diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,37530 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 5352,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005605381165919282,
+      "grad_norm": 0.978611306363841,
+      "learning_rate": 3.7313432835820895e-07,
+      "loss": 1.3562,
+      "step": 1
+    },
+    {
+      "epoch": 0.0011210762331838565,
+      "grad_norm": 0.9785669708543357,
+      "learning_rate": 7.462686567164179e-07,
+      "loss": 1.3832,
+      "step": 2
+    },
+    {
+      "epoch": 0.0016816143497757848,
+      "grad_norm": 0.965928509838546,
+      "learning_rate": 1.119402985074627e-06,
+      "loss": 1.3498,
+      "step": 3
+    },
+    {
+      "epoch": 0.002242152466367713,
+      "grad_norm": 1.0011476156785488,
+      "learning_rate": 1.4925373134328358e-06,
+      "loss": 1.3665,
+      "step": 4
+    },
+    {
+      "epoch": 0.002802690582959641,
+      "grad_norm": 0.9743028706157322,
+      "learning_rate": 1.8656716417910446e-06,
+      "loss": 1.3718,
+      "step": 5
+    },
+    {
+      "epoch": 0.0033632286995515697,
+      "grad_norm": 0.9549639453848878,
+      "learning_rate": 2.238805970149254e-06,
+      "loss": 1.355,
+      "step": 6
+    },
+    {
+      "epoch": 0.003923766816143498,
+      "grad_norm": 0.9444168662872774,
+      "learning_rate": 2.6119402985074627e-06,
+      "loss": 1.3296,
+      "step": 7
+    },
+    {
+      "epoch": 0.004484304932735426,
+      "grad_norm": 0.9653130750500606,
+      "learning_rate": 2.9850746268656716e-06,
+      "loss": 1.3505,
+      "step": 8
+    },
+    {
+      "epoch": 0.005044843049327354,
+      "grad_norm": 0.9728971645466663,
+      "learning_rate": 3.358208955223881e-06,
+      "loss": 1.3525,
+      "step": 9
+    },
+    {
+      "epoch": 0.005605381165919282,
+      "grad_norm": 0.9263086036248187,
+      "learning_rate": 3.7313432835820893e-06,
+      "loss": 1.3388,
+      "step": 10
+    },
+    {
+      "epoch": 0.00616591928251121,
+      "grad_norm": 0.9753471433575638,
+      "learning_rate": 4.1044776119402985e-06,
+      "loss": 1.3496,
+      "step": 11
+    },
+    {
+      "epoch": 0.006726457399103139,
+      "grad_norm": 0.9040200360486869,
+      "learning_rate": 4.477611940298508e-06,
+      "loss": 1.3239,
+      "step": 12
+    },
+    {
+      "epoch": 0.0072869955156950675,
+      "grad_norm": 0.9437511706878353,
+      "learning_rate": 4.850746268656717e-06,
+      "loss": 1.3111,
+      "step": 13
+    },
+    {
+      "epoch": 0.007847533632286996,
+      "grad_norm": 0.9494404428795452,
+      "learning_rate": 5.2238805970149255e-06,
+      "loss": 1.3522,
+      "step": 14
+    },
+    {
+      "epoch": 0.008408071748878924,
+      "grad_norm": 0.9317807129575663,
+      "learning_rate": 5.597014925373135e-06,
+      "loss": 1.3261,
+      "step": 15
+    },
+    {
+      "epoch": 0.008968609865470852,
+      "grad_norm": 0.9503925528940235,
+      "learning_rate": 5.970149253731343e-06,
+      "loss": 1.3358,
+      "step": 16
+    },
+    {
+      "epoch": 0.00952914798206278,
+      "grad_norm": 0.8692982982016761,
+      "learning_rate": 6.343283582089552e-06,
+      "loss": 1.2674,
+      "step": 17
+    },
+    {
+      "epoch": 0.010089686098654708,
+      "grad_norm": 0.8512916351445661,
+      "learning_rate": 6.716417910447762e-06,
+      "loss": 1.261,
+      "step": 18
+    },
+    {
+      "epoch": 0.010650224215246636,
+      "grad_norm": 0.8710189287442345,
+      "learning_rate": 7.08955223880597e-06,
+      "loss": 1.2842,
+      "step": 19
+    },
+    {
+      "epoch": 0.011210762331838564,
+      "grad_norm": 0.781668523585148,
+      "learning_rate": 7.4626865671641785e-06,
+      "loss": 1.2373,
+      "step": 20
+    },
+    {
+      "epoch": 0.011771300448430493,
+      "grad_norm": 0.7381052580031785,
+      "learning_rate": 7.835820895522389e-06,
+      "loss": 1.1989,
+      "step": 21
+    },
+    {
+      "epoch": 0.01233183856502242,
+      "grad_norm": 0.6770393690721677,
+      "learning_rate": 8.208955223880597e-06,
+      "loss": 1.1672,
+      "step": 22
+    },
+    {
+      "epoch": 0.01289237668161435,
+      "grad_norm": 0.6191395018917473,
+      "learning_rate": 8.582089552238805e-06,
+      "loss": 1.1619,
+      "step": 23
+    },
+    {
+      "epoch": 0.013452914798206279,
+      "grad_norm": 0.6289526892086984,
+      "learning_rate": 8.955223880597016e-06,
+      "loss": 1.1558,
+      "step": 24
+    },
+    {
+      "epoch": 0.014013452914798207,
+      "grad_norm": 0.5682916094085656,
+      "learning_rate": 9.328358208955226e-06,
+      "loss": 1.1096,
+      "step": 25
+    },
+    {
+      "epoch": 0.014573991031390135,
+      "grad_norm": 0.5649795252033589,
+      "learning_rate": 9.701492537313434e-06,
+      "loss": 1.1183,
+      "step": 26
+    },
+    {
+      "epoch": 0.015134529147982063,
+      "grad_norm": 0.5592638391416755,
+      "learning_rate": 1.0074626865671643e-05,
+      "loss": 1.1039,
+      "step": 27
+    },
+    {
+      "epoch": 0.01569506726457399,
+      "grad_norm": 0.5453326006757038,
+      "learning_rate": 1.0447761194029851e-05,
+      "loss": 1.0674,
+      "step": 28
+    },
+    {
+      "epoch": 0.016255605381165918,
+      "grad_norm": 0.5439999602347558,
+      "learning_rate": 1.082089552238806e-05,
+      "loss": 1.0574,
+      "step": 29
+    },
+    {
+      "epoch": 0.016816143497757848,
+      "grad_norm": 0.5511396127286549,
+      "learning_rate": 1.119402985074627e-05,
+      "loss": 1.013,
+      "step": 30
+    },
+    {
+      "epoch": 0.017376681614349777,
+      "grad_norm": 0.5702221080642447,
+      "learning_rate": 1.1567164179104478e-05,
+      "loss": 0.9702,
+      "step": 31
+    },
+    {
+      "epoch": 0.017937219730941704,
+      "grad_norm": 0.5720559627118424,
+      "learning_rate": 1.1940298507462686e-05,
+      "loss": 0.9662,
+      "step": 32
+    },
+    {
+      "epoch": 0.018497757847533634,
+      "grad_norm": 0.5611113226975845,
+      "learning_rate": 1.2313432835820896e-05,
+      "loss": 0.9274,
+      "step": 33
+    },
+    {
+      "epoch": 0.01905829596412556,
+      "grad_norm": 0.6090665517488381,
+      "learning_rate": 1.2686567164179105e-05,
+      "loss": 0.9,
+      "step": 34
+    },
+    {
+      "epoch": 0.01961883408071749,
+      "grad_norm": 0.5839791809624392,
+      "learning_rate": 1.3059701492537313e-05,
+      "loss": 0.8903,
+      "step": 35
+    },
+    {
+      "epoch": 0.020179372197309416,
+      "grad_norm": 0.5736113015335764,
+      "learning_rate": 1.3432835820895523e-05,
+      "loss": 0.8887,
+      "step": 36
+    },
+    {
+      "epoch": 0.020739910313901346,
+      "grad_norm": 0.5190623064152249,
+      "learning_rate": 1.3805970149253733e-05,
+      "loss": 0.8273,
+      "step": 37
+    },
+    {
+      "epoch": 0.021300448430493273,
+      "grad_norm": 0.553768945628212,
+      "learning_rate": 1.417910447761194e-05,
+      "loss": 0.8283,
+      "step": 38
+    },
+    {
+      "epoch": 0.021860986547085202,
+      "grad_norm": 0.534909748905568,
+      "learning_rate": 1.455223880597015e-05,
+      "loss": 0.782,
+      "step": 39
+    },
+    {
+      "epoch": 0.02242152466367713,
+      "grad_norm": 0.5401435696197063,
+      "learning_rate": 1.4925373134328357e-05,
+      "loss": 0.7611,
+      "step": 40
+    },
+    {
+      "epoch": 0.02298206278026906,
+      "grad_norm": 0.528685630974113,
+      "learning_rate": 1.529850746268657e-05,
+      "loss": 0.7325,
+      "step": 41
+    },
+    {
+      "epoch": 0.023542600896860985,
+      "grad_norm": 0.49093077695142545,
+      "learning_rate": 1.5671641791044777e-05,
+      "loss": 0.6981,
+      "step": 42
+    },
+    {
+      "epoch": 0.024103139013452915,
+      "grad_norm": 0.4438139439428313,
+      "learning_rate": 1.6044776119402986e-05,
+      "loss": 0.6774,
+      "step": 43
+    },
+    {
+      "epoch": 0.02466367713004484,
+      "grad_norm": 0.4482263523387966,
+      "learning_rate": 1.6417910447761194e-05,
+      "loss": 0.6645,
+      "step": 44
+    },
+    {
+      "epoch": 0.02522421524663677,
+      "grad_norm": 0.4540231120607396,
+      "learning_rate": 1.6791044776119406e-05,
+      "loss": 0.6123,
+      "step": 45
+    },
+    {
+      "epoch": 0.0257847533632287,
+      "grad_norm": 0.4326299184847521,
+      "learning_rate": 1.716417910447761e-05,
+      "loss": 0.5931,
+      "step": 46
+    },
+    {
+      "epoch": 0.026345291479820628,
+      "grad_norm": 0.3024749933538969,
+      "learning_rate": 1.7537313432835823e-05,
+      "loss": 0.6052,
+      "step": 47
+    },
+    {
+      "epoch": 0.026905829596412557,
+      "grad_norm": 0.2818759060120326,
+      "learning_rate": 1.791044776119403e-05,
+      "loss": 0.5998,
+      "step": 48
+    },
+    {
+      "epoch": 0.027466367713004484,
+      "grad_norm": 0.27406340113751826,
+      "learning_rate": 1.828358208955224e-05,
+      "loss": 0.5961,
+      "step": 49
+    },
+    {
+      "epoch": 0.028026905829596414,
+      "grad_norm": 0.2479023351788548,
+      "learning_rate": 1.865671641791045e-05,
+      "loss": 0.5627,
+      "step": 50
+    },
+    {
+      "epoch": 0.02858744394618834,
+      "grad_norm": 0.23908909726117478,
+      "learning_rate": 1.9029850746268656e-05,
+      "loss": 0.5535,
+      "step": 51
+    },
+    {
+      "epoch": 0.02914798206278027,
+      "grad_norm": 0.21890178164904825,
+      "learning_rate": 1.9402985074626868e-05,
+      "loss": 0.5498,
+      "step": 52
+    },
+    {
+      "epoch": 0.029708520179372196,
+      "grad_norm": 0.20444438580516272,
+      "learning_rate": 1.9776119402985073e-05,
+      "loss": 0.5661,
+      "step": 53
+    },
+    {
+      "epoch": 0.030269058295964126,
+      "grad_norm": 0.20254090583306722,
+      "learning_rate": 2.0149253731343285e-05,
+      "loss": 0.5653,
+      "step": 54
+    },
+    {
+      "epoch": 0.030829596412556053,
+      "grad_norm": 0.21987649104111895,
+      "learning_rate": 2.0522388059701493e-05,
+      "loss": 0.5406,
+      "step": 55
+    },
+    {
+      "epoch": 0.03139013452914798,
+      "grad_norm": 0.18690597707242654,
+      "learning_rate": 2.0895522388059702e-05,
+      "loss": 0.5544,
+      "step": 56
+    },
+    {
+      "epoch": 0.03195067264573991,
+      "grad_norm": 0.18441579639270703,
+      "learning_rate": 2.126865671641791e-05,
+      "loss": 0.5324,
+      "step": 57
+    },
+    {
+      "epoch": 0.032511210762331835,
+      "grad_norm": 0.17795328095595792,
+      "learning_rate": 2.164179104477612e-05,
+      "loss": 0.5284,
+      "step": 58
+    },
+    {
+      "epoch": 0.033071748878923765,
+      "grad_norm": 0.17770157145787804,
+      "learning_rate": 2.201492537313433e-05,
+      "loss": 0.5199,
+      "step": 59
+    },
+    {
+      "epoch": 0.033632286995515695,
+      "grad_norm": 0.19713002587880057,
+      "learning_rate": 2.238805970149254e-05,
+      "loss": 0.4949,
+      "step": 60
+    },
+    {
+      "epoch": 0.034192825112107625,
+      "grad_norm": 0.1936909800596092,
+      "learning_rate": 2.2761194029850747e-05,
+      "loss": 0.5079,
+      "step": 61
+    },
+    {
+      "epoch": 0.034753363228699555,
+      "grad_norm": 0.1766770198216385,
+      "learning_rate": 2.3134328358208956e-05,
+      "loss": 0.5172,
+      "step": 62
+    },
+    {
+      "epoch": 0.03531390134529148,
+      "grad_norm": 0.16215448560164902,
+      "learning_rate": 2.3507462686567168e-05,
+      "loss": 0.515,
+      "step": 63
+    },
+    {
+      "epoch": 0.03587443946188341,
+      "grad_norm": 0.16097779703313178,
+      "learning_rate": 2.3880597014925373e-05,
+      "loss": 0.4931,
+      "step": 64
+    },
+    {
+      "epoch": 0.03643497757847534,
+      "grad_norm": 0.1679553680550244,
+      "learning_rate": 2.4253731343283584e-05,
+      "loss": 0.5,
+      "step": 65
+    },
+    {
+      "epoch": 0.03699551569506727,
+      "grad_norm": 0.16770477974378512,
+      "learning_rate": 2.4626865671641793e-05,
+      "loss": 0.5052,
+      "step": 66
+    },
+    {
+      "epoch": 0.03755605381165919,
+      "grad_norm": 0.15086103753068986,
+      "learning_rate": 2.5e-05,
+      "loss": 0.4932,
+      "step": 67
+    },
+    {
+      "epoch": 0.03811659192825112,
+      "grad_norm": 0.1448393261526423,
+      "learning_rate": 2.537313432835821e-05,
+      "loss": 0.4965,
+      "step": 68
+    },
+    {
+      "epoch": 0.03867713004484305,
+      "grad_norm": 0.13536710404495228,
+      "learning_rate": 2.574626865671642e-05,
+      "loss": 0.4949,
+      "step": 69
+    },
+    {
+      "epoch": 0.03923766816143498,
+      "grad_norm": 0.1318960286316755,
+      "learning_rate": 2.6119402985074626e-05,
+      "loss": 0.4717,
+      "step": 70
+    },
+    {
+      "epoch": 0.0397982062780269,
+      "grad_norm": 0.13908346439467814,
+      "learning_rate": 2.6492537313432835e-05,
+      "loss": 0.4886,
+      "step": 71
+    },
+    {
+      "epoch": 0.04035874439461883,
+      "grad_norm": 0.13274477559301887,
+      "learning_rate": 2.6865671641791047e-05,
+      "loss": 0.478,
+      "step": 72
+    },
+    {
+      "epoch": 0.04091928251121076,
+      "grad_norm": 0.12187263958584976,
+      "learning_rate": 2.7238805970149255e-05,
+      "loss": 0.481,
+      "step": 73
+    },
+    {
+      "epoch": 0.04147982062780269,
+      "grad_norm": 0.12112276934943426,
+      "learning_rate": 2.7611940298507467e-05,
+      "loss": 0.4629,
+      "step": 74
+    },
+    {
+      "epoch": 0.04204035874439462,
+      "grad_norm": 0.12784869143888544,
+      "learning_rate": 2.7985074626865672e-05,
+      "loss": 0.4667,
+      "step": 75
+    },
+    {
+      "epoch": 0.042600896860986545,
+      "grad_norm": 0.12366913602187946,
+      "learning_rate": 2.835820895522388e-05,
+      "loss": 0.4663,
+      "step": 76
+    },
+    {
+      "epoch": 0.043161434977578475,
+      "grad_norm": 0.13038172476851312,
+      "learning_rate": 2.8731343283582092e-05,
+      "loss": 0.4644,
+      "step": 77
+    },
+    {
+      "epoch": 0.043721973094170405,
+      "grad_norm": 0.11454023239597903,
+      "learning_rate": 2.91044776119403e-05,
+      "loss": 0.4528,
+      "step": 78
+    },
+    {
+      "epoch": 0.044282511210762335,
+      "grad_norm": 0.11601962977381795,
+      "learning_rate": 2.9477611940298512e-05,
+      "loss": 0.462,
+      "step": 79
+    },
+    {
+      "epoch": 0.04484304932735426,
+      "grad_norm": 0.1219882451392154,
+      "learning_rate": 2.9850746268656714e-05,
+      "loss": 0.4791,
+      "step": 80
+    },
+    {
+      "epoch": 0.04540358744394619,
+      "grad_norm": 0.1293231926178124,
+      "learning_rate": 3.0223880597014926e-05,
+      "loss": 0.4684,
+      "step": 81
+    },
+    {
+      "epoch": 0.04596412556053812,
+      "grad_norm": 0.11996052800463039,
+      "learning_rate": 3.059701492537314e-05,
+      "loss": 0.4566,
+      "step": 82
+    },
+    {
+      "epoch": 0.04652466367713005,
+      "grad_norm": 0.12246511147890005,
+      "learning_rate": 3.0970149253731346e-05,
+      "loss": 0.4681,
+      "step": 83
+    },
+    {
+      "epoch": 0.04708520179372197,
+      "grad_norm": 0.11218525045205559,
+      "learning_rate": 3.1343283582089554e-05,
+      "loss": 0.4633,
+      "step": 84
+    },
+    {
+      "epoch": 0.0476457399103139,
+      "grad_norm": 0.11506805026698533,
+      "learning_rate": 3.171641791044776e-05,
+      "loss": 0.4764,
+      "step": 85
+    },
+    {
+      "epoch": 0.04820627802690583,
+      "grad_norm": 0.1278539909669664,
+      "learning_rate": 3.208955223880597e-05,
+      "loss": 0.4578,
+      "step": 86
+    },
+    {
+      "epoch": 0.04876681614349776,
+      "grad_norm": 0.11325282709726044,
+      "learning_rate": 3.246268656716418e-05,
+      "loss": 0.436,
+      "step": 87
+    },
+    {
+      "epoch": 0.04932735426008968,
+      "grad_norm": 0.11889933717086941,
+      "learning_rate": 3.283582089552239e-05,
+      "loss": 0.45,
+      "step": 88
+    },
+    {
+      "epoch": 0.04988789237668161,
+      "grad_norm": 0.12864292199145877,
+      "learning_rate": 3.32089552238806e-05,
+      "loss": 0.4465,
+      "step": 89
+    },
+    {
+      "epoch": 0.05044843049327354,
+      "grad_norm": 0.12367148564811485,
+      "learning_rate": 3.358208955223881e-05,
+      "loss": 0.4442,
+      "step": 90
+    },
+    {
+      "epoch": 0.05100896860986547,
+      "grad_norm": 0.11840931265139464,
+      "learning_rate": 3.395522388059701e-05,
+      "loss": 0.4342,
+      "step": 91
+    },
+    {
+      "epoch": 0.0515695067264574,
+      "grad_norm": 0.12460573190500202,
+      "learning_rate": 3.432835820895522e-05,
+      "loss": 0.4704,
+      "step": 92
+    },
+    {
+      "epoch": 0.052130044843049325,
+      "grad_norm": 0.11275332417661361,
+      "learning_rate": 3.470149253731344e-05,
+      "loss": 0.4566,
+      "step": 93
+    },
+    {
+      "epoch": 0.052690582959641255,
+      "grad_norm": 0.12027472702499808,
+      "learning_rate": 3.5074626865671645e-05,
+      "loss": 0.4683,
+      "step": 94
+    },
+    {
+      "epoch": 0.053251121076233185,
+      "grad_norm": 0.12166680224485274,
+      "learning_rate": 3.5447761194029854e-05,
+      "loss": 0.4502,
+      "step": 95
+    },
+    {
+      "epoch": 0.053811659192825115,
+      "grad_norm": 0.1143149054981622,
+      "learning_rate": 3.582089552238806e-05,
+      "loss": 0.4384,
+      "step": 96
+    },
+    {
+      "epoch": 0.05437219730941704,
+      "grad_norm": 0.11701314141339264,
+      "learning_rate": 3.619402985074627e-05,
+      "loss": 0.4361,
+      "step": 97
+    },
+    {
+      "epoch": 0.05493273542600897,
+      "grad_norm": 0.12407417894757372,
+      "learning_rate": 3.656716417910448e-05,
+      "loss": 0.4498,
+      "step": 98
+    },
+    {
+      "epoch": 0.0554932735426009,
+      "grad_norm": 0.1300165206876751,
+      "learning_rate": 3.694029850746269e-05,
+      "loss": 0.4563,
+      "step": 99
+    },
+    {
+      "epoch": 0.05605381165919283,
+      "grad_norm": 0.11518521501317551,
+      "learning_rate": 3.73134328358209e-05,
+      "loss": 0.4466,
+      "step": 100
+    },
+    {
+      "epoch": 0.05661434977578475,
+      "grad_norm": 0.11754858539944062,
+      "learning_rate": 3.7686567164179104e-05,
+      "loss": 0.4267,
+      "step": 101
+    },
+    {
+      "epoch": 0.05717488789237668,
+      "grad_norm": 0.12077083786609252,
+      "learning_rate": 3.805970149253731e-05,
+      "loss": 0.4313,
+      "step": 102
+    },
+    {
+      "epoch": 0.05773542600896861,
+      "grad_norm": 0.12080980975033845,
+      "learning_rate": 3.843283582089552e-05,
+      "loss": 0.4335,
+      "step": 103
+    },
+    {
+      "epoch": 0.05829596412556054,
+      "grad_norm": 0.1242265803727591,
+      "learning_rate": 3.8805970149253736e-05,
+      "loss": 0.4461,
+      "step": 104
+    },
+    {
+      "epoch": 0.05885650224215247,
+      "grad_norm": 0.13003496252871916,
+      "learning_rate": 3.9179104477611945e-05,
+      "loss": 0.4447,
+      "step": 105
+    },
+    {
+      "epoch": 0.05941704035874439,
+      "grad_norm": 0.12908146749282087,
+      "learning_rate": 3.9552238805970146e-05,
+      "loss": 0.4219,
+      "step": 106
+    },
+    {
+      "epoch": 0.05997757847533632,
+      "grad_norm": 0.11582509313092396,
+      "learning_rate": 3.992537313432836e-05,
+      "loss": 0.4108,
+      "step": 107
+    },
+    {
+      "epoch": 0.06053811659192825,
+      "grad_norm": 0.12643265441930768,
+      "learning_rate": 4.029850746268657e-05,
+      "loss": 0.4229,
+      "step": 108
+    },
+    {
+      "epoch": 0.06109865470852018,
+      "grad_norm": 0.13820675934864066,
+      "learning_rate": 4.067164179104478e-05,
+      "loss": 0.4275,
+      "step": 109
+    },
+    {
+      "epoch": 0.061659192825112105,
+      "grad_norm": 0.1360538838053136,
+      "learning_rate": 4.104477611940299e-05,
+      "loss": 0.4312,
+      "step": 110
+    },
+    {
+      "epoch": 0.062219730941704035,
+      "grad_norm": 0.12554029300674785,
+      "learning_rate": 4.1417910447761195e-05,
+      "loss": 0.4326,
+      "step": 111
+    },
+    {
+      "epoch": 0.06278026905829596,
+      "grad_norm": 0.13890555381256833,
+      "learning_rate": 4.1791044776119404e-05,
+      "loss": 0.4339,
+      "step": 112
+    },
+    {
+      "epoch": 0.0633408071748879,
+      "grad_norm": 0.13227702018718224,
+      "learning_rate": 4.216417910447761e-05,
+      "loss": 0.414,
+      "step": 113
+    },
+    {
+      "epoch": 0.06390134529147982,
+      "grad_norm": 0.1376833510856438,
+      "learning_rate": 4.253731343283582e-05,
+      "loss": 0.4179,
+      "step": 114
+    },
+    {
+      "epoch": 0.06446188340807175,
+      "grad_norm": 0.1416082156028862,
+      "learning_rate": 4.2910447761194036e-05,
+      "loss": 0.4207,
+      "step": 115
+    },
+    {
+      "epoch": 0.06502242152466367,
+      "grad_norm": 0.1365191037664384,
+      "learning_rate": 4.328358208955224e-05,
+      "loss": 0.4243,
+      "step": 116
+    },
+    {
+      "epoch": 0.0655829596412556,
+      "grad_norm": 0.14062318849572733,
+      "learning_rate": 4.3656716417910446e-05,
+      "loss": 0.413,
+      "step": 117
+    },
+    {
+      "epoch": 0.06614349775784753,
+      "grad_norm": 0.14187004014123297,
+      "learning_rate": 4.402985074626866e-05,
+      "loss": 0.4135,
+      "step": 118
+    },
+    {
+      "epoch": 0.06670403587443946,
+      "grad_norm": 0.14531335764922046,
+      "learning_rate": 4.440298507462687e-05,
+      "loss": 0.4147,
+      "step": 119
+    },
+    {
+      "epoch": 0.06726457399103139,
+      "grad_norm": 0.13755295535271247,
+      "learning_rate": 4.477611940298508e-05,
+      "loss": 0.433,
+      "step": 120
+    },
+    {
+      "epoch": 0.06782511210762332,
+      "grad_norm": 0.1432819698901696,
+      "learning_rate": 4.5149253731343286e-05,
+      "loss": 0.4096,
+      "step": 121
+    },
+    {
+      "epoch": 0.06838565022421525,
+      "grad_norm": 0.14193457086049052,
+      "learning_rate": 4.5522388059701495e-05,
+      "loss": 0.4245,
+      "step": 122
+    },
+    {
+      "epoch": 0.06894618834080718,
+      "grad_norm": 0.13928060042054785,
+      "learning_rate": 4.58955223880597e-05,
+      "loss": 0.4262,
+      "step": 123
+    },
+    {
+      "epoch": 0.06950672645739911,
+      "grad_norm": 0.14003170515118332,
+      "learning_rate": 4.626865671641791e-05,
+      "loss": 0.4035,
+      "step": 124
+    },
+    {
+      "epoch": 0.07006726457399103,
+      "grad_norm": 0.1331650387336229,
+      "learning_rate": 4.664179104477612e-05,
+      "loss": 0.4203,
+      "step": 125
+    },
+    {
+      "epoch": 0.07062780269058296,
+      "grad_norm": 0.13498074727847828,
+      "learning_rate": 4.7014925373134335e-05,
+      "loss": 0.3935,
+      "step": 126
+    },
+    {
+      "epoch": 0.07118834080717489,
+      "grad_norm": 0.13666502845282,
+      "learning_rate": 4.738805970149254e-05,
+      "loss": 0.4042,
+      "step": 127
+    },
+    {
+      "epoch": 0.07174887892376682,
+      "grad_norm": 0.14758309572816788,
+      "learning_rate": 4.7761194029850745e-05,
+      "loss": 0.4058,
+      "step": 128
+    },
+    {
+      "epoch": 0.07230941704035874,
+      "grad_norm": 0.14492402892274034,
+      "learning_rate": 4.813432835820896e-05,
+      "loss": 0.41,
+      "step": 129
+    },
+    {
+      "epoch": 0.07286995515695067,
+      "grad_norm": 0.14695792097356924,
+      "learning_rate": 4.850746268656717e-05,
+      "loss": 0.3948,
+      "step": 130
+    },
+    {
+      "epoch": 0.0734304932735426,
+      "grad_norm": 0.15151969250879915,
+      "learning_rate": 4.888059701492538e-05,
+      "loss": 0.3948,
+      "step": 131
+    },
+    {
+      "epoch": 0.07399103139013453,
+      "grad_norm": 0.13754418875073354,
+      "learning_rate": 4.9253731343283586e-05,
+      "loss": 0.4022,
+      "step": 132
+    },
+    {
+      "epoch": 0.07455156950672645,
+      "grad_norm": 0.15254984318473536,
+      "learning_rate": 4.9626865671641794e-05,
+      "loss": 0.4013,
+      "step": 133
+    },
+    {
+      "epoch": 0.07511210762331838,
+      "grad_norm": 0.14547684214076642,
+      "learning_rate": 5e-05,
+      "loss": 0.4063,
+      "step": 134
+    },
+    {
+      "epoch": 0.07567264573991031,
+      "grad_norm": 0.14922739042225802,
+      "learning_rate": 5.0373134328358204e-05,
+      "loss": 0.4101,
+      "step": 135
+    },
+    {
+      "epoch": 0.07623318385650224,
+      "grad_norm": 0.15098699044966904,
+      "learning_rate": 5.074626865671642e-05,
+      "loss": 0.4047,
+      "step": 136
+    },
+    {
+      "epoch": 0.07679372197309417,
+      "grad_norm": 0.1452335515819299,
+      "learning_rate": 5.111940298507463e-05,
+      "loss": 0.4048,
+      "step": 137
+    },
+    {
+      "epoch": 0.0773542600896861,
+      "grad_norm": 0.15079187782364642,
+      "learning_rate": 5.149253731343284e-05,
+      "loss": 0.4024,
+      "step": 138
+    },
+    {
+      "epoch": 0.07791479820627803,
+      "grad_norm": 0.14434756712244956,
+      "learning_rate": 5.1865671641791044e-05,
+      "loss": 0.405,
+      "step": 139
+    },
+    {
+      "epoch": 0.07847533632286996,
+      "grad_norm": 0.13254131018383866,
+      "learning_rate": 5.223880597014925e-05,
+      "loss": 0.4001,
+      "step": 140
+    },
+    {
+      "epoch": 0.07903587443946189,
+      "grad_norm": 0.1417441767722713,
+      "learning_rate": 5.261194029850747e-05,
+      "loss": 0.3927,
+      "step": 141
+    },
+    {
+      "epoch": 0.0795964125560538,
+      "grad_norm": 0.15274755766224463,
+      "learning_rate": 5.298507462686567e-05,
+      "loss": 0.4009,
+      "step": 142
+    },
+    {
+      "epoch": 0.08015695067264574,
+      "grad_norm": 0.14245430733786046,
+      "learning_rate": 5.3358208955223885e-05,
+      "loss": 0.3854,
+      "step": 143
+    },
+    {
+      "epoch": 0.08071748878923767,
+      "grad_norm": 0.1536214882615985,
+      "learning_rate": 5.373134328358209e-05,
+      "loss": 0.3851,
+      "step": 144
+    },
+    {
+      "epoch": 0.0812780269058296,
+      "grad_norm": 0.19139835581037787,
+      "learning_rate": 5.4104477611940295e-05,
+      "loss": 0.4039,
+      "step": 145
+    },
+    {
+      "epoch": 0.08183856502242152,
+      "grad_norm": 0.16363703483744255,
+      "learning_rate": 5.447761194029851e-05,
+      "loss": 0.4022,
+      "step": 146
+    },
+    {
+      "epoch": 0.08239910313901345,
+      "grad_norm": 0.15294575979455324,
+      "learning_rate": 5.485074626865672e-05,
+      "loss": 0.3867,
+      "step": 147
+    },
+    {
+      "epoch": 0.08295964125560538,
+      "grad_norm": 0.16677125259280787,
+      "learning_rate": 5.5223880597014934e-05,
+      "loss": 0.3919,
+      "step": 148
+    },
+    {
+      "epoch": 0.08352017937219731,
+      "grad_norm": 0.16265087004682827,
+      "learning_rate": 5.5597014925373135e-05,
+      "loss": 0.3864,
+      "step": 149
+    },
+    {
+      "epoch": 0.08408071748878924,
+      "grad_norm": 0.14272101706953824,
+      "learning_rate": 5.5970149253731344e-05,
+      "loss": 0.3827,
+      "step": 150
+    },
+    {
+      "epoch": 0.08464125560538116,
+      "grad_norm": 0.16824308627818732,
+      "learning_rate": 5.634328358208956e-05,
+      "loss": 0.4,
+      "step": 151
+    },
+    {
+      "epoch": 0.08520179372197309,
+      "grad_norm": 0.16732040229196365,
+      "learning_rate": 5.671641791044776e-05,
+      "loss": 0.3701,
+      "step": 152
+    },
+    {
+      "epoch": 0.08576233183856502,
+      "grad_norm": 0.14882351488471363,
+      "learning_rate": 5.7089552238805976e-05,
+      "loss": 0.3972,
+      "step": 153
+    },
+    {
+      "epoch": 0.08632286995515695,
+      "grad_norm": 0.16796214736400641,
+      "learning_rate": 5.7462686567164184e-05,
+      "loss": 0.4063,
+      "step": 154
+    },
+    {
+      "epoch": 0.08688340807174888,
+      "grad_norm": 0.16090778047008664,
+      "learning_rate": 5.7835820895522386e-05,
+      "loss": 0.4082,
+      "step": 155
+    },
+    {
+      "epoch": 0.08744394618834081,
+      "grad_norm": 0.1515617591452116,
+      "learning_rate": 5.82089552238806e-05,
+      "loss": 0.3936,
+      "step": 156
+    },
+    {
+      "epoch": 0.08800448430493274,
+      "grad_norm": 0.14765500679843194,
+      "learning_rate": 5.85820895522388e-05,
+      "loss": 0.3889,
+      "step": 157
+    },
+    {
+      "epoch": 0.08856502242152467,
+      "grad_norm": 0.15544914165747062,
+      "learning_rate": 5.8955223880597025e-05,
+      "loss": 0.38,
+      "step": 158
+    },
+    {
+      "epoch": 0.08912556053811659,
+      "grad_norm": 0.16721420657142755,
+      "learning_rate": 5.9328358208955226e-05,
+      "loss": 0.3811,
+      "step": 159
+    },
+    {
+      "epoch": 0.08968609865470852,
+      "grad_norm": 0.16794811927217704,
+      "learning_rate": 5.970149253731343e-05,
+      "loss": 0.3909,
+      "step": 160
+    },
+    {
+      "epoch": 0.09024663677130045,
+      "grad_norm": 0.15165371471110536,
+      "learning_rate": 6.007462686567164e-05,
+      "loss": 0.3836,
+      "step": 161
+    },
+    {
+      "epoch": 0.09080717488789238,
+      "grad_norm": 0.15262228871854122,
+      "learning_rate": 6.044776119402985e-05,
+      "loss": 0.367,
+      "step": 162
+    },
+    {
+      "epoch": 0.0913677130044843,
+      "grad_norm": 0.16250954049606353,
+      "learning_rate": 6.082089552238807e-05,
+      "loss": 0.3842,
+      "step": 163
+    },
+    {
+      "epoch": 0.09192825112107623,
+      "grad_norm": 0.16164141273342517,
+      "learning_rate": 6.119402985074628e-05,
+      "loss": 0.4001,
+      "step": 164
+    },
+    {
+      "epoch": 0.09248878923766816,
+      "grad_norm": 0.14821752027103804,
+      "learning_rate": 6.156716417910448e-05,
+      "loss": 0.391,
+      "step": 165
+    },
+    {
+      "epoch": 0.0930493273542601,
+      "grad_norm": 0.1649310143100295,
+      "learning_rate": 6.194029850746269e-05,
+      "loss": 0.3831,
+      "step": 166
+    },
+    {
+      "epoch": 0.09360986547085202,
+      "grad_norm": 0.16182061113902127,
+      "learning_rate": 6.23134328358209e-05,
+      "loss": 0.3779,
+      "step": 167
+    },
+    {
+      "epoch": 0.09417040358744394,
+      "grad_norm": 0.16190445609076318,
+      "learning_rate": 6.268656716417911e-05,
+      "loss": 0.3897,
+      "step": 168
+    },
+    {
+      "epoch": 0.09473094170403587,
+      "grad_norm": 0.15156515470873255,
+      "learning_rate": 6.305970149253731e-05,
+      "loss": 0.3764,
+      "step": 169
+    },
+    {
+      "epoch": 0.0952914798206278,
+      "grad_norm": 0.17156111651448278,
+      "learning_rate": 6.343283582089553e-05,
+      "loss": 0.3938,
+      "step": 170
+    },
+    {
+      "epoch": 0.09585201793721973,
+      "grad_norm": 0.15513896968402013,
+      "learning_rate": 6.380597014925374e-05,
+      "loss": 0.387,
+      "step": 171
+    },
+    {
+      "epoch": 0.09641255605381166,
+      "grad_norm": 0.1788700878157143,
+      "learning_rate": 6.417910447761194e-05,
+      "loss": 0.3749,
+      "step": 172
+    },
+    {
+      "epoch": 0.09697309417040359,
+      "grad_norm": 0.1843006493151444,
+      "learning_rate": 6.455223880597016e-05,
+      "loss": 0.3857,
+      "step": 173
+    },
+    {
+      "epoch": 0.09753363228699552,
+      "grad_norm": 0.16235020723013396,
+      "learning_rate": 6.492537313432836e-05,
+      "loss": 0.3819,
+      "step": 174
+    },
+    {
+      "epoch": 0.09809417040358745,
+      "grad_norm": 0.1767763613709395,
+      "learning_rate": 6.529850746268657e-05,
+      "loss": 0.3761,
+      "step": 175
+    },
+    {
+      "epoch": 0.09865470852017937,
+      "grad_norm": 0.17256885471732497,
+      "learning_rate": 6.567164179104478e-05,
+      "loss": 0.3895,
+      "step": 176
+    },
+    {
+      "epoch": 0.0992152466367713,
+      "grad_norm": 0.16950015988649256,
+      "learning_rate": 6.604477611940298e-05,
+      "loss": 0.3767,
+      "step": 177
+    },
+    {
+      "epoch": 0.09977578475336323,
+      "grad_norm": 0.19591518959576854,
+      "learning_rate": 6.64179104477612e-05,
+      "loss": 0.4027,
+      "step": 178
+    },
+    {
+      "epoch": 0.10033632286995516,
+      "grad_norm": 0.15667226214733038,
+      "learning_rate": 6.679104477611941e-05,
+      "loss": 0.3821,
+      "step": 179
+    },
+    {
+      "epoch": 0.10089686098654709,
+      "grad_norm": 0.17160752708809768,
+      "learning_rate": 6.716417910447762e-05,
+      "loss": 0.3685,
+      "step": 180
+    },
+    {
+      "epoch": 0.10145739910313901,
+      "grad_norm": 0.15121233958417601,
+      "learning_rate": 6.753731343283583e-05,
+      "loss": 0.3677,
+      "step": 181
+    },
+    {
+      "epoch": 0.10201793721973094,
+      "grad_norm": 0.15733202495872775,
+      "learning_rate": 6.791044776119403e-05,
+      "loss": 0.3841,
+      "step": 182
+    },
+    {
+      "epoch": 0.10257847533632287,
+      "grad_norm": 0.17626355608916983,
+      "learning_rate": 6.828358208955224e-05,
+      "loss": 0.3889,
+      "step": 183
+    },
+    {
+      "epoch": 0.1031390134529148,
+      "grad_norm": 0.17412687579414454,
+      "learning_rate": 6.865671641791044e-05,
+      "loss": 0.3793,
+      "step": 184
+    },
+    {
+      "epoch": 0.10369955156950672,
+      "grad_norm": 0.16411490181034571,
+      "learning_rate": 6.902985074626866e-05,
+      "loss": 0.3651,
+      "step": 185
+    },
+    {
+      "epoch": 0.10426008968609865,
+      "grad_norm": 0.1809656119702917,
+      "learning_rate": 6.940298507462687e-05,
+      "loss": 0.3785,
+      "step": 186
+    },
+    {
+      "epoch": 0.10482062780269058,
+      "grad_norm": 0.1703795524339579,
+      "learning_rate": 6.977611940298508e-05,
+      "loss": 0.3742,
+      "step": 187
+    },
+    {
+      "epoch": 0.10538116591928251,
+      "grad_norm": 0.17116149539584355,
+      "learning_rate": 7.014925373134329e-05,
+      "loss": 0.3729,
+      "step": 188
+    },
+    {
+      "epoch": 0.10594170403587444,
+      "grad_norm": 0.17956733029995822,
+      "learning_rate": 7.052238805970149e-05,
+      "loss": 0.3775,
+      "step": 189
+    },
+    {
+      "epoch": 0.10650224215246637,
+      "grad_norm": 0.16346388961754899,
+      "learning_rate": 7.089552238805971e-05,
+      "loss": 0.3873,
+      "step": 190
+    },
+    {
+      "epoch": 0.1070627802690583,
+      "grad_norm": 0.1568726807138821,
+      "learning_rate": 7.126865671641791e-05,
+      "loss": 0.373,
+      "step": 191
+    },
+    {
+      "epoch": 0.10762331838565023,
+      "grad_norm": 0.1796852200923323,
+      "learning_rate": 7.164179104477612e-05,
+      "loss": 0.3804,
+      "step": 192
+    },
+    {
+      "epoch": 0.10818385650224215,
+      "grad_norm": 0.16135711806174538,
+      "learning_rate": 7.201492537313434e-05,
+      "loss": 0.3736,
+      "step": 193
+    },
+    {
+      "epoch": 0.10874439461883408,
+      "grad_norm": 0.16102305151660706,
+      "learning_rate": 7.238805970149254e-05,
+      "loss": 0.3658,
+      "step": 194
+    },
+    {
+      "epoch": 0.109304932735426,
+      "grad_norm": 0.16781137288468057,
+      "learning_rate": 7.276119402985076e-05,
+      "loss": 0.3737,
+      "step": 195
+    },
+    {
+      "epoch": 0.10986547085201794,
+      "grad_norm": 0.18545424393368118,
+      "learning_rate": 7.313432835820896e-05,
+      "loss": 0.3695,
+      "step": 196
+    },
+    {
+      "epoch": 0.11042600896860987,
+      "grad_norm": 0.17958819480714575,
+      "learning_rate": 7.350746268656716e-05,
+      "loss": 0.3759,
+      "step": 197
+    },
+    {
+      "epoch": 0.1109865470852018,
+      "grad_norm": 0.20219498453199727,
+      "learning_rate": 7.388059701492537e-05,
+      "loss": 0.3864,
+      "step": 198
+    },
+    {
+      "epoch": 0.11154708520179372,
+      "grad_norm": 0.1726692529003045,
+      "learning_rate": 7.425373134328359e-05,
+      "loss": 0.3635,
+      "step": 199
+    },
+    {
+      "epoch": 0.11210762331838565,
+      "grad_norm": 0.1746472625241535,
+      "learning_rate": 7.46268656716418e-05,
+      "loss": 0.3774,
+      "step": 200
+    },
+    {
+      "epoch": 0.11266816143497758,
+      "grad_norm": 0.18311083159929262,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.3736,
+      "step": 201
+    },
+    {
+      "epoch": 0.1132286995515695,
+      "grad_norm": 0.18726919900702563,
+      "learning_rate": 7.537313432835821e-05,
+      "loss": 0.3817,
+      "step": 202
+    },
+    {
+      "epoch": 0.11378923766816143,
+      "grad_norm": 0.18987151892129903,
+      "learning_rate": 7.574626865671642e-05,
+      "loss": 0.3794,
+      "step": 203
+    },
+    {
+      "epoch": 0.11434977578475336,
+      "grad_norm": 0.18795067963203516,
+      "learning_rate": 7.611940298507463e-05,
+      "loss": 0.3674,
+      "step": 204
+    },
+    {
+      "epoch": 0.11491031390134529,
+      "grad_norm": 0.19242197215916712,
+      "learning_rate": 7.649253731343284e-05,
+      "loss": 0.364,
+      "step": 205
+    },
+    {
+      "epoch": 0.11547085201793722,
+      "grad_norm": 0.17251419491478528,
+      "learning_rate": 7.686567164179104e-05,
+      "loss": 0.3632,
+      "step": 206
+    },
+    {
+      "epoch": 0.11603139013452915,
+      "grad_norm": 0.1728889950589284,
+      "learning_rate": 7.723880597014926e-05,
+      "loss": 0.3591,
+      "step": 207
+    },
+    {
+      "epoch": 0.11659192825112108,
+      "grad_norm": 0.1899590050084193,
+      "learning_rate": 7.761194029850747e-05,
+      "loss": 0.3759,
+      "step": 208
+    },
+    {
+      "epoch": 0.11715246636771301,
+      "grad_norm": 0.1789564355228219,
+      "learning_rate": 7.798507462686567e-05,
+      "loss": 0.3669,
+      "step": 209
+    },
+    {
+      "epoch": 0.11771300448430494,
+      "grad_norm": 0.18910618441974555,
+      "learning_rate": 7.835820895522389e-05,
+      "loss": 0.3679,
+      "step": 210
+    },
+    {
+      "epoch": 0.11827354260089686,
+      "grad_norm": 0.1768485362984264,
+      "learning_rate": 7.873134328358209e-05,
+      "loss": 0.3668,
+      "step": 211
+    },
+    {
+      "epoch": 0.11883408071748879,
+      "grad_norm": 0.19534942551965084,
+      "learning_rate": 7.910447761194029e-05,
+      "loss": 0.3635,
+      "step": 212
+    },
+    {
+      "epoch": 0.11939461883408072,
+      "grad_norm": 0.17520388878220655,
+      "learning_rate": 7.947761194029851e-05,
+      "loss": 0.3543,
+      "step": 213
+    },
+    {
+      "epoch": 0.11995515695067265,
+      "grad_norm": 0.18235087513990209,
+      "learning_rate": 7.985074626865672e-05,
+      "loss": 0.3658,
+      "step": 214
+    },
+    {
+      "epoch": 0.12051569506726457,
+      "grad_norm": 0.1827333194205262,
+      "learning_rate": 8.022388059701494e-05,
+      "loss": 0.364,
+      "step": 215
+    },
+    {
+      "epoch": 0.1210762331838565,
+      "grad_norm": 0.19175747163007564,
+      "learning_rate": 8.059701492537314e-05,
+      "loss": 0.3903,
+      "step": 216
+    },
+    {
+      "epoch": 0.12163677130044843,
+      "grad_norm": 0.18369610261399918,
+      "learning_rate": 8.097014925373134e-05,
+      "loss": 0.3936,
+      "step": 217
+    },
+    {
+      "epoch": 0.12219730941704036,
+      "grad_norm": 0.17145574679985418,
+      "learning_rate": 8.134328358208956e-05,
+      "loss": 0.3567,
+      "step": 218
+    },
+    {
+      "epoch": 0.12275784753363228,
+      "grad_norm": 0.1921713146643893,
+      "learning_rate": 8.171641791044776e-05,
+      "loss": 0.3674,
+      "step": 219
+    },
+    {
+      "epoch": 0.12331838565022421,
+      "grad_norm": 0.1570865798975667,
+      "learning_rate": 8.208955223880597e-05,
+      "loss": 0.3773,
+      "step": 220
+    },
+    {
+      "epoch": 0.12387892376681614,
+      "grad_norm": 0.1763016822133998,
+      "learning_rate": 8.246268656716419e-05,
+      "loss": 0.3757,
+      "step": 221
+    },
+    {
+      "epoch": 0.12443946188340807,
+      "grad_norm": 0.18174355080196333,
+      "learning_rate": 8.283582089552239e-05,
+      "loss": 0.3653,
+      "step": 222
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 0.16621511712083203,
+      "learning_rate": 8.32089552238806e-05,
+      "loss": 0.3678,
+      "step": 223
+    },
+    {
+      "epoch": 0.12556053811659193,
+      "grad_norm": 0.1663459391630373,
+      "learning_rate": 8.358208955223881e-05,
+      "loss": 0.3705,
+      "step": 224
+    },
+    {
+      "epoch": 0.12612107623318386,
+      "grad_norm": 0.1725075785062485,
+      "learning_rate": 8.395522388059702e-05,
+      "loss": 0.3609,
+      "step": 225
+    },
+    {
+      "epoch": 0.1266816143497758,
+      "grad_norm": 0.15966589014814622,
+      "learning_rate": 8.432835820895522e-05,
+      "loss": 0.36,
+      "step": 226
+    },
+    {
+      "epoch": 0.12724215246636772,
+      "grad_norm": 0.17842238761202062,
+      "learning_rate": 8.470149253731343e-05,
+      "loss": 0.3564,
+      "step": 227
+    },
+    {
+      "epoch": 0.12780269058295965,
+      "grad_norm": 0.18978029743720834,
+      "learning_rate": 8.507462686567164e-05,
+      "loss": 0.3845,
+      "step": 228
+    },
+    {
+      "epoch": 0.12836322869955158,
+      "grad_norm": 0.17577228283856472,
+      "learning_rate": 8.544776119402986e-05,
+      "loss": 0.3494,
+      "step": 229
+    },
+    {
+      "epoch": 0.1289237668161435,
+      "grad_norm": 0.17628453743082703,
+      "learning_rate": 8.582089552238807e-05,
+      "loss": 0.36,
+      "step": 230
+    },
+    {
+      "epoch": 0.12948430493273544,
+      "grad_norm": 0.16428265399110947,
+      "learning_rate": 8.619402985074627e-05,
+      "loss": 0.3647,
+      "step": 231
+    },
+    {
+      "epoch": 0.13004484304932734,
+      "grad_norm": 0.1836827615926638,
+      "learning_rate": 8.656716417910447e-05,
+      "loss": 0.3648,
+      "step": 232
+    },
+    {
+      "epoch": 0.13060538116591927,
+      "grad_norm": 0.1866251710481314,
+      "learning_rate": 8.694029850746269e-05,
+      "loss": 0.3596,
+      "step": 233
+    },
+    {
+      "epoch": 0.1311659192825112,
+      "grad_norm": 0.18372353476625689,
+      "learning_rate": 8.731343283582089e-05,
+      "loss": 0.3581,
+      "step": 234
+    },
+    {
+      "epoch": 0.13172645739910313,
+      "grad_norm": 0.17254888167050808,
+      "learning_rate": 8.76865671641791e-05,
+      "loss": 0.3736,
+      "step": 235
+    },
+    {
+      "epoch": 0.13228699551569506,
+      "grad_norm": 0.1682612354618672,
+      "learning_rate": 8.805970149253732e-05,
+      "loss": 0.3627,
+      "step": 236
+    },
+    {
+      "epoch": 0.132847533632287,
+      "grad_norm": 0.1684459539424527,
+      "learning_rate": 8.843283582089554e-05,
+      "loss": 0.3663,
+      "step": 237
+    },
+    {
+      "epoch": 0.13340807174887892,
+      "grad_norm": 0.16975246091905197,
+      "learning_rate": 8.880597014925374e-05,
+      "loss": 0.3608,
+      "step": 238
+    },
+    {
+      "epoch": 0.13396860986547085,
+      "grad_norm": 0.15473984678290356,
+      "learning_rate": 8.917910447761194e-05,
+      "loss": 0.354,
+      "step": 239
+    },
+    {
+      "epoch": 0.13452914798206278,
+      "grad_norm": 0.18185884309316858,
+      "learning_rate": 8.955223880597016e-05,
+      "loss": 0.37,
+      "step": 240
+    },
+    {
+      "epoch": 0.1350896860986547,
+      "grad_norm": 0.14927991709699787,
+      "learning_rate": 8.992537313432836e-05,
+      "loss": 0.3647,
+      "step": 241
+    },
+    {
+      "epoch": 0.13565022421524664,
+      "grad_norm": 0.1636477741733644,
+      "learning_rate": 9.029850746268657e-05,
+      "loss": 0.353,
+      "step": 242
+    },
+    {
+      "epoch": 0.13621076233183857,
+      "grad_norm": 0.15567916635474852,
+      "learning_rate": 9.067164179104479e-05,
+      "loss": 0.3495,
+      "step": 243
+    },
+    {
+      "epoch": 0.1367713004484305,
+      "grad_norm": 0.1586917656306562,
+      "learning_rate": 9.104477611940299e-05,
+      "loss": 0.3586,
+      "step": 244
+    },
+    {
+      "epoch": 0.13733183856502243,
+      "grad_norm": 0.1658072814748706,
+      "learning_rate": 9.14179104477612e-05,
+      "loss": 0.377,
+      "step": 245
+    },
+    {
+      "epoch": 0.13789237668161436,
+      "grad_norm": 0.17172300258830786,
+      "learning_rate": 9.17910447761194e-05,
+      "loss": 0.3629,
+      "step": 246
+    },
+    {
+      "epoch": 0.1384529147982063,
+      "grad_norm": 0.16886456245374368,
+      "learning_rate": 9.216417910447762e-05,
+      "loss": 0.3633,
+      "step": 247
+    },
+    {
+      "epoch": 0.13901345291479822,
+      "grad_norm": 0.1684761343496583,
+      "learning_rate": 9.253731343283582e-05,
+      "loss": 0.3415,
+      "step": 248
+    },
+    {
+      "epoch": 0.13957399103139012,
+      "grad_norm": 0.16925015900566961,
+      "learning_rate": 9.291044776119402e-05,
+      "loss": 0.3461,
+      "step": 249
+    },
+    {
+      "epoch": 0.14013452914798205,
+      "grad_norm": 0.15843094471805708,
+      "learning_rate": 9.328358208955224e-05,
+      "loss": 0.3551,
+      "step": 250
+    },
+    {
+      "epoch": 0.14069506726457398,
+      "grad_norm": 0.1696294696406677,
+      "learning_rate": 9.365671641791045e-05,
+      "loss": 0.3458,
+      "step": 251
+    },
+    {
+      "epoch": 0.1412556053811659,
+      "grad_norm": 0.17601667106314525,
+      "learning_rate": 9.402985074626867e-05,
+      "loss": 0.3639,
+      "step": 252
+    },
+    {
+      "epoch": 0.14181614349775784,
+      "grad_norm": 0.15595609006027045,
+      "learning_rate": 9.440298507462687e-05,
+      "loss": 0.3735,
+      "step": 253
+    },
+    {
+      "epoch": 0.14237668161434977,
+      "grad_norm": 0.16626523046712888,
+      "learning_rate": 9.477611940298507e-05,
+      "loss": 0.3541,
+      "step": 254
+    },
+    {
+      "epoch": 0.1429372197309417,
+      "grad_norm": 0.16235816753333587,
+      "learning_rate": 9.514925373134329e-05,
+      "loss": 0.3501,
+      "step": 255
+    },
+    {
+      "epoch": 0.14349775784753363,
+      "grad_norm": 0.15176637857752753,
+      "learning_rate": 9.552238805970149e-05,
+      "loss": 0.3619,
+      "step": 256
+    },
+    {
+      "epoch": 0.14405829596412556,
+      "grad_norm": 0.1688461940896348,
+      "learning_rate": 9.58955223880597e-05,
+      "loss": 0.3539,
+      "step": 257
+    },
+    {
+      "epoch": 0.1446188340807175,
+      "grad_norm": 0.15435614225238878,
+      "learning_rate": 9.626865671641792e-05,
+      "loss": 0.3543,
+      "step": 258
+    },
+    {
+      "epoch": 0.14517937219730942,
+      "grad_norm": 0.1795558992512674,
+      "learning_rate": 9.664179104477612e-05,
+      "loss": 0.3631,
+      "step": 259
+    },
+    {
+      "epoch": 0.14573991031390135,
+      "grad_norm": 0.15941611102169298,
+      "learning_rate": 9.701492537313434e-05,
+      "loss": 0.3417,
+      "step": 260
+    },
+    {
+      "epoch": 0.14630044843049328,
+      "grad_norm": 0.16017055644477737,
+      "learning_rate": 9.738805970149254e-05,
+      "loss": 0.3624,
+      "step": 261
+    },
+    {
+      "epoch": 0.1468609865470852,
+      "grad_norm": 0.1702042759017935,
+      "learning_rate": 9.776119402985075e-05,
+      "loss": 0.3675,
+      "step": 262
+    },
+    {
+      "epoch": 0.14742152466367714,
+      "grad_norm": 0.15574217721760533,
+      "learning_rate": 9.813432835820896e-05,
+      "loss": 0.3537,
+      "step": 263
+    },
+    {
+      "epoch": 0.14798206278026907,
+      "grad_norm": 0.15595756003052536,
+      "learning_rate": 9.850746268656717e-05,
+      "loss": 0.3413,
+      "step": 264
+    },
+    {
+      "epoch": 0.148542600896861,
+      "grad_norm": 0.16407156402134312,
+      "learning_rate": 9.888059701492539e-05,
+      "loss": 0.348,
+      "step": 265
+    },
+    {
+      "epoch": 0.1491031390134529,
+      "grad_norm": 0.16740889130311884,
+      "learning_rate": 9.925373134328359e-05,
+      "loss": 0.3434,
+      "step": 266
+    },
+    {
+      "epoch": 0.14966367713004483,
+      "grad_norm": 0.16159905903153549,
+      "learning_rate": 9.96268656716418e-05,
+      "loss": 0.3585,
+      "step": 267
+    },
+    {
+      "epoch": 0.15022421524663676,
+      "grad_norm": 0.16452829367524685,
+      "learning_rate": 0.0001,
+      "loss": 0.36,
+      "step": 268
+    },
+    {
+      "epoch": 0.1507847533632287,
+      "grad_norm": 0.16212933723251338,
+      "learning_rate": 0.00010037313432835822,
+      "loss": 0.3536,
+      "step": 269
+    },
+    {
+      "epoch": 0.15134529147982062,
+      "grad_norm": 0.1677596853734516,
+      "learning_rate": 0.00010074626865671641,
+      "loss": 0.3461,
+      "step": 270
+    },
+    {
+      "epoch": 0.15190582959641255,
+      "grad_norm": 0.14092955431703147,
+      "learning_rate": 0.00010111940298507462,
+      "loss": 0.3514,
+      "step": 271
+    },
+    {
+      "epoch": 0.15246636771300448,
+      "grad_norm": 0.15504121203886792,
+      "learning_rate": 0.00010149253731343284,
+      "loss": 0.3601,
+      "step": 272
+    },
+    {
+      "epoch": 0.1530269058295964,
+      "grad_norm": 0.16086274774375903,
+      "learning_rate": 0.00010186567164179107,
+      "loss": 0.3683,
+      "step": 273
+    },
+    {
+      "epoch": 0.15358744394618834,
+      "grad_norm": 0.1604018025859141,
+      "learning_rate": 0.00010223880597014926,
+      "loss": 0.3633,
+      "step": 274
+    },
+    {
+      "epoch": 0.15414798206278027,
+      "grad_norm": 0.15692121791747665,
+      "learning_rate": 0.00010261194029850747,
+      "loss": 0.332,
+      "step": 275
+    },
+    {
+      "epoch": 0.1547085201793722,
+      "grad_norm": 0.16795441518784324,
+      "learning_rate": 0.00010298507462686569,
+      "loss": 0.3644,
+      "step": 276
+    },
+    {
+      "epoch": 0.15526905829596413,
+      "grad_norm": 0.17131877689821567,
+      "learning_rate": 0.00010335820895522387,
+      "loss": 0.3434,
+      "step": 277
+    },
+    {
+      "epoch": 0.15582959641255606,
+      "grad_norm": 0.1607628706106527,
+      "learning_rate": 0.00010373134328358209,
+      "loss": 0.3626,
+      "step": 278
+    },
+    {
+      "epoch": 0.156390134529148,
+      "grad_norm": 0.16269328407163647,
+      "learning_rate": 0.0001041044776119403,
+      "loss": 0.3586,
+      "step": 279
+    },
+    {
+      "epoch": 0.15695067264573992,
+      "grad_norm": 0.17535923762892863,
+      "learning_rate": 0.0001044776119402985,
+      "loss": 0.3644,
+      "step": 280
+    },
+    {
+      "epoch": 0.15751121076233185,
+      "grad_norm": 0.15924213345397006,
+      "learning_rate": 0.00010485074626865672,
+      "loss": 0.3443,
+      "step": 281
+    },
+    {
+      "epoch": 0.15807174887892378,
+      "grad_norm": 0.16200755025292304,
+      "learning_rate": 0.00010522388059701494,
+      "loss": 0.3451,
+      "step": 282
+    },
+    {
+      "epoch": 0.1586322869955157,
+      "grad_norm": 0.16240644121268424,
+      "learning_rate": 0.00010559701492537315,
+      "loss": 0.3433,
+      "step": 283
+    },
+    {
+      "epoch": 0.1591928251121076,
+      "grad_norm": 0.1599085397839329,
+      "learning_rate": 0.00010597014925373134,
+      "loss": 0.3445,
+      "step": 284
+    },
+    {
+      "epoch": 0.15975336322869954,
+      "grad_norm": 0.16881740152266073,
+      "learning_rate": 0.00010634328358208955,
+      "loss": 0.3587,
+      "step": 285
+    },
+    {
+      "epoch": 0.16031390134529147,
+      "grad_norm": 0.15197850146245018,
+      "learning_rate": 0.00010671641791044777,
+      "loss": 0.3581,
+      "step": 286
+    },
+    {
+      "epoch": 0.1608744394618834,
+      "grad_norm": 0.15979984233256023,
+      "learning_rate": 0.00010708955223880597,
+      "loss": 0.3325,
+      "step": 287
+    },
+    {
+      "epoch": 0.16143497757847533,
+      "grad_norm": 0.14785386490221808,
+      "learning_rate": 0.00010746268656716419,
+      "loss": 0.3523,
+      "step": 288
+    },
+    {
+      "epoch": 0.16199551569506726,
+      "grad_norm": 0.16121360229564613,
+      "learning_rate": 0.0001078358208955224,
+      "loss": 0.3532,
+      "step": 289
+    },
+    {
+      "epoch": 0.1625560538116592,
+      "grad_norm": 0.1666278996368264,
+      "learning_rate": 0.00010820895522388059,
+      "loss": 0.3344,
+      "step": 290
+    },
+    {
+      "epoch": 0.16311659192825112,
+      "grad_norm": 0.1658946129973587,
+      "learning_rate": 0.0001085820895522388,
+      "loss": 0.3619,
+      "step": 291
+    },
+    {
+      "epoch": 0.16367713004484305,
+      "grad_norm": 0.15391246174723924,
+      "learning_rate": 0.00010895522388059702,
+      "loss": 0.3437,
+      "step": 292
+    },
+    {
+      "epoch": 0.16423766816143498,
+      "grad_norm": 0.17369812945799734,
+      "learning_rate": 0.00010932835820895524,
+      "loss": 0.3435,
+      "step": 293
+    },
+    {
+      "epoch": 0.1647982062780269,
+      "grad_norm": 0.16097988268775634,
+      "learning_rate": 0.00010970149253731344,
+      "loss": 0.3458,
+      "step": 294
+    },
+    {
+      "epoch": 0.16535874439461884,
+      "grad_norm": 0.14845524173879637,
+      "learning_rate": 0.00011007462686567165,
+      "loss": 0.3617,
+      "step": 295
+    },
+    {
+      "epoch": 0.16591928251121077,
+      "grad_norm": 0.16204126238942138,
+      "learning_rate": 0.00011044776119402987,
+      "loss": 0.3639,
+      "step": 296
+    },
+    {
+      "epoch": 0.1664798206278027,
+      "grad_norm": 0.15918335661791597,
+      "learning_rate": 0.00011082089552238806,
+      "loss": 0.3556,
+      "step": 297
+    },
+    {
+      "epoch": 0.16704035874439463,
+      "grad_norm": 0.1508624404514423,
+      "learning_rate": 0.00011119402985074627,
+      "loss": 0.3331,
+      "step": 298
+    },
+    {
+      "epoch": 0.16760089686098656,
+      "grad_norm": 0.15001100888742291,
+      "learning_rate": 0.00011156716417910449,
+      "loss": 0.3336,
+      "step": 299
+    },
+    {
+      "epoch": 0.1681614349775785,
+      "grad_norm": 0.1565482297002879,
+      "learning_rate": 0.00011194029850746269,
+      "loss": 0.3475,
+      "step": 300
+    },
+    {
+      "epoch": 0.1687219730941704,
+      "grad_norm": 0.16549120599034722,
+      "learning_rate": 0.0001123134328358209,
+      "loss": 0.3558,
+      "step": 301
+    },
+    {
+      "epoch": 0.16928251121076232,
+      "grad_norm": 0.1505863588805418,
+      "learning_rate": 0.00011268656716417912,
+      "loss": 0.3664,
+      "step": 302
+    },
+    {
+      "epoch": 0.16984304932735425,
+      "grad_norm": 0.15323937742218322,
+      "learning_rate": 0.00011305970149253733,
+      "loss": 0.3552,
+      "step": 303
+    },
+    {
+      "epoch": 0.17040358744394618,
+      "grad_norm": 0.15854127887583344,
+      "learning_rate": 0.00011343283582089552,
+      "loss": 0.358,
+      "step": 304
+    },
+    {
+      "epoch": 0.1709641255605381,
+      "grad_norm": 0.16425026656892816,
+      "learning_rate": 0.00011380597014925374,
+      "loss": 0.3385,
+      "step": 305
+    },
+    {
+      "epoch": 0.17152466367713004,
+      "grad_norm": 0.14453136672166597,
+      "learning_rate": 0.00011417910447761195,
+      "loss": 0.3485,
+      "step": 306
+    },
+    {
+      "epoch": 0.17208520179372197,
+      "grad_norm": 0.14521817335331513,
+      "learning_rate": 0.00011455223880597015,
+      "loss": 0.3379,
+      "step": 307
+    },
+    {
+      "epoch": 0.1726457399103139,
+      "grad_norm": 0.15447478262387904,
+      "learning_rate": 0.00011492537313432837,
+      "loss": 0.3401,
+      "step": 308
+    },
+    {
+      "epoch": 0.17320627802690583,
+      "grad_norm": 0.14543845819057685,
+      "learning_rate": 0.00011529850746268658,
+      "loss": 0.3531,
+      "step": 309
+    },
+    {
+      "epoch": 0.17376681614349776,
+      "grad_norm": 0.14848515874034607,
+      "learning_rate": 0.00011567164179104477,
+      "loss": 0.3408,
+      "step": 310
+    },
+    {
+      "epoch": 0.1743273542600897,
+      "grad_norm": 0.1468921528006467,
+      "learning_rate": 0.00011604477611940299,
+      "loss": 0.3539,
+      "step": 311
+    },
+    {
+      "epoch": 0.17488789237668162,
+      "grad_norm": 0.14226517621252568,
+      "learning_rate": 0.0001164179104477612,
+      "loss": 0.3406,
+      "step": 312
+    },
+    {
+      "epoch": 0.17544843049327355,
+      "grad_norm": 0.14388093245906539,
+      "learning_rate": 0.00011679104477611942,
+      "loss": 0.3419,
+      "step": 313
+    },
+    {
+      "epoch": 0.17600896860986548,
+      "grad_norm": 0.1447233515128136,
+      "learning_rate": 0.0001171641791044776,
+      "loss": 0.336,
+      "step": 314
+    },
+    {
+      "epoch": 0.1765695067264574,
+      "grad_norm": 0.1455324803503982,
+      "learning_rate": 0.00011753731343283582,
+      "loss": 0.3465,
+      "step": 315
+    },
+    {
+      "epoch": 0.17713004484304934,
+      "grad_norm": 0.14770033589219328,
+      "learning_rate": 0.00011791044776119405,
+      "loss": 0.3405,
+      "step": 316
+    },
+    {
+      "epoch": 0.17769058295964127,
+      "grad_norm": 0.13999832382983,
+      "learning_rate": 0.00011828358208955224,
+      "loss": 0.3469,
+      "step": 317
+    },
+    {
+      "epoch": 0.17825112107623317,
+      "grad_norm": 0.13762285639869806,
+      "learning_rate": 0.00011865671641791045,
+      "loss": 0.3416,
+      "step": 318
+    },
+    {
+      "epoch": 0.1788116591928251,
+      "grad_norm": 0.14693857139845815,
+      "learning_rate": 0.00011902985074626867,
+      "loss": 0.3411,
+      "step": 319
+    },
+    {
+      "epoch": 0.17937219730941703,
+      "grad_norm": 0.1508866834864334,
+      "learning_rate": 0.00011940298507462686,
+      "loss": 0.3629,
+      "step": 320
+    },
+    {
+      "epoch": 0.17993273542600896,
+      "grad_norm": 0.14654222377365972,
+      "learning_rate": 0.00011977611940298507,
+      "loss": 0.3385,
+      "step": 321
+    },
+    {
+      "epoch": 0.1804932735426009,
+      "grad_norm": 0.14817029949391497,
+      "learning_rate": 0.00012014925373134329,
+      "loss": 0.3187,
+      "step": 322
+    },
+    {
+      "epoch": 0.18105381165919282,
+      "grad_norm": 0.14496977911609008,
+      "learning_rate": 0.0001205223880597015,
+      "loss": 0.334,
+      "step": 323
+    },
+    {
+      "epoch": 0.18161434977578475,
+      "grad_norm": 0.14616862773211267,
+      "learning_rate": 0.0001208955223880597,
+      "loss": 0.3353,
+      "step": 324
+    },
+    {
+      "epoch": 0.18217488789237668,
+      "grad_norm": 0.139574242854469,
+      "learning_rate": 0.00012126865671641792,
+      "loss": 0.3444,
+      "step": 325
+    },
+    {
+      "epoch": 0.1827354260089686,
+      "grad_norm": 0.14678711716707,
+      "learning_rate": 0.00012164179104477613,
+      "loss": 0.3406,
+      "step": 326
+    },
+    {
+      "epoch": 0.18329596412556054,
+      "grad_norm": 0.1383242188369039,
+      "learning_rate": 0.00012201492537313432,
+      "loss": 0.3462,
+      "step": 327
+    },
+    {
+      "epoch": 0.18385650224215247,
+      "grad_norm": 0.14074334987634868,
+      "learning_rate": 0.00012238805970149255,
+      "loss": 0.34,
+      "step": 328
+    },
+    {
+      "epoch": 0.1844170403587444,
+      "grad_norm": 0.13221503607649598,
+      "learning_rate": 0.00012276119402985077,
+      "loss": 0.3382,
+      "step": 329
+    },
+    {
+      "epoch": 0.18497757847533633,
+      "grad_norm": 0.14504252934701425,
+      "learning_rate": 0.00012313432835820895,
+      "loss": 0.3356,
+      "step": 330
+    },
+    {
+      "epoch": 0.18553811659192826,
+      "grad_norm": 0.14189066468483738,
+      "learning_rate": 0.00012350746268656717,
+      "loss": 0.3543,
+      "step": 331
+    },
+    {
+      "epoch": 0.1860986547085202,
+      "grad_norm": 0.14575030702579406,
+      "learning_rate": 0.00012388059701492538,
+      "loss": 0.3327,
+      "step": 332
+    },
+    {
+      "epoch": 0.18665919282511212,
+      "grad_norm": 0.1518908122266155,
+      "learning_rate": 0.0001242537313432836,
+      "loss": 0.3427,
+      "step": 333
+    },
+    {
+      "epoch": 0.18721973094170405,
+      "grad_norm": 0.13073540683964113,
+      "learning_rate": 0.0001246268656716418,
+      "loss": 0.3323,
+      "step": 334
+    },
+    {
+      "epoch": 0.18778026905829595,
+      "grad_norm": 0.14654813791251525,
+      "learning_rate": 0.000125,
+      "loss": 0.3513,
+      "step": 335
+    },
+    {
+      "epoch": 0.18834080717488788,
+      "grad_norm": 0.14395303052531452,
+      "learning_rate": 0.00012537313432835822,
+      "loss": 0.3563,
+      "step": 336
+    },
+    {
+      "epoch": 0.1889013452914798,
+      "grad_norm": 0.1361516821315323,
+      "learning_rate": 0.0001257462686567164,
+      "loss": 0.3263,
+      "step": 337
+    },
+    {
+      "epoch": 0.18946188340807174,
+      "grad_norm": 0.1419463217755675,
+      "learning_rate": 0.00012611940298507462,
+      "loss": 0.3297,
+      "step": 338
+    },
+    {
+      "epoch": 0.19002242152466367,
+      "grad_norm": 0.1655300104293705,
+      "learning_rate": 0.00012649253731343284,
+      "loss": 0.3393,
+      "step": 339
+    },
+    {
+      "epoch": 0.1905829596412556,
+      "grad_norm": 0.15281292173494365,
+      "learning_rate": 0.00012686567164179105,
+      "loss": 0.3433,
+      "step": 340
+    },
+    {
+      "epoch": 0.19114349775784753,
+      "grad_norm": 0.16592140919555418,
+      "learning_rate": 0.00012723880597014927,
+      "loss": 0.3285,
+      "step": 341
+    },
+    {
+      "epoch": 0.19170403587443946,
+      "grad_norm": 0.14634846678856345,
+      "learning_rate": 0.00012761194029850748,
+      "loss": 0.3353,
+      "step": 342
+    },
+    {
+      "epoch": 0.1922645739910314,
+      "grad_norm": 0.14725326795722657,
+      "learning_rate": 0.0001279850746268657,
+      "loss": 0.347,
+      "step": 343
+    },
+    {
+      "epoch": 0.19282511210762332,
+      "grad_norm": 0.1410427654180543,
+      "learning_rate": 0.00012835820895522389,
+      "loss": 0.348,
+      "step": 344
+    },
+    {
+      "epoch": 0.19338565022421525,
+      "grad_norm": 0.14665245126527168,
+      "learning_rate": 0.0001287313432835821,
+      "loss": 0.3442,
+      "step": 345
+    },
+    {
+      "epoch": 0.19394618834080718,
+      "grad_norm": 0.149124223155523,
+      "learning_rate": 0.00012910447761194032,
+      "loss": 0.3358,
+      "step": 346
+    },
+    {
+      "epoch": 0.1945067264573991,
+      "grad_norm": 0.1395297425768304,
+      "learning_rate": 0.0001294776119402985,
+      "loss": 0.3265,
+      "step": 347
+    },
+    {
+      "epoch": 0.19506726457399104,
+      "grad_norm": 0.15476619855626284,
+      "learning_rate": 0.00012985074626865672,
+      "loss": 0.342,
+      "step": 348
+    },
+    {
+      "epoch": 0.19562780269058297,
+      "grad_norm": 0.14943977200050382,
+      "learning_rate": 0.00013022388059701493,
+      "loss": 0.347,
+      "step": 349
+    },
+    {
+      "epoch": 0.1961883408071749,
+      "grad_norm": 0.13087846809773412,
+      "learning_rate": 0.00013059701492537315,
+      "loss": 0.3273,
+      "step": 350
+    },
+    {
+      "epoch": 0.19674887892376683,
+      "grad_norm": 0.14118185286815782,
+      "learning_rate": 0.00013097014925373134,
+      "loss": 0.3314,
+      "step": 351
+    },
+    {
+      "epoch": 0.19730941704035873,
+      "grad_norm": 0.1486868542745911,
+      "learning_rate": 0.00013134328358208955,
+      "loss": 0.3327,
+      "step": 352
+    },
+    {
+      "epoch": 0.19786995515695066,
+      "grad_norm": 0.13026695041967967,
+      "learning_rate": 0.00013171641791044777,
+      "loss": 0.3339,
+      "step": 353
+    },
+    {
+      "epoch": 0.1984304932735426,
+      "grad_norm": 0.14206153326263493,
+      "learning_rate": 0.00013208955223880596,
+      "loss": 0.3428,
+      "step": 354
+    },
+    {
+      "epoch": 0.19899103139013452,
+      "grad_norm": 0.14564897855047862,
+      "learning_rate": 0.0001324626865671642,
+      "loss": 0.3502,
+      "step": 355
+    },
+    {
+      "epoch": 0.19955156950672645,
+      "grad_norm": 0.13257702188973255,
+      "learning_rate": 0.0001328358208955224,
+      "loss": 0.3417,
+      "step": 356
+    },
+    {
+      "epoch": 0.20011210762331838,
+      "grad_norm": 0.12707052181073036,
+      "learning_rate": 0.0001332089552238806,
+      "loss": 0.3379,
+      "step": 357
+    },
+    {
+      "epoch": 0.2006726457399103,
+      "grad_norm": 0.14558961583149116,
+      "learning_rate": 0.00013358208955223882,
+      "loss": 0.3446,
+      "step": 358
+    },
+    {
+      "epoch": 0.20123318385650224,
+      "grad_norm": 0.13403130997892224,
+      "learning_rate": 0.00013395522388059703,
+      "loss": 0.3386,
+      "step": 359
+    },
+    {
+      "epoch": 0.20179372197309417,
+      "grad_norm": 0.1383693750622578,
+      "learning_rate": 0.00013432835820895525,
+      "loss": 0.3294,
+      "step": 360
+    },
+    {
+      "epoch": 0.2023542600896861,
+      "grad_norm": 0.14914836437591436,
+      "learning_rate": 0.00013470149253731343,
+      "loss": 0.336,
+      "step": 361
+    },
+    {
+      "epoch": 0.20291479820627803,
+      "grad_norm": 0.14807475904717177,
+      "learning_rate": 0.00013507462686567165,
+      "loss": 0.3458,
+      "step": 362
+    },
+    {
+      "epoch": 0.20347533632286996,
+      "grad_norm": 0.14607368191149564,
+      "learning_rate": 0.00013544776119402987,
+      "loss": 0.3418,
+      "step": 363
+    },
+    {
+      "epoch": 0.2040358744394619,
+      "grad_norm": 0.14366790334710722,
+      "learning_rate": 0.00013582089552238805,
+      "loss": 0.3274,
+      "step": 364
+    },
+    {
+      "epoch": 0.20459641255605382,
+      "grad_norm": 0.13680550785389592,
+      "learning_rate": 0.00013619402985074627,
+      "loss": 0.3123,
+      "step": 365
+    },
+    {
+      "epoch": 0.20515695067264575,
+      "grad_norm": 0.13993820313914998,
+      "learning_rate": 0.00013656716417910448,
+      "loss": 0.3382,
+      "step": 366
+    },
+    {
+      "epoch": 0.20571748878923768,
+      "grad_norm": 0.13377861071671815,
+      "learning_rate": 0.00013694029850746267,
+      "loss": 0.33,
+      "step": 367
+    },
+    {
+      "epoch": 0.2062780269058296,
+      "grad_norm": 0.13704463750464957,
+      "learning_rate": 0.0001373134328358209,
+      "loss": 0.3523,
+      "step": 368
+    },
+    {
+      "epoch": 0.2068385650224215,
+      "grad_norm": 0.13513020711651452,
+      "learning_rate": 0.0001376865671641791,
+      "loss": 0.3416,
+      "step": 369
+    },
+    {
+      "epoch": 0.20739910313901344,
+      "grad_norm": 0.14131252420969243,
+      "learning_rate": 0.00013805970149253732,
+      "loss": 0.3316,
+      "step": 370
+    },
+    {
+      "epoch": 0.20795964125560537,
+      "grad_norm": 0.15006108842697338,
+      "learning_rate": 0.00013843283582089553,
+      "loss": 0.3319,
+      "step": 371
+    },
+    {
+      "epoch": 0.2085201793721973,
+      "grad_norm": 0.13713602623692356,
+      "learning_rate": 0.00013880597014925375,
+      "loss": 0.3334,
+      "step": 372
+    },
+    {
+      "epoch": 0.20908071748878923,
+      "grad_norm": 0.14038281003520495,
+      "learning_rate": 0.00013917910447761196,
+      "loss": 0.3459,
+      "step": 373
+    },
+    {
+      "epoch": 0.20964125560538116,
+      "grad_norm": 0.12250736037089766,
+      "learning_rate": 0.00013955223880597015,
+      "loss": 0.3432,
+      "step": 374
+    },
+    {
+      "epoch": 0.2102017937219731,
+      "grad_norm": 0.12958496353681398,
+      "learning_rate": 0.00013992537313432837,
+      "loss": 0.3406,
+      "step": 375
+    },
+    {
+      "epoch": 0.21076233183856502,
+      "grad_norm": 0.12931345924298085,
+      "learning_rate": 0.00014029850746268658,
+      "loss": 0.344,
+      "step": 376
+    },
+    {
+      "epoch": 0.21132286995515695,
+      "grad_norm": 0.14903529712216704,
+      "learning_rate": 0.00014067164179104477,
+      "loss": 0.3318,
+      "step": 377
+    },
+    {
+      "epoch": 0.21188340807174888,
+      "grad_norm": 0.13170430329554297,
+      "learning_rate": 0.00014104477611940298,
+      "loss": 0.3317,
+      "step": 378
+    },
+    {
+      "epoch": 0.2124439461883408,
+      "grad_norm": 0.12667419201802455,
+      "learning_rate": 0.0001414179104477612,
+      "loss": 0.3273,
+      "step": 379
+    },
+    {
+      "epoch": 0.21300448430493274,
+      "grad_norm": 0.13178478862435777,
+      "learning_rate": 0.00014179104477611942,
+      "loss": 0.3381,
+      "step": 380
+    },
+    {
+      "epoch": 0.21356502242152467,
+      "grad_norm": 0.1291010897335178,
+      "learning_rate": 0.0001421641791044776,
+      "loss": 0.3375,
+      "step": 381
+    },
+    {
+      "epoch": 0.2141255605381166,
+      "grad_norm": 0.13394707835363362,
+      "learning_rate": 0.00014253731343283582,
+      "loss": 0.3326,
+      "step": 382
+    },
+    {
+      "epoch": 0.21468609865470853,
+      "grad_norm": 0.12853384755084213,
+      "learning_rate": 0.00014291044776119403,
+      "loss": 0.345,
+      "step": 383
+    },
+    {
+      "epoch": 0.21524663677130046,
+      "grad_norm": 0.12994048579956208,
+      "learning_rate": 0.00014328358208955225,
+      "loss": 0.3156,
+      "step": 384
+    },
+    {
+      "epoch": 0.2158071748878924,
+      "grad_norm": 0.1368026539480669,
+      "learning_rate": 0.00014365671641791046,
+      "loss": 0.3342,
+      "step": 385
+    },
+    {
+      "epoch": 0.2163677130044843,
+      "grad_norm": 0.1380109557323045,
+      "learning_rate": 0.00014402985074626868,
+      "loss": 0.3292,
+      "step": 386
+    },
+    {
+      "epoch": 0.21692825112107622,
+      "grad_norm": 0.13514092013237952,
+      "learning_rate": 0.00014440298507462687,
+      "loss": 0.3285,
+      "step": 387
+    },
+    {
+      "epoch": 0.21748878923766815,
+      "grad_norm": 0.12733082443982355,
+      "learning_rate": 0.00014477611940298508,
+      "loss": 0.3315,
+      "step": 388
+    },
+    {
+      "epoch": 0.21804932735426008,
+      "grad_norm": 0.1294327190388571,
+      "learning_rate": 0.0001451492537313433,
+      "loss": 0.3373,
+      "step": 389
+    },
+    {
+      "epoch": 0.218609865470852,
+      "grad_norm": 0.12943400016143503,
+      "learning_rate": 0.0001455223880597015,
+      "loss": 0.3286,
+      "step": 390
+    },
+    {
+      "epoch": 0.21917040358744394,
+      "grad_norm": 0.12626830903738212,
+      "learning_rate": 0.0001458955223880597,
+      "loss": 0.3448,
+      "step": 391
+    },
+    {
+      "epoch": 0.21973094170403587,
+      "grad_norm": 0.12459201714786171,
+      "learning_rate": 0.00014626865671641792,
+      "loss": 0.3314,
+      "step": 392
+    },
+    {
+      "epoch": 0.2202914798206278,
+      "grad_norm": 0.12647395067356484,
+      "learning_rate": 0.00014664179104477613,
+      "loss": 0.3395,
+      "step": 393
+    },
+    {
+      "epoch": 0.22085201793721973,
+      "grad_norm": 0.12867348213574867,
+      "learning_rate": 0.00014701492537313432,
+      "loss": 0.3423,
+      "step": 394
+    },
+    {
+      "epoch": 0.22141255605381166,
+      "grad_norm": 0.1221176661464511,
+      "learning_rate": 0.00014738805970149253,
+      "loss": 0.3301,
+      "step": 395
+    },
+    {
+      "epoch": 0.2219730941704036,
+      "grad_norm": 0.1372287955348562,
+      "learning_rate": 0.00014776119402985075,
+      "loss": 0.343,
+      "step": 396
+    },
+    {
+      "epoch": 0.22253363228699552,
+      "grad_norm": 0.13066443642622122,
+      "learning_rate": 0.00014813432835820894,
+      "loss": 0.3337,
+      "step": 397
+    },
+    {
+      "epoch": 0.22309417040358745,
+      "grad_norm": 0.12841879469681047,
+      "learning_rate": 0.00014850746268656718,
+      "loss": 0.323,
+      "step": 398
+    },
+    {
+      "epoch": 0.22365470852017938,
+      "grad_norm": 0.12639845139011982,
+      "learning_rate": 0.0001488805970149254,
+      "loss": 0.3181,
+      "step": 399
+    },
+    {
+      "epoch": 0.2242152466367713,
+      "grad_norm": 0.13152331106311524,
+      "learning_rate": 0.0001492537313432836,
+      "loss": 0.3311,
+      "step": 400
+    },
+    {
+      "epoch": 0.22477578475336324,
+      "grad_norm": 0.1257243364510895,
+      "learning_rate": 0.0001496268656716418,
+      "loss": 0.3473,
+      "step": 401
+    },
+    {
+      "epoch": 0.22533632286995517,
+      "grad_norm": 0.11961559661149512,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.3302,
+      "step": 402
+    },
+    {
+      "epoch": 0.2258968609865471,
+      "grad_norm": 0.1285158772976168,
+      "learning_rate": 0.00015037313432835823,
+      "loss": 0.3303,
+      "step": 403
+    },
+    {
+      "epoch": 0.226457399103139,
+      "grad_norm": 0.12700948859909972,
+      "learning_rate": 0.00015074626865671642,
+      "loss": 0.346,
+      "step": 404
+    },
+    {
+      "epoch": 0.22701793721973093,
+      "grad_norm": 0.12493785729680966,
+      "learning_rate": 0.00015111940298507463,
+      "loss": 0.3228,
+      "step": 405
+    },
+    {
+      "epoch": 0.22757847533632286,
+      "grad_norm": 0.1387796666197342,
+      "learning_rate": 0.00015149253731343285,
+      "loss": 0.3419,
+      "step": 406
+    },
+    {
+      "epoch": 0.2281390134529148,
+      "grad_norm": 0.12168730216608055,
+      "learning_rate": 0.00015186567164179106,
+      "loss": 0.3269,
+      "step": 407
+    },
+    {
+      "epoch": 0.22869955156950672,
+      "grad_norm": 0.12054043851765807,
+      "learning_rate": 0.00015223880597014925,
+      "loss": 0.3417,
+      "step": 408
+    },
+    {
+      "epoch": 0.22926008968609865,
+      "grad_norm": 0.1262020840793984,
+      "learning_rate": 0.00015261194029850747,
+      "loss": 0.3351,
+      "step": 409
+    },
+    {
+      "epoch": 0.22982062780269058,
+      "grad_norm": 0.12646633604287688,
+      "learning_rate": 0.00015298507462686568,
+      "loss": 0.332,
+      "step": 410
+    },
+    {
+      "epoch": 0.2303811659192825,
+      "grad_norm": 0.12678814574552924,
+      "learning_rate": 0.00015335820895522387,
+      "loss": 0.3274,
+      "step": 411
+    },
+    {
+      "epoch": 0.23094170403587444,
+      "grad_norm": 0.1247749319102667,
+      "learning_rate": 0.00015373134328358208,
+      "loss": 0.3293,
+      "step": 412
+    },
+    {
+      "epoch": 0.23150224215246637,
+      "grad_norm": 0.1319512977808111,
+      "learning_rate": 0.0001541044776119403,
+      "loss": 0.3475,
+      "step": 413
+    },
+    {
+      "epoch": 0.2320627802690583,
+      "grad_norm": 0.1291927456921302,
+      "learning_rate": 0.00015447761194029851,
+      "loss": 0.3318,
+      "step": 414
+    },
+    {
+      "epoch": 0.23262331838565023,
+      "grad_norm": 0.1326160886521916,
+      "learning_rate": 0.00015485074626865673,
+      "loss": 0.3398,
+      "step": 415
+    },
+    {
+      "epoch": 0.23318385650224216,
+      "grad_norm": 0.13783299680690347,
+      "learning_rate": 0.00015522388059701495,
+      "loss": 0.3145,
+      "step": 416
+    },
+    {
+      "epoch": 0.2337443946188341,
+      "grad_norm": 0.14278625424021602,
+      "learning_rate": 0.00015559701492537316,
+      "loss": 0.3237,
+      "step": 417
+    },
+    {
+      "epoch": 0.23430493273542602,
+      "grad_norm": 0.1255560465560579,
+      "learning_rate": 0.00015597014925373135,
+      "loss": 0.318,
+      "step": 418
+    },
+    {
+      "epoch": 0.23486547085201795,
+      "grad_norm": 0.1293448524834811,
+      "learning_rate": 0.00015634328358208956,
+      "loss": 0.3244,
+      "step": 419
+    },
+    {
+      "epoch": 0.23542600896860988,
+      "grad_norm": 0.12125345580304638,
+      "learning_rate": 0.00015671641791044778,
+      "loss": 0.3269,
+      "step": 420
+    },
+    {
+      "epoch": 0.23598654708520178,
+      "grad_norm": 0.1301695815606069,
+      "learning_rate": 0.00015708955223880597,
+      "loss": 0.3452,
+      "step": 421
+    },
+    {
+      "epoch": 0.2365470852017937,
+      "grad_norm": 0.11543487301027802,
+      "learning_rate": 0.00015746268656716418,
+      "loss": 0.3108,
+      "step": 422
+    },
+    {
+      "epoch": 0.23710762331838564,
+      "grad_norm": 0.11808058684873649,
+      "learning_rate": 0.0001578358208955224,
+      "loss": 0.304,
+      "step": 423
+    },
+    {
+      "epoch": 0.23766816143497757,
+      "grad_norm": 0.11914598540750668,
+      "learning_rate": 0.00015820895522388059,
+      "loss": 0.3307,
+      "step": 424
+    },
+    {
+      "epoch": 0.2382286995515695,
+      "grad_norm": 0.11718635825048222,
+      "learning_rate": 0.0001585820895522388,
+      "loss": 0.3282,
+      "step": 425
+    },
+    {
+      "epoch": 0.23878923766816143,
+      "grad_norm": 0.12828602408853326,
+      "learning_rate": 0.00015895522388059702,
+      "loss": 0.344,
+      "step": 426
+    },
+    {
+      "epoch": 0.23934977578475336,
+      "grad_norm": 0.12241044394498629,
+      "learning_rate": 0.00015932835820895523,
+      "loss": 0.3263,
+      "step": 427
+    },
+    {
+      "epoch": 0.2399103139013453,
+      "grad_norm": 0.12628652174142493,
+      "learning_rate": 0.00015970149253731345,
+      "loss": 0.3193,
+      "step": 428
+    },
+    {
+      "epoch": 0.24047085201793722,
+      "grad_norm": 0.12350703208546007,
+      "learning_rate": 0.00016007462686567166,
+      "loss": 0.33,
+      "step": 429
+    },
+    {
+      "epoch": 0.24103139013452915,
+      "grad_norm": 0.1260052637488716,
+      "learning_rate": 0.00016044776119402988,
+      "loss": 0.3201,
+      "step": 430
+    },
+    {
+      "epoch": 0.24159192825112108,
+      "grad_norm": 0.1281992427766547,
+      "learning_rate": 0.00016082089552238806,
+      "loss": 0.3278,
+      "step": 431
+    },
+    {
+      "epoch": 0.242152466367713,
+      "grad_norm": 0.12262788847949457,
+      "learning_rate": 0.00016119402985074628,
+      "loss": 0.3321,
+      "step": 432
+    },
+    {
+      "epoch": 0.24271300448430494,
+      "grad_norm": 0.12259307695789733,
+      "learning_rate": 0.0001615671641791045,
+      "loss": 0.3334,
+      "step": 433
+    },
+    {
+      "epoch": 0.24327354260089687,
+      "grad_norm": 0.13187011912485422,
+      "learning_rate": 0.00016194029850746268,
+      "loss": 0.3377,
+      "step": 434
+    },
+    {
+      "epoch": 0.2438340807174888,
+      "grad_norm": 0.12047127318376988,
+      "learning_rate": 0.0001623134328358209,
+      "loss": 0.3327,
+      "step": 435
+    },
+    {
+      "epoch": 0.24439461883408073,
+      "grad_norm": 0.12254124825937357,
+      "learning_rate": 0.00016268656716417911,
+      "loss": 0.3251,
+      "step": 436
+    },
+    {
+      "epoch": 0.24495515695067266,
+      "grad_norm": 0.11229576952987816,
+      "learning_rate": 0.00016305970149253733,
+      "loss": 0.3247,
+      "step": 437
+    },
+    {
+      "epoch": 0.24551569506726456,
+      "grad_norm": 0.11316247053608615,
+      "learning_rate": 0.00016343283582089552,
+      "loss": 0.3362,
+      "step": 438
+    },
+    {
+      "epoch": 0.2460762331838565,
+      "grad_norm": 0.12670848030646706,
+      "learning_rate": 0.00016380597014925373,
+      "loss": 0.3324,
+      "step": 439
+    },
+    {
+      "epoch": 0.24663677130044842,
+      "grad_norm": 0.11407342067194146,
+      "learning_rate": 0.00016417910447761195,
+      "loss": 0.33,
+      "step": 440
+    },
+    {
+      "epoch": 0.24719730941704035,
+      "grad_norm": 0.12686252667414058,
+      "learning_rate": 0.00016455223880597016,
+      "loss": 0.321,
+      "step": 441
+    },
+    {
+      "epoch": 0.24775784753363228,
+      "grad_norm": 0.12257208306795794,
+      "learning_rate": 0.00016492537313432838,
+      "loss": 0.3359,
+      "step": 442
+    },
+    {
+      "epoch": 0.2483183856502242,
+      "grad_norm": 0.12416215291655391,
+      "learning_rate": 0.0001652985074626866,
+      "loss": 0.3391,
+      "step": 443
+    },
+    {
+      "epoch": 0.24887892376681614,
+      "grad_norm": 0.11842325228755514,
+      "learning_rate": 0.00016567164179104478,
+      "loss": 0.3228,
+      "step": 444
+    },
+    {
+      "epoch": 0.24943946188340807,
+      "grad_norm": 0.11557531510118868,
+      "learning_rate": 0.000166044776119403,
+      "loss": 0.3307,
+      "step": 445
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.12170820760450851,
+      "learning_rate": 0.0001664179104477612,
+      "loss": 0.3482,
+      "step": 446
+    },
+    {
+      "epoch": 0.2505605381165919,
+      "grad_norm": 0.1293312049559174,
+      "learning_rate": 0.00016679104477611943,
+      "loss": 0.3348,
+      "step": 447
+    },
+    {
+      "epoch": 0.25112107623318386,
+      "grad_norm": 0.1129033052648288,
+      "learning_rate": 0.00016716417910447761,
+      "loss": 0.3312,
+      "step": 448
+    },
+    {
+      "epoch": 0.25168161434977576,
+      "grad_norm": 0.1215834233409738,
+      "learning_rate": 0.00016753731343283583,
+      "loss": 0.3255,
+      "step": 449
+    },
+    {
+      "epoch": 0.2522421524663677,
+      "grad_norm": 0.12180545876740287,
+      "learning_rate": 0.00016791044776119405,
+      "loss": 0.3339,
+      "step": 450
+    },
+    {
+      "epoch": 0.2528026905829596,
+      "grad_norm": 0.11871718449220875,
+      "learning_rate": 0.00016828358208955223,
+      "loss": 0.3342,
+      "step": 451
+    },
+    {
+      "epoch": 0.2533632286995516,
+      "grad_norm": 0.10955442215647969,
+      "learning_rate": 0.00016865671641791045,
+      "loss": 0.3183,
+      "step": 452
+    },
+    {
+      "epoch": 0.2539237668161435,
+      "grad_norm": 0.12141904390740511,
+      "learning_rate": 0.00016902985074626866,
+      "loss": 0.3338,
+      "step": 453
+    },
+    {
+      "epoch": 0.25448430493273544,
+      "grad_norm": 0.109363161276611,
+      "learning_rate": 0.00016940298507462685,
+      "loss": 0.3275,
+      "step": 454
+    },
+    {
+      "epoch": 0.25504484304932734,
+      "grad_norm": 0.10861033736457545,
+      "learning_rate": 0.00016977611940298507,
+      "loss": 0.3202,
+      "step": 455
+    },
+    {
+      "epoch": 0.2556053811659193,
+      "grad_norm": 0.11701255773856319,
+      "learning_rate": 0.00017014925373134328,
+      "loss": 0.3259,
+      "step": 456
+    },
+    {
+      "epoch": 0.2561659192825112,
+      "grad_norm": 0.1237391208716396,
+      "learning_rate": 0.0001705223880597015,
+      "loss": 0.3367,
+      "step": 457
+    },
+    {
+      "epoch": 0.25672645739910316,
+      "grad_norm": 0.12193150919073976,
+      "learning_rate": 0.0001708955223880597,
+      "loss": 0.3064,
+      "step": 458
+    },
+    {
+      "epoch": 0.25728699551569506,
+      "grad_norm": 0.11247330351826199,
+      "learning_rate": 0.00017126865671641793,
+      "loss": 0.337,
+      "step": 459
+    },
+    {
+      "epoch": 0.257847533632287,
+      "grad_norm": 0.12608345820517644,
+      "learning_rate": 0.00017164179104477614,
+      "loss": 0.3307,
+      "step": 460
+    },
+    {
+      "epoch": 0.2584080717488789,
+      "grad_norm": 0.10981245125607718,
+      "learning_rate": 0.00017201492537313433,
+      "loss": 0.3249,
+      "step": 461
+    },
+    {
+      "epoch": 0.2589686098654709,
+      "grad_norm": 0.12160721153034673,
+      "learning_rate": 0.00017238805970149255,
+      "loss": 0.3148,
+      "step": 462
+    },
+    {
+      "epoch": 0.2595291479820628,
+      "grad_norm": 0.11499630830458173,
+      "learning_rate": 0.00017276119402985076,
+      "loss": 0.3247,
+      "step": 463
+    },
+    {
+      "epoch": 0.2600896860986547,
+      "grad_norm": 0.11773276443927631,
+      "learning_rate": 0.00017313432835820895,
+      "loss": 0.3409,
+      "step": 464
+    },
+    {
+      "epoch": 0.26065022421524664,
+      "grad_norm": 0.1163894352235071,
+      "learning_rate": 0.00017350746268656716,
+      "loss": 0.3312,
+      "step": 465
+    },
+    {
+      "epoch": 0.26121076233183854,
+      "grad_norm": 0.11990564012827774,
+      "learning_rate": 0.00017388059701492538,
+      "loss": 0.3253,
+      "step": 466
+    },
+    {
+      "epoch": 0.2617713004484305,
+      "grad_norm": 0.11171211187369197,
+      "learning_rate": 0.0001742537313432836,
+      "loss": 0.3254,
+      "step": 467
+    },
+    {
+      "epoch": 0.2623318385650224,
+      "grad_norm": 0.11139288133558127,
+      "learning_rate": 0.00017462686567164178,
+      "loss": 0.3384,
+      "step": 468
+    },
+    {
+      "epoch": 0.26289237668161436,
+      "grad_norm": 0.10673322543340892,
+      "learning_rate": 0.000175,
+      "loss": 0.3214,
+      "step": 469
+    },
+    {
+      "epoch": 0.26345291479820626,
+      "grad_norm": 0.11082988585060753,
+      "learning_rate": 0.0001753731343283582,
+      "loss": 0.3326,
+      "step": 470
+    },
+    {
+      "epoch": 0.2640134529147982,
+      "grad_norm": 0.11109033950550597,
+      "learning_rate": 0.00017574626865671643,
+      "loss": 0.3234,
+      "step": 471
+    },
+    {
+      "epoch": 0.2645739910313901,
+      "grad_norm": 0.10788131857393239,
+      "learning_rate": 0.00017611940298507464,
+      "loss": 0.3144,
+      "step": 472
+    },
+    {
+      "epoch": 0.2651345291479821,
+      "grad_norm": 0.11526598585641781,
+      "learning_rate": 0.00017649253731343286,
+      "loss": 0.3208,
+      "step": 473
+    },
+    {
+      "epoch": 0.265695067264574,
+      "grad_norm": 0.12174354420730553,
+      "learning_rate": 0.00017686567164179107,
+      "loss": 0.3173,
+      "step": 474
+    },
+    {
+      "epoch": 0.26625560538116594,
+      "grad_norm": 0.11021769520247941,
+      "learning_rate": 0.00017723880597014926,
+      "loss": 0.3194,
+      "step": 475
+    },
+    {
+      "epoch": 0.26681614349775784,
+      "grad_norm": 0.11133392006178908,
+      "learning_rate": 0.00017761194029850748,
+      "loss": 0.318,
+      "step": 476
+    },
+    {
+      "epoch": 0.2673766816143498,
+      "grad_norm": 0.11254586812321232,
+      "learning_rate": 0.0001779850746268657,
+      "loss": 0.3293,
+      "step": 477
+    },
+    {
+      "epoch": 0.2679372197309417,
+      "grad_norm": 0.11183122193309668,
+      "learning_rate": 0.00017835820895522388,
+      "loss": 0.3368,
+      "step": 478
+    },
+    {
+      "epoch": 0.26849775784753366,
+      "grad_norm": 0.10711717418436024,
+      "learning_rate": 0.0001787313432835821,
+      "loss": 0.3153,
+      "step": 479
+    },
+    {
+      "epoch": 0.26905829596412556,
+      "grad_norm": 0.11448137615640368,
+      "learning_rate": 0.0001791044776119403,
+      "loss": 0.3471,
+      "step": 480
+    },
+    {
+      "epoch": 0.26961883408071746,
+      "grad_norm": 0.11914801671735352,
+      "learning_rate": 0.0001794776119402985,
+      "loss": 0.3223,
+      "step": 481
+    },
+    {
+      "epoch": 0.2701793721973094,
+      "grad_norm": 0.11137282750934051,
+      "learning_rate": 0.00017985074626865671,
+      "loss": 0.3211,
+      "step": 482
+    },
+    {
+      "epoch": 0.2707399103139013,
+      "grad_norm": 0.10832969827373858,
+      "learning_rate": 0.00018022388059701493,
+      "loss": 0.3085,
+      "step": 483
+    },
+    {
+      "epoch": 0.2713004484304933,
+      "grad_norm": 0.11085902902463791,
+      "learning_rate": 0.00018059701492537314,
+      "loss": 0.3175,
+      "step": 484
+    },
+    {
+      "epoch": 0.2718609865470852,
+      "grad_norm": 0.11819460294134253,
+      "learning_rate": 0.00018097014925373136,
+      "loss": 0.3248,
+      "step": 485
+    },
+    {
+      "epoch": 0.27242152466367714,
+      "grad_norm": 0.11313174097954944,
+      "learning_rate": 0.00018134328358208958,
+      "loss": 0.3199,
+      "step": 486
+    },
+    {
+      "epoch": 0.27298206278026904,
+      "grad_norm": 0.12095925240596792,
+      "learning_rate": 0.0001817164179104478,
+      "loss": 0.3225,
+      "step": 487
+    },
+    {
+      "epoch": 0.273542600896861,
+      "grad_norm": 0.11014289563863666,
+      "learning_rate": 0.00018208955223880598,
+      "loss": 0.3286,
+      "step": 488
+    },
+    {
+      "epoch": 0.2741031390134529,
+      "grad_norm": 0.11204674994833956,
+      "learning_rate": 0.0001824626865671642,
+      "loss": 0.3108,
+      "step": 489
+    },
+    {
+      "epoch": 0.27466367713004486,
+      "grad_norm": 0.1157331496584953,
+      "learning_rate": 0.0001828358208955224,
+      "loss": 0.3351,
+      "step": 490
+    },
+    {
+      "epoch": 0.27522421524663676,
+      "grad_norm": 0.10566204971680675,
+      "learning_rate": 0.0001832089552238806,
+      "loss": 0.3116,
+      "step": 491
+    },
+    {
+      "epoch": 0.2757847533632287,
+      "grad_norm": 0.11759218211668784,
+      "learning_rate": 0.0001835820895522388,
+      "loss": 0.3258,
+      "step": 492
+    },
+    {
+      "epoch": 0.2763452914798206,
+      "grad_norm": 0.11586768756096413,
+      "learning_rate": 0.00018395522388059703,
+      "loss": 0.3313,
+      "step": 493
+    },
+    {
+      "epoch": 0.2769058295964126,
+      "grad_norm": 0.1113792378824031,
+      "learning_rate": 0.00018432835820895524,
+      "loss": 0.3164,
+      "step": 494
+    },
+    {
+      "epoch": 0.2774663677130045,
+      "grad_norm": 0.10801651632926149,
+      "learning_rate": 0.00018470149253731343,
+      "loss": 0.3189,
+      "step": 495
+    },
+    {
+      "epoch": 0.27802690582959644,
+      "grad_norm": 0.11938118233555668,
+      "learning_rate": 0.00018507462686567165,
+      "loss": 0.3203,
+      "step": 496
+    },
+    {
+      "epoch": 0.27858744394618834,
+      "grad_norm": 0.11022565270186098,
+      "learning_rate": 0.00018544776119402986,
+      "loss": 0.3284,
+      "step": 497
+    },
+    {
+      "epoch": 0.27914798206278024,
+      "grad_norm": 0.11945531573485274,
+      "learning_rate": 0.00018582089552238805,
+      "loss": 0.3271,
+      "step": 498
+    },
+    {
+      "epoch": 0.2797085201793722,
+      "grad_norm": 0.12370474328505932,
+      "learning_rate": 0.00018619402985074626,
+      "loss": 0.3387,
+      "step": 499
+    },
+    {
+      "epoch": 0.2802690582959641,
+      "grad_norm": 0.11262416779259585,
+      "learning_rate": 0.00018656716417910448,
+      "loss": 0.3191,
+      "step": 500
+    },
+    {
+      "epoch": 0.28082959641255606,
+      "grad_norm": 0.11273380446520367,
+      "learning_rate": 0.0001869402985074627,
+      "loss": 0.3211,
+      "step": 501
+    },
+    {
+      "epoch": 0.28139013452914796,
+      "grad_norm": 0.12015390797410297,
+      "learning_rate": 0.0001873134328358209,
+      "loss": 0.3214,
+      "step": 502
+    },
+    {
+      "epoch": 0.2819506726457399,
+      "grad_norm": 0.12206397188877245,
+      "learning_rate": 0.00018768656716417913,
+      "loss": 0.3162,
+      "step": 503
+    },
+    {
+      "epoch": 0.2825112107623318,
+      "grad_norm": 0.10816697719872387,
+      "learning_rate": 0.00018805970149253734,
+      "loss": 0.3301,
+      "step": 504
+    },
+    {
+      "epoch": 0.2830717488789238,
+      "grad_norm": 0.11848480710071109,
+      "learning_rate": 0.00018843283582089553,
+      "loss": 0.3247,
+      "step": 505
+    },
+    {
+      "epoch": 0.2836322869955157,
+      "grad_norm": 0.11298401980526003,
+      "learning_rate": 0.00018880597014925374,
+      "loss": 0.3279,
+      "step": 506
+    },
+    {
+      "epoch": 0.28419282511210764,
+      "grad_norm": 0.11953959263626984,
+      "learning_rate": 0.00018917910447761196,
+      "loss": 0.3299,
+      "step": 507
+    },
+    {
+      "epoch": 0.28475336322869954,
+      "grad_norm": 0.10894689280623014,
+      "learning_rate": 0.00018955223880597015,
+      "loss": 0.3166,
+      "step": 508
+    },
+    {
+      "epoch": 0.2853139013452915,
+      "grad_norm": 0.11704978977855075,
+      "learning_rate": 0.00018992537313432836,
+      "loss": 0.3288,
+      "step": 509
+    },
+    {
+      "epoch": 0.2858744394618834,
+      "grad_norm": 0.11017058286438122,
+      "learning_rate": 0.00019029850746268658,
+      "loss": 0.324,
+      "step": 510
+    },
+    {
+      "epoch": 0.28643497757847536,
+      "grad_norm": 0.10023837779249684,
+      "learning_rate": 0.00019067164179104477,
+      "loss": 0.3136,
+      "step": 511
+    },
+    {
+      "epoch": 0.28699551569506726,
+      "grad_norm": 0.10568197659375772,
+      "learning_rate": 0.00019104477611940298,
+      "loss": 0.316,
+      "step": 512
+    },
+    {
+      "epoch": 0.2875560538116592,
+      "grad_norm": 0.11314477582838516,
+      "learning_rate": 0.0001914179104477612,
+      "loss": 0.332,
+      "step": 513
+    },
+    {
+      "epoch": 0.2881165919282511,
+      "grad_norm": 0.10223986780338,
+      "learning_rate": 0.0001917910447761194,
+      "loss": 0.3263,
+      "step": 514
+    },
+    {
+      "epoch": 0.288677130044843,
+      "grad_norm": 0.11621140033209205,
+      "learning_rate": 0.00019216417910447763,
+      "loss": 0.3251,
+      "step": 515
+    },
+    {
+      "epoch": 0.289237668161435,
+      "grad_norm": 0.10557208094551189,
+      "learning_rate": 0.00019253731343283584,
+      "loss": 0.3101,
+      "step": 516
+    },
+    {
+      "epoch": 0.2897982062780269,
+      "grad_norm": 0.10613098222526403,
+      "learning_rate": 0.00019291044776119406,
+      "loss": 0.3081,
+      "step": 517
+    },
+    {
+      "epoch": 0.29035874439461884,
+      "grad_norm": 0.10511912052331052,
+      "learning_rate": 0.00019328358208955224,
+      "loss": 0.3085,
+      "step": 518
+    },
+    {
+      "epoch": 0.29091928251121074,
+      "grad_norm": 0.10622882777429886,
+      "learning_rate": 0.00019365671641791046,
+      "loss": 0.3211,
+      "step": 519
+    },
+    {
+      "epoch": 0.2914798206278027,
+      "grad_norm": 0.11402567747806869,
+      "learning_rate": 0.00019402985074626867,
+      "loss": 0.3209,
+      "step": 520
+    },
+    {
+      "epoch": 0.2920403587443946,
+      "grad_norm": 0.10923405553131683,
+      "learning_rate": 0.00019440298507462686,
+      "loss": 0.3273,
+      "step": 521
+    },
+    {
+      "epoch": 0.29260089686098656,
+      "grad_norm": 0.11579444570145475,
+      "learning_rate": 0.00019477611940298508,
+      "loss": 0.3161,
+      "step": 522
+    },
+    {
+      "epoch": 0.29316143497757846,
+      "grad_norm": 0.10559557707401233,
+      "learning_rate": 0.0001951492537313433,
+      "loss": 0.3248,
+      "step": 523
+    },
+    {
+      "epoch": 0.2937219730941704,
+      "grad_norm": 0.1122893033865488,
+      "learning_rate": 0.0001955223880597015,
+      "loss": 0.3284,
+      "step": 524
+    },
+    {
+      "epoch": 0.2942825112107623,
+      "grad_norm": 0.10802888290038817,
+      "learning_rate": 0.0001958955223880597,
+      "loss": 0.3337,
+      "step": 525
+    },
+    {
+      "epoch": 0.2948430493273543,
+      "grad_norm": 0.1093293512128306,
+      "learning_rate": 0.0001962686567164179,
+      "loss": 0.3238,
+      "step": 526
+    },
+    {
+      "epoch": 0.2954035874439462,
+      "grad_norm": 0.11270917361257277,
+      "learning_rate": 0.00019664179104477613,
+      "loss": 0.3306,
+      "step": 527
+    },
+    {
+      "epoch": 0.29596412556053814,
+      "grad_norm": 0.10963247890500617,
+      "learning_rate": 0.00019701492537313434,
+      "loss": 0.3187,
+      "step": 528
+    },
+    {
+      "epoch": 0.29652466367713004,
+      "grad_norm": 0.10687024928485456,
+      "learning_rate": 0.00019738805970149256,
+      "loss": 0.3185,
+      "step": 529
+    },
+    {
+      "epoch": 0.297085201793722,
+      "grad_norm": 0.10717585930306575,
+      "learning_rate": 0.00019776119402985077,
+      "loss": 0.3133,
+      "step": 530
+    },
+    {
+      "epoch": 0.2976457399103139,
+      "grad_norm": 0.11014802532876093,
+      "learning_rate": 0.00019813432835820896,
+      "loss": 0.3213,
+      "step": 531
+    },
+    {
+      "epoch": 0.2982062780269058,
+      "grad_norm": 0.11159846382648866,
+      "learning_rate": 0.00019850746268656718,
+      "loss": 0.3272,
+      "step": 532
+    },
+    {
+      "epoch": 0.29876681614349776,
+      "grad_norm": 0.11310481463261243,
+      "learning_rate": 0.0001988805970149254,
+      "loss": 0.3087,
+      "step": 533
+    },
+    {
+      "epoch": 0.29932735426008966,
+      "grad_norm": 0.110102311935057,
+      "learning_rate": 0.0001992537313432836,
+      "loss": 0.3342,
+      "step": 534
+    },
+    {
+      "epoch": 0.2998878923766816,
+      "grad_norm": 0.10480909108854215,
+      "learning_rate": 0.0001996268656716418,
+      "loss": 0.3119,
+      "step": 535
+    },
+    {
+      "epoch": 0.3004484304932735,
+      "grad_norm": 0.11468645363518458,
+      "learning_rate": 0.0002,
+      "loss": 0.3187,
+      "step": 536
+    },
+    {
+      "epoch": 0.3010089686098655,
+      "grad_norm": 0.10153097094841049,
+      "learning_rate": 0.00019999997872366705,
+      "loss": 0.318,
+      "step": 537
+    },
+    {
+      "epoch": 0.3015695067264574,
+      "grad_norm": 0.1007684303802919,
+      "learning_rate": 0.00019999991489467726,
+      "loss": 0.3213,
+      "step": 538
+    },
+    {
+      "epoch": 0.30213004484304934,
+      "grad_norm": 0.10014895411046884,
+      "learning_rate": 0.00019999980851305782,
+      "loss": 0.3076,
+      "step": 539
+    },
+    {
+      "epoch": 0.30269058295964124,
+      "grad_norm": 0.10585004893494934,
+      "learning_rate": 0.00019999965957885393,
+      "loss": 0.3292,
+      "step": 540
+    },
+    {
+      "epoch": 0.3032511210762332,
+      "grad_norm": 0.1152760618726801,
+      "learning_rate": 0.00019999946809212904,
+      "loss": 0.3221,
+      "step": 541
+    },
+    {
+      "epoch": 0.3038116591928251,
+      "grad_norm": 0.09959725256443848,
+      "learning_rate": 0.00019999923405296458,
+      "loss": 0.3352,
+      "step": 542
+    },
+    {
+      "epoch": 0.30437219730941706,
+      "grad_norm": 0.10277553131827694,
+      "learning_rate": 0.0001999989574614602,
+      "loss": 0.3336,
+      "step": 543
+    },
+    {
+      "epoch": 0.30493273542600896,
+      "grad_norm": 0.10378740301586237,
+      "learning_rate": 0.0001999986383177335,
+      "loss": 0.3305,
+      "step": 544
+    },
+    {
+      "epoch": 0.3054932735426009,
+      "grad_norm": 0.10570039748170808,
+      "learning_rate": 0.00019999827662192033,
+      "loss": 0.3141,
+      "step": 545
+    },
+    {
+      "epoch": 0.3060538116591928,
+      "grad_norm": 0.10751566390236701,
+      "learning_rate": 0.00019999787237417468,
+      "loss": 0.3078,
+      "step": 546
+    },
+    {
+      "epoch": 0.3066143497757848,
+      "grad_norm": 0.10953137196388753,
+      "learning_rate": 0.00019999742557466846,
+      "loss": 0.3109,
+      "step": 547
+    },
+    {
+      "epoch": 0.3071748878923767,
+      "grad_norm": 0.10836313700897207,
+      "learning_rate": 0.00019999693622359184,
+      "loss": 0.3309,
+      "step": 548
+    },
+    {
+      "epoch": 0.3077354260089686,
+      "grad_norm": 0.11964787223857871,
+      "learning_rate": 0.00019999640432115303,
+      "loss": 0.321,
+      "step": 549
+    },
+    {
+      "epoch": 0.30829596412556054,
+      "grad_norm": 0.09679420757442243,
+      "learning_rate": 0.00019999582986757842,
+      "loss": 0.328,
+      "step": 550
+    },
+    {
+      "epoch": 0.30885650224215244,
+      "grad_norm": 0.10483981868839679,
+      "learning_rate": 0.00019999521286311238,
+      "loss": 0.3277,
+      "step": 551
+    },
+    {
+      "epoch": 0.3094170403587444,
+      "grad_norm": 0.10724254550549712,
+      "learning_rate": 0.0001999945533080175,
+      "loss": 0.3121,
+      "step": 552
+    },
+    {
+      "epoch": 0.3099775784753363,
+      "grad_norm": 0.09635511885821291,
+      "learning_rate": 0.00019999385120257447,
+      "loss": 0.3146,
+      "step": 553
+    },
+    {
+      "epoch": 0.31053811659192826,
+      "grad_norm": 0.10613771585662128,
+      "learning_rate": 0.00019999310654708204,
+      "loss": 0.3218,
+      "step": 554
+    },
+    {
+      "epoch": 0.31109865470852016,
+      "grad_norm": 0.0957282225898911,
+      "learning_rate": 0.00019999231934185704,
+      "loss": 0.3157,
+      "step": 555
+    },
+    {
+      "epoch": 0.3116591928251121,
+      "grad_norm": 0.0969081583603865,
+      "learning_rate": 0.00019999148958723447,
+      "loss": 0.3187,
+      "step": 556
+    },
+    {
+      "epoch": 0.312219730941704,
+      "grad_norm": 0.09891256442125447,
+      "learning_rate": 0.00019999061728356743,
+      "loss": 0.3054,
+      "step": 557
+    },
+    {
+      "epoch": 0.312780269058296,
+      "grad_norm": 0.1122183619977053,
+      "learning_rate": 0.0001999897024312271,
+      "loss": 0.3112,
+      "step": 558
+    },
+    {
+      "epoch": 0.3133408071748879,
+      "grad_norm": 0.11277342986397464,
+      "learning_rate": 0.00019998874503060273,
+      "loss": 0.3296,
+      "step": 559
+    },
+    {
+      "epoch": 0.31390134529147984,
+      "grad_norm": 0.0970263777528716,
+      "learning_rate": 0.0001999877450821018,
+      "loss": 0.3154,
+      "step": 560
+    },
+    {
+      "epoch": 0.31446188340807174,
+      "grad_norm": 0.10002989014994842,
+      "learning_rate": 0.00019998670258614975,
+      "loss": 0.3147,
+      "step": 561
+    },
+    {
+      "epoch": 0.3150224215246637,
+      "grad_norm": 0.10261105992382952,
+      "learning_rate": 0.00019998561754319024,
+      "loss": 0.3174,
+      "step": 562
+    },
+    {
+      "epoch": 0.3155829596412556,
+      "grad_norm": 0.0972262981200895,
+      "learning_rate": 0.0001999844899536849,
+      "loss": 0.3211,
+      "step": 563
+    },
+    {
+      "epoch": 0.31614349775784756,
+      "grad_norm": 0.09662314153307829,
+      "learning_rate": 0.00019998331981811366,
+      "loss": 0.3069,
+      "step": 564
+    },
+    {
+      "epoch": 0.31670403587443946,
+      "grad_norm": 0.09693479137435353,
+      "learning_rate": 0.00019998210713697437,
+      "loss": 0.3248,
+      "step": 565
+    },
+    {
+      "epoch": 0.3172645739910314,
+      "grad_norm": 0.10673502252972808,
+      "learning_rate": 0.0001999808519107831,
+      "loss": 0.3077,
+      "step": 566
+    },
+    {
+      "epoch": 0.3178251121076233,
+      "grad_norm": 0.10270536213698937,
+      "learning_rate": 0.00019997955414007392,
+      "loss": 0.3177,
+      "step": 567
+    },
+    {
+      "epoch": 0.3183856502242152,
+      "grad_norm": 0.10141916843663253,
+      "learning_rate": 0.0001999782138253991,
+      "loss": 0.304,
+      "step": 568
+    },
+    {
+      "epoch": 0.3189461883408072,
+      "grad_norm": 0.10545679225726372,
+      "learning_rate": 0.00019997683096732906,
+      "loss": 0.3126,
+      "step": 569
+    },
+    {
+      "epoch": 0.3195067264573991,
+      "grad_norm": 0.10698494377190094,
+      "learning_rate": 0.00019997540556645208,
+      "loss": 0.3197,
+      "step": 570
+    },
+    {
+      "epoch": 0.32006726457399104,
+      "grad_norm": 0.09791580882147409,
+      "learning_rate": 0.00019997393762337487,
+      "loss": 0.3313,
+      "step": 571
+    },
+    {
+      "epoch": 0.32062780269058294,
+      "grad_norm": 0.09848352076082446,
+      "learning_rate": 0.00019997242713872196,
+      "loss": 0.3106,
+      "step": 572
+    },
+    {
+      "epoch": 0.3211883408071749,
+      "grad_norm": 0.10380107770857379,
+      "learning_rate": 0.00019997087411313617,
+      "loss": 0.3213,
+      "step": 573
+    },
+    {
+      "epoch": 0.3217488789237668,
+      "grad_norm": 0.09518341108716867,
+      "learning_rate": 0.0001999692785472783,
+      "loss": 0.3141,
+      "step": 574
+    },
+    {
+      "epoch": 0.32230941704035876,
+      "grad_norm": 0.09490611547488066,
+      "learning_rate": 0.00019996764044182737,
+      "loss": 0.3022,
+      "step": 575
+    },
+    {
+      "epoch": 0.32286995515695066,
+      "grad_norm": 0.10238074080827257,
+      "learning_rate": 0.00019996595979748037,
+      "loss": 0.324,
+      "step": 576
+    },
+    {
+      "epoch": 0.3234304932735426,
+      "grad_norm": 0.09636959644273263,
+      "learning_rate": 0.00019996423661495252,
+      "loss": 0.3029,
+      "step": 577
+    },
+    {
+      "epoch": 0.3239910313901345,
+      "grad_norm": 0.0906416563210388,
+      "learning_rate": 0.00019996247089497704,
+      "loss": 0.3046,
+      "step": 578
+    },
+    {
+      "epoch": 0.3245515695067265,
+      "grad_norm": 0.09867816493236885,
+      "learning_rate": 0.00019996066263830531,
+      "loss": 0.3034,
+      "step": 579
+    },
+    {
+      "epoch": 0.3251121076233184,
+      "grad_norm": 0.10144669492274641,
+      "learning_rate": 0.00019995881184570676,
+      "loss": 0.3231,
+      "step": 580
+    },
+    {
+      "epoch": 0.32567264573991034,
+      "grad_norm": 0.09336944858099906,
+      "learning_rate": 0.000199956918517969,
+      "loss": 0.322,
+      "step": 581
+    },
+    {
+      "epoch": 0.32623318385650224,
+      "grad_norm": 0.09516852389764888,
+      "learning_rate": 0.00019995498265589764,
+      "loss": 0.3153,
+      "step": 582
+    },
+    {
+      "epoch": 0.3267937219730942,
+      "grad_norm": 0.09579667771014898,
+      "learning_rate": 0.00019995300426031652,
+      "loss": 0.3192,
+      "step": 583
+    },
+    {
+      "epoch": 0.3273542600896861,
+      "grad_norm": 0.09465110693869315,
+      "learning_rate": 0.00019995098333206742,
+      "loss": 0.3021,
+      "step": 584
+    },
+    {
+      "epoch": 0.327914798206278,
+      "grad_norm": 0.09649031325116542,
+      "learning_rate": 0.00019994891987201033,
+      "loss": 0.3237,
+      "step": 585
+    },
+    {
+      "epoch": 0.32847533632286996,
+      "grad_norm": 0.09981989468985598,
+      "learning_rate": 0.00019994681388102329,
+      "loss": 0.3122,
+      "step": 586
+    },
+    {
+      "epoch": 0.32903587443946186,
+      "grad_norm": 0.09679065544912857,
+      "learning_rate": 0.00019994466536000247,
+      "loss": 0.3058,
+      "step": 587
+    },
+    {
+      "epoch": 0.3295964125560538,
+      "grad_norm": 0.09342549112119153,
+      "learning_rate": 0.00019994247430986213,
+      "loss": 0.3076,
+      "step": 588
+    },
+    {
+      "epoch": 0.3301569506726457,
+      "grad_norm": 0.09804709111681088,
+      "learning_rate": 0.0001999402407315346,
+      "loss": 0.3151,
+      "step": 589
+    },
+    {
+      "epoch": 0.3307174887892377,
+      "grad_norm": 0.09797177899408367,
+      "learning_rate": 0.00019993796462597038,
+      "loss": 0.3159,
+      "step": 590
+    },
+    {
+      "epoch": 0.3312780269058296,
+      "grad_norm": 0.0930412294160317,
+      "learning_rate": 0.00019993564599413792,
+      "loss": 0.3112,
+      "step": 591
+    },
+    {
+      "epoch": 0.33183856502242154,
+      "grad_norm": 0.09188878510394342,
+      "learning_rate": 0.00019993328483702393,
+      "loss": 0.3053,
+      "step": 592
+    },
+    {
+      "epoch": 0.33239910313901344,
+      "grad_norm": 0.09712951064347462,
+      "learning_rate": 0.00019993088115563318,
+      "loss": 0.3104,
+      "step": 593
+    },
+    {
+      "epoch": 0.3329596412556054,
+      "grad_norm": 0.09704324731264691,
+      "learning_rate": 0.00019992843495098838,
+      "loss": 0.3116,
+      "step": 594
+    },
+    {
+      "epoch": 0.3335201793721973,
+      "grad_norm": 0.09257904567873011,
+      "learning_rate": 0.00019992594622413056,
+      "loss": 0.3197,
+      "step": 595
+    },
+    {
+      "epoch": 0.33408071748878926,
+      "grad_norm": 0.09909579185505196,
+      "learning_rate": 0.0001999234149761187,
+      "loss": 0.3296,
+      "step": 596
+    },
+    {
+      "epoch": 0.33464125560538116,
+      "grad_norm": 0.09585706197264815,
+      "learning_rate": 0.0001999208412080299,
+      "loss": 0.3062,
+      "step": 597
+    },
+    {
+      "epoch": 0.3352017937219731,
+      "grad_norm": 0.09786691067245115,
+      "learning_rate": 0.00019991822492095943,
+      "loss": 0.312,
+      "step": 598
+    },
+    {
+      "epoch": 0.335762331838565,
+      "grad_norm": 0.09425861974210471,
+      "learning_rate": 0.0001999155661160205,
+      "loss": 0.321,
+      "step": 599
+    },
+    {
+      "epoch": 0.336322869955157,
+      "grad_norm": 0.09807000412215684,
+      "learning_rate": 0.00019991286479434454,
+      "loss": 0.3177,
+      "step": 600
+    },
+    {
+      "epoch": 0.3368834080717489,
+      "grad_norm": 0.09292352823747375,
+      "learning_rate": 0.00019991012095708105,
+      "loss": 0.3054,
+      "step": 601
+    },
+    {
+      "epoch": 0.3374439461883408,
+      "grad_norm": 0.09778994065142038,
+      "learning_rate": 0.00019990733460539762,
+      "loss": 0.3286,
+      "step": 602
+    },
+    {
+      "epoch": 0.33800448430493274,
+      "grad_norm": 0.10450814592202336,
+      "learning_rate": 0.0001999045057404799,
+      "loss": 0.3001,
+      "step": 603
+    },
+    {
+      "epoch": 0.33856502242152464,
+      "grad_norm": 0.09520669536670996,
+      "learning_rate": 0.0001999016343635316,
+      "loss": 0.3078,
+      "step": 604
+    },
+    {
+      "epoch": 0.3391255605381166,
+      "grad_norm": 0.09123463098371748,
+      "learning_rate": 0.00019989872047577464,
+      "loss": 0.3164,
+      "step": 605
+    },
+    {
+      "epoch": 0.3396860986547085,
+      "grad_norm": 0.092099612562847,
+      "learning_rate": 0.00019989576407844893,
+      "loss": 0.3006,
+      "step": 606
+    },
+    {
+      "epoch": 0.34024663677130046,
+      "grad_norm": 0.0994434074200813,
+      "learning_rate": 0.00019989276517281247,
+      "loss": 0.3149,
+      "step": 607
+    },
+    {
+      "epoch": 0.34080717488789236,
+      "grad_norm": 0.09630693591932951,
+      "learning_rate": 0.00019988972376014142,
+      "loss": 0.3106,
+      "step": 608
+    },
+    {
+      "epoch": 0.3413677130044843,
+      "grad_norm": 0.09685835074291246,
+      "learning_rate": 0.00019988663984172992,
+      "loss": 0.3247,
+      "step": 609
+    },
+    {
+      "epoch": 0.3419282511210762,
+      "grad_norm": 0.09391183577383291,
+      "learning_rate": 0.00019988351341889034,
+      "loss": 0.3035,
+      "step": 610
+    },
+    {
+      "epoch": 0.3424887892376682,
+      "grad_norm": 0.09525770323709616,
+      "learning_rate": 0.00019988034449295298,
+      "loss": 0.3104,
+      "step": 611
+    },
+    {
+      "epoch": 0.3430493273542601,
+      "grad_norm": 0.09130352692181182,
+      "learning_rate": 0.00019987713306526638,
+      "loss": 0.3021,
+      "step": 612
+    },
+    {
+      "epoch": 0.34360986547085204,
+      "grad_norm": 0.09620468186426591,
+      "learning_rate": 0.00019987387913719698,
+      "loss": 0.3019,
+      "step": 613
+    },
+    {
+      "epoch": 0.34417040358744394,
+      "grad_norm": 0.09840324784808555,
+      "learning_rate": 0.00019987058271012952,
+      "loss": 0.3218,
+      "step": 614
+    },
+    {
+      "epoch": 0.3447309417040359,
+      "grad_norm": 0.09636745952863293,
+      "learning_rate": 0.0001998672437854667,
+      "loss": 0.3052,
+      "step": 615
+    },
+    {
+      "epoch": 0.3452914798206278,
+      "grad_norm": 0.09479726421382038,
+      "learning_rate": 0.00019986386236462924,
+      "loss": 0.3198,
+      "step": 616
+    },
+    {
+      "epoch": 0.34585201793721976,
+      "grad_norm": 0.10083947783070678,
+      "learning_rate": 0.00019986043844905612,
+      "loss": 0.308,
+      "step": 617
+    },
+    {
+      "epoch": 0.34641255605381166,
+      "grad_norm": 0.09657101971626757,
+      "learning_rate": 0.00019985697204020423,
+      "loss": 0.3103,
+      "step": 618
+    },
+    {
+      "epoch": 0.34697309417040356,
+      "grad_norm": 0.1075570262110066,
+      "learning_rate": 0.00019985346313954868,
+      "loss": 0.3134,
+      "step": 619
+    },
+    {
+      "epoch": 0.3475336322869955,
+      "grad_norm": 0.09711004477313603,
+      "learning_rate": 0.00019984991174858257,
+      "loss": 0.325,
+      "step": 620
+    },
+    {
+      "epoch": 0.3480941704035874,
+      "grad_norm": 0.09418690943467588,
+      "learning_rate": 0.00019984631786881715,
+      "loss": 0.3147,
+      "step": 621
+    },
+    {
+      "epoch": 0.3486547085201794,
+      "grad_norm": 0.09724763178040019,
+      "learning_rate": 0.00019984268150178167,
+      "loss": 0.3235,
+      "step": 622
+    },
+    {
+      "epoch": 0.3492152466367713,
+      "grad_norm": 0.09182375734289457,
+      "learning_rate": 0.00019983900264902352,
+      "loss": 0.3138,
+      "step": 623
+    },
+    {
+      "epoch": 0.34977578475336324,
+      "grad_norm": 0.10118783544252928,
+      "learning_rate": 0.00019983528131210812,
+      "loss": 0.3165,
+      "step": 624
+    },
+    {
+      "epoch": 0.35033632286995514,
+      "grad_norm": 0.10138373780823115,
+      "learning_rate": 0.00019983151749261905,
+      "loss": 0.3214,
+      "step": 625
+    },
+    {
+      "epoch": 0.3508968609865471,
+      "grad_norm": 0.09104715454857507,
+      "learning_rate": 0.00019982771119215784,
+      "loss": 0.3144,
+      "step": 626
+    },
+    {
+      "epoch": 0.351457399103139,
+      "grad_norm": 0.09908394515423055,
+      "learning_rate": 0.00019982386241234424,
+      "loss": 0.3109,
+      "step": 627
+    },
+    {
+      "epoch": 0.35201793721973096,
+      "grad_norm": 0.09498301385503512,
+      "learning_rate": 0.00019981997115481602,
+      "loss": 0.3007,
+      "step": 628
+    },
+    {
+      "epoch": 0.35257847533632286,
+      "grad_norm": 0.09740150558388727,
+      "learning_rate": 0.00019981603742122894,
+      "loss": 0.3149,
+      "step": 629
+    },
+    {
+      "epoch": 0.3531390134529148,
+      "grad_norm": 0.09447092201896107,
+      "learning_rate": 0.00019981206121325696,
+      "loss": 0.3032,
+      "step": 630
+    },
+    {
+      "epoch": 0.3536995515695067,
+      "grad_norm": 0.09734285460983631,
+      "learning_rate": 0.00019980804253259205,
+      "loss": 0.3132,
+      "step": 631
+    },
+    {
+      "epoch": 0.3542600896860987,
+      "grad_norm": 0.09301208400220677,
+      "learning_rate": 0.00019980398138094428,
+      "loss": 0.3192,
+      "step": 632
+    },
+    {
+      "epoch": 0.3548206278026906,
+      "grad_norm": 0.08989075009688506,
+      "learning_rate": 0.00019979987776004178,
+      "loss": 0.308,
+      "step": 633
+    },
+    {
+      "epoch": 0.35538116591928254,
+      "grad_norm": 0.09501691186329138,
+      "learning_rate": 0.0001997957316716307,
+      "loss": 0.3105,
+      "step": 634
+    },
+    {
+      "epoch": 0.35594170403587444,
+      "grad_norm": 0.08750421147011359,
+      "learning_rate": 0.00019979154311747536,
+      "loss": 0.3176,
+      "step": 635
+    },
+    {
+      "epoch": 0.35650224215246634,
+      "grad_norm": 0.08913056153223253,
+      "learning_rate": 0.0001997873120993581,
+      "loss": 0.3177,
+      "step": 636
+    },
+    {
+      "epoch": 0.3570627802690583,
+      "grad_norm": 0.09047371146376425,
+      "learning_rate": 0.00019978303861907932,
+      "loss": 0.3037,
+      "step": 637
+    },
+    {
+      "epoch": 0.3576233183856502,
+      "grad_norm": 0.0980302332632284,
+      "learning_rate": 0.0001997787226784575,
+      "loss": 0.3222,
+      "step": 638
+    },
+    {
+      "epoch": 0.35818385650224216,
+      "grad_norm": 0.09807009199677308,
+      "learning_rate": 0.0001997743642793292,
+      "loss": 0.3159,
+      "step": 639
+    },
+    {
+      "epoch": 0.35874439461883406,
+      "grad_norm": 0.0936984161703018,
+      "learning_rate": 0.00019976996342354898,
+      "loss": 0.3131,
+      "step": 640
+    },
+    {
+      "epoch": 0.359304932735426,
+      "grad_norm": 0.09710519116360979,
+      "learning_rate": 0.0001997655201129896,
+      "loss": 0.3102,
+      "step": 641
+    },
+    {
+      "epoch": 0.3598654708520179,
+      "grad_norm": 0.09834199646182591,
+      "learning_rate": 0.00019976103434954175,
+      "loss": 0.2985,
+      "step": 642
+    },
+    {
+      "epoch": 0.3604260089686099,
+      "grad_norm": 0.09318331458756994,
+      "learning_rate": 0.00019975650613511428,
+      "loss": 0.3105,
+      "step": 643
+    },
+    {
+      "epoch": 0.3609865470852018,
+      "grad_norm": 0.09060134422957127,
+      "learning_rate": 0.00019975193547163404,
+      "loss": 0.3025,
+      "step": 644
+    },
+    {
+      "epoch": 0.36154708520179374,
+      "grad_norm": 0.1001621229650619,
+      "learning_rate": 0.00019974732236104596,
+      "loss": 0.318,
+      "step": 645
+    },
+    {
+      "epoch": 0.36210762331838564,
+      "grad_norm": 0.09533176887067027,
+      "learning_rate": 0.00019974266680531307,
+      "loss": 0.3188,
+      "step": 646
+    },
+    {
+      "epoch": 0.3626681614349776,
+      "grad_norm": 0.09447592612994977,
+      "learning_rate": 0.00019973796880641645,
+      "loss": 0.3231,
+      "step": 647
+    },
+    {
+      "epoch": 0.3632286995515695,
+      "grad_norm": 0.10038905685358805,
+      "learning_rate": 0.00019973322836635518,
+      "loss": 0.3057,
+      "step": 648
+    },
+    {
+      "epoch": 0.36378923766816146,
+      "grad_norm": 0.09837039723338849,
+      "learning_rate": 0.00019972844548714648,
+      "loss": 0.3097,
+      "step": 649
+    },
+    {
+      "epoch": 0.36434977578475336,
+      "grad_norm": 0.09746096847482318,
+      "learning_rate": 0.00019972362017082554,
+      "loss": 0.2969,
+      "step": 650
+    },
+    {
+      "epoch": 0.3649103139013453,
+      "grad_norm": 0.09046357217513716,
+      "learning_rate": 0.0001997187524194457,
+      "loss": 0.3075,
+      "step": 651
+    },
+    {
+      "epoch": 0.3654708520179372,
+      "grad_norm": 0.08981985275088883,
+      "learning_rate": 0.0001997138422350783,
+      "loss": 0.3137,
+      "step": 652
+    },
+    {
+      "epoch": 0.3660313901345291,
+      "grad_norm": 0.0973371615853409,
+      "learning_rate": 0.0001997088896198128,
+      "loss": 0.3079,
+      "step": 653
+    },
+    {
+      "epoch": 0.3665919282511211,
+      "grad_norm": 0.09641053284915638,
+      "learning_rate": 0.0001997038945757566,
+      "loss": 0.3233,
+      "step": 654
+    },
+    {
+      "epoch": 0.367152466367713,
+      "grad_norm": 0.09083531796513775,
+      "learning_rate": 0.0001996988571050353,
+      "loss": 0.3042,
+      "step": 655
+    },
+    {
+      "epoch": 0.36771300448430494,
+      "grad_norm": 0.0923886180677156,
+      "learning_rate": 0.00019969377720979237,
+      "loss": 0.3095,
+      "step": 656
+    },
+    {
+      "epoch": 0.36827354260089684,
+      "grad_norm": 0.09268763132645462,
+      "learning_rate": 0.0001996886548921896,
+      "loss": 0.3157,
+      "step": 657
+    },
+    {
+      "epoch": 0.3688340807174888,
+      "grad_norm": 0.096762877643564,
+      "learning_rate": 0.00019968349015440652,
+      "loss": 0.2992,
+      "step": 658
+    },
+    {
+      "epoch": 0.3693946188340807,
+      "grad_norm": 0.09588852152964636,
+      "learning_rate": 0.00019967828299864094,
+      "loss": 0.3226,
+      "step": 659
+    },
+    {
+      "epoch": 0.36995515695067266,
+      "grad_norm": 0.09855560751484342,
+      "learning_rate": 0.00019967303342710864,
+      "loss": 0.3172,
+      "step": 660
+    },
+    {
+      "epoch": 0.37051569506726456,
+      "grad_norm": 0.09124675338819187,
+      "learning_rate": 0.00019966774144204342,
+      "loss": 0.3114,
+      "step": 661
+    },
+    {
+      "epoch": 0.3710762331838565,
+      "grad_norm": 0.09497679213022318,
+      "learning_rate": 0.0001996624070456972,
+      "loss": 0.3119,
+      "step": 662
+    },
+    {
+      "epoch": 0.3716367713004484,
+      "grad_norm": 0.09383513529910878,
+      "learning_rate": 0.00019965703024033988,
+      "loss": 0.315,
+      "step": 663
+    },
+    {
+      "epoch": 0.3721973094170404,
+      "grad_norm": 0.09187405365219889,
+      "learning_rate": 0.00019965161102825945,
+      "loss": 0.2989,
+      "step": 664
+    },
+    {
+      "epoch": 0.3727578475336323,
+      "grad_norm": 0.09846630219244981,
+      "learning_rate": 0.00019964614941176195,
+      "loss": 0.3162,
+      "step": 665
+    },
+    {
+      "epoch": 0.37331838565022424,
+      "grad_norm": 0.10258783996322045,
+      "learning_rate": 0.00019964064539317137,
+      "loss": 0.3116,
+      "step": 666
+    },
+    {
+      "epoch": 0.37387892376681614,
+      "grad_norm": 0.09694456938304831,
+      "learning_rate": 0.00019963509897482986,
+      "loss": 0.3019,
+      "step": 667
+    },
+    {
+      "epoch": 0.3744394618834081,
+      "grad_norm": 0.08723571528519997,
+      "learning_rate": 0.0001996295101590976,
+      "loss": 0.2958,
+      "step": 668
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 0.0902327190955776,
+      "learning_rate": 0.00019962387894835275,
+      "loss": 0.3125,
+      "step": 669
+    },
+    {
+      "epoch": 0.3755605381165919,
+      "grad_norm": 0.10497903692593351,
+      "learning_rate": 0.00019961820534499154,
+      "loss": 0.3145,
+      "step": 670
+    },
+    {
+      "epoch": 0.37612107623318386,
+      "grad_norm": 0.08794974221680721,
+      "learning_rate": 0.00019961248935142825,
+      "loss": 0.304,
+      "step": 671
+    },
+    {
+      "epoch": 0.37668161434977576,
+      "grad_norm": 0.08592111650477056,
+      "learning_rate": 0.00019960673097009518,
+      "loss": 0.3295,
+      "step": 672
+    },
+    {
+      "epoch": 0.3772421524663677,
+      "grad_norm": 0.09416973778233857,
+      "learning_rate": 0.00019960093020344265,
+      "loss": 0.3078,
+      "step": 673
+    },
+    {
+      "epoch": 0.3778026905829596,
+      "grad_norm": 0.09882554225698,
+      "learning_rate": 0.0001995950870539391,
+      "loss": 0.3208,
+      "step": 674
+    },
+    {
+      "epoch": 0.3783632286995516,
+      "grad_norm": 0.09056733013559291,
+      "learning_rate": 0.00019958920152407088,
+      "loss": 0.3057,
+      "step": 675
+    },
+    {
+      "epoch": 0.3789237668161435,
+      "grad_norm": 0.093680820250419,
+      "learning_rate": 0.00019958327361634248,
+      "loss": 0.305,
+      "step": 676
+    },
+    {
+      "epoch": 0.37948430493273544,
+      "grad_norm": 0.09117543360175803,
+      "learning_rate": 0.00019957730333327637,
+      "loss": 0.3079,
+      "step": 677
+    },
+    {
+      "epoch": 0.38004484304932734,
+      "grad_norm": 0.09058289965652437,
+      "learning_rate": 0.00019957129067741308,
+      "loss": 0.3224,
+      "step": 678
+    },
+    {
+      "epoch": 0.3806053811659193,
+      "grad_norm": 0.09434118281344171,
+      "learning_rate": 0.00019956523565131115,
+      "loss": 0.3071,
+      "step": 679
+    },
+    {
+      "epoch": 0.3811659192825112,
+      "grad_norm": 0.09689254622157607,
+      "learning_rate": 0.00019955913825754713,
+      "loss": 0.3175,
+      "step": 680
+    },
+    {
+      "epoch": 0.38172645739910316,
+      "grad_norm": 0.09033082100701963,
+      "learning_rate": 0.00019955299849871568,
+      "loss": 0.3153,
+      "step": 681
+    },
+    {
+      "epoch": 0.38228699551569506,
+      "grad_norm": 0.09300510466497924,
+      "learning_rate": 0.00019954681637742933,
+      "loss": 0.3182,
+      "step": 682
+    },
+    {
+      "epoch": 0.382847533632287,
+      "grad_norm": 0.08982220320560935,
+      "learning_rate": 0.00019954059189631883,
+      "loss": 0.2985,
+      "step": 683
+    },
+    {
+      "epoch": 0.3834080717488789,
+      "grad_norm": 0.09212901813961062,
+      "learning_rate": 0.00019953432505803286,
+      "loss": 0.3079,
+      "step": 684
+    },
+    {
+      "epoch": 0.3839686098654709,
+      "grad_norm": 0.09640137098440237,
+      "learning_rate": 0.00019952801586523808,
+      "loss": 0.3063,
+      "step": 685
+    },
+    {
+      "epoch": 0.3845291479820628,
+      "grad_norm": 0.09408208141715627,
+      "learning_rate": 0.00019952166432061924,
+      "loss": 0.3168,
+      "step": 686
+    },
+    {
+      "epoch": 0.3850896860986547,
+      "grad_norm": 0.09168586104891628,
+      "learning_rate": 0.0001995152704268791,
+      "loss": 0.3168,
+      "step": 687
+    },
+    {
+      "epoch": 0.38565022421524664,
+      "grad_norm": 0.09216058156012896,
+      "learning_rate": 0.0001995088341867384,
+      "loss": 0.3058,
+      "step": 688
+    },
+    {
+      "epoch": 0.38621076233183854,
+      "grad_norm": 0.09314968709238407,
+      "learning_rate": 0.000199502355602936,
+      "loss": 0.3074,
+      "step": 689
+    },
+    {
+      "epoch": 0.3867713004484305,
+      "grad_norm": 0.08840063438652919,
+      "learning_rate": 0.0001994958346782286,
+      "loss": 0.3001,
+      "step": 690
+    },
+    {
+      "epoch": 0.3873318385650224,
+      "grad_norm": 0.09621405562039968,
+      "learning_rate": 0.00019948927141539113,
+      "loss": 0.3212,
+      "step": 691
+    },
+    {
+      "epoch": 0.38789237668161436,
+      "grad_norm": 0.08842288128281502,
+      "learning_rate": 0.00019948266581721642,
+      "loss": 0.2957,
+      "step": 692
+    },
+    {
+      "epoch": 0.38845291479820626,
+      "grad_norm": 0.09329535326280906,
+      "learning_rate": 0.00019947601788651527,
+      "loss": 0.3059,
+      "step": 693
+    },
+    {
+      "epoch": 0.3890134529147982,
+      "grad_norm": 0.09776146544452814,
+      "learning_rate": 0.00019946932762611658,
+      "loss": 0.3207,
+      "step": 694
+    },
+    {
+      "epoch": 0.3895739910313901,
+      "grad_norm": 0.09346752724479822,
+      "learning_rate": 0.0001994625950388673,
+      "loss": 0.2949,
+      "step": 695
+    },
+    {
+      "epoch": 0.3901345291479821,
+      "grad_norm": 0.09772096851212224,
+      "learning_rate": 0.0001994558201276322,
+      "loss": 0.3142,
+      "step": 696
+    },
+    {
+      "epoch": 0.390695067264574,
+      "grad_norm": 0.0938987828017779,
+      "learning_rate": 0.00019944900289529425,
+      "loss": 0.3056,
+      "step": 697
+    },
+    {
+      "epoch": 0.39125560538116594,
+      "grad_norm": 0.0965497026787379,
+      "learning_rate": 0.00019944214334475442,
+      "loss": 0.3088,
+      "step": 698
+    },
+    {
+      "epoch": 0.39181614349775784,
+      "grad_norm": 0.09602660877350971,
+      "learning_rate": 0.00019943524147893153,
+      "loss": 0.316,
+      "step": 699
+    },
+    {
+      "epoch": 0.3923766816143498,
+      "grad_norm": 0.09339139467599078,
+      "learning_rate": 0.00019942829730076257,
+      "loss": 0.3134,
+      "step": 700
+    },
+    {
+      "epoch": 0.3929372197309417,
+      "grad_norm": 0.09273886736295357,
+      "learning_rate": 0.00019942131081320246,
+      "loss": 0.312,
+      "step": 701
+    },
+    {
+      "epoch": 0.39349775784753366,
+      "grad_norm": 0.09490328698430993,
+      "learning_rate": 0.00019941428201922413,
+      "loss": 0.3095,
+      "step": 702
+    },
+    {
+      "epoch": 0.39405829596412556,
+      "grad_norm": 0.09384106880964664,
+      "learning_rate": 0.00019940721092181853,
+      "loss": 0.3223,
+      "step": 703
+    },
+    {
+      "epoch": 0.39461883408071746,
+      "grad_norm": 0.09133086939387193,
+      "learning_rate": 0.0001994000975239946,
+      "loss": 0.2988,
+      "step": 704
+    },
+    {
+      "epoch": 0.3951793721973094,
+      "grad_norm": 0.09553642977701621,
+      "learning_rate": 0.00019939294182877925,
+      "loss": 0.3193,
+      "step": 705
+    },
+    {
+      "epoch": 0.3957399103139013,
+      "grad_norm": 0.08857932146083475,
+      "learning_rate": 0.0001993857438392175,
+      "loss": 0.305,
+      "step": 706
+    },
+    {
+      "epoch": 0.3963004484304933,
+      "grad_norm": 0.08771824644520687,
+      "learning_rate": 0.00019937850355837217,
+      "loss": 0.2984,
+      "step": 707
+    },
+    {
+      "epoch": 0.3968609865470852,
+      "grad_norm": 0.0903763377102578,
+      "learning_rate": 0.00019937122098932428,
+      "loss": 0.298,
+      "step": 708
+    },
+    {
+      "epoch": 0.39742152466367714,
+      "grad_norm": 0.09034740590815869,
+      "learning_rate": 0.0001993638961351727,
+      "loss": 0.3112,
+      "step": 709
+    },
+    {
+      "epoch": 0.39798206278026904,
+      "grad_norm": 0.09708930191408309,
+      "learning_rate": 0.00019935652899903442,
+      "loss": 0.3106,
+      "step": 710
+    },
+    {
+      "epoch": 0.398542600896861,
+      "grad_norm": 0.09763701767129569,
+      "learning_rate": 0.00019934911958404428,
+      "loss": 0.3092,
+      "step": 711
+    },
+    {
+      "epoch": 0.3991031390134529,
+      "grad_norm": 0.09257194185153657,
+      "learning_rate": 0.00019934166789335525,
+      "loss": 0.3091,
+      "step": 712
+    },
+    {
+      "epoch": 0.39966367713004486,
+      "grad_norm": 0.09116327725227731,
+      "learning_rate": 0.00019933417393013815,
+      "loss": 0.3076,
+      "step": 713
+    },
+    {
+      "epoch": 0.40022421524663676,
+      "grad_norm": 0.08918440964551518,
+      "learning_rate": 0.00019932663769758194,
+      "loss": 0.321,
+      "step": 714
+    },
+    {
+      "epoch": 0.4007847533632287,
+      "grad_norm": 0.09328211535654479,
+      "learning_rate": 0.00019931905919889342,
+      "loss": 0.3016,
+      "step": 715
+    },
+    {
+      "epoch": 0.4013452914798206,
+      "grad_norm": 0.08758524473066116,
+      "learning_rate": 0.00019931143843729748,
+      "loss": 0.3019,
+      "step": 716
+    },
+    {
+      "epoch": 0.4019058295964126,
+      "grad_norm": 0.08923669644041521,
+      "learning_rate": 0.00019930377541603695,
+      "loss": 0.3133,
+      "step": 717
+    },
+    {
+      "epoch": 0.4024663677130045,
+      "grad_norm": 0.08958452005045,
+      "learning_rate": 0.0001992960701383727,
+      "loss": 0.304,
+      "step": 718
+    },
+    {
+      "epoch": 0.40302690582959644,
+      "grad_norm": 0.08916290118406439,
+      "learning_rate": 0.0001992883226075834,
+      "loss": 0.299,
+      "step": 719
+    },
+    {
+      "epoch": 0.40358744394618834,
+      "grad_norm": 0.09392346415005111,
+      "learning_rate": 0.00019928053282696596,
+      "loss": 0.3075,
+      "step": 720
+    },
+    {
+      "epoch": 0.40414798206278024,
+      "grad_norm": 0.09133173321569467,
+      "learning_rate": 0.00019927270079983506,
+      "loss": 0.3074,
+      "step": 721
+    },
+    {
+      "epoch": 0.4047085201793722,
+      "grad_norm": 0.08909159921128257,
+      "learning_rate": 0.00019926482652952347,
+      "loss": 0.3136,
+      "step": 722
+    },
+    {
+      "epoch": 0.4052690582959641,
+      "grad_norm": 0.09527893792172565,
+      "learning_rate": 0.0001992569100193819,
+      "loss": 0.3156,
+      "step": 723
+    },
+    {
+      "epoch": 0.40582959641255606,
+      "grad_norm": 0.08969969555334135,
+      "learning_rate": 0.00019924895127277907,
+      "loss": 0.3086,
+      "step": 724
+    },
+    {
+      "epoch": 0.40639013452914796,
+      "grad_norm": 0.09241404516413201,
+      "learning_rate": 0.00019924095029310158,
+      "loss": 0.301,
+      "step": 725
+    },
+    {
+      "epoch": 0.4069506726457399,
+      "grad_norm": 0.09069006933938728,
+      "learning_rate": 0.00019923290708375412,
+      "loss": 0.3148,
+      "step": 726
+    },
+    {
+      "epoch": 0.4075112107623318,
+      "grad_norm": 0.09890533901369626,
+      "learning_rate": 0.0001992248216481592,
+      "loss": 0.3048,
+      "step": 727
+    },
+    {
+      "epoch": 0.4080717488789238,
+      "grad_norm": 0.09313363812448927,
+      "learning_rate": 0.00019921669398975745,
+      "loss": 0.3013,
+      "step": 728
+    },
+    {
+      "epoch": 0.4086322869955157,
+      "grad_norm": 0.08711918573325704,
+      "learning_rate": 0.0001992085241120074,
+      "loss": 0.2915,
+      "step": 729
+    },
+    {
+      "epoch": 0.40919282511210764,
+      "grad_norm": 0.09498553317795624,
+      "learning_rate": 0.00019920031201838557,
+      "loss": 0.3079,
+      "step": 730
+    },
+    {
+      "epoch": 0.40975336322869954,
+      "grad_norm": 0.08829302260299005,
+      "learning_rate": 0.00019919205771238638,
+      "loss": 0.308,
+      "step": 731
+    },
+    {
+      "epoch": 0.4103139013452915,
+      "grad_norm": 0.09193666567406178,
+      "learning_rate": 0.0001991837611975223,
+      "loss": 0.3077,
+      "step": 732
+    },
+    {
+      "epoch": 0.4108744394618834,
+      "grad_norm": 0.0877409102296216,
+      "learning_rate": 0.0001991754224773237,
+      "loss": 0.3021,
+      "step": 733
+    },
+    {
+      "epoch": 0.41143497757847536,
+      "grad_norm": 0.09163527694355844,
+      "learning_rate": 0.0001991670415553389,
+      "loss": 0.3157,
+      "step": 734
+    },
+    {
+      "epoch": 0.41199551569506726,
+      "grad_norm": 0.08851308902476697,
+      "learning_rate": 0.00019915861843513425,
+      "loss": 0.2987,
+      "step": 735
+    },
+    {
+      "epoch": 0.4125560538116592,
+      "grad_norm": 0.08865061811742718,
+      "learning_rate": 0.000199150153120294,
+      "loss": 0.318,
+      "step": 736
+    },
+    {
+      "epoch": 0.4131165919282511,
+      "grad_norm": 0.0916424076710549,
+      "learning_rate": 0.00019914164561442036,
+      "loss": 0.3042,
+      "step": 737
+    },
+    {
+      "epoch": 0.413677130044843,
+      "grad_norm": 0.09075462969970004,
+      "learning_rate": 0.00019913309592113347,
+      "loss": 0.3093,
+      "step": 738
+    },
+    {
+      "epoch": 0.414237668161435,
+      "grad_norm": 0.09227473483449715,
+      "learning_rate": 0.0001991245040440715,
+      "loss": 0.3072,
+      "step": 739
+    },
+    {
+      "epoch": 0.4147982062780269,
+      "grad_norm": 0.09014553712868298,
+      "learning_rate": 0.0001991158699868905,
+      "loss": 0.3028,
+      "step": 740
+    },
+    {
+      "epoch": 0.41535874439461884,
+      "grad_norm": 0.0923974795060702,
+      "learning_rate": 0.00019910719375326453,
+      "loss": 0.2984,
+      "step": 741
+    },
+    {
+      "epoch": 0.41591928251121074,
+      "grad_norm": 0.08943840942515693,
+      "learning_rate": 0.00019909847534688553,
+      "loss": 0.3004,
+      "step": 742
+    },
+    {
+      "epoch": 0.4164798206278027,
+      "grad_norm": 0.08567226895006697,
+      "learning_rate": 0.00019908971477146338,
+      "loss": 0.2944,
+      "step": 743
+    },
+    {
+      "epoch": 0.4170403587443946,
+      "grad_norm": 0.08701439358407168,
+      "learning_rate": 0.00019908091203072598,
+      "loss": 0.2941,
+      "step": 744
+    },
+    {
+      "epoch": 0.41760089686098656,
+      "grad_norm": 0.08850206841755387,
+      "learning_rate": 0.00019907206712841915,
+      "loss": 0.3077,
+      "step": 745
+    },
+    {
+      "epoch": 0.41816143497757846,
+      "grad_norm": 0.08721106462150639,
+      "learning_rate": 0.00019906318006830657,
+      "loss": 0.3151,
+      "step": 746
+    },
+    {
+      "epoch": 0.4187219730941704,
+      "grad_norm": 0.09393278041736332,
+      "learning_rate": 0.00019905425085416995,
+      "loss": 0.3068,
+      "step": 747
+    },
+    {
+      "epoch": 0.4192825112107623,
+      "grad_norm": 0.09324005796413601,
+      "learning_rate": 0.00019904527948980894,
+      "loss": 0.3047,
+      "step": 748
+    },
+    {
+      "epoch": 0.4198430493273543,
+      "grad_norm": 0.0888935722749391,
+      "learning_rate": 0.00019903626597904105,
+      "loss": 0.2967,
+      "step": 749
+    },
+    {
+      "epoch": 0.4204035874439462,
+      "grad_norm": 0.08673797173238905,
+      "learning_rate": 0.00019902721032570176,
+      "loss": 0.2977,
+      "step": 750
+    },
+    {
+      "epoch": 0.42096412556053814,
+      "grad_norm": 0.08806981508498309,
+      "learning_rate": 0.00019901811253364456,
+      "loss": 0.318,
+      "step": 751
+    },
+    {
+      "epoch": 0.42152466367713004,
+      "grad_norm": 0.08881109920967563,
+      "learning_rate": 0.00019900897260674073,
+      "loss": 0.2999,
+      "step": 752
+    },
+    {
+      "epoch": 0.422085201793722,
+      "grad_norm": 0.09538416584410722,
+      "learning_rate": 0.00019899979054887964,
+      "loss": 0.3203,
+      "step": 753
+    },
+    {
+      "epoch": 0.4226457399103139,
+      "grad_norm": 0.08952151997022839,
+      "learning_rate": 0.0001989905663639684,
+      "loss": 0.3004,
+      "step": 754
+    },
+    {
+      "epoch": 0.4232062780269058,
+      "grad_norm": 0.09001015247373621,
+      "learning_rate": 0.00019898130005593218,
+      "loss": 0.2847,
+      "step": 755
+    },
+    {
+      "epoch": 0.42376681614349776,
+      "grad_norm": 0.09281590608787986,
+      "learning_rate": 0.00019897199162871408,
+      "loss": 0.307,
+      "step": 756
+    },
+    {
+      "epoch": 0.42432735426008966,
+      "grad_norm": 0.08672644067463742,
+      "learning_rate": 0.00019896264108627506,
+      "loss": 0.3116,
+      "step": 757
+    },
+    {
+      "epoch": 0.4248878923766816,
+      "grad_norm": 0.08503075443969747,
+      "learning_rate": 0.000198953248432594,
+      "loss": 0.303,
+      "step": 758
+    },
+    {
+      "epoch": 0.4254484304932735,
+      "grad_norm": 0.09051207207933547,
+      "learning_rate": 0.00019894381367166773,
+      "loss": 0.3055,
+      "step": 759
+    },
+    {
+      "epoch": 0.4260089686098655,
+      "grad_norm": 0.08723584743342241,
+      "learning_rate": 0.00019893433680751103,
+      "loss": 0.3102,
+      "step": 760
+    },
+    {
+      "epoch": 0.4265695067264574,
+      "grad_norm": 0.08507901240916076,
+      "learning_rate": 0.00019892481784415653,
+      "loss": 0.3009,
+      "step": 761
+    },
+    {
+      "epoch": 0.42713004484304934,
+      "grad_norm": 0.08923668828516959,
+      "learning_rate": 0.0001989152567856548,
+      "loss": 0.3039,
+      "step": 762
+    },
+    {
+      "epoch": 0.42769058295964124,
+      "grad_norm": 0.08780375279840583,
+      "learning_rate": 0.00019890565363607436,
+      "loss": 0.3181,
+      "step": 763
+    },
+    {
+      "epoch": 0.4282511210762332,
+      "grad_norm": 0.08725479512572638,
+      "learning_rate": 0.00019889600839950155,
+      "loss": 0.296,
+      "step": 764
+    },
+    {
+      "epoch": 0.4288116591928251,
+      "grad_norm": 0.08288503384141521,
+      "learning_rate": 0.00019888632108004074,
+      "loss": 0.2958,
+      "step": 765
+    },
+    {
+      "epoch": 0.42937219730941706,
+      "grad_norm": 0.08968254486397406,
+      "learning_rate": 0.0001988765916818141,
+      "loss": 0.3125,
+      "step": 766
+    },
+    {
+      "epoch": 0.42993273542600896,
+      "grad_norm": 0.08205694133360567,
+      "learning_rate": 0.00019886682020896172,
+      "loss": 0.292,
+      "step": 767
+    },
+    {
+      "epoch": 0.4304932735426009,
+      "grad_norm": 0.0897100401146557,
+      "learning_rate": 0.0001988570066656417,
+      "loss": 0.2924,
+      "step": 768
+    },
+    {
+      "epoch": 0.4310538116591928,
+      "grad_norm": 0.08738986931277319,
+      "learning_rate": 0.00019884715105602992,
+      "loss": 0.2935,
+      "step": 769
+    },
+    {
+      "epoch": 0.4316143497757848,
+      "grad_norm": 0.08997555138739971,
+      "learning_rate": 0.00019883725338432017,
+      "loss": 0.321,
+      "step": 770
+    },
+    {
+      "epoch": 0.4321748878923767,
+      "grad_norm": 0.09148998862478722,
+      "learning_rate": 0.00019882731365472424,
+      "loss": 0.3118,
+      "step": 771
+    },
+    {
+      "epoch": 0.4327354260089686,
+      "grad_norm": 0.08668291843813691,
+      "learning_rate": 0.00019881733187147171,
+      "loss": 0.321,
+      "step": 772
+    },
+    {
+      "epoch": 0.43329596412556054,
+      "grad_norm": 0.08776191807915673,
+      "learning_rate": 0.00019880730803881013,
+      "loss": 0.2999,
+      "step": 773
+    },
+    {
+      "epoch": 0.43385650224215244,
+      "grad_norm": 0.08987310100223357,
+      "learning_rate": 0.00019879724216100486,
+      "loss": 0.3162,
+      "step": 774
+    },
+    {
+      "epoch": 0.4344170403587444,
+      "grad_norm": 0.08760893018411792,
+      "learning_rate": 0.00019878713424233924,
+      "loss": 0.313,
+      "step": 775
+    },
+    {
+      "epoch": 0.4349775784753363,
+      "grad_norm": 0.09028271206037046,
+      "learning_rate": 0.00019877698428711442,
+      "loss": 0.3076,
+      "step": 776
+    },
+    {
+      "epoch": 0.43553811659192826,
+      "grad_norm": 0.08422567534720686,
+      "learning_rate": 0.00019876679229964949,
+      "loss": 0.2988,
+      "step": 777
+    },
+    {
+      "epoch": 0.43609865470852016,
+      "grad_norm": 0.08974811464181615,
+      "learning_rate": 0.00019875655828428145,
+      "loss": 0.311,
+      "step": 778
+    },
+    {
+      "epoch": 0.4366591928251121,
+      "grad_norm": 0.08446337267991294,
+      "learning_rate": 0.00019874628224536513,
+      "loss": 0.2995,
+      "step": 779
+    },
+    {
+      "epoch": 0.437219730941704,
+      "grad_norm": 0.09443337966417452,
+      "learning_rate": 0.0001987359641872732,
+      "loss": 0.3195,
+      "step": 780
+    },
+    {
+      "epoch": 0.437780269058296,
+      "grad_norm": 0.08831834843301448,
+      "learning_rate": 0.00019872560411439633,
+      "loss": 0.3051,
+      "step": 781
+    },
+    {
+      "epoch": 0.4383408071748879,
+      "grad_norm": 0.08717347016508245,
+      "learning_rate": 0.000198715202031143,
+      "loss": 0.3095,
+      "step": 782
+    },
+    {
+      "epoch": 0.43890134529147984,
+      "grad_norm": 0.08143361331223166,
+      "learning_rate": 0.00019870475794193956,
+      "loss": 0.3034,
+      "step": 783
+    },
+    {
+      "epoch": 0.43946188340807174,
+      "grad_norm": 0.0872812460626501,
+      "learning_rate": 0.00019869427185123027,
+      "loss": 0.3037,
+      "step": 784
+    },
+    {
+      "epoch": 0.4400224215246637,
+      "grad_norm": 0.08663859380296493,
+      "learning_rate": 0.0001986837437634772,
+      "loss": 0.3115,
+      "step": 785
+    },
+    {
+      "epoch": 0.4405829596412556,
+      "grad_norm": 0.09141249344984331,
+      "learning_rate": 0.00019867317368316037,
+      "loss": 0.3036,
+      "step": 786
+    },
+    {
+      "epoch": 0.44114349775784756,
+      "grad_norm": 0.08653436715401916,
+      "learning_rate": 0.0001986625616147776,
+      "loss": 0.2992,
+      "step": 787
+    },
+    {
+      "epoch": 0.44170403587443946,
+      "grad_norm": 0.0875322052920452,
+      "learning_rate": 0.00019865190756284467,
+      "loss": 0.3012,
+      "step": 788
+    },
+    {
+      "epoch": 0.4422645739910314,
+      "grad_norm": 0.09067851337305519,
+      "learning_rate": 0.0001986412115318951,
+      "loss": 0.2952,
+      "step": 789
+    },
+    {
+      "epoch": 0.4428251121076233,
+      "grad_norm": 0.0882649566344737,
+      "learning_rate": 0.00019863047352648033,
+      "loss": 0.3093,
+      "step": 790
+    },
+    {
+      "epoch": 0.4433856502242152,
+      "grad_norm": 0.08513561123668154,
+      "learning_rate": 0.0001986196935511697,
+      "loss": 0.3136,
+      "step": 791
+    },
+    {
+      "epoch": 0.4439461883408072,
+      "grad_norm": 0.09622214580200847,
+      "learning_rate": 0.00019860887161055038,
+      "loss": 0.3108,
+      "step": 792
+    },
+    {
+      "epoch": 0.4445067264573991,
+      "grad_norm": 0.08687900491585457,
+      "learning_rate": 0.0001985980077092274,
+      "loss": 0.2986,
+      "step": 793
+    },
+    {
+      "epoch": 0.44506726457399104,
+      "grad_norm": 0.08869706121836343,
+      "learning_rate": 0.0001985871018518236,
+      "loss": 0.31,
+      "step": 794
+    },
+    {
+      "epoch": 0.44562780269058294,
+      "grad_norm": 0.08728306694736086,
+      "learning_rate": 0.00019857615404297974,
+      "loss": 0.3074,
+      "step": 795
+    },
+    {
+      "epoch": 0.4461883408071749,
+      "grad_norm": 0.08713815025560269,
+      "learning_rate": 0.0001985651642873544,
+      "loss": 0.3044,
+      "step": 796
+    },
+    {
+      "epoch": 0.4467488789237668,
+      "grad_norm": 0.08358102481973519,
+      "learning_rate": 0.00019855413258962402,
+      "loss": 0.3005,
+      "step": 797
+    },
+    {
+      "epoch": 0.44730941704035876,
+      "grad_norm": 0.08558742897488306,
+      "learning_rate": 0.00019854305895448287,
+      "loss": 0.3125,
+      "step": 798
+    },
+    {
+      "epoch": 0.44786995515695066,
+      "grad_norm": 0.09278562896889817,
+      "learning_rate": 0.00019853194338664308,
+      "loss": 0.3007,
+      "step": 799
+    },
+    {
+      "epoch": 0.4484304932735426,
+      "grad_norm": 0.08837103157047665,
+      "learning_rate": 0.00019852078589083466,
+      "loss": 0.3196,
+      "step": 800
+    },
+    {
+      "epoch": 0.4489910313901345,
+      "grad_norm": 0.0877899829779131,
+      "learning_rate": 0.00019850958647180534,
+      "loss": 0.2914,
+      "step": 801
+    },
+    {
+      "epoch": 0.4495515695067265,
+      "grad_norm": 0.08740164316166062,
+      "learning_rate": 0.00019849834513432083,
+      "loss": 0.2977,
+      "step": 802
+    },
+    {
+      "epoch": 0.4501121076233184,
+      "grad_norm": 0.08735279292449416,
+      "learning_rate": 0.00019848706188316465,
+      "loss": 0.3261,
+      "step": 803
+    },
+    {
+      "epoch": 0.45067264573991034,
+      "grad_norm": 0.08828556023556346,
+      "learning_rate": 0.00019847573672313802,
+      "loss": 0.3054,
+      "step": 804
+    },
+    {
+      "epoch": 0.45123318385650224,
+      "grad_norm": 0.08800560863232093,
+      "learning_rate": 0.0001984643696590602,
+      "loss": 0.308,
+      "step": 805
+    },
+    {
+      "epoch": 0.4517937219730942,
+      "grad_norm": 0.08208728640708014,
+      "learning_rate": 0.00019845296069576809,
+      "loss": 0.3014,
+      "step": 806
+    },
+    {
+      "epoch": 0.4523542600896861,
+      "grad_norm": 0.08271451540214034,
+      "learning_rate": 0.00019844150983811657,
+      "loss": 0.3001,
+      "step": 807
+    },
+    {
+      "epoch": 0.452914798206278,
+      "grad_norm": 0.09035018600164363,
+      "learning_rate": 0.0001984300170909783,
+      "loss": 0.3024,
+      "step": 808
+    },
+    {
+      "epoch": 0.45347533632286996,
+      "grad_norm": 0.08771905523544332,
+      "learning_rate": 0.0001984184824592437,
+      "loss": 0.3183,
+      "step": 809
+    },
+    {
+      "epoch": 0.45403587443946186,
+      "grad_norm": 0.08752544096797525,
+      "learning_rate": 0.00019840690594782109,
+      "loss": 0.2999,
+      "step": 810
+    },
+    {
+      "epoch": 0.4545964125560538,
+      "grad_norm": 0.08578415848259245,
+      "learning_rate": 0.00019839528756163656,
+      "loss": 0.3091,
+      "step": 811
+    },
+    {
+      "epoch": 0.4551569506726457,
+      "grad_norm": 0.08779365509665701,
+      "learning_rate": 0.00019838362730563406,
+      "loss": 0.3019,
+      "step": 812
+    },
+    {
+      "epoch": 0.4557174887892377,
+      "grad_norm": 0.08759195135247771,
+      "learning_rate": 0.00019837192518477536,
+      "loss": 0.309,
+      "step": 813
+    },
+    {
+      "epoch": 0.4562780269058296,
+      "grad_norm": 0.0853426049035136,
+      "learning_rate": 0.00019836018120404002,
+      "loss": 0.3054,
+      "step": 814
+    },
+    {
+      "epoch": 0.45683856502242154,
+      "grad_norm": 0.08576349200936671,
+      "learning_rate": 0.00019834839536842536,
+      "loss": 0.3098,
+      "step": 815
+    },
+    {
+      "epoch": 0.45739910313901344,
+      "grad_norm": 0.08619979660754418,
+      "learning_rate": 0.00019833656768294662,
+      "loss": 0.2964,
+      "step": 816
+    },
+    {
+      "epoch": 0.4579596412556054,
+      "grad_norm": 0.09528083875601881,
+      "learning_rate": 0.0001983246981526368,
+      "loss": 0.2959,
+      "step": 817
+    },
+    {
+      "epoch": 0.4585201793721973,
+      "grad_norm": 0.08344925075952278,
+      "learning_rate": 0.0001983127867825467,
+      "loss": 0.3069,
+      "step": 818
+    },
+    {
+      "epoch": 0.45908071748878926,
+      "grad_norm": 0.08460194518898766,
+      "learning_rate": 0.00019830083357774486,
+      "loss": 0.3042,
+      "step": 819
+    },
+    {
+      "epoch": 0.45964125560538116,
+      "grad_norm": 0.08721789967399277,
+      "learning_rate": 0.00019828883854331776,
+      "loss": 0.3177,
+      "step": 820
+    },
+    {
+      "epoch": 0.4602017937219731,
+      "grad_norm": 0.08722294812002279,
+      "learning_rate": 0.0001982768016843696,
+      "loss": 0.2854,
+      "step": 821
+    },
+    {
+      "epoch": 0.460762331838565,
+      "grad_norm": 0.08689503300840559,
+      "learning_rate": 0.00019826472300602237,
+      "loss": 0.2978,
+      "step": 822
+    },
+    {
+      "epoch": 0.461322869955157,
+      "grad_norm": 0.08779777368725004,
+      "learning_rate": 0.00019825260251341587,
+      "loss": 0.2997,
+      "step": 823
+    },
+    {
+      "epoch": 0.4618834080717489,
+      "grad_norm": 0.08693707059653778,
+      "learning_rate": 0.0001982404402117077,
+      "loss": 0.2945,
+      "step": 824
+    },
+    {
+      "epoch": 0.4624439461883408,
+      "grad_norm": 0.08650668902086212,
+      "learning_rate": 0.0001982282361060732,
+      "loss": 0.2937,
+      "step": 825
+    },
+    {
+      "epoch": 0.46300448430493274,
+      "grad_norm": 0.09077683624522533,
+      "learning_rate": 0.0001982159902017056,
+      "loss": 0.3028,
+      "step": 826
+    },
+    {
+      "epoch": 0.46356502242152464,
+      "grad_norm": 0.08647148809856209,
+      "learning_rate": 0.00019820370250381585,
+      "loss": 0.299,
+      "step": 827
+    },
+    {
+      "epoch": 0.4641255605381166,
+      "grad_norm": 0.08243284869484088,
+      "learning_rate": 0.00019819137301763267,
+      "loss": 0.3057,
+      "step": 828
+    },
+    {
+      "epoch": 0.4646860986547085,
+      "grad_norm": 0.08572317691098481,
+      "learning_rate": 0.00019817900174840257,
+      "loss": 0.3066,
+      "step": 829
+    },
+    {
+      "epoch": 0.46524663677130046,
+      "grad_norm": 0.08511427435690537,
+      "learning_rate": 0.0001981665887013899,
+      "loss": 0.3064,
+      "step": 830
+    },
+    {
+      "epoch": 0.46580717488789236,
+      "grad_norm": 0.08716461609535324,
+      "learning_rate": 0.00019815413388187672,
+      "loss": 0.3134,
+      "step": 831
+    },
+    {
+      "epoch": 0.4663677130044843,
+      "grad_norm": 0.08720503249671421,
+      "learning_rate": 0.00019814163729516292,
+      "loss": 0.2941,
+      "step": 832
+    },
+    {
+      "epoch": 0.4669282511210762,
+      "grad_norm": 0.08863179581270388,
+      "learning_rate": 0.00019812909894656607,
+      "loss": 0.3091,
+      "step": 833
+    },
+    {
+      "epoch": 0.4674887892376682,
+      "grad_norm": 0.08659608487256064,
+      "learning_rate": 0.00019811651884142162,
+      "loss": 0.2929,
+      "step": 834
+    },
+    {
+      "epoch": 0.4680493273542601,
+      "grad_norm": 0.09212980135470979,
+      "learning_rate": 0.0001981038969850827,
+      "loss": 0.3138,
+      "step": 835
+    },
+    {
+      "epoch": 0.46860986547085204,
+      "grad_norm": 0.08663807470643362,
+      "learning_rate": 0.0001980912333829203,
+      "loss": 0.3109,
+      "step": 836
+    },
+    {
+      "epoch": 0.46917040358744394,
+      "grad_norm": 0.08056886951111863,
+      "learning_rate": 0.00019807852804032305,
+      "loss": 0.3056,
+      "step": 837
+    },
+    {
+      "epoch": 0.4697309417040359,
+      "grad_norm": 0.08447064309312803,
+      "learning_rate": 0.0001980657809626975,
+      "loss": 0.305,
+      "step": 838
+    },
+    {
+      "epoch": 0.4702914798206278,
+      "grad_norm": 0.08525248902671248,
+      "learning_rate": 0.00019805299215546778,
+      "loss": 0.3085,
+      "step": 839
+    },
+    {
+      "epoch": 0.47085201793721976,
+      "grad_norm": 0.08299100105530506,
+      "learning_rate": 0.0001980401616240759,
+      "loss": 0.3034,
+      "step": 840
+    },
+    {
+      "epoch": 0.47141255605381166,
+      "grad_norm": 0.08789070494074572,
+      "learning_rate": 0.00019802728937398165,
+      "loss": 0.3021,
+      "step": 841
+    },
+    {
+      "epoch": 0.47197309417040356,
+      "grad_norm": 0.08685735259602445,
+      "learning_rate": 0.00019801437541066243,
+      "loss": 0.3009,
+      "step": 842
+    },
+    {
+      "epoch": 0.4725336322869955,
+      "grad_norm": 0.0879939615906324,
+      "learning_rate": 0.00019800141973961357,
+      "loss": 0.3014,
+      "step": 843
+    },
+    {
+      "epoch": 0.4730941704035874,
+      "grad_norm": 0.08780966857867269,
+      "learning_rate": 0.00019798842236634797,
+      "loss": 0.2952,
+      "step": 844
+    },
+    {
+      "epoch": 0.4736547085201794,
+      "grad_norm": 0.08871799147041039,
+      "learning_rate": 0.0001979753832963964,
+      "loss": 0.2958,
+      "step": 845
+    },
+    {
+      "epoch": 0.4742152466367713,
+      "grad_norm": 0.08597502962700726,
+      "learning_rate": 0.00019796230253530728,
+      "loss": 0.3081,
+      "step": 846
+    },
+    {
+      "epoch": 0.47477578475336324,
+      "grad_norm": 0.08323592889114863,
+      "learning_rate": 0.00019794918008864687,
+      "loss": 0.304,
+      "step": 847
+    },
+    {
+      "epoch": 0.47533632286995514,
+      "grad_norm": 0.08512574100191798,
+      "learning_rate": 0.00019793601596199912,
+      "loss": 0.3064,
+      "step": 848
+    },
+    {
+      "epoch": 0.4758968609865471,
+      "grad_norm": 0.08203530412255575,
+      "learning_rate": 0.00019792281016096572,
+      "loss": 0.3063,
+      "step": 849
+    },
+    {
+      "epoch": 0.476457399103139,
+      "grad_norm": 0.08334682011724658,
+      "learning_rate": 0.0001979095626911661,
+      "loss": 0.2988,
+      "step": 850
+    },
+    {
+      "epoch": 0.47701793721973096,
+      "grad_norm": 0.08518984832948939,
+      "learning_rate": 0.00019789627355823735,
+      "loss": 0.2968,
+      "step": 851
+    },
+    {
+      "epoch": 0.47757847533632286,
+      "grad_norm": 0.08791084921603258,
+      "learning_rate": 0.00019788294276783442,
+      "loss": 0.2973,
+      "step": 852
+    },
+    {
+      "epoch": 0.4781390134529148,
+      "grad_norm": 0.08269680827258992,
+      "learning_rate": 0.00019786957032562986,
+      "loss": 0.2892,
+      "step": 853
+    },
+    {
+      "epoch": 0.4786995515695067,
+      "grad_norm": 0.08784029867945722,
+      "learning_rate": 0.00019785615623731407,
+      "loss": 0.311,
+      "step": 854
+    },
+    {
+      "epoch": 0.4792600896860987,
+      "grad_norm": 0.08797940188677301,
+      "learning_rate": 0.00019784270050859503,
+      "loss": 0.2975,
+      "step": 855
+    },
+    {
+      "epoch": 0.4798206278026906,
+      "grad_norm": 0.08287170236731753,
+      "learning_rate": 0.00019782920314519856,
+      "loss": 0.307,
+      "step": 856
+    },
+    {
+      "epoch": 0.48038116591928254,
+      "grad_norm": 0.08748965260811624,
+      "learning_rate": 0.00019781566415286812,
+      "loss": 0.2999,
+      "step": 857
+    },
+    {
+      "epoch": 0.48094170403587444,
+      "grad_norm": 0.08510434026271985,
+      "learning_rate": 0.00019780208353736495,
+      "loss": 0.299,
+      "step": 858
+    },
+    {
+      "epoch": 0.48150224215246634,
+      "grad_norm": 0.08667037835816382,
+      "learning_rate": 0.00019778846130446792,
+      "loss": 0.2967,
+      "step": 859
+    },
+    {
+      "epoch": 0.4820627802690583,
+      "grad_norm": 0.08547251398480658,
+      "learning_rate": 0.00019777479745997366,
+      "loss": 0.3007,
+      "step": 860
+    },
+    {
+      "epoch": 0.4826233183856502,
+      "grad_norm": 0.08527076483126479,
+      "learning_rate": 0.0001977610920096965,
+      "loss": 0.3066,
+      "step": 861
+    },
+    {
+      "epoch": 0.48318385650224216,
+      "grad_norm": 0.0840641801252334,
+      "learning_rate": 0.0001977473449594685,
+      "loss": 0.303,
+      "step": 862
+    },
+    {
+      "epoch": 0.48374439461883406,
+      "grad_norm": 0.08931134590055506,
+      "learning_rate": 0.00019773355631513942,
+      "loss": 0.3126,
+      "step": 863
+    },
+    {
+      "epoch": 0.484304932735426,
+      "grad_norm": 0.08820459512945028,
+      "learning_rate": 0.00019771972608257659,
+      "loss": 0.3026,
+      "step": 864
+    },
+    {
+      "epoch": 0.4848654708520179,
+      "grad_norm": 0.08414892835297764,
+      "learning_rate": 0.00019770585426766527,
+      "loss": 0.303,
+      "step": 865
+    },
+    {
+      "epoch": 0.4854260089686099,
+      "grad_norm": 0.08554246377401148,
+      "learning_rate": 0.00019769194087630818,
+      "loss": 0.2914,
+      "step": 866
+    },
+    {
+      "epoch": 0.4859865470852018,
+      "grad_norm": 0.08733454850313473,
+      "learning_rate": 0.0001976779859144259,
+      "loss": 0.3,
+      "step": 867
+    },
+    {
+      "epoch": 0.48654708520179374,
+      "grad_norm": 0.08234421007740982,
+      "learning_rate": 0.00019766398938795662,
+      "loss": 0.2991,
+      "step": 868
+    },
+    {
+      "epoch": 0.48710762331838564,
+      "grad_norm": 0.08306767371159449,
+      "learning_rate": 0.00019764995130285625,
+      "loss": 0.3053,
+      "step": 869
+    },
+    {
+      "epoch": 0.4876681614349776,
+      "grad_norm": 0.0870799104666167,
+      "learning_rate": 0.00019763587166509835,
+      "loss": 0.3016,
+      "step": 870
+    },
+    {
+      "epoch": 0.4882286995515695,
+      "grad_norm": 0.08881348677675693,
+      "learning_rate": 0.0001976217504806742,
+      "loss": 0.2944,
+      "step": 871
+    },
+    {
+      "epoch": 0.48878923766816146,
+      "grad_norm": 0.0881222931146758,
+      "learning_rate": 0.00019760758775559274,
+      "loss": 0.313,
+      "step": 872
+    },
+    {
+      "epoch": 0.48934977578475336,
+      "grad_norm": 0.07774576318742801,
+      "learning_rate": 0.00019759338349588054,
+      "loss": 0.2827,
+      "step": 873
+    },
+    {
+      "epoch": 0.4899103139013453,
+      "grad_norm": 0.08820544631924236,
+      "learning_rate": 0.00019757913770758196,
+      "loss": 0.2971,
+      "step": 874
+    },
+    {
+      "epoch": 0.4904708520179372,
+      "grad_norm": 0.0832526339338868,
+      "learning_rate": 0.0001975648503967589,
+      "loss": 0.2941,
+      "step": 875
+    },
+    {
+      "epoch": 0.4910313901345291,
+      "grad_norm": 0.08545950056971909,
+      "learning_rate": 0.00019755052156949105,
+      "loss": 0.311,
+      "step": 876
+    },
+    {
+      "epoch": 0.4915919282511211,
+      "grad_norm": 0.08417839227555864,
+      "learning_rate": 0.00019753615123187568,
+      "loss": 0.3109,
+      "step": 877
+    },
+    {
+      "epoch": 0.492152466367713,
+      "grad_norm": 0.08054406355358931,
+      "learning_rate": 0.00019752173939002776,
+      "loss": 0.3008,
+      "step": 878
+    },
+    {
+      "epoch": 0.49271300448430494,
+      "grad_norm": 0.08678396311747016,
+      "learning_rate": 0.0001975072860500799,
+      "loss": 0.2913,
+      "step": 879
+    },
+    {
+      "epoch": 0.49327354260089684,
+      "grad_norm": 0.08665877370645897,
+      "learning_rate": 0.00019749279121818235,
+      "loss": 0.2974,
+      "step": 880
+    },
+    {
+      "epoch": 0.4938340807174888,
+      "grad_norm": 0.08501744931535897,
+      "learning_rate": 0.00019747825490050314,
+      "loss": 0.3122,
+      "step": 881
+    },
+    {
+      "epoch": 0.4943946188340807,
+      "grad_norm": 0.0849881245549837,
+      "learning_rate": 0.00019746367710322778,
+      "loss": 0.3032,
+      "step": 882
+    },
+    {
+      "epoch": 0.49495515695067266,
+      "grad_norm": 0.0843430176577945,
+      "learning_rate": 0.00019744905783255953,
+      "loss": 0.2987,
+      "step": 883
+    },
+    {
+      "epoch": 0.49551569506726456,
+      "grad_norm": 0.0861446110677906,
+      "learning_rate": 0.0001974343970947193,
+      "loss": 0.2856,
+      "step": 884
+    },
+    {
+      "epoch": 0.4960762331838565,
+      "grad_norm": 0.09074356324891023,
+      "learning_rate": 0.0001974196948959456,
+      "loss": 0.3141,
+      "step": 885
+    },
+    {
+      "epoch": 0.4966367713004484,
+      "grad_norm": 0.08187514785841825,
+      "learning_rate": 0.0001974049512424946,
+      "loss": 0.3033,
+      "step": 886
+    },
+    {
+      "epoch": 0.4971973094170404,
+      "grad_norm": 0.08379922826990938,
+      "learning_rate": 0.00019739016614064018,
+      "loss": 0.3116,
+      "step": 887
+    },
+    {
+      "epoch": 0.4977578475336323,
+      "grad_norm": 0.08078480933878579,
+      "learning_rate": 0.0001973753395966737,
+      "loss": 0.3014,
+      "step": 888
+    },
+    {
+      "epoch": 0.49831838565022424,
+      "grad_norm": 0.08269206583626729,
+      "learning_rate": 0.00019736047161690435,
+      "loss": 0.2912,
+      "step": 889
+    },
+    {
+      "epoch": 0.49887892376681614,
+      "grad_norm": 0.08455761204795444,
+      "learning_rate": 0.00019734556220765877,
+      "loss": 0.301,
+      "step": 890
+    },
+    {
+      "epoch": 0.4994394618834081,
+      "grad_norm": 0.0883603760082935,
+      "learning_rate": 0.00019733061137528136,
+      "loss": 0.3056,
+      "step": 891
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.08345844537621519,
+      "learning_rate": 0.00019731561912613406,
+      "loss": 0.2918,
+      "step": 892
+    },
+    {
+      "epoch": 0.500560538116592,
+      "grad_norm": 0.08379990914186539,
+      "learning_rate": 0.00019730058546659653,
+      "loss": 0.3041,
+      "step": 893
+    },
+    {
+      "epoch": 0.5011210762331838,
+      "grad_norm": 0.08126863320078369,
+      "learning_rate": 0.00019728551040306593,
+      "loss": 0.3049,
+      "step": 894
+    },
+    {
+      "epoch": 0.5016816143497758,
+      "grad_norm": 0.08687314500462376,
+      "learning_rate": 0.0001972703939419571,
+      "loss": 0.3043,
+      "step": 895
+    },
+    {
+      "epoch": 0.5022421524663677,
+      "grad_norm": 0.0890477112161628,
+      "learning_rate": 0.00019725523608970255,
+      "loss": 0.3097,
+      "step": 896
+    },
+    {
+      "epoch": 0.5028026905829597,
+      "grad_norm": 0.08299004293296505,
+      "learning_rate": 0.00019724003685275235,
+      "loss": 0.3032,
+      "step": 897
+    },
+    {
+      "epoch": 0.5033632286995515,
+      "grad_norm": 0.08938387229759379,
+      "learning_rate": 0.00019722479623757413,
+      "loss": 0.2923,
+      "step": 898
+    },
+    {
+      "epoch": 0.5039237668161435,
+      "grad_norm": 0.0862274235241268,
+      "learning_rate": 0.00019720951425065318,
+      "loss": 0.2916,
+      "step": 899
+    },
+    {
+      "epoch": 0.5044843049327354,
+      "grad_norm": 0.08413531724910275,
+      "learning_rate": 0.00019719419089849247,
+      "loss": 0.3093,
+      "step": 900
+    },
+    {
+      "epoch": 0.5050448430493274,
+      "grad_norm": 0.08505393388446382,
+      "learning_rate": 0.0001971788261876124,
+      "loss": 0.2996,
+      "step": 901
+    },
+    {
+      "epoch": 0.5056053811659192,
+      "grad_norm": 0.08595061886024928,
+      "learning_rate": 0.00019716342012455112,
+      "loss": 0.2948,
+      "step": 902
+    },
+    {
+      "epoch": 0.5061659192825112,
+      "grad_norm": 0.086781454846237,
+      "learning_rate": 0.00019714797271586432,
+      "loss": 0.3035,
+      "step": 903
+    },
+    {
+      "epoch": 0.5067264573991032,
+      "grad_norm": 0.08615481383556181,
+      "learning_rate": 0.00019713248396812524,
+      "loss": 0.3002,
+      "step": 904
+    },
+    {
+      "epoch": 0.5072869955156951,
+      "grad_norm": 0.08185668542667167,
+      "learning_rate": 0.0001971169538879248,
+      "loss": 0.2985,
+      "step": 905
+    },
+    {
+      "epoch": 0.507847533632287,
+      "grad_norm": 0.08487130316266801,
+      "learning_rate": 0.00019710138248187143,
+      "loss": 0.3145,
+      "step": 906
+    },
+    {
+      "epoch": 0.5084080717488789,
+      "grad_norm": 0.08473969458264732,
+      "learning_rate": 0.00019708576975659123,
+      "loss": 0.2968,
+      "step": 907
+    },
+    {
+      "epoch": 0.5089686098654709,
+      "grad_norm": 0.08487478002006031,
+      "learning_rate": 0.00019707011571872777,
+      "loss": 0.3149,
+      "step": 908
+    },
+    {
+      "epoch": 0.5095291479820628,
+      "grad_norm": 0.07891403399740557,
+      "learning_rate": 0.0001970544203749423,
+      "loss": 0.3028,
+      "step": 909
+    },
+    {
+      "epoch": 0.5100896860986547,
+      "grad_norm": 0.08186472706182046,
+      "learning_rate": 0.00019703868373191358,
+      "loss": 0.31,
+      "step": 910
+    },
+    {
+      "epoch": 0.5106502242152466,
+      "grad_norm": 0.08271364945301952,
+      "learning_rate": 0.00019702290579633799,
+      "loss": 0.2998,
+      "step": 911
+    },
+    {
+      "epoch": 0.5112107623318386,
+      "grad_norm": 0.08408491039003725,
+      "learning_rate": 0.00019700708657492948,
+      "loss": 0.2946,
+      "step": 912
+    },
+    {
+      "epoch": 0.5117713004484304,
+      "grad_norm": 0.08217553132602926,
+      "learning_rate": 0.0001969912260744195,
+      "loss": 0.2966,
+      "step": 913
+    },
+    {
+      "epoch": 0.5123318385650224,
+      "grad_norm": 0.08527351868710076,
+      "learning_rate": 0.00019697532430155716,
+      "loss": 0.2953,
+      "step": 914
+    },
+    {
+      "epoch": 0.5128923766816144,
+      "grad_norm": 0.08430499254522876,
+      "learning_rate": 0.00019695938126310908,
+      "loss": 0.3144,
+      "step": 915
+    },
+    {
+      "epoch": 0.5134529147982063,
+      "grad_norm": 0.08747472712751019,
+      "learning_rate": 0.00019694339696585942,
+      "loss": 0.2923,
+      "step": 916
+    },
+    {
+      "epoch": 0.5140134529147982,
+      "grad_norm": 0.0838268115142445,
+      "learning_rate": 0.00019692737141660996,
+      "loss": 0.3056,
+      "step": 917
+    },
+    {
+      "epoch": 0.5145739910313901,
+      "grad_norm": 0.08324548511507619,
+      "learning_rate": 0.00019691130462217996,
+      "loss": 0.2978,
+      "step": 918
+    },
+    {
+      "epoch": 0.5151345291479821,
+      "grad_norm": 0.0856078729888167,
+      "learning_rate": 0.0001968951965894063,
+      "loss": 0.3045,
+      "step": 919
+    },
+    {
+      "epoch": 0.515695067264574,
+      "grad_norm": 0.07967163778859834,
+      "learning_rate": 0.0001968790473251434,
+      "loss": 0.2954,
+      "step": 920
+    },
+    {
+      "epoch": 0.5162556053811659,
+      "grad_norm": 0.08354138058819514,
+      "learning_rate": 0.00019686285683626314,
+      "loss": 0.2843,
+      "step": 921
+    },
+    {
+      "epoch": 0.5168161434977578,
+      "grad_norm": 0.07968434796052404,
+      "learning_rate": 0.00019684662512965505,
+      "loss": 0.2966,
+      "step": 922
+    },
+    {
+      "epoch": 0.5173766816143498,
+      "grad_norm": 0.08155084666307773,
+      "learning_rate": 0.00019683035221222618,
+      "loss": 0.303,
+      "step": 923
+    },
+    {
+      "epoch": 0.5179372197309418,
+      "grad_norm": 0.08600324315161532,
+      "learning_rate": 0.00019681403809090097,
+      "loss": 0.329,
+      "step": 924
+    },
+    {
+      "epoch": 0.5184977578475336,
+      "grad_norm": 0.08025233555071982,
+      "learning_rate": 0.00019679768277262164,
+      "loss": 0.2973,
+      "step": 925
+    },
+    {
+      "epoch": 0.5190582959641256,
+      "grad_norm": 0.08371492254683732,
+      "learning_rate": 0.00019678128626434777,
+      "loss": 0.2909,
+      "step": 926
+    },
+    {
+      "epoch": 0.5196188340807175,
+      "grad_norm": 0.08506849111333878,
+      "learning_rate": 0.00019676484857305654,
+      "loss": 0.3015,
+      "step": 927
+    },
+    {
+      "epoch": 0.5201793721973094,
+      "grad_norm": 0.0831534850052501,
+      "learning_rate": 0.00019674836970574254,
+      "loss": 0.3045,
+      "step": 928
+    },
+    {
+      "epoch": 0.5207399103139013,
+      "grad_norm": 0.08092206783572063,
+      "learning_rate": 0.00019673184966941803,
+      "loss": 0.2963,
+      "step": 929
+    },
+    {
+      "epoch": 0.5213004484304933,
+      "grad_norm": 0.08408705293952237,
+      "learning_rate": 0.00019671528847111275,
+      "loss": 0.2879,
+      "step": 930
+    },
+    {
+      "epoch": 0.5218609865470852,
+      "grad_norm": 0.08576490959085466,
+      "learning_rate": 0.00019669868611787387,
+      "loss": 0.2994,
+      "step": 931
+    },
+    {
+      "epoch": 0.5224215246636771,
+      "grad_norm": 0.08854960466302186,
+      "learning_rate": 0.00019668204261676618,
+      "loss": 0.3071,
+      "step": 932
+    },
+    {
+      "epoch": 0.522982062780269,
+      "grad_norm": 0.08110885464158614,
+      "learning_rate": 0.00019666535797487194,
+      "loss": 0.2964,
+      "step": 933
+    },
+    {
+      "epoch": 0.523542600896861,
+      "grad_norm": 0.08031550556933642,
+      "learning_rate": 0.00019664863219929086,
+      "loss": 0.2966,
+      "step": 934
+    },
+    {
+      "epoch": 0.524103139013453,
+      "grad_norm": 0.08238932652082077,
+      "learning_rate": 0.0001966318652971402,
+      "loss": 0.3089,
+      "step": 935
+    },
+    {
+      "epoch": 0.5246636771300448,
+      "grad_norm": 0.08422789627755416,
+      "learning_rate": 0.00019661505727555482,
+      "loss": 0.303,
+      "step": 936
+    },
+    {
+      "epoch": 0.5252242152466368,
+      "grad_norm": 0.08734140193215176,
+      "learning_rate": 0.0001965982081416869,
+      "loss": 0.3074,
+      "step": 937
+    },
+    {
+      "epoch": 0.5257847533632287,
+      "grad_norm": 0.08127075155698334,
+      "learning_rate": 0.0001965813179027062,
+      "loss": 0.297,
+      "step": 938
+    },
+    {
+      "epoch": 0.5263452914798207,
+      "grad_norm": 0.08500582631589884,
+      "learning_rate": 0.00019656438656579997,
+      "loss": 0.2927,
+      "step": 939
+    },
+    {
+      "epoch": 0.5269058295964125,
+      "grad_norm": 0.08228510145684727,
+      "learning_rate": 0.00019654741413817296,
+      "loss": 0.2939,
+      "step": 940
+    },
+    {
+      "epoch": 0.5274663677130045,
+      "grad_norm": 0.08647748094621052,
+      "learning_rate": 0.00019653040062704737,
+      "loss": 0.2997,
+      "step": 941
+    },
+    {
+      "epoch": 0.5280269058295964,
+      "grad_norm": 0.07839774222368344,
+      "learning_rate": 0.00019651334603966295,
+      "loss": 0.3014,
+      "step": 942
+    },
+    {
+      "epoch": 0.5285874439461884,
+      "grad_norm": 0.08693378206066758,
+      "learning_rate": 0.00019649625038327683,
+      "loss": 0.3021,
+      "step": 943
+    },
+    {
+      "epoch": 0.5291479820627802,
+      "grad_norm": 0.08215386654261363,
+      "learning_rate": 0.0001964791136651637,
+      "loss": 0.3067,
+      "step": 944
+    },
+    {
+      "epoch": 0.5297085201793722,
+      "grad_norm": 0.08369155709381948,
+      "learning_rate": 0.00019646193589261565,
+      "loss": 0.3126,
+      "step": 945
+    },
+    {
+      "epoch": 0.5302690582959642,
+      "grad_norm": 0.08236756787936445,
+      "learning_rate": 0.00019644471707294233,
+      "loss": 0.3179,
+      "step": 946
+    },
+    {
+      "epoch": 0.530829596412556,
+      "grad_norm": 0.08268734154037605,
+      "learning_rate": 0.00019642745721347077,
+      "loss": 0.2978,
+      "step": 947
+    },
+    {
+      "epoch": 0.531390134529148,
+      "grad_norm": 0.08359759524474471,
+      "learning_rate": 0.00019641015632154552,
+      "loss": 0.3032,
+      "step": 948
+    },
+    {
+      "epoch": 0.5319506726457399,
+      "grad_norm": 0.08415225130718881,
+      "learning_rate": 0.00019639281440452856,
+      "loss": 0.2989,
+      "step": 949
+    },
+    {
+      "epoch": 0.5325112107623319,
+      "grad_norm": 0.08454145862518143,
+      "learning_rate": 0.00019637543146979939,
+      "loss": 0.2933,
+      "step": 950
+    },
+    {
+      "epoch": 0.5330717488789237,
+      "grad_norm": 0.08178203425002459,
+      "learning_rate": 0.0001963580075247548,
+      "loss": 0.3045,
+      "step": 951
+    },
+    {
+      "epoch": 0.5336322869955157,
+      "grad_norm": 0.08121384674961249,
+      "learning_rate": 0.00019634054257680923,
+      "loss": 0.292,
+      "step": 952
+    },
+    {
+      "epoch": 0.5341928251121076,
+      "grad_norm": 0.08281600642346805,
+      "learning_rate": 0.00019632303663339444,
+      "loss": 0.3008,
+      "step": 953
+    },
+    {
+      "epoch": 0.5347533632286996,
+      "grad_norm": 0.08062824486513642,
+      "learning_rate": 0.00019630548970195975,
+      "loss": 0.2911,
+      "step": 954
+    },
+    {
+      "epoch": 0.5353139013452914,
+      "grad_norm": 0.082421799974641,
+      "learning_rate": 0.00019628790178997173,
+      "loss": 0.2964,
+      "step": 955
+    },
+    {
+      "epoch": 0.5358744394618834,
+      "grad_norm": 0.08450222653145384,
+      "learning_rate": 0.00019627027290491458,
+      "loss": 0.3043,
+      "step": 956
+    },
+    {
+      "epoch": 0.5364349775784754,
+      "grad_norm": 0.08761895213193802,
+      "learning_rate": 0.00019625260305428989,
+      "loss": 0.3059,
+      "step": 957
+    },
+    {
+      "epoch": 0.5369955156950673,
+      "grad_norm": 0.08492338268252649,
+      "learning_rate": 0.00019623489224561657,
+      "loss": 0.2994,
+      "step": 958
+    },
+    {
+      "epoch": 0.5375560538116592,
+      "grad_norm": 0.08135785679281451,
+      "learning_rate": 0.0001962171404864311,
+      "loss": 0.2966,
+      "step": 959
+    },
+    {
+      "epoch": 0.5381165919282511,
+      "grad_norm": 0.08044172378823879,
+      "learning_rate": 0.0001961993477842873,
+      "loss": 0.2983,
+      "step": 960
+    },
+    {
+      "epoch": 0.5386771300448431,
+      "grad_norm": 0.08185763804368876,
+      "learning_rate": 0.00019618151414675644,
+      "loss": 0.3054,
+      "step": 961
+    },
+    {
+      "epoch": 0.5392376681614349,
+      "grad_norm": 0.08366828054161651,
+      "learning_rate": 0.00019616363958142722,
+      "loss": 0.2985,
+      "step": 962
+    },
+    {
+      "epoch": 0.5397982062780269,
+      "grad_norm": 0.08518406054877001,
+      "learning_rate": 0.00019614572409590574,
+      "loss": 0.3043,
+      "step": 963
+    },
+    {
+      "epoch": 0.5403587443946188,
+      "grad_norm": 0.07975472875203916,
+      "learning_rate": 0.00019612776769781554,
+      "loss": 0.3046,
+      "step": 964
+    },
+    {
+      "epoch": 0.5409192825112108,
+      "grad_norm": 0.08010969525048041,
+      "learning_rate": 0.00019610977039479746,
+      "loss": 0.3025,
+      "step": 965
+    },
+    {
+      "epoch": 0.5414798206278026,
+      "grad_norm": 0.08372359413046519,
+      "learning_rate": 0.00019609173219450998,
+      "loss": 0.2944,
+      "step": 966
+    },
+    {
+      "epoch": 0.5420403587443946,
+      "grad_norm": 0.08073693338331876,
+      "learning_rate": 0.00019607365310462868,
+      "loss": 0.2926,
+      "step": 967
+    },
+    {
+      "epoch": 0.5426008968609866,
+      "grad_norm": 0.08239761710797047,
+      "learning_rate": 0.0001960555331328468,
+      "loss": 0.2919,
+      "step": 968
+    },
+    {
+      "epoch": 0.5431614349775785,
+      "grad_norm": 0.08080189571564042,
+      "learning_rate": 0.0001960373722868748,
+      "loss": 0.288,
+      "step": 969
+    },
+    {
+      "epoch": 0.5437219730941704,
+      "grad_norm": 0.0836435259389605,
+      "learning_rate": 0.00019601917057444072,
+      "loss": 0.2961,
+      "step": 970
+    },
+    {
+      "epoch": 0.5442825112107623,
+      "grad_norm": 0.08086009589923894,
+      "learning_rate": 0.0001960009280032897,
+      "loss": 0.2924,
+      "step": 971
+    },
+    {
+      "epoch": 0.5448430493273543,
+      "grad_norm": 0.08296429379715402,
+      "learning_rate": 0.00019598264458118458,
+      "loss": 0.2983,
+      "step": 972
+    },
+    {
+      "epoch": 0.5454035874439462,
+      "grad_norm": 0.08308733138527796,
+      "learning_rate": 0.0001959643203159054,
+      "loss": 0.2903,
+      "step": 973
+    },
+    {
+      "epoch": 0.5459641255605381,
+      "grad_norm": 0.07872166292022645,
+      "learning_rate": 0.0001959459552152496,
+      "loss": 0.2929,
+      "step": 974
+    },
+    {
+      "epoch": 0.54652466367713,
+      "grad_norm": 0.07986976819269802,
+      "learning_rate": 0.00019592754928703205,
+      "loss": 0.301,
+      "step": 975
+    },
+    {
+      "epoch": 0.547085201793722,
+      "grad_norm": 0.08205647747003766,
+      "learning_rate": 0.00019590910253908494,
+      "loss": 0.2976,
+      "step": 976
+    },
+    {
+      "epoch": 0.547645739910314,
+      "grad_norm": 0.0839728790871989,
+      "learning_rate": 0.0001958906149792579,
+      "loss": 0.3089,
+      "step": 977
+    },
+    {
+      "epoch": 0.5482062780269058,
+      "grad_norm": 0.0803251806561804,
+      "learning_rate": 0.00019587208661541784,
+      "loss": 0.3064,
+      "step": 978
+    },
+    {
+      "epoch": 0.5487668161434978,
+      "grad_norm": 0.07985275513420223,
+      "learning_rate": 0.00019585351745544905,
+      "loss": 0.2882,
+      "step": 979
+    },
+    {
+      "epoch": 0.5493273542600897,
+      "grad_norm": 0.08140438201801921,
+      "learning_rate": 0.00019583490750725325,
+      "loss": 0.3104,
+      "step": 980
+    },
+    {
+      "epoch": 0.5498878923766816,
+      "grad_norm": 0.08307980787136376,
+      "learning_rate": 0.00019581625677874944,
+      "loss": 0.298,
+      "step": 981
+    },
+    {
+      "epoch": 0.5504484304932735,
+      "grad_norm": 0.0797856319440976,
+      "learning_rate": 0.00019579756527787404,
+      "loss": 0.3056,
+      "step": 982
+    },
+    {
+      "epoch": 0.5510089686098655,
+      "grad_norm": 0.080113786389902,
+      "learning_rate": 0.0001957788330125807,
+      "loss": 0.2981,
+      "step": 983
+    },
+    {
+      "epoch": 0.5515695067264574,
+      "grad_norm": 0.08553049329528893,
+      "learning_rate": 0.0001957600599908406,
+      "loss": 0.2968,
+      "step": 984
+    },
+    {
+      "epoch": 0.5521300448430493,
+      "grad_norm": 0.08937914865124305,
+      "learning_rate": 0.00019574124622064208,
+      "loss": 0.3042,
+      "step": 985
+    },
+    {
+      "epoch": 0.5526905829596412,
+      "grad_norm": 0.0808160952842041,
+      "learning_rate": 0.00019572239170999098,
+      "loss": 0.2942,
+      "step": 986
+    },
+    {
+      "epoch": 0.5532511210762332,
+      "grad_norm": 0.08097659068475872,
+      "learning_rate": 0.00019570349646691034,
+      "loss": 0.2847,
+      "step": 987
+    },
+    {
+      "epoch": 0.5538116591928252,
+      "grad_norm": 0.08091683580684034,
+      "learning_rate": 0.0001956845604994406,
+      "loss": 0.2972,
+      "step": 988
+    },
+    {
+      "epoch": 0.554372197309417,
+      "grad_norm": 0.07621229573195737,
+      "learning_rate": 0.0001956655838156395,
+      "loss": 0.2955,
+      "step": 989
+    },
+    {
+      "epoch": 0.554932735426009,
+      "grad_norm": 0.08338821667309904,
+      "learning_rate": 0.00019564656642358217,
+      "loss": 0.2993,
+      "step": 990
+    },
+    {
+      "epoch": 0.5554932735426009,
+      "grad_norm": 0.0820814226050958,
+      "learning_rate": 0.00019562750833136097,
+      "loss": 0.2948,
+      "step": 991
+    },
+    {
+      "epoch": 0.5560538116591929,
+      "grad_norm": 0.08424537008272272,
+      "learning_rate": 0.00019560840954708565,
+      "loss": 0.2931,
+      "step": 992
+    },
+    {
+      "epoch": 0.5566143497757847,
+      "grad_norm": 0.08630715850803786,
+      "learning_rate": 0.00019558927007888328,
+      "loss": 0.2893,
+      "step": 993
+    },
+    {
+      "epoch": 0.5571748878923767,
+      "grad_norm": 0.08787717405940232,
+      "learning_rate": 0.00019557008993489815,
+      "loss": 0.3017,
+      "step": 994
+    },
+    {
+      "epoch": 0.5577354260089686,
+      "grad_norm": 0.08219811787822397,
+      "learning_rate": 0.00019555086912329198,
+      "loss": 0.2987,
+      "step": 995
+    },
+    {
+      "epoch": 0.5582959641255605,
+      "grad_norm": 0.08257802571035627,
+      "learning_rate": 0.00019553160765224372,
+      "loss": 0.2968,
+      "step": 996
+    },
+    {
+      "epoch": 0.5588565022421524,
+      "grad_norm": 0.08232718737486464,
+      "learning_rate": 0.0001955123055299496,
+      "loss": 0.2927,
+      "step": 997
+    },
+    {
+      "epoch": 0.5594170403587444,
+      "grad_norm": 0.08472102275377649,
+      "learning_rate": 0.00019549296276462325,
+      "loss": 0.2967,
+      "step": 998
+    },
+    {
+      "epoch": 0.5599775784753364,
+      "grad_norm": 0.08162396782133255,
+      "learning_rate": 0.0001954735793644955,
+      "loss": 0.3083,
+      "step": 999
+    },
+    {
+      "epoch": 0.5605381165919282,
+      "grad_norm": 0.08124050128833456,
+      "learning_rate": 0.0001954541553378145,
+      "loss": 0.2942,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5610986547085202,
+      "grad_norm": 0.08292789662007756,
+      "learning_rate": 0.00019543469069284572,
+      "loss": 0.3047,
+      "step": 1001
+    },
+    {
+      "epoch": 0.5616591928251121,
+      "grad_norm": 0.0802762167797529,
+      "learning_rate": 0.00019541518543787184,
+      "loss": 0.2997,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5622197309417041,
+      "grad_norm": 0.08099010862176718,
+      "learning_rate": 0.00019539563958119292,
+      "loss": 0.3083,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5627802690582959,
+      "grad_norm": 0.08287263273356754,
+      "learning_rate": 0.0001953760531311262,
+      "loss": 0.3062,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5633408071748879,
+      "grad_norm": 0.0822128744799154,
+      "learning_rate": 0.00019535642609600623,
+      "loss": 0.2965,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5639013452914798,
+      "grad_norm": 0.08091755046384955,
+      "learning_rate": 0.00019533675848418488,
+      "loss": 0.2865,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5644618834080718,
+      "grad_norm": 0.0824953280005843,
+      "learning_rate": 0.00019531705030403123,
+      "loss": 0.3058,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5650224215246636,
+      "grad_norm": 0.08139511280822345,
+      "learning_rate": 0.0001952973015639316,
+      "loss": 0.3009,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5655829596412556,
+      "grad_norm": 0.08380566540204436,
+      "learning_rate": 0.00019527751227228963,
+      "loss": 0.3094,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5661434977578476,
+      "grad_norm": 0.08396668460521664,
+      "learning_rate": 0.0001952576824375262,
+      "loss": 0.3011,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5667040358744395,
+      "grad_norm": 0.08324030871087482,
+      "learning_rate": 0.00019523781206807944,
+      "loss": 0.2996,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5672645739910314,
+      "grad_norm": 0.08074950884817227,
+      "learning_rate": 0.0001952179011724047,
+      "loss": 0.3091,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5678251121076233,
+      "grad_norm": 0.08156566020093663,
+      "learning_rate": 0.0001951979497589746,
+      "loss": 0.3074,
+      "step": 1013
+    },
+    {
+      "epoch": 0.5683856502242153,
+      "grad_norm": 0.08130419215015418,
+      "learning_rate": 0.000195177957836279,
+      "loss": 0.3047,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5689461883408071,
+      "grad_norm": 0.07816132167890519,
+      "learning_rate": 0.00019515792541282504,
+      "loss": 0.2919,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5695067264573991,
+      "grad_norm": 0.08065245102568831,
+      "learning_rate": 0.00019513785249713697,
+      "loss": 0.2886,
+      "step": 1016
+    },
+    {
+      "epoch": 0.570067264573991,
+      "grad_norm": 0.0811814245119974,
+      "learning_rate": 0.00019511773909775638,
+      "loss": 0.2964,
+      "step": 1017
+    },
+    {
+      "epoch": 0.570627802690583,
+      "grad_norm": 0.08086825544168999,
+      "learning_rate": 0.00019509758522324208,
+      "loss": 0.2962,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5711883408071748,
+      "grad_norm": 0.07905304183732399,
+      "learning_rate": 0.00019507739088217007,
+      "loss": 0.2919,
+      "step": 1019
+    },
+    {
+      "epoch": 0.5717488789237668,
+      "grad_norm": 0.08154893266510474,
+      "learning_rate": 0.00019505715608313359,
+      "loss": 0.294,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5723094170403588,
+      "grad_norm": 0.08621625302958354,
+      "learning_rate": 0.00019503688083474306,
+      "loss": 0.3072,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5728699551569507,
+      "grad_norm": 0.08102355358650781,
+      "learning_rate": 0.00019501656514562616,
+      "loss": 0.2977,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5734304932735426,
+      "grad_norm": 0.08352356346903747,
+      "learning_rate": 0.00019499620902442777,
+      "loss": 0.2973,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5739910313901345,
+      "grad_norm": 0.08645298103063442,
+      "learning_rate": 0.00019497581247980992,
+      "loss": 0.3009,
+      "step": 1024
+    },
+    {
+      "epoch": 0.5745515695067265,
+      "grad_norm": 0.08215864116992247,
+      "learning_rate": 0.0001949553755204519,
+      "loss": 0.2986,
+      "step": 1025
+    },
+    {
+      "epoch": 0.5751121076233184,
+      "grad_norm": 0.08300242923574697,
+      "learning_rate": 0.00019493489815505018,
+      "loss": 0.2979,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5756726457399103,
+      "grad_norm": 0.08274552344212312,
+      "learning_rate": 0.00019491438039231847,
+      "loss": 0.2897,
+      "step": 1027
+    },
+    {
+      "epoch": 0.5762331838565022,
+      "grad_norm": 0.08198176538321253,
+      "learning_rate": 0.0001948938222409876,
+      "loss": 0.2911,
+      "step": 1028
+    },
+    {
+      "epoch": 0.5767937219730942,
+      "grad_norm": 0.08268500303931227,
+      "learning_rate": 0.00019487322370980557,
+      "loss": 0.2982,
+      "step": 1029
+    },
+    {
+      "epoch": 0.577354260089686,
+      "grad_norm": 0.08292390243228843,
+      "learning_rate": 0.00019485258480753763,
+      "loss": 0.2953,
+      "step": 1030
+    },
+    {
+      "epoch": 0.577914798206278,
+      "grad_norm": 0.08017134599658629,
+      "learning_rate": 0.0001948319055429662,
+      "loss": 0.2978,
+      "step": 1031
+    },
+    {
+      "epoch": 0.57847533632287,
+      "grad_norm": 0.07872226328959364,
+      "learning_rate": 0.00019481118592489086,
+      "loss": 0.2938,
+      "step": 1032
+    },
+    {
+      "epoch": 0.5790358744394619,
+      "grad_norm": 0.08151729117815305,
+      "learning_rate": 0.0001947904259621283,
+      "loss": 0.3008,
+      "step": 1033
+    },
+    {
+      "epoch": 0.5795964125560538,
+      "grad_norm": 0.0827677338829837,
+      "learning_rate": 0.0001947696256635125,
+      "loss": 0.3031,
+      "step": 1034
+    },
+    {
+      "epoch": 0.5801569506726457,
+      "grad_norm": 0.08327695153084624,
+      "learning_rate": 0.00019474878503789457,
+      "loss": 0.2808,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5807174887892377,
+      "grad_norm": 0.08225192058168172,
+      "learning_rate": 0.00019472790409414266,
+      "loss": 0.2992,
+      "step": 1036
+    },
+    {
+      "epoch": 0.5812780269058296,
+      "grad_norm": 0.08073922503190531,
+      "learning_rate": 0.00019470698284114221,
+      "loss": 0.2888,
+      "step": 1037
+    },
+    {
+      "epoch": 0.5818385650224215,
+      "grad_norm": 0.08066445833016037,
+      "learning_rate": 0.0001946860212877958,
+      "loss": 0.2967,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5823991031390134,
+      "grad_norm": 0.082569667570103,
+      "learning_rate": 0.0001946650194430231,
+      "loss": 0.2997,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5829596412556054,
+      "grad_norm": 0.07835736029464292,
+      "learning_rate": 0.00019464397731576094,
+      "loss": 0.2891,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5835201793721974,
+      "grad_norm": 0.07887715261284907,
+      "learning_rate": 0.00019462289491496335,
+      "loss": 0.2984,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5840807174887892,
+      "grad_norm": 0.08162349914427701,
+      "learning_rate": 0.0001946017722496014,
+      "loss": 0.2891,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5846412556053812,
+      "grad_norm": 0.08446767920897999,
+      "learning_rate": 0.00019458060932866342,
+      "loss": 0.2769,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5852017937219731,
+      "grad_norm": 0.0809568234632934,
+      "learning_rate": 0.00019455940616115472,
+      "loss": 0.3003,
+      "step": 1044
+    },
+    {
+      "epoch": 0.5857623318385651,
+      "grad_norm": 0.08308651523590681,
+      "learning_rate": 0.00019453816275609786,
+      "loss": 0.3016,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5863228699551569,
+      "grad_norm": 0.0818178078741461,
+      "learning_rate": 0.00019451687912253247,
+      "loss": 0.2997,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5868834080717489,
+      "grad_norm": 0.0807603243720443,
+      "learning_rate": 0.00019449555526951528,
+      "loss": 0.2866,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5874439461883408,
+      "grad_norm": 0.0813055583399256,
+      "learning_rate": 0.00019447419120612017,
+      "loss": 0.281,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5880044843049327,
+      "grad_norm": 0.07993839104960439,
+      "learning_rate": 0.00019445278694143813,
+      "loss": 0.2958,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5885650224215246,
+      "grad_norm": 0.08372831791778379,
+      "learning_rate": 0.00019443134248457727,
+      "loss": 0.3057,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5891255605381166,
+      "grad_norm": 0.07820095062069352,
+      "learning_rate": 0.0001944098578446627,
+      "loss": 0.3036,
+      "step": 1051
+    },
+    {
+      "epoch": 0.5896860986547086,
+      "grad_norm": 0.08196388786253347,
+      "learning_rate": 0.00019438833303083678,
+      "loss": 0.3065,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5902466367713004,
+      "grad_norm": 0.08085677677082478,
+      "learning_rate": 0.00019436676805225885,
+      "loss": 0.3027,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5908071748878924,
+      "grad_norm": 0.08049968008603187,
+      "learning_rate": 0.0001943451629181054,
+      "loss": 0.2936,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5913677130044843,
+      "grad_norm": 0.08110490365762947,
+      "learning_rate": 0.00019432351763756998,
+      "loss": 0.2946,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5919282511210763,
+      "grad_norm": 0.08020434181505288,
+      "learning_rate": 0.00019430183221986325,
+      "loss": 0.2905,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5924887892376681,
+      "grad_norm": 0.08109196937394578,
+      "learning_rate": 0.00019428010667421294,
+      "loss": 0.296,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5930493273542601,
+      "grad_norm": 0.08521408961392643,
+      "learning_rate": 0.0001942583410098638,
+      "loss": 0.3011,
+      "step": 1058
+    },
+    {
+      "epoch": 0.593609865470852,
+      "grad_norm": 0.08178596609634749,
+      "learning_rate": 0.00019423653523607776,
+      "loss": 0.2995,
+      "step": 1059
+    },
+    {
+      "epoch": 0.594170403587444,
+      "grad_norm": 0.08275727913240233,
+      "learning_rate": 0.0001942146893621337,
+      "loss": 0.3057,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5947309417040358,
+      "grad_norm": 0.07869497119472335,
+      "learning_rate": 0.0001941928033973277,
+      "loss": 0.2803,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5952914798206278,
+      "grad_norm": 0.07938151254570681,
+      "learning_rate": 0.00019417087735097276,
+      "loss": 0.3008,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5958520179372198,
+      "grad_norm": 0.08096247717015172,
+      "learning_rate": 0.00019414891123239902,
+      "loss": 0.3009,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5964125560538116,
+      "grad_norm": 0.08188219012976368,
+      "learning_rate": 0.00019412690505095365,
+      "loss": 0.2834,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5969730941704036,
+      "grad_norm": 0.07950204205382994,
+      "learning_rate": 0.00019410485881600083,
+      "loss": 0.3028,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5975336322869955,
+      "grad_norm": 0.07961964154651452,
+      "learning_rate": 0.00019408277253692187,
+      "loss": 0.299,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5980941704035875,
+      "grad_norm": 0.08141812652243936,
+      "learning_rate": 0.00019406064622311503,
+      "loss": 0.3103,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5986547085201793,
+      "grad_norm": 0.08272941016317377,
+      "learning_rate": 0.0001940384798839957,
+      "loss": 0.3019,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5992152466367713,
+      "grad_norm": 0.08134911105415703,
+      "learning_rate": 0.00019401627352899617,
+      "loss": 0.2926,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5997757847533632,
+      "grad_norm": 0.0802752797624018,
+      "learning_rate": 0.00019399402716756593,
+      "loss": 0.2991,
+      "step": 1070
+    },
+    {
+      "epoch": 0.6003363228699552,
+      "grad_norm": 0.08196460690449331,
+      "learning_rate": 0.00019397174080917133,
+      "loss": 0.2874,
+      "step": 1071
+    },
+    {
+      "epoch": 0.600896860986547,
+      "grad_norm": 0.08062758419866864,
+      "learning_rate": 0.00019394941446329583,
+      "loss": 0.2924,
+      "step": 1072
+    },
+    {
+      "epoch": 0.601457399103139,
+      "grad_norm": 0.08103845174553295,
+      "learning_rate": 0.00019392704813943988,
+      "loss": 0.303,
+      "step": 1073
+    },
+    {
+      "epoch": 0.602017937219731,
+      "grad_norm": 0.07931275903974408,
+      "learning_rate": 0.00019390464184712095,
+      "loss": 0.2887,
+      "step": 1074
+    },
+    {
+      "epoch": 0.6025784753363229,
+      "grad_norm": 0.07809852492116702,
+      "learning_rate": 0.00019388219559587352,
+      "loss": 0.2954,
+      "step": 1075
+    },
+    {
+      "epoch": 0.6031390134529148,
+      "grad_norm": 0.07695128462165139,
+      "learning_rate": 0.0001938597093952491,
+      "loss": 0.2937,
+      "step": 1076
+    },
+    {
+      "epoch": 0.6036995515695067,
+      "grad_norm": 0.07968305921044003,
+      "learning_rate": 0.00019383718325481611,
+      "loss": 0.2899,
+      "step": 1077
+    },
+    {
+      "epoch": 0.6042600896860987,
+      "grad_norm": 0.0766371362280237,
+      "learning_rate": 0.00019381461718416003,
+      "loss": 0.2887,
+      "step": 1078
+    },
+    {
+      "epoch": 0.6048206278026906,
+      "grad_norm": 0.08224654145621542,
+      "learning_rate": 0.00019379201119288335,
+      "loss": 0.298,
+      "step": 1079
+    },
+    {
+      "epoch": 0.6053811659192825,
+      "grad_norm": 0.08935417113127495,
+      "learning_rate": 0.00019376936529060554,
+      "loss": 0.2903,
+      "step": 1080
+    },
+    {
+      "epoch": 0.6059417040358744,
+      "grad_norm": 0.07924901485783611,
+      "learning_rate": 0.000193746679486963,
+      "loss": 0.3027,
+      "step": 1081
+    },
+    {
+      "epoch": 0.6065022421524664,
+      "grad_norm": 0.08002093282893484,
+      "learning_rate": 0.00019372395379160912,
+      "loss": 0.2897,
+      "step": 1082
+    },
+    {
+      "epoch": 0.6070627802690582,
+      "grad_norm": 0.07970239275692652,
+      "learning_rate": 0.00019370118821421435,
+      "loss": 0.3055,
+      "step": 1083
+    },
+    {
+      "epoch": 0.6076233183856502,
+      "grad_norm": 0.07952328389630461,
+      "learning_rate": 0.000193678382764466,
+      "loss": 0.2951,
+      "step": 1084
+    },
+    {
+      "epoch": 0.6081838565022422,
+      "grad_norm": 0.07836077662821399,
+      "learning_rate": 0.00019365553745206846,
+      "loss": 0.3058,
+      "step": 1085
+    },
+    {
+      "epoch": 0.6087443946188341,
+      "grad_norm": 0.07837725396868078,
+      "learning_rate": 0.00019363265228674296,
+      "loss": 0.2992,
+      "step": 1086
+    },
+    {
+      "epoch": 0.609304932735426,
+      "grad_norm": 0.07958884714718022,
+      "learning_rate": 0.00019360972727822774,
+      "loss": 0.2788,
+      "step": 1087
+    },
+    {
+      "epoch": 0.6098654708520179,
+      "grad_norm": 0.07932542886598398,
+      "learning_rate": 0.00019358676243627808,
+      "loss": 0.2883,
+      "step": 1088
+    },
+    {
+      "epoch": 0.6104260089686099,
+      "grad_norm": 0.08309045755745338,
+      "learning_rate": 0.00019356375777066604,
+      "loss": 0.3058,
+      "step": 1089
+    },
+    {
+      "epoch": 0.6109865470852018,
+      "grad_norm": 0.08453098433315742,
+      "learning_rate": 0.0001935407132911808,
+      "loss": 0.2861,
+      "step": 1090
+    },
+    {
+      "epoch": 0.6115470852017937,
+      "grad_norm": 0.07726523523099504,
+      "learning_rate": 0.00019351762900762833,
+      "loss": 0.2975,
+      "step": 1091
+    },
+    {
+      "epoch": 0.6121076233183856,
+      "grad_norm": 0.08583559568404842,
+      "learning_rate": 0.00019349450492983164,
+      "loss": 0.2962,
+      "step": 1092
+    },
+    {
+      "epoch": 0.6126681614349776,
+      "grad_norm": 0.08279385055747263,
+      "learning_rate": 0.00019347134106763062,
+      "loss": 0.3077,
+      "step": 1093
+    },
+    {
+      "epoch": 0.6132286995515696,
+      "grad_norm": 0.07913164196006647,
+      "learning_rate": 0.00019344813743088217,
+      "loss": 0.306,
+      "step": 1094
+    },
+    {
+      "epoch": 0.6137892376681614,
+      "grad_norm": 0.08142004526828342,
+      "learning_rate": 0.00019342489402945998,
+      "loss": 0.3104,
+      "step": 1095
+    },
+    {
+      "epoch": 0.6143497757847534,
+      "grad_norm": 0.07725508609340993,
+      "learning_rate": 0.0001934016108732548,
+      "loss": 0.2832,
+      "step": 1096
+    },
+    {
+      "epoch": 0.6149103139013453,
+      "grad_norm": 0.07701666006903832,
+      "learning_rate": 0.0001933782879721742,
+      "loss": 0.3037,
+      "step": 1097
+    },
+    {
+      "epoch": 0.6154708520179372,
+      "grad_norm": 0.0796016432331057,
+      "learning_rate": 0.00019335492533614272,
+      "loss": 0.288,
+      "step": 1098
+    },
+    {
+      "epoch": 0.6160313901345291,
+      "grad_norm": 0.08038386277262886,
+      "learning_rate": 0.00019333152297510176,
+      "loss": 0.2958,
+      "step": 1099
+    },
+    {
+      "epoch": 0.6165919282511211,
+      "grad_norm": 0.08285440899869845,
+      "learning_rate": 0.00019330808089900963,
+      "loss": 0.3047,
+      "step": 1100
+    },
+    {
+      "epoch": 0.617152466367713,
+      "grad_norm": 0.08012111876088956,
+      "learning_rate": 0.00019328459911784163,
+      "loss": 0.2938,
+      "step": 1101
+    },
+    {
+      "epoch": 0.6177130044843049,
+      "grad_norm": 0.08236582150599499,
+      "learning_rate": 0.00019326107764158982,
+      "loss": 0.2902,
+      "step": 1102
+    },
+    {
+      "epoch": 0.6182735426008968,
+      "grad_norm": 0.08255214806492887,
+      "learning_rate": 0.0001932375164802632,
+      "loss": 0.3007,
+      "step": 1103
+    },
+    {
+      "epoch": 0.6188340807174888,
+      "grad_norm": 0.08146766932112492,
+      "learning_rate": 0.00019321391564388775,
+      "loss": 0.2992,
+      "step": 1104
+    },
+    {
+      "epoch": 0.6193946188340808,
+      "grad_norm": 0.08066260480208594,
+      "learning_rate": 0.00019319027514250618,
+      "loss": 0.2795,
+      "step": 1105
+    },
+    {
+      "epoch": 0.6199551569506726,
+      "grad_norm": 0.08387491653054607,
+      "learning_rate": 0.0001931665949861782,
+      "loss": 0.2908,
+      "step": 1106
+    },
+    {
+      "epoch": 0.6205156950672646,
+      "grad_norm": 0.07852227844176,
+      "learning_rate": 0.00019314287518498033,
+      "loss": 0.2857,
+      "step": 1107
+    },
+    {
+      "epoch": 0.6210762331838565,
+      "grad_norm": 0.07612580999607896,
+      "learning_rate": 0.00019311911574900598,
+      "loss": 0.294,
+      "step": 1108
+    },
+    {
+      "epoch": 0.6216367713004485,
+      "grad_norm": 0.0765034734546166,
+      "learning_rate": 0.00019309531668836545,
+      "loss": 0.3059,
+      "step": 1109
+    },
+    {
+      "epoch": 0.6221973094170403,
+      "grad_norm": 0.07948666602882432,
+      "learning_rate": 0.00019307147801318585,
+      "loss": 0.3032,
+      "step": 1110
+    },
+    {
+      "epoch": 0.6227578475336323,
+      "grad_norm": 0.08120273308887432,
+      "learning_rate": 0.00019304759973361112,
+      "loss": 0.2921,
+      "step": 1111
+    },
+    {
+      "epoch": 0.6233183856502242,
+      "grad_norm": 0.07598800597650929,
+      "learning_rate": 0.00019302368185980217,
+      "loss": 0.2807,
+      "step": 1112
+    },
+    {
+      "epoch": 0.6238789237668162,
+      "grad_norm": 0.08107964441077349,
+      "learning_rate": 0.00019299972440193672,
+      "loss": 0.299,
+      "step": 1113
+    },
+    {
+      "epoch": 0.624439461883408,
+      "grad_norm": 0.08025669159681502,
+      "learning_rate": 0.00019297572737020922,
+      "loss": 0.3001,
+      "step": 1114
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.0808282612672915,
+      "learning_rate": 0.0001929516907748311,
+      "loss": 0.2929,
+      "step": 1115
+    },
+    {
+      "epoch": 0.625560538116592,
+      "grad_norm": 0.07907580039207018,
+      "learning_rate": 0.00019292761462603056,
+      "loss": 0.2986,
+      "step": 1116
+    },
+    {
+      "epoch": 0.6261210762331838,
+      "grad_norm": 0.0849296890291424,
+      "learning_rate": 0.00019290349893405268,
+      "loss": 0.3079,
+      "step": 1117
+    },
+    {
+      "epoch": 0.6266816143497758,
+      "grad_norm": 0.07739679164512617,
+      "learning_rate": 0.00019287934370915925,
+      "loss": 0.2929,
+      "step": 1118
+    },
+    {
+      "epoch": 0.6272421524663677,
+      "grad_norm": 0.07656728336817563,
+      "learning_rate": 0.00019285514896162905,
+      "loss": 0.3,
+      "step": 1119
+    },
+    {
+      "epoch": 0.6278026905829597,
+      "grad_norm": 0.07690746172050378,
+      "learning_rate": 0.00019283091470175754,
+      "loss": 0.2943,
+      "step": 1120
+    },
+    {
+      "epoch": 0.6283632286995515,
+      "grad_norm": 0.07997516497280334,
+      "learning_rate": 0.00019280664093985705,
+      "loss": 0.3095,
+      "step": 1121
+    },
+    {
+      "epoch": 0.6289237668161435,
+      "grad_norm": 0.07824202192119754,
+      "learning_rate": 0.0001927823276862567,
+      "loss": 0.306,
+      "step": 1122
+    },
+    {
+      "epoch": 0.6294843049327354,
+      "grad_norm": 0.07528993585215925,
+      "learning_rate": 0.00019275797495130247,
+      "loss": 0.2947,
+      "step": 1123
+    },
+    {
+      "epoch": 0.6300448430493274,
+      "grad_norm": 0.07902576493284015,
+      "learning_rate": 0.00019273358274535704,
+      "loss": 0.2933,
+      "step": 1124
+    },
+    {
+      "epoch": 0.6306053811659192,
+      "grad_norm": 0.08134667819638486,
+      "learning_rate": 0.0001927091510788,
+      "loss": 0.3097,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6311659192825112,
+      "grad_norm": 0.07938084005101839,
+      "learning_rate": 0.00019268467996202762,
+      "loss": 0.304,
+      "step": 1126
+    },
+    {
+      "epoch": 0.6317264573991032,
+      "grad_norm": 0.08270307342787032,
+      "learning_rate": 0.00019266016940545306,
+      "loss": 0.2941,
+      "step": 1127
+    },
+    {
+      "epoch": 0.6322869955156951,
+      "grad_norm": 0.08035414600387734,
+      "learning_rate": 0.00019263561941950622,
+      "loss": 0.2929,
+      "step": 1128
+    },
+    {
+      "epoch": 0.632847533632287,
+      "grad_norm": 0.08408923218561712,
+      "learning_rate": 0.0001926110300146337,
+      "loss": 0.3006,
+      "step": 1129
+    },
+    {
+      "epoch": 0.6334080717488789,
+      "grad_norm": 0.0781064497769012,
+      "learning_rate": 0.00019258640120129906,
+      "loss": 0.2943,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6339686098654709,
+      "grad_norm": 0.08038236922867095,
+      "learning_rate": 0.00019256173298998243,
+      "loss": 0.2885,
+      "step": 1131
+    },
+    {
+      "epoch": 0.6345291479820628,
+      "grad_norm": 0.08268602877577204,
+      "learning_rate": 0.0001925370253911808,
+      "loss": 0.288,
+      "step": 1132
+    },
+    {
+      "epoch": 0.6350896860986547,
+      "grad_norm": 0.07860721805885536,
+      "learning_rate": 0.00019251227841540796,
+      "loss": 0.3001,
+      "step": 1133
+    },
+    {
+      "epoch": 0.6356502242152466,
+      "grad_norm": 0.07708782842231296,
+      "learning_rate": 0.00019248749207319437,
+      "loss": 0.2845,
+      "step": 1134
+    },
+    {
+      "epoch": 0.6362107623318386,
+      "grad_norm": 0.08111114715442702,
+      "learning_rate": 0.00019246266637508726,
+      "loss": 0.2925,
+      "step": 1135
+    },
+    {
+      "epoch": 0.6367713004484304,
+      "grad_norm": 0.07924848980093907,
+      "learning_rate": 0.00019243780133165067,
+      "loss": 0.2973,
+      "step": 1136
+    },
+    {
+      "epoch": 0.6373318385650224,
+      "grad_norm": 0.08002464215813118,
+      "learning_rate": 0.00019241289695346532,
+      "loss": 0.2891,
+      "step": 1137
+    },
+    {
+      "epoch": 0.6378923766816144,
+      "grad_norm": 0.07998453079671493,
+      "learning_rate": 0.0001923879532511287,
+      "loss": 0.2924,
+      "step": 1138
+    },
+    {
+      "epoch": 0.6384529147982063,
+      "grad_norm": 0.0796813445505197,
+      "learning_rate": 0.00019236297023525497,
+      "loss": 0.2953,
+      "step": 1139
+    },
+    {
+      "epoch": 0.6390134529147982,
+      "grad_norm": 0.07958559913224303,
+      "learning_rate": 0.00019233794791647516,
+      "loss": 0.2946,
+      "step": 1140
+    },
+    {
+      "epoch": 0.6395739910313901,
+      "grad_norm": 0.07820666152885129,
+      "learning_rate": 0.00019231288630543685,
+      "loss": 0.2925,
+      "step": 1141
+    },
+    {
+      "epoch": 0.6401345291479821,
+      "grad_norm": 0.08113787442901467,
+      "learning_rate": 0.00019228778541280445,
+      "loss": 0.2936,
+      "step": 1142
+    },
+    {
+      "epoch": 0.640695067264574,
+      "grad_norm": 0.07979761874717892,
+      "learning_rate": 0.0001922626452492591,
+      "loss": 0.2929,
+      "step": 1143
+    },
+    {
+      "epoch": 0.6412556053811659,
+      "grad_norm": 0.07936109418562343,
+      "learning_rate": 0.00019223746582549853,
+      "loss": 0.2958,
+      "step": 1144
+    },
+    {
+      "epoch": 0.6418161434977578,
+      "grad_norm": 0.07638652262545426,
+      "learning_rate": 0.00019221224715223732,
+      "loss": 0.2862,
+      "step": 1145
+    },
+    {
+      "epoch": 0.6423766816143498,
+      "grad_norm": 0.08046334380154133,
+      "learning_rate": 0.00019218698924020668,
+      "loss": 0.2801,
+      "step": 1146
+    },
+    {
+      "epoch": 0.6429372197309418,
+      "grad_norm": 0.08288744800397352,
+      "learning_rate": 0.00019216169210015452,
+      "loss": 0.3026,
+      "step": 1147
+    },
+    {
+      "epoch": 0.6434977578475336,
+      "grad_norm": 0.08371225533628113,
+      "learning_rate": 0.0001921363557428454,
+      "loss": 0.3043,
+      "step": 1148
+    },
+    {
+      "epoch": 0.6440582959641256,
+      "grad_norm": 0.07998489281693971,
+      "learning_rate": 0.0001921109801790607,
+      "loss": 0.3037,
+      "step": 1149
+    },
+    {
+      "epoch": 0.6446188340807175,
+      "grad_norm": 0.0774369203642628,
+      "learning_rate": 0.00019208556541959834,
+      "loss": 0.2968,
+      "step": 1150
+    },
+    {
+      "epoch": 0.6451793721973094,
+      "grad_norm": 0.07994420039040068,
+      "learning_rate": 0.00019206011147527297,
+      "loss": 0.2973,
+      "step": 1151
+    },
+    {
+      "epoch": 0.6457399103139013,
+      "grad_norm": 0.07681632416455374,
+      "learning_rate": 0.00019203461835691594,
+      "loss": 0.2896,
+      "step": 1152
+    },
+    {
+      "epoch": 0.6463004484304933,
+      "grad_norm": 0.0802178755635615,
+      "learning_rate": 0.0001920090860753753,
+      "loss": 0.3013,
+      "step": 1153
+    },
+    {
+      "epoch": 0.6468609865470852,
+      "grad_norm": 0.08001251593046894,
+      "learning_rate": 0.00019198351464151564,
+      "loss": 0.3064,
+      "step": 1154
+    },
+    {
+      "epoch": 0.6474215246636771,
+      "grad_norm": 0.07999620091460043,
+      "learning_rate": 0.00019195790406621832,
+      "loss": 0.2889,
+      "step": 1155
+    },
+    {
+      "epoch": 0.647982062780269,
+      "grad_norm": 0.07437850211384636,
+      "learning_rate": 0.00019193225436038133,
+      "loss": 0.2904,
+      "step": 1156
+    },
+    {
+      "epoch": 0.648542600896861,
+      "grad_norm": 0.07719958033834659,
+      "learning_rate": 0.0001919065655349193,
+      "loss": 0.295,
+      "step": 1157
+    },
+    {
+      "epoch": 0.649103139013453,
+      "grad_norm": 0.07809909218661798,
+      "learning_rate": 0.0001918808376007635,
+      "loss": 0.2928,
+      "step": 1158
+    },
+    {
+      "epoch": 0.6496636771300448,
+      "grad_norm": 0.07869917717937468,
+      "learning_rate": 0.00019185507056886184,
+      "loss": 0.2918,
+      "step": 1159
+    },
+    {
+      "epoch": 0.6502242152466368,
+      "grad_norm": 0.0772141480892248,
+      "learning_rate": 0.00019182926445017893,
+      "loss": 0.2907,
+      "step": 1160
+    },
+    {
+      "epoch": 0.6507847533632287,
+      "grad_norm": 0.07727526271027191,
+      "learning_rate": 0.00019180341925569588,
+      "loss": 0.2862,
+      "step": 1161
+    },
+    {
+      "epoch": 0.6513452914798207,
+      "grad_norm": 0.07935252002939296,
+      "learning_rate": 0.0001917775349964106,
+      "loss": 0.2931,
+      "step": 1162
+    },
+    {
+      "epoch": 0.6519058295964125,
+      "grad_norm": 0.08399651091444187,
+      "learning_rate": 0.00019175161168333751,
+      "loss": 0.2846,
+      "step": 1163
+    },
+    {
+      "epoch": 0.6524663677130045,
+      "grad_norm": 0.08077552818708185,
+      "learning_rate": 0.0001917256493275076,
+      "loss": 0.297,
+      "step": 1164
+    },
+    {
+      "epoch": 0.6530269058295964,
+      "grad_norm": 0.07900938204339188,
+      "learning_rate": 0.0001916996479399686,
+      "loss": 0.3003,
+      "step": 1165
+    },
+    {
+      "epoch": 0.6535874439461884,
+      "grad_norm": 0.07531638452434798,
+      "learning_rate": 0.00019167360753178481,
+      "loss": 0.2813,
+      "step": 1166
+    },
+    {
+      "epoch": 0.6541479820627802,
+      "grad_norm": 0.08224864578997634,
+      "learning_rate": 0.00019164752811403707,
+      "loss": 0.3053,
+      "step": 1167
+    },
+    {
+      "epoch": 0.6547085201793722,
+      "grad_norm": 0.08063253736908517,
+      "learning_rate": 0.00019162140969782292,
+      "loss": 0.3042,
+      "step": 1168
+    },
+    {
+      "epoch": 0.6552690582959642,
+      "grad_norm": 0.07750934853255069,
+      "learning_rate": 0.00019159525229425642,
+      "loss": 0.2838,
+      "step": 1169
+    },
+    {
+      "epoch": 0.655829596412556,
+      "grad_norm": 0.07901906811626463,
+      "learning_rate": 0.0001915690559144682,
+      "loss": 0.3118,
+      "step": 1170
+    },
+    {
+      "epoch": 0.656390134529148,
+      "grad_norm": 0.07609691917887312,
+      "learning_rate": 0.00019154282056960557,
+      "loss": 0.2807,
+      "step": 1171
+    },
+    {
+      "epoch": 0.6569506726457399,
+      "grad_norm": 0.07638632839140339,
+      "learning_rate": 0.00019151654627083236,
+      "loss": 0.2903,
+      "step": 1172
+    },
+    {
+      "epoch": 0.6575112107623319,
+      "grad_norm": 0.07639299406238732,
+      "learning_rate": 0.000191490233029329,
+      "loss": 0.2836,
+      "step": 1173
+    },
+    {
+      "epoch": 0.6580717488789237,
+      "grad_norm": 0.08206956385097149,
+      "learning_rate": 0.00019146388085629245,
+      "loss": 0.297,
+      "step": 1174
+    },
+    {
+      "epoch": 0.6586322869955157,
+      "grad_norm": 0.07832471243632444,
+      "learning_rate": 0.00019143748976293624,
+      "loss": 0.2876,
+      "step": 1175
+    },
+    {
+      "epoch": 0.6591928251121076,
+      "grad_norm": 0.0796515723229912,
+      "learning_rate": 0.00019141105976049053,
+      "loss": 0.2932,
+      "step": 1176
+    },
+    {
+      "epoch": 0.6597533632286996,
+      "grad_norm": 0.080297126566589,
+      "learning_rate": 0.00019138459086020198,
+      "loss": 0.2994,
+      "step": 1177
+    },
+    {
+      "epoch": 0.6603139013452914,
+      "grad_norm": 0.08028040228837845,
+      "learning_rate": 0.0001913580830733338,
+      "loss": 0.2948,
+      "step": 1178
+    },
+    {
+      "epoch": 0.6608744394618834,
+      "grad_norm": 0.07586611021097159,
+      "learning_rate": 0.00019133153641116577,
+      "loss": 0.2926,
+      "step": 1179
+    },
+    {
+      "epoch": 0.6614349775784754,
+      "grad_norm": 0.07839649698548022,
+      "learning_rate": 0.0001913049508849942,
+      "loss": 0.2816,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6619955156950673,
+      "grad_norm": 0.0804896491931343,
+      "learning_rate": 0.00019127832650613189,
+      "loss": 0.3017,
+      "step": 1181
+    },
+    {
+      "epoch": 0.6625560538116592,
+      "grad_norm": 0.07642599468111257,
+      "learning_rate": 0.00019125166328590832,
+      "loss": 0.2928,
+      "step": 1182
+    },
+    {
+      "epoch": 0.6631165919282511,
+      "grad_norm": 0.07494985488009158,
+      "learning_rate": 0.00019122496123566933,
+      "loss": 0.2883,
+      "step": 1183
+    },
+    {
+      "epoch": 0.6636771300448431,
+      "grad_norm": 0.07897164442319689,
+      "learning_rate": 0.00019119822036677738,
+      "loss": 0.2932,
+      "step": 1184
+    },
+    {
+      "epoch": 0.6642376681614349,
+      "grad_norm": 0.07725523600064436,
+      "learning_rate": 0.0001911714406906114,
+      "loss": 0.2939,
+      "step": 1185
+    },
+    {
+      "epoch": 0.6647982062780269,
+      "grad_norm": 0.07857775076834542,
+      "learning_rate": 0.00019114462221856692,
+      "loss": 0.2966,
+      "step": 1186
+    },
+    {
+      "epoch": 0.6653587443946188,
+      "grad_norm": 0.07874979095357006,
+      "learning_rate": 0.0001911177649620558,
+      "loss": 0.2887,
+      "step": 1187
+    },
+    {
+      "epoch": 0.6659192825112108,
+      "grad_norm": 0.07562173420640422,
+      "learning_rate": 0.00019109086893250664,
+      "loss": 0.2713,
+      "step": 1188
+    },
+    {
+      "epoch": 0.6664798206278026,
+      "grad_norm": 0.0751433940022504,
+      "learning_rate": 0.00019106393414136438,
+      "loss": 0.2852,
+      "step": 1189
+    },
+    {
+      "epoch": 0.6670403587443946,
+      "grad_norm": 0.08060658317525636,
+      "learning_rate": 0.00019103696060009044,
+      "loss": 0.2955,
+      "step": 1190
+    },
+    {
+      "epoch": 0.6676008968609866,
+      "grad_norm": 0.08126966081941613,
+      "learning_rate": 0.00019100994832016283,
+      "loss": 0.2961,
+      "step": 1191
+    },
+    {
+      "epoch": 0.6681614349775785,
+      "grad_norm": 0.08060783778066744,
+      "learning_rate": 0.000190982897313076,
+      "loss": 0.2978,
+      "step": 1192
+    },
+    {
+      "epoch": 0.6687219730941704,
+      "grad_norm": 0.07930156320662468,
+      "learning_rate": 0.00019095580759034082,
+      "loss": 0.2962,
+      "step": 1193
+    },
+    {
+      "epoch": 0.6692825112107623,
+      "grad_norm": 0.07672510005379797,
+      "learning_rate": 0.00019092867916348477,
+      "loss": 0.2879,
+      "step": 1194
+    },
+    {
+      "epoch": 0.6698430493273543,
+      "grad_norm": 0.08058371505436204,
+      "learning_rate": 0.00019090151204405166,
+      "loss": 0.3008,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6704035874439462,
+      "grad_norm": 0.07833685452897783,
+      "learning_rate": 0.0001908743062436018,
+      "loss": 0.2969,
+      "step": 1196
+    },
+    {
+      "epoch": 0.6709641255605381,
+      "grad_norm": 0.07925289543162384,
+      "learning_rate": 0.00019084706177371208,
+      "loss": 0.3002,
+      "step": 1197
+    },
+    {
+      "epoch": 0.67152466367713,
+      "grad_norm": 0.07591225635300325,
+      "learning_rate": 0.00019081977864597564,
+      "loss": 0.2951,
+      "step": 1198
+    },
+    {
+      "epoch": 0.672085201793722,
+      "grad_norm": 0.07889895529910429,
+      "learning_rate": 0.00019079245687200227,
+      "loss": 0.2989,
+      "step": 1199
+    },
+    {
+      "epoch": 0.672645739910314,
+      "grad_norm": 0.07647347213481327,
+      "learning_rate": 0.00019076509646341806,
+      "loss": 0.2875,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6732062780269058,
+      "grad_norm": 0.07966217356892244,
+      "learning_rate": 0.0001907376974318656,
+      "loss": 0.2948,
+      "step": 1201
+    },
+    {
+      "epoch": 0.6737668161434978,
+      "grad_norm": 0.07782921796907775,
+      "learning_rate": 0.00019071025978900392,
+      "loss": 0.288,
+      "step": 1202
+    },
+    {
+      "epoch": 0.6743273542600897,
+      "grad_norm": 0.08166371192255209,
+      "learning_rate": 0.00019068278354650845,
+      "loss": 0.2845,
+      "step": 1203
+    },
+    {
+      "epoch": 0.6748878923766816,
+      "grad_norm": 0.08168203664130542,
+      "learning_rate": 0.00019065526871607112,
+      "loss": 0.296,
+      "step": 1204
+    },
+    {
+      "epoch": 0.6754484304932735,
+      "grad_norm": 0.07828404784030776,
+      "learning_rate": 0.00019062771530940013,
+      "loss": 0.2908,
+      "step": 1205
+    },
+    {
+      "epoch": 0.6760089686098655,
+      "grad_norm": 0.07878905522865796,
+      "learning_rate": 0.00019060012333822025,
+      "loss": 0.286,
+      "step": 1206
+    },
+    {
+      "epoch": 0.6765695067264574,
+      "grad_norm": 0.07729516979929453,
+      "learning_rate": 0.0001905724928142726,
+      "loss": 0.2958,
+      "step": 1207
+    },
+    {
+      "epoch": 0.6771300448430493,
+      "grad_norm": 0.08155782664594782,
+      "learning_rate": 0.00019054482374931467,
+      "loss": 0.305,
+      "step": 1208
+    },
+    {
+      "epoch": 0.6776905829596412,
+      "grad_norm": 0.07754570918200666,
+      "learning_rate": 0.00019051711615512043,
+      "loss": 0.2808,
+      "step": 1209
+    },
+    {
+      "epoch": 0.6782511210762332,
+      "grad_norm": 0.07994832399248969,
+      "learning_rate": 0.00019048937004348016,
+      "loss": 0.2873,
+      "step": 1210
+    },
+    {
+      "epoch": 0.6788116591928252,
+      "grad_norm": 0.08079659164992742,
+      "learning_rate": 0.0001904615854262006,
+      "loss": 0.3056,
+      "step": 1211
+    },
+    {
+      "epoch": 0.679372197309417,
+      "grad_norm": 0.07699454605752862,
+      "learning_rate": 0.00019043376231510484,
+      "loss": 0.309,
+      "step": 1212
+    },
+    {
+      "epoch": 0.679932735426009,
+      "grad_norm": 0.07958776554217052,
+      "learning_rate": 0.00019040590072203232,
+      "loss": 0.2927,
+      "step": 1213
+    },
+    {
+      "epoch": 0.6804932735426009,
+      "grad_norm": 0.0761266135368868,
+      "learning_rate": 0.00019037800065883895,
+      "loss": 0.3034,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6810538116591929,
+      "grad_norm": 0.07525512019293601,
+      "learning_rate": 0.0001903500621373969,
+      "loss": 0.2969,
+      "step": 1215
+    },
+    {
+      "epoch": 0.6816143497757847,
+      "grad_norm": 0.0850182475030861,
+      "learning_rate": 0.0001903220851695948,
+      "loss": 0.3009,
+      "step": 1216
+    },
+    {
+      "epoch": 0.6821748878923767,
+      "grad_norm": 0.0793591392970288,
+      "learning_rate": 0.00019029406976733756,
+      "loss": 0.2913,
+      "step": 1217
+    },
+    {
+      "epoch": 0.6827354260089686,
+      "grad_norm": 0.07563554698353232,
+      "learning_rate": 0.0001902660159425465,
+      "loss": 0.2886,
+      "step": 1218
+    },
+    {
+      "epoch": 0.6832959641255605,
+      "grad_norm": 0.08294638253174616,
+      "learning_rate": 0.00019023792370715924,
+      "loss": 0.292,
+      "step": 1219
+    },
+    {
+      "epoch": 0.6838565022421524,
+      "grad_norm": 0.0808696941836712,
+      "learning_rate": 0.0001902097930731298,
+      "loss": 0.3064,
+      "step": 1220
+    },
+    {
+      "epoch": 0.6844170403587444,
+      "grad_norm": 0.0800785386075129,
+      "learning_rate": 0.0001901816240524285,
+      "loss": 0.3081,
+      "step": 1221
+    },
+    {
+      "epoch": 0.6849775784753364,
+      "grad_norm": 0.08033149677757245,
+      "learning_rate": 0.00019015341665704205,
+      "loss": 0.3001,
+      "step": 1222
+    },
+    {
+      "epoch": 0.6855381165919282,
+      "grad_norm": 0.08082502481771073,
+      "learning_rate": 0.0001901251708989734,
+      "loss": 0.2999,
+      "step": 1223
+    },
+    {
+      "epoch": 0.6860986547085202,
+      "grad_norm": 0.0752550962403608,
+      "learning_rate": 0.0001900968867902419,
+      "loss": 0.2929,
+      "step": 1224
+    },
+    {
+      "epoch": 0.6866591928251121,
+      "grad_norm": 0.07769334337032176,
+      "learning_rate": 0.00019006856434288324,
+      "loss": 0.2912,
+      "step": 1225
+    },
+    {
+      "epoch": 0.6872197309417041,
+      "grad_norm": 0.07802959161878885,
+      "learning_rate": 0.00019004020356894926,
+      "loss": 0.2872,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6877802690582959,
+      "grad_norm": 0.07834574297659148,
+      "learning_rate": 0.00019001180448050827,
+      "loss": 0.287,
+      "step": 1227
+    },
+    {
+      "epoch": 0.6883408071748879,
+      "grad_norm": 0.08160535681558054,
+      "learning_rate": 0.00018998336708964488,
+      "loss": 0.3056,
+      "step": 1228
+    },
+    {
+      "epoch": 0.6889013452914798,
+      "grad_norm": 0.07690740685562249,
+      "learning_rate": 0.00018995489140845995,
+      "loss": 0.2938,
+      "step": 1229
+    },
+    {
+      "epoch": 0.6894618834080718,
+      "grad_norm": 0.07817464809682581,
+      "learning_rate": 0.00018992637744907063,
+      "loss": 0.2829,
+      "step": 1230
+    },
+    {
+      "epoch": 0.6900224215246636,
+      "grad_norm": 0.07889620909426741,
+      "learning_rate": 0.00018989782522361033,
+      "loss": 0.2973,
+      "step": 1231
+    },
+    {
+      "epoch": 0.6905829596412556,
+      "grad_norm": 0.07909642834814787,
+      "learning_rate": 0.00018986923474422884,
+      "loss": 0.284,
+      "step": 1232
+    },
+    {
+      "epoch": 0.6911434977578476,
+      "grad_norm": 0.07876951617721326,
+      "learning_rate": 0.00018984060602309213,
+      "loss": 0.2904,
+      "step": 1233
+    },
+    {
+      "epoch": 0.6917040358744395,
+      "grad_norm": 0.07842618999673406,
+      "learning_rate": 0.00018981193907238253,
+      "loss": 0.299,
+      "step": 1234
+    },
+    {
+      "epoch": 0.6922645739910314,
+      "grad_norm": 0.07561517151383235,
+      "learning_rate": 0.00018978323390429855,
+      "loss": 0.2884,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6928251121076233,
+      "grad_norm": 0.07578068063055152,
+      "learning_rate": 0.00018975449053105505,
+      "loss": 0.2844,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6933856502242153,
+      "grad_norm": 0.07790798516880275,
+      "learning_rate": 0.00018972570896488305,
+      "loss": 0.2982,
+      "step": 1237
+    },
+    {
+      "epoch": 0.6939461883408071,
+      "grad_norm": 0.07914549155279324,
+      "learning_rate": 0.00018969688921802988,
+      "loss": 0.2878,
+      "step": 1238
+    },
+    {
+      "epoch": 0.6945067264573991,
+      "grad_norm": 0.07404575322266842,
+      "learning_rate": 0.00018966803130275915,
+      "loss": 0.2868,
+      "step": 1239
+    },
+    {
+      "epoch": 0.695067264573991,
+      "grad_norm": 0.08120303825599713,
+      "learning_rate": 0.0001896391352313506,
+      "loss": 0.2945,
+      "step": 1240
+    },
+    {
+      "epoch": 0.695627802690583,
+      "grad_norm": 0.07437289579864437,
+      "learning_rate": 0.00018961020101610038,
+      "loss": 0.2905,
+      "step": 1241
+    },
+    {
+      "epoch": 0.6961883408071748,
+      "grad_norm": 0.07927624379164572,
+      "learning_rate": 0.00018958122866932067,
+      "loss": 0.2873,
+      "step": 1242
+    },
+    {
+      "epoch": 0.6967488789237668,
+      "grad_norm": 0.07683562036131639,
+      "learning_rate": 0.00018955221820334008,
+      "loss": 0.2995,
+      "step": 1243
+    },
+    {
+      "epoch": 0.6973094170403588,
+      "grad_norm": 0.07562879219671298,
+      "learning_rate": 0.00018952316963050328,
+      "loss": 0.2967,
+      "step": 1244
+    },
+    {
+      "epoch": 0.6978699551569507,
+      "grad_norm": 0.07988497736399475,
+      "learning_rate": 0.00018949408296317115,
+      "loss": 0.2907,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6984304932735426,
+      "grad_norm": 0.0787077251985745,
+      "learning_rate": 0.00018946495821372094,
+      "loss": 0.3093,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6989910313901345,
+      "grad_norm": 0.07735407110297393,
+      "learning_rate": 0.000189435795394546,
+      "loss": 0.2997,
+      "step": 1247
+    },
+    {
+      "epoch": 0.6995515695067265,
+      "grad_norm": 0.07741020308703161,
+      "learning_rate": 0.0001894065945180558,
+      "loss": 0.2926,
+      "step": 1248
+    },
+    {
+      "epoch": 0.7001121076233184,
+      "grad_norm": 0.07976914290152838,
+      "learning_rate": 0.00018937735559667617,
+      "loss": 0.293,
+      "step": 1249
+    },
+    {
+      "epoch": 0.7006726457399103,
+      "grad_norm": 0.07510775283599667,
+      "learning_rate": 0.00018934807864284903,
+      "loss": 0.2943,
+      "step": 1250
+    },
+    {
+      "epoch": 0.7012331838565022,
+      "grad_norm": 0.07593201972228869,
+      "learning_rate": 0.00018931876366903253,
+      "loss": 0.2917,
+      "step": 1251
+    },
+    {
+      "epoch": 0.7017937219730942,
+      "grad_norm": 0.07451377135318003,
+      "learning_rate": 0.00018928941068770093,
+      "loss": 0.2939,
+      "step": 1252
+    },
+    {
+      "epoch": 0.702354260089686,
+      "grad_norm": 0.07813147177851087,
+      "learning_rate": 0.0001892600197113447,
+      "loss": 0.2937,
+      "step": 1253
+    },
+    {
+      "epoch": 0.702914798206278,
+      "grad_norm": 0.0770543925802459,
+      "learning_rate": 0.00018923059075247054,
+      "loss": 0.2852,
+      "step": 1254
+    },
+    {
+      "epoch": 0.70347533632287,
+      "grad_norm": 0.0786103794148151,
+      "learning_rate": 0.0001892011238236012,
+      "loss": 0.2872,
+      "step": 1255
+    },
+    {
+      "epoch": 0.7040358744394619,
+      "grad_norm": 0.07774864788902053,
+      "learning_rate": 0.0001891716189372757,
+      "loss": 0.2985,
+      "step": 1256
+    },
+    {
+      "epoch": 0.7045964125560538,
+      "grad_norm": 0.0769095326661384,
+      "learning_rate": 0.0001891420761060491,
+      "loss": 0.2797,
+      "step": 1257
+    },
+    {
+      "epoch": 0.7051569506726457,
+      "grad_norm": 0.07782531883080296,
+      "learning_rate": 0.00018911249534249268,
+      "loss": 0.293,
+      "step": 1258
+    },
+    {
+      "epoch": 0.7057174887892377,
+      "grad_norm": 0.0800402316054872,
+      "learning_rate": 0.00018908287665919384,
+      "loss": 0.2991,
+      "step": 1259
+    },
+    {
+      "epoch": 0.7062780269058296,
+      "grad_norm": 0.0775037455296807,
+      "learning_rate": 0.00018905322006875617,
+      "loss": 0.2993,
+      "step": 1260
+    },
+    {
+      "epoch": 0.7068385650224215,
+      "grad_norm": 0.07450553709538528,
+      "learning_rate": 0.00018902352558379924,
+      "loss": 0.2798,
+      "step": 1261
+    },
+    {
+      "epoch": 0.7073991031390134,
+      "grad_norm": 0.07518676284466798,
+      "learning_rate": 0.00018899379321695895,
+      "loss": 0.2907,
+      "step": 1262
+    },
+    {
+      "epoch": 0.7079596412556054,
+      "grad_norm": 0.07842392987449064,
+      "learning_rate": 0.00018896402298088715,
+      "loss": 0.274,
+      "step": 1263
+    },
+    {
+      "epoch": 0.7085201793721974,
+      "grad_norm": 0.08030036395322773,
+      "learning_rate": 0.0001889342148882519,
+      "loss": 0.2914,
+      "step": 1264
+    },
+    {
+      "epoch": 0.7090807174887892,
+      "grad_norm": 0.07846212324557195,
+      "learning_rate": 0.00018890436895173732,
+      "loss": 0.2991,
+      "step": 1265
+    },
+    {
+      "epoch": 0.7096412556053812,
+      "grad_norm": 0.08157597085045058,
+      "learning_rate": 0.00018887448518404364,
+      "loss": 0.2864,
+      "step": 1266
+    },
+    {
+      "epoch": 0.7102017937219731,
+      "grad_norm": 0.079973832983049,
+      "learning_rate": 0.00018884456359788724,
+      "loss": 0.2945,
+      "step": 1267
+    },
+    {
+      "epoch": 0.7107623318385651,
+      "grad_norm": 0.08100813683795925,
+      "learning_rate": 0.0001888146042060005,
+      "loss": 0.3062,
+      "step": 1268
+    },
+    {
+      "epoch": 0.7113228699551569,
+      "grad_norm": 0.07850234211695056,
+      "learning_rate": 0.000188784607021132,
+      "loss": 0.2998,
+      "step": 1269
+    },
+    {
+      "epoch": 0.7118834080717489,
+      "grad_norm": 0.07651407446888785,
+      "learning_rate": 0.00018875457205604632,
+      "loss": 0.2757,
+      "step": 1270
+    },
+    {
+      "epoch": 0.7124439461883408,
+      "grad_norm": 0.07609078214940616,
+      "learning_rate": 0.0001887244993235241,
+      "loss": 0.2946,
+      "step": 1271
+    },
+    {
+      "epoch": 0.7130044843049327,
+      "grad_norm": 0.07746617973765456,
+      "learning_rate": 0.00018869438883636214,
+      "loss": 0.2887,
+      "step": 1272
+    },
+    {
+      "epoch": 0.7135650224215246,
+      "grad_norm": 0.0789499722820708,
+      "learning_rate": 0.0001886642406073732,
+      "loss": 0.3051,
+      "step": 1273
+    },
+    {
+      "epoch": 0.7141255605381166,
+      "grad_norm": 0.07521197453858279,
+      "learning_rate": 0.00018863405464938622,
+      "loss": 0.2913,
+      "step": 1274
+    },
+    {
+      "epoch": 0.7146860986547086,
+      "grad_norm": 0.07898055919018118,
+      "learning_rate": 0.00018860383097524608,
+      "loss": 0.3067,
+      "step": 1275
+    },
+    {
+      "epoch": 0.7152466367713004,
+      "grad_norm": 0.0773309100267759,
+      "learning_rate": 0.00018857356959781378,
+      "loss": 0.2968,
+      "step": 1276
+    },
+    {
+      "epoch": 0.7158071748878924,
+      "grad_norm": 0.07631060611281504,
+      "learning_rate": 0.00018854327052996635,
+      "loss": 0.3048,
+      "step": 1277
+    },
+    {
+      "epoch": 0.7163677130044843,
+      "grad_norm": 0.07770995667275178,
+      "learning_rate": 0.00018851293378459685,
+      "loss": 0.3072,
+      "step": 1278
+    },
+    {
+      "epoch": 0.7169282511210763,
+      "grad_norm": 0.07669477056566686,
+      "learning_rate": 0.00018848255937461435,
+      "loss": 0.2976,
+      "step": 1279
+    },
+    {
+      "epoch": 0.7174887892376681,
+      "grad_norm": 0.07592043732806922,
+      "learning_rate": 0.000188452147312944,
+      "loss": 0.2923,
+      "step": 1280
+    },
+    {
+      "epoch": 0.7180493273542601,
+      "grad_norm": 0.07864646912832403,
+      "learning_rate": 0.0001884216976125269,
+      "loss": 0.2875,
+      "step": 1281
+    },
+    {
+      "epoch": 0.718609865470852,
+      "grad_norm": 0.08284975313147833,
+      "learning_rate": 0.00018839121028632025,
+      "loss": 0.2828,
+      "step": 1282
+    },
+    {
+      "epoch": 0.719170403587444,
+      "grad_norm": 0.07969891678461695,
+      "learning_rate": 0.00018836068534729722,
+      "loss": 0.2845,
+      "step": 1283
+    },
+    {
+      "epoch": 0.7197309417040358,
+      "grad_norm": 0.07680522024641628,
+      "learning_rate": 0.00018833012280844699,
+      "loss": 0.2883,
+      "step": 1284
+    },
+    {
+      "epoch": 0.7202914798206278,
+      "grad_norm": 0.0765633573161656,
+      "learning_rate": 0.0001882995226827747,
+      "loss": 0.303,
+      "step": 1285
+    },
+    {
+      "epoch": 0.7208520179372198,
+      "grad_norm": 0.07838781066391033,
+      "learning_rate": 0.00018826888498330155,
+      "loss": 0.2804,
+      "step": 1286
+    },
+    {
+      "epoch": 0.7214125560538116,
+      "grad_norm": 0.07719297221136888,
+      "learning_rate": 0.00018823820972306468,
+      "loss": 0.3005,
+      "step": 1287
+    },
+    {
+      "epoch": 0.7219730941704036,
+      "grad_norm": 0.07800903389182938,
+      "learning_rate": 0.0001882074969151172,
+      "loss": 0.2864,
+      "step": 1288
+    },
+    {
+      "epoch": 0.7225336322869955,
+      "grad_norm": 0.07996779675024149,
+      "learning_rate": 0.00018817674657252832,
+      "loss": 0.2973,
+      "step": 1289
+    },
+    {
+      "epoch": 0.7230941704035875,
+      "grad_norm": 0.07733184379911967,
+      "learning_rate": 0.00018814595870838305,
+      "loss": 0.2982,
+      "step": 1290
+    },
+    {
+      "epoch": 0.7236547085201793,
+      "grad_norm": 0.07731490662415895,
+      "learning_rate": 0.0001881151333357825,
+      "loss": 0.29,
+      "step": 1291
+    },
+    {
+      "epoch": 0.7242152466367713,
+      "grad_norm": 0.0782262630148421,
+      "learning_rate": 0.00018808427046784366,
+      "loss": 0.3026,
+      "step": 1292
+    },
+    {
+      "epoch": 0.7247757847533632,
+      "grad_norm": 0.07651090647832731,
+      "learning_rate": 0.00018805337011769947,
+      "loss": 0.2906,
+      "step": 1293
+    },
+    {
+      "epoch": 0.7253363228699552,
+      "grad_norm": 0.07872373873076764,
+      "learning_rate": 0.00018802243229849893,
+      "loss": 0.3079,
+      "step": 1294
+    },
+    {
+      "epoch": 0.725896860986547,
+      "grad_norm": 0.07884950443304208,
+      "learning_rate": 0.0001879914570234068,
+      "loss": 0.2991,
+      "step": 1295
+    },
+    {
+      "epoch": 0.726457399103139,
+      "grad_norm": 0.07497108857838872,
+      "learning_rate": 0.000187960444305604,
+      "loss": 0.2894,
+      "step": 1296
+    },
+    {
+      "epoch": 0.727017937219731,
+      "grad_norm": 0.07587875237581815,
+      "learning_rate": 0.0001879293941582872,
+      "loss": 0.2899,
+      "step": 1297
+    },
+    {
+      "epoch": 0.7275784753363229,
+      "grad_norm": 0.07369102143037222,
+      "learning_rate": 0.00018789830659466912,
+      "loss": 0.2827,
+      "step": 1298
+    },
+    {
+      "epoch": 0.7281390134529148,
+      "grad_norm": 0.07789720105922017,
+      "learning_rate": 0.00018786718162797826,
+      "loss": 0.2756,
+      "step": 1299
+    },
+    {
+      "epoch": 0.7286995515695067,
+      "grad_norm": 0.0760606785136823,
+      "learning_rate": 0.0001878360192714592,
+      "loss": 0.2964,
+      "step": 1300
+    },
+    {
+      "epoch": 0.7292600896860987,
+      "grad_norm": 0.07437994591108035,
+      "learning_rate": 0.00018780481953837233,
+      "loss": 0.2876,
+      "step": 1301
+    },
+    {
+      "epoch": 0.7298206278026906,
+      "grad_norm": 0.07608950743250191,
+      "learning_rate": 0.00018777358244199393,
+      "loss": 0.2796,
+      "step": 1302
+    },
+    {
+      "epoch": 0.7303811659192825,
+      "grad_norm": 0.07805452353954782,
+      "learning_rate": 0.0001877423079956163,
+      "loss": 0.2701,
+      "step": 1303
+    },
+    {
+      "epoch": 0.7309417040358744,
+      "grad_norm": 0.07573651354847788,
+      "learning_rate": 0.00018771099621254746,
+      "loss": 0.2833,
+      "step": 1304
+    },
+    {
+      "epoch": 0.7315022421524664,
+      "grad_norm": 0.07806005779452138,
+      "learning_rate": 0.00018767964710611148,
+      "loss": 0.2855,
+      "step": 1305
+    },
+    {
+      "epoch": 0.7320627802690582,
+      "grad_norm": 0.07467295002820801,
+      "learning_rate": 0.0001876482606896482,
+      "loss": 0.2926,
+      "step": 1306
+    },
+    {
+      "epoch": 0.7326233183856502,
+      "grad_norm": 0.07553584065478107,
+      "learning_rate": 0.0001876168369765134,
+      "loss": 0.2943,
+      "step": 1307
+    },
+    {
+      "epoch": 0.7331838565022422,
+      "grad_norm": 0.07591966668596306,
+      "learning_rate": 0.00018758537598007868,
+      "loss": 0.2967,
+      "step": 1308
+    },
+    {
+      "epoch": 0.7337443946188341,
+      "grad_norm": 0.07711605811857049,
+      "learning_rate": 0.00018755387771373155,
+      "loss": 0.2785,
+      "step": 1309
+    },
+    {
+      "epoch": 0.734304932735426,
+      "grad_norm": 0.07804229988629692,
+      "learning_rate": 0.00018752234219087538,
+      "loss": 0.3016,
+      "step": 1310
+    },
+    {
+      "epoch": 0.7348654708520179,
+      "grad_norm": 0.07724641664074823,
+      "learning_rate": 0.00018749076942492935,
+      "loss": 0.2996,
+      "step": 1311
+    },
+    {
+      "epoch": 0.7354260089686099,
+      "grad_norm": 0.07674865115715372,
+      "learning_rate": 0.0001874591594293285,
+      "loss": 0.3016,
+      "step": 1312
+    },
+    {
+      "epoch": 0.7359865470852018,
+      "grad_norm": 0.07493025188227236,
+      "learning_rate": 0.00018742751221752376,
+      "loss": 0.3033,
+      "step": 1313
+    },
+    {
+      "epoch": 0.7365470852017937,
+      "grad_norm": 0.07935121801870736,
+      "learning_rate": 0.00018739582780298187,
+      "loss": 0.2942,
+      "step": 1314
+    },
+    {
+      "epoch": 0.7371076233183856,
+      "grad_norm": 0.07792308909891997,
+      "learning_rate": 0.00018736410619918535,
+      "loss": 0.2847,
+      "step": 1315
+    },
+    {
+      "epoch": 0.7376681614349776,
+      "grad_norm": 0.07543353574385893,
+      "learning_rate": 0.0001873323474196326,
+      "loss": 0.285,
+      "step": 1316
+    },
+    {
+      "epoch": 0.7382286995515696,
+      "grad_norm": 0.07826412511500047,
+      "learning_rate": 0.00018730055147783787,
+      "loss": 0.2843,
+      "step": 1317
+    },
+    {
+      "epoch": 0.7387892376681614,
+      "grad_norm": 0.07543515537108975,
+      "learning_rate": 0.00018726871838733113,
+      "loss": 0.2952,
+      "step": 1318
+    },
+    {
+      "epoch": 0.7393497757847534,
+      "grad_norm": 0.07872077356370108,
+      "learning_rate": 0.0001872368481616582,
+      "loss": 0.292,
+      "step": 1319
+    },
+    {
+      "epoch": 0.7399103139013453,
+      "grad_norm": 0.07649789977559601,
+      "learning_rate": 0.00018720494081438078,
+      "loss": 0.3022,
+      "step": 1320
+    },
+    {
+      "epoch": 0.7404708520179372,
+      "grad_norm": 0.07589209125514779,
+      "learning_rate": 0.0001871729963590762,
+      "loss": 0.2882,
+      "step": 1321
+    },
+    {
+      "epoch": 0.7410313901345291,
+      "grad_norm": 0.07559430157312046,
+      "learning_rate": 0.0001871410148093378,
+      "loss": 0.2962,
+      "step": 1322
+    },
+    {
+      "epoch": 0.7415919282511211,
+      "grad_norm": 0.07351979593308759,
+      "learning_rate": 0.00018710899617877446,
+      "loss": 0.2772,
+      "step": 1323
+    },
+    {
+      "epoch": 0.742152466367713,
+      "grad_norm": 0.07629143680804043,
+      "learning_rate": 0.00018707694048101104,
+      "loss": 0.2916,
+      "step": 1324
+    },
+    {
+      "epoch": 0.7427130044843049,
+      "grad_norm": 0.0737835599349803,
+      "learning_rate": 0.00018704484772968808,
+      "loss": 0.2859,
+      "step": 1325
+    },
+    {
+      "epoch": 0.7432735426008968,
+      "grad_norm": 0.07927090998639666,
+      "learning_rate": 0.00018701271793846185,
+      "loss": 0.2983,
+      "step": 1326
+    },
+    {
+      "epoch": 0.7438340807174888,
+      "grad_norm": 0.07550225413500154,
+      "learning_rate": 0.0001869805511210045,
+      "loss": 0.2852,
+      "step": 1327
+    },
+    {
+      "epoch": 0.7443946188340808,
+      "grad_norm": 0.07460152949029368,
+      "learning_rate": 0.00018694834729100386,
+      "loss": 0.2906,
+      "step": 1328
+    },
+    {
+      "epoch": 0.7449551569506726,
+      "grad_norm": 0.07676090452029702,
+      "learning_rate": 0.00018691610646216344,
+      "loss": 0.2834,
+      "step": 1329
+    },
+    {
+      "epoch": 0.7455156950672646,
+      "grad_norm": 0.07803325358661241,
+      "learning_rate": 0.00018688382864820267,
+      "loss": 0.2902,
+      "step": 1330
+    },
+    {
+      "epoch": 0.7460762331838565,
+      "grad_norm": 0.0769711200908372,
+      "learning_rate": 0.0001868515138628566,
+      "loss": 0.3062,
+      "step": 1331
+    },
+    {
+      "epoch": 0.7466367713004485,
+      "grad_norm": 0.07433534825617263,
+      "learning_rate": 0.00018681916211987597,
+      "loss": 0.2976,
+      "step": 1332
+    },
+    {
+      "epoch": 0.7471973094170403,
+      "grad_norm": 0.0776685315458737,
+      "learning_rate": 0.00018678677343302738,
+      "loss": 0.2978,
+      "step": 1333
+    },
+    {
+      "epoch": 0.7477578475336323,
+      "grad_norm": 0.07545609129455182,
+      "learning_rate": 0.00018675434781609303,
+      "loss": 0.2898,
+      "step": 1334
+    },
+    {
+      "epoch": 0.7483183856502242,
+      "grad_norm": 0.07572405945464439,
+      "learning_rate": 0.00018672188528287093,
+      "loss": 0.2802,
+      "step": 1335
+    },
+    {
+      "epoch": 0.7488789237668162,
+      "grad_norm": 0.07719842275732951,
+      "learning_rate": 0.00018668938584717471,
+      "loss": 0.2949,
+      "step": 1336
+    },
+    {
+      "epoch": 0.749439461883408,
+      "grad_norm": 0.07690169045731857,
+      "learning_rate": 0.0001866568495228338,
+      "loss": 0.2982,
+      "step": 1337
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07467124607579745,
+      "learning_rate": 0.0001866242763236932,
+      "loss": 0.2906,
+      "step": 1338
+    },
+    {
+      "epoch": 0.750560538116592,
+      "grad_norm": 0.0811234171484504,
+      "learning_rate": 0.00018659166626361375,
+      "loss": 0.2949,
+      "step": 1339
+    },
+    {
+      "epoch": 0.7511210762331838,
+      "grad_norm": 0.07717253760610217,
+      "learning_rate": 0.00018655901935647187,
+      "loss": 0.3011,
+      "step": 1340
+    },
+    {
+      "epoch": 0.7516816143497758,
+      "grad_norm": 0.07491111542612311,
+      "learning_rate": 0.00018652633561615963,
+      "loss": 0.2933,
+      "step": 1341
+    },
+    {
+      "epoch": 0.7522421524663677,
+      "grad_norm": 0.07812280967678403,
+      "learning_rate": 0.00018649361505658493,
+      "loss": 0.2802,
+      "step": 1342
+    },
+    {
+      "epoch": 0.7528026905829597,
+      "grad_norm": 0.0739088336544016,
+      "learning_rate": 0.00018646085769167119,
+      "loss": 0.2775,
+      "step": 1343
+    },
+    {
+      "epoch": 0.7533632286995515,
+      "grad_norm": 0.07744656519534383,
+      "learning_rate": 0.00018642806353535754,
+      "loss": 0.2902,
+      "step": 1344
+    },
+    {
+      "epoch": 0.7539237668161435,
+      "grad_norm": 0.07741913855187414,
+      "learning_rate": 0.0001863952326015988,
+      "loss": 0.2921,
+      "step": 1345
+    },
+    {
+      "epoch": 0.7544843049327354,
+      "grad_norm": 0.07621173786924237,
+      "learning_rate": 0.00018636236490436535,
+      "loss": 0.2938,
+      "step": 1346
+    },
+    {
+      "epoch": 0.7550448430493274,
+      "grad_norm": 0.07952963227808484,
+      "learning_rate": 0.00018632946045764334,
+      "loss": 0.2954,
+      "step": 1347
+    },
+    {
+      "epoch": 0.7556053811659192,
+      "grad_norm": 0.07906435746949275,
+      "learning_rate": 0.00018629651927543447,
+      "loss": 0.2832,
+      "step": 1348
+    },
+    {
+      "epoch": 0.7561659192825112,
+      "grad_norm": 0.07540400579101485,
+      "learning_rate": 0.00018626354137175603,
+      "loss": 0.2875,
+      "step": 1349
+    },
+    {
+      "epoch": 0.7567264573991032,
+      "grad_norm": 0.07616969867605854,
+      "learning_rate": 0.00018623052676064106,
+      "loss": 0.2839,
+      "step": 1350
+    },
+    {
+      "epoch": 0.7572869955156951,
+      "grad_norm": 0.08035148075312004,
+      "learning_rate": 0.00018619747545613814,
+      "loss": 0.2909,
+      "step": 1351
+    },
+    {
+      "epoch": 0.757847533632287,
+      "grad_norm": 0.07969136405752344,
+      "learning_rate": 0.00018616438747231148,
+      "loss": 0.2902,
+      "step": 1352
+    },
+    {
+      "epoch": 0.7584080717488789,
+      "grad_norm": 0.07753625952755534,
+      "learning_rate": 0.00018613126282324092,
+      "loss": 0.2853,
+      "step": 1353
+    },
+    {
+      "epoch": 0.7589686098654709,
+      "grad_norm": 0.07604892980998836,
+      "learning_rate": 0.00018609810152302183,
+      "loss": 0.2993,
+      "step": 1354
+    },
+    {
+      "epoch": 0.7595291479820628,
+      "grad_norm": 0.07896616442677123,
+      "learning_rate": 0.0001860649035857653,
+      "loss": 0.3027,
+      "step": 1355
+    },
+    {
+      "epoch": 0.7600896860986547,
+      "grad_norm": 0.07940682063157535,
+      "learning_rate": 0.00018603166902559783,
+      "loss": 0.2941,
+      "step": 1356
+    },
+    {
+      "epoch": 0.7606502242152466,
+      "grad_norm": 0.07279709841837333,
+      "learning_rate": 0.00018599839785666172,
+      "loss": 0.2784,
+      "step": 1357
+    },
+    {
+      "epoch": 0.7612107623318386,
+      "grad_norm": 0.07287156515708634,
+      "learning_rate": 0.00018596509009311473,
+      "loss": 0.2955,
+      "step": 1358
+    },
+    {
+      "epoch": 0.7617713004484304,
+      "grad_norm": 0.07910667735159985,
+      "learning_rate": 0.00018593174574913014,
+      "loss": 0.2918,
+      "step": 1359
+    },
+    {
+      "epoch": 0.7623318385650224,
+      "grad_norm": 0.07535123141943245,
+      "learning_rate": 0.00018589836483889687,
+      "loss": 0.3043,
+      "step": 1360
+    },
+    {
+      "epoch": 0.7628923766816144,
+      "grad_norm": 0.07079144742106978,
+      "learning_rate": 0.00018586494737661942,
+      "loss": 0.2906,
+      "step": 1361
+    },
+    {
+      "epoch": 0.7634529147982063,
+      "grad_norm": 0.07646124561271408,
+      "learning_rate": 0.0001858314933765178,
+      "loss": 0.2753,
+      "step": 1362
+    },
+    {
+      "epoch": 0.7640134529147982,
+      "grad_norm": 0.08000962596500587,
+      "learning_rate": 0.00018579800285282758,
+      "loss": 0.2888,
+      "step": 1363
+    },
+    {
+      "epoch": 0.7645739910313901,
+      "grad_norm": 0.07737151093016979,
+      "learning_rate": 0.00018576447581979984,
+      "loss": 0.2866,
+      "step": 1364
+    },
+    {
+      "epoch": 0.7651345291479821,
+      "grad_norm": 0.07922437567858846,
+      "learning_rate": 0.00018573091229170125,
+      "loss": 0.2859,
+      "step": 1365
+    },
+    {
+      "epoch": 0.765695067264574,
+      "grad_norm": 0.07781993765796033,
+      "learning_rate": 0.00018569731228281402,
+      "loss": 0.293,
+      "step": 1366
+    },
+    {
+      "epoch": 0.7662556053811659,
+      "grad_norm": 0.0786279322732407,
+      "learning_rate": 0.00018566367580743578,
+      "loss": 0.2894,
+      "step": 1367
+    },
+    {
+      "epoch": 0.7668161434977578,
+      "grad_norm": 0.07748284406177973,
+      "learning_rate": 0.0001856300028798798,
+      "loss": 0.2869,
+      "step": 1368
+    },
+    {
+      "epoch": 0.7673766816143498,
+      "grad_norm": 0.07679842315728289,
+      "learning_rate": 0.00018559629351447477,
+      "loss": 0.2971,
+      "step": 1369
+    },
+    {
+      "epoch": 0.7679372197309418,
+      "grad_norm": 0.0736803367460062,
+      "learning_rate": 0.00018556254772556497,
+      "loss": 0.2841,
+      "step": 1370
+    },
+    {
+      "epoch": 0.7684977578475336,
+      "grad_norm": 0.07769094809947598,
+      "learning_rate": 0.0001855287655275101,
+      "loss": 0.2985,
+      "step": 1371
+    },
+    {
+      "epoch": 0.7690582959641256,
+      "grad_norm": 0.08724476283650384,
+      "learning_rate": 0.0001854949469346854,
+      "loss": 0.2887,
+      "step": 1372
+    },
+    {
+      "epoch": 0.7696188340807175,
+      "grad_norm": 0.073526400118211,
+      "learning_rate": 0.00018546109196148158,
+      "loss": 0.2815,
+      "step": 1373
+    },
+    {
+      "epoch": 0.7701793721973094,
+      "grad_norm": 0.07062244148529655,
+      "learning_rate": 0.00018542720062230482,
+      "loss": 0.2967,
+      "step": 1374
+    },
+    {
+      "epoch": 0.7707399103139013,
+      "grad_norm": 0.07671357170768117,
+      "learning_rate": 0.0001853932729315768,
+      "loss": 0.2898,
+      "step": 1375
+    },
+    {
+      "epoch": 0.7713004484304933,
+      "grad_norm": 0.07347117491199325,
+      "learning_rate": 0.00018535930890373466,
+      "loss": 0.2876,
+      "step": 1376
+    },
+    {
+      "epoch": 0.7718609865470852,
+      "grad_norm": 0.07571055973685425,
+      "learning_rate": 0.00018532530855323097,
+      "loss": 0.2925,
+      "step": 1377
+    },
+    {
+      "epoch": 0.7724215246636771,
+      "grad_norm": 0.07569006514744296,
+      "learning_rate": 0.00018529127189453382,
+      "loss": 0.2875,
+      "step": 1378
+    },
+    {
+      "epoch": 0.772982062780269,
+      "grad_norm": 0.07964297912638683,
+      "learning_rate": 0.00018525719894212675,
+      "loss": 0.2993,
+      "step": 1379
+    },
+    {
+      "epoch": 0.773542600896861,
+      "grad_norm": 0.0783041774718746,
+      "learning_rate": 0.00018522308971050865,
+      "loss": 0.2989,
+      "step": 1380
+    },
+    {
+      "epoch": 0.774103139013453,
+      "grad_norm": 0.0758786416845156,
+      "learning_rate": 0.0001851889442141939,
+      "loss": 0.2894,
+      "step": 1381
+    },
+    {
+      "epoch": 0.7746636771300448,
+      "grad_norm": 0.07541063650038844,
+      "learning_rate": 0.00018515476246771232,
+      "loss": 0.3034,
+      "step": 1382
+    },
+    {
+      "epoch": 0.7752242152466368,
+      "grad_norm": 0.07650126337312455,
+      "learning_rate": 0.0001851205444856092,
+      "loss": 0.2902,
+      "step": 1383
+    },
+    {
+      "epoch": 0.7757847533632287,
+      "grad_norm": 0.07588693687731386,
+      "learning_rate": 0.00018508629028244519,
+      "loss": 0.2958,
+      "step": 1384
+    },
+    {
+      "epoch": 0.7763452914798207,
+      "grad_norm": 0.07726162355363445,
+      "learning_rate": 0.00018505199987279634,
+      "loss": 0.2828,
+      "step": 1385
+    },
+    {
+      "epoch": 0.7769058295964125,
+      "grad_norm": 0.07688664333995268,
+      "learning_rate": 0.00018501767327125417,
+      "loss": 0.2927,
+      "step": 1386
+    },
+    {
+      "epoch": 0.7774663677130045,
+      "grad_norm": 0.07463864822174102,
+      "learning_rate": 0.00018498331049242553,
+      "loss": 0.2824,
+      "step": 1387
+    },
+    {
+      "epoch": 0.7780269058295964,
+      "grad_norm": 0.07270587460945237,
+      "learning_rate": 0.00018494891155093274,
+      "loss": 0.285,
+      "step": 1388
+    },
+    {
+      "epoch": 0.7785874439461884,
+      "grad_norm": 0.07472118172733451,
+      "learning_rate": 0.00018491447646141337,
+      "loss": 0.2838,
+      "step": 1389
+    },
+    {
+      "epoch": 0.7791479820627802,
+      "grad_norm": 0.07906313104393052,
+      "learning_rate": 0.0001848800052385206,
+      "loss": 0.2993,
+      "step": 1390
+    },
+    {
+      "epoch": 0.7797085201793722,
+      "grad_norm": 0.08000296545660737,
+      "learning_rate": 0.00018484549789692277,
+      "loss": 0.3023,
+      "step": 1391
+    },
+    {
+      "epoch": 0.7802690582959642,
+      "grad_norm": 0.08070960466410868,
+      "learning_rate": 0.0001848109544513037,
+      "loss": 0.2888,
+      "step": 1392
+    },
+    {
+      "epoch": 0.780829596412556,
+      "grad_norm": 0.07772837035425759,
+      "learning_rate": 0.00018477637491636254,
+      "loss": 0.2924,
+      "step": 1393
+    },
+    {
+      "epoch": 0.781390134529148,
+      "grad_norm": 0.07328464817430952,
+      "learning_rate": 0.0001847417593068138,
+      "loss": 0.2844,
+      "step": 1394
+    },
+    {
+      "epoch": 0.7819506726457399,
+      "grad_norm": 0.07588594193230476,
+      "learning_rate": 0.00018470710763738736,
+      "loss": 0.3031,
+      "step": 1395
+    },
+    {
+      "epoch": 0.7825112107623319,
+      "grad_norm": 0.07594699223222112,
+      "learning_rate": 0.00018467241992282843,
+      "loss": 0.2872,
+      "step": 1396
+    },
+    {
+      "epoch": 0.7830717488789237,
+      "grad_norm": 0.0718626645875063,
+      "learning_rate": 0.00018463769617789756,
+      "loss": 0.2935,
+      "step": 1397
+    },
+    {
+      "epoch": 0.7836322869955157,
+      "grad_norm": 0.0739120716985456,
+      "learning_rate": 0.00018460293641737056,
+      "loss": 0.2837,
+      "step": 1398
+    },
+    {
+      "epoch": 0.7841928251121076,
+      "grad_norm": 0.07631399704508741,
+      "learning_rate": 0.00018456814065603874,
+      "loss": 0.2838,
+      "step": 1399
+    },
+    {
+      "epoch": 0.7847533632286996,
+      "grad_norm": 0.07527887722291483,
+      "learning_rate": 0.00018453330890870855,
+      "loss": 0.2941,
+      "step": 1400
+    },
+    {
+      "epoch": 0.7853139013452914,
+      "grad_norm": 0.07687638348874318,
+      "learning_rate": 0.00018449844119020187,
+      "loss": 0.2958,
+      "step": 1401
+    },
+    {
+      "epoch": 0.7858744394618834,
+      "grad_norm": 0.0751731950384892,
+      "learning_rate": 0.00018446353751535585,
+      "loss": 0.2891,
+      "step": 1402
+    },
+    {
+      "epoch": 0.7864349775784754,
+      "grad_norm": 0.0771625419362304,
+      "learning_rate": 0.00018442859789902287,
+      "loss": 0.2844,
+      "step": 1403
+    },
+    {
+      "epoch": 0.7869955156950673,
+      "grad_norm": 0.07948109827598755,
+      "learning_rate": 0.0001843936223560707,
+      "loss": 0.2908,
+      "step": 1404
+    },
+    {
+      "epoch": 0.7875560538116592,
+      "grad_norm": 0.07731481396768788,
+      "learning_rate": 0.0001843586109013824,
+      "loss": 0.2877,
+      "step": 1405
+    },
+    {
+      "epoch": 0.7881165919282511,
+      "grad_norm": 0.07740649519667313,
+      "learning_rate": 0.00018432356354985624,
+      "loss": 0.2901,
+      "step": 1406
+    },
+    {
+      "epoch": 0.7886771300448431,
+      "grad_norm": 0.07299887622297382,
+      "learning_rate": 0.00018428848031640582,
+      "loss": 0.2857,
+      "step": 1407
+    },
+    {
+      "epoch": 0.7892376681614349,
+      "grad_norm": 0.07281704715766726,
+      "learning_rate": 0.00018425336121596,
+      "loss": 0.2789,
+      "step": 1408
+    },
+    {
+      "epoch": 0.7897982062780269,
+      "grad_norm": 0.07442852265460735,
+      "learning_rate": 0.00018421820626346287,
+      "loss": 0.2862,
+      "step": 1409
+    },
+    {
+      "epoch": 0.7903587443946188,
+      "grad_norm": 0.07253215159296206,
+      "learning_rate": 0.0001841830154738738,
+      "loss": 0.3008,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7909192825112108,
+      "grad_norm": 0.07303673507071831,
+      "learning_rate": 0.00018414778886216744,
+      "loss": 0.2745,
+      "step": 1411
+    },
+    {
+      "epoch": 0.7914798206278026,
+      "grad_norm": 0.07633390907693668,
+      "learning_rate": 0.00018411252644333362,
+      "loss": 0.2841,
+      "step": 1412
+    },
+    {
+      "epoch": 0.7920403587443946,
+      "grad_norm": 0.07473600112958628,
+      "learning_rate": 0.0001840772282323774,
+      "loss": 0.2941,
+      "step": 1413
+    },
+    {
+      "epoch": 0.7926008968609866,
+      "grad_norm": 0.07371548412534,
+      "learning_rate": 0.00018404189424431921,
+      "loss": 0.2832,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7931614349775785,
+      "grad_norm": 0.07692758307826633,
+      "learning_rate": 0.00018400652449419456,
+      "loss": 0.2868,
+      "step": 1415
+    },
+    {
+      "epoch": 0.7937219730941704,
+      "grad_norm": 0.07644275048553147,
+      "learning_rate": 0.00018397111899705419,
+      "loss": 0.2889,
+      "step": 1416
+    },
+    {
+      "epoch": 0.7942825112107623,
+      "grad_norm": 0.07900690964527601,
+      "learning_rate": 0.00018393567776796413,
+      "loss": 0.3043,
+      "step": 1417
+    },
+    {
+      "epoch": 0.7948430493273543,
+      "grad_norm": 0.07358862281444084,
+      "learning_rate": 0.0001839002008220055,
+      "loss": 0.2901,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7954035874439462,
+      "grad_norm": 0.07391016621207044,
+      "learning_rate": 0.0001838646881742748,
+      "loss": 0.2867,
+      "step": 1419
+    },
+    {
+      "epoch": 0.7959641255605381,
+      "grad_norm": 0.07941176651737829,
+      "learning_rate": 0.00018382913983988348,
+      "loss": 0.2867,
+      "step": 1420
+    },
+    {
+      "epoch": 0.79652466367713,
+      "grad_norm": 0.07572282829815266,
+      "learning_rate": 0.00018379355583395842,
+      "loss": 0.2934,
+      "step": 1421
+    },
+    {
+      "epoch": 0.797085201793722,
+      "grad_norm": 0.074960350680707,
+      "learning_rate": 0.00018375793617164145,
+      "loss": 0.309,
+      "step": 1422
+    },
+    {
+      "epoch": 0.797645739910314,
+      "grad_norm": 0.07575566300387161,
+      "learning_rate": 0.00018372228086808979,
+      "loss": 0.2862,
+      "step": 1423
+    },
+    {
+      "epoch": 0.7982062780269058,
+      "grad_norm": 0.07308428000549903,
+      "learning_rate": 0.00018368658993847566,
+      "loss": 0.2909,
+      "step": 1424
+    },
+    {
+      "epoch": 0.7987668161434978,
+      "grad_norm": 0.07549301760709565,
+      "learning_rate": 0.0001836508633979865,
+      "loss": 0.2864,
+      "step": 1425
+    },
+    {
+      "epoch": 0.7993273542600897,
+      "grad_norm": 0.0731274376673481,
+      "learning_rate": 0.00018361510126182492,
+      "loss": 0.2789,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7998878923766816,
+      "grad_norm": 0.07738565135023746,
+      "learning_rate": 0.0001835793035452087,
+      "loss": 0.3048,
+      "step": 1427
+    },
+    {
+      "epoch": 0.8004484304932735,
+      "grad_norm": 0.07177820742743628,
+      "learning_rate": 0.00018354347026337066,
+      "loss": 0.2907,
+      "step": 1428
+    },
+    {
+      "epoch": 0.8010089686098655,
+      "grad_norm": 0.07640689818899002,
+      "learning_rate": 0.00018350760143155884,
+      "loss": 0.2863,
+      "step": 1429
+    },
+    {
+      "epoch": 0.8015695067264574,
+      "grad_norm": 0.07672084445856638,
+      "learning_rate": 0.0001834716970650364,
+      "loss": 0.2965,
+      "step": 1430
+    },
+    {
+      "epoch": 0.8021300448430493,
+      "grad_norm": 0.07627208815081751,
+      "learning_rate": 0.00018343575717908158,
+      "loss": 0.296,
+      "step": 1431
+    },
+    {
+      "epoch": 0.8026905829596412,
+      "grad_norm": 0.07380746132396154,
+      "learning_rate": 0.0001833997817889878,
+      "loss": 0.2719,
+      "step": 1432
+    },
+    {
+      "epoch": 0.8032511210762332,
+      "grad_norm": 0.07973900555459396,
+      "learning_rate": 0.00018336377091006351,
+      "loss": 0.2961,
+      "step": 1433
+    },
+    {
+      "epoch": 0.8038116591928252,
+      "grad_norm": 0.07299240702694566,
+      "learning_rate": 0.00018332772455763232,
+      "loss": 0.2886,
+      "step": 1434
+    },
+    {
+      "epoch": 0.804372197309417,
+      "grad_norm": 0.07489555184832908,
+      "learning_rate": 0.00018329164274703287,
+      "loss": 0.2934,
+      "step": 1435
+    },
+    {
+      "epoch": 0.804932735426009,
+      "grad_norm": 0.0768283435332101,
+      "learning_rate": 0.000183255525493619,
+      "loss": 0.2975,
+      "step": 1436
+    },
+    {
+      "epoch": 0.8054932735426009,
+      "grad_norm": 0.07530896474211601,
+      "learning_rate": 0.00018321937281275951,
+      "loss": 0.2772,
+      "step": 1437
+    },
+    {
+      "epoch": 0.8060538116591929,
+      "grad_norm": 0.07343763669424976,
+      "learning_rate": 0.00018318318471983837,
+      "loss": 0.2905,
+      "step": 1438
+    },
+    {
+      "epoch": 0.8066143497757847,
+      "grad_norm": 0.0759174954877864,
+      "learning_rate": 0.00018314696123025454,
+      "loss": 0.2756,
+      "step": 1439
+    },
+    {
+      "epoch": 0.8071748878923767,
+      "grad_norm": 0.07954362881356519,
+      "learning_rate": 0.0001831107023594221,
+      "loss": 0.2906,
+      "step": 1440
+    },
+    {
+      "epoch": 0.8077354260089686,
+      "grad_norm": 0.07722907000273242,
+      "learning_rate": 0.00018307440812277017,
+      "loss": 0.2927,
+      "step": 1441
+    },
+    {
+      "epoch": 0.8082959641255605,
+      "grad_norm": 0.0761978990934437,
+      "learning_rate": 0.0001830380785357429,
+      "loss": 0.2854,
+      "step": 1442
+    },
+    {
+      "epoch": 0.8088565022421524,
+      "grad_norm": 0.07192499399813279,
+      "learning_rate": 0.00018300171361379953,
+      "loss": 0.2751,
+      "step": 1443
+    },
+    {
+      "epoch": 0.8094170403587444,
+      "grad_norm": 0.07672166540018124,
+      "learning_rate": 0.00018296531337241425,
+      "loss": 0.2867,
+      "step": 1444
+    },
+    {
+      "epoch": 0.8099775784753364,
+      "grad_norm": 0.07432234888523945,
+      "learning_rate": 0.0001829288778270764,
+      "loss": 0.28,
+      "step": 1445
+    },
+    {
+      "epoch": 0.8105381165919282,
+      "grad_norm": 0.07552370929739224,
+      "learning_rate": 0.0001828924069932902,
+      "loss": 0.3054,
+      "step": 1446
+    },
+    {
+      "epoch": 0.8110986547085202,
+      "grad_norm": 0.07243138995072261,
+      "learning_rate": 0.00018285590088657503,
+      "loss": 0.2743,
+      "step": 1447
+    },
+    {
+      "epoch": 0.8116591928251121,
+      "grad_norm": 0.0744121758908257,
+      "learning_rate": 0.00018281935952246518,
+      "loss": 0.2936,
+      "step": 1448
+    },
+    {
+      "epoch": 0.8122197309417041,
+      "grad_norm": 0.07618031580109315,
+      "learning_rate": 0.00018278278291650998,
+      "loss": 0.2869,
+      "step": 1449
+    },
+    {
+      "epoch": 0.8127802690582959,
+      "grad_norm": 0.07364536149850888,
+      "learning_rate": 0.00018274617108427374,
+      "loss": 0.2872,
+      "step": 1450
+    },
+    {
+      "epoch": 0.8133408071748879,
+      "grad_norm": 0.0731242088928229,
+      "learning_rate": 0.0001827095240413358,
+      "loss": 0.2908,
+      "step": 1451
+    },
+    {
+      "epoch": 0.8139013452914798,
+      "grad_norm": 0.07910289495768733,
+      "learning_rate": 0.0001826728418032904,
+      "loss": 0.2999,
+      "step": 1452
+    },
+    {
+      "epoch": 0.8144618834080718,
+      "grad_norm": 0.07491422734436892,
+      "learning_rate": 0.0001826361243857469,
+      "loss": 0.2918,
+      "step": 1453
+    },
+    {
+      "epoch": 0.8150224215246636,
+      "grad_norm": 0.07069684199857261,
+      "learning_rate": 0.00018259937180432943,
+      "loss": 0.2825,
+      "step": 1454
+    },
+    {
+      "epoch": 0.8155829596412556,
+      "grad_norm": 0.07461081918013844,
+      "learning_rate": 0.0001825625840746773,
+      "loss": 0.2996,
+      "step": 1455
+    },
+    {
+      "epoch": 0.8161434977578476,
+      "grad_norm": 0.07456817318255526,
+      "learning_rate": 0.00018252576121244456,
+      "loss": 0.291,
+      "step": 1456
+    },
+    {
+      "epoch": 0.8167040358744395,
+      "grad_norm": 0.0752009880568552,
+      "learning_rate": 0.00018248890323330037,
+      "loss": 0.2879,
+      "step": 1457
+    },
+    {
+      "epoch": 0.8172645739910314,
+      "grad_norm": 0.07976424882963153,
+      "learning_rate": 0.00018245201015292884,
+      "loss": 0.291,
+      "step": 1458
+    },
+    {
+      "epoch": 0.8178251121076233,
+      "grad_norm": 0.07391074644209736,
+      "learning_rate": 0.00018241508198702888,
+      "loss": 0.2795,
+      "step": 1459
+    },
+    {
+      "epoch": 0.8183856502242153,
+      "grad_norm": 0.0744088522851671,
+      "learning_rate": 0.00018237811875131444,
+      "loss": 0.2982,
+      "step": 1460
+    },
+    {
+      "epoch": 0.8189461883408071,
+      "grad_norm": 0.07261544628193192,
+      "learning_rate": 0.00018234112046151436,
+      "loss": 0.28,
+      "step": 1461
+    },
+    {
+      "epoch": 0.8195067264573991,
+      "grad_norm": 0.07734099041728819,
+      "learning_rate": 0.00018230408713337242,
+      "loss": 0.295,
+      "step": 1462
+    },
+    {
+      "epoch": 0.820067264573991,
+      "grad_norm": 0.07363549616953434,
+      "learning_rate": 0.00018226701878264724,
+      "loss": 0.2817,
+      "step": 1463
+    },
+    {
+      "epoch": 0.820627802690583,
+      "grad_norm": 0.07466610010605963,
+      "learning_rate": 0.00018222991542511246,
+      "loss": 0.2985,
+      "step": 1464
+    },
+    {
+      "epoch": 0.8211883408071748,
+      "grad_norm": 0.07301472031476657,
+      "learning_rate": 0.00018219277707655644,
+      "loss": 0.281,
+      "step": 1465
+    },
+    {
+      "epoch": 0.8217488789237668,
+      "grad_norm": 0.07369725422981575,
+      "learning_rate": 0.00018215560375278264,
+      "loss": 0.2927,
+      "step": 1466
+    },
+    {
+      "epoch": 0.8223094170403588,
+      "grad_norm": 0.0720833906344199,
+      "learning_rate": 0.00018211839546960928,
+      "loss": 0.2913,
+      "step": 1467
+    },
+    {
+      "epoch": 0.8228699551569507,
+      "grad_norm": 0.07282264057414481,
+      "learning_rate": 0.00018208115224286947,
+      "loss": 0.2887,
+      "step": 1468
+    },
+    {
+      "epoch": 0.8234304932735426,
+      "grad_norm": 0.07209233132516016,
+      "learning_rate": 0.0001820438740884111,
+      "loss": 0.2936,
+      "step": 1469
+    },
+    {
+      "epoch": 0.8239910313901345,
+      "grad_norm": 0.07431361686281528,
+      "learning_rate": 0.00018200656102209718,
+      "loss": 0.2798,
+      "step": 1470
+    },
+    {
+      "epoch": 0.8245515695067265,
+      "grad_norm": 0.0729069874816868,
+      "learning_rate": 0.00018196921305980532,
+      "loss": 0.2888,
+      "step": 1471
+    },
+    {
+      "epoch": 0.8251121076233184,
+      "grad_norm": 0.07514945154645199,
+      "learning_rate": 0.0001819318302174281,
+      "loss": 0.2898,
+      "step": 1472
+    },
+    {
+      "epoch": 0.8256726457399103,
+      "grad_norm": 0.07526812354818167,
+      "learning_rate": 0.00018189441251087292,
+      "loss": 0.2896,
+      "step": 1473
+    },
+    {
+      "epoch": 0.8262331838565022,
+      "grad_norm": 0.07367208114650646,
+      "learning_rate": 0.00018185695995606195,
+      "loss": 0.2906,
+      "step": 1474
+    },
+    {
+      "epoch": 0.8267937219730942,
+      "grad_norm": 0.07347381591983008,
+      "learning_rate": 0.00018181947256893234,
+      "loss": 0.2877,
+      "step": 1475
+    },
+    {
+      "epoch": 0.827354260089686,
+      "grad_norm": 0.07537713414133736,
+      "learning_rate": 0.00018178195036543592,
+      "loss": 0.2842,
+      "step": 1476
+    },
+    {
+      "epoch": 0.827914798206278,
+      "grad_norm": 0.075994983600787,
+      "learning_rate": 0.00018174439336153943,
+      "loss": 0.2867,
+      "step": 1477
+    },
+    {
+      "epoch": 0.82847533632287,
+      "grad_norm": 0.07483465441397896,
+      "learning_rate": 0.00018170680157322434,
+      "loss": 0.2858,
+      "step": 1478
+    },
+    {
+      "epoch": 0.8290358744394619,
+      "grad_norm": 0.07805373208430955,
+      "learning_rate": 0.00018166917501648695,
+      "loss": 0.3151,
+      "step": 1479
+    },
+    {
+      "epoch": 0.8295964125560538,
+      "grad_norm": 0.07482220546270187,
+      "learning_rate": 0.00018163151370733838,
+      "loss": 0.2926,
+      "step": 1480
+    },
+    {
+      "epoch": 0.8301569506726457,
+      "grad_norm": 0.07368764143513033,
+      "learning_rate": 0.00018159381766180452,
+      "loss": 0.292,
+      "step": 1481
+    },
+    {
+      "epoch": 0.8307174887892377,
+      "grad_norm": 0.08143484239387924,
+      "learning_rate": 0.00018155608689592604,
+      "loss": 0.2885,
+      "step": 1482
+    },
+    {
+      "epoch": 0.8312780269058296,
+      "grad_norm": 0.07849511557171265,
+      "learning_rate": 0.00018151832142575838,
+      "loss": 0.2884,
+      "step": 1483
+    },
+    {
+      "epoch": 0.8318385650224215,
+      "grad_norm": 0.0739668970466124,
+      "learning_rate": 0.00018148052126737177,
+      "loss": 0.29,
+      "step": 1484
+    },
+    {
+      "epoch": 0.8323991031390134,
+      "grad_norm": 0.0741235222041159,
+      "learning_rate": 0.00018144268643685118,
+      "loss": 0.2897,
+      "step": 1485
+    },
+    {
+      "epoch": 0.8329596412556054,
+      "grad_norm": 0.0751369067793912,
+      "learning_rate": 0.00018140481695029634,
+      "loss": 0.2708,
+      "step": 1486
+    },
+    {
+      "epoch": 0.8335201793721974,
+      "grad_norm": 0.07468639268265492,
+      "learning_rate": 0.0001813669128238217,
+      "loss": 0.286,
+      "step": 1487
+    },
+    {
+      "epoch": 0.8340807174887892,
+      "grad_norm": 0.07405511155465547,
+      "learning_rate": 0.00018132897407355657,
+      "loss": 0.275,
+      "step": 1488
+    },
+    {
+      "epoch": 0.8346412556053812,
+      "grad_norm": 0.07183006998574253,
+      "learning_rate": 0.00018129100071564476,
+      "loss": 0.2774,
+      "step": 1489
+    },
+    {
+      "epoch": 0.8352017937219731,
+      "grad_norm": 0.07455414281594977,
+      "learning_rate": 0.00018125299276624504,
+      "loss": 0.2878,
+      "step": 1490
+    },
+    {
+      "epoch": 0.8357623318385651,
+      "grad_norm": 0.07342341772225446,
+      "learning_rate": 0.0001812149502415308,
+      "loss": 0.2942,
+      "step": 1491
+    },
+    {
+      "epoch": 0.8363228699551569,
+      "grad_norm": 0.07247308389338286,
+      "learning_rate": 0.00018117687315769007,
+      "loss": 0.2736,
+      "step": 1492
+    },
+    {
+      "epoch": 0.8368834080717489,
+      "grad_norm": 0.07331788336274414,
+      "learning_rate": 0.00018113876153092576,
+      "loss": 0.2924,
+      "step": 1493
+    },
+    {
+      "epoch": 0.8374439461883408,
+      "grad_norm": 0.07636980653536268,
+      "learning_rate": 0.00018110061537745536,
+      "loss": 0.285,
+      "step": 1494
+    },
+    {
+      "epoch": 0.8380044843049327,
+      "grad_norm": 0.07379623135697953,
+      "learning_rate": 0.00018106243471351105,
+      "loss": 0.2902,
+      "step": 1495
+    },
+    {
+      "epoch": 0.8385650224215246,
+      "grad_norm": 0.07550751475407677,
+      "learning_rate": 0.00018102421955533974,
+      "loss": 0.298,
+      "step": 1496
+    },
+    {
+      "epoch": 0.8391255605381166,
+      "grad_norm": 0.07360096778024856,
+      "learning_rate": 0.00018098596991920297,
+      "loss": 0.2967,
+      "step": 1497
+    },
+    {
+      "epoch": 0.8396860986547086,
+      "grad_norm": 0.07242540875769851,
+      "learning_rate": 0.000180947685821377,
+      "loss": 0.2817,
+      "step": 1498
+    },
+    {
+      "epoch": 0.8402466367713004,
+      "grad_norm": 0.07314544504553466,
+      "learning_rate": 0.00018090936727815278,
+      "loss": 0.2801,
+      "step": 1499
+    },
+    {
+      "epoch": 0.8408071748878924,
+      "grad_norm": 0.07591653427559013,
+      "learning_rate": 0.00018087101430583577,
+      "loss": 0.296,
+      "step": 1500
+    },
+    {
+      "epoch": 0.8413677130044843,
+      "grad_norm": 0.07416486332512015,
+      "learning_rate": 0.00018083262692074627,
+      "loss": 0.2748,
+      "step": 1501
+    },
+    {
+      "epoch": 0.8419282511210763,
+      "grad_norm": 0.07317031892892478,
+      "learning_rate": 0.00018079420513921913,
+      "loss": 0.2779,
+      "step": 1502
+    },
+    {
+      "epoch": 0.8424887892376681,
+      "grad_norm": 0.07485099910739548,
+      "learning_rate": 0.00018075574897760376,
+      "loss": 0.2821,
+      "step": 1503
+    },
+    {
+      "epoch": 0.8430493273542601,
+      "grad_norm": 0.0733565558947233,
+      "learning_rate": 0.00018071725845226436,
+      "loss": 0.2899,
+      "step": 1504
+    },
+    {
+      "epoch": 0.843609865470852,
+      "grad_norm": 0.07634656510574725,
+      "learning_rate": 0.00018067873357957968,
+      "loss": 0.2963,
+      "step": 1505
+    },
+    {
+      "epoch": 0.844170403587444,
+      "grad_norm": 0.0716798945396636,
+      "learning_rate": 0.00018064017437594303,
+      "loss": 0.2887,
+      "step": 1506
+    },
+    {
+      "epoch": 0.8447309417040358,
+      "grad_norm": 0.07494626920072107,
+      "learning_rate": 0.0001806015808577624,
+      "loss": 0.3005,
+      "step": 1507
+    },
+    {
+      "epoch": 0.8452914798206278,
+      "grad_norm": 0.07371906529487111,
+      "learning_rate": 0.0001805629530414604,
+      "loss": 0.2945,
+      "step": 1508
+    },
+    {
+      "epoch": 0.8458520179372198,
+      "grad_norm": 0.07331479262974581,
+      "learning_rate": 0.00018052429094347411,
+      "loss": 0.2838,
+      "step": 1509
+    },
+    {
+      "epoch": 0.8464125560538116,
+      "grad_norm": 0.07540102259039154,
+      "learning_rate": 0.00018048559458025537,
+      "loss": 0.287,
+      "step": 1510
+    },
+    {
+      "epoch": 0.8469730941704036,
+      "grad_norm": 0.07457926552589283,
+      "learning_rate": 0.00018044686396827047,
+      "loss": 0.2755,
+      "step": 1511
+    },
+    {
+      "epoch": 0.8475336322869955,
+      "grad_norm": 0.07497185379846803,
+      "learning_rate": 0.0001804080991240003,
+      "loss": 0.2867,
+      "step": 1512
+    },
+    {
+      "epoch": 0.8480941704035875,
+      "grad_norm": 0.07785367889006829,
+      "learning_rate": 0.00018036930006394038,
+      "loss": 0.2963,
+      "step": 1513
+    },
+    {
+      "epoch": 0.8486547085201793,
+      "grad_norm": 0.07251561565241364,
+      "learning_rate": 0.00018033046680460073,
+      "loss": 0.2871,
+      "step": 1514
+    },
+    {
+      "epoch": 0.8492152466367713,
+      "grad_norm": 0.07316362504655254,
+      "learning_rate": 0.00018029159936250593,
+      "loss": 0.2903,
+      "step": 1515
+    },
+    {
+      "epoch": 0.8497757847533632,
+      "grad_norm": 0.07273848774249944,
+      "learning_rate": 0.00018025269775419507,
+      "loss": 0.2784,
+      "step": 1516
+    },
+    {
+      "epoch": 0.8503363228699552,
+      "grad_norm": 0.07357435732694743,
+      "learning_rate": 0.0001802137619962219,
+      "loss": 0.293,
+      "step": 1517
+    },
+    {
+      "epoch": 0.850896860986547,
+      "grad_norm": 0.07301027893663752,
+      "learning_rate": 0.00018017479210515462,
+      "loss": 0.2813,
+      "step": 1518
+    },
+    {
+      "epoch": 0.851457399103139,
+      "grad_norm": 0.07417040740255613,
+      "learning_rate": 0.0001801357880975759,
+      "loss": 0.2815,
+      "step": 1519
+    },
+    {
+      "epoch": 0.852017937219731,
+      "grad_norm": 0.07201234029862826,
+      "learning_rate": 0.000180096749990083,
+      "loss": 0.2837,
+      "step": 1520
+    },
+    {
+      "epoch": 0.8525784753363229,
+      "grad_norm": 0.07654746052709169,
+      "learning_rate": 0.00018005767779928768,
+      "loss": 0.3035,
+      "step": 1521
+    },
+    {
+      "epoch": 0.8531390134529148,
+      "grad_norm": 0.07559009198121788,
+      "learning_rate": 0.00018001857154181626,
+      "loss": 0.2953,
+      "step": 1522
+    },
+    {
+      "epoch": 0.8536995515695067,
+      "grad_norm": 0.0742104436963455,
+      "learning_rate": 0.00017997943123430936,
+      "loss": 0.2733,
+      "step": 1523
+    },
+    {
+      "epoch": 0.8542600896860987,
+      "grad_norm": 0.07313887641047588,
+      "learning_rate": 0.00017994025689342235,
+      "loss": 0.283,
+      "step": 1524
+    },
+    {
+      "epoch": 0.8548206278026906,
+      "grad_norm": 0.07282347483864758,
+      "learning_rate": 0.00017990104853582493,
+      "loss": 0.2806,
+      "step": 1525
+    },
+    {
+      "epoch": 0.8553811659192825,
+      "grad_norm": 0.07364438692256288,
+      "learning_rate": 0.00017986180617820123,
+      "loss": 0.2899,
+      "step": 1526
+    },
+    {
+      "epoch": 0.8559417040358744,
+      "grad_norm": 0.07649758437982163,
+      "learning_rate": 0.00017982252983725,
+      "loss": 0.2906,
+      "step": 1527
+    },
+    {
+      "epoch": 0.8565022421524664,
+      "grad_norm": 0.07423743401057138,
+      "learning_rate": 0.00017978321952968434,
+      "loss": 0.2868,
+      "step": 1528
+    },
+    {
+      "epoch": 0.8570627802690582,
+      "grad_norm": 0.0758700713210416,
+      "learning_rate": 0.00017974387527223184,
+      "loss": 0.2981,
+      "step": 1529
+    },
+    {
+      "epoch": 0.8576233183856502,
+      "grad_norm": 0.07786873063275755,
+      "learning_rate": 0.00017970449708163452,
+      "loss": 0.2919,
+      "step": 1530
+    },
+    {
+      "epoch": 0.8581838565022422,
+      "grad_norm": 0.07565227507775568,
+      "learning_rate": 0.0001796650849746489,
+      "loss": 0.2949,
+      "step": 1531
+    },
+    {
+      "epoch": 0.8587443946188341,
+      "grad_norm": 0.0762723363189425,
+      "learning_rate": 0.00017962563896804578,
+      "loss": 0.2958,
+      "step": 1532
+    },
+    {
+      "epoch": 0.859304932735426,
+      "grad_norm": 0.07144478659817405,
+      "learning_rate": 0.00017958615907861055,
+      "loss": 0.2893,
+      "step": 1533
+    },
+    {
+      "epoch": 0.8598654708520179,
+      "grad_norm": 0.07507825955322966,
+      "learning_rate": 0.00017954664532314295,
+      "loss": 0.2847,
+      "step": 1534
+    },
+    {
+      "epoch": 0.8604260089686099,
+      "grad_norm": 0.07424681293346497,
+      "learning_rate": 0.0001795070977184572,
+      "loss": 0.2944,
+      "step": 1535
+    },
+    {
+      "epoch": 0.8609865470852018,
+      "grad_norm": 0.07497557650932773,
+      "learning_rate": 0.00017946751628138174,
+      "loss": 0.2698,
+      "step": 1536
+    },
+    {
+      "epoch": 0.8615470852017937,
+      "grad_norm": 0.07066603839213563,
+      "learning_rate": 0.0001794279010287596,
+      "loss": 0.2856,
+      "step": 1537
+    },
+    {
+      "epoch": 0.8621076233183856,
+      "grad_norm": 0.07503927606305691,
+      "learning_rate": 0.00017938825197744807,
+      "loss": 0.2764,
+      "step": 1538
+    },
+    {
+      "epoch": 0.8626681614349776,
+      "grad_norm": 0.07871173230887407,
+      "learning_rate": 0.00017934856914431899,
+      "loss": 0.2844,
+      "step": 1539
+    },
+    {
+      "epoch": 0.8632286995515696,
+      "grad_norm": 0.07386649820562965,
+      "learning_rate": 0.00017930885254625832,
+      "loss": 0.2912,
+      "step": 1540
+    },
+    {
+      "epoch": 0.8637892376681614,
+      "grad_norm": 0.07584706875354484,
+      "learning_rate": 0.00017926910220016667,
+      "loss": 0.3029,
+      "step": 1541
+    },
+    {
+      "epoch": 0.8643497757847534,
+      "grad_norm": 0.07615015482798214,
+      "learning_rate": 0.0001792293181229588,
+      "loss": 0.2887,
+      "step": 1542
+    },
+    {
+      "epoch": 0.8649103139013453,
+      "grad_norm": 0.07157336214839691,
+      "learning_rate": 0.00017918950033156384,
+      "loss": 0.283,
+      "step": 1543
+    },
+    {
+      "epoch": 0.8654708520179372,
+      "grad_norm": 0.07478197968145954,
+      "learning_rate": 0.00017914964884292544,
+      "loss": 0.2822,
+      "step": 1544
+    },
+    {
+      "epoch": 0.8660313901345291,
+      "grad_norm": 0.07670048948114096,
+      "learning_rate": 0.0001791097636740014,
+      "loss": 0.2895,
+      "step": 1545
+    },
+    {
+      "epoch": 0.8665919282511211,
+      "grad_norm": 0.0756401758028447,
+      "learning_rate": 0.0001790698448417639,
+      "loss": 0.2807,
+      "step": 1546
+    },
+    {
+      "epoch": 0.867152466367713,
+      "grad_norm": 0.07682176437281332,
+      "learning_rate": 0.00017902989236319954,
+      "loss": 0.29,
+      "step": 1547
+    },
+    {
+      "epoch": 0.8677130044843049,
+      "grad_norm": 0.07527893824548262,
+      "learning_rate": 0.0001789899062553091,
+      "loss": 0.3,
+      "step": 1548
+    },
+    {
+      "epoch": 0.8682735426008968,
+      "grad_norm": 0.0734658991643186,
+      "learning_rate": 0.0001789498865351078,
+      "loss": 0.2888,
+      "step": 1549
+    },
+    {
+      "epoch": 0.8688340807174888,
+      "grad_norm": 0.07338358151768215,
+      "learning_rate": 0.00017890983321962501,
+      "loss": 0.2897,
+      "step": 1550
+    },
+    {
+      "epoch": 0.8693946188340808,
+      "grad_norm": 0.07335106174051759,
+      "learning_rate": 0.0001788697463259046,
+      "loss": 0.2854,
+      "step": 1551
+    },
+    {
+      "epoch": 0.8699551569506726,
+      "grad_norm": 0.07173539150981001,
+      "learning_rate": 0.0001788296258710045,
+      "loss": 0.2847,
+      "step": 1552
+    },
+    {
+      "epoch": 0.8705156950672646,
+      "grad_norm": 0.07242438890257322,
+      "learning_rate": 0.0001787894718719971,
+      "loss": 0.2948,
+      "step": 1553
+    },
+    {
+      "epoch": 0.8710762331838565,
+      "grad_norm": 0.0733302232690697,
+      "learning_rate": 0.00017874928434596896,
+      "loss": 0.2907,
+      "step": 1554
+    },
+    {
+      "epoch": 0.8716367713004485,
+      "grad_norm": 0.07290729598111374,
+      "learning_rate": 0.00017870906331002098,
+      "loss": 0.2935,
+      "step": 1555
+    },
+    {
+      "epoch": 0.8721973094170403,
+      "grad_norm": 0.07418692156886422,
+      "learning_rate": 0.00017866880878126824,
+      "loss": 0.2956,
+      "step": 1556
+    },
+    {
+      "epoch": 0.8727578475336323,
+      "grad_norm": 0.07510192303785496,
+      "learning_rate": 0.00017862852077684015,
+      "loss": 0.2922,
+      "step": 1557
+    },
+    {
+      "epoch": 0.8733183856502242,
+      "grad_norm": 0.07293741181847814,
+      "learning_rate": 0.00017858819931388032,
+      "loss": 0.2729,
+      "step": 1558
+    },
+    {
+      "epoch": 0.8738789237668162,
+      "grad_norm": 0.0727239038880651,
+      "learning_rate": 0.0001785478444095466,
+      "loss": 0.2871,
+      "step": 1559
+    },
+    {
+      "epoch": 0.874439461883408,
+      "grad_norm": 0.07136868555596285,
+      "learning_rate": 0.0001785074560810111,
+      "loss": 0.2824,
+      "step": 1560
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 0.07365218153638509,
+      "learning_rate": 0.0001784670343454601,
+      "loss": 0.2884,
+      "step": 1561
+    },
+    {
+      "epoch": 0.875560538116592,
+      "grad_norm": 0.0744107435266324,
+      "learning_rate": 0.00017842657922009415,
+      "loss": 0.2941,
+      "step": 1562
+    },
+    {
+      "epoch": 0.8761210762331838,
+      "grad_norm": 0.07692989521022398,
+      "learning_rate": 0.00017838609072212794,
+      "loss": 0.2856,
+      "step": 1563
+    },
+    {
+      "epoch": 0.8766816143497758,
+      "grad_norm": 0.07615571543745912,
+      "learning_rate": 0.00017834556886879045,
+      "loss": 0.2987,
+      "step": 1564
+    },
+    {
+      "epoch": 0.8772421524663677,
+      "grad_norm": 0.07303922127135919,
+      "learning_rate": 0.00017830501367732484,
+      "loss": 0.291,
+      "step": 1565
+    },
+    {
+      "epoch": 0.8778026905829597,
+      "grad_norm": 0.07672307658101149,
+      "learning_rate": 0.00017826442516498837,
+      "loss": 0.2976,
+      "step": 1566
+    },
+    {
+      "epoch": 0.8783632286995515,
+      "grad_norm": 0.07441023986700969,
+      "learning_rate": 0.00017822380334905251,
+      "loss": 0.2955,
+      "step": 1567
+    },
+    {
+      "epoch": 0.8789237668161435,
+      "grad_norm": 0.07523563038497909,
+      "learning_rate": 0.000178183148246803,
+      "loss": 0.2719,
+      "step": 1568
+    },
+    {
+      "epoch": 0.8794843049327354,
+      "grad_norm": 0.07415243357870113,
+      "learning_rate": 0.00017814245987553962,
+      "loss": 0.2809,
+      "step": 1569
+    },
+    {
+      "epoch": 0.8800448430493274,
+      "grad_norm": 0.07525867364687405,
+      "learning_rate": 0.00017810173825257635,
+      "loss": 0.2736,
+      "step": 1570
+    },
+    {
+      "epoch": 0.8806053811659192,
+      "grad_norm": 0.074790495663368,
+      "learning_rate": 0.00017806098339524136,
+      "loss": 0.2864,
+      "step": 1571
+    },
+    {
+      "epoch": 0.8811659192825112,
+      "grad_norm": 0.07251174855671035,
+      "learning_rate": 0.00017802019532087694,
+      "loss": 0.2903,
+      "step": 1572
+    },
+    {
+      "epoch": 0.8817264573991032,
+      "grad_norm": 0.07246612243866497,
+      "learning_rate": 0.00017797937404683944,
+      "loss": 0.2878,
+      "step": 1573
+    },
+    {
+      "epoch": 0.8822869955156951,
+      "grad_norm": 0.0738823540378283,
+      "learning_rate": 0.00017793851959049944,
+      "loss": 0.3034,
+      "step": 1574
+    },
+    {
+      "epoch": 0.882847533632287,
+      "grad_norm": 0.07284419168932472,
+      "learning_rate": 0.00017789763196924163,
+      "loss": 0.2848,
+      "step": 1575
+    },
+    {
+      "epoch": 0.8834080717488789,
+      "grad_norm": 0.07194220334685288,
+      "learning_rate": 0.00017785671120046473,
+      "loss": 0.2915,
+      "step": 1576
+    },
+    {
+      "epoch": 0.8839686098654709,
+      "grad_norm": 0.07103917207526902,
+      "learning_rate": 0.00017781575730158164,
+      "loss": 0.2676,
+      "step": 1577
+    },
+    {
+      "epoch": 0.8845291479820628,
+      "grad_norm": 0.07243230180954226,
+      "learning_rate": 0.00017777477029001933,
+      "loss": 0.2903,
+      "step": 1578
+    },
+    {
+      "epoch": 0.8850896860986547,
+      "grad_norm": 0.07703232437862763,
+      "learning_rate": 0.00017773375018321886,
+      "loss": 0.3024,
+      "step": 1579
+    },
+    {
+      "epoch": 0.8856502242152466,
+      "grad_norm": 0.07412623943492334,
+      "learning_rate": 0.00017769269699863542,
+      "loss": 0.2859,
+      "step": 1580
+    },
+    {
+      "epoch": 0.8862107623318386,
+      "grad_norm": 0.07321862402476478,
+      "learning_rate": 0.00017765161075373816,
+      "loss": 0.2807,
+      "step": 1581
+    },
+    {
+      "epoch": 0.8867713004484304,
+      "grad_norm": 0.07472200715797371,
+      "learning_rate": 0.00017761049146601047,
+      "loss": 0.2852,
+      "step": 1582
+    },
+    {
+      "epoch": 0.8873318385650224,
+      "grad_norm": 0.07208903613257223,
+      "learning_rate": 0.00017756933915294963,
+      "loss": 0.279,
+      "step": 1583
+    },
+    {
+      "epoch": 0.8878923766816144,
+      "grad_norm": 0.07307170811980906,
+      "learning_rate": 0.00017752815383206705,
+      "loss": 0.2728,
+      "step": 1584
+    },
+    {
+      "epoch": 0.8884529147982063,
+      "grad_norm": 0.07370855397672015,
+      "learning_rate": 0.00017748693552088822,
+      "loss": 0.2887,
+      "step": 1585
+    },
+    {
+      "epoch": 0.8890134529147982,
+      "grad_norm": 0.07496967397687773,
+      "learning_rate": 0.00017744568423695259,
+      "loss": 0.295,
+      "step": 1586
+    },
+    {
+      "epoch": 0.8895739910313901,
+      "grad_norm": 0.07189932754460883,
+      "learning_rate": 0.0001774043999978137,
+      "loss": 0.2905,
+      "step": 1587
+    },
+    {
+      "epoch": 0.8901345291479821,
+      "grad_norm": 0.07126333198542388,
+      "learning_rate": 0.00017736308282103908,
+      "loss": 0.2844,
+      "step": 1588
+    },
+    {
+      "epoch": 0.890695067264574,
+      "grad_norm": 0.0718213911978105,
+      "learning_rate": 0.0001773217327242103,
+      "loss": 0.2861,
+      "step": 1589
+    },
+    {
+      "epoch": 0.8912556053811659,
+      "grad_norm": 0.07343198943939576,
+      "learning_rate": 0.00017728034972492297,
+      "loss": 0.3002,
+      "step": 1590
+    },
+    {
+      "epoch": 0.8918161434977578,
+      "grad_norm": 0.07275212591986045,
+      "learning_rate": 0.0001772389338407866,
+      "loss": 0.2821,
+      "step": 1591
+    },
+    {
+      "epoch": 0.8923766816143498,
+      "grad_norm": 0.07531081626251544,
+      "learning_rate": 0.0001771974850894248,
+      "loss": 0.2887,
+      "step": 1592
+    },
+    {
+      "epoch": 0.8929372197309418,
+      "grad_norm": 0.07279085656429406,
+      "learning_rate": 0.00017715600348847506,
+      "loss": 0.2877,
+      "step": 1593
+    },
+    {
+      "epoch": 0.8934977578475336,
+      "grad_norm": 0.07301727573273224,
+      "learning_rate": 0.00017711448905558897,
+      "loss": 0.2716,
+      "step": 1594
+    },
+    {
+      "epoch": 0.8940582959641256,
+      "grad_norm": 0.0733169780867396,
+      "learning_rate": 0.00017707294180843196,
+      "loss": 0.287,
+      "step": 1595
+    },
+    {
+      "epoch": 0.8946188340807175,
+      "grad_norm": 0.07721743970204584,
+      "learning_rate": 0.00017703136176468355,
+      "loss": 0.2905,
+      "step": 1596
+    },
+    {
+      "epoch": 0.8951793721973094,
+      "grad_norm": 0.07462719746638331,
+      "learning_rate": 0.0001769897489420371,
+      "loss": 0.2942,
+      "step": 1597
+    },
+    {
+      "epoch": 0.8957399103139013,
+      "grad_norm": 0.07088383450363162,
+      "learning_rate": 0.00017694810335820008,
+      "loss": 0.2832,
+      "step": 1598
+    },
+    {
+      "epoch": 0.8963004484304933,
+      "grad_norm": 0.07341824885173188,
+      "learning_rate": 0.0001769064250308937,
+      "loss": 0.2951,
+      "step": 1599
+    },
+    {
+      "epoch": 0.8968609865470852,
+      "grad_norm": 0.07462788854070883,
+      "learning_rate": 0.0001768647139778532,
+      "loss": 0.2829,
+      "step": 1600
+    },
+    {
+      "epoch": 0.8974215246636771,
+      "grad_norm": 0.07150287687830631,
+      "learning_rate": 0.0001768229702168278,
+      "loss": 0.2885,
+      "step": 1601
+    },
+    {
+      "epoch": 0.897982062780269,
+      "grad_norm": 0.07109980541289707,
+      "learning_rate": 0.00017678119376558055,
+      "loss": 0.2884,
+      "step": 1602
+    },
+    {
+      "epoch": 0.898542600896861,
+      "grad_norm": 0.0724317268039019,
+      "learning_rate": 0.00017673938464188847,
+      "loss": 0.3012,
+      "step": 1603
+    },
+    {
+      "epoch": 0.899103139013453,
+      "grad_norm": 0.0722355061570106,
+      "learning_rate": 0.00017669754286354241,
+      "loss": 0.293,
+      "step": 1604
+    },
+    {
+      "epoch": 0.8996636771300448,
+      "grad_norm": 0.07023265944433922,
+      "learning_rate": 0.00017665566844834717,
+      "loss": 0.2792,
+      "step": 1605
+    },
+    {
+      "epoch": 0.9002242152466368,
+      "grad_norm": 0.07453291308877033,
+      "learning_rate": 0.0001766137614141215,
+      "loss": 0.3005,
+      "step": 1606
+    },
+    {
+      "epoch": 0.9007847533632287,
+      "grad_norm": 0.0781899935366532,
+      "learning_rate": 0.00017657182177869787,
+      "loss": 0.2982,
+      "step": 1607
+    },
+    {
+      "epoch": 0.9013452914798207,
+      "grad_norm": 0.07384604421356011,
+      "learning_rate": 0.00017652984955992277,
+      "loss": 0.2874,
+      "step": 1608
+    },
+    {
+      "epoch": 0.9019058295964125,
+      "grad_norm": 0.07543178994690473,
+      "learning_rate": 0.00017648784477565648,
+      "loss": 0.2823,
+      "step": 1609
+    },
+    {
+      "epoch": 0.9024663677130045,
+      "grad_norm": 0.07647822461198672,
+      "learning_rate": 0.0001764458074437731,
+      "loss": 0.2946,
+      "step": 1610
+    },
+    {
+      "epoch": 0.9030269058295964,
+      "grad_norm": 0.07427298006381126,
+      "learning_rate": 0.00017640373758216077,
+      "loss": 0.2801,
+      "step": 1611
+    },
+    {
+      "epoch": 0.9035874439461884,
+      "grad_norm": 0.07085855311195798,
+      "learning_rate": 0.00017636163520872122,
+      "loss": 0.2822,
+      "step": 1612
+    },
+    {
+      "epoch": 0.9041479820627802,
+      "grad_norm": 0.07275776548540917,
+      "learning_rate": 0.00017631950034137015,
+      "loss": 0.2782,
+      "step": 1613
+    },
+    {
+      "epoch": 0.9047085201793722,
+      "grad_norm": 0.07289352731265318,
+      "learning_rate": 0.00017627733299803712,
+      "loss": 0.2964,
+      "step": 1614
+    },
+    {
+      "epoch": 0.9052690582959642,
+      "grad_norm": 0.07577448675281252,
+      "learning_rate": 0.00017623513319666543,
+      "loss": 0.2892,
+      "step": 1615
+    },
+    {
+      "epoch": 0.905829596412556,
+      "grad_norm": 0.07269049085417104,
+      "learning_rate": 0.0001761929009552122,
+      "loss": 0.2931,
+      "step": 1616
+    },
+    {
+      "epoch": 0.906390134529148,
+      "grad_norm": 0.07216742524982867,
+      "learning_rate": 0.00017615063629164838,
+      "loss": 0.2916,
+      "step": 1617
+    },
+    {
+      "epoch": 0.9069506726457399,
+      "grad_norm": 0.07521766155167502,
+      "learning_rate": 0.00017610833922395878,
+      "loss": 0.2777,
+      "step": 1618
+    },
+    {
+      "epoch": 0.9075112107623319,
+      "grad_norm": 0.07783378306532199,
+      "learning_rate": 0.00017606600977014184,
+      "loss": 0.297,
+      "step": 1619
+    },
+    {
+      "epoch": 0.9080717488789237,
+      "grad_norm": 0.07533264268788593,
+      "learning_rate": 0.0001760236479482099,
+      "loss": 0.2859,
+      "step": 1620
+    },
+    {
+      "epoch": 0.9086322869955157,
+      "grad_norm": 0.07778834560707679,
+      "learning_rate": 0.00017598125377618905,
+      "loss": 0.2865,
+      "step": 1621
+    },
+    {
+      "epoch": 0.9091928251121076,
+      "grad_norm": 0.07532574463796318,
+      "learning_rate": 0.00017593882727211916,
+      "loss": 0.2877,
+      "step": 1622
+    },
+    {
+      "epoch": 0.9097533632286996,
+      "grad_norm": 0.07550661399566316,
+      "learning_rate": 0.00017589636845405376,
+      "loss": 0.2888,
+      "step": 1623
+    },
+    {
+      "epoch": 0.9103139013452914,
+      "grad_norm": 0.07215447422793152,
+      "learning_rate": 0.00017585387734006034,
+      "loss": 0.2877,
+      "step": 1624
+    },
+    {
+      "epoch": 0.9108744394618834,
+      "grad_norm": 0.07394756569605147,
+      "learning_rate": 0.0001758113539482199,
+      "loss": 0.2784,
+      "step": 1625
+    },
+    {
+      "epoch": 0.9114349775784754,
+      "grad_norm": 0.07481048058195179,
+      "learning_rate": 0.00017576879829662732,
+      "loss": 0.2859,
+      "step": 1626
+    },
+    {
+      "epoch": 0.9119955156950673,
+      "grad_norm": 0.07495227012462921,
+      "learning_rate": 0.00017572621040339113,
+      "loss": 0.2835,
+      "step": 1627
+    },
+    {
+      "epoch": 0.9125560538116592,
+      "grad_norm": 0.07215411798909717,
+      "learning_rate": 0.00017568359028663364,
+      "loss": 0.2848,
+      "step": 1628
+    },
+    {
+      "epoch": 0.9131165919282511,
+      "grad_norm": 0.07477079461228578,
+      "learning_rate": 0.00017564093796449087,
+      "loss": 0.2922,
+      "step": 1629
+    },
+    {
+      "epoch": 0.9136771300448431,
+      "grad_norm": 0.07122569878337474,
+      "learning_rate": 0.00017559825345511243,
+      "loss": 0.2841,
+      "step": 1630
+    },
+    {
+      "epoch": 0.9142376681614349,
+      "grad_norm": 0.07319260835369037,
+      "learning_rate": 0.00017555553677666184,
+      "loss": 0.2829,
+      "step": 1631
+    },
+    {
+      "epoch": 0.9147982062780269,
+      "grad_norm": 0.0730504020630298,
+      "learning_rate": 0.00017551278794731607,
+      "loss": 0.2879,
+      "step": 1632
+    },
+    {
+      "epoch": 0.9153587443946188,
+      "grad_norm": 0.07413813123792716,
+      "learning_rate": 0.00017547000698526596,
+      "loss": 0.2905,
+      "step": 1633
+    },
+    {
+      "epoch": 0.9159192825112108,
+      "grad_norm": 0.07395622659326352,
+      "learning_rate": 0.00017542719390871593,
+      "loss": 0.2998,
+      "step": 1634
+    },
+    {
+      "epoch": 0.9164798206278026,
+      "grad_norm": 0.07323445118900361,
+      "learning_rate": 0.00017538434873588408,
+      "loss": 0.2828,
+      "step": 1635
+    },
+    {
+      "epoch": 0.9170403587443946,
+      "grad_norm": 0.06841709465287436,
+      "learning_rate": 0.0001753414714850022,
+      "loss": 0.2903,
+      "step": 1636
+    },
+    {
+      "epoch": 0.9176008968609866,
+      "grad_norm": 0.07413478072911547,
+      "learning_rate": 0.00017529856217431567,
+      "loss": 0.2839,
+      "step": 1637
+    },
+    {
+      "epoch": 0.9181614349775785,
+      "grad_norm": 0.07754698812595993,
+      "learning_rate": 0.00017525562082208355,
+      "loss": 0.2876,
+      "step": 1638
+    },
+    {
+      "epoch": 0.9187219730941704,
+      "grad_norm": 0.07327624458482546,
+      "learning_rate": 0.00017521264744657856,
+      "loss": 0.2837,
+      "step": 1639
+    },
+    {
+      "epoch": 0.9192825112107623,
+      "grad_norm": 0.07558291745709474,
+      "learning_rate": 0.00017516964206608696,
+      "loss": 0.2863,
+      "step": 1640
+    },
+    {
+      "epoch": 0.9198430493273543,
+      "grad_norm": 0.07558419350710056,
+      "learning_rate": 0.0001751266046989087,
+      "loss": 0.3001,
+      "step": 1641
+    },
+    {
+      "epoch": 0.9204035874439462,
+      "grad_norm": 0.0782416546651362,
+      "learning_rate": 0.0001750835353633574,
+      "loss": 0.3039,
+      "step": 1642
+    },
+    {
+      "epoch": 0.9209641255605381,
+      "grad_norm": 0.07546183397911319,
+      "learning_rate": 0.00017504043407776015,
+      "loss": 0.2862,
+      "step": 1643
+    },
+    {
+      "epoch": 0.92152466367713,
+      "grad_norm": 0.0756112604900152,
+      "learning_rate": 0.00017499730086045767,
+      "loss": 0.3063,
+      "step": 1644
+    },
+    {
+      "epoch": 0.922085201793722,
+      "grad_norm": 0.07215444054397309,
+      "learning_rate": 0.00017495413572980435,
+      "loss": 0.2893,
+      "step": 1645
+    },
+    {
+      "epoch": 0.922645739910314,
+      "grad_norm": 0.07046996461513384,
+      "learning_rate": 0.00017491093870416807,
+      "loss": 0.2848,
+      "step": 1646
+    },
+    {
+      "epoch": 0.9232062780269058,
+      "grad_norm": 0.07099289832624417,
+      "learning_rate": 0.00017486770980193033,
+      "loss": 0.2837,
+      "step": 1647
+    },
+    {
+      "epoch": 0.9237668161434978,
+      "grad_norm": 0.07204428919620674,
+      "learning_rate": 0.00017482444904148617,
+      "loss": 0.282,
+      "step": 1648
+    },
+    {
+      "epoch": 0.9243273542600897,
+      "grad_norm": 0.07432187501781345,
+      "learning_rate": 0.00017478115644124423,
+      "loss": 0.2943,
+      "step": 1649
+    },
+    {
+      "epoch": 0.9248878923766816,
+      "grad_norm": 0.06906577445089956,
+      "learning_rate": 0.00017473783201962665,
+      "loss": 0.2882,
+      "step": 1650
+    },
+    {
+      "epoch": 0.9254484304932735,
+      "grad_norm": 0.07260085951472438,
+      "learning_rate": 0.00017469447579506907,
+      "loss": 0.2848,
+      "step": 1651
+    },
+    {
+      "epoch": 0.9260089686098655,
+      "grad_norm": 0.07075547934294112,
+      "learning_rate": 0.0001746510877860208,
+      "loss": 0.2852,
+      "step": 1652
+    },
+    {
+      "epoch": 0.9265695067264574,
+      "grad_norm": 0.07525858681227156,
+      "learning_rate": 0.00017460766801094454,
+      "loss": 0.2809,
+      "step": 1653
+    },
+    {
+      "epoch": 0.9271300448430493,
+      "grad_norm": 0.07234387482659316,
+      "learning_rate": 0.00017456421648831655,
+      "loss": 0.2876,
+      "step": 1654
+    },
+    {
+      "epoch": 0.9276905829596412,
+      "grad_norm": 0.07487929033557668,
+      "learning_rate": 0.0001745207332366267,
+      "loss": 0.2899,
+      "step": 1655
+    },
+    {
+      "epoch": 0.9282511210762332,
+      "grad_norm": 0.07259129422906563,
+      "learning_rate": 0.0001744772182743782,
+      "loss": 0.295,
+      "step": 1656
+    },
+    {
+      "epoch": 0.9288116591928252,
+      "grad_norm": 0.07078856411822994,
+      "learning_rate": 0.00017443367162008785,
+      "loss": 0.2786,
+      "step": 1657
+    },
+    {
+      "epoch": 0.929372197309417,
+      "grad_norm": 0.07194721171963898,
+      "learning_rate": 0.00017439009329228586,
+      "loss": 0.2639,
+      "step": 1658
+    },
+    {
+      "epoch": 0.929932735426009,
+      "grad_norm": 0.07178484282976831,
+      "learning_rate": 0.00017434648330951605,
+      "loss": 0.2871,
+      "step": 1659
+    },
+    {
+      "epoch": 0.9304932735426009,
+      "grad_norm": 0.07340063223561963,
+      "learning_rate": 0.0001743028416903356,
+      "loss": 0.2806,
+      "step": 1660
+    },
+    {
+      "epoch": 0.9310538116591929,
+      "grad_norm": 0.07566542410927178,
+      "learning_rate": 0.00017425916845331517,
+      "loss": 0.2896,
+      "step": 1661
+    },
+    {
+      "epoch": 0.9316143497757847,
+      "grad_norm": 0.07486497673865919,
+      "learning_rate": 0.0001742154636170389,
+      "loss": 0.2911,
+      "step": 1662
+    },
+    {
+      "epoch": 0.9321748878923767,
+      "grad_norm": 0.07511359840039579,
+      "learning_rate": 0.00017417172720010434,
+      "loss": 0.2989,
+      "step": 1663
+    },
+    {
+      "epoch": 0.9327354260089686,
+      "grad_norm": 0.07557817479695081,
+      "learning_rate": 0.00017412795922112253,
+      "loss": 0.2752,
+      "step": 1664
+    },
+    {
+      "epoch": 0.9332959641255605,
+      "grad_norm": 0.07461501444770283,
+      "learning_rate": 0.0001740841596987179,
+      "loss": 0.2875,
+      "step": 1665
+    },
+    {
+      "epoch": 0.9338565022421524,
+      "grad_norm": 0.0693166526056427,
+      "learning_rate": 0.00017404032865152834,
+      "loss": 0.2842,
+      "step": 1666
+    },
+    {
+      "epoch": 0.9344170403587444,
+      "grad_norm": 0.07322495815613736,
+      "learning_rate": 0.00017399646609820505,
+      "loss": 0.2825,
+      "step": 1667
+    },
+    {
+      "epoch": 0.9349775784753364,
+      "grad_norm": 0.07665243071009871,
+      "learning_rate": 0.0001739525720574128,
+      "loss": 0.2915,
+      "step": 1668
+    },
+    {
+      "epoch": 0.9355381165919282,
+      "grad_norm": 0.07737383882355353,
+      "learning_rate": 0.00017390864654782964,
+      "loss": 0.2998,
+      "step": 1669
+    },
+    {
+      "epoch": 0.9360986547085202,
+      "grad_norm": 0.0737494468927531,
+      "learning_rate": 0.00017386468958814706,
+      "loss": 0.2878,
+      "step": 1670
+    },
+    {
+      "epoch": 0.9366591928251121,
+      "grad_norm": 0.0714066787744356,
+      "learning_rate": 0.00017382070119706988,
+      "loss": 0.2813,
+      "step": 1671
+    },
+    {
+      "epoch": 0.9372197309417041,
+      "grad_norm": 0.07043150060159434,
+      "learning_rate": 0.0001737766813933164,
+      "loss": 0.2784,
+      "step": 1672
+    },
+    {
+      "epoch": 0.9377802690582959,
+      "grad_norm": 0.07227826616766787,
+      "learning_rate": 0.00017373263019561814,
+      "loss": 0.2836,
+      "step": 1673
+    },
+    {
+      "epoch": 0.9383408071748879,
+      "grad_norm": 0.0733661404805429,
+      "learning_rate": 0.00017368854762272014,
+      "loss": 0.2799,
+      "step": 1674
+    },
+    {
+      "epoch": 0.9389013452914798,
+      "grad_norm": 0.07071892951110591,
+      "learning_rate": 0.00017364443369338064,
+      "loss": 0.2678,
+      "step": 1675
+    },
+    {
+      "epoch": 0.9394618834080718,
+      "grad_norm": 0.07470850463250955,
+      "learning_rate": 0.0001736002884263713,
+      "loss": 0.2949,
+      "step": 1676
+    },
+    {
+      "epoch": 0.9400224215246636,
+      "grad_norm": 0.07278768556263866,
+      "learning_rate": 0.00017355611184047718,
+      "loss": 0.274,
+      "step": 1677
+    },
+    {
+      "epoch": 0.9405829596412556,
+      "grad_norm": 0.07258556909282828,
+      "learning_rate": 0.00017351190395449651,
+      "loss": 0.282,
+      "step": 1678
+    },
+    {
+      "epoch": 0.9411434977578476,
+      "grad_norm": 0.07444418686302096,
+      "learning_rate": 0.000173467664787241,
+      "loss": 0.2888,
+      "step": 1679
+    },
+    {
+      "epoch": 0.9417040358744395,
+      "grad_norm": 0.07728184646618526,
+      "learning_rate": 0.00017342339435753553,
+      "loss": 0.3073,
+      "step": 1680
+    },
+    {
+      "epoch": 0.9422645739910314,
+      "grad_norm": 0.07412477436974288,
+      "learning_rate": 0.00017337909268421835,
+      "loss": 0.2878,
+      "step": 1681
+    },
+    {
+      "epoch": 0.9428251121076233,
+      "grad_norm": 0.0741828737276978,
+      "learning_rate": 0.00017333475978614107,
+      "loss": 0.301,
+      "step": 1682
+    },
+    {
+      "epoch": 0.9433856502242153,
+      "grad_norm": 0.07254070658542294,
+      "learning_rate": 0.00017329039568216844,
+      "loss": 0.2745,
+      "step": 1683
+    },
+    {
+      "epoch": 0.9439461883408071,
+      "grad_norm": 0.07217470374090927,
+      "learning_rate": 0.00017324600039117863,
+      "loss": 0.2835,
+      "step": 1684
+    },
+    {
+      "epoch": 0.9445067264573991,
+      "grad_norm": 0.07196992715424201,
+      "learning_rate": 0.00017320157393206298,
+      "loss": 0.2831,
+      "step": 1685
+    },
+    {
+      "epoch": 0.945067264573991,
+      "grad_norm": 0.07168226957448584,
+      "learning_rate": 0.00017315711632372613,
+      "loss": 0.2803,
+      "step": 1686
+    },
+    {
+      "epoch": 0.945627802690583,
+      "grad_norm": 0.07358645350177724,
+      "learning_rate": 0.000173112627585086,
+      "loss": 0.2819,
+      "step": 1687
+    },
+    {
+      "epoch": 0.9461883408071748,
+      "grad_norm": 0.0748629407326928,
+      "learning_rate": 0.00017306810773507376,
+      "loss": 0.2838,
+      "step": 1688
+    },
+    {
+      "epoch": 0.9467488789237668,
+      "grad_norm": 0.07535040596126334,
+      "learning_rate": 0.00017302355679263377,
+      "loss": 0.2957,
+      "step": 1689
+    },
+    {
+      "epoch": 0.9473094170403588,
+      "grad_norm": 0.07008243670534321,
+      "learning_rate": 0.0001729789747767236,
+      "loss": 0.2927,
+      "step": 1690
+    },
+    {
+      "epoch": 0.9478699551569507,
+      "grad_norm": 0.0729678214876866,
+      "learning_rate": 0.00017293436170631415,
+      "loss": 0.2896,
+      "step": 1691
+    },
+    {
+      "epoch": 0.9484304932735426,
+      "grad_norm": 0.07339696701392524,
+      "learning_rate": 0.00017288971760038942,
+      "loss": 0.289,
+      "step": 1692
+    },
+    {
+      "epoch": 0.9489910313901345,
+      "grad_norm": 0.07197335218511028,
+      "learning_rate": 0.00017284504247794667,
+      "loss": 0.2933,
+      "step": 1693
+    },
+    {
+      "epoch": 0.9495515695067265,
+      "grad_norm": 0.07378140210096644,
+      "learning_rate": 0.0001728003363579964,
+      "loss": 0.2797,
+      "step": 1694
+    },
+    {
+      "epoch": 0.9501121076233184,
+      "grad_norm": 0.07116653651425976,
+      "learning_rate": 0.00017275559925956227,
+      "loss": 0.2728,
+      "step": 1695
+    },
+    {
+      "epoch": 0.9506726457399103,
+      "grad_norm": 0.06946675108565646,
+      "learning_rate": 0.00017271083120168102,
+      "loss": 0.2784,
+      "step": 1696
+    },
+    {
+      "epoch": 0.9512331838565022,
+      "grad_norm": 0.07416590486296964,
+      "learning_rate": 0.0001726660322034027,
+      "loss": 0.2873,
+      "step": 1697
+    },
+    {
+      "epoch": 0.9517937219730942,
+      "grad_norm": 0.07593399139838054,
+      "learning_rate": 0.0001726212022837905,
+      "loss": 0.2857,
+      "step": 1698
+    },
+    {
+      "epoch": 0.952354260089686,
+      "grad_norm": 0.0727408230288732,
+      "learning_rate": 0.00017257634146192072,
+      "loss": 0.2691,
+      "step": 1699
+    },
+    {
+      "epoch": 0.952914798206278,
+      "grad_norm": 0.07460940164156943,
+      "learning_rate": 0.00017253144975688285,
+      "loss": 0.2873,
+      "step": 1700
+    },
+    {
+      "epoch": 0.95347533632287,
+      "grad_norm": 0.07429803178082536,
+      "learning_rate": 0.0001724865271877795,
+      "loss": 0.2869,
+      "step": 1701
+    },
+    {
+      "epoch": 0.9540358744394619,
+      "grad_norm": 0.07134851108937684,
+      "learning_rate": 0.00017244157377372638,
+      "loss": 0.275,
+      "step": 1702
+    },
+    {
+      "epoch": 0.9545964125560538,
+      "grad_norm": 0.06986109810489552,
+      "learning_rate": 0.00017239658953385246,
+      "loss": 0.2837,
+      "step": 1703
+    },
+    {
+      "epoch": 0.9551569506726457,
+      "grad_norm": 0.06856388411678808,
+      "learning_rate": 0.00017235157448729967,
+      "loss": 0.2777,
+      "step": 1704
+    },
+    {
+      "epoch": 0.9557174887892377,
+      "grad_norm": 0.07019980323520368,
+      "learning_rate": 0.00017230652865322309,
+      "loss": 0.2855,
+      "step": 1705
+    },
+    {
+      "epoch": 0.9562780269058296,
+      "grad_norm": 0.07028149928057296,
+      "learning_rate": 0.00017226145205079095,
+      "loss": 0.2845,
+      "step": 1706
+    },
+    {
+      "epoch": 0.9568385650224215,
+      "grad_norm": 0.0722077813669913,
+      "learning_rate": 0.00017221634469918458,
+      "loss": 0.2819,
+      "step": 1707
+    },
+    {
+      "epoch": 0.9573991031390134,
+      "grad_norm": 0.07222438978417997,
+      "learning_rate": 0.00017217120661759832,
+      "loss": 0.2865,
+      "step": 1708
+    },
+    {
+      "epoch": 0.9579596412556054,
+      "grad_norm": 0.07357381335046854,
+      "learning_rate": 0.00017212603782523964,
+      "loss": 0.2786,
+      "step": 1709
+    },
+    {
+      "epoch": 0.9585201793721974,
+      "grad_norm": 0.07635328706561935,
+      "learning_rate": 0.00017208083834132905,
+      "loss": 0.2837,
+      "step": 1710
+    },
+    {
+      "epoch": 0.9590807174887892,
+      "grad_norm": 0.07295089505628365,
+      "learning_rate": 0.00017203560818510017,
+      "loss": 0.2917,
+      "step": 1711
+    },
+    {
+      "epoch": 0.9596412556053812,
+      "grad_norm": 0.07045358913278182,
+      "learning_rate": 0.0001719903473757996,
+      "loss": 0.2833,
+      "step": 1712
+    },
+    {
+      "epoch": 0.9602017937219731,
+      "grad_norm": 0.07228807028747083,
+      "learning_rate": 0.00017194505593268704,
+      "loss": 0.2873,
+      "step": 1713
+    },
+    {
+      "epoch": 0.9607623318385651,
+      "grad_norm": 0.07603221313096017,
+      "learning_rate": 0.00017189973387503522,
+      "loss": 0.2889,
+      "step": 1714
+    },
+    {
+      "epoch": 0.9613228699551569,
+      "grad_norm": 0.07307077879941619,
+      "learning_rate": 0.00017185438122212983,
+      "loss": 0.2933,
+      "step": 1715
+    },
+    {
+      "epoch": 0.9618834080717489,
+      "grad_norm": 0.07322017184361226,
+      "learning_rate": 0.0001718089979932697,
+      "loss": 0.2955,
+      "step": 1716
+    },
+    {
+      "epoch": 0.9624439461883408,
+      "grad_norm": 0.07295962133338377,
+      "learning_rate": 0.00017176358420776654,
+      "loss": 0.2873,
+      "step": 1717
+    },
+    {
+      "epoch": 0.9630044843049327,
+      "grad_norm": 0.07133986025044317,
+      "learning_rate": 0.00017171813988494522,
+      "loss": 0.2842,
+      "step": 1718
+    },
+    {
+      "epoch": 0.9635650224215246,
+      "grad_norm": 0.07403222948184751,
+      "learning_rate": 0.00017167266504414342,
+      "loss": 0.2913,
+      "step": 1719
+    },
+    {
+      "epoch": 0.9641255605381166,
+      "grad_norm": 0.07074358995016189,
+      "learning_rate": 0.0001716271597047119,
+      "loss": 0.2868,
+      "step": 1720
+    },
+    {
+      "epoch": 0.9646860986547086,
+      "grad_norm": 0.07377182669211226,
+      "learning_rate": 0.00017158162388601443,
+      "loss": 0.283,
+      "step": 1721
+    },
+    {
+      "epoch": 0.9652466367713004,
+      "grad_norm": 0.0703874372314276,
+      "learning_rate": 0.00017153605760742777,
+      "loss": 0.2791,
+      "step": 1722
+    },
+    {
+      "epoch": 0.9658071748878924,
+      "grad_norm": 0.07113289112728188,
+      "learning_rate": 0.00017149046088834146,
+      "loss": 0.2754,
+      "step": 1723
+    },
+    {
+      "epoch": 0.9663677130044843,
+      "grad_norm": 0.07261030259423171,
+      "learning_rate": 0.0001714448337481582,
+      "loss": 0.2946,
+      "step": 1724
+    },
+    {
+      "epoch": 0.9669282511210763,
+      "grad_norm": 0.07566204586329739,
+      "learning_rate": 0.00017139917620629356,
+      "loss": 0.3112,
+      "step": 1725
+    },
+    {
+      "epoch": 0.9674887892376681,
+      "grad_norm": 0.07371387803374542,
+      "learning_rate": 0.000171353488282176,
+      "loss": 0.2928,
+      "step": 1726
+    },
+    {
+      "epoch": 0.9680493273542601,
+      "grad_norm": 0.07108812790567268,
+      "learning_rate": 0.00017130776999524697,
+      "loss": 0.289,
+      "step": 1727
+    },
+    {
+      "epoch": 0.968609865470852,
+      "grad_norm": 0.07129036585172886,
+      "learning_rate": 0.0001712620213649608,
+      "loss": 0.2923,
+      "step": 1728
+    },
+    {
+      "epoch": 0.969170403587444,
+      "grad_norm": 0.07419267575581269,
+      "learning_rate": 0.00017121624241078477,
+      "loss": 0.2817,
+      "step": 1729
+    },
+    {
+      "epoch": 0.9697309417040358,
+      "grad_norm": 0.06964552695593264,
+      "learning_rate": 0.0001711704331521991,
+      "loss": 0.2839,
+      "step": 1730
+    },
+    {
+      "epoch": 0.9702914798206278,
+      "grad_norm": 0.0704547099569811,
+      "learning_rate": 0.00017112459360869674,
+      "loss": 0.2908,
+      "step": 1731
+    },
+    {
+      "epoch": 0.9708520179372198,
+      "grad_norm": 0.07062667985179868,
+      "learning_rate": 0.00017107872379978374,
+      "loss": 0.2902,
+      "step": 1732
+    },
+    {
+      "epoch": 0.9714125560538116,
+      "grad_norm": 0.0749374524137223,
+      "learning_rate": 0.00017103282374497883,
+      "loss": 0.2826,
+      "step": 1733
+    },
+    {
+      "epoch": 0.9719730941704036,
+      "grad_norm": 0.07667026279177375,
+      "learning_rate": 0.0001709868934638138,
+      "loss": 0.2977,
+      "step": 1734
+    },
+    {
+      "epoch": 0.9725336322869955,
+      "grad_norm": 0.07429141839562514,
+      "learning_rate": 0.00017094093297583316,
+      "loss": 0.293,
+      "step": 1735
+    },
+    {
+      "epoch": 0.9730941704035875,
+      "grad_norm": 0.07384385025776573,
+      "learning_rate": 0.00017089494230059432,
+      "loss": 0.2769,
+      "step": 1736
+    },
+    {
+      "epoch": 0.9736547085201793,
+      "grad_norm": 0.07170487252525123,
+      "learning_rate": 0.00017084892145766755,
+      "loss": 0.2964,
+      "step": 1737
+    },
+    {
+      "epoch": 0.9742152466367713,
+      "grad_norm": 0.07015361855556726,
+      "learning_rate": 0.00017080287046663596,
+      "loss": 0.2792,
+      "step": 1738
+    },
+    {
+      "epoch": 0.9747757847533632,
+      "grad_norm": 0.06983521640448365,
+      "learning_rate": 0.00017075678934709543,
+      "loss": 0.2738,
+      "step": 1739
+    },
+    {
+      "epoch": 0.9753363228699552,
+      "grad_norm": 0.07268793667525557,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 0.291,
+      "step": 1740
+    },
+    {
+      "epoch": 0.975896860986547,
+      "grad_norm": 0.0695631561511414,
+      "learning_rate": 0.00017066453680093547,
+      "loss": 0.2709,
+      "step": 1741
+    },
+    {
+      "epoch": 0.976457399103139,
+      "grad_norm": 0.06951619746345279,
+      "learning_rate": 0.00017061836541357192,
+      "loss": 0.2691,
+      "step": 1742
+    },
+    {
+      "epoch": 0.977017937219731,
+      "grad_norm": 0.07315230646262866,
+      "learning_rate": 0.0001705721639762113,
+      "loss": 0.2909,
+      "step": 1743
+    },
+    {
+      "epoch": 0.9775784753363229,
+      "grad_norm": 0.07655221744584981,
+      "learning_rate": 0.0001705259325085135,
+      "loss": 0.2979,
+      "step": 1744
+    },
+    {
+      "epoch": 0.9781390134529148,
+      "grad_norm": 0.07423686823518584,
+      "learning_rate": 0.00017047967103015133,
+      "loss": 0.2936,
+      "step": 1745
+    },
+    {
+      "epoch": 0.9786995515695067,
+      "grad_norm": 0.07187254002922631,
+      "learning_rate": 0.0001704333795608102,
+      "loss": 0.2802,
+      "step": 1746
+    },
+    {
+      "epoch": 0.9792600896860987,
+      "grad_norm": 0.07187169839889325,
+      "learning_rate": 0.00017038705812018833,
+      "loss": 0.2761,
+      "step": 1747
+    },
+    {
+      "epoch": 0.9798206278026906,
+      "grad_norm": 0.07175327447886823,
+      "learning_rate": 0.00017034070672799684,
+      "loss": 0.2718,
+      "step": 1748
+    },
+    {
+      "epoch": 0.9803811659192825,
+      "grad_norm": 0.07325962417801318,
+      "learning_rate": 0.00017029432540395943,
+      "loss": 0.282,
+      "step": 1749
+    },
+    {
+      "epoch": 0.9809417040358744,
+      "grad_norm": 0.07220968769746226,
+      "learning_rate": 0.00017024791416781257,
+      "loss": 0.2868,
+      "step": 1750
+    },
+    {
+      "epoch": 0.9815022421524664,
+      "grad_norm": 0.07092038680130973,
+      "learning_rate": 0.00017020147303930554,
+      "loss": 0.2911,
+      "step": 1751
+    },
+    {
+      "epoch": 0.9820627802690582,
+      "grad_norm": 0.06939572948741551,
+      "learning_rate": 0.00017015500203820022,
+      "loss": 0.2689,
+      "step": 1752
+    },
+    {
+      "epoch": 0.9826233183856502,
+      "grad_norm": 0.07479910602139032,
+      "learning_rate": 0.00017010850118427125,
+      "loss": 0.2915,
+      "step": 1753
+    },
+    {
+      "epoch": 0.9831838565022422,
+      "grad_norm": 0.0768870712403653,
+      "learning_rate": 0.00017006197049730602,
+      "loss": 0.3008,
+      "step": 1754
+    },
+    {
+      "epoch": 0.9837443946188341,
+      "grad_norm": 0.0720958557658177,
+      "learning_rate": 0.00017001540999710458,
+      "loss": 0.2901,
+      "step": 1755
+    },
+    {
+      "epoch": 0.984304932735426,
+      "grad_norm": 0.07102148035823705,
+      "learning_rate": 0.00016996881970347962,
+      "loss": 0.2782,
+      "step": 1756
+    },
+    {
+      "epoch": 0.9848654708520179,
+      "grad_norm": 0.07326873236410598,
+      "learning_rate": 0.00016992219963625659,
+      "loss": 0.2864,
+      "step": 1757
+    },
+    {
+      "epoch": 0.9854260089686099,
+      "grad_norm": 0.07239571035831743,
+      "learning_rate": 0.00016987554981527357,
+      "loss": 0.2878,
+      "step": 1758
+    },
+    {
+      "epoch": 0.9859865470852018,
+      "grad_norm": 0.06905207533713582,
+      "learning_rate": 0.00016982887026038132,
+      "loss": 0.2718,
+      "step": 1759
+    },
+    {
+      "epoch": 0.9865470852017937,
+      "grad_norm": 0.0737159363857612,
+      "learning_rate": 0.0001697821609914432,
+      "loss": 0.2871,
+      "step": 1760
+    },
+    {
+      "epoch": 0.9871076233183856,
+      "grad_norm": 0.07430574604340062,
+      "learning_rate": 0.00016973542202833528,
+      "loss": 0.2799,
+      "step": 1761
+    },
+    {
+      "epoch": 0.9876681614349776,
+      "grad_norm": 0.07367466835976355,
+      "learning_rate": 0.00016968865339094617,
+      "loss": 0.2833,
+      "step": 1762
+    },
+    {
+      "epoch": 0.9882286995515696,
+      "grad_norm": 0.07167264334942111,
+      "learning_rate": 0.00016964185509917725,
+      "loss": 0.2848,
+      "step": 1763
+    },
+    {
+      "epoch": 0.9887892376681614,
+      "grad_norm": 0.07155343589224945,
+      "learning_rate": 0.00016959502717294242,
+      "loss": 0.2809,
+      "step": 1764
+    },
+    {
+      "epoch": 0.9893497757847534,
+      "grad_norm": 0.07381396163039367,
+      "learning_rate": 0.00016954816963216817,
+      "loss": 0.2929,
+      "step": 1765
+    },
+    {
+      "epoch": 0.9899103139013453,
+      "grad_norm": 0.07316072572490417,
+      "learning_rate": 0.00016950128249679366,
+      "loss": 0.2827,
+      "step": 1766
+    },
+    {
+      "epoch": 0.9904708520179372,
+      "grad_norm": 0.06881598592091107,
+      "learning_rate": 0.00016945436578677065,
+      "loss": 0.2724,
+      "step": 1767
+    },
+    {
+      "epoch": 0.9910313901345291,
+      "grad_norm": 0.0737642707714976,
+      "learning_rate": 0.0001694074195220634,
+      "loss": 0.2831,
+      "step": 1768
+    },
+    {
+      "epoch": 0.9915919282511211,
+      "grad_norm": 0.07126875761505212,
+      "learning_rate": 0.0001693604437226488,
+      "loss": 0.2787,
+      "step": 1769
+    },
+    {
+      "epoch": 0.992152466367713,
+      "grad_norm": 0.07055264246428153,
+      "learning_rate": 0.00016931343840851634,
+      "loss": 0.2782,
+      "step": 1770
+    },
+    {
+      "epoch": 0.9927130044843049,
+      "grad_norm": 0.07393618688943292,
+      "learning_rate": 0.00016926640359966807,
+      "loss": 0.2862,
+      "step": 1771
+    },
+    {
+      "epoch": 0.9932735426008968,
+      "grad_norm": 0.07555718134908355,
+      "learning_rate": 0.0001692193393161184,
+      "loss": 0.2927,
+      "step": 1772
+    },
+    {
+      "epoch": 0.9938340807174888,
+      "grad_norm": 0.07570857481309304,
+      "learning_rate": 0.0001691722455778946,
+      "loss": 0.2824,
+      "step": 1773
+    },
+    {
+      "epoch": 0.9943946188340808,
+      "grad_norm": 0.07465306801537522,
+      "learning_rate": 0.00016912512240503625,
+      "loss": 0.2947,
+      "step": 1774
+    },
+    {
+      "epoch": 0.9949551569506726,
+      "grad_norm": 0.07004367926501669,
+      "learning_rate": 0.0001690779698175955,
+      "loss": 0.2744,
+      "step": 1775
+    },
+    {
+      "epoch": 0.9955156950672646,
+      "grad_norm": 0.07297217404931539,
+      "learning_rate": 0.000169030787835637,
+      "loss": 0.2685,
+      "step": 1776
+    },
+    {
+      "epoch": 0.9960762331838565,
+      "grad_norm": 0.07402597749753811,
+      "learning_rate": 0.0001689835764792381,
+      "loss": 0.3014,
+      "step": 1777
+    },
+    {
+      "epoch": 0.9966367713004485,
+      "grad_norm": 0.06893864854981804,
+      "learning_rate": 0.00016893633576848827,
+      "loss": 0.2673,
+      "step": 1778
+    },
+    {
+      "epoch": 0.9971973094170403,
+      "grad_norm": 0.07454637620033523,
+      "learning_rate": 0.00016888906572348988,
+      "loss": 0.3062,
+      "step": 1779
+    },
+    {
+      "epoch": 0.9977578475336323,
+      "grad_norm": 0.06892855385802012,
+      "learning_rate": 0.00016884176636435748,
+      "loss": 0.2849,
+      "step": 1780
+    },
+    {
+      "epoch": 0.9983183856502242,
+      "grad_norm": 0.07142555303088303,
+      "learning_rate": 0.00016879443771121826,
+      "loss": 0.2927,
+      "step": 1781
+    },
+    {
+      "epoch": 0.9988789237668162,
+      "grad_norm": 0.06913628944171628,
+      "learning_rate": 0.0001687470797842118,
+      "loss": 0.2833,
+      "step": 1782
+    },
+    {
+      "epoch": 0.999439461883408,
+      "grad_norm": 0.07329335793223181,
+      "learning_rate": 0.00016869969260349018,
+      "loss": 0.2758,
+      "step": 1783
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0731286536565907,
+      "learning_rate": 0.00016865227618921788,
+      "loss": 0.2882,
+      "step": 1784
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.28521716594696045,
+      "eval_runtime": 350.0541,
+      "eval_samples_per_second": 34.329,
+      "eval_steps_per_second": 1.074,
+      "step": 1784
+    },
+    {
+      "epoch": 1.0005605381165918,
+      "grad_norm": 0.07293250489340956,
+      "learning_rate": 0.00016860483056157187,
+      "loss": 0.2731,
+      "step": 1785
+    },
+    {
+      "epoch": 1.001121076233184,
+      "grad_norm": 0.07287987344760258,
+      "learning_rate": 0.00016855735574074153,
+      "loss": 0.2777,
+      "step": 1786
+    },
+    {
+      "epoch": 1.0016816143497758,
+      "grad_norm": 0.07024587084667601,
+      "learning_rate": 0.00016850985174692867,
+      "loss": 0.2686,
+      "step": 1787
+    },
+    {
+      "epoch": 1.0022421524663676,
+      "grad_norm": 0.07061542132913487,
+      "learning_rate": 0.00016846231860034747,
+      "loss": 0.2683,
+      "step": 1788
+    },
+    {
+      "epoch": 1.0028026905829597,
+      "grad_norm": 0.07234478832965124,
+      "learning_rate": 0.0001684147563212246,
+      "loss": 0.2668,
+      "step": 1789
+    },
+    {
+      "epoch": 1.0033632286995515,
+      "grad_norm": 0.07539927212010297,
+      "learning_rate": 0.00016836716492979903,
+      "loss": 0.2974,
+      "step": 1790
+    },
+    {
+      "epoch": 1.0039237668161436,
+      "grad_norm": 0.07191466876978402,
+      "learning_rate": 0.0001683195444463222,
+      "loss": 0.2721,
+      "step": 1791
+    },
+    {
+      "epoch": 1.0044843049327354,
+      "grad_norm": 0.07516331104667469,
+      "learning_rate": 0.00016827189489105788,
+      "loss": 0.2865,
+      "step": 1792
+    },
+    {
+      "epoch": 1.0050448430493273,
+      "grad_norm": 0.07705362114824074,
+      "learning_rate": 0.00016822421628428223,
+      "loss": 0.2779,
+      "step": 1793
+    },
+    {
+      "epoch": 1.0056053811659194,
+      "grad_norm": 0.07478865240503156,
+      "learning_rate": 0.00016817650864628375,
+      "loss": 0.2761,
+      "step": 1794
+    },
+    {
+      "epoch": 1.0061659192825112,
+      "grad_norm": 0.0720000034295649,
+      "learning_rate": 0.00016812877199736333,
+      "loss": 0.2579,
+      "step": 1795
+    },
+    {
+      "epoch": 1.006726457399103,
+      "grad_norm": 0.07435548623394256,
+      "learning_rate": 0.00016808100635783423,
+      "loss": 0.264,
+      "step": 1796
+    },
+    {
+      "epoch": 1.0072869955156951,
+      "grad_norm": 0.07183268780358426,
+      "learning_rate": 0.00016803321174802194,
+      "loss": 0.2706,
+      "step": 1797
+    },
+    {
+      "epoch": 1.007847533632287,
+      "grad_norm": 0.07248897595226891,
+      "learning_rate": 0.00016798538818826435,
+      "loss": 0.2815,
+      "step": 1798
+    },
+    {
+      "epoch": 1.008408071748879,
+      "grad_norm": 0.0761853411440106,
+      "learning_rate": 0.00016793753569891164,
+      "loss": 0.274,
+      "step": 1799
+    },
+    {
+      "epoch": 1.0089686098654709,
+      "grad_norm": 0.07087616772664414,
+      "learning_rate": 0.00016788965430032638,
+      "loss": 0.2587,
+      "step": 1800
+    },
+    {
+      "epoch": 1.0095291479820627,
+      "grad_norm": 0.07533376703956196,
+      "learning_rate": 0.00016784174401288335,
+      "loss": 0.2822,
+      "step": 1801
+    },
+    {
+      "epoch": 1.0100896860986548,
+      "grad_norm": 0.07790901551081032,
+      "learning_rate": 0.00016779380485696966,
+      "loss": 0.2876,
+      "step": 1802
+    },
+    {
+      "epoch": 1.0106502242152466,
+      "grad_norm": 0.075412403632775,
+      "learning_rate": 0.00016774583685298468,
+      "loss": 0.2778,
+      "step": 1803
+    },
+    {
+      "epoch": 1.0112107623318385,
+      "grad_norm": 0.0794674500124423,
+      "learning_rate": 0.00016769784002134008,
+      "loss": 0.2735,
+      "step": 1804
+    },
+    {
+      "epoch": 1.0117713004484306,
+      "grad_norm": 0.07574994258448081,
+      "learning_rate": 0.00016764981438245982,
+      "loss": 0.2652,
+      "step": 1805
+    },
+    {
+      "epoch": 1.0123318385650224,
+      "grad_norm": 0.07626438458999782,
+      "learning_rate": 0.00016760175995678007,
+      "loss": 0.2665,
+      "step": 1806
+    },
+    {
+      "epoch": 1.0128923766816142,
+      "grad_norm": 0.07928434348537178,
+      "learning_rate": 0.00016755367676474925,
+      "loss": 0.295,
+      "step": 1807
+    },
+    {
+      "epoch": 1.0134529147982063,
+      "grad_norm": 0.07586826519871802,
+      "learning_rate": 0.00016750556482682805,
+      "loss": 0.2882,
+      "step": 1808
+    },
+    {
+      "epoch": 1.0140134529147982,
+      "grad_norm": 0.07201222606650312,
+      "learning_rate": 0.0001674574241634894,
+      "loss": 0.2734,
+      "step": 1809
+    },
+    {
+      "epoch": 1.0145739910313902,
+      "grad_norm": 0.07214175221268386,
+      "learning_rate": 0.00016740925479521846,
+      "loss": 0.2801,
+      "step": 1810
+    },
+    {
+      "epoch": 1.015134529147982,
+      "grad_norm": 0.0705783431084273,
+      "learning_rate": 0.00016736105674251253,
+      "loss": 0.2704,
+      "step": 1811
+    },
+    {
+      "epoch": 1.015695067264574,
+      "grad_norm": 0.07063542893083063,
+      "learning_rate": 0.0001673128300258812,
+      "loss": 0.2742,
+      "step": 1812
+    },
+    {
+      "epoch": 1.016255605381166,
+      "grad_norm": 0.06794217535359018,
+      "learning_rate": 0.00016726457466584616,
+      "loss": 0.2531,
+      "step": 1813
+    },
+    {
+      "epoch": 1.0168161434977578,
+      "grad_norm": 0.0692309014154081,
+      "learning_rate": 0.00016721629068294143,
+      "loss": 0.2715,
+      "step": 1814
+    },
+    {
+      "epoch": 1.0173766816143497,
+      "grad_norm": 0.07322768519959756,
+      "learning_rate": 0.00016716797809771309,
+      "loss": 0.2504,
+      "step": 1815
+    },
+    {
+      "epoch": 1.0179372197309418,
+      "grad_norm": 0.07434577440239028,
+      "learning_rate": 0.00016711963693071943,
+      "loss": 0.2832,
+      "step": 1816
+    },
+    {
+      "epoch": 1.0184977578475336,
+      "grad_norm": 0.07442519185698841,
+      "learning_rate": 0.00016707126720253096,
+      "loss": 0.2627,
+      "step": 1817
+    },
+    {
+      "epoch": 1.0190582959641254,
+      "grad_norm": 0.07349019336629117,
+      "learning_rate": 0.00016702286893373021,
+      "loss": 0.2696,
+      "step": 1818
+    },
+    {
+      "epoch": 1.0196188340807175,
+      "grad_norm": 0.07431849333936506,
+      "learning_rate": 0.000166974442144912,
+      "loss": 0.2685,
+      "step": 1819
+    },
+    {
+      "epoch": 1.0201793721973094,
+      "grad_norm": 0.07372154678436714,
+      "learning_rate": 0.00016692598685668318,
+      "loss": 0.2634,
+      "step": 1820
+    },
+    {
+      "epoch": 1.0207399103139014,
+      "grad_norm": 0.07412288889160065,
+      "learning_rate": 0.00016687750308966277,
+      "loss": 0.2812,
+      "step": 1821
+    },
+    {
+      "epoch": 1.0213004484304933,
+      "grad_norm": 0.07447017805279436,
+      "learning_rate": 0.0001668289908644819,
+      "loss": 0.2831,
+      "step": 1822
+    },
+    {
+      "epoch": 1.0218609865470851,
+      "grad_norm": 0.07509167912997043,
+      "learning_rate": 0.00016678045020178386,
+      "loss": 0.2756,
+      "step": 1823
+    },
+    {
+      "epoch": 1.0224215246636772,
+      "grad_norm": 0.07329819074630305,
+      "learning_rate": 0.00016673188112222394,
+      "loss": 0.2805,
+      "step": 1824
+    },
+    {
+      "epoch": 1.022982062780269,
+      "grad_norm": 0.07110181482666514,
+      "learning_rate": 0.00016668328364646964,
+      "loss": 0.2615,
+      "step": 1825
+    },
+    {
+      "epoch": 1.0235426008968609,
+      "grad_norm": 0.07097793223431766,
+      "learning_rate": 0.0001666346577952004,
+      "loss": 0.2625,
+      "step": 1826
+    },
+    {
+      "epoch": 1.024103139013453,
+      "grad_norm": 0.07347791538232488,
+      "learning_rate": 0.0001665860035891079,
+      "loss": 0.2671,
+      "step": 1827
+    },
+    {
+      "epoch": 1.0246636771300448,
+      "grad_norm": 0.07439211790067399,
+      "learning_rate": 0.00016653732104889572,
+      "loss": 0.285,
+      "step": 1828
+    },
+    {
+      "epoch": 1.0252242152466369,
+      "grad_norm": 0.0715778928034537,
+      "learning_rate": 0.00016648861019527965,
+      "loss": 0.2617,
+      "step": 1829
+    },
+    {
+      "epoch": 1.0257847533632287,
+      "grad_norm": 0.07373282835850924,
+      "learning_rate": 0.0001664398710489874,
+      "loss": 0.2626,
+      "step": 1830
+    },
+    {
+      "epoch": 1.0263452914798206,
+      "grad_norm": 0.07377690998400391,
+      "learning_rate": 0.00016639110363075884,
+      "loss": 0.2733,
+      "step": 1831
+    },
+    {
+      "epoch": 1.0269058295964126,
+      "grad_norm": 0.07729672713511841,
+      "learning_rate": 0.00016634230796134576,
+      "loss": 0.2838,
+      "step": 1832
+    },
+    {
+      "epoch": 1.0274663677130045,
+      "grad_norm": 0.07282556514626554,
+      "learning_rate": 0.000166293484061512,
+      "loss": 0.2799,
+      "step": 1833
+    },
+    {
+      "epoch": 1.0280269058295963,
+      "grad_norm": 0.07403167397836082,
+      "learning_rate": 0.00016624463195203347,
+      "loss": 0.2679,
+      "step": 1834
+    },
+    {
+      "epoch": 1.0285874439461884,
+      "grad_norm": 0.07215275103758523,
+      "learning_rate": 0.00016619575165369805,
+      "loss": 0.2581,
+      "step": 1835
+    },
+    {
+      "epoch": 1.0291479820627802,
+      "grad_norm": 0.07344180895601204,
+      "learning_rate": 0.0001661468431873056,
+      "loss": 0.2795,
+      "step": 1836
+    },
+    {
+      "epoch": 1.029708520179372,
+      "grad_norm": 0.07432497378747537,
+      "learning_rate": 0.00016609790657366798,
+      "loss": 0.2837,
+      "step": 1837
+    },
+    {
+      "epoch": 1.0302690582959642,
+      "grad_norm": 0.07441357225587134,
+      "learning_rate": 0.000166048941833609,
+      "loss": 0.2717,
+      "step": 1838
+    },
+    {
+      "epoch": 1.030829596412556,
+      "grad_norm": 0.07590935476605204,
+      "learning_rate": 0.00016599994898796444,
+      "loss": 0.2851,
+      "step": 1839
+    },
+    {
+      "epoch": 1.031390134529148,
+      "grad_norm": 0.07354179140119582,
+      "learning_rate": 0.0001659509280575821,
+      "loss": 0.2688,
+      "step": 1840
+    },
+    {
+      "epoch": 1.03195067264574,
+      "grad_norm": 0.07439261859279374,
+      "learning_rate": 0.00016590187906332176,
+      "loss": 0.2727,
+      "step": 1841
+    },
+    {
+      "epoch": 1.0325112107623318,
+      "grad_norm": 0.07342703110129624,
+      "learning_rate": 0.00016585280202605497,
+      "loss": 0.2737,
+      "step": 1842
+    },
+    {
+      "epoch": 1.0330717488789238,
+      "grad_norm": 0.07525213264906828,
+      "learning_rate": 0.00016580369696666533,
+      "loss": 0.2732,
+      "step": 1843
+    },
+    {
+      "epoch": 1.0336322869955157,
+      "grad_norm": 0.07665551497568929,
+      "learning_rate": 0.0001657545639060484,
+      "loss": 0.2717,
+      "step": 1844
+    },
+    {
+      "epoch": 1.0341928251121075,
+      "grad_norm": 0.07479429520952934,
+      "learning_rate": 0.0001657054028651116,
+      "loss": 0.2826,
+      "step": 1845
+    },
+    {
+      "epoch": 1.0347533632286996,
+      "grad_norm": 0.07161623450510525,
+      "learning_rate": 0.00016565621386477423,
+      "loss": 0.2676,
+      "step": 1846
+    },
+    {
+      "epoch": 1.0353139013452914,
+      "grad_norm": 0.07600791631557516,
+      "learning_rate": 0.0001656069969259675,
+      "loss": 0.2631,
+      "step": 1847
+    },
+    {
+      "epoch": 1.0358744394618835,
+      "grad_norm": 0.07376485296076876,
+      "learning_rate": 0.0001655577520696346,
+      "loss": 0.2741,
+      "step": 1848
+    },
+    {
+      "epoch": 1.0364349775784754,
+      "grad_norm": 0.0728118805040438,
+      "learning_rate": 0.0001655084793167305,
+      "loss": 0.2683,
+      "step": 1849
+    },
+    {
+      "epoch": 1.0369955156950672,
+      "grad_norm": 0.07729576875042328,
+      "learning_rate": 0.00016545917868822203,
+      "loss": 0.2757,
+      "step": 1850
+    },
+    {
+      "epoch": 1.0375560538116593,
+      "grad_norm": 0.07378878336261326,
+      "learning_rate": 0.000165409850205088,
+      "loss": 0.2717,
+      "step": 1851
+    },
+    {
+      "epoch": 1.0381165919282511,
+      "grad_norm": 0.07242644750935845,
+      "learning_rate": 0.00016536049388831894,
+      "loss": 0.2782,
+      "step": 1852
+    },
+    {
+      "epoch": 1.038677130044843,
+      "grad_norm": 0.07858958551614088,
+      "learning_rate": 0.00016531110975891728,
+      "loss": 0.2698,
+      "step": 1853
+    },
+    {
+      "epoch": 1.039237668161435,
+      "grad_norm": 0.07400067409429323,
+      "learning_rate": 0.00016526169783789732,
+      "loss": 0.2685,
+      "step": 1854
+    },
+    {
+      "epoch": 1.0397982062780269,
+      "grad_norm": 0.07421871003510273,
+      "learning_rate": 0.00016521225814628506,
+      "loss": 0.2791,
+      "step": 1855
+    },
+    {
+      "epoch": 1.0403587443946187,
+      "grad_norm": 0.06969685301603386,
+      "learning_rate": 0.00016516279070511854,
+      "loss": 0.2709,
+      "step": 1856
+    },
+    {
+      "epoch": 1.0409192825112108,
+      "grad_norm": 0.0702092212314942,
+      "learning_rate": 0.0001651132955354474,
+      "loss": 0.2663,
+      "step": 1857
+    },
+    {
+      "epoch": 1.0414798206278026,
+      "grad_norm": 0.07554527085023427,
+      "learning_rate": 0.00016506377265833314,
+      "loss": 0.2837,
+      "step": 1858
+    },
+    {
+      "epoch": 1.0420403587443947,
+      "grad_norm": 0.07409275626702215,
+      "learning_rate": 0.00016501422209484908,
+      "loss": 0.2735,
+      "step": 1859
+    },
+    {
+      "epoch": 1.0426008968609866,
+      "grad_norm": 0.07507628014574805,
+      "learning_rate": 0.0001649646438660803,
+      "loss": 0.2788,
+      "step": 1860
+    },
+    {
+      "epoch": 1.0431614349775784,
+      "grad_norm": 0.07450700255195067,
+      "learning_rate": 0.0001649150379931237,
+      "loss": 0.2868,
+      "step": 1861
+    },
+    {
+      "epoch": 1.0437219730941705,
+      "grad_norm": 0.07140069146454657,
+      "learning_rate": 0.00016486540449708783,
+      "loss": 0.2744,
+      "step": 1862
+    },
+    {
+      "epoch": 1.0442825112107623,
+      "grad_norm": 0.07658793289262474,
+      "learning_rate": 0.0001648157433990931,
+      "loss": 0.2829,
+      "step": 1863
+    },
+    {
+      "epoch": 1.0448430493273542,
+      "grad_norm": 0.0766190372006443,
+      "learning_rate": 0.00016476605472027172,
+      "loss": 0.2869,
+      "step": 1864
+    },
+    {
+      "epoch": 1.0454035874439462,
+      "grad_norm": 0.07550648904207007,
+      "learning_rate": 0.00016471633848176738,
+      "loss": 0.2711,
+      "step": 1865
+    },
+    {
+      "epoch": 1.045964125560538,
+      "grad_norm": 0.07764537487796822,
+      "learning_rate": 0.00016466659470473579,
+      "loss": 0.295,
+      "step": 1866
+    },
+    {
+      "epoch": 1.0465246636771302,
+      "grad_norm": 0.0744227562253243,
+      "learning_rate": 0.0001646168234103442,
+      "loss": 0.2734,
+      "step": 1867
+    },
+    {
+      "epoch": 1.047085201793722,
+      "grad_norm": 0.07383161532210666,
+      "learning_rate": 0.0001645670246197716,
+      "loss": 0.2767,
+      "step": 1868
+    },
+    {
+      "epoch": 1.0476457399103138,
+      "grad_norm": 0.07862854408338234,
+      "learning_rate": 0.00016451719835420877,
+      "loss": 0.2921,
+      "step": 1869
+    },
+    {
+      "epoch": 1.048206278026906,
+      "grad_norm": 0.07240748787709966,
+      "learning_rate": 0.0001644673446348581,
+      "loss": 0.2779,
+      "step": 1870
+    },
+    {
+      "epoch": 1.0487668161434978,
+      "grad_norm": 0.0728487907630601,
+      "learning_rate": 0.00016441746348293363,
+      "loss": 0.2881,
+      "step": 1871
+    },
+    {
+      "epoch": 1.0493273542600896,
+      "grad_norm": 0.0696023534040059,
+      "learning_rate": 0.00016436755491966115,
+      "loss": 0.2517,
+      "step": 1872
+    },
+    {
+      "epoch": 1.0498878923766817,
+      "grad_norm": 0.07345203238219637,
+      "learning_rate": 0.00016431761896627806,
+      "loss": 0.2612,
+      "step": 1873
+    },
+    {
+      "epoch": 1.0504484304932735,
+      "grad_norm": 0.07513713484721089,
+      "learning_rate": 0.0001642676556440335,
+      "loss": 0.2793,
+      "step": 1874
+    },
+    {
+      "epoch": 1.0510089686098654,
+      "grad_norm": 0.07452318706610823,
+      "learning_rate": 0.00016421766497418816,
+      "loss": 0.2749,
+      "step": 1875
+    },
+    {
+      "epoch": 1.0515695067264574,
+      "grad_norm": 0.074578312418863,
+      "learning_rate": 0.00016416764697801438,
+      "loss": 0.2841,
+      "step": 1876
+    },
+    {
+      "epoch": 1.0521300448430493,
+      "grad_norm": 0.0732401610664788,
+      "learning_rate": 0.00016411760167679617,
+      "loss": 0.2664,
+      "step": 1877
+    },
+    {
+      "epoch": 1.0526905829596414,
+      "grad_norm": 0.07456699601331833,
+      "learning_rate": 0.00016406752909182916,
+      "loss": 0.2637,
+      "step": 1878
+    },
+    {
+      "epoch": 1.0532511210762332,
+      "grad_norm": 0.07671286998164872,
+      "learning_rate": 0.00016401742924442055,
+      "loss": 0.2855,
+      "step": 1879
+    },
+    {
+      "epoch": 1.053811659192825,
+      "grad_norm": 0.07677523499569627,
+      "learning_rate": 0.00016396730215588915,
+      "loss": 0.2869,
+      "step": 1880
+    },
+    {
+      "epoch": 1.0543721973094171,
+      "grad_norm": 0.07277576355711973,
+      "learning_rate": 0.00016391714784756538,
+      "loss": 0.2755,
+      "step": 1881
+    },
+    {
+      "epoch": 1.054932735426009,
+      "grad_norm": 0.07247938594655151,
+      "learning_rate": 0.00016386696634079125,
+      "loss": 0.2726,
+      "step": 1882
+    },
+    {
+      "epoch": 1.0554932735426008,
+      "grad_norm": 0.07569041734695106,
+      "learning_rate": 0.00016381675765692028,
+      "loss": 0.2686,
+      "step": 1883
+    },
+    {
+      "epoch": 1.0560538116591929,
+      "grad_norm": 0.07799774429884256,
+      "learning_rate": 0.00016376652181731769,
+      "loss": 0.2773,
+      "step": 1884
+    },
+    {
+      "epoch": 1.0566143497757847,
+      "grad_norm": 0.07321293018316793,
+      "learning_rate": 0.0001637162588433601,
+      "loss": 0.2702,
+      "step": 1885
+    },
+    {
+      "epoch": 1.0571748878923768,
+      "grad_norm": 0.07436428152382965,
+      "learning_rate": 0.00016366596875643576,
+      "loss": 0.2764,
+      "step": 1886
+    },
+    {
+      "epoch": 1.0577354260089686,
+      "grad_norm": 0.0761426898434651,
+      "learning_rate": 0.00016361565157794447,
+      "loss": 0.277,
+      "step": 1887
+    },
+    {
+      "epoch": 1.0582959641255605,
+      "grad_norm": 0.07208787257540157,
+      "learning_rate": 0.0001635653073292975,
+      "loss": 0.2701,
+      "step": 1888
+    },
+    {
+      "epoch": 1.0588565022421526,
+      "grad_norm": 0.07687084904278584,
+      "learning_rate": 0.00016351493603191766,
+      "loss": 0.2796,
+      "step": 1889
+    },
+    {
+      "epoch": 1.0594170403587444,
+      "grad_norm": 0.07593381504905147,
+      "learning_rate": 0.0001634645377072393,
+      "loss": 0.2749,
+      "step": 1890
+    },
+    {
+      "epoch": 1.0599775784753362,
+      "grad_norm": 0.0727610271335173,
+      "learning_rate": 0.00016341411237670827,
+      "loss": 0.2664,
+      "step": 1891
+    },
+    {
+      "epoch": 1.0605381165919283,
+      "grad_norm": 0.0753329496258447,
+      "learning_rate": 0.00016336366006178187,
+      "loss": 0.2714,
+      "step": 1892
+    },
+    {
+      "epoch": 1.0610986547085202,
+      "grad_norm": 0.07624308074621539,
+      "learning_rate": 0.0001633131807839289,
+      "loss": 0.2917,
+      "step": 1893
+    },
+    {
+      "epoch": 1.061659192825112,
+      "grad_norm": 0.07363716680200075,
+      "learning_rate": 0.00016326267456462964,
+      "loss": 0.2736,
+      "step": 1894
+    },
+    {
+      "epoch": 1.062219730941704,
+      "grad_norm": 0.07894494493318059,
+      "learning_rate": 0.00016321214142537584,
+      "loss": 0.2745,
+      "step": 1895
+    },
+    {
+      "epoch": 1.062780269058296,
+      "grad_norm": 0.07539024803923519,
+      "learning_rate": 0.0001631615813876707,
+      "loss": 0.2883,
+      "step": 1896
+    },
+    {
+      "epoch": 1.063340807174888,
+      "grad_norm": 0.07434098765575238,
+      "learning_rate": 0.00016311099447302886,
+      "loss": 0.279,
+      "step": 1897
+    },
+    {
+      "epoch": 1.0639013452914798,
+      "grad_norm": 0.07412931566846713,
+      "learning_rate": 0.00016306038070297641,
+      "loss": 0.2646,
+      "step": 1898
+    },
+    {
+      "epoch": 1.0644618834080717,
+      "grad_norm": 0.0740234768281963,
+      "learning_rate": 0.00016300974009905085,
+      "loss": 0.2676,
+      "step": 1899
+    },
+    {
+      "epoch": 1.0650224215246638,
+      "grad_norm": 0.07557766774395729,
+      "learning_rate": 0.00016295907268280109,
+      "loss": 0.2683,
+      "step": 1900
+    },
+    {
+      "epoch": 1.0655829596412556,
+      "grad_norm": 0.07281936621870358,
+      "learning_rate": 0.0001629083784757875,
+      "loss": 0.262,
+      "step": 1901
+    },
+    {
+      "epoch": 1.0661434977578474,
+      "grad_norm": 0.07508960385501345,
+      "learning_rate": 0.0001628576574995818,
+      "loss": 0.2735,
+      "step": 1902
+    },
+    {
+      "epoch": 1.0667040358744395,
+      "grad_norm": 0.07507273932238448,
+      "learning_rate": 0.0001628069097757671,
+      "loss": 0.2723,
+      "step": 1903
+    },
+    {
+      "epoch": 1.0672645739910314,
+      "grad_norm": 0.07377330757731597,
+      "learning_rate": 0.0001627561353259379,
+      "loss": 0.2571,
+      "step": 1904
+    },
+    {
+      "epoch": 1.0678251121076232,
+      "grad_norm": 0.07402325427347224,
+      "learning_rate": 0.00016270533417170015,
+      "loss": 0.277,
+      "step": 1905
+    },
+    {
+      "epoch": 1.0683856502242153,
+      "grad_norm": 0.07323515158476367,
+      "learning_rate": 0.00016265450633467105,
+      "loss": 0.2719,
+      "step": 1906
+    },
+    {
+      "epoch": 1.0689461883408071,
+      "grad_norm": 0.07296175866704731,
+      "learning_rate": 0.0001626036518364792,
+      "loss": 0.2708,
+      "step": 1907
+    },
+    {
+      "epoch": 1.0695067264573992,
+      "grad_norm": 0.07374203094566148,
+      "learning_rate": 0.00016255277069876454,
+      "loss": 0.2691,
+      "step": 1908
+    },
+    {
+      "epoch": 1.070067264573991,
+      "grad_norm": 0.07203784919823066,
+      "learning_rate": 0.00016250186294317835,
+      "loss": 0.27,
+      "step": 1909
+    },
+    {
+      "epoch": 1.0706278026905829,
+      "grad_norm": 0.0720051439648434,
+      "learning_rate": 0.00016245092859138328,
+      "loss": 0.2664,
+      "step": 1910
+    },
+    {
+      "epoch": 1.071188340807175,
+      "grad_norm": 0.07352081307789217,
+      "learning_rate": 0.0001623999676650532,
+      "loss": 0.289,
+      "step": 1911
+    },
+    {
+      "epoch": 1.0717488789237668,
+      "grad_norm": 0.0734531363351306,
+      "learning_rate": 0.00016234898018587337,
+      "loss": 0.2719,
+      "step": 1912
+    },
+    {
+      "epoch": 1.0723094170403586,
+      "grad_norm": 0.0716115495065179,
+      "learning_rate": 0.00016229796617554028,
+      "loss": 0.2723,
+      "step": 1913
+    },
+    {
+      "epoch": 1.0728699551569507,
+      "grad_norm": 0.07629494921644377,
+      "learning_rate": 0.00016224692565576184,
+      "loss": 0.2792,
+      "step": 1914
+    },
+    {
+      "epoch": 1.0734304932735426,
+      "grad_norm": 0.0767134313243209,
+      "learning_rate": 0.00016219585864825706,
+      "loss": 0.2797,
+      "step": 1915
+    },
+    {
+      "epoch": 1.0739910313901346,
+      "grad_norm": 0.07529608505081263,
+      "learning_rate": 0.00016214476517475637,
+      "loss": 0.2766,
+      "step": 1916
+    },
+    {
+      "epoch": 1.0745515695067265,
+      "grad_norm": 0.07366744989513864,
+      "learning_rate": 0.00016209364525700138,
+      "loss": 0.2695,
+      "step": 1917
+    },
+    {
+      "epoch": 1.0751121076233183,
+      "grad_norm": 0.07597379970681774,
+      "learning_rate": 0.00016204249891674496,
+      "loss": 0.2603,
+      "step": 1918
+    },
+    {
+      "epoch": 1.0756726457399104,
+      "grad_norm": 0.07499576204333906,
+      "learning_rate": 0.0001619913261757513,
+      "loss": 0.2767,
+      "step": 1919
+    },
+    {
+      "epoch": 1.0762331838565022,
+      "grad_norm": 0.07629421939936472,
+      "learning_rate": 0.00016194012705579572,
+      "loss": 0.2857,
+      "step": 1920
+    },
+    {
+      "epoch": 1.076793721973094,
+      "grad_norm": 0.07384185136588617,
+      "learning_rate": 0.00016188890157866484,
+      "loss": 0.2859,
+      "step": 1921
+    },
+    {
+      "epoch": 1.0773542600896862,
+      "grad_norm": 0.07211707283690752,
+      "learning_rate": 0.0001618376497661564,
+      "loss": 0.2737,
+      "step": 1922
+    },
+    {
+      "epoch": 1.077914798206278,
+      "grad_norm": 0.07424252476059912,
+      "learning_rate": 0.00016178637164007947,
+      "loss": 0.2699,
+      "step": 1923
+    },
+    {
+      "epoch": 1.07847533632287,
+      "grad_norm": 0.07580353014079456,
+      "learning_rate": 0.00016173506722225428,
+      "loss": 0.2649,
+      "step": 1924
+    },
+    {
+      "epoch": 1.079035874439462,
+      "grad_norm": 0.07337276159961045,
+      "learning_rate": 0.00016168373653451218,
+      "loss": 0.279,
+      "step": 1925
+    },
+    {
+      "epoch": 1.0795964125560538,
+      "grad_norm": 0.07704345689952927,
+      "learning_rate": 0.00016163237959869578,
+      "loss": 0.2751,
+      "step": 1926
+    },
+    {
+      "epoch": 1.0801569506726458,
+      "grad_norm": 0.07165282706648543,
+      "learning_rate": 0.00016158099643665878,
+      "loss": 0.2726,
+      "step": 1927
+    },
+    {
+      "epoch": 1.0807174887892377,
+      "grad_norm": 0.07313133774074929,
+      "learning_rate": 0.00016152958707026614,
+      "loss": 0.2687,
+      "step": 1928
+    },
+    {
+      "epoch": 1.0812780269058295,
+      "grad_norm": 0.07463579158475091,
+      "learning_rate": 0.00016147815152139385,
+      "loss": 0.2736,
+      "step": 1929
+    },
+    {
+      "epoch": 1.0818385650224216,
+      "grad_norm": 0.07633363811644278,
+      "learning_rate": 0.00016142668981192917,
+      "loss": 0.2728,
+      "step": 1930
+    },
+    {
+      "epoch": 1.0823991031390134,
+      "grad_norm": 0.07215708677994055,
+      "learning_rate": 0.00016137520196377042,
+      "loss": 0.2689,
+      "step": 1931
+    },
+    {
+      "epoch": 1.0829596412556053,
+      "grad_norm": 0.0728669146380796,
+      "learning_rate": 0.00016132368799882704,
+      "loss": 0.2599,
+      "step": 1932
+    },
+    {
+      "epoch": 1.0835201793721974,
+      "grad_norm": 0.07403038728621991,
+      "learning_rate": 0.00016127214793901958,
+      "loss": 0.2776,
+      "step": 1933
+    },
+    {
+      "epoch": 1.0840807174887892,
+      "grad_norm": 0.078488015679292,
+      "learning_rate": 0.0001612205818062797,
+      "loss": 0.268,
+      "step": 1934
+    },
+    {
+      "epoch": 1.0846412556053813,
+      "grad_norm": 0.07270573932284943,
+      "learning_rate": 0.0001611689896225502,
+      "loss": 0.2606,
+      "step": 1935
+    },
+    {
+      "epoch": 1.0852017937219731,
+      "grad_norm": 0.07633324067651374,
+      "learning_rate": 0.00016111737140978494,
+      "loss": 0.2733,
+      "step": 1936
+    },
+    {
+      "epoch": 1.085762331838565,
+      "grad_norm": 0.07314871818107127,
+      "learning_rate": 0.0001610657271899488,
+      "loss": 0.2725,
+      "step": 1937
+    },
+    {
+      "epoch": 1.086322869955157,
+      "grad_norm": 0.07436002582849571,
+      "learning_rate": 0.00016101405698501782,
+      "loss": 0.2733,
+      "step": 1938
+    },
+    {
+      "epoch": 1.0868834080717489,
+      "grad_norm": 0.07258638981000642,
+      "learning_rate": 0.000160962360816979,
+      "loss": 0.2599,
+      "step": 1939
+    },
+    {
+      "epoch": 1.0874439461883407,
+      "grad_norm": 0.0779396234130977,
+      "learning_rate": 0.00016091063870783047,
+      "loss": 0.2828,
+      "step": 1940
+    },
+    {
+      "epoch": 1.0880044843049328,
+      "grad_norm": 0.07387292587429643,
+      "learning_rate": 0.00016085889067958136,
+      "loss": 0.2579,
+      "step": 1941
+    },
+    {
+      "epoch": 1.0885650224215246,
+      "grad_norm": 0.07477688016800593,
+      "learning_rate": 0.0001608071167542518,
+      "loss": 0.2773,
+      "step": 1942
+    },
+    {
+      "epoch": 1.0891255605381165,
+      "grad_norm": 0.0716554336645514,
+      "learning_rate": 0.00016075531695387303,
+      "loss": 0.2573,
+      "step": 1943
+    },
+    {
+      "epoch": 1.0896860986547086,
+      "grad_norm": 0.07642593341952433,
+      "learning_rate": 0.00016070349130048724,
+      "loss": 0.2862,
+      "step": 1944
+    },
+    {
+      "epoch": 1.0902466367713004,
+      "grad_norm": 0.07517365458647388,
+      "learning_rate": 0.00016065163981614764,
+      "loss": 0.2798,
+      "step": 1945
+    },
+    {
+      "epoch": 1.0908071748878925,
+      "grad_norm": 0.07694130433057056,
+      "learning_rate": 0.00016059976252291835,
+      "loss": 0.2821,
+      "step": 1946
+    },
+    {
+      "epoch": 1.0913677130044843,
+      "grad_norm": 0.07370821250050272,
+      "learning_rate": 0.0001605478594428746,
+      "loss": 0.2774,
+      "step": 1947
+    },
+    {
+      "epoch": 1.0919282511210762,
+      "grad_norm": 0.07303370270961763,
+      "learning_rate": 0.00016049593059810248,
+      "loss": 0.2793,
+      "step": 1948
+    },
+    {
+      "epoch": 1.0924887892376682,
+      "grad_norm": 0.07339460012365612,
+      "learning_rate": 0.00016044397601069918,
+      "loss": 0.2707,
+      "step": 1949
+    },
+    {
+      "epoch": 1.09304932735426,
+      "grad_norm": 0.08404334083821938,
+      "learning_rate": 0.0001603919957027727,
+      "loss": 0.2785,
+      "step": 1950
+    },
+    {
+      "epoch": 1.093609865470852,
+      "grad_norm": 0.0769906958486627,
+      "learning_rate": 0.00016033998969644205,
+      "loss": 0.285,
+      "step": 1951
+    },
+    {
+      "epoch": 1.094170403587444,
+      "grad_norm": 0.07644487362969105,
+      "learning_rate": 0.00016028795801383718,
+      "loss": 0.2911,
+      "step": 1952
+    },
+    {
+      "epoch": 1.0947309417040358,
+      "grad_norm": 0.0724238461624434,
+      "learning_rate": 0.00016023590067709898,
+      "loss": 0.2783,
+      "step": 1953
+    },
+    {
+      "epoch": 1.0952914798206277,
+      "grad_norm": 0.0740558866631432,
+      "learning_rate": 0.00016018381770837922,
+      "loss": 0.2789,
+      "step": 1954
+    },
+    {
+      "epoch": 1.0958520179372198,
+      "grad_norm": 0.07226000369479146,
+      "learning_rate": 0.00016013170912984058,
+      "loss": 0.2719,
+      "step": 1955
+    },
+    {
+      "epoch": 1.0964125560538116,
+      "grad_norm": 0.07753612392525226,
+      "learning_rate": 0.00016007957496365667,
+      "loss": 0.2772,
+      "step": 1956
+    },
+    {
+      "epoch": 1.0969730941704037,
+      "grad_norm": 0.07491630542893295,
+      "learning_rate": 0.00016002741523201195,
+      "loss": 0.2764,
+      "step": 1957
+    },
+    {
+      "epoch": 1.0975336322869955,
+      "grad_norm": 0.074447892607347,
+      "learning_rate": 0.00015997522995710178,
+      "loss": 0.2831,
+      "step": 1958
+    },
+    {
+      "epoch": 1.0980941704035874,
+      "grad_norm": 0.07609596858643143,
+      "learning_rate": 0.00015992301916113242,
+      "loss": 0.2852,
+      "step": 1959
+    },
+    {
+      "epoch": 1.0986547085201794,
+      "grad_norm": 0.07332301350280312,
+      "learning_rate": 0.0001598707828663209,
+      "loss": 0.2705,
+      "step": 1960
+    },
+    {
+      "epoch": 1.0992152466367713,
+      "grad_norm": 0.07545889966915205,
+      "learning_rate": 0.00015981852109489517,
+      "loss": 0.2536,
+      "step": 1961
+    },
+    {
+      "epoch": 1.0997757847533631,
+      "grad_norm": 0.07421315465410608,
+      "learning_rate": 0.000159766233869094,
+      "loss": 0.277,
+      "step": 1962
+    },
+    {
+      "epoch": 1.1003363228699552,
+      "grad_norm": 0.0746229326858004,
+      "learning_rate": 0.00015971392121116705,
+      "loss": 0.2673,
+      "step": 1963
+    },
+    {
+      "epoch": 1.100896860986547,
+      "grad_norm": 0.07803795188828119,
+      "learning_rate": 0.00015966158314337472,
+      "loss": 0.2787,
+      "step": 1964
+    },
+    {
+      "epoch": 1.1014573991031391,
+      "grad_norm": 0.07507220910349977,
+      "learning_rate": 0.00015960921968798824,
+      "loss": 0.2706,
+      "step": 1965
+    },
+    {
+      "epoch": 1.102017937219731,
+      "grad_norm": 0.07245626260377289,
+      "learning_rate": 0.00015955683086728962,
+      "loss": 0.2726,
+      "step": 1966
+    },
+    {
+      "epoch": 1.1025784753363228,
+      "grad_norm": 0.07147169574943817,
+      "learning_rate": 0.0001595044167035718,
+      "loss": 0.265,
+      "step": 1967
+    },
+    {
+      "epoch": 1.1031390134529149,
+      "grad_norm": 0.07721066159888125,
+      "learning_rate": 0.00015945197721913833,
+      "loss": 0.2751,
+      "step": 1968
+    },
+    {
+      "epoch": 1.1036995515695067,
+      "grad_norm": 0.07437528064108397,
+      "learning_rate": 0.00015939951243630363,
+      "loss": 0.2745,
+      "step": 1969
+    },
+    {
+      "epoch": 1.1042600896860986,
+      "grad_norm": 0.07158845788098418,
+      "learning_rate": 0.00015934702237739288,
+      "loss": 0.2622,
+      "step": 1970
+    },
+    {
+      "epoch": 1.1048206278026906,
+      "grad_norm": 0.07339457836335966,
+      "learning_rate": 0.00015929450706474198,
+      "loss": 0.2773,
+      "step": 1971
+    },
+    {
+      "epoch": 1.1053811659192825,
+      "grad_norm": 0.07430073797855495,
+      "learning_rate": 0.00015924196652069758,
+      "loss": 0.2654,
+      "step": 1972
+    },
+    {
+      "epoch": 1.1059417040358746,
+      "grad_norm": 0.07093762722174926,
+      "learning_rate": 0.0001591894007676171,
+      "loss": 0.2625,
+      "step": 1973
+    },
+    {
+      "epoch": 1.1065022421524664,
+      "grad_norm": 0.07460068410984273,
+      "learning_rate": 0.00015913680982786868,
+      "loss": 0.2739,
+      "step": 1974
+    },
+    {
+      "epoch": 1.1070627802690582,
+      "grad_norm": 0.073403458352427,
+      "learning_rate": 0.00015908419372383112,
+      "loss": 0.2671,
+      "step": 1975
+    },
+    {
+      "epoch": 1.1076233183856503,
+      "grad_norm": 0.07492098467625789,
+      "learning_rate": 0.00015903155247789404,
+      "loss": 0.2703,
+      "step": 1976
+    },
+    {
+      "epoch": 1.1081838565022422,
+      "grad_norm": 0.07128177484776693,
+      "learning_rate": 0.00015897888611245766,
+      "loss": 0.2754,
+      "step": 1977
+    },
+    {
+      "epoch": 1.108744394618834,
+      "grad_norm": 0.07383092451473902,
+      "learning_rate": 0.00015892619464993293,
+      "loss": 0.2706,
+      "step": 1978
+    },
+    {
+      "epoch": 1.109304932735426,
+      "grad_norm": 0.07263557664368579,
+      "learning_rate": 0.00015887347811274145,
+      "loss": 0.2668,
+      "step": 1979
+    },
+    {
+      "epoch": 1.109865470852018,
+      "grad_norm": 0.07115233393958298,
+      "learning_rate": 0.00015882073652331556,
+      "loss": 0.2734,
+      "step": 1980
+    },
+    {
+      "epoch": 1.1104260089686098,
+      "grad_norm": 0.07555708936788974,
+      "learning_rate": 0.00015876796990409815,
+      "loss": 0.2795,
+      "step": 1981
+    },
+    {
+      "epoch": 1.1109865470852018,
+      "grad_norm": 0.0753979028353171,
+      "learning_rate": 0.00015871517827754285,
+      "loss": 0.2809,
+      "step": 1982
+    },
+    {
+      "epoch": 1.1115470852017937,
+      "grad_norm": 0.0770002797947777,
+      "learning_rate": 0.00015866236166611395,
+      "loss": 0.2765,
+      "step": 1983
+    },
+    {
+      "epoch": 1.1121076233183858,
+      "grad_norm": 0.07575667063290138,
+      "learning_rate": 0.00015860952009228625,
+      "loss": 0.2779,
+      "step": 1984
+    },
+    {
+      "epoch": 1.1126681614349776,
+      "grad_norm": 0.07787975694349888,
+      "learning_rate": 0.0001585566535785453,
+      "loss": 0.2719,
+      "step": 1985
+    },
+    {
+      "epoch": 1.1132286995515694,
+      "grad_norm": 0.0740094391058316,
+      "learning_rate": 0.0001585037621473872,
+      "loss": 0.2615,
+      "step": 1986
+    },
+    {
+      "epoch": 1.1137892376681615,
+      "grad_norm": 0.07279280444286443,
+      "learning_rate": 0.00015845084582131867,
+      "loss": 0.2739,
+      "step": 1987
+    },
+    {
+      "epoch": 1.1143497757847534,
+      "grad_norm": 0.07483962761701597,
+      "learning_rate": 0.00015839790462285696,
+      "loss": 0.2704,
+      "step": 1988
+    },
+    {
+      "epoch": 1.1149103139013452,
+      "grad_norm": 0.0730593124018396,
+      "learning_rate": 0.00015834493857453007,
+      "loss": 0.2726,
+      "step": 1989
+    },
+    {
+      "epoch": 1.1154708520179373,
+      "grad_norm": 0.07462548117811857,
+      "learning_rate": 0.00015829194769887634,
+      "loss": 0.2759,
+      "step": 1990
+    },
+    {
+      "epoch": 1.1160313901345291,
+      "grad_norm": 0.07509532867389848,
+      "learning_rate": 0.0001582389320184449,
+      "loss": 0.2692,
+      "step": 1991
+    },
+    {
+      "epoch": 1.116591928251121,
+      "grad_norm": 0.07305637870162947,
+      "learning_rate": 0.0001581858915557953,
+      "loss": 0.2737,
+      "step": 1992
+    },
+    {
+      "epoch": 1.117152466367713,
+      "grad_norm": 0.07287159088842872,
+      "learning_rate": 0.00015813282633349765,
+      "loss": 0.2709,
+      "step": 1993
+    },
+    {
+      "epoch": 1.1177130044843049,
+      "grad_norm": 0.07435947185943617,
+      "learning_rate": 0.00015807973637413264,
+      "loss": 0.2737,
+      "step": 1994
+    },
+    {
+      "epoch": 1.118273542600897,
+      "grad_norm": 0.07278272531265187,
+      "learning_rate": 0.00015802662170029148,
+      "loss": 0.2719,
+      "step": 1995
+    },
+    {
+      "epoch": 1.1188340807174888,
+      "grad_norm": 0.07382141303330053,
+      "learning_rate": 0.00015797348233457584,
+      "loss": 0.2772,
+      "step": 1996
+    },
+    {
+      "epoch": 1.1193946188340806,
+      "grad_norm": 0.07397075999132298,
+      "learning_rate": 0.000157920318299598,
+      "loss": 0.2607,
+      "step": 1997
+    },
+    {
+      "epoch": 1.1199551569506727,
+      "grad_norm": 0.07183826883376702,
+      "learning_rate": 0.0001578671296179806,
+      "loss": 0.2813,
+      "step": 1998
+    },
+    {
+      "epoch": 1.1205156950672646,
+      "grad_norm": 0.07436399136538775,
+      "learning_rate": 0.00015781391631235686,
+      "loss": 0.2734,
+      "step": 1999
+    },
+    {
+      "epoch": 1.1210762331838564,
+      "grad_norm": 0.07192922286500203,
+      "learning_rate": 0.0001577606784053705,
+      "loss": 0.2568,
+      "step": 2000
+    },
+    {
+      "epoch": 1.1216367713004485,
+      "grad_norm": 0.07617022498805927,
+      "learning_rate": 0.00015770741591967559,
+      "loss": 0.2762,
+      "step": 2001
+    },
+    {
+      "epoch": 1.1221973094170403,
+      "grad_norm": 0.07651159820540177,
+      "learning_rate": 0.00015765412887793682,
+      "loss": 0.2798,
+      "step": 2002
+    },
+    {
+      "epoch": 1.1227578475336322,
+      "grad_norm": 0.07712755698411447,
+      "learning_rate": 0.00015760081730282924,
+      "loss": 0.2749,
+      "step": 2003
+    },
+    {
+      "epoch": 1.1233183856502242,
+      "grad_norm": 0.0777130799899898,
+      "learning_rate": 0.0001575474812170383,
+      "loss": 0.2744,
+      "step": 2004
+    },
+    {
+      "epoch": 1.123878923766816,
+      "grad_norm": 0.07297566247431461,
+      "learning_rate": 0.00015749412064325994,
+      "loss": 0.2614,
+      "step": 2005
+    },
+    {
+      "epoch": 1.1244394618834082,
+      "grad_norm": 0.07287115726268938,
+      "learning_rate": 0.00015744073560420053,
+      "loss": 0.273,
+      "step": 2006
+    },
+    {
+      "epoch": 1.125,
+      "grad_norm": 0.07543796690093375,
+      "learning_rate": 0.0001573873261225768,
+      "loss": 0.2744,
+      "step": 2007
+    },
+    {
+      "epoch": 1.1255605381165918,
+      "grad_norm": 0.07125092095866423,
+      "learning_rate": 0.00015733389222111592,
+      "loss": 0.2858,
+      "step": 2008
+    },
+    {
+      "epoch": 1.126121076233184,
+      "grad_norm": 0.07115468366189806,
+      "learning_rate": 0.00015728043392255545,
+      "loss": 0.2716,
+      "step": 2009
+    },
+    {
+      "epoch": 1.1266816143497758,
+      "grad_norm": 0.07453062732128388,
+      "learning_rate": 0.0001572269512496433,
+      "loss": 0.2727,
+      "step": 2010
+    },
+    {
+      "epoch": 1.1272421524663678,
+      "grad_norm": 0.07513388744332164,
+      "learning_rate": 0.0001571734442251378,
+      "loss": 0.2909,
+      "step": 2011
+    },
+    {
+      "epoch": 1.1278026905829597,
+      "grad_norm": 0.07494383499721882,
+      "learning_rate": 0.0001571199128718076,
+      "loss": 0.2781,
+      "step": 2012
+    },
+    {
+      "epoch": 1.1283632286995515,
+      "grad_norm": 0.07504409999077581,
+      "learning_rate": 0.00015706635721243173,
+      "loss": 0.289,
+      "step": 2013
+    },
+    {
+      "epoch": 1.1289237668161436,
+      "grad_norm": 0.07166225475336385,
+      "learning_rate": 0.00015701277726979952,
+      "loss": 0.2682,
+      "step": 2014
+    },
+    {
+      "epoch": 1.1294843049327354,
+      "grad_norm": 0.0742582638600298,
+      "learning_rate": 0.00015695917306671067,
+      "loss": 0.2607,
+      "step": 2015
+    },
+    {
+      "epoch": 1.1300448430493273,
+      "grad_norm": 0.0747725454096467,
+      "learning_rate": 0.00015690554462597522,
+      "loss": 0.2729,
+      "step": 2016
+    },
+    {
+      "epoch": 1.1306053811659194,
+      "grad_norm": 0.07511852957362249,
+      "learning_rate": 0.0001568518919704135,
+      "loss": 0.2818,
+      "step": 2017
+    },
+    {
+      "epoch": 1.1311659192825112,
+      "grad_norm": 0.07314543195251337,
+      "learning_rate": 0.00015679821512285615,
+      "loss": 0.2709,
+      "step": 2018
+    },
+    {
+      "epoch": 1.131726457399103,
+      "grad_norm": 0.07433168459280702,
+      "learning_rate": 0.00015674451410614405,
+      "loss": 0.2771,
+      "step": 2019
+    },
+    {
+      "epoch": 1.1322869955156951,
+      "grad_norm": 0.07200847633930003,
+      "learning_rate": 0.00015669078894312848,
+      "loss": 0.2532,
+      "step": 2020
+    },
+    {
+      "epoch": 1.132847533632287,
+      "grad_norm": 0.07400974519073239,
+      "learning_rate": 0.0001566370396566709,
+      "loss": 0.277,
+      "step": 2021
+    },
+    {
+      "epoch": 1.133408071748879,
+      "grad_norm": 0.07450630557819417,
+      "learning_rate": 0.00015658326626964301,
+      "loss": 0.278,
+      "step": 2022
+    },
+    {
+      "epoch": 1.1339686098654709,
+      "grad_norm": 0.07698578648706768,
+      "learning_rate": 0.00015652946880492693,
+      "loss": 0.2714,
+      "step": 2023
+    },
+    {
+      "epoch": 1.1345291479820627,
+      "grad_norm": 0.07482987816471644,
+      "learning_rate": 0.00015647564728541485,
+      "loss": 0.2766,
+      "step": 2024
+    },
+    {
+      "epoch": 1.1350896860986548,
+      "grad_norm": 0.07267521546994753,
+      "learning_rate": 0.0001564218017340093,
+      "loss": 0.274,
+      "step": 2025
+    },
+    {
+      "epoch": 1.1356502242152466,
+      "grad_norm": 0.07414428261073383,
+      "learning_rate": 0.00015636793217362288,
+      "loss": 0.2738,
+      "step": 2026
+    },
+    {
+      "epoch": 1.1362107623318385,
+      "grad_norm": 0.07282687696485579,
+      "learning_rate": 0.0001563140386271787,
+      "loss": 0.2713,
+      "step": 2027
+    },
+    {
+      "epoch": 1.1367713004484306,
+      "grad_norm": 0.07659535033965022,
+      "learning_rate": 0.00015626012111760975,
+      "loss": 0.2754,
+      "step": 2028
+    },
+    {
+      "epoch": 1.1373318385650224,
+      "grad_norm": 0.07396044636197424,
+      "learning_rate": 0.00015620617966785946,
+      "loss": 0.262,
+      "step": 2029
+    },
+    {
+      "epoch": 1.1378923766816142,
+      "grad_norm": 0.07397253747286966,
+      "learning_rate": 0.00015615221430088133,
+      "loss": 0.2769,
+      "step": 2030
+    },
+    {
+      "epoch": 1.1384529147982063,
+      "grad_norm": 0.07202784009774799,
+      "learning_rate": 0.00015609822503963907,
+      "loss": 0.2698,
+      "step": 2031
+    },
+    {
+      "epoch": 1.1390134529147982,
+      "grad_norm": 0.07085916746893994,
+      "learning_rate": 0.0001560442119071065,
+      "loss": 0.2731,
+      "step": 2032
+    },
+    {
+      "epoch": 1.1395739910313902,
+      "grad_norm": 0.07146571140318538,
+      "learning_rate": 0.00015599017492626773,
+      "loss": 0.2719,
+      "step": 2033
+    },
+    {
+      "epoch": 1.140134529147982,
+      "grad_norm": 0.07223059545245242,
+      "learning_rate": 0.00015593611412011686,
+      "loss": 0.2724,
+      "step": 2034
+    },
+    {
+      "epoch": 1.140695067264574,
+      "grad_norm": 0.07503730230596423,
+      "learning_rate": 0.00015588202951165824,
+      "loss": 0.2846,
+      "step": 2035
+    },
+    {
+      "epoch": 1.141255605381166,
+      "grad_norm": 0.07870846771326531,
+      "learning_rate": 0.0001558279211239063,
+      "loss": 0.2877,
+      "step": 2036
+    },
+    {
+      "epoch": 1.1418161434977578,
+      "grad_norm": 0.07129188141657869,
+      "learning_rate": 0.0001557737889798856,
+      "loss": 0.2593,
+      "step": 2037
+    },
+    {
+      "epoch": 1.1423766816143497,
+      "grad_norm": 0.07298394346129729,
+      "learning_rate": 0.00015571963310263086,
+      "loss": 0.2703,
+      "step": 2038
+    },
+    {
+      "epoch": 1.1429372197309418,
+      "grad_norm": 0.07475042100280528,
+      "learning_rate": 0.0001556654535151868,
+      "loss": 0.2766,
+      "step": 2039
+    },
+    {
+      "epoch": 1.1434977578475336,
+      "grad_norm": 0.07522057409664021,
+      "learning_rate": 0.00015561125024060826,
+      "loss": 0.2919,
+      "step": 2040
+    },
+    {
+      "epoch": 1.1440582959641254,
+      "grad_norm": 0.07525585281821849,
+      "learning_rate": 0.00015555702330196023,
+      "loss": 0.2753,
+      "step": 2041
+    },
+    {
+      "epoch": 1.1446188340807175,
+      "grad_norm": 0.07307159500608724,
+      "learning_rate": 0.00015550277272231768,
+      "loss": 0.2662,
+      "step": 2042
+    },
+    {
+      "epoch": 1.1451793721973094,
+      "grad_norm": 0.07316600541297696,
+      "learning_rate": 0.00015544849852476572,
+      "loss": 0.2695,
+      "step": 2043
+    },
+    {
+      "epoch": 1.1457399103139014,
+      "grad_norm": 0.07328754262887907,
+      "learning_rate": 0.00015539420073239942,
+      "loss": 0.2728,
+      "step": 2044
+    },
+    {
+      "epoch": 1.1463004484304933,
+      "grad_norm": 0.07449548962149974,
+      "learning_rate": 0.00015533987936832398,
+      "loss": 0.2757,
+      "step": 2045
+    },
+    {
+      "epoch": 1.1468609865470851,
+      "grad_norm": 0.0717644803480842,
+      "learning_rate": 0.00015528553445565453,
+      "loss": 0.2692,
+      "step": 2046
+    },
+    {
+      "epoch": 1.1474215246636772,
+      "grad_norm": 0.0760828551305723,
+      "learning_rate": 0.00015523116601751636,
+      "loss": 0.2705,
+      "step": 2047
+    },
+    {
+      "epoch": 1.147982062780269,
+      "grad_norm": 0.07217028849499157,
+      "learning_rate": 0.0001551767740770446,
+      "loss": 0.2624,
+      "step": 2048
+    },
+    {
+      "epoch": 1.148542600896861,
+      "grad_norm": 0.07294470052669451,
+      "learning_rate": 0.00015512235865738455,
+      "loss": 0.2736,
+      "step": 2049
+    },
+    {
+      "epoch": 1.149103139013453,
+      "grad_norm": 0.07264040780197499,
+      "learning_rate": 0.00015506791978169137,
+      "loss": 0.2675,
+      "step": 2050
+    },
+    {
+      "epoch": 1.1496636771300448,
+      "grad_norm": 0.07380138969502384,
+      "learning_rate": 0.00015501345747313027,
+      "loss": 0.2672,
+      "step": 2051
+    },
+    {
+      "epoch": 1.1502242152466366,
+      "grad_norm": 0.07393005364478991,
+      "learning_rate": 0.00015495897175487645,
+      "loss": 0.2728,
+      "step": 2052
+    },
+    {
+      "epoch": 1.1507847533632287,
+      "grad_norm": 0.07228127557031969,
+      "learning_rate": 0.00015490446265011495,
+      "loss": 0.2614,
+      "step": 2053
+    },
+    {
+      "epoch": 1.1513452914798206,
+      "grad_norm": 0.07395836335811319,
+      "learning_rate": 0.00015484993018204094,
+      "loss": 0.2619,
+      "step": 2054
+    },
+    {
+      "epoch": 1.1519058295964126,
+      "grad_norm": 0.07793273892744321,
+      "learning_rate": 0.00015479537437385938,
+      "loss": 0.2872,
+      "step": 2055
+    },
+    {
+      "epoch": 1.1524663677130045,
+      "grad_norm": 0.07701208241018277,
+      "learning_rate": 0.00015474079524878525,
+      "loss": 0.2789,
+      "step": 2056
+    },
+    {
+      "epoch": 1.1530269058295963,
+      "grad_norm": 0.07303503277383908,
+      "learning_rate": 0.0001546861928300434,
+      "loss": 0.2643,
+      "step": 2057
+    },
+    {
+      "epoch": 1.1535874439461884,
+      "grad_norm": 0.07499397816821558,
+      "learning_rate": 0.00015463156714086863,
+      "loss": 0.2839,
+      "step": 2058
+    },
+    {
+      "epoch": 1.1541479820627802,
+      "grad_norm": 0.07245367094461294,
+      "learning_rate": 0.00015457691820450564,
+      "loss": 0.2741,
+      "step": 2059
+    },
+    {
+      "epoch": 1.1547085201793723,
+      "grad_norm": 0.07284724044775001,
+      "learning_rate": 0.00015452224604420897,
+      "loss": 0.2774,
+      "step": 2060
+    },
+    {
+      "epoch": 1.1552690582959642,
+      "grad_norm": 0.07538937638420962,
+      "learning_rate": 0.0001544675506832431,
+      "loss": 0.2701,
+      "step": 2061
+    },
+    {
+      "epoch": 1.155829596412556,
+      "grad_norm": 0.071544697213512,
+      "learning_rate": 0.0001544128321448824,
+      "loss": 0.2726,
+      "step": 2062
+    },
+    {
+      "epoch": 1.156390134529148,
+      "grad_norm": 0.07273860295876183,
+      "learning_rate": 0.000154358090452411,
+      "loss": 0.2686,
+      "step": 2063
+    },
+    {
+      "epoch": 1.15695067264574,
+      "grad_norm": 0.07362864858475911,
+      "learning_rate": 0.000154303325629123,
+      "loss": 0.2832,
+      "step": 2064
+    },
+    {
+      "epoch": 1.1575112107623318,
+      "grad_norm": 0.07280685817498073,
+      "learning_rate": 0.00015424853769832226,
+      "loss": 0.269,
+      "step": 2065
+    },
+    {
+      "epoch": 1.1580717488789238,
+      "grad_norm": 0.07646770256689027,
+      "learning_rate": 0.00015419372668332254,
+      "loss": 0.2726,
+      "step": 2066
+    },
+    {
+      "epoch": 1.1586322869955157,
+      "grad_norm": 0.07101844516132096,
+      "learning_rate": 0.00015413889260744735,
+      "loss": 0.2636,
+      "step": 2067
+    },
+    {
+      "epoch": 1.1591928251121075,
+      "grad_norm": 0.07379766462631883,
+      "learning_rate": 0.0001540840354940301,
+      "loss": 0.2728,
+      "step": 2068
+    },
+    {
+      "epoch": 1.1597533632286996,
+      "grad_norm": 0.07221682533650452,
+      "learning_rate": 0.0001540291553664139,
+      "loss": 0.2747,
+      "step": 2069
+    },
+    {
+      "epoch": 1.1603139013452914,
+      "grad_norm": 0.07401261609018353,
+      "learning_rate": 0.00015397425224795177,
+      "loss": 0.2756,
+      "step": 2070
+    },
+    {
+      "epoch": 1.1608744394618835,
+      "grad_norm": 0.07339268743499448,
+      "learning_rate": 0.0001539193261620064,
+      "loss": 0.2817,
+      "step": 2071
+    },
+    {
+      "epoch": 1.1614349775784754,
+      "grad_norm": 0.07129692574775143,
+      "learning_rate": 0.0001538643771319503,
+      "loss": 0.2743,
+      "step": 2072
+    },
+    {
+      "epoch": 1.1619955156950672,
+      "grad_norm": 0.07368552168044658,
+      "learning_rate": 0.00015380940518116578,
+      "loss": 0.2711,
+      "step": 2073
+    },
+    {
+      "epoch": 1.1625560538116593,
+      "grad_norm": 0.0741087351189812,
+      "learning_rate": 0.00015375441033304484,
+      "loss": 0.2617,
+      "step": 2074
+    },
+    {
+      "epoch": 1.1631165919282511,
+      "grad_norm": 0.07145873287096843,
+      "learning_rate": 0.00015369939261098927,
+      "loss": 0.2648,
+      "step": 2075
+    },
+    {
+      "epoch": 1.163677130044843,
+      "grad_norm": 0.07312692487144262,
+      "learning_rate": 0.00015364435203841058,
+      "loss": 0.284,
+      "step": 2076
+    },
+    {
+      "epoch": 1.164237668161435,
+      "grad_norm": 0.07408197867575649,
+      "learning_rate": 0.00015358928863872998,
+      "loss": 0.263,
+      "step": 2077
+    },
+    {
+      "epoch": 1.1647982062780269,
+      "grad_norm": 0.07227449052330495,
+      "learning_rate": 0.00015353420243537848,
+      "loss": 0.2583,
+      "step": 2078
+    },
+    {
+      "epoch": 1.1653587443946187,
+      "grad_norm": 0.07218673290807645,
+      "learning_rate": 0.00015347909345179666,
+      "loss": 0.2701,
+      "step": 2079
+    },
+    {
+      "epoch": 1.1659192825112108,
+      "grad_norm": 0.07277001388474791,
+      "learning_rate": 0.00015342396171143488,
+      "loss": 0.2738,
+      "step": 2080
+    },
+    {
+      "epoch": 1.1664798206278026,
+      "grad_norm": 0.07384023078084269,
+      "learning_rate": 0.00015336880723775312,
+      "loss": 0.2792,
+      "step": 2081
+    },
+    {
+      "epoch": 1.1670403587443947,
+      "grad_norm": 0.07221786179926024,
+      "learning_rate": 0.00015331363005422117,
+      "loss": 0.2737,
+      "step": 2082
+    },
+    {
+      "epoch": 1.1676008968609866,
+      "grad_norm": 0.07454906070908028,
+      "learning_rate": 0.00015325843018431835,
+      "loss": 0.2711,
+      "step": 2083
+    },
+    {
+      "epoch": 1.1681614349775784,
+      "grad_norm": 0.07217768915976,
+      "learning_rate": 0.00015320320765153367,
+      "loss": 0.2842,
+      "step": 2084
+    },
+    {
+      "epoch": 1.1687219730941705,
+      "grad_norm": 0.07307954618021212,
+      "learning_rate": 0.00015314796247936578,
+      "loss": 0.2691,
+      "step": 2085
+    },
+    {
+      "epoch": 1.1692825112107623,
+      "grad_norm": 0.07459096911237502,
+      "learning_rate": 0.000153092694691323,
+      "loss": 0.2688,
+      "step": 2086
+    },
+    {
+      "epoch": 1.1698430493273542,
+      "grad_norm": 0.07051881368437275,
+      "learning_rate": 0.00015303740431092325,
+      "loss": 0.2673,
+      "step": 2087
+    },
+    {
+      "epoch": 1.1704035874439462,
+      "grad_norm": 0.07355848353297499,
+      "learning_rate": 0.00015298209136169403,
+      "loss": 0.2695,
+      "step": 2088
+    },
+    {
+      "epoch": 1.170964125560538,
+      "grad_norm": 0.0739564640053159,
+      "learning_rate": 0.00015292675586717246,
+      "loss": 0.2745,
+      "step": 2089
+    },
+    {
+      "epoch": 1.17152466367713,
+      "grad_norm": 0.07679015985636238,
+      "learning_rate": 0.00015287139785090533,
+      "loss": 0.2779,
+      "step": 2090
+    },
+    {
+      "epoch": 1.172085201793722,
+      "grad_norm": 0.07606643203847724,
+      "learning_rate": 0.00015281601733644894,
+      "loss": 0.2675,
+      "step": 2091
+    },
+    {
+      "epoch": 1.1726457399103138,
+      "grad_norm": 0.0756602189290465,
+      "learning_rate": 0.00015276061434736914,
+      "loss": 0.2693,
+      "step": 2092
+    },
+    {
+      "epoch": 1.173206278026906,
+      "grad_norm": 0.07538211370389455,
+      "learning_rate": 0.00015270518890724138,
+      "loss": 0.2822,
+      "step": 2093
+    },
+    {
+      "epoch": 1.1737668161434978,
+      "grad_norm": 0.07369372022215448,
+      "learning_rate": 0.00015264974103965068,
+      "loss": 0.2779,
+      "step": 2094
+    },
+    {
+      "epoch": 1.1743273542600896,
+      "grad_norm": 0.07313457952230602,
+      "learning_rate": 0.0001525942707681916,
+      "loss": 0.2709,
+      "step": 2095
+    },
+    {
+      "epoch": 1.1748878923766817,
+      "grad_norm": 0.0772087090084902,
+      "learning_rate": 0.00015253877811646817,
+      "loss": 0.2867,
+      "step": 2096
+    },
+    {
+      "epoch": 1.1754484304932735,
+      "grad_norm": 0.07470780457162388,
+      "learning_rate": 0.00015248326310809404,
+      "loss": 0.2798,
+      "step": 2097
+    },
+    {
+      "epoch": 1.1760089686098656,
+      "grad_norm": 0.07119790551418703,
+      "learning_rate": 0.00015242772576669236,
+      "loss": 0.2758,
+      "step": 2098
+    },
+    {
+      "epoch": 1.1765695067264574,
+      "grad_norm": 0.07251655224837092,
+      "learning_rate": 0.00015237216611589563,
+      "loss": 0.2738,
+      "step": 2099
+    },
+    {
+      "epoch": 1.1771300448430493,
+      "grad_norm": 0.07382764770345313,
+      "learning_rate": 0.00015231658417934606,
+      "loss": 0.2707,
+      "step": 2100
+    },
+    {
+      "epoch": 1.1776905829596414,
+      "grad_norm": 0.07487957934819602,
+      "learning_rate": 0.0001522609799806952,
+      "loss": 0.2783,
+      "step": 2101
+    },
+    {
+      "epoch": 1.1782511210762332,
+      "grad_norm": 0.07022136303111838,
+      "learning_rate": 0.00015220535354360415,
+      "loss": 0.2607,
+      "step": 2102
+    },
+    {
+      "epoch": 1.178811659192825,
+      "grad_norm": 0.0740416630693838,
+      "learning_rate": 0.00015214970489174341,
+      "loss": 0.2809,
+      "step": 2103
+    },
+    {
+      "epoch": 1.1793721973094171,
+      "grad_norm": 0.07274708528402801,
+      "learning_rate": 0.00015209403404879303,
+      "loss": 0.2728,
+      "step": 2104
+    },
+    {
+      "epoch": 1.179932735426009,
+      "grad_norm": 0.07598974070073093,
+      "learning_rate": 0.00015203834103844237,
+      "loss": 0.2745,
+      "step": 2105
+    },
+    {
+      "epoch": 1.1804932735426008,
+      "grad_norm": 0.07312980386250663,
+      "learning_rate": 0.00015198262588439032,
+      "loss": 0.2696,
+      "step": 2106
+    },
+    {
+      "epoch": 1.1810538116591929,
+      "grad_norm": 0.0735721273097422,
+      "learning_rate": 0.00015192688861034515,
+      "loss": 0.2749,
+      "step": 2107
+    },
+    {
+      "epoch": 1.1816143497757847,
+      "grad_norm": 0.07301663598658625,
+      "learning_rate": 0.00015187112924002456,
+      "loss": 0.2671,
+      "step": 2108
+    },
+    {
+      "epoch": 1.1821748878923768,
+      "grad_norm": 0.07044857634256689,
+      "learning_rate": 0.00015181534779715565,
+      "loss": 0.2529,
+      "step": 2109
+    },
+    {
+      "epoch": 1.1827354260089686,
+      "grad_norm": 0.0742907419683679,
+      "learning_rate": 0.0001517595443054749,
+      "loss": 0.2661,
+      "step": 2110
+    },
+    {
+      "epoch": 1.1832959641255605,
+      "grad_norm": 0.07310441013458707,
+      "learning_rate": 0.00015170371878872818,
+      "loss": 0.2743,
+      "step": 2111
+    },
+    {
+      "epoch": 1.1838565022421526,
+      "grad_norm": 0.07238894594001412,
+      "learning_rate": 0.0001516478712706708,
+      "loss": 0.2562,
+      "step": 2112
+    },
+    {
+      "epoch": 1.1844170403587444,
+      "grad_norm": 0.07786287525190146,
+      "learning_rate": 0.00015159200177506727,
+      "loss": 0.2697,
+      "step": 2113
+    },
+    {
+      "epoch": 1.1849775784753362,
+      "grad_norm": 0.07261831937645209,
+      "learning_rate": 0.0001515361103256916,
+      "loss": 0.2695,
+      "step": 2114
+    },
+    {
+      "epoch": 1.1855381165919283,
+      "grad_norm": 0.07532954820401011,
+      "learning_rate": 0.00015148019694632715,
+      "loss": 0.2771,
+      "step": 2115
+    },
+    {
+      "epoch": 1.1860986547085202,
+      "grad_norm": 0.07658085746908065,
+      "learning_rate": 0.00015142426166076645,
+      "loss": 0.2858,
+      "step": 2116
+    },
+    {
+      "epoch": 1.186659192825112,
+      "grad_norm": 0.07542217177630721,
+      "learning_rate": 0.00015136830449281148,
+      "loss": 0.2784,
+      "step": 2117
+    },
+    {
+      "epoch": 1.187219730941704,
+      "grad_norm": 0.07528234769052114,
+      "learning_rate": 0.00015131232546627355,
+      "loss": 0.2657,
+      "step": 2118
+    },
+    {
+      "epoch": 1.187780269058296,
+      "grad_norm": 0.07515155676111435,
+      "learning_rate": 0.0001512563246049732,
+      "loss": 0.2703,
+      "step": 2119
+    },
+    {
+      "epoch": 1.188340807174888,
+      "grad_norm": 0.07361613613334834,
+      "learning_rate": 0.00015120030193274027,
+      "loss": 0.2627,
+      "step": 2120
+    },
+    {
+      "epoch": 1.1889013452914798,
+      "grad_norm": 0.07367647072905742,
+      "learning_rate": 0.00015114425747341396,
+      "loss": 0.2695,
+      "step": 2121
+    },
+    {
+      "epoch": 1.1894618834080717,
+      "grad_norm": 0.07517684710602675,
+      "learning_rate": 0.00015108819125084262,
+      "loss": 0.2734,
+      "step": 2122
+    },
+    {
+      "epoch": 1.1900224215246638,
+      "grad_norm": 0.07459082377476396,
+      "learning_rate": 0.00015103210328888396,
+      "loss": 0.2687,
+      "step": 2123
+    },
+    {
+      "epoch": 1.1905829596412556,
+      "grad_norm": 0.07449003919367506,
+      "learning_rate": 0.00015097599361140487,
+      "loss": 0.2722,
+      "step": 2124
+    },
+    {
+      "epoch": 1.1911434977578474,
+      "grad_norm": 0.07212325460713669,
+      "learning_rate": 0.00015091986224228157,
+      "loss": 0.2658,
+      "step": 2125
+    },
+    {
+      "epoch": 1.1917040358744395,
+      "grad_norm": 0.07171874994936649,
+      "learning_rate": 0.00015086370920539937,
+      "loss": 0.2776,
+      "step": 2126
+    },
+    {
+      "epoch": 1.1922645739910314,
+      "grad_norm": 0.0724722953900823,
+      "learning_rate": 0.00015080753452465296,
+      "loss": 0.2786,
+      "step": 2127
+    },
+    {
+      "epoch": 1.1928251121076232,
+      "grad_norm": 0.07513205905519629,
+      "learning_rate": 0.00015075133822394613,
+      "loss": 0.2717,
+      "step": 2128
+    },
+    {
+      "epoch": 1.1933856502242153,
+      "grad_norm": 0.07262251960721408,
+      "learning_rate": 0.0001506951203271919,
+      "loss": 0.2719,
+      "step": 2129
+    },
+    {
+      "epoch": 1.1939461883408071,
+      "grad_norm": 0.07317864940142353,
+      "learning_rate": 0.0001506388808583125,
+      "loss": 0.2725,
+      "step": 2130
+    },
+    {
+      "epoch": 1.1945067264573992,
+      "grad_norm": 0.07543270870371434,
+      "learning_rate": 0.0001505826198412393,
+      "loss": 0.274,
+      "step": 2131
+    },
+    {
+      "epoch": 1.195067264573991,
+      "grad_norm": 0.07626307700070344,
+      "learning_rate": 0.00015052633729991294,
+      "loss": 0.285,
+      "step": 2132
+    },
+    {
+      "epoch": 1.1956278026905829,
+      "grad_norm": 0.07192479462550591,
+      "learning_rate": 0.00015047003325828305,
+      "loss": 0.2703,
+      "step": 2133
+    },
+    {
+      "epoch": 1.196188340807175,
+      "grad_norm": 0.0750745523992153,
+      "learning_rate": 0.0001504137077403085,
+      "loss": 0.2778,
+      "step": 2134
+    },
+    {
+      "epoch": 1.1967488789237668,
+      "grad_norm": 0.07457514985822909,
+      "learning_rate": 0.00015035736076995736,
+      "loss": 0.2593,
+      "step": 2135
+    },
+    {
+      "epoch": 1.1973094170403586,
+      "grad_norm": 0.07286343319243911,
+      "learning_rate": 0.00015030099237120674,
+      "loss": 0.2761,
+      "step": 2136
+    },
+    {
+      "epoch": 1.1978699551569507,
+      "grad_norm": 0.07453411808354928,
+      "learning_rate": 0.0001502446025680429,
+      "loss": 0.267,
+      "step": 2137
+    },
+    {
+      "epoch": 1.1984304932735426,
+      "grad_norm": 0.0745072188165511,
+      "learning_rate": 0.0001501881913844612,
+      "loss": 0.2767,
+      "step": 2138
+    },
+    {
+      "epoch": 1.1989910313901344,
+      "grad_norm": 0.07249510969676805,
+      "learning_rate": 0.0001501317588444661,
+      "loss": 0.2637,
+      "step": 2139
+    },
+    {
+      "epoch": 1.1995515695067265,
+      "grad_norm": 0.07717624953611066,
+      "learning_rate": 0.00015007530497207117,
+      "loss": 0.2794,
+      "step": 2140
+    },
+    {
+      "epoch": 1.2001121076233183,
+      "grad_norm": 0.07387397226639023,
+      "learning_rate": 0.00015001882979129899,
+      "loss": 0.2743,
+      "step": 2141
+    },
+    {
+      "epoch": 1.2006726457399104,
+      "grad_norm": 0.07496227177042705,
+      "learning_rate": 0.00014996233332618128,
+      "loss": 0.2755,
+      "step": 2142
+    },
+    {
+      "epoch": 1.2012331838565022,
+      "grad_norm": 0.07496313798382635,
+      "learning_rate": 0.00014990581560075881,
+      "loss": 0.2721,
+      "step": 2143
+    },
+    {
+      "epoch": 1.201793721973094,
+      "grad_norm": 0.07439376297487851,
+      "learning_rate": 0.00014984927663908137,
+      "loss": 0.2708,
+      "step": 2144
+    },
+    {
+      "epoch": 1.2023542600896862,
+      "grad_norm": 0.07523485667912887,
+      "learning_rate": 0.00014979271646520782,
+      "loss": 0.2612,
+      "step": 2145
+    },
+    {
+      "epoch": 1.202914798206278,
+      "grad_norm": 0.0716685011793692,
+      "learning_rate": 0.00014973613510320594,
+      "loss": 0.264,
+      "step": 2146
+    },
+    {
+      "epoch": 1.20347533632287,
+      "grad_norm": 0.07371711472779705,
+      "learning_rate": 0.00014967953257715268,
+      "loss": 0.2754,
+      "step": 2147
+    },
+    {
+      "epoch": 1.204035874439462,
+      "grad_norm": 0.0752351159209486,
+      "learning_rate": 0.00014962290891113394,
+      "loss": 0.2821,
+      "step": 2148
+    },
+    {
+      "epoch": 1.2045964125560538,
+      "grad_norm": 0.0732745052311195,
+      "learning_rate": 0.00014956626412924453,
+      "loss": 0.2775,
+      "step": 2149
+    },
+    {
+      "epoch": 1.2051569506726458,
+      "grad_norm": 0.07324326890999037,
+      "learning_rate": 0.00014950959825558836,
+      "loss": 0.2751,
+      "step": 2150
+    },
+    {
+      "epoch": 1.2057174887892377,
+      "grad_norm": 0.07409978346285448,
+      "learning_rate": 0.00014945291131427825,
+      "loss": 0.2834,
+      "step": 2151
+    },
+    {
+      "epoch": 1.2062780269058295,
+      "grad_norm": 0.07404126723380783,
+      "learning_rate": 0.00014939620332943604,
+      "loss": 0.2646,
+      "step": 2152
+    },
+    {
+      "epoch": 1.2068385650224216,
+      "grad_norm": 0.07189288322301148,
+      "learning_rate": 0.00014933947432519245,
+      "loss": 0.276,
+      "step": 2153
+    },
+    {
+      "epoch": 1.2073991031390134,
+      "grad_norm": 0.0711885201660691,
+      "learning_rate": 0.0001492827243256872,
+      "loss": 0.2629,
+      "step": 2154
+    },
+    {
+      "epoch": 1.2079596412556053,
+      "grad_norm": 0.07357263793071628,
+      "learning_rate": 0.00014922595335506892,
+      "loss": 0.2739,
+      "step": 2155
+    },
+    {
+      "epoch": 1.2085201793721974,
+      "grad_norm": 0.07372589669407172,
+      "learning_rate": 0.00014916916143749518,
+      "loss": 0.2762,
+      "step": 2156
+    },
+    {
+      "epoch": 1.2090807174887892,
+      "grad_norm": 0.07925057873266328,
+      "learning_rate": 0.00014911234859713243,
+      "loss": 0.2636,
+      "step": 2157
+    },
+    {
+      "epoch": 1.2096412556053813,
+      "grad_norm": 0.07224035604882845,
+      "learning_rate": 0.0001490555148581561,
+      "loss": 0.2728,
+      "step": 2158
+    },
+    {
+      "epoch": 1.2102017937219731,
+      "grad_norm": 0.07328029375527327,
+      "learning_rate": 0.00014899866024475043,
+      "loss": 0.2714,
+      "step": 2159
+    },
+    {
+      "epoch": 1.210762331838565,
+      "grad_norm": 0.07440276422602377,
+      "learning_rate": 0.00014894178478110857,
+      "loss": 0.2811,
+      "step": 2160
+    },
+    {
+      "epoch": 1.211322869955157,
+      "grad_norm": 0.07292885635959033,
+      "learning_rate": 0.00014888488849143253,
+      "loss": 0.2671,
+      "step": 2161
+    },
+    {
+      "epoch": 1.2118834080717489,
+      "grad_norm": 0.07235886439051883,
+      "learning_rate": 0.00014882797139993326,
+      "loss": 0.2651,
+      "step": 2162
+    },
+    {
+      "epoch": 1.2124439461883407,
+      "grad_norm": 0.07311515589613908,
+      "learning_rate": 0.00014877103353083042,
+      "loss": 0.2785,
+      "step": 2163
+    },
+    {
+      "epoch": 1.2130044843049328,
+      "grad_norm": 0.0742640259694533,
+      "learning_rate": 0.00014871407490835262,
+      "loss": 0.2679,
+      "step": 2164
+    },
+    {
+      "epoch": 1.2135650224215246,
+      "grad_norm": 0.07121133370611302,
+      "learning_rate": 0.00014865709555673734,
+      "loss": 0.2678,
+      "step": 2165
+    },
+    {
+      "epoch": 1.2141255605381165,
+      "grad_norm": 0.07120529979390215,
+      "learning_rate": 0.00014860009550023072,
+      "loss": 0.2689,
+      "step": 2166
+    },
+    {
+      "epoch": 1.2146860986547086,
+      "grad_norm": 0.07297595302908803,
+      "learning_rate": 0.00014854307476308781,
+      "loss": 0.2754,
+      "step": 2167
+    },
+    {
+      "epoch": 1.2152466367713004,
+      "grad_norm": 0.0765807805862152,
+      "learning_rate": 0.00014848603336957251,
+      "loss": 0.2811,
+      "step": 2168
+    },
+    {
+      "epoch": 1.2158071748878925,
+      "grad_norm": 0.08047048222225542,
+      "learning_rate": 0.00014842897134395743,
+      "loss": 0.2671,
+      "step": 2169
+    },
+    {
+      "epoch": 1.2163677130044843,
+      "grad_norm": 0.08551996410336832,
+      "learning_rate": 0.000148371888710524,
+      "loss": 0.266,
+      "step": 2170
+    },
+    {
+      "epoch": 1.2169282511210762,
+      "grad_norm": 0.07120052575877524,
+      "learning_rate": 0.00014831478549356234,
+      "loss": 0.2497,
+      "step": 2171
+    },
+    {
+      "epoch": 1.2174887892376682,
+      "grad_norm": 0.0722776407246042,
+      "learning_rate": 0.00014825766171737146,
+      "loss": 0.2654,
+      "step": 2172
+    },
+    {
+      "epoch": 1.21804932735426,
+      "grad_norm": 0.07458138290458914,
+      "learning_rate": 0.00014820051740625903,
+      "loss": 0.271,
+      "step": 2173
+    },
+    {
+      "epoch": 1.218609865470852,
+      "grad_norm": 0.0729714914697601,
+      "learning_rate": 0.00014814335258454145,
+      "loss": 0.2665,
+      "step": 2174
+    },
+    {
+      "epoch": 1.219170403587444,
+      "grad_norm": 0.07460522134944941,
+      "learning_rate": 0.0001480861672765439,
+      "loss": 0.2817,
+      "step": 2175
+    },
+    {
+      "epoch": 1.2197309417040358,
+      "grad_norm": 0.07836658995501097,
+      "learning_rate": 0.00014802896150660022,
+      "loss": 0.2754,
+      "step": 2176
+    },
+    {
+      "epoch": 1.2202914798206277,
+      "grad_norm": 0.0736678787021427,
+      "learning_rate": 0.00014797173529905306,
+      "loss": 0.2798,
+      "step": 2177
+    },
+    {
+      "epoch": 1.2208520179372198,
+      "grad_norm": 0.07071270337372172,
+      "learning_rate": 0.00014791448867825365,
+      "loss": 0.2584,
+      "step": 2178
+    },
+    {
+      "epoch": 1.2214125560538116,
+      "grad_norm": 0.07231434933962502,
+      "learning_rate": 0.00014785722166856194,
+      "loss": 0.2769,
+      "step": 2179
+    },
+    {
+      "epoch": 1.2219730941704037,
+      "grad_norm": 0.07094203957441038,
+      "learning_rate": 0.00014779993429434659,
+      "loss": 0.2635,
+      "step": 2180
+    },
+    {
+      "epoch": 1.2225336322869955,
+      "grad_norm": 0.07033983329472304,
+      "learning_rate": 0.00014774262657998491,
+      "loss": 0.2675,
+      "step": 2181
+    },
+    {
+      "epoch": 1.2230941704035874,
+      "grad_norm": 0.07425672814540628,
+      "learning_rate": 0.00014768529854986286,
+      "loss": 0.2773,
+      "step": 2182
+    },
+    {
+      "epoch": 1.2236547085201794,
+      "grad_norm": 0.07118439642612227,
+      "learning_rate": 0.00014762795022837504,
+      "loss": 0.2678,
+      "step": 2183
+    },
+    {
+      "epoch": 1.2242152466367713,
+      "grad_norm": 0.07350427202585197,
+      "learning_rate": 0.00014757058163992464,
+      "loss": 0.2746,
+      "step": 2184
+    },
+    {
+      "epoch": 1.2247757847533634,
+      "grad_norm": 0.07384173873934805,
+      "learning_rate": 0.00014751319280892366,
+      "loss": 0.2707,
+      "step": 2185
+    },
+    {
+      "epoch": 1.2253363228699552,
+      "grad_norm": 0.07511276355337908,
+      "learning_rate": 0.00014745578375979245,
+      "loss": 0.2671,
+      "step": 2186
+    },
+    {
+      "epoch": 1.225896860986547,
+      "grad_norm": 0.07542390591320873,
+      "learning_rate": 0.00014739835451696011,
+      "loss": 0.2796,
+      "step": 2187
+    },
+    {
+      "epoch": 1.226457399103139,
+      "grad_norm": 0.07318198970241921,
+      "learning_rate": 0.00014734090510486433,
+      "loss": 0.2765,
+      "step": 2188
+    },
+    {
+      "epoch": 1.227017937219731,
+      "grad_norm": 0.07221627487433964,
+      "learning_rate": 0.00014728343554795142,
+      "loss": 0.274,
+      "step": 2189
+    },
+    {
+      "epoch": 1.2275784753363228,
+      "grad_norm": 0.07302762715277836,
+      "learning_rate": 0.0001472259458706761,
+      "loss": 0.2735,
+      "step": 2190
+    },
+    {
+      "epoch": 1.2281390134529149,
+      "grad_norm": 0.07071971217399264,
+      "learning_rate": 0.00014716843609750187,
+      "loss": 0.2597,
+      "step": 2191
+    },
+    {
+      "epoch": 1.2286995515695067,
+      "grad_norm": 0.07064134460184776,
+      "learning_rate": 0.00014711090625290057,
+      "loss": 0.2742,
+      "step": 2192
+    },
+    {
+      "epoch": 1.2292600896860986,
+      "grad_norm": 0.07175876715421317,
+      "learning_rate": 0.0001470533563613528,
+      "loss": 0.2774,
+      "step": 2193
+    },
+    {
+      "epoch": 1.2298206278026906,
+      "grad_norm": 0.07201379572339128,
+      "learning_rate": 0.00014699578644734746,
+      "loss": 0.267,
+      "step": 2194
+    },
+    {
+      "epoch": 1.2303811659192825,
+      "grad_norm": 0.07433355312968802,
+      "learning_rate": 0.00014693819653538215,
+      "loss": 0.2627,
+      "step": 2195
+    },
+    {
+      "epoch": 1.2309417040358746,
+      "grad_norm": 0.07390570798339077,
+      "learning_rate": 0.00014688058664996285,
+      "loss": 0.2661,
+      "step": 2196
+    },
+    {
+      "epoch": 1.2315022421524664,
+      "grad_norm": 0.073286120109269,
+      "learning_rate": 0.0001468229568156042,
+      "loss": 0.2797,
+      "step": 2197
+    },
+    {
+      "epoch": 1.2320627802690582,
+      "grad_norm": 0.07485715475066163,
+      "learning_rate": 0.00014676530705682914,
+      "loss": 0.2539,
+      "step": 2198
+    },
+    {
+      "epoch": 1.2326233183856503,
+      "grad_norm": 0.07433746289621238,
+      "learning_rate": 0.00014670763739816923,
+      "loss": 0.2656,
+      "step": 2199
+    },
+    {
+      "epoch": 1.2331838565022422,
+      "grad_norm": 0.07812464782683104,
+      "learning_rate": 0.0001466499478641644,
+      "loss": 0.2801,
+      "step": 2200
+    },
+    {
+      "epoch": 1.233744394618834,
+      "grad_norm": 0.07406661426771381,
+      "learning_rate": 0.00014659223847936315,
+      "loss": 0.2726,
+      "step": 2201
+    },
+    {
+      "epoch": 1.234304932735426,
+      "grad_norm": 0.07283861596645486,
+      "learning_rate": 0.00014653450926832234,
+      "loss": 0.27,
+      "step": 2202
+    },
+    {
+      "epoch": 1.234865470852018,
+      "grad_norm": 0.07475440658412534,
+      "learning_rate": 0.00014647676025560726,
+      "loss": 0.2769,
+      "step": 2203
+    },
+    {
+      "epoch": 1.2354260089686098,
+      "grad_norm": 0.07276401211295648,
+      "learning_rate": 0.00014641899146579168,
+      "loss": 0.2684,
+      "step": 2204
+    },
+    {
+      "epoch": 1.2359865470852018,
+      "grad_norm": 0.07576724537918178,
+      "learning_rate": 0.00014636120292345773,
+      "loss": 0.2743,
+      "step": 2205
+    },
+    {
+      "epoch": 1.2365470852017937,
+      "grad_norm": 0.07233726605379628,
+      "learning_rate": 0.00014630339465319603,
+      "loss": 0.2692,
+      "step": 2206
+    },
+    {
+      "epoch": 1.2371076233183858,
+      "grad_norm": 0.07410354461819409,
+      "learning_rate": 0.00014624556667960548,
+      "loss": 0.2688,
+      "step": 2207
+    },
+    {
+      "epoch": 1.2376681614349776,
+      "grad_norm": 0.07233910524759327,
+      "learning_rate": 0.00014618771902729342,
+      "loss": 0.2805,
+      "step": 2208
+    },
+    {
+      "epoch": 1.2382286995515694,
+      "grad_norm": 0.0703341510133088,
+      "learning_rate": 0.00014612985172087565,
+      "loss": 0.2602,
+      "step": 2209
+    },
+    {
+      "epoch": 1.2387892376681615,
+      "grad_norm": 0.0728511550381156,
+      "learning_rate": 0.0001460719647849762,
+      "loss": 0.2635,
+      "step": 2210
+    },
+    {
+      "epoch": 1.2393497757847534,
+      "grad_norm": 0.0724761427933535,
+      "learning_rate": 0.0001460140582442275,
+      "loss": 0.2596,
+      "step": 2211
+    },
+    {
+      "epoch": 1.2399103139013452,
+      "grad_norm": 0.07432563748277206,
+      "learning_rate": 0.00014595613212327032,
+      "loss": 0.2765,
+      "step": 2212
+    },
+    {
+      "epoch": 1.2404708520179373,
+      "grad_norm": 0.07635799672042924,
+      "learning_rate": 0.00014589818644675378,
+      "loss": 0.2831,
+      "step": 2213
+    },
+    {
+      "epoch": 1.2410313901345291,
+      "grad_norm": 0.07608695610001634,
+      "learning_rate": 0.0001458402212393353,
+      "loss": 0.2757,
+      "step": 2214
+    },
+    {
+      "epoch": 1.241591928251121,
+      "grad_norm": 0.0751592324960594,
+      "learning_rate": 0.00014578223652568067,
+      "loss": 0.269,
+      "step": 2215
+    },
+    {
+      "epoch": 1.242152466367713,
+      "grad_norm": 0.07439220809306073,
+      "learning_rate": 0.00014572423233046386,
+      "loss": 0.2718,
+      "step": 2216
+    },
+    {
+      "epoch": 1.2427130044843049,
+      "grad_norm": 0.07575224592435262,
+      "learning_rate": 0.00014566620867836725,
+      "loss": 0.2715,
+      "step": 2217
+    },
+    {
+      "epoch": 1.243273542600897,
+      "grad_norm": 0.07518921890484784,
+      "learning_rate": 0.00014560816559408142,
+      "loss": 0.2735,
+      "step": 2218
+    },
+    {
+      "epoch": 1.2438340807174888,
+      "grad_norm": 0.07499287456456347,
+      "learning_rate": 0.0001455501031023053,
+      "loss": 0.27,
+      "step": 2219
+    },
+    {
+      "epoch": 1.2443946188340806,
+      "grad_norm": 0.07479889085555115,
+      "learning_rate": 0.00014549202122774596,
+      "loss": 0.2938,
+      "step": 2220
+    },
+    {
+      "epoch": 1.2449551569506727,
+      "grad_norm": 0.07352899020051516,
+      "learning_rate": 0.0001454339199951188,
+      "loss": 0.2765,
+      "step": 2221
+    },
+    {
+      "epoch": 1.2455156950672646,
+      "grad_norm": 0.07284971509791976,
+      "learning_rate": 0.00014537579942914752,
+      "loss": 0.2659,
+      "step": 2222
+    },
+    {
+      "epoch": 1.2460762331838564,
+      "grad_norm": 0.07305313033571034,
+      "learning_rate": 0.00014531765955456388,
+      "loss": 0.2717,
+      "step": 2223
+    },
+    {
+      "epoch": 1.2466367713004485,
+      "grad_norm": 0.07373772134837558,
+      "learning_rate": 0.000145259500396108,
+      "loss": 0.2711,
+      "step": 2224
+    },
+    {
+      "epoch": 1.2471973094170403,
+      "grad_norm": 0.07269328693647831,
+      "learning_rate": 0.00014520132197852812,
+      "loss": 0.2643,
+      "step": 2225
+    },
+    {
+      "epoch": 1.2477578475336322,
+      "grad_norm": 0.07349888227263872,
+      "learning_rate": 0.00014514312432658072,
+      "loss": 0.2868,
+      "step": 2226
+    },
+    {
+      "epoch": 1.2483183856502242,
+      "grad_norm": 0.0759531850087204,
+      "learning_rate": 0.00014508490746503044,
+      "loss": 0.2709,
+      "step": 2227
+    },
+    {
+      "epoch": 1.248878923766816,
+      "grad_norm": 0.07320229472749225,
+      "learning_rate": 0.00014502667141865015,
+      "loss": 0.2735,
+      "step": 2228
+    },
+    {
+      "epoch": 1.2494394618834082,
+      "grad_norm": 0.07410626056346635,
+      "learning_rate": 0.00014496841621222076,
+      "loss": 0.2712,
+      "step": 2229
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.07402976456458424,
+      "learning_rate": 0.00014491014187053148,
+      "loss": 0.2761,
+      "step": 2230
+    },
+    {
+      "epoch": 1.2505605381165918,
+      "grad_norm": 0.07312087350695935,
+      "learning_rate": 0.0001448518484183796,
+      "loss": 0.2681,
+      "step": 2231
+    },
+    {
+      "epoch": 1.251121076233184,
+      "grad_norm": 0.07213741476143876,
+      "learning_rate": 0.00014479353588057052,
+      "loss": 0.2553,
+      "step": 2232
+    },
+    {
+      "epoch": 1.2516816143497758,
+      "grad_norm": 0.07125397595984921,
+      "learning_rate": 0.00014473520428191775,
+      "loss": 0.2757,
+      "step": 2233
+    },
+    {
+      "epoch": 1.2522421524663678,
+      "grad_norm": 0.07434629780509557,
+      "learning_rate": 0.00014467685364724298,
+      "loss": 0.2751,
+      "step": 2234
+    },
+    {
+      "epoch": 1.2528026905829597,
+      "grad_norm": 0.0715831021024079,
+      "learning_rate": 0.00014461848400137595,
+      "loss": 0.2611,
+      "step": 2235
+    },
+    {
+      "epoch": 1.2533632286995515,
+      "grad_norm": 0.07675312304907582,
+      "learning_rate": 0.00014456009536915448,
+      "loss": 0.2675,
+      "step": 2236
+    },
+    {
+      "epoch": 1.2539237668161434,
+      "grad_norm": 0.07200440188423189,
+      "learning_rate": 0.0001445016877754245,
+      "loss": 0.273,
+      "step": 2237
+    },
+    {
+      "epoch": 1.2544843049327354,
+      "grad_norm": 0.07061599291190361,
+      "learning_rate": 0.00014444326124504002,
+      "loss": 0.2595,
+      "step": 2238
+    },
+    {
+      "epoch": 1.2550448430493273,
+      "grad_norm": 0.069665333881225,
+      "learning_rate": 0.0001443848158028631,
+      "loss": 0.2527,
+      "step": 2239
+    },
+    {
+      "epoch": 1.2556053811659194,
+      "grad_norm": 0.07085060283078363,
+      "learning_rate": 0.00014432635147376376,
+      "loss": 0.26,
+      "step": 2240
+    },
+    {
+      "epoch": 1.2561659192825112,
+      "grad_norm": 0.07287943724145639,
+      "learning_rate": 0.00014426786828262018,
+      "loss": 0.2734,
+      "step": 2241
+    },
+    {
+      "epoch": 1.256726457399103,
+      "grad_norm": 0.07600559046718026,
+      "learning_rate": 0.00014420936625431853,
+      "loss": 0.2737,
+      "step": 2242
+    },
+    {
+      "epoch": 1.2572869955156951,
+      "grad_norm": 0.07263207168045012,
+      "learning_rate": 0.00014415084541375295,
+      "loss": 0.2739,
+      "step": 2243
+    },
+    {
+      "epoch": 1.257847533632287,
+      "grad_norm": 0.07481167414877608,
+      "learning_rate": 0.00014409230578582566,
+      "loss": 0.2825,
+      "step": 2244
+    },
+    {
+      "epoch": 1.258408071748879,
+      "grad_norm": 0.07248565996941632,
+      "learning_rate": 0.00014403374739544678,
+      "loss": 0.2774,
+      "step": 2245
+    },
+    {
+      "epoch": 1.2589686098654709,
+      "grad_norm": 0.07526214400489531,
+      "learning_rate": 0.0001439751702675345,
+      "loss": 0.2659,
+      "step": 2246
+    },
+    {
+      "epoch": 1.2595291479820627,
+      "grad_norm": 0.07199083013001102,
+      "learning_rate": 0.00014391657442701494,
+      "loss": 0.2725,
+      "step": 2247
+    },
+    {
+      "epoch": 1.2600896860986546,
+      "grad_norm": 0.07185921680402563,
+      "learning_rate": 0.00014385795989882221,
+      "loss": 0.2764,
+      "step": 2248
+    },
+    {
+      "epoch": 1.2606502242152466,
+      "grad_norm": 0.0745909469659899,
+      "learning_rate": 0.00014379932670789832,
+      "loss": 0.2834,
+      "step": 2249
+    },
+    {
+      "epoch": 1.2612107623318385,
+      "grad_norm": 0.07356235271331986,
+      "learning_rate": 0.00014374067487919322,
+      "loss": 0.2841,
+      "step": 2250
+    },
+    {
+      "epoch": 1.2617713004484306,
+      "grad_norm": 0.07432903247307711,
+      "learning_rate": 0.00014368200443766495,
+      "loss": 0.2729,
+      "step": 2251
+    },
+    {
+      "epoch": 1.2623318385650224,
+      "grad_norm": 0.07183795198830267,
+      "learning_rate": 0.00014362331540827928,
+      "loss": 0.2679,
+      "step": 2252
+    },
+    {
+      "epoch": 1.2628923766816142,
+      "grad_norm": 0.0728036017529027,
+      "learning_rate": 0.00014356460781600992,
+      "loss": 0.2752,
+      "step": 2253
+    },
+    {
+      "epoch": 1.2634529147982063,
+      "grad_norm": 0.07085391731397918,
+      "learning_rate": 0.00014350588168583856,
+      "loss": 0.2721,
+      "step": 2254
+    },
+    {
+      "epoch": 1.2640134529147982,
+      "grad_norm": 0.0710387486000272,
+      "learning_rate": 0.00014344713704275472,
+      "loss": 0.2601,
+      "step": 2255
+    },
+    {
+      "epoch": 1.2645739910313902,
+      "grad_norm": 0.07169809121267175,
+      "learning_rate": 0.00014338837391175582,
+      "loss": 0.2703,
+      "step": 2256
+    },
+    {
+      "epoch": 1.265134529147982,
+      "grad_norm": 0.07053674888474745,
+      "learning_rate": 0.00014332959231784712,
+      "loss": 0.2643,
+      "step": 2257
+    },
+    {
+      "epoch": 1.265695067264574,
+      "grad_norm": 0.07205890920821309,
+      "learning_rate": 0.00014327079228604176,
+      "loss": 0.2798,
+      "step": 2258
+    },
+    {
+      "epoch": 1.266255605381166,
+      "grad_norm": 0.07386825344155837,
+      "learning_rate": 0.0001432119738413608,
+      "loss": 0.2747,
+      "step": 2259
+    },
+    {
+      "epoch": 1.2668161434977578,
+      "grad_norm": 0.0721786940621619,
+      "learning_rate": 0.00014315313700883294,
+      "loss": 0.2558,
+      "step": 2260
+    },
+    {
+      "epoch": 1.26737668161435,
+      "grad_norm": 0.07226327141121044,
+      "learning_rate": 0.00014309428181349484,
+      "loss": 0.279,
+      "step": 2261
+    },
+    {
+      "epoch": 1.2679372197309418,
+      "grad_norm": 0.07417550526920404,
+      "learning_rate": 0.00014303540828039098,
+      "loss": 0.282,
+      "step": 2262
+    },
+    {
+      "epoch": 1.2684977578475336,
+      "grad_norm": 0.07055665704719952,
+      "learning_rate": 0.00014297651643457366,
+      "loss": 0.2752,
+      "step": 2263
+    },
+    {
+      "epoch": 1.2690582959641254,
+      "grad_norm": 0.07145927025909532,
+      "learning_rate": 0.00014291760630110288,
+      "loss": 0.2721,
+      "step": 2264
+    },
+    {
+      "epoch": 1.2696188340807175,
+      "grad_norm": 0.07055847344841475,
+      "learning_rate": 0.0001428586779050465,
+      "loss": 0.2575,
+      "step": 2265
+    },
+    {
+      "epoch": 1.2701793721973094,
+      "grad_norm": 0.0706480831312279,
+      "learning_rate": 0.00014279973127148004,
+      "loss": 0.2786,
+      "step": 2266
+    },
+    {
+      "epoch": 1.2707399103139014,
+      "grad_norm": 0.07007179862259319,
+      "learning_rate": 0.000142740766425487,
+      "loss": 0.2619,
+      "step": 2267
+    },
+    {
+      "epoch": 1.2713004484304933,
+      "grad_norm": 0.07131325441368183,
+      "learning_rate": 0.00014268178339215838,
+      "loss": 0.2698,
+      "step": 2268
+    },
+    {
+      "epoch": 1.2718609865470851,
+      "grad_norm": 0.07415336161002052,
+      "learning_rate": 0.0001426227821965931,
+      "loss": 0.2729,
+      "step": 2269
+    },
+    {
+      "epoch": 1.2724215246636772,
+      "grad_norm": 0.07306777949703797,
+      "learning_rate": 0.00014256376286389769,
+      "loss": 0.2692,
+      "step": 2270
+    },
+    {
+      "epoch": 1.272982062780269,
+      "grad_norm": 0.07156311353403604,
+      "learning_rate": 0.0001425047254191865,
+      "loss": 0.2725,
+      "step": 2271
+    },
+    {
+      "epoch": 1.273542600896861,
+      "grad_norm": 0.07526729163791998,
+      "learning_rate": 0.00014244566988758152,
+      "loss": 0.2835,
+      "step": 2272
+    },
+    {
+      "epoch": 1.274103139013453,
+      "grad_norm": 0.07258970663293911,
+      "learning_rate": 0.00014238659629421245,
+      "loss": 0.2709,
+      "step": 2273
+    },
+    {
+      "epoch": 1.2746636771300448,
+      "grad_norm": 0.0750406240649443,
+      "learning_rate": 0.00014232750466421665,
+      "loss": 0.2783,
+      "step": 2274
+    },
+    {
+      "epoch": 1.2752242152466366,
+      "grad_norm": 0.07456835751881956,
+      "learning_rate": 0.0001422683950227392,
+      "loss": 0.2831,
+      "step": 2275
+    },
+    {
+      "epoch": 1.2757847533632287,
+      "grad_norm": 0.0738641247695516,
+      "learning_rate": 0.00014220926739493288,
+      "loss": 0.2862,
+      "step": 2276
+    },
+    {
+      "epoch": 1.2763452914798206,
+      "grad_norm": 0.07447914433347605,
+      "learning_rate": 0.00014215012180595802,
+      "loss": 0.2725,
+      "step": 2277
+    },
+    {
+      "epoch": 1.2769058295964126,
+      "grad_norm": 0.072193602978639,
+      "learning_rate": 0.00014209095828098263,
+      "loss": 0.2648,
+      "step": 2278
+    },
+    {
+      "epoch": 1.2774663677130045,
+      "grad_norm": 0.0706664606726907,
+      "learning_rate": 0.00014203177684518243,
+      "loss": 0.258,
+      "step": 2279
+    },
+    {
+      "epoch": 1.2780269058295963,
+      "grad_norm": 0.07333496681068567,
+      "learning_rate": 0.0001419725775237406,
+      "loss": 0.2773,
+      "step": 2280
+    },
+    {
+      "epoch": 1.2785874439461884,
+      "grad_norm": 0.07243360497997042,
+      "learning_rate": 0.00014191336034184818,
+      "loss": 0.2741,
+      "step": 2281
+    },
+    {
+      "epoch": 1.2791479820627802,
+      "grad_norm": 0.06998674397688194,
+      "learning_rate": 0.0001418541253247035,
+      "loss": 0.2667,
+      "step": 2282
+    },
+    {
+      "epoch": 1.2797085201793723,
+      "grad_norm": 0.06880900181815017,
+      "learning_rate": 0.0001417948724975127,
+      "loss": 0.2514,
+      "step": 2283
+    },
+    {
+      "epoch": 1.2802690582959642,
+      "grad_norm": 0.07143344551366512,
+      "learning_rate": 0.00014173560188548948,
+      "loss": 0.27,
+      "step": 2284
+    },
+    {
+      "epoch": 1.280829596412556,
+      "grad_norm": 0.07293076040915937,
+      "learning_rate": 0.00014167631351385504,
+      "loss": 0.2678,
+      "step": 2285
+    },
+    {
+      "epoch": 1.2813901345291479,
+      "grad_norm": 0.07382402327007528,
+      "learning_rate": 0.00014161700740783815,
+      "loss": 0.2812,
+      "step": 2286
+    },
+    {
+      "epoch": 1.28195067264574,
+      "grad_norm": 0.07046114694580449,
+      "learning_rate": 0.00014155768359267511,
+      "loss": 0.2603,
+      "step": 2287
+    },
+    {
+      "epoch": 1.2825112107623318,
+      "grad_norm": 0.07294584210310635,
+      "learning_rate": 0.00014149834209360986,
+      "loss": 0.2695,
+      "step": 2288
+    },
+    {
+      "epoch": 1.2830717488789238,
+      "grad_norm": 0.07311525705372239,
+      "learning_rate": 0.00014143898293589373,
+      "loss": 0.2697,
+      "step": 2289
+    },
+    {
+      "epoch": 1.2836322869955157,
+      "grad_norm": 0.07151234033113946,
+      "learning_rate": 0.00014137960614478564,
+      "loss": 0.2639,
+      "step": 2290
+    },
+    {
+      "epoch": 1.2841928251121075,
+      "grad_norm": 0.06883676087253877,
+      "learning_rate": 0.00014132021174555198,
+      "loss": 0.266,
+      "step": 2291
+    },
+    {
+      "epoch": 1.2847533632286996,
+      "grad_norm": 0.07323764340793483,
+      "learning_rate": 0.0001412607997634667,
+      "loss": 0.281,
+      "step": 2292
+    },
+    {
+      "epoch": 1.2853139013452914,
+      "grad_norm": 0.06994806350440688,
+      "learning_rate": 0.00014120137022381117,
+      "loss": 0.2607,
+      "step": 2293
+    },
+    {
+      "epoch": 1.2858744394618835,
+      "grad_norm": 0.07302141344099933,
+      "learning_rate": 0.0001411419231518742,
+      "loss": 0.2666,
+      "step": 2294
+    },
+    {
+      "epoch": 1.2864349775784754,
+      "grad_norm": 0.07577632080911738,
+      "learning_rate": 0.0001410824585729521,
+      "loss": 0.2797,
+      "step": 2295
+    },
+    {
+      "epoch": 1.2869955156950672,
+      "grad_norm": 0.07252737827488943,
+      "learning_rate": 0.0001410229765123487,
+      "loss": 0.2752,
+      "step": 2296
+    },
+    {
+      "epoch": 1.2875560538116593,
+      "grad_norm": 0.07270549855949322,
+      "learning_rate": 0.00014096347699537516,
+      "loss": 0.2787,
+      "step": 2297
+    },
+    {
+      "epoch": 1.2881165919282511,
+      "grad_norm": 0.07168693628490974,
+      "learning_rate": 0.0001409039600473501,
+      "loss": 0.2583,
+      "step": 2298
+    },
+    {
+      "epoch": 1.288677130044843,
+      "grad_norm": 0.07304145070152694,
+      "learning_rate": 0.00014084442569359964,
+      "loss": 0.2747,
+      "step": 2299
+    },
+    {
+      "epoch": 1.289237668161435,
+      "grad_norm": 0.07304492367720695,
+      "learning_rate": 0.00014078487395945713,
+      "loss": 0.2793,
+      "step": 2300
+    },
+    {
+      "epoch": 1.2897982062780269,
+      "grad_norm": 0.07261463886806618,
+      "learning_rate": 0.00014072530487026347,
+      "loss": 0.2679,
+      "step": 2301
+    },
+    {
+      "epoch": 1.2903587443946187,
+      "grad_norm": 0.07501550696532452,
+      "learning_rate": 0.00014066571845136692,
+      "loss": 0.2829,
+      "step": 2302
+    },
+    {
+      "epoch": 1.2909192825112108,
+      "grad_norm": 0.0725622550946889,
+      "learning_rate": 0.000140606114728123,
+      "loss": 0.2868,
+      "step": 2303
+    },
+    {
+      "epoch": 1.2914798206278026,
+      "grad_norm": 0.0715450409641427,
+      "learning_rate": 0.00014054649372589482,
+      "loss": 0.271,
+      "step": 2304
+    },
+    {
+      "epoch": 1.2920403587443947,
+      "grad_norm": 0.07332758764704696,
+      "learning_rate": 0.0001404868554700526,
+      "loss": 0.2819,
+      "step": 2305
+    },
+    {
+      "epoch": 1.2926008968609866,
+      "grad_norm": 0.0691637338358271,
+      "learning_rate": 0.00014042719998597409,
+      "loss": 0.2628,
+      "step": 2306
+    },
+    {
+      "epoch": 1.2931614349775784,
+      "grad_norm": 0.07196187566198209,
+      "learning_rate": 0.00014036752729904418,
+      "loss": 0.2773,
+      "step": 2307
+    },
+    {
+      "epoch": 1.2937219730941705,
+      "grad_norm": 0.07556174369975631,
+      "learning_rate": 0.00014030783743465528,
+      "loss": 0.2833,
+      "step": 2308
+    },
+    {
+      "epoch": 1.2942825112107623,
+      "grad_norm": 0.07430317200840218,
+      "learning_rate": 0.00014024813041820699,
+      "loss": 0.2831,
+      "step": 2309
+    },
+    {
+      "epoch": 1.2948430493273544,
+      "grad_norm": 0.07428510115993528,
+      "learning_rate": 0.00014018840627510622,
+      "loss": 0.2674,
+      "step": 2310
+    },
+    {
+      "epoch": 1.2954035874439462,
+      "grad_norm": 0.07122164159460866,
+      "learning_rate": 0.00014012866503076721,
+      "loss": 0.2672,
+      "step": 2311
+    },
+    {
+      "epoch": 1.295964125560538,
+      "grad_norm": 0.0729114329375746,
+      "learning_rate": 0.00014006890671061143,
+      "loss": 0.277,
+      "step": 2312
+    },
+    {
+      "epoch": 1.29652466367713,
+      "grad_norm": 0.07042398703873136,
+      "learning_rate": 0.00014000913134006767,
+      "loss": 0.2559,
+      "step": 2313
+    },
+    {
+      "epoch": 1.297085201793722,
+      "grad_norm": 0.07393397653669982,
+      "learning_rate": 0.00013994933894457192,
+      "loss": 0.2548,
+      "step": 2314
+    },
+    {
+      "epoch": 1.2976457399103138,
+      "grad_norm": 0.07287978252403222,
+      "learning_rate": 0.00013988952954956745,
+      "loss": 0.2757,
+      "step": 2315
+    },
+    {
+      "epoch": 1.298206278026906,
+      "grad_norm": 0.07234603630097806,
+      "learning_rate": 0.0001398297031805047,
+      "loss": 0.2665,
+      "step": 2316
+    },
+    {
+      "epoch": 1.2987668161434978,
+      "grad_norm": 0.07375355118824485,
+      "learning_rate": 0.00013976985986284147,
+      "loss": 0.2733,
+      "step": 2317
+    },
+    {
+      "epoch": 1.2993273542600896,
+      "grad_norm": 0.07087325341820375,
+      "learning_rate": 0.00013970999962204265,
+      "loss": 0.259,
+      "step": 2318
+    },
+    {
+      "epoch": 1.2998878923766817,
+      "grad_norm": 0.07220953544823676,
+      "learning_rate": 0.00013965012248358036,
+      "loss": 0.2755,
+      "step": 2319
+    },
+    {
+      "epoch": 1.3004484304932735,
+      "grad_norm": 0.07119809272659058,
+      "learning_rate": 0.00013959022847293391,
+      "loss": 0.2636,
+      "step": 2320
+    },
+    {
+      "epoch": 1.3010089686098656,
+      "grad_norm": 0.07253157824321244,
+      "learning_rate": 0.00013953031761558982,
+      "loss": 0.2729,
+      "step": 2321
+    },
+    {
+      "epoch": 1.3015695067264574,
+      "grad_norm": 0.07186140529816074,
+      "learning_rate": 0.00013947038993704177,
+      "loss": 0.2741,
+      "step": 2322
+    },
+    {
+      "epoch": 1.3021300448430493,
+      "grad_norm": 0.07143050947797319,
+      "learning_rate": 0.00013941044546279054,
+      "loss": 0.2593,
+      "step": 2323
+    },
+    {
+      "epoch": 1.3026905829596411,
+      "grad_norm": 0.07322050298555188,
+      "learning_rate": 0.0001393504842183441,
+      "loss": 0.2849,
+      "step": 2324
+    },
+    {
+      "epoch": 1.3032511210762332,
+      "grad_norm": 0.07245247220336955,
+      "learning_rate": 0.00013929050622921762,
+      "loss": 0.2681,
+      "step": 2325
+    },
+    {
+      "epoch": 1.303811659192825,
+      "grad_norm": 0.0721492767322242,
+      "learning_rate": 0.0001392305115209333,
+      "loss": 0.2769,
+      "step": 2326
+    },
+    {
+      "epoch": 1.3043721973094171,
+      "grad_norm": 0.07130441768921186,
+      "learning_rate": 0.00013917050011902048,
+      "loss": 0.2734,
+      "step": 2327
+    },
+    {
+      "epoch": 1.304932735426009,
+      "grad_norm": 0.07381937785338555,
+      "learning_rate": 0.0001391104720490156,
+      "loss": 0.2767,
+      "step": 2328
+    },
+    {
+      "epoch": 1.3054932735426008,
+      "grad_norm": 0.07235614217671463,
+      "learning_rate": 0.00013905042733646224,
+      "loss": 0.2711,
+      "step": 2329
+    },
+    {
+      "epoch": 1.3060538116591929,
+      "grad_norm": 0.07019467785727469,
+      "learning_rate": 0.000138990366006911,
+      "loss": 0.2624,
+      "step": 2330
+    },
+    {
+      "epoch": 1.3066143497757847,
+      "grad_norm": 0.07240908628062101,
+      "learning_rate": 0.0001389302880859196,
+      "loss": 0.2868,
+      "step": 2331
+    },
+    {
+      "epoch": 1.3071748878923768,
+      "grad_norm": 0.07257995860684734,
+      "learning_rate": 0.00013887019359905275,
+      "loss": 0.2715,
+      "step": 2332
+    },
+    {
+      "epoch": 1.3077354260089686,
+      "grad_norm": 0.07237788870869241,
+      "learning_rate": 0.0001388100825718823,
+      "loss": 0.2779,
+      "step": 2333
+    },
+    {
+      "epoch": 1.3082959641255605,
+      "grad_norm": 0.07339006898519476,
+      "learning_rate": 0.00013874995502998706,
+      "loss": 0.2797,
+      "step": 2334
+    },
+    {
+      "epoch": 1.3088565022421523,
+      "grad_norm": 0.07075462432399776,
+      "learning_rate": 0.00013868981099895294,
+      "loss": 0.2601,
+      "step": 2335
+    },
+    {
+      "epoch": 1.3094170403587444,
+      "grad_norm": 0.07162694408756295,
+      "learning_rate": 0.0001386296505043728,
+      "loss": 0.263,
+      "step": 2336
+    },
+    {
+      "epoch": 1.3099775784753362,
+      "grad_norm": 0.06924353205055772,
+      "learning_rate": 0.00013856947357184657,
+      "loss": 0.2719,
+      "step": 2337
+    },
+    {
+      "epoch": 1.3105381165919283,
+      "grad_norm": 0.07313976800317197,
+      "learning_rate": 0.00013850928022698112,
+      "loss": 0.2779,
+      "step": 2338
+    },
+    {
+      "epoch": 1.3110986547085202,
+      "grad_norm": 0.07219278412249301,
+      "learning_rate": 0.0001384490704953903,
+      "loss": 0.2662,
+      "step": 2339
+    },
+    {
+      "epoch": 1.311659192825112,
+      "grad_norm": 0.07088047183199304,
+      "learning_rate": 0.00013838884440269496,
+      "loss": 0.2623,
+      "step": 2340
+    },
+    {
+      "epoch": 1.312219730941704,
+      "grad_norm": 0.07202159403822177,
+      "learning_rate": 0.00013832860197452294,
+      "loss": 0.2628,
+      "step": 2341
+    },
+    {
+      "epoch": 1.312780269058296,
+      "grad_norm": 0.07289564709942851,
+      "learning_rate": 0.000138268343236509,
+      "loss": 0.2837,
+      "step": 2342
+    },
+    {
+      "epoch": 1.313340807174888,
+      "grad_norm": 0.06880975974949648,
+      "learning_rate": 0.0001382080682142948,
+      "loss": 0.26,
+      "step": 2343
+    },
+    {
+      "epoch": 1.3139013452914798,
+      "grad_norm": 0.07268727279675526,
+      "learning_rate": 0.000138147776933529,
+      "loss": 0.2749,
+      "step": 2344
+    },
+    {
+      "epoch": 1.3144618834080717,
+      "grad_norm": 0.07258524138923673,
+      "learning_rate": 0.00013808746941986708,
+      "loss": 0.2746,
+      "step": 2345
+    },
+    {
+      "epoch": 1.3150224215246638,
+      "grad_norm": 0.07379419595042286,
+      "learning_rate": 0.00013802714569897162,
+      "loss": 0.2699,
+      "step": 2346
+    },
+    {
+      "epoch": 1.3155829596412556,
+      "grad_norm": 0.07166246045813811,
+      "learning_rate": 0.00013796680579651187,
+      "loss": 0.2713,
+      "step": 2347
+    },
+    {
+      "epoch": 1.3161434977578477,
+      "grad_norm": 0.07269806792322606,
+      "learning_rate": 0.0001379064497381641,
+      "loss": 0.2637,
+      "step": 2348
+    },
+    {
+      "epoch": 1.3167040358744395,
+      "grad_norm": 0.0748217429176605,
+      "learning_rate": 0.0001378460775496114,
+      "loss": 0.2867,
+      "step": 2349
+    },
+    {
+      "epoch": 1.3172645739910314,
+      "grad_norm": 0.07115741698647755,
+      "learning_rate": 0.00013778568925654382,
+      "loss": 0.2629,
+      "step": 2350
+    },
+    {
+      "epoch": 1.3178251121076232,
+      "grad_norm": 0.07347852046495931,
+      "learning_rate": 0.0001377252848846581,
+      "loss": 0.2665,
+      "step": 2351
+    },
+    {
+      "epoch": 1.3183856502242153,
+      "grad_norm": 0.07287568379605906,
+      "learning_rate": 0.00013766486445965795,
+      "loss": 0.273,
+      "step": 2352
+    },
+    {
+      "epoch": 1.3189461883408071,
+      "grad_norm": 0.07487119510405997,
+      "learning_rate": 0.00013760442800725387,
+      "loss": 0.2804,
+      "step": 2353
+    },
+    {
+      "epoch": 1.3195067264573992,
+      "grad_norm": 0.07129724434208816,
+      "learning_rate": 0.00013754397555316322,
+      "loss": 0.268,
+      "step": 2354
+    },
+    {
+      "epoch": 1.320067264573991,
+      "grad_norm": 0.07413545960455435,
+      "learning_rate": 0.00013748350712311004,
+      "loss": 0.2809,
+      "step": 2355
+    },
+    {
+      "epoch": 1.3206278026905829,
+      "grad_norm": 0.07167169675379284,
+      "learning_rate": 0.00013742302274282533,
+      "loss": 0.2781,
+      "step": 2356
+    },
+    {
+      "epoch": 1.321188340807175,
+      "grad_norm": 0.07308218187278753,
+      "learning_rate": 0.00013736252243804677,
+      "loss": 0.2783,
+      "step": 2357
+    },
+    {
+      "epoch": 1.3217488789237668,
+      "grad_norm": 0.07226496753721331,
+      "learning_rate": 0.00013730200623451888,
+      "loss": 0.2732,
+      "step": 2358
+    },
+    {
+      "epoch": 1.3223094170403589,
+      "grad_norm": 0.07190735515069895,
+      "learning_rate": 0.00013724147415799292,
+      "loss": 0.2672,
+      "step": 2359
+    },
+    {
+      "epoch": 1.3228699551569507,
+      "grad_norm": 0.07240624410270693,
+      "learning_rate": 0.00013718092623422686,
+      "loss": 0.2718,
+      "step": 2360
+    },
+    {
+      "epoch": 1.3234304932735426,
+      "grad_norm": 0.07338334184086785,
+      "learning_rate": 0.0001371203624889855,
+      "loss": 0.2784,
+      "step": 2361
+    },
+    {
+      "epoch": 1.3239910313901344,
+      "grad_norm": 0.07082306820891138,
+      "learning_rate": 0.00013705978294804028,
+      "loss": 0.2585,
+      "step": 2362
+    },
+    {
+      "epoch": 1.3245515695067265,
+      "grad_norm": 0.07207401693906636,
+      "learning_rate": 0.0001369991876371695,
+      "loss": 0.2876,
+      "step": 2363
+    },
+    {
+      "epoch": 1.3251121076233183,
+      "grad_norm": 0.0739441911284224,
+      "learning_rate": 0.000136938576582158,
+      "loss": 0.2788,
+      "step": 2364
+    },
+    {
+      "epoch": 1.3256726457399104,
+      "grad_norm": 0.07423962072526538,
+      "learning_rate": 0.0001368779498087974,
+      "loss": 0.2803,
+      "step": 2365
+    },
+    {
+      "epoch": 1.3262331838565022,
+      "grad_norm": 0.07505191493550833,
+      "learning_rate": 0.00013681730734288605,
+      "loss": 0.2706,
+      "step": 2366
+    },
+    {
+      "epoch": 1.326793721973094,
+      "grad_norm": 0.07051531038088685,
+      "learning_rate": 0.0001367566492102289,
+      "loss": 0.2647,
+      "step": 2367
+    },
+    {
+      "epoch": 1.3273542600896862,
+      "grad_norm": 0.07118623141697969,
+      "learning_rate": 0.00013669597543663762,
+      "loss": 0.278,
+      "step": 2368
+    },
+    {
+      "epoch": 1.327914798206278,
+      "grad_norm": 0.07214664899734435,
+      "learning_rate": 0.0001366352860479305,
+      "loss": 0.2586,
+      "step": 2369
+    },
+    {
+      "epoch": 1.32847533632287,
+      "grad_norm": 0.07179236909265237,
+      "learning_rate": 0.00013657458106993258,
+      "loss": 0.2739,
+      "step": 2370
+    },
+    {
+      "epoch": 1.329035874439462,
+      "grad_norm": 0.07101289143586781,
+      "learning_rate": 0.00013651386052847533,
+      "loss": 0.2697,
+      "step": 2371
+    },
+    {
+      "epoch": 1.3295964125560538,
+      "grad_norm": 0.0725617724663863,
+      "learning_rate": 0.000136453124449397,
+      "loss": 0.2604,
+      "step": 2372
+    },
+    {
+      "epoch": 1.3301569506726456,
+      "grad_norm": 0.076506082639709,
+      "learning_rate": 0.00013639237285854243,
+      "loss": 0.2778,
+      "step": 2373
+    },
+    {
+      "epoch": 1.3307174887892377,
+      "grad_norm": 0.07319661031867133,
+      "learning_rate": 0.000136331605781763,
+      "loss": 0.2664,
+      "step": 2374
+    },
+    {
+      "epoch": 1.3312780269058295,
+      "grad_norm": 0.0719402160984244,
+      "learning_rate": 0.00013627082324491678,
+      "loss": 0.2619,
+      "step": 2375
+    },
+    {
+      "epoch": 1.3318385650224216,
+      "grad_norm": 0.07019774909320625,
+      "learning_rate": 0.00013621002527386834,
+      "loss": 0.252,
+      "step": 2376
+    },
+    {
+      "epoch": 1.3323991031390134,
+      "grad_norm": 0.07035640757768981,
+      "learning_rate": 0.00013614921189448879,
+      "loss": 0.2748,
+      "step": 2377
+    },
+    {
+      "epoch": 1.3329596412556053,
+      "grad_norm": 0.07268993860380529,
+      "learning_rate": 0.00013608838313265587,
+      "loss": 0.2734,
+      "step": 2378
+    },
+    {
+      "epoch": 1.3335201793721974,
+      "grad_norm": 0.07116808690159099,
+      "learning_rate": 0.0001360275390142539,
+      "loss": 0.2758,
+      "step": 2379
+    },
+    {
+      "epoch": 1.3340807174887892,
+      "grad_norm": 0.0705630739577653,
+      "learning_rate": 0.0001359666795651736,
+      "loss": 0.2754,
+      "step": 2380
+    },
+    {
+      "epoch": 1.3346412556053813,
+      "grad_norm": 0.07124532253903049,
+      "learning_rate": 0.0001359058048113123,
+      "loss": 0.2669,
+      "step": 2381
+    },
+    {
+      "epoch": 1.3352017937219731,
+      "grad_norm": 0.07120827955893569,
+      "learning_rate": 0.00013584491477857384,
+      "loss": 0.262,
+      "step": 2382
+    },
+    {
+      "epoch": 1.335762331838565,
+      "grad_norm": 0.07179104584486316,
+      "learning_rate": 0.00013578400949286855,
+      "loss": 0.2719,
+      "step": 2383
+    },
+    {
+      "epoch": 1.336322869955157,
+      "grad_norm": 0.07098980417413621,
+      "learning_rate": 0.0001357230889801133,
+      "loss": 0.2617,
+      "step": 2384
+    },
+    {
+      "epoch": 1.3368834080717489,
+      "grad_norm": 0.07208511500860665,
+      "learning_rate": 0.0001356621532662313,
+      "loss": 0.2639,
+      "step": 2385
+    },
+    {
+      "epoch": 1.3374439461883407,
+      "grad_norm": 0.07294348606152207,
+      "learning_rate": 0.00013560120237715242,
+      "loss": 0.2694,
+      "step": 2386
+    },
+    {
+      "epoch": 1.3380044843049328,
+      "grad_norm": 0.072253436486753,
+      "learning_rate": 0.0001355402363388128,
+      "loss": 0.2727,
+      "step": 2387
+    },
+    {
+      "epoch": 1.3385650224215246,
+      "grad_norm": 0.07179010157489436,
+      "learning_rate": 0.00013547925517715519,
+      "loss": 0.2638,
+      "step": 2388
+    },
+    {
+      "epoch": 1.3391255605381165,
+      "grad_norm": 0.07031658021354985,
+      "learning_rate": 0.00013541825891812863,
+      "loss": 0.2732,
+      "step": 2389
+    },
+    {
+      "epoch": 1.3396860986547086,
+      "grad_norm": 0.07361561315970444,
+      "learning_rate": 0.00013535724758768867,
+      "loss": 0.2683,
+      "step": 2390
+    },
+    {
+      "epoch": 1.3402466367713004,
+      "grad_norm": 0.07007627948386362,
+      "learning_rate": 0.00013529622121179733,
+      "loss": 0.2651,
+      "step": 2391
+    },
+    {
+      "epoch": 1.3408071748878925,
+      "grad_norm": 0.07248825788078303,
+      "learning_rate": 0.00013523517981642286,
+      "loss": 0.2757,
+      "step": 2392
+    },
+    {
+      "epoch": 1.3413677130044843,
+      "grad_norm": 0.07486248813509154,
+      "learning_rate": 0.0001351741234275401,
+      "loss": 0.2791,
+      "step": 2393
+    },
+    {
+      "epoch": 1.3419282511210762,
+      "grad_norm": 0.07215523400320231,
+      "learning_rate": 0.0001351130520711301,
+      "loss": 0.2822,
+      "step": 2394
+    },
+    {
+      "epoch": 1.3424887892376682,
+      "grad_norm": 0.07206609859268151,
+      "learning_rate": 0.0001350519657731803,
+      "loss": 0.2719,
+      "step": 2395
+    },
+    {
+      "epoch": 1.34304932735426,
+      "grad_norm": 0.07094754593911295,
+      "learning_rate": 0.00013499086455968467,
+      "loss": 0.2672,
+      "step": 2396
+    },
+    {
+      "epoch": 1.3436098654708521,
+      "grad_norm": 0.07345131253742171,
+      "learning_rate": 0.00013492974845664336,
+      "loss": 0.2646,
+      "step": 2397
+    },
+    {
+      "epoch": 1.344170403587444,
+      "grad_norm": 0.07148448933462602,
+      "learning_rate": 0.00013486861749006286,
+      "loss": 0.2631,
+      "step": 2398
+    },
+    {
+      "epoch": 1.3447309417040358,
+      "grad_norm": 0.0734079306257349,
+      "learning_rate": 0.0001348074716859561,
+      "loss": 0.2803,
+      "step": 2399
+    },
+    {
+      "epoch": 1.3452914798206277,
+      "grad_norm": 0.07510957327668555,
+      "learning_rate": 0.0001347463110703422,
+      "loss": 0.2732,
+      "step": 2400
+    },
+    {
+      "epoch": 1.3458520179372198,
+      "grad_norm": 0.07361881542352745,
+      "learning_rate": 0.00013468513566924662,
+      "loss": 0.2682,
+      "step": 2401
+    },
+    {
+      "epoch": 1.3464125560538116,
+      "grad_norm": 0.07337817081027358,
+      "learning_rate": 0.00013462394550870115,
+      "loss": 0.2746,
+      "step": 2402
+    },
+    {
+      "epoch": 1.3469730941704037,
+      "grad_norm": 0.07321876065712256,
+      "learning_rate": 0.00013456274061474384,
+      "loss": 0.2618,
+      "step": 2403
+    },
+    {
+      "epoch": 1.3475336322869955,
+      "grad_norm": 0.07395144294198633,
+      "learning_rate": 0.00013450152101341896,
+      "loss": 0.2704,
+      "step": 2404
+    },
+    {
+      "epoch": 1.3480941704035874,
+      "grad_norm": 0.07216160241493295,
+      "learning_rate": 0.00013444028673077716,
+      "loss": 0.2855,
+      "step": 2405
+    },
+    {
+      "epoch": 1.3486547085201794,
+      "grad_norm": 0.06951922593507989,
+      "learning_rate": 0.0001343790377928752,
+      "loss": 0.2627,
+      "step": 2406
+    },
+    {
+      "epoch": 1.3492152466367713,
+      "grad_norm": 0.07141496023438036,
+      "learning_rate": 0.00013431777422577614,
+      "loss": 0.2676,
+      "step": 2407
+    },
+    {
+      "epoch": 1.3497757847533634,
+      "grad_norm": 0.07112079206079516,
+      "learning_rate": 0.00013425649605554928,
+      "loss": 0.2668,
+      "step": 2408
+    },
+    {
+      "epoch": 1.3503363228699552,
+      "grad_norm": 0.07165333061044075,
+      "learning_rate": 0.0001341952033082701,
+      "loss": 0.2698,
+      "step": 2409
+    },
+    {
+      "epoch": 1.350896860986547,
+      "grad_norm": 0.0731320853414597,
+      "learning_rate": 0.00013413389601002034,
+      "loss": 0.2647,
+      "step": 2410
+    },
+    {
+      "epoch": 1.351457399103139,
+      "grad_norm": 0.07198129210042335,
+      "learning_rate": 0.0001340725741868878,
+      "loss": 0.2662,
+      "step": 2411
+    },
+    {
+      "epoch": 1.352017937219731,
+      "grad_norm": 0.0684396964613639,
+      "learning_rate": 0.00013401123786496664,
+      "loss": 0.2539,
+      "step": 2412
+    },
+    {
+      "epoch": 1.3525784753363228,
+      "grad_norm": 0.07309220050943435,
+      "learning_rate": 0.00013394988707035707,
+      "loss": 0.2577,
+      "step": 2413
+    },
+    {
+      "epoch": 1.3531390134529149,
+      "grad_norm": 0.07483948334841695,
+      "learning_rate": 0.00013388852182916544,
+      "loss": 0.2702,
+      "step": 2414
+    },
+    {
+      "epoch": 1.3536995515695067,
+      "grad_norm": 0.07026305718568777,
+      "learning_rate": 0.00013382714216750438,
+      "loss": 0.2591,
+      "step": 2415
+    },
+    {
+      "epoch": 1.3542600896860986,
+      "grad_norm": 0.07500682600556109,
+      "learning_rate": 0.00013376574811149253,
+      "loss": 0.2829,
+      "step": 2416
+    },
+    {
+      "epoch": 1.3548206278026906,
+      "grad_norm": 0.07317671520929299,
+      "learning_rate": 0.00013370433968725468,
+      "loss": 0.2786,
+      "step": 2417
+    },
+    {
+      "epoch": 1.3553811659192825,
+      "grad_norm": 0.07092215065383145,
+      "learning_rate": 0.00013364291692092182,
+      "loss": 0.2724,
+      "step": 2418
+    },
+    {
+      "epoch": 1.3559417040358746,
+      "grad_norm": 0.0722465926401936,
+      "learning_rate": 0.00013358147983863087,
+      "loss": 0.276,
+      "step": 2419
+    },
+    {
+      "epoch": 1.3565022421524664,
+      "grad_norm": 0.0689147359853441,
+      "learning_rate": 0.00013352002846652504,
+      "loss": 0.262,
+      "step": 2420
+    },
+    {
+      "epoch": 1.3570627802690582,
+      "grad_norm": 0.06956653911163503,
+      "learning_rate": 0.00013345856283075347,
+      "loss": 0.2684,
+      "step": 2421
+    },
+    {
+      "epoch": 1.35762331838565,
+      "grad_norm": 0.07095296721528385,
+      "learning_rate": 0.00013339708295747146,
+      "loss": 0.2624,
+      "step": 2422
+    },
+    {
+      "epoch": 1.3581838565022422,
+      "grad_norm": 0.07132575762686803,
+      "learning_rate": 0.0001333355888728403,
+      "loss": 0.2811,
+      "step": 2423
+    },
+    {
+      "epoch": 1.358744394618834,
+      "grad_norm": 0.07301929011953896,
+      "learning_rate": 0.00013327408060302738,
+      "loss": 0.2703,
+      "step": 2424
+    },
+    {
+      "epoch": 1.359304932735426,
+      "grad_norm": 0.07186179738526437,
+      "learning_rate": 0.00013321255817420614,
+      "loss": 0.2708,
+      "step": 2425
+    },
+    {
+      "epoch": 1.359865470852018,
+      "grad_norm": 0.07158317572330604,
+      "learning_rate": 0.000133151021612556,
+      "loss": 0.2715,
+      "step": 2426
+    },
+    {
+      "epoch": 1.3604260089686098,
+      "grad_norm": 0.07282108957574474,
+      "learning_rate": 0.00013308947094426237,
+      "loss": 0.278,
+      "step": 2427
+    },
+    {
+      "epoch": 1.3609865470852018,
+      "grad_norm": 0.07211048808596557,
+      "learning_rate": 0.00013302790619551674,
+      "loss": 0.2582,
+      "step": 2428
+    },
+    {
+      "epoch": 1.3615470852017937,
+      "grad_norm": 0.07199512540150688,
+      "learning_rate": 0.00013296632739251649,
+      "loss": 0.2616,
+      "step": 2429
+    },
+    {
+      "epoch": 1.3621076233183858,
+      "grad_norm": 0.07376143293365058,
+      "learning_rate": 0.00013290473456146513,
+      "loss": 0.271,
+      "step": 2430
+    },
+    {
+      "epoch": 1.3626681614349776,
+      "grad_norm": 0.07289892751397548,
+      "learning_rate": 0.00013284312772857197,
+      "loss": 0.2709,
+      "step": 2431
+    },
+    {
+      "epoch": 1.3632286995515694,
+      "grad_norm": 0.07210804572294657,
+      "learning_rate": 0.00013278150692005243,
+      "loss": 0.2838,
+      "step": 2432
+    },
+    {
+      "epoch": 1.3637892376681615,
+      "grad_norm": 0.07043118964102611,
+      "learning_rate": 0.0001327198721621278,
+      "loss": 0.2609,
+      "step": 2433
+    },
+    {
+      "epoch": 1.3643497757847534,
+      "grad_norm": 0.07031977851384845,
+      "learning_rate": 0.00013265822348102526,
+      "loss": 0.2752,
+      "step": 2434
+    },
+    {
+      "epoch": 1.3649103139013454,
+      "grad_norm": 0.0726635579400134,
+      "learning_rate": 0.000132596560902978,
+      "loss": 0.2747,
+      "step": 2435
+    },
+    {
+      "epoch": 1.3654708520179373,
+      "grad_norm": 0.07282388910028605,
+      "learning_rate": 0.00013253488445422507,
+      "loss": 0.2769,
+      "step": 2436
+    },
+    {
+      "epoch": 1.3660313901345291,
+      "grad_norm": 0.07105098521583782,
+      "learning_rate": 0.00013247319416101146,
+      "loss": 0.2562,
+      "step": 2437
+    },
+    {
+      "epoch": 1.366591928251121,
+      "grad_norm": 0.0722914203984814,
+      "learning_rate": 0.00013241149004958807,
+      "loss": 0.277,
+      "step": 2438
+    },
+    {
+      "epoch": 1.367152466367713,
+      "grad_norm": 0.07130305361271941,
+      "learning_rate": 0.00013234977214621158,
+      "loss": 0.2679,
+      "step": 2439
+    },
+    {
+      "epoch": 1.3677130044843049,
+      "grad_norm": 0.07382272031436721,
+      "learning_rate": 0.00013228804047714463,
+      "loss": 0.2838,
+      "step": 2440
+    },
+    {
+      "epoch": 1.368273542600897,
+      "grad_norm": 0.07168330955222213,
+      "learning_rate": 0.00013222629506865572,
+      "loss": 0.2589,
+      "step": 2441
+    },
+    {
+      "epoch": 1.3688340807174888,
+      "grad_norm": 0.07176541308751692,
+      "learning_rate": 0.00013216453594701912,
+      "loss": 0.2676,
+      "step": 2442
+    },
+    {
+      "epoch": 1.3693946188340806,
+      "grad_norm": 0.07160344434933923,
+      "learning_rate": 0.000132102763138515,
+      "loss": 0.2722,
+      "step": 2443
+    },
+    {
+      "epoch": 1.3699551569506727,
+      "grad_norm": 0.07152436426939439,
+      "learning_rate": 0.00013204097666942932,
+      "loss": 0.2858,
+      "step": 2444
+    },
+    {
+      "epoch": 1.3705156950672646,
+      "grad_norm": 0.07669489731715146,
+      "learning_rate": 0.0001319791765660539,
+      "loss": 0.2747,
+      "step": 2445
+    },
+    {
+      "epoch": 1.3710762331838566,
+      "grad_norm": 0.07425976800563304,
+      "learning_rate": 0.00013191736285468638,
+      "loss": 0.2834,
+      "step": 2446
+    },
+    {
+      "epoch": 1.3716367713004485,
+      "grad_norm": 0.07296885862444086,
+      "learning_rate": 0.00013185553556163,
+      "loss": 0.2713,
+      "step": 2447
+    },
+    {
+      "epoch": 1.3721973094170403,
+      "grad_norm": 0.07428666702438311,
+      "learning_rate": 0.00013179369471319404,
+      "loss": 0.2842,
+      "step": 2448
+    },
+    {
+      "epoch": 1.3727578475336322,
+      "grad_norm": 0.07296014058514794,
+      "learning_rate": 0.00013173184033569342,
+      "loss": 0.2745,
+      "step": 2449
+    },
+    {
+      "epoch": 1.3733183856502242,
+      "grad_norm": 0.07174745038851872,
+      "learning_rate": 0.00013166997245544877,
+      "loss": 0.2783,
+      "step": 2450
+    },
+    {
+      "epoch": 1.373878923766816,
+      "grad_norm": 0.07147497552727385,
+      "learning_rate": 0.00013160809109878655,
+      "loss": 0.2721,
+      "step": 2451
+    },
+    {
+      "epoch": 1.3744394618834082,
+      "grad_norm": 0.07150141944074127,
+      "learning_rate": 0.00013154619629203893,
+      "loss": 0.2598,
+      "step": 2452
+    },
+    {
+      "epoch": 1.375,
+      "grad_norm": 0.07181043821190135,
+      "learning_rate": 0.00013148428806154382,
+      "loss": 0.2667,
+      "step": 2453
+    },
+    {
+      "epoch": 1.3755605381165918,
+      "grad_norm": 0.07051746386550256,
+      "learning_rate": 0.0001314223664336448,
+      "loss": 0.2742,
+      "step": 2454
+    },
+    {
+      "epoch": 1.376121076233184,
+      "grad_norm": 0.07253062832481688,
+      "learning_rate": 0.00013136043143469116,
+      "loss": 0.2634,
+      "step": 2455
+    },
+    {
+      "epoch": 1.3766816143497758,
+      "grad_norm": 0.07432425293979825,
+      "learning_rate": 0.0001312984830910379,
+      "loss": 0.2663,
+      "step": 2456
+    },
+    {
+      "epoch": 1.3772421524663678,
+      "grad_norm": 0.0711176297415943,
+      "learning_rate": 0.00013123652142904574,
+      "loss": 0.2719,
+      "step": 2457
+    },
+    {
+      "epoch": 1.3778026905829597,
+      "grad_norm": 0.07257757758574657,
+      "learning_rate": 0.00013117454647508094,
+      "loss": 0.2681,
+      "step": 2458
+    },
+    {
+      "epoch": 1.3783632286995515,
+      "grad_norm": 0.0708100378992227,
+      "learning_rate": 0.00013111255825551556,
+      "loss": 0.2698,
+      "step": 2459
+    },
+    {
+      "epoch": 1.3789237668161434,
+      "grad_norm": 0.06870018997269857,
+      "learning_rate": 0.0001310505567967272,
+      "loss": 0.2718,
+      "step": 2460
+    },
+    {
+      "epoch": 1.3794843049327354,
+      "grad_norm": 0.07203519243056784,
+      "learning_rate": 0.00013098854212509917,
+      "loss": 0.2796,
+      "step": 2461
+    },
+    {
+      "epoch": 1.3800448430493273,
+      "grad_norm": 0.07187680311401265,
+      "learning_rate": 0.00013092651426702034,
+      "loss": 0.2684,
+      "step": 2462
+    },
+    {
+      "epoch": 1.3806053811659194,
+      "grad_norm": 0.06981221618911998,
+      "learning_rate": 0.0001308644732488852,
+      "loss": 0.2496,
+      "step": 2463
+    },
+    {
+      "epoch": 1.3811659192825112,
+      "grad_norm": 0.0731266258637114,
+      "learning_rate": 0.00013080241909709387,
+      "loss": 0.2763,
+      "step": 2464
+    },
+    {
+      "epoch": 1.381726457399103,
+      "grad_norm": 0.07584163817780647,
+      "learning_rate": 0.00013074035183805209,
+      "loss": 0.2723,
+      "step": 2465
+    },
+    {
+      "epoch": 1.3822869955156951,
+      "grad_norm": 0.07430429344317484,
+      "learning_rate": 0.00013067827149817112,
+      "loss": 0.2689,
+      "step": 2466
+    },
+    {
+      "epoch": 1.382847533632287,
+      "grad_norm": 0.07265967034126478,
+      "learning_rate": 0.00013061617810386774,
+      "loss": 0.2596,
+      "step": 2467
+    },
+    {
+      "epoch": 1.383408071748879,
+      "grad_norm": 0.07203365339853766,
+      "learning_rate": 0.00013055407168156437,
+      "loss": 0.2613,
+      "step": 2468
+    },
+    {
+      "epoch": 1.3839686098654709,
+      "grad_norm": 0.06996162619517453,
+      "learning_rate": 0.00013049195225768898,
+      "loss": 0.2726,
+      "step": 2469
+    },
+    {
+      "epoch": 1.3845291479820627,
+      "grad_norm": 0.07104204707391903,
+      "learning_rate": 0.00013042981985867503,
+      "loss": 0.2646,
+      "step": 2470
+    },
+    {
+      "epoch": 1.3850896860986546,
+      "grad_norm": 0.07096724598851205,
+      "learning_rate": 0.00013036767451096148,
+      "loss": 0.2522,
+      "step": 2471
+    },
+    {
+      "epoch": 1.3856502242152466,
+      "grad_norm": 0.07160993411400399,
+      "learning_rate": 0.00013030551624099287,
+      "loss": 0.2751,
+      "step": 2472
+    },
+    {
+      "epoch": 1.3862107623318385,
+      "grad_norm": 0.07107463944019773,
+      "learning_rate": 0.0001302433450752192,
+      "loss": 0.2805,
+      "step": 2473
+    },
+    {
+      "epoch": 1.3867713004484306,
+      "grad_norm": 0.07394302746868758,
+      "learning_rate": 0.0001301811610400959,
+      "loss": 0.3,
+      "step": 2474
+    },
+    {
+      "epoch": 1.3873318385650224,
+      "grad_norm": 0.07500896698184124,
+      "learning_rate": 0.00013011896416208405,
+      "loss": 0.261,
+      "step": 2475
+    },
+    {
+      "epoch": 1.3878923766816142,
+      "grad_norm": 0.07284475433767182,
+      "learning_rate": 0.00013005675446764998,
+      "loss": 0.279,
+      "step": 2476
+    },
+    {
+      "epoch": 1.3884529147982063,
+      "grad_norm": 0.07059100190411745,
+      "learning_rate": 0.00012999453198326557,
+      "loss": 0.2681,
+      "step": 2477
+    },
+    {
+      "epoch": 1.3890134529147982,
+      "grad_norm": 0.07201489034288251,
+      "learning_rate": 0.00012993229673540822,
+      "loss": 0.2748,
+      "step": 2478
+    },
+    {
+      "epoch": 1.3895739910313902,
+      "grad_norm": 0.06947626310788592,
+      "learning_rate": 0.0001298700487505606,
+      "loss": 0.2692,
+      "step": 2479
+    },
+    {
+      "epoch": 1.390134529147982,
+      "grad_norm": 0.07303319834804392,
+      "learning_rate": 0.000129807788055211,
+      "loss": 0.2737,
+      "step": 2480
+    },
+    {
+      "epoch": 1.390695067264574,
+      "grad_norm": 0.07431794317957002,
+      "learning_rate": 0.0001297455146758529,
+      "loss": 0.2615,
+      "step": 2481
+    },
+    {
+      "epoch": 1.391255605381166,
+      "grad_norm": 0.07137612781800744,
+      "learning_rate": 0.00012968322863898533,
+      "loss": 0.262,
+      "step": 2482
+    },
+    {
+      "epoch": 1.3918161434977578,
+      "grad_norm": 0.07266858820294593,
+      "learning_rate": 0.00012962092997111265,
+      "loss": 0.2845,
+      "step": 2483
+    },
+    {
+      "epoch": 1.39237668161435,
+      "grad_norm": 0.07229964808963044,
+      "learning_rate": 0.0001295586186987446,
+      "loss": 0.2717,
+      "step": 2484
+    },
+    {
+      "epoch": 1.3929372197309418,
+      "grad_norm": 0.07265995318964269,
+      "learning_rate": 0.0001294962948483963,
+      "loss": 0.264,
+      "step": 2485
+    },
+    {
+      "epoch": 1.3934977578475336,
+      "grad_norm": 0.07191498135629393,
+      "learning_rate": 0.00012943395844658821,
+      "loss": 0.2534,
+      "step": 2486
+    },
+    {
+      "epoch": 1.3940582959641254,
+      "grad_norm": 0.07209557545021433,
+      "learning_rate": 0.0001293716095198461,
+      "loss": 0.2685,
+      "step": 2487
+    },
+    {
+      "epoch": 1.3946188340807175,
+      "grad_norm": 0.07239319810802483,
+      "learning_rate": 0.00012930924809470115,
+      "loss": 0.2682,
+      "step": 2488
+    },
+    {
+      "epoch": 1.3951793721973094,
+      "grad_norm": 0.07657128034518046,
+      "learning_rate": 0.00012924687419768976,
+      "loss": 0.2841,
+      "step": 2489
+    },
+    {
+      "epoch": 1.3957399103139014,
+      "grad_norm": 0.07293139721131948,
+      "learning_rate": 0.0001291844878553537,
+      "loss": 0.2743,
+      "step": 2490
+    },
+    {
+      "epoch": 1.3963004484304933,
+      "grad_norm": 0.07139628602035565,
+      "learning_rate": 0.00012912208909424006,
+      "loss": 0.2649,
+      "step": 2491
+    },
+    {
+      "epoch": 1.3968609865470851,
+      "grad_norm": 0.0713037784138301,
+      "learning_rate": 0.00012905967794090114,
+      "loss": 0.2672,
+      "step": 2492
+    },
+    {
+      "epoch": 1.3974215246636772,
+      "grad_norm": 0.07197814435876061,
+      "learning_rate": 0.00012899725442189457,
+      "loss": 0.2792,
+      "step": 2493
+    },
+    {
+      "epoch": 1.397982062780269,
+      "grad_norm": 0.07242621985982685,
+      "learning_rate": 0.00012893481856378317,
+      "loss": 0.2821,
+      "step": 2494
+    },
+    {
+      "epoch": 1.398542600896861,
+      "grad_norm": 0.07114519859944791,
+      "learning_rate": 0.00012887237039313514,
+      "loss": 0.2684,
+      "step": 2495
+    },
+    {
+      "epoch": 1.399103139013453,
+      "grad_norm": 0.07012615325623704,
+      "learning_rate": 0.00012880990993652377,
+      "loss": 0.2417,
+      "step": 2496
+    },
+    {
+      "epoch": 1.3996636771300448,
+      "grad_norm": 0.07294149638436452,
+      "learning_rate": 0.00012874743722052768,
+      "loss": 0.2749,
+      "step": 2497
+    },
+    {
+      "epoch": 1.4002242152466366,
+      "grad_norm": 0.07591298177509882,
+      "learning_rate": 0.00012868495227173068,
+      "loss": 0.2781,
+      "step": 2498
+    },
+    {
+      "epoch": 1.4007847533632287,
+      "grad_norm": 0.07304669174863514,
+      "learning_rate": 0.0001286224551167218,
+      "loss": 0.288,
+      "step": 2499
+    },
+    {
+      "epoch": 1.4013452914798206,
+      "grad_norm": 0.07383601777071674,
+      "learning_rate": 0.00012855994578209526,
+      "loss": 0.2607,
+      "step": 2500
+    },
+    {
+      "epoch": 1.4019058295964126,
+      "grad_norm": 0.07508444870924733,
+      "learning_rate": 0.00012849742429445034,
+      "loss": 0.2819,
+      "step": 2501
+    },
+    {
+      "epoch": 1.4024663677130045,
+      "grad_norm": 0.07250717828673553,
+      "learning_rate": 0.0001284348906803917,
+      "loss": 0.2658,
+      "step": 2502
+    },
+    {
+      "epoch": 1.4030269058295963,
+      "grad_norm": 0.07236648227488171,
+      "learning_rate": 0.000128372344966529,
+      "loss": 0.2692,
+      "step": 2503
+    },
+    {
+      "epoch": 1.4035874439461884,
+      "grad_norm": 0.06937327028257002,
+      "learning_rate": 0.00012830978717947718,
+      "loss": 0.2588,
+      "step": 2504
+    },
+    {
+      "epoch": 1.4041479820627802,
+      "grad_norm": 0.0691096829483645,
+      "learning_rate": 0.00012824721734585622,
+      "loss": 0.277,
+      "step": 2505
+    },
+    {
+      "epoch": 1.4047085201793723,
+      "grad_norm": 0.07041324357200651,
+      "learning_rate": 0.00012818463549229121,
+      "loss": 0.2714,
+      "step": 2506
+    },
+    {
+      "epoch": 1.4052690582959642,
+      "grad_norm": 0.07115900727837653,
+      "learning_rate": 0.00012812204164541245,
+      "loss": 0.2712,
+      "step": 2507
+    },
+    {
+      "epoch": 1.405829596412556,
+      "grad_norm": 0.07322315484925514,
+      "learning_rate": 0.00012805943583185525,
+      "loss": 0.2706,
+      "step": 2508
+    },
+    {
+      "epoch": 1.4063901345291479,
+      "grad_norm": 0.0702243899051264,
+      "learning_rate": 0.00012799681807826004,
+      "loss": 0.2638,
+      "step": 2509
+    },
+    {
+      "epoch": 1.40695067264574,
+      "grad_norm": 0.07119816545279681,
+      "learning_rate": 0.0001279341884112724,
+      "loss": 0.2767,
+      "step": 2510
+    },
+    {
+      "epoch": 1.4075112107623318,
+      "grad_norm": 0.07027483611000442,
+      "learning_rate": 0.0001278715468575429,
+      "loss": 0.2634,
+      "step": 2511
+    },
+    {
+      "epoch": 1.4080717488789238,
+      "grad_norm": 0.07416164006334076,
+      "learning_rate": 0.00012780889344372718,
+      "loss": 0.2767,
+      "step": 2512
+    },
+    {
+      "epoch": 1.4086322869955157,
+      "grad_norm": 0.07371935302881523,
+      "learning_rate": 0.00012774622819648597,
+      "loss": 0.2747,
+      "step": 2513
+    },
+    {
+      "epoch": 1.4091928251121075,
+      "grad_norm": 0.07491047306138772,
+      "learning_rate": 0.00012768355114248494,
+      "loss": 0.278,
+      "step": 2514
+    },
+    {
+      "epoch": 1.4097533632286996,
+      "grad_norm": 0.07330923773402105,
+      "learning_rate": 0.0001276208623083949,
+      "loss": 0.2779,
+      "step": 2515
+    },
+    {
+      "epoch": 1.4103139013452914,
+      "grad_norm": 0.07084274284445143,
+      "learning_rate": 0.00012755816172089164,
+      "loss": 0.2668,
+      "step": 2516
+    },
+    {
+      "epoch": 1.4108744394618835,
+      "grad_norm": 0.07244971643006003,
+      "learning_rate": 0.00012749544940665586,
+      "loss": 0.2762,
+      "step": 2517
+    },
+    {
+      "epoch": 1.4114349775784754,
+      "grad_norm": 0.07143764784972916,
+      "learning_rate": 0.00012743272539237333,
+      "loss": 0.2678,
+      "step": 2518
+    },
+    {
+      "epoch": 1.4119955156950672,
+      "grad_norm": 0.07111406798867922,
+      "learning_rate": 0.00012736998970473487,
+      "loss": 0.2773,
+      "step": 2519
+    },
+    {
+      "epoch": 1.4125560538116593,
+      "grad_norm": 0.07067982889597485,
+      "learning_rate": 0.00012730724237043615,
+      "loss": 0.2558,
+      "step": 2520
+    },
+    {
+      "epoch": 1.4131165919282511,
+      "grad_norm": 0.07053306522161905,
+      "learning_rate": 0.00012724448341617776,
+      "loss": 0.2609,
+      "step": 2521
+    },
+    {
+      "epoch": 1.413677130044843,
+      "grad_norm": 0.07151634488386535,
+      "learning_rate": 0.00012718171286866538,
+      "loss": 0.2818,
+      "step": 2522
+    },
+    {
+      "epoch": 1.414237668161435,
+      "grad_norm": 0.06953363186812431,
+      "learning_rate": 0.00012711893075460957,
+      "loss": 0.2667,
+      "step": 2523
+    },
+    {
+      "epoch": 1.4147982062780269,
+      "grad_norm": 0.07024275109725835,
+      "learning_rate": 0.00012705613710072575,
+      "loss": 0.2721,
+      "step": 2524
+    },
+    {
+      "epoch": 1.4153587443946187,
+      "grad_norm": 0.06947733509275524,
+      "learning_rate": 0.0001269933319337343,
+      "loss": 0.2525,
+      "step": 2525
+    },
+    {
+      "epoch": 1.4159192825112108,
+      "grad_norm": 0.07274290192513362,
+      "learning_rate": 0.00012693051528036051,
+      "loss": 0.2674,
+      "step": 2526
+    },
+    {
+      "epoch": 1.4164798206278026,
+      "grad_norm": 0.07325339688881154,
+      "learning_rate": 0.00012686768716733453,
+      "loss": 0.2783,
+      "step": 2527
+    },
+    {
+      "epoch": 1.4170403587443947,
+      "grad_norm": 0.06868100889040213,
+      "learning_rate": 0.0001268048476213914,
+      "loss": 0.2537,
+      "step": 2528
+    },
+    {
+      "epoch": 1.4176008968609866,
+      "grad_norm": 0.06976187011388851,
+      "learning_rate": 0.000126741996669271,
+      "loss": 0.2703,
+      "step": 2529
+    },
+    {
+      "epoch": 1.4181614349775784,
+      "grad_norm": 0.07257196857537393,
+      "learning_rate": 0.0001266791343377181,
+      "loss": 0.2663,
+      "step": 2530
+    },
+    {
+      "epoch": 1.4187219730941705,
+      "grad_norm": 0.07363162939396198,
+      "learning_rate": 0.0001266162606534823,
+      "loss": 0.2704,
+      "step": 2531
+    },
+    {
+      "epoch": 1.4192825112107623,
+      "grad_norm": 0.07297789157037428,
+      "learning_rate": 0.00012655337564331805,
+      "loss": 0.2682,
+      "step": 2532
+    },
+    {
+      "epoch": 1.4198430493273544,
+      "grad_norm": 0.07347585669911441,
+      "learning_rate": 0.0001264904793339846,
+      "loss": 0.2738,
+      "step": 2533
+    },
+    {
+      "epoch": 1.4204035874439462,
+      "grad_norm": 0.07234283051398997,
+      "learning_rate": 0.00012642757175224595,
+      "loss": 0.2645,
+      "step": 2534
+    },
+    {
+      "epoch": 1.420964125560538,
+      "grad_norm": 0.07127456429361785,
+      "learning_rate": 0.000126364652924871,
+      "loss": 0.2506,
+      "step": 2535
+    },
+    {
+      "epoch": 1.42152466367713,
+      "grad_norm": 0.07176298472511297,
+      "learning_rate": 0.0001263017228786334,
+      "loss": 0.2691,
+      "step": 2536
+    },
+    {
+      "epoch": 1.422085201793722,
+      "grad_norm": 0.07257759103677881,
+      "learning_rate": 0.0001262387816403115,
+      "loss": 0.2707,
+      "step": 2537
+    },
+    {
+      "epoch": 1.4226457399103138,
+      "grad_norm": 0.07112861165554513,
+      "learning_rate": 0.00012617582923668853,
+      "loss": 0.2711,
+      "step": 2538
+    },
+    {
+      "epoch": 1.423206278026906,
+      "grad_norm": 0.07450628624044056,
+      "learning_rate": 0.0001261128656945524,
+      "loss": 0.2669,
+      "step": 2539
+    },
+    {
+      "epoch": 1.4237668161434978,
+      "grad_norm": 0.07405807598729093,
+      "learning_rate": 0.0001260498910406958,
+      "loss": 0.2754,
+      "step": 2540
+    },
+    {
+      "epoch": 1.4243273542600896,
+      "grad_norm": 0.07248947588583825,
+      "learning_rate": 0.00012598690530191608,
+      "loss": 0.2776,
+      "step": 2541
+    },
+    {
+      "epoch": 1.4248878923766817,
+      "grad_norm": 0.07310931331855644,
+      "learning_rate": 0.00012592390850501537,
+      "loss": 0.2632,
+      "step": 2542
+    },
+    {
+      "epoch": 1.4254484304932735,
+      "grad_norm": 0.07061925291194063,
+      "learning_rate": 0.00012586090067680047,
+      "loss": 0.2632,
+      "step": 2543
+    },
+    {
+      "epoch": 1.4260089686098656,
+      "grad_norm": 0.07131670537493585,
+      "learning_rate": 0.00012579788184408295,
+      "loss": 0.2809,
+      "step": 2544
+    },
+    {
+      "epoch": 1.4265695067264574,
+      "grad_norm": 0.07106765132212761,
+      "learning_rate": 0.00012573485203367895,
+      "loss": 0.2518,
+      "step": 2545
+    },
+    {
+      "epoch": 1.4271300448430493,
+      "grad_norm": 0.07117179198796343,
+      "learning_rate": 0.00012567181127240933,
+      "loss": 0.2665,
+      "step": 2546
+    },
+    {
+      "epoch": 1.4276905829596411,
+      "grad_norm": 0.07266380140678072,
+      "learning_rate": 0.00012560875958709963,
+      "loss": 0.2749,
+      "step": 2547
+    },
+    {
+      "epoch": 1.4282511210762332,
+      "grad_norm": 0.0722153633269687,
+      "learning_rate": 0.00012554569700458002,
+      "loss": 0.2721,
+      "step": 2548
+    },
+    {
+      "epoch": 1.428811659192825,
+      "grad_norm": 0.06952925783993638,
+      "learning_rate": 0.00012548262355168533,
+      "loss": 0.2715,
+      "step": 2549
+    },
+    {
+      "epoch": 1.4293721973094171,
+      "grad_norm": 0.07215173118012304,
+      "learning_rate": 0.000125419539255255,
+      "loss": 0.2763,
+      "step": 2550
+    },
+    {
+      "epoch": 1.429932735426009,
+      "grad_norm": 0.07150045454122866,
+      "learning_rate": 0.000125356444142133,
+      "loss": 0.2676,
+      "step": 2551
+    },
+    {
+      "epoch": 1.4304932735426008,
+      "grad_norm": 0.07037204873905802,
+      "learning_rate": 0.00012529333823916807,
+      "loss": 0.2846,
+      "step": 2552
+    },
+    {
+      "epoch": 1.4310538116591929,
+      "grad_norm": 0.06942295359738748,
+      "learning_rate": 0.00012523022157321346,
+      "loss": 0.2693,
+      "step": 2553
+    },
+    {
+      "epoch": 1.4316143497757847,
+      "grad_norm": 0.07052826277122247,
+      "learning_rate": 0.00012516709417112693,
+      "loss": 0.2789,
+      "step": 2554
+    },
+    {
+      "epoch": 1.4321748878923768,
+      "grad_norm": 0.06766539896480486,
+      "learning_rate": 0.00012510395605977087,
+      "loss": 0.2757,
+      "step": 2555
+    },
+    {
+      "epoch": 1.4327354260089686,
+      "grad_norm": 0.07142433812281358,
+      "learning_rate": 0.00012504080726601232,
+      "loss": 0.2619,
+      "step": 2556
+    },
+    {
+      "epoch": 1.4332959641255605,
+      "grad_norm": 0.07044191277742376,
+      "learning_rate": 0.0001249776478167227,
+      "loss": 0.2825,
+      "step": 2557
+    },
+    {
+      "epoch": 1.4338565022421523,
+      "grad_norm": 0.0701961555356744,
+      "learning_rate": 0.00012491447773877804,
+      "loss": 0.2756,
+      "step": 2558
+    },
+    {
+      "epoch": 1.4344170403587444,
+      "grad_norm": 0.07089835656914216,
+      "learning_rate": 0.00012485129705905893,
+      "loss": 0.2662,
+      "step": 2559
+    },
+    {
+      "epoch": 1.4349775784753362,
+      "grad_norm": 0.07252791283428492,
+      "learning_rate": 0.0001247881058044504,
+      "loss": 0.271,
+      "step": 2560
+    },
+    {
+      "epoch": 1.4355381165919283,
+      "grad_norm": 0.07217047354376249,
+      "learning_rate": 0.00012472490400184205,
+      "loss": 0.2622,
+      "step": 2561
+    },
+    {
+      "epoch": 1.4360986547085202,
+      "grad_norm": 0.07126349475146648,
+      "learning_rate": 0.0001246616916781279,
+      "loss": 0.2703,
+      "step": 2562
+    },
+    {
+      "epoch": 1.436659192825112,
+      "grad_norm": 0.07133291996222735,
+      "learning_rate": 0.00012459846886020643,
+      "loss": 0.2612,
+      "step": 2563
+    },
+    {
+      "epoch": 1.437219730941704,
+      "grad_norm": 0.06766315324492028,
+      "learning_rate": 0.00012453523557498075,
+      "loss": 0.2591,
+      "step": 2564
+    },
+    {
+      "epoch": 1.437780269058296,
+      "grad_norm": 0.07217502057141792,
+      "learning_rate": 0.00012447199184935823,
+      "loss": 0.2815,
+      "step": 2565
+    },
+    {
+      "epoch": 1.438340807174888,
+      "grad_norm": 0.07416700940497611,
+      "learning_rate": 0.0001244087377102508,
+      "loss": 0.2598,
+      "step": 2566
+    },
+    {
+      "epoch": 1.4389013452914798,
+      "grad_norm": 0.0701666560713026,
+      "learning_rate": 0.00012434547318457474,
+      "loss": 0.2718,
+      "step": 2567
+    },
+    {
+      "epoch": 1.4394618834080717,
+      "grad_norm": 0.07068547220314525,
+      "learning_rate": 0.00012428219829925083,
+      "loss": 0.2653,
+      "step": 2568
+    },
+    {
+      "epoch": 1.4400224215246638,
+      "grad_norm": 0.07092853131496765,
+      "learning_rate": 0.0001242189130812042,
+      "loss": 0.2641,
+      "step": 2569
+    },
+    {
+      "epoch": 1.4405829596412556,
+      "grad_norm": 0.07325103216390964,
+      "learning_rate": 0.0001241556175573644,
+      "loss": 0.2786,
+      "step": 2570
+    },
+    {
+      "epoch": 1.4411434977578477,
+      "grad_norm": 0.07007544062017068,
+      "learning_rate": 0.00012409231175466537,
+      "loss": 0.2706,
+      "step": 2571
+    },
+    {
+      "epoch": 1.4417040358744395,
+      "grad_norm": 0.06931592432697589,
+      "learning_rate": 0.00012402899570004543,
+      "loss": 0.2616,
+      "step": 2572
+    },
+    {
+      "epoch": 1.4422645739910314,
+      "grad_norm": 0.06984288584717985,
+      "learning_rate": 0.00012396566942044724,
+      "loss": 0.2694,
+      "step": 2573
+    },
+    {
+      "epoch": 1.4428251121076232,
+      "grad_norm": 0.07034742549246696,
+      "learning_rate": 0.0001239023329428178,
+      "loss": 0.2583,
+      "step": 2574
+    },
+    {
+      "epoch": 1.4433856502242153,
+      "grad_norm": 0.07156428602049808,
+      "learning_rate": 0.00012383898629410843,
+      "loss": 0.2718,
+      "step": 2575
+    },
+    {
+      "epoch": 1.4439461883408071,
+      "grad_norm": 0.07435589986711433,
+      "learning_rate": 0.00012377562950127493,
+      "loss": 0.2879,
+      "step": 2576
+    },
+    {
+      "epoch": 1.4445067264573992,
+      "grad_norm": 0.07248407339009183,
+      "learning_rate": 0.00012371226259127725,
+      "loss": 0.268,
+      "step": 2577
+    },
+    {
+      "epoch": 1.445067264573991,
+      "grad_norm": 0.07149452390292611,
+      "learning_rate": 0.00012364888559107966,
+      "loss": 0.2625,
+      "step": 2578
+    },
+    {
+      "epoch": 1.4456278026905829,
+      "grad_norm": 0.07385282012847322,
+      "learning_rate": 0.00012358549852765083,
+      "loss": 0.2687,
+      "step": 2579
+    },
+    {
+      "epoch": 1.446188340807175,
+      "grad_norm": 0.07214885170023577,
+      "learning_rate": 0.0001235221014279636,
+      "loss": 0.2601,
+      "step": 2580
+    },
+    {
+      "epoch": 1.4467488789237668,
+      "grad_norm": 0.07356871078510978,
+      "learning_rate": 0.0001234586943189951,
+      "loss": 0.263,
+      "step": 2581
+    },
+    {
+      "epoch": 1.4473094170403589,
+      "grad_norm": 0.07215878633263344,
+      "learning_rate": 0.00012339527722772683,
+      "loss": 0.2722,
+      "step": 2582
+    },
+    {
+      "epoch": 1.4478699551569507,
+      "grad_norm": 0.0714950757352865,
+      "learning_rate": 0.00012333185018114439,
+      "loss": 0.2782,
+      "step": 2583
+    },
+    {
+      "epoch": 1.4484304932735426,
+      "grad_norm": 0.06893091139036768,
+      "learning_rate": 0.00012326841320623767,
+      "loss": 0.2582,
+      "step": 2584
+    },
+    {
+      "epoch": 1.4489910313901344,
+      "grad_norm": 0.07247272132126951,
+      "learning_rate": 0.00012320496633000088,
+      "loss": 0.2705,
+      "step": 2585
+    },
+    {
+      "epoch": 1.4495515695067265,
+      "grad_norm": 0.0742463707609409,
+      "learning_rate": 0.00012314150957943226,
+      "loss": 0.257,
+      "step": 2586
+    },
+    {
+      "epoch": 1.4501121076233183,
+      "grad_norm": 0.07289183935471263,
+      "learning_rate": 0.0001230780429815344,
+      "loss": 0.2785,
+      "step": 2587
+    },
+    {
+      "epoch": 1.4506726457399104,
+      "grad_norm": 0.07100913828665414,
+      "learning_rate": 0.00012301456656331402,
+      "loss": 0.2545,
+      "step": 2588
+    },
+    {
+      "epoch": 1.4512331838565022,
+      "grad_norm": 0.07359079977697225,
+      "learning_rate": 0.000122951080351782,
+      "loss": 0.2752,
+      "step": 2589
+    },
+    {
+      "epoch": 1.451793721973094,
+      "grad_norm": 0.07124953126346799,
+      "learning_rate": 0.00012288758437395343,
+      "loss": 0.262,
+      "step": 2590
+    },
+    {
+      "epoch": 1.4523542600896862,
+      "grad_norm": 0.07301714224898355,
+      "learning_rate": 0.00012282407865684758,
+      "loss": 0.2786,
+      "step": 2591
+    },
+    {
+      "epoch": 1.452914798206278,
+      "grad_norm": 0.0728659319865238,
+      "learning_rate": 0.00012276056322748778,
+      "loss": 0.254,
+      "step": 2592
+    },
+    {
+      "epoch": 1.45347533632287,
+      "grad_norm": 0.07110814045711723,
+      "learning_rate": 0.0001226970381129016,
+      "loss": 0.2677,
+      "step": 2593
+    },
+    {
+      "epoch": 1.454035874439462,
+      "grad_norm": 0.07231421794163116,
+      "learning_rate": 0.0001226335033401206,
+      "loss": 0.2743,
+      "step": 2594
+    },
+    {
+      "epoch": 1.4545964125560538,
+      "grad_norm": 0.07283555116735829,
+      "learning_rate": 0.00012256995893618054,
+      "loss": 0.2642,
+      "step": 2595
+    },
+    {
+      "epoch": 1.4551569506726456,
+      "grad_norm": 0.073614984693919,
+      "learning_rate": 0.0001225064049281212,
+      "loss": 0.2709,
+      "step": 2596
+    },
+    {
+      "epoch": 1.4557174887892377,
+      "grad_norm": 0.07712966688124136,
+      "learning_rate": 0.00012244284134298666,
+      "loss": 0.2737,
+      "step": 2597
+    },
+    {
+      "epoch": 1.4562780269058295,
+      "grad_norm": 0.07027837764258521,
+      "learning_rate": 0.00012237926820782478,
+      "loss": 0.2582,
+      "step": 2598
+    },
+    {
+      "epoch": 1.4568385650224216,
+      "grad_norm": 0.07190987399585402,
+      "learning_rate": 0.00012231568554968767,
+      "loss": 0.2709,
+      "step": 2599
+    },
+    {
+      "epoch": 1.4573991031390134,
+      "grad_norm": 0.07164571881496312,
+      "learning_rate": 0.00012225209339563145,
+      "loss": 0.2721,
+      "step": 2600
+    },
+    {
+      "epoch": 1.4579596412556053,
+      "grad_norm": 0.07157687722194844,
+      "learning_rate": 0.00012218849177271626,
+      "loss": 0.2641,
+      "step": 2601
+    },
+    {
+      "epoch": 1.4585201793721974,
+      "grad_norm": 0.07083119161099244,
+      "learning_rate": 0.00012212488070800635,
+      "loss": 0.2759,
+      "step": 2602
+    },
+    {
+      "epoch": 1.4590807174887892,
+      "grad_norm": 0.07050244131689339,
+      "learning_rate": 0.00012206126022856984,
+      "loss": 0.2668,
+      "step": 2603
+    },
+    {
+      "epoch": 1.4596412556053813,
+      "grad_norm": 0.07271707727122122,
+      "learning_rate": 0.00012199763036147895,
+      "loss": 0.2731,
+      "step": 2604
+    },
+    {
+      "epoch": 1.4602017937219731,
+      "grad_norm": 0.06951004638988882,
+      "learning_rate": 0.00012193399113380994,
+      "loss": 0.2617,
+      "step": 2605
+    },
+    {
+      "epoch": 1.460762331838565,
+      "grad_norm": 0.0723026309843755,
+      "learning_rate": 0.00012187034257264297,
+      "loss": 0.2689,
+      "step": 2606
+    },
+    {
+      "epoch": 1.461322869955157,
+      "grad_norm": 0.07046208655478085,
+      "learning_rate": 0.0001218066847050622,
+      "loss": 0.257,
+      "step": 2607
+    },
+    {
+      "epoch": 1.4618834080717489,
+      "grad_norm": 0.07079115538340826,
+      "learning_rate": 0.00012174301755815571,
+      "loss": 0.2782,
+      "step": 2608
+    },
+    {
+      "epoch": 1.4624439461883407,
+      "grad_norm": 0.07405499837804955,
+      "learning_rate": 0.00012167934115901563,
+      "loss": 0.2658,
+      "step": 2609
+    },
+    {
+      "epoch": 1.4630044843049328,
+      "grad_norm": 0.0724381236796581,
+      "learning_rate": 0.00012161565553473792,
+      "loss": 0.2753,
+      "step": 2610
+    },
+    {
+      "epoch": 1.4635650224215246,
+      "grad_norm": 0.07152512654173403,
+      "learning_rate": 0.00012155196071242254,
+      "loss": 0.2737,
+      "step": 2611
+    },
+    {
+      "epoch": 1.4641255605381165,
+      "grad_norm": 0.07565898894319906,
+      "learning_rate": 0.00012148825671917334,
+      "loss": 0.2762,
+      "step": 2612
+    },
+    {
+      "epoch": 1.4646860986547086,
+      "grad_norm": 0.07006862453203005,
+      "learning_rate": 0.00012142454358209803,
+      "loss": 0.2585,
+      "step": 2613
+    },
+    {
+      "epoch": 1.4652466367713004,
+      "grad_norm": 0.06943282943633416,
+      "learning_rate": 0.00012136082132830828,
+      "loss": 0.2619,
+      "step": 2614
+    },
+    {
+      "epoch": 1.4658071748878925,
+      "grad_norm": 0.07151809193225557,
+      "learning_rate": 0.0001212970899849196,
+      "loss": 0.2721,
+      "step": 2615
+    },
+    {
+      "epoch": 1.4663677130044843,
+      "grad_norm": 0.06957586106604624,
+      "learning_rate": 0.0001212333495790514,
+      "loss": 0.2617,
+      "step": 2616
+    },
+    {
+      "epoch": 1.4669282511210762,
+      "grad_norm": 0.0682323905413523,
+      "learning_rate": 0.00012116960013782684,
+      "loss": 0.2675,
+      "step": 2617
+    },
+    {
+      "epoch": 1.4674887892376682,
+      "grad_norm": 0.07157429220500997,
+      "learning_rate": 0.00012110584168837309,
+      "loss": 0.274,
+      "step": 2618
+    },
+    {
+      "epoch": 1.46804932735426,
+      "grad_norm": 0.07082774922467017,
+      "learning_rate": 0.00012104207425782104,
+      "loss": 0.2691,
+      "step": 2619
+    },
+    {
+      "epoch": 1.4686098654708521,
+      "grad_norm": 0.07172556028530615,
+      "learning_rate": 0.00012097829787330544,
+      "loss": 0.2782,
+      "step": 2620
+    },
+    {
+      "epoch": 1.469170403587444,
+      "grad_norm": 0.07282110633980718,
+      "learning_rate": 0.00012091451256196484,
+      "loss": 0.2737,
+      "step": 2621
+    },
+    {
+      "epoch": 1.4697309417040358,
+      "grad_norm": 0.07215127145326976,
+      "learning_rate": 0.0001208507183509416,
+      "loss": 0.2644,
+      "step": 2622
+    },
+    {
+      "epoch": 1.4702914798206277,
+      "grad_norm": 0.07070637957573725,
+      "learning_rate": 0.00012078691526738181,
+      "loss": 0.2675,
+      "step": 2623
+    },
+    {
+      "epoch": 1.4708520179372198,
+      "grad_norm": 0.07102377989515693,
+      "learning_rate": 0.00012072310333843544,
+      "loss": 0.2641,
+      "step": 2624
+    },
+    {
+      "epoch": 1.4714125560538116,
+      "grad_norm": 0.07221309044790118,
+      "learning_rate": 0.00012065928259125611,
+      "loss": 0.2614,
+      "step": 2625
+    },
+    {
+      "epoch": 1.4719730941704037,
+      "grad_norm": 0.07180070926361873,
+      "learning_rate": 0.0001205954530530013,
+      "loss": 0.2736,
+      "step": 2626
+    },
+    {
+      "epoch": 1.4725336322869955,
+      "grad_norm": 0.06971237885977041,
+      "learning_rate": 0.0001205316147508322,
+      "loss": 0.2717,
+      "step": 2627
+    },
+    {
+      "epoch": 1.4730941704035874,
+      "grad_norm": 0.07367921734383244,
+      "learning_rate": 0.00012046776771191366,
+      "loss": 0.2777,
+      "step": 2628
+    },
+    {
+      "epoch": 1.4736547085201794,
+      "grad_norm": 0.07097022201909618,
+      "learning_rate": 0.00012040391196341427,
+      "loss": 0.2712,
+      "step": 2629
+    },
+    {
+      "epoch": 1.4742152466367713,
+      "grad_norm": 0.06842246960972916,
+      "learning_rate": 0.00012034004753250643,
+      "loss": 0.2625,
+      "step": 2630
+    },
+    {
+      "epoch": 1.4747757847533634,
+      "grad_norm": 0.07260291316386094,
+      "learning_rate": 0.00012027617444636612,
+      "loss": 0.2567,
+      "step": 2631
+    },
+    {
+      "epoch": 1.4753363228699552,
+      "grad_norm": 0.06995572021067038,
+      "learning_rate": 0.00012021229273217302,
+      "loss": 0.2641,
+      "step": 2632
+    },
+    {
+      "epoch": 1.475896860986547,
+      "grad_norm": 0.07559509626086572,
+      "learning_rate": 0.00012014840241711054,
+      "loss": 0.2786,
+      "step": 2633
+    },
+    {
+      "epoch": 1.476457399103139,
+      "grad_norm": 0.07400623872966176,
+      "learning_rate": 0.00012008450352836572,
+      "loss": 0.2664,
+      "step": 2634
+    },
+    {
+      "epoch": 1.477017937219731,
+      "grad_norm": 0.06892022781762777,
+      "learning_rate": 0.00012002059609312917,
+      "loss": 0.2598,
+      "step": 2635
+    },
+    {
+      "epoch": 1.4775784753363228,
+      "grad_norm": 0.07318586459323616,
+      "learning_rate": 0.00011995668013859529,
+      "loss": 0.2739,
+      "step": 2636
+    },
+    {
+      "epoch": 1.4781390134529149,
+      "grad_norm": 0.0724655659139408,
+      "learning_rate": 0.00011989275569196194,
+      "loss": 0.2732,
+      "step": 2637
+    },
+    {
+      "epoch": 1.4786995515695067,
+      "grad_norm": 0.0708265950975277,
+      "learning_rate": 0.00011982882278043077,
+      "loss": 0.263,
+      "step": 2638
+    },
+    {
+      "epoch": 1.4792600896860986,
+      "grad_norm": 0.0715074657441884,
+      "learning_rate": 0.00011976488143120687,
+      "loss": 0.2741,
+      "step": 2639
+    },
+    {
+      "epoch": 1.4798206278026906,
+      "grad_norm": 0.06812472045590152,
+      "learning_rate": 0.00011970093167149905,
+      "loss": 0.2696,
+      "step": 2640
+    },
+    {
+      "epoch": 1.4803811659192825,
+      "grad_norm": 0.06926495675281664,
+      "learning_rate": 0.00011963697352851955,
+      "loss": 0.2715,
+      "step": 2641
+    },
+    {
+      "epoch": 1.4809417040358746,
+      "grad_norm": 0.06789046197779881,
+      "learning_rate": 0.00011957300702948435,
+      "loss": 0.2641,
+      "step": 2642
+    },
+    {
+      "epoch": 1.4815022421524664,
+      "grad_norm": 0.07183132691344266,
+      "learning_rate": 0.00011950903220161285,
+      "loss": 0.275,
+      "step": 2643
+    },
+    {
+      "epoch": 1.4820627802690582,
+      "grad_norm": 0.07124962976203886,
+      "learning_rate": 0.00011944504907212804,
+      "loss": 0.2667,
+      "step": 2644
+    },
+    {
+      "epoch": 1.48262331838565,
+      "grad_norm": 0.07381585212265701,
+      "learning_rate": 0.0001193810576682565,
+      "loss": 0.2736,
+      "step": 2645
+    },
+    {
+      "epoch": 1.4831838565022422,
+      "grad_norm": 0.07207926726683282,
+      "learning_rate": 0.00011931705801722818,
+      "loss": 0.2667,
+      "step": 2646
+    },
+    {
+      "epoch": 1.483744394618834,
+      "grad_norm": 0.07106474229779323,
+      "learning_rate": 0.00011925305014627678,
+      "loss": 0.2588,
+      "step": 2647
+    },
+    {
+      "epoch": 1.484304932735426,
+      "grad_norm": 0.07148674970861063,
+      "learning_rate": 0.00011918903408263924,
+      "loss": 0.2798,
+      "step": 2648
+    },
+    {
+      "epoch": 1.484865470852018,
+      "grad_norm": 0.06948382996852732,
+      "learning_rate": 0.00011912500985355614,
+      "loss": 0.2712,
+      "step": 2649
+    },
+    {
+      "epoch": 1.4854260089686098,
+      "grad_norm": 0.07199061909052544,
+      "learning_rate": 0.00011906097748627149,
+      "loss": 0.278,
+      "step": 2650
+    },
+    {
+      "epoch": 1.4859865470852018,
+      "grad_norm": 0.07159046390315446,
+      "learning_rate": 0.00011899693700803278,
+      "loss": 0.2717,
+      "step": 2651
+    },
+    {
+      "epoch": 1.4865470852017937,
+      "grad_norm": 0.07056011106519934,
+      "learning_rate": 0.00011893288844609094,
+      "loss": 0.274,
+      "step": 2652
+    },
+    {
+      "epoch": 1.4871076233183858,
+      "grad_norm": 0.07348995833777743,
+      "learning_rate": 0.00011886883182770035,
+      "loss": 0.2579,
+      "step": 2653
+    },
+    {
+      "epoch": 1.4876681614349776,
+      "grad_norm": 0.07263591105598495,
+      "learning_rate": 0.00011880476718011877,
+      "loss": 0.2615,
+      "step": 2654
+    },
+    {
+      "epoch": 1.4882286995515694,
+      "grad_norm": 0.07070466411503919,
+      "learning_rate": 0.00011874069453060746,
+      "loss": 0.2527,
+      "step": 2655
+    },
+    {
+      "epoch": 1.4887892376681615,
+      "grad_norm": 0.07060666180313581,
+      "learning_rate": 0.000118676613906431,
+      "loss": 0.2723,
+      "step": 2656
+    },
+    {
+      "epoch": 1.4893497757847534,
+      "grad_norm": 0.07293653223106318,
+      "learning_rate": 0.00011861252533485742,
+      "loss": 0.2621,
+      "step": 2657
+    },
+    {
+      "epoch": 1.4899103139013454,
+      "grad_norm": 0.07305538068005077,
+      "learning_rate": 0.00011854842884315813,
+      "loss": 0.2736,
+      "step": 2658
+    },
+    {
+      "epoch": 1.4904708520179373,
+      "grad_norm": 0.07138618891331834,
+      "learning_rate": 0.00011848432445860789,
+      "loss": 0.2794,
+      "step": 2659
+    },
+    {
+      "epoch": 1.4910313901345291,
+      "grad_norm": 0.07073266031117438,
+      "learning_rate": 0.00011842021220848486,
+      "loss": 0.2644,
+      "step": 2660
+    },
+    {
+      "epoch": 1.491591928251121,
+      "grad_norm": 0.0765358318870987,
+      "learning_rate": 0.00011835609212007042,
+      "loss": 0.2958,
+      "step": 2661
+    },
+    {
+      "epoch": 1.492152466367713,
+      "grad_norm": 0.07162018165171527,
+      "learning_rate": 0.00011829196422064943,
+      "loss": 0.2662,
+      "step": 2662
+    },
+    {
+      "epoch": 1.4927130044843049,
+      "grad_norm": 0.07290240396375056,
+      "learning_rate": 0.00011822782853751002,
+      "loss": 0.2688,
+      "step": 2663
+    },
+    {
+      "epoch": 1.493273542600897,
+      "grad_norm": 0.07229718840329626,
+      "learning_rate": 0.00011816368509794364,
+      "loss": 0.2691,
+      "step": 2664
+    },
+    {
+      "epoch": 1.4938340807174888,
+      "grad_norm": 0.07353094536852173,
+      "learning_rate": 0.00011809953392924504,
+      "loss": 0.2605,
+      "step": 2665
+    },
+    {
+      "epoch": 1.4943946188340806,
+      "grad_norm": 0.07089474069888169,
+      "learning_rate": 0.00011803537505871225,
+      "loss": 0.2573,
+      "step": 2666
+    },
+    {
+      "epoch": 1.4949551569506727,
+      "grad_norm": 0.07282249799098499,
+      "learning_rate": 0.00011797120851364653,
+      "loss": 0.2755,
+      "step": 2667
+    },
+    {
+      "epoch": 1.4955156950672646,
+      "grad_norm": 0.07238850475224404,
+      "learning_rate": 0.00011790703432135253,
+      "loss": 0.273,
+      "step": 2668
+    },
+    {
+      "epoch": 1.4960762331838566,
+      "grad_norm": 0.07410776643227834,
+      "learning_rate": 0.00011784285250913802,
+      "loss": 0.274,
+      "step": 2669
+    },
+    {
+      "epoch": 1.4966367713004485,
+      "grad_norm": 0.07013595725707807,
+      "learning_rate": 0.00011777866310431409,
+      "loss": 0.2643,
+      "step": 2670
+    },
+    {
+      "epoch": 1.4971973094170403,
+      "grad_norm": 0.07312148011415108,
+      "learning_rate": 0.00011771446613419508,
+      "loss": 0.2622,
+      "step": 2671
+    },
+    {
+      "epoch": 1.4977578475336322,
+      "grad_norm": 0.0715652009409089,
+      "learning_rate": 0.00011765026162609847,
+      "loss": 0.2753,
+      "step": 2672
+    },
+    {
+      "epoch": 1.4983183856502242,
+      "grad_norm": 0.06865430293035583,
+      "learning_rate": 0.00011758604960734499,
+      "loss": 0.2624,
+      "step": 2673
+    },
+    {
+      "epoch": 1.498878923766816,
+      "grad_norm": 0.071011427325384,
+      "learning_rate": 0.0001175218301052586,
+      "loss": 0.2753,
+      "step": 2674
+    },
+    {
+      "epoch": 1.4994394618834082,
+      "grad_norm": 0.07501914248030692,
+      "learning_rate": 0.00011745760314716636,
+      "loss": 0.2846,
+      "step": 2675
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.07315247184854988,
+      "learning_rate": 0.00011739336876039859,
+      "loss": 0.2536,
+      "step": 2676
+    },
+    {
+      "epoch": 1.5005605381165918,
+      "grad_norm": 0.07157590122387843,
+      "learning_rate": 0.00011732912697228872,
+      "loss": 0.2767,
+      "step": 2677
+    },
+    {
+      "epoch": 1.5011210762331837,
+      "grad_norm": 0.07109590528600179,
+      "learning_rate": 0.00011726487781017337,
+      "loss": 0.2694,
+      "step": 2678
+    },
+    {
+      "epoch": 1.5016816143497758,
+      "grad_norm": 0.07068316674858972,
+      "learning_rate": 0.0001172006213013922,
+      "loss": 0.2658,
+      "step": 2679
+    },
+    {
+      "epoch": 1.5022421524663678,
+      "grad_norm": 0.07250084006014036,
+      "learning_rate": 0.00011713635747328818,
+      "loss": 0.2706,
+      "step": 2680
+    },
+    {
+      "epoch": 1.5028026905829597,
+      "grad_norm": 0.07366391637259316,
+      "learning_rate": 0.00011707208635320718,
+      "loss": 0.2674,
+      "step": 2681
+    },
+    {
+      "epoch": 1.5033632286995515,
+      "grad_norm": 0.0699000966109222,
+      "learning_rate": 0.00011700780796849833,
+      "loss": 0.2713,
+      "step": 2682
+    },
+    {
+      "epoch": 1.5039237668161434,
+      "grad_norm": 0.07257458686854877,
+      "learning_rate": 0.00011694352234651373,
+      "loss": 0.2849,
+      "step": 2683
+    },
+    {
+      "epoch": 1.5044843049327354,
+      "grad_norm": 0.07190670499483498,
+      "learning_rate": 0.00011687922951460872,
+      "loss": 0.2657,
+      "step": 2684
+    },
+    {
+      "epoch": 1.5050448430493275,
+      "grad_norm": 0.07117959792590416,
+      "learning_rate": 0.00011681492950014157,
+      "loss": 0.266,
+      "step": 2685
+    },
+    {
+      "epoch": 1.5056053811659194,
+      "grad_norm": 0.07294991022142684,
+      "learning_rate": 0.00011675062233047364,
+      "loss": 0.2763,
+      "step": 2686
+    },
+    {
+      "epoch": 1.5061659192825112,
+      "grad_norm": 0.07142237945995769,
+      "learning_rate": 0.00011668630803296939,
+      "loss": 0.2802,
+      "step": 2687
+    },
+    {
+      "epoch": 1.506726457399103,
+      "grad_norm": 0.0735994024551922,
+      "learning_rate": 0.00011662198663499619,
+      "loss": 0.2659,
+      "step": 2688
+    },
+    {
+      "epoch": 1.5072869955156951,
+      "grad_norm": 0.07203572611511369,
+      "learning_rate": 0.00011655765816392457,
+      "loss": 0.2687,
+      "step": 2689
+    },
+    {
+      "epoch": 1.507847533632287,
+      "grad_norm": 0.071135139809204,
+      "learning_rate": 0.00011649332264712798,
+      "loss": 0.2687,
+      "step": 2690
+    },
+    {
+      "epoch": 1.508408071748879,
+      "grad_norm": 0.07379585959916812,
+      "learning_rate": 0.00011642898011198288,
+      "loss": 0.2756,
+      "step": 2691
+    },
+    {
+      "epoch": 1.5089686098654709,
+      "grad_norm": 0.07231405879270913,
+      "learning_rate": 0.00011636463058586881,
+      "loss": 0.2556,
+      "step": 2692
+    },
+    {
+      "epoch": 1.5095291479820627,
+      "grad_norm": 0.07157443774998543,
+      "learning_rate": 0.00011630027409616817,
+      "loss": 0.2653,
+      "step": 2693
+    },
+    {
+      "epoch": 1.5100896860986546,
+      "grad_norm": 0.06944155962361898,
+      "learning_rate": 0.00011623591067026636,
+      "loss": 0.2728,
+      "step": 2694
+    },
+    {
+      "epoch": 1.5106502242152466,
+      "grad_norm": 0.07350506179889904,
+      "learning_rate": 0.00011617154033555169,
+      "loss": 0.2753,
+      "step": 2695
+    },
+    {
+      "epoch": 1.5112107623318387,
+      "grad_norm": 0.06980443587081692,
+      "learning_rate": 0.0001161071631194155,
+      "loss": 0.2537,
+      "step": 2696
+    },
+    {
+      "epoch": 1.5117713004484306,
+      "grad_norm": 0.07382751069626999,
+      "learning_rate": 0.000116042779049252,
+      "loss": 0.289,
+      "step": 2697
+    },
+    {
+      "epoch": 1.5123318385650224,
+      "grad_norm": 0.06844084556266467,
+      "learning_rate": 0.00011597838815245836,
+      "loss": 0.2514,
+      "step": 2698
+    },
+    {
+      "epoch": 1.5128923766816142,
+      "grad_norm": 0.07038537623927177,
+      "learning_rate": 0.00011591399045643455,
+      "loss": 0.2789,
+      "step": 2699
+    },
+    {
+      "epoch": 1.5134529147982063,
+      "grad_norm": 0.07115950011898217,
+      "learning_rate": 0.00011584958598858359,
+      "loss": 0.2532,
+      "step": 2700
+    },
+    {
+      "epoch": 1.5140134529147982,
+      "grad_norm": 0.07091988456732695,
+      "learning_rate": 0.00011578517477631125,
+      "loss": 0.2649,
+      "step": 2701
+    },
+    {
+      "epoch": 1.5145739910313902,
+      "grad_norm": 0.07208183094581253,
+      "learning_rate": 0.00011572075684702624,
+      "loss": 0.2725,
+      "step": 2702
+    },
+    {
+      "epoch": 1.515134529147982,
+      "grad_norm": 0.07021910344644247,
+      "learning_rate": 0.00011565633222814005,
+      "loss": 0.2623,
+      "step": 2703
+    },
+    {
+      "epoch": 1.515695067264574,
+      "grad_norm": 0.07114186929092427,
+      "learning_rate": 0.00011559190094706714,
+      "loss": 0.2557,
+      "step": 2704
+    },
+    {
+      "epoch": 1.5162556053811658,
+      "grad_norm": 0.07215828197766797,
+      "learning_rate": 0.0001155274630312247,
+      "loss": 0.2781,
+      "step": 2705
+    },
+    {
+      "epoch": 1.5168161434977578,
+      "grad_norm": 0.07282199701609747,
+      "learning_rate": 0.00011546301850803282,
+      "loss": 0.2737,
+      "step": 2706
+    },
+    {
+      "epoch": 1.51737668161435,
+      "grad_norm": 0.07083573120703515,
+      "learning_rate": 0.00011539856740491432,
+      "loss": 0.2695,
+      "step": 2707
+    },
+    {
+      "epoch": 1.5179372197309418,
+      "grad_norm": 0.07259718479360823,
+      "learning_rate": 0.0001153341097492949,
+      "loss": 0.2805,
+      "step": 2708
+    },
+    {
+      "epoch": 1.5184977578475336,
+      "grad_norm": 0.07061212408384449,
+      "learning_rate": 0.00011526964556860298,
+      "loss": 0.2661,
+      "step": 2709
+    },
+    {
+      "epoch": 1.5190582959641254,
+      "grad_norm": 0.06926870430173317,
+      "learning_rate": 0.0001152051748902698,
+      "loss": 0.2656,
+      "step": 2710
+    },
+    {
+      "epoch": 1.5196188340807175,
+      "grad_norm": 0.07171120149804051,
+      "learning_rate": 0.00011514069774172936,
+      "loss": 0.2747,
+      "step": 2711
+    },
+    {
+      "epoch": 1.5201793721973094,
+      "grad_norm": 0.07436517776992174,
+      "learning_rate": 0.00011507621415041837,
+      "loss": 0.2773,
+      "step": 2712
+    },
+    {
+      "epoch": 1.5207399103139014,
+      "grad_norm": 0.0688617758581786,
+      "learning_rate": 0.00011501172414377634,
+      "loss": 0.2563,
+      "step": 2713
+    },
+    {
+      "epoch": 1.5213004484304933,
+      "grad_norm": 0.07049212794757495,
+      "learning_rate": 0.00011494722774924554,
+      "loss": 0.2645,
+      "step": 2714
+    },
+    {
+      "epoch": 1.5218609865470851,
+      "grad_norm": 0.07169040941860719,
+      "learning_rate": 0.0001148827249942708,
+      "loss": 0.2821,
+      "step": 2715
+    },
+    {
+      "epoch": 1.522421524663677,
+      "grad_norm": 0.07251652451590532,
+      "learning_rate": 0.00011481821590629985,
+      "loss": 0.2593,
+      "step": 2716
+    },
+    {
+      "epoch": 1.522982062780269,
+      "grad_norm": 0.07270256957010714,
+      "learning_rate": 0.00011475370051278298,
+      "loss": 0.2657,
+      "step": 2717
+    },
+    {
+      "epoch": 1.523542600896861,
+      "grad_norm": 0.07386026335593725,
+      "learning_rate": 0.00011468917884117323,
+      "loss": 0.2707,
+      "step": 2718
+    },
+    {
+      "epoch": 1.524103139013453,
+      "grad_norm": 0.07115321344068759,
+      "learning_rate": 0.0001146246509189263,
+      "loss": 0.2782,
+      "step": 2719
+    },
+    {
+      "epoch": 1.5246636771300448,
+      "grad_norm": 0.07029243312234992,
+      "learning_rate": 0.00011456011677350051,
+      "loss": 0.2701,
+      "step": 2720
+    },
+    {
+      "epoch": 1.5252242152466366,
+      "grad_norm": 0.07110567655941383,
+      "learning_rate": 0.00011449557643235686,
+      "loss": 0.2646,
+      "step": 2721
+    },
+    {
+      "epoch": 1.5257847533632287,
+      "grad_norm": 0.07330966340004928,
+      "learning_rate": 0.00011443102992295904,
+      "loss": 0.2872,
+      "step": 2722
+    },
+    {
+      "epoch": 1.5263452914798208,
+      "grad_norm": 0.07141975317438559,
+      "learning_rate": 0.00011436647727277326,
+      "loss": 0.2665,
+      "step": 2723
+    },
+    {
+      "epoch": 1.5269058295964126,
+      "grad_norm": 0.07109655928452469,
+      "learning_rate": 0.00011430191850926837,
+      "loss": 0.2806,
+      "step": 2724
+    },
+    {
+      "epoch": 1.5274663677130045,
+      "grad_norm": 0.07087037740185703,
+      "learning_rate": 0.0001142373536599159,
+      "loss": 0.2751,
+      "step": 2725
+    },
+    {
+      "epoch": 1.5280269058295963,
+      "grad_norm": 0.07042245583234814,
+      "learning_rate": 0.0001141727827521899,
+      "loss": 0.2726,
+      "step": 2726
+    },
+    {
+      "epoch": 1.5285874439461884,
+      "grad_norm": 0.06983269779012252,
+      "learning_rate": 0.00011410820581356705,
+      "loss": 0.2716,
+      "step": 2727
+    },
+    {
+      "epoch": 1.5291479820627802,
+      "grad_norm": 0.07007730001343894,
+      "learning_rate": 0.00011404362287152646,
+      "loss": 0.2693,
+      "step": 2728
+    },
+    {
+      "epoch": 1.5297085201793723,
+      "grad_norm": 0.06948591281865128,
+      "learning_rate": 0.00011397903395354996,
+      "loss": 0.2668,
+      "step": 2729
+    },
+    {
+      "epoch": 1.5302690582959642,
+      "grad_norm": 0.06983825876951026,
+      "learning_rate": 0.00011391443908712185,
+      "loss": 0.2685,
+      "step": 2730
+    },
+    {
+      "epoch": 1.530829596412556,
+      "grad_norm": 0.0703247996027728,
+      "learning_rate": 0.00011384983829972898,
+      "loss": 0.2661,
+      "step": 2731
+    },
+    {
+      "epoch": 1.5313901345291479,
+      "grad_norm": 0.07069065438339443,
+      "learning_rate": 0.00011378523161886066,
+      "loss": 0.2603,
+      "step": 2732
+    },
+    {
+      "epoch": 1.53195067264574,
+      "grad_norm": 0.07320307867717173,
+      "learning_rate": 0.00011372061907200881,
+      "loss": 0.2632,
+      "step": 2733
+    },
+    {
+      "epoch": 1.532511210762332,
+      "grad_norm": 0.07502228465481751,
+      "learning_rate": 0.0001136560006866678,
+      "loss": 0.2759,
+      "step": 2734
+    },
+    {
+      "epoch": 1.5330717488789238,
+      "grad_norm": 0.07260114925117225,
+      "learning_rate": 0.0001135913764903344,
+      "loss": 0.2656,
+      "step": 2735
+    },
+    {
+      "epoch": 1.5336322869955157,
+      "grad_norm": 0.07233958278729473,
+      "learning_rate": 0.00011352674651050796,
+      "loss": 0.2648,
+      "step": 2736
+    },
+    {
+      "epoch": 1.5341928251121075,
+      "grad_norm": 0.07047796696966747,
+      "learning_rate": 0.00011346211077469029,
+      "loss": 0.2653,
+      "step": 2737
+    },
+    {
+      "epoch": 1.5347533632286996,
+      "grad_norm": 0.06960266451046976,
+      "learning_rate": 0.00011339746931038562,
+      "loss": 0.2604,
+      "step": 2738
+    },
+    {
+      "epoch": 1.5353139013452914,
+      "grad_norm": 0.07251433886476083,
+      "learning_rate": 0.00011333282214510057,
+      "loss": 0.269,
+      "step": 2739
+    },
+    {
+      "epoch": 1.5358744394618835,
+      "grad_norm": 0.07198280804520825,
+      "learning_rate": 0.00011326816930634427,
+      "loss": 0.2739,
+      "step": 2740
+    },
+    {
+      "epoch": 1.5364349775784754,
+      "grad_norm": 0.06817081809098341,
+      "learning_rate": 0.00011320351082162821,
+      "loss": 0.2704,
+      "step": 2741
+    },
+    {
+      "epoch": 1.5369955156950672,
+      "grad_norm": 0.07273992337057941,
+      "learning_rate": 0.00011313884671846631,
+      "loss": 0.2663,
+      "step": 2742
+    },
+    {
+      "epoch": 1.537556053811659,
+      "grad_norm": 0.07221271416711049,
+      "learning_rate": 0.00011307417702437486,
+      "loss": 0.264,
+      "step": 2743
+    },
+    {
+      "epoch": 1.5381165919282511,
+      "grad_norm": 0.06943371837414848,
+      "learning_rate": 0.00011300950176687255,
+      "loss": 0.2567,
+      "step": 2744
+    },
+    {
+      "epoch": 1.5386771300448432,
+      "grad_norm": 0.07117596406391344,
+      "learning_rate": 0.00011294482097348041,
+      "loss": 0.2744,
+      "step": 2745
+    },
+    {
+      "epoch": 1.539237668161435,
+      "grad_norm": 0.07227818845611876,
+      "learning_rate": 0.00011288013467172184,
+      "loss": 0.2778,
+      "step": 2746
+    },
+    {
+      "epoch": 1.5397982062780269,
+      "grad_norm": 0.07042713884682089,
+      "learning_rate": 0.00011281544288912264,
+      "loss": 0.2618,
+      "step": 2747
+    },
+    {
+      "epoch": 1.5403587443946187,
+      "grad_norm": 0.07304650101652455,
+      "learning_rate": 0.0001127507456532108,
+      "loss": 0.2732,
+      "step": 2748
+    },
+    {
+      "epoch": 1.5409192825112108,
+      "grad_norm": 0.07328343982768012,
+      "learning_rate": 0.00011268604299151677,
+      "loss": 0.279,
+      "step": 2749
+    },
+    {
+      "epoch": 1.5414798206278026,
+      "grad_norm": 0.07318699184086801,
+      "learning_rate": 0.00011262133493157327,
+      "loss": 0.2713,
+      "step": 2750
+    },
+    {
+      "epoch": 1.5420403587443947,
+      "grad_norm": 0.07337036632547349,
+      "learning_rate": 0.00011255662150091526,
+      "loss": 0.2566,
+      "step": 2751
+    },
+    {
+      "epoch": 1.5426008968609866,
+      "grad_norm": 0.0695308768661344,
+      "learning_rate": 0.00011249190272708008,
+      "loss": 0.2624,
+      "step": 2752
+    },
+    {
+      "epoch": 1.5431614349775784,
+      "grad_norm": 0.0705214090456011,
+      "learning_rate": 0.00011242717863760723,
+      "loss": 0.2628,
+      "step": 2753
+    },
+    {
+      "epoch": 1.5437219730941703,
+      "grad_norm": 0.07129424585278271,
+      "learning_rate": 0.00011236244926003865,
+      "loss": 0.2788,
+      "step": 2754
+    },
+    {
+      "epoch": 1.5442825112107623,
+      "grad_norm": 0.07344043028329685,
+      "learning_rate": 0.0001122977146219183,
+      "loss": 0.2627,
+      "step": 2755
+    },
+    {
+      "epoch": 1.5448430493273544,
+      "grad_norm": 0.07176293018560918,
+      "learning_rate": 0.00011223297475079251,
+      "loss": 0.2728,
+      "step": 2756
+    },
+    {
+      "epoch": 1.5454035874439462,
+      "grad_norm": 0.07338480257588491,
+      "learning_rate": 0.00011216822967420985,
+      "loss": 0.2771,
+      "step": 2757
+    },
+    {
+      "epoch": 1.545964125560538,
+      "grad_norm": 0.06922021966047334,
+      "learning_rate": 0.00011210347941972108,
+      "loss": 0.2745,
+      "step": 2758
+    },
+    {
+      "epoch": 1.54652466367713,
+      "grad_norm": 0.0707187359354573,
+      "learning_rate": 0.00011203872401487916,
+      "loss": 0.2674,
+      "step": 2759
+    },
+    {
+      "epoch": 1.547085201793722,
+      "grad_norm": 0.07189796571809781,
+      "learning_rate": 0.00011197396348723923,
+      "loss": 0.2662,
+      "step": 2760
+    },
+    {
+      "epoch": 1.547645739910314,
+      "grad_norm": 0.07125606905672409,
+      "learning_rate": 0.00011190919786435863,
+      "loss": 0.2771,
+      "step": 2761
+    },
+    {
+      "epoch": 1.548206278026906,
+      "grad_norm": 0.07189254363456374,
+      "learning_rate": 0.00011184442717379686,
+      "loss": 0.27,
+      "step": 2762
+    },
+    {
+      "epoch": 1.5487668161434978,
+      "grad_norm": 0.06947296257811135,
+      "learning_rate": 0.00011177965144311556,
+      "loss": 0.2813,
+      "step": 2763
+    },
+    {
+      "epoch": 1.5493273542600896,
+      "grad_norm": 0.07347781274131902,
+      "learning_rate": 0.00011171487069987851,
+      "loss": 0.2789,
+      "step": 2764
+    },
+    {
+      "epoch": 1.5498878923766815,
+      "grad_norm": 0.06778496555499179,
+      "learning_rate": 0.00011165008497165168,
+      "loss": 0.2601,
+      "step": 2765
+    },
+    {
+      "epoch": 1.5504484304932735,
+      "grad_norm": 0.07252596866984758,
+      "learning_rate": 0.00011158529428600313,
+      "loss": 0.2689,
+      "step": 2766
+    },
+    {
+      "epoch": 1.5510089686098656,
+      "grad_norm": 0.0733734824623474,
+      "learning_rate": 0.00011152049867050305,
+      "loss": 0.266,
+      "step": 2767
+    },
+    {
+      "epoch": 1.5515695067264574,
+      "grad_norm": 0.06979653415887406,
+      "learning_rate": 0.0001114556981527236,
+      "loss": 0.2585,
+      "step": 2768
+    },
+    {
+      "epoch": 1.5521300448430493,
+      "grad_norm": 0.06970372354236257,
+      "learning_rate": 0.00011139089276023919,
+      "loss": 0.2684,
+      "step": 2769
+    },
+    {
+      "epoch": 1.5526905829596411,
+      "grad_norm": 0.07104856291168664,
+      "learning_rate": 0.00011132608252062629,
+      "loss": 0.2744,
+      "step": 2770
+    },
+    {
+      "epoch": 1.5532511210762332,
+      "grad_norm": 0.07079411878105876,
+      "learning_rate": 0.0001112612674614633,
+      "loss": 0.2643,
+      "step": 2771
+    },
+    {
+      "epoch": 1.5538116591928253,
+      "grad_norm": 0.0681075421107277,
+      "learning_rate": 0.00011119644761033078,
+      "loss": 0.2678,
+      "step": 2772
+    },
+    {
+      "epoch": 1.5543721973094171,
+      "grad_norm": 0.0721166405276055,
+      "learning_rate": 0.00011113162299481134,
+      "loss": 0.2744,
+      "step": 2773
+    },
+    {
+      "epoch": 1.554932735426009,
+      "grad_norm": 0.07146514304558317,
+      "learning_rate": 0.00011106679364248957,
+      "loss": 0.2532,
+      "step": 2774
+    },
+    {
+      "epoch": 1.5554932735426008,
+      "grad_norm": 0.06874137621174602,
+      "learning_rate": 0.00011100195958095208,
+      "loss": 0.2676,
+      "step": 2775
+    },
+    {
+      "epoch": 1.5560538116591929,
+      "grad_norm": 0.0709707177707054,
+      "learning_rate": 0.00011093712083778746,
+      "loss": 0.2658,
+      "step": 2776
+    },
+    {
+      "epoch": 1.5566143497757847,
+      "grad_norm": 0.07155378484789673,
+      "learning_rate": 0.00011087227744058637,
+      "loss": 0.2703,
+      "step": 2777
+    },
+    {
+      "epoch": 1.5571748878923768,
+      "grad_norm": 0.07099205099644741,
+      "learning_rate": 0.00011080742941694136,
+      "loss": 0.2521,
+      "step": 2778
+    },
+    {
+      "epoch": 1.5577354260089686,
+      "grad_norm": 0.07149758730548149,
+      "learning_rate": 0.00011074257679444702,
+      "loss": 0.2676,
+      "step": 2779
+    },
+    {
+      "epoch": 1.5582959641255605,
+      "grad_norm": 0.07460605906429514,
+      "learning_rate": 0.00011067771960069991,
+      "loss": 0.2743,
+      "step": 2780
+    },
+    {
+      "epoch": 1.5588565022421523,
+      "grad_norm": 0.07055552984018025,
+      "learning_rate": 0.0001106128578632984,
+      "loss": 0.2762,
+      "step": 2781
+    },
+    {
+      "epoch": 1.5594170403587444,
+      "grad_norm": 0.06945617974553053,
+      "learning_rate": 0.000110547991609843,
+      "loss": 0.2601,
+      "step": 2782
+    },
+    {
+      "epoch": 1.5599775784753365,
+      "grad_norm": 0.07204499448870326,
+      "learning_rate": 0.00011048312086793593,
+      "loss": 0.2736,
+      "step": 2783
+    },
+    {
+      "epoch": 1.5605381165919283,
+      "grad_norm": 0.06979439642918006,
+      "learning_rate": 0.00011041824566518146,
+      "loss": 0.2624,
+      "step": 2784
+    },
+    {
+      "epoch": 1.5610986547085202,
+      "grad_norm": 0.0715784064127982,
+      "learning_rate": 0.00011035336602918575,
+      "loss": 0.2789,
+      "step": 2785
+    },
+    {
+      "epoch": 1.561659192825112,
+      "grad_norm": 0.06969271980528327,
+      "learning_rate": 0.00011028848198755674,
+      "loss": 0.2662,
+      "step": 2786
+    },
+    {
+      "epoch": 1.562219730941704,
+      "grad_norm": 0.07031677060499117,
+      "learning_rate": 0.00011022359356790444,
+      "loss": 0.2613,
+      "step": 2787
+    },
+    {
+      "epoch": 1.562780269058296,
+      "grad_norm": 0.0713995278588566,
+      "learning_rate": 0.00011015870079784048,
+      "loss": 0.2679,
+      "step": 2788
+    },
+    {
+      "epoch": 1.563340807174888,
+      "grad_norm": 0.07023125811789262,
+      "learning_rate": 0.00011009380370497851,
+      "loss": 0.2739,
+      "step": 2789
+    },
+    {
+      "epoch": 1.5639013452914798,
+      "grad_norm": 0.0711916126481959,
+      "learning_rate": 0.00011002890231693395,
+      "loss": 0.2637,
+      "step": 2790
+    },
+    {
+      "epoch": 1.5644618834080717,
+      "grad_norm": 0.07193346331917248,
+      "learning_rate": 0.00010996399666132411,
+      "loss": 0.2602,
+      "step": 2791
+    },
+    {
+      "epoch": 1.5650224215246635,
+      "grad_norm": 0.07058975312724415,
+      "learning_rate": 0.00010989908676576807,
+      "loss": 0.2622,
+      "step": 2792
+    },
+    {
+      "epoch": 1.5655829596412556,
+      "grad_norm": 0.07184971261985763,
+      "learning_rate": 0.00010983417265788673,
+      "loss": 0.2715,
+      "step": 2793
+    },
+    {
+      "epoch": 1.5661434977578477,
+      "grad_norm": 0.070371241492397,
+      "learning_rate": 0.00010976925436530275,
+      "loss": 0.2638,
+      "step": 2794
+    },
+    {
+      "epoch": 1.5667040358744395,
+      "grad_norm": 0.06995184283050783,
+      "learning_rate": 0.00010970433191564058,
+      "loss": 0.2468,
+      "step": 2795
+    },
+    {
+      "epoch": 1.5672645739910314,
+      "grad_norm": 0.07167981001709199,
+      "learning_rate": 0.00010963940533652648,
+      "loss": 0.2804,
+      "step": 2796
+    },
+    {
+      "epoch": 1.5678251121076232,
+      "grad_norm": 0.0736796357197792,
+      "learning_rate": 0.00010957447465558844,
+      "loss": 0.2732,
+      "step": 2797
+    },
+    {
+      "epoch": 1.5683856502242153,
+      "grad_norm": 0.07222731874484677,
+      "learning_rate": 0.00010950953990045615,
+      "loss": 0.2703,
+      "step": 2798
+    },
+    {
+      "epoch": 1.5689461883408071,
+      "grad_norm": 0.07294943644454632,
+      "learning_rate": 0.00010944460109876116,
+      "loss": 0.2614,
+      "step": 2799
+    },
+    {
+      "epoch": 1.5695067264573992,
+      "grad_norm": 0.0697380808648057,
+      "learning_rate": 0.00010937965827813661,
+      "loss": 0.2551,
+      "step": 2800
+    },
+    {
+      "epoch": 1.570067264573991,
+      "grad_norm": 0.06959792050929359,
+      "learning_rate": 0.00010931471146621743,
+      "loss": 0.264,
+      "step": 2801
+    },
+    {
+      "epoch": 1.5706278026905829,
+      "grad_norm": 0.07226178235763066,
+      "learning_rate": 0.00010924976069064017,
+      "loss": 0.2611,
+      "step": 2802
+    },
+    {
+      "epoch": 1.5711883408071747,
+      "grad_norm": 0.06982425711033327,
+      "learning_rate": 0.00010918480597904317,
+      "loss": 0.2624,
+      "step": 2803
+    },
+    {
+      "epoch": 1.5717488789237668,
+      "grad_norm": 0.07167713506916018,
+      "learning_rate": 0.00010911984735906635,
+      "loss": 0.2766,
+      "step": 2804
+    },
+    {
+      "epoch": 1.5723094170403589,
+      "grad_norm": 0.07335757508509609,
+      "learning_rate": 0.00010905488485835138,
+      "loss": 0.2651,
+      "step": 2805
+    },
+    {
+      "epoch": 1.5728699551569507,
+      "grad_norm": 0.0690048074688932,
+      "learning_rate": 0.00010898991850454148,
+      "loss": 0.2609,
+      "step": 2806
+    },
+    {
+      "epoch": 1.5734304932735426,
+      "grad_norm": 0.06899119672362618,
+      "learning_rate": 0.00010892494832528161,
+      "loss": 0.2449,
+      "step": 2807
+    },
+    {
+      "epoch": 1.5739910313901344,
+      "grad_norm": 0.06809427165175322,
+      "learning_rate": 0.00010885997434821831,
+      "loss": 0.2741,
+      "step": 2808
+    },
+    {
+      "epoch": 1.5745515695067265,
+      "grad_norm": 0.07024340254718424,
+      "learning_rate": 0.0001087949966009997,
+      "loss": 0.2712,
+      "step": 2809
+    },
+    {
+      "epoch": 1.5751121076233185,
+      "grad_norm": 0.0713010211256685,
+      "learning_rate": 0.00010873001511127556,
+      "loss": 0.2719,
+      "step": 2810
+    },
+    {
+      "epoch": 1.5756726457399104,
+      "grad_norm": 0.07275783822804599,
+      "learning_rate": 0.0001086650299066973,
+      "loss": 0.2846,
+      "step": 2811
+    },
+    {
+      "epoch": 1.5762331838565022,
+      "grad_norm": 0.0708113444975407,
+      "learning_rate": 0.00010860004101491779,
+      "loss": 0.2727,
+      "step": 2812
+    },
+    {
+      "epoch": 1.576793721973094,
+      "grad_norm": 0.07054937568232769,
+      "learning_rate": 0.00010853504846359157,
+      "loss": 0.2584,
+      "step": 2813
+    },
+    {
+      "epoch": 1.577354260089686,
+      "grad_norm": 0.06924677616308327,
+      "learning_rate": 0.0001084700522803747,
+      "loss": 0.2642,
+      "step": 2814
+    },
+    {
+      "epoch": 1.577914798206278,
+      "grad_norm": 0.07141196823985863,
+      "learning_rate": 0.00010840505249292476,
+      "loss": 0.273,
+      "step": 2815
+    },
+    {
+      "epoch": 1.57847533632287,
+      "grad_norm": 0.07110415841307303,
+      "learning_rate": 0.00010834004912890092,
+      "loss": 0.2704,
+      "step": 2816
+    },
+    {
+      "epoch": 1.579035874439462,
+      "grad_norm": 0.07098262554380318,
+      "learning_rate": 0.00010827504221596387,
+      "loss": 0.284,
+      "step": 2817
+    },
+    {
+      "epoch": 1.5795964125560538,
+      "grad_norm": 0.06938465067367643,
+      "learning_rate": 0.0001082100317817757,
+      "loss": 0.2654,
+      "step": 2818
+    },
+    {
+      "epoch": 1.5801569506726456,
+      "grad_norm": 0.07177108422535652,
+      "learning_rate": 0.00010814501785400017,
+      "loss": 0.2643,
+      "step": 2819
+    },
+    {
+      "epoch": 1.5807174887892377,
+      "grad_norm": 0.07017833742767383,
+      "learning_rate": 0.0001080800004603024,
+      "loss": 0.253,
+      "step": 2820
+    },
+    {
+      "epoch": 1.5812780269058297,
+      "grad_norm": 0.07174402594855155,
+      "learning_rate": 0.00010801497962834906,
+      "loss": 0.264,
+      "step": 2821
+    },
+    {
+      "epoch": 1.5818385650224216,
+      "grad_norm": 0.0738102730992614,
+      "learning_rate": 0.00010794995538580819,
+      "loss": 0.2597,
+      "step": 2822
+    },
+    {
+      "epoch": 1.5823991031390134,
+      "grad_norm": 0.07128680620070806,
+      "learning_rate": 0.00010788492776034935,
+      "loss": 0.2648,
+      "step": 2823
+    },
+    {
+      "epoch": 1.5829596412556053,
+      "grad_norm": 0.07157932013154396,
+      "learning_rate": 0.00010781989677964355,
+      "loss": 0.2589,
+      "step": 2824
+    },
+    {
+      "epoch": 1.5835201793721974,
+      "grad_norm": 0.07198657642299293,
+      "learning_rate": 0.00010775486247136322,
+      "loss": 0.2638,
+      "step": 2825
+    },
+    {
+      "epoch": 1.5840807174887892,
+      "grad_norm": 0.07131804998406553,
+      "learning_rate": 0.00010768982486318216,
+      "loss": 0.2574,
+      "step": 2826
+    },
+    {
+      "epoch": 1.5846412556053813,
+      "grad_norm": 0.06973227350515016,
+      "learning_rate": 0.00010762478398277563,
+      "loss": 0.2503,
+      "step": 2827
+    },
+    {
+      "epoch": 1.5852017937219731,
+      "grad_norm": 0.07223748123285785,
+      "learning_rate": 0.00010755973985782022,
+      "loss": 0.2601,
+      "step": 2828
+    },
+    {
+      "epoch": 1.585762331838565,
+      "grad_norm": 0.07199270000671919,
+      "learning_rate": 0.000107494692515994,
+      "loss": 0.2676,
+      "step": 2829
+    },
+    {
+      "epoch": 1.5863228699551568,
+      "grad_norm": 0.07102152544902327,
+      "learning_rate": 0.00010742964198497629,
+      "loss": 0.2771,
+      "step": 2830
+    },
+    {
+      "epoch": 1.5868834080717489,
+      "grad_norm": 0.06943510787252358,
+      "learning_rate": 0.00010736458829244785,
+      "loss": 0.2591,
+      "step": 2831
+    },
+    {
+      "epoch": 1.587443946188341,
+      "grad_norm": 0.07229759136086263,
+      "learning_rate": 0.00010729953146609076,
+      "loss": 0.2679,
+      "step": 2832
+    },
+    {
+      "epoch": 1.5880044843049328,
+      "grad_norm": 0.07382303361218785,
+      "learning_rate": 0.00010723447153358843,
+      "loss": 0.2748,
+      "step": 2833
+    },
+    {
+      "epoch": 1.5885650224215246,
+      "grad_norm": 0.0722381606461495,
+      "learning_rate": 0.00010716940852262564,
+      "loss": 0.2711,
+      "step": 2834
+    },
+    {
+      "epoch": 1.5891255605381165,
+      "grad_norm": 0.07174213151662394,
+      "learning_rate": 0.00010710434246088834,
+      "loss": 0.2786,
+      "step": 2835
+    },
+    {
+      "epoch": 1.5896860986547086,
+      "grad_norm": 0.07072008633108894,
+      "learning_rate": 0.00010703927337606396,
+      "loss": 0.266,
+      "step": 2836
+    },
+    {
+      "epoch": 1.5902466367713004,
+      "grad_norm": 0.06802868005905105,
+      "learning_rate": 0.00010697420129584108,
+      "loss": 0.2702,
+      "step": 2837
+    },
+    {
+      "epoch": 1.5908071748878925,
+      "grad_norm": 0.0706938270071413,
+      "learning_rate": 0.00010690912624790966,
+      "loss": 0.2606,
+      "step": 2838
+    },
+    {
+      "epoch": 1.5913677130044843,
+      "grad_norm": 0.07081635104087369,
+      "learning_rate": 0.00010684404825996079,
+      "loss": 0.2636,
+      "step": 2839
+    },
+    {
+      "epoch": 1.5919282511210762,
+      "grad_norm": 0.07197010979102217,
+      "learning_rate": 0.00010677896735968693,
+      "loss": 0.2752,
+      "step": 2840
+    },
+    {
+      "epoch": 1.592488789237668,
+      "grad_norm": 0.07214003327677651,
+      "learning_rate": 0.00010671388357478179,
+      "loss": 0.2859,
+      "step": 2841
+    },
+    {
+      "epoch": 1.59304932735426,
+      "grad_norm": 0.07586786214389095,
+      "learning_rate": 0.00010664879693294017,
+      "loss": 0.2514,
+      "step": 2842
+    },
+    {
+      "epoch": 1.5936098654708521,
+      "grad_norm": 0.07060520451179678,
+      "learning_rate": 0.00010658370746185817,
+      "loss": 0.2679,
+      "step": 2843
+    },
+    {
+      "epoch": 1.594170403587444,
+      "grad_norm": 0.07028483497476899,
+      "learning_rate": 0.00010651861518923319,
+      "loss": 0.2706,
+      "step": 2844
+    },
+    {
+      "epoch": 1.5947309417040358,
+      "grad_norm": 0.06893105853432323,
+      "learning_rate": 0.00010645352014276364,
+      "loss": 0.2566,
+      "step": 2845
+    },
+    {
+      "epoch": 1.5952914798206277,
+      "grad_norm": 0.07169038763566891,
+      "learning_rate": 0.00010638842235014924,
+      "loss": 0.2579,
+      "step": 2846
+    },
+    {
+      "epoch": 1.5958520179372198,
+      "grad_norm": 0.07049140886089875,
+      "learning_rate": 0.0001063233218390908,
+      "loss": 0.2701,
+      "step": 2847
+    },
+    {
+      "epoch": 1.5964125560538116,
+      "grad_norm": 0.07083561940110712,
+      "learning_rate": 0.00010625821863729036,
+      "loss": 0.2764,
+      "step": 2848
+    },
+    {
+      "epoch": 1.5969730941704037,
+      "grad_norm": 0.07107293763979249,
+      "learning_rate": 0.00010619311277245104,
+      "loss": 0.2575,
+      "step": 2849
+    },
+    {
+      "epoch": 1.5975336322869955,
+      "grad_norm": 0.0701489084443605,
+      "learning_rate": 0.00010612800427227714,
+      "loss": 0.2686,
+      "step": 2850
+    },
+    {
+      "epoch": 1.5980941704035874,
+      "grad_norm": 0.07270388701694894,
+      "learning_rate": 0.00010606289316447406,
+      "loss": 0.2764,
+      "step": 2851
+    },
+    {
+      "epoch": 1.5986547085201792,
+      "grad_norm": 0.0703825627000912,
+      "learning_rate": 0.00010599777947674829,
+      "loss": 0.26,
+      "step": 2852
+    },
+    {
+      "epoch": 1.5992152466367713,
+      "grad_norm": 0.07263488626169894,
+      "learning_rate": 0.00010593266323680749,
+      "loss": 0.2674,
+      "step": 2853
+    },
+    {
+      "epoch": 1.5997757847533634,
+      "grad_norm": 0.0717526815615539,
+      "learning_rate": 0.00010586754447236031,
+      "loss": 0.2695,
+      "step": 2854
+    },
+    {
+      "epoch": 1.6003363228699552,
+      "grad_norm": 0.0693101009253264,
+      "learning_rate": 0.00010580242321111653,
+      "loss": 0.2574,
+      "step": 2855
+    },
+    {
+      "epoch": 1.600896860986547,
+      "grad_norm": 0.06893498612964494,
+      "learning_rate": 0.00010573729948078699,
+      "loss": 0.2598,
+      "step": 2856
+    },
+    {
+      "epoch": 1.601457399103139,
+      "grad_norm": 0.07264346152373237,
+      "learning_rate": 0.00010567217330908357,
+      "loss": 0.268,
+      "step": 2857
+    },
+    {
+      "epoch": 1.602017937219731,
+      "grad_norm": 0.07215474208576995,
+      "learning_rate": 0.00010560704472371919,
+      "loss": 0.2705,
+      "step": 2858
+    },
+    {
+      "epoch": 1.602578475336323,
+      "grad_norm": 0.07014321948741521,
+      "learning_rate": 0.0001055419137524078,
+      "loss": 0.2805,
+      "step": 2859
+    },
+    {
+      "epoch": 1.6031390134529149,
+      "grad_norm": 0.07162201942423144,
+      "learning_rate": 0.00010547678042286436,
+      "loss": 0.2602,
+      "step": 2860
+    },
+    {
+      "epoch": 1.6036995515695067,
+      "grad_norm": 0.0691817907815888,
+      "learning_rate": 0.00010541164476280487,
+      "loss": 0.2706,
+      "step": 2861
+    },
+    {
+      "epoch": 1.6042600896860986,
+      "grad_norm": 0.06968361003295294,
+      "learning_rate": 0.00010534650679994627,
+      "loss": 0.2656,
+      "step": 2862
+    },
+    {
+      "epoch": 1.6048206278026906,
+      "grad_norm": 0.07128643674238011,
+      "learning_rate": 0.00010528136656200647,
+      "loss": 0.2787,
+      "step": 2863
+    },
+    {
+      "epoch": 1.6053811659192825,
+      "grad_norm": 0.06960925004641096,
+      "learning_rate": 0.00010521622407670439,
+      "loss": 0.2698,
+      "step": 2864
+    },
+    {
+      "epoch": 1.6059417040358746,
+      "grad_norm": 0.07042370337542646,
+      "learning_rate": 0.00010515107937175995,
+      "loss": 0.2716,
+      "step": 2865
+    },
+    {
+      "epoch": 1.6065022421524664,
+      "grad_norm": 0.07097628524248398,
+      "learning_rate": 0.00010508593247489389,
+      "loss": 0.2716,
+      "step": 2866
+    },
+    {
+      "epoch": 1.6070627802690582,
+      "grad_norm": 0.06916798698525788,
+      "learning_rate": 0.00010502078341382797,
+      "loss": 0.2671,
+      "step": 2867
+    },
+    {
+      "epoch": 1.60762331838565,
+      "grad_norm": 0.07064265752587373,
+      "learning_rate": 0.00010495563221628486,
+      "loss": 0.2617,
+      "step": 2868
+    },
+    {
+      "epoch": 1.6081838565022422,
+      "grad_norm": 0.06920364415237995,
+      "learning_rate": 0.00010489047890998816,
+      "loss": 0.2647,
+      "step": 2869
+    },
+    {
+      "epoch": 1.6087443946188342,
+      "grad_norm": 0.07259613304903145,
+      "learning_rate": 0.00010482532352266227,
+      "loss": 0.2796,
+      "step": 2870
+    },
+    {
+      "epoch": 1.609304932735426,
+      "grad_norm": 0.07272303135725403,
+      "learning_rate": 0.0001047601660820326,
+      "loss": 0.2742,
+      "step": 2871
+    },
+    {
+      "epoch": 1.609865470852018,
+      "grad_norm": 0.07268996656414749,
+      "learning_rate": 0.00010469500661582536,
+      "loss": 0.2598,
+      "step": 2872
+    },
+    {
+      "epoch": 1.6104260089686098,
+      "grad_norm": 0.07097297630819888,
+      "learning_rate": 0.00010462984515176764,
+      "loss": 0.2794,
+      "step": 2873
+    },
+    {
+      "epoch": 1.6109865470852018,
+      "grad_norm": 0.07015583243603275,
+      "learning_rate": 0.0001045646817175874,
+      "loss": 0.2764,
+      "step": 2874
+    },
+    {
+      "epoch": 1.6115470852017937,
+      "grad_norm": 0.07061977707648315,
+      "learning_rate": 0.00010449951634101338,
+      "loss": 0.268,
+      "step": 2875
+    },
+    {
+      "epoch": 1.6121076233183858,
+      "grad_norm": 0.07129773048093416,
+      "learning_rate": 0.00010443434904977518,
+      "loss": 0.2815,
+      "step": 2876
+    },
+    {
+      "epoch": 1.6126681614349776,
+      "grad_norm": 0.07530695328656455,
+      "learning_rate": 0.00010436917987160328,
+      "loss": 0.273,
+      "step": 2877
+    },
+    {
+      "epoch": 1.6132286995515694,
+      "grad_norm": 0.06912379645418971,
+      "learning_rate": 0.00010430400883422886,
+      "loss": 0.2727,
+      "step": 2878
+    },
+    {
+      "epoch": 1.6137892376681613,
+      "grad_norm": 0.0703260999096315,
+      "learning_rate": 0.00010423883596538395,
+      "loss": 0.2696,
+      "step": 2879
+    },
+    {
+      "epoch": 1.6143497757847534,
+      "grad_norm": 0.07116881586283529,
+      "learning_rate": 0.00010417366129280133,
+      "loss": 0.2648,
+      "step": 2880
+    },
+    {
+      "epoch": 1.6149103139013454,
+      "grad_norm": 0.06900435851378543,
+      "learning_rate": 0.00010410848484421454,
+      "loss": 0.2646,
+      "step": 2881
+    },
+    {
+      "epoch": 1.6154708520179373,
+      "grad_norm": 0.07252026805511815,
+      "learning_rate": 0.00010404330664735796,
+      "loss": 0.2634,
+      "step": 2882
+    },
+    {
+      "epoch": 1.6160313901345291,
+      "grad_norm": 0.0704851611576344,
+      "learning_rate": 0.00010397812672996658,
+      "loss": 0.2651,
+      "step": 2883
+    },
+    {
+      "epoch": 1.616591928251121,
+      "grad_norm": 0.06915927703579841,
+      "learning_rate": 0.00010391294511977623,
+      "loss": 0.2615,
+      "step": 2884
+    },
+    {
+      "epoch": 1.617152466367713,
+      "grad_norm": 0.07130739293659118,
+      "learning_rate": 0.0001038477618445234,
+      "loss": 0.2555,
+      "step": 2885
+    },
+    {
+      "epoch": 1.6177130044843049,
+      "grad_norm": 0.0712896928495532,
+      "learning_rate": 0.00010378257693194535,
+      "loss": 0.2785,
+      "step": 2886
+    },
+    {
+      "epoch": 1.618273542600897,
+      "grad_norm": 0.07182611511899527,
+      "learning_rate": 0.00010371739040978,
+      "loss": 0.2774,
+      "step": 2887
+    },
+    {
+      "epoch": 1.6188340807174888,
+      "grad_norm": 0.07218038843947314,
+      "learning_rate": 0.0001036522023057659,
+      "loss": 0.2696,
+      "step": 2888
+    },
+    {
+      "epoch": 1.6193946188340806,
+      "grad_norm": 0.0708372643358692,
+      "learning_rate": 0.00010358701264764234,
+      "loss": 0.2632,
+      "step": 2889
+    },
+    {
+      "epoch": 1.6199551569506725,
+      "grad_norm": 0.07305156201554344,
+      "learning_rate": 0.00010352182146314931,
+      "loss": 0.2802,
+      "step": 2890
+    },
+    {
+      "epoch": 1.6205156950672646,
+      "grad_norm": 0.06901337263656344,
+      "learning_rate": 0.00010345662878002733,
+      "loss": 0.2647,
+      "step": 2891
+    },
+    {
+      "epoch": 1.6210762331838566,
+      "grad_norm": 0.07305095994353092,
+      "learning_rate": 0.00010339143462601768,
+      "loss": 0.2698,
+      "step": 2892
+    },
+    {
+      "epoch": 1.6216367713004485,
+      "grad_norm": 0.0718301702444084,
+      "learning_rate": 0.00010332623902886214,
+      "loss": 0.2693,
+      "step": 2893
+    },
+    {
+      "epoch": 1.6221973094170403,
+      "grad_norm": 0.07207048037056661,
+      "learning_rate": 0.00010326104201630326,
+      "loss": 0.2813,
+      "step": 2894
+    },
+    {
+      "epoch": 1.6227578475336322,
+      "grad_norm": 0.07257496358147895,
+      "learning_rate": 0.00010319584361608407,
+      "loss": 0.278,
+      "step": 2895
+    },
+    {
+      "epoch": 1.6233183856502242,
+      "grad_norm": 0.07193843740836481,
+      "learning_rate": 0.00010313064385594822,
+      "loss": 0.266,
+      "step": 2896
+    },
+    {
+      "epoch": 1.6238789237668163,
+      "grad_norm": 0.06886935555901082,
+      "learning_rate": 0.00010306544276363992,
+      "loss": 0.272,
+      "step": 2897
+    },
+    {
+      "epoch": 1.6244394618834082,
+      "grad_norm": 0.07197003880910544,
+      "learning_rate": 0.00010300024036690402,
+      "loss": 0.2625,
+      "step": 2898
+    },
+    {
+      "epoch": 1.625,
+      "grad_norm": 0.06942898499670558,
+      "learning_rate": 0.00010293503669348586,
+      "loss": 0.269,
+      "step": 2899
+    },
+    {
+      "epoch": 1.6255605381165918,
+      "grad_norm": 0.0705762892144056,
+      "learning_rate": 0.00010286983177113135,
+      "loss": 0.2698,
+      "step": 2900
+    },
+    {
+      "epoch": 1.6261210762331837,
+      "grad_norm": 0.06944684274046396,
+      "learning_rate": 0.0001028046256275869,
+      "loss": 0.2689,
+      "step": 2901
+    },
+    {
+      "epoch": 1.6266816143497758,
+      "grad_norm": 0.07229536193834671,
+      "learning_rate": 0.0001027394182905995,
+      "loss": 0.2604,
+      "step": 2902
+    },
+    {
+      "epoch": 1.6272421524663678,
+      "grad_norm": 0.07167360624516773,
+      "learning_rate": 0.00010267420978791657,
+      "loss": 0.2532,
+      "step": 2903
+    },
+    {
+      "epoch": 1.6278026905829597,
+      "grad_norm": 0.06910521396979334,
+      "learning_rate": 0.0001026090001472861,
+      "loss": 0.2666,
+      "step": 2904
+    },
+    {
+      "epoch": 1.6283632286995515,
+      "grad_norm": 0.0708341927339222,
+      "learning_rate": 0.00010254378939645648,
+      "loss": 0.2763,
+      "step": 2905
+    },
+    {
+      "epoch": 1.6289237668161434,
+      "grad_norm": 0.07150730049544997,
+      "learning_rate": 0.00010247857756317666,
+      "loss": 0.2734,
+      "step": 2906
+    },
+    {
+      "epoch": 1.6294843049327354,
+      "grad_norm": 0.06995895983103713,
+      "learning_rate": 0.00010241336467519604,
+      "loss": 0.2657,
+      "step": 2907
+    },
+    {
+      "epoch": 1.6300448430493275,
+      "grad_norm": 0.07108159682180759,
+      "learning_rate": 0.00010234815076026442,
+      "loss": 0.2645,
+      "step": 2908
+    },
+    {
+      "epoch": 1.6306053811659194,
+      "grad_norm": 0.07106786705261911,
+      "learning_rate": 0.00010228293584613203,
+      "loss": 0.2742,
+      "step": 2909
+    },
+    {
+      "epoch": 1.6311659192825112,
+      "grad_norm": 0.07098809892260202,
+      "learning_rate": 0.00010221771996054958,
+      "loss": 0.276,
+      "step": 2910
+    },
+    {
+      "epoch": 1.631726457399103,
+      "grad_norm": 0.0716960153271508,
+      "learning_rate": 0.00010215250313126817,
+      "loss": 0.2619,
+      "step": 2911
+    },
+    {
+      "epoch": 1.6322869955156951,
+      "grad_norm": 0.07146872339606528,
+      "learning_rate": 0.00010208728538603929,
+      "loss": 0.2685,
+      "step": 2912
+    },
+    {
+      "epoch": 1.632847533632287,
+      "grad_norm": 0.0693907623417422,
+      "learning_rate": 0.00010202206675261484,
+      "loss": 0.2715,
+      "step": 2913
+    },
+    {
+      "epoch": 1.633408071748879,
+      "grad_norm": 0.07049327137235907,
+      "learning_rate": 0.00010195684725874706,
+      "loss": 0.2639,
+      "step": 2914
+    },
+    {
+      "epoch": 1.6339686098654709,
+      "grad_norm": 0.07056680465844478,
+      "learning_rate": 0.00010189162693218864,
+      "loss": 0.2688,
+      "step": 2915
+    },
+    {
+      "epoch": 1.6345291479820627,
+      "grad_norm": 0.07165852741318098,
+      "learning_rate": 0.0001018264058006925,
+      "loss": 0.2876,
+      "step": 2916
+    },
+    {
+      "epoch": 1.6350896860986546,
+      "grad_norm": 0.07032431728646045,
+      "learning_rate": 0.00010176118389201201,
+      "loss": 0.2641,
+      "step": 2917
+    },
+    {
+      "epoch": 1.6356502242152466,
+      "grad_norm": 0.07229702801072894,
+      "learning_rate": 0.00010169596123390082,
+      "loss": 0.2783,
+      "step": 2918
+    },
+    {
+      "epoch": 1.6362107623318387,
+      "grad_norm": 0.06945481415257722,
+      "learning_rate": 0.0001016307378541129,
+      "loss": 0.2619,
+      "step": 2919
+    },
+    {
+      "epoch": 1.6367713004484306,
+      "grad_norm": 0.06876431346048285,
+      "learning_rate": 0.00010156551378040258,
+      "loss": 0.2782,
+      "step": 2920
+    },
+    {
+      "epoch": 1.6373318385650224,
+      "grad_norm": 0.07014776896764986,
+      "learning_rate": 0.0001015002890405244,
+      "loss": 0.2644,
+      "step": 2921
+    },
+    {
+      "epoch": 1.6378923766816142,
+      "grad_norm": 0.06792938310515426,
+      "learning_rate": 0.00010143506366223323,
+      "loss": 0.2728,
+      "step": 2922
+    },
+    {
+      "epoch": 1.6384529147982063,
+      "grad_norm": 0.06949802951722193,
+      "learning_rate": 0.00010136983767328422,
+      "loss": 0.2537,
+      "step": 2923
+    },
+    {
+      "epoch": 1.6390134529147982,
+      "grad_norm": 0.06706218003585812,
+      "learning_rate": 0.00010130461110143277,
+      "loss": 0.2631,
+      "step": 2924
+    },
+    {
+      "epoch": 1.6395739910313902,
+      "grad_norm": 0.06924991154675039,
+      "learning_rate": 0.00010123938397443451,
+      "loss": 0.2664,
+      "step": 2925
+    },
+    {
+      "epoch": 1.640134529147982,
+      "grad_norm": 0.07235014008202201,
+      "learning_rate": 0.0001011741563200453,
+      "loss": 0.2656,
+      "step": 2926
+    },
+    {
+      "epoch": 1.640695067264574,
+      "grad_norm": 0.06985174256664821,
+      "learning_rate": 0.0001011089281660213,
+      "loss": 0.2682,
+      "step": 2927
+    },
+    {
+      "epoch": 1.6412556053811658,
+      "grad_norm": 0.0700158681377797,
+      "learning_rate": 0.00010104369954011883,
+      "loss": 0.2668,
+      "step": 2928
+    },
+    {
+      "epoch": 1.6418161434977578,
+      "grad_norm": 0.07156922368606,
+      "learning_rate": 0.00010097847047009437,
+      "loss": 0.2554,
+      "step": 2929
+    },
+    {
+      "epoch": 1.64237668161435,
+      "grad_norm": 0.0719107263817642,
+      "learning_rate": 0.00010091324098370458,
+      "loss": 0.2701,
+      "step": 2930
+    },
+    {
+      "epoch": 1.6429372197309418,
+      "grad_norm": 0.07293399490486906,
+      "learning_rate": 0.00010084801110870648,
+      "loss": 0.2671,
+      "step": 2931
+    },
+    {
+      "epoch": 1.6434977578475336,
+      "grad_norm": 0.07003004697968263,
+      "learning_rate": 0.000100782780872857,
+      "loss": 0.2653,
+      "step": 2932
+    },
+    {
+      "epoch": 1.6440582959641254,
+      "grad_norm": 0.06935644692435886,
+      "learning_rate": 0.0001007175503039134,
+      "loss": 0.2501,
+      "step": 2933
+    },
+    {
+      "epoch": 1.6446188340807175,
+      "grad_norm": 0.06864500191077569,
+      "learning_rate": 0.00010065231942963305,
+      "loss": 0.2634,
+      "step": 2934
+    },
+    {
+      "epoch": 1.6451793721973094,
+      "grad_norm": 0.06940400777998981,
+      "learning_rate": 0.00010058708827777335,
+      "loss": 0.262,
+      "step": 2935
+    },
+    {
+      "epoch": 1.6457399103139014,
+      "grad_norm": 0.07022841818358101,
+      "learning_rate": 0.00010052185687609197,
+      "loss": 0.2675,
+      "step": 2936
+    },
+    {
+      "epoch": 1.6463004484304933,
+      "grad_norm": 0.06986936343805794,
+      "learning_rate": 0.00010045662525234656,
+      "loss": 0.2682,
+      "step": 2937
+    },
+    {
+      "epoch": 1.6468609865470851,
+      "grad_norm": 0.06864227560784018,
+      "learning_rate": 0.00010039139343429492,
+      "loss": 0.257,
+      "step": 2938
+    },
+    {
+      "epoch": 1.647421524663677,
+      "grad_norm": 0.07076917806402258,
+      "learning_rate": 0.00010032616144969494,
+      "loss": 0.2672,
+      "step": 2939
+    },
+    {
+      "epoch": 1.647982062780269,
+      "grad_norm": 0.06974530633846089,
+      "learning_rate": 0.00010026092932630457,
+      "loss": 0.2626,
+      "step": 2940
+    },
+    {
+      "epoch": 1.648542600896861,
+      "grad_norm": 0.06737481489290036,
+      "learning_rate": 0.00010019569709188186,
+      "loss": 0.2492,
+      "step": 2941
+    },
+    {
+      "epoch": 1.649103139013453,
+      "grad_norm": 0.07101003567831017,
+      "learning_rate": 0.00010013046477418475,
+      "loss": 0.257,
+      "step": 2942
+    },
+    {
+      "epoch": 1.6496636771300448,
+      "grad_norm": 0.07198482669475338,
+      "learning_rate": 0.00010006523240097146,
+      "loss": 0.2745,
+      "step": 2943
+    },
+    {
+      "epoch": 1.6502242152466366,
+      "grad_norm": 0.07116817086657783,
+      "learning_rate": 0.0001,
+      "loss": 0.2639,
+      "step": 2944
+    },
+    {
+      "epoch": 1.6507847533632287,
+      "grad_norm": 0.07064895989898635,
+      "learning_rate": 9.99347675990286e-05,
+      "loss": 0.264,
+      "step": 2945
+    },
+    {
+      "epoch": 1.6513452914798208,
+      "grad_norm": 0.07006740044190507,
+      "learning_rate": 9.986953522581526e-05,
+      "loss": 0.2744,
+      "step": 2946
+    },
+    {
+      "epoch": 1.6519058295964126,
+      "grad_norm": 0.06985943320704728,
+      "learning_rate": 9.980430290811818e-05,
+      "loss": 0.2547,
+      "step": 2947
+    },
+    {
+      "epoch": 1.6524663677130045,
+      "grad_norm": 0.07137157676555536,
+      "learning_rate": 9.973907067369543e-05,
+      "loss": 0.2724,
+      "step": 2948
+    },
+    {
+      "epoch": 1.6530269058295963,
+      "grad_norm": 0.07279180942706098,
+      "learning_rate": 9.967383855030509e-05,
+      "loss": 0.2759,
+      "step": 2949
+    },
+    {
+      "epoch": 1.6535874439461884,
+      "grad_norm": 0.0704488424146485,
+      "learning_rate": 9.960860656570509e-05,
+      "loss": 0.2751,
+      "step": 2950
+    },
+    {
+      "epoch": 1.6541479820627802,
+      "grad_norm": 0.07104883829550132,
+      "learning_rate": 9.954337474765347e-05,
+      "loss": 0.2755,
+      "step": 2951
+    },
+    {
+      "epoch": 1.6547085201793723,
+      "grad_norm": 0.07269547521425603,
+      "learning_rate": 9.947814312390808e-05,
+      "loss": 0.2648,
+      "step": 2952
+    },
+    {
+      "epoch": 1.6552690582959642,
+      "grad_norm": 0.07173620153129019,
+      "learning_rate": 9.941291172222666e-05,
+      "loss": 0.277,
+      "step": 2953
+    },
+    {
+      "epoch": 1.655829596412556,
+      "grad_norm": 0.07091369461295653,
+      "learning_rate": 9.934768057036699e-05,
+      "loss": 0.2728,
+      "step": 2954
+    },
+    {
+      "epoch": 1.6563901345291479,
+      "grad_norm": 0.06941523353081054,
+      "learning_rate": 9.928244969608659e-05,
+      "loss": 0.2687,
+      "step": 2955
+    },
+    {
+      "epoch": 1.65695067264574,
+      "grad_norm": 0.07021619384640268,
+      "learning_rate": 9.921721912714301e-05,
+      "loss": 0.2576,
+      "step": 2956
+    },
+    {
+      "epoch": 1.657511210762332,
+      "grad_norm": 0.07064173212862243,
+      "learning_rate": 9.915198889129353e-05,
+      "loss": 0.2635,
+      "step": 2957
+    },
+    {
+      "epoch": 1.6580717488789238,
+      "grad_norm": 0.07017826202142279,
+      "learning_rate": 9.908675901629543e-05,
+      "loss": 0.2563,
+      "step": 2958
+    },
+    {
+      "epoch": 1.6586322869955157,
+      "grad_norm": 0.07167348464219234,
+      "learning_rate": 9.902152952990568e-05,
+      "loss": 0.2757,
+      "step": 2959
+    },
+    {
+      "epoch": 1.6591928251121075,
+      "grad_norm": 0.07084110287213552,
+      "learning_rate": 9.89563004598812e-05,
+      "loss": 0.2662,
+      "step": 2960
+    },
+    {
+      "epoch": 1.6597533632286996,
+      "grad_norm": 0.07005592040282468,
+      "learning_rate": 9.889107183397872e-05,
+      "loss": 0.2584,
+      "step": 2961
+    },
+    {
+      "epoch": 1.6603139013452914,
+      "grad_norm": 0.06939271278116278,
+      "learning_rate": 9.88258436799547e-05,
+      "loss": 0.2642,
+      "step": 2962
+    },
+    {
+      "epoch": 1.6608744394618835,
+      "grad_norm": 0.07270946229402037,
+      "learning_rate": 9.876061602556552e-05,
+      "loss": 0.2709,
+      "step": 2963
+    },
+    {
+      "epoch": 1.6614349775784754,
+      "grad_norm": 0.0725881565646497,
+      "learning_rate": 9.869538889856723e-05,
+      "loss": 0.2761,
+      "step": 2964
+    },
+    {
+      "epoch": 1.6619955156950672,
+      "grad_norm": 0.07158674809084947,
+      "learning_rate": 9.86301623267158e-05,
+      "loss": 0.2689,
+      "step": 2965
+    },
+    {
+      "epoch": 1.662556053811659,
+      "grad_norm": 0.07060696614329968,
+      "learning_rate": 9.856493633776682e-05,
+      "loss": 0.2577,
+      "step": 2966
+    },
+    {
+      "epoch": 1.6631165919282511,
+      "grad_norm": 0.07180296984368345,
+      "learning_rate": 9.849971095947562e-05,
+      "loss": 0.2628,
+      "step": 2967
+    },
+    {
+      "epoch": 1.6636771300448432,
+      "grad_norm": 0.06809368516988418,
+      "learning_rate": 9.843448621959745e-05,
+      "loss": 0.2569,
+      "step": 2968
+    },
+    {
+      "epoch": 1.664237668161435,
+      "grad_norm": 0.06953623240111217,
+      "learning_rate": 9.83692621458871e-05,
+      "loss": 0.2587,
+      "step": 2969
+    },
+    {
+      "epoch": 1.6647982062780269,
+      "grad_norm": 0.07132151939492777,
+      "learning_rate": 9.830403876609922e-05,
+      "loss": 0.2677,
+      "step": 2970
+    },
+    {
+      "epoch": 1.6653587443946187,
+      "grad_norm": 0.06760232918482255,
+      "learning_rate": 9.823881610798804e-05,
+      "loss": 0.2539,
+      "step": 2971
+    },
+    {
+      "epoch": 1.6659192825112108,
+      "grad_norm": 0.06873085708011183,
+      "learning_rate": 9.817359419930751e-05,
+      "loss": 0.2645,
+      "step": 2972
+    },
+    {
+      "epoch": 1.6664798206278026,
+      "grad_norm": 0.06715588054507451,
+      "learning_rate": 9.810837306781141e-05,
+      "loss": 0.2527,
+      "step": 2973
+    },
+    {
+      "epoch": 1.6670403587443947,
+      "grad_norm": 0.07141325757822094,
+      "learning_rate": 9.804315274125295e-05,
+      "loss": 0.2632,
+      "step": 2974
+    },
+    {
+      "epoch": 1.6676008968609866,
+      "grad_norm": 0.07067485424500766,
+      "learning_rate": 9.797793324738519e-05,
+      "loss": 0.2781,
+      "step": 2975
+    },
+    {
+      "epoch": 1.6681614349775784,
+      "grad_norm": 0.07027156452442415,
+      "learning_rate": 9.79127146139607e-05,
+      "loss": 0.2716,
+      "step": 2976
+    },
+    {
+      "epoch": 1.6687219730941703,
+      "grad_norm": 0.06902926780891283,
+      "learning_rate": 9.784749686873185e-05,
+      "loss": 0.2685,
+      "step": 2977
+    },
+    {
+      "epoch": 1.6692825112107623,
+      "grad_norm": 0.07009559217339502,
+      "learning_rate": 9.778228003945047e-05,
+      "loss": 0.2579,
+      "step": 2978
+    },
+    {
+      "epoch": 1.6698430493273544,
+      "grad_norm": 0.07290110289519239,
+      "learning_rate": 9.7717064153868e-05,
+      "loss": 0.2685,
+      "step": 2979
+    },
+    {
+      "epoch": 1.6704035874439462,
+      "grad_norm": 0.07219199438100524,
+      "learning_rate": 9.765184923973561e-05,
+      "loss": 0.2663,
+      "step": 2980
+    },
+    {
+      "epoch": 1.670964125560538,
+      "grad_norm": 0.07173612240528146,
+      "learning_rate": 9.758663532480395e-05,
+      "loss": 0.262,
+      "step": 2981
+    },
+    {
+      "epoch": 1.67152466367713,
+      "grad_norm": 0.06999620724925222,
+      "learning_rate": 9.752142243682335e-05,
+      "loss": 0.2646,
+      "step": 2982
+    },
+    {
+      "epoch": 1.672085201793722,
+      "grad_norm": 0.07583412286829273,
+      "learning_rate": 9.745621060354353e-05,
+      "loss": 0.2684,
+      "step": 2983
+    },
+    {
+      "epoch": 1.672645739910314,
+      "grad_norm": 0.0717555811476974,
+      "learning_rate": 9.739099985271394e-05,
+      "loss": 0.2688,
+      "step": 2984
+    },
+    {
+      "epoch": 1.673206278026906,
+      "grad_norm": 0.071767650515776,
+      "learning_rate": 9.732579021208348e-05,
+      "loss": 0.2727,
+      "step": 2985
+    },
+    {
+      "epoch": 1.6737668161434978,
+      "grad_norm": 0.0716331600302753,
+      "learning_rate": 9.726058170940053e-05,
+      "loss": 0.2628,
+      "step": 2986
+    },
+    {
+      "epoch": 1.6743273542600896,
+      "grad_norm": 0.07198001567738212,
+      "learning_rate": 9.719537437241312e-05,
+      "loss": 0.2689,
+      "step": 2987
+    },
+    {
+      "epoch": 1.6748878923766815,
+      "grad_norm": 0.07180028215768763,
+      "learning_rate": 9.713016822886866e-05,
+      "loss": 0.2868,
+      "step": 2988
+    },
+    {
+      "epoch": 1.6754484304932735,
+      "grad_norm": 0.07328396935140088,
+      "learning_rate": 9.706496330651415e-05,
+      "loss": 0.2475,
+      "step": 2989
+    },
+    {
+      "epoch": 1.6760089686098656,
+      "grad_norm": 0.06894723066689226,
+      "learning_rate": 9.699975963309599e-05,
+      "loss": 0.266,
+      "step": 2990
+    },
+    {
+      "epoch": 1.6765695067264574,
+      "grad_norm": 0.06736843118572723,
+      "learning_rate": 9.693455723636011e-05,
+      "loss": 0.2521,
+      "step": 2991
+    },
+    {
+      "epoch": 1.6771300448430493,
+      "grad_norm": 0.07280393560519785,
+      "learning_rate": 9.686935614405183e-05,
+      "loss": 0.279,
+      "step": 2992
+    },
+    {
+      "epoch": 1.6776905829596411,
+      "grad_norm": 0.07175747202684021,
+      "learning_rate": 9.680415638391594e-05,
+      "loss": 0.2689,
+      "step": 2993
+    },
+    {
+      "epoch": 1.6782511210762332,
+      "grad_norm": 0.07098623385981147,
+      "learning_rate": 9.673895798369676e-05,
+      "loss": 0.2702,
+      "step": 2994
+    },
+    {
+      "epoch": 1.6788116591928253,
+      "grad_norm": 0.07088433029769355,
+      "learning_rate": 9.667376097113786e-05,
+      "loss": 0.2702,
+      "step": 2995
+    },
+    {
+      "epoch": 1.6793721973094171,
+      "grad_norm": 0.0704308492547985,
+      "learning_rate": 9.660856537398235e-05,
+      "loss": 0.2655,
+      "step": 2996
+    },
+    {
+      "epoch": 1.679932735426009,
+      "grad_norm": 0.06944793765819227,
+      "learning_rate": 9.654337121997266e-05,
+      "loss": 0.2773,
+      "step": 2997
+    },
+    {
+      "epoch": 1.6804932735426008,
+      "grad_norm": 0.06882161335319166,
+      "learning_rate": 9.647817853685072e-05,
+      "loss": 0.2725,
+      "step": 2998
+    },
+    {
+      "epoch": 1.6810538116591929,
+      "grad_norm": 0.06916524968769376,
+      "learning_rate": 9.641298735235768e-05,
+      "loss": 0.2716,
+      "step": 2999
+    },
+    {
+      "epoch": 1.6816143497757847,
+      "grad_norm": 0.0690816299425424,
+      "learning_rate": 9.63477976942341e-05,
+      "loss": 0.2726,
+      "step": 3000
+    },
+    {
+      "epoch": 1.6821748878923768,
+      "grad_norm": 0.07408137401644892,
+      "learning_rate": 9.628260959022004e-05,
+      "loss": 0.2793,
+      "step": 3001
+    },
+    {
+      "epoch": 1.6827354260089686,
+      "grad_norm": 0.07362027591922002,
+      "learning_rate": 9.621742306805465e-05,
+      "loss": 0.2682,
+      "step": 3002
+    },
+    {
+      "epoch": 1.6832959641255605,
+      "grad_norm": 0.07145697481318186,
+      "learning_rate": 9.615223815547662e-05,
+      "loss": 0.263,
+      "step": 3003
+    },
+    {
+      "epoch": 1.6838565022421523,
+      "grad_norm": 0.07114809621469859,
+      "learning_rate": 9.608705488022378e-05,
+      "loss": 0.2571,
+      "step": 3004
+    },
+    {
+      "epoch": 1.6844170403587444,
+      "grad_norm": 0.07275872617108507,
+      "learning_rate": 9.602187327003344e-05,
+      "loss": 0.2592,
+      "step": 3005
+    },
+    {
+      "epoch": 1.6849775784753365,
+      "grad_norm": 0.07080794287021444,
+      "learning_rate": 9.59566933526421e-05,
+      "loss": 0.2728,
+      "step": 3006
+    },
+    {
+      "epoch": 1.6855381165919283,
+      "grad_norm": 0.07142005610893062,
+      "learning_rate": 9.589151515578547e-05,
+      "loss": 0.2737,
+      "step": 3007
+    },
+    {
+      "epoch": 1.6860986547085202,
+      "grad_norm": 0.07124654403647934,
+      "learning_rate": 9.582633870719871e-05,
+      "loss": 0.2762,
+      "step": 3008
+    },
+    {
+      "epoch": 1.686659192825112,
+      "grad_norm": 0.07381357998712451,
+      "learning_rate": 9.576116403461606e-05,
+      "loss": 0.2647,
+      "step": 3009
+    },
+    {
+      "epoch": 1.687219730941704,
+      "grad_norm": 0.06931885142630366,
+      "learning_rate": 9.569599116577116e-05,
+      "loss": 0.2453,
+      "step": 3010
+    },
+    {
+      "epoch": 1.687780269058296,
+      "grad_norm": 0.07379506914416824,
+      "learning_rate": 9.563082012839676e-05,
+      "loss": 0.2739,
+      "step": 3011
+    },
+    {
+      "epoch": 1.688340807174888,
+      "grad_norm": 0.0745453623059807,
+      "learning_rate": 9.556565095022483e-05,
+      "loss": 0.2698,
+      "step": 3012
+    },
+    {
+      "epoch": 1.6889013452914798,
+      "grad_norm": 0.07108561066703549,
+      "learning_rate": 9.550048365898666e-05,
+      "loss": 0.2632,
+      "step": 3013
+    },
+    {
+      "epoch": 1.6894618834080717,
+      "grad_norm": 0.07032612375066644,
+      "learning_rate": 9.543531828241262e-05,
+      "loss": 0.2625,
+      "step": 3014
+    },
+    {
+      "epoch": 1.6900224215246635,
+      "grad_norm": 0.07015202217184319,
+      "learning_rate": 9.53701548482324e-05,
+      "loss": 0.266,
+      "step": 3015
+    },
+    {
+      "epoch": 1.6905829596412556,
+      "grad_norm": 0.07018880177928077,
+      "learning_rate": 9.530499338417465e-05,
+      "loss": 0.2697,
+      "step": 3016
+    },
+    {
+      "epoch": 1.6911434977578477,
+      "grad_norm": 0.06954117778871866,
+      "learning_rate": 9.523983391796741e-05,
+      "loss": 0.2709,
+      "step": 3017
+    },
+    {
+      "epoch": 1.6917040358744395,
+      "grad_norm": 0.06944908965735618,
+      "learning_rate": 9.517467647733776e-05,
+      "loss": 0.2636,
+      "step": 3018
+    },
+    {
+      "epoch": 1.6922645739910314,
+      "grad_norm": 0.07100586273481777,
+      "learning_rate": 9.510952109001188e-05,
+      "loss": 0.2577,
+      "step": 3019
+    },
+    {
+      "epoch": 1.6928251121076232,
+      "grad_norm": 0.07116860959799011,
+      "learning_rate": 9.504436778371515e-05,
+      "loss": 0.258,
+      "step": 3020
+    },
+    {
+      "epoch": 1.6933856502242153,
+      "grad_norm": 0.07067580662515252,
+      "learning_rate": 9.497921658617202e-05,
+      "loss": 0.2707,
+      "step": 3021
+    },
+    {
+      "epoch": 1.6939461883408071,
+      "grad_norm": 0.06929238656794127,
+      "learning_rate": 9.491406752510615e-05,
+      "loss": 0.2569,
+      "step": 3022
+    },
+    {
+      "epoch": 1.6945067264573992,
+      "grad_norm": 0.06884714893774431,
+      "learning_rate": 9.484892062824006e-05,
+      "loss": 0.2548,
+      "step": 3023
+    },
+    {
+      "epoch": 1.695067264573991,
+      "grad_norm": 0.0724877318697893,
+      "learning_rate": 9.478377592329563e-05,
+      "loss": 0.2722,
+      "step": 3024
+    },
+    {
+      "epoch": 1.6956278026905829,
+      "grad_norm": 0.07015228462062964,
+      "learning_rate": 9.471863343799357e-05,
+      "loss": 0.2699,
+      "step": 3025
+    },
+    {
+      "epoch": 1.6961883408071747,
+      "grad_norm": 0.07015815284285637,
+      "learning_rate": 9.465349320005376e-05,
+      "loss": 0.2586,
+      "step": 3026
+    },
+    {
+      "epoch": 1.6967488789237668,
+      "grad_norm": 0.06988909279333051,
+      "learning_rate": 9.458835523719515e-05,
+      "loss": 0.2649,
+      "step": 3027
+    },
+    {
+      "epoch": 1.6973094170403589,
+      "grad_norm": 0.07348641110049924,
+      "learning_rate": 9.452321957713564e-05,
+      "loss": 0.2895,
+      "step": 3028
+    },
+    {
+      "epoch": 1.6978699551569507,
+      "grad_norm": 0.07119626809216405,
+      "learning_rate": 9.445808624759222e-05,
+      "loss": 0.26,
+      "step": 3029
+    },
+    {
+      "epoch": 1.6984304932735426,
+      "grad_norm": 0.06891101524214079,
+      "learning_rate": 9.439295527628081e-05,
+      "loss": 0.2561,
+      "step": 3030
+    },
+    {
+      "epoch": 1.6989910313901344,
+      "grad_norm": 0.0723019273672455,
+      "learning_rate": 9.432782669091645e-05,
+      "loss": 0.2688,
+      "step": 3031
+    },
+    {
+      "epoch": 1.6995515695067265,
+      "grad_norm": 0.06983747657931758,
+      "learning_rate": 9.426270051921304e-05,
+      "loss": 0.2571,
+      "step": 3032
+    },
+    {
+      "epoch": 1.7001121076233185,
+      "grad_norm": 0.07132127600929675,
+      "learning_rate": 9.419757678888348e-05,
+      "loss": 0.2553,
+      "step": 3033
+    },
+    {
+      "epoch": 1.7006726457399104,
+      "grad_norm": 0.06718582215606907,
+      "learning_rate": 9.413245552763972e-05,
+      "loss": 0.252,
+      "step": 3034
+    },
+    {
+      "epoch": 1.7012331838565022,
+      "grad_norm": 0.07159437990405979,
+      "learning_rate": 9.406733676319252e-05,
+      "loss": 0.2591,
+      "step": 3035
+    },
+    {
+      "epoch": 1.701793721973094,
+      "grad_norm": 0.07098325897458493,
+      "learning_rate": 9.400222052325174e-05,
+      "loss": 0.2703,
+      "step": 3036
+    },
+    {
+      "epoch": 1.702354260089686,
+      "grad_norm": 0.06998738777762464,
+      "learning_rate": 9.393710683552596e-05,
+      "loss": 0.2649,
+      "step": 3037
+    },
+    {
+      "epoch": 1.702914798206278,
+      "grad_norm": 0.07062642596208016,
+      "learning_rate": 9.387199572772289e-05,
+      "loss": 0.2617,
+      "step": 3038
+    },
+    {
+      "epoch": 1.70347533632287,
+      "grad_norm": 0.06869741170969536,
+      "learning_rate": 9.3806887227549e-05,
+      "loss": 0.266,
+      "step": 3039
+    },
+    {
+      "epoch": 1.704035874439462,
+      "grad_norm": 0.06995979410621993,
+      "learning_rate": 9.374178136270966e-05,
+      "loss": 0.2624,
+      "step": 3040
+    },
+    {
+      "epoch": 1.7045964125560538,
+      "grad_norm": 0.07060379407487441,
+      "learning_rate": 9.367667816090923e-05,
+      "loss": 0.2584,
+      "step": 3041
+    },
+    {
+      "epoch": 1.7051569506726456,
+      "grad_norm": 0.07162770183311762,
+      "learning_rate": 9.361157764985077e-05,
+      "loss": 0.2606,
+      "step": 3042
+    },
+    {
+      "epoch": 1.7057174887892377,
+      "grad_norm": 0.07077137055125471,
+      "learning_rate": 9.354647985723639e-05,
+      "loss": 0.2609,
+      "step": 3043
+    },
+    {
+      "epoch": 1.7062780269058297,
+      "grad_norm": 0.06853811613387817,
+      "learning_rate": 9.348138481076682e-05,
+      "loss": 0.2461,
+      "step": 3044
+    },
+    {
+      "epoch": 1.7068385650224216,
+      "grad_norm": 0.07198854797865258,
+      "learning_rate": 9.341629253814185e-05,
+      "loss": 0.2545,
+      "step": 3045
+    },
+    {
+      "epoch": 1.7073991031390134,
+      "grad_norm": 0.0702773269196386,
+      "learning_rate": 9.335120306705988e-05,
+      "loss": 0.2502,
+      "step": 3046
+    },
+    {
+      "epoch": 1.7079596412556053,
+      "grad_norm": 0.07042076930798195,
+      "learning_rate": 9.328611642521824e-05,
+      "loss": 0.2615,
+      "step": 3047
+    },
+    {
+      "epoch": 1.7085201793721974,
+      "grad_norm": 0.06992299100185177,
+      "learning_rate": 9.32210326403131e-05,
+      "loss": 0.2536,
+      "step": 3048
+    },
+    {
+      "epoch": 1.7090807174887892,
+      "grad_norm": 0.07290105226536828,
+      "learning_rate": 9.315595174003922e-05,
+      "loss": 0.2902,
+      "step": 3049
+    },
+    {
+      "epoch": 1.7096412556053813,
+      "grad_norm": 0.07359894082978928,
+      "learning_rate": 9.309087375209038e-05,
+      "loss": 0.2791,
+      "step": 3050
+    },
+    {
+      "epoch": 1.7102017937219731,
+      "grad_norm": 0.07384148555694922,
+      "learning_rate": 9.302579870415891e-05,
+      "loss": 0.2857,
+      "step": 3051
+    },
+    {
+      "epoch": 1.710762331838565,
+      "grad_norm": 0.07045667499614068,
+      "learning_rate": 9.296072662393607e-05,
+      "loss": 0.281,
+      "step": 3052
+    },
+    {
+      "epoch": 1.7113228699551568,
+      "grad_norm": 0.07106247579603404,
+      "learning_rate": 9.289565753911168e-05,
+      "loss": 0.2687,
+      "step": 3053
+    },
+    {
+      "epoch": 1.7118834080717489,
+      "grad_norm": 0.07188165141857443,
+      "learning_rate": 9.283059147737438e-05,
+      "loss": 0.2695,
+      "step": 3054
+    },
+    {
+      "epoch": 1.712443946188341,
+      "grad_norm": 0.07101824771296773,
+      "learning_rate": 9.276552846641159e-05,
+      "loss": 0.2737,
+      "step": 3055
+    },
+    {
+      "epoch": 1.7130044843049328,
+      "grad_norm": 0.07179545911967246,
+      "learning_rate": 9.270046853390925e-05,
+      "loss": 0.2541,
+      "step": 3056
+    },
+    {
+      "epoch": 1.7135650224215246,
+      "grad_norm": 0.070350875098147,
+      "learning_rate": 9.263541170755219e-05,
+      "loss": 0.2696,
+      "step": 3057
+    },
+    {
+      "epoch": 1.7141255605381165,
+      "grad_norm": 0.0718526150502596,
+      "learning_rate": 9.257035801502374e-05,
+      "loss": 0.2719,
+      "step": 3058
+    },
+    {
+      "epoch": 1.7146860986547086,
+      "grad_norm": 0.07337203808787283,
+      "learning_rate": 9.250530748400603e-05,
+      "loss": 0.274,
+      "step": 3059
+    },
+    {
+      "epoch": 1.7152466367713004,
+      "grad_norm": 0.07024342762836264,
+      "learning_rate": 9.244026014217981e-05,
+      "loss": 0.2758,
+      "step": 3060
+    },
+    {
+      "epoch": 1.7158071748878925,
+      "grad_norm": 0.0674875710121854,
+      "learning_rate": 9.237521601722441e-05,
+      "loss": 0.2693,
+      "step": 3061
+    },
+    {
+      "epoch": 1.7163677130044843,
+      "grad_norm": 0.07053448479785376,
+      "learning_rate": 9.231017513681787e-05,
+      "loss": 0.2702,
+      "step": 3062
+    },
+    {
+      "epoch": 1.7169282511210762,
+      "grad_norm": 0.07003506100204805,
+      "learning_rate": 9.224513752863678e-05,
+      "loss": 0.2689,
+      "step": 3063
+    },
+    {
+      "epoch": 1.717488789237668,
+      "grad_norm": 0.06930945698949534,
+      "learning_rate": 9.218010322035647e-05,
+      "loss": 0.2737,
+      "step": 3064
+    },
+    {
+      "epoch": 1.71804932735426,
+      "grad_norm": 0.07060988877570926,
+      "learning_rate": 9.211507223965068e-05,
+      "loss": 0.2809,
+      "step": 3065
+    },
+    {
+      "epoch": 1.7186098654708521,
+      "grad_norm": 0.07016350349691232,
+      "learning_rate": 9.205004461419183e-05,
+      "loss": 0.2654,
+      "step": 3066
+    },
+    {
+      "epoch": 1.719170403587444,
+      "grad_norm": 0.07060481235737551,
+      "learning_rate": 9.198502037165099e-05,
+      "loss": 0.2627,
+      "step": 3067
+    },
+    {
+      "epoch": 1.7197309417040358,
+      "grad_norm": 0.07180419301130528,
+      "learning_rate": 9.19199995396976e-05,
+      "loss": 0.2655,
+      "step": 3068
+    },
+    {
+      "epoch": 1.7202914798206277,
+      "grad_norm": 0.07012083865224054,
+      "learning_rate": 9.185498214599986e-05,
+      "loss": 0.2609,
+      "step": 3069
+    },
+    {
+      "epoch": 1.7208520179372198,
+      "grad_norm": 0.07075057742476638,
+      "learning_rate": 9.17899682182243e-05,
+      "loss": 0.2651,
+      "step": 3070
+    },
+    {
+      "epoch": 1.7214125560538116,
+      "grad_norm": 0.07130713360383625,
+      "learning_rate": 9.172495778403616e-05,
+      "loss": 0.2594,
+      "step": 3071
+    },
+    {
+      "epoch": 1.7219730941704037,
+      "grad_norm": 0.07164578949602156,
+      "learning_rate": 9.165995087109911e-05,
+      "loss": 0.2628,
+      "step": 3072
+    },
+    {
+      "epoch": 1.7225336322869955,
+      "grad_norm": 0.07054279941227877,
+      "learning_rate": 9.159494750707526e-05,
+      "loss": 0.2672,
+      "step": 3073
+    },
+    {
+      "epoch": 1.7230941704035874,
+      "grad_norm": 0.07438277723282749,
+      "learning_rate": 9.152994771962534e-05,
+      "loss": 0.2721,
+      "step": 3074
+    },
+    {
+      "epoch": 1.7236547085201792,
+      "grad_norm": 0.0702279516844392,
+      "learning_rate": 9.146495153640843e-05,
+      "loss": 0.2763,
+      "step": 3075
+    },
+    {
+      "epoch": 1.7242152466367713,
+      "grad_norm": 0.07202764128466099,
+      "learning_rate": 9.139995898508223e-05,
+      "loss": 0.2651,
+      "step": 3076
+    },
+    {
+      "epoch": 1.7247757847533634,
+      "grad_norm": 0.0702625810768685,
+      "learning_rate": 9.13349700933027e-05,
+      "loss": 0.2529,
+      "step": 3077
+    },
+    {
+      "epoch": 1.7253363228699552,
+      "grad_norm": 0.07210497296936554,
+      "learning_rate": 9.126998488872445e-05,
+      "loss": 0.2716,
+      "step": 3078
+    },
+    {
+      "epoch": 1.725896860986547,
+      "grad_norm": 0.06899408142272777,
+      "learning_rate": 9.120500339900034e-05,
+      "loss": 0.2722,
+      "step": 3079
+    },
+    {
+      "epoch": 1.726457399103139,
+      "grad_norm": 0.06848898381848725,
+      "learning_rate": 9.114002565178172e-05,
+      "loss": 0.2563,
+      "step": 3080
+    },
+    {
+      "epoch": 1.727017937219731,
+      "grad_norm": 0.06860275806120515,
+      "learning_rate": 9.107505167471842e-05,
+      "loss": 0.2679,
+      "step": 3081
+    },
+    {
+      "epoch": 1.727578475336323,
+      "grad_norm": 0.06798358697582003,
+      "learning_rate": 9.101008149545853e-05,
+      "loss": 0.2631,
+      "step": 3082
+    },
+    {
+      "epoch": 1.7281390134529149,
+      "grad_norm": 0.06660575541074609,
+      "learning_rate": 9.094511514164865e-05,
+      "loss": 0.2656,
+      "step": 3083
+    },
+    {
+      "epoch": 1.7286995515695067,
+      "grad_norm": 0.07050352657958711,
+      "learning_rate": 9.088015264093365e-05,
+      "loss": 0.2577,
+      "step": 3084
+    },
+    {
+      "epoch": 1.7292600896860986,
+      "grad_norm": 0.06775868625679991,
+      "learning_rate": 9.081519402095686e-05,
+      "loss": 0.2547,
+      "step": 3085
+    },
+    {
+      "epoch": 1.7298206278026906,
+      "grad_norm": 0.07065844362425264,
+      "learning_rate": 9.075023930935986e-05,
+      "loss": 0.2622,
+      "step": 3086
+    },
+    {
+      "epoch": 1.7303811659192825,
+      "grad_norm": 0.06864780010863963,
+      "learning_rate": 9.068528853378259e-05,
+      "loss": 0.2649,
+      "step": 3087
+    },
+    {
+      "epoch": 1.7309417040358746,
+      "grad_norm": 0.06857036119308015,
+      "learning_rate": 9.062034172186341e-05,
+      "loss": 0.254,
+      "step": 3088
+    },
+    {
+      "epoch": 1.7315022421524664,
+      "grad_norm": 0.07050490876479314,
+      "learning_rate": 9.055539890123884e-05,
+      "loss": 0.2561,
+      "step": 3089
+    },
+    {
+      "epoch": 1.7320627802690582,
+      "grad_norm": 0.07090601308162996,
+      "learning_rate": 9.049046009954386e-05,
+      "loss": 0.2698,
+      "step": 3090
+    },
+    {
+      "epoch": 1.73262331838565,
+      "grad_norm": 0.06789021117057008,
+      "learning_rate": 9.042552534441158e-05,
+      "loss": 0.2612,
+      "step": 3091
+    },
+    {
+      "epoch": 1.7331838565022422,
+      "grad_norm": 0.0699997888142276,
+      "learning_rate": 9.036059466347354e-05,
+      "loss": 0.2642,
+      "step": 3092
+    },
+    {
+      "epoch": 1.7337443946188342,
+      "grad_norm": 0.06998588770115205,
+      "learning_rate": 9.029566808435947e-05,
+      "loss": 0.2645,
+      "step": 3093
+    },
+    {
+      "epoch": 1.734304932735426,
+      "grad_norm": 0.07462279254263574,
+      "learning_rate": 9.023074563469729e-05,
+      "loss": 0.2744,
+      "step": 3094
+    },
+    {
+      "epoch": 1.734865470852018,
+      "grad_norm": 0.0702336911996288,
+      "learning_rate": 9.016582734211329e-05,
+      "loss": 0.2625,
+      "step": 3095
+    },
+    {
+      "epoch": 1.7354260089686098,
+      "grad_norm": 0.0688000072136554,
+      "learning_rate": 9.010091323423192e-05,
+      "loss": 0.2587,
+      "step": 3096
+    },
+    {
+      "epoch": 1.7359865470852018,
+      "grad_norm": 0.07136428883923977,
+      "learning_rate": 9.00360033386759e-05,
+      "loss": 0.2684,
+      "step": 3097
+    },
+    {
+      "epoch": 1.7365470852017937,
+      "grad_norm": 0.06878614529989564,
+      "learning_rate": 8.997109768306607e-05,
+      "loss": 0.2563,
+      "step": 3098
+    },
+    {
+      "epoch": 1.7371076233183858,
+      "grad_norm": 0.07023414067608712,
+      "learning_rate": 8.990619629502151e-05,
+      "loss": 0.2684,
+      "step": 3099
+    },
+    {
+      "epoch": 1.7376681614349776,
+      "grad_norm": 0.07091769076032442,
+      "learning_rate": 8.984129920215959e-05,
+      "loss": 0.2577,
+      "step": 3100
+    },
+    {
+      "epoch": 1.7382286995515694,
+      "grad_norm": 0.07165391074392734,
+      "learning_rate": 8.97764064320956e-05,
+      "loss": 0.2652,
+      "step": 3101
+    },
+    {
+      "epoch": 1.7387892376681613,
+      "grad_norm": 0.06756220303399435,
+      "learning_rate": 8.971151801244328e-05,
+      "loss": 0.2483,
+      "step": 3102
+    },
+    {
+      "epoch": 1.7393497757847534,
+      "grad_norm": 0.07083035336483604,
+      "learning_rate": 8.964663397081427e-05,
+      "loss": 0.2633,
+      "step": 3103
+    },
+    {
+      "epoch": 1.7399103139013454,
+      "grad_norm": 0.06898172477509441,
+      "learning_rate": 8.958175433481855e-05,
+      "loss": 0.2535,
+      "step": 3104
+    },
+    {
+      "epoch": 1.7404708520179373,
+      "grad_norm": 0.07441487339332731,
+      "learning_rate": 8.951687913206412e-05,
+      "loss": 0.2773,
+      "step": 3105
+    },
+    {
+      "epoch": 1.7410313901345291,
+      "grad_norm": 0.07092710174526971,
+      "learning_rate": 8.945200839015704e-05,
+      "loss": 0.2683,
+      "step": 3106
+    },
+    {
+      "epoch": 1.741591928251121,
+      "grad_norm": 0.07098459363992686,
+      "learning_rate": 8.938714213670161e-05,
+      "loss": 0.2712,
+      "step": 3107
+    },
+    {
+      "epoch": 1.742152466367713,
+      "grad_norm": 0.07073516711074394,
+      "learning_rate": 8.93222803993001e-05,
+      "loss": 0.2629,
+      "step": 3108
+    },
+    {
+      "epoch": 1.7427130044843049,
+      "grad_norm": 0.07019086756585276,
+      "learning_rate": 8.925742320555299e-05,
+      "loss": 0.2646,
+      "step": 3109
+    },
+    {
+      "epoch": 1.743273542600897,
+      "grad_norm": 0.07198470245921795,
+      "learning_rate": 8.919257058305865e-05,
+      "loss": 0.2711,
+      "step": 3110
+    },
+    {
+      "epoch": 1.7438340807174888,
+      "grad_norm": 0.07072514162027327,
+      "learning_rate": 8.912772255941366e-05,
+      "loss": 0.271,
+      "step": 3111
+    },
+    {
+      "epoch": 1.7443946188340806,
+      "grad_norm": 0.07267649599137432,
+      "learning_rate": 8.906287916221259e-05,
+      "loss": 0.2743,
+      "step": 3112
+    },
+    {
+      "epoch": 1.7449551569506725,
+      "grad_norm": 0.067678361706435,
+      "learning_rate": 8.899804041904795e-05,
+      "loss": 0.2591,
+      "step": 3113
+    },
+    {
+      "epoch": 1.7455156950672646,
+      "grad_norm": 0.0706536193841887,
+      "learning_rate": 8.893320635751046e-05,
+      "loss": 0.2698,
+      "step": 3114
+    },
+    {
+      "epoch": 1.7460762331838566,
+      "grad_norm": 0.07012229400985279,
+      "learning_rate": 8.886837700518867e-05,
+      "loss": 0.2673,
+      "step": 3115
+    },
+    {
+      "epoch": 1.7466367713004485,
+      "grad_norm": 0.07104279219256676,
+      "learning_rate": 8.880355238966923e-05,
+      "loss": 0.268,
+      "step": 3116
+    },
+    {
+      "epoch": 1.7471973094170403,
+      "grad_norm": 0.07219760103078716,
+      "learning_rate": 8.873873253853671e-05,
+      "loss": 0.2645,
+      "step": 3117
+    },
+    {
+      "epoch": 1.7477578475336322,
+      "grad_norm": 0.07006623090942019,
+      "learning_rate": 8.867391747937375e-05,
+      "loss": 0.2667,
+      "step": 3118
+    },
+    {
+      "epoch": 1.7483183856502242,
+      "grad_norm": 0.07025344403982556,
+      "learning_rate": 8.860910723976082e-05,
+      "loss": 0.2689,
+      "step": 3119
+    },
+    {
+      "epoch": 1.7488789237668163,
+      "grad_norm": 0.07179975973608532,
+      "learning_rate": 8.85443018472764e-05,
+      "loss": 0.2741,
+      "step": 3120
+    },
+    {
+      "epoch": 1.7494394618834082,
+      "grad_norm": 0.06699987628425405,
+      "learning_rate": 8.8479501329497e-05,
+      "loss": 0.2583,
+      "step": 3121
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.070404933115905,
+      "learning_rate": 8.841470571399685e-05,
+      "loss": 0.2612,
+      "step": 3122
+    },
+    {
+      "epoch": 1.7505605381165918,
+      "grad_norm": 0.0709969851086739,
+      "learning_rate": 8.834991502834834e-05,
+      "loss": 0.2675,
+      "step": 3123
+    },
+    {
+      "epoch": 1.7511210762331837,
+      "grad_norm": 0.07039249756051098,
+      "learning_rate": 8.82851293001215e-05,
+      "loss": 0.255,
+      "step": 3124
+    },
+    {
+      "epoch": 1.7516816143497758,
+      "grad_norm": 0.07152062412816956,
+      "learning_rate": 8.822034855688447e-05,
+      "loss": 0.2696,
+      "step": 3125
+    },
+    {
+      "epoch": 1.7522421524663678,
+      "grad_norm": 0.06780549183692385,
+      "learning_rate": 8.815557282620319e-05,
+      "loss": 0.2663,
+      "step": 3126
+    },
+    {
+      "epoch": 1.7528026905829597,
+      "grad_norm": 0.0711071595044796,
+      "learning_rate": 8.809080213564138e-05,
+      "loss": 0.2634,
+      "step": 3127
+    },
+    {
+      "epoch": 1.7533632286995515,
+      "grad_norm": 0.07121032137984734,
+      "learning_rate": 8.802603651276078e-05,
+      "loss": 0.2636,
+      "step": 3128
+    },
+    {
+      "epoch": 1.7539237668161434,
+      "grad_norm": 0.0721488260548215,
+      "learning_rate": 8.796127598512083e-05,
+      "loss": 0.2729,
+      "step": 3129
+    },
+    {
+      "epoch": 1.7544843049327354,
+      "grad_norm": 0.07154160384180487,
+      "learning_rate": 8.789652058027893e-05,
+      "loss": 0.2637,
+      "step": 3130
+    },
+    {
+      "epoch": 1.7550448430493275,
+      "grad_norm": 0.07031801849017648,
+      "learning_rate": 8.783177032579016e-05,
+      "loss": 0.2466,
+      "step": 3131
+    },
+    {
+      "epoch": 1.7556053811659194,
+      "grad_norm": 0.07176124650491238,
+      "learning_rate": 8.77670252492075e-05,
+      "loss": 0.2682,
+      "step": 3132
+    },
+    {
+      "epoch": 1.7561659192825112,
+      "grad_norm": 0.06964915955219239,
+      "learning_rate": 8.770228537808176e-05,
+      "loss": 0.256,
+      "step": 3133
+    },
+    {
+      "epoch": 1.756726457399103,
+      "grad_norm": 0.0723981284459236,
+      "learning_rate": 8.763755073996138e-05,
+      "loss": 0.2527,
+      "step": 3134
+    },
+    {
+      "epoch": 1.7572869955156951,
+      "grad_norm": 0.07128482933734852,
+      "learning_rate": 8.757282136239278e-05,
+      "loss": 0.2718,
+      "step": 3135
+    },
+    {
+      "epoch": 1.757847533632287,
+      "grad_norm": 0.0693705470171095,
+      "learning_rate": 8.750809727291995e-05,
+      "loss": 0.2546,
+      "step": 3136
+    },
+    {
+      "epoch": 1.758408071748879,
+      "grad_norm": 0.06992921067552417,
+      "learning_rate": 8.744337849908475e-05,
+      "loss": 0.2674,
+      "step": 3137
+    },
+    {
+      "epoch": 1.7589686098654709,
+      "grad_norm": 0.06856428122737501,
+      "learning_rate": 8.737866506842678e-05,
+      "loss": 0.2648,
+      "step": 3138
+    },
+    {
+      "epoch": 1.7595291479820627,
+      "grad_norm": 0.07136748675182866,
+      "learning_rate": 8.731395700848325e-05,
+      "loss": 0.2671,
+      "step": 3139
+    },
+    {
+      "epoch": 1.7600896860986546,
+      "grad_norm": 0.06964346593155742,
+      "learning_rate": 8.724925434678923e-05,
+      "loss": 0.2747,
+      "step": 3140
+    },
+    {
+      "epoch": 1.7606502242152466,
+      "grad_norm": 0.07362594572182111,
+      "learning_rate": 8.718455711087738e-05,
+      "loss": 0.2778,
+      "step": 3141
+    },
+    {
+      "epoch": 1.7612107623318387,
+      "grad_norm": 0.07093386207693751,
+      "learning_rate": 8.711986532827818e-05,
+      "loss": 0.2651,
+      "step": 3142
+    },
+    {
+      "epoch": 1.7617713004484306,
+      "grad_norm": 0.06804162479263667,
+      "learning_rate": 8.70551790265196e-05,
+      "loss": 0.2685,
+      "step": 3143
+    },
+    {
+      "epoch": 1.7623318385650224,
+      "grad_norm": 0.06947569097608512,
+      "learning_rate": 8.699049823312748e-05,
+      "loss": 0.2634,
+      "step": 3144
+    },
+    {
+      "epoch": 1.7628923766816142,
+      "grad_norm": 0.07080636522352704,
+      "learning_rate": 8.692582297562517e-05,
+      "loss": 0.2524,
+      "step": 3145
+    },
+    {
+      "epoch": 1.7634529147982063,
+      "grad_norm": 0.07231716313744631,
+      "learning_rate": 8.68611532815337e-05,
+      "loss": 0.262,
+      "step": 3146
+    },
+    {
+      "epoch": 1.7640134529147982,
+      "grad_norm": 0.07357365761579043,
+      "learning_rate": 8.679648917837183e-05,
+      "loss": 0.2728,
+      "step": 3147
+    },
+    {
+      "epoch": 1.7645739910313902,
+      "grad_norm": 0.07469380384489484,
+      "learning_rate": 8.673183069365574e-05,
+      "loss": 0.2812,
+      "step": 3148
+    },
+    {
+      "epoch": 1.765134529147982,
+      "grad_norm": 0.07139128759640609,
+      "learning_rate": 8.666717785489946e-05,
+      "loss": 0.2785,
+      "step": 3149
+    },
+    {
+      "epoch": 1.765695067264574,
+      "grad_norm": 0.07104476454836627,
+      "learning_rate": 8.660253068961439e-05,
+      "loss": 0.2652,
+      "step": 3150
+    },
+    {
+      "epoch": 1.7662556053811658,
+      "grad_norm": 0.06873181865782436,
+      "learning_rate": 8.653788922530972e-05,
+      "loss": 0.256,
+      "step": 3151
+    },
+    {
+      "epoch": 1.7668161434977578,
+      "grad_norm": 0.07082589489465703,
+      "learning_rate": 8.647325348949206e-05,
+      "loss": 0.2729,
+      "step": 3152
+    },
+    {
+      "epoch": 1.76737668161435,
+      "grad_norm": 0.06892885881911734,
+      "learning_rate": 8.640862350966561e-05,
+      "loss": 0.2586,
+      "step": 3153
+    },
+    {
+      "epoch": 1.7679372197309418,
+      "grad_norm": 0.07005436255428589,
+      "learning_rate": 8.634399931333226e-05,
+      "loss": 0.2698,
+      "step": 3154
+    },
+    {
+      "epoch": 1.7684977578475336,
+      "grad_norm": 0.06872973363314161,
+      "learning_rate": 8.62793809279912e-05,
+      "loss": 0.2658,
+      "step": 3155
+    },
+    {
+      "epoch": 1.7690582959641254,
+      "grad_norm": 0.06728216748363262,
+      "learning_rate": 8.621476838113937e-05,
+      "loss": 0.2581,
+      "step": 3156
+    },
+    {
+      "epoch": 1.7696188340807175,
+      "grad_norm": 0.07111028155058144,
+      "learning_rate": 8.615016170027105e-05,
+      "loss": 0.2602,
+      "step": 3157
+    },
+    {
+      "epoch": 1.7701793721973094,
+      "grad_norm": 0.06848589057565128,
+      "learning_rate": 8.608556091287816e-05,
+      "loss": 0.2607,
+      "step": 3158
+    },
+    {
+      "epoch": 1.7707399103139014,
+      "grad_norm": 0.07143064266989885,
+      "learning_rate": 8.602096604645009e-05,
+      "loss": 0.2593,
+      "step": 3159
+    },
+    {
+      "epoch": 1.7713004484304933,
+      "grad_norm": 0.06741105563703916,
+      "learning_rate": 8.595637712847358e-05,
+      "loss": 0.2593,
+      "step": 3160
+    },
+    {
+      "epoch": 1.7718609865470851,
+      "grad_norm": 0.06956810942187591,
+      "learning_rate": 8.5891794186433e-05,
+      "loss": 0.2605,
+      "step": 3161
+    },
+    {
+      "epoch": 1.772421524663677,
+      "grad_norm": 0.07109752640084324,
+      "learning_rate": 8.582721724781009e-05,
+      "loss": 0.2699,
+      "step": 3162
+    },
+    {
+      "epoch": 1.772982062780269,
+      "grad_norm": 0.07249070737198052,
+      "learning_rate": 8.576264634008413e-05,
+      "loss": 0.2589,
+      "step": 3163
+    },
+    {
+      "epoch": 1.773542600896861,
+      "grad_norm": 0.0704968621730139,
+      "learning_rate": 8.569808149073163e-05,
+      "loss": 0.2623,
+      "step": 3164
+    },
+    {
+      "epoch": 1.774103139013453,
+      "grad_norm": 0.06983540091279966,
+      "learning_rate": 8.563352272722678e-05,
+      "loss": 0.2748,
+      "step": 3165
+    },
+    {
+      "epoch": 1.7746636771300448,
+      "grad_norm": 0.07084953142000845,
+      "learning_rate": 8.556897007704101e-05,
+      "loss": 0.2666,
+      "step": 3166
+    },
+    {
+      "epoch": 1.7752242152466366,
+      "grad_norm": 0.07175664841858653,
+      "learning_rate": 8.550442356764314e-05,
+      "loss": 0.2766,
+      "step": 3167
+    },
+    {
+      "epoch": 1.7757847533632287,
+      "grad_norm": 0.07035713859999491,
+      "learning_rate": 8.543988322649954e-05,
+      "loss": 0.2624,
+      "step": 3168
+    },
+    {
+      "epoch": 1.7763452914798208,
+      "grad_norm": 0.06880609784502839,
+      "learning_rate": 8.537534908107373e-05,
+      "loss": 0.2585,
+      "step": 3169
+    },
+    {
+      "epoch": 1.7769058295964126,
+      "grad_norm": 0.07240736256877163,
+      "learning_rate": 8.531082115882679e-05,
+      "loss": 0.2684,
+      "step": 3170
+    },
+    {
+      "epoch": 1.7774663677130045,
+      "grad_norm": 0.07040923173470649,
+      "learning_rate": 8.524629948721701e-05,
+      "loss": 0.2647,
+      "step": 3171
+    },
+    {
+      "epoch": 1.7780269058295963,
+      "grad_norm": 0.07112543261971123,
+      "learning_rate": 8.518178409370017e-05,
+      "loss": 0.2592,
+      "step": 3172
+    },
+    {
+      "epoch": 1.7785874439461884,
+      "grad_norm": 0.07326574347624072,
+      "learning_rate": 8.511727500572923e-05,
+      "loss": 0.2722,
+      "step": 3173
+    },
+    {
+      "epoch": 1.7791479820627802,
+      "grad_norm": 0.06914195983997955,
+      "learning_rate": 8.505277225075449e-05,
+      "loss": 0.2562,
+      "step": 3174
+    },
+    {
+      "epoch": 1.7797085201793723,
+      "grad_norm": 0.07049569779725122,
+      "learning_rate": 8.498827585622368e-05,
+      "loss": 0.2817,
+      "step": 3175
+    },
+    {
+      "epoch": 1.7802690582959642,
+      "grad_norm": 0.0722256599595862,
+      "learning_rate": 8.492378584958164e-05,
+      "loss": 0.2677,
+      "step": 3176
+    },
+    {
+      "epoch": 1.780829596412556,
+      "grad_norm": 0.06888788626499369,
+      "learning_rate": 8.485930225827067e-05,
+      "loss": 0.2668,
+      "step": 3177
+    },
+    {
+      "epoch": 1.7813901345291479,
+      "grad_norm": 0.06999229867299883,
+      "learning_rate": 8.47948251097302e-05,
+      "loss": 0.26,
+      "step": 3178
+    },
+    {
+      "epoch": 1.78195067264574,
+      "grad_norm": 0.07105075343537055,
+      "learning_rate": 8.473035443139704e-05,
+      "loss": 0.2686,
+      "step": 3179
+    },
+    {
+      "epoch": 1.782511210762332,
+      "grad_norm": 0.06985758955852764,
+      "learning_rate": 8.466589025070513e-05,
+      "loss": 0.2545,
+      "step": 3180
+    },
+    {
+      "epoch": 1.7830717488789238,
+      "grad_norm": 0.06889826978128506,
+      "learning_rate": 8.460143259508569e-05,
+      "loss": 0.2624,
+      "step": 3181
+    },
+    {
+      "epoch": 1.7836322869955157,
+      "grad_norm": 0.06665555794591822,
+      "learning_rate": 8.45369814919672e-05,
+      "loss": 0.2429,
+      "step": 3182
+    },
+    {
+      "epoch": 1.7841928251121075,
+      "grad_norm": 0.06866345022967978,
+      "learning_rate": 8.447253696877529e-05,
+      "loss": 0.2576,
+      "step": 3183
+    },
+    {
+      "epoch": 1.7847533632286996,
+      "grad_norm": 0.07451630210940821,
+      "learning_rate": 8.440809905293288e-05,
+      "loss": 0.2729,
+      "step": 3184
+    },
+    {
+      "epoch": 1.7853139013452914,
+      "grad_norm": 0.07147842306947091,
+      "learning_rate": 8.434366777185999e-05,
+      "loss": 0.2519,
+      "step": 3185
+    },
+    {
+      "epoch": 1.7858744394618835,
+      "grad_norm": 0.07091977993095579,
+      "learning_rate": 8.42792431529738e-05,
+      "loss": 0.2714,
+      "step": 3186
+    },
+    {
+      "epoch": 1.7864349775784754,
+      "grad_norm": 0.06926473478972674,
+      "learning_rate": 8.42148252236888e-05,
+      "loss": 0.2642,
+      "step": 3187
+    },
+    {
+      "epoch": 1.7869955156950672,
+      "grad_norm": 0.07117021031936455,
+      "learning_rate": 8.415041401141642e-05,
+      "loss": 0.255,
+      "step": 3188
+    },
+    {
+      "epoch": 1.787556053811659,
+      "grad_norm": 0.07006286138561699,
+      "learning_rate": 8.408600954356548e-05,
+      "loss": 0.2663,
+      "step": 3189
+    },
+    {
+      "epoch": 1.7881165919282511,
+      "grad_norm": 0.0794701824361796,
+      "learning_rate": 8.402161184754167e-05,
+      "loss": 0.2498,
+      "step": 3190
+    },
+    {
+      "epoch": 1.7886771300448432,
+      "grad_norm": 0.0710591466676107,
+      "learning_rate": 8.395722095074802e-05,
+      "loss": 0.2641,
+      "step": 3191
+    },
+    {
+      "epoch": 1.789237668161435,
+      "grad_norm": 0.06852756842054991,
+      "learning_rate": 8.389283688058454e-05,
+      "loss": 0.2615,
+      "step": 3192
+    },
+    {
+      "epoch": 1.7897982062780269,
+      "grad_norm": 0.06920898255021925,
+      "learning_rate": 8.382845966444835e-05,
+      "loss": 0.2693,
+      "step": 3193
+    },
+    {
+      "epoch": 1.7903587443946187,
+      "grad_norm": 0.06997216796862439,
+      "learning_rate": 8.376408932973368e-05,
+      "loss": 0.2739,
+      "step": 3194
+    },
+    {
+      "epoch": 1.7909192825112108,
+      "grad_norm": 0.07271591113089486,
+      "learning_rate": 8.369972590383184e-05,
+      "loss": 0.2687,
+      "step": 3195
+    },
+    {
+      "epoch": 1.7914798206278026,
+      "grad_norm": 0.06965632588414579,
+      "learning_rate": 8.363536941413121e-05,
+      "loss": 0.2685,
+      "step": 3196
+    },
+    {
+      "epoch": 1.7920403587443947,
+      "grad_norm": 0.06924543112339039,
+      "learning_rate": 8.35710198880171e-05,
+      "loss": 0.2752,
+      "step": 3197
+    },
+    {
+      "epoch": 1.7926008968609866,
+      "grad_norm": 0.07127109557181958,
+      "learning_rate": 8.350667735287204e-05,
+      "loss": 0.2739,
+      "step": 3198
+    },
+    {
+      "epoch": 1.7931614349775784,
+      "grad_norm": 0.07082010008262021,
+      "learning_rate": 8.344234183607548e-05,
+      "loss": 0.2784,
+      "step": 3199
+    },
+    {
+      "epoch": 1.7937219730941703,
+      "grad_norm": 0.06889189876720758,
+      "learning_rate": 8.337801336500383e-05,
+      "loss": 0.2679,
+      "step": 3200
+    },
+    {
+      "epoch": 1.7942825112107623,
+      "grad_norm": 0.07111508795108079,
+      "learning_rate": 8.331369196703067e-05,
+      "loss": 0.2666,
+      "step": 3201
+    },
+    {
+      "epoch": 1.7948430493273544,
+      "grad_norm": 0.07047617931067245,
+      "learning_rate": 8.324937766952638e-05,
+      "loss": 0.2714,
+      "step": 3202
+    },
+    {
+      "epoch": 1.7954035874439462,
+      "grad_norm": 0.06829611267382257,
+      "learning_rate": 8.318507049985844e-05,
+      "loss": 0.256,
+      "step": 3203
+    },
+    {
+      "epoch": 1.795964125560538,
+      "grad_norm": 0.0714382509113834,
+      "learning_rate": 8.312077048539126e-05,
+      "loss": 0.2641,
+      "step": 3204
+    },
+    {
+      "epoch": 1.79652466367713,
+      "grad_norm": 0.07127447028914165,
+      "learning_rate": 8.305647765348628e-05,
+      "loss": 0.2702,
+      "step": 3205
+    },
+    {
+      "epoch": 1.797085201793722,
+      "grad_norm": 0.07082535121720834,
+      "learning_rate": 8.299219203150172e-05,
+      "loss": 0.2533,
+      "step": 3206
+    },
+    {
+      "epoch": 1.797645739910314,
+      "grad_norm": 0.06724754996951594,
+      "learning_rate": 8.292791364679284e-05,
+      "loss": 0.2639,
+      "step": 3207
+    },
+    {
+      "epoch": 1.798206278026906,
+      "grad_norm": 0.07388332110536336,
+      "learning_rate": 8.286364252671187e-05,
+      "loss": 0.2783,
+      "step": 3208
+    },
+    {
+      "epoch": 1.7987668161434978,
+      "grad_norm": 0.07227585955328372,
+      "learning_rate": 8.27993786986078e-05,
+      "loss": 0.2666,
+      "step": 3209
+    },
+    {
+      "epoch": 1.7993273542600896,
+      "grad_norm": 0.07033468100949353,
+      "learning_rate": 8.273512218982667e-05,
+      "loss": 0.2706,
+      "step": 3210
+    },
+    {
+      "epoch": 1.7998878923766815,
+      "grad_norm": 0.06979323578283012,
+      "learning_rate": 8.267087302771127e-05,
+      "loss": 0.2672,
+      "step": 3211
+    },
+    {
+      "epoch": 1.8004484304932735,
+      "grad_norm": 0.06808999277656827,
+      "learning_rate": 8.260663123960143e-05,
+      "loss": 0.2583,
+      "step": 3212
+    },
+    {
+      "epoch": 1.8010089686098656,
+      "grad_norm": 0.06981524626852194,
+      "learning_rate": 8.254239685283369e-05,
+      "loss": 0.2621,
+      "step": 3213
+    },
+    {
+      "epoch": 1.8015695067264574,
+      "grad_norm": 0.06768463125622527,
+      "learning_rate": 8.247816989474144e-05,
+      "loss": 0.259,
+      "step": 3214
+    },
+    {
+      "epoch": 1.8021300448430493,
+      "grad_norm": 0.07103339223893138,
+      "learning_rate": 8.241395039265504e-05,
+      "loss": 0.2707,
+      "step": 3215
+    },
+    {
+      "epoch": 1.8026905829596411,
+      "grad_norm": 0.07168521210758859,
+      "learning_rate": 8.234973837390154e-05,
+      "loss": 0.2655,
+      "step": 3216
+    },
+    {
+      "epoch": 1.8032511210762332,
+      "grad_norm": 0.07160453714399535,
+      "learning_rate": 8.228553386580496e-05,
+      "loss": 0.2698,
+      "step": 3217
+    },
+    {
+      "epoch": 1.8038116591928253,
+      "grad_norm": 0.07050039742585491,
+      "learning_rate": 8.22213368956859e-05,
+      "loss": 0.2751,
+      "step": 3218
+    },
+    {
+      "epoch": 1.8043721973094171,
+      "grad_norm": 0.07036432486773841,
+      "learning_rate": 8.215714749086199e-05,
+      "loss": 0.2742,
+      "step": 3219
+    },
+    {
+      "epoch": 1.804932735426009,
+      "grad_norm": 0.06606636055771409,
+      "learning_rate": 8.209296567864752e-05,
+      "loss": 0.2424,
+      "step": 3220
+    },
+    {
+      "epoch": 1.8054932735426008,
+      "grad_norm": 0.0698273036311671,
+      "learning_rate": 8.202879148635348e-05,
+      "loss": 0.2662,
+      "step": 3221
+    },
+    {
+      "epoch": 1.8060538116591929,
+      "grad_norm": 0.06902402511065545,
+      "learning_rate": 8.196462494128778e-05,
+      "loss": 0.2527,
+      "step": 3222
+    },
+    {
+      "epoch": 1.8066143497757847,
+      "grad_norm": 0.06966987801069195,
+      "learning_rate": 8.190046607075496e-05,
+      "loss": 0.2513,
+      "step": 3223
+    },
+    {
+      "epoch": 1.8071748878923768,
+      "grad_norm": 0.0708178147378975,
+      "learning_rate": 8.183631490205637e-05,
+      "loss": 0.259,
+      "step": 3224
+    },
+    {
+      "epoch": 1.8077354260089686,
+      "grad_norm": 0.07219415365250346,
+      "learning_rate": 8.177217146249001e-05,
+      "loss": 0.272,
+      "step": 3225
+    },
+    {
+      "epoch": 1.8082959641255605,
+      "grad_norm": 0.07092431502983654,
+      "learning_rate": 8.17080357793506e-05,
+      "loss": 0.2686,
+      "step": 3226
+    },
+    {
+      "epoch": 1.8088565022421523,
+      "grad_norm": 0.06940473245167077,
+      "learning_rate": 8.164390787992963e-05,
+      "loss": 0.2592,
+      "step": 3227
+    },
+    {
+      "epoch": 1.8094170403587444,
+      "grad_norm": 0.07063170321896624,
+      "learning_rate": 8.157978779151518e-05,
+      "loss": 0.2706,
+      "step": 3228
+    },
+    {
+      "epoch": 1.8099775784753365,
+      "grad_norm": 0.0691899810144605,
+      "learning_rate": 8.151567554139213e-05,
+      "loss": 0.257,
+      "step": 3229
+    },
+    {
+      "epoch": 1.8105381165919283,
+      "grad_norm": 0.07043625516430235,
+      "learning_rate": 8.145157115684188e-05,
+      "loss": 0.2695,
+      "step": 3230
+    },
+    {
+      "epoch": 1.8110986547085202,
+      "grad_norm": 0.07121692930243351,
+      "learning_rate": 8.138747466514258e-05,
+      "loss": 0.2638,
+      "step": 3231
+    },
+    {
+      "epoch": 1.811659192825112,
+      "grad_norm": 0.07112343850379138,
+      "learning_rate": 8.132338609356904e-05,
+      "loss": 0.2748,
+      "step": 3232
+    },
+    {
+      "epoch": 1.812219730941704,
+      "grad_norm": 0.0708272835416039,
+      "learning_rate": 8.125930546939258e-05,
+      "loss": 0.278,
+      "step": 3233
+    },
+    {
+      "epoch": 1.812780269058296,
+      "grad_norm": 0.07002810189428951,
+      "learning_rate": 8.119523281988128e-05,
+      "loss": 0.261,
+      "step": 3234
+    },
+    {
+      "epoch": 1.813340807174888,
+      "grad_norm": 0.06857860753845983,
+      "learning_rate": 8.113116817229969e-05,
+      "loss": 0.2699,
+      "step": 3235
+    },
+    {
+      "epoch": 1.8139013452914798,
+      "grad_norm": 0.06873621338240818,
+      "learning_rate": 8.106711155390908e-05,
+      "loss": 0.2602,
+      "step": 3236
+    },
+    {
+      "epoch": 1.8144618834080717,
+      "grad_norm": 0.06951754756328997,
+      "learning_rate": 8.100306299196722e-05,
+      "loss": 0.275,
+      "step": 3237
+    },
+    {
+      "epoch": 1.8150224215246635,
+      "grad_norm": 0.07079477677031519,
+      "learning_rate": 8.093902251372853e-05,
+      "loss": 0.262,
+      "step": 3238
+    },
+    {
+      "epoch": 1.8155829596412556,
+      "grad_norm": 0.06994273981841134,
+      "learning_rate": 8.087499014644388e-05,
+      "loss": 0.267,
+      "step": 3239
+    },
+    {
+      "epoch": 1.8161434977578477,
+      "grad_norm": 0.07012393285154203,
+      "learning_rate": 8.081096591736076e-05,
+      "loss": 0.268,
+      "step": 3240
+    },
+    {
+      "epoch": 1.8167040358744395,
+      "grad_norm": 0.07179251474163152,
+      "learning_rate": 8.074694985372327e-05,
+      "loss": 0.2495,
+      "step": 3241
+    },
+    {
+      "epoch": 1.8172645739910314,
+      "grad_norm": 0.06895965921480107,
+      "learning_rate": 8.068294198277181e-05,
+      "loss": 0.2609,
+      "step": 3242
+    },
+    {
+      "epoch": 1.8178251121076232,
+      "grad_norm": 0.07374500362428504,
+      "learning_rate": 8.061894233174354e-05,
+      "loss": 0.2707,
+      "step": 3243
+    },
+    {
+      "epoch": 1.8183856502242153,
+      "grad_norm": 0.07016354106209098,
+      "learning_rate": 8.055495092787196e-05,
+      "loss": 0.267,
+      "step": 3244
+    },
+    {
+      "epoch": 1.8189461883408071,
+      "grad_norm": 0.0702743851240875,
+      "learning_rate": 8.049096779838719e-05,
+      "loss": 0.2507,
+      "step": 3245
+    },
+    {
+      "epoch": 1.8195067264573992,
+      "grad_norm": 0.06953644702458327,
+      "learning_rate": 8.04269929705157e-05,
+      "loss": 0.27,
+      "step": 3246
+    },
+    {
+      "epoch": 1.820067264573991,
+      "grad_norm": 0.07111965090248204,
+      "learning_rate": 8.036302647148046e-05,
+      "loss": 0.2688,
+      "step": 3247
+    },
+    {
+      "epoch": 1.8206278026905829,
+      "grad_norm": 0.07225789021090594,
+      "learning_rate": 8.029906832850098e-05,
+      "loss": 0.2749,
+      "step": 3248
+    },
+    {
+      "epoch": 1.8211883408071747,
+      "grad_norm": 0.06865477466078054,
+      "learning_rate": 8.023511856879312e-05,
+      "loss": 0.2762,
+      "step": 3249
+    },
+    {
+      "epoch": 1.8217488789237668,
+      "grad_norm": 0.06957466780206226,
+      "learning_rate": 8.017117721956925e-05,
+      "loss": 0.2647,
+      "step": 3250
+    },
+    {
+      "epoch": 1.8223094170403589,
+      "grad_norm": 0.07143351345817087,
+      "learning_rate": 8.010724430803806e-05,
+      "loss": 0.2736,
+      "step": 3251
+    },
+    {
+      "epoch": 1.8228699551569507,
+      "grad_norm": 0.07217209061752763,
+      "learning_rate": 8.004331986140474e-05,
+      "loss": 0.2569,
+      "step": 3252
+    },
+    {
+      "epoch": 1.8234304932735426,
+      "grad_norm": 0.06995558984458415,
+      "learning_rate": 7.997940390687087e-05,
+      "loss": 0.2701,
+      "step": 3253
+    },
+    {
+      "epoch": 1.8239910313901344,
+      "grad_norm": 0.06795951316620413,
+      "learning_rate": 7.991549647163432e-05,
+      "loss": 0.2661,
+      "step": 3254
+    },
+    {
+      "epoch": 1.8245515695067265,
+      "grad_norm": 0.07162393917965158,
+      "learning_rate": 7.985159758288947e-05,
+      "loss": 0.2733,
+      "step": 3255
+    },
+    {
+      "epoch": 1.8251121076233185,
+      "grad_norm": 0.06940977813543703,
+      "learning_rate": 7.978770726782697e-05,
+      "loss": 0.2561,
+      "step": 3256
+    },
+    {
+      "epoch": 1.8256726457399104,
+      "grad_norm": 0.07077292043103751,
+      "learning_rate": 7.972382555363391e-05,
+      "loss": 0.2691,
+      "step": 3257
+    },
+    {
+      "epoch": 1.8262331838565022,
+      "grad_norm": 0.06910878263615959,
+      "learning_rate": 7.965995246749357e-05,
+      "loss": 0.2598,
+      "step": 3258
+    },
+    {
+      "epoch": 1.826793721973094,
+      "grad_norm": 0.06971989571511918,
+      "learning_rate": 7.959608803658575e-05,
+      "loss": 0.2673,
+      "step": 3259
+    },
+    {
+      "epoch": 1.827354260089686,
+      "grad_norm": 0.07088971305325505,
+      "learning_rate": 7.953223228808639e-05,
+      "loss": 0.2689,
+      "step": 3260
+    },
+    {
+      "epoch": 1.827914798206278,
+      "grad_norm": 0.06957420254574456,
+      "learning_rate": 7.94683852491678e-05,
+      "loss": 0.2587,
+      "step": 3261
+    },
+    {
+      "epoch": 1.82847533632287,
+      "grad_norm": 0.06939132116929006,
+      "learning_rate": 7.94045469469987e-05,
+      "loss": 0.26,
+      "step": 3262
+    },
+    {
+      "epoch": 1.829035874439462,
+      "grad_norm": 0.06917830409614653,
+      "learning_rate": 7.93407174087439e-05,
+      "loss": 0.2587,
+      "step": 3263
+    },
+    {
+      "epoch": 1.8295964125560538,
+      "grad_norm": 0.07049559381730097,
+      "learning_rate": 7.927689666156458e-05,
+      "loss": 0.2564,
+      "step": 3264
+    },
+    {
+      "epoch": 1.8301569506726456,
+      "grad_norm": 0.07032746320631154,
+      "learning_rate": 7.92130847326182e-05,
+      "loss": 0.2599,
+      "step": 3265
+    },
+    {
+      "epoch": 1.8307174887892377,
+      "grad_norm": 0.06991754386154504,
+      "learning_rate": 7.914928164905844e-05,
+      "loss": 0.2727,
+      "step": 3266
+    },
+    {
+      "epoch": 1.8312780269058297,
+      "grad_norm": 0.0690391811553293,
+      "learning_rate": 7.90854874380352e-05,
+      "loss": 0.2484,
+      "step": 3267
+    },
+    {
+      "epoch": 1.8318385650224216,
+      "grad_norm": 0.06895022301849879,
+      "learning_rate": 7.902170212669457e-05,
+      "loss": 0.2558,
+      "step": 3268
+    },
+    {
+      "epoch": 1.8323991031390134,
+      "grad_norm": 0.07026848639407661,
+      "learning_rate": 7.895792574217897e-05,
+      "loss": 0.263,
+      "step": 3269
+    },
+    {
+      "epoch": 1.8329596412556053,
+      "grad_norm": 0.06950520145984855,
+      "learning_rate": 7.88941583116269e-05,
+      "loss": 0.2696,
+      "step": 3270
+    },
+    {
+      "epoch": 1.8335201793721974,
+      "grad_norm": 0.07198396734595838,
+      "learning_rate": 7.883039986217319e-05,
+      "loss": 0.2649,
+      "step": 3271
+    },
+    {
+      "epoch": 1.8340807174887892,
+      "grad_norm": 0.07260195426208052,
+      "learning_rate": 7.876665042094867e-05,
+      "loss": 0.2709,
+      "step": 3272
+    },
+    {
+      "epoch": 1.8346412556053813,
+      "grad_norm": 0.06863939703963023,
+      "learning_rate": 7.870291001508041e-05,
+      "loss": 0.2625,
+      "step": 3273
+    },
+    {
+      "epoch": 1.8352017937219731,
+      "grad_norm": 0.06777536080393645,
+      "learning_rate": 7.863917867169174e-05,
+      "loss": 0.2679,
+      "step": 3274
+    },
+    {
+      "epoch": 1.835762331838565,
+      "grad_norm": 0.06994768384030978,
+      "learning_rate": 7.857545641790198e-05,
+      "loss": 0.2687,
+      "step": 3275
+    },
+    {
+      "epoch": 1.8363228699551568,
+      "grad_norm": 0.06773691764880604,
+      "learning_rate": 7.851174328082669e-05,
+      "loss": 0.2617,
+      "step": 3276
+    },
+    {
+      "epoch": 1.8368834080717489,
+      "grad_norm": 0.06936073208434729,
+      "learning_rate": 7.844803928757746e-05,
+      "loss": 0.2611,
+      "step": 3277
+    },
+    {
+      "epoch": 1.837443946188341,
+      "grad_norm": 0.06944387446971986,
+      "learning_rate": 7.83843444652621e-05,
+      "loss": 0.2637,
+      "step": 3278
+    },
+    {
+      "epoch": 1.8380044843049328,
+      "grad_norm": 0.06992787302602949,
+      "learning_rate": 7.832065884098442e-05,
+      "loss": 0.2655,
+      "step": 3279
+    },
+    {
+      "epoch": 1.8385650224215246,
+      "grad_norm": 0.06937655616251745,
+      "learning_rate": 7.825698244184431e-05,
+      "loss": 0.2714,
+      "step": 3280
+    },
+    {
+      "epoch": 1.8391255605381165,
+      "grad_norm": 0.07068469553748749,
+      "learning_rate": 7.819331529493785e-05,
+      "loss": 0.2761,
+      "step": 3281
+    },
+    {
+      "epoch": 1.8396860986547086,
+      "grad_norm": 0.06934047764865918,
+      "learning_rate": 7.812965742735704e-05,
+      "loss": 0.2568,
+      "step": 3282
+    },
+    {
+      "epoch": 1.8402466367713004,
+      "grad_norm": 0.06871391216147392,
+      "learning_rate": 7.806600886619008e-05,
+      "loss": 0.2664,
+      "step": 3283
+    },
+    {
+      "epoch": 1.8408071748878925,
+      "grad_norm": 0.07283045631824599,
+      "learning_rate": 7.800236963852106e-05,
+      "loss": 0.2782,
+      "step": 3284
+    },
+    {
+      "epoch": 1.8413677130044843,
+      "grad_norm": 0.06928698187482213,
+      "learning_rate": 7.793873977143019e-05,
+      "loss": 0.2601,
+      "step": 3285
+    },
+    {
+      "epoch": 1.8419282511210762,
+      "grad_norm": 0.07072451148724478,
+      "learning_rate": 7.78751192919937e-05,
+      "loss": 0.2676,
+      "step": 3286
+    },
+    {
+      "epoch": 1.842488789237668,
+      "grad_norm": 0.07076600383735644,
+      "learning_rate": 7.781150822728373e-05,
+      "loss": 0.2799,
+      "step": 3287
+    },
+    {
+      "epoch": 1.84304932735426,
+      "grad_norm": 0.07131993457810994,
+      "learning_rate": 7.774790660436858e-05,
+      "loss": 0.2623,
+      "step": 3288
+    },
+    {
+      "epoch": 1.8436098654708521,
+      "grad_norm": 0.06881675330410725,
+      "learning_rate": 7.768431445031233e-05,
+      "loss": 0.2702,
+      "step": 3289
+    },
+    {
+      "epoch": 1.844170403587444,
+      "grad_norm": 0.07152766006078261,
+      "learning_rate": 7.762073179217526e-05,
+      "loss": 0.2803,
+      "step": 3290
+    },
+    {
+      "epoch": 1.8447309417040358,
+      "grad_norm": 0.0700623448658665,
+      "learning_rate": 7.755715865701337e-05,
+      "loss": 0.264,
+      "step": 3291
+    },
+    {
+      "epoch": 1.8452914798206277,
+      "grad_norm": 0.07361118318553973,
+      "learning_rate": 7.749359507187882e-05,
+      "loss": 0.2755,
+      "step": 3292
+    },
+    {
+      "epoch": 1.8458520179372198,
+      "grad_norm": 0.07158716462650891,
+      "learning_rate": 7.743004106381952e-05,
+      "loss": 0.2709,
+      "step": 3293
+    },
+    {
+      "epoch": 1.8464125560538116,
+      "grad_norm": 0.069650654147604,
+      "learning_rate": 7.736649665987943e-05,
+      "loss": 0.2622,
+      "step": 3294
+    },
+    {
+      "epoch": 1.8469730941704037,
+      "grad_norm": 0.06838339865969636,
+      "learning_rate": 7.730296188709844e-05,
+      "loss": 0.264,
+      "step": 3295
+    },
+    {
+      "epoch": 1.8475336322869955,
+      "grad_norm": 0.06915450823913075,
+      "learning_rate": 7.723943677251222e-05,
+      "loss": 0.2539,
+      "step": 3296
+    },
+    {
+      "epoch": 1.8480941704035874,
+      "grad_norm": 0.07051579087788801,
+      "learning_rate": 7.717592134315243e-05,
+      "loss": 0.2598,
+      "step": 3297
+    },
+    {
+      "epoch": 1.8486547085201792,
+      "grad_norm": 0.07275806676175088,
+      "learning_rate": 7.711241562604655e-05,
+      "loss": 0.2714,
+      "step": 3298
+    },
+    {
+      "epoch": 1.8492152466367713,
+      "grad_norm": 0.07072076011263187,
+      "learning_rate": 7.704891964821802e-05,
+      "loss": 0.2626,
+      "step": 3299
+    },
+    {
+      "epoch": 1.8497757847533634,
+      "grad_norm": 0.07077532083248372,
+      "learning_rate": 7.698543343668602e-05,
+      "loss": 0.2604,
+      "step": 3300
+    },
+    {
+      "epoch": 1.8503363228699552,
+      "grad_norm": 0.06818733312627143,
+      "learning_rate": 7.69219570184656e-05,
+      "loss": 0.245,
+      "step": 3301
+    },
+    {
+      "epoch": 1.850896860986547,
+      "grad_norm": 0.07023083270256576,
+      "learning_rate": 7.685849042056776e-05,
+      "loss": 0.2601,
+      "step": 3302
+    },
+    {
+      "epoch": 1.851457399103139,
+      "grad_norm": 0.07171723482332348,
+      "learning_rate": 7.679503366999913e-05,
+      "loss": 0.2659,
+      "step": 3303
+    },
+    {
+      "epoch": 1.852017937219731,
+      "grad_norm": 0.07000108111565913,
+      "learning_rate": 7.673158679376234e-05,
+      "loss": 0.2566,
+      "step": 3304
+    },
+    {
+      "epoch": 1.852578475336323,
+      "grad_norm": 0.0703263400695774,
+      "learning_rate": 7.666814981885562e-05,
+      "loss": 0.2495,
+      "step": 3305
+    },
+    {
+      "epoch": 1.8531390134529149,
+      "grad_norm": 0.0722579247585532,
+      "learning_rate": 7.66047227722732e-05,
+      "loss": 0.2694,
+      "step": 3306
+    },
+    {
+      "epoch": 1.8536995515695067,
+      "grad_norm": 0.07136350358402029,
+      "learning_rate": 7.654130568100492e-05,
+      "loss": 0.2578,
+      "step": 3307
+    },
+    {
+      "epoch": 1.8542600896860986,
+      "grad_norm": 0.06908740131278682,
+      "learning_rate": 7.647789857203645e-05,
+      "loss": 0.2481,
+      "step": 3308
+    },
+    {
+      "epoch": 1.8548206278026906,
+      "grad_norm": 0.07265326145159552,
+      "learning_rate": 7.64145014723492e-05,
+      "loss": 0.2628,
+      "step": 3309
+    },
+    {
+      "epoch": 1.8553811659192825,
+      "grad_norm": 0.07148769815898748,
+      "learning_rate": 7.635111440892034e-05,
+      "loss": 0.2749,
+      "step": 3310
+    },
+    {
+      "epoch": 1.8559417040358746,
+      "grad_norm": 0.07087077594439263,
+      "learning_rate": 7.62877374087228e-05,
+      "loss": 0.2622,
+      "step": 3311
+    },
+    {
+      "epoch": 1.8565022421524664,
+      "grad_norm": 0.07034259844404242,
+      "learning_rate": 7.622437049872512e-05,
+      "loss": 0.2683,
+      "step": 3312
+    },
+    {
+      "epoch": 1.8570627802690582,
+      "grad_norm": 0.07194898546729013,
+      "learning_rate": 7.616101370589158e-05,
+      "loss": 0.2643,
+      "step": 3313
+    },
+    {
+      "epoch": 1.85762331838565,
+      "grad_norm": 0.07097837155946217,
+      "learning_rate": 7.609766705718225e-05,
+      "loss": 0.26,
+      "step": 3314
+    },
+    {
+      "epoch": 1.8581838565022422,
+      "grad_norm": 0.07060544817745804,
+      "learning_rate": 7.60343305795528e-05,
+      "loss": 0.2648,
+      "step": 3315
+    },
+    {
+      "epoch": 1.8587443946188342,
+      "grad_norm": 0.06912082320374903,
+      "learning_rate": 7.59710042999546e-05,
+      "loss": 0.2684,
+      "step": 3316
+    },
+    {
+      "epoch": 1.859304932735426,
+      "grad_norm": 0.07062735182228592,
+      "learning_rate": 7.590768824533463e-05,
+      "loss": 0.2708,
+      "step": 3317
+    },
+    {
+      "epoch": 1.859865470852018,
+      "grad_norm": 0.07106775905431914,
+      "learning_rate": 7.584438244263561e-05,
+      "loss": 0.2665,
+      "step": 3318
+    },
+    {
+      "epoch": 1.8604260089686098,
+      "grad_norm": 0.07179109023822038,
+      "learning_rate": 7.578108691879584e-05,
+      "loss": 0.2766,
+      "step": 3319
+    },
+    {
+      "epoch": 1.8609865470852018,
+      "grad_norm": 0.07177401584888915,
+      "learning_rate": 7.57178017007492e-05,
+      "loss": 0.2614,
+      "step": 3320
+    },
+    {
+      "epoch": 1.8615470852017937,
+      "grad_norm": 0.07334474144983744,
+      "learning_rate": 7.565452681542529e-05,
+      "loss": 0.2722,
+      "step": 3321
+    },
+    {
+      "epoch": 1.8621076233183858,
+      "grad_norm": 0.07217197580352912,
+      "learning_rate": 7.559126228974921e-05,
+      "loss": 0.2593,
+      "step": 3322
+    },
+    {
+      "epoch": 1.8626681614349776,
+      "grad_norm": 0.06946442838759735,
+      "learning_rate": 7.55280081506418e-05,
+      "loss": 0.2667,
+      "step": 3323
+    },
+    {
+      "epoch": 1.8632286995515694,
+      "grad_norm": 0.07131625752713222,
+      "learning_rate": 7.546476442501926e-05,
+      "loss": 0.2655,
+      "step": 3324
+    },
+    {
+      "epoch": 1.8637892376681613,
+      "grad_norm": 0.06996968895541275,
+      "learning_rate": 7.54015311397936e-05,
+      "loss": 0.2723,
+      "step": 3325
+    },
+    {
+      "epoch": 1.8643497757847534,
+      "grad_norm": 0.07084335490981197,
+      "learning_rate": 7.533830832187216e-05,
+      "loss": 0.2677,
+      "step": 3326
+    },
+    {
+      "epoch": 1.8649103139013454,
+      "grad_norm": 0.07049422582762938,
+      "learning_rate": 7.527509599815799e-05,
+      "loss": 0.2654,
+      "step": 3327
+    },
+    {
+      "epoch": 1.8654708520179373,
+      "grad_norm": 0.07020752735237043,
+      "learning_rate": 7.521189419554963e-05,
+      "loss": 0.2606,
+      "step": 3328
+    },
+    {
+      "epoch": 1.8660313901345291,
+      "grad_norm": 0.06884771378378535,
+      "learning_rate": 7.51487029409411e-05,
+      "loss": 0.2564,
+      "step": 3329
+    },
+    {
+      "epoch": 1.866591928251121,
+      "grad_norm": 0.06850733209924952,
+      "learning_rate": 7.508552226122197e-05,
+      "loss": 0.267,
+      "step": 3330
+    },
+    {
+      "epoch": 1.867152466367713,
+      "grad_norm": 0.06958197189734695,
+      "learning_rate": 7.502235218327731e-05,
+      "loss": 0.2654,
+      "step": 3331
+    },
+    {
+      "epoch": 1.8677130044843049,
+      "grad_norm": 0.06861922633796758,
+      "learning_rate": 7.49591927339877e-05,
+      "loss": 0.2704,
+      "step": 3332
+    },
+    {
+      "epoch": 1.868273542600897,
+      "grad_norm": 0.06588015335176929,
+      "learning_rate": 7.489604394022914e-05,
+      "loss": 0.2449,
+      "step": 3333
+    },
+    {
+      "epoch": 1.8688340807174888,
+      "grad_norm": 0.07808355023131226,
+      "learning_rate": 7.483290582887308e-05,
+      "loss": 0.2591,
+      "step": 3334
+    },
+    {
+      "epoch": 1.8693946188340806,
+      "grad_norm": 0.06726693978009231,
+      "learning_rate": 7.476977842678659e-05,
+      "loss": 0.2628,
+      "step": 3335
+    },
+    {
+      "epoch": 1.8699551569506725,
+      "grad_norm": 0.07212035137203555,
+      "learning_rate": 7.470666176083192e-05,
+      "loss": 0.2805,
+      "step": 3336
+    },
+    {
+      "epoch": 1.8705156950672646,
+      "grad_norm": 0.06904850809163075,
+      "learning_rate": 7.464355585786702e-05,
+      "loss": 0.2665,
+      "step": 3337
+    },
+    {
+      "epoch": 1.8710762331838566,
+      "grad_norm": 0.07195609236414281,
+      "learning_rate": 7.458046074474504e-05,
+      "loss": 0.2543,
+      "step": 3338
+    },
+    {
+      "epoch": 1.8716367713004485,
+      "grad_norm": 0.06899222034013164,
+      "learning_rate": 7.451737644831469e-05,
+      "loss": 0.2781,
+      "step": 3339
+    },
+    {
+      "epoch": 1.8721973094170403,
+      "grad_norm": 0.06897223641407911,
+      "learning_rate": 7.445430299542002e-05,
+      "loss": 0.2535,
+      "step": 3340
+    },
+    {
+      "epoch": 1.8727578475336322,
+      "grad_norm": 0.07179626227747822,
+      "learning_rate": 7.43912404129004e-05,
+      "loss": 0.268,
+      "step": 3341
+    },
+    {
+      "epoch": 1.8733183856502242,
+      "grad_norm": 0.0668777419822922,
+      "learning_rate": 7.432818872759071e-05,
+      "loss": 0.2578,
+      "step": 3342
+    },
+    {
+      "epoch": 1.8738789237668163,
+      "grad_norm": 0.06907163869149917,
+      "learning_rate": 7.426514796632108e-05,
+      "loss": 0.2726,
+      "step": 3343
+    },
+    {
+      "epoch": 1.8744394618834082,
+      "grad_norm": 0.0706533991015148,
+      "learning_rate": 7.420211815591709e-05,
+      "loss": 0.2559,
+      "step": 3344
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 0.07173026287207163,
+      "learning_rate": 7.413909932319952e-05,
+      "loss": 0.2709,
+      "step": 3345
+    },
+    {
+      "epoch": 1.8755605381165918,
+      "grad_norm": 0.07111268467729995,
+      "learning_rate": 7.407609149498467e-05,
+      "loss": 0.2538,
+      "step": 3346
+    },
+    {
+      "epoch": 1.8761210762331837,
+      "grad_norm": 0.0684648010044524,
+      "learning_rate": 7.401309469808395e-05,
+      "loss": 0.2636,
+      "step": 3347
+    },
+    {
+      "epoch": 1.8766816143497758,
+      "grad_norm": 0.06904085368570378,
+      "learning_rate": 7.395010895930421e-05,
+      "loss": 0.2732,
+      "step": 3348
+    },
+    {
+      "epoch": 1.8772421524663678,
+      "grad_norm": 0.06744956753540765,
+      "learning_rate": 7.388713430544763e-05,
+      "loss": 0.2682,
+      "step": 3349
+    },
+    {
+      "epoch": 1.8778026905829597,
+      "grad_norm": 0.0685403697581941,
+      "learning_rate": 7.382417076331147e-05,
+      "loss": 0.258,
+      "step": 3350
+    },
+    {
+      "epoch": 1.8783632286995515,
+      "grad_norm": 0.07099772248656062,
+      "learning_rate": 7.376121835968851e-05,
+      "loss": 0.2672,
+      "step": 3351
+    },
+    {
+      "epoch": 1.8789237668161434,
+      "grad_norm": 0.07080807687251338,
+      "learning_rate": 7.369827712136661e-05,
+      "loss": 0.2637,
+      "step": 3352
+    },
+    {
+      "epoch": 1.8794843049327354,
+      "grad_norm": 0.06846463607950867,
+      "learning_rate": 7.363534707512901e-05,
+      "loss": 0.2525,
+      "step": 3353
+    },
+    {
+      "epoch": 1.8800448430493275,
+      "grad_norm": 0.06777615635017255,
+      "learning_rate": 7.357242824775406e-05,
+      "loss": 0.2568,
+      "step": 3354
+    },
+    {
+      "epoch": 1.8806053811659194,
+      "grad_norm": 0.07173455763218528,
+      "learning_rate": 7.35095206660154e-05,
+      "loss": 0.2659,
+      "step": 3355
+    },
+    {
+      "epoch": 1.8811659192825112,
+      "grad_norm": 0.06946326908201468,
+      "learning_rate": 7.344662435668196e-05,
+      "loss": 0.2714,
+      "step": 3356
+    },
+    {
+      "epoch": 1.881726457399103,
+      "grad_norm": 0.07026772989639453,
+      "learning_rate": 7.338373934651768e-05,
+      "loss": 0.266,
+      "step": 3357
+    },
+    {
+      "epoch": 1.8822869955156951,
+      "grad_norm": 0.07253312539345642,
+      "learning_rate": 7.332086566228194e-05,
+      "loss": 0.2583,
+      "step": 3358
+    },
+    {
+      "epoch": 1.882847533632287,
+      "grad_norm": 0.072342685380318,
+      "learning_rate": 7.325800333072904e-05,
+      "loss": 0.2689,
+      "step": 3359
+    },
+    {
+      "epoch": 1.883408071748879,
+      "grad_norm": 0.07121501443116349,
+      "learning_rate": 7.319515237860864e-05,
+      "loss": 0.2664,
+      "step": 3360
+    },
+    {
+      "epoch": 1.8839686098654709,
+      "grad_norm": 0.0706672213069749,
+      "learning_rate": 7.313231283266551e-05,
+      "loss": 0.262,
+      "step": 3361
+    },
+    {
+      "epoch": 1.8845291479820627,
+      "grad_norm": 0.0708292342381674,
+      "learning_rate": 7.306948471963951e-05,
+      "loss": 0.2592,
+      "step": 3362
+    },
+    {
+      "epoch": 1.8850896860986546,
+      "grad_norm": 0.07074562930059831,
+      "learning_rate": 7.300666806626572e-05,
+      "loss": 0.2658,
+      "step": 3363
+    },
+    {
+      "epoch": 1.8856502242152466,
+      "grad_norm": 0.0690218491119218,
+      "learning_rate": 7.294386289927425e-05,
+      "loss": 0.2495,
+      "step": 3364
+    },
+    {
+      "epoch": 1.8862107623318387,
+      "grad_norm": 0.07256808673582511,
+      "learning_rate": 7.288106924539045e-05,
+      "loss": 0.2754,
+      "step": 3365
+    },
+    {
+      "epoch": 1.8867713004484306,
+      "grad_norm": 0.06936835280441091,
+      "learning_rate": 7.281828713133463e-05,
+      "loss": 0.2624,
+      "step": 3366
+    },
+    {
+      "epoch": 1.8873318385650224,
+      "grad_norm": 0.07098030153719274,
+      "learning_rate": 7.275551658382224e-05,
+      "loss": 0.2786,
+      "step": 3367
+    },
+    {
+      "epoch": 1.8878923766816142,
+      "grad_norm": 0.06914782827830317,
+      "learning_rate": 7.269275762956392e-05,
+      "loss": 0.2598,
+      "step": 3368
+    },
+    {
+      "epoch": 1.8884529147982063,
+      "grad_norm": 0.07038723810265705,
+      "learning_rate": 7.263001029526514e-05,
+      "loss": 0.269,
+      "step": 3369
+    },
+    {
+      "epoch": 1.8890134529147982,
+      "grad_norm": 0.07129477470082596,
+      "learning_rate": 7.256727460762669e-05,
+      "loss": 0.2781,
+      "step": 3370
+    },
+    {
+      "epoch": 1.8895739910313902,
+      "grad_norm": 0.0679187090080146,
+      "learning_rate": 7.250455059334417e-05,
+      "loss": 0.254,
+      "step": 3371
+    },
+    {
+      "epoch": 1.890134529147982,
+      "grad_norm": 0.0701264604194011,
+      "learning_rate": 7.24418382791084e-05,
+      "loss": 0.268,
+      "step": 3372
+    },
+    {
+      "epoch": 1.890695067264574,
+      "grad_norm": 0.06989599911204056,
+      "learning_rate": 7.237913769160514e-05,
+      "loss": 0.2768,
+      "step": 3373
+    },
+    {
+      "epoch": 1.8912556053811658,
+      "grad_norm": 0.07112791500392587,
+      "learning_rate": 7.231644885751507e-05,
+      "loss": 0.2685,
+      "step": 3374
+    },
+    {
+      "epoch": 1.8918161434977578,
+      "grad_norm": 0.07253441231150679,
+      "learning_rate": 7.225377180351406e-05,
+      "loss": 0.2726,
+      "step": 3375
+    },
+    {
+      "epoch": 1.89237668161435,
+      "grad_norm": 0.07037726472225406,
+      "learning_rate": 7.219110655627281e-05,
+      "loss": 0.258,
+      "step": 3376
+    },
+    {
+      "epoch": 1.8929372197309418,
+      "grad_norm": 0.07196987542914432,
+      "learning_rate": 7.212845314245712e-05,
+      "loss": 0.2604,
+      "step": 3377
+    },
+    {
+      "epoch": 1.8934977578475336,
+      "grad_norm": 0.07431132206256182,
+      "learning_rate": 7.20658115887276e-05,
+      "loss": 0.2638,
+      "step": 3378
+    },
+    {
+      "epoch": 1.8940582959641254,
+      "grad_norm": 0.06932179226439326,
+      "learning_rate": 7.200318192173998e-05,
+      "loss": 0.2572,
+      "step": 3379
+    },
+    {
+      "epoch": 1.8946188340807175,
+      "grad_norm": 0.06984265728570498,
+      "learning_rate": 7.194056416814481e-05,
+      "loss": 0.2607,
+      "step": 3380
+    },
+    {
+      "epoch": 1.8951793721973094,
+      "grad_norm": 0.06794315477215011,
+      "learning_rate": 7.187795835458759e-05,
+      "loss": 0.2553,
+      "step": 3381
+    },
+    {
+      "epoch": 1.8957399103139014,
+      "grad_norm": 0.07257280209516982,
+      "learning_rate": 7.181536450770882e-05,
+      "loss": 0.2658,
+      "step": 3382
+    },
+    {
+      "epoch": 1.8963004484304933,
+      "grad_norm": 0.06998263085993002,
+      "learning_rate": 7.17527826541438e-05,
+      "loss": 0.2737,
+      "step": 3383
+    },
+    {
+      "epoch": 1.8968609865470851,
+      "grad_norm": 0.06850490431660242,
+      "learning_rate": 7.169021282052283e-05,
+      "loss": 0.2538,
+      "step": 3384
+    },
+    {
+      "epoch": 1.897421524663677,
+      "grad_norm": 0.07011061332069132,
+      "learning_rate": 7.162765503347097e-05,
+      "loss": 0.2574,
+      "step": 3385
+    },
+    {
+      "epoch": 1.897982062780269,
+      "grad_norm": 0.06744822387404592,
+      "learning_rate": 7.156510931960833e-05,
+      "loss": 0.2502,
+      "step": 3386
+    },
+    {
+      "epoch": 1.898542600896861,
+      "grad_norm": 0.0682590772522164,
+      "learning_rate": 7.15025757055497e-05,
+      "loss": 0.267,
+      "step": 3387
+    },
+    {
+      "epoch": 1.899103139013453,
+      "grad_norm": 0.0718727765124095,
+      "learning_rate": 7.144005421790479e-05,
+      "loss": 0.2671,
+      "step": 3388
+    },
+    {
+      "epoch": 1.8996636771300448,
+      "grad_norm": 0.06914143322440189,
+      "learning_rate": 7.137754488327822e-05,
+      "loss": 0.2546,
+      "step": 3389
+    },
+    {
+      "epoch": 1.9002242152466366,
+      "grad_norm": 0.07013261679104858,
+      "learning_rate": 7.131504772826931e-05,
+      "loss": 0.2576,
+      "step": 3390
+    },
+    {
+      "epoch": 1.9007847533632287,
+      "grad_norm": 0.07149179608649352,
+      "learning_rate": 7.125256277947234e-05,
+      "loss": 0.2658,
+      "step": 3391
+    },
+    {
+      "epoch": 1.9013452914798208,
+      "grad_norm": 0.07162361854306432,
+      "learning_rate": 7.119009006347625e-05,
+      "loss": 0.2733,
+      "step": 3392
+    },
+    {
+      "epoch": 1.9019058295964126,
+      "grad_norm": 0.0705232636082105,
+      "learning_rate": 7.112762960686489e-05,
+      "loss": 0.2611,
+      "step": 3393
+    },
+    {
+      "epoch": 1.9024663677130045,
+      "grad_norm": 0.07068874339562098,
+      "learning_rate": 7.106518143621687e-05,
+      "loss": 0.2559,
+      "step": 3394
+    },
+    {
+      "epoch": 1.9030269058295963,
+      "grad_norm": 0.07028228605770458,
+      "learning_rate": 7.100274557810546e-05,
+      "loss": 0.2596,
+      "step": 3395
+    },
+    {
+      "epoch": 1.9035874439461884,
+      "grad_norm": 0.07031247943732717,
+      "learning_rate": 7.094032205909888e-05,
+      "loss": 0.261,
+      "step": 3396
+    },
+    {
+      "epoch": 1.9041479820627802,
+      "grad_norm": 0.07276064957321741,
+      "learning_rate": 7.087791090575995e-05,
+      "loss": 0.2605,
+      "step": 3397
+    },
+    {
+      "epoch": 1.9047085201793723,
+      "grad_norm": 0.06930737962177644,
+      "learning_rate": 7.081551214464632e-05,
+      "loss": 0.2637,
+      "step": 3398
+    },
+    {
+      "epoch": 1.9052690582959642,
+      "grad_norm": 0.07065158486563466,
+      "learning_rate": 7.075312580231027e-05,
+      "loss": 0.2577,
+      "step": 3399
+    },
+    {
+      "epoch": 1.905829596412556,
+      "grad_norm": 0.07105371941610006,
+      "learning_rate": 7.069075190529888e-05,
+      "loss": 0.2595,
+      "step": 3400
+    },
+    {
+      "epoch": 1.9063901345291479,
+      "grad_norm": 0.07097615339552256,
+      "learning_rate": 7.062839048015392e-05,
+      "loss": 0.2739,
+      "step": 3401
+    },
+    {
+      "epoch": 1.90695067264574,
+      "grad_norm": 0.07166368974905003,
+      "learning_rate": 7.056604155341182e-05,
+      "loss": 0.2549,
+      "step": 3402
+    },
+    {
+      "epoch": 1.907511210762332,
+      "grad_norm": 0.07083603818333438,
+      "learning_rate": 7.050370515160373e-05,
+      "loss": 0.2734,
+      "step": 3403
+    },
+    {
+      "epoch": 1.9080717488789238,
+      "grad_norm": 0.06761603470174596,
+      "learning_rate": 7.044138130125542e-05,
+      "loss": 0.2634,
+      "step": 3404
+    },
+    {
+      "epoch": 1.9086322869955157,
+      "grad_norm": 0.06724505576057976,
+      "learning_rate": 7.037907002888738e-05,
+      "loss": 0.2612,
+      "step": 3405
+    },
+    {
+      "epoch": 1.9091928251121075,
+      "grad_norm": 0.06824851044340624,
+      "learning_rate": 7.031677136101472e-05,
+      "loss": 0.2649,
+      "step": 3406
+    },
+    {
+      "epoch": 1.9097533632286996,
+      "grad_norm": 0.06851903037823158,
+      "learning_rate": 7.025448532414712e-05,
+      "loss": 0.2628,
+      "step": 3407
+    },
+    {
+      "epoch": 1.9103139013452914,
+      "grad_norm": 0.06840303525882,
+      "learning_rate": 7.019221194478904e-05,
+      "loss": 0.2495,
+      "step": 3408
+    },
+    {
+      "epoch": 1.9108744394618835,
+      "grad_norm": 0.06964956062777737,
+      "learning_rate": 7.012995124943937e-05,
+      "loss": 0.2671,
+      "step": 3409
+    },
+    {
+      "epoch": 1.9114349775784754,
+      "grad_norm": 0.06620059709148725,
+      "learning_rate": 7.006770326459182e-05,
+      "loss": 0.2596,
+      "step": 3410
+    },
+    {
+      "epoch": 1.9119955156950672,
+      "grad_norm": 0.06864741139145739,
+      "learning_rate": 7.000546801673444e-05,
+      "loss": 0.2497,
+      "step": 3411
+    },
+    {
+      "epoch": 1.912556053811659,
+      "grad_norm": 0.06845870815715167,
+      "learning_rate": 6.994324553235006e-05,
+      "loss": 0.2628,
+      "step": 3412
+    },
+    {
+      "epoch": 1.9131165919282511,
+      "grad_norm": 0.06991522454333711,
+      "learning_rate": 6.9881035837916e-05,
+      "loss": 0.2671,
+      "step": 3413
+    },
+    {
+      "epoch": 1.9136771300448432,
+      "grad_norm": 0.06790449947644525,
+      "learning_rate": 6.981883895990409e-05,
+      "loss": 0.2568,
+      "step": 3414
+    },
+    {
+      "epoch": 1.914237668161435,
+      "grad_norm": 0.06895546513207257,
+      "learning_rate": 6.975665492478084e-05,
+      "loss": 0.2735,
+      "step": 3415
+    },
+    {
+      "epoch": 1.9147982062780269,
+      "grad_norm": 0.06997529341199701,
+      "learning_rate": 6.969448375900715e-05,
+      "loss": 0.2641,
+      "step": 3416
+    },
+    {
+      "epoch": 1.9153587443946187,
+      "grad_norm": 0.07031479138242287,
+      "learning_rate": 6.963232548903853e-05,
+      "loss": 0.2773,
+      "step": 3417
+    },
+    {
+      "epoch": 1.9159192825112108,
+      "grad_norm": 0.06988837242062663,
+      "learning_rate": 6.957018014132498e-05,
+      "loss": 0.2659,
+      "step": 3418
+    },
+    {
+      "epoch": 1.9164798206278026,
+      "grad_norm": 0.07048631773028599,
+      "learning_rate": 6.950804774231104e-05,
+      "loss": 0.2858,
+      "step": 3419
+    },
+    {
+      "epoch": 1.9170403587443947,
+      "grad_norm": 0.07065037724997916,
+      "learning_rate": 6.944592831843566e-05,
+      "loss": 0.2673,
+      "step": 3420
+    },
+    {
+      "epoch": 1.9176008968609866,
+      "grad_norm": 0.07091648817424527,
+      "learning_rate": 6.938382189613228e-05,
+      "loss": 0.2614,
+      "step": 3421
+    },
+    {
+      "epoch": 1.9181614349775784,
+      "grad_norm": 0.0701086193186055,
+      "learning_rate": 6.932172850182893e-05,
+      "loss": 0.2559,
+      "step": 3422
+    },
+    {
+      "epoch": 1.9187219730941703,
+      "grad_norm": 0.07150135750957044,
+      "learning_rate": 6.925964816194791e-05,
+      "loss": 0.2741,
+      "step": 3423
+    },
+    {
+      "epoch": 1.9192825112107623,
+      "grad_norm": 0.07037303016788464,
+      "learning_rate": 6.919758090290614e-05,
+      "loss": 0.2661,
+      "step": 3424
+    },
+    {
+      "epoch": 1.9198430493273544,
+      "grad_norm": 0.06989314969355041,
+      "learning_rate": 6.913552675111481e-05,
+      "loss": 0.2559,
+      "step": 3425
+    },
+    {
+      "epoch": 1.9204035874439462,
+      "grad_norm": 0.07333797503236086,
+      "learning_rate": 6.90734857329797e-05,
+      "loss": 0.2666,
+      "step": 3426
+    },
+    {
+      "epoch": 1.920964125560538,
+      "grad_norm": 0.07056701656580355,
+      "learning_rate": 6.901145787490087e-05,
+      "loss": 0.264,
+      "step": 3427
+    },
+    {
+      "epoch": 1.92152466367713,
+      "grad_norm": 0.07091650241937678,
+      "learning_rate": 6.894944320327281e-05,
+      "loss": 0.2711,
+      "step": 3428
+    },
+    {
+      "epoch": 1.922085201793722,
+      "grad_norm": 0.069847449242454,
+      "learning_rate": 6.888744174448446e-05,
+      "loss": 0.2573,
+      "step": 3429
+    },
+    {
+      "epoch": 1.922645739910314,
+      "grad_norm": 0.07089561215141386,
+      "learning_rate": 6.882545352491904e-05,
+      "loss": 0.2525,
+      "step": 3430
+    },
+    {
+      "epoch": 1.923206278026906,
+      "grad_norm": 0.07095533170948423,
+      "learning_rate": 6.87634785709543e-05,
+      "loss": 0.264,
+      "step": 3431
+    },
+    {
+      "epoch": 1.9237668161434978,
+      "grad_norm": 0.0710438893524046,
+      "learning_rate": 6.870151690896209e-05,
+      "loss": 0.2558,
+      "step": 3432
+    },
+    {
+      "epoch": 1.9243273542600896,
+      "grad_norm": 0.06895392739783383,
+      "learning_rate": 6.863956856530885e-05,
+      "loss": 0.2574,
+      "step": 3433
+    },
+    {
+      "epoch": 1.9248878923766815,
+      "grad_norm": 0.06666994015063393,
+      "learning_rate": 6.857763356635525e-05,
+      "loss": 0.2494,
+      "step": 3434
+    },
+    {
+      "epoch": 1.9254484304932735,
+      "grad_norm": 0.06995428661714155,
+      "learning_rate": 6.851571193845619e-05,
+      "loss": 0.2792,
+      "step": 3435
+    },
+    {
+      "epoch": 1.9260089686098656,
+      "grad_norm": 0.07121034669068539,
+      "learning_rate": 6.845380370796111e-05,
+      "loss": 0.2717,
+      "step": 3436
+    },
+    {
+      "epoch": 1.9265695067264574,
+      "grad_norm": 0.06806246914144769,
+      "learning_rate": 6.839190890121348e-05,
+      "loss": 0.2529,
+      "step": 3437
+    },
+    {
+      "epoch": 1.9271300448430493,
+      "grad_norm": 0.06986789929975527,
+      "learning_rate": 6.833002754455125e-05,
+      "loss": 0.2584,
+      "step": 3438
+    },
+    {
+      "epoch": 1.9276905829596411,
+      "grad_norm": 0.07238871603645033,
+      "learning_rate": 6.826815966430664e-05,
+      "loss": 0.2556,
+      "step": 3439
+    },
+    {
+      "epoch": 1.9282511210762332,
+      "grad_norm": 0.07283599960176848,
+      "learning_rate": 6.820630528680597e-05,
+      "loss": 0.2631,
+      "step": 3440
+    },
+    {
+      "epoch": 1.9288116591928253,
+      "grad_norm": 0.07207116428072398,
+      "learning_rate": 6.814446443837001e-05,
+      "loss": 0.2581,
+      "step": 3441
+    },
+    {
+      "epoch": 1.9293721973094171,
+      "grad_norm": 0.0719993417784609,
+      "learning_rate": 6.808263714531364e-05,
+      "loss": 0.2664,
+      "step": 3442
+    },
+    {
+      "epoch": 1.929932735426009,
+      "grad_norm": 0.06970555438156936,
+      "learning_rate": 6.802082343394611e-05,
+      "loss": 0.2598,
+      "step": 3443
+    },
+    {
+      "epoch": 1.9304932735426008,
+      "grad_norm": 0.07226176670628195,
+      "learning_rate": 6.795902333057067e-05,
+      "loss": 0.2635,
+      "step": 3444
+    },
+    {
+      "epoch": 1.9310538116591929,
+      "grad_norm": 0.06961802825931704,
+      "learning_rate": 6.789723686148502e-05,
+      "loss": 0.2614,
+      "step": 3445
+    },
+    {
+      "epoch": 1.9316143497757847,
+      "grad_norm": 0.06982975699214118,
+      "learning_rate": 6.783546405298094e-05,
+      "loss": 0.2531,
+      "step": 3446
+    },
+    {
+      "epoch": 1.9321748878923768,
+      "grad_norm": 0.07114606328300568,
+      "learning_rate": 6.777370493134431e-05,
+      "loss": 0.256,
+      "step": 3447
+    },
+    {
+      "epoch": 1.9327354260089686,
+      "grad_norm": 0.07222637872927773,
+      "learning_rate": 6.77119595228554e-05,
+      "loss": 0.2545,
+      "step": 3448
+    },
+    {
+      "epoch": 1.9332959641255605,
+      "grad_norm": 0.0691511910091209,
+      "learning_rate": 6.765022785378845e-05,
+      "loss": 0.2637,
+      "step": 3449
+    },
+    {
+      "epoch": 1.9338565022421523,
+      "grad_norm": 0.07122319030048847,
+      "learning_rate": 6.758850995041197e-05,
+      "loss": 0.2635,
+      "step": 3450
+    },
+    {
+      "epoch": 1.9344170403587444,
+      "grad_norm": 0.06893474155708976,
+      "learning_rate": 6.752680583898853e-05,
+      "loss": 0.2565,
+      "step": 3451
+    },
+    {
+      "epoch": 1.9349775784753365,
+      "grad_norm": 0.07132881195848476,
+      "learning_rate": 6.746511554577497e-05,
+      "loss": 0.2702,
+      "step": 3452
+    },
+    {
+      "epoch": 1.9355381165919283,
+      "grad_norm": 0.07196555459747075,
+      "learning_rate": 6.740343909702205e-05,
+      "loss": 0.278,
+      "step": 3453
+    },
+    {
+      "epoch": 1.9360986547085202,
+      "grad_norm": 0.07010298135227981,
+      "learning_rate": 6.734177651897475e-05,
+      "loss": 0.2616,
+      "step": 3454
+    },
+    {
+      "epoch": 1.936659192825112,
+      "grad_norm": 0.06726790355415212,
+      "learning_rate": 6.728012783787224e-05,
+      "loss": 0.2599,
+      "step": 3455
+    },
+    {
+      "epoch": 1.937219730941704,
+      "grad_norm": 0.06985074552731459,
+      "learning_rate": 6.721849307994756e-05,
+      "loss": 0.2661,
+      "step": 3456
+    },
+    {
+      "epoch": 1.937780269058296,
+      "grad_norm": 0.0683660746443686,
+      "learning_rate": 6.715687227142804e-05,
+      "loss": 0.2457,
+      "step": 3457
+    },
+    {
+      "epoch": 1.938340807174888,
+      "grad_norm": 0.06873087698366843,
+      "learning_rate": 6.709526543853489e-05,
+      "loss": 0.2642,
+      "step": 3458
+    },
+    {
+      "epoch": 1.9389013452914798,
+      "grad_norm": 0.06885891373281881,
+      "learning_rate": 6.703367260748352e-05,
+      "loss": 0.2658,
+      "step": 3459
+    },
+    {
+      "epoch": 1.9394618834080717,
+      "grad_norm": 0.0711839966087522,
+      "learning_rate": 6.697209380448333e-05,
+      "loss": 0.2787,
+      "step": 3460
+    },
+    {
+      "epoch": 1.9400224215246635,
+      "grad_norm": 0.07203560142958441,
+      "learning_rate": 6.691052905573766e-05,
+      "loss": 0.2779,
+      "step": 3461
+    },
+    {
+      "epoch": 1.9405829596412556,
+      "grad_norm": 0.07064585839757186,
+      "learning_rate": 6.684897838744403e-05,
+      "loss": 0.2639,
+      "step": 3462
+    },
+    {
+      "epoch": 1.9411434977578477,
+      "grad_norm": 0.06814353501211662,
+      "learning_rate": 6.678744182579384e-05,
+      "loss": 0.2565,
+      "step": 3463
+    },
+    {
+      "epoch": 1.9417040358744395,
+      "grad_norm": 0.07081774051467991,
+      "learning_rate": 6.672591939697261e-05,
+      "loss": 0.273,
+      "step": 3464
+    },
+    {
+      "epoch": 1.9422645739910314,
+      "grad_norm": 0.06898429138012982,
+      "learning_rate": 6.66644111271597e-05,
+      "loss": 0.2561,
+      "step": 3465
+    },
+    {
+      "epoch": 1.9428251121076232,
+      "grad_norm": 0.06893255463540728,
+      "learning_rate": 6.660291704252855e-05,
+      "loss": 0.2623,
+      "step": 3466
+    },
+    {
+      "epoch": 1.9433856502242153,
+      "grad_norm": 0.06918468020767439,
+      "learning_rate": 6.654143716924656e-05,
+      "loss": 0.2546,
+      "step": 3467
+    },
+    {
+      "epoch": 1.9439461883408071,
+      "grad_norm": 0.0694607758966827,
+      "learning_rate": 6.647997153347498e-05,
+      "loss": 0.2608,
+      "step": 3468
+    },
+    {
+      "epoch": 1.9445067264573992,
+      "grad_norm": 0.06975092141192037,
+      "learning_rate": 6.641852016136916e-05,
+      "loss": 0.2609,
+      "step": 3469
+    },
+    {
+      "epoch": 1.945067264573991,
+      "grad_norm": 0.06913760605936561,
+      "learning_rate": 6.635708307907822e-05,
+      "loss": 0.2623,
+      "step": 3470
+    },
+    {
+      "epoch": 1.9456278026905829,
+      "grad_norm": 0.0875483678068613,
+      "learning_rate": 6.629566031274532e-05,
+      "loss": 0.2584,
+      "step": 3471
+    },
+    {
+      "epoch": 1.9461883408071747,
+      "grad_norm": 0.07058202455417682,
+      "learning_rate": 6.623425188850746e-05,
+      "loss": 0.2629,
+      "step": 3472
+    },
+    {
+      "epoch": 1.9467488789237668,
+      "grad_norm": 0.07038343018874284,
+      "learning_rate": 6.617285783249563e-05,
+      "loss": 0.2625,
+      "step": 3473
+    },
+    {
+      "epoch": 1.9473094170403589,
+      "grad_norm": 0.06941074905461624,
+      "learning_rate": 6.611147817083456e-05,
+      "loss": 0.2712,
+      "step": 3474
+    },
+    {
+      "epoch": 1.9478699551569507,
+      "grad_norm": 0.0692743253769447,
+      "learning_rate": 6.605011292964297e-05,
+      "loss": 0.2483,
+      "step": 3475
+    },
+    {
+      "epoch": 1.9484304932735426,
+      "grad_norm": 0.06808507809897883,
+      "learning_rate": 6.598876213503339e-05,
+      "loss": 0.2628,
+      "step": 3476
+    },
+    {
+      "epoch": 1.9489910313901344,
+      "grad_norm": 0.0667729391352458,
+      "learning_rate": 6.592742581311221e-05,
+      "loss": 0.2493,
+      "step": 3477
+    },
+    {
+      "epoch": 1.9495515695067265,
+      "grad_norm": 0.07055567641483064,
+      "learning_rate": 6.58661039899797e-05,
+      "loss": 0.2667,
+      "step": 3478
+    },
+    {
+      "epoch": 1.9501121076233185,
+      "grad_norm": 0.0718589113081936,
+      "learning_rate": 6.580479669172989e-05,
+      "loss": 0.273,
+      "step": 3479
+    },
+    {
+      "epoch": 1.9506726457399104,
+      "grad_norm": 0.06705308605935648,
+      "learning_rate": 6.574350394445074e-05,
+      "loss": 0.2557,
+      "step": 3480
+    },
+    {
+      "epoch": 1.9512331838565022,
+      "grad_norm": 0.07119398618447678,
+      "learning_rate": 6.568222577422389e-05,
+      "loss": 0.2691,
+      "step": 3481
+    },
+    {
+      "epoch": 1.951793721973094,
+      "grad_norm": 0.06839092796226122,
+      "learning_rate": 6.562096220712482e-05,
+      "loss": 0.2626,
+      "step": 3482
+    },
+    {
+      "epoch": 1.952354260089686,
+      "grad_norm": 0.07330272242824666,
+      "learning_rate": 6.555971326922286e-05,
+      "loss": 0.2766,
+      "step": 3483
+    },
+    {
+      "epoch": 1.952914798206278,
+      "grad_norm": 0.06811195070717267,
+      "learning_rate": 6.549847898658102e-05,
+      "loss": 0.2589,
+      "step": 3484
+    },
+    {
+      "epoch": 1.95347533632287,
+      "grad_norm": 0.0700191072296125,
+      "learning_rate": 6.54372593852562e-05,
+      "loss": 0.2665,
+      "step": 3485
+    },
+    {
+      "epoch": 1.954035874439462,
+      "grad_norm": 0.06806656260491024,
+      "learning_rate": 6.537605449129888e-05,
+      "loss": 0.2716,
+      "step": 3486
+    },
+    {
+      "epoch": 1.9545964125560538,
+      "grad_norm": 0.06902008174469523,
+      "learning_rate": 6.531486433075339e-05,
+      "loss": 0.2712,
+      "step": 3487
+    },
+    {
+      "epoch": 1.9551569506726456,
+      "grad_norm": 0.06943678410535838,
+      "learning_rate": 6.525368892965784e-05,
+      "loss": 0.2554,
+      "step": 3488
+    },
+    {
+      "epoch": 1.9557174887892377,
+      "grad_norm": 0.068487552278527,
+      "learning_rate": 6.519252831404392e-05,
+      "loss": 0.2624,
+      "step": 3489
+    },
+    {
+      "epoch": 1.9562780269058297,
+      "grad_norm": 0.07075715030821601,
+      "learning_rate": 6.513138250993716e-05,
+      "loss": 0.2714,
+      "step": 3490
+    },
+    {
+      "epoch": 1.9568385650224216,
+      "grad_norm": 0.07190539931059477,
+      "learning_rate": 6.507025154335666e-05,
+      "loss": 0.2693,
+      "step": 3491
+    },
+    {
+      "epoch": 1.9573991031390134,
+      "grad_norm": 0.06926257962240638,
+      "learning_rate": 6.500913544031534e-05,
+      "loss": 0.2444,
+      "step": 3492
+    },
+    {
+      "epoch": 1.9579596412556053,
+      "grad_norm": 0.07127571826060886,
+      "learning_rate": 6.494803422681972e-05,
+      "loss": 0.262,
+      "step": 3493
+    },
+    {
+      "epoch": 1.9585201793721974,
+      "grad_norm": 0.07075473239262661,
+      "learning_rate": 6.488694792886996e-05,
+      "loss": 0.2637,
+      "step": 3494
+    },
+    {
+      "epoch": 1.9590807174887892,
+      "grad_norm": 0.0718492394912454,
+      "learning_rate": 6.482587657245994e-05,
+      "loss": 0.2638,
+      "step": 3495
+    },
+    {
+      "epoch": 1.9596412556053813,
+      "grad_norm": 0.06909919217224249,
+      "learning_rate": 6.476482018357713e-05,
+      "loss": 0.2572,
+      "step": 3496
+    },
+    {
+      "epoch": 1.9602017937219731,
+      "grad_norm": 0.06955860991608306,
+      "learning_rate": 6.47037787882027e-05,
+      "loss": 0.2627,
+      "step": 3497
+    },
+    {
+      "epoch": 1.960762331838565,
+      "grad_norm": 0.06980360814637783,
+      "learning_rate": 6.464275241231132e-05,
+      "loss": 0.2633,
+      "step": 3498
+    },
+    {
+      "epoch": 1.9613228699551568,
+      "grad_norm": 0.06713565112040079,
+      "learning_rate": 6.458174108187139e-05,
+      "loss": 0.2634,
+      "step": 3499
+    },
+    {
+      "epoch": 1.9618834080717489,
+      "grad_norm": 0.06954400272833183,
+      "learning_rate": 6.452074482284487e-05,
+      "loss": 0.264,
+      "step": 3500
+    },
+    {
+      "epoch": 1.962443946188341,
+      "grad_norm": 0.06805182739731903,
+      "learning_rate": 6.445976366118722e-05,
+      "loss": 0.2667,
+      "step": 3501
+    },
+    {
+      "epoch": 1.9630044843049328,
+      "grad_norm": 0.0685719547389835,
+      "learning_rate": 6.439879762284763e-05,
+      "loss": 0.2487,
+      "step": 3502
+    },
+    {
+      "epoch": 1.9635650224215246,
+      "grad_norm": 0.07172031683536766,
+      "learning_rate": 6.43378467337687e-05,
+      "loss": 0.2704,
+      "step": 3503
+    },
+    {
+      "epoch": 1.9641255605381165,
+      "grad_norm": 0.06963178535603862,
+      "learning_rate": 6.427691101988673e-05,
+      "loss": 0.2498,
+      "step": 3504
+    },
+    {
+      "epoch": 1.9646860986547086,
+      "grad_norm": 0.07098641070628586,
+      "learning_rate": 6.421599050713144e-05,
+      "loss": 0.2725,
+      "step": 3505
+    },
+    {
+      "epoch": 1.9652466367713004,
+      "grad_norm": 0.06828147116270429,
+      "learning_rate": 6.415508522142619e-05,
+      "loss": 0.2574,
+      "step": 3506
+    },
+    {
+      "epoch": 1.9658071748878925,
+      "grad_norm": 0.06957147042938308,
+      "learning_rate": 6.409419518868775e-05,
+      "loss": 0.2633,
+      "step": 3507
+    },
+    {
+      "epoch": 1.9663677130044843,
+      "grad_norm": 0.07109989970710494,
+      "learning_rate": 6.403332043482643e-05,
+      "loss": 0.273,
+      "step": 3508
+    },
+    {
+      "epoch": 1.9669282511210762,
+      "grad_norm": 0.0672142330811063,
+      "learning_rate": 6.397246098574615e-05,
+      "loss": 0.254,
+      "step": 3509
+    },
+    {
+      "epoch": 1.967488789237668,
+      "grad_norm": 0.06840471354209353,
+      "learning_rate": 6.391161686734413e-05,
+      "loss": 0.2506,
+      "step": 3510
+    },
+    {
+      "epoch": 1.96804932735426,
+      "grad_norm": 0.06920685502036582,
+      "learning_rate": 6.385078810551124e-05,
+      "loss": 0.2705,
+      "step": 3511
+    },
+    {
+      "epoch": 1.9686098654708521,
+      "grad_norm": 0.07107145289071963,
+      "learning_rate": 6.378997472613169e-05,
+      "loss": 0.2765,
+      "step": 3512
+    },
+    {
+      "epoch": 1.969170403587444,
+      "grad_norm": 0.06773951318983387,
+      "learning_rate": 6.372917675508324e-05,
+      "loss": 0.2552,
+      "step": 3513
+    },
+    {
+      "epoch": 1.9697309417040358,
+      "grad_norm": 0.07167256187046933,
+      "learning_rate": 6.366839421823702e-05,
+      "loss": 0.2688,
+      "step": 3514
+    },
+    {
+      "epoch": 1.9702914798206277,
+      "grad_norm": 0.06966715196630331,
+      "learning_rate": 6.360762714145761e-05,
+      "loss": 0.2431,
+      "step": 3515
+    },
+    {
+      "epoch": 1.9708520179372198,
+      "grad_norm": 0.07022803064435848,
+      "learning_rate": 6.354687555060302e-05,
+      "loss": 0.2622,
+      "step": 3516
+    },
+    {
+      "epoch": 1.9714125560538116,
+      "grad_norm": 0.070004797446652,
+      "learning_rate": 6.348613947152468e-05,
+      "loss": 0.2513,
+      "step": 3517
+    },
+    {
+      "epoch": 1.9719730941704037,
+      "grad_norm": 0.06784047449113982,
+      "learning_rate": 6.342541893006746e-05,
+      "loss": 0.2563,
+      "step": 3518
+    },
+    {
+      "epoch": 1.9725336322869955,
+      "grad_norm": 0.07028830379687871,
+      "learning_rate": 6.336471395206946e-05,
+      "loss": 0.2626,
+      "step": 3519
+    },
+    {
+      "epoch": 1.9730941704035874,
+      "grad_norm": 0.07144555424165255,
+      "learning_rate": 6.330402456336237e-05,
+      "loss": 0.2586,
+      "step": 3520
+    },
+    {
+      "epoch": 1.9736547085201792,
+      "grad_norm": 0.0712109303411046,
+      "learning_rate": 6.324335078977112e-05,
+      "loss": 0.2705,
+      "step": 3521
+    },
+    {
+      "epoch": 1.9742152466367713,
+      "grad_norm": 0.07047632646386391,
+      "learning_rate": 6.318269265711398e-05,
+      "loss": 0.2659,
+      "step": 3522
+    },
+    {
+      "epoch": 1.9747757847533634,
+      "grad_norm": 0.07160627079639259,
+      "learning_rate": 6.312205019120262e-05,
+      "loss": 0.279,
+      "step": 3523
+    },
+    {
+      "epoch": 1.9753363228699552,
+      "grad_norm": 0.06996500426013655,
+      "learning_rate": 6.306142341784202e-05,
+      "loss": 0.251,
+      "step": 3524
+    },
+    {
+      "epoch": 1.975896860986547,
+      "grad_norm": 0.07115808243675245,
+      "learning_rate": 6.300081236283053e-05,
+      "loss": 0.2588,
+      "step": 3525
+    },
+    {
+      "epoch": 1.976457399103139,
+      "grad_norm": 0.07118630719531856,
+      "learning_rate": 6.294021705195974e-05,
+      "loss": 0.2557,
+      "step": 3526
+    },
+    {
+      "epoch": 1.977017937219731,
+      "grad_norm": 0.07209774976734813,
+      "learning_rate": 6.287963751101454e-05,
+      "loss": 0.268,
+      "step": 3527
+    },
+    {
+      "epoch": 1.977578475336323,
+      "grad_norm": 0.07024002145899598,
+      "learning_rate": 6.281907376577316e-05,
+      "loss": 0.2547,
+      "step": 3528
+    },
+    {
+      "epoch": 1.9781390134529149,
+      "grad_norm": 0.0730987438721767,
+      "learning_rate": 6.27585258420071e-05,
+      "loss": 0.2533,
+      "step": 3529
+    },
+    {
+      "epoch": 1.9786995515695067,
+      "grad_norm": 0.06986966309781407,
+      "learning_rate": 6.269799376548116e-05,
+      "loss": 0.2598,
+      "step": 3530
+    },
+    {
+      "epoch": 1.9792600896860986,
+      "grad_norm": 0.06984335708696456,
+      "learning_rate": 6.263747756195324e-05,
+      "loss": 0.27,
+      "step": 3531
+    },
+    {
+      "epoch": 1.9798206278026906,
+      "grad_norm": 0.07196028280505191,
+      "learning_rate": 6.257697725717468e-05,
+      "loss": 0.2614,
+      "step": 3532
+    },
+    {
+      "epoch": 1.9803811659192825,
+      "grad_norm": 0.07156502426950263,
+      "learning_rate": 6.251649287688999e-05,
+      "loss": 0.2699,
+      "step": 3533
+    },
+    {
+      "epoch": 1.9809417040358746,
+      "grad_norm": 0.07167329694846092,
+      "learning_rate": 6.245602444683681e-05,
+      "loss": 0.2541,
+      "step": 3534
+    },
+    {
+      "epoch": 1.9815022421524664,
+      "grad_norm": 0.06847158656031785,
+      "learning_rate": 6.239557199274615e-05,
+      "loss": 0.2589,
+      "step": 3535
+    },
+    {
+      "epoch": 1.9820627802690582,
+      "grad_norm": 0.07134674822776889,
+      "learning_rate": 6.233513554034204e-05,
+      "loss": 0.2695,
+      "step": 3536
+    },
+    {
+      "epoch": 1.98262331838565,
+      "grad_norm": 0.07024867462243686,
+      "learning_rate": 6.227471511534191e-05,
+      "loss": 0.2613,
+      "step": 3537
+    },
+    {
+      "epoch": 1.9831838565022422,
+      "grad_norm": 0.06914310788811501,
+      "learning_rate": 6.221431074345618e-05,
+      "loss": 0.2662,
+      "step": 3538
+    },
+    {
+      "epoch": 1.9837443946188342,
+      "grad_norm": 0.07049521652395627,
+      "learning_rate": 6.21539224503886e-05,
+      "loss": 0.265,
+      "step": 3539
+    },
+    {
+      "epoch": 1.984304932735426,
+      "grad_norm": 0.06930403337463462,
+      "learning_rate": 6.209355026183594e-05,
+      "loss": 0.2531,
+      "step": 3540
+    },
+    {
+      "epoch": 1.984865470852018,
+      "grad_norm": 0.06784220137590388,
+      "learning_rate": 6.203319420348814e-05,
+      "loss": 0.2552,
+      "step": 3541
+    },
+    {
+      "epoch": 1.9854260089686098,
+      "grad_norm": 0.06979882931331614,
+      "learning_rate": 6.197285430102843e-05,
+      "loss": 0.2578,
+      "step": 3542
+    },
+    {
+      "epoch": 1.9859865470852018,
+      "grad_norm": 0.06946776088389077,
+      "learning_rate": 6.191253058013292e-05,
+      "loss": 0.2682,
+      "step": 3543
+    },
+    {
+      "epoch": 1.9865470852017937,
+      "grad_norm": 0.06973110505777286,
+      "learning_rate": 6.185222306647105e-05,
+      "loss": 0.2684,
+      "step": 3544
+    },
+    {
+      "epoch": 1.9871076233183858,
+      "grad_norm": 0.06677543280650287,
+      "learning_rate": 6.179193178570521e-05,
+      "loss": 0.2413,
+      "step": 3545
+    },
+    {
+      "epoch": 1.9876681614349776,
+      "grad_norm": 0.07212011602930748,
+      "learning_rate": 6.173165676349103e-05,
+      "loss": 0.2622,
+      "step": 3546
+    },
+    {
+      "epoch": 1.9882286995515694,
+      "grad_norm": 0.06899791297894088,
+      "learning_rate": 6.167139802547709e-05,
+      "loss": 0.2572,
+      "step": 3547
+    },
+    {
+      "epoch": 1.9887892376681613,
+      "grad_norm": 0.06933956269035081,
+      "learning_rate": 6.161115559730505e-05,
+      "loss": 0.2497,
+      "step": 3548
+    },
+    {
+      "epoch": 1.9893497757847534,
+      "grad_norm": 0.07089229995175773,
+      "learning_rate": 6.155092950460972e-05,
+      "loss": 0.2674,
+      "step": 3549
+    },
+    {
+      "epoch": 1.9899103139013454,
+      "grad_norm": 0.07146566008039273,
+      "learning_rate": 6.149071977301889e-05,
+      "loss": 0.2594,
+      "step": 3550
+    },
+    {
+      "epoch": 1.9904708520179373,
+      "grad_norm": 0.0709130469045352,
+      "learning_rate": 6.143052642815344e-05,
+      "loss": 0.2648,
+      "step": 3551
+    },
+    {
+      "epoch": 1.9910313901345291,
+      "grad_norm": 0.06939505763940963,
+      "learning_rate": 6.137034949562719e-05,
+      "loss": 0.2637,
+      "step": 3552
+    },
+    {
+      "epoch": 1.991591928251121,
+      "grad_norm": 0.06919568454464115,
+      "learning_rate": 6.131018900104705e-05,
+      "loss": 0.2643,
+      "step": 3553
+    },
+    {
+      "epoch": 1.992152466367713,
+      "grad_norm": 0.06839363754516936,
+      "learning_rate": 6.125004497001297e-05,
+      "loss": 0.2638,
+      "step": 3554
+    },
+    {
+      "epoch": 1.9927130044843049,
+      "grad_norm": 0.06941072012337744,
+      "learning_rate": 6.118991742811773e-05,
+      "loss": 0.2485,
+      "step": 3555
+    },
+    {
+      "epoch": 1.993273542600897,
+      "grad_norm": 0.06919893078365716,
+      "learning_rate": 6.112980640094728e-05,
+      "loss": 0.267,
+      "step": 3556
+    },
+    {
+      "epoch": 1.9938340807174888,
+      "grad_norm": 0.06988259270592627,
+      "learning_rate": 6.106971191408042e-05,
+      "loss": 0.2623,
+      "step": 3557
+    },
+    {
+      "epoch": 1.9943946188340806,
+      "grad_norm": 0.07389775122760188,
+      "learning_rate": 6.1009633993089023e-05,
+      "loss": 0.28,
+      "step": 3558
+    },
+    {
+      "epoch": 1.9949551569506725,
+      "grad_norm": 0.07096910171506648,
+      "learning_rate": 6.094957266353776e-05,
+      "loss": 0.2644,
+      "step": 3559
+    },
+    {
+      "epoch": 1.9955156950672646,
+      "grad_norm": 0.0698603598767232,
+      "learning_rate": 6.0889527950984416e-05,
+      "loss": 0.2604,
+      "step": 3560
+    },
+    {
+      "epoch": 1.9960762331838566,
+      "grad_norm": 0.07024849389536654,
+      "learning_rate": 6.082949988097954e-05,
+      "loss": 0.2652,
+      "step": 3561
+    },
+    {
+      "epoch": 1.9966367713004485,
+      "grad_norm": 0.06949137843652796,
+      "learning_rate": 6.0769488479066706e-05,
+      "loss": 0.2584,
+      "step": 3562
+    },
+    {
+      "epoch": 1.9971973094170403,
+      "grad_norm": 0.07090870799176502,
+      "learning_rate": 6.07094937707824e-05,
+      "loss": 0.2627,
+      "step": 3563
+    },
+    {
+      "epoch": 1.9977578475336322,
+      "grad_norm": 0.06914428548303446,
+      "learning_rate": 6.06495157816559e-05,
+      "loss": 0.2642,
+      "step": 3564
+    },
+    {
+      "epoch": 1.9983183856502242,
+      "grad_norm": 0.06946977098314562,
+      "learning_rate": 6.058955453720949e-05,
+      "loss": 0.261,
+      "step": 3565
+    },
+    {
+      "epoch": 1.9988789237668163,
+      "grad_norm": 0.06896759326172326,
+      "learning_rate": 6.052961006295824e-05,
+      "loss": 0.2573,
+      "step": 3566
+    },
+    {
+      "epoch": 1.9994394618834082,
+      "grad_norm": 0.06819216529004883,
+      "learning_rate": 6.0469682384410195e-05,
+      "loss": 0.2576,
+      "step": 3567
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.06871967606933381,
+      "learning_rate": 6.040977152706613e-05,
+      "loss": 0.257,
+      "step": 3568
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.2704615294933319,
+      "eval_runtime": 342.5117,
+      "eval_samples_per_second": 35.085,
+      "eval_steps_per_second": 1.098,
+      "step": 3568
+    },
+    {
+      "epoch": 2.000560538116592,
+      "grad_norm": 0.06881331512094488,
+      "learning_rate": 6.034987751641967e-05,
+      "loss": 0.2532,
+      "step": 3569
+    },
+    {
+      "epoch": 2.0011210762331837,
+      "grad_norm": 0.06683374304992575,
+      "learning_rate": 6.029000037795738e-05,
+      "loss": 0.258,
+      "step": 3570
+    },
+    {
+      "epoch": 2.001681614349776,
+      "grad_norm": 0.0697166774924195,
+      "learning_rate": 6.023014013715853e-05,
+      "loss": 0.253,
+      "step": 3571
+    },
+    {
+      "epoch": 2.002242152466368,
+      "grad_norm": 0.07162643839747408,
+      "learning_rate": 6.017029681949531e-05,
+      "loss": 0.2525,
+      "step": 3572
+    },
+    {
+      "epoch": 2.0028026905829597,
+      "grad_norm": 0.06976413816504087,
+      "learning_rate": 6.01104704504326e-05,
+      "loss": 0.2425,
+      "step": 3573
+    },
+    {
+      "epoch": 2.0033632286995515,
+      "grad_norm": 0.06861865906982345,
+      "learning_rate": 6.005066105542809e-05,
+      "loss": 0.2374,
+      "step": 3574
+    },
+    {
+      "epoch": 2.0039237668161434,
+      "grad_norm": 0.07142251589075896,
+      "learning_rate": 5.999086865993236e-05,
+      "loss": 0.2382,
+      "step": 3575
+    },
+    {
+      "epoch": 2.004484304932735,
+      "grad_norm": 0.07156604660459492,
+      "learning_rate": 5.9931093289388576e-05,
+      "loss": 0.2458,
+      "step": 3576
+    },
+    {
+      "epoch": 2.0050448430493275,
+      "grad_norm": 0.07280334146812427,
+      "learning_rate": 5.987133496923281e-05,
+      "loss": 0.2481,
+      "step": 3577
+    },
+    {
+      "epoch": 2.0056053811659194,
+      "grad_norm": 0.07492042606750819,
+      "learning_rate": 5.981159372489378e-05,
+      "loss": 0.2453,
+      "step": 3578
+    },
+    {
+      "epoch": 2.006165919282511,
+      "grad_norm": 0.0758618656234295,
+      "learning_rate": 5.975186958179304e-05,
+      "loss": 0.2584,
+      "step": 3579
+    },
+    {
+      "epoch": 2.006726457399103,
+      "grad_norm": 0.07644821585164174,
+      "learning_rate": 5.9692162565344755e-05,
+      "loss": 0.2498,
+      "step": 3580
+    },
+    {
+      "epoch": 2.007286995515695,
+      "grad_norm": 0.07455617340795175,
+      "learning_rate": 5.9632472700955846e-05,
+      "loss": 0.2466,
+      "step": 3581
+    },
+    {
+      "epoch": 2.007847533632287,
+      "grad_norm": 0.07172595519992646,
+      "learning_rate": 5.957280001402595e-05,
+      "loss": 0.2308,
+      "step": 3582
+    },
+    {
+      "epoch": 2.008408071748879,
+      "grad_norm": 0.07367491550452068,
+      "learning_rate": 5.951314452994738e-05,
+      "loss": 0.2506,
+      "step": 3583
+    },
+    {
+      "epoch": 2.008968609865471,
+      "grad_norm": 0.072440305357167,
+      "learning_rate": 5.94535062741052e-05,
+      "loss": 0.2336,
+      "step": 3584
+    },
+    {
+      "epoch": 2.0095291479820627,
+      "grad_norm": 0.0735028091121321,
+      "learning_rate": 5.939388527187697e-05,
+      "loss": 0.2429,
+      "step": 3585
+    },
+    {
+      "epoch": 2.0100896860986546,
+      "grad_norm": 0.07316342029447456,
+      "learning_rate": 5.9334281548633106e-05,
+      "loss": 0.2564,
+      "step": 3586
+    },
+    {
+      "epoch": 2.0106502242152464,
+      "grad_norm": 0.0734512260587328,
+      "learning_rate": 5.927469512973656e-05,
+      "loss": 0.2542,
+      "step": 3587
+    },
+    {
+      "epoch": 2.0112107623318387,
+      "grad_norm": 0.07391007018946201,
+      "learning_rate": 5.921512604054289e-05,
+      "loss": 0.2423,
+      "step": 3588
+    },
+    {
+      "epoch": 2.0117713004484306,
+      "grad_norm": 0.07519596716498071,
+      "learning_rate": 5.9155574306400395e-05,
+      "loss": 0.2535,
+      "step": 3589
+    },
+    {
+      "epoch": 2.0123318385650224,
+      "grad_norm": 0.0729143913336409,
+      "learning_rate": 5.9096039952649876e-05,
+      "loss": 0.2478,
+      "step": 3590
+    },
+    {
+      "epoch": 2.0128923766816142,
+      "grad_norm": 0.07194680849829693,
+      "learning_rate": 5.903652300462485e-05,
+      "loss": 0.2474,
+      "step": 3591
+    },
+    {
+      "epoch": 2.013452914798206,
+      "grad_norm": 0.07168630931317939,
+      "learning_rate": 5.897702348765129e-05,
+      "loss": 0.2453,
+      "step": 3592
+    },
+    {
+      "epoch": 2.0140134529147984,
+      "grad_norm": 0.06919176725888804,
+      "learning_rate": 5.891754142704791e-05,
+      "loss": 0.2354,
+      "step": 3593
+    },
+    {
+      "epoch": 2.0145739910313902,
+      "grad_norm": 0.07349036326309909,
+      "learning_rate": 5.885807684812584e-05,
+      "loss": 0.2501,
+      "step": 3594
+    },
+    {
+      "epoch": 2.015134529147982,
+      "grad_norm": 0.0724354831554613,
+      "learning_rate": 5.879862977618886e-05,
+      "loss": 0.242,
+      "step": 3595
+    },
+    {
+      "epoch": 2.015695067264574,
+      "grad_norm": 0.07670077811205034,
+      "learning_rate": 5.873920023653332e-05,
+      "loss": 0.2384,
+      "step": 3596
+    },
+    {
+      "epoch": 2.0162556053811658,
+      "grad_norm": 0.072267893958526,
+      "learning_rate": 5.867978825444802e-05,
+      "loss": 0.25,
+      "step": 3597
+    },
+    {
+      "epoch": 2.016816143497758,
+      "grad_norm": 0.07544364867089007,
+      "learning_rate": 5.8620393855214384e-05,
+      "loss": 0.2553,
+      "step": 3598
+    },
+    {
+      "epoch": 2.01737668161435,
+      "grad_norm": 0.07298918720019351,
+      "learning_rate": 5.856101706410628e-05,
+      "loss": 0.2356,
+      "step": 3599
+    },
+    {
+      "epoch": 2.0179372197309418,
+      "grad_norm": 0.07275824547402367,
+      "learning_rate": 5.8501657906390175e-05,
+      "loss": 0.2354,
+      "step": 3600
+    },
+    {
+      "epoch": 2.0184977578475336,
+      "grad_norm": 0.0746848707308173,
+      "learning_rate": 5.8442316407324895e-05,
+      "loss": 0.2375,
+      "step": 3601
+    },
+    {
+      "epoch": 2.0190582959641254,
+      "grad_norm": 0.07612609715007827,
+      "learning_rate": 5.838299259216187e-05,
+      "loss": 0.245,
+      "step": 3602
+    },
+    {
+      "epoch": 2.0196188340807173,
+      "grad_norm": 0.07619694838706523,
+      "learning_rate": 5.832368648614499e-05,
+      "loss": 0.247,
+      "step": 3603
+    },
+    {
+      "epoch": 2.0201793721973096,
+      "grad_norm": 0.0740977327631378,
+      "learning_rate": 5.826439811451052e-05,
+      "loss": 0.2423,
+      "step": 3604
+    },
+    {
+      "epoch": 2.0207399103139014,
+      "grad_norm": 0.0731684928055168,
+      "learning_rate": 5.820512750248731e-05,
+      "loss": 0.2431,
+      "step": 3605
+    },
+    {
+      "epoch": 2.0213004484304933,
+      "grad_norm": 0.07577288505530066,
+      "learning_rate": 5.814587467529652e-05,
+      "loss": 0.2434,
+      "step": 3606
+    },
+    {
+      "epoch": 2.021860986547085,
+      "grad_norm": 0.07507269911073491,
+      "learning_rate": 5.808663965815188e-05,
+      "loss": 0.2275,
+      "step": 3607
+    },
+    {
+      "epoch": 2.022421524663677,
+      "grad_norm": 0.07513388383068403,
+      "learning_rate": 5.8027422476259385e-05,
+      "loss": 0.2419,
+      "step": 3608
+    },
+    {
+      "epoch": 2.0229820627802693,
+      "grad_norm": 0.0741382324111763,
+      "learning_rate": 5.796822315481758e-05,
+      "loss": 0.2493,
+      "step": 3609
+    },
+    {
+      "epoch": 2.023542600896861,
+      "grad_norm": 0.07569066865477501,
+      "learning_rate": 5.7909041719017385e-05,
+      "loss": 0.2367,
+      "step": 3610
+    },
+    {
+      "epoch": 2.024103139013453,
+      "grad_norm": 0.07450803937671906,
+      "learning_rate": 5.7849878194042e-05,
+      "loss": 0.2493,
+      "step": 3611
+    },
+    {
+      "epoch": 2.024663677130045,
+      "grad_norm": 0.07204647458005459,
+      "learning_rate": 5.779073260506713e-05,
+      "loss": 0.239,
+      "step": 3612
+    },
+    {
+      "epoch": 2.0252242152466366,
+      "grad_norm": 0.07400742953868136,
+      "learning_rate": 5.773160497726082e-05,
+      "loss": 0.2439,
+      "step": 3613
+    },
+    {
+      "epoch": 2.0257847533632285,
+      "grad_norm": 0.07234116837916794,
+      "learning_rate": 5.767249533578338e-05,
+      "loss": 0.2388,
+      "step": 3614
+    },
+    {
+      "epoch": 2.026345291479821,
+      "grad_norm": 0.0729479198673305,
+      "learning_rate": 5.76134037057876e-05,
+      "loss": 0.2447,
+      "step": 3615
+    },
+    {
+      "epoch": 2.0269058295964126,
+      "grad_norm": 0.07548523547570704,
+      "learning_rate": 5.755433011241851e-05,
+      "loss": 0.2467,
+      "step": 3616
+    },
+    {
+      "epoch": 2.0274663677130045,
+      "grad_norm": 0.07356258567090886,
+      "learning_rate": 5.7495274580813494e-05,
+      "loss": 0.2451,
+      "step": 3617
+    },
+    {
+      "epoch": 2.0280269058295963,
+      "grad_norm": 0.07503584154196515,
+      "learning_rate": 5.743623713610229e-05,
+      "loss": 0.2521,
+      "step": 3618
+    },
+    {
+      "epoch": 2.028587443946188,
+      "grad_norm": 0.07503181429152264,
+      "learning_rate": 5.7377217803406925e-05,
+      "loss": 0.2602,
+      "step": 3619
+    },
+    {
+      "epoch": 2.0291479820627805,
+      "grad_norm": 0.07497700035198634,
+      "learning_rate": 5.7318216607841644e-05,
+      "loss": 0.2538,
+      "step": 3620
+    },
+    {
+      "epoch": 2.0297085201793723,
+      "grad_norm": 0.07440507985772586,
+      "learning_rate": 5.7259233574513025e-05,
+      "loss": 0.2401,
+      "step": 3621
+    },
+    {
+      "epoch": 2.030269058295964,
+      "grad_norm": 0.0754323185863951,
+      "learning_rate": 5.720026872851998e-05,
+      "loss": 0.2368,
+      "step": 3622
+    },
+    {
+      "epoch": 2.030829596412556,
+      "grad_norm": 0.07543000929160433,
+      "learning_rate": 5.714132209495354e-05,
+      "loss": 0.2475,
+      "step": 3623
+    },
+    {
+      "epoch": 2.031390134529148,
+      "grad_norm": 0.07470759129408015,
+      "learning_rate": 5.7082393698897166e-05,
+      "loss": 0.2535,
+      "step": 3624
+    },
+    {
+      "epoch": 2.0319506726457397,
+      "grad_norm": 0.0738447933102809,
+      "learning_rate": 5.702348356542635e-05,
+      "loss": 0.239,
+      "step": 3625
+    },
+    {
+      "epoch": 2.032511210762332,
+      "grad_norm": 0.07437147445974224,
+      "learning_rate": 5.696459171960899e-05,
+      "loss": 0.249,
+      "step": 3626
+    },
+    {
+      "epoch": 2.033071748878924,
+      "grad_norm": 0.07427215677895992,
+      "learning_rate": 5.6905718186505185e-05,
+      "loss": 0.2385,
+      "step": 3627
+    },
+    {
+      "epoch": 2.0336322869955157,
+      "grad_norm": 0.072072620232772,
+      "learning_rate": 5.684686299116709e-05,
+      "loss": 0.2401,
+      "step": 3628
+    },
+    {
+      "epoch": 2.0341928251121075,
+      "grad_norm": 0.07418752073168021,
+      "learning_rate": 5.678802615863925e-05,
+      "loss": 0.2432,
+      "step": 3629
+    },
+    {
+      "epoch": 2.0347533632286994,
+      "grad_norm": 0.07376478735097676,
+      "learning_rate": 5.672920771395822e-05,
+      "loss": 0.2476,
+      "step": 3630
+    },
+    {
+      "epoch": 2.0353139013452917,
+      "grad_norm": 0.07317600360599674,
+      "learning_rate": 5.6670407682152906e-05,
+      "loss": 0.247,
+      "step": 3631
+    },
+    {
+      "epoch": 2.0358744394618835,
+      "grad_norm": 0.07548459192437802,
+      "learning_rate": 5.6611626088244194e-05,
+      "loss": 0.2518,
+      "step": 3632
+    },
+    {
+      "epoch": 2.0364349775784754,
+      "grad_norm": 0.07673378194190912,
+      "learning_rate": 5.655286295724528e-05,
+      "loss": 0.2504,
+      "step": 3633
+    },
+    {
+      "epoch": 2.036995515695067,
+      "grad_norm": 0.07594685544914863,
+      "learning_rate": 5.649411831416147e-05,
+      "loss": 0.2519,
+      "step": 3634
+    },
+    {
+      "epoch": 2.037556053811659,
+      "grad_norm": 0.0789259362724179,
+      "learning_rate": 5.643539218399009e-05,
+      "loss": 0.2558,
+      "step": 3635
+    },
+    {
+      "epoch": 2.038116591928251,
+      "grad_norm": 0.07513976547343512,
+      "learning_rate": 5.6376684591720766e-05,
+      "loss": 0.2471,
+      "step": 3636
+    },
+    {
+      "epoch": 2.038677130044843,
+      "grad_norm": 0.07673956293110795,
+      "learning_rate": 5.6317995562335055e-05,
+      "loss": 0.2542,
+      "step": 3637
+    },
+    {
+      "epoch": 2.039237668161435,
+      "grad_norm": 0.07422344910909859,
+      "learning_rate": 5.625932512080678e-05,
+      "loss": 0.2357,
+      "step": 3638
+    },
+    {
+      "epoch": 2.039798206278027,
+      "grad_norm": 0.0770687977508266,
+      "learning_rate": 5.620067329210172e-05,
+      "loss": 0.2477,
+      "step": 3639
+    },
+    {
+      "epoch": 2.0403587443946187,
+      "grad_norm": 0.07477312610911233,
+      "learning_rate": 5.614204010117785e-05,
+      "loss": 0.2458,
+      "step": 3640
+    },
+    {
+      "epoch": 2.0409192825112106,
+      "grad_norm": 0.07703501507752882,
+      "learning_rate": 5.608342557298508e-05,
+      "loss": 0.2691,
+      "step": 3641
+    },
+    {
+      "epoch": 2.041479820627803,
+      "grad_norm": 0.07425971126281866,
+      "learning_rate": 5.60248297324655e-05,
+      "loss": 0.2518,
+      "step": 3642
+    },
+    {
+      "epoch": 2.0420403587443947,
+      "grad_norm": 0.07409210468771125,
+      "learning_rate": 5.596625260455324e-05,
+      "loss": 0.2459,
+      "step": 3643
+    },
+    {
+      "epoch": 2.0426008968609866,
+      "grad_norm": 0.07743348048971965,
+      "learning_rate": 5.5907694214174344e-05,
+      "loss": 0.2474,
+      "step": 3644
+    },
+    {
+      "epoch": 2.0431614349775784,
+      "grad_norm": 0.07555808441932708,
+      "learning_rate": 5.584915458624706e-05,
+      "loss": 0.2417,
+      "step": 3645
+    },
+    {
+      "epoch": 2.0437219730941703,
+      "grad_norm": 0.07646567090721604,
+      "learning_rate": 5.5790633745681475e-05,
+      "loss": 0.2443,
+      "step": 3646
+    },
+    {
+      "epoch": 2.0442825112107625,
+      "grad_norm": 0.07391358861055916,
+      "learning_rate": 5.573213171737983e-05,
+      "loss": 0.2551,
+      "step": 3647
+    },
+    {
+      "epoch": 2.0448430493273544,
+      "grad_norm": 0.07467413970510883,
+      "learning_rate": 5.567364852623629e-05,
+      "loss": 0.2453,
+      "step": 3648
+    },
+    {
+      "epoch": 2.0454035874439462,
+      "grad_norm": 0.0755515448790851,
+      "learning_rate": 5.561518419713695e-05,
+      "loss": 0.2385,
+      "step": 3649
+    },
+    {
+      "epoch": 2.045964125560538,
+      "grad_norm": 0.07519782812949223,
+      "learning_rate": 5.555673875495999e-05,
+      "loss": 0.2473,
+      "step": 3650
+    },
+    {
+      "epoch": 2.04652466367713,
+      "grad_norm": 0.07838271136523178,
+      "learning_rate": 5.549831222457549e-05,
+      "loss": 0.2515,
+      "step": 3651
+    },
+    {
+      "epoch": 2.0470852017937218,
+      "grad_norm": 0.08057752662737491,
+      "learning_rate": 5.543990463084554e-05,
+      "loss": 0.2409,
+      "step": 3652
+    },
+    {
+      "epoch": 2.047645739910314,
+      "grad_norm": 0.07357099627993306,
+      "learning_rate": 5.538151599862407e-05,
+      "loss": 0.2443,
+      "step": 3653
+    },
+    {
+      "epoch": 2.048206278026906,
+      "grad_norm": 0.07528720685237601,
+      "learning_rate": 5.532314635275705e-05,
+      "loss": 0.2471,
+      "step": 3654
+    },
+    {
+      "epoch": 2.0487668161434978,
+      "grad_norm": 0.07640859527958416,
+      "learning_rate": 5.52647957180823e-05,
+      "loss": 0.2572,
+      "step": 3655
+    },
+    {
+      "epoch": 2.0493273542600896,
+      "grad_norm": 0.07577679780106567,
+      "learning_rate": 5.520646411942951e-05,
+      "loss": 0.2488,
+      "step": 3656
+    },
+    {
+      "epoch": 2.0498878923766815,
+      "grad_norm": 0.07390965585458116,
+      "learning_rate": 5.514815158162041e-05,
+      "loss": 0.2461,
+      "step": 3657
+    },
+    {
+      "epoch": 2.0504484304932737,
+      "grad_norm": 0.07605508395958041,
+      "learning_rate": 5.50898581294685e-05,
+      "loss": 0.2418,
+      "step": 3658
+    },
+    {
+      "epoch": 2.0510089686098656,
+      "grad_norm": 0.07787568561042851,
+      "learning_rate": 5.503158378777924e-05,
+      "loss": 0.2527,
+      "step": 3659
+    },
+    {
+      "epoch": 2.0515695067264574,
+      "grad_norm": 0.07379881671794046,
+      "learning_rate": 5.497332858134992e-05,
+      "loss": 0.2462,
+      "step": 3660
+    },
+    {
+      "epoch": 2.0521300448430493,
+      "grad_norm": 0.07474063782729076,
+      "learning_rate": 5.491509253496958e-05,
+      "loss": 0.2439,
+      "step": 3661
+    },
+    {
+      "epoch": 2.052690582959641,
+      "grad_norm": 0.07515187953468082,
+      "learning_rate": 5.4856875673419326e-05,
+      "loss": 0.251,
+      "step": 3662
+    },
+    {
+      "epoch": 2.053251121076233,
+      "grad_norm": 0.07412649718740476,
+      "learning_rate": 5.479867802147191e-05,
+      "loss": 0.2416,
+      "step": 3663
+    },
+    {
+      "epoch": 2.0538116591928253,
+      "grad_norm": 0.0763874254973806,
+      "learning_rate": 5.474049960389205e-05,
+      "loss": 0.2505,
+      "step": 3664
+    },
+    {
+      "epoch": 2.054372197309417,
+      "grad_norm": 0.07464565136436936,
+      "learning_rate": 5.468234044543614e-05,
+      "loss": 0.2457,
+      "step": 3665
+    },
+    {
+      "epoch": 2.054932735426009,
+      "grad_norm": 0.07412344875406843,
+      "learning_rate": 5.462420057085249e-05,
+      "loss": 0.2521,
+      "step": 3666
+    },
+    {
+      "epoch": 2.055493273542601,
+      "grad_norm": 0.07306081151791718,
+      "learning_rate": 5.456608000488119e-05,
+      "loss": 0.2385,
+      "step": 3667
+    },
+    {
+      "epoch": 2.0560538116591927,
+      "grad_norm": 0.07730984821947563,
+      "learning_rate": 5.450797877225404e-05,
+      "loss": 0.2531,
+      "step": 3668
+    },
+    {
+      "epoch": 2.056614349775785,
+      "grad_norm": 0.07521617598903653,
+      "learning_rate": 5.4449896897694744e-05,
+      "loss": 0.2406,
+      "step": 3669
+    },
+    {
+      "epoch": 2.057174887892377,
+      "grad_norm": 0.07760150487580726,
+      "learning_rate": 5.4391834405918574e-05,
+      "loss": 0.2517,
+      "step": 3670
+    },
+    {
+      "epoch": 2.0577354260089686,
+      "grad_norm": 0.07765654951932434,
+      "learning_rate": 5.433379132163279e-05,
+      "loss": 0.2479,
+      "step": 3671
+    },
+    {
+      "epoch": 2.0582959641255605,
+      "grad_norm": 0.07459978677925008,
+      "learning_rate": 5.4275767669536146e-05,
+      "loss": 0.2302,
+      "step": 3672
+    },
+    {
+      "epoch": 2.0588565022421523,
+      "grad_norm": 0.0740377335860988,
+      "learning_rate": 5.421776347431937e-05,
+      "loss": 0.2457,
+      "step": 3673
+    },
+    {
+      "epoch": 2.059417040358744,
+      "grad_norm": 0.07306074248882465,
+      "learning_rate": 5.41597787606647e-05,
+      "loss": 0.2416,
+      "step": 3674
+    },
+    {
+      "epoch": 2.0599775784753365,
+      "grad_norm": 0.07448946926245875,
+      "learning_rate": 5.410181355324622e-05,
+      "loss": 0.2455,
+      "step": 3675
+    },
+    {
+      "epoch": 2.0605381165919283,
+      "grad_norm": 0.07665634256411681,
+      "learning_rate": 5.40438678767297e-05,
+      "loss": 0.2576,
+      "step": 3676
+    },
+    {
+      "epoch": 2.06109865470852,
+      "grad_norm": 0.07741339767568457,
+      "learning_rate": 5.398594175577252e-05,
+      "loss": 0.2497,
+      "step": 3677
+    },
+    {
+      "epoch": 2.061659192825112,
+      "grad_norm": 0.0771709029185898,
+      "learning_rate": 5.392803521502383e-05,
+      "loss": 0.2486,
+      "step": 3678
+    },
+    {
+      "epoch": 2.062219730941704,
+      "grad_norm": 0.07577514077984976,
+      "learning_rate": 5.387014827912434e-05,
+      "loss": 0.2458,
+      "step": 3679
+    },
+    {
+      "epoch": 2.062780269058296,
+      "grad_norm": 0.07395310574854698,
+      "learning_rate": 5.381228097270659e-05,
+      "loss": 0.2309,
+      "step": 3680
+    },
+    {
+      "epoch": 2.063340807174888,
+      "grad_norm": 0.07710178899027828,
+      "learning_rate": 5.375443332039458e-05,
+      "loss": 0.2584,
+      "step": 3681
+    },
+    {
+      "epoch": 2.06390134529148,
+      "grad_norm": 0.074934507801412,
+      "learning_rate": 5.369660534680402e-05,
+      "loss": 0.248,
+      "step": 3682
+    },
+    {
+      "epoch": 2.0644618834080717,
+      "grad_norm": 0.07345878202054623,
+      "learning_rate": 5.363879707654228e-05,
+      "loss": 0.2408,
+      "step": 3683
+    },
+    {
+      "epoch": 2.0650224215246635,
+      "grad_norm": 0.07603644683417077,
+      "learning_rate": 5.3581008534208334e-05,
+      "loss": 0.2451,
+      "step": 3684
+    },
+    {
+      "epoch": 2.0655829596412554,
+      "grad_norm": 0.07407798941233894,
+      "learning_rate": 5.352323974439275e-05,
+      "loss": 0.2324,
+      "step": 3685
+    },
+    {
+      "epoch": 2.0661434977578477,
+      "grad_norm": 0.07589936133155524,
+      "learning_rate": 5.3465490731677655e-05,
+      "loss": 0.2523,
+      "step": 3686
+    },
+    {
+      "epoch": 2.0667040358744395,
+      "grad_norm": 0.07476780548570601,
+      "learning_rate": 5.3407761520636845e-05,
+      "loss": 0.2416,
+      "step": 3687
+    },
+    {
+      "epoch": 2.0672645739910314,
+      "grad_norm": 0.07620405663914806,
+      "learning_rate": 5.3350052135835616e-05,
+      "loss": 0.2473,
+      "step": 3688
+    },
+    {
+      "epoch": 2.067825112107623,
+      "grad_norm": 0.0767194329863688,
+      "learning_rate": 5.32923626018308e-05,
+      "loss": 0.2497,
+      "step": 3689
+    },
+    {
+      "epoch": 2.068385650224215,
+      "grad_norm": 0.07495906496924769,
+      "learning_rate": 5.3234692943170874e-05,
+      "loss": 0.243,
+      "step": 3690
+    },
+    {
+      "epoch": 2.0689461883408073,
+      "grad_norm": 0.07649211839470273,
+      "learning_rate": 5.31770431843958e-05,
+      "loss": 0.2563,
+      "step": 3691
+    },
+    {
+      "epoch": 2.069506726457399,
+      "grad_norm": 0.07558525414436873,
+      "learning_rate": 5.311941335003715e-05,
+      "loss": 0.2414,
+      "step": 3692
+    },
+    {
+      "epoch": 2.070067264573991,
+      "grad_norm": 0.0757366907227106,
+      "learning_rate": 5.306180346461786e-05,
+      "loss": 0.2457,
+      "step": 3693
+    },
+    {
+      "epoch": 2.070627802690583,
+      "grad_norm": 0.07728198121553063,
+      "learning_rate": 5.300421355265257e-05,
+      "loss": 0.2549,
+      "step": 3694
+    },
+    {
+      "epoch": 2.0711883408071747,
+      "grad_norm": 0.07666444202365309,
+      "learning_rate": 5.294664363864725e-05,
+      "loss": 0.2533,
+      "step": 3695
+    },
+    {
+      "epoch": 2.071748878923767,
+      "grad_norm": 0.073571632939005,
+      "learning_rate": 5.2889093747099427e-05,
+      "loss": 0.2451,
+      "step": 3696
+    },
+    {
+      "epoch": 2.072309417040359,
+      "grad_norm": 0.07441857755812561,
+      "learning_rate": 5.283156390249817e-05,
+      "loss": 0.2478,
+      "step": 3697
+    },
+    {
+      "epoch": 2.0728699551569507,
+      "grad_norm": 0.07209931004529369,
+      "learning_rate": 5.27740541293239e-05,
+      "loss": 0.2337,
+      "step": 3698
+    },
+    {
+      "epoch": 2.0734304932735426,
+      "grad_norm": 0.07434303853065642,
+      "learning_rate": 5.27165644520486e-05,
+      "loss": 0.2389,
+      "step": 3699
+    },
+    {
+      "epoch": 2.0739910313901344,
+      "grad_norm": 0.07538332019807867,
+      "learning_rate": 5.265909489513567e-05,
+      "loss": 0.2346,
+      "step": 3700
+    },
+    {
+      "epoch": 2.0745515695067263,
+      "grad_norm": 0.07563528299165076,
+      "learning_rate": 5.2601645483039896e-05,
+      "loss": 0.2503,
+      "step": 3701
+    },
+    {
+      "epoch": 2.0751121076233185,
+      "grad_norm": 0.07550535897521962,
+      "learning_rate": 5.25442162402076e-05,
+      "loss": 0.2457,
+      "step": 3702
+    },
+    {
+      "epoch": 2.0756726457399104,
+      "grad_norm": 0.07708504890256196,
+      "learning_rate": 5.248680719107636e-05,
+      "loss": 0.2437,
+      "step": 3703
+    },
+    {
+      "epoch": 2.0762331838565022,
+      "grad_norm": 0.07926456227678434,
+      "learning_rate": 5.242941836007536e-05,
+      "loss": 0.2365,
+      "step": 3704
+    },
+    {
+      "epoch": 2.076793721973094,
+      "grad_norm": 0.07803839890484698,
+      "learning_rate": 5.237204977162498e-05,
+      "loss": 0.2367,
+      "step": 3705
+    },
+    {
+      "epoch": 2.077354260089686,
+      "grad_norm": 0.0779619945978881,
+      "learning_rate": 5.231470145013717e-05,
+      "loss": 0.2462,
+      "step": 3706
+    },
+    {
+      "epoch": 2.0779147982062782,
+      "grad_norm": 0.07725602039695345,
+      "learning_rate": 5.22573734200151e-05,
+      "loss": 0.2392,
+      "step": 3707
+    },
+    {
+      "epoch": 2.07847533632287,
+      "grad_norm": 0.07703201254512783,
+      "learning_rate": 5.220006570565341e-05,
+      "loss": 0.2483,
+      "step": 3708
+    },
+    {
+      "epoch": 2.079035874439462,
+      "grad_norm": 0.07834171005721667,
+      "learning_rate": 5.214277833143808e-05,
+      "loss": 0.2476,
+      "step": 3709
+    },
+    {
+      "epoch": 2.0795964125560538,
+      "grad_norm": 0.07754509610713888,
+      "learning_rate": 5.208551132174637e-05,
+      "loss": 0.2491,
+      "step": 3710
+    },
+    {
+      "epoch": 2.0801569506726456,
+      "grad_norm": 0.07803061197291622,
+      "learning_rate": 5.202826470094697e-05,
+      "loss": 0.2523,
+      "step": 3711
+    },
+    {
+      "epoch": 2.0807174887892375,
+      "grad_norm": 0.07506148068870845,
+      "learning_rate": 5.197103849339978e-05,
+      "loss": 0.2506,
+      "step": 3712
+    },
+    {
+      "epoch": 2.0812780269058297,
+      "grad_norm": 0.07685695921965507,
+      "learning_rate": 5.1913832723456144e-05,
+      "loss": 0.2492,
+      "step": 3713
+    },
+    {
+      "epoch": 2.0818385650224216,
+      "grad_norm": 0.07827560089693103,
+      "learning_rate": 5.185664741545861e-05,
+      "loss": 0.2478,
+      "step": 3714
+    },
+    {
+      "epoch": 2.0823991031390134,
+      "grad_norm": 0.0772674108978287,
+      "learning_rate": 5.179948259374102e-05,
+      "loss": 0.2453,
+      "step": 3715
+    },
+    {
+      "epoch": 2.0829596412556053,
+      "grad_norm": 0.07533389028706951,
+      "learning_rate": 5.174233828262855e-05,
+      "loss": 0.2378,
+      "step": 3716
+    },
+    {
+      "epoch": 2.083520179372197,
+      "grad_norm": 0.07285465726019562,
+      "learning_rate": 5.1685214506437654e-05,
+      "loss": 0.2393,
+      "step": 3717
+    },
+    {
+      "epoch": 2.0840807174887894,
+      "grad_norm": 0.07656244214074619,
+      "learning_rate": 5.162811128947602e-05,
+      "loss": 0.2458,
+      "step": 3718
+    },
+    {
+      "epoch": 2.0846412556053813,
+      "grad_norm": 0.07510983004184547,
+      "learning_rate": 5.157102865604255e-05,
+      "loss": 0.2477,
+      "step": 3719
+    },
+    {
+      "epoch": 2.085201793721973,
+      "grad_norm": 0.07571182232761092,
+      "learning_rate": 5.151396663042749e-05,
+      "loss": 0.2394,
+      "step": 3720
+    },
+    {
+      "epoch": 2.085762331838565,
+      "grad_norm": 0.07738104482933404,
+      "learning_rate": 5.145692523691222e-05,
+      "loss": 0.2423,
+      "step": 3721
+    },
+    {
+      "epoch": 2.086322869955157,
+      "grad_norm": 0.07424042187491937,
+      "learning_rate": 5.139990449976933e-05,
+      "loss": 0.2552,
+      "step": 3722
+    },
+    {
+      "epoch": 2.086883408071749,
+      "grad_norm": 0.07712723505263727,
+      "learning_rate": 5.1342904443262686e-05,
+      "loss": 0.2429,
+      "step": 3723
+    },
+    {
+      "epoch": 2.087443946188341,
+      "grad_norm": 0.07744484870522164,
+      "learning_rate": 5.128592509164736e-05,
+      "loss": 0.2541,
+      "step": 3724
+    },
+    {
+      "epoch": 2.088004484304933,
+      "grad_norm": 0.07568429844380314,
+      "learning_rate": 5.122896646916959e-05,
+      "loss": 0.24,
+      "step": 3725
+    },
+    {
+      "epoch": 2.0885650224215246,
+      "grad_norm": 0.07667519386845806,
+      "learning_rate": 5.1172028600066757e-05,
+      "loss": 0.2478,
+      "step": 3726
+    },
+    {
+      "epoch": 2.0891255605381165,
+      "grad_norm": 0.07309287635617359,
+      "learning_rate": 5.1115111508567484e-05,
+      "loss": 0.2295,
+      "step": 3727
+    },
+    {
+      "epoch": 2.0896860986547083,
+      "grad_norm": 0.07595344765462039,
+      "learning_rate": 5.105821521889147e-05,
+      "loss": 0.2515,
+      "step": 3728
+    },
+    {
+      "epoch": 2.0902466367713006,
+      "grad_norm": 0.07729452661945743,
+      "learning_rate": 5.100133975524959e-05,
+      "loss": 0.2412,
+      "step": 3729
+    },
+    {
+      "epoch": 2.0908071748878925,
+      "grad_norm": 0.07521225983949126,
+      "learning_rate": 5.094448514184393e-05,
+      "loss": 0.2444,
+      "step": 3730
+    },
+    {
+      "epoch": 2.0913677130044843,
+      "grad_norm": 0.07579948195096538,
+      "learning_rate": 5.0887651402867576e-05,
+      "loss": 0.232,
+      "step": 3731
+    },
+    {
+      "epoch": 2.091928251121076,
+      "grad_norm": 0.07935544607466706,
+      "learning_rate": 5.0830838562504835e-05,
+      "loss": 0.2501,
+      "step": 3732
+    },
+    {
+      "epoch": 2.092488789237668,
+      "grad_norm": 0.07751926544596244,
+      "learning_rate": 5.0774046644931074e-05,
+      "loss": 0.2401,
+      "step": 3733
+    },
+    {
+      "epoch": 2.0930493273542603,
+      "grad_norm": 0.07389189242772543,
+      "learning_rate": 5.0717275674312814e-05,
+      "loss": 0.2245,
+      "step": 3734
+    },
+    {
+      "epoch": 2.093609865470852,
+      "grad_norm": 0.0769607643153231,
+      "learning_rate": 5.066052567480759e-05,
+      "loss": 0.2351,
+      "step": 3735
+    },
+    {
+      "epoch": 2.094170403587444,
+      "grad_norm": 0.0759419778431755,
+      "learning_rate": 5.060379667056399e-05,
+      "loss": 0.2326,
+      "step": 3736
+    },
+    {
+      "epoch": 2.094730941704036,
+      "grad_norm": 0.0792415087895477,
+      "learning_rate": 5.054708868572178e-05,
+      "loss": 0.2551,
+      "step": 3737
+    },
+    {
+      "epoch": 2.0952914798206277,
+      "grad_norm": 0.07855042508720948,
+      "learning_rate": 5.049040174441166e-05,
+      "loss": 0.248,
+      "step": 3738
+    },
+    {
+      "epoch": 2.0958520179372195,
+      "grad_norm": 0.07613422022280052,
+      "learning_rate": 5.043373587075551e-05,
+      "loss": 0.2508,
+      "step": 3739
+    },
+    {
+      "epoch": 2.096412556053812,
+      "grad_norm": 0.0789895223067926,
+      "learning_rate": 5.0377091088866094e-05,
+      "loss": 0.2467,
+      "step": 3740
+    },
+    {
+      "epoch": 2.0969730941704037,
+      "grad_norm": 0.08027646715524697,
+      "learning_rate": 5.032046742284731e-05,
+      "loss": 0.2468,
+      "step": 3741
+    },
+    {
+      "epoch": 2.0975336322869955,
+      "grad_norm": 0.0778152627142349,
+      "learning_rate": 5.026386489679408e-05,
+      "loss": 0.2538,
+      "step": 3742
+    },
+    {
+      "epoch": 2.0980941704035874,
+      "grad_norm": 0.07677584154005475,
+      "learning_rate": 5.0207283534792205e-05,
+      "loss": 0.2481,
+      "step": 3743
+    },
+    {
+      "epoch": 2.098654708520179,
+      "grad_norm": 0.07547205995345732,
+      "learning_rate": 5.015072336091866e-05,
+      "loss": 0.2482,
+      "step": 3744
+    },
+    {
+      "epoch": 2.0992152466367715,
+      "grad_norm": 0.07596555024492226,
+      "learning_rate": 5.0094184399241196e-05,
+      "loss": 0.2442,
+      "step": 3745
+    },
+    {
+      "epoch": 2.0997757847533634,
+      "grad_norm": 0.0760207349377979,
+      "learning_rate": 5.003766667381875e-05,
+      "loss": 0.2504,
+      "step": 3746
+    },
+    {
+      "epoch": 2.100336322869955,
+      "grad_norm": 0.07549347569273997,
+      "learning_rate": 4.998117020870108e-05,
+      "loss": 0.2417,
+      "step": 3747
+    },
+    {
+      "epoch": 2.100896860986547,
+      "grad_norm": 0.07484700862466377,
+      "learning_rate": 4.992469502792889e-05,
+      "loss": 0.2414,
+      "step": 3748
+    },
+    {
+      "epoch": 2.101457399103139,
+      "grad_norm": 0.07366947561702218,
+      "learning_rate": 4.986824115553392e-05,
+      "loss": 0.2335,
+      "step": 3749
+    },
+    {
+      "epoch": 2.1020179372197307,
+      "grad_norm": 0.07484692603803325,
+      "learning_rate": 4.98118086155388e-05,
+      "loss": 0.2469,
+      "step": 3750
+    },
+    {
+      "epoch": 2.102578475336323,
+      "grad_norm": 0.07464427055286087,
+      "learning_rate": 4.9755397431957116e-05,
+      "loss": 0.2523,
+      "step": 3751
+    },
+    {
+      "epoch": 2.103139013452915,
+      "grad_norm": 0.07636046363902961,
+      "learning_rate": 4.969900762879325e-05,
+      "loss": 0.2364,
+      "step": 3752
+    },
+    {
+      "epoch": 2.1036995515695067,
+      "grad_norm": 0.07565136512051622,
+      "learning_rate": 4.9642639230042654e-05,
+      "loss": 0.2394,
+      "step": 3753
+    },
+    {
+      "epoch": 2.1042600896860986,
+      "grad_norm": 0.07499809113500086,
+      "learning_rate": 4.958629225969153e-05,
+      "loss": 0.2441,
+      "step": 3754
+    },
+    {
+      "epoch": 2.1048206278026904,
+      "grad_norm": 0.0728745125772784,
+      "learning_rate": 4.952996674171698e-05,
+      "loss": 0.2364,
+      "step": 3755
+    },
+    {
+      "epoch": 2.1053811659192827,
+      "grad_norm": 0.0776436778018926,
+      "learning_rate": 4.947366270008707e-05,
+      "loss": 0.2356,
+      "step": 3756
+    },
+    {
+      "epoch": 2.1059417040358746,
+      "grad_norm": 0.0748259736118957,
+      "learning_rate": 4.9417380158760663e-05,
+      "loss": 0.2414,
+      "step": 3757
+    },
+    {
+      "epoch": 2.1065022421524664,
+      "grad_norm": 0.07740404251257217,
+      "learning_rate": 4.936111914168749e-05,
+      "loss": 0.2377,
+      "step": 3758
+    },
+    {
+      "epoch": 2.1070627802690582,
+      "grad_norm": 0.07602820206573185,
+      "learning_rate": 4.930487967280809e-05,
+      "loss": 0.2447,
+      "step": 3759
+    },
+    {
+      "epoch": 2.10762331838565,
+      "grad_norm": 0.07744549973889647,
+      "learning_rate": 4.924866177605389e-05,
+      "loss": 0.2449,
+      "step": 3760
+    },
+    {
+      "epoch": 2.108183856502242,
+      "grad_norm": 0.07865211127766736,
+      "learning_rate": 4.919246547534708e-05,
+      "loss": 0.2435,
+      "step": 3761
+    },
+    {
+      "epoch": 2.1087443946188342,
+      "grad_norm": 0.07601670072773913,
+      "learning_rate": 4.913629079460065e-05,
+      "loss": 0.2525,
+      "step": 3762
+    },
+    {
+      "epoch": 2.109304932735426,
+      "grad_norm": 0.07449968206673598,
+      "learning_rate": 4.908013775771849e-05,
+      "loss": 0.2336,
+      "step": 3763
+    },
+    {
+      "epoch": 2.109865470852018,
+      "grad_norm": 0.07913432716177185,
+      "learning_rate": 4.9024006388595155e-05,
+      "loss": 0.2491,
+      "step": 3764
+    },
+    {
+      "epoch": 2.1104260089686098,
+      "grad_norm": 0.0783133059555131,
+      "learning_rate": 4.896789671111606e-05,
+      "loss": 0.2505,
+      "step": 3765
+    },
+    {
+      "epoch": 2.1109865470852016,
+      "grad_norm": 0.0770458655210177,
+      "learning_rate": 4.891180874915737e-05,
+      "loss": 0.2391,
+      "step": 3766
+    },
+    {
+      "epoch": 2.111547085201794,
+      "grad_norm": 0.07804450235053058,
+      "learning_rate": 4.885574252658607e-05,
+      "loss": 0.2512,
+      "step": 3767
+    },
+    {
+      "epoch": 2.1121076233183858,
+      "grad_norm": 0.07719063199576529,
+      "learning_rate": 4.8799698067259757e-05,
+      "loss": 0.2496,
+      "step": 3768
+    },
+    {
+      "epoch": 2.1126681614349776,
+      "grad_norm": 0.07509637258198873,
+      "learning_rate": 4.8743675395026836e-05,
+      "loss": 0.2318,
+      "step": 3769
+    },
+    {
+      "epoch": 2.1132286995515694,
+      "grad_norm": 0.07880164737334229,
+      "learning_rate": 4.868767453372649e-05,
+      "loss": 0.251,
+      "step": 3770
+    },
+    {
+      "epoch": 2.1137892376681613,
+      "grad_norm": 0.07472006746718912,
+      "learning_rate": 4.863169550718855e-05,
+      "loss": 0.2386,
+      "step": 3771
+    },
+    {
+      "epoch": 2.1143497757847536,
+      "grad_norm": 0.07503838079189648,
+      "learning_rate": 4.857573833923361e-05,
+      "loss": 0.2451,
+      "step": 3772
+    },
+    {
+      "epoch": 2.1149103139013454,
+      "grad_norm": 0.07783665493828258,
+      "learning_rate": 4.85198030536729e-05,
+      "loss": 0.2469,
+      "step": 3773
+    },
+    {
+      "epoch": 2.1154708520179373,
+      "grad_norm": 0.07252369190069591,
+      "learning_rate": 4.8463889674308386e-05,
+      "loss": 0.237,
+      "step": 3774
+    },
+    {
+      "epoch": 2.116031390134529,
+      "grad_norm": 0.0767332797547159,
+      "learning_rate": 4.8407998224932746e-05,
+      "loss": 0.2397,
+      "step": 3775
+    },
+    {
+      "epoch": 2.116591928251121,
+      "grad_norm": 0.07314039073294969,
+      "learning_rate": 4.8352128729329226e-05,
+      "loss": 0.241,
+      "step": 3776
+    },
+    {
+      "epoch": 2.117152466367713,
+      "grad_norm": 0.07798493184421496,
+      "learning_rate": 4.8296281211271845e-05,
+      "loss": 0.2569,
+      "step": 3777
+    },
+    {
+      "epoch": 2.117713004484305,
+      "grad_norm": 0.07640627922643582,
+      "learning_rate": 4.824045569452512e-05,
+      "loss": 0.2545,
+      "step": 3778
+    },
+    {
+      "epoch": 2.118273542600897,
+      "grad_norm": 0.07698966097835544,
+      "learning_rate": 4.81846522028444e-05,
+      "loss": 0.2494,
+      "step": 3779
+    },
+    {
+      "epoch": 2.118834080717489,
+      "grad_norm": 0.0777952701462906,
+      "learning_rate": 4.8128870759975474e-05,
+      "loss": 0.2569,
+      "step": 3780
+    },
+    {
+      "epoch": 2.1193946188340806,
+      "grad_norm": 0.07933407709088636,
+      "learning_rate": 4.8073111389654904e-05,
+      "loss": 0.2553,
+      "step": 3781
+    },
+    {
+      "epoch": 2.1199551569506725,
+      "grad_norm": 0.0736801240831016,
+      "learning_rate": 4.8017374115609705e-05,
+      "loss": 0.2325,
+      "step": 3782
+    },
+    {
+      "epoch": 2.120515695067265,
+      "grad_norm": 0.07656959783693124,
+      "learning_rate": 4.796165896155762e-05,
+      "loss": 0.2369,
+      "step": 3783
+    },
+    {
+      "epoch": 2.1210762331838566,
+      "grad_norm": 0.07589478437817554,
+      "learning_rate": 4.790596595120699e-05,
+      "loss": 0.2475,
+      "step": 3784
+    },
+    {
+      "epoch": 2.1216367713004485,
+      "grad_norm": 0.07832222185990736,
+      "learning_rate": 4.785029510825656e-05,
+      "loss": 0.2439,
+      "step": 3785
+    },
+    {
+      "epoch": 2.1221973094170403,
+      "grad_norm": 0.07661442828983898,
+      "learning_rate": 4.7794646456395864e-05,
+      "loss": 0.2286,
+      "step": 3786
+    },
+    {
+      "epoch": 2.122757847533632,
+      "grad_norm": 0.07948573208800389,
+      "learning_rate": 4.7739020019304836e-05,
+      "loss": 0.258,
+      "step": 3787
+    },
+    {
+      "epoch": 2.123318385650224,
+      "grad_norm": 0.07676654468744926,
+      "learning_rate": 4.7683415820653976e-05,
+      "loss": 0.2359,
+      "step": 3788
+    },
+    {
+      "epoch": 2.1238789237668163,
+      "grad_norm": 0.07949196052434777,
+      "learning_rate": 4.7627833884104376e-05,
+      "loss": 0.2539,
+      "step": 3789
+    },
+    {
+      "epoch": 2.124439461883408,
+      "grad_norm": 0.0808029184848327,
+      "learning_rate": 4.757227423330766e-05,
+      "loss": 0.2465,
+      "step": 3790
+    },
+    {
+      "epoch": 2.125,
+      "grad_norm": 0.07684233703423032,
+      "learning_rate": 4.751673689190596e-05,
+      "loss": 0.2465,
+      "step": 3791
+    },
+    {
+      "epoch": 2.125560538116592,
+      "grad_norm": 0.07582523958146972,
+      "learning_rate": 4.746122188353182e-05,
+      "loss": 0.2389,
+      "step": 3792
+    },
+    {
+      "epoch": 2.1261210762331837,
+      "grad_norm": 0.07770405301933939,
+      "learning_rate": 4.740572923180843e-05,
+      "loss": 0.2462,
+      "step": 3793
+    },
+    {
+      "epoch": 2.126681614349776,
+      "grad_norm": 0.07630725806189007,
+      "learning_rate": 4.7350258960349345e-05,
+      "loss": 0.2418,
+      "step": 3794
+    },
+    {
+      "epoch": 2.127242152466368,
+      "grad_norm": 0.08095571713148267,
+      "learning_rate": 4.729481109275864e-05,
+      "loss": 0.2349,
+      "step": 3795
+    },
+    {
+      "epoch": 2.1278026905829597,
+      "grad_norm": 0.07515324286245524,
+      "learning_rate": 4.723938565263091e-05,
+      "loss": 0.2338,
+      "step": 3796
+    },
+    {
+      "epoch": 2.1283632286995515,
+      "grad_norm": 0.07576355358240246,
+      "learning_rate": 4.718398266355109e-05,
+      "loss": 0.2365,
+      "step": 3797
+    },
+    {
+      "epoch": 2.1289237668161434,
+      "grad_norm": 0.08129801619576742,
+      "learning_rate": 4.712860214909466e-05,
+      "loss": 0.2562,
+      "step": 3798
+    },
+    {
+      "epoch": 2.1294843049327357,
+      "grad_norm": 0.07792485345307101,
+      "learning_rate": 4.707324413282751e-05,
+      "loss": 0.2452,
+      "step": 3799
+    },
+    {
+      "epoch": 2.1300448430493275,
+      "grad_norm": 0.07677282962544202,
+      "learning_rate": 4.7017908638305995e-05,
+      "loss": 0.2474,
+      "step": 3800
+    },
+    {
+      "epoch": 2.1306053811659194,
+      "grad_norm": 0.07859457790958559,
+      "learning_rate": 4.6962595689076796e-05,
+      "loss": 0.2527,
+      "step": 3801
+    },
+    {
+      "epoch": 2.131165919282511,
+      "grad_norm": 0.07490852181293503,
+      "learning_rate": 4.6907305308677005e-05,
+      "loss": 0.2365,
+      "step": 3802
+    },
+    {
+      "epoch": 2.131726457399103,
+      "grad_norm": 0.07599246717101346,
+      "learning_rate": 4.685203752063425e-05,
+      "loss": 0.2399,
+      "step": 3803
+    },
+    {
+      "epoch": 2.132286995515695,
+      "grad_norm": 0.07659072204029127,
+      "learning_rate": 4.6796792348466356e-05,
+      "loss": 0.2357,
+      "step": 3804
+    },
+    {
+      "epoch": 2.132847533632287,
+      "grad_norm": 0.07676255779618765,
+      "learning_rate": 4.6741569815681685e-05,
+      "loss": 0.2373,
+      "step": 3805
+    },
+    {
+      "epoch": 2.133408071748879,
+      "grad_norm": 0.07471477190494208,
+      "learning_rate": 4.668636994577884e-05,
+      "loss": 0.2352,
+      "step": 3806
+    },
+    {
+      "epoch": 2.133968609865471,
+      "grad_norm": 0.07791178793759876,
+      "learning_rate": 4.663119276224688e-05,
+      "loss": 0.2506,
+      "step": 3807
+    },
+    {
+      "epoch": 2.1345291479820627,
+      "grad_norm": 0.07603256569795648,
+      "learning_rate": 4.657603828856517e-05,
+      "loss": 0.2459,
+      "step": 3808
+    },
+    {
+      "epoch": 2.1350896860986546,
+      "grad_norm": 0.07359062577381702,
+      "learning_rate": 4.652090654820337e-05,
+      "loss": 0.2343,
+      "step": 3809
+    },
+    {
+      "epoch": 2.1356502242152464,
+      "grad_norm": 0.07916302803967806,
+      "learning_rate": 4.646579756462156e-05,
+      "loss": 0.246,
+      "step": 3810
+    },
+    {
+      "epoch": 2.1362107623318387,
+      "grad_norm": 0.07706270016322592,
+      "learning_rate": 4.641071136127001e-05,
+      "loss": 0.2462,
+      "step": 3811
+    },
+    {
+      "epoch": 2.1367713004484306,
+      "grad_norm": 0.07634153045319463,
+      "learning_rate": 4.635564796158945e-05,
+      "loss": 0.233,
+      "step": 3812
+    },
+    {
+      "epoch": 2.1373318385650224,
+      "grad_norm": 0.07927724125092082,
+      "learning_rate": 4.6300607389010744e-05,
+      "loss": 0.2519,
+      "step": 3813
+    },
+    {
+      "epoch": 2.1378923766816142,
+      "grad_norm": 0.07872347424418544,
+      "learning_rate": 4.62455896669552e-05,
+      "loss": 0.2443,
+      "step": 3814
+    },
+    {
+      "epoch": 2.138452914798206,
+      "grad_norm": 0.07626789547393499,
+      "learning_rate": 4.619059481883425e-05,
+      "loss": 0.2394,
+      "step": 3815
+    },
+    {
+      "epoch": 2.1390134529147984,
+      "grad_norm": 0.07598290348951336,
+      "learning_rate": 4.61356228680497e-05,
+      "loss": 0.2404,
+      "step": 3816
+    },
+    {
+      "epoch": 2.1395739910313902,
+      "grad_norm": 0.07637304699987817,
+      "learning_rate": 4.608067383799363e-05,
+      "loss": 0.2299,
+      "step": 3817
+    },
+    {
+      "epoch": 2.140134529147982,
+      "grad_norm": 0.07640569892654692,
+      "learning_rate": 4.602574775204823e-05,
+      "loss": 0.2452,
+      "step": 3818
+    },
+    {
+      "epoch": 2.140695067264574,
+      "grad_norm": 0.07826351124159894,
+      "learning_rate": 4.59708446335861e-05,
+      "loss": 0.2533,
+      "step": 3819
+    },
+    {
+      "epoch": 2.1412556053811658,
+      "grad_norm": 0.0755196721551444,
+      "learning_rate": 4.59159645059699e-05,
+      "loss": 0.2392,
+      "step": 3820
+    },
+    {
+      "epoch": 2.141816143497758,
+      "grad_norm": 0.07857330298730242,
+      "learning_rate": 4.586110739255266e-05,
+      "loss": 0.2466,
+      "step": 3821
+    },
+    {
+      "epoch": 2.14237668161435,
+      "grad_norm": 0.07687845949230376,
+      "learning_rate": 4.580627331667747e-05,
+      "loss": 0.2484,
+      "step": 3822
+    },
+    {
+      "epoch": 2.1429372197309418,
+      "grad_norm": 0.07632208103163783,
+      "learning_rate": 4.575146230167773e-05,
+      "loss": 0.2339,
+      "step": 3823
+    },
+    {
+      "epoch": 2.1434977578475336,
+      "grad_norm": 0.07970092291202537,
+      "learning_rate": 4.569667437087702e-05,
+      "loss": 0.2515,
+      "step": 3824
+    },
+    {
+      "epoch": 2.1440582959641254,
+      "grad_norm": 0.07747091344484212,
+      "learning_rate": 4.5641909547589e-05,
+      "loss": 0.2395,
+      "step": 3825
+    },
+    {
+      "epoch": 2.1446188340807173,
+      "grad_norm": 0.07660521329433977,
+      "learning_rate": 4.558716785511764e-05,
+      "loss": 0.2499,
+      "step": 3826
+    },
+    {
+      "epoch": 2.1451793721973096,
+      "grad_norm": 0.07656104461730631,
+      "learning_rate": 4.553244931675694e-05,
+      "loss": 0.2388,
+      "step": 3827
+    },
+    {
+      "epoch": 2.1457399103139014,
+      "grad_norm": 0.07737457472872494,
+      "learning_rate": 4.547775395579106e-05,
+      "loss": 0.2294,
+      "step": 3828
+    },
+    {
+      "epoch": 2.1463004484304933,
+      "grad_norm": 0.07474592555709123,
+      "learning_rate": 4.542308179549442e-05,
+      "loss": 0.2377,
+      "step": 3829
+    },
+    {
+      "epoch": 2.146860986547085,
+      "grad_norm": 0.07629029026990886,
+      "learning_rate": 4.5368432859131395e-05,
+      "loss": 0.246,
+      "step": 3830
+    },
+    {
+      "epoch": 2.147421524663677,
+      "grad_norm": 0.08308993641254947,
+      "learning_rate": 4.5313807169956604e-05,
+      "loss": 0.251,
+      "step": 3831
+    },
+    {
+      "epoch": 2.1479820627802693,
+      "grad_norm": 0.07435536864433591,
+      "learning_rate": 4.5259204751214743e-05,
+      "loss": 0.2422,
+      "step": 3832
+    },
+    {
+      "epoch": 2.148542600896861,
+      "grad_norm": 0.0750009085658999,
+      "learning_rate": 4.520462562614063e-05,
+      "loss": 0.2458,
+      "step": 3833
+    },
+    {
+      "epoch": 2.149103139013453,
+      "grad_norm": 0.07542735638758706,
+      "learning_rate": 4.515006981795909e-05,
+      "loss": 0.2362,
+      "step": 3834
+    },
+    {
+      "epoch": 2.149663677130045,
+      "grad_norm": 0.07705255610502441,
+      "learning_rate": 4.5095537349885055e-05,
+      "loss": 0.2429,
+      "step": 3835
+    },
+    {
+      "epoch": 2.1502242152466366,
+      "grad_norm": 0.0759154050047277,
+      "learning_rate": 4.50410282451236e-05,
+      "loss": 0.2407,
+      "step": 3836
+    },
+    {
+      "epoch": 2.1507847533632285,
+      "grad_norm": 0.07820174924523798,
+      "learning_rate": 4.498654252686975e-05,
+      "loss": 0.2489,
+      "step": 3837
+    },
+    {
+      "epoch": 2.151345291479821,
+      "grad_norm": 0.07898805879350591,
+      "learning_rate": 4.493208021830867e-05,
+      "loss": 0.2479,
+      "step": 3838
+    },
+    {
+      "epoch": 2.1519058295964126,
+      "grad_norm": 0.07834726373003945,
+      "learning_rate": 4.487764134261549e-05,
+      "loss": 0.2428,
+      "step": 3839
+    },
+    {
+      "epoch": 2.1524663677130045,
+      "grad_norm": 0.07730436457149517,
+      "learning_rate": 4.48232259229554e-05,
+      "loss": 0.2513,
+      "step": 3840
+    },
+    {
+      "epoch": 2.1530269058295963,
+      "grad_norm": 0.07633422950781828,
+      "learning_rate": 4.4768833982483694e-05,
+      "loss": 0.2443,
+      "step": 3841
+    },
+    {
+      "epoch": 2.153587443946188,
+      "grad_norm": 0.07596439538807573,
+      "learning_rate": 4.471446554434548e-05,
+      "loss": 0.2506,
+      "step": 3842
+    },
+    {
+      "epoch": 2.1541479820627805,
+      "grad_norm": 0.07582511713890523,
+      "learning_rate": 4.466012063167607e-05,
+      "loss": 0.2319,
+      "step": 3843
+    },
+    {
+      "epoch": 2.1547085201793723,
+      "grad_norm": 0.07835604156296745,
+      "learning_rate": 4.460579926760059e-05,
+      "loss": 0.247,
+      "step": 3844
+    },
+    {
+      "epoch": 2.155269058295964,
+      "grad_norm": 0.07985378092104663,
+      "learning_rate": 4.455150147523431e-05,
+      "loss": 0.244,
+      "step": 3845
+    },
+    {
+      "epoch": 2.155829596412556,
+      "grad_norm": 0.07705146405540986,
+      "learning_rate": 4.449722727768233e-05,
+      "loss": 0.251,
+      "step": 3846
+    },
+    {
+      "epoch": 2.156390134529148,
+      "grad_norm": 0.07930047758046223,
+      "learning_rate": 4.444297669803981e-05,
+      "loss": 0.248,
+      "step": 3847
+    },
+    {
+      "epoch": 2.15695067264574,
+      "grad_norm": 0.078031348593723,
+      "learning_rate": 4.4388749759391754e-05,
+      "loss": 0.2436,
+      "step": 3848
+    },
+    {
+      "epoch": 2.157511210762332,
+      "grad_norm": 0.07754219449849102,
+      "learning_rate": 4.433454648481321e-05,
+      "loss": 0.2418,
+      "step": 3849
+    },
+    {
+      "epoch": 2.158071748878924,
+      "grad_norm": 0.07283117307017277,
+      "learning_rate": 4.4280366897369165e-05,
+      "loss": 0.2351,
+      "step": 3850
+    },
+    {
+      "epoch": 2.1586322869955157,
+      "grad_norm": 0.07617950654479107,
+      "learning_rate": 4.422621102011438e-05,
+      "loss": 0.2477,
+      "step": 3851
+    },
+    {
+      "epoch": 2.1591928251121075,
+      "grad_norm": 0.07638222595701658,
+      "learning_rate": 4.417207887609372e-05,
+      "loss": 0.233,
+      "step": 3852
+    },
+    {
+      "epoch": 2.1597533632286994,
+      "grad_norm": 0.07651077585222588,
+      "learning_rate": 4.411797048834179e-05,
+      "loss": 0.2258,
+      "step": 3853
+    },
+    {
+      "epoch": 2.1603139013452917,
+      "grad_norm": 0.07936726239590704,
+      "learning_rate": 4.4063885879883184e-05,
+      "loss": 0.2394,
+      "step": 3854
+    },
+    {
+      "epoch": 2.1608744394618835,
+      "grad_norm": 0.07520368120379198,
+      "learning_rate": 4.40098250737323e-05,
+      "loss": 0.2387,
+      "step": 3855
+    },
+    {
+      "epoch": 2.1614349775784754,
+      "grad_norm": 0.07928479974436303,
+      "learning_rate": 4.395578809289349e-05,
+      "loss": 0.2446,
+      "step": 3856
+    },
+    {
+      "epoch": 2.161995515695067,
+      "grad_norm": 0.07559575442400272,
+      "learning_rate": 4.3901774960360964e-05,
+      "loss": 0.2403,
+      "step": 3857
+    },
+    {
+      "epoch": 2.162556053811659,
+      "grad_norm": 0.07804502998940659,
+      "learning_rate": 4.384778569911867e-05,
+      "loss": 0.236,
+      "step": 3858
+    },
+    {
+      "epoch": 2.163116591928251,
+      "grad_norm": 0.0784335877150939,
+      "learning_rate": 4.379382033214055e-05,
+      "loss": 0.2541,
+      "step": 3859
+    },
+    {
+      "epoch": 2.163677130044843,
+      "grad_norm": 0.07989086111430377,
+      "learning_rate": 4.373987888239024e-05,
+      "loss": 0.2524,
+      "step": 3860
+    },
+    {
+      "epoch": 2.164237668161435,
+      "grad_norm": 0.07613239404875734,
+      "learning_rate": 4.3685961372821336e-05,
+      "loss": 0.2289,
+      "step": 3861
+    },
+    {
+      "epoch": 2.164798206278027,
+      "grad_norm": 0.08097441844842519,
+      "learning_rate": 4.363206782637714e-05,
+      "loss": 0.2547,
+      "step": 3862
+    },
+    {
+      "epoch": 2.1653587443946187,
+      "grad_norm": 0.0792652512977873,
+      "learning_rate": 4.3578198265990765e-05,
+      "loss": 0.2502,
+      "step": 3863
+    },
+    {
+      "epoch": 2.1659192825112106,
+      "grad_norm": 0.07492364153520117,
+      "learning_rate": 4.352435271458516e-05,
+      "loss": 0.242,
+      "step": 3864
+    },
+    {
+      "epoch": 2.166479820627803,
+      "grad_norm": 0.07749362513129567,
+      "learning_rate": 4.347053119507306e-05,
+      "loss": 0.2349,
+      "step": 3865
+    },
+    {
+      "epoch": 2.1670403587443947,
+      "grad_norm": 0.07511795817075329,
+      "learning_rate": 4.341673373035698e-05,
+      "loss": 0.2283,
+      "step": 3866
+    },
+    {
+      "epoch": 2.1676008968609866,
+      "grad_norm": 0.076819330087108,
+      "learning_rate": 4.336296034332912e-05,
+      "loss": 0.2445,
+      "step": 3867
+    },
+    {
+      "epoch": 2.1681614349775784,
+      "grad_norm": 0.07600638908748635,
+      "learning_rate": 4.3309211056871546e-05,
+      "loss": 0.2452,
+      "step": 3868
+    },
+    {
+      "epoch": 2.1687219730941703,
+      "grad_norm": 0.07838996109633528,
+      "learning_rate": 4.3255485893855985e-05,
+      "loss": 0.2417,
+      "step": 3869
+    },
+    {
+      "epoch": 2.1692825112107625,
+      "grad_norm": 0.07612712342833701,
+      "learning_rate": 4.320178487714389e-05,
+      "loss": 0.2517,
+      "step": 3870
+    },
+    {
+      "epoch": 2.1698430493273544,
+      "grad_norm": 0.07750713654226672,
+      "learning_rate": 4.3148108029586545e-05,
+      "loss": 0.2283,
+      "step": 3871
+    },
+    {
+      "epoch": 2.1704035874439462,
+      "grad_norm": 0.07811498338255723,
+      "learning_rate": 4.30944553740248e-05,
+      "loss": 0.2564,
+      "step": 3872
+    },
+    {
+      "epoch": 2.170964125560538,
+      "grad_norm": 0.0761621345285335,
+      "learning_rate": 4.3040826933289335e-05,
+      "loss": 0.236,
+      "step": 3873
+    },
+    {
+      "epoch": 2.17152466367713,
+      "grad_norm": 0.07563388758275286,
+      "learning_rate": 4.2987222730200515e-05,
+      "loss": 0.2462,
+      "step": 3874
+    },
+    {
+      "epoch": 2.1720852017937218,
+      "grad_norm": 0.07708477410440934,
+      "learning_rate": 4.2933642787568293e-05,
+      "loss": 0.2379,
+      "step": 3875
+    },
+    {
+      "epoch": 2.172645739910314,
+      "grad_norm": 0.07971412225618608,
+      "learning_rate": 4.288008712819243e-05,
+      "loss": 0.2413,
+      "step": 3876
+    },
+    {
+      "epoch": 2.173206278026906,
+      "grad_norm": 0.07893113768036324,
+      "learning_rate": 4.282655577486221e-05,
+      "loss": 0.2461,
+      "step": 3877
+    },
+    {
+      "epoch": 2.1737668161434978,
+      "grad_norm": 0.0758942562754993,
+      "learning_rate": 4.2773048750356716e-05,
+      "loss": 0.2466,
+      "step": 3878
+    },
+    {
+      "epoch": 2.1743273542600896,
+      "grad_norm": 0.07729882209380876,
+      "learning_rate": 4.2719566077444565e-05,
+      "loss": 0.2308,
+      "step": 3879
+    },
+    {
+      "epoch": 2.1748878923766815,
+      "grad_norm": 0.07908419518265027,
+      "learning_rate": 4.2666107778884065e-05,
+      "loss": 0.2475,
+      "step": 3880
+    },
+    {
+      "epoch": 2.1754484304932737,
+      "grad_norm": 0.07714408912918791,
+      "learning_rate": 4.261267387742323e-05,
+      "loss": 0.2386,
+      "step": 3881
+    },
+    {
+      "epoch": 2.1760089686098656,
+      "grad_norm": 0.07845478885021406,
+      "learning_rate": 4.255926439579948e-05,
+      "loss": 0.2409,
+      "step": 3882
+    },
+    {
+      "epoch": 2.1765695067264574,
+      "grad_norm": 0.07785859360485685,
+      "learning_rate": 4.250587935674009e-05,
+      "loss": 0.2349,
+      "step": 3883
+    },
+    {
+      "epoch": 2.1771300448430493,
+      "grad_norm": 0.07751788254387426,
+      "learning_rate": 4.245251878296171e-05,
+      "loss": 0.2372,
+      "step": 3884
+    },
+    {
+      "epoch": 2.177690582959641,
+      "grad_norm": 0.07727111253762252,
+      "learning_rate": 4.2399182697170806e-05,
+      "loss": 0.2511,
+      "step": 3885
+    },
+    {
+      "epoch": 2.178251121076233,
+      "grad_norm": 0.0813134242476208,
+      "learning_rate": 4.234587112206317e-05,
+      "loss": 0.2483,
+      "step": 3886
+    },
+    {
+      "epoch": 2.1788116591928253,
+      "grad_norm": 0.08104585169424193,
+      "learning_rate": 4.2292584080324424e-05,
+      "loss": 0.2412,
+      "step": 3887
+    },
+    {
+      "epoch": 2.179372197309417,
+      "grad_norm": 0.07784345841521798,
+      "learning_rate": 4.223932159462954e-05,
+      "loss": 0.237,
+      "step": 3888
+    },
+    {
+      "epoch": 2.179932735426009,
+      "grad_norm": 0.07412396478730746,
+      "learning_rate": 4.218608368764314e-05,
+      "loss": 0.2407,
+      "step": 3889
+    },
+    {
+      "epoch": 2.180493273542601,
+      "grad_norm": 0.07413010645194912,
+      "learning_rate": 4.213287038201943e-05,
+      "loss": 0.2491,
+      "step": 3890
+    },
+    {
+      "epoch": 2.1810538116591927,
+      "grad_norm": 0.08052501546387542,
+      "learning_rate": 4.207968170040202e-05,
+      "loss": 0.235,
+      "step": 3891
+    },
+    {
+      "epoch": 2.181614349775785,
+      "grad_norm": 0.07749474283314743,
+      "learning_rate": 4.202651766542416e-05,
+      "loss": 0.2441,
+      "step": 3892
+    },
+    {
+      "epoch": 2.182174887892377,
+      "grad_norm": 0.07857814645642697,
+      "learning_rate": 4.197337829970852e-05,
+      "loss": 0.2527,
+      "step": 3893
+    },
+    {
+      "epoch": 2.1827354260089686,
+      "grad_norm": 0.07543219070462921,
+      "learning_rate": 4.1920263625867364e-05,
+      "loss": 0.2422,
+      "step": 3894
+    },
+    {
+      "epoch": 2.1832959641255605,
+      "grad_norm": 0.07617937414399167,
+      "learning_rate": 4.1867173666502393e-05,
+      "loss": 0.2334,
+      "step": 3895
+    },
+    {
+      "epoch": 2.1838565022421523,
+      "grad_norm": 0.07993840930599429,
+      "learning_rate": 4.181410844420474e-05,
+      "loss": 0.2561,
+      "step": 3896
+    },
+    {
+      "epoch": 2.1844170403587446,
+      "grad_norm": 0.07924758171821018,
+      "learning_rate": 4.1761067981555114e-05,
+      "loss": 0.247,
+      "step": 3897
+    },
+    {
+      "epoch": 2.1849775784753365,
+      "grad_norm": 0.07495089799120876,
+      "learning_rate": 4.170805230112366e-05,
+      "loss": 0.2383,
+      "step": 3898
+    },
+    {
+      "epoch": 2.1855381165919283,
+      "grad_norm": 0.07727828477574174,
+      "learning_rate": 4.1655061425469976e-05,
+      "loss": 0.2459,
+      "step": 3899
+    },
+    {
+      "epoch": 2.18609865470852,
+      "grad_norm": 0.07555327292191727,
+      "learning_rate": 4.160209537714304e-05,
+      "loss": 0.2469,
+      "step": 3900
+    },
+    {
+      "epoch": 2.186659192825112,
+      "grad_norm": 0.07592763275951135,
+      "learning_rate": 4.154915417868137e-05,
+      "loss": 0.2331,
+      "step": 3901
+    },
+    {
+      "epoch": 2.187219730941704,
+      "grad_norm": 0.07644350710027373,
+      "learning_rate": 4.149623785261284e-05,
+      "loss": 0.2364,
+      "step": 3902
+    },
+    {
+      "epoch": 2.187780269058296,
+      "grad_norm": 0.07419899351884267,
+      "learning_rate": 4.1443346421454724e-05,
+      "loss": 0.239,
+      "step": 3903
+    },
+    {
+      "epoch": 2.188340807174888,
+      "grad_norm": 0.07910748606280418,
+      "learning_rate": 4.139047990771378e-05,
+      "loss": 0.254,
+      "step": 3904
+    },
+    {
+      "epoch": 2.18890134529148,
+      "grad_norm": 0.07852864663680223,
+      "learning_rate": 4.133763833388609e-05,
+      "loss": 0.2353,
+      "step": 3905
+    },
+    {
+      "epoch": 2.1894618834080717,
+      "grad_norm": 0.07932087877745334,
+      "learning_rate": 4.128482172245715e-05,
+      "loss": 0.2437,
+      "step": 3906
+    },
+    {
+      "epoch": 2.1900224215246635,
+      "grad_norm": 0.0758056247353558,
+      "learning_rate": 4.123203009590185e-05,
+      "loss": 0.2443,
+      "step": 3907
+    },
+    {
+      "epoch": 2.1905829596412554,
+      "grad_norm": 0.07805058809414696,
+      "learning_rate": 4.1179263476684474e-05,
+      "loss": 0.2469,
+      "step": 3908
+    },
+    {
+      "epoch": 2.1911434977578477,
+      "grad_norm": 0.07674673233219874,
+      "learning_rate": 4.112652188725859e-05,
+      "loss": 0.2429,
+      "step": 3909
+    },
+    {
+      "epoch": 2.1917040358744395,
+      "grad_norm": 0.07350573510538515,
+      "learning_rate": 4.1073805350067096e-05,
+      "loss": 0.2293,
+      "step": 3910
+    },
+    {
+      "epoch": 2.1922645739910314,
+      "grad_norm": 0.07935859742013034,
+      "learning_rate": 4.102111388754238e-05,
+      "loss": 0.2517,
+      "step": 3911
+    },
+    {
+      "epoch": 2.192825112107623,
+      "grad_norm": 0.07596313105262871,
+      "learning_rate": 4.096844752210598e-05,
+      "loss": 0.2479,
+      "step": 3912
+    },
+    {
+      "epoch": 2.193385650224215,
+      "grad_norm": 0.07985349632362626,
+      "learning_rate": 4.091580627616888e-05,
+      "loss": 0.237,
+      "step": 3913
+    },
+    {
+      "epoch": 2.1939461883408073,
+      "grad_norm": 0.07661541764495726,
+      "learning_rate": 4.0863190172131364e-05,
+      "loss": 0.2542,
+      "step": 3914
+    },
+    {
+      "epoch": 2.194506726457399,
+      "grad_norm": 0.07762540071070838,
+      "learning_rate": 4.0810599232382916e-05,
+      "loss": 0.2454,
+      "step": 3915
+    },
+    {
+      "epoch": 2.195067264573991,
+      "grad_norm": 0.07787073005395762,
+      "learning_rate": 4.075803347930245e-05,
+      "loss": 0.2408,
+      "step": 3916
+    },
+    {
+      "epoch": 2.195627802690583,
+      "grad_norm": 0.07657366919537804,
+      "learning_rate": 4.070549293525804e-05,
+      "loss": 0.2447,
+      "step": 3917
+    },
+    {
+      "epoch": 2.1961883408071747,
+      "grad_norm": 0.07724671822423297,
+      "learning_rate": 4.0652977622607145e-05,
+      "loss": 0.2481,
+      "step": 3918
+    },
+    {
+      "epoch": 2.196748878923767,
+      "grad_norm": 0.07524027479414251,
+      "learning_rate": 4.0600487563696364e-05,
+      "loss": 0.2358,
+      "step": 3919
+    },
+    {
+      "epoch": 2.197309417040359,
+      "grad_norm": 0.077028529564304,
+      "learning_rate": 4.054802278086168e-05,
+      "loss": 0.2359,
+      "step": 3920
+    },
+    {
+      "epoch": 2.1978699551569507,
+      "grad_norm": 0.0785710538200137,
+      "learning_rate": 4.0495583296428205e-05,
+      "loss": 0.2427,
+      "step": 3921
+    },
+    {
+      "epoch": 2.1984304932735426,
+      "grad_norm": 0.07816972861165873,
+      "learning_rate": 4.044316913271036e-05,
+      "loss": 0.2414,
+      "step": 3922
+    },
+    {
+      "epoch": 2.1989910313901344,
+      "grad_norm": 0.07692316529475866,
+      "learning_rate": 4.03907803120118e-05,
+      "loss": 0.2464,
+      "step": 3923
+    },
+    {
+      "epoch": 2.1995515695067263,
+      "grad_norm": 0.07748236072229979,
+      "learning_rate": 4.0338416856625294e-05,
+      "loss": 0.2331,
+      "step": 3924
+    },
+    {
+      "epoch": 2.2001121076233185,
+      "grad_norm": 0.0769074810556553,
+      "learning_rate": 4.028607878883297e-05,
+      "loss": 0.2453,
+      "step": 3925
+    },
+    {
+      "epoch": 2.2006726457399104,
+      "grad_norm": 0.07879946810562749,
+      "learning_rate": 4.023376613090599e-05,
+      "loss": 0.2473,
+      "step": 3926
+    },
+    {
+      "epoch": 2.2012331838565022,
+      "grad_norm": 0.07608325617460276,
+      "learning_rate": 4.018147890510486e-05,
+      "loss": 0.2277,
+      "step": 3927
+    },
+    {
+      "epoch": 2.201793721973094,
+      "grad_norm": 0.0784418091050367,
+      "learning_rate": 4.012921713367916e-05,
+      "loss": 0.2517,
+      "step": 3928
+    },
+    {
+      "epoch": 2.202354260089686,
+      "grad_norm": 0.07802357173766879,
+      "learning_rate": 4.0076980838867625e-05,
+      "loss": 0.2444,
+      "step": 3929
+    },
+    {
+      "epoch": 2.2029147982062782,
+      "grad_norm": 0.08036596867295931,
+      "learning_rate": 4.0024770042898215e-05,
+      "loss": 0.2368,
+      "step": 3930
+    },
+    {
+      "epoch": 2.20347533632287,
+      "grad_norm": 0.07627521653760704,
+      "learning_rate": 3.997258476798804e-05,
+      "loss": 0.2478,
+      "step": 3931
+    },
+    {
+      "epoch": 2.204035874439462,
+      "grad_norm": 0.07637266150754995,
+      "learning_rate": 3.9920425036343344e-05,
+      "loss": 0.2396,
+      "step": 3932
+    },
+    {
+      "epoch": 2.2045964125560538,
+      "grad_norm": 0.0757979635557969,
+      "learning_rate": 3.9868290870159405e-05,
+      "loss": 0.2372,
+      "step": 3933
+    },
+    {
+      "epoch": 2.2051569506726456,
+      "grad_norm": 0.0758271305367863,
+      "learning_rate": 3.98161822916208e-05,
+      "loss": 0.2315,
+      "step": 3934
+    },
+    {
+      "epoch": 2.2057174887892375,
+      "grad_norm": 0.07840895001099792,
+      "learning_rate": 3.9764099322901047e-05,
+      "loss": 0.243,
+      "step": 3935
+    },
+    {
+      "epoch": 2.2062780269058297,
+      "grad_norm": 0.07792711393841163,
+      "learning_rate": 3.971204198616284e-05,
+      "loss": 0.2458,
+      "step": 3936
+    },
+    {
+      "epoch": 2.2068385650224216,
+      "grad_norm": 0.07779689081296363,
+      "learning_rate": 3.9660010303558005e-05,
+      "loss": 0.256,
+      "step": 3937
+    },
+    {
+      "epoch": 2.2073991031390134,
+      "grad_norm": 0.07770239643017164,
+      "learning_rate": 3.960800429722734e-05,
+      "loss": 0.2391,
+      "step": 3938
+    },
+    {
+      "epoch": 2.2079596412556053,
+      "grad_norm": 0.07759603142305241,
+      "learning_rate": 3.955602398930084e-05,
+      "loss": 0.2393,
+      "step": 3939
+    },
+    {
+      "epoch": 2.208520179372197,
+      "grad_norm": 0.07864454186406615,
+      "learning_rate": 3.9504069401897505e-05,
+      "loss": 0.2356,
+      "step": 3940
+    },
+    {
+      "epoch": 2.2090807174887894,
+      "grad_norm": 0.07917369685068788,
+      "learning_rate": 3.9452140557125435e-05,
+      "loss": 0.243,
+      "step": 3941
+    },
+    {
+      "epoch": 2.2096412556053813,
+      "grad_norm": 0.0769382735363664,
+      "learning_rate": 3.940023747708169e-05,
+      "loss": 0.2377,
+      "step": 3942
+    },
+    {
+      "epoch": 2.210201793721973,
+      "grad_norm": 0.07825002371469891,
+      "learning_rate": 3.934836018385239e-05,
+      "loss": 0.2526,
+      "step": 3943
+    },
+    {
+      "epoch": 2.210762331838565,
+      "grad_norm": 0.07734985454503165,
+      "learning_rate": 3.929650869951278e-05,
+      "loss": 0.2437,
+      "step": 3944
+    },
+    {
+      "epoch": 2.211322869955157,
+      "grad_norm": 0.07900761978227461,
+      "learning_rate": 3.924468304612696e-05,
+      "loss": 0.2369,
+      "step": 3945
+    },
+    {
+      "epoch": 2.211883408071749,
+      "grad_norm": 0.07796886832267623,
+      "learning_rate": 3.9192883245748194e-05,
+      "loss": 0.2529,
+      "step": 3946
+    },
+    {
+      "epoch": 2.212443946188341,
+      "grad_norm": 0.07774089387545376,
+      "learning_rate": 3.914110932041865e-05,
+      "loss": 0.2356,
+      "step": 3947
+    },
+    {
+      "epoch": 2.213004484304933,
+      "grad_norm": 0.07669595893560695,
+      "learning_rate": 3.908936129216955e-05,
+      "loss": 0.2441,
+      "step": 3948
+    },
+    {
+      "epoch": 2.2135650224215246,
+      "grad_norm": 0.07694441354189155,
+      "learning_rate": 3.903763918302104e-05,
+      "loss": 0.2418,
+      "step": 3949
+    },
+    {
+      "epoch": 2.2141255605381165,
+      "grad_norm": 0.0784777046425361,
+      "learning_rate": 3.898594301498221e-05,
+      "loss": 0.248,
+      "step": 3950
+    },
+    {
+      "epoch": 2.2146860986547083,
+      "grad_norm": 0.07805312335295722,
+      "learning_rate": 3.893427281005122e-05,
+      "loss": 0.2381,
+      "step": 3951
+    },
+    {
+      "epoch": 2.2152466367713006,
+      "grad_norm": 0.08146147369591868,
+      "learning_rate": 3.8882628590215074e-05,
+      "loss": 0.2508,
+      "step": 3952
+    },
+    {
+      "epoch": 2.2158071748878925,
+      "grad_norm": 0.07740833226057425,
+      "learning_rate": 3.8831010377449816e-05,
+      "loss": 0.2498,
+      "step": 3953
+    },
+    {
+      "epoch": 2.2163677130044843,
+      "grad_norm": 0.07565218389561658,
+      "learning_rate": 3.877941819372031e-05,
+      "loss": 0.2353,
+      "step": 3954
+    },
+    {
+      "epoch": 2.216928251121076,
+      "grad_norm": 0.0787770429773584,
+      "learning_rate": 3.8727852060980444e-05,
+      "loss": 0.2499,
+      "step": 3955
+    },
+    {
+      "epoch": 2.217488789237668,
+      "grad_norm": 0.0799115347001249,
+      "learning_rate": 3.8676312001173e-05,
+      "loss": 0.2496,
+      "step": 3956
+    },
+    {
+      "epoch": 2.21804932735426,
+      "grad_norm": 0.07721801439002256,
+      "learning_rate": 3.862479803622958e-05,
+      "loss": 0.2416,
+      "step": 3957
+    },
+    {
+      "epoch": 2.218609865470852,
+      "grad_norm": 0.07818729106979565,
+      "learning_rate": 3.8573310188070845e-05,
+      "loss": 0.2546,
+      "step": 3958
+    },
+    {
+      "epoch": 2.219170403587444,
+      "grad_norm": 0.07966364760612964,
+      "learning_rate": 3.852184847860615e-05,
+      "loss": 0.2352,
+      "step": 3959
+    },
+    {
+      "epoch": 2.219730941704036,
+      "grad_norm": 0.0773622586431013,
+      "learning_rate": 3.84704129297339e-05,
+      "loss": 0.2435,
+      "step": 3960
+    },
+    {
+      "epoch": 2.2202914798206277,
+      "grad_norm": 0.07706245453058086,
+      "learning_rate": 3.841900356334127e-05,
+      "loss": 0.2413,
+      "step": 3961
+    },
+    {
+      "epoch": 2.2208520179372195,
+      "grad_norm": 0.07925653446803047,
+      "learning_rate": 3.836762040130426e-05,
+      "loss": 0.2443,
+      "step": 3962
+    },
+    {
+      "epoch": 2.221412556053812,
+      "grad_norm": 0.07597662986644867,
+      "learning_rate": 3.8316263465487834e-05,
+      "loss": 0.248,
+      "step": 3963
+    },
+    {
+      "epoch": 2.2219730941704037,
+      "grad_norm": 0.07965193025099362,
+      "learning_rate": 3.826493277774572e-05,
+      "loss": 0.2519,
+      "step": 3964
+    },
+    {
+      "epoch": 2.2225336322869955,
+      "grad_norm": 0.07974783423100505,
+      "learning_rate": 3.821362835992053e-05,
+      "loss": 0.2491,
+      "step": 3965
+    },
+    {
+      "epoch": 2.2230941704035874,
+      "grad_norm": 0.07617788752007221,
+      "learning_rate": 3.81623502338436e-05,
+      "loss": 0.2458,
+      "step": 3966
+    },
+    {
+      "epoch": 2.223654708520179,
+      "grad_norm": 0.07716877427571417,
+      "learning_rate": 3.81110984213352e-05,
+      "loss": 0.2365,
+      "step": 3967
+    },
+    {
+      "epoch": 2.2242152466367715,
+      "grad_norm": 0.07925341243024847,
+      "learning_rate": 3.8059872944204324e-05,
+      "loss": 0.2443,
+      "step": 3968
+    },
+    {
+      "epoch": 2.2247757847533634,
+      "grad_norm": 0.07883561549575804,
+      "learning_rate": 3.800867382424872e-05,
+      "loss": 0.2469,
+      "step": 3969
+    },
+    {
+      "epoch": 2.225336322869955,
+      "grad_norm": 0.07655529733887344,
+      "learning_rate": 3.7957501083255065e-05,
+      "loss": 0.2546,
+      "step": 3970
+    },
+    {
+      "epoch": 2.225896860986547,
+      "grad_norm": 0.07559675606024821,
+      "learning_rate": 3.7906354742998654e-05,
+      "loss": 0.2391,
+      "step": 3971
+    },
+    {
+      "epoch": 2.226457399103139,
+      "grad_norm": 0.0784403741646514,
+      "learning_rate": 3.7855234825243644e-05,
+      "loss": 0.2564,
+      "step": 3972
+    },
+    {
+      "epoch": 2.2270179372197307,
+      "grad_norm": 0.07677942678759365,
+      "learning_rate": 3.7804141351742925e-05,
+      "loss": 0.2369,
+      "step": 3973
+    },
+    {
+      "epoch": 2.227578475336323,
+      "grad_norm": 0.07786561126409022,
+      "learning_rate": 3.775307434423818e-05,
+      "loss": 0.2488,
+      "step": 3974
+    },
+    {
+      "epoch": 2.228139013452915,
+      "grad_norm": 0.07733142719653782,
+      "learning_rate": 3.770203382445974e-05,
+      "loss": 0.2412,
+      "step": 3975
+    },
+    {
+      "epoch": 2.2286995515695067,
+      "grad_norm": 0.07510479482687211,
+      "learning_rate": 3.7651019814126654e-05,
+      "loss": 0.2404,
+      "step": 3976
+    },
+    {
+      "epoch": 2.2292600896860986,
+      "grad_norm": 0.0793733886983191,
+      "learning_rate": 3.760003233494683e-05,
+      "loss": 0.2479,
+      "step": 3977
+    },
+    {
+      "epoch": 2.2298206278026904,
+      "grad_norm": 0.07943863183490149,
+      "learning_rate": 3.754907140861674e-05,
+      "loss": 0.2377,
+      "step": 3978
+    },
+    {
+      "epoch": 2.2303811659192827,
+      "grad_norm": 0.07476932550523543,
+      "learning_rate": 3.7498137056821634e-05,
+      "loss": 0.2298,
+      "step": 3979
+    },
+    {
+      "epoch": 2.2309417040358746,
+      "grad_norm": 0.07930513376330914,
+      "learning_rate": 3.7447229301235445e-05,
+      "loss": 0.2424,
+      "step": 3980
+    },
+    {
+      "epoch": 2.2315022421524664,
+      "grad_norm": 0.07769688844282986,
+      "learning_rate": 3.739634816352081e-05,
+      "loss": 0.2449,
+      "step": 3981
+    },
+    {
+      "epoch": 2.2320627802690582,
+      "grad_norm": 0.07778542276378456,
+      "learning_rate": 3.734549366532898e-05,
+      "loss": 0.2431,
+      "step": 3982
+    },
+    {
+      "epoch": 2.23262331838565,
+      "grad_norm": 0.07605549064102234,
+      "learning_rate": 3.7294665828299856e-05,
+      "loss": 0.2453,
+      "step": 3983
+    },
+    {
+      "epoch": 2.233183856502242,
+      "grad_norm": 0.07560550393605212,
+      "learning_rate": 3.724386467406211e-05,
+      "loss": 0.236,
+      "step": 3984
+    },
+    {
+      "epoch": 2.2337443946188342,
+      "grad_norm": 0.0778699111104112,
+      "learning_rate": 3.719309022423293e-05,
+      "loss": 0.2453,
+      "step": 3985
+    },
+    {
+      "epoch": 2.234304932735426,
+      "grad_norm": 0.07691655621653959,
+      "learning_rate": 3.7142342500418256e-05,
+      "loss": 0.245,
+      "step": 3986
+    },
+    {
+      "epoch": 2.234865470852018,
+      "grad_norm": 0.07839221092373565,
+      "learning_rate": 3.709162152421253e-05,
+      "loss": 0.2417,
+      "step": 3987
+    },
+    {
+      "epoch": 2.2354260089686098,
+      "grad_norm": 0.07837476833048565,
+      "learning_rate": 3.704092731719892e-05,
+      "loss": 0.2396,
+      "step": 3988
+    },
+    {
+      "epoch": 2.2359865470852016,
+      "grad_norm": 0.07892332509063256,
+      "learning_rate": 3.699025990094919e-05,
+      "loss": 0.236,
+      "step": 3989
+    },
+    {
+      "epoch": 2.236547085201794,
+      "grad_norm": 0.07867162636838011,
+      "learning_rate": 3.6939619297023595e-05,
+      "loss": 0.2428,
+      "step": 3990
+    },
+    {
+      "epoch": 2.2371076233183858,
+      "grad_norm": 0.07895061423186442,
+      "learning_rate": 3.688900552697115e-05,
+      "loss": 0.2637,
+      "step": 3991
+    },
+    {
+      "epoch": 2.2376681614349776,
+      "grad_norm": 0.07699857519011867,
+      "learning_rate": 3.6838418612329305e-05,
+      "loss": 0.2445,
+      "step": 3992
+    },
+    {
+      "epoch": 2.2382286995515694,
+      "grad_norm": 0.07663795542669885,
+      "learning_rate": 3.6787858574624176e-05,
+      "loss": 0.2426,
+      "step": 3993
+    },
+    {
+      "epoch": 2.2387892376681613,
+      "grad_norm": 0.07848411722853481,
+      "learning_rate": 3.6737325435370374e-05,
+      "loss": 0.2631,
+      "step": 3994
+    },
+    {
+      "epoch": 2.2393497757847536,
+      "grad_norm": 0.07971443492816431,
+      "learning_rate": 3.668681921607113e-05,
+      "loss": 0.2494,
+      "step": 3995
+    },
+    {
+      "epoch": 2.2399103139013454,
+      "grad_norm": 0.07852676713453574,
+      "learning_rate": 3.663633993821816e-05,
+      "loss": 0.2538,
+      "step": 3996
+    },
+    {
+      "epoch": 2.2404708520179373,
+      "grad_norm": 0.07840943264658039,
+      "learning_rate": 3.658588762329174e-05,
+      "loss": 0.2365,
+      "step": 3997
+    },
+    {
+      "epoch": 2.241031390134529,
+      "grad_norm": 0.07743865110849929,
+      "learning_rate": 3.6535462292760715e-05,
+      "loss": 0.2517,
+      "step": 3998
+    },
+    {
+      "epoch": 2.241591928251121,
+      "grad_norm": 0.07988589540027931,
+      "learning_rate": 3.6485063968082344e-05,
+      "loss": 0.2498,
+      "step": 3999
+    },
+    {
+      "epoch": 2.242152466367713,
+      "grad_norm": 0.07613715602980295,
+      "learning_rate": 3.6434692670702545e-05,
+      "loss": 0.2418,
+      "step": 4000
+    },
+    {
+      "epoch": 2.242713004484305,
+      "grad_norm": 0.07606313247057905,
+      "learning_rate": 3.638434842205558e-05,
+      "loss": 0.2306,
+      "step": 4001
+    },
+    {
+      "epoch": 2.243273542600897,
+      "grad_norm": 0.07720011259500538,
+      "learning_rate": 3.633403124356426e-05,
+      "loss": 0.2387,
+      "step": 4002
+    },
+    {
+      "epoch": 2.243834080717489,
+      "grad_norm": 0.07846912534834977,
+      "learning_rate": 3.628374115663995e-05,
+      "loss": 0.2506,
+      "step": 4003
+    },
+    {
+      "epoch": 2.2443946188340806,
+      "grad_norm": 0.07776662326654782,
+      "learning_rate": 3.6233478182682345e-05,
+      "loss": 0.2345,
+      "step": 4004
+    },
+    {
+      "epoch": 2.2449551569506725,
+      "grad_norm": 0.07540971408911927,
+      "learning_rate": 3.618324234307973e-05,
+      "loss": 0.2373,
+      "step": 4005
+    },
+    {
+      "epoch": 2.2455156950672643,
+      "grad_norm": 0.077880691647363,
+      "learning_rate": 3.613303365920877e-05,
+      "loss": 0.2412,
+      "step": 4006
+    },
+    {
+      "epoch": 2.2460762331838566,
+      "grad_norm": 0.08161547694039394,
+      "learning_rate": 3.6082852152434646e-05,
+      "loss": 0.2541,
+      "step": 4007
+    },
+    {
+      "epoch": 2.2466367713004485,
+      "grad_norm": 0.07710773873508162,
+      "learning_rate": 3.60326978441109e-05,
+      "loss": 0.2378,
+      "step": 4008
+    },
+    {
+      "epoch": 2.2471973094170403,
+      "grad_norm": 0.07698233339569654,
+      "learning_rate": 3.598257075557948e-05,
+      "loss": 0.2324,
+      "step": 4009
+    },
+    {
+      "epoch": 2.247757847533632,
+      "grad_norm": 0.07960604665158132,
+      "learning_rate": 3.593247090817088e-05,
+      "loss": 0.2459,
+      "step": 4010
+    },
+    {
+      "epoch": 2.248318385650224,
+      "grad_norm": 0.07303690685007934,
+      "learning_rate": 3.5882398323203834e-05,
+      "loss": 0.2406,
+      "step": 4011
+    },
+    {
+      "epoch": 2.2488789237668163,
+      "grad_norm": 0.07989948416525376,
+      "learning_rate": 3.583235302198562e-05,
+      "loss": 0.2444,
+      "step": 4012
+    },
+    {
+      "epoch": 2.249439461883408,
+      "grad_norm": 0.0768515749305458,
+      "learning_rate": 3.578233502581183e-05,
+      "loss": 0.2423,
+      "step": 4013
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.07962415092933411,
+      "learning_rate": 3.5732344355966494e-05,
+      "loss": 0.2408,
+      "step": 4014
+    },
+    {
+      "epoch": 2.250560538116592,
+      "grad_norm": 0.07901083134605137,
+      "learning_rate": 3.5682381033721944e-05,
+      "loss": 0.2372,
+      "step": 4015
+    },
+    {
+      "epoch": 2.2511210762331837,
+      "grad_norm": 0.08134667818885108,
+      "learning_rate": 3.563244508033887e-05,
+      "loss": 0.2536,
+      "step": 4016
+    },
+    {
+      "epoch": 2.251681614349776,
+      "grad_norm": 0.07714038989261836,
+      "learning_rate": 3.558253651706641e-05,
+      "loss": 0.2412,
+      "step": 4017
+    },
+    {
+      "epoch": 2.252242152466368,
+      "grad_norm": 0.07646345861859105,
+      "learning_rate": 3.5532655365141934e-05,
+      "loss": 0.2328,
+      "step": 4018
+    },
+    {
+      "epoch": 2.2528026905829597,
+      "grad_norm": 0.07513654052258976,
+      "learning_rate": 3.548280164579126e-05,
+      "loss": 0.2421,
+      "step": 4019
+    },
+    {
+      "epoch": 2.2533632286995515,
+      "grad_norm": 0.0767600702398758,
+      "learning_rate": 3.543297538022842e-05,
+      "loss": 0.2507,
+      "step": 4020
+    },
+    {
+      "epoch": 2.2539237668161434,
+      "grad_norm": 0.07794866454848276,
+      "learning_rate": 3.538317658965583e-05,
+      "loss": 0.2428,
+      "step": 4021
+    },
+    {
+      "epoch": 2.2544843049327357,
+      "grad_norm": 0.07751718657953899,
+      "learning_rate": 3.533340529526426e-05,
+      "loss": 0.2464,
+      "step": 4022
+    },
+    {
+      "epoch": 2.2550448430493275,
+      "grad_norm": 0.07890045797164695,
+      "learning_rate": 3.5283661518232635e-05,
+      "loss": 0.2464,
+      "step": 4023
+    },
+    {
+      "epoch": 2.2556053811659194,
+      "grad_norm": 0.07771274932781744,
+      "learning_rate": 3.523394527972833e-05,
+      "loss": 0.2446,
+      "step": 4024
+    },
+    {
+      "epoch": 2.256165919282511,
+      "grad_norm": 0.07799172469865923,
+      "learning_rate": 3.5184256600906885e-05,
+      "loss": 0.2492,
+      "step": 4025
+    },
+    {
+      "epoch": 2.256726457399103,
+      "grad_norm": 0.07770715943505946,
+      "learning_rate": 3.513459550291219e-05,
+      "loss": 0.2519,
+      "step": 4026
+    },
+    {
+      "epoch": 2.257286995515695,
+      "grad_norm": 0.08006532274437009,
+      "learning_rate": 3.508496200687633e-05,
+      "loss": 0.2526,
+      "step": 4027
+    },
+    {
+      "epoch": 2.257847533632287,
+      "grad_norm": 0.07499003121362369,
+      "learning_rate": 3.503535613391973e-05,
+      "loss": 0.2264,
+      "step": 4028
+    },
+    {
+      "epoch": 2.258408071748879,
+      "grad_norm": 0.07983255025461201,
+      "learning_rate": 3.498577790515095e-05,
+      "loss": 0.2325,
+      "step": 4029
+    },
+    {
+      "epoch": 2.258968609865471,
+      "grad_norm": 0.08010545958973918,
+      "learning_rate": 3.493622734166688e-05,
+      "loss": 0.2532,
+      "step": 4030
+    },
+    {
+      "epoch": 2.2595291479820627,
+      "grad_norm": 0.076604938351707,
+      "learning_rate": 3.4886704464552635e-05,
+      "loss": 0.2427,
+      "step": 4031
+    },
+    {
+      "epoch": 2.2600896860986546,
+      "grad_norm": 0.07817015284831182,
+      "learning_rate": 3.4837209294881467e-05,
+      "loss": 0.2406,
+      "step": 4032
+    },
+    {
+      "epoch": 2.2606502242152464,
+      "grad_norm": 0.07822138725494229,
+      "learning_rate": 3.478774185371494e-05,
+      "loss": 0.2405,
+      "step": 4033
+    },
+    {
+      "epoch": 2.2612107623318387,
+      "grad_norm": 0.07386648586034597,
+      "learning_rate": 3.473830216210271e-05,
+      "loss": 0.2356,
+      "step": 4034
+    },
+    {
+      "epoch": 2.2617713004484306,
+      "grad_norm": 0.07914046299505863,
+      "learning_rate": 3.468889024108275e-05,
+      "loss": 0.2425,
+      "step": 4035
+    },
+    {
+      "epoch": 2.2623318385650224,
+      "grad_norm": 0.07792349004039284,
+      "learning_rate": 3.463950611168111e-05,
+      "loss": 0.2333,
+      "step": 4036
+    },
+    {
+      "epoch": 2.2628923766816142,
+      "grad_norm": 0.08293485491792615,
+      "learning_rate": 3.459014979491203e-05,
+      "loss": 0.2426,
+      "step": 4037
+    },
+    {
+      "epoch": 2.263452914798206,
+      "grad_norm": 0.07648159565150504,
+      "learning_rate": 3.454082131177797e-05,
+      "loss": 0.2391,
+      "step": 4038
+    },
+    {
+      "epoch": 2.2640134529147984,
+      "grad_norm": 0.07604430237799964,
+      "learning_rate": 3.449152068326951e-05,
+      "loss": 0.2347,
+      "step": 4039
+    },
+    {
+      "epoch": 2.2645739910313902,
+      "grad_norm": 0.08036645521485034,
+      "learning_rate": 3.4442247930365426e-05,
+      "loss": 0.2496,
+      "step": 4040
+    },
+    {
+      "epoch": 2.265134529147982,
+      "grad_norm": 0.07581613443950523,
+      "learning_rate": 3.439300307403254e-05,
+      "loss": 0.2427,
+      "step": 4041
+    },
+    {
+      "epoch": 2.265695067264574,
+      "grad_norm": 0.0795612256663696,
+      "learning_rate": 3.434378613522582e-05,
+      "loss": 0.2444,
+      "step": 4042
+    },
+    {
+      "epoch": 2.2662556053811658,
+      "grad_norm": 0.07958407744060352,
+      "learning_rate": 3.429459713488846e-05,
+      "loss": 0.2491,
+      "step": 4043
+    },
+    {
+      "epoch": 2.266816143497758,
+      "grad_norm": 0.08036965217293619,
+      "learning_rate": 3.424543609395162e-05,
+      "loss": 0.2401,
+      "step": 4044
+    },
+    {
+      "epoch": 2.26737668161435,
+      "grad_norm": 0.07644904034051768,
+      "learning_rate": 3.419630303333466e-05,
+      "loss": 0.2329,
+      "step": 4045
+    },
+    {
+      "epoch": 2.2679372197309418,
+      "grad_norm": 0.07495170930821564,
+      "learning_rate": 3.4147197973945035e-05,
+      "loss": 0.2373,
+      "step": 4046
+    },
+    {
+      "epoch": 2.2684977578475336,
+      "grad_norm": 0.07948384758363594,
+      "learning_rate": 3.409812093667826e-05,
+      "loss": 0.2321,
+      "step": 4047
+    },
+    {
+      "epoch": 2.2690582959641254,
+      "grad_norm": 0.08064903002594626,
+      "learning_rate": 3.4049071942417896e-05,
+      "loss": 0.2487,
+      "step": 4048
+    },
+    {
+      "epoch": 2.2696188340807173,
+      "grad_norm": 0.07667457565003931,
+      "learning_rate": 3.400005101203557e-05,
+      "loss": 0.2445,
+      "step": 4049
+    },
+    {
+      "epoch": 2.2701793721973096,
+      "grad_norm": 0.08096075945625165,
+      "learning_rate": 3.395105816639106e-05,
+      "loss": 0.2482,
+      "step": 4050
+    },
+    {
+      "epoch": 2.2707399103139014,
+      "grad_norm": 0.0802842618493245,
+      "learning_rate": 3.390209342633205e-05,
+      "loss": 0.2364,
+      "step": 4051
+    },
+    {
+      "epoch": 2.2713004484304933,
+      "grad_norm": 0.07799608919902139,
+      "learning_rate": 3.385315681269443e-05,
+      "loss": 0.2484,
+      "step": 4052
+    },
+    {
+      "epoch": 2.271860986547085,
+      "grad_norm": 0.07986379349658052,
+      "learning_rate": 3.380424834630196e-05,
+      "loss": 0.234,
+      "step": 4053
+    },
+    {
+      "epoch": 2.272421524663677,
+      "grad_norm": 0.07868640209787364,
+      "learning_rate": 3.375536804796652e-05,
+      "loss": 0.2475,
+      "step": 4054
+    },
+    {
+      "epoch": 2.272982062780269,
+      "grad_norm": 0.07977117505183345,
+      "learning_rate": 3.370651593848802e-05,
+      "loss": 0.2426,
+      "step": 4055
+    },
+    {
+      "epoch": 2.273542600896861,
+      "grad_norm": 0.07978111905869915,
+      "learning_rate": 3.365769203865425e-05,
+      "loss": 0.2493,
+      "step": 4056
+    },
+    {
+      "epoch": 2.274103139013453,
+      "grad_norm": 0.07815652824547649,
+      "learning_rate": 3.3608896369241196e-05,
+      "loss": 0.2525,
+      "step": 4057
+    },
+    {
+      "epoch": 2.274663677130045,
+      "grad_norm": 0.07897625992101927,
+      "learning_rate": 3.356012895101259e-05,
+      "loss": 0.2411,
+      "step": 4058
+    },
+    {
+      "epoch": 2.2752242152466366,
+      "grad_norm": 0.07796984568897937,
+      "learning_rate": 3.351138980472038e-05,
+      "loss": 0.2469,
+      "step": 4059
+    },
+    {
+      "epoch": 2.2757847533632285,
+      "grad_norm": 0.07554704922149767,
+      "learning_rate": 3.346267895110429e-05,
+      "loss": 0.2318,
+      "step": 4060
+    },
+    {
+      "epoch": 2.276345291479821,
+      "grad_norm": 0.07803333492071318,
+      "learning_rate": 3.3413996410892143e-05,
+      "loss": 0.255,
+      "step": 4061
+    },
+    {
+      "epoch": 2.2769058295964126,
+      "grad_norm": 0.07812744126921167,
+      "learning_rate": 3.336534220479961e-05,
+      "loss": 0.2368,
+      "step": 4062
+    },
+    {
+      "epoch": 2.2774663677130045,
+      "grad_norm": 0.08184618250853364,
+      "learning_rate": 3.331671635353037e-05,
+      "loss": 0.2407,
+      "step": 4063
+    },
+    {
+      "epoch": 2.2780269058295963,
+      "grad_norm": 0.07650302276882859,
+      "learning_rate": 3.3268118877776066e-05,
+      "loss": 0.24,
+      "step": 4064
+    },
+    {
+      "epoch": 2.278587443946188,
+      "grad_norm": 0.07648967362677596,
+      "learning_rate": 3.3219549798216145e-05,
+      "loss": 0.2434,
+      "step": 4065
+    },
+    {
+      "epoch": 2.2791479820627805,
+      "grad_norm": 0.07454490616444838,
+      "learning_rate": 3.317100913551812e-05,
+      "loss": 0.2352,
+      "step": 4066
+    },
+    {
+      "epoch": 2.2797085201793723,
+      "grad_norm": 0.07706418209665264,
+      "learning_rate": 3.3122496910337245e-05,
+      "loss": 0.2398,
+      "step": 4067
+    },
+    {
+      "epoch": 2.280269058295964,
+      "grad_norm": 0.07769346623390297,
+      "learning_rate": 3.307401314331686e-05,
+      "loss": 0.2463,
+      "step": 4068
+    },
+    {
+      "epoch": 2.280829596412556,
+      "grad_norm": 0.07761138116810193,
+      "learning_rate": 3.302555785508802e-05,
+      "loss": 0.2474,
+      "step": 4069
+    },
+    {
+      "epoch": 2.281390134529148,
+      "grad_norm": 0.07802716154311114,
+      "learning_rate": 3.297713106626978e-05,
+      "loss": 0.2421,
+      "step": 4070
+    },
+    {
+      "epoch": 2.28195067264574,
+      "grad_norm": 0.07665101677824855,
+      "learning_rate": 3.292873279746906e-05,
+      "loss": 0.2353,
+      "step": 4071
+    },
+    {
+      "epoch": 2.282511210762332,
+      "grad_norm": 0.08033235546011366,
+      "learning_rate": 3.288036306928055e-05,
+      "loss": 0.2482,
+      "step": 4072
+    },
+    {
+      "epoch": 2.283071748878924,
+      "grad_norm": 0.07805004891580285,
+      "learning_rate": 3.283202190228692e-05,
+      "loss": 0.2437,
+      "step": 4073
+    },
+    {
+      "epoch": 2.2836322869955157,
+      "grad_norm": 0.07677009588242084,
+      "learning_rate": 3.2783709317058575e-05,
+      "loss": 0.2425,
+      "step": 4074
+    },
+    {
+      "epoch": 2.2841928251121075,
+      "grad_norm": 0.07843372585241097,
+      "learning_rate": 3.273542533415386e-05,
+      "loss": 0.2429,
+      "step": 4075
+    },
+    {
+      "epoch": 2.2847533632286994,
+      "grad_norm": 0.0780651584816468,
+      "learning_rate": 3.268716997411886e-05,
+      "loss": 0.2466,
+      "step": 4076
+    },
+    {
+      "epoch": 2.2853139013452917,
+      "grad_norm": 0.07784304696239416,
+      "learning_rate": 3.26389432574875e-05,
+      "loss": 0.2421,
+      "step": 4077
+    },
+    {
+      "epoch": 2.2858744394618835,
+      "grad_norm": 0.07751781105207137,
+      "learning_rate": 3.2590745204781534e-05,
+      "loss": 0.244,
+      "step": 4078
+    },
+    {
+      "epoch": 2.2864349775784754,
+      "grad_norm": 0.07810555174866082,
+      "learning_rate": 3.2542575836510556e-05,
+      "loss": 0.2481,
+      "step": 4079
+    },
+    {
+      "epoch": 2.286995515695067,
+      "grad_norm": 0.07740809999962645,
+      "learning_rate": 3.249443517317194e-05,
+      "loss": 0.246,
+      "step": 4080
+    },
+    {
+      "epoch": 2.287556053811659,
+      "grad_norm": 0.07497482770784392,
+      "learning_rate": 3.244632323525074e-05,
+      "loss": 0.2355,
+      "step": 4081
+    },
+    {
+      "epoch": 2.288116591928251,
+      "grad_norm": 0.07732100446605851,
+      "learning_rate": 3.239824004321995e-05,
+      "loss": 0.2454,
+      "step": 4082
+    },
+    {
+      "epoch": 2.288677130044843,
+      "grad_norm": 0.07840414159510177,
+      "learning_rate": 3.235018561754022e-05,
+      "loss": 0.2391,
+      "step": 4083
+    },
+    {
+      "epoch": 2.289237668161435,
+      "grad_norm": 0.07943565519059993,
+      "learning_rate": 3.230215997865993e-05,
+      "loss": 0.2434,
+      "step": 4084
+    },
+    {
+      "epoch": 2.289798206278027,
+      "grad_norm": 0.0799020551470384,
+      "learning_rate": 3.225416314701537e-05,
+      "loss": 0.2497,
+      "step": 4085
+    },
+    {
+      "epoch": 2.2903587443946187,
+      "grad_norm": 0.07862871878240238,
+      "learning_rate": 3.220619514303037e-05,
+      "loss": 0.2257,
+      "step": 4086
+    },
+    {
+      "epoch": 2.2909192825112106,
+      "grad_norm": 0.07736652159487113,
+      "learning_rate": 3.2158255987116656e-05,
+      "loss": 0.2295,
+      "step": 4087
+    },
+    {
+      "epoch": 2.291479820627803,
+      "grad_norm": 0.08090013758807421,
+      "learning_rate": 3.211034569967365e-05,
+      "loss": 0.2512,
+      "step": 4088
+    },
+    {
+      "epoch": 2.2920403587443947,
+      "grad_norm": 0.07723418842114685,
+      "learning_rate": 3.2062464301088366e-05,
+      "loss": 0.2491,
+      "step": 4089
+    },
+    {
+      "epoch": 2.2926008968609866,
+      "grad_norm": 0.07906858623019813,
+      "learning_rate": 3.2014611811735695e-05,
+      "loss": 0.2602,
+      "step": 4090
+    },
+    {
+      "epoch": 2.2931614349775784,
+      "grad_norm": 0.07912332090476885,
+      "learning_rate": 3.19667882519781e-05,
+      "loss": 0.2425,
+      "step": 4091
+    },
+    {
+      "epoch": 2.2937219730941703,
+      "grad_norm": 0.07859642615535287,
+      "learning_rate": 3.191899364216581e-05,
+      "loss": 0.2433,
+      "step": 4092
+    },
+    {
+      "epoch": 2.2942825112107625,
+      "grad_norm": 0.0796324284676464,
+      "learning_rate": 3.187122800263667e-05,
+      "loss": 0.2522,
+      "step": 4093
+    },
+    {
+      "epoch": 2.2948430493273544,
+      "grad_norm": 0.0744524204632941,
+      "learning_rate": 3.182349135371627e-05,
+      "loss": 0.2306,
+      "step": 4094
+    },
+    {
+      "epoch": 2.2954035874439462,
+      "grad_norm": 0.07777123457265987,
+      "learning_rate": 3.17757837157178e-05,
+      "loss": 0.2513,
+      "step": 4095
+    },
+    {
+      "epoch": 2.295964125560538,
+      "grad_norm": 0.07524779879726812,
+      "learning_rate": 3.172810510894213e-05,
+      "loss": 0.243,
+      "step": 4096
+    },
+    {
+      "epoch": 2.29652466367713,
+      "grad_norm": 0.08221702157434065,
+      "learning_rate": 3.1680455553677824e-05,
+      "loss": 0.2523,
+      "step": 4097
+    },
+    {
+      "epoch": 2.297085201793722,
+      "grad_norm": 0.07822118201964864,
+      "learning_rate": 3.1632835070200975e-05,
+      "loss": 0.2396,
+      "step": 4098
+    },
+    {
+      "epoch": 2.297645739910314,
+      "grad_norm": 0.07892494483631408,
+      "learning_rate": 3.158524367877543e-05,
+      "loss": 0.2366,
+      "step": 4099
+    },
+    {
+      "epoch": 2.298206278026906,
+      "grad_norm": 0.07793237181027066,
+      "learning_rate": 3.153768139965253e-05,
+      "loss": 0.2343,
+      "step": 4100
+    },
+    {
+      "epoch": 2.2987668161434978,
+      "grad_norm": 0.07775126927184879,
+      "learning_rate": 3.1490148253071364e-05,
+      "loss": 0.2347,
+      "step": 4101
+    },
+    {
+      "epoch": 2.2993273542600896,
+      "grad_norm": 0.0810981916746797,
+      "learning_rate": 3.144264425925847e-05,
+      "loss": 0.2508,
+      "step": 4102
+    },
+    {
+      "epoch": 2.2998878923766815,
+      "grad_norm": 0.0775882006991532,
+      "learning_rate": 3.139516943842812e-05,
+      "loss": 0.2441,
+      "step": 4103
+    },
+    {
+      "epoch": 2.3004484304932733,
+      "grad_norm": 0.08040060914547632,
+      "learning_rate": 3.1347723810782134e-05,
+      "loss": 0.2564,
+      "step": 4104
+    },
+    {
+      "epoch": 2.3010089686098656,
+      "grad_norm": 0.07888930461738493,
+      "learning_rate": 3.130030739650983e-05,
+      "loss": 0.2364,
+      "step": 4105
+    },
+    {
+      "epoch": 2.3015695067264574,
+      "grad_norm": 0.07745309053089348,
+      "learning_rate": 3.125292021578822e-05,
+      "loss": 0.2493,
+      "step": 4106
+    },
+    {
+      "epoch": 2.3021300448430493,
+      "grad_norm": 0.07826243616532201,
+      "learning_rate": 3.120556228878174e-05,
+      "loss": 0.2446,
+      "step": 4107
+    },
+    {
+      "epoch": 2.302690582959641,
+      "grad_norm": 0.07712325781994052,
+      "learning_rate": 3.115823363564254e-05,
+      "loss": 0.2477,
+      "step": 4108
+    },
+    {
+      "epoch": 2.303251121076233,
+      "grad_norm": 0.07904786797151007,
+      "learning_rate": 3.111093427651016e-05,
+      "loss": 0.2404,
+      "step": 4109
+    },
+    {
+      "epoch": 2.3038116591928253,
+      "grad_norm": 0.07768571598879805,
+      "learning_rate": 3.1063664231511737e-05,
+      "loss": 0.2434,
+      "step": 4110
+    },
+    {
+      "epoch": 2.304372197309417,
+      "grad_norm": 0.08103527833027029,
+      "learning_rate": 3.101642352076194e-05,
+      "loss": 0.2432,
+      "step": 4111
+    },
+    {
+      "epoch": 2.304932735426009,
+      "grad_norm": 0.07757052847999166,
+      "learning_rate": 3.0969212164362957e-05,
+      "loss": 0.2328,
+      "step": 4112
+    },
+    {
+      "epoch": 2.305493273542601,
+      "grad_norm": 0.07919100705158838,
+      "learning_rate": 3.092203018240453e-05,
+      "loss": 0.2495,
+      "step": 4113
+    },
+    {
+      "epoch": 2.3060538116591927,
+      "grad_norm": 0.08109971919640106,
+      "learning_rate": 3.087487759496377e-05,
+      "loss": 0.2458,
+      "step": 4114
+    },
+    {
+      "epoch": 2.306614349775785,
+      "grad_norm": 0.08030104502170897,
+      "learning_rate": 3.0827754422105416e-05,
+      "loss": 0.2462,
+      "step": 4115
+    },
+    {
+      "epoch": 2.307174887892377,
+      "grad_norm": 0.0792602287700327,
+      "learning_rate": 3.078066068388162e-05,
+      "loss": 0.2449,
+      "step": 4116
+    },
+    {
+      "epoch": 2.3077354260089686,
+      "grad_norm": 0.07674638180789069,
+      "learning_rate": 3.0733596400331985e-05,
+      "loss": 0.2321,
+      "step": 4117
+    },
+    {
+      "epoch": 2.3082959641255605,
+      "grad_norm": 0.0767444128071632,
+      "learning_rate": 3.0686561591483675e-05,
+      "loss": 0.2326,
+      "step": 4118
+    },
+    {
+      "epoch": 2.3088565022421523,
+      "grad_norm": 0.07942404273918463,
+      "learning_rate": 3.063955627735121e-05,
+      "loss": 0.2364,
+      "step": 4119
+    },
+    {
+      "epoch": 2.3094170403587446,
+      "grad_norm": 0.07881660767792163,
+      "learning_rate": 3.059258047793661e-05,
+      "loss": 0.2507,
+      "step": 4120
+    },
+    {
+      "epoch": 2.3099775784753365,
+      "grad_norm": 0.07788673391022796,
+      "learning_rate": 3.0545634213229344e-05,
+      "loss": 0.2413,
+      "step": 4121
+    },
+    {
+      "epoch": 2.3105381165919283,
+      "grad_norm": 0.07771950950622547,
+      "learning_rate": 3.0498717503206343e-05,
+      "loss": 0.2361,
+      "step": 4122
+    },
+    {
+      "epoch": 2.31109865470852,
+      "grad_norm": 0.07753886183078841,
+      "learning_rate": 3.0451830367831858e-05,
+      "loss": 0.2541,
+      "step": 4123
+    },
+    {
+      "epoch": 2.311659192825112,
+      "grad_norm": 0.0776569441223699,
+      "learning_rate": 3.040497282705761e-05,
+      "loss": 0.2443,
+      "step": 4124
+    },
+    {
+      "epoch": 2.312219730941704,
+      "grad_norm": 0.07951166059416201,
+      "learning_rate": 3.0358144900822782e-05,
+      "loss": 0.2426,
+      "step": 4125
+    },
+    {
+      "epoch": 2.312780269058296,
+      "grad_norm": 0.07821582041906866,
+      "learning_rate": 3.0311346609053838e-05,
+      "loss": 0.2492,
+      "step": 4126
+    },
+    {
+      "epoch": 2.313340807174888,
+      "grad_norm": 0.0769999045840295,
+      "learning_rate": 3.0264577971664764e-05,
+      "loss": 0.2416,
+      "step": 4127
+    },
+    {
+      "epoch": 2.31390134529148,
+      "grad_norm": 0.08098583886406661,
+      "learning_rate": 3.0217839008556816e-05,
+      "loss": 0.2548,
+      "step": 4128
+    },
+    {
+      "epoch": 2.3144618834080717,
+      "grad_norm": 0.07957081430188001,
+      "learning_rate": 3.0171129739618676e-05,
+      "loss": 0.2447,
+      "step": 4129
+    },
+    {
+      "epoch": 2.3150224215246635,
+      "grad_norm": 0.08109236841779928,
+      "learning_rate": 3.0124450184726426e-05,
+      "loss": 0.2498,
+      "step": 4130
+    },
+    {
+      "epoch": 2.3155829596412554,
+      "grad_norm": 0.07542664082322645,
+      "learning_rate": 3.0077800363743404e-05,
+      "loss": 0.2414,
+      "step": 4131
+    },
+    {
+      "epoch": 2.3161434977578477,
+      "grad_norm": 0.07811373621644717,
+      "learning_rate": 3.003118029652041e-05,
+      "loss": 0.2423,
+      "step": 4132
+    },
+    {
+      "epoch": 2.3167040358744395,
+      "grad_norm": 0.07683433823903178,
+      "learning_rate": 2.998459000289545e-05,
+      "loss": 0.2392,
+      "step": 4133
+    },
+    {
+      "epoch": 2.3172645739910314,
+      "grad_norm": 0.07650638507528401,
+      "learning_rate": 2.993802950269402e-05,
+      "loss": 0.2324,
+      "step": 4134
+    },
+    {
+      "epoch": 2.317825112107623,
+      "grad_norm": 0.0781286363655784,
+      "learning_rate": 2.989149881572878e-05,
+      "loss": 0.2376,
+      "step": 4135
+    },
+    {
+      "epoch": 2.318385650224215,
+      "grad_norm": 0.08039557730752402,
+      "learning_rate": 2.9844997961799814e-05,
+      "loss": 0.2494,
+      "step": 4136
+    },
+    {
+      "epoch": 2.3189461883408073,
+      "grad_norm": 0.07599771755317568,
+      "learning_rate": 2.9798526960694496e-05,
+      "loss": 0.2334,
+      "step": 4137
+    },
+    {
+      "epoch": 2.319506726457399,
+      "grad_norm": 0.07976903775833245,
+      "learning_rate": 2.9752085832187416e-05,
+      "loss": 0.2456,
+      "step": 4138
+    },
+    {
+      "epoch": 2.320067264573991,
+      "grad_norm": 0.07863155156129803,
+      "learning_rate": 2.970567459604059e-05,
+      "loss": 0.2383,
+      "step": 4139
+    },
+    {
+      "epoch": 2.320627802690583,
+      "grad_norm": 0.077354580932718,
+      "learning_rate": 2.9659293272003164e-05,
+      "loss": 0.2392,
+      "step": 4140
+    },
+    {
+      "epoch": 2.3211883408071747,
+      "grad_norm": 0.07702481247877152,
+      "learning_rate": 2.9612941879811684e-05,
+      "loss": 0.2529,
+      "step": 4141
+    },
+    {
+      "epoch": 2.321748878923767,
+      "grad_norm": 0.07938052518841564,
+      "learning_rate": 2.9566620439189874e-05,
+      "loss": 0.2526,
+      "step": 4142
+    },
+    {
+      "epoch": 2.322309417040359,
+      "grad_norm": 0.07983346286425103,
+      "learning_rate": 2.952032896984871e-05,
+      "loss": 0.2456,
+      "step": 4143
+    },
+    {
+      "epoch": 2.3228699551569507,
+      "grad_norm": 0.07883227984065583,
+      "learning_rate": 2.947406749148649e-05,
+      "loss": 0.2391,
+      "step": 4144
+    },
+    {
+      "epoch": 2.3234304932735426,
+      "grad_norm": 0.07631071687041674,
+      "learning_rate": 2.942783602378869e-05,
+      "loss": 0.2404,
+      "step": 4145
+    },
+    {
+      "epoch": 2.3239910313901344,
+      "grad_norm": 0.07880079020974173,
+      "learning_rate": 2.9381634586428085e-05,
+      "loss": 0.2428,
+      "step": 4146
+    },
+    {
+      "epoch": 2.3245515695067267,
+      "grad_norm": 0.07806630823958408,
+      "learning_rate": 2.933546319906453e-05,
+      "loss": 0.2347,
+      "step": 4147
+    },
+    {
+      "epoch": 2.3251121076233185,
+      "grad_norm": 0.07890647078523001,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 0.2418,
+      "step": 4148
+    },
+    {
+      "epoch": 2.3256726457399104,
+      "grad_norm": 0.08074047596923722,
+      "learning_rate": 2.9243210652904596e-05,
+      "loss": 0.2469,
+      "step": 4149
+    },
+    {
+      "epoch": 2.3262331838565022,
+      "grad_norm": 0.08036244888123181,
+      "learning_rate": 2.9197129533364065e-05,
+      "loss": 0.2597,
+      "step": 4150
+    },
+    {
+      "epoch": 2.326793721973094,
+      "grad_norm": 0.07550127743518344,
+      "learning_rate": 2.9151078542332478e-05,
+      "loss": 0.2384,
+      "step": 4151
+    },
+    {
+      "epoch": 2.327354260089686,
+      "grad_norm": 0.07672098868048535,
+      "learning_rate": 2.9105057699405702e-05,
+      "loss": 0.2481,
+      "step": 4152
+    },
+    {
+      "epoch": 2.327914798206278,
+      "grad_norm": 0.0783208984911028,
+      "learning_rate": 2.9059067024166854e-05,
+      "loss": 0.2542,
+      "step": 4153
+    },
+    {
+      "epoch": 2.32847533632287,
+      "grad_norm": 0.08095674337572743,
+      "learning_rate": 2.9013106536186186e-05,
+      "loss": 0.2465,
+      "step": 4154
+    },
+    {
+      "epoch": 2.329035874439462,
+      "grad_norm": 0.07680151887015477,
+      "learning_rate": 2.8967176255021167e-05,
+      "loss": 0.2315,
+      "step": 4155
+    },
+    {
+      "epoch": 2.3295964125560538,
+      "grad_norm": 0.0780413711790148,
+      "learning_rate": 2.8921276200216296e-05,
+      "loss": 0.2419,
+      "step": 4156
+    },
+    {
+      "epoch": 2.3301569506726456,
+      "grad_norm": 0.08115479226965608,
+      "learning_rate": 2.8875406391303263e-05,
+      "loss": 0.2324,
+      "step": 4157
+    },
+    {
+      "epoch": 2.3307174887892375,
+      "grad_norm": 0.08035250801214854,
+      "learning_rate": 2.8829566847800948e-05,
+      "loss": 0.24,
+      "step": 4158
+    },
+    {
+      "epoch": 2.3312780269058297,
+      "grad_norm": 0.07953425457381018,
+      "learning_rate": 2.878375758921522e-05,
+      "loss": 0.2514,
+      "step": 4159
+    },
+    {
+      "epoch": 2.3318385650224216,
+      "grad_norm": 0.07773993048067764,
+      "learning_rate": 2.873797863503923e-05,
+      "loss": 0.2341,
+      "step": 4160
+    },
+    {
+      "epoch": 2.3323991031390134,
+      "grad_norm": 0.0784552985488261,
+      "learning_rate": 2.8692230004753063e-05,
+      "loss": 0.2489,
+      "step": 4161
+    },
+    {
+      "epoch": 2.3329596412556053,
+      "grad_norm": 0.07722982653058977,
+      "learning_rate": 2.8646511717824022e-05,
+      "loss": 0.24,
+      "step": 4162
+    },
+    {
+      "epoch": 2.333520179372197,
+      "grad_norm": 0.0796887994841046,
+      "learning_rate": 2.8600823793706476e-05,
+      "loss": 0.2467,
+      "step": 4163
+    },
+    {
+      "epoch": 2.3340807174887894,
+      "grad_norm": 0.07927708767540144,
+      "learning_rate": 2.8555166251841802e-05,
+      "loss": 0.2376,
+      "step": 4164
+    },
+    {
+      "epoch": 2.3346412556053813,
+      "grad_norm": 0.07721321287976032,
+      "learning_rate": 2.850953911165857e-05,
+      "loss": 0.2494,
+      "step": 4165
+    },
+    {
+      "epoch": 2.335201793721973,
+      "grad_norm": 0.07789760540328405,
+      "learning_rate": 2.846394239257226e-05,
+      "loss": 0.2395,
+      "step": 4166
+    },
+    {
+      "epoch": 2.335762331838565,
+      "grad_norm": 0.07920922048422345,
+      "learning_rate": 2.841837611398558e-05,
+      "loss": 0.2515,
+      "step": 4167
+    },
+    {
+      "epoch": 2.336322869955157,
+      "grad_norm": 0.08017300403346657,
+      "learning_rate": 2.8372840295288106e-05,
+      "loss": 0.2473,
+      "step": 4168
+    },
+    {
+      "epoch": 2.336883408071749,
+      "grad_norm": 0.0787310910925733,
+      "learning_rate": 2.8327334955856598e-05,
+      "loss": 0.2441,
+      "step": 4169
+    },
+    {
+      "epoch": 2.337443946188341,
+      "grad_norm": 0.07895596255184466,
+      "learning_rate": 2.8281860115054815e-05,
+      "loss": 0.2377,
+      "step": 4170
+    },
+    {
+      "epoch": 2.338004484304933,
+      "grad_norm": 0.0756965468129164,
+      "learning_rate": 2.823641579223344e-05,
+      "loss": 0.2416,
+      "step": 4171
+    },
+    {
+      "epoch": 2.3385650224215246,
+      "grad_norm": 0.07708556236193148,
+      "learning_rate": 2.8191002006730328e-05,
+      "loss": 0.2574,
+      "step": 4172
+    },
+    {
+      "epoch": 2.3391255605381165,
+      "grad_norm": 0.07907645792962631,
+      "learning_rate": 2.8145618777870176e-05,
+      "loss": 0.2405,
+      "step": 4173
+    },
+    {
+      "epoch": 2.3396860986547083,
+      "grad_norm": 0.07762214440323861,
+      "learning_rate": 2.8100266124964824e-05,
+      "loss": 0.2352,
+      "step": 4174
+    },
+    {
+      "epoch": 2.3402466367713006,
+      "grad_norm": 0.07667205585008668,
+      "learning_rate": 2.8054944067313005e-05,
+      "loss": 0.2387,
+      "step": 4175
+    },
+    {
+      "epoch": 2.3408071748878925,
+      "grad_norm": 0.07914785379616622,
+      "learning_rate": 2.800965262420043e-05,
+      "loss": 0.2569,
+      "step": 4176
+    },
+    {
+      "epoch": 2.3413677130044843,
+      "grad_norm": 0.07786459051819472,
+      "learning_rate": 2.796439181489985e-05,
+      "loss": 0.2348,
+      "step": 4177
+    },
+    {
+      "epoch": 2.341928251121076,
+      "grad_norm": 0.07729522574347791,
+      "learning_rate": 2.7919161658670945e-05,
+      "loss": 0.2408,
+      "step": 4178
+    },
+    {
+      "epoch": 2.342488789237668,
+      "grad_norm": 0.07639465568166552,
+      "learning_rate": 2.787396217476038e-05,
+      "loss": 0.2362,
+      "step": 4179
+    },
+    {
+      "epoch": 2.34304932735426,
+      "grad_norm": 0.0763207392197241,
+      "learning_rate": 2.7828793382401685e-05,
+      "loss": 0.2334,
+      "step": 4180
+    },
+    {
+      "epoch": 2.343609865470852,
+      "grad_norm": 0.08115765241592474,
+      "learning_rate": 2.7783655300815447e-05,
+      "loss": 0.2497,
+      "step": 4181
+    },
+    {
+      "epoch": 2.344170403587444,
+      "grad_norm": 0.07869998341763018,
+      "learning_rate": 2.7738547949209082e-05,
+      "loss": 0.2406,
+      "step": 4182
+    },
+    {
+      "epoch": 2.344730941704036,
+      "grad_norm": 0.07820693522758504,
+      "learning_rate": 2.7693471346776944e-05,
+      "loss": 0.2441,
+      "step": 4183
+    },
+    {
+      "epoch": 2.3452914798206277,
+      "grad_norm": 0.07758533438790251,
+      "learning_rate": 2.7648425512700393e-05,
+      "loss": 0.2376,
+      "step": 4184
+    },
+    {
+      "epoch": 2.3458520179372195,
+      "grad_norm": 0.08007894475897577,
+      "learning_rate": 2.7603410466147572e-05,
+      "loss": 0.2525,
+      "step": 4185
+    },
+    {
+      "epoch": 2.346412556053812,
+      "grad_norm": 0.07936192949458998,
+      "learning_rate": 2.7558426226273615e-05,
+      "loss": 0.2499,
+      "step": 4186
+    },
+    {
+      "epoch": 2.3469730941704037,
+      "grad_norm": 0.07874011422451643,
+      "learning_rate": 2.751347281222051e-05,
+      "loss": 0.2378,
+      "step": 4187
+    },
+    {
+      "epoch": 2.3475336322869955,
+      "grad_norm": 0.07898987392037829,
+      "learning_rate": 2.7468550243117165e-05,
+      "loss": 0.2374,
+      "step": 4188
+    },
+    {
+      "epoch": 2.3480941704035874,
+      "grad_norm": 0.07785486151463969,
+      "learning_rate": 2.74236585380793e-05,
+      "loss": 0.2372,
+      "step": 4189
+    },
+    {
+      "epoch": 2.348654708520179,
+      "grad_norm": 0.07853561330172434,
+      "learning_rate": 2.7378797716209503e-05,
+      "loss": 0.2428,
+      "step": 4190
+    },
+    {
+      "epoch": 2.3492152466367715,
+      "grad_norm": 0.07875410139224716,
+      "learning_rate": 2.7333967796597315e-05,
+      "loss": 0.2493,
+      "step": 4191
+    },
+    {
+      "epoch": 2.3497757847533634,
+      "grad_norm": 0.0804659168601273,
+      "learning_rate": 2.7289168798318997e-05,
+      "loss": 0.2378,
+      "step": 4192
+    },
+    {
+      "epoch": 2.350336322869955,
+      "grad_norm": 0.0793124955256602,
+      "learning_rate": 2.724440074043778e-05,
+      "loss": 0.2439,
+      "step": 4193
+    },
+    {
+      "epoch": 2.350896860986547,
+      "grad_norm": 0.07862219014425169,
+      "learning_rate": 2.7199663642003603e-05,
+      "loss": 0.2363,
+      "step": 4194
+    },
+    {
+      "epoch": 2.351457399103139,
+      "grad_norm": 0.07942827124797995,
+      "learning_rate": 2.7154957522053316e-05,
+      "loss": 0.2403,
+      "step": 4195
+    },
+    {
+      "epoch": 2.352017937219731,
+      "grad_norm": 0.07925172483902786,
+      "learning_rate": 2.711028239961061e-05,
+      "loss": 0.2401,
+      "step": 4196
+    },
+    {
+      "epoch": 2.352578475336323,
+      "grad_norm": 0.0794265937335694,
+      "learning_rate": 2.706563829368587e-05,
+      "loss": 0.2566,
+      "step": 4197
+    },
+    {
+      "epoch": 2.353139013452915,
+      "grad_norm": 0.07786233837941513,
+      "learning_rate": 2.702102522327642e-05,
+      "loss": 0.2415,
+      "step": 4198
+    },
+    {
+      "epoch": 2.3536995515695067,
+      "grad_norm": 0.07655076644210121,
+      "learning_rate": 2.6976443207366255e-05,
+      "loss": 0.2405,
+      "step": 4199
+    },
+    {
+      "epoch": 2.3542600896860986,
+      "grad_norm": 0.0795775900145881,
+      "learning_rate": 2.693189226492625e-05,
+      "loss": 0.2418,
+      "step": 4200
+    },
+    {
+      "epoch": 2.3548206278026904,
+      "grad_norm": 0.07792781835497384,
+      "learning_rate": 2.688737241491398e-05,
+      "loss": 0.2346,
+      "step": 4201
+    },
+    {
+      "epoch": 2.3553811659192827,
+      "grad_norm": 0.07977929198265812,
+      "learning_rate": 2.6842883676273857e-05,
+      "loss": 0.2431,
+      "step": 4202
+    },
+    {
+      "epoch": 2.3559417040358746,
+      "grad_norm": 0.08037315181607053,
+      "learning_rate": 2.6798426067937045e-05,
+      "loss": 0.2364,
+      "step": 4203
+    },
+    {
+      "epoch": 2.3565022421524664,
+      "grad_norm": 0.07552057334313014,
+      "learning_rate": 2.675399960882138e-05,
+      "loss": 0.2435,
+      "step": 4204
+    },
+    {
+      "epoch": 2.3570627802690582,
+      "grad_norm": 0.07822281390195752,
+      "learning_rate": 2.6709604317831583e-05,
+      "loss": 0.2395,
+      "step": 4205
+    },
+    {
+      "epoch": 2.35762331838565,
+      "grad_norm": 0.07508628675605872,
+      "learning_rate": 2.6665240213858946e-05,
+      "loss": 0.2329,
+      "step": 4206
+    },
+    {
+      "epoch": 2.358183856502242,
+      "grad_norm": 0.07824720213732886,
+      "learning_rate": 2.6620907315781662e-05,
+      "loss": 0.2546,
+      "step": 4207
+    },
+    {
+      "epoch": 2.3587443946188342,
+      "grad_norm": 0.07708763304847759,
+      "learning_rate": 2.657660564246449e-05,
+      "loss": 0.2388,
+      "step": 4208
+    },
+    {
+      "epoch": 2.359304932735426,
+      "grad_norm": 0.07979685337298173,
+      "learning_rate": 2.653233521275904e-05,
+      "loss": 0.2557,
+      "step": 4209
+    },
+    {
+      "epoch": 2.359865470852018,
+      "grad_norm": 0.07847072190202505,
+      "learning_rate": 2.6488096045503485e-05,
+      "loss": 0.2403,
+      "step": 4210
+    },
+    {
+      "epoch": 2.3604260089686098,
+      "grad_norm": 0.08144716623859771,
+      "learning_rate": 2.6443888159522823e-05,
+      "loss": 0.2565,
+      "step": 4211
+    },
+    {
+      "epoch": 2.3609865470852016,
+      "grad_norm": 0.0761119790546523,
+      "learning_rate": 2.6399711573628704e-05,
+      "loss": 0.2463,
+      "step": 4212
+    },
+    {
+      "epoch": 2.361547085201794,
+      "grad_norm": 0.07790333307748155,
+      "learning_rate": 2.6355566306619373e-05,
+      "loss": 0.2383,
+      "step": 4213
+    },
+    {
+      "epoch": 2.3621076233183858,
+      "grad_norm": 0.07959313799088172,
+      "learning_rate": 2.63114523772799e-05,
+      "loss": 0.2404,
+      "step": 4214
+    },
+    {
+      "epoch": 2.3626681614349776,
+      "grad_norm": 0.08043596452556043,
+      "learning_rate": 2.626736980438189e-05,
+      "loss": 0.2498,
+      "step": 4215
+    },
+    {
+      "epoch": 2.3632286995515694,
+      "grad_norm": 0.07833916486793696,
+      "learning_rate": 2.6223318606683645e-05,
+      "loss": 0.2379,
+      "step": 4216
+    },
+    {
+      "epoch": 2.3637892376681613,
+      "grad_norm": 0.07900490105132915,
+      "learning_rate": 2.6179298802930154e-05,
+      "loss": 0.2514,
+      "step": 4217
+    },
+    {
+      "epoch": 2.3643497757847536,
+      "grad_norm": 0.07616742985296307,
+      "learning_rate": 2.6135310411852977e-05,
+      "loss": 0.2343,
+      "step": 4218
+    },
+    {
+      "epoch": 2.3649103139013454,
+      "grad_norm": 0.07710329656177863,
+      "learning_rate": 2.6091353452170375e-05,
+      "loss": 0.2253,
+      "step": 4219
+    },
+    {
+      "epoch": 2.3654708520179373,
+      "grad_norm": 0.07824264240298706,
+      "learning_rate": 2.60474279425872e-05,
+      "loss": 0.2391,
+      "step": 4220
+    },
+    {
+      "epoch": 2.366031390134529,
+      "grad_norm": 0.07744570500088467,
+      "learning_rate": 2.6003533901794962e-05,
+      "loss": 0.2376,
+      "step": 4221
+    },
+    {
+      "epoch": 2.366591928251121,
+      "grad_norm": 0.0792097248625774,
+      "learning_rate": 2.5959671348471715e-05,
+      "loss": 0.2443,
+      "step": 4222
+    },
+    {
+      "epoch": 2.367152466367713,
+      "grad_norm": 0.08155545890005843,
+      "learning_rate": 2.5915840301282114e-05,
+      "loss": 0.2414,
+      "step": 4223
+    },
+    {
+      "epoch": 2.367713004484305,
+      "grad_norm": 0.08197006402100236,
+      "learning_rate": 2.5872040778877503e-05,
+      "loss": 0.2547,
+      "step": 4224
+    },
+    {
+      "epoch": 2.368273542600897,
+      "grad_norm": 0.0788384367686174,
+      "learning_rate": 2.582827279989568e-05,
+      "loss": 0.2546,
+      "step": 4225
+    },
+    {
+      "epoch": 2.368834080717489,
+      "grad_norm": 0.08117189338626334,
+      "learning_rate": 2.5784536382961145e-05,
+      "loss": 0.2393,
+      "step": 4226
+    },
+    {
+      "epoch": 2.3693946188340806,
+      "grad_norm": 0.07847918315769185,
+      "learning_rate": 2.5740831546684853e-05,
+      "loss": 0.2453,
+      "step": 4227
+    },
+    {
+      "epoch": 2.3699551569506725,
+      "grad_norm": 0.07660854798589885,
+      "learning_rate": 2.5697158309664404e-05,
+      "loss": 0.2268,
+      "step": 4228
+    },
+    {
+      "epoch": 2.3705156950672643,
+      "grad_norm": 0.07993643725998426,
+      "learning_rate": 2.565351669048397e-05,
+      "loss": 0.2334,
+      "step": 4229
+    },
+    {
+      "epoch": 2.3710762331838566,
+      "grad_norm": 0.0788627487921801,
+      "learning_rate": 2.5609906707714137e-05,
+      "loss": 0.2464,
+      "step": 4230
+    },
+    {
+      "epoch": 2.3716367713004485,
+      "grad_norm": 0.07878287045150213,
+      "learning_rate": 2.5566328379912196e-05,
+      "loss": 0.238,
+      "step": 4231
+    },
+    {
+      "epoch": 2.3721973094170403,
+      "grad_norm": 0.08726667034774241,
+      "learning_rate": 2.5522781725621813e-05,
+      "loss": 0.2599,
+      "step": 4232
+    },
+    {
+      "epoch": 2.372757847533632,
+      "grad_norm": 0.07917378228374336,
+      "learning_rate": 2.547926676337333e-05,
+      "loss": 0.2437,
+      "step": 4233
+    },
+    {
+      "epoch": 2.373318385650224,
+      "grad_norm": 0.07998427079224918,
+      "learning_rate": 2.5435783511683443e-05,
+      "loss": 0.246,
+      "step": 4234
+    },
+    {
+      "epoch": 2.3738789237668163,
+      "grad_norm": 0.08095502398783906,
+      "learning_rate": 2.5392331989055486e-05,
+      "loss": 0.2427,
+      "step": 4235
+    },
+    {
+      "epoch": 2.374439461883408,
+      "grad_norm": 0.0818390218436475,
+      "learning_rate": 2.5348912213979235e-05,
+      "loss": 0.2426,
+      "step": 4236
+    },
+    {
+      "epoch": 2.375,
+      "grad_norm": 0.0811426481293932,
+      "learning_rate": 2.530552420493094e-05,
+      "loss": 0.2431,
+      "step": 4237
+    },
+    {
+      "epoch": 2.375560538116592,
+      "grad_norm": 0.0785544440402398,
+      "learning_rate": 2.5262167980373395e-05,
+      "loss": 0.2467,
+      "step": 4238
+    },
+    {
+      "epoch": 2.3761210762331837,
+      "grad_norm": 0.07859659167847614,
+      "learning_rate": 2.5218843558755778e-05,
+      "loss": 0.244,
+      "step": 4239
+    },
+    {
+      "epoch": 2.376681614349776,
+      "grad_norm": 0.08042249061124941,
+      "learning_rate": 2.5175550958513837e-05,
+      "loss": 0.2393,
+      "step": 4240
+    },
+    {
+      "epoch": 2.377242152466368,
+      "grad_norm": 0.07820509183110318,
+      "learning_rate": 2.5132290198069675e-05,
+      "loss": 0.2405,
+      "step": 4241
+    },
+    {
+      "epoch": 2.3778026905829597,
+      "grad_norm": 0.07719627068710612,
+      "learning_rate": 2.508906129583195e-05,
+      "loss": 0.2404,
+      "step": 4242
+    },
+    {
+      "epoch": 2.3783632286995515,
+      "grad_norm": 0.07934775266755617,
+      "learning_rate": 2.5045864270195675e-05,
+      "loss": 0.2398,
+      "step": 4243
+    },
+    {
+      "epoch": 2.3789237668161434,
+      "grad_norm": 0.07976585651151132,
+      "learning_rate": 2.500269913954233e-05,
+      "loss": 0.2515,
+      "step": 4244
+    },
+    {
+      "epoch": 2.3794843049327357,
+      "grad_norm": 0.07674324206891203,
+      "learning_rate": 2.495956592223988e-05,
+      "loss": 0.2368,
+      "step": 4245
+    },
+    {
+      "epoch": 2.3800448430493275,
+      "grad_norm": 0.0804676516461419,
+      "learning_rate": 2.491646463664261e-05,
+      "loss": 0.2535,
+      "step": 4246
+    },
+    {
+      "epoch": 2.3806053811659194,
+      "grad_norm": 0.07980334937569152,
+      "learning_rate": 2.4873395301091306e-05,
+      "loss": 0.2437,
+      "step": 4247
+    },
+    {
+      "epoch": 2.381165919282511,
+      "grad_norm": 0.078867433642548,
+      "learning_rate": 2.4830357933913063e-05,
+      "loss": 0.2546,
+      "step": 4248
+    },
+    {
+      "epoch": 2.381726457399103,
+      "grad_norm": 0.07672055027397391,
+      "learning_rate": 2.4787352553421493e-05,
+      "loss": 0.2313,
+      "step": 4249
+    },
+    {
+      "epoch": 2.382286995515695,
+      "grad_norm": 0.0787764644600143,
+      "learning_rate": 2.4744379177916498e-05,
+      "loss": 0.2392,
+      "step": 4250
+    },
+    {
+      "epoch": 2.382847533632287,
+      "grad_norm": 0.07969681313403791,
+      "learning_rate": 2.470143782568436e-05,
+      "loss": 0.2389,
+      "step": 4251
+    },
+    {
+      "epoch": 2.383408071748879,
+      "grad_norm": 0.0809229643912626,
+      "learning_rate": 2.4658528514997815e-05,
+      "loss": 0.245,
+      "step": 4252
+    },
+    {
+      "epoch": 2.383968609865471,
+      "grad_norm": 0.07820701687792098,
+      "learning_rate": 2.4615651264115903e-05,
+      "loss": 0.2258,
+      "step": 4253
+    },
+    {
+      "epoch": 2.3845291479820627,
+      "grad_norm": 0.07881482865000651,
+      "learning_rate": 2.4572806091284073e-05,
+      "loss": 0.247,
+      "step": 4254
+    },
+    {
+      "epoch": 2.3850896860986546,
+      "grad_norm": 0.07810169967599107,
+      "learning_rate": 2.452999301473403e-05,
+      "loss": 0.2477,
+      "step": 4255
+    },
+    {
+      "epoch": 2.3856502242152464,
+      "grad_norm": 0.07822537831160745,
+      "learning_rate": 2.448721205268395e-05,
+      "loss": 0.239,
+      "step": 4256
+    },
+    {
+      "epoch": 2.3862107623318387,
+      "grad_norm": 0.08059235644032076,
+      "learning_rate": 2.444446322333821e-05,
+      "loss": 0.2576,
+      "step": 4257
+    },
+    {
+      "epoch": 2.3867713004484306,
+      "grad_norm": 0.07991668386232427,
+      "learning_rate": 2.4401746544887584e-05,
+      "loss": 0.2508,
+      "step": 4258
+    },
+    {
+      "epoch": 2.3873318385650224,
+      "grad_norm": 0.08189843528545956,
+      "learning_rate": 2.435906203550916e-05,
+      "loss": 0.2457,
+      "step": 4259
+    },
+    {
+      "epoch": 2.3878923766816142,
+      "grad_norm": 0.07693101994417906,
+      "learning_rate": 2.4316409713366352e-05,
+      "loss": 0.248,
+      "step": 4260
+    },
+    {
+      "epoch": 2.388452914798206,
+      "grad_norm": 0.07645113103596757,
+      "learning_rate": 2.4273789596608887e-05,
+      "loss": 0.2281,
+      "step": 4261
+    },
+    {
+      "epoch": 2.3890134529147984,
+      "grad_norm": 0.0756294332759964,
+      "learning_rate": 2.423120170337272e-05,
+      "loss": 0.2398,
+      "step": 4262
+    },
+    {
+      "epoch": 2.3895739910313902,
+      "grad_norm": 0.08038861463910725,
+      "learning_rate": 2.4188646051780117e-05,
+      "loss": 0.2453,
+      "step": 4263
+    },
+    {
+      "epoch": 2.390134529147982,
+      "grad_norm": 0.0807597476574884,
+      "learning_rate": 2.4146122659939686e-05,
+      "loss": 0.2369,
+      "step": 4264
+    },
+    {
+      "epoch": 2.390695067264574,
+      "grad_norm": 0.07738844331988792,
+      "learning_rate": 2.4103631545946225e-05,
+      "loss": 0.2322,
+      "step": 4265
+    },
+    {
+      "epoch": 2.3912556053811658,
+      "grad_norm": 0.08144564840558986,
+      "learning_rate": 2.4061172727880886e-05,
+      "loss": 0.2498,
+      "step": 4266
+    },
+    {
+      "epoch": 2.391816143497758,
+      "grad_norm": 0.08014202723678344,
+      "learning_rate": 2.4018746223810974e-05,
+      "loss": 0.2515,
+      "step": 4267
+    },
+    {
+      "epoch": 2.39237668161435,
+      "grad_norm": 0.08100509140202429,
+      "learning_rate": 2.3976352051790117e-05,
+      "loss": 0.2495,
+      "step": 4268
+    },
+    {
+      "epoch": 2.3929372197309418,
+      "grad_norm": 0.07794262279111899,
+      "learning_rate": 2.3933990229858193e-05,
+      "loss": 0.239,
+      "step": 4269
+    },
+    {
+      "epoch": 2.3934977578475336,
+      "grad_norm": 0.08012132421360765,
+      "learning_rate": 2.3891660776041247e-05,
+      "loss": 0.249,
+      "step": 4270
+    },
+    {
+      "epoch": 2.3940582959641254,
+      "grad_norm": 0.07733188014743615,
+      "learning_rate": 2.3849363708351625e-05,
+      "loss": 0.2406,
+      "step": 4271
+    },
+    {
+      "epoch": 2.3946188340807173,
+      "grad_norm": 0.07792694206848627,
+      "learning_rate": 2.3807099044787818e-05,
+      "loss": 0.2393,
+      "step": 4272
+    },
+    {
+      "epoch": 2.3951793721973096,
+      "grad_norm": 0.07738852943741609,
+      "learning_rate": 2.3764866803334606e-05,
+      "loss": 0.2354,
+      "step": 4273
+    },
+    {
+      "epoch": 2.3957399103139014,
+      "grad_norm": 0.07792379658822556,
+      "learning_rate": 2.3722667001962896e-05,
+      "loss": 0.2349,
+      "step": 4274
+    },
+    {
+      "epoch": 2.3963004484304933,
+      "grad_norm": 0.07742232260896334,
+      "learning_rate": 2.3680499658629874e-05,
+      "loss": 0.2382,
+      "step": 4275
+    },
+    {
+      "epoch": 2.396860986547085,
+      "grad_norm": 0.07856630204479857,
+      "learning_rate": 2.363836479127881e-05,
+      "loss": 0.2444,
+      "step": 4276
+    },
+    {
+      "epoch": 2.397421524663677,
+      "grad_norm": 0.07927930061326396,
+      "learning_rate": 2.3596262417839255e-05,
+      "loss": 0.2419,
+      "step": 4277
+    },
+    {
+      "epoch": 2.397982062780269,
+      "grad_norm": 0.07836612250875499,
+      "learning_rate": 2.3554192556226896e-05,
+      "loss": 0.2313,
+      "step": 4278
+    },
+    {
+      "epoch": 2.398542600896861,
+      "grad_norm": 0.07940189246279593,
+      "learning_rate": 2.3512155224343546e-05,
+      "loss": 0.2567,
+      "step": 4279
+    },
+    {
+      "epoch": 2.399103139013453,
+      "grad_norm": 0.07828878517899063,
+      "learning_rate": 2.3470150440077266e-05,
+      "loss": 0.2458,
+      "step": 4280
+    },
+    {
+      "epoch": 2.399663677130045,
+      "grad_norm": 0.07912544986227371,
+      "learning_rate": 2.3428178221302144e-05,
+      "loss": 0.2367,
+      "step": 4281
+    },
+    {
+      "epoch": 2.4002242152466366,
+      "grad_norm": 0.07855078011819978,
+      "learning_rate": 2.3386238585878538e-05,
+      "loss": 0.2495,
+      "step": 4282
+    },
+    {
+      "epoch": 2.4007847533632285,
+      "grad_norm": 0.0823602874783878,
+      "learning_rate": 2.3344331551652854e-05,
+      "loss": 0.2465,
+      "step": 4283
+    },
+    {
+      "epoch": 2.401345291479821,
+      "grad_norm": 0.08065013835008998,
+      "learning_rate": 2.3302457136457623e-05,
+      "loss": 0.2555,
+      "step": 4284
+    },
+    {
+      "epoch": 2.4019058295964126,
+      "grad_norm": 0.0782231546469997,
+      "learning_rate": 2.326061535811156e-05,
+      "loss": 0.238,
+      "step": 4285
+    },
+    {
+      "epoch": 2.4024663677130045,
+      "grad_norm": 0.08010215013422635,
+      "learning_rate": 2.3218806234419443e-05,
+      "loss": 0.2476,
+      "step": 4286
+    },
+    {
+      "epoch": 2.4030269058295963,
+      "grad_norm": 0.07845562409630219,
+      "learning_rate": 2.317702978317221e-05,
+      "loss": 0.2486,
+      "step": 4287
+    },
+    {
+      "epoch": 2.403587443946188,
+      "grad_norm": 0.07969730646413398,
+      "learning_rate": 2.3135286022146785e-05,
+      "loss": 0.2485,
+      "step": 4288
+    },
+    {
+      "epoch": 2.4041479820627805,
+      "grad_norm": 0.07663272981994358,
+      "learning_rate": 2.3093574969106323e-05,
+      "loss": 0.2406,
+      "step": 4289
+    },
+    {
+      "epoch": 2.4047085201793723,
+      "grad_norm": 0.0776438435990643,
+      "learning_rate": 2.3051896641799952e-05,
+      "loss": 0.2424,
+      "step": 4290
+    },
+    {
+      "epoch": 2.405269058295964,
+      "grad_norm": 0.08018596927300335,
+      "learning_rate": 2.3010251057962883e-05,
+      "loss": 0.25,
+      "step": 4291
+    },
+    {
+      "epoch": 2.405829596412556,
+      "grad_norm": 0.07966056261094619,
+      "learning_rate": 2.2968638235316466e-05,
+      "loss": 0.2378,
+      "step": 4292
+    },
+    {
+      "epoch": 2.406390134529148,
+      "grad_norm": 0.0775690470940192,
+      "learning_rate": 2.292705819156803e-05,
+      "loss": 0.2362,
+      "step": 4293
+    },
+    {
+      "epoch": 2.40695067264574,
+      "grad_norm": 0.07927168083563292,
+      "learning_rate": 2.288551094441106e-05,
+      "loss": 0.2533,
+      "step": 4294
+    },
+    {
+      "epoch": 2.407511210762332,
+      "grad_norm": 0.07857339041746221,
+      "learning_rate": 2.2843996511524934e-05,
+      "loss": 0.2495,
+      "step": 4295
+    },
+    {
+      "epoch": 2.408071748878924,
+      "grad_norm": 0.07967302591749084,
+      "learning_rate": 2.2802514910575223e-05,
+      "loss": 0.2557,
+      "step": 4296
+    },
+    {
+      "epoch": 2.4086322869955157,
+      "grad_norm": 0.07861300943021306,
+      "learning_rate": 2.2761066159213417e-05,
+      "loss": 0.246,
+      "step": 4297
+    },
+    {
+      "epoch": 2.4091928251121075,
+      "grad_norm": 0.07786911660528349,
+      "learning_rate": 2.271965027507704e-05,
+      "loss": 0.2381,
+      "step": 4298
+    },
+    {
+      "epoch": 2.4097533632286994,
+      "grad_norm": 0.07775500781116534,
+      "learning_rate": 2.2678267275789712e-05,
+      "loss": 0.2415,
+      "step": 4299
+    },
+    {
+      "epoch": 2.4103139013452917,
+      "grad_norm": 0.0792791489016953,
+      "learning_rate": 2.2636917178960938e-05,
+      "loss": 0.2473,
+      "step": 4300
+    },
+    {
+      "epoch": 2.4108744394618835,
+      "grad_norm": 0.0819511269755226,
+      "learning_rate": 2.259560000218631e-05,
+      "loss": 0.2487,
+      "step": 4301
+    },
+    {
+      "epoch": 2.4114349775784754,
+      "grad_norm": 0.08299400506533303,
+      "learning_rate": 2.255431576304744e-05,
+      "loss": 0.2529,
+      "step": 4302
+    },
+    {
+      "epoch": 2.411995515695067,
+      "grad_norm": 0.07929795528474204,
+      "learning_rate": 2.25130644791118e-05,
+      "loss": 0.2341,
+      "step": 4303
+    },
+    {
+      "epoch": 2.412556053811659,
+      "grad_norm": 0.07743424489324233,
+      "learning_rate": 2.2471846167932975e-05,
+      "loss": 0.241,
+      "step": 4304
+    },
+    {
+      "epoch": 2.413116591928251,
+      "grad_norm": 0.07995720980201747,
+      "learning_rate": 2.243066084705039e-05,
+      "loss": 0.2373,
+      "step": 4305
+    },
+    {
+      "epoch": 2.413677130044843,
+      "grad_norm": 0.07970112758822016,
+      "learning_rate": 2.2389508533989555e-05,
+      "loss": 0.2378,
+      "step": 4306
+    },
+    {
+      "epoch": 2.414237668161435,
+      "grad_norm": 0.0778690871138127,
+      "learning_rate": 2.2348389246261837e-05,
+      "loss": 0.2435,
+      "step": 4307
+    },
+    {
+      "epoch": 2.414798206278027,
+      "grad_norm": 0.07685825327763682,
+      "learning_rate": 2.230730300136461e-05,
+      "loss": 0.2305,
+      "step": 4308
+    },
+    {
+      "epoch": 2.4153587443946187,
+      "grad_norm": 0.07469704153860139,
+      "learning_rate": 2.226624981678115e-05,
+      "loss": 0.237,
+      "step": 4309
+    },
+    {
+      "epoch": 2.4159192825112106,
+      "grad_norm": 0.07975969402431803,
+      "learning_rate": 2.2225229709980676e-05,
+      "loss": 0.2426,
+      "step": 4310
+    },
+    {
+      "epoch": 2.416479820627803,
+      "grad_norm": 0.08007789008632174,
+      "learning_rate": 2.218424269841838e-05,
+      "loss": 0.2452,
+      "step": 4311
+    },
+    {
+      "epoch": 2.4170403587443947,
+      "grad_norm": 0.07812765917880916,
+      "learning_rate": 2.214328879953528e-05,
+      "loss": 0.2202,
+      "step": 4312
+    },
+    {
+      "epoch": 2.4176008968609866,
+      "grad_norm": 0.07833743456277244,
+      "learning_rate": 2.210236803075839e-05,
+      "loss": 0.2446,
+      "step": 4313
+    },
+    {
+      "epoch": 2.4181614349775784,
+      "grad_norm": 0.07971713235725623,
+      "learning_rate": 2.2061480409500556e-05,
+      "loss": 0.2307,
+      "step": 4314
+    },
+    {
+      "epoch": 2.4187219730941703,
+      "grad_norm": 0.07736258676653242,
+      "learning_rate": 2.2020625953160577e-05,
+      "loss": 0.2437,
+      "step": 4315
+    },
+    {
+      "epoch": 2.4192825112107625,
+      "grad_norm": 0.07961912114769379,
+      "learning_rate": 2.1979804679123106e-05,
+      "loss": 0.2483,
+      "step": 4316
+    },
+    {
+      "epoch": 2.4198430493273544,
+      "grad_norm": 0.07935805691038576,
+      "learning_rate": 2.1939016604758656e-05,
+      "loss": 0.2363,
+      "step": 4317
+    },
+    {
+      "epoch": 2.4204035874439462,
+      "grad_norm": 0.08065309659353526,
+      "learning_rate": 2.1898261747423655e-05,
+      "loss": 0.2531,
+      "step": 4318
+    },
+    {
+      "epoch": 2.420964125560538,
+      "grad_norm": 0.07754738254742989,
+      "learning_rate": 2.1857540124460397e-05,
+      "loss": 0.2325,
+      "step": 4319
+    },
+    {
+      "epoch": 2.42152466367713,
+      "grad_norm": 0.07660030820169896,
+      "learning_rate": 2.181685175319702e-05,
+      "loss": 0.2401,
+      "step": 4320
+    },
+    {
+      "epoch": 2.422085201793722,
+      "grad_norm": 0.0793339551644271,
+      "learning_rate": 2.177619665094749e-05,
+      "loss": 0.2354,
+      "step": 4321
+    },
+    {
+      "epoch": 2.422645739910314,
+      "grad_norm": 0.07763747831226875,
+      "learning_rate": 2.1735574835011664e-05,
+      "loss": 0.248,
+      "step": 4322
+    },
+    {
+      "epoch": 2.423206278026906,
+      "grad_norm": 0.08188821089100286,
+      "learning_rate": 2.1694986322675202e-05,
+      "loss": 0.2517,
+      "step": 4323
+    },
+    {
+      "epoch": 2.4237668161434978,
+      "grad_norm": 0.07903829794138365,
+      "learning_rate": 2.1654431131209553e-05,
+      "loss": 0.2451,
+      "step": 4324
+    },
+    {
+      "epoch": 2.4243273542600896,
+      "grad_norm": 0.0813434473736274,
+      "learning_rate": 2.1613909277872056e-05,
+      "loss": 0.2546,
+      "step": 4325
+    },
+    {
+      "epoch": 2.4248878923766815,
+      "grad_norm": 0.08003868992303634,
+      "learning_rate": 2.157342077990586e-05,
+      "loss": 0.2489,
+      "step": 4326
+    },
+    {
+      "epoch": 2.4254484304932733,
+      "grad_norm": 0.07693609909216155,
+      "learning_rate": 2.1532965654539915e-05,
+      "loss": 0.2474,
+      "step": 4327
+    },
+    {
+      "epoch": 2.4260089686098656,
+      "grad_norm": 0.07950874352056599,
+      "learning_rate": 2.1492543918988907e-05,
+      "loss": 0.2525,
+      "step": 4328
+    },
+    {
+      "epoch": 2.4265695067264574,
+      "grad_norm": 0.08214382775918407,
+      "learning_rate": 2.1452155590453404e-05,
+      "loss": 0.2429,
+      "step": 4329
+    },
+    {
+      "epoch": 2.4271300448430493,
+      "grad_norm": 0.0758595129042081,
+      "learning_rate": 2.141180068611971e-05,
+      "loss": 0.2265,
+      "step": 4330
+    },
+    {
+      "epoch": 2.427690582959641,
+      "grad_norm": 0.08153205909614925,
+      "learning_rate": 2.1371479223159862e-05,
+      "loss": 0.2433,
+      "step": 4331
+    },
+    {
+      "epoch": 2.428251121076233,
+      "grad_norm": 0.0786338410238144,
+      "learning_rate": 2.1331191218731783e-05,
+      "loss": 0.2348,
+      "step": 4332
+    },
+    {
+      "epoch": 2.4288116591928253,
+      "grad_norm": 0.07898342934560665,
+      "learning_rate": 2.1290936689979047e-05,
+      "loss": 0.2588,
+      "step": 4333
+    },
+    {
+      "epoch": 2.429372197309417,
+      "grad_norm": 0.08115594101252847,
+      "learning_rate": 2.125071565403104e-05,
+      "loss": 0.2553,
+      "step": 4334
+    },
+    {
+      "epoch": 2.429932735426009,
+      "grad_norm": 0.08148196147954538,
+      "learning_rate": 2.1210528128002904e-05,
+      "loss": 0.2451,
+      "step": 4335
+    },
+    {
+      "epoch": 2.430493273542601,
+      "grad_norm": 0.07611515539965039,
+      "learning_rate": 2.1170374128995507e-05,
+      "loss": 0.2385,
+      "step": 4336
+    },
+    {
+      "epoch": 2.4310538116591927,
+      "grad_norm": 0.07760886178051588,
+      "learning_rate": 2.1130253674095435e-05,
+      "loss": 0.2397,
+      "step": 4337
+    },
+    {
+      "epoch": 2.431614349775785,
+      "grad_norm": 0.07889059788244317,
+      "learning_rate": 2.1090166780374975e-05,
+      "loss": 0.2472,
+      "step": 4338
+    },
+    {
+      "epoch": 2.432174887892377,
+      "grad_norm": 0.07976461126208553,
+      "learning_rate": 2.105011346489224e-05,
+      "loss": 0.2487,
+      "step": 4339
+    },
+    {
+      "epoch": 2.4327354260089686,
+      "grad_norm": 0.07951958202780855,
+      "learning_rate": 2.1010093744690908e-05,
+      "loss": 0.2333,
+      "step": 4340
+    },
+    {
+      "epoch": 2.4332959641255605,
+      "grad_norm": 0.07962954473906425,
+      "learning_rate": 2.0970107636800495e-05,
+      "loss": 0.2429,
+      "step": 4341
+    },
+    {
+      "epoch": 2.4338565022421523,
+      "grad_norm": 0.0794323731347504,
+      "learning_rate": 2.093015515823612e-05,
+      "loss": 0.2413,
+      "step": 4342
+    },
+    {
+      "epoch": 2.4344170403587446,
+      "grad_norm": 0.07952549060784866,
+      "learning_rate": 2.0890236325998635e-05,
+      "loss": 0.2373,
+      "step": 4343
+    },
+    {
+      "epoch": 2.4349775784753365,
+      "grad_norm": 0.07917474803448454,
+      "learning_rate": 2.0850351157074598e-05,
+      "loss": 0.2429,
+      "step": 4344
+    },
+    {
+      "epoch": 2.4355381165919283,
+      "grad_norm": 0.07913531651724788,
+      "learning_rate": 2.0810499668436166e-05,
+      "loss": 0.235,
+      "step": 4345
+    },
+    {
+      "epoch": 2.43609865470852,
+      "grad_norm": 0.08034038891466093,
+      "learning_rate": 2.0770681877041253e-05,
+      "loss": 0.2482,
+      "step": 4346
+    },
+    {
+      "epoch": 2.436659192825112,
+      "grad_norm": 0.07891006005804807,
+      "learning_rate": 2.0730897799833348e-05,
+      "loss": 0.2391,
+      "step": 4347
+    },
+    {
+      "epoch": 2.437219730941704,
+      "grad_norm": 0.08217772709245133,
+      "learning_rate": 2.0691147453741687e-05,
+      "loss": 0.2473,
+      "step": 4348
+    },
+    {
+      "epoch": 2.437780269058296,
+      "grad_norm": 0.07853254914239965,
+      "learning_rate": 2.0651430855681064e-05,
+      "loss": 0.2496,
+      "step": 4349
+    },
+    {
+      "epoch": 2.438340807174888,
+      "grad_norm": 0.08087459568714833,
+      "learning_rate": 2.0611748022551936e-05,
+      "loss": 0.2319,
+      "step": 4350
+    },
+    {
+      "epoch": 2.43890134529148,
+      "grad_norm": 0.08140349728833127,
+      "learning_rate": 2.057209897124043e-05,
+      "loss": 0.2495,
+      "step": 4351
+    },
+    {
+      "epoch": 2.4394618834080717,
+      "grad_norm": 0.08125526010914347,
+      "learning_rate": 2.0532483718618267e-05,
+      "loss": 0.2531,
+      "step": 4352
+    },
+    {
+      "epoch": 2.4400224215246635,
+      "grad_norm": 0.07734467440134396,
+      "learning_rate": 2.0492902281542836e-05,
+      "loss": 0.2414,
+      "step": 4353
+    },
+    {
+      "epoch": 2.4405829596412554,
+      "grad_norm": 0.07660232671184296,
+      "learning_rate": 2.045335467685703e-05,
+      "loss": 0.2381,
+      "step": 4354
+    },
+    {
+      "epoch": 2.4411434977578477,
+      "grad_norm": 0.08019024531719846,
+      "learning_rate": 2.041384092138946e-05,
+      "loss": 0.2464,
+      "step": 4355
+    },
+    {
+      "epoch": 2.4417040358744395,
+      "grad_norm": 0.07706511284846577,
+      "learning_rate": 2.037436103195426e-05,
+      "loss": 0.2318,
+      "step": 4356
+    },
+    {
+      "epoch": 2.4422645739910314,
+      "grad_norm": 0.08246826832148706,
+      "learning_rate": 2.0334915025351142e-05,
+      "loss": 0.2538,
+      "step": 4357
+    },
+    {
+      "epoch": 2.442825112107623,
+      "grad_norm": 0.08263485684952807,
+      "learning_rate": 2.0295502918365472e-05,
+      "loss": 0.2487,
+      "step": 4358
+    },
+    {
+      "epoch": 2.443385650224215,
+      "grad_norm": 0.07687059585038453,
+      "learning_rate": 2.0256124727768143e-05,
+      "loss": 0.2405,
+      "step": 4359
+    },
+    {
+      "epoch": 2.4439461883408073,
+      "grad_norm": 0.0765570939471235,
+      "learning_rate": 2.0216780470315655e-05,
+      "loss": 0.2242,
+      "step": 4360
+    },
+    {
+      "epoch": 2.444506726457399,
+      "grad_norm": 0.07908731441849903,
+      "learning_rate": 2.017747016274999e-05,
+      "loss": 0.2421,
+      "step": 4361
+    },
+    {
+      "epoch": 2.445067264573991,
+      "grad_norm": 0.08001846364376108,
+      "learning_rate": 2.013819382179878e-05,
+      "loss": 0.2452,
+      "step": 4362
+    },
+    {
+      "epoch": 2.445627802690583,
+      "grad_norm": 0.07592919032742099,
+      "learning_rate": 2.009895146417512e-05,
+      "loss": 0.2372,
+      "step": 4363
+    },
+    {
+      "epoch": 2.4461883408071747,
+      "grad_norm": 0.07982023578857524,
+      "learning_rate": 2.0059743106577654e-05,
+      "loss": 0.2551,
+      "step": 4364
+    },
+    {
+      "epoch": 2.446748878923767,
+      "grad_norm": 0.08070692659023758,
+      "learning_rate": 2.002056876569066e-05,
+      "loss": 0.2476,
+      "step": 4365
+    },
+    {
+      "epoch": 2.447309417040359,
+      "grad_norm": 0.08002500517767909,
+      "learning_rate": 1.9981428458183792e-05,
+      "loss": 0.2384,
+      "step": 4366
+    },
+    {
+      "epoch": 2.4478699551569507,
+      "grad_norm": 0.07873885223419161,
+      "learning_rate": 1.9942322200712315e-05,
+      "loss": 0.2433,
+      "step": 4367
+    },
+    {
+      "epoch": 2.4484304932735426,
+      "grad_norm": 0.08066244265538584,
+      "learning_rate": 1.9903250009916997e-05,
+      "loss": 0.2543,
+      "step": 4368
+    },
+    {
+      "epoch": 2.4489910313901344,
+      "grad_norm": 0.08182267861637552,
+      "learning_rate": 1.9864211902424123e-05,
+      "loss": 0.2464,
+      "step": 4369
+    },
+    {
+      "epoch": 2.4495515695067267,
+      "grad_norm": 0.07793138319209078,
+      "learning_rate": 1.982520789484541e-05,
+      "loss": 0.2365,
+      "step": 4370
+    },
+    {
+      "epoch": 2.4501121076233185,
+      "grad_norm": 0.07729500337141762,
+      "learning_rate": 1.978623800377809e-05,
+      "loss": 0.238,
+      "step": 4371
+    },
+    {
+      "epoch": 2.4506726457399104,
+      "grad_norm": 0.0785283191450988,
+      "learning_rate": 1.9747302245804945e-05,
+      "loss": 0.2551,
+      "step": 4372
+    },
+    {
+      "epoch": 2.4512331838565022,
+      "grad_norm": 0.07754012827939691,
+      "learning_rate": 1.9708400637494105e-05,
+      "loss": 0.2407,
+      "step": 4373
+    },
+    {
+      "epoch": 2.451793721973094,
+      "grad_norm": 0.07908752512751777,
+      "learning_rate": 1.9669533195399316e-05,
+      "loss": 0.2491,
+      "step": 4374
+    },
+    {
+      "epoch": 2.452354260089686,
+      "grad_norm": 0.07769777019386097,
+      "learning_rate": 1.963069993605964e-05,
+      "loss": 0.2495,
+      "step": 4375
+    },
+    {
+      "epoch": 2.452914798206278,
+      "grad_norm": 0.07953808071860707,
+      "learning_rate": 1.9591900875999703e-05,
+      "loss": 0.2493,
+      "step": 4376
+    },
+    {
+      "epoch": 2.45347533632287,
+      "grad_norm": 0.0767251163133952,
+      "learning_rate": 1.955313603172957e-05,
+      "loss": 0.2366,
+      "step": 4377
+    },
+    {
+      "epoch": 2.454035874439462,
+      "grad_norm": 0.0772876732653098,
+      "learning_rate": 1.9514405419744654e-05,
+      "loss": 0.2377,
+      "step": 4378
+    },
+    {
+      "epoch": 2.4545964125560538,
+      "grad_norm": 0.07802199616568897,
+      "learning_rate": 1.9475709056525905e-05,
+      "loss": 0.2407,
+      "step": 4379
+    },
+    {
+      "epoch": 2.4551569506726456,
+      "grad_norm": 0.08125062605646792,
+      "learning_rate": 1.943704695853963e-05,
+      "loss": 0.2457,
+      "step": 4380
+    },
+    {
+      "epoch": 2.4557174887892375,
+      "grad_norm": 0.07950019725415046,
+      "learning_rate": 1.939841914223761e-05,
+      "loss": 0.2322,
+      "step": 4381
+    },
+    {
+      "epoch": 2.4562780269058297,
+      "grad_norm": 0.07970714171355352,
+      "learning_rate": 1.935982562405698e-05,
+      "loss": 0.2428,
+      "step": 4382
+    },
+    {
+      "epoch": 2.4568385650224216,
+      "grad_norm": 0.07860283694525858,
+      "learning_rate": 1.932126642042035e-05,
+      "loss": 0.2372,
+      "step": 4383
+    },
+    {
+      "epoch": 2.4573991031390134,
+      "grad_norm": 0.0799753943160765,
+      "learning_rate": 1.9282741547735637e-05,
+      "loss": 0.2363,
+      "step": 4384
+    },
+    {
+      "epoch": 2.4579596412556053,
+      "grad_norm": 0.07884053181074349,
+      "learning_rate": 1.9244251022396233e-05,
+      "loss": 0.2486,
+      "step": 4385
+    },
+    {
+      "epoch": 2.458520179372197,
+      "grad_norm": 0.0818453974140935,
+      "learning_rate": 1.9205794860780914e-05,
+      "loss": 0.2538,
+      "step": 4386
+    },
+    {
+      "epoch": 2.4590807174887894,
+      "grad_norm": 0.07779196140984937,
+      "learning_rate": 1.9167373079253727e-05,
+      "loss": 0.2409,
+      "step": 4387
+    },
+    {
+      "epoch": 2.4596412556053813,
+      "grad_norm": 0.08201345727344617,
+      "learning_rate": 1.9128985694164237e-05,
+      "loss": 0.2426,
+      "step": 4388
+    },
+    {
+      "epoch": 2.460201793721973,
+      "grad_norm": 0.08178188182088565,
+      "learning_rate": 1.909063272184727e-05,
+      "loss": 0.2403,
+      "step": 4389
+    },
+    {
+      "epoch": 2.460762331838565,
+      "grad_norm": 0.07567328625492045,
+      "learning_rate": 1.9052314178623008e-05,
+      "loss": 0.236,
+      "step": 4390
+    },
+    {
+      "epoch": 2.461322869955157,
+      "grad_norm": 0.07804704174016891,
+      "learning_rate": 1.901403008079704e-05,
+      "loss": 0.2406,
+      "step": 4391
+    },
+    {
+      "epoch": 2.461883408071749,
+      "grad_norm": 0.08055788967776399,
+      "learning_rate": 1.8975780444660273e-05,
+      "loss": 0.2454,
+      "step": 4392
+    },
+    {
+      "epoch": 2.462443946188341,
+      "grad_norm": 0.07939038698067213,
+      "learning_rate": 1.8937565286488966e-05,
+      "loss": 0.2451,
+      "step": 4393
+    },
+    {
+      "epoch": 2.463004484304933,
+      "grad_norm": 0.07808079204750822,
+      "learning_rate": 1.8899384622544646e-05,
+      "loss": 0.2437,
+      "step": 4394
+    },
+    {
+      "epoch": 2.4635650224215246,
+      "grad_norm": 0.07847211429675839,
+      "learning_rate": 1.8861238469074248e-05,
+      "loss": 0.249,
+      "step": 4395
+    },
+    {
+      "epoch": 2.4641255605381165,
+      "grad_norm": 0.08213501625298139,
+      "learning_rate": 1.8823126842309956e-05,
+      "loss": 0.2408,
+      "step": 4396
+    },
+    {
+      "epoch": 2.4646860986547083,
+      "grad_norm": 0.07848382733862806,
+      "learning_rate": 1.8785049758469254e-05,
+      "loss": 0.2408,
+      "step": 4397
+    },
+    {
+      "epoch": 2.4652466367713006,
+      "grad_norm": 0.07943947355086714,
+      "learning_rate": 1.8747007233755e-05,
+      "loss": 0.2443,
+      "step": 4398
+    },
+    {
+      "epoch": 2.4658071748878925,
+      "grad_norm": 0.08090414909413521,
+      "learning_rate": 1.8708999284355266e-05,
+      "loss": 0.246,
+      "step": 4399
+    },
+    {
+      "epoch": 2.4663677130044843,
+      "grad_norm": 0.07786272828440925,
+      "learning_rate": 1.8671025926443465e-05,
+      "loss": 0.2367,
+      "step": 4400
+    },
+    {
+      "epoch": 2.466928251121076,
+      "grad_norm": 0.08054905391377483,
+      "learning_rate": 1.8633087176178276e-05,
+      "loss": 0.2455,
+      "step": 4401
+    },
+    {
+      "epoch": 2.467488789237668,
+      "grad_norm": 0.08390852101212544,
+      "learning_rate": 1.8595183049703668e-05,
+      "loss": 0.2475,
+      "step": 4402
+    },
+    {
+      "epoch": 2.46804932735426,
+      "grad_norm": 0.08319907705555271,
+      "learning_rate": 1.8557313563148847e-05,
+      "loss": 0.251,
+      "step": 4403
+    },
+    {
+      "epoch": 2.468609865470852,
+      "grad_norm": 0.07819796437479647,
+      "learning_rate": 1.8519478732628247e-05,
+      "loss": 0.2385,
+      "step": 4404
+    },
+    {
+      "epoch": 2.469170403587444,
+      "grad_norm": 0.08300631738946596,
+      "learning_rate": 1.8481678574241646e-05,
+      "loss": 0.2566,
+      "step": 4405
+    },
+    {
+      "epoch": 2.469730941704036,
+      "grad_norm": 0.07987810854223557,
+      "learning_rate": 1.8443913104073983e-05,
+      "loss": 0.2356,
+      "step": 4406
+    },
+    {
+      "epoch": 2.4702914798206277,
+      "grad_norm": 0.07729763414437077,
+      "learning_rate": 1.840618233819552e-05,
+      "loss": 0.226,
+      "step": 4407
+    },
+    {
+      "epoch": 2.4708520179372195,
+      "grad_norm": 0.0788044560119795,
+      "learning_rate": 1.8368486292661657e-05,
+      "loss": 0.2546,
+      "step": 4408
+    },
+    {
+      "epoch": 2.471412556053812,
+      "grad_norm": 0.07870547075686708,
+      "learning_rate": 1.8330824983513072e-05,
+      "loss": 0.2271,
+      "step": 4409
+    },
+    {
+      "epoch": 2.4719730941704037,
+      "grad_norm": 0.07736564087421319,
+      "learning_rate": 1.829319842677569e-05,
+      "loss": 0.2381,
+      "step": 4410
+    },
+    {
+      "epoch": 2.4725336322869955,
+      "grad_norm": 0.07917595786659042,
+      "learning_rate": 1.8255606638460576e-05,
+      "loss": 0.2427,
+      "step": 4411
+    },
+    {
+      "epoch": 2.4730941704035874,
+      "grad_norm": 0.07929542482453235,
+      "learning_rate": 1.8218049634564082e-05,
+      "loss": 0.2315,
+      "step": 4412
+    },
+    {
+      "epoch": 2.473654708520179,
+      "grad_norm": 0.08353627003581249,
+      "learning_rate": 1.818052743106766e-05,
+      "loss": 0.2601,
+      "step": 4413
+    },
+    {
+      "epoch": 2.4742152466367715,
+      "grad_norm": 0.07782654990584914,
+      "learning_rate": 1.8143040043938054e-05,
+      "loss": 0.2491,
+      "step": 4414
+    },
+    {
+      "epoch": 2.4747757847533634,
+      "grad_norm": 0.07936809537464609,
+      "learning_rate": 1.8105587489127106e-05,
+      "loss": 0.2323,
+      "step": 4415
+    },
+    {
+      "epoch": 2.475336322869955,
+      "grad_norm": 0.07816231296613192,
+      "learning_rate": 1.806816978257192e-05,
+      "loss": 0.241,
+      "step": 4416
+    },
+    {
+      "epoch": 2.475896860986547,
+      "grad_norm": 0.07749425776392747,
+      "learning_rate": 1.8030786940194688e-05,
+      "loss": 0.243,
+      "step": 4417
+    },
+    {
+      "epoch": 2.476457399103139,
+      "grad_norm": 0.08105678697332463,
+      "learning_rate": 1.799343897790282e-05,
+      "loss": 0.244,
+      "step": 4418
+    },
+    {
+      "epoch": 2.477017937219731,
+      "grad_norm": 0.07967362420984159,
+      "learning_rate": 1.7956125911588893e-05,
+      "loss": 0.2442,
+      "step": 4419
+    },
+    {
+      "epoch": 2.477578475336323,
+      "grad_norm": 0.07924519560102106,
+      "learning_rate": 1.7918847757130575e-05,
+      "loss": 0.2378,
+      "step": 4420
+    },
+    {
+      "epoch": 2.478139013452915,
+      "grad_norm": 0.07848148392860445,
+      "learning_rate": 1.788160453039075e-05,
+      "loss": 0.2345,
+      "step": 4421
+    },
+    {
+      "epoch": 2.4786995515695067,
+      "grad_norm": 0.07788126157653065,
+      "learning_rate": 1.7844396247217354e-05,
+      "loss": 0.2431,
+      "step": 4422
+    },
+    {
+      "epoch": 2.4792600896860986,
+      "grad_norm": 0.07936994105987126,
+      "learning_rate": 1.7807222923443567e-05,
+      "loss": 0.2478,
+      "step": 4423
+    },
+    {
+      "epoch": 2.4798206278026904,
+      "grad_norm": 0.07983013467764592,
+      "learning_rate": 1.7770084574887567e-05,
+      "loss": 0.2601,
+      "step": 4424
+    },
+    {
+      "epoch": 2.4803811659192827,
+      "grad_norm": 0.08100937582379807,
+      "learning_rate": 1.773298121735275e-05,
+      "loss": 0.242,
+      "step": 4425
+    },
+    {
+      "epoch": 2.4809417040358746,
+      "grad_norm": 0.07860663922366933,
+      "learning_rate": 1.7695912866627595e-05,
+      "loss": 0.2522,
+      "step": 4426
+    },
+    {
+      "epoch": 2.4815022421524664,
+      "grad_norm": 0.07825753850876958,
+      "learning_rate": 1.7658879538485628e-05,
+      "loss": 0.239,
+      "step": 4427
+    },
+    {
+      "epoch": 2.4820627802690582,
+      "grad_norm": 0.0841305919680388,
+      "learning_rate": 1.762188124868557e-05,
+      "loss": 0.2531,
+      "step": 4428
+    },
+    {
+      "epoch": 2.48262331838565,
+      "grad_norm": 0.08114721060009097,
+      "learning_rate": 1.758491801297114e-05,
+      "loss": 0.241,
+      "step": 4429
+    },
+    {
+      "epoch": 2.483183856502242,
+      "grad_norm": 0.07930211089835233,
+      "learning_rate": 1.7547989847071178e-05,
+      "loss": 0.2486,
+      "step": 4430
+    },
+    {
+      "epoch": 2.4837443946188342,
+      "grad_norm": 0.07808225378957155,
+      "learning_rate": 1.7511096766699643e-05,
+      "loss": 0.235,
+      "step": 4431
+    },
+    {
+      "epoch": 2.484304932735426,
+      "grad_norm": 0.08022252134741344,
+      "learning_rate": 1.7474238787555476e-05,
+      "loss": 0.2421,
+      "step": 4432
+    },
+    {
+      "epoch": 2.484865470852018,
+      "grad_norm": 0.08119256917900608,
+      "learning_rate": 1.7437415925322743e-05,
+      "loss": 0.2342,
+      "step": 4433
+    },
+    {
+      "epoch": 2.4854260089686098,
+      "grad_norm": 0.07989719398209497,
+      "learning_rate": 1.7400628195670578e-05,
+      "loss": 0.2607,
+      "step": 4434
+    },
+    {
+      "epoch": 2.4859865470852016,
+      "grad_norm": 0.0833588246480297,
+      "learning_rate": 1.7363875614253135e-05,
+      "loss": 0.2464,
+      "step": 4435
+    },
+    {
+      "epoch": 2.486547085201794,
+      "grad_norm": 0.07816092133763146,
+      "learning_rate": 1.7327158196709613e-05,
+      "loss": 0.2388,
+      "step": 4436
+    },
+    {
+      "epoch": 2.4871076233183858,
+      "grad_norm": 0.08122459364730421,
+      "learning_rate": 1.729047595866422e-05,
+      "loss": 0.2383,
+      "step": 4437
+    },
+    {
+      "epoch": 2.4876681614349776,
+      "grad_norm": 0.0815383675823079,
+      "learning_rate": 1.725382891572629e-05,
+      "loss": 0.2602,
+      "step": 4438
+    },
+    {
+      "epoch": 2.4882286995515694,
+      "grad_norm": 0.0802074151304083,
+      "learning_rate": 1.7217217083490044e-05,
+      "loss": 0.2468,
+      "step": 4439
+    },
+    {
+      "epoch": 2.4887892376681613,
+      "grad_norm": 0.07881996617835439,
+      "learning_rate": 1.7180640477534847e-05,
+      "loss": 0.2472,
+      "step": 4440
+    },
+    {
+      "epoch": 2.4893497757847536,
+      "grad_norm": 0.07896209214879413,
+      "learning_rate": 1.7144099113424984e-05,
+      "loss": 0.2496,
+      "step": 4441
+    },
+    {
+      "epoch": 2.4899103139013454,
+      "grad_norm": 0.07970756377176035,
+      "learning_rate": 1.7107593006709798e-05,
+      "loss": 0.2469,
+      "step": 4442
+    },
+    {
+      "epoch": 2.4904708520179373,
+      "grad_norm": 0.07799921125634339,
+      "learning_rate": 1.7071122172923636e-05,
+      "loss": 0.2391,
+      "step": 4443
+    },
+    {
+      "epoch": 2.491031390134529,
+      "grad_norm": 0.07928685093016447,
+      "learning_rate": 1.7034686627585762e-05,
+      "loss": 0.241,
+      "step": 4444
+    },
+    {
+      "epoch": 2.491591928251121,
+      "grad_norm": 0.07984796699906008,
+      "learning_rate": 1.6998286386200503e-05,
+      "loss": 0.2365,
+      "step": 4445
+    },
+    {
+      "epoch": 2.492152466367713,
+      "grad_norm": 0.08129675139741374,
+      "learning_rate": 1.6961921464257114e-05,
+      "loss": 0.24,
+      "step": 4446
+    },
+    {
+      "epoch": 2.492713004484305,
+      "grad_norm": 0.0791719916919716,
+      "learning_rate": 1.6925591877229863e-05,
+      "loss": 0.2448,
+      "step": 4447
+    },
+    {
+      "epoch": 2.493273542600897,
+      "grad_norm": 0.08149933503813549,
+      "learning_rate": 1.6889297640577915e-05,
+      "loss": 0.2466,
+      "step": 4448
+    },
+    {
+      "epoch": 2.493834080717489,
+      "grad_norm": 0.08132523919018882,
+      "learning_rate": 1.6853038769745467e-05,
+      "loss": 0.2304,
+      "step": 4449
+    },
+    {
+      "epoch": 2.4943946188340806,
+      "grad_norm": 0.07958570040183402,
+      "learning_rate": 1.681681528016166e-05,
+      "loss": 0.2495,
+      "step": 4450
+    },
+    {
+      "epoch": 2.4949551569506725,
+      "grad_norm": 0.07854599401912843,
+      "learning_rate": 1.6780627187240493e-05,
+      "loss": 0.2408,
+      "step": 4451
+    },
+    {
+      "epoch": 2.4955156950672643,
+      "grad_norm": 0.07854173988150537,
+      "learning_rate": 1.6744474506381023e-05,
+      "loss": 0.2357,
+      "step": 4452
+    },
+    {
+      "epoch": 2.4960762331838566,
+      "grad_norm": 0.07606243551369524,
+      "learning_rate": 1.670835725296713e-05,
+      "loss": 0.2399,
+      "step": 4453
+    },
+    {
+      "epoch": 2.4966367713004485,
+      "grad_norm": 0.08173527176087826,
+      "learning_rate": 1.667227544236771e-05,
+      "loss": 0.2392,
+      "step": 4454
+    },
+    {
+      "epoch": 2.4971973094170403,
+      "grad_norm": 0.08425758700292559,
+      "learning_rate": 1.6636229089936496e-05,
+      "loss": 0.251,
+      "step": 4455
+    },
+    {
+      "epoch": 2.497757847533632,
+      "grad_norm": 0.0786189663188184,
+      "learning_rate": 1.660021821101222e-05,
+      "loss": 0.2376,
+      "step": 4456
+    },
+    {
+      "epoch": 2.498318385650224,
+      "grad_norm": 0.07850242465631987,
+      "learning_rate": 1.6564242820918418e-05,
+      "loss": 0.2444,
+      "step": 4457
+    },
+    {
+      "epoch": 2.4988789237668163,
+      "grad_norm": 0.08449997838843436,
+      "learning_rate": 1.65283029349636e-05,
+      "loss": 0.2484,
+      "step": 4458
+    },
+    {
+      "epoch": 2.499439461883408,
+      "grad_norm": 0.07851891897220185,
+      "learning_rate": 1.649239856844117e-05,
+      "loss": 0.2407,
+      "step": 4459
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.0765513225178551,
+      "learning_rate": 1.6456529736629345e-05,
+      "loss": 0.2254,
+      "step": 4460
+    },
+    {
+      "epoch": 2.500560538116592,
+      "grad_norm": 0.08363321984849857,
+      "learning_rate": 1.6420696454791328e-05,
+      "loss": 0.2535,
+      "step": 4461
+    },
+    {
+      "epoch": 2.5011210762331837,
+      "grad_norm": 0.07737282235847463,
+      "learning_rate": 1.638489873817508e-05,
+      "loss": 0.2353,
+      "step": 4462
+    },
+    {
+      "epoch": 2.501681614349776,
+      "grad_norm": 0.08120031383422203,
+      "learning_rate": 1.6349136602013527e-05,
+      "loss": 0.2469,
+      "step": 4463
+    },
+    {
+      "epoch": 2.502242152466368,
+      "grad_norm": 0.07968539336795967,
+      "learning_rate": 1.6313410061524393e-05,
+      "loss": 0.2435,
+      "step": 4464
+    },
+    {
+      "epoch": 2.5028026905829597,
+      "grad_norm": 0.07712043740829289,
+      "learning_rate": 1.627771913191024e-05,
+      "loss": 0.2362,
+      "step": 4465
+    },
+    {
+      "epoch": 2.5033632286995515,
+      "grad_norm": 0.07857567724201052,
+      "learning_rate": 1.6242063828358544e-05,
+      "loss": 0.2334,
+      "step": 4466
+    },
+    {
+      "epoch": 2.5039237668161434,
+      "grad_norm": 0.08151544220668971,
+      "learning_rate": 1.620644416604159e-05,
+      "loss": 0.249,
+      "step": 4467
+    },
+    {
+      "epoch": 2.5044843049327357,
+      "grad_norm": 0.07972918683534676,
+      "learning_rate": 1.617086016011652e-05,
+      "loss": 0.2444,
+      "step": 4468
+    },
+    {
+      "epoch": 2.5050448430493275,
+      "grad_norm": 0.07812113755382295,
+      "learning_rate": 1.6135311825725208e-05,
+      "loss": 0.2418,
+      "step": 4469
+    },
+    {
+      "epoch": 2.5056053811659194,
+      "grad_norm": 0.07748075153834869,
+      "learning_rate": 1.609979917799449e-05,
+      "loss": 0.2316,
+      "step": 4470
+    },
+    {
+      "epoch": 2.506165919282511,
+      "grad_norm": 0.08065539286292779,
+      "learning_rate": 1.60643222320359e-05,
+      "loss": 0.2405,
+      "step": 4471
+    },
+    {
+      "epoch": 2.506726457399103,
+      "grad_norm": 0.08030272895310879,
+      "learning_rate": 1.6028881002945818e-05,
+      "loss": 0.2536,
+      "step": 4472
+    },
+    {
+      "epoch": 2.5072869955156953,
+      "grad_norm": 0.07938423577475674,
+      "learning_rate": 1.5993475505805467e-05,
+      "loss": 0.2413,
+      "step": 4473
+    },
+    {
+      "epoch": 2.5078475336322867,
+      "grad_norm": 0.07728689118560435,
+      "learning_rate": 1.5958105755680795e-05,
+      "loss": 0.2467,
+      "step": 4474
+    },
+    {
+      "epoch": 2.508408071748879,
+      "grad_norm": 0.08172159601220094,
+      "learning_rate": 1.5922771767622592e-05,
+      "loss": 0.2443,
+      "step": 4475
+    },
+    {
+      "epoch": 2.508968609865471,
+      "grad_norm": 0.08162774951123969,
+      "learning_rate": 1.588747355666642e-05,
+      "loss": 0.2461,
+      "step": 4476
+    },
+    {
+      "epoch": 2.5095291479820627,
+      "grad_norm": 0.08042985764171678,
+      "learning_rate": 1.5852211137832583e-05,
+      "loss": 0.2419,
+      "step": 4477
+    },
+    {
+      "epoch": 2.5100896860986546,
+      "grad_norm": 0.08105304381057965,
+      "learning_rate": 1.5816984526126222e-05,
+      "loss": 0.2359,
+      "step": 4478
+    },
+    {
+      "epoch": 2.5106502242152464,
+      "grad_norm": 0.07763512135806293,
+      "learning_rate": 1.5781793736537143e-05,
+      "loss": 0.2441,
+      "step": 4479
+    },
+    {
+      "epoch": 2.5112107623318387,
+      "grad_norm": 0.08043991943257807,
+      "learning_rate": 1.5746638784040025e-05,
+      "loss": 0.2343,
+      "step": 4480
+    },
+    {
+      "epoch": 2.5117713004484306,
+      "grad_norm": 0.07809864052161128,
+      "learning_rate": 1.5711519683594188e-05,
+      "loss": 0.2352,
+      "step": 4481
+    },
+    {
+      "epoch": 2.5123318385650224,
+      "grad_norm": 0.07785867561014251,
+      "learning_rate": 1.567643645014376e-05,
+      "loss": 0.2389,
+      "step": 4482
+    },
+    {
+      "epoch": 2.5128923766816142,
+      "grad_norm": 0.07777156654471881,
+      "learning_rate": 1.564138909861762e-05,
+      "loss": 0.23,
+      "step": 4483
+    },
+    {
+      "epoch": 2.513452914798206,
+      "grad_norm": 0.07725422661411521,
+      "learning_rate": 1.5606377643929304e-05,
+      "loss": 0.2295,
+      "step": 4484
+    },
+    {
+      "epoch": 2.5140134529147984,
+      "grad_norm": 0.07834660994788373,
+      "learning_rate": 1.5571402100977163e-05,
+      "loss": 0.2324,
+      "step": 4485
+    },
+    {
+      "epoch": 2.5145739910313902,
+      "grad_norm": 0.0805634437751587,
+      "learning_rate": 1.5536462484644187e-05,
+      "loss": 0.2349,
+      "step": 4486
+    },
+    {
+      "epoch": 2.515134529147982,
+      "grad_norm": 0.07766852791708763,
+      "learning_rate": 1.5501558809798154e-05,
+      "loss": 0.2266,
+      "step": 4487
+    },
+    {
+      "epoch": 2.515695067264574,
+      "grad_norm": 0.08147062811926131,
+      "learning_rate": 1.5466691091291454e-05,
+      "loss": 0.2446,
+      "step": 4488
+    },
+    {
+      "epoch": 2.5162556053811658,
+      "grad_norm": 0.0795981439577369,
+      "learning_rate": 1.5431859343961284e-05,
+      "loss": 0.2267,
+      "step": 4489
+    },
+    {
+      "epoch": 2.516816143497758,
+      "grad_norm": 0.0814700308689247,
+      "learning_rate": 1.5397063582629445e-05,
+      "loss": 0.246,
+      "step": 4490
+    },
+    {
+      "epoch": 2.51737668161435,
+      "grad_norm": 0.07933582847866993,
+      "learning_rate": 1.5362303822102466e-05,
+      "loss": 0.2444,
+      "step": 4491
+    },
+    {
+      "epoch": 2.5179372197309418,
+      "grad_norm": 0.07714052608508398,
+      "learning_rate": 1.5327580077171587e-05,
+      "loss": 0.2359,
+      "step": 4492
+    },
+    {
+      "epoch": 2.5184977578475336,
+      "grad_norm": 0.0790934573509968,
+      "learning_rate": 1.5292892362612642e-05,
+      "loss": 0.2356,
+      "step": 4493
+    },
+    {
+      "epoch": 2.5190582959641254,
+      "grad_norm": 0.0803605935968346,
+      "learning_rate": 1.525824069318621e-05,
+      "loss": 0.2407,
+      "step": 4494
+    },
+    {
+      "epoch": 2.5196188340807177,
+      "grad_norm": 0.08167257081183345,
+      "learning_rate": 1.5223625083637471e-05,
+      "loss": 0.2335,
+      "step": 4495
+    },
+    {
+      "epoch": 2.520179372197309,
+      "grad_norm": 0.07919293163948786,
+      "learning_rate": 1.5189045548696323e-05,
+      "loss": 0.2418,
+      "step": 4496
+    },
+    {
+      "epoch": 2.5207399103139014,
+      "grad_norm": 0.07895571155816208,
+      "learning_rate": 1.5154502103077261e-05,
+      "loss": 0.2444,
+      "step": 4497
+    },
+    {
+      "epoch": 2.5213004484304933,
+      "grad_norm": 0.08321216825372495,
+      "learning_rate": 1.5119994761479427e-05,
+      "loss": 0.2598,
+      "step": 4498
+    },
+    {
+      "epoch": 2.521860986547085,
+      "grad_norm": 0.0804118459280637,
+      "learning_rate": 1.5085523538586632e-05,
+      "loss": 0.2378,
+      "step": 4499
+    },
+    {
+      "epoch": 2.522421524663677,
+      "grad_norm": 0.08262966477108134,
+      "learning_rate": 1.5051088449067285e-05,
+      "loss": 0.2532,
+      "step": 4500
+    },
+    {
+      "epoch": 2.522982062780269,
+      "grad_norm": 0.080991477269032,
+      "learning_rate": 1.5016689507574488e-05,
+      "loss": 0.2493,
+      "step": 4501
+    },
+    {
+      "epoch": 2.523542600896861,
+      "grad_norm": 0.08076073788813379,
+      "learning_rate": 1.4982326728745843e-05,
+      "loss": 0.2388,
+      "step": 4502
+    },
+    {
+      "epoch": 2.524103139013453,
+      "grad_norm": 0.08012089860702121,
+      "learning_rate": 1.4948000127203666e-05,
+      "loss": 0.2462,
+      "step": 4503
+    },
+    {
+      "epoch": 2.524663677130045,
+      "grad_norm": 0.07946309313108386,
+      "learning_rate": 1.4913709717554836e-05,
+      "loss": 0.2399,
+      "step": 4504
+    },
+    {
+      "epoch": 2.5252242152466366,
+      "grad_norm": 0.07700977642894151,
+      "learning_rate": 1.4879455514390816e-05,
+      "loss": 0.236,
+      "step": 4505
+    },
+    {
+      "epoch": 2.5257847533632285,
+      "grad_norm": 0.08192900961330732,
+      "learning_rate": 1.4845237532287704e-05,
+      "loss": 0.2569,
+      "step": 4506
+    },
+    {
+      "epoch": 2.526345291479821,
+      "grad_norm": 0.07732857927297841,
+      "learning_rate": 1.4811055785806138e-05,
+      "loss": 0.2301,
+      "step": 4507
+    },
+    {
+      "epoch": 2.5269058295964126,
+      "grad_norm": 0.07881332488983218,
+      "learning_rate": 1.4776910289491385e-05,
+      "loss": 0.2385,
+      "step": 4508
+    },
+    {
+      "epoch": 2.5274663677130045,
+      "grad_norm": 0.0798930794260508,
+      "learning_rate": 1.4742801057873257e-05,
+      "loss": 0.2475,
+      "step": 4509
+    },
+    {
+      "epoch": 2.5280269058295963,
+      "grad_norm": 0.08099140742978495,
+      "learning_rate": 1.4708728105466163e-05,
+      "loss": 0.2605,
+      "step": 4510
+    },
+    {
+      "epoch": 2.528587443946188,
+      "grad_norm": 0.08039335763500817,
+      "learning_rate": 1.467469144676904e-05,
+      "loss": 0.2542,
+      "step": 4511
+    },
+    {
+      "epoch": 2.5291479820627805,
+      "grad_norm": 0.08032503381937363,
+      "learning_rate": 1.4640691096265358e-05,
+      "loss": 0.235,
+      "step": 4512
+    },
+    {
+      "epoch": 2.5297085201793723,
+      "grad_norm": 0.07911745865309457,
+      "learning_rate": 1.460672706842323e-05,
+      "loss": 0.2452,
+      "step": 4513
+    },
+    {
+      "epoch": 2.530269058295964,
+      "grad_norm": 0.07906123109904009,
+      "learning_rate": 1.45727993776952e-05,
+      "loss": 0.2375,
+      "step": 4514
+    },
+    {
+      "epoch": 2.530829596412556,
+      "grad_norm": 0.08147152535880466,
+      "learning_rate": 1.4538908038518428e-05,
+      "loss": 0.2424,
+      "step": 4515
+    },
+    {
+      "epoch": 2.531390134529148,
+      "grad_norm": 0.07972598814904243,
+      "learning_rate": 1.4505053065314611e-05,
+      "loss": 0.2451,
+      "step": 4516
+    },
+    {
+      "epoch": 2.53195067264574,
+      "grad_norm": 0.07841395503999708,
+      "learning_rate": 1.4471234472489892e-05,
+      "loss": 0.2425,
+      "step": 4517
+    },
+    {
+      "epoch": 2.532511210762332,
+      "grad_norm": 0.08000426459048555,
+      "learning_rate": 1.4437452274435037e-05,
+      "loss": 0.2436,
+      "step": 4518
+    },
+    {
+      "epoch": 2.533071748878924,
+      "grad_norm": 0.08122733185769104,
+      "learning_rate": 1.4403706485525225e-05,
+      "loss": 0.2433,
+      "step": 4519
+    },
+    {
+      "epoch": 2.5336322869955157,
+      "grad_norm": 0.0786567013200513,
+      "learning_rate": 1.4369997120120227e-05,
+      "loss": 0.2234,
+      "step": 4520
+    },
+    {
+      "epoch": 2.5341928251121075,
+      "grad_norm": 0.07672949608216117,
+      "learning_rate": 1.4336324192564232e-05,
+      "loss": 0.2482,
+      "step": 4521
+    },
+    {
+      "epoch": 2.5347533632287,
+      "grad_norm": 0.07997039882139345,
+      "learning_rate": 1.4302687717186014e-05,
+      "loss": 0.2311,
+      "step": 4522
+    },
+    {
+      "epoch": 2.535313901345291,
+      "grad_norm": 0.07938003456651439,
+      "learning_rate": 1.4269087708298755e-05,
+      "loss": 0.2455,
+      "step": 4523
+    },
+    {
+      "epoch": 2.5358744394618835,
+      "grad_norm": 0.07928533776455994,
+      "learning_rate": 1.4235524180200166e-05,
+      "loss": 0.2432,
+      "step": 4524
+    },
+    {
+      "epoch": 2.5364349775784754,
+      "grad_norm": 0.08194063261809265,
+      "learning_rate": 1.4201997147172453e-05,
+      "loss": 0.2521,
+      "step": 4525
+    },
+    {
+      "epoch": 2.536995515695067,
+      "grad_norm": 0.07946429322564628,
+      "learning_rate": 1.4168506623482202e-05,
+      "loss": 0.2298,
+      "step": 4526
+    },
+    {
+      "epoch": 2.537556053811659,
+      "grad_norm": 0.08106337835210105,
+      "learning_rate": 1.4135052623380596e-05,
+      "loss": 0.2518,
+      "step": 4527
+    },
+    {
+      "epoch": 2.538116591928251,
+      "grad_norm": 0.07957732161781982,
+      "learning_rate": 1.4101635161103132e-05,
+      "loss": 0.2368,
+      "step": 4528
+    },
+    {
+      "epoch": 2.538677130044843,
+      "grad_norm": 0.07883091820953315,
+      "learning_rate": 1.4068254250869895e-05,
+      "loss": 0.2348,
+      "step": 4529
+    },
+    {
+      "epoch": 2.539237668161435,
+      "grad_norm": 0.08068664223577357,
+      "learning_rate": 1.4034909906885308e-05,
+      "loss": 0.2479,
+      "step": 4530
+    },
+    {
+      "epoch": 2.539798206278027,
+      "grad_norm": 0.0786028594120897,
+      "learning_rate": 1.4001602143338277e-05,
+      "loss": 0.2478,
+      "step": 4531
+    },
+    {
+      "epoch": 2.5403587443946187,
+      "grad_norm": 0.08154351406638181,
+      "learning_rate": 1.3968330974402166e-05,
+      "loss": 0.2276,
+      "step": 4532
+    },
+    {
+      "epoch": 2.5409192825112106,
+      "grad_norm": 0.07803866700309083,
+      "learning_rate": 1.3935096414234728e-05,
+      "loss": 0.2349,
+      "step": 4533
+    },
+    {
+      "epoch": 2.541479820627803,
+      "grad_norm": 0.08112742036593473,
+      "learning_rate": 1.390189847697818e-05,
+      "loss": 0.2348,
+      "step": 4534
+    },
+    {
+      "epoch": 2.5420403587443947,
+      "grad_norm": 0.07816987304461699,
+      "learning_rate": 1.3868737176759106e-05,
+      "loss": 0.2319,
+      "step": 4535
+    },
+    {
+      "epoch": 2.5426008968609866,
+      "grad_norm": 0.08162227289399063,
+      "learning_rate": 1.3835612527688536e-05,
+      "loss": 0.2415,
+      "step": 4536
+    },
+    {
+      "epoch": 2.5431614349775784,
+      "grad_norm": 0.07834022741575637,
+      "learning_rate": 1.3802524543861894e-05,
+      "loss": 0.2402,
+      "step": 4537
+    },
+    {
+      "epoch": 2.5437219730941703,
+      "grad_norm": 0.07705607029820112,
+      "learning_rate": 1.3769473239358965e-05,
+      "loss": 0.2384,
+      "step": 4538
+    },
+    {
+      "epoch": 2.5442825112107625,
+      "grad_norm": 0.08090779714416466,
+      "learning_rate": 1.3736458628244008e-05,
+      "loss": 0.2371,
+      "step": 4539
+    },
+    {
+      "epoch": 2.5448430493273544,
+      "grad_norm": 0.07997151891543049,
+      "learning_rate": 1.3703480724565577e-05,
+      "loss": 0.2497,
+      "step": 4540
+    },
+    {
+      "epoch": 2.5454035874439462,
+      "grad_norm": 0.0775203468999416,
+      "learning_rate": 1.3670539542356664e-05,
+      "loss": 0.2373,
+      "step": 4541
+    },
+    {
+      "epoch": 2.545964125560538,
+      "grad_norm": 0.07965788466092727,
+      "learning_rate": 1.3637635095634626e-05,
+      "loss": 0.2484,
+      "step": 4542
+    },
+    {
+      "epoch": 2.54652466367713,
+      "grad_norm": 0.07914207365379075,
+      "learning_rate": 1.3604767398401208e-05,
+      "loss": 0.2377,
+      "step": 4543
+    },
+    {
+      "epoch": 2.547085201793722,
+      "grad_norm": 0.08037478126061315,
+      "learning_rate": 1.3571936464642466e-05,
+      "loss": 0.2328,
+      "step": 4544
+    },
+    {
+      "epoch": 2.547645739910314,
+      "grad_norm": 0.08045814110102133,
+      "learning_rate": 1.353914230832881e-05,
+      "loss": 0.2465,
+      "step": 4545
+    },
+    {
+      "epoch": 2.548206278026906,
+      "grad_norm": 0.08201626906455045,
+      "learning_rate": 1.3506384943415084e-05,
+      "loss": 0.2508,
+      "step": 4546
+    },
+    {
+      "epoch": 2.5487668161434978,
+      "grad_norm": 0.07948458979602914,
+      "learning_rate": 1.3473664383840367e-05,
+      "loss": 0.2457,
+      "step": 4547
+    },
+    {
+      "epoch": 2.5493273542600896,
+      "grad_norm": 0.07907056490097471,
+      "learning_rate": 1.3440980643528144e-05,
+      "loss": 0.2385,
+      "step": 4548
+    },
+    {
+      "epoch": 2.5498878923766815,
+      "grad_norm": 0.07622766870336829,
+      "learning_rate": 1.3408333736386236e-05,
+      "loss": 0.2344,
+      "step": 4549
+    },
+    {
+      "epoch": 2.5504484304932733,
+      "grad_norm": 0.08030827437419959,
+      "learning_rate": 1.3375723676306783e-05,
+      "loss": 0.2343,
+      "step": 4550
+    },
+    {
+      "epoch": 2.5510089686098656,
+      "grad_norm": 0.07961449378768444,
+      "learning_rate": 1.3343150477166222e-05,
+      "loss": 0.2433,
+      "step": 4551
+    },
+    {
+      "epoch": 2.5515695067264574,
+      "grad_norm": 0.0806096282528269,
+      "learning_rate": 1.3310614152825273e-05,
+      "loss": 0.2318,
+      "step": 4552
+    },
+    {
+      "epoch": 2.5521300448430493,
+      "grad_norm": 0.08180752786629299,
+      "learning_rate": 1.3278114717129087e-05,
+      "loss": 0.2422,
+      "step": 4553
+    },
+    {
+      "epoch": 2.552690582959641,
+      "grad_norm": 0.0806910267206613,
+      "learning_rate": 1.3245652183906965e-05,
+      "loss": 0.2427,
+      "step": 4554
+    },
+    {
+      "epoch": 2.553251121076233,
+      "grad_norm": 0.08057267283346009,
+      "learning_rate": 1.3213226566972647e-05,
+      "loss": 0.2307,
+      "step": 4555
+    },
+    {
+      "epoch": 2.5538116591928253,
+      "grad_norm": 0.08034367523185901,
+      "learning_rate": 1.3180837880124041e-05,
+      "loss": 0.242,
+      "step": 4556
+    },
+    {
+      "epoch": 2.554372197309417,
+      "grad_norm": 0.07940310528470543,
+      "learning_rate": 1.3148486137143423e-05,
+      "loss": 0.2405,
+      "step": 4557
+    },
+    {
+      "epoch": 2.554932735426009,
+      "grad_norm": 0.0813902006779356,
+      "learning_rate": 1.3116171351797336e-05,
+      "loss": 0.2556,
+      "step": 4558
+    },
+    {
+      "epoch": 2.555493273542601,
+      "grad_norm": 0.08056421516881529,
+      "learning_rate": 1.3083893537836556e-05,
+      "loss": 0.2451,
+      "step": 4559
+    },
+    {
+      "epoch": 2.5560538116591927,
+      "grad_norm": 0.07870155170299742,
+      "learning_rate": 1.3051652708996177e-05,
+      "loss": 0.2381,
+      "step": 4560
+    },
+    {
+      "epoch": 2.556614349775785,
+      "grad_norm": 0.08021855892759801,
+      "learning_rate": 1.3019448878995499e-05,
+      "loss": 0.2356,
+      "step": 4561
+    },
+    {
+      "epoch": 2.557174887892377,
+      "grad_norm": 0.0792681406793524,
+      "learning_rate": 1.2987282061538164e-05,
+      "loss": 0.2424,
+      "step": 4562
+    },
+    {
+      "epoch": 2.5577354260089686,
+      "grad_norm": 0.08254078751573686,
+      "learning_rate": 1.2955152270311966e-05,
+      "loss": 0.2461,
+      "step": 4563
+    },
+    {
+      "epoch": 2.5582959641255605,
+      "grad_norm": 0.07926710899890296,
+      "learning_rate": 1.2923059518988979e-05,
+      "loss": 0.2397,
+      "step": 4564
+    },
+    {
+      "epoch": 2.5588565022421523,
+      "grad_norm": 0.08008160747822592,
+      "learning_rate": 1.2891003821225545e-05,
+      "loss": 0.2526,
+      "step": 4565
+    },
+    {
+      "epoch": 2.5594170403587446,
+      "grad_norm": 0.08247390082805088,
+      "learning_rate": 1.285898519066221e-05,
+      "loss": 0.2406,
+      "step": 4566
+    },
+    {
+      "epoch": 2.5599775784753365,
+      "grad_norm": 0.08026679579463669,
+      "learning_rate": 1.2827003640923784e-05,
+      "loss": 0.2374,
+      "step": 4567
+    },
+    {
+      "epoch": 2.5605381165919283,
+      "grad_norm": 0.08066291654705528,
+      "learning_rate": 1.2795059185619229e-05,
+      "loss": 0.2318,
+      "step": 4568
+    },
+    {
+      "epoch": 2.56109865470852,
+      "grad_norm": 0.07971569258124504,
+      "learning_rate": 1.2763151838341802e-05,
+      "loss": 0.239,
+      "step": 4569
+    },
+    {
+      "epoch": 2.561659192825112,
+      "grad_norm": 0.07783877390894758,
+      "learning_rate": 1.273128161266891e-05,
+      "loss": 0.238,
+      "step": 4570
+    },
+    {
+      "epoch": 2.5622197309417043,
+      "grad_norm": 0.07880881028458633,
+      "learning_rate": 1.2699448522162161e-05,
+      "loss": 0.2369,
+      "step": 4571
+    },
+    {
+      "epoch": 2.5627802690582957,
+      "grad_norm": 0.07789180805884843,
+      "learning_rate": 1.2667652580367417e-05,
+      "loss": 0.225,
+      "step": 4572
+    },
+    {
+      "epoch": 2.563340807174888,
+      "grad_norm": 0.07905274035098453,
+      "learning_rate": 1.2635893800814669e-05,
+      "loss": 0.2441,
+      "step": 4573
+    },
+    {
+      "epoch": 2.56390134529148,
+      "grad_norm": 0.08161497148242826,
+      "learning_rate": 1.260417219701815e-05,
+      "loss": 0.2481,
+      "step": 4574
+    },
+    {
+      "epoch": 2.5644618834080717,
+      "grad_norm": 0.08324032759367675,
+      "learning_rate": 1.2572487782476228e-05,
+      "loss": 0.251,
+      "step": 4575
+    },
+    {
+      "epoch": 2.5650224215246635,
+      "grad_norm": 0.0782085524442898,
+      "learning_rate": 1.2540840570671497e-05,
+      "loss": 0.2401,
+      "step": 4576
+    },
+    {
+      "epoch": 2.5655829596412554,
+      "grad_norm": 0.07768410537529105,
+      "learning_rate": 1.2509230575070686e-05,
+      "loss": 0.2539,
+      "step": 4577
+    },
+    {
+      "epoch": 2.5661434977578477,
+      "grad_norm": 0.08105766213368072,
+      "learning_rate": 1.2477657809124631e-05,
+      "loss": 0.242,
+      "step": 4578
+    },
+    {
+      "epoch": 2.5667040358744395,
+      "grad_norm": 0.07995084175421051,
+      "learning_rate": 1.2446122286268469e-05,
+      "loss": 0.2433,
+      "step": 4579
+    },
+    {
+      "epoch": 2.5672645739910314,
+      "grad_norm": 0.07746120906424067,
+      "learning_rate": 1.241462401992134e-05,
+      "loss": 0.2303,
+      "step": 4580
+    },
+    {
+      "epoch": 2.567825112107623,
+      "grad_norm": 0.07821609713443424,
+      "learning_rate": 1.238316302348661e-05,
+      "loss": 0.2362,
+      "step": 4581
+    },
+    {
+      "epoch": 2.568385650224215,
+      "grad_norm": 0.07921550521842298,
+      "learning_rate": 1.2351739310351795e-05,
+      "loss": 0.2365,
+      "step": 4582
+    },
+    {
+      "epoch": 2.5689461883408073,
+      "grad_norm": 0.07781707559113729,
+      "learning_rate": 1.2320352893888532e-05,
+      "loss": 0.2434,
+      "step": 4583
+    },
+    {
+      "epoch": 2.569506726457399,
+      "grad_norm": 0.07812556330316417,
+      "learning_rate": 1.2289003787452557e-05,
+      "loss": 0.2217,
+      "step": 4584
+    },
+    {
+      "epoch": 2.570067264573991,
+      "grad_norm": 0.07958740262817013,
+      "learning_rate": 1.2257692004383725e-05,
+      "loss": 0.2416,
+      "step": 4585
+    },
+    {
+      "epoch": 2.570627802690583,
+      "grad_norm": 0.0769627442836369,
+      "learning_rate": 1.2226417558006087e-05,
+      "loss": 0.2319,
+      "step": 4586
+    },
+    {
+      "epoch": 2.5711883408071747,
+      "grad_norm": 0.07679104601067815,
+      "learning_rate": 1.2195180461627698e-05,
+      "loss": 0.2337,
+      "step": 4587
+    },
+    {
+      "epoch": 2.571748878923767,
+      "grad_norm": 0.07473221010039457,
+      "learning_rate": 1.2163980728540835e-05,
+      "loss": 0.2287,
+      "step": 4588
+    },
+    {
+      "epoch": 2.572309417040359,
+      "grad_norm": 0.0782509976939727,
+      "learning_rate": 1.2132818372021759e-05,
+      "loss": 0.2338,
+      "step": 4589
+    },
+    {
+      "epoch": 2.5728699551569507,
+      "grad_norm": 0.08023409978733587,
+      "learning_rate": 1.2101693405330906e-05,
+      "loss": 0.2474,
+      "step": 4590
+    },
+    {
+      "epoch": 2.5734304932735426,
+      "grad_norm": 0.0815156134427342,
+      "learning_rate": 1.2070605841712813e-05,
+      "loss": 0.2418,
+      "step": 4591
+    },
+    {
+      "epoch": 2.5739910313901344,
+      "grad_norm": 0.08174500722411769,
+      "learning_rate": 1.203955569439601e-05,
+      "loss": 0.2518,
+      "step": 4592
+    },
+    {
+      "epoch": 2.5745515695067267,
+      "grad_norm": 0.08030947691507194,
+      "learning_rate": 1.2008542976593206e-05,
+      "loss": 0.2388,
+      "step": 4593
+    },
+    {
+      "epoch": 2.5751121076233185,
+      "grad_norm": 0.08130695240203926,
+      "learning_rate": 1.1977567701501103e-05,
+      "loss": 0.2355,
+      "step": 4594
+    },
+    {
+      "epoch": 2.5756726457399104,
+      "grad_norm": 0.0790884021363608,
+      "learning_rate": 1.1946629882300553e-05,
+      "loss": 0.2389,
+      "step": 4595
+    },
+    {
+      "epoch": 2.5762331838565022,
+      "grad_norm": 0.07893046150354383,
+      "learning_rate": 1.1915729532156372e-05,
+      "loss": 0.2414,
+      "step": 4596
+    },
+    {
+      "epoch": 2.576793721973094,
+      "grad_norm": 0.07761726989109233,
+      "learning_rate": 1.188486666421753e-05,
+      "loss": 0.2386,
+      "step": 4597
+    },
+    {
+      "epoch": 2.577354260089686,
+      "grad_norm": 0.0797890917727982,
+      "learning_rate": 1.1854041291616946e-05,
+      "loss": 0.2404,
+      "step": 4598
+    },
+    {
+      "epoch": 2.577914798206278,
+      "grad_norm": 0.0827881777224284,
+      "learning_rate": 1.1823253427471681e-05,
+      "loss": 0.2495,
+      "step": 4599
+    },
+    {
+      "epoch": 2.57847533632287,
+      "grad_norm": 0.08082628941808752,
+      "learning_rate": 1.1792503084882789e-05,
+      "loss": 0.2428,
+      "step": 4600
+    },
+    {
+      "epoch": 2.579035874439462,
+      "grad_norm": 0.08058979589937264,
+      "learning_rate": 1.1761790276935336e-05,
+      "loss": 0.2373,
+      "step": 4601
+    },
+    {
+      "epoch": 2.5795964125560538,
+      "grad_norm": 0.07695649037026123,
+      "learning_rate": 1.173111501669848e-05,
+      "loss": 0.2341,
+      "step": 4602
+    },
+    {
+      "epoch": 2.5801569506726456,
+      "grad_norm": 0.08087859212820767,
+      "learning_rate": 1.1700477317225334e-05,
+      "loss": 0.2385,
+      "step": 4603
+    },
+    {
+      "epoch": 2.5807174887892375,
+      "grad_norm": 0.08287720887802386,
+      "learning_rate": 1.1669877191553035e-05,
+      "loss": 0.2263,
+      "step": 4604
+    },
+    {
+      "epoch": 2.5812780269058297,
+      "grad_norm": 0.07816250950995993,
+      "learning_rate": 1.1639314652702793e-05,
+      "loss": 0.2398,
+      "step": 4605
+    },
+    {
+      "epoch": 2.5818385650224216,
+      "grad_norm": 0.08091750663182318,
+      "learning_rate": 1.1608789713679757e-05,
+      "loss": 0.2629,
+      "step": 4606
+    },
+    {
+      "epoch": 2.5823991031390134,
+      "grad_norm": 0.07660575513511397,
+      "learning_rate": 1.1578302387473105e-05,
+      "loss": 0.233,
+      "step": 4607
+    },
+    {
+      "epoch": 2.5829596412556053,
+      "grad_norm": 0.07788562205501486,
+      "learning_rate": 1.1547852687056015e-05,
+      "loss": 0.2337,
+      "step": 4608
+    },
+    {
+      "epoch": 2.583520179372197,
+      "grad_norm": 0.08016084118623913,
+      "learning_rate": 1.1517440625385667e-05,
+      "loss": 0.2384,
+      "step": 4609
+    },
+    {
+      "epoch": 2.5840807174887894,
+      "grad_norm": 0.0783022785313455,
+      "learning_rate": 1.1487066215403186e-05,
+      "loss": 0.2407,
+      "step": 4610
+    },
+    {
+      "epoch": 2.5846412556053813,
+      "grad_norm": 0.08113824685483437,
+      "learning_rate": 1.1456729470033667e-05,
+      "loss": 0.255,
+      "step": 4611
+    },
+    {
+      "epoch": 2.585201793721973,
+      "grad_norm": 0.08211068433770122,
+      "learning_rate": 1.142643040218624e-05,
+      "loss": 0.241,
+      "step": 4612
+    },
+    {
+      "epoch": 2.585762331838565,
+      "grad_norm": 0.07804110098165112,
+      "learning_rate": 1.1396169024753933e-05,
+      "loss": 0.2357,
+      "step": 4613
+    },
+    {
+      "epoch": 2.586322869955157,
+      "grad_norm": 0.07843205512424804,
+      "learning_rate": 1.1365945350613793e-05,
+      "loss": 0.2296,
+      "step": 4614
+    },
+    {
+      "epoch": 2.586883408071749,
+      "grad_norm": 0.07740813803202314,
+      "learning_rate": 1.1335759392626798e-05,
+      "loss": 0.2467,
+      "step": 4615
+    },
+    {
+      "epoch": 2.587443946188341,
+      "grad_norm": 0.07890370072179591,
+      "learning_rate": 1.1305611163637886e-05,
+      "loss": 0.2442,
+      "step": 4616
+    },
+    {
+      "epoch": 2.588004484304933,
+      "grad_norm": 0.08049726094139927,
+      "learning_rate": 1.1275500676475925e-05,
+      "loss": 0.2461,
+      "step": 4617
+    },
+    {
+      "epoch": 2.5885650224215246,
+      "grad_norm": 0.07677261217838514,
+      "learning_rate": 1.12454279439537e-05,
+      "loss": 0.2388,
+      "step": 4618
+    },
+    {
+      "epoch": 2.5891255605381165,
+      "grad_norm": 0.07969531796060425,
+      "learning_rate": 1.121539297886801e-05,
+      "loss": 0.2294,
+      "step": 4619
+    },
+    {
+      "epoch": 2.589686098654709,
+      "grad_norm": 0.07793063555972102,
+      "learning_rate": 1.1185395793999497e-05,
+      "loss": 0.2382,
+      "step": 4620
+    },
+    {
+      "epoch": 2.5902466367713,
+      "grad_norm": 0.08041496562426886,
+      "learning_rate": 1.1155436402112785e-05,
+      "loss": 0.245,
+      "step": 4621
+    },
+    {
+      "epoch": 2.5908071748878925,
+      "grad_norm": 0.07736623150867082,
+      "learning_rate": 1.1125514815956361e-05,
+      "loss": 0.2309,
+      "step": 4622
+    },
+    {
+      "epoch": 2.5913677130044843,
+      "grad_norm": 0.08058079119287281,
+      "learning_rate": 1.109563104826269e-05,
+      "loss": 0.2416,
+      "step": 4623
+    },
+    {
+      "epoch": 2.591928251121076,
+      "grad_norm": 0.07671705061752368,
+      "learning_rate": 1.1065785111748117e-05,
+      "loss": 0.2445,
+      "step": 4624
+    },
+    {
+      "epoch": 2.592488789237668,
+      "grad_norm": 0.08091256841818938,
+      "learning_rate": 1.1035977019112852e-05,
+      "loss": 0.2375,
+      "step": 4625
+    },
+    {
+      "epoch": 2.59304932735426,
+      "grad_norm": 0.07834204684578358,
+      "learning_rate": 1.1006206783041063e-05,
+      "loss": 0.2313,
+      "step": 4626
+    },
+    {
+      "epoch": 2.593609865470852,
+      "grad_norm": 0.0812937113539261,
+      "learning_rate": 1.0976474416200755e-05,
+      "loss": 0.2392,
+      "step": 4627
+    },
+    {
+      "epoch": 2.594170403587444,
+      "grad_norm": 0.0782731877233754,
+      "learning_rate": 1.0946779931243866e-05,
+      "loss": 0.2321,
+      "step": 4628
+    },
+    {
+      "epoch": 2.594730941704036,
+      "grad_norm": 0.07883648296355882,
+      "learning_rate": 1.0917123340806168e-05,
+      "loss": 0.2396,
+      "step": 4629
+    },
+    {
+      "epoch": 2.5952914798206277,
+      "grad_norm": 0.078362956128506,
+      "learning_rate": 1.0887504657507353e-05,
+      "loss": 0.2233,
+      "step": 4630
+    },
+    {
+      "epoch": 2.5958520179372195,
+      "grad_norm": 0.08003834880521533,
+      "learning_rate": 1.0857923893950928e-05,
+      "loss": 0.247,
+      "step": 4631
+    },
+    {
+      "epoch": 2.596412556053812,
+      "grad_norm": 0.08025058062989013,
+      "learning_rate": 1.0828381062724324e-05,
+      "loss": 0.2494,
+      "step": 4632
+    },
+    {
+      "epoch": 2.5969730941704037,
+      "grad_norm": 0.07880394678428947,
+      "learning_rate": 1.079887617639881e-05,
+      "loss": 0.2317,
+      "step": 4633
+    },
+    {
+      "epoch": 2.5975336322869955,
+      "grad_norm": 0.08150930138545963,
+      "learning_rate": 1.0769409247529472e-05,
+      "loss": 0.2536,
+      "step": 4634
+    },
+    {
+      "epoch": 2.5980941704035874,
+      "grad_norm": 0.07985651384036674,
+      "learning_rate": 1.0739980288655316e-05,
+      "loss": 0.2349,
+      "step": 4635
+    },
+    {
+      "epoch": 2.598654708520179,
+      "grad_norm": 0.08150527738856496,
+      "learning_rate": 1.0710589312299091e-05,
+      "loss": 0.2333,
+      "step": 4636
+    },
+    {
+      "epoch": 2.5992152466367715,
+      "grad_norm": 0.08416068011414829,
+      "learning_rate": 1.0681236330967503e-05,
+      "loss": 0.2492,
+      "step": 4637
+    },
+    {
+      "epoch": 2.5997757847533634,
+      "grad_norm": 0.07921301034048962,
+      "learning_rate": 1.0651921357150996e-05,
+      "loss": 0.2423,
+      "step": 4638
+    },
+    {
+      "epoch": 2.600336322869955,
+      "grad_norm": 0.08012881168134223,
+      "learning_rate": 1.0622644403323844e-05,
+      "loss": 0.2299,
+      "step": 4639
+    },
+    {
+      "epoch": 2.600896860986547,
+      "grad_norm": 0.07948077463770052,
+      "learning_rate": 1.0593405481944208e-05,
+      "loss": 0.2475,
+      "step": 4640
+    },
+    {
+      "epoch": 2.601457399103139,
+      "grad_norm": 0.08134179424726441,
+      "learning_rate": 1.0564204605454032e-05,
+      "loss": 0.234,
+      "step": 4641
+    },
+    {
+      "epoch": 2.602017937219731,
+      "grad_norm": 0.08318924334977452,
+      "learning_rate": 1.0535041786279066e-05,
+      "loss": 0.2461,
+      "step": 4642
+    },
+    {
+      "epoch": 2.602578475336323,
+      "grad_norm": 0.08136989551036139,
+      "learning_rate": 1.050591703682886e-05,
+      "loss": 0.2488,
+      "step": 4643
+    },
+    {
+      "epoch": 2.603139013452915,
+      "grad_norm": 0.08134620876824172,
+      "learning_rate": 1.0476830369496759e-05,
+      "loss": 0.2448,
+      "step": 4644
+    },
+    {
+      "epoch": 2.6036995515695067,
+      "grad_norm": 0.07831996360486439,
+      "learning_rate": 1.0447781796659938e-05,
+      "loss": 0.2336,
+      "step": 4645
+    },
+    {
+      "epoch": 2.6042600896860986,
+      "grad_norm": 0.07755786350418256,
+      "learning_rate": 1.0418771330679311e-05,
+      "loss": 0.2342,
+      "step": 4646
+    },
+    {
+      "epoch": 2.604820627802691,
+      "grad_norm": 0.07775354144623337,
+      "learning_rate": 1.0389798983899624e-05,
+      "loss": 0.2188,
+      "step": 4647
+    },
+    {
+      "epoch": 2.6053811659192823,
+      "grad_norm": 0.07888127674381684,
+      "learning_rate": 1.0360864768649381e-05,
+      "loss": 0.2418,
+      "step": 4648
+    },
+    {
+      "epoch": 2.6059417040358746,
+      "grad_norm": 0.08047717524674732,
+      "learning_rate": 1.0331968697240879e-05,
+      "loss": 0.254,
+      "step": 4649
+    },
+    {
+      "epoch": 2.6065022421524664,
+      "grad_norm": 0.07762362263219512,
+      "learning_rate": 1.030311078197015e-05,
+      "loss": 0.2384,
+      "step": 4650
+    },
+    {
+      "epoch": 2.6070627802690582,
+      "grad_norm": 0.07799042816569264,
+      "learning_rate": 1.027429103511698e-05,
+      "loss": 0.2302,
+      "step": 4651
+    },
+    {
+      "epoch": 2.60762331838565,
+      "grad_norm": 0.08085930304246164,
+      "learning_rate": 1.0245509468944992e-05,
+      "loss": 0.2446,
+      "step": 4652
+    },
+    {
+      "epoch": 2.608183856502242,
+      "grad_norm": 0.07988787874092403,
+      "learning_rate": 1.0216766095701457e-05,
+      "loss": 0.2502,
+      "step": 4653
+    },
+    {
+      "epoch": 2.6087443946188342,
+      "grad_norm": 0.0809710576534792,
+      "learning_rate": 1.0188060927617494e-05,
+      "loss": 0.2501,
+      "step": 4654
+    },
+    {
+      "epoch": 2.609304932735426,
+      "grad_norm": 0.0782794833603739,
+      "learning_rate": 1.0159393976907871e-05,
+      "loss": 0.238,
+      "step": 4655
+    },
+    {
+      "epoch": 2.609865470852018,
+      "grad_norm": 0.07852642768667292,
+      "learning_rate": 1.0130765255771169e-05,
+      "loss": 0.2489,
+      "step": 4656
+    },
+    {
+      "epoch": 2.6104260089686098,
+      "grad_norm": 0.07980893402914446,
+      "learning_rate": 1.0102174776389683e-05,
+      "loss": 0.2392,
+      "step": 4657
+    },
+    {
+      "epoch": 2.6109865470852016,
+      "grad_norm": 0.07734153573236843,
+      "learning_rate": 1.0073622550929395e-05,
+      "loss": 0.2383,
+      "step": 4658
+    },
+    {
+      "epoch": 2.611547085201794,
+      "grad_norm": 0.07860451703068655,
+      "learning_rate": 1.0045108591540075e-05,
+      "loss": 0.2289,
+      "step": 4659
+    },
+    {
+      "epoch": 2.6121076233183858,
+      "grad_norm": 0.0795175394237589,
+      "learning_rate": 1.0016632910355117e-05,
+      "loss": 0.2327,
+      "step": 4660
+    },
+    {
+      "epoch": 2.6126681614349776,
+      "grad_norm": 0.08074684826967744,
+      "learning_rate": 9.988195519491739e-06,
+      "loss": 0.2309,
+      "step": 4661
+    },
+    {
+      "epoch": 2.6132286995515694,
+      "grad_norm": 0.07752388356246212,
+      "learning_rate": 9.959796431050772e-06,
+      "loss": 0.2322,
+      "step": 4662
+    },
+    {
+      "epoch": 2.6137892376681613,
+      "grad_norm": 0.07847828544892076,
+      "learning_rate": 9.931435657116817e-06,
+      "loss": 0.2352,
+      "step": 4663
+    },
+    {
+      "epoch": 2.6143497757847536,
+      "grad_norm": 0.07897018834734854,
+      "learning_rate": 9.903113209758096e-06,
+      "loss": 0.2309,
+      "step": 4664
+    },
+    {
+      "epoch": 2.6149103139013454,
+      "grad_norm": 0.08033580219940079,
+      "learning_rate": 9.874829101026584e-06,
+      "loss": 0.238,
+      "step": 4665
+    },
+    {
+      "epoch": 2.6154708520179373,
+      "grad_norm": 0.0756377449056898,
+      "learning_rate": 9.84658334295796e-06,
+      "loss": 0.2369,
+      "step": 4666
+    },
+    {
+      "epoch": 2.616031390134529,
+      "grad_norm": 0.08068419590807634,
+      "learning_rate": 9.81837594757149e-06,
+      "loss": 0.2356,
+      "step": 4667
+    },
+    {
+      "epoch": 2.616591928251121,
+      "grad_norm": 0.07996958970321559,
+      "learning_rate": 9.790206926870215e-06,
+      "loss": 0.2352,
+      "step": 4668
+    },
+    {
+      "epoch": 2.6171524663677133,
+      "grad_norm": 0.07998580082674038,
+      "learning_rate": 9.762076292840783e-06,
+      "loss": 0.2426,
+      "step": 4669
+    },
+    {
+      "epoch": 2.6177130044843047,
+      "grad_norm": 0.08341483581380287,
+      "learning_rate": 9.733984057453538e-06,
+      "loss": 0.2456,
+      "step": 4670
+    },
+    {
+      "epoch": 2.618273542600897,
+      "grad_norm": 0.08034491533586992,
+      "learning_rate": 9.705930232662453e-06,
+      "loss": 0.244,
+      "step": 4671
+    },
+    {
+      "epoch": 2.618834080717489,
+      "grad_norm": 0.07772304435861314,
+      "learning_rate": 9.67791483040521e-06,
+      "loss": 0.2432,
+      "step": 4672
+    },
+    {
+      "epoch": 2.6193946188340806,
+      "grad_norm": 0.08057563850747478,
+      "learning_rate": 9.649937862603099e-06,
+      "loss": 0.2347,
+      "step": 4673
+    },
+    {
+      "epoch": 2.6199551569506725,
+      "grad_norm": 0.08122556936148323,
+      "learning_rate": 9.621999341161047e-06,
+      "loss": 0.2462,
+      "step": 4674
+    },
+    {
+      "epoch": 2.6205156950672643,
+      "grad_norm": 0.07976444756404565,
+      "learning_rate": 9.594099277967683e-06,
+      "loss": 0.235,
+      "step": 4675
+    },
+    {
+      "epoch": 2.6210762331838566,
+      "grad_norm": 0.08050677289630043,
+      "learning_rate": 9.566237684895174e-06,
+      "loss": 0.2367,
+      "step": 4676
+    },
+    {
+      "epoch": 2.6216367713004485,
+      "grad_norm": 0.08007169774187989,
+      "learning_rate": 9.538414573799414e-06,
+      "loss": 0.2463,
+      "step": 4677
+    },
+    {
+      "epoch": 2.6221973094170403,
+      "grad_norm": 0.07968327117755601,
+      "learning_rate": 9.510629956519868e-06,
+      "loss": 0.2369,
+      "step": 4678
+    },
+    {
+      "epoch": 2.622757847533632,
+      "grad_norm": 0.08057650051681482,
+      "learning_rate": 9.482883844879597e-06,
+      "loss": 0.2416,
+      "step": 4679
+    },
+    {
+      "epoch": 2.623318385650224,
+      "grad_norm": 0.07976406048750015,
+      "learning_rate": 9.455176250685338e-06,
+      "loss": 0.247,
+      "step": 4680
+    },
+    {
+      "epoch": 2.6238789237668163,
+      "grad_norm": 0.08036405300601392,
+      "learning_rate": 9.427507185727413e-06,
+      "loss": 0.2436,
+      "step": 4681
+    },
+    {
+      "epoch": 2.624439461883408,
+      "grad_norm": 0.07820639019388559,
+      "learning_rate": 9.399876661779771e-06,
+      "loss": 0.2465,
+      "step": 4682
+    },
+    {
+      "epoch": 2.625,
+      "grad_norm": 0.08169916753855645,
+      "learning_rate": 9.372284690599887e-06,
+      "loss": 0.2526,
+      "step": 4683
+    },
+    {
+      "epoch": 2.625560538116592,
+      "grad_norm": 0.0805764573820981,
+      "learning_rate": 9.34473128392892e-06,
+      "loss": 0.2362,
+      "step": 4684
+    },
+    {
+      "epoch": 2.6261210762331837,
+      "grad_norm": 0.08014393480570253,
+      "learning_rate": 9.317216453491562e-06,
+      "loss": 0.2467,
+      "step": 4685
+    },
+    {
+      "epoch": 2.626681614349776,
+      "grad_norm": 0.0820739945849279,
+      "learning_rate": 9.28974021099609e-06,
+      "loss": 0.2556,
+      "step": 4686
+    },
+    {
+      "epoch": 2.627242152466368,
+      "grad_norm": 0.08136757509280972,
+      "learning_rate": 9.262302568134418e-06,
+      "loss": 0.2538,
+      "step": 4687
+    },
+    {
+      "epoch": 2.6278026905829597,
+      "grad_norm": 0.07774133860201993,
+      "learning_rate": 9.234903536581952e-06,
+      "loss": 0.2273,
+      "step": 4688
+    },
+    {
+      "epoch": 2.6283632286995515,
+      "grad_norm": 0.08084569565280489,
+      "learning_rate": 9.207543127997731e-06,
+      "loss": 0.2422,
+      "step": 4689
+    },
+    {
+      "epoch": 2.6289237668161434,
+      "grad_norm": 0.07986444650344848,
+      "learning_rate": 9.180221354024354e-06,
+      "loss": 0.2358,
+      "step": 4690
+    },
+    {
+      "epoch": 2.6294843049327357,
+      "grad_norm": 0.07962366862394384,
+      "learning_rate": 9.152938226287932e-06,
+      "loss": 0.2319,
+      "step": 4691
+    },
+    {
+      "epoch": 2.6300448430493275,
+      "grad_norm": 0.08022289749822759,
+      "learning_rate": 9.125693756398202e-06,
+      "loss": 0.2356,
+      "step": 4692
+    },
+    {
+      "epoch": 2.6306053811659194,
+      "grad_norm": 0.07875499080453484,
+      "learning_rate": 9.098487955948364e-06,
+      "loss": 0.2377,
+      "step": 4693
+    },
+    {
+      "epoch": 2.631165919282511,
+      "grad_norm": 0.08077688843109554,
+      "learning_rate": 9.071320836515262e-06,
+      "loss": 0.2456,
+      "step": 4694
+    },
+    {
+      "epoch": 2.631726457399103,
+      "grad_norm": 0.08026063523711201,
+      "learning_rate": 9.04419240965918e-06,
+      "loss": 0.2376,
+      "step": 4695
+    },
+    {
+      "epoch": 2.6322869955156953,
+      "grad_norm": 0.07839831164392934,
+      "learning_rate": 9.017102686924028e-06,
+      "loss": 0.2375,
+      "step": 4696
+    },
+    {
+      "epoch": 2.6328475336322867,
+      "grad_norm": 0.07939925362615312,
+      "learning_rate": 8.990051679837175e-06,
+      "loss": 0.2474,
+      "step": 4697
+    },
+    {
+      "epoch": 2.633408071748879,
+      "grad_norm": 0.07863814740863895,
+      "learning_rate": 8.963039399909556e-06,
+      "loss": 0.2394,
+      "step": 4698
+    },
+    {
+      "epoch": 2.633968609865471,
+      "grad_norm": 0.08038397945388753,
+      "learning_rate": 8.936065858635633e-06,
+      "loss": 0.2442,
+      "step": 4699
+    },
+    {
+      "epoch": 2.6345291479820627,
+      "grad_norm": 0.08153688943104939,
+      "learning_rate": 8.909131067493348e-06,
+      "loss": 0.2429,
+      "step": 4700
+    },
+    {
+      "epoch": 2.6350896860986546,
+      "grad_norm": 0.07933631782552886,
+      "learning_rate": 8.882235037944186e-06,
+      "loss": 0.2487,
+      "step": 4701
+    },
+    {
+      "epoch": 2.6356502242152464,
+      "grad_norm": 0.08350820800736432,
+      "learning_rate": 8.855377781433094e-06,
+      "loss": 0.2455,
+      "step": 4702
+    },
+    {
+      "epoch": 2.6362107623318387,
+      "grad_norm": 0.08225559708493778,
+      "learning_rate": 8.828559309388596e-06,
+      "loss": 0.2457,
+      "step": 4703
+    },
+    {
+      "epoch": 2.6367713004484306,
+      "grad_norm": 0.07950349828915644,
+      "learning_rate": 8.80177963322263e-06,
+      "loss": 0.2397,
+      "step": 4704
+    },
+    {
+      "epoch": 2.6373318385650224,
+      "grad_norm": 0.0833246707543202,
+      "learning_rate": 8.775038764330679e-06,
+      "loss": 0.2345,
+      "step": 4705
+    },
+    {
+      "epoch": 2.6378923766816142,
+      "grad_norm": 0.08210156398956797,
+      "learning_rate": 8.748336714091698e-06,
+      "loss": 0.2456,
+      "step": 4706
+    },
+    {
+      "epoch": 2.638452914798206,
+      "grad_norm": 0.07923756575214472,
+      "learning_rate": 8.72167349386811e-06,
+      "loss": 0.2396,
+      "step": 4707
+    },
+    {
+      "epoch": 2.6390134529147984,
+      "grad_norm": 0.08042949943095787,
+      "learning_rate": 8.695049115005837e-06,
+      "loss": 0.2322,
+      "step": 4708
+    },
+    {
+      "epoch": 2.6395739910313902,
+      "grad_norm": 0.08324991921517294,
+      "learning_rate": 8.668463588834253e-06,
+      "loss": 0.2446,
+      "step": 4709
+    },
+    {
+      "epoch": 2.640134529147982,
+      "grad_norm": 0.07962396678104898,
+      "learning_rate": 8.641916926666216e-06,
+      "loss": 0.2445,
+      "step": 4710
+    },
+    {
+      "epoch": 2.640695067264574,
+      "grad_norm": 0.08128990462058089,
+      "learning_rate": 8.615409139798048e-06,
+      "loss": 0.2504,
+      "step": 4711
+    },
+    {
+      "epoch": 2.6412556053811658,
+      "grad_norm": 0.08049269636217003,
+      "learning_rate": 8.58894023950948e-06,
+      "loss": 0.2468,
+      "step": 4712
+    },
+    {
+      "epoch": 2.641816143497758,
+      "grad_norm": 0.07924177779021946,
+      "learning_rate": 8.562510237063758e-06,
+      "loss": 0.242,
+      "step": 4713
+    },
+    {
+      "epoch": 2.64237668161435,
+      "grad_norm": 0.08155667402234759,
+      "learning_rate": 8.536119143707555e-06,
+      "loss": 0.2481,
+      "step": 4714
+    },
+    {
+      "epoch": 2.6429372197309418,
+      "grad_norm": 0.07820082385607566,
+      "learning_rate": 8.509766970671007e-06,
+      "loss": 0.2346,
+      "step": 4715
+    },
+    {
+      "epoch": 2.6434977578475336,
+      "grad_norm": 0.0825180322805504,
+      "learning_rate": 8.483453729167622e-06,
+      "loss": 0.2517,
+      "step": 4716
+    },
+    {
+      "epoch": 2.6440582959641254,
+      "grad_norm": 0.08108549758779822,
+      "learning_rate": 8.457179430394424e-06,
+      "loss": 0.2372,
+      "step": 4717
+    },
+    {
+      "epoch": 2.6446188340807177,
+      "grad_norm": 0.07790912705506942,
+      "learning_rate": 8.430944085531811e-06,
+      "loss": 0.2436,
+      "step": 4718
+    },
+    {
+      "epoch": 2.645179372197309,
+      "grad_norm": 0.07977766511688403,
+      "learning_rate": 8.40474770574361e-06,
+      "loss": 0.2528,
+      "step": 4719
+    },
+    {
+      "epoch": 2.6457399103139014,
+      "grad_norm": 0.08123725272003492,
+      "learning_rate": 8.378590302177102e-06,
+      "loss": 0.2384,
+      "step": 4720
+    },
+    {
+      "epoch": 2.6463004484304933,
+      "grad_norm": 0.08140475056257848,
+      "learning_rate": 8.352471885962931e-06,
+      "loss": 0.2347,
+      "step": 4721
+    },
+    {
+      "epoch": 2.646860986547085,
+      "grad_norm": 0.07923734139983633,
+      "learning_rate": 8.326392468215205e-06,
+      "loss": 0.231,
+      "step": 4722
+    },
+    {
+      "epoch": 2.647421524663677,
+      "grad_norm": 0.0833880681940613,
+      "learning_rate": 8.300352060031391e-06,
+      "loss": 0.2429,
+      "step": 4723
+    },
+    {
+      "epoch": 2.647982062780269,
+      "grad_norm": 0.08357552737977247,
+      "learning_rate": 8.274350672492415e-06,
+      "loss": 0.2437,
+      "step": 4724
+    },
+    {
+      "epoch": 2.648542600896861,
+      "grad_norm": 0.0805449550726762,
+      "learning_rate": 8.248388316662525e-06,
+      "loss": 0.2585,
+      "step": 4725
+    },
+    {
+      "epoch": 2.649103139013453,
+      "grad_norm": 0.07891704758599948,
+      "learning_rate": 8.222465003589398e-06,
+      "loss": 0.2374,
+      "step": 4726
+    },
+    {
+      "epoch": 2.649663677130045,
+      "grad_norm": 0.08106681371740672,
+      "learning_rate": 8.196580744304116e-06,
+      "loss": 0.2322,
+      "step": 4727
+    },
+    {
+      "epoch": 2.6502242152466366,
+      "grad_norm": 0.08009115068030585,
+      "learning_rate": 8.170735549821085e-06,
+      "loss": 0.232,
+      "step": 4728
+    },
+    {
+      "epoch": 2.6507847533632285,
+      "grad_norm": 0.07874987050584663,
+      "learning_rate": 8.14492943113817e-06,
+      "loss": 0.2399,
+      "step": 4729
+    },
+    {
+      "epoch": 2.651345291479821,
+      "grad_norm": 0.07662010629486314,
+      "learning_rate": 8.119162399236513e-06,
+      "loss": 0.2422,
+      "step": 4730
+    },
+    {
+      "epoch": 2.6519058295964126,
+      "grad_norm": 0.08352877307067955,
+      "learning_rate": 8.093434465080706e-06,
+      "loss": 0.2427,
+      "step": 4731
+    },
+    {
+      "epoch": 2.6524663677130045,
+      "grad_norm": 0.08123913507433109,
+      "learning_rate": 8.067745639618684e-06,
+      "loss": 0.2453,
+      "step": 4732
+    },
+    {
+      "epoch": 2.6530269058295963,
+      "grad_norm": 0.08063630580939841,
+      "learning_rate": 8.04209593378168e-06,
+      "loss": 0.2476,
+      "step": 4733
+    },
+    {
+      "epoch": 2.653587443946188,
+      "grad_norm": 0.07806425792440652,
+      "learning_rate": 8.016485358484383e-06,
+      "loss": 0.2288,
+      "step": 4734
+    },
+    {
+      "epoch": 2.6541479820627805,
+      "grad_norm": 0.07707749516236907,
+      "learning_rate": 7.990913924624722e-06,
+      "loss": 0.2381,
+      "step": 4735
+    },
+    {
+      "epoch": 2.6547085201793723,
+      "grad_norm": 0.07900612663841797,
+      "learning_rate": 7.96538164308407e-06,
+      "loss": 0.2319,
+      "step": 4736
+    },
+    {
+      "epoch": 2.655269058295964,
+      "grad_norm": 0.07954983331838897,
+      "learning_rate": 7.939888524727047e-06,
+      "loss": 0.2349,
+      "step": 4737
+    },
+    {
+      "epoch": 2.655829596412556,
+      "grad_norm": 0.08021331901320931,
+      "learning_rate": 7.914434580401686e-06,
+      "loss": 0.2399,
+      "step": 4738
+    },
+    {
+      "epoch": 2.656390134529148,
+      "grad_norm": 0.07905180362069561,
+      "learning_rate": 7.889019820939325e-06,
+      "loss": 0.2332,
+      "step": 4739
+    },
+    {
+      "epoch": 2.65695067264574,
+      "grad_norm": 0.08130104002495893,
+      "learning_rate": 7.8636442571546e-06,
+      "loss": 0.2483,
+      "step": 4740
+    },
+    {
+      "epoch": 2.657511210762332,
+      "grad_norm": 0.07953731009728611,
+      "learning_rate": 7.838307899845509e-06,
+      "loss": 0.238,
+      "step": 4741
+    },
+    {
+      "epoch": 2.658071748878924,
+      "grad_norm": 0.08075923076714953,
+      "learning_rate": 7.813010759793326e-06,
+      "loss": 0.2399,
+      "step": 4742
+    },
+    {
+      "epoch": 2.6586322869955157,
+      "grad_norm": 0.07912799503613634,
+      "learning_rate": 7.787752847762685e-06,
+      "loss": 0.244,
+      "step": 4743
+    },
+    {
+      "epoch": 2.6591928251121075,
+      "grad_norm": 0.08262334060021577,
+      "learning_rate": 7.76253417450149e-06,
+      "loss": 0.2528,
+      "step": 4744
+    },
+    {
+      "epoch": 2.6597533632287,
+      "grad_norm": 0.08379137046638221,
+      "learning_rate": 7.737354750740933e-06,
+      "loss": 0.2445,
+      "step": 4745
+    },
+    {
+      "epoch": 2.660313901345291,
+      "grad_norm": 0.07959878967597425,
+      "learning_rate": 7.712214587195554e-06,
+      "loss": 0.249,
+      "step": 4746
+    },
+    {
+      "epoch": 2.6608744394618835,
+      "grad_norm": 0.07851889305169113,
+      "learning_rate": 7.687113694563153e-06,
+      "loss": 0.2325,
+      "step": 4747
+    },
+    {
+      "epoch": 2.6614349775784754,
+      "grad_norm": 0.07951174972995802,
+      "learning_rate": 7.662052083524863e-06,
+      "loss": 0.2445,
+      "step": 4748
+    },
+    {
+      "epoch": 2.661995515695067,
+      "grad_norm": 0.07800438368537502,
+      "learning_rate": 7.637029764745019e-06,
+      "loss": 0.2463,
+      "step": 4749
+    },
+    {
+      "epoch": 2.662556053811659,
+      "grad_norm": 0.07793478332510123,
+      "learning_rate": 7.612046748871327e-06,
+      "loss": 0.2143,
+      "step": 4750
+    },
+    {
+      "epoch": 2.663116591928251,
+      "grad_norm": 0.07937211104180285,
+      "learning_rate": 7.587103046534705e-06,
+      "loss": 0.2376,
+      "step": 4751
+    },
+    {
+      "epoch": 2.663677130044843,
+      "grad_norm": 0.08041006055644226,
+      "learning_rate": 7.562198668349352e-06,
+      "loss": 0.2336,
+      "step": 4752
+    },
+    {
+      "epoch": 2.664237668161435,
+      "grad_norm": 0.07947393992836635,
+      "learning_rate": 7.537333624912768e-06,
+      "loss": 0.2347,
+      "step": 4753
+    },
+    {
+      "epoch": 2.664798206278027,
+      "grad_norm": 0.0809116546041903,
+      "learning_rate": 7.512507926805668e-06,
+      "loss": 0.2624,
+      "step": 4754
+    },
+    {
+      "epoch": 2.6653587443946187,
+      "grad_norm": 0.0796819941226107,
+      "learning_rate": 7.4877215845920555e-06,
+      "loss": 0.2412,
+      "step": 4755
+    },
+    {
+      "epoch": 2.6659192825112106,
+      "grad_norm": 0.07864802644574229,
+      "learning_rate": 7.462974608819196e-06,
+      "loss": 0.2371,
+      "step": 4756
+    },
+    {
+      "epoch": 2.666479820627803,
+      "grad_norm": 0.07734365155763259,
+      "learning_rate": 7.438267010017585e-06,
+      "loss": 0.2375,
+      "step": 4757
+    },
+    {
+      "epoch": 2.6670403587443947,
+      "grad_norm": 0.08022431608682906,
+      "learning_rate": 7.41359879870096e-06,
+      "loss": 0.2364,
+      "step": 4758
+    },
+    {
+      "epoch": 2.6676008968609866,
+      "grad_norm": 0.07997424733047849,
+      "learning_rate": 7.38896998536629e-06,
+      "loss": 0.2376,
+      "step": 4759
+    },
+    {
+      "epoch": 2.6681614349775784,
+      "grad_norm": 0.08197462441510985,
+      "learning_rate": 7.364380580493813e-06,
+      "loss": 0.2461,
+      "step": 4760
+    },
+    {
+      "epoch": 2.6687219730941703,
+      "grad_norm": 0.07633156868865107,
+      "learning_rate": 7.339830594546937e-06,
+      "loss": 0.2323,
+      "step": 4761
+    },
+    {
+      "epoch": 2.6692825112107625,
+      "grad_norm": 0.07682717870384738,
+      "learning_rate": 7.315320037972395e-06,
+      "loss": 0.2386,
+      "step": 4762
+    },
+    {
+      "epoch": 2.6698430493273544,
+      "grad_norm": 0.07828778526955535,
+      "learning_rate": 7.290848921200022e-06,
+      "loss": 0.2313,
+      "step": 4763
+    },
+    {
+      "epoch": 2.6704035874439462,
+      "grad_norm": 0.08058231110695564,
+      "learning_rate": 7.2664172546429655e-06,
+      "loss": 0.2478,
+      "step": 4764
+    },
+    {
+      "epoch": 2.670964125560538,
+      "grad_norm": 0.07953252462034587,
+      "learning_rate": 7.242025048697565e-06,
+      "loss": 0.2338,
+      "step": 4765
+    },
+    {
+      "epoch": 2.67152466367713,
+      "grad_norm": 0.0782901538887304,
+      "learning_rate": 7.217672313743306e-06,
+      "loss": 0.2451,
+      "step": 4766
+    },
+    {
+      "epoch": 2.672085201793722,
+      "grad_norm": 0.08036684426565673,
+      "learning_rate": 7.193359060142979e-06,
+      "loss": 0.2339,
+      "step": 4767
+    },
+    {
+      "epoch": 2.672645739910314,
+      "grad_norm": 0.07749707686250977,
+      "learning_rate": 7.169085298242473e-06,
+      "loss": 0.2552,
+      "step": 4768
+    },
+    {
+      "epoch": 2.673206278026906,
+      "grad_norm": 0.08274416518853744,
+      "learning_rate": 7.1448510383709696e-06,
+      "loss": 0.2565,
+      "step": 4769
+    },
+    {
+      "epoch": 2.6737668161434978,
+      "grad_norm": 0.07804543456423374,
+      "learning_rate": 7.120656290840744e-06,
+      "loss": 0.2453,
+      "step": 4770
+    },
+    {
+      "epoch": 2.6743273542600896,
+      "grad_norm": 0.08115831948075995,
+      "learning_rate": 7.0965010659473256e-06,
+      "loss": 0.2312,
+      "step": 4771
+    },
+    {
+      "epoch": 2.6748878923766815,
+      "grad_norm": 0.08128287586424483,
+      "learning_rate": 7.0723853739694364e-06,
+      "loss": 0.2452,
+      "step": 4772
+    },
+    {
+      "epoch": 2.6754484304932733,
+      "grad_norm": 0.08119408695590151,
+      "learning_rate": 7.048309225168903e-06,
+      "loss": 0.2358,
+      "step": 4773
+    },
+    {
+      "epoch": 2.6760089686098656,
+      "grad_norm": 0.07941834280290176,
+      "learning_rate": 7.024272629790795e-06,
+      "loss": 0.2387,
+      "step": 4774
+    },
+    {
+      "epoch": 2.6765695067264574,
+      "grad_norm": 0.08015615594526529,
+      "learning_rate": 7.000275598063299e-06,
+      "loss": 0.2363,
+      "step": 4775
+    },
+    {
+      "epoch": 2.6771300448430493,
+      "grad_norm": 0.0778155107576842,
+      "learning_rate": 6.976318140197835e-06,
+      "loss": 0.2285,
+      "step": 4776
+    },
+    {
+      "epoch": 2.677690582959641,
+      "grad_norm": 0.08062179382102118,
+      "learning_rate": 6.952400266388903e-06,
+      "loss": 0.2529,
+      "step": 4777
+    },
+    {
+      "epoch": 2.678251121076233,
+      "grad_norm": 0.08224849028132702,
+      "learning_rate": 6.928521986814196e-06,
+      "loss": 0.2471,
+      "step": 4778
+    },
+    {
+      "epoch": 2.6788116591928253,
+      "grad_norm": 0.08003155990063,
+      "learning_rate": 6.9046833116345635e-06,
+      "loss": 0.2469,
+      "step": 4779
+    },
+    {
+      "epoch": 2.679372197309417,
+      "grad_norm": 0.08078888522682524,
+      "learning_rate": 6.8808842509940015e-06,
+      "loss": 0.2409,
+      "step": 4780
+    },
+    {
+      "epoch": 2.679932735426009,
+      "grad_norm": 0.08159683352788072,
+      "learning_rate": 6.857124815019666e-06,
+      "loss": 0.252,
+      "step": 4781
+    },
+    {
+      "epoch": 2.680493273542601,
+      "grad_norm": 0.07765536446895811,
+      "learning_rate": 6.833405013821792e-06,
+      "loss": 0.2462,
+      "step": 4782
+    },
+    {
+      "epoch": 2.6810538116591927,
+      "grad_norm": 0.08021593389332181,
+      "learning_rate": 6.809724857493826e-06,
+      "loss": 0.2399,
+      "step": 4783
+    },
+    {
+      "epoch": 2.681614349775785,
+      "grad_norm": 0.08141240401544701,
+      "learning_rate": 6.7860843561122765e-06,
+      "loss": 0.2467,
+      "step": 4784
+    },
+    {
+      "epoch": 2.682174887892377,
+      "grad_norm": 0.07766797150732227,
+      "learning_rate": 6.762483519736806e-06,
+      "loss": 0.2423,
+      "step": 4785
+    },
+    {
+      "epoch": 2.6827354260089686,
+      "grad_norm": 0.0829476374142876,
+      "learning_rate": 6.7389223584102265e-06,
+      "loss": 0.249,
+      "step": 4786
+    },
+    {
+      "epoch": 2.6832959641255605,
+      "grad_norm": 0.0790995203648318,
+      "learning_rate": 6.715400882158396e-06,
+      "loss": 0.2493,
+      "step": 4787
+    },
+    {
+      "epoch": 2.6838565022421523,
+      "grad_norm": 0.08060900565937387,
+      "learning_rate": 6.6919191009903734e-06,
+      "loss": 0.2499,
+      "step": 4788
+    },
+    {
+      "epoch": 2.6844170403587446,
+      "grad_norm": 0.07701822265961265,
+      "learning_rate": 6.668477024898257e-06,
+      "loss": 0.2381,
+      "step": 4789
+    },
+    {
+      "epoch": 2.6849775784753365,
+      "grad_norm": 0.07947128695041439,
+      "learning_rate": 6.645074663857298e-06,
+      "loss": 0.2512,
+      "step": 4790
+    },
+    {
+      "epoch": 2.6855381165919283,
+      "grad_norm": 0.07912370472664916,
+      "learning_rate": 6.621712027825811e-06,
+      "loss": 0.243,
+      "step": 4791
+    },
+    {
+      "epoch": 2.68609865470852,
+      "grad_norm": 0.08015187760994366,
+      "learning_rate": 6.598389126745208e-06,
+      "loss": 0.2242,
+      "step": 4792
+    },
+    {
+      "epoch": 2.686659192825112,
+      "grad_norm": 0.08021705459077672,
+      "learning_rate": 6.5751059705400295e-06,
+      "loss": 0.2475,
+      "step": 4793
+    },
+    {
+      "epoch": 2.6872197309417043,
+      "grad_norm": 0.0793126903156517,
+      "learning_rate": 6.55186256911785e-06,
+      "loss": 0.246,
+      "step": 4794
+    },
+    {
+      "epoch": 2.6877802690582957,
+      "grad_norm": 0.08237946052523472,
+      "learning_rate": 6.5286589323693914e-06,
+      "loss": 0.2473,
+      "step": 4795
+    },
+    {
+      "epoch": 2.688340807174888,
+      "grad_norm": 0.08298713615830196,
+      "learning_rate": 6.505495070168388e-06,
+      "loss": 0.2422,
+      "step": 4796
+    },
+    {
+      "epoch": 2.68890134529148,
+      "grad_norm": 0.08190739726043625,
+      "learning_rate": 6.482370992371689e-06,
+      "loss": 0.2451,
+      "step": 4797
+    },
+    {
+      "epoch": 2.6894618834080717,
+      "grad_norm": 0.08095815081778734,
+      "learning_rate": 6.459286708819234e-06,
+      "loss": 0.2406,
+      "step": 4798
+    },
+    {
+      "epoch": 2.6900224215246635,
+      "grad_norm": 0.08024020442517155,
+      "learning_rate": 6.4362422293339665e-06,
+      "loss": 0.2324,
+      "step": 4799
+    },
+    {
+      "epoch": 2.6905829596412554,
+      "grad_norm": 0.08199655127025579,
+      "learning_rate": 6.413237563721941e-06,
+      "loss": 0.24,
+      "step": 4800
+    },
+    {
+      "epoch": 2.6911434977578477,
+      "grad_norm": 0.07907318573785835,
+      "learning_rate": 6.39027272177225e-06,
+      "loss": 0.243,
+      "step": 4801
+    },
+    {
+      "epoch": 2.6917040358744395,
+      "grad_norm": 0.0815893672235332,
+      "learning_rate": 6.367347713257066e-06,
+      "loss": 0.2477,
+      "step": 4802
+    },
+    {
+      "epoch": 2.6922645739910314,
+      "grad_norm": 0.0807876069026921,
+      "learning_rate": 6.344462547931551e-06,
+      "loss": 0.241,
+      "step": 4803
+    },
+    {
+      "epoch": 2.692825112107623,
+      "grad_norm": 0.08201034304267234,
+      "learning_rate": 6.321617235533983e-06,
+      "loss": 0.2547,
+      "step": 4804
+    },
+    {
+      "epoch": 2.693385650224215,
+      "grad_norm": 0.0807326355114875,
+      "learning_rate": 6.298811785785663e-06,
+      "loss": 0.2432,
+      "step": 4805
+    },
+    {
+      "epoch": 2.6939461883408073,
+      "grad_norm": 0.08143303028404608,
+      "learning_rate": 6.276046208390873e-06,
+      "loss": 0.2505,
+      "step": 4806
+    },
+    {
+      "epoch": 2.694506726457399,
+      "grad_norm": 0.07886102191315485,
+      "learning_rate": 6.253320513037031e-06,
+      "loss": 0.226,
+      "step": 4807
+    },
+    {
+      "epoch": 2.695067264573991,
+      "grad_norm": 0.07878016928054814,
+      "learning_rate": 6.230634709394478e-06,
+      "loss": 0.2366,
+      "step": 4808
+    },
+    {
+      "epoch": 2.695627802690583,
+      "grad_norm": 0.07975266815564028,
+      "learning_rate": 6.207988807116649e-06,
+      "loss": 0.2365,
+      "step": 4809
+    },
+    {
+      "epoch": 2.6961883408071747,
+      "grad_norm": 0.07899986348448437,
+      "learning_rate": 6.185382815839969e-06,
+      "loss": 0.2476,
+      "step": 4810
+    },
+    {
+      "epoch": 2.696748878923767,
+      "grad_norm": 0.08157078034556416,
+      "learning_rate": 6.162816745183919e-06,
+      "loss": 0.2398,
+      "step": 4811
+    },
+    {
+      "epoch": 2.697309417040359,
+      "grad_norm": 0.07931318136489432,
+      "learning_rate": 6.14029060475092e-06,
+      "loss": 0.243,
+      "step": 4812
+    },
+    {
+      "epoch": 2.6978699551569507,
+      "grad_norm": 0.07772738146435056,
+      "learning_rate": 6.117804404126459e-06,
+      "loss": 0.2353,
+      "step": 4813
+    },
+    {
+      "epoch": 2.6984304932735426,
+      "grad_norm": 0.07937697193267051,
+      "learning_rate": 6.095358152879049e-06,
+      "loss": 0.2406,
+      "step": 4814
+    },
+    {
+      "epoch": 2.6989910313901344,
+      "grad_norm": 0.08013247564199258,
+      "learning_rate": 6.072951860560128e-06,
+      "loss": 0.2418,
+      "step": 4815
+    },
+    {
+      "epoch": 2.6995515695067267,
+      "grad_norm": 0.08280738561451664,
+      "learning_rate": 6.0505855367041895e-06,
+      "loss": 0.2458,
+      "step": 4816
+    },
+    {
+      "epoch": 2.7001121076233185,
+      "grad_norm": 0.07795701064327104,
+      "learning_rate": 6.0282591908287e-06,
+      "loss": 0.2318,
+      "step": 4817
+    },
+    {
+      "epoch": 2.7006726457399104,
+      "grad_norm": 0.07787441036672033,
+      "learning_rate": 6.005972832434093e-06,
+      "loss": 0.2318,
+      "step": 4818
+    },
+    {
+      "epoch": 2.7012331838565022,
+      "grad_norm": 0.08228921367300045,
+      "learning_rate": 5.983726471003836e-06,
+      "loss": 0.2431,
+      "step": 4819
+    },
+    {
+      "epoch": 2.701793721973094,
+      "grad_norm": 0.0815026313218986,
+      "learning_rate": 5.961520116004327e-06,
+      "loss": 0.2432,
+      "step": 4820
+    },
+    {
+      "epoch": 2.702354260089686,
+      "grad_norm": 0.08253122671677174,
+      "learning_rate": 5.93935377688497e-06,
+      "loss": 0.2347,
+      "step": 4821
+    },
+    {
+      "epoch": 2.702914798206278,
+      "grad_norm": 0.08356273204249452,
+      "learning_rate": 5.917227463078146e-06,
+      "loss": 0.2467,
+      "step": 4822
+    },
+    {
+      "epoch": 2.70347533632287,
+      "grad_norm": 0.0797360105512306,
+      "learning_rate": 5.895141183999187e-06,
+      "loss": 0.239,
+      "step": 4823
+    },
+    {
+      "epoch": 2.704035874439462,
+      "grad_norm": 0.07962457970507808,
+      "learning_rate": 5.873094949046387e-06,
+      "loss": 0.2305,
+      "step": 4824
+    },
+    {
+      "epoch": 2.7045964125560538,
+      "grad_norm": 0.07778230843941036,
+      "learning_rate": 5.851088767600998e-06,
+      "loss": 0.2299,
+      "step": 4825
+    },
+    {
+      "epoch": 2.7051569506726456,
+      "grad_norm": 0.0817028845104827,
+      "learning_rate": 5.8291226490272526e-06,
+      "loss": 0.2458,
+      "step": 4826
+    },
+    {
+      "epoch": 2.7057174887892375,
+      "grad_norm": 0.08313270905328339,
+      "learning_rate": 5.807196602672305e-06,
+      "loss": 0.2521,
+      "step": 4827
+    },
+    {
+      "epoch": 2.7062780269058297,
+      "grad_norm": 0.07929295703556682,
+      "learning_rate": 5.785310637866304e-06,
+      "loss": 0.2248,
+      "step": 4828
+    },
+    {
+      "epoch": 2.7068385650224216,
+      "grad_norm": 0.08038151631744177,
+      "learning_rate": 5.763464763922255e-06,
+      "loss": 0.2377,
+      "step": 4829
+    },
+    {
+      "epoch": 2.7073991031390134,
+      "grad_norm": 0.08086742912267261,
+      "learning_rate": 5.7416589901362115e-06,
+      "loss": 0.2326,
+      "step": 4830
+    },
+    {
+      "epoch": 2.7079596412556053,
+      "grad_norm": 0.0812195687445791,
+      "learning_rate": 5.7198933257870955e-06,
+      "loss": 0.2537,
+      "step": 4831
+    },
+    {
+      "epoch": 2.708520179372197,
+      "grad_norm": 0.08277110352553523,
+      "learning_rate": 5.698167780136765e-06,
+      "loss": 0.239,
+      "step": 4832
+    },
+    {
+      "epoch": 2.7090807174887894,
+      "grad_norm": 0.08032941302057249,
+      "learning_rate": 5.676482362430047e-06,
+      "loss": 0.2381,
+      "step": 4833
+    },
+    {
+      "epoch": 2.7096412556053813,
+      "grad_norm": 0.08059047081602719,
+      "learning_rate": 5.654837081894626e-06,
+      "loss": 0.2473,
+      "step": 4834
+    },
+    {
+      "epoch": 2.710201793721973,
+      "grad_norm": 0.08055343672565488,
+      "learning_rate": 5.63323194774118e-06,
+      "loss": 0.2437,
+      "step": 4835
+    },
+    {
+      "epoch": 2.710762331838565,
+      "grad_norm": 0.08329435670479612,
+      "learning_rate": 5.611666969163243e-06,
+      "loss": 0.2446,
+      "step": 4836
+    },
+    {
+      "epoch": 2.711322869955157,
+      "grad_norm": 0.0778452362302262,
+      "learning_rate": 5.590142155337308e-06,
+      "loss": 0.2412,
+      "step": 4837
+    },
+    {
+      "epoch": 2.711883408071749,
+      "grad_norm": 0.08349358126462086,
+      "learning_rate": 5.568657515422759e-06,
+      "loss": 0.2406,
+      "step": 4838
+    },
+    {
+      "epoch": 2.712443946188341,
+      "grad_norm": 0.08061624483910564,
+      "learning_rate": 5.547213058561862e-06,
+      "loss": 0.2474,
+      "step": 4839
+    },
+    {
+      "epoch": 2.713004484304933,
+      "grad_norm": 0.08583768508047582,
+      "learning_rate": 5.525808793879838e-06,
+      "loss": 0.2499,
+      "step": 4840
+    },
+    {
+      "epoch": 2.7135650224215246,
+      "grad_norm": 0.08190295770275775,
+      "learning_rate": 5.504444730484726e-06,
+      "loss": 0.2297,
+      "step": 4841
+    },
+    {
+      "epoch": 2.7141255605381165,
+      "grad_norm": 0.08046549479817044,
+      "learning_rate": 5.4831208774675515e-06,
+      "loss": 0.2517,
+      "step": 4842
+    },
+    {
+      "epoch": 2.714686098654709,
+      "grad_norm": 0.08275112958692439,
+      "learning_rate": 5.461837243902146e-06,
+      "loss": 0.2411,
+      "step": 4843
+    },
+    {
+      "epoch": 2.7152466367713,
+      "grad_norm": 0.08320547371796622,
+      "learning_rate": 5.440593838845287e-06,
+      "loss": 0.2412,
+      "step": 4844
+    },
+    {
+      "epoch": 2.7158071748878925,
+      "grad_norm": 0.08583192312348593,
+      "learning_rate": 5.4193906713366e-06,
+      "loss": 0.2509,
+      "step": 4845
+    },
+    {
+      "epoch": 2.7163677130044843,
+      "grad_norm": 0.08095407175439231,
+      "learning_rate": 5.398227750398588e-06,
+      "loss": 0.2436,
+      "step": 4846
+    },
+    {
+      "epoch": 2.716928251121076,
+      "grad_norm": 0.0817826587660967,
+      "learning_rate": 5.377105085036671e-06,
+      "loss": 0.244,
+      "step": 4847
+    },
+    {
+      "epoch": 2.717488789237668,
+      "grad_norm": 0.07997392241960607,
+      "learning_rate": 5.3560226842390596e-06,
+      "loss": 0.2412,
+      "step": 4848
+    },
+    {
+      "epoch": 2.71804932735426,
+      "grad_norm": 0.0806758305455095,
+      "learning_rate": 5.33498055697692e-06,
+      "loss": 0.241,
+      "step": 4849
+    },
+    {
+      "epoch": 2.718609865470852,
+      "grad_norm": 0.08151772787644262,
+      "learning_rate": 5.313978712204215e-06,
+      "loss": 0.2521,
+      "step": 4850
+    },
+    {
+      "epoch": 2.719170403587444,
+      "grad_norm": 0.08042888903419694,
+      "learning_rate": 5.293017158857804e-06,
+      "loss": 0.2378,
+      "step": 4851
+    },
+    {
+      "epoch": 2.719730941704036,
+      "grad_norm": 0.0817197201429122,
+      "learning_rate": 5.2720959058573775e-06,
+      "loss": 0.2431,
+      "step": 4852
+    },
+    {
+      "epoch": 2.7202914798206277,
+      "grad_norm": 0.07911262496270136,
+      "learning_rate": 5.251214962105466e-06,
+      "loss": 0.2446,
+      "step": 4853
+    },
+    {
+      "epoch": 2.7208520179372195,
+      "grad_norm": 0.08148581568289714,
+      "learning_rate": 5.230374336487498e-06,
+      "loss": 0.2453,
+      "step": 4854
+    },
+    {
+      "epoch": 2.721412556053812,
+      "grad_norm": 0.07972619160712666,
+      "learning_rate": 5.209574037871701e-06,
+      "loss": 0.2345,
+      "step": 4855
+    },
+    {
+      "epoch": 2.7219730941704037,
+      "grad_norm": 0.08018928423055902,
+      "learning_rate": 5.188814075109172e-06,
+      "loss": 0.2544,
+      "step": 4856
+    },
+    {
+      "epoch": 2.7225336322869955,
+      "grad_norm": 0.07909166543769797,
+      "learning_rate": 5.168094457033801e-06,
+      "loss": 0.2293,
+      "step": 4857
+    },
+    {
+      "epoch": 2.7230941704035874,
+      "grad_norm": 0.08227752919006537,
+      "learning_rate": 5.147415192462379e-06,
+      "loss": 0.2425,
+      "step": 4858
+    },
+    {
+      "epoch": 2.723654708520179,
+      "grad_norm": 0.08192267781634209,
+      "learning_rate": 5.1267762901944575e-06,
+      "loss": 0.2437,
+      "step": 4859
+    },
+    {
+      "epoch": 2.7242152466367715,
+      "grad_norm": 0.0790172993354068,
+      "learning_rate": 5.106177759012421e-06,
+      "loss": 0.2341,
+      "step": 4860
+    },
+    {
+      "epoch": 2.7247757847533634,
+      "grad_norm": 0.07960463122203566,
+      "learning_rate": 5.085619607681524e-06,
+      "loss": 0.2444,
+      "step": 4861
+    },
+    {
+      "epoch": 2.725336322869955,
+      "grad_norm": 0.07758129627669537,
+      "learning_rate": 5.065101844949794e-06,
+      "loss": 0.2439,
+      "step": 4862
+    },
+    {
+      "epoch": 2.725896860986547,
+      "grad_norm": 0.08202261406444565,
+      "learning_rate": 5.044624479548099e-06,
+      "loss": 0.2402,
+      "step": 4863
+    },
+    {
+      "epoch": 2.726457399103139,
+      "grad_norm": 0.07979434006296819,
+      "learning_rate": 5.024187520190104e-06,
+      "loss": 0.2336,
+      "step": 4864
+    },
+    {
+      "epoch": 2.727017937219731,
+      "grad_norm": 0.0816635584131755,
+      "learning_rate": 5.003790975572253e-06,
+      "loss": 0.239,
+      "step": 4865
+    },
+    {
+      "epoch": 2.727578475336323,
+      "grad_norm": 0.08024566579238525,
+      "learning_rate": 4.983434854373858e-06,
+      "loss": 0.2442,
+      "step": 4866
+    },
+    {
+      "epoch": 2.728139013452915,
+      "grad_norm": 0.08106903881757009,
+      "learning_rate": 4.9631191652569465e-06,
+      "loss": 0.2543,
+      "step": 4867
+    },
+    {
+      "epoch": 2.7286995515695067,
+      "grad_norm": 0.08347005456982191,
+      "learning_rate": 4.942843916866435e-06,
+      "loss": 0.2449,
+      "step": 4868
+    },
+    {
+      "epoch": 2.7292600896860986,
+      "grad_norm": 0.08170344258225214,
+      "learning_rate": 4.922609117829946e-06,
+      "loss": 0.2391,
+      "step": 4869
+    },
+    {
+      "epoch": 2.729820627802691,
+      "grad_norm": 0.0797622584057896,
+      "learning_rate": 4.902414776757924e-06,
+      "loss": 0.228,
+      "step": 4870
+    },
+    {
+      "epoch": 2.7303811659192823,
+      "grad_norm": 0.08062118472366343,
+      "learning_rate": 4.88226090224364e-06,
+      "loss": 0.2357,
+      "step": 4871
+    },
+    {
+      "epoch": 2.7309417040358746,
+      "grad_norm": 0.08068485482047623,
+      "learning_rate": 4.862147502863057e-06,
+      "loss": 0.247,
+      "step": 4872
+    },
+    {
+      "epoch": 2.7315022421524664,
+      "grad_norm": 0.07981616688176582,
+      "learning_rate": 4.842074587175005e-06,
+      "loss": 0.2365,
+      "step": 4873
+    },
+    {
+      "epoch": 2.7320627802690582,
+      "grad_norm": 0.07694542358170804,
+      "learning_rate": 4.8220421637209965e-06,
+      "loss": 0.2302,
+      "step": 4874
+    },
+    {
+      "epoch": 2.73262331838565,
+      "grad_norm": 0.08008372601815365,
+      "learning_rate": 4.802050241025413e-06,
+      "loss": 0.2368,
+      "step": 4875
+    },
+    {
+      "epoch": 2.733183856502242,
+      "grad_norm": 0.07925762672758445,
+      "learning_rate": 4.7820988275953045e-06,
+      "loss": 0.2423,
+      "step": 4876
+    },
+    {
+      "epoch": 2.7337443946188342,
+      "grad_norm": 0.08165250402638714,
+      "learning_rate": 4.762187931920581e-06,
+      "loss": 0.2471,
+      "step": 4877
+    },
+    {
+      "epoch": 2.734304932735426,
+      "grad_norm": 0.0783592756245194,
+      "learning_rate": 4.742317562473797e-06,
+      "loss": 0.2415,
+      "step": 4878
+    },
+    {
+      "epoch": 2.734865470852018,
+      "grad_norm": 0.07911870695642753,
+      "learning_rate": 4.722487727710368e-06,
+      "loss": 0.2318,
+      "step": 4879
+    },
+    {
+      "epoch": 2.7354260089686098,
+      "grad_norm": 0.08067448083031267,
+      "learning_rate": 4.7026984360684205e-06,
+      "loss": 0.2455,
+      "step": 4880
+    },
+    {
+      "epoch": 2.7359865470852016,
+      "grad_norm": 0.07891274357847908,
+      "learning_rate": 4.6829496959687855e-06,
+      "loss": 0.2363,
+      "step": 4881
+    },
+    {
+      "epoch": 2.736547085201794,
+      "grad_norm": 0.07844329612266418,
+      "learning_rate": 4.663241515815131e-06,
+      "loss": 0.2333,
+      "step": 4882
+    },
+    {
+      "epoch": 2.7371076233183858,
+      "grad_norm": 0.07957756570103335,
+      "learning_rate": 4.64357390399377e-06,
+      "loss": 0.2385,
+      "step": 4883
+    },
+    {
+      "epoch": 2.7376681614349776,
+      "grad_norm": 0.08250791774164136,
+      "learning_rate": 4.623946868873819e-06,
+      "loss": 0.2438,
+      "step": 4884
+    },
+    {
+      "epoch": 2.7382286995515694,
+      "grad_norm": 0.078887470344168,
+      "learning_rate": 4.604360418807108e-06,
+      "loss": 0.2367,
+      "step": 4885
+    },
+    {
+      "epoch": 2.7387892376681613,
+      "grad_norm": 0.0799538042776628,
+      "learning_rate": 4.584814562128159e-06,
+      "loss": 0.2355,
+      "step": 4886
+    },
+    {
+      "epoch": 2.7393497757847536,
+      "grad_norm": 0.07890725023986828,
+      "learning_rate": 4.565309307154286e-06,
+      "loss": 0.2317,
+      "step": 4887
+    },
+    {
+      "epoch": 2.7399103139013454,
+      "grad_norm": 0.08039816000780453,
+      "learning_rate": 4.5458446621854945e-06,
+      "loss": 0.2449,
+      "step": 4888
+    },
+    {
+      "epoch": 2.7404708520179373,
+      "grad_norm": 0.08266772222645542,
+      "learning_rate": 4.526420635504502e-06,
+      "loss": 0.2388,
+      "step": 4889
+    },
+    {
+      "epoch": 2.741031390134529,
+      "grad_norm": 0.07847399945650287,
+      "learning_rate": 4.507037235376754e-06,
+      "loss": 0.2403,
+      "step": 4890
+    },
+    {
+      "epoch": 2.741591928251121,
+      "grad_norm": 0.0779826784899804,
+      "learning_rate": 4.487694470050408e-06,
+      "loss": 0.2375,
+      "step": 4891
+    },
+    {
+      "epoch": 2.7421524663677133,
+      "grad_norm": 0.07834341959171157,
+      "learning_rate": 4.468392347756312e-06,
+      "loss": 0.2252,
+      "step": 4892
+    },
+    {
+      "epoch": 2.7427130044843047,
+      "grad_norm": 0.081313239708545,
+      "learning_rate": 4.44913087670803e-06,
+      "loss": 0.2454,
+      "step": 4893
+    },
+    {
+      "epoch": 2.743273542600897,
+      "grad_norm": 0.0813023161291648,
+      "learning_rate": 4.42991006510185e-06,
+      "loss": 0.2528,
+      "step": 4894
+    },
+    {
+      "epoch": 2.743834080717489,
+      "grad_norm": 0.08061347706009954,
+      "learning_rate": 4.41072992111673e-06,
+      "loss": 0.2375,
+      "step": 4895
+    },
+    {
+      "epoch": 2.7443946188340806,
+      "grad_norm": 0.08317858229039553,
+      "learning_rate": 4.391590452914352e-06,
+      "loss": 0.2491,
+      "step": 4896
+    },
+    {
+      "epoch": 2.7449551569506725,
+      "grad_norm": 0.0829078123381164,
+      "learning_rate": 4.372491668639034e-06,
+      "loss": 0.237,
+      "step": 4897
+    },
+    {
+      "epoch": 2.7455156950672643,
+      "grad_norm": 0.08108923857329119,
+      "learning_rate": 4.3534335764178536e-06,
+      "loss": 0.2399,
+      "step": 4898
+    },
+    {
+      "epoch": 2.7460762331838566,
+      "grad_norm": 0.07967048316211159,
+      "learning_rate": 4.334416184360512e-06,
+      "loss": 0.2489,
+      "step": 4899
+    },
+    {
+      "epoch": 2.7466367713004485,
+      "grad_norm": 0.07937350424281729,
+      "learning_rate": 4.315439500559426e-06,
+      "loss": 0.2461,
+      "step": 4900
+    },
+    {
+      "epoch": 2.7471973094170403,
+      "grad_norm": 0.07761947036514329,
+      "learning_rate": 4.29650353308968e-06,
+      "loss": 0.2271,
+      "step": 4901
+    },
+    {
+      "epoch": 2.747757847533632,
+      "grad_norm": 0.07857183843296289,
+      "learning_rate": 4.277608290009027e-06,
+      "loss": 0.2383,
+      "step": 4902
+    },
+    {
+      "epoch": 2.748318385650224,
+      "grad_norm": 0.08140988844295936,
+      "learning_rate": 4.258753779357904e-06,
+      "loss": 0.2497,
+      "step": 4903
+    },
+    {
+      "epoch": 2.7488789237668163,
+      "grad_norm": 0.07894403613544215,
+      "learning_rate": 4.2399400091594154e-06,
+      "loss": 0.2321,
+      "step": 4904
+    },
+    {
+      "epoch": 2.749439461883408,
+      "grad_norm": 0.08116152376529585,
+      "learning_rate": 4.221166987419289e-06,
+      "loss": 0.2508,
+      "step": 4905
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.08105910638593729,
+      "learning_rate": 4.202434722125992e-06,
+      "loss": 0.2388,
+      "step": 4906
+    },
+    {
+      "epoch": 2.750560538116592,
+      "grad_norm": 0.08004397586876048,
+      "learning_rate": 4.183743221250569e-06,
+      "loss": 0.2395,
+      "step": 4907
+    },
+    {
+      "epoch": 2.7511210762331837,
+      "grad_norm": 0.08312500342586408,
+      "learning_rate": 4.16509249274678e-06,
+      "loss": 0.2434,
+      "step": 4908
+    },
+    {
+      "epoch": 2.751681614349776,
+      "grad_norm": 0.08453968282838248,
+      "learning_rate": 4.146482544550967e-06,
+      "loss": 0.2543,
+      "step": 4909
+    },
+    {
+      "epoch": 2.752242152466368,
+      "grad_norm": 0.08298475355558854,
+      "learning_rate": 4.127913384582205e-06,
+      "loss": 0.257,
+      "step": 4910
+    },
+    {
+      "epoch": 2.7528026905829597,
+      "grad_norm": 0.0841434311601157,
+      "learning_rate": 4.109385020742118e-06,
+      "loss": 0.2459,
+      "step": 4911
+    },
+    {
+      "epoch": 2.7533632286995515,
+      "grad_norm": 0.07996229214808133,
+      "learning_rate": 4.090897460915055e-06,
+      "loss": 0.2409,
+      "step": 4912
+    },
+    {
+      "epoch": 2.7539237668161434,
+      "grad_norm": 0.08045599167074863,
+      "learning_rate": 4.0724507129679676e-06,
+      "loss": 0.2466,
+      "step": 4913
+    },
+    {
+      "epoch": 2.7544843049327357,
+      "grad_norm": 0.07993007824493323,
+      "learning_rate": 4.0540447847504105e-06,
+      "loss": 0.2461,
+      "step": 4914
+    },
+    {
+      "epoch": 2.7550448430493275,
+      "grad_norm": 0.08018779590258114,
+      "learning_rate": 4.0356796840946286e-06,
+      "loss": 0.245,
+      "step": 4915
+    },
+    {
+      "epoch": 2.7556053811659194,
+      "grad_norm": 0.08176447835814545,
+      "learning_rate": 4.017355418815427e-06,
+      "loss": 0.2493,
+      "step": 4916
+    },
+    {
+      "epoch": 2.756165919282511,
+      "grad_norm": 0.07955554783291073,
+      "learning_rate": 3.999071996710313e-06,
+      "loss": 0.2432,
+      "step": 4917
+    },
+    {
+      "epoch": 2.756726457399103,
+      "grad_norm": 0.08050780902204767,
+      "learning_rate": 3.980829425559329e-06,
+      "loss": 0.2359,
+      "step": 4918
+    },
+    {
+      "epoch": 2.7572869955156953,
+      "grad_norm": 0.0808025743924573,
+      "learning_rate": 3.962627713125189e-06,
+      "loss": 0.2469,
+      "step": 4919
+    },
+    {
+      "epoch": 2.7578475336322867,
+      "grad_norm": 0.0800776363391883,
+      "learning_rate": 3.944466867153218e-06,
+      "loss": 0.246,
+      "step": 4920
+    },
+    {
+      "epoch": 2.758408071748879,
+      "grad_norm": 0.08208451543057813,
+      "learning_rate": 3.926346895371313e-06,
+      "loss": 0.2507,
+      "step": 4921
+    },
+    {
+      "epoch": 2.758968609865471,
+      "grad_norm": 0.08027660809899927,
+      "learning_rate": 3.908267805490051e-06,
+      "loss": 0.2283,
+      "step": 4922
+    },
+    {
+      "epoch": 2.7595291479820627,
+      "grad_norm": 0.0808482546151297,
+      "learning_rate": 3.890229605202522e-06,
+      "loss": 0.2481,
+      "step": 4923
+    },
+    {
+      "epoch": 2.7600896860986546,
+      "grad_norm": 0.07994689251204488,
+      "learning_rate": 3.872232302184487e-06,
+      "loss": 0.2373,
+      "step": 4924
+    },
+    {
+      "epoch": 2.7606502242152464,
+      "grad_norm": 0.07699473357642464,
+      "learning_rate": 3.8542759040942734e-06,
+      "loss": 0.2351,
+      "step": 4925
+    },
+    {
+      "epoch": 2.7612107623318387,
+      "grad_norm": 0.08243198715586861,
+      "learning_rate": 3.836360418572793e-06,
+      "loss": 0.2461,
+      "step": 4926
+    },
+    {
+      "epoch": 2.7617713004484306,
+      "grad_norm": 0.07884934050552518,
+      "learning_rate": 3.81848585324357e-06,
+      "loss": 0.2292,
+      "step": 4927
+    },
+    {
+      "epoch": 2.7623318385650224,
+      "grad_norm": 0.08077545761857474,
+      "learning_rate": 3.8006522157127078e-06,
+      "loss": 0.2308,
+      "step": 4928
+    },
+    {
+      "epoch": 2.7628923766816142,
+      "grad_norm": 0.08021670897868287,
+      "learning_rate": 3.782859513568915e-06,
+      "loss": 0.2404,
+      "step": 4929
+    },
+    {
+      "epoch": 2.763452914798206,
+      "grad_norm": 0.08416006061751108,
+      "learning_rate": 3.7651077543834346e-06,
+      "loss": 0.2592,
+      "step": 4930
+    },
+    {
+      "epoch": 2.7640134529147984,
+      "grad_norm": 0.083454558547178,
+      "learning_rate": 3.7473969457101356e-06,
+      "loss": 0.2442,
+      "step": 4931
+    },
+    {
+      "epoch": 2.7645739910313902,
+      "grad_norm": 0.08028477042441148,
+      "learning_rate": 3.729727095085422e-06,
+      "loss": 0.242,
+      "step": 4932
+    },
+    {
+      "epoch": 2.765134529147982,
+      "grad_norm": 0.08061222581489345,
+      "learning_rate": 3.712098210028281e-06,
+      "loss": 0.2427,
+      "step": 4933
+    },
+    {
+      "epoch": 2.765695067264574,
+      "grad_norm": 0.0784356871893266,
+      "learning_rate": 3.694510298040288e-06,
+      "loss": 0.2312,
+      "step": 4934
+    },
+    {
+      "epoch": 2.7662556053811658,
+      "grad_norm": 0.07930054523012602,
+      "learning_rate": 3.676963366605557e-06,
+      "loss": 0.2401,
+      "step": 4935
+    },
+    {
+      "epoch": 2.766816143497758,
+      "grad_norm": 0.08276631730613505,
+      "learning_rate": 3.659457423190782e-06,
+      "loss": 0.2507,
+      "step": 4936
+    },
+    {
+      "epoch": 2.76737668161435,
+      "grad_norm": 0.07760277060054778,
+      "learning_rate": 3.641992475245204e-06,
+      "loss": 0.2373,
+      "step": 4937
+    },
+    {
+      "epoch": 2.7679372197309418,
+      "grad_norm": 0.07999184808020934,
+      "learning_rate": 3.6245685302006447e-06,
+      "loss": 0.2373,
+      "step": 4938
+    },
+    {
+      "epoch": 2.7684977578475336,
+      "grad_norm": 0.07937605545005519,
+      "learning_rate": 3.6071855954714406e-06,
+      "loss": 0.2305,
+      "step": 4939
+    },
+    {
+      "epoch": 2.7690582959641254,
+      "grad_norm": 0.0776053555961307,
+      "learning_rate": 3.5898436784544854e-06,
+      "loss": 0.2403,
+      "step": 4940
+    },
+    {
+      "epoch": 2.7696188340807177,
+      "grad_norm": 0.07961889367097014,
+      "learning_rate": 3.572542786529243e-06,
+      "loss": 0.2321,
+      "step": 4941
+    },
+    {
+      "epoch": 2.770179372197309,
+      "grad_norm": 0.0793727827627126,
+      "learning_rate": 3.5552829270576792e-06,
+      "loss": 0.2332,
+      "step": 4942
+    },
+    {
+      "epoch": 2.7707399103139014,
+      "grad_norm": 0.08138300071359784,
+      "learning_rate": 3.5380641073843645e-06,
+      "loss": 0.2343,
+      "step": 4943
+    },
+    {
+      "epoch": 2.7713004484304933,
+      "grad_norm": 0.08105653688643481,
+      "learning_rate": 3.5208863348363263e-06,
+      "loss": 0.2344,
+      "step": 4944
+    },
+    {
+      "epoch": 2.771860986547085,
+      "grad_norm": 0.08219759076109563,
+      "learning_rate": 3.503749616723173e-06,
+      "loss": 0.2441,
+      "step": 4945
+    },
+    {
+      "epoch": 2.772421524663677,
+      "grad_norm": 0.07948917468362943,
+      "learning_rate": 3.4866539603370605e-06,
+      "loss": 0.2536,
+      "step": 4946
+    },
+    {
+      "epoch": 2.772982062780269,
+      "grad_norm": 0.07940583597981922,
+      "learning_rate": 3.4695993729526254e-06,
+      "loss": 0.2375,
+      "step": 4947
+    },
+    {
+      "epoch": 2.773542600896861,
+      "grad_norm": 0.07809825453838833,
+      "learning_rate": 3.4525858618270625e-06,
+      "loss": 0.2334,
+      "step": 4948
+    },
+    {
+      "epoch": 2.774103139013453,
+      "grad_norm": 0.0790630400089733,
+      "learning_rate": 3.4356134342000467e-06,
+      "loss": 0.2334,
+      "step": 4949
+    },
+    {
+      "epoch": 2.774663677130045,
+      "grad_norm": 0.07968962642906242,
+      "learning_rate": 3.418682097293835e-06,
+      "loss": 0.2412,
+      "step": 4950
+    },
+    {
+      "epoch": 2.7752242152466366,
+      "grad_norm": 0.081772312236787,
+      "learning_rate": 3.4017918583131414e-06,
+      "loss": 0.2537,
+      "step": 4951
+    },
+    {
+      "epoch": 2.7757847533632285,
+      "grad_norm": 0.07791353466848155,
+      "learning_rate": 3.384942724445195e-06,
+      "loss": 0.2387,
+      "step": 4952
+    },
+    {
+      "epoch": 2.776345291479821,
+      "grad_norm": 0.07982942110717695,
+      "learning_rate": 3.368134702859782e-06,
+      "loss": 0.2351,
+      "step": 4953
+    },
+    {
+      "epoch": 2.7769058295964126,
+      "grad_norm": 0.0794594563417086,
+      "learning_rate": 3.3513678007091596e-06,
+      "loss": 0.2436,
+      "step": 4954
+    },
+    {
+      "epoch": 2.7774663677130045,
+      "grad_norm": 0.0809108170869888,
+      "learning_rate": 3.3346420251280876e-06,
+      "loss": 0.2414,
+      "step": 4955
+    },
+    {
+      "epoch": 2.7780269058295963,
+      "grad_norm": 0.0799346895705784,
+      "learning_rate": 3.317957383233816e-06,
+      "loss": 0.2338,
+      "step": 4956
+    },
+    {
+      "epoch": 2.778587443946188,
+      "grad_norm": 0.08209863494745351,
+      "learning_rate": 3.3013138821261336e-06,
+      "loss": 0.2361,
+      "step": 4957
+    },
+    {
+      "epoch": 2.7791479820627805,
+      "grad_norm": 0.08173173423039823,
+      "learning_rate": 3.284711528887274e-06,
+      "loss": 0.2452,
+      "step": 4958
+    },
+    {
+      "epoch": 2.7797085201793723,
+      "grad_norm": 0.07815383566511941,
+      "learning_rate": 3.268150330581976e-06,
+      "loss": 0.225,
+      "step": 4959
+    },
+    {
+      "epoch": 2.780269058295964,
+      "grad_norm": 0.0768024051910523,
+      "learning_rate": 3.2516302942574793e-06,
+      "loss": 0.2328,
+      "step": 4960
+    },
+    {
+      "epoch": 2.780829596412556,
+      "grad_norm": 0.08145432874610853,
+      "learning_rate": 3.2351514269434945e-06,
+      "loss": 0.2395,
+      "step": 4961
+    },
+    {
+      "epoch": 2.781390134529148,
+      "grad_norm": 0.07791961234583099,
+      "learning_rate": 3.2187137356522346e-06,
+      "loss": 0.2363,
+      "step": 4962
+    },
+    {
+      "epoch": 2.78195067264574,
+      "grad_norm": 0.08034275923951512,
+      "learning_rate": 3.2023172273783486e-06,
+      "loss": 0.2483,
+      "step": 4963
+    },
+    {
+      "epoch": 2.782511210762332,
+      "grad_norm": 0.08040564210423004,
+      "learning_rate": 3.1859619090990222e-06,
+      "loss": 0.2408,
+      "step": 4964
+    },
+    {
+      "epoch": 2.783071748878924,
+      "grad_norm": 0.08026860440316434,
+      "learning_rate": 3.169647787773866e-06,
+      "loss": 0.2346,
+      "step": 4965
+    },
+    {
+      "epoch": 2.7836322869955157,
+      "grad_norm": 0.07995657738869455,
+      "learning_rate": 3.1533748703449494e-06,
+      "loss": 0.2388,
+      "step": 4966
+    },
+    {
+      "epoch": 2.7841928251121075,
+      "grad_norm": 0.07816803033206252,
+      "learning_rate": 3.1371431637368665e-06,
+      "loss": 0.2401,
+      "step": 4967
+    },
+    {
+      "epoch": 2.7847533632287,
+      "grad_norm": 0.08108139592489091,
+      "learning_rate": 3.120952674856614e-06,
+      "loss": 0.2377,
+      "step": 4968
+    },
+    {
+      "epoch": 2.785313901345291,
+      "grad_norm": 0.07894030303325457,
+      "learning_rate": 3.104803410593693e-06,
+      "loss": 0.2385,
+      "step": 4969
+    },
+    {
+      "epoch": 2.7858744394618835,
+      "grad_norm": 0.07766078483546503,
+      "learning_rate": 3.0886953778200277e-06,
+      "loss": 0.2283,
+      "step": 4970
+    },
+    {
+      "epoch": 2.7864349775784754,
+      "grad_norm": 0.08313766543340664,
+      "learning_rate": 3.0726285833900583e-06,
+      "loss": 0.2443,
+      "step": 4971
+    },
+    {
+      "epoch": 2.786995515695067,
+      "grad_norm": 0.0783837382802284,
+      "learning_rate": 3.0566030341405925e-06,
+      "loss": 0.2358,
+      "step": 4972
+    },
+    {
+      "epoch": 2.787556053811659,
+      "grad_norm": 0.07980212385524099,
+      "learning_rate": 3.0406187368909435e-06,
+      "loss": 0.2521,
+      "step": 4973
+    },
+    {
+      "epoch": 2.788116591928251,
+      "grad_norm": 0.08193651596707059,
+      "learning_rate": 3.0246756984428582e-06,
+      "loss": 0.2464,
+      "step": 4974
+    },
+    {
+      "epoch": 2.788677130044843,
+      "grad_norm": 0.07896535840835087,
+      "learning_rate": 3.0087739255804993e-06,
+      "loss": 0.2417,
+      "step": 4975
+    },
+    {
+      "epoch": 2.789237668161435,
+      "grad_norm": 0.08002503693440731,
+      "learning_rate": 2.9929134250705427e-06,
+      "loss": 0.2398,
+      "step": 4976
+    },
+    {
+      "epoch": 2.789798206278027,
+      "grad_norm": 0.08082508620144269,
+      "learning_rate": 2.977094203662012e-06,
+      "loss": 0.2407,
+      "step": 4977
+    },
+    {
+      "epoch": 2.7903587443946187,
+      "grad_norm": 0.08292065152762618,
+      "learning_rate": 2.9613162680864224e-06,
+      "loss": 0.2462,
+      "step": 4978
+    },
+    {
+      "epoch": 2.7909192825112106,
+      "grad_norm": 0.08050626144783643,
+      "learning_rate": 2.945579625057715e-06,
+      "loss": 0.2475,
+      "step": 4979
+    },
+    {
+      "epoch": 2.791479820627803,
+      "grad_norm": 0.08182481488605035,
+      "learning_rate": 2.9298842812722327e-06,
+      "loss": 0.2523,
+      "step": 4980
+    },
+    {
+      "epoch": 2.7920403587443947,
+      "grad_norm": 0.08249741359834577,
+      "learning_rate": 2.914230243408789e-06,
+      "loss": 0.2354,
+      "step": 4981
+    },
+    {
+      "epoch": 2.7926008968609866,
+      "grad_norm": 0.08086698775520357,
+      "learning_rate": 2.898617518128566e-06,
+      "loss": 0.2386,
+      "step": 4982
+    },
+    {
+      "epoch": 2.7931614349775784,
+      "grad_norm": 0.08314237452585792,
+      "learning_rate": 2.8830461120752163e-06,
+      "loss": 0.2424,
+      "step": 4983
+    },
+    {
+      "epoch": 2.7937219730941703,
+      "grad_norm": 0.0791477644478879,
+      "learning_rate": 2.8675160318747727e-06,
+      "loss": 0.238,
+      "step": 4984
+    },
+    {
+      "epoch": 2.7942825112107625,
+      "grad_norm": 0.08031181024209258,
+      "learning_rate": 2.8520272841357055e-06,
+      "loss": 0.2421,
+      "step": 4985
+    },
+    {
+      "epoch": 2.7948430493273544,
+      "grad_norm": 0.0810401880716221,
+      "learning_rate": 2.836579875448886e-06,
+      "loss": 0.2407,
+      "step": 4986
+    },
+    {
+      "epoch": 2.7954035874439462,
+      "grad_norm": 0.08004090489087526,
+      "learning_rate": 2.8211738123876006e-06,
+      "loss": 0.2478,
+      "step": 4987
+    },
+    {
+      "epoch": 2.795964125560538,
+      "grad_norm": 0.08115702823252946,
+      "learning_rate": 2.8058091015075394e-06,
+      "loss": 0.2436,
+      "step": 4988
+    },
+    {
+      "epoch": 2.79652466367713,
+      "grad_norm": 0.08034753708061859,
+      "learning_rate": 2.790485749346805e-06,
+      "loss": 0.2381,
+      "step": 4989
+    },
+    {
+      "epoch": 2.797085201793722,
+      "grad_norm": 0.08439626147729173,
+      "learning_rate": 2.775203762425882e-06,
+      "loss": 0.2462,
+      "step": 4990
+    },
+    {
+      "epoch": 2.797645739910314,
+      "grad_norm": 0.07875450477637529,
+      "learning_rate": 2.7599631472476683e-06,
+      "loss": 0.2319,
+      "step": 4991
+    },
+    {
+      "epoch": 2.798206278026906,
+      "grad_norm": 0.08115777020139331,
+      "learning_rate": 2.7447639102974434e-06,
+      "loss": 0.231,
+      "step": 4992
+    },
+    {
+      "epoch": 2.7987668161434978,
+      "grad_norm": 0.08316558015368568,
+      "learning_rate": 2.7296060580428885e-06,
+      "loss": 0.2424,
+      "step": 4993
+    },
+    {
+      "epoch": 2.7993273542600896,
+      "grad_norm": 0.08079652698760205,
+      "learning_rate": 2.714489596934089e-06,
+      "loss": 0.2371,
+      "step": 4994
+    },
+    {
+      "epoch": 2.7998878923766815,
+      "grad_norm": 0.0812179868460853,
+      "learning_rate": 2.6994145334034994e-06,
+      "loss": 0.2427,
+      "step": 4995
+    },
+    {
+      "epoch": 2.8004484304932733,
+      "grad_norm": 0.0812805787555618,
+      "learning_rate": 2.6843808738659324e-06,
+      "loss": 0.2479,
+      "step": 4996
+    },
+    {
+      "epoch": 2.8010089686098656,
+      "grad_norm": 0.07843320292716413,
+      "learning_rate": 2.6693886247186605e-06,
+      "loss": 0.236,
+      "step": 4997
+    },
+    {
+      "epoch": 2.8015695067264574,
+      "grad_norm": 0.08061411163696575,
+      "learning_rate": 2.6544377923412465e-06,
+      "loss": 0.2504,
+      "step": 4998
+    },
+    {
+      "epoch": 2.8021300448430493,
+      "grad_norm": 0.07930131566618949,
+      "learning_rate": 2.6395283830956686e-06,
+      "loss": 0.2377,
+      "step": 4999
+    },
+    {
+      "epoch": 2.802690582959641,
+      "grad_norm": 0.08071531139627736,
+      "learning_rate": 2.6246604033262954e-06,
+      "loss": 0.2421,
+      "step": 5000
+    },
+    {
+      "epoch": 2.803251121076233,
+      "grad_norm": 0.07931495086159082,
+      "learning_rate": 2.6098338593598447e-06,
+      "loss": 0.2258,
+      "step": 5001
+    },
+    {
+      "epoch": 2.8038116591928253,
+      "grad_norm": 0.08094859462018693,
+      "learning_rate": 2.595048757505392e-06,
+      "loss": 0.2392,
+      "step": 5002
+    },
+    {
+      "epoch": 2.804372197309417,
+      "grad_norm": 0.07992420918468417,
+      "learning_rate": 2.5803051040544146e-06,
+      "loss": 0.2275,
+      "step": 5003
+    },
+    {
+      "epoch": 2.804932735426009,
+      "grad_norm": 0.0799245025277946,
+      "learning_rate": 2.565602905280717e-06,
+      "loss": 0.2401,
+      "step": 5004
+    },
+    {
+      "epoch": 2.805493273542601,
+      "grad_norm": 0.07856487841272483,
+      "learning_rate": 2.5509421674404844e-06,
+      "loss": 0.2313,
+      "step": 5005
+    },
+    {
+      "epoch": 2.8060538116591927,
+      "grad_norm": 0.07885956821847813,
+      "learning_rate": 2.5363228967722364e-06,
+      "loss": 0.2449,
+      "step": 5006
+    },
+    {
+      "epoch": 2.806614349775785,
+      "grad_norm": 0.07722484258529054,
+      "learning_rate": 2.521745099496886e-06,
+      "loss": 0.2427,
+      "step": 5007
+    },
+    {
+      "epoch": 2.807174887892377,
+      "grad_norm": 0.07920463080690886,
+      "learning_rate": 2.5072087818176382e-06,
+      "loss": 0.2326,
+      "step": 5008
+    },
+    {
+      "epoch": 2.8077354260089686,
+      "grad_norm": 0.07947998344169381,
+      "learning_rate": 2.4927139499201225e-06,
+      "loss": 0.23,
+      "step": 5009
+    },
+    {
+      "epoch": 2.8082959641255605,
+      "grad_norm": 0.08173714135267525,
+      "learning_rate": 2.4782606099722606e-06,
+      "loss": 0.247,
+      "step": 5010
+    },
+    {
+      "epoch": 2.8088565022421523,
+      "grad_norm": 0.07981986410717254,
+      "learning_rate": 2.4638487681243215e-06,
+      "loss": 0.2236,
+      "step": 5011
+    },
+    {
+      "epoch": 2.8094170403587446,
+      "grad_norm": 0.07642627727368918,
+      "learning_rate": 2.4494784305089557e-06,
+      "loss": 0.238,
+      "step": 5012
+    },
+    {
+      "epoch": 2.8099775784753365,
+      "grad_norm": 0.08232066180554665,
+      "learning_rate": 2.4351496032410938e-06,
+      "loss": 0.2414,
+      "step": 5013
+    },
+    {
+      "epoch": 2.8105381165919283,
+      "grad_norm": 0.08064050801326456,
+      "learning_rate": 2.4208622924180578e-06,
+      "loss": 0.2406,
+      "step": 5014
+    },
+    {
+      "epoch": 2.81109865470852,
+      "grad_norm": 0.08195562292195246,
+      "learning_rate": 2.406616504119463e-06,
+      "loss": 0.2476,
+      "step": 5015
+    },
+    {
+      "epoch": 2.811659192825112,
+      "grad_norm": 0.08059547939756304,
+      "learning_rate": 2.392412244407294e-06,
+      "loss": 0.2513,
+      "step": 5016
+    },
+    {
+      "epoch": 2.8122197309417043,
+      "grad_norm": 0.08216416650635741,
+      "learning_rate": 2.3782495193258147e-06,
+      "loss": 0.2558,
+      "step": 5017
+    },
+    {
+      "epoch": 2.8127802690582957,
+      "grad_norm": 0.08068733347935715,
+      "learning_rate": 2.3641283349016607e-06,
+      "loss": 0.258,
+      "step": 5018
+    },
+    {
+      "epoch": 2.813340807174888,
+      "grad_norm": 0.08050818728921794,
+      "learning_rate": 2.3500486971437587e-06,
+      "loss": 0.2456,
+      "step": 5019
+    },
+    {
+      "epoch": 2.81390134529148,
+      "grad_norm": 0.08292945331512773,
+      "learning_rate": 2.336010612043382e-06,
+      "loss": 0.2437,
+      "step": 5020
+    },
+    {
+      "epoch": 2.8144618834080717,
+      "grad_norm": 0.07795781503050385,
+      "learning_rate": 2.322014085574109e-06,
+      "loss": 0.2386,
+      "step": 5021
+    },
+    {
+      "epoch": 2.8150224215246635,
+      "grad_norm": 0.08022008972424731,
+      "learning_rate": 2.3080591236918303e-06,
+      "loss": 0.2336,
+      "step": 5022
+    },
+    {
+      "epoch": 2.8155829596412554,
+      "grad_norm": 0.08247247207217011,
+      "learning_rate": 2.2941457323347627e-06,
+      "loss": 0.2506,
+      "step": 5023
+    },
+    {
+      "epoch": 2.8161434977578477,
+      "grad_norm": 0.08018762360263491,
+      "learning_rate": 2.2802739174234146e-06,
+      "loss": 0.2406,
+      "step": 5024
+    },
+    {
+      "epoch": 2.8167040358744395,
+      "grad_norm": 0.08380126760379156,
+      "learning_rate": 2.2664436848606194e-06,
+      "loss": 0.265,
+      "step": 5025
+    },
+    {
+      "epoch": 2.8172645739910314,
+      "grad_norm": 0.0800422086942214,
+      "learning_rate": 2.252655040531493e-06,
+      "loss": 0.2322,
+      "step": 5026
+    },
+    {
+      "epoch": 2.817825112107623,
+      "grad_norm": 0.07979590612087926,
+      "learning_rate": 2.238907990303496e-06,
+      "loss": 0.2359,
+      "step": 5027
+    },
+    {
+      "epoch": 2.818385650224215,
+      "grad_norm": 0.08201699136530352,
+      "learning_rate": 2.225202540026361e-06,
+      "loss": 0.2594,
+      "step": 5028
+    },
+    {
+      "epoch": 2.8189461883408073,
+      "grad_norm": 0.0830963486106769,
+      "learning_rate": 2.2115386955321004e-06,
+      "loss": 0.2448,
+      "step": 5029
+    },
+    {
+      "epoch": 2.819506726457399,
+      "grad_norm": 0.08119421109654049,
+      "learning_rate": 2.1979164626350748e-06,
+      "loss": 0.2481,
+      "step": 5030
+    },
+    {
+      "epoch": 2.820067264573991,
+      "grad_norm": 0.0809885995298249,
+      "learning_rate": 2.1843358471318908e-06,
+      "loss": 0.2436,
+      "step": 5031
+    },
+    {
+      "epoch": 2.820627802690583,
+      "grad_norm": 0.07844458178452507,
+      "learning_rate": 2.170796854801449e-06,
+      "loss": 0.242,
+      "step": 5032
+    },
+    {
+      "epoch": 2.8211883408071747,
+      "grad_norm": 0.07898023469165534,
+      "learning_rate": 2.1572994914049847e-06,
+      "loss": 0.234,
+      "step": 5033
+    },
+    {
+      "epoch": 2.821748878923767,
+      "grad_norm": 0.0823097354542868,
+      "learning_rate": 2.1438437626859487e-06,
+      "loss": 0.2403,
+      "step": 5034
+    },
+    {
+      "epoch": 2.822309417040359,
+      "grad_norm": 0.07985392689741619,
+      "learning_rate": 2.130429674370138e-06,
+      "loss": 0.227,
+      "step": 5035
+    },
+    {
+      "epoch": 2.8228699551569507,
+      "grad_norm": 0.08264550665738286,
+      "learning_rate": 2.1170572321655868e-06,
+      "loss": 0.2488,
+      "step": 5036
+    },
+    {
+      "epoch": 2.8234304932735426,
+      "grad_norm": 0.08096897766330863,
+      "learning_rate": 2.1037264417626544e-06,
+      "loss": 0.2502,
+      "step": 5037
+    },
+    {
+      "epoch": 2.8239910313901344,
+      "grad_norm": 0.08129709995477422,
+      "learning_rate": 2.0904373088339367e-06,
+      "loss": 0.2506,
+      "step": 5038
+    },
+    {
+      "epoch": 2.8245515695067267,
+      "grad_norm": 0.08062035437913692,
+      "learning_rate": 2.077189839034288e-06,
+      "loss": 0.2321,
+      "step": 5039
+    },
+    {
+      "epoch": 2.8251121076233185,
+      "grad_norm": 0.07923689907747232,
+      "learning_rate": 2.063984038000888e-06,
+      "loss": 0.2402,
+      "step": 5040
+    },
+    {
+      "epoch": 2.8256726457399104,
+      "grad_norm": 0.08120919706097315,
+      "learning_rate": 2.0508199113531414e-06,
+      "loss": 0.2444,
+      "step": 5041
+    },
+    {
+      "epoch": 2.8262331838565022,
+      "grad_norm": 0.08176687509396283,
+      "learning_rate": 2.037697464692756e-06,
+      "loss": 0.2376,
+      "step": 5042
+    },
+    {
+      "epoch": 2.826793721973094,
+      "grad_norm": 0.08126695881107068,
+      "learning_rate": 2.0246167036036543e-06,
+      "loss": 0.2372,
+      "step": 5043
+    },
+    {
+      "epoch": 2.827354260089686,
+      "grad_norm": 0.08124315561587099,
+      "learning_rate": 2.011577633652062e-06,
+      "loss": 0.2484,
+      "step": 5044
+    },
+    {
+      "epoch": 2.827914798206278,
+      "grad_norm": 0.0792594597102603,
+      "learning_rate": 1.9985802603864624e-06,
+      "loss": 0.2401,
+      "step": 5045
+    },
+    {
+      "epoch": 2.82847533632287,
+      "grad_norm": 0.08248713684791316,
+      "learning_rate": 1.9856245893375645e-06,
+      "loss": 0.241,
+      "step": 5046
+    },
+    {
+      "epoch": 2.829035874439462,
+      "grad_norm": 0.08223085773508047,
+      "learning_rate": 1.9727106260183704e-06,
+      "loss": 0.2318,
+      "step": 5047
+    },
+    {
+      "epoch": 2.8295964125560538,
+      "grad_norm": 0.08181175375406184,
+      "learning_rate": 1.9598383759240946e-06,
+      "loss": 0.2573,
+      "step": 5048
+    },
+    {
+      "epoch": 2.8301569506726456,
+      "grad_norm": 0.07889165704259438,
+      "learning_rate": 1.947007844532245e-06,
+      "loss": 0.2198,
+      "step": 5049
+    },
+    {
+      "epoch": 2.8307174887892375,
+      "grad_norm": 0.07950468343923922,
+      "learning_rate": 1.9342190373025313e-06,
+      "loss": 0.2343,
+      "step": 5050
+    },
+    {
+      "epoch": 2.8312780269058297,
+      "grad_norm": 0.07981220514558676,
+      "learning_rate": 1.921471959676957e-06,
+      "loss": 0.2442,
+      "step": 5051
+    },
+    {
+      "epoch": 2.8318385650224216,
+      "grad_norm": 0.079618243775057,
+      "learning_rate": 1.9087666170797267e-06,
+      "loss": 0.24,
+      "step": 5052
+    },
+    {
+      "epoch": 2.8323991031390134,
+      "grad_norm": 0.0785857151532339,
+      "learning_rate": 1.8961030149173054e-06,
+      "loss": 0.2476,
+      "step": 5053
+    },
+    {
+      "epoch": 2.8329596412556053,
+      "grad_norm": 0.08025682912235395,
+      "learning_rate": 1.883481158578404e-06,
+      "loss": 0.2477,
+      "step": 5054
+    },
+    {
+      "epoch": 2.833520179372197,
+      "grad_norm": 0.0788992016416964,
+      "learning_rate": 1.8709010534339378e-06,
+      "loss": 0.2346,
+      "step": 5055
+    },
+    {
+      "epoch": 2.8340807174887894,
+      "grad_norm": 0.07886702908057779,
+      "learning_rate": 1.8583627048371022e-06,
+      "loss": 0.2378,
+      "step": 5056
+    },
+    {
+      "epoch": 2.8346412556053813,
+      "grad_norm": 0.07768139483717154,
+      "learning_rate": 1.8458661181232739e-06,
+      "loss": 0.2207,
+      "step": 5057
+    },
+    {
+      "epoch": 2.835201793721973,
+      "grad_norm": 0.07668230535027755,
+      "learning_rate": 1.8334112986100992e-06,
+      "loss": 0.2359,
+      "step": 5058
+    },
+    {
+      "epoch": 2.835762331838565,
+      "grad_norm": 0.07892917109696834,
+      "learning_rate": 1.8209982515974277e-06,
+      "loss": 0.2346,
+      "step": 5059
+    },
+    {
+      "epoch": 2.836322869955157,
+      "grad_norm": 0.07841402349758807,
+      "learning_rate": 1.8086269823673563e-06,
+      "loss": 0.2339,
+      "step": 5060
+    },
+    {
+      "epoch": 2.836883408071749,
+      "grad_norm": 0.07847361926914799,
+      "learning_rate": 1.7962974961841738e-06,
+      "loss": 0.2315,
+      "step": 5061
+    },
+    {
+      "epoch": 2.837443946188341,
+      "grad_norm": 0.08110310021840367,
+      "learning_rate": 1.784009798294406e-06,
+      "loss": 0.2511,
+      "step": 5062
+    },
+    {
+      "epoch": 2.838004484304933,
+      "grad_norm": 0.08141770914014684,
+      "learning_rate": 1.7717638939268145e-06,
+      "loss": 0.237,
+      "step": 5063
+    },
+    {
+      "epoch": 2.8385650224215246,
+      "grad_norm": 0.08032709195173478,
+      "learning_rate": 1.7595597882923309e-06,
+      "loss": 0.2359,
+      "step": 5064
+    },
+    {
+      "epoch": 2.8391255605381165,
+      "grad_norm": 0.08112292362244095,
+      "learning_rate": 1.7473974865841569e-06,
+      "loss": 0.2299,
+      "step": 5065
+    },
+    {
+      "epoch": 2.839686098654709,
+      "grad_norm": 0.07803159972214607,
+      "learning_rate": 1.7352769939776526e-06,
+      "loss": 0.2328,
+      "step": 5066
+    },
+    {
+      "epoch": 2.8402466367713,
+      "grad_norm": 0.0785507265144912,
+      "learning_rate": 1.7231983156304144e-06,
+      "loss": 0.2367,
+      "step": 5067
+    },
+    {
+      "epoch": 2.8408071748878925,
+      "grad_norm": 0.0817662928210577,
+      "learning_rate": 1.711161456682242e-06,
+      "loss": 0.2434,
+      "step": 5068
+    },
+    {
+      "epoch": 2.8413677130044843,
+      "grad_norm": 0.07645522785547558,
+      "learning_rate": 1.6991664222551495e-06,
+      "loss": 0.2245,
+      "step": 5069
+    },
+    {
+      "epoch": 2.841928251121076,
+      "grad_norm": 0.08026739599895902,
+      "learning_rate": 1.6872132174533427e-06,
+      "loss": 0.2441,
+      "step": 5070
+    },
+    {
+      "epoch": 2.842488789237668,
+      "grad_norm": 0.08211132386376416,
+      "learning_rate": 1.6753018473632087e-06,
+      "loss": 0.2506,
+      "step": 5071
+    },
+    {
+      "epoch": 2.84304932735426,
+      "grad_norm": 0.08170934677712485,
+      "learning_rate": 1.6634323170533928e-06,
+      "loss": 0.2446,
+      "step": 5072
+    },
+    {
+      "epoch": 2.843609865470852,
+      "grad_norm": 0.07739677551438127,
+      "learning_rate": 1.6516046315746659e-06,
+      "loss": 0.2428,
+      "step": 5073
+    },
+    {
+      "epoch": 2.844170403587444,
+      "grad_norm": 0.08151001977790998,
+      "learning_rate": 1.639818795960013e-06,
+      "loss": 0.2469,
+      "step": 5074
+    },
+    {
+      "epoch": 2.844730941704036,
+      "grad_norm": 0.08105458687991796,
+      "learning_rate": 1.6280748152246562e-06,
+      "loss": 0.255,
+      "step": 5075
+    },
+    {
+      "epoch": 2.8452914798206277,
+      "grad_norm": 0.08235718212563223,
+      "learning_rate": 1.6163726943659419e-06,
+      "loss": 0.2458,
+      "step": 5076
+    },
+    {
+      "epoch": 2.8458520179372195,
+      "grad_norm": 0.07878870549768287,
+      "learning_rate": 1.6047124383634537e-06,
+      "loss": 0.2382,
+      "step": 5077
+    },
+    {
+      "epoch": 2.846412556053812,
+      "grad_norm": 0.08324852404587757,
+      "learning_rate": 1.593094052178945e-06,
+      "loss": 0.2397,
+      "step": 5078
+    },
+    {
+      "epoch": 2.8469730941704037,
+      "grad_norm": 0.08146821956937725,
+      "learning_rate": 1.5815175407563165e-06,
+      "loss": 0.233,
+      "step": 5079
+    },
+    {
+      "epoch": 2.8475336322869955,
+      "grad_norm": 0.08099050094226903,
+      "learning_rate": 1.5699829090217278e-06,
+      "loss": 0.2408,
+      "step": 5080
+    },
+    {
+      "epoch": 2.8480941704035874,
+      "grad_norm": 0.07791236653043693,
+      "learning_rate": 1.5584901618834301e-06,
+      "loss": 0.2382,
+      "step": 5081
+    },
+    {
+      "epoch": 2.848654708520179,
+      "grad_norm": 0.08058011888194505,
+      "learning_rate": 1.5470393042319232e-06,
+      "loss": 0.238,
+      "step": 5082
+    },
+    {
+      "epoch": 2.8492152466367715,
+      "grad_norm": 0.0821254894649332,
+      "learning_rate": 1.535630340939842e-06,
+      "loss": 0.2481,
+      "step": 5083
+    },
+    {
+      "epoch": 2.8497757847533634,
+      "grad_norm": 0.08206897357919915,
+      "learning_rate": 1.5242632768619925e-06,
+      "loss": 0.2369,
+      "step": 5084
+    },
+    {
+      "epoch": 2.850336322869955,
+      "grad_norm": 0.0786350321647868,
+      "learning_rate": 1.512938116835394e-06,
+      "loss": 0.24,
+      "step": 5085
+    },
+    {
+      "epoch": 2.850896860986547,
+      "grad_norm": 0.08234829259429416,
+      "learning_rate": 1.5016548656791697e-06,
+      "loss": 0.2566,
+      "step": 5086
+    },
+    {
+      "epoch": 2.851457399103139,
+      "grad_norm": 0.08187607263505849,
+      "learning_rate": 1.4904135281946673e-06,
+      "loss": 0.2459,
+      "step": 5087
+    },
+    {
+      "epoch": 2.852017937219731,
+      "grad_norm": 0.07914659198871633,
+      "learning_rate": 1.4792141091653612e-06,
+      "loss": 0.2366,
+      "step": 5088
+    },
+    {
+      "epoch": 2.852578475336323,
+      "grad_norm": 0.08124554504977664,
+      "learning_rate": 1.4680566133569162e-06,
+      "loss": 0.2418,
+      "step": 5089
+    },
+    {
+      "epoch": 2.853139013452915,
+      "grad_norm": 0.07868278459909037,
+      "learning_rate": 1.4569410455171351e-06,
+      "loss": 0.246,
+      "step": 5090
+    },
+    {
+      "epoch": 2.8536995515695067,
+      "grad_norm": 0.0795003399865061,
+      "learning_rate": 1.4458674103759894e-06,
+      "loss": 0.2565,
+      "step": 5091
+    },
+    {
+      "epoch": 2.8542600896860986,
+      "grad_norm": 0.08041450301498791,
+      "learning_rate": 1.4348357126456102e-06,
+      "loss": 0.2321,
+      "step": 5092
+    },
+    {
+      "epoch": 2.854820627802691,
+      "grad_norm": 0.08077531018250238,
+      "learning_rate": 1.4238459570202644e-06,
+      "loss": 0.2468,
+      "step": 5093
+    },
+    {
+      "epoch": 2.8553811659192823,
+      "grad_norm": 0.08090510985480488,
+      "learning_rate": 1.4128981481764115e-06,
+      "loss": 0.2482,
+      "step": 5094
+    },
+    {
+      "epoch": 2.8559417040358746,
+      "grad_norm": 0.0822852376229242,
+      "learning_rate": 1.4019922907726136e-06,
+      "loss": 0.2482,
+      "step": 5095
+    },
+    {
+      "epoch": 2.8565022421524664,
+      "grad_norm": 0.07866308000957997,
+      "learning_rate": 1.3911283894496253e-06,
+      "loss": 0.2267,
+      "step": 5096
+    },
+    {
+      "epoch": 2.8570627802690582,
+      "grad_norm": 0.08146261608076158,
+      "learning_rate": 1.380306448830293e-06,
+      "loss": 0.2414,
+      "step": 5097
+    },
+    {
+      "epoch": 2.85762331838565,
+      "grad_norm": 0.08291387306663744,
+      "learning_rate": 1.3695264735196778e-06,
+      "loss": 0.2613,
+      "step": 5098
+    },
+    {
+      "epoch": 2.858183856502242,
+      "grad_norm": 0.07746619006915174,
+      "learning_rate": 1.3587884681049322e-06,
+      "loss": 0.229,
+      "step": 5099
+    },
+    {
+      "epoch": 2.8587443946188342,
+      "grad_norm": 0.0811871605233219,
+      "learning_rate": 1.348092437155346e-06,
+      "loss": 0.2532,
+      "step": 5100
+    },
+    {
+      "epoch": 2.859304932735426,
+      "grad_norm": 0.08048551935917159,
+      "learning_rate": 1.3374383852223892e-06,
+      "loss": 0.2342,
+      "step": 5101
+    },
+    {
+      "epoch": 2.859865470852018,
+      "grad_norm": 0.08047930535275231,
+      "learning_rate": 1.3268263168396245e-06,
+      "loss": 0.2468,
+      "step": 5102
+    },
+    {
+      "epoch": 2.8604260089686098,
+      "grad_norm": 0.07982049190502127,
+      "learning_rate": 1.316256236522806e-06,
+      "loss": 0.2399,
+      "step": 5103
+    },
+    {
+      "epoch": 2.8609865470852016,
+      "grad_norm": 0.0808219184841395,
+      "learning_rate": 1.305728148769736e-06,
+      "loss": 0.2442,
+      "step": 5104
+    },
+    {
+      "epoch": 2.861547085201794,
+      "grad_norm": 0.08077620103395643,
+      "learning_rate": 1.295242058060442e-06,
+      "loss": 0.2438,
+      "step": 5105
+    },
+    {
+      "epoch": 2.8621076233183858,
+      "grad_norm": 0.07761821979349973,
+      "learning_rate": 1.28479796885701e-06,
+      "loss": 0.2398,
+      "step": 5106
+    },
+    {
+      "epoch": 2.8626681614349776,
+      "grad_norm": 0.0788037658984251,
+      "learning_rate": 1.2743958856036743e-06,
+      "loss": 0.2391,
+      "step": 5107
+    },
+    {
+      "epoch": 2.8632286995515694,
+      "grad_norm": 0.07908527218812027,
+      "learning_rate": 1.2640358127268049e-06,
+      "loss": 0.2407,
+      "step": 5108
+    },
+    {
+      "epoch": 2.8637892376681613,
+      "grad_norm": 0.07836661894689284,
+      "learning_rate": 1.2537177546348978e-06,
+      "loss": 0.2346,
+      "step": 5109
+    },
+    {
+      "epoch": 2.8643497757847536,
+      "grad_norm": 0.08349840583336998,
+      "learning_rate": 1.2434417157185519e-06,
+      "loss": 0.2485,
+      "step": 5110
+    },
+    {
+      "epoch": 2.8649103139013454,
+      "grad_norm": 0.07926798817133278,
+      "learning_rate": 1.2332077003505027e-06,
+      "loss": 0.2349,
+      "step": 5111
+    },
+    {
+      "epoch": 2.8654708520179373,
+      "grad_norm": 0.08018566682994367,
+      "learning_rate": 1.223015712885589e-06,
+      "loss": 0.2415,
+      "step": 5112
+    },
+    {
+      "epoch": 2.866031390134529,
+      "grad_norm": 0.0815649615936414,
+      "learning_rate": 1.2128657576607861e-06,
+      "loss": 0.2394,
+      "step": 5113
+    },
+    {
+      "epoch": 2.866591928251121,
+      "grad_norm": 0.07900698383651082,
+      "learning_rate": 1.2027578389951499e-06,
+      "loss": 0.2319,
+      "step": 5114
+    },
+    {
+      "epoch": 2.8671524663677133,
+      "grad_norm": 0.0781644852483143,
+      "learning_rate": 1.1926919611898847e-06,
+      "loss": 0.2418,
+      "step": 5115
+    },
+    {
+      "epoch": 2.8677130044843047,
+      "grad_norm": 0.07902751814612542,
+      "learning_rate": 1.182668128528286e-06,
+      "loss": 0.2289,
+      "step": 5116
+    },
+    {
+      "epoch": 2.868273542600897,
+      "grad_norm": 0.07763216154627357,
+      "learning_rate": 1.1726863452757642e-06,
+      "loss": 0.2432,
+      "step": 5117
+    },
+    {
+      "epoch": 2.868834080717489,
+      "grad_norm": 0.07987496597951721,
+      "learning_rate": 1.1627466156798328e-06,
+      "loss": 0.2366,
+      "step": 5118
+    },
+    {
+      "epoch": 2.8693946188340806,
+      "grad_norm": 0.08033334021918984,
+      "learning_rate": 1.1528489439701085e-06,
+      "loss": 0.2439,
+      "step": 5119
+    },
+    {
+      "epoch": 2.8699551569506725,
+      "grad_norm": 0.08015640437497674,
+      "learning_rate": 1.142993334358311e-06,
+      "loss": 0.2416,
+      "step": 5120
+    },
+    {
+      "epoch": 2.8705156950672643,
+      "grad_norm": 0.08066441239988657,
+      "learning_rate": 1.1331797910382747e-06,
+      "loss": 0.2408,
+      "step": 5121
+    },
+    {
+      "epoch": 2.8710762331838566,
+      "grad_norm": 0.08384191065283618,
+      "learning_rate": 1.1234083181859256e-06,
+      "loss": 0.2384,
+      "step": 5122
+    },
+    {
+      "epoch": 2.8716367713004485,
+      "grad_norm": 0.08123361697642735,
+      "learning_rate": 1.1136789199592713e-06,
+      "loss": 0.2462,
+      "step": 5123
+    },
+    {
+      "epoch": 2.8721973094170403,
+      "grad_norm": 0.0794233829323199,
+      "learning_rate": 1.1039916004984441e-06,
+      "loss": 0.2418,
+      "step": 5124
+    },
+    {
+      "epoch": 2.872757847533632,
+      "grad_norm": 0.07897174463108514,
+      "learning_rate": 1.094346363925647e-06,
+      "loss": 0.2377,
+      "step": 5125
+    },
+    {
+      "epoch": 2.873318385650224,
+      "grad_norm": 0.08348418792163165,
+      "learning_rate": 1.0847432143451962e-06,
+      "loss": 0.2421,
+      "step": 5126
+    },
+    {
+      "epoch": 2.8738789237668163,
+      "grad_norm": 0.08161778924749036,
+      "learning_rate": 1.0751821558434793e-06,
+      "loss": 0.2459,
+      "step": 5127
+    },
+    {
+      "epoch": 2.874439461883408,
+      "grad_norm": 0.08116353598027902,
+      "learning_rate": 1.0656631924889749e-06,
+      "loss": 0.2478,
+      "step": 5128
+    },
+    {
+      "epoch": 2.875,
+      "grad_norm": 0.08347565726708857,
+      "learning_rate": 1.0561863283322759e-06,
+      "loss": 0.2393,
+      "step": 5129
+    },
+    {
+      "epoch": 2.875560538116592,
+      "grad_norm": 0.08033654412257757,
+      "learning_rate": 1.0467515674060236e-06,
+      "loss": 0.2417,
+      "step": 5130
+    },
+    {
+      "epoch": 2.8761210762331837,
+      "grad_norm": 0.07684616191101404,
+      "learning_rate": 1.037358913724973e-06,
+      "loss": 0.2362,
+      "step": 5131
+    },
+    {
+      "epoch": 2.876681614349776,
+      "grad_norm": 0.08189726163249501,
+      "learning_rate": 1.028008371285938e-06,
+      "loss": 0.2407,
+      "step": 5132
+    },
+    {
+      "epoch": 2.877242152466368,
+      "grad_norm": 0.08280875568688238,
+      "learning_rate": 1.0186999440678246e-06,
+      "loss": 0.2496,
+      "step": 5133
+    },
+    {
+      "epoch": 2.8778026905829597,
+      "grad_norm": 0.07971865121690089,
+      "learning_rate": 1.0094336360316202e-06,
+      "loss": 0.2363,
+      "step": 5134
+    },
+    {
+      "epoch": 2.8783632286995515,
+      "grad_norm": 0.08195841973986688,
+      "learning_rate": 1.0002094511203819e-06,
+      "loss": 0.2407,
+      "step": 5135
+    },
+    {
+      "epoch": 2.8789237668161434,
+      "grad_norm": 0.08239469761028677,
+      "learning_rate": 9.910273932592584e-07,
+      "loss": 0.2561,
+      "step": 5136
+    },
+    {
+      "epoch": 2.8794843049327357,
+      "grad_norm": 0.08177280278997268,
+      "learning_rate": 9.818874663554357e-07,
+      "loss": 0.2459,
+      "step": 5137
+    },
+    {
+      "epoch": 2.8800448430493275,
+      "grad_norm": 0.07983579704437194,
+      "learning_rate": 9.727896742982245e-07,
+      "loss": 0.2486,
+      "step": 5138
+    },
+    {
+      "epoch": 2.8806053811659194,
+      "grad_norm": 0.08053571258340099,
+      "learning_rate": 9.63734020958973e-07,
+      "loss": 0.2365,
+      "step": 5139
+    },
+    {
+      "epoch": 2.881165919282511,
+      "grad_norm": 0.07728186842334782,
+      "learning_rate": 9.54720510191076e-07,
+      "loss": 0.2355,
+      "step": 5140
+    },
+    {
+      "epoch": 2.881726457399103,
+      "grad_norm": 0.08057560656759212,
+      "learning_rate": 9.457491458300549e-07,
+      "loss": 0.2421,
+      "step": 5141
+    },
+    {
+      "epoch": 2.8822869955156953,
+      "grad_norm": 0.08003610843585328,
+      "learning_rate": 9.368199316934445e-07,
+      "loss": 0.2478,
+      "step": 5142
+    },
+    {
+      "epoch": 2.8828475336322867,
+      "grad_norm": 0.08280976479529986,
+      "learning_rate": 9.279328715808722e-07,
+      "loss": 0.2435,
+      "step": 5143
+    },
+    {
+      "epoch": 2.883408071748879,
+      "grad_norm": 0.07809939277009989,
+      "learning_rate": 9.190879692740128e-07,
+      "loss": 0.2337,
+      "step": 5144
+    },
+    {
+      "epoch": 2.883968609865471,
+      "grad_norm": 0.07922551065074915,
+      "learning_rate": 9.102852285366226e-07,
+      "loss": 0.2446,
+      "step": 5145
+    },
+    {
+      "epoch": 2.8845291479820627,
+      "grad_norm": 0.07825885666601355,
+      "learning_rate": 9.015246531144939e-07,
+      "loss": 0.2306,
+      "step": 5146
+    },
+    {
+      "epoch": 2.8850896860986546,
+      "grad_norm": 0.07878833678038266,
+      "learning_rate": 8.92806246735467e-07,
+      "loss": 0.2341,
+      "step": 5147
+    },
+    {
+      "epoch": 2.8856502242152464,
+      "grad_norm": 0.08181587406985132,
+      "learning_rate": 8.841300131094854e-07,
+      "loss": 0.2457,
+      "step": 5148
+    },
+    {
+      "epoch": 2.8862107623318387,
+      "grad_norm": 0.0805969143015379,
+      "learning_rate": 8.75495955928507e-07,
+      "loss": 0.2351,
+      "step": 5149
+    },
+    {
+      "epoch": 2.8867713004484306,
+      "grad_norm": 0.07843158968423362,
+      "learning_rate": 8.669040788665372e-07,
+      "loss": 0.2324,
+      "step": 5150
+    },
+    {
+      "epoch": 2.8873318385650224,
+      "grad_norm": 0.07739285893068815,
+      "learning_rate": 8.583543855796738e-07,
+      "loss": 0.2319,
+      "step": 5151
+    },
+    {
+      "epoch": 2.8878923766816142,
+      "grad_norm": 0.08471175286687038,
+      "learning_rate": 8.498468797060289e-07,
+      "loss": 0.2366,
+      "step": 5152
+    },
+    {
+      "epoch": 2.888452914798206,
+      "grad_norm": 0.08302254160230785,
+      "learning_rate": 8.413815648657731e-07,
+      "loss": 0.2525,
+      "step": 5153
+    },
+    {
+      "epoch": 2.8890134529147984,
+      "grad_norm": 0.07967514164685403,
+      "learning_rate": 8.329584446611138e-07,
+      "loss": 0.2447,
+      "step": 5154
+    },
+    {
+      "epoch": 2.8895739910313902,
+      "grad_norm": 0.08256158297069675,
+      "learning_rate": 8.245775226763397e-07,
+      "loss": 0.2476,
+      "step": 5155
+    },
+    {
+      "epoch": 2.890134529147982,
+      "grad_norm": 0.08179878115799412,
+      "learning_rate": 8.162388024777201e-07,
+      "loss": 0.2411,
+      "step": 5156
+    },
+    {
+      "epoch": 2.890695067264574,
+      "grad_norm": 0.08052932090181747,
+      "learning_rate": 8.079422876136388e-07,
+      "loss": 0.237,
+      "step": 5157
+    },
+    {
+      "epoch": 2.8912556053811658,
+      "grad_norm": 0.08060278046000689,
+      "learning_rate": 7.996879816144498e-07,
+      "loss": 0.2423,
+      "step": 5158
+    },
+    {
+      "epoch": 2.891816143497758,
+      "grad_norm": 0.08077338111923703,
+      "learning_rate": 7.914758879925988e-07,
+      "loss": 0.2445,
+      "step": 5159
+    },
+    {
+      "epoch": 2.89237668161435,
+      "grad_norm": 0.07978404180548836,
+      "learning_rate": 7.833060102425682e-07,
+      "loss": 0.2481,
+      "step": 5160
+    },
+    {
+      "epoch": 2.8929372197309418,
+      "grad_norm": 0.08072306353866018,
+      "learning_rate": 7.751783518408218e-07,
+      "loss": 0.2392,
+      "step": 5161
+    },
+    {
+      "epoch": 2.8934977578475336,
+      "grad_norm": 0.07596315491150957,
+      "learning_rate": 7.670929162459261e-07,
+      "loss": 0.2229,
+      "step": 5162
+    },
+    {
+      "epoch": 2.8940582959641254,
+      "grad_norm": 0.08204698511684655,
+      "learning_rate": 7.590497068984293e-07,
+      "loss": 0.2464,
+      "step": 5163
+    },
+    {
+      "epoch": 2.8946188340807177,
+      "grad_norm": 0.08109058564013136,
+      "learning_rate": 7.510487272209377e-07,
+      "loss": 0.2437,
+      "step": 5164
+    },
+    {
+      "epoch": 2.895179372197309,
+      "grad_norm": 0.08325131574882993,
+      "learning_rate": 7.430899806180835e-07,
+      "loss": 0.245,
+      "step": 5165
+    },
+    {
+      "epoch": 2.8957399103139014,
+      "grad_norm": 0.08101001328605403,
+      "learning_rate": 7.351734704765245e-07,
+      "loss": 0.2377,
+      "step": 5166
+    },
+    {
+      "epoch": 2.8963004484304933,
+      "grad_norm": 0.08026757155181756,
+      "learning_rate": 7.272992001649436e-07,
+      "loss": 0.2284,
+      "step": 5167
+    },
+    {
+      "epoch": 2.896860986547085,
+      "grad_norm": 0.07914441858849464,
+      "learning_rate": 7.194671730340608e-07,
+      "loss": 0.2449,
+      "step": 5168
+    },
+    {
+      "epoch": 2.897421524663677,
+      "grad_norm": 0.08445317725781687,
+      "learning_rate": 7.116773924166098e-07,
+      "loss": 0.2426,
+      "step": 5169
+    },
+    {
+      "epoch": 2.897982062780269,
+      "grad_norm": 0.07822137255581857,
+      "learning_rate": 7.039298616273393e-07,
+      "loss": 0.2333,
+      "step": 5170
+    },
+    {
+      "epoch": 2.898542600896861,
+      "grad_norm": 0.07983348122960583,
+      "learning_rate": 6.962245839630455e-07,
+      "loss": 0.2417,
+      "step": 5171
+    },
+    {
+      "epoch": 2.899103139013453,
+      "grad_norm": 0.08133611004648454,
+      "learning_rate": 6.885615627025166e-07,
+      "loss": 0.2406,
+      "step": 5172
+    },
+    {
+      "epoch": 2.899663677130045,
+      "grad_norm": 0.07786241853242937,
+      "learning_rate": 6.809408011065887e-07,
+      "loss": 0.2348,
+      "step": 5173
+    },
+    {
+      "epoch": 2.9002242152466366,
+      "grad_norm": 0.08319128466847514,
+      "learning_rate": 6.733623024180791e-07,
+      "loss": 0.251,
+      "step": 5174
+    },
+    {
+      "epoch": 2.9007847533632285,
+      "grad_norm": 0.0806610722398245,
+      "learning_rate": 6.658260698618524e-07,
+      "loss": 0.2428,
+      "step": 5175
+    },
+    {
+      "epoch": 2.901345291479821,
+      "grad_norm": 0.07884296766934859,
+      "learning_rate": 6.583321066447656e-07,
+      "loss": 0.2354,
+      "step": 5176
+    },
+    {
+      "epoch": 2.9019058295964126,
+      "grad_norm": 0.08325168236695274,
+      "learning_rate": 6.508804159557236e-07,
+      "loss": 0.2576,
+      "step": 5177
+    },
+    {
+      "epoch": 2.9024663677130045,
+      "grad_norm": 0.08127027999981284,
+      "learning_rate": 6.434710009656008e-07,
+      "loss": 0.2307,
+      "step": 5178
+    },
+    {
+      "epoch": 2.9030269058295963,
+      "grad_norm": 0.07950439689511878,
+      "learning_rate": 6.361038648273088e-07,
+      "loss": 0.238,
+      "step": 5179
+    },
+    {
+      "epoch": 2.903587443946188,
+      "grad_norm": 0.07941759994163064,
+      "learning_rate": 6.287790106757396e-07,
+      "loss": 0.2459,
+      "step": 5180
+    },
+    {
+      "epoch": 2.9041479820627805,
+      "grad_norm": 0.08089931654603481,
+      "learning_rate": 6.214964416278445e-07,
+      "loss": 0.2446,
+      "step": 5181
+    },
+    {
+      "epoch": 2.9047085201793723,
+      "grad_norm": 0.08088896415189534,
+      "learning_rate": 6.142561607825337e-07,
+      "loss": 0.2353,
+      "step": 5182
+    },
+    {
+      "epoch": 2.905269058295964,
+      "grad_norm": 0.08117874836660299,
+      "learning_rate": 6.070581712207424e-07,
+      "loss": 0.2469,
+      "step": 5183
+    },
+    {
+      "epoch": 2.905829596412556,
+      "grad_norm": 0.08065707726695893,
+      "learning_rate": 5.999024760054095e-07,
+      "loss": 0.2387,
+      "step": 5184
+    },
+    {
+      "epoch": 2.906390134529148,
+      "grad_norm": 0.08049225142767744,
+      "learning_rate": 5.927890781814661e-07,
+      "loss": 0.2415,
+      "step": 5185
+    },
+    {
+      "epoch": 2.90695067264574,
+      "grad_norm": 0.08113091376622063,
+      "learning_rate": 5.857179807758684e-07,
+      "loss": 0.2491,
+      "step": 5186
+    },
+    {
+      "epoch": 2.907511210762332,
+      "grad_norm": 0.07695739944725659,
+      "learning_rate": 5.78689186797543e-07,
+      "loss": 0.2306,
+      "step": 5187
+    },
+    {
+      "epoch": 2.908071748878924,
+      "grad_norm": 0.08066498499230515,
+      "learning_rate": 5.717026992374308e-07,
+      "loss": 0.2472,
+      "step": 5188
+    },
+    {
+      "epoch": 2.9086322869955157,
+      "grad_norm": 0.08083582416430926,
+      "learning_rate": 5.647585210684758e-07,
+      "loss": 0.2547,
+      "step": 5189
+    },
+    {
+      "epoch": 2.9091928251121075,
+      "grad_norm": 0.0813592996568508,
+      "learning_rate": 5.578566552456032e-07,
+      "loss": 0.2411,
+      "step": 5190
+    },
+    {
+      "epoch": 2.9097533632287,
+      "grad_norm": 0.07931356603718039,
+      "learning_rate": 5.509971047057416e-07,
+      "loss": 0.2411,
+      "step": 5191
+    },
+    {
+      "epoch": 2.910313901345291,
+      "grad_norm": 0.08105103356021948,
+      "learning_rate": 5.441798723678115e-07,
+      "loss": 0.2415,
+      "step": 5192
+    },
+    {
+      "epoch": 2.9108744394618835,
+      "grad_norm": 0.07855446840127679,
+      "learning_rate": 5.37404961132737e-07,
+      "loss": 0.2408,
+      "step": 5193
+    },
+    {
+      "epoch": 2.9114349775784754,
+      "grad_norm": 0.07995087383105443,
+      "learning_rate": 5.306723738834119e-07,
+      "loss": 0.2389,
+      "step": 5194
+    },
+    {
+      "epoch": 2.911995515695067,
+      "grad_norm": 0.08029067563327771,
+      "learning_rate": 5.239821134847445e-07,
+      "loss": 0.2341,
+      "step": 5195
+    },
+    {
+      "epoch": 2.912556053811659,
+      "grad_norm": 0.07820034799966565,
+      "learning_rate": 5.173341827836021e-07,
+      "loss": 0.2436,
+      "step": 5196
+    },
+    {
+      "epoch": 2.913116591928251,
+      "grad_norm": 0.07774134929015715,
+      "learning_rate": 5.107285846088772e-07,
+      "loss": 0.2465,
+      "step": 5197
+    },
+    {
+      "epoch": 2.913677130044843,
+      "grad_norm": 0.081846016749724,
+      "learning_rate": 5.041653217713993e-07,
+      "loss": 0.2551,
+      "step": 5198
+    },
+    {
+      "epoch": 2.914237668161435,
+      "grad_norm": 0.08222288904358514,
+      "learning_rate": 4.976443970640343e-07,
+      "loss": 0.2476,
+      "step": 5199
+    },
+    {
+      "epoch": 2.914798206278027,
+      "grad_norm": 0.07929364672043442,
+      "learning_rate": 4.91165813261607e-07,
+      "loss": 0.2474,
+      "step": 5200
+    },
+    {
+      "epoch": 2.9153587443946187,
+      "grad_norm": 0.0796620167903564,
+      "learning_rate": 4.847295731209234e-07,
+      "loss": 0.2489,
+      "step": 5201
+    },
+    {
+      "epoch": 2.9159192825112106,
+      "grad_norm": 0.08027609274529898,
+      "learning_rate": 4.783356793807814e-07,
+      "loss": 0.2394,
+      "step": 5202
+    },
+    {
+      "epoch": 2.916479820627803,
+      "grad_norm": 0.07993725490400207,
+      "learning_rate": 4.7198413476193804e-07,
+      "loss": 0.2492,
+      "step": 5203
+    },
+    {
+      "epoch": 2.9170403587443947,
+      "grad_norm": 0.0803466866394501,
+      "learning_rate": 4.6567494196715354e-07,
+      "loss": 0.2511,
+      "step": 5204
+    },
+    {
+      "epoch": 2.9176008968609866,
+      "grad_norm": 0.0800223749066712,
+      "learning_rate": 4.5940810368116924e-07,
+      "loss": 0.2418,
+      "step": 5205
+    },
+    {
+      "epoch": 2.9181614349775784,
+      "grad_norm": 0.08007370509835651,
+      "learning_rate": 4.5318362257067426e-07,
+      "loss": 0.2277,
+      "step": 5206
+    },
+    {
+      "epoch": 2.9187219730941703,
+      "grad_norm": 0.08043926279562774,
+      "learning_rate": 4.4700150128436094e-07,
+      "loss": 0.2492,
+      "step": 5207
+    },
+    {
+      "epoch": 2.9192825112107625,
+      "grad_norm": 0.08165038215580994,
+      "learning_rate": 4.4086174245288047e-07,
+      "loss": 0.2419,
+      "step": 5208
+    },
+    {
+      "epoch": 2.9198430493273544,
+      "grad_norm": 0.08080196846154497,
+      "learning_rate": 4.347643486888653e-07,
+      "loss": 0.2433,
+      "step": 5209
+    },
+    {
+      "epoch": 2.9204035874439462,
+      "grad_norm": 0.07920220502048722,
+      "learning_rate": 4.287093225869288e-07,
+      "loss": 0.2443,
+      "step": 5210
+    },
+    {
+      "epoch": 2.920964125560538,
+      "grad_norm": 0.08143488760783185,
+      "learning_rate": 4.226966667236321e-07,
+      "loss": 0.2463,
+      "step": 5211
+    },
+    {
+      "epoch": 2.92152466367713,
+      "grad_norm": 0.08247489520940915,
+      "learning_rate": 4.167263836575286e-07,
+      "loss": 0.247,
+      "step": 5212
+    },
+    {
+      "epoch": 2.922085201793722,
+      "grad_norm": 0.0807093966938727,
+      "learning_rate": 4.107984759291306e-07,
+      "loss": 0.2392,
+      "step": 5213
+    },
+    {
+      "epoch": 2.922645739910314,
+      "grad_norm": 0.0795617388431577,
+      "learning_rate": 4.0491294606093135e-07,
+      "loss": 0.2384,
+      "step": 5214
+    },
+    {
+      "epoch": 2.923206278026906,
+      "grad_norm": 0.07962850511281837,
+      "learning_rate": 3.990697965573609e-07,
+      "loss": 0.2374,
+      "step": 5215
+    },
+    {
+      "epoch": 2.9237668161434978,
+      "grad_norm": 0.07993802378102899,
+      "learning_rate": 3.9326902990484136e-07,
+      "loss": 0.2432,
+      "step": 5216
+    },
+    {
+      "epoch": 2.9243273542600896,
+      "grad_norm": 0.08015230120043666,
+      "learning_rate": 3.87510648571765e-07,
+      "loss": 0.2484,
+      "step": 5217
+    },
+    {
+      "epoch": 2.9248878923766815,
+      "grad_norm": 0.07999735915052116,
+      "learning_rate": 3.8179465500846057e-07,
+      "loss": 0.2297,
+      "step": 5218
+    },
+    {
+      "epoch": 2.9254484304932733,
+      "grad_norm": 0.07941900333132416,
+      "learning_rate": 3.761210516472602e-07,
+      "loss": 0.2383,
+      "step": 5219
+    },
+    {
+      "epoch": 2.9260089686098656,
+      "grad_norm": 0.08122180837584679,
+      "learning_rate": 3.7048984090239934e-07,
+      "loss": 0.2384,
+      "step": 5220
+    },
+    {
+      "epoch": 2.9265695067264574,
+      "grad_norm": 0.07721862593318037,
+      "learning_rate": 3.64901025170139e-07,
+      "loss": 0.2316,
+      "step": 5221
+    },
+    {
+      "epoch": 2.9271300448430493,
+      "grad_norm": 0.07826050021759158,
+      "learning_rate": 3.593546068286435e-07,
+      "loss": 0.2359,
+      "step": 5222
+    },
+    {
+      "epoch": 2.927690582959641,
+      "grad_norm": 0.08215849382135554,
+      "learning_rate": 3.5385058823809156e-07,
+      "loss": 0.241,
+      "step": 5223
+    },
+    {
+      "epoch": 2.928251121076233,
+      "grad_norm": 0.07902599499630579,
+      "learning_rate": 3.4838897174055417e-07,
+      "loss": 0.2371,
+      "step": 5224
+    },
+    {
+      "epoch": 2.9288116591928253,
+      "grad_norm": 0.08411121884071082,
+      "learning_rate": 3.429697596601278e-07,
+      "loss": 0.2457,
+      "step": 5225
+    },
+    {
+      "epoch": 2.929372197309417,
+      "grad_norm": 0.07626874289616034,
+      "learning_rate": 3.3759295430281223e-07,
+      "loss": 0.2375,
+      "step": 5226
+    },
+    {
+      "epoch": 2.929932735426009,
+      "grad_norm": 0.07894069331815652,
+      "learning_rate": 3.3225855795658845e-07,
+      "loss": 0.2371,
+      "step": 5227
+    },
+    {
+      "epoch": 2.930493273542601,
+      "grad_norm": 0.0828622145940112,
+      "learning_rate": 3.26966572891374e-07,
+      "loss": 0.2532,
+      "step": 5228
+    },
+    {
+      "epoch": 2.9310538116591927,
+      "grad_norm": 0.08139724099673272,
+      "learning_rate": 3.2171700135906756e-07,
+      "loss": 0.2362,
+      "step": 5229
+    },
+    {
+      "epoch": 2.931614349775785,
+      "grad_norm": 0.07847946908193802,
+      "learning_rate": 3.1650984559349337e-07,
+      "loss": 0.2404,
+      "step": 5230
+    },
+    {
+      "epoch": 2.932174887892377,
+      "grad_norm": 0.08144523311452388,
+      "learning_rate": 3.1134510781042347e-07,
+      "loss": 0.2577,
+      "step": 5231
+    },
+    {
+      "epoch": 2.9327354260089686,
+      "grad_norm": 0.08331031210937782,
+      "learning_rate": 3.062227902076109e-07,
+      "loss": 0.2371,
+      "step": 5232
+    },
+    {
+      "epoch": 2.9332959641255605,
+      "grad_norm": 0.08001492880179598,
+      "learning_rate": 3.011428949647233e-07,
+      "loss": 0.2433,
+      "step": 5233
+    },
+    {
+      "epoch": 2.9338565022421523,
+      "grad_norm": 0.0807533308463706,
+      "learning_rate": 2.9610542424339803e-07,
+      "loss": 0.2407,
+      "step": 5234
+    },
+    {
+      "epoch": 2.9344170403587446,
+      "grad_norm": 0.08240554474517298,
+      "learning_rate": 2.911103801872206e-07,
+      "loss": 0.2454,
+      "step": 5235
+    },
+    {
+      "epoch": 2.9349775784753365,
+      "grad_norm": 0.07734753742670465,
+      "learning_rate": 2.8615776492170176e-07,
+      "loss": 0.2379,
+      "step": 5236
+    },
+    {
+      "epoch": 2.9355381165919283,
+      "grad_norm": 0.078549890262253,
+      "learning_rate": 2.812475805543224e-07,
+      "loss": 0.2395,
+      "step": 5237
+    },
+    {
+      "epoch": 2.93609865470852,
+      "grad_norm": 0.07995050894732823,
+      "learning_rate": 2.763798291744779e-07,
+      "loss": 0.2401,
+      "step": 5238
+    },
+    {
+      "epoch": 2.936659192825112,
+      "grad_norm": 0.07807092079681063,
+      "learning_rate": 2.715545128535557e-07,
+      "loss": 0.2356,
+      "step": 5239
+    },
+    {
+      "epoch": 2.9372197309417043,
+      "grad_norm": 0.07855991185553275,
+      "learning_rate": 2.667716336448356e-07,
+      "loss": 0.245,
+      "step": 5240
+    },
+    {
+      "epoch": 2.9377802690582957,
+      "grad_norm": 0.07944252318522475,
+      "learning_rate": 2.6203119358356733e-07,
+      "loss": 0.2428,
+      "step": 5241
+    },
+    {
+      "epoch": 2.938340807174888,
+      "grad_norm": 0.08289075442761087,
+      "learning_rate": 2.573331946869262e-07,
+      "loss": 0.2293,
+      "step": 5242
+    },
+    {
+      "epoch": 2.93890134529148,
+      "grad_norm": 0.0815455144238909,
+      "learning_rate": 2.526776389540353e-07,
+      "loss": 0.2254,
+      "step": 5243
+    },
+    {
+      "epoch": 2.9394618834080717,
+      "grad_norm": 0.0803571095603025,
+      "learning_rate": 2.480645283659766e-07,
+      "loss": 0.2322,
+      "step": 5244
+    },
+    {
+      "epoch": 2.9400224215246635,
+      "grad_norm": 0.08474456628701002,
+      "learning_rate": 2.4349386488574654e-07,
+      "loss": 0.2532,
+      "step": 5245
+    },
+    {
+      "epoch": 2.9405829596412554,
+      "grad_norm": 0.0805078766355413,
+      "learning_rate": 2.3896565045826714e-07,
+      "loss": 0.2369,
+      "step": 5246
+    },
+    {
+      "epoch": 2.9411434977578477,
+      "grad_norm": 0.08180737048391117,
+      "learning_rate": 2.3447988701043034e-07,
+      "loss": 0.2467,
+      "step": 5247
+    },
+    {
+      "epoch": 2.9417040358744395,
+      "grad_norm": 0.07958616382853968,
+      "learning_rate": 2.300365764510315e-07,
+      "loss": 0.236,
+      "step": 5248
+    },
+    {
+      "epoch": 2.9422645739910314,
+      "grad_norm": 0.08049069112082367,
+      "learning_rate": 2.2563572067083595e-07,
+      "loss": 0.2414,
+      "step": 5249
+    },
+    {
+      "epoch": 2.942825112107623,
+      "grad_norm": 0.0781498909935061,
+      "learning_rate": 2.2127732154251235e-07,
+      "loss": 0.24,
+      "step": 5250
+    },
+    {
+      "epoch": 2.943385650224215,
+      "grad_norm": 0.07528173107924589,
+      "learning_rate": 2.169613809206883e-07,
+      "loss": 0.2294,
+      "step": 5251
+    },
+    {
+      "epoch": 2.9439461883408073,
+      "grad_norm": 0.07999346420773303,
+      "learning_rate": 2.126879006419058e-07,
+      "loss": 0.2254,
+      "step": 5252
+    },
+    {
+      "epoch": 2.944506726457399,
+      "grad_norm": 0.08168423635911633,
+      "learning_rate": 2.0845688252464357e-07,
+      "loss": 0.25,
+      "step": 5253
+    },
+    {
+      "epoch": 2.945067264573991,
+      "grad_norm": 0.08056444463902272,
+      "learning_rate": 2.0426832836930587e-07,
+      "loss": 0.2392,
+      "step": 5254
+    },
+    {
+      "epoch": 2.945627802690583,
+      "grad_norm": 0.07982966048495936,
+      "learning_rate": 2.0012223995824474e-07,
+      "loss": 0.2382,
+      "step": 5255
+    },
+    {
+      "epoch": 2.9461883408071747,
+      "grad_norm": 0.08063967191486175,
+      "learning_rate": 1.9601861905572672e-07,
+      "loss": 0.2341,
+      "step": 5256
+    },
+    {
+      "epoch": 2.946748878923767,
+      "grad_norm": 0.07939613336880443,
+      "learning_rate": 1.9195746740795495e-07,
+      "loss": 0.2485,
+      "step": 5257
+    },
+    {
+      "epoch": 2.947309417040359,
+      "grad_norm": 0.08065516975169129,
+      "learning_rate": 1.879387867430471e-07,
+      "loss": 0.2509,
+      "step": 5258
+    },
+    {
+      "epoch": 2.9478699551569507,
+      "grad_norm": 0.08133260163247694,
+      "learning_rate": 1.839625787710686e-07,
+      "loss": 0.2409,
+      "step": 5259
+    },
+    {
+      "epoch": 2.9484304932735426,
+      "grad_norm": 0.07895193321754743,
+      "learning_rate": 1.8002884518401041e-07,
+      "loss": 0.2223,
+      "step": 5260
+    },
+    {
+      "epoch": 2.9489910313901344,
+      "grad_norm": 0.07988661288054187,
+      "learning_rate": 1.7613758765576692e-07,
+      "loss": 0.2478,
+      "step": 5261
+    },
+    {
+      "epoch": 2.9495515695067267,
+      "grad_norm": 0.07916704842128662,
+      "learning_rate": 1.7228880784216915e-07,
+      "loss": 0.2297,
+      "step": 5262
+    },
+    {
+      "epoch": 2.9501121076233185,
+      "grad_norm": 0.08033690224000986,
+      "learning_rate": 1.684825073809848e-07,
+      "loss": 0.2371,
+      "step": 5263
+    },
+    {
+      "epoch": 2.9506726457399104,
+      "grad_norm": 0.08024692582305298,
+      "learning_rate": 1.6471868789189603e-07,
+      "loss": 0.2296,
+      "step": 5264
+    },
+    {
+      "epoch": 2.9512331838565022,
+      "grad_norm": 0.07631516485832594,
+      "learning_rate": 1.6099735097651058e-07,
+      "loss": 0.2249,
+      "step": 5265
+    },
+    {
+      "epoch": 2.951793721973094,
+      "grad_norm": 0.08230465193965804,
+      "learning_rate": 1.5731849821833954e-07,
+      "loss": 0.2406,
+      "step": 5266
+    },
+    {
+      "epoch": 2.952354260089686,
+      "grad_norm": 0.08151279907385826,
+      "learning_rate": 1.536821311828529e-07,
+      "loss": 0.2415,
+      "step": 5267
+    },
+    {
+      "epoch": 2.952914798206278,
+      "grad_norm": 0.08051643231472558,
+      "learning_rate": 1.50088251417424e-07,
+      "loss": 0.2321,
+      "step": 5268
+    },
+    {
+      "epoch": 2.95347533632287,
+      "grad_norm": 0.0813535094180172,
+      "learning_rate": 1.4653686045131843e-07,
+      "loss": 0.2327,
+      "step": 5269
+    },
+    {
+      "epoch": 2.954035874439462,
+      "grad_norm": 0.08177174334741082,
+      "learning_rate": 1.4302795979577177e-07,
+      "loss": 0.2573,
+      "step": 5270
+    },
+    {
+      "epoch": 2.9545964125560538,
+      "grad_norm": 0.08309563398890388,
+      "learning_rate": 1.395615509439119e-07,
+      "loss": 0.2405,
+      "step": 5271
+    },
+    {
+      "epoch": 2.9551569506726456,
+      "grad_norm": 0.07977820548089735,
+      "learning_rate": 1.3613763537078105e-07,
+      "loss": 0.2393,
+      "step": 5272
+    },
+    {
+      "epoch": 2.9557174887892375,
+      "grad_norm": 0.07727833251814736,
+      "learning_rate": 1.3275621453333608e-07,
+      "loss": 0.2396,
+      "step": 5273
+    },
+    {
+      "epoch": 2.9562780269058297,
+      "grad_norm": 0.07994854813799306,
+      "learning_rate": 1.294172898704815e-07,
+      "loss": 0.2255,
+      "step": 5274
+    },
+    {
+      "epoch": 2.9568385650224216,
+      "grad_norm": 0.08086972983216732,
+      "learning_rate": 1.2612086280302527e-07,
+      "loss": 0.2412,
+      "step": 5275
+    },
+    {
+      "epoch": 2.9573991031390134,
+      "grad_norm": 0.07839291898304604,
+      "learning_rate": 1.228669347336564e-07,
+      "loss": 0.2306,
+      "step": 5276
+    },
+    {
+      "epoch": 2.9579596412556053,
+      "grad_norm": 0.08084565279047115,
+      "learning_rate": 1.1965550704702288e-07,
+      "loss": 0.2517,
+      "step": 5277
+    },
+    {
+      "epoch": 2.958520179372197,
+      "grad_norm": 0.08194291771179481,
+      "learning_rate": 1.1648658110967603e-07,
+      "loss": 0.2494,
+      "step": 5278
+    },
+    {
+      "epoch": 2.9590807174887894,
+      "grad_norm": 0.08006505008080902,
+      "learning_rate": 1.1336015827008161e-07,
+      "loss": 0.243,
+      "step": 5279
+    },
+    {
+      "epoch": 2.9596412556053813,
+      "grad_norm": 0.07754436239138039,
+      "learning_rate": 1.102762398586088e-07,
+      "loss": 0.2288,
+      "step": 5280
+    },
+    {
+      "epoch": 2.960201793721973,
+      "grad_norm": 0.07944347964397883,
+      "learning_rate": 1.0723482718754118e-07,
+      "loss": 0.2366,
+      "step": 5281
+    },
+    {
+      "epoch": 2.960762331838565,
+      "grad_norm": 0.07924613103277234,
+      "learning_rate": 1.0423592155108797e-07,
+      "loss": 0.2485,
+      "step": 5282
+    },
+    {
+      "epoch": 2.961322869955157,
+      "grad_norm": 0.0775394939832644,
+      "learning_rate": 1.0127952422536169e-07,
+      "loss": 0.2306,
+      "step": 5283
+    },
+    {
+      "epoch": 2.961883408071749,
+      "grad_norm": 0.08111721430949032,
+      "learning_rate": 9.836563646840047e-08,
+      "loss": 0.2451,
+      "step": 5284
+    },
+    {
+      "epoch": 2.962443946188341,
+      "grad_norm": 0.07987107450762725,
+      "learning_rate": 9.549425952012358e-08,
+      "loss": 0.2284,
+      "step": 5285
+    },
+    {
+      "epoch": 2.963004484304933,
+      "grad_norm": 0.08141807380302148,
+      "learning_rate": 9.266539460238699e-08,
+      "loss": 0.2419,
+      "step": 5286
+    },
+    {
+      "epoch": 2.9635650224215246,
+      "grad_norm": 0.07786866507150199,
+      "learning_rate": 8.98790429189389e-08,
+      "loss": 0.2376,
+      "step": 5287
+    },
+    {
+      "epoch": 2.9641255605381165,
+      "grad_norm": 0.0832605088128385,
+      "learning_rate": 8.713520565546419e-08,
+      "loss": 0.237,
+      "step": 5288
+    },
+    {
+      "epoch": 2.964686098654709,
+      "grad_norm": 0.08094358918281826,
+      "learning_rate": 8.44338839795289e-08,
+      "loss": 0.241,
+      "step": 5289
+    },
+    {
+      "epoch": 2.9652466367713,
+      "grad_norm": 0.08049074033771993,
+      "learning_rate": 8.177507904060244e-08,
+      "loss": 0.2351,
+      "step": 5290
+    },
+    {
+      "epoch": 2.9658071748878925,
+      "grad_norm": 0.08298281914638202,
+      "learning_rate": 7.915879197010201e-08,
+      "loss": 0.2554,
+      "step": 5291
+    },
+    {
+      "epoch": 2.9663677130044843,
+      "grad_norm": 0.0773610087962181,
+      "learning_rate": 7.658502388131483e-08,
+      "loss": 0.2312,
+      "step": 5292
+    },
+    {
+      "epoch": 2.966928251121076,
+      "grad_norm": 0.07854609379353328,
+      "learning_rate": 7.405377586945372e-08,
+      "loss": 0.2344,
+      "step": 5293
+    },
+    {
+      "epoch": 2.967488789237668,
+      "grad_norm": 0.07953023109582347,
+      "learning_rate": 7.156504901162375e-08,
+      "loss": 0.2404,
+      "step": 5294
+    },
+    {
+      "epoch": 2.96804932735426,
+      "grad_norm": 0.08025893705482982,
+      "learning_rate": 6.911884436685556e-08,
+      "loss": 0.2316,
+      "step": 5295
+    },
+    {
+      "epoch": 2.968609865470852,
+      "grad_norm": 0.07994746443161596,
+      "learning_rate": 6.671516297606095e-08,
+      "loss": 0.2371,
+      "step": 5296
+    },
+    {
+      "epoch": 2.969170403587444,
+      "grad_norm": 0.08187138973611886,
+      "learning_rate": 6.43540058620884e-08,
+      "loss": 0.2413,
+      "step": 5297
+    },
+    {
+      "epoch": 2.969730941704036,
+      "grad_norm": 0.08147503619062962,
+      "learning_rate": 6.203537402965643e-08,
+      "loss": 0.2468,
+      "step": 5298
+    },
+    {
+      "epoch": 2.9702914798206277,
+      "grad_norm": 0.08068868467183059,
+      "learning_rate": 5.975926846540914e-08,
+      "loss": 0.2349,
+      "step": 5299
+    },
+    {
+      "epoch": 2.9708520179372195,
+      "grad_norm": 0.07894573332874899,
+      "learning_rate": 5.752569013788289e-08,
+      "loss": 0.2315,
+      "step": 5300
+    },
+    {
+      "epoch": 2.971412556053812,
+      "grad_norm": 0.08129324141411996,
+      "learning_rate": 5.533463999755073e-08,
+      "loss": 0.2423,
+      "step": 5301
+    },
+    {
+      "epoch": 2.9719730941704037,
+      "grad_norm": 0.08115487487147469,
+      "learning_rate": 5.318611897673353e-08,
+      "loss": 0.2485,
+      "step": 5302
+    },
+    {
+      "epoch": 2.9725336322869955,
+      "grad_norm": 0.0806793046784492,
+      "learning_rate": 5.1080127989699966e-08,
+      "loss": 0.2495,
+      "step": 5303
+    },
+    {
+      "epoch": 2.9730941704035874,
+      "grad_norm": 0.07952902349286228,
+      "learning_rate": 4.901666793261095e-08,
+      "loss": 0.2433,
+      "step": 5304
+    },
+    {
+      "epoch": 2.973654708520179,
+      "grad_norm": 0.07782880919830099,
+      "learning_rate": 4.6995739683508564e-08,
+      "loss": 0.2478,
+      "step": 5305
+    },
+    {
+      "epoch": 2.9742152466367715,
+      "grad_norm": 0.07836102983454873,
+      "learning_rate": 4.501734410234937e-08,
+      "loss": 0.2317,
+      "step": 5306
+    },
+    {
+      "epoch": 2.9747757847533634,
+      "grad_norm": 0.08228369591277344,
+      "learning_rate": 4.3081482031015476e-08,
+      "loss": 0.241,
+      "step": 5307
+    },
+    {
+      "epoch": 2.975336322869955,
+      "grad_norm": 0.07932916963866564,
+      "learning_rate": 4.118815429324796e-08,
+      "loss": 0.2463,
+      "step": 5308
+    },
+    {
+      "epoch": 2.975896860986547,
+      "grad_norm": 0.08058193160876528,
+      "learning_rate": 3.933736169471347e-08,
+      "loss": 0.2398,
+      "step": 5309
+    },
+    {
+      "epoch": 2.976457399103139,
+      "grad_norm": 0.07912455799954479,
+      "learning_rate": 3.752910502297091e-08,
+      "loss": 0.2348,
+      "step": 5310
+    },
+    {
+      "epoch": 2.977017937219731,
+      "grad_norm": 0.07955512966216624,
+      "learning_rate": 3.576338504749366e-08,
+      "loss": 0.2499,
+      "step": 5311
+    },
+    {
+      "epoch": 2.977578475336323,
+      "grad_norm": 0.08242015311617172,
+      "learning_rate": 3.404020251963624e-08,
+      "loss": 0.2299,
+      "step": 5312
+    },
+    {
+      "epoch": 2.978139013452915,
+      "grad_norm": 0.07910613197284071,
+      "learning_rate": 3.235955817264546e-08,
+      "loss": 0.2408,
+      "step": 5313
+    },
+    {
+      "epoch": 2.9786995515695067,
+      "grad_norm": 0.07865882269700655,
+      "learning_rate": 3.0721452721704794e-08,
+      "loss": 0.2257,
+      "step": 5314
+    },
+    {
+      "epoch": 2.9792600896860986,
+      "grad_norm": 0.0815141709448338,
+      "learning_rate": 2.912588686384554e-08,
+      "loss": 0.2486,
+      "step": 5315
+    },
+    {
+      "epoch": 2.979820627802691,
+      "grad_norm": 0.075898357644101,
+      "learning_rate": 2.7572861278046814e-08,
+      "loss": 0.2239,
+      "step": 5316
+    },
+    {
+      "epoch": 2.9803811659192823,
+      "grad_norm": 0.08081346619433816,
+      "learning_rate": 2.6062376625146658e-08,
+      "loss": 0.2434,
+      "step": 5317
+    },
+    {
+      "epoch": 2.9809417040358746,
+      "grad_norm": 0.08136868711161611,
+      "learning_rate": 2.4594433547908692e-08,
+      "loss": 0.2495,
+      "step": 5318
+    },
+    {
+      "epoch": 2.9815022421524664,
+      "grad_norm": 0.08028070958974504,
+      "learning_rate": 2.3169032670966595e-08,
+      "loss": 0.2526,
+      "step": 5319
+    },
+    {
+      "epoch": 2.9820627802690582,
+      "grad_norm": 0.07911738094906717,
+      "learning_rate": 2.1786174600879617e-08,
+      "loss": 0.2445,
+      "step": 5320
+    },
+    {
+      "epoch": 2.98262331838565,
+      "grad_norm": 0.07899780552489176,
+      "learning_rate": 2.044585992608816e-08,
+      "loss": 0.2375,
+      "step": 5321
+    },
+    {
+      "epoch": 2.983183856502242,
+      "grad_norm": 0.07786507902858977,
+      "learning_rate": 1.9148089216936006e-08,
+      "loss": 0.239,
+      "step": 5322
+    },
+    {
+      "epoch": 2.9837443946188342,
+      "grad_norm": 0.08380194012054783,
+      "learning_rate": 1.7892863025648078e-08,
+      "loss": 0.2341,
+      "step": 5323
+    },
+    {
+      "epoch": 2.984304932735426,
+      "grad_norm": 0.08168210844370141,
+      "learning_rate": 1.6680181886352676e-08,
+      "loss": 0.2274,
+      "step": 5324
+    },
+    {
+      "epoch": 2.984865470852018,
+      "grad_norm": 0.08387252489451523,
+      "learning_rate": 1.5510046315092563e-08,
+      "loss": 0.2578,
+      "step": 5325
+    },
+    {
+      "epoch": 2.9854260089686098,
+      "grad_norm": 0.08217803833684102,
+      "learning_rate": 1.4382456809791667e-08,
+      "loss": 0.2474,
+      "step": 5326
+    },
+    {
+      "epoch": 2.9859865470852016,
+      "grad_norm": 0.07870568632951037,
+      "learning_rate": 1.329741385025507e-08,
+      "loss": 0.2403,
+      "step": 5327
+    },
+    {
+      "epoch": 2.986547085201794,
+      "grad_norm": 0.08120854289722704,
+      "learning_rate": 1.2254917898213425e-08,
+      "loss": 0.2336,
+      "step": 5328
+    },
+    {
+      "epoch": 2.9871076233183858,
+      "grad_norm": 0.08275810981080454,
+      "learning_rate": 1.1254969397267446e-08,
+      "loss": 0.246,
+      "step": 5329
+    },
+    {
+      "epoch": 2.9876681614349776,
+      "grad_norm": 0.07922153622750067,
+      "learning_rate": 1.0297568772921208e-08,
+      "loss": 0.2353,
+      "step": 5330
+    },
+    {
+      "epoch": 2.9882286995515694,
+      "grad_norm": 0.0801170106962022,
+      "learning_rate": 9.38271643258215e-09,
+      "loss": 0.2437,
+      "step": 5331
+    },
+    {
+      "epoch": 2.9887892376681613,
+      "grad_norm": 0.08220155343207602,
+      "learning_rate": 8.510412765538877e-09,
+      "loss": 0.2507,
+      "step": 5332
+    },
+    {
+      "epoch": 2.9893497757847536,
+      "grad_norm": 0.08188243844050613,
+      "learning_rate": 7.680658142972252e-09,
+      "loss": 0.2398,
+      "step": 5333
+    },
+    {
+      "epoch": 2.9899103139013454,
+      "grad_norm": 0.08040070756209229,
+      "learning_rate": 6.893452917977606e-09,
+      "loss": 0.2559,
+      "step": 5334
+    },
+    {
+      "epoch": 2.9904708520179373,
+      "grad_norm": 0.08049706669376877,
+      "learning_rate": 6.148797425520325e-09,
+      "loss": 0.2479,
+      "step": 5335
+    },
+    {
+      "epoch": 2.991031390134529,
+      "grad_norm": 0.07909212031292856,
+      "learning_rate": 5.44669198249137e-09,
+      "loss": 0.2419,
+      "step": 5336
+    },
+    {
+      "epoch": 2.991591928251121,
+      "grad_norm": 0.0798282239268473,
+      "learning_rate": 4.787136887629551e-09,
+      "loss": 0.2371,
+      "step": 5337
+    },
+    {
+      "epoch": 2.9921524663677133,
+      "grad_norm": 0.08070436214896615,
+      "learning_rate": 4.170132421610351e-09,
+      "loss": 0.2366,
+      "step": 5338
+    },
+    {
+      "epoch": 2.9927130044843047,
+      "grad_norm": 0.0779271714653247,
+      "learning_rate": 3.595678846979311e-09,
+      "loss": 0.2428,
+      "step": 5339
+    },
+    {
+      "epoch": 2.993273542600897,
+      "grad_norm": 0.08259654736944913,
+      "learning_rate": 3.0637764081853372e-09,
+      "loss": 0.2532,
+      "step": 5340
+    },
+    {
+      "epoch": 2.993834080717489,
+      "grad_norm": 0.07982008564088164,
+      "learning_rate": 2.574425331558494e-09,
+      "loss": 0.2365,
+      "step": 5341
+    },
+    {
+      "epoch": 2.9943946188340806,
+      "grad_norm": 0.0788106786967822,
+      "learning_rate": 2.127625825343316e-09,
+      "loss": 0.2401,
+      "step": 5342
+    },
+    {
+      "epoch": 2.9949551569506725,
+      "grad_norm": 0.08042716587308342,
+      "learning_rate": 1.723378079654392e-09,
+      "loss": 0.2358,
+      "step": 5343
+    },
+    {
+      "epoch": 2.9955156950672643,
+      "grad_norm": 0.08040778822418321,
+      "learning_rate": 1.3616822665096785e-09,
+      "loss": 0.2349,
+      "step": 5344
+    },
+    {
+      "epoch": 2.9960762331838566,
+      "grad_norm": 0.0804256952244607,
+      "learning_rate": 1.0425385398304955e-09,
+      "loss": 0.2333,
+      "step": 5345
+    },
+    {
+      "epoch": 2.9966367713004485,
+      "grad_norm": 0.08424987235173159,
+      "learning_rate": 7.659470354193232e-10,
+      "loss": 0.2492,
+      "step": 5346
+    },
+    {
+      "epoch": 2.9971973094170403,
+      "grad_norm": 0.08132774570248197,
+      "learning_rate": 5.319078709709047e-10,
+      "loss": 0.2483,
+      "step": 5347
+    },
+    {
+      "epoch": 2.997757847533632,
+      "grad_norm": 0.08075649495863021,
+      "learning_rate": 3.4042114606114284e-10,
+      "loss": 0.2315,
+      "step": 5348
+    },
+    {
+      "epoch": 2.998318385650224,
+      "grad_norm": 0.07950158267209038,
+      "learning_rate": 1.9148694219150997e-10,
+      "loss": 0.2441,
+      "step": 5349
+    },
+    {
+      "epoch": 2.9988789237668163,
+      "grad_norm": 0.08212722437926628,
+      "learning_rate": 8.510532273353633e-11,
+      "loss": 0.2496,
+      "step": 5350
+    },
+    {
+      "epoch": 2.999439461883408,
+      "grad_norm": 0.07977824407540027,
+      "learning_rate": 2.127633295101461e-11,
+      "loss": 0.2393,
+      "step": 5351
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.0776265795211927,
+      "learning_rate": 0.0,
+      "loss": 0.2329,
+      "step": 5352
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 0.26840582489967346,
+      "eval_runtime": 342.4131,
+      "eval_samples_per_second": 35.095,
+      "eval_steps_per_second": 1.098,
+      "step": 5352
+    },
+    {
+      "epoch": 3.0,
+      "step": 5352,
+      "total_flos": 1.7156250092814991e+18,
+      "train_loss": 0.28245328673299386,
+      "train_runtime": 54133.2624,
+      "train_samples_per_second": 12.652,
+      "train_steps_per_second": 0.099
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 5352,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "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": 1.7156250092814991e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}