diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5017 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.5781153767193594,
+  "eval_steps": 88,
+  "global_step": 704,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.000821186614658181,
+      "grad_norm": 0.480191707611084,
+      "learning_rate": 2e-05,
+      "loss": 1.332,
+      "step": 1
+    },
+    {
+      "epoch": 0.001642373229316362,
+      "grad_norm": 0.43048733472824097,
+      "learning_rate": 4e-05,
+      "loss": 1.2784,
+      "step": 2
+    },
+    {
+      "epoch": 0.002463559843974543,
+      "grad_norm": 0.33739173412323,
+      "learning_rate": 6e-05,
+      "loss": 1.3286,
+      "step": 3
+    },
+    {
+      "epoch": 0.003284746458632724,
+      "grad_norm": 0.432579904794693,
+      "learning_rate": 8e-05,
+      "loss": 1.3079,
+      "step": 4
+    },
+    {
+      "epoch": 0.0041059330732909054,
+      "grad_norm": 0.3490436375141144,
+      "learning_rate": 0.0001,
+      "loss": 1.2182,
+      "step": 5
+    },
+    {
+      "epoch": 0.004927119687949086,
+      "grad_norm": 0.20206260681152344,
+      "learning_rate": 9.997257268239166e-05,
+      "loss": 1.2828,
+      "step": 6
+    },
+    {
+      "epoch": 0.005748306302607267,
+      "grad_norm": 0.15355628728866577,
+      "learning_rate": 9.994514536478333e-05,
+      "loss": 1.1616,
+      "step": 7
+    },
+    {
+      "epoch": 0.006569492917265448,
+      "grad_norm": 0.16756780445575714,
+      "learning_rate": 9.9917718047175e-05,
+      "loss": 1.0935,
+      "step": 8
+    },
+    {
+      "epoch": 0.00739067953192363,
+      "grad_norm": 0.17429664731025696,
+      "learning_rate": 9.989029072956665e-05,
+      "loss": 0.9694,
+      "step": 9
+    },
+    {
+      "epoch": 0.008211866146581811,
+      "grad_norm": 0.22355175018310547,
+      "learning_rate": 9.986286341195832e-05,
+      "loss": 0.9407,
+      "step": 10
+    },
+    {
+      "epoch": 0.009033052761239991,
+      "grad_norm": 0.33407703042030334,
+      "learning_rate": 9.983543609434997e-05,
+      "loss": 0.7717,
+      "step": 11
+    },
+    {
+      "epoch": 0.009854239375898173,
+      "grad_norm": 0.47473278641700745,
+      "learning_rate": 9.980800877674164e-05,
+      "loss": 0.7763,
+      "step": 12
+    },
+    {
+      "epoch": 0.010675425990556354,
+      "grad_norm": 0.2812059819698334,
+      "learning_rate": 9.978058145913331e-05,
+      "loss": 0.6258,
+      "step": 13
+    },
+    {
+      "epoch": 0.011496612605214535,
+      "grad_norm": 0.23547925055027008,
+      "learning_rate": 9.975315414152496e-05,
+      "loss": 0.5968,
+      "step": 14
+    },
+    {
+      "epoch": 0.012317799219872716,
+      "grad_norm": 0.18453630805015564,
+      "learning_rate": 9.972572682391662e-05,
+      "loss": 0.5368,
+      "step": 15
+    },
+    {
+      "epoch": 0.013138985834530896,
+      "grad_norm": 0.16103577613830566,
+      "learning_rate": 9.969829950630828e-05,
+      "loss": 0.4909,
+      "step": 16
+    },
+    {
+      "epoch": 0.013960172449189078,
+      "grad_norm": 0.18651455640792847,
+      "learning_rate": 9.967087218869995e-05,
+      "loss": 0.5135,
+      "step": 17
+    },
+    {
+      "epoch": 0.01478135906384726,
+      "grad_norm": 0.11300642043352127,
+      "learning_rate": 9.96434448710916e-05,
+      "loss": 0.5069,
+      "step": 18
+    },
+    {
+      "epoch": 0.01560254567850544,
+      "grad_norm": 0.10415703803300858,
+      "learning_rate": 9.961601755348327e-05,
+      "loss": 0.4851,
+      "step": 19
+    },
+    {
+      "epoch": 0.016423732293163622,
+      "grad_norm": 0.11693017929792404,
+      "learning_rate": 9.958859023587493e-05,
+      "loss": 0.4625,
+      "step": 20
+    },
+    {
+      "epoch": 0.017244918907821802,
+      "grad_norm": 0.10035043954849243,
+      "learning_rate": 9.95611629182666e-05,
+      "loss": 0.4822,
+      "step": 21
+    },
+    {
+      "epoch": 0.018066105522479982,
+      "grad_norm": 0.10483390837907791,
+      "learning_rate": 9.953373560065826e-05,
+      "loss": 0.4342,
+      "step": 22
+    },
+    {
+      "epoch": 0.018887292137138165,
+      "grad_norm": 2.8405802249908447,
+      "learning_rate": 9.950630828304992e-05,
+      "loss": 0.4664,
+      "step": 23
+    },
+    {
+      "epoch": 0.019708478751796345,
+      "grad_norm": 0.13821998238563538,
+      "learning_rate": 9.947888096544159e-05,
+      "loss": 0.4468,
+      "step": 24
+    },
+    {
+      "epoch": 0.020529665366454525,
+      "grad_norm": 0.1991378366947174,
+      "learning_rate": 9.945145364783325e-05,
+      "loss": 0.4605,
+      "step": 25
+    },
+    {
+      "epoch": 0.02135085198111271,
+      "grad_norm": 0.07619134336709976,
+      "learning_rate": 9.942402633022491e-05,
+      "loss": 0.4597,
+      "step": 26
+    },
+    {
+      "epoch": 0.02217203859577089,
+      "grad_norm": 0.13373583555221558,
+      "learning_rate": 9.939659901261658e-05,
+      "loss": 0.4626,
+      "step": 27
+    },
+    {
+      "epoch": 0.02299322521042907,
+      "grad_norm": 0.09962721168994904,
+      "learning_rate": 9.936917169500823e-05,
+      "loss": 0.4638,
+      "step": 28
+    },
+    {
+      "epoch": 0.023814411825087253,
+      "grad_norm": 0.09395964443683624,
+      "learning_rate": 9.93417443773999e-05,
+      "loss": 0.4569,
+      "step": 29
+    },
+    {
+      "epoch": 0.024635598439745433,
+      "grad_norm": 0.09109952300786972,
+      "learning_rate": 9.931431705979157e-05,
+      "loss": 0.4439,
+      "step": 30
+    },
+    {
+      "epoch": 0.025456785054403613,
+      "grad_norm": 0.10370515286922455,
+      "learning_rate": 9.928688974218322e-05,
+      "loss": 0.4425,
+      "step": 31
+    },
+    {
+      "epoch": 0.026277971669061793,
+      "grad_norm": 0.2153477966785431,
+      "learning_rate": 9.925946242457488e-05,
+      "loss": 0.4503,
+      "step": 32
+    },
+    {
+      "epoch": 0.027099158283719976,
+      "grad_norm": 0.08772841095924377,
+      "learning_rate": 9.923203510696654e-05,
+      "loss": 0.419,
+      "step": 33
+    },
+    {
+      "epoch": 0.027920344898378156,
+      "grad_norm": 0.10951374471187592,
+      "learning_rate": 9.920460778935821e-05,
+      "loss": 0.4353,
+      "step": 34
+    },
+    {
+      "epoch": 0.028741531513036336,
+      "grad_norm": 0.09190870076417923,
+      "learning_rate": 9.917718047174987e-05,
+      "loss": 0.5196,
+      "step": 35
+    },
+    {
+      "epoch": 0.02956271812769452,
+      "grad_norm": 0.07667124271392822,
+      "learning_rate": 9.914975315414153e-05,
+      "loss": 0.4358,
+      "step": 36
+    },
+    {
+      "epoch": 0.0303839047423527,
+      "grad_norm": 0.1514267474412918,
+      "learning_rate": 9.912232583653319e-05,
+      "loss": 0.411,
+      "step": 37
+    },
+    {
+      "epoch": 0.03120509135701088,
+      "grad_norm": 0.09086549282073975,
+      "learning_rate": 9.909489851892486e-05,
+      "loss": 0.4003,
+      "step": 38
+    },
+    {
+      "epoch": 0.032026277971669063,
+      "grad_norm": 0.2616782486438751,
+      "learning_rate": 9.906747120131652e-05,
+      "loss": 0.4842,
+      "step": 39
+    },
+    {
+      "epoch": 0.032847464586327244,
+      "grad_norm": 0.11908283084630966,
+      "learning_rate": 9.904004388370818e-05,
+      "loss": 0.4143,
+      "step": 40
+    },
+    {
+      "epoch": 0.033668651200985424,
+      "grad_norm": 0.07770542800426483,
+      "learning_rate": 9.901261656609983e-05,
+      "loss": 0.3873,
+      "step": 41
+    },
+    {
+      "epoch": 0.034489837815643604,
+      "grad_norm": 0.08934606611728668,
+      "learning_rate": 9.89851892484915e-05,
+      "loss": 0.4235,
+      "step": 42
+    },
+    {
+      "epoch": 0.035311024430301784,
+      "grad_norm": 0.09303563088178635,
+      "learning_rate": 9.895776193088317e-05,
+      "loss": 0.4103,
+      "step": 43
+    },
+    {
+      "epoch": 0.036132211044959964,
+      "grad_norm": 0.08622181415557861,
+      "learning_rate": 9.893033461327482e-05,
+      "loss": 0.448,
+      "step": 44
+    },
+    {
+      "epoch": 0.03695339765961815,
+      "grad_norm": 0.08822862058877945,
+      "learning_rate": 9.890290729566649e-05,
+      "loss": 0.3855,
+      "step": 45
+    },
+    {
+      "epoch": 0.03777458427427633,
+      "grad_norm": 0.08557698875665665,
+      "learning_rate": 9.887547997805814e-05,
+      "loss": 0.3945,
+      "step": 46
+    },
+    {
+      "epoch": 0.03859577088893451,
+      "grad_norm": 0.07540106773376465,
+      "learning_rate": 9.884805266044981e-05,
+      "loss": 0.4192,
+      "step": 47
+    },
+    {
+      "epoch": 0.03941695750359269,
+      "grad_norm": 0.1023702397942543,
+      "learning_rate": 9.882062534284148e-05,
+      "loss": 0.4126,
+      "step": 48
+    },
+    {
+      "epoch": 0.04023814411825087,
+      "grad_norm": 0.07779772579669952,
+      "learning_rate": 9.879319802523313e-05,
+      "loss": 0.4244,
+      "step": 49
+    },
+    {
+      "epoch": 0.04105933073290905,
+      "grad_norm": 0.08826564252376556,
+      "learning_rate": 9.876577070762479e-05,
+      "loss": 0.415,
+      "step": 50
+    },
+    {
+      "epoch": 0.04188051734756724,
+      "grad_norm": 0.08254576474428177,
+      "learning_rate": 9.873834339001646e-05,
+      "loss": 0.4346,
+      "step": 51
+    },
+    {
+      "epoch": 0.04270170396222542,
+      "grad_norm": 0.08287151902914047,
+      "learning_rate": 9.871091607240812e-05,
+      "loss": 0.4142,
+      "step": 52
+    },
+    {
+      "epoch": 0.0435228905768836,
+      "grad_norm": 0.08196476101875305,
+      "learning_rate": 9.868348875479978e-05,
+      "loss": 0.3822,
+      "step": 53
+    },
+    {
+      "epoch": 0.04434407719154178,
+      "grad_norm": 0.08654092252254486,
+      "learning_rate": 9.865606143719145e-05,
+      "loss": 0.3937,
+      "step": 54
+    },
+    {
+      "epoch": 0.04516526380619996,
+      "grad_norm": 0.1102684736251831,
+      "learning_rate": 9.86286341195831e-05,
+      "loss": 0.4508,
+      "step": 55
+    },
+    {
+      "epoch": 0.04598645042085814,
+      "grad_norm": 0.08240954577922821,
+      "learning_rate": 9.860120680197478e-05,
+      "loss": 0.4069,
+      "step": 56
+    },
+    {
+      "epoch": 0.04680763703551632,
+      "grad_norm": 0.08542217314243317,
+      "learning_rate": 9.857377948436644e-05,
+      "loss": 0.5002,
+      "step": 57
+    },
+    {
+      "epoch": 0.047628823650174505,
+      "grad_norm": 0.08390172570943832,
+      "learning_rate": 9.854635216675809e-05,
+      "loss": 0.3851,
+      "step": 58
+    },
+    {
+      "epoch": 0.048450010264832685,
+      "grad_norm": 0.10694168508052826,
+      "learning_rate": 9.851892484914976e-05,
+      "loss": 0.4026,
+      "step": 59
+    },
+    {
+      "epoch": 0.049271196879490865,
+      "grad_norm": 0.0852806493639946,
+      "learning_rate": 9.849149753154143e-05,
+      "loss": 0.424,
+      "step": 60
+    },
+    {
+      "epoch": 0.050092383494149045,
+      "grad_norm": 0.15425831079483032,
+      "learning_rate": 9.846407021393308e-05,
+      "loss": 0.4307,
+      "step": 61
+    },
+    {
+      "epoch": 0.050913570108807225,
+      "grad_norm": 0.08546218276023865,
+      "learning_rate": 9.843664289632475e-05,
+      "loss": 0.4386,
+      "step": 62
+    },
+    {
+      "epoch": 0.051734756723465405,
+      "grad_norm": 0.08588795363903046,
+      "learning_rate": 9.84092155787164e-05,
+      "loss": 0.4433,
+      "step": 63
+    },
+    {
+      "epoch": 0.052555943338123585,
+      "grad_norm": 0.08901514858007431,
+      "learning_rate": 9.838178826110807e-05,
+      "loss": 0.4263,
+      "step": 64
+    },
+    {
+      "epoch": 0.05337712995278177,
+      "grad_norm": 0.17668095231056213,
+      "learning_rate": 9.835436094349974e-05,
+      "loss": 0.407,
+      "step": 65
+    },
+    {
+      "epoch": 0.05419831656743995,
+      "grad_norm": 0.081763856112957,
+      "learning_rate": 9.83269336258914e-05,
+      "loss": 0.374,
+      "step": 66
+    },
+    {
+      "epoch": 0.05501950318209813,
+      "grad_norm": 0.09063572436571121,
+      "learning_rate": 9.829950630828305e-05,
+      "loss": 0.3839,
+      "step": 67
+    },
+    {
+      "epoch": 0.05584068979675631,
+      "grad_norm": 0.08264392614364624,
+      "learning_rate": 9.827207899067472e-05,
+      "loss": 0.4676,
+      "step": 68
+    },
+    {
+      "epoch": 0.05666187641141449,
+      "grad_norm": 0.0827123150229454,
+      "learning_rate": 9.824465167306638e-05,
+      "loss": 0.3801,
+      "step": 69
+    },
+    {
+      "epoch": 0.05748306302607267,
+      "grad_norm": 0.07972189038991928,
+      "learning_rate": 9.821722435545804e-05,
+      "loss": 0.3758,
+      "step": 70
+    },
+    {
+      "epoch": 0.05830424964073085,
+      "grad_norm": 0.08738942444324493,
+      "learning_rate": 9.81897970378497e-05,
+      "loss": 0.3634,
+      "step": 71
+    },
+    {
+      "epoch": 0.05912543625538904,
+      "grad_norm": 0.08442792296409607,
+      "learning_rate": 9.816236972024136e-05,
+      "loss": 0.3803,
+      "step": 72
+    },
+    {
+      "epoch": 0.05994662287004722,
+      "grad_norm": 0.2751137316226959,
+      "learning_rate": 9.813494240263303e-05,
+      "loss": 0.3869,
+      "step": 73
+    },
+    {
+      "epoch": 0.0607678094847054,
+      "grad_norm": 0.19429296255111694,
+      "learning_rate": 9.81075150850247e-05,
+      "loss": 0.3992,
+      "step": 74
+    },
+    {
+      "epoch": 0.06158899609936358,
+      "grad_norm": 0.08392605930566788,
+      "learning_rate": 9.808008776741635e-05,
+      "loss": 0.361,
+      "step": 75
+    },
+    {
+      "epoch": 0.06241018271402176,
+      "grad_norm": 2.01043963432312,
+      "learning_rate": 9.8052660449808e-05,
+      "loss": 0.3787,
+      "step": 76
+    },
+    {
+      "epoch": 0.06323136932867994,
+      "grad_norm": 0.0808538943529129,
+      "learning_rate": 9.802523313219967e-05,
+      "loss": 0.4013,
+      "step": 77
+    },
+    {
+      "epoch": 0.06405255594333813,
+      "grad_norm": 0.07846518605947495,
+      "learning_rate": 9.799780581459134e-05,
+      "loss": 0.3808,
+      "step": 78
+    },
+    {
+      "epoch": 0.0648737425579963,
+      "grad_norm": 0.08152970671653748,
+      "learning_rate": 9.7970378496983e-05,
+      "loss": 0.4076,
+      "step": 79
+    },
+    {
+      "epoch": 0.06569492917265449,
+      "grad_norm": 0.07745791226625443,
+      "learning_rate": 9.794295117937466e-05,
+      "loss": 0.3981,
+      "step": 80
+    },
+    {
+      "epoch": 0.06651611578731266,
+      "grad_norm": 0.0842173770070076,
+      "learning_rate": 9.791552386176632e-05,
+      "loss": 0.4469,
+      "step": 81
+    },
+    {
+      "epoch": 0.06733730240197085,
+      "grad_norm": 0.07920663058757782,
+      "learning_rate": 9.788809654415799e-05,
+      "loss": 0.3988,
+      "step": 82
+    },
+    {
+      "epoch": 0.06815848901662903,
+      "grad_norm": 0.07899456471204758,
+      "learning_rate": 9.786066922654965e-05,
+      "loss": 0.4418,
+      "step": 83
+    },
+    {
+      "epoch": 0.06897967563128721,
+      "grad_norm": 0.07229727506637573,
+      "learning_rate": 9.783324190894131e-05,
+      "loss": 0.4016,
+      "step": 84
+    },
+    {
+      "epoch": 0.0698008622459454,
+      "grad_norm": 0.08062436431646347,
+      "learning_rate": 9.780581459133296e-05,
+      "loss": 0.3762,
+      "step": 85
+    },
+    {
+      "epoch": 0.07062204886060357,
+      "grad_norm": 0.08440960198640823,
+      "learning_rate": 9.777838727372464e-05,
+      "loss": 0.4362,
+      "step": 86
+    },
+    {
+      "epoch": 0.07144323547526175,
+      "grad_norm": 0.07544733583927155,
+      "learning_rate": 9.77509599561163e-05,
+      "loss": 0.4135,
+      "step": 87
+    },
+    {
+      "epoch": 0.07226442208991993,
+      "grad_norm": 0.1760656237602234,
+      "learning_rate": 9.772353263850797e-05,
+      "loss": 0.3887,
+      "step": 88
+    },
+    {
+      "epoch": 0.07226442208991993,
+      "eval_runtime": 494.3052,
+      "eval_samples_per_second": 0.399,
+      "eval_steps_per_second": 0.2,
+      "step": 88
+    },
+    {
+      "epoch": 0.07308560870457811,
+      "grad_norm": 0.08049104362726212,
+      "learning_rate": 9.769610532089962e-05,
+      "loss": 0.4272,
+      "step": 89
+    },
+    {
+      "epoch": 0.0739067953192363,
+      "grad_norm": 0.07617965340614319,
+      "learning_rate": 9.766867800329129e-05,
+      "loss": 0.3873,
+      "step": 90
+    },
+    {
+      "epoch": 0.07472798193389447,
+      "grad_norm": 0.07975462824106216,
+      "learning_rate": 9.764125068568296e-05,
+      "loss": 0.3903,
+      "step": 91
+    },
+    {
+      "epoch": 0.07554916854855266,
+      "grad_norm": 0.08299189805984497,
+      "learning_rate": 9.761382336807461e-05,
+      "loss": 0.404,
+      "step": 92
+    },
+    {
+      "epoch": 0.07637035516321083,
+      "grad_norm": 0.08578819036483765,
+      "learning_rate": 9.758639605046626e-05,
+      "loss": 0.3983,
+      "step": 93
+    },
+    {
+      "epoch": 0.07719154177786902,
+      "grad_norm": 0.10872071981430054,
+      "learning_rate": 9.755896873285793e-05,
+      "loss": 0.3553,
+      "step": 94
+    },
+    {
+      "epoch": 0.07801272839252721,
+      "grad_norm": 0.08121436834335327,
+      "learning_rate": 9.75315414152496e-05,
+      "loss": 0.408,
+      "step": 95
+    },
+    {
+      "epoch": 0.07883391500718538,
+      "grad_norm": 0.09175996482372284,
+      "learning_rate": 9.750411409764125e-05,
+      "loss": 0.4008,
+      "step": 96
+    },
+    {
+      "epoch": 0.07965510162184357,
+      "grad_norm": 0.08122789114713669,
+      "learning_rate": 9.747668678003292e-05,
+      "loss": 0.4045,
+      "step": 97
+    },
+    {
+      "epoch": 0.08047628823650174,
+      "grad_norm": 0.08402436226606369,
+      "learning_rate": 9.744925946242458e-05,
+      "loss": 0.3814,
+      "step": 98
+    },
+    {
+      "epoch": 0.08129747485115993,
+      "grad_norm": 0.08454861491918564,
+      "learning_rate": 9.742183214481624e-05,
+      "loss": 0.3904,
+      "step": 99
+    },
+    {
+      "epoch": 0.0821186614658181,
+      "grad_norm": 0.08125888556241989,
+      "learning_rate": 9.739440482720791e-05,
+      "loss": 0.3681,
+      "step": 100
+    },
+    {
+      "epoch": 0.08293984808047629,
+      "grad_norm": 0.08544078469276428,
+      "learning_rate": 9.736697750959957e-05,
+      "loss": 0.367,
+      "step": 101
+    },
+    {
+      "epoch": 0.08376103469513448,
+      "grad_norm": 0.07539089769124985,
+      "learning_rate": 9.733955019199122e-05,
+      "loss": 0.3821,
+      "step": 102
+    },
+    {
+      "epoch": 0.08458222130979265,
+      "grad_norm": 0.07529085874557495,
+      "learning_rate": 9.731212287438289e-05,
+      "loss": 0.4169,
+      "step": 103
+    },
+    {
+      "epoch": 0.08540340792445084,
+      "grad_norm": 0.07588034868240356,
+      "learning_rate": 9.728469555677456e-05,
+      "loss": 0.4211,
+      "step": 104
+    },
+    {
+      "epoch": 0.08622459453910901,
+      "grad_norm": 0.08019097149372101,
+      "learning_rate": 9.725726823916621e-05,
+      "loss": 0.4033,
+      "step": 105
+    },
+    {
+      "epoch": 0.0870457811537672,
+      "grad_norm": 0.07878712564706802,
+      "learning_rate": 9.722984092155788e-05,
+      "loss": 0.3842,
+      "step": 106
+    },
+    {
+      "epoch": 0.08786696776842537,
+      "grad_norm": 0.08083963394165039,
+      "learning_rate": 9.720241360394953e-05,
+      "loss": 0.3558,
+      "step": 107
+    },
+    {
+      "epoch": 0.08868815438308356,
+      "grad_norm": 0.08340411633253098,
+      "learning_rate": 9.71749862863412e-05,
+      "loss": 0.4037,
+      "step": 108
+    },
+    {
+      "epoch": 0.08950934099774174,
+      "grad_norm": 0.09317754209041595,
+      "learning_rate": 9.714755896873287e-05,
+      "loss": 0.3874,
+      "step": 109
+    },
+    {
+      "epoch": 0.09033052761239992,
+      "grad_norm": 0.07143125683069229,
+      "learning_rate": 9.712013165112452e-05,
+      "loss": 0.3441,
+      "step": 110
+    },
+    {
+      "epoch": 0.0911517142270581,
+      "grad_norm": 0.0909111350774765,
+      "learning_rate": 9.709270433351618e-05,
+      "loss": 0.3552,
+      "step": 111
+    },
+    {
+      "epoch": 0.09197290084171628,
+      "grad_norm": 0.08041603863239288,
+      "learning_rate": 9.706527701590785e-05,
+      "loss": 0.394,
+      "step": 112
+    },
+    {
+      "epoch": 0.09279408745637446,
+      "grad_norm": 0.08147992193698883,
+      "learning_rate": 9.703784969829951e-05,
+      "loss": 0.3916,
+      "step": 113
+    },
+    {
+      "epoch": 0.09361527407103264,
+      "grad_norm": 0.08365318924188614,
+      "learning_rate": 9.701042238069117e-05,
+      "loss": 0.4075,
+      "step": 114
+    },
+    {
+      "epoch": 0.09443646068569082,
+      "grad_norm": 0.083246149122715,
+      "learning_rate": 9.698299506308284e-05,
+      "loss": 0.3566,
+      "step": 115
+    },
+    {
+      "epoch": 0.09525764730034901,
+      "grad_norm": 0.0942649245262146,
+      "learning_rate": 9.695556774547449e-05,
+      "loss": 0.3516,
+      "step": 116
+    },
+    {
+      "epoch": 0.09607883391500718,
+      "grad_norm": 0.08110091835260391,
+      "learning_rate": 9.692814042786616e-05,
+      "loss": 0.3902,
+      "step": 117
+    },
+    {
+      "epoch": 0.09690002052966537,
+      "grad_norm": 0.43124014139175415,
+      "learning_rate": 9.690071311025783e-05,
+      "loss": 0.3956,
+      "step": 118
+    },
+    {
+      "epoch": 0.09772120714432354,
+      "grad_norm": 0.08204706013202667,
+      "learning_rate": 9.687328579264948e-05,
+      "loss": 0.3865,
+      "step": 119
+    },
+    {
+      "epoch": 0.09854239375898173,
+      "grad_norm": 0.08262762427330017,
+      "learning_rate": 9.684585847504113e-05,
+      "loss": 0.3777,
+      "step": 120
+    },
+    {
+      "epoch": 0.0993635803736399,
+      "grad_norm": 0.08233962953090668,
+      "learning_rate": 9.681843115743282e-05,
+      "loss": 0.3976,
+      "step": 121
+    },
+    {
+      "epoch": 0.10018476698829809,
+      "grad_norm": 0.7428120970726013,
+      "learning_rate": 9.679100383982447e-05,
+      "loss": 0.377,
+      "step": 122
+    },
+    {
+      "epoch": 0.10100595360295628,
+      "grad_norm": 0.07909400761127472,
+      "learning_rate": 9.676357652221614e-05,
+      "loss": 0.3654,
+      "step": 123
+    },
+    {
+      "epoch": 0.10182714021761445,
+      "grad_norm": 0.08714035898447037,
+      "learning_rate": 9.673614920460779e-05,
+      "loss": 0.4181,
+      "step": 124
+    },
+    {
+      "epoch": 0.10264832683227264,
+      "grad_norm": 0.08017311245203018,
+      "learning_rate": 9.670872188699946e-05,
+      "loss": 0.4052,
+      "step": 125
+    },
+    {
+      "epoch": 0.10346951344693081,
+      "grad_norm": 0.13821078836917877,
+      "learning_rate": 9.668129456939113e-05,
+      "loss": 0.3733,
+      "step": 126
+    },
+    {
+      "epoch": 0.104290700061589,
+      "grad_norm": 0.1609969437122345,
+      "learning_rate": 9.665386725178278e-05,
+      "loss": 0.3686,
+      "step": 127
+    },
+    {
+      "epoch": 0.10511188667624717,
+      "grad_norm": 0.1280309557914734,
+      "learning_rate": 9.662643993417444e-05,
+      "loss": 0.4036,
+      "step": 128
+    },
+    {
+      "epoch": 0.10593307329090536,
+      "grad_norm": 0.08747898787260056,
+      "learning_rate": 9.65990126165661e-05,
+      "loss": 0.3587,
+      "step": 129
+    },
+    {
+      "epoch": 0.10675425990556354,
+      "grad_norm": 0.1252209097146988,
+      "learning_rate": 9.657158529895777e-05,
+      "loss": 0.3626,
+      "step": 130
+    },
+    {
+      "epoch": 0.10757544652022172,
+      "grad_norm": 0.09374388307332993,
+      "learning_rate": 9.654415798134943e-05,
+      "loss": 0.3824,
+      "step": 131
+    },
+    {
+      "epoch": 0.1083966331348799,
+      "grad_norm": 0.08730709552764893,
+      "learning_rate": 9.65167306637411e-05,
+      "loss": 0.3827,
+      "step": 132
+    },
+    {
+      "epoch": 0.10921781974953808,
+      "grad_norm": 0.07719024270772934,
+      "learning_rate": 9.648930334613275e-05,
+      "loss": 0.3762,
+      "step": 133
+    },
+    {
+      "epoch": 0.11003900636419627,
+      "grad_norm": 0.08843278139829636,
+      "learning_rate": 9.646187602852442e-05,
+      "loss": 0.3843,
+      "step": 134
+    },
+    {
+      "epoch": 0.11086019297885444,
+      "grad_norm": 0.08568207919597626,
+      "learning_rate": 9.643444871091608e-05,
+      "loss": 0.3594,
+      "step": 135
+    },
+    {
+      "epoch": 0.11168137959351263,
+      "grad_norm": 0.08556952327489853,
+      "learning_rate": 9.640702139330774e-05,
+      "loss": 0.3886,
+      "step": 136
+    },
+    {
+      "epoch": 0.11250256620817081,
+      "grad_norm": 0.09016801416873932,
+      "learning_rate": 9.63795940756994e-05,
+      "loss": 0.4023,
+      "step": 137
+    },
+    {
+      "epoch": 0.11332375282282899,
+      "grad_norm": 0.08133590966463089,
+      "learning_rate": 9.635216675809106e-05,
+      "loss": 0.3634,
+      "step": 138
+    },
+    {
+      "epoch": 0.11414493943748717,
+      "grad_norm": 0.2211730182170868,
+      "learning_rate": 9.632473944048273e-05,
+      "loss": 0.3518,
+      "step": 139
+    },
+    {
+      "epoch": 0.11496612605214535,
+      "grad_norm": 0.08816584199666977,
+      "learning_rate": 9.629731212287438e-05,
+      "loss": 0.3727,
+      "step": 140
+    },
+    {
+      "epoch": 0.11578731266680353,
+      "grad_norm": 0.11618969589471817,
+      "learning_rate": 9.626988480526605e-05,
+      "loss": 0.4247,
+      "step": 141
+    },
+    {
+      "epoch": 0.1166084992814617,
+      "grad_norm": 0.09243030101060867,
+      "learning_rate": 9.62424574876577e-05,
+      "loss": 0.396,
+      "step": 142
+    },
+    {
+      "epoch": 0.11742968589611989,
+      "grad_norm": 0.08566376566886902,
+      "learning_rate": 9.621503017004937e-05,
+      "loss": 0.3852,
+      "step": 143
+    },
+    {
+      "epoch": 0.11825087251077808,
+      "grad_norm": 0.08220973610877991,
+      "learning_rate": 9.618760285244104e-05,
+      "loss": 0.3961,
+      "step": 144
+    },
+    {
+      "epoch": 0.11907205912543625,
+      "grad_norm": 0.08240345865488052,
+      "learning_rate": 9.61601755348327e-05,
+      "loss": 0.3518,
+      "step": 145
+    },
+    {
+      "epoch": 0.11989324574009444,
+      "grad_norm": 0.08472532778978348,
+      "learning_rate": 9.613274821722435e-05,
+      "loss": 0.3586,
+      "step": 146
+    },
+    {
+      "epoch": 0.12071443235475261,
+      "grad_norm": 0.08407485485076904,
+      "learning_rate": 9.610532089961602e-05,
+      "loss": 0.3797,
+      "step": 147
+    },
+    {
+      "epoch": 0.1215356189694108,
+      "grad_norm": 0.09284385293722153,
+      "learning_rate": 9.607789358200769e-05,
+      "loss": 0.3499,
+      "step": 148
+    },
+    {
+      "epoch": 0.12235680558406897,
+      "grad_norm": 0.08499818295240402,
+      "learning_rate": 9.605046626439934e-05,
+      "loss": 0.3722,
+      "step": 149
+    },
+    {
+      "epoch": 0.12317799219872716,
+      "grad_norm": 0.080271415412426,
+      "learning_rate": 9.602303894679101e-05,
+      "loss": 0.3871,
+      "step": 150
+    },
+    {
+      "epoch": 0.12399917881338535,
+      "grad_norm": 0.07850060611963272,
+      "learning_rate": 9.599561162918266e-05,
+      "loss": 0.3679,
+      "step": 151
+    },
+    {
+      "epoch": 0.12482036542804352,
+      "grad_norm": 0.07685016840696335,
+      "learning_rate": 9.596818431157433e-05,
+      "loss": 0.4078,
+      "step": 152
+    },
+    {
+      "epoch": 0.1256415520427017,
+      "grad_norm": 0.09402357786893845,
+      "learning_rate": 9.5940756993966e-05,
+      "loss": 0.3996,
+      "step": 153
+    },
+    {
+      "epoch": 0.12646273865735988,
+      "grad_norm": 0.08445476740598679,
+      "learning_rate": 9.591332967635765e-05,
+      "loss": 0.3841,
+      "step": 154
+    },
+    {
+      "epoch": 0.12728392527201807,
+      "grad_norm": 0.08233911544084549,
+      "learning_rate": 9.588590235874932e-05,
+      "loss": 0.3529,
+      "step": 155
+    },
+    {
+      "epoch": 0.12810511188667625,
+      "grad_norm": 0.07896068692207336,
+      "learning_rate": 9.585847504114099e-05,
+      "loss": 0.3585,
+      "step": 156
+    },
+    {
+      "epoch": 0.12892629850133444,
+      "grad_norm": 0.0822276696562767,
+      "learning_rate": 9.583104772353264e-05,
+      "loss": 0.3962,
+      "step": 157
+    },
+    {
+      "epoch": 0.1297474851159926,
+      "grad_norm": 0.07977598905563354,
+      "learning_rate": 9.580362040592431e-05,
+      "loss": 0.454,
+      "step": 158
+    },
+    {
+      "epoch": 0.1305686717306508,
+      "grad_norm": 0.0857616737484932,
+      "learning_rate": 9.577619308831597e-05,
+      "loss": 0.3954,
+      "step": 159
+    },
+    {
+      "epoch": 0.13138985834530897,
+      "grad_norm": 0.0874355211853981,
+      "learning_rate": 9.574876577070763e-05,
+      "loss": 0.3591,
+      "step": 160
+    },
+    {
+      "epoch": 0.13221104495996716,
+      "grad_norm": 0.07877468317747116,
+      "learning_rate": 9.57213384530993e-05,
+      "loss": 0.348,
+      "step": 161
+    },
+    {
+      "epoch": 0.13303223157462532,
+      "grad_norm": 0.08618593961000443,
+      "learning_rate": 9.569391113549096e-05,
+      "loss": 0.3723,
+      "step": 162
+    },
+    {
+      "epoch": 0.1338534181892835,
+      "grad_norm": 0.08144336938858032,
+      "learning_rate": 9.566648381788261e-05,
+      "loss": 0.4146,
+      "step": 163
+    },
+    {
+      "epoch": 0.1346746048039417,
+      "grad_norm": 0.07322760671377182,
+      "learning_rate": 9.563905650027428e-05,
+      "loss": 0.3107,
+      "step": 164
+    },
+    {
+      "epoch": 0.13549579141859988,
+      "grad_norm": 0.08007095754146576,
+      "learning_rate": 9.561162918266595e-05,
+      "loss": 0.4128,
+      "step": 165
+    },
+    {
+      "epoch": 0.13631697803325807,
+      "grad_norm": 0.09636646509170532,
+      "learning_rate": 9.55842018650576e-05,
+      "loss": 0.4089,
+      "step": 166
+    },
+    {
+      "epoch": 0.13713816464791623,
+      "grad_norm": 0.08381053060293198,
+      "learning_rate": 9.555677454744927e-05,
+      "loss": 0.3624,
+      "step": 167
+    },
+    {
+      "epoch": 0.13795935126257441,
+      "grad_norm": 0.07476504147052765,
+      "learning_rate": 9.552934722984092e-05,
+      "loss": 0.3906,
+      "step": 168
+    },
+    {
+      "epoch": 0.1387805378772326,
+      "grad_norm": 0.0901239663362503,
+      "learning_rate": 9.550191991223259e-05,
+      "loss": 0.3378,
+      "step": 169
+    },
+    {
+      "epoch": 0.1396017244918908,
+      "grad_norm": 0.0813356265425682,
+      "learning_rate": 9.547449259462426e-05,
+      "loss": 0.3627,
+      "step": 170
+    },
+    {
+      "epoch": 0.14042291110654898,
+      "grad_norm": 0.14319093525409698,
+      "learning_rate": 9.544706527701591e-05,
+      "loss": 0.3512,
+      "step": 171
+    },
+    {
+      "epoch": 0.14124409772120713,
+      "grad_norm": 0.13329866528511047,
+      "learning_rate": 9.541963795940757e-05,
+      "loss": 0.3809,
+      "step": 172
+    },
+    {
+      "epoch": 0.14206528433586532,
+      "grad_norm": 0.0815596953034401,
+      "learning_rate": 9.539221064179923e-05,
+      "loss": 0.3444,
+      "step": 173
+    },
+    {
+      "epoch": 0.1428864709505235,
+      "grad_norm": 0.08646956831216812,
+      "learning_rate": 9.53647833241909e-05,
+      "loss": 0.3699,
+      "step": 174
+    },
+    {
+      "epoch": 0.1437076575651817,
+      "grad_norm": 0.09374339133501053,
+      "learning_rate": 9.533735600658256e-05,
+      "loss": 0.3569,
+      "step": 175
+    },
+    {
+      "epoch": 0.14452884417983985,
+      "grad_norm": 0.0834718644618988,
+      "learning_rate": 9.530992868897422e-05,
+      "loss": 0.3708,
+      "step": 176
+    },
+    {
+      "epoch": 0.14452884417983985,
+      "eval_runtime": 493.9539,
+      "eval_samples_per_second": 0.399,
+      "eval_steps_per_second": 0.2,
+      "step": 176
+    },
+    {
+      "epoch": 0.14535003079449804,
+      "grad_norm": 0.08648290485143661,
+      "learning_rate": 9.528250137136588e-05,
+      "loss": 0.379,
+      "step": 177
+    },
+    {
+      "epoch": 0.14617121740915623,
+      "grad_norm": 0.08577203750610352,
+      "learning_rate": 9.525507405375755e-05,
+      "loss": 0.4368,
+      "step": 178
+    },
+    {
+      "epoch": 0.14699240402381442,
+      "grad_norm": 0.1023576483130455,
+      "learning_rate": 9.522764673614921e-05,
+      "loss": 0.3553,
+      "step": 179
+    },
+    {
+      "epoch": 0.1478135906384726,
+      "grad_norm": 0.08062634617090225,
+      "learning_rate": 9.520021941854087e-05,
+      "loss": 0.3616,
+      "step": 180
+    },
+    {
+      "epoch": 0.14863477725313076,
+      "grad_norm": 0.07487751543521881,
+      "learning_rate": 9.517279210093252e-05,
+      "loss": 0.4413,
+      "step": 181
+    },
+    {
+      "epoch": 0.14945596386778895,
+      "grad_norm": 0.07332492619752884,
+      "learning_rate": 9.514536478332419e-05,
+      "loss": 0.3166,
+      "step": 182
+    },
+    {
+      "epoch": 0.15027715048244714,
+      "grad_norm": 0.08658608049154282,
+      "learning_rate": 9.511793746571586e-05,
+      "loss": 0.332,
+      "step": 183
+    },
+    {
+      "epoch": 0.15109833709710532,
+      "grad_norm": 0.10461894422769547,
+      "learning_rate": 9.509051014810751e-05,
+      "loss": 0.3653,
+      "step": 184
+    },
+    {
+      "epoch": 0.1519195237117635,
+      "grad_norm": 0.19619494676589966,
+      "learning_rate": 9.506308283049918e-05,
+      "loss": 0.4224,
+      "step": 185
+    },
+    {
+      "epoch": 0.15274071032642167,
+      "grad_norm": 0.08313202857971191,
+      "learning_rate": 9.503565551289084e-05,
+      "loss": 0.3512,
+      "step": 186
+    },
+    {
+      "epoch": 0.15356189694107986,
+      "grad_norm": 0.08686342090368271,
+      "learning_rate": 9.50082281952825e-05,
+      "loss": 0.3352,
+      "step": 187
+    },
+    {
+      "epoch": 0.15438308355573804,
+      "grad_norm": 0.09020522236824036,
+      "learning_rate": 9.498080087767417e-05,
+      "loss": 0.4138,
+      "step": 188
+    },
+    {
+      "epoch": 0.15520427017039623,
+      "grad_norm": 0.0799839124083519,
+      "learning_rate": 9.495337356006583e-05,
+      "loss": 0.3436,
+      "step": 189
+    },
+    {
+      "epoch": 0.15602545678505442,
+      "grad_norm": 0.08851379156112671,
+      "learning_rate": 9.49259462424575e-05,
+      "loss": 0.3933,
+      "step": 190
+    },
+    {
+      "epoch": 0.15684664339971258,
+      "grad_norm": 0.08521082252264023,
+      "learning_rate": 9.489851892484916e-05,
+      "loss": 0.3667,
+      "step": 191
+    },
+    {
+      "epoch": 0.15766783001437076,
+      "grad_norm": 0.09019312262535095,
+      "learning_rate": 9.487109160724082e-05,
+      "loss": 0.3556,
+      "step": 192
+    },
+    {
+      "epoch": 0.15848901662902895,
+      "grad_norm": 0.08162654936313629,
+      "learning_rate": 9.484366428963248e-05,
+      "loss": 0.3864,
+      "step": 193
+    },
+    {
+      "epoch": 0.15931020324368714,
+      "grad_norm": 0.08963490277528763,
+      "learning_rate": 9.481623697202414e-05,
+      "loss": 0.391,
+      "step": 194
+    },
+    {
+      "epoch": 0.1601313898583453,
+      "grad_norm": 0.10057719051837921,
+      "learning_rate": 9.47888096544158e-05,
+      "loss": 0.3599,
+      "step": 195
+    },
+    {
+      "epoch": 0.16095257647300348,
+      "grad_norm": 0.1694149523973465,
+      "learning_rate": 9.476138233680747e-05,
+      "loss": 0.3612,
+      "step": 196
+    },
+    {
+      "epoch": 0.16177376308766167,
+      "grad_norm": 0.08255323767662048,
+      "learning_rate": 9.473395501919913e-05,
+      "loss": 0.3441,
+      "step": 197
+    },
+    {
+      "epoch": 0.16259494970231986,
+      "grad_norm": 0.09228333085775375,
+      "learning_rate": 9.470652770159078e-05,
+      "loss": 0.3432,
+      "step": 198
+    },
+    {
+      "epoch": 0.16341613631697804,
+      "grad_norm": 0.11736617982387543,
+      "learning_rate": 9.467910038398245e-05,
+      "loss": 0.3408,
+      "step": 199
+    },
+    {
+      "epoch": 0.1642373229316362,
+      "grad_norm": 0.16636626422405243,
+      "learning_rate": 9.465167306637412e-05,
+      "loss": 0.3639,
+      "step": 200
+    },
+    {
+      "epoch": 0.1650585095462944,
+      "grad_norm": 0.0912085771560669,
+      "learning_rate": 9.462424574876577e-05,
+      "loss": 0.4055,
+      "step": 201
+    },
+    {
+      "epoch": 0.16587969616095258,
+      "grad_norm": 0.08452475070953369,
+      "learning_rate": 9.459681843115744e-05,
+      "loss": 0.3497,
+      "step": 202
+    },
+    {
+      "epoch": 0.16670088277561076,
+      "grad_norm": 0.09225429594516754,
+      "learning_rate": 9.45693911135491e-05,
+      "loss": 0.3765,
+      "step": 203
+    },
+    {
+      "epoch": 0.16752206939026895,
+      "grad_norm": 0.08500406891107559,
+      "learning_rate": 9.454196379594076e-05,
+      "loss": 0.3622,
+      "step": 204
+    },
+    {
+      "epoch": 0.1683432560049271,
+      "grad_norm": 0.08277002722024918,
+      "learning_rate": 9.451453647833243e-05,
+      "loss": 0.3596,
+      "step": 205
+    },
+    {
+      "epoch": 0.1691644426195853,
+      "grad_norm": 0.0855122059583664,
+      "learning_rate": 9.448710916072408e-05,
+      "loss": 0.3495,
+      "step": 206
+    },
+    {
+      "epoch": 0.16998562923424348,
+      "grad_norm": 0.11409081518650055,
+      "learning_rate": 9.445968184311574e-05,
+      "loss": 0.3123,
+      "step": 207
+    },
+    {
+      "epoch": 0.17080681584890167,
+      "grad_norm": 0.08898866921663284,
+      "learning_rate": 9.443225452550741e-05,
+      "loss": 0.3383,
+      "step": 208
+    },
+    {
+      "epoch": 0.17162800246355983,
+      "grad_norm": 0.07967101782560349,
+      "learning_rate": 9.440482720789908e-05,
+      "loss": 0.3927,
+      "step": 209
+    },
+    {
+      "epoch": 0.17244918907821802,
+      "grad_norm": 0.08267655968666077,
+      "learning_rate": 9.437739989029073e-05,
+      "loss": 0.3503,
+      "step": 210
+    },
+    {
+      "epoch": 0.1732703756928762,
+      "grad_norm": 0.1024966835975647,
+      "learning_rate": 9.43499725726824e-05,
+      "loss": 0.343,
+      "step": 211
+    },
+    {
+      "epoch": 0.1740915623075344,
+      "grad_norm": 0.11825034767389297,
+      "learning_rate": 9.432254525507405e-05,
+      "loss": 0.3694,
+      "step": 212
+    },
+    {
+      "epoch": 0.17491274892219258,
+      "grad_norm": 0.08225111663341522,
+      "learning_rate": 9.429511793746572e-05,
+      "loss": 0.3625,
+      "step": 213
+    },
+    {
+      "epoch": 0.17573393553685074,
+      "grad_norm": 0.082975834608078,
+      "learning_rate": 9.426769061985739e-05,
+      "loss": 0.3692,
+      "step": 214
+    },
+    {
+      "epoch": 0.17655512215150893,
+      "grad_norm": 0.09659875184297562,
+      "learning_rate": 9.424026330224904e-05,
+      "loss": 0.3857,
+      "step": 215
+    },
+    {
+      "epoch": 0.1773763087661671,
+      "grad_norm": 0.08381886035203934,
+      "learning_rate": 9.42128359846407e-05,
+      "loss": 0.3658,
+      "step": 216
+    },
+    {
+      "epoch": 0.1781974953808253,
+      "grad_norm": 0.08097488433122635,
+      "learning_rate": 9.418540866703238e-05,
+      "loss": 0.3556,
+      "step": 217
+    },
+    {
+      "epoch": 0.1790186819954835,
+      "grad_norm": 0.08661879599094391,
+      "learning_rate": 9.415798134942403e-05,
+      "loss": 0.3495,
+      "step": 218
+    },
+    {
+      "epoch": 0.17983986861014165,
+      "grad_norm": 0.09621778875589371,
+      "learning_rate": 9.413055403181569e-05,
+      "loss": 0.3488,
+      "step": 219
+    },
+    {
+      "epoch": 0.18066105522479983,
+      "grad_norm": 0.0867924615740776,
+      "learning_rate": 9.410312671420735e-05,
+      "loss": 0.3496,
+      "step": 220
+    },
+    {
+      "epoch": 0.18148224183945802,
+      "grad_norm": 0.09928230196237564,
+      "learning_rate": 9.407569939659901e-05,
+      "loss": 0.411,
+      "step": 221
+    },
+    {
+      "epoch": 0.1823034284541162,
+      "grad_norm": 0.08545473217964172,
+      "learning_rate": 9.404827207899069e-05,
+      "loss": 0.3707,
+      "step": 222
+    },
+    {
+      "epoch": 0.18312461506877437,
+      "grad_norm": 0.1317296177148819,
+      "learning_rate": 9.402084476138234e-05,
+      "loss": 0.4021,
+      "step": 223
+    },
+    {
+      "epoch": 0.18394580168343255,
+      "grad_norm": 0.09009065479040146,
+      "learning_rate": 9.3993417443774e-05,
+      "loss": 0.3451,
+      "step": 224
+    },
+    {
+      "epoch": 0.18476698829809074,
+      "grad_norm": 0.08332253247499466,
+      "learning_rate": 9.396599012616567e-05,
+      "loss": 0.3457,
+      "step": 225
+    },
+    {
+      "epoch": 0.18558817491274893,
+      "grad_norm": 0.08737312257289886,
+      "learning_rate": 9.393856280855733e-05,
+      "loss": 0.3721,
+      "step": 226
+    },
+    {
+      "epoch": 0.1864093615274071,
+      "grad_norm": 0.07472239434719086,
+      "learning_rate": 9.391113549094899e-05,
+      "loss": 0.3542,
+      "step": 227
+    },
+    {
+      "epoch": 0.18723054814206527,
+      "grad_norm": 0.08319877088069916,
+      "learning_rate": 9.388370817334066e-05,
+      "loss": 0.361,
+      "step": 228
+    },
+    {
+      "epoch": 0.18805173475672346,
+      "grad_norm": 0.08192326873540878,
+      "learning_rate": 9.385628085573231e-05,
+      "loss": 0.344,
+      "step": 229
+    },
+    {
+      "epoch": 0.18887292137138165,
+      "grad_norm": 0.11502642929553986,
+      "learning_rate": 9.382885353812398e-05,
+      "loss": 0.363,
+      "step": 230
+    },
+    {
+      "epoch": 0.18969410798603983,
+      "grad_norm": 0.08567750453948975,
+      "learning_rate": 9.380142622051565e-05,
+      "loss": 0.3565,
+      "step": 231
+    },
+    {
+      "epoch": 0.19051529460069802,
+      "grad_norm": 0.08059141039848328,
+      "learning_rate": 9.37739989029073e-05,
+      "loss": 0.3722,
+      "step": 232
+    },
+    {
+      "epoch": 0.19133648121535618,
+      "grad_norm": 0.07670270651578903,
+      "learning_rate": 9.374657158529896e-05,
+      "loss": 0.3493,
+      "step": 233
+    },
+    {
+      "epoch": 0.19215766783001437,
+      "grad_norm": 0.08022642135620117,
+      "learning_rate": 9.371914426769062e-05,
+      "loss": 0.4142,
+      "step": 234
+    },
+    {
+      "epoch": 0.19297885444467255,
+      "grad_norm": 0.08015397936105728,
+      "learning_rate": 9.369171695008229e-05,
+      "loss": 0.3442,
+      "step": 235
+    },
+    {
+      "epoch": 0.19380004105933074,
+      "grad_norm": 0.08329442143440247,
+      "learning_rate": 9.366428963247395e-05,
+      "loss": 0.3587,
+      "step": 236
+    },
+    {
+      "epoch": 0.1946212276739889,
+      "grad_norm": 0.08311276882886887,
+      "learning_rate": 9.363686231486561e-05,
+      "loss": 0.3792,
+      "step": 237
+    },
+    {
+      "epoch": 0.1954424142886471,
+      "grad_norm": 0.1332862675189972,
+      "learning_rate": 9.360943499725727e-05,
+      "loss": 0.3451,
+      "step": 238
+    },
+    {
+      "epoch": 0.19626360090330527,
+      "grad_norm": 0.081804558634758,
+      "learning_rate": 9.358200767964894e-05,
+      "loss": 0.4212,
+      "step": 239
+    },
+    {
+      "epoch": 0.19708478751796346,
+      "grad_norm": 0.08397019654512405,
+      "learning_rate": 9.35545803620406e-05,
+      "loss": 0.3482,
+      "step": 240
+    },
+    {
+      "epoch": 0.19790597413262165,
+      "grad_norm": 0.09495637565851212,
+      "learning_rate": 9.352715304443226e-05,
+      "loss": 0.3432,
+      "step": 241
+    },
+    {
+      "epoch": 0.1987271607472798,
+      "grad_norm": 0.09187504649162292,
+      "learning_rate": 9.349972572682391e-05,
+      "loss": 0.3938,
+      "step": 242
+    },
+    {
+      "epoch": 0.199548347361938,
+      "grad_norm": 0.100834921002388,
+      "learning_rate": 9.347229840921558e-05,
+      "loss": 0.3567,
+      "step": 243
+    },
+    {
+      "epoch": 0.20036953397659618,
+      "grad_norm": 0.09292273223400116,
+      "learning_rate": 9.344487109160725e-05,
+      "loss": 0.3802,
+      "step": 244
+    },
+    {
+      "epoch": 0.20119072059125437,
+      "grad_norm": 0.07785986363887787,
+      "learning_rate": 9.34174437739989e-05,
+      "loss": 0.3832,
+      "step": 245
+    },
+    {
+      "epoch": 0.20201190720591256,
+      "grad_norm": 0.081189826130867,
+      "learning_rate": 9.339001645639057e-05,
+      "loss": 0.3448,
+      "step": 246
+    },
+    {
+      "epoch": 0.20283309382057071,
+      "grad_norm": 0.08209879696369171,
+      "learning_rate": 9.336258913878222e-05,
+      "loss": 0.3602,
+      "step": 247
+    },
+    {
+      "epoch": 0.2036542804352289,
+      "grad_norm": 0.0813421905040741,
+      "learning_rate": 9.333516182117389e-05,
+      "loss": 0.3477,
+      "step": 248
+    },
+    {
+      "epoch": 0.2044754670498871,
+      "grad_norm": 0.08188773691654205,
+      "learning_rate": 9.330773450356556e-05,
+      "loss": 0.3466,
+      "step": 249
+    },
+    {
+      "epoch": 0.20529665366454528,
+      "grad_norm": 0.09966633468866348,
+      "learning_rate": 9.328030718595721e-05,
+      "loss": 0.3776,
+      "step": 250
+    },
+    {
+      "epoch": 0.20611784027920346,
+      "grad_norm": 0.08154954016208649,
+      "learning_rate": 9.325287986834887e-05,
+      "loss": 0.3727,
+      "step": 251
+    },
+    {
+      "epoch": 0.20693902689386162,
+      "grad_norm": 0.07942931354045868,
+      "learning_rate": 9.322545255074055e-05,
+      "loss": 0.3468,
+      "step": 252
+    },
+    {
+      "epoch": 0.2077602135085198,
+      "grad_norm": 0.07944433391094208,
+      "learning_rate": 9.31980252331322e-05,
+      "loss": 0.3512,
+      "step": 253
+    },
+    {
+      "epoch": 0.208581400123178,
+      "grad_norm": 0.08637971431016922,
+      "learning_rate": 9.317059791552387e-05,
+      "loss": 0.3397,
+      "step": 254
+    },
+    {
+      "epoch": 0.20940258673783618,
+      "grad_norm": 0.08189195394515991,
+      "learning_rate": 9.314317059791553e-05,
+      "loss": 0.4105,
+      "step": 255
+    },
+    {
+      "epoch": 0.21022377335249434,
+      "grad_norm": 0.08190836012363434,
+      "learning_rate": 9.31157432803072e-05,
+      "loss": 0.346,
+      "step": 256
+    },
+    {
+      "epoch": 0.21104495996715253,
+      "grad_norm": 0.08331865072250366,
+      "learning_rate": 9.308831596269886e-05,
+      "loss": 0.3655,
+      "step": 257
+    },
+    {
+      "epoch": 0.21186614658181072,
+      "grad_norm": 0.0822620540857315,
+      "learning_rate": 9.306088864509052e-05,
+      "loss": 0.372,
+      "step": 258
+    },
+    {
+      "epoch": 0.2126873331964689,
+      "grad_norm": 0.08011777698993683,
+      "learning_rate": 9.303346132748217e-05,
+      "loss": 0.3304,
+      "step": 259
+    },
+    {
+      "epoch": 0.2135085198111271,
+      "grad_norm": 0.08395062386989594,
+      "learning_rate": 9.300603400987384e-05,
+      "loss": 0.3903,
+      "step": 260
+    },
+    {
+      "epoch": 0.21432970642578525,
+      "grad_norm": 0.08101452887058258,
+      "learning_rate": 9.297860669226551e-05,
+      "loss": 0.3579,
+      "step": 261
+    },
+    {
+      "epoch": 0.21515089304044344,
+      "grad_norm": 0.11504925042390823,
+      "learning_rate": 9.295117937465716e-05,
+      "loss": 0.3723,
+      "step": 262
+    },
+    {
+      "epoch": 0.21597207965510162,
+      "grad_norm": 0.0704338401556015,
+      "learning_rate": 9.292375205704883e-05,
+      "loss": 0.3341,
+      "step": 263
+    },
+    {
+      "epoch": 0.2167932662697598,
+      "grad_norm": 0.07944470643997192,
+      "learning_rate": 9.289632473944048e-05,
+      "loss": 0.3273,
+      "step": 264
+    },
+    {
+      "epoch": 0.2167932662697598,
+      "eval_runtime": 507.3366,
+      "eval_samples_per_second": 0.388,
+      "eval_steps_per_second": 0.195,
+      "step": 264
+    },
+    {
+      "epoch": 0.217614452884418,
+      "grad_norm": 0.0778859481215477,
+      "learning_rate": 9.286889742183215e-05,
+      "loss": 0.3707,
+      "step": 265
+    },
+    {
+      "epoch": 0.21843563949907616,
+      "grad_norm": 0.07564377039670944,
+      "learning_rate": 9.284147010422382e-05,
+      "loss": 0.334,
+      "step": 266
+    },
+    {
+      "epoch": 0.21925682611373434,
+      "grad_norm": 0.07744503021240234,
+      "learning_rate": 9.281404278661547e-05,
+      "loss": 0.3521,
+      "step": 267
+    },
+    {
+      "epoch": 0.22007801272839253,
+      "grad_norm": 0.08139525353908539,
+      "learning_rate": 9.278661546900713e-05,
+      "loss": 0.4069,
+      "step": 268
+    },
+    {
+      "epoch": 0.22089919934305072,
+      "grad_norm": 0.08162204176187515,
+      "learning_rate": 9.27591881513988e-05,
+      "loss": 0.3292,
+      "step": 269
+    },
+    {
+      "epoch": 0.22172038595770888,
+      "grad_norm": 0.07798879593610764,
+      "learning_rate": 9.273176083379046e-05,
+      "loss": 0.3441,
+      "step": 270
+    },
+    {
+      "epoch": 0.22254157257236706,
+      "grad_norm": 0.07629597932100296,
+      "learning_rate": 9.270433351618212e-05,
+      "loss": 0.3989,
+      "step": 271
+    },
+    {
+      "epoch": 0.22336275918702525,
+      "grad_norm": 0.07695835083723068,
+      "learning_rate": 9.267690619857379e-05,
+      "loss": 0.3506,
+      "step": 272
+    },
+    {
+      "epoch": 0.22418394580168344,
+      "grad_norm": 0.07938527315855026,
+      "learning_rate": 9.264947888096544e-05,
+      "loss": 0.327,
+      "step": 273
+    },
+    {
+      "epoch": 0.22500513241634162,
+      "grad_norm": 0.08125253766775131,
+      "learning_rate": 9.262205156335711e-05,
+      "loss": 0.3749,
+      "step": 274
+    },
+    {
+      "epoch": 0.22582631903099978,
+      "grad_norm": 0.07831385731697083,
+      "learning_rate": 9.259462424574878e-05,
+      "loss": 0.3477,
+      "step": 275
+    },
+    {
+      "epoch": 0.22664750564565797,
+      "grad_norm": 0.08408527821302414,
+      "learning_rate": 9.256719692814043e-05,
+      "loss": 0.3488,
+      "step": 276
+    },
+    {
+      "epoch": 0.22746869226031616,
+      "grad_norm": 0.0800737589597702,
+      "learning_rate": 9.253976961053208e-05,
+      "loss": 0.3607,
+      "step": 277
+    },
+    {
+      "epoch": 0.22828987887497434,
+      "grad_norm": 0.07969113439321518,
+      "learning_rate": 9.251234229292375e-05,
+      "loss": 0.38,
+      "step": 278
+    },
+    {
+      "epoch": 0.22911106548963253,
+      "grad_norm": 0.08066916465759277,
+      "learning_rate": 9.248491497531542e-05,
+      "loss": 0.3469,
+      "step": 279
+    },
+    {
+      "epoch": 0.2299322521042907,
+      "grad_norm": 0.07686860859394073,
+      "learning_rate": 9.245748765770708e-05,
+      "loss": 0.3311,
+      "step": 280
+    },
+    {
+      "epoch": 0.23075343871894888,
+      "grad_norm": 0.13063722848892212,
+      "learning_rate": 9.243006034009874e-05,
+      "loss": 0.3091,
+      "step": 281
+    },
+    {
+      "epoch": 0.23157462533360706,
+      "grad_norm": 0.08465570211410522,
+      "learning_rate": 9.24026330224904e-05,
+      "loss": 0.3548,
+      "step": 282
+    },
+    {
+      "epoch": 0.23239581194826525,
+      "grad_norm": 0.07824988663196564,
+      "learning_rate": 9.237520570488207e-05,
+      "loss": 0.3485,
+      "step": 283
+    },
+    {
+      "epoch": 0.2332169985629234,
+      "grad_norm": 0.087236687541008,
+      "learning_rate": 9.234777838727373e-05,
+      "loss": 0.3564,
+      "step": 284
+    },
+    {
+      "epoch": 0.2340381851775816,
+      "grad_norm": 0.09531310200691223,
+      "learning_rate": 9.232035106966539e-05,
+      "loss": 0.3763,
+      "step": 285
+    },
+    {
+      "epoch": 0.23485937179223978,
+      "grad_norm": 0.0783536285161972,
+      "learning_rate": 9.229292375205704e-05,
+      "loss": 0.3479,
+      "step": 286
+    },
+    {
+      "epoch": 0.23568055840689797,
+      "grad_norm": 0.08209957927465439,
+      "learning_rate": 9.226549643444872e-05,
+      "loss": 0.3563,
+      "step": 287
+    },
+    {
+      "epoch": 0.23650174502155616,
+      "grad_norm": 0.08195500820875168,
+      "learning_rate": 9.223806911684038e-05,
+      "loss": 0.3879,
+      "step": 288
+    },
+    {
+      "epoch": 0.23732293163621432,
+      "grad_norm": 0.0796385332942009,
+      "learning_rate": 9.221064179923205e-05,
+      "loss": 0.3787,
+      "step": 289
+    },
+    {
+      "epoch": 0.2381441182508725,
+      "grad_norm": 0.08523210138082504,
+      "learning_rate": 9.21832144816237e-05,
+      "loss": 0.3608,
+      "step": 290
+    },
+    {
+      "epoch": 0.2389653048655307,
+      "grad_norm": 0.08573302626609802,
+      "learning_rate": 9.215578716401537e-05,
+      "loss": 0.3366,
+      "step": 291
+    },
+    {
+      "epoch": 0.23978649148018888,
+      "grad_norm": 0.08111650496721268,
+      "learning_rate": 9.212835984640704e-05,
+      "loss": 0.3331,
+      "step": 292
+    },
+    {
+      "epoch": 0.24060767809484707,
+      "grad_norm": 0.07877853512763977,
+      "learning_rate": 9.210093252879869e-05,
+      "loss": 0.3346,
+      "step": 293
+    },
+    {
+      "epoch": 0.24142886470950523,
+      "grad_norm": 0.0814940333366394,
+      "learning_rate": 9.207350521119034e-05,
+      "loss": 0.3701,
+      "step": 294
+    },
+    {
+      "epoch": 0.2422500513241634,
+      "grad_norm": 0.0843539908528328,
+      "learning_rate": 9.204607789358201e-05,
+      "loss": 0.3339,
+      "step": 295
+    },
+    {
+      "epoch": 0.2430712379388216,
+      "grad_norm": 0.0875789001584053,
+      "learning_rate": 9.201865057597368e-05,
+      "loss": 0.362,
+      "step": 296
+    },
+    {
+      "epoch": 0.2438924245534798,
+      "grad_norm": 0.08261071145534515,
+      "learning_rate": 9.199122325836533e-05,
+      "loss": 0.3578,
+      "step": 297
+    },
+    {
+      "epoch": 0.24471361116813795,
+      "grad_norm": 0.09160041064023972,
+      "learning_rate": 9.1963795940757e-05,
+      "loss": 0.387,
+      "step": 298
+    },
+    {
+      "epoch": 0.24553479778279613,
+      "grad_norm": 0.08152728527784348,
+      "learning_rate": 9.193636862314866e-05,
+      "loss": 0.3282,
+      "step": 299
+    },
+    {
+      "epoch": 0.24635598439745432,
+      "grad_norm": 0.079656682908535,
+      "learning_rate": 9.190894130554032e-05,
+      "loss": 0.3238,
+      "step": 300
+    },
+    {
+      "epoch": 0.2471771710121125,
+      "grad_norm": 0.0797642394900322,
+      "learning_rate": 9.188151398793199e-05,
+      "loss": 0.3506,
+      "step": 301
+    },
+    {
+      "epoch": 0.2479983576267707,
+      "grad_norm": 0.08688945323228836,
+      "learning_rate": 9.185408667032365e-05,
+      "loss": 0.4177,
+      "step": 302
+    },
+    {
+      "epoch": 0.24881954424142885,
+      "grad_norm": 0.106851726770401,
+      "learning_rate": 9.18266593527153e-05,
+      "loss": 0.3527,
+      "step": 303
+    },
+    {
+      "epoch": 0.24964073085608704,
+      "grad_norm": 0.08586329221725464,
+      "learning_rate": 9.179923203510697e-05,
+      "loss": 0.3763,
+      "step": 304
+    },
+    {
+      "epoch": 0.25046191747074525,
+      "grad_norm": 0.1141560897231102,
+      "learning_rate": 9.177180471749864e-05,
+      "loss": 0.3451,
+      "step": 305
+    },
+    {
+      "epoch": 0.2512831040854034,
+      "grad_norm": 0.08728976547718048,
+      "learning_rate": 9.174437739989029e-05,
+      "loss": 0.3444,
+      "step": 306
+    },
+    {
+      "epoch": 0.2521042907000616,
+      "grad_norm": 0.07848083972930908,
+      "learning_rate": 9.171695008228196e-05,
+      "loss": 0.344,
+      "step": 307
+    },
+    {
+      "epoch": 0.25292547731471976,
+      "grad_norm": 0.08140011131763458,
+      "learning_rate": 9.168952276467361e-05,
+      "loss": 0.3338,
+      "step": 308
+    },
+    {
+      "epoch": 0.25374666392937795,
+      "grad_norm": 0.07934946566820145,
+      "learning_rate": 9.166209544706528e-05,
+      "loss": 0.3869,
+      "step": 309
+    },
+    {
+      "epoch": 0.25456785054403613,
+      "grad_norm": 0.08028525859117508,
+      "learning_rate": 9.163466812945695e-05,
+      "loss": 0.3141,
+      "step": 310
+    },
+    {
+      "epoch": 0.2553890371586943,
+      "grad_norm": 0.08121496438980103,
+      "learning_rate": 9.16072408118486e-05,
+      "loss": 0.4098,
+      "step": 311
+    },
+    {
+      "epoch": 0.2562102237733525,
+      "grad_norm": 0.09907019138336182,
+      "learning_rate": 9.157981349424026e-05,
+      "loss": 0.3594,
+      "step": 312
+    },
+    {
+      "epoch": 0.2570314103880107,
+      "grad_norm": 0.10270131379365921,
+      "learning_rate": 9.155238617663193e-05,
+      "loss": 0.3713,
+      "step": 313
+    },
+    {
+      "epoch": 0.2578525970026689,
+      "grad_norm": 0.08354981243610382,
+      "learning_rate": 9.15249588590236e-05,
+      "loss": 0.3943,
+      "step": 314
+    },
+    {
+      "epoch": 0.258673783617327,
+      "grad_norm": 0.09153895825147629,
+      "learning_rate": 9.149753154141525e-05,
+      "loss": 0.3499,
+      "step": 315
+    },
+    {
+      "epoch": 0.2594949702319852,
+      "grad_norm": 0.07772937417030334,
+      "learning_rate": 9.147010422380692e-05,
+      "loss": 0.3323,
+      "step": 316
+    },
+    {
+      "epoch": 0.2603161568466434,
+      "grad_norm": 0.08953258395195007,
+      "learning_rate": 9.144267690619857e-05,
+      "loss": 0.3684,
+      "step": 317
+    },
+    {
+      "epoch": 0.2611373434613016,
+      "grad_norm": 0.08204226195812225,
+      "learning_rate": 9.141524958859024e-05,
+      "loss": 0.3387,
+      "step": 318
+    },
+    {
+      "epoch": 0.26195853007595976,
+      "grad_norm": 0.07604125887155533,
+      "learning_rate": 9.13878222709819e-05,
+      "loss": 0.3104,
+      "step": 319
+    },
+    {
+      "epoch": 0.26277971669061795,
+      "grad_norm": 0.07614021003246307,
+      "learning_rate": 9.136039495337356e-05,
+      "loss": 0.329,
+      "step": 320
+    },
+    {
+      "epoch": 0.26360090330527614,
+      "grad_norm": 0.19391578435897827,
+      "learning_rate": 9.133296763576523e-05,
+      "loss": 0.349,
+      "step": 321
+    },
+    {
+      "epoch": 0.2644220899199343,
+      "grad_norm": 0.07927168160676956,
+      "learning_rate": 9.13055403181569e-05,
+      "loss": 0.3495,
+      "step": 322
+    },
+    {
+      "epoch": 0.2652432765345925,
+      "grad_norm": 0.08091975003480911,
+      "learning_rate": 9.127811300054855e-05,
+      "loss": 0.3563,
+      "step": 323
+    },
+    {
+      "epoch": 0.26606446314925064,
+      "grad_norm": 0.07865800708532333,
+      "learning_rate": 9.125068568294022e-05,
+      "loss": 0.36,
+      "step": 324
+    },
+    {
+      "epoch": 0.2668856497639088,
+      "grad_norm": 0.08946891874074936,
+      "learning_rate": 9.122325836533187e-05,
+      "loss": 0.3434,
+      "step": 325
+    },
+    {
+      "epoch": 0.267706836378567,
+      "grad_norm": 0.07987753301858902,
+      "learning_rate": 9.119583104772354e-05,
+      "loss": 0.3593,
+      "step": 326
+    },
+    {
+      "epoch": 0.2685280229932252,
+      "grad_norm": 0.07894265651702881,
+      "learning_rate": 9.116840373011521e-05,
+      "loss": 0.3746,
+      "step": 327
+    },
+    {
+      "epoch": 0.2693492096078834,
+      "grad_norm": 0.08121046423912048,
+      "learning_rate": 9.114097641250686e-05,
+      "loss": 0.357,
+      "step": 328
+    },
+    {
+      "epoch": 0.2701703962225416,
+      "grad_norm": 0.08320184051990509,
+      "learning_rate": 9.111354909489852e-05,
+      "loss": 0.3783,
+      "step": 329
+    },
+    {
+      "epoch": 0.27099158283719976,
+      "grad_norm": 0.10911049693822861,
+      "learning_rate": 9.108612177729018e-05,
+      "loss": 0.3699,
+      "step": 330
+    },
+    {
+      "epoch": 0.27181276945185795,
+      "grad_norm": 0.09739838540554047,
+      "learning_rate": 9.105869445968185e-05,
+      "loss": 0.3524,
+      "step": 331
+    },
+    {
+      "epoch": 0.27263395606651614,
+      "grad_norm": 0.0874953344464302,
+      "learning_rate": 9.103126714207351e-05,
+      "loss": 0.3807,
+      "step": 332
+    },
+    {
+      "epoch": 0.2734551426811743,
+      "grad_norm": 0.09758396446704865,
+      "learning_rate": 9.100383982446518e-05,
+      "loss": 0.3876,
+      "step": 333
+    },
+    {
+      "epoch": 0.27427632929583246,
+      "grad_norm": 0.08628170937299728,
+      "learning_rate": 9.097641250685683e-05,
+      "loss": 0.3632,
+      "step": 334
+    },
+    {
+      "epoch": 0.27509751591049064,
+      "grad_norm": 0.08430638164281845,
+      "learning_rate": 9.09489851892485e-05,
+      "loss": 0.3448,
+      "step": 335
+    },
+    {
+      "epoch": 0.27591870252514883,
+      "grad_norm": 0.0725574865937233,
+      "learning_rate": 9.092155787164017e-05,
+      "loss": 0.2914,
+      "step": 336
+    },
+    {
+      "epoch": 0.276739889139807,
+      "grad_norm": 0.08421092480421066,
+      "learning_rate": 9.089413055403182e-05,
+      "loss": 0.3676,
+      "step": 337
+    },
+    {
+      "epoch": 0.2775610757544652,
+      "grad_norm": 0.08702068030834198,
+      "learning_rate": 9.086670323642347e-05,
+      "loss": 0.345,
+      "step": 338
+    },
+    {
+      "epoch": 0.2783822623691234,
+      "grad_norm": 0.08053667843341827,
+      "learning_rate": 9.083927591881514e-05,
+      "loss": 0.324,
+      "step": 339
+    },
+    {
+      "epoch": 0.2792034489837816,
+      "grad_norm": 0.08294267952442169,
+      "learning_rate": 9.081184860120681e-05,
+      "loss": 0.348,
+      "step": 340
+    },
+    {
+      "epoch": 0.28002463559843976,
+      "grad_norm": 0.18557026982307434,
+      "learning_rate": 9.078442128359846e-05,
+      "loss": 0.3483,
+      "step": 341
+    },
+    {
+      "epoch": 0.28084582221309795,
+      "grad_norm": 0.08436847478151321,
+      "learning_rate": 9.075699396599013e-05,
+      "loss": 0.3641,
+      "step": 342
+    },
+    {
+      "epoch": 0.2816670088277561,
+      "grad_norm": 0.08285505324602127,
+      "learning_rate": 9.072956664838179e-05,
+      "loss": 0.3641,
+      "step": 343
+    },
+    {
+      "epoch": 0.28248819544241427,
+      "grad_norm": 0.08681569993495941,
+      "learning_rate": 9.070213933077345e-05,
+      "loss": 0.342,
+      "step": 344
+    },
+    {
+      "epoch": 0.28330938205707246,
+      "grad_norm": 0.08668892085552216,
+      "learning_rate": 9.067471201316512e-05,
+      "loss": 0.3514,
+      "step": 345
+    },
+    {
+      "epoch": 0.28413056867173064,
+      "grad_norm": 0.08393285423517227,
+      "learning_rate": 9.064728469555678e-05,
+      "loss": 0.3531,
+      "step": 346
+    },
+    {
+      "epoch": 0.28495175528638883,
+      "grad_norm": 0.07783684879541397,
+      "learning_rate": 9.061985737794843e-05,
+      "loss": 0.3792,
+      "step": 347
+    },
+    {
+      "epoch": 0.285772941901047,
+      "grad_norm": 0.08259117603302002,
+      "learning_rate": 9.059243006034011e-05,
+      "loss": 0.375,
+      "step": 348
+    },
+    {
+      "epoch": 0.2865941285157052,
+      "grad_norm": 0.0969453975558281,
+      "learning_rate": 9.056500274273177e-05,
+      "loss": 0.3776,
+      "step": 349
+    },
+    {
+      "epoch": 0.2874153151303634,
+      "grad_norm": 0.08154502511024475,
+      "learning_rate": 9.053757542512342e-05,
+      "loss": 0.3577,
+      "step": 350
+    },
+    {
+      "epoch": 0.2882365017450216,
+      "grad_norm": 0.09322196245193481,
+      "learning_rate": 9.051014810751509e-05,
+      "loss": 0.3774,
+      "step": 351
+    },
+    {
+      "epoch": 0.2890576883596797,
+      "grad_norm": 0.0804852545261383,
+      "learning_rate": 9.048272078990674e-05,
+      "loss": 0.333,
+      "step": 352
+    },
+    {
+      "epoch": 0.2890576883596797,
+      "eval_runtime": 493.668,
+      "eval_samples_per_second": 0.399,
+      "eval_steps_per_second": 0.201,
+      "step": 352
+    },
+    {
+      "epoch": 0.2898788749743379,
+      "grad_norm": 0.08082845062017441,
+      "learning_rate": 9.045529347229841e-05,
+      "loss": 0.3111,
+      "step": 353
+    },
+    {
+      "epoch": 0.2907000615889961,
+      "grad_norm": 0.08396881073713303,
+      "learning_rate": 9.042786615469008e-05,
+      "loss": 0.3429,
+      "step": 354
+    },
+    {
+      "epoch": 0.29152124820365427,
+      "grad_norm": 0.08037351071834564,
+      "learning_rate": 9.040043883708173e-05,
+      "loss": 0.349,
+      "step": 355
+    },
+    {
+      "epoch": 0.29234243481831246,
+      "grad_norm": 0.10421521961688995,
+      "learning_rate": 9.03730115194734e-05,
+      "loss": 0.3228,
+      "step": 356
+    },
+    {
+      "epoch": 0.29316362143297064,
+      "grad_norm": 0.08268436044454575,
+      "learning_rate": 9.034558420186507e-05,
+      "loss": 0.3414,
+      "step": 357
+    },
+    {
+      "epoch": 0.29398480804762883,
+      "grad_norm": 0.08203061670064926,
+      "learning_rate": 9.031815688425672e-05,
+      "loss": 0.3596,
+      "step": 358
+    },
+    {
+      "epoch": 0.294805994662287,
+      "grad_norm": 0.07778481394052505,
+      "learning_rate": 9.029072956664839e-05,
+      "loss": 0.3337,
+      "step": 359
+    },
+    {
+      "epoch": 0.2956271812769452,
+      "grad_norm": 0.14194349944591522,
+      "learning_rate": 9.026330224904005e-05,
+      "loss": 0.3613,
+      "step": 360
+    },
+    {
+      "epoch": 0.2964483678916034,
+      "grad_norm": 0.09854190051555634,
+      "learning_rate": 9.023587493143171e-05,
+      "loss": 0.3862,
+      "step": 361
+    },
+    {
+      "epoch": 0.2972695545062615,
+      "grad_norm": 0.07878783345222473,
+      "learning_rate": 9.020844761382338e-05,
+      "loss": 0.3294,
+      "step": 362
+    },
+    {
+      "epoch": 0.2980907411209197,
+      "grad_norm": 0.07692375034093857,
+      "learning_rate": 9.018102029621504e-05,
+      "loss": 0.3083,
+      "step": 363
+    },
+    {
+      "epoch": 0.2989119277355779,
+      "grad_norm": 0.0772586315870285,
+      "learning_rate": 9.015359297860669e-05,
+      "loss": 0.3086,
+      "step": 364
+    },
+    {
+      "epoch": 0.2997331143502361,
+      "grad_norm": 0.0739215686917305,
+      "learning_rate": 9.012616566099836e-05,
+      "loss": 0.3103,
+      "step": 365
+    },
+    {
+      "epoch": 0.30055430096489427,
+      "grad_norm": 0.08434160053730011,
+      "learning_rate": 9.009873834339003e-05,
+      "loss": 0.3485,
+      "step": 366
+    },
+    {
+      "epoch": 0.30137548757955246,
+      "grad_norm": 0.08038529008626938,
+      "learning_rate": 9.007131102578168e-05,
+      "loss": 0.3421,
+      "step": 367
+    },
+    {
+      "epoch": 0.30219667419421065,
+      "grad_norm": 0.08303964138031006,
+      "learning_rate": 9.004388370817335e-05,
+      "loss": 0.3349,
+      "step": 368
+    },
+    {
+      "epoch": 0.30301786080886883,
+      "grad_norm": 0.08672379702329636,
+      "learning_rate": 9.0016456390565e-05,
+      "loss": 0.3228,
+      "step": 369
+    },
+    {
+      "epoch": 0.303839047423527,
+      "grad_norm": 0.08417553454637527,
+      "learning_rate": 8.998902907295667e-05,
+      "loss": 0.3312,
+      "step": 370
+    },
+    {
+      "epoch": 0.30466023403818515,
+      "grad_norm": 0.0842244029045105,
+      "learning_rate": 8.996160175534834e-05,
+      "loss": 0.3545,
+      "step": 371
+    },
+    {
+      "epoch": 0.30548142065284334,
+      "grad_norm": 0.08479683846235275,
+      "learning_rate": 8.993417443773999e-05,
+      "loss": 0.3446,
+      "step": 372
+    },
+    {
+      "epoch": 0.3063026072675015,
+      "grad_norm": 0.07726123929023743,
+      "learning_rate": 8.990674712013165e-05,
+      "loss": 0.3116,
+      "step": 373
+    },
+    {
+      "epoch": 0.3071237938821597,
+      "grad_norm": 0.08586291968822479,
+      "learning_rate": 8.987931980252331e-05,
+      "loss": 0.3617,
+      "step": 374
+    },
+    {
+      "epoch": 0.3079449804968179,
+      "grad_norm": 0.07475403696298599,
+      "learning_rate": 8.985189248491498e-05,
+      "loss": 0.3481,
+      "step": 375
+    },
+    {
+      "epoch": 0.3087661671114761,
+      "grad_norm": 0.08379320055246353,
+      "learning_rate": 8.982446516730664e-05,
+      "loss": 0.3434,
+      "step": 376
+    },
+    {
+      "epoch": 0.3095873537261343,
+      "grad_norm": 0.07624776661396027,
+      "learning_rate": 8.97970378496983e-05,
+      "loss": 0.3262,
+      "step": 377
+    },
+    {
+      "epoch": 0.31040854034079246,
+      "grad_norm": 0.07878948748111725,
+      "learning_rate": 8.976961053208996e-05,
+      "loss": 0.3366,
+      "step": 378
+    },
+    {
+      "epoch": 0.31122972695545065,
+      "grad_norm": 0.08924887329339981,
+      "learning_rate": 8.974218321448163e-05,
+      "loss": 0.3175,
+      "step": 379
+    },
+    {
+      "epoch": 0.31205091357010883,
+      "grad_norm": 0.08927389234304428,
+      "learning_rate": 8.97147558968733e-05,
+      "loss": 0.3338,
+      "step": 380
+    },
+    {
+      "epoch": 0.31287210018476697,
+      "grad_norm": 0.08250463008880615,
+      "learning_rate": 8.968732857926495e-05,
+      "loss": 0.3314,
+      "step": 381
+    },
+    {
+      "epoch": 0.31369328679942515,
+      "grad_norm": 0.0977792888879776,
+      "learning_rate": 8.96599012616566e-05,
+      "loss": 0.3222,
+      "step": 382
+    },
+    {
+      "epoch": 0.31451447341408334,
+      "grad_norm": 0.08197218924760818,
+      "learning_rate": 8.963247394404828e-05,
+      "loss": 0.3288,
+      "step": 383
+    },
+    {
+      "epoch": 0.3153356600287415,
+      "grad_norm": 0.12502075731754303,
+      "learning_rate": 8.960504662643994e-05,
+      "loss": 0.3053,
+      "step": 384
+    },
+    {
+      "epoch": 0.3161568466433997,
+      "grad_norm": 0.08058564364910126,
+      "learning_rate": 8.95776193088316e-05,
+      "loss": 0.3292,
+      "step": 385
+    },
+    {
+      "epoch": 0.3169780332580579,
+      "grad_norm": 0.08365239202976227,
+      "learning_rate": 8.955019199122326e-05,
+      "loss": 0.3747,
+      "step": 386
+    },
+    {
+      "epoch": 0.3177992198727161,
+      "grad_norm": 0.07539816945791245,
+      "learning_rate": 8.952276467361493e-05,
+      "loss": 0.3179,
+      "step": 387
+    },
+    {
+      "epoch": 0.3186204064873743,
+      "grad_norm": 0.08136752992868423,
+      "learning_rate": 8.94953373560066e-05,
+      "loss": 0.3647,
+      "step": 388
+    },
+    {
+      "epoch": 0.31944159310203246,
+      "grad_norm": 0.09546865522861481,
+      "learning_rate": 8.946791003839825e-05,
+      "loss": 0.4056,
+      "step": 389
+    },
+    {
+      "epoch": 0.3202627797166906,
+      "grad_norm": 0.08654627203941345,
+      "learning_rate": 8.94404827207899e-05,
+      "loss": 0.3853,
+      "step": 390
+    },
+    {
+      "epoch": 0.3210839663313488,
+      "grad_norm": 0.0822821855545044,
+      "learning_rate": 8.941305540318157e-05,
+      "loss": 0.327,
+      "step": 391
+    },
+    {
+      "epoch": 0.32190515294600697,
+      "grad_norm": 0.08066459000110626,
+      "learning_rate": 8.938562808557324e-05,
+      "loss": 0.3585,
+      "step": 392
+    },
+    {
+      "epoch": 0.32272633956066515,
+      "grad_norm": 0.08845409005880356,
+      "learning_rate": 8.93582007679649e-05,
+      "loss": 0.3419,
+      "step": 393
+    },
+    {
+      "epoch": 0.32354752617532334,
+      "grad_norm": 0.08482253551483154,
+      "learning_rate": 8.933077345035656e-05,
+      "loss": 0.3263,
+      "step": 394
+    },
+    {
+      "epoch": 0.32436871278998153,
+      "grad_norm": 0.0801006481051445,
+      "learning_rate": 8.930334613274822e-05,
+      "loss": 0.3713,
+      "step": 395
+    },
+    {
+      "epoch": 0.3251898994046397,
+      "grad_norm": 0.08470006287097931,
+      "learning_rate": 8.927591881513989e-05,
+      "loss": 0.3631,
+      "step": 396
+    },
+    {
+      "epoch": 0.3260110860192979,
+      "grad_norm": 0.08457473665475845,
+      "learning_rate": 8.924849149753155e-05,
+      "loss": 0.339,
+      "step": 397
+    },
+    {
+      "epoch": 0.3268322726339561,
+      "grad_norm": 0.14836351573467255,
+      "learning_rate": 8.922106417992321e-05,
+      "loss": 0.3039,
+      "step": 398
+    },
+    {
+      "epoch": 0.3276534592486142,
+      "grad_norm": 0.08391403406858444,
+      "learning_rate": 8.919363686231486e-05,
+      "loss": 0.3415,
+      "step": 399
+    },
+    {
+      "epoch": 0.3284746458632724,
+      "grad_norm": 0.07998157292604446,
+      "learning_rate": 8.916620954470653e-05,
+      "loss": 0.2943,
+      "step": 400
+    },
+    {
+      "epoch": 0.3292958324779306,
+      "grad_norm": 0.07682473957538605,
+      "learning_rate": 8.91387822270982e-05,
+      "loss": 0.3413,
+      "step": 401
+    },
+    {
+      "epoch": 0.3301170190925888,
+      "grad_norm": 0.09960923343896866,
+      "learning_rate": 8.911135490948985e-05,
+      "loss": 0.356,
+      "step": 402
+    },
+    {
+      "epoch": 0.33093820570724697,
+      "grad_norm": 0.08010439574718475,
+      "learning_rate": 8.908392759188152e-05,
+      "loss": 0.3408,
+      "step": 403
+    },
+    {
+      "epoch": 0.33175939232190516,
+      "grad_norm": 0.0811014398932457,
+      "learning_rate": 8.905650027427318e-05,
+      "loss": 0.3159,
+      "step": 404
+    },
+    {
+      "epoch": 0.33258057893656334,
+      "grad_norm": 0.08086768537759781,
+      "learning_rate": 8.902907295666484e-05,
+      "loss": 0.3145,
+      "step": 405
+    },
+    {
+      "epoch": 0.33340176555122153,
+      "grad_norm": 0.0778844803571701,
+      "learning_rate": 8.900164563905651e-05,
+      "loss": 0.3271,
+      "step": 406
+    },
+    {
+      "epoch": 0.3342229521658797,
+      "grad_norm": 0.08636100590229034,
+      "learning_rate": 8.897421832144817e-05,
+      "loss": 0.395,
+      "step": 407
+    },
+    {
+      "epoch": 0.3350441387805379,
+      "grad_norm": 0.08080980181694031,
+      "learning_rate": 8.894679100383982e-05,
+      "loss": 0.3819,
+      "step": 408
+    },
+    {
+      "epoch": 0.33586532539519603,
+      "grad_norm": 0.08312460780143738,
+      "learning_rate": 8.891936368623149e-05,
+      "loss": 0.3413,
+      "step": 409
+    },
+    {
+      "epoch": 0.3366865120098542,
+      "grad_norm": 0.07980356365442276,
+      "learning_rate": 8.889193636862316e-05,
+      "loss": 0.3546,
+      "step": 410
+    },
+    {
+      "epoch": 0.3375076986245124,
+      "grad_norm": 0.0868874043226242,
+      "learning_rate": 8.886450905101481e-05,
+      "loss": 0.3354,
+      "step": 411
+    },
+    {
+      "epoch": 0.3383288852391706,
+      "grad_norm": 0.08026042580604553,
+      "learning_rate": 8.883708173340648e-05,
+      "loss": 0.355,
+      "step": 412
+    },
+    {
+      "epoch": 0.3391500718538288,
+      "grad_norm": 0.0777151882648468,
+      "learning_rate": 8.880965441579813e-05,
+      "loss": 0.2835,
+      "step": 413
+    },
+    {
+      "epoch": 0.33997125846848697,
+      "grad_norm": 0.08007453382015228,
+      "learning_rate": 8.87822270981898e-05,
+      "loss": 0.3042,
+      "step": 414
+    },
+    {
+      "epoch": 0.34079244508314516,
+      "grad_norm": 0.08239760994911194,
+      "learning_rate": 8.875479978058147e-05,
+      "loss": 0.4139,
+      "step": 415
+    },
+    {
+      "epoch": 0.34161363169780334,
+      "grad_norm": 0.08022072911262512,
+      "learning_rate": 8.872737246297312e-05,
+      "loss": 0.319,
+      "step": 416
+    },
+    {
+      "epoch": 0.34243481831246153,
+      "grad_norm": 0.09205158054828644,
+      "learning_rate": 8.869994514536478e-05,
+      "loss": 0.3283,
+      "step": 417
+    },
+    {
+      "epoch": 0.34325600492711966,
+      "grad_norm": 0.07974108308553696,
+      "learning_rate": 8.867251782775646e-05,
+      "loss": 0.3331,
+      "step": 418
+    },
+    {
+      "epoch": 0.34407719154177785,
+      "grad_norm": 0.0810483992099762,
+      "learning_rate": 8.864509051014811e-05,
+      "loss": 0.4005,
+      "step": 419
+    },
+    {
+      "epoch": 0.34489837815643604,
+      "grad_norm": 0.08448805660009384,
+      "learning_rate": 8.861766319253978e-05,
+      "loss": 0.3327,
+      "step": 420
+    },
+    {
+      "epoch": 0.3457195647710942,
+      "grad_norm": 0.0865749716758728,
+      "learning_rate": 8.859023587493143e-05,
+      "loss": 0.4459,
+      "step": 421
+    },
+    {
+      "epoch": 0.3465407513857524,
+      "grad_norm": 0.08855796605348587,
+      "learning_rate": 8.85628085573231e-05,
+      "loss": 0.3566,
+      "step": 422
+    },
+    {
+      "epoch": 0.3473619380004106,
+      "grad_norm": 0.08421412855386734,
+      "learning_rate": 8.853538123971477e-05,
+      "loss": 0.3248,
+      "step": 423
+    },
+    {
+      "epoch": 0.3481831246150688,
+      "grad_norm": 0.07903390377759933,
+      "learning_rate": 8.850795392210642e-05,
+      "loss": 0.3608,
+      "step": 424
+    },
+    {
+      "epoch": 0.34900431122972697,
+      "grad_norm": 0.09979040920734406,
+      "learning_rate": 8.848052660449808e-05,
+      "loss": 0.3079,
+      "step": 425
+    },
+    {
+      "epoch": 0.34982549784438516,
+      "grad_norm": 0.08713112026453018,
+      "learning_rate": 8.845309928688975e-05,
+      "loss": 0.3462,
+      "step": 426
+    },
+    {
+      "epoch": 0.35064668445904335,
+      "grad_norm": 0.07836464792490005,
+      "learning_rate": 8.842567196928141e-05,
+      "loss": 0.3467,
+      "step": 427
+    },
+    {
+      "epoch": 0.3514678710737015,
+      "grad_norm": 0.07553427666425705,
+      "learning_rate": 8.839824465167307e-05,
+      "loss": 0.3325,
+      "step": 428
+    },
+    {
+      "epoch": 0.35228905768835966,
+      "grad_norm": 0.08221621066331863,
+      "learning_rate": 8.837081733406474e-05,
+      "loss": 0.3366,
+      "step": 429
+    },
+    {
+      "epoch": 0.35311024430301785,
+      "grad_norm": 0.09283772855997086,
+      "learning_rate": 8.834339001645639e-05,
+      "loss": 0.3518,
+      "step": 430
+    },
+    {
+      "epoch": 0.35393143091767604,
+      "grad_norm": 0.08067970722913742,
+      "learning_rate": 8.831596269884806e-05,
+      "loss": 0.3521,
+      "step": 431
+    },
+    {
+      "epoch": 0.3547526175323342,
+      "grad_norm": 0.08192427456378937,
+      "learning_rate": 8.828853538123973e-05,
+      "loss": 0.3794,
+      "step": 432
+    },
+    {
+      "epoch": 0.3555738041469924,
+      "grad_norm": 0.08332204073667526,
+      "learning_rate": 8.826110806363138e-05,
+      "loss": 0.3243,
+      "step": 433
+    },
+    {
+      "epoch": 0.3563949907616506,
+      "grad_norm": 0.08099368214607239,
+      "learning_rate": 8.823368074602304e-05,
+      "loss": 0.3409,
+      "step": 434
+    },
+    {
+      "epoch": 0.3572161773763088,
+      "grad_norm": 0.0801381841301918,
+      "learning_rate": 8.82062534284147e-05,
+      "loss": 0.3489,
+      "step": 435
+    },
+    {
+      "epoch": 0.358037363990967,
+      "grad_norm": 0.0828125849366188,
+      "learning_rate": 8.817882611080637e-05,
+      "loss": 0.3332,
+      "step": 436
+    },
+    {
+      "epoch": 0.3588585506056251,
+      "grad_norm": 0.08463995903730392,
+      "learning_rate": 8.815139879319803e-05,
+      "loss": 0.3532,
+      "step": 437
+    },
+    {
+      "epoch": 0.3596797372202833,
+      "grad_norm": 0.08150382339954376,
+      "learning_rate": 8.81239714755897e-05,
+      "loss": 0.3271,
+      "step": 438
+    },
+    {
+      "epoch": 0.3605009238349415,
+      "grad_norm": 0.08224772661924362,
+      "learning_rate": 8.809654415798135e-05,
+      "loss": 0.3847,
+      "step": 439
+    },
+    {
+      "epoch": 0.36132211044959966,
+      "grad_norm": 0.07509076595306396,
+      "learning_rate": 8.806911684037302e-05,
+      "loss": 0.346,
+      "step": 440
+    },
+    {
+      "epoch": 0.36132211044959966,
+      "eval_runtime": 493.5837,
+      "eval_samples_per_second": 0.399,
+      "eval_steps_per_second": 0.201,
+      "step": 440
+    },
+    {
+      "epoch": 0.36214329706425785,
+      "grad_norm": 0.07801686972379684,
+      "learning_rate": 8.804168952276468e-05,
+      "loss": 0.3352,
+      "step": 441
+    },
+    {
+      "epoch": 0.36296448367891604,
+      "grad_norm": 0.08195506781339645,
+      "learning_rate": 8.801426220515634e-05,
+      "loss": 0.3076,
+      "step": 442
+    },
+    {
+      "epoch": 0.3637856702935742,
+      "grad_norm": 0.08178213983774185,
+      "learning_rate": 8.798683488754799e-05,
+      "loss": 0.3565,
+      "step": 443
+    },
+    {
+      "epoch": 0.3646068569082324,
+      "grad_norm": 0.07905025035142899,
+      "learning_rate": 8.795940756993966e-05,
+      "loss": 0.3513,
+      "step": 444
+    },
+    {
+      "epoch": 0.3654280435228906,
+      "grad_norm": 0.1251908391714096,
+      "learning_rate": 8.793198025233133e-05,
+      "loss": 0.3656,
+      "step": 445
+    },
+    {
+      "epoch": 0.36624923013754873,
+      "grad_norm": 0.08813164383172989,
+      "learning_rate": 8.790455293472298e-05,
+      "loss": 0.3352,
+      "step": 446
+    },
+    {
+      "epoch": 0.3670704167522069,
+      "grad_norm": 0.0773930549621582,
+      "learning_rate": 8.787712561711465e-05,
+      "loss": 0.3241,
+      "step": 447
+    },
+    {
+      "epoch": 0.3678916033668651,
+      "grad_norm": 0.08977091312408447,
+      "learning_rate": 8.78496982995063e-05,
+      "loss": 0.3308,
+      "step": 448
+    },
+    {
+      "epoch": 0.3687127899815233,
+      "grad_norm": 0.0813109427690506,
+      "learning_rate": 8.782227098189797e-05,
+      "loss": 0.344,
+      "step": 449
+    },
+    {
+      "epoch": 0.3695339765961815,
+      "grad_norm": 0.08452528715133667,
+      "learning_rate": 8.779484366428964e-05,
+      "loss": 0.2985,
+      "step": 450
+    },
+    {
+      "epoch": 0.37035516321083967,
+      "grad_norm": 0.08880957961082458,
+      "learning_rate": 8.77674163466813e-05,
+      "loss": 0.3674,
+      "step": 451
+    },
+    {
+      "epoch": 0.37117634982549785,
+      "grad_norm": 0.08020122349262238,
+      "learning_rate": 8.773998902907295e-05,
+      "loss": 0.3487,
+      "step": 452
+    },
+    {
+      "epoch": 0.37199753644015604,
+      "grad_norm": 0.09180288016796112,
+      "learning_rate": 8.771256171146463e-05,
+      "loss": 0.364,
+      "step": 453
+    },
+    {
+      "epoch": 0.3728187230548142,
+      "grad_norm": 0.08132430911064148,
+      "learning_rate": 8.768513439385628e-05,
+      "loss": 0.3428,
+      "step": 454
+    },
+    {
+      "epoch": 0.3736399096694724,
+      "grad_norm": 0.08801679313182831,
+      "learning_rate": 8.765770707624795e-05,
+      "loss": 0.3287,
+      "step": 455
+    },
+    {
+      "epoch": 0.37446109628413055,
+      "grad_norm": 0.088813915848732,
+      "learning_rate": 8.763027975863961e-05,
+      "loss": 0.3508,
+      "step": 456
+    },
+    {
+      "epoch": 0.37528228289878873,
+      "grad_norm": 0.07454930245876312,
+      "learning_rate": 8.760285244103128e-05,
+      "loss": 0.3267,
+      "step": 457
+    },
+    {
+      "epoch": 0.3761034695134469,
+      "grad_norm": 0.0816444680094719,
+      "learning_rate": 8.757542512342294e-05,
+      "loss": 0.3237,
+      "step": 458
+    },
+    {
+      "epoch": 0.3769246561281051,
+      "grad_norm": 0.08916571736335754,
+      "learning_rate": 8.75479978058146e-05,
+      "loss": 0.3599,
+      "step": 459
+    },
+    {
+      "epoch": 0.3777458427427633,
+      "grad_norm": 0.08580256998538971,
+      "learning_rate": 8.752057048820625e-05,
+      "loss": 0.351,
+      "step": 460
+    },
+    {
+      "epoch": 0.3785670293574215,
+      "grad_norm": 0.0849752202630043,
+      "learning_rate": 8.749314317059792e-05,
+      "loss": 0.3616,
+      "step": 461
+    },
+    {
+      "epoch": 0.37938821597207967,
+      "grad_norm": 0.11230983585119247,
+      "learning_rate": 8.746571585298959e-05,
+      "loss": 0.3401,
+      "step": 462
+    },
+    {
+      "epoch": 0.38020940258673785,
+      "grad_norm": 0.08614380657672882,
+      "learning_rate": 8.743828853538124e-05,
+      "loss": 0.3603,
+      "step": 463
+    },
+    {
+      "epoch": 0.38103058920139604,
+      "grad_norm": 0.08322657644748688,
+      "learning_rate": 8.741086121777291e-05,
+      "loss": 0.3087,
+      "step": 464
+    },
+    {
+      "epoch": 0.3818517758160542,
+      "grad_norm": 0.08578125387430191,
+      "learning_rate": 8.738343390016456e-05,
+      "loss": 0.3483,
+      "step": 465
+    },
+    {
+      "epoch": 0.38267296243071236,
+      "grad_norm": 0.08472792059183121,
+      "learning_rate": 8.735600658255623e-05,
+      "loss": 0.3445,
+      "step": 466
+    },
+    {
+      "epoch": 0.38349414904537055,
+      "grad_norm": 0.08300035446882248,
+      "learning_rate": 8.73285792649479e-05,
+      "loss": 0.3384,
+      "step": 467
+    },
+    {
+      "epoch": 0.38431533566002873,
+      "grad_norm": 0.08582165092229843,
+      "learning_rate": 8.730115194733955e-05,
+      "loss": 0.3497,
+      "step": 468
+    },
+    {
+      "epoch": 0.3851365222746869,
+      "grad_norm": 0.28932541608810425,
+      "learning_rate": 8.727372462973121e-05,
+      "loss": 0.3091,
+      "step": 469
+    },
+    {
+      "epoch": 0.3859577088893451,
+      "grad_norm": 0.08396289497613907,
+      "learning_rate": 8.724629731212288e-05,
+      "loss": 0.3472,
+      "step": 470
+    },
+    {
+      "epoch": 0.3867788955040033,
+      "grad_norm": 0.08243526518344879,
+      "learning_rate": 8.721886999451454e-05,
+      "loss": 0.3641,
+      "step": 471
+    },
+    {
+      "epoch": 0.3876000821186615,
+      "grad_norm": 0.08014799654483795,
+      "learning_rate": 8.71914426769062e-05,
+      "loss": 0.3389,
+      "step": 472
+    },
+    {
+      "epoch": 0.38842126873331967,
+      "grad_norm": 0.08125128597021103,
+      "learning_rate": 8.716401535929787e-05,
+      "loss": 0.3293,
+      "step": 473
+    },
+    {
+      "epoch": 0.3892424553479778,
+      "grad_norm": 0.08730563521385193,
+      "learning_rate": 8.713658804168952e-05,
+      "loss": 0.3102,
+      "step": 474
+    },
+    {
+      "epoch": 0.390063641962636,
+      "grad_norm": 0.08945506811141968,
+      "learning_rate": 8.710916072408119e-05,
+      "loss": 0.3496,
+      "step": 475
+    },
+    {
+      "epoch": 0.3908848285772942,
+      "grad_norm": 0.09083441644906998,
+      "learning_rate": 8.708173340647286e-05,
+      "loss": 0.3411,
+      "step": 476
+    },
+    {
+      "epoch": 0.39170601519195236,
+      "grad_norm": 0.08667677640914917,
+      "learning_rate": 8.705430608886451e-05,
+      "loss": 0.3289,
+      "step": 477
+    },
+    {
+      "epoch": 0.39252720180661055,
+      "grad_norm": 0.08793242275714874,
+      "learning_rate": 8.702687877125617e-05,
+      "loss": 0.3542,
+      "step": 478
+    },
+    {
+      "epoch": 0.39334838842126874,
+      "grad_norm": 0.08769497275352478,
+      "learning_rate": 8.699945145364785e-05,
+      "loss": 0.3207,
+      "step": 479
+    },
+    {
+      "epoch": 0.3941695750359269,
+      "grad_norm": 0.08303678780794144,
+      "learning_rate": 8.69720241360395e-05,
+      "loss": 0.343,
+      "step": 480
+    },
+    {
+      "epoch": 0.3949907616505851,
+      "grad_norm": 0.07943850755691528,
+      "learning_rate": 8.694459681843116e-05,
+      "loss": 0.3273,
+      "step": 481
+    },
+    {
+      "epoch": 0.3958119482652433,
+      "grad_norm": 0.08450903743505478,
+      "learning_rate": 8.691716950082282e-05,
+      "loss": 0.3895,
+      "step": 482
+    },
+    {
+      "epoch": 0.3966331348799015,
+      "grad_norm": 0.07759539037942886,
+      "learning_rate": 8.688974218321448e-05,
+      "loss": 0.3426,
+      "step": 483
+    },
+    {
+      "epoch": 0.3974543214945596,
+      "grad_norm": 0.09013447165489197,
+      "learning_rate": 8.686231486560615e-05,
+      "loss": 0.336,
+      "step": 484
+    },
+    {
+      "epoch": 0.3982755081092178,
+      "grad_norm": 0.08396545797586441,
+      "learning_rate": 8.683488754799781e-05,
+      "loss": 0.3647,
+      "step": 485
+    },
+    {
+      "epoch": 0.399096694723876,
+      "grad_norm": 0.08143208920955658,
+      "learning_rate": 8.680746023038947e-05,
+      "loss": 0.3239,
+      "step": 486
+    },
+    {
+      "epoch": 0.3999178813385342,
+      "grad_norm": 0.07628796994686127,
+      "learning_rate": 8.678003291278114e-05,
+      "loss": 0.3389,
+      "step": 487
+    },
+    {
+      "epoch": 0.40073906795319236,
+      "grad_norm": 0.08144954591989517,
+      "learning_rate": 8.67526055951728e-05,
+      "loss": 0.3258,
+      "step": 488
+    },
+    {
+      "epoch": 0.40156025456785055,
+      "grad_norm": 0.08240176737308502,
+      "learning_rate": 8.672517827756446e-05,
+      "loss": 0.3316,
+      "step": 489
+    },
+    {
+      "epoch": 0.40238144118250874,
+      "grad_norm": 0.09237710386514664,
+      "learning_rate": 8.669775095995613e-05,
+      "loss": 0.3435,
+      "step": 490
+    },
+    {
+      "epoch": 0.4032026277971669,
+      "grad_norm": 0.07773042470216751,
+      "learning_rate": 8.667032364234778e-05,
+      "loss": 0.3479,
+      "step": 491
+    },
+    {
+      "epoch": 0.4040238144118251,
+      "grad_norm": 0.08410782366991043,
+      "learning_rate": 8.664289632473945e-05,
+      "loss": 0.3792,
+      "step": 492
+    },
+    {
+      "epoch": 0.40484500102648324,
+      "grad_norm": 0.08251044154167175,
+      "learning_rate": 8.661546900713112e-05,
+      "loss": 0.3388,
+      "step": 493
+    },
+    {
+      "epoch": 0.40566618764114143,
+      "grad_norm": 0.08804575353860855,
+      "learning_rate": 8.658804168952277e-05,
+      "loss": 0.3388,
+      "step": 494
+    },
+    {
+      "epoch": 0.4064873742557996,
+      "grad_norm": 0.10321494936943054,
+      "learning_rate": 8.656061437191442e-05,
+      "loss": 0.3348,
+      "step": 495
+    },
+    {
+      "epoch": 0.4073085608704578,
+      "grad_norm": 0.07642900943756104,
+      "learning_rate": 8.653318705430609e-05,
+      "loss": 0.3155,
+      "step": 496
+    },
+    {
+      "epoch": 0.408129747485116,
+      "grad_norm": 0.08629249781370163,
+      "learning_rate": 8.650575973669776e-05,
+      "loss": 0.3303,
+      "step": 497
+    },
+    {
+      "epoch": 0.4089509340997742,
+      "grad_norm": 0.07933122664690018,
+      "learning_rate": 8.647833241908941e-05,
+      "loss": 0.3399,
+      "step": 498
+    },
+    {
+      "epoch": 0.40977212071443236,
+      "grad_norm": 0.0802338495850563,
+      "learning_rate": 8.645090510148108e-05,
+      "loss": 0.3623,
+      "step": 499
+    },
+    {
+      "epoch": 0.41059330732909055,
+      "grad_norm": 0.07566798478364944,
+      "learning_rate": 8.642347778387274e-05,
+      "loss": 0.3195,
+      "step": 500
+    },
+    {
+      "epoch": 0.41141449394374874,
+      "grad_norm": 0.0790770947933197,
+      "learning_rate": 8.63960504662644e-05,
+      "loss": 0.3478,
+      "step": 501
+    },
+    {
+      "epoch": 0.4122356805584069,
+      "grad_norm": 0.08219192922115326,
+      "learning_rate": 8.636862314865607e-05,
+      "loss": 0.4037,
+      "step": 502
+    },
+    {
+      "epoch": 0.41305686717306506,
+      "grad_norm": 0.09459681063890457,
+      "learning_rate": 8.634119583104773e-05,
+      "loss": 0.3604,
+      "step": 503
+    },
+    {
+      "epoch": 0.41387805378772324,
+      "grad_norm": 0.08201157301664352,
+      "learning_rate": 8.631376851343938e-05,
+      "loss": 0.3561,
+      "step": 504
+    },
+    {
+      "epoch": 0.41469924040238143,
+      "grad_norm": 0.08181356638669968,
+      "learning_rate": 8.628634119583105e-05,
+      "loss": 0.3359,
+      "step": 505
+    },
+    {
+      "epoch": 0.4155204270170396,
+      "grad_norm": 0.08242693543434143,
+      "learning_rate": 8.625891387822272e-05,
+      "loss": 0.3501,
+      "step": 506
+    },
+    {
+      "epoch": 0.4163416136316978,
+      "grad_norm": 0.08045560866594315,
+      "learning_rate": 8.623148656061437e-05,
+      "loss": 0.3188,
+      "step": 507
+    },
+    {
+      "epoch": 0.417162800246356,
+      "grad_norm": 0.08392325043678284,
+      "learning_rate": 8.620405924300604e-05,
+      "loss": 0.3607,
+      "step": 508
+    },
+    {
+      "epoch": 0.4179839868610142,
+      "grad_norm": 0.0920289009809494,
+      "learning_rate": 8.61766319253977e-05,
+      "loss": 0.3425,
+      "step": 509
+    },
+    {
+      "epoch": 0.41880517347567237,
+      "grad_norm": 0.08921059966087341,
+      "learning_rate": 8.614920460778936e-05,
+      "loss": 0.3386,
+      "step": 510
+    },
+    {
+      "epoch": 0.41962636009033055,
+      "grad_norm": 0.08129517734050751,
+      "learning_rate": 8.612177729018103e-05,
+      "loss": 0.3631,
+      "step": 511
+    },
+    {
+      "epoch": 0.4204475467049887,
+      "grad_norm": 0.19061192870140076,
+      "learning_rate": 8.609434997257268e-05,
+      "loss": 0.3285,
+      "step": 512
+    },
+    {
+      "epoch": 0.42126873331964687,
+      "grad_norm": 0.08446541428565979,
+      "learning_rate": 8.606692265496434e-05,
+      "loss": 0.4133,
+      "step": 513
+    },
+    {
+      "epoch": 0.42208991993430506,
+      "grad_norm": 0.07793641835451126,
+      "learning_rate": 8.603949533735602e-05,
+      "loss": 0.3166,
+      "step": 514
+    },
+    {
+      "epoch": 0.42291110654896324,
+      "grad_norm": 0.07862336188554764,
+      "learning_rate": 8.601206801974767e-05,
+      "loss": 0.3448,
+      "step": 515
+    },
+    {
+      "epoch": 0.42373229316362143,
+      "grad_norm": 0.07776626199483871,
+      "learning_rate": 8.598464070213933e-05,
+      "loss": 0.3717,
+      "step": 516
+    },
+    {
+      "epoch": 0.4245534797782796,
+      "grad_norm": 0.08171868324279785,
+      "learning_rate": 8.5957213384531e-05,
+      "loss": 0.3384,
+      "step": 517
+    },
+    {
+      "epoch": 0.4253746663929378,
+      "grad_norm": 0.08469153940677643,
+      "learning_rate": 8.592978606692266e-05,
+      "loss": 0.3238,
+      "step": 518
+    },
+    {
+      "epoch": 0.426195853007596,
+      "grad_norm": 0.0739326924085617,
+      "learning_rate": 8.590235874931433e-05,
+      "loss": 0.3033,
+      "step": 519
+    },
+    {
+      "epoch": 0.4270170396222542,
+      "grad_norm": 0.08084508031606674,
+      "learning_rate": 8.587493143170599e-05,
+      "loss": 0.3557,
+      "step": 520
+    },
+    {
+      "epoch": 0.4278382262369123,
+      "grad_norm": 0.079572893679142,
+      "learning_rate": 8.584750411409764e-05,
+      "loss": 0.3369,
+      "step": 521
+    },
+    {
+      "epoch": 0.4286594128515705,
+      "grad_norm": 0.09056146442890167,
+      "learning_rate": 8.582007679648931e-05,
+      "loss": 0.3362,
+      "step": 522
+    },
+    {
+      "epoch": 0.4294805994662287,
+      "grad_norm": 0.09055227041244507,
+      "learning_rate": 8.579264947888098e-05,
+      "loss": 0.3457,
+      "step": 523
+    },
+    {
+      "epoch": 0.43030178608088687,
+      "grad_norm": 0.11645710468292236,
+      "learning_rate": 8.576522216127263e-05,
+      "loss": 0.3378,
+      "step": 524
+    },
+    {
+      "epoch": 0.43112297269554506,
+      "grad_norm": 0.08169250190258026,
+      "learning_rate": 8.57377948436643e-05,
+      "loss": 0.3406,
+      "step": 525
+    },
+    {
+      "epoch": 0.43194415931020325,
+      "grad_norm": 0.08108235895633698,
+      "learning_rate": 8.571036752605595e-05,
+      "loss": 0.3253,
+      "step": 526
+    },
+    {
+      "epoch": 0.43276534592486143,
+      "grad_norm": 0.07971398532390594,
+      "learning_rate": 8.568294020844762e-05,
+      "loss": 0.3389,
+      "step": 527
+    },
+    {
+      "epoch": 0.4335865325395196,
+      "grad_norm": 0.08281555771827698,
+      "learning_rate": 8.565551289083929e-05,
+      "loss": 0.3289,
+      "step": 528
+    },
+    {
+      "epoch": 0.4335865325395196,
+      "eval_runtime": 493.1791,
+      "eval_samples_per_second": 0.399,
+      "eval_steps_per_second": 0.201,
+      "step": 528
+    },
+    {
+      "epoch": 0.4344077191541778,
+      "grad_norm": 0.08380738645792007,
+      "learning_rate": 8.562808557323094e-05,
+      "loss": 0.3445,
+      "step": 529
+    },
+    {
+      "epoch": 0.435228905768836,
+      "grad_norm": 0.07987751066684723,
+      "learning_rate": 8.56006582556226e-05,
+      "loss": 0.3149,
+      "step": 530
+    },
+    {
+      "epoch": 0.4360500923834941,
+      "grad_norm": 0.08102700114250183,
+      "learning_rate": 8.557323093801427e-05,
+      "loss": 0.3309,
+      "step": 531
+    },
+    {
+      "epoch": 0.4368712789981523,
+      "grad_norm": 0.08056002855300903,
+      "learning_rate": 8.554580362040593e-05,
+      "loss": 0.3579,
+      "step": 532
+    },
+    {
+      "epoch": 0.4376924656128105,
+      "grad_norm": 0.08129940181970596,
+      "learning_rate": 8.551837630279759e-05,
+      "loss": 0.3414,
+      "step": 533
+    },
+    {
+      "epoch": 0.4385136522274687,
+      "grad_norm": 0.07720111310482025,
+      "learning_rate": 8.549094898518926e-05,
+      "loss": 0.3066,
+      "step": 534
+    },
+    {
+      "epoch": 0.4393348388421269,
+      "grad_norm": 0.08578766137361526,
+      "learning_rate": 8.546352166758091e-05,
+      "loss": 0.4056,
+      "step": 535
+    },
+    {
+      "epoch": 0.44015602545678506,
+      "grad_norm": 0.14550408720970154,
+      "learning_rate": 8.543609434997258e-05,
+      "loss": 0.353,
+      "step": 536
+    },
+    {
+      "epoch": 0.44097721207144325,
+      "grad_norm": 0.07624532282352448,
+      "learning_rate": 8.540866703236425e-05,
+      "loss": 0.323,
+      "step": 537
+    },
+    {
+      "epoch": 0.44179839868610143,
+      "grad_norm": 0.07774993777275085,
+      "learning_rate": 8.53812397147559e-05,
+      "loss": 0.3244,
+      "step": 538
+    },
+    {
+      "epoch": 0.4426195853007596,
+      "grad_norm": 0.08061113953590393,
+      "learning_rate": 8.535381239714755e-05,
+      "loss": 0.3209,
+      "step": 539
+    },
+    {
+      "epoch": 0.44344077191541775,
+      "grad_norm": 0.07771355658769608,
+      "learning_rate": 8.532638507953922e-05,
+      "loss": 0.3604,
+      "step": 540
+    },
+    {
+      "epoch": 0.44426195853007594,
+      "grad_norm": 0.08160841464996338,
+      "learning_rate": 8.529895776193089e-05,
+      "loss": 0.3373,
+      "step": 541
+    },
+    {
+      "epoch": 0.4450831451447341,
+      "grad_norm": 0.0819459930062294,
+      "learning_rate": 8.527153044432254e-05,
+      "loss": 0.3097,
+      "step": 542
+    },
+    {
+      "epoch": 0.4459043317593923,
+      "grad_norm": 0.08429809659719467,
+      "learning_rate": 8.524410312671421e-05,
+      "loss": 0.3538,
+      "step": 543
+    },
+    {
+      "epoch": 0.4467255183740505,
+      "grad_norm": 0.08519409596920013,
+      "learning_rate": 8.521667580910587e-05,
+      "loss": 0.3486,
+      "step": 544
+    },
+    {
+      "epoch": 0.4475467049887087,
+      "grad_norm": 0.08900930732488632,
+      "learning_rate": 8.518924849149753e-05,
+      "loss": 0.3339,
+      "step": 545
+    },
+    {
+      "epoch": 0.4483678916033669,
+      "grad_norm": 0.08914072066545486,
+      "learning_rate": 8.51618211738892e-05,
+      "loss": 0.2933,
+      "step": 546
+    },
+    {
+      "epoch": 0.44918907821802506,
+      "grad_norm": 0.08086919039487839,
+      "learning_rate": 8.513439385628086e-05,
+      "loss": 0.3367,
+      "step": 547
+    },
+    {
+      "epoch": 0.45001026483268325,
+      "grad_norm": 0.08358173817396164,
+      "learning_rate": 8.510696653867251e-05,
+      "loss": 0.3443,
+      "step": 548
+    },
+    {
+      "epoch": 0.45083145144734144,
+      "grad_norm": 0.0807274803519249,
+      "learning_rate": 8.507953922106419e-05,
+      "loss": 0.3332,
+      "step": 549
+    },
+    {
+      "epoch": 0.45165263806199957,
+      "grad_norm": 0.08126116544008255,
+      "learning_rate": 8.505211190345585e-05,
+      "loss": 0.3485,
+      "step": 550
+    },
+    {
+      "epoch": 0.45247382467665775,
+      "grad_norm": 0.08835196495056152,
+      "learning_rate": 8.50246845858475e-05,
+      "loss": 0.3739,
+      "step": 551
+    },
+    {
+      "epoch": 0.45329501129131594,
+      "grad_norm": 0.07614720612764359,
+      "learning_rate": 8.499725726823917e-05,
+      "loss": 0.3464,
+      "step": 552
+    },
+    {
+      "epoch": 0.45411619790597413,
+      "grad_norm": 0.07444674521684647,
+      "learning_rate": 8.496982995063084e-05,
+      "loss": 0.3553,
+      "step": 553
+    },
+    {
+      "epoch": 0.4549373845206323,
+      "grad_norm": 0.0741548240184784,
+      "learning_rate": 8.49424026330225e-05,
+      "loss": 0.341,
+      "step": 554
+    },
+    {
+      "epoch": 0.4557585711352905,
+      "grad_norm": 0.0820796862244606,
+      "learning_rate": 8.491497531541416e-05,
+      "loss": 0.3359,
+      "step": 555
+    },
+    {
+      "epoch": 0.4565797577499487,
+      "grad_norm": 0.08371058106422424,
+      "learning_rate": 8.488754799780581e-05,
+      "loss": 0.3339,
+      "step": 556
+    },
+    {
+      "epoch": 0.4574009443646069,
+      "grad_norm": 0.08354955166578293,
+      "learning_rate": 8.486012068019748e-05,
+      "loss": 0.3619,
+      "step": 557
+    },
+    {
+      "epoch": 0.45822213097926506,
+      "grad_norm": 0.08976832777261734,
+      "learning_rate": 8.483269336258915e-05,
+      "loss": 0.353,
+      "step": 558
+    },
+    {
+      "epoch": 0.4590433175939232,
+      "grad_norm": 0.07933076471090317,
+      "learning_rate": 8.48052660449808e-05,
+      "loss": 0.3764,
+      "step": 559
+    },
+    {
+      "epoch": 0.4598645042085814,
+      "grad_norm": 0.08624580502510071,
+      "learning_rate": 8.477783872737247e-05,
+      "loss": 0.3372,
+      "step": 560
+    },
+    {
+      "epoch": 0.46068569082323957,
+      "grad_norm": 0.08301227539777756,
+      "learning_rate": 8.475041140976413e-05,
+      "loss": 0.3748,
+      "step": 561
+    },
+    {
+      "epoch": 0.46150687743789776,
+      "grad_norm": 0.07408706098794937,
+      "learning_rate": 8.47229840921558e-05,
+      "loss": 0.2622,
+      "step": 562
+    },
+    {
+      "epoch": 0.46232806405255594,
+      "grad_norm": 0.07799118012189865,
+      "learning_rate": 8.469555677454746e-05,
+      "loss": 0.3614,
+      "step": 563
+    },
+    {
+      "epoch": 0.46314925066721413,
+      "grad_norm": 0.08185340464115143,
+      "learning_rate": 8.466812945693912e-05,
+      "loss": 0.3423,
+      "step": 564
+    },
+    {
+      "epoch": 0.4639704372818723,
+      "grad_norm": 0.0817733034491539,
+      "learning_rate": 8.464070213933077e-05,
+      "loss": 0.3519,
+      "step": 565
+    },
+    {
+      "epoch": 0.4647916238965305,
+      "grad_norm": 0.08026120066642761,
+      "learning_rate": 8.461327482172244e-05,
+      "loss": 0.3372,
+      "step": 566
+    },
+    {
+      "epoch": 0.4656128105111887,
+      "grad_norm": 0.07413862645626068,
+      "learning_rate": 8.45858475041141e-05,
+      "loss": 0.3017,
+      "step": 567
+    },
+    {
+      "epoch": 0.4664339971258468,
+      "grad_norm": 0.09256873279809952,
+      "learning_rate": 8.455842018650576e-05,
+      "loss": 0.3495,
+      "step": 568
+    },
+    {
+      "epoch": 0.467255183740505,
+      "grad_norm": 0.08319979161024094,
+      "learning_rate": 8.453099286889743e-05,
+      "loss": 0.3339,
+      "step": 569
+    },
+    {
+      "epoch": 0.4680763703551632,
+      "grad_norm": 0.08824891597032547,
+      "learning_rate": 8.450356555128908e-05,
+      "loss": 0.3272,
+      "step": 570
+    },
+    {
+      "epoch": 0.4688975569698214,
+      "grad_norm": 0.0790686085820198,
+      "learning_rate": 8.447613823368075e-05,
+      "loss": 0.3446,
+      "step": 571
+    },
+    {
+      "epoch": 0.46971874358447957,
+      "grad_norm": 0.08730736374855042,
+      "learning_rate": 8.444871091607242e-05,
+      "loss": 0.3731,
+      "step": 572
+    },
+    {
+      "epoch": 0.47053993019913776,
+      "grad_norm": 0.08023872971534729,
+      "learning_rate": 8.442128359846407e-05,
+      "loss": 0.3226,
+      "step": 573
+    },
+    {
+      "epoch": 0.47136111681379594,
+      "grad_norm": 0.09242340922355652,
+      "learning_rate": 8.439385628085573e-05,
+      "loss": 0.3229,
+      "step": 574
+    },
+    {
+      "epoch": 0.47218230342845413,
+      "grad_norm": 0.07871225476264954,
+      "learning_rate": 8.43664289632474e-05,
+      "loss": 0.3112,
+      "step": 575
+    },
+    {
+      "epoch": 0.4730034900431123,
+      "grad_norm": 0.08368773013353348,
+      "learning_rate": 8.433900164563906e-05,
+      "loss": 0.3357,
+      "step": 576
+    },
+    {
+      "epoch": 0.4738246766577705,
+      "grad_norm": 0.07561603933572769,
+      "learning_rate": 8.431157432803072e-05,
+      "loss": 0.3125,
+      "step": 577
+    },
+    {
+      "epoch": 0.47464586327242864,
+      "grad_norm": 0.08441564440727234,
+      "learning_rate": 8.428414701042238e-05,
+      "loss": 0.3211,
+      "step": 578
+    },
+    {
+      "epoch": 0.4754670498870868,
+      "grad_norm": 0.09772761911153793,
+      "learning_rate": 8.425671969281404e-05,
+      "loss": 0.3285,
+      "step": 579
+    },
+    {
+      "epoch": 0.476288236501745,
+      "grad_norm": 0.07900305837392807,
+      "learning_rate": 8.422929237520571e-05,
+      "loss": 0.344,
+      "step": 580
+    },
+    {
+      "epoch": 0.4771094231164032,
+      "grad_norm": 0.13551943004131317,
+      "learning_rate": 8.420186505759737e-05,
+      "loss": 0.3515,
+      "step": 581
+    },
+    {
+      "epoch": 0.4779306097310614,
+      "grad_norm": 0.07592803239822388,
+      "learning_rate": 8.417443773998903e-05,
+      "loss": 0.3377,
+      "step": 582
+    },
+    {
+      "epoch": 0.47875179634571957,
+      "grad_norm": 0.08301271498203278,
+      "learning_rate": 8.414701042238068e-05,
+      "loss": 0.365,
+      "step": 583
+    },
+    {
+      "epoch": 0.47957298296037776,
+      "grad_norm": 0.11750177294015884,
+      "learning_rate": 8.411958310477237e-05,
+      "loss": 0.3159,
+      "step": 584
+    },
+    {
+      "epoch": 0.48039416957503595,
+      "grad_norm": 0.08506017178297043,
+      "learning_rate": 8.409215578716402e-05,
+      "loss": 0.3381,
+      "step": 585
+    },
+    {
+      "epoch": 0.48121535618969413,
+      "grad_norm": 0.0827595517039299,
+      "learning_rate": 8.406472846955569e-05,
+      "loss": 0.3497,
+      "step": 586
+    },
+    {
+      "epoch": 0.48203654280435226,
+      "grad_norm": 0.07703883200883865,
+      "learning_rate": 8.403730115194734e-05,
+      "loss": 0.2989,
+      "step": 587
+    },
+    {
+      "epoch": 0.48285772941901045,
+      "grad_norm": 0.07976412773132324,
+      "learning_rate": 8.400987383433901e-05,
+      "loss": 0.3617,
+      "step": 588
+    },
+    {
+      "epoch": 0.48367891603366864,
+      "grad_norm": 0.09424680471420288,
+      "learning_rate": 8.398244651673068e-05,
+      "loss": 0.3623,
+      "step": 589
+    },
+    {
+      "epoch": 0.4845001026483268,
+      "grad_norm": 0.08831481635570526,
+      "learning_rate": 8.395501919912233e-05,
+      "loss": 0.3526,
+      "step": 590
+    },
+    {
+      "epoch": 0.485321289262985,
+      "grad_norm": 0.07811284810304642,
+      "learning_rate": 8.392759188151399e-05,
+      "loss": 0.3721,
+      "step": 591
+    },
+    {
+      "epoch": 0.4861424758776432,
+      "grad_norm": 0.08139461278915405,
+      "learning_rate": 8.390016456390565e-05,
+      "loss": 0.33,
+      "step": 592
+    },
+    {
+      "epoch": 0.4869636624923014,
+      "grad_norm": 0.0807470753788948,
+      "learning_rate": 8.387273724629732e-05,
+      "loss": 0.372,
+      "step": 593
+    },
+    {
+      "epoch": 0.4877848491069596,
+      "grad_norm": 0.08856850117444992,
+      "learning_rate": 8.384530992868898e-05,
+      "loss": 0.3288,
+      "step": 594
+    },
+    {
+      "epoch": 0.48860603572161776,
+      "grad_norm": 0.08670219779014587,
+      "learning_rate": 8.381788261108064e-05,
+      "loss": 0.4113,
+      "step": 595
+    },
+    {
+      "epoch": 0.4894272223362759,
+      "grad_norm": 0.0814259722828865,
+      "learning_rate": 8.37904552934723e-05,
+      "loss": 0.3735,
+      "step": 596
+    },
+    {
+      "epoch": 0.4902484089509341,
+      "grad_norm": 0.08457124978303909,
+      "learning_rate": 8.376302797586397e-05,
+      "loss": 0.3516,
+      "step": 597
+    },
+    {
+      "epoch": 0.49106959556559227,
+      "grad_norm": 0.0774846151471138,
+      "learning_rate": 8.373560065825563e-05,
+      "loss": 0.3604,
+      "step": 598
+    },
+    {
+      "epoch": 0.49189078218025045,
+      "grad_norm": 0.07951226830482483,
+      "learning_rate": 8.370817334064729e-05,
+      "loss": 0.3016,
+      "step": 599
+    },
+    {
+      "epoch": 0.49271196879490864,
+      "grad_norm": 0.08235177397727966,
+      "learning_rate": 8.368074602303894e-05,
+      "loss": 0.3502,
+      "step": 600
+    },
+    {
+      "epoch": 0.4935331554095668,
+      "grad_norm": 0.08112053573131561,
+      "learning_rate": 8.365331870543061e-05,
+      "loss": 0.3122,
+      "step": 601
+    },
+    {
+      "epoch": 0.494354342024225,
+      "grad_norm": 0.07517421990633011,
+      "learning_rate": 8.362589138782228e-05,
+      "loss": 0.3482,
+      "step": 602
+    },
+    {
+      "epoch": 0.4951755286388832,
+      "grad_norm": 0.08243374526500702,
+      "learning_rate": 8.359846407021393e-05,
+      "loss": 0.3455,
+      "step": 603
+    },
+    {
+      "epoch": 0.4959967152535414,
+      "grad_norm": 0.08028637617826462,
+      "learning_rate": 8.35710367526056e-05,
+      "loss": 0.3454,
+      "step": 604
+    },
+    {
+      "epoch": 0.4968179018681996,
+      "grad_norm": 0.08006370067596436,
+      "learning_rate": 8.354360943499726e-05,
+      "loss": 0.3216,
+      "step": 605
+    },
+    {
+      "epoch": 0.4976390884828577,
+      "grad_norm": 0.1940547525882721,
+      "learning_rate": 8.351618211738892e-05,
+      "loss": 0.3017,
+      "step": 606
+    },
+    {
+      "epoch": 0.4984602750975159,
+      "grad_norm": 0.07875222712755203,
+      "learning_rate": 8.348875479978059e-05,
+      "loss": 0.3295,
+      "step": 607
+    },
+    {
+      "epoch": 0.4992814617121741,
+      "grad_norm": 0.07511032372713089,
+      "learning_rate": 8.346132748217225e-05,
+      "loss": 0.328,
+      "step": 608
+    },
+    {
+      "epoch": 0.5001026483268323,
+      "grad_norm": 0.08197803795337677,
+      "learning_rate": 8.34339001645639e-05,
+      "loss": 0.3232,
+      "step": 609
+    },
+    {
+      "epoch": 0.5009238349414905,
+      "grad_norm": 0.08152435719966888,
+      "learning_rate": 8.340647284695558e-05,
+      "loss": 0.3546,
+      "step": 610
+    },
+    {
+      "epoch": 0.5017450215561486,
+      "grad_norm": 0.0796707347035408,
+      "learning_rate": 8.337904552934724e-05,
+      "loss": 0.3518,
+      "step": 611
+    },
+    {
+      "epoch": 0.5025662081708068,
+      "grad_norm": 0.07969588786363602,
+      "learning_rate": 8.335161821173889e-05,
+      "loss": 0.3195,
+      "step": 612
+    },
+    {
+      "epoch": 0.503387394785465,
+      "grad_norm": 0.08270443230867386,
+      "learning_rate": 8.332419089413056e-05,
+      "loss": 0.3244,
+      "step": 613
+    },
+    {
+      "epoch": 0.5042085814001231,
+      "grad_norm": 0.08303502202033997,
+      "learning_rate": 8.329676357652221e-05,
+      "loss": 0.3386,
+      "step": 614
+    },
+    {
+      "epoch": 0.5050297680147814,
+      "grad_norm": 0.07783270627260208,
+      "learning_rate": 8.326933625891388e-05,
+      "loss": 0.3046,
+      "step": 615
+    },
+    {
+      "epoch": 0.5058509546294395,
+      "grad_norm": 0.08112157136201859,
+      "learning_rate": 8.324190894130555e-05,
+      "loss": 0.3244,
+      "step": 616
+    },
+    {
+      "epoch": 0.5058509546294395,
+      "eval_runtime": 494.4481,
+      "eval_samples_per_second": 0.398,
+      "eval_steps_per_second": 0.2,
+      "step": 616
+    },
+    {
+      "epoch": 0.5066721412440978,
+      "grad_norm": 0.08969878405332565,
+      "learning_rate": 8.32144816236972e-05,
+      "loss": 0.3365,
+      "step": 617
+    },
+    {
+      "epoch": 0.5074933278587559,
+      "grad_norm": 0.08142217993736267,
+      "learning_rate": 8.318705430608886e-05,
+      "loss": 0.3413,
+      "step": 618
+    },
+    {
+      "epoch": 0.5083145144734141,
+      "grad_norm": 0.08319269865751266,
+      "learning_rate": 8.315962698848054e-05,
+      "loss": 0.3458,
+      "step": 619
+    },
+    {
+      "epoch": 0.5091357010880723,
+      "grad_norm": 0.08104688674211502,
+      "learning_rate": 8.313219967087219e-05,
+      "loss": 0.3286,
+      "step": 620
+    },
+    {
+      "epoch": 0.5099568877027304,
+      "grad_norm": 0.08447278290987015,
+      "learning_rate": 8.310477235326386e-05,
+      "loss": 0.3485,
+      "step": 621
+    },
+    {
+      "epoch": 0.5107780743173886,
+      "grad_norm": 0.08879225701093674,
+      "learning_rate": 8.307734503565551e-05,
+      "loss": 0.2986,
+      "step": 622
+    },
+    {
+      "epoch": 0.5115992609320468,
+      "grad_norm": 0.07878706604242325,
+      "learning_rate": 8.304991771804718e-05,
+      "loss": 0.3301,
+      "step": 623
+    },
+    {
+      "epoch": 0.512420447546705,
+      "grad_norm": 0.08089771121740341,
+      "learning_rate": 8.302249040043885e-05,
+      "loss": 0.3183,
+      "step": 624
+    },
+    {
+      "epoch": 0.5132416341613631,
+      "grad_norm": 0.08130493015050888,
+      "learning_rate": 8.29950630828305e-05,
+      "loss": 0.3559,
+      "step": 625
+    },
+    {
+      "epoch": 0.5140628207760214,
+      "grad_norm": 0.0880448967218399,
+      "learning_rate": 8.296763576522216e-05,
+      "loss": 0.3567,
+      "step": 626
+    },
+    {
+      "epoch": 0.5148840073906795,
+      "grad_norm": 0.087735615670681,
+      "learning_rate": 8.294020844761383e-05,
+      "loss": 0.3123,
+      "step": 627
+    },
+    {
+      "epoch": 0.5157051940053378,
+      "grad_norm": 0.08659686893224716,
+      "learning_rate": 8.29127811300055e-05,
+      "loss": 0.3212,
+      "step": 628
+    },
+    {
+      "epoch": 0.5165263806199959,
+      "grad_norm": 0.0853281244635582,
+      "learning_rate": 8.288535381239715e-05,
+      "loss": 0.3337,
+      "step": 629
+    },
+    {
+      "epoch": 0.517347567234654,
+      "grad_norm": 0.07861575484275818,
+      "learning_rate": 8.285792649478882e-05,
+      "loss": 0.3666,
+      "step": 630
+    },
+    {
+      "epoch": 0.5181687538493123,
+      "grad_norm": 0.08911813795566559,
+      "learning_rate": 8.283049917718047e-05,
+      "loss": 0.349,
+      "step": 631
+    },
+    {
+      "epoch": 0.5189899404639704,
+      "grad_norm": 0.08234114944934845,
+      "learning_rate": 8.280307185957214e-05,
+      "loss": 0.3354,
+      "step": 632
+    },
+    {
+      "epoch": 0.5198111270786286,
+      "grad_norm": 0.08694402128458023,
+      "learning_rate": 8.277564454196381e-05,
+      "loss": 0.3321,
+      "step": 633
+    },
+    {
+      "epoch": 0.5206323136932868,
+      "grad_norm": 0.08186987787485123,
+      "learning_rate": 8.274821722435546e-05,
+      "loss": 0.3595,
+      "step": 634
+    },
+    {
+      "epoch": 0.521453500307945,
+      "grad_norm": 0.08425264060497284,
+      "learning_rate": 8.272078990674712e-05,
+      "loss": 0.3641,
+      "step": 635
+    },
+    {
+      "epoch": 0.5222746869226031,
+      "grad_norm": 0.1728411465883255,
+      "learning_rate": 8.269336258913878e-05,
+      "loss": 0.3646,
+      "step": 636
+    },
+    {
+      "epoch": 0.5230958735372614,
+      "grad_norm": 0.09049440920352936,
+      "learning_rate": 8.266593527153045e-05,
+      "loss": 0.3368,
+      "step": 637
+    },
+    {
+      "epoch": 0.5239170601519195,
+      "grad_norm": 0.081216000020504,
+      "learning_rate": 8.26385079539221e-05,
+      "loss": 0.3814,
+      "step": 638
+    },
+    {
+      "epoch": 0.5247382467665777,
+      "grad_norm": 0.08497685194015503,
+      "learning_rate": 8.261108063631377e-05,
+      "loss": 0.3315,
+      "step": 639
+    },
+    {
+      "epoch": 0.5255594333812359,
+      "grad_norm": 0.08426640927791595,
+      "learning_rate": 8.258365331870543e-05,
+      "loss": 0.3366,
+      "step": 640
+    },
+    {
+      "epoch": 0.526380619995894,
+      "grad_norm": 0.08354829251766205,
+      "learning_rate": 8.25562260010971e-05,
+      "loss": 0.3439,
+      "step": 641
+    },
+    {
+      "epoch": 0.5272018066105523,
+      "grad_norm": 0.08015681803226471,
+      "learning_rate": 8.252879868348876e-05,
+      "loss": 0.3401,
+      "step": 642
+    },
+    {
+      "epoch": 0.5280229932252104,
+      "grad_norm": 0.08374017477035522,
+      "learning_rate": 8.250137136588042e-05,
+      "loss": 0.3533,
+      "step": 643
+    },
+    {
+      "epoch": 0.5288441798398686,
+      "grad_norm": 0.09182158857584,
+      "learning_rate": 8.247394404827207e-05,
+      "loss": 0.2983,
+      "step": 644
+    },
+    {
+      "epoch": 0.5296653664545268,
+      "grad_norm": 0.08268212527036667,
+      "learning_rate": 8.244651673066375e-05,
+      "loss": 0.331,
+      "step": 645
+    },
+    {
+      "epoch": 0.530486553069185,
+      "grad_norm": 0.08059857040643692,
+      "learning_rate": 8.241908941305541e-05,
+      "loss": 0.3448,
+      "step": 646
+    },
+    {
+      "epoch": 0.5313077396838431,
+      "grad_norm": 0.08811867982149124,
+      "learning_rate": 8.239166209544706e-05,
+      "loss": 0.3678,
+      "step": 647
+    },
+    {
+      "epoch": 0.5321289262985013,
+      "grad_norm": 0.07894317060709,
+      "learning_rate": 8.236423477783873e-05,
+      "loss": 0.3456,
+      "step": 648
+    },
+    {
+      "epoch": 0.5329501129131595,
+      "grad_norm": 0.07806131988763809,
+      "learning_rate": 8.23368074602304e-05,
+      "loss": 0.3286,
+      "step": 649
+    },
+    {
+      "epoch": 0.5337712995278177,
+      "grad_norm": 0.08501176536083221,
+      "learning_rate": 8.230938014262205e-05,
+      "loss": 0.3383,
+      "step": 650
+    },
+    {
+      "epoch": 0.5345924861424759,
+      "grad_norm": 0.12988948822021484,
+      "learning_rate": 8.228195282501372e-05,
+      "loss": 0.3384,
+      "step": 651
+    },
+    {
+      "epoch": 0.535413672757134,
+      "grad_norm": 0.08005869388580322,
+      "learning_rate": 8.225452550740537e-05,
+      "loss": 0.323,
+      "step": 652
+    },
+    {
+      "epoch": 0.5362348593717923,
+      "grad_norm": 0.07902330160140991,
+      "learning_rate": 8.222709818979704e-05,
+      "loss": 0.3248,
+      "step": 653
+    },
+    {
+      "epoch": 0.5370560459864504,
+      "grad_norm": 0.08319025486707687,
+      "learning_rate": 8.219967087218871e-05,
+      "loss": 0.3145,
+      "step": 654
+    },
+    {
+      "epoch": 0.5378772326011086,
+      "grad_norm": 0.08141138404607773,
+      "learning_rate": 8.217224355458037e-05,
+      "loss": 0.3105,
+      "step": 655
+    },
+    {
+      "epoch": 0.5386984192157668,
+      "grad_norm": 0.07592695206403732,
+      "learning_rate": 8.214481623697203e-05,
+      "loss": 0.3184,
+      "step": 656
+    },
+    {
+      "epoch": 0.5395196058304249,
+      "grad_norm": 0.16522720456123352,
+      "learning_rate": 8.211738891936369e-05,
+      "loss": 0.3492,
+      "step": 657
+    },
+    {
+      "epoch": 0.5403407924450832,
+      "grad_norm": 0.0965307429432869,
+      "learning_rate": 8.208996160175536e-05,
+      "loss": 0.3012,
+      "step": 658
+    },
+    {
+      "epoch": 0.5411619790597413,
+      "grad_norm": 0.08132067322731018,
+      "learning_rate": 8.206253428414702e-05,
+      "loss": 0.3256,
+      "step": 659
+    },
+    {
+      "epoch": 0.5419831656743995,
+      "grad_norm": 0.08183111250400543,
+      "learning_rate": 8.203510696653868e-05,
+      "loss": 0.3591,
+      "step": 660
+    },
+    {
+      "epoch": 0.5428043522890577,
+      "grad_norm": 0.07674401253461838,
+      "learning_rate": 8.200767964893033e-05,
+      "loss": 0.3273,
+      "step": 661
+    },
+    {
+      "epoch": 0.5436255389037159,
+      "grad_norm": 0.07937440276145935,
+      "learning_rate": 8.1980252331322e-05,
+      "loss": 0.3028,
+      "step": 662
+    },
+    {
+      "epoch": 0.544446725518374,
+      "grad_norm": 0.07886291295289993,
+      "learning_rate": 8.195282501371367e-05,
+      "loss": 0.3452,
+      "step": 663
+    },
+    {
+      "epoch": 0.5452679121330323,
+      "grad_norm": 0.078894704580307,
+      "learning_rate": 8.192539769610532e-05,
+      "loss": 0.3447,
+      "step": 664
+    },
+    {
+      "epoch": 0.5460890987476904,
+      "grad_norm": 0.0840119943022728,
+      "learning_rate": 8.189797037849699e-05,
+      "loss": 0.3348,
+      "step": 665
+    },
+    {
+      "epoch": 0.5469102853623486,
+      "grad_norm": 0.08318603038787842,
+      "learning_rate": 8.187054306088864e-05,
+      "loss": 0.3523,
+      "step": 666
+    },
+    {
+      "epoch": 0.5477314719770068,
+      "grad_norm": 0.08060989528894424,
+      "learning_rate": 8.184311574328031e-05,
+      "loss": 0.3211,
+      "step": 667
+    },
+    {
+      "epoch": 0.5485526585916649,
+      "grad_norm": 0.07878737151622772,
+      "learning_rate": 8.181568842567198e-05,
+      "loss": 0.3865,
+      "step": 668
+    },
+    {
+      "epoch": 0.5493738452063232,
+      "grad_norm": 0.08277317881584167,
+      "learning_rate": 8.178826110806363e-05,
+      "loss": 0.3507,
+      "step": 669
+    },
+    {
+      "epoch": 0.5501950318209813,
+      "grad_norm": 0.15883035957813263,
+      "learning_rate": 8.176083379045529e-05,
+      "loss": 0.3428,
+      "step": 670
+    },
+    {
+      "epoch": 0.5510162184356395,
+      "grad_norm": 0.08117318898439407,
+      "learning_rate": 8.173340647284696e-05,
+      "loss": 0.3297,
+      "step": 671
+    },
+    {
+      "epoch": 0.5518374050502977,
+      "grad_norm": 0.07886844873428345,
+      "learning_rate": 8.170597915523862e-05,
+      "loss": 0.3489,
+      "step": 672
+    },
+    {
+      "epoch": 0.5526585916649559,
+      "grad_norm": 0.0774691253900528,
+      "learning_rate": 8.167855183763028e-05,
+      "loss": 0.3438,
+      "step": 673
+    },
+    {
+      "epoch": 0.553479778279614,
+      "grad_norm": 0.07944227755069733,
+      "learning_rate": 8.165112452002195e-05,
+      "loss": 0.3183,
+      "step": 674
+    },
+    {
+      "epoch": 0.5543009648942723,
+      "grad_norm": 0.08308850973844528,
+      "learning_rate": 8.16236972024136e-05,
+      "loss": 0.3268,
+      "step": 675
+    },
+    {
+      "epoch": 0.5551221515089304,
+      "grad_norm": 0.07913679629564285,
+      "learning_rate": 8.159626988480527e-05,
+      "loss": 0.3358,
+      "step": 676
+    },
+    {
+      "epoch": 0.5559433381235885,
+      "grad_norm": 0.08472824096679688,
+      "learning_rate": 8.156884256719694e-05,
+      "loss": 0.3619,
+      "step": 677
+    },
+    {
+      "epoch": 0.5567645247382468,
+      "grad_norm": 0.07870602607727051,
+      "learning_rate": 8.154141524958859e-05,
+      "loss": 0.3372,
+      "step": 678
+    },
+    {
+      "epoch": 0.5575857113529049,
+      "grad_norm": 0.07889708876609802,
+      "learning_rate": 8.151398793198025e-05,
+      "loss": 0.34,
+      "step": 679
+    },
+    {
+      "epoch": 0.5584068979675632,
+      "grad_norm": 0.0856415405869484,
+      "learning_rate": 8.148656061437193e-05,
+      "loss": 0.3132,
+      "step": 680
+    },
+    {
+      "epoch": 0.5592280845822213,
+      "grad_norm": 0.0840810164809227,
+      "learning_rate": 8.145913329676358e-05,
+      "loss": 0.3451,
+      "step": 681
+    },
+    {
+      "epoch": 0.5600492711968795,
+      "grad_norm": 0.07759146392345428,
+      "learning_rate": 8.143170597915524e-05,
+      "loss": 0.3452,
+      "step": 682
+    },
+    {
+      "epoch": 0.5608704578115377,
+      "grad_norm": 0.08038268238306046,
+      "learning_rate": 8.14042786615469e-05,
+      "loss": 0.3103,
+      "step": 683
+    },
+    {
+      "epoch": 0.5616916444261959,
+      "grad_norm": 0.07695576548576355,
+      "learning_rate": 8.137685134393857e-05,
+      "loss": 0.342,
+      "step": 684
+    },
+    {
+      "epoch": 0.562512831040854,
+      "grad_norm": 0.08458397537469864,
+      "learning_rate": 8.134942402633024e-05,
+      "loss": 0.344,
+      "step": 685
+    },
+    {
+      "epoch": 0.5633340176555122,
+      "grad_norm": 0.08124341070652008,
+      "learning_rate": 8.13219967087219e-05,
+      "loss": 0.3235,
+      "step": 686
+    },
+    {
+      "epoch": 0.5641552042701704,
+      "grad_norm": 0.09097409248352051,
+      "learning_rate": 8.129456939111355e-05,
+      "loss": 0.3142,
+      "step": 687
+    },
+    {
+      "epoch": 0.5649763908848285,
+      "grad_norm": 0.07753387093544006,
+      "learning_rate": 8.126714207350522e-05,
+      "loss": 0.3295,
+      "step": 688
+    },
+    {
+      "epoch": 0.5657975774994868,
+      "grad_norm": 0.07960804551839828,
+      "learning_rate": 8.123971475589688e-05,
+      "loss": 0.3156,
+      "step": 689
+    },
+    {
+      "epoch": 0.5666187641141449,
+      "grad_norm": 0.0779690146446228,
+      "learning_rate": 8.121228743828854e-05,
+      "loss": 0.3224,
+      "step": 690
+    },
+    {
+      "epoch": 0.5674399507288032,
+      "grad_norm": 0.07964863628149033,
+      "learning_rate": 8.11848601206802e-05,
+      "loss": 0.3846,
+      "step": 691
+    },
+    {
+      "epoch": 0.5682611373434613,
+      "grad_norm": 0.08143079280853271,
+      "learning_rate": 8.115743280307186e-05,
+      "loss": 0.3268,
+      "step": 692
+    },
+    {
+      "epoch": 0.5690823239581195,
+      "grad_norm": 0.08909327536821365,
+      "learning_rate": 8.113000548546353e-05,
+      "loss": 0.3354,
+      "step": 693
+    },
+    {
+      "epoch": 0.5699035105727777,
+      "grad_norm": 0.07217192649841309,
+      "learning_rate": 8.11025781678552e-05,
+      "loss": 0.3077,
+      "step": 694
+    },
+    {
+      "epoch": 0.5707246971874358,
+      "grad_norm": 0.07756021618843079,
+      "learning_rate": 8.107515085024685e-05,
+      "loss": 0.3244,
+      "step": 695
+    },
+    {
+      "epoch": 0.571545883802094,
+      "grad_norm": 0.08227197080850601,
+      "learning_rate": 8.10477235326385e-05,
+      "loss": 0.34,
+      "step": 696
+    },
+    {
+      "epoch": 0.5723670704167522,
+      "grad_norm": 0.07765581458806992,
+      "learning_rate": 8.102029621503017e-05,
+      "loss": 0.2898,
+      "step": 697
+    },
+    {
+      "epoch": 0.5731882570314104,
+      "grad_norm": 0.07710105925798416,
+      "learning_rate": 8.099286889742184e-05,
+      "loss": 0.3056,
+      "step": 698
+    },
+    {
+      "epoch": 0.5740094436460685,
+      "grad_norm": 0.07832709699869156,
+      "learning_rate": 8.09654415798135e-05,
+      "loss": 0.3338,
+      "step": 699
+    },
+    {
+      "epoch": 0.5748306302607268,
+      "grad_norm": 0.08406300097703934,
+      "learning_rate": 8.093801426220516e-05,
+      "loss": 0.3483,
+      "step": 700
+    },
+    {
+      "epoch": 0.5756518168753849,
+      "grad_norm": 0.07969654351472855,
+      "learning_rate": 8.091058694459682e-05,
+      "loss": 0.3645,
+      "step": 701
+    },
+    {
+      "epoch": 0.5764730034900432,
+      "grad_norm": 0.07744769752025604,
+      "learning_rate": 8.088315962698848e-05,
+      "loss": 0.3324,
+      "step": 702
+    },
+    {
+      "epoch": 0.5772941901047013,
+      "grad_norm": 0.07306936383247375,
+      "learning_rate": 8.085573230938015e-05,
+      "loss": 0.3011,
+      "step": 703
+    },
+    {
+      "epoch": 0.5781153767193594,
+      "grad_norm": 0.09406828135251999,
+      "learning_rate": 8.082830499177181e-05,
+      "loss": 0.3518,
+      "step": 704
+    },
+    {
+      "epoch": 0.5781153767193594,
+      "eval_runtime": 494.5472,
+      "eval_samples_per_second": 0.398,
+      "eval_steps_per_second": 0.2,
+      "step": 704
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3651,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 88,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.1068114755114435e+19,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}