 1.6382252559726963e-06,
+      "loss": 0.8797,
+      "step": 12
+    },
+    {
+      "epoch": 0.02218430034129693,
+      "grad_norm": 2.3381353337879407,
+      "learning_rate": 1.7747440273037543e-06,
+      "loss": 0.7645,
+      "step": 13
+    },
+    {
+      "epoch": 0.023890784982935155,
+      "grad_norm": 2.2908419509729003,
+      "learning_rate": 1.9112627986348124e-06,
+      "loss": 0.7652,
+      "step": 14
+    },
+    {
+      "epoch": 0.025597269624573378,
+      "grad_norm": 2.1380170510125645,
+      "learning_rate": 2.0477815699658705e-06,
+      "loss": 0.8098,
+      "step": 15
+    },
+    {
+      "epoch": 0.027303754266211604,
+      "grad_norm": 1.9451355783607645,
+      "learning_rate": 2.1843003412969285e-06,
+      "loss": 0.7624,
+      "step": 16
+    },
+    {
+      "epoch": 0.02901023890784983,
+      "grad_norm": 2.4088929607724725,
+      "learning_rate": 2.3208191126279866e-06,
+      "loss": 0.747,
+      "step": 17
+    },
+    {
+      "epoch": 0.030716723549488054,
+      "grad_norm": 3.075601225318945,
+      "learning_rate": 2.4573378839590446e-06,
+      "loss": 0.7657,
+      "step": 18
+    },
+    {
+      "epoch": 0.032423208191126277,
+      "grad_norm": 3.216111415255342,
+      "learning_rate": 2.5938566552901023e-06,
+      "loss": 0.7787,
+      "step": 19
+    },
+    {
+      "epoch": 0.034129692832764506,
+      "grad_norm": 2.93934194565305,
+      "learning_rate": 2.7303754266211608e-06,
+      "loss": 0.7258,
+      "step": 20
+    },
+    {
+      "epoch": 0.03583617747440273,
+      "grad_norm": 2.8243675906534116,
+      "learning_rate": 2.8668941979522184e-06,
+      "loss": 0.7321,
+      "step": 21
+    },
+    {
+      "epoch": 0.03754266211604096,
+      "grad_norm": 2.782095780866637,
+      "learning_rate": 3.003412969283277e-06,
+      "loss": 0.7147,
+      "step": 22
+    },
+    {
+      "epoch": 0.03924914675767918,
+      "grad_norm": 2.1270711160975204,
+      "learning_rate": 3.139931740614335e-06,
+      "loss": 0.6972,
+      "step": 23
+    },
+    {
+      "epoch": 0.040955631399317405,
+      "grad_norm": 1.6607845981555187,
+      "learning_rate": 3.2764505119453926e-06,
+      "loss": 0.7115,
+      "step": 24
+    },
+    {
+      "epoch": 0.042662116040955635,
+      "grad_norm": 1.4607261467117063,
+      "learning_rate": 3.412969283276451e-06,
+      "loss": 0.6944,
+      "step": 25
+    },
+    {
+      "epoch": 0.04436860068259386,
+      "grad_norm": 1.1103212511747629,
+      "learning_rate": 3.5494880546075087e-06,
+      "loss": 0.6621,
+      "step": 26
+    },
+    {
+      "epoch": 0.04607508532423208,
+      "grad_norm": 1.0962073209819554,
+      "learning_rate": 3.6860068259385667e-06,
+      "loss": 0.633,
+      "step": 27
+    },
+    {
+      "epoch": 0.04778156996587031,
+      "grad_norm": 1.2994991639042826,
+      "learning_rate": 3.822525597269625e-06,
+      "loss": 0.6254,
+      "step": 28
+    },
+    {
+      "epoch": 0.04948805460750853,
+      "grad_norm": 1.24110998413488,
+      "learning_rate": 3.959044368600683e-06,
+      "loss": 0.6478,
+      "step": 29
+    },
+    {
+      "epoch": 0.051194539249146756,
+      "grad_norm": 1.377261198618834,
+      "learning_rate": 4.095563139931741e-06,
+      "loss": 0.692,
+      "step": 30
+    },
+    {
+      "epoch": 0.052901023890784986,
+      "grad_norm": 1.0661079292122833,
+      "learning_rate": 4.232081911262799e-06,
+      "loss": 0.6809,
+      "step": 31
+    },
+    {
+      "epoch": 0.05460750853242321,
+      "grad_norm": 0.9607749961051216,
+      "learning_rate": 4.368600682593857e-06,
+      "loss": 0.6222,
+      "step": 32
+    },
+    {
+      "epoch": 0.05631399317406143,
+      "grad_norm": 0.8427434655009416,
+      "learning_rate": 4.505119453924915e-06,
+      "loss": 0.6109,
+      "step": 33
+    },
+    {
+      "epoch": 0.05802047781569966,
+      "grad_norm": 0.7136777520094485,
+      "learning_rate": 4.641638225255973e-06,
+      "loss": 0.6084,
+      "step": 34
+    },
+    {
+      "epoch": 0.059726962457337884,
+      "grad_norm": 0.8210756508341455,
+      "learning_rate": 4.778156996587031e-06,
+      "loss": 0.6158,
+      "step": 35
+    },
+    {
+      "epoch": 0.06143344709897611,
+      "grad_norm": 1.0150335798080576,
+      "learning_rate": 4.914675767918089e-06,
+      "loss": 0.6199,
+      "step": 36
+    },
+    {
+      "epoch": 0.06313993174061433,
+      "grad_norm": 0.9063806967726155,
+      "learning_rate": 5.051194539249147e-06,
+      "loss": 0.6453,
+      "step": 37
+    },
+    {
+      "epoch": 0.06484641638225255,
+      "grad_norm": 0.7068201160652776,
+      "learning_rate": 5.1877133105802046e-06,
+      "loss": 0.6104,
+      "step": 38
+    },
+    {
+      "epoch": 0.06655290102389079,
+      "grad_norm": 0.638263361178943,
+      "learning_rate": 5.324232081911264e-06,
+      "loss": 0.59,
+      "step": 39
+    },
+    {
+      "epoch": 0.06825938566552901,
+      "grad_norm": 0.6916379175425308,
+      "learning_rate": 5.4607508532423215e-06,
+      "loss": 0.5712,
+      "step": 40
+    },
+    {
+      "epoch": 0.06996587030716724,
+      "grad_norm": 0.8611281068730197,
+      "learning_rate": 5.597269624573379e-06,
+      "loss": 0.6005,
+      "step": 41
+    },
+    {
+      "epoch": 0.07167235494880546,
+      "grad_norm": 0.8014386934383825,
+      "learning_rate": 5.733788395904437e-06,
+      "loss": 0.573,
+      "step": 42
+    },
+    {
+      "epoch": 0.07337883959044368,
+      "grad_norm": 0.6815974545922857,
+      "learning_rate": 5.870307167235495e-06,
+      "loss": 0.6091,
+      "step": 43
+    },
+    {
+      "epoch": 0.07508532423208192,
+      "grad_norm": 0.601937526209857,
+      "learning_rate": 6.006825938566554e-06,
+      "loss": 0.5869,
+      "step": 44
+    },
+    {
+      "epoch": 0.07679180887372014,
+      "grad_norm": 0.6872760313892801,
+      "learning_rate": 6.143344709897611e-06,
+      "loss": 0.6399,
+      "step": 45
+    },
+    {
+      "epoch": 0.07849829351535836,
+      "grad_norm": 0.5584926915208507,
+      "learning_rate": 6.27986348122867e-06,
+      "loss": 0.6031,
+      "step": 46
+    },
+    {
+      "epoch": 0.08020477815699659,
+      "grad_norm": 0.7413461854532756,
+      "learning_rate": 6.4163822525597275e-06,
+      "loss": 0.6285,
+      "step": 47
+    },
+    {
+      "epoch": 0.08191126279863481,
+      "grad_norm": 0.7484828621746611,
+      "learning_rate": 6.552901023890785e-06,
+      "loss": 0.6211,
+      "step": 48
+    },
+    {
+      "epoch": 0.08361774744027303,
+      "grad_norm": 0.59460985132373,
+      "learning_rate": 6.689419795221843e-06,
+      "loss": 0.6042,
+      "step": 49
+    },
+    {
+      "epoch": 0.08532423208191127,
+      "grad_norm": 0.5303831678889555,
+      "learning_rate": 6.825938566552902e-06,
+      "loss": 0.5623,
+      "step": 50
+    },
+    {
+      "epoch": 0.08703071672354949,
+      "grad_norm": 0.5993589782434914,
+      "learning_rate": 6.96245733788396e-06,
+      "loss": 0.6152,
+      "step": 51
+    },
+    {
+      "epoch": 0.08873720136518772,
+      "grad_norm": 0.7099366121307583,
+      "learning_rate": 7.098976109215017e-06,
+      "loss": 0.5996,
+      "step": 52
+    },
+    {
+      "epoch": 0.09044368600682594,
+      "grad_norm": 0.5195563613459195,
+      "learning_rate": 7.235494880546076e-06,
+      "loss": 0.586,
+      "step": 53
+    },
+    {
+      "epoch": 0.09215017064846416,
+      "grad_norm": 0.5328364263084424,
+      "learning_rate": 7.3720136518771335e-06,
+      "loss": 0.5986,
+      "step": 54
+    },
+    {
+      "epoch": 0.09385665529010238,
+      "grad_norm": 0.59504684976116,
+      "learning_rate": 7.508532423208191e-06,
+      "loss": 0.5781,
+      "step": 55
+    },
+    {
+      "epoch": 0.09556313993174062,
+      "grad_norm": 0.5423417430735504,
+      "learning_rate": 7.64505119453925e-06,
+      "loss": 0.5317,
+      "step": 56
+    },
+    {
+      "epoch": 0.09726962457337884,
+      "grad_norm": 0.6595692463709092,
+      "learning_rate": 7.781569965870308e-06,
+      "loss": 0.5683,
+      "step": 57
+    },
+    {
+      "epoch": 0.09897610921501707,
+      "grad_norm": 0.7062945077387907,
+      "learning_rate": 7.918088737201367e-06,
+      "loss": 0.6021,
+      "step": 58
+    },
+    {
+      "epoch": 0.10068259385665529,
+      "grad_norm": 0.5117294569725397,
+      "learning_rate": 8.054607508532423e-06,
+      "loss": 0.573,
+      "step": 59
+    },
+    {
+      "epoch": 0.10238907849829351,
+      "grad_norm": 0.5628977910968079,
+      "learning_rate": 8.191126279863482e-06,
+      "loss": 0.5784,
+      "step": 60
+    },
+    {
+      "epoch": 0.10409556313993173,
+      "grad_norm": 0.561096384808619,
+      "learning_rate": 8.327645051194539e-06,
+      "loss": 0.5733,
+      "step": 61
+    },
+    {
+      "epoch": 0.10580204778156997,
+      "grad_norm": 0.6314272293109521,
+      "learning_rate": 8.464163822525599e-06,
+      "loss": 0.5638,
+      "step": 62
+    },
+    {
+      "epoch": 0.1075085324232082,
+      "grad_norm": 0.6462442946319007,
+      "learning_rate": 8.600682593856656e-06,
+      "loss": 0.5847,
+      "step": 63
+    },
+    {
+      "epoch": 0.10921501706484642,
+      "grad_norm": 0.4906413352665859,
+      "learning_rate": 8.737201365187714e-06,
+      "loss": 0.5353,
+      "step": 64
+    },
+    {
+      "epoch": 0.11092150170648464,
+      "grad_norm": 0.5876681283539944,
+      "learning_rate": 8.873720136518773e-06,
+      "loss": 0.5841,
+      "step": 65
+    },
+    {
+      "epoch": 0.11262798634812286,
+      "grad_norm": 0.5448387276272628,
+      "learning_rate": 9.01023890784983e-06,
+      "loss": 0.5838,
+      "step": 66
+    },
+    {
+      "epoch": 0.11433447098976109,
+      "grad_norm": 0.5664187342399059,
+      "learning_rate": 9.146757679180888e-06,
+      "loss": 0.5755,
+      "step": 67
+    },
+    {
+      "epoch": 0.11604095563139932,
+      "grad_norm": 0.5701448535010463,
+      "learning_rate": 9.283276450511946e-06,
+      "loss": 0.6011,
+      "step": 68
+    },
+    {
+      "epoch": 0.11774744027303755,
+      "grad_norm": 0.5958601312411905,
+      "learning_rate": 9.419795221843005e-06,
+      "loss": 0.5972,
+      "step": 69
+    },
+    {
+      "epoch": 0.11945392491467577,
+      "grad_norm": 0.5505177998669957,
+      "learning_rate": 9.556313993174062e-06,
+      "loss": 0.5856,
+      "step": 70
+    },
+    {
+      "epoch": 0.12116040955631399,
+      "grad_norm": 0.6211369988588423,
+      "learning_rate": 9.69283276450512e-06,
+      "loss": 0.5607,
+      "step": 71
+    },
+    {
+      "epoch": 0.12286689419795221,
+      "grad_norm": 0.5830089207049569,
+      "learning_rate": 9.829351535836179e-06,
+      "loss": 0.5609,
+      "step": 72
+    },
+    {
+      "epoch": 0.12457337883959044,
+      "grad_norm": 0.5771202366609306,
+      "learning_rate": 9.965870307167235e-06,
+      "loss": 0.5753,
+      "step": 73
+    },
+    {
+      "epoch": 0.12627986348122866,
+      "grad_norm": 0.5512514871122207,
+      "learning_rate": 1.0102389078498294e-05,
+      "loss": 0.5501,
+      "step": 74
+    },
+    {
+      "epoch": 0.12798634812286688,
+      "grad_norm": 0.5921576744642906,
+      "learning_rate": 1.0238907849829352e-05,
+      "loss": 0.588,
+      "step": 75
+    },
+    {
+      "epoch": 0.1296928327645051,
+      "grad_norm": 0.6279629362229029,
+      "learning_rate": 1.0375426621160409e-05,
+      "loss": 0.596,
+      "step": 76
+    },
+    {
+      "epoch": 0.13139931740614336,
+      "grad_norm": 0.5328651531767897,
+      "learning_rate": 1.051194539249147e-05,
+      "loss": 0.6041,
+      "step": 77
+    },
+    {
+      "epoch": 0.13310580204778158,
+      "grad_norm": 0.5245734699897833,
+      "learning_rate": 1.0648464163822528e-05,
+      "loss": 0.5617,
+      "step": 78
+    },
+    {
+      "epoch": 0.1348122866894198,
+      "grad_norm": 0.5820231240535302,
+      "learning_rate": 1.0784982935153585e-05,
+      "loss": 0.5996,
+      "step": 79
+    },
+    {
+      "epoch": 0.13651877133105803,
+      "grad_norm": 0.5787551695285894,
+      "learning_rate": 1.0921501706484643e-05,
+      "loss": 0.5913,
+      "step": 80
+    },
+    {
+      "epoch": 0.13822525597269625,
+      "grad_norm": 0.5775924449404031,
+      "learning_rate": 1.1058020477815702e-05,
+      "loss": 0.5504,
+      "step": 81
+    },
+    {
+      "epoch": 0.13993174061433447,
+      "grad_norm": 0.6000730341856517,
+      "learning_rate": 1.1194539249146758e-05,
+      "loss": 0.5753,
+      "step": 82
+    },
+    {
+      "epoch": 0.1416382252559727,
+      "grad_norm": 0.5027518145184813,
+      "learning_rate": 1.1331058020477817e-05,
+      "loss": 0.5215,
+      "step": 83
+    },
+    {
+      "epoch": 0.14334470989761092,
+      "grad_norm": 0.665396255158122,
+      "learning_rate": 1.1467576791808874e-05,
+      "loss": 0.5909,
+      "step": 84
+    },
+    {
+      "epoch": 0.14505119453924914,
+      "grad_norm": 0.6507897383778756,
+      "learning_rate": 1.1604095563139932e-05,
+      "loss": 0.5719,
+      "step": 85
+    },
+    {
+      "epoch": 0.14675767918088736,
+      "grad_norm": 0.5608661072691324,
+      "learning_rate": 1.174061433447099e-05,
+      "loss": 0.5437,
+      "step": 86
+    },
+    {
+      "epoch": 0.14846416382252559,
+      "grad_norm": 0.5915603711419299,
+      "learning_rate": 1.1877133105802047e-05,
+      "loss": 0.5533,
+      "step": 87
+    },
+    {
+      "epoch": 0.15017064846416384,
+      "grad_norm": 0.6332333514132811,
+      "learning_rate": 1.2013651877133108e-05,
+      "loss": 0.5432,
+      "step": 88
+    },
+    {
+      "epoch": 0.15187713310580206,
+      "grad_norm": 0.5868221722267082,
+      "learning_rate": 1.2150170648464166e-05,
+      "loss": 0.5495,
+      "step": 89
+    },
+    {
+      "epoch": 0.15358361774744028,
+      "grad_norm": 0.7081815015621716,
+      "learning_rate": 1.2286689419795223e-05,
+      "loss": 0.5258,
+      "step": 90
+    },
+    {
+      "epoch": 0.1552901023890785,
+      "grad_norm": 0.71392633614486,
+      "learning_rate": 1.2423208191126281e-05,
+      "loss": 0.5762,
+      "step": 91
+    },
+    {
+      "epoch": 0.15699658703071673,
+      "grad_norm": 0.6981523026906576,
+      "learning_rate": 1.255972696245734e-05,
+      "loss": 0.582,
+      "step": 92
+    },
+    {
+      "epoch": 0.15870307167235495,
+      "grad_norm": 0.6171926262629204,
+      "learning_rate": 1.2696245733788397e-05,
+      "loss": 0.5607,
+      "step": 93
+    },
+    {
+      "epoch": 0.16040955631399317,
+      "grad_norm": 0.5545169364183319,
+      "learning_rate": 1.2832764505119455e-05,
+      "loss": 0.5247,
+      "step": 94
+    },
+    {
+      "epoch": 0.1621160409556314,
+      "grad_norm": 0.7302617590838634,
+      "learning_rate": 1.2969283276450513e-05,
+      "loss": 0.5377,
+      "step": 95
+    },
+    {
+      "epoch": 0.16382252559726962,
+      "grad_norm": 0.6212142146844655,
+      "learning_rate": 1.310580204778157e-05,
+      "loss": 0.5573,
+      "step": 96
+    },
+    {
+      "epoch": 0.16552901023890784,
+      "grad_norm": 0.7259534885617567,
+      "learning_rate": 1.3242320819112629e-05,
+      "loss": 0.5702,
+      "step": 97
+    },
+    {
+      "epoch": 0.16723549488054607,
+      "grad_norm": 0.690308401698924,
+      "learning_rate": 1.3378839590443686e-05,
+      "loss": 0.4928,
+      "step": 98
+    },
+    {
+      "epoch": 0.1689419795221843,
+      "grad_norm": 0.5541121051324966,
+      "learning_rate": 1.3515358361774744e-05,
+      "loss": 0.5138,
+      "step": 99
+    },
+    {
+      "epoch": 0.17064846416382254,
+      "grad_norm": 0.6225627341053582,
+      "learning_rate": 1.3651877133105804e-05,
+      "loss": 0.5467,
+      "step": 100
+    },
+    {
+      "epoch": 0.17235494880546076,
+      "grad_norm": 0.6549365302580259,
+      "learning_rate": 1.3788395904436863e-05,
+      "loss": 0.5784,
+      "step": 101
+    },
+    {
+      "epoch": 0.17406143344709898,
+      "grad_norm": 0.5738375615533333,
+      "learning_rate": 1.392491467576792e-05,
+      "loss": 0.5478,
+      "step": 102
+    },
+    {
+      "epoch": 0.1757679180887372,
+      "grad_norm": 0.7174652077251786,
+      "learning_rate": 1.4061433447098978e-05,
+      "loss": 0.5154,
+      "step": 103
+    },
+    {
+      "epoch": 0.17747440273037543,
+      "grad_norm": 0.6259420624867417,
+      "learning_rate": 1.4197952218430035e-05,
+      "loss": 0.5328,
+      "step": 104
+    },
+    {
+      "epoch": 0.17918088737201365,
+      "grad_norm": 0.6980024100865326,
+      "learning_rate": 1.4334470989761093e-05,
+      "loss": 0.5543,
+      "step": 105
+    },
+    {
+      "epoch": 0.18088737201365188,
+      "grad_norm": 0.592418945044328,
+      "learning_rate": 1.4470989761092152e-05,
+      "loss": 0.5487,
+      "step": 106
+    },
+    {
+      "epoch": 0.1825938566552901,
+      "grad_norm": 0.7138884575404962,
+      "learning_rate": 1.4607508532423209e-05,
+      "loss": 0.5359,
+      "step": 107
+    },
+    {
+      "epoch": 0.18430034129692832,
+      "grad_norm": 0.601790646337058,
+      "learning_rate": 1.4744027303754267e-05,
+      "loss": 0.6079,
+      "step": 108
+    },
+    {
+      "epoch": 0.18600682593856654,
+      "grad_norm": 0.7129207866642714,
+      "learning_rate": 1.4880546075085325e-05,
+      "loss": 0.5701,
+      "step": 109
+    },
+    {
+      "epoch": 0.18771331058020477,
+      "grad_norm": 0.6382641762018992,
+      "learning_rate": 1.5017064846416382e-05,
+      "loss": 0.5662,
+      "step": 110
+    },
+    {
+      "epoch": 0.189419795221843,
+      "grad_norm": 0.697222342232993,
+      "learning_rate": 1.515358361774744e-05,
+      "loss": 0.5631,
+      "step": 111
+    },
+    {
+      "epoch": 0.19112627986348124,
+      "grad_norm": 0.627950035362892,
+      "learning_rate": 1.52901023890785e-05,
+      "loss": 0.5299,
+      "step": 112
+    },
+    {
+      "epoch": 0.19283276450511946,
+      "grad_norm": 0.7254030376716648,
+      "learning_rate": 1.5426621160409558e-05,
+      "loss": 0.5587,
+      "step": 113
+    },
+    {
+      "epoch": 0.1945392491467577,
+      "grad_norm": 0.6339029349597561,
+      "learning_rate": 1.5563139931740616e-05,
+      "loss": 0.5233,
+      "step": 114
+    },
+    {
+      "epoch": 0.1962457337883959,
+      "grad_norm": 0.6136035997518353,
+      "learning_rate": 1.5699658703071675e-05,
+      "loss": 0.5658,
+      "step": 115
+    },
+    {
+      "epoch": 0.19795221843003413,
+      "grad_norm": 0.638647805167733,
+      "learning_rate": 1.5836177474402733e-05,
+      "loss": 0.5164,
+      "step": 116
+    },
+    {
+      "epoch": 0.19965870307167236,
+      "grad_norm": 0.5708056648144988,
+      "learning_rate": 1.5972696245733788e-05,
+      "loss": 0.5721,
+      "step": 117
+    },
+    {
+      "epoch": 0.20136518771331058,
+      "grad_norm": 0.582867595107491,
+      "learning_rate": 1.6109215017064847e-05,
+      "loss": 0.5087,
+      "step": 118
+    },
+    {
+      "epoch": 0.2030716723549488,
+      "grad_norm": 0.7041608007844309,
+      "learning_rate": 1.6245733788395905e-05,
+      "loss": 0.5429,
+      "step": 119
+    },
+    {
+      "epoch": 0.20477815699658702,
+      "grad_norm": 0.7296029016815011,
+      "learning_rate": 1.6382252559726964e-05,
+      "loss": 0.523,
+      "step": 120
+    },
+    {
+      "epoch": 0.20648464163822525,
+      "grad_norm": 0.7187410525850392,
+      "learning_rate": 1.6518771331058022e-05,
+      "loss": 0.5277,
+      "step": 121
+    },
+    {
+      "epoch": 0.20819112627986347,
+      "grad_norm": 0.7371160136610254,
+      "learning_rate": 1.6655290102389077e-05,
+      "loss": 0.5416,
+      "step": 122
+    },
+    {
+      "epoch": 0.2098976109215017,
+      "grad_norm": 0.705271247094977,
+      "learning_rate": 1.6791808873720136e-05,
+      "loss": 0.5919,
+      "step": 123
+    },
+    {
+      "epoch": 0.21160409556313994,
+      "grad_norm": 0.6215027705274428,
+      "learning_rate": 1.6928327645051198e-05,
+      "loss": 0.5458,
+      "step": 124
+    },
+    {
+      "epoch": 0.21331058020477817,
+      "grad_norm": 0.5530765715474965,
+      "learning_rate": 1.7064846416382256e-05,
+      "loss": 0.5468,
+      "step": 125
+    },
+    {
+      "epoch": 0.2150170648464164,
+      "grad_norm": 0.6835601219469564,
+      "learning_rate": 1.720136518771331e-05,
+      "loss": 0.506,
+      "step": 126
+    },
+    {
+      "epoch": 0.2167235494880546,
+      "grad_norm": 0.677833683372293,
+      "learning_rate": 1.733788395904437e-05,
+      "loss": 0.5408,
+      "step": 127
+    },
+    {
+      "epoch": 0.21843003412969283,
+      "grad_norm": 0.6494275011141496,
+      "learning_rate": 1.7474402730375428e-05,
+      "loss": 0.555,
+      "step": 128
+    },
+    {
+      "epoch": 0.22013651877133106,
+      "grad_norm": 0.726392896816137,
+      "learning_rate": 1.7610921501706487e-05,
+      "loss": 0.5315,
+      "step": 129
+    },
+    {
+      "epoch": 0.22184300341296928,
+      "grad_norm": 0.5611722977626344,
+      "learning_rate": 1.7747440273037545e-05,
+      "loss": 0.5089,
+      "step": 130
+    },
+    {
+      "epoch": 0.2235494880546075,
+      "grad_norm": 0.6652759423536284,
+      "learning_rate": 1.78839590443686e-05,
+      "loss": 0.5568,
+      "step": 131
+    },
+    {
+      "epoch": 0.22525597269624573,
+      "grad_norm": 0.7259929707728502,
+      "learning_rate": 1.802047781569966e-05,
+      "loss": 0.5351,
+      "step": 132
+    },
+    {
+      "epoch": 0.22696245733788395,
+      "grad_norm": 0.6196339136069524,
+      "learning_rate": 1.8156996587030717e-05,
+      "loss": 0.5078,
+      "step": 133
+    },
+    {
+      "epoch": 0.22866894197952217,
+      "grad_norm": 0.6352424805125334,
+      "learning_rate": 1.8293515358361776e-05,
+      "loss": 0.5305,
+      "step": 134
+    },
+    {
+      "epoch": 0.23037542662116042,
+      "grad_norm": 0.6756053919720748,
+      "learning_rate": 1.8430034129692834e-05,
+      "loss": 0.595,
+      "step": 135
+    },
+    {
+      "epoch": 0.23208191126279865,
+      "grad_norm": 0.7257648478517666,
+      "learning_rate": 1.8566552901023893e-05,
+      "loss": 0.5047,
+      "step": 136
+    },
+    {
+      "epoch": 0.23378839590443687,
+      "grad_norm": 0.6020515565047796,
+      "learning_rate": 1.870307167235495e-05,
+      "loss": 0.51,
+      "step": 137
+    },
+    {
+      "epoch": 0.2354948805460751,
+      "grad_norm": 0.6805295832228313,
+      "learning_rate": 1.883959044368601e-05,
+      "loss": 0.5564,
+      "step": 138
+    },
+    {
+      "epoch": 0.23720136518771331,
+      "grad_norm": 0.6016742179261152,
+      "learning_rate": 1.8976109215017068e-05,
+      "loss": 0.5324,
+      "step": 139
+    },
+    {
+      "epoch": 0.23890784982935154,
+      "grad_norm": 0.7159623015806224,
+      "learning_rate": 1.9112627986348123e-05,
+      "loss": 0.5604,
+      "step": 140
+    },
+    {
+      "epoch": 0.24061433447098976,
+      "grad_norm": 0.5731729805511779,
+      "learning_rate": 1.924914675767918e-05,
+      "loss": 0.5536,
+      "step": 141
+    },
+    {
+      "epoch": 0.24232081911262798,
+      "grad_norm": 0.6701475664152404,
+      "learning_rate": 1.938566552901024e-05,
+      "loss": 0.5562,
+      "step": 142
+    },
+    {
+      "epoch": 0.2440273037542662,
+      "grad_norm": 0.6546936591609688,
+      "learning_rate": 1.95221843003413e-05,
+      "loss": 0.5069,
+      "step": 143
+    },
+    {
+      "epoch": 0.24573378839590443,
+      "grad_norm": 0.6563483278900194,
+      "learning_rate": 1.9658703071672357e-05,
+      "loss": 0.5073,
+      "step": 144
+    },
+    {
+      "epoch": 0.24744027303754265,
+      "grad_norm": 0.6528440297406726,
+      "learning_rate": 1.9795221843003412e-05,
+      "loss": 0.5237,
+      "step": 145
+    },
+    {
+      "epoch": 0.24914675767918087,
+      "grad_norm": 0.8771673210309611,
+      "learning_rate": 1.993174061433447e-05,
+      "loss": 0.5513,
+      "step": 146
+    },
+    {
+      "epoch": 0.2508532423208191,
+      "grad_norm": 0.6642617286794137,
+      "learning_rate": 2.0068259385665533e-05,
+      "loss": 0.5514,
+      "step": 147
+    },
+    {
+      "epoch": 0.2525597269624573,
+      "grad_norm": 0.6735728295163484,
+      "learning_rate": 2.0204778156996588e-05,
+      "loss": 0.4947,
+      "step": 148
+    },
+    {
+      "epoch": 0.25426621160409557,
+      "grad_norm": 0.705298539331165,
+      "learning_rate": 2.0341296928327646e-05,
+      "loss": 0.5239,
+      "step": 149
+    },
+    {
+      "epoch": 0.25597269624573377,
+      "grad_norm": 0.8255167567788236,
+      "learning_rate": 2.0477815699658705e-05,
+      "loss": 0.5451,
+      "step": 150
+    },
+    {
+      "epoch": 0.257679180887372,
+      "grad_norm": 0.6253757777768068,
+      "learning_rate": 2.0614334470989763e-05,
+      "loss": 0.5471,
+      "step": 151
+    },
+    {
+      "epoch": 0.2593856655290102,
+      "grad_norm": 0.6655165402469364,
+      "learning_rate": 2.0750853242320818e-05,
+      "loss": 0.5169,
+      "step": 152
+    },
+    {
+      "epoch": 0.26109215017064846,
+      "grad_norm": 0.7135634503327206,
+      "learning_rate": 2.088737201365188e-05,
+      "loss": 0.5764,
+      "step": 153
+    },
+    {
+      "epoch": 0.2627986348122867,
+      "grad_norm": 0.557034065124935,
+      "learning_rate": 2.102389078498294e-05,
+      "loss": 0.5264,
+      "step": 154
+    },
+    {
+      "epoch": 0.2645051194539249,
+      "grad_norm": 0.7903317566711652,
+      "learning_rate": 2.1160409556313994e-05,
+      "loss": 0.5408,
+      "step": 155
+    },
+    {
+      "epoch": 0.26621160409556316,
+      "grad_norm": 0.5705963001825619,
+      "learning_rate": 2.1296928327645056e-05,
+      "loss": 0.5145,
+      "step": 156
+    },
+    {
+      "epoch": 0.26791808873720135,
+      "grad_norm": 0.7379094063463202,
+      "learning_rate": 2.143344709897611e-05,
+      "loss": 0.5544,
+      "step": 157
+    },
+    {
+      "epoch": 0.2696245733788396,
+      "grad_norm": 0.5092582654098079,
+      "learning_rate": 2.156996587030717e-05,
+      "loss": 0.5334,
+      "step": 158
+    },
+    {
+      "epoch": 0.2713310580204778,
+      "grad_norm": 0.7637584549009019,
+      "learning_rate": 2.1706484641638224e-05,
+      "loss": 0.5629,
+      "step": 159
+    },
+    {
+      "epoch": 0.27303754266211605,
+      "grad_norm": 0.6039638458104883,
+      "learning_rate": 2.1843003412969286e-05,
+      "loss": 0.4974,
+      "step": 160
+    },
+    {
+      "epoch": 0.27474402730375425,
+      "grad_norm": 0.6339412678654934,
+      "learning_rate": 2.197952218430034e-05,
+      "loss": 0.5186,
+      "step": 161
+    },
+    {
+      "epoch": 0.2764505119453925,
+      "grad_norm": 0.6411901060597484,
+      "learning_rate": 2.2116040955631403e-05,
+      "loss": 0.5544,
+      "step": 162
+    },
+    {
+      "epoch": 0.2781569965870307,
+      "grad_norm": 0.5670224727349668,
+      "learning_rate": 2.2252559726962458e-05,
+      "loss": 0.5144,
+      "step": 163
+    },
+    {
+      "epoch": 0.27986348122866894,
+      "grad_norm": 0.6239260800233626,
+      "learning_rate": 2.2389078498293517e-05,
+      "loss": 0.5236,
+      "step": 164
+    },
+    {
+      "epoch": 0.2815699658703072,
+      "grad_norm": 0.6347144073249235,
+      "learning_rate": 2.252559726962458e-05,
+      "loss": 0.5449,
+      "step": 165
+    },
+    {
+      "epoch": 0.2832764505119454,
+      "grad_norm": 0.5902846060642897,
+      "learning_rate": 2.2662116040955634e-05,
+      "loss": 0.5467,
+      "step": 166
+    },
+    {
+      "epoch": 0.28498293515358364,
+      "grad_norm": 0.6702489702179578,
+      "learning_rate": 2.2798634812286692e-05,
+      "loss": 0.5508,
+      "step": 167
+    },
+    {
+      "epoch": 0.28668941979522183,
+      "grad_norm": 0.521860586936388,
+      "learning_rate": 2.2935153583617747e-05,
+      "loss": 0.4819,
+      "step": 168
+    },
+    {
+      "epoch": 0.2883959044368601,
+      "grad_norm": 0.6757634905743073,
+      "learning_rate": 2.307167235494881e-05,
+      "loss": 0.5019,
+      "step": 169
+    },
+    {
+      "epoch": 0.2901023890784983,
+      "grad_norm": 0.643206610737337,
+      "learning_rate": 2.3208191126279864e-05,
+      "loss": 0.5483,
+      "step": 170
+    },
+    {
+      "epoch": 0.29180887372013653,
+      "grad_norm": 0.6980931391123527,
+      "learning_rate": 2.3344709897610926e-05,
+      "loss": 0.5236,
+      "step": 171
+    },
+    {
+      "epoch": 0.2935153583617747,
+      "grad_norm": 0.7843824537207991,
+      "learning_rate": 2.348122866894198e-05,
+      "loss": 0.5377,
+      "step": 172
+    },
+    {
+      "epoch": 0.295221843003413,
+      "grad_norm": 0.6537672086818195,
+      "learning_rate": 2.361774744027304e-05,
+      "loss": 0.5315,
+      "step": 173
+    },
+    {
+      "epoch": 0.29692832764505117,
+      "grad_norm": 1.0082111848546023,
+      "learning_rate": 2.3754266211604095e-05,
+      "loss": 0.5339,
+      "step": 174
+    },
+    {
+      "epoch": 0.2986348122866894,
+      "grad_norm": 0.9355730184429999,
+      "learning_rate": 2.3890784982935157e-05,
+      "loss": 0.545,
+      "step": 175
+    },
+    {
+      "epoch": 0.3003412969283277,
+      "grad_norm": 0.8312045452721931,
+      "learning_rate": 2.4027303754266215e-05,
+      "loss": 0.5138,
+      "step": 176
+    },
+    {
+      "epoch": 0.30204778156996587,
+      "grad_norm": 0.7766276707556691,
+      "learning_rate": 2.416382252559727e-05,
+      "loss": 0.5596,
+      "step": 177
+    },
+    {
+      "epoch": 0.3037542662116041,
+      "grad_norm": 0.7036513475468814,
+      "learning_rate": 2.4300341296928332e-05,
+      "loss": 0.5588,
+      "step": 178
+    },
+    {
+      "epoch": 0.3054607508532423,
+      "grad_norm": 0.8199422623918116,
+      "learning_rate": 2.4436860068259387e-05,
+      "loss": 0.5156,
+      "step": 179
+    },
+    {
+      "epoch": 0.30716723549488056,
+      "grad_norm": 0.718906109811401,
+      "learning_rate": 2.4573378839590446e-05,
+      "loss": 0.5513,
+      "step": 180
+    },
+    {
+      "epoch": 0.30887372013651876,
+      "grad_norm": 0.8086075980393629,
+      "learning_rate": 2.4709897610921504e-05,
+      "loss": 0.5273,
+      "step": 181
+    },
+    {
+      "epoch": 0.310580204778157,
+      "grad_norm": 0.6153099660722545,
+      "learning_rate": 2.4846416382252563e-05,
+      "loss": 0.5369,
+      "step": 182
+    },
+    {
+      "epoch": 0.3122866894197952,
+      "grad_norm": 0.8095994740813184,
+      "learning_rate": 2.4982935153583618e-05,
+      "loss": 0.5493,
+      "step": 183
+    },
+    {
+      "epoch": 0.31399317406143346,
+      "grad_norm": 0.5922586913811912,
+      "learning_rate": 2.511945392491468e-05,
+      "loss": 0.5209,
+      "step": 184
+    },
+    {
+      "epoch": 0.31569965870307165,
+      "grad_norm": 0.7562463377738635,
+      "learning_rate": 2.5255972696245735e-05,
+      "loss": 0.5187,
+      "step": 185
+    },
+    {
+      "epoch": 0.3174061433447099,
+      "grad_norm": 0.6423657920096462,
+      "learning_rate": 2.5392491467576793e-05,
+      "loss": 0.4866,
+      "step": 186
+    },
+    {
+      "epoch": 0.3191126279863481,
+      "grad_norm": 0.7239640355125908,
+      "learning_rate": 2.5529010238907848e-05,
+      "loss": 0.5409,
+      "step": 187
+    },
+    {
+      "epoch": 0.32081911262798635,
+      "grad_norm": 0.7326794140013588,
+      "learning_rate": 2.566552901023891e-05,
+      "loss": 0.5832,
+      "step": 188
+    },
+    {
+      "epoch": 0.3225255972696246,
+      "grad_norm": 0.6853080082888039,
+      "learning_rate": 2.580204778156997e-05,
+      "loss": 0.5526,
+      "step": 189
+    },
+    {
+      "epoch": 0.3242320819112628,
+      "grad_norm": 0.7850430606444814,
+      "learning_rate": 2.5938566552901027e-05,
+      "loss": 0.5489,
+      "step": 190
+    },
+    {
+      "epoch": 0.32593856655290104,
+      "grad_norm": 0.6849175762613122,
+      "learning_rate": 2.6075085324232085e-05,
+      "loss": 0.5107,
+      "step": 191
+    },
+    {
+      "epoch": 0.32764505119453924,
+      "grad_norm": 0.7255246675949937,
+      "learning_rate": 2.621160409556314e-05,
+      "loss": 0.5333,
+      "step": 192
+    },
+    {
+      "epoch": 0.3293515358361775,
+      "grad_norm": 0.6490845962032434,
+      "learning_rate": 2.6348122866894202e-05,
+      "loss": 0.4867,
+      "step": 193
+    },
+    {
+      "epoch": 0.3310580204778157,
+      "grad_norm": 0.6075482567970028,
+      "learning_rate": 2.6484641638225258e-05,
+      "loss": 0.5384,
+      "step": 194
+    },
+    {
+      "epoch": 0.33276450511945393,
+      "grad_norm": 0.6837704922056123,
+      "learning_rate": 2.6621160409556316e-05,
+      "loss": 0.5357,
+      "step": 195
+    },
+    {
+      "epoch": 0.33447098976109213,
+      "grad_norm": 0.569359613071419,
+      "learning_rate": 2.675767918088737e-05,
+      "loss": 0.5543,
+      "step": 196
+    },
+    {
+      "epoch": 0.3361774744027304,
+      "grad_norm": 0.6092108859734203,
+      "learning_rate": 2.6894197952218433e-05,
+      "loss": 0.5334,
+      "step": 197
+    },
+    {
+      "epoch": 0.3378839590443686,
+      "grad_norm": 0.524514989396669,
+      "learning_rate": 2.7030716723549488e-05,
+      "loss": 0.6771,
+      "step": 198
+    },
+    {
+      "epoch": 0.3395904436860068,
+      "grad_norm": 0.5628727302131958,
+      "learning_rate": 2.716723549488055e-05,
+      "loss": 0.5679,
+      "step": 199
+    },
+    {
+      "epoch": 0.3412969283276451,
+      "grad_norm": 0.5947009175582536,
+      "learning_rate": 2.730375426621161e-05,
+      "loss": 0.5075,
+      "step": 200
+    },
+    {
+      "epoch": 0.3430034129692833,
+      "grad_norm": 0.5624210026982975,
+      "learning_rate": 2.7440273037542664e-05,
+      "loss": 0.5353,
+      "step": 201
+    },
+    {
+      "epoch": 0.3447098976109215,
+      "grad_norm": 0.6721813831160073,
+      "learning_rate": 2.7576791808873725e-05,
+      "loss": 0.5301,
+      "step": 202
+    },
+    {
+      "epoch": 0.3464163822525597,
+      "grad_norm": 0.5492717552749656,
+      "learning_rate": 2.771331058020478e-05,
+      "loss": 0.5593,
+      "step": 203
+    },
+    {
+      "epoch": 0.34812286689419797,
+      "grad_norm": 0.5813072784903112,
+      "learning_rate": 2.784982935153584e-05,
+      "loss": 0.5434,
+      "step": 204
+    },
+    {
+      "epoch": 0.34982935153583616,
+      "grad_norm": 0.6527862562597682,
+      "learning_rate": 2.7986348122866894e-05,
+      "loss": 0.532,
+      "step": 205
+    },
+    {
+      "epoch": 0.3515358361774744,
+      "grad_norm": 0.682141809401472,
+      "learning_rate": 2.8122866894197956e-05,
+      "loss": 0.5652,
+      "step": 206
+    },
+    {
+      "epoch": 0.3532423208191126,
+      "grad_norm": 0.7797701437737451,
+      "learning_rate": 2.825938566552901e-05,
+      "loss": 0.5081,
+      "step": 207
+    },
+    {
+      "epoch": 0.35494880546075086,
+      "grad_norm": 0.7324241925774362,
+      "learning_rate": 2.839590443686007e-05,
+      "loss": 0.5356,
+      "step": 208
+    },
+    {
+      "epoch": 0.35665529010238906,
+      "grad_norm": 0.6576419396573902,
+      "learning_rate": 2.8532423208191128e-05,
+      "loss": 0.4907,
+      "step": 209
+    },
+    {
+      "epoch": 0.3583617747440273,
+      "grad_norm": 0.6867342503402145,
+      "learning_rate": 2.8668941979522186e-05,
+      "loss": 0.5129,
+      "step": 210
+    },
+    {
+      "epoch": 0.36006825938566556,
+      "grad_norm": 0.6549644929381108,
+      "learning_rate": 2.880546075085325e-05,
+      "loss": 0.5486,
+      "step": 211
+    },
+    {
+      "epoch": 0.36177474402730375,
+      "grad_norm": 0.662278431802467,
+      "learning_rate": 2.8941979522184303e-05,
+      "loss": 0.5403,
+      "step": 212
+    },
+    {
+      "epoch": 0.363481228668942,
+      "grad_norm": 0.526648439460945,
+      "learning_rate": 2.9078498293515362e-05,
+      "loss": 0.4915,
+      "step": 213
+    },
+    {
+      "epoch": 0.3651877133105802,
+      "grad_norm": 0.6386994912227945,
+      "learning_rate": 2.9215017064846417e-05,
+      "loss": 0.5583,
+      "step": 214
+    },
+    {
+      "epoch": 0.36689419795221845,
+      "grad_norm": 0.5749334384233534,
+      "learning_rate": 2.935153583617748e-05,
+      "loss": 0.5254,
+      "step": 215
+    },
+    {
+      "epoch": 0.36860068259385664,
+      "grad_norm": 0.7850556328751596,
+      "learning_rate": 2.9488054607508534e-05,
+      "loss": 0.574,
+      "step": 216
+    },
+    {
+      "epoch": 0.3703071672354949,
+      "grad_norm": 0.5715196045398925,
+      "learning_rate": 2.9624573378839592e-05,
+      "loss": 0.5269,
+      "step": 217
+    },
+    {
+      "epoch": 0.3720136518771331,
+      "grad_norm": 0.7461124241882047,
+      "learning_rate": 2.976109215017065e-05,
+      "loss": 0.5514,
+      "step": 218
+    },
+    {
+      "epoch": 0.37372013651877134,
+      "grad_norm": 0.6191886408557776,
+      "learning_rate": 2.989761092150171e-05,
+      "loss": 0.5224,
+      "step": 219
+    },
+    {
+      "epoch": 0.37542662116040953,
+      "grad_norm": 0.6005041736662892,
+      "learning_rate": 3.0034129692832765e-05,
+      "loss": 0.4765,
+      "step": 220
+    },
+    {
+      "epoch": 0.3771331058020478,
+      "grad_norm": 0.724603231134017,
+      "learning_rate": 3.0170648464163826e-05,
+      "loss": 0.5405,
+      "step": 221
+    },
+    {
+      "epoch": 0.378839590443686,
+      "grad_norm": 0.5826856216983127,
+      "learning_rate": 3.030716723549488e-05,
+      "loss": 0.5345,
+      "step": 222
+    },
+    {
+      "epoch": 0.38054607508532423,
+      "grad_norm": 0.7198123114976283,
+      "learning_rate": 3.044368600682594e-05,
+      "loss": 0.5247,
+      "step": 223
+    },
+    {
+      "epoch": 0.3822525597269625,
+      "grad_norm": 0.5762779778908053,
+      "learning_rate": 3.0580204778157e-05,
+      "loss": 0.5119,
+      "step": 224
+    },
+    {
+      "epoch": 0.3839590443686007,
+      "grad_norm": 0.7830870037164162,
+      "learning_rate": 3.0716723549488054e-05,
+      "loss": 0.5465,
+      "step": 225
+    },
+    {
+      "epoch": 0.3856655290102389,
+      "grad_norm": 0.5420792016735733,
+      "learning_rate": 3.0853242320819115e-05,
+      "loss": 0.4986,
+      "step": 226
+    },
+    {
+      "epoch": 0.3873720136518771,
+      "grad_norm": 0.7105286095689686,
+      "learning_rate": 3.098976109215017e-05,
+      "loss": 0.4984,
+      "step": 227
+    },
+    {
+      "epoch": 0.3890784982935154,
+      "grad_norm": 0.6107287514130099,
+      "learning_rate": 3.112627986348123e-05,
+      "loss": 0.5422,
+      "step": 228
+    },
+    {
+      "epoch": 0.39078498293515357,
+      "grad_norm": 0.6725313318953229,
+      "learning_rate": 3.126279863481229e-05,
+      "loss": 0.5472,
+      "step": 229
+    },
+    {
+      "epoch": 0.3924914675767918,
+      "grad_norm": 0.6667927283774585,
+      "learning_rate": 3.139931740614335e-05,
+      "loss": 0.5205,
+      "step": 230
+    },
+    {
+      "epoch": 0.39419795221843,
+      "grad_norm": 0.5771264923474617,
+      "learning_rate": 3.1535836177474404e-05,
+      "loss": 0.5335,
+      "step": 231
+    },
+    {
+      "epoch": 0.39590443686006827,
+      "grad_norm": 0.5743686892022483,
+      "learning_rate": 3.1672354948805466e-05,
+      "loss": 0.567,
+      "step": 232
+    },
+    {
+      "epoch": 0.39761092150170646,
+      "grad_norm": 0.642334479475538,
+      "learning_rate": 3.180887372013652e-05,
+      "loss": 0.5661,
+      "step": 233
+    },
+    {
+      "epoch": 0.3993174061433447,
+      "grad_norm": 0.5708609452890699,
+      "learning_rate": 3.1945392491467577e-05,
+      "loss": 0.5227,
+      "step": 234
+    },
+    {
+      "epoch": 0.40102389078498296,
+      "grad_norm": 0.6687187653936159,
+      "learning_rate": 3.208191126279864e-05,
+      "loss": 0.5339,
+      "step": 235
+    },
+    {
+      "epoch": 0.40273037542662116,
+      "grad_norm": 0.6645444871635297,
+      "learning_rate": 3.2218430034129693e-05,
+      "loss": 0.5457,
+      "step": 236
+    },
+    {
+      "epoch": 0.4044368600682594,
+      "grad_norm": 0.5313583267845954,
+      "learning_rate": 3.2354948805460755e-05,
+      "loss": 0.5193,
+      "step": 237
+    },
+    {
+      "epoch": 0.4061433447098976,
+      "grad_norm": 0.6166169566756322,
+      "learning_rate": 3.249146757679181e-05,
+      "loss": 0.509,
+      "step": 238
+    },
+    {
+      "epoch": 0.40784982935153585,
+      "grad_norm": 0.4855449343720997,
+      "learning_rate": 3.262798634812287e-05,
+      "loss": 0.5065,
+      "step": 239
+    },
+    {
+      "epoch": 0.40955631399317405,
+      "grad_norm": 0.6096286686967779,
+      "learning_rate": 3.276450511945393e-05,
+      "loss": 0.5372,
+      "step": 240
+    },
+    {
+      "epoch": 0.4112627986348123,
+      "grad_norm": 0.6881439297443295,
+      "learning_rate": 3.290102389078499e-05,
+      "loss": 0.5462,
+      "step": 241
+    },
+    {
+      "epoch": 0.4129692832764505,
+      "grad_norm": 0.6556036894082107,
+      "learning_rate": 3.3037542662116044e-05,
+      "loss": 0.546,
+      "step": 242
+    },
+    {
+      "epoch": 0.41467576791808874,
+      "grad_norm": 0.5420472627486871,
+      "learning_rate": 3.31740614334471e-05,
+      "loss": 0.5107,
+      "step": 243
+    },
+    {
+      "epoch": 0.41638225255972694,
+      "grad_norm": 0.55567844718282,
+      "learning_rate": 3.3310580204778155e-05,
+      "loss": 0.5174,
+      "step": 244
+    },
+    {
+      "epoch": 0.4180887372013652,
+      "grad_norm": 0.6166867575282067,
+      "learning_rate": 3.3447098976109216e-05,
+      "loss": 0.5108,
+      "step": 245
+    },
+    {
+      "epoch": 0.4197952218430034,
+      "grad_norm": 0.5185587182303334,
+      "learning_rate": 3.358361774744027e-05,
+      "loss": 0.5152,
+      "step": 246
+    },
+    {
+      "epoch": 0.42150170648464164,
+      "grad_norm": 0.6043195051507727,
+      "learning_rate": 3.3720136518771333e-05,
+      "loss": 0.5329,
+      "step": 247
+    },
+    {
+      "epoch": 0.4232081911262799,
+      "grad_norm": 0.6320652628525706,
+      "learning_rate": 3.3856655290102395e-05,
+      "loss": 0.5258,
+      "step": 248
+    },
+    {
+      "epoch": 0.4249146757679181,
+      "grad_norm": 0.5445988520895617,
+      "learning_rate": 3.399317406143345e-05,
+      "loss": 0.5106,
+      "step": 249
+    },
+    {
+      "epoch": 0.42662116040955633,
+      "grad_norm": 0.6001675298572623,
+      "learning_rate": 3.412969283276451e-05,
+      "loss": 0.5319,
+      "step": 250
+    },
+    {
+      "epoch": 0.4283276450511945,
+      "grad_norm": 0.5259131674323454,
+      "learning_rate": 3.426621160409557e-05,
+      "loss": 0.4957,
+      "step": 251
+    },
+    {
+      "epoch": 0.4300341296928328,
+      "grad_norm": 0.723456289561426,
+      "learning_rate": 3.440273037542662e-05,
+      "loss": 0.5187,
+      "step": 252
+    },
+    {
+      "epoch": 0.431740614334471,
+      "grad_norm": 0.6062601161437798,
+      "learning_rate": 3.453924914675768e-05,
+      "loss": 0.5288,
+      "step": 253
+    },
+    {
+      "epoch": 0.4334470989761092,
+      "grad_norm": 0.6357304201610898,
+      "learning_rate": 3.467576791808874e-05,
+      "loss": 0.4921,
+      "step": 254
+    },
+    {
+      "epoch": 0.4351535836177474,
+      "grad_norm": 0.6726290740589141,
+      "learning_rate": 3.4812286689419794e-05,
+      "loss": 0.5185,
+      "step": 255
+    },
+    {
+      "epoch": 0.43686006825938567,
+      "grad_norm": 0.7087160989243053,
+      "learning_rate": 3.4948805460750856e-05,
+      "loss": 0.533,
+      "step": 256
+    },
+    {
+      "epoch": 0.43856655290102387,
+      "grad_norm": 0.666647645226124,
+      "learning_rate": 3.508532423208191e-05,
+      "loss": 0.5353,
+      "step": 257
+    },
+    {
+      "epoch": 0.4402730375426621,
+      "grad_norm": 0.6270896669729984,
+      "learning_rate": 3.522184300341297e-05,
+      "loss": 0.4978,
+      "step": 258
+    },
+    {
+      "epoch": 0.44197952218430037,
+      "grad_norm": 0.6759682019764348,
+      "learning_rate": 3.5358361774744035e-05,
+      "loss": 0.5251,
+      "step": 259
+    },
+    {
+      "epoch": 0.44368600682593856,
+      "grad_norm": 0.6676073872340431,
+      "learning_rate": 3.549488054607509e-05,
+      "loss": 0.4391,
+      "step": 260
+    },
+    {
+      "epoch": 0.4453924914675768,
+      "grad_norm": 0.6956499802081617,
+      "learning_rate": 3.5631399317406145e-05,
+      "loss": 0.5426,
+      "step": 261
+    },
+    {
+      "epoch": 0.447098976109215,
+      "grad_norm": 0.8039012535909295,
+      "learning_rate": 3.57679180887372e-05,
+      "loss": 0.5619,
+      "step": 262
+    },
+    {
+      "epoch": 0.44880546075085326,
+      "grad_norm": 0.6701469728153912,
+      "learning_rate": 3.590443686006826e-05,
+      "loss": 0.5252,
+      "step": 263
+    },
+    {
+      "epoch": 0.45051194539249145,
+      "grad_norm": 0.6490618284705951,
+      "learning_rate": 3.604095563139932e-05,
+      "loss": 0.5219,
+      "step": 264
+    },
+    {
+      "epoch": 0.4522184300341297,
+      "grad_norm": 0.721560807463733,
+      "learning_rate": 3.617747440273038e-05,
+      "loss": 0.5359,
+      "step": 265
+    },
+    {
+      "epoch": 0.4539249146757679,
+      "grad_norm": 0.5561673893746749,
+      "learning_rate": 3.6313993174061434e-05,
+      "loss": 0.6168,
+      "step": 266
+    },
+    {
+      "epoch": 0.45563139931740615,
+      "grad_norm": 0.7584229071720511,
+      "learning_rate": 3.6450511945392496e-05,
+      "loss": 0.5315,
+      "step": 267
+    },
+    {
+      "epoch": 0.45733788395904434,
+      "grad_norm": 0.5649763445593471,
+      "learning_rate": 3.658703071672355e-05,
+      "loss": 0.5245,
+      "step": 268
+    },
+    {
+      "epoch": 0.4590443686006826,
+      "grad_norm": 0.6056674738757325,
+      "learning_rate": 3.672354948805461e-05,
+      "loss": 0.4945,
+      "step": 269
+    },
+    {
+      "epoch": 0.46075085324232085,
+      "grad_norm": 0.5421799420542381,
+      "learning_rate": 3.686006825938567e-05,
+      "loss": 0.518,
+      "step": 270
+    },
+    {
+      "epoch": 0.46245733788395904,
+      "grad_norm": 0.5751187796420072,
+      "learning_rate": 3.6996587030716723e-05,
+      "loss": 0.5393,
+      "step": 271
+    },
+    {
+      "epoch": 0.4641638225255973,
+      "grad_norm": 0.5859129942902703,
+      "learning_rate": 3.7133105802047785e-05,
+      "loss": 0.5293,
+      "step": 272
+    },
+    {
+      "epoch": 0.4658703071672355,
+      "grad_norm": 0.5938721485777587,
+      "learning_rate": 3.726962457337884e-05,
+      "loss": 0.5388,
+      "step": 273
+    },
+    {
+      "epoch": 0.46757679180887374,
+      "grad_norm": 0.5218225688533202,
+      "learning_rate": 3.74061433447099e-05,
+      "loss": 0.5415,
+      "step": 274
+    },
+    {
+      "epoch": 0.46928327645051193,
+      "grad_norm": 0.6449052599396903,
+      "learning_rate": 3.754266211604096e-05,
+      "loss": 0.5264,
+      "step": 275
+    },
+    {
+      "epoch": 0.4709897610921502,
+      "grad_norm": 0.732037053998216,
+      "learning_rate": 3.767918088737202e-05,
+      "loss": 0.571,
+      "step": 276
+    },
+    {
+      "epoch": 0.4726962457337884,
+      "grad_norm": 0.5440753439319131,
+      "learning_rate": 3.7815699658703074e-05,
+      "loss": 0.5385,
+      "step": 277
+    },
+    {
+      "epoch": 0.47440273037542663,
+      "grad_norm": 0.6329161957095896,
+      "learning_rate": 3.7952218430034136e-05,
+      "loss": 0.5456,
+      "step": 278
+    },
+    {
+      "epoch": 0.4761092150170648,
+      "grad_norm": 0.6528500511006126,
+      "learning_rate": 3.808873720136519e-05,
+      "loss": 0.5448,
+      "step": 279
+    },
+    {
+      "epoch": 0.4778156996587031,
+      "grad_norm": 0.5545510586675965,
+      "learning_rate": 3.8225255972696246e-05,
+      "loss": 0.4856,
+      "step": 280
+    },
+    {
+      "epoch": 0.47952218430034127,
+      "grad_norm": 0.5786537016897892,
+      "learning_rate": 3.83617747440273e-05,
+      "loss": 0.536,
+      "step": 281
+    },
+    {
+      "epoch": 0.4812286689419795,
+      "grad_norm": 0.641252312841796,
+      "learning_rate": 3.849829351535836e-05,
+      "loss": 0.5591,
+      "step": 282
+    },
+    {
+      "epoch": 0.48293515358361777,
+      "grad_norm": 0.5838397942330007,
+      "learning_rate": 3.8634812286689425e-05,
+      "loss": 0.5596,
+      "step": 283
+    },
+    {
+      "epoch": 0.48464163822525597,
+      "grad_norm": 0.6148987590733005,
+      "learning_rate": 3.877133105802048e-05,
+      "loss": 0.4954,
+      "step": 284
+    },
+    {
+      "epoch": 0.4863481228668942,
+      "grad_norm": 0.6518937049519072,
+      "learning_rate": 3.890784982935154e-05,
+      "loss": 0.5341,
+      "step": 285
+    },
+    {
+      "epoch": 0.4880546075085324,
+      "grad_norm": 0.5410865955347726,
+      "learning_rate": 3.90443686006826e-05,
+      "loss": 0.5382,
+      "step": 286
+    },
+    {
+      "epoch": 0.48976109215017066,
+      "grad_norm": 0.7101085537879301,
+      "learning_rate": 3.918088737201366e-05,
+      "loss": 0.592,
+      "step": 287
+    },
+    {
+      "epoch": 0.49146757679180886,
+      "grad_norm": 0.6240221422741028,
+      "learning_rate": 3.9317406143344714e-05,
+      "loss": 0.5299,
+      "step": 288
+    },
+    {
+      "epoch": 0.4931740614334471,
+      "grad_norm": 0.7583190153817212,
+      "learning_rate": 3.945392491467577e-05,
+      "loss": 0.5454,
+      "step": 289
+    },
+    {
+      "epoch": 0.4948805460750853,
+      "grad_norm": 0.6085094169710514,
+      "learning_rate": 3.9590443686006824e-05,
+      "loss": 0.5159,
+      "step": 290
+    },
+    {
+      "epoch": 0.49658703071672355,
+      "grad_norm": 0.549608958818555,
+      "learning_rate": 3.9726962457337886e-05,
+      "loss": 0.5009,
+      "step": 291
+    },
+    {
+      "epoch": 0.49829351535836175,
+      "grad_norm": 0.6214117294853628,
+      "learning_rate": 3.986348122866894e-05,
+      "loss": 0.5484,
+      "step": 292
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5328158045651903,
+      "learning_rate": 4e-05,
+      "loss": 0.5332,
+      "step": 293
+    },
+    {
+      "epoch": 0.5017064846416383,
+      "grad_norm": 0.5554681326449042,
+      "learning_rate": 3.9999985806829025e-05,
+      "loss": 0.5124,
+      "step": 294
+    },
+    {
+      "epoch": 0.5034129692832765,
+      "grad_norm": 0.5918511463290707,
+      "learning_rate": 3.999994322733625e-05,
+      "loss": 0.5466,
+      "step": 295
+    },
+    {
+      "epoch": 0.5051194539249146,
+      "grad_norm": 0.6742018044203787,
+      "learning_rate": 3.99998722615821e-05,
+      "loss": 0.5973,
+      "step": 296
+    },
+    {
+      "epoch": 0.5068259385665529,
+      "grad_norm": 0.6783398500413388,
+      "learning_rate": 3.999977290966729e-05,
+      "loss": 0.5627,
+      "step": 297
+    },
+    {
+      "epoch": 0.5085324232081911,
+      "grad_norm": 0.6443965373732842,
+      "learning_rate": 3.999964517173286e-05,
+      "loss": 0.502,
+      "step": 298
+    },
+    {
+      "epoch": 0.5102389078498294,
+      "grad_norm": 0.75173297356595,
+      "learning_rate": 3.999948904796009e-05,
+      "loss": 0.5753,
+      "step": 299
+    },
+    {
+      "epoch": 0.5119453924914675,
+      "grad_norm": 0.5578886470083074,
+      "learning_rate": 3.9999304538570564e-05,
+      "loss": 0.5828,
+      "step": 300
+    },
+    {
+      "epoch": 0.5136518771331058,
+      "grad_norm": 0.646565069167516,
+      "learning_rate": 3.9999091643826175e-05,
+      "loss": 0.5104,
+      "step": 301
+    },
+    {
+      "epoch": 0.515358361774744,
+      "grad_norm": 0.601549331413928,
+      "learning_rate": 3.999885036402908e-05,
+      "loss": 0.5578,
+      "step": 302
+    },
+    {
+      "epoch": 0.5170648464163823,
+      "grad_norm": 0.6385380164309451,
+      "learning_rate": 3.999858069952173e-05,
+      "loss": 0.542,
+      "step": 303
+    },
+    {
+      "epoch": 0.5187713310580204,
+      "grad_norm": 0.6533143947463091,
+      "learning_rate": 3.999828265068687e-05,
+      "loss": 0.4904,
+      "step": 304
+    },
+    {
+      "epoch": 0.5204778156996587,
+      "grad_norm": 0.6452412797735231,
+      "learning_rate": 3.9997956217947525e-05,
+      "loss": 0.5721,
+      "step": 305
+    },
+    {
+      "epoch": 0.5221843003412969,
+      "grad_norm": 0.7700498239283877,
+      "learning_rate": 3.999760140176701e-05,
+      "loss": 0.5299,
+      "step": 306
+    },
+    {
+      "epoch": 0.5238907849829352,
+      "grad_norm": 0.5347847319347675,
+      "learning_rate": 3.999721820264891e-05,
+      "loss": 0.4806,
+      "step": 307
+    },
+    {
+      "epoch": 0.5255972696245734,
+      "grad_norm": 0.7201943625961811,
+      "learning_rate": 3.999680662113711e-05,
+      "loss": 0.5431,
+      "step": 308
+    },
+    {
+      "epoch": 0.5273037542662116,
+      "grad_norm": 0.6587459195924609,
+      "learning_rate": 3.9996366657815784e-05,
+      "loss": 0.4872,
+      "step": 309
+    },
+    {
+      "epoch": 0.5290102389078498,
+      "grad_norm": 0.568648667150652,
+      "learning_rate": 3.999589831330937e-05,
+      "loss": 0.5303,
+      "step": 310
+    },
+    {
+      "epoch": 0.5307167235494881,
+      "grad_norm": 0.755987795579431,
+      "learning_rate": 3.99954015882826e-05,
+      "loss": 0.5326,
+      "step": 311
+    },
+    {
+      "epoch": 0.5324232081911263,
+      "grad_norm": 0.7495800369506979,
+      "learning_rate": 3.9994876483440483e-05,
+      "loss": 0.5329,
+      "step": 312
+    },
+    {
+      "epoch": 0.5341296928327645,
+      "grad_norm": 0.5368426553441042,
+      "learning_rate": 3.999432299952831e-05,
+      "loss": 0.5349,
+      "step": 313
+    },
+    {
+      "epoch": 0.5358361774744027,
+      "grad_norm": 0.5423808226328886,
+      "learning_rate": 3.999374113733165e-05,
+      "loss": 0.4707,
+      "step": 314
+    },
+    {
+      "epoch": 0.537542662116041,
+      "grad_norm": 0.4979739764128441,
+      "learning_rate": 3.999313089767635e-05,
+      "loss": 0.5378,
+      "step": 315
+    },
+    {
+      "epoch": 0.5392491467576792,
+      "grad_norm": 0.5607008750404271,
+      "learning_rate": 3.999249228142854e-05,
+      "loss": 0.5719,
+      "step": 316
+    },
+    {
+      "epoch": 0.5409556313993175,
+      "grad_norm": 0.5635028320952156,
+      "learning_rate": 3.999182528949462e-05,
+      "loss": 0.5007,
+      "step": 317
+    },
+    {
+      "epoch": 0.5426621160409556,
+      "grad_norm": 0.4774574283561093,
+      "learning_rate": 3.9991129922821244e-05,
+      "loss": 0.5356,
+      "step": 318
+    },
+    {
+      "epoch": 0.5443686006825939,
+      "grad_norm": 0.6841699434711654,
+      "learning_rate": 3.999040618239537e-05,
+      "loss": 0.5723,
+      "step": 319
+    },
+    {
+      "epoch": 0.5460750853242321,
+      "grad_norm": 0.536155548533752,
+      "learning_rate": 3.998965406924422e-05,
+      "loss": 0.5195,
+      "step": 320
+    },
+    {
+      "epoch": 0.5477815699658704,
+      "grad_norm": 0.5995649286161172,
+      "learning_rate": 3.998887358443528e-05,
+      "loss": 0.5412,
+      "step": 321
+    },
+    {
+      "epoch": 0.5494880546075085,
+      "grad_norm": 0.5112550518958299,
+      "learning_rate": 3.99880647290763e-05,
+      "loss": 0.5296,
+      "step": 322
+    },
+    {
+      "epoch": 0.5511945392491467,
+      "grad_norm": 0.5700086851182184,
+      "learning_rate": 3.9987227504315295e-05,
+      "loss": 0.593,
+      "step": 323
+    },
+    {
+      "epoch": 0.552901023890785,
+      "grad_norm": 0.5291855952380958,
+      "learning_rate": 3.998636191134057e-05,
+      "loss": 0.542,
+      "step": 324
+    },
+    {
+      "epoch": 0.5546075085324232,
+      "grad_norm": 0.5119590767300913,
+      "learning_rate": 3.9985467951380666e-05,
+      "loss": 0.584,
+      "step": 325
+    },
+    {
+      "epoch": 0.5563139931740614,
+      "grad_norm": 0.5594566122074516,
+      "learning_rate": 3.9984545625704396e-05,
+      "loss": 0.5336,
+      "step": 326
+    },
+    {
+      "epoch": 0.5580204778156996,
+      "grad_norm": 0.4637946091135965,
+      "learning_rate": 3.9983594935620835e-05,
+      "loss": 0.5618,
+      "step": 327
+    },
+    {
+      "epoch": 0.5597269624573379,
+      "grad_norm": 0.6359265452032311,
+      "learning_rate": 3.998261588247931e-05,
+      "loss": 0.5676,
+      "step": 328
+    },
+    {
+      "epoch": 0.5614334470989761,
+      "grad_norm": 0.47797621922130384,
+      "learning_rate": 3.998160846766941e-05,
+      "loss": 0.5135,
+      "step": 329
+    },
+    {
+      "epoch": 0.5631399317406144,
+      "grad_norm": 0.5906982722443385,
+      "learning_rate": 3.998057269262099e-05,
+      "loss": 0.5358,
+      "step": 330
+    },
+    {
+      "epoch": 0.5648464163822525,
+      "grad_norm": 0.5203977398556296,
+      "learning_rate": 3.997950855880411e-05,
+      "loss": 0.5072,
+      "step": 331
+    },
+    {
+      "epoch": 0.5665529010238908,
+      "grad_norm": 0.6067712351750322,
+      "learning_rate": 3.997841606772914e-05,
+      "loss": 0.5626,
+      "step": 332
+    },
+    {
+      "epoch": 0.568259385665529,
+      "grad_norm": 0.6157080556517592,
+      "learning_rate": 3.997729522094667e-05,
+      "loss": 0.5352,
+      "step": 333
+    },
+    {
+      "epoch": 0.5699658703071673,
+      "grad_norm": 0.590679814892138,
+      "learning_rate": 3.997614602004752e-05,
+      "loss": 0.4935,
+      "step": 334
+    },
+    {
+      "epoch": 0.5716723549488054,
+      "grad_norm": 0.6323221642452715,
+      "learning_rate": 3.997496846666279e-05,
+      "loss": 0.5341,
+      "step": 335
+    },
+    {
+      "epoch": 0.5733788395904437,
+      "grad_norm": 0.6151594886516842,
+      "learning_rate": 3.997376256246379e-05,
+      "loss": 0.5615,
+      "step": 336
+    },
+    {
+      "epoch": 0.5750853242320819,
+      "grad_norm": 0.5846274003132382,
+      "learning_rate": 3.9972528309162086e-05,
+      "loss": 0.5096,
+      "step": 337
+    },
+    {
+      "epoch": 0.5767918088737202,
+      "grad_norm": 0.6245632779110738,
+      "learning_rate": 3.997126570850947e-05,
+      "loss": 0.5144,
+      "step": 338
+    },
+    {
+      "epoch": 0.5784982935153583,
+      "grad_norm": 0.6498707890711793,
+      "learning_rate": 3.9969974762297974e-05,
+      "loss": 0.5371,
+      "step": 339
+    },
+    {
+      "epoch": 0.5802047781569966,
+      "grad_norm": 0.6551387671037868,
+      "learning_rate": 3.996865547235987e-05,
+      "loss": 0.5,
+      "step": 340
+    },
+    {
+      "epoch": 0.5819112627986348,
+      "grad_norm": 0.5854419220167429,
+      "learning_rate": 3.996730784056763e-05,
+      "loss": 0.5355,
+      "step": 341
+    },
+    {
+      "epoch": 0.5836177474402731,
+      "grad_norm": 0.6766082992863565,
+      "learning_rate": 3.9965931868833984e-05,
+      "loss": 0.5334,
+      "step": 342
+    },
+    {
+      "epoch": 0.5853242320819113,
+      "grad_norm": 0.5653804448656627,
+      "learning_rate": 3.996452755911187e-05,
+      "loss": 0.492,
+      "step": 343
+    },
+    {
+      "epoch": 0.5870307167235495,
+      "grad_norm": 0.6505973904692599,
+      "learning_rate": 3.996309491339445e-05,
+      "loss": 0.5765,
+      "step": 344
+    },
+    {
+      "epoch": 0.5887372013651877,
+      "grad_norm": 0.5830522080149915,
+      "learning_rate": 3.99616339337151e-05,
+      "loss": 0.5502,
+      "step": 345
+    },
+    {
+      "epoch": 0.590443686006826,
+      "grad_norm": 0.6184398266691763,
+      "learning_rate": 3.996014462214741e-05,
+      "loss": 0.5188,
+      "step": 346
+    },
+    {
+      "epoch": 0.5921501706484642,
+      "grad_norm": 0.5602745636382449,
+      "learning_rate": 3.99586269808052e-05,
+      "loss": 0.579,
+      "step": 347
+    },
+    {
+      "epoch": 0.5938566552901023,
+      "grad_norm": 0.5417892827926399,
+      "learning_rate": 3.995708101184246e-05,
+      "loss": 0.5066,
+      "step": 348
+    },
+    {
+      "epoch": 0.5955631399317406,
+      "grad_norm": 0.5635084844223573,
+      "learning_rate": 3.995550671745343e-05,
+      "loss": 0.5273,
+      "step": 349
+    },
+    {
+      "epoch": 0.5972696245733788,
+      "grad_norm": 0.49850321668504116,
+      "learning_rate": 3.9953904099872525e-05,
+      "loss": 0.5225,
+      "step": 350
+    },
+    {
+      "epoch": 0.5989761092150171,
+      "grad_norm": 0.5897835179612174,
+      "learning_rate": 3.9952273161374366e-05,
+      "loss": 0.5169,
+      "step": 351
+    },
+    {
+      "epoch": 0.6006825938566553,
+      "grad_norm": 0.5115903280203131,
+      "learning_rate": 3.9950613904273786e-05,
+      "loss": 0.5376,
+      "step": 352
+    },
+    {
+      "epoch": 0.6023890784982935,
+      "grad_norm": 0.5655155488320919,
+      "learning_rate": 3.9948926330925775e-05,
+      "loss": 0.5826,
+      "step": 353
+    },
+    {
+      "epoch": 0.6040955631399317,
+      "grad_norm": 0.5548295358049624,
+      "learning_rate": 3.994721044372555e-05,
+      "loss": 0.5233,
+      "step": 354
+    },
+    {
+      "epoch": 0.60580204778157,
+      "grad_norm": 0.49586722317702214,
+      "learning_rate": 3.994546624510849e-05,
+      "loss": 0.4997,
+      "step": 355
+    },
+    {
+      "epoch": 0.6075085324232082,
+      "grad_norm": 0.7159231199083884,
+      "learning_rate": 3.994369373755018e-05,
+      "loss": 0.5276,
+      "step": 356
+    },
+    {
+      "epoch": 0.6092150170648464,
+      "grad_norm": 0.5695196393270819,
+      "learning_rate": 3.9941892923566354e-05,
+      "loss": 0.5091,
+      "step": 357
+    },
+    {
+      "epoch": 0.6109215017064846,
+      "grad_norm": 0.63179271946944,
+      "learning_rate": 3.994006380571295e-05,
+      "loss": 0.4945,
+      "step": 358
+    },
+    {
+      "epoch": 0.6126279863481229,
+      "grad_norm": 0.4900888065252049,
+      "learning_rate": 3.993820638658606e-05,
+      "loss": 0.4921,
+      "step": 359
+    },
+    {
+      "epoch": 0.6143344709897611,
+      "grad_norm": 0.5130463209318009,
+      "learning_rate": 3.993632066882195e-05,
+      "loss": 0.5001,
+      "step": 360
+    },
+    {
+      "epoch": 0.6160409556313993,
+      "grad_norm": 0.5950557347893138,
+      "learning_rate": 3.9934406655097055e-05,
+      "loss": 0.5462,
+      "step": 361
+    },
+    {
+      "epoch": 0.6177474402730375,
+      "grad_norm": 0.48765122887867945,
+      "learning_rate": 3.9932464348127965e-05,
+      "loss": 0.5383,
+      "step": 362
+    },
+    {
+      "epoch": 0.6194539249146758,
+      "grad_norm": 0.6614230862754317,
+      "learning_rate": 3.993049375067143e-05,
+      "loss": 0.5632,
+      "step": 363
+    },
+    {
+      "epoch": 0.621160409556314,
+      "grad_norm": 0.5271955351184203,
+      "learning_rate": 3.992849486552435e-05,
+      "loss": 0.5603,
+      "step": 364
+    },
+    {
+      "epoch": 0.6228668941979523,
+      "grad_norm": 0.6424373509828155,
+      "learning_rate": 3.992646769552379e-05,
+      "loss": 0.5454,
+      "step": 365
+    },
+    {
+      "epoch": 0.6245733788395904,
+      "grad_norm": 0.5912704776713205,
+      "learning_rate": 3.992441224354693e-05,
+      "loss": 0.5441,
+      "step": 366
+    },
+    {
+      "epoch": 0.6262798634812287,
+      "grad_norm": 0.5659728833578781,
+      "learning_rate": 3.9922328512511114e-05,
+      "loss": 0.533,
+      "step": 367
+    },
+    {
+      "epoch": 0.6279863481228669,
+      "grad_norm": 0.5759290370965763,
+      "learning_rate": 3.992021650537382e-05,
+      "loss": 0.5595,
+      "step": 368
+    },
+    {
+      "epoch": 0.6296928327645052,
+      "grad_norm": 0.47616510524297306,
+      "learning_rate": 3.991807622513266e-05,
+      "loss": 0.5279,
+      "step": 369
+    },
+    {
+      "epoch": 0.6313993174061433,
+      "grad_norm": 0.5825601686484897,
+      "learning_rate": 3.9915907674825356e-05,
+      "loss": 0.5413,
+      "step": 370
+    },
+    {
+      "epoch": 0.6331058020477816,
+      "grad_norm": 0.4483345812815246,
+      "learning_rate": 3.9913710857529784e-05,
+      "loss": 0.4962,
+      "step": 371
+    },
+    {
+      "epoch": 0.6348122866894198,
+      "grad_norm": 0.6775345195695525,
+      "learning_rate": 3.991148577636391e-05,
+      "loss": 0.5157,
+      "step": 372
+    },
+    {
+      "epoch": 0.636518771331058,
+      "grad_norm": 0.4624283233342401,
+      "learning_rate": 3.9909232434485836e-05,
+      "loss": 0.5205,
+      "step": 373
+    },
+    {
+      "epoch": 0.6382252559726962,
+      "grad_norm": 0.6699100743983422,
+      "learning_rate": 3.990695083509378e-05,
+      "loss": 0.5256,
+      "step": 374
+    },
+    {
+      "epoch": 0.6399317406143344,
+      "grad_norm": 0.47189183222271325,
+      "learning_rate": 3.990464098142604e-05,
+      "loss": 0.5237,
+      "step": 375
+    },
+    {
+      "epoch": 0.6416382252559727,
+      "grad_norm": 0.7343933179790109,
+      "learning_rate": 3.990230287676103e-05,
+      "loss": 0.5695,
+      "step": 376
+    },
+    {
+      "epoch": 0.643344709897611,
+      "grad_norm": 0.47194980144112525,
+      "learning_rate": 3.9899936524417274e-05,
+      "loss": 0.4663,
+      "step": 377
+    },
+    {
+      "epoch": 0.6450511945392492,
+      "grad_norm": 0.7087714839868563,
+      "learning_rate": 3.9897541927753365e-05,
+      "loss": 0.5266,
+      "step": 378
+    },
+    {
+      "epoch": 0.6467576791808873,
+      "grad_norm": 0.5193346775206192,
+      "learning_rate": 3.9895119090168e-05,
+      "loss": 0.528,
+      "step": 379
+    },
+    {
+      "epoch": 0.6484641638225256,
+      "grad_norm": 0.6482737816606934,
+      "learning_rate": 3.989266801509996e-05,
+      "loss": 0.5674,
+      "step": 380
+    },
+    {
+      "epoch": 0.6501706484641638,
+      "grad_norm": 0.5638383348296886,
+      "learning_rate": 3.9890188706028084e-05,
+      "loss": 0.4937,
+      "step": 381
+    },
+    {
+      "epoch": 0.6518771331058021,
+      "grad_norm": 0.4752960584294783,
+      "learning_rate": 3.988768116647131e-05,
+      "loss": 0.5186,
+      "step": 382
+    },
+    {
+      "epoch": 0.6535836177474402,
+      "grad_norm": 0.5731454087459066,
+      "learning_rate": 3.988514539998862e-05,
+      "loss": 0.5342,
+      "step": 383
+    },
+    {
+      "epoch": 0.6552901023890785,
+      "grad_norm": 0.616267780212079,
+      "learning_rate": 3.988258141017909e-05,
+      "loss": 0.5393,
+      "step": 384
+    },
+    {
+      "epoch": 0.6569965870307167,
+      "grad_norm": 0.5773096287704862,
+      "learning_rate": 3.987998920068181e-05,
+      "loss": 0.536,
+      "step": 385
+    },
+    {
+      "epoch": 0.658703071672355,
+      "grad_norm": 0.6696008091283474,
+      "learning_rate": 3.987736877517597e-05,
+      "loss": 0.5336,
+      "step": 386
+    },
+    {
+      "epoch": 0.6604095563139932,
+      "grad_norm": 0.5434061720204441,
+      "learning_rate": 3.987472013738076e-05,
+      "loss": 0.5243,
+      "step": 387
+    },
+    {
+      "epoch": 0.6621160409556314,
+      "grad_norm": 0.6591616426149787,
+      "learning_rate": 3.987204329105547e-05,
+      "loss": 0.4863,
+      "step": 388
+    },
+    {
+      "epoch": 0.6638225255972696,
+      "grad_norm": 0.5679402238967721,
+      "learning_rate": 3.986933823999936e-05,
+      "loss": 0.4915,
+      "step": 389
+    },
+    {
+      "epoch": 0.6655290102389079,
+      "grad_norm": 0.556965760174508,
+      "learning_rate": 3.986660498805177e-05,
+      "loss": 0.524,
+      "step": 390
+    },
+    {
+      "epoch": 0.6672354948805461,
+      "grad_norm": 0.6852120576628149,
+      "learning_rate": 3.986384353909205e-05,
+      "loss": 0.5152,
+      "step": 391
+    },
+    {
+      "epoch": 0.6689419795221843,
+      "grad_norm": 0.46228477287702213,
+      "learning_rate": 3.9861053897039585e-05,
+      "loss": 0.5128,
+      "step": 392
+    },
+    {
+      "epoch": 0.6706484641638225,
+      "grad_norm": 0.5520827992904221,
+      "learning_rate": 3.9858236065853745e-05,
+      "loss": 0.5262,
+      "step": 393
+    },
+    {
+      "epoch": 0.6723549488054608,
+      "grad_norm": 0.5637411541812409,
+      "learning_rate": 3.985539004953393e-05,
+      "loss": 0.5347,
+      "step": 394
+    },
+    {
+      "epoch": 0.674061433447099,
+      "grad_norm": 0.48720007300892015,
+      "learning_rate": 3.9852515852119535e-05,
+      "loss": 0.4894,
+      "step": 395
+    },
+    {
+      "epoch": 0.6757679180887372,
+      "grad_norm": 0.5142995634230677,
+      "learning_rate": 3.9849613477689964e-05,
+      "loss": 0.5536,
+      "step": 396
+    },
+    {
+      "epoch": 0.6774744027303754,
+      "grad_norm": 0.5518178426385149,
+      "learning_rate": 3.9846682930364614e-05,
+      "loss": 0.5666,
+      "step": 397
+    },
+    {
+      "epoch": 0.6791808873720137,
+      "grad_norm": 0.46116045794249394,
+      "learning_rate": 3.9843724214302844e-05,
+      "loss": 0.5038,
+      "step": 398
+    },
+    {
+      "epoch": 0.6808873720136519,
+      "grad_norm": 0.4856736298162217,
+      "learning_rate": 3.984073733370402e-05,
+      "loss": 0.4917,
+      "step": 399
+    },
+    {
+      "epoch": 0.6825938566552902,
+      "grad_norm": 0.4968308292207791,
+      "learning_rate": 3.9837722292807465e-05,
+      "loss": 0.4928,
+      "step": 400
+    },
+    {
+      "epoch": 0.6843003412969283,
+      "grad_norm": 0.4375164806803662,
+      "learning_rate": 3.9834679095892494e-05,
+      "loss": 0.5173,
+      "step": 401
+    },
+    {
+      "epoch": 0.6860068259385665,
+      "grad_norm": 0.4632871905277133,
+      "learning_rate": 3.983160774727836e-05,
+      "loss": 0.5085,
+      "step": 402
+    },
+    {
+      "epoch": 0.6877133105802048,
+      "grad_norm": 0.5600303956613327,
+      "learning_rate": 3.982850825132428e-05,
+      "loss": 0.5597,
+      "step": 403
+    },
+    {
+      "epoch": 0.689419795221843,
+      "grad_norm": 0.4806062870610982,
+      "learning_rate": 3.982538061242941e-05,
+      "loss": 0.5102,
+      "step": 404
+    },
+    {
+      "epoch": 0.6911262798634812,
+      "grad_norm": 0.6085239375371078,
+      "learning_rate": 3.982222483503288e-05,
+      "loss": 0.4918,
+      "step": 405
+    },
+    {
+      "epoch": 0.6928327645051194,
+      "grad_norm": 0.530816198322567,
+      "learning_rate": 3.9819040923613734e-05,
+      "loss": 0.5419,
+      "step": 406
+    },
+    {
+      "epoch": 0.6945392491467577,
+      "grad_norm": 0.5500182844719029,
+      "learning_rate": 3.981582888269094e-05,
+      "loss": 0.5077,
+      "step": 407
+    },
+    {
+      "epoch": 0.6962457337883959,
+      "grad_norm": 0.5420512450147166,
+      "learning_rate": 3.9812588716823424e-05,
+      "loss": 0.5057,
+      "step": 408
+    },
+    {
+      "epoch": 0.6979522184300341,
+      "grad_norm": 0.5157060237685485,
+      "learning_rate": 3.980932043060999e-05,
+      "loss": 0.5346,
+      "step": 409
+    },
+    {
+      "epoch": 0.6996587030716723,
+      "grad_norm": 0.651581330036418,
+      "learning_rate": 3.9806024028689376e-05,
+      "loss": 0.5202,
+      "step": 410
+    },
+    {
+      "epoch": 0.7013651877133106,
+      "grad_norm": 0.5661892497248172,
+      "learning_rate": 3.980269951574022e-05,
+      "loss": 0.5427,
+      "step": 411
+    },
+    {
+      "epoch": 0.7030716723549488,
+      "grad_norm": 0.5967504503872553,
+      "learning_rate": 3.979934689648108e-05,
+      "loss": 0.5249,
+      "step": 412
+    },
+    {
+      "epoch": 0.7047781569965871,
+      "grad_norm": 0.5592298812958022,
+      "learning_rate": 3.979596617567036e-05,
+      "loss": 0.5261,
+      "step": 413
+    },
+    {
+      "epoch": 0.7064846416382252,
+      "grad_norm": 0.558091676610057,
+      "learning_rate": 3.9792557358106385e-05,
+      "loss": 0.4912,
+      "step": 414
+    },
+    {
+      "epoch": 0.7081911262798635,
+      "grad_norm": 0.5022628834006392,
+      "learning_rate": 3.978912044862735e-05,
+      "loss": 0.4906,
+      "step": 415
+    },
+    {
+      "epoch": 0.7098976109215017,
+      "grad_norm": 0.5017534110957725,
+      "learning_rate": 3.978565545211132e-05,
+      "loss": 0.5587,
+      "step": 416
+    },
+    {
+      "epoch": 0.71160409556314,
+      "grad_norm": 0.46967662094610146,
+      "learning_rate": 3.978216237347622e-05,
+      "loss": 0.5136,
+      "step": 417
+    },
+    {
+      "epoch": 0.7133105802047781,
+      "grad_norm": 0.4744131857049406,
+      "learning_rate": 3.977864121767985e-05,
+      "loss": 0.5402,
+      "step": 418
+    },
+    {
+      "epoch": 0.7150170648464164,
+      "grad_norm": 0.4959077190551545,
+      "learning_rate": 3.977509198971982e-05,
+      "loss": 0.5232,
+      "step": 419
+    },
+    {
+      "epoch": 0.7167235494880546,
+      "grad_norm": 0.45485104221578665,
+      "learning_rate": 3.977151469463363e-05,
+      "loss": 0.5463,
+      "step": 420
+    },
+    {
+      "epoch": 0.7184300341296929,
+      "grad_norm": 0.4693252104059602,
+      "learning_rate": 3.9767909337498584e-05,
+      "loss": 0.5141,
+      "step": 421
+    },
+    {
+      "epoch": 0.7201365187713311,
+      "grad_norm": 0.38234318138230633,
+      "learning_rate": 3.9764275923431836e-05,
+      "loss": 0.4884,
+      "step": 422
+    },
+    {
+      "epoch": 0.7218430034129693,
+      "grad_norm": 0.483426578271764,
+      "learning_rate": 3.976061445759035e-05,
+      "loss": 0.5446,
+      "step": 423
+    },
+    {
+      "epoch": 0.7235494880546075,
+      "grad_norm": 0.4313890083656951,
+      "learning_rate": 3.9756924945170914e-05,
+      "loss": 0.4995,
+      "step": 424
+    },
+    {
+      "epoch": 0.7252559726962458,
+      "grad_norm": 0.45736306408385247,
+      "learning_rate": 3.97532073914101e-05,
+      "loss": 0.4908,
+      "step": 425
+    },
+    {
+      "epoch": 0.726962457337884,
+      "grad_norm": 0.4431863764164522,
+      "learning_rate": 3.974946180158431e-05,
+      "loss": 0.5272,
+      "step": 426
+    },
+    {
+      "epoch": 0.7286689419795221,
+      "grad_norm": 0.5360713837969308,
+      "learning_rate": 3.9745688181009716e-05,
+      "loss": 0.537,
+      "step": 427
+    },
+    {
+      "epoch": 0.7303754266211604,
+      "grad_norm": 0.45208450010601464,
+      "learning_rate": 3.974188653504229e-05,
+      "loss": 0.5191,
+      "step": 428
+    },
+    {
+      "epoch": 0.7320819112627986,
+      "grad_norm": 0.5989479913873644,
+      "learning_rate": 3.973805686907777e-05,
+      "loss": 0.5399,
+      "step": 429
+    },
+    {
+      "epoch": 0.7337883959044369,
+      "grad_norm": 0.4208920353633511,
+      "learning_rate": 3.9734199188551655e-05,
+      "loss": 0.5136,
+      "step": 430
+    },
+    {
+      "epoch": 0.735494880546075,
+      "grad_norm": 0.5169436211718849,
+      "learning_rate": 3.9730313498939225e-05,
+      "loss": 0.5047,
+      "step": 431
+    },
+    {
+      "epoch": 0.7372013651877133,
+      "grad_norm": 0.45544132069369514,
+      "learning_rate": 3.972639980575552e-05,
+      "loss": 0.5041,
+      "step": 432
+    },
+    {
+      "epoch": 0.7389078498293515,
+      "grad_norm": 0.45687457737266457,
+      "learning_rate": 3.972245811455529e-05,
+      "loss": 0.4646,
+      "step": 433
+    },
+    {
+      "epoch": 0.7406143344709898,
+      "grad_norm": 0.5039590590912099,
+      "learning_rate": 3.971848843093305e-05,
+      "loss": 0.5065,
+      "step": 434
+    },
+    {
+      "epoch": 0.742320819112628,
+      "grad_norm": 0.4656271645725468,
+      "learning_rate": 3.971449076052305e-05,
+      "loss": 0.5204,
+      "step": 435
+    },
+    {
+      "epoch": 0.7440273037542662,
+      "grad_norm": 0.4755396482692981,
+      "learning_rate": 3.9710465108999245e-05,
+      "loss": 0.5004,
+      "step": 436
+    },
+    {
+      "epoch": 0.7457337883959044,
+      "grad_norm": 0.5280369390574324,
+      "learning_rate": 3.9706411482075304e-05,
+      "loss": 0.5469,
+      "step": 437
+    },
+    {
+      "epoch": 0.7474402730375427,
+      "grad_norm": 0.4588996567849872,
+      "learning_rate": 3.970232988550462e-05,
+      "loss": 0.5059,
+      "step": 438
+    },
+    {
+      "epoch": 0.7491467576791809,
+      "grad_norm": 0.5067536606016981,
+      "learning_rate": 3.9698220325080275e-05,
+      "loss": 0.5132,
+      "step": 439
+    },
+    {
+      "epoch": 0.7508532423208191,
+      "grad_norm": 0.4701245104251668,
+      "learning_rate": 3.9694082806635026e-05,
+      "loss": 0.4901,
+      "step": 440
+    },
+    {
+      "epoch": 0.7525597269624573,
+      "grad_norm": 0.46968979252829884,
+      "learning_rate": 3.9689917336041336e-05,
+      "loss": 0.5426,
+      "step": 441
+    },
+    {
+      "epoch": 0.7542662116040956,
+      "grad_norm": 0.49805975671632624,
+      "learning_rate": 3.9685723919211316e-05,
+      "loss": 0.5136,
+      "step": 442
+    },
+    {
+      "epoch": 0.7559726962457338,
+      "grad_norm": 0.4726176407036461,
+      "learning_rate": 3.9681502562096764e-05,
+      "loss": 0.5295,
+      "step": 443
+    },
+    {
+      "epoch": 0.757679180887372,
+      "grad_norm": 0.5199428138549627,
+      "learning_rate": 3.9677253270689116e-05,
+      "loss": 0.4818,
+      "step": 444
+    },
+    {
+      "epoch": 0.7593856655290102,
+      "grad_norm": 0.40331331997993186,
+      "learning_rate": 3.9672976051019477e-05,
+      "loss": 0.5205,
+      "step": 445
+    },
+    {
+      "epoch": 0.7610921501706485,
+      "grad_norm": 0.5327355482683795,
+      "learning_rate": 3.9668670909158565e-05,
+      "loss": 0.5009,
+      "step": 446
+    },
+    {
+      "epoch": 0.7627986348122867,
+      "grad_norm": 0.43032587423532564,
+      "learning_rate": 3.966433785121675e-05,
+      "loss": 0.5193,
+      "step": 447
+    },
+    {
+      "epoch": 0.764505119453925,
+      "grad_norm": 0.4506217825601562,
+      "learning_rate": 3.965997688334401e-05,
+      "loss": 0.5237,
+      "step": 448
+    },
+    {
+      "epoch": 0.7662116040955631,
+      "grad_norm": 0.4492241327004955,
+      "learning_rate": 3.965558801172994e-05,
+      "loss": 0.5027,
+      "step": 449
+    },
+    {
+      "epoch": 0.7679180887372014,
+      "grad_norm": 0.5212413159647389,
+      "learning_rate": 3.9651171242603746e-05,
+      "loss": 0.6042,
+      "step": 450
+    },
+    {
+      "epoch": 0.7696245733788396,
+      "grad_norm": 0.5075927470071195,
+      "learning_rate": 3.964672658223422e-05,
+      "loss": 0.5733,
+      "step": 451
+    },
+    {
+      "epoch": 0.7713310580204779,
+      "grad_norm": 0.5239354700021109,
+      "learning_rate": 3.964225403692975e-05,
+      "loss": 0.5156,
+      "step": 452
+    },
+    {
+      "epoch": 0.773037542662116,
+      "grad_norm": 0.4531072066408163,
+      "learning_rate": 3.963775361303829e-05,
+      "loss": 0.5297,
+      "step": 453
+    },
+    {
+      "epoch": 0.7747440273037542,
+      "grad_norm": 0.4599191961695215,
+      "learning_rate": 3.963322531694737e-05,
+      "loss": 0.5257,
+      "step": 454
+    },
+    {
+      "epoch": 0.7764505119453925,
+      "grad_norm": 0.5001588299554403,
+      "learning_rate": 3.962866915508408e-05,
+      "loss": 0.4549,
+      "step": 455
+    },
+    {
+      "epoch": 0.7781569965870307,
+      "grad_norm": 0.5038191454985079,
+      "learning_rate": 3.962408513391505e-05,
+      "loss": 0.5941,
+      "step": 456
+    },
+    {
+      "epoch": 0.7798634812286689,
+      "grad_norm": 0.5772536453395402,
+      "learning_rate": 3.961947325994648e-05,
+      "loss": 0.6,
+      "step": 457
+    },
+    {
+      "epoch": 0.7815699658703071,
+      "grad_norm": 0.4475196746073441,
+      "learning_rate": 3.961483353972406e-05,
+      "loss": 0.535,
+      "step": 458
+    },
+    {
+      "epoch": 0.7832764505119454,
+      "grad_norm": 0.5310562522442643,
+      "learning_rate": 3.961016597983303e-05,
+      "loss": 0.5483,
+      "step": 459
+    },
+    {
+      "epoch": 0.7849829351535836,
+      "grad_norm": 0.4589968743514119,
+      "learning_rate": 3.960547058689814e-05,
+      "loss": 0.5876,
+      "step": 460
+    },
+    {
+      "epoch": 0.7866894197952219,
+      "grad_norm": 0.5334530981422132,
+      "learning_rate": 3.960074736758365e-05,
+      "loss": 0.5141,
+      "step": 461
+    },
+    {
+      "epoch": 0.78839590443686,
+      "grad_norm": 0.530776151808494,
+      "learning_rate": 3.9595996328593293e-05,
+      "loss": 0.5312,
+      "step": 462
+    },
+    {
+      "epoch": 0.7901023890784983,
+      "grad_norm": 0.4857358421044376,
+      "learning_rate": 3.9591217476670306e-05,
+      "loss": 0.5749,
+      "step": 463
+    },
+    {
+      "epoch": 0.7918088737201365,
+      "grad_norm": 0.48944429444835524,
+      "learning_rate": 3.958641081859739e-05,
+      "loss": 0.4869,
+      "step": 464
+    },
+    {
+      "epoch": 0.7935153583617748,
+      "grad_norm": 0.5512252561788058,
+      "learning_rate": 3.958157636119672e-05,
+      "loss": 0.5265,
+      "step": 465
+    },
+    {
+      "epoch": 0.7952218430034129,
+      "grad_norm": 0.5493204923836218,
+      "learning_rate": 3.9576714111329926e-05,
+      "loss": 0.479,
+      "step": 466
+    },
+    {
+      "epoch": 0.7969283276450512,
+      "grad_norm": 0.42907462218706693,
+      "learning_rate": 3.957182407589809e-05,
+      "loss": 0.5077,
+      "step": 467
+    },
+    {
+      "epoch": 0.7986348122866894,
+      "grad_norm": 0.5833402640110258,
+      "learning_rate": 3.9566906261841694e-05,
+      "loss": 0.5156,
+      "step": 468
+    },
+    {
+      "epoch": 0.8003412969283277,
+      "grad_norm": 0.5027207242408734,
+      "learning_rate": 3.956196067614071e-05,
+      "loss": 0.5389,
+      "step": 469
+    },
+    {
+      "epoch": 0.8020477815699659,
+      "grad_norm": 0.5903775732262893,
+      "learning_rate": 3.9556987325814474e-05,
+      "loss": 0.5127,
+      "step": 470
+    },
+    {
+      "epoch": 0.8037542662116041,
+      "grad_norm": 0.4489790026763905,
+      "learning_rate": 3.9551986217921755e-05,
+      "loss": 0.5033,
+      "step": 471
+    },
+    {
+      "epoch": 0.8054607508532423,
+      "grad_norm": 0.6737608327501232,
+      "learning_rate": 3.9546957359560704e-05,
+      "loss": 0.4755,
+      "step": 472
+    },
+    {
+      "epoch": 0.8071672354948806,
+      "grad_norm": 0.4866463991441708,
+      "learning_rate": 3.954190075786887e-05,
+      "loss": 0.5537,
+      "step": 473
+    },
+    {
+      "epoch": 0.8088737201365188,
+      "grad_norm": 0.6150681127724358,
+      "learning_rate": 3.953681642002317e-05,
+      "loss": 0.5031,
+      "step": 474
+    },
+    {
+      "epoch": 0.810580204778157,
+      "grad_norm": 0.43728211056504895,
+      "learning_rate": 3.9531704353239895e-05,
+      "loss": 0.5052,
+      "step": 475
+    },
+    {
+      "epoch": 0.8122866894197952,
+      "grad_norm": 0.510925067172793,
+      "learning_rate": 3.9526564564774685e-05,
+      "loss": 0.5382,
+      "step": 476
+    },
+    {
+      "epoch": 0.8139931740614335,
+      "grad_norm": 0.6163408488743702,
+      "learning_rate": 3.9521397061922536e-05,
+      "loss": 0.5657,
+      "step": 477
+    },
+    {
+      "epoch": 0.8156996587030717,
+      "grad_norm": 0.5176686859913134,
+      "learning_rate": 3.951620185201777e-05,
+      "loss": 0.521,
+      "step": 478
+    },
+    {
+      "epoch": 0.8174061433447098,
+      "grad_norm": 0.4271267095214705,
+      "learning_rate": 3.951097894243404e-05,
+      "loss": 0.5036,
+      "step": 479
+    },
+    {
+      "epoch": 0.8191126279863481,
+      "grad_norm": 0.6514774607052338,
+      "learning_rate": 3.9505728340584305e-05,
+      "loss": 0.5813,
+      "step": 480
+    },
+    {
+      "epoch": 0.8208191126279863,
+      "grad_norm": 0.4368536071383127,
+      "learning_rate": 3.950045005392084e-05,
+      "loss": 0.5416,
+      "step": 481
+    },
+    {
+      "epoch": 0.8225255972696246,
+      "grad_norm": 0.5044516773539471,
+      "learning_rate": 3.94951440899352e-05,
+      "loss": 0.4848,
+      "step": 482
+    },
+    {
+      "epoch": 0.8242320819112628,
+      "grad_norm": 0.5344943071649257,
+      "learning_rate": 3.948981045615823e-05,
+      "loss": 0.6499,
+      "step": 483
+    },
+    {
+      "epoch": 0.825938566552901,
+      "grad_norm": 0.515701849186856,
+      "learning_rate": 3.9484449160160064e-05,
+      "loss": 0.5208,
+      "step": 484
+    },
+    {
+      "epoch": 0.8276450511945392,
+      "grad_norm": 0.518510086224629,
+      "learning_rate": 3.9479060209550066e-05,
+      "loss": 0.5357,
+      "step": 485
+    },
+    {
+      "epoch": 0.8293515358361775,
+      "grad_norm": 0.4916579823521775,
+      "learning_rate": 3.947364361197687e-05,
+      "loss": 0.5195,
+      "step": 486
+    },
+    {
+      "epoch": 0.8310580204778157,
+      "grad_norm": 0.5022037731108061,
+      "learning_rate": 3.946819937512835e-05,
+      "loss": 0.6218,
+      "step": 487
+    },
+    {
+      "epoch": 0.8327645051194539,
+      "grad_norm": 0.5325920446887586,
+      "learning_rate": 3.9462727506731584e-05,
+      "loss": 0.5368,
+      "step": 488
+    },
+    {
+      "epoch": 0.8344709897610921,
+      "grad_norm": 0.47314747320582723,
+      "learning_rate": 3.9457228014552916e-05,
+      "loss": 0.4832,
+      "step": 489
+    },
+    {
+      "epoch": 0.8361774744027304,
+      "grad_norm": 0.5165856181868046,
+      "learning_rate": 3.9451700906397855e-05,
+      "loss": 0.5393,
+      "step": 490
+    },
+    {
+      "epoch": 0.8378839590443686,
+      "grad_norm": 0.604855358248186,
+      "learning_rate": 3.944614619011112e-05,
+      "loss": 0.5501,
+      "step": 491
+    },
+    {
+      "epoch": 0.8395904436860068,
+      "grad_norm": 0.5374469090634256,
+      "learning_rate": 3.944056387357662e-05,
+      "loss": 0.5336,
+      "step": 492
+    },
+    {
+      "epoch": 0.841296928327645,
+      "grad_norm": 0.4921963833425323,
+      "learning_rate": 3.9434953964717424e-05,
+      "loss": 0.5294,
+      "step": 493
+    },
+    {
+      "epoch": 0.8430034129692833,
+      "grad_norm": 0.5319944552023902,
+      "learning_rate": 3.9429316471495777e-05,
+      "loss": 0.4934,
+      "step": 494
+    },
+    {
+      "epoch": 0.8447098976109215,
+      "grad_norm": 0.4481641231154688,
+      "learning_rate": 3.9423651401913074e-05,
+      "loss": 0.5038,
+      "step": 495
+    },
+    {
+      "epoch": 0.8464163822525598,
+      "grad_norm": 0.6046057595807798,
+      "learning_rate": 3.941795876400984e-05,
+      "loss": 0.5661,
+      "step": 496
+    },
+    {
+      "epoch": 0.8481228668941979,
+      "grad_norm": 0.4588428131899579,
+      "learning_rate": 3.941223856586573e-05,
+      "loss": 0.5179,
+      "step": 497
+    },
+    {
+      "epoch": 0.8498293515358362,
+      "grad_norm": 0.49445101667923497,
+      "learning_rate": 3.940649081559953e-05,
+      "loss": 0.5383,
+      "step": 498
+    },
+    {
+      "epoch": 0.8515358361774744,
+      "grad_norm": 0.47792535762506255,
+      "learning_rate": 3.9400715521369106e-05,
+      "loss": 0.494,
+      "step": 499
+    },
+    {
+      "epoch": 0.8532423208191127,
+      "grad_norm": 0.4709617414602991,
+      "learning_rate": 3.939491269137144e-05,
+      "loss": 0.5158,
+      "step": 500
+    },
+    {
+      "epoch": 0.8549488054607508,
+      "grad_norm": 0.4287421664931594,
+      "learning_rate": 3.938908233384259e-05,
+      "loss": 0.5037,
+      "step": 501
+    },
+    {
+      "epoch": 0.856655290102389,
+      "grad_norm": 0.4554952824428464,
+      "learning_rate": 3.9383224457057676e-05,
+      "loss": 0.5316,
+      "step": 502
+    },
+    {
+      "epoch": 0.8583617747440273,
+      "grad_norm": 0.47415170222164305,
+      "learning_rate": 3.937733906933089e-05,
+      "loss": 0.5595,
+      "step": 503
+    },
+    {
+      "epoch": 0.8600682593856656,
+      "grad_norm": 0.4784593159132243,
+      "learning_rate": 3.937142617901545e-05,
+      "loss": 0.5256,
+      "step": 504
+    },
+    {
+      "epoch": 0.8617747440273038,
+      "grad_norm": 0.5224405316117319,
+      "learning_rate": 3.936548579450364e-05,
+      "loss": 0.5356,
+      "step": 505
+    },
+    {
+      "epoch": 0.863481228668942,
+      "grad_norm": 0.5062461175533289,
+      "learning_rate": 3.9359517924226734e-05,
+      "loss": 0.4793,
+      "step": 506
+    },
+    {
+      "epoch": 0.8651877133105802,
+      "grad_norm": 0.46137996691684924,
+      "learning_rate": 3.9353522576655045e-05,
+      "loss": 0.5226,
+      "step": 507
+    },
+    {
+      "epoch": 0.8668941979522184,
+      "grad_norm": 0.57463265864944,
+      "learning_rate": 3.9347499760297864e-05,
+      "loss": 0.541,
+      "step": 508
+    },
+    {
+      "epoch": 0.8686006825938567,
+      "grad_norm": 0.446650800221027,
+      "learning_rate": 3.9341449483703474e-05,
+      "loss": 0.5278,
+      "step": 509
+    },
+    {
+      "epoch": 0.8703071672354948,
+      "grad_norm": 0.6374975100826807,
+      "learning_rate": 3.933537175545914e-05,
+      "loss": 0.487,
+      "step": 510
+    },
+    {
+      "epoch": 0.8720136518771331,
+      "grad_norm": 0.5051408760304202,
+      "learning_rate": 3.93292665841911e-05,
+      "loss": 0.5465,
+      "step": 511
+    },
+    {
+      "epoch": 0.8737201365187713,
+      "grad_norm": 0.4877550532041799,
+      "learning_rate": 3.9323133978564506e-05,
+      "loss": 0.4792,
+      "step": 512
+    },
+    {
+      "epoch": 0.8754266211604096,
+      "grad_norm": 0.47071774112823295,
+      "learning_rate": 3.931697394728348e-05,
+      "loss": 0.5222,
+      "step": 513
+    },
+    {
+      "epoch": 0.8771331058020477,
+      "grad_norm": 0.4832616488734822,
+      "learning_rate": 3.9310786499091055e-05,
+      "loss": 0.5238,
+      "step": 514
+    },
+    {
+      "epoch": 0.878839590443686,
+      "grad_norm": 0.4622626510648489,
+      "learning_rate": 3.9304571642769194e-05,
+      "loss": 0.5331,
+      "step": 515
+    },
+    {
+      "epoch": 0.8805460750853242,
+      "grad_norm": 0.49609456268044955,
+      "learning_rate": 3.9298329387138735e-05,
+      "loss": 0.5017,
+      "step": 516
+    },
+    {
+      "epoch": 0.8822525597269625,
+      "grad_norm": 0.3894121360741194,
+      "learning_rate": 3.9292059741059426e-05,
+      "loss": 0.5245,
+      "step": 517
+    },
+    {
+      "epoch": 0.8839590443686007,
+      "grad_norm": 0.5017588035273783,
+      "learning_rate": 3.928576271342988e-05,
+      "loss": 0.5081,
+      "step": 518
+    },
+    {
+      "epoch": 0.8856655290102389,
+      "grad_norm": 0.4077781952167856,
+      "learning_rate": 3.927943831318757e-05,
+      "loss": 0.4843,
+      "step": 519
+    },
+    {
+      "epoch": 0.8873720136518771,
+      "grad_norm": 0.47941716917477173,
+      "learning_rate": 3.927308654930884e-05,
+      "loss": 0.5113,
+      "step": 520
+    },
+    {
+      "epoch": 0.8890784982935154,
+      "grad_norm": 0.48013869845261975,
+      "learning_rate": 3.9266707430808845e-05,
+      "loss": 0.5065,
+      "step": 521
+    },
+    {
+      "epoch": 0.8907849829351536,
+      "grad_norm": 0.469542715885672,
+      "learning_rate": 3.926030096674159e-05,
+      "loss": 0.5461,
+      "step": 522
+    },
+    {
+      "epoch": 0.8924914675767918,
+      "grad_norm": 0.4060262709657794,
+      "learning_rate": 3.925386716619986e-05,
+      "loss": 0.502,
+      "step": 523
+    },
+    {
+      "epoch": 0.89419795221843,
+      "grad_norm": 0.48621669422457514,
+      "learning_rate": 3.9247406038315274e-05,
+      "loss": 0.5422,
+      "step": 524
+    },
+    {
+      "epoch": 0.8959044368600683,
+      "grad_norm": 0.4199510705050333,
+      "learning_rate": 3.924091759225821e-05,
+      "loss": 0.5113,
+      "step": 525
+    },
+    {
+      "epoch": 0.8976109215017065,
+      "grad_norm": 0.450553534259471,
+      "learning_rate": 3.9234401837237846e-05,
+      "loss": 0.5366,
+      "step": 526
+    },
+    {
+      "epoch": 0.8993174061433447,
+      "grad_norm": 0.40428967397574317,
+      "learning_rate": 3.9227858782502084e-05,
+      "loss": 0.4976,
+      "step": 527
+    },
+    {
+      "epoch": 0.9010238907849829,
+      "grad_norm": 0.4956441669290184,
+      "learning_rate": 3.92212884373376e-05,
+      "loss": 0.4949,
+      "step": 528
+    },
+    {
+      "epoch": 0.9027303754266212,
+      "grad_norm": 0.421791557031393,
+      "learning_rate": 3.9214690811069814e-05,
+      "loss": 0.485,
+      "step": 529
+    },
+    {
+      "epoch": 0.9044368600682594,
+      "grad_norm": 0.5055936934279345,
+      "learning_rate": 3.9208065913062824e-05,
+      "loss": 0.5163,
+      "step": 530
+    },
+    {
+      "epoch": 0.9061433447098977,
+      "grad_norm": 0.5545843080358152,
+      "learning_rate": 3.9201413752719484e-05,
+      "loss": 0.556,
+      "step": 531
+    },
+    {
+      "epoch": 0.9078498293515358,
+      "grad_norm": 0.4589640114774441,
+      "learning_rate": 3.9194734339481304e-05,
+      "loss": 0.4826,
+      "step": 532
+    },
+    {
+      "epoch": 0.909556313993174,
+      "grad_norm": 0.5207402038709883,
+      "learning_rate": 3.9188027682828494e-05,
+      "loss": 0.5142,
+      "step": 533
+    },
+    {
+      "epoch": 0.9112627986348123,
+      "grad_norm": 0.44635435968918563,
+      "learning_rate": 3.918129379227992e-05,
+      "loss": 0.5301,
+      "step": 534
+    },
+    {
+      "epoch": 0.9129692832764505,
+      "grad_norm": 0.4126548402669452,
+      "learning_rate": 3.917453267739313e-05,
+      "loss": 0.4616,
+      "step": 535
+    },
+    {
+      "epoch": 0.9146757679180887,
+      "grad_norm": 0.40596036295652865,
+      "learning_rate": 3.916774434776426e-05,
+      "loss": 0.5175,
+      "step": 536
+    },
+    {
+      "epoch": 0.9163822525597269,
+      "grad_norm": 0.4071549783983065,
+      "learning_rate": 3.916092881302812e-05,
+      "loss": 0.4828,
+      "step": 537
+    },
+    {
+      "epoch": 0.9180887372013652,
+      "grad_norm": 0.41815613090929515,
+      "learning_rate": 3.915408608285812e-05,
+      "loss": 0.5181,
+      "step": 538
+    },
+    {
+      "epoch": 0.9197952218430034,
+      "grad_norm": 0.4646522949101528,
+      "learning_rate": 3.914721616696625e-05,
+      "loss": 0.5372,
+      "step": 539
+    },
+    {
+      "epoch": 0.9215017064846417,
+      "grad_norm": 0.386392680741229,
+      "learning_rate": 3.9140319075103105e-05,
+      "loss": 0.4834,
+      "step": 540
+    },
+    {
+      "epoch": 0.9232081911262798,
+      "grad_norm": 0.4248500201352696,
+      "learning_rate": 3.9133394817057844e-05,
+      "loss": 0.5204,
+      "step": 541
+    },
+    {
+      "epoch": 0.9249146757679181,
+      "grad_norm": 0.4582581997897516,
+      "learning_rate": 3.912644340265819e-05,
+      "loss": 0.4969,
+      "step": 542
+    },
+    {
+      "epoch": 0.9266211604095563,
+      "grad_norm": 0.3959557100960288,
+      "learning_rate": 3.91194648417704e-05,
+      "loss": 0.5118,
+      "step": 543
+    },
+    {
+      "epoch": 0.9283276450511946,
+      "grad_norm": 0.47939921367511734,
+      "learning_rate": 3.9112459144299255e-05,
+      "loss": 0.5207,
+      "step": 544
+    },
+    {
+      "epoch": 0.9300341296928327,
+      "grad_norm": 0.43658501873671096,
+      "learning_rate": 3.910542632018808e-05,
+      "loss": 0.502,
+      "step": 545
+    },
+    {
+      "epoch": 0.931740614334471,
+      "grad_norm": 0.4608285168964317,
+      "learning_rate": 3.909836637941867e-05,
+      "loss": 0.4967,
+      "step": 546
+    },
+    {
+      "epoch": 0.9334470989761092,
+      "grad_norm": 0.4452967998329967,
+      "learning_rate": 3.909127933201133e-05,
+      "loss": 0.5175,
+      "step": 547
+    },
+    {
+      "epoch": 0.9351535836177475,
+      "grad_norm": 0.42903249907341506,
+      "learning_rate": 3.908416518802481e-05,
+      "loss": 0.4919,
+      "step": 548
+    },
+    {
+      "epoch": 0.9368600682593856,
+      "grad_norm": 0.4983362738620335,
+      "learning_rate": 3.907702395755636e-05,
+      "loss": 0.5544,
+      "step": 549
+    },
+    {
+      "epoch": 0.9385665529010239,
+      "grad_norm": 0.48137760016486825,
+      "learning_rate": 3.906985565074163e-05,
+      "loss": 0.4978,
+      "step": 550
+    },
+    {
+      "epoch": 0.9402730375426621,
+      "grad_norm": 0.42814076114002747,
+      "learning_rate": 3.9062660277754726e-05,
+      "loss": 0.5696,
+      "step": 551
+    },
+    {
+      "epoch": 0.9419795221843004,
+      "grad_norm": 0.480834195450671,
+      "learning_rate": 3.905543784880817e-05,
+      "loss": 0.5441,
+      "step": 552
+    },
+    {
+      "epoch": 0.9436860068259386,
+      "grad_norm": 0.46675322544792136,
+      "learning_rate": 3.9048188374152875e-05,
+      "loss": 0.5334,
+      "step": 553
+    },
+    {
+      "epoch": 0.9453924914675768,
+      "grad_norm": 0.459122809136642,
+      "learning_rate": 3.904091186407815e-05,
+      "loss": 0.4805,
+      "step": 554
+    },
+    {
+      "epoch": 0.947098976109215,
+      "grad_norm": 0.45818196676392814,
+      "learning_rate": 3.9033608328911655e-05,
+      "loss": 0.5095,
+      "step": 555
+    },
+    {
+      "epoch": 0.9488054607508533,
+      "grad_norm": 0.441105921685137,
+      "learning_rate": 3.9026277779019434e-05,
+      "loss": 0.4756,
+      "step": 556
+    },
+    {
+      "epoch": 0.9505119453924915,
+      "grad_norm": 0.5008988029978901,
+      "learning_rate": 3.901892022480586e-05,
+      "loss": 0.5224,
+      "step": 557
+    },
+    {
+      "epoch": 0.9522184300341296,
+      "grad_norm": 0.47988835378586664,
+      "learning_rate": 3.9011535676713636e-05,
+      "loss": 0.4853,
+      "step": 558
+    },
+    {
+      "epoch": 0.9539249146757679,
+      "grad_norm": 0.4372032012177308,
+      "learning_rate": 3.900412414522378e-05,
+      "loss": 0.4946,
+      "step": 559
+    },
+    {
+      "epoch": 0.9556313993174061,
+      "grad_norm": 0.5358214314599237,
+      "learning_rate": 3.899668564085559e-05,
+      "loss": 0.5006,
+      "step": 560
+    },
+    {
+      "epoch": 0.9573378839590444,
+      "grad_norm": 0.45911660732913245,
+      "learning_rate": 3.898922017416668e-05,
+      "loss": 0.5374,
+      "step": 561
+    },
+    {
+      "epoch": 0.9590443686006825,
+      "grad_norm": 0.44658419169074315,
+      "learning_rate": 3.898172775575291e-05,
+      "loss": 0.5209,
+      "step": 562
+    },
+    {
+      "epoch": 0.9607508532423208,
+      "grad_norm": 0.4310985352672766,
+      "learning_rate": 3.89742083962484e-05,
+      "loss": 0.4737,
+      "step": 563
+    },
+    {
+      "epoch": 0.962457337883959,
+      "grad_norm": 0.4122180714819184,
+      "learning_rate": 3.8966662106325495e-05,
+      "loss": 0.5104,
+      "step": 564
+    },
+    {
+      "epoch": 0.9641638225255973,
+      "grad_norm": 0.44427959014053525,
+      "learning_rate": 3.8959088896694785e-05,
+      "loss": 0.4915,
+      "step": 565
+    },
+    {
+      "epoch": 0.9658703071672355,
+      "grad_norm": 0.44001548707309424,
+      "learning_rate": 3.8951488778105054e-05,
+      "loss": 0.5102,
+      "step": 566
+    },
+    {
+      "epoch": 0.9675767918088737,
+      "grad_norm": 0.47205644918158723,
+      "learning_rate": 3.894386176134327e-05,
+      "loss": 0.5043,
+      "step": 567
+    },
+    {
+      "epoch": 0.9692832764505119,
+      "grad_norm": 0.4799588163560951,
+      "learning_rate": 3.89362078572346e-05,
+      "loss": 0.5007,
+      "step": 568
+    },
+    {
+      "epoch": 0.9709897610921502,
+      "grad_norm": 0.4371472093668799,
+      "learning_rate": 3.892852707664235e-05,
+      "loss": 0.5157,
+      "step": 569
+    },
+    {
+      "epoch": 0.9726962457337884,
+      "grad_norm": 0.45620099061021785,
+      "learning_rate": 3.892081943046799e-05,
+      "loss": 0.5083,
+      "step": 570
+    },
+    {
+      "epoch": 0.9744027303754266,
+      "grad_norm": 0.4902069995205142,
+      "learning_rate": 3.891308492965112e-05,
+      "loss": 0.5263,
+      "step": 571
+    },
+    {
+      "epoch": 0.9761092150170648,
+      "grad_norm": 0.37562437603564053,
+      "learning_rate": 3.890532358516944e-05,
+      "loss": 0.4775,
+      "step": 572
+    },
+    {
+      "epoch": 0.9778156996587031,
+      "grad_norm": 0.45149513920661594,
+      "learning_rate": 3.889753540803876e-05,
+      "loss": 0.5118,
+      "step": 573
+    },
+    {
+      "epoch": 0.9795221843003413,
+      "grad_norm": 0.45070569518442194,
+      "learning_rate": 3.888972040931299e-05,
+      "loss": 0.4998,
+      "step": 574
+    },
+    {
+      "epoch": 0.9812286689419796,
+      "grad_norm": 0.4554162489722807,
+      "learning_rate": 3.8881878600084053e-05,
+      "loss": 0.5143,
+      "step": 575
+    },
+    {
+      "epoch": 0.9829351535836177,
+      "grad_norm": 0.46808574128865893,
+      "learning_rate": 3.8874009991482e-05,
+      "loss": 0.5106,
+      "step": 576
+    },
+    {
+      "epoch": 0.984641638225256,
+      "grad_norm": 0.44264827285152525,
+      "learning_rate": 3.8866114594674865e-05,
+      "loss": 0.5096,
+      "step": 577
+    },
+    {
+      "epoch": 0.9863481228668942,
+      "grad_norm": 0.38689828803482745,
+      "learning_rate": 3.885819242086872e-05,
+      "loss": 0.4652,
+      "step": 578
+    },
+    {
+      "epoch": 0.9880546075085325,
+      "grad_norm": 0.44863248742872985,
+      "learning_rate": 3.885024348130765e-05,
+      "loss": 0.5021,
+      "step": 579
+    },
+    {
+      "epoch": 0.9897610921501706,
+      "grad_norm": 0.4376711926299938,
+      "learning_rate": 3.884226778727371e-05,
+      "loss": 0.4894,
+      "step": 580
+    },
+    {
+      "epoch": 0.9914675767918089,
+      "grad_norm": 0.4836618691935011,
+      "learning_rate": 3.883426535008694e-05,
+      "loss": 0.4912,
+      "step": 581
+    },
+    {
+      "epoch": 0.9931740614334471,
+      "grad_norm": 0.4634093982781147,
+      "learning_rate": 3.8826236181105344e-05,
+      "loss": 0.5249,
+      "step": 582
+    },
+    {
+      "epoch": 0.9948805460750854,
+      "grad_norm": 0.45182596219251353,
+      "learning_rate": 3.8818180291724855e-05,
+      "loss": 0.4644,
+      "step": 583
+    },
+    {
+      "epoch": 0.9965870307167235,
+      "grad_norm": 0.4524376905770041,
+      "learning_rate": 3.8810097693379336e-05,
+      "loss": 0.5234,
+      "step": 584
+    },
+    {
+      "epoch": 0.9982935153583617,
+      "grad_norm": 0.41527674638945133,
+      "learning_rate": 3.8801988397540554e-05,
+      "loss": 0.5019,
+      "step": 585
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.45259487503676415,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.527,
+      "step": 586
+    },
+    {
+      "epoch": 1.0017064846416381,
+      "grad_norm": 0.5439507243208022,
+      "learning_rate": 3.878568975945973e-05,
+      "loss": 0.4537,
+      "step": 587
+    },
+    {
+      "epoch": 1.0034129692832765,
+      "grad_norm": 0.5332516474258484,
+      "learning_rate": 3.877750044035062e-05,
+      "loss": 0.4203,
+      "step": 588
+    },
+    {
+      "epoch": 1.0051194539249146,
+      "grad_norm": 0.48766492268460837,
+      "learning_rate": 3.876928447001409e-05,
+      "loss": 0.4501,
+      "step": 589
+    },
+    {
+      "epoch": 1.006825938566553,
+      "grad_norm": 0.4831041742336393,
+      "learning_rate": 3.8761041860111206e-05,
+      "loss": 0.3917,
+      "step": 590
+    },
+    {
+      "epoch": 1.0085324232081911,
+      "grad_norm": 0.5381540681172681,
+      "learning_rate": 3.875277262234083e-05,
+      "loss": 0.4446,
+      "step": 591
+    },
+    {
+      "epoch": 1.0102389078498293,
+      "grad_norm": 0.5320198688078498,
+      "learning_rate": 3.874447676843966e-05,
+      "loss": 0.4561,
+      "step": 592
+    },
+    {
+      "epoch": 1.0119453924914676,
+      "grad_norm": 0.46283902661104653,
+      "learning_rate": 3.873615431018213e-05,
+      "loss": 0.4123,
+      "step": 593
+    },
+    {
+      "epoch": 1.0136518771331058,
+      "grad_norm": 0.4538227503029545,
+      "learning_rate": 3.872780525938044e-05,
+      "loss": 0.4069,
+      "step": 594
+    },
+    {
+      "epoch": 1.015358361774744,
+      "grad_norm": 0.44200735879492625,
+      "learning_rate": 3.8719429627884544e-05,
+      "loss": 0.4557,
+      "step": 595
+    },
+    {
+      "epoch": 1.0170648464163823,
+      "grad_norm": 0.46052859455189543,
+      "learning_rate": 3.8711027427582126e-05,
+      "loss": 0.3977,
+      "step": 596
+    },
+    {
+      "epoch": 1.0187713310580204,
+      "grad_norm": 0.46912938323050846,
+      "learning_rate": 3.870259867039857e-05,
+      "loss": 0.4253,
+      "step": 597
+    },
+    {
+      "epoch": 1.0204778156996588,
+      "grad_norm": 0.48104790336408815,
+      "learning_rate": 3.869414336829695e-05,
+      "loss": 0.4317,
+      "step": 598
+    },
+    {
+      "epoch": 1.022184300341297,
+      "grad_norm": 0.5070473741322639,
+      "learning_rate": 3.8685661533278026e-05,
+      "loss": 0.4017,
+      "step": 599
+    },
+    {
+      "epoch": 1.023890784982935,
+      "grad_norm": 0.5337220776817938,
+      "learning_rate": 3.8677153177380206e-05,
+      "loss": 0.3987,
+      "step": 600
+    },
+    {
+      "epoch": 1.0255972696245734,
+      "grad_norm": 0.5046087234277966,
+      "learning_rate": 3.8668618312679556e-05,
+      "loss": 0.4662,
+      "step": 601
+    },
+    {
+      "epoch": 1.0273037542662116,
+      "grad_norm": 0.5209637119395667,
+      "learning_rate": 3.866005695128974e-05,
+      "loss": 0.4637,
+      "step": 602
+    },
+    {
+      "epoch": 1.02901023890785,
+      "grad_norm": 0.4826434962536877,
+      "learning_rate": 3.865146910536206e-05,
+      "loss": 0.4091,
+      "step": 603
+    },
+    {
+      "epoch": 1.030716723549488,
+      "grad_norm": 0.47968781942819244,
+      "learning_rate": 3.864285478708538e-05,
+      "loss": 0.4089,
+      "step": 604
+    },
+    {
+      "epoch": 1.0324232081911262,
+      "grad_norm": 0.46153618460883944,
+      "learning_rate": 3.8634214008686155e-05,
+      "loss": 0.4268,
+      "step": 605
+    },
+    {
+      "epoch": 1.0341296928327646,
+      "grad_norm": 0.4255108453767918,
+      "learning_rate": 3.862554678242839e-05,
+      "loss": 0.4401,
+      "step": 606
+    },
+    {
+      "epoch": 1.0358361774744027,
+      "grad_norm": 0.4091807950411335,
+      "learning_rate": 3.8616853120613634e-05,
+      "loss": 0.4352,
+      "step": 607
+    },
+    {
+      "epoch": 1.0375426621160408,
+      "grad_norm": 0.4837900126036535,
+      "learning_rate": 3.860813303558093e-05,
+      "loss": 0.4419,
+      "step": 608
+    },
+    {
+      "epoch": 1.0392491467576792,
+      "grad_norm": 0.43491684501985955,
+      "learning_rate": 3.8599386539706866e-05,
+      "loss": 0.4003,
+      "step": 609
+    },
+    {
+      "epoch": 1.0409556313993173,
+      "grad_norm": 0.4400347750814093,
+      "learning_rate": 3.859061364540548e-05,
+      "loss": 0.4547,
+      "step": 610
+    },
+    {
+      "epoch": 1.0426621160409557,
+      "grad_norm": 0.48834670025862204,
+      "learning_rate": 3.858181436512829e-05,
+      "loss": 0.4637,
+      "step": 611
+    },
+    {
+      "epoch": 1.0443686006825939,
+      "grad_norm": 0.5126681768767233,
+      "learning_rate": 3.8572988711364275e-05,
+      "loss": 0.4379,
+      "step": 612
+    },
+    {
+      "epoch": 1.046075085324232,
+      "grad_norm": 0.4626678296479312,
+      "learning_rate": 3.8564136696639826e-05,
+      "loss": 0.4137,
+      "step": 613
+    },
+    {
+      "epoch": 1.0477815699658704,
+      "grad_norm": 0.5528573683773701,
+      "learning_rate": 3.855525833351876e-05,
+      "loss": 0.4286,
+      "step": 614
+    },
+    {
+      "epoch": 1.0494880546075085,
+      "grad_norm": 0.495307267501836,
+      "learning_rate": 3.85463536346023e-05,
+      "loss": 0.4075,
+      "step": 615
+    },
+    {
+      "epoch": 1.0511945392491469,
+      "grad_norm": 0.5650705271490135,
+      "learning_rate": 3.8537422612529025e-05,
+      "loss": 0.4348,
+      "step": 616
+    },
+    {
+      "epoch": 1.052901023890785,
+      "grad_norm": 0.5586265260826336,
+      "learning_rate": 3.85284652799749e-05,
+      "loss": 0.4509,
+      "step": 617
+    },
+    {
+      "epoch": 1.0546075085324231,
+      "grad_norm": 0.5266504778950317,
+      "learning_rate": 3.851948164965321e-05,
+      "loss": 0.4733,
+      "step": 618
+    },
+    {
+      "epoch": 1.0563139931740615,
+      "grad_norm": 0.5712865595016923,
+      "learning_rate": 3.851047173431458e-05,
+      "loss": 0.4183,
+      "step": 619
+    },
+    {
+      "epoch": 1.0580204778156996,
+      "grad_norm": 0.471589194011068,
+      "learning_rate": 3.8501435546746926e-05,
+      "loss": 0.4552,
+      "step": 620
+    },
+    {
+      "epoch": 1.0597269624573378,
+      "grad_norm": 0.46799408444546003,
+      "learning_rate": 3.849237309977548e-05,
+      "loss": 0.3926,
+      "step": 621
+    },
+    {
+      "epoch": 1.0614334470989761,
+      "grad_norm": 0.512683696291688,
+      "learning_rate": 3.848328440626271e-05,
+      "loss": 0.423,
+      "step": 622
+    },
+    {
+      "epoch": 1.0631399317406143,
+      "grad_norm": 0.40009463640214077,
+      "learning_rate": 3.847416947910837e-05,
+      "loss": 0.4254,
+      "step": 623
+    },
+    {
+      "epoch": 1.0648464163822526,
+      "grad_norm": 0.46965072279488435,
+      "learning_rate": 3.846502833124943e-05,
+      "loss": 0.5496,
+      "step": 624
+    },
+    {
+      "epoch": 1.0665529010238908,
+      "grad_norm": 0.4447820975915501,
+      "learning_rate": 3.8455860975660073e-05,
+      "loss": 0.4308,
+      "step": 625
+    },
+    {
+      "epoch": 1.068259385665529,
+      "grad_norm": 0.482882463484246,
+      "learning_rate": 3.844666742535168e-05,
+      "loss": 0.4017,
+      "step": 626
+    },
+    {
+      "epoch": 1.0699658703071673,
+      "grad_norm": 0.46911565381234427,
+      "learning_rate": 3.843744769337282e-05,
+      "loss": 0.4043,
+      "step": 627
+    },
+    {
+      "epoch": 1.0716723549488054,
+      "grad_norm": 0.4615084891322283,
+      "learning_rate": 3.8428201792809213e-05,
+      "loss": 0.431,
+      "step": 628
+    },
+    {
+      "epoch": 1.0733788395904438,
+      "grad_norm": 0.4654575319131965,
+      "learning_rate": 3.841892973678373e-05,
+      "loss": 0.4445,
+      "step": 629
+    },
+    {
+      "epoch": 1.075085324232082,
+      "grad_norm": 0.4556300734049713,
+      "learning_rate": 3.840963153845635e-05,
+      "loss": 0.4035,
+      "step": 630
+    },
+    {
+      "epoch": 1.07679180887372,
+      "grad_norm": 0.44366473510398396,
+      "learning_rate": 3.840030721102417e-05,
+      "loss": 0.4491,
+      "step": 631
+    },
+    {
+      "epoch": 1.0784982935153584,
+      "grad_norm": 0.482008116003779,
+      "learning_rate": 3.839095676772137e-05,
+      "loss": 0.4003,
+      "step": 632
+    },
+    {
+      "epoch": 1.0802047781569966,
+      "grad_norm": 0.48158786376442975,
+      "learning_rate": 3.838158022181918e-05,
+      "loss": 0.424,
+      "step": 633
+    },
+    {
+      "epoch": 1.0819112627986347,
+      "grad_norm": 0.4672233450530056,
+      "learning_rate": 3.837217758662592e-05,
+      "loss": 0.4267,
+      "step": 634
+    },
+    {
+      "epoch": 1.083617747440273,
+      "grad_norm": 0.624999999598963,
+      "learning_rate": 3.836274887548688e-05,
+      "loss": 0.4857,
+      "step": 635
+    },
+    {
+      "epoch": 1.0853242320819112,
+      "grad_norm": 0.4778073909596797,
+      "learning_rate": 3.83532941017844e-05,
+      "loss": 0.4125,
+      "step": 636
+    },
+    {
+      "epoch": 1.0870307167235496,
+      "grad_norm": 0.5669499222940174,
+      "learning_rate": 3.8343813278937815e-05,
+      "loss": 0.4105,
+      "step": 637
+    },
+    {
+      "epoch": 1.0887372013651877,
+      "grad_norm": 0.5457371005641741,
+      "learning_rate": 3.8334306420403404e-05,
+      "loss": 0.4825,
+      "step": 638
+    },
+    {
+      "epoch": 1.0904436860068258,
+      "grad_norm": 0.3923472466345296,
+      "learning_rate": 3.832477353967442e-05,
+      "loss": 0.4099,
+      "step": 639
+    },
+    {
+      "epoch": 1.0921501706484642,
+      "grad_norm": 0.6081290992214772,
+      "learning_rate": 3.8315214650281045e-05,
+      "loss": 0.4539,
+      "step": 640
+    },
+    {
+      "epoch": 1.0938566552901023,
+      "grad_norm": 0.5100099873776106,
+      "learning_rate": 3.830562976579038e-05,
+      "loss": 0.4124,
+      "step": 641
+    },
+    {
+      "epoch": 1.0955631399317407,
+      "grad_norm": 0.5143473347132098,
+      "learning_rate": 3.82960188998064e-05,
+      "loss": 0.4049,
+      "step": 642
+    },
+    {
+      "epoch": 1.0972696245733788,
+      "grad_norm": 0.5584803377698523,
+      "learning_rate": 3.828638206596998e-05,
+      "loss": 0.4204,
+      "step": 643
+    },
+    {
+      "epoch": 1.098976109215017,
+      "grad_norm": 0.5338983168276245,
+      "learning_rate": 3.8276719277958847e-05,
+      "loss": 0.4217,
+      "step": 644
+    },
+    {
+      "epoch": 1.1006825938566553,
+      "grad_norm": 0.4978895498293814,
+      "learning_rate": 3.8267030549487546e-05,
+      "loss": 0.4365,
+      "step": 645
+    },
+    {
+      "epoch": 1.1023890784982935,
+      "grad_norm": 0.5306887611710531,
+      "learning_rate": 3.8257315894307474e-05,
+      "loss": 0.4129,
+      "step": 646
+    },
+    {
+      "epoch": 1.1040955631399316,
+      "grad_norm": 0.5311987023404002,
+      "learning_rate": 3.8247575326206795e-05,
+      "loss": 0.4556,
+      "step": 647
+    },
+    {
+      "epoch": 1.10580204778157,
+      "grad_norm": 0.42657444240974196,
+      "learning_rate": 3.823780885901047e-05,
+      "loss": 0.5134,
+      "step": 648
+    },
+    {
+      "epoch": 1.1075085324232081,
+      "grad_norm": 0.5998945716495959,
+      "learning_rate": 3.8228016506580215e-05,
+      "loss": 0.4621,
+      "step": 649
+    },
+    {
+      "epoch": 1.1092150170648465,
+      "grad_norm": 0.5196911656376497,
+      "learning_rate": 3.821819828281447e-05,
+      "loss": 0.4487,
+      "step": 650
+    },
+    {
+      "epoch": 1.1109215017064846,
+      "grad_norm": 0.5323892395331172,
+      "learning_rate": 3.820835420164842e-05,
+      "loss": 0.4337,
+      "step": 651
+    },
+    {
+      "epoch": 1.1126279863481228,
+      "grad_norm": 0.49936369160051053,
+      "learning_rate": 3.819848427705393e-05,
+      "loss": 0.4733,
+      "step": 652
+    },
+    {
+      "epoch": 1.1143344709897611,
+      "grad_norm": 0.43757305047961276,
+      "learning_rate": 3.8188588523039575e-05,
+      "loss": 0.4172,
+      "step": 653
+    },
+    {
+      "epoch": 1.1160409556313993,
+      "grad_norm": 0.5284302122594836,
+      "learning_rate": 3.817866695365053e-05,
+      "loss": 0.4687,
+      "step": 654
+    },
+    {
+      "epoch": 1.1177474402730376,
+      "grad_norm": 0.5129813630203046,
+      "learning_rate": 3.8168719582968676e-05,
+      "loss": 0.4799,
+      "step": 655
+    },
+    {
+      "epoch": 1.1194539249146758,
+      "grad_norm": 0.5120020068358729,
+      "learning_rate": 3.8158746425112484e-05,
+      "loss": 0.4151,
+      "step": 656
+    },
+    {
+      "epoch": 1.121160409556314,
+      "grad_norm": 0.526318315169048,
+      "learning_rate": 3.814874749423701e-05,
+      "loss": 0.4262,
+      "step": 657
+    },
+    {
+      "epoch": 1.1228668941979523,
+      "grad_norm": 0.4801676547137466,
+      "learning_rate": 3.8138722804533924e-05,
+      "loss": 0.4371,
+      "step": 658
+    },
+    {
+      "epoch": 1.1245733788395904,
+      "grad_norm": 0.5183019208351913,
+      "learning_rate": 3.8128672370231437e-05,
+      "loss": 0.4571,
+      "step": 659
+    },
+    {
+      "epoch": 1.1262798634812285,
+      "grad_norm": 0.5112906104349612,
+      "learning_rate": 3.811859620559429e-05,
+      "loss": 0.4344,
+      "step": 660
+    },
+    {
+      "epoch": 1.127986348122867,
+      "grad_norm": 0.4419751955606906,
+      "learning_rate": 3.8108494324923776e-05,
+      "loss": 0.4553,
+      "step": 661
+    },
+    {
+      "epoch": 1.129692832764505,
+      "grad_norm": 0.46052945892140157,
+      "learning_rate": 3.8098366742557655e-05,
+      "loss": 0.4115,
+      "step": 662
+    },
+    {
+      "epoch": 1.1313993174061434,
+      "grad_norm": 0.41139701881709145,
+      "learning_rate": 3.8088213472870184e-05,
+      "loss": 0.4382,
+      "step": 663
+    },
+    {
+      "epoch": 1.1331058020477816,
+      "grad_norm": 0.42760230770614577,
+      "learning_rate": 3.8078034530272064e-05,
+      "loss": 0.4446,
+      "step": 664
+    },
+    {
+      "epoch": 1.13481228668942,
+      "grad_norm": 0.42098905402498127,
+      "learning_rate": 3.806782992921044e-05,
+      "loss": 0.4481,
+      "step": 665
+    },
+    {
+      "epoch": 1.136518771331058,
+      "grad_norm": 0.4730154695502598,
+      "learning_rate": 3.8057599684168885e-05,
+      "loss": 0.4374,
+      "step": 666
+    },
+    {
+      "epoch": 1.1382252559726962,
+      "grad_norm": 0.40018138361332156,
+      "learning_rate": 3.8047343809667364e-05,
+      "loss": 0.4039,
+      "step": 667
+    },
+    {
+      "epoch": 1.1399317406143346,
+      "grad_norm": 0.4506043190981934,
+      "learning_rate": 3.803706232026221e-05,
+      "loss": 0.4259,
+      "step": 668
+    },
+    {
+      "epoch": 1.1416382252559727,
+      "grad_norm": 0.4233013092094586,
+      "learning_rate": 3.802675523054611e-05,
+      "loss": 0.4213,
+      "step": 669
+    },
+    {
+      "epoch": 1.1433447098976108,
+      "grad_norm": 0.5067727131656333,
+      "learning_rate": 3.8016422555148095e-05,
+      "loss": 0.4307,
+      "step": 670
+    },
+    {
+      "epoch": 1.1450511945392492,
+      "grad_norm": 0.443308195447026,
+      "learning_rate": 3.8006064308733525e-05,
+      "loss": 0.4743,
+      "step": 671
+    },
+    {
+      "epoch": 1.1467576791808873,
+      "grad_norm": 0.4638145022775026,
+      "learning_rate": 3.7995680506004016e-05,
+      "loss": 0.4194,
+      "step": 672
+    },
+    {
+      "epoch": 1.1484641638225255,
+      "grad_norm": 0.43737972177867845,
+      "learning_rate": 3.7985271161697476e-05,
+      "loss": 0.4589,
+      "step": 673
+    },
+    {
+      "epoch": 1.1501706484641638,
+      "grad_norm": 0.38056511468491,
+      "learning_rate": 3.797483629058809e-05,
+      "loss": 0.421,
+      "step": 674
+    },
+    {
+      "epoch": 1.151877133105802,
+      "grad_norm": 0.42431704532297027,
+      "learning_rate": 3.796437590748622e-05,
+      "loss": 0.4154,
+      "step": 675
+    },
+    {
+      "epoch": 1.1535836177474403,
+      "grad_norm": 0.41259449298798845,
+      "learning_rate": 3.795389002723848e-05,
+      "loss": 0.4534,
+      "step": 676
+    },
+    {
+      "epoch": 1.1552901023890785,
+      "grad_norm": 0.43351259341884885,
+      "learning_rate": 3.7943378664727665e-05,
+      "loss": 0.3991,
+      "step": 677
+    },
+    {
+      "epoch": 1.1569965870307168,
+      "grad_norm": 0.4144554120400783,
+      "learning_rate": 3.7932841834872714e-05,
+      "loss": 0.457,
+      "step": 678
+    },
+    {
+      "epoch": 1.158703071672355,
+      "grad_norm": 0.4493500592051183,
+      "learning_rate": 3.792227955262875e-05,
+      "loss": 0.4329,
+      "step": 679
+    },
+    {
+      "epoch": 1.1604095563139931,
+      "grad_norm": 0.46932545805628595,
+      "learning_rate": 3.7911691832986986e-05,
+      "loss": 0.4137,
+      "step": 680
+    },
+    {
+      "epoch": 1.1621160409556315,
+      "grad_norm": 0.4339728498300188,
+      "learning_rate": 3.790107869097475e-05,
+      "loss": 0.4419,
+      "step": 681
+    },
+    {
+      "epoch": 1.1638225255972696,
+      "grad_norm": 0.6040022644435886,
+      "learning_rate": 3.789044014165548e-05,
+      "loss": 0.4257,
+      "step": 682
+    },
+    {
+      "epoch": 1.1655290102389078,
+      "grad_norm": 0.47726214656604243,
+      "learning_rate": 3.787977620012863e-05,
+      "loss": 0.4569,
+      "step": 683
+    },
+    {
+      "epoch": 1.1672354948805461,
+      "grad_norm": 0.3996425683267564,
+      "learning_rate": 3.786908688152971e-05,
+      "loss": 0.419,
+      "step": 684
+    },
+    {
+      "epoch": 1.1689419795221843,
+      "grad_norm": 0.5629467403716398,
+      "learning_rate": 3.785837220103027e-05,
+      "loss": 0.4321,
+      "step": 685
+    },
+    {
+      "epoch": 1.1706484641638226,
+      "grad_norm": 0.3980783427094222,
+      "learning_rate": 3.784763217383783e-05,
+      "loss": 0.4458,
+      "step": 686
+    },
+    {
+      "epoch": 1.1723549488054608,
+      "grad_norm": 0.4265806939898511,
+      "learning_rate": 3.7836866815195896e-05,
+      "loss": 0.4646,
+      "step": 687
+    },
+    {
+      "epoch": 1.174061433447099,
+      "grad_norm": 0.4434867718245713,
+      "learning_rate": 3.782607614038393e-05,
+      "loss": 0.464,
+      "step": 688
+    },
+    {
+      "epoch": 1.1757679180887373,
+      "grad_norm": 0.44452649741611283,
+      "learning_rate": 3.7815260164717314e-05,
+      "loss": 0.4209,
+      "step": 689
+    },
+    {
+      "epoch": 1.1774744027303754,
+      "grad_norm": 0.41431565381119073,
+      "learning_rate": 3.780441890354735e-05,
+      "loss": 0.4593,
+      "step": 690
+    },
+    {
+      "epoch": 1.1791808873720138,
+      "grad_norm": 0.41728408825524926,
+      "learning_rate": 3.779355237226122e-05,
+      "loss": 0.4554,
+      "step": 691
+    },
+    {
+      "epoch": 1.180887372013652,
+      "grad_norm": 0.4321526604917808,
+      "learning_rate": 3.778266058628199e-05,
+      "loss": 0.4583,
+      "step": 692
+    },
+    {
+      "epoch": 1.18259385665529,
+      "grad_norm": 0.37239568412353885,
+      "learning_rate": 3.7771743561068546e-05,
+      "loss": 0.4296,
+      "step": 693
+    },
+    {
+      "epoch": 1.1843003412969284,
+      "grad_norm": 0.4197587486467048,
+      "learning_rate": 3.776080131211561e-05,
+      "loss": 0.4455,
+      "step": 694
+    },
+    {
+      "epoch": 1.1860068259385665,
+      "grad_norm": 0.3726077401897924,
+      "learning_rate": 3.7749833854953714e-05,
+      "loss": 0.4437,
+      "step": 695
+    },
+    {
+      "epoch": 1.1877133105802047,
+      "grad_norm": 0.5031733990320283,
+      "learning_rate": 3.773884120514915e-05,
+      "loss": 0.4137,
+      "step": 696
+    },
+    {
+      "epoch": 1.189419795221843,
+      "grad_norm": 0.4265025413297216,
+      "learning_rate": 3.7727823378303974e-05,
+      "loss": 0.4479,
+      "step": 697
+    },
+    {
+      "epoch": 1.1911262798634812,
+      "grad_norm": 0.4569892386107676,
+      "learning_rate": 3.771678039005597e-05,
+      "loss": 0.4362,
+      "step": 698
+    },
+    {
+      "epoch": 1.1928327645051195,
+      "grad_norm": 0.424109578056102,
+      "learning_rate": 3.770571225607865e-05,
+      "loss": 0.4381,
+      "step": 699
+    },
+    {
+      "epoch": 1.1945392491467577,
+      "grad_norm": 0.42906844196264826,
+      "learning_rate": 3.76946189920812e-05,
+      "loss": 0.3963,
+      "step": 700
+    },
+    {
+      "epoch": 1.1962457337883958,
+      "grad_norm": 0.4890031891369991,
+      "learning_rate": 3.768350061380848e-05,
+      "loss": 0.4269,
+      "step": 701
+    },
+    {
+      "epoch": 1.1979522184300342,
+      "grad_norm": 0.4350188302553573,
+      "learning_rate": 3.7672357137041e-05,
+      "loss": 0.4631,
+      "step": 702
+    },
+    {
+      "epoch": 1.1996587030716723,
+      "grad_norm": 0.4852383658977233,
+      "learning_rate": 3.7661188577594875e-05,
+      "loss": 0.4529,
+      "step": 703
+    },
+    {
+      "epoch": 1.2013651877133107,
+      "grad_norm": 0.4697744180206133,
+      "learning_rate": 3.764999495132185e-05,
+      "loss": 0.4469,
+      "step": 704
+    },
+    {
+      "epoch": 1.2030716723549488,
+      "grad_norm": 0.4445829932767305,
+      "learning_rate": 3.763877627410921e-05,
+      "loss": 0.4167,
+      "step": 705
+    },
+    {
+      "epoch": 1.204778156996587,
+      "grad_norm": 0.4676689122099299,
+      "learning_rate": 3.7627532561879833e-05,
+      "loss": 0.4339,
+      "step": 706
+    },
+    {
+      "epoch": 1.2064846416382253,
+      "grad_norm": 0.47678804282414183,
+      "learning_rate": 3.761626383059209e-05,
+      "loss": 0.4514,
+      "step": 707
+    },
+    {
+      "epoch": 1.2081911262798635,
+      "grad_norm": 0.4433655984117378,
+      "learning_rate": 3.760497009623991e-05,
+      "loss": 0.4444,
+      "step": 708
+    },
+    {
+      "epoch": 1.2098976109215016,
+      "grad_norm": 0.41479676956479195,
+      "learning_rate": 3.759365137485267e-05,
+      "loss": 0.4388,
+      "step": 709
+    },
+    {
+      "epoch": 1.21160409556314,
+      "grad_norm": 0.5205221037848315,
+      "learning_rate": 3.7582307682495225e-05,
+      "loss": 0.4375,
+      "step": 710
+    },
+    {
+      "epoch": 1.213310580204778,
+      "grad_norm": 0.40919927859417016,
+      "learning_rate": 3.757093903526788e-05,
+      "loss": 0.4171,
+      "step": 711
+    },
+    {
+      "epoch": 1.2150170648464165,
+      "grad_norm": 0.5071053572893132,
+      "learning_rate": 3.755954544930633e-05,
+      "loss": 0.4026,
+      "step": 712
+    },
+    {
+      "epoch": 1.2167235494880546,
+      "grad_norm": 0.4573968991390098,
+      "learning_rate": 3.754812694078171e-05,
+      "loss": 0.455,
+      "step": 713
+    },
+    {
+      "epoch": 1.2184300341296928,
+      "grad_norm": 0.39043408323060524,
+      "learning_rate": 3.753668352590049e-05,
+      "loss": 0.4048,
+      "step": 714
+    },
+    {
+      "epoch": 1.2201365187713311,
+      "grad_norm": 0.4590394037207356,
+      "learning_rate": 3.752521522090451e-05,
+      "loss": 0.4259,
+      "step": 715
+    },
+    {
+      "epoch": 1.2218430034129693,
+      "grad_norm": 0.42435704513968003,
+      "learning_rate": 3.751372204207093e-05,
+      "loss": 0.4332,
+      "step": 716
+    },
+    {
+      "epoch": 1.2235494880546076,
+      "grad_norm": 0.40530298060100095,
+      "learning_rate": 3.750220400571221e-05,
+      "loss": 0.4482,
+      "step": 717
+    },
+    {
+      "epoch": 1.2252559726962458,
+      "grad_norm": 0.4190752701415436,
+      "learning_rate": 3.7490661128176105e-05,
+      "loss": 0.4168,
+      "step": 718
+    },
+    {
+      "epoch": 1.226962457337884,
+      "grad_norm": 0.40080779836791597,
+      "learning_rate": 3.747909342584561e-05,
+      "loss": 0.5669,
+      "step": 719
+    },
+    {
+      "epoch": 1.2286689419795223,
+      "grad_norm": 0.3987684782960706,
+      "learning_rate": 3.746750091513897e-05,
+      "loss": 0.4324,
+      "step": 720
+    },
+    {
+      "epoch": 1.2303754266211604,
+      "grad_norm": 0.40562391090602207,
+      "learning_rate": 3.745588361250963e-05,
+      "loss": 0.4643,
+      "step": 721
+    },
+    {
+      "epoch": 1.2320819112627985,
+      "grad_norm": 0.4474485919797283,
+      "learning_rate": 3.744424153444623e-05,
+      "loss": 0.4573,
+      "step": 722
+    },
+    {
+      "epoch": 1.233788395904437,
+      "grad_norm": 0.4004606169913817,
+      "learning_rate": 3.7432574697472564e-05,
+      "loss": 0.4734,
+      "step": 723
+    },
+    {
+      "epoch": 1.235494880546075,
+      "grad_norm": 0.41481424422596996,
+      "learning_rate": 3.742088311814758e-05,
+      "loss": 0.4689,
+      "step": 724
+    },
+    {
+      "epoch": 1.2372013651877134,
+      "grad_norm": 0.4613744985595309,
+      "learning_rate": 3.740916681306533e-05,
+      "loss": 0.421,
+      "step": 725
+    },
+    {
+      "epoch": 1.2389078498293515,
+      "grad_norm": 0.4349483795031661,
+      "learning_rate": 3.7397425798854964e-05,
+      "loss": 0.4414,
+      "step": 726
+    },
+    {
+      "epoch": 1.2406143344709897,
+      "grad_norm": 0.4444615372530873,
+      "learning_rate": 3.738566009218071e-05,
+      "loss": 0.4778,
+      "step": 727
+    },
+    {
+      "epoch": 1.242320819112628,
+      "grad_norm": 0.5550194560618841,
+      "learning_rate": 3.737386970974185e-05,
+      "loss": 0.4493,
+      "step": 728
+    },
+    {
+      "epoch": 1.2440273037542662,
+      "grad_norm": 0.4731437171419562,
+      "learning_rate": 3.736205466827265e-05,
+      "loss": 0.4664,
+      "step": 729
+    },
+    {
+      "epoch": 1.2457337883959045,
+      "grad_norm": 0.47858053424822616,
+      "learning_rate": 3.7350214984542416e-05,
+      "loss": 0.4302,
+      "step": 730
+    },
+    {
+      "epoch": 1.2474402730375427,
+      "grad_norm": 0.5045909061179693,
+      "learning_rate": 3.73383506753554e-05,
+      "loss": 0.4638,
+      "step": 731
+    },
+    {
+      "epoch": 1.2491467576791808,
+      "grad_norm": 0.42161732820421655,
+      "learning_rate": 3.732646175755084e-05,
+      "loss": 0.4389,
+      "step": 732
+    },
+    {
+      "epoch": 1.2508532423208192,
+      "grad_norm": 0.5085164211620015,
+      "learning_rate": 3.731454824800286e-05,
+      "loss": 0.4667,
+      "step": 733
+    },
+    {
+      "epoch": 1.2525597269624573,
+      "grad_norm": 0.5141788654510248,
+      "learning_rate": 3.730261016362052e-05,
+      "loss": 0.4422,
+      "step": 734
+    },
+    {
+      "epoch": 1.2542662116040955,
+      "grad_norm": 0.44945335088525795,
+      "learning_rate": 3.729064752134774e-05,
+      "loss": 0.4203,
+      "step": 735
+    },
+    {
+      "epoch": 1.2559726962457338,
+      "grad_norm": 0.46033952481751483,
+      "learning_rate": 3.727866033816331e-05,
+      "loss": 0.4015,
+      "step": 736
+    },
+    {
+      "epoch": 1.257679180887372,
+      "grad_norm": 0.4558626734368758,
+      "learning_rate": 3.726664863108084e-05,
+      "loss": 0.3963,
+      "step": 737
+    },
+    {
+      "epoch": 1.25938566552901,
+      "grad_norm": 0.40734766802181244,
+      "learning_rate": 3.7254612417148744e-05,
+      "loss": 0.4098,
+      "step": 738
+    },
+    {
+      "epoch": 1.2610921501706485,
+      "grad_norm": 0.5049229183639216,
+      "learning_rate": 3.724255171345024e-05,
+      "loss": 0.4601,
+      "step": 739
+    },
+    {
+      "epoch": 1.2627986348122868,
+      "grad_norm": 0.3723550240432109,
+      "learning_rate": 3.723046653710329e-05,
+      "loss": 0.435,
+      "step": 740
+    },
+    {
+      "epoch": 1.264505119453925,
+      "grad_norm": 0.4834866392138083,
+      "learning_rate": 3.7218356905260576e-05,
+      "loss": 0.4851,
+      "step": 741
+    },
+    {
+      "epoch": 1.266211604095563,
+      "grad_norm": 0.47059437880819743,
+      "learning_rate": 3.7206222835109525e-05,
+      "loss": 0.4371,
+      "step": 742
+    },
+    {
+      "epoch": 1.2679180887372015,
+      "grad_norm": 0.4488476614201091,
+      "learning_rate": 3.719406434387221e-05,
+      "loss": 0.4293,
+      "step": 743
+    },
+    {
+      "epoch": 1.2696245733788396,
+      "grad_norm": 0.4769670725825094,
+      "learning_rate": 3.7181881448805407e-05,
+      "loss": 0.4587,
+      "step": 744
+    },
+    {
+      "epoch": 1.2713310580204777,
+      "grad_norm": 0.4558559954017582,
+      "learning_rate": 3.716967416720049e-05,
+      "loss": 0.4418,
+      "step": 745
+    },
+    {
+      "epoch": 1.273037542662116,
+      "grad_norm": 0.45087258818774806,
+      "learning_rate": 3.715744251638347e-05,
+      "loss": 0.4146,
+      "step": 746
+    },
+    {
+      "epoch": 1.2747440273037542,
+      "grad_norm": 0.4534888156553091,
+      "learning_rate": 3.714518651371494e-05,
+      "loss": 0.4686,
+      "step": 747
+    },
+    {
+      "epoch": 1.2764505119453924,
+      "grad_norm": 0.462002171339421,
+      "learning_rate": 3.713290617659005e-05,
+      "loss": 0.4693,
+      "step": 748
+    },
+    {
+      "epoch": 1.2781569965870307,
+      "grad_norm": 0.5096654338817803,
+      "learning_rate": 3.712060152243849e-05,
+      "loss": 0.4496,
+      "step": 749
+    },
+    {
+      "epoch": 1.2798634812286689,
+      "grad_norm": 0.4369904424706895,
+      "learning_rate": 3.710827256872447e-05,
+      "loss": 0.4306,
+      "step": 750
+    },
+    {
+      "epoch": 1.2815699658703072,
+      "grad_norm": 0.4775228575976149,
+      "learning_rate": 3.7095919332946693e-05,
+      "loss": 0.4427,
+      "step": 751
+    },
+    {
+      "epoch": 1.2832764505119454,
+      "grad_norm": 0.4394095447413763,
+      "learning_rate": 3.7083541832638304e-05,
+      "loss": 0.4592,
+      "step": 752
+    },
+    {
+      "epoch": 1.2849829351535837,
+      "grad_norm": 0.4461061424192823,
+      "learning_rate": 3.70711400853669e-05,
+      "loss": 0.4175,
+      "step": 753
+    },
+    {
+      "epoch": 1.286689419795222,
+      "grad_norm": 0.4143666448961568,
+      "learning_rate": 3.7058714108734503e-05,
+      "loss": 0.4174,
+      "step": 754
+    },
+    {
+      "epoch": 1.28839590443686,
+      "grad_norm": 0.3840367032155938,
+      "learning_rate": 3.704626392037751e-05,
+      "loss": 0.4461,
+      "step": 755
+    },
+    {
+      "epoch": 1.2901023890784984,
+      "grad_norm": 0.46008333986661104,
+      "learning_rate": 3.703378953796669e-05,
+      "loss": 0.4372,
+      "step": 756
+    },
+    {
+      "epoch": 1.2918088737201365,
+      "grad_norm": 0.4913797925551799,
+      "learning_rate": 3.702129097920715e-05,
+      "loss": 0.4848,
+      "step": 757
+    },
+    {
+      "epoch": 1.2935153583617747,
+      "grad_norm": 0.4137229879559694,
+      "learning_rate": 3.700876826183829e-05,
+      "loss": 0.4168,
+      "step": 758
+    },
+    {
+      "epoch": 1.295221843003413,
+      "grad_norm": 0.44606644673173096,
+      "learning_rate": 3.699622140363383e-05,
+      "loss": 0.4635,
+      "step": 759
+    },
+    {
+      "epoch": 1.2969283276450512,
+      "grad_norm": 0.4237165781520206,
+      "learning_rate": 3.6983650422401744e-05,
+      "loss": 0.4529,
+      "step": 760
+    },
+    {
+      "epoch": 1.2986348122866893,
+      "grad_norm": 0.42451416478377646,
+      "learning_rate": 3.697105533598423e-05,
+      "loss": 0.4914,
+      "step": 761
+    },
+    {
+      "epoch": 1.3003412969283277,
+      "grad_norm": 0.3878869863717339,
+      "learning_rate": 3.695843616225772e-05,
+      "loss": 0.4716,
+      "step": 762
+    },
+    {
+      "epoch": 1.3020477815699658,
+      "grad_norm": 0.4134415686037431,
+      "learning_rate": 3.694579291913282e-05,
+      "loss": 0.4601,
+      "step": 763
+    },
+    {
+      "epoch": 1.3037542662116042,
+      "grad_norm": 0.8003369338048761,
+      "learning_rate": 3.693312562455429e-05,
+      "loss": 0.4911,
+      "step": 764
+    },
+    {
+      "epoch": 1.3054607508532423,
+      "grad_norm": 0.37537128276066495,
+      "learning_rate": 3.692043429650105e-05,
+      "loss": 0.4469,
+      "step": 765
+    },
+    {
+      "epoch": 1.3071672354948807,
+      "grad_norm": 0.37529715858826523,
+      "learning_rate": 3.690771895298612e-05,
+      "loss": 0.4242,
+      "step": 766
+    },
+    {
+      "epoch": 1.3088737201365188,
+      "grad_norm": 0.38984349190821227,
+      "learning_rate": 3.6894979612056596e-05,
+      "loss": 0.4187,
+      "step": 767
+    },
+    {
+      "epoch": 1.310580204778157,
+      "grad_norm": 0.4055181076986435,
+      "learning_rate": 3.688221629179365e-05,
+      "loss": 0.4613,
+      "step": 768
+    },
+    {
+      "epoch": 1.3122866894197953,
+      "grad_norm": 0.4029344250787094,
+      "learning_rate": 3.686942901031247e-05,
+      "loss": 0.4316,
+      "step": 769
+    },
+    {
+      "epoch": 1.3139931740614335,
+      "grad_norm": 0.4079829615017142,
+      "learning_rate": 3.6856617785762286e-05,
+      "loss": 0.4329,
+      "step": 770
+    },
+    {
+      "epoch": 1.3156996587030716,
+      "grad_norm": 0.4069157151626814,
+      "learning_rate": 3.6843782636326256e-05,
+      "loss": 0.3958,
+      "step": 771
+    },
+    {
+      "epoch": 1.31740614334471,
+      "grad_norm": 0.47706876658562686,
+      "learning_rate": 3.6830923580221556e-05,
+      "loss": 0.4329,
+      "step": 772
+    },
+    {
+      "epoch": 1.319112627986348,
+      "grad_norm": 1.2357596289802686,
+      "learning_rate": 3.6818040635699245e-05,
+      "loss": 0.4568,
+      "step": 773
+    },
+    {
+      "epoch": 1.3208191126279862,
+      "grad_norm": 0.43601104388944756,
+      "learning_rate": 3.680513382104432e-05,
+      "loss": 0.4458,
+      "step": 774
+    },
+    {
+      "epoch": 1.3225255972696246,
+      "grad_norm": 0.39840852817000055,
+      "learning_rate": 3.679220315457563e-05,
+      "loss": 0.449,
+      "step": 775
+    },
+    {
+      "epoch": 1.3242320819112627,
+      "grad_norm": 0.44289749853193855,
+      "learning_rate": 3.67792486546459e-05,
+      "loss": 0.4745,
+      "step": 776
+    },
+    {
+      "epoch": 1.325938566552901,
+      "grad_norm": 0.38501530038108883,
+      "learning_rate": 3.676627033964167e-05,
+      "loss": 0.4284,
+      "step": 777
+    },
+    {
+      "epoch": 1.3276450511945392,
+      "grad_norm": 0.4436655072824138,
+      "learning_rate": 3.675326822798329e-05,
+      "loss": 0.4182,
+      "step": 778
+    },
+    {
+      "epoch": 1.3293515358361776,
+      "grad_norm": 0.3984948609435545,
+      "learning_rate": 3.674024233812487e-05,
+      "loss": 0.4206,
+      "step": 779
+    },
+    {
+      "epoch": 1.3310580204778157,
+      "grad_norm": 0.4387932025256469,
+      "learning_rate": 3.672719268855429e-05,
+      "loss": 0.4298,
+      "step": 780
+    },
+    {
+      "epoch": 1.3327645051194539,
+      "grad_norm": 0.4090726444133126,
+      "learning_rate": 3.671411929779313e-05,
+      "loss": 0.4216,
+      "step": 781
+    },
+    {
+      "epoch": 1.3344709897610922,
+      "grad_norm": 0.5003333111778476,
+      "learning_rate": 3.670102218439669e-05,
+      "loss": 0.4496,
+      "step": 782
+    },
+    {
+      "epoch": 1.3361774744027304,
+      "grad_norm": 0.3999097982807462,
+      "learning_rate": 3.66879013669539e-05,
+      "loss": 0.4106,
+      "step": 783
+    },
+    {
+      "epoch": 1.3378839590443685,
+      "grad_norm": 0.39528289174297737,
+      "learning_rate": 3.667475686408739e-05,
+      "loss": 0.4192,
+      "step": 784
+    },
+    {
+      "epoch": 1.3395904436860069,
+      "grad_norm": 0.3530174875714856,
+      "learning_rate": 3.666158869445336e-05,
+      "loss": 0.4202,
+      "step": 785
+    },
+    {
+      "epoch": 1.341296928327645,
+      "grad_norm": 0.42240313951429387,
+      "learning_rate": 3.664839687674163e-05,
+      "loss": 0.461,
+      "step": 786
+    },
+    {
+      "epoch": 1.3430034129692832,
+      "grad_norm": 0.37962678593801646,
+      "learning_rate": 3.663518142967557e-05,
+      "loss": 0.4435,
+      "step": 787
+    },
+    {
+      "epoch": 1.3447098976109215,
+      "grad_norm": 0.4508809400632154,
+      "learning_rate": 3.662194237201208e-05,
+      "loss": 0.4367,
+      "step": 788
+    },
+    {
+      "epoch": 1.3464163822525597,
+      "grad_norm": 0.517558585482028,
+      "learning_rate": 3.660867972254159e-05,
+      "loss": 0.4544,
+      "step": 789
+    },
+    {
+      "epoch": 1.348122866894198,
+      "grad_norm": 0.42198247927274773,
+      "learning_rate": 3.6595393500088e-05,
+      "loss": 0.4426,
+      "step": 790
+    },
+    {
+      "epoch": 1.3498293515358362,
+      "grad_norm": 0.38044452785973637,
+      "learning_rate": 3.658208372350868e-05,
+      "loss": 0.4183,
+      "step": 791
+    },
+    {
+      "epoch": 1.3515358361774745,
+      "grad_norm": 0.38931030207281575,
+      "learning_rate": 3.656875041169442e-05,
+      "loss": 0.419,
+      "step": 792
+    },
+    {
+      "epoch": 1.3532423208191127,
+      "grad_norm": 0.3871251663984767,
+      "learning_rate": 3.655539358356941e-05,
+      "loss": 0.4161,
+      "step": 793
+    },
+    {
+      "epoch": 1.3549488054607508,
+      "grad_norm": 0.41111760528386887,
+      "learning_rate": 3.6542013258091236e-05,
+      "loss": 0.3992,
+      "step": 794
+    },
+    {
+      "epoch": 1.3566552901023892,
+      "grad_norm": 0.39052845005499515,
+      "learning_rate": 3.652860945425082e-05,
+      "loss": 0.4208,
+      "step": 795
+    },
+    {
+      "epoch": 1.3583617747440273,
+      "grad_norm": 0.38865687702951424,
+      "learning_rate": 3.65151821910724e-05,
+      "loss": 0.4049,
+      "step": 796
+    },
+    {
+      "epoch": 1.3600682593856654,
+      "grad_norm": 0.40903580121063793,
+      "learning_rate": 3.650173148761353e-05,
+      "loss": 0.4149,
+      "step": 797
+    },
+    {
+      "epoch": 1.3617747440273038,
+      "grad_norm": 0.3868920258274055,
+      "learning_rate": 3.6488257362965026e-05,
+      "loss": 0.4377,
+      "step": 798
+    },
+    {
+      "epoch": 1.363481228668942,
+      "grad_norm": 0.4033212911143925,
+      "learning_rate": 3.6474759836250936e-05,
+      "loss": 0.4143,
+      "step": 799
+    },
+    {
+      "epoch": 1.36518771331058,
+      "grad_norm": 0.45515120908056866,
+      "learning_rate": 3.646123892662854e-05,
+      "loss": 0.4009,
+      "step": 800
+    },
+    {
+      "epoch": 1.3668941979522184,
+      "grad_norm": 0.4308414454379085,
+      "learning_rate": 3.644769465328828e-05,
+      "loss": 0.4518,
+      "step": 801
+    },
+    {
+      "epoch": 1.3686006825938566,
+      "grad_norm": 0.4619818515515126,
+      "learning_rate": 3.643412703545378e-05,
+      "loss": 0.4365,
+      "step": 802
+    },
+    {
+      "epoch": 1.370307167235495,
+      "grad_norm": 0.38985140685038283,
+      "learning_rate": 3.642053609238181e-05,
+      "loss": 0.44,
+      "step": 803
+    },
+    {
+      "epoch": 1.372013651877133,
+      "grad_norm": 0.4499721307861487,
+      "learning_rate": 3.640692184336221e-05,
+      "loss": 0.4532,
+      "step": 804
+    },
+    {
+      "epoch": 1.3737201365187715,
+      "grad_norm": 0.4228258907033133,
+      "learning_rate": 3.639328430771792e-05,
+      "loss": 0.4286,
+      "step": 805
+    },
+    {
+      "epoch": 1.3754266211604096,
+      "grad_norm": 0.43820754146061625,
+      "learning_rate": 3.637962350480492e-05,
+      "loss": 0.4308,
+      "step": 806
+    },
+    {
+      "epoch": 1.3771331058020477,
+      "grad_norm": 0.4104514343792071,
+      "learning_rate": 3.636593945401224e-05,
+      "loss": 0.4435,
+      "step": 807
+    },
+    {
+      "epoch": 1.378839590443686,
+      "grad_norm": 0.4523514499643776,
+      "learning_rate": 3.6352232174761865e-05,
+      "loss": 0.4238,
+      "step": 808
+    },
+    {
+      "epoch": 1.3805460750853242,
+      "grad_norm": 0.4055406278240278,
+      "learning_rate": 3.633850168650879e-05,
+      "loss": 0.4741,
+      "step": 809
+    },
+    {
+      "epoch": 1.3822525597269624,
+      "grad_norm": 0.43571497578619445,
+      "learning_rate": 3.6324748008740925e-05,
+      "loss": 0.4182,
+      "step": 810
+    },
+    {
+      "epoch": 1.3839590443686007,
+      "grad_norm": 0.38896961713941225,
+      "learning_rate": 3.63109711609791e-05,
+      "loss": 0.4329,
+      "step": 811
+    },
+    {
+      "epoch": 1.3856655290102389,
+      "grad_norm": 0.4011016710632721,
+      "learning_rate": 3.629717116277702e-05,
+      "loss": 0.4561,
+      "step": 812
+    },
+    {
+      "epoch": 1.387372013651877,
+      "grad_norm": 0.3774279212199751,
+      "learning_rate": 3.628334803372127e-05,
+      "loss": 0.4574,
+      "step": 813
+    },
+    {
+      "epoch": 1.3890784982935154,
+      "grad_norm": 0.3854368012109252,
+      "learning_rate": 3.626950179343126e-05,
+      "loss": 0.4419,
+      "step": 814
+    },
+    {
+      "epoch": 1.3907849829351535,
+      "grad_norm": 0.4029903450946548,
+      "learning_rate": 3.6255632461559176e-05,
+      "loss": 0.4679,
+      "step": 815
+    },
+    {
+      "epoch": 1.3924914675767919,
+      "grad_norm": 0.4146348888054304,
+      "learning_rate": 3.624174005779002e-05,
+      "loss": 0.4171,
+      "step": 816
+    },
+    {
+      "epoch": 1.39419795221843,
+      "grad_norm": 0.42837264640372247,
+      "learning_rate": 3.62278246018415e-05,
+      "loss": 0.4318,
+      "step": 817
+    },
+    {
+      "epoch": 1.3959044368600684,
+      "grad_norm": 0.4144169325480012,
+      "learning_rate": 3.621388611346407e-05,
+      "loss": 0.4142,
+      "step": 818
+    },
+    {
+      "epoch": 1.3976109215017065,
+      "grad_norm": 0.42446552773541035,
+      "learning_rate": 3.6199924612440855e-05,
+      "loss": 0.4448,
+      "step": 819
+    },
+    {
+      "epoch": 1.3993174061433447,
+      "grad_norm": 0.3969940864973572,
+      "learning_rate": 3.6185940118587673e-05,
+      "loss": 0.4332,
+      "step": 820
+    },
+    {
+      "epoch": 1.401023890784983,
+      "grad_norm": 0.41046129721730956,
+      "learning_rate": 3.617193265175293e-05,
+      "loss": 0.4448,
+      "step": 821
+    },
+    {
+      "epoch": 1.4027303754266212,
+      "grad_norm": 0.3735310996078358,
+      "learning_rate": 3.615790223181768e-05,
+      "loss": 0.3887,
+      "step": 822
+    },
+    {
+      "epoch": 1.4044368600682593,
+      "grad_norm": 0.4665973159613071,
+      "learning_rate": 3.614384887869553e-05,
+      "loss": 0.4463,
+      "step": 823
+    },
+    {
+      "epoch": 1.4061433447098977,
+      "grad_norm": 0.3937368085046607,
+      "learning_rate": 3.612977261233265e-05,
+      "loss": 0.4331,
+      "step": 824
+    },
+    {
+      "epoch": 1.4078498293515358,
+      "grad_norm": 0.4617674156449415,
+      "learning_rate": 3.611567345270772e-05,
+      "loss": 0.4387,
+      "step": 825
+    },
+    {
+      "epoch": 1.409556313993174,
+      "grad_norm": 0.4203672742229765,
+      "learning_rate": 3.610155141983192e-05,
+      "loss": 0.4434,
+      "step": 826
+    },
+    {
+      "epoch": 1.4112627986348123,
+      "grad_norm": 0.40848430637835254,
+      "learning_rate": 3.608740653374889e-05,
+      "loss": 0.4229,
+      "step": 827
+    },
+    {
+      "epoch": 1.4129692832764504,
+      "grad_norm": 0.47426540651765864,
+      "learning_rate": 3.607323881453472e-05,
+      "loss": 0.4154,
+      "step": 828
+    },
+    {
+      "epoch": 1.4146757679180888,
+      "grad_norm": 0.41797039048076445,
+      "learning_rate": 3.6059048282297887e-05,
+      "loss": 0.4241,
+      "step": 829
+    },
+    {
+      "epoch": 1.416382252559727,
+      "grad_norm": 0.4135512178844431,
+      "learning_rate": 3.604483495717926e-05,
+      "loss": 0.4031,
+      "step": 830
+    },
+    {
+      "epoch": 1.4180887372013653,
+      "grad_norm": 0.41837197750420363,
+      "learning_rate": 3.603059885935205e-05,
+      "loss": 0.4146,
+      "step": 831
+    },
+    {
+      "epoch": 1.4197952218430034,
+      "grad_norm": 0.40027620826967797,
+      "learning_rate": 3.601634000902179e-05,
+      "loss": 0.423,
+      "step": 832
+    },
+    {
+      "epoch": 1.4215017064846416,
+      "grad_norm": 0.4256842466443175,
+      "learning_rate": 3.600205842642632e-05,
+      "loss": 0.4551,
+      "step": 833
+    },
+    {
+      "epoch": 1.42320819112628,
+      "grad_norm": 0.4550854880984585,
+      "learning_rate": 3.598775413183573e-05,
+      "loss": 0.4599,
+      "step": 834
+    },
+    {
+      "epoch": 1.424914675767918,
+      "grad_norm": 0.4815447932080825,
+      "learning_rate": 3.597342714555235e-05,
+      "loss": 0.4036,
+      "step": 835
+    },
+    {
+      "epoch": 1.4266211604095562,
+      "grad_norm": 0.43871354888787345,
+      "learning_rate": 3.595907748791071e-05,
+      "loss": 0.4119,
+      "step": 836
+    },
+    {
+      "epoch": 1.4283276450511946,
+      "grad_norm": 0.46235121968022785,
+      "learning_rate": 3.594470517927755e-05,
+      "loss": 0.4477,
+      "step": 837
+    },
+    {
+      "epoch": 1.4300341296928327,
+      "grad_norm": 0.5716499129913892,
+      "learning_rate": 3.59303102400517e-05,
+      "loss": 0.4315,
+      "step": 838
+    },
+    {
+      "epoch": 1.4317406143344709,
+      "grad_norm": 0.4513547768347061,
+      "learning_rate": 3.591589269066416e-05,
+      "loss": 0.4293,
+      "step": 839
+    },
+    {
+      "epoch": 1.4334470989761092,
+      "grad_norm": 0.5258567217520957,
+      "learning_rate": 3.5901452551578e-05,
+      "loss": 0.4333,
+      "step": 840
+    },
+    {
+      "epoch": 1.4351535836177474,
+      "grad_norm": 0.4446607367154777,
+      "learning_rate": 3.5886989843288364e-05,
+      "loss": 0.4607,
+      "step": 841
+    },
+    {
+      "epoch": 1.4368600682593857,
+      "grad_norm": 0.4557637105087839,
+      "learning_rate": 3.587250458632241e-05,
+      "loss": 0.4472,
+      "step": 842
+    },
+    {
+      "epoch": 1.4385665529010239,
+      "grad_norm": 0.4577223794812873,
+      "learning_rate": 3.585799680123932e-05,
+      "loss": 0.4487,
+      "step": 843
+    },
+    {
+      "epoch": 1.4402730375426622,
+      "grad_norm": 0.42430168888349257,
+      "learning_rate": 3.584346650863024e-05,
+      "loss": 0.4452,
+      "step": 844
+    },
+    {
+      "epoch": 1.4419795221843004,
+      "grad_norm": 0.37398876259118063,
+      "learning_rate": 3.582891372911825e-05,
+      "loss": 0.3806,
+      "step": 845
+    },
+    {
+      "epoch": 1.4436860068259385,
+      "grad_norm": 0.416566462619093,
+      "learning_rate": 3.581433848335838e-05,
+      "loss": 0.4733,
+      "step": 846
+    },
+    {
+      "epoch": 1.4453924914675769,
+      "grad_norm": 0.36234070654329864,
+      "learning_rate": 3.5799740792037515e-05,
+      "loss": 0.4224,
+      "step": 847
+    },
+    {
+      "epoch": 1.447098976109215,
+      "grad_norm": 0.4129583494033684,
+      "learning_rate": 3.578512067587441e-05,
+      "loss": 0.4305,
+      "step": 848
+    },
+    {
+      "epoch": 1.4488054607508531,
+      "grad_norm": 0.40331761774448877,
+      "learning_rate": 3.5770478155619636e-05,
+      "loss": 0.4727,
+      "step": 849
+    },
+    {
+      "epoch": 1.4505119453924915,
+      "grad_norm": 0.5056379260288237,
+      "learning_rate": 3.575581325205558e-05,
+      "loss": 0.4231,
+      "step": 850
+    },
+    {
+      "epoch": 1.4522184300341296,
+      "grad_norm": 0.3654321179751699,
+      "learning_rate": 3.574112598599639e-05,
+      "loss": 0.4269,
+      "step": 851
+    },
+    {
+      "epoch": 1.4539249146757678,
+      "grad_norm": 0.4250546017967828,
+      "learning_rate": 3.5726416378287965e-05,
+      "loss": 0.4627,
+      "step": 852
+    },
+    {
+      "epoch": 1.4556313993174061,
+      "grad_norm": 0.3678010879072778,
+      "learning_rate": 3.571168444980788e-05,
+      "loss": 0.3916,
+      "step": 853
+    },
+    {
+      "epoch": 1.4573378839590443,
+      "grad_norm": 0.39832118428914814,
+      "learning_rate": 3.5696930221465427e-05,
+      "loss": 0.4594,
+      "step": 854
+    },
+    {
+      "epoch": 1.4590443686006827,
+      "grad_norm": 0.35873588194403805,
+      "learning_rate": 3.568215371420153e-05,
+      "loss": 0.4414,
+      "step": 855
+    },
+    {
+      "epoch": 1.4607508532423208,
+      "grad_norm": 0.42738781031288464,
+      "learning_rate": 3.566735494898875e-05,
+      "loss": 0.4259,
+      "step": 856
+    },
+    {
+      "epoch": 1.4624573378839592,
+      "grad_norm": 0.4421005033450358,
+      "learning_rate": 3.565253394683121e-05,
+      "loss": 0.4236,
+      "step": 857
+    },
+    {
+      "epoch": 1.4641638225255973,
+      "grad_norm": 0.4006965058599362,
+      "learning_rate": 3.563769072876463e-05,
+      "loss": 0.4778,
+      "step": 858
+    },
+    {
+      "epoch": 1.4658703071672354,
+      "grad_norm": 0.42058567843765865,
+      "learning_rate": 3.5622825315856223e-05,
+      "loss": 0.4762,
+      "step": 859
+    },
+    {
+      "epoch": 1.4675767918088738,
+      "grad_norm": 0.3584682257159212,
+      "learning_rate": 3.560793772920474e-05,
+      "loss": 0.4568,
+      "step": 860
+    },
+    {
+      "epoch": 1.469283276450512,
+      "grad_norm": 0.40671444966184894,
+      "learning_rate": 3.559302798994038e-05,
+      "loss": 0.5237,
+      "step": 861
+    },
+    {
+      "epoch": 1.47098976109215,
+      "grad_norm": 0.4258033641750018,
+      "learning_rate": 3.557809611922479e-05,
+      "loss": 0.4448,
+      "step": 862
+    },
+    {
+      "epoch": 1.4726962457337884,
+      "grad_norm": 0.49700500782089985,
+      "learning_rate": 3.556314213825103e-05,
+      "loss": 0.4191,
+      "step": 863
+    },
+    {
+      "epoch": 1.4744027303754266,
+      "grad_norm": 0.4343913869278087,
+      "learning_rate": 3.5548166068243554e-05,
+      "loss": 0.4241,
+      "step": 864
+    },
+    {
+      "epoch": 1.4761092150170647,
+      "grad_norm": 0.4111341986973055,
+      "learning_rate": 3.553316793045813e-05,
+      "loss": 0.4802,
+      "step": 865
+    },
+    {
+      "epoch": 1.477815699658703,
+      "grad_norm": 0.4258209373245728,
+      "learning_rate": 3.551814774618189e-05,
+      "loss": 0.4434,
+      "step": 866
+    },
+    {
+      "epoch": 1.4795221843003412,
+      "grad_norm": 0.46306345542355654,
+      "learning_rate": 3.550310553673323e-05,
+      "loss": 0.4619,
+      "step": 867
+    },
+    {
+      "epoch": 1.4812286689419796,
+      "grad_norm": 0.4581807591999424,
+      "learning_rate": 3.548804132346182e-05,
+      "loss": 0.4149,
+      "step": 868
+    },
+    {
+      "epoch": 1.4829351535836177,
+      "grad_norm": 0.3949912398526118,
+      "learning_rate": 3.547295512774855e-05,
+      "loss": 0.4335,
+      "step": 869
+    },
+    {
+      "epoch": 1.484641638225256,
+      "grad_norm": 0.4220703031918266,
+      "learning_rate": 3.545784697100551e-05,
+      "loss": 0.3977,
+      "step": 870
+    },
+    {
+      "epoch": 1.4863481228668942,
+      "grad_norm": 0.46599454689196135,
+      "learning_rate": 3.544271687467599e-05,
+      "loss": 0.4523,
+      "step": 871
+    },
+    {
+      "epoch": 1.4880546075085324,
+      "grad_norm": 0.401862878134278,
+      "learning_rate": 3.542756486023437e-05,
+      "loss": 0.4362,
+      "step": 872
+    },
+    {
+      "epoch": 1.4897610921501707,
+      "grad_norm": 0.4813492793507341,
+      "learning_rate": 3.541239094918617e-05,
+      "loss": 0.4271,
+      "step": 873
+    },
+    {
+      "epoch": 1.4914675767918089,
+      "grad_norm": 0.4182023307175445,
+      "learning_rate": 3.5397195163067985e-05,
+      "loss": 0.4476,
+      "step": 874
+    },
+    {
+      "epoch": 1.493174061433447,
+      "grad_norm": 0.42276483733512216,
+      "learning_rate": 3.5381977523447454e-05,
+      "loss": 0.4163,
+      "step": 875
+    },
+    {
+      "epoch": 1.4948805460750854,
+      "grad_norm": 0.47632383003430656,
+      "learning_rate": 3.536673805192323e-05,
+      "loss": 0.4698,
+      "step": 876
+    },
+    {
+      "epoch": 1.4965870307167235,
+      "grad_norm": 0.36931128067340063,
+      "learning_rate": 3.535147677012495e-05,
+      "loss": 0.4486,
+      "step": 877
+    },
+    {
+      "epoch": 1.4982935153583616,
+      "grad_norm": 0.4604522960110994,
+      "learning_rate": 3.533619369971322e-05,
+      "loss": 0.4119,
+      "step": 878
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3975754189073352,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.4479,
+      "step": 879
+    },
+    {
+      "epoch": 1.5017064846416384,
+      "grad_norm": 0.37769858386666094,
+      "learning_rate": 3.530556227984639e-05,
+      "loss": 0.4059,
+      "step": 880
+    },
+    {
+      "epoch": 1.5034129692832765,
+      "grad_norm": 0.43939814983028674,
+      "learning_rate": 3.5290213973867e-05,
+      "loss": 0.448,
+      "step": 881
+    },
+    {
+      "epoch": 1.5051194539249146,
+      "grad_norm": 0.3868394972875524,
+      "learning_rate": 3.527484396622548e-05,
+      "loss": 0.4237,
+      "step": 882
+    },
+    {
+      "epoch": 1.506825938566553,
+      "grad_norm": 0.3887541284770232,
+      "learning_rate": 3.525945227873676e-05,
+      "loss": 0.4658,
+      "step": 883
+    },
+    {
+      "epoch": 1.5085324232081911,
+      "grad_norm": 0.37052252214316006,
+      "learning_rate": 3.524403893324653e-05,
+      "loss": 0.4039,
+      "step": 884
+    },
+    {
+      "epoch": 1.5102389078498293,
+      "grad_norm": 0.39225305033832947,
+      "learning_rate": 3.52286039516312e-05,
+      "loss": 0.4296,
+      "step": 885
+    },
+    {
+      "epoch": 1.5119453924914676,
+      "grad_norm": 0.4668921452833352,
+      "learning_rate": 3.52131473557979e-05,
+      "loss": 0.4836,
+      "step": 886
+    },
+    {
+      "epoch": 1.5136518771331058,
+      "grad_norm": 0.41657018354420566,
+      "learning_rate": 3.519766916768447e-05,
+      "loss": 0.4585,
+      "step": 887
+    },
+    {
+      "epoch": 1.515358361774744,
+      "grad_norm": 0.4013173259210855,
+      "learning_rate": 3.518216940925934e-05,
+      "loss": 0.4553,
+      "step": 888
+    },
+    {
+      "epoch": 1.5170648464163823,
+      "grad_norm": 0.5024901693039066,
+      "learning_rate": 3.516664810252159e-05,
+      "loss": 0.4347,
+      "step": 889
+    },
+    {
+      "epoch": 1.5187713310580204,
+      "grad_norm": 0.37789225448719616,
+      "learning_rate": 3.5151105269500876e-05,
+      "loss": 0.4156,
+      "step": 890
+    },
+    {
+      "epoch": 1.5204778156996586,
+      "grad_norm": 0.508454307999391,
+      "learning_rate": 3.513554093225741e-05,
+      "loss": 0.4029,
+      "step": 891
+    },
+    {
+      "epoch": 1.522184300341297,
+      "grad_norm": 0.45808430716907134,
+      "learning_rate": 3.511995511288191e-05,
+      "loss": 0.4261,
+      "step": 892
+    },
+    {
+      "epoch": 1.5238907849829353,
+      "grad_norm": 0.48105430594252147,
+      "learning_rate": 3.510434783349562e-05,
+      "loss": 0.4382,
+      "step": 893
+    },
+    {
+      "epoch": 1.5255972696245734,
+      "grad_norm": 0.40020881894592886,
+      "learning_rate": 3.50887191162502e-05,
+      "loss": 0.4129,
+      "step": 894
+    },
+    {
+      "epoch": 1.5273037542662116,
+      "grad_norm": 0.46194774910126624,
+      "learning_rate": 3.507306898332775e-05,
+      "loss": 0.4352,
+      "step": 895
+    },
+    {
+      "epoch": 1.52901023890785,
+      "grad_norm": 0.4170030314335261,
+      "learning_rate": 3.5057397456940786e-05,
+      "loss": 0.417,
+      "step": 896
+    },
+    {
+      "epoch": 1.530716723549488,
+      "grad_norm": 0.5223695395475298,
+      "learning_rate": 3.504170455933216e-05,
+      "loss": 0.4974,
+      "step": 897
+    },
+    {
+      "epoch": 1.5324232081911262,
+      "grad_norm": 0.40890728841524154,
+      "learning_rate": 3.502599031277509e-05,
+      "loss": 0.4783,
+      "step": 898
+    },
+    {
+      "epoch": 1.5341296928327646,
+      "grad_norm": 0.4870710591339736,
+      "learning_rate": 3.501025473957305e-05,
+      "loss": 0.4178,
+      "step": 899
+    },
+    {
+      "epoch": 1.5358361774744027,
+      "grad_norm": 0.42576048970858366,
+      "learning_rate": 3.4994497862059824e-05,
+      "loss": 0.4206,
+      "step": 900
+    },
+    {
+      "epoch": 1.5375426621160408,
+      "grad_norm": 0.38954421729882216,
+      "learning_rate": 3.497871970259942e-05,
+      "loss": 0.4397,
+      "step": 901
+    },
+    {
+      "epoch": 1.5392491467576792,
+      "grad_norm": 0.41905274040841317,
+      "learning_rate": 3.496292028358604e-05,
+      "loss": 0.4368,
+      "step": 902
+    },
+    {
+      "epoch": 1.5409556313993176,
+      "grad_norm": 0.43303839375027164,
+      "learning_rate": 3.4947099627444074e-05,
+      "loss": 0.4106,
+      "step": 903
+    },
+    {
+      "epoch": 1.5426621160409555,
+      "grad_norm": 0.450215779773176,
+      "learning_rate": 3.493125775662805e-05,
+      "loss": 0.4352,
+      "step": 904
+    },
+    {
+      "epoch": 1.5443686006825939,
+      "grad_norm": 0.47692891229434176,
+      "learning_rate": 3.49153946936226e-05,
+      "loss": 0.4565,
+      "step": 905
+    },
+    {
+      "epoch": 1.5460750853242322,
+      "grad_norm": 0.4373631729864461,
+      "learning_rate": 3.489951046094245e-05,
+      "loss": 0.4176,
+      "step": 906
+    },
+    {
+      "epoch": 1.5477815699658704,
+      "grad_norm": 0.45636172992483093,
+      "learning_rate": 3.488360508113235e-05,
+      "loss": 0.4953,
+      "step": 907
+    },
+    {
+      "epoch": 1.5494880546075085,
+      "grad_norm": 0.39474307753636256,
+      "learning_rate": 3.4867678576767093e-05,
+      "loss": 0.4087,
+      "step": 908
+    },
+    {
+      "epoch": 1.5511945392491469,
+      "grad_norm": 0.4661296639102082,
+      "learning_rate": 3.4851730970451434e-05,
+      "loss": 0.4117,
+      "step": 909
+    },
+    {
+      "epoch": 1.552901023890785,
+      "grad_norm": 0.45467468666026545,
+      "learning_rate": 3.483576228482008e-05,
+      "loss": 0.4424,
+      "step": 910
+    },
+    {
+      "epoch": 1.5546075085324231,
+      "grad_norm": 0.48820547483802595,
+      "learning_rate": 3.481977254253765e-05,
+      "loss": 0.4964,
+      "step": 911
+    },
+    {
+      "epoch": 1.5563139931740615,
+      "grad_norm": 0.4204745027701614,
+      "learning_rate": 3.480376176629868e-05,
+      "loss": 0.4365,
+      "step": 912
+    },
+    {
+      "epoch": 1.5580204778156996,
+      "grad_norm": 0.42272646261560387,
+      "learning_rate": 3.478772997882753e-05,
+      "loss": 0.4425,
+      "step": 913
+    },
+    {
+      "epoch": 1.5597269624573378,
+      "grad_norm": 0.4468729774453571,
+      "learning_rate": 3.4771677202878385e-05,
+      "loss": 0.5113,
+      "step": 914
+    },
+    {
+      "epoch": 1.5614334470989761,
+      "grad_norm": 0.4375932249905856,
+      "learning_rate": 3.475560346123523e-05,
+      "loss": 0.4551,
+      "step": 915
+    },
+    {
+      "epoch": 1.5631399317406145,
+      "grad_norm": 0.34994039025065354,
+      "learning_rate": 3.473950877671179e-05,
+      "loss": 0.4147,
+      "step": 916
+    },
+    {
+      "epoch": 1.5648464163822524,
+      "grad_norm": 0.4038765449860422,
+      "learning_rate": 3.472339317215154e-05,
+      "loss": 0.4225,
+      "step": 917
+    },
+    {
+      "epoch": 1.5665529010238908,
+      "grad_norm": 0.4761160218072004,
+      "learning_rate": 3.4707256670427627e-05,
+      "loss": 0.4501,
+      "step": 918
+    },
+    {
+      "epoch": 1.5682593856655291,
+      "grad_norm": 0.3461893030643718,
+      "learning_rate": 3.4691099294442864e-05,
+      "loss": 0.4178,
+      "step": 919
+    },
+    {
+      "epoch": 1.5699658703071673,
+      "grad_norm": 0.39888863696078497,
+      "learning_rate": 3.467492106712969e-05,
+      "loss": 0.4386,
+      "step": 920
+    },
+    {
+      "epoch": 1.5716723549488054,
+      "grad_norm": 0.4003656769181341,
+      "learning_rate": 3.4658722011450145e-05,
+      "loss": 0.481,
+      "step": 921
+    },
+    {
+      "epoch": 1.5733788395904438,
+      "grad_norm": 0.42337446898808284,
+      "learning_rate": 3.464250215039582e-05,
+      "loss": 0.3887,
+      "step": 922
+    },
+    {
+      "epoch": 1.575085324232082,
+      "grad_norm": 0.4076560575146216,
+      "learning_rate": 3.4626261506987834e-05,
+      "loss": 0.4886,
+      "step": 923
+    },
+    {
+      "epoch": 1.57679180887372,
+      "grad_norm": 0.4039618681457414,
+      "learning_rate": 3.461000010427683e-05,
+      "loss": 0.4292,
+      "step": 924
+    },
+    {
+      "epoch": 1.5784982935153584,
+      "grad_norm": 0.41161267348770914,
+      "learning_rate": 3.4593717965342884e-05,
+      "loss": 0.4227,
+      "step": 925
+    },
+    {
+      "epoch": 1.5802047781569966,
+      "grad_norm": 0.4363220450461537,
+      "learning_rate": 3.457741511329551e-05,
+      "loss": 0.3945,
+      "step": 926
+    },
+    {
+      "epoch": 1.5819112627986347,
+      "grad_norm": 0.4212142249085022,
+      "learning_rate": 3.4561091571273625e-05,
+      "loss": 0.4064,
+      "step": 927
+    },
+    {
+      "epoch": 1.583617747440273,
+      "grad_norm": 0.4073724699295551,
+      "learning_rate": 3.4544747362445524e-05,
+      "loss": 0.4875,
+      "step": 928
+    },
+    {
+      "epoch": 1.5853242320819114,
+      "grad_norm": 0.48761202784212476,
+      "learning_rate": 3.45283825100088e-05,
+      "loss": 0.4615,
+      "step": 929
+    },
+    {
+      "epoch": 1.5870307167235493,
+      "grad_norm": 0.3988491731909571,
+      "learning_rate": 3.451199703719039e-05,
+      "loss": 0.4145,
+      "step": 930
+    },
+    {
+      "epoch": 1.5887372013651877,
+      "grad_norm": 0.42670855330547913,
+      "learning_rate": 3.449559096724646e-05,
+      "loss": 0.452,
+      "step": 931
+    },
+    {
+      "epoch": 1.590443686006826,
+      "grad_norm": 0.389587683305919,
+      "learning_rate": 3.4479164323462436e-05,
+      "loss": 0.4067,
+      "step": 932
+    },
+    {
+      "epoch": 1.5921501706484642,
+      "grad_norm": 0.41461434900801203,
+      "learning_rate": 3.446271712915294e-05,
+      "loss": 0.4555,
+      "step": 933
+    },
+    {
+      "epoch": 1.5938566552901023,
+      "grad_norm": 0.40604885019880904,
+      "learning_rate": 3.444624940766173e-05,
+      "loss": 0.4417,
+      "step": 934
+    },
+    {
+      "epoch": 1.5955631399317407,
+      "grad_norm": 0.3909617504686526,
+      "learning_rate": 3.442976118236175e-05,
+      "loss": 0.4226,
+      "step": 935
+    },
+    {
+      "epoch": 1.5972696245733788,
+      "grad_norm": 0.4369790065778818,
+      "learning_rate": 3.4413252476655e-05,
+      "loss": 0.4459,
+      "step": 936
+    },
+    {
+      "epoch": 1.598976109215017,
+      "grad_norm": 0.4160896979655948,
+      "learning_rate": 3.439672331397259e-05,
+      "loss": 0.4324,
+      "step": 937
+    },
+    {
+      "epoch": 1.6006825938566553,
+      "grad_norm": 0.40068228702102676,
+      "learning_rate": 3.4380173717774635e-05,
+      "loss": 0.4231,
+      "step": 938
+    },
+    {
+      "epoch": 1.6023890784982935,
+      "grad_norm": 0.5049147324987331,
+      "learning_rate": 3.436360371155025e-05,
+      "loss": 0.4489,
+      "step": 939
+    },
+    {
+      "epoch": 1.6040955631399316,
+      "grad_norm": 0.41968825833903556,
+      "learning_rate": 3.434701331881754e-05,
+      "loss": 0.4231,
+      "step": 940
+    },
+    {
+      "epoch": 1.60580204778157,
+      "grad_norm": 0.45615630046040506,
+      "learning_rate": 3.433040256312352e-05,
+      "loss": 0.4533,
+      "step": 941
+    },
+    {
+      "epoch": 1.6075085324232083,
+      "grad_norm": 0.42769386304041224,
+      "learning_rate": 3.431377146804414e-05,
+      "loss": 0.4653,
+      "step": 942
+    },
+    {
+      "epoch": 1.6092150170648463,
+      "grad_norm": 0.41982860108854264,
+      "learning_rate": 3.429712005718417e-05,
+      "loss": 0.4273,
+      "step": 943
+    },
+    {
+      "epoch": 1.6109215017064846,
+      "grad_norm": 0.4127398132688913,
+      "learning_rate": 3.4280448354177275e-05,
+      "loss": 0.4324,
+      "step": 944
+    },
+    {
+      "epoch": 1.612627986348123,
+      "grad_norm": 0.4386242108058767,
+      "learning_rate": 3.426375638268586e-05,
+      "loss": 0.4409,
+      "step": 945
+    },
+    {
+      "epoch": 1.6143344709897611,
+      "grad_norm": 0.41443492371586427,
+      "learning_rate": 3.424704416640115e-05,
+      "loss": 0.4001,
+      "step": 946
+    },
+    {
+      "epoch": 1.6160409556313993,
+      "grad_norm": 0.41419820104479615,
+      "learning_rate": 3.423031172904305e-05,
+      "loss": 0.4275,
+      "step": 947
+    },
+    {
+      "epoch": 1.6177474402730376,
+      "grad_norm": 0.4171958425823162,
+      "learning_rate": 3.421355909436022e-05,
+      "loss": 0.4578,
+      "step": 948
+    },
+    {
+      "epoch": 1.6194539249146758,
+      "grad_norm": 0.35257128369438767,
+      "learning_rate": 3.4196786286129945e-05,
+      "loss": 0.4287,
+      "step": 949
+    },
+    {
+      "epoch": 1.621160409556314,
+      "grad_norm": 0.4523630638066935,
+      "learning_rate": 3.417999332815817e-05,
+      "loss": 0.4475,
+      "step": 950
+    },
+    {
+      "epoch": 1.6228668941979523,
+      "grad_norm": 0.38733410491674525,
+      "learning_rate": 3.416318024427942e-05,
+      "loss": 0.4415,
+      "step": 951
+    },
+    {
+      "epoch": 1.6245733788395904,
+      "grad_norm": 0.4853033858137463,
+      "learning_rate": 3.414634705835679e-05,
+      "loss": 0.4385,
+      "step": 952
+    },
+    {
+      "epoch": 1.6262798634812285,
+      "grad_norm": 0.398748208356677,
+      "learning_rate": 3.412949379428192e-05,
+      "loss": 0.4077,
+      "step": 953
+    },
+    {
+      "epoch": 1.627986348122867,
+      "grad_norm": 0.4275818403560876,
+      "learning_rate": 3.411262047597492e-05,
+      "loss": 0.4107,
+      "step": 954
+    },
+    {
+      "epoch": 1.6296928327645053,
+      "grad_norm": 0.4535435912627168,
+      "learning_rate": 3.40957271273844e-05,
+      "loss": 0.4263,
+      "step": 955
+    },
+    {
+      "epoch": 1.6313993174061432,
+      "grad_norm": 0.40462878667784596,
+      "learning_rate": 3.407881377248736e-05,
+      "loss": 0.4792,
+      "step": 956
+    },
+    {
+      "epoch": 1.6331058020477816,
+      "grad_norm": 0.37436342835477465,
+      "learning_rate": 3.4061880435289214e-05,
+      "loss": 0.4293,
+      "step": 957
+    },
+    {
+      "epoch": 1.63481228668942,
+      "grad_norm": 0.41145182800772445,
+      "learning_rate": 3.404492713982375e-05,
+      "loss": 0.4532,
+      "step": 958
+    },
+    {
+      "epoch": 1.636518771331058,
+      "grad_norm": 0.47069268076938797,
+      "learning_rate": 3.402795391015307e-05,
+      "loss": 0.4421,
+      "step": 959
+    },
+    {
+      "epoch": 1.6382252559726962,
+      "grad_norm": 0.3894286690190037,
+      "learning_rate": 3.401096077036755e-05,
+      "loss": 0.4508,
+      "step": 960
+    },
+    {
+      "epoch": 1.6399317406143346,
+      "grad_norm": 0.4354063816331828,
+      "learning_rate": 3.399394774458586e-05,
+      "loss": 0.464,
+      "step": 961
+    },
+    {
+      "epoch": 1.6416382252559727,
+      "grad_norm": 0.4388846590151156,
+      "learning_rate": 3.3976914856954876e-05,
+      "loss": 0.4518,
+      "step": 962
+    },
+    {
+      "epoch": 1.6433447098976108,
+      "grad_norm": 0.45091502302612824,
+      "learning_rate": 3.3959862131649665e-05,
+      "loss": 0.4192,
+      "step": 963
+    },
+    {
+      "epoch": 1.6450511945392492,
+      "grad_norm": 0.4651907434306189,
+      "learning_rate": 3.3942789592873454e-05,
+      "loss": 0.5461,
+      "step": 964
+    },
+    {
+      "epoch": 1.6467576791808873,
+      "grad_norm": 0.4245109764212635,
+      "learning_rate": 3.392569726485759e-05,
+      "loss": 0.4212,
+      "step": 965
+    },
+    {
+      "epoch": 1.6484641638225255,
+      "grad_norm": 0.4288455867724874,
+      "learning_rate": 3.390858517186149e-05,
+      "loss": 0.421,
+      "step": 966
+    },
+    {
+      "epoch": 1.6501706484641638,
+      "grad_norm": 0.4864298431309527,
+      "learning_rate": 3.389145333817266e-05,
+      "loss": 0.4871,
+      "step": 967
+    },
+    {
+      "epoch": 1.6518771331058022,
+      "grad_norm": 0.4545088336033929,
+      "learning_rate": 3.387430178810661e-05,
+      "loss": 0.4202,
+      "step": 968
+    },
+    {
+      "epoch": 1.6535836177474401,
+      "grad_norm": 0.37090624906992437,
+      "learning_rate": 3.38571305460068e-05,
+      "loss": 0.4182,
+      "step": 969
+    },
+    {
+      "epoch": 1.6552901023890785,
+      "grad_norm": 0.3815866609881445,
+      "learning_rate": 3.383993963624469e-05,
+      "loss": 0.4115,
+      "step": 970
+    },
+    {
+      "epoch": 1.6569965870307168,
+      "grad_norm": 0.39516260065390635,
+      "learning_rate": 3.3822729083219635e-05,
+      "loss": 0.4106,
+      "step": 971
+    },
+    {
+      "epoch": 1.658703071672355,
+      "grad_norm": 0.41165425795653404,
+      "learning_rate": 3.380549891135884e-05,
+      "loss": 0.4509,
+      "step": 972
+    },
+    {
+      "epoch": 1.6604095563139931,
+      "grad_norm": 0.4106639445516941,
+      "learning_rate": 3.378824914511741e-05,
+      "loss": 0.4139,
+      "step": 973
+    },
+    {
+      "epoch": 1.6621160409556315,
+      "grad_norm": 0.42566925025520447,
+      "learning_rate": 3.3770979808978225e-05,
+      "loss": 0.456,
+      "step": 974
+    },
+    {
+      "epoch": 1.6638225255972696,
+      "grad_norm": 0.39512181124894374,
+      "learning_rate": 3.375369092745195e-05,
+      "loss": 0.4079,
+      "step": 975
+    },
+    {
+      "epoch": 1.6655290102389078,
+      "grad_norm": 0.38230396858721344,
+      "learning_rate": 3.373638252507698e-05,
+      "loss": 0.4228,
+      "step": 976
+    },
+    {
+      "epoch": 1.6672354948805461,
+      "grad_norm": 0.47063681118009787,
+      "learning_rate": 3.371905462641944e-05,
+      "loss": 0.4472,
+      "step": 977
+    },
+    {
+      "epoch": 1.6689419795221843,
+      "grad_norm": 0.4447077825481116,
+      "learning_rate": 3.3701707256073105e-05,
+      "loss": 0.4496,
+      "step": 978
+    },
+    {
+      "epoch": 1.6706484641638224,
+      "grad_norm": 0.4060212414485413,
+      "learning_rate": 3.3684340438659405e-05,
+      "loss": 0.4272,
+      "step": 979
+    },
+    {
+      "epoch": 1.6723549488054608,
+      "grad_norm": 0.3795742306274751,
+      "learning_rate": 3.366695419882734e-05,
+      "loss": 0.4231,
+      "step": 980
+    },
+    {
+      "epoch": 1.6740614334470991,
+      "grad_norm": 0.43991563290482527,
+      "learning_rate": 3.364954856125351e-05,
+      "loss": 0.4345,
+      "step": 981
+    },
+    {
+      "epoch": 1.675767918088737,
+      "grad_norm": 0.3877476024110567,
+      "learning_rate": 3.363212355064205e-05,
+      "loss": 0.4349,
+      "step": 982
+    },
+    {
+      "epoch": 1.6774744027303754,
+      "grad_norm": 0.4179531198525114,
+      "learning_rate": 3.361467919172454e-05,
+      "loss": 0.4507,
+      "step": 983
+    },
+    {
+      "epoch": 1.6791808873720138,
+      "grad_norm": 0.39344144979568285,
+      "learning_rate": 3.3597215509260086e-05,
+      "loss": 0.444,
+      "step": 984
+    },
+    {
+      "epoch": 1.680887372013652,
+      "grad_norm": 0.3724567972038973,
+      "learning_rate": 3.357973252803518e-05,
+      "loss": 0.4416,
+      "step": 985
+    },
+    {
+      "epoch": 1.68259385665529,
+      "grad_norm": 0.3595142334695815,
+      "learning_rate": 3.356223027286372e-05,
+      "loss": 0.4182,
+      "step": 986
+    },
+    {
+      "epoch": 1.6843003412969284,
+      "grad_norm": 0.4257994094890164,
+      "learning_rate": 3.354470876858695e-05,
+      "loss": 0.4591,
+      "step": 987
+    },
+    {
+      "epoch": 1.6860068259385665,
+      "grad_norm": 0.38949058168413975,
+      "learning_rate": 3.3527168040073446e-05,
+      "loss": 0.4545,
+      "step": 988
+    },
+    {
+      "epoch": 1.6877133105802047,
+      "grad_norm": 0.42007113519918304,
+      "learning_rate": 3.3509608112219055e-05,
+      "loss": 0.4369,
+      "step": 989
+    },
+    {
+      "epoch": 1.689419795221843,
+      "grad_norm": 0.36248815732923834,
+      "learning_rate": 3.34920290099469e-05,
+      "loss": 0.4129,
+      "step": 990
+    },
+    {
+      "epoch": 1.6911262798634812,
+      "grad_norm": 0.4458775071227498,
+      "learning_rate": 3.347443075820729e-05,
+      "loss": 0.4485,
+      "step": 991
+    },
+    {
+      "epoch": 1.6928327645051193,
+      "grad_norm": 0.3893563280856113,
+      "learning_rate": 3.345681338197772e-05,
+      "loss": 0.432,
+      "step": 992
+    },
+    {
+      "epoch": 1.6945392491467577,
+      "grad_norm": 0.3739107046460591,
+      "learning_rate": 3.3439176906262835e-05,
+      "loss": 0.4681,
+      "step": 993
+    },
+    {
+      "epoch": 1.696245733788396,
+      "grad_norm": 0.36502308371838477,
+      "learning_rate": 3.34215213560944e-05,
+      "loss": 0.4231,
+      "step": 994
+    },
+    {
+      "epoch": 1.697952218430034,
+      "grad_norm": 0.40461580892457993,
+      "learning_rate": 3.340384675653123e-05,
+      "loss": 0.4812,
+      "step": 995
+    },
+    {
+      "epoch": 1.6996587030716723,
+      "grad_norm": 0.4621930292777579,
+      "learning_rate": 3.3386153132659184e-05,
+      "loss": 0.4498,
+      "step": 996
+    },
+    {
+      "epoch": 1.7013651877133107,
+      "grad_norm": 0.3505765117737797,
+      "learning_rate": 3.336844050959113e-05,
+      "loss": 0.4362,
+      "step": 997
+    },
+    {
+      "epoch": 1.7030716723549488,
+      "grad_norm": 0.4085279897692518,
+      "learning_rate": 3.335070891246689e-05,
+      "loss": 0.5195,
+      "step": 998
+    },
+    {
+      "epoch": 1.704778156996587,
+      "grad_norm": 0.4011492088215556,
+      "learning_rate": 3.3332958366453225e-05,
+      "loss": 0.429,
+      "step": 999
+    },
+    {
+      "epoch": 1.7064846416382253,
+      "grad_norm": 0.3873576769554521,
+      "learning_rate": 3.3315188896743796e-05,
+      "loss": 0.4301,
+      "step": 1000
+    },
+    {
+      "epoch": 1.7081911262798635,
+      "grad_norm": 0.43132909483485776,
+      "learning_rate": 3.32974005285591e-05,
+      "loss": 0.4484,
+      "step": 1001
+    },
+    {
+      "epoch": 1.7098976109215016,
+      "grad_norm": 0.34576970317395267,
+      "learning_rate": 3.327959328714649e-05,
+      "loss": 0.4332,
+      "step": 1002
+    },
+    {
+      "epoch": 1.71160409556314,
+      "grad_norm": 0.42511892675815227,
+      "learning_rate": 3.326176719778008e-05,
+      "loss": 0.4389,
+      "step": 1003
+    },
+    {
+      "epoch": 1.713310580204778,
+      "grad_norm": 0.3821873635784118,
+      "learning_rate": 3.3243922285760736e-05,
+      "loss": 0.4691,
+      "step": 1004
+    },
+    {
+      "epoch": 1.7150170648464163,
+      "grad_norm": 0.3892273970414259,
+      "learning_rate": 3.322605857641606e-05,
+      "loss": 0.406,
+      "step": 1005
+    },
+    {
+      "epoch": 1.7167235494880546,
+      "grad_norm": 0.4004541387857402,
+      "learning_rate": 3.320817609510032e-05,
+      "loss": 0.3913,
+      "step": 1006
+    },
+    {
+      "epoch": 1.718430034129693,
+      "grad_norm": 0.4130824978333927,
+      "learning_rate": 3.319027486719441e-05,
+      "loss": 0.4311,
+      "step": 1007
+    },
+    {
+      "epoch": 1.7201365187713311,
+      "grad_norm": 0.35282306819720366,
+      "learning_rate": 3.3172354918105864e-05,
+      "loss": 0.4356,
+      "step": 1008
+    },
+    {
+      "epoch": 1.7218430034129693,
+      "grad_norm": 0.4109825192006032,
+      "learning_rate": 3.3154416273268766e-05,
+      "loss": 0.4107,
+      "step": 1009
+    },
+    {
+      "epoch": 1.7235494880546076,
+      "grad_norm": 0.3951503042835757,
+      "learning_rate": 3.313645895814375e-05,
+      "loss": 0.4129,
+      "step": 1010
+    },
+    {
+      "epoch": 1.7252559726962458,
+      "grad_norm": 0.381791539552012,
+      "learning_rate": 3.311848299821793e-05,
+      "loss": 0.4365,
+      "step": 1011
+    },
+    {
+      "epoch": 1.726962457337884,
+      "grad_norm": 0.47677304464409714,
+      "learning_rate": 3.31004884190049e-05,
+      "loss": 0.4561,
+      "step": 1012
+    },
+    {
+      "epoch": 1.7286689419795223,
+      "grad_norm": 0.3537629215873041,
+      "learning_rate": 3.3082475246044666e-05,
+      "loss": 0.4624,
+      "step": 1013
+    },
+    {
+      "epoch": 1.7303754266211604,
+      "grad_norm": 0.37454400532288973,
+      "learning_rate": 3.306444350490364e-05,
+      "loss": 0.4184,
+      "step": 1014
+    },
+    {
+      "epoch": 1.7320819112627985,
+      "grad_norm": 0.4527335732070691,
+      "learning_rate": 3.3046393221174584e-05,
+      "loss": 0.4619,
+      "step": 1015
+    },
+    {
+      "epoch": 1.733788395904437,
+      "grad_norm": 0.34280779884824736,
+      "learning_rate": 3.302832442047656e-05,
+      "loss": 0.3938,
+      "step": 1016
+    },
+    {
+      "epoch": 1.735494880546075,
+      "grad_norm": 0.39422292233838113,
+      "learning_rate": 3.301023712845494e-05,
+      "loss": 0.4844,
+      "step": 1017
+    },
+    {
+      "epoch": 1.7372013651877132,
+      "grad_norm": 0.40200103116961866,
+      "learning_rate": 3.2992131370781324e-05,
+      "loss": 0.4306,
+      "step": 1018
+    },
+    {
+      "epoch": 1.7389078498293515,
+      "grad_norm": 0.3432667076140728,
+      "learning_rate": 3.297400717315351e-05,
+      "loss": 0.4444,
+      "step": 1019
+    },
+    {
+      "epoch": 1.74061433447099,
+      "grad_norm": 0.39592905119657307,
+      "learning_rate": 3.29558645612955e-05,
+      "loss": 0.4496,
+      "step": 1020
+    },
+    {
+      "epoch": 1.742320819112628,
+      "grad_norm": 0.3732006670729543,
+      "learning_rate": 3.2937703560957405e-05,
+      "loss": 0.4519,
+      "step": 1021
+    },
+    {
+      "epoch": 1.7440273037542662,
+      "grad_norm": 0.4016176887258834,
+      "learning_rate": 3.2919524197915436e-05,
+      "loss": 0.4372,
+      "step": 1022
+    },
+    {
+      "epoch": 1.7457337883959045,
+      "grad_norm": 0.4392872292719323,
+      "learning_rate": 3.290132649797188e-05,
+      "loss": 0.4682,
+      "step": 1023
+    },
+    {
+      "epoch": 1.7474402730375427,
+      "grad_norm": 0.3769716755751727,
+      "learning_rate": 3.288311048695506e-05,
+      "loss": 0.416,
+      "step": 1024
+    },
+    {
+      "epoch": 1.7491467576791808,
+      "grad_norm": 0.4341939032109703,
+      "learning_rate": 3.2864876190719245e-05,
+      "loss": 0.4574,
+      "step": 1025
+    },
+    {
+      "epoch": 1.7508532423208192,
+      "grad_norm": 0.38497051556421485,
+      "learning_rate": 3.28466236351447e-05,
+      "loss": 0.4247,
+      "step": 1026
+    },
+    {
+      "epoch": 1.7525597269624573,
+      "grad_norm": 0.38460545318593387,
+      "learning_rate": 3.282835284613759e-05,
+      "loss": 0.4178,
+      "step": 1027
+    },
+    {
+      "epoch": 1.7542662116040955,
+      "grad_norm": 0.44963309509926697,
+      "learning_rate": 3.281006384962994e-05,
+      "loss": 0.4662,
+      "step": 1028
+    },
+    {
+      "epoch": 1.7559726962457338,
+      "grad_norm": 0.39444392040125487,
+      "learning_rate": 3.279175667157966e-05,
+      "loss": 0.4409,
+      "step": 1029
+    },
+    {
+      "epoch": 1.757679180887372,
+      "grad_norm": 0.379860254515413,
+      "learning_rate": 3.277343133797042e-05,
+      "loss": 0.4346,
+      "step": 1030
+    },
+    {
+      "epoch": 1.75938566552901,
+      "grad_norm": 0.3928991006625797,
+      "learning_rate": 3.2755087874811696e-05,
+      "loss": 0.4365,
+      "step": 1031
+    },
+    {
+      "epoch": 1.7610921501706485,
+      "grad_norm": 0.3758416937166645,
+      "learning_rate": 3.2736726308138666e-05,
+      "loss": 0.436,
+      "step": 1032
+    },
+    {
+      "epoch": 1.7627986348122868,
+      "grad_norm": 0.3984699724279319,
+      "learning_rate": 3.271834666401222e-05,
+      "loss": 0.4159,
+      "step": 1033
+    },
+    {
+      "epoch": 1.764505119453925,
+      "grad_norm": 0.40347824679120803,
+      "learning_rate": 3.2699948968518905e-05,
+      "loss": 0.4212,
+      "step": 1034
+    },
+    {
+      "epoch": 1.766211604095563,
+      "grad_norm": 0.41223483424837204,
+      "learning_rate": 3.268153324777088e-05,
+      "loss": 0.4342,
+      "step": 1035
+    },
+    {
+      "epoch": 1.7679180887372015,
+      "grad_norm": 0.37001044900354435,
+      "learning_rate": 3.26630995279059e-05,
+      "loss": 0.434,
+      "step": 1036
+    },
+    {
+      "epoch": 1.7696245733788396,
+      "grad_norm": 0.39536129614409976,
+      "learning_rate": 3.264464783508724e-05,
+      "loss": 0.4762,
+      "step": 1037
+    },
+    {
+      "epoch": 1.7713310580204777,
+      "grad_norm": 0.386527238328391,
+      "learning_rate": 3.2626178195503725e-05,
+      "loss": 0.4104,
+      "step": 1038
+    },
+    {
+      "epoch": 1.773037542662116,
+      "grad_norm": 0.415262031466017,
+      "learning_rate": 3.260769063536962e-05,
+      "loss": 0.477,
+      "step": 1039
+    },
+    {
+      "epoch": 1.7747440273037542,
+      "grad_norm": 0.37361364115284673,
+      "learning_rate": 3.2589185180924634e-05,
+      "loss": 0.4211,
+      "step": 1040
+    },
+    {
+      "epoch": 1.7764505119453924,
+      "grad_norm": 0.41212711102387606,
+      "learning_rate": 3.257066185843388e-05,
+      "loss": 0.4284,
+      "step": 1041
+    },
+    {
+      "epoch": 1.7781569965870307,
+      "grad_norm": 0.369270748557867,
+      "learning_rate": 3.255212069418782e-05,
+      "loss": 0.4463,
+      "step": 1042
+    },
+    {
+      "epoch": 1.7798634812286689,
+      "grad_norm": 0.3712719038302196,
+      "learning_rate": 3.253356171450225e-05,
+      "loss": 0.4596,
+      "step": 1043
+    },
+    {
+      "epoch": 1.781569965870307,
+      "grad_norm": 0.3792687716129807,
+      "learning_rate": 3.251498494571825e-05,
+      "loss": 0.4536,
+      "step": 1044
+    },
+    {
+      "epoch": 1.7832764505119454,
+      "grad_norm": 0.3548862654831631,
+      "learning_rate": 3.249639041420214e-05,
+      "loss": 0.4211,
+      "step": 1045
+    },
+    {
+      "epoch": 1.7849829351535837,
+      "grad_norm": 0.35616337807709136,
+      "learning_rate": 3.247777814634545e-05,
+      "loss": 0.4152,
+      "step": 1046
+    },
+    {
+      "epoch": 1.786689419795222,
+      "grad_norm": 0.38087475817301647,
+      "learning_rate": 3.245914816856491e-05,
+      "loss": 0.398,
+      "step": 1047
+    },
+    {
+      "epoch": 1.78839590443686,
+      "grad_norm": 0.4399343175733236,
+      "learning_rate": 3.244050050730235e-05,
+      "loss": 0.4572,
+      "step": 1048
+    },
+    {
+      "epoch": 1.7901023890784984,
+      "grad_norm": 0.33778622101738565,
+      "learning_rate": 3.242183518902471e-05,
+      "loss": 0.4066,
+      "step": 1049
+    },
+    {
+      "epoch": 1.7918088737201365,
+      "grad_norm": 0.34462130214733205,
+      "learning_rate": 3.2403152240224016e-05,
+      "loss": 0.4171,
+      "step": 1050
+    },
+    {
+      "epoch": 1.7935153583617747,
+      "grad_norm": 0.3696951944918031,
+      "learning_rate": 3.238445168741728e-05,
+      "loss": 0.4322,
+      "step": 1051
+    },
+    {
+      "epoch": 1.795221843003413,
+      "grad_norm": 0.34184919207550546,
+      "learning_rate": 3.2365733557146524e-05,
+      "loss": 0.4431,
+      "step": 1052
+    },
+    {
+      "epoch": 1.7969283276450512,
+      "grad_norm": 0.38993811879144025,
+      "learning_rate": 3.23469978759787e-05,
+      "loss": 0.4215,
+      "step": 1053
+    },
+    {
+      "epoch": 1.7986348122866893,
+      "grad_norm": 0.327127168964827,
+      "learning_rate": 3.232824467050569e-05,
+      "loss": 0.4107,
+      "step": 1054
+    },
+    {
+      "epoch": 1.8003412969283277,
+      "grad_norm": 0.3996487580489501,
+      "learning_rate": 3.2309473967344246e-05,
+      "loss": 0.4827,
+      "step": 1055
+    },
+    {
+      "epoch": 1.802047781569966,
+      "grad_norm": 0.3674756187328674,
+      "learning_rate": 3.229068579313593e-05,
+      "loss": 0.4398,
+      "step": 1056
+    },
+    {
+      "epoch": 1.803754266211604,
+      "grad_norm": 0.41368242038508635,
+      "learning_rate": 3.227188017454713e-05,
+      "loss": 0.4392,
+      "step": 1057
+    },
+    {
+      "epoch": 1.8054607508532423,
+      "grad_norm": 0.39890941941714153,
+      "learning_rate": 3.225305713826898e-05,
+      "loss": 0.4428,
+      "step": 1058
+    },
+    {
+      "epoch": 1.8071672354948807,
+      "grad_norm": 0.41038117710239236,
+      "learning_rate": 3.223421671101734e-05,
+      "loss": 0.4228,
+      "step": 1059
+    },
+    {
+      "epoch": 1.8088737201365188,
+      "grad_norm": 0.4226904975833523,
+      "learning_rate": 3.2215358919532735e-05,
+      "loss": 0.4105,
+      "step": 1060
+    },
+    {
+      "epoch": 1.810580204778157,
+      "grad_norm": 0.3882678446191072,
+      "learning_rate": 3.219648379058037e-05,
+      "loss": 0.4476,
+      "step": 1061
+    },
+    {
+      "epoch": 1.8122866894197953,
+      "grad_norm": 0.4087205311385779,
+      "learning_rate": 3.217759135095004e-05,
+      "loss": 0.429,
+      "step": 1062
+    },
+    {
+      "epoch": 1.8139931740614335,
+      "grad_norm": 0.3340537469358327,
+      "learning_rate": 3.215868162745609e-05,
+      "loss": 0.4568,
+      "step": 1063
+    },
+    {
+      "epoch": 1.8156996587030716,
+      "grad_norm": 0.37739535451555306,
+      "learning_rate": 3.213975464693743e-05,
+      "loss": 0.4448,
+      "step": 1064
+    },
+    {
+      "epoch": 1.81740614334471,
+      "grad_norm": 0.359159589987258,
+      "learning_rate": 3.2120810436257435e-05,
+      "loss": 0.4423,
+      "step": 1065
+    },
+    {
+      "epoch": 1.819112627986348,
+      "grad_norm": 0.3493876724755677,
+      "learning_rate": 3.2101849022303955e-05,
+      "loss": 0.4196,
+      "step": 1066
+    },
+    {
+      "epoch": 1.8208191126279862,
+      "grad_norm": 0.3574344041797343,
+      "learning_rate": 3.2082870431989245e-05,
+      "loss": 0.4215,
+      "step": 1067
+    },
+    {
+      "epoch": 1.8225255972696246,
+      "grad_norm": 0.3449199113564368,
+      "learning_rate": 3.2063874692249947e-05,
+      "loss": 0.4101,
+      "step": 1068
+    },
+    {
+      "epoch": 1.824232081911263,
+      "grad_norm": 0.3778493662630658,
+      "learning_rate": 3.204486183004703e-05,
+      "loss": 0.4136,
+      "step": 1069
+    },
+    {
+      "epoch": 1.8259385665529009,
+      "grad_norm": 0.3524828063713649,
+      "learning_rate": 3.2025831872365784e-05,
+      "loss": 0.4536,
+      "step": 1070
+    },
+    {
+      "epoch": 1.8276450511945392,
+      "grad_norm": 0.33278250688476424,
+      "learning_rate": 3.200678484621575e-05,
+      "loss": 0.4134,
+      "step": 1071
+    },
+    {
+      "epoch": 1.8293515358361776,
+      "grad_norm": 0.40248605788959163,
+      "learning_rate": 3.19877207786307e-05,
+      "loss": 0.4232,
+      "step": 1072
+    },
+    {
+      "epoch": 1.8310580204778157,
+      "grad_norm": 0.3703730711217583,
+      "learning_rate": 3.1968639696668584e-05,
+      "loss": 0.4307,
+      "step": 1073
+    },
+    {
+      "epoch": 1.8327645051194539,
+      "grad_norm": 0.3614945164676251,
+      "learning_rate": 3.194954162741152e-05,
+      "loss": 0.4467,
+      "step": 1074
+    },
+    {
+      "epoch": 1.8344709897610922,
+      "grad_norm": 0.3540166492601531,
+      "learning_rate": 3.1930426597965714e-05,
+      "loss": 0.4411,
+      "step": 1075
+    },
+    {
+      "epoch": 1.8361774744027304,
+      "grad_norm": 0.3696926170561594,
+      "learning_rate": 3.1911294635461455e-05,
+      "loss": 0.4348,
+      "step": 1076
+    },
+    {
+      "epoch": 1.8378839590443685,
+      "grad_norm": 0.3662622788229689,
+      "learning_rate": 3.189214576705307e-05,
+      "loss": 0.4277,
+      "step": 1077
+    },
+    {
+      "epoch": 1.8395904436860069,
+      "grad_norm": 0.3895871706871898,
+      "learning_rate": 3.1872980019918864e-05,
+      "loss": 0.4163,
+      "step": 1078
+    },
+    {
+      "epoch": 1.841296928327645,
+      "grad_norm": 0.3438737423599471,
+      "learning_rate": 3.1853797421261125e-05,
+      "loss": 0.4037,
+      "step": 1079
+    },
+    {
+      "epoch": 1.8430034129692832,
+      "grad_norm": 0.40485789953961415,
+      "learning_rate": 3.183459799830603e-05,
+      "loss": 0.4358,
+      "step": 1080
+    },
+    {
+      "epoch": 1.8447098976109215,
+      "grad_norm": 0.3678614520421831,
+      "learning_rate": 3.181538177830366e-05,
+      "loss": 0.4076,
+      "step": 1081
+    },
+    {
+      "epoch": 1.8464163822525599,
+      "grad_norm": 0.3891439721848829,
+      "learning_rate": 3.179614878852792e-05,
+      "loss": 0.4644,
+      "step": 1082
+    },
+    {
+      "epoch": 1.8481228668941978,
+      "grad_norm": 0.5129982908158024,
+      "learning_rate": 3.177689905627651e-05,
+      "loss": 0.4325,
+      "step": 1083
+    },
+    {
+      "epoch": 1.8498293515358362,
+      "grad_norm": 0.43424400708864125,
+      "learning_rate": 3.1757632608870915e-05,
+      "loss": 0.4242,
+      "step": 1084
+    },
+    {
+      "epoch": 1.8515358361774745,
+      "grad_norm": 0.4108442271574386,
+      "learning_rate": 3.173834947365634e-05,
+      "loss": 0.4156,
+      "step": 1085
+    },
+    {
+      "epoch": 1.8532423208191127,
+      "grad_norm": 0.3703682959673418,
+      "learning_rate": 3.171904967800166e-05,
+      "loss": 0.423,
+      "step": 1086
+    },
+    {
+      "epoch": 1.8549488054607508,
+      "grad_norm": 0.3731841091953381,
+      "learning_rate": 3.1699733249299395e-05,
+      "loss": 0.4447,
+      "step": 1087
+    },
+    {
+      "epoch": 1.8566552901023892,
+      "grad_norm": 0.3924445032580443,
+      "learning_rate": 3.16804002149657e-05,
+      "loss": 0.4099,
+      "step": 1088
+    },
+    {
+      "epoch": 1.8583617747440273,
+      "grad_norm": 0.36438230574683184,
+      "learning_rate": 3.166105060244029e-05,
+      "loss": 0.4359,
+      "step": 1089
+    },
+    {
+      "epoch": 1.8600682593856654,
+      "grad_norm": 0.3672867920210704,
+      "learning_rate": 3.164168443918636e-05,
+      "loss": 0.4371,
+      "step": 1090
+    },
+    {
+      "epoch": 1.8617747440273038,
+      "grad_norm": 0.3619548462607363,
+      "learning_rate": 3.1622301752690675e-05,
+      "loss": 0.4287,
+      "step": 1091
+    },
+    {
+      "epoch": 1.863481228668942,
+      "grad_norm": 0.3989562928398588,
+      "learning_rate": 3.1602902570463396e-05,
+      "loss": 0.3808,
+      "step": 1092
+    },
+    {
+      "epoch": 1.86518771331058,
+      "grad_norm": 0.3400880684976596,
+      "learning_rate": 3.158348692003812e-05,
+      "loss": 0.4552,
+      "step": 1093
+    },
+    {
+      "epoch": 1.8668941979522184,
+      "grad_norm": 0.3884228545028214,
+      "learning_rate": 3.156405482897181e-05,
+      "loss": 0.4424,
+      "step": 1094
+    },
+    {
+      "epoch": 1.8686006825938568,
+      "grad_norm": 0.47001264022373496,
+      "learning_rate": 3.154460632484477e-05,
+      "loss": 0.4518,
+      "step": 1095
+    },
+    {
+      "epoch": 1.8703071672354947,
+      "grad_norm": 0.3709605969510673,
+      "learning_rate": 3.152514143526058e-05,
+      "loss": 0.4298,
+      "step": 1096
+    },
+    {
+      "epoch": 1.872013651877133,
+      "grad_norm": 0.4383716509647053,
+      "learning_rate": 3.15056601878461e-05,
+      "loss": 0.4678,
+      "step": 1097
+    },
+    {
+      "epoch": 1.8737201365187715,
+      "grad_norm": 0.38594310190286124,
+      "learning_rate": 3.1486162610251405e-05,
+      "loss": 0.4393,
+      "step": 1098
+    },
+    {
+      "epoch": 1.8754266211604096,
+      "grad_norm": 0.40721100356765266,
+      "learning_rate": 3.146664873014973e-05,
+      "loss": 0.435,
+      "step": 1099
+    },
+    {
+      "epoch": 1.8771331058020477,
+      "grad_norm": 0.49506906473357676,
+      "learning_rate": 3.144711857523746e-05,
+      "loss": 0.4317,
+      "step": 1100
+    },
+    {
+      "epoch": 1.878839590443686,
+      "grad_norm": 0.4090982341013302,
+      "learning_rate": 3.142757217323408e-05,
+      "loss": 0.4682,
+      "step": 1101
+    },
+    {
+      "epoch": 1.8805460750853242,
+      "grad_norm": 0.39453830650395644,
+      "learning_rate": 3.140800955188213e-05,
+      "loss": 0.4127,
+      "step": 1102
+    },
+    {
+      "epoch": 1.8822525597269624,
+      "grad_norm": 0.3727138401061236,
+      "learning_rate": 3.138843073894717e-05,
+      "loss": 0.4639,
+      "step": 1103
+    },
+    {
+      "epoch": 1.8839590443686007,
+      "grad_norm": 0.3867992352973858,
+      "learning_rate": 3.1368835762217755e-05,
+      "loss": 0.4368,
+      "step": 1104
+    },
+    {
+      "epoch": 1.8856655290102389,
+      "grad_norm": 0.3946509268348863,
+      "learning_rate": 3.1349224649505366e-05,
+      "loss": 0.4406,
+      "step": 1105
+    },
+    {
+      "epoch": 1.887372013651877,
+      "grad_norm": 0.36077305576822366,
+      "learning_rate": 3.132959742864438e-05,
+      "loss": 0.4051,
+      "step": 1106
+    },
+    {
+      "epoch": 1.8890784982935154,
+      "grad_norm": 0.40475493268433543,
+      "learning_rate": 3.130995412749206e-05,
+      "loss": 0.4724,
+      "step": 1107
+    },
+    {
+      "epoch": 1.8907849829351537,
+      "grad_norm": 0.3787893148385328,
+      "learning_rate": 3.129029477392848e-05,
+      "loss": 0.4264,
+      "step": 1108
+    },
+    {
+      "epoch": 1.8924914675767917,
+      "grad_norm": 0.3886535714745907,
+      "learning_rate": 3.127061939585649e-05,
+      "loss": 0.4208,
+      "step": 1109
+    },
+    {
+      "epoch": 1.89419795221843,
+      "grad_norm": 0.40410505742416974,
+      "learning_rate": 3.125092802120169e-05,
+      "loss": 0.4343,
+      "step": 1110
+    },
+    {
+      "epoch": 1.8959044368600684,
+      "grad_norm": 0.3793707170377886,
+      "learning_rate": 3.123122067791238e-05,
+      "loss": 0.3967,
+      "step": 1111
+    },
+    {
+      "epoch": 1.8976109215017065,
+      "grad_norm": 0.38968057558625735,
+      "learning_rate": 3.1211497393959546e-05,
+      "loss": 0.3915,
+      "step": 1112
+    },
+    {
+      "epoch": 1.8993174061433447,
+      "grad_norm": 0.43184149778702685,
+      "learning_rate": 3.119175819733677e-05,
+      "loss": 0.4112,
+      "step": 1113
+    },
+    {
+      "epoch": 1.901023890784983,
+      "grad_norm": 0.4547391887609193,
+      "learning_rate": 3.117200311606023e-05,
+      "loss": 0.4073,
+      "step": 1114
+    },
+    {
+      "epoch": 1.9027303754266212,
+      "grad_norm": 0.4023504384319604,
+      "learning_rate": 3.1152232178168655e-05,
+      "loss": 0.4562,
+      "step": 1115
+    },
+    {
+      "epoch": 1.9044368600682593,
+      "grad_norm": 0.43798414339531605,
+      "learning_rate": 3.113244541172328e-05,
+      "loss": 0.4463,
+      "step": 1116
+    },
+    {
+      "epoch": 1.9061433447098977,
+      "grad_norm": 0.43205765378414585,
+      "learning_rate": 3.111264284480779e-05,
+      "loss": 0.4811,
+      "step": 1117
+    },
+    {
+      "epoch": 1.9078498293515358,
+      "grad_norm": 0.40707407192136424,
+      "learning_rate": 3.109282450552831e-05,
+      "loss": 0.4366,
+      "step": 1118
+    },
+    {
+      "epoch": 1.909556313993174,
+      "grad_norm": 0.41441282044999433,
+      "learning_rate": 3.1072990422013354e-05,
+      "loss": 0.4344,
+      "step": 1119
+    },
+    {
+      "epoch": 1.9112627986348123,
+      "grad_norm": 0.3471608850863164,
+      "learning_rate": 3.105314062241377e-05,
+      "loss": 0.4064,
+      "step": 1120
+    },
+    {
+      "epoch": 1.9129692832764507,
+      "grad_norm": 0.407103583087269,
+      "learning_rate": 3.1033275134902714e-05,
+      "loss": 0.4307,
+      "step": 1121
+    },
+    {
+      "epoch": 1.9146757679180886,
+      "grad_norm": 0.37090540027247704,
+      "learning_rate": 3.1013393987675624e-05,
+      "loss": 0.4482,
+      "step": 1122
+    },
+    {
+      "epoch": 1.916382252559727,
+      "grad_norm": 0.3905360275094505,
+      "learning_rate": 3.099349720895015e-05,
+      "loss": 0.4102,
+      "step": 1123
+    },
+    {
+      "epoch": 1.9180887372013653,
+      "grad_norm": 0.44609768346190243,
+      "learning_rate": 3.0973584826966114e-05,
+      "loss": 0.4424,
+      "step": 1124
+    },
+    {
+      "epoch": 1.9197952218430034,
+      "grad_norm": 0.3742218681777803,
+      "learning_rate": 3.095365686998552e-05,
+      "loss": 0.4411,
+      "step": 1125
+    },
+    {
+      "epoch": 1.9215017064846416,
+      "grad_norm": 0.4251254043386818,
+      "learning_rate": 3.093371336629245e-05,
+      "loss": 0.4514,
+      "step": 1126
+    },
+    {
+      "epoch": 1.92320819112628,
+      "grad_norm": 0.36785879492426965,
+      "learning_rate": 3.091375434419306e-05,
+      "loss": 0.4343,
+      "step": 1127
+    },
+    {
+      "epoch": 1.924914675767918,
+      "grad_norm": 0.3703783175452118,
+      "learning_rate": 3.089377983201553e-05,
+      "loss": 0.4405,
+      "step": 1128
+    },
+    {
+      "epoch": 1.9266211604095562,
+      "grad_norm": 0.3659750932274629,
+      "learning_rate": 3.0873789858110037e-05,
+      "loss": 0.4056,
+      "step": 1129
+    },
+    {
+      "epoch": 1.9283276450511946,
+      "grad_norm": 0.3853425455044206,
+      "learning_rate": 3.085378445084868e-05,
+      "loss": 0.4262,
+      "step": 1130
+    },
+    {
+      "epoch": 1.9300341296928327,
+      "grad_norm": 0.3851004913621087,
+      "learning_rate": 3.0833763638625466e-05,
+      "loss": 0.447,
+      "step": 1131
+    },
+    {
+      "epoch": 1.9317406143344709,
+      "grad_norm": 0.4442090248664147,
+      "learning_rate": 3.0813727449856305e-05,
+      "loss": 0.467,
+      "step": 1132
+    },
+    {
+      "epoch": 1.9334470989761092,
+      "grad_norm": 0.3786585427874933,
+      "learning_rate": 3.0793675912978875e-05,
+      "loss": 0.4297,
+      "step": 1133
+    },
+    {
+      "epoch": 1.9351535836177476,
+      "grad_norm": 0.38519490340606855,
+      "learning_rate": 3.0773609056452683e-05,
+      "loss": 0.4627,
+      "step": 1134
+    },
+    {
+      "epoch": 1.9368600682593855,
+      "grad_norm": 0.35724071458801504,
+      "learning_rate": 3.0753526908758956e-05,
+      "loss": 0.4645,
+      "step": 1135
+    },
+    {
+      "epoch": 1.9385665529010239,
+      "grad_norm": 0.426621147363609,
+      "learning_rate": 3.073342949840063e-05,
+      "loss": 0.4512,
+      "step": 1136
+    },
+    {
+      "epoch": 1.9402730375426622,
+      "grad_norm": 0.37380802533576596,
+      "learning_rate": 3.0713316853902296e-05,
+      "loss": 0.4411,
+      "step": 1137
+    },
+    {
+      "epoch": 1.9419795221843004,
+      "grad_norm": 0.4066743908256873,
+      "learning_rate": 3.069318900381019e-05,
+      "loss": 0.4259,
+      "step": 1138
+    },
+    {
+      "epoch": 1.9436860068259385,
+      "grad_norm": 0.39421066591401493,
+      "learning_rate": 3.0673045976692095e-05,
+      "loss": 0.4505,
+      "step": 1139
+    },
+    {
+      "epoch": 1.9453924914675769,
+      "grad_norm": 0.4285964084137661,
+      "learning_rate": 3.0652887801137365e-05,
+      "loss": 0.6663,
+      "step": 1140
+    },
+    {
+      "epoch": 1.947098976109215,
+      "grad_norm": 0.39704888749269823,
+      "learning_rate": 3.063271450575685e-05,
+      "loss": 0.4357,
+      "step": 1141
+    },
+    {
+      "epoch": 1.9488054607508531,
+      "grad_norm": 0.40234313660149934,
+      "learning_rate": 3.061252611918283e-05,
+      "loss": 0.4256,
+      "step": 1142
+    },
+    {
+      "epoch": 1.9505119453924915,
+      "grad_norm": 0.39051818078690365,
+      "learning_rate": 3.0592322670069044e-05,
+      "loss": 0.4429,
+      "step": 1143
+    },
+    {
+      "epoch": 1.9522184300341296,
+      "grad_norm": 0.4453280616407959,
+      "learning_rate": 3.05721041870906e-05,
+      "loss": 0.4346,
+      "step": 1144
+    },
+    {
+      "epoch": 1.9539249146757678,
+      "grad_norm": 0.39418177898309475,
+      "learning_rate": 3.055187069894392e-05,
+      "loss": 0.4532,
+      "step": 1145
+    },
+    {
+      "epoch": 1.9556313993174061,
+      "grad_norm": 0.40025147717303106,
+      "learning_rate": 3.0531622234346747e-05,
+      "loss": 0.444,
+      "step": 1146
+    },
+    {
+      "epoch": 1.9573378839590445,
+      "grad_norm": 0.37661192218548234,
+      "learning_rate": 3.0511358822038075e-05,
+      "loss": 0.4127,
+      "step": 1147
+    },
+    {
+      "epoch": 1.9590443686006824,
+      "grad_norm": 0.38856559022753606,
+      "learning_rate": 3.0491080490778105e-05,
+      "loss": 0.4829,
+      "step": 1148
+    },
+    {
+      "epoch": 1.9607508532423208,
+      "grad_norm": 0.3888783912215676,
+      "learning_rate": 3.0470787269348218e-05,
+      "loss": 0.4567,
+      "step": 1149
+    },
+    {
+      "epoch": 1.9624573378839592,
+      "grad_norm": 0.41945768107269354,
+      "learning_rate": 3.0450479186550948e-05,
+      "loss": 0.451,
+      "step": 1150
+    },
+    {
+      "epoch": 1.9641638225255973,
+      "grad_norm": 0.4513414837155506,
+      "learning_rate": 3.043015627120989e-05,
+      "loss": 0.4557,
+      "step": 1151
+    },
+    {
+      "epoch": 1.9658703071672354,
+      "grad_norm": 0.3882336660858084,
+      "learning_rate": 3.04098185521697e-05,
+      "loss": 0.4433,
+      "step": 1152
+    },
+    {
+      "epoch": 1.9675767918088738,
+      "grad_norm": 0.4117897053320961,
+      "learning_rate": 3.038946605829606e-05,
+      "loss": 0.427,
+      "step": 1153
+    },
+    {
+      "epoch": 1.969283276450512,
+      "grad_norm": 0.39225920778552686,
+      "learning_rate": 3.0369098818475612e-05,
+      "loss": 0.4669,
+      "step": 1154
+    },
+    {
+      "epoch": 1.97098976109215,
+      "grad_norm": 0.3642477857528382,
+      "learning_rate": 3.0348716861615917e-05,
+      "loss": 0.4248,
+      "step": 1155
+    },
+    {
+      "epoch": 1.9726962457337884,
+      "grad_norm": 0.35814160996825456,
+      "learning_rate": 3.032832021664544e-05,
+      "loss": 0.4495,
+      "step": 1156
+    },
+    {
+      "epoch": 1.9744027303754266,
+      "grad_norm": 0.3724344732843097,
+      "learning_rate": 3.0307908912513507e-05,
+      "loss": 0.4421,
+      "step": 1157
+    },
+    {
+      "epoch": 1.9761092150170647,
+      "grad_norm": 0.3981452207001675,
+      "learning_rate": 3.0287482978190207e-05,
+      "loss": 0.4369,
+      "step": 1158
+    },
+    {
+      "epoch": 1.977815699658703,
+      "grad_norm": 0.3830660314911436,
+      "learning_rate": 3.0267042442666423e-05,
+      "loss": 0.4401,
+      "step": 1159
+    },
+    {
+      "epoch": 1.9795221843003414,
+      "grad_norm": 0.37721324646779664,
+      "learning_rate": 3.0246587334953772e-05,
+      "loss": 0.3923,
+      "step": 1160
+    },
+    {
+      "epoch": 1.9812286689419796,
+      "grad_norm": 0.37845672699312793,
+      "learning_rate": 3.022611768408451e-05,
+      "loss": 0.4174,
+      "step": 1161
+    },
+    {
+      "epoch": 1.9829351535836177,
+      "grad_norm": 0.43344727294576857,
+      "learning_rate": 3.0205633519111583e-05,
+      "loss": 0.4488,
+      "step": 1162
+    },
+    {
+      "epoch": 1.984641638225256,
+      "grad_norm": 0.4359501540294188,
+      "learning_rate": 3.018513486910852e-05,
+      "loss": 0.4113,
+      "step": 1163
+    },
+    {
+      "epoch": 1.9863481228668942,
+      "grad_norm": 0.35369746565652127,
+      "learning_rate": 3.0164621763169384e-05,
+      "loss": 0.4292,
+      "step": 1164
+    },
+    {
+      "epoch": 1.9880546075085324,
+      "grad_norm": 0.40236101912475597,
+      "learning_rate": 3.0144094230408796e-05,
+      "loss": 0.427,
+      "step": 1165
+    },
+    {
+      "epoch": 1.9897610921501707,
+      "grad_norm": 0.4326007117434534,
+      "learning_rate": 3.012355229996183e-05,
+      "loss": 0.4336,
+      "step": 1166
+    },
+    {
+      "epoch": 1.9914675767918089,
+      "grad_norm": 0.4036010077074162,
+      "learning_rate": 3.0102996000983993e-05,
+      "loss": 0.4203,
+      "step": 1167
+    },
+    {
+      "epoch": 1.993174061433447,
+      "grad_norm": 0.39351860542297634,
+      "learning_rate": 3.0082425362651197e-05,
+      "loss": 0.4511,
+      "step": 1168
+    },
+    {
+      "epoch": 1.9948805460750854,
+      "grad_norm": 0.38401207247759234,
+      "learning_rate": 3.00618404141597e-05,
+      "loss": 0.4477,
+      "step": 1169
+    },
+    {
+      "epoch": 1.9965870307167235,
+      "grad_norm": 0.36355677275483467,
+      "learning_rate": 3.004124118472607e-05,
+      "loss": 0.4213,
+      "step": 1170
+    },
+    {
+      "epoch": 1.9982935153583616,
+      "grad_norm": 0.38159766177942894,
+      "learning_rate": 3.0020627703587154e-05,
+      "loss": 0.4753,
+      "step": 1171
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.43975034858095224,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.4433,
+      "step": 1172
+    },
+    {
+      "epoch": 2.0017064846416384,
+      "grad_norm": 0.457128745992015,
+      "learning_rate": 2.997935810324188e-05,
+      "loss": 0.3122,
+      "step": 1173
+    },
+    {
+      "epoch": 2.0034129692832763,
+      "grad_norm": 0.4139908575686257,
+      "learning_rate": 2.9958702042610176e-05,
+      "loss": 0.3422,
+      "step": 1174
+    },
+    {
+      "epoch": 2.0051194539249146,
+      "grad_norm": 0.6828806786379927,
+      "learning_rate": 2.9938031847422395e-05,
+      "loss": 0.3496,
+      "step": 1175
+    },
+    {
+      "epoch": 2.006825938566553,
+      "grad_norm": 0.4453379166984815,
+      "learning_rate": 2.99173475470161e-05,
+      "loss": 0.3221,
+      "step": 1176
+    },
+    {
+      "epoch": 2.008532423208191,
+      "grad_norm": 0.481091150644822,
+      "learning_rate": 2.9896649170748864e-05,
+      "loss": 0.2986,
+      "step": 1177
+    },
+    {
+      "epoch": 2.0102389078498293,
+      "grad_norm": 0.4273047300781224,
+      "learning_rate": 2.987593674799826e-05,
+      "loss": 0.3229,
+      "step": 1178
+    },
+    {
+      "epoch": 2.0119453924914676,
+      "grad_norm": 0.43016109425591,
+      "learning_rate": 2.985521030816177e-05,
+      "loss": 0.3332,
+      "step": 1179
+    },
+    {
+      "epoch": 2.013651877133106,
+      "grad_norm": 0.43062017675061975,
+      "learning_rate": 2.983446988065679e-05,
+      "loss": 0.3085,
+      "step": 1180
+    },
+    {
+      "epoch": 2.015358361774744,
+      "grad_norm": 0.42147762229991387,
+      "learning_rate": 2.9813715494920568e-05,
+      "loss": 0.3715,
+      "step": 1181
+    },
+    {
+      "epoch": 2.0170648464163823,
+      "grad_norm": 0.411732748179639,
+      "learning_rate": 2.9792947180410146e-05,
+      "loss": 0.3308,
+      "step": 1182
+    },
+    {
+      "epoch": 2.0187713310580206,
+      "grad_norm": 0.43182507847266327,
+      "learning_rate": 2.9772164966602362e-05,
+      "loss": 0.3562,
+      "step": 1183
+    },
+    {
+      "epoch": 2.0204778156996586,
+      "grad_norm": 0.39303469490960924,
+      "learning_rate": 2.9751368882993765e-05,
+      "loss": 0.3442,
+      "step": 1184
+    },
+    {
+      "epoch": 2.022184300341297,
+      "grad_norm": 0.44901601334467767,
+      "learning_rate": 2.9730558959100585e-05,
+      "loss": 0.3258,
+      "step": 1185
+    },
+    {
+      "epoch": 2.0238907849829353,
+      "grad_norm": 0.4234197234589005,
+      "learning_rate": 2.9709735224458703e-05,
+      "loss": 0.3233,
+      "step": 1186
+    },
+    {
+      "epoch": 2.025597269624573,
+      "grad_norm": 0.4145126381242284,
+      "learning_rate": 2.968889770862361e-05,
+      "loss": 0.3139,
+      "step": 1187
+    },
+    {
+      "epoch": 2.0273037542662116,
+      "grad_norm": 0.4650028395470894,
+      "learning_rate": 2.9668046441170338e-05,
+      "loss": 0.3258,
+      "step": 1188
+    },
+    {
+      "epoch": 2.02901023890785,
+      "grad_norm": 0.4138016958262979,
+      "learning_rate": 2.9647181451693456e-05,
+      "loss": 0.3418,
+      "step": 1189
+    },
+    {
+      "epoch": 2.030716723549488,
+      "grad_norm": 0.44527009359826686,
+      "learning_rate": 2.962630276980699e-05,
+      "loss": 0.3341,
+      "step": 1190
+    },
+    {
+      "epoch": 2.032423208191126,
+      "grad_norm": 0.449841389006582,
+      "learning_rate": 2.960541042514443e-05,
+      "loss": 0.3972,
+      "step": 1191
+    },
+    {
+      "epoch": 2.0341296928327646,
+      "grad_norm": 0.4114068718397478,
+      "learning_rate": 2.9584504447358617e-05,
+      "loss": 0.3244,
+      "step": 1192
+    },
+    {
+      "epoch": 2.035836177474403,
+      "grad_norm": 0.44429247596201654,
+      "learning_rate": 2.956358486612177e-05,
+      "loss": 0.3535,
+      "step": 1193
+    },
+    {
+      "epoch": 2.037542662116041,
+      "grad_norm": 0.44617564995683495,
+      "learning_rate": 2.9542651711125413e-05,
+      "loss": 0.2977,
+      "step": 1194
+    },
+    {
+      "epoch": 2.039249146757679,
+      "grad_norm": 0.4096953046125742,
+      "learning_rate": 2.9521705012080326e-05,
+      "loss": 0.3218,
+      "step": 1195
+    },
+    {
+      "epoch": 2.0409556313993176,
+      "grad_norm": 0.3979622530922502,
+      "learning_rate": 2.9500744798716515e-05,
+      "loss": 0.3509,
+      "step": 1196
+    },
+    {
+      "epoch": 2.0426621160409555,
+      "grad_norm": 0.40047495999040483,
+      "learning_rate": 2.947977110078317e-05,
+      "loss": 0.3017,
+      "step": 1197
+    },
+    {
+      "epoch": 2.044368600682594,
+      "grad_norm": 0.3847757055516993,
+      "learning_rate": 2.945878394804863e-05,
+      "loss": 0.3071,
+      "step": 1198
+    },
+    {
+      "epoch": 2.046075085324232,
+      "grad_norm": 0.3600892843081187,
+      "learning_rate": 2.9437783370300302e-05,
+      "loss": 0.3573,
+      "step": 1199
+    },
+    {
+      "epoch": 2.04778156996587,
+      "grad_norm": 0.38126019571864667,
+      "learning_rate": 2.9416769397344685e-05,
+      "loss": 0.3506,
+      "step": 1200
+    },
+    {
+      "epoch": 2.0494880546075085,
+      "grad_norm": 0.3554100311173552,
+      "learning_rate": 2.939574205900725e-05,
+      "loss": 0.3072,
+      "step": 1201
+    },
+    {
+      "epoch": 2.051194539249147,
+      "grad_norm": 0.38334979368635375,
+      "learning_rate": 2.9374701385132472e-05,
+      "loss": 0.3259,
+      "step": 1202
+    },
+    {
+      "epoch": 2.0529010238907848,
+      "grad_norm": 0.3703136644266881,
+      "learning_rate": 2.935364740558373e-05,
+      "loss": 0.315,
+      "step": 1203
+    },
+    {
+      "epoch": 2.054607508532423,
+      "grad_norm": 0.40500653297603784,
+      "learning_rate": 2.93325801502433e-05,
+      "loss": 0.3408,
+      "step": 1204
+    },
+    {
+      "epoch": 2.0563139931740615,
+      "grad_norm": 0.4062486109717856,
+      "learning_rate": 2.9311499649012304e-05,
+      "loss": 0.2987,
+      "step": 1205
+    },
+    {
+      "epoch": 2.0580204778157,
+      "grad_norm": 0.3616658692205193,
+      "learning_rate": 2.929040593181065e-05,
+      "loss": 0.3149,
+      "step": 1206
+    },
+    {
+      "epoch": 2.0597269624573378,
+      "grad_norm": 0.3848905374239776,
+      "learning_rate": 2.9269299028577016e-05,
+      "loss": 0.3253,
+      "step": 1207
+    },
+    {
+      "epoch": 2.061433447098976,
+      "grad_norm": 0.4298981718447302,
+      "learning_rate": 2.924817896926879e-05,
+      "loss": 0.3256,
+      "step": 1208
+    },
+    {
+      "epoch": 2.0631399317406145,
+      "grad_norm": 0.3534847095397029,
+      "learning_rate": 2.9227045783862026e-05,
+      "loss": 0.3597,
+      "step": 1209
+    },
+    {
+      "epoch": 2.0648464163822524,
+      "grad_norm": 0.33638127804053347,
+      "learning_rate": 2.9205899502351427e-05,
+      "loss": 0.3273,
+      "step": 1210
+    },
+    {
+      "epoch": 2.0665529010238908,
+      "grad_norm": 0.3855231208992533,
+      "learning_rate": 2.9184740154750265e-05,
+      "loss": 0.3138,
+      "step": 1211
+    },
+    {
+      "epoch": 2.068259385665529,
+      "grad_norm": 0.381774736275018,
+      "learning_rate": 2.9163567771090368e-05,
+      "loss": 0.3043,
+      "step": 1212
+    },
+    {
+      "epoch": 2.069965870307167,
+      "grad_norm": 0.37464619845398417,
+      "learning_rate": 2.9142382381422058e-05,
+      "loss": 0.316,
+      "step": 1213
+    },
+    {
+      "epoch": 2.0716723549488054,
+      "grad_norm": 0.4126972583021004,
+      "learning_rate": 2.912118401581412e-05,
+      "loss": 0.2867,
+      "step": 1214
+    },
+    {
+      "epoch": 2.073378839590444,
+      "grad_norm": 0.4314536077963715,
+      "learning_rate": 2.9099972704353763e-05,
+      "loss": 0.327,
+      "step": 1215
+    },
+    {
+      "epoch": 2.0750853242320817,
+      "grad_norm": 0.36047470235701123,
+      "learning_rate": 2.9078748477146552e-05,
+      "loss": 0.3445,
+      "step": 1216
+    },
+    {
+      "epoch": 2.07679180887372,
+      "grad_norm": 0.4301966987859649,
+      "learning_rate": 2.905751136431641e-05,
+      "loss": 0.3275,
+      "step": 1217
+    },
+    {
+      "epoch": 2.0784982935153584,
+      "grad_norm": 0.40268655007936277,
+      "learning_rate": 2.9036261396005526e-05,
+      "loss": 0.3288,
+      "step": 1218
+    },
+    {
+      "epoch": 2.080204778156997,
+      "grad_norm": 0.3992363790721564,
+      "learning_rate": 2.9014998602374345e-05,
+      "loss": 0.363,
+      "step": 1219
+    },
+    {
+      "epoch": 2.0819112627986347,
+      "grad_norm": 0.3883051261851283,
+      "learning_rate": 2.899372301360152e-05,
+      "loss": 0.3166,
+      "step": 1220
+    },
+    {
+      "epoch": 2.083617747440273,
+      "grad_norm": 0.4050737657795587,
+      "learning_rate": 2.8972434659883847e-05,
+      "loss": 0.3298,
+      "step": 1221
+    },
+    {
+      "epoch": 2.0853242320819114,
+      "grad_norm": 0.3984900692632608,
+      "learning_rate": 2.8951133571436255e-05,
+      "loss": 0.3272,
+      "step": 1222
+    },
+    {
+      "epoch": 2.0870307167235493,
+      "grad_norm": 0.35897624852421484,
+      "learning_rate": 2.8929819778491736e-05,
+      "loss": 0.3574,
+      "step": 1223
+    },
+    {
+      "epoch": 2.0887372013651877,
+      "grad_norm": 0.37441217352575806,
+      "learning_rate": 2.8908493311301336e-05,
+      "loss": 0.3392,
+      "step": 1224
+    },
+    {
+      "epoch": 2.090443686006826,
+      "grad_norm": 0.3806877568546792,
+      "learning_rate": 2.8887154200134066e-05,
+      "loss": 0.3405,
+      "step": 1225
+    },
+    {
+      "epoch": 2.092150170648464,
+      "grad_norm": 0.3635010398502254,
+      "learning_rate": 2.8865802475276888e-05,
+      "loss": 0.3308,
+      "step": 1226
+    },
+    {
+      "epoch": 2.0938566552901023,
+      "grad_norm": 0.393071249205216,
+      "learning_rate": 2.8844438167034675e-05,
+      "loss": 0.3389,
+      "step": 1227
+    },
+    {
+      "epoch": 2.0955631399317407,
+      "grad_norm": 0.3623276745352252,
+      "learning_rate": 2.8823061305730154e-05,
+      "loss": 0.3172,
+      "step": 1228
+    },
+    {
+      "epoch": 2.0972696245733786,
+      "grad_norm": 0.3840492877615427,
+      "learning_rate": 2.8801671921703875e-05,
+      "loss": 0.385,
+      "step": 1229
+    },
+    {
+      "epoch": 2.098976109215017,
+      "grad_norm": 0.35674517742092077,
+      "learning_rate": 2.878027004531414e-05,
+      "loss": 0.3354,
+      "step": 1230
+    },
+    {
+      "epoch": 2.1006825938566553,
+      "grad_norm": 0.3724437131382105,
+      "learning_rate": 2.8758855706937015e-05,
+      "loss": 0.3244,
+      "step": 1231
+    },
+    {
+      "epoch": 2.1023890784982937,
+      "grad_norm": 0.42648734108694186,
+      "learning_rate": 2.873742893696623e-05,
+      "loss": 0.3322,
+      "step": 1232
+    },
+    {
+      "epoch": 2.1040955631399316,
+      "grad_norm": 0.3731360928602656,
+      "learning_rate": 2.871598976581317e-05,
+      "loss": 0.2982,
+      "step": 1233
+    },
+    {
+      "epoch": 2.10580204778157,
+      "grad_norm": 0.40421291050408936,
+      "learning_rate": 2.8694538223906812e-05,
+      "loss": 0.354,
+      "step": 1234
+    },
+    {
+      "epoch": 2.1075085324232083,
+      "grad_norm": 0.3816196487587043,
+      "learning_rate": 2.8673074341693698e-05,
+      "loss": 0.3645,
+      "step": 1235
+    },
+    {
+      "epoch": 2.1092150170648463,
+      "grad_norm": 0.37710300585084694,
+      "learning_rate": 2.865159814963788e-05,
+      "loss": 0.3484,
+      "step": 1236
+    },
+    {
+      "epoch": 2.1109215017064846,
+      "grad_norm": 0.43153582922337635,
+      "learning_rate": 2.863010967822089e-05,
+      "loss": 0.3541,
+      "step": 1237
+    },
+    {
+      "epoch": 2.112627986348123,
+      "grad_norm": 0.38541541269540536,
+      "learning_rate": 2.8608608957941677e-05,
+      "loss": 0.3148,
+      "step": 1238
+    },
+    {
+      "epoch": 2.114334470989761,
+      "grad_norm": 0.3976182070732549,
+      "learning_rate": 2.8587096019316588e-05,
+      "loss": 0.3631,
+      "step": 1239
+    },
+    {
+      "epoch": 2.1160409556313993,
+      "grad_norm": 0.4170776432996827,
+      "learning_rate": 2.8565570892879308e-05,
+      "loss": 0.3037,
+      "step": 1240
+    },
+    {
+      "epoch": 2.1177474402730376,
+      "grad_norm": 0.4067765892230644,
+      "learning_rate": 2.8544033609180797e-05,
+      "loss": 0.3303,
+      "step": 1241
+    },
+    {
+      "epoch": 2.1194539249146755,
+      "grad_norm": 0.35639261403736644,
+      "learning_rate": 2.8522484198789308e-05,
+      "loss": 0.3323,
+      "step": 1242
+    },
+    {
+      "epoch": 2.121160409556314,
+      "grad_norm": 0.39487236200655806,
+      "learning_rate": 2.8500922692290284e-05,
+      "loss": 0.3783,
+      "step": 1243
+    },
+    {
+      "epoch": 2.1228668941979523,
+      "grad_norm": 0.35821615465041945,
+      "learning_rate": 2.8479349120286337e-05,
+      "loss": 0.3407,
+      "step": 1244
+    },
+    {
+      "epoch": 2.1245733788395906,
+      "grad_norm": 0.35341164274325937,
+      "learning_rate": 2.8457763513397206e-05,
+      "loss": 0.3313,
+      "step": 1245
+    },
+    {
+      "epoch": 2.1262798634812285,
+      "grad_norm": 0.3957886110355856,
+      "learning_rate": 2.8436165902259717e-05,
+      "loss": 0.38,
+      "step": 1246
+    },
+    {
+      "epoch": 2.127986348122867,
+      "grad_norm": 0.36024142111095514,
+      "learning_rate": 2.8414556317527722e-05,
+      "loss": 0.3296,
+      "step": 1247
+    },
+    {
+      "epoch": 2.1296928327645053,
+      "grad_norm": 0.39155019349124676,
+      "learning_rate": 2.839293478987208e-05,
+      "loss": 0.3411,
+      "step": 1248
+    },
+    {
+      "epoch": 2.131399317406143,
+      "grad_norm": 0.39297329760609745,
+      "learning_rate": 2.8371301349980593e-05,
+      "loss": 0.3202,
+      "step": 1249
+    },
+    {
+      "epoch": 2.1331058020477816,
+      "grad_norm": 0.41730207596559865,
+      "learning_rate": 2.834965602855797e-05,
+      "loss": 0.3689,
+      "step": 1250
+    },
+    {
+      "epoch": 2.13481228668942,
+      "grad_norm": 0.3628482280926389,
+      "learning_rate": 2.8327998856325788e-05,
+      "loss": 0.345,
+      "step": 1251
+    },
+    {
+      "epoch": 2.136518771331058,
+      "grad_norm": 0.3744238919178526,
+      "learning_rate": 2.8306329864022446e-05,
+      "loss": 0.3209,
+      "step": 1252
+    },
+    {
+      "epoch": 2.138225255972696,
+      "grad_norm": 0.3947466710094417,
+      "learning_rate": 2.8284649082403107e-05,
+      "loss": 0.345,
+      "step": 1253
+    },
+    {
+      "epoch": 2.1399317406143346,
+      "grad_norm": 0.38032210318900517,
+      "learning_rate": 2.8262956542239678e-05,
+      "loss": 0.3365,
+      "step": 1254
+    },
+    {
+      "epoch": 2.1416382252559725,
+      "grad_norm": 0.34228417774558967,
+      "learning_rate": 2.8241252274320753e-05,
+      "loss": 0.3256,
+      "step": 1255
+    },
+    {
+      "epoch": 2.143344709897611,
+      "grad_norm": 0.42236716597469687,
+      "learning_rate": 2.8219536309451566e-05,
+      "loss": 0.3158,
+      "step": 1256
+    },
+    {
+      "epoch": 2.145051194539249,
+      "grad_norm": 0.3638834569248063,
+      "learning_rate": 2.8197808678453965e-05,
+      "loss": 0.3294,
+      "step": 1257
+    },
+    {
+      "epoch": 2.1467576791808876,
+      "grad_norm": 0.38632598483966657,
+      "learning_rate": 2.8176069412166345e-05,
+      "loss": 0.3295,
+      "step": 1258
+    },
+    {
+      "epoch": 2.1484641638225255,
+      "grad_norm": 0.4170935336781794,
+      "learning_rate": 2.815431854144362e-05,
+      "loss": 0.3097,
+      "step": 1259
+    },
+    {
+      "epoch": 2.150170648464164,
+      "grad_norm": 0.3525829213286291,
+      "learning_rate": 2.813255609715717e-05,
+      "loss": 0.353,
+      "step": 1260
+    },
+    {
+      "epoch": 2.151877133105802,
+      "grad_norm": 0.3854869447902711,
+      "learning_rate": 2.81107821101948e-05,
+      "loss": 0.3359,
+      "step": 1261
+    },
+    {
+      "epoch": 2.15358361774744,
+      "grad_norm": 0.3413451536237944,
+      "learning_rate": 2.808899661146072e-05,
+      "loss": 0.2912,
+      "step": 1262
+    },
+    {
+      "epoch": 2.1552901023890785,
+      "grad_norm": 0.3819099640985915,
+      "learning_rate": 2.806719963187543e-05,
+      "loss": 0.3383,
+      "step": 1263
+    },
+    {
+      "epoch": 2.156996587030717,
+      "grad_norm": 0.37496052907993016,
+      "learning_rate": 2.804539120237578e-05,
+      "loss": 0.3249,
+      "step": 1264
+    },
+    {
+      "epoch": 2.1587030716723548,
+      "grad_norm": 0.3831512033150945,
+      "learning_rate": 2.8023571353914846e-05,
+      "loss": 0.319,
+      "step": 1265
+    },
+    {
+      "epoch": 2.160409556313993,
+      "grad_norm": 0.396726145433138,
+      "learning_rate": 2.80017401174619e-05,
+      "loss": 0.3739,
+      "step": 1266
+    },
+    {
+      "epoch": 2.1621160409556315,
+      "grad_norm": 0.40561507479220216,
+      "learning_rate": 2.79798975240024e-05,
+      "loss": 0.3481,
+      "step": 1267
+    },
+    {
+      "epoch": 2.1638225255972694,
+      "grad_norm": 0.3275781604015692,
+      "learning_rate": 2.795804360453791e-05,
+      "loss": 0.2998,
+      "step": 1268
+    },
+    {
+      "epoch": 2.1655290102389078,
+      "grad_norm": 0.3522773371443922,
+      "learning_rate": 2.793617839008606e-05,
+      "loss": 0.3093,
+      "step": 1269
+    },
+    {
+      "epoch": 2.167235494880546,
+      "grad_norm": 0.3887568155863699,
+      "learning_rate": 2.7914301911680535e-05,
+      "loss": 0.2977,
+      "step": 1270
+    },
+    {
+      "epoch": 2.1689419795221845,
+      "grad_norm": 0.3648550969703528,
+      "learning_rate": 2.7892414200371e-05,
+      "loss": 0.3187,
+      "step": 1271
+    },
+    {
+      "epoch": 2.1706484641638224,
+      "grad_norm": 0.43367590800895783,
+      "learning_rate": 2.7870515287223043e-05,
+      "loss": 0.3514,
+      "step": 1272
+    },
+    {
+      "epoch": 2.1723549488054608,
+      "grad_norm": 0.417822649774299,
+      "learning_rate": 2.7848605203318177e-05,
+      "loss": 0.3289,
+      "step": 1273
+    },
+    {
+      "epoch": 2.174061433447099,
+      "grad_norm": 0.3626889608693872,
+      "learning_rate": 2.7826683979753753e-05,
+      "loss": 0.4058,
+      "step": 1274
+    },
+    {
+      "epoch": 2.175767918088737,
+      "grad_norm": 0.3782938440184352,
+      "learning_rate": 2.780475164764294e-05,
+      "loss": 0.3226,
+      "step": 1275
+    },
+    {
+      "epoch": 2.1774744027303754,
+      "grad_norm": 0.40971833875259145,
+      "learning_rate": 2.778280823811467e-05,
+      "loss": 0.3582,
+      "step": 1276
+    },
+    {
+      "epoch": 2.1791808873720138,
+      "grad_norm": 0.3757058724125517,
+      "learning_rate": 2.7760853782313598e-05,
+      "loss": 0.3343,
+      "step": 1277
+    },
+    {
+      "epoch": 2.1808873720136517,
+      "grad_norm": 0.37586921864323336,
+      "learning_rate": 2.7738888311400066e-05,
+      "loss": 0.3334,
+      "step": 1278
+    },
+    {
+      "epoch": 2.18259385665529,
+      "grad_norm": 0.336446479477883,
+      "learning_rate": 2.7716911856550036e-05,
+      "loss": 0.3061,
+      "step": 1279
+    },
+    {
+      "epoch": 2.1843003412969284,
+      "grad_norm": 0.4029523350484026,
+      "learning_rate": 2.7694924448955072e-05,
+      "loss": 0.3611,
+      "step": 1280
+    },
+    {
+      "epoch": 2.1860068259385663,
+      "grad_norm": 0.39868971446555357,
+      "learning_rate": 2.7672926119822272e-05,
+      "loss": 0.3715,
+      "step": 1281
+    },
+    {
+      "epoch": 2.1877133105802047,
+      "grad_norm": 0.3642945128867239,
+      "learning_rate": 2.7650916900374238e-05,
+      "loss": 0.3316,
+      "step": 1282
+    },
+    {
+      "epoch": 2.189419795221843,
+      "grad_norm": 0.38044114907318405,
+      "learning_rate": 2.762889682184904e-05,
+      "loss": 0.3496,
+      "step": 1283
+    },
+    {
+      "epoch": 2.1911262798634814,
+      "grad_norm": 0.3756040633783301,
+      "learning_rate": 2.7606865915500148e-05,
+      "loss": 0.3428,
+      "step": 1284
+    },
+    {
+      "epoch": 2.1928327645051193,
+      "grad_norm": 0.3766036687251031,
+      "learning_rate": 2.7584824212596396e-05,
+      "loss": 0.33,
+      "step": 1285
+    },
+    {
+      "epoch": 2.1945392491467577,
+      "grad_norm": 0.3793415403883485,
+      "learning_rate": 2.7562771744421974e-05,
+      "loss": 0.3257,
+      "step": 1286
+    },
+    {
+      "epoch": 2.196245733788396,
+      "grad_norm": 0.38676018867366924,
+      "learning_rate": 2.7540708542276297e-05,
+      "loss": 0.3334,
+      "step": 1287
+    },
+    {
+      "epoch": 2.197952218430034,
+      "grad_norm": 0.34957567057729205,
+      "learning_rate": 2.7518634637474063e-05,
+      "loss": 0.3429,
+      "step": 1288
+    },
+    {
+      "epoch": 2.1996587030716723,
+      "grad_norm": 0.3686311749994957,
+      "learning_rate": 2.7496550061345138e-05,
+      "loss": 0.3222,
+      "step": 1289
+    },
+    {
+      "epoch": 2.2013651877133107,
+      "grad_norm": 0.35075494112369326,
+      "learning_rate": 2.7474454845234534e-05,
+      "loss": 0.3607,
+      "step": 1290
+    },
+    {
+      "epoch": 2.2030716723549486,
+      "grad_norm": 0.3490460231118274,
+      "learning_rate": 2.7452349020502377e-05,
+      "loss": 0.3203,
+      "step": 1291
+    },
+    {
+      "epoch": 2.204778156996587,
+      "grad_norm": 0.4159082349899918,
+      "learning_rate": 2.7430232618523846e-05,
+      "loss": 0.3329,
+      "step": 1292
+    },
+    {
+      "epoch": 2.2064846416382253,
+      "grad_norm": 0.3935325819376942,
+      "learning_rate": 2.7408105670689114e-05,
+      "loss": 0.3626,
+      "step": 1293
+    },
+    {
+      "epoch": 2.2081911262798632,
+      "grad_norm": 0.356625458585148,
+      "learning_rate": 2.7385968208403343e-05,
+      "loss": 0.3621,
+      "step": 1294
+    },
+    {
+      "epoch": 2.2098976109215016,
+      "grad_norm": 0.3823241707511623,
+      "learning_rate": 2.7363820263086616e-05,
+      "loss": 0.3767,
+      "step": 1295
+    },
+    {
+      "epoch": 2.21160409556314,
+      "grad_norm": 0.35899165701810964,
+      "learning_rate": 2.7341661866173882e-05,
+      "loss": 0.321,
+      "step": 1296
+    },
+    {
+      "epoch": 2.2133105802047783,
+      "grad_norm": 0.3291105534040846,
+      "learning_rate": 2.7319493049114937e-05,
+      "loss": 0.2973,
+      "step": 1297
+    },
+    {
+      "epoch": 2.2150170648464163,
+      "grad_norm": 0.41428730038665074,
+      "learning_rate": 2.7297313843374364e-05,
+      "loss": 0.3517,
+      "step": 1298
+    },
+    {
+      "epoch": 2.2167235494880546,
+      "grad_norm": 0.6282577075784936,
+      "learning_rate": 2.7275124280431492e-05,
+      "loss": 0.3486,
+      "step": 1299
+    },
+    {
+      "epoch": 2.218430034129693,
+      "grad_norm": 0.3819749469181485,
+      "learning_rate": 2.7252924391780338e-05,
+      "loss": 0.358,
+      "step": 1300
+    },
+    {
+      "epoch": 2.220136518771331,
+      "grad_norm": 0.3734912146512999,
+      "learning_rate": 2.723071420892959e-05,
+      "loss": 0.3751,
+      "step": 1301
+    },
+    {
+      "epoch": 2.2218430034129693,
+      "grad_norm": 0.4060889025671746,
+      "learning_rate": 2.7208493763402538e-05,
+      "loss": 0.3393,
+      "step": 1302
+    },
+    {
+      "epoch": 2.2235494880546076,
+      "grad_norm": 0.3542469612727659,
+      "learning_rate": 2.7186263086737034e-05,
+      "loss": 0.3743,
+      "step": 1303
+    },
+    {
+      "epoch": 2.2252559726962455,
+      "grad_norm": 0.3690400547180425,
+      "learning_rate": 2.7164022210485468e-05,
+      "loss": 0.3269,
+      "step": 1304
+    },
+    {
+      "epoch": 2.226962457337884,
+      "grad_norm": 0.4127396663588053,
+      "learning_rate": 2.7141771166214694e-05,
+      "loss": 0.3322,
+      "step": 1305
+    },
+    {
+      "epoch": 2.2286689419795223,
+      "grad_norm": 0.3606825518166459,
+      "learning_rate": 2.7119509985505997e-05,
+      "loss": 0.3413,
+      "step": 1306
+    },
+    {
+      "epoch": 2.2303754266211606,
+      "grad_norm": 0.37041989688669297,
+      "learning_rate": 2.709723869995505e-05,
+      "loss": 0.3264,
+      "step": 1307
+    },
+    {
+      "epoch": 2.2320819112627985,
+      "grad_norm": 0.3960536252124662,
+      "learning_rate": 2.7074957341171874e-05,
+      "loss": 0.3375,
+      "step": 1308
+    },
+    {
+      "epoch": 2.233788395904437,
+      "grad_norm": 0.3818893032667079,
+      "learning_rate": 2.705266594078078e-05,
+      "loss": 0.368,
+      "step": 1309
+    },
+    {
+      "epoch": 2.2354948805460753,
+      "grad_norm": 0.3664428592188226,
+      "learning_rate": 2.703036453042033e-05,
+      "loss": 0.3223,
+      "step": 1310
+    },
+    {
+      "epoch": 2.237201365187713,
+      "grad_norm": 0.3561441401803114,
+      "learning_rate": 2.7008053141743298e-05,
+      "loss": 0.3245,
+      "step": 1311
+    },
+    {
+      "epoch": 2.2389078498293515,
+      "grad_norm": 0.3485024959150887,
+      "learning_rate": 2.6985731806416623e-05,
+      "loss": 0.3191,
+      "step": 1312
+    },
+    {
+      "epoch": 2.24061433447099,
+      "grad_norm": 0.3809419114703872,
+      "learning_rate": 2.6963400556121362e-05,
+      "loss": 0.3182,
+      "step": 1313
+    },
+    {
+      "epoch": 2.242320819112628,
+      "grad_norm": 0.41760905832690415,
+      "learning_rate": 2.6941059422552635e-05,
+      "loss": 0.3032,
+      "step": 1314
+    },
+    {
+      "epoch": 2.244027303754266,
+      "grad_norm": 0.39218141888484626,
+      "learning_rate": 2.691870843741959e-05,
+      "loss": 0.335,
+      "step": 1315
+    },
+    {
+      "epoch": 2.2457337883959045,
+      "grad_norm": 0.3728659579751814,
+      "learning_rate": 2.689634763244537e-05,
+      "loss": 0.3345,
+      "step": 1316
+    },
+    {
+      "epoch": 2.2474402730375425,
+      "grad_norm": 0.37450328090186463,
+      "learning_rate": 2.687397703936704e-05,
+      "loss": 0.3302,
+      "step": 1317
+    },
+    {
+      "epoch": 2.249146757679181,
+      "grad_norm": 0.36789679345836473,
+      "learning_rate": 2.6851596689935574e-05,
+      "loss": 0.333,
+      "step": 1318
+    },
+    {
+      "epoch": 2.250853242320819,
+      "grad_norm": 0.390239075495621,
+      "learning_rate": 2.682920661591578e-05,
+      "loss": 0.3322,
+      "step": 1319
+    },
+    {
+      "epoch": 2.252559726962457,
+      "grad_norm": 0.39697227505570915,
+      "learning_rate": 2.6806806849086276e-05,
+      "loss": 0.3276,
+      "step": 1320
+    },
+    {
+      "epoch": 2.2542662116040955,
+      "grad_norm": 0.41570925076951337,
+      "learning_rate": 2.678439742123943e-05,
+      "loss": 0.3717,
+      "step": 1321
+    },
+    {
+      "epoch": 2.255972696245734,
+      "grad_norm": 0.38928174172708657,
+      "learning_rate": 2.6761978364181323e-05,
+      "loss": 0.3765,
+      "step": 1322
+    },
+    {
+      "epoch": 2.257679180887372,
+      "grad_norm": 0.37549967121944744,
+      "learning_rate": 2.673954970973172e-05,
+      "loss": 0.3589,
+      "step": 1323
+    },
+    {
+      "epoch": 2.25938566552901,
+      "grad_norm": 0.37697727591598484,
+      "learning_rate": 2.671711148972398e-05,
+      "loss": 0.3661,
+      "step": 1324
+    },
+    {
+      "epoch": 2.2610921501706485,
+      "grad_norm": 0.38798438711320105,
+      "learning_rate": 2.6694663736005054e-05,
+      "loss": 0.3421,
+      "step": 1325
+    },
+    {
+      "epoch": 2.262798634812287,
+      "grad_norm": 0.3406417472710425,
+      "learning_rate": 2.6672206480435433e-05,
+      "loss": 0.3245,
+      "step": 1326
+    },
+    {
+      "epoch": 2.2645051194539247,
+      "grad_norm": 0.3639364853861221,
+      "learning_rate": 2.664973975488907e-05,
+      "loss": 0.36,
+      "step": 1327
+    },
+    {
+      "epoch": 2.266211604095563,
+      "grad_norm": 0.38183582673797606,
+      "learning_rate": 2.6627263591253382e-05,
+      "loss": 0.3161,
+      "step": 1328
+    },
+    {
+      "epoch": 2.2679180887372015,
+      "grad_norm": 0.3596190854379249,
+      "learning_rate": 2.6604778021429164e-05,
+      "loss": 0.3381,
+      "step": 1329
+    },
+    {
+      "epoch": 2.26962457337884,
+      "grad_norm": 0.380270558721218,
+      "learning_rate": 2.6582283077330582e-05,
+      "loss": 0.3403,
+      "step": 1330
+    },
+    {
+      "epoch": 2.2713310580204777,
+      "grad_norm": 0.3745524915536513,
+      "learning_rate": 2.6559778790885084e-05,
+      "loss": 0.3428,
+      "step": 1331
+    },
+    {
+      "epoch": 2.273037542662116,
+      "grad_norm": 0.35135954321568064,
+      "learning_rate": 2.653726519403339e-05,
+      "loss": 0.3196,
+      "step": 1332
+    },
+    {
+      "epoch": 2.274744027303754,
+      "grad_norm": 0.327555590634362,
+      "learning_rate": 2.6514742318729445e-05,
+      "loss": 0.351,
+      "step": 1333
+    },
+    {
+      "epoch": 2.2764505119453924,
+      "grad_norm": 0.36208360425157576,
+      "learning_rate": 2.649221019694033e-05,
+      "loss": 0.3515,
+      "step": 1334
+    },
+    {
+      "epoch": 2.2781569965870307,
+      "grad_norm": 0.3848771308126638,
+      "learning_rate": 2.646966886064629e-05,
+      "loss": 0.3119,
+      "step": 1335
+    },
+    {
+      "epoch": 2.279863481228669,
+      "grad_norm": 0.3621027257971774,
+      "learning_rate": 2.644711834184062e-05,
+      "loss": 0.3243,
+      "step": 1336
+    },
+    {
+      "epoch": 2.281569965870307,
+      "grad_norm": 0.39052460845925924,
+      "learning_rate": 2.6424558672529648e-05,
+      "loss": 0.3249,
+      "step": 1337
+    },
+    {
+      "epoch": 2.2832764505119454,
+      "grad_norm": 0.362282332219868,
+      "learning_rate": 2.6401989884732716e-05,
+      "loss": 0.3268,
+      "step": 1338
+    },
+    {
+      "epoch": 2.2849829351535837,
+      "grad_norm": 0.35499716624898975,
+      "learning_rate": 2.6379412010482087e-05,
+      "loss": 0.3052,
+      "step": 1339
+    },
+    {
+      "epoch": 2.2866894197952217,
+      "grad_norm": 0.38850349063964607,
+      "learning_rate": 2.635682508182291e-05,
+      "loss": 0.3151,
+      "step": 1340
+    },
+    {
+      "epoch": 2.28839590443686,
+      "grad_norm": 0.5053926075264994,
+      "learning_rate": 2.6334229130813212e-05,
+      "loss": 0.3476,
+      "step": 1341
+    },
+    {
+      "epoch": 2.2901023890784984,
+      "grad_norm": 0.445190693428244,
+      "learning_rate": 2.6311624189523818e-05,
+      "loss": 0.357,
+      "step": 1342
+    },
+    {
+      "epoch": 2.2918088737201368,
+      "grad_norm": 0.37232420913577263,
+      "learning_rate": 2.6289010290038287e-05,
+      "loss": 0.3304,
+      "step": 1343
+    },
+    {
+      "epoch": 2.2935153583617747,
+      "grad_norm": 0.4092725423806077,
+      "learning_rate": 2.6266387464452926e-05,
+      "loss": 0.3307,
+      "step": 1344
+    },
+    {
+      "epoch": 2.295221843003413,
+      "grad_norm": 0.44424690257197885,
+      "learning_rate": 2.6243755744876706e-05,
+      "loss": 0.3268,
+      "step": 1345
+    },
+    {
+      "epoch": 2.296928327645051,
+      "grad_norm": 0.3537529521792447,
+      "learning_rate": 2.62211151634312e-05,
+      "loss": 0.3202,
+      "step": 1346
+    },
+    {
+      "epoch": 2.2986348122866893,
+      "grad_norm": 0.353289356438286,
+      "learning_rate": 2.6198465752250575e-05,
+      "loss": 0.3319,
+      "step": 1347
+    },
+    {
+      "epoch": 2.3003412969283277,
+      "grad_norm": 0.42972592961031836,
+      "learning_rate": 2.6175807543481533e-05,
+      "loss": 0.5392,
+      "step": 1348
+    },
+    {
+      "epoch": 2.302047781569966,
+      "grad_norm": 0.4006275659437903,
+      "learning_rate": 2.615314056928325e-05,
+      "loss": 0.311,
+      "step": 1349
+    },
+    {
+      "epoch": 2.303754266211604,
+      "grad_norm": 0.3513733934426921,
+      "learning_rate": 2.6130464861827355e-05,
+      "loss": 0.2919,
+      "step": 1350
+    },
+    {
+      "epoch": 2.3054607508532423,
+      "grad_norm": 0.35056098323764556,
+      "learning_rate": 2.6107780453297867e-05,
+      "loss": 0.3157,
+      "step": 1351
+    },
+    {
+      "epoch": 2.3071672354948807,
+      "grad_norm": 0.4774862257969973,
+      "learning_rate": 2.6085087375891148e-05,
+      "loss": 0.3239,
+      "step": 1352
+    },
+    {
+      "epoch": 2.3088737201365186,
+      "grad_norm": 0.4243137448825834,
+      "learning_rate": 2.6062385661815883e-05,
+      "loss": 0.3867,
+      "step": 1353
+    },
+    {
+      "epoch": 2.310580204778157,
+      "grad_norm": 0.4005873582549145,
+      "learning_rate": 2.6039675343293e-05,
+      "loss": 0.3492,
+      "step": 1354
+    },
+    {
+      "epoch": 2.3122866894197953,
+      "grad_norm": 0.4082423081612512,
+      "learning_rate": 2.6016956452555634e-05,
+      "loss": 0.513,
+      "step": 1355
+    },
+    {
+      "epoch": 2.3139931740614337,
+      "grad_norm": 0.3992966329238439,
+      "learning_rate": 2.5994229021849098e-05,
+      "loss": 0.352,
+      "step": 1356
+    },
+    {
+      "epoch": 2.3156996587030716,
+      "grad_norm": 0.42991402649117133,
+      "learning_rate": 2.597149308343083e-05,
+      "loss": 0.3406,
+      "step": 1357
+    },
+    {
+      "epoch": 2.31740614334471,
+      "grad_norm": 0.3662392973111248,
+      "learning_rate": 2.5948748669570325e-05,
+      "loss": 0.3365,
+      "step": 1358
+    },
+    {
+      "epoch": 2.319112627986348,
+      "grad_norm": 0.3885234377213803,
+      "learning_rate": 2.5925995812549126e-05,
+      "loss": 0.3511,
+      "step": 1359
+    },
+    {
+      "epoch": 2.3208191126279862,
+      "grad_norm": 0.35696602212521295,
+      "learning_rate": 2.5903234544660755e-05,
+      "loss": 0.2986,
+      "step": 1360
+    },
+    {
+      "epoch": 2.3225255972696246,
+      "grad_norm": 0.40269050230313447,
+      "learning_rate": 2.588046489821066e-05,
+      "loss": 0.3347,
+      "step": 1361
+    },
+    {
+      "epoch": 2.324232081911263,
+      "grad_norm": 0.38282631602250244,
+      "learning_rate": 2.5857686905516195e-05,
+      "loss": 0.3403,
+      "step": 1362
+    },
+    {
+      "epoch": 2.325938566552901,
+      "grad_norm": 0.3630691576269162,
+      "learning_rate": 2.5834900598906557e-05,
+      "loss": 0.2834,
+      "step": 1363
+    },
+    {
+      "epoch": 2.3276450511945392,
+      "grad_norm": 0.3654111923242123,
+      "learning_rate": 2.5812106010722732e-05,
+      "loss": 0.386,
+      "step": 1364
+    },
+    {
+      "epoch": 2.3293515358361776,
+      "grad_norm": 0.3962095056625424,
+      "learning_rate": 2.578930317331747e-05,
+      "loss": 0.4225,
+      "step": 1365
+    },
+    {
+      "epoch": 2.3310580204778155,
+      "grad_norm": 0.4003552112952632,
+      "learning_rate": 2.5766492119055237e-05,
+      "loss": 0.3027,
+      "step": 1366
+    },
+    {
+      "epoch": 2.332764505119454,
+      "grad_norm": 0.3867896985438215,
+      "learning_rate": 2.5743672880312152e-05,
+      "loss": 0.3704,
+      "step": 1367
+    },
+    {
+      "epoch": 2.3344709897610922,
+      "grad_norm": 0.3886665243250387,
+      "learning_rate": 2.5720845489475935e-05,
+      "loss": 0.3538,
+      "step": 1368
+    },
+    {
+      "epoch": 2.3361774744027306,
+      "grad_norm": 0.3986068443722344,
+      "learning_rate": 2.569800997894591e-05,
+      "loss": 0.3632,
+      "step": 1369
+    },
+    {
+      "epoch": 2.3378839590443685,
+      "grad_norm": 0.35079424216905164,
+      "learning_rate": 2.5675166381132895e-05,
+      "loss": 0.3129,
+      "step": 1370
+    },
+    {
+      "epoch": 2.339590443686007,
+      "grad_norm": 0.3548076543694736,
+      "learning_rate": 2.5652314728459207e-05,
+      "loss": 0.3136,
+      "step": 1371
+    },
+    {
+      "epoch": 2.3412969283276452,
+      "grad_norm": 0.38468021466593455,
+      "learning_rate": 2.5629455053358582e-05,
+      "loss": 0.3387,
+      "step": 1372
+    },
+    {
+      "epoch": 2.343003412969283,
+      "grad_norm": 0.39646317065531794,
+      "learning_rate": 2.5606587388276153e-05,
+      "loss": 0.3569,
+      "step": 1373
+    },
+    {
+      "epoch": 2.3447098976109215,
+      "grad_norm": 0.34637816706673646,
+      "learning_rate": 2.558371176566839e-05,
+      "loss": 0.3191,
+      "step": 1374
+    },
+    {
+      "epoch": 2.34641638225256,
+      "grad_norm": 0.3455783144141557,
+      "learning_rate": 2.556082821800304e-05,
+      "loss": 0.3044,
+      "step": 1375
+    },
+    {
+      "epoch": 2.348122866894198,
+      "grad_norm": 0.39109613972413526,
+      "learning_rate": 2.5537936777759137e-05,
+      "loss": 0.3628,
+      "step": 1376
+    },
+    {
+      "epoch": 2.349829351535836,
+      "grad_norm": 0.36307973446643577,
+      "learning_rate": 2.5515037477426865e-05,
+      "loss": 0.2975,
+      "step": 1377
+    },
+    {
+      "epoch": 2.3515358361774745,
+      "grad_norm": 0.3938124454749116,
+      "learning_rate": 2.5492130349507615e-05,
+      "loss": 0.3196,
+      "step": 1378
+    },
+    {
+      "epoch": 2.3532423208191124,
+      "grad_norm": 0.4033717777247042,
+      "learning_rate": 2.546921542651386e-05,
+      "loss": 0.3157,
+      "step": 1379
+    },
+    {
+      "epoch": 2.354948805460751,
+      "grad_norm": 0.353514246210565,
+      "learning_rate": 2.5446292740969137e-05,
+      "loss": 0.3563,
+      "step": 1380
+    },
+    {
+      "epoch": 2.356655290102389,
+      "grad_norm": 0.3874177630384359,
+      "learning_rate": 2.5423362325408012e-05,
+      "loss": 0.342,
+      "step": 1381
+    },
+    {
+      "epoch": 2.3583617747440275,
+      "grad_norm": 0.38665862860940753,
+      "learning_rate": 2.5400424212376016e-05,
+      "loss": 0.3283,
+      "step": 1382
+    },
+    {
+      "epoch": 2.3600682593856654,
+      "grad_norm": 0.3899975436857683,
+      "learning_rate": 2.5377478434429597e-05,
+      "loss": 0.3314,
+      "step": 1383
+    },
+    {
+      "epoch": 2.361774744027304,
+      "grad_norm": 0.39935415423996007,
+      "learning_rate": 2.535452502413609e-05,
+      "loss": 0.3253,
+      "step": 1384
+    },
+    {
+      "epoch": 2.363481228668942,
+      "grad_norm": 0.3971800997252506,
+      "learning_rate": 2.533156401407367e-05,
+      "loss": 0.3645,
+      "step": 1385
+    },
+    {
+      "epoch": 2.36518771331058,
+      "grad_norm": 0.3712698641390973,
+      "learning_rate": 2.5308595436831293e-05,
+      "loss": 0.3369,
+      "step": 1386
+    },
+    {
+      "epoch": 2.3668941979522184,
+      "grad_norm": 0.4292161442786262,
+      "learning_rate": 2.5285619325008642e-05,
+      "loss": 0.3321,
+      "step": 1387
+    },
+    {
+      "epoch": 2.368600682593857,
+      "grad_norm": 0.32124725800414833,
+      "learning_rate": 2.526263571121612e-05,
+      "loss": 0.3016,
+      "step": 1388
+    },
+    {
+      "epoch": 2.3703071672354947,
+      "grad_norm": 0.42628150818210847,
+      "learning_rate": 2.5239644628074753e-05,
+      "loss": 0.3302,
+      "step": 1389
+    },
+    {
+      "epoch": 2.372013651877133,
+      "grad_norm": 0.41628538728162245,
+      "learning_rate": 2.5216646108216178e-05,
+      "loss": 0.3614,
+      "step": 1390
+    },
+    {
+      "epoch": 2.3737201365187715,
+      "grad_norm": 0.3644672453853354,
+      "learning_rate": 2.519364018428259e-05,
+      "loss": 0.3288,
+      "step": 1391
+    },
+    {
+      "epoch": 2.3754266211604094,
+      "grad_norm": 0.4152115385707028,
+      "learning_rate": 2.517062688892669e-05,
+      "loss": 0.3291,
+      "step": 1392
+    },
+    {
+      "epoch": 2.3771331058020477,
+      "grad_norm": 0.3594864619523619,
+      "learning_rate": 2.5147606254811644e-05,
+      "loss": 0.3225,
+      "step": 1393
+    },
+    {
+      "epoch": 2.378839590443686,
+      "grad_norm": 0.38511608243923456,
+      "learning_rate": 2.5124578314611028e-05,
+      "loss": 0.3375,
+      "step": 1394
+    },
+    {
+      "epoch": 2.3805460750853245,
+      "grad_norm": 0.3547306658802128,
+      "learning_rate": 2.5101543101008795e-05,
+      "loss": 0.3311,
+      "step": 1395
+    },
+    {
+      "epoch": 2.3822525597269624,
+      "grad_norm": 0.348365710623611,
+      "learning_rate": 2.507850064669921e-05,
+      "loss": 0.3309,
+      "step": 1396
+    },
+    {
+      "epoch": 2.3839590443686007,
+      "grad_norm": 0.39395625941786366,
+      "learning_rate": 2.5055450984386828e-05,
+      "loss": 0.3446,
+      "step": 1397
+    },
+    {
+      "epoch": 2.385665529010239,
+      "grad_norm": 0.37869370932349095,
+      "learning_rate": 2.5032394146786434e-05,
+      "loss": 0.3721,
+      "step": 1398
+    },
+    {
+      "epoch": 2.387372013651877,
+      "grad_norm": 0.416001557948223,
+      "learning_rate": 2.500933016662298e-05,
+      "loss": 0.3184,
+      "step": 1399
+    },
+    {
+      "epoch": 2.3890784982935154,
+      "grad_norm": 0.3708619882788283,
+      "learning_rate": 2.498625907663158e-05,
+      "loss": 0.3231,
+      "step": 1400
+    },
+    {
+      "epoch": 2.3907849829351537,
+      "grad_norm": 0.3734903623021443,
+      "learning_rate": 2.4963180909557413e-05,
+      "loss": 0.337,
+      "step": 1401
+    },
+    {
+      "epoch": 2.3924914675767917,
+      "grad_norm": 0.3619258193640416,
+      "learning_rate": 2.4940095698155728e-05,
+      "loss": 0.3325,
+      "step": 1402
+    },
+    {
+      "epoch": 2.39419795221843,
+      "grad_norm": 0.43457933401171456,
+      "learning_rate": 2.4917003475191752e-05,
+      "loss": 0.3566,
+      "step": 1403
+    },
+    {
+      "epoch": 2.3959044368600684,
+      "grad_norm": 0.36789493194294154,
+      "learning_rate": 2.4893904273440677e-05,
+      "loss": 0.3305,
+      "step": 1404
+    },
+    {
+      "epoch": 2.3976109215017063,
+      "grad_norm": 0.3702806744907535,
+      "learning_rate": 2.487079812568759e-05,
+      "loss": 0.339,
+      "step": 1405
+    },
+    {
+      "epoch": 2.3993174061433447,
+      "grad_norm": 0.34765241619067994,
+      "learning_rate": 2.484768506472745e-05,
+      "loss": 0.3303,
+      "step": 1406
+    },
+    {
+      "epoch": 2.401023890784983,
+      "grad_norm": 0.3691851322616644,
+      "learning_rate": 2.4824565123365013e-05,
+      "loss": 0.3561,
+      "step": 1407
+    },
+    {
+      "epoch": 2.4027303754266214,
+      "grad_norm": 0.33489601093114274,
+      "learning_rate": 2.4801438334414808e-05,
+      "loss": 0.3086,
+      "step": 1408
+    },
+    {
+      "epoch": 2.4044368600682593,
+      "grad_norm": 0.4886959502080782,
+      "learning_rate": 2.477830473070108e-05,
+      "loss": 0.3248,
+      "step": 1409
+    },
+    {
+      "epoch": 2.4061433447098977,
+      "grad_norm": 0.4439476834476966,
+      "learning_rate": 2.4755164345057754e-05,
+      "loss": 0.3346,
+      "step": 1410
+    },
+    {
+      "epoch": 2.407849829351536,
+      "grad_norm": 0.34873223446808826,
+      "learning_rate": 2.473201721032837e-05,
+      "loss": 0.3368,
+      "step": 1411
+    },
+    {
+      "epoch": 2.409556313993174,
+      "grad_norm": 0.41435991243131887,
+      "learning_rate": 2.4708863359366056e-05,
+      "loss": 0.3307,
+      "step": 1412
+    },
+    {
+      "epoch": 2.4112627986348123,
+      "grad_norm": 0.36887177211123173,
+      "learning_rate": 2.4685702825033464e-05,
+      "loss": 0.3461,
+      "step": 1413
+    },
+    {
+      "epoch": 2.4129692832764507,
+      "grad_norm": 0.3812688309491389,
+      "learning_rate": 2.4662535640202737e-05,
+      "loss": 0.329,
+      "step": 1414
+    },
+    {
+      "epoch": 2.4146757679180886,
+      "grad_norm": 0.37755571817285283,
+      "learning_rate": 2.4639361837755463e-05,
+      "loss": 0.3127,
+      "step": 1415
+    },
+    {
+      "epoch": 2.416382252559727,
+      "grad_norm": 0.38208082202786353,
+      "learning_rate": 2.4616181450582613e-05,
+      "loss": 0.3714,
+      "step": 1416
+    },
+    {
+      "epoch": 2.4180887372013653,
+      "grad_norm": 0.424806909408548,
+      "learning_rate": 2.459299451158449e-05,
+      "loss": 0.3207,
+      "step": 1417
+    },
+    {
+      "epoch": 2.419795221843003,
+      "grad_norm": 0.394188998633754,
+      "learning_rate": 2.456980105367074e-05,
+      "loss": 0.3147,
+      "step": 1418
+    },
+    {
+      "epoch": 2.4215017064846416,
+      "grad_norm": 0.35728493684993823,
+      "learning_rate": 2.4546601109760223e-05,
+      "loss": 0.3321,
+      "step": 1419
+    },
+    {
+      "epoch": 2.42320819112628,
+      "grad_norm": 0.3390171300960544,
+      "learning_rate": 2.4523394712781014e-05,
+      "loss": 0.3382,
+      "step": 1420
+    },
+    {
+      "epoch": 2.4249146757679183,
+      "grad_norm": 0.41077102645216684,
+      "learning_rate": 2.4500181895670353e-05,
+      "loss": 0.311,
+      "step": 1421
+    },
+    {
+      "epoch": 2.426621160409556,
+      "grad_norm": 0.3670602580780559,
+      "learning_rate": 2.4476962691374582e-05,
+      "loss": 0.314,
+      "step": 1422
+    },
+    {
+      "epoch": 2.4283276450511946,
+      "grad_norm": 0.3537707122343819,
+      "learning_rate": 2.445373713284912e-05,
+      "loss": 0.3366,
+      "step": 1423
+    },
+    {
+      "epoch": 2.430034129692833,
+      "grad_norm": 0.3337392371509326,
+      "learning_rate": 2.4430505253058394e-05,
+      "loss": 0.3547,
+      "step": 1424
+    },
+    {
+      "epoch": 2.431740614334471,
+      "grad_norm": 0.3933018784100391,
+      "learning_rate": 2.4407267084975815e-05,
+      "loss": 0.314,
+      "step": 1425
+    },
+    {
+      "epoch": 2.4334470989761092,
+      "grad_norm": 0.37056767427899595,
+      "learning_rate": 2.4384022661583705e-05,
+      "loss": 0.3111,
+      "step": 1426
+    },
+    {
+      "epoch": 2.4351535836177476,
+      "grad_norm": 0.4441057268863796,
+      "learning_rate": 2.4360772015873274e-05,
+      "loss": 0.3791,
+      "step": 1427
+    },
+    {
+      "epoch": 2.4368600682593855,
+      "grad_norm": 0.3429781436405533,
+      "learning_rate": 2.4337515180844573e-05,
+      "loss": 0.3561,
+      "step": 1428
+    },
+    {
+      "epoch": 2.438566552901024,
+      "grad_norm": 0.36718679917444935,
+      "learning_rate": 2.4314252189506408e-05,
+      "loss": 0.3323,
+      "step": 1429
+    },
+    {
+      "epoch": 2.4402730375426622,
+      "grad_norm": 0.31672706515917426,
+      "learning_rate": 2.429098307487635e-05,
+      "loss": 0.3236,
+      "step": 1430
+    },
+    {
+      "epoch": 2.4419795221843,
+      "grad_norm": 0.3919414240234656,
+      "learning_rate": 2.4267707869980646e-05,
+      "loss": 0.3345,
+      "step": 1431
+    },
+    {
+      "epoch": 2.4436860068259385,
+      "grad_norm": 0.3372894842089935,
+      "learning_rate": 2.4244426607854193e-05,
+      "loss": 0.3024,
+      "step": 1432
+    },
+    {
+      "epoch": 2.445392491467577,
+      "grad_norm": 0.3816449925186258,
+      "learning_rate": 2.422113932154049e-05,
+      "loss": 0.3394,
+      "step": 1433
+    },
+    {
+      "epoch": 2.4470989761092152,
+      "grad_norm": 0.3669988886690805,
+      "learning_rate": 2.4197846044091585e-05,
+      "loss": 0.3791,
+      "step": 1434
+    },
+    {
+      "epoch": 2.448805460750853,
+      "grad_norm": 0.34702738131934063,
+      "learning_rate": 2.417454680856801e-05,
+      "loss": 0.3294,
+      "step": 1435
+    },
+    {
+      "epoch": 2.4505119453924915,
+      "grad_norm": 0.36254456020321973,
+      "learning_rate": 2.415124164803877e-05,
+      "loss": 0.3303,
+      "step": 1436
+    },
+    {
+      "epoch": 2.45221843003413,
+      "grad_norm": 0.3419956969338929,
+      "learning_rate": 2.4127930595581285e-05,
+      "loss": 0.3223,
+      "step": 1437
+    },
+    {
+      "epoch": 2.453924914675768,
+      "grad_norm": 0.37916769686625507,
+      "learning_rate": 2.410461368428133e-05,
+      "loss": 0.2957,
+      "step": 1438
+    },
+    {
+      "epoch": 2.455631399317406,
+      "grad_norm": 0.39593285539957496,
+      "learning_rate": 2.4081290947232993e-05,
+      "loss": 0.3332,
+      "step": 1439
+    },
+    {
+      "epoch": 2.4573378839590445,
+      "grad_norm": 0.3641399475866781,
+      "learning_rate": 2.405796241753864e-05,
+      "loss": 0.3252,
+      "step": 1440
+    },
+    {
+      "epoch": 2.4590443686006824,
+      "grad_norm": 0.37308343106539177,
+      "learning_rate": 2.4034628128308844e-05,
+      "loss": 0.296,
+      "step": 1441
+    },
+    {
+      "epoch": 2.460750853242321,
+      "grad_norm": 0.3569807734609345,
+      "learning_rate": 2.4011288112662364e-05,
+      "loss": 0.2946,
+      "step": 1442
+    },
+    {
+      "epoch": 2.462457337883959,
+      "grad_norm": 0.4090553354996803,
+      "learning_rate": 2.398794240372608e-05,
+      "loss": 0.3319,
+      "step": 1443
+    },
+    {
+      "epoch": 2.464163822525597,
+      "grad_norm": 0.3730810745041935,
+      "learning_rate": 2.396459103463496e-05,
+      "loss": 0.3222,
+      "step": 1444
+    },
+    {
+      "epoch": 2.4658703071672354,
+      "grad_norm": 0.3379437485115968,
+      "learning_rate": 2.3941234038532e-05,
+      "loss": 0.3389,
+      "step": 1445
+    },
+    {
+      "epoch": 2.467576791808874,
+      "grad_norm": 0.3638076547668745,
+      "learning_rate": 2.3917871448568182e-05,
+      "loss": 0.3443,
+      "step": 1446
+    },
+    {
+      "epoch": 2.469283276450512,
+      "grad_norm": 0.37797090219784407,
+      "learning_rate": 2.3894503297902437e-05,
+      "loss": 0.3314,
+      "step": 1447
+    },
+    {
+      "epoch": 2.47098976109215,
+      "grad_norm": 0.3454814124459251,
+      "learning_rate": 2.387112961970157e-05,
+      "loss": 0.3179,
+      "step": 1448
+    },
+    {
+      "epoch": 2.4726962457337884,
+      "grad_norm": 0.3890717934183557,
+      "learning_rate": 2.384775044714025e-05,
+      "loss": 0.3438,
+      "step": 1449
+    },
+    {
+      "epoch": 2.474402730375427,
+      "grad_norm": 0.3985238454597805,
+      "learning_rate": 2.3824365813400934e-05,
+      "loss": 0.3945,
+      "step": 1450
+    },
+    {
+      "epoch": 2.4761092150170647,
+      "grad_norm": 0.41221527808061414,
+      "learning_rate": 2.3800975751673825e-05,
+      "loss": 0.3986,
+      "step": 1451
+    },
+    {
+      "epoch": 2.477815699658703,
+      "grad_norm": 0.3969985176310822,
+      "learning_rate": 2.377758029515685e-05,
+      "loss": 0.3191,
+      "step": 1452
+    },
+    {
+      "epoch": 2.4795221843003414,
+      "grad_norm": 0.3880421847396879,
+      "learning_rate": 2.3754179477055576e-05,
+      "loss": 0.3541,
+      "step": 1453
+    },
+    {
+      "epoch": 2.4812286689419794,
+      "grad_norm": 0.3715763075289722,
+      "learning_rate": 2.3730773330583183e-05,
+      "loss": 0.3288,
+      "step": 1454
+    },
+    {
+      "epoch": 2.4829351535836177,
+      "grad_norm": 0.4031941085825435,
+      "learning_rate": 2.3707361888960413e-05,
+      "loss": 0.3153,
+      "step": 1455
+    },
+    {
+      "epoch": 2.484641638225256,
+      "grad_norm": 0.33837450641052536,
+      "learning_rate": 2.3683945185415528e-05,
+      "loss": 0.2986,
+      "step": 1456
+    },
+    {
+      "epoch": 2.486348122866894,
+      "grad_norm": 0.39223149797213314,
+      "learning_rate": 2.3660523253184254e-05,
+      "loss": 0.3654,
+      "step": 1457
+    },
+    {
+      "epoch": 2.4880546075085324,
+      "grad_norm": 0.4584458087319988,
+      "learning_rate": 2.3637096125509737e-05,
+      "loss": 0.3326,
+      "step": 1458
+    },
+    {
+      "epoch": 2.4897610921501707,
+      "grad_norm": 0.35742942089729385,
+      "learning_rate": 2.3613663835642515e-05,
+      "loss": 0.3503,
+      "step": 1459
+    },
+    {
+      "epoch": 2.491467576791809,
+      "grad_norm": 0.4003287666095139,
+      "learning_rate": 2.3590226416840415e-05,
+      "loss": 0.3748,
+      "step": 1460
+    },
+    {
+      "epoch": 2.493174061433447,
+      "grad_norm": 0.38386986612118523,
+      "learning_rate": 2.3566783902368586e-05,
+      "loss": 0.3567,
+      "step": 1461
+    },
+    {
+      "epoch": 2.4948805460750854,
+      "grad_norm": 0.3898899370166114,
+      "learning_rate": 2.354333632549938e-05,
+      "loss": 0.3233,
+      "step": 1462
+    },
+    {
+      "epoch": 2.4965870307167237,
+      "grad_norm": 0.34310571632457215,
+      "learning_rate": 2.3519883719512345e-05,
+      "loss": 0.3556,
+      "step": 1463
+    },
+    {
+      "epoch": 2.4982935153583616,
+      "grad_norm": 0.3367770889277449,
+      "learning_rate": 2.349642611769416e-05,
+      "loss": 0.3526,
+      "step": 1464
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.38653236875272007,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.4039,
+      "step": 1465
+    },
+    {
+      "epoch": 2.5017064846416384,
+      "grad_norm": 0.3797790080161611,
+      "learning_rate": 2.3449496059746508e-05,
+      "loss": 0.3071,
+      "step": 1466
+    },
+    {
+      "epoch": 2.5034129692832767,
+      "grad_norm": 0.4026662760948025,
+      "learning_rate": 2.3426023670225674e-05,
+      "loss": 0.4287,
+      "step": 1467
+    },
+    {
+      "epoch": 2.5051194539249146,
+      "grad_norm": 0.36107542161294814,
+      "learning_rate": 2.340254641809087e-05,
+      "loss": 0.297,
+      "step": 1468
+    },
+    {
+      "epoch": 2.506825938566553,
+      "grad_norm": 0.35554618094449175,
+      "learning_rate": 2.3379064336663754e-05,
+      "loss": 0.3542,
+      "step": 1469
+    },
+    {
+      "epoch": 2.508532423208191,
+      "grad_norm": 0.36658748821399895,
+      "learning_rate": 2.3355577459272856e-05,
+      "loss": 0.352,
+      "step": 1470
+    },
+    {
+      "epoch": 2.5102389078498293,
+      "grad_norm": 0.40957016863765,
+      "learning_rate": 2.3332085819253494e-05,
+      "loss": 0.3212,
+      "step": 1471
+    },
+    {
+      "epoch": 2.5119453924914676,
+      "grad_norm": 0.3330358802773706,
+      "learning_rate": 2.330858944994776e-05,
+      "loss": 0.2995,
+      "step": 1472
+    },
+    {
+      "epoch": 2.513651877133106,
+      "grad_norm": 0.37169872025122885,
+      "learning_rate": 2.328508838470445e-05,
+      "loss": 0.3471,
+      "step": 1473
+    },
+    {
+      "epoch": 2.515358361774744,
+      "grad_norm": 0.4068903182174979,
+      "learning_rate": 2.326158265687903e-05,
+      "loss": 0.3515,
+      "step": 1474
+    },
+    {
+      "epoch": 2.5170648464163823,
+      "grad_norm": 0.40512073054173614,
+      "learning_rate": 2.3238072299833584e-05,
+      "loss": 0.3025,
+      "step": 1475
+    },
+    {
+      "epoch": 2.51877133105802,
+      "grad_norm": 0.35816508402342645,
+      "learning_rate": 2.3214557346936755e-05,
+      "loss": 0.3154,
+      "step": 1476
+    },
+    {
+      "epoch": 2.5204778156996586,
+      "grad_norm": 0.45982283442918426,
+      "learning_rate": 2.3191037831563727e-05,
+      "loss": 0.345,
+      "step": 1477
+    },
+    {
+      "epoch": 2.522184300341297,
+      "grad_norm": 0.4173484682436333,
+      "learning_rate": 2.316751378709614e-05,
+      "loss": 0.3406,
+      "step": 1478
+    },
+    {
+      "epoch": 2.5238907849829353,
+      "grad_norm": 0.36961090311820594,
+      "learning_rate": 2.3143985246922077e-05,
+      "loss": 0.3778,
+      "step": 1479
+    },
+    {
+      "epoch": 2.5255972696245736,
+      "grad_norm": 0.37553258774310944,
+      "learning_rate": 2.3120452244436e-05,
+      "loss": 0.3119,
+      "step": 1480
+    },
+    {
+      "epoch": 2.5273037542662116,
+      "grad_norm": 0.3750417092831229,
+      "learning_rate": 2.309691481303871e-05,
+      "loss": 0.3246,
+      "step": 1481
+    },
+    {
+      "epoch": 2.52901023890785,
+      "grad_norm": 0.3141087844075222,
+      "learning_rate": 2.3073372986137274e-05,
+      "loss": 0.344,
+      "step": 1482
+    },
+    {
+      "epoch": 2.530716723549488,
+      "grad_norm": 0.3378382304761785,
+      "learning_rate": 2.3049826797145002e-05,
+      "loss": 0.3533,
+      "step": 1483
+    },
+    {
+      "epoch": 2.532423208191126,
+      "grad_norm": 0.3202302348119175,
+      "learning_rate": 2.302627627948142e-05,
+      "loss": 0.332,
+      "step": 1484
+    },
+    {
+      "epoch": 2.5341296928327646,
+      "grad_norm": 0.3417342609194838,
+      "learning_rate": 2.3002721466572168e-05,
+      "loss": 0.3396,
+      "step": 1485
+    },
+    {
+      "epoch": 2.535836177474403,
+      "grad_norm": 0.35991105559347536,
+      "learning_rate": 2.2979162391849003e-05,
+      "loss": 0.3366,
+      "step": 1486
+    },
+    {
+      "epoch": 2.537542662116041,
+      "grad_norm": 0.3355428033489997,
+      "learning_rate": 2.2955599088749722e-05,
+      "loss": 0.4948,
+      "step": 1487
+    },
+    {
+      "epoch": 2.539249146757679,
+      "grad_norm": 0.36609471776900404,
+      "learning_rate": 2.2932031590718116e-05,
+      "loss": 0.3213,
+      "step": 1488
+    },
+    {
+      "epoch": 2.5409556313993176,
+      "grad_norm": 0.35296771491544104,
+      "learning_rate": 2.2908459931203947e-05,
+      "loss": 0.3242,
+      "step": 1489
+    },
+    {
+      "epoch": 2.5426621160409555,
+      "grad_norm": 0.3853064302037463,
+      "learning_rate": 2.2884884143662867e-05,
+      "loss": 0.3418,
+      "step": 1490
+    },
+    {
+      "epoch": 2.544368600682594,
+      "grad_norm": 0.3185441446222499,
+      "learning_rate": 2.2861304261556393e-05,
+      "loss": 0.3416,
+      "step": 1491
+    },
+    {
+      "epoch": 2.546075085324232,
+      "grad_norm": 0.3427064024986631,
+      "learning_rate": 2.2837720318351866e-05,
+      "loss": 0.348,
+      "step": 1492
+    },
+    {
+      "epoch": 2.5477815699658706,
+      "grad_norm": 0.36539764520827644,
+      "learning_rate": 2.2814132347522375e-05,
+      "loss": 0.327,
+      "step": 1493
+    },
+    {
+      "epoch": 2.5494880546075085,
+      "grad_norm": 0.34389922932611533,
+      "learning_rate": 2.2790540382546724e-05,
+      "loss": 0.3053,
+      "step": 1494
+    },
+    {
+      "epoch": 2.551194539249147,
+      "grad_norm": 0.39591972206641485,
+      "learning_rate": 2.27669444569094e-05,
+      "loss": 0.3255,
+      "step": 1495
+    },
+    {
+      "epoch": 2.5529010238907848,
+      "grad_norm": 0.36868469745591775,
+      "learning_rate": 2.27433446041005e-05,
+      "loss": 0.3016,
+      "step": 1496
+    },
+    {
+      "epoch": 2.554607508532423,
+      "grad_norm": 0.37489919340094,
+      "learning_rate": 2.2719740857615697e-05,
+      "loss": 0.3004,
+      "step": 1497
+    },
+    {
+      "epoch": 2.5563139931740615,
+      "grad_norm": 0.3511340577664287,
+      "learning_rate": 2.2696133250956192e-05,
+      "loss": 0.3285,
+      "step": 1498
+    },
+    {
+      "epoch": 2.5580204778157,
+      "grad_norm": 0.3454939870308825,
+      "learning_rate": 2.267252181762867e-05,
+      "loss": 0.347,
+      "step": 1499
+    },
+    {
+      "epoch": 2.5597269624573378,
+      "grad_norm": 0.4154911702244027,
+      "learning_rate": 2.2648906591145238e-05,
+      "loss": 0.3419,
+      "step": 1500
+    },
+    {
+      "epoch": 2.561433447098976,
+      "grad_norm": 0.3570111180021174,
+      "learning_rate": 2.2625287605023392e-05,
+      "loss": 0.3769,
+      "step": 1501
+    },
+    {
+      "epoch": 2.5631399317406145,
+      "grad_norm": 0.3401737285458632,
+      "learning_rate": 2.260166489278597e-05,
+      "loss": 0.3474,
+      "step": 1502
+    },
+    {
+      "epoch": 2.5648464163822524,
+      "grad_norm": 0.44603488244193745,
+      "learning_rate": 2.2578038487961075e-05,
+      "loss": 0.3603,
+      "step": 1503
+    },
+    {
+      "epoch": 2.5665529010238908,
+      "grad_norm": 0.33278179482227377,
+      "learning_rate": 2.2554408424082075e-05,
+      "loss": 0.3017,
+      "step": 1504
+    },
+    {
+      "epoch": 2.568259385665529,
+      "grad_norm": 0.33587061726853057,
+      "learning_rate": 2.2530774734687525e-05,
+      "loss": 0.3323,
+      "step": 1505
+    },
+    {
+      "epoch": 2.5699658703071675,
+      "grad_norm": 0.4400712644895227,
+      "learning_rate": 2.2507137453321125e-05,
+      "loss": 0.356,
+      "step": 1506
+    },
+    {
+      "epoch": 2.5716723549488054,
+      "grad_norm": 0.3571140421651242,
+      "learning_rate": 2.248349661353167e-05,
+      "loss": 0.3155,
+      "step": 1507
+    },
+    {
+      "epoch": 2.573378839590444,
+      "grad_norm": 0.364079170575246,
+      "learning_rate": 2.2459852248873012e-05,
+      "loss": 0.3482,
+      "step": 1508
+    },
+    {
+      "epoch": 2.5750853242320817,
+      "grad_norm": 0.42970346416624294,
+      "learning_rate": 2.2436204392904006e-05,
+      "loss": 0.3525,
+      "step": 1509
+    },
+    {
+      "epoch": 2.57679180887372,
+      "grad_norm": 0.39056825360322367,
+      "learning_rate": 2.241255307918844e-05,
+      "loss": 0.3062,
+      "step": 1510
+    },
+    {
+      "epoch": 2.5784982935153584,
+      "grad_norm": 0.34620813919722987,
+      "learning_rate": 2.2388898341295053e-05,
+      "loss": 0.4253,
+      "step": 1511
+    },
+    {
+      "epoch": 2.580204778156997,
+      "grad_norm": 0.39111293003093417,
+      "learning_rate": 2.2365240212797397e-05,
+      "loss": 0.3122,
+      "step": 1512
+    },
+    {
+      "epoch": 2.5819112627986347,
+      "grad_norm": 0.36781307725420737,
+      "learning_rate": 2.234157872727387e-05,
+      "loss": 0.3157,
+      "step": 1513
+    },
+    {
+      "epoch": 2.583617747440273,
+      "grad_norm": 0.3509137695917543,
+      "learning_rate": 2.2317913918307616e-05,
+      "loss": 0.32,
+      "step": 1514
+    },
+    {
+      "epoch": 2.5853242320819114,
+      "grad_norm": 0.36137579661930413,
+      "learning_rate": 2.2294245819486515e-05,
+      "loss": 0.3439,
+      "step": 1515
+    },
+    {
+      "epoch": 2.5870307167235493,
+      "grad_norm": 0.37850530442502667,
+      "learning_rate": 2.227057446440309e-05,
+      "loss": 0.3305,
+      "step": 1516
+    },
+    {
+      "epoch": 2.5887372013651877,
+      "grad_norm": 0.3450905118912204,
+      "learning_rate": 2.2246899886654512e-05,
+      "loss": 0.332,
+      "step": 1517
+    },
+    {
+      "epoch": 2.590443686006826,
+      "grad_norm": 0.3540659388313431,
+      "learning_rate": 2.2223222119842505e-05,
+      "loss": 0.3285,
+      "step": 1518
+    },
+    {
+      "epoch": 2.5921501706484644,
+      "grad_norm": 0.33810482101335076,
+      "learning_rate": 2.219954119757333e-05,
+      "loss": 0.3101,
+      "step": 1519
+    },
+    {
+      "epoch": 2.5938566552901023,
+      "grad_norm": 0.4027706707970301,
+      "learning_rate": 2.2175857153457733e-05,
+      "loss": 0.3008,
+      "step": 1520
+    },
+    {
+      "epoch": 2.5955631399317407,
+      "grad_norm": 0.3850159272249956,
+      "learning_rate": 2.2152170021110876e-05,
+      "loss": 0.3737,
+      "step": 1521
+    },
+    {
+      "epoch": 2.5972696245733786,
+      "grad_norm": 0.3420595026622498,
+      "learning_rate": 2.2128479834152303e-05,
+      "loss": 0.3571,
+      "step": 1522
+    },
+    {
+      "epoch": 2.598976109215017,
+      "grad_norm": 0.36231382375440196,
+      "learning_rate": 2.2104786626205915e-05,
+      "loss": 0.3179,
+      "step": 1523
+    },
+    {
+      "epoch": 2.6006825938566553,
+      "grad_norm": 0.35324028732062923,
+      "learning_rate": 2.208109043089988e-05,
+      "loss": 0.3779,
+      "step": 1524
+    },
+    {
+      "epoch": 2.6023890784982937,
+      "grad_norm": 0.3599439627039347,
+      "learning_rate": 2.2057391281866617e-05,
+      "loss": 0.3225,
+      "step": 1525
+    },
+    {
+      "epoch": 2.6040955631399316,
+      "grad_norm": 0.4297972491076275,
+      "learning_rate": 2.203368921274273e-05,
+      "loss": 0.3475,
+      "step": 1526
+    },
+    {
+      "epoch": 2.60580204778157,
+      "grad_norm": 0.32920065294422945,
+      "learning_rate": 2.2009984257168978e-05,
+      "loss": 0.3201,
+      "step": 1527
+    },
+    {
+      "epoch": 2.6075085324232083,
+      "grad_norm": 0.36500147382442766,
+      "learning_rate": 2.19862764487902e-05,
+      "loss": 0.3312,
+      "step": 1528
+    },
+    {
+      "epoch": 2.6092150170648463,
+      "grad_norm": 0.3520609399732075,
+      "learning_rate": 2.196256582125529e-05,
+      "loss": 0.2805,
+      "step": 1529
+    },
+    {
+      "epoch": 2.6109215017064846,
+      "grad_norm": 0.35033065080394465,
+      "learning_rate": 2.1938852408217168e-05,
+      "loss": 0.37,
+      "step": 1530
+    },
+    {
+      "epoch": 2.612627986348123,
+      "grad_norm": 0.34259172335469495,
+      "learning_rate": 2.1915136243332662e-05,
+      "loss": 0.3181,
+      "step": 1531
+    },
+    {
+      "epoch": 2.6143344709897613,
+      "grad_norm": 0.37189251013844604,
+      "learning_rate": 2.189141736026255e-05,
+      "loss": 0.3168,
+      "step": 1532
+    },
+    {
+      "epoch": 2.6160409556313993,
+      "grad_norm": 0.3721659335744151,
+      "learning_rate": 2.186769579267144e-05,
+      "loss": 0.3202,
+      "step": 1533
+    },
+    {
+      "epoch": 2.6177474402730376,
+      "grad_norm": 0.3639098058840575,
+      "learning_rate": 2.1843971574227755e-05,
+      "loss": 0.3074,
+      "step": 1534
+    },
+    {
+      "epoch": 2.6194539249146755,
+      "grad_norm": 0.3769605832514304,
+      "learning_rate": 2.1820244738603686e-05,
+      "loss": 0.356,
+      "step": 1535
+    },
+    {
+      "epoch": 2.621160409556314,
+      "grad_norm": 0.3655217175141363,
+      "learning_rate": 2.1796515319475144e-05,
+      "loss": 0.3438,
+      "step": 1536
+    },
+    {
+      "epoch": 2.6228668941979523,
+      "grad_norm": 0.35103512956620203,
+      "learning_rate": 2.177278335052169e-05,
+      "loss": 0.3687,
+      "step": 1537
+    },
+    {
+      "epoch": 2.6245733788395906,
+      "grad_norm": 0.3645165351023154,
+      "learning_rate": 2.174904886542651e-05,
+      "loss": 0.3403,
+      "step": 1538
+    },
+    {
+      "epoch": 2.6262798634812285,
+      "grad_norm": 0.33120848612975773,
+      "learning_rate": 2.172531189787638e-05,
+      "loss": 0.3352,
+      "step": 1539
+    },
+    {
+      "epoch": 2.627986348122867,
+      "grad_norm": 0.34308548415086215,
+      "learning_rate": 2.1701572481561574e-05,
+      "loss": 0.3373,
+      "step": 1540
+    },
+    {
+      "epoch": 2.6296928327645053,
+      "grad_norm": 0.3624467272308664,
+      "learning_rate": 2.167783065017585e-05,
+      "loss": 0.3129,
+      "step": 1541
+    },
+    {
+      "epoch": 2.631399317406143,
+      "grad_norm": 0.3275048899706405,
+      "learning_rate": 2.1654086437416394e-05,
+      "loss": 0.376,
+      "step": 1542
+    },
+    {
+      "epoch": 2.6331058020477816,
+      "grad_norm": 0.34876972572695425,
+      "learning_rate": 2.1630339876983783e-05,
+      "loss": 0.3044,
+      "step": 1543
+    },
+    {
+      "epoch": 2.63481228668942,
+      "grad_norm": 0.354065295382463,
+      "learning_rate": 2.160659100258191e-05,
+      "loss": 0.3273,
+      "step": 1544
+    },
+    {
+      "epoch": 2.6365187713310583,
+      "grad_norm": 0.3605089626247716,
+      "learning_rate": 2.1582839847917954e-05,
+      "loss": 0.3437,
+      "step": 1545
+    },
+    {
+      "epoch": 2.638225255972696,
+      "grad_norm": 0.39176557422172903,
+      "learning_rate": 2.155908644670234e-05,
+      "loss": 0.3292,
+      "step": 1546
+    },
+    {
+      "epoch": 2.6399317406143346,
+      "grad_norm": 0.34142414891699385,
+      "learning_rate": 2.1535330832648677e-05,
+      "loss": 0.3296,
+      "step": 1547
+    },
+    {
+      "epoch": 2.6416382252559725,
+      "grad_norm": 0.3628720766562326,
+      "learning_rate": 2.151157303947371e-05,
+      "loss": 0.3651,
+      "step": 1548
+    },
+    {
+      "epoch": 2.643344709897611,
+      "grad_norm": 0.35653396070721,
+      "learning_rate": 2.1487813100897287e-05,
+      "loss": 0.3165,
+      "step": 1549
+    },
+    {
+      "epoch": 2.645051194539249,
+      "grad_norm": 0.37500779656869304,
+      "learning_rate": 2.146405105064229e-05,
+      "loss": 0.3008,
+      "step": 1550
+    },
+    {
+      "epoch": 2.6467576791808876,
+      "grad_norm": 0.3638063646129447,
+      "learning_rate": 2.1440286922434604e-05,
+      "loss": 0.3527,
+      "step": 1551
+    },
+    {
+      "epoch": 2.6484641638225255,
+      "grad_norm": 0.3975840120709893,
+      "learning_rate": 2.1416520750003065e-05,
+      "loss": 0.339,
+      "step": 1552
+    },
+    {
+      "epoch": 2.650170648464164,
+      "grad_norm": 0.3997896639599607,
+      "learning_rate": 2.139275256707941e-05,
+      "loss": 0.3032,
+      "step": 1553
+    },
+    {
+      "epoch": 2.651877133105802,
+      "grad_norm": 0.3379880488943614,
+      "learning_rate": 2.1368982407398218e-05,
+      "loss": 0.3407,
+      "step": 1554
+    },
+    {
+      "epoch": 2.65358361774744,
+      "grad_norm": 0.3584110053295769,
+      "learning_rate": 2.13452103046969e-05,
+      "loss": 0.2919,
+      "step": 1555
+    },
+    {
+      "epoch": 2.6552901023890785,
+      "grad_norm": 0.32809341209082954,
+      "learning_rate": 2.1321436292715587e-05,
+      "loss": 0.3373,
+      "step": 1556
+    },
+    {
+      "epoch": 2.656996587030717,
+      "grad_norm": 0.34693589171988704,
+      "learning_rate": 2.1297660405197155e-05,
+      "loss": 0.302,
+      "step": 1557
+    },
+    {
+      "epoch": 2.658703071672355,
+      "grad_norm": 0.417036926378175,
+      "learning_rate": 2.1273882675887122e-05,
+      "loss": 0.3456,
+      "step": 1558
+    },
+    {
+      "epoch": 2.660409556313993,
+      "grad_norm": 0.3987269943492533,
+      "learning_rate": 2.125010313853362e-05,
+      "loss": 0.3279,
+      "step": 1559
+    },
+    {
+      "epoch": 2.6621160409556315,
+      "grad_norm": 0.3725601004877647,
+      "learning_rate": 2.1226321826887368e-05,
+      "loss": 0.3287,
+      "step": 1560
+    },
+    {
+      "epoch": 2.6638225255972694,
+      "grad_norm": 0.3603161553659546,
+      "learning_rate": 2.120253877470158e-05,
+      "loss": 0.3253,
+      "step": 1561
+    },
+    {
+      "epoch": 2.6655290102389078,
+      "grad_norm": 0.37497768836582135,
+      "learning_rate": 2.1178754015731945e-05,
+      "loss": 0.3633,
+      "step": 1562
+    },
+    {
+      "epoch": 2.667235494880546,
+      "grad_norm": 0.34696377080955054,
+      "learning_rate": 2.1154967583736584e-05,
+      "loss": 0.344,
+      "step": 1563
+    },
+    {
+      "epoch": 2.6689419795221845,
+      "grad_norm": 0.3839859277931613,
+      "learning_rate": 2.113117951247598e-05,
+      "loss": 0.3626,
+      "step": 1564
+    },
+    {
+      "epoch": 2.6706484641638224,
+      "grad_norm": 0.3575518356176394,
+      "learning_rate": 2.1107389835712955e-05,
+      "loss": 0.324,
+      "step": 1565
+    },
+    {
+      "epoch": 2.6723549488054608,
+      "grad_norm": 0.30616261327690975,
+      "learning_rate": 2.1083598587212605e-05,
+      "loss": 0.3274,
+      "step": 1566
+    },
+    {
+      "epoch": 2.674061433447099,
+      "grad_norm": 0.34876946647688606,
+      "learning_rate": 2.105980580074226e-05,
+      "loss": 0.3529,
+      "step": 1567
+    },
+    {
+      "epoch": 2.675767918088737,
+      "grad_norm": 0.36759236932565875,
+      "learning_rate": 2.1036011510071416e-05,
+      "loss": 0.2979,
+      "step": 1568
+    },
+    {
+      "epoch": 2.6774744027303754,
+      "grad_norm": 0.36493991987989527,
+      "learning_rate": 2.101221574897172e-05,
+      "loss": 0.3081,
+      "step": 1569
+    },
+    {
+      "epoch": 2.6791808873720138,
+      "grad_norm": 0.37896500199069055,
+      "learning_rate": 2.0988418551216912e-05,
+      "loss": 0.3458,
+      "step": 1570
+    },
+    {
+      "epoch": 2.680887372013652,
+      "grad_norm": 1.1167198478547995,
+      "learning_rate": 2.0964619950582747e-05,
+      "loss": 0.3232,
+      "step": 1571
+    },
+    {
+      "epoch": 2.68259385665529,
+      "grad_norm": 0.35455315066160753,
+      "learning_rate": 2.0940819980846992e-05,
+      "loss": 0.3542,
+      "step": 1572
+    },
+    {
+      "epoch": 2.6843003412969284,
+      "grad_norm": 0.4023373744928067,
+      "learning_rate": 2.091701867578936e-05,
+      "loss": 0.3243,
+      "step": 1573
+    },
+    {
+      "epoch": 2.6860068259385663,
+      "grad_norm": 0.36866692733463213,
+      "learning_rate": 2.0893216069191437e-05,
+      "loss": 0.3368,
+      "step": 1574
+    },
+    {
+      "epoch": 2.6877133105802047,
+      "grad_norm": 0.38032693691732394,
+      "learning_rate": 2.0869412194836677e-05,
+      "loss": 0.3318,
+      "step": 1575
+    },
+    {
+      "epoch": 2.689419795221843,
+      "grad_norm": 0.3310774792170441,
+      "learning_rate": 2.084560708651033e-05,
+      "loss": 0.3083,
+      "step": 1576
+    },
+    {
+      "epoch": 2.6911262798634814,
+      "grad_norm": 0.34387431070918356,
+      "learning_rate": 2.082180077799938e-05,
+      "loss": 0.3443,
+      "step": 1577
+    },
+    {
+      "epoch": 2.6928327645051193,
+      "grad_norm": 0.4166481093910801,
+      "learning_rate": 2.079799330309254e-05,
+      "loss": 0.3494,
+      "step": 1578
+    },
+    {
+      "epoch": 2.6945392491467577,
+      "grad_norm": 0.3728399532458675,
+      "learning_rate": 2.077418469558015e-05,
+      "loss": 0.3397,
+      "step": 1579
+    },
+    {
+      "epoch": 2.696245733788396,
+      "grad_norm": 0.33987847638407687,
+      "learning_rate": 2.07503749892542e-05,
+      "loss": 0.3516,
+      "step": 1580
+    },
+    {
+      "epoch": 2.697952218430034,
+      "grad_norm": 0.3625117258574951,
+      "learning_rate": 2.0726564217908188e-05,
+      "loss": 0.3135,
+      "step": 1581
+    },
+    {
+      "epoch": 2.6996587030716723,
+      "grad_norm": 0.3215747834204709,
+      "learning_rate": 2.070275241533716e-05,
+      "loss": 0.3459,
+      "step": 1582
+    },
+    {
+      "epoch": 2.7013651877133107,
+      "grad_norm": 0.36407703958316107,
+      "learning_rate": 2.0678939615337625e-05,
+      "loss": 0.3124,
+      "step": 1583
+    },
+    {
+      "epoch": 2.703071672354949,
+      "grad_norm": 0.3841532131926999,
+      "learning_rate": 2.065512585170747e-05,
+      "loss": 0.3196,
+      "step": 1584
+    },
+    {
+      "epoch": 2.704778156996587,
+      "grad_norm": 0.3984499240967021,
+      "learning_rate": 2.0631311158246002e-05,
+      "loss": 0.2921,
+      "step": 1585
+    },
+    {
+      "epoch": 2.7064846416382253,
+      "grad_norm": 0.36633412889607153,
+      "learning_rate": 2.060749556875381e-05,
+      "loss": 0.3278,
+      "step": 1586
+    },
+    {
+      "epoch": 2.7081911262798632,
+      "grad_norm": 0.34262907420461297,
+      "learning_rate": 2.058367911703277e-05,
+      "loss": 0.3335,
+      "step": 1587
+    },
+    {
+      "epoch": 2.7098976109215016,
+      "grad_norm": 0.32516339887309825,
+      "learning_rate": 2.055986183688598e-05,
+      "loss": 0.3561,
+      "step": 1588
+    },
+    {
+      "epoch": 2.71160409556314,
+      "grad_norm": 0.36094019810397276,
+      "learning_rate": 2.0536043762117717e-05,
+      "loss": 0.3347,
+      "step": 1589
+    },
+    {
+      "epoch": 2.7133105802047783,
+      "grad_norm": 0.3678192871584571,
+      "learning_rate": 2.0512224926533375e-05,
+      "loss": 0.317,
+      "step": 1590
+    },
+    {
+      "epoch": 2.7150170648464163,
+      "grad_norm": 0.35626547301694567,
+      "learning_rate": 2.0488405363939434e-05,
+      "loss": 0.3646,
+      "step": 1591
+    },
+    {
+      "epoch": 2.7167235494880546,
+      "grad_norm": 0.3876448802387714,
+      "learning_rate": 2.046458510814341e-05,
+      "loss": 0.3623,
+      "step": 1592
+    },
+    {
+      "epoch": 2.718430034129693,
+      "grad_norm": 0.3737984481229139,
+      "learning_rate": 2.0440764192953805e-05,
+      "loss": 0.357,
+      "step": 1593
+    },
+    {
+      "epoch": 2.720136518771331,
+      "grad_norm": 0.47120557756429443,
+      "learning_rate": 2.0416942652180037e-05,
+      "loss": 0.3219,
+      "step": 1594
+    },
+    {
+      "epoch": 2.7218430034129693,
+      "grad_norm": 0.34760756372777724,
+      "learning_rate": 2.0393120519632444e-05,
+      "loss": 0.3269,
+      "step": 1595
+    },
+    {
+      "epoch": 2.7235494880546076,
+      "grad_norm": 0.3365726738286948,
+      "learning_rate": 2.0369297829122168e-05,
+      "loss": 0.3386,
+      "step": 1596
+    },
+    {
+      "epoch": 2.725255972696246,
+      "grad_norm": 0.3602677916560659,
+      "learning_rate": 2.034547461446117e-05,
+      "loss": 0.3206,
+      "step": 1597
+    },
+    {
+      "epoch": 2.726962457337884,
+      "grad_norm": 0.3730257272181104,
+      "learning_rate": 2.0321650909462144e-05,
+      "loss": 0.3596,
+      "step": 1598
+    },
+    {
+      "epoch": 2.7286689419795223,
+      "grad_norm": 0.30881143810988376,
+      "learning_rate": 2.0297826747938483e-05,
+      "loss": 0.3277,
+      "step": 1599
+    },
+    {
+      "epoch": 2.73037542662116,
+      "grad_norm": 0.36938060678999873,
+      "learning_rate": 2.0274002163704226e-05,
+      "loss": 0.3485,
+      "step": 1600
+    },
+    {
+      "epoch": 2.7320819112627985,
+      "grad_norm": 0.3369892272185491,
+      "learning_rate": 2.0250177190574023e-05,
+      "loss": 0.342,
+      "step": 1601
+    },
+    {
+      "epoch": 2.733788395904437,
+      "grad_norm": 0.34992532725689274,
+      "learning_rate": 2.0226351862363043e-05,
+      "loss": 0.3325,
+      "step": 1602
+    },
+    {
+      "epoch": 2.7354948805460753,
+      "grad_norm": 0.3501014851111137,
+      "learning_rate": 2.0202526212887003e-05,
+      "loss": 0.3738,
+      "step": 1603
+    },
+    {
+      "epoch": 2.737201365187713,
+      "grad_norm": 0.3476924097993854,
+      "learning_rate": 2.0178700275962044e-05,
+      "loss": 0.3377,
+      "step": 1604
+    },
+    {
+      "epoch": 2.7389078498293515,
+      "grad_norm": 0.33874329422942734,
+      "learning_rate": 2.0154874085404724e-05,
+      "loss": 0.3583,
+      "step": 1605
+    },
+    {
+      "epoch": 2.74061433447099,
+      "grad_norm": 0.36032833957102245,
+      "learning_rate": 2.013104767503197e-05,
+      "loss": 0.3462,
+      "step": 1606
+    },
+    {
+      "epoch": 2.742320819112628,
+      "grad_norm": 0.3326759957576823,
+      "learning_rate": 2.0107221078661016e-05,
+      "loss": 0.3117,
+      "step": 1607
+    },
+    {
+      "epoch": 2.744027303754266,
+      "grad_norm": 0.40181150515698444,
+      "learning_rate": 2.008339433010934e-05,
+      "loss": 0.3226,
+      "step": 1608
+    },
+    {
+      "epoch": 2.7457337883959045,
+      "grad_norm": 0.3620467965946172,
+      "learning_rate": 2.0059567463194675e-05,
+      "loss": 0.3106,
+      "step": 1609
+    },
+    {
+      "epoch": 2.747440273037543,
+      "grad_norm": 0.3574953575687224,
+      "learning_rate": 2.0035740511734892e-05,
+      "loss": 0.3368,
+      "step": 1610
+    },
+    {
+      "epoch": 2.749146757679181,
+      "grad_norm": 0.3794964523542467,
+      "learning_rate": 2.0011913509547983e-05,
+      "loss": 0.3577,
+      "step": 1611
+    },
+    {
+      "epoch": 2.750853242320819,
+      "grad_norm": 0.3306234888248439,
+      "learning_rate": 1.9988086490452027e-05,
+      "loss": 0.3071,
+      "step": 1612
+    },
+    {
+      "epoch": 2.752559726962457,
+      "grad_norm": 0.3293412215545551,
+      "learning_rate": 1.996425948826512e-05,
+      "loss": 0.3205,
+      "step": 1613
+    },
+    {
+      "epoch": 2.7542662116040955,
+      "grad_norm": 0.3504073187990615,
+      "learning_rate": 1.9940432536805332e-05,
+      "loss": 0.3686,
+      "step": 1614
+    },
+    {
+      "epoch": 2.755972696245734,
+      "grad_norm": 0.38665914863114975,
+      "learning_rate": 1.9916605669890662e-05,
+      "loss": 0.3348,
+      "step": 1615
+    },
+    {
+      "epoch": 2.757679180887372,
+      "grad_norm": 0.351258212938779,
+      "learning_rate": 1.9892778921338994e-05,
+      "loss": 0.3601,
+      "step": 1616
+    },
+    {
+      "epoch": 2.75938566552901,
+      "grad_norm": 0.3397611048333248,
+      "learning_rate": 1.986895232496803e-05,
+      "loss": 0.3313,
+      "step": 1617
+    },
+    {
+      "epoch": 2.7610921501706485,
+      "grad_norm": 0.3458145745845894,
+      "learning_rate": 1.9845125914595283e-05,
+      "loss": 0.3112,
+      "step": 1618
+    },
+    {
+      "epoch": 2.762798634812287,
+      "grad_norm": 0.3659654358977534,
+      "learning_rate": 1.9821299724037963e-05,
+      "loss": 0.313,
+      "step": 1619
+    },
+    {
+      "epoch": 2.7645051194539247,
+      "grad_norm": 0.38214107263163416,
+      "learning_rate": 1.9797473787113004e-05,
+      "loss": 0.3483,
+      "step": 1620
+    },
+    {
+      "epoch": 2.766211604095563,
+      "grad_norm": 0.364102572664005,
+      "learning_rate": 1.977364813763696e-05,
+      "loss": 0.3597,
+      "step": 1621
+    },
+    {
+      "epoch": 2.7679180887372015,
+      "grad_norm": 0.34318240802525257,
+      "learning_rate": 1.9749822809425984e-05,
+      "loss": 0.3345,
+      "step": 1622
+    },
+    {
+      "epoch": 2.76962457337884,
+      "grad_norm": 0.3389410304007314,
+      "learning_rate": 1.9725997836295774e-05,
+      "loss": 0.341,
+      "step": 1623
+    },
+    {
+      "epoch": 2.7713310580204777,
+      "grad_norm": 0.35781642598500735,
+      "learning_rate": 1.9702173252061523e-05,
+      "loss": 0.3224,
+      "step": 1624
+    },
+    {
+      "epoch": 2.773037542662116,
+      "grad_norm": 0.3323242259119811,
+      "learning_rate": 1.9678349090537863e-05,
+      "loss": 0.3277,
+      "step": 1625
+    },
+    {
+      "epoch": 2.774744027303754,
+      "grad_norm": 0.3864060483057039,
+      "learning_rate": 1.9654525385538835e-05,
+      "loss": 0.3512,
+      "step": 1626
+    },
+    {
+      "epoch": 2.7764505119453924,
+      "grad_norm": 0.335503887449825,
+      "learning_rate": 1.963070217087784e-05,
+      "loss": 0.3151,
+      "step": 1627
+    },
+    {
+      "epoch": 2.7781569965870307,
+      "grad_norm": 0.3273278255736187,
+      "learning_rate": 1.9606879480367566e-05,
+      "loss": 0.3143,
+      "step": 1628
+    },
+    {
+      "epoch": 2.779863481228669,
+      "grad_norm": 0.31165697054945707,
+      "learning_rate": 1.9583057347819966e-05,
+      "loss": 0.356,
+      "step": 1629
+    },
+    {
+      "epoch": 2.781569965870307,
+      "grad_norm": 0.7722818020380438,
+      "learning_rate": 1.95592358070462e-05,
+      "loss": 0.3595,
+      "step": 1630
+    },
+    {
+      "epoch": 2.7832764505119454,
+      "grad_norm": 0.3569523719652423,
+      "learning_rate": 1.9535414891856594e-05,
+      "loss": 0.3354,
+      "step": 1631
+    },
+    {
+      "epoch": 2.7849829351535837,
+      "grad_norm": 0.41552819027089377,
+      "learning_rate": 1.9511594636060572e-05,
+      "loss": 0.3636,
+      "step": 1632
+    },
+    {
+      "epoch": 2.7866894197952217,
+      "grad_norm": 0.37048344346462647,
+      "learning_rate": 1.9487775073466632e-05,
+      "loss": 0.354,
+      "step": 1633
+    },
+    {
+      "epoch": 2.78839590443686,
+      "grad_norm": 0.34630389475085166,
+      "learning_rate": 1.9463956237882286e-05,
+      "loss": 0.2968,
+      "step": 1634
+    },
+    {
+      "epoch": 2.7901023890784984,
+      "grad_norm": 0.3428750144611517,
+      "learning_rate": 1.944013816311402e-05,
+      "loss": 0.3554,
+      "step": 1635
+    },
+    {
+      "epoch": 2.7918088737201368,
+      "grad_norm": 0.35617944498917076,
+      "learning_rate": 1.941632088296723e-05,
+      "loss": 0.3153,
+      "step": 1636
+    },
+    {
+      "epoch": 2.7935153583617747,
+      "grad_norm": 0.3612441828462749,
+      "learning_rate": 1.93925044312462e-05,
+      "loss": 0.3321,
+      "step": 1637
+    },
+    {
+      "epoch": 2.795221843003413,
+      "grad_norm": 0.3656782507777216,
+      "learning_rate": 1.9368688841754004e-05,
+      "loss": 0.332,
+      "step": 1638
+    },
+    {
+      "epoch": 2.796928327645051,
+      "grad_norm": 0.3753492388272683,
+      "learning_rate": 1.9344874148292535e-05,
+      "loss": 0.3284,
+      "step": 1639
+    },
+    {
+      "epoch": 2.7986348122866893,
+      "grad_norm": 0.3832829508141039,
+      "learning_rate": 1.9321060384662386e-05,
+      "loss": 0.3678,
+      "step": 1640
+    },
+    {
+      "epoch": 2.8003412969283277,
+      "grad_norm": 0.3266071892760176,
+      "learning_rate": 1.929724758466284e-05,
+      "loss": 0.3722,
+      "step": 1641
+    },
+    {
+      "epoch": 2.802047781569966,
+      "grad_norm": 0.35389760032277373,
+      "learning_rate": 1.9273435782091815e-05,
+      "loss": 0.3357,
+      "step": 1642
+    },
+    {
+      "epoch": 2.803754266211604,
+      "grad_norm": 0.40707048073722657,
+      "learning_rate": 1.9249625010745814e-05,
+      "loss": 0.3267,
+      "step": 1643
+    },
+    {
+      "epoch": 2.8054607508532423,
+      "grad_norm": 0.3310891060298458,
+      "learning_rate": 1.9225815304419856e-05,
+      "loss": 0.3418,
+      "step": 1644
+    },
+    {
+      "epoch": 2.8071672354948807,
+      "grad_norm": 0.4085721022507227,
+      "learning_rate": 1.920200669690747e-05,
+      "loss": 0.3156,
+      "step": 1645
+    },
+    {
+      "epoch": 2.8088737201365186,
+      "grad_norm": 0.37154394826080867,
+      "learning_rate": 1.9178199222000626e-05,
+      "loss": 0.3323,
+      "step": 1646
+    },
+    {
+      "epoch": 2.810580204778157,
+      "grad_norm": 0.36333177012949025,
+      "learning_rate": 1.9154392913489677e-05,
+      "loss": 0.3432,
+      "step": 1647
+    },
+    {
+      "epoch": 2.8122866894197953,
+      "grad_norm": 0.382956873945717,
+      "learning_rate": 1.9130587805163323e-05,
+      "loss": 0.3371,
+      "step": 1648
+    },
+    {
+      "epoch": 2.8139931740614337,
+      "grad_norm": 0.40629424241154394,
+      "learning_rate": 1.9106783930808573e-05,
+      "loss": 0.3458,
+      "step": 1649
+    },
+    {
+      "epoch": 2.8156996587030716,
+      "grad_norm": 0.35089230236120816,
+      "learning_rate": 1.908298132421065e-05,
+      "loss": 0.3479,
+      "step": 1650
+    },
+    {
+      "epoch": 2.81740614334471,
+      "grad_norm": 0.39715671994578994,
+      "learning_rate": 1.9059180019153015e-05,
+      "loss": 0.3408,
+      "step": 1651
+    },
+    {
+      "epoch": 2.819112627986348,
+      "grad_norm": 0.34745493026514485,
+      "learning_rate": 1.9035380049417257e-05,
+      "loss": 0.3659,
+      "step": 1652
+    },
+    {
+      "epoch": 2.8208191126279862,
+      "grad_norm": 0.3389434273750528,
+      "learning_rate": 1.9011581448783098e-05,
+      "loss": 0.3319,
+      "step": 1653
+    },
+    {
+      "epoch": 2.8225255972696246,
+      "grad_norm": 0.38343909047401437,
+      "learning_rate": 1.8987784251028284e-05,
+      "loss": 0.337,
+      "step": 1654
+    },
+    {
+      "epoch": 2.824232081911263,
+      "grad_norm": 0.35147212457480564,
+      "learning_rate": 1.8963988489928594e-05,
+      "loss": 0.3379,
+      "step": 1655
+    },
+    {
+      "epoch": 2.825938566552901,
+      "grad_norm": 0.38603576817115554,
+      "learning_rate": 1.894019419925775e-05,
+      "loss": 0.3364,
+      "step": 1656
+    },
+    {
+      "epoch": 2.8276450511945392,
+      "grad_norm": 0.37816149358219,
+      "learning_rate": 1.89164014127874e-05,
+      "loss": 0.3402,
+      "step": 1657
+    },
+    {
+      "epoch": 2.8293515358361776,
+      "grad_norm": 0.3744403182409511,
+      "learning_rate": 1.8892610164287048e-05,
+      "loss": 0.3789,
+      "step": 1658
+    },
+    {
+      "epoch": 2.8310580204778155,
+      "grad_norm": 0.34421138287281045,
+      "learning_rate": 1.8868820487524022e-05,
+      "loss": 0.3275,
+      "step": 1659
+    },
+    {
+      "epoch": 2.832764505119454,
+      "grad_norm": 0.3817176866198686,
+      "learning_rate": 1.884503241626342e-05,
+      "loss": 0.3106,
+      "step": 1660
+    },
+    {
+      "epoch": 2.8344709897610922,
+      "grad_norm": 0.38572147672185014,
+      "learning_rate": 1.8821245984268065e-05,
+      "loss": 0.3052,
+      "step": 1661
+    },
+    {
+      "epoch": 2.8361774744027306,
+      "grad_norm": 0.3575545479916945,
+      "learning_rate": 1.879746122529843e-05,
+      "loss": 0.3366,
+      "step": 1662
+    },
+    {
+      "epoch": 2.8378839590443685,
+      "grad_norm": 0.37463725801175146,
+      "learning_rate": 1.877367817311264e-05,
+      "loss": 0.2992,
+      "step": 1663
+    },
+    {
+      "epoch": 2.839590443686007,
+      "grad_norm": 0.36923264203118955,
+      "learning_rate": 1.8749896861466382e-05,
+      "loss": 0.3733,
+      "step": 1664
+    },
+    {
+      "epoch": 2.841296928327645,
+      "grad_norm": 0.7918302733905488,
+      "learning_rate": 1.8726117324112888e-05,
+      "loss": 0.415,
+      "step": 1665
+    },
+    {
+      "epoch": 2.843003412969283,
+      "grad_norm": 0.3842100597264359,
+      "learning_rate": 1.870233959480285e-05,
+      "loss": 0.3697,
+      "step": 1666
+    },
+    {
+      "epoch": 2.8447098976109215,
+      "grad_norm": 0.3682276116500696,
+      "learning_rate": 1.8678563707284413e-05,
+      "loss": 0.3252,
+      "step": 1667
+    },
+    {
+      "epoch": 2.84641638225256,
+      "grad_norm": 0.34451700978454036,
+      "learning_rate": 1.865478969530311e-05,
+      "loss": 0.3486,
+      "step": 1668
+    },
+    {
+      "epoch": 2.848122866894198,
+      "grad_norm": 0.3724225173806301,
+      "learning_rate": 1.8631017592601785e-05,
+      "loss": 0.3416,
+      "step": 1669
+    },
+    {
+      "epoch": 2.849829351535836,
+      "grad_norm": 0.38269679188125283,
+      "learning_rate": 1.8607247432920595e-05,
+      "loss": 0.355,
+      "step": 1670
+    },
+    {
+      "epoch": 2.8515358361774745,
+      "grad_norm": 0.3332971620180318,
+      "learning_rate": 1.8583479249996938e-05,
+      "loss": 0.3152,
+      "step": 1671
+    },
+    {
+      "epoch": 2.8532423208191124,
+      "grad_norm": 0.3815614723642687,
+      "learning_rate": 1.85597130775654e-05,
+      "loss": 0.3375,
+      "step": 1672
+    },
+    {
+      "epoch": 2.854948805460751,
+      "grad_norm": 0.37556318578572173,
+      "learning_rate": 1.8535948949357713e-05,
+      "loss": 0.3649,
+      "step": 1673
+    },
+    {
+      "epoch": 2.856655290102389,
+      "grad_norm": 0.33362137910336215,
+      "learning_rate": 1.8512186899102723e-05,
+      "loss": 0.3464,
+      "step": 1674
+    },
+    {
+      "epoch": 2.8583617747440275,
+      "grad_norm": 0.37871549210871974,
+      "learning_rate": 1.8488426960526297e-05,
+      "loss": 0.3092,
+      "step": 1675
+    },
+    {
+      "epoch": 2.8600682593856654,
+      "grad_norm": 0.3569331714652457,
+      "learning_rate": 1.8464669167351333e-05,
+      "loss": 0.3316,
+      "step": 1676
+    },
+    {
+      "epoch": 2.861774744027304,
+      "grad_norm": 0.3996301598780719,
+      "learning_rate": 1.8440913553297666e-05,
+      "loss": 0.3309,
+      "step": 1677
+    },
+    {
+      "epoch": 2.8634812286689417,
+      "grad_norm": 0.3716298479462723,
+      "learning_rate": 1.8417160152082053e-05,
+      "loss": 0.3446,
+      "step": 1678
+    },
+    {
+      "epoch": 2.86518771331058,
+      "grad_norm": 0.33360159229287817,
+      "learning_rate": 1.8393408997418098e-05,
+      "loss": 0.3164,
+      "step": 1679
+    },
+    {
+      "epoch": 2.8668941979522184,
+      "grad_norm": 0.3524761121384744,
+      "learning_rate": 1.8369660123016227e-05,
+      "loss": 0.3283,
+      "step": 1680
+    },
+    {
+      "epoch": 2.868600682593857,
+      "grad_norm": 0.3101343794372301,
+      "learning_rate": 1.834591356258361e-05,
+      "loss": 0.3213,
+      "step": 1681
+    },
+    {
+      "epoch": 2.8703071672354947,
+      "grad_norm": 0.38322792602580813,
+      "learning_rate": 1.8322169349824157e-05,
+      "loss": 0.3386,
+      "step": 1682
+    },
+    {
+      "epoch": 2.872013651877133,
+      "grad_norm": 0.3526255439636646,
+      "learning_rate": 1.8298427518438433e-05,
+      "loss": 0.3619,
+      "step": 1683
+    },
+    {
+      "epoch": 2.8737201365187715,
+      "grad_norm": 0.34664419024831267,
+      "learning_rate": 1.8274688102123622e-05,
+      "loss": 0.333,
+      "step": 1684
+    },
+    {
+      "epoch": 2.8754266211604094,
+      "grad_norm": 0.3696462343494238,
+      "learning_rate": 1.825095113457349e-05,
+      "loss": 0.327,
+      "step": 1685
+    },
+    {
+      "epoch": 2.8771331058020477,
+      "grad_norm": 0.35472191617978754,
+      "learning_rate": 1.822721664947832e-05,
+      "loss": 0.3333,
+      "step": 1686
+    },
+    {
+      "epoch": 2.878839590443686,
+      "grad_norm": 0.33826733032433637,
+      "learning_rate": 1.8203484680524863e-05,
+      "loss": 0.3613,
+      "step": 1687
+    },
+    {
+      "epoch": 2.8805460750853245,
+      "grad_norm": 0.3063764356850511,
+      "learning_rate": 1.8179755261396318e-05,
+      "loss": 0.3553,
+      "step": 1688
+    },
+    {
+      "epoch": 2.8822525597269624,
+      "grad_norm": 0.35193178090811733,
+      "learning_rate": 1.815602842577225e-05,
+      "loss": 0.3063,
+      "step": 1689
+    },
+    {
+      "epoch": 2.8839590443686007,
+      "grad_norm": 0.3516029177930882,
+      "learning_rate": 1.8132304207328566e-05,
+      "loss": 0.3261,
+      "step": 1690
+    },
+    {
+      "epoch": 2.8856655290102387,
+      "grad_norm": 0.3456046758424527,
+      "learning_rate": 1.8108582639737455e-05,
+      "loss": 0.2941,
+      "step": 1691
+    },
+    {
+      "epoch": 2.887372013651877,
+      "grad_norm": 0.334817789772746,
+      "learning_rate": 1.808486375666734e-05,
+      "loss": 0.3105,
+      "step": 1692
+    },
+    {
+      "epoch": 2.8890784982935154,
+      "grad_norm": 0.38880289358555925,
+      "learning_rate": 1.8061147591782842e-05,
+      "loss": 0.2991,
+      "step": 1693
+    },
+    {
+      "epoch": 2.8907849829351537,
+      "grad_norm": 0.38970751626981504,
+      "learning_rate": 1.8037434178744712e-05,
+      "loss": 0.3139,
+      "step": 1694
+    },
+    {
+      "epoch": 2.8924914675767917,
+      "grad_norm": 0.36346955054676744,
+      "learning_rate": 1.801372355120981e-05,
+      "loss": 0.3581,
+      "step": 1695
+    },
+    {
+      "epoch": 2.89419795221843,
+      "grad_norm": 0.3554126585708228,
+      "learning_rate": 1.799001574283103e-05,
+      "loss": 0.3126,
+      "step": 1696
+    },
+    {
+      "epoch": 2.8959044368600684,
+      "grad_norm": 0.35115634959577685,
+      "learning_rate": 1.796631078725727e-05,
+      "loss": 0.3256,
+      "step": 1697
+    },
+    {
+      "epoch": 2.8976109215017063,
+      "grad_norm": 0.3568065378240726,
+      "learning_rate": 1.794260871813339e-05,
+      "loss": 0.3343,
+      "step": 1698
+    },
+    {
+      "epoch": 2.8993174061433447,
+      "grad_norm": 0.34053271734915225,
+      "learning_rate": 1.7918909569100126e-05,
+      "loss": 0.3137,
+      "step": 1699
+    },
+    {
+      "epoch": 2.901023890784983,
+      "grad_norm": 0.2999742909350015,
+      "learning_rate": 1.789521337379409e-05,
+      "loss": 0.3283,
+      "step": 1700
+    },
+    {
+      "epoch": 2.9027303754266214,
+      "grad_norm": 0.379850473840393,
+      "learning_rate": 1.7871520165847704e-05,
+      "loss": 0.3117,
+      "step": 1701
+    },
+    {
+      "epoch": 2.9044368600682593,
+      "grad_norm": 0.4541116141212034,
+      "learning_rate": 1.7847829978889134e-05,
+      "loss": 0.346,
+      "step": 1702
+    },
+    {
+      "epoch": 2.9061433447098977,
+      "grad_norm": 0.3513194920970446,
+      "learning_rate": 1.782414284654227e-05,
+      "loss": 0.3692,
+      "step": 1703
+    },
+    {
+      "epoch": 2.9078498293515356,
+      "grad_norm": 0.3433079443982604,
+      "learning_rate": 1.780045880242667e-05,
+      "loss": 0.3309,
+      "step": 1704
+    },
+    {
+      "epoch": 2.909556313993174,
+      "grad_norm": 0.3711508304255128,
+      "learning_rate": 1.77767778801575e-05,
+      "loss": 0.3495,
+      "step": 1705
+    },
+    {
+      "epoch": 2.9112627986348123,
+      "grad_norm": 0.33974962485669125,
+      "learning_rate": 1.7753100113345495e-05,
+      "loss": 0.3435,
+      "step": 1706
+    },
+    {
+      "epoch": 2.9129692832764507,
+      "grad_norm": 0.35878700876429676,
+      "learning_rate": 1.7729425535596915e-05,
+      "loss": 0.3442,
+      "step": 1707
+    },
+    {
+      "epoch": 2.9146757679180886,
+      "grad_norm": 0.3514467106273154,
+      "learning_rate": 1.7705754180513492e-05,
+      "loss": 0.3625,
+      "step": 1708
+    },
+    {
+      "epoch": 2.916382252559727,
+      "grad_norm": 0.3266246564680988,
+      "learning_rate": 1.7682086081692384e-05,
+      "loss": 0.3608,
+      "step": 1709
+    },
+    {
+      "epoch": 2.9180887372013653,
+      "grad_norm": 0.3870515552127354,
+      "learning_rate": 1.7658421272726135e-05,
+      "loss": 0.328,
+      "step": 1710
+    },
+    {
+      "epoch": 2.919795221843003,
+      "grad_norm": 0.3295969332551737,
+      "learning_rate": 1.7634759787202616e-05,
+      "loss": 0.3068,
+      "step": 1711
+    },
+    {
+      "epoch": 2.9215017064846416,
+      "grad_norm": 0.3576330018796421,
+      "learning_rate": 1.7611101658704957e-05,
+      "loss": 0.3187,
+      "step": 1712
+    },
+    {
+      "epoch": 2.92320819112628,
+      "grad_norm": 0.3390664369715398,
+      "learning_rate": 1.7587446920811563e-05,
+      "loss": 0.3558,
+      "step": 1713
+    },
+    {
+      "epoch": 2.9249146757679183,
+      "grad_norm": 0.34732048934447574,
+      "learning_rate": 1.7563795607096e-05,
+      "loss": 0.3278,
+      "step": 1714
+    },
+    {
+      "epoch": 2.926621160409556,
+      "grad_norm": 0.35650434645488815,
+      "learning_rate": 1.7540147751126988e-05,
+      "loss": 0.329,
+      "step": 1715
+    },
+    {
+      "epoch": 2.9283276450511946,
+      "grad_norm": 0.36947783277043117,
+      "learning_rate": 1.7516503386468332e-05,
+      "loss": 0.3957,
+      "step": 1716
+    },
+    {
+      "epoch": 2.9300341296928325,
+      "grad_norm": 0.3782148284432186,
+      "learning_rate": 1.7492862546678885e-05,
+      "loss": 0.3592,
+      "step": 1717
+    },
+    {
+      "epoch": 2.931740614334471,
+      "grad_norm": 0.35540209250899896,
+      "learning_rate": 1.7469225265312485e-05,
+      "loss": 0.3571,
+      "step": 1718
+    },
+    {
+      "epoch": 2.9334470989761092,
+      "grad_norm": 0.34653886642231185,
+      "learning_rate": 1.744559157591793e-05,
+      "loss": 0.3562,
+      "step": 1719
+    },
+    {
+      "epoch": 2.9351535836177476,
+      "grad_norm": 0.3258636850661726,
+      "learning_rate": 1.7421961512038935e-05,
+      "loss": 0.3267,
+      "step": 1720
+    },
+    {
+      "epoch": 2.9368600682593855,
+      "grad_norm": 0.34482993782625104,
+      "learning_rate": 1.739833510721404e-05,
+      "loss": 0.3169,
+      "step": 1721
+    },
+    {
+      "epoch": 2.938566552901024,
+      "grad_norm": 0.37757768964280464,
+      "learning_rate": 1.737471239497661e-05,
+      "loss": 0.322,
+      "step": 1722
+    },
+    {
+      "epoch": 2.9402730375426622,
+      "grad_norm": 0.34239014533985274,
+      "learning_rate": 1.7351093408854772e-05,
+      "loss": 0.33,
+      "step": 1723
+    },
+    {
+      "epoch": 2.9419795221843,
+      "grad_norm": 0.3646764896000102,
+      "learning_rate": 1.7327478182371336e-05,
+      "loss": 0.3115,
+      "step": 1724
+    },
+    {
+      "epoch": 2.9436860068259385,
+      "grad_norm": 0.3231747953703291,
+      "learning_rate": 1.7303866749043814e-05,
+      "loss": 0.3576,
+      "step": 1725
+    },
+    {
+      "epoch": 2.945392491467577,
+      "grad_norm": 0.344845352925776,
+      "learning_rate": 1.728025914238431e-05,
+      "loss": 0.3588,
+      "step": 1726
+    },
+    {
+      "epoch": 2.9470989761092152,
+      "grad_norm": 0.35901698400371646,
+      "learning_rate": 1.7256655395899504e-05,
+      "loss": 0.3351,
+      "step": 1727
+    },
+    {
+      "epoch": 2.948805460750853,
+      "grad_norm": 0.35284111975516075,
+      "learning_rate": 1.7233055543090603e-05,
+      "loss": 0.3909,
+      "step": 1728
+    },
+    {
+      "epoch": 2.9505119453924915,
+      "grad_norm": 0.32083376472315694,
+      "learning_rate": 1.7209459617453286e-05,
+      "loss": 0.3374,
+      "step": 1729
+    },
+    {
+      "epoch": 2.9522184300341294,
+      "grad_norm": 0.36171096746041753,
+      "learning_rate": 1.7185867652477635e-05,
+      "loss": 0.3213,
+      "step": 1730
+    },
+    {
+      "epoch": 2.953924914675768,
+      "grad_norm": 0.36505654380300484,
+      "learning_rate": 1.716227968164814e-05,
+      "loss": 0.3379,
+      "step": 1731
+    },
+    {
+      "epoch": 2.955631399317406,
+      "grad_norm": 0.40374314394985816,
+      "learning_rate": 1.713869573844361e-05,
+      "loss": 0.3448,
+      "step": 1732
+    },
+    {
+      "epoch": 2.9573378839590445,
+      "grad_norm": 0.33805725343076615,
+      "learning_rate": 1.7115115856337136e-05,
+      "loss": 0.3998,
+      "step": 1733
+    },
+    {
+      "epoch": 2.9590443686006824,
+      "grad_norm": 0.34102697790356706,
+      "learning_rate": 1.7091540068796057e-05,
+      "loss": 0.3498,
+      "step": 1734
+    },
+    {
+      "epoch": 2.960750853242321,
+      "grad_norm": 0.3040015821061206,
+      "learning_rate": 1.7067968409281884e-05,
+      "loss": 0.3166,
+      "step": 1735
+    },
+    {
+      "epoch": 2.962457337883959,
+      "grad_norm": 0.3590008059595252,
+      "learning_rate": 1.704440091125029e-05,
+      "loss": 0.3588,
+      "step": 1736
+    },
+    {
+      "epoch": 2.964163822525597,
+      "grad_norm": 0.3493592992297364,
+      "learning_rate": 1.7020837608151e-05,
+      "loss": 0.3254,
+      "step": 1737
+    },
+    {
+      "epoch": 2.9658703071672354,
+      "grad_norm": 0.31100361374726077,
+      "learning_rate": 1.6997278533427835e-05,
+      "loss": 0.3252,
+      "step": 1738
+    },
+    {
+      "epoch": 2.967576791808874,
+      "grad_norm": 0.3598789653436826,
+      "learning_rate": 1.6973723720518588e-05,
+      "loss": 0.3709,
+      "step": 1739
+    },
+    {
+      "epoch": 2.969283276450512,
+      "grad_norm": 0.34279502689764196,
+      "learning_rate": 1.6950173202854998e-05,
+      "loss": 0.2954,
+      "step": 1740
+    },
+    {
+      "epoch": 2.97098976109215,
+      "grad_norm": 0.32542441687699925,
+      "learning_rate": 1.692662701386273e-05,
+      "loss": 0.3274,
+      "step": 1741
+    },
+    {
+      "epoch": 2.9726962457337884,
+      "grad_norm": 0.3601029282375435,
+      "learning_rate": 1.69030851869613e-05,
+      "loss": 0.3437,
+      "step": 1742
+    },
+    {
+      "epoch": 2.9744027303754264,
+      "grad_norm": 0.34962987836091736,
+      "learning_rate": 1.6879547755564002e-05,
+      "loss": 0.3137,
+      "step": 1743
+    },
+    {
+      "epoch": 2.9761092150170647,
+      "grad_norm": 0.34961152926425815,
+      "learning_rate": 1.6856014753077926e-05,
+      "loss": 0.3414,
+      "step": 1744
+    },
+    {
+      "epoch": 2.977815699658703,
+      "grad_norm": 0.36859741563002496,
+      "learning_rate": 1.6832486212903866e-05,
+      "loss": 0.3391,
+      "step": 1745
+    },
+    {
+      "epoch": 2.9795221843003414,
+      "grad_norm": 0.3765022977379053,
+      "learning_rate": 1.6808962168436283e-05,
+      "loss": 0.3483,
+      "step": 1746
+    },
+    {
+      "epoch": 2.98122866894198,
+      "grad_norm": 0.3677976039002658,
+      "learning_rate": 1.6785442653063248e-05,
+      "loss": 0.3081,
+      "step": 1747
+    },
+    {
+      "epoch": 2.9829351535836177,
+      "grad_norm": 0.3467096282103713,
+      "learning_rate": 1.6761927700166426e-05,
+      "loss": 0.2911,
+      "step": 1748
+    },
+    {
+      "epoch": 2.984641638225256,
+      "grad_norm": 0.3324479185783913,
+      "learning_rate": 1.6738417343120977e-05,
+      "loss": 0.3371,
+      "step": 1749
+    },
+    {
+      "epoch": 2.986348122866894,
+      "grad_norm": 0.3853773281218314,
+      "learning_rate": 1.6714911615295556e-05,
+      "loss": 0.3381,
+      "step": 1750
+    },
+    {
+      "epoch": 2.9880546075085324,
+      "grad_norm": 0.3832177388719795,
+      "learning_rate": 1.6691410550052247e-05,
+      "loss": 0.3019,
+      "step": 1751
+    },
+    {
+      "epoch": 2.9897610921501707,
+      "grad_norm": 0.4141979351193689,
+      "learning_rate": 1.6667914180746512e-05,
+      "loss": 0.296,
+      "step": 1752
+    },
+    {
+      "epoch": 2.991467576791809,
+      "grad_norm": 0.3230747420619479,
+      "learning_rate": 1.664442254072715e-05,
+      "loss": 0.3179,
+      "step": 1753
+    },
+    {
+      "epoch": 2.993174061433447,
+      "grad_norm": 0.3696078043835276,
+      "learning_rate": 1.6620935663336256e-05,
+      "loss": 0.3791,
+      "step": 1754
+    },
+    {
+      "epoch": 2.9948805460750854,
+      "grad_norm": 0.3503901883425271,
+      "learning_rate": 1.659745358190914e-05,
+      "loss": 0.3818,
+      "step": 1755
+    },
+    {
+      "epoch": 2.9965870307167233,
+      "grad_norm": 0.31827258026790567,
+      "learning_rate": 1.6573976329774333e-05,
+      "loss": 0.316,
+      "step": 1756
+    },
+    {
+      "epoch": 2.9982935153583616,
+      "grad_norm": 0.35080157401751505,
+      "learning_rate": 1.6550503940253495e-05,
+      "loss": 0.318,
+      "step": 1757
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.3239978521112306,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.3189,
+      "step": 1758
+    },
+    {
+      "epoch": 3.0017064846416384,
+      "grad_norm": 0.5023542824981905,
+      "learning_rate": 1.6503573882305844e-05,
+      "loss": 0.2291,
+      "step": 1759
+    },
+    {
+      "epoch": 3.0034129692832763,
+      "grad_norm": 0.3622111307251838,
+      "learning_rate": 1.6480116280487668e-05,
+      "loss": 0.261,
+      "step": 1760
+    },
+    {
+      "epoch": 3.0051194539249146,
+      "grad_norm": 0.4661124368191468,
+      "learning_rate": 1.6456663674500627e-05,
+      "loss": 0.2145,
+      "step": 1761
+    },
+    {
+      "epoch": 3.006825938566553,
+      "grad_norm": 0.5106325649604367,
+      "learning_rate": 1.643321609763142e-05,
+      "loss": 0.2429,
+      "step": 1762
+    },
+    {
+      "epoch": 3.008532423208191,
+      "grad_norm": 0.3938572896434494,
+      "learning_rate": 1.6409773583159588e-05,
+      "loss": 0.2594,
+      "step": 1763
+    },
+    {
+      "epoch": 3.0102389078498293,
+      "grad_norm": 0.430218824863444,
+      "learning_rate": 1.638633616435749e-05,
+      "loss": 0.2301,
+      "step": 1764
+    },
+    {
+      "epoch": 3.0119453924914676,
+      "grad_norm": 0.44749877614295497,
+      "learning_rate": 1.6362903874490263e-05,
+      "loss": 0.2239,
+      "step": 1765
+    },
+    {
+      "epoch": 3.013651877133106,
+      "grad_norm": 0.3865086318406251,
+      "learning_rate": 1.6339476746815756e-05,
+      "loss": 0.2755,
+      "step": 1766
+    },
+    {
+      "epoch": 3.015358361774744,
+      "grad_norm": 0.3662040320567563,
+      "learning_rate": 1.6316054814584483e-05,
+      "loss": 0.2271,
+      "step": 1767
+    },
+    {
+      "epoch": 3.0170648464163823,
+      "grad_norm": 0.39800939298770655,
+      "learning_rate": 1.6292638111039597e-05,
+      "loss": 0.2185,
+      "step": 1768
+    },
+    {
+      "epoch": 3.0187713310580206,
+      "grad_norm": 0.4240591084534608,
+      "learning_rate": 1.6269226669416824e-05,
+      "loss": 0.2504,
+      "step": 1769
+    },
+    {
+      "epoch": 3.0204778156996586,
+      "grad_norm": 0.37156845230020247,
+      "learning_rate": 1.6245820522944427e-05,
+      "loss": 0.2239,
+      "step": 1770
+    },
+    {
+      "epoch": 3.022184300341297,
+      "grad_norm": 0.37450054665593224,
+      "learning_rate": 1.6222419704843154e-05,
+      "loss": 0.2453,
+      "step": 1771
+    },
+    {
+      "epoch": 3.0238907849829353,
+      "grad_norm": 0.3278205522266145,
+      "learning_rate": 1.6199024248326175e-05,
+      "loss": 0.2295,
+      "step": 1772
+    },
+    {
+      "epoch": 3.025597269624573,
+      "grad_norm": 0.36357733046965546,
+      "learning_rate": 1.6175634186599076e-05,
+      "loss": 0.2129,
+      "step": 1773
+    },
+    {
+      "epoch": 3.0273037542662116,
+      "grad_norm": 0.3563255311678045,
+      "learning_rate": 1.6152249552859758e-05,
+      "loss": 0.2638,
+      "step": 1774
+    },
+    {
+      "epoch": 3.02901023890785,
+      "grad_norm": 0.3420762406482644,
+      "learning_rate": 1.6128870380298436e-05,
+      "loss": 0.2645,
+      "step": 1775
+    },
+    {
+      "epoch": 3.030716723549488,
+      "grad_norm": 0.493384909929465,
+      "learning_rate": 1.610549670209757e-05,
+      "loss": 0.2558,
+      "step": 1776
+    },
+    {
+      "epoch": 3.032423208191126,
+      "grad_norm": 0.3565257496434803,
+      "learning_rate": 1.6082128551431818e-05,
+      "loss": 0.265,
+      "step": 1777
+    },
+    {
+      "epoch": 3.0341296928327646,
+      "grad_norm": 0.333066759624748,
+      "learning_rate": 1.6058765961468e-05,
+      "loss": 0.2147,
+      "step": 1778
+    },
+    {
+      "epoch": 3.035836177474403,
+      "grad_norm": 0.46879563123055723,
+      "learning_rate": 1.6035408965365043e-05,
+      "loss": 0.2278,
+      "step": 1779
+    },
+    {
+      "epoch": 3.037542662116041,
+      "grad_norm": 0.35351271163216785,
+      "learning_rate": 1.6012057596273923e-05,
+      "loss": 0.2514,
+      "step": 1780
+    },
+    {
+      "epoch": 3.039249146757679,
+      "grad_norm": 0.33527172164589053,
+      "learning_rate": 1.5988711887337643e-05,
+      "loss": 0.2607,
+      "step": 1781
+    },
+    {
+      "epoch": 3.0409556313993176,
+      "grad_norm": 0.39044995830132523,
+      "learning_rate": 1.596537187169116e-05,
+      "loss": 0.2226,
+      "step": 1782
+    },
+    {
+      "epoch": 3.0426621160409555,
+      "grad_norm": 0.3892517718830978,
+      "learning_rate": 1.594203758246136e-05,
+      "loss": 0.243,
+      "step": 1783
+    },
+    {
+      "epoch": 3.044368600682594,
+      "grad_norm": 0.3495506564595455,
+      "learning_rate": 1.5918709052767004e-05,
+      "loss": 0.2549,
+      "step": 1784
+    },
+    {
+      "epoch": 3.046075085324232,
+      "grad_norm": 0.3506167624324571,
+      "learning_rate": 1.5895386315718675e-05,
+      "loss": 0.2154,
+      "step": 1785
+    },
+    {
+      "epoch": 3.04778156996587,
+      "grad_norm": 0.390251454799263,
+      "learning_rate": 1.587206940441872e-05,
+      "loss": 0.2167,
+      "step": 1786
+    },
+    {
+      "epoch": 3.0494880546075085,
+      "grad_norm": 0.32700113070011627,
+      "learning_rate": 1.584875835196124e-05,
+      "loss": 0.2425,
+      "step": 1787
+    },
+    {
+      "epoch": 3.051194539249147,
+      "grad_norm": 0.36416392434798417,
+      "learning_rate": 1.5825453191432e-05,
+      "loss": 0.23,
+      "step": 1788
+    },
+    {
+      "epoch": 3.0529010238907848,
+      "grad_norm": 0.3494229842343675,
+      "learning_rate": 1.5802153955908425e-05,
+      "loss": 0.2101,
+      "step": 1789
+    },
+    {
+      "epoch": 3.054607508532423,
+      "grad_norm": 0.35812200045695386,
+      "learning_rate": 1.577886067845951e-05,
+      "loss": 0.2645,
+      "step": 1790
+    },
+    {
+      "epoch": 3.0563139931740615,
+      "grad_norm": 0.3669156298724796,
+      "learning_rate": 1.5755573392145814e-05,
+      "loss": 0.245,
+      "step": 1791
+    },
+    {
+      "epoch": 3.0580204778157,
+      "grad_norm": 0.3750095044042606,
+      "learning_rate": 1.573229213001936e-05,
+      "loss": 0.2709,
+      "step": 1792
+    },
+    {
+      "epoch": 3.0597269624573378,
+      "grad_norm": 0.3634012555607957,
+      "learning_rate": 1.5709016925123658e-05,
+      "loss": 0.2636,
+      "step": 1793
+    },
+    {
+      "epoch": 3.061433447098976,
+      "grad_norm": 0.3292429504701851,
+      "learning_rate": 1.5685747810493596e-05,
+      "loss": 0.2754,
+      "step": 1794
+    },
+    {
+      "epoch": 3.0631399317406145,
+      "grad_norm": 0.34796794426658706,
+      "learning_rate": 1.5662484819155434e-05,
+      "loss": 0.2431,
+      "step": 1795
+    },
+    {
+      "epoch": 3.0648464163822524,
+      "grad_norm": 0.3916271190395842,
+      "learning_rate": 1.5639227984126722e-05,
+      "loss": 0.2253,
+      "step": 1796
+    },
+    {
+      "epoch": 3.0665529010238908,
+      "grad_norm": 0.3617107429258443,
+      "learning_rate": 1.5615977338416305e-05,
+      "loss": 0.2331,
+      "step": 1797
+    },
+    {
+      "epoch": 3.068259385665529,
+      "grad_norm": 0.33604907347977464,
+      "learning_rate": 1.55927329150242e-05,
+      "loss": 0.2338,
+      "step": 1798
+    },
+    {
+      "epoch": 3.069965870307167,
+      "grad_norm": 0.35007626103547385,
+      "learning_rate": 1.5569494746941613e-05,
+      "loss": 0.2087,
+      "step": 1799
+    },
+    {
+      "epoch": 3.0716723549488054,
+      "grad_norm": 0.33924243027213385,
+      "learning_rate": 1.5546262867150888e-05,
+      "loss": 0.2349,
+      "step": 1800
+    },
+    {
+      "epoch": 3.073378839590444,
+      "grad_norm": 0.3241741854770072,
+      "learning_rate": 1.5523037308625424e-05,
+      "loss": 0.2304,
+      "step": 1801
+    },
+    {
+      "epoch": 3.0750853242320817,
+      "grad_norm": 0.3261934992254656,
+      "learning_rate": 1.549981810432965e-05,
+      "loss": 0.2494,
+      "step": 1802
+    },
+    {
+      "epoch": 3.07679180887372,
+      "grad_norm": 0.3559673508201443,
+      "learning_rate": 1.5476605287218997e-05,
+      "loss": 0.2463,
+      "step": 1803
+    },
+    {
+      "epoch": 3.0784982935153584,
+      "grad_norm": 0.3685695289688876,
+      "learning_rate": 1.5453398890239784e-05,
+      "loss": 0.239,
+      "step": 1804
+    },
+    {
+      "epoch": 3.080204778156997,
+      "grad_norm": 0.4127238074822048,
+      "learning_rate": 1.5430198946329266e-05,
+      "loss": 0.256,
+      "step": 1805
+    },
+    {
+      "epoch": 3.0819112627986347,
+      "grad_norm": 0.35952788137173813,
+      "learning_rate": 1.540700548841551e-05,
+      "loss": 0.2453,
+      "step": 1806
+    },
+    {
+      "epoch": 3.083617747440273,
+      "grad_norm": 0.369170728501119,
+      "learning_rate": 1.5383818549417397e-05,
+      "loss": 0.2218,
+      "step": 1807
+    },
+    {
+      "epoch": 3.0853242320819114,
+      "grad_norm": 0.3575433147530976,
+      "learning_rate": 1.536063816224454e-05,
+      "loss": 0.2065,
+      "step": 1808
+    },
+    {
+      "epoch": 3.0870307167235493,
+      "grad_norm": 0.3343538027842782,
+      "learning_rate": 1.533746435979726e-05,
+      "loss": 0.2187,
+      "step": 1809
+    },
+    {
+      "epoch": 3.0887372013651877,
+      "grad_norm": 0.3685384119739101,
+      "learning_rate": 1.5314297174966543e-05,
+      "loss": 0.2275,
+      "step": 1810
+    },
+    {
+      "epoch": 3.090443686006826,
+      "grad_norm": 0.37265269793468075,
+      "learning_rate": 1.529113664063395e-05,
+      "loss": 0.2544,
+      "step": 1811
+    },
+    {
+      "epoch": 3.092150170648464,
+      "grad_norm": 0.34520980463000883,
+      "learning_rate": 1.5267982789671636e-05,
+      "loss": 0.2696,
+      "step": 1812
+    },
+    {
+      "epoch": 3.0938566552901023,
+      "grad_norm": 0.3744550897419855,
+      "learning_rate": 1.5244835654942252e-05,
+      "loss": 0.2778,
+      "step": 1813
+    },
+    {
+      "epoch": 3.0955631399317407,
+      "grad_norm": 0.3508486292671562,
+      "learning_rate": 1.5221695269298918e-05,
+      "loss": 0.2563,
+      "step": 1814
+    },
+    {
+      "epoch": 3.0972696245733786,
+      "grad_norm": 0.33880674873198074,
+      "learning_rate": 1.5198561665585192e-05,
+      "loss": 0.2554,
+      "step": 1815
+    },
+    {
+      "epoch": 3.098976109215017,
+      "grad_norm": 0.37550221100309866,
+      "learning_rate": 1.5175434876634994e-05,
+      "loss": 0.2367,
+      "step": 1816
+    },
+    {
+      "epoch": 3.1006825938566553,
+      "grad_norm": 0.3547039204864811,
+      "learning_rate": 1.5152314935272556e-05,
+      "loss": 0.3123,
+      "step": 1817
+    },
+    {
+      "epoch": 3.1023890784982937,
+      "grad_norm": 0.31596627546541817,
+      "learning_rate": 1.5129201874312414e-05,
+      "loss": 0.2216,
+      "step": 1818
+    },
+    {
+      "epoch": 3.1040955631399316,
+      "grad_norm": 0.36454602271429526,
+      "learning_rate": 1.5106095726559328e-05,
+      "loss": 0.2528,
+      "step": 1819
+    },
+    {
+      "epoch": 3.10580204778157,
+      "grad_norm": 0.3710957779992305,
+      "learning_rate": 1.5082996524808251e-05,
+      "loss": 0.234,
+      "step": 1820
+    },
+    {
+      "epoch": 3.1075085324232083,
+      "grad_norm": 0.3640748644132522,
+      "learning_rate": 1.5059904301844272e-05,
+      "loss": 0.2506,
+      "step": 1821
+    },
+    {
+      "epoch": 3.1092150170648463,
+      "grad_norm": 0.37604614593850916,
+      "learning_rate": 1.5036819090442594e-05,
+      "loss": 0.2482,
+      "step": 1822
+    },
+    {
+      "epoch": 3.1109215017064846,
+      "grad_norm": 0.3408757814876876,
+      "learning_rate": 1.501374092336843e-05,
+      "loss": 0.2459,
+      "step": 1823
+    },
+    {
+      "epoch": 3.112627986348123,
+      "grad_norm": 0.3591411983387152,
+      "learning_rate": 1.4990669833377025e-05,
+      "loss": 0.2352,
+      "step": 1824
+    },
+    {
+      "epoch": 3.114334470989761,
+      "grad_norm": 0.32486352500740007,
+      "learning_rate": 1.4967605853213573e-05,
+      "loss": 0.2346,
+      "step": 1825
+    },
+    {
+      "epoch": 3.1160409556313993,
+      "grad_norm": 0.36102677067398153,
+      "learning_rate": 1.4944549015613175e-05,
+      "loss": 0.2392,
+      "step": 1826
+    },
+    {
+      "epoch": 3.1177474402730376,
+      "grad_norm": 0.32565962763060924,
+      "learning_rate": 1.4921499353300795e-05,
+      "loss": 0.2402,
+      "step": 1827
+    },
+    {
+      "epoch": 3.1194539249146755,
+      "grad_norm": 0.35285773270868653,
+      "learning_rate": 1.4898456898991216e-05,
+      "loss": 0.2221,
+      "step": 1828
+    },
+    {
+      "epoch": 3.121160409556314,
+      "grad_norm": 0.3509794023449979,
+      "learning_rate": 1.487542168538898e-05,
+      "loss": 0.2344,
+      "step": 1829
+    },
+    {
+      "epoch": 3.1228668941979523,
+      "grad_norm": 0.3503379450918871,
+      "learning_rate": 1.4852393745188365e-05,
+      "loss": 0.2318,
+      "step": 1830
+    },
+    {
+      "epoch": 3.1245733788395906,
+      "grad_norm": 0.3570192207186589,
+      "learning_rate": 1.4829373111073318e-05,
+      "loss": 0.2502,
+      "step": 1831
+    },
+    {
+      "epoch": 3.1262798634812285,
+      "grad_norm": 0.36819319677663465,
+      "learning_rate": 1.4806359815717416e-05,
+      "loss": 0.2375,
+      "step": 1832
+    },
+    {
+      "epoch": 3.127986348122867,
+      "grad_norm": 0.3421945346630767,
+      "learning_rate": 1.4783353891783829e-05,
+      "loss": 0.2418,
+      "step": 1833
+    },
+    {
+      "epoch": 3.1296928327645053,
+      "grad_norm": 0.33862533559137625,
+      "learning_rate": 1.4760355371925257e-05,
+      "loss": 0.2483,
+      "step": 1834
+    },
+    {
+      "epoch": 3.131399317406143,
+      "grad_norm": 0.38210124875299717,
+      "learning_rate": 1.4737364288783888e-05,
+      "loss": 0.2633,
+      "step": 1835
+    },
+    {
+      "epoch": 3.1331058020477816,
+      "grad_norm": 0.3530735041752546,
+      "learning_rate": 1.4714380674991362e-05,
+      "loss": 0.2592,
+      "step": 1836
+    },
+    {
+      "epoch": 3.13481228668942,
+      "grad_norm": 0.3502007412316067,
+      "learning_rate": 1.4691404563168714e-05,
+      "loss": 0.2427,
+      "step": 1837
+    },
+    {
+      "epoch": 3.136518771331058,
+      "grad_norm": 0.33108415507707345,
+      "learning_rate": 1.4668435985926333e-05,
+      "loss": 0.2272,
+      "step": 1838
+    },
+    {
+      "epoch": 3.138225255972696,
+      "grad_norm": 0.3837508654592469,
+      "learning_rate": 1.4645474975863914e-05,
+      "loss": 0.2858,
+      "step": 1839
+    },
+    {
+      "epoch": 3.1399317406143346,
+      "grad_norm": 0.3649616719606301,
+      "learning_rate": 1.4622521565570416e-05,
+      "loss": 0.2362,
+      "step": 1840
+    },
+    {
+      "epoch": 3.1416382252559725,
+      "grad_norm": 0.3615495063252914,
+      "learning_rate": 1.4599575787623996e-05,
+      "loss": 0.2286,
+      "step": 1841
+    },
+    {
+      "epoch": 3.143344709897611,
+      "grad_norm": 0.3403965739967832,
+      "learning_rate": 1.4576637674591994e-05,
+      "loss": 0.2601,
+      "step": 1842
+    },
+    {
+      "epoch": 3.145051194539249,
+      "grad_norm": 0.35068584616614407,
+      "learning_rate": 1.4553707259030868e-05,
+      "loss": 0.2581,
+      "step": 1843
+    },
+    {
+      "epoch": 3.1467576791808876,
+      "grad_norm": 0.37626957059735494,
+      "learning_rate": 1.4530784573486145e-05,
+      "loss": 0.2172,
+      "step": 1844
+    },
+    {
+      "epoch": 3.1484641638225255,
+      "grad_norm": 0.3379602035732661,
+      "learning_rate": 1.4507869650492388e-05,
+      "loss": 0.2283,
+      "step": 1845
+    },
+    {
+      "epoch": 3.150170648464164,
+      "grad_norm": 0.3446202357655234,
+      "learning_rate": 1.4484962522573139e-05,
+      "loss": 0.2756,
+      "step": 1846
+    },
+    {
+      "epoch": 3.151877133105802,
+      "grad_norm": 0.3587557392379511,
+      "learning_rate": 1.4462063222240876e-05,
+      "loss": 0.2863,
+      "step": 1847
+    },
+    {
+      "epoch": 3.15358361774744,
+      "grad_norm": 0.34599281859684994,
+      "learning_rate": 1.4439171781996963e-05,
+      "loss": 0.2164,
+      "step": 1848
+    },
+    {
+      "epoch": 3.1552901023890785,
+      "grad_norm": 0.36023909082037764,
+      "learning_rate": 1.4416288234331619e-05,
+      "loss": 0.2281,
+      "step": 1849
+    },
+    {
+      "epoch": 3.156996587030717,
+      "grad_norm": 0.32250540221869584,
+      "learning_rate": 1.439341261172385e-05,
+      "loss": 0.2823,
+      "step": 1850
+    },
+    {
+      "epoch": 3.1587030716723548,
+      "grad_norm": 0.36124174178421486,
+      "learning_rate": 1.4370544946641417e-05,
+      "loss": 0.2247,
+      "step": 1851
+    },
+    {
+      "epoch": 3.160409556313993,
+      "grad_norm": 0.3452908730969287,
+      "learning_rate": 1.4347685271540796e-05,
+      "loss": 0.2418,
+      "step": 1852
+    },
+    {
+      "epoch": 3.1621160409556315,
+      "grad_norm": 0.33620773915850727,
+      "learning_rate": 1.4324833618867109e-05,
+      "loss": 0.219,
+      "step": 1853
+    },
+    {
+      "epoch": 3.1638225255972694,
+      "grad_norm": 0.33711611163939365,
+      "learning_rate": 1.4301990021054097e-05,
+      "loss": 0.2226,
+      "step": 1854
+    },
+    {
+      "epoch": 3.1655290102389078,
+      "grad_norm": 0.3348941686990772,
+      "learning_rate": 1.4279154510524067e-05,
+      "loss": 0.2029,
+      "step": 1855
+    },
+    {
+      "epoch": 3.167235494880546,
+      "grad_norm": 0.3653135426903386,
+      "learning_rate": 1.4256327119687856e-05,
+      "loss": 0.2376,
+      "step": 1856
+    },
+    {
+      "epoch": 3.1689419795221845,
+      "grad_norm": 0.3361146834254551,
+      "learning_rate": 1.4233507880944763e-05,
+      "loss": 0.2553,
+      "step": 1857
+    },
+    {
+      "epoch": 3.1706484641638224,
+      "grad_norm": 0.359233530077925,
+      "learning_rate": 1.4210696826682528e-05,
+      "loss": 0.271,
+      "step": 1858
+    },
+    {
+      "epoch": 3.1723549488054608,
+      "grad_norm": 0.34714011682073453,
+      "learning_rate": 1.4187893989277276e-05,
+      "loss": 0.2314,
+      "step": 1859
+    },
+    {
+      "epoch": 3.174061433447099,
+      "grad_norm": 0.34784780024400985,
+      "learning_rate": 1.4165099401093451e-05,
+      "loss": 0.2265,
+      "step": 1860
+    },
+    {
+      "epoch": 3.175767918088737,
+      "grad_norm": 0.3348934881845762,
+      "learning_rate": 1.4142313094483809e-05,
+      "loss": 0.2383,
+      "step": 1861
+    },
+    {
+      "epoch": 3.1774744027303754,
+      "grad_norm": 0.37631062261917697,
+      "learning_rate": 1.4119535101789343e-05,
+      "loss": 0.248,
+      "step": 1862
+    },
+    {
+      "epoch": 3.1791808873720138,
+      "grad_norm": 0.35434134240920667,
+      "learning_rate": 1.409676545533925e-05,
+      "loss": 0.247,
+      "step": 1863
+    },
+    {
+      "epoch": 3.1808873720136517,
+      "grad_norm": 0.31177329554546984,
+      "learning_rate": 1.4074004187450875e-05,
+      "loss": 0.2699,
+      "step": 1864
+    },
+    {
+      "epoch": 3.18259385665529,
+      "grad_norm": 0.3692566348232966,
+      "learning_rate": 1.4051251330429687e-05,
+      "loss": 0.2614,
+      "step": 1865
+    },
+    {
+      "epoch": 3.1843003412969284,
+      "grad_norm": 0.3591545979896418,
+      "learning_rate": 1.402850691656918e-05,
+      "loss": 0.222,
+      "step": 1866
+    },
+    {
+      "epoch": 3.1860068259385663,
+      "grad_norm": 0.32658477622950105,
+      "learning_rate": 1.4005770978150908e-05,
+      "loss": 0.2384,
+      "step": 1867
+    },
+    {
+      "epoch": 3.1877133105802047,
+      "grad_norm": 0.34270865867485484,
+      "learning_rate": 1.3983043547444372e-05,
+      "loss": 0.2669,
+      "step": 1868
+    },
+    {
+      "epoch": 3.189419795221843,
+      "grad_norm": 0.3266859760124682,
+      "learning_rate": 1.3960324656707007e-05,
+      "loss": 0.235,
+      "step": 1869
+    },
+    {
+      "epoch": 3.1911262798634814,
+      "grad_norm": 0.36711686303123836,
+      "learning_rate": 1.3937614338184118e-05,
+      "loss": 0.2142,
+      "step": 1870
+    },
+    {
+      "epoch": 3.1928327645051193,
+      "grad_norm": 0.3429333059841405,
+      "learning_rate": 1.3914912624108859e-05,
+      "loss": 0.2263,
+      "step": 1871
+    },
+    {
+      "epoch": 3.1945392491467577,
+      "grad_norm": 0.3384884011293629,
+      "learning_rate": 1.3892219546702146e-05,
+      "loss": 0.2171,
+      "step": 1872
+    },
+    {
+      "epoch": 3.196245733788396,
+      "grad_norm": 0.3275230728976038,
+      "learning_rate": 1.386953513817265e-05,
+      "loss": 0.2254,
+      "step": 1873
+    },
+    {
+      "epoch": 3.197952218430034,
+      "grad_norm": 0.33125861049669947,
+      "learning_rate": 1.3846859430716754e-05,
+      "loss": 0.2393,
+      "step": 1874
+    },
+    {
+      "epoch": 3.1996587030716723,
+      "grad_norm": 0.3172722221260696,
+      "learning_rate": 1.3824192456518473e-05,
+      "loss": 0.2813,
+      "step": 1875
+    },
+    {
+      "epoch": 3.2013651877133107,
+      "grad_norm": 0.3227587267254146,
+      "learning_rate": 1.3801534247749429e-05,
+      "loss": 0.2456,
+      "step": 1876
+    },
+    {
+      "epoch": 3.2030716723549486,
+      "grad_norm": 0.3382551334984173,
+      "learning_rate": 1.3778884836568805e-05,
+      "loss": 0.2408,
+      "step": 1877
+    },
+    {
+      "epoch": 3.204778156996587,
+      "grad_norm": 0.34909231303508853,
+      "learning_rate": 1.3756244255123306e-05,
+      "loss": 0.2433,
+      "step": 1878
+    },
+    {
+      "epoch": 3.2064846416382253,
+      "grad_norm": 0.3525890311139006,
+      "learning_rate": 1.3733612535547079e-05,
+      "loss": 0.2273,
+      "step": 1879
+    },
+    {
+      "epoch": 3.2081911262798632,
+      "grad_norm": 0.4012175842687751,
+      "learning_rate": 1.3710989709961715e-05,
+      "loss": 0.2228,
+      "step": 1880
+    },
+    {
+      "epoch": 3.2098976109215016,
+      "grad_norm": 0.34042317206269157,
+      "learning_rate": 1.3688375810476187e-05,
+      "loss": 0.2442,
+      "step": 1881
+    },
+    {
+      "epoch": 3.21160409556314,
+      "grad_norm": 0.2984314761610252,
+      "learning_rate": 1.3665770869186786e-05,
+      "loss": 0.2519,
+      "step": 1882
+    },
+    {
+      "epoch": 3.2133105802047783,
+      "grad_norm": 0.36221064689138793,
+      "learning_rate": 1.3643174918177087e-05,
+      "loss": 0.2739,
+      "step": 1883
+    },
+    {
+      "epoch": 3.2150170648464163,
+      "grad_norm": 0.34011365766330337,
+      "learning_rate": 1.3620587989517923e-05,
+      "loss": 0.2128,
+      "step": 1884
+    },
+    {
+      "epoch": 3.2167235494880546,
+      "grad_norm": 0.3374640431132146,
+      "learning_rate": 1.3598010115267291e-05,
+      "loss": 0.2466,
+      "step": 1885
+    },
+    {
+      "epoch": 3.218430034129693,
+      "grad_norm": 0.36025845239555393,
+      "learning_rate": 1.3575441327470355e-05,
+      "loss": 0.2444,
+      "step": 1886
+    },
+    {
+      "epoch": 3.220136518771331,
+      "grad_norm": 0.34504373418480144,
+      "learning_rate": 1.3552881658159387e-05,
+      "loss": 0.2134,
+      "step": 1887
+    },
+    {
+      "epoch": 3.2218430034129693,
+      "grad_norm": 0.3602740625682377,
+      "learning_rate": 1.3530331139353714e-05,
+      "loss": 0.2213,
+      "step": 1888
+    },
+    {
+      "epoch": 3.2235494880546076,
+      "grad_norm": 0.3733850968683052,
+      "learning_rate": 1.3507789803059668e-05,
+      "loss": 0.2458,
+      "step": 1889
+    },
+    {
+      "epoch": 3.2252559726962455,
+      "grad_norm": 0.34137926019999193,
+      "learning_rate": 1.3485257681270566e-05,
+      "loss": 0.2335,
+      "step": 1890
+    },
+    {
+      "epoch": 3.226962457337884,
+      "grad_norm": 0.34130405558457455,
+      "learning_rate": 1.3462734805966613e-05,
+      "loss": 0.2215,
+      "step": 1891
+    },
+    {
+      "epoch": 3.2286689419795223,
+      "grad_norm": 0.3575521878547094,
+      "learning_rate": 1.3440221209114923e-05,
+      "loss": 0.2184,
+      "step": 1892
+    },
+    {
+      "epoch": 3.2303754266211606,
+      "grad_norm": 0.3551736749480817,
+      "learning_rate": 1.3417716922669426e-05,
+      "loss": 0.2428,
+      "step": 1893
+    },
+    {
+      "epoch": 3.2320819112627985,
+      "grad_norm": 0.3446580440186442,
+      "learning_rate": 1.3395221978570838e-05,
+      "loss": 0.2775,
+      "step": 1894
+    },
+    {
+      "epoch": 3.233788395904437,
+      "grad_norm": 0.3627323605040583,
+      "learning_rate": 1.3372736408746621e-05,
+      "loss": 0.2369,
+      "step": 1895
+    },
+    {
+      "epoch": 3.2354948805460753,
+      "grad_norm": 0.3498572644209508,
+      "learning_rate": 1.3350260245110937e-05,
+      "loss": 0.2322,
+      "step": 1896
+    },
+    {
+      "epoch": 3.237201365187713,
+      "grad_norm": 0.36722623239785107,
+      "learning_rate": 1.3327793519564578e-05,
+      "loss": 0.2467,
+      "step": 1897
+    },
+    {
+      "epoch": 3.2389078498293515,
+      "grad_norm": 0.34721827634004465,
+      "learning_rate": 1.330533626399495e-05,
+      "loss": 0.2481,
+      "step": 1898
+    },
+    {
+      "epoch": 3.24061433447099,
+      "grad_norm": 0.37345264955293433,
+      "learning_rate": 1.3282888510276026e-05,
+      "loss": 0.2369,
+      "step": 1899
+    },
+    {
+      "epoch": 3.242320819112628,
+      "grad_norm": 0.29310257573596366,
+      "learning_rate": 1.3260450290268287e-05,
+      "loss": 0.2529,
+      "step": 1900
+    },
+    {
+      "epoch": 3.244027303754266,
+      "grad_norm": 0.33081350524602654,
+      "learning_rate": 1.3238021635818678e-05,
+      "loss": 0.2188,
+      "step": 1901
+    },
+    {
+      "epoch": 3.2457337883959045,
+      "grad_norm": 0.3465974170610297,
+      "learning_rate": 1.3215602578760577e-05,
+      "loss": 0.2382,
+      "step": 1902
+    },
+    {
+      "epoch": 3.2474402730375425,
+      "grad_norm": 0.35455396116047055,
+      "learning_rate": 1.3193193150913733e-05,
+      "loss": 0.2152,
+      "step": 1903
+    },
+    {
+      "epoch": 3.249146757679181,
+      "grad_norm": 0.37453236806803086,
+      "learning_rate": 1.3170793384084225e-05,
+      "loss": 0.2339,
+      "step": 1904
+    },
+    {
+      "epoch": 3.250853242320819,
+      "grad_norm": 0.35469744803904324,
+      "learning_rate": 1.3148403310064433e-05,
+      "loss": 0.249,
+      "step": 1905
+    },
+    {
+      "epoch": 3.252559726962457,
+      "grad_norm": 0.36018751994429793,
+      "learning_rate": 1.3126022960632967e-05,
+      "loss": 0.2274,
+      "step": 1906
+    },
+    {
+      "epoch": 3.2542662116040955,
+      "grad_norm": 0.34554921252142795,
+      "learning_rate": 1.3103652367554638e-05,
+      "loss": 0.2548,
+      "step": 1907
+    },
+    {
+      "epoch": 3.255972696245734,
+      "grad_norm": 0.3406967968970072,
+      "learning_rate": 1.308129156258042e-05,
+      "loss": 0.2647,
+      "step": 1908
+    },
+    {
+      "epoch": 3.257679180887372,
+      "grad_norm": 0.34705004086797836,
+      "learning_rate": 1.3058940577447377e-05,
+      "loss": 0.2791,
+      "step": 1909
+    },
+    {
+      "epoch": 3.25938566552901,
+      "grad_norm": 0.32784231106076733,
+      "learning_rate": 1.3036599443878646e-05,
+      "loss": 0.2373,
+      "step": 1910
+    },
+    {
+      "epoch": 3.2610921501706485,
+      "grad_norm": 0.3069148460143671,
+      "learning_rate": 1.3014268193583379e-05,
+      "loss": 0.2348,
+      "step": 1911
+    },
+    {
+      "epoch": 3.262798634812287,
+      "grad_norm": 0.337863834585334,
+      "learning_rate": 1.2991946858256706e-05,
+      "loss": 0.2419,
+      "step": 1912
+    },
+    {
+      "epoch": 3.2645051194539247,
+      "grad_norm": 0.33560728930860845,
+      "learning_rate": 1.2969635469579678e-05,
+      "loss": 0.2447,
+      "step": 1913
+    },
+    {
+      "epoch": 3.266211604095563,
+      "grad_norm": 0.35065106962297304,
+      "learning_rate": 1.2947334059219228e-05,
+      "loss": 0.2296,
+      "step": 1914
+    },
+    {
+      "epoch": 3.2679180887372015,
+      "grad_norm": 0.34540691636026183,
+      "learning_rate": 1.2925042658828133e-05,
+      "loss": 0.2553,
+      "step": 1915
+    },
+    {
+      "epoch": 3.26962457337884,
+      "grad_norm": 0.3371687676374983,
+      "learning_rate": 1.2902761300044955e-05,
+      "loss": 0.2365,
+      "step": 1916
+    },
+    {
+      "epoch": 3.2713310580204777,
+      "grad_norm": 0.3113923658360549,
+      "learning_rate": 1.2880490014494007e-05,
+      "loss": 0.2299,
+      "step": 1917
+    },
+    {
+      "epoch": 3.273037542662116,
+      "grad_norm": 0.3279319065598936,
+      "learning_rate": 1.285822883378531e-05,
+      "loss": 0.2393,
+      "step": 1918
+    },
+    {
+      "epoch": 3.274744027303754,
+      "grad_norm": 0.33565710343698707,
+      "learning_rate": 1.2835977789514534e-05,
+      "loss": 0.2496,
+      "step": 1919
+    },
+    {
+      "epoch": 3.2764505119453924,
+      "grad_norm": 0.3284855761396808,
+      "learning_rate": 1.2813736913262966e-05,
+      "loss": 0.245,
+      "step": 1920
+    },
+    {
+      "epoch": 3.2781569965870307,
+      "grad_norm": 0.3520433933608297,
+      "learning_rate": 1.279150623659747e-05,
+      "loss": 0.2443,
+      "step": 1921
+    },
+    {
+      "epoch": 3.279863481228669,
+      "grad_norm": 0.3064093157127476,
+      "learning_rate": 1.2769285791070418e-05,
+      "loss": 0.222,
+      "step": 1922
+    },
+    {
+      "epoch": 3.281569965870307,
+      "grad_norm": 0.31084278254740383,
+      "learning_rate": 1.2747075608219669e-05,
+      "loss": 0.2589,
+      "step": 1923
+    },
+    {
+      "epoch": 3.2832764505119454,
+      "grad_norm": 0.3346063879894871,
+      "learning_rate": 1.2724875719568513e-05,
+      "loss": 0.2461,
+      "step": 1924
+    },
+    {
+      "epoch": 3.2849829351535837,
+      "grad_norm": 0.3297140693650807,
+      "learning_rate": 1.270268615662564e-05,
+      "loss": 0.2338,
+      "step": 1925
+    },
+    {
+      "epoch": 3.2866894197952217,
+      "grad_norm": 0.3531062928841853,
+      "learning_rate": 1.2680506950885065e-05,
+      "loss": 0.2804,
+      "step": 1926
+    },
+    {
+      "epoch": 3.28839590443686,
+      "grad_norm": 0.39237415889349203,
+      "learning_rate": 1.2658338133826126e-05,
+      "loss": 0.3519,
+      "step": 1927
+    },
+    {
+      "epoch": 3.2901023890784984,
+      "grad_norm": 0.3655010611659869,
+      "learning_rate": 1.2636179736913392e-05,
+      "loss": 0.2332,
+      "step": 1928
+    },
+    {
+      "epoch": 3.2918088737201368,
+      "grad_norm": 0.3526753256875661,
+      "learning_rate": 1.2614031791596663e-05,
+      "loss": 0.2602,
+      "step": 1929
+    },
+    {
+      "epoch": 3.2935153583617747,
+      "grad_norm": 0.3363745793702122,
+      "learning_rate": 1.2591894329310895e-05,
+      "loss": 0.2583,
+      "step": 1930
+    },
+    {
+      "epoch": 3.295221843003413,
+      "grad_norm": 0.3256636494750763,
+      "learning_rate": 1.2569767381476161e-05,
+      "loss": 0.2633,
+      "step": 1931
+    },
+    {
+      "epoch": 3.296928327645051,
+      "grad_norm": 0.3502170397655713,
+      "learning_rate": 1.2547650979497623e-05,
+      "loss": 0.2397,
+      "step": 1932
+    },
+    {
+      "epoch": 3.2986348122866893,
+      "grad_norm": 0.33003496997319587,
+      "learning_rate": 1.2525545154765471e-05,
+      "loss": 0.2458,
+      "step": 1933
+    },
+    {
+      "epoch": 3.3003412969283277,
+      "grad_norm": 0.3579124619667584,
+      "learning_rate": 1.250344993865487e-05,
+      "loss": 0.235,
+      "step": 1934
+    },
+    {
+      "epoch": 3.302047781569966,
+      "grad_norm": 0.34465714702561495,
+      "learning_rate": 1.2481365362525944e-05,
+      "loss": 0.2427,
+      "step": 1935
+    },
+    {
+      "epoch": 3.303754266211604,
+      "grad_norm": 0.35744696182435814,
+      "learning_rate": 1.2459291457723708e-05,
+      "loss": 0.2204,
+      "step": 1936
+    },
+    {
+      "epoch": 3.3054607508532423,
+      "grad_norm": 0.35339834478638066,
+      "learning_rate": 1.2437228255578036e-05,
+      "loss": 0.227,
+      "step": 1937
+    },
+    {
+      "epoch": 3.3071672354948807,
+      "grad_norm": 0.3701262370914736,
+      "learning_rate": 1.2415175787403602e-05,
+      "loss": 0.237,
+      "step": 1938
+    },
+    {
+      "epoch": 3.3088737201365186,
+      "grad_norm": 0.33875916075918705,
+      "learning_rate": 1.239313408449986e-05,
+      "loss": 0.239,
+      "step": 1939
+    },
+    {
+      "epoch": 3.310580204778157,
+      "grad_norm": 0.34045558186040226,
+      "learning_rate": 1.2371103178150965e-05,
+      "loss": 0.3002,
+      "step": 1940
+    },
+    {
+      "epoch": 3.3122866894197953,
+      "grad_norm": 0.338262407610433,
+      "learning_rate": 1.2349083099625764e-05,
+      "loss": 0.2131,
+      "step": 1941
+    },
+    {
+      "epoch": 3.3139931740614337,
+      "grad_norm": 0.34534813428288663,
+      "learning_rate": 1.2327073880177735e-05,
+      "loss": 0.2468,
+      "step": 1942
+    },
+    {
+      "epoch": 3.3156996587030716,
+      "grad_norm": 0.3854984150875541,
+      "learning_rate": 1.2305075551044934e-05,
+      "loss": 0.2383,
+      "step": 1943
+    },
+    {
+      "epoch": 3.31740614334471,
+      "grad_norm": 0.33135716332895926,
+      "learning_rate": 1.2283088143449966e-05,
+      "loss": 0.2351,
+      "step": 1944
+    },
+    {
+      "epoch": 3.319112627986348,
+      "grad_norm": 0.35284557516011333,
+      "learning_rate": 1.2261111688599944e-05,
+      "loss": 0.2827,
+      "step": 1945
+    },
+    {
+      "epoch": 3.3208191126279862,
+      "grad_norm": 0.34011517179545575,
+      "learning_rate": 1.223914621768641e-05,
+      "loss": 0.2559,
+      "step": 1946
+    },
+    {
+      "epoch": 3.3225255972696246,
+      "grad_norm": 0.3265219606050788,
+      "learning_rate": 1.2217191761885339e-05,
+      "loss": 0.2481,
+      "step": 1947
+    },
+    {
+      "epoch": 3.324232081911263,
+      "grad_norm": 0.34878977356026486,
+      "learning_rate": 1.2195248352357067e-05,
+      "loss": 0.2245,
+      "step": 1948
+    },
+    {
+      "epoch": 3.325938566552901,
+      "grad_norm": 0.3375538846496755,
+      "learning_rate": 1.217331602024625e-05,
+      "loss": 0.2398,
+      "step": 1949
+    },
+    {
+      "epoch": 3.3276450511945392,
+      "grad_norm": 0.34580986789699825,
+      "learning_rate": 1.2151394796681826e-05,
+      "loss": 0.2367,
+      "step": 1950
+    },
+    {
+      "epoch": 3.3293515358361776,
+      "grad_norm": 0.3442862008039689,
+      "learning_rate": 1.2129484712776955e-05,
+      "loss": 0.2309,
+      "step": 1951
+    },
+    {
+      "epoch": 3.3310580204778155,
+      "grad_norm": 0.3541233175172183,
+      "learning_rate": 1.2107585799629009e-05,
+      "loss": 0.2557,
+      "step": 1952
+    },
+    {
+      "epoch": 3.332764505119454,
+      "grad_norm": 0.3805049307635654,
+      "learning_rate": 1.2085698088319468e-05,
+      "loss": 0.2529,
+      "step": 1953
+    },
+    {
+      "epoch": 3.3344709897610922,
+      "grad_norm": 0.3577101787428832,
+      "learning_rate": 1.2063821609913941e-05,
+      "loss": 0.2449,
+      "step": 1954
+    },
+    {
+      "epoch": 3.3361774744027306,
+      "grad_norm": 0.32569582925826074,
+      "learning_rate": 1.2041956395462098e-05,
+      "loss": 0.2151,
+      "step": 1955
+    },
+    {
+      "epoch": 3.3378839590443685,
+      "grad_norm": 0.3474692955307898,
+      "learning_rate": 1.20201024759976e-05,
+      "loss": 0.2637,
+      "step": 1956
+    },
+    {
+      "epoch": 3.339590443686007,
+      "grad_norm": 0.34904041376292966,
+      "learning_rate": 1.19982598825381e-05,
+      "loss": 0.2235,
+      "step": 1957
+    },
+    {
+      "epoch": 3.3412969283276452,
+      "grad_norm": 0.3842071665617674,
+      "learning_rate": 1.1976428646085163e-05,
+      "loss": 0.251,
+      "step": 1958
+    },
+    {
+      "epoch": 3.343003412969283,
+      "grad_norm": 0.35094761850865375,
+      "learning_rate": 1.1954608797624225e-05,
+      "loss": 0.2362,
+      "step": 1959
+    },
+    {
+      "epoch": 3.3447098976109215,
+      "grad_norm": 0.32792563908726285,
+      "learning_rate": 1.1932800368124578e-05,
+      "loss": 0.2448,
+      "step": 1960
+    },
+    {
+      "epoch": 3.34641638225256,
+      "grad_norm": 0.36989155687533665,
+      "learning_rate": 1.1911003388539291e-05,
+      "loss": 0.2548,
+      "step": 1961
+    },
+    {
+      "epoch": 3.348122866894198,
+      "grad_norm": 0.34510986347654204,
+      "learning_rate": 1.18892178898052e-05,
+      "loss": 0.2762,
+      "step": 1962
+    },
+    {
+      "epoch": 3.349829351535836,
+      "grad_norm": 0.3267422575767667,
+      "learning_rate": 1.1867443902842832e-05,
+      "loss": 0.2522,
+      "step": 1963
+    },
+    {
+      "epoch": 3.3515358361774745,
+      "grad_norm": 0.3044265017602334,
+      "learning_rate": 1.1845681458556389e-05,
+      "loss": 0.2503,
+      "step": 1964
+    },
+    {
+      "epoch": 3.3532423208191124,
+      "grad_norm": 0.3319703442252075,
+      "learning_rate": 1.1823930587833661e-05,
+      "loss": 0.2703,
+      "step": 1965
+    },
+    {
+      "epoch": 3.354948805460751,
+      "grad_norm": 0.3518177021180182,
+      "learning_rate": 1.1802191321546042e-05,
+      "loss": 0.2934,
+      "step": 1966
+    },
+    {
+      "epoch": 3.356655290102389,
+      "grad_norm": 0.33236123352508423,
+      "learning_rate": 1.1780463690548439e-05,
+      "loss": 0.2333,
+      "step": 1967
+    },
+    {
+      "epoch": 3.3583617747440275,
+      "grad_norm": 0.31062378619814635,
+      "learning_rate": 1.1758747725679252e-05,
+      "loss": 0.2506,
+      "step": 1968
+    },
+    {
+      "epoch": 3.3600682593856654,
+      "grad_norm": 0.3582359855071637,
+      "learning_rate": 1.1737043457760327e-05,
+      "loss": 0.2527,
+      "step": 1969
+    },
+    {
+      "epoch": 3.361774744027304,
+      "grad_norm": 0.3512082083980241,
+      "learning_rate": 1.1715350917596905e-05,
+      "loss": 0.2484,
+      "step": 1970
+    },
+    {
+      "epoch": 3.363481228668942,
+      "grad_norm": 0.35249205053243116,
+      "learning_rate": 1.1693670135977564e-05,
+      "loss": 0.2408,
+      "step": 1971
+    },
+    {
+      "epoch": 3.36518771331058,
+      "grad_norm": 0.3625026984233858,
+      "learning_rate": 1.1672001143674212e-05,
+      "loss": 0.2577,
+      "step": 1972
+    },
+    {
+      "epoch": 3.3668941979522184,
+      "grad_norm": 0.3032013995413711,
+      "learning_rate": 1.1650343971442035e-05,
+      "loss": 0.2443,
+      "step": 1973
+    },
+    {
+      "epoch": 3.368600682593857,
+      "grad_norm": 0.358853915448751,
+      "learning_rate": 1.162869865001941e-05,
+      "loss": 0.2429,
+      "step": 1974
+    },
+    {
+      "epoch": 3.3703071672354947,
+      "grad_norm": 0.3595995927875013,
+      "learning_rate": 1.1607065210127924e-05,
+      "loss": 0.2487,
+      "step": 1975
+    },
+    {
+      "epoch": 3.372013651877133,
+      "grad_norm": 0.3455297346187812,
+      "learning_rate": 1.1585443682472286e-05,
+      "loss": 0.2706,
+      "step": 1976
+    },
+    {
+      "epoch": 3.3737201365187715,
+      "grad_norm": 0.32154204618152804,
+      "learning_rate": 1.156383409774029e-05,
+      "loss": 0.2409,
+      "step": 1977
+    },
+    {
+      "epoch": 3.3754266211604094,
+      "grad_norm": 0.3249442534415948,
+      "learning_rate": 1.1542236486602803e-05,
+      "loss": 0.2343,
+      "step": 1978
+    },
+    {
+      "epoch": 3.3771331058020477,
+      "grad_norm": 0.3597188145479468,
+      "learning_rate": 1.1520650879713667e-05,
+      "loss": 0.2506,
+      "step": 1979
+    },
+    {
+      "epoch": 3.378839590443686,
+      "grad_norm": 0.32518884717067953,
+      "learning_rate": 1.1499077307709723e-05,
+      "loss": 0.2527,
+      "step": 1980
+    },
+    {
+      "epoch": 3.3805460750853245,
+      "grad_norm": 0.3530381665715562,
+      "learning_rate": 1.1477515801210695e-05,
+      "loss": 0.244,
+      "step": 1981
+    },
+    {
+      "epoch": 3.3822525597269624,
+      "grad_norm": 0.3463271542168002,
+      "learning_rate": 1.1455966390819207e-05,
+      "loss": 0.2601,
+      "step": 1982
+    },
+    {
+      "epoch": 3.3839590443686007,
+      "grad_norm": 0.34753476438002995,
+      "learning_rate": 1.1434429107120706e-05,
+      "loss": 0.2605,
+      "step": 1983
+    },
+    {
+      "epoch": 3.385665529010239,
+      "grad_norm": 0.3222606859771089,
+      "learning_rate": 1.1412903980683412e-05,
+      "loss": 0.2643,
+      "step": 1984
+    },
+    {
+      "epoch": 3.387372013651877,
+      "grad_norm": 0.3235759469848687,
+      "learning_rate": 1.1391391042058326e-05,
+      "loss": 0.2409,
+      "step": 1985
+    },
+    {
+      "epoch": 3.3890784982935154,
+      "grad_norm": 0.34987651338927844,
+      "learning_rate": 1.1369890321779111e-05,
+      "loss": 0.2952,
+      "step": 1986
+    },
+    {
+      "epoch": 3.3907849829351537,
+      "grad_norm": 0.3546307720901437,
+      "learning_rate": 1.1348401850362123e-05,
+      "loss": 0.2693,
+      "step": 1987
+    },
+    {
+      "epoch": 3.3924914675767917,
+      "grad_norm": 0.3197376481900684,
+      "learning_rate": 1.1326925658306305e-05,
+      "loss": 0.2303,
+      "step": 1988
+    },
+    {
+      "epoch": 3.39419795221843,
+      "grad_norm": 0.3276668625330492,
+      "learning_rate": 1.1305461776093201e-05,
+      "loss": 0.2117,
+      "step": 1989
+    },
+    {
+      "epoch": 3.3959044368600684,
+      "grad_norm": 0.36466775304996984,
+      "learning_rate": 1.1284010234186837e-05,
+      "loss": 0.2889,
+      "step": 1990
+    },
+    {
+      "epoch": 3.3976109215017063,
+      "grad_norm": 0.3574113143457065,
+      "learning_rate": 1.126257106303377e-05,
+      "loss": 0.231,
+      "step": 1991
+    },
+    {
+      "epoch": 3.3993174061433447,
+      "grad_norm": 0.3191063881050614,
+      "learning_rate": 1.1241144293062987e-05,
+      "loss": 0.2291,
+      "step": 1992
+    },
+    {
+      "epoch": 3.401023890784983,
+      "grad_norm": 0.36727546588401,
+      "learning_rate": 1.1219729954685859e-05,
+      "loss": 0.2558,
+      "step": 1993
+    },
+    {
+      "epoch": 3.4027303754266214,
+      "grad_norm": 0.3488165644525964,
+      "learning_rate": 1.1198328078296132e-05,
+      "loss": 0.2258,
+      "step": 1994
+    },
+    {
+      "epoch": 3.4044368600682593,
+      "grad_norm": 0.3358403662067082,
+      "learning_rate": 1.1176938694269852e-05,
+      "loss": 0.2697,
+      "step": 1995
+    },
+    {
+      "epoch": 3.4061433447098977,
+      "grad_norm": 0.33128698103955495,
+      "learning_rate": 1.1155561832965333e-05,
+      "loss": 0.2049,
+      "step": 1996
+    },
+    {
+      "epoch": 3.407849829351536,
+      "grad_norm": 0.3380251770320218,
+      "learning_rate": 1.1134197524723119e-05,
+      "loss": 0.2352,
+      "step": 1997
+    },
+    {
+      "epoch": 3.409556313993174,
+      "grad_norm": 0.323003213828888,
+      "learning_rate": 1.1112845799865939e-05,
+      "loss": 0.2637,
+      "step": 1998
+    },
+    {
+      "epoch": 3.4112627986348123,
+      "grad_norm": 0.3582840796815567,
+      "learning_rate": 1.1091506688698668e-05,
+      "loss": 0.2423,
+      "step": 1999
+    },
+    {
+      "epoch": 3.4129692832764507,
+      "grad_norm": 0.3531054241705733,
+      "learning_rate": 1.1070180221508262e-05,
+      "loss": 0.2426,
+      "step": 2000
+    },
+    {
+      "epoch": 3.4146757679180886,
+      "grad_norm": 0.3712259718653561,
+      "learning_rate": 1.104886642856376e-05,
+      "loss": 0.206,
+      "step": 2001
+    },
+    {
+      "epoch": 3.416382252559727,
+      "grad_norm": 0.3660185687278045,
+      "learning_rate": 1.1027565340116161e-05,
+      "loss": 0.328,
+      "step": 2002
+    },
+    {
+      "epoch": 3.4180887372013653,
+      "grad_norm": 0.33910687317691934,
+      "learning_rate": 1.1006276986398494e-05,
+      "loss": 0.2502,
+      "step": 2003
+    },
+    {
+      "epoch": 3.419795221843003,
+      "grad_norm": 0.3566924270411525,
+      "learning_rate": 1.0985001397625656e-05,
+      "loss": 0.2381,
+      "step": 2004
+    },
+    {
+      "epoch": 3.4215017064846416,
+      "grad_norm": 0.3663963696457625,
+      "learning_rate": 1.0963738603994472e-05,
+      "loss": 0.2192,
+      "step": 2005
+    },
+    {
+      "epoch": 3.42320819112628,
+      "grad_norm": 0.31861520786270336,
+      "learning_rate": 1.0942488635683593e-05,
+      "loss": 0.2296,
+      "step": 2006
+    },
+    {
+      "epoch": 3.4249146757679183,
+      "grad_norm": 0.3480641618683181,
+      "learning_rate": 1.0921251522853451e-05,
+      "loss": 0.2474,
+      "step": 2007
+    },
+    {
+      "epoch": 3.426621160409556,
+      "grad_norm": 0.3390906993385012,
+      "learning_rate": 1.090002729564625e-05,
+      "loss": 0.2539,
+      "step": 2008
+    },
+    {
+      "epoch": 3.4283276450511946,
+      "grad_norm": 0.4183257078823607,
+      "learning_rate": 1.0878815984185885e-05,
+      "loss": 0.2501,
+      "step": 2009
+    },
+    {
+      "epoch": 3.430034129692833,
+      "grad_norm": 0.34479899996859115,
+      "learning_rate": 1.0857617618577952e-05,
+      "loss": 0.2154,
+      "step": 2010
+    },
+    {
+      "epoch": 3.431740614334471,
+      "grad_norm": 0.36386280275147787,
+      "learning_rate": 1.0836432228909635e-05,
+      "loss": 0.2516,
+      "step": 2011
+    },
+    {
+      "epoch": 3.4334470989761092,
+      "grad_norm": 0.3302625894981507,
+      "learning_rate": 1.0815259845249732e-05,
+      "loss": 0.2189,
+      "step": 2012
+    },
+    {
+      "epoch": 3.4351535836177476,
+      "grad_norm": 0.3762940139922422,
+      "learning_rate": 1.0794100497648583e-05,
+      "loss": 0.2359,
+      "step": 2013
+    },
+    {
+      "epoch": 3.4368600682593855,
+      "grad_norm": 0.36364143241072366,
+      "learning_rate": 1.0772954216137976e-05,
+      "loss": 0.2126,
+      "step": 2014
+    },
+    {
+      "epoch": 3.438566552901024,
+      "grad_norm": 0.3432791677901663,
+      "learning_rate": 1.075182103073122e-05,
+      "loss": 0.2494,
+      "step": 2015
+    },
+    {
+      "epoch": 3.4402730375426622,
+      "grad_norm": 0.37294533024769344,
+      "learning_rate": 1.0730700971422987e-05,
+      "loss": 0.2829,
+      "step": 2016
+    },
+    {
+      "epoch": 3.4419795221843,
+      "grad_norm": 0.3241303234593021,
+      "learning_rate": 1.0709594068189358e-05,
+      "loss": 0.2291,
+      "step": 2017
+    },
+    {
+      "epoch": 3.4436860068259385,
+      "grad_norm": 0.33216601505329185,
+      "learning_rate": 1.0688500350987698e-05,
+      "loss": 0.2066,
+      "step": 2018
+    },
+    {
+      "epoch": 3.445392491467577,
+      "grad_norm": 0.3116101017638262,
+      "learning_rate": 1.0667419849756694e-05,
+      "loss": 0.2605,
+      "step": 2019
+    },
+    {
+      "epoch": 3.4470989761092152,
+      "grad_norm": 0.3460080728945283,
+      "learning_rate": 1.0646352594416281e-05,
+      "loss": 0.2388,
+      "step": 2020
+    },
+    {
+      "epoch": 3.448805460750853,
+      "grad_norm": 0.3376151726400388,
+      "learning_rate": 1.0625298614867536e-05,
+      "loss": 0.2088,
+      "step": 2021
+    },
+    {
+      "epoch": 3.4505119453924915,
+      "grad_norm": 0.33395702364836977,
+      "learning_rate": 1.0604257940992757e-05,
+      "loss": 0.2462,
+      "step": 2022
+    },
+    {
+      "epoch": 3.45221843003413,
+      "grad_norm": 0.3668360281337767,
+      "learning_rate": 1.0583230602655324e-05,
+      "loss": 0.2518,
+      "step": 2023
+    },
+    {
+      "epoch": 3.453924914675768,
+      "grad_norm": 0.34744879848627575,
+      "learning_rate": 1.0562216629699701e-05,
+      "loss": 0.2125,
+      "step": 2024
+    },
+    {
+      "epoch": 3.455631399317406,
+      "grad_norm": 0.3826813486094946,
+      "learning_rate": 1.0541216051951374e-05,
+      "loss": 0.2667,
+      "step": 2025
+    },
+    {
+      "epoch": 3.4573378839590445,
+      "grad_norm": 0.3377490570187433,
+      "learning_rate": 1.052022889921683e-05,
+      "loss": 0.2041,
+      "step": 2026
+    },
+    {
+      "epoch": 3.4590443686006824,
+      "grad_norm": 0.3761261493220408,
+      "learning_rate": 1.0499255201283493e-05,
+      "loss": 0.2508,
+      "step": 2027
+    },
+    {
+      "epoch": 3.460750853242321,
+      "grad_norm": 0.31098876312161366,
+      "learning_rate": 1.047829498791968e-05,
+      "loss": 0.2243,
+      "step": 2028
+    },
+    {
+      "epoch": 3.462457337883959,
+      "grad_norm": 0.3636135752348594,
+      "learning_rate": 1.0457348288874595e-05,
+      "loss": 0.2202,
+      "step": 2029
+    },
+    {
+      "epoch": 3.464163822525597,
+      "grad_norm": 0.32537791997034504,
+      "learning_rate": 1.0436415133878233e-05,
+      "loss": 0.2314,
+      "step": 2030
+    },
+    {
+      "epoch": 3.4658703071672354,
+      "grad_norm": 0.3735986322438891,
+      "learning_rate": 1.041549555264139e-05,
+      "loss": 0.2142,
+      "step": 2031
+    },
+    {
+      "epoch": 3.467576791808874,
+      "grad_norm": 0.3293558147258305,
+      "learning_rate": 1.0394589574855583e-05,
+      "loss": 0.2327,
+      "step": 2032
+    },
+    {
+      "epoch": 3.469283276450512,
+      "grad_norm": 0.3476570831175653,
+      "learning_rate": 1.037369723019301e-05,
+      "loss": 0.2343,
+      "step": 2033
+    },
+    {
+      "epoch": 3.47098976109215,
+      "grad_norm": 0.36212969651527965,
+      "learning_rate": 1.0352818548306554e-05,
+      "loss": 0.2288,
+      "step": 2034
+    },
+    {
+      "epoch": 3.4726962457337884,
+      "grad_norm": 0.3856953437181421,
+      "learning_rate": 1.0331953558829663e-05,
+      "loss": 0.1916,
+      "step": 2035
+    },
+    {
+      "epoch": 3.474402730375427,
+      "grad_norm": 0.34621353791171144,
+      "learning_rate": 1.03111022913764e-05,
+      "loss": 0.1997,
+      "step": 2036
+    },
+    {
+      "epoch": 3.4761092150170647,
+      "grad_norm": 0.3663934082576361,
+      "learning_rate": 1.0290264775541297e-05,
+      "loss": 0.2376,
+      "step": 2037
+    },
+    {
+      "epoch": 3.477815699658703,
+      "grad_norm": 0.35422542247938005,
+      "learning_rate": 1.0269441040899422e-05,
+      "loss": 0.2333,
+      "step": 2038
+    },
+    {
+      "epoch": 3.4795221843003414,
+      "grad_norm": 0.3751510764701517,
+      "learning_rate": 1.0248631117006243e-05,
+      "loss": 0.2322,
+      "step": 2039
+    },
+    {
+      "epoch": 3.4812286689419794,
+      "grad_norm": 0.31375231625716266,
+      "learning_rate": 1.0227835033397638e-05,
+      "loss": 0.2511,
+      "step": 2040
+    },
+    {
+      "epoch": 3.4829351535836177,
+      "grad_norm": 0.31947992957786364,
+      "learning_rate": 1.0207052819589855e-05,
+      "loss": 0.2655,
+      "step": 2041
+    },
+    {
+      "epoch": 3.484641638225256,
+      "grad_norm": 0.3254591679354409,
+      "learning_rate": 1.0186284505079435e-05,
+      "loss": 0.2314,
+      "step": 2042
+    },
+    {
+      "epoch": 3.486348122866894,
+      "grad_norm": 0.3481138793380576,
+      "learning_rate": 1.0165530119343214e-05,
+      "loss": 0.264,
+      "step": 2043
+    },
+    {
+      "epoch": 3.4880546075085324,
+      "grad_norm": 0.345451971216344,
+      "learning_rate": 1.0144789691838239e-05,
+      "loss": 0.2707,
+      "step": 2044
+    },
+    {
+      "epoch": 3.4897610921501707,
+      "grad_norm": 0.3201471895581909,
+      "learning_rate": 1.0124063252001745e-05,
+      "loss": 0.2688,
+      "step": 2045
+    },
+    {
+      "epoch": 3.491467576791809,
+      "grad_norm": 0.32945610263536973,
+      "learning_rate": 1.010335082925114e-05,
+      "loss": 0.2336,
+      "step": 2046
+    },
+    {
+      "epoch": 3.493174061433447,
+      "grad_norm": 0.37693022505129287,
+      "learning_rate": 1.0082652452983902e-05,
+      "loss": 0.2251,
+      "step": 2047
+    },
+    {
+      "epoch": 3.4948805460750854,
+      "grad_norm": 0.3724222324989047,
+      "learning_rate": 1.006196815257761e-05,
+      "loss": 0.2363,
+      "step": 2048
+    },
+    {
+      "epoch": 3.4965870307167237,
+      "grad_norm": 0.3427774678713566,
+      "learning_rate": 1.0041297957389826e-05,
+      "loss": 0.249,
+      "step": 2049
+    },
+    {
+      "epoch": 3.4982935153583616,
+      "grad_norm": 0.33088959732927303,
+      "learning_rate": 1.0020641896758127e-05,
+      "loss": 0.237,
+      "step": 2050
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.36103123726832864,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.2115,
+      "step": 2051
+    },
+    {
+      "epoch": 3.5017064846416384,
+      "grad_norm": 0.33607964075042507,
+      "learning_rate": 9.97937229641285e-06,
+      "loss": 0.2769,
+      "step": 2052
+    },
+    {
+      "epoch": 3.5034129692832767,
+      "grad_norm": 0.3160307593162131,
+      "learning_rate": 9.958758815273932e-06,
+      "loss": 0.264,
+      "step": 2053
+    },
+    {
+      "epoch": 3.5051194539249146,
+      "grad_norm": 0.3394635698134081,
+      "learning_rate": 9.9381595858403e-06,
+      "loss": 0.2331,
+      "step": 2054
+    },
+    {
+      "epoch": 3.506825938566553,
+      "grad_norm": 0.34937552515141296,
+      "learning_rate": 9.917574637348806e-06,
+      "loss": 0.2599,
+      "step": 2055
+    },
+    {
+      "epoch": 3.508532423208191,
+      "grad_norm": 0.4101937972209689,
+      "learning_rate": 9.897003999016006e-06,
+      "loss": 0.2768,
+      "step": 2056
+    },
+    {
+      "epoch": 3.5102389078498293,
+      "grad_norm": 0.35787731128662253,
+      "learning_rate": 9.876447700038175e-06,
+      "loss": 0.2661,
+      "step": 2057
+    },
+    {
+      "epoch": 3.5119453924914676,
+      "grad_norm": 0.3443731772940693,
+      "learning_rate": 9.85590576959121e-06,
+      "loss": 0.2497,
+      "step": 2058
+    },
+    {
+      "epoch": 3.513651877133106,
+      "grad_norm": 0.3993109958736273,
+      "learning_rate": 9.835378236830618e-06,
+      "loss": 0.22,
+      "step": 2059
+    },
+    {
+      "epoch": 3.515358361774744,
+      "grad_norm": 0.3605083597229052,
+      "learning_rate": 9.814865130891489e-06,
+      "loss": 0.2394,
+      "step": 2060
+    },
+    {
+      "epoch": 3.5170648464163823,
+      "grad_norm": 0.3608372999361292,
+      "learning_rate": 9.794366480888415e-06,
+      "loss": 0.271,
+      "step": 2061
+    },
+    {
+      "epoch": 3.51877133105802,
+      "grad_norm": 0.31714881311600096,
+      "learning_rate": 9.773882315915494e-06,
+      "loss": 0.2183,
+      "step": 2062
+    },
+    {
+      "epoch": 3.5204778156996586,
+      "grad_norm": 0.3373085642605433,
+      "learning_rate": 9.75341266504624e-06,
+      "loss": 0.2594,
+      "step": 2063
+    },
+    {
+      "epoch": 3.522184300341297,
+      "grad_norm": 0.3335303524801546,
+      "learning_rate": 9.732957557333575e-06,
+      "loss": 0.2658,
+      "step": 2064
+    },
+    {
+      "epoch": 3.5238907849829353,
+      "grad_norm": 0.34937036857673354,
+      "learning_rate": 9.712517021809798e-06,
+      "loss": 0.2468,
+      "step": 2065
+    },
+    {
+      "epoch": 3.5255972696245736,
+      "grad_norm": 0.3087690690490283,
+      "learning_rate": 9.692091087486495e-06,
+      "loss": 0.2542,
+      "step": 2066
+    },
+    {
+      "epoch": 3.5273037542662116,
+      "grad_norm": 0.36445015588519614,
+      "learning_rate": 9.671679783354557e-06,
+      "loss": 0.2388,
+      "step": 2067
+    },
+    {
+      "epoch": 3.52901023890785,
+      "grad_norm": 0.35192548669396095,
+      "learning_rate": 9.651283138384084e-06,
+      "loss": 0.2722,
+      "step": 2068
+    },
+    {
+      "epoch": 3.530716723549488,
+      "grad_norm": 0.352238143899704,
+      "learning_rate": 9.630901181524406e-06,
+      "loss": 0.2539,
+      "step": 2069
+    },
+    {
+      "epoch": 3.532423208191126,
+      "grad_norm": 0.3254943879594676,
+      "learning_rate": 9.61053394170395e-06,
+      "loss": 0.2523,
+      "step": 2070
+    },
+    {
+      "epoch": 3.5341296928327646,
+      "grad_norm": 0.3471578549897025,
+      "learning_rate": 9.590181447830305e-06,
+      "loss": 0.2196,
+      "step": 2071
+    },
+    {
+      "epoch": 3.535836177474403,
+      "grad_norm": 0.3428321981949603,
+      "learning_rate": 9.56984372879012e-06,
+      "loss": 0.2667,
+      "step": 2072
+    },
+    {
+      "epoch": 3.537542662116041,
+      "grad_norm": 0.3351455769276011,
+      "learning_rate": 9.549520813449053e-06,
+      "loss": 0.2729,
+      "step": 2073
+    },
+    {
+      "epoch": 3.539249146757679,
+      "grad_norm": 0.32223649268587223,
+      "learning_rate": 9.52921273065178e-06,
+      "loss": 0.2213,
+      "step": 2074
+    },
+    {
+      "epoch": 3.5409556313993176,
+      "grad_norm": 0.3420021574309786,
+      "learning_rate": 9.508919509221903e-06,
+      "loss": 0.2326,
+      "step": 2075
+    },
+    {
+      "epoch": 3.5426621160409555,
+      "grad_norm": 0.3453782665927399,
+      "learning_rate": 9.488641177961939e-06,
+      "loss": 0.2556,
+      "step": 2076
+    },
+    {
+      "epoch": 3.544368600682594,
+      "grad_norm": 0.3203712108150496,
+      "learning_rate": 9.46837776565326e-06,
+      "loss": 0.2867,
+      "step": 2077
+    },
+    {
+      "epoch": 3.546075085324232,
+      "grad_norm": 0.3287923565115119,
+      "learning_rate": 9.448129301056083e-06,
+      "loss": 0.2315,
+      "step": 2078
+    },
+    {
+      "epoch": 3.5477815699658706,
+      "grad_norm": 0.3418083531162421,
+      "learning_rate": 9.427895812909406e-06,
+      "loss": 0.2489,
+      "step": 2079
+    },
+    {
+      "epoch": 3.5494880546075085,
+      "grad_norm": 0.31476289542989877,
+      "learning_rate": 9.407677329930953e-06,
+      "loss": 0.2333,
+      "step": 2080
+    },
+    {
+      "epoch": 3.551194539249147,
+      "grad_norm": 0.294170684685756,
+      "learning_rate": 9.387473880817182e-06,
+      "loss": 0.2683,
+      "step": 2081
+    },
+    {
+      "epoch": 3.5529010238907848,
+      "grad_norm": 0.30412989872613483,
+      "learning_rate": 9.367285494243164e-06,
+      "loss": 0.2215,
+      "step": 2082
+    },
+    {
+      "epoch": 3.554607508532423,
+      "grad_norm": 0.32414311515130295,
+      "learning_rate": 9.347112198862645e-06,
+      "loss": 0.2154,
+      "step": 2083
+    },
+    {
+      "epoch": 3.5563139931740615,
+      "grad_norm": 0.32488273820018754,
+      "learning_rate": 9.32695402330791e-06,
+      "loss": 0.2352,
+      "step": 2084
+    },
+    {
+      "epoch": 3.5580204778157,
+      "grad_norm": 0.3312057994129262,
+      "learning_rate": 9.306810996189823e-06,
+      "loss": 0.2377,
+      "step": 2085
+    },
+    {
+      "epoch": 3.5597269624573378,
+      "grad_norm": 0.34094265625638326,
+      "learning_rate": 9.286683146097705e-06,
+      "loss": 0.2474,
+      "step": 2086
+    },
+    {
+      "epoch": 3.561433447098976,
+      "grad_norm": 0.3714211231506183,
+      "learning_rate": 9.266570501599372e-06,
+      "loss": 0.2181,
+      "step": 2087
+    },
+    {
+      "epoch": 3.5631399317406145,
+      "grad_norm": 0.33872908334548024,
+      "learning_rate": 9.246473091241056e-06,
+      "loss": 0.2087,
+      "step": 2088
+    },
+    {
+      "epoch": 3.5648464163822524,
+      "grad_norm": 0.36716903367936693,
+      "learning_rate": 9.226390943547322e-06,
+      "loss": 0.2201,
+      "step": 2089
+    },
+    {
+      "epoch": 3.5665529010238908,
+      "grad_norm": 0.32987230116430716,
+      "learning_rate": 9.206324087021132e-06,
+      "loss": 0.2371,
+      "step": 2090
+    },
+    {
+      "epoch": 3.568259385665529,
+      "grad_norm": 0.3650003583650618,
+      "learning_rate": 9.186272550143702e-06,
+      "loss": 0.229,
+      "step": 2091
+    },
+    {
+      "epoch": 3.5699658703071675,
+      "grad_norm": 0.3323941837267181,
+      "learning_rate": 9.166236361374539e-06,
+      "loss": 0.2271,
+      "step": 2092
+    },
+    {
+      "epoch": 3.5716723549488054,
+      "grad_norm": 0.34853938656176586,
+      "learning_rate": 9.14621554915133e-06,
+      "loss": 0.2366,
+      "step": 2093
+    },
+    {
+      "epoch": 3.573378839590444,
+      "grad_norm": 0.31455666309193503,
+      "learning_rate": 9.126210141889974e-06,
+      "loss": 0.247,
+      "step": 2094
+    },
+    {
+      "epoch": 3.5750853242320817,
+      "grad_norm": 0.3179121796258658,
+      "learning_rate": 9.106220167984474e-06,
+      "loss": 0.2248,
+      "step": 2095
+    },
+    {
+      "epoch": 3.57679180887372,
+      "grad_norm": 0.3216632641238737,
+      "learning_rate": 9.08624565580694e-06,
+      "loss": 0.2492,
+      "step": 2096
+    },
+    {
+      "epoch": 3.5784982935153584,
+      "grad_norm": 0.32704473572032683,
+      "learning_rate": 9.066286633707552e-06,
+      "loss": 0.2413,
+      "step": 2097
+    },
+    {
+      "epoch": 3.580204778156997,
+      "grad_norm": 0.4293241626059611,
+      "learning_rate": 9.04634313001448e-06,
+      "loss": 0.2424,
+      "step": 2098
+    },
+    {
+      "epoch": 3.5819112627986347,
+      "grad_norm": 0.33355020607853786,
+      "learning_rate": 9.026415173033886e-06,
+      "loss": 0.2911,
+      "step": 2099
+    },
+    {
+      "epoch": 3.583617747440273,
+      "grad_norm": 0.32899497678876044,
+      "learning_rate": 9.006502791049861e-06,
+      "loss": 0.2401,
+      "step": 2100
+    },
+    {
+      "epoch": 3.5853242320819114,
+      "grad_norm": 0.33560992790127164,
+      "learning_rate": 8.986606012324376e-06,
+      "loss": 0.2224,
+      "step": 2101
+    },
+    {
+      "epoch": 3.5870307167235493,
+      "grad_norm": 0.3149671347227518,
+      "learning_rate": 8.96672486509729e-06,
+      "loss": 0.2616,
+      "step": 2102
+    },
+    {
+      "epoch": 3.5887372013651877,
+      "grad_norm": 0.3229912121146895,
+      "learning_rate": 8.946859377586236e-06,
+      "loss": 0.2551,
+      "step": 2103
+    },
+    {
+      "epoch": 3.590443686006826,
+      "grad_norm": 0.33545725996995696,
+      "learning_rate": 8.927009577986654e-06,
+      "loss": 0.2136,
+      "step": 2104
+    },
+    {
+      "epoch": 3.5921501706484644,
+      "grad_norm": 0.31534595869214316,
+      "learning_rate": 8.907175494471693e-06,
+      "loss": 0.2394,
+      "step": 2105
+    },
+    {
+      "epoch": 3.5938566552901023,
+      "grad_norm": 0.3527301933284398,
+      "learning_rate": 8.887357155192218e-06,
+      "loss": 0.2323,
+      "step": 2106
+    },
+    {
+      "epoch": 3.5955631399317407,
+      "grad_norm": 0.36778167133207873,
+      "learning_rate": 8.867554588276732e-06,
+      "loss": 0.2074,
+      "step": 2107
+    },
+    {
+      "epoch": 3.5972696245733786,
+      "grad_norm": 0.3269001726303635,
+      "learning_rate": 8.847767821831347e-06,
+      "loss": 0.2668,
+      "step": 2108
+    },
+    {
+      "epoch": 3.598976109215017,
+      "grad_norm": 0.3380568718247642,
+      "learning_rate": 8.827996883939779e-06,
+      "loss": 0.2431,
+      "step": 2109
+    },
+    {
+      "epoch": 3.6006825938566553,
+      "grad_norm": 0.3529140072056332,
+      "learning_rate": 8.808241802663236e-06,
+      "loss": 0.2095,
+      "step": 2110
+    },
+    {
+      "epoch": 3.6023890784982937,
+      "grad_norm": 0.31665187204189843,
+      "learning_rate": 8.78850260604046e-06,
+      "loss": 0.2277,
+      "step": 2111
+    },
+    {
+      "epoch": 3.6040955631399316,
+      "grad_norm": 0.3412747364772627,
+      "learning_rate": 8.768779322087626e-06,
+      "loss": 0.2406,
+      "step": 2112
+    },
+    {
+      "epoch": 3.60580204778157,
+      "grad_norm": 0.33124871502489983,
+      "learning_rate": 8.749071978798319e-06,
+      "loss": 0.2282,
+      "step": 2113
+    },
+    {
+      "epoch": 3.6075085324232083,
+      "grad_norm": 0.35546659291499355,
+      "learning_rate": 8.72938060414352e-06,
+      "loss": 0.2705,
+      "step": 2114
+    },
+    {
+      "epoch": 3.6092150170648463,
+      "grad_norm": 0.34706721829273074,
+      "learning_rate": 8.709705226071526e-06,
+      "loss": 0.2359,
+      "step": 2115
+    },
+    {
+      "epoch": 3.6109215017064846,
+      "grad_norm": 0.35923389833000235,
+      "learning_rate": 8.690045872507944e-06,
+      "loss": 0.2538,
+      "step": 2116
+    },
+    {
+      "epoch": 3.612627986348123,
+      "grad_norm": 0.30490467326500825,
+      "learning_rate": 8.67040257135562e-06,
+      "loss": 0.2699,
+      "step": 2117
+    },
+    {
+      "epoch": 3.6143344709897613,
+      "grad_norm": 0.3216326103324856,
+      "learning_rate": 8.650775350494643e-06,
+      "loss": 0.2473,
+      "step": 2118
+    },
+    {
+      "epoch": 3.6160409556313993,
+      "grad_norm": 0.337849661154815,
+      "learning_rate": 8.631164237782253e-06,
+      "loss": 0.2286,
+      "step": 2119
+    },
+    {
+      "epoch": 3.6177474402730376,
+      "grad_norm": 0.3357868273033272,
+      "learning_rate": 8.611569261052833e-06,
+      "loss": 0.2234,
+      "step": 2120
+    },
+    {
+      "epoch": 3.6194539249146755,
+      "grad_norm": 0.31790624894855846,
+      "learning_rate": 8.59199044811788e-06,
+      "loss": 0.249,
+      "step": 2121
+    },
+    {
+      "epoch": 3.621160409556314,
+      "grad_norm": 0.3253336788323475,
+      "learning_rate": 8.572427826765926e-06,
+      "loss": 0.2522,
+      "step": 2122
+    },
+    {
+      "epoch": 3.6228668941979523,
+      "grad_norm": 0.3437765066648847,
+      "learning_rate": 8.55288142476255e-06,
+      "loss": 0.2145,
+      "step": 2123
+    },
+    {
+      "epoch": 3.6245733788395906,
+      "grad_norm": 0.3694147767833111,
+      "learning_rate": 8.533351269850273e-06,
+      "loss": 0.2315,
+      "step": 2124
+    },
+    {
+      "epoch": 3.6262798634812285,
+      "grad_norm": 0.33413722060276174,
+      "learning_rate": 8.5138373897486e-06,
+      "loss": 0.259,
+      "step": 2125
+    },
+    {
+      "epoch": 3.627986348122867,
+      "grad_norm": 0.32437481899148385,
+      "learning_rate": 8.494339812153905e-06,
+      "loss": 0.2442,
+      "step": 2126
+    },
+    {
+      "epoch": 3.6296928327645053,
+      "grad_norm": 0.3611094623293024,
+      "learning_rate": 8.474858564739423e-06,
+      "loss": 0.2128,
+      "step": 2127
+    },
+    {
+      "epoch": 3.631399317406143,
+      "grad_norm": 0.3138052807257615,
+      "learning_rate": 8.455393675155239e-06,
+      "loss": 0.2337,
+      "step": 2128
+    },
+    {
+      "epoch": 3.6331058020477816,
+      "grad_norm": 0.33835867675852266,
+      "learning_rate": 8.43594517102819e-06,
+      "loss": 0.216,
+      "step": 2129
+    },
+    {
+      "epoch": 3.63481228668942,
+      "grad_norm": 0.34307857648303824,
+      "learning_rate": 8.41651307996188e-06,
+      "loss": 0.2323,
+      "step": 2130
+    },
+    {
+      "epoch": 3.6365187713310583,
+      "grad_norm": 0.31969100992776645,
+      "learning_rate": 8.39709742953661e-06,
+      "loss": 0.2518,
+      "step": 2131
+    },
+    {
+      "epoch": 3.638225255972696,
+      "grad_norm": 0.34710848014672774,
+      "learning_rate": 8.377698247309327e-06,
+      "loss": 0.2374,
+      "step": 2132
+    },
+    {
+      "epoch": 3.6399317406143346,
+      "grad_norm": 0.389195318887567,
+      "learning_rate": 8.358315560813642e-06,
+      "loss": 0.2039,
+      "step": 2133
+    },
+    {
+      "epoch": 3.6416382252559725,
+      "grad_norm": 0.35978442598178156,
+      "learning_rate": 8.33894939755972e-06,
+      "loss": 0.2332,
+      "step": 2134
+    },
+    {
+      "epoch": 3.643344709897611,
+      "grad_norm": 0.31619041833521977,
+      "learning_rate": 8.319599785034296e-06,
+      "loss": 0.2508,
+      "step": 2135
+    },
+    {
+      "epoch": 3.645051194539249,
+      "grad_norm": 0.3851432754060637,
+      "learning_rate": 8.300266750700598e-06,
+      "loss": 0.3253,
+      "step": 2136
+    },
+    {
+      "epoch": 3.6467576791808876,
+      "grad_norm": 0.3017952759413884,
+      "learning_rate": 8.28095032199835e-06,
+      "loss": 0.2465,
+      "step": 2137
+    },
+    {
+      "epoch": 3.6484641638225255,
+      "grad_norm": 0.33636218269981427,
+      "learning_rate": 8.261650526343665e-06,
+      "loss": 0.212,
+      "step": 2138
+    },
+    {
+      "epoch": 3.650170648464164,
+      "grad_norm": 0.4071704606431761,
+      "learning_rate": 8.242367391129082e-06,
+      "loss": 0.2115,
+      "step": 2139
+    },
+    {
+      "epoch": 3.651877133105802,
+      "grad_norm": 0.3892103197274746,
+      "learning_rate": 8.223100943723494e-06,
+      "loss": 0.2438,
+      "step": 2140
+    },
+    {
+      "epoch": 3.65358361774744,
+      "grad_norm": 0.34583333849310055,
+      "learning_rate": 8.203851211472088e-06,
+      "loss": 0.2224,
+      "step": 2141
+    },
+    {
+      "epoch": 3.6552901023890785,
+      "grad_norm": 0.35043746125060526,
+      "learning_rate": 8.184618221696346e-06,
+      "loss": 0.2478,
+      "step": 2142
+    },
+    {
+      "epoch": 3.656996587030717,
+      "grad_norm": 0.352112680488658,
+      "learning_rate": 8.165402001693976e-06,
+      "loss": 0.2115,
+      "step": 2143
+    },
+    {
+      "epoch": 3.658703071672355,
+      "grad_norm": 0.31977179892520446,
+      "learning_rate": 8.146202578738887e-06,
+      "loss": 0.2366,
+      "step": 2144
+    },
+    {
+      "epoch": 3.660409556313993,
+      "grad_norm": 0.38326623640423296,
+      "learning_rate": 8.127019980081141e-06,
+      "loss": 0.2525,
+      "step": 2145
+    },
+    {
+      "epoch": 3.6621160409556315,
+      "grad_norm": 0.3179398491073991,
+      "learning_rate": 8.107854232946937e-06,
+      "loss": 0.2643,
+      "step": 2146
+    },
+    {
+      "epoch": 3.6638225255972694,
+      "grad_norm": 0.3351679457817645,
+      "learning_rate": 8.088705364538552e-06,
+      "loss": 0.2265,
+      "step": 2147
+    },
+    {
+      "epoch": 3.6655290102389078,
+      "grad_norm": 0.33559257148440325,
+      "learning_rate": 8.06957340203429e-06,
+      "loss": 0.2277,
+      "step": 2148
+    },
+    {
+      "epoch": 3.667235494880546,
+      "grad_norm": 0.315222850177876,
+      "learning_rate": 8.050458372588493e-06,
+      "loss": 0.2481,
+      "step": 2149
+    },
+    {
+      "epoch": 3.6689419795221845,
+      "grad_norm": 0.316354698233717,
+      "learning_rate": 8.031360303331419e-06,
+      "loss": 0.2461,
+      "step": 2150
+    },
+    {
+      "epoch": 3.6706484641638224,
+      "grad_norm": 0.3299376154281702,
+      "learning_rate": 8.012279221369308e-06,
+      "loss": 0.2764,
+      "step": 2151
+    },
+    {
+      "epoch": 3.6723549488054608,
+      "grad_norm": 0.3358898591519496,
+      "learning_rate": 7.993215153784254e-06,
+      "loss": 0.2412,
+      "step": 2152
+    },
+    {
+      "epoch": 3.674061433447099,
+      "grad_norm": 0.34779848701641763,
+      "learning_rate": 7.974168127634214e-06,
+      "loss": 0.2123,
+      "step": 2153
+    },
+    {
+      "epoch": 3.675767918088737,
+      "grad_norm": 0.34578906731774767,
+      "learning_rate": 7.955138169952972e-06,
+      "loss": 0.2425,
+      "step": 2154
+    },
+    {
+      "epoch": 3.6774744027303754,
+      "grad_norm": 0.3179010754096789,
+      "learning_rate": 7.936125307750062e-06,
+      "loss": 0.2584,
+      "step": 2155
+    },
+    {
+      "epoch": 3.6791808873720138,
+      "grad_norm": 0.3284611461610992,
+      "learning_rate": 7.917129568010764e-06,
+      "loss": 0.3071,
+      "step": 2156
+    },
+    {
+      "epoch": 3.680887372013652,
+      "grad_norm": 0.3492684100554768,
+      "learning_rate": 7.898150977696051e-06,
+      "loss": 0.2251,
+      "step": 2157
+    },
+    {
+      "epoch": 3.68259385665529,
+      "grad_norm": 0.33288502140176757,
+      "learning_rate": 7.879189563742574e-06,
+      "loss": 0.2009,
+      "step": 2158
+    },
+    {
+      "epoch": 3.6843003412969284,
+      "grad_norm": 0.34974732061580127,
+      "learning_rate": 7.860245353062575e-06,
+      "loss": 0.2255,
+      "step": 2159
+    },
+    {
+      "epoch": 3.6860068259385663,
+      "grad_norm": 0.3324104074612852,
+      "learning_rate": 7.841318372543906e-06,
+      "loss": 0.2708,
+      "step": 2160
+    },
+    {
+      "epoch": 3.6877133105802047,
+      "grad_norm": 0.3081862433519352,
+      "learning_rate": 7.822408649049963e-06,
+      "loss": 0.2203,
+      "step": 2161
+    },
+    {
+      "epoch": 3.689419795221843,
+      "grad_norm": 0.31222019978876575,
+      "learning_rate": 7.803516209419631e-06,
+      "loss": 0.2464,
+      "step": 2162
+    },
+    {
+      "epoch": 3.6911262798634814,
+      "grad_norm": 0.3073306272647333,
+      "learning_rate": 7.784641080467272e-06,
+      "loss": 0.2685,
+      "step": 2163
+    },
+    {
+      "epoch": 3.6928327645051193,
+      "grad_norm": 0.35335232359685936,
+      "learning_rate": 7.76578328898267e-06,
+      "loss": 0.2664,
+      "step": 2164
+    },
+    {
+      "epoch": 3.6945392491467577,
+      "grad_norm": 0.3272278578330255,
+      "learning_rate": 7.74694286173103e-06,
+      "loss": 0.2718,
+      "step": 2165
+    },
+    {
+      "epoch": 3.696245733788396,
+      "grad_norm": 0.3329431402767,
+      "learning_rate": 7.728119825452875e-06,
+      "loss": 0.2178,
+      "step": 2166
+    },
+    {
+      "epoch": 3.697952218430034,
+      "grad_norm": 0.3365044939427897,
+      "learning_rate": 7.709314206864071e-06,
+      "loss": 0.2396,
+      "step": 2167
+    },
+    {
+      "epoch": 3.6996587030716723,
+      "grad_norm": 0.3415383532921338,
+      "learning_rate": 7.690526032655768e-06,
+      "loss": 0.266,
+      "step": 2168
+    },
+    {
+      "epoch": 3.7013651877133107,
+      "grad_norm": 0.3204323451928413,
+      "learning_rate": 7.671755329494312e-06,
+      "loss": 0.1995,
+      "step": 2169
+    },
+    {
+      "epoch": 3.703071672354949,
+      "grad_norm": 0.32460920939744076,
+      "learning_rate": 7.653002124021307e-06,
+      "loss": 0.2149,
+      "step": 2170
+    },
+    {
+      "epoch": 3.704778156996587,
+      "grad_norm": 0.33696350421488724,
+      "learning_rate": 7.634266442853485e-06,
+      "loss": 0.2153,
+      "step": 2171
+    },
+    {
+      "epoch": 3.7064846416382253,
+      "grad_norm": 0.34418317167429263,
+      "learning_rate": 7.615548312582728e-06,
+      "loss": 0.2084,
+      "step": 2172
+    },
+    {
+      "epoch": 3.7081911262798632,
+      "grad_norm": 0.34758003716187136,
+      "learning_rate": 7.596847759775987e-06,
+      "loss": 0.2628,
+      "step": 2173
+    },
+    {
+      "epoch": 3.7098976109215016,
+      "grad_norm": 0.32713967444234604,
+      "learning_rate": 7.5781648109752904e-06,
+      "loss": 0.2353,
+      "step": 2174
+    },
+    {
+      "epoch": 3.71160409556314,
+      "grad_norm": 0.4392955751004257,
+      "learning_rate": 7.559499492697662e-06,
+      "loss": 0.231,
+      "step": 2175
+    },
+    {
+      "epoch": 3.7133105802047783,
+      "grad_norm": 0.37151134656871065,
+      "learning_rate": 7.540851831435097e-06,
+      "loss": 0.241,
+      "step": 2176
+    },
+    {
+      "epoch": 3.7150170648464163,
+      "grad_norm": 0.3108353138523825,
+      "learning_rate": 7.522221853654554e-06,
+      "loss": 0.2597,
+      "step": 2177
+    },
+    {
+      "epoch": 3.7167235494880546,
+      "grad_norm": 0.34212178021994255,
+      "learning_rate": 7.503609585797866e-06,
+      "loss": 0.2795,
+      "step": 2178
+    },
+    {
+      "epoch": 3.718430034129693,
+      "grad_norm": 0.33501937974174234,
+      "learning_rate": 7.485015054281757e-06,
+      "loss": 0.2221,
+      "step": 2179
+    },
+    {
+      "epoch": 3.720136518771331,
+      "grad_norm": 0.33226356966856707,
+      "learning_rate": 7.4664382854977564e-06,
+      "loss": 0.2105,
+      "step": 2180
+    },
+    {
+      "epoch": 3.7218430034129693,
+      "grad_norm": 0.35084184784766803,
+      "learning_rate": 7.447879305812184e-06,
+      "loss": 0.2219,
+      "step": 2181
+    },
+    {
+      "epoch": 3.7235494880546076,
+      "grad_norm": 0.370333932074055,
+      "learning_rate": 7.429338141566129e-06,
+      "loss": 0.2341,
+      "step": 2182
+    },
+    {
+      "epoch": 3.725255972696246,
+      "grad_norm": 0.3485737212375025,
+      "learning_rate": 7.410814819075369e-06,
+      "loss": 0.2254,
+      "step": 2183
+    },
+    {
+      "epoch": 3.726962457337884,
+      "grad_norm": 0.35474527253456567,
+      "learning_rate": 7.392309364630388e-06,
+      "loss": 0.2371,
+      "step": 2184
+    },
+    {
+      "epoch": 3.7286689419795223,
+      "grad_norm": 0.3239195091455572,
+      "learning_rate": 7.373821804496277e-06,
+      "loss": 0.2137,
+      "step": 2185
+    },
+    {
+      "epoch": 3.73037542662116,
+      "grad_norm": 0.37912144087539207,
+      "learning_rate": 7.355352164912763e-06,
+      "loss": 0.2174,
+      "step": 2186
+    },
+    {
+      "epoch": 3.7320819112627985,
+      "grad_norm": 0.34093396600499953,
+      "learning_rate": 7.336900472094113e-06,
+      "loss": 0.2144,
+      "step": 2187
+    },
+    {
+      "epoch": 3.733788395904437,
+      "grad_norm": 0.36912720858317266,
+      "learning_rate": 7.318466752229123e-06,
+      "loss": 0.2506,
+      "step": 2188
+    },
+    {
+      "epoch": 3.7354948805460753,
+      "grad_norm": 0.3400048147976415,
+      "learning_rate": 7.300051031481101e-06,
+      "loss": 0.2558,
+      "step": 2189
+    },
+    {
+      "epoch": 3.737201365187713,
+      "grad_norm": 0.3616647897221543,
+      "learning_rate": 7.281653335987782e-06,
+      "loss": 0.2323,
+      "step": 2190
+    },
+    {
+      "epoch": 3.7389078498293515,
+      "grad_norm": 0.33049342906173645,
+      "learning_rate": 7.263273691861341e-06,
+      "loss": 0.2316,
+      "step": 2191
+    },
+    {
+      "epoch": 3.74061433447099,
+      "grad_norm": 0.35189500703155224,
+      "learning_rate": 7.244912125188308e-06,
+      "loss": 0.2412,
+      "step": 2192
+    },
+    {
+      "epoch": 3.742320819112628,
+      "grad_norm": 0.3864810014202174,
+      "learning_rate": 7.226568662029583e-06,
+      "loss": 0.226,
+      "step": 2193
+    },
+    {
+      "epoch": 3.744027303754266,
+      "grad_norm": 0.3372212223722344,
+      "learning_rate": 7.208243328420348e-06,
+      "loss": 0.2394,
+      "step": 2194
+    },
+    {
+      "epoch": 3.7457337883959045,
+      "grad_norm": 0.33764911560400823,
+      "learning_rate": 7.189936150370061e-06,
+      "loss": 0.2161,
+      "step": 2195
+    },
+    {
+      "epoch": 3.747440273037543,
+      "grad_norm": 0.29995034049716035,
+      "learning_rate": 7.17164715386242e-06,
+      "loss": 0.2302,
+      "step": 2196
+    },
+    {
+      "epoch": 3.749146757679181,
+      "grad_norm": 0.30065277674622454,
+      "learning_rate": 7.153376364855298e-06,
+      "loss": 0.2355,
+      "step": 2197
+    },
+    {
+      "epoch": 3.750853242320819,
+      "grad_norm": 0.3233014807766729,
+      "learning_rate": 7.135123809280755e-06,
+      "loss": 0.2206,
+      "step": 2198
+    },
+    {
+      "epoch": 3.752559726962457,
+      "grad_norm": 0.3085572665587782,
+      "learning_rate": 7.116889513044947e-06,
+      "loss": 0.2431,
+      "step": 2199
+    },
+    {
+      "epoch": 3.7542662116040955,
+      "grad_norm": 0.3099328328660361,
+      "learning_rate": 7.098673502028115e-06,
+      "loss": 0.2432,
+      "step": 2200
+    },
+    {
+      "epoch": 3.755972696245734,
+      "grad_norm": 0.3453231654873741,
+      "learning_rate": 7.08047580208457e-06,
+      "loss": 0.2463,
+      "step": 2201
+    },
+    {
+      "epoch": 3.757679180887372,
+      "grad_norm": 0.34884193800760066,
+      "learning_rate": 7.062296439042602e-06,
+      "loss": 0.2295,
+      "step": 2202
+    },
+    {
+      "epoch": 3.75938566552901,
+      "grad_norm": 0.3118987064147418,
+      "learning_rate": 7.044135438704509e-06,
+      "loss": 0.2682,
+      "step": 2203
+    },
+    {
+      "epoch": 3.7610921501706485,
+      "grad_norm": 0.35472956867938155,
+      "learning_rate": 7.025992826846493e-06,
+      "loss": 0.2351,
+      "step": 2204
+    },
+    {
+      "epoch": 3.762798634812287,
+      "grad_norm": 0.30905315147129464,
+      "learning_rate": 7.007868629218686e-06,
+      "loss": 0.2411,
+      "step": 2205
+    },
+    {
+      "epoch": 3.7645051194539247,
+      "grad_norm": 0.336875372461944,
+      "learning_rate": 6.989762871545069e-06,
+      "loss": 0.2832,
+      "step": 2206
+    },
+    {
+      "epoch": 3.766211604095563,
+      "grad_norm": 0.5530281983642463,
+      "learning_rate": 6.971675579523443e-06,
+      "loss": 0.3134,
+      "step": 2207
+    },
+    {
+      "epoch": 3.7679180887372015,
+      "grad_norm": 0.3346356161529318,
+      "learning_rate": 6.953606778825426e-06,
+      "loss": 0.2121,
+      "step": 2208
+    },
+    {
+      "epoch": 3.76962457337884,
+      "grad_norm": 0.3457212843139903,
+      "learning_rate": 6.9355564950963606e-06,
+      "loss": 0.2487,
+      "step": 2209
+    },
+    {
+      "epoch": 3.7713310580204777,
+      "grad_norm": 0.30021566156978446,
+      "learning_rate": 6.917524753955338e-06,
+      "loss": 0.2632,
+      "step": 2210
+    },
+    {
+      "epoch": 3.773037542662116,
+      "grad_norm": 0.37622942535137993,
+      "learning_rate": 6.899511580995111e-06,
+      "loss": 0.2448,
+      "step": 2211
+    },
+    {
+      "epoch": 3.774744027303754,
+      "grad_norm": 0.33896457300845667,
+      "learning_rate": 6.881517001782074e-06,
+      "loss": 0.2577,
+      "step": 2212
+    },
+    {
+      "epoch": 3.7764505119453924,
+      "grad_norm": 0.3362204559926037,
+      "learning_rate": 6.8635410418562585e-06,
+      "loss": 0.2014,
+      "step": 2213
+    },
+    {
+      "epoch": 3.7781569965870307,
+      "grad_norm": 0.36751468114360947,
+      "learning_rate": 6.845583726731236e-06,
+      "loss": 0.2032,
+      "step": 2214
+    },
+    {
+      "epoch": 3.779863481228669,
+      "grad_norm": 0.3390314224721075,
+      "learning_rate": 6.827645081894141e-06,
+      "loss": 0.2356,
+      "step": 2215
+    },
+    {
+      "epoch": 3.781569965870307,
+      "grad_norm": 0.3571072257977754,
+      "learning_rate": 6.809725132805591e-06,
+      "loss": 0.2418,
+      "step": 2216
+    },
+    {
+      "epoch": 3.7832764505119454,
+      "grad_norm": 0.350285070405977,
+      "learning_rate": 6.791823904899695e-06,
+      "loss": 0.2256,
+      "step": 2217
+    },
+    {
+      "epoch": 3.7849829351535837,
+      "grad_norm": 0.32614851056606337,
+      "learning_rate": 6.773941423583945e-06,
+      "loss": 0.2377,
+      "step": 2218
+    },
+    {
+      "epoch": 3.7866894197952217,
+      "grad_norm": 0.3467495156868907,
+      "learning_rate": 6.756077714239264e-06,
+      "loss": 0.2482,
+      "step": 2219
+    },
+    {
+      "epoch": 3.78839590443686,
+      "grad_norm": 0.3416010808769771,
+      "learning_rate": 6.7382328022199265e-06,
+      "loss": 0.2546,
+      "step": 2220
+    },
+    {
+      "epoch": 3.7901023890784984,
+      "grad_norm": 0.3177041264236061,
+      "learning_rate": 6.720406712853511e-06,
+      "loss": 0.2522,
+      "step": 2221
+    },
+    {
+      "epoch": 3.7918088737201368,
+      "grad_norm": 0.34743021750533976,
+      "learning_rate": 6.7025994714409004e-06,
+      "loss": 0.2487,
+      "step": 2222
+    },
+    {
+      "epoch": 3.7935153583617747,
+      "grad_norm": 0.3337937791850411,
+      "learning_rate": 6.684811103256215e-06,
+      "loss": 0.2148,
+      "step": 2223
+    },
+    {
+      "epoch": 3.795221843003413,
+      "grad_norm": 0.32356246884646855,
+      "learning_rate": 6.667041633546785e-06,
+      "loss": 0.2401,
+      "step": 2224
+    },
+    {
+      "epoch": 3.796928327645051,
+      "grad_norm": 0.3113785634712896,
+      "learning_rate": 6.649291087533119e-06,
+      "loss": 0.2329,
+      "step": 2225
+    },
+    {
+      "epoch": 3.7986348122866893,
+      "grad_norm": 0.3424029121429665,
+      "learning_rate": 6.631559490408874e-06,
+      "loss": 0.2345,
+      "step": 2226
+    },
+    {
+      "epoch": 3.8003412969283277,
+      "grad_norm": 0.299235607502441,
+      "learning_rate": 6.613846867340821e-06,
+      "loss": 0.2521,
+      "step": 2227
+    },
+    {
+      "epoch": 3.802047781569966,
+      "grad_norm": 0.3321546834321874,
+      "learning_rate": 6.5961532434687704e-06,
+      "loss": 0.2202,
+      "step": 2228
+    },
+    {
+      "epoch": 3.803754266211604,
+      "grad_norm": 0.3061966973035992,
+      "learning_rate": 6.578478643905601e-06,
+      "loss": 0.2475,
+      "step": 2229
+    },
+    {
+      "epoch": 3.8054607508532423,
+      "grad_norm": 0.33110608821165544,
+      "learning_rate": 6.560823093737165e-06,
+      "loss": 0.248,
+      "step": 2230
+    },
+    {
+      "epoch": 3.8071672354948807,
+      "grad_norm": 0.3080709292000393,
+      "learning_rate": 6.54318661802229e-06,
+      "loss": 0.2633,
+      "step": 2231
+    },
+    {
+      "epoch": 3.8088737201365186,
+      "grad_norm": 0.35665467275094986,
+      "learning_rate": 6.52556924179272e-06,
+      "loss": 0.2292,
+      "step": 2232
+    },
+    {
+      "epoch": 3.810580204778157,
+      "grad_norm": 0.33053205317040113,
+      "learning_rate": 6.507970990053103e-06,
+      "loss": 0.2418,
+      "step": 2233
+    },
+    {
+      "epoch": 3.8122866894197953,
+      "grad_norm": 0.30625806374467224,
+      "learning_rate": 6.490391887780947e-06,
+      "loss": 0.2344,
+      "step": 2234
+    },
+    {
+      "epoch": 3.8139931740614337,
+      "grad_norm": 0.30583391518437175,
+      "learning_rate": 6.472831959926558e-06,
+      "loss": 0.2452,
+      "step": 2235
+    },
+    {
+      "epoch": 3.8156996587030716,
+      "grad_norm": 0.3135255557559026,
+      "learning_rate": 6.4552912314130614e-06,
+      "loss": 0.2172,
+      "step": 2236
+    },
+    {
+      "epoch": 3.81740614334471,
+      "grad_norm": 0.3307861502183436,
+      "learning_rate": 6.43776972713629e-06,
+      "loss": 0.2244,
+      "step": 2237
+    },
+    {
+      "epoch": 3.819112627986348,
+      "grad_norm": 0.3206168661451337,
+      "learning_rate": 6.420267471964829e-06,
+      "loss": 0.2305,
+      "step": 2238
+    },
+    {
+      "epoch": 3.8208191126279862,
+      "grad_norm": 0.29036335019893844,
+      "learning_rate": 6.40278449073992e-06,
+      "loss": 0.2733,
+      "step": 2239
+    },
+    {
+      "epoch": 3.8225255972696246,
+      "grad_norm": 0.3094904433899518,
+      "learning_rate": 6.385320808275459e-06,
+      "loss": 0.229,
+      "step": 2240
+    },
+    {
+      "epoch": 3.824232081911263,
+      "grad_norm": 0.3121188827480497,
+      "learning_rate": 6.36787644935796e-06,
+      "loss": 0.2651,
+      "step": 2241
+    },
+    {
+      "epoch": 3.825938566552901,
+      "grad_norm": 0.3329944582433921,
+      "learning_rate": 6.3504514387464925e-06,
+      "loss": 0.2161,
+      "step": 2242
+    },
+    {
+      "epoch": 3.8276450511945392,
+      "grad_norm": 0.38151076966358494,
+      "learning_rate": 6.333045801172668e-06,
+      "loss": 0.2189,
+      "step": 2243
+    },
+    {
+      "epoch": 3.8293515358361776,
+      "grad_norm": 0.34522340292194326,
+      "learning_rate": 6.315659561340606e-06,
+      "loss": 0.2924,
+      "step": 2244
+    },
+    {
+      "epoch": 3.8310580204778155,
+      "grad_norm": 0.34706992737004416,
+      "learning_rate": 6.298292743926901e-06,
+      "loss": 0.2191,
+      "step": 2245
+    },
+    {
+      "epoch": 3.832764505119454,
+      "grad_norm": 0.29693843677243775,
+      "learning_rate": 6.280945373580563e-06,
+      "loss": 0.244,
+      "step": 2246
+    },
+    {
+      "epoch": 3.8344709897610922,
+      "grad_norm": 0.35326438823935363,
+      "learning_rate": 6.263617474923019e-06,
+      "loss": 0.2571,
+      "step": 2247
+    },
+    {
+      "epoch": 3.8361774744027306,
+      "grad_norm": 0.33319605929871543,
+      "learning_rate": 6.246309072548062e-06,
+      "loss": 0.2292,
+      "step": 2248
+    },
+    {
+      "epoch": 3.8378839590443685,
+      "grad_norm": 0.37060000034961466,
+      "learning_rate": 6.22902019102178e-06,
+      "loss": 0.2536,
+      "step": 2249
+    },
+    {
+      "epoch": 3.839590443686007,
+      "grad_norm": 0.35893056819086877,
+      "learning_rate": 6.211750854882594e-06,
+      "loss": 0.327,
+      "step": 2250
+    },
+    {
+      "epoch": 3.841296928327645,
+      "grad_norm": 0.354582764238136,
+      "learning_rate": 6.19450108864116e-06,
+      "loss": 0.22,
+      "step": 2251
+    },
+    {
+      "epoch": 3.843003412969283,
+      "grad_norm": 0.3232038907232125,
+      "learning_rate": 6.177270916780378e-06,
+      "loss": 0.2261,
+      "step": 2252
+    },
+    {
+      "epoch": 3.8447098976109215,
+      "grad_norm": 0.34937331755549195,
+      "learning_rate": 6.160060363755311e-06,
+      "loss": 0.2511,
+      "step": 2253
+    },
+    {
+      "epoch": 3.84641638225256,
+      "grad_norm": 0.3288608821927333,
+      "learning_rate": 6.142869453993203e-06,
+      "loss": 0.229,
+      "step": 2254
+    },
+    {
+      "epoch": 3.848122866894198,
+      "grad_norm": 0.3414056843048828,
+      "learning_rate": 6.125698211893403e-06,
+      "loss": 0.243,
+      "step": 2255
+    },
+    {
+      "epoch": 3.849829351535836,
+      "grad_norm": 0.3665089696862573,
+      "learning_rate": 6.108546661827339e-06,
+      "loss": 0.2791,
+      "step": 2256
+    },
+    {
+      "epoch": 3.8515358361774745,
+      "grad_norm": 0.33870992023890073,
+      "learning_rate": 6.0914148281385134e-06,
+      "loss": 0.2369,
+      "step": 2257
+    },
+    {
+      "epoch": 3.8532423208191124,
+      "grad_norm": 0.37553628654683074,
+      "learning_rate": 6.074302735142419e-06,
+      "loss": 0.2446,
+      "step": 2258
+    },
+    {
+      "epoch": 3.854948805460751,
+      "grad_norm": 0.350783647429083,
+      "learning_rate": 6.057210407126552e-06,
+      "loss": 0.2203,
+      "step": 2259
+    },
+    {
+      "epoch": 3.856655290102389,
+      "grad_norm": 0.3459254788092183,
+      "learning_rate": 6.040137868350342e-06,
+      "loss": 0.2236,
+      "step": 2260
+    },
+    {
+      "epoch": 3.8583617747440275,
+      "grad_norm": 0.3601623602081703,
+      "learning_rate": 6.0230851430451265e-06,
+      "loss": 0.2306,
+      "step": 2261
+    },
+    {
+      "epoch": 3.8600682593856654,
+      "grad_norm": 0.33960158131789747,
+      "learning_rate": 6.006052255414145e-06,
+      "loss": 0.2158,
+      "step": 2262
+    },
+    {
+      "epoch": 3.861774744027304,
+      "grad_norm": 0.305629609959642,
+      "learning_rate": 5.989039229632454e-06,
+      "loss": 0.2609,
+      "step": 2263
+    },
+    {
+      "epoch": 3.8634812286689417,
+      "grad_norm": 0.332401760716566,
+      "learning_rate": 5.972046089846941e-06,
+      "loss": 0.2387,
+      "step": 2264
+    },
+    {
+      "epoch": 3.86518771331058,
+      "grad_norm": 0.3265704471848609,
+      "learning_rate": 5.95507286017625e-06,
+      "loss": 0.2874,
+      "step": 2265
+    },
+    {
+      "epoch": 3.8668941979522184,
+      "grad_norm": 0.3080184933089667,
+      "learning_rate": 5.938119564710787e-06,
+      "loss": 0.2034,
+      "step": 2266
+    },
+    {
+      "epoch": 3.868600682593857,
+      "grad_norm": 0.3289585999305886,
+      "learning_rate": 5.92118622751265e-06,
+      "loss": 0.2406,
+      "step": 2267
+    },
+    {
+      "epoch": 3.8703071672354947,
+      "grad_norm": 0.3088107402469846,
+      "learning_rate": 5.904272872615606e-06,
+      "loss": 0.2669,
+      "step": 2268
+    },
+    {
+      "epoch": 3.872013651877133,
+      "grad_norm": 0.33557172655003475,
+      "learning_rate": 5.887379524025083e-06,
+      "loss": 0.2157,
+      "step": 2269
+    },
+    {
+      "epoch": 3.8737201365187715,
+      "grad_norm": 0.30262976875520914,
+      "learning_rate": 5.870506205718085e-06,
+      "loss": 0.236,
+      "step": 2270
+    },
+    {
+      "epoch": 3.8754266211604094,
+      "grad_norm": 0.343779664813241,
+      "learning_rate": 5.853652941643213e-06,
+      "loss": 0.2378,
+      "step": 2271
+    },
+    {
+      "epoch": 3.8771331058020477,
+      "grad_norm": 0.36997023186316547,
+      "learning_rate": 5.836819755720584e-06,
+      "loss": 0.2414,
+      "step": 2272
+    },
+    {
+      "epoch": 3.878839590443686,
+      "grad_norm": 0.32154972520442693,
+      "learning_rate": 5.820006671841836e-06,
+      "loss": 0.2221,
+      "step": 2273
+    },
+    {
+      "epoch": 3.8805460750853245,
+      "grad_norm": 0.31894735071449115,
+      "learning_rate": 5.803213713870059e-06,
+      "loss": 0.2388,
+      "step": 2274
+    },
+    {
+      "epoch": 3.8822525597269624,
+      "grad_norm": 0.32642540892236,
+      "learning_rate": 5.786440905639785e-06,
+      "loss": 0.2871,
+      "step": 2275
+    },
+    {
+      "epoch": 3.8839590443686007,
+      "grad_norm": 0.34316752007851276,
+      "learning_rate": 5.769688270956955e-06,
+      "loss": 0.2549,
+      "step": 2276
+    },
+    {
+      "epoch": 3.8856655290102387,
+      "grad_norm": 0.33117472072087556,
+      "learning_rate": 5.7529558335988565e-06,
+      "loss": 0.1963,
+      "step": 2277
+    },
+    {
+      "epoch": 3.887372013651877,
+      "grad_norm": 0.3466509582332143,
+      "learning_rate": 5.736243617314141e-06,
+      "loss": 0.2527,
+      "step": 2278
+    },
+    {
+      "epoch": 3.8890784982935154,
+      "grad_norm": 0.36290764683736154,
+      "learning_rate": 5.719551645822732e-06,
+      "loss": 0.2143,
+      "step": 2279
+    },
+    {
+      "epoch": 3.8907849829351537,
+      "grad_norm": 0.3077642498955797,
+      "learning_rate": 5.702879942815827e-06,
+      "loss": 0.2556,
+      "step": 2280
+    },
+    {
+      "epoch": 3.8924914675767917,
+      "grad_norm": 0.29458913299329875,
+      "learning_rate": 5.686228531955868e-06,
+      "loss": 0.2473,
+      "step": 2281
+    },
+    {
+      "epoch": 3.89419795221843,
+      "grad_norm": 0.37471296755406247,
+      "learning_rate": 5.6695974368764795e-06,
+      "loss": 0.2266,
+      "step": 2282
+    },
+    {
+      "epoch": 3.8959044368600684,
+      "grad_norm": 0.32745265445978655,
+      "learning_rate": 5.652986681182469e-06,
+      "loss": 0.2553,
+      "step": 2283
+    },
+    {
+      "epoch": 3.8976109215017063,
+      "grad_norm": 0.32490435963202086,
+      "learning_rate": 5.6363962884497525e-06,
+      "loss": 0.2368,
+      "step": 2284
+    },
+    {
+      "epoch": 3.8993174061433447,
+      "grad_norm": 0.3216794269515537,
+      "learning_rate": 5.619826282225374e-06,
+      "loss": 0.2542,
+      "step": 2285
+    },
+    {
+      "epoch": 3.901023890784983,
+      "grad_norm": 0.3277939858056872,
+      "learning_rate": 5.603276686027415e-06,
+      "loss": 0.2526,
+      "step": 2286
+    },
+    {
+      "epoch": 3.9027303754266214,
+      "grad_norm": 0.30229387775788785,
+      "learning_rate": 5.586747523345e-06,
+      "loss": 0.257,
+      "step": 2287
+    },
+    {
+      "epoch": 3.9044368600682593,
+      "grad_norm": 0.2910221901714991,
+      "learning_rate": 5.570238817638261e-06,
+      "loss": 0.2414,
+      "step": 2288
+    },
+    {
+      "epoch": 3.9061433447098977,
+      "grad_norm": 0.3176948209851626,
+      "learning_rate": 5.553750592338274e-06,
+      "loss": 0.2415,
+      "step": 2289
+    },
+    {
+      "epoch": 3.9078498293515356,
+      "grad_norm": 0.37527095766258994,
+      "learning_rate": 5.537282870847071e-06,
+      "loss": 0.2223,
+      "step": 2290
+    },
+    {
+      "epoch": 3.909556313993174,
+      "grad_norm": 0.3196489250939008,
+      "learning_rate": 5.520835676537568e-06,
+      "loss": 0.2302,
+      "step": 2291
+    },
+    {
+      "epoch": 3.9112627986348123,
+      "grad_norm": 0.32730931571869526,
+      "learning_rate": 5.504409032753539e-06,
+      "loss": 0.2634,
+      "step": 2292
+    },
+    {
+      "epoch": 3.9129692832764507,
+      "grad_norm": 0.31263443085071635,
+      "learning_rate": 5.4880029628096154e-06,
+      "loss": 0.297,
+      "step": 2293
+    },
+    {
+      "epoch": 3.9146757679180886,
+      "grad_norm": 0.3062321902336411,
+      "learning_rate": 5.471617489991199e-06,
+      "loss": 0.2359,
+      "step": 2294
+    },
+    {
+      "epoch": 3.916382252559727,
+      "grad_norm": 0.3064753602125174,
+      "learning_rate": 5.455252637554485e-06,
+      "loss": 0.2169,
+      "step": 2295
+    },
+    {
+      "epoch": 3.9180887372013653,
+      "grad_norm": 0.3258559818378836,
+      "learning_rate": 5.438908428726375e-06,
+      "loss": 0.2459,
+      "step": 2296
+    },
+    {
+      "epoch": 3.919795221843003,
+      "grad_norm": 0.31517624160939745,
+      "learning_rate": 5.422584886704503e-06,
+      "loss": 0.2418,
+      "step": 2297
+    },
+    {
+      "epoch": 3.9215017064846416,
+      "grad_norm": 0.3254440422368787,
+      "learning_rate": 5.406282034657124e-06,
+      "loss": 0.2363,
+      "step": 2298
+    },
+    {
+      "epoch": 3.92320819112628,
+      "grad_norm": 0.35719119206231137,
+      "learning_rate": 5.389999895723171e-06,
+      "loss": 0.2463,
+      "step": 2299
+    },
+    {
+      "epoch": 3.9249146757679183,
+      "grad_norm": 0.34033229275445037,
+      "learning_rate": 5.3737384930121664e-06,
+      "loss": 0.2377,
+      "step": 2300
+    },
+    {
+      "epoch": 3.926621160409556,
+      "grad_norm": 0.31523156058285745,
+      "learning_rate": 5.357497849604185e-06,
+      "loss": 0.269,
+      "step": 2301
+    },
+    {
+      "epoch": 3.9283276450511946,
+      "grad_norm": 0.3346505640474263,
+      "learning_rate": 5.341277988549863e-06,
+      "loss": 0.2719,
+      "step": 2302
+    },
+    {
+      "epoch": 3.9300341296928325,
+      "grad_norm": 0.3224474214926856,
+      "learning_rate": 5.325078932870311e-06,
+      "loss": 0.1977,
+      "step": 2303
+    },
+    {
+      "epoch": 3.931740614334471,
+      "grad_norm": 0.3295470584186461,
+      "learning_rate": 5.308900705557147e-06,
+      "loss": 0.2422,
+      "step": 2304
+    },
+    {
+      "epoch": 3.9334470989761092,
+      "grad_norm": 0.3411314928029215,
+      "learning_rate": 5.292743329572381e-06,
+      "loss": 0.3056,
+      "step": 2305
+    },
+    {
+      "epoch": 3.9351535836177476,
+      "grad_norm": 0.31089877772154434,
+      "learning_rate": 5.276606827848463e-06,
+      "loss": 0.2731,
+      "step": 2306
+    },
+    {
+      "epoch": 3.9368600682593855,
+      "grad_norm": 0.3202486943451624,
+      "learning_rate": 5.2604912232882156e-06,
+      "loss": 0.223,
+      "step": 2307
+    },
+    {
+      "epoch": 3.938566552901024,
+      "grad_norm": 0.35034122491047537,
+      "learning_rate": 5.244396538764775e-06,
+      "loss": 0.2836,
+      "step": 2308
+    },
+    {
+      "epoch": 3.9402730375426622,
+      "grad_norm": 0.3493013375936862,
+      "learning_rate": 5.228322797121619e-06,
+      "loss": 0.2349,
+      "step": 2309
+    },
+    {
+      "epoch": 3.9419795221843,
+      "grad_norm": 0.34220479674079957,
+      "learning_rate": 5.212270021172477e-06,
+      "loss": 0.206,
+      "step": 2310
+    },
+    {
+      "epoch": 3.9436860068259385,
+      "grad_norm": 0.3280333198375722,
+      "learning_rate": 5.196238233701325e-06,
+      "loss": 0.2146,
+      "step": 2311
+    },
+    {
+      "epoch": 3.945392491467577,
+      "grad_norm": 0.3030451926278049,
+      "learning_rate": 5.18022745746235e-06,
+      "loss": 0.2659,
+      "step": 2312
+    },
+    {
+      "epoch": 3.9470989761092152,
+      "grad_norm": 0.3034563225242389,
+      "learning_rate": 5.164237715179925e-06,
+      "loss": 0.266,
+      "step": 2313
+    },
+    {
+      "epoch": 3.948805460750853,
+      "grad_norm": 0.33878900097053566,
+      "learning_rate": 5.148269029548571e-06,
+      "loss": 0.2592,
+      "step": 2314
+    },
+    {
+      "epoch": 3.9505119453924915,
+      "grad_norm": 0.29709169171885424,
+      "learning_rate": 5.132321423232906e-06,
+      "loss": 0.2514,
+      "step": 2315
+    },
+    {
+      "epoch": 3.9522184300341294,
+      "grad_norm": 0.30617037457440516,
+      "learning_rate": 5.116394918867655e-06,
+      "loss": 0.2784,
+      "step": 2316
+    },
+    {
+      "epoch": 3.953924914675768,
+      "grad_norm": 0.3277167449507937,
+      "learning_rate": 5.100489539057558e-06,
+      "loss": 0.2547,
+      "step": 2317
+    },
+    {
+      "epoch": 3.955631399317406,
+      "grad_norm": 0.3652591747948567,
+      "learning_rate": 5.084605306377408e-06,
+      "loss": 0.2121,
+      "step": 2318
+    },
+    {
+      "epoch": 3.9573378839590445,
+      "grad_norm": 0.33869441229654995,
+      "learning_rate": 5.068742243371958e-06,
+      "loss": 0.2413,
+      "step": 2319
+    },
+    {
+      "epoch": 3.9590443686006824,
+      "grad_norm": 0.3200939822375758,
+      "learning_rate": 5.0529003725559336e-06,
+      "loss": 0.2938,
+      "step": 2320
+    },
+    {
+      "epoch": 3.960750853242321,
+      "grad_norm": 0.3255871580066888,
+      "learning_rate": 5.037079716413962e-06,
+      "loss": 0.2306,
+      "step": 2321
+    },
+    {
+      "epoch": 3.962457337883959,
+      "grad_norm": 0.33548813018959317,
+      "learning_rate": 5.021280297400584e-06,
+      "loss": 0.2539,
+      "step": 2322
+    },
+    {
+      "epoch": 3.964163822525597,
+      "grad_norm": 0.3237828408910613,
+      "learning_rate": 5.005502137940179e-06,
+      "loss": 0.2591,
+      "step": 2323
+    },
+    {
+      "epoch": 3.9658703071672354,
+      "grad_norm": 0.3858997700192493,
+      "learning_rate": 4.989745260426952e-06,
+      "loss": 0.2894,
+      "step": 2324
+    },
+    {
+      "epoch": 3.967576791808874,
+      "grad_norm": 0.3107408904982704,
+      "learning_rate": 4.974009687224919e-06,
+      "loss": 0.2434,
+      "step": 2325
+    },
+    {
+      "epoch": 3.969283276450512,
+      "grad_norm": 0.2898058224896841,
+      "learning_rate": 4.95829544066784e-06,
+      "loss": 0.2727,
+      "step": 2326
+    },
+    {
+      "epoch": 3.97098976109215,
+      "grad_norm": 0.3488440701342725,
+      "learning_rate": 4.942602543059223e-06,
+      "loss": 0.228,
+      "step": 2327
+    },
+    {
+      "epoch": 3.9726962457337884,
+      "grad_norm": 0.3258586338371556,
+      "learning_rate": 4.926931016672259e-06,
+      "loss": 0.2544,
+      "step": 2328
+    },
+    {
+      "epoch": 3.9744027303754264,
+      "grad_norm": 0.330967486944079,
+      "learning_rate": 4.91128088374981e-06,
+      "loss": 0.2427,
+      "step": 2329
+    },
+    {
+      "epoch": 3.9761092150170647,
+      "grad_norm": 0.3520407541519079,
+      "learning_rate": 4.895652166504388e-06,
+      "loss": 0.3261,
+      "step": 2330
+    },
+    {
+      "epoch": 3.977815699658703,
+      "grad_norm": 0.3209888995589713,
+      "learning_rate": 4.880044887118087e-06,
+      "loss": 0.2429,
+      "step": 2331
+    },
+    {
+      "epoch": 3.9795221843003414,
+      "grad_norm": 0.33335280650284016,
+      "learning_rate": 4.864459067742595e-06,
+      "loss": 0.2401,
+      "step": 2332
+    },
+    {
+      "epoch": 3.98122866894198,
+      "grad_norm": 0.3202836880283363,
+      "learning_rate": 4.848894730499125e-06,
+      "loss": 0.2485,
+      "step": 2333
+    },
+    {
+      "epoch": 3.9829351535836177,
+      "grad_norm": 0.34117405364873554,
+      "learning_rate": 4.833351897478413e-06,
+      "loss": 0.2533,
+      "step": 2334
+    },
+    {
+      "epoch": 3.984641638225256,
+      "grad_norm": 0.30368478133877486,
+      "learning_rate": 4.817830590740666e-06,
+      "loss": 0.23,
+      "step": 2335
+    },
+    {
+      "epoch": 3.986348122866894,
+      "grad_norm": 0.34107955341438967,
+      "learning_rate": 4.802330832315534e-06,
+      "loss": 0.2074,
+      "step": 2336
+    },
+    {
+      "epoch": 3.9880546075085324,
+      "grad_norm": 0.3952011415516192,
+      "learning_rate": 4.786852644202098e-06,
+      "loss": 0.2755,
+      "step": 2337
+    },
+    {
+      "epoch": 3.9897610921501707,
+      "grad_norm": 0.32140344822370476,
+      "learning_rate": 4.771396048368806e-06,
+      "loss": 0.2435,
+      "step": 2338
+    },
+    {
+      "epoch": 3.991467576791809,
+      "grad_norm": 0.33607156170289026,
+      "learning_rate": 4.7559610667534806e-06,
+      "loss": 0.243,
+      "step": 2339
+    },
+    {
+      "epoch": 3.993174061433447,
+      "grad_norm": 0.309209732048636,
+      "learning_rate": 4.7405477212632404e-06,
+      "loss": 0.2509,
+      "step": 2340
+    },
+    {
+      "epoch": 3.9948805460750854,
+      "grad_norm": 0.32793268229295297,
+      "learning_rate": 4.725156033774523e-06,
+      "loss": 0.2386,
+      "step": 2341
+    },
+    {
+      "epoch": 3.9965870307167233,
+      "grad_norm": 0.33004739302554115,
+      "learning_rate": 4.70978602613301e-06,
+      "loss": 0.22,
+      "step": 2342
+    },
+    {
+      "epoch": 3.9982935153583616,
+      "grad_norm": 0.3779354844844152,
+      "learning_rate": 4.6944377201536085e-06,
+      "loss": 0.2241,
+      "step": 2343
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.3290126249506407,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.2136,
+      "step": 2344
+    },
+    {
+      "epoch": 4.001706484641638,
+      "grad_norm": 0.4195821381484528,
+      "learning_rate": 4.663806300286781e-06,
+      "loss": 0.181,
+      "step": 2345
+    },
+    {
+      "epoch": 4.003412969283277,
+      "grad_norm": 0.380339643137854,
+      "learning_rate": 4.648523229875057e-06,
+      "loss": 0.1879,
+      "step": 2346
+    },
+    {
+      "epoch": 4.005119453924915,
+      "grad_norm": 0.3186044596112882,
+      "learning_rate": 4.633261948076782e-06,
+      "loss": 0.1376,
+      "step": 2347
+    },
+    {
+      "epoch": 4.006825938566553,
+      "grad_norm": 0.31859418181041815,
+      "learning_rate": 4.618022476552553e-06,
+      "loss": 0.1901,
+      "step": 2348
+    },
+    {
+      "epoch": 4.008532423208191,
+      "grad_norm": 0.3635020562324159,
+      "learning_rate": 4.6028048369320195e-06,
+      "loss": 0.1927,
+      "step": 2349
+    },
+    {
+      "epoch": 4.010238907849829,
+      "grad_norm": 0.39829087215604436,
+      "learning_rate": 4.5876090508138305e-06,
+      "loss": 0.1687,
+      "step": 2350
+    },
+    {
+      "epoch": 4.011945392491468,
+      "grad_norm": 0.3747006339515895,
+      "learning_rate": 4.572435139765637e-06,
+      "loss": 0.1651,
+      "step": 2351
+    },
+    {
+      "epoch": 4.013651877133106,
+      "grad_norm": 0.35305428459036114,
+      "learning_rate": 4.557283125324012e-06,
+      "loss": 0.1892,
+      "step": 2352
+    },
+    {
+      "epoch": 4.015358361774744,
+      "grad_norm": 0.3267597738262304,
+      "learning_rate": 4.542153028994487e-06,
+      "loss": 0.1759,
+      "step": 2353
+    },
+    {
+      "epoch": 4.017064846416382,
+      "grad_norm": 0.34994523827678176,
+      "learning_rate": 4.527044872251458e-06,
+      "loss": 0.1868,
+      "step": 2354
+    },
+    {
+      "epoch": 4.01877133105802,
+      "grad_norm": 0.38266239654081796,
+      "learning_rate": 4.511958676538186e-06,
+      "loss": 0.1628,
+      "step": 2355
+    },
+    {
+      "epoch": 4.020477815699659,
+      "grad_norm": 0.40574871016875586,
+      "learning_rate": 4.4968944632667764e-06,
+      "loss": 0.1634,
+      "step": 2356
+    },
+    {
+      "epoch": 4.022184300341297,
+      "grad_norm": 0.3570647899059514,
+      "learning_rate": 4.481852253818113e-06,
+      "loss": 0.186,
+      "step": 2357
+    },
+    {
+      "epoch": 4.023890784982935,
+      "grad_norm": 0.35037574886916495,
+      "learning_rate": 4.4668320695418736e-06,
+      "loss": 0.1886,
+      "step": 2358
+    },
+    {
+      "epoch": 4.025597269624574,
+      "grad_norm": 0.3230046933285519,
+      "learning_rate": 4.451833931756457e-06,
+      "loss": 0.1969,
+      "step": 2359
+    },
+    {
+      "epoch": 4.027303754266212,
+      "grad_norm": 0.3246664508384852,
+      "learning_rate": 4.436857861748969e-06,
+      "loss": 0.184,
+      "step": 2360
+    },
+    {
+      "epoch": 4.0290102389078495,
+      "grad_norm": 0.3382136883702398,
+      "learning_rate": 4.4219038807752135e-06,
+      "loss": 0.1727,
+      "step": 2361
+    },
+    {
+      "epoch": 4.030716723549488,
+      "grad_norm": 0.3574461598883982,
+      "learning_rate": 4.406972010059623e-06,
+      "loss": 0.1693,
+      "step": 2362
+    },
+    {
+      "epoch": 4.032423208191126,
+      "grad_norm": 0.3159101539493147,
+      "learning_rate": 4.3920622707952635e-06,
+      "loss": 0.1863,
+      "step": 2363
+    },
+    {
+      "epoch": 4.034129692832765,
+      "grad_norm": 0.3027313233728572,
+      "learning_rate": 4.3771746841437765e-06,
+      "loss": 0.1543,
+      "step": 2364
+    },
+    {
+      "epoch": 4.035836177474403,
+      "grad_norm": 0.29884331247701995,
+      "learning_rate": 4.362309271235374e-06,
+      "loss": 0.1762,
+      "step": 2365
+    },
+    {
+      "epoch": 4.037542662116041,
+      "grad_norm": 0.31625900200267343,
+      "learning_rate": 4.3474660531687915e-06,
+      "loss": 0.179,
+      "step": 2366
+    },
+    {
+      "epoch": 4.039249146757679,
+      "grad_norm": 0.32403350412788073,
+      "learning_rate": 4.332645051011253e-06,
+      "loss": 0.2008,
+      "step": 2367
+    },
+    {
+      "epoch": 4.040955631399317,
+      "grad_norm": 0.30085829152297205,
+      "learning_rate": 4.3178462857984705e-06,
+      "loss": 0.1471,
+      "step": 2368
+    },
+    {
+      "epoch": 4.0426621160409555,
+      "grad_norm": 0.3434443386369388,
+      "learning_rate": 4.303069778534574e-06,
+      "loss": 0.1585,
+      "step": 2369
+    },
+    {
+      "epoch": 4.044368600682594,
+      "grad_norm": 0.3194979721641485,
+      "learning_rate": 4.288315550192126e-06,
+      "loss": 0.164,
+      "step": 2370
+    },
+    {
+      "epoch": 4.046075085324232,
+      "grad_norm": 0.31954843401356814,
+      "learning_rate": 4.273583621712041e-06,
+      "loss": 0.1791,
+      "step": 2371
+    },
+    {
+      "epoch": 4.047781569965871,
+      "grad_norm": 0.30033373224280846,
+      "learning_rate": 4.258874014003616e-06,
+      "loss": 0.1512,
+      "step": 2372
+    },
+    {
+      "epoch": 4.049488054607509,
+      "grad_norm": 0.33525231379828074,
+      "learning_rate": 4.244186747944425e-06,
+      "loss": 0.157,
+      "step": 2373
+    },
+    {
+      "epoch": 4.051194539249146,
+      "grad_norm": 0.3391082338188415,
+      "learning_rate": 4.2295218443803686e-06,
+      "loss": 0.131,
+      "step": 2374
+    },
+    {
+      "epoch": 4.052901023890785,
+      "grad_norm": 0.3236061516370624,
+      "learning_rate": 4.214879324125601e-06,
+      "loss": 0.1891,
+      "step": 2375
+    },
+    {
+      "epoch": 4.054607508532423,
+      "grad_norm": 0.30222161809585246,
+      "learning_rate": 4.200259207962487e-06,
+      "loss": 0.1932,
+      "step": 2376
+    },
+    {
+      "epoch": 4.0563139931740615,
+      "grad_norm": 0.2995732617456124,
+      "learning_rate": 4.185661516641622e-06,
+      "loss": 0.233,
+      "step": 2377
+    },
+    {
+      "epoch": 4.0580204778157,
+      "grad_norm": 0.2935346367299693,
+      "learning_rate": 4.171086270881752e-06,
+      "loss": 0.1659,
+      "step": 2378
+    },
+    {
+      "epoch": 4.059726962457338,
+      "grad_norm": 0.31560070678665797,
+      "learning_rate": 4.156533491369772e-06,
+      "loss": 0.1645,
+      "step": 2379
+    },
+    {
+      "epoch": 4.061433447098976,
+      "grad_norm": 0.30346193217864753,
+      "learning_rate": 4.142003198760685e-06,
+      "loss": 0.1734,
+      "step": 2380
+    },
+    {
+      "epoch": 4.063139931740614,
+      "grad_norm": 0.30241178077042385,
+      "learning_rate": 4.127495413677592e-06,
+      "loss": 0.1854,
+      "step": 2381
+    },
+    {
+      "epoch": 4.064846416382252,
+      "grad_norm": 0.30940900115979975,
+      "learning_rate": 4.1130101567116435e-06,
+      "loss": 0.225,
+      "step": 2382
+    },
+    {
+      "epoch": 4.066552901023891,
+      "grad_norm": 0.2980665181561457,
+      "learning_rate": 4.0985474484219986e-06,
+      "loss": 0.1982,
+      "step": 2383
+    },
+    {
+      "epoch": 4.068259385665529,
+      "grad_norm": 0.29077509347783315,
+      "learning_rate": 4.08410730933585e-06,
+      "loss": 0.1957,
+      "step": 2384
+    },
+    {
+      "epoch": 4.0699658703071675,
+      "grad_norm": 0.32745222882898095,
+      "learning_rate": 4.069689759948308e-06,
+      "loss": 0.1886,
+      "step": 2385
+    },
+    {
+      "epoch": 4.071672354948806,
+      "grad_norm": 0.30247319005231443,
+      "learning_rate": 4.055294820722462e-06,
+      "loss": 0.1575,
+      "step": 2386
+    },
+    {
+      "epoch": 4.073378839590443,
+      "grad_norm": 0.30760195988207906,
+      "learning_rate": 4.040922512089287e-06,
+      "loss": 0.1918,
+      "step": 2387
+    },
+    {
+      "epoch": 4.075085324232082,
+      "grad_norm": 0.36117705419328205,
+      "learning_rate": 4.02657285444765e-06,
+      "loss": 0.1721,
+      "step": 2388
+    },
+    {
+      "epoch": 4.07679180887372,
+      "grad_norm": 0.3094996833260368,
+      "learning_rate": 4.012245868164273e-06,
+      "loss": 0.198,
+      "step": 2389
+    },
+    {
+      "epoch": 4.078498293515358,
+      "grad_norm": 0.317353138990636,
+      "learning_rate": 3.997941573573685e-06,
+      "loss": 0.2555,
+      "step": 2390
+    },
+    {
+      "epoch": 4.080204778156997,
+      "grad_norm": 0.28796017944724317,
+      "learning_rate": 3.983659990978217e-06,
+      "loss": 0.1837,
+      "step": 2391
+    },
+    {
+      "epoch": 4.081911262798635,
+      "grad_norm": 0.31493694678711875,
+      "learning_rate": 3.969401140647957e-06,
+      "loss": 0.1891,
+      "step": 2392
+    },
+    {
+      "epoch": 4.083617747440273,
+      "grad_norm": 0.3113813772180343,
+      "learning_rate": 3.955165042820748e-06,
+      "loss": 0.1497,
+      "step": 2393
+    },
+    {
+      "epoch": 4.085324232081911,
+      "grad_norm": 0.28613972465653337,
+      "learning_rate": 3.940951717702115e-06,
+      "loss": 0.1808,
+      "step": 2394
+    },
+    {
+      "epoch": 4.087030716723549,
+      "grad_norm": 0.3173301121228953,
+      "learning_rate": 3.926761185465277e-06,
+      "loss": 0.1959,
+      "step": 2395
+    },
+    {
+      "epoch": 4.088737201365188,
+      "grad_norm": 0.28874789354399394,
+      "learning_rate": 3.912593466251111e-06,
+      "loss": 0.2134,
+      "step": 2396
+    },
+    {
+      "epoch": 4.090443686006826,
+      "grad_norm": 0.3179204985820542,
+      "learning_rate": 3.898448580168084e-06,
+      "loss": 0.2062,
+      "step": 2397
+    },
+    {
+      "epoch": 4.092150170648464,
+      "grad_norm": 0.3498702154663877,
+      "learning_rate": 3.8843265472922874e-06,
+      "loss": 0.1824,
+      "step": 2398
+    },
+    {
+      "epoch": 4.093856655290103,
+      "grad_norm": 0.3181203277230868,
+      "learning_rate": 3.870227387667355e-06,
+      "loss": 0.1751,
+      "step": 2399
+    },
+    {
+      "epoch": 4.09556313993174,
+      "grad_norm": 0.32833596189353553,
+      "learning_rate": 3.856151121304477e-06,
+      "loss": 0.1827,
+      "step": 2400
+    },
+    {
+      "epoch": 4.097269624573379,
+      "grad_norm": 0.3097861506682424,
+      "learning_rate": 3.842097768182324e-06,
+      "loss": 0.1596,
+      "step": 2401
+    },
+    {
+      "epoch": 4.098976109215017,
+      "grad_norm": 0.29847487443923326,
+      "learning_rate": 3.828067348247076e-06,
+      "loss": 0.1888,
+      "step": 2402
+    },
+    {
+      "epoch": 4.100682593856655,
+      "grad_norm": 0.32331562735973857,
+      "learning_rate": 3.8140598814123374e-06,
+      "loss": 0.1648,
+      "step": 2403
+    },
+    {
+      "epoch": 4.102389078498294,
+      "grad_norm": 0.31911722607997095,
+      "learning_rate": 3.8000753875591455e-06,
+      "loss": 0.1648,
+      "step": 2404
+    },
+    {
+      "epoch": 4.104095563139932,
+      "grad_norm": 0.3168256761024783,
+      "learning_rate": 3.7861138865359383e-06,
+      "loss": 0.1689,
+      "step": 2405
+    },
+    {
+      "epoch": 4.1058020477815695,
+      "grad_norm": 0.3201312099698668,
+      "learning_rate": 3.772175398158504e-06,
+      "loss": 0.1677,
+      "step": 2406
+    },
+    {
+      "epoch": 4.107508532423208,
+      "grad_norm": 0.3088327738845882,
+      "learning_rate": 3.7582599422099873e-06,
+      "loss": 0.1412,
+      "step": 2407
+    },
+    {
+      "epoch": 4.109215017064846,
+      "grad_norm": 0.30505958330845273,
+      "learning_rate": 3.744367538440823e-06,
+      "loss": 0.1517,
+      "step": 2408
+    },
+    {
+      "epoch": 4.110921501706485,
+      "grad_norm": 0.32653606387534573,
+      "learning_rate": 3.7304982065687447e-06,
+      "loss": 0.1906,
+      "step": 2409
+    },
+    {
+      "epoch": 4.112627986348123,
+      "grad_norm": 0.332667019462028,
+      "learning_rate": 3.7166519662787327e-06,
+      "loss": 0.1678,
+      "step": 2410
+    },
+    {
+      "epoch": 4.114334470989761,
+      "grad_norm": 0.2994498270660051,
+      "learning_rate": 3.7028288372229825e-06,
+      "loss": 0.1931,
+      "step": 2411
+    },
+    {
+      "epoch": 4.1160409556314,
+      "grad_norm": 0.28551721653406226,
+      "learning_rate": 3.6890288390209093e-06,
+      "loss": 0.1698,
+      "step": 2412
+    },
+    {
+      "epoch": 4.117747440273037,
+      "grad_norm": 0.31365733078967856,
+      "learning_rate": 3.675251991259079e-06,
+      "loss": 0.2001,
+      "step": 2413
+    },
+    {
+      "epoch": 4.1194539249146755,
+      "grad_norm": 0.31358514885701966,
+      "learning_rate": 3.661498313491214e-06,
+      "loss": 0.1492,
+      "step": 2414
+    },
+    {
+      "epoch": 4.121160409556314,
+      "grad_norm": 0.29088423041712363,
+      "learning_rate": 3.6477678252381375e-06,
+      "loss": 0.1716,
+      "step": 2415
+    },
+    {
+      "epoch": 4.122866894197952,
+      "grad_norm": 0.31473294787636624,
+      "learning_rate": 3.6340605459877675e-06,
+      "loss": 0.1633,
+      "step": 2416
+    },
+    {
+      "epoch": 4.124573378839591,
+      "grad_norm": 0.337747855502917,
+      "learning_rate": 3.6203764951950836e-06,
+      "loss": 0.184,
+      "step": 2417
+    },
+    {
+      "epoch": 4.126279863481229,
+      "grad_norm": 0.3060822525059725,
+      "learning_rate": 3.6067156922820877e-06,
+      "loss": 0.1865,
+      "step": 2418
+    },
+    {
+      "epoch": 4.1279863481228665,
+      "grad_norm": 0.3430301730706332,
+      "learning_rate": 3.593078156637797e-06,
+      "loss": 0.1935,
+      "step": 2419
+    },
+    {
+      "epoch": 4.129692832764505,
+      "grad_norm": 0.30029683914160743,
+      "learning_rate": 3.5794639076181924e-06,
+      "loss": 0.1868,
+      "step": 2420
+    },
+    {
+      "epoch": 4.131399317406143,
+      "grad_norm": 0.3452712998400332,
+      "learning_rate": 3.5658729645462175e-06,
+      "loss": 0.1679,
+      "step": 2421
+    },
+    {
+      "epoch": 4.1331058020477816,
+      "grad_norm": 0.2948759494429693,
+      "learning_rate": 3.5523053467117287e-06,
+      "loss": 0.1866,
+      "step": 2422
+    },
+    {
+      "epoch": 4.13481228668942,
+      "grad_norm": 0.346765414439053,
+      "learning_rate": 3.5387610733714685e-06,
+      "loss": 0.1677,
+      "step": 2423
+    },
+    {
+      "epoch": 4.136518771331058,
+      "grad_norm": 0.3307964280323542,
+      "learning_rate": 3.5252401637490683e-06,
+      "loss": 0.1778,
+      "step": 2424
+    },
+    {
+      "epoch": 4.138225255972697,
+      "grad_norm": 0.2995036414158578,
+      "learning_rate": 3.5117426370349763e-06,
+      "loss": 0.1825,
+      "step": 2425
+    },
+    {
+      "epoch": 4.139931740614334,
+      "grad_norm": 0.32137369482037303,
+      "learning_rate": 3.4982685123864712e-06,
+      "loss": 0.1963,
+      "step": 2426
+    },
+    {
+      "epoch": 4.1416382252559725,
+      "grad_norm": 0.30769668129711114,
+      "learning_rate": 3.484817808927605e-06,
+      "loss": 0.1787,
+      "step": 2427
+    },
+    {
+      "epoch": 4.143344709897611,
+      "grad_norm": 0.33124297434962124,
+      "learning_rate": 3.471390545749187e-06,
+      "loss": 0.1649,
+      "step": 2428
+    },
+    {
+      "epoch": 4.145051194539249,
+      "grad_norm": 0.3157941746181479,
+      "learning_rate": 3.4579867419087696e-06,
+      "loss": 0.1941,
+      "step": 2429
+    },
+    {
+      "epoch": 4.146757679180888,
+      "grad_norm": 0.31393737449028575,
+      "learning_rate": 3.444606416430594e-06,
+      "loss": 0.1619,
+      "step": 2430
+    },
+    {
+      "epoch": 4.148464163822526,
+      "grad_norm": 0.29530891100487067,
+      "learning_rate": 3.4312495883055898e-06,
+      "loss": 0.1539,
+      "step": 2431
+    },
+    {
+      "epoch": 4.150170648464163,
+      "grad_norm": 0.3001566132614726,
+      "learning_rate": 3.417916276491324e-06,
+      "loss": 0.1701,
+      "step": 2432
+    },
+    {
+      "epoch": 4.151877133105802,
+      "grad_norm": 0.31297145051268754,
+      "learning_rate": 3.404606499912004e-06,
+      "loss": 0.2238,
+      "step": 2433
+    },
+    {
+      "epoch": 4.15358361774744,
+      "grad_norm": 0.3008453140731978,
+      "learning_rate": 3.3913202774584187e-06,
+      "loss": 0.1826,
+      "step": 2434
+    },
+    {
+      "epoch": 4.1552901023890785,
+      "grad_norm": 0.33268967523986326,
+      "learning_rate": 3.378057627987925e-06,
+      "loss": 0.2026,
+      "step": 2435
+    },
+    {
+      "epoch": 4.156996587030717,
+      "grad_norm": 0.31920950857284724,
+      "learning_rate": 3.3648185703244396e-06,
+      "loss": 0.1999,
+      "step": 2436
+    },
+    {
+      "epoch": 4.158703071672355,
+      "grad_norm": 0.30205249689919317,
+      "learning_rate": 3.3516031232583737e-06,
+      "loss": 0.1965,
+      "step": 2437
+    },
+    {
+      "epoch": 4.160409556313994,
+      "grad_norm": 0.3074081500263038,
+      "learning_rate": 3.3384113055466428e-06,
+      "loss": 0.2154,
+      "step": 2438
+    },
+    {
+      "epoch": 4.162116040955631,
+      "grad_norm": 0.34729139465173037,
+      "learning_rate": 3.3252431359126147e-06,
+      "loss": 0.1595,
+      "step": 2439
+    },
+    {
+      "epoch": 4.163822525597269,
+      "grad_norm": 0.3054807917727276,
+      "learning_rate": 3.3120986330461036e-06,
+      "loss": 0.149,
+      "step": 2440
+    },
+    {
+      "epoch": 4.165529010238908,
+      "grad_norm": 0.30757842212638165,
+      "learning_rate": 3.2989778156033257e-06,
+      "loss": 0.1673,
+      "step": 2441
+    },
+    {
+      "epoch": 4.167235494880546,
+      "grad_norm": 0.3142441653237269,
+      "learning_rate": 3.285880702206874e-06,
+      "loss": 0.1913,
+      "step": 2442
+    },
+    {
+      "epoch": 4.1689419795221845,
+      "grad_norm": 0.3152934803543005,
+      "learning_rate": 3.272807311445716e-06,
+      "loss": 0.1665,
+      "step": 2443
+    },
+    {
+      "epoch": 4.170648464163823,
+      "grad_norm": 0.3286907418748899,
+      "learning_rate": 3.259757661875129e-06,
+      "loss": 0.1797,
+      "step": 2444
+    },
+    {
+      "epoch": 4.172354948805461,
+      "grad_norm": 0.30941258169800717,
+      "learning_rate": 3.2467317720167135e-06,
+      "loss": 0.1973,
+      "step": 2445
+    },
+    {
+      "epoch": 4.174061433447099,
+      "grad_norm": 0.3055585061508135,
+      "learning_rate": 3.2337296603583336e-06,
+      "loss": 0.1947,
+      "step": 2446
+    },
+    {
+      "epoch": 4.175767918088737,
+      "grad_norm": 0.34521820630003597,
+      "learning_rate": 3.2207513453541027e-06,
+      "loss": 0.176,
+      "step": 2447
+    },
+    {
+      "epoch": 4.177474402730375,
+      "grad_norm": 0.2809607338036196,
+      "learning_rate": 3.2077968454243757e-06,
+      "loss": 0.181,
+      "step": 2448
+    },
+    {
+      "epoch": 4.179180887372014,
+      "grad_norm": 0.33299427285978445,
+      "learning_rate": 3.1948661789556844e-06,
+      "loss": 0.1574,
+      "step": 2449
+    },
+    {
+      "epoch": 4.180887372013652,
+      "grad_norm": 0.30328282365657666,
+      "learning_rate": 3.1819593643007574e-06,
+      "loss": 0.1883,
+      "step": 2450
+    },
+    {
+      "epoch": 4.1825938566552905,
+      "grad_norm": 0.33795654538569,
+      "learning_rate": 3.1690764197784453e-06,
+      "loss": 0.1714,
+      "step": 2451
+    },
+    {
+      "epoch": 4.184300341296928,
+      "grad_norm": 0.30990167385479767,
+      "learning_rate": 3.156217363673748e-06,
+      "loss": 0.1939,
+      "step": 2452
+    },
+    {
+      "epoch": 4.186006825938566,
+      "grad_norm": 0.31593916081353907,
+      "learning_rate": 3.1433822142377222e-06,
+      "loss": 0.1479,
+      "step": 2453
+    },
+    {
+      "epoch": 4.187713310580205,
+      "grad_norm": 0.30879416774852353,
+      "learning_rate": 3.1305709896875267e-06,
+      "loss": 0.2035,
+      "step": 2454
+    },
+    {
+      "epoch": 4.189419795221843,
+      "grad_norm": 0.31106106097285646,
+      "learning_rate": 3.1177837082063565e-06,
+      "loss": 0.1769,
+      "step": 2455
+    },
+    {
+      "epoch": 4.191126279863481,
+      "grad_norm": 0.3285079312818511,
+      "learning_rate": 3.105020387943405e-06,
+      "loss": 0.1665,
+      "step": 2456
+    },
+    {
+      "epoch": 4.19283276450512,
+      "grad_norm": 0.35199684412836346,
+      "learning_rate": 3.0922810470138854e-06,
+      "loss": 0.1617,
+      "step": 2457
+    },
+    {
+      "epoch": 4.194539249146757,
+      "grad_norm": 0.31623575446419255,
+      "learning_rate": 3.079565703498957e-06,
+      "loss": 0.1905,
+      "step": 2458
+    },
+    {
+      "epoch": 4.196245733788396,
+      "grad_norm": 0.31649344891525233,
+      "learning_rate": 3.0668743754457207e-06,
+      "loss": 0.1663,
+      "step": 2459
+    },
+    {
+      "epoch": 4.197952218430034,
+      "grad_norm": 0.3283474471234332,
+      "learning_rate": 3.054207080867191e-06,
+      "loss": 0.1781,
+      "step": 2460
+    },
+    {
+      "epoch": 4.199658703071672,
+      "grad_norm": 0.30898068543856966,
+      "learning_rate": 3.0415638377422853e-06,
+      "loss": 0.2372,
+      "step": 2461
+    },
+    {
+      "epoch": 4.201365187713311,
+      "grad_norm": 0.31695022179054894,
+      "learning_rate": 3.0289446640157736e-06,
+      "loss": 0.1655,
+      "step": 2462
+    },
+    {
+      "epoch": 4.203071672354949,
+      "grad_norm": 0.2830330422400834,
+      "learning_rate": 3.016349577598261e-06,
+      "loss": 0.1886,
+      "step": 2463
+    },
+    {
+      "epoch": 4.204778156996587,
+      "grad_norm": 0.2892447590865804,
+      "learning_rate": 3.003778596366178e-06,
+      "loss": 0.1727,
+      "step": 2464
+    },
+    {
+      "epoch": 4.206484641638225,
+      "grad_norm": 0.3289191933004354,
+      "learning_rate": 2.991231738161717e-06,
+      "loss": 0.1761,
+      "step": 2465
+    },
+    {
+      "epoch": 4.208191126279863,
+      "grad_norm": 0.29484730013264737,
+      "learning_rate": 2.9787090207928627e-06,
+      "loss": 0.1913,
+      "step": 2466
+    },
+    {
+      "epoch": 4.209897610921502,
+      "grad_norm": 0.3409795446165211,
+      "learning_rate": 2.9662104620333122e-06,
+      "loss": 0.1706,
+      "step": 2467
+    },
+    {
+      "epoch": 4.21160409556314,
+      "grad_norm": 0.29846831357916814,
+      "learning_rate": 2.953736079622487e-06,
+      "loss": 0.2099,
+      "step": 2468
+    },
+    {
+      "epoch": 4.213310580204778,
+      "grad_norm": 0.32455537188833833,
+      "learning_rate": 2.9412858912654973e-06,
+      "loss": 0.1622,
+      "step": 2469
+    },
+    {
+      "epoch": 4.215017064846417,
+      "grad_norm": 0.2924663053130245,
+      "learning_rate": 2.9288599146331043e-06,
+      "loss": 0.1695,
+      "step": 2470
+    },
+    {
+      "epoch": 4.216723549488055,
+      "grad_norm": 0.31636765141272133,
+      "learning_rate": 2.916458167361709e-06,
+      "loss": 0.1685,
+      "step": 2471
+    },
+    {
+      "epoch": 4.2184300341296925,
+      "grad_norm": 0.29890500727297004,
+      "learning_rate": 2.904080667053315e-06,
+      "loss": 0.1848,
+      "step": 2472
+    },
+    {
+      "epoch": 4.220136518771331,
+      "grad_norm": 0.3159287635158706,
+      "learning_rate": 2.891727431275535e-06,
+      "loss": 0.166,
+      "step": 2473
+    },
+    {
+      "epoch": 4.221843003412969,
+      "grad_norm": 0.32374950088405524,
+      "learning_rate": 2.879398477561515e-06,
+      "loss": 0.1511,
+      "step": 2474
+    },
+    {
+      "epoch": 4.223549488054608,
+      "grad_norm": 0.29710815719678113,
+      "learning_rate": 2.8670938234099544e-06,
+      "loss": 0.216,
+      "step": 2475
+    },
+    {
+      "epoch": 4.225255972696246,
+      "grad_norm": 0.30868037596993714,
+      "learning_rate": 2.854813486285066e-06,
+      "loss": 0.21,
+      "step": 2476
+    },
+    {
+      "epoch": 4.226962457337884,
+      "grad_norm": 0.3152187498011452,
+      "learning_rate": 2.8425574836165347e-06,
+      "loss": 0.1402,
+      "step": 2477
+    },
+    {
+      "epoch": 4.228668941979522,
+      "grad_norm": 0.31038840558797376,
+      "learning_rate": 2.8303258327995164e-06,
+      "loss": 0.1557,
+      "step": 2478
+    },
+    {
+      "epoch": 4.23037542662116,
+      "grad_norm": 0.34017375830113183,
+      "learning_rate": 2.8181185511945997e-06,
+      "loss": 0.1758,
+      "step": 2479
+    },
+    {
+      "epoch": 4.2320819112627985,
+      "grad_norm": 0.31419921149308355,
+      "learning_rate": 2.805935656127794e-06,
+      "loss": 0.179,
+      "step": 2480
+    },
+    {
+      "epoch": 4.233788395904437,
+      "grad_norm": 0.3087576948893619,
+      "learning_rate": 2.793777164890481e-06,
+      "loss": 0.2254,
+      "step": 2481
+    },
+    {
+      "epoch": 4.235494880546075,
+      "grad_norm": 0.2897950303673048,
+      "learning_rate": 2.7816430947394234e-06,
+      "loss": 0.1704,
+      "step": 2482
+    },
+    {
+      "epoch": 4.237201365187714,
+      "grad_norm": 0.30271889939946317,
+      "learning_rate": 2.7695334628967186e-06,
+      "loss": 0.1715,
+      "step": 2483
+    },
+    {
+      "epoch": 4.238907849829351,
+      "grad_norm": 0.32471275307318515,
+      "learning_rate": 2.757448286549762e-06,
+      "loss": 0.1731,
+      "step": 2484
+    },
+    {
+      "epoch": 4.2406143344709895,
+      "grad_norm": 0.30425370243267175,
+      "learning_rate": 2.74538758285126e-06,
+      "loss": 0.1662,
+      "step": 2485
+    },
+    {
+      "epoch": 4.242320819112628,
+      "grad_norm": 0.3310020579739023,
+      "learning_rate": 2.733351368919166e-06,
+      "loss": 0.1838,
+      "step": 2486
+    },
+    {
+      "epoch": 4.244027303754266,
+      "grad_norm": 0.3229193120957275,
+      "learning_rate": 2.7213396618366973e-06,
+      "loss": 0.1635,
+      "step": 2487
+    },
+    {
+      "epoch": 4.2457337883959045,
+      "grad_norm": 0.3057093282694775,
+      "learning_rate": 2.709352478652263e-06,
+      "loss": 0.1788,
+      "step": 2488
+    },
+    {
+      "epoch": 4.247440273037543,
+      "grad_norm": 0.32147087828457116,
+      "learning_rate": 2.697389836379487e-06,
+      "loss": 0.1454,
+      "step": 2489
+    },
+    {
+      "epoch": 4.249146757679181,
+      "grad_norm": 0.2910842326849101,
+      "learning_rate": 2.685451751997148e-06,
+      "loss": 0.178,
+      "step": 2490
+    },
+    {
+      "epoch": 4.250853242320819,
+      "grad_norm": 0.3054245893841373,
+      "learning_rate": 2.6735382424491675e-06,
+      "loss": 0.1415,
+      "step": 2491
+    },
+    {
+      "epoch": 4.252559726962457,
+      "grad_norm": 0.3052755561261774,
+      "learning_rate": 2.661649324644604e-06,
+      "loss": 0.1798,
+      "step": 2492
+    },
+    {
+      "epoch": 4.2542662116040955,
+      "grad_norm": 0.3182082318814107,
+      "learning_rate": 2.649785015457591e-06,
+      "loss": 0.1748,
+      "step": 2493
+    },
+    {
+      "epoch": 4.255972696245734,
+      "grad_norm": 0.29962345203407825,
+      "learning_rate": 2.637945331727356e-06,
+      "loss": 0.1745,
+      "step": 2494
+    },
+    {
+      "epoch": 4.257679180887372,
+      "grad_norm": 0.32292803056798236,
+      "learning_rate": 2.6261302902581597e-06,
+      "loss": 0.1657,
+      "step": 2495
+    },
+    {
+      "epoch": 4.2593856655290105,
+      "grad_norm": 0.3132811007902274,
+      "learning_rate": 2.6143399078192877e-06,
+      "loss": 0.1414,
+      "step": 2496
+    },
+    {
+      "epoch": 4.261092150170649,
+      "grad_norm": 0.33403720279621085,
+      "learning_rate": 2.6025742011450406e-06,
+      "loss": 0.1657,
+      "step": 2497
+    },
+    {
+      "epoch": 4.262798634812286,
+      "grad_norm": 0.32096632949756077,
+      "learning_rate": 2.590833186934676e-06,
+      "loss": 0.1731,
+      "step": 2498
+    },
+    {
+      "epoch": 4.264505119453925,
+      "grad_norm": 0.29495802159678225,
+      "learning_rate": 2.5791168818524283e-06,
+      "loss": 0.1773,
+      "step": 2499
+    },
+    {
+      "epoch": 4.266211604095563,
+      "grad_norm": 0.2944433363612561,
+      "learning_rate": 2.5674253025274396e-06,
+      "loss": 0.198,
+      "step": 2500
+    },
+    {
+      "epoch": 4.2679180887372015,
+      "grad_norm": 0.3176713163870496,
+      "learning_rate": 2.5557584655537746e-06,
+      "loss": 0.1539,
+      "step": 2501
+    },
+    {
+      "epoch": 4.26962457337884,
+      "grad_norm": 0.2853734440639459,
+      "learning_rate": 2.5441163874903742e-06,
+      "loss": 0.1919,
+      "step": 2502
+    },
+    {
+      "epoch": 4.271331058020478,
+      "grad_norm": 0.32089791476011936,
+      "learning_rate": 2.532499084861033e-06,
+      "loss": 0.1641,
+      "step": 2503
+    },
+    {
+      "epoch": 4.273037542662116,
+      "grad_norm": 0.3061287654995635,
+      "learning_rate": 2.5209065741543936e-06,
+      "loss": 0.1725,
+      "step": 2504
+    },
+    {
+      "epoch": 4.274744027303754,
+      "grad_norm": 0.27879825597833974,
+      "learning_rate": 2.5093388718238987e-06,
+      "loss": 0.139,
+      "step": 2505
+    },
+    {
+      "epoch": 4.276450511945392,
+      "grad_norm": 0.31504174633544063,
+      "learning_rate": 2.497795994287795e-06,
+      "loss": 0.1702,
+      "step": 2506
+    },
+    {
+      "epoch": 4.278156996587031,
+      "grad_norm": 0.35568855400749266,
+      "learning_rate": 2.4862779579290797e-06,
+      "loss": 0.1431,
+      "step": 2507
+    },
+    {
+      "epoch": 4.279863481228669,
+      "grad_norm": 0.31786679741994495,
+      "learning_rate": 2.474784779095496e-06,
+      "loss": 0.1591,
+      "step": 2508
+    },
+    {
+      "epoch": 4.2815699658703075,
+      "grad_norm": 0.29747899647345816,
+      "learning_rate": 2.4633164740995154e-06,
+      "loss": 0.2301,
+      "step": 2509
+    },
+    {
+      "epoch": 4.283276450511945,
+      "grad_norm": 0.29501765904829275,
+      "learning_rate": 2.4518730592182926e-06,
+      "loss": 0.1597,
+      "step": 2510
+    },
+    {
+      "epoch": 4.284982935153583,
+      "grad_norm": 0.2955468959025003,
+      "learning_rate": 2.4404545506936716e-06,
+      "loss": 0.2034,
+      "step": 2511
+    },
+    {
+      "epoch": 4.286689419795222,
+      "grad_norm": 0.30088377990607584,
+      "learning_rate": 2.429060964732126e-06,
+      "loss": 0.2104,
+      "step": 2512
+    },
+    {
+      "epoch": 4.28839590443686,
+      "grad_norm": 0.3020198754204438,
+      "learning_rate": 2.4176923175047763e-06,
+      "loss": 0.1778,
+      "step": 2513
+    },
+    {
+      "epoch": 4.290102389078498,
+      "grad_norm": 0.3102244929124343,
+      "learning_rate": 2.4063486251473344e-06,
+      "loss": 0.1793,
+      "step": 2514
+    },
+    {
+      "epoch": 4.291808873720137,
+      "grad_norm": 0.3253556220298246,
+      "learning_rate": 2.395029903760091e-06,
+      "loss": 0.1489,
+      "step": 2515
+    },
+    {
+      "epoch": 4.293515358361775,
+      "grad_norm": 0.3410550422933944,
+      "learning_rate": 2.3837361694079107e-06,
+      "loss": 0.1485,
+      "step": 2516
+    },
+    {
+      "epoch": 4.295221843003413,
+      "grad_norm": 0.3130313366088956,
+      "learning_rate": 2.372467438120174e-06,
+      "loss": 0.1833,
+      "step": 2517
+    },
+    {
+      "epoch": 4.296928327645051,
+      "grad_norm": 0.3120907160997788,
+      "learning_rate": 2.3612237258907957e-06,
+      "loss": 0.1779,
+      "step": 2518
+    },
+    {
+      "epoch": 4.298634812286689,
+      "grad_norm": 0.2922644753111327,
+      "learning_rate": 2.3500050486781566e-06,
+      "loss": 0.1672,
+      "step": 2519
+    },
+    {
+      "epoch": 4.300341296928328,
+      "grad_norm": 0.3227869521532051,
+      "learning_rate": 2.338811422405127e-06,
+      "loss": 0.1571,
+      "step": 2520
+    },
+    {
+      "epoch": 4.302047781569966,
+      "grad_norm": 0.3034733437198407,
+      "learning_rate": 2.3276428629590075e-06,
+      "loss": 0.2009,
+      "step": 2521
+    },
+    {
+      "epoch": 4.303754266211604,
+      "grad_norm": 0.3263899032863054,
+      "learning_rate": 2.316499386191522e-06,
+      "loss": 0.1496,
+      "step": 2522
+    },
+    {
+      "epoch": 4.305460750853243,
+      "grad_norm": 0.30830203624371233,
+      "learning_rate": 2.3053810079188057e-06,
+      "loss": 0.1644,
+      "step": 2523
+    },
+    {
+      "epoch": 4.30716723549488,
+      "grad_norm": 0.3050694718628497,
+      "learning_rate": 2.2942877439213528e-06,
+      "loss": 0.1994,
+      "step": 2524
+    },
+    {
+      "epoch": 4.308873720136519,
+      "grad_norm": 0.29845586994255363,
+      "learning_rate": 2.283219609944034e-06,
+      "loss": 0.1786,
+      "step": 2525
+    },
+    {
+      "epoch": 4.310580204778157,
+      "grad_norm": 0.3418072194790865,
+      "learning_rate": 2.272176621696034e-06,
+      "loss": 0.1735,
+      "step": 2526
+    },
+    {
+      "epoch": 4.312286689419795,
+      "grad_norm": 0.29544505841648233,
+      "learning_rate": 2.261158794850853e-06,
+      "loss": 0.1902,
+      "step": 2527
+    },
+    {
+      "epoch": 4.313993174061434,
+      "grad_norm": 0.31134334943651026,
+      "learning_rate": 2.2501661450462886e-06,
+      "loss": 0.179,
+      "step": 2528
+    },
+    {
+      "epoch": 4.315699658703072,
+      "grad_norm": 0.301310338429921,
+      "learning_rate": 2.2391986878843876e-06,
+      "loss": 0.1682,
+      "step": 2529
+    },
+    {
+      "epoch": 4.3174061433447095,
+      "grad_norm": 0.29330584727835124,
+      "learning_rate": 2.2282564389314576e-06,
+      "loss": 0.1794,
+      "step": 2530
+    },
+    {
+      "epoch": 4.319112627986348,
+      "grad_norm": 0.3070215194233873,
+      "learning_rate": 2.217339413718014e-06,
+      "loss": 0.1417,
+      "step": 2531
+    },
+    {
+      "epoch": 4.320819112627986,
+      "grad_norm": 0.30821668681842695,
+      "learning_rate": 2.2064476277387858e-06,
+      "loss": 0.1736,
+      "step": 2532
+    },
+    {
+      "epoch": 4.322525597269625,
+      "grad_norm": 0.3184121253545697,
+      "learning_rate": 2.1955810964526593e-06,
+      "loss": 0.2005,
+      "step": 2533
+    },
+    {
+      "epoch": 4.324232081911263,
+      "grad_norm": 0.31780555744096844,
+      "learning_rate": 2.18473983528269e-06,
+      "loss": 0.1924,
+      "step": 2534
+    },
+    {
+      "epoch": 4.325938566552901,
+      "grad_norm": 0.3072965040016955,
+      "learning_rate": 2.173923859616076e-06,
+      "loss": 0.18,
+      "step": 2535
+    },
+    {
+      "epoch": 4.327645051194539,
+      "grad_norm": 0.2824134182741914,
+      "learning_rate": 2.1631331848041025e-06,
+      "loss": 0.16,
+      "step": 2536
+    },
+    {
+      "epoch": 4.329351535836177,
+      "grad_norm": 0.31297368541999177,
+      "learning_rate": 2.1523678261621715e-06,
+      "loss": 0.1593,
+      "step": 2537
+    },
+    {
+      "epoch": 4.3310580204778155,
+      "grad_norm": 0.32119774641052234,
+      "learning_rate": 2.1416277989697344e-06,
+      "loss": 0.157,
+      "step": 2538
+    },
+    {
+      "epoch": 4.332764505119454,
+      "grad_norm": 0.2886738863134529,
+      "learning_rate": 2.130913118470297e-06,
+      "loss": 0.2005,
+      "step": 2539
+    },
+    {
+      "epoch": 4.334470989761092,
+      "grad_norm": 0.2872363431291567,
+      "learning_rate": 2.1202237998713814e-06,
+      "loss": 0.1568,
+      "step": 2540
+    },
+    {
+      "epoch": 4.336177474402731,
+      "grad_norm": 0.2913090673455017,
+      "learning_rate": 2.1095598583445255e-06,
+      "loss": 0.1927,
+      "step": 2541
+    },
+    {
+      "epoch": 4.337883959044369,
+      "grad_norm": 0.3110281755536224,
+      "learning_rate": 2.09892130902525e-06,
+      "loss": 0.1517,
+      "step": 2542
+    },
+    {
+      "epoch": 4.339590443686006,
+      "grad_norm": 0.2896544715959582,
+      "learning_rate": 2.0883081670130202e-06,
+      "loss": 0.1644,
+      "step": 2543
+    },
+    {
+      "epoch": 4.341296928327645,
+      "grad_norm": 0.29564797273531257,
+      "learning_rate": 2.0777204473712564e-06,
+      "loss": 0.1809,
+      "step": 2544
+    },
+    {
+      "epoch": 4.343003412969283,
+      "grad_norm": 0.30223501178504436,
+      "learning_rate": 2.06715816512729e-06,
+      "loss": 0.1618,
+      "step": 2545
+    },
+    {
+      "epoch": 4.3447098976109215,
+      "grad_norm": 0.2796922695719279,
+      "learning_rate": 2.056621335272344e-06,
+      "loss": 0.1909,
+      "step": 2546
+    },
+    {
+      "epoch": 4.34641638225256,
+      "grad_norm": 0.3059516004912932,
+      "learning_rate": 2.046109972761523e-06,
+      "loss": 0.2023,
+      "step": 2547
+    },
+    {
+      "epoch": 4.348122866894198,
+      "grad_norm": 0.2909794324825243,
+      "learning_rate": 2.0356240925137816e-06,
+      "loss": 0.14,
+      "step": 2548
+    },
+    {
+      "epoch": 4.349829351535837,
+      "grad_norm": 0.30160162314385436,
+      "learning_rate": 2.025163709411917e-06,
+      "loss": 0.2013,
+      "step": 2549
+    },
+    {
+      "epoch": 4.351535836177474,
+      "grad_norm": 0.2830562144601256,
+      "learning_rate": 2.0147288383025197e-06,
+      "loss": 0.1638,
+      "step": 2550
+    },
+    {
+      "epoch": 4.353242320819112,
+      "grad_norm": 0.3332359895409523,
+      "learning_rate": 2.004319493995992e-06,
+      "loss": 0.1801,
+      "step": 2551
+    },
+    {
+      "epoch": 4.354948805460751,
+      "grad_norm": 0.3047315291970305,
+      "learning_rate": 1.993935691266482e-06,
+      "loss": 0.1702,
+      "step": 2552
+    },
+    {
+      "epoch": 4.356655290102389,
+      "grad_norm": 0.3091788010415166,
+      "learning_rate": 1.9835774448519075e-06,
+      "loss": 0.1949,
+      "step": 2553
+    },
+    {
+      "epoch": 4.3583617747440275,
+      "grad_norm": 0.29617004681223497,
+      "learning_rate": 1.973244769453897e-06,
+      "loss": 0.1979,
+      "step": 2554
+    },
+    {
+      "epoch": 4.360068259385666,
+      "grad_norm": 0.29985425485567585,
+      "learning_rate": 1.9629376797378e-06,
+      "loss": 0.2125,
+      "step": 2555
+    },
+    {
+      "epoch": 4.361774744027303,
+      "grad_norm": 0.29373314664152933,
+      "learning_rate": 1.95265619033264e-06,
+      "loss": 0.1832,
+      "step": 2556
+    },
+    {
+      "epoch": 4.363481228668942,
+      "grad_norm": 0.30556795498559286,
+      "learning_rate": 1.9424003158311187e-06,
+      "loss": 0.1862,
+      "step": 2557
+    },
+    {
+      "epoch": 4.36518771331058,
+      "grad_norm": 0.3373225031070908,
+      "learning_rate": 1.9321700707895672e-06,
+      "loss": 0.1708,
+      "step": 2558
+    },
+    {
+      "epoch": 4.3668941979522184,
+      "grad_norm": 0.31772294187974187,
+      "learning_rate": 1.9219654697279443e-06,
+      "loss": 0.2783,
+      "step": 2559
+    },
+    {
+      "epoch": 4.368600682593857,
+      "grad_norm": 0.29033686690165184,
+      "learning_rate": 1.9117865271298264e-06,
+      "loss": 0.1807,
+      "step": 2560
+    },
+    {
+      "epoch": 4.370307167235495,
+      "grad_norm": 0.2997392442069027,
+      "learning_rate": 1.9016332574423479e-06,
+      "loss": 0.1626,
+      "step": 2561
+    },
+    {
+      "epoch": 4.372013651877133,
+      "grad_norm": 0.30622537739214434,
+      "learning_rate": 1.8915056750762261e-06,
+      "loss": 0.1916,
+      "step": 2562
+    },
+    {
+      "epoch": 4.373720136518771,
+      "grad_norm": 0.31749443847272735,
+      "learning_rate": 1.8814037944057117e-06,
+      "loss": 0.1934,
+      "step": 2563
+    },
+    {
+      "epoch": 4.375426621160409,
+      "grad_norm": 0.3081795435254109,
+      "learning_rate": 1.8713276297685712e-06,
+      "loss": 0.1489,
+      "step": 2564
+    },
+    {
+      "epoch": 4.377133105802048,
+      "grad_norm": 0.3030137285901555,
+      "learning_rate": 1.8612771954660825e-06,
+      "loss": 0.1535,
+      "step": 2565
+    },
+    {
+      "epoch": 4.378839590443686,
+      "grad_norm": 0.31679080381563246,
+      "learning_rate": 1.851252505762995e-06,
+      "loss": 0.1608,
+      "step": 2566
+    },
+    {
+      "epoch": 4.3805460750853245,
+      "grad_norm": 0.3178725577997585,
+      "learning_rate": 1.841253574887527e-06,
+      "loss": 0.2116,
+      "step": 2567
+    },
+    {
+      "epoch": 4.382252559726963,
+      "grad_norm": 0.3120407911523594,
+      "learning_rate": 1.831280417031327e-06,
+      "loss": 0.1739,
+      "step": 2568
+    },
+    {
+      "epoch": 4.3839590443686,
+      "grad_norm": 0.3156116274518043,
+      "learning_rate": 1.8213330463494738e-06,
+      "loss": 0.134,
+      "step": 2569
+    },
+    {
+      "epoch": 4.385665529010239,
+      "grad_norm": 0.3129024986980865,
+      "learning_rate": 1.8114114769604363e-06,
+      "loss": 0.1667,
+      "step": 2570
+    },
+    {
+      "epoch": 4.387372013651877,
+      "grad_norm": 0.3251618265488506,
+      "learning_rate": 1.8015157229460656e-06,
+      "loss": 0.1775,
+      "step": 2571
+    },
+    {
+      "epoch": 4.389078498293515,
+      "grad_norm": 0.30972584219226884,
+      "learning_rate": 1.7916457983515822e-06,
+      "loss": 0.1906,
+      "step": 2572
+    },
+    {
+      "epoch": 4.390784982935154,
+      "grad_norm": 0.3167115292804992,
+      "learning_rate": 1.7818017171855318e-06,
+      "loss": 0.1734,
+      "step": 2573
+    },
+    {
+      "epoch": 4.392491467576792,
+      "grad_norm": 0.3096780871319159,
+      "learning_rate": 1.771983493419791e-06,
+      "loss": 0.1772,
+      "step": 2574
+    },
+    {
+      "epoch": 4.3941979522184305,
+      "grad_norm": 0.29489330937597885,
+      "learning_rate": 1.7621911409895332e-06,
+      "loss": 0.1867,
+      "step": 2575
+    },
+    {
+      "epoch": 4.395904436860068,
+      "grad_norm": 0.3133041098716859,
+      "learning_rate": 1.7524246737932072e-06,
+      "loss": 0.1606,
+      "step": 2576
+    },
+    {
+      "epoch": 4.397610921501706,
+      "grad_norm": 0.3117573213645619,
+      "learning_rate": 1.7426841056925315e-06,
+      "loss": 0.1984,
+      "step": 2577
+    },
+    {
+      "epoch": 4.399317406143345,
+      "grad_norm": 0.29972628372325183,
+      "learning_rate": 1.732969450512456e-06,
+      "loss": 0.2101,
+      "step": 2578
+    },
+    {
+      "epoch": 4.401023890784983,
+      "grad_norm": 0.3092356765885778,
+      "learning_rate": 1.7232807220411629e-06,
+      "loss": 0.1716,
+      "step": 2579
+    },
+    {
+      "epoch": 4.402730375426621,
+      "grad_norm": 0.3062746650892749,
+      "learning_rate": 1.713617934030023e-06,
+      "loss": 0.14,
+      "step": 2580
+    },
+    {
+      "epoch": 4.40443686006826,
+      "grad_norm": 0.2734822135973346,
+      "learning_rate": 1.7039811001936056e-06,
+      "loss": 0.2018,
+      "step": 2581
+    },
+    {
+      "epoch": 4.406143344709897,
+      "grad_norm": 0.3173603903458222,
+      "learning_rate": 1.694370234209628e-06,
+      "loss": 0.1619,
+      "step": 2582
+    },
+    {
+      "epoch": 4.407849829351536,
+      "grad_norm": 0.320973709722284,
+      "learning_rate": 1.6847853497189538e-06,
+      "loss": 0.2777,
+      "step": 2583
+    },
+    {
+      "epoch": 4.409556313993174,
+      "grad_norm": 0.2884858345193101,
+      "learning_rate": 1.675226460325583e-06,
+      "loss": 0.1649,
+      "step": 2584
+    },
+    {
+      "epoch": 4.411262798634812,
+      "grad_norm": 0.3112597061860702,
+      "learning_rate": 1.6656935795965989e-06,
+      "loss": 0.1983,
+      "step": 2585
+    },
+    {
+      "epoch": 4.412969283276451,
+      "grad_norm": 0.3059564652787005,
+      "learning_rate": 1.6561867210621918e-06,
+      "loss": 0.1928,
+      "step": 2586
+    },
+    {
+      "epoch": 4.414675767918089,
+      "grad_norm": 0.2982886465732699,
+      "learning_rate": 1.6467058982156015e-06,
+      "loss": 0.1839,
+      "step": 2587
+    },
+    {
+      "epoch": 4.4163822525597265,
+      "grad_norm": 0.3149448537193366,
+      "learning_rate": 1.6372511245131285e-06,
+      "loss": 0.1489,
+      "step": 2588
+    },
+    {
+      "epoch": 4.418088737201365,
+      "grad_norm": 0.31512048989873015,
+      "learning_rate": 1.6278224133740917e-06,
+      "loss": 0.193,
+      "step": 2589
+    },
+    {
+      "epoch": 4.419795221843003,
+      "grad_norm": 0.3286822254843685,
+      "learning_rate": 1.6184197781808197e-06,
+      "loss": 0.1561,
+      "step": 2590
+    },
+    {
+      "epoch": 4.421501706484642,
+      "grad_norm": 0.35473759580754277,
+      "learning_rate": 1.6090432322786375e-06,
+      "loss": 0.1887,
+      "step": 2591
+    },
+    {
+      "epoch": 4.42320819112628,
+      "grad_norm": 0.3042147572286529,
+      "learning_rate": 1.5996927889758307e-06,
+      "loss": 0.1905,
+      "step": 2592
+    },
+    {
+      "epoch": 4.424914675767918,
+      "grad_norm": 0.2966243777562204,
+      "learning_rate": 1.5903684615436542e-06,
+      "loss": 0.191,
+      "step": 2593
+    },
+    {
+      "epoch": 4.426621160409557,
+      "grad_norm": 0.31376177368311275,
+      "learning_rate": 1.5810702632162755e-06,
+      "loss": 0.1643,
+      "step": 2594
+    },
+    {
+      "epoch": 4.428327645051194,
+      "grad_norm": 0.31289484267866075,
+      "learning_rate": 1.571798207190789e-06,
+      "loss": 0.1781,
+      "step": 2595
+    },
+    {
+      "epoch": 4.4300341296928325,
+      "grad_norm": 0.29694775918388083,
+      "learning_rate": 1.5625523066271852e-06,
+      "loss": 0.1659,
+      "step": 2596
+    },
+    {
+      "epoch": 4.431740614334471,
+      "grad_norm": 0.29147605821295364,
+      "learning_rate": 1.553332574648323e-06,
+      "loss": 0.1593,
+      "step": 2597
+    },
+    {
+      "epoch": 4.433447098976109,
+      "grad_norm": 0.3057342409234324,
+      "learning_rate": 1.5441390243399345e-06,
+      "loss": 0.1639,
+      "step": 2598
+    },
+    {
+      "epoch": 4.435153583617748,
+      "grad_norm": 0.28642803934421673,
+      "learning_rate": 1.5349716687505733e-06,
+      "loss": 0.1626,
+      "step": 2599
+    },
+    {
+      "epoch": 4.436860068259386,
+      "grad_norm": 0.285548146724235,
+      "learning_rate": 1.5258305208916314e-06,
+      "loss": 0.1606,
+      "step": 2600
+    },
+    {
+      "epoch": 4.438566552901024,
+      "grad_norm": 0.32479682544844474,
+      "learning_rate": 1.5167155937372947e-06,
+      "loss": 0.1888,
+      "step": 2601
+    },
+    {
+      "epoch": 4.440273037542662,
+      "grad_norm": 0.3378938837647975,
+      "learning_rate": 1.5076269002245304e-06,
+      "loss": 0.1761,
+      "step": 2602
+    },
+    {
+      "epoch": 4.4419795221843,
+      "grad_norm": 0.29406065899669787,
+      "learning_rate": 1.4985644532530819e-06,
+      "loss": 0.2741,
+      "step": 2603
+    },
+    {
+      "epoch": 4.4436860068259385,
+      "grad_norm": 0.291056615154014,
+      "learning_rate": 1.4895282656854293e-06,
+      "loss": 0.1709,
+      "step": 2604
+    },
+    {
+      "epoch": 4.445392491467577,
+      "grad_norm": 0.3077417839945242,
+      "learning_rate": 1.4805183503467979e-06,
+      "loss": 0.201,
+      "step": 2605
+    },
+    {
+      "epoch": 4.447098976109215,
+      "grad_norm": 0.31781018398972843,
+      "learning_rate": 1.4715347200251052e-06,
+      "loss": 0.1696,
+      "step": 2606
+    },
+    {
+      "epoch": 4.448805460750854,
+      "grad_norm": 0.3163121595251067,
+      "learning_rate": 1.462577387470978e-06,
+      "loss": 0.1956,
+      "step": 2607
+    },
+    {
+      "epoch": 4.450511945392491,
+      "grad_norm": 0.29257827516208407,
+      "learning_rate": 1.4536463653977028e-06,
+      "loss": 0.1598,
+      "step": 2608
+    },
+    {
+      "epoch": 4.452218430034129,
+      "grad_norm": 0.31633914007976754,
+      "learning_rate": 1.4447416664812374e-06,
+      "loss": 0.1907,
+      "step": 2609
+    },
+    {
+      "epoch": 4.453924914675768,
+      "grad_norm": 0.33350557899838706,
+      "learning_rate": 1.4358633033601788e-06,
+      "loss": 0.1561,
+      "step": 2610
+    },
+    {
+      "epoch": 4.455631399317406,
+      "grad_norm": 0.3167449232019445,
+      "learning_rate": 1.427011288635729e-06,
+      "loss": 0.1559,
+      "step": 2611
+    },
+    {
+      "epoch": 4.4573378839590445,
+      "grad_norm": 0.30530600953842574,
+      "learning_rate": 1.418185634871716e-06,
+      "loss": 0.1641,
+      "step": 2612
+    },
+    {
+      "epoch": 4.459044368600683,
+      "grad_norm": 0.3131984591858436,
+      "learning_rate": 1.4093863545945263e-06,
+      "loss": 0.177,
+      "step": 2613
+    },
+    {
+      "epoch": 4.460750853242321,
+      "grad_norm": 0.3380908183071507,
+      "learning_rate": 1.4006134602931408e-06,
+      "loss": 0.1825,
+      "step": 2614
+    },
+    {
+      "epoch": 4.462457337883959,
+      "grad_norm": 0.28330310082518084,
+      "learning_rate": 1.3918669644190708e-06,
+      "loss": 0.1774,
+      "step": 2615
+    },
+    {
+      "epoch": 4.464163822525597,
+      "grad_norm": 0.31826886606580584,
+      "learning_rate": 1.3831468793863701e-06,
+      "loss": 0.1836,
+      "step": 2616
+    },
+    {
+      "epoch": 4.465870307167235,
+      "grad_norm": 0.28804015352123924,
+      "learning_rate": 1.3744532175716098e-06,
+      "loss": 0.241,
+      "step": 2617
+    },
+    {
+      "epoch": 4.467576791808874,
+      "grad_norm": 0.3430602007229569,
+      "learning_rate": 1.3657859913138437e-06,
+      "loss": 0.2021,
+      "step": 2618
+    },
+    {
+      "epoch": 4.469283276450512,
+      "grad_norm": 0.30126030775145457,
+      "learning_rate": 1.3571452129146234e-06,
+      "loss": 0.1724,
+      "step": 2619
+    },
+    {
+      "epoch": 4.4709897610921505,
+      "grad_norm": 0.31609525596538085,
+      "learning_rate": 1.348530894637945e-06,
+      "loss": 0.1823,
+      "step": 2620
+    },
+    {
+      "epoch": 4.472696245733788,
+      "grad_norm": 0.2911441143738094,
+      "learning_rate": 1.3399430487102638e-06,
+      "loss": 0.1721,
+      "step": 2621
+    },
+    {
+      "epoch": 4.474402730375426,
+      "grad_norm": 0.29289498103999745,
+      "learning_rate": 1.33138168732045e-06,
+      "loss": 0.2007,
+      "step": 2622
+    },
+    {
+      "epoch": 4.476109215017065,
+      "grad_norm": 0.3081640723691369,
+      "learning_rate": 1.3228468226197944e-06,
+      "loss": 0.1615,
+      "step": 2623
+    },
+    {
+      "epoch": 4.477815699658703,
+      "grad_norm": 0.2785466270617825,
+      "learning_rate": 1.3143384667219783e-06,
+      "loss": 0.1776,
+      "step": 2624
+    },
+    {
+      "epoch": 4.479522184300341,
+      "grad_norm": 0.2884258060547518,
+      "learning_rate": 1.3058566317030551e-06,
+      "loss": 0.1906,
+      "step": 2625
+    },
+    {
+      "epoch": 4.48122866894198,
+      "grad_norm": 0.31444204672002174,
+      "learning_rate": 1.2974013296014376e-06,
+      "loss": 0.1669,
+      "step": 2626
+    },
+    {
+      "epoch": 4.482935153583618,
+      "grad_norm": 0.3195633943847116,
+      "learning_rate": 1.288972572417877e-06,
+      "loss": 0.2045,
+      "step": 2627
+    },
+    {
+      "epoch": 4.484641638225256,
+      "grad_norm": 0.31559512088158087,
+      "learning_rate": 1.2805703721154594e-06,
+      "loss": 0.1702,
+      "step": 2628
+    },
+    {
+      "epoch": 4.486348122866894,
+      "grad_norm": 0.3166673710140947,
+      "learning_rate": 1.2721947406195657e-06,
+      "loss": 0.1788,
+      "step": 2629
+    },
+    {
+      "epoch": 4.488054607508532,
+      "grad_norm": 0.3060831513316056,
+      "learning_rate": 1.2638456898178752e-06,
+      "loss": 0.1873,
+      "step": 2630
+    },
+    {
+      "epoch": 4.489761092150171,
+      "grad_norm": 0.31154060582160964,
+      "learning_rate": 1.2555232315603449e-06,
+      "loss": 0.2374,
+      "step": 2631
+    },
+    {
+      "epoch": 4.491467576791809,
+      "grad_norm": 0.3076884170379762,
+      "learning_rate": 1.247227377659168e-06,
+      "loss": 0.1839,
+      "step": 2632
+    },
+    {
+      "epoch": 4.493174061433447,
+      "grad_norm": 0.2985982338062527,
+      "learning_rate": 1.238958139888804e-06,
+      "loss": 0.175,
+      "step": 2633
+    },
+    {
+      "epoch": 4.494880546075085,
+      "grad_norm": 0.31920058660397355,
+      "learning_rate": 1.2307155299859153e-06,
+      "loss": 0.2152,
+      "step": 2634
+    },
+    {
+      "epoch": 4.496587030716723,
+      "grad_norm": 0.34402587537787804,
+      "learning_rate": 1.222499559649386e-06,
+      "loss": 0.1808,
+      "step": 2635
+    },
+    {
+      "epoch": 4.498293515358362,
+      "grad_norm": 0.317583860848277,
+      "learning_rate": 1.2143102405402751e-06,
+      "loss": 0.2015,
+      "step": 2636
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.3282545929788059,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.1948,
+      "step": 2637
+    },
+    {
+      "epoch": 4.501706484641638,
+      "grad_norm": 0.2838840587396505,
+      "learning_rate": 1.1980116024594524e-06,
+      "loss": 0.1838,
+      "step": 2638
+    },
+    {
+      "epoch": 4.503412969283277,
+      "grad_norm": 0.2987084175711107,
+      "learning_rate": 1.1899023066206671e-06,
+      "loss": 0.1957,
+      "step": 2639
+    },
+    {
+      "epoch": 4.505119453924914,
+      "grad_norm": 0.3013170240642986,
+      "learning_rate": 1.1818197082751493e-06,
+      "loss": 0.1787,
+      "step": 2640
+    },
+    {
+      "epoch": 4.506825938566553,
+      "grad_norm": 0.31558048418216567,
+      "learning_rate": 1.1737638188946577e-06,
+      "loss": 0.1733,
+      "step": 2641
+    },
+    {
+      "epoch": 4.508532423208191,
+      "grad_norm": 0.2906170661542559,
+      "learning_rate": 1.165734649913064e-06,
+      "loss": 0.1824,
+      "step": 2642
+    },
+    {
+      "epoch": 4.510238907849829,
+      "grad_norm": 0.30098580797968205,
+      "learning_rate": 1.157732212726299e-06,
+      "loss": 0.1826,
+      "step": 2643
+    },
+    {
+      "epoch": 4.511945392491468,
+      "grad_norm": 0.31994314706625765,
+      "learning_rate": 1.1497565186923575e-06,
+      "loss": 0.2408,
+      "step": 2644
+    },
+    {
+      "epoch": 4.513651877133106,
+      "grad_norm": 0.3106424840527709,
+      "learning_rate": 1.1418075791312843e-06,
+      "loss": 0.1755,
+      "step": 2645
+    },
+    {
+      "epoch": 4.515358361774744,
+      "grad_norm": 0.3027234786599201,
+      "learning_rate": 1.133885405325139e-06,
+      "loss": 0.2042,
+      "step": 2646
+    },
+    {
+      "epoch": 4.517064846416382,
+      "grad_norm": 0.2923515699191684,
+      "learning_rate": 1.1259900085180054e-06,
+      "loss": 0.1585,
+      "step": 2647
+    },
+    {
+      "epoch": 4.51877133105802,
+      "grad_norm": 0.2817260535186955,
+      "learning_rate": 1.1181213999159458e-06,
+      "loss": 0.1941,
+      "step": 2648
+    },
+    {
+      "epoch": 4.520477815699659,
+      "grad_norm": 0.31735840851041724,
+      "learning_rate": 1.1102795906870223e-06,
+      "loss": 0.1669,
+      "step": 2649
+    },
+    {
+      "epoch": 4.522184300341297,
+      "grad_norm": 0.3055142775670233,
+      "learning_rate": 1.1024645919612386e-06,
+      "loss": 0.1757,
+      "step": 2650
+    },
+    {
+      "epoch": 4.523890784982935,
+      "grad_norm": 0.3314596012225001,
+      "learning_rate": 1.09467641483056e-06,
+      "loss": 0.149,
+      "step": 2651
+    },
+    {
+      "epoch": 4.525597269624574,
+      "grad_norm": 0.2962199127617063,
+      "learning_rate": 1.0869150703488818e-06,
+      "loss": 0.1703,
+      "step": 2652
+    },
+    {
+      "epoch": 4.527303754266212,
+      "grad_norm": 0.3144511395605482,
+      "learning_rate": 1.079180569532008e-06,
+      "loss": 0.2131,
+      "step": 2653
+    },
+    {
+      "epoch": 4.5290102389078495,
+      "grad_norm": 0.3059647228514497,
+      "learning_rate": 1.0714729233576526e-06,
+      "loss": 0.1623,
+      "step": 2654
+    },
+    {
+      "epoch": 4.530716723549488,
+      "grad_norm": 0.3198382002725939,
+      "learning_rate": 1.0637921427654052e-06,
+      "loss": 0.1648,
+      "step": 2655
+    },
+    {
+      "epoch": 4.532423208191126,
+      "grad_norm": 0.313476882585119,
+      "learning_rate": 1.0561382386567342e-06,
+      "loss": 0.1728,
+      "step": 2656
+    },
+    {
+      "epoch": 4.534129692832765,
+      "grad_norm": 0.32461608946860504,
+      "learning_rate": 1.0485112218949544e-06,
+      "loss": 0.1991,
+      "step": 2657
+    },
+    {
+      "epoch": 4.535836177474403,
+      "grad_norm": 0.3067653776366247,
+      "learning_rate": 1.0409111033052154e-06,
+      "loss": 0.1965,
+      "step": 2658
+    },
+    {
+      "epoch": 4.537542662116041,
+      "grad_norm": 0.3210134260484315,
+      "learning_rate": 1.0333378936745064e-06,
+      "loss": 0.1726,
+      "step": 2659
+    },
+    {
+      "epoch": 4.53924914675768,
+      "grad_norm": 0.2937600173082321,
+      "learning_rate": 1.0257916037516025e-06,
+      "loss": 0.1459,
+      "step": 2660
+    },
+    {
+      "epoch": 4.540955631399317,
+      "grad_norm": 0.28590331785936346,
+      "learning_rate": 1.0182722442470894e-06,
+      "loss": 0.1525,
+      "step": 2661
+    },
+    {
+      "epoch": 4.5426621160409555,
+      "grad_norm": 0.2849589712836118,
+      "learning_rate": 1.0107798258333213e-06,
+      "loss": 0.1642,
+      "step": 2662
+    },
+    {
+      "epoch": 4.544368600682594,
+      "grad_norm": 0.2850599390185159,
+      "learning_rate": 1.0033143591444116e-06,
+      "loss": 0.1999,
+      "step": 2663
+    },
+    {
+      "epoch": 4.546075085324232,
+      "grad_norm": 0.3246837318250444,
+      "learning_rate": 9.958758547762292e-07,
+      "loss": 0.1728,
+      "step": 2664
+    },
+    {
+      "epoch": 4.547781569965871,
+      "grad_norm": 0.31034182132595006,
+      "learning_rate": 9.884643232863666e-07,
+      "loss": 0.1603,
+      "step": 2665
+    },
+    {
+      "epoch": 4.549488054607508,
+      "grad_norm": 0.2699307113298506,
+      "learning_rate": 9.810797751941448e-07,
+      "loss": 0.173,
+      "step": 2666
+    },
+    {
+      "epoch": 4.551194539249146,
+      "grad_norm": 0.3164873422498817,
+      "learning_rate": 9.737222209805686e-07,
+      "loss": 0.1479,
+      "step": 2667
+    },
+    {
+      "epoch": 4.552901023890785,
+      "grad_norm": 0.3101121665742394,
+      "learning_rate": 9.663916710883493e-07,
+      "loss": 0.1798,
+      "step": 2668
+    },
+    {
+      "epoch": 4.554607508532423,
+      "grad_norm": 0.31326479172343186,
+      "learning_rate": 9.590881359218595e-07,
+      "loss": 0.1951,
+      "step": 2669
+    },
+    {
+      "epoch": 4.5563139931740615,
+      "grad_norm": 0.29438538514221124,
+      "learning_rate": 9.518116258471254e-07,
+      "loss": 0.1652,
+      "step": 2670
+    },
+    {
+      "epoch": 4.5580204778157,
+      "grad_norm": 0.2991309107110425,
+      "learning_rate": 9.445621511918324e-07,
+      "loss": 0.1897,
+      "step": 2671
+    },
+    {
+      "epoch": 4.559726962457338,
+      "grad_norm": 0.2955512942850018,
+      "learning_rate": 9.373397222452741e-07,
+      "loss": 0.1792,
+      "step": 2672
+    },
+    {
+      "epoch": 4.561433447098976,
+      "grad_norm": 0.31376758353537465,
+      "learning_rate": 9.301443492583751e-07,
+      "loss": 0.1648,
+      "step": 2673
+    },
+    {
+      "epoch": 4.563139931740614,
+      "grad_norm": 0.3204276037498276,
+      "learning_rate": 9.229760424436462e-07,
+      "loss": 0.1761,
+      "step": 2674
+    },
+    {
+      "epoch": 4.564846416382252,
+      "grad_norm": 0.3012058029999409,
+      "learning_rate": 9.158348119751892e-07,
+      "loss": 0.1626,
+      "step": 2675
+    },
+    {
+      "epoch": 4.566552901023891,
+      "grad_norm": 0.287669382311409,
+      "learning_rate": 9.087206679886762e-07,
+      "loss": 0.1712,
+      "step": 2676
+    },
+    {
+      "epoch": 4.568259385665529,
+      "grad_norm": 0.2784332470218125,
+      "learning_rate": 9.016336205813303e-07,
+      "loss": 0.1856,
+      "step": 2677
+    },
+    {
+      "epoch": 4.5699658703071675,
+      "grad_norm": 0.3157487210825833,
+      "learning_rate": 8.945736798119253e-07,
+      "loss": 0.1801,
+      "step": 2678
+    },
+    {
+      "epoch": 4.571672354948806,
+      "grad_norm": 0.33192547669048056,
+      "learning_rate": 8.875408557007459e-07,
+      "loss": 0.1917,
+      "step": 2679
+    },
+    {
+      "epoch": 4.573378839590443,
+      "grad_norm": 0.29465625247378635,
+      "learning_rate": 8.805351582296118e-07,
+      "loss": 0.181,
+      "step": 2680
+    },
+    {
+      "epoch": 4.575085324232082,
+      "grad_norm": 0.2975756901413667,
+      "learning_rate": 8.735565973418181e-07,
+      "loss": 0.1888,
+      "step": 2681
+    },
+    {
+      "epoch": 4.57679180887372,
+      "grad_norm": 0.30854480493941716,
+      "learning_rate": 8.666051829421596e-07,
+      "loss": 0.159,
+      "step": 2682
+    },
+    {
+      "epoch": 4.578498293515358,
+      "grad_norm": 0.3145222152129316,
+      "learning_rate": 8.596809248968996e-07,
+      "loss": 0.169,
+      "step": 2683
+    },
+    {
+      "epoch": 4.580204778156997,
+      "grad_norm": 0.3063202595906734,
+      "learning_rate": 8.527838330337524e-07,
+      "loss": 0.1644,
+      "step": 2684
+    },
+    {
+      "epoch": 4.581911262798635,
+      "grad_norm": 0.31189663379964455,
+      "learning_rate": 8.459139171418851e-07,
+      "loss": 0.1621,
+      "step": 2685
+    },
+    {
+      "epoch": 4.5836177474402735,
+      "grad_norm": 0.2904213106630353,
+      "learning_rate": 8.390711869718782e-07,
+      "loss": 0.1588,
+      "step": 2686
+    },
+    {
+      "epoch": 4.585324232081911,
+      "grad_norm": 0.3179560834185949,
+      "learning_rate": 8.322556522357427e-07,
+      "loss": 0.1677,
+      "step": 2687
+    },
+    {
+      "epoch": 4.587030716723549,
+      "grad_norm": 0.31293041696083573,
+      "learning_rate": 8.254673226068788e-07,
+      "loss": 0.1641,
+      "step": 2688
+    },
+    {
+      "epoch": 4.588737201365188,
+      "grad_norm": 0.33577700473897765,
+      "learning_rate": 8.18706207720077e-07,
+      "loss": 0.1561,
+      "step": 2689
+    },
+    {
+      "epoch": 4.590443686006826,
+      "grad_norm": 0.3157403650864591,
+      "learning_rate": 8.119723171715122e-07,
+      "loss": 0.1641,
+      "step": 2690
+    },
+    {
+      "epoch": 4.592150170648464,
+      "grad_norm": 0.2862777973290719,
+      "learning_rate": 8.052656605187015e-07,
+      "loss": 0.1494,
+      "step": 2691
+    },
+    {
+      "epoch": 4.593856655290102,
+      "grad_norm": 0.30784437239644985,
+      "learning_rate": 7.985862472805217e-07,
+      "loss": 0.1907,
+      "step": 2692
+    },
+    {
+      "epoch": 4.59556313993174,
+      "grad_norm": 0.28642816392311105,
+      "learning_rate": 7.919340869371783e-07,
+      "loss": 0.1514,
+      "step": 2693
+    },
+    {
+      "epoch": 4.597269624573379,
+      "grad_norm": 0.29941313106236994,
+      "learning_rate": 7.853091889301944e-07,
+      "loss": 0.168,
+      "step": 2694
+    },
+    {
+      "epoch": 4.598976109215017,
+      "grad_norm": 0.2916049076030227,
+      "learning_rate": 7.78711562662402e-07,
+      "loss": 0.1596,
+      "step": 2695
+    },
+    {
+      "epoch": 4.600682593856655,
+      "grad_norm": 0.2919761829480753,
+      "learning_rate": 7.721412174979214e-07,
+      "loss": 0.1494,
+      "step": 2696
+    },
+    {
+      "epoch": 4.602389078498294,
+      "grad_norm": 0.2860719533668131,
+      "learning_rate": 7.655981627621645e-07,
+      "loss": 0.1636,
+      "step": 2697
+    },
+    {
+      "epoch": 4.604095563139932,
+      "grad_norm": 0.3015545961815166,
+      "learning_rate": 7.590824077417913e-07,
+      "loss": 0.2241,
+      "step": 2698
+    },
+    {
+      "epoch": 4.6058020477815695,
+      "grad_norm": 0.2890766624921712,
+      "learning_rate": 7.525939616847333e-07,
+      "loss": 0.136,
+      "step": 2699
+    },
+    {
+      "epoch": 4.607508532423208,
+      "grad_norm": 0.3042706431257455,
+      "learning_rate": 7.461328338001417e-07,
+      "loss": 0.146,
+      "step": 2700
+    },
+    {
+      "epoch": 4.609215017064846,
+      "grad_norm": 0.2947006269527326,
+      "learning_rate": 7.396990332584164e-07,
+      "loss": 0.183,
+      "step": 2701
+    },
+    {
+      "epoch": 4.610921501706485,
+      "grad_norm": 0.31901339926876754,
+      "learning_rate": 7.33292569191153e-07,
+      "loss": 0.1826,
+      "step": 2702
+    },
+    {
+      "epoch": 4.612627986348123,
+      "grad_norm": 0.31810941062398185,
+      "learning_rate": 7.269134506911579e-07,
+      "loss": 0.1568,
+      "step": 2703
+    },
+    {
+      "epoch": 4.614334470989761,
+      "grad_norm": 0.30449542970454335,
+      "learning_rate": 7.205616868124288e-07,
+      "loss": 0.1724,
+      "step": 2704
+    },
+    {
+      "epoch": 4.6160409556314,
+      "grad_norm": 0.2955978987021187,
+      "learning_rate": 7.142372865701253e-07,
+      "loss": 0.1761,
+      "step": 2705
+    },
+    {
+      "epoch": 4.617747440273037,
+      "grad_norm": 0.291363412093148,
+      "learning_rate": 7.079402589405804e-07,
+      "loss": 0.1659,
+      "step": 2706
+    },
+    {
+      "epoch": 4.6194539249146755,
+      "grad_norm": 0.28902122490739873,
+      "learning_rate": 7.016706128612694e-07,
+      "loss": 0.2135,
+      "step": 2707
+    },
+    {
+      "epoch": 4.621160409556314,
+      "grad_norm": 0.2915347981940905,
+      "learning_rate": 6.954283572308118e-07,
+      "loss": 0.1962,
+      "step": 2708
+    },
+    {
+      "epoch": 4.622866894197952,
+      "grad_norm": 0.32533380821156654,
+      "learning_rate": 6.892135009089451e-07,
+      "loss": 0.1733,
+      "step": 2709
+    },
+    {
+      "epoch": 4.624573378839591,
+      "grad_norm": 0.3271404140899551,
+      "learning_rate": 6.830260527165222e-07,
+      "loss": 0.196,
+      "step": 2710
+    },
+    {
+      "epoch": 4.626279863481229,
+      "grad_norm": 0.31781236648785877,
+      "learning_rate": 6.768660214355005e-07,
+      "loss": 0.1554,
+      "step": 2711
+    },
+    {
+      "epoch": 4.627986348122867,
+      "grad_norm": 0.3217830186925165,
+      "learning_rate": 6.707334158089063e-07,
+      "loss": 0.159,
+      "step": 2712
+    },
+    {
+      "epoch": 4.629692832764505,
+      "grad_norm": 0.31911002436978775,
+      "learning_rate": 6.646282445408591e-07,
+      "loss": 0.2502,
+      "step": 2713
+    },
+    {
+      "epoch": 4.631399317406143,
+      "grad_norm": 0.3143470280291693,
+      "learning_rate": 6.5855051629653e-07,
+      "loss": 0.1961,
+      "step": 2714
+    },
+    {
+      "epoch": 4.6331058020477816,
+      "grad_norm": 0.30478601257389953,
+      "learning_rate": 6.525002397021451e-07,
+      "loss": 0.1847,
+      "step": 2715
+    },
+    {
+      "epoch": 4.63481228668942,
+      "grad_norm": 0.27885189050541664,
+      "learning_rate": 6.464774233449622e-07,
+      "loss": 0.1727,
+      "step": 2716
+    },
+    {
+      "epoch": 4.636518771331058,
+      "grad_norm": 0.2767544904678458,
+      "learning_rate": 6.4048207577327e-07,
+      "loss": 0.1964,
+      "step": 2717
+    },
+    {
+      "epoch": 4.638225255972696,
+      "grad_norm": 0.3440955133555953,
+      "learning_rate": 6.345142054963682e-07,
+      "loss": 0.1421,
+      "step": 2718
+    },
+    {
+      "epoch": 4.639931740614334,
+      "grad_norm": 0.3081120729111336,
+      "learning_rate": 6.285738209845527e-07,
+      "loss": 0.2058,
+      "step": 2719
+    },
+    {
+      "epoch": 4.6416382252559725,
+      "grad_norm": 0.29037800407941694,
+      "learning_rate": 6.226609306691189e-07,
+      "loss": 0.1725,
+      "step": 2720
+    },
+    {
+      "epoch": 4.643344709897611,
+      "grad_norm": 0.2972819480569761,
+      "learning_rate": 6.167755429423272e-07,
+      "loss": 0.174,
+      "step": 2721
+    },
+    {
+      "epoch": 4.645051194539249,
+      "grad_norm": 0.31806866417226826,
+      "learning_rate": 6.109176661574134e-07,
+      "loss": 0.173,
+      "step": 2722
+    },
+    {
+      "epoch": 4.646757679180888,
+      "grad_norm": 0.29740846813435423,
+      "learning_rate": 6.050873086285602e-07,
+      "loss": 0.214,
+      "step": 2723
+    },
+    {
+      "epoch": 4.648464163822526,
+      "grad_norm": 0.3242311357095905,
+      "learning_rate": 5.992844786308971e-07,
+      "loss": 0.1989,
+      "step": 2724
+    },
+    {
+      "epoch": 4.650170648464163,
+      "grad_norm": 0.31894716450438126,
+      "learning_rate": 5.935091844004759e-07,
+      "loss": 0.1722,
+      "step": 2725
+    },
+    {
+      "epoch": 4.651877133105802,
+      "grad_norm": 0.28662564392079753,
+      "learning_rate": 5.877614341342708e-07,
+      "loss": 0.1811,
+      "step": 2726
+    },
+    {
+      "epoch": 4.65358361774744,
+      "grad_norm": 0.304397679136295,
+      "learning_rate": 5.820412359901629e-07,
+      "loss": 0.1791,
+      "step": 2727
+    },
+    {
+      "epoch": 4.6552901023890785,
+      "grad_norm": 0.29630041278839964,
+      "learning_rate": 5.763485980869265e-07,
+      "loss": 0.1843,
+      "step": 2728
+    },
+    {
+      "epoch": 4.656996587030717,
+      "grad_norm": 0.3165607930599475,
+      "learning_rate": 5.706835285042233e-07,
+      "loss": 0.1731,
+      "step": 2729
+    },
+    {
+      "epoch": 4.658703071672355,
+      "grad_norm": 0.29492874795878143,
+      "learning_rate": 5.650460352825793e-07,
+      "loss": 0.1825,
+      "step": 2730
+    },
+    {
+      "epoch": 4.660409556313994,
+      "grad_norm": 0.3243112959246551,
+      "learning_rate": 5.594361264233849e-07,
+      "loss": 0.1783,
+      "step": 2731
+    },
+    {
+      "epoch": 4.662116040955631,
+      "grad_norm": 0.3032385848030944,
+      "learning_rate": 5.538538098888846e-07,
+      "loss": 0.1832,
+      "step": 2732
+    },
+    {
+      "epoch": 4.663822525597269,
+      "grad_norm": 0.31122906686810375,
+      "learning_rate": 5.482990936021493e-07,
+      "loss": 0.1775,
+      "step": 2733
+    },
+    {
+      "epoch": 4.665529010238908,
+      "grad_norm": 0.3128666110482303,
+      "learning_rate": 5.427719854470881e-07,
+      "loss": 0.1573,
+      "step": 2734
+    },
+    {
+      "epoch": 4.667235494880546,
+      "grad_norm": 0.3072523033104867,
+      "learning_rate": 5.37272493268417e-07,
+      "loss": 0.1822,
+      "step": 2735
+    },
+    {
+      "epoch": 4.6689419795221845,
+      "grad_norm": 0.31487425009550746,
+      "learning_rate": 5.318006248716589e-07,
+      "loss": 0.1713,
+      "step": 2736
+    },
+    {
+      "epoch": 4.670648464163823,
+      "grad_norm": 0.31277740188190023,
+      "learning_rate": 5.263563880231348e-07,
+      "loss": 0.16,
+      "step": 2737
+    },
+    {
+      "epoch": 4.672354948805461,
+      "grad_norm": 0.31728623002589723,
+      "learning_rate": 5.209397904499369e-07,
+      "loss": 0.1844,
+      "step": 2738
+    },
+    {
+      "epoch": 4.674061433447099,
+      "grad_norm": 0.31231543604810585,
+      "learning_rate": 5.155508398399378e-07,
+      "loss": 0.1937,
+      "step": 2739
+    },
+    {
+      "epoch": 4.675767918088737,
+      "grad_norm": 0.29585155268736635,
+      "learning_rate": 5.101895438417659e-07,
+      "loss": 0.176,
+      "step": 2740
+    },
+    {
+      "epoch": 4.677474402730375,
+      "grad_norm": 0.30976890927663414,
+      "learning_rate": 5.048559100648054e-07,
+      "loss": 0.1624,
+      "step": 2741
+    },
+    {
+      "epoch": 4.679180887372014,
+      "grad_norm": 0.33309511940289294,
+      "learning_rate": 4.995499460791675e-07,
+      "loss": 0.1892,
+      "step": 2742
+    },
+    {
+      "epoch": 4.680887372013652,
+      "grad_norm": 0.31102619428772654,
+      "learning_rate": 4.942716594156993e-07,
+      "loss": 0.1675,
+      "step": 2743
+    },
+    {
+      "epoch": 4.6825938566552905,
+      "grad_norm": 0.31624748102064804,
+      "learning_rate": 4.89021057565966e-07,
+      "loss": 0.1485,
+      "step": 2744
+    },
+    {
+      "epoch": 4.684300341296928,
+      "grad_norm": 0.298870727149146,
+      "learning_rate": 4.837981479822307e-07,
+      "loss": 0.2,
+      "step": 2745
+    },
+    {
+      "epoch": 4.686006825938566,
+      "grad_norm": 0.30261983612884613,
+      "learning_rate": 4.78602938077466e-07,
+      "loss": 0.1727,
+      "step": 2746
+    },
+    {
+      "epoch": 4.687713310580205,
+      "grad_norm": 0.29766006341673873,
+      "learning_rate": 4.7343543522531563e-07,
+      "loss": 0.1545,
+      "step": 2747
+    },
+    {
+      "epoch": 4.689419795221843,
+      "grad_norm": 0.30366717370402896,
+      "learning_rate": 4.6829564676011076e-07,
+      "loss": 0.1776,
+      "step": 2748
+    },
+    {
+      "epoch": 4.691126279863481,
+      "grad_norm": 0.3122046083113992,
+      "learning_rate": 4.6318357997683583e-07,
+      "loss": 0.1541,
+      "step": 2749
+    },
+    {
+      "epoch": 4.69283276450512,
+      "grad_norm": 0.295379614278616,
+      "learning_rate": 4.580992421311359e-07,
+      "loss": 0.1702,
+      "step": 2750
+    },
+    {
+      "epoch": 4.694539249146757,
+      "grad_norm": 0.28431269348999033,
+      "learning_rate": 4.530426404393007e-07,
+      "loss": 0.1764,
+      "step": 2751
+    },
+    {
+      "epoch": 4.696245733788396,
+      "grad_norm": 0.3081064890767569,
+      "learning_rate": 4.480137820782493e-07,
+      "loss": 0.1709,
+      "step": 2752
+    },
+    {
+      "epoch": 4.697952218430034,
+      "grad_norm": 0.331527140545914,
+      "learning_rate": 4.4301267418552786e-07,
+      "loss": 0.1802,
+      "step": 2753
+    },
+    {
+      "epoch": 4.699658703071672,
+      "grad_norm": 0.2918451725241063,
+      "learning_rate": 4.380393238592917e-07,
+      "loss": 0.1924,
+      "step": 2754
+    },
+    {
+      "epoch": 4.701365187713311,
+      "grad_norm": 0.3260580720198137,
+      "learning_rate": 4.3309373815830334e-07,
+      "loss": 0.175,
+      "step": 2755
+    },
+    {
+      "epoch": 4.703071672354949,
+      "grad_norm": 0.2960709556526829,
+      "learning_rate": 4.281759241019212e-07,
+      "loss": 0.1755,
+      "step": 2756
+    },
+    {
+      "epoch": 4.704778156996587,
+      "grad_norm": 0.306517715233806,
+      "learning_rate": 4.2328588867007526e-07,
+      "loss": 0.1767,
+      "step": 2757
+    },
+    {
+      "epoch": 4.706484641638225,
+      "grad_norm": 0.3270552723774275,
+      "learning_rate": 4.184236388032825e-07,
+      "loss": 0.1829,
+      "step": 2758
+    },
+    {
+      "epoch": 4.708191126279863,
+      "grad_norm": 0.3213406072411129,
+      "learning_rate": 4.1358918140261385e-07,
+      "loss": 0.1905,
+      "step": 2759
+    },
+    {
+      "epoch": 4.709897610921502,
+      "grad_norm": 0.3154723719671582,
+      "learning_rate": 4.0878252332970046e-07,
+      "loss": 0.1803,
+      "step": 2760
+    },
+    {
+      "epoch": 4.71160409556314,
+      "grad_norm": 0.30352072765361676,
+      "learning_rate": 4.040036714067119e-07,
+      "loss": 0.1895,
+      "step": 2761
+    },
+    {
+      "epoch": 4.713310580204778,
+      "grad_norm": 0.2948796231530349,
+      "learning_rate": 3.992526324163537e-07,
+      "loss": 0.1802,
+      "step": 2762
+    },
+    {
+      "epoch": 4.715017064846417,
+      "grad_norm": 0.31508476168355987,
+      "learning_rate": 3.945294131018584e-07,
+      "loss": 0.192,
+      "step": 2763
+    },
+    {
+      "epoch": 4.716723549488055,
+      "grad_norm": 0.29702285597614453,
+      "learning_rate": 3.898340201669726e-07,
+      "loss": 0.1668,
+      "step": 2764
+    },
+    {
+      "epoch": 4.7184300341296925,
+      "grad_norm": 0.31669356081958894,
+      "learning_rate": 3.851664602759453e-07,
+      "loss": 0.1841,
+      "step": 2765
+    },
+    {
+      "epoch": 4.720136518771331,
+      "grad_norm": 0.3068568373032543,
+      "learning_rate": 3.805267400535262e-07,
+      "loss": 0.1633,
+      "step": 2766
+    },
+    {
+      "epoch": 4.721843003412969,
+      "grad_norm": 0.2865610931038237,
+      "learning_rate": 3.759148660849521e-07,
+      "loss": 0.1592,
+      "step": 2767
+    },
+    {
+      "epoch": 4.723549488054608,
+      "grad_norm": 0.31918610916972,
+      "learning_rate": 3.71330844915927e-07,
+      "loss": 0.1925,
+      "step": 2768
+    },
+    {
+      "epoch": 4.725255972696246,
+      "grad_norm": 0.3350479762390741,
+      "learning_rate": 3.667746830526331e-07,
+      "loss": 0.1523,
+      "step": 2769
+    },
+    {
+      "epoch": 4.726962457337884,
+      "grad_norm": 0.3294918911488467,
+      "learning_rate": 3.622463869617154e-07,
+      "loss": 0.1693,
+      "step": 2770
+    },
+    {
+      "epoch": 4.728668941979522,
+      "grad_norm": 0.3373615262899913,
+      "learning_rate": 3.577459630702551e-07,
+      "loss": 0.1717,
+      "step": 2771
+    },
+    {
+      "epoch": 4.73037542662116,
+      "grad_norm": 0.31649543808060154,
+      "learning_rate": 3.5327341776578263e-07,
+      "loss": 0.1744,
+      "step": 2772
+    },
+    {
+      "epoch": 4.7320819112627985,
+      "grad_norm": 0.3015120607731921,
+      "learning_rate": 3.488287573962601e-07,
+      "loss": 0.1697,
+      "step": 2773
+    },
+    {
+      "epoch": 4.733788395904437,
+      "grad_norm": 0.34245813217671023,
+      "learning_rate": 3.444119882700658e-07,
+      "loss": 0.1776,
+      "step": 2774
+    },
+    {
+      "epoch": 4.735494880546075,
+      "grad_norm": 0.3141969796974187,
+      "learning_rate": 3.400231166559986e-07,
+      "loss": 0.1855,
+      "step": 2775
+    },
+    {
+      "epoch": 4.737201365187714,
+      "grad_norm": 0.27139431295185523,
+      "learning_rate": 3.3566214878325564e-07,
+      "loss": 0.1515,
+      "step": 2776
+    },
+    {
+      "epoch": 4.738907849829351,
+      "grad_norm": 0.2907069336102253,
+      "learning_rate": 3.3132909084143906e-07,
+      "loss": 0.1714,
+      "step": 2777
+    },
+    {
+      "epoch": 4.7406143344709895,
+      "grad_norm": 0.2925091837910021,
+      "learning_rate": 3.2702394898052936e-07,
+      "loss": 0.1865,
+      "step": 2778
+    },
+    {
+      "epoch": 4.742320819112628,
+      "grad_norm": 0.33924125096350244,
+      "learning_rate": 3.2274672931088766e-07,
+      "loss": 0.1804,
+      "step": 2779
+    },
+    {
+      "epoch": 4.744027303754266,
+      "grad_norm": 0.3012524961675928,
+      "learning_rate": 3.184974379032424e-07,
+      "loss": 0.1508,
+      "step": 2780
+    },
+    {
+      "epoch": 4.7457337883959045,
+      "grad_norm": 0.28754242174783323,
+      "learning_rate": 3.1427608078869133e-07,
+      "loss": 0.1791,
+      "step": 2781
+    },
+    {
+      "epoch": 4.747440273037543,
+      "grad_norm": 0.30860274649893404,
+      "learning_rate": 3.100826639586707e-07,
+      "loss": 0.154,
+      "step": 2782
+    },
+    {
+      "epoch": 4.749146757679181,
+      "grad_norm": 0.3113431887945359,
+      "learning_rate": 3.059171933649752e-07,
+      "loss": 0.1662,
+      "step": 2783
+    },
+    {
+      "epoch": 4.750853242320819,
+      "grad_norm": 0.32115456934856906,
+      "learning_rate": 3.0177967491972884e-07,
+      "loss": 0.1577,
+      "step": 2784
+    },
+    {
+      "epoch": 4.752559726962457,
+      "grad_norm": 0.3028379125307131,
+      "learning_rate": 2.976701144953786e-07,
+      "loss": 0.18,
+      "step": 2785
+    },
+    {
+      "epoch": 4.7542662116040955,
+      "grad_norm": 0.28721084226302274,
+      "learning_rate": 2.9358851792469665e-07,
+      "loss": 0.2204,
+      "step": 2786
+    },
+    {
+      "epoch": 4.755972696245734,
+      "grad_norm": 0.3092608536419546,
+      "learning_rate": 2.8953489100076003e-07,
+      "loss": 0.1989,
+      "step": 2787
+    },
+    {
+      "epoch": 4.757679180887372,
+      "grad_norm": 0.3056259123470543,
+      "learning_rate": 2.855092394769532e-07,
+      "loss": 0.1785,
+      "step": 2788
+    },
+    {
+      "epoch": 4.7593856655290105,
+      "grad_norm": 0.2815407935497288,
+      "learning_rate": 2.815115690669501e-07,
+      "loss": 0.1791,
+      "step": 2789
+    },
+    {
+      "epoch": 4.761092150170649,
+      "grad_norm": 0.31139205545351045,
+      "learning_rate": 2.7754188544471426e-07,
+      "loss": 0.1717,
+      "step": 2790
+    },
+    {
+      "epoch": 4.762798634812286,
+      "grad_norm": 0.30346067170537583,
+      "learning_rate": 2.7360019424448545e-07,
+      "loss": 0.1797,
+      "step": 2791
+    },
+    {
+      "epoch": 4.764505119453925,
+      "grad_norm": 0.31944855771688185,
+      "learning_rate": 2.6968650106077296e-07,
+      "loss": 0.1546,
+      "step": 2792
+    },
+    {
+      "epoch": 4.766211604095563,
+      "grad_norm": 0.320352778892074,
+      "learning_rate": 2.6580081144834903e-07,
+      "loss": 0.1647,
+      "step": 2793
+    },
+    {
+      "epoch": 4.7679180887372015,
+      "grad_norm": 0.3221202846879678,
+      "learning_rate": 2.6194313092223756e-07,
+      "loss": 0.1541,
+      "step": 2794
+    },
+    {
+      "epoch": 4.76962457337884,
+      "grad_norm": 0.31167416292528777,
+      "learning_rate": 2.5811346495771436e-07,
+      "loss": 0.194,
+      "step": 2795
+    },
+    {
+      "epoch": 4.771331058020478,
+      "grad_norm": 0.2748895521787072,
+      "learning_rate": 2.5431181899028267e-07,
+      "loss": 0.2068,
+      "step": 2796
+    },
+    {
+      "epoch": 4.773037542662116,
+      "grad_norm": 0.30999551577917733,
+      "learning_rate": 2.5053819841569295e-07,
+      "loss": 0.1688,
+      "step": 2797
+    },
+    {
+      "epoch": 4.774744027303754,
+      "grad_norm": 0.2905927479723217,
+      "learning_rate": 2.4679260858990306e-07,
+      "loss": 0.1815,
+      "step": 2798
+    },
+    {
+      "epoch": 4.776450511945392,
+      "grad_norm": 0.33157245543394975,
+      "learning_rate": 2.4307505482909166e-07,
+      "loss": 0.2107,
+      "step": 2799
+    },
+    {
+      "epoch": 4.778156996587031,
+      "grad_norm": 0.2928265507497819,
+      "learning_rate": 2.393855424096514e-07,
+      "loss": 0.2077,
+      "step": 2800
+    },
+    {
+      "epoch": 4.779863481228669,
+      "grad_norm": 0.32357915474682025,
+      "learning_rate": 2.3572407656816676e-07,
+      "loss": 0.1586,
+      "step": 2801
+    },
+    {
+      "epoch": 4.7815699658703075,
+      "grad_norm": 0.32482969682952206,
+      "learning_rate": 2.3209066250142077e-07,
+      "loss": 0.1842,
+      "step": 2802
+    },
+    {
+      "epoch": 4.783276450511945,
+      "grad_norm": 0.318429985615523,
+      "learning_rate": 2.2848530536637713e-07,
+      "loss": 0.157,
+      "step": 2803
+    },
+    {
+      "epoch": 4.784982935153583,
+      "grad_norm": 0.3027980996848901,
+      "learning_rate": 2.2490801028018704e-07,
+      "loss": 0.1555,
+      "step": 2804
+    },
+    {
+      "epoch": 4.786689419795222,
+      "grad_norm": 0.3243052592109917,
+      "learning_rate": 2.2135878232016016e-07,
+      "loss": 0.1785,
+      "step": 2805
+    },
+    {
+      "epoch": 4.78839590443686,
+      "grad_norm": 0.28715838094365875,
+      "learning_rate": 2.1783762652377806e-07,
+      "loss": 0.2175,
+      "step": 2806
+    },
+    {
+      "epoch": 4.790102389078498,
+      "grad_norm": 0.28236179247227394,
+      "learning_rate": 2.1434454788867854e-07,
+      "loss": 0.1893,
+      "step": 2807
+    },
+    {
+      "epoch": 4.791808873720137,
+      "grad_norm": 0.30194532666071067,
+      "learning_rate": 2.1087955137264694e-07,
+      "loss": 0.1522,
+      "step": 2808
+    },
+    {
+      "epoch": 4.793515358361775,
+      "grad_norm": 0.3300862918201864,
+      "learning_rate": 2.0744264189361373e-07,
+      "loss": 0.1923,
+      "step": 2809
+    },
+    {
+      "epoch": 4.795221843003413,
+      "grad_norm": 0.28551516089001727,
+      "learning_rate": 2.0403382432964358e-07,
+      "loss": 0.1659,
+      "step": 2810
+    },
+    {
+      "epoch": 4.796928327645051,
+      "grad_norm": 0.304688556538885,
+      "learning_rate": 2.006531035189241e-07,
+      "loss": 0.1434,
+      "step": 2811
+    },
+    {
+      "epoch": 4.798634812286689,
+      "grad_norm": 0.2985728529439217,
+      "learning_rate": 1.97300484259777e-07,
+      "loss": 0.1734,
+      "step": 2812
+    },
+    {
+      "epoch": 4.800341296928328,
+      "grad_norm": 0.28064308184830183,
+      "learning_rate": 1.9397597131062929e-07,
+      "loss": 0.1966,
+      "step": 2813
+    },
+    {
+      "epoch": 4.802047781569966,
+      "grad_norm": 0.3211313780680713,
+      "learning_rate": 1.9067956939001763e-07,
+      "loss": 0.2012,
+      "step": 2814
+    },
+    {
+      "epoch": 4.803754266211604,
+      "grad_norm": 0.29065414759287017,
+      "learning_rate": 1.8741128317658176e-07,
+      "loss": 0.1672,
+      "step": 2815
+    },
+    {
+      "epoch": 4.805460750853243,
+      "grad_norm": 0.3228224201642899,
+      "learning_rate": 1.841711173090599e-07,
+      "loss": 0.1898,
+      "step": 2816
+    },
+    {
+      "epoch": 4.80716723549488,
+      "grad_norm": 0.28289833310229995,
+      "learning_rate": 1.809590763862712e-07,
+      "loss": 0.1695,
+      "step": 2817
+    },
+    {
+      "epoch": 4.808873720136519,
+      "grad_norm": 0.31084333570315364,
+      "learning_rate": 1.777751649671222e-07,
+      "loss": 0.208,
+      "step": 2818
+    },
+    {
+      "epoch": 4.810580204778157,
+      "grad_norm": 0.3262453918891251,
+      "learning_rate": 1.746193875705915e-07,
+      "loss": 0.187,
+      "step": 2819
+    },
+    {
+      "epoch": 4.812286689419795,
+      "grad_norm": 0.2989793802404444,
+      "learning_rate": 1.7149174867572725e-07,
+      "loss": 0.2378,
+      "step": 2820
+    },
+    {
+      "epoch": 4.813993174061434,
+      "grad_norm": 0.3005972736647702,
+      "learning_rate": 1.6839225272164306e-07,
+      "loss": 0.182,
+      "step": 2821
+    },
+    {
+      "epoch": 4.815699658703072,
+      "grad_norm": 0.28847734685992565,
+      "learning_rate": 1.6532090410750656e-07,
+      "loss": 0.1665,
+      "step": 2822
+    },
+    {
+      "epoch": 4.8174061433447095,
+      "grad_norm": 0.27575510564221406,
+      "learning_rate": 1.6227770719253299e-07,
+      "loss": 0.1784,
+      "step": 2823
+    },
+    {
+      "epoch": 4.819112627986348,
+      "grad_norm": 0.33633905780001094,
+      "learning_rate": 1.5926266629598507e-07,
+      "loss": 0.182,
+      "step": 2824
+    },
+    {
+      "epoch": 4.820819112627986,
+      "grad_norm": 0.31140841186578966,
+      "learning_rate": 1.5627578569715974e-07,
+      "loss": 0.1858,
+      "step": 2825
+    },
+    {
+      "epoch": 4.822525597269625,
+      "grad_norm": 0.3140449726136613,
+      "learning_rate": 1.533170696353925e-07,
+      "loss": 0.1453,
+      "step": 2826
+    },
+    {
+      "epoch": 4.824232081911263,
+      "grad_norm": 0.3025865907438611,
+      "learning_rate": 1.5038652231003759e-07,
+      "loss": 0.2066,
+      "step": 2827
+    },
+    {
+      "epoch": 4.825938566552901,
+      "grad_norm": 0.3117915855683816,
+      "learning_rate": 1.4748414788046783e-07,
+      "loss": 0.2041,
+      "step": 2828
+    },
+    {
+      "epoch": 4.827645051194539,
+      "grad_norm": 0.3105049324517198,
+      "learning_rate": 1.4460995046607694e-07,
+      "loss": 0.167,
+      "step": 2829
+    },
+    {
+      "epoch": 4.829351535836177,
+      "grad_norm": 0.30577258623016473,
+      "learning_rate": 1.4176393414625956e-07,
+      "loss": 0.2138,
+      "step": 2830
+    },
+    {
+      "epoch": 4.8310580204778155,
+      "grad_norm": 0.2904177630274955,
+      "learning_rate": 1.3894610296041776e-07,
+      "loss": 0.1692,
+      "step": 2831
+    },
+    {
+      "epoch": 4.832764505119454,
+      "grad_norm": 0.3126283626493536,
+      "learning_rate": 1.3615646090794575e-07,
+      "loss": 0.2796,
+      "step": 2832
+    },
+    {
+      "epoch": 4.834470989761092,
+      "grad_norm": 0.3167202436833116,
+      "learning_rate": 1.333950119482319e-07,
+      "loss": 0.1588,
+      "step": 2833
+    },
+    {
+      "epoch": 4.836177474402731,
+      "grad_norm": 0.30706499524181907,
+      "learning_rate": 1.3066176000064545e-07,
+      "loss": 0.1853,
+      "step": 2834
+    },
+    {
+      "epoch": 4.837883959044369,
+      "grad_norm": 0.31719703891509415,
+      "learning_rate": 1.279567089445388e-07,
+      "loss": 0.1661,
+      "step": 2835
+    },
+    {
+      "epoch": 4.839590443686006,
+      "grad_norm": 0.3105845229605263,
+      "learning_rate": 1.2527986261923863e-07,
+      "loss": 0.1982,
+      "step": 2836
+    },
+    {
+      "epoch": 4.841296928327645,
+      "grad_norm": 0.29004040836338585,
+      "learning_rate": 1.2263122482403688e-07,
+      "loss": 0.175,
+      "step": 2837
+    },
+    {
+      "epoch": 4.843003412969283,
+      "grad_norm": 0.31375723515635084,
+      "learning_rate": 1.2001079931819093e-07,
+      "loss": 0.2174,
+      "step": 2838
+    },
+    {
+      "epoch": 4.8447098976109215,
+      "grad_norm": 0.2965738464238479,
+      "learning_rate": 1.1741858982091459e-07,
+      "loss": 0.1822,
+      "step": 2839
+    },
+    {
+      "epoch": 4.84641638225256,
+      "grad_norm": 0.3146061658375514,
+      "learning_rate": 1.1485460001137816e-07,
+      "loss": 0.1661,
+      "step": 2840
+    },
+    {
+      "epoch": 4.848122866894198,
+      "grad_norm": 0.3090491363527829,
+      "learning_rate": 1.1231883352869288e-07,
+      "loss": 0.1805,
+      "step": 2841
+    },
+    {
+      "epoch": 4.849829351535837,
+      "grad_norm": 0.3111542712054954,
+      "learning_rate": 1.0981129397191759e-07,
+      "loss": 0.1617,
+      "step": 2842
+    },
+    {
+      "epoch": 4.851535836177474,
+      "grad_norm": 0.3095343867608167,
+      "learning_rate": 1.0733198490004537e-07,
+      "loss": 0.1437,
+      "step": 2843
+    },
+    {
+      "epoch": 4.853242320819112,
+      "grad_norm": 0.27464120927047125,
+      "learning_rate": 1.0488090983199917e-07,
+      "loss": 0.2008,
+      "step": 2844
+    },
+    {
+      "epoch": 4.854948805460751,
+      "grad_norm": 0.2960470457039208,
+      "learning_rate": 1.0245807224663839e-07,
+      "loss": 0.1984,
+      "step": 2845
+    },
+    {
+      "epoch": 4.856655290102389,
+      "grad_norm": 0.31142625347213476,
+      "learning_rate": 1.0006347558273011e-07,
+      "loss": 0.1721,
+      "step": 2846
+    },
+    {
+      "epoch": 4.8583617747440275,
+      "grad_norm": 0.2941382931020921,
+      "learning_rate": 9.76971232389734e-08,
+      "loss": 0.1897,
+      "step": 2847
+    },
+    {
+      "epoch": 4.860068259385666,
+      "grad_norm": 0.32342748889464856,
+      "learning_rate": 9.535901857396612e-08,
+      "loss": 0.1814,
+      "step": 2848
+    },
+    {
+      "epoch": 4.861774744027304,
+      "grad_norm": 0.31555805635068296,
+      "learning_rate": 9.304916490622484e-08,
+      "loss": 0.1559,
+      "step": 2849
+    },
+    {
+      "epoch": 4.863481228668942,
+      "grad_norm": 0.29670548942782626,
+      "learning_rate": 9.076756551416266e-08,
+      "loss": 0.1854,
+      "step": 2850
+    },
+    {
+      "epoch": 4.86518771331058,
+      "grad_norm": 0.303231056138658,
+      "learning_rate": 8.851422363609363e-08,
+      "loss": 0.1793,
+      "step": 2851
+    },
+    {
+      "epoch": 4.8668941979522184,
+      "grad_norm": 0.29573043410037203,
+      "learning_rate": 8.628914247022168e-08,
+      "loss": 0.1831,
+      "step": 2852
+    },
+    {
+      "epoch": 4.868600682593857,
+      "grad_norm": 0.323917051878582,
+      "learning_rate": 8.409232517464727e-08,
+      "loss": 0.1927,
+      "step": 2853
+    },
+    {
+      "epoch": 4.870307167235495,
+      "grad_norm": 0.3124109856277249,
+      "learning_rate": 8.192377486734516e-08,
+      "loss": 0.1396,
+      "step": 2854
+    },
+    {
+      "epoch": 4.872013651877133,
+      "grad_norm": 0.2886688196130614,
+      "learning_rate": 7.978349462617996e-08,
+      "loss": 0.175,
+      "step": 2855
+    },
+    {
+      "epoch": 4.873720136518771,
+      "grad_norm": 0.3114793111963149,
+      "learning_rate": 7.76714874888862e-08,
+      "loss": 0.2133,
+      "step": 2856
+    },
+    {
+      "epoch": 4.875426621160409,
+      "grad_norm": 0.2938654172339554,
+      "learning_rate": 7.55877564530727e-08,
+      "loss": 0.1905,
+      "step": 2857
+    },
+    {
+      "epoch": 4.877133105802048,
+      "grad_norm": 0.2840451874449184,
+      "learning_rate": 7.353230447621373e-08,
+      "loss": 0.1889,
+      "step": 2858
+    },
+    {
+      "epoch": 4.878839590443686,
+      "grad_norm": 0.31742584514956634,
+      "learning_rate": 7.1505134475649e-08,
+      "loss": 0.1533,
+      "step": 2859
+    },
+    {
+      "epoch": 4.8805460750853245,
+      "grad_norm": 0.3036746756978537,
+      "learning_rate": 6.950624932857253e-08,
+      "loss": 0.1991,
+      "step": 2860
+    },
+    {
+      "epoch": 4.882252559726963,
+      "grad_norm": 0.4697302986930764,
+      "learning_rate": 6.753565187203937e-08,
+      "loss": 0.1602,
+      "step": 2861
+    },
+    {
+      "epoch": 4.8839590443686,
+      "grad_norm": 0.2984718074892485,
+      "learning_rate": 6.559334490294778e-08,
+      "loss": 0.1782,
+      "step": 2862
+    },
+    {
+      "epoch": 4.885665529010239,
+      "grad_norm": 0.2927837797491726,
+      "learning_rate": 6.367933117805258e-08,
+      "loss": 0.1844,
+      "step": 2863
+    },
+    {
+      "epoch": 4.887372013651877,
+      "grad_norm": 0.32155411946981005,
+      "learning_rate": 6.179361341394297e-08,
+      "loss": 0.1636,
+      "step": 2864
+    },
+    {
+      "epoch": 4.889078498293515,
+      "grad_norm": 0.3008982510208331,
+      "learning_rate": 5.993619428705355e-08,
+      "loss": 0.1883,
+      "step": 2865
+    },
+    {
+      "epoch": 4.890784982935154,
+      "grad_norm": 0.3004924236775075,
+      "learning_rate": 5.810707643364666e-08,
+      "loss": 0.1823,
+      "step": 2866
+    },
+    {
+      "epoch": 4.892491467576792,
+      "grad_norm": 0.28338462990858665,
+      "learning_rate": 5.6306262449823403e-08,
+      "loss": 0.1724,
+      "step": 2867
+    },
+    {
+      "epoch": 4.8941979522184305,
+      "grad_norm": 0.30429095342862417,
+      "learning_rate": 5.453375489150814e-08,
+      "loss": 0.1987,
+      "step": 2868
+    },
+    {
+      "epoch": 4.895904436860068,
+      "grad_norm": 0.3075393001944788,
+      "learning_rate": 5.2789556274452925e-08,
+      "loss": 0.1801,
+      "step": 2869
+    },
+    {
+      "epoch": 4.897610921501706,
+      "grad_norm": 0.3030911501777876,
+      "learning_rate": 5.1073669074228616e-08,
+      "loss": 0.1859,
+      "step": 2870
+    },
+    {
+      "epoch": 4.899317406143345,
+      "grad_norm": 0.283894859345853,
+      "learning_rate": 4.938609572622044e-08,
+      "loss": 0.1939,
+      "step": 2871
+    },
+    {
+      "epoch": 4.901023890784983,
+      "grad_norm": 0.29741908715480087,
+      "learning_rate": 4.772683862563465e-08,
+      "loss": 0.1489,
+      "step": 2872
+    },
+    {
+      "epoch": 4.902730375426621,
+      "grad_norm": 0.3081236236789463,
+      "learning_rate": 4.609590012747856e-08,
+      "loss": 0.1609,
+      "step": 2873
+    },
+    {
+      "epoch": 4.90443686006826,
+      "grad_norm": 0.2912861896734738,
+      "learning_rate": 4.4493282546573815e-08,
+      "loss": 0.1732,
+      "step": 2874
+    },
+    {
+      "epoch": 4.906143344709898,
+      "grad_norm": 0.29829247907729073,
+      "learning_rate": 4.291898815754314e-08,
+      "loss": 0.2248,
+      "step": 2875
+    },
+    {
+      "epoch": 4.907849829351536,
+      "grad_norm": 0.32330446741770824,
+      "learning_rate": 4.1373019194808074e-08,
+      "loss": 0.1876,
+      "step": 2876
+    },
+    {
+      "epoch": 4.909556313993174,
+      "grad_norm": 0.3082796970775283,
+      "learning_rate": 3.985537785259119e-08,
+      "loss": 0.1795,
+      "step": 2877
+    },
+    {
+      "epoch": 4.911262798634812,
+      "grad_norm": 0.3089579069750301,
+      "learning_rate": 3.83660662849028e-08,
+      "loss": 0.1678,
+      "step": 2878
+    },
+    {
+      "epoch": 4.912969283276451,
+      "grad_norm": 0.3021216791497954,
+      "learning_rate": 3.690508660555203e-08,
+      "loss": 0.1623,
+      "step": 2879
+    },
+    {
+      "epoch": 4.914675767918089,
+      "grad_norm": 0.3155362336068173,
+      "learning_rate": 3.547244088812907e-08,
+      "loss": 0.1788,
+      "step": 2880
+    },
+    {
+      "epoch": 4.9163822525597265,
+      "grad_norm": 0.31759061550143486,
+      "learning_rate": 3.4068131166016264e-08,
+      "loss": 0.149,
+      "step": 2881
+    },
+    {
+      "epoch": 4.918088737201365,
+      "grad_norm": 0.30188078406571833,
+      "learning_rate": 3.2692159432370364e-08,
+      "loss": 0.1841,
+      "step": 2882
+    },
+    {
+      "epoch": 4.919795221843003,
+      "grad_norm": 0.3105937487156836,
+      "learning_rate": 3.134452764013363e-08,
+      "loss": 0.1682,
+      "step": 2883
+    },
+    {
+      "epoch": 4.921501706484642,
+      "grad_norm": 0.3280878154771452,
+      "learning_rate": 3.002523770202492e-08,
+      "loss": 0.1934,
+      "step": 2884
+    },
+    {
+      "epoch": 4.92320819112628,
+      "grad_norm": 0.30362182221058903,
+      "learning_rate": 2.8734291490530863e-08,
+      "loss": 0.1708,
+      "step": 2885
+    },
+    {
+      "epoch": 4.924914675767918,
+      "grad_norm": 0.31904967974094284,
+      "learning_rate": 2.7471690837916897e-08,
+      "loss": 0.1638,
+      "step": 2886
+    },
+    {
+      "epoch": 4.926621160409557,
+      "grad_norm": 0.2850030617567953,
+      "learning_rate": 2.6237437536211774e-08,
+      "loss": 0.1527,
+      "step": 2887
+    },
+    {
+      "epoch": 4.928327645051194,
+      "grad_norm": 0.30847599156420746,
+      "learning_rate": 2.5031533337211978e-08,
+      "loss": 0.1799,
+      "step": 2888
+    },
+    {
+      "epoch": 4.9300341296928325,
+      "grad_norm": 0.3091155118863489,
+      "learning_rate": 2.3853979952481733e-08,
+      "loss": 0.1766,
+      "step": 2889
+    },
+    {
+      "epoch": 4.931740614334471,
+      "grad_norm": 0.3172787205950717,
+      "learning_rate": 2.2704779053337456e-08,
+      "loss": 0.1852,
+      "step": 2890
+    },
+    {
+      "epoch": 4.933447098976109,
+      "grad_norm": 0.29594053893996564,
+      "learning_rate": 2.1583932270863307e-08,
+      "loss": 0.159,
+      "step": 2891
+    },
+    {
+      "epoch": 4.935153583617748,
+      "grad_norm": 0.3084174629664152,
+      "learning_rate": 2.0491441195893412e-08,
+      "loss": 0.2037,
+      "step": 2892
+    },
+    {
+      "epoch": 4.936860068259386,
+      "grad_norm": 0.3192737298258064,
+      "learning_rate": 1.9427307379020765e-08,
+      "loss": 0.1704,
+      "step": 2893
+    },
+    {
+      "epoch": 4.938566552901024,
+      "grad_norm": 0.3081067773374951,
+      "learning_rate": 1.8391532330590544e-08,
+      "loss": 0.2199,
+      "step": 2894
+    },
+    {
+      "epoch": 4.940273037542662,
+      "grad_norm": 0.3143303371306659,
+      "learning_rate": 1.7384117520691246e-08,
+      "loss": 0.1593,
+      "step": 2895
+    },
+    {
+      "epoch": 4.9419795221843,
+      "grad_norm": 0.2841289933841045,
+      "learning_rate": 1.640506437917022e-08,
+      "loss": 0.1685,
+      "step": 2896
+    },
+    {
+      "epoch": 4.9436860068259385,
+      "grad_norm": 0.2947377064229445,
+      "learning_rate": 1.545437429560703e-08,
+      "loss": 0.1894,
+      "step": 2897
+    },
+    {
+      "epoch": 4.945392491467577,
+      "grad_norm": 0.37895918120738786,
+      "learning_rate": 1.453204861933788e-08,
+      "loss": 0.2802,
+      "step": 2898
+    },
+    {
+      "epoch": 4.947098976109215,
+      "grad_norm": 0.30593636302043853,
+      "learning_rate": 1.363808865943339e-08,
+      "loss": 0.1536,
+      "step": 2899
+    },
+    {
+      "epoch": 4.948805460750854,
+      "grad_norm": 0.32220876524546754,
+      "learning_rate": 1.277249568470751e-08,
+      "loss": 0.1849,
+      "step": 2900
+    },
+    {
+      "epoch": 4.950511945392492,
+      "grad_norm": 0.2937325209008791,
+      "learning_rate": 1.1935270923708609e-08,
+      "loss": 0.2105,
+      "step": 2901
+    },
+    {
+      "epoch": 4.952218430034129,
+      "grad_norm": 0.31260924234270576,
+      "learning_rate": 1.1126415564726157e-08,
+      "loss": 0.1763,
+      "step": 2902
+    },
+    {
+      "epoch": 4.953924914675768,
+      "grad_norm": 0.30275187594512826,
+      "learning_rate": 1.034593075578183e-08,
+      "loss": 0.1669,
+      "step": 2903
+    },
+    {
+      "epoch": 4.955631399317406,
+      "grad_norm": 0.3368758053021886,
+      "learning_rate": 9.59381760463174e-09,
+      "loss": 0.1716,
+      "step": 2904
+    },
+    {
+      "epoch": 4.9573378839590445,
+      "grad_norm": 0.29851385572075545,
+      "learning_rate": 8.870077178761981e-09,
+      "loss": 0.1782,
+      "step": 2905
+    },
+    {
+      "epoch": 4.959044368600683,
+      "grad_norm": 0.2764806802342295,
+      "learning_rate": 8.17471050538865e-09,
+      "loss": 0.2035,
+      "step": 2906
+    },
+    {
+      "epoch": 4.96075085324232,
+      "grad_norm": 0.31223657501904756,
+      "learning_rate": 7.507718571460044e-09,
+      "loss": 0.1478,
+      "step": 2907
+    },
+    {
+      "epoch": 4.962457337883959,
+      "grad_norm": 0.32092822818481126,
+      "learning_rate": 6.8691023236477914e-09,
+      "loss": 0.18,
+      "step": 2908
+    },
+    {
+      "epoch": 4.964163822525597,
+      "grad_norm": 0.3139808834377375,
+      "learning_rate": 6.258862668351296e-09,
+      "loss": 0.1992,
+      "step": 2909
+    },
+    {
+      "epoch": 4.965870307167235,
+      "grad_norm": 0.36154629547941414,
+      "learning_rate": 5.677000471693283e-09,
+      "loss": 0.1739,
+      "step": 2910
+    },
+    {
+      "epoch": 4.967576791808874,
+      "grad_norm": 0.29872209839584657,
+      "learning_rate": 5.123516559522035e-09,
+      "loss": 0.1777,
+      "step": 2911
+    },
+    {
+      "epoch": 4.969283276450512,
+      "grad_norm": 0.2961551390547971,
+      "learning_rate": 4.598411717404716e-09,
+      "loss": 0.1676,
+      "step": 2912
+    },
+    {
+      "epoch": 4.9709897610921505,
+      "grad_norm": 0.30619626425180235,
+      "learning_rate": 4.1016866906340435e-09,
+      "loss": 0.1828,
+      "step": 2913
+    },
+    {
+      "epoch": 4.972696245733788,
+      "grad_norm": 0.3010614571479018,
+      "learning_rate": 3.6333421842194015e-09,
+      "loss": 0.1393,
+      "step": 2914
+    },
+    {
+      "epoch": 4.974402730375426,
+      "grad_norm": 0.2809770847281617,
+      "learning_rate": 3.193378862891283e-09,
+      "loss": 0.1452,
+      "step": 2915
+    },
+    {
+      "epoch": 4.976109215017065,
+      "grad_norm": 0.29638133049217613,
+      "learning_rate": 2.7817973510946284e-09,
+      "loss": 0.2233,
+      "step": 2916
+    },
+    {
+      "epoch": 4.977815699658703,
+      "grad_norm": 0.30868966487380983,
+      "learning_rate": 2.398598232995486e-09,
+      "loss": 0.1718,
+      "step": 2917
+    },
+    {
+      "epoch": 4.979522184300341,
+      "grad_norm": 0.2928606482705561,
+      "learning_rate": 2.0437820524743524e-09,
+      "loss": 0.1662,
+      "step": 2918
+    },
+    {
+      "epoch": 4.98122866894198,
+      "grad_norm": 0.32612854136308245,
+      "learning_rate": 1.7173493131283914e-09,
+      "loss": 0.1693,
+      "step": 2919
+    },
+    {
+      "epoch": 4.982935153583618,
+      "grad_norm": 0.28808533238959344,
+      "learning_rate": 1.4193004782692144e-09,
+      "loss": 0.1895,
+      "step": 2920
+    },
+    {
+      "epoch": 4.984641638225256,
+      "grad_norm": 0.3318512216485675,
+      "learning_rate": 1.1496359709228798e-09,
+      "loss": 0.181,
+      "step": 2921
+    },
+    {
+      "epoch": 4.986348122866894,
+      "grad_norm": 0.3745280144589005,
+      "learning_rate": 9.083561738276736e-10,
+      "loss": 0.2369,
+      "step": 2922
+    },
+    {
+      "epoch": 4.988054607508532,
+      "grad_norm": 0.30803026547551043,
+      "learning_rate": 6.954614294385486e-10,
+      "loss": 0.1786,
+      "step": 2923
+    },
+    {
+      "epoch": 4.989761092150171,
+      "grad_norm": 0.30026959439019923,
+      "learning_rate": 5.109520399182443e-10,
+      "loss": 0.1634,
+      "step": 2924
+    },
+    {
+      "epoch": 4.991467576791809,
+      "grad_norm": 0.2966519292573069,
+      "learning_rate": 3.5482826714394733e-10,
+      "loss": 0.1842,
+      "step": 2925
+    },
+    {
+      "epoch": 4.993174061433447,
+      "grad_norm": 0.29530756983955236,
+      "learning_rate": 2.2709033270729154e-10,
+      "loss": 0.1511,
+      "step": 2926
+    },
+    {
+      "epoch": 4.994880546075086,
+      "grad_norm": 0.3529263677322087,
+      "learning_rate": 1.2773841790769682e-10,
+      "loss": 0.1369,
+      "step": 2927
+    },
+    {
+      "epoch": 4.996587030716723,
+      "grad_norm": 0.3007431179010985,
+      "learning_rate": 5.677266375458956e-11,
+      "loss": 0.2011,
+      "step": 2928
+    },
+    {
+      "epoch": 4.998293515358362,
+      "grad_norm": 0.32241652480938615,
+      "learning_rate": 1.4193170974063918e-11,
+      "loss": 0.1771,
+      "step": 2929
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.3247430104362262,
+      "learning_rate": 0.0,
+      "loss": 0.1839,
+      "step": 2930
+    },
+    {
+      "epoch": 5.0,
+      "step": 2930,
+      "total_flos": 2698707716407296.0,
+      "train_loss": 0.34839511024463704,
+      "train_runtime": 52686.9412,
+      "train_samples_per_second": 7.118,
+      "train_steps_per_second": 0.056
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2930,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2698707716407296.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}