diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,17080 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 30,
+  "global_step": 2346,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00128,
+      "grad_norm": 0.02388751693069935,
+      "learning_rate": 0.0,
+      "loss": 0.0005,
+      "step": 1
+    },
+    {
+      "epoch": 0.00256,
+      "grad_norm": 0.02199248969554901,
+      "learning_rate": 5.633802816901409e-07,
+      "loss": 0.0004,
+      "step": 2
+    },
+    {
+      "epoch": 0.00384,
+      "grad_norm": 0.019646070897579193,
+      "learning_rate": 1.1267605633802817e-06,
+      "loss": 0.0004,
+      "step": 3
+    },
+    {
+      "epoch": 0.00512,
+      "grad_norm": 0.020258139818906784,
+      "learning_rate": 1.6901408450704227e-06,
+      "loss": 0.0004,
+      "step": 4
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 0.01908482424914837,
+      "learning_rate": 2.2535211267605635e-06,
+      "loss": 0.0004,
+      "step": 5
+    },
+    {
+      "epoch": 0.00768,
+      "grad_norm": 0.015439272858202457,
+      "learning_rate": 2.8169014084507046e-06,
+      "loss": 0.0004,
+      "step": 6
+    },
+    {
+      "epoch": 0.00896,
+      "grad_norm": 0.013040938414633274,
+      "learning_rate": 3.3802816901408454e-06,
+      "loss": 0.0004,
+      "step": 7
+    },
+    {
+      "epoch": 0.01024,
+      "grad_norm": 0.01280244067311287,
+      "learning_rate": 3.943661971830986e-06,
+      "loss": 0.0004,
+      "step": 8
+    },
+    {
+      "epoch": 0.01152,
+      "grad_norm": 0.010930157266557217,
+      "learning_rate": 4.507042253521127e-06,
+      "loss": 0.0004,
+      "step": 9
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 0.011020738631486893,
+      "learning_rate": 5.070422535211268e-06,
+      "loss": 0.0004,
+      "step": 10
+    },
+    {
+      "epoch": 0.01408,
+      "grad_norm": 0.011919221840798855,
+      "learning_rate": 5.633802816901409e-06,
+      "loss": 0.0004,
+      "step": 11
+    },
+    {
+      "epoch": 0.01536,
+      "grad_norm": 0.016522666439414024,
+      "learning_rate": 6.197183098591549e-06,
+      "loss": 0.0004,
+      "step": 12
+    },
+    {
+      "epoch": 0.01664,
+      "grad_norm": 0.016699526458978653,
+      "learning_rate": 6.760563380281691e-06,
+      "loss": 0.0004,
+      "step": 13
+    },
+    {
+      "epoch": 0.01792,
+      "grad_norm": 0.01769128441810608,
+      "learning_rate": 7.3239436619718316e-06,
+      "loss": 0.0004,
+      "step": 14
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 0.018868286162614822,
+      "learning_rate": 7.887323943661972e-06,
+      "loss": 0.0004,
+      "step": 15
+    },
+    {
+      "epoch": 0.02048,
+      "grad_norm": 0.017053455114364624,
+      "learning_rate": 8.450704225352114e-06,
+      "loss": 0.0004,
+      "step": 16
+    },
+    {
+      "epoch": 0.02176,
+      "grad_norm": 0.01509174332022667,
+      "learning_rate": 9.014084507042254e-06,
+      "loss": 0.0004,
+      "step": 17
+    },
+    {
+      "epoch": 0.02304,
+      "grad_norm": 0.011613738723099232,
+      "learning_rate": 9.577464788732394e-06,
+      "loss": 0.0003,
+      "step": 18
+    },
+    {
+      "epoch": 0.02432,
+      "grad_norm": 0.006196283735334873,
+      "learning_rate": 1.0140845070422535e-05,
+      "loss": 0.0004,
+      "step": 19
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 0.0016393736004829407,
+      "learning_rate": 1.0704225352112675e-05,
+      "loss": 0.0003,
+      "step": 20
+    },
+    {
+      "epoch": 0.02688,
+      "grad_norm": 0.004450325388461351,
+      "learning_rate": 1.1267605633802819e-05,
+      "loss": 0.0003,
+      "step": 21
+    },
+    {
+      "epoch": 0.02816,
+      "grad_norm": 0.010002792812883854,
+      "learning_rate": 1.1830985915492958e-05,
+      "loss": 0.0003,
+      "step": 22
+    },
+    {
+      "epoch": 0.02944,
+      "grad_norm": 0.011458332650363445,
+      "learning_rate": 1.2394366197183098e-05,
+      "loss": 0.0003,
+      "step": 23
+    },
+    {
+      "epoch": 0.03072,
+      "grad_norm": 0.011857348494231701,
+      "learning_rate": 1.2957746478873242e-05,
+      "loss": 0.0003,
+      "step": 24
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 0.012468946166336536,
+      "learning_rate": 1.3521126760563382e-05,
+      "loss": 0.0003,
+      "step": 25
+    },
+    {
+      "epoch": 0.03328,
+      "grad_norm": 0.010764599777758121,
+      "learning_rate": 1.4084507042253522e-05,
+      "loss": 0.0003,
+      "step": 26
+    },
+    {
+      "epoch": 0.03456,
+      "grad_norm": 0.007388254161924124,
+      "learning_rate": 1.4647887323943663e-05,
+      "loss": 0.0003,
+      "step": 27
+    },
+    {
+      "epoch": 0.03584,
+      "grad_norm": 0.002933664945885539,
+      "learning_rate": 1.5211267605633803e-05,
+      "loss": 0.0003,
+      "step": 28
+    },
+    {
+      "epoch": 0.03712,
+      "grad_norm": 0.001106169424019754,
+      "learning_rate": 1.5774647887323945e-05,
+      "loss": 0.0003,
+      "step": 29
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 0.004174572881311178,
+      "learning_rate": 1.6338028169014086e-05,
+      "loss": 0.0003,
+      "step": 30
+    },
+    {
+      "epoch": 0.0384,
+      "eval_loss": 1.3955299854278564,
+      "eval_runtime": 45.1592,
+      "eval_samples_per_second": 11.116,
+      "eval_steps_per_second": 1.395,
+      "step": 30
+    },
+    {
+      "epoch": 0.03968,
+      "grad_norm": 0.0063258083537220955,
+      "learning_rate": 1.6901408450704228e-05,
+      "loss": 0.0003,
+      "step": 31
+    },
+    {
+      "epoch": 0.04096,
+      "grad_norm": 0.007857787422835827,
+      "learning_rate": 1.7464788732394366e-05,
+      "loss": 0.0003,
+      "step": 32
+    },
+    {
+      "epoch": 0.04224,
+      "grad_norm": 0.00830362644046545,
+      "learning_rate": 1.8028169014084508e-05,
+      "loss": 0.0003,
+      "step": 33
+    },
+    {
+      "epoch": 0.04352,
+      "grad_norm": 0.007050361018627882,
+      "learning_rate": 1.859154929577465e-05,
+      "loss": 0.0003,
+      "step": 34
+    },
+    {
+      "epoch": 0.0448,
+      "grad_norm": 0.004895492922514677,
+      "learning_rate": 1.9154929577464788e-05,
+      "loss": 0.0003,
+      "step": 35
+    },
+    {
+      "epoch": 0.04608,
+      "grad_norm": 0.0021485399920493364,
+      "learning_rate": 1.9718309859154933e-05,
+      "loss": 0.0003,
+      "step": 36
+    },
+    {
+      "epoch": 0.04736,
+      "grad_norm": 0.0009553782292641699,
+      "learning_rate": 2.028169014084507e-05,
+      "loss": 0.0003,
+      "step": 37
+    },
+    {
+      "epoch": 0.04864,
+      "grad_norm": 0.0037018507719039917,
+      "learning_rate": 2.0845070422535212e-05,
+      "loss": 0.0003,
+      "step": 38
+    },
+    {
+      "epoch": 0.04992,
+      "grad_norm": 0.006912225391715765,
+      "learning_rate": 2.140845070422535e-05,
+      "loss": 0.0003,
+      "step": 39
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 0.004844357259571552,
+      "learning_rate": 2.1971830985915496e-05,
+      "loss": 0.0002,
+      "step": 40
+    },
+    {
+      "epoch": 0.05248,
+      "grad_norm": 0.005295700393617153,
+      "learning_rate": 2.2535211267605637e-05,
+      "loss": 0.0003,
+      "step": 41
+    },
+    {
+      "epoch": 0.05376,
+      "grad_norm": 0.003574399510398507,
+      "learning_rate": 2.3098591549295775e-05,
+      "loss": 0.0002,
+      "step": 42
+    },
+    {
+      "epoch": 0.05504,
+      "grad_norm": 0.001562813762575388,
+      "learning_rate": 2.3661971830985917e-05,
+      "loss": 0.0003,
+      "step": 43
+    },
+    {
+      "epoch": 0.05632,
+      "grad_norm": 0.0006343662971630692,
+      "learning_rate": 2.422535211267606e-05,
+      "loss": 0.0003,
+      "step": 44
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 0.00290254526771605,
+      "learning_rate": 2.4788732394366197e-05,
+      "loss": 0.0003,
+      "step": 45
+    },
+    {
+      "epoch": 0.05888,
+      "grad_norm": 0.0033715504687279463,
+      "learning_rate": 2.5352112676056342e-05,
+      "loss": 0.0003,
+      "step": 46
+    },
+    {
+      "epoch": 0.06016,
+      "grad_norm": 0.0034714580979198217,
+      "learning_rate": 2.5915492957746483e-05,
+      "loss": 0.0002,
+      "step": 47
+    },
+    {
+      "epoch": 0.06144,
+      "grad_norm": 0.003244715742766857,
+      "learning_rate": 2.647887323943662e-05,
+      "loss": 0.0003,
+      "step": 48
+    },
+    {
+      "epoch": 0.06272,
+      "grad_norm": 0.001325926510617137,
+      "learning_rate": 2.7042253521126763e-05,
+      "loss": 0.0002,
+      "step": 49
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 0.00045873725321143866,
+      "learning_rate": 2.76056338028169e-05,
+      "loss": 0.0003,
+      "step": 50
+    },
+    {
+      "epoch": 0.06528,
+      "grad_norm": 0.0007484716479666531,
+      "learning_rate": 2.8169014084507043e-05,
+      "loss": 0.0003,
+      "step": 51
+    },
+    {
+      "epoch": 0.06656,
+      "grad_norm": 0.0027240754570811987,
+      "learning_rate": 2.8732394366197188e-05,
+      "loss": 0.0003,
+      "step": 52
+    },
+    {
+      "epoch": 0.06784,
+      "grad_norm": 0.002920588944107294,
+      "learning_rate": 2.9295774647887326e-05,
+      "loss": 0.0003,
+      "step": 53
+    },
+    {
+      "epoch": 0.06912,
+      "grad_norm": 0.0020773897413164377,
+      "learning_rate": 2.9859154929577468e-05,
+      "loss": 0.0003,
+      "step": 54
+    },
+    {
+      "epoch": 0.0704,
+      "grad_norm": 0.0013842222979292274,
+      "learning_rate": 3.0422535211267606e-05,
+      "loss": 0.0003,
+      "step": 55
+    },
+    {
+      "epoch": 0.07168,
+      "grad_norm": 0.0004421753983478993,
+      "learning_rate": 3.0985915492957744e-05,
+      "loss": 0.0003,
+      "step": 56
+    },
+    {
+      "epoch": 0.07296,
+      "grad_norm": 0.0009825655724853277,
+      "learning_rate": 3.154929577464789e-05,
+      "loss": 0.0002,
+      "step": 57
+    },
+    {
+      "epoch": 0.07424,
+      "grad_norm": 0.0016538031632080674,
+      "learning_rate": 3.2112676056338034e-05,
+      "loss": 0.0003,
+      "step": 58
+    },
+    {
+      "epoch": 0.07552,
+      "grad_norm": 0.0019267118768766522,
+      "learning_rate": 3.267605633802817e-05,
+      "loss": 0.0003,
+      "step": 59
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 0.001848129671998322,
+      "learning_rate": 3.323943661971831e-05,
+      "loss": 0.0002,
+      "step": 60
+    },
+    {
+      "epoch": 0.0768,
+      "eval_loss": 1.2682050466537476,
+      "eval_runtime": 40.917,
+      "eval_samples_per_second": 12.269,
+      "eval_steps_per_second": 1.54,
+      "step": 60
+    },
+    {
+      "epoch": 0.07808,
+      "grad_norm": 0.0008087851456366479,
+      "learning_rate": 3.3802816901408456e-05,
+      "loss": 0.0002,
+      "step": 61
+    },
+    {
+      "epoch": 0.07936,
+      "grad_norm": 0.00034242391120642424,
+      "learning_rate": 3.4366197183098594e-05,
+      "loss": 0.0003,
+      "step": 62
+    },
+    {
+      "epoch": 0.08064,
+      "grad_norm": 0.0011758297914639115,
+      "learning_rate": 3.492957746478873e-05,
+      "loss": 0.0002,
+      "step": 63
+    },
+    {
+      "epoch": 0.08192,
+      "grad_norm": 0.0014905749121680856,
+      "learning_rate": 3.549295774647888e-05,
+      "loss": 0.0002,
+      "step": 64
+    },
+    {
+      "epoch": 0.0832,
+      "grad_norm": 0.0013011845294386148,
+      "learning_rate": 3.6056338028169015e-05,
+      "loss": 0.0002,
+      "step": 65
+    },
+    {
+      "epoch": 0.08448,
+      "grad_norm": 0.0011716658482328057,
+      "learning_rate": 3.661971830985916e-05,
+      "loss": 0.0003,
+      "step": 66
+    },
+    {
+      "epoch": 0.08576,
+      "grad_norm": 0.00024143150949385017,
+      "learning_rate": 3.71830985915493e-05,
+      "loss": 0.0002,
+      "step": 67
+    },
+    {
+      "epoch": 0.08704,
+      "grad_norm": 0.0006570953410118818,
+      "learning_rate": 3.774647887323944e-05,
+      "loss": 0.0002,
+      "step": 68
+    },
+    {
+      "epoch": 0.08832,
+      "grad_norm": 0.0012625143863260746,
+      "learning_rate": 3.8309859154929575e-05,
+      "loss": 0.0003,
+      "step": 69
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 0.0007922302465885878,
+      "learning_rate": 3.887323943661972e-05,
+      "loss": 0.0003,
+      "step": 70
+    },
+    {
+      "epoch": 0.09088,
+      "grad_norm": 0.0008460975368507206,
+      "learning_rate": 3.9436619718309865e-05,
+      "loss": 0.0002,
+      "step": 71
+    },
+    {
+      "epoch": 0.09216,
+      "grad_norm": 0.00019190074817743152,
+      "learning_rate": 4e-05,
+      "loss": 0.0003,
+      "step": 72
+    },
+    {
+      "epoch": 0.09344,
+      "grad_norm": 0.00025078540784306824,
+      "learning_rate": 3.99999828375461e-05,
+      "loss": 0.0002,
+      "step": 73
+    },
+    {
+      "epoch": 0.09472,
+      "grad_norm": 0.0007434043800458312,
+      "learning_rate": 3.999993135021711e-05,
+      "loss": 0.0003,
+      "step": 74
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 0.0010244562290608883,
+      "learning_rate": 3.999984553811122e-05,
+      "loss": 0.0003,
+      "step": 75
+    },
+    {
+      "epoch": 0.09728,
+      "grad_norm": 0.000786862859968096,
+      "learning_rate": 3.999972540139207e-05,
+      "loss": 0.0003,
+      "step": 76
+    },
+    {
+      "epoch": 0.09856,
+      "grad_norm": 0.0001873165019787848,
+      "learning_rate": 3.9999570940288754e-05,
+      "loss": 0.0002,
+      "step": 77
+    },
+    {
+      "epoch": 0.09984,
+      "grad_norm": 0.000646409927867353,
+      "learning_rate": 3.999938215509581e-05,
+      "loss": 0.0003,
+      "step": 78
+    },
+    {
+      "epoch": 0.10112,
+      "grad_norm": 0.0005671542603522539,
+      "learning_rate": 3.999915904617326e-05,
+      "loss": 0.0003,
+      "step": 79
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 0.00029301070026122034,
+      "learning_rate": 3.999890161394654e-05,
+      "loss": 0.0002,
+      "step": 80
+    },
+    {
+      "epoch": 0.10368,
+      "grad_norm": 0.0006586903473362327,
+      "learning_rate": 3.9998609858906565e-05,
+      "loss": 0.0002,
+      "step": 81
+    },
+    {
+      "epoch": 0.10496,
+      "grad_norm": 0.00024197845777962357,
+      "learning_rate": 3.99982837816097e-05,
+      "loss": 0.0003,
+      "step": 82
+    },
+    {
+      "epoch": 0.10624,
+      "grad_norm": 0.0005213619442656636,
+      "learning_rate": 3.999792338267774e-05,
+      "loss": 0.0002,
+      "step": 83
+    },
+    {
+      "epoch": 0.10752,
+      "grad_norm": 0.0006407280452549458,
+      "learning_rate": 3.9997528662797956e-05,
+      "loss": 0.0002,
+      "step": 84
+    },
+    {
+      "epoch": 0.1088,
+      "grad_norm": 0.00033941137371584773,
+      "learning_rate": 3.999709962272305e-05,
+      "loss": 0.0003,
+      "step": 85
+    },
+    {
+      "epoch": 0.11008,
+      "grad_norm": 0.00016794790280982852,
+      "learning_rate": 3.9996636263271184e-05,
+      "loss": 0.0002,
+      "step": 86
+    },
+    {
+      "epoch": 0.11136,
+      "grad_norm": 0.0004396963631734252,
+      "learning_rate": 3.9996138585325935e-05,
+      "loss": 0.0002,
+      "step": 87
+    },
+    {
+      "epoch": 0.11264,
+      "grad_norm": 0.00031045766081660986,
+      "learning_rate": 3.999560658983637e-05,
+      "loss": 0.0002,
+      "step": 88
+    },
+    {
+      "epoch": 0.11392,
+      "grad_norm": 0.00012366034206934273,
+      "learning_rate": 3.9995040277816956e-05,
+      "loss": 0.0002,
+      "step": 89
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 0.00011263033957220614,
+      "learning_rate": 3.999443965034762e-05,
+      "loss": 0.0002,
+      "step": 90
+    },
+    {
+      "epoch": 0.1152,
+      "eval_loss": 1.2403218746185303,
+      "eval_runtime": 40.6319,
+      "eval_samples_per_second": 12.355,
+      "eval_steps_per_second": 1.551,
+      "step": 90
+    },
+    {
+      "epoch": 0.11648,
+      "grad_norm": 0.0001996486244024709,
+      "learning_rate": 3.9993804708573724e-05,
+      "loss": 0.0003,
+      "step": 91
+    },
+    {
+      "epoch": 0.11776,
+      "grad_norm": 0.0002045943110715598,
+      "learning_rate": 3.999313545370607e-05,
+      "loss": 0.0002,
+      "step": 92
+    },
+    {
+      "epoch": 0.11904,
+      "grad_norm": 0.00012093196710338816,
+      "learning_rate": 3.999243188702086e-05,
+      "loss": 0.0002,
+      "step": 93
+    },
+    {
+      "epoch": 0.12032,
+      "grad_norm": 0.00030544595210812986,
+      "learning_rate": 3.9991694009859784e-05,
+      "loss": 0.0002,
+      "step": 94
+    },
+    {
+      "epoch": 0.1216,
+      "grad_norm": 0.0002122317673638463,
+      "learning_rate": 3.9990921823629916e-05,
+      "loss": 0.0002,
+      "step": 95
+    },
+    {
+      "epoch": 0.12288,
+      "grad_norm": 0.0001355641143163666,
+      "learning_rate": 3.999011532980377e-05,
+      "loss": 0.0002,
+      "step": 96
+    },
+    {
+      "epoch": 0.12416,
+      "grad_norm": 0.0002007813163800165,
+      "learning_rate": 3.998927452991927e-05,
+      "loss": 0.0003,
+      "step": 97
+    },
+    {
+      "epoch": 0.12544,
+      "grad_norm": 0.0002493359206710011,
+      "learning_rate": 3.998839942557979e-05,
+      "loss": 0.0002,
+      "step": 98
+    },
+    {
+      "epoch": 0.12672,
+      "grad_norm": 0.0001009925763355568,
+      "learning_rate": 3.9987490018454085e-05,
+      "loss": 0.0002,
+      "step": 99
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 0.00012175164738437161,
+      "learning_rate": 3.998654631027635e-05,
+      "loss": 0.0002,
+      "step": 100
+    },
+    {
+      "epoch": 0.12928,
+      "grad_norm": 0.0003556807350832969,
+      "learning_rate": 3.9985568302846166e-05,
+      "loss": 0.0002,
+      "step": 101
+    },
+    {
+      "epoch": 0.13056,
+      "grad_norm": 0.0003202245570719242,
+      "learning_rate": 3.9984555998028544e-05,
+      "loss": 0.0003,
+      "step": 102
+    },
+    {
+      "epoch": 0.13184,
+      "grad_norm": 0.0003585947270039469,
+      "learning_rate": 3.998350939775389e-05,
+      "loss": 0.0002,
+      "step": 103
+    },
+    {
+      "epoch": 0.13312,
+      "grad_norm": 0.00013465459051076323,
+      "learning_rate": 3.9982428504018006e-05,
+      "loss": 0.0003,
+      "step": 104
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 0.00012200856144772843,
+      "learning_rate": 3.998131331888208e-05,
+      "loss": 0.0002,
+      "step": 105
+    },
+    {
+      "epoch": 0.13568,
+      "grad_norm": 0.000345123145962134,
+      "learning_rate": 3.998016384447271e-05,
+      "loss": 0.0002,
+      "step": 106
+    },
+    {
+      "epoch": 0.13696,
+      "grad_norm": 0.00037582783261314034,
+      "learning_rate": 3.997898008298188e-05,
+      "loss": 0.0003,
+      "step": 107
+    },
+    {
+      "epoch": 0.13824,
+      "grad_norm": 0.00011709384125424549,
+      "learning_rate": 3.997776203666694e-05,
+      "loss": 0.0002,
+      "step": 108
+    },
+    {
+      "epoch": 0.13952,
+      "grad_norm": 0.0005938383401371539,
+      "learning_rate": 3.9976509707850645e-05,
+      "loss": 0.0002,
+      "step": 109
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 0.0001890748826554045,
+      "learning_rate": 3.9975223098921094e-05,
+      "loss": 0.0002,
+      "step": 110
+    },
+    {
+      "epoch": 0.14208,
+      "grad_norm": 0.00018463779997546226,
+      "learning_rate": 3.997390221233178e-05,
+      "loss": 0.0002,
+      "step": 111
+    },
+    {
+      "epoch": 0.14336,
+      "grad_norm": 0.0003491342649795115,
+      "learning_rate": 3.997254705060155e-05,
+      "loss": 0.0002,
+      "step": 112
+    },
+    {
+      "epoch": 0.14464,
+      "grad_norm": 0.00032335062860511243,
+      "learning_rate": 3.997115761631462e-05,
+      "loss": 0.0002,
+      "step": 113
+    },
+    {
+      "epoch": 0.14592,
+      "grad_norm": 0.00013327212946023792,
+      "learning_rate": 3.996973391212056e-05,
+      "loss": 0.0002,
+      "step": 114
+    },
+    {
+      "epoch": 0.1472,
+      "grad_norm": 0.0003823594015557319,
+      "learning_rate": 3.996827594073428e-05,
+      "loss": 0.0002,
+      "step": 115
+    },
+    {
+      "epoch": 0.14848,
+      "grad_norm": 0.00010233603097731248,
+      "learning_rate": 3.996678370493604e-05,
+      "loss": 0.0002,
+      "step": 116
+    },
+    {
+      "epoch": 0.14976,
+      "grad_norm": 0.00015697085473220795,
+      "learning_rate": 3.996525720757145e-05,
+      "loss": 0.0002,
+      "step": 117
+    },
+    {
+      "epoch": 0.15104,
+      "grad_norm": 9.921671880874783e-05,
+      "learning_rate": 3.996369645155145e-05,
+      "loss": 0.0002,
+      "step": 118
+    },
+    {
+      "epoch": 0.15232,
+      "grad_norm": 0.00010817526344908401,
+      "learning_rate": 3.99621014398523e-05,
+      "loss": 0.0002,
+      "step": 119
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 0.0003328252932988107,
+      "learning_rate": 3.9960472175515596e-05,
+      "loss": 0.0003,
+      "step": 120
+    },
+    {
+      "epoch": 0.1536,
+      "eval_loss": 1.2306902408599854,
+      "eval_runtime": 41.0777,
+      "eval_samples_per_second": 12.221,
+      "eval_steps_per_second": 1.534,
+      "step": 120
+    },
+    {
+      "epoch": 0.15488,
+      "grad_norm": 0.0001208317480632104,
+      "learning_rate": 3.995880866164824e-05,
+      "loss": 0.0002,
+      "step": 121
+    },
+    {
+      "epoch": 0.15616,
+      "grad_norm": 0.00010686482710298151,
+      "learning_rate": 3.995711090142246e-05,
+      "loss": 0.0002,
+      "step": 122
+    },
+    {
+      "epoch": 0.15744,
+      "grad_norm": 0.000191230108612217,
+      "learning_rate": 3.9955378898075776e-05,
+      "loss": 0.0002,
+      "step": 123
+    },
+    {
+      "epoch": 0.15872,
+      "grad_norm": 0.00043578416807577014,
+      "learning_rate": 3.995361265491102e-05,
+      "loss": 0.0002,
+      "step": 124
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.00012811734632123262,
+      "learning_rate": 3.99518121752963e-05,
+      "loss": 0.0002,
+      "step": 125
+    },
+    {
+      "epoch": 0.16128,
+      "grad_norm": 0.00011695229477481917,
+      "learning_rate": 3.994997746266502e-05,
+      "loss": 0.0002,
+      "step": 126
+    },
+    {
+      "epoch": 0.16256,
+      "grad_norm": 0.00026602434809319675,
+      "learning_rate": 3.994810852051589e-05,
+      "loss": 0.0002,
+      "step": 127
+    },
+    {
+      "epoch": 0.16384,
+      "grad_norm": 0.00020878778013866395,
+      "learning_rate": 3.9946205352412836e-05,
+      "loss": 0.0002,
+      "step": 128
+    },
+    {
+      "epoch": 0.16512,
+      "grad_norm": 0.00013080754433758557,
+      "learning_rate": 3.994426796198511e-05,
+      "loss": 0.0002,
+      "step": 129
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 0.00017335913435090333,
+      "learning_rate": 3.994229635292718e-05,
+      "loss": 0.0002,
+      "step": 130
+    },
+    {
+      "epoch": 0.16768,
+      "grad_norm": 0.0001642439019633457,
+      "learning_rate": 3.99402905289988e-05,
+      "loss": 0.0002,
+      "step": 131
+    },
+    {
+      "epoch": 0.16896,
+      "grad_norm": 0.00011353264562785625,
+      "learning_rate": 3.993825049402494e-05,
+      "loss": 0.0002,
+      "step": 132
+    },
+    {
+      "epoch": 0.17024,
+      "grad_norm": 0.0003552866692189127,
+      "learning_rate": 3.993617625189584e-05,
+      "loss": 0.0002,
+      "step": 133
+    },
+    {
+      "epoch": 0.17152,
+      "grad_norm": 0.0002389036671957001,
+      "learning_rate": 3.993406780656694e-05,
+      "loss": 0.0002,
+      "step": 134
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 0.00018962110334541649,
+      "learning_rate": 3.993192516205892e-05,
+      "loss": 0.0002,
+      "step": 135
+    },
+    {
+      "epoch": 0.17408,
+      "grad_norm": 0.000491205370053649,
+      "learning_rate": 3.9929748322457674e-05,
+      "loss": 0.0002,
+      "step": 136
+    },
+    {
+      "epoch": 0.17536,
+      "grad_norm": 0.00025262351846322417,
+      "learning_rate": 3.992753729191431e-05,
+      "loss": 0.0002,
+      "step": 137
+    },
+    {
+      "epoch": 0.17664,
+      "grad_norm": 0.0002617812715470791,
+      "learning_rate": 3.992529207464512e-05,
+      "loss": 0.0003,
+      "step": 138
+    },
+    {
+      "epoch": 0.17792,
+      "grad_norm": 0.0002854976919479668,
+      "learning_rate": 3.992301267493159e-05,
+      "loss": 0.0002,
+      "step": 139
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 0.0001424322254024446,
+      "learning_rate": 3.9920699097120414e-05,
+      "loss": 0.0002,
+      "step": 140
+    },
+    {
+      "epoch": 0.18048,
+      "grad_norm": 0.00023581937421113253,
+      "learning_rate": 3.991835134562344e-05,
+      "loss": 0.0002,
+      "step": 141
+    },
+    {
+      "epoch": 0.18176,
+      "grad_norm": 0.0003542240883689374,
+      "learning_rate": 3.991596942491768e-05,
+      "loss": 0.0002,
+      "step": 142
+    },
+    {
+      "epoch": 0.18304,
+      "grad_norm": 0.0003476674319244921,
+      "learning_rate": 3.9913553339545315e-05,
+      "loss": 0.0002,
+      "step": 143
+    },
+    {
+      "epoch": 0.18432,
+      "grad_norm": 0.00037666381103917956,
+      "learning_rate": 3.991110309411368e-05,
+      "loss": 0.0003,
+      "step": 144
+    },
+    {
+      "epoch": 0.1856,
+      "grad_norm": 0.00011861531675094739,
+      "learning_rate": 3.990861869329524e-05,
+      "loss": 0.0002,
+      "step": 145
+    },
+    {
+      "epoch": 0.18688,
+      "grad_norm": 0.0004715830145869404,
+      "learning_rate": 3.99061001418276e-05,
+      "loss": 0.0002,
+      "step": 146
+    },
+    {
+      "epoch": 0.18816,
+      "grad_norm": 0.00014487896987702698,
+      "learning_rate": 3.990354744451347e-05,
+      "loss": 0.0002,
+      "step": 147
+    },
+    {
+      "epoch": 0.18944,
+      "grad_norm": 0.00035398933687247336,
+      "learning_rate": 3.990096060622071e-05,
+      "loss": 0.0002,
+      "step": 148
+    },
+    {
+      "epoch": 0.19072,
+      "grad_norm": 0.0005021999822929502,
+      "learning_rate": 3.989833963188225e-05,
+      "loss": 0.0002,
+      "step": 149
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 0.00015732903557363898,
+      "learning_rate": 3.989568452649613e-05,
+      "loss": 0.0002,
+      "step": 150
+    },
+    {
+      "epoch": 0.192,
+      "eval_loss": 1.2276506423950195,
+      "eval_runtime": 41.6549,
+      "eval_samples_per_second": 12.051,
+      "eval_steps_per_second": 1.512,
+      "step": 150
+    },
+    {
+      "epoch": 0.19328,
+      "grad_norm": 0.0005945006851106882,
+      "learning_rate": 3.989299529512548e-05,
+      "loss": 0.0002,
+      "step": 151
+    },
+    {
+      "epoch": 0.19456,
+      "grad_norm": 0.0004206049197819084,
+      "learning_rate": 3.98902719428985e-05,
+      "loss": 0.0002,
+      "step": 152
+    },
+    {
+      "epoch": 0.19584,
+      "grad_norm": 0.0005308086983859539,
+      "learning_rate": 3.988751447500845e-05,
+      "loss": 0.0002,
+      "step": 153
+    },
+    {
+      "epoch": 0.19712,
+      "grad_norm": 0.0003311938198748976,
+      "learning_rate": 3.988472289671367e-05,
+      "loss": 0.0002,
+      "step": 154
+    },
+    {
+      "epoch": 0.1984,
+      "grad_norm": 0.00020781578496098518,
+      "learning_rate": 3.988189721333751e-05,
+      "loss": 0.0003,
+      "step": 155
+    },
+    {
+      "epoch": 0.19968,
+      "grad_norm": 0.0002495672379154712,
+      "learning_rate": 3.9879037430268394e-05,
+      "loss": 0.0002,
+      "step": 156
+    },
+    {
+      "epoch": 0.20096,
+      "grad_norm": 0.0003674014296848327,
+      "learning_rate": 3.987614355295975e-05,
+      "loss": 0.0002,
+      "step": 157
+    },
+    {
+      "epoch": 0.20224,
+      "grad_norm": 0.00011461845133453608,
+      "learning_rate": 3.987321558693002e-05,
+      "loss": 0.0002,
+      "step": 158
+    },
+    {
+      "epoch": 0.20352,
+      "grad_norm": 0.00014985799498390406,
+      "learning_rate": 3.9870253537762666e-05,
+      "loss": 0.0002,
+      "step": 159
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 0.00023186200996860862,
+      "learning_rate": 3.9867257411106126e-05,
+      "loss": 0.0002,
+      "step": 160
+    },
+    {
+      "epoch": 0.20608,
+      "grad_norm": 0.00011972602806054056,
+      "learning_rate": 3.986422721267384e-05,
+      "loss": 0.0002,
+      "step": 161
+    },
+    {
+      "epoch": 0.20736,
+      "grad_norm": 0.0001768232323229313,
+      "learning_rate": 3.986116294824421e-05,
+      "loss": 0.0002,
+      "step": 162
+    },
+    {
+      "epoch": 0.20864,
+      "grad_norm": 0.00023504432465415448,
+      "learning_rate": 3.985806462366061e-05,
+      "loss": 0.0002,
+      "step": 163
+    },
+    {
+      "epoch": 0.20992,
+      "grad_norm": 0.00016327628691215068,
+      "learning_rate": 3.9854932244831334e-05,
+      "loss": 0.0002,
+      "step": 164
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 0.00029715322307311,
+      "learning_rate": 3.985176581772967e-05,
+      "loss": 0.0002,
+      "step": 165
+    },
+    {
+      "epoch": 0.21248,
+      "grad_norm": 0.00016738087288103998,
+      "learning_rate": 3.984856534839378e-05,
+      "loss": 0.0002,
+      "step": 166
+    },
+    {
+      "epoch": 0.21376,
+      "grad_norm": 0.0007520260405726731,
+      "learning_rate": 3.984533084292677e-05,
+      "loss": 0.0002,
+      "step": 167
+    },
+    {
+      "epoch": 0.21504,
+      "grad_norm": 0.0003500195743981749,
+      "learning_rate": 3.984206230749664e-05,
+      "loss": 0.0002,
+      "step": 168
+    },
+    {
+      "epoch": 0.21632,
+      "grad_norm": 0.0006380021804943681,
+      "learning_rate": 3.9838759748336306e-05,
+      "loss": 0.0002,
+      "step": 169
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 0.0008250039536505938,
+      "learning_rate": 3.983542317174354e-05,
+      "loss": 0.0003,
+      "step": 170
+    },
+    {
+      "epoch": 0.21888,
+      "grad_norm": 0.00023163011064752936,
+      "learning_rate": 3.983205258408097e-05,
+      "loss": 0.0002,
+      "step": 171
+    },
+    {
+      "epoch": 0.22016,
+      "grad_norm": 0.0009266913402825594,
+      "learning_rate": 3.982864799177614e-05,
+      "loss": 0.0002,
+      "step": 172
+    },
+    {
+      "epoch": 0.22144,
+      "grad_norm": 0.0007819707971066236,
+      "learning_rate": 3.9825209401321375e-05,
+      "loss": 0.0002,
+      "step": 173
+    },
+    {
+      "epoch": 0.22272,
+      "grad_norm": 0.0009147466043941677,
+      "learning_rate": 3.9821736819273865e-05,
+      "loss": 0.0002,
+      "step": 174
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 0.0006302861729636788,
+      "learning_rate": 3.9818230252255614e-05,
+      "loss": 0.0002,
+      "step": 175
+    },
+    {
+      "epoch": 0.22528,
+      "grad_norm": 0.0007533888565376401,
+      "learning_rate": 3.981468970695343e-05,
+      "loss": 0.0002,
+      "step": 176
+    },
+    {
+      "epoch": 0.22656,
+      "grad_norm": 0.0003364177537150681,
+      "learning_rate": 3.981111519011892e-05,
+      "loss": 0.0002,
+      "step": 177
+    },
+    {
+      "epoch": 0.22784,
+      "grad_norm": 0.0015544546768069267,
+      "learning_rate": 3.9807506708568476e-05,
+      "loss": 0.0002,
+      "step": 178
+    },
+    {
+      "epoch": 0.22912,
+      "grad_norm": 0.00048260428593494,
+      "learning_rate": 3.9803864269183244e-05,
+      "loss": 0.0002,
+      "step": 179
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 0.0009548643138259649,
+      "learning_rate": 3.980018787890914e-05,
+      "loss": 0.0002,
+      "step": 180
+    },
+    {
+      "epoch": 0.2304,
+      "eval_loss": 1.2241315841674805,
+      "eval_runtime": 63.3318,
+      "eval_samples_per_second": 7.927,
+      "eval_steps_per_second": 0.995,
+      "step": 180
+    },
+    {
+      "epoch": 0.23168,
+      "grad_norm": 0.0013649332104250789,
+      "learning_rate": 3.9796477544756814e-05,
+      "loss": 0.0002,
+      "step": 181
+    },
+    {
+      "epoch": 0.23296,
+      "grad_norm": 0.000332722527673468,
+      "learning_rate": 3.979273327380165e-05,
+      "loss": 0.0002,
+      "step": 182
+    },
+    {
+      "epoch": 0.23424,
+      "grad_norm": 0.0010766334598883986,
+      "learning_rate": 3.9788955073183747e-05,
+      "loss": 0.0002,
+      "step": 183
+    },
+    {
+      "epoch": 0.23552,
+      "grad_norm": 0.0004133693582843989,
+      "learning_rate": 3.97851429501079e-05,
+      "loss": 0.0002,
+      "step": 184
+    },
+    {
+      "epoch": 0.2368,
+      "grad_norm": 0.00047085783444345,
+      "learning_rate": 3.9781296911843606e-05,
+      "loss": 0.0002,
+      "step": 185
+    },
+    {
+      "epoch": 0.23808,
+      "grad_norm": 0.000769197242334485,
+      "learning_rate": 3.977741696572501e-05,
+      "loss": 0.0002,
+      "step": 186
+    },
+    {
+      "epoch": 0.23936,
+      "grad_norm": 0.0002509950427338481,
+      "learning_rate": 3.9773503119150955e-05,
+      "loss": 0.0002,
+      "step": 187
+    },
+    {
+      "epoch": 0.24064,
+      "grad_norm": 0.0010357450228184462,
+      "learning_rate": 3.976955537958489e-05,
+      "loss": 0.0002,
+      "step": 188
+    },
+    {
+      "epoch": 0.24192,
+      "grad_norm": 0.0006930662202648818,
+      "learning_rate": 3.9765573754554924e-05,
+      "loss": 0.0002,
+      "step": 189
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 0.0009113008854910731,
+      "learning_rate": 3.9761558251653774e-05,
+      "loss": 0.0002,
+      "step": 190
+    },
+    {
+      "epoch": 0.24448,
+      "grad_norm": 0.0005375745822675526,
+      "learning_rate": 3.975750887853876e-05,
+      "loss": 0.0002,
+      "step": 191
+    },
+    {
+      "epoch": 0.24576,
+      "grad_norm": 0.0002424489357508719,
+      "learning_rate": 3.9753425642931786e-05,
+      "loss": 0.0002,
+      "step": 192
+    },
+    {
+      "epoch": 0.24704,
+      "grad_norm": 0.000845865230076015,
+      "learning_rate": 3.974930855261935e-05,
+      "loss": 0.0002,
+      "step": 193
+    },
+    {
+      "epoch": 0.24832,
+      "grad_norm": 0.0004812710976693779,
+      "learning_rate": 3.974515761545247e-05,
+      "loss": 0.0002,
+      "step": 194
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 0.00042488076724112034,
+      "learning_rate": 3.974097283934675e-05,
+      "loss": 0.0002,
+      "step": 195
+    },
+    {
+      "epoch": 0.25088,
+      "grad_norm": 0.00011899595847353339,
+      "learning_rate": 3.9736754232282295e-05,
+      "loss": 0.0002,
+      "step": 196
+    },
+    {
+      "epoch": 0.25216,
+      "grad_norm": 0.000173148131580092,
+      "learning_rate": 3.9732501802303735e-05,
+      "loss": 0.0002,
+      "step": 197
+    },
+    {
+      "epoch": 0.25344,
+      "grad_norm": 0.0010396325960755348,
+      "learning_rate": 3.9728215557520194e-05,
+      "loss": 0.0002,
+      "step": 198
+    },
+    {
+      "epoch": 0.25472,
+      "grad_norm": 0.0009005111642181873,
+      "learning_rate": 3.972389550610528e-05,
+      "loss": 0.0002,
+      "step": 199
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 0.000827074924018234,
+      "learning_rate": 3.9719541656297076e-05,
+      "loss": 0.0002,
+      "step": 200
+    },
+    {
+      "epoch": 0.25728,
+      "grad_norm": 0.00014234821719583124,
+      "learning_rate": 3.9715154016398114e-05,
+      "loss": 0.0002,
+      "step": 201
+    },
+    {
+      "epoch": 0.25856,
+      "grad_norm": 0.0020800912752747536,
+      "learning_rate": 3.9710732594775334e-05,
+      "loss": 0.0002,
+      "step": 202
+    },
+    {
+      "epoch": 0.25984,
+      "grad_norm": 0.0010654254583641887,
+      "learning_rate": 3.970627739986014e-05,
+      "loss": 0.0002,
+      "step": 203
+    },
+    {
+      "epoch": 0.26112,
+      "grad_norm": 0.001251160865649581,
+      "learning_rate": 3.970178844014831e-05,
+      "loss": 0.0002,
+      "step": 204
+    },
+    {
+      "epoch": 0.2624,
+      "grad_norm": 0.0004991651512682438,
+      "learning_rate": 3.969726572420002e-05,
+      "loss": 0.0002,
+      "step": 205
+    },
+    {
+      "epoch": 0.26368,
+      "grad_norm": 0.0012509931111708283,
+      "learning_rate": 3.9692709260639815e-05,
+      "loss": 0.0002,
+      "step": 206
+    },
+    {
+      "epoch": 0.26496,
+      "grad_norm": 0.00045681357732973993,
+      "learning_rate": 3.968811905815659e-05,
+      "loss": 0.0003,
+      "step": 207
+    },
+    {
+      "epoch": 0.26624,
+      "grad_norm": 0.00115011737216264,
+      "learning_rate": 3.968349512550359e-05,
+      "loss": 0.0002,
+      "step": 208
+    },
+    {
+      "epoch": 0.26752,
+      "grad_norm": 0.00021435614326037467,
+      "learning_rate": 3.967883747149837e-05,
+      "loss": 0.0002,
+      "step": 209
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 0.0015478826826438308,
+      "learning_rate": 3.9674146105022795e-05,
+      "loss": 0.0002,
+      "step": 210
+    },
+    {
+      "epoch": 0.2688,
+      "eval_loss": 1.2069844007492065,
+      "eval_runtime": 42.1506,
+      "eval_samples_per_second": 11.91,
+      "eval_steps_per_second": 1.495,
+      "step": 210
+    },
+    {
+      "epoch": 0.27008,
+      "grad_norm": 0.00039821109385229647,
+      "learning_rate": 3.966942103502301e-05,
+      "loss": 0.0002,
+      "step": 211
+    },
+    {
+      "epoch": 0.27136,
+      "grad_norm": 0.0013712712097913027,
+      "learning_rate": 3.966466227050945e-05,
+      "loss": 0.0002,
+      "step": 212
+    },
+    {
+      "epoch": 0.27264,
+      "grad_norm": 0.0006687208078801632,
+      "learning_rate": 3.965986982055677e-05,
+      "loss": 0.0002,
+      "step": 213
+    },
+    {
+      "epoch": 0.27392,
+      "grad_norm": 0.0006046419148333371,
+      "learning_rate": 3.96550436943039e-05,
+      "loss": 0.0002,
+      "step": 214
+    },
+    {
+      "epoch": 0.2752,
+      "grad_norm": 0.0005272756097838283,
+      "learning_rate": 3.965018390095396e-05,
+      "loss": 0.0002,
+      "step": 215
+    },
+    {
+      "epoch": 0.27648,
+      "grad_norm": 0.00027221845812164247,
+      "learning_rate": 3.964529044977429e-05,
+      "loss": 0.0002,
+      "step": 216
+    },
+    {
+      "epoch": 0.27776,
+      "grad_norm": 0.00022938432812225074,
+      "learning_rate": 3.9640363350096396e-05,
+      "loss": 0.0002,
+      "step": 217
+    },
+    {
+      "epoch": 0.27904,
+      "grad_norm": 0.0005330294952727854,
+      "learning_rate": 3.963540261131596e-05,
+      "loss": 0.0002,
+      "step": 218
+    },
+    {
+      "epoch": 0.28032,
+      "grad_norm": 0.0008340986678376794,
+      "learning_rate": 3.963040824289281e-05,
+      "loss": 0.0002,
+      "step": 219
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 0.00021592361736111343,
+      "learning_rate": 3.96253802543509e-05,
+      "loss": 0.0002,
+      "step": 220
+    },
+    {
+      "epoch": 0.28288,
+      "grad_norm": 0.0010405798675492406,
+      "learning_rate": 3.9620318655278315e-05,
+      "loss": 0.0002,
+      "step": 221
+    },
+    {
+      "epoch": 0.28416,
+      "grad_norm": 0.00037682682159356773,
+      "learning_rate": 3.96152234553272e-05,
+      "loss": 0.0002,
+      "step": 222
+    },
+    {
+      "epoch": 0.28544,
+      "grad_norm": 0.0005944096483290195,
+      "learning_rate": 3.961009466421379e-05,
+      "loss": 0.0002,
+      "step": 223
+    },
+    {
+      "epoch": 0.28672,
+      "grad_norm": 0.0008876877254806459,
+      "learning_rate": 3.9604932291718396e-05,
+      "loss": 0.0002,
+      "step": 224
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 0.0004229796468280256,
+      "learning_rate": 3.959973634768534e-05,
+      "loss": 0.0002,
+      "step": 225
+    },
+    {
+      "epoch": 0.28928,
+      "grad_norm": 0.0004196965601295233,
+      "learning_rate": 3.959450684202296e-05,
+      "loss": 0.0002,
+      "step": 226
+    },
+    {
+      "epoch": 0.29056,
+      "grad_norm": 0.0007056231843307614,
+      "learning_rate": 3.958924378470363e-05,
+      "loss": 0.0002,
+      "step": 227
+    },
+    {
+      "epoch": 0.29184,
+      "grad_norm": 0.0003414407547097653,
+      "learning_rate": 3.958394718576366e-05,
+      "loss": 0.0002,
+      "step": 228
+    },
+    {
+      "epoch": 0.29312,
+      "grad_norm": 0.00045374143519438803,
+      "learning_rate": 3.957861705530335e-05,
+      "loss": 0.0002,
+      "step": 229
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 0.00032401952194049954,
+      "learning_rate": 3.9573253403486944e-05,
+      "loss": 0.0002,
+      "step": 230
+    },
+    {
+      "epoch": 0.29568,
+      "grad_norm": 0.00038617721293121576,
+      "learning_rate": 3.956785624054259e-05,
+      "loss": 0.0002,
+      "step": 231
+    },
+    {
+      "epoch": 0.29696,
+      "grad_norm": 0.0002595918485894799,
+      "learning_rate": 3.956242557676235e-05,
+      "loss": 0.0002,
+      "step": 232
+    },
+    {
+      "epoch": 0.29824,
+      "grad_norm": 0.0007822422194294631,
+      "learning_rate": 3.9556961422502176e-05,
+      "loss": 0.0002,
+      "step": 233
+    },
+    {
+      "epoch": 0.29952,
+      "grad_norm": 0.0009492220124229789,
+      "learning_rate": 3.955146378818188e-05,
+      "loss": 0.0002,
+      "step": 234
+    },
+    {
+      "epoch": 0.3008,
+      "grad_norm": 0.00015348431770689785,
+      "learning_rate": 3.9545932684285105e-05,
+      "loss": 0.0002,
+      "step": 235
+    },
+    {
+      "epoch": 0.30208,
+      "grad_norm": 0.0009721404640004039,
+      "learning_rate": 3.954036812135934e-05,
+      "loss": 0.0002,
+      "step": 236
+    },
+    {
+      "epoch": 0.30336,
+      "grad_norm": 0.0006310763419605792,
+      "learning_rate": 3.9534770110015876e-05,
+      "loss": 0.0002,
+      "step": 237
+    },
+    {
+      "epoch": 0.30464,
+      "grad_norm": 0.00045944383600726724,
+      "learning_rate": 3.9529138660929764e-05,
+      "loss": 0.0002,
+      "step": 238
+    },
+    {
+      "epoch": 0.30592,
+      "grad_norm": 0.0009684551623649895,
+      "learning_rate": 3.9523473784839846e-05,
+      "loss": 0.0002,
+      "step": 239
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 0.0011421306990087032,
+      "learning_rate": 3.95177754925487e-05,
+      "loss": 0.0002,
+      "step": 240
+    },
+    {
+      "epoch": 0.3072,
+      "eval_loss": 1.1997703313827515,
+      "eval_runtime": 42.6282,
+      "eval_samples_per_second": 11.776,
+      "eval_steps_per_second": 1.478,
+      "step": 240
+    },
+    {
+      "epoch": 0.30848,
+      "grad_norm": 0.0007120371446944773,
+      "learning_rate": 3.9512043794922614e-05,
+      "loss": 0.0002,
+      "step": 241
+    },
+    {
+      "epoch": 0.30976,
+      "grad_norm": 0.00023797668109182268,
+      "learning_rate": 3.9506278702891594e-05,
+      "loss": 0.0002,
+      "step": 242
+    },
+    {
+      "epoch": 0.31104,
+      "grad_norm": 0.0002528477052692324,
+      "learning_rate": 3.9500480227449316e-05,
+      "loss": 0.0002,
+      "step": 243
+    },
+    {
+      "epoch": 0.31232,
+      "grad_norm": 0.0001783139305189252,
+      "learning_rate": 3.949464837965313e-05,
+      "loss": 0.0002,
+      "step": 244
+    },
+    {
+      "epoch": 0.3136,
+      "grad_norm": 0.0013822782784700394,
+      "learning_rate": 3.9488783170624e-05,
+      "loss": 0.0002,
+      "step": 245
+    },
+    {
+      "epoch": 0.31488,
+      "grad_norm": 0.0027772309258580208,
+      "learning_rate": 3.948288461154654e-05,
+      "loss": 0.0002,
+      "step": 246
+    },
+    {
+      "epoch": 0.31616,
+      "grad_norm": 0.0026890826411545277,
+      "learning_rate": 3.947695271366894e-05,
+      "loss": 0.0002,
+      "step": 247
+    },
+    {
+      "epoch": 0.31744,
+      "grad_norm": 0.0009026412153616548,
+      "learning_rate": 3.947098748830296e-05,
+      "loss": 0.0002,
+      "step": 248
+    },
+    {
+      "epoch": 0.31872,
+      "grad_norm": 0.0014957863604649901,
+      "learning_rate": 3.946498894682394e-05,
+      "loss": 0.0002,
+      "step": 249
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0031202055979520082,
+      "learning_rate": 3.945895710067072e-05,
+      "loss": 0.0002,
+      "step": 250
+    },
+    {
+      "epoch": 0.32128,
+      "grad_norm": 0.0037764040753245354,
+      "learning_rate": 3.9452891961345666e-05,
+      "loss": 0.0002,
+      "step": 251
+    },
+    {
+      "epoch": 0.32256,
+      "grad_norm": 0.003359750611707568,
+      "learning_rate": 3.944679354041464e-05,
+      "loss": 0.0002,
+      "step": 252
+    },
+    {
+      "epoch": 0.32384,
+      "grad_norm": 0.002477930160239339,
+      "learning_rate": 3.944066184950695e-05,
+      "loss": 0.0002,
+      "step": 253
+    },
+    {
+      "epoch": 0.32512,
+      "grad_norm": 0.002216948429122567,
+      "learning_rate": 3.9434496900315364e-05,
+      "loss": 0.0002,
+      "step": 254
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 0.0025624523404985666,
+      "learning_rate": 3.942829870459606e-05,
+      "loss": 0.0002,
+      "step": 255
+    },
+    {
+      "epoch": 0.32768,
+      "grad_norm": 0.002335009863600135,
+      "learning_rate": 3.9422067274168634e-05,
+      "loss": 0.0002,
+      "step": 256
+    },
+    {
+      "epoch": 0.32896,
+      "grad_norm": 0.0010357327992096543,
+      "learning_rate": 3.941580262091603e-05,
+      "loss": 0.0002,
+      "step": 257
+    },
+    {
+      "epoch": 0.33024,
+      "grad_norm": 0.0002824495895765722,
+      "learning_rate": 3.940950475678459e-05,
+      "loss": 0.0002,
+      "step": 258
+    },
+    {
+      "epoch": 0.33152,
+      "grad_norm": 0.000942192564252764,
+      "learning_rate": 3.940317369378391e-05,
+      "loss": 0.0002,
+      "step": 259
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 0.0009988127276301384,
+      "learning_rate": 3.939680944398699e-05,
+      "loss": 0.0002,
+      "step": 260
+    },
+    {
+      "epoch": 0.33408,
+      "grad_norm": 0.0008213394903577864,
+      "learning_rate": 3.9390412019530046e-05,
+      "loss": 0.0002,
+      "step": 261
+    },
+    {
+      "epoch": 0.33536,
+      "grad_norm": 0.00039377162465825677,
+      "learning_rate": 3.938398143261258e-05,
+      "loss": 0.0002,
+      "step": 262
+    },
+    {
+      "epoch": 0.33664,
+      "grad_norm": 0.0008121069986373186,
+      "learning_rate": 3.937751769549733e-05,
+      "loss": 0.0002,
+      "step": 263
+    },
+    {
+      "epoch": 0.33792,
+      "grad_norm": 0.002108998131006956,
+      "learning_rate": 3.937102082051026e-05,
+      "loss": 0.0002,
+      "step": 264
+    },
+    {
+      "epoch": 0.3392,
+      "grad_norm": 0.003128256183117628,
+      "learning_rate": 3.9364490820040506e-05,
+      "loss": 0.0002,
+      "step": 265
+    },
+    {
+      "epoch": 0.34048,
+      "grad_norm": 0.003833349095657468,
+      "learning_rate": 3.935792770654039e-05,
+      "loss": 0.0002,
+      "step": 266
+    },
+    {
+      "epoch": 0.34176,
+      "grad_norm": 0.005173802375793457,
+      "learning_rate": 3.935133149252536e-05,
+      "loss": 0.0002,
+      "step": 267
+    },
+    {
+      "epoch": 0.34304,
+      "grad_norm": 0.008912354707717896,
+      "learning_rate": 3.934470219057401e-05,
+      "loss": 0.0002,
+      "step": 268
+    },
+    {
+      "epoch": 0.34432,
+      "grad_norm": 0.011784413829445839,
+      "learning_rate": 3.933803981332801e-05,
+      "loss": 0.0002,
+      "step": 269
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 0.0093743447214365,
+      "learning_rate": 3.933134437349211e-05,
+      "loss": 0.0002,
+      "step": 270
+    },
+    {
+      "epoch": 0.3456,
+      "eval_loss": 1.220482349395752,
+      "eval_runtime": 43.5388,
+      "eval_samples_per_second": 11.53,
+      "eval_steps_per_second": 1.447,
+      "step": 270
+    },
+    {
+      "epoch": 0.34688,
+      "grad_norm": 0.003992204554378986,
+      "learning_rate": 3.9324615883834103e-05,
+      "loss": 0.0002,
+      "step": 271
+    },
+    {
+      "epoch": 0.34816,
+      "grad_norm": 0.002077931072562933,
+      "learning_rate": 3.9317854357184815e-05,
+      "loss": 0.0002,
+      "step": 272
+    },
+    {
+      "epoch": 0.34944,
+      "grad_norm": 0.006616830825805664,
+      "learning_rate": 3.9311059806438065e-05,
+      "loss": 0.0002,
+      "step": 273
+    },
+    {
+      "epoch": 0.35072,
+      "grad_norm": 0.0070667387917637825,
+      "learning_rate": 3.930423224455065e-05,
+      "loss": 0.0002,
+      "step": 274
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 0.0025246667210012674,
+      "learning_rate": 3.929737168454232e-05,
+      "loss": 0.0002,
+      "step": 275
+    },
+    {
+      "epoch": 0.35328,
+      "grad_norm": 0.003516591852530837,
+      "learning_rate": 3.929047813949575e-05,
+      "loss": 0.0002,
+      "step": 276
+    },
+    {
+      "epoch": 0.35456,
+      "grad_norm": 0.005735184997320175,
+      "learning_rate": 3.928355162255649e-05,
+      "loss": 0.0002,
+      "step": 277
+    },
+    {
+      "epoch": 0.35584,
+      "grad_norm": 0.003947160672396421,
+      "learning_rate": 3.9276592146933004e-05,
+      "loss": 0.0002,
+      "step": 278
+    },
+    {
+      "epoch": 0.35712,
+      "grad_norm": 0.0008040797547437251,
+      "learning_rate": 3.92695997258966e-05,
+      "loss": 0.0002,
+      "step": 279
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 0.0020477804355323315,
+      "learning_rate": 3.9262574372781383e-05,
+      "loss": 0.0002,
+      "step": 280
+    },
+    {
+      "epoch": 0.35968,
+      "grad_norm": 0.003859796328470111,
+      "learning_rate": 3.9255516100984285e-05,
+      "loss": 0.0002,
+      "step": 281
+    },
+    {
+      "epoch": 0.36096,
+      "grad_norm": 0.003972633741796017,
+      "learning_rate": 3.924842492396499e-05,
+      "loss": 0.0002,
+      "step": 282
+    },
+    {
+      "epoch": 0.36224,
+      "grad_norm": 0.0018808410968631506,
+      "learning_rate": 3.924130085524596e-05,
+      "loss": 0.0002,
+      "step": 283
+    },
+    {
+      "epoch": 0.36352,
+      "grad_norm": 0.0007452194113284349,
+      "learning_rate": 3.923414390841235e-05,
+      "loss": 0.0002,
+      "step": 284
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 0.0026158560067415237,
+      "learning_rate": 3.922695409711202e-05,
+      "loss": 0.0002,
+      "step": 285
+    },
+    {
+      "epoch": 0.36608,
+      "grad_norm": 0.003892722772434354,
+      "learning_rate": 3.921973143505552e-05,
+      "loss": 0.0002,
+      "step": 286
+    },
+    {
+      "epoch": 0.36736,
+      "grad_norm": 0.004123721271753311,
+      "learning_rate": 3.921247593601601e-05,
+      "loss": 0.0002,
+      "step": 287
+    },
+    {
+      "epoch": 0.36864,
+      "grad_norm": 0.0037682715337723494,
+      "learning_rate": 3.92051876138293e-05,
+      "loss": 0.0002,
+      "step": 288
+    },
+    {
+      "epoch": 0.36992,
+      "grad_norm": 0.002960368525236845,
+      "learning_rate": 3.919786648239377e-05,
+      "loss": 0.0002,
+      "step": 289
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 0.0015276469057425857,
+      "learning_rate": 3.9190512555670374e-05,
+      "loss": 0.0002,
+      "step": 290
+    },
+    {
+      "epoch": 0.37248,
+      "grad_norm": 0.00017276291328016669,
+      "learning_rate": 3.918312584768261e-05,
+      "loss": 0.0002,
+      "step": 291
+    },
+    {
+      "epoch": 0.37376,
+      "grad_norm": 0.0009839094709604979,
+      "learning_rate": 3.917570637251647e-05,
+      "loss": 0.0002,
+      "step": 292
+    },
+    {
+      "epoch": 0.37504,
+      "grad_norm": 0.001523509039543569,
+      "learning_rate": 3.916825414432046e-05,
+      "loss": 0.0002,
+      "step": 293
+    },
+    {
+      "epoch": 0.37632,
+      "grad_norm": 0.0014921202091500163,
+      "learning_rate": 3.916076917730552e-05,
+      "loss": 0.0002,
+      "step": 294
+    },
+    {
+      "epoch": 0.3776,
+      "grad_norm": 0.001001936150714755,
+      "learning_rate": 3.915325148574502e-05,
+      "loss": 0.0002,
+      "step": 295
+    },
+    {
+      "epoch": 0.37888,
+      "grad_norm": 0.0005724078509956598,
+      "learning_rate": 3.914570108397474e-05,
+      "loss": 0.0002,
+      "step": 296
+    },
+    {
+      "epoch": 0.38016,
+      "grad_norm": 0.001378412009216845,
+      "learning_rate": 3.9138117986392856e-05,
+      "loss": 0.0002,
+      "step": 297
+    },
+    {
+      "epoch": 0.38144,
+      "grad_norm": 0.0024734761100262403,
+      "learning_rate": 3.913050220745986e-05,
+      "loss": 0.0002,
+      "step": 298
+    },
+    {
+      "epoch": 0.38272,
+      "grad_norm": 0.0031675021164119244,
+      "learning_rate": 3.9122853761698595e-05,
+      "loss": 0.0002,
+      "step": 299
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.0029399548657238483,
+      "learning_rate": 3.911517266369416e-05,
+      "loss": 0.0002,
+      "step": 300
+    },
+    {
+      "epoch": 0.384,
+      "eval_loss": 1.1859415769577026,
+      "eval_runtime": 42.3844,
+      "eval_samples_per_second": 11.844,
+      "eval_steps_per_second": 1.486,
+      "step": 300
+    },
+    {
+      "epoch": 0.38528,
+      "grad_norm": 0.0010896556777879596,
+      "learning_rate": 3.910745892809396e-05,
+      "loss": 0.0002,
+      "step": 301
+    },
+    {
+      "epoch": 0.38656,
+      "grad_norm": 0.0017165422905236483,
+      "learning_rate": 3.9099712569607614e-05,
+      "loss": 0.0002,
+      "step": 302
+    },
+    {
+      "epoch": 0.38784,
+      "grad_norm": 0.004119697492569685,
+      "learning_rate": 3.9091933603006964e-05,
+      "loss": 0.0002,
+      "step": 303
+    },
+    {
+      "epoch": 0.38912,
+      "grad_norm": 0.006195970810949802,
+      "learning_rate": 3.9084122043126026e-05,
+      "loss": 0.0002,
+      "step": 304
+    },
+    {
+      "epoch": 0.3904,
+      "grad_norm": 0.007094519678503275,
+      "learning_rate": 3.9076277904860965e-05,
+      "loss": 0.0002,
+      "step": 305
+    },
+    {
+      "epoch": 0.39168,
+      "grad_norm": 0.005756770726293325,
+      "learning_rate": 3.906840120317008e-05,
+      "loss": 0.0002,
+      "step": 306
+    },
+    {
+      "epoch": 0.39296,
+      "grad_norm": 0.0035659433342516422,
+      "learning_rate": 3.906049195307376e-05,
+      "loss": 0.0002,
+      "step": 307
+    },
+    {
+      "epoch": 0.39424,
+      "grad_norm": 0.0014575119130313396,
+      "learning_rate": 3.905255016965447e-05,
+      "loss": 0.0002,
+      "step": 308
+    },
+    {
+      "epoch": 0.39552,
+      "grad_norm": 0.0005094299558550119,
+      "learning_rate": 3.9044575868056705e-05,
+      "loss": 0.0002,
+      "step": 309
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 0.002270213095471263,
+      "learning_rate": 3.903656906348698e-05,
+      "loss": 0.0002,
+      "step": 310
+    },
+    {
+      "epoch": 0.39808,
+      "grad_norm": 0.003336275927722454,
+      "learning_rate": 3.902852977121378e-05,
+      "loss": 0.0002,
+      "step": 311
+    },
+    {
+      "epoch": 0.39936,
+      "grad_norm": 0.002990037202835083,
+      "learning_rate": 3.902045800656755e-05,
+      "loss": 0.0002,
+      "step": 312
+    },
+    {
+      "epoch": 0.40064,
+      "grad_norm": 0.002952362410724163,
+      "learning_rate": 3.901235378494065e-05,
+      "loss": 0.0002,
+      "step": 313
+    },
+    {
+      "epoch": 0.40192,
+      "grad_norm": 0.0025355329271405935,
+      "learning_rate": 3.9004217121787354e-05,
+      "loss": 0.0002,
+      "step": 314
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 0.002028323709964752,
+      "learning_rate": 3.899604803262377e-05,
+      "loss": 0.0002,
+      "step": 315
+    },
+    {
+      "epoch": 0.40448,
+      "grad_norm": 0.001996732084080577,
+      "learning_rate": 3.898784653302786e-05,
+      "loss": 0.0002,
+      "step": 316
+    },
+    {
+      "epoch": 0.40576,
+      "grad_norm": 0.0018237540498375893,
+      "learning_rate": 3.897961263863938e-05,
+      "loss": 0.0002,
+      "step": 317
+    },
+    {
+      "epoch": 0.40704,
+      "grad_norm": 0.002562148030847311,
+      "learning_rate": 3.897134636515987e-05,
+      "loss": 0.0002,
+      "step": 318
+    },
+    {
+      "epoch": 0.40832,
+      "grad_norm": 0.00395054230466485,
+      "learning_rate": 3.896304772835263e-05,
+      "loss": 0.0002,
+      "step": 319
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 0.004901579115539789,
+      "learning_rate": 3.895471674404264e-05,
+      "loss": 0.0002,
+      "step": 320
+    },
+    {
+      "epoch": 0.41088,
+      "grad_norm": 0.0055976551957428455,
+      "learning_rate": 3.894635342811657e-05,
+      "loss": 0.0002,
+      "step": 321
+    },
+    {
+      "epoch": 0.41216,
+      "grad_norm": 0.006354920100420713,
+      "learning_rate": 3.893795779652278e-05,
+      "loss": 0.0002,
+      "step": 322
+    },
+    {
+      "epoch": 0.41344,
+      "grad_norm": 0.008057119324803352,
+      "learning_rate": 3.892952986527122e-05,
+      "loss": 0.0002,
+      "step": 323
+    },
+    {
+      "epoch": 0.41472,
+      "grad_norm": 0.00950048491358757,
+      "learning_rate": 3.892106965043344e-05,
+      "loss": 0.0002,
+      "step": 324
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 0.009841698221862316,
+      "learning_rate": 3.891257716814256e-05,
+      "loss": 0.0002,
+      "step": 325
+    },
+    {
+      "epoch": 0.41728,
+      "grad_norm": 0.008819937705993652,
+      "learning_rate": 3.890405243459322e-05,
+      "loss": 0.0002,
+      "step": 326
+    },
+    {
+      "epoch": 0.41856,
+      "grad_norm": 0.004502292722463608,
+      "learning_rate": 3.889549546604159e-05,
+      "loss": 0.0002,
+      "step": 327
+    },
+    {
+      "epoch": 0.41984,
+      "grad_norm": 0.001475517638027668,
+      "learning_rate": 3.888690627880527e-05,
+      "loss": 0.0002,
+      "step": 328
+    },
+    {
+      "epoch": 0.42112,
+      "grad_norm": 0.004811195190995932,
+      "learning_rate": 3.8878284889263326e-05,
+      "loss": 0.0002,
+      "step": 329
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 0.003931612242013216,
+      "learning_rate": 3.886963131385623e-05,
+      "loss": 0.0002,
+      "step": 330
+    },
+    {
+      "epoch": 0.4224,
+      "eval_loss": 1.1920558214187622,
+      "eval_runtime": 43.3287,
+      "eval_samples_per_second": 11.586,
+      "eval_steps_per_second": 1.454,
+      "step": 330
+    },
+    {
+      "epoch": 0.42368,
+      "grad_norm": 0.00033131142845377326,
+      "learning_rate": 3.886094556908581e-05,
+      "loss": 0.0002,
+      "step": 331
+    },
+    {
+      "epoch": 0.42496,
+      "grad_norm": 0.003252562368288636,
+      "learning_rate": 3.8852227671515274e-05,
+      "loss": 0.0002,
+      "step": 332
+    },
+    {
+      "epoch": 0.42624,
+      "grad_norm": 0.005851993802934885,
+      "learning_rate": 3.8843477637769106e-05,
+      "loss": 0.0002,
+      "step": 333
+    },
+    {
+      "epoch": 0.42752,
+      "grad_norm": 0.00618679728358984,
+      "learning_rate": 3.8834695484533104e-05,
+      "loss": 0.0002,
+      "step": 334
+    },
+    {
+      "epoch": 0.4288,
+      "grad_norm": 0.0034405761398375034,
+      "learning_rate": 3.8825881228554295e-05,
+      "loss": 0.0002,
+      "step": 335
+    },
+    {
+      "epoch": 0.43008,
+      "grad_norm": 0.0008351071155630052,
+      "learning_rate": 3.8817034886640924e-05,
+      "loss": 0.0002,
+      "step": 336
+    },
+    {
+      "epoch": 0.43136,
+      "grad_norm": 0.0040811290964484215,
+      "learning_rate": 3.880815647566244e-05,
+      "loss": 0.0002,
+      "step": 337
+    },
+    {
+      "epoch": 0.43264,
+      "grad_norm": 0.0059448895044624805,
+      "learning_rate": 3.879924601254943e-05,
+      "loss": 0.0002,
+      "step": 338
+    },
+    {
+      "epoch": 0.43392,
+      "grad_norm": 0.006282978691160679,
+      "learning_rate": 3.87903035142936e-05,
+      "loss": 0.0002,
+      "step": 339
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 0.00391614343971014,
+      "learning_rate": 3.878132899794776e-05,
+      "loss": 0.0002,
+      "step": 340
+    },
+    {
+      "epoch": 0.43648,
+      "grad_norm": 0.0001543579564895481,
+      "learning_rate": 3.877232248062576e-05,
+      "loss": 0.0002,
+      "step": 341
+    },
+    {
+      "epoch": 0.43776,
+      "grad_norm": 0.0030169414822012186,
+      "learning_rate": 3.876328397950249e-05,
+      "loss": 0.0002,
+      "step": 342
+    },
+    {
+      "epoch": 0.43904,
+      "grad_norm": 0.004486779682338238,
+      "learning_rate": 3.8754213511813815e-05,
+      "loss": 0.0002,
+      "step": 343
+    },
+    {
+      "epoch": 0.44032,
+      "grad_norm": 0.004293914418667555,
+      "learning_rate": 3.874511109485658e-05,
+      "loss": 0.0002,
+      "step": 344
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 0.0029261987656354904,
+      "learning_rate": 3.873597674598853e-05,
+      "loss": 0.0002,
+      "step": 345
+    },
+    {
+      "epoch": 0.44288,
+      "grad_norm": 0.0002709278487600386,
+      "learning_rate": 3.872681048262832e-05,
+      "loss": 0.0002,
+      "step": 346
+    },
+    {
+      "epoch": 0.44416,
+      "grad_norm": 0.0035741126630455256,
+      "learning_rate": 3.871761232225544e-05,
+      "loss": 0.0002,
+      "step": 347
+    },
+    {
+      "epoch": 0.44544,
+      "grad_norm": 0.005375510081648827,
+      "learning_rate": 3.870838228241025e-05,
+      "loss": 0.0002,
+      "step": 348
+    },
+    {
+      "epoch": 0.44672,
+      "grad_norm": 0.004629251081496477,
+      "learning_rate": 3.869912038069387e-05,
+      "loss": 0.0002,
+      "step": 349
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 0.0019839161541312933,
+      "learning_rate": 3.868982663476817e-05,
+      "loss": 0.0002,
+      "step": 350
+    },
+    {
+      "epoch": 0.44928,
+      "grad_norm": 0.0009309860761277378,
+      "learning_rate": 3.868050106235578e-05,
+      "loss": 0.0002,
+      "step": 351
+    },
+    {
+      "epoch": 0.45056,
+      "grad_norm": 0.00313609279692173,
+      "learning_rate": 3.867114368123998e-05,
+      "loss": 0.0002,
+      "step": 352
+    },
+    {
+      "epoch": 0.45184,
+      "grad_norm": 0.0037451910320669413,
+      "learning_rate": 3.866175450926474e-05,
+      "loss": 0.0002,
+      "step": 353
+    },
+    {
+      "epoch": 0.45312,
+      "grad_norm": 0.002794757019728422,
+      "learning_rate": 3.865233356433464e-05,
+      "loss": 0.0002,
+      "step": 354
+    },
+    {
+      "epoch": 0.4544,
+      "grad_norm": 0.0012516654096543789,
+      "learning_rate": 3.864288086441485e-05,
+      "loss": 0.0002,
+      "step": 355
+    },
+    {
+      "epoch": 0.45568,
+      "grad_norm": 0.00018904400349128991,
+      "learning_rate": 3.8633396427531096e-05,
+      "loss": 0.0002,
+      "step": 356
+    },
+    {
+      "epoch": 0.45696,
+      "grad_norm": 0.0005176289123483002,
+      "learning_rate": 3.862388027176962e-05,
+      "loss": 0.0002,
+      "step": 357
+    },
+    {
+      "epoch": 0.45824,
+      "grad_norm": 0.00018633867148309946,
+      "learning_rate": 3.8614332415277155e-05,
+      "loss": 0.0002,
+      "step": 358
+    },
+    {
+      "epoch": 0.45952,
+      "grad_norm": 0.0004992458852939308,
+      "learning_rate": 3.8604752876260885e-05,
+      "loss": 0.0002,
+      "step": 359
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 0.0002147257764590904,
+      "learning_rate": 3.859514167298841e-05,
+      "loss": 0.0002,
+      "step": 360
+    },
+    {
+      "epoch": 0.4608,
+      "eval_loss": 1.1968976259231567,
+      "eval_runtime": 43.4271,
+      "eval_samples_per_second": 11.56,
+      "eval_steps_per_second": 1.451,
+      "step": 360
+    },
+    {
+      "epoch": 0.46208,
+      "grad_norm": 0.0015842154389247298,
+      "learning_rate": 3.8585498823787707e-05,
+      "loss": 0.0002,
+      "step": 361
+    },
+    {
+      "epoch": 0.46336,
+      "grad_norm": 0.003454265184700489,
+      "learning_rate": 3.8575824347047115e-05,
+      "loss": 0.0002,
+      "step": 362
+    },
+    {
+      "epoch": 0.46464,
+      "grad_norm": 0.004604188725352287,
+      "learning_rate": 3.856611826121526e-05,
+      "loss": 0.0002,
+      "step": 363
+    },
+    {
+      "epoch": 0.46592,
+      "grad_norm": 0.0045526353642344475,
+      "learning_rate": 3.8556380584801075e-05,
+      "loss": 0.0002,
+      "step": 364
+    },
+    {
+      "epoch": 0.4672,
+      "grad_norm": 0.0031676136422902346,
+      "learning_rate": 3.85466113363737e-05,
+      "loss": 0.0002,
+      "step": 365
+    },
+    {
+      "epoch": 0.46848,
+      "grad_norm": 0.0005085321608930826,
+      "learning_rate": 3.853681053456252e-05,
+      "loss": 0.0002,
+      "step": 366
+    },
+    {
+      "epoch": 0.46976,
+      "grad_norm": 0.0029830520506948233,
+      "learning_rate": 3.852697819805705e-05,
+      "loss": 0.0002,
+      "step": 367
+    },
+    {
+      "epoch": 0.47104,
+      "grad_norm": 0.0048976195976138115,
+      "learning_rate": 3.851711434560697e-05,
+      "loss": 0.0002,
+      "step": 368
+    },
+    {
+      "epoch": 0.47232,
+      "grad_norm": 0.005438893102109432,
+      "learning_rate": 3.850721899602204e-05,
+      "loss": 0.0002,
+      "step": 369
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 0.0041366140358150005,
+      "learning_rate": 3.84972921681721e-05,
+      "loss": 0.0002,
+      "step": 370
+    },
+    {
+      "epoch": 0.47488,
+      "grad_norm": 0.0002860951062757522,
+      "learning_rate": 3.8487333880987e-05,
+      "loss": 0.0002,
+      "step": 371
+    },
+    {
+      "epoch": 0.47616,
+      "grad_norm": 0.004208084661513567,
+      "learning_rate": 3.84773441534566e-05,
+      "loss": 0.0002,
+      "step": 372
+    },
+    {
+      "epoch": 0.47744,
+      "grad_norm": 0.006803308613598347,
+      "learning_rate": 3.8467323004630695e-05,
+      "loss": 0.0002,
+      "step": 373
+    },
+    {
+      "epoch": 0.47872,
+      "grad_norm": 0.006709063891321421,
+      "learning_rate": 3.8457270453619e-05,
+      "loss": 0.0002,
+      "step": 374
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.005010607186704874,
+      "learning_rate": 3.844718651959115e-05,
+      "loss": 0.0002,
+      "step": 375
+    },
+    {
+      "epoch": 0.48128,
+      "grad_norm": 0.002244663191959262,
+      "learning_rate": 3.843707122177656e-05,
+      "loss": 0.0002,
+      "step": 376
+    },
+    {
+      "epoch": 0.48256,
+      "grad_norm": 0.001649985439144075,
+      "learning_rate": 3.842692457946452e-05,
+      "loss": 0.0002,
+      "step": 377
+    },
+    {
+      "epoch": 0.48384,
+      "grad_norm": 0.004389249719679356,
+      "learning_rate": 3.841674661200403e-05,
+      "loss": 0.0002,
+      "step": 378
+    },
+    {
+      "epoch": 0.48512,
+      "grad_norm": 0.004841000307351351,
+      "learning_rate": 3.840653733880388e-05,
+      "loss": 0.0002,
+      "step": 379
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 0.003227783599868417,
+      "learning_rate": 3.839629677933253e-05,
+      "loss": 0.0002,
+      "step": 380
+    },
+    {
+      "epoch": 0.48768,
+      "grad_norm": 0.00019390507077332586,
+      "learning_rate": 3.8386024953118105e-05,
+      "loss": 0.0002,
+      "step": 381
+    },
+    {
+      "epoch": 0.48896,
+      "grad_norm": 0.002688514068722725,
+      "learning_rate": 3.837572187974835e-05,
+      "loss": 0.0002,
+      "step": 382
+    },
+    {
+      "epoch": 0.49024,
+      "grad_norm": 0.003430665237829089,
+      "learning_rate": 3.83653875788706e-05,
+      "loss": 0.0002,
+      "step": 383
+    },
+    {
+      "epoch": 0.49152,
+      "grad_norm": 0.0023852745071053505,
+      "learning_rate": 3.8355022070191745e-05,
+      "loss": 0.0002,
+      "step": 384
+    },
+    {
+      "epoch": 0.4928,
+      "grad_norm": 0.00032828684197738767,
+      "learning_rate": 3.834462537347817e-05,
+      "loss": 0.0002,
+      "step": 385
+    },
+    {
+      "epoch": 0.49408,
+      "grad_norm": 0.001501849154010415,
+      "learning_rate": 3.833419750855577e-05,
+      "loss": 0.0002,
+      "step": 386
+    },
+    {
+      "epoch": 0.49536,
+      "grad_norm": 0.002512734616175294,
+      "learning_rate": 3.832373849530982e-05,
+      "loss": 0.0002,
+      "step": 387
+    },
+    {
+      "epoch": 0.49664,
+      "grad_norm": 0.0017667177598923445,
+      "learning_rate": 3.831324835368505e-05,
+      "loss": 0.0002,
+      "step": 388
+    },
+    {
+      "epoch": 0.49792,
+      "grad_norm": 0.00017293778364546597,
+      "learning_rate": 3.83027271036855e-05,
+      "loss": 0.0002,
+      "step": 389
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 0.0009446240146644413,
+      "learning_rate": 3.829217476537457e-05,
+      "loss": 0.0002,
+      "step": 390
+    },
+    {
+      "epoch": 0.4992,
+      "eval_loss": 1.193045735359192,
+      "eval_runtime": 43.3135,
+      "eval_samples_per_second": 11.59,
+      "eval_steps_per_second": 1.455,
+      "step": 390
+    },
+    {
+      "epoch": 0.50048,
+      "grad_norm": 0.0007563307299278677,
+      "learning_rate": 3.828159135887493e-05,
+      "loss": 0.0002,
+      "step": 391
+    },
+    {
+      "epoch": 0.50176,
+      "grad_norm": 0.000485950440634042,
+      "learning_rate": 3.827097690436848e-05,
+      "loss": 0.0002,
+      "step": 392
+    },
+    {
+      "epoch": 0.50304,
+      "grad_norm": 0.0014980818377807736,
+      "learning_rate": 3.826033142209636e-05,
+      "loss": 0.0002,
+      "step": 393
+    },
+    {
+      "epoch": 0.50432,
+      "grad_norm": 0.002262465190142393,
+      "learning_rate": 3.824965493235885e-05,
+      "loss": 0.0002,
+      "step": 394
+    },
+    {
+      "epoch": 0.5056,
+      "grad_norm": 0.0031683961860835552,
+      "learning_rate": 3.823894745551536e-05,
+      "loss": 0.0002,
+      "step": 395
+    },
+    {
+      "epoch": 0.50688,
+      "grad_norm": 0.0036339478101581335,
+      "learning_rate": 3.8228209011984404e-05,
+      "loss": 0.0002,
+      "step": 396
+    },
+    {
+      "epoch": 0.50816,
+      "grad_norm": 0.003546744817867875,
+      "learning_rate": 3.821743962224355e-05,
+      "loss": 0.0002,
+      "step": 397
+    },
+    {
+      "epoch": 0.50944,
+      "grad_norm": 0.002480078022927046,
+      "learning_rate": 3.820663930682936e-05,
+      "loss": 0.0002,
+      "step": 398
+    },
+    {
+      "epoch": 0.51072,
+      "grad_norm": 0.0004905156674794853,
+      "learning_rate": 3.819580808633739e-05,
+      "loss": 0.0002,
+      "step": 399
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 0.0015791412442922592,
+      "learning_rate": 3.818494598142211e-05,
+      "loss": 0.0002,
+      "step": 400
+    },
+    {
+      "epoch": 0.51328,
+      "grad_norm": 0.0024596957955509424,
+      "learning_rate": 3.81740530127969e-05,
+      "loss": 0.0002,
+      "step": 401
+    },
+    {
+      "epoch": 0.51456,
+      "grad_norm": 0.002662543673068285,
+      "learning_rate": 3.8163129201234e-05,
+      "loss": 0.0002,
+      "step": 402
+    },
+    {
+      "epoch": 0.51584,
+      "grad_norm": 0.002260034205392003,
+      "learning_rate": 3.815217456756444e-05,
+      "loss": 0.0002,
+      "step": 403
+    },
+    {
+      "epoch": 0.51712,
+      "grad_norm": 0.0009084303746931255,
+      "learning_rate": 3.8141189132678045e-05,
+      "loss": 0.0002,
+      "step": 404
+    },
+    {
+      "epoch": 0.5184,
+      "grad_norm": 0.0014930436154827476,
+      "learning_rate": 3.813017291752338e-05,
+      "loss": 0.0002,
+      "step": 405
+    },
+    {
+      "epoch": 0.51968,
+      "grad_norm": 0.004289396107196808,
+      "learning_rate": 3.8119125943107694e-05,
+      "loss": 0.0002,
+      "step": 406
+    },
+    {
+      "epoch": 0.52096,
+      "grad_norm": 0.006647216156125069,
+      "learning_rate": 3.8108048230496896e-05,
+      "loss": 0.0002,
+      "step": 407
+    },
+    {
+      "epoch": 0.52224,
+      "grad_norm": 0.008235771209001541,
+      "learning_rate": 3.8096939800815514e-05,
+      "loss": 0.0002,
+      "step": 408
+    },
+    {
+      "epoch": 0.52352,
+      "grad_norm": 0.007829473353922367,
+      "learning_rate": 3.808580067524665e-05,
+      "loss": 0.0002,
+      "step": 409
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 0.005282750353217125,
+      "learning_rate": 3.807463087503194e-05,
+      "loss": 0.0002,
+      "step": 410
+    },
+    {
+      "epoch": 0.52608,
+      "grad_norm": 0.0014312955318018794,
+      "learning_rate": 3.806343042147151e-05,
+      "loss": 0.0002,
+      "step": 411
+    },
+    {
+      "epoch": 0.52736,
+      "grad_norm": 0.0028434335254132748,
+      "learning_rate": 3.805219933592396e-05,
+      "loss": 0.0002,
+      "step": 412
+    },
+    {
+      "epoch": 0.52864,
+      "grad_norm": 0.005742455366998911,
+      "learning_rate": 3.804093763980627e-05,
+      "loss": 0.0002,
+      "step": 413
+    },
+    {
+      "epoch": 0.52992,
+      "grad_norm": 0.006840975023806095,
+      "learning_rate": 3.8029645354593825e-05,
+      "loss": 0.0002,
+      "step": 414
+    },
+    {
+      "epoch": 0.5312,
+      "grad_norm": 0.004969259258359671,
+      "learning_rate": 3.8018322501820314e-05,
+      "loss": 0.0002,
+      "step": 415
+    },
+    {
+      "epoch": 0.53248,
+      "grad_norm": 0.0005858050426468253,
+      "learning_rate": 3.8006969103077736e-05,
+      "loss": 0.0002,
+      "step": 416
+    },
+    {
+      "epoch": 0.53376,
+      "grad_norm": 0.0043167779222130775,
+      "learning_rate": 3.799558518001635e-05,
+      "loss": 0.0002,
+      "step": 417
+    },
+    {
+      "epoch": 0.53504,
+      "grad_norm": 0.006647319067269564,
+      "learning_rate": 3.7984170754344584e-05,
+      "loss": 0.0002,
+      "step": 418
+    },
+    {
+      "epoch": 0.53632,
+      "grad_norm": 0.004266531206667423,
+      "learning_rate": 3.797272584782906e-05,
+      "loss": 0.0002,
+      "step": 419
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 0.00030707393307238817,
+      "learning_rate": 3.796125048229453e-05,
+      "loss": 0.0002,
+      "step": 420
+    },
+    {
+      "epoch": 0.5376,
+      "eval_loss": 1.2100694179534912,
+      "eval_runtime": 44.2412,
+      "eval_samples_per_second": 11.347,
+      "eval_steps_per_second": 1.424,
+      "step": 420
+    },
+    {
+      "epoch": 0.53888,
+      "grad_norm": 0.004020696971565485,
+      "learning_rate": 3.794974467962382e-05,
+      "loss": 0.0002,
+      "step": 421
+    },
+    {
+      "epoch": 0.54016,
+      "grad_norm": 0.005686239339411259,
+      "learning_rate": 3.793820846175778e-05,
+      "loss": 0.0002,
+      "step": 422
+    },
+    {
+      "epoch": 0.54144,
+      "grad_norm": 0.004134351387619972,
+      "learning_rate": 3.79266418506953e-05,
+      "loss": 0.0002,
+      "step": 423
+    },
+    {
+      "epoch": 0.54272,
+      "grad_norm": 0.0006439160206355155,
+      "learning_rate": 3.791504486849318e-05,
+      "loss": 0.0002,
+      "step": 424
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 0.0027023518923670053,
+      "learning_rate": 3.790341753726618e-05,
+      "loss": 0.0002,
+      "step": 425
+    },
+    {
+      "epoch": 0.54528,
+      "grad_norm": 0.0032724770717322826,
+      "learning_rate": 3.7891759879186924e-05,
+      "loss": 0.0002,
+      "step": 426
+    },
+    {
+      "epoch": 0.54656,
+      "grad_norm": 0.0012540655443444848,
+      "learning_rate": 3.788007191648583e-05,
+      "loss": 0.0002,
+      "step": 427
+    },
+    {
+      "epoch": 0.54784,
+      "grad_norm": 0.0018252013251185417,
+      "learning_rate": 3.786835367145116e-05,
+      "loss": 0.0002,
+      "step": 428
+    },
+    {
+      "epoch": 0.54912,
+      "grad_norm": 0.004245402291417122,
+      "learning_rate": 3.785660516642888e-05,
+      "loss": 0.0002,
+      "step": 429
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 0.004345291759818792,
+      "learning_rate": 3.7844826423822685e-05,
+      "loss": 0.0002,
+      "step": 430
+    },
+    {
+      "epoch": 0.55168,
+      "grad_norm": 0.0018071354134008288,
+      "learning_rate": 3.783301746609391e-05,
+      "loss": 0.0002,
+      "step": 431
+    },
+    {
+      "epoch": 0.55296,
+      "grad_norm": 0.00209009344689548,
+      "learning_rate": 3.782117831576153e-05,
+      "loss": 0.0002,
+      "step": 432
+    },
+    {
+      "epoch": 0.55424,
+      "grad_norm": 0.004698239266872406,
+      "learning_rate": 3.7809308995402104e-05,
+      "loss": 0.0002,
+      "step": 433
+    },
+    {
+      "epoch": 0.55552,
+      "grad_norm": 0.004285912495106459,
+      "learning_rate": 3.779740952764968e-05,
+      "loss": 0.0002,
+      "step": 434
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 0.001513406983576715,
+      "learning_rate": 3.778547993519582e-05,
+      "loss": 0.0002,
+      "step": 435
+    },
+    {
+      "epoch": 0.55808,
+      "grad_norm": 0.0010023791110143065,
+      "learning_rate": 3.777352024078955e-05,
+      "loss": 0.0002,
+      "step": 436
+    },
+    {
+      "epoch": 0.55936,
+      "grad_norm": 0.001528212451376021,
+      "learning_rate": 3.776153046723727e-05,
+      "loss": 0.0002,
+      "step": 437
+    },
+    {
+      "epoch": 0.56064,
+      "grad_norm": 0.0007561501697637141,
+      "learning_rate": 3.774951063740275e-05,
+      "loss": 0.0002,
+      "step": 438
+    },
+    {
+      "epoch": 0.56192,
+      "grad_norm": 0.0004925573011860251,
+      "learning_rate": 3.7737460774207094e-05,
+      "loss": 0.0002,
+      "step": 439
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 0.0008101520943455398,
+      "learning_rate": 3.772538090062863e-05,
+      "loss": 0.0002,
+      "step": 440
+    },
+    {
+      "epoch": 0.56448,
+      "grad_norm": 0.0004090173461008817,
+      "learning_rate": 3.771327103970297e-05,
+      "loss": 0.0002,
+      "step": 441
+    },
+    {
+      "epoch": 0.56576,
+      "grad_norm": 0.0022296542301774025,
+      "learning_rate": 3.770113121452287e-05,
+      "loss": 0.0002,
+      "step": 442
+    },
+    {
+      "epoch": 0.56704,
+      "grad_norm": 0.003427501767873764,
+      "learning_rate": 3.768896144823825e-05,
+      "loss": 0.0002,
+      "step": 443
+    },
+    {
+      "epoch": 0.56832,
+      "grad_norm": 0.003229739610105753,
+      "learning_rate": 3.76767617640561e-05,
+      "loss": 0.0002,
+      "step": 444
+    },
+    {
+      "epoch": 0.5696,
+      "grad_norm": 0.0014694056008011103,
+      "learning_rate": 3.76645321852405e-05,
+      "loss": 0.0002,
+      "step": 445
+    },
+    {
+      "epoch": 0.57088,
+      "grad_norm": 0.0008542861323803663,
+      "learning_rate": 3.7652272735112494e-05,
+      "loss": 0.0002,
+      "step": 446
+    },
+    {
+      "epoch": 0.57216,
+      "grad_norm": 0.0027981793973594904,
+      "learning_rate": 3.763998343705012e-05,
+      "loss": 0.0002,
+      "step": 447
+    },
+    {
+      "epoch": 0.57344,
+      "grad_norm": 0.003584227291867137,
+      "learning_rate": 3.7627664314488324e-05,
+      "loss": 0.0002,
+      "step": 448
+    },
+    {
+      "epoch": 0.57472,
+      "grad_norm": 0.002387993037700653,
+      "learning_rate": 3.761531539091892e-05,
+      "loss": 0.0002,
+      "step": 449
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.00032775092404335737,
+      "learning_rate": 3.760293668989055e-05,
+      "loss": 0.0002,
+      "step": 450
+    },
+    {
+      "epoch": 0.576,
+      "eval_loss": 1.2023624181747437,
+      "eval_runtime": 43.8775,
+      "eval_samples_per_second": 11.441,
+      "eval_steps_per_second": 1.436,
+      "step": 450
+    },
+    {
+      "epoch": 0.57728,
+      "grad_norm": 0.0023032776080071926,
+      "learning_rate": 3.7590528235008677e-05,
+      "loss": 0.0002,
+      "step": 451
+    },
+    {
+      "epoch": 0.57856,
+      "grad_norm": 0.0027386387810111046,
+      "learning_rate": 3.757809004993543e-05,
+      "loss": 0.0002,
+      "step": 452
+    },
+    {
+      "epoch": 0.57984,
+      "grad_norm": 0.001904963981360197,
+      "learning_rate": 3.75656221583897e-05,
+      "loss": 0.0002,
+      "step": 453
+    },
+    {
+      "epoch": 0.58112,
+      "grad_norm": 0.0004317044804338366,
+      "learning_rate": 3.7553124584146986e-05,
+      "loss": 0.0002,
+      "step": 454
+    },
+    {
+      "epoch": 0.5824,
+      "grad_norm": 0.0009737997897900641,
+      "learning_rate": 3.754059735103941e-05,
+      "loss": 0.0002,
+      "step": 455
+    },
+    {
+      "epoch": 0.58368,
+      "grad_norm": 0.0012084428453817964,
+      "learning_rate": 3.7528040482955655e-05,
+      "loss": 0.0002,
+      "step": 456
+    },
+    {
+      "epoch": 0.58496,
+      "grad_norm": 0.0005159890279173851,
+      "learning_rate": 3.751545400384089e-05,
+      "loss": 0.0002,
+      "step": 457
+    },
+    {
+      "epoch": 0.58624,
+      "grad_norm": 0.0005341696669347584,
+      "learning_rate": 3.7502837937696776e-05,
+      "loss": 0.0002,
+      "step": 458
+    },
+    {
+      "epoch": 0.58752,
+      "grad_norm": 0.0014383515808731318,
+      "learning_rate": 3.749019230858138e-05,
+      "loss": 0.0002,
+      "step": 459
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 0.002283843234181404,
+      "learning_rate": 3.747751714060915e-05,
+      "loss": 0.0002,
+      "step": 460
+    },
+    {
+      "epoch": 0.59008,
+      "grad_norm": 0.0026318333111703396,
+      "learning_rate": 3.746481245795087e-05,
+      "loss": 0.0002,
+      "step": 461
+    },
+    {
+      "epoch": 0.59136,
+      "grad_norm": 0.0022059311158955097,
+      "learning_rate": 3.745207828483359e-05,
+      "loss": 0.0002,
+      "step": 462
+    },
+    {
+      "epoch": 0.59264,
+      "grad_norm": 0.0014173684176057577,
+      "learning_rate": 3.743931464554061e-05,
+      "loss": 0.0002,
+      "step": 463
+    },
+    {
+      "epoch": 0.59392,
+      "grad_norm": 0.00021343116532079875,
+      "learning_rate": 3.742652156441142e-05,
+      "loss": 0.0002,
+      "step": 464
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 0.0014344080118462443,
+      "learning_rate": 3.741369906584164e-05,
+      "loss": 0.0002,
+      "step": 465
+    },
+    {
+      "epoch": 0.59648,
+      "grad_norm": 0.0022634316701442003,
+      "learning_rate": 3.7400847174283004e-05,
+      "loss": 0.0002,
+      "step": 466
+    },
+    {
+      "epoch": 0.59776,
+      "grad_norm": 0.0016528741689398885,
+      "learning_rate": 3.738796591424328e-05,
+      "loss": 0.0002,
+      "step": 467
+    },
+    {
+      "epoch": 0.59904,
+      "grad_norm": 0.00032261331216432154,
+      "learning_rate": 3.7375055310286266e-05,
+      "loss": 0.0002,
+      "step": 468
+    },
+    {
+      "epoch": 0.60032,
+      "grad_norm": 0.001875245594419539,
+      "learning_rate": 3.736211538703169e-05,
+      "loss": 0.0002,
+      "step": 469
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 0.0023898689541965723,
+      "learning_rate": 3.7349146169155194e-05,
+      "loss": 0.0002,
+      "step": 470
+    },
+    {
+      "epoch": 0.60288,
+      "grad_norm": 0.0013310150243341923,
+      "learning_rate": 3.7336147681388294e-05,
+      "loss": 0.0002,
+      "step": 471
+    },
+    {
+      "epoch": 0.60416,
+      "grad_norm": 0.00098335649818182,
+      "learning_rate": 3.732311994851832e-05,
+      "loss": 0.0002,
+      "step": 472
+    },
+    {
+      "epoch": 0.60544,
+      "grad_norm": 0.0034482080955058336,
+      "learning_rate": 3.731006299538837e-05,
+      "loss": 0.0002,
+      "step": 473
+    },
+    {
+      "epoch": 0.60672,
+      "grad_norm": 0.004604153335094452,
+      "learning_rate": 3.7296976846897266e-05,
+      "loss": 0.0002,
+      "step": 474
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 0.002799773123115301,
+      "learning_rate": 3.728386152799949e-05,
+      "loss": 0.0002,
+      "step": 475
+    },
+    {
+      "epoch": 0.60928,
+      "grad_norm": 0.0007603497942909598,
+      "learning_rate": 3.727071706370516e-05,
+      "loss": 0.0002,
+      "step": 476
+    },
+    {
+      "epoch": 0.61056,
+      "grad_norm": 0.0037390943616628647,
+      "learning_rate": 3.725754347907997e-05,
+      "loss": 0.0002,
+      "step": 477
+    },
+    {
+      "epoch": 0.61184,
+      "grad_norm": 0.004827678669244051,
+      "learning_rate": 3.724434079924515e-05,
+      "loss": 0.0002,
+      "step": 478
+    },
+    {
+      "epoch": 0.61312,
+      "grad_norm": 0.002908482449129224,
+      "learning_rate": 3.7231109049377414e-05,
+      "loss": 0.0002,
+      "step": 479
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 0.0007244603475555778,
+      "learning_rate": 3.721784825470889e-05,
+      "loss": 0.0002,
+      "step": 480
+    },
+    {
+      "epoch": 0.6144,
+      "eval_loss": 1.1689307689666748,
+      "eval_runtime": 42.2181,
+      "eval_samples_per_second": 11.891,
+      "eval_steps_per_second": 1.492,
+      "step": 480
+    },
+    {
+      "epoch": 0.61568,
+      "grad_norm": 0.004279144573956728,
+      "learning_rate": 3.7204558440527136e-05,
+      "loss": 0.0002,
+      "step": 481
+    },
+    {
+      "epoch": 0.61696,
+      "grad_norm": 0.0062600066885352135,
+      "learning_rate": 3.7191239632175e-05,
+      "loss": 0.0002,
+      "step": 482
+    },
+    {
+      "epoch": 0.61824,
+      "grad_norm": 0.005994696635752916,
+      "learning_rate": 3.717789185505064e-05,
+      "loss": 0.0002,
+      "step": 483
+    },
+    {
+      "epoch": 0.61952,
+      "grad_norm": 0.0035205462481826544,
+      "learning_rate": 3.7164515134607475e-05,
+      "loss": 0.0002,
+      "step": 484
+    },
+    {
+      "epoch": 0.6208,
+      "grad_norm": 0.00023122667334973812,
+      "learning_rate": 3.7151109496354086e-05,
+      "loss": 0.0002,
+      "step": 485
+    },
+    {
+      "epoch": 0.62208,
+      "grad_norm": 0.0031270990148186684,
+      "learning_rate": 3.713767496585421e-05,
+      "loss": 0.0002,
+      "step": 486
+    },
+    {
+      "epoch": 0.62336,
+      "grad_norm": 0.00399185623973608,
+      "learning_rate": 3.712421156872668e-05,
+      "loss": 0.0002,
+      "step": 487
+    },
+    {
+      "epoch": 0.62464,
+      "grad_norm": 0.0028176584746688604,
+      "learning_rate": 3.711071933064539e-05,
+      "loss": 0.0002,
+      "step": 488
+    },
+    {
+      "epoch": 0.62592,
+      "grad_norm": 0.00030795088969171047,
+      "learning_rate": 3.7097198277339216e-05,
+      "loss": 0.0002,
+      "step": 489
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 0.0023664638865739107,
+      "learning_rate": 3.708364843459198e-05,
+      "loss": 0.0002,
+      "step": 490
+    },
+    {
+      "epoch": 0.62848,
+      "grad_norm": 0.004323100205510855,
+      "learning_rate": 3.70700698282424e-05,
+      "loss": 0.0002,
+      "step": 491
+    },
+    {
+      "epoch": 0.62976,
+      "grad_norm": 0.0050489455461502075,
+      "learning_rate": 3.705646248418408e-05,
+      "loss": 0.0002,
+      "step": 492
+    },
+    {
+      "epoch": 0.63104,
+      "grad_norm": 0.0033712410368025303,
+      "learning_rate": 3.704282642836539e-05,
+      "loss": 0.0002,
+      "step": 493
+    },
+    {
+      "epoch": 0.63232,
+      "grad_norm": 0.0005509121110662818,
+      "learning_rate": 3.702916168678945e-05,
+      "loss": 0.0002,
+      "step": 494
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 0.0014277957379817963,
+      "learning_rate": 3.70154682855141e-05,
+      "loss": 0.0002,
+      "step": 495
+    },
+    {
+      "epoch": 0.63488,
+      "grad_norm": 0.001731383497826755,
+      "learning_rate": 3.700174625065183e-05,
+      "loss": 0.0002,
+      "step": 496
+    },
+    {
+      "epoch": 0.63616,
+      "grad_norm": 0.0005305620143190026,
+      "learning_rate": 3.698799560836972e-05,
+      "loss": 0.0002,
+      "step": 497
+    },
+    {
+      "epoch": 0.63744,
+      "grad_norm": 0.0011219007428735495,
+      "learning_rate": 3.697421638488941e-05,
+      "loss": 0.0002,
+      "step": 498
+    },
+    {
+      "epoch": 0.63872,
+      "grad_norm": 0.002674842020496726,
+      "learning_rate": 3.696040860648706e-05,
+      "loss": 0.0002,
+      "step": 499
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.003906667232513428,
+      "learning_rate": 3.694657229949324e-05,
+      "loss": 0.0002,
+      "step": 500
+    },
+    {
+      "epoch": 0.64128,
+      "grad_norm": 0.0040111360140144825,
+      "learning_rate": 3.6932707490292965e-05,
+      "loss": 0.0002,
+      "step": 501
+    },
+    {
+      "epoch": 0.64256,
+      "grad_norm": 0.002312349621206522,
+      "learning_rate": 3.691881420532558e-05,
+      "loss": 0.0002,
+      "step": 502
+    },
+    {
+      "epoch": 0.64384,
+      "grad_norm": 0.0004572660254780203,
+      "learning_rate": 3.690489247108474e-05,
+      "loss": 0.0002,
+      "step": 503
+    },
+    {
+      "epoch": 0.64512,
+      "grad_norm": 0.0024236240424215794,
+      "learning_rate": 3.689094231411834e-05,
+      "loss": 0.0002,
+      "step": 504
+    },
+    {
+      "epoch": 0.6464,
+      "grad_norm": 0.002606478286907077,
+      "learning_rate": 3.6876963761028486e-05,
+      "loss": 0.0002,
+      "step": 505
+    },
+    {
+      "epoch": 0.64768,
+      "grad_norm": 0.001922594616189599,
+      "learning_rate": 3.686295683847144e-05,
+      "loss": 0.0002,
+      "step": 506
+    },
+    {
+      "epoch": 0.64896,
+      "grad_norm": 0.0006554339197464287,
+      "learning_rate": 3.684892157315754e-05,
+      "loss": 0.0002,
+      "step": 507
+    },
+    {
+      "epoch": 0.65024,
+      "grad_norm": 0.0007631480693817139,
+      "learning_rate": 3.683485799185119e-05,
+      "loss": 0.0002,
+      "step": 508
+    },
+    {
+      "epoch": 0.65152,
+      "grad_norm": 0.0010840974282473326,
+      "learning_rate": 3.682076612137079e-05,
+      "loss": 0.0002,
+      "step": 509
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 0.000218061642954126,
+      "learning_rate": 3.680664598858868e-05,
+      "loss": 0.0002,
+      "step": 510
+    },
+    {
+      "epoch": 0.6528,
+      "eval_loss": 1.1986668109893799,
+      "eval_runtime": 43.8576,
+      "eval_samples_per_second": 11.446,
+      "eval_steps_per_second": 1.436,
+      "step": 510
+    },
+    {
+      "epoch": 0.65408,
+      "grad_norm": 0.00179583253338933,
+      "learning_rate": 3.679249762043109e-05,
+      "loss": 0.0002,
+      "step": 511
+    },
+    {
+      "epoch": 0.65536,
+      "grad_norm": 0.0026319276075810194,
+      "learning_rate": 3.67783210438781e-05,
+      "loss": 0.0002,
+      "step": 512
+    },
+    {
+      "epoch": 0.65664,
+      "grad_norm": 0.0023609804920852184,
+      "learning_rate": 3.6764116285963595e-05,
+      "loss": 0.0002,
+      "step": 513
+    },
+    {
+      "epoch": 0.65792,
+      "grad_norm": 0.0012038566637784243,
+      "learning_rate": 3.6749883373775174e-05,
+      "loss": 0.0002,
+      "step": 514
+    },
+    {
+      "epoch": 0.6592,
+      "grad_norm": 0.00014525132428389043,
+      "learning_rate": 3.6735622334454145e-05,
+      "loss": 0.0002,
+      "step": 515
+    },
+    {
+      "epoch": 0.66048,
+      "grad_norm": 0.0007447360549122095,
+      "learning_rate": 3.6721333195195436e-05,
+      "loss": 0.0002,
+      "step": 516
+    },
+    {
+      "epoch": 0.66176,
+      "grad_norm": 0.0010004477808251977,
+      "learning_rate": 3.6707015983247575e-05,
+      "loss": 0.0002,
+      "step": 517
+    },
+    {
+      "epoch": 0.66304,
+      "grad_norm": 0.0005548037588596344,
+      "learning_rate": 3.6692670725912616e-05,
+      "loss": 0.0002,
+      "step": 518
+    },
+    {
+      "epoch": 0.66432,
+      "grad_norm": 0.0002495903172530234,
+      "learning_rate": 3.667829745054609e-05,
+      "loss": 0.0002,
+      "step": 519
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 0.0008480948745273054,
+      "learning_rate": 3.666389618455696e-05,
+      "loss": 0.0002,
+      "step": 520
+    },
+    {
+      "epoch": 0.66688,
+      "grad_norm": 0.00144975446164608,
+      "learning_rate": 3.6649466955407576e-05,
+      "loss": 0.0002,
+      "step": 521
+    },
+    {
+      "epoch": 0.66816,
+      "grad_norm": 0.0022027993109077215,
+      "learning_rate": 3.663500979061359e-05,
+      "loss": 0.0002,
+      "step": 522
+    },
+    {
+      "epoch": 0.66944,
+      "grad_norm": 0.002505333162844181,
+      "learning_rate": 3.6620524717743944e-05,
+      "loss": 0.0002,
+      "step": 523
+    },
+    {
+      "epoch": 0.67072,
+      "grad_norm": 0.0019165552221238613,
+      "learning_rate": 3.66060117644208e-05,
+      "loss": 0.0002,
+      "step": 524
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 0.00015419794362969697,
+      "learning_rate": 3.6591470958319465e-05,
+      "loss": 0.0002,
+      "step": 525
+    },
+    {
+      "epoch": 0.67328,
+      "grad_norm": 0.0018403524300083518,
+      "learning_rate": 3.657690232716839e-05,
+      "loss": 0.0002,
+      "step": 526
+    },
+    {
+      "epoch": 0.67456,
+      "grad_norm": 0.0028898706659674644,
+      "learning_rate": 3.656230589874905e-05,
+      "loss": 0.0002,
+      "step": 527
+    },
+    {
+      "epoch": 0.67584,
+      "grad_norm": 0.0030745165422558784,
+      "learning_rate": 3.6547681700895976e-05,
+      "loss": 0.0002,
+      "step": 528
+    },
+    {
+      "epoch": 0.67712,
+      "grad_norm": 0.002408272586762905,
+      "learning_rate": 3.65330297614966e-05,
+      "loss": 0.0002,
+      "step": 529
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 0.0008627455099485815,
+      "learning_rate": 3.65183501084913e-05,
+      "loss": 0.0002,
+      "step": 530
+    },
+    {
+      "epoch": 0.67968,
+      "grad_norm": 0.0015142913907766342,
+      "learning_rate": 3.650364276987327e-05,
+      "loss": 0.0002,
+      "step": 531
+    },
+    {
+      "epoch": 0.68096,
+      "grad_norm": 0.00383846671320498,
+      "learning_rate": 3.648890777368852e-05,
+      "loss": 0.0002,
+      "step": 532
+    },
+    {
+      "epoch": 0.68224,
+      "grad_norm": 0.005133279133588076,
+      "learning_rate": 3.6474145148035786e-05,
+      "loss": 0.0002,
+      "step": 533
+    },
+    {
+      "epoch": 0.68352,
+      "grad_norm": 0.0045053777284920216,
+      "learning_rate": 3.645935492106651e-05,
+      "loss": 0.0002,
+      "step": 534
+    },
+    {
+      "epoch": 0.6848,
+      "grad_norm": 0.0019303944427520037,
+      "learning_rate": 3.644453712098475e-05,
+      "loss": 0.0002,
+      "step": 535
+    },
+    {
+      "epoch": 0.68608,
+      "grad_norm": 0.0006616184837184846,
+      "learning_rate": 3.6429691776047165e-05,
+      "loss": 0.0002,
+      "step": 536
+    },
+    {
+      "epoch": 0.68736,
+      "grad_norm": 0.0017754195723682642,
+      "learning_rate": 3.6414818914562905e-05,
+      "loss": 0.0002,
+      "step": 537
+    },
+    {
+      "epoch": 0.68864,
+      "grad_norm": 0.0016088245902210474,
+      "learning_rate": 3.639991856489363e-05,
+      "loss": 0.0002,
+      "step": 538
+    },
+    {
+      "epoch": 0.68992,
+      "grad_norm": 0.0012247121194377542,
+      "learning_rate": 3.6384990755453396e-05,
+      "loss": 0.0002,
+      "step": 539
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 0.0010904190130531788,
+      "learning_rate": 3.637003551470863e-05,
+      "loss": 0.0002,
+      "step": 540
+    },
+    {
+      "epoch": 0.6912,
+      "eval_loss": 1.180873990058899,
+      "eval_runtime": 42.9212,
+      "eval_samples_per_second": 11.696,
+      "eval_steps_per_second": 1.468,
+      "step": 540
+    },
+    {
+      "epoch": 0.69248,
+      "grad_norm": 0.001041168114170432,
+      "learning_rate": 3.635505287117807e-05,
+      "loss": 0.0002,
+      "step": 541
+    },
+    {
+      "epoch": 0.69376,
+      "grad_norm": 0.00024401405244134367,
+      "learning_rate": 3.634004285343271e-05,
+      "loss": 0.0002,
+      "step": 542
+    },
+    {
+      "epoch": 0.69504,
+      "grad_norm": 0.0008669750532135367,
+      "learning_rate": 3.6325005490095746e-05,
+      "loss": 0.0002,
+      "step": 543
+    },
+    {
+      "epoch": 0.69632,
+      "grad_norm": 0.0014739528996869922,
+      "learning_rate": 3.630994080984251e-05,
+      "loss": 0.0002,
+      "step": 544
+    },
+    {
+      "epoch": 0.6976,
+      "grad_norm": 0.0015093573601916432,
+      "learning_rate": 3.629484884140044e-05,
+      "loss": 0.0002,
+      "step": 545
+    },
+    {
+      "epoch": 0.69888,
+      "grad_norm": 0.0003642103401944041,
+      "learning_rate": 3.627972961354899e-05,
+      "loss": 0.0002,
+      "step": 546
+    },
+    {
+      "epoch": 0.70016,
+      "grad_norm": 0.001804412342607975,
+      "learning_rate": 3.626458315511963e-05,
+      "loss": 0.0002,
+      "step": 547
+    },
+    {
+      "epoch": 0.70144,
+      "grad_norm": 0.0034682515542954206,
+      "learning_rate": 3.6249409494995724e-05,
+      "loss": 0.0002,
+      "step": 548
+    },
+    {
+      "epoch": 0.70272,
+      "grad_norm": 0.004247341770678759,
+      "learning_rate": 3.6234208662112524e-05,
+      "loss": 0.0002,
+      "step": 549
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 0.004289446864277124,
+      "learning_rate": 3.621898068545711e-05,
+      "loss": 0.0002,
+      "step": 550
+    },
+    {
+      "epoch": 0.70528,
+      "grad_norm": 0.0029438326600939035,
+      "learning_rate": 3.620372559406828e-05,
+      "loss": 0.0002,
+      "step": 551
+    },
+    {
+      "epoch": 0.70656,
+      "grad_norm": 0.0008173590758815408,
+      "learning_rate": 3.618844341703659e-05,
+      "loss": 0.0002,
+      "step": 552
+    },
+    {
+      "epoch": 0.70784,
+      "grad_norm": 0.00509173097088933,
+      "learning_rate": 3.617313418350422e-05,
+      "loss": 0.0002,
+      "step": 553
+    },
+    {
+      "epoch": 0.70912,
+      "grad_norm": 0.0070475186221301556,
+      "learning_rate": 3.6157797922664946e-05,
+      "loss": 0.0002,
+      "step": 554
+    },
+    {
+      "epoch": 0.7104,
+      "grad_norm": 0.0052854400128126144,
+      "learning_rate": 3.614243466376409e-05,
+      "loss": 0.0002,
+      "step": 555
+    },
+    {
+      "epoch": 0.71168,
+      "grad_norm": 0.00032702955650165677,
+      "learning_rate": 3.6127044436098445e-05,
+      "loss": 0.0002,
+      "step": 556
+    },
+    {
+      "epoch": 0.71296,
+      "grad_norm": 0.006080043967813253,
+      "learning_rate": 3.611162726901626e-05,
+      "loss": 0.0002,
+      "step": 557
+    },
+    {
+      "epoch": 0.71424,
+      "grad_norm": 0.007932756096124649,
+      "learning_rate": 3.609618319191712e-05,
+      "loss": 0.0002,
+      "step": 558
+    },
+    {
+      "epoch": 0.71552,
+      "grad_norm": 0.006158908829092979,
+      "learning_rate": 3.608071223425195e-05,
+      "loss": 0.0002,
+      "step": 559
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 0.002953893970698118,
+      "learning_rate": 3.606521442552293e-05,
+      "loss": 0.0002,
+      "step": 560
+    },
+    {
+      "epoch": 0.71808,
+      "grad_norm": 0.0007969742291606963,
+      "learning_rate": 3.604968979528345e-05,
+      "loss": 0.0002,
+      "step": 561
+    },
+    {
+      "epoch": 0.71936,
+      "grad_norm": 0.0039662644267082214,
+      "learning_rate": 3.603413837313801e-05,
+      "loss": 0.0002,
+      "step": 562
+    },
+    {
+      "epoch": 0.72064,
+      "grad_norm": 0.005224650260061026,
+      "learning_rate": 3.601856018874224e-05,
+      "loss": 0.0002,
+      "step": 563
+    },
+    {
+      "epoch": 0.72192,
+      "grad_norm": 0.004226877354085445,
+      "learning_rate": 3.600295527180281e-05,
+      "loss": 0.0002,
+      "step": 564
+    },
+    {
+      "epoch": 0.7232,
+      "grad_norm": 0.0011959822149947286,
+      "learning_rate": 3.5987323652077326e-05,
+      "loss": 0.0002,
+      "step": 565
+    },
+    {
+      "epoch": 0.72448,
+      "grad_norm": 0.0025563903618603945,
+      "learning_rate": 3.597166535937436e-05,
+      "loss": 0.0002,
+      "step": 566
+    },
+    {
+      "epoch": 0.72576,
+      "grad_norm": 0.005399333778768778,
+      "learning_rate": 3.595598042355331e-05,
+      "loss": 0.0002,
+      "step": 567
+    },
+    {
+      "epoch": 0.72704,
+      "grad_norm": 0.005194221623241901,
+      "learning_rate": 3.5940268874524396e-05,
+      "loss": 0.0002,
+      "step": 568
+    },
+    {
+      "epoch": 0.72832,
+      "grad_norm": 0.0018591524567455053,
+      "learning_rate": 3.5924530742248595e-05,
+      "loss": 0.0002,
+      "step": 569
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 0.003616099478676915,
+      "learning_rate": 3.590876605673758e-05,
+      "loss": 0.0002,
+      "step": 570
+    },
+    {
+      "epoch": 0.7296,
+      "eval_loss": 1.2382491827011108,
+      "eval_runtime": 45.2439,
+      "eval_samples_per_second": 11.095,
+      "eval_steps_per_second": 1.392,
+      "step": 570
+    },
+    {
+      "epoch": 0.73088,
+      "grad_norm": 0.007196330465376377,
+      "learning_rate": 3.589297484805363e-05,
+      "loss": 0.0002,
+      "step": 571
+    },
+    {
+      "epoch": 0.73216,
+      "grad_norm": 0.005559357814490795,
+      "learning_rate": 3.587715714630963e-05,
+      "loss": 0.0002,
+      "step": 572
+    },
+    {
+      "epoch": 0.73344,
+      "grad_norm": 0.0007927911938168108,
+      "learning_rate": 3.5861312981668984e-05,
+      "loss": 0.0002,
+      "step": 573
+    },
+    {
+      "epoch": 0.73472,
+      "grad_norm": 0.0043013193644583225,
+      "learning_rate": 3.5845442384345546e-05,
+      "loss": 0.0002,
+      "step": 574
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 0.0068841902539134026,
+      "learning_rate": 3.582954538460359e-05,
+      "loss": 0.0002,
+      "step": 575
+    },
+    {
+      "epoch": 0.73728,
+      "grad_norm": 0.005568365100771189,
+      "learning_rate": 3.581362201275771e-05,
+      "loss": 0.0002,
+      "step": 576
+    },
+    {
+      "epoch": 0.73856,
+      "grad_norm": 0.0008979937992990017,
+      "learning_rate": 3.579767229917284e-05,
+      "loss": 0.0002,
+      "step": 577
+    },
+    {
+      "epoch": 0.73984,
+      "grad_norm": 0.005714059807360172,
+      "learning_rate": 3.5781696274264083e-05,
+      "loss": 0.0002,
+      "step": 578
+    },
+    {
+      "epoch": 0.74112,
+      "grad_norm": 0.00888694729655981,
+      "learning_rate": 3.576569396849678e-05,
+      "loss": 0.0002,
+      "step": 579
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 0.005648651625961065,
+      "learning_rate": 3.574966541238633e-05,
+      "loss": 0.0002,
+      "step": 580
+    },
+    {
+      "epoch": 0.74368,
+      "grad_norm": 0.0008465887513011694,
+      "learning_rate": 3.573361063649823e-05,
+      "loss": 0.0002,
+      "step": 581
+    },
+    {
+      "epoch": 0.74496,
+      "grad_norm": 0.00557989114895463,
+      "learning_rate": 3.571752967144797e-05,
+      "loss": 0.0002,
+      "step": 582
+    },
+    {
+      "epoch": 0.74624,
+      "grad_norm": 0.00599457835778594,
+      "learning_rate": 3.570142254790097e-05,
+      "loss": 0.0002,
+      "step": 583
+    },
+    {
+      "epoch": 0.74752,
+      "grad_norm": 0.002908288734033704,
+      "learning_rate": 3.568528929657252e-05,
+      "loss": 0.0002,
+      "step": 584
+    },
+    {
+      "epoch": 0.7488,
+      "grad_norm": 0.002017195103690028,
+      "learning_rate": 3.566912994822778e-05,
+      "loss": 0.0002,
+      "step": 585
+    },
+    {
+      "epoch": 0.75008,
+      "grad_norm": 0.00458879116922617,
+      "learning_rate": 3.565294453368162e-05,
+      "loss": 0.0002,
+      "step": 586
+    },
+    {
+      "epoch": 0.75136,
+      "grad_norm": 0.0023251990787684917,
+      "learning_rate": 3.563673308379867e-05,
+      "loss": 0.0002,
+      "step": 587
+    },
+    {
+      "epoch": 0.75264,
+      "grad_norm": 0.0018551320536062121,
+      "learning_rate": 3.562049562949316e-05,
+      "loss": 0.0002,
+      "step": 588
+    },
+    {
+      "epoch": 0.75392,
+      "grad_norm": 0.004012602381408215,
+      "learning_rate": 3.560423220172894e-05,
+      "loss": 0.0002,
+      "step": 589
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 0.0016248335596174002,
+      "learning_rate": 3.558794283151938e-05,
+      "loss": 0.0002,
+      "step": 590
+    },
+    {
+      "epoch": 0.75648,
+      "grad_norm": 0.0024472202640026808,
+      "learning_rate": 3.5571627549927316e-05,
+      "loss": 0.0002,
+      "step": 591
+    },
+    {
+      "epoch": 0.75776,
+      "grad_norm": 0.003877311712130904,
+      "learning_rate": 3.5555286388065004e-05,
+      "loss": 0.0002,
+      "step": 592
+    },
+    {
+      "epoch": 0.75904,
+      "grad_norm": 0.002112805377691984,
+      "learning_rate": 3.553891937709404e-05,
+      "loss": 0.0002,
+      "step": 593
+    },
+    {
+      "epoch": 0.76032,
+      "grad_norm": 0.0009606420644558966,
+      "learning_rate": 3.5522526548225326e-05,
+      "loss": 0.0002,
+      "step": 594
+    },
+    {
+      "epoch": 0.7616,
+      "grad_norm": 0.00314708286896348,
+      "learning_rate": 3.5506107932719005e-05,
+      "loss": 0.0002,
+      "step": 595
+    },
+    {
+      "epoch": 0.76288,
+      "grad_norm": 0.003082283539697528,
+      "learning_rate": 3.5489663561884366e-05,
+      "loss": 0.0002,
+      "step": 596
+    },
+    {
+      "epoch": 0.76416,
+      "grad_norm": 0.0001470320567023009,
+      "learning_rate": 3.5473193467079824e-05,
+      "loss": 0.0002,
+      "step": 597
+    },
+    {
+      "epoch": 0.76544,
+      "grad_norm": 0.004448519553989172,
+      "learning_rate": 3.545669767971286e-05,
+      "loss": 0.0002,
+      "step": 598
+    },
+    {
+      "epoch": 0.76672,
+      "grad_norm": 0.005195180885493755,
+      "learning_rate": 3.544017623123993e-05,
+      "loss": 0.0002,
+      "step": 599
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.0006394287920556962,
+      "learning_rate": 3.542362915316644e-05,
+      "loss": 0.0002,
+      "step": 600
+    },
+    {
+      "epoch": 0.768,
+      "eval_loss": 1.164539098739624,
+      "eval_runtime": 41.5569,
+      "eval_samples_per_second": 12.08,
+      "eval_steps_per_second": 1.516,
+      "step": 600
+    },
+    {
+      "epoch": 0.76928,
+      "grad_norm": 0.004153765272349119,
+      "learning_rate": 3.540705647704667e-05,
+      "loss": 0.0002,
+      "step": 601
+    },
+    {
+      "epoch": 0.77056,
+      "grad_norm": 0.00440931273624301,
+      "learning_rate": 3.539045823448369e-05,
+      "loss": 0.0002,
+      "step": 602
+    },
+    {
+      "epoch": 0.77184,
+      "grad_norm": 0.00035680842120200396,
+      "learning_rate": 3.537383445712936e-05,
+      "loss": 0.0002,
+      "step": 603
+    },
+    {
+      "epoch": 0.77312,
+      "grad_norm": 0.004429604858160019,
+      "learning_rate": 3.535718517668421e-05,
+      "loss": 0.0002,
+      "step": 604
+    },
+    {
+      "epoch": 0.7744,
+      "grad_norm": 0.006617020815610886,
+      "learning_rate": 3.534051042489739e-05,
+      "loss": 0.0002,
+      "step": 605
+    },
+    {
+      "epoch": 0.77568,
+      "grad_norm": 0.005061544477939606,
+      "learning_rate": 3.532381023356666e-05,
+      "loss": 0.0002,
+      "step": 606
+    },
+    {
+      "epoch": 0.77696,
+      "grad_norm": 0.0011133073130622506,
+      "learning_rate": 3.530708463453827e-05,
+      "loss": 0.0002,
+      "step": 607
+    },
+    {
+      "epoch": 0.77824,
+      "grad_norm": 0.002393580274656415,
+      "learning_rate": 3.5290333659706915e-05,
+      "loss": 0.0002,
+      "step": 608
+    },
+    {
+      "epoch": 0.77952,
+      "grad_norm": 0.0032681473530828953,
+      "learning_rate": 3.527355734101569e-05,
+      "loss": 0.0002,
+      "step": 609
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 0.0013635074719786644,
+      "learning_rate": 3.525675571045602e-05,
+      "loss": 0.0002,
+      "step": 610
+    },
+    {
+      "epoch": 0.78208,
+      "grad_norm": 0.0019797049462795258,
+      "learning_rate": 3.52399288000676e-05,
+      "loss": 0.0002,
+      "step": 611
+    },
+    {
+      "epoch": 0.78336,
+      "grad_norm": 0.0038020156789571047,
+      "learning_rate": 3.522307664193831e-05,
+      "loss": 0.0002,
+      "step": 612
+    },
+    {
+      "epoch": 0.78464,
+      "grad_norm": 0.0025022916961461306,
+      "learning_rate": 3.5206199268204207e-05,
+      "loss": 0.0002,
+      "step": 613
+    },
+    {
+      "epoch": 0.78592,
+      "grad_norm": 0.0010878605535253882,
+      "learning_rate": 3.5189296711049405e-05,
+      "loss": 0.0002,
+      "step": 614
+    },
+    {
+      "epoch": 0.7872,
+      "grad_norm": 0.003955469932407141,
+      "learning_rate": 3.517236900270608e-05,
+      "loss": 0.0002,
+      "step": 615
+    },
+    {
+      "epoch": 0.78848,
+      "grad_norm": 0.0026833557058125734,
+      "learning_rate": 3.515541617545432e-05,
+      "loss": 0.0002,
+      "step": 616
+    },
+    {
+      "epoch": 0.78976,
+      "grad_norm": 0.0021759020164608955,
+      "learning_rate": 3.513843826162214e-05,
+      "loss": 0.0002,
+      "step": 617
+    },
+    {
+      "epoch": 0.79104,
+      "grad_norm": 0.005862012039870024,
+      "learning_rate": 3.512143529358541e-05,
+      "loss": 0.0002,
+      "step": 618
+    },
+    {
+      "epoch": 0.79232,
+      "grad_norm": 0.004435022361576557,
+      "learning_rate": 3.510440730376775e-05,
+      "loss": 0.0002,
+      "step": 619
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 0.0007107479032129049,
+      "learning_rate": 3.508735432464049e-05,
+      "loss": 0.0002,
+      "step": 620
+    },
+    {
+      "epoch": 0.79488,
+      "grad_norm": 0.005531948991119862,
+      "learning_rate": 3.507027638872264e-05,
+      "loss": 0.0002,
+      "step": 621
+    },
+    {
+      "epoch": 0.79616,
+      "grad_norm": 0.006097315810620785,
+      "learning_rate": 3.505317352858078e-05,
+      "loss": 0.0002,
+      "step": 622
+    },
+    {
+      "epoch": 0.79744,
+      "grad_norm": 0.0015222890069708228,
+      "learning_rate": 3.503604577682903e-05,
+      "loss": 0.0002,
+      "step": 623
+    },
+    {
+      "epoch": 0.79872,
+      "grad_norm": 0.004580902401357889,
+      "learning_rate": 3.5018893166128984e-05,
+      "loss": 0.0002,
+      "step": 624
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.006305031478404999,
+      "learning_rate": 3.500171572918961e-05,
+      "loss": 0.0002,
+      "step": 625
+    },
+    {
+      "epoch": 0.80128,
+      "grad_norm": 0.002156647155061364,
+      "learning_rate": 3.498451349876725e-05,
+      "loss": 0.0002,
+      "step": 626
+    },
+    {
+      "epoch": 0.80256,
+      "grad_norm": 0.003837764263153076,
+      "learning_rate": 3.496728650766552e-05,
+      "loss": 0.0002,
+      "step": 627
+    },
+    {
+      "epoch": 0.80384,
+      "grad_norm": 0.005848568864166737,
+      "learning_rate": 3.495003478873523e-05,
+      "loss": 0.0002,
+      "step": 628
+    },
+    {
+      "epoch": 0.80512,
+      "grad_norm": 0.0019309833878651261,
+      "learning_rate": 3.493275837487437e-05,
+      "loss": 0.0002,
+      "step": 629
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 0.0037213312461972237,
+      "learning_rate": 3.491545729902801e-05,
+      "loss": 0.0002,
+      "step": 630
+    },
+    {
+      "epoch": 0.8064,
+      "eval_loss": 1.227569580078125,
+      "eval_runtime": 45.2126,
+      "eval_samples_per_second": 11.103,
+      "eval_steps_per_second": 1.393,
+      "step": 630
+    },
+    {
+      "epoch": 0.80768,
+      "grad_norm": 0.00538895046338439,
+      "learning_rate": 3.489813159418826e-05,
+      "loss": 0.0002,
+      "step": 631
+    },
+    {
+      "epoch": 0.80896,
+      "grad_norm": 0.0008948746835812926,
+      "learning_rate": 3.488078129339418e-05,
+      "loss": 0.0002,
+      "step": 632
+    },
+    {
+      "epoch": 0.81024,
+      "grad_norm": 0.004895337857306004,
+      "learning_rate": 3.486340642973174e-05,
+      "loss": 0.0002,
+      "step": 633
+    },
+    {
+      "epoch": 0.81152,
+      "grad_norm": 0.004855775274336338,
+      "learning_rate": 3.484600703633376e-05,
+      "loss": 0.0002,
+      "step": 634
+    },
+    {
+      "epoch": 0.8128,
+      "grad_norm": 0.0011731626000255346,
+      "learning_rate": 3.4828583146379817e-05,
+      "loss": 0.0002,
+      "step": 635
+    },
+    {
+      "epoch": 0.81408,
+      "grad_norm": 0.006135343573987484,
+      "learning_rate": 3.4811134793096216e-05,
+      "loss": 0.0002,
+      "step": 636
+    },
+    {
+      "epoch": 0.81536,
+      "grad_norm": 0.003618490183725953,
+      "learning_rate": 3.479366200975591e-05,
+      "loss": 0.0002,
+      "step": 637
+    },
+    {
+      "epoch": 0.81664,
+      "grad_norm": 0.0027814011555165052,
+      "learning_rate": 3.477616482967844e-05,
+      "loss": 0.0002,
+      "step": 638
+    },
+    {
+      "epoch": 0.81792,
+      "grad_norm": 0.006283624563366175,
+      "learning_rate": 3.475864328622985e-05,
+      "loss": 0.0002,
+      "step": 639
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 0.004275933373719454,
+      "learning_rate": 3.474109741282267e-05,
+      "loss": 0.0002,
+      "step": 640
+    },
+    {
+      "epoch": 0.82048,
+      "grad_norm": 0.0015624063089489937,
+      "learning_rate": 3.472352724291583e-05,
+      "loss": 0.0002,
+      "step": 641
+    },
+    {
+      "epoch": 0.82176,
+      "grad_norm": 0.005364539101719856,
+      "learning_rate": 3.4705932810014555e-05,
+      "loss": 0.0002,
+      "step": 642
+    },
+    {
+      "epoch": 0.82304,
+      "grad_norm": 0.00375459436327219,
+      "learning_rate": 3.468831414767038e-05,
+      "loss": 0.0002,
+      "step": 643
+    },
+    {
+      "epoch": 0.82432,
+      "grad_norm": 0.001632419298402965,
+      "learning_rate": 3.467067128948101e-05,
+      "loss": 0.0002,
+      "step": 644
+    },
+    {
+      "epoch": 0.8256,
+      "grad_norm": 0.005220402032136917,
+      "learning_rate": 3.465300426909032e-05,
+      "loss": 0.0002,
+      "step": 645
+    },
+    {
+      "epoch": 0.82688,
+      "grad_norm": 0.0028709019534289837,
+      "learning_rate": 3.463531312018823e-05,
+      "loss": 0.0002,
+      "step": 646
+    },
+    {
+      "epoch": 0.82816,
+      "grad_norm": 0.0024156903382390738,
+      "learning_rate": 3.4617597876510694e-05,
+      "loss": 0.0002,
+      "step": 647
+    },
+    {
+      "epoch": 0.82944,
+      "grad_norm": 0.004406792111694813,
+      "learning_rate": 3.4599858571839606e-05,
+      "loss": 0.0002,
+      "step": 648
+    },
+    {
+      "epoch": 0.83072,
+      "grad_norm": 0.0010315208928659558,
+      "learning_rate": 3.458209524000275e-05,
+      "loss": 0.0002,
+      "step": 649
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 0.003976741805672646,
+      "learning_rate": 3.4564307914873716e-05,
+      "loss": 0.0002,
+      "step": 650
+    },
+    {
+      "epoch": 0.83328,
+      "grad_norm": 0.005697767250239849,
+      "learning_rate": 3.4546496630371854e-05,
+      "loss": 0.0002,
+      "step": 651
+    },
+    {
+      "epoch": 0.83456,
+      "grad_norm": 0.0018598579335957766,
+      "learning_rate": 3.452866142046221e-05,
+      "loss": 0.0002,
+      "step": 652
+    },
+    {
+      "epoch": 0.83584,
+      "grad_norm": 0.00341482344083488,
+      "learning_rate": 3.4510802319155435e-05,
+      "loss": 0.0002,
+      "step": 653
+    },
+    {
+      "epoch": 0.83712,
+      "grad_norm": 0.004788935650140047,
+      "learning_rate": 3.4492919360507754e-05,
+      "loss": 0.0002,
+      "step": 654
+    },
+    {
+      "epoch": 0.8384,
+      "grad_norm": 0.0021135262213647366,
+      "learning_rate": 3.4475012578620895e-05,
+      "loss": 0.0002,
+      "step": 655
+    },
+    {
+      "epoch": 0.83968,
+      "grad_norm": 0.0020841804798692465,
+      "learning_rate": 3.4457082007641996e-05,
+      "loss": 0.0002,
+      "step": 656
+    },
+    {
+      "epoch": 0.84096,
+      "grad_norm": 0.004247150383889675,
+      "learning_rate": 3.4439127681763566e-05,
+      "loss": 0.0002,
+      "step": 657
+    },
+    {
+      "epoch": 0.84224,
+      "grad_norm": 0.002661739941686392,
+      "learning_rate": 3.4421149635223416e-05,
+      "loss": 0.0002,
+      "step": 658
+    },
+    {
+      "epoch": 0.84352,
+      "grad_norm": 0.0008635800913907588,
+      "learning_rate": 3.4403147902304596e-05,
+      "loss": 0.0002,
+      "step": 659
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 0.0035655642859637737,
+      "learning_rate": 3.438512251733532e-05,
+      "loss": 0.0002,
+      "step": 660
+    },
+    {
+      "epoch": 0.8448,
+      "eval_loss": 1.2087175846099854,
+      "eval_runtime": 44.3231,
+      "eval_samples_per_second": 11.326,
+      "eval_steps_per_second": 1.421,
+      "step": 660
+    },
+    {
+      "epoch": 0.84608,
+      "grad_norm": 0.0034810348879545927,
+      "learning_rate": 3.436707351468889e-05,
+      "loss": 0.0002,
+      "step": 661
+    },
+    {
+      "epoch": 0.84736,
+      "grad_norm": 0.0005423646070994437,
+      "learning_rate": 3.434900092878367e-05,
+      "loss": 0.0002,
+      "step": 662
+    },
+    {
+      "epoch": 0.84864,
+      "grad_norm": 0.0021196764428168535,
+      "learning_rate": 3.433090479408298e-05,
+      "loss": 0.0002,
+      "step": 663
+    },
+    {
+      "epoch": 0.84992,
+      "grad_norm": 0.0016890447586774826,
+      "learning_rate": 3.431278514509505e-05,
+      "loss": 0.0002,
+      "step": 664
+    },
+    {
+      "epoch": 0.8512,
+      "grad_norm": 0.0009622541256248951,
+      "learning_rate": 3.4294642016372954e-05,
+      "loss": 0.0002,
+      "step": 665
+    },
+    {
+      "epoch": 0.85248,
+      "grad_norm": 0.003417698899284005,
+      "learning_rate": 3.4276475442514536e-05,
+      "loss": 0.0002,
+      "step": 666
+    },
+    {
+      "epoch": 0.85376,
+      "grad_norm": 0.003611312946304679,
+      "learning_rate": 3.425828545816235e-05,
+      "loss": 0.0002,
+      "step": 667
+    },
+    {
+      "epoch": 0.85504,
+      "grad_norm": 0.0006507485522888601,
+      "learning_rate": 3.42400720980036e-05,
+      "loss": 0.0002,
+      "step": 668
+    },
+    {
+      "epoch": 0.85632,
+      "grad_norm": 0.003617242444306612,
+      "learning_rate": 3.422183539677005e-05,
+      "loss": 0.0002,
+      "step": 669
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 0.004010677337646484,
+      "learning_rate": 3.420357538923798e-05,
+      "loss": 0.0002,
+      "step": 670
+    },
+    {
+      "epoch": 0.85888,
+      "grad_norm": 0.0006595096201635897,
+      "learning_rate": 3.418529211022812e-05,
+      "loss": 0.0002,
+      "step": 671
+    },
+    {
+      "epoch": 0.86016,
+      "grad_norm": 0.004336192272603512,
+      "learning_rate": 3.416698559460558e-05,
+      "loss": 0.0002,
+      "step": 672
+    },
+    {
+      "epoch": 0.86144,
+      "grad_norm": 0.003125009825453162,
+      "learning_rate": 3.414865587727977e-05,
+      "loss": 0.0002,
+      "step": 673
+    },
+    {
+      "epoch": 0.86272,
+      "grad_norm": 0.0010173583868891,
+      "learning_rate": 3.4130302993204345e-05,
+      "loss": 0.0002,
+      "step": 674
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 0.003576030256226659,
+      "learning_rate": 3.411192697737715e-05,
+      "loss": 0.0002,
+      "step": 675
+    },
+    {
+      "epoch": 0.86528,
+      "grad_norm": 0.0021791106555610895,
+      "learning_rate": 3.409352786484012e-05,
+      "loss": 0.0002,
+      "step": 676
+    },
+    {
+      "epoch": 0.86656,
+      "grad_norm": 0.001492811250500381,
+      "learning_rate": 3.407510569067926e-05,
+      "loss": 0.0002,
+      "step": 677
+    },
+    {
+      "epoch": 0.86784,
+      "grad_norm": 0.0030872321221977472,
+      "learning_rate": 3.4056660490024524e-05,
+      "loss": 0.0002,
+      "step": 678
+    },
+    {
+      "epoch": 0.86912,
+      "grad_norm": 0.0008694932330399752,
+      "learning_rate": 3.40381922980498e-05,
+      "loss": 0.0002,
+      "step": 679
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 0.0015984508208930492,
+      "learning_rate": 3.4019701149972806e-05,
+      "loss": 0.0002,
+      "step": 680
+    },
+    {
+      "epoch": 0.87168,
+      "grad_norm": 0.002277463674545288,
+      "learning_rate": 3.400118708105505e-05,
+      "loss": 0.0002,
+      "step": 681
+    },
+    {
+      "epoch": 0.87296,
+      "grad_norm": 0.001001443131826818,
+      "learning_rate": 3.398265012660172e-05,
+      "loss": 0.0002,
+      "step": 682
+    },
+    {
+      "epoch": 0.87424,
+      "grad_norm": 0.0014973531942814589,
+      "learning_rate": 3.396409032196167e-05,
+      "loss": 0.0002,
+      "step": 683
+    },
+    {
+      "epoch": 0.87552,
+      "grad_norm": 0.0019171397434547544,
+      "learning_rate": 3.3945507702527325e-05,
+      "loss": 0.0002,
+      "step": 684
+    },
+    {
+      "epoch": 0.8768,
+      "grad_norm": 0.00016306435281876475,
+      "learning_rate": 3.3926902303734605e-05,
+      "loss": 0.0002,
+      "step": 685
+    },
+    {
+      "epoch": 0.87808,
+      "grad_norm": 0.002407992724329233,
+      "learning_rate": 3.390827416106289e-05,
+      "loss": 0.0002,
+      "step": 686
+    },
+    {
+      "epoch": 0.87936,
+      "grad_norm": 0.002446410246193409,
+      "learning_rate": 3.388962331003491e-05,
+      "loss": 0.0002,
+      "step": 687
+    },
+    {
+      "epoch": 0.88064,
+      "grad_norm": 0.000253974343650043,
+      "learning_rate": 3.38709497862167e-05,
+      "loss": 0.0002,
+      "step": 688
+    },
+    {
+      "epoch": 0.88192,
+      "grad_norm": 0.0016051908023655415,
+      "learning_rate": 3.3852253625217555e-05,
+      "loss": 0.0002,
+      "step": 689
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 0.0011209291405975819,
+      "learning_rate": 3.38335348626899e-05,
+      "loss": 0.0002,
+      "step": 690
+    },
+    {
+      "epoch": 0.8832,
+      "eval_loss": 1.1801164150238037,
+      "eval_runtime": 42.922,
+      "eval_samples_per_second": 11.696,
+      "eval_steps_per_second": 1.468,
+      "step": 690
+    },
+    {
+      "epoch": 0.88448,
+      "grad_norm": 0.0005705386283807456,
+      "learning_rate": 3.381479353432929e-05,
+      "loss": 0.0002,
+      "step": 691
+    },
+    {
+      "epoch": 0.88576,
+      "grad_norm": 0.0013472288846969604,
+      "learning_rate": 3.37960296758743e-05,
+      "loss": 0.0002,
+      "step": 692
+    },
+    {
+      "epoch": 0.88704,
+      "grad_norm": 0.0012993423733860254,
+      "learning_rate": 3.377724332310647e-05,
+      "loss": 0.0002,
+      "step": 693
+    },
+    {
+      "epoch": 0.88832,
+      "grad_norm": 0.00032160914270207286,
+      "learning_rate": 3.3758434511850224e-05,
+      "loss": 0.0002,
+      "step": 694
+    },
+    {
+      "epoch": 0.8896,
+      "grad_norm": 0.0009844450978562236,
+      "learning_rate": 3.3739603277972835e-05,
+      "loss": 0.0002,
+      "step": 695
+    },
+    {
+      "epoch": 0.89088,
+      "grad_norm": 0.0020079065579921007,
+      "learning_rate": 3.372074965738433e-05,
+      "loss": 0.0002,
+      "step": 696
+    },
+    {
+      "epoch": 0.89216,
+      "grad_norm": 0.0014246099162846804,
+      "learning_rate": 3.37018736860374e-05,
+      "loss": 0.0002,
+      "step": 697
+    },
+    {
+      "epoch": 0.89344,
+      "grad_norm": 0.00041524734115228057,
+      "learning_rate": 3.3682975399927386e-05,
+      "loss": 0.0002,
+      "step": 698
+    },
+    {
+      "epoch": 0.89472,
+      "grad_norm": 0.0022344680037349463,
+      "learning_rate": 3.3664054835092184e-05,
+      "loss": 0.0002,
+      "step": 699
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.0023142280988395214,
+      "learning_rate": 3.364511202761214e-05,
+      "loss": 0.0002,
+      "step": 700
+    },
+    {
+      "epoch": 0.89728,
+      "grad_norm": 0.0005925571895204484,
+      "learning_rate": 3.3626147013610066e-05,
+      "loss": 0.0002,
+      "step": 701
+    },
+    {
+      "epoch": 0.89856,
+      "grad_norm": 0.003085650270804763,
+      "learning_rate": 3.360715982925107e-05,
+      "loss": 0.0002,
+      "step": 702
+    },
+    {
+      "epoch": 0.89984,
+      "grad_norm": 0.002100037643685937,
+      "learning_rate": 3.358815051074257e-05,
+      "loss": 0.0002,
+      "step": 703
+    },
+    {
+      "epoch": 0.90112,
+      "grad_norm": 0.0018144150963053107,
+      "learning_rate": 3.356911909433419e-05,
+      "loss": 0.0002,
+      "step": 704
+    },
+    {
+      "epoch": 0.9024,
+      "grad_norm": 0.004643341060727835,
+      "learning_rate": 3.355006561631767e-05,
+      "loss": 0.0002,
+      "step": 705
+    },
+    {
+      "epoch": 0.90368,
+      "grad_norm": 0.003487049136310816,
+      "learning_rate": 3.353099011302685e-05,
+      "loss": 0.0002,
+      "step": 706
+    },
+    {
+      "epoch": 0.90496,
+      "grad_norm": 0.0005208138027228415,
+      "learning_rate": 3.3511892620837554e-05,
+      "loss": 0.0002,
+      "step": 707
+    },
+    {
+      "epoch": 0.90624,
+      "grad_norm": 0.003496536985039711,
+      "learning_rate": 3.349277317616754e-05,
+      "loss": 0.0002,
+      "step": 708
+    },
+    {
+      "epoch": 0.90752,
+      "grad_norm": 0.002766245976090431,
+      "learning_rate": 3.347363181547642e-05,
+      "loss": 0.0002,
+      "step": 709
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 0.0004007960669696331,
+      "learning_rate": 3.345446857526561e-05,
+      "loss": 0.0002,
+      "step": 710
+    },
+    {
+      "epoch": 0.91008,
+      "grad_norm": 0.0024029084015637636,
+      "learning_rate": 3.343528349207827e-05,
+      "loss": 0.0002,
+      "step": 711
+    },
+    {
+      "epoch": 0.91136,
+      "grad_norm": 0.001649223268032074,
+      "learning_rate": 3.341607660249916e-05,
+      "loss": 0.0002,
+      "step": 712
+    },
+    {
+      "epoch": 0.91264,
+      "grad_norm": 0.0008164893370121717,
+      "learning_rate": 3.339684794315467e-05,
+      "loss": 0.0002,
+      "step": 713
+    },
+    {
+      "epoch": 0.91392,
+      "grad_norm": 0.0026240425650030375,
+      "learning_rate": 3.3377597550712676e-05,
+      "loss": 0.0002,
+      "step": 714
+    },
+    {
+      "epoch": 0.9152,
+      "grad_norm": 0.00138380890712142,
+      "learning_rate": 3.3358325461882516e-05,
+      "loss": 0.0002,
+      "step": 715
+    },
+    {
+      "epoch": 0.91648,
+      "grad_norm": 0.0012182299979031086,
+      "learning_rate": 3.333903171341489e-05,
+      "loss": 0.0002,
+      "step": 716
+    },
+    {
+      "epoch": 0.91776,
+      "grad_norm": 0.0022654475178569555,
+      "learning_rate": 3.3319716342101804e-05,
+      "loss": 0.0002,
+      "step": 717
+    },
+    {
+      "epoch": 0.91904,
+      "grad_norm": 0.0006078749429434538,
+      "learning_rate": 3.330037938477651e-05,
+      "loss": 0.0002,
+      "step": 718
+    },
+    {
+      "epoch": 0.92032,
+      "grad_norm": 0.00262492336332798,
+      "learning_rate": 3.3281020878313396e-05,
+      "loss": 0.0002,
+      "step": 719
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 0.0032508145086467266,
+      "learning_rate": 3.3261640859627964e-05,
+      "loss": 0.0002,
+      "step": 720
+    },
+    {
+      "epoch": 0.9216,
+      "eval_loss": 1.1878145933151245,
+      "eval_runtime": 43.4753,
+      "eval_samples_per_second": 11.547,
+      "eval_steps_per_second": 1.449,
+      "step": 720
+    },
+    {
+      "epoch": 0.92288,
+      "grad_norm": 0.00020908880105707794,
+      "learning_rate": 3.324223936567675e-05,
+      "loss": 0.0002,
+      "step": 721
+    },
+    {
+      "epoch": 0.92416,
+      "grad_norm": 0.0035989386960864067,
+      "learning_rate": 3.3222816433457205e-05,
+      "loss": 0.0002,
+      "step": 722
+    },
+    {
+      "epoch": 0.92544,
+      "grad_norm": 0.004371747840195894,
+      "learning_rate": 3.3203372100007694e-05,
+      "loss": 0.0002,
+      "step": 723
+    },
+    {
+      "epoch": 0.92672,
+      "grad_norm": 0.0012274390319362283,
+      "learning_rate": 3.318390640240738e-05,
+      "loss": 0.0002,
+      "step": 724
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 0.002447142032906413,
+      "learning_rate": 3.316441937777615e-05,
+      "loss": 0.0002,
+      "step": 725
+    },
+    {
+      "epoch": 0.92928,
+      "grad_norm": 0.0030363500118255615,
+      "learning_rate": 3.314491106327461e-05,
+      "loss": 0.0002,
+      "step": 726
+    },
+    {
+      "epoch": 0.93056,
+      "grad_norm": 0.00019563132082112134,
+      "learning_rate": 3.3125381496103905e-05,
+      "loss": 0.0002,
+      "step": 727
+    },
+    {
+      "epoch": 0.93184,
+      "grad_norm": 0.003880007890984416,
+      "learning_rate": 3.3105830713505745e-05,
+      "loss": 0.0002,
+      "step": 728
+    },
+    {
+      "epoch": 0.93312,
+      "grad_norm": 0.0046269530430436134,
+      "learning_rate": 3.308625875276228e-05,
+      "loss": 0.0002,
+      "step": 729
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 0.0020352972205728292,
+      "learning_rate": 3.306666565119606e-05,
+      "loss": 0.0002,
+      "step": 730
+    },
+    {
+      "epoch": 0.93568,
+      "grad_norm": 0.0020175424870103598,
+      "learning_rate": 3.304705144616994e-05,
+      "loss": 0.0002,
+      "step": 731
+    },
+    {
+      "epoch": 0.93696,
+      "grad_norm": 0.0037658843211829662,
+      "learning_rate": 3.3027416175087015e-05,
+      "loss": 0.0002,
+      "step": 732
+    },
+    {
+      "epoch": 0.93824,
+      "grad_norm": 0.0012228352716192603,
+      "learning_rate": 3.3007759875390554e-05,
+      "loss": 0.0002,
+      "step": 733
+    },
+    {
+      "epoch": 0.93952,
+      "grad_norm": 0.0029969573952257633,
+      "learning_rate": 3.298808258456393e-05,
+      "loss": 0.0002,
+      "step": 734
+    },
+    {
+      "epoch": 0.9408,
+      "grad_norm": 0.004848573822528124,
+      "learning_rate": 3.296838434013055e-05,
+      "loss": 0.0002,
+      "step": 735
+    },
+    {
+      "epoch": 0.94208,
+      "grad_norm": 0.0023214390967041254,
+      "learning_rate": 3.294866517965375e-05,
+      "loss": 0.0002,
+      "step": 736
+    },
+    {
+      "epoch": 0.94336,
+      "grad_norm": 0.00205830205231905,
+      "learning_rate": 3.2928925140736805e-05,
+      "loss": 0.0002,
+      "step": 737
+    },
+    {
+      "epoch": 0.94464,
+      "grad_norm": 0.0036876010708510876,
+      "learning_rate": 3.2909164261022745e-05,
+      "loss": 0.0002,
+      "step": 738
+    },
+    {
+      "epoch": 0.94592,
+      "grad_norm": 0.0014606869081035256,
+      "learning_rate": 3.2889382578194374e-05,
+      "loss": 0.0002,
+      "step": 739
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 0.0022556378971785307,
+      "learning_rate": 3.286958012997416e-05,
+      "loss": 0.0002,
+      "step": 740
+    },
+    {
+      "epoch": 0.94848,
+      "grad_norm": 0.00396134564653039,
+      "learning_rate": 3.284975695412418e-05,
+      "loss": 0.0002,
+      "step": 741
+    },
+    {
+      "epoch": 0.94976,
+      "grad_norm": 0.0020983486901968718,
+      "learning_rate": 3.282991308844602e-05,
+      "loss": 0.0002,
+      "step": 742
+    },
+    {
+      "epoch": 0.95104,
+      "grad_norm": 0.0017063944833353162,
+      "learning_rate": 3.2810048570780725e-05,
+      "loss": 0.0002,
+      "step": 743
+    },
+    {
+      "epoch": 0.95232,
+      "grad_norm": 0.0034495610743761063,
+      "learning_rate": 3.2790163439008735e-05,
+      "loss": 0.0002,
+      "step": 744
+    },
+    {
+      "epoch": 0.9536,
+      "grad_norm": 0.00145576277282089,
+      "learning_rate": 3.277025773104978e-05,
+      "loss": 0.0002,
+      "step": 745
+    },
+    {
+      "epoch": 0.95488,
+      "grad_norm": 0.0014225257327780128,
+      "learning_rate": 3.2750331484862844e-05,
+      "loss": 0.0002,
+      "step": 746
+    },
+    {
+      "epoch": 0.95616,
+      "grad_norm": 0.002464696764945984,
+      "learning_rate": 3.273038473844606e-05,
+      "loss": 0.0002,
+      "step": 747
+    },
+    {
+      "epoch": 0.95744,
+      "grad_norm": 0.0012042096350342035,
+      "learning_rate": 3.271041752983667e-05,
+      "loss": 0.0002,
+      "step": 748
+    },
+    {
+      "epoch": 0.95872,
+      "grad_norm": 0.0007343910983763635,
+      "learning_rate": 3.2690429897110925e-05,
+      "loss": 0.0002,
+      "step": 749
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0017276835860684514,
+      "learning_rate": 3.2670421878384044e-05,
+      "loss": 0.0002,
+      "step": 750
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 1.192034363746643,
+      "eval_runtime": 43.5448,
+      "eval_samples_per_second": 11.528,
+      "eval_steps_per_second": 1.447,
+      "step": 750
+    },
+    {
+      "epoch": 0.96128,
+      "grad_norm": 0.001222791033796966,
+      "learning_rate": 3.2650393511810086e-05,
+      "loss": 0.0002,
+      "step": 751
+    },
+    {
+      "epoch": 0.96256,
+      "grad_norm": 0.0004405670042615384,
+      "learning_rate": 3.2630344835581935e-05,
+      "loss": 0.0002,
+      "step": 752
+    },
+    {
+      "epoch": 0.96384,
+      "grad_norm": 0.0020185222383588552,
+      "learning_rate": 3.26102758879312e-05,
+      "loss": 0.0002,
+      "step": 753
+    },
+    {
+      "epoch": 0.96512,
+      "grad_norm": 0.0022137605119496584,
+      "learning_rate": 3.2590186707128156e-05,
+      "loss": 0.0002,
+      "step": 754
+    },
+    {
+      "epoch": 0.9664,
+      "grad_norm": 0.0005233174888417125,
+      "learning_rate": 3.257007733148163e-05,
+      "loss": 0.0002,
+      "step": 755
+    },
+    {
+      "epoch": 0.96768,
+      "grad_norm": 0.0011725371005013585,
+      "learning_rate": 3.254994779933901e-05,
+      "loss": 0.0002,
+      "step": 756
+    },
+    {
+      "epoch": 0.96896,
+      "grad_norm": 0.0009187847608700395,
+      "learning_rate": 3.2529798149086075e-05,
+      "loss": 0.0002,
+      "step": 757
+    },
+    {
+      "epoch": 0.97024,
+      "grad_norm": 0.0007130370941013098,
+      "learning_rate": 3.2509628419146984e-05,
+      "loss": 0.0002,
+      "step": 758
+    },
+    {
+      "epoch": 0.97152,
+      "grad_norm": 0.0014700796455144882,
+      "learning_rate": 3.248943864798419e-05,
+      "loss": 0.0002,
+      "step": 759
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 0.0002132820663973689,
+      "learning_rate": 3.246922887409837e-05,
+      "loss": 0.0002,
+      "step": 760
+    },
+    {
+      "epoch": 0.97408,
+      "grad_norm": 0.0012054058024659753,
+      "learning_rate": 3.2448999136028326e-05,
+      "loss": 0.0002,
+      "step": 761
+    },
+    {
+      "epoch": 0.97536,
+      "grad_norm": 0.00087908492423594,
+      "learning_rate": 3.242874947235095e-05,
+      "loss": 0.0002,
+      "step": 762
+    },
+    {
+      "epoch": 0.97664,
+      "grad_norm": 0.00034825917100533843,
+      "learning_rate": 3.240847992168111e-05,
+      "loss": 0.0002,
+      "step": 763
+    },
+    {
+      "epoch": 0.97792,
+      "grad_norm": 0.0009228674462065101,
+      "learning_rate": 3.238819052267162e-05,
+      "loss": 0.0002,
+      "step": 764
+    },
+    {
+      "epoch": 0.9792,
+      "grad_norm": 0.00041521593811921775,
+      "learning_rate": 3.236788131401313e-05,
+      "loss": 0.0002,
+      "step": 765
+    },
+    {
+      "epoch": 0.98048,
+      "grad_norm": 0.0007673377986066043,
+      "learning_rate": 3.234755233443406e-05,
+      "loss": 0.0002,
+      "step": 766
+    },
+    {
+      "epoch": 0.98176,
+      "grad_norm": 0.0011025749845430255,
+      "learning_rate": 3.232720362270057e-05,
+      "loss": 0.0002,
+      "step": 767
+    },
+    {
+      "epoch": 0.98304,
+      "grad_norm": 0.0006265212432481349,
+      "learning_rate": 3.2306835217616374e-05,
+      "loss": 0.0002,
+      "step": 768
+    },
+    {
+      "epoch": 0.98432,
+      "grad_norm": 0.0001603246491868049,
+      "learning_rate": 3.228644715802282e-05,
+      "loss": 0.0002,
+      "step": 769
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 0.0005604405887424946,
+      "learning_rate": 3.226603948279868e-05,
+      "loss": 0.0002,
+      "step": 770
+    },
+    {
+      "epoch": 0.98688,
+      "grad_norm": 0.000603645050432533,
+      "learning_rate": 3.224561223086017e-05,
+      "loss": 0.0002,
+      "step": 771
+    },
+    {
+      "epoch": 0.98816,
+      "grad_norm": 0.0004497454210650176,
+      "learning_rate": 3.222516544116081e-05,
+      "loss": 0.0002,
+      "step": 772
+    },
+    {
+      "epoch": 0.98944,
+      "grad_norm": 0.0003486175846774131,
+      "learning_rate": 3.220469915269138e-05,
+      "loss": 0.0002,
+      "step": 773
+    },
+    {
+      "epoch": 0.99072,
+      "grad_norm": 0.0009738111984916031,
+      "learning_rate": 3.218421340447986e-05,
+      "loss": 0.0002,
+      "step": 774
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 0.001073844963684678,
+      "learning_rate": 3.216370823559133e-05,
+      "loss": 0.0002,
+      "step": 775
+    },
+    {
+      "epoch": 0.99328,
+      "grad_norm": 0.0004396456351969391,
+      "learning_rate": 3.21431836851279e-05,
+      "loss": 0.0002,
+      "step": 776
+    },
+    {
+      "epoch": 0.99456,
+      "grad_norm": 0.00045633409172296524,
+      "learning_rate": 3.2122639792228645e-05,
+      "loss": 0.0002,
+      "step": 777
+    },
+    {
+      "epoch": 0.99584,
+      "grad_norm": 0.0005275820149108768,
+      "learning_rate": 3.210207659606951e-05,
+      "loss": 0.0002,
+      "step": 778
+    },
+    {
+      "epoch": 0.99712,
+      "grad_norm": 0.00014961262058932334,
+      "learning_rate": 3.208149413586329e-05,
+      "loss": 0.0002,
+      "step": 779
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 0.0009144467767328024,
+      "learning_rate": 3.206089245085945e-05,
+      "loss": 0.0002,
+      "step": 780
+    },
+    {
+      "epoch": 0.9984,
+      "eval_loss": 1.194998025894165,
+      "eval_runtime": 43.8434,
+      "eval_samples_per_second": 11.45,
+      "eval_steps_per_second": 1.437,
+      "step": 780
+    },
+    {
+      "epoch": 0.99968,
+      "grad_norm": 0.0014543908182531595,
+      "learning_rate": 3.2040271580344176e-05,
+      "loss": 0.0002,
+      "step": 781
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0022573259193450212,
+      "learning_rate": 3.201963156364022e-05,
+      "loss": 0.0002,
+      "step": 782
+    },
+    {
+      "epoch": 1.00128,
+      "grad_norm": 0.0029144599102437496,
+      "learning_rate": 3.1998972440106833e-05,
+      "loss": 0.0002,
+      "step": 783
+    },
+    {
+      "epoch": 1.00256,
+      "grad_norm": 0.0017878921935334802,
+      "learning_rate": 3.197829424913971e-05,
+      "loss": 0.0002,
+      "step": 784
+    },
+    {
+      "epoch": 1.00384,
+      "grad_norm": 0.0008101496496237814,
+      "learning_rate": 3.195759703017091e-05,
+      "loss": 0.0002,
+      "step": 785
+    },
+    {
+      "epoch": 1.00512,
+      "grad_norm": 0.003011012217029929,
+      "learning_rate": 3.193688082266878e-05,
+      "loss": 0.0002,
+      "step": 786
+    },
+    {
+      "epoch": 1.0064,
+      "grad_norm": 0.002561653731390834,
+      "learning_rate": 3.191614566613785e-05,
+      "loss": 0.0002,
+      "step": 787
+    },
+    {
+      "epoch": 1.00768,
+      "grad_norm": 0.0004529696889221668,
+      "learning_rate": 3.189539160011882e-05,
+      "loss": 0.0002,
+      "step": 788
+    },
+    {
+      "epoch": 1.00896,
+      "grad_norm": 0.003300096606835723,
+      "learning_rate": 3.1874618664188435e-05,
+      "loss": 0.0002,
+      "step": 789
+    },
+    {
+      "epoch": 1.01024,
+      "grad_norm": 0.0036370474845170975,
+      "learning_rate": 3.185382689795943e-05,
+      "loss": 0.0002,
+      "step": 790
+    },
+    {
+      "epoch": 1.01152,
+      "grad_norm": 0.0012020476860925555,
+      "learning_rate": 3.1833016341080424e-05,
+      "loss": 0.0002,
+      "step": 791
+    },
+    {
+      "epoch": 1.0128,
+      "grad_norm": 0.001177908037789166,
+      "learning_rate": 3.18121870332359e-05,
+      "loss": 0.0002,
+      "step": 792
+    },
+    {
+      "epoch": 1.01408,
+      "grad_norm": 0.0015154675347730517,
+      "learning_rate": 3.1791339014146075e-05,
+      "loss": 0.0002,
+      "step": 793
+    },
+    {
+      "epoch": 1.01536,
+      "grad_norm": 0.0005656080902554095,
+      "learning_rate": 3.1770472323566864e-05,
+      "loss": 0.0002,
+      "step": 794
+    },
+    {
+      "epoch": 1.01664,
+      "grad_norm": 0.0005200884770601988,
+      "learning_rate": 3.174958700128977e-05,
+      "loss": 0.0002,
+      "step": 795
+    },
+    {
+      "epoch": 1.01792,
+      "grad_norm": 0.0005907294107601047,
+      "learning_rate": 3.172868308714185e-05,
+      "loss": 0.0002,
+      "step": 796
+    },
+    {
+      "epoch": 1.0192,
+      "grad_norm": 0.0002785873075481504,
+      "learning_rate": 3.170776062098559e-05,
+      "loss": 0.0002,
+      "step": 797
+    },
+    {
+      "epoch": 1.02048,
+      "grad_norm": 0.00015250472642946988,
+      "learning_rate": 3.168681964271886e-05,
+      "loss": 0.0002,
+      "step": 798
+    },
+    {
+      "epoch": 1.02176,
+      "grad_norm": 0.00017732252308633178,
+      "learning_rate": 3.166586019227485e-05,
+      "loss": 0.0002,
+      "step": 799
+    },
+    {
+      "epoch": 1.02304,
+      "grad_norm": 0.00015901085862424225,
+      "learning_rate": 3.164488230962194e-05,
+      "loss": 0.0002,
+      "step": 800
+    },
+    {
+      "epoch": 1.02432,
+      "grad_norm": 0.0001894240704132244,
+      "learning_rate": 3.16238860347637e-05,
+      "loss": 0.0002,
+      "step": 801
+    },
+    {
+      "epoch": 1.0256,
+      "grad_norm": 0.00015613554569426924,
+      "learning_rate": 3.160287140773873e-05,
+      "loss": 0.0002,
+      "step": 802
+    },
+    {
+      "epoch": 1.02688,
+      "grad_norm": 0.00070232676807791,
+      "learning_rate": 3.1581838468620674e-05,
+      "loss": 0.0002,
+      "step": 803
+    },
+    {
+      "epoch": 1.02816,
+      "grad_norm": 0.0006290404126048088,
+      "learning_rate": 3.156078725751805e-05,
+      "loss": 0.0002,
+      "step": 804
+    },
+    {
+      "epoch": 1.02944,
+      "grad_norm": 0.00029896237538196146,
+      "learning_rate": 3.1539717814574244e-05,
+      "loss": 0.0002,
+      "step": 805
+    },
+    {
+      "epoch": 1.03072,
+      "grad_norm": 0.0009222381049767137,
+      "learning_rate": 3.151863017996741e-05,
+      "loss": 0.0002,
+      "step": 806
+    },
+    {
+      "epoch": 1.032,
+      "grad_norm": 0.001593570108525455,
+      "learning_rate": 3.1497524393910396e-05,
+      "loss": 0.0002,
+      "step": 807
+    },
+    {
+      "epoch": 1.03328,
+      "grad_norm": 0.000536854553502053,
+      "learning_rate": 3.147640049665063e-05,
+      "loss": 0.0002,
+      "step": 808
+    },
+    {
+      "epoch": 1.03456,
+      "grad_norm": 0.0011051252949982882,
+      "learning_rate": 3.145525852847013e-05,
+      "loss": 0.0002,
+      "step": 809
+    },
+    {
+      "epoch": 1.03584,
+      "grad_norm": 0.0017763072391971946,
+      "learning_rate": 3.143409852968534e-05,
+      "loss": 0.0002,
+      "step": 810
+    },
+    {
+      "epoch": 1.03584,
+      "eval_loss": 1.1909286975860596,
+      "eval_runtime": 43.7864,
+      "eval_samples_per_second": 11.465,
+      "eval_steps_per_second": 1.439,
+      "step": 810
+    },
+    {
+      "epoch": 1.03712,
+      "grad_norm": 0.0015747741563245654,
+      "learning_rate": 3.141292054064707e-05,
+      "loss": 0.0002,
+      "step": 811
+    },
+    {
+      "epoch": 1.0384,
+      "grad_norm": 0.0007148962467908859,
+      "learning_rate": 3.139172460174049e-05,
+      "loss": 0.0002,
+      "step": 812
+    },
+    {
+      "epoch": 1.03968,
+      "grad_norm": 0.001513791736215353,
+      "learning_rate": 3.137051075338496e-05,
+      "loss": 0.0002,
+      "step": 813
+    },
+    {
+      "epoch": 1.04096,
+      "grad_norm": 0.0031881192699074745,
+      "learning_rate": 3.134927903603399e-05,
+      "loss": 0.0002,
+      "step": 814
+    },
+    {
+      "epoch": 1.04224,
+      "grad_norm": 0.002331627532839775,
+      "learning_rate": 3.132802949017519e-05,
+      "loss": 0.0002,
+      "step": 815
+    },
+    {
+      "epoch": 1.04352,
+      "grad_norm": 0.00039511098293587565,
+      "learning_rate": 3.130676215633016e-05,
+      "loss": 0.0002,
+      "step": 816
+    },
+    {
+      "epoch": 1.0448,
+      "grad_norm": 0.0027486959006637335,
+      "learning_rate": 3.128547707505439e-05,
+      "loss": 0.0002,
+      "step": 817
+    },
+    {
+      "epoch": 1.04608,
+      "grad_norm": 0.002707613864913583,
+      "learning_rate": 3.126417428693726e-05,
+      "loss": 0.0002,
+      "step": 818
+    },
+    {
+      "epoch": 1.04736,
+      "grad_norm": 0.00034304489963687956,
+      "learning_rate": 3.12428538326019e-05,
+      "loss": 0.0002,
+      "step": 819
+    },
+    {
+      "epoch": 1.04864,
+      "grad_norm": 0.0018327307188883424,
+      "learning_rate": 3.12215157527051e-05,
+      "loss": 0.0002,
+      "step": 820
+    },
+    {
+      "epoch": 1.04992,
+      "grad_norm": 0.0018494493560865521,
+      "learning_rate": 3.1200160087937304e-05,
+      "loss": 0.0002,
+      "step": 821
+    },
+    {
+      "epoch": 1.0512,
+      "grad_norm": 0.0004779093724209815,
+      "learning_rate": 3.117878687902245e-05,
+      "loss": 0.0002,
+      "step": 822
+    },
+    {
+      "epoch": 1.05248,
+      "grad_norm": 0.0009879459394142032,
+      "learning_rate": 3.115739616671796e-05,
+      "loss": 0.0002,
+      "step": 823
+    },
+    {
+      "epoch": 1.05376,
+      "grad_norm": 0.0014417411293834448,
+      "learning_rate": 3.1135987991814635e-05,
+      "loss": 0.0002,
+      "step": 824
+    },
+    {
+      "epoch": 1.05504,
+      "grad_norm": 0.00022263139544520527,
+      "learning_rate": 3.1114562395136545e-05,
+      "loss": 0.0002,
+      "step": 825
+    },
+    {
+      "epoch": 1.05632,
+      "grad_norm": 0.002068109344691038,
+      "learning_rate": 3.109311941754102e-05,
+      "loss": 0.0002,
+      "step": 826
+    },
+    {
+      "epoch": 1.0576,
+      "grad_norm": 0.0023081593681126833,
+      "learning_rate": 3.107165909991851e-05,
+      "loss": 0.0002,
+      "step": 827
+    },
+    {
+      "epoch": 1.05888,
+      "grad_norm": 0.00045680778566747904,
+      "learning_rate": 3.105018148319254e-05,
+      "loss": 0.0002,
+      "step": 828
+    },
+    {
+      "epoch": 1.06016,
+      "grad_norm": 0.0015283108223229647,
+      "learning_rate": 3.102868660831962e-05,
+      "loss": 0.0002,
+      "step": 829
+    },
+    {
+      "epoch": 1.06144,
+      "grad_norm": 0.001918208785355091,
+      "learning_rate": 3.100717451628917e-05,
+      "loss": 0.0002,
+      "step": 830
+    },
+    {
+      "epoch": 1.06272,
+      "grad_norm": 0.0006047256174497306,
+      "learning_rate": 3.098564524812345e-05,
+      "loss": 0.0002,
+      "step": 831
+    },
+    {
+      "epoch": 1.064,
+      "grad_norm": 0.0013389823725447059,
+      "learning_rate": 3.0964098844877464e-05,
+      "loss": 0.0002,
+      "step": 832
+    },
+    {
+      "epoch": 1.06528,
+      "grad_norm": 0.002758131129667163,
+      "learning_rate": 3.0942535347638896e-05,
+      "loss": 0.0002,
+      "step": 833
+    },
+    {
+      "epoch": 1.06656,
+      "grad_norm": 0.00223194039426744,
+      "learning_rate": 3.092095479752803e-05,
+      "loss": 0.0002,
+      "step": 834
+    },
+    {
+      "epoch": 1.06784,
+      "grad_norm": 0.0002103773003909737,
+      "learning_rate": 3.0899357235697665e-05,
+      "loss": 0.0002,
+      "step": 835
+    },
+    {
+      "epoch": 1.06912,
+      "grad_norm": 0.0022379762958735228,
+      "learning_rate": 3.0877742703333034e-05,
+      "loss": 0.0002,
+      "step": 836
+    },
+    {
+      "epoch": 1.0704,
+      "grad_norm": 0.0028583749663084745,
+      "learning_rate": 3.085611124165175e-05,
+      "loss": 0.0002,
+      "step": 837
+    },
+    {
+      "epoch": 1.07168,
+      "grad_norm": 0.000792765524238348,
+      "learning_rate": 3.083446289190369e-05,
+      "loss": 0.0002,
+      "step": 838
+    },
+    {
+      "epoch": 1.07296,
+      "grad_norm": 0.001578961149789393,
+      "learning_rate": 3.081279769537095e-05,
+      "loss": 0.0002,
+      "step": 839
+    },
+    {
+      "epoch": 1.07424,
+      "grad_norm": 0.0018366475123912096,
+      "learning_rate": 3.0791115693367725e-05,
+      "loss": 0.0002,
+      "step": 840
+    },
+    {
+      "epoch": 1.07424,
+      "eval_loss": 1.183817982673645,
+      "eval_runtime": 43.479,
+      "eval_samples_per_second": 11.546,
+      "eval_steps_per_second": 1.449,
+      "step": 840
+    },
+    {
+      "epoch": 1.07552,
+      "grad_norm": 0.00039021187694743276,
+      "learning_rate": 3.076941692724031e-05,
+      "loss": 0.0002,
+      "step": 841
+    },
+    {
+      "epoch": 1.0768,
+      "grad_norm": 0.0007836410659365356,
+      "learning_rate": 3.0747701438366906e-05,
+      "loss": 0.0002,
+      "step": 842
+    },
+    {
+      "epoch": 1.07808,
+      "grad_norm": 0.00045217288425192237,
+      "learning_rate": 3.072596926815766e-05,
+      "loss": 0.0002,
+      "step": 843
+    },
+    {
+      "epoch": 1.07936,
+      "grad_norm": 0.0005364661919884384,
+      "learning_rate": 3.070422045805448e-05,
+      "loss": 0.0002,
+      "step": 844
+    },
+    {
+      "epoch": 1.08064,
+      "grad_norm": 0.0008205175399780273,
+      "learning_rate": 3.068245504953103e-05,
+      "loss": 0.0002,
+      "step": 845
+    },
+    {
+      "epoch": 1.08192,
+      "grad_norm": 0.0001271996443392709,
+      "learning_rate": 3.066067308409262e-05,
+      "loss": 0.0002,
+      "step": 846
+    },
+    {
+      "epoch": 1.0832,
+      "grad_norm": 0.0013580002123489976,
+      "learning_rate": 3.063887460327616e-05,
+      "loss": 0.0002,
+      "step": 847
+    },
+    {
+      "epoch": 1.08448,
+      "grad_norm": 0.002191178034991026,
+      "learning_rate": 3.061705964865e-05,
+      "loss": 0.0002,
+      "step": 848
+    },
+    {
+      "epoch": 1.08576,
+      "grad_norm": 0.0014392256271094084,
+      "learning_rate": 3.059522826181396e-05,
+      "loss": 0.0002,
+      "step": 849
+    },
+    {
+      "epoch": 1.08704,
+      "grad_norm": 0.000954207091126591,
+      "learning_rate": 3.0573380484399155e-05,
+      "loss": 0.0002,
+      "step": 850
+    },
+    {
+      "epoch": 1.08832,
+      "grad_norm": 0.0029570641927421093,
+      "learning_rate": 3.055151635806797e-05,
+      "loss": 0.0002,
+      "step": 851
+    },
+    {
+      "epoch": 1.0896,
+      "grad_norm": 0.002237113891169429,
+      "learning_rate": 3.0529635924513974e-05,
+      "loss": 0.0002,
+      "step": 852
+    },
+    {
+      "epoch": 1.09088,
+      "grad_norm": 0.0007386120851151645,
+      "learning_rate": 3.0507739225461827e-05,
+      "loss": 0.0002,
+      "step": 853
+    },
+    {
+      "epoch": 1.09216,
+      "grad_norm": 0.0026111765764653683,
+      "learning_rate": 3.0485826302667206e-05,
+      "loss": 0.0002,
+      "step": 854
+    },
+    {
+      "epoch": 1.09344,
+      "grad_norm": 0.0016683172434568405,
+      "learning_rate": 3.046389719791672e-05,
+      "loss": 0.0002,
+      "step": 855
+    },
+    {
+      "epoch": 1.09472,
+      "grad_norm": 0.0005721906200051308,
+      "learning_rate": 3.044195195302784e-05,
+      "loss": 0.0002,
+      "step": 856
+    },
+    {
+      "epoch": 1.096,
+      "grad_norm": 0.0019186767749488354,
+      "learning_rate": 3.041999060984882e-05,
+      "loss": 0.0002,
+      "step": 857
+    },
+    {
+      "epoch": 1.09728,
+      "grad_norm": 0.0013161582173779607,
+      "learning_rate": 3.0398013210258607e-05,
+      "loss": 0.0002,
+      "step": 858
+    },
+    {
+      "epoch": 1.09856,
+      "grad_norm": 0.0007004044600762427,
+      "learning_rate": 3.037601979616677e-05,
+      "loss": 0.0002,
+      "step": 859
+    },
+    {
+      "epoch": 1.09984,
+      "grad_norm": 0.0017194673418998718,
+      "learning_rate": 3.0354010409513418e-05,
+      "loss": 0.0002,
+      "step": 860
+    },
+    {
+      "epoch": 1.10112,
+      "grad_norm": 0.0009747726726345718,
+      "learning_rate": 3.0331985092269107e-05,
+      "loss": 0.0002,
+      "step": 861
+    },
+    {
+      "epoch": 1.1024,
+      "grad_norm": 0.0005672164843417704,
+      "learning_rate": 3.0309943886434796e-05,
+      "loss": 0.0002,
+      "step": 862
+    },
+    {
+      "epoch": 1.10368,
+      "grad_norm": 0.0017328979447484016,
+      "learning_rate": 3.028788683404171e-05,
+      "loss": 0.0002,
+      "step": 863
+    },
+    {
+      "epoch": 1.10496,
+      "grad_norm": 0.00160455540753901,
+      "learning_rate": 3.026581397715132e-05,
+      "loss": 0.0002,
+      "step": 864
+    },
+    {
+      "epoch": 1.1062400000000001,
+      "grad_norm": 0.00018998782616108656,
+      "learning_rate": 3.024372535785522e-05,
+      "loss": 0.0002,
+      "step": 865
+    },
+    {
+      "epoch": 1.10752,
+      "grad_norm": 0.0013784171314910054,
+      "learning_rate": 3.0221621018275067e-05,
+      "loss": 0.0002,
+      "step": 866
+    },
+    {
+      "epoch": 1.1088,
+      "grad_norm": 0.0012219077907502651,
+      "learning_rate": 3.0199501000562498e-05,
+      "loss": 0.0002,
+      "step": 867
+    },
+    {
+      "epoch": 1.11008,
+      "grad_norm": 0.00020059949019923806,
+      "learning_rate": 3.0177365346899045e-05,
+      "loss": 0.0002,
+      "step": 868
+    },
+    {
+      "epoch": 1.11136,
+      "grad_norm": 0.001082933391444385,
+      "learning_rate": 3.015521409949605e-05,
+      "loss": 0.0002,
+      "step": 869
+    },
+    {
+      "epoch": 1.11264,
+      "grad_norm": 0.0016314086969941854,
+      "learning_rate": 3.0133047300594615e-05,
+      "loss": 0.0002,
+      "step": 870
+    },
+    {
+      "epoch": 1.11264,
+      "eval_loss": 1.1762357950210571,
+      "eval_runtime": 42.7757,
+      "eval_samples_per_second": 11.736,
+      "eval_steps_per_second": 1.473,
+      "step": 870
+    },
+    {
+      "epoch": 1.11392,
+      "grad_norm": 0.0017652047099545598,
+      "learning_rate": 3.011086499246546e-05,
+      "loss": 0.0002,
+      "step": 871
+    },
+    {
+      "epoch": 1.1152,
+      "grad_norm": 0.0012991164112463593,
+      "learning_rate": 3.0088667217408907e-05,
+      "loss": 0.0002,
+      "step": 872
+    },
+    {
+      "epoch": 1.11648,
+      "grad_norm": 0.0004405010840855539,
+      "learning_rate": 3.0066454017754768e-05,
+      "loss": 0.0002,
+      "step": 873
+    },
+    {
+      "epoch": 1.11776,
+      "grad_norm": 0.000430526357376948,
+      "learning_rate": 3.0044225435862268e-05,
+      "loss": 0.0002,
+      "step": 874
+    },
+    {
+      "epoch": 1.11904,
+      "grad_norm": 0.00026241704472340643,
+      "learning_rate": 3.0021981514119966e-05,
+      "loss": 0.0002,
+      "step": 875
+    },
+    {
+      "epoch": 1.12032,
+      "grad_norm": 0.0005373109597712755,
+      "learning_rate": 2.9999722294945665e-05,
+      "loss": 0.0002,
+      "step": 876
+    },
+    {
+      "epoch": 1.1216,
+      "grad_norm": 0.0006974454736337066,
+      "learning_rate": 2.9977447820786348e-05,
+      "loss": 0.0002,
+      "step": 877
+    },
+    {
+      "epoch": 1.12288,
+      "grad_norm": 0.0003311955078970641,
+      "learning_rate": 2.9955158134118085e-05,
+      "loss": 0.0002,
+      "step": 878
+    },
+    {
+      "epoch": 1.12416,
+      "grad_norm": 0.0018197010504081845,
+      "learning_rate": 2.9932853277445958e-05,
+      "loss": 0.0002,
+      "step": 879
+    },
+    {
+      "epoch": 1.12544,
+      "grad_norm": 0.001579714473336935,
+      "learning_rate": 2.9910533293303974e-05,
+      "loss": 0.0002,
+      "step": 880
+    },
+    {
+      "epoch": 1.12672,
+      "grad_norm": 0.0002464255376253277,
+      "learning_rate": 2.9888198224254986e-05,
+      "loss": 0.0002,
+      "step": 881
+    },
+    {
+      "epoch": 1.1280000000000001,
+      "grad_norm": 0.001337314024567604,
+      "learning_rate": 2.986584811289062e-05,
+      "loss": 0.0002,
+      "step": 882
+    },
+    {
+      "epoch": 1.12928,
+      "grad_norm": 0.0007345476187765598,
+      "learning_rate": 2.984348300183118e-05,
+      "loss": 0.0002,
+      "step": 883
+    },
+    {
+      "epoch": 1.13056,
+      "grad_norm": 0.0006227174308151007,
+      "learning_rate": 2.982110293372558e-05,
+      "loss": 0.0002,
+      "step": 884
+    },
+    {
+      "epoch": 1.13184,
+      "grad_norm": 0.0008132438524626195,
+      "learning_rate": 2.9798707951251243e-05,
+      "loss": 0.0002,
+      "step": 885
+    },
+    {
+      "epoch": 1.13312,
+      "grad_norm": 0.0004234490043018013,
+      "learning_rate": 2.9776298097114048e-05,
+      "loss": 0.0002,
+      "step": 886
+    },
+    {
+      "epoch": 1.1344,
+      "grad_norm": 0.0015641709323972464,
+      "learning_rate": 2.9753873414048223e-05,
+      "loss": 0.0002,
+      "step": 887
+    },
+    {
+      "epoch": 1.13568,
+      "grad_norm": 0.00038816570304334164,
+      "learning_rate": 2.9731433944816282e-05,
+      "loss": 0.0002,
+      "step": 888
+    },
+    {
+      "epoch": 1.13696,
+      "grad_norm": 0.0018072250531986356,
+      "learning_rate": 2.9708979732208933e-05,
+      "loss": 0.0002,
+      "step": 889
+    },
+    {
+      "epoch": 1.13824,
+      "grad_norm": 0.0019957402255386114,
+      "learning_rate": 2.9686510819044988e-05,
+      "loss": 0.0002,
+      "step": 890
+    },
+    {
+      "epoch": 1.13952,
+      "grad_norm": 0.0002762428193818778,
+      "learning_rate": 2.966402724817131e-05,
+      "loss": 0.0002,
+      "step": 891
+    },
+    {
+      "epoch": 1.1408,
+      "grad_norm": 0.0021149199455976486,
+      "learning_rate": 2.9641529062462703e-05,
+      "loss": 0.0002,
+      "step": 892
+    },
+    {
+      "epoch": 1.14208,
+      "grad_norm": 0.002743937075138092,
+      "learning_rate": 2.9619016304821837e-05,
+      "loss": 0.0002,
+      "step": 893
+    },
+    {
+      "epoch": 1.14336,
+      "grad_norm": 0.00023781094932928681,
+      "learning_rate": 2.959648901817918e-05,
+      "loss": 0.0002,
+      "step": 894
+    },
+    {
+      "epoch": 1.1446399999999999,
+      "grad_norm": 0.002317876322194934,
+      "learning_rate": 2.95739472454929e-05,
+      "loss": 0.0002,
+      "step": 895
+    },
+    {
+      "epoch": 1.14592,
+      "grad_norm": 0.0012434066738933325,
+      "learning_rate": 2.9551391029748785e-05,
+      "loss": 0.0002,
+      "step": 896
+    },
+    {
+      "epoch": 1.1472,
+      "grad_norm": 0.002554555656388402,
+      "learning_rate": 2.9528820413960172e-05,
+      "loss": 0.0002,
+      "step": 897
+    },
+    {
+      "epoch": 1.14848,
+      "grad_norm": 0.004035282880067825,
+      "learning_rate": 2.9506235441167862e-05,
+      "loss": 0.0002,
+      "step": 898
+    },
+    {
+      "epoch": 1.1497600000000001,
+      "grad_norm": 0.0009114958229474723,
+      "learning_rate": 2.9483636154440024e-05,
+      "loss": 0.0002,
+      "step": 899
+    },
+    {
+      "epoch": 1.15104,
+      "grad_norm": 0.003329591127112508,
+      "learning_rate": 2.9461022596872125e-05,
+      "loss": 0.0002,
+      "step": 900
+    },
+    {
+      "epoch": 1.15104,
+      "eval_loss": 1.214892864227295,
+      "eval_runtime": 44.5411,
+      "eval_samples_per_second": 11.27,
+      "eval_steps_per_second": 1.414,
+      "step": 900
+    },
+    {
+      "epoch": 1.15232,
+      "grad_norm": 0.0036533991806209087,
+      "learning_rate": 2.943839481158684e-05,
+      "loss": 0.0002,
+      "step": 901
+    },
+    {
+      "epoch": 1.1536,
+      "grad_norm": 0.00112155731767416,
+      "learning_rate": 2.9415752841734003e-05,
+      "loss": 0.0002,
+      "step": 902
+    },
+    {
+      "epoch": 1.15488,
+      "grad_norm": 0.005300252232700586,
+      "learning_rate": 2.9393096730490454e-05,
+      "loss": 0.0002,
+      "step": 903
+    },
+    {
+      "epoch": 1.15616,
+      "grad_norm": 0.0034816202241927385,
+      "learning_rate": 2.9370426521060036e-05,
+      "loss": 0.0002,
+      "step": 904
+    },
+    {
+      "epoch": 1.15744,
+      "grad_norm": 0.002600653562694788,
+      "learning_rate": 2.9347742256673455e-05,
+      "loss": 0.0002,
+      "step": 905
+    },
+    {
+      "epoch": 1.15872,
+      "grad_norm": 0.007056380622088909,
+      "learning_rate": 2.9325043980588233e-05,
+      "loss": 0.0002,
+      "step": 906
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.004707319661974907,
+      "learning_rate": 2.9302331736088603e-05,
+      "loss": 0.0002,
+      "step": 907
+    },
+    {
+      "epoch": 1.16128,
+      "grad_norm": 0.002767306286841631,
+      "learning_rate": 2.9279605566485437e-05,
+      "loss": 0.0002,
+      "step": 908
+    },
+    {
+      "epoch": 1.16256,
+      "grad_norm": 0.008173026144504547,
+      "learning_rate": 2.925686551511616e-05,
+      "loss": 0.0002,
+      "step": 909
+    },
+    {
+      "epoch": 1.16384,
+      "grad_norm": 0.006080700550228357,
+      "learning_rate": 2.923411162534467e-05,
+      "loss": 0.0002,
+      "step": 910
+    },
+    {
+      "epoch": 1.16512,
+      "grad_norm": 0.0016829116502776742,
+      "learning_rate": 2.9211343940561265e-05,
+      "loss": 0.0002,
+      "step": 911
+    },
+    {
+      "epoch": 1.1663999999999999,
+      "grad_norm": 0.006520835217088461,
+      "learning_rate": 2.918856250418252e-05,
+      "loss": 0.0002,
+      "step": 912
+    },
+    {
+      "epoch": 1.16768,
+      "grad_norm": 0.004198993556201458,
+      "learning_rate": 2.9165767359651275e-05,
+      "loss": 0.0002,
+      "step": 913
+    },
+    {
+      "epoch": 1.16896,
+      "grad_norm": 0.0021160051692277193,
+      "learning_rate": 2.9142958550436457e-05,
+      "loss": 0.0002,
+      "step": 914
+    },
+    {
+      "epoch": 1.17024,
+      "grad_norm": 0.005272620357573032,
+      "learning_rate": 2.9120136120033104e-05,
+      "loss": 0.0002,
+      "step": 915
+    },
+    {
+      "epoch": 1.1715200000000001,
+      "grad_norm": 0.0017011539312079549,
+      "learning_rate": 2.9097300111962197e-05,
+      "loss": 0.0002,
+      "step": 916
+    },
+    {
+      "epoch": 1.1728,
+      "grad_norm": 0.0037756350357085466,
+      "learning_rate": 2.9074450569770628e-05,
+      "loss": 0.0002,
+      "step": 917
+    },
+    {
+      "epoch": 1.17408,
+      "grad_norm": 0.004187179729342461,
+      "learning_rate": 2.9051587537031075e-05,
+      "loss": 0.0002,
+      "step": 918
+    },
+    {
+      "epoch": 1.17536,
+      "grad_norm": 0.0008325865492224693,
+      "learning_rate": 2.9028711057341964e-05,
+      "loss": 0.0002,
+      "step": 919
+    },
+    {
+      "epoch": 1.17664,
+      "grad_norm": 0.0047426228411495686,
+      "learning_rate": 2.9005821174327343e-05,
+      "loss": 0.0002,
+      "step": 920
+    },
+    {
+      "epoch": 1.17792,
+      "grad_norm": 0.0034046040382236242,
+      "learning_rate": 2.8982917931636844e-05,
+      "loss": 0.0002,
+      "step": 921
+    },
+    {
+      "epoch": 1.1792,
+      "grad_norm": 0.0014802594669163227,
+      "learning_rate": 2.896000137294554e-05,
+      "loss": 0.0002,
+      "step": 922
+    },
+    {
+      "epoch": 1.18048,
+      "grad_norm": 0.003786710323765874,
+      "learning_rate": 2.893707154195394e-05,
+      "loss": 0.0002,
+      "step": 923
+    },
+    {
+      "epoch": 1.18176,
+      "grad_norm": 0.00018919807916972786,
+      "learning_rate": 2.891412848238783e-05,
+      "loss": 0.0002,
+      "step": 924
+    },
+    {
+      "epoch": 1.18304,
+      "grad_norm": 0.004215307533740997,
+      "learning_rate": 2.8891172237998236e-05,
+      "loss": 0.0002,
+      "step": 925
+    },
+    {
+      "epoch": 1.18432,
+      "grad_norm": 0.002527839969843626,
+      "learning_rate": 2.8868202852561316e-05,
+      "loss": 0.0002,
+      "step": 926
+    },
+    {
+      "epoch": 1.1856,
+      "grad_norm": 0.002203546231612563,
+      "learning_rate": 2.884522036987831e-05,
+      "loss": 0.0002,
+      "step": 927
+    },
+    {
+      "epoch": 1.18688,
+      "grad_norm": 0.003203033236786723,
+      "learning_rate": 2.8822224833775393e-05,
+      "loss": 0.0002,
+      "step": 928
+    },
+    {
+      "epoch": 1.1881599999999999,
+      "grad_norm": 0.0009034183458425105,
+      "learning_rate": 2.879921628810368e-05,
+      "loss": 0.0002,
+      "step": 929
+    },
+    {
+      "epoch": 1.18944,
+      "grad_norm": 0.004178234841674566,
+      "learning_rate": 2.8776194776739056e-05,
+      "loss": 0.0002,
+      "step": 930
+    },
+    {
+      "epoch": 1.18944,
+      "eval_loss": 1.175118088722229,
+      "eval_runtime": 43.0279,
+      "eval_samples_per_second": 11.667,
+      "eval_steps_per_second": 1.464,
+      "step": 930
+    },
+    {
+      "epoch": 1.19072,
+      "grad_norm": 0.0013845686335116625,
+      "learning_rate": 2.8753160343582162e-05,
+      "loss": 0.0002,
+      "step": 931
+    },
+    {
+      "epoch": 1.192,
+      "grad_norm": 0.004039150197058916,
+      "learning_rate": 2.8730113032558253e-05,
+      "loss": 0.0002,
+      "step": 932
+    },
+    {
+      "epoch": 1.1932800000000001,
+      "grad_norm": 0.005750681739300489,
+      "learning_rate": 2.870705288761715e-05,
+      "loss": 0.0002,
+      "step": 933
+    },
+    {
+      "epoch": 1.19456,
+      "grad_norm": 0.0012514875270426273,
+      "learning_rate": 2.868397995273316e-05,
+      "loss": 0.0002,
+      "step": 934
+    },
+    {
+      "epoch": 1.19584,
+      "grad_norm": 0.005147598218172789,
+      "learning_rate": 2.8660894271904958e-05,
+      "loss": 0.0002,
+      "step": 935
+    },
+    {
+      "epoch": 1.19712,
+      "grad_norm": 0.004961889237165451,
+      "learning_rate": 2.8637795889155552e-05,
+      "loss": 0.0002,
+      "step": 936
+    },
+    {
+      "epoch": 1.1984,
+      "grad_norm": 0.002034895122051239,
+      "learning_rate": 2.8614684848532147e-05,
+      "loss": 0.0002,
+      "step": 937
+    },
+    {
+      "epoch": 1.19968,
+      "grad_norm": 0.007024592719972134,
+      "learning_rate": 2.8591561194106096e-05,
+      "loss": 0.0002,
+      "step": 938
+    },
+    {
+      "epoch": 1.20096,
+      "grad_norm": 0.004905078560113907,
+      "learning_rate": 2.8568424969972805e-05,
+      "loss": 0.0002,
+      "step": 939
+    },
+    {
+      "epoch": 1.20224,
+      "grad_norm": 0.0008516943780705333,
+      "learning_rate": 2.854527622025165e-05,
+      "loss": 0.0002,
+      "step": 940
+    },
+    {
+      "epoch": 1.20352,
+      "grad_norm": 0.0038896326441317797,
+      "learning_rate": 2.852211498908589e-05,
+      "loss": 0.0002,
+      "step": 941
+    },
+    {
+      "epoch": 1.2048,
+      "grad_norm": 0.0018528762739151716,
+      "learning_rate": 2.8498941320642592e-05,
+      "loss": 0.0002,
+      "step": 942
+    },
+    {
+      "epoch": 1.20608,
+      "grad_norm": 0.002043849555775523,
+      "learning_rate": 2.8475755259112528e-05,
+      "loss": 0.0002,
+      "step": 943
+    },
+    {
+      "epoch": 1.20736,
+      "grad_norm": 0.0025435269344598055,
+      "learning_rate": 2.845255684871012e-05,
+      "loss": 0.0002,
+      "step": 944
+    },
+    {
+      "epoch": 1.20864,
+      "grad_norm": 0.0008807358099147677,
+      "learning_rate": 2.842934613367332e-05,
+      "loss": 0.0002,
+      "step": 945
+    },
+    {
+      "epoch": 1.2099199999999999,
+      "grad_norm": 0.003347573336213827,
+      "learning_rate": 2.840612315826356e-05,
+      "loss": 0.0002,
+      "step": 946
+    },
+    {
+      "epoch": 1.2112,
+      "grad_norm": 0.0014326430391520262,
+      "learning_rate": 2.8382887966765645e-05,
+      "loss": 0.0002,
+      "step": 947
+    },
+    {
+      "epoch": 1.21248,
+      "grad_norm": 0.0026219149585813284,
+      "learning_rate": 2.8359640603487673e-05,
+      "loss": 0.0002,
+      "step": 948
+    },
+    {
+      "epoch": 1.21376,
+      "grad_norm": 0.003363543888553977,
+      "learning_rate": 2.833638111276095e-05,
+      "loss": 0.0002,
+      "step": 949
+    },
+    {
+      "epoch": 1.2150400000000001,
+      "grad_norm": 0.00021797107183374465,
+      "learning_rate": 2.8313109538939918e-05,
+      "loss": 0.0002,
+      "step": 950
+    },
+    {
+      "epoch": 1.21632,
+      "grad_norm": 0.0037933571729809046,
+      "learning_rate": 2.828982592640207e-05,
+      "loss": 0.0002,
+      "step": 951
+    },
+    {
+      "epoch": 1.2176,
+      "grad_norm": 0.0028030104003846645,
+      "learning_rate": 2.826653031954781e-05,
+      "loss": 0.0002,
+      "step": 952
+    },
+    {
+      "epoch": 1.21888,
+      "grad_norm": 0.0015061424346640706,
+      "learning_rate": 2.824322276280048e-05,
+      "loss": 0.0002,
+      "step": 953
+    },
+    {
+      "epoch": 1.22016,
+      "grad_norm": 0.002614412922412157,
+      "learning_rate": 2.821990330060616e-05,
+      "loss": 0.0002,
+      "step": 954
+    },
+    {
+      "epoch": 1.22144,
+      "grad_norm": 0.00022668491874355823,
+      "learning_rate": 2.819657197743365e-05,
+      "loss": 0.0002,
+      "step": 955
+    },
+    {
+      "epoch": 1.22272,
+      "grad_norm": 0.0018148926319554448,
+      "learning_rate": 2.8173228837774365e-05,
+      "loss": 0.0002,
+      "step": 956
+    },
+    {
+      "epoch": 1.224,
+      "grad_norm": 0.0004227437893860042,
+      "learning_rate": 2.814987392614227e-05,
+      "loss": 0.0002,
+      "step": 957
+    },
+    {
+      "epoch": 1.22528,
+      "grad_norm": 0.0014729988761246204,
+      "learning_rate": 2.812650728707375e-05,
+      "loss": 0.0002,
+      "step": 958
+    },
+    {
+      "epoch": 1.22656,
+      "grad_norm": 0.0012458571000024676,
+      "learning_rate": 2.8103128965127574e-05,
+      "loss": 0.0002,
+      "step": 959
+    },
+    {
+      "epoch": 1.22784,
+      "grad_norm": 0.00014713928976561874,
+      "learning_rate": 2.8079739004884783e-05,
+      "loss": 0.0002,
+      "step": 960
+    },
+    {
+      "epoch": 1.22784,
+      "eval_loss": 1.1813075542449951,
+      "eval_runtime": 43.1318,
+      "eval_samples_per_second": 11.639,
+      "eval_steps_per_second": 1.461,
+      "step": 960
+    },
+    {
+      "epoch": 1.22912,
+      "grad_norm": 0.0003031356609426439,
+      "learning_rate": 2.805633745094861e-05,
+      "loss": 0.0002,
+      "step": 961
+    },
+    {
+      "epoch": 1.2304,
+      "grad_norm": 0.00014785327948629856,
+      "learning_rate": 2.8032924347944394e-05,
+      "loss": 0.0002,
+      "step": 962
+    },
+    {
+      "epoch": 1.2316799999999999,
+      "grad_norm": 0.00023444350517820567,
+      "learning_rate": 2.8009499740519514e-05,
+      "loss": 0.0002,
+      "step": 963
+    },
+    {
+      "epoch": 1.23296,
+      "grad_norm": 0.00013383221812546253,
+      "learning_rate": 2.7986063673343263e-05,
+      "loss": 0.0002,
+      "step": 964
+    },
+    {
+      "epoch": 1.23424,
+      "grad_norm": 0.0003675081243272871,
+      "learning_rate": 2.796261619110681e-05,
+      "loss": 0.0002,
+      "step": 965
+    },
+    {
+      "epoch": 1.23552,
+      "grad_norm": 0.00046441424638032913,
+      "learning_rate": 2.7939157338523074e-05,
+      "loss": 0.0002,
+      "step": 966
+    },
+    {
+      "epoch": 1.2368000000000001,
+      "grad_norm": 0.00019773181702475995,
+      "learning_rate": 2.791568716032666e-05,
+      "loss": 0.0002,
+      "step": 967
+    },
+    {
+      "epoch": 1.23808,
+      "grad_norm": 0.00011621037992881611,
+      "learning_rate": 2.7892205701273784e-05,
+      "loss": 0.0002,
+      "step": 968
+    },
+    {
+      "epoch": 1.23936,
+      "grad_norm": 0.001013370230793953,
+      "learning_rate": 2.7868713006142156e-05,
+      "loss": 0.0002,
+      "step": 969
+    },
+    {
+      "epoch": 1.24064,
+      "grad_norm": 0.0011750737903639674,
+      "learning_rate": 2.784520911973092e-05,
+      "loss": 0.0002,
+      "step": 970
+    },
+    {
+      "epoch": 1.24192,
+      "grad_norm": 0.0007235942757688463,
+      "learning_rate": 2.7821694086860568e-05,
+      "loss": 0.0002,
+      "step": 971
+    },
+    {
+      "epoch": 1.2432,
+      "grad_norm": 0.001979984575882554,
+      "learning_rate": 2.779816795237283e-05,
+      "loss": 0.0002,
+      "step": 972
+    },
+    {
+      "epoch": 1.24448,
+      "grad_norm": 0.0006019955035299063,
+      "learning_rate": 2.777463076113063e-05,
+      "loss": 0.0002,
+      "step": 973
+    },
+    {
+      "epoch": 1.24576,
+      "grad_norm": 0.0014207311905920506,
+      "learning_rate": 2.7751082558017953e-05,
+      "loss": 0.0002,
+      "step": 974
+    },
+    {
+      "epoch": 1.24704,
+      "grad_norm": 0.0008990809437818825,
+      "learning_rate": 2.77275233879398e-05,
+      "loss": 0.0002,
+      "step": 975
+    },
+    {
+      "epoch": 1.24832,
+      "grad_norm": 0.0007700049318373203,
+      "learning_rate": 2.770395329582208e-05,
+      "loss": 0.0002,
+      "step": 976
+    },
+    {
+      "epoch": 1.2496,
+      "grad_norm": 0.0007865463849157095,
+      "learning_rate": 2.7680372326611517e-05,
+      "loss": 0.0002,
+      "step": 977
+    },
+    {
+      "epoch": 1.25088,
+      "grad_norm": 0.0001616239023860544,
+      "learning_rate": 2.7656780525275598e-05,
+      "loss": 0.0002,
+      "step": 978
+    },
+    {
+      "epoch": 1.25216,
+      "grad_norm": 0.00020688146469183266,
+      "learning_rate": 2.763317793680245e-05,
+      "loss": 0.0002,
+      "step": 979
+    },
+    {
+      "epoch": 1.2534399999999999,
+      "grad_norm": 0.0004241189162712544,
+      "learning_rate": 2.7609564606200788e-05,
+      "loss": 0.0002,
+      "step": 980
+    },
+    {
+      "epoch": 1.25472,
+      "grad_norm": 0.0008062803535722196,
+      "learning_rate": 2.7585940578499792e-05,
+      "loss": 0.0002,
+      "step": 981
+    },
+    {
+      "epoch": 1.256,
+      "grad_norm": 0.0005234317504800856,
+      "learning_rate": 2.7562305898749054e-05,
+      "loss": 0.0002,
+      "step": 982
+    },
+    {
+      "epoch": 1.25728,
+      "grad_norm": 0.00017792911967262626,
+      "learning_rate": 2.7538660612018477e-05,
+      "loss": 0.0002,
+      "step": 983
+    },
+    {
+      "epoch": 1.2585600000000001,
+      "grad_norm": 0.00015215743042062968,
+      "learning_rate": 2.7515004763398172e-05,
+      "loss": 0.0002,
+      "step": 984
+    },
+    {
+      "epoch": 1.25984,
+      "grad_norm": 0.0001460699422750622,
+      "learning_rate": 2.749133839799843e-05,
+      "loss": 0.0002,
+      "step": 985
+    },
+    {
+      "epoch": 1.26112,
+      "grad_norm": 0.00027044335729442537,
+      "learning_rate": 2.746766156094955e-05,
+      "loss": 0.0002,
+      "step": 986
+    },
+    {
+      "epoch": 1.2624,
+      "grad_norm": 0.0005267596570774913,
+      "learning_rate": 2.7443974297401842e-05,
+      "loss": 0.0002,
+      "step": 987
+    },
+    {
+      "epoch": 1.26368,
+      "grad_norm": 0.000697027484420687,
+      "learning_rate": 2.742027665252547e-05,
+      "loss": 0.0002,
+      "step": 988
+    },
+    {
+      "epoch": 1.2649599999999999,
+      "grad_norm": 0.00040344882290810347,
+      "learning_rate": 2.739656867151042e-05,
+      "loss": 0.0002,
+      "step": 989
+    },
+    {
+      "epoch": 1.26624,
+      "grad_norm": 0.0004076792683918029,
+      "learning_rate": 2.737285039956635e-05,
+      "loss": 0.0002,
+      "step": 990
+    },
+    {
+      "epoch": 1.26624,
+      "eval_loss": 1.1873022317886353,
+      "eval_runtime": 43.5805,
+      "eval_samples_per_second": 11.519,
+      "eval_steps_per_second": 1.446,
+      "step": 990
+    },
+    {
+      "epoch": 1.26752,
+      "grad_norm": 0.0003270470770075917,
+      "learning_rate": 2.734912188192258e-05,
+      "loss": 0.0002,
+      "step": 991
+    },
+    {
+      "epoch": 1.2688,
+      "grad_norm": 0.0003373113286215812,
+      "learning_rate": 2.7325383163827947e-05,
+      "loss": 0.0002,
+      "step": 992
+    },
+    {
+      "epoch": 1.27008,
+      "grad_norm": 0.00026391519349999726,
+      "learning_rate": 2.7301634290550762e-05,
+      "loss": 0.0002,
+      "step": 993
+    },
+    {
+      "epoch": 1.27136,
+      "grad_norm": 0.000451684114523232,
+      "learning_rate": 2.727787530737866e-05,
+      "loss": 0.0002,
+      "step": 994
+    },
+    {
+      "epoch": 1.27264,
+      "grad_norm": 0.0008735348237678409,
+      "learning_rate": 2.7254106259618604e-05,
+      "loss": 0.0002,
+      "step": 995
+    },
+    {
+      "epoch": 1.27392,
+      "grad_norm": 0.0006068554357625544,
+      "learning_rate": 2.723032719259671e-05,
+      "loss": 0.0002,
+      "step": 996
+    },
+    {
+      "epoch": 1.2752,
+      "grad_norm": 0.0001193479256471619,
+      "learning_rate": 2.7206538151658222e-05,
+      "loss": 0.0002,
+      "step": 997
+    },
+    {
+      "epoch": 1.27648,
+      "grad_norm": 0.0007612605695612729,
+      "learning_rate": 2.718273918216739e-05,
+      "loss": 0.0002,
+      "step": 998
+    },
+    {
+      "epoch": 1.27776,
+      "grad_norm": 0.000665926025249064,
+      "learning_rate": 2.715893032950742e-05,
+      "loss": 0.0002,
+      "step": 999
+    },
+    {
+      "epoch": 1.27904,
+      "grad_norm": 0.0006610708078369498,
+      "learning_rate": 2.7135111639080335e-05,
+      "loss": 0.0002,
+      "step": 1000
+    },
+    {
+      "epoch": 1.2803200000000001,
+      "grad_norm": 0.0011780158383771777,
+      "learning_rate": 2.7111283156306957e-05,
+      "loss": 0.0002,
+      "step": 1001
+    },
+    {
+      "epoch": 1.2816,
+      "grad_norm": 0.0001723014866001904,
+      "learning_rate": 2.7087444926626717e-05,
+      "loss": 0.0002,
+      "step": 1002
+    },
+    {
+      "epoch": 1.28288,
+      "grad_norm": 0.0009809026960283518,
+      "learning_rate": 2.70635969954977e-05,
+      "loss": 0.0002,
+      "step": 1003
+    },
+    {
+      "epoch": 1.28416,
+      "grad_norm": 0.000595583813264966,
+      "learning_rate": 2.7039739408396456e-05,
+      "loss": 0.0002,
+      "step": 1004
+    },
+    {
+      "epoch": 1.28544,
+      "grad_norm": 0.0006542736082337797,
+      "learning_rate": 2.7015872210817956e-05,
+      "loss": 0.0002,
+      "step": 1005
+    },
+    {
+      "epoch": 1.2867199999999999,
+      "grad_norm": 0.0009241512161679566,
+      "learning_rate": 2.6991995448275506e-05,
+      "loss": 0.0002,
+      "step": 1006
+    },
+    {
+      "epoch": 1.288,
+      "grad_norm": 0.00016221008263528347,
+      "learning_rate": 2.696810916630063e-05,
+      "loss": 0.0002,
+      "step": 1007
+    },
+    {
+      "epoch": 1.28928,
+      "grad_norm": 0.0008979769772849977,
+      "learning_rate": 2.6944213410443026e-05,
+      "loss": 0.0002,
+      "step": 1008
+    },
+    {
+      "epoch": 1.29056,
+      "grad_norm": 0.0011557216057553887,
+      "learning_rate": 2.6920308226270448e-05,
+      "loss": 0.0002,
+      "step": 1009
+    },
+    {
+      "epoch": 1.29184,
+      "grad_norm": 0.0005811048904433846,
+      "learning_rate": 2.6896393659368637e-05,
+      "loss": 0.0002,
+      "step": 1010
+    },
+    {
+      "epoch": 1.29312,
+      "grad_norm": 0.000264172675088048,
+      "learning_rate": 2.6872469755341213e-05,
+      "loss": 0.0002,
+      "step": 1011
+    },
+    {
+      "epoch": 1.2944,
+      "grad_norm": 0.00032463358365930617,
+      "learning_rate": 2.684853655980962e-05,
+      "loss": 0.0002,
+      "step": 1012
+    },
+    {
+      "epoch": 1.29568,
+      "grad_norm": 0.0004930752329528332,
+      "learning_rate": 2.6824594118412998e-05,
+      "loss": 0.0002,
+      "step": 1013
+    },
+    {
+      "epoch": 1.29696,
+      "grad_norm": 0.00100604142062366,
+      "learning_rate": 2.680064247680813e-05,
+      "loss": 0.0002,
+      "step": 1014
+    },
+    {
+      "epoch": 1.29824,
+      "grad_norm": 0.000345872831530869,
+      "learning_rate": 2.6776681680669353e-05,
+      "loss": 0.0002,
+      "step": 1015
+    },
+    {
+      "epoch": 1.29952,
+      "grad_norm": 0.0003196639590896666,
+      "learning_rate": 2.6752711775688452e-05,
+      "loss": 0.0002,
+      "step": 1016
+    },
+    {
+      "epoch": 1.3008,
+      "grad_norm": 0.00015214429004117846,
+      "learning_rate": 2.6728732807574566e-05,
+      "loss": 0.0002,
+      "step": 1017
+    },
+    {
+      "epoch": 1.3020800000000001,
+      "grad_norm": 0.0003405305033084005,
+      "learning_rate": 2.6704744822054155e-05,
+      "loss": 0.0002,
+      "step": 1018
+    },
+    {
+      "epoch": 1.30336,
+      "grad_norm": 0.0002859738888218999,
+      "learning_rate": 2.6680747864870823e-05,
+      "loss": 0.0002,
+      "step": 1019
+    },
+    {
+      "epoch": 1.30464,
+      "grad_norm": 0.0010289245983585715,
+      "learning_rate": 2.665674198178534e-05,
+      "loss": 0.0002,
+      "step": 1020
+    },
+    {
+      "epoch": 1.30464,
+      "eval_loss": 1.1872708797454834,
+      "eval_runtime": 43.5075,
+      "eval_samples_per_second": 11.538,
+      "eval_steps_per_second": 1.448,
+      "step": 1020
+    },
+    {
+      "epoch": 1.30592,
+      "grad_norm": 0.0006891186349093914,
+      "learning_rate": 2.6632727218575446e-05,
+      "loss": 0.0002,
+      "step": 1021
+    },
+    {
+      "epoch": 1.3072,
+      "grad_norm": 0.0004253540828358382,
+      "learning_rate": 2.6608703621035853e-05,
+      "loss": 0.0002,
+      "step": 1022
+    },
+    {
+      "epoch": 1.3084799999999999,
+      "grad_norm": 0.001381613314151764,
+      "learning_rate": 2.6584671234978094e-05,
+      "loss": 0.0002,
+      "step": 1023
+    },
+    {
+      "epoch": 1.30976,
+      "grad_norm": 0.0010794890113174915,
+      "learning_rate": 2.6560630106230478e-05,
+      "loss": 0.0002,
+      "step": 1024
+    },
+    {
+      "epoch": 1.31104,
+      "grad_norm": 0.0005325808888301253,
+      "learning_rate": 2.6536580280637975e-05,
+      "loss": 0.0002,
+      "step": 1025
+    },
+    {
+      "epoch": 1.31232,
+      "grad_norm": 0.0013000130420550704,
+      "learning_rate": 2.651252180406214e-05,
+      "loss": 0.0002,
+      "step": 1026
+    },
+    {
+      "epoch": 1.3136,
+      "grad_norm": 0.00042323453817516565,
+      "learning_rate": 2.6488454722381037e-05,
+      "loss": 0.0002,
+      "step": 1027
+    },
+    {
+      "epoch": 1.31488,
+      "grad_norm": 0.001034266664646566,
+      "learning_rate": 2.646437908148912e-05,
+      "loss": 0.0002,
+      "step": 1028
+    },
+    {
+      "epoch": 1.31616,
+      "grad_norm": 0.001303969300352037,
+      "learning_rate": 2.6440294927297185e-05,
+      "loss": 0.0002,
+      "step": 1029
+    },
+    {
+      "epoch": 1.31744,
+      "grad_norm": 0.00026662839809432626,
+      "learning_rate": 2.6416202305732248e-05,
+      "loss": 0.0002,
+      "step": 1030
+    },
+    {
+      "epoch": 1.31872,
+      "grad_norm": 0.0006337203085422516,
+      "learning_rate": 2.6392101262737474e-05,
+      "loss": 0.0002,
+      "step": 1031
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.00018651703430805355,
+      "learning_rate": 2.6367991844272095e-05,
+      "loss": 0.0002,
+      "step": 1032
+    },
+    {
+      "epoch": 1.32128,
+      "grad_norm": 0.0011202350724488497,
+      "learning_rate": 2.6343874096311308e-05,
+      "loss": 0.0002,
+      "step": 1033
+    },
+    {
+      "epoch": 1.32256,
+      "grad_norm": 0.001055483939126134,
+      "learning_rate": 2.631974806484619e-05,
+      "loss": 0.0002,
+      "step": 1034
+    },
+    {
+      "epoch": 1.3238400000000001,
+      "grad_norm": 0.00020614058303181082,
+      "learning_rate": 2.6295613795883624e-05,
+      "loss": 0.0002,
+      "step": 1035
+    },
+    {
+      "epoch": 1.32512,
+      "grad_norm": 0.0007595567149110138,
+      "learning_rate": 2.6271471335446194e-05,
+      "loss": 0.0002,
+      "step": 1036
+    },
+    {
+      "epoch": 1.3264,
+      "grad_norm": 0.0004523808602243662,
+      "learning_rate": 2.624732072957211e-05,
+      "loss": 0.0002,
+      "step": 1037
+    },
+    {
+      "epoch": 1.32768,
+      "grad_norm": 0.00021755503257736564,
+      "learning_rate": 2.62231620243151e-05,
+      "loss": 0.0002,
+      "step": 1038
+    },
+    {
+      "epoch": 1.32896,
+      "grad_norm": 0.00024666677927598357,
+      "learning_rate": 2.6198995265744374e-05,
+      "loss": 0.0002,
+      "step": 1039
+    },
+    {
+      "epoch": 1.3302399999999999,
+      "grad_norm": 0.0002588001952972263,
+      "learning_rate": 2.6174820499944446e-05,
+      "loss": 0.0002,
+      "step": 1040
+    },
+    {
+      "epoch": 1.33152,
+      "grad_norm": 0.00016227785090450197,
+      "learning_rate": 2.615063777301515e-05,
+      "loss": 0.0002,
+      "step": 1041
+    },
+    {
+      "epoch": 1.3328,
+      "grad_norm": 0.0006037519196979702,
+      "learning_rate": 2.6126447131071467e-05,
+      "loss": 0.0002,
+      "step": 1042
+    },
+    {
+      "epoch": 1.33408,
+      "grad_norm": 0.0009266502456739545,
+      "learning_rate": 2.610224862024349e-05,
+      "loss": 0.0002,
+      "step": 1043
+    },
+    {
+      "epoch": 1.33536,
+      "grad_norm": 0.00027118143043480814,
+      "learning_rate": 2.60780422866763e-05,
+      "loss": 0.0002,
+      "step": 1044
+    },
+    {
+      "epoch": 1.33664,
+      "grad_norm": 0.0013858125312253833,
+      "learning_rate": 2.6053828176529924e-05,
+      "loss": 0.0002,
+      "step": 1045
+    },
+    {
+      "epoch": 1.33792,
+      "grad_norm": 0.0015585527289658785,
+      "learning_rate": 2.6029606335979178e-05,
+      "loss": 0.0002,
+      "step": 1046
+    },
+    {
+      "epoch": 1.3392,
+      "grad_norm": 0.00019692144996952266,
+      "learning_rate": 2.600537681121366e-05,
+      "loss": 0.0002,
+      "step": 1047
+    },
+    {
+      "epoch": 1.34048,
+      "grad_norm": 0.0013894090661779046,
+      "learning_rate": 2.59811396484376e-05,
+      "loss": 0.0002,
+      "step": 1048
+    },
+    {
+      "epoch": 1.34176,
+      "grad_norm": 0.00041332849650643766,
+      "learning_rate": 2.595689489386979e-05,
+      "loss": 0.0002,
+      "step": 1049
+    },
+    {
+      "epoch": 1.34304,
+      "grad_norm": 0.001263888319954276,
+      "learning_rate": 2.593264259374352e-05,
+      "loss": 0.0002,
+      "step": 1050
+    },
+    {
+      "epoch": 1.34304,
+      "eval_loss": 1.1927013397216797,
+      "eval_runtime": 43.9413,
+      "eval_samples_per_second": 11.424,
+      "eval_steps_per_second": 1.434,
+      "step": 1050
+    },
+    {
+      "epoch": 1.34432,
+      "grad_norm": 0.001191023038700223,
+      "learning_rate": 2.5908382794306435e-05,
+      "loss": 0.0002,
+      "step": 1051
+    },
+    {
+      "epoch": 1.3456000000000001,
+      "grad_norm": 0.0004447832179721445,
+      "learning_rate": 2.5884115541820514e-05,
+      "loss": 0.0002,
+      "step": 1052
+    },
+    {
+      "epoch": 1.34688,
+      "grad_norm": 0.0015104867052286863,
+      "learning_rate": 2.585984088256193e-05,
+      "loss": 0.0002,
+      "step": 1053
+    },
+    {
+      "epoch": 1.34816,
+      "grad_norm": 0.0002054869692074135,
+      "learning_rate": 2.5835558862820997e-05,
+      "loss": 0.0002,
+      "step": 1054
+    },
+    {
+      "epoch": 1.34944,
+      "grad_norm": 0.0016293632797896862,
+      "learning_rate": 2.581126952890203e-05,
+      "loss": 0.0002,
+      "step": 1055
+    },
+    {
+      "epoch": 1.35072,
+      "grad_norm": 0.0007887612446211278,
+      "learning_rate": 2.5786972927123333e-05,
+      "loss": 0.0002,
+      "step": 1056
+    },
+    {
+      "epoch": 1.3519999999999999,
+      "grad_norm": 0.0021560071036219597,
+      "learning_rate": 2.576266910381705e-05,
+      "loss": 0.0002,
+      "step": 1057
+    },
+    {
+      "epoch": 1.35328,
+      "grad_norm": 0.003101208945736289,
+      "learning_rate": 2.573835810532909e-05,
+      "loss": 0.0002,
+      "step": 1058
+    },
+    {
+      "epoch": 1.35456,
+      "grad_norm": 0.0003672559978440404,
+      "learning_rate": 2.5714039978019062e-05,
+      "loss": 0.0002,
+      "step": 1059
+    },
+    {
+      "epoch": 1.35584,
+      "grad_norm": 0.003368792589753866,
+      "learning_rate": 2.568971476826015e-05,
+      "loss": 0.0002,
+      "step": 1060
+    },
+    {
+      "epoch": 1.35712,
+      "grad_norm": 0.0014343790244311094,
+      "learning_rate": 2.5665382522439053e-05,
+      "loss": 0.0002,
+      "step": 1061
+    },
+    {
+      "epoch": 1.3584,
+      "grad_norm": 0.00228679645806551,
+      "learning_rate": 2.5641043286955896e-05,
+      "loss": 0.0002,
+      "step": 1062
+    },
+    {
+      "epoch": 1.35968,
+      "grad_norm": 0.0023334973957389593,
+      "learning_rate": 2.5616697108224116e-05,
+      "loss": 0.0002,
+      "step": 1063
+    },
+    {
+      "epoch": 1.36096,
+      "grad_norm": 0.001103361020796001,
+      "learning_rate": 2.5592344032670406e-05,
+      "loss": 0.0002,
+      "step": 1064
+    },
+    {
+      "epoch": 1.36224,
+      "grad_norm": 0.0027988275978714228,
+      "learning_rate": 2.5567984106734597e-05,
+      "loss": 0.0002,
+      "step": 1065
+    },
+    {
+      "epoch": 1.36352,
+      "grad_norm": 0.00014616409316658974,
+      "learning_rate": 2.5543617376869584e-05,
+      "loss": 0.0002,
+      "step": 1066
+    },
+    {
+      "epoch": 1.3648,
+      "grad_norm": 0.0036822843831032515,
+      "learning_rate": 2.551924388954126e-05,
+      "loss": 0.0002,
+      "step": 1067
+    },
+    {
+      "epoch": 1.36608,
+      "grad_norm": 0.0036023741122335196,
+      "learning_rate": 2.549486369122837e-05,
+      "loss": 0.0002,
+      "step": 1068
+    },
+    {
+      "epoch": 1.3673600000000001,
+      "grad_norm": 0.00010829493112396449,
+      "learning_rate": 2.5470476828422482e-05,
+      "loss": 0.0002,
+      "step": 1069
+    },
+    {
+      "epoch": 1.36864,
+      "grad_norm": 0.003002248238772154,
+      "learning_rate": 2.5446083347627865e-05,
+      "loss": 0.0002,
+      "step": 1070
+    },
+    {
+      "epoch": 1.36992,
+      "grad_norm": 0.0024013391230255365,
+      "learning_rate": 2.5421683295361396e-05,
+      "loss": 0.0002,
+      "step": 1071
+    },
+    {
+      "epoch": 1.3712,
+      "grad_norm": 0.0012497822754085064,
+      "learning_rate": 2.5397276718152503e-05,
+      "loss": 0.0002,
+      "step": 1072
+    },
+    {
+      "epoch": 1.37248,
+      "grad_norm": 0.0032194824889302254,
+      "learning_rate": 2.5372863662543047e-05,
+      "loss": 0.0002,
+      "step": 1073
+    },
+    {
+      "epoch": 1.3737599999999999,
+      "grad_norm": 0.0007761308224871755,
+      "learning_rate": 2.534844417508724e-05,
+      "loss": 0.0002,
+      "step": 1074
+    },
+    {
+      "epoch": 1.37504,
+      "grad_norm": 0.0025674847420305014,
+      "learning_rate": 2.5324018302351564e-05,
+      "loss": 0.0002,
+      "step": 1075
+    },
+    {
+      "epoch": 1.37632,
+      "grad_norm": 0.0020916294306516647,
+      "learning_rate": 2.529958609091468e-05,
+      "loss": 0.0002,
+      "step": 1076
+    },
+    {
+      "epoch": 1.3776,
+      "grad_norm": 0.0012239001225680113,
+      "learning_rate": 2.5275147587367337e-05,
+      "loss": 0.0002,
+      "step": 1077
+    },
+    {
+      "epoch": 1.37888,
+      "grad_norm": 0.0021880425047129393,
+      "learning_rate": 2.5250702838312263e-05,
+      "loss": 0.0002,
+      "step": 1078
+    },
+    {
+      "epoch": 1.38016,
+      "grad_norm": 0.00032185690361075103,
+      "learning_rate": 2.5226251890364122e-05,
+      "loss": 0.0002,
+      "step": 1079
+    },
+    {
+      "epoch": 1.38144,
+      "grad_norm": 0.0026376349851489067,
+      "learning_rate": 2.5201794790149383e-05,
+      "loss": 0.0002,
+      "step": 1080
+    },
+    {
+      "epoch": 1.38144,
+      "eval_loss": 1.1927109956741333,
+      "eval_runtime": 43.8681,
+      "eval_samples_per_second": 11.443,
+      "eval_steps_per_second": 1.436,
+      "step": 1080
+    },
+    {
+      "epoch": 1.38272,
+      "grad_norm": 0.0016633147606626153,
+      "learning_rate": 2.517733158430626e-05,
+      "loss": 0.0002,
+      "step": 1081
+    },
+    {
+      "epoch": 1.384,
+      "grad_norm": 0.0013083289377391338,
+      "learning_rate": 2.5152862319484596e-05,
+      "loss": 0.0002,
+      "step": 1082
+    },
+    {
+      "epoch": 1.38528,
+      "grad_norm": 0.002328772097826004,
+      "learning_rate": 2.5128387042345792e-05,
+      "loss": 0.0002,
+      "step": 1083
+    },
+    {
+      "epoch": 1.38656,
+      "grad_norm": 0.00023164039885159582,
+      "learning_rate": 2.5103905799562723e-05,
+      "loss": 0.0002,
+      "step": 1084
+    },
+    {
+      "epoch": 1.38784,
+      "grad_norm": 0.0023417614866048098,
+      "learning_rate": 2.507941863781963e-05,
+      "loss": 0.0002,
+      "step": 1085
+    },
+    {
+      "epoch": 1.3891200000000001,
+      "grad_norm": 0.001744166831485927,
+      "learning_rate": 2.5054925603812042e-05,
+      "loss": 0.0002,
+      "step": 1086
+    },
+    {
+      "epoch": 1.3904,
+      "grad_norm": 0.0015349910827353597,
+      "learning_rate": 2.503042674424669e-05,
+      "loss": 0.0002,
+      "step": 1087
+    },
+    {
+      "epoch": 1.39168,
+      "grad_norm": 0.0033147349022328854,
+      "learning_rate": 2.500592210584143e-05,
+      "loss": 0.0002,
+      "step": 1088
+    },
+    {
+      "epoch": 1.39296,
+      "grad_norm": 0.0007261090795509517,
+      "learning_rate": 2.4981411735325088e-05,
+      "loss": 0.0002,
+      "step": 1089
+    },
+    {
+      "epoch": 1.39424,
+      "grad_norm": 0.003090859390795231,
+      "learning_rate": 2.495689567943748e-05,
+      "loss": 0.0002,
+      "step": 1090
+    },
+    {
+      "epoch": 1.3955199999999999,
+      "grad_norm": 0.002288298448547721,
+      "learning_rate": 2.493237398492922e-05,
+      "loss": 0.0002,
+      "step": 1091
+    },
+    {
+      "epoch": 1.3968,
+      "grad_norm": 0.0014610658399760723,
+      "learning_rate": 2.4907846698561704e-05,
+      "loss": 0.0002,
+      "step": 1092
+    },
+    {
+      "epoch": 1.39808,
+      "grad_norm": 0.002210103441029787,
+      "learning_rate": 2.488331386710697e-05,
+      "loss": 0.0002,
+      "step": 1093
+    },
+    {
+      "epoch": 1.39936,
+      "grad_norm": 0.00047967262798920274,
+      "learning_rate": 2.4858775537347654e-05,
+      "loss": 0.0002,
+      "step": 1094
+    },
+    {
+      "epoch": 1.40064,
+      "grad_norm": 0.0018254382302984595,
+      "learning_rate": 2.4834231756076833e-05,
+      "loss": 0.0002,
+      "step": 1095
+    },
+    {
+      "epoch": 1.40192,
+      "grad_norm": 0.00032902159728109837,
+      "learning_rate": 2.480968257009804e-05,
+      "loss": 0.0002,
+      "step": 1096
+    },
+    {
+      "epoch": 1.4032,
+      "grad_norm": 0.0019898158498108387,
+      "learning_rate": 2.4785128026225062e-05,
+      "loss": 0.0002,
+      "step": 1097
+    },
+    {
+      "epoch": 1.40448,
+      "grad_norm": 0.0008954960503615439,
+      "learning_rate": 2.4760568171281934e-05,
+      "loss": 0.0002,
+      "step": 1098
+    },
+    {
+      "epoch": 1.40576,
+      "grad_norm": 0.0008229295490309596,
+      "learning_rate": 2.4736003052102816e-05,
+      "loss": 0.0002,
+      "step": 1099
+    },
+    {
+      "epoch": 1.40704,
+      "grad_norm": 0.0005774835590273142,
+      "learning_rate": 2.4711432715531874e-05,
+      "loss": 0.0002,
+      "step": 1100
+    },
+    {
+      "epoch": 1.40832,
+      "grad_norm": 0.0005937288515269756,
+      "learning_rate": 2.4686857208423276e-05,
+      "loss": 0.0002,
+      "step": 1101
+    },
+    {
+      "epoch": 1.4096,
+      "grad_norm": 0.00021285265393089503,
+      "learning_rate": 2.466227657764101e-05,
+      "loss": 0.0002,
+      "step": 1102
+    },
+    {
+      "epoch": 1.4108800000000001,
+      "grad_norm": 0.000575678946916014,
+      "learning_rate": 2.463769087005885e-05,
+      "loss": 0.0002,
+      "step": 1103
+    },
+    {
+      "epoch": 1.41216,
+      "grad_norm": 0.0004997372161597013,
+      "learning_rate": 2.461310013256025e-05,
+      "loss": 0.0002,
+      "step": 1104
+    },
+    {
+      "epoch": 1.41344,
+      "grad_norm": 0.00018097950669471174,
+      "learning_rate": 2.4588504412038253e-05,
+      "loss": 0.0002,
+      "step": 1105
+    },
+    {
+      "epoch": 1.41472,
+      "grad_norm": 0.0005268906243145466,
+      "learning_rate": 2.45639037553954e-05,
+      "loss": 0.0002,
+      "step": 1106
+    },
+    {
+      "epoch": 1.416,
+      "grad_norm": 0.00013170884631108493,
+      "learning_rate": 2.453929820954366e-05,
+      "loss": 0.0002,
+      "step": 1107
+    },
+    {
+      "epoch": 1.4172799999999999,
+      "grad_norm": 0.00040802324656397104,
+      "learning_rate": 2.451468782140431e-05,
+      "loss": 0.0002,
+      "step": 1108
+    },
+    {
+      "epoch": 1.41856,
+      "grad_norm": 0.0009418302797712386,
+      "learning_rate": 2.4490072637907873e-05,
+      "loss": 0.0002,
+      "step": 1109
+    },
+    {
+      "epoch": 1.41984,
+      "grad_norm": 0.0011874447809532285,
+      "learning_rate": 2.4465452705994e-05,
+      "loss": 0.0002,
+      "step": 1110
+    },
+    {
+      "epoch": 1.41984,
+      "eval_loss": 1.1833388805389404,
+      "eval_runtime": 43.5138,
+      "eval_samples_per_second": 11.537,
+      "eval_steps_per_second": 1.448,
+      "step": 1110
+    },
+    {
+      "epoch": 1.42112,
+      "grad_norm": 0.0001316447596764192,
+      "learning_rate": 2.444082807261142e-05,
+      "loss": 0.0002,
+      "step": 1111
+    },
+    {
+      "epoch": 1.4224,
+      "grad_norm": 0.0008458861848339438,
+      "learning_rate": 2.44161987847178e-05,
+      "loss": 0.0002,
+      "step": 1112
+    },
+    {
+      "epoch": 1.42368,
+      "grad_norm": 0.0002578686107881367,
+      "learning_rate": 2.4391564889279712e-05,
+      "loss": 0.0002,
+      "step": 1113
+    },
+    {
+      "epoch": 1.42496,
+      "grad_norm": 0.001641182112507522,
+      "learning_rate": 2.4366926433272494e-05,
+      "loss": 0.0002,
+      "step": 1114
+    },
+    {
+      "epoch": 1.42624,
+      "grad_norm": 0.0012906165793538094,
+      "learning_rate": 2.434228346368018e-05,
+      "loss": 0.0002,
+      "step": 1115
+    },
+    {
+      "epoch": 1.42752,
+      "grad_norm": 0.0002661029575392604,
+      "learning_rate": 2.4317636027495426e-05,
+      "loss": 0.0002,
+      "step": 1116
+    },
+    {
+      "epoch": 1.4288,
+      "grad_norm": 0.0005313612637110054,
+      "learning_rate": 2.42929841717194e-05,
+      "loss": 0.0002,
+      "step": 1117
+    },
+    {
+      "epoch": 1.43008,
+      "grad_norm": 0.0009100149036385119,
+      "learning_rate": 2.4268327943361688e-05,
+      "loss": 0.0002,
+      "step": 1118
+    },
+    {
+      "epoch": 1.43136,
+      "grad_norm": 0.0018073428655043244,
+      "learning_rate": 2.4243667389440222e-05,
+      "loss": 0.0002,
+      "step": 1119
+    },
+    {
+      "epoch": 1.4326400000000001,
+      "grad_norm": 0.0005083298892714083,
+      "learning_rate": 2.4219002556981184e-05,
+      "loss": 0.0002,
+      "step": 1120
+    },
+    {
+      "epoch": 1.43392,
+      "grad_norm": 0.0010776943527162075,
+      "learning_rate": 2.4194333493018912e-05,
+      "loss": 0.0002,
+      "step": 1121
+    },
+    {
+      "epoch": 1.4352,
+      "grad_norm": 0.00042680808110162616,
+      "learning_rate": 2.4169660244595818e-05,
+      "loss": 0.0002,
+      "step": 1122
+    },
+    {
+      "epoch": 1.43648,
+      "grad_norm": 0.0024712400045245886,
+      "learning_rate": 2.414498285876228e-05,
+      "loss": 0.0002,
+      "step": 1123
+    },
+    {
+      "epoch": 1.43776,
+      "grad_norm": 0.0015264041721820831,
+      "learning_rate": 2.4120301382576593e-05,
+      "loss": 0.0002,
+      "step": 1124
+    },
+    {
+      "epoch": 1.4390399999999999,
+      "grad_norm": 0.0009439815767109394,
+      "learning_rate": 2.4095615863104812e-05,
+      "loss": 0.0002,
+      "step": 1125
+    },
+    {
+      "epoch": 1.44032,
+      "grad_norm": 0.0013193822233006358,
+      "learning_rate": 2.4070926347420746e-05,
+      "loss": 0.0002,
+      "step": 1126
+    },
+    {
+      "epoch": 1.4416,
+      "grad_norm": 0.000317655794788152,
+      "learning_rate": 2.404623288260578e-05,
+      "loss": 0.0002,
+      "step": 1127
+    },
+    {
+      "epoch": 1.44288,
+      "grad_norm": 0.0012924104230478406,
+      "learning_rate": 2.402153551574888e-05,
+      "loss": 0.0002,
+      "step": 1128
+    },
+    {
+      "epoch": 1.44416,
+      "grad_norm": 0.0003663348325062543,
+      "learning_rate": 2.399683429394641e-05,
+      "loss": 0.0002,
+      "step": 1129
+    },
+    {
+      "epoch": 1.44544,
+      "grad_norm": 0.0015283627435564995,
+      "learning_rate": 2.3972129264302103e-05,
+      "loss": 0.0002,
+      "step": 1130
+    },
+    {
+      "epoch": 1.44672,
+      "grad_norm": 0.0014445302076637745,
+      "learning_rate": 2.3947420473926943e-05,
+      "loss": 0.0002,
+      "step": 1131
+    },
+    {
+      "epoch": 1.448,
+      "grad_norm": 0.0007018298492766917,
+      "learning_rate": 2.3922707969939115e-05,
+      "loss": 0.0002,
+      "step": 1132
+    },
+    {
+      "epoch": 1.44928,
+      "grad_norm": 0.0018520744051784277,
+      "learning_rate": 2.3897991799463847e-05,
+      "loss": 0.0002,
+      "step": 1133
+    },
+    {
+      "epoch": 1.45056,
+      "grad_norm": 0.00043054608977399766,
+      "learning_rate": 2.387327200963339e-05,
+      "loss": 0.0002,
+      "step": 1134
+    },
+    {
+      "epoch": 1.45184,
+      "grad_norm": 0.0013373054098337889,
+      "learning_rate": 2.384854864758686e-05,
+      "loss": 0.0002,
+      "step": 1135
+    },
+    {
+      "epoch": 1.45312,
+      "grad_norm": 0.00040695464122109115,
+      "learning_rate": 2.3823821760470237e-05,
+      "loss": 0.0002,
+      "step": 1136
+    },
+    {
+      "epoch": 1.4544000000000001,
+      "grad_norm": 0.0019905297085642815,
+      "learning_rate": 2.3799091395436172e-05,
+      "loss": 0.0002,
+      "step": 1137
+    },
+    {
+      "epoch": 1.45568,
+      "grad_norm": 0.00207939138635993,
+      "learning_rate": 2.3774357599643985e-05,
+      "loss": 0.0002,
+      "step": 1138
+    },
+    {
+      "epoch": 1.45696,
+      "grad_norm": 0.00039109544013626873,
+      "learning_rate": 2.374962042025952e-05,
+      "loss": 0.0002,
+      "step": 1139
+    },
+    {
+      "epoch": 1.45824,
+      "grad_norm": 0.0014972800854593515,
+      "learning_rate": 2.3724879904455072e-05,
+      "loss": 0.0002,
+      "step": 1140
+    },
+    {
+      "epoch": 1.45824,
+      "eval_loss": 1.1820496320724487,
+      "eval_runtime": 43.3596,
+      "eval_samples_per_second": 11.578,
+      "eval_steps_per_second": 1.453,
+      "step": 1140
+    },
+    {
+      "epoch": 1.45952,
+      "grad_norm": 0.00017591100186109543,
+      "learning_rate": 2.370013609940931e-05,
+      "loss": 0.0002,
+      "step": 1141
+    },
+    {
+      "epoch": 1.4607999999999999,
+      "grad_norm": 0.002228839322924614,
+      "learning_rate": 2.3675389052307165e-05,
+      "loss": 0.0002,
+      "step": 1142
+    },
+    {
+      "epoch": 1.46208,
+      "grad_norm": 0.0017025255365297198,
+      "learning_rate": 2.3650638810339755e-05,
+      "loss": 0.0002,
+      "step": 1143
+    },
+    {
+      "epoch": 1.46336,
+      "grad_norm": 0.0016512767178937793,
+      "learning_rate": 2.3625885420704283e-05,
+      "loss": 0.0002,
+      "step": 1144
+    },
+    {
+      "epoch": 1.46464,
+      "grad_norm": 0.0030168932862579823,
+      "learning_rate": 2.360112893060398e-05,
+      "loss": 0.0002,
+      "step": 1145
+    },
+    {
+      "epoch": 1.4659200000000001,
+      "grad_norm": 0.00015002665168140084,
+      "learning_rate": 2.357636938724795e-05,
+      "loss": 0.0002,
+      "step": 1146
+    },
+    {
+      "epoch": 1.4672,
+      "grad_norm": 0.0025957797188311815,
+      "learning_rate": 2.355160683785115e-05,
+      "loss": 0.0002,
+      "step": 1147
+    },
+    {
+      "epoch": 1.46848,
+      "grad_norm": 0.0005568786873482168,
+      "learning_rate": 2.3526841329634258e-05,
+      "loss": 0.0002,
+      "step": 1148
+    },
+    {
+      "epoch": 1.46976,
+      "grad_norm": 0.0029370211996138096,
+      "learning_rate": 2.3502072909823598e-05,
+      "loss": 0.0002,
+      "step": 1149
+    },
+    {
+      "epoch": 1.47104,
+      "grad_norm": 0.001947471289895475,
+      "learning_rate": 2.3477301625651033e-05,
+      "loss": 0.0002,
+      "step": 1150
+    },
+    {
+      "epoch": 1.47232,
+      "grad_norm": 0.0019220055546611547,
+      "learning_rate": 2.3452527524353913e-05,
+      "loss": 0.0002,
+      "step": 1151
+    },
+    {
+      "epoch": 1.4736,
+      "grad_norm": 0.0024576217401772738,
+      "learning_rate": 2.342775065317494e-05,
+      "loss": 0.0002,
+      "step": 1152
+    },
+    {
+      "epoch": 1.47488,
+      "grad_norm": 0.0007621421827934682,
+      "learning_rate": 2.34029710593621e-05,
+      "loss": 0.0002,
+      "step": 1153
+    },
+    {
+      "epoch": 1.4761600000000001,
+      "grad_norm": 0.002741249743849039,
+      "learning_rate": 2.3378188790168576e-05,
+      "loss": 0.0002,
+      "step": 1154
+    },
+    {
+      "epoch": 1.47744,
+      "grad_norm": 0.0012215838069096208,
+      "learning_rate": 2.335340389285266e-05,
+      "loss": 0.0002,
+      "step": 1155
+    },
+    {
+      "epoch": 1.47872,
+      "grad_norm": 0.0014048725133761764,
+      "learning_rate": 2.3328616414677633e-05,
+      "loss": 0.0002,
+      "step": 1156
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.0012951638782396913,
+      "learning_rate": 2.330382640291173e-05,
+      "loss": 0.0002,
+      "step": 1157
+    },
+    {
+      "epoch": 1.48128,
+      "grad_norm": 0.001137186773121357,
+      "learning_rate": 2.3279033904827983e-05,
+      "loss": 0.0002,
+      "step": 1158
+    },
+    {
+      "epoch": 1.4825599999999999,
+      "grad_norm": 0.0011690363753587008,
+      "learning_rate": 2.3254238967704184e-05,
+      "loss": 0.0002,
+      "step": 1159
+    },
+    {
+      "epoch": 1.48384,
+      "grad_norm": 0.0013390486128628254,
+      "learning_rate": 2.3229441638822783e-05,
+      "loss": 0.0002,
+      "step": 1160
+    },
+    {
+      "epoch": 1.48512,
+      "grad_norm": 0.001537360018119216,
+      "learning_rate": 2.3204641965470773e-05,
+      "loss": 0.0002,
+      "step": 1161
+    },
+    {
+      "epoch": 1.4864,
+      "grad_norm": 0.000685350620187819,
+      "learning_rate": 2.317983999493963e-05,
+      "loss": 0.0002,
+      "step": 1162
+    },
+    {
+      "epoch": 1.4876800000000001,
+      "grad_norm": 0.0017830547876656055,
+      "learning_rate": 2.3155035774525206e-05,
+      "loss": 0.0002,
+      "step": 1163
+    },
+    {
+      "epoch": 1.48896,
+      "grad_norm": 0.0004691873327828944,
+      "learning_rate": 2.3130229351527645e-05,
+      "loss": 0.0002,
+      "step": 1164
+    },
+    {
+      "epoch": 1.49024,
+      "grad_norm": 0.001537265139631927,
+      "learning_rate": 2.310542077325129e-05,
+      "loss": 0.0002,
+      "step": 1165
+    },
+    {
+      "epoch": 1.49152,
+      "grad_norm": 0.0013536819024011493,
+      "learning_rate": 2.308061008700459e-05,
+      "loss": 0.0002,
+      "step": 1166
+    },
+    {
+      "epoch": 1.4928,
+      "grad_norm": 0.00046399055281654,
+      "learning_rate": 2.3055797340100024e-05,
+      "loss": 0.0002,
+      "step": 1167
+    },
+    {
+      "epoch": 1.49408,
+      "grad_norm": 0.0013303180458024144,
+      "learning_rate": 2.3030982579853996e-05,
+      "loss": 0.0002,
+      "step": 1168
+    },
+    {
+      "epoch": 1.49536,
+      "grad_norm": 0.000706115213688463,
+      "learning_rate": 2.3006165853586747e-05,
+      "loss": 0.0002,
+      "step": 1169
+    },
+    {
+      "epoch": 1.49664,
+      "grad_norm": 0.00045828090514987707,
+      "learning_rate": 2.298134720862227e-05,
+      "loss": 0.0002,
+      "step": 1170
+    },
+    {
+      "epoch": 1.49664,
+      "eval_loss": 1.1880303621292114,
+      "eval_runtime": 43.6913,
+      "eval_samples_per_second": 11.49,
+      "eval_steps_per_second": 1.442,
+      "step": 1170
+    },
+    {
+      "epoch": 1.49792,
+      "grad_norm": 0.0005636084824800491,
+      "learning_rate": 2.2956526692288197e-05,
+      "loss": 0.0002,
+      "step": 1171
+    },
+    {
+      "epoch": 1.4992,
+      "grad_norm": 0.0004983227117918432,
+      "learning_rate": 2.2931704351915776e-05,
+      "loss": 0.0002,
+      "step": 1172
+    },
+    {
+      "epoch": 1.50048,
+      "grad_norm": 0.0012799539836123586,
+      "learning_rate": 2.2906880234839676e-05,
+      "loss": 0.0002,
+      "step": 1173
+    },
+    {
+      "epoch": 1.50176,
+      "grad_norm": 0.0003243404207751155,
+      "learning_rate": 2.2882054388398006e-05,
+      "loss": 0.0002,
+      "step": 1174
+    },
+    {
+      "epoch": 1.50304,
+      "grad_norm": 0.0009844460291787982,
+      "learning_rate": 2.2857226859932134e-05,
+      "loss": 0.0002,
+      "step": 1175
+    },
+    {
+      "epoch": 1.5043199999999999,
+      "grad_norm": 0.00018364368588663638,
+      "learning_rate": 2.2832397696786653e-05,
+      "loss": 0.0002,
+      "step": 1176
+    },
+    {
+      "epoch": 1.5056,
+      "grad_norm": 0.001054965308867395,
+      "learning_rate": 2.2807566946309273e-05,
+      "loss": 0.0002,
+      "step": 1177
+    },
+    {
+      "epoch": 1.50688,
+      "grad_norm": 0.00043028456275351346,
+      "learning_rate": 2.2782734655850727e-05,
+      "loss": 0.0002,
+      "step": 1178
+    },
+    {
+      "epoch": 1.50816,
+      "grad_norm": 0.0005619335570372641,
+      "learning_rate": 2.275790087276468e-05,
+      "loss": 0.0002,
+      "step": 1179
+    },
+    {
+      "epoch": 1.5094400000000001,
+      "grad_norm": 0.0001896164903882891,
+      "learning_rate": 2.273306564440767e-05,
+      "loss": 0.0002,
+      "step": 1180
+    },
+    {
+      "epoch": 1.51072,
+      "grad_norm": 0.000446715101134032,
+      "learning_rate": 2.2708229018138946e-05,
+      "loss": 0.0002,
+      "step": 1181
+    },
+    {
+      "epoch": 1.512,
+      "grad_norm": 0.0004963649553246796,
+      "learning_rate": 2.268339104132046e-05,
+      "loss": 0.0002,
+      "step": 1182
+    },
+    {
+      "epoch": 1.51328,
+      "grad_norm": 0.0008193458197638392,
+      "learning_rate": 2.2658551761316723e-05,
+      "loss": 0.0002,
+      "step": 1183
+    },
+    {
+      "epoch": 1.51456,
+      "grad_norm": 0.0004890802665613592,
+      "learning_rate": 2.263371122549474e-05,
+      "loss": 0.0002,
+      "step": 1184
+    },
+    {
+      "epoch": 1.5158399999999999,
+      "grad_norm": 0.0001344592310488224,
+      "learning_rate": 2.2608869481223898e-05,
+      "loss": 0.0002,
+      "step": 1185
+    },
+    {
+      "epoch": 1.51712,
+      "grad_norm": 0.00012116249126847833,
+      "learning_rate": 2.2584026575875902e-05,
+      "loss": 0.0002,
+      "step": 1186
+    },
+    {
+      "epoch": 1.5184,
+      "grad_norm": 0.0006150749395601451,
+      "learning_rate": 2.2559182556824667e-05,
+      "loss": 0.0002,
+      "step": 1187
+    },
+    {
+      "epoch": 1.5196800000000001,
+      "grad_norm": 0.0006445617182180285,
+      "learning_rate": 2.2534337471446223e-05,
+      "loss": 0.0002,
+      "step": 1188
+    },
+    {
+      "epoch": 1.52096,
+      "grad_norm": 0.00024265323008876294,
+      "learning_rate": 2.2509491367118642e-05,
+      "loss": 0.0002,
+      "step": 1189
+    },
+    {
+      "epoch": 1.52224,
+      "grad_norm": 0.0001686473551671952,
+      "learning_rate": 2.2484644291221933e-05,
+      "loss": 0.0002,
+      "step": 1190
+    },
+    {
+      "epoch": 1.52352,
+      "grad_norm": 0.0006335959769785404,
+      "learning_rate": 2.245979629113797e-05,
+      "loss": 0.0002,
+      "step": 1191
+    },
+    {
+      "epoch": 1.5248,
+      "grad_norm": 0.0006540421163663268,
+      "learning_rate": 2.2434947414250387e-05,
+      "loss": 0.0002,
+      "step": 1192
+    },
+    {
+      "epoch": 1.5260799999999999,
+      "grad_norm": 0.00016801041783764958,
+      "learning_rate": 2.241009770794447e-05,
+      "loss": 0.0002,
+      "step": 1193
+    },
+    {
+      "epoch": 1.52736,
+      "grad_norm": 0.0006320360698737204,
+      "learning_rate": 2.238524721960711e-05,
+      "loss": 0.0002,
+      "step": 1194
+    },
+    {
+      "epoch": 1.52864,
+      "grad_norm": 0.0003093292471021414,
+      "learning_rate": 2.236039599662667e-05,
+      "loss": 0.0002,
+      "step": 1195
+    },
+    {
+      "epoch": 1.52992,
+      "grad_norm": 0.0003343690186738968,
+      "learning_rate": 2.233554408639294e-05,
+      "loss": 0.0002,
+      "step": 1196
+    },
+    {
+      "epoch": 1.5312000000000001,
+      "grad_norm": 0.00028043152997270226,
+      "learning_rate": 2.2310691536296995e-05,
+      "loss": 0.0002,
+      "step": 1197
+    },
+    {
+      "epoch": 1.53248,
+      "grad_norm": 0.000370222725905478,
+      "learning_rate": 2.2285838393731146e-05,
+      "loss": 0.0002,
+      "step": 1198
+    },
+    {
+      "epoch": 1.53376,
+      "grad_norm": 0.00022848798835184425,
+      "learning_rate": 2.226098470608882e-05,
+      "loss": 0.0002,
+      "step": 1199
+    },
+    {
+      "epoch": 1.53504,
+      "grad_norm": 0.0006417376571334898,
+      "learning_rate": 2.2236130520764493e-05,
+      "loss": 0.0002,
+      "step": 1200
+    },
+    {
+      "epoch": 1.53504,
+      "eval_loss": 1.18635094165802,
+      "eval_runtime": 43.537,
+      "eval_samples_per_second": 11.53,
+      "eval_steps_per_second": 1.447,
+      "step": 1200
+    },
+    {
+      "epoch": 1.53632,
+      "grad_norm": 0.0005315435701049864,
+      "learning_rate": 2.2211275885153594e-05,
+      "loss": 0.0002,
+      "step": 1201
+    },
+    {
+      "epoch": 1.5375999999999999,
+      "grad_norm": 0.0007035636226646602,
+      "learning_rate": 2.21864208466524e-05,
+      "loss": 0.0002,
+      "step": 1202
+    },
+    {
+      "epoch": 1.53888,
+      "grad_norm": 0.0015890513313934207,
+      "learning_rate": 2.2161565452657964e-05,
+      "loss": 0.0002,
+      "step": 1203
+    },
+    {
+      "epoch": 1.54016,
+      "grad_norm": 0.0005806463886983693,
+      "learning_rate": 2.2136709750568023e-05,
+      "loss": 0.0002,
+      "step": 1204
+    },
+    {
+      "epoch": 1.5414400000000001,
+      "grad_norm": 0.0012760835234075785,
+      "learning_rate": 2.2111853787780864e-05,
+      "loss": 0.0002,
+      "step": 1205
+    },
+    {
+      "epoch": 1.54272,
+      "grad_norm": 0.0006499637383967638,
+      "learning_rate": 2.208699761169533e-05,
+      "loss": 0.0002,
+      "step": 1206
+    },
+    {
+      "epoch": 1.544,
+      "grad_norm": 0.0016380108427256346,
+      "learning_rate": 2.2062141269710627e-05,
+      "loss": 0.0002,
+      "step": 1207
+    },
+    {
+      "epoch": 1.54528,
+      "grad_norm": 0.001415404723957181,
+      "learning_rate": 2.2037284809226295e-05,
+      "loss": 0.0002,
+      "step": 1208
+    },
+    {
+      "epoch": 1.54656,
+      "grad_norm": 0.0009337143274024129,
+      "learning_rate": 2.2012428277642085e-05,
+      "loss": 0.0002,
+      "step": 1209
+    },
+    {
+      "epoch": 1.5478399999999999,
+      "grad_norm": 0.0014679753221571445,
+      "learning_rate": 2.198757172235792e-05,
+      "loss": 0.0002,
+      "step": 1210
+    },
+    {
+      "epoch": 1.54912,
+      "grad_norm": 0.00028598069911822677,
+      "learning_rate": 2.196271519077371e-05,
+      "loss": 0.0002,
+      "step": 1211
+    },
+    {
+      "epoch": 1.5504,
+      "grad_norm": 0.0016987015260383487,
+      "learning_rate": 2.193785873028938e-05,
+      "loss": 0.0002,
+      "step": 1212
+    },
+    {
+      "epoch": 1.55168,
+      "grad_norm": 0.0011778499465435743,
+      "learning_rate": 2.191300238830468e-05,
+      "loss": 0.0002,
+      "step": 1213
+    },
+    {
+      "epoch": 1.5529600000000001,
+      "grad_norm": 0.0002532570797484368,
+      "learning_rate": 2.1888146212219142e-05,
+      "loss": 0.0002,
+      "step": 1214
+    },
+    {
+      "epoch": 1.55424,
+      "grad_norm": 0.0004746295453514904,
+      "learning_rate": 2.1863290249431993e-05,
+      "loss": 0.0002,
+      "step": 1215
+    },
+    {
+      "epoch": 1.55552,
+      "grad_norm": 0.00015324894047807902,
+      "learning_rate": 2.183843454734204e-05,
+      "loss": 0.0002,
+      "step": 1216
+    },
+    {
+      "epoch": 1.5568,
+      "grad_norm": 0.00023059602244757116,
+      "learning_rate": 2.1813579153347604e-05,
+      "loss": 0.0002,
+      "step": 1217
+    },
+    {
+      "epoch": 1.55808,
+      "grad_norm": 0.00017352044233120978,
+      "learning_rate": 2.1788724114846408e-05,
+      "loss": 0.0002,
+      "step": 1218
+    },
+    {
+      "epoch": 1.5593599999999999,
+      "grad_norm": 0.0008680567843839526,
+      "learning_rate": 2.1763869479235512e-05,
+      "loss": 0.0002,
+      "step": 1219
+    },
+    {
+      "epoch": 1.56064,
+      "grad_norm": 0.0011944602010771632,
+      "learning_rate": 2.173901529391119e-05,
+      "loss": 0.0002,
+      "step": 1220
+    },
+    {
+      "epoch": 1.56192,
+      "grad_norm": 0.0003440252039581537,
+      "learning_rate": 2.1714161606268863e-05,
+      "loss": 0.0002,
+      "step": 1221
+    },
+    {
+      "epoch": 1.5632000000000001,
+      "grad_norm": 0.001474754768423736,
+      "learning_rate": 2.168930846370301e-05,
+      "loss": 0.0002,
+      "step": 1222
+    },
+    {
+      "epoch": 1.56448,
+      "grad_norm": 0.0006110747344791889,
+      "learning_rate": 2.1664455913607063e-05,
+      "loss": 0.0002,
+      "step": 1223
+    },
+    {
+      "epoch": 1.56576,
+      "grad_norm": 0.00047909640124998987,
+      "learning_rate": 2.1639604003373332e-05,
+      "loss": 0.0002,
+      "step": 1224
+    },
+    {
+      "epoch": 1.56704,
+      "grad_norm": 0.0001748676149873063,
+      "learning_rate": 2.16147527803929e-05,
+      "loss": 0.0002,
+      "step": 1225
+    },
+    {
+      "epoch": 1.56832,
+      "grad_norm": 0.000838764535728842,
+      "learning_rate": 2.1589902292055534e-05,
+      "loss": 0.0002,
+      "step": 1226
+    },
+    {
+      "epoch": 1.5695999999999999,
+      "grad_norm": 0.0002627510402817279,
+      "learning_rate": 2.1565052585749626e-05,
+      "loss": 0.0002,
+      "step": 1227
+    },
+    {
+      "epoch": 1.57088,
+      "grad_norm": 0.0006864861934445798,
+      "learning_rate": 2.154020370886203e-05,
+      "loss": 0.0002,
+      "step": 1228
+    },
+    {
+      "epoch": 1.57216,
+      "grad_norm": 0.00024974255939014256,
+      "learning_rate": 2.1515355708778072e-05,
+      "loss": 0.0002,
+      "step": 1229
+    },
+    {
+      "epoch": 1.57344,
+      "grad_norm": 0.0008880963432602584,
+      "learning_rate": 2.149050863288137e-05,
+      "loss": 0.0002,
+      "step": 1230
+    },
+    {
+      "epoch": 1.57344,
+      "eval_loss": 1.184080719947815,
+      "eval_runtime": 43.5512,
+      "eval_samples_per_second": 11.527,
+      "eval_steps_per_second": 1.447,
+      "step": 1230
+    },
+    {
+      "epoch": 1.5747200000000001,
+      "grad_norm": 0.0002939118421636522,
+      "learning_rate": 2.1465662528553787e-05,
+      "loss": 0.0002,
+      "step": 1231
+    },
+    {
+      "epoch": 1.576,
+      "grad_norm": 0.001100562745705247,
+      "learning_rate": 2.1440817443175342e-05,
+      "loss": 0.0002,
+      "step": 1232
+    },
+    {
+      "epoch": 1.57728,
+      "grad_norm": 0.0009249002323485911,
+      "learning_rate": 2.14159734241241e-05,
+      "loss": 0.0002,
+      "step": 1233
+    },
+    {
+      "epoch": 1.57856,
+      "grad_norm": 0.0005277720047160983,
+      "learning_rate": 2.1391130518776104e-05,
+      "loss": 0.0002,
+      "step": 1234
+    },
+    {
+      "epoch": 1.57984,
+      "grad_norm": 0.0010325635084882379,
+      "learning_rate": 2.1366288774505263e-05,
+      "loss": 0.0002,
+      "step": 1235
+    },
+    {
+      "epoch": 1.5811199999999999,
+      "grad_norm": 0.00023603586305398494,
+      "learning_rate": 2.134144823868328e-05,
+      "loss": 0.0002,
+      "step": 1236
+    },
+    {
+      "epoch": 1.5824,
+      "grad_norm": 0.0016086595132946968,
+      "learning_rate": 2.1316608958679547e-05,
+      "loss": 0.0002,
+      "step": 1237
+    },
+    {
+      "epoch": 1.58368,
+      "grad_norm": 0.0009198304614983499,
+      "learning_rate": 2.129177098186106e-05,
+      "loss": 0.0002,
+      "step": 1238
+    },
+    {
+      "epoch": 1.5849600000000001,
+      "grad_norm": 0.0010304874740540981,
+      "learning_rate": 2.1266934355592337e-05,
+      "loss": 0.0002,
+      "step": 1239
+    },
+    {
+      "epoch": 1.58624,
+      "grad_norm": 0.0009441455476917326,
+      "learning_rate": 2.124209912723532e-05,
+      "loss": 0.0002,
+      "step": 1240
+    },
+    {
+      "epoch": 1.58752,
+      "grad_norm": 0.0005426053539849818,
+      "learning_rate": 2.121726534414928e-05,
+      "loss": 0.0002,
+      "step": 1241
+    },
+    {
+      "epoch": 1.5888,
+      "grad_norm": 0.00025539376656524837,
+      "learning_rate": 2.1192433053690743e-05,
+      "loss": 0.0002,
+      "step": 1242
+    },
+    {
+      "epoch": 1.59008,
+      "grad_norm": 0.0010780163574963808,
+      "learning_rate": 2.1167602303213352e-05,
+      "loss": 0.0002,
+      "step": 1243
+    },
+    {
+      "epoch": 1.5913599999999999,
+      "grad_norm": 0.00113013107329607,
+      "learning_rate": 2.114277314006788e-05,
+      "loss": 0.0002,
+      "step": 1244
+    },
+    {
+      "epoch": 1.5926399999999998,
+      "grad_norm": 0.00017947869491763413,
+      "learning_rate": 2.1117945611602e-05,
+      "loss": 0.0002,
+      "step": 1245
+    },
+    {
+      "epoch": 1.59392,
+      "grad_norm": 0.001032519736327231,
+      "learning_rate": 2.109311976516033e-05,
+      "loss": 0.0002,
+      "step": 1246
+    },
+    {
+      "epoch": 1.5952,
+      "grad_norm": 0.0004488160484470427,
+      "learning_rate": 2.1068295648084233e-05,
+      "loss": 0.0002,
+      "step": 1247
+    },
+    {
+      "epoch": 1.5964800000000001,
+      "grad_norm": 0.0001314591063419357,
+      "learning_rate": 2.1043473307711805e-05,
+      "loss": 0.0002,
+      "step": 1248
+    },
+    {
+      "epoch": 1.59776,
+      "grad_norm": 0.0003850892826449126,
+      "learning_rate": 2.1018652791377744e-05,
+      "loss": 0.0002,
+      "step": 1249
+    },
+    {
+      "epoch": 1.59904,
+      "grad_norm": 0.00025546064716763794,
+      "learning_rate": 2.099383414641326e-05,
+      "loss": 0.0002,
+      "step": 1250
+    },
+    {
+      "epoch": 1.60032,
+      "grad_norm": 0.0003797194513026625,
+      "learning_rate": 2.096901742014601e-05,
+      "loss": 0.0002,
+      "step": 1251
+    },
+    {
+      "epoch": 1.6016,
+      "grad_norm": 0.00018792178889270872,
+      "learning_rate": 2.0944202659899978e-05,
+      "loss": 0.0002,
+      "step": 1252
+    },
+    {
+      "epoch": 1.6028799999999999,
+      "grad_norm": 0.0005130458157509565,
+      "learning_rate": 2.0919389912995416e-05,
+      "loss": 0.0002,
+      "step": 1253
+    },
+    {
+      "epoch": 1.60416,
+      "grad_norm": 0.00027721276273950934,
+      "learning_rate": 2.0894579226748722e-05,
+      "loss": 0.0002,
+      "step": 1254
+    },
+    {
+      "epoch": 1.60544,
+      "grad_norm": 0.0006829426274634898,
+      "learning_rate": 2.0869770648472364e-05,
+      "loss": 0.0002,
+      "step": 1255
+    },
+    {
+      "epoch": 1.6067200000000001,
+      "grad_norm": 0.0011762769427150488,
+      "learning_rate": 2.08449642254748e-05,
+      "loss": 0.0002,
+      "step": 1256
+    },
+    {
+      "epoch": 1.608,
+      "grad_norm": 0.0001786458888091147,
+      "learning_rate": 2.0820160005060376e-05,
+      "loss": 0.0002,
+      "step": 1257
+    },
+    {
+      "epoch": 1.60928,
+      "grad_norm": 0.001003823708742857,
+      "learning_rate": 2.0795358034529233e-05,
+      "loss": 0.0002,
+      "step": 1258
+    },
+    {
+      "epoch": 1.61056,
+      "grad_norm": 0.0005043644341640174,
+      "learning_rate": 2.077055836117723e-05,
+      "loss": 0.0002,
+      "step": 1259
+    },
+    {
+      "epoch": 1.61184,
+      "grad_norm": 0.0005947434692643583,
+      "learning_rate": 2.074576103229582e-05,
+      "loss": 0.0002,
+      "step": 1260
+    },
+    {
+      "epoch": 1.61184,
+      "eval_loss": 1.1842581033706665,
+      "eval_runtime": 43.4749,
+      "eval_samples_per_second": 11.547,
+      "eval_steps_per_second": 1.449,
+      "step": 1260
+    },
+    {
+      "epoch": 1.6131199999999999,
+      "grad_norm": 0.0007518928032368422,
+      "learning_rate": 2.0720966095172033e-05,
+      "loss": 0.0002,
+      "step": 1261
+    },
+    {
+      "epoch": 1.6143999999999998,
+      "grad_norm": 0.0005646328208968043,
+      "learning_rate": 2.0696173597088283e-05,
+      "loss": 0.0002,
+      "step": 1262
+    },
+    {
+      "epoch": 1.61568,
+      "grad_norm": 0.0002734081353992224,
+      "learning_rate": 2.0671383585322372e-05,
+      "loss": 0.0002,
+      "step": 1263
+    },
+    {
+      "epoch": 1.61696,
+      "grad_norm": 0.0007313131354749203,
+      "learning_rate": 2.064659610714735e-05,
+      "loss": 0.0002,
+      "step": 1264
+    },
+    {
+      "epoch": 1.6182400000000001,
+      "grad_norm": 0.00136158661916852,
+      "learning_rate": 2.062181120983143e-05,
+      "loss": 0.0002,
+      "step": 1265
+    },
+    {
+      "epoch": 1.61952,
+      "grad_norm": 0.00017255271086469293,
+      "learning_rate": 2.0597028940637907e-05,
+      "loss": 0.0002,
+      "step": 1266
+    },
+    {
+      "epoch": 1.6208,
+      "grad_norm": 0.00156474020332098,
+      "learning_rate": 2.0572249346825067e-05,
+      "loss": 0.0002,
+      "step": 1267
+    },
+    {
+      "epoch": 1.62208,
+      "grad_norm": 0.0006997156888246536,
+      "learning_rate": 2.0547472475646093e-05,
+      "loss": 0.0002,
+      "step": 1268
+    },
+    {
+      "epoch": 1.62336,
+      "grad_norm": 0.0015759170055389404,
+      "learning_rate": 2.052269837434897e-05,
+      "loss": 0.0002,
+      "step": 1269
+    },
+    {
+      "epoch": 1.6246399999999999,
+      "grad_norm": 0.0018613300053402781,
+      "learning_rate": 2.0497927090176408e-05,
+      "loss": 0.0002,
+      "step": 1270
+    },
+    {
+      "epoch": 1.62592,
+      "grad_norm": 0.0004975548945367336,
+      "learning_rate": 2.0473158670365745e-05,
+      "loss": 0.0002,
+      "step": 1271
+    },
+    {
+      "epoch": 1.6272,
+      "grad_norm": 0.0012139310128986835,
+      "learning_rate": 2.0448393162148852e-05,
+      "loss": 0.0002,
+      "step": 1272
+    },
+    {
+      "epoch": 1.6284800000000001,
+      "grad_norm": 0.0007971475715748966,
+      "learning_rate": 2.0423630612752054e-05,
+      "loss": 0.0002,
+      "step": 1273
+    },
+    {
+      "epoch": 1.62976,
+      "grad_norm": 0.0012991585535928607,
+      "learning_rate": 2.0398871069396023e-05,
+      "loss": 0.0002,
+      "step": 1274
+    },
+    {
+      "epoch": 1.63104,
+      "grad_norm": 0.0008941097185015678,
+      "learning_rate": 2.0374114579295723e-05,
+      "loss": 0.0002,
+      "step": 1275
+    },
+    {
+      "epoch": 1.63232,
+      "grad_norm": 0.0021978786680847406,
+      "learning_rate": 2.0349361189660247e-05,
+      "loss": 0.0002,
+      "step": 1276
+    },
+    {
+      "epoch": 1.6336,
+      "grad_norm": 0.0004447043174877763,
+      "learning_rate": 2.032461094769284e-05,
+      "loss": 0.0002,
+      "step": 1277
+    },
+    {
+      "epoch": 1.6348799999999999,
+      "grad_norm": 0.0012047910131514072,
+      "learning_rate": 2.0299863900590702e-05,
+      "loss": 0.0002,
+      "step": 1278
+    },
+    {
+      "epoch": 1.6361599999999998,
+      "grad_norm": 0.00027334975311532617,
+      "learning_rate": 2.027512009554493e-05,
+      "loss": 0.0002,
+      "step": 1279
+    },
+    {
+      "epoch": 1.63744,
+      "grad_norm": 0.0020765531808137894,
+      "learning_rate": 2.025037957974049e-05,
+      "loss": 0.0002,
+      "step": 1280
+    },
+    {
+      "epoch": 1.63872,
+      "grad_norm": 0.0007688076002523303,
+      "learning_rate": 2.0225642400356018e-05,
+      "loss": 0.0002,
+      "step": 1281
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 0.0015095536364242435,
+      "learning_rate": 2.0200908604563834e-05,
+      "loss": 0.0002,
+      "step": 1282
+    },
+    {
+      "epoch": 1.64128,
+      "grad_norm": 0.000709025131072849,
+      "learning_rate": 2.0176178239529775e-05,
+      "loss": 0.0002,
+      "step": 1283
+    },
+    {
+      "epoch": 1.64256,
+      "grad_norm": 0.001618406968191266,
+      "learning_rate": 2.0151451352413144e-05,
+      "loss": 0.0002,
+      "step": 1284
+    },
+    {
+      "epoch": 1.64384,
+      "grad_norm": 0.0016390075907111168,
+      "learning_rate": 2.0126727990366625e-05,
+      "loss": 0.0002,
+      "step": 1285
+    },
+    {
+      "epoch": 1.64512,
+      "grad_norm": 0.000761967443395406,
+      "learning_rate": 2.0102008200536155e-05,
+      "loss": 0.0002,
+      "step": 1286
+    },
+    {
+      "epoch": 1.6463999999999999,
+      "grad_norm": 0.0018765556160360575,
+      "learning_rate": 2.0077292030060894e-05,
+      "loss": 0.0002,
+      "step": 1287
+    },
+    {
+      "epoch": 1.64768,
+      "grad_norm": 0.00017738196766003966,
+      "learning_rate": 2.005257952607306e-05,
+      "loss": 0.0002,
+      "step": 1288
+    },
+    {
+      "epoch": 1.64896,
+      "grad_norm": 0.001326532568782568,
+      "learning_rate": 2.0027870735697906e-05,
+      "loss": 0.0002,
+      "step": 1289
+    },
+    {
+      "epoch": 1.6502400000000002,
+      "grad_norm": 0.0003637004701886326,
+      "learning_rate": 2.0003165706053603e-05,
+      "loss": 0.0002,
+      "step": 1290
+    },
+    {
+      "epoch": 1.6502400000000002,
+      "eval_loss": 1.1752021312713623,
+      "eval_runtime": 43.1151,
+      "eval_samples_per_second": 11.643,
+      "eval_steps_per_second": 1.461,
+      "step": 1290
+    },
+    {
+      "epoch": 1.65152,
+      "grad_norm": 0.0008304680814035237,
+      "learning_rate": 1.9978464484251125e-05,
+      "loss": 0.0002,
+      "step": 1291
+    },
+    {
+      "epoch": 1.6528,
+      "grad_norm": 0.00021488082711584866,
+      "learning_rate": 1.9953767117394225e-05,
+      "loss": 0.0002,
+      "step": 1292
+    },
+    {
+      "epoch": 1.65408,
+      "grad_norm": 0.0006303602713160217,
+      "learning_rate": 1.992907365257926e-05,
+      "loss": 0.0002,
+      "step": 1293
+    },
+    {
+      "epoch": 1.65536,
+      "grad_norm": 0.00033922336297109723,
+      "learning_rate": 1.9904384136895193e-05,
+      "loss": 0.0002,
+      "step": 1294
+    },
+    {
+      "epoch": 1.65664,
+      "grad_norm": 0.00041787928785197437,
+      "learning_rate": 1.9879698617423416e-05,
+      "loss": 0.0002,
+      "step": 1295
+    },
+    {
+      "epoch": 1.6579199999999998,
+      "grad_norm": 0.0005025964928790927,
+      "learning_rate": 1.985501714123772e-05,
+      "loss": 0.0002,
+      "step": 1296
+    },
+    {
+      "epoch": 1.6592,
+      "grad_norm": 0.00022155040642246604,
+      "learning_rate": 1.983033975540419e-05,
+      "loss": 0.0002,
+      "step": 1297
+    },
+    {
+      "epoch": 1.66048,
+      "grad_norm": 0.0002778613707050681,
+      "learning_rate": 1.980566650698109e-05,
+      "loss": 0.0002,
+      "step": 1298
+    },
+    {
+      "epoch": 1.6617600000000001,
+      "grad_norm": 0.0003134211292490363,
+      "learning_rate": 1.9780997443018822e-05,
+      "loss": 0.0002,
+      "step": 1299
+    },
+    {
+      "epoch": 1.66304,
+      "grad_norm": 0.0006997901946306229,
+      "learning_rate": 1.9756332610559787e-05,
+      "loss": 0.0002,
+      "step": 1300
+    },
+    {
+      "epoch": 1.66432,
+      "grad_norm": 0.0010301823494955897,
+      "learning_rate": 1.973167205663832e-05,
+      "loss": 0.0002,
+      "step": 1301
+    },
+    {
+      "epoch": 1.6656,
+      "grad_norm": 0.00035824833321385086,
+      "learning_rate": 1.9707015828280607e-05,
+      "loss": 0.0002,
+      "step": 1302
+    },
+    {
+      "epoch": 1.66688,
+      "grad_norm": 0.0009457553969696164,
+      "learning_rate": 1.9682363972504577e-05,
+      "loss": 0.0002,
+      "step": 1303
+    },
+    {
+      "epoch": 1.6681599999999999,
+      "grad_norm": 0.0007050599670037627,
+      "learning_rate": 1.9657716536319826e-05,
+      "loss": 0.0002,
+      "step": 1304
+    },
+    {
+      "epoch": 1.66944,
+      "grad_norm": 0.00086369359632954,
+      "learning_rate": 1.9633073566727515e-05,
+      "loss": 0.0002,
+      "step": 1305
+    },
+    {
+      "epoch": 1.67072,
+      "grad_norm": 0.0013965156394988298,
+      "learning_rate": 1.9608435110720294e-05,
+      "loss": 0.0002,
+      "step": 1306
+    },
+    {
+      "epoch": 1.6720000000000002,
+      "grad_norm": 0.0014690575189888477,
+      "learning_rate": 1.958380121528221e-05,
+      "loss": 0.0002,
+      "step": 1307
+    },
+    {
+      "epoch": 1.67328,
+      "grad_norm": 0.0012357474770396948,
+      "learning_rate": 1.9559171927388586e-05,
+      "loss": 0.0002,
+      "step": 1308
+    },
+    {
+      "epoch": 1.67456,
+      "grad_norm": 0.0019563697278499603,
+      "learning_rate": 1.9534547294006006e-05,
+      "loss": 0.0002,
+      "step": 1309
+    },
+    {
+      "epoch": 1.67584,
+      "grad_norm": 0.00038007416878826916,
+      "learning_rate": 1.9509927362092133e-05,
+      "loss": 0.0002,
+      "step": 1310
+    },
+    {
+      "epoch": 1.67712,
+      "grad_norm": 0.0010130003793165088,
+      "learning_rate": 1.94853121785957e-05,
+      "loss": 0.0002,
+      "step": 1311
+    },
+    {
+      "epoch": 1.6784,
+      "grad_norm": 0.0007123137474991381,
+      "learning_rate": 1.9460701790456348e-05,
+      "loss": 0.0002,
+      "step": 1312
+    },
+    {
+      "epoch": 1.6796799999999998,
+      "grad_norm": 0.0009491070522926748,
+      "learning_rate": 1.9436096244604607e-05,
+      "loss": 0.0002,
+      "step": 1313
+    },
+    {
+      "epoch": 1.68096,
+      "grad_norm": 0.0008381984662264585,
+      "learning_rate": 1.941149558796176e-05,
+      "loss": 0.0002,
+      "step": 1314
+    },
+    {
+      "epoch": 1.68224,
+      "grad_norm": 0.001452930853702128,
+      "learning_rate": 1.9386899867439756e-05,
+      "loss": 0.0002,
+      "step": 1315
+    },
+    {
+      "epoch": 1.6835200000000001,
+      "grad_norm": 0.0004416204756125808,
+      "learning_rate": 1.9362309129941157e-05,
+      "loss": 0.0002,
+      "step": 1316
+    },
+    {
+      "epoch": 1.6848,
+      "grad_norm": 0.0013925581006333232,
+      "learning_rate": 1.9337723422358992e-05,
+      "loss": 0.0002,
+      "step": 1317
+    },
+    {
+      "epoch": 1.68608,
+      "grad_norm": 0.00019246626470703632,
+      "learning_rate": 1.9313142791576733e-05,
+      "loss": 0.0002,
+      "step": 1318
+    },
+    {
+      "epoch": 1.68736,
+      "grad_norm": 0.0008070362964645028,
+      "learning_rate": 1.9288567284468128e-05,
+      "loss": 0.0002,
+      "step": 1319
+    },
+    {
+      "epoch": 1.68864,
+      "grad_norm": 0.0005856645875610411,
+      "learning_rate": 1.9263996947897193e-05,
+      "loss": 0.0002,
+      "step": 1320
+    },
+    {
+      "epoch": 1.68864,
+      "eval_loss": 1.1746774911880493,
+      "eval_runtime": 43.1038,
+      "eval_samples_per_second": 11.646,
+      "eval_steps_per_second": 1.462,
+      "step": 1320
+    },
+    {
+      "epoch": 1.6899199999999999,
+      "grad_norm": 0.0009006186737678945,
+      "learning_rate": 1.9239431828718068e-05,
+      "loss": 0.0002,
+      "step": 1321
+    },
+    {
+      "epoch": 1.6912,
+      "grad_norm": 0.0007656152010895312,
+      "learning_rate": 1.9214871973774936e-05,
+      "loss": 0.0002,
+      "step": 1322
+    },
+    {
+      "epoch": 1.69248,
+      "grad_norm": 0.0007279204437509179,
+      "learning_rate": 1.9190317429901965e-05,
+      "loss": 0.0002,
+      "step": 1323
+    },
+    {
+      "epoch": 1.6937600000000002,
+      "grad_norm": 0.0014004282420501113,
+      "learning_rate": 1.9165768243923172e-05,
+      "loss": 0.0002,
+      "step": 1324
+    },
+    {
+      "epoch": 1.69504,
+      "grad_norm": 0.0016127376584336162,
+      "learning_rate": 1.9141224462652355e-05,
+      "loss": 0.0002,
+      "step": 1325
+    },
+    {
+      "epoch": 1.69632,
+      "grad_norm": 0.0012736310018226504,
+      "learning_rate": 1.9116686132893037e-05,
+      "loss": 0.0002,
+      "step": 1326
+    },
+    {
+      "epoch": 1.6976,
+      "grad_norm": 0.002573431469500065,
+      "learning_rate": 1.9092153301438295e-05,
+      "loss": 0.0002,
+      "step": 1327
+    },
+    {
+      "epoch": 1.69888,
+      "grad_norm": 0.000511201040353626,
+      "learning_rate": 1.9067626015070784e-05,
+      "loss": 0.0002,
+      "step": 1328
+    },
+    {
+      "epoch": 1.70016,
+      "grad_norm": 0.002701991470530629,
+      "learning_rate": 1.9043104320562534e-05,
+      "loss": 0.0002,
+      "step": 1329
+    },
+    {
+      "epoch": 1.7014399999999998,
+      "grad_norm": 0.00016028250684030354,
+      "learning_rate": 1.9018588264674918e-05,
+      "loss": 0.0002,
+      "step": 1330
+    },
+    {
+      "epoch": 1.70272,
+      "grad_norm": 0.002182245021685958,
+      "learning_rate": 1.8994077894158583e-05,
+      "loss": 0.0002,
+      "step": 1331
+    },
+    {
+      "epoch": 1.704,
+      "grad_norm": 0.0001369698002235964,
+      "learning_rate": 1.896957325575331e-05,
+      "loss": 0.0002,
+      "step": 1332
+    },
+    {
+      "epoch": 1.7052800000000001,
+      "grad_norm": 0.0019207252189517021,
+      "learning_rate": 1.894507439618796e-05,
+      "loss": 0.0002,
+      "step": 1333
+    },
+    {
+      "epoch": 1.70656,
+      "grad_norm": 0.0006946883513592184,
+      "learning_rate": 1.8920581362180376e-05,
+      "loss": 0.0002,
+      "step": 1334
+    },
+    {
+      "epoch": 1.70784,
+      "grad_norm": 0.0023960743565112352,
+      "learning_rate": 1.8896094200437286e-05,
+      "loss": 0.0002,
+      "step": 1335
+    },
+    {
+      "epoch": 1.70912,
+      "grad_norm": 0.0003100070753134787,
+      "learning_rate": 1.8871612957654214e-05,
+      "loss": 0.0002,
+      "step": 1336
+    },
+    {
+      "epoch": 1.7104,
+      "grad_norm": 0.0013245106674730778,
+      "learning_rate": 1.8847137680515407e-05,
+      "loss": 0.0002,
+      "step": 1337
+    },
+    {
+      "epoch": 1.7116799999999999,
+      "grad_norm": 0.00015704419638495892,
+      "learning_rate": 1.882266841569375e-05,
+      "loss": 0.0002,
+      "step": 1338
+    },
+    {
+      "epoch": 1.71296,
+      "grad_norm": 0.0016549899009987712,
+      "learning_rate": 1.879820520985062e-05,
+      "loss": 0.0002,
+      "step": 1339
+    },
+    {
+      "epoch": 1.71424,
+      "grad_norm": 0.000723134318832308,
+      "learning_rate": 1.8773748109635887e-05,
+      "loss": 0.0002,
+      "step": 1340
+    },
+    {
+      "epoch": 1.7155200000000002,
+      "grad_norm": 0.0005959108239039779,
+      "learning_rate": 1.874929716168775e-05,
+      "loss": 0.0002,
+      "step": 1341
+    },
+    {
+      "epoch": 1.7168,
+      "grad_norm": 0.00012082476314390078,
+      "learning_rate": 1.872485241263267e-05,
+      "loss": 0.0002,
+      "step": 1342
+    },
+    {
+      "epoch": 1.71808,
+      "grad_norm": 0.0008101105922833085,
+      "learning_rate": 1.8700413909085324e-05,
+      "loss": 0.0002,
+      "step": 1343
+    },
+    {
+      "epoch": 1.71936,
+      "grad_norm": 0.0002989626955240965,
+      "learning_rate": 1.8675981697648435e-05,
+      "loss": 0.0002,
+      "step": 1344
+    },
+    {
+      "epoch": 1.72064,
+      "grad_norm": 0.00021426856983453035,
+      "learning_rate": 1.865155582491277e-05,
+      "loss": 0.0002,
+      "step": 1345
+    },
+    {
+      "epoch": 1.72192,
+      "grad_norm": 0.00018660198838915676,
+      "learning_rate": 1.8627136337456956e-05,
+      "loss": 0.0002,
+      "step": 1346
+    },
+    {
+      "epoch": 1.7231999999999998,
+      "grad_norm": 0.0006224081735126674,
+      "learning_rate": 1.8602723281847503e-05,
+      "loss": 0.0002,
+      "step": 1347
+    },
+    {
+      "epoch": 1.72448,
+      "grad_norm": 0.0008301949710585177,
+      "learning_rate": 1.857831670463861e-05,
+      "loss": 0.0002,
+      "step": 1348
+    },
+    {
+      "epoch": 1.72576,
+      "grad_norm": 0.00014336922322399914,
+      "learning_rate": 1.8553916652372144e-05,
+      "loss": 0.0002,
+      "step": 1349
+    },
+    {
+      "epoch": 1.7270400000000001,
+      "grad_norm": 0.0011552822543308139,
+      "learning_rate": 1.8529523171577524e-05,
+      "loss": 0.0002,
+      "step": 1350
+    },
+    {
+      "epoch": 1.7270400000000001,
+      "eval_loss": 1.1862709522247314,
+      "eval_runtime": 43.6506,
+      "eval_samples_per_second": 11.5,
+      "eval_steps_per_second": 1.443,
+      "step": 1350
+    },
+    {
+      "epoch": 1.72832,
+      "grad_norm": 0.0010331370867788792,
+      "learning_rate": 1.8505136308771633e-05,
+      "loss": 0.0002,
+      "step": 1351
+    },
+    {
+      "epoch": 1.7296,
+      "grad_norm": 0.00012964263441972435,
+      "learning_rate": 1.8480756110458745e-05,
+      "loss": 0.0002,
+      "step": 1352
+    },
+    {
+      "epoch": 1.73088,
+      "grad_norm": 0.0005008718580938876,
+      "learning_rate": 1.8456382623130422e-05,
+      "loss": 0.0002,
+      "step": 1353
+    },
+    {
+      "epoch": 1.73216,
+      "grad_norm": 0.0001546658604638651,
+      "learning_rate": 1.843201589326541e-05,
+      "loss": 0.0002,
+      "step": 1354
+    },
+    {
+      "epoch": 1.7334399999999999,
+      "grad_norm": 0.0003648756246548146,
+      "learning_rate": 1.8407655967329603e-05,
+      "loss": 0.0002,
+      "step": 1355
+    },
+    {
+      "epoch": 1.73472,
+      "grad_norm": 0.0002198543370468542,
+      "learning_rate": 1.8383302891775887e-05,
+      "loss": 0.0002,
+      "step": 1356
+    },
+    {
+      "epoch": 1.736,
+      "grad_norm": 0.00030617695301771164,
+      "learning_rate": 1.835895671304411e-05,
+      "loss": 0.0002,
+      "step": 1357
+    },
+    {
+      "epoch": 1.7372800000000002,
+      "grad_norm": 0.0005935181980021298,
+      "learning_rate": 1.8334617477560956e-05,
+      "loss": 0.0002,
+      "step": 1358
+    },
+    {
+      "epoch": 1.73856,
+      "grad_norm": 0.00011263234773650765,
+      "learning_rate": 1.8310285231739855e-05,
+      "loss": 0.0002,
+      "step": 1359
+    },
+    {
+      "epoch": 1.73984,
+      "grad_norm": 0.0007660306873731315,
+      "learning_rate": 1.828596002198095e-05,
+      "loss": 0.0002,
+      "step": 1360
+    },
+    {
+      "epoch": 1.74112,
+      "grad_norm": 0.0001412270503351465,
+      "learning_rate": 1.826164189467091e-05,
+      "loss": 0.0002,
+      "step": 1361
+    },
+    {
+      "epoch": 1.7424,
+      "grad_norm": 0.0006607779650948942,
+      "learning_rate": 1.823733089618296e-05,
+      "loss": 0.0002,
+      "step": 1362
+    },
+    {
+      "epoch": 1.74368,
+      "grad_norm": 0.0003936325665563345,
+      "learning_rate": 1.8213027072876665e-05,
+      "loss": 0.0002,
+      "step": 1363
+    },
+    {
+      "epoch": 1.7449599999999998,
+      "grad_norm": 0.0006705633131787181,
+      "learning_rate": 1.8188730471097976e-05,
+      "loss": 0.0002,
+      "step": 1364
+    },
+    {
+      "epoch": 1.74624,
+      "grad_norm": 0.0003361855342518538,
+      "learning_rate": 1.8164441137179016e-05,
+      "loss": 0.0002,
+      "step": 1365
+    },
+    {
+      "epoch": 1.74752,
+      "grad_norm": 0.0005639787414111197,
+      "learning_rate": 1.8140159117438073e-05,
+      "loss": 0.0002,
+      "step": 1366
+    },
+    {
+      "epoch": 1.7488000000000001,
+      "grad_norm": 0.00037033509579487145,
+      "learning_rate": 1.8115884458179495e-05,
+      "loss": 0.0002,
+      "step": 1367
+    },
+    {
+      "epoch": 1.75008,
+      "grad_norm": 0.0008438365184701979,
+      "learning_rate": 1.8091617205693574e-05,
+      "loss": 0.0002,
+      "step": 1368
+    },
+    {
+      "epoch": 1.75136,
+      "grad_norm": 0.000185017182957381,
+      "learning_rate": 1.806735740625649e-05,
+      "loss": 0.0002,
+      "step": 1369
+    },
+    {
+      "epoch": 1.75264,
+      "grad_norm": 0.0009372641798108816,
+      "learning_rate": 1.8043105106130213e-05,
+      "loss": 0.0002,
+      "step": 1370
+    },
+    {
+      "epoch": 1.75392,
+      "grad_norm": 0.0003428539785090834,
+      "learning_rate": 1.8018860351562406e-05,
+      "loss": 0.0002,
+      "step": 1371
+    },
+    {
+      "epoch": 1.7551999999999999,
+      "grad_norm": 0.001404768554493785,
+      "learning_rate": 1.7994623188786346e-05,
+      "loss": 0.0002,
+      "step": 1372
+    },
+    {
+      "epoch": 1.75648,
+      "grad_norm": 0.0002842540852725506,
+      "learning_rate": 1.7970393664020824e-05,
+      "loss": 0.0002,
+      "step": 1373
+    },
+    {
+      "epoch": 1.75776,
+      "grad_norm": 0.0010203004349023104,
+      "learning_rate": 1.7946171823470085e-05,
+      "loss": 0.0002,
+      "step": 1374
+    },
+    {
+      "epoch": 1.7590400000000002,
+      "grad_norm": 0.0005073043284937739,
+      "learning_rate": 1.7921957713323695e-05,
+      "loss": 0.0002,
+      "step": 1375
+    },
+    {
+      "epoch": 1.76032,
+      "grad_norm": 0.00044870551209896803,
+      "learning_rate": 1.7897751379756515e-05,
+      "loss": 0.0002,
+      "step": 1376
+    },
+    {
+      "epoch": 1.7616,
+      "grad_norm": 0.0001787729124771431,
+      "learning_rate": 1.787355286892854e-05,
+      "loss": 0.0002,
+      "step": 1377
+    },
+    {
+      "epoch": 1.76288,
+      "grad_norm": 0.0007749973447062075,
+      "learning_rate": 1.7849362226984852e-05,
+      "loss": 0.0002,
+      "step": 1378
+    },
+    {
+      "epoch": 1.76416,
+      "grad_norm": 0.0006996009033173323,
+      "learning_rate": 1.7825179500055556e-05,
+      "loss": 0.0002,
+      "step": 1379
+    },
+    {
+      "epoch": 1.76544,
+      "grad_norm": 0.0005011953180655837,
+      "learning_rate": 1.7801004734255632e-05,
+      "loss": 0.0002,
+      "step": 1380
+    },
+    {
+      "epoch": 1.76544,
+      "eval_loss": 1.1896673440933228,
+      "eval_runtime": 43.8766,
+      "eval_samples_per_second": 11.441,
+      "eval_steps_per_second": 1.436,
+      "step": 1380
+    },
+    {
+      "epoch": 1.7667199999999998,
+      "grad_norm": 0.0009371109772473574,
+      "learning_rate": 1.7776837975684902e-05,
+      "loss": 0.0002,
+      "step": 1381
+    },
+    {
+      "epoch": 1.768,
+      "grad_norm": 0.000150485968333669,
+      "learning_rate": 1.7752679270427902e-05,
+      "loss": 0.0002,
+      "step": 1382
+    },
+    {
+      "epoch": 1.76928,
+      "grad_norm": 0.0002970367786474526,
+      "learning_rate": 1.7728528664553812e-05,
+      "loss": 0.0002,
+      "step": 1383
+    },
+    {
+      "epoch": 1.7705600000000001,
+      "grad_norm": 0.000249633303610608,
+      "learning_rate": 1.7704386204116385e-05,
+      "loss": 0.0002,
+      "step": 1384
+    },
+    {
+      "epoch": 1.77184,
+      "grad_norm": 0.000201058792299591,
+      "learning_rate": 1.7680251935153813e-05,
+      "loss": 0.0002,
+      "step": 1385
+    },
+    {
+      "epoch": 1.77312,
+      "grad_norm": 0.00021982115868013352,
+      "learning_rate": 1.7656125903688698e-05,
+      "loss": 0.0002,
+      "step": 1386
+    },
+    {
+      "epoch": 1.7744,
+      "grad_norm": 0.00044147908920422196,
+      "learning_rate": 1.763200815572791e-05,
+      "loss": 0.0002,
+      "step": 1387
+    },
+    {
+      "epoch": 1.77568,
+      "grad_norm": 0.00031368451891466975,
+      "learning_rate": 1.7607898737262528e-05,
+      "loss": 0.0002,
+      "step": 1388
+    },
+    {
+      "epoch": 1.7769599999999999,
+      "grad_norm": 0.00021405424922704697,
+      "learning_rate": 1.7583797694267754e-05,
+      "loss": 0.0002,
+      "step": 1389
+    },
+    {
+      "epoch": 1.77824,
+      "grad_norm": 0.00014162302250042558,
+      "learning_rate": 1.7559705072702814e-05,
+      "loss": 0.0002,
+      "step": 1390
+    },
+    {
+      "epoch": 1.77952,
+      "grad_norm": 0.000270137214101851,
+      "learning_rate": 1.7535620918510883e-05,
+      "loss": 0.0002,
+      "step": 1391
+    },
+    {
+      "epoch": 1.7808000000000002,
+      "grad_norm": 0.0003177259932272136,
+      "learning_rate": 1.7511545277618965e-05,
+      "loss": 0.0002,
+      "step": 1392
+    },
+    {
+      "epoch": 1.78208,
+      "grad_norm": 0.0005201466847211123,
+      "learning_rate": 1.7487478195937865e-05,
+      "loss": 0.0002,
+      "step": 1393
+    },
+    {
+      "epoch": 1.78336,
+      "grad_norm": 0.0006870387587696314,
+      "learning_rate": 1.7463419719362034e-05,
+      "loss": 0.0002,
+      "step": 1394
+    },
+    {
+      "epoch": 1.78464,
+      "grad_norm": 0.00019761078874580562,
+      "learning_rate": 1.7439369893769528e-05,
+      "loss": 0.0002,
+      "step": 1395
+    },
+    {
+      "epoch": 1.78592,
+      "grad_norm": 0.0012992613483220339,
+      "learning_rate": 1.7415328765021912e-05,
+      "loss": 0.0002,
+      "step": 1396
+    },
+    {
+      "epoch": 1.7872,
+      "grad_norm": 0.0006985355284996331,
+      "learning_rate": 1.7391296378964156e-05,
+      "loss": 0.0002,
+      "step": 1397
+    },
+    {
+      "epoch": 1.7884799999999998,
+      "grad_norm": 0.0004908322589471936,
+      "learning_rate": 1.7367272781424566e-05,
+      "loss": 0.0002,
+      "step": 1398
+    },
+    {
+      "epoch": 1.78976,
+      "grad_norm": 0.00016040352056734264,
+      "learning_rate": 1.7343258018214673e-05,
+      "loss": 0.0002,
+      "step": 1399
+    },
+    {
+      "epoch": 1.79104,
+      "grad_norm": 0.00038101038080640137,
+      "learning_rate": 1.731925213512918e-05,
+      "loss": 0.0002,
+      "step": 1400
+    },
+    {
+      "epoch": 1.7923200000000001,
+      "grad_norm": 0.00041198916733264923,
+      "learning_rate": 1.7295255177945858e-05,
+      "loss": 0.0002,
+      "step": 1401
+    },
+    {
+      "epoch": 1.7936,
+      "grad_norm": 0.0008410747977904975,
+      "learning_rate": 1.7271267192425436e-05,
+      "loss": 0.0002,
+      "step": 1402
+    },
+    {
+      "epoch": 1.79488,
+      "grad_norm": 0.00016792569658719003,
+      "learning_rate": 1.7247288224311557e-05,
+      "loss": 0.0002,
+      "step": 1403
+    },
+    {
+      "epoch": 1.79616,
+      "grad_norm": 0.00025911914417520165,
+      "learning_rate": 1.722331831933065e-05,
+      "loss": 0.0002,
+      "step": 1404
+    },
+    {
+      "epoch": 1.79744,
+      "grad_norm": 0.0005258122109808028,
+      "learning_rate": 1.719935752319187e-05,
+      "loss": 0.0002,
+      "step": 1405
+    },
+    {
+      "epoch": 1.7987199999999999,
+      "grad_norm": 0.000921487167943269,
+      "learning_rate": 1.7175405881587015e-05,
+      "loss": 0.0002,
+      "step": 1406
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.00017789006233215332,
+      "learning_rate": 1.7151463440190387e-05,
+      "loss": 0.0002,
+      "step": 1407
+    },
+    {
+      "epoch": 1.80128,
+      "grad_norm": 0.0008785719983279705,
+      "learning_rate": 1.7127530244658796e-05,
+      "loss": 0.0002,
+      "step": 1408
+    },
+    {
+      "epoch": 1.8025600000000002,
+      "grad_norm": 0.0005776337930001318,
+      "learning_rate": 1.7103606340631365e-05,
+      "loss": 0.0002,
+      "step": 1409
+    },
+    {
+      "epoch": 1.80384,
+      "grad_norm": 0.00021976965945214033,
+      "learning_rate": 1.7079691773729558e-05,
+      "loss": 0.0002,
+      "step": 1410
+    },
+    {
+      "epoch": 1.80384,
+      "eval_loss": 1.1821080446243286,
+      "eval_runtime": 43.496,
+      "eval_samples_per_second": 11.541,
+      "eval_steps_per_second": 1.448,
+      "step": 1410
+    },
+    {
+      "epoch": 1.80512,
+      "grad_norm": 0.00015776725194882601,
+      "learning_rate": 1.7055786589556983e-05,
+      "loss": 0.0002,
+      "step": 1411
+    },
+    {
+      "epoch": 1.8064,
+      "grad_norm": 0.0003731239412445575,
+      "learning_rate": 1.703189083369938e-05,
+      "loss": 0.0002,
+      "step": 1412
+    },
+    {
+      "epoch": 1.80768,
+      "grad_norm": 0.0003114558639936149,
+      "learning_rate": 1.7008004551724503e-05,
+      "loss": 0.0002,
+      "step": 1413
+    },
+    {
+      "epoch": 1.80896,
+      "grad_norm": 0.0005815111217088997,
+      "learning_rate": 1.6984127789182046e-05,
+      "loss": 0.0002,
+      "step": 1414
+    },
+    {
+      "epoch": 1.8102399999999998,
+      "grad_norm": 0.00024505917099304497,
+      "learning_rate": 1.6960260591603553e-05,
+      "loss": 0.0002,
+      "step": 1415
+    },
+    {
+      "epoch": 1.81152,
+      "grad_norm": 0.0004912762087769806,
+      "learning_rate": 1.6936403004502303e-05,
+      "loss": 0.0002,
+      "step": 1416
+    },
+    {
+      "epoch": 1.8128,
+      "grad_norm": 0.0006611712742596865,
+      "learning_rate": 1.6912555073373292e-05,
+      "loss": 0.0002,
+      "step": 1417
+    },
+    {
+      "epoch": 1.8140800000000001,
+      "grad_norm": 0.0011671832762658596,
+      "learning_rate": 1.688871684369306e-05,
+      "loss": 0.0002,
+      "step": 1418
+    },
+    {
+      "epoch": 1.81536,
+      "grad_norm": 0.0006492765387520194,
+      "learning_rate": 1.6864888360919664e-05,
+      "loss": 0.0002,
+      "step": 1419
+    },
+    {
+      "epoch": 1.81664,
+      "grad_norm": 0.0014940252294763923,
+      "learning_rate": 1.6841069670492584e-05,
+      "loss": 0.0002,
+      "step": 1420
+    },
+    {
+      "epoch": 1.81792,
+      "grad_norm": 0.00017851243319455534,
+      "learning_rate": 1.681726081783261e-05,
+      "loss": 0.0002,
+      "step": 1421
+    },
+    {
+      "epoch": 1.8192,
+      "grad_norm": 0.0008618085994385183,
+      "learning_rate": 1.6793461848341787e-05,
+      "loss": 0.0002,
+      "step": 1422
+    },
+    {
+      "epoch": 1.8204799999999999,
+      "grad_norm": 0.0009502249886281788,
+      "learning_rate": 1.67696728074033e-05,
+      "loss": 0.0002,
+      "step": 1423
+    },
+    {
+      "epoch": 1.82176,
+      "grad_norm": 0.0014715096913278103,
+      "learning_rate": 1.6745893740381402e-05,
+      "loss": 0.0002,
+      "step": 1424
+    },
+    {
+      "epoch": 1.82304,
+      "grad_norm": 0.0006264635012485087,
+      "learning_rate": 1.6722124692621348e-05,
+      "loss": 0.0002,
+      "step": 1425
+    },
+    {
+      "epoch": 1.8243200000000002,
+      "grad_norm": 0.0013680956326425076,
+      "learning_rate": 1.6698365709449243e-05,
+      "loss": 0.0002,
+      "step": 1426
+    },
+    {
+      "epoch": 1.8256000000000001,
+      "grad_norm": 0.0005238053272478282,
+      "learning_rate": 1.6674616836172055e-05,
+      "loss": 0.0002,
+      "step": 1427
+    },
+    {
+      "epoch": 1.82688,
+      "grad_norm": 0.001148419687524438,
+      "learning_rate": 1.6650878118077426e-05,
+      "loss": 0.0002,
+      "step": 1428
+    },
+    {
+      "epoch": 1.82816,
+      "grad_norm": 0.0004914168966934085,
+      "learning_rate": 1.6627149600433656e-05,
+      "loss": 0.0002,
+      "step": 1429
+    },
+    {
+      "epoch": 1.82944,
+      "grad_norm": 0.0009969781385734677,
+      "learning_rate": 1.660343132848959e-05,
+      "loss": 0.0002,
+      "step": 1430
+    },
+    {
+      "epoch": 1.83072,
+      "grad_norm": 0.0003050584637094289,
+      "learning_rate": 1.657972334747453e-05,
+      "loss": 0.0002,
+      "step": 1431
+    },
+    {
+      "epoch": 1.8319999999999999,
+      "grad_norm": 0.0008236157591454685,
+      "learning_rate": 1.6556025702598163e-05,
+      "loss": 0.0002,
+      "step": 1432
+    },
+    {
+      "epoch": 1.83328,
+      "grad_norm": 0.00024275497708003968,
+      "learning_rate": 1.6532338439050453e-05,
+      "loss": 0.0002,
+      "step": 1433
+    },
+    {
+      "epoch": 1.83456,
+      "grad_norm": 0.0009991588303819299,
+      "learning_rate": 1.650866160200158e-05,
+      "loss": 0.0002,
+      "step": 1434
+    },
+    {
+      "epoch": 1.8358400000000001,
+      "grad_norm": 0.0002861864340957254,
+      "learning_rate": 1.648499523660183e-05,
+      "loss": 0.0002,
+      "step": 1435
+    },
+    {
+      "epoch": 1.83712,
+      "grad_norm": 0.0007088854908943176,
+      "learning_rate": 1.6461339387981535e-05,
+      "loss": 0.0002,
+      "step": 1436
+    },
+    {
+      "epoch": 1.8384,
+      "grad_norm": 0.0005638033035211265,
+      "learning_rate": 1.643769410125095e-05,
+      "loss": 0.0002,
+      "step": 1437
+    },
+    {
+      "epoch": 1.83968,
+      "grad_norm": 0.00013802683679386973,
+      "learning_rate": 1.6414059421500207e-05,
+      "loss": 0.0002,
+      "step": 1438
+    },
+    {
+      "epoch": 1.84096,
+      "grad_norm": 0.00027752117603085935,
+      "learning_rate": 1.6390435393799214e-05,
+      "loss": 0.0002,
+      "step": 1439
+    },
+    {
+      "epoch": 1.8422399999999999,
+      "grad_norm": 0.0008538371766917408,
+      "learning_rate": 1.6366822063197556e-05,
+      "loss": 0.0002,
+      "step": 1440
+    },
+    {
+      "epoch": 1.8422399999999999,
+      "eval_loss": 1.1808280944824219,
+      "eval_runtime": 43.3456,
+      "eval_samples_per_second": 11.581,
+      "eval_steps_per_second": 1.453,
+      "step": 1440
+    },
+    {
+      "epoch": 1.84352,
+      "grad_norm": 0.00022036586597096175,
+      "learning_rate": 1.6343219474724404e-05,
+      "loss": 0.0002,
+      "step": 1441
+    },
+    {
+      "epoch": 1.8448,
+      "grad_norm": 0.0010862440103664994,
+      "learning_rate": 1.6319627673388495e-05,
+      "loss": 0.0002,
+      "step": 1442
+    },
+    {
+      "epoch": 1.8460800000000002,
+      "grad_norm": 0.00029390587587840855,
+      "learning_rate": 1.6296046704177927e-05,
+      "loss": 0.0002,
+      "step": 1443
+    },
+    {
+      "epoch": 1.8473600000000001,
+      "grad_norm": 0.0015011231880635023,
+      "learning_rate": 1.627247661206021e-05,
+      "loss": 0.0002,
+      "step": 1444
+    },
+    {
+      "epoch": 1.84864,
+      "grad_norm": 0.0004352701944299042,
+      "learning_rate": 1.6248917441982045e-05,
+      "loss": 0.0002,
+      "step": 1445
+    },
+    {
+      "epoch": 1.84992,
+      "grad_norm": 0.0007848394452594221,
+      "learning_rate": 1.6225369238869374e-05,
+      "loss": 0.0002,
+      "step": 1446
+    },
+    {
+      "epoch": 1.8512,
+      "grad_norm": 0.00015477949636988342,
+      "learning_rate": 1.6201832047627174e-05,
+      "loss": 0.0002,
+      "step": 1447
+    },
+    {
+      "epoch": 1.85248,
+      "grad_norm": 0.0010525579564273357,
+      "learning_rate": 1.6178305913139438e-05,
+      "loss": 0.0002,
+      "step": 1448
+    },
+    {
+      "epoch": 1.8537599999999999,
+      "grad_norm": 0.00019572787277866155,
+      "learning_rate": 1.6154790880269083e-05,
+      "loss": 0.0002,
+      "step": 1449
+    },
+    {
+      "epoch": 1.85504,
+      "grad_norm": 0.001078969449736178,
+      "learning_rate": 1.613128699385785e-05,
+      "loss": 0.0002,
+      "step": 1450
+    },
+    {
+      "epoch": 1.85632,
+      "grad_norm": 0.0001666359748924151,
+      "learning_rate": 1.6107794298726222e-05,
+      "loss": 0.0002,
+      "step": 1451
+    },
+    {
+      "epoch": 1.8576000000000001,
+      "grad_norm": 0.0004069862188771367,
+      "learning_rate": 1.6084312839673346e-05,
+      "loss": 0.0002,
+      "step": 1452
+    },
+    {
+      "epoch": 1.85888,
+      "grad_norm": 0.0008127239998430014,
+      "learning_rate": 1.6060842661476935e-05,
+      "loss": 0.0002,
+      "step": 1453
+    },
+    {
+      "epoch": 1.86016,
+      "grad_norm": 0.0008960436680354178,
+      "learning_rate": 1.6037383808893195e-05,
+      "loss": 0.0002,
+      "step": 1454
+    },
+    {
+      "epoch": 1.86144,
+      "grad_norm": 0.00027456111274659634,
+      "learning_rate": 1.6013936326656736e-05,
+      "loss": 0.0002,
+      "step": 1455
+    },
+    {
+      "epoch": 1.86272,
+      "grad_norm": 0.0003418760607019067,
+      "learning_rate": 1.5990500259480492e-05,
+      "loss": 0.0002,
+      "step": 1456
+    },
+    {
+      "epoch": 1.8639999999999999,
+      "grad_norm": 0.0004984101396985352,
+      "learning_rate": 1.5967075652055616e-05,
+      "loss": 0.0002,
+      "step": 1457
+    },
+    {
+      "epoch": 1.86528,
+      "grad_norm": 0.0002873337361961603,
+      "learning_rate": 1.59436625490514e-05,
+      "loss": 0.0002,
+      "step": 1458
+    },
+    {
+      "epoch": 1.86656,
+      "grad_norm": 0.0005379713838919997,
+      "learning_rate": 1.592026099511523e-05,
+      "loss": 0.0002,
+      "step": 1459
+    },
+    {
+      "epoch": 1.86784,
+      "grad_norm": 0.0003407693584449589,
+      "learning_rate": 1.589687103487243e-05,
+      "loss": 0.0002,
+      "step": 1460
+    },
+    {
+      "epoch": 1.8691200000000001,
+      "grad_norm": 0.0007091562729328871,
+      "learning_rate": 1.587349271292626e-05,
+      "loss": 0.0002,
+      "step": 1461
+    },
+    {
+      "epoch": 1.8704,
+      "grad_norm": 0.0007454613223671913,
+      "learning_rate": 1.5850126073857733e-05,
+      "loss": 0.0002,
+      "step": 1462
+    },
+    {
+      "epoch": 1.87168,
+      "grad_norm": 0.00040651875315234065,
+      "learning_rate": 1.5826771162225637e-05,
+      "loss": 0.0002,
+      "step": 1463
+    },
+    {
+      "epoch": 1.87296,
+      "grad_norm": 0.0006714593037031591,
+      "learning_rate": 1.580342802256636e-05,
+      "loss": 0.0002,
+      "step": 1464
+    },
+    {
+      "epoch": 1.87424,
+      "grad_norm": 0.0006067923968657851,
+      "learning_rate": 1.5780096699393845e-05,
+      "loss": 0.0002,
+      "step": 1465
+    },
+    {
+      "epoch": 1.8755199999999999,
+      "grad_norm": 0.0007768379291519523,
+      "learning_rate": 1.5756777237199527e-05,
+      "loss": 0.0002,
+      "step": 1466
+    },
+    {
+      "epoch": 1.8768,
+      "grad_norm": 0.0004379746678750962,
+      "learning_rate": 1.573346968045219e-05,
+      "loss": 0.0002,
+      "step": 1467
+    },
+    {
+      "epoch": 1.87808,
+      "grad_norm": 0.00031985618988983333,
+      "learning_rate": 1.5710174073597944e-05,
+      "loss": 0.0002,
+      "step": 1468
+    },
+    {
+      "epoch": 1.8793600000000001,
+      "grad_norm": 0.0005030058673582971,
+      "learning_rate": 1.5686890461060084e-05,
+      "loss": 0.0002,
+      "step": 1469
+    },
+    {
+      "epoch": 1.88064,
+      "grad_norm": 0.0002572862431406975,
+      "learning_rate": 1.5663618887239053e-05,
+      "loss": 0.0002,
+      "step": 1470
+    },
+    {
+      "epoch": 1.88064,
+      "eval_loss": 1.1787371635437012,
+      "eval_runtime": 43.1586,
+      "eval_samples_per_second": 11.632,
+      "eval_steps_per_second": 1.46,
+      "step": 1470
+    },
+    {
+      "epoch": 1.88192,
+      "grad_norm": 0.0005245308275334537,
+      "learning_rate": 1.564035939651234e-05,
+      "loss": 0.0002,
+      "step": 1471
+    },
+    {
+      "epoch": 1.8832,
+      "grad_norm": 0.00030732108280062675,
+      "learning_rate": 1.5617112033234357e-05,
+      "loss": 0.0002,
+      "step": 1472
+    },
+    {
+      "epoch": 1.88448,
+      "grad_norm": 0.0007221931009553373,
+      "learning_rate": 1.559387684173644e-05,
+      "loss": 0.0002,
+      "step": 1473
+    },
+    {
+      "epoch": 1.8857599999999999,
+      "grad_norm": 0.0006306172581389546,
+      "learning_rate": 1.5570653866326685e-05,
+      "loss": 0.0002,
+      "step": 1474
+    },
+    {
+      "epoch": 1.88704,
+      "grad_norm": 0.0006269289879128337,
+      "learning_rate": 1.5547443151289887e-05,
+      "loss": 0.0002,
+      "step": 1475
+    },
+    {
+      "epoch": 1.88832,
+      "grad_norm": 0.0005559783894568682,
+      "learning_rate": 1.552424474088748e-05,
+      "loss": 0.0002,
+      "step": 1476
+    },
+    {
+      "epoch": 1.8896,
+      "grad_norm": 0.00046742561971768737,
+      "learning_rate": 1.5501058679357413e-05,
+      "loss": 0.0002,
+      "step": 1477
+    },
+    {
+      "epoch": 1.8908800000000001,
+      "grad_norm": 0.00011757310858229175,
+      "learning_rate": 1.5477885010914116e-05,
+      "loss": 0.0002,
+      "step": 1478
+    },
+    {
+      "epoch": 1.89216,
+      "grad_norm": 0.0009887436171993613,
+      "learning_rate": 1.5454723779748352e-05,
+      "loss": 0.0002,
+      "step": 1479
+    },
+    {
+      "epoch": 1.89344,
+      "grad_norm": 0.00039360520895570517,
+      "learning_rate": 1.5431575030027204e-05,
+      "loss": 0.0002,
+      "step": 1480
+    },
+    {
+      "epoch": 1.89472,
+      "grad_norm": 0.0007880927878431976,
+      "learning_rate": 1.540843880589391e-05,
+      "loss": 0.0002,
+      "step": 1481
+    },
+    {
+      "epoch": 1.896,
+      "grad_norm": 0.0004839149478357285,
+      "learning_rate": 1.5385315151467856e-05,
+      "loss": 0.0002,
+      "step": 1482
+    },
+    {
+      "epoch": 1.8972799999999999,
+      "grad_norm": 0.0007827369263395667,
+      "learning_rate": 1.5362204110844454e-05,
+      "loss": 0.0002,
+      "step": 1483
+    },
+    {
+      "epoch": 1.89856,
+      "grad_norm": 0.0004319619038142264,
+      "learning_rate": 1.533910572809504e-05,
+      "loss": 0.0002,
+      "step": 1484
+    },
+    {
+      "epoch": 1.89984,
+      "grad_norm": 0.00044077992788515985,
+      "learning_rate": 1.531602004726685e-05,
+      "loss": 0.0002,
+      "step": 1485
+    },
+    {
+      "epoch": 1.9011200000000001,
+      "grad_norm": 0.0003705875133164227,
+      "learning_rate": 1.529294711238286e-05,
+      "loss": 0.0002,
+      "step": 1486
+    },
+    {
+      "epoch": 1.9024,
+      "grad_norm": 0.0005771559081040323,
+      "learning_rate": 1.5269886967441756e-05,
+      "loss": 0.0002,
+      "step": 1487
+    },
+    {
+      "epoch": 1.90368,
+      "grad_norm": 0.00013823727203998715,
+      "learning_rate": 1.5246839656417845e-05,
+      "loss": 0.0002,
+      "step": 1488
+    },
+    {
+      "epoch": 1.90496,
+      "grad_norm": 0.0001308949722442776,
+      "learning_rate": 1.5223805223260946e-05,
+      "loss": 0.0002,
+      "step": 1489
+    },
+    {
+      "epoch": 1.90624,
+      "grad_norm": 0.00029769743559882045,
+      "learning_rate": 1.5200783711896327e-05,
+      "loss": 0.0002,
+      "step": 1490
+    },
+    {
+      "epoch": 1.9075199999999999,
+      "grad_norm": 0.0002313807635800913,
+      "learning_rate": 1.5177775166224606e-05,
+      "loss": 0.0002,
+      "step": 1491
+    },
+    {
+      "epoch": 1.9088,
+      "grad_norm": 0.00012953733676113188,
+      "learning_rate": 1.5154779630121703e-05,
+      "loss": 0.0002,
+      "step": 1492
+    },
+    {
+      "epoch": 1.91008,
+      "grad_norm": 0.0003203785454388708,
+      "learning_rate": 1.5131797147438692e-05,
+      "loss": 0.0002,
+      "step": 1493
+    },
+    {
+      "epoch": 1.91136,
+      "grad_norm": 0.00015288084978237748,
+      "learning_rate": 1.5108827762001772e-05,
+      "loss": 0.0002,
+      "step": 1494
+    },
+    {
+      "epoch": 1.9126400000000001,
+      "grad_norm": 0.00027298854547552764,
+      "learning_rate": 1.5085871517612178e-05,
+      "loss": 0.0002,
+      "step": 1495
+    },
+    {
+      "epoch": 1.91392,
+      "grad_norm": 0.0003144654037896544,
+      "learning_rate": 1.5062928458046066e-05,
+      "loss": 0.0002,
+      "step": 1496
+    },
+    {
+      "epoch": 1.9152,
+      "grad_norm": 0.0005439766100607812,
+      "learning_rate": 1.5039998627054462e-05,
+      "loss": 0.0002,
+      "step": 1497
+    },
+    {
+      "epoch": 1.91648,
+      "grad_norm": 0.00028937109163962305,
+      "learning_rate": 1.5017082068363168e-05,
+      "loss": 0.0002,
+      "step": 1498
+    },
+    {
+      "epoch": 1.91776,
+      "grad_norm": 0.00034966770908795297,
+      "learning_rate": 1.4994178825672663e-05,
+      "loss": 0.0002,
+      "step": 1499
+    },
+    {
+      "epoch": 1.9190399999999999,
+      "grad_norm": 0.000162362091941759,
+      "learning_rate": 1.4971288942658046e-05,
+      "loss": 0.0002,
+      "step": 1500
+    },
+    {
+      "epoch": 1.9190399999999999,
+      "eval_loss": 1.18230402469635,
+      "eval_runtime": 43.4353,
+      "eval_samples_per_second": 11.557,
+      "eval_steps_per_second": 1.45,
+      "step": 1500
+    },
+    {
+      "epoch": 1.92032,
+      "grad_norm": 0.00014286025543697178,
+      "learning_rate": 1.4948412462968929e-05,
+      "loss": 0.0002,
+      "step": 1501
+    },
+    {
+      "epoch": 1.9216,
+      "grad_norm": 0.00026032765163108706,
+      "learning_rate": 1.4925549430229378e-05,
+      "loss": 0.0002,
+      "step": 1502
+    },
+    {
+      "epoch": 1.9228800000000001,
+      "grad_norm": 0.000396853982238099,
+      "learning_rate": 1.49026998880378e-05,
+      "loss": 0.0002,
+      "step": 1503
+    },
+    {
+      "epoch": 1.92416,
+      "grad_norm": 0.00029547963640652597,
+      "learning_rate": 1.4879863879966903e-05,
+      "loss": 0.0002,
+      "step": 1504
+    },
+    {
+      "epoch": 1.92544,
+      "grad_norm": 0.0006995249423198402,
+      "learning_rate": 1.4857041449563552e-05,
+      "loss": 0.0002,
+      "step": 1505
+    },
+    {
+      "epoch": 1.92672,
+      "grad_norm": 0.00016520809731446207,
+      "learning_rate": 1.4834232640348733e-05,
+      "loss": 0.0002,
+      "step": 1506
+    },
+    {
+      "epoch": 1.928,
+      "grad_norm": 0.00048660230822861195,
+      "learning_rate": 1.4811437495817486e-05,
+      "loss": 0.0002,
+      "step": 1507
+    },
+    {
+      "epoch": 1.9292799999999999,
+      "grad_norm": 0.0006820852868258953,
+      "learning_rate": 1.4788656059438742e-05,
+      "loss": 0.0002,
+      "step": 1508
+    },
+    {
+      "epoch": 1.93056,
+      "grad_norm": 0.0003663533425424248,
+      "learning_rate": 1.4765888374655335e-05,
+      "loss": 0.0002,
+      "step": 1509
+    },
+    {
+      "epoch": 1.93184,
+      "grad_norm": 0.0011120438575744629,
+      "learning_rate": 1.4743134484883847e-05,
+      "loss": 0.0002,
+      "step": 1510
+    },
+    {
+      "epoch": 1.93312,
+      "grad_norm": 0.00039282278157770634,
+      "learning_rate": 1.4720394433514572e-05,
+      "loss": 0.0002,
+      "step": 1511
+    },
+    {
+      "epoch": 1.9344000000000001,
+      "grad_norm": 0.0013644680147990584,
+      "learning_rate": 1.4697668263911408e-05,
+      "loss": 0.0002,
+      "step": 1512
+    },
+    {
+      "epoch": 1.93568,
+      "grad_norm": 0.0001902208023238927,
+      "learning_rate": 1.4674956019411775e-05,
+      "loss": 0.0002,
+      "step": 1513
+    },
+    {
+      "epoch": 1.93696,
+      "grad_norm": 0.0017309424001723528,
+      "learning_rate": 1.4652257743326557e-05,
+      "loss": 0.0002,
+      "step": 1514
+    },
+    {
+      "epoch": 1.93824,
+      "grad_norm": 0.00018405367154628038,
+      "learning_rate": 1.4629573478939976e-05,
+      "loss": 0.0002,
+      "step": 1515
+    },
+    {
+      "epoch": 1.93952,
+      "grad_norm": 0.0017507137963548303,
+      "learning_rate": 1.4606903269509554e-05,
+      "loss": 0.0002,
+      "step": 1516
+    },
+    {
+      "epoch": 1.9407999999999999,
+      "grad_norm": 0.00046362378634512424,
+      "learning_rate": 1.4584247158266008e-05,
+      "loss": 0.0002,
+      "step": 1517
+    },
+    {
+      "epoch": 1.94208,
+      "grad_norm": 0.0012449919013306499,
+      "learning_rate": 1.456160518841316e-05,
+      "loss": 0.0002,
+      "step": 1518
+    },
+    {
+      "epoch": 1.94336,
+      "grad_norm": 0.00016421821783296764,
+      "learning_rate": 1.453897740312788e-05,
+      "loss": 0.0002,
+      "step": 1519
+    },
+    {
+      "epoch": 1.9446400000000001,
+      "grad_norm": 0.0012402897700667381,
+      "learning_rate": 1.451636384555998e-05,
+      "loss": 0.0002,
+      "step": 1520
+    },
+    {
+      "epoch": 1.94592,
+      "grad_norm": 0.00031017547007650137,
+      "learning_rate": 1.449376455883214e-05,
+      "loss": 0.0002,
+      "step": 1521
+    },
+    {
+      "epoch": 1.9472,
+      "grad_norm": 0.0014095036312937737,
+      "learning_rate": 1.4471179586039833e-05,
+      "loss": 0.0002,
+      "step": 1522
+    },
+    {
+      "epoch": 1.94848,
+      "grad_norm": 0.000539817672688514,
+      "learning_rate": 1.444860897025122e-05,
+      "loss": 0.0002,
+      "step": 1523
+    },
+    {
+      "epoch": 1.94976,
+      "grad_norm": 0.001391338068060577,
+      "learning_rate": 1.4426052754507108e-05,
+      "loss": 0.0002,
+      "step": 1524
+    },
+    {
+      "epoch": 1.9510399999999999,
+      "grad_norm": 0.0005815787590108812,
+      "learning_rate": 1.4403510981820823e-05,
+      "loss": 0.0002,
+      "step": 1525
+    },
+    {
+      "epoch": 1.95232,
+      "grad_norm": 0.000504453491885215,
+      "learning_rate": 1.438098369517817e-05,
+      "loss": 0.0002,
+      "step": 1526
+    },
+    {
+      "epoch": 1.9536,
+      "grad_norm": 0.0015071576926857233,
+      "learning_rate": 1.4358470937537306e-05,
+      "loss": 0.0002,
+      "step": 1527
+    },
+    {
+      "epoch": 1.95488,
+      "grad_norm": 0.0005614665569737554,
+      "learning_rate": 1.4335972751828693e-05,
+      "loss": 0.0002,
+      "step": 1528
+    },
+    {
+      "epoch": 1.9561600000000001,
+      "grad_norm": 0.0013540390646085143,
+      "learning_rate": 1.4313489180955017e-05,
+      "loss": 0.0002,
+      "step": 1529
+    },
+    {
+      "epoch": 1.95744,
+      "grad_norm": 0.0001240475830854848,
+      "learning_rate": 1.4291020267791073e-05,
+      "loss": 0.0002,
+      "step": 1530
+    },
+    {
+      "epoch": 1.95744,
+      "eval_loss": 1.1908667087554932,
+      "eval_runtime": 43.9335,
+      "eval_samples_per_second": 11.426,
+      "eval_steps_per_second": 1.434,
+      "step": 1530
+    },
+    {
+      "epoch": 1.95872,
+      "grad_norm": 0.0014987153699621558,
+      "learning_rate": 1.4268566055183723e-05,
+      "loss": 0.0002,
+      "step": 1531
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.00023755997244734317,
+      "learning_rate": 1.4246126585951779e-05,
+      "loss": 0.0002,
+      "step": 1532
+    },
+    {
+      "epoch": 1.96128,
+      "grad_norm": 0.001222156686708331,
+      "learning_rate": 1.422370190288596e-05,
+      "loss": 0.0002,
+      "step": 1533
+    },
+    {
+      "epoch": 1.9625599999999999,
+      "grad_norm": 0.0004482627846300602,
+      "learning_rate": 1.420129204874877e-05,
+      "loss": 0.0002,
+      "step": 1534
+    },
+    {
+      "epoch": 1.96384,
+      "grad_norm": 0.0009276245255023241,
+      "learning_rate": 1.4178897066274427e-05,
+      "loss": 0.0002,
+      "step": 1535
+    },
+    {
+      "epoch": 1.96512,
+      "grad_norm": 0.00022700500267092139,
+      "learning_rate": 1.4156516998168824e-05,
+      "loss": 0.0002,
+      "step": 1536
+    },
+    {
+      "epoch": 1.9664000000000001,
+      "grad_norm": 0.0006968076340854168,
+      "learning_rate": 1.4134151887109381e-05,
+      "loss": 0.0002,
+      "step": 1537
+    },
+    {
+      "epoch": 1.96768,
+      "grad_norm": 0.00012518244329839945,
+      "learning_rate": 1.4111801775745022e-05,
+      "loss": 0.0002,
+      "step": 1538
+    },
+    {
+      "epoch": 1.96896,
+      "grad_norm": 0.0006591297569684684,
+      "learning_rate": 1.4089466706696034e-05,
+      "loss": 0.0002,
+      "step": 1539
+    },
+    {
+      "epoch": 1.97024,
+      "grad_norm": 0.00018697626364883035,
+      "learning_rate": 1.4067146722554048e-05,
+      "loss": 0.0002,
+      "step": 1540
+    },
+    {
+      "epoch": 1.97152,
+      "grad_norm": 0.00033598518348298967,
+      "learning_rate": 1.4044841865881926e-05,
+      "loss": 0.0002,
+      "step": 1541
+    },
+    {
+      "epoch": 1.9727999999999999,
+      "grad_norm": 0.000304957153275609,
+      "learning_rate": 1.4022552179213658e-05,
+      "loss": 0.0002,
+      "step": 1542
+    },
+    {
+      "epoch": 1.9740799999999998,
+      "grad_norm": 0.0005245900247246027,
+      "learning_rate": 1.400027770505434e-05,
+      "loss": 0.0002,
+      "step": 1543
+    },
+    {
+      "epoch": 1.97536,
+      "grad_norm": 0.00034147215774282813,
+      "learning_rate": 1.3978018485880037e-05,
+      "loss": 0.0002,
+      "step": 1544
+    },
+    {
+      "epoch": 1.97664,
+      "grad_norm": 0.0003990849363617599,
+      "learning_rate": 1.3955774564137736e-05,
+      "loss": 0.0002,
+      "step": 1545
+    },
+    {
+      "epoch": 1.9779200000000001,
+      "grad_norm": 0.0005777077749371529,
+      "learning_rate": 1.3933545982245236e-05,
+      "loss": 0.0002,
+      "step": 1546
+    },
+    {
+      "epoch": 1.9792,
+      "grad_norm": 0.0005894573987461627,
+      "learning_rate": 1.3911332782591093e-05,
+      "loss": 0.0002,
+      "step": 1547
+    },
+    {
+      "epoch": 1.98048,
+      "grad_norm": 0.0003304107813164592,
+      "learning_rate": 1.3889135007534549e-05,
+      "loss": 0.0002,
+      "step": 1548
+    },
+    {
+      "epoch": 1.98176,
+      "grad_norm": 0.0006092807743698359,
+      "learning_rate": 1.3866952699405394e-05,
+      "loss": 0.0002,
+      "step": 1549
+    },
+    {
+      "epoch": 1.98304,
+      "grad_norm": 0.0005102291470393538,
+      "learning_rate": 1.3844785900503946e-05,
+      "loss": 0.0002,
+      "step": 1550
+    },
+    {
+      "epoch": 1.9843199999999999,
+      "grad_norm": 0.0012099050218239427,
+      "learning_rate": 1.382263465310096e-05,
+      "loss": 0.0002,
+      "step": 1551
+    },
+    {
+      "epoch": 1.9856,
+      "grad_norm": 0.00048186135245487094,
+      "learning_rate": 1.3800498999437504e-05,
+      "loss": 0.0002,
+      "step": 1552
+    },
+    {
+      "epoch": 1.98688,
+      "grad_norm": 0.00017949033644981682,
+      "learning_rate": 1.377837898172494e-05,
+      "loss": 0.0002,
+      "step": 1553
+    },
+    {
+      "epoch": 1.9881600000000001,
+      "grad_norm": 0.0003641147050075233,
+      "learning_rate": 1.375627464214479e-05,
+      "loss": 0.0002,
+      "step": 1554
+    },
+    {
+      "epoch": 1.98944,
+      "grad_norm": 0.00032777784508652985,
+      "learning_rate": 1.3734186022848688e-05,
+      "loss": 0.0002,
+      "step": 1555
+    },
+    {
+      "epoch": 1.99072,
+      "grad_norm": 0.0004519484646152705,
+      "learning_rate": 1.3712113165958302e-05,
+      "loss": 0.0002,
+      "step": 1556
+    },
+    {
+      "epoch": 1.992,
+      "grad_norm": 0.0003106287040282041,
+      "learning_rate": 1.3690056113565215e-05,
+      "loss": 0.0002,
+      "step": 1557
+    },
+    {
+      "epoch": 1.99328,
+      "grad_norm": 0.0003117345040664077,
+      "learning_rate": 1.3668014907730895e-05,
+      "loss": 0.0002,
+      "step": 1558
+    },
+    {
+      "epoch": 1.9945599999999999,
+      "grad_norm": 0.00021027673210483044,
+      "learning_rate": 1.3645989590486586e-05,
+      "loss": 0.0002,
+      "step": 1559
+    },
+    {
+      "epoch": 1.9958399999999998,
+      "grad_norm": 0.0006823004223406315,
+      "learning_rate": 1.3623980203833233e-05,
+      "loss": 0.0002,
+      "step": 1560
+    },
+    {
+      "epoch": 1.9958399999999998,
+      "eval_loss": 1.1802235841751099,
+      "eval_runtime": 43.2595,
+      "eval_samples_per_second": 11.604,
+      "eval_steps_per_second": 1.456,
+      "step": 1560
+    },
+    {
+      "epoch": 1.99712,
+      "grad_norm": 0.00023292088008020073,
+      "learning_rate": 1.3601986789741397e-05,
+      "loss": 0.0002,
+      "step": 1561
+    },
+    {
+      "epoch": 1.9984,
+      "grad_norm": 0.0006186995888128877,
+      "learning_rate": 1.3580009390151185e-05,
+      "loss": 0.0002,
+      "step": 1562
+    },
+    {
+      "epoch": 1.9996800000000001,
+      "grad_norm": 0.0003250852460041642,
+      "learning_rate": 1.3558048046972169e-05,
+      "loss": 0.0002,
+      "step": 1563
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.001266122329980135,
+      "learning_rate": 1.3536102802083288e-05,
+      "loss": 0.0002,
+      "step": 1564
+    },
+    {
+      "epoch": 2.00128,
+      "grad_norm": 0.0008162333979271352,
+      "learning_rate": 1.3514173697332804e-05,
+      "loss": 0.0002,
+      "step": 1565
+    },
+    {
+      "epoch": 2.00256,
+      "grad_norm": 0.0008856783388182521,
+      "learning_rate": 1.349226077453817e-05,
+      "loss": 0.0002,
+      "step": 1566
+    },
+    {
+      "epoch": 2.00384,
+      "grad_norm": 0.0009964342461898923,
+      "learning_rate": 1.3470364075486027e-05,
+      "loss": 0.0002,
+      "step": 1567
+    },
+    {
+      "epoch": 2.00512,
+      "grad_norm": 0.00029980260296724737,
+      "learning_rate": 1.3448483641932037e-05,
+      "loss": 0.0002,
+      "step": 1568
+    },
+    {
+      "epoch": 2.0064,
+      "grad_norm": 0.00043958862079307437,
+      "learning_rate": 1.3426619515600854e-05,
+      "loss": 0.0002,
+      "step": 1569
+    },
+    {
+      "epoch": 2.00768,
+      "grad_norm": 0.0012856496032327414,
+      "learning_rate": 1.3404771738186047e-05,
+      "loss": 0.0002,
+      "step": 1570
+    },
+    {
+      "epoch": 2.00896,
+      "grad_norm": 0.0005549627239815891,
+      "learning_rate": 1.3382940351350003e-05,
+      "loss": 0.0002,
+      "step": 1571
+    },
+    {
+      "epoch": 2.01024,
+      "grad_norm": 0.0012308419682085514,
+      "learning_rate": 1.336112539672385e-05,
+      "loss": 0.0002,
+      "step": 1572
+    },
+    {
+      "epoch": 2.01152,
+      "grad_norm": 0.0009879851713776588,
+      "learning_rate": 1.333932691590738e-05,
+      "loss": 0.0002,
+      "step": 1573
+    },
+    {
+      "epoch": 2.0128,
+      "grad_norm": 0.0006125083309598267,
+      "learning_rate": 1.331754495046898e-05,
+      "loss": 0.0002,
+      "step": 1574
+    },
+    {
+      "epoch": 2.01408,
+      "grad_norm": 0.0011476647341623902,
+      "learning_rate": 1.3295779541945535e-05,
+      "loss": 0.0002,
+      "step": 1575
+    },
+    {
+      "epoch": 2.01536,
+      "grad_norm": 0.00018577600712887943,
+      "learning_rate": 1.3274030731842349e-05,
+      "loss": 0.0002,
+      "step": 1576
+    },
+    {
+      "epoch": 2.01664,
+      "grad_norm": 0.0008417818462476134,
+      "learning_rate": 1.3252298561633093e-05,
+      "loss": 0.0002,
+      "step": 1577
+    },
+    {
+      "epoch": 2.01792,
+      "grad_norm": 0.00107707513961941,
+      "learning_rate": 1.3230583072759693e-05,
+      "loss": 0.0002,
+      "step": 1578
+    },
+    {
+      "epoch": 2.0192,
+      "grad_norm": 0.000682896061334759,
+      "learning_rate": 1.3208884306632274e-05,
+      "loss": 0.0002,
+      "step": 1579
+    },
+    {
+      "epoch": 2.02048,
+      "grad_norm": 0.0016809995286166668,
+      "learning_rate": 1.3187202304629066e-05,
+      "loss": 0.0002,
+      "step": 1580
+    },
+    {
+      "epoch": 2.02176,
+      "grad_norm": 0.0006951438263058662,
+      "learning_rate": 1.3165537108096311e-05,
+      "loss": 0.0002,
+      "step": 1581
+    },
+    {
+      "epoch": 2.02304,
+      "grad_norm": 0.001414105063304305,
+      "learning_rate": 1.3143888758348256e-05,
+      "loss": 0.0002,
+      "step": 1582
+    },
+    {
+      "epoch": 2.02432,
+      "grad_norm": 0.00011542163701960817,
+      "learning_rate": 1.3122257296666966e-05,
+      "loss": 0.0002,
+      "step": 1583
+    },
+    {
+      "epoch": 2.0256,
+      "grad_norm": 0.0013241195119917393,
+      "learning_rate": 1.3100642764302344e-05,
+      "loss": 0.0002,
+      "step": 1584
+    },
+    {
+      "epoch": 2.02688,
+      "grad_norm": 0.0009679512586444616,
+      "learning_rate": 1.3079045202471977e-05,
+      "loss": 0.0002,
+      "step": 1585
+    },
+    {
+      "epoch": 2.02816,
+      "grad_norm": 0.002106900094076991,
+      "learning_rate": 1.3057464652361106e-05,
+      "loss": 0.0002,
+      "step": 1586
+    },
+    {
+      "epoch": 2.02944,
+      "grad_norm": 0.00037472325493581593,
+      "learning_rate": 1.3035901155122547e-05,
+      "loss": 0.0002,
+      "step": 1587
+    },
+    {
+      "epoch": 2.03072,
+      "grad_norm": 0.0017979259137064219,
+      "learning_rate": 1.3014354751876557e-05,
+      "loss": 0.0002,
+      "step": 1588
+    },
+    {
+      "epoch": 2.032,
+      "grad_norm": 0.0005142168374732137,
+      "learning_rate": 1.2992825483710837e-05,
+      "loss": 0.0002,
+      "step": 1589
+    },
+    {
+      "epoch": 2.03328,
+      "grad_norm": 0.0016031867126002908,
+      "learning_rate": 1.2971313391680384e-05,
+      "loss": 0.0002,
+      "step": 1590
+    },
+    {
+      "epoch": 2.03328,
+      "eval_loss": 1.1785191297531128,
+      "eval_runtime": 43.1988,
+      "eval_samples_per_second": 11.621,
+      "eval_steps_per_second": 1.458,
+      "step": 1590
+    },
+    {
+      "epoch": 2.03456,
+      "grad_norm": 0.0003188488190062344,
+      "learning_rate": 1.2949818516807468e-05,
+      "loss": 0.0002,
+      "step": 1591
+    },
+    {
+      "epoch": 2.03584,
+      "grad_norm": 0.000818319502286613,
+      "learning_rate": 1.2928340900081495e-05,
+      "loss": 0.0002,
+      "step": 1592
+    },
+    {
+      "epoch": 2.03712,
+      "grad_norm": 0.0006152302958071232,
+      "learning_rate": 1.2906880582458983e-05,
+      "loss": 0.0002,
+      "step": 1593
+    },
+    {
+      "epoch": 2.0384,
+      "grad_norm": 0.0004554502957034856,
+      "learning_rate": 1.2885437604863458e-05,
+      "loss": 0.0002,
+      "step": 1594
+    },
+    {
+      "epoch": 2.03968,
+      "grad_norm": 0.0007213074713945389,
+      "learning_rate": 1.2864012008185371e-05,
+      "loss": 0.0002,
+      "step": 1595
+    },
+    {
+      "epoch": 2.04096,
+      "grad_norm": 0.0003609023697208613,
+      "learning_rate": 1.284260383328204e-05,
+      "loss": 0.0002,
+      "step": 1596
+    },
+    {
+      "epoch": 2.04224,
+      "grad_norm": 0.000855445337947458,
+      "learning_rate": 1.2821213120977558e-05,
+      "loss": 0.0002,
+      "step": 1597
+    },
+    {
+      "epoch": 2.04352,
+      "grad_norm": 0.00029147861641831696,
+      "learning_rate": 1.2799839912062708e-05,
+      "loss": 0.0002,
+      "step": 1598
+    },
+    {
+      "epoch": 2.0448,
+      "grad_norm": 0.0007133941981010139,
+      "learning_rate": 1.2778484247294909e-05,
+      "loss": 0.0002,
+      "step": 1599
+    },
+    {
+      "epoch": 2.04608,
+      "grad_norm": 0.000201601127628237,
+      "learning_rate": 1.2757146167398103e-05,
+      "loss": 0.0002,
+      "step": 1600
+    },
+    {
+      "epoch": 2.04736,
+      "grad_norm": 0.00014098279643803835,
+      "learning_rate": 1.2735825713062738e-05,
+      "loss": 0.0002,
+      "step": 1601
+    },
+    {
+      "epoch": 2.04864,
+      "grad_norm": 0.0002673087583389133,
+      "learning_rate": 1.271452292494561e-05,
+      "loss": 0.0002,
+      "step": 1602
+    },
+    {
+      "epoch": 2.04992,
+      "grad_norm": 0.00015182622883003205,
+      "learning_rate": 1.2693237843669852e-05,
+      "loss": 0.0002,
+      "step": 1603
+    },
+    {
+      "epoch": 2.0512,
+      "grad_norm": 0.00028255832148715854,
+      "learning_rate": 1.267197050982481e-05,
+      "loss": 0.0002,
+      "step": 1604
+    },
+    {
+      "epoch": 2.05248,
+      "grad_norm": 0.00029869540594518185,
+      "learning_rate": 1.265072096396601e-05,
+      "loss": 0.0002,
+      "step": 1605
+    },
+    {
+      "epoch": 2.05376,
+      "grad_norm": 0.0001590411993674934,
+      "learning_rate": 1.2629489246615052e-05,
+      "loss": 0.0002,
+      "step": 1606
+    },
+    {
+      "epoch": 2.05504,
+      "grad_norm": 0.00010362736065872014,
+      "learning_rate": 1.2608275398259516e-05,
+      "loss": 0.0002,
+      "step": 1607
+    },
+    {
+      "epoch": 2.05632,
+      "grad_norm": 0.00034905390930362046,
+      "learning_rate": 1.2587079459352931e-05,
+      "loss": 0.0002,
+      "step": 1608
+    },
+    {
+      "epoch": 2.0576,
+      "grad_norm": 0.0002614253608044237,
+      "learning_rate": 1.2565901470314677e-05,
+      "loss": 0.0002,
+      "step": 1609
+    },
+    {
+      "epoch": 2.05888,
+      "grad_norm": 0.00016580188821535558,
+      "learning_rate": 1.2544741471529873e-05,
+      "loss": 0.0002,
+      "step": 1610
+    },
+    {
+      "epoch": 2.06016,
+      "grad_norm": 0.0003083675110246986,
+      "learning_rate": 1.2523599503349368e-05,
+      "loss": 0.0002,
+      "step": 1611
+    },
+    {
+      "epoch": 2.06144,
+      "grad_norm": 0.00018635000742506236,
+      "learning_rate": 1.250247560608961e-05,
+      "loss": 0.0002,
+      "step": 1612
+    },
+    {
+      "epoch": 2.06272,
+      "grad_norm": 0.0003100994508713484,
+      "learning_rate": 1.2481369820032594e-05,
+      "loss": 0.0002,
+      "step": 1613
+    },
+    {
+      "epoch": 2.064,
+      "grad_norm": 0.00014046489377506077,
+      "learning_rate": 1.2460282185425767e-05,
+      "loss": 0.0002,
+      "step": 1614
+    },
+    {
+      "epoch": 2.06528,
+      "grad_norm": 0.00012373473145999014,
+      "learning_rate": 1.2439212742481957e-05,
+      "loss": 0.0002,
+      "step": 1615
+    },
+    {
+      "epoch": 2.06656,
+      "grad_norm": 0.0002941812272183597,
+      "learning_rate": 1.2418161531379336e-05,
+      "loss": 0.0002,
+      "step": 1616
+    },
+    {
+      "epoch": 2.06784,
+      "grad_norm": 0.0004038626793771982,
+      "learning_rate": 1.2397128592261271e-05,
+      "loss": 0.0002,
+      "step": 1617
+    },
+    {
+      "epoch": 2.06912,
+      "grad_norm": 0.00012880339636467397,
+      "learning_rate": 1.2376113965236311e-05,
+      "loss": 0.0002,
+      "step": 1618
+    },
+    {
+      "epoch": 2.0704,
+      "grad_norm": 0.00032092208857648075,
+      "learning_rate": 1.2355117690378057e-05,
+      "loss": 0.0002,
+      "step": 1619
+    },
+    {
+      "epoch": 2.07168,
+      "grad_norm": 0.00014724551874678582,
+      "learning_rate": 1.2334139807725158e-05,
+      "loss": 0.0002,
+      "step": 1620
+    },
+    {
+      "epoch": 2.07168,
+      "eval_loss": 1.181175708770752,
+      "eval_runtime": 43.3984,
+      "eval_samples_per_second": 11.567,
+      "eval_steps_per_second": 1.452,
+      "step": 1620
+    },
+    {
+      "epoch": 2.07296,
+      "grad_norm": 0.00035675789695233107,
+      "learning_rate": 1.2313180357281146e-05,
+      "loss": 0.0002,
+      "step": 1621
+    },
+    {
+      "epoch": 2.07424,
+      "grad_norm": 0.0004117561038583517,
+      "learning_rate": 1.2292239379014417e-05,
+      "loss": 0.0002,
+      "step": 1622
+    },
+    {
+      "epoch": 2.07552,
+      "grad_norm": 0.00019115378381684422,
+      "learning_rate": 1.2271316912858155e-05,
+      "loss": 0.0002,
+      "step": 1623
+    },
+    {
+      "epoch": 2.0768,
+      "grad_norm": 0.0001364218769595027,
+      "learning_rate": 1.225041299871023e-05,
+      "loss": 0.0002,
+      "step": 1624
+    },
+    {
+      "epoch": 2.07808,
+      "grad_norm": 0.00017668337386567146,
+      "learning_rate": 1.2229527676433147e-05,
+      "loss": 0.0002,
+      "step": 1625
+    },
+    {
+      "epoch": 2.07936,
+      "grad_norm": 0.00018027937039732933,
+      "learning_rate": 1.2208660985853934e-05,
+      "loss": 0.0002,
+      "step": 1626
+    },
+    {
+      "epoch": 2.08064,
+      "grad_norm": 0.0001851502020144835,
+      "learning_rate": 1.2187812966764105e-05,
+      "loss": 0.0002,
+      "step": 1627
+    },
+    {
+      "epoch": 2.08192,
+      "grad_norm": 0.0002945510495919734,
+      "learning_rate": 1.2166983658919587e-05,
+      "loss": 0.0002,
+      "step": 1628
+    },
+    {
+      "epoch": 2.0832,
+      "grad_norm": 0.00013975801994092762,
+      "learning_rate": 1.2146173102040577e-05,
+      "loss": 0.0002,
+      "step": 1629
+    },
+    {
+      "epoch": 2.08448,
+      "grad_norm": 0.0004168010491412133,
+      "learning_rate": 1.2125381335811564e-05,
+      "loss": 0.0002,
+      "step": 1630
+    },
+    {
+      "epoch": 2.08576,
+      "grad_norm": 0.00022464545327238739,
+      "learning_rate": 1.210460839988118e-05,
+      "loss": 0.0002,
+      "step": 1631
+    },
+    {
+      "epoch": 2.08704,
+      "grad_norm": 0.00011976376117672771,
+      "learning_rate": 1.2083854333862157e-05,
+      "loss": 0.0002,
+      "step": 1632
+    },
+    {
+      "epoch": 2.08832,
+      "grad_norm": 0.0001563438563607633,
+      "learning_rate": 1.2063119177331237e-05,
+      "loss": 0.0002,
+      "step": 1633
+    },
+    {
+      "epoch": 2.0896,
+      "grad_norm": 0.00014362685033120215,
+      "learning_rate": 1.204240296982909e-05,
+      "loss": 0.0002,
+      "step": 1634
+    },
+    {
+      "epoch": 2.09088,
+      "grad_norm": 0.0001592675835127011,
+      "learning_rate": 1.2021705750860295e-05,
+      "loss": 0.0002,
+      "step": 1635
+    },
+    {
+      "epoch": 2.09216,
+      "grad_norm": 0.00012455842806957662,
+      "learning_rate": 1.2001027559893172e-05,
+      "loss": 0.0002,
+      "step": 1636
+    },
+    {
+      "epoch": 2.09344,
+      "grad_norm": 0.00014256317808758467,
+      "learning_rate": 1.198036843635979e-05,
+      "loss": 0.0002,
+      "step": 1637
+    },
+    {
+      "epoch": 2.09472,
+      "grad_norm": 0.0006191166467033327,
+      "learning_rate": 1.1959728419655828e-05,
+      "loss": 0.0002,
+      "step": 1638
+    },
+    {
+      "epoch": 2.096,
+      "grad_norm": 0.000110753440822009,
+      "learning_rate": 1.1939107549140557e-05,
+      "loss": 0.0002,
+      "step": 1639
+    },
+    {
+      "epoch": 2.09728,
+      "grad_norm": 0.0005446019931696355,
+      "learning_rate": 1.1918505864136728e-05,
+      "loss": 0.0002,
+      "step": 1640
+    },
+    {
+      "epoch": 2.09856,
+      "grad_norm": 0.0005599053110927343,
+      "learning_rate": 1.1897923403930491e-05,
+      "loss": 0.0002,
+      "step": 1641
+    },
+    {
+      "epoch": 2.09984,
+      "grad_norm": 0.000916913675609976,
+      "learning_rate": 1.1877360207771363e-05,
+      "loss": 0.0002,
+      "step": 1642
+    },
+    {
+      "epoch": 2.10112,
+      "grad_norm": 0.0005046268925070763,
+      "learning_rate": 1.1856816314872108e-05,
+      "loss": 0.0002,
+      "step": 1643
+    },
+    {
+      "epoch": 2.1024,
+      "grad_norm": 0.0007644724100828171,
+      "learning_rate": 1.1836291764408672e-05,
+      "loss": 0.0002,
+      "step": 1644
+    },
+    {
+      "epoch": 2.1036799999999998,
+      "grad_norm": 0.0008394183241762221,
+      "learning_rate": 1.181578659552014e-05,
+      "loss": 0.0002,
+      "step": 1645
+    },
+    {
+      "epoch": 2.10496,
+      "grad_norm": 0.0007981343660503626,
+      "learning_rate": 1.1795300847308621e-05,
+      "loss": 0.0002,
+      "step": 1646
+    },
+    {
+      "epoch": 2.10624,
+      "grad_norm": 0.00044534794869832695,
+      "learning_rate": 1.1774834558839198e-05,
+      "loss": 0.0002,
+      "step": 1647
+    },
+    {
+      "epoch": 2.10752,
+      "grad_norm": 0.00014556580572389066,
+      "learning_rate": 1.1754387769139835e-05,
+      "loss": 0.0002,
+      "step": 1648
+    },
+    {
+      "epoch": 2.1088,
+      "grad_norm": 0.0006214206805452704,
+      "learning_rate": 1.1733960517201316e-05,
+      "loss": 0.0002,
+      "step": 1649
+    },
+    {
+      "epoch": 2.11008,
+      "grad_norm": 0.0003172321885358542,
+      "learning_rate": 1.1713552841977185e-05,
+      "loss": 0.0002,
+      "step": 1650
+    },
+    {
+      "epoch": 2.11008,
+      "eval_loss": 1.1766486167907715,
+      "eval_runtime": 43.0906,
+      "eval_samples_per_second": 11.65,
+      "eval_steps_per_second": 1.462,
+      "step": 1650
+    },
+    {
+      "epoch": 2.11136,
+      "grad_norm": 0.0007667154422961175,
+      "learning_rate": 1.1693164782383628e-05,
+      "loss": 0.0002,
+      "step": 1651
+    },
+    {
+      "epoch": 2.11264,
+      "grad_norm": 0.0004064768727403134,
+      "learning_rate": 1.1672796377299444e-05,
+      "loss": 0.0002,
+      "step": 1652
+    },
+    {
+      "epoch": 2.11392,
+      "grad_norm": 0.000624951149802655,
+      "learning_rate": 1.1652447665565941e-05,
+      "loss": 0.0002,
+      "step": 1653
+    },
+    {
+      "epoch": 2.1152,
+      "grad_norm": 0.0003828768967650831,
+      "learning_rate": 1.1632118685986875e-05,
+      "loss": 0.0002,
+      "step": 1654
+    },
+    {
+      "epoch": 2.11648,
+      "grad_norm": 0.0006233001477085054,
+      "learning_rate": 1.1611809477328388e-05,
+      "loss": 0.0002,
+      "step": 1655
+    },
+    {
+      "epoch": 2.11776,
+      "grad_norm": 0.00017231212405022234,
+      "learning_rate": 1.1591520078318898e-05,
+      "loss": 0.0002,
+      "step": 1656
+    },
+    {
+      "epoch": 2.11904,
+      "grad_norm": 0.00027037342078983784,
+      "learning_rate": 1.1571250527649061e-05,
+      "loss": 0.0002,
+      "step": 1657
+    },
+    {
+      "epoch": 2.12032,
+      "grad_norm": 0.00028422908508218825,
+      "learning_rate": 1.1551000863971677e-05,
+      "loss": 0.0002,
+      "step": 1658
+    },
+    {
+      "epoch": 2.1216,
+      "grad_norm": 0.00034367659827694297,
+      "learning_rate": 1.1530771125901641e-05,
+      "loss": 0.0002,
+      "step": 1659
+    },
+    {
+      "epoch": 2.12288,
+      "grad_norm": 0.00011658842413453385,
+      "learning_rate": 1.1510561352015817e-05,
+      "loss": 0.0002,
+      "step": 1660
+    },
+    {
+      "epoch": 2.12416,
+      "grad_norm": 0.00011873227049363777,
+      "learning_rate": 1.1490371580853021e-05,
+      "loss": 0.0002,
+      "step": 1661
+    },
+    {
+      "epoch": 2.12544,
+      "grad_norm": 0.00020507187582552433,
+      "learning_rate": 1.1470201850913935e-05,
+      "loss": 0.0002,
+      "step": 1662
+    },
+    {
+      "epoch": 2.12672,
+      "grad_norm": 0.00015456117398571223,
+      "learning_rate": 1.1450052200660997e-05,
+      "loss": 0.0002,
+      "step": 1663
+    },
+    {
+      "epoch": 2.128,
+      "grad_norm": 0.00029631159850396216,
+      "learning_rate": 1.1429922668518367e-05,
+      "loss": 0.0002,
+      "step": 1664
+    },
+    {
+      "epoch": 2.12928,
+      "grad_norm": 0.0005491878837347031,
+      "learning_rate": 1.1409813292871848e-05,
+      "loss": 0.0002,
+      "step": 1665
+    },
+    {
+      "epoch": 2.13056,
+      "grad_norm": 0.0009095377754420042,
+      "learning_rate": 1.1389724112068803e-05,
+      "loss": 0.0002,
+      "step": 1666
+    },
+    {
+      "epoch": 2.13184,
+      "grad_norm": 0.00017942706472240388,
+      "learning_rate": 1.1369655164418076e-05,
+      "loss": 0.0002,
+      "step": 1667
+    },
+    {
+      "epoch": 2.13312,
+      "grad_norm": 0.0008740307530388236,
+      "learning_rate": 1.1349606488189923e-05,
+      "loss": 0.0002,
+      "step": 1668
+    },
+    {
+      "epoch": 2.1344,
+      "grad_norm": 0.00012318171502556652,
+      "learning_rate": 1.1329578121615961e-05,
+      "loss": 0.0002,
+      "step": 1669
+    },
+    {
+      "epoch": 2.13568,
+      "grad_norm": 0.0005075764493085444,
+      "learning_rate": 1.130957010288907e-05,
+      "loss": 0.0002,
+      "step": 1670
+    },
+    {
+      "epoch": 2.13696,
+      "grad_norm": 0.0003700858214870095,
+      "learning_rate": 1.1289582470163336e-05,
+      "loss": 0.0002,
+      "step": 1671
+    },
+    {
+      "epoch": 2.13824,
+      "grad_norm": 0.0010242791613563895,
+      "learning_rate": 1.1269615261553946e-05,
+      "loss": 0.0002,
+      "step": 1672
+    },
+    {
+      "epoch": 2.13952,
+      "grad_norm": 0.00033670556149445474,
+      "learning_rate": 1.1249668515137162e-05,
+      "loss": 0.0002,
+      "step": 1673
+    },
+    {
+      "epoch": 2.1408,
+      "grad_norm": 0.0005608523497357965,
+      "learning_rate": 1.122974226895023e-05,
+      "loss": 0.0002,
+      "step": 1674
+    },
+    {
+      "epoch": 2.14208,
+      "grad_norm": 0.00029023917159065604,
+      "learning_rate": 1.1209836560991274e-05,
+      "loss": 0.0002,
+      "step": 1675
+    },
+    {
+      "epoch": 2.14336,
+      "grad_norm": 0.00019341774168424308,
+      "learning_rate": 1.1189951429219276e-05,
+      "loss": 0.0002,
+      "step": 1676
+    },
+    {
+      "epoch": 2.14464,
+      "grad_norm": 0.00017067784210667014,
+      "learning_rate": 1.1170086911553981e-05,
+      "loss": 0.0002,
+      "step": 1677
+    },
+    {
+      "epoch": 2.14592,
+      "grad_norm": 0.00016009536921046674,
+      "learning_rate": 1.1150243045875824e-05,
+      "loss": 0.0002,
+      "step": 1678
+    },
+    {
+      "epoch": 2.1471999999999998,
+      "grad_norm": 0.00041830449481494725,
+      "learning_rate": 1.1130419870025848e-05,
+      "loss": 0.0002,
+      "step": 1679
+    },
+    {
+      "epoch": 2.14848,
+      "grad_norm": 0.0001825682702474296,
+      "learning_rate": 1.111061742180563e-05,
+      "loss": 0.0002,
+      "step": 1680
+    },
+    {
+      "epoch": 2.14848,
+      "eval_loss": 1.178200125694275,
+      "eval_runtime": 43.3215,
+      "eval_samples_per_second": 11.588,
+      "eval_steps_per_second": 1.454,
+      "step": 1680
+    },
+    {
+      "epoch": 2.14976,
+      "grad_norm": 0.00045886740554124117,
+      "learning_rate": 1.1090835738977264e-05,
+      "loss": 0.0002,
+      "step": 1681
+    },
+    {
+      "epoch": 2.15104,
+      "grad_norm": 0.00017417485651094466,
+      "learning_rate": 1.10710748592632e-05,
+      "loss": 0.0002,
+      "step": 1682
+    },
+    {
+      "epoch": 2.15232,
+      "grad_norm": 0.0005541493883356452,
+      "learning_rate": 1.105133482034625e-05,
+      "loss": 0.0002,
+      "step": 1683
+    },
+    {
+      "epoch": 2.1536,
+      "grad_norm": 0.00013444297655951232,
+      "learning_rate": 1.1031615659869456e-05,
+      "loss": 0.0002,
+      "step": 1684
+    },
+    {
+      "epoch": 2.15488,
+      "grad_norm": 0.0001350865204585716,
+      "learning_rate": 1.1011917415436071e-05,
+      "loss": 0.0002,
+      "step": 1685
+    },
+    {
+      "epoch": 2.15616,
+      "grad_norm": 0.00012103471817681566,
+      "learning_rate": 1.0992240124609452e-05,
+      "loss": 0.0002,
+      "step": 1686
+    },
+    {
+      "epoch": 2.15744,
+      "grad_norm": 0.0001908185367938131,
+      "learning_rate": 1.0972583824912993e-05,
+      "loss": 0.0002,
+      "step": 1687
+    },
+    {
+      "epoch": 2.15872,
+      "grad_norm": 0.0002593412937130779,
+      "learning_rate": 1.0952948553830063e-05,
+      "loss": 0.0002,
+      "step": 1688
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.0004132689500693232,
+      "learning_rate": 1.093333434880394e-05,
+      "loss": 0.0002,
+      "step": 1689
+    },
+    {
+      "epoch": 2.16128,
+      "grad_norm": 0.000668655673507601,
+      "learning_rate": 1.0913741247237724e-05,
+      "loss": 0.0002,
+      "step": 1690
+    },
+    {
+      "epoch": 2.16256,
+      "grad_norm": 0.0007550397422164679,
+      "learning_rate": 1.0894169286494263e-05,
+      "loss": 0.0002,
+      "step": 1691
+    },
+    {
+      "epoch": 2.16384,
+      "grad_norm": 0.0006890227668918669,
+      "learning_rate": 1.0874618503896099e-05,
+      "loss": 0.0002,
+      "step": 1692
+    },
+    {
+      "epoch": 2.16512,
+      "grad_norm": 0.0010281478753313422,
+      "learning_rate": 1.08550889367254e-05,
+      "loss": 0.0002,
+      "step": 1693
+    },
+    {
+      "epoch": 2.1664,
+      "grad_norm": 0.0003120306064374745,
+      "learning_rate": 1.083558062222385e-05,
+      "loss": 0.0002,
+      "step": 1694
+    },
+    {
+      "epoch": 2.16768,
+      "grad_norm": 0.0009151713456958532,
+      "learning_rate": 1.0816093597592627e-05,
+      "loss": 0.0002,
+      "step": 1695
+    },
+    {
+      "epoch": 2.16896,
+      "grad_norm": 0.0002048113674391061,
+      "learning_rate": 1.0796627899992313e-05,
+      "loss": 0.0002,
+      "step": 1696
+    },
+    {
+      "epoch": 2.17024,
+      "grad_norm": 0.0007615440990775824,
+      "learning_rate": 1.0777183566542799e-05,
+      "loss": 0.0002,
+      "step": 1697
+    },
+    {
+      "epoch": 2.17152,
+      "grad_norm": 0.00036416234797798097,
+      "learning_rate": 1.075776063432326e-05,
+      "loss": 0.0002,
+      "step": 1698
+    },
+    {
+      "epoch": 2.1728,
+      "grad_norm": 0.0003339081595186144,
+      "learning_rate": 1.0738359140372033e-05,
+      "loss": 0.0002,
+      "step": 1699
+    },
+    {
+      "epoch": 2.17408,
+      "grad_norm": 0.00048636324936524034,
+      "learning_rate": 1.0718979121686613e-05,
+      "loss": 0.0002,
+      "step": 1700
+    },
+    {
+      "epoch": 2.17536,
+      "grad_norm": 0.00012778060045093298,
+      "learning_rate": 1.0699620615223499e-05,
+      "loss": 0.0002,
+      "step": 1701
+    },
+    {
+      "epoch": 2.17664,
+      "grad_norm": 0.0005591086810454726,
+      "learning_rate": 1.0680283657898205e-05,
+      "loss": 0.0002,
+      "step": 1702
+    },
+    {
+      "epoch": 2.17792,
+      "grad_norm": 0.00011497444211272523,
+      "learning_rate": 1.066096828658512e-05,
+      "loss": 0.0002,
+      "step": 1703
+    },
+    {
+      "epoch": 2.1792,
+      "grad_norm": 0.00012892419181298465,
+      "learning_rate": 1.0641674538117491e-05,
+      "loss": 0.0002,
+      "step": 1704
+    },
+    {
+      "epoch": 2.18048,
+      "grad_norm": 0.0007491412106901407,
+      "learning_rate": 1.0622402449287335e-05,
+      "loss": 0.0002,
+      "step": 1705
+    },
+    {
+      "epoch": 2.18176,
+      "grad_norm": 0.0006464109173975885,
+      "learning_rate": 1.0603152056845339e-05,
+      "loss": 0.0002,
+      "step": 1706
+    },
+    {
+      "epoch": 2.18304,
+      "grad_norm": 0.00023252949176821858,
+      "learning_rate": 1.0583923397500841e-05,
+      "loss": 0.0002,
+      "step": 1707
+    },
+    {
+      "epoch": 2.18432,
+      "grad_norm": 0.0005042919074185193,
+      "learning_rate": 1.056471650792174e-05,
+      "loss": 0.0002,
+      "step": 1708
+    },
+    {
+      "epoch": 2.1856,
+      "grad_norm": 0.0007390492246486247,
+      "learning_rate": 1.0545531424734387e-05,
+      "loss": 0.0002,
+      "step": 1709
+    },
+    {
+      "epoch": 2.18688,
+      "grad_norm": 0.0006836414686404169,
+      "learning_rate": 1.0526368184523585e-05,
+      "loss": 0.0002,
+      "step": 1710
+    },
+    {
+      "epoch": 2.18688,
+      "eval_loss": 1.1836718320846558,
+      "eval_runtime": 43.5042,
+      "eval_samples_per_second": 11.539,
+      "eval_steps_per_second": 1.448,
+      "step": 1710
+    },
+    {
+      "epoch": 2.18816,
+      "grad_norm": 0.0005014064372517169,
+      "learning_rate": 1.0507226823832469e-05,
+      "loss": 0.0002,
+      "step": 1711
+    },
+    {
+      "epoch": 2.18944,
+      "grad_norm": 0.00038360522012226284,
+      "learning_rate": 1.0488107379162453e-05,
+      "loss": 0.0002,
+      "step": 1712
+    },
+    {
+      "epoch": 2.19072,
+      "grad_norm": 0.0002872354525607079,
+      "learning_rate": 1.046900988697316e-05,
+      "loss": 0.0002,
+      "step": 1713
+    },
+    {
+      "epoch": 2.192,
+      "grad_norm": 0.00022756788530386984,
+      "learning_rate": 1.0449934383682328e-05,
+      "loss": 0.0002,
+      "step": 1714
+    },
+    {
+      "epoch": 2.19328,
+      "grad_norm": 0.000334077951265499,
+      "learning_rate": 1.0430880905665814e-05,
+      "loss": 0.0002,
+      "step": 1715
+    },
+    {
+      "epoch": 2.19456,
+      "grad_norm": 0.00034818786662071943,
+      "learning_rate": 1.0411849489257428e-05,
+      "loss": 0.0002,
+      "step": 1716
+    },
+    {
+      "epoch": 2.19584,
+      "grad_norm": 0.000506854965351522,
+      "learning_rate": 1.0392840170748934e-05,
+      "loss": 0.0002,
+      "step": 1717
+    },
+    {
+      "epoch": 2.19712,
+      "grad_norm": 0.0004816302680410445,
+      "learning_rate": 1.0373852986389934e-05,
+      "loss": 0.0002,
+      "step": 1718
+    },
+    {
+      "epoch": 2.1984,
+      "grad_norm": 0.0006793339271098375,
+      "learning_rate": 1.0354887972387858e-05,
+      "loss": 0.0002,
+      "step": 1719
+    },
+    {
+      "epoch": 2.19968,
+      "grad_norm": 0.00035194880911149085,
+      "learning_rate": 1.0335945164907827e-05,
+      "loss": 0.0002,
+      "step": 1720
+    },
+    {
+      "epoch": 2.20096,
+      "grad_norm": 0.00023243727628141642,
+      "learning_rate": 1.0317024600072618e-05,
+      "loss": 0.0002,
+      "step": 1721
+    },
+    {
+      "epoch": 2.20224,
+      "grad_norm": 0.0008212224929593503,
+      "learning_rate": 1.0298126313962604e-05,
+      "loss": 0.0002,
+      "step": 1722
+    },
+    {
+      "epoch": 2.20352,
+      "grad_norm": 0.00029432986048050225,
+      "learning_rate": 1.0279250342615676e-05,
+      "loss": 0.0002,
+      "step": 1723
+    },
+    {
+      "epoch": 2.2048,
+      "grad_norm": 0.00025335565442219377,
+      "learning_rate": 1.0260396722027168e-05,
+      "loss": 0.0002,
+      "step": 1724
+    },
+    {
+      "epoch": 2.20608,
+      "grad_norm": 0.0008946171146817505,
+      "learning_rate": 1.0241565488149781e-05,
+      "loss": 0.0002,
+      "step": 1725
+    },
+    {
+      "epoch": 2.20736,
+      "grad_norm": 0.0006439693388529122,
+      "learning_rate": 1.0222756676893536e-05,
+      "loss": 0.0002,
+      "step": 1726
+    },
+    {
+      "epoch": 2.20864,
+      "grad_norm": 0.000815407547634095,
+      "learning_rate": 1.0203970324125708e-05,
+      "loss": 0.0002,
+      "step": 1727
+    },
+    {
+      "epoch": 2.20992,
+      "grad_norm": 0.0002080225822282955,
+      "learning_rate": 1.0185206465670712e-05,
+      "loss": 0.0002,
+      "step": 1728
+    },
+    {
+      "epoch": 2.2112,
+      "grad_norm": 0.0007100445800460875,
+      "learning_rate": 1.01664651373101e-05,
+      "loss": 0.0002,
+      "step": 1729
+    },
+    {
+      "epoch": 2.2124800000000002,
+      "grad_norm": 0.000299918872769922,
+      "learning_rate": 1.0147746374782445e-05,
+      "loss": 0.0002,
+      "step": 1730
+    },
+    {
+      "epoch": 2.21376,
+      "grad_norm": 0.0006271378370001912,
+      "learning_rate": 1.0129050213783299e-05,
+      "loss": 0.0002,
+      "step": 1731
+    },
+    {
+      "epoch": 2.21504,
+      "grad_norm": 0.0003948029479943216,
+      "learning_rate": 1.0110376689965099e-05,
+      "loss": 0.0002,
+      "step": 1732
+    },
+    {
+      "epoch": 2.21632,
+      "grad_norm": 0.00014382618246600032,
+      "learning_rate": 1.0091725838937112e-05,
+      "loss": 0.0002,
+      "step": 1733
+    },
+    {
+      "epoch": 2.2176,
+      "grad_norm": 0.0008930534822866321,
+      "learning_rate": 1.0073097696265398e-05,
+      "loss": 0.0002,
+      "step": 1734
+    },
+    {
+      "epoch": 2.21888,
+      "grad_norm": 0.0003659480717033148,
+      "learning_rate": 1.0054492297472684e-05,
+      "loss": 0.0002,
+      "step": 1735
+    },
+    {
+      "epoch": 2.22016,
+      "grad_norm": 0.00012835237430408597,
+      "learning_rate": 1.0035909678038344e-05,
+      "loss": 0.0002,
+      "step": 1736
+    },
+    {
+      "epoch": 2.22144,
+      "grad_norm": 0.0002881828404497355,
+      "learning_rate": 1.0017349873398295e-05,
+      "loss": 0.0002,
+      "step": 1737
+    },
+    {
+      "epoch": 2.22272,
+      "grad_norm": 0.00011149987403769046,
+      "learning_rate": 9.998812918944963e-06,
+      "loss": 0.0002,
+      "step": 1738
+    },
+    {
+      "epoch": 2.224,
+      "grad_norm": 0.00010650986223481596,
+      "learning_rate": 9.980298850027201e-06,
+      "loss": 0.0002,
+      "step": 1739
+    },
+    {
+      "epoch": 2.22528,
+      "grad_norm": 0.0003959020250476897,
+      "learning_rate": 9.961807701950207e-06,
+      "loss": 0.0002,
+      "step": 1740
+    },
+    {
+      "epoch": 2.22528,
+      "eval_loss": 1.1779879331588745,
+      "eval_runtime": 43.2666,
+      "eval_samples_per_second": 11.602,
+      "eval_steps_per_second": 1.456,
+      "step": 1740
+    },
+    {
+      "epoch": 2.22656,
+      "grad_norm": 0.0005422495305538177,
+      "learning_rate": 9.943339509975481e-06,
+      "loss": 0.0002,
+      "step": 1741
+    },
+    {
+      "epoch": 2.22784,
+      "grad_norm": 0.0004364894703030586,
+      "learning_rate": 9.924894309320754e-06,
+      "loss": 0.0002,
+      "step": 1742
+    },
+    {
+      "epoch": 2.22912,
+      "grad_norm": 0.0008182907477021217,
+      "learning_rate": 9.906472135159886e-06,
+      "loss": 0.0002,
+      "step": 1743
+    },
+    {
+      "epoch": 2.2304,
+      "grad_norm": 0.0003204997337888926,
+      "learning_rate": 9.888073022622858e-06,
+      "loss": 0.0002,
+      "step": 1744
+    },
+    {
+      "epoch": 2.23168,
+      "grad_norm": 0.0007177082588896155,
+      "learning_rate": 9.869697006795657e-06,
+      "loss": 0.0002,
+      "step": 1745
+    },
+    {
+      "epoch": 2.23296,
+      "grad_norm": 0.00014134106459096074,
+      "learning_rate": 9.851344122720236e-06,
+      "loss": 0.0002,
+      "step": 1746
+    },
+    {
+      "epoch": 2.23424,
+      "grad_norm": 0.0009650987340137362,
+      "learning_rate": 9.833014405394423e-06,
+      "loss": 0.0002,
+      "step": 1747
+    },
+    {
+      "epoch": 2.23552,
+      "grad_norm": 0.00016414733545389026,
+      "learning_rate": 9.81470788977188e-06,
+      "loss": 0.0002,
+      "step": 1748
+    },
+    {
+      "epoch": 2.2368,
+      "grad_norm": 0.0002189509687013924,
+      "learning_rate": 9.79642461076203e-06,
+      "loss": 0.0002,
+      "step": 1749
+    },
+    {
+      "epoch": 2.23808,
+      "grad_norm": 0.001173550495877862,
+      "learning_rate": 9.778164603229959e-06,
+      "loss": 0.0002,
+      "step": 1750
+    },
+    {
+      "epoch": 2.23936,
+      "grad_norm": 0.0001450089766876772,
+      "learning_rate": 9.75992790199641e-06,
+      "loss": 0.0002,
+      "step": 1751
+    },
+    {
+      "epoch": 2.24064,
+      "grad_norm": 0.001029281527735293,
+      "learning_rate": 9.74171454183765e-06,
+      "loss": 0.0002,
+      "step": 1752
+    },
+    {
+      "epoch": 2.24192,
+      "grad_norm": 0.0006216371548362076,
+      "learning_rate": 9.72352455748547e-06,
+      "loss": 0.0002,
+      "step": 1753
+    },
+    {
+      "epoch": 2.2432,
+      "grad_norm": 0.00053853151621297,
+      "learning_rate": 9.705357983627057e-06,
+      "loss": 0.0002,
+      "step": 1754
+    },
+    {
+      "epoch": 2.24448,
+      "grad_norm": 0.0008826253470033407,
+      "learning_rate": 9.687214854904959e-06,
+      "loss": 0.0002,
+      "step": 1755
+    },
+    {
+      "epoch": 2.24576,
+      "grad_norm": 0.00010876509622903541,
+      "learning_rate": 9.66909520591703e-06,
+      "loss": 0.0002,
+      "step": 1756
+    },
+    {
+      "epoch": 2.24704,
+      "grad_norm": 0.0006186162354424596,
+      "learning_rate": 9.650999071216338e-06,
+      "loss": 0.0002,
+      "step": 1757
+    },
+    {
+      "epoch": 2.24832,
+      "grad_norm": 0.00030843057902529836,
+      "learning_rate": 9.632926485311119e-06,
+      "loss": 0.0002,
+      "step": 1758
+    },
+    {
+      "epoch": 2.2496,
+      "grad_norm": 0.0003102632472291589,
+      "learning_rate": 9.614877482664688e-06,
+      "loss": 0.0002,
+      "step": 1759
+    },
+    {
+      "epoch": 2.25088,
+      "grad_norm": 0.0004681867139879614,
+      "learning_rate": 9.596852097695404e-06,
+      "loss": 0.0002,
+      "step": 1760
+    },
+    {
+      "epoch": 2.25216,
+      "grad_norm": 0.0002742313372436911,
+      "learning_rate": 9.578850364776588e-06,
+      "loss": 0.0002,
+      "step": 1761
+    },
+    {
+      "epoch": 2.25344,
+      "grad_norm": 0.0005104625597596169,
+      "learning_rate": 9.560872318236437e-06,
+      "loss": 0.0002,
+      "step": 1762
+    },
+    {
+      "epoch": 2.25472,
+      "grad_norm": 0.0002644178457558155,
+      "learning_rate": 9.542917992358008e-06,
+      "loss": 0.0002,
+      "step": 1763
+    },
+    {
+      "epoch": 2.2560000000000002,
+      "grad_norm": 0.00029693092801608145,
+      "learning_rate": 9.524987421379105e-06,
+      "loss": 0.0002,
+      "step": 1764
+    },
+    {
+      "epoch": 2.25728,
+      "grad_norm": 0.00042807578574866056,
+      "learning_rate": 9.507080639492248e-06,
+      "loss": 0.0002,
+      "step": 1765
+    },
+    {
+      "epoch": 2.25856,
+      "grad_norm": 0.0006727955769747496,
+      "learning_rate": 9.489197680844575e-06,
+      "loss": 0.0002,
+      "step": 1766
+    },
+    {
+      "epoch": 2.25984,
+      "grad_norm": 0.0006015878752805293,
+      "learning_rate": 9.471338579537794e-06,
+      "loss": 0.0002,
+      "step": 1767
+    },
+    {
+      "epoch": 2.26112,
+      "grad_norm": 0.0007681053830310702,
+      "learning_rate": 9.45350336962815e-06,
+      "loss": 0.0002,
+      "step": 1768
+    },
+    {
+      "epoch": 2.2624,
+      "grad_norm": 0.00019292229262646288,
+      "learning_rate": 9.435692085126286e-06,
+      "loss": 0.0002,
+      "step": 1769
+    },
+    {
+      "epoch": 2.26368,
+      "grad_norm": 0.00048759227502159774,
+      "learning_rate": 9.417904759997257e-06,
+      "loss": 0.0002,
+      "step": 1770
+    },
+    {
+      "epoch": 2.26368,
+      "eval_loss": 1.1767559051513672,
+      "eval_runtime": 43.1357,
+      "eval_samples_per_second": 11.638,
+      "eval_steps_per_second": 1.461,
+      "step": 1770
+    },
+    {
+      "epoch": 2.26496,
+      "grad_norm": 0.0008960186969488859,
+      "learning_rate": 9.400141428160397e-06,
+      "loss": 0.0002,
+      "step": 1771
+    },
+    {
+      "epoch": 2.26624,
+      "grad_norm": 0.00046001203008927405,
+      "learning_rate": 9.382402123489312e-06,
+      "loss": 0.0002,
+      "step": 1772
+    },
+    {
+      "epoch": 2.26752,
+      "grad_norm": 0.0006008041673339903,
+      "learning_rate": 9.364686879811781e-06,
+      "loss": 0.0002,
+      "step": 1773
+    },
+    {
+      "epoch": 2.2688,
+      "grad_norm": 0.0003917776921298355,
+      "learning_rate": 9.346995730909692e-06,
+      "loss": 0.0002,
+      "step": 1774
+    },
+    {
+      "epoch": 2.27008,
+      "grad_norm": 0.00011662714678095654,
+      "learning_rate": 9.329328710518992e-06,
+      "loss": 0.0002,
+      "step": 1775
+    },
+    {
+      "epoch": 2.27136,
+      "grad_norm": 0.000828468007966876,
+      "learning_rate": 9.311685852329622e-06,
+      "loss": 0.0002,
+      "step": 1776
+    },
+    {
+      "epoch": 2.27264,
+      "grad_norm": 0.0005161363515071571,
+      "learning_rate": 9.294067189985446e-06,
+      "loss": 0.0002,
+      "step": 1777
+    },
+    {
+      "epoch": 2.27392,
+      "grad_norm": 0.000594046083278954,
+      "learning_rate": 9.276472757084176e-06,
+      "loss": 0.0002,
+      "step": 1778
+    },
+    {
+      "epoch": 2.2752,
+      "grad_norm": 0.0010922584915533662,
+      "learning_rate": 9.258902587177328e-06,
+      "loss": 0.0002,
+      "step": 1779
+    },
+    {
+      "epoch": 2.27648,
+      "grad_norm": 0.0004009727854281664,
+      "learning_rate": 9.241356713770157e-06,
+      "loss": 0.0002,
+      "step": 1780
+    },
+    {
+      "epoch": 2.27776,
+      "grad_norm": 0.0010642922716215253,
+      "learning_rate": 9.22383517032157e-06,
+      "loss": 0.0002,
+      "step": 1781
+    },
+    {
+      "epoch": 2.27904,
+      "grad_norm": 0.00014497771917376667,
+      "learning_rate": 9.206337990244096e-06,
+      "loss": 0.0002,
+      "step": 1782
+    },
+    {
+      "epoch": 2.28032,
+      "grad_norm": 0.00046217776252888143,
+      "learning_rate": 9.188865206903788e-06,
+      "loss": 0.0002,
+      "step": 1783
+    },
+    {
+      "epoch": 2.2816,
+      "grad_norm": 0.0007038525654934347,
+      "learning_rate": 9.171416853620184e-06,
+      "loss": 0.0002,
+      "step": 1784
+    },
+    {
+      "epoch": 2.28288,
+      "grad_norm": 0.0004476465692277998,
+      "learning_rate": 9.153992963666247e-06,
+      "loss": 0.0002,
+      "step": 1785
+    },
+    {
+      "epoch": 2.28416,
+      "grad_norm": 0.00021213979925960302,
+      "learning_rate": 9.136593570268261e-06,
+      "loss": 0.0002,
+      "step": 1786
+    },
+    {
+      "epoch": 2.28544,
+      "grad_norm": 0.00044914140016771853,
+      "learning_rate": 9.119218706605823e-06,
+      "loss": 0.0002,
+      "step": 1787
+    },
+    {
+      "epoch": 2.28672,
+      "grad_norm": 0.0005167127237655222,
+      "learning_rate": 9.10186840581174e-06,
+      "loss": 0.0002,
+      "step": 1788
+    },
+    {
+      "epoch": 2.288,
+      "grad_norm": 0.00013319379650056362,
+      "learning_rate": 9.084542700971993e-06,
+      "loss": 0.0002,
+      "step": 1789
+    },
+    {
+      "epoch": 2.2892799999999998,
+      "grad_norm": 0.0002639666781760752,
+      "learning_rate": 9.067241625125635e-06,
+      "loss": 0.0002,
+      "step": 1790
+    },
+    {
+      "epoch": 2.29056,
+      "grad_norm": 0.0002221256581833586,
+      "learning_rate": 9.049965211264773e-06,
+      "loss": 0.0002,
+      "step": 1791
+    },
+    {
+      "epoch": 2.29184,
+      "grad_norm": 0.0003682387759909034,
+      "learning_rate": 9.03271349233449e-06,
+      "loss": 0.0002,
+      "step": 1792
+    },
+    {
+      "epoch": 2.29312,
+      "grad_norm": 0.00022319938580039889,
+      "learning_rate": 9.015486501232751e-06,
+      "loss": 0.0002,
+      "step": 1793
+    },
+    {
+      "epoch": 2.2944,
+      "grad_norm": 0.00020367535762488842,
+      "learning_rate": 8.998284270810392e-06,
+      "loss": 0.0002,
+      "step": 1794
+    },
+    {
+      "epoch": 2.29568,
+      "grad_norm": 0.00025371191441081464,
+      "learning_rate": 8.981106833871025e-06,
+      "loss": 0.0002,
+      "step": 1795
+    },
+    {
+      "epoch": 2.29696,
+      "grad_norm": 0.00015242303197737783,
+      "learning_rate": 8.963954223170972e-06,
+      "loss": 0.0002,
+      "step": 1796
+    },
+    {
+      "epoch": 2.29824,
+      "grad_norm": 0.0005362336523830891,
+      "learning_rate": 8.94682647141923e-06,
+      "loss": 0.0002,
+      "step": 1797
+    },
+    {
+      "epoch": 2.2995200000000002,
+      "grad_norm": 0.0002387440181337297,
+      "learning_rate": 8.929723611277366e-06,
+      "loss": 0.0002,
+      "step": 1798
+    },
+    {
+      "epoch": 2.3008,
+      "grad_norm": 0.0001875937741715461,
+      "learning_rate": 8.91264567535952e-06,
+      "loss": 0.0002,
+      "step": 1799
+    },
+    {
+      "epoch": 2.30208,
+      "grad_norm": 0.0002941049460787326,
+      "learning_rate": 8.89559269623226e-06,
+      "loss": 0.0002,
+      "step": 1800
+    },
+    {
+      "epoch": 2.30208,
+      "eval_loss": 1.1789199113845825,
+      "eval_runtime": 43.3392,
+      "eval_samples_per_second": 11.583,
+      "eval_steps_per_second": 1.454,
+      "step": 1800
+    },
+    {
+      "epoch": 2.30336,
+      "grad_norm": 0.00032305315835401416,
+      "learning_rate": 8.878564706414597e-06,
+      "loss": 0.0002,
+      "step": 1801
+    },
+    {
+      "epoch": 2.30464,
+      "grad_norm": 0.00010795753041747957,
+      "learning_rate": 8.861561738377861e-06,
+      "loss": 0.0002,
+      "step": 1802
+    },
+    {
+      "epoch": 2.30592,
+      "grad_norm": 0.00043354270746931434,
+      "learning_rate": 8.844583824545688e-06,
+      "loss": 0.0002,
+      "step": 1803
+    },
+    {
+      "epoch": 2.3072,
+      "grad_norm": 0.00028120525530539453,
+      "learning_rate": 8.827630997293933e-06,
+      "loss": 0.0002,
+      "step": 1804
+    },
+    {
+      "epoch": 2.30848,
+      "grad_norm": 0.0008327866671606898,
+      "learning_rate": 8.810703288950596e-06,
+      "loss": 0.0002,
+      "step": 1805
+    },
+    {
+      "epoch": 2.30976,
+      "grad_norm": 0.0002845035633072257,
+      "learning_rate": 8.7938007317958e-06,
+      "loss": 0.0002,
+      "step": 1806
+    },
+    {
+      "epoch": 2.31104,
+      "grad_norm": 0.000504064722917974,
+      "learning_rate": 8.7769233580617e-06,
+      "loss": 0.0002,
+      "step": 1807
+    },
+    {
+      "epoch": 2.31232,
+      "grad_norm": 0.0002677366719581187,
+      "learning_rate": 8.76007119993241e-06,
+      "loss": 0.0002,
+      "step": 1808
+    },
+    {
+      "epoch": 2.3136,
+      "grad_norm": 0.00022501955390907824,
+      "learning_rate": 8.74324428954398e-06,
+      "loss": 0.0002,
+      "step": 1809
+    },
+    {
+      "epoch": 2.31488,
+      "grad_norm": 0.0006158575415611267,
+      "learning_rate": 8.72644265898431e-06,
+      "loss": 0.0002,
+      "step": 1810
+    },
+    {
+      "epoch": 2.31616,
+      "grad_norm": 0.00011165841715410352,
+      "learning_rate": 8.70966634029309e-06,
+      "loss": 0.0002,
+      "step": 1811
+    },
+    {
+      "epoch": 2.31744,
+      "grad_norm": 0.0008404940599575639,
+      "learning_rate": 8.692915365461739e-06,
+      "loss": 0.0002,
+      "step": 1812
+    },
+    {
+      "epoch": 2.31872,
+      "grad_norm": 0.0002906533482018858,
+      "learning_rate": 8.67618976643334e-06,
+      "loss": 0.0002,
+      "step": 1813
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.0006887954659759998,
+      "learning_rate": 8.659489575102617e-06,
+      "loss": 0.0002,
+      "step": 1814
+    },
+    {
+      "epoch": 2.32128,
+      "grad_norm": 0.00022589090804103762,
+      "learning_rate": 8.642814823315804e-06,
+      "loss": 0.0002,
+      "step": 1815
+    },
+    {
+      "epoch": 2.32256,
+      "grad_norm": 0.0006809495971538126,
+      "learning_rate": 8.626165542870651e-06,
+      "loss": 0.0002,
+      "step": 1816
+    },
+    {
+      "epoch": 2.32384,
+      "grad_norm": 0.0008959157858043909,
+      "learning_rate": 8.60954176551631e-06,
+      "loss": 0.0002,
+      "step": 1817
+    },
+    {
+      "epoch": 2.32512,
+      "grad_norm": 0.0008075773948803544,
+      "learning_rate": 8.592943522953339e-06,
+      "loss": 0.0002,
+      "step": 1818
+    },
+    {
+      "epoch": 2.3264,
+      "grad_norm": 0.00017589032358955592,
+      "learning_rate": 8.576370846833567e-06,
+      "loss": 0.0002,
+      "step": 1819
+    },
+    {
+      "epoch": 2.32768,
+      "grad_norm": 0.00101184262894094,
+      "learning_rate": 8.559823768760078e-06,
+      "loss": 0.0002,
+      "step": 1820
+    },
+    {
+      "epoch": 2.32896,
+      "grad_norm": 0.00037295141373761,
+      "learning_rate": 8.543302320287151e-06,
+      "loss": 0.0002,
+      "step": 1821
+    },
+    {
+      "epoch": 2.33024,
+      "grad_norm": 0.000557410647161305,
+      "learning_rate": 8.52680653292018e-06,
+      "loss": 0.0002,
+      "step": 1822
+    },
+    {
+      "epoch": 2.33152,
+      "grad_norm": 0.0007255637901835144,
+      "learning_rate": 8.510336438115643e-06,
+      "loss": 0.0002,
+      "step": 1823
+    },
+    {
+      "epoch": 2.3327999999999998,
+      "grad_norm": 0.0003043143660761416,
+      "learning_rate": 8.493892067281e-06,
+      "loss": 0.0002,
+      "step": 1824
+    },
+    {
+      "epoch": 2.33408,
+      "grad_norm": 0.0008695904980413616,
+      "learning_rate": 8.477473451774671e-06,
+      "loss": 0.0002,
+      "step": 1825
+    },
+    {
+      "epoch": 2.33536,
+      "grad_norm": 0.0004969065776094794,
+      "learning_rate": 8.461080622905966e-06,
+      "loss": 0.0002,
+      "step": 1826
+    },
+    {
+      "epoch": 2.33664,
+      "grad_norm": 0.000555870239622891,
+      "learning_rate": 8.444713611935005e-06,
+      "loss": 0.0002,
+      "step": 1827
+    },
+    {
+      "epoch": 2.33792,
+      "grad_norm": 0.0005065601435489953,
+      "learning_rate": 8.42837245007269e-06,
+      "loss": 0.0002,
+      "step": 1828
+    },
+    {
+      "epoch": 2.3392,
+      "grad_norm": 0.00013234195648692548,
+      "learning_rate": 8.412057168480623e-06,
+      "loss": 0.0002,
+      "step": 1829
+    },
+    {
+      "epoch": 2.34048,
+      "grad_norm": 0.0008337848703376949,
+      "learning_rate": 8.395767798271065e-06,
+      "loss": 0.0002,
+      "step": 1830
+    },
+    {
+      "epoch": 2.34048,
+      "eval_loss": 1.1797373294830322,
+      "eval_runtime": 43.3554,
+      "eval_samples_per_second": 11.579,
+      "eval_steps_per_second": 1.453,
+      "step": 1830
+    },
+    {
+      "epoch": 2.34176,
+      "grad_norm": 0.00016388027870561928,
+      "learning_rate": 8.379504370506847e-06,
+      "loss": 0.0002,
+      "step": 1831
+    },
+    {
+      "epoch": 2.3430400000000002,
+      "grad_norm": 0.0010740353027358651,
+      "learning_rate": 8.363266916201331e-06,
+      "loss": 0.0002,
+      "step": 1832
+    },
+    {
+      "epoch": 2.34432,
+      "grad_norm": 0.00013404090714175254,
+      "learning_rate": 8.34705546631838e-06,
+      "loss": 0.0002,
+      "step": 1833
+    },
+    {
+      "epoch": 2.3456,
+      "grad_norm": 0.0005119265988469124,
+      "learning_rate": 8.330870051772226e-06,
+      "loss": 0.0002,
+      "step": 1834
+    },
+    {
+      "epoch": 2.34688,
+      "grad_norm": 0.000775667664129287,
+      "learning_rate": 8.314710703427487e-06,
+      "loss": 0.0002,
+      "step": 1835
+    },
+    {
+      "epoch": 2.34816,
+      "grad_norm": 0.00010712215589592233,
+      "learning_rate": 8.29857745209904e-06,
+      "loss": 0.0002,
+      "step": 1836
+    },
+    {
+      "epoch": 2.34944,
+      "grad_norm": 0.0007510327268391848,
+      "learning_rate": 8.282470328552037e-06,
+      "loss": 0.0002,
+      "step": 1837
+    },
+    {
+      "epoch": 2.35072,
+      "grad_norm": 0.0007226445595733821,
+      "learning_rate": 8.266389363501775e-06,
+      "loss": 0.0002,
+      "step": 1838
+    },
+    {
+      "epoch": 2.352,
+      "grad_norm": 0.0006508962833322585,
+      "learning_rate": 8.250334587613677e-06,
+      "loss": 0.0002,
+      "step": 1839
+    },
+    {
+      "epoch": 2.35328,
+      "grad_norm": 0.0007670673076063395,
+      "learning_rate": 8.23430603150323e-06,
+      "loss": 0.0002,
+      "step": 1840
+    },
+    {
+      "epoch": 2.35456,
+      "grad_norm": 0.0002744818921200931,
+      "learning_rate": 8.21830372573592e-06,
+      "loss": 0.0002,
+      "step": 1841
+    },
+    {
+      "epoch": 2.35584,
+      "grad_norm": 0.0005801216466352344,
+      "learning_rate": 8.202327700827171e-06,
+      "loss": 0.0002,
+      "step": 1842
+    },
+    {
+      "epoch": 2.35712,
+      "grad_norm": 0.0003591123386286199,
+      "learning_rate": 8.18637798724229e-06,
+      "loss": 0.0002,
+      "step": 1843
+    },
+    {
+      "epoch": 2.3584,
+      "grad_norm": 0.00020390120334923267,
+      "learning_rate": 8.170454615396412e-06,
+      "loss": 0.0002,
+      "step": 1844
+    },
+    {
+      "epoch": 2.35968,
+      "grad_norm": 0.0007962146191857755,
+      "learning_rate": 8.154557615654455e-06,
+      "loss": 0.0002,
+      "step": 1845
+    },
+    {
+      "epoch": 2.36096,
+      "grad_norm": 0.0001696029503364116,
+      "learning_rate": 8.138687018331016e-06,
+      "loss": 0.0002,
+      "step": 1846
+    },
+    {
+      "epoch": 2.36224,
+      "grad_norm": 0.0004607769660651684,
+      "learning_rate": 8.122842853690368e-06,
+      "loss": 0.0002,
+      "step": 1847
+    },
+    {
+      "epoch": 2.36352,
+      "grad_norm": 0.0005677232402376831,
+      "learning_rate": 8.107025151946376e-06,
+      "loss": 0.0002,
+      "step": 1848
+    },
+    {
+      "epoch": 2.3648,
+      "grad_norm": 0.00037837939453311265,
+      "learning_rate": 8.091233943262426e-06,
+      "loss": 0.0002,
+      "step": 1849
+    },
+    {
+      "epoch": 2.36608,
+      "grad_norm": 0.0009469679207541049,
+      "learning_rate": 8.075469257751409e-06,
+      "loss": 0.0002,
+      "step": 1850
+    },
+    {
+      "epoch": 2.36736,
+      "grad_norm": 0.0005874616326764226,
+      "learning_rate": 8.059731125475605e-06,
+      "loss": 0.0002,
+      "step": 1851
+    },
+    {
+      "epoch": 2.36864,
+      "grad_norm": 0.0008352144504897296,
+      "learning_rate": 8.044019576446698e-06,
+      "loss": 0.0002,
+      "step": 1852
+    },
+    {
+      "epoch": 2.36992,
+      "grad_norm": 0.00018226521206088364,
+      "learning_rate": 8.02833464062565e-06,
+      "loss": 0.0002,
+      "step": 1853
+    },
+    {
+      "epoch": 2.3712,
+      "grad_norm": 0.00012786927982233465,
+      "learning_rate": 8.012676347922676e-06,
+      "loss": 0.0002,
+      "step": 1854
+    },
+    {
+      "epoch": 2.37248,
+      "grad_norm": 0.0005658522713929415,
+      "learning_rate": 7.997044728197196e-06,
+      "loss": 0.0002,
+      "step": 1855
+    },
+    {
+      "epoch": 2.37376,
+      "grad_norm": 0.00010807412036228925,
+      "learning_rate": 7.981439811257758e-06,
+      "loss": 0.0002,
+      "step": 1856
+    },
+    {
+      "epoch": 2.37504,
+      "grad_norm": 0.0005290827830322087,
+      "learning_rate": 7.965861626861999e-06,
+      "loss": 0.0002,
+      "step": 1857
+    },
+    {
+      "epoch": 2.3763199999999998,
+      "grad_norm": 0.00011951862688874826,
+      "learning_rate": 7.950310204716562e-06,
+      "loss": 0.0002,
+      "step": 1858
+    },
+    {
+      "epoch": 2.3776,
+      "grad_norm": 0.00024504747125320137,
+      "learning_rate": 7.934785574477069e-06,
+      "loss": 0.0002,
+      "step": 1859
+    },
+    {
+      "epoch": 2.37888,
+      "grad_norm": 0.00025921399355866015,
+      "learning_rate": 7.919287765748052e-06,
+      "loss": 0.0002,
+      "step": 1860
+    },
+    {
+      "epoch": 2.37888,
+      "eval_loss": 1.1839673519134521,
+      "eval_runtime": 43.5719,
+      "eval_samples_per_second": 11.521,
+      "eval_steps_per_second": 1.446,
+      "step": 1860
+    },
+    {
+      "epoch": 2.38016,
+      "grad_norm": 0.0006775157526135445,
+      "learning_rate": 7.903816808082883e-06,
+      "loss": 0.0002,
+      "step": 1861
+    },
+    {
+      "epoch": 2.38144,
+      "grad_norm": 0.0004763702745549381,
+      "learning_rate": 7.888372730983744e-06,
+      "loss": 0.0002,
+      "step": 1862
+    },
+    {
+      "epoch": 2.38272,
+      "grad_norm": 0.00022730005730409175,
+      "learning_rate": 7.872955563901554e-06,
+      "loss": 0.0002,
+      "step": 1863
+    },
+    {
+      "epoch": 2.384,
+      "grad_norm": 0.0005117187974974513,
+      "learning_rate": 7.85756533623592e-06,
+      "loss": 0.0002,
+      "step": 1864
+    },
+    {
+      "epoch": 2.38528,
+      "grad_norm": 0.00012259108189027756,
+      "learning_rate": 7.842202077335063e-06,
+      "loss": 0.0002,
+      "step": 1865
+    },
+    {
+      "epoch": 2.3865600000000002,
+      "grad_norm": 0.0010777805000543594,
+      "learning_rate": 7.826865816495784e-06,
+      "loss": 0.0002,
+      "step": 1866
+    },
+    {
+      "epoch": 2.38784,
+      "grad_norm": 0.0001915521133923903,
+      "learning_rate": 7.811556582963415e-06,
+      "loss": 0.0002,
+      "step": 1867
+    },
+    {
+      "epoch": 2.38912,
+      "grad_norm": 0.0004820209287572652,
+      "learning_rate": 7.796274405931725e-06,
+      "loss": 0.0002,
+      "step": 1868
+    },
+    {
+      "epoch": 2.3904,
+      "grad_norm": 0.0007654948276467621,
+      "learning_rate": 7.781019314542904e-06,
+      "loss": 0.0002,
+      "step": 1869
+    },
+    {
+      "epoch": 2.39168,
+      "grad_norm": 0.0003211789298802614,
+      "learning_rate": 7.765791337887477e-06,
+      "loss": 0.0002,
+      "step": 1870
+    },
+    {
+      "epoch": 2.39296,
+      "grad_norm": 0.00039891450433060527,
+      "learning_rate": 7.750590505004278e-06,
+      "loss": 0.0002,
+      "step": 1871
+    },
+    {
+      "epoch": 2.39424,
+      "grad_norm": 0.0004278056148905307,
+      "learning_rate": 7.735416844880377e-06,
+      "loss": 0.0002,
+      "step": 1872
+    },
+    {
+      "epoch": 2.39552,
+      "grad_norm": 0.00041603570571169257,
+      "learning_rate": 7.720270386451012e-06,
+      "loss": 0.0002,
+      "step": 1873
+    },
+    {
+      "epoch": 2.3968,
+      "grad_norm": 0.0001487014233134687,
+      "learning_rate": 7.705151158599568e-06,
+      "loss": 0.0002,
+      "step": 1874
+    },
+    {
+      "epoch": 2.39808,
+      "grad_norm": 0.0007464838563464582,
+      "learning_rate": 7.690059190157493e-06,
+      "loss": 0.0002,
+      "step": 1875
+    },
+    {
+      "epoch": 2.39936,
+      "grad_norm": 0.0002430266613373533,
+      "learning_rate": 7.67499450990426e-06,
+      "loss": 0.0002,
+      "step": 1876
+    },
+    {
+      "epoch": 2.40064,
+      "grad_norm": 0.00044445652747526765,
+      "learning_rate": 7.659957146567292e-06,
+      "loss": 0.0002,
+      "step": 1877
+    },
+    {
+      "epoch": 2.40192,
+      "grad_norm": 0.000578554579988122,
+      "learning_rate": 7.644947128821931e-06,
+      "loss": 0.0002,
+      "step": 1878
+    },
+    {
+      "epoch": 2.4032,
+      "grad_norm": 0.00038803924690000713,
+      "learning_rate": 7.629964485291377e-06,
+      "loss": 0.0002,
+      "step": 1879
+    },
+    {
+      "epoch": 2.40448,
+      "grad_norm": 0.0005333941080607474,
+      "learning_rate": 7.615009244546611e-06,
+      "loss": 0.0002,
+      "step": 1880
+    },
+    {
+      "epoch": 2.40576,
+      "grad_norm": 0.00010084795212605968,
+      "learning_rate": 7.600081435106375e-06,
+      "loss": 0.0002,
+      "step": 1881
+    },
+    {
+      "epoch": 2.40704,
+      "grad_norm": 0.00016836215218063444,
+      "learning_rate": 7.585181085437101e-06,
+      "loss": 0.0002,
+      "step": 1882
+    },
+    {
+      "epoch": 2.40832,
+      "grad_norm": 0.0006275451160036027,
+      "learning_rate": 7.570308223952843e-06,
+      "loss": 0.0002,
+      "step": 1883
+    },
+    {
+      "epoch": 2.4096,
+      "grad_norm": 0.000125435835798271,
+      "learning_rate": 7.55546287901525e-06,
+      "loss": 0.0002,
+      "step": 1884
+    },
+    {
+      "epoch": 2.41088,
+      "grad_norm": 0.00022631444153375924,
+      "learning_rate": 7.540645078933491e-06,
+      "loss": 0.0002,
+      "step": 1885
+    },
+    {
+      "epoch": 2.41216,
+      "grad_norm": 0.00045561956358142197,
+      "learning_rate": 7.525854851964214e-06,
+      "loss": 0.0002,
+      "step": 1886
+    },
+    {
+      "epoch": 2.41344,
+      "grad_norm": 0.00031555627356283367,
+      "learning_rate": 7.511092226311482e-06,
+      "loss": 0.0002,
+      "step": 1887
+    },
+    {
+      "epoch": 2.41472,
+      "grad_norm": 0.0005395089974626899,
+      "learning_rate": 7.4963572301267364e-06,
+      "loss": 0.0002,
+      "step": 1888
+    },
+    {
+      "epoch": 2.416,
+      "grad_norm": 0.0007261098944582045,
+      "learning_rate": 7.481649891508706e-06,
+      "loss": 0.0002,
+      "step": 1889
+    },
+    {
+      "epoch": 2.41728,
+      "grad_norm": 0.00010826917423401028,
+      "learning_rate": 7.466970238503399e-06,
+      "loss": 0.0002,
+      "step": 1890
+    },
+    {
+      "epoch": 2.41728,
+      "eval_loss": 1.1747924089431763,
+      "eval_runtime": 43.0535,
+      "eval_samples_per_second": 11.66,
+      "eval_steps_per_second": 1.463,
+      "step": 1890
+    },
+    {
+      "epoch": 2.41856,
+      "grad_norm": 0.0011698183370754123,
+      "learning_rate": 7.45231829910403e-06,
+      "loss": 0.0002,
+      "step": 1891
+    },
+    {
+      "epoch": 2.4198399999999998,
+      "grad_norm": 0.0003063059411942959,
+      "learning_rate": 7.437694101250949e-06,
+      "loss": 0.0002,
+      "step": 1892
+    },
+    {
+      "epoch": 2.42112,
+      "grad_norm": 0.00037533161230385303,
+      "learning_rate": 7.423097672831616e-06,
+      "loss": 0.0002,
+      "step": 1893
+    },
+    {
+      "epoch": 2.4224,
+      "grad_norm": 0.0013914547162130475,
+      "learning_rate": 7.4085290416805385e-06,
+      "loss": 0.0002,
+      "step": 1894
+    },
+    {
+      "epoch": 2.42368,
+      "grad_norm": 0.0003663707757368684,
+      "learning_rate": 7.393988235579208e-06,
+      "loss": 0.0002,
+      "step": 1895
+    },
+    {
+      "epoch": 2.42496,
+      "grad_norm": 0.0010497913463041186,
+      "learning_rate": 7.379475282256059e-06,
+      "loss": 0.0002,
+      "step": 1896
+    },
+    {
+      "epoch": 2.42624,
+      "grad_norm": 0.00026213075034320354,
+      "learning_rate": 7.364990209386413e-06,
+      "loss": 0.0002,
+      "step": 1897
+    },
+    {
+      "epoch": 2.42752,
+      "grad_norm": 0.000615556025877595,
+      "learning_rate": 7.350533044592434e-06,
+      "loss": 0.0002,
+      "step": 1898
+    },
+    {
+      "epoch": 2.4288,
+      "grad_norm": 0.0003180488711223006,
+      "learning_rate": 7.33610381544305e-06,
+      "loss": 0.0002,
+      "step": 1899
+    },
+    {
+      "epoch": 2.4300800000000002,
+      "grad_norm": 0.00014440002269111574,
+      "learning_rate": 7.321702549453922e-06,
+      "loss": 0.0002,
+      "step": 1900
+    },
+    {
+      "epoch": 2.43136,
+      "grad_norm": 0.0003721504472196102,
+      "learning_rate": 7.307329274087394e-06,
+      "loss": 0.0002,
+      "step": 1901
+    },
+    {
+      "epoch": 2.43264,
+      "grad_norm": 0.00012881176371593028,
+      "learning_rate": 7.2929840167524295e-06,
+      "loss": 0.0002,
+      "step": 1902
+    },
+    {
+      "epoch": 2.43392,
+      "grad_norm": 0.00012962603068444878,
+      "learning_rate": 7.278666804804571e-06,
+      "loss": 0.0002,
+      "step": 1903
+    },
+    {
+      "epoch": 2.4352,
+      "grad_norm": 0.0001382453046971932,
+      "learning_rate": 7.264377665545861e-06,
+      "loss": 0.0002,
+      "step": 1904
+    },
+    {
+      "epoch": 2.43648,
+      "grad_norm": 0.00035853637382388115,
+      "learning_rate": 7.2501166262248265e-06,
+      "loss": 0.0002,
+      "step": 1905
+    },
+    {
+      "epoch": 2.43776,
+      "grad_norm": 0.00015113272820599377,
+      "learning_rate": 7.235883714036411e-06,
+      "loss": 0.0002,
+      "step": 1906
+    },
+    {
+      "epoch": 2.43904,
+      "grad_norm": 0.0004522831295616925,
+      "learning_rate": 7.221678956121903e-06,
+      "loss": 0.0002,
+      "step": 1907
+    },
+    {
+      "epoch": 2.44032,
+      "grad_norm": 0.00030768956639803946,
+      "learning_rate": 7.207502379568916e-06,
+      "loss": 0.0002,
+      "step": 1908
+    },
+    {
+      "epoch": 2.4416,
+      "grad_norm": 0.0003462491149548441,
+      "learning_rate": 7.1933540114113266e-06,
+      "loss": 0.0002,
+      "step": 1909
+    },
+    {
+      "epoch": 2.44288,
+      "grad_norm": 0.0003502328472677618,
+      "learning_rate": 7.179233878629217e-06,
+      "loss": 0.0002,
+      "step": 1910
+    },
+    {
+      "epoch": 2.44416,
+      "grad_norm": 0.00020817614858970046,
+      "learning_rate": 7.165142008148818e-06,
+      "loss": 0.0002,
+      "step": 1911
+    },
+    {
+      "epoch": 2.44544,
+      "grad_norm": 0.000591093732509762,
+      "learning_rate": 7.1510784268424635e-06,
+      "loss": 0.0002,
+      "step": 1912
+    },
+    {
+      "epoch": 2.44672,
+      "grad_norm": 0.000337381410645321,
+      "learning_rate": 7.137043161528567e-06,
+      "loss": 0.0002,
+      "step": 1913
+    },
+    {
+      "epoch": 2.448,
+      "grad_norm": 0.00018345269199926406,
+      "learning_rate": 7.123036238971517e-06,
+      "loss": 0.0002,
+      "step": 1914
+    },
+    {
+      "epoch": 2.44928,
+      "grad_norm": 0.0006533708074130118,
+      "learning_rate": 7.109057685881668e-06,
+      "loss": 0.0002,
+      "step": 1915
+    },
+    {
+      "epoch": 2.45056,
+      "grad_norm": 0.00033750629518181086,
+      "learning_rate": 7.095107528915265e-06,
+      "loss": 0.0002,
+      "step": 1916
+    },
+    {
+      "epoch": 2.45184,
+      "grad_norm": 0.0001718109124340117,
+      "learning_rate": 7.081185794674425e-06,
+      "loss": 0.0002,
+      "step": 1917
+    },
+    {
+      "epoch": 2.45312,
+      "grad_norm": 0.0006910200463607907,
+      "learning_rate": 7.0672925097070446e-06,
+      "loss": 0.0002,
+      "step": 1918
+    },
+    {
+      "epoch": 2.4544,
+      "grad_norm": 0.0003057036083191633,
+      "learning_rate": 7.053427700506767e-06,
+      "loss": 0.0002,
+      "step": 1919
+    },
+    {
+      "epoch": 2.45568,
+      "grad_norm": 0.0002737011236604303,
+      "learning_rate": 7.03959139351295e-06,
+      "loss": 0.0002,
+      "step": 1920
+    },
+    {
+      "epoch": 2.45568,
+      "eval_loss": 1.1830275058746338,
+      "eval_runtime": 43.4832,
+      "eval_samples_per_second": 11.545,
+      "eval_steps_per_second": 1.449,
+      "step": 1920
+    },
+    {
+      "epoch": 2.45696,
+      "grad_norm": 0.0003738821833394468,
+      "learning_rate": 7.02578361511059e-06,
+      "loss": 0.0002,
+      "step": 1921
+    },
+    {
+      "epoch": 2.45824,
+      "grad_norm": 0.00014257086149882525,
+      "learning_rate": 7.0120043916302885e-06,
+      "loss": 0.0002,
+      "step": 1922
+    },
+    {
+      "epoch": 2.45952,
+      "grad_norm": 0.00015353388153016567,
+      "learning_rate": 6.998253749348178e-06,
+      "loss": 0.0002,
+      "step": 1923
+    },
+    {
+      "epoch": 2.4608,
+      "grad_norm": 0.0002355447504669428,
+      "learning_rate": 6.984531714485905e-06,
+      "loss": 0.0002,
+      "step": 1924
+    },
+    {
+      "epoch": 2.46208,
+      "grad_norm": 0.00034142841468565166,
+      "learning_rate": 6.970838313210557e-06,
+      "loss": 0.0002,
+      "step": 1925
+    },
+    {
+      "epoch": 2.4633599999999998,
+      "grad_norm": 0.0005062466952949762,
+      "learning_rate": 6.957173571634617e-06,
+      "loss": 0.0002,
+      "step": 1926
+    },
+    {
+      "epoch": 2.46464,
+      "grad_norm": 0.0003805873275268823,
+      "learning_rate": 6.943537515815919e-06,
+      "loss": 0.0002,
+      "step": 1927
+    },
+    {
+      "epoch": 2.46592,
+      "grad_norm": 0.00024458515690639615,
+      "learning_rate": 6.929930171757595e-06,
+      "loss": 0.0002,
+      "step": 1928
+    },
+    {
+      "epoch": 2.4672,
+      "grad_norm": 0.0004816846631001681,
+      "learning_rate": 6.916351565408028e-06,
+      "loss": 0.0002,
+      "step": 1929
+    },
+    {
+      "epoch": 2.46848,
+      "grad_norm": 0.00022859709861222655,
+      "learning_rate": 6.902801722660793e-06,
+      "loss": 0.0002,
+      "step": 1930
+    },
+    {
+      "epoch": 2.46976,
+      "grad_norm": 0.0002492612402420491,
+      "learning_rate": 6.88928066935461e-06,
+      "loss": 0.0002,
+      "step": 1931
+    },
+    {
+      "epoch": 2.47104,
+      "grad_norm": 0.00032544921850785613,
+      "learning_rate": 6.875788431273322e-06,
+      "loss": 0.0002,
+      "step": 1932
+    },
+    {
+      "epoch": 2.47232,
+      "grad_norm": 0.0005566601175814867,
+      "learning_rate": 6.862325034145796e-06,
+      "loss": 0.0002,
+      "step": 1933
+    },
+    {
+      "epoch": 2.4736000000000002,
+      "grad_norm": 0.00034675264032557607,
+      "learning_rate": 6.848890503645924e-06,
+      "loss": 0.0002,
+      "step": 1934
+    },
+    {
+      "epoch": 2.47488,
+      "grad_norm": 0.0004174388595856726,
+      "learning_rate": 6.83548486539253e-06,
+      "loss": 0.0002,
+      "step": 1935
+    },
+    {
+      "epoch": 2.47616,
+      "grad_norm": 0.00031951870187185705,
+      "learning_rate": 6.822108144949359e-06,
+      "loss": 0.0002,
+      "step": 1936
+    },
+    {
+      "epoch": 2.47744,
+      "grad_norm": 0.0002985461615025997,
+      "learning_rate": 6.808760367825009e-06,
+      "loss": 0.0002,
+      "step": 1937
+    },
+    {
+      "epoch": 2.47872,
+      "grad_norm": 0.00015834596706554294,
+      "learning_rate": 6.79544155947287e-06,
+      "loss": 0.0002,
+      "step": 1938
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.0006239261128939688,
+      "learning_rate": 6.782151745291108e-06,
+      "loss": 0.0002,
+      "step": 1939
+    },
+    {
+      "epoch": 2.48128,
+      "grad_norm": 0.00012638726911973208,
+      "learning_rate": 6.768890950622596e-06,
+      "loss": 0.0002,
+      "step": 1940
+    },
+    {
+      "epoch": 2.48256,
+      "grad_norm": 0.0003220559738110751,
+      "learning_rate": 6.755659200754856e-06,
+      "loss": 0.0002,
+      "step": 1941
+    },
+    {
+      "epoch": 2.48384,
+      "grad_norm": 0.00022451272525358945,
+      "learning_rate": 6.742456520920035e-06,
+      "loss": 0.0002,
+      "step": 1942
+    },
+    {
+      "epoch": 2.48512,
+      "grad_norm": 0.00015170875121839345,
+      "learning_rate": 6.729282936294846e-06,
+      "loss": 0.0002,
+      "step": 1943
+    },
+    {
+      "epoch": 2.4864,
+      "grad_norm": 0.0003331229672767222,
+      "learning_rate": 6.716138472000517e-06,
+      "loss": 0.0002,
+      "step": 1944
+    },
+    {
+      "epoch": 2.48768,
+      "grad_norm": 0.0001102744645322673,
+      "learning_rate": 6.703023153102738e-06,
+      "loss": 0.0002,
+      "step": 1945
+    },
+    {
+      "epoch": 2.48896,
+      "grad_norm": 0.00012537582369986922,
+      "learning_rate": 6.6899370046116265e-06,
+      "loss": 0.0002,
+      "step": 1946
+    },
+    {
+      "epoch": 2.49024,
+      "grad_norm": 0.0002181735326303169,
+      "learning_rate": 6.67688005148168e-06,
+      "loss": 0.0002,
+      "step": 1947
+    },
+    {
+      "epoch": 2.49152,
+      "grad_norm": 0.0004947243141941726,
+      "learning_rate": 6.6638523186117084e-06,
+      "loss": 0.0002,
+      "step": 1948
+    },
+    {
+      "epoch": 2.4928,
+      "grad_norm": 0.00017499204841442406,
+      "learning_rate": 6.650853830844817e-06,
+      "loss": 0.0002,
+      "step": 1949
+    },
+    {
+      "epoch": 2.49408,
+      "grad_norm": 0.0003295360365882516,
+      "learning_rate": 6.637884612968315e-06,
+      "loss": 0.0002,
+      "step": 1950
+    },
+    {
+      "epoch": 2.49408,
+      "eval_loss": 1.1815475225448608,
+      "eval_runtime": 43.3933,
+      "eval_samples_per_second": 11.569,
+      "eval_steps_per_second": 1.452,
+      "step": 1950
+    },
+    {
+      "epoch": 2.49536,
+      "grad_norm": 0.0003929525555577129,
+      "learning_rate": 6.624944689713739e-06,
+      "loss": 0.0002,
+      "step": 1951
+    },
+    {
+      "epoch": 2.49664,
+      "grad_norm": 0.00013083942758385092,
+      "learning_rate": 6.6120340857567226e-06,
+      "loss": 0.0002,
+      "step": 1952
+    },
+    {
+      "epoch": 2.49792,
+      "grad_norm": 0.0006896118284203112,
+      "learning_rate": 6.599152825717001e-06,
+      "loss": 0.0002,
+      "step": 1953
+    },
+    {
+      "epoch": 2.4992,
+      "grad_norm": 0.0003455237310845405,
+      "learning_rate": 6.586300934158363e-06,
+      "loss": 0.0002,
+      "step": 1954
+    },
+    {
+      "epoch": 2.50048,
+      "grad_norm": 0.0005406014970503747,
+      "learning_rate": 6.573478435588581e-06,
+      "loss": 0.0002,
+      "step": 1955
+    },
+    {
+      "epoch": 2.50176,
+      "grad_norm": 0.00023619551211595535,
+      "learning_rate": 6.5606853544593914e-06,
+      "loss": 0.0002,
+      "step": 1956
+    },
+    {
+      "epoch": 2.50304,
+      "grad_norm": 0.0004901570500805974,
+      "learning_rate": 6.547921715166411e-06,
+      "loss": 0.0002,
+      "step": 1957
+    },
+    {
+      "epoch": 2.50432,
+      "grad_norm": 0.00059032830176875,
+      "learning_rate": 6.535187542049132e-06,
+      "loss": 0.0002,
+      "step": 1958
+    },
+    {
+      "epoch": 2.5056000000000003,
+      "grad_norm": 0.00040494927088730037,
+      "learning_rate": 6.522482859390853e-06,
+      "loss": 0.0002,
+      "step": 1959
+    },
+    {
+      "epoch": 2.5068799999999998,
+      "grad_norm": 0.00031261437106877565,
+      "learning_rate": 6.509807691418626e-06,
+      "loss": 0.0002,
+      "step": 1960
+    },
+    {
+      "epoch": 2.50816,
+      "grad_norm": 0.00013683347788173705,
+      "learning_rate": 6.497162062303229e-06,
+      "loss": 0.0002,
+      "step": 1961
+    },
+    {
+      "epoch": 2.50944,
+      "grad_norm": 0.0001638557150727138,
+      "learning_rate": 6.484545996159114e-06,
+      "loss": 0.0002,
+      "step": 1962
+    },
+    {
+      "epoch": 2.51072,
+      "grad_norm": 0.0002301942149642855,
+      "learning_rate": 6.47195951704435e-06,
+      "loss": 0.0002,
+      "step": 1963
+    },
+    {
+      "epoch": 2.512,
+      "grad_norm": 0.0001310958032263443,
+      "learning_rate": 6.459402648960594e-06,
+      "loss": 0.0002,
+      "step": 1964
+    },
+    {
+      "epoch": 2.51328,
+      "grad_norm": 0.0005266906809993088,
+      "learning_rate": 6.446875415853016e-06,
+      "loss": 0.0002,
+      "step": 1965
+    },
+    {
+      "epoch": 2.51456,
+      "grad_norm": 0.0002377618511673063,
+      "learning_rate": 6.434377841610306e-06,
+      "loss": 0.0002,
+      "step": 1966
+    },
+    {
+      "epoch": 2.51584,
+      "grad_norm": 0.0003869717475026846,
+      "learning_rate": 6.421909950064575e-06,
+      "loss": 0.0002,
+      "step": 1967
+    },
+    {
+      "epoch": 2.5171200000000002,
+      "grad_norm": 0.0007507234113290906,
+      "learning_rate": 6.409471764991335e-06,
+      "loss": 0.0002,
+      "step": 1968
+    },
+    {
+      "epoch": 2.5183999999999997,
+      "grad_norm": 0.00019987816631328315,
+      "learning_rate": 6.397063310109448e-06,
+      "loss": 0.0002,
+      "step": 1969
+    },
+    {
+      "epoch": 2.51968,
+      "grad_norm": 0.0004650322371162474,
+      "learning_rate": 6.384684609081084e-06,
+      "loss": 0.0002,
+      "step": 1970
+    },
+    {
+      "epoch": 2.52096,
+      "grad_norm": 0.0006089420639909804,
+      "learning_rate": 6.372335685511685e-06,
+      "loss": 0.0002,
+      "step": 1971
+    },
+    {
+      "epoch": 2.52224,
+      "grad_norm": 0.0001694520324235782,
+      "learning_rate": 6.360016562949884e-06,
+      "loss": 0.0002,
+      "step": 1972
+    },
+    {
+      "epoch": 2.52352,
+      "grad_norm": 0.0004142711404711008,
+      "learning_rate": 6.3477272648875084e-06,
+      "loss": 0.0002,
+      "step": 1973
+    },
+    {
+      "epoch": 2.5248,
+      "grad_norm": 0.000749512983020395,
+      "learning_rate": 6.335467814759505e-06,
+      "loss": 0.0002,
+      "step": 1974
+    },
+    {
+      "epoch": 2.52608,
+      "grad_norm": 0.0002557408297434449,
+      "learning_rate": 6.3232382359439e-06,
+      "loss": 0.0002,
+      "step": 1975
+    },
+    {
+      "epoch": 2.52736,
+      "grad_norm": 0.00038810624391771853,
+      "learning_rate": 6.311038551761755e-06,
+      "loss": 0.0002,
+      "step": 1976
+    },
+    {
+      "epoch": 2.52864,
+      "grad_norm": 0.0002354103489778936,
+      "learning_rate": 6.298868785477132e-06,
+      "loss": 0.0002,
+      "step": 1977
+    },
+    {
+      "epoch": 2.5299199999999997,
+      "grad_norm": 0.00012855166278313845,
+      "learning_rate": 6.286728960297036e-06,
+      "loss": 0.0002,
+      "step": 1978
+    },
+    {
+      "epoch": 2.5312,
+      "grad_norm": 0.00019848457304760814,
+      "learning_rate": 6.274619099371373e-06,
+      "loss": 0.0002,
+      "step": 1979
+    },
+    {
+      "epoch": 2.53248,
+      "grad_norm": 0.00019530925783328712,
+      "learning_rate": 6.262539225792914e-06,
+      "loss": 0.0002,
+      "step": 1980
+    },
+    {
+      "epoch": 2.53248,
+      "eval_loss": 1.1816483736038208,
+      "eval_runtime": 43.6059,
+      "eval_samples_per_second": 11.512,
+      "eval_steps_per_second": 1.445,
+      "step": 1980
+    },
+    {
+      "epoch": 2.53376,
+      "grad_norm": 0.00015873344091232866,
+      "learning_rate": 6.250489362597252e-06,
+      "loss": 0.0002,
+      "step": 1981
+    },
+    {
+      "epoch": 2.53504,
+      "grad_norm": 0.00015742145478725433,
+      "learning_rate": 6.238469532762737e-06,
+      "loss": 0.0002,
+      "step": 1982
+    },
+    {
+      "epoch": 2.53632,
+      "grad_norm": 0.00015187107783276588,
+      "learning_rate": 6.22647975921046e-06,
+      "loss": 0.0002,
+      "step": 1983
+    },
+    {
+      "epoch": 2.5376,
+      "grad_norm": 0.00018921127775684,
+      "learning_rate": 6.214520064804183e-06,
+      "loss": 0.0002,
+      "step": 1984
+    },
+    {
+      "epoch": 2.53888,
+      "grad_norm": 8.872054604580626e-05,
+      "learning_rate": 6.20259047235033e-06,
+      "loss": 0.0002,
+      "step": 1985
+    },
+    {
+      "epoch": 2.54016,
+      "grad_norm": 0.0001130819073296152,
+      "learning_rate": 6.190691004597901e-06,
+      "loss": 0.0002,
+      "step": 1986
+    },
+    {
+      "epoch": 2.54144,
+      "grad_norm": 0.0002832153986673802,
+      "learning_rate": 6.178821684238467e-06,
+      "loss": 0.0002,
+      "step": 1987
+    },
+    {
+      "epoch": 2.54272,
+      "grad_norm": 0.00026950897881761193,
+      "learning_rate": 6.166982533906092e-06,
+      "loss": 0.0002,
+      "step": 1988
+    },
+    {
+      "epoch": 2.544,
+      "grad_norm": 0.0001317868591286242,
+      "learning_rate": 6.155173576177321e-06,
+      "loss": 0.0002,
+      "step": 1989
+    },
+    {
+      "epoch": 2.54528,
+      "grad_norm": 0.000146243633935228,
+      "learning_rate": 6.143394833571127e-06,
+      "loss": 0.0002,
+      "step": 1990
+    },
+    {
+      "epoch": 2.54656,
+      "grad_norm": 0.00030357291689142585,
+      "learning_rate": 6.131646328548846e-06,
+      "loss": 0.0002,
+      "step": 1991
+    },
+    {
+      "epoch": 2.54784,
+      "grad_norm": 0.00015373178757727146,
+      "learning_rate": 6.11992808351417e-06,
+      "loss": 0.0002,
+      "step": 1992
+    },
+    {
+      "epoch": 2.5491200000000003,
+      "grad_norm": 0.0001959829533006996,
+      "learning_rate": 6.108240120813085e-06,
+      "loss": 0.0002,
+      "step": 1993
+    },
+    {
+      "epoch": 2.5504,
+      "grad_norm": 0.00013423572818282992,
+      "learning_rate": 6.096582462733818e-06,
+      "loss": 0.0002,
+      "step": 1994
+    },
+    {
+      "epoch": 2.55168,
+      "grad_norm": 0.0005714610451832414,
+      "learning_rate": 6.08495513150682e-06,
+      "loss": 0.0002,
+      "step": 1995
+    },
+    {
+      "epoch": 2.55296,
+      "grad_norm": 0.00018218479817733169,
+      "learning_rate": 6.073358149304708e-06,
+      "loss": 0.0002,
+      "step": 1996
+    },
+    {
+      "epoch": 2.55424,
+      "grad_norm": 0.0009697470813989639,
+      "learning_rate": 6.061791538242224e-06,
+      "loss": 0.0002,
+      "step": 1997
+    },
+    {
+      "epoch": 2.55552,
+      "grad_norm": 0.00021223712246865034,
+      "learning_rate": 6.05025532037619e-06,
+      "loss": 0.0002,
+      "step": 1998
+    },
+    {
+      "epoch": 2.5568,
+      "grad_norm": 0.0006042190361768007,
+      "learning_rate": 6.038749517705469e-06,
+      "loss": 0.0002,
+      "step": 1999
+    },
+    {
+      "epoch": 2.55808,
+      "grad_norm": 0.0008896634681150317,
+      "learning_rate": 6.027274152170941e-06,
+      "loss": 0.0002,
+      "step": 2000
+    },
+    {
+      "epoch": 2.55936,
+      "grad_norm": 0.00023286855139303952,
+      "learning_rate": 6.01582924565542e-06,
+      "loss": 0.0002,
+      "step": 2001
+    },
+    {
+      "epoch": 2.5606400000000002,
+      "grad_norm": 0.00044719918514601886,
+      "learning_rate": 6.0044148199836595e-06,
+      "loss": 0.0002,
+      "step": 2002
+    },
+    {
+      "epoch": 2.5619199999999998,
+      "grad_norm": 0.0006025312468409538,
+      "learning_rate": 5.993030896922266e-06,
+      "loss": 0.0002,
+      "step": 2003
+    },
+    {
+      "epoch": 2.5632,
+      "grad_norm": 0.00023673917166888714,
+      "learning_rate": 5.981677498179692e-06,
+      "loss": 0.0002,
+      "step": 2004
+    },
+    {
+      "epoch": 2.56448,
+      "grad_norm": 0.00039530108915641904,
+      "learning_rate": 5.970354645406189e-06,
+      "loss": 0.0002,
+      "step": 2005
+    },
+    {
+      "epoch": 2.56576,
+      "grad_norm": 0.0005158588173799217,
+      "learning_rate": 5.959062360193738e-06,
+      "loss": 0.0002,
+      "step": 2006
+    },
+    {
+      "epoch": 2.56704,
+      "grad_norm": 0.00020971003687009215,
+      "learning_rate": 5.947800664076047e-06,
+      "loss": 0.0002,
+      "step": 2007
+    },
+    {
+      "epoch": 2.56832,
+      "grad_norm": 0.0007004168583080173,
+      "learning_rate": 5.9365695785284906e-06,
+      "loss": 0.0002,
+      "step": 2008
+    },
+    {
+      "epoch": 2.5696,
+      "grad_norm": 0.0006223475793376565,
+      "learning_rate": 5.925369124968066e-06,
+      "loss": 0.0002,
+      "step": 2009
+    },
+    {
+      "epoch": 2.57088,
+      "grad_norm": 0.00013660028344020247,
+      "learning_rate": 5.9141993247533525e-06,
+      "loss": 0.0002,
+      "step": 2010
+    },
+    {
+      "epoch": 2.57088,
+      "eval_loss": 1.1760672330856323,
+      "eval_runtime": 43.1068,
+      "eval_samples_per_second": 11.645,
+      "eval_steps_per_second": 1.461,
+      "step": 2010
+    },
+    {
+      "epoch": 2.5721600000000002,
+      "grad_norm": 0.0007261909195221961,
+      "learning_rate": 5.903060199184487e-06,
+      "loss": 0.0002,
+      "step": 2011
+    },
+    {
+      "epoch": 2.5734399999999997,
+      "grad_norm": 0.000268265837803483,
+      "learning_rate": 5.891951769503107e-06,
+      "loss": 0.0002,
+      "step": 2012
+    },
+    {
+      "epoch": 2.57472,
+      "grad_norm": 0.0002705653023440391,
+      "learning_rate": 5.880874056892308e-06,
+      "loss": 0.0002,
+      "step": 2013
+    },
+    {
+      "epoch": 2.576,
+      "grad_norm": 0.00012832919310312718,
+      "learning_rate": 5.869827082476625e-06,
+      "loss": 0.0002,
+      "step": 2014
+    },
+    {
+      "epoch": 2.57728,
+      "grad_norm": 0.0006553751300089061,
+      "learning_rate": 5.858810867321955e-06,
+      "loss": 0.0002,
+      "step": 2015
+    },
+    {
+      "epoch": 2.57856,
+      "grad_norm": 0.00020814413437619805,
+      "learning_rate": 5.847825432435566e-06,
+      "loss": 0.0002,
+      "step": 2016
+    },
+    {
+      "epoch": 2.57984,
+      "grad_norm": 0.0006322242552414536,
+      "learning_rate": 5.836870798766008e-06,
+      "loss": 0.0002,
+      "step": 2017
+    },
+    {
+      "epoch": 2.58112,
+      "grad_norm": 0.0002930402697529644,
+      "learning_rate": 5.8259469872031006e-06,
+      "loss": 0.0002,
+      "step": 2018
+    },
+    {
+      "epoch": 2.5824,
+      "grad_norm": 0.0005195149569772184,
+      "learning_rate": 5.815054018577894e-06,
+      "loss": 0.0002,
+      "step": 2019
+    },
+    {
+      "epoch": 2.58368,
+      "grad_norm": 0.0004457169270608574,
+      "learning_rate": 5.8041919136626144e-06,
+      "loss": 0.0002,
+      "step": 2020
+    },
+    {
+      "epoch": 2.58496,
+      "grad_norm": 9.912128007272258e-05,
+      "learning_rate": 5.793360693170646e-06,
+      "loss": 0.0002,
+      "step": 2021
+    },
+    {
+      "epoch": 2.58624,
+      "grad_norm": 0.0005960392882116139,
+      "learning_rate": 5.782560377756459e-06,
+      "loss": 0.0002,
+      "step": 2022
+    },
+    {
+      "epoch": 2.58752,
+      "grad_norm": 0.0007441943162120879,
+      "learning_rate": 5.771790988015601e-06,
+      "loss": 0.0002,
+      "step": 2023
+    },
+    {
+      "epoch": 2.5888,
+      "grad_norm": 0.0006059888401068747,
+      "learning_rate": 5.761052544484652e-06,
+      "loss": 0.0002,
+      "step": 2024
+    },
+    {
+      "epoch": 2.59008,
+      "grad_norm": 0.000811684294603765,
+      "learning_rate": 5.750345067641161e-06,
+      "loss": 0.0002,
+      "step": 2025
+    },
+    {
+      "epoch": 2.59136,
+      "grad_norm": 0.00020101373957004398,
+      "learning_rate": 5.739668577903646e-06,
+      "loss": 0.0002,
+      "step": 2026
+    },
+    {
+      "epoch": 2.59264,
+      "grad_norm": 0.0002765175304375589,
+      "learning_rate": 5.729023095631524e-06,
+      "loss": 0.0002,
+      "step": 2027
+    },
+    {
+      "epoch": 2.59392,
+      "grad_norm": 0.0007142999093048275,
+      "learning_rate": 5.71840864112508e-06,
+      "loss": 0.0002,
+      "step": 2028
+    },
+    {
+      "epoch": 2.5952,
+      "grad_norm": 0.0005172202945686877,
+      "learning_rate": 5.7078252346254395e-06,
+      "loss": 0.0002,
+      "step": 2029
+    },
+    {
+      "epoch": 2.59648,
+      "grad_norm": 0.0005064026918262243,
+      "learning_rate": 5.6972728963145055e-06,
+      "loss": 0.0002,
+      "step": 2030
+    },
+    {
+      "epoch": 2.59776,
+      "grad_norm": 0.0007822649786248803,
+      "learning_rate": 5.686751646314961e-06,
+      "loss": 0.0002,
+      "step": 2031
+    },
+    {
+      "epoch": 2.59904,
+      "grad_norm": 0.00016675966617185622,
+      "learning_rate": 5.676261504690184e-06,
+      "loss": 0.0002,
+      "step": 2032
+    },
+    {
+      "epoch": 2.60032,
+      "grad_norm": 0.0004352393443696201,
+      "learning_rate": 5.665802491444239e-06,
+      "loss": 0.0002,
+      "step": 2033
+    },
+    {
+      "epoch": 2.6016,
+      "grad_norm": 0.00018516472482588142,
+      "learning_rate": 5.65537462652183e-06,
+      "loss": 0.0002,
+      "step": 2034
+    },
+    {
+      "epoch": 2.60288,
+      "grad_norm": 0.00020108661556150764,
+      "learning_rate": 5.64497792980826e-06,
+      "loss": 0.0002,
+      "step": 2035
+    },
+    {
+      "epoch": 2.6041600000000003,
+      "grad_norm": 0.0005044660647399724,
+      "learning_rate": 5.6346124211294074e-06,
+      "loss": 0.0002,
+      "step": 2036
+    },
+    {
+      "epoch": 2.6054399999999998,
+      "grad_norm": 0.0001886051904875785,
+      "learning_rate": 5.624278120251657e-06,
+      "loss": 0.0002,
+      "step": 2037
+    },
+    {
+      "epoch": 2.60672,
+      "grad_norm": 0.0007088842103257775,
+      "learning_rate": 5.613975046881899e-06,
+      "loss": 0.0002,
+      "step": 2038
+    },
+    {
+      "epoch": 2.608,
+      "grad_norm": 0.0003815259551629424,
+      "learning_rate": 5.603703220667473e-06,
+      "loss": 0.0002,
+      "step": 2039
+    },
+    {
+      "epoch": 2.60928,
+      "grad_norm": 0.0004345429188106209,
+      "learning_rate": 5.593462661196119e-06,
+      "loss": 0.0002,
+      "step": 2040
+    },
+    {
+      "epoch": 2.60928,
+      "eval_loss": 1.1763349771499634,
+      "eval_runtime": 43.2109,
+      "eval_samples_per_second": 11.617,
+      "eval_steps_per_second": 1.458,
+      "step": 2040
+    },
+    {
+      "epoch": 2.61056,
+      "grad_norm": 0.0006033392855897546,
+      "learning_rate": 5.583253387995969e-06,
+      "loss": 0.0002,
+      "step": 2041
+    },
+    {
+      "epoch": 2.61184,
+      "grad_norm": 0.00014597387053072453,
+      "learning_rate": 5.573075420535488e-06,
+      "loss": 0.0002,
+      "step": 2042
+    },
+    {
+      "epoch": 2.61312,
+      "grad_norm": 0.00015920022269710898,
+      "learning_rate": 5.562928778223439e-06,
+      "loss": 0.0002,
+      "step": 2043
+    },
+    {
+      "epoch": 2.6144,
+      "grad_norm": 0.00037969640106894076,
+      "learning_rate": 5.552813480408856e-06,
+      "loss": 0.0002,
+      "step": 2044
+    },
+    {
+      "epoch": 2.6156800000000002,
+      "grad_norm": 0.0005262352642603219,
+      "learning_rate": 5.542729546380997e-06,
+      "loss": 0.0002,
+      "step": 2045
+    },
+    {
+      "epoch": 2.6169599999999997,
+      "grad_norm": 0.0007127207354642451,
+      "learning_rate": 5.532676995369314e-06,
+      "loss": 0.0002,
+      "step": 2046
+    },
+    {
+      "epoch": 2.61824,
+      "grad_norm": 0.00018934982654172927,
+      "learning_rate": 5.522655846543408e-06,
+      "loss": 0.0002,
+      "step": 2047
+    },
+    {
+      "epoch": 2.61952,
+      "grad_norm": 0.0003394143714103848,
+      "learning_rate": 5.512666119013005e-06,
+      "loss": 0.0002,
+      "step": 2048
+    },
+    {
+      "epoch": 2.6208,
+      "grad_norm": 0.00015175512817222625,
+      "learning_rate": 5.5027078318279035e-06,
+      "loss": 0.0002,
+      "step": 2049
+    },
+    {
+      "epoch": 2.62208,
+      "grad_norm": 0.00025432585971429944,
+      "learning_rate": 5.492781003977965e-06,
+      "loss": 0.0002,
+      "step": 2050
+    },
+    {
+      "epoch": 2.62336,
+      "grad_norm": 0.0002585637557785958,
+      "learning_rate": 5.482885654393039e-06,
+      "loss": 0.0002,
+      "step": 2051
+    },
+    {
+      "epoch": 2.62464,
+      "grad_norm": 0.000490740523673594,
+      "learning_rate": 5.473021801942955e-06,
+      "loss": 0.0002,
+      "step": 2052
+    },
+    {
+      "epoch": 2.62592,
+      "grad_norm": 0.0001708458730718121,
+      "learning_rate": 5.463189465437484e-06,
+      "loss": 0.0002,
+      "step": 2053
+    },
+    {
+      "epoch": 2.6272,
+      "grad_norm": 0.0001390179677400738,
+      "learning_rate": 5.4533886636262955e-06,
+      "loss": 0.0002,
+      "step": 2054
+    },
+    {
+      "epoch": 2.62848,
+      "grad_norm": 0.0006850110366940498,
+      "learning_rate": 5.443619415198929e-06,
+      "loss": 0.0002,
+      "step": 2055
+    },
+    {
+      "epoch": 2.62976,
+      "grad_norm": 9.433117520529777e-05,
+      "learning_rate": 5.433881738784739e-06,
+      "loss": 0.0002,
+      "step": 2056
+    },
+    {
+      "epoch": 2.63104,
+      "grad_norm": 0.0007274181698448956,
+      "learning_rate": 5.424175652952887e-06,
+      "loss": 0.0002,
+      "step": 2057
+    },
+    {
+      "epoch": 2.63232,
+      "grad_norm": 0.00011057154915761203,
+      "learning_rate": 5.414501176212295e-06,
+      "loss": 0.0002,
+      "step": 2058
+    },
+    {
+      "epoch": 2.6336,
+      "grad_norm": 0.00011248882947256789,
+      "learning_rate": 5.4048583270115966e-06,
+      "loss": 0.0002,
+      "step": 2059
+    },
+    {
+      "epoch": 2.63488,
+      "grad_norm": 0.0005161191220395267,
+      "learning_rate": 5.395247123739119e-06,
+      "loss": 0.0002,
+      "step": 2060
+    },
+    {
+      "epoch": 2.63616,
+      "grad_norm": 0.0005980413407087326,
+      "learning_rate": 5.385667584722849e-06,
+      "loss": 0.0002,
+      "step": 2061
+    },
+    {
+      "epoch": 2.63744,
+      "grad_norm": 0.00018621369963511825,
+      "learning_rate": 5.376119728230387e-06,
+      "loss": 0.0002,
+      "step": 2062
+    },
+    {
+      "epoch": 2.63872,
+      "grad_norm": 0.0007834093994461,
+      "learning_rate": 5.3666035724689135e-06,
+      "loss": 0.0002,
+      "step": 2063
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.0007524779066443443,
+      "learning_rate": 5.357119135585153e-06,
+      "loss": 0.0002,
+      "step": 2064
+    },
+    {
+      "epoch": 2.64128,
+      "grad_norm": 0.00033790417364798486,
+      "learning_rate": 5.347666435665363e-06,
+      "loss": 0.0002,
+      "step": 2065
+    },
+    {
+      "epoch": 2.64256,
+      "grad_norm": 0.0007513008895330131,
+      "learning_rate": 5.338245490735263e-06,
+      "loss": 0.0002,
+      "step": 2066
+    },
+    {
+      "epoch": 2.64384,
+      "grad_norm": 0.0007832477567717433,
+      "learning_rate": 5.328856318760026e-06,
+      "loss": 0.0002,
+      "step": 2067
+    },
+    {
+      "epoch": 2.64512,
+      "grad_norm": 0.00011210219963686541,
+      "learning_rate": 5.319498937644228e-06,
+      "loss": 0.0002,
+      "step": 2068
+    },
+    {
+      "epoch": 2.6464,
+      "grad_norm": 0.0006179630290716887,
+      "learning_rate": 5.31017336523183e-06,
+      "loss": 0.0002,
+      "step": 2069
+    },
+    {
+      "epoch": 2.6476800000000003,
+      "grad_norm": 0.000690981512889266,
+      "learning_rate": 5.300879619306135e-06,
+      "loss": 0.0002,
+      "step": 2070
+    },
+    {
+      "epoch": 2.6476800000000003,
+      "eval_loss": 1.1787551641464233,
+      "eval_runtime": 43.2975,
+      "eval_samples_per_second": 11.594,
+      "eval_steps_per_second": 1.455,
+      "step": 2070
+    },
+    {
+      "epoch": 2.6489599999999998,
+      "grad_norm": 8.601046283729374e-05,
+      "learning_rate": 5.29161771758975e-06,
+      "loss": 0.0002,
+      "step": 2071
+    },
+    {
+      "epoch": 2.65024,
+      "grad_norm": 0.0004467715334612876,
+      "learning_rate": 5.282387677744559e-06,
+      "loss": 0.0002,
+      "step": 2072
+    },
+    {
+      "epoch": 2.65152,
+      "grad_norm": 0.00033754404284991324,
+      "learning_rate": 5.273189517371689e-06,
+      "loss": 0.0002,
+      "step": 2073
+    },
+    {
+      "epoch": 2.6528,
+      "grad_norm": 9.469128417549655e-05,
+      "learning_rate": 5.264023254011476e-06,
+      "loss": 0.0002,
+      "step": 2074
+    },
+    {
+      "epoch": 2.65408,
+      "grad_norm": 0.00046040205052122474,
+      "learning_rate": 5.254888905143425e-06,
+      "loss": 0.0002,
+      "step": 2075
+    },
+    {
+      "epoch": 2.65536,
+      "grad_norm": 0.00017660969751887023,
+      "learning_rate": 5.245786488186183e-06,
+      "loss": 0.0002,
+      "step": 2076
+    },
+    {
+      "epoch": 2.65664,
+      "grad_norm": 0.0002686119987629354,
+      "learning_rate": 5.236716020497515e-06,
+      "loss": 0.0002,
+      "step": 2077
+    },
+    {
+      "epoch": 2.65792,
+      "grad_norm": 0.0005117706023156643,
+      "learning_rate": 5.227677519374243e-06,
+      "loss": 0.0002,
+      "step": 2078
+    },
+    {
+      "epoch": 2.6592000000000002,
+      "grad_norm": 0.00013892765855416656,
+      "learning_rate": 5.2186710020522435e-06,
+      "loss": 0.0002,
+      "step": 2079
+    },
+    {
+      "epoch": 2.6604799999999997,
+      "grad_norm": 0.0001570197200635448,
+      "learning_rate": 5.209696485706404e-06,
+      "loss": 0.0002,
+      "step": 2080
+    },
+    {
+      "epoch": 2.66176,
+      "grad_norm": 0.0002135578397428617,
+      "learning_rate": 5.200753987450575e-06,
+      "loss": 0.0002,
+      "step": 2081
+    },
+    {
+      "epoch": 2.66304,
+      "grad_norm": 0.0005491600022651255,
+      "learning_rate": 5.1918435243375646e-06,
+      "loss": 0.0002,
+      "step": 2082
+    },
+    {
+      "epoch": 2.66432,
+      "grad_norm": 0.00010938257037196308,
+      "learning_rate": 5.182965113359075e-06,
+      "loss": 0.0002,
+      "step": 2083
+    },
+    {
+      "epoch": 2.6656,
+      "grad_norm": 0.00019497588800732046,
+      "learning_rate": 5.174118771445711e-06,
+      "loss": 0.0002,
+      "step": 2084
+    },
+    {
+      "epoch": 2.66688,
+      "grad_norm": 0.0001784377236617729,
+      "learning_rate": 5.165304515466897e-06,
+      "loss": 0.0002,
+      "step": 2085
+    },
+    {
+      "epoch": 2.66816,
+      "grad_norm": 0.0004209493927191943,
+      "learning_rate": 5.156522362230895e-06,
+      "loss": 0.0002,
+      "step": 2086
+    },
+    {
+      "epoch": 2.66944,
+      "grad_norm": 0.00045996581320650876,
+      "learning_rate": 5.147772328484731e-06,
+      "loss": 0.0002,
+      "step": 2087
+    },
+    {
+      "epoch": 2.67072,
+      "grad_norm": 0.00036037570680491626,
+      "learning_rate": 5.139054430914192e-06,
+      "loss": 0.0002,
+      "step": 2088
+    },
+    {
+      "epoch": 2.672,
+      "grad_norm": 0.00041035652975551784,
+      "learning_rate": 5.13036868614378e-06,
+      "loss": 0.0002,
+      "step": 2089
+    },
+    {
+      "epoch": 2.67328,
+      "grad_norm": 0.00014587903569918126,
+      "learning_rate": 5.121715110736679e-06,
+      "loss": 0.0002,
+      "step": 2090
+    },
+    {
+      "epoch": 2.67456,
+      "grad_norm": 0.00012942550529260188,
+      "learning_rate": 5.113093721194738e-06,
+      "loss": 0.0002,
+      "step": 2091
+    },
+    {
+      "epoch": 2.67584,
+      "grad_norm": 0.00011897664808202535,
+      "learning_rate": 5.104504533958419e-06,
+      "loss": 0.0002,
+      "step": 2092
+    },
+    {
+      "epoch": 2.67712,
+      "grad_norm": 0.00014001535600982606,
+      "learning_rate": 5.095947565406784e-06,
+      "loss": 0.0002,
+      "step": 2093
+    },
+    {
+      "epoch": 2.6784,
+      "grad_norm": 0.0005521393613889813,
+      "learning_rate": 5.087422831857449e-06,
+      "loss": 0.0002,
+      "step": 2094
+    },
+    {
+      "epoch": 2.67968,
+      "grad_norm": 0.0001651291677262634,
+      "learning_rate": 5.0789303495665664e-06,
+      "loss": 0.0002,
+      "step": 2095
+    },
+    {
+      "epoch": 2.68096,
+      "grad_norm": 0.0004139514931011945,
+      "learning_rate": 5.070470134728789e-06,
+      "loss": 0.0002,
+      "step": 2096
+    },
+    {
+      "epoch": 2.68224,
+      "grad_norm": 0.0003664352698251605,
+      "learning_rate": 5.062042203477227e-06,
+      "loss": 0.0002,
+      "step": 2097
+    },
+    {
+      "epoch": 2.68352,
+      "grad_norm": 9.784977009985596e-05,
+      "learning_rate": 5.053646571883431e-06,
+      "loss": 0.0002,
+      "step": 2098
+    },
+    {
+      "epoch": 2.6848,
+      "grad_norm": 0.0004598863306455314,
+      "learning_rate": 5.045283255957371e-06,
+      "loss": 0.0002,
+      "step": 2099
+    },
+    {
+      "epoch": 2.68608,
+      "grad_norm": 0.00018233667651657015,
+      "learning_rate": 5.036952271647375e-06,
+      "loss": 0.0002,
+      "step": 2100
+    },
+    {
+      "epoch": 2.68608,
+      "eval_loss": 1.1784363985061646,
+      "eval_runtime": 43.2782,
+      "eval_samples_per_second": 11.599,
+      "eval_steps_per_second": 1.456,
+      "step": 2100
+    },
+    {
+      "epoch": 2.68736,
+      "grad_norm": 0.0004271143116056919,
+      "learning_rate": 5.02865363484013e-06,
+      "loss": 0.0002,
+      "step": 2101
+    },
+    {
+      "epoch": 2.68864,
+      "grad_norm": 0.00019682984566316009,
+      "learning_rate": 5.020387361360624e-06,
+      "loss": 0.0002,
+      "step": 2102
+    },
+    {
+      "epoch": 2.68992,
+      "grad_norm": 0.00015684337995480746,
+      "learning_rate": 5.01215346697215e-06,
+      "loss": 0.0002,
+      "step": 2103
+    },
+    {
+      "epoch": 2.6912000000000003,
+      "grad_norm": 0.00022414051636587828,
+      "learning_rate": 5.003951967376241e-06,
+      "loss": 0.0002,
+      "step": 2104
+    },
+    {
+      "epoch": 2.6924799999999998,
+      "grad_norm": 0.0005924166762270033,
+      "learning_rate": 4.995782878212654e-06,
+      "loss": 0.0002,
+      "step": 2105
+    },
+    {
+      "epoch": 2.69376,
+      "grad_norm": 0.00012100968888262287,
+      "learning_rate": 4.9876462150593515e-06,
+      "loss": 0.0002,
+      "step": 2106
+    },
+    {
+      "epoch": 2.69504,
+      "grad_norm": 0.0006111165857873857,
+      "learning_rate": 4.979541993432454e-06,
+      "loss": 0.0002,
+      "step": 2107
+    },
+    {
+      "epoch": 2.69632,
+      "grad_norm": 0.00018479368009138852,
+      "learning_rate": 4.971470228786226e-06,
+      "loss": 0.0002,
+      "step": 2108
+    },
+    {
+      "epoch": 2.6976,
+      "grad_norm": 0.00019101858197245747,
+      "learning_rate": 4.963430936513024e-06,
+      "loss": 0.0002,
+      "step": 2109
+    },
+    {
+      "epoch": 2.69888,
+      "grad_norm": 0.0001369758101645857,
+      "learning_rate": 4.955424131943296e-06,
+      "loss": 0.0002,
+      "step": 2110
+    },
+    {
+      "epoch": 2.70016,
+      "grad_norm": 0.0006212744629010558,
+      "learning_rate": 4.947449830345536e-06,
+      "loss": 0.0002,
+      "step": 2111
+    },
+    {
+      "epoch": 2.70144,
+      "grad_norm": 0.00013725864118896425,
+      "learning_rate": 4.939508046926245e-06,
+      "loss": 0.0002,
+      "step": 2112
+    },
+    {
+      "epoch": 2.7027200000000002,
+      "grad_norm": 0.0003938393492717296,
+      "learning_rate": 4.931598796829925e-06,
+      "loss": 0.0002,
+      "step": 2113
+    },
+    {
+      "epoch": 2.7039999999999997,
+      "grad_norm": 0.0003750926407519728,
+      "learning_rate": 4.923722095139039e-06,
+      "loss": 0.0002,
+      "step": 2114
+    },
+    {
+      "epoch": 2.70528,
+      "grad_norm": 0.00019795184198301286,
+      "learning_rate": 4.915877956873979e-06,
+      "loss": 0.0002,
+      "step": 2115
+    },
+    {
+      "epoch": 2.70656,
+      "grad_norm": 0.0003393131191842258,
+      "learning_rate": 4.9080663969930395e-06,
+      "loss": 0.0002,
+      "step": 2116
+    },
+    {
+      "epoch": 2.70784,
+      "grad_norm": 0.0005614679539576173,
+      "learning_rate": 4.900287430392386e-06,
+      "loss": 0.0002,
+      "step": 2117
+    },
+    {
+      "epoch": 2.70912,
+      "grad_norm": 0.0001567447034176439,
+      "learning_rate": 4.892541071906042e-06,
+      "loss": 0.0002,
+      "step": 2118
+    },
+    {
+      "epoch": 2.7104,
+      "grad_norm": 0.00024158720043487847,
+      "learning_rate": 4.884827336305843e-06,
+      "loss": 0.0002,
+      "step": 2119
+    },
+    {
+      "epoch": 2.71168,
+      "grad_norm": 0.00022995199833530933,
+      "learning_rate": 4.8771462383014155e-06,
+      "loss": 0.0002,
+      "step": 2120
+    },
+    {
+      "epoch": 2.71296,
+      "grad_norm": 0.00029130213079042733,
+      "learning_rate": 4.86949779254014e-06,
+      "loss": 0.0002,
+      "step": 2121
+    },
+    {
+      "epoch": 2.71424,
+      "grad_norm": 0.0005213206168264151,
+      "learning_rate": 4.861882013607146e-06,
+      "loss": 0.0002,
+      "step": 2122
+    },
+    {
+      "epoch": 2.71552,
+      "grad_norm": 0.0003036532725673169,
+      "learning_rate": 4.854298916025262e-06,
+      "loss": 0.0002,
+      "step": 2123
+    },
+    {
+      "epoch": 2.7168,
+      "grad_norm": 0.0004486961697693914,
+      "learning_rate": 4.846748514254989e-06,
+      "loss": 0.0002,
+      "step": 2124
+    },
+    {
+      "epoch": 2.71808,
+      "grad_norm": 0.0008269405807368457,
+      "learning_rate": 4.839230822694487e-06,
+      "loss": 0.0002,
+      "step": 2125
+    },
+    {
+      "epoch": 2.71936,
+      "grad_norm": 0.00011923944111913443,
+      "learning_rate": 4.831745855679545e-06,
+      "loss": 0.0002,
+      "step": 2126
+    },
+    {
+      "epoch": 2.72064,
+      "grad_norm": 0.0005785307730548084,
+      "learning_rate": 4.824293627483531e-06,
+      "loss": 0.0002,
+      "step": 2127
+    },
+    {
+      "epoch": 2.72192,
+      "grad_norm": 0.000523311726283282,
+      "learning_rate": 4.816874152317396e-06,
+      "loss": 0.0002,
+      "step": 2128
+    },
+    {
+      "epoch": 2.7232,
+      "grad_norm": 0.0002992335648741573,
+      "learning_rate": 4.809487444329629e-06,
+      "loss": 0.0002,
+      "step": 2129
+    },
+    {
+      "epoch": 2.72448,
+      "grad_norm": 0.00010770787775982171,
+      "learning_rate": 4.802133517606236e-06,
+      "loss": 0.0002,
+      "step": 2130
+    },
+    {
+      "epoch": 2.72448,
+      "eval_loss": 1.1832425594329834,
+      "eval_runtime": 43.5517,
+      "eval_samples_per_second": 11.527,
+      "eval_steps_per_second": 1.447,
+      "step": 2130
+    },
+    {
+      "epoch": 2.72576,
+      "grad_norm": 0.0007023107027634978,
+      "learning_rate": 4.794812386170704e-06,
+      "loss": 0.0002,
+      "step": 2131
+    },
+    {
+      "epoch": 2.72704,
+      "grad_norm": 0.0002617916907183826,
+      "learning_rate": 4.787524063983993e-06,
+      "loss": 0.0002,
+      "step": 2132
+    },
+    {
+      "epoch": 2.72832,
+      "grad_norm": 0.00022266076121013612,
+      "learning_rate": 4.780268564944484e-06,
+      "loss": 0.0002,
+      "step": 2133
+    },
+    {
+      "epoch": 2.7296,
+      "grad_norm": 0.00012528970546554774,
+      "learning_rate": 4.773045902887979e-06,
+      "loss": 0.0002,
+      "step": 2134
+    },
+    {
+      "epoch": 2.73088,
+      "grad_norm": 0.00026173575315624475,
+      "learning_rate": 4.765856091587654e-06,
+      "loss": 0.0002,
+      "step": 2135
+    },
+    {
+      "epoch": 2.73216,
+      "grad_norm": 0.00034735462395474315,
+      "learning_rate": 4.758699144754043e-06,
+      "loss": 0.0002,
+      "step": 2136
+    },
+    {
+      "epoch": 2.73344,
+      "grad_norm": 0.00010721830039983615,
+      "learning_rate": 4.751575076035009e-06,
+      "loss": 0.0002,
+      "step": 2137
+    },
+    {
+      "epoch": 2.7347200000000003,
+      "grad_norm": 0.0005374618922360241,
+      "learning_rate": 4.744483899015722e-06,
+      "loss": 0.0002,
+      "step": 2138
+    },
+    {
+      "epoch": 2.7359999999999998,
+      "grad_norm": 0.00016130245057865977,
+      "learning_rate": 4.7374256272186205e-06,
+      "loss": 0.0002,
+      "step": 2139
+    },
+    {
+      "epoch": 2.73728,
+      "grad_norm": 0.00034650624729692936,
+      "learning_rate": 4.730400274103404e-06,
+      "loss": 0.0002,
+      "step": 2140
+    },
+    {
+      "epoch": 2.73856,
+      "grad_norm": 0.00032838908373378217,
+      "learning_rate": 4.723407853066994e-06,
+      "loss": 0.0002,
+      "step": 2141
+    },
+    {
+      "epoch": 2.73984,
+      "grad_norm": 9.886346379062161e-05,
+      "learning_rate": 4.716448377443515e-06,
+      "loss": 0.0002,
+      "step": 2142
+    },
+    {
+      "epoch": 2.74112,
+      "grad_norm": 0.0004354401316959411,
+      "learning_rate": 4.709521860504261e-06,
+      "loss": 0.0002,
+      "step": 2143
+    },
+    {
+      "epoch": 2.7424,
+      "grad_norm": 0.00011024817649740726,
+      "learning_rate": 4.702628315457683e-06,
+      "loss": 0.0002,
+      "step": 2144
+    },
+    {
+      "epoch": 2.74368,
+      "grad_norm": 0.00012188892287667841,
+      "learning_rate": 4.695767755449353e-06,
+      "loss": 0.0002,
+      "step": 2145
+    },
+    {
+      "epoch": 2.74496,
+      "grad_norm": 0.0003627574478741735,
+      "learning_rate": 4.68894019356194e-06,
+      "loss": 0.0002,
+      "step": 2146
+    },
+    {
+      "epoch": 2.7462400000000002,
+      "grad_norm": 0.0003473378310445696,
+      "learning_rate": 4.682145642815193e-06,
+      "loss": 0.0002,
+      "step": 2147
+    },
+    {
+      "epoch": 2.7475199999999997,
+      "grad_norm": 0.00028617039788514376,
+      "learning_rate": 4.675384116165903e-06,
+      "loss": 0.0002,
+      "step": 2148
+    },
+    {
+      "epoch": 2.7488,
+      "grad_norm": 0.0002804455580189824,
+      "learning_rate": 4.668655626507898e-06,
+      "loss": 0.0002,
+      "step": 2149
+    },
+    {
+      "epoch": 2.75008,
+      "grad_norm": 0.00034440791932865977,
+      "learning_rate": 4.661960186671997e-06,
+      "loss": 0.0002,
+      "step": 2150
+    },
+    {
+      "epoch": 2.75136,
+      "grad_norm": 0.00016433031123597175,
+      "learning_rate": 4.655297809425995e-06,
+      "loss": 0.0002,
+      "step": 2151
+    },
+    {
+      "epoch": 2.75264,
+      "grad_norm": 0.00031462355400435627,
+      "learning_rate": 4.648668507474642e-06,
+      "loss": 0.0002,
+      "step": 2152
+    },
+    {
+      "epoch": 2.75392,
+      "grad_norm": 0.0002593990648165345,
+      "learning_rate": 4.642072293459616e-06,
+      "loss": 0.0002,
+      "step": 2153
+    },
+    {
+      "epoch": 2.7552,
+      "grad_norm": 0.0001353595725959167,
+      "learning_rate": 4.635509179959497e-06,
+      "loss": 0.0002,
+      "step": 2154
+    },
+    {
+      "epoch": 2.75648,
+      "grad_norm": 0.00020742540073115379,
+      "learning_rate": 4.628979179489745e-06,
+      "loss": 0.0002,
+      "step": 2155
+    },
+    {
+      "epoch": 2.75776,
+      "grad_norm": 0.00020888213475700468,
+      "learning_rate": 4.62248230450267e-06,
+      "loss": 0.0002,
+      "step": 2156
+    },
+    {
+      "epoch": 2.75904,
+      "grad_norm": 0.0002596872509457171,
+      "learning_rate": 4.616018567387425e-06,
+      "loss": 0.0002,
+      "step": 2157
+    },
+    {
+      "epoch": 2.76032,
+      "grad_norm": 0.00011134509259136394,
+      "learning_rate": 4.609587980469958e-06,
+      "loss": 0.0002,
+      "step": 2158
+    },
+    {
+      "epoch": 2.7616,
+      "grad_norm": 0.0002468661987222731,
+      "learning_rate": 4.603190556013013e-06,
+      "loss": 0.0002,
+      "step": 2159
+    },
+    {
+      "epoch": 2.76288,
+      "grad_norm": 0.00023062645050231367,
+      "learning_rate": 4.596826306216087e-06,
+      "loss": 0.0002,
+      "step": 2160
+    },
+    {
+      "epoch": 2.76288,
+      "eval_loss": 1.179226279258728,
+      "eval_runtime": 43.5134,
+      "eval_samples_per_second": 11.537,
+      "eval_steps_per_second": 1.448,
+      "step": 2160
+    },
+    {
+      "epoch": 2.76416,
+      "grad_norm": 0.00036494573578238487,
+      "learning_rate": 4.590495243215422e-06,
+      "loss": 0.0002,
+      "step": 2161
+    },
+    {
+      "epoch": 2.76544,
+      "grad_norm": 0.00019542551308404654,
+      "learning_rate": 4.584197379083969e-06,
+      "loss": 0.0002,
+      "step": 2162
+    },
+    {
+      "epoch": 2.76672,
+      "grad_norm": 0.00019057006284128875,
+      "learning_rate": 4.577932725831367e-06,
+      "loss": 0.0002,
+      "step": 2163
+    },
+    {
+      "epoch": 2.768,
+      "grad_norm": 0.00048471824266016483,
+      "learning_rate": 4.571701295403939e-06,
+      "loss": 0.0002,
+      "step": 2164
+    },
+    {
+      "epoch": 2.76928,
+      "grad_norm": 0.00010409307287773117,
+      "learning_rate": 4.565503099684641e-06,
+      "loss": 0.0002,
+      "step": 2165
+    },
+    {
+      "epoch": 2.77056,
+      "grad_norm": 0.00016285658057313412,
+      "learning_rate": 4.559338150493056e-06,
+      "loss": 0.0002,
+      "step": 2166
+    },
+    {
+      "epoch": 2.77184,
+      "grad_norm": 0.00011712354171322659,
+      "learning_rate": 4.553206459585366e-06,
+      "loss": 0.0002,
+      "step": 2167
+    },
+    {
+      "epoch": 2.77312,
+      "grad_norm": 0.0002646244829520583,
+      "learning_rate": 4.547108038654336e-06,
+      "loss": 0.0002,
+      "step": 2168
+    },
+    {
+      "epoch": 2.7744,
+      "grad_norm": 0.0004395518044475466,
+      "learning_rate": 4.541042899329287e-06,
+      "loss": 0.0002,
+      "step": 2169
+    },
+    {
+      "epoch": 2.77568,
+      "grad_norm": 0.00015305139822885394,
+      "learning_rate": 4.535011053176065e-06,
+      "loss": 0.0002,
+      "step": 2170
+    },
+    {
+      "epoch": 2.77696,
+      "grad_norm": 0.00021129405649844557,
+      "learning_rate": 4.52901251169704e-06,
+      "loss": 0.0002,
+      "step": 2171
+    },
+    {
+      "epoch": 2.7782400000000003,
+      "grad_norm": 0.0003616804606281221,
+      "learning_rate": 4.523047286331064e-06,
+      "loss": 0.0002,
+      "step": 2172
+    },
+    {
+      "epoch": 2.7795199999999998,
+      "grad_norm": 0.00016815500566735864,
+      "learning_rate": 4.517115388453462e-06,
+      "loss": 0.0002,
+      "step": 2173
+    },
+    {
+      "epoch": 2.7808,
+      "grad_norm": 0.00023923354456201196,
+      "learning_rate": 4.511216829376001e-06,
+      "loss": 0.0002,
+      "step": 2174
+    },
+    {
+      "epoch": 2.78208,
+      "grad_norm": 0.00020838297496084124,
+      "learning_rate": 4.505351620346878e-06,
+      "loss": 0.0002,
+      "step": 2175
+    },
+    {
+      "epoch": 2.78336,
+      "grad_norm": 0.00019640865502879024,
+      "learning_rate": 4.49951977255069e-06,
+      "loss": 0.0002,
+      "step": 2176
+    },
+    {
+      "epoch": 2.78464,
+      "grad_norm": 0.0001642953575355932,
+      "learning_rate": 4.493721297108412e-06,
+      "loss": 0.0002,
+      "step": 2177
+    },
+    {
+      "epoch": 2.78592,
+      "grad_norm": 0.0002643915649969131,
+      "learning_rate": 4.487956205077392e-06,
+      "loss": 0.0002,
+      "step": 2178
+    },
+    {
+      "epoch": 2.7872,
+      "grad_norm": 0.00023071031318977475,
+      "learning_rate": 4.482224507451309e-06,
+      "loss": 0.0002,
+      "step": 2179
+    },
+    {
+      "epoch": 2.78848,
+      "grad_norm": 0.0002969163761008531,
+      "learning_rate": 4.476526215160157e-06,
+      "loss": 0.0002,
+      "step": 2180
+    },
+    {
+      "epoch": 2.7897600000000002,
+      "grad_norm": 0.00031866165227256715,
+      "learning_rate": 4.470861339070243e-06,
+      "loss": 0.0002,
+      "step": 2181
+    },
+    {
+      "epoch": 2.7910399999999997,
+      "grad_norm": 0.0005999641725793481,
+      "learning_rate": 4.465229889984131e-06,
+      "loss": 0.0002,
+      "step": 2182
+    },
+    {
+      "epoch": 2.79232,
+      "grad_norm": 0.0003952770493924618,
+      "learning_rate": 4.459631878640661e-06,
+      "loss": 0.0002,
+      "step": 2183
+    },
+    {
+      "epoch": 2.7936,
+      "grad_norm": 0.0008258552988991141,
+      "learning_rate": 4.454067315714897e-06,
+      "loss": 0.0002,
+      "step": 2184
+    },
+    {
+      "epoch": 2.79488,
+      "grad_norm": 0.00019526865798979998,
+      "learning_rate": 4.448536211818127e-06,
+      "loss": 0.0002,
+      "step": 2185
+    },
+    {
+      "epoch": 2.79616,
+      "grad_norm": 0.0006488322978839278,
+      "learning_rate": 4.443038577497828e-06,
+      "loss": 0.0002,
+      "step": 2186
+    },
+    {
+      "epoch": 2.79744,
+      "grad_norm": 0.000724302779417485,
+      "learning_rate": 4.437574423237652e-06,
+      "loss": 0.0002,
+      "step": 2187
+    },
+    {
+      "epoch": 2.79872,
+      "grad_norm": 0.00013176088395994157,
+      "learning_rate": 4.432143759457415e-06,
+      "loss": 0.0002,
+      "step": 2188
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.0008997769327834249,
+      "learning_rate": 4.42674659651306e-06,
+      "loss": 0.0002,
+      "step": 2189
+    },
+    {
+      "epoch": 2.80128,
+      "grad_norm": 0.0004246947355568409,
+      "learning_rate": 4.4213829446966495e-06,
+      "loss": 0.0002,
+      "step": 2190
+    },
+    {
+      "epoch": 2.80128,
+      "eval_loss": 1.1821544170379639,
+      "eval_runtime": 43.5085,
+      "eval_samples_per_second": 11.538,
+      "eval_steps_per_second": 1.448,
+      "step": 2190
+    },
+    {
+      "epoch": 2.80256,
+      "grad_norm": 0.00038387291715480387,
+      "learning_rate": 4.416052814236344e-06,
+      "loss": 0.0002,
+      "step": 2191
+    },
+    {
+      "epoch": 2.80384,
+      "grad_norm": 0.00042236316949129105,
+      "learning_rate": 4.410756215296375e-06,
+      "loss": 0.0002,
+      "step": 2192
+    },
+    {
+      "epoch": 2.80512,
+      "grad_norm": 0.00043121661292389035,
+      "learning_rate": 4.405493157977039e-06,
+      "loss": 0.0002,
+      "step": 2193
+    },
+    {
+      "epoch": 2.8064,
+      "grad_norm": 0.00012004785821773112,
+      "learning_rate": 4.400263652314666e-06,
+      "loss": 0.0002,
+      "step": 2194
+    },
+    {
+      "epoch": 2.80768,
+      "grad_norm": 0.0002733418659772724,
+      "learning_rate": 4.395067708281608e-06,
+      "loss": 0.0002,
+      "step": 2195
+    },
+    {
+      "epoch": 2.80896,
+      "grad_norm": 0.00019762851297855377,
+      "learning_rate": 4.389905335786212e-06,
+      "loss": 0.0002,
+      "step": 2196
+    },
+    {
+      "epoch": 2.81024,
+      "grad_norm": 0.00014854299661237746,
+      "learning_rate": 4.3847765446728054e-06,
+      "loss": 0.0002,
+      "step": 2197
+    },
+    {
+      "epoch": 2.81152,
+      "grad_norm": 0.0003144988731946796,
+      "learning_rate": 4.3796813447216906e-06,
+      "loss": 0.0002,
+      "step": 2198
+    },
+    {
+      "epoch": 2.8128,
+      "grad_norm": 0.0001517291384516284,
+      "learning_rate": 4.374619745649099e-06,
+      "loss": 0.0002,
+      "step": 2199
+    },
+    {
+      "epoch": 2.81408,
+      "grad_norm": 0.0002452972694300115,
+      "learning_rate": 4.369591757107193e-06,
+      "loss": 0.0002,
+      "step": 2200
+    },
+    {
+      "epoch": 2.81536,
+      "grad_norm": 0.0003286633000243455,
+      "learning_rate": 4.3645973886840434e-06,
+      "loss": 0.0002,
+      "step": 2201
+    },
+    {
+      "epoch": 2.81664,
+      "grad_norm": 0.0004903704393655062,
+      "learning_rate": 4.359636649903608e-06,
+      "loss": 0.0002,
+      "step": 2202
+    },
+    {
+      "epoch": 2.81792,
+      "grad_norm": 0.00029815713060088456,
+      "learning_rate": 4.354709550225714e-06,
+      "loss": 0.0002,
+      "step": 2203
+    },
+    {
+      "epoch": 2.8192,
+      "grad_norm": 0.00015331384201999754,
+      "learning_rate": 4.34981609904604e-06,
+      "loss": 0.0002,
+      "step": 2204
+    },
+    {
+      "epoch": 2.82048,
+      "grad_norm": 0.000682519341353327,
+      "learning_rate": 4.344956305696104e-06,
+      "loss": 0.0002,
+      "step": 2205
+    },
+    {
+      "epoch": 2.8217600000000003,
+      "grad_norm": 0.0003466587222646922,
+      "learning_rate": 4.3401301794432325e-06,
+      "loss": 0.0002,
+      "step": 2206
+    },
+    {
+      "epoch": 2.8230399999999998,
+      "grad_norm": 0.0006293237674981356,
+      "learning_rate": 4.335337729490559e-06,
+      "loss": 0.0002,
+      "step": 2207
+    },
+    {
+      "epoch": 2.82432,
+      "grad_norm": 0.0004514796892181039,
+      "learning_rate": 4.330578964976992e-06,
+      "loss": 0.0002,
+      "step": 2208
+    },
+    {
+      "epoch": 2.8256,
+      "grad_norm": 9.345506259705871e-05,
+      "learning_rate": 4.32585389497721e-06,
+      "loss": 0.0002,
+      "step": 2209
+    },
+    {
+      "epoch": 2.82688,
+      "grad_norm": 0.0005342678050510585,
+      "learning_rate": 4.321162528501635e-06,
+      "loss": 0.0002,
+      "step": 2210
+    },
+    {
+      "epoch": 2.82816,
+      "grad_norm": 0.00011190603981958702,
+      "learning_rate": 4.316504874496412e-06,
+      "loss": 0.0002,
+      "step": 2211
+    },
+    {
+      "epoch": 2.82944,
+      "grad_norm": 0.00010925935202976689,
+      "learning_rate": 4.3118809418434106e-06,
+      "loss": 0.0002,
+      "step": 2212
+    },
+    {
+      "epoch": 2.83072,
+      "grad_norm": 0.00012517714640125632,
+      "learning_rate": 4.307290739360186e-06,
+      "loss": 0.0002,
+      "step": 2213
+    },
+    {
+      "epoch": 2.832,
+      "grad_norm": 0.00015721505042165518,
+      "learning_rate": 4.302734275799981e-06,
+      "loss": 0.0002,
+      "step": 2214
+    },
+    {
+      "epoch": 2.8332800000000002,
+      "grad_norm": 0.00013470579870045185,
+      "learning_rate": 4.2982115598516915e-06,
+      "loss": 0.0002,
+      "step": 2215
+    },
+    {
+      "epoch": 2.8345599999999997,
+      "grad_norm": 0.0003435101534705609,
+      "learning_rate": 4.293722600139862e-06,
+      "loss": 0.0002,
+      "step": 2216
+    },
+    {
+      "epoch": 2.83584,
+      "grad_norm": 0.00033612651168368757,
+      "learning_rate": 4.289267405224666e-06,
+      "loss": 0.0002,
+      "step": 2217
+    },
+    {
+      "epoch": 2.83712,
+      "grad_norm": 0.0004498395719565451,
+      "learning_rate": 4.284845983601892e-06,
+      "loss": 0.0002,
+      "step": 2218
+    },
+    {
+      "epoch": 2.8384,
+      "grad_norm": 0.00033877856913022697,
+      "learning_rate": 4.280458343702924e-06,
+      "loss": 0.0002,
+      "step": 2219
+    },
+    {
+      "epoch": 2.83968,
+      "grad_norm": 0.00011193512909812853,
+      "learning_rate": 4.2761044938947214e-06,
+      "loss": 0.0002,
+      "step": 2220
+    },
+    {
+      "epoch": 2.83968,
+      "eval_loss": 1.1815558671951294,
+      "eval_runtime": 43.5068,
+      "eval_samples_per_second": 11.538,
+      "eval_steps_per_second": 1.448,
+      "step": 2220
+    },
+    {
+      "epoch": 2.84096,
+      "grad_norm": 0.0005381443188525736,
+      "learning_rate": 4.271784442479811e-06,
+      "loss": 0.0002,
+      "step": 2221
+    },
+    {
+      "epoch": 2.84224,
+      "grad_norm": 0.000222622329602018,
+      "learning_rate": 4.2674981976962724e-06,
+      "loss": 0.0002,
+      "step": 2222
+    },
+    {
+      "epoch": 2.84352,
+      "grad_norm": 0.0003087735385634005,
+      "learning_rate": 4.263245767717712e-06,
+      "loss": 0.0002,
+      "step": 2223
+    },
+    {
+      "epoch": 2.8448,
+      "grad_norm": 0.00019644832354970276,
+      "learning_rate": 4.259027160653258e-06,
+      "loss": 0.0002,
+      "step": 2224
+    },
+    {
+      "epoch": 2.84608,
+      "grad_norm": 0.00019042703206650913,
+      "learning_rate": 4.254842384547535e-06,
+      "loss": 0.0002,
+      "step": 2225
+    },
+    {
+      "epoch": 2.84736,
+      "grad_norm": 0.0001009497937047854,
+      "learning_rate": 4.250691447380659e-06,
+      "loss": 0.0002,
+      "step": 2226
+    },
+    {
+      "epoch": 2.84864,
+      "grad_norm": 0.0005751587450504303,
+      "learning_rate": 4.2465743570682134e-06,
+      "loss": 0.0002,
+      "step": 2227
+    },
+    {
+      "epoch": 2.84992,
+      "grad_norm": 0.0003292016335763037,
+      "learning_rate": 4.242491121461242e-06,
+      "loss": 0.0002,
+      "step": 2228
+    },
+    {
+      "epoch": 2.8512,
+      "grad_norm": 0.0004692550573963672,
+      "learning_rate": 4.23844174834623e-06,
+      "loss": 0.0002,
+      "step": 2229
+    },
+    {
+      "epoch": 2.85248,
+      "grad_norm": 0.000195763394003734,
+      "learning_rate": 4.234426245445078e-06,
+      "loss": 0.0002,
+      "step": 2230
+    },
+    {
+      "epoch": 2.85376,
+      "grad_norm": 0.00020026469428557903,
+      "learning_rate": 4.230444620415114e-06,
+      "loss": 0.0002,
+      "step": 2231
+    },
+    {
+      "epoch": 2.85504,
+      "grad_norm": 0.0003763276035897434,
+      "learning_rate": 4.226496880849052e-06,
+      "loss": 0.0002,
+      "step": 2232
+    },
+    {
+      "epoch": 2.85632,
+      "grad_norm": 0.00046748522436246276,
+      "learning_rate": 4.222583034274991e-06,
+      "loss": 0.0002,
+      "step": 2233
+    },
+    {
+      "epoch": 2.8576,
+      "grad_norm": 0.0001498359051765874,
+      "learning_rate": 4.218703088156401e-06,
+      "loss": 0.0002,
+      "step": 2234
+    },
+    {
+      "epoch": 2.85888,
+      "grad_norm": 0.0005568740307353437,
+      "learning_rate": 4.214857049892103e-06,
+      "loss": 0.0002,
+      "step": 2235
+    },
+    {
+      "epoch": 2.86016,
+      "grad_norm": 0.0005128760822117329,
+      "learning_rate": 4.211044926816259e-06,
+      "loss": 0.0002,
+      "step": 2236
+    },
+    {
+      "epoch": 2.86144,
+      "grad_norm": 0.0001381375186610967,
+      "learning_rate": 4.207266726198356e-06,
+      "loss": 0.0002,
+      "step": 2237
+    },
+    {
+      "epoch": 2.86272,
+      "grad_norm": 0.00025074410950765014,
+      "learning_rate": 4.203522455243192e-06,
+      "loss": 0.0002,
+      "step": 2238
+    },
+    {
+      "epoch": 2.864,
+      "grad_norm": 0.0009641392389312387,
+      "learning_rate": 4.199812121090866e-06,
+      "loss": 0.0002,
+      "step": 2239
+    },
+    {
+      "epoch": 2.8652800000000003,
+      "grad_norm": 0.00031504526850767434,
+      "learning_rate": 4.196135730816762e-06,
+      "loss": 0.0002,
+      "step": 2240
+    },
+    {
+      "epoch": 2.8665599999999998,
+      "grad_norm": 0.00048518256517127156,
+      "learning_rate": 4.19249329143153e-06,
+      "loss": 0.0002,
+      "step": 2241
+    },
+    {
+      "epoch": 2.86784,
+      "grad_norm": 0.0004062238149344921,
+      "learning_rate": 4.18888480988108e-06,
+      "loss": 0.0002,
+      "step": 2242
+    },
+    {
+      "epoch": 2.86912,
+      "grad_norm": 0.0003679286455735564,
+      "learning_rate": 4.185310293046573e-06,
+      "loss": 0.0002,
+      "step": 2243
+    },
+    {
+      "epoch": 2.8704,
+      "grad_norm": 0.00010955840116366744,
+      "learning_rate": 4.1817697477443895e-06,
+      "loss": 0.0002,
+      "step": 2244
+    },
+    {
+      "epoch": 2.87168,
+      "grad_norm": 0.00031511441920883954,
+      "learning_rate": 4.178263180726138e-06,
+      "loss": 0.0002,
+      "step": 2245
+    },
+    {
+      "epoch": 2.87296,
+      "grad_norm": 0.0006907782517373562,
+      "learning_rate": 4.1747905986786295e-06,
+      "loss": 0.0002,
+      "step": 2246
+    },
+    {
+      "epoch": 2.87424,
+      "grad_norm": 0.00010518222552491352,
+      "learning_rate": 4.171352008223863e-06,
+      "loss": 0.0002,
+      "step": 2247
+    },
+    {
+      "epoch": 2.87552,
+      "grad_norm": 0.0002229425881523639,
+      "learning_rate": 4.167947415919027e-06,
+      "loss": 0.0002,
+      "step": 2248
+    },
+    {
+      "epoch": 2.8768000000000002,
+      "grad_norm": 0.0005626695929095149,
+      "learning_rate": 4.1645768282564715e-06,
+      "loss": 0.0002,
+      "step": 2249
+    },
+    {
+      "epoch": 2.8780799999999997,
+      "grad_norm": 0.0001849379186751321,
+      "learning_rate": 4.1612402516637e-06,
+      "loss": 0.0002,
+      "step": 2250
+    },
+    {
+      "epoch": 2.8780799999999997,
+      "eval_loss": 1.1795929670333862,
+      "eval_runtime": 43.3675,
+      "eval_samples_per_second": 11.575,
+      "eval_steps_per_second": 1.453,
+      "step": 2250
+    },
+    {
+      "epoch": 2.87936,
+      "grad_norm": 0.00012294574116822332,
+      "learning_rate": 4.157937692503361e-06,
+      "loss": 0.0002,
+      "step": 2251
+    },
+    {
+      "epoch": 2.88064,
+      "grad_norm": 0.00041280241566710174,
+      "learning_rate": 4.154669157073239e-06,
+      "loss": 0.0002,
+      "step": 2252
+    },
+    {
+      "epoch": 2.88192,
+      "grad_norm": 0.00011976512178080156,
+      "learning_rate": 4.15143465160623e-06,
+      "loss": 0.0002,
+      "step": 2253
+    },
+    {
+      "epoch": 2.8832,
+      "grad_norm": 0.0003270368033554405,
+      "learning_rate": 4.148234182270339e-06,
+      "loss": 0.0002,
+      "step": 2254
+    },
+    {
+      "epoch": 2.88448,
+      "grad_norm": 0.0001754332333803177,
+      "learning_rate": 4.145067755168667e-06,
+      "loss": 0.0002,
+      "step": 2255
+    },
+    {
+      "epoch": 2.88576,
+      "grad_norm": 0.00020488614973146468,
+      "learning_rate": 4.141935376339401e-06,
+      "loss": 0.0002,
+      "step": 2256
+    },
+    {
+      "epoch": 2.88704,
+      "grad_norm": 0.0003626338439062238,
+      "learning_rate": 4.138837051755794e-06,
+      "loss": 0.0002,
+      "step": 2257
+    },
+    {
+      "epoch": 2.88832,
+      "grad_norm": 0.0001282071170862764,
+      "learning_rate": 4.135772787326164e-06,
+      "loss": 0.0002,
+      "step": 2258
+    },
+    {
+      "epoch": 2.8895999999999997,
+      "grad_norm": 0.000176410234416835,
+      "learning_rate": 4.132742588893878e-06,
+      "loss": 0.0002,
+      "step": 2259
+    },
+    {
+      "epoch": 2.89088,
+      "grad_norm": 9.673011663835496e-05,
+      "learning_rate": 4.129746462237341e-06,
+      "loss": 0.0002,
+      "step": 2260
+    },
+    {
+      "epoch": 2.89216,
+      "grad_norm": 0.0004994468181394041,
+      "learning_rate": 4.126784413069985e-06,
+      "loss": 0.0002,
+      "step": 2261
+    },
+    {
+      "epoch": 2.89344,
+      "grad_norm": 0.0002942743303719908,
+      "learning_rate": 4.123856447040254e-06,
+      "loss": 0.0002,
+      "step": 2262
+    },
+    {
+      "epoch": 2.89472,
+      "grad_norm": 0.0006747826701030135,
+      "learning_rate": 4.120962569731607e-06,
+      "loss": 0.0002,
+      "step": 2263
+    },
+    {
+      "epoch": 2.896,
+      "grad_norm": 0.0002655930002219975,
+      "learning_rate": 4.118102786662489e-06,
+      "loss": 0.0002,
+      "step": 2264
+    },
+    {
+      "epoch": 2.89728,
+      "grad_norm": 0.00012897778651677072,
+      "learning_rate": 4.115277103286335e-06,
+      "loss": 0.0002,
+      "step": 2265
+    },
+    {
+      "epoch": 2.89856,
+      "grad_norm": 0.0007293322705663741,
+      "learning_rate": 4.11248552499155e-06,
+      "loss": 0.0002,
+      "step": 2266
+    },
+    {
+      "epoch": 2.89984,
+      "grad_norm": 0.0006220631767064333,
+      "learning_rate": 4.1097280571015044e-06,
+      "loss": 0.0002,
+      "step": 2267
+    },
+    {
+      "epoch": 2.90112,
+      "grad_norm": 0.00011390577856218442,
+      "learning_rate": 4.107004704874524e-06,
+      "loss": 0.0002,
+      "step": 2268
+    },
+    {
+      "epoch": 2.9024,
+      "grad_norm": 0.0009089137311093509,
+      "learning_rate": 4.104315473503875e-06,
+      "loss": 0.0002,
+      "step": 2269
+    },
+    {
+      "epoch": 2.90368,
+      "grad_norm": 0.0007660266710445285,
+      "learning_rate": 4.101660368117756e-06,
+      "loss": 0.0002,
+      "step": 2270
+    },
+    {
+      "epoch": 2.90496,
+      "grad_norm": 0.00010220236435998231,
+      "learning_rate": 4.099039393779296e-06,
+      "loss": 0.0002,
+      "step": 2271
+    },
+    {
+      "epoch": 2.90624,
+      "grad_norm": 0.0007261789869517088,
+      "learning_rate": 4.096452555486533e-06,
+      "loss": 0.0002,
+      "step": 2272
+    },
+    {
+      "epoch": 2.90752,
+      "grad_norm": 0.0006142670172266662,
+      "learning_rate": 4.0938998581724085e-06,
+      "loss": 0.0002,
+      "step": 2273
+    },
+    {
+      "epoch": 2.9088000000000003,
+      "grad_norm": 0.00012708887516055256,
+      "learning_rate": 4.0913813067047645e-06,
+      "loss": 0.0002,
+      "step": 2274
+    },
+    {
+      "epoch": 2.91008,
+      "grad_norm": 0.00042775258771143854,
+      "learning_rate": 4.088896905886322e-06,
+      "loss": 0.0002,
+      "step": 2275
+    },
+    {
+      "epoch": 2.91136,
+      "grad_norm": 0.0002019358944380656,
+      "learning_rate": 4.086446660454687e-06,
+      "loss": 0.0002,
+      "step": 2276
+    },
+    {
+      "epoch": 2.91264,
+      "grad_norm": 0.0001991699100472033,
+      "learning_rate": 4.084030575082325e-06,
+      "loss": 0.0002,
+      "step": 2277
+    },
+    {
+      "epoch": 2.91392,
+      "grad_norm": 0.0001914306340040639,
+      "learning_rate": 4.081648654376567e-06,
+      "loss": 0.0002,
+      "step": 2278
+    },
+    {
+      "epoch": 2.9152,
+      "grad_norm": 0.00037433000397868454,
+      "learning_rate": 4.07930090287959e-06,
+      "loss": 0.0002,
+      "step": 2279
+    },
+    {
+      "epoch": 2.91648,
+      "grad_norm": 0.00021021462453063577,
+      "learning_rate": 4.076987325068415e-06,
+      "loss": 0.0002,
+      "step": 2280
+    },
+    {
+      "epoch": 2.91648,
+      "eval_loss": 1.180129051208496,
+      "eval_runtime": 43.4225,
+      "eval_samples_per_second": 11.561,
+      "eval_steps_per_second": 1.451,
+      "step": 2280
+    },
+    {
+      "epoch": 2.91776,
+      "grad_norm": 0.00023146902094595134,
+      "learning_rate": 4.074707925354891e-06,
+      "loss": 0.0002,
+      "step": 2281
+    },
+    {
+      "epoch": 2.91904,
+      "grad_norm": 0.0002690425608307123,
+      "learning_rate": 4.072462708085697e-06,
+      "loss": 0.0002,
+      "step": 2282
+    },
+    {
+      "epoch": 2.9203200000000002,
+      "grad_norm": 0.00017321344057563692,
+      "learning_rate": 4.07025167754233e-06,
+      "loss": 0.0002,
+      "step": 2283
+    },
+    {
+      "epoch": 2.9215999999999998,
+      "grad_norm": 0.0005131001817062497,
+      "learning_rate": 4.068074837941084e-06,
+      "loss": 0.0002,
+      "step": 2284
+    },
+    {
+      "epoch": 2.92288,
+      "grad_norm": 0.0001270236971322447,
+      "learning_rate": 4.065932193433064e-06,
+      "loss": 0.0002,
+      "step": 2285
+    },
+    {
+      "epoch": 2.92416,
+      "grad_norm": 0.00018355673819314688,
+      "learning_rate": 4.063823748104164e-06,
+      "loss": 0.0002,
+      "step": 2286
+    },
+    {
+      "epoch": 2.92544,
+      "grad_norm": 0.00037108006654307246,
+      "learning_rate": 4.0617495059750585e-06,
+      "loss": 0.0002,
+      "step": 2287
+    },
+    {
+      "epoch": 2.92672,
+      "grad_norm": 0.0005037841037847102,
+      "learning_rate": 4.059709471001203e-06,
+      "loss": 0.0002,
+      "step": 2288
+    },
+    {
+      "epoch": 2.928,
+      "grad_norm": 0.00011262594489380717,
+      "learning_rate": 4.057703647072821e-06,
+      "loss": 0.0002,
+      "step": 2289
+    },
+    {
+      "epoch": 2.92928,
+      "grad_norm": 0.0007866703672334552,
+      "learning_rate": 4.055732038014895e-06,
+      "loss": 0.0002,
+      "step": 2290
+    },
+    {
+      "epoch": 2.93056,
+      "grad_norm": 0.00035812563146464527,
+      "learning_rate": 4.053794647587166e-06,
+      "loss": 0.0002,
+      "step": 2291
+    },
+    {
+      "epoch": 2.9318400000000002,
+      "grad_norm": 0.00012094627163605765,
+      "learning_rate": 4.051891479484118e-06,
+      "loss": 0.0002,
+      "step": 2292
+    },
+    {
+      "epoch": 2.9331199999999997,
+      "grad_norm": 0.00018673752492759377,
+      "learning_rate": 4.050022537334979e-06,
+      "loss": 0.0002,
+      "step": 2293
+    },
+    {
+      "epoch": 2.9344,
+      "grad_norm": 0.0005387719720602036,
+      "learning_rate": 4.048187824703708e-06,
+      "loss": 0.0002,
+      "step": 2294
+    },
+    {
+      "epoch": 2.93568,
+      "grad_norm": 0.000565039983484894,
+      "learning_rate": 4.046387345088988e-06,
+      "loss": 0.0002,
+      "step": 2295
+    },
+    {
+      "epoch": 2.93696,
+      "grad_norm": 0.0003313857887405902,
+      "learning_rate": 4.044621101924227e-06,
+      "loss": 0.0002,
+      "step": 2296
+    },
+    {
+      "epoch": 2.93824,
+      "grad_norm": 0.00014667944924440235,
+      "learning_rate": 4.042889098577545e-06,
+      "loss": 0.0002,
+      "step": 2297
+    },
+    {
+      "epoch": 2.93952,
+      "grad_norm": 0.0001700992143014446,
+      "learning_rate": 4.041191338351762e-06,
+      "loss": 0.0002,
+      "step": 2298
+    },
+    {
+      "epoch": 2.9408,
+      "grad_norm": 0.000125421880511567,
+      "learning_rate": 4.03952782448441e-06,
+      "loss": 0.0002,
+      "step": 2299
+    },
+    {
+      "epoch": 2.94208,
+      "grad_norm": 0.0002756211906671524,
+      "learning_rate": 4.037898560147704e-06,
+      "loss": 0.0002,
+      "step": 2300
+    },
+    {
+      "epoch": 2.94336,
+      "grad_norm": 0.00016799187869764864,
+      "learning_rate": 4.036303548448556e-06,
+      "loss": 0.0002,
+      "step": 2301
+    },
+    {
+      "epoch": 2.94464,
+      "grad_norm": 0.0005784616805613041,
+      "learning_rate": 4.034742792428553e-06,
+      "loss": 0.0002,
+      "step": 2302
+    },
+    {
+      "epoch": 2.94592,
+      "grad_norm": 0.00011791891301982105,
+      "learning_rate": 4.033216295063964e-06,
+      "loss": 0.0002,
+      "step": 2303
+    },
+    {
+      "epoch": 2.9472,
+      "grad_norm": 0.00036018676473759115,
+      "learning_rate": 4.031724059265726e-06,
+      "loss": 0.0002,
+      "step": 2304
+    },
+    {
+      "epoch": 2.94848,
+      "grad_norm": 0.00035507127176970243,
+      "learning_rate": 4.0302660878794435e-06,
+      "loss": 0.0002,
+      "step": 2305
+    },
+    {
+      "epoch": 2.94976,
+      "grad_norm": 0.0001190779876196757,
+      "learning_rate": 4.0288423836853805e-06,
+      "loss": 0.0002,
+      "step": 2306
+    },
+    {
+      "epoch": 2.95104,
+      "grad_norm": 0.0005441626999527216,
+      "learning_rate": 4.02745294939845e-06,
+      "loss": 0.0002,
+      "step": 2307
+    },
+    {
+      "epoch": 2.9523200000000003,
+      "grad_norm": 0.0001164350105682388,
+      "learning_rate": 4.026097787668224e-06,
+      "loss": 0.0002,
+      "step": 2308
+    },
+    {
+      "epoch": 2.9536,
+      "grad_norm": 0.0001243723090738058,
+      "learning_rate": 4.0247769010789095e-06,
+      "loss": 0.0002,
+      "step": 2309
+    },
+    {
+      "epoch": 2.95488,
+      "grad_norm": 0.00012079760199412704,
+      "learning_rate": 4.023490292149359e-06,
+      "loss": 0.0002,
+      "step": 2310
+    },
+    {
+      "epoch": 2.95488,
+      "eval_loss": 1.1803983449935913,
+      "eval_runtime": 43.384,
+      "eval_samples_per_second": 11.571,
+      "eval_steps_per_second": 1.452,
+      "step": 2310
+    },
+    {
+      "epoch": 2.95616,
+      "grad_norm": 0.00018000841373577714,
+      "learning_rate": 4.022237963333059e-06,
+      "loss": 0.0002,
+      "step": 2311
+    },
+    {
+      "epoch": 2.95744,
+      "grad_norm": 0.00032567320158705115,
+      "learning_rate": 4.021019917018121e-06,
+      "loss": 0.0002,
+      "step": 2312
+    },
+    {
+      "epoch": 2.95872,
+      "grad_norm": 0.0001935055770445615,
+      "learning_rate": 4.01983615552729e-06,
+      "loss": 0.0002,
+      "step": 2313
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.00019914534641429782,
+      "learning_rate": 4.0186866811179235e-06,
+      "loss": 0.0002,
+      "step": 2314
+    },
+    {
+      "epoch": 2.96128,
+      "grad_norm": 0.0004246193275321275,
+      "learning_rate": 4.017571495982e-06,
+      "loss": 0.0002,
+      "step": 2315
+    },
+    {
+      "epoch": 2.96256,
+      "grad_norm": 0.0001828927779570222,
+      "learning_rate": 4.016490602246111e-06,
+      "loss": 0.0002,
+      "step": 2316
+    },
+    {
+      "epoch": 2.9638400000000003,
+      "grad_norm": 0.00032234840909950435,
+      "learning_rate": 4.015444001971456e-06,
+      "loss": 0.0002,
+      "step": 2317
+    },
+    {
+      "epoch": 2.9651199999999998,
+      "grad_norm": 0.0006089527742005885,
+      "learning_rate": 4.014431697153837e-06,
+      "loss": 0.0002,
+      "step": 2318
+    },
+    {
+      "epoch": 2.9664,
+      "grad_norm": 0.00029877014458179474,
+      "learning_rate": 4.013453689723657e-06,
+      "loss": 0.0002,
+      "step": 2319
+    },
+    {
+      "epoch": 2.96768,
+      "grad_norm": 0.0003506283101160079,
+      "learning_rate": 4.01250998154592e-06,
+      "loss": 0.0002,
+      "step": 2320
+    },
+    {
+      "epoch": 2.96896,
+      "grad_norm": 0.0006300249951891601,
+      "learning_rate": 4.011600574420216e-06,
+      "loss": 0.0002,
+      "step": 2321
+    },
+    {
+      "epoch": 2.97024,
+      "grad_norm": 0.0005357094341889024,
+      "learning_rate": 4.010725470080733e-06,
+      "loss": 0.0002,
+      "step": 2322
+    },
+    {
+      "epoch": 2.97152,
+      "grad_norm": 0.00011800980428233743,
+      "learning_rate": 4.009884670196239e-06,
+      "loss": 0.0002,
+      "step": 2323
+    },
+    {
+      "epoch": 2.9728,
+      "grad_norm": 0.0004599698877427727,
+      "learning_rate": 4.009078176370089e-06,
+      "loss": 0.0002,
+      "step": 2324
+    },
+    {
+      "epoch": 2.97408,
+      "grad_norm": 0.0005400791415013373,
+      "learning_rate": 4.008305990140219e-06,
+      "loss": 0.0002,
+      "step": 2325
+    },
+    {
+      "epoch": 2.9753600000000002,
+      "grad_norm": 0.00037052438710816205,
+      "learning_rate": 4.00756811297914e-06,
+      "loss": 0.0002,
+      "step": 2326
+    },
+    {
+      "epoch": 2.9766399999999997,
+      "grad_norm": 0.00015403024735860527,
+      "learning_rate": 4.006864546293941e-06,
+      "loss": 0.0002,
+      "step": 2327
+    },
+    {
+      "epoch": 2.97792,
+      "grad_norm": 0.0002792374580167234,
+      "learning_rate": 4.00619529142628e-06,
+      "loss": 0.0002,
+      "step": 2328
+    },
+    {
+      "epoch": 2.9792,
+      "grad_norm": 0.0011096937814727426,
+      "learning_rate": 4.005560349652384e-06,
+      "loss": 0.0002,
+      "step": 2329
+    },
+    {
+      "epoch": 2.98048,
+      "grad_norm": 0.00017472790204919875,
+      "learning_rate": 4.00495972218305e-06,
+      "loss": 0.0002,
+      "step": 2330
+    },
+    {
+      "epoch": 2.98176,
+      "grad_norm": 0.0001532040478195995,
+      "learning_rate": 4.004393410163635e-06,
+      "loss": 0.0002,
+      "step": 2331
+    },
+    {
+      "epoch": 2.98304,
+      "grad_norm": 0.000590935698710382,
+      "learning_rate": 4.003861414674068e-06,
+      "loss": 0.0002,
+      "step": 2332
+    },
+    {
+      "epoch": 2.98432,
+      "grad_norm": 0.0006752461194992065,
+      "learning_rate": 4.003363736728825e-06,
+      "loss": 0.0002,
+      "step": 2333
+    },
+    {
+      "epoch": 2.9856,
+      "grad_norm": 0.00026696111308410764,
+      "learning_rate": 4.002900377276953e-06,
+      "loss": 0.0002,
+      "step": 2334
+    },
+    {
+      "epoch": 2.98688,
+      "grad_norm": 0.00044541782699525356,
+      "learning_rate": 4.002471337202048e-06,
+      "loss": 0.0002,
+      "step": 2335
+    },
+    {
+      "epoch": 2.98816,
+      "grad_norm": 0.0009002209990285337,
+      "learning_rate": 4.002076617322264e-06,
+      "loss": 0.0002,
+      "step": 2336
+    },
+    {
+      "epoch": 2.98944,
+      "grad_norm": 0.00013183237751945853,
+      "learning_rate": 4.001716218390306e-06,
+      "loss": 0.0002,
+      "step": 2337
+    },
+    {
+      "epoch": 2.99072,
+      "grad_norm": 0.00016939207853283733,
+      "learning_rate": 4.001390141093436e-06,
+      "loss": 0.0002,
+      "step": 2338
+    },
+    {
+      "epoch": 2.992,
+      "grad_norm": 0.0005103266448713839,
+      "learning_rate": 4.0010983860534634e-06,
+      "loss": 0.0002,
+      "step": 2339
+    },
+    {
+      "epoch": 2.99328,
+      "grad_norm": 0.0006204844103194773,
+      "learning_rate": 4.000840953826744e-06,
+      "loss": 0.0002,
+      "step": 2340
+    },
+    {
+      "epoch": 2.99328,
+      "eval_loss": 1.179864525794983,
+      "eval_runtime": 43.3601,
+      "eval_samples_per_second": 11.577,
+      "eval_steps_per_second": 1.453,
+      "step": 2340
+    },
+    {
+      "epoch": 2.99456,
+      "grad_norm": 0.00016618035442661494,
+      "learning_rate": 4.000617844904188e-06,
+      "loss": 0.0002,
+      "step": 2341
+    },
+    {
+      "epoch": 2.99584,
+      "grad_norm": 0.0002050574403256178,
+      "learning_rate": 4.0004290597112505e-06,
+      "loss": 0.0002,
+      "step": 2342
+    },
+    {
+      "epoch": 2.99712,
+      "grad_norm": 0.0006428991910070181,
+      "learning_rate": 4.0002745986079325e-06,
+      "loss": 0.0002,
+      "step": 2343
+    },
+    {
+      "epoch": 2.9984,
+      "grad_norm": 0.0003186491085216403,
+      "learning_rate": 4.0001544618887825e-06,
+      "loss": 0.0002,
+      "step": 2344
+    },
+    {
+      "epoch": 2.99968,
+      "grad_norm": 0.0002871889155358076,
+      "learning_rate": 4.000068649782895e-06,
+      "loss": 0.0002,
+      "step": 2345
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.0007445301744155586,
+      "learning_rate": 4.000017162453906e-06,
+      "loss": 0.0002,
+      "step": 2346
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2346,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0606903825807704e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}