{
  "best_metric": 3.130511999130249,
  "best_model_checkpoint": "miner_id_24/checkpoint-500",
  "epoch": 0.10231749117511639,
  "eval_steps": 100,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00020463498235023277,
      "grad_norm": 17.940895080566406,
      "learning_rate": 5e-06,
      "loss": 14.0433,
      "step": 1
    },
    {
      "epoch": 0.00020463498235023277,
      "eval_loss": 3.9562315940856934,
      "eval_runtime": 47.1989,
      "eval_samples_per_second": 174.368,
      "eval_steps_per_second": 43.603,
      "step": 1
    },
    {
      "epoch": 0.00040926996470046554,
      "grad_norm": 22.070707321166992,
      "learning_rate": 1e-05,
      "loss": 14.7688,
      "step": 2
    },
    {
      "epoch": 0.0006139049470506984,
      "grad_norm": 20.095142364501953,
      "learning_rate": 1.5e-05,
      "loss": 14.6774,
      "step": 3
    },
    {
      "epoch": 0.0008185399294009311,
      "grad_norm": 17.93838119506836,
      "learning_rate": 2e-05,
      "loss": 14.672,
      "step": 4
    },
    {
      "epoch": 0.001023174911751164,
      "grad_norm": 13.304869651794434,
      "learning_rate": 2.5e-05,
      "loss": 14.7464,
      "step": 5
    },
    {
      "epoch": 0.0012278098941013967,
      "grad_norm": 10.018509864807129,
      "learning_rate": 3e-05,
      "loss": 14.7743,
      "step": 6
    },
    {
      "epoch": 0.0014324448764516293,
      "grad_norm": 8.531241416931152,
      "learning_rate": 3.5e-05,
      "loss": 14.7664,
      "step": 7
    },
    {
      "epoch": 0.0016370798588018621,
      "grad_norm": 8.279265403747559,
      "learning_rate": 4e-05,
      "loss": 14.8116,
      "step": 8
    },
    {
      "epoch": 0.001841714841152095,
      "grad_norm": 7.955838680267334,
      "learning_rate": 4.5e-05,
      "loss": 14.0356,
      "step": 9
    },
    {
      "epoch": 0.002046349823502328,
      "grad_norm": 8.396589279174805,
      "learning_rate": 5e-05,
      "loss": 14.4371,
      "step": 10
    },
    {
      "epoch": 0.0022509848058525606,
      "grad_norm": 7.477896213531494,
      "learning_rate": 5.500000000000001e-05,
      "loss": 14.3411,
      "step": 11
    },
    {
      "epoch": 0.0024556197882027934,
      "grad_norm": 8.143397331237793,
      "learning_rate": 6e-05,
      "loss": 14.5638,
      "step": 12
    },
    {
      "epoch": 0.0026602547705530263,
      "grad_norm": 7.183845043182373,
      "learning_rate": 6.500000000000001e-05,
      "loss": 14.4296,
      "step": 13
    },
    {
      "epoch": 0.0028648897529032586,
      "grad_norm": 7.372915267944336,
      "learning_rate": 7e-05,
      "loss": 14.7023,
      "step": 14
    },
    {
      "epoch": 0.0030695247352534915,
      "grad_norm": 8.364891052246094,
      "learning_rate": 7.500000000000001e-05,
      "loss": 14.6036,
      "step": 15
    },
    {
      "epoch": 0.0032741597176037243,
      "grad_norm": 8.31488037109375,
      "learning_rate": 8e-05,
      "loss": 14.7258,
      "step": 16
    },
    {
      "epoch": 0.003478794699953957,
      "grad_norm": 7.918307304382324,
      "learning_rate": 8.5e-05,
      "loss": 14.2108,
      "step": 17
    },
    {
      "epoch": 0.00368342968230419,
      "grad_norm": 7.043214797973633,
      "learning_rate": 9e-05,
      "loss": 14.4846,
      "step": 18
    },
    {
      "epoch": 0.0038880646646544227,
      "grad_norm": 7.622882843017578,
      "learning_rate": 9.5e-05,
      "loss": 14.7337,
      "step": 19
    },
    {
      "epoch": 0.004092699647004656,
      "grad_norm": 7.523791790008545,
      "learning_rate": 0.0001,
      "loss": 13.6743,
      "step": 20
    },
    {
      "epoch": 0.004297334629354888,
      "grad_norm": 7.598966598510742,
      "learning_rate": 9.999892908320647e-05,
      "loss": 13.6939,
      "step": 21
    },
    {
      "epoch": 0.004501969611705121,
      "grad_norm": 7.914957046508789,
      "learning_rate": 9.999571637870036e-05,
      "loss": 14.0858,
      "step": 22
    },
    {
      "epoch": 0.004706604594055354,
      "grad_norm": 8.37746810913086,
      "learning_rate": 9.999036202410325e-05,
      "loss": 14.1071,
      "step": 23
    },
    {
      "epoch": 0.004911239576405587,
      "grad_norm": 8.19976806640625,
      "learning_rate": 9.998286624877786e-05,
      "loss": 14.4376,
      "step": 24
    },
    {
      "epoch": 0.00511587455875582,
      "grad_norm": 8.429183959960938,
      "learning_rate": 9.997322937381829e-05,
      "loss": 13.8618,
      "step": 25
    },
    {
      "epoch": 0.0053205095411060525,
      "grad_norm": 8.871038436889648,
      "learning_rate": 9.996145181203615e-05,
      "loss": 14.0325,
      "step": 26
    },
    {
      "epoch": 0.0055251445234562845,
      "grad_norm": 9.305150985717773,
      "learning_rate": 9.994753406794301e-05,
      "loss": 14.309,
      "step": 27
    },
    {
      "epoch": 0.005729779505806517,
      "grad_norm": 8.649129867553711,
      "learning_rate": 9.99314767377287e-05,
      "loss": 14.174,
      "step": 28
    },
    {
      "epoch": 0.00593441448815675,
      "grad_norm": 9.096344947814941,
      "learning_rate": 9.991328050923581e-05,
      "loss": 14.1257,
      "step": 29
    },
    {
      "epoch": 0.006139049470506983,
      "grad_norm": 9.188639640808105,
      "learning_rate": 9.989294616193017e-05,
      "loss": 13.7195,
      "step": 30
    },
    {
      "epoch": 0.006343684452857216,
      "grad_norm": 9.031623840332031,
      "learning_rate": 9.98704745668676e-05,
      "loss": 13.2224,
      "step": 31
    },
    {
      "epoch": 0.0065483194352074486,
      "grad_norm": 10.63833999633789,
      "learning_rate": 9.98458666866564e-05,
      "loss": 14.6318,
      "step": 32
    },
    {
      "epoch": 0.006752954417557681,
      "grad_norm": 9.114713668823242,
      "learning_rate": 9.981912357541627e-05,
      "loss": 12.9636,
      "step": 33
    },
    {
      "epoch": 0.006957589399907914,
      "grad_norm": 10.119890213012695,
      "learning_rate": 9.97902463787331e-05,
      "loss": 14.3062,
      "step": 34
    },
    {
      "epoch": 0.007162224382258147,
      "grad_norm": 9.861045837402344,
      "learning_rate": 9.975923633360985e-05,
      "loss": 14.4585,
      "step": 35
    },
    {
      "epoch": 0.00736685936460838,
      "grad_norm": 9.876028060913086,
      "learning_rate": 9.972609476841367e-05,
      "loss": 13.4113,
      "step": 36
    },
    {
      "epoch": 0.007571494346958613,
      "grad_norm": 9.652708053588867,
      "learning_rate": 9.969082310281891e-05,
      "loss": 13.4341,
      "step": 37
    },
    {
      "epoch": 0.0077761293293088455,
      "grad_norm": 10.080284118652344,
      "learning_rate": 9.965342284774632e-05,
      "loss": 13.4885,
      "step": 38
    },
    {
      "epoch": 0.007980764311659078,
      "grad_norm": 10.708706855773926,
      "learning_rate": 9.961389560529836e-05,
      "loss": 13.8326,
      "step": 39
    },
    {
      "epoch": 0.008185399294009311,
      "grad_norm": 10.825308799743652,
      "learning_rate": 9.957224306869053e-05,
      "loss": 14.1761,
      "step": 40
    },
    {
      "epoch": 0.008390034276359544,
      "grad_norm": 11.960084915161133,
      "learning_rate": 9.952846702217886e-05,
      "loss": 13.7566,
      "step": 41
    },
    {
      "epoch": 0.008594669258709777,
      "grad_norm": 13.216733932495117,
      "learning_rate": 9.948256934098352e-05,
      "loss": 13.3996,
      "step": 42
    },
    {
      "epoch": 0.00879930424106001,
      "grad_norm": 11.897379875183105,
      "learning_rate": 9.943455199120837e-05,
      "loss": 13.2895,
      "step": 43
    },
    {
      "epoch": 0.009003939223410242,
      "grad_norm": 13.036720275878906,
      "learning_rate": 9.938441702975689e-05,
      "loss": 13.2497,
      "step": 44
    },
    {
      "epoch": 0.009208574205760475,
      "grad_norm": 12.781022071838379,
      "learning_rate": 9.933216660424395e-05,
      "loss": 13.1251,
      "step": 45
    },
    {
      "epoch": 0.009413209188110708,
      "grad_norm": 13.262389183044434,
      "learning_rate": 9.927780295290389e-05,
      "loss": 13.5698,
      "step": 46
    },
    {
      "epoch": 0.009617844170460941,
      "grad_norm": 15.804819107055664,
      "learning_rate": 9.922132840449459e-05,
      "loss": 14.0475,
      "step": 47
    },
    {
      "epoch": 0.009822479152811174,
      "grad_norm": 17.065067291259766,
      "learning_rate": 9.916274537819775e-05,
      "loss": 13.7782,
      "step": 48
    },
    {
      "epoch": 0.010027114135161407,
      "grad_norm": 19.974628448486328,
      "learning_rate": 9.91020563835152e-05,
      "loss": 13.1872,
      "step": 49
    },
    {
      "epoch": 0.01023174911751164,
      "grad_norm": 33.5842399597168,
      "learning_rate": 9.903926402016153e-05,
      "loss": 14.529,
      "step": 50
    },
    {
      "epoch": 0.010436384099861872,
      "grad_norm": 27.80715560913086,
      "learning_rate": 9.897437097795257e-05,
      "loss": 15.1481,
      "step": 51
    },
    {
      "epoch": 0.010641019082212105,
      "grad_norm": 25.08864974975586,
      "learning_rate": 9.890738003669029e-05,
      "loss": 15.3311,
      "step": 52
    },
    {
      "epoch": 0.010845654064562336,
      "grad_norm": 20.487937927246094,
      "learning_rate": 9.883829406604363e-05,
      "loss": 14.983,
      "step": 53
    },
    {
      "epoch": 0.011050289046912569,
      "grad_norm": 18.901628494262695,
      "learning_rate": 9.876711602542563e-05,
      "loss": 15.1327,
      "step": 54
    },
    {
      "epoch": 0.011254924029262802,
      "grad_norm": 14.395322799682617,
      "learning_rate": 9.869384896386668e-05,
      "loss": 14.1538,
      "step": 55
    },
    {
      "epoch": 0.011459559011613035,
      "grad_norm": 10.616168022155762,
      "learning_rate": 9.861849601988383e-05,
      "loss": 14.1126,
      "step": 56
    },
    {
      "epoch": 0.011664193993963267,
      "grad_norm": 8.293721199035645,
      "learning_rate": 9.854106042134641e-05,
      "loss": 13.6647,
      "step": 57
    },
    {
      "epoch": 0.0118688289763135,
      "grad_norm": 7.054158687591553,
      "learning_rate": 9.846154548533773e-05,
      "loss": 13.3569,
      "step": 58
    },
    {
      "epoch": 0.012073463958663733,
      "grad_norm": 7.207637310028076,
      "learning_rate": 9.837995461801299e-05,
      "loss": 13.5518,
      "step": 59
    },
    {
      "epoch": 0.012278098941013966,
      "grad_norm": 7.297451496124268,
      "learning_rate": 9.829629131445342e-05,
      "loss": 13.9634,
      "step": 60
    },
    {
      "epoch": 0.012482733923364199,
      "grad_norm": 7.142579555511475,
      "learning_rate": 9.821055915851647e-05,
      "loss": 13.9129,
      "step": 61
    },
    {
      "epoch": 0.012687368905714431,
      "grad_norm": 6.920661926269531,
      "learning_rate": 9.812276182268236e-05,
      "loss": 13.7893,
      "step": 62
    },
    {
      "epoch": 0.012892003888064664,
      "grad_norm": 7.224207401275635,
      "learning_rate": 9.803290306789676e-05,
      "loss": 13.4208,
      "step": 63
    },
    {
      "epoch": 0.013096638870414897,
      "grad_norm": 6.482351779937744,
      "learning_rate": 9.794098674340965e-05,
      "loss": 14.011,
      "step": 64
    },
    {
      "epoch": 0.01330127385276513,
      "grad_norm": 5.605070114135742,
      "learning_rate": 9.784701678661045e-05,
      "loss": 13.362,
      "step": 65
    },
    {
      "epoch": 0.013505908835115363,
      "grad_norm": 6.630495548248291,
      "learning_rate": 9.775099722285935e-05,
      "loss": 13.4117,
      "step": 66
    },
    {
      "epoch": 0.013710543817465596,
      "grad_norm": 6.491345405578613,
      "learning_rate": 9.765293216531486e-05,
      "loss": 13.6704,
      "step": 67
    },
    {
      "epoch": 0.013915178799815828,
      "grad_norm": 6.191098690032959,
      "learning_rate": 9.755282581475769e-05,
      "loss": 13.405,
      "step": 68
    },
    {
      "epoch": 0.014119813782166061,
      "grad_norm": 5.732306480407715,
      "learning_rate": 9.74506824594107e-05,
      "loss": 13.2353,
      "step": 69
    },
    {
      "epoch": 0.014324448764516294,
      "grad_norm": 5.888903617858887,
      "learning_rate": 9.73465064747553e-05,
      "loss": 13.3435,
      "step": 70
    },
    {
      "epoch": 0.014529083746866527,
      "grad_norm": 5.951428413391113,
      "learning_rate": 9.724030232334391e-05,
      "loss": 13.6627,
      "step": 71
    },
    {
      "epoch": 0.01473371872921676,
      "grad_norm": 6.11599063873291,
      "learning_rate": 9.713207455460894e-05,
      "loss": 13.5377,
      "step": 72
    },
    {
      "epoch": 0.014938353711566993,
      "grad_norm": 6.475672245025635,
      "learning_rate": 9.702182780466775e-05,
      "loss": 13.5349,
      "step": 73
    },
    {
      "epoch": 0.015142988693917225,
      "grad_norm": 6.255554676055908,
      "learning_rate": 9.690956679612421e-05,
      "loss": 13.427,
      "step": 74
    },
    {
      "epoch": 0.015347623676267458,
      "grad_norm": 6.450949668884277,
      "learning_rate": 9.67952963378663e-05,
      "loss": 13.1908,
      "step": 75
    },
    {
      "epoch": 0.015552258658617691,
      "grad_norm": 6.359993934631348,
      "learning_rate": 9.667902132486009e-05,
      "loss": 13.6013,
      "step": 76
    },
    {
      "epoch": 0.015756893640967922,
      "grad_norm": 6.952643394470215,
      "learning_rate": 9.656074673794018e-05,
      "loss": 12.9901,
      "step": 77
    },
    {
      "epoch": 0.015961528623318157,
      "grad_norm": 6.661192417144775,
      "learning_rate": 9.644047764359622e-05,
      "loss": 13.4725,
      "step": 78
    },
    {
      "epoch": 0.016166163605668388,
      "grad_norm": 6.839961051940918,
      "learning_rate": 9.631821919375591e-05,
      "loss": 13.0454,
      "step": 79
    },
    {
      "epoch": 0.016370798588018622,
      "grad_norm": 7.225835800170898,
      "learning_rate": 9.619397662556435e-05,
      "loss": 13.5887,
      "step": 80
    },
    {
      "epoch": 0.016575433570368853,
      "grad_norm": 7.208718776702881,
      "learning_rate": 9.606775526115963e-05,
      "loss": 12.9486,
      "step": 81
    },
    {
      "epoch": 0.016780068552719088,
      "grad_norm": 7.323213577270508,
      "learning_rate": 9.593956050744492e-05,
      "loss": 13.15,
      "step": 82
    },
    {
      "epoch": 0.01698470353506932,
      "grad_norm": 7.82949161529541,
      "learning_rate": 9.580939785585681e-05,
      "loss": 13.2066,
      "step": 83
    },
    {
      "epoch": 0.017189338517419554,
      "grad_norm": 7.900018692016602,
      "learning_rate": 9.567727288213005e-05,
      "loss": 13.556,
      "step": 84
    },
    {
      "epoch": 0.017393973499769785,
      "grad_norm": 8.467442512512207,
      "learning_rate": 9.554319124605879e-05,
      "loss": 13.7811,
      "step": 85
    },
    {
      "epoch": 0.01759860848212002,
      "grad_norm": 8.75790786743164,
      "learning_rate": 9.540715869125407e-05,
      "loss": 13.5011,
      "step": 86
    },
    {
      "epoch": 0.01780324346447025,
      "grad_norm": 8.657136917114258,
      "learning_rate": 9.526918104489777e-05,
      "loss": 12.9077,
      "step": 87
    },
    {
      "epoch": 0.018007878446820485,
      "grad_norm": 8.441506385803223,
      "learning_rate": 9.512926421749304e-05,
      "loss": 13.4052,
      "step": 88
    },
    {
      "epoch": 0.018212513429170716,
      "grad_norm": 8.36744499206543,
      "learning_rate": 9.498741420261108e-05,
      "loss": 12.8423,
      "step": 89
    },
    {
      "epoch": 0.01841714841152095,
      "grad_norm": 9.058586120605469,
      "learning_rate": 9.484363707663442e-05,
      "loss": 13.437,
      "step": 90
    },
    {
      "epoch": 0.01862178339387118,
      "grad_norm": 9.995050430297852,
      "learning_rate": 9.469793899849661e-05,
      "loss": 13.485,
      "step": 91
    },
    {
      "epoch": 0.018826418376221416,
      "grad_norm": 9.687874794006348,
      "learning_rate": 9.45503262094184e-05,
      "loss": 12.3617,
      "step": 92
    },
    {
      "epoch": 0.019031053358571647,
      "grad_norm": 9.374463081359863,
      "learning_rate": 9.440080503264037e-05,
      "loss": 12.6718,
      "step": 93
    },
    {
      "epoch": 0.019235688340921882,
      "grad_norm": 10.396638870239258,
      "learning_rate": 9.42493818731521e-05,
      "loss": 12.5755,
      "step": 94
    },
    {
      "epoch": 0.019440323323272113,
      "grad_norm": 11.383647918701172,
      "learning_rate": 9.409606321741775e-05,
      "loss": 12.4185,
      "step": 95
    },
    {
      "epoch": 0.019644958305622347,
      "grad_norm": 10.607769012451172,
      "learning_rate": 9.394085563309827e-05,
      "loss": 12.5911,
      "step": 96
    },
    {
      "epoch": 0.01984959328797258,
      "grad_norm": 13.467524528503418,
      "learning_rate": 9.378376576876999e-05,
      "loss": 14.1048,
      "step": 97
    },
    {
      "epoch": 0.020054228270322813,
      "grad_norm": 14.150450706481934,
      "learning_rate": 9.362480035363986e-05,
      "loss": 12.6339,
      "step": 98
    },
    {
      "epoch": 0.020258863252673044,
      "grad_norm": 15.242779731750488,
      "learning_rate": 9.34639661972572e-05,
      "loss": 13.5117,
      "step": 99
    },
    {
      "epoch": 0.02046349823502328,
      "grad_norm": 21.984615325927734,
      "learning_rate": 9.330127018922194e-05,
      "loss": 13.9761,
      "step": 100
    },
    {
      "epoch": 0.02046349823502328,
      "eval_loss": 3.370993137359619,
      "eval_runtime": 47.2126,
      "eval_samples_per_second": 174.318,
      "eval_steps_per_second": 43.59,
      "step": 100
    },
    {
      "epoch": 0.02066813321737351,
      "grad_norm": 11.214513778686523,
      "learning_rate": 9.31367192988896e-05,
      "loss": 14.0912,
      "step": 101
    },
    {
      "epoch": 0.020872768199723744,
      "grad_norm": 10.00228214263916,
      "learning_rate": 9.297032057507264e-05,
      "loss": 13.9889,
      "step": 102
    },
    {
      "epoch": 0.021077403182073975,
      "grad_norm": 10.31716251373291,
      "learning_rate": 9.280208114573859e-05,
      "loss": 14.1193,
      "step": 103
    },
    {
      "epoch": 0.02128203816442421,
      "grad_norm": 9.314844131469727,
      "learning_rate": 9.263200821770461e-05,
      "loss": 13.7396,
      "step": 104
    },
    {
      "epoch": 0.02148667314677444,
      "grad_norm": 7.567698001861572,
      "learning_rate": 9.246010907632895e-05,
      "loss": 13.7879,
      "step": 105
    },
    {
      "epoch": 0.021691308129124672,
      "grad_norm": 6.531108856201172,
      "learning_rate": 9.228639108519868e-05,
      "loss": 13.5185,
      "step": 106
    },
    {
      "epoch": 0.021895943111474907,
      "grad_norm": 5.202017307281494,
      "learning_rate": 9.211086168581433e-05,
      "loss": 13.3491,
      "step": 107
    },
    {
      "epoch": 0.022100578093825138,
      "grad_norm": 4.530038356781006,
      "learning_rate": 9.193352839727121e-05,
      "loss": 13.2004,
      "step": 108
    },
    {
      "epoch": 0.022305213076175372,
      "grad_norm": 4.831387996673584,
      "learning_rate": 9.175439881593716e-05,
      "loss": 13.4205,
      "step": 109
    },
    {
      "epoch": 0.022509848058525603,
      "grad_norm": 4.692884922027588,
      "learning_rate": 9.157348061512727e-05,
      "loss": 13.4912,
      "step": 110
    },
    {
      "epoch": 0.022714483040875838,
      "grad_norm": 5.204988479614258,
      "learning_rate": 9.139078154477512e-05,
      "loss": 13.1214,
      "step": 111
    },
    {
      "epoch": 0.02291911802322607,
      "grad_norm": 4.781569004058838,
      "learning_rate": 9.120630943110077e-05,
      "loss": 12.6118,
      "step": 112
    },
    {
      "epoch": 0.023123753005576304,
      "grad_norm": 4.754026412963867,
      "learning_rate": 9.102007217627568e-05,
      "loss": 13.186,
      "step": 113
    },
    {
      "epoch": 0.023328387987926535,
      "grad_norm": 5.035665035247803,
      "learning_rate": 9.083207775808396e-05,
      "loss": 12.7322,
      "step": 114
    },
    {
      "epoch": 0.02353302297027677,
      "grad_norm": 5.12575626373291,
      "learning_rate": 9.064233422958077e-05,
      "loss": 13.0182,
      "step": 115
    },
    {
      "epoch": 0.023737657952627,
      "grad_norm": 5.39860200881958,
      "learning_rate": 9.045084971874738e-05,
      "loss": 13.4676,
      "step": 116
    },
    {
      "epoch": 0.023942292934977235,
      "grad_norm": 5.005839824676514,
      "learning_rate": 9.025763242814291e-05,
      "loss": 13.0532,
      "step": 117
    },
    {
      "epoch": 0.024146927917327466,
      "grad_norm": 5.046457290649414,
      "learning_rate": 9.006269063455304e-05,
      "loss": 13.336,
      "step": 118
    },
    {
      "epoch": 0.0243515628996777,
      "grad_norm": 4.951815128326416,
      "learning_rate": 8.986603268863536e-05,
      "loss": 13.1308,
      "step": 119
    },
    {
      "epoch": 0.02455619788202793,
      "grad_norm": 5.16800594329834,
      "learning_rate": 8.966766701456177e-05,
      "loss": 12.7553,
      "step": 120
    },
    {
      "epoch": 0.024760832864378166,
      "grad_norm": 5.190509796142578,
      "learning_rate": 8.94676021096575e-05,
      "loss": 13.2239,
      "step": 121
    },
    {
      "epoch": 0.024965467846728397,
      "grad_norm": 5.662418365478516,
      "learning_rate": 8.926584654403724e-05,
      "loss": 13.2593,
      "step": 122
    },
    {
      "epoch": 0.025170102829078632,
      "grad_norm": 5.604646682739258,
      "learning_rate": 8.906240896023794e-05,
      "loss": 13.2693,
      "step": 123
    },
    {
      "epoch": 0.025374737811428863,
      "grad_norm": 5.807793140411377,
      "learning_rate": 8.885729807284856e-05,
      "loss": 13.465,
      "step": 124
    },
    {
      "epoch": 0.025579372793779098,
      "grad_norm": 6.032169818878174,
      "learning_rate": 8.865052266813685e-05,
      "loss": 13.0197,
      "step": 125
    },
    {
      "epoch": 0.02578400777612933,
      "grad_norm": 5.969254970550537,
      "learning_rate": 8.844209160367299e-05,
      "loss": 12.91,
      "step": 126
    },
    {
      "epoch": 0.025988642758479563,
      "grad_norm": 5.627323627471924,
      "learning_rate": 8.823201380795001e-05,
      "loss": 12.9693,
      "step": 127
    },
    {
      "epoch": 0.026193277740829794,
      "grad_norm": 5.775904655456543,
      "learning_rate": 8.802029828000156e-05,
      "loss": 13.3716,
      "step": 128
    },
    {
      "epoch": 0.02639791272318003,
      "grad_norm": 6.050631999969482,
      "learning_rate": 8.780695408901613e-05,
      "loss": 12.9946,
      "step": 129
    },
    {
      "epoch": 0.02660254770553026,
      "grad_norm": 6.608086109161377,
      "learning_rate": 8.759199037394887e-05,
      "loss": 12.7268,
      "step": 130
    },
    {
      "epoch": 0.026807182687880494,
      "grad_norm": 6.4099202156066895,
      "learning_rate": 8.737541634312985e-05,
      "loss": 13.3797,
      "step": 131
    },
    {
      "epoch": 0.027011817670230726,
      "grad_norm": 6.958422660827637,
      "learning_rate": 8.715724127386972e-05,
      "loss": 13.2627,
      "step": 132
    },
    {
      "epoch": 0.02721645265258096,
      "grad_norm": 6.657001495361328,
      "learning_rate": 8.693747451206232e-05,
      "loss": 13.1662,
      "step": 133
    },
    {
      "epoch": 0.02742108763493119,
      "grad_norm": 6.775047302246094,
      "learning_rate": 8.671612547178428e-05,
      "loss": 12.8757,
      "step": 134
    },
    {
      "epoch": 0.027625722617281426,
      "grad_norm": 6.7623419761657715,
      "learning_rate": 8.649320363489179e-05,
      "loss": 12.5799,
      "step": 135
    },
    {
      "epoch": 0.027830357599631657,
      "grad_norm": 7.408362865447998,
      "learning_rate": 8.626871855061438e-05,
      "loss": 13.8727,
      "step": 136
    },
    {
      "epoch": 0.02803499258198189,
      "grad_norm": 6.984137535095215,
      "learning_rate": 8.604267983514594e-05,
      "loss": 12.6957,
      "step": 137
    },
    {
      "epoch": 0.028239627564332122,
      "grad_norm": 7.494143486022949,
      "learning_rate": 8.581509717123273e-05,
      "loss": 13.5292,
      "step": 138
    },
    {
      "epoch": 0.028444262546682357,
      "grad_norm": 7.043254375457764,
      "learning_rate": 8.558598030775857e-05,
      "loss": 12.5103,
      "step": 139
    },
    {
      "epoch": 0.028648897529032588,
      "grad_norm": 7.2675957679748535,
      "learning_rate": 8.535533905932738e-05,
      "loss": 12.8951,
      "step": 140
    },
    {
      "epoch": 0.02885353251138282,
      "grad_norm": 7.874957084655762,
      "learning_rate": 8.51231833058426e-05,
      "loss": 12.9737,
      "step": 141
    },
    {
      "epoch": 0.029058167493733054,
      "grad_norm": 8.002019882202148,
      "learning_rate": 8.488952299208401e-05,
      "loss": 12.8148,
      "step": 142
    },
    {
      "epoch": 0.029262802476083285,
      "grad_norm": 8.36933422088623,
      "learning_rate": 8.46543681272818e-05,
      "loss": 12.4946,
      "step": 143
    },
    {
      "epoch": 0.02946743745843352,
      "grad_norm": 9.498835563659668,
      "learning_rate": 8.44177287846877e-05,
      "loss": 13.271,
      "step": 144
    },
    {
      "epoch": 0.02967207244078375,
      "grad_norm": 8.976995468139648,
      "learning_rate": 8.417961510114356e-05,
      "loss": 12.5241,
      "step": 145
    },
    {
      "epoch": 0.029876707423133985,
      "grad_norm": 9.178775787353516,
      "learning_rate": 8.39400372766471e-05,
      "loss": 12.4166,
      "step": 146
    },
    {
      "epoch": 0.030081342405484216,
      "grad_norm": 10.875651359558105,
      "learning_rate": 8.36990055739149e-05,
      "loss": 12.7323,
      "step": 147
    },
    {
      "epoch": 0.03028597738783445,
      "grad_norm": 11.843050003051758,
      "learning_rate": 8.345653031794292e-05,
      "loss": 12.6294,
      "step": 148
    },
    {
      "epoch": 0.030490612370184682,
      "grad_norm": 12.797874450683594,
      "learning_rate": 8.321262189556409e-05,
      "loss": 11.9468,
      "step": 149
    },
    {
      "epoch": 0.030695247352534916,
      "grad_norm": 21.556180953979492,
      "learning_rate": 8.296729075500344e-05,
      "loss": 14.375,
      "step": 150
    },
    {
      "epoch": 0.030899882334885147,
      "grad_norm": 5.878223419189453,
      "learning_rate": 8.272054740543052e-05,
      "loss": 13.2625,
      "step": 151
    },
    {
      "epoch": 0.031104517317235382,
      "grad_norm": 6.683862209320068,
      "learning_rate": 8.247240241650918e-05,
      "loss": 13.4469,
      "step": 152
    },
    {
      "epoch": 0.03130915229958561,
      "grad_norm": 6.695138931274414,
      "learning_rate": 8.222286641794488e-05,
      "loss": 13.8935,
      "step": 153
    },
    {
      "epoch": 0.031513787281935844,
      "grad_norm": 6.529450416564941,
      "learning_rate": 8.197195009902924e-05,
      "loss": 13.2827,
      "step": 154
    },
    {
      "epoch": 0.03171842226428608,
      "grad_norm": 5.889492034912109,
      "learning_rate": 8.171966420818228e-05,
      "loss": 13.2815,
      "step": 155
    },
    {
      "epoch": 0.03192305724663631,
      "grad_norm": 5.005529403686523,
      "learning_rate": 8.146601955249188e-05,
      "loss": 13.1348,
      "step": 156
    },
    {
      "epoch": 0.032127692228986544,
      "grad_norm": 4.527781009674072,
      "learning_rate": 8.121102699725089e-05,
      "loss": 12.9616,
      "step": 157
    },
    {
      "epoch": 0.032332327211336775,
      "grad_norm": 3.992450714111328,
      "learning_rate": 8.095469746549172e-05,
      "loss": 13.2171,
      "step": 158
    },
    {
      "epoch": 0.032536962193687013,
      "grad_norm": 3.9536304473876953,
      "learning_rate": 8.069704193751832e-05,
      "loss": 13.5083,
      "step": 159
    },
    {
      "epoch": 0.032741597176037245,
      "grad_norm": 4.0044264793396,
      "learning_rate": 8.043807145043604e-05,
      "loss": 13.5044,
      "step": 160
    },
    {
      "epoch": 0.032946232158387476,
      "grad_norm": 4.166686058044434,
      "learning_rate": 8.017779709767858e-05,
      "loss": 12.9416,
      "step": 161
    },
    {
      "epoch": 0.03315086714073771,
      "grad_norm": 4.292598724365234,
      "learning_rate": 7.991623002853296e-05,
      "loss": 12.932,
      "step": 162
    },
    {
      "epoch": 0.033355502123087945,
      "grad_norm": 4.622048377990723,
      "learning_rate": 7.965338144766186e-05,
      "loss": 13.1667,
      "step": 163
    },
    {
      "epoch": 0.033560137105438176,
      "grad_norm": 4.218106746673584,
      "learning_rate": 7.938926261462366e-05,
      "loss": 12.9429,
      "step": 164
    },
    {
      "epoch": 0.03376477208778841,
      "grad_norm": 4.615002155303955,
      "learning_rate": 7.912388484339012e-05,
      "loss": 13.4383,
      "step": 165
    },
    {
      "epoch": 0.03396940707013864,
      "grad_norm": 4.371853828430176,
      "learning_rate": 7.88572595018617e-05,
      "loss": 12.6596,
      "step": 166
    },
    {
      "epoch": 0.034174042052488876,
      "grad_norm": 4.507296562194824,
      "learning_rate": 7.858939801138061e-05,
      "loss": 13.0297,
      "step": 167
    },
    {
      "epoch": 0.03437867703483911,
      "grad_norm": 4.6610941886901855,
      "learning_rate": 7.832031184624164e-05,
      "loss": 12.6801,
      "step": 168
    },
    {
      "epoch": 0.03458331201718934,
      "grad_norm": 4.3974714279174805,
      "learning_rate": 7.80500125332005e-05,
      "loss": 12.6394,
      "step": 169
    },
    {
      "epoch": 0.03478794699953957,
      "grad_norm": 4.65360689163208,
      "learning_rate": 7.777851165098012e-05,
      "loss": 13.2642,
      "step": 170
    },
    {
      "epoch": 0.03499258198188981,
      "grad_norm": 4.651695251464844,
      "learning_rate": 7.750582082977467e-05,
      "loss": 13.3055,
      "step": 171
    },
    {
      "epoch": 0.03519721696424004,
      "grad_norm": 5.114010810852051,
      "learning_rate": 7.723195175075136e-05,
      "loss": 13.0045,
      "step": 172
    },
    {
      "epoch": 0.03540185194659027,
      "grad_norm": 5.113755702972412,
      "learning_rate": 7.695691614555003e-05,
      "loss": 12.9366,
      "step": 173
    },
    {
      "epoch": 0.0356064869289405,
      "grad_norm": 5.089533805847168,
      "learning_rate": 7.668072579578058e-05,
      "loss": 12.959,
      "step": 174
    },
    {
      "epoch": 0.03581112191129073,
      "grad_norm": 5.559483051300049,
      "learning_rate": 7.64033925325184e-05,
      "loss": 13.2842,
      "step": 175
    },
    {
      "epoch": 0.03601575689364097,
      "grad_norm": 5.3359761238098145,
      "learning_rate": 7.612492823579745e-05,
      "loss": 13.0262,
      "step": 176
    },
    {
      "epoch": 0.0362203918759912,
      "grad_norm": 5.409842014312744,
      "learning_rate": 7.584534483410137e-05,
      "loss": 13.0076,
      "step": 177
    },
    {
      "epoch": 0.03642502685834143,
      "grad_norm": 5.253081321716309,
      "learning_rate": 7.55646543038526e-05,
      "loss": 11.9703,
      "step": 178
    },
    {
      "epoch": 0.03662966184069166,
      "grad_norm": 5.482647895812988,
      "learning_rate": 7.528286866889924e-05,
      "loss": 12.6692,
      "step": 179
    },
    {
      "epoch": 0.0368342968230419,
      "grad_norm": 5.659306049346924,
      "learning_rate": 7.500000000000001e-05,
      "loss": 13.0874,
      "step": 180
    },
    {
      "epoch": 0.03703893180539213,
      "grad_norm": 5.71022891998291,
      "learning_rate": 7.471606041430723e-05,
      "loss": 12.9602,
      "step": 181
    },
    {
      "epoch": 0.03724356678774236,
      "grad_norm": 6.031240940093994,
      "learning_rate": 7.443106207484776e-05,
      "loss": 12.8276,
      "step": 182
    },
    {
      "epoch": 0.037448201770092594,
      "grad_norm": 5.916280746459961,
      "learning_rate": 7.414501719000187e-05,
      "loss": 12.7036,
      "step": 183
    },
    {
      "epoch": 0.03765283675244283,
      "grad_norm": 6.090421676635742,
      "learning_rate": 7.385793801298042e-05,
      "loss": 12.5362,
      "step": 184
    },
    {
      "epoch": 0.03785747173479306,
      "grad_norm": 6.97968053817749,
      "learning_rate": 7.35698368412999e-05,
      "loss": 13.1994,
      "step": 185
    },
    {
      "epoch": 0.038062106717143294,
      "grad_norm": 6.6946587562561035,
      "learning_rate": 7.328072601625557e-05,
      "loss": 12.9428,
      "step": 186
    },
    {
      "epoch": 0.038266741699493526,
      "grad_norm": 6.86458158493042,
      "learning_rate": 7.2990617922393e-05,
      "loss": 13.5336,
      "step": 187
    },
    {
      "epoch": 0.038471376681843764,
      "grad_norm": 7.41053581237793,
      "learning_rate": 7.269952498697734e-05,
      "loss": 13.1752,
      "step": 188
    },
    {
      "epoch": 0.038676011664193995,
      "grad_norm": 6.769413948059082,
      "learning_rate": 7.240745967946113e-05,
      "loss": 12.3618,
      "step": 189
    },
    {
      "epoch": 0.038880646646544226,
      "grad_norm": 8.171807289123535,
      "learning_rate": 7.211443451095007e-05,
      "loss": 13.4612,
      "step": 190
    },
    {
      "epoch": 0.03908528162889446,
      "grad_norm": 7.6870598793029785,
      "learning_rate": 7.18204620336671e-05,
      "loss": 12.8721,
      "step": 191
    },
    {
      "epoch": 0.039289916611244695,
      "grad_norm": 7.984126567840576,
      "learning_rate": 7.152555484041476e-05,
      "loss": 12.5025,
      "step": 192
    },
    {
      "epoch": 0.039494551593594926,
      "grad_norm": 8.748424530029297,
      "learning_rate": 7.122972556403567e-05,
      "loss": 12.5803,
      "step": 193
    },
    {
      "epoch": 0.03969918657594516,
      "grad_norm": 8.19789981842041,
      "learning_rate": 7.09329868768714e-05,
      "loss": 13.0793,
      "step": 194
    },
    {
      "epoch": 0.03990382155829539,
      "grad_norm": 8.25755786895752,
      "learning_rate": 7.063535149021973e-05,
      "loss": 13.2436,
      "step": 195
    },
    {
      "epoch": 0.040108456540645626,
      "grad_norm": 10.084080696105957,
      "learning_rate": 7.033683215379002e-05,
      "loss": 12.4769,
      "step": 196
    },
    {
      "epoch": 0.04031309152299586,
      "grad_norm": 11.04244327545166,
      "learning_rate": 7.003744165515705e-05,
      "loss": 13.3229,
      "step": 197
    },
    {
      "epoch": 0.04051772650534609,
      "grad_norm": 10.718149185180664,
      "learning_rate": 6.973719281921335e-05,
      "loss": 13.0458,
      "step": 198
    },
    {
      "epoch": 0.04072236148769632,
      "grad_norm": 12.596996307373047,
      "learning_rate": 6.943609850761979e-05,
      "loss": 13.2156,
      "step": 199
    },
    {
      "epoch": 0.04092699647004656,
      "grad_norm": 16.626497268676758,
      "learning_rate": 6.91341716182545e-05,
      "loss": 12.1366,
      "step": 200
    },
    {
      "epoch": 0.04092699647004656,
      "eval_loss": 3.208616256713867,
      "eval_runtime": 47.3335,
      "eval_samples_per_second": 173.872,
      "eval_steps_per_second": 43.479,
      "step": 200
    },
    {
      "epoch": 0.04113163145239679,
      "grad_norm": 3.9187700748443604,
      "learning_rate": 6.883142508466054e-05,
      "loss": 13.2198,
      "step": 201
    },
    {
      "epoch": 0.04133626643474702,
      "grad_norm": 5.088418006896973,
      "learning_rate": 6.852787187549182e-05,
      "loss": 12.934,
      "step": 202
    },
    {
      "epoch": 0.04154090141709725,
      "grad_norm": 4.74566125869751,
      "learning_rate": 6.82235249939575e-05,
      "loss": 13.3393,
      "step": 203
    },
    {
      "epoch": 0.04174553639944749,
      "grad_norm": 4.898460865020752,
      "learning_rate": 6.7918397477265e-05,
      "loss": 13.5989,
      "step": 204
    },
    {
      "epoch": 0.04195017138179772,
      "grad_norm": 4.619757652282715,
      "learning_rate": 6.761250239606169e-05,
      "loss": 13.0342,
      "step": 205
    },
    {
      "epoch": 0.04215480636414795,
      "grad_norm": 4.482340335845947,
      "learning_rate": 6.730585285387465e-05,
      "loss": 13.0489,
      "step": 206
    },
    {
      "epoch": 0.04235944134649818,
      "grad_norm": 4.1753644943237305,
      "learning_rate": 6.699846198654971e-05,
      "loss": 13.2165,
      "step": 207
    },
    {
      "epoch": 0.04256407632884842,
      "grad_norm": 3.9566304683685303,
      "learning_rate": 6.669034296168855e-05,
      "loss": 13.2601,
      "step": 208
    },
    {
      "epoch": 0.04276871131119865,
      "grad_norm": 4.045615196228027,
      "learning_rate": 6.638150897808468e-05,
      "loss": 13.0854,
      "step": 209
    },
    {
      "epoch": 0.04297334629354888,
      "grad_norm": 3.9672138690948486,
      "learning_rate": 6.607197326515808e-05,
      "loss": 13.5277,
      "step": 210
    },
    {
      "epoch": 0.04317798127589911,
      "grad_norm": 3.8964602947235107,
      "learning_rate": 6.57617490823885e-05,
      "loss": 13.3445,
      "step": 211
    },
    {
      "epoch": 0.043382616258249344,
      "grad_norm": 3.9119648933410645,
      "learning_rate": 6.545084971874738e-05,
      "loss": 12.763,
      "step": 212
    },
    {
      "epoch": 0.04358725124059958,
      "grad_norm": 4.09339714050293,
      "learning_rate": 6.513928849212873e-05,
      "loss": 13.1653,
      "step": 213
    },
    {
      "epoch": 0.043791886222949813,
      "grad_norm": 4.33394193649292,
      "learning_rate": 6.482707874877854e-05,
      "loss": 13.1689,
      "step": 214
    },
    {
      "epoch": 0.043996521205300045,
      "grad_norm": 4.071203231811523,
      "learning_rate": 6.451423386272312e-05,
      "loss": 12.9157,
      "step": 215
    },
    {
      "epoch": 0.044201156187650276,
      "grad_norm": 4.155096054077148,
      "learning_rate": 6.420076723519614e-05,
      "loss": 12.9944,
      "step": 216
    },
    {
      "epoch": 0.044405791170000514,
      "grad_norm": 4.474510669708252,
      "learning_rate": 6.388669229406462e-05,
      "loss": 12.9211,
      "step": 217
    },
    {
      "epoch": 0.044610426152350745,
      "grad_norm": 4.203741550445557,
      "learning_rate": 6.357202249325371e-05,
      "loss": 12.5732,
      "step": 218
    },
    {
      "epoch": 0.044815061134700976,
      "grad_norm": 4.361083984375,
      "learning_rate": 6.32567713121704e-05,
      "loss": 13.1363,
      "step": 219
    },
    {
      "epoch": 0.04501969611705121,
      "grad_norm": 4.626219749450684,
      "learning_rate": 6.294095225512603e-05,
      "loss": 13.1044,
      "step": 220
    },
    {
      "epoch": 0.045224331099401445,
      "grad_norm": 4.69849967956543,
      "learning_rate": 6.26245788507579e-05,
      "loss": 12.7933,
      "step": 221
    },
    {
      "epoch": 0.045428966081751676,
      "grad_norm": 4.679666996002197,
      "learning_rate": 6.230766465144967e-05,
      "loss": 13.1581,
      "step": 222
    },
    {
      "epoch": 0.04563360106410191,
      "grad_norm": 4.953638553619385,
      "learning_rate": 6.199022323275083e-05,
      "loss": 13.0212,
      "step": 223
    },
    {
      "epoch": 0.04583823604645214,
      "grad_norm": 4.850236415863037,
      "learning_rate": 6.167226819279528e-05,
      "loss": 13.0019,
      "step": 224
    },
    {
      "epoch": 0.046042871028802376,
      "grad_norm": 4.989190578460693,
      "learning_rate": 6.135381315171867e-05,
      "loss": 12.2553,
      "step": 225
    },
    {
      "epoch": 0.04624750601115261,
      "grad_norm": 4.897017478942871,
      "learning_rate": 6.103487175107507e-05,
      "loss": 13.0096,
      "step": 226
    },
    {
      "epoch": 0.04645214099350284,
      "grad_norm": 4.998581886291504,
      "learning_rate": 6.071545765325254e-05,
      "loss": 12.4746,
      "step": 227
    },
    {
      "epoch": 0.04665677597585307,
      "grad_norm": 5.277119159698486,
      "learning_rate": 6.0395584540887963e-05,
      "loss": 12.6303,
      "step": 228
    },
    {
      "epoch": 0.04686141095820331,
      "grad_norm": 5.548853874206543,
      "learning_rate": 6.007526611628086e-05,
      "loss": 12.8215,
      "step": 229
    },
    {
      "epoch": 0.04706604594055354,
      "grad_norm": 5.378997325897217,
      "learning_rate": 5.9754516100806423e-05,
      "loss": 13.1972,
      "step": 230
    },
    {
      "epoch": 0.04727068092290377,
      "grad_norm": 5.815462589263916,
      "learning_rate": 5.9433348234327765e-05,
      "loss": 13.2821,
      "step": 231
    },
    {
      "epoch": 0.047475315905254,
      "grad_norm": 5.872306823730469,
      "learning_rate": 5.911177627460739e-05,
      "loss": 13.199,
      "step": 232
    },
    {
      "epoch": 0.04767995088760424,
      "grad_norm": 5.859600067138672,
      "learning_rate": 5.8789813996717736e-05,
      "loss": 12.6879,
      "step": 233
    },
    {
      "epoch": 0.04788458586995447,
      "grad_norm": 6.172786712646484,
      "learning_rate": 5.8467475192451226e-05,
      "loss": 13.2259,
      "step": 234
    },
    {
      "epoch": 0.0480892208523047,
      "grad_norm": 6.31078577041626,
      "learning_rate": 5.814477366972945e-05,
      "loss": 12.3852,
      "step": 235
    },
    {
      "epoch": 0.04829385583465493,
      "grad_norm": 6.534313201904297,
      "learning_rate": 5.782172325201155e-05,
      "loss": 13.3609,
      "step": 236
    },
    {
      "epoch": 0.04849849081700517,
      "grad_norm": 6.585941314697266,
      "learning_rate": 5.749833777770225e-05,
      "loss": 12.9351,
      "step": 237
    },
    {
      "epoch": 0.0487031257993554,
      "grad_norm": 6.7909159660339355,
      "learning_rate": 5.717463109955896e-05,
      "loss": 12.2785,
      "step": 238
    },
    {
      "epoch": 0.04890776078170563,
      "grad_norm": 7.011481761932373,
      "learning_rate": 5.685061708409841e-05,
      "loss": 13.0815,
      "step": 239
    },
    {
      "epoch": 0.04911239576405586,
      "grad_norm": 7.400633335113525,
      "learning_rate": 5.6526309611002594e-05,
      "loss": 12.5917,
      "step": 240
    },
    {
      "epoch": 0.0493170307464061,
      "grad_norm": 6.944792747497559,
      "learning_rate": 5.6201722572524275e-05,
      "loss": 12.8307,
      "step": 241
    },
    {
      "epoch": 0.04952166572875633,
      "grad_norm": 9.408550262451172,
      "learning_rate": 5.587686987289189e-05,
      "loss": 13.0507,
      "step": 242
    },
    {
      "epoch": 0.049726300711106564,
      "grad_norm": 8.005476951599121,
      "learning_rate": 5.5551765427713884e-05,
      "loss": 12.5769,
      "step": 243
    },
    {
      "epoch": 0.049930935693456795,
      "grad_norm": 8.190591812133789,
      "learning_rate": 5.522642316338268e-05,
      "loss": 12.2517,
      "step": 244
    },
    {
      "epoch": 0.050135570675807026,
      "grad_norm": 9.257965087890625,
      "learning_rate": 5.490085701647805e-05,
      "loss": 13.0658,
      "step": 245
    },
    {
      "epoch": 0.050340205658157264,
      "grad_norm": 9.173686027526855,
      "learning_rate": 5.457508093317013e-05,
      "loss": 12.6359,
      "step": 246
    },
    {
      "epoch": 0.050544840640507495,
      "grad_norm": 9.59607982635498,
      "learning_rate": 5.4249108868622086e-05,
      "loss": 13.0733,
      "step": 247
    },
    {
      "epoch": 0.050749475622857726,
      "grad_norm": 10.842556953430176,
      "learning_rate": 5.392295478639225e-05,
      "loss": 12.6057,
      "step": 248
    },
    {
      "epoch": 0.05095411060520796,
      "grad_norm": 11.467256546020508,
      "learning_rate": 5.359663265783598e-05,
      "loss": 12.7772,
      "step": 249
    },
    {
      "epoch": 0.051158745587558195,
      "grad_norm": 16.473846435546875,
      "learning_rate": 5.327015646150716e-05,
      "loss": 12.2446,
      "step": 250
    },
    {
      "epoch": 0.051363380569908426,
      "grad_norm": 2.952035903930664,
      "learning_rate": 5.294354018255945e-05,
      "loss": 13.1735,
      "step": 251
    },
    {
      "epoch": 0.05156801555225866,
      "grad_norm": 3.22007417678833,
      "learning_rate": 5.26167978121472e-05,
      "loss": 13.0792,
      "step": 252
    },
    {
      "epoch": 0.05177265053460889,
      "grad_norm": 3.549088716506958,
      "learning_rate": 5.228994334682604e-05,
      "loss": 12.7948,
      "step": 253
    },
    {
      "epoch": 0.051977285516959126,
      "grad_norm": 3.728848695755005,
      "learning_rate": 5.196299078795344e-05,
      "loss": 13.0125,
      "step": 254
    },
    {
      "epoch": 0.05218192049930936,
      "grad_norm": 3.749281167984009,
      "learning_rate": 5.1635954141088813e-05,
      "loss": 13.2622,
      "step": 255
    },
    {
      "epoch": 0.05238655548165959,
      "grad_norm": 3.824709415435791,
      "learning_rate": 5.1308847415393666e-05,
      "loss": 12.9234,
      "step": 256
    },
    {
      "epoch": 0.05259119046400982,
      "grad_norm": 3.764427661895752,
      "learning_rate": 5.0981684623031415e-05,
      "loss": 13.2274,
      "step": 257
    },
    {
      "epoch": 0.05279582544636006,
      "grad_norm": 3.7923285961151123,
      "learning_rate": 5.0654479778567223e-05,
      "loss": 13.1988,
      "step": 258
    },
    {
      "epoch": 0.05300046042871029,
      "grad_norm": 3.733365535736084,
      "learning_rate": 5.0327246898367597e-05,
      "loss": 12.5384,
      "step": 259
    },
    {
      "epoch": 0.05320509541106052,
      "grad_norm": 3.8030307292938232,
      "learning_rate": 5e-05,
      "loss": 12.991,
      "step": 260
    },
    {
      "epoch": 0.05340973039341075,
      "grad_norm": 3.7069780826568604,
      "learning_rate": 4.9672753101632415e-05,
      "loss": 12.7534,
      "step": 261
    },
    {
      "epoch": 0.05361436537576099,
      "grad_norm": 3.764336109161377,
      "learning_rate": 4.934552022143279e-05,
      "loss": 12.8254,
      "step": 262
    },
    {
      "epoch": 0.05381900035811122,
      "grad_norm": 3.753891944885254,
      "learning_rate": 4.901831537696859e-05,
      "loss": 12.6035,
      "step": 263
    },
    {
      "epoch": 0.05402363534046145,
      "grad_norm": 3.9714443683624268,
      "learning_rate": 4.869115258460635e-05,
      "loss": 12.6499,
      "step": 264
    },
    {
      "epoch": 0.05422827032281168,
      "grad_norm": 3.999743938446045,
      "learning_rate": 4.83640458589112e-05,
      "loss": 13.0991,
      "step": 265
    },
    {
      "epoch": 0.05443290530516192,
      "grad_norm": 3.9809932708740234,
      "learning_rate": 4.8037009212046586e-05,
      "loss": 12.6208,
      "step": 266
    },
    {
      "epoch": 0.05463754028751215,
      "grad_norm": 4.165307521820068,
      "learning_rate": 4.7710056653173976e-05,
      "loss": 13.0348,
      "step": 267
    },
    {
      "epoch": 0.05484217526986238,
      "grad_norm": 4.428051948547363,
      "learning_rate": 4.738320218785281e-05,
      "loss": 13.4851,
      "step": 268
    },
    {
      "epoch": 0.055046810252212613,
      "grad_norm": 4.276752948760986,
      "learning_rate": 4.7056459817440544e-05,
      "loss": 12.8883,
      "step": 269
    },
    {
      "epoch": 0.05525144523456285,
      "grad_norm": 4.741238594055176,
      "learning_rate": 4.6729843538492847e-05,
      "loss": 13.3597,
      "step": 270
    },
    {
      "epoch": 0.05545608021691308,
      "grad_norm": 4.348086833953857,
      "learning_rate": 4.640336734216403e-05,
      "loss": 12.9206,
      "step": 271
    },
    {
      "epoch": 0.055660715199263314,
      "grad_norm": 4.487641334533691,
      "learning_rate": 4.607704521360776e-05,
      "loss": 12.7532,
      "step": 272
    },
    {
      "epoch": 0.055865350181613545,
      "grad_norm": 4.975533485412598,
      "learning_rate": 4.575089113137792e-05,
      "loss": 12.5924,
      "step": 273
    },
    {
      "epoch": 0.05606998516396378,
      "grad_norm": 4.721080303192139,
      "learning_rate": 4.542491906682989e-05,
      "loss": 12.9426,
      "step": 274
    },
    {
      "epoch": 0.056274620146314014,
      "grad_norm": 4.957543849945068,
      "learning_rate": 4.509914298352197e-05,
      "loss": 12.5506,
      "step": 275
    },
    {
      "epoch": 0.056479255128664245,
      "grad_norm": 4.958243370056152,
      "learning_rate": 4.477357683661734e-05,
      "loss": 12.7699,
      "step": 276
    },
    {
      "epoch": 0.056683890111014476,
      "grad_norm": 5.35684061050415,
      "learning_rate": 4.444823457228612e-05,
      "loss": 12.8455,
      "step": 277
    },
    {
      "epoch": 0.056888525093364714,
      "grad_norm": 5.440086364746094,
      "learning_rate": 4.412313012710813e-05,
      "loss": 13.7743,
      "step": 278
    },
    {
      "epoch": 0.057093160075714945,
      "grad_norm": 5.20829439163208,
      "learning_rate": 4.379827742747575e-05,
      "loss": 13.318,
      "step": 279
    },
    {
      "epoch": 0.057297795058065176,
      "grad_norm": 5.2258405685424805,
      "learning_rate": 4.347369038899744e-05,
      "loss": 13.0874,
      "step": 280
    },
    {
      "epoch": 0.05750243004041541,
      "grad_norm": 5.654691219329834,
      "learning_rate": 4.3149382915901606e-05,
      "loss": 12.4869,
      "step": 281
    },
    {
      "epoch": 0.05770706502276564,
      "grad_norm": 5.957024097442627,
      "learning_rate": 4.282536890044104e-05,
      "loss": 12.8174,
      "step": 282
    },
    {
      "epoch": 0.057911700005115876,
      "grad_norm": 6.341736316680908,
      "learning_rate": 4.250166222229774e-05,
      "loss": 12.6841,
      "step": 283
    },
    {
      "epoch": 0.05811633498746611,
      "grad_norm": 6.56013822555542,
      "learning_rate": 4.2178276747988446e-05,
      "loss": 13.0789,
      "step": 284
    },
    {
      "epoch": 0.05832096996981634,
      "grad_norm": 6.450329780578613,
      "learning_rate": 4.185522633027057e-05,
      "loss": 12.6028,
      "step": 285
    },
    {
      "epoch": 0.05852560495216657,
      "grad_norm": 6.356710433959961,
      "learning_rate": 4.153252480754877e-05,
      "loss": 13.0871,
      "step": 286
    },
    {
      "epoch": 0.05873023993451681,
      "grad_norm": 6.647814750671387,
      "learning_rate": 4.1210186003282275e-05,
      "loss": 12.6458,
      "step": 287
    },
    {
      "epoch": 0.05893487491686704,
      "grad_norm": 6.441559314727783,
      "learning_rate": 4.088822372539263e-05,
      "loss": 12.2483,
      "step": 288
    },
    {
      "epoch": 0.05913950989921727,
      "grad_norm": 8.019023895263672,
      "learning_rate": 4.0566651765672246e-05,
      "loss": 12.7241,
      "step": 289
    },
    {
      "epoch": 0.0593441448815675,
      "grad_norm": 7.507869720458984,
      "learning_rate": 4.0245483899193595e-05,
      "loss": 13.3737,
      "step": 290
    },
    {
      "epoch": 0.05954877986391774,
      "grad_norm": 7.296957015991211,
      "learning_rate": 3.992473388371915e-05,
      "loss": 12.6952,
      "step": 291
    },
    {
      "epoch": 0.05975341484626797,
      "grad_norm": 8.110812187194824,
      "learning_rate": 3.960441545911204e-05,
      "loss": 12.2771,
      "step": 292
    },
    {
      "epoch": 0.0599580498286182,
      "grad_norm": 8.923057556152344,
      "learning_rate": 3.928454234674747e-05,
      "loss": 12.7108,
      "step": 293
    },
    {
      "epoch": 0.06016268481096843,
      "grad_norm": 10.090682983398438,
      "learning_rate": 3.896512824892495e-05,
      "loss": 13.1477,
      "step": 294
    },
    {
      "epoch": 0.06036731979331867,
      "grad_norm": 8.957847595214844,
      "learning_rate": 3.864618684828134e-05,
      "loss": 11.7159,
      "step": 295
    },
    {
      "epoch": 0.0605719547756689,
      "grad_norm": 10.131745338439941,
      "learning_rate": 3.832773180720475e-05,
      "loss": 12.191,
      "step": 296
    },
    {
      "epoch": 0.06077658975801913,
      "grad_norm": 10.587480545043945,
      "learning_rate": 3.800977676724919e-05,
      "loss": 12.6343,
      "step": 297
    },
    {
      "epoch": 0.060981224740369364,
      "grad_norm": 11.739582061767578,
      "learning_rate": 3.769233534855035e-05,
      "loss": 12.5622,
      "step": 298
    },
    {
      "epoch": 0.0611858597227196,
      "grad_norm": 14.421760559082031,
      "learning_rate": 3.73754211492421e-05,
      "loss": 12.8323,
      "step": 299
    },
    {
      "epoch": 0.06139049470506983,
      "grad_norm": 17.0710391998291,
      "learning_rate": 3.705904774487396e-05,
      "loss": 11.9398,
      "step": 300
    },
    {
      "epoch": 0.06139049470506983,
      "eval_loss": 3.1531736850738525,
      "eval_runtime": 47.3782,
      "eval_samples_per_second": 173.709,
      "eval_steps_per_second": 43.438,
      "step": 300
    },
    {
      "epoch": 0.061595129687420064,
      "grad_norm": 2.744182825088501,
      "learning_rate": 3.6743228687829595e-05,
      "loss": 12.8172,
      "step": 301
    },
    {
      "epoch": 0.061799764669770295,
      "grad_norm": 3.1303441524505615,
      "learning_rate": 3.642797750674629e-05,
      "loss": 12.6422,
      "step": 302
    },
    {
      "epoch": 0.06200439965212053,
      "grad_norm": 3.3844480514526367,
      "learning_rate": 3.6113307705935396e-05,
      "loss": 13.0422,
      "step": 303
    },
    {
      "epoch": 0.062209034634470764,
      "grad_norm": 3.383885383605957,
      "learning_rate": 3.579923276480387e-05,
      "loss": 12.6751,
      "step": 304
    },
    {
      "epoch": 0.062413669616820995,
      "grad_norm": 3.3772575855255127,
      "learning_rate": 3.5485766137276894e-05,
      "loss": 13.055,
      "step": 305
    },
    {
      "epoch": 0.06261830459917123,
      "grad_norm": 3.6094918251037598,
      "learning_rate": 3.5172921251221455e-05,
      "loss": 13.1467,
      "step": 306
    },
    {
      "epoch": 0.06282293958152146,
      "grad_norm": 3.674668073654175,
      "learning_rate": 3.486071150787128e-05,
      "loss": 12.7998,
      "step": 307
    },
    {
      "epoch": 0.06302757456387169,
      "grad_norm": 3.914242744445801,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 13.6751,
      "step": 308
    },
    {
      "epoch": 0.06323220954622193,
      "grad_norm": 3.7367589473724365,
      "learning_rate": 3.423825091761153e-05,
      "loss": 13.0127,
      "step": 309
    },
    {
      "epoch": 0.06343684452857216,
      "grad_norm": 3.7376673221588135,
      "learning_rate": 3.392802673484193e-05,
      "loss": 13.3006,
      "step": 310
    },
    {
      "epoch": 0.06364147951092239,
      "grad_norm": 3.9828696250915527,
      "learning_rate": 3.361849102191533e-05,
      "loss": 13.0535,
      "step": 311
    },
    {
      "epoch": 0.06384611449327263,
      "grad_norm": 3.7258951663970947,
      "learning_rate": 3.330965703831146e-05,
      "loss": 12.8206,
      "step": 312
    },
    {
      "epoch": 0.06405074947562286,
      "grad_norm": 3.842252254486084,
      "learning_rate": 3.300153801345028e-05,
      "loss": 13.0683,
      "step": 313
    },
    {
      "epoch": 0.06425538445797309,
      "grad_norm": 3.9074199199676514,
      "learning_rate": 3.2694147146125345e-05,
      "loss": 12.9179,
      "step": 314
    },
    {
      "epoch": 0.06446001944032333,
      "grad_norm": 3.9515974521636963,
      "learning_rate": 3.2387497603938326e-05,
      "loss": 13.3211,
      "step": 315
    },
    {
      "epoch": 0.06466465442267355,
      "grad_norm": 4.135197162628174,
      "learning_rate": 3.2081602522734986e-05,
      "loss": 13.1106,
      "step": 316
    },
    {
      "epoch": 0.06486928940502379,
      "grad_norm": 4.115512371063232,
      "learning_rate": 3.177647500604252e-05,
      "loss": 12.8488,
      "step": 317
    },
    {
      "epoch": 0.06507392438737403,
      "grad_norm": 4.200262069702148,
      "learning_rate": 3.147212812450819e-05,
      "loss": 12.7581,
      "step": 318
    },
    {
      "epoch": 0.06527855936972425,
      "grad_norm": 4.211337089538574,
      "learning_rate": 3.116857491533947e-05,
      "loss": 12.8883,
      "step": 319
    },
    {
      "epoch": 0.06548319435207449,
      "grad_norm": 4.417909622192383,
      "learning_rate": 3.086582838174551e-05,
      "loss": 12.9176,
      "step": 320
    },
    {
      "epoch": 0.06568782933442471,
      "grad_norm": 4.327807903289795,
      "learning_rate": 3.056390149238022e-05,
      "loss": 12.5289,
      "step": 321
    },
    {
      "epoch": 0.06589246431677495,
      "grad_norm": 4.564601898193359,
      "learning_rate": 3.0262807180786647e-05,
      "loss": 12.8324,
      "step": 322
    },
    {
      "epoch": 0.06609709929912519,
      "grad_norm": 4.540707111358643,
      "learning_rate": 2.996255834484296e-05,
      "loss": 12.1696,
      "step": 323
    },
    {
      "epoch": 0.06630173428147541,
      "grad_norm": 4.797798156738281,
      "learning_rate": 2.9663167846209998e-05,
      "loss": 12.73,
      "step": 324
    },
    {
      "epoch": 0.06650636926382565,
      "grad_norm": 4.712722301483154,
      "learning_rate": 2.936464850978027e-05,
      "loss": 12.439,
      "step": 325
    },
    {
      "epoch": 0.06671100424617589,
      "grad_norm": 4.9245781898498535,
      "learning_rate": 2.9067013123128613e-05,
      "loss": 12.8693,
      "step": 326
    },
    {
      "epoch": 0.06691563922852611,
      "grad_norm": 5.3487467765808105,
      "learning_rate": 2.8770274435964355e-05,
      "loss": 12.7175,
      "step": 327
    },
    {
      "epoch": 0.06712027421087635,
      "grad_norm": 5.03184175491333,
      "learning_rate": 2.8474445159585235e-05,
      "loss": 12.3854,
      "step": 328
    },
    {
      "epoch": 0.06732490919322658,
      "grad_norm": 4.980981349945068,
      "learning_rate": 2.8179537966332887e-05,
      "loss": 13.0615,
      "step": 329
    },
    {
      "epoch": 0.06752954417557681,
      "grad_norm": 5.728270053863525,
      "learning_rate": 2.7885565489049946e-05,
      "loss": 13.4382,
      "step": 330
    },
    {
      "epoch": 0.06773417915792705,
      "grad_norm": 5.375972270965576,
      "learning_rate": 2.759254032053888e-05,
      "loss": 12.7093,
      "step": 331
    },
    {
      "epoch": 0.06793881414027728,
      "grad_norm": 6.161487579345703,
      "learning_rate": 2.7300475013022663e-05,
      "loss": 13.2182,
      "step": 332
    },
    {
      "epoch": 0.06814344912262751,
      "grad_norm": 5.7192254066467285,
      "learning_rate": 2.700938207760701e-05,
      "loss": 13.2617,
      "step": 333
    },
    {
      "epoch": 0.06834808410497775,
      "grad_norm": 5.916773796081543,
      "learning_rate": 2.671927398374443e-05,
      "loss": 13.1881,
      "step": 334
    },
    {
      "epoch": 0.06855271908732798,
      "grad_norm": 6.228060722351074,
      "learning_rate": 2.6430163158700115e-05,
      "loss": 12.904,
      "step": 335
    },
    {
      "epoch": 0.06875735406967821,
      "grad_norm": 7.778335094451904,
      "learning_rate": 2.6142061987019577e-05,
      "loss": 13.1079,
      "step": 336
    },
    {
      "epoch": 0.06896198905202844,
      "grad_norm": 6.622939109802246,
      "learning_rate": 2.5854982809998153e-05,
      "loss": 12.9644,
      "step": 337
    },
    {
      "epoch": 0.06916662403437868,
      "grad_norm": 6.916367053985596,
      "learning_rate": 2.556893792515227e-05,
      "loss": 13.016,
      "step": 338
    },
    {
      "epoch": 0.06937125901672891,
      "grad_norm": 6.418735980987549,
      "learning_rate": 2.5283939585692783e-05,
      "loss": 12.5322,
      "step": 339
    },
    {
      "epoch": 0.06957589399907914,
      "grad_norm": 7.215633392333984,
      "learning_rate": 2.500000000000001e-05,
      "loss": 12.5916,
      "step": 340
    },
    {
      "epoch": 0.06978052898142938,
      "grad_norm": 7.442222595214844,
      "learning_rate": 2.471713133110078e-05,
      "loss": 13.7867,
      "step": 341
    },
    {
      "epoch": 0.06998516396377961,
      "grad_norm": 8.242687225341797,
      "learning_rate": 2.4435345696147403e-05,
      "loss": 12.5457,
      "step": 342
    },
    {
      "epoch": 0.07018979894612984,
      "grad_norm": 8.024588584899902,
      "learning_rate": 2.4154655165898627e-05,
      "loss": 13.2987,
      "step": 343
    },
    {
      "epoch": 0.07039443392848008,
      "grad_norm": 8.097381591796875,
      "learning_rate": 2.3875071764202563e-05,
      "loss": 12.2318,
      "step": 344
    },
    {
      "epoch": 0.0705990689108303,
      "grad_norm": 8.730584144592285,
      "learning_rate": 2.3596607467481603e-05,
      "loss": 12.7858,
      "step": 345
    },
    {
      "epoch": 0.07080370389318054,
      "grad_norm": 10.03264045715332,
      "learning_rate": 2.3319274204219428e-05,
      "loss": 12.4164,
      "step": 346
    },
    {
      "epoch": 0.07100833887553078,
      "grad_norm": 10.419230461120605,
      "learning_rate": 2.3043083854449988e-05,
      "loss": 11.9891,
      "step": 347
    },
    {
      "epoch": 0.071212973857881,
      "grad_norm": 10.628999710083008,
      "learning_rate": 2.2768048249248648e-05,
      "loss": 13.0758,
      "step": 348
    },
    {
      "epoch": 0.07141760884023124,
      "grad_norm": 12.918512344360352,
      "learning_rate": 2.2494179170225333e-05,
      "loss": 12.2184,
      "step": 349
    },
    {
      "epoch": 0.07162224382258146,
      "grad_norm": 17.08102798461914,
      "learning_rate": 2.2221488349019903e-05,
      "loss": 13.2989,
      "step": 350
    },
    {
      "epoch": 0.0718268788049317,
      "grad_norm": 2.621025800704956,
      "learning_rate": 2.194998746679952e-05,
      "loss": 13.1545,
      "step": 351
    },
    {
      "epoch": 0.07203151378728194,
      "grad_norm": 2.733877182006836,
      "learning_rate": 2.167968815375837e-05,
      "loss": 12.7721,
      "step": 352
    },
    {
      "epoch": 0.07223614876963216,
      "grad_norm": 2.7492332458496094,
      "learning_rate": 2.1410601988619394e-05,
      "loss": 12.9304,
      "step": 353
    },
    {
      "epoch": 0.0724407837519824,
      "grad_norm": 3.0048248767852783,
      "learning_rate": 2.1142740498138324e-05,
      "loss": 12.5243,
      "step": 354
    },
    {
      "epoch": 0.07264541873433264,
      "grad_norm": 3.2829930782318115,
      "learning_rate": 2.08761151566099e-05,
      "loss": 12.8763,
      "step": 355
    },
    {
      "epoch": 0.07285005371668286,
      "grad_norm": 3.4896061420440674,
      "learning_rate": 2.061073738537635e-05,
      "loss": 12.6124,
      "step": 356
    },
    {
      "epoch": 0.0730546886990331,
      "grad_norm": 3.766759157180786,
      "learning_rate": 2.034661855233815e-05,
      "loss": 12.8122,
      "step": 357
    },
    {
      "epoch": 0.07325932368138333,
      "grad_norm": 3.564368486404419,
      "learning_rate": 2.008376997146705e-05,
      "loss": 12.5289,
      "step": 358
    },
    {
      "epoch": 0.07346395866373356,
      "grad_norm": 3.5410869121551514,
      "learning_rate": 1.982220290232143e-05,
      "loss": 12.9323,
      "step": 359
    },
    {
      "epoch": 0.0736685936460838,
      "grad_norm": 3.591470718383789,
      "learning_rate": 1.9561928549563968e-05,
      "loss": 12.6307,
      "step": 360
    },
    {
      "epoch": 0.07387322862843403,
      "grad_norm": 3.747335910797119,
      "learning_rate": 1.9302958062481673e-05,
      "loss": 12.2941,
      "step": 361
    },
    {
      "epoch": 0.07407786361078426,
      "grad_norm": 3.776078224182129,
      "learning_rate": 1.9045302534508297e-05,
      "loss": 12.6067,
      "step": 362
    },
    {
      "epoch": 0.0742824985931345,
      "grad_norm": 4.040799617767334,
      "learning_rate": 1.8788973002749112e-05,
      "loss": 12.9593,
      "step": 363
    },
    {
      "epoch": 0.07448713357548473,
      "grad_norm": 3.7775509357452393,
      "learning_rate": 1.8533980447508137e-05,
      "loss": 12.3113,
      "step": 364
    },
    {
      "epoch": 0.07469176855783496,
      "grad_norm": 3.858264446258545,
      "learning_rate": 1.8280335791817733e-05,
      "loss": 12.9729,
      "step": 365
    },
    {
      "epoch": 0.07489640354018519,
      "grad_norm": 4.055905342102051,
      "learning_rate": 1.8028049900970767e-05,
      "loss": 12.5952,
      "step": 366
    },
    {
      "epoch": 0.07510103852253543,
      "grad_norm": 4.153656959533691,
      "learning_rate": 1.777713358205514e-05,
      "loss": 13.1257,
      "step": 367
    },
    {
      "epoch": 0.07530567350488566,
      "grad_norm": 4.324829578399658,
      "learning_rate": 1.7527597583490822e-05,
      "loss": 12.7625,
      "step": 368
    },
    {
      "epoch": 0.07551030848723589,
      "grad_norm": 4.624112129211426,
      "learning_rate": 1.7279452594569483e-05,
      "loss": 12.7958,
      "step": 369
    },
    {
      "epoch": 0.07571494346958613,
      "grad_norm": 4.383573055267334,
      "learning_rate": 1.703270924499656e-05,
      "loss": 12.8363,
      "step": 370
    },
    {
      "epoch": 0.07591957845193636,
      "grad_norm": 4.42855167388916,
      "learning_rate": 1.678737810443593e-05,
      "loss": 12.8649,
      "step": 371
    },
    {
      "epoch": 0.07612421343428659,
      "grad_norm": 4.5845947265625,
      "learning_rate": 1.6543469682057106e-05,
      "loss": 12.901,
      "step": 372
    },
    {
      "epoch": 0.07632884841663683,
      "grad_norm": 4.834083080291748,
      "learning_rate": 1.6300994426085103e-05,
      "loss": 13.347,
      "step": 373
    },
    {
      "epoch": 0.07653348339898705,
      "grad_norm": 4.795494079589844,
      "learning_rate": 1.605996272335291e-05,
      "loss": 12.8918,
      "step": 374
    },
    {
      "epoch": 0.07673811838133729,
      "grad_norm": 4.895383358001709,
      "learning_rate": 1.5820384898856434e-05,
      "loss": 13.162,
      "step": 375
    },
    {
      "epoch": 0.07694275336368753,
      "grad_norm": 4.894996166229248,
      "learning_rate": 1.5582271215312294e-05,
      "loss": 12.701,
      "step": 376
    },
    {
      "epoch": 0.07714738834603775,
      "grad_norm": 5.065547943115234,
      "learning_rate": 1.5345631872718214e-05,
      "loss": 12.8654,
      "step": 377
    },
    {
      "epoch": 0.07735202332838799,
      "grad_norm": 5.112913608551025,
      "learning_rate": 1.5110477007916001e-05,
      "loss": 12.766,
      "step": 378
    },
    {
      "epoch": 0.07755665831073823,
      "grad_norm": 5.340709209442139,
      "learning_rate": 1.4876816694157419e-05,
      "loss": 12.9361,
      "step": 379
    },
    {
      "epoch": 0.07776129329308845,
      "grad_norm": 5.618555068969727,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 13.0778,
      "step": 380
    },
    {
      "epoch": 0.07796592827543869,
      "grad_norm": 5.697518348693848,
      "learning_rate": 1.4414019692241437e-05,
      "loss": 12.582,
      "step": 381
    },
    {
      "epoch": 0.07817056325778891,
      "grad_norm": 5.6424241065979,
      "learning_rate": 1.4184902828767287e-05,
      "loss": 12.7521,
      "step": 382
    },
    {
      "epoch": 0.07837519824013915,
      "grad_norm": 6.131405830383301,
      "learning_rate": 1.3957320164854059e-05,
      "loss": 13.2849,
      "step": 383
    },
    {
      "epoch": 0.07857983322248939,
      "grad_norm": 5.919434547424316,
      "learning_rate": 1.373128144938563e-05,
      "loss": 13.0198,
      "step": 384
    },
    {
      "epoch": 0.07878446820483961,
      "grad_norm": 6.204239368438721,
      "learning_rate": 1.3506796365108232e-05,
      "loss": 13.2723,
      "step": 385
    },
    {
      "epoch": 0.07898910318718985,
      "grad_norm": 6.187657833099365,
      "learning_rate": 1.3283874528215733e-05,
      "loss": 12.9363,
      "step": 386
    },
    {
      "epoch": 0.07919373816954008,
      "grad_norm": 6.771162509918213,
      "learning_rate": 1.3062525487937699e-05,
      "loss": 12.9462,
      "step": 387
    },
    {
      "epoch": 0.07939837315189031,
      "grad_norm": 7.12640380859375,
      "learning_rate": 1.2842758726130283e-05,
      "loss": 13.2795,
      "step": 388
    },
    {
      "epoch": 0.07960300813424055,
      "grad_norm": 6.75380802154541,
      "learning_rate": 1.2624583656870154e-05,
      "loss": 12.9841,
      "step": 389
    },
    {
      "epoch": 0.07980764311659078,
      "grad_norm": 7.025509357452393,
      "learning_rate": 1.2408009626051137e-05,
      "loss": 12.963,
      "step": 390
    },
    {
      "epoch": 0.08001227809894101,
      "grad_norm": 7.841353893280029,
      "learning_rate": 1.2193045910983863e-05,
      "loss": 12.8249,
      "step": 391
    },
    {
      "epoch": 0.08021691308129125,
      "grad_norm": 7.905152320861816,
      "learning_rate": 1.1979701719998453e-05,
      "loss": 12.7228,
      "step": 392
    },
    {
      "epoch": 0.08042154806364148,
      "grad_norm": 7.849075794219971,
      "learning_rate": 1.1767986192049984e-05,
      "loss": 12.9185,
      "step": 393
    },
    {
      "epoch": 0.08062618304599171,
      "grad_norm": 9.404616355895996,
      "learning_rate": 1.1557908396327028e-05,
      "loss": 12.9398,
      "step": 394
    },
    {
      "epoch": 0.08083081802834194,
      "grad_norm": 9.282879829406738,
      "learning_rate": 1.134947733186315e-05,
      "loss": 12.8917,
      "step": 395
    },
    {
      "epoch": 0.08103545301069218,
      "grad_norm": 10.77182388305664,
      "learning_rate": 1.1142701927151456e-05,
      "loss": 13.7875,
      "step": 396
    },
    {
      "epoch": 0.08124008799304241,
      "grad_norm": 11.910287857055664,
      "learning_rate": 1.0937591039762085e-05,
      "loss": 13.1089,
      "step": 397
    },
    {
      "epoch": 0.08144472297539264,
      "grad_norm": 11.270553588867188,
      "learning_rate": 1.0734153455962765e-05,
      "loss": 12.87,
      "step": 398
    },
    {
      "epoch": 0.08164935795774288,
      "grad_norm": 12.47913646697998,
      "learning_rate": 1.0532397890342505e-05,
      "loss": 12.6082,
      "step": 399
    },
    {
      "epoch": 0.08185399294009311,
      "grad_norm": 19.859079360961914,
      "learning_rate": 1.0332332985438248e-05,
      "loss": 12.4463,
      "step": 400
    },
    {
      "epoch": 0.08185399294009311,
      "eval_loss": 3.1334729194641113,
      "eval_runtime": 47.3815,
      "eval_samples_per_second": 173.697,
      "eval_steps_per_second": 43.435,
      "step": 400
    },
    {
      "epoch": 0.08205862792244334,
      "grad_norm": 2.140819549560547,
      "learning_rate": 1.013396731136465e-05,
      "loss": 12.6548,
      "step": 401
    },
    {
      "epoch": 0.08226326290479358,
      "grad_norm": 2.7390999794006348,
      "learning_rate": 9.937309365446973e-06,
      "loss": 12.7752,
      "step": 402
    },
    {
      "epoch": 0.0824678978871438,
      "grad_norm": 2.9735829830169678,
      "learning_rate": 9.742367571857091e-06,
      "loss": 13.2948,
      "step": 403
    },
    {
      "epoch": 0.08267253286949404,
      "grad_norm": 2.9806323051452637,
      "learning_rate": 9.549150281252633e-06,
      "loss": 12.9468,
      "step": 404
    },
    {
      "epoch": 0.08287716785184428,
      "grad_norm": 2.9686119556427,
      "learning_rate": 9.357665770419244e-06,
      "loss": 12.8879,
      "step": 405
    },
    {
      "epoch": 0.0830818028341945,
      "grad_norm": 3.127723217010498,
      "learning_rate": 9.167922241916055e-06,
      "loss": 12.6892,
      "step": 406
    },
    {
      "epoch": 0.08328643781654474,
      "grad_norm": 3.3917040824890137,
      "learning_rate": 8.97992782372432e-06,
      "loss": 13.2438,
      "step": 407
    },
    {
      "epoch": 0.08349107279889498,
      "grad_norm": 3.28285551071167,
      "learning_rate": 8.793690568899216e-06,
      "loss": 12.7453,
      "step": 408
    },
    {
      "epoch": 0.0836957077812452,
      "grad_norm": 3.364295482635498,
      "learning_rate": 8.609218455224893e-06,
      "loss": 13.0133,
      "step": 409
    },
    {
      "epoch": 0.08390034276359544,
      "grad_norm": 3.374210834503174,
      "learning_rate": 8.426519384872733e-06,
      "loss": 12.79,
      "step": 410
    },
    {
      "epoch": 0.08410497774594566,
      "grad_norm": 3.6572799682617188,
      "learning_rate": 8.245601184062852e-06,
      "loss": 12.6433,
      "step": 411
    },
    {
      "epoch": 0.0843096127282959,
      "grad_norm": 3.5379762649536133,
      "learning_rate": 8.066471602728803e-06,
      "loss": 12.6151,
      "step": 412
    },
    {
      "epoch": 0.08451424771064614,
      "grad_norm": 3.6328377723693848,
      "learning_rate": 7.889138314185678e-06,
      "loss": 13.0676,
      "step": 413
    },
    {
      "epoch": 0.08471888269299636,
      "grad_norm": 4.000665664672852,
      "learning_rate": 7.71360891480134e-06,
      "loss": 13.2594,
      "step": 414
    },
    {
      "epoch": 0.0849235176753466,
      "grad_norm": 4.031978130340576,
      "learning_rate": 7.539890923671062e-06,
      "loss": 12.738,
      "step": 415
    },
    {
      "epoch": 0.08512815265769684,
      "grad_norm": 3.724813461303711,
      "learning_rate": 7.367991782295391e-06,
      "loss": 12.2267,
      "step": 416
    },
    {
      "epoch": 0.08533278764004706,
      "grad_norm": 4.113027095794678,
      "learning_rate": 7.197918854261432e-06,
      "loss": 12.7985,
      "step": 417
    },
    {
      "epoch": 0.0855374226223973,
      "grad_norm": 4.147072792053223,
      "learning_rate": 7.029679424927365e-06,
      "loss": 13.3496,
      "step": 418
    },
    {
      "epoch": 0.08574205760474753,
      "grad_norm": 4.0234527587890625,
      "learning_rate": 6.863280701110408e-06,
      "loss": 13.023,
      "step": 419
    },
    {
      "epoch": 0.08594669258709776,
      "grad_norm": 4.267332077026367,
      "learning_rate": 6.698729810778065e-06,
      "loss": 13.0938,
      "step": 420
    },
    {
      "epoch": 0.086151327569448,
      "grad_norm": 4.35993766784668,
      "learning_rate": 6.536033802742813e-06,
      "loss": 12.7928,
      "step": 421
    },
    {
      "epoch": 0.08635596255179823,
      "grad_norm": 4.47703218460083,
      "learning_rate": 6.375199646360142e-06,
      "loss": 12.9274,
      "step": 422
    },
    {
      "epoch": 0.08656059753414846,
      "grad_norm": 4.716027736663818,
      "learning_rate": 6.216234231230012e-06,
      "loss": 13.2614,
      "step": 423
    },
    {
      "epoch": 0.08676523251649869,
      "grad_norm": 4.96610164642334,
      "learning_rate": 6.059144366901736e-06,
      "loss": 12.2165,
      "step": 424
    },
    {
      "epoch": 0.08696986749884893,
      "grad_norm": 4.611530780792236,
      "learning_rate": 5.903936782582253e-06,
      "loss": 12.7755,
      "step": 425
    },
    {
      "epoch": 0.08717450248119916,
      "grad_norm": 4.9565300941467285,
      "learning_rate": 5.750618126847912e-06,
      "loss": 13.0633,
      "step": 426
    },
    {
      "epoch": 0.08737913746354939,
      "grad_norm": 5.047351837158203,
      "learning_rate": 5.599194967359639e-06,
      "loss": 12.7034,
      "step": 427
    },
    {
      "epoch": 0.08758377244589963,
      "grad_norm": 4.904860019683838,
      "learning_rate": 5.449673790581611e-06,
      "loss": 12.5122,
      "step": 428
    },
    {
      "epoch": 0.08778840742824986,
      "grad_norm": 5.229506015777588,
      "learning_rate": 5.302061001503394e-06,
      "loss": 13.0031,
      "step": 429
    },
    {
      "epoch": 0.08799304241060009,
      "grad_norm": 5.458662509918213,
      "learning_rate": 5.156362923365588e-06,
      "loss": 12.8912,
      "step": 430
    },
    {
      "epoch": 0.08819767739295033,
      "grad_norm": 5.457494735717773,
      "learning_rate": 5.012585797388936e-06,
      "loss": 13.3934,
      "step": 431
    },
    {
      "epoch": 0.08840231237530055,
      "grad_norm": 5.259174823760986,
      "learning_rate": 4.87073578250698e-06,
      "loss": 12.2529,
      "step": 432
    },
    {
      "epoch": 0.08860694735765079,
      "grad_norm": 5.968120574951172,
      "learning_rate": 4.730818955102234e-06,
      "loss": 12.54,
      "step": 433
    },
    {
      "epoch": 0.08881158234000103,
      "grad_norm": 6.257950782775879,
      "learning_rate": 4.592841308745932e-06,
      "loss": 12.6848,
      "step": 434
    },
    {
      "epoch": 0.08901621732235125,
      "grad_norm": 6.009523868560791,
      "learning_rate": 4.456808753941205e-06,
      "loss": 13.0703,
      "step": 435
    },
    {
      "epoch": 0.08922085230470149,
      "grad_norm": 6.272739887237549,
      "learning_rate": 4.322727117869951e-06,
      "loss": 12.3788,
      "step": 436
    },
    {
      "epoch": 0.08942548728705173,
      "grad_norm": 6.675440788269043,
      "learning_rate": 4.190602144143207e-06,
      "loss": 13.0829,
      "step": 437
    },
    {
      "epoch": 0.08963012226940195,
      "grad_norm": 6.91643762588501,
      "learning_rate": 4.06043949255509e-06,
      "loss": 13.3085,
      "step": 438
    },
    {
      "epoch": 0.08983475725175219,
      "grad_norm": 7.155692100524902,
      "learning_rate": 3.932244738840379e-06,
      "loss": 12.7579,
      "step": 439
    },
    {
      "epoch": 0.09003939223410241,
      "grad_norm": 6.8719940185546875,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 12.96,
      "step": 440
    },
    {
      "epoch": 0.09024402721645265,
      "grad_norm": 7.757195949554443,
      "learning_rate": 3.681780806244095e-06,
      "loss": 12.8733,
      "step": 441
    },
    {
      "epoch": 0.09044866219880289,
      "grad_norm": 7.0664215087890625,
      "learning_rate": 3.5595223564037884e-06,
      "loss": 12.6288,
      "step": 442
    },
    {
      "epoch": 0.09065329718115311,
      "grad_norm": 8.440971374511719,
      "learning_rate": 3.4392532620598216e-06,
      "loss": 13.3932,
      "step": 443
    },
    {
      "epoch": 0.09085793216350335,
      "grad_norm": 8.004168510437012,
      "learning_rate": 3.3209786751399187e-06,
      "loss": 12.6128,
      "step": 444
    },
    {
      "epoch": 0.09106256714585359,
      "grad_norm": 8.411291122436523,
      "learning_rate": 3.2047036621337236e-06,
      "loss": 13.4134,
      "step": 445
    },
    {
      "epoch": 0.09126720212820381,
      "grad_norm": 9.257599830627441,
      "learning_rate": 3.0904332038757977e-06,
      "loss": 12.8954,
      "step": 446
    },
    {
      "epoch": 0.09147183711055405,
      "grad_norm": 9.609465599060059,
      "learning_rate": 2.978172195332263e-06,
      "loss": 13.3849,
      "step": 447
    },
    {
      "epoch": 0.09167647209290428,
      "grad_norm": 11.43130111694336,
      "learning_rate": 2.8679254453910785e-06,
      "loss": 12.0155,
      "step": 448
    },
    {
      "epoch": 0.09188110707525451,
      "grad_norm": 12.757843971252441,
      "learning_rate": 2.759697676656098e-06,
      "loss": 12.183,
      "step": 449
    },
    {
      "epoch": 0.09208574205760475,
      "grad_norm": 17.703031539916992,
      "learning_rate": 2.653493525244721e-06,
      "loss": 13.8012,
      "step": 450
    },
    {
      "epoch": 0.09229037703995498,
      "grad_norm": 2.267075300216675,
      "learning_rate": 2.549317540589308e-06,
      "loss": 12.4868,
      "step": 451
    },
    {
      "epoch": 0.09249501202230521,
      "grad_norm": 2.5427181720733643,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 12.5128,
      "step": 452
    },
    {
      "epoch": 0.09269964700465545,
      "grad_norm": 2.599229097366333,
      "learning_rate": 2.3470678346851518e-06,
      "loss": 12.9201,
      "step": 453
    },
    {
      "epoch": 0.09290428198700568,
      "grad_norm": 2.6054906845092773,
      "learning_rate": 2.2490027771406687e-06,
      "loss": 12.6195,
      "step": 454
    },
    {
      "epoch": 0.09310891696935591,
      "grad_norm": 2.9035027027130127,
      "learning_rate": 2.152983213389559e-06,
      "loss": 12.7726,
      "step": 455
    },
    {
      "epoch": 0.09331355195170614,
      "grad_norm": 3.060668468475342,
      "learning_rate": 2.0590132565903476e-06,
      "loss": 13.0542,
      "step": 456
    },
    {
      "epoch": 0.09351818693405638,
      "grad_norm": 3.1342105865478516,
      "learning_rate": 1.9670969321032407e-06,
      "loss": 12.9032,
      "step": 457
    },
    {
      "epoch": 0.09372282191640662,
      "grad_norm": 3.196115493774414,
      "learning_rate": 1.8772381773176417e-06,
      "loss": 13.1882,
      "step": 458
    },
    {
      "epoch": 0.09392745689875684,
      "grad_norm": 3.232515335083008,
      "learning_rate": 1.7894408414835362e-06,
      "loss": 12.8594,
      "step": 459
    },
    {
      "epoch": 0.09413209188110708,
      "grad_norm": 3.358297348022461,
      "learning_rate": 1.70370868554659e-06,
      "loss": 12.909,
      "step": 460
    },
    {
      "epoch": 0.0943367268634573,
      "grad_norm": 3.49599027633667,
      "learning_rate": 1.620045381987012e-06,
      "loss": 12.7092,
      "step": 461
    },
    {
      "epoch": 0.09454136184580754,
      "grad_norm": 3.4245402812957764,
      "learning_rate": 1.5384545146622852e-06,
      "loss": 12.656,
      "step": 462
    },
    {
      "epoch": 0.09474599682815778,
      "grad_norm": 3.7484071254730225,
      "learning_rate": 1.4589395786535953e-06,
      "loss": 12.6428,
      "step": 463
    },
    {
      "epoch": 0.094950631810508,
      "grad_norm": 3.6963798999786377,
      "learning_rate": 1.3815039801161721e-06,
      "loss": 12.75,
      "step": 464
    },
    {
      "epoch": 0.09515526679285824,
      "grad_norm": 3.9733211994171143,
      "learning_rate": 1.3061510361333185e-06,
      "loss": 12.9671,
      "step": 465
    },
    {
      "epoch": 0.09535990177520848,
      "grad_norm": 3.8950541019439697,
      "learning_rate": 1.232883974574367e-06,
      "loss": 12.7971,
      "step": 466
    },
    {
      "epoch": 0.0955645367575587,
      "grad_norm": 4.091780185699463,
      "learning_rate": 1.1617059339563807e-06,
      "loss": 12.4321,
      "step": 467
    },
    {
      "epoch": 0.09576917173990894,
      "grad_norm": 4.282808303833008,
      "learning_rate": 1.0926199633097157e-06,
      "loss": 13.4677,
      "step": 468
    },
    {
      "epoch": 0.09597380672225916,
      "grad_norm": 4.246225357055664,
      "learning_rate": 1.0256290220474307e-06,
      "loss": 12.7789,
      "step": 469
    },
    {
      "epoch": 0.0961784417046094,
      "grad_norm": 4.31233549118042,
      "learning_rate": 9.607359798384785e-07,
      "loss": 12.7813,
      "step": 470
    },
    {
      "epoch": 0.09638307668695964,
      "grad_norm": 4.715573787689209,
      "learning_rate": 8.979436164848088e-07,
      "loss": 13.1816,
      "step": 471
    },
    {
      "epoch": 0.09658771166930986,
      "grad_norm": 4.476752281188965,
      "learning_rate": 8.372546218022747e-07,
      "loss": 13.1693,
      "step": 472
    },
    {
      "epoch": 0.0967923466516601,
      "grad_norm": 4.613335132598877,
      "learning_rate": 7.786715955054203e-07,
      "loss": 12.6692,
      "step": 473
    },
    {
      "epoch": 0.09699698163401034,
      "grad_norm": 4.564615249633789,
      "learning_rate": 7.221970470961125e-07,
      "loss": 12.6351,
      "step": 474
    },
    {
      "epoch": 0.09720161661636056,
      "grad_norm": 4.650303840637207,
      "learning_rate": 6.678333957560512e-07,
      "loss": 13.1352,
      "step": 475
    },
    {
      "epoch": 0.0974062515987108,
      "grad_norm": 4.742563724517822,
      "learning_rate": 6.15582970243117e-07,
      "loss": 12.8478,
      "step": 476
    },
    {
      "epoch": 0.09761088658106103,
      "grad_norm": 5.02756929397583,
      "learning_rate": 5.654480087916303e-07,
      "loss": 12.4434,
      "step": 477
    },
    {
      "epoch": 0.09781552156341126,
      "grad_norm": 5.11809778213501,
      "learning_rate": 5.174306590164879e-07,
      "loss": 12.5754,
      "step": 478
    },
    {
      "epoch": 0.0980201565457615,
      "grad_norm": 5.026023864746094,
      "learning_rate": 4.715329778211375e-07,
      "loss": 12.7589,
      "step": 479
    },
    {
      "epoch": 0.09822479152811173,
      "grad_norm": 5.174000263214111,
      "learning_rate": 4.277569313094809e-07,
      "loss": 12.4781,
      "step": 480
    },
    {
      "epoch": 0.09842942651046196,
      "grad_norm": 5.73392391204834,
      "learning_rate": 3.8610439470164737e-07,
      "loss": 13.2662,
      "step": 481
    },
    {
      "epoch": 0.0986340614928122,
      "grad_norm": 5.5650224685668945,
      "learning_rate": 3.465771522536854e-07,
      "loss": 12.698,
      "step": 482
    },
    {
      "epoch": 0.09883869647516243,
      "grad_norm": 5.415269374847412,
      "learning_rate": 3.09176897181096e-07,
      "loss": 12.4908,
      "step": 483
    },
    {
      "epoch": 0.09904333145751266,
      "grad_norm": 5.8570170402526855,
      "learning_rate": 2.7390523158633554e-07,
      "loss": 12.5447,
      "step": 484
    },
    {
      "epoch": 0.09924796643986289,
      "grad_norm": 6.1962361335754395,
      "learning_rate": 2.407636663901591e-07,
      "loss": 13.4094,
      "step": 485
    },
    {
      "epoch": 0.09945260142221313,
      "grad_norm": 6.143553256988525,
      "learning_rate": 2.0975362126691712e-07,
      "loss": 12.9071,
      "step": 486
    },
    {
      "epoch": 0.09965723640456337,
      "grad_norm": 6.178109645843506,
      "learning_rate": 1.8087642458373134e-07,
      "loss": 12.7015,
      "step": 487
    },
    {
      "epoch": 0.09986187138691359,
      "grad_norm": 6.477899551391602,
      "learning_rate": 1.5413331334360182e-07,
      "loss": 12.9638,
      "step": 488
    },
    {
      "epoch": 0.10006650636926383,
      "grad_norm": 6.375460624694824,
      "learning_rate": 1.2952543313240472e-07,
      "loss": 12.3259,
      "step": 489
    },
    {
      "epoch": 0.10027114135161405,
      "grad_norm": 7.033268928527832,
      "learning_rate": 1.0705383806982606e-07,
      "loss": 12.8778,
      "step": 490
    },
    {
      "epoch": 0.10047577633396429,
      "grad_norm": 6.597715377807617,
      "learning_rate": 8.671949076420882e-08,
      "loss": 12.0536,
      "step": 491
    },
    {
      "epoch": 0.10068041131631453,
      "grad_norm": 7.840211868286133,
      "learning_rate": 6.852326227130834e-08,
      "loss": 13.5376,
      "step": 492
    },
    {
      "epoch": 0.10088504629866475,
      "grad_norm": 7.423345565795898,
      "learning_rate": 5.246593205699424e-08,
      "loss": 12.8476,
      "step": 493
    },
    {
      "epoch": 0.10108968128101499,
      "grad_norm": 8.046225547790527,
      "learning_rate": 3.8548187963854956e-08,
      "loss": 12.5719,
      "step": 494
    },
    {
      "epoch": 0.10129431626336523,
      "grad_norm": 8.249234199523926,
      "learning_rate": 2.6770626181715773e-08,
      "loss": 12.1992,
      "step": 495
    },
    {
      "epoch": 0.10149895124571545,
      "grad_norm": 9.472099304199219,
      "learning_rate": 1.7133751222137007e-08,
      "loss": 12.8579,
      "step": 496
    },
    {
      "epoch": 0.10170358622806569,
      "grad_norm": 9.204952239990234,
      "learning_rate": 9.637975896759077e-09,
      "loss": 12.0369,
      "step": 497
    },
    {
      "epoch": 0.10190822121041591,
      "grad_norm": 11.747493743896484,
      "learning_rate": 4.2836212996499865e-09,
      "loss": 13.0472,
      "step": 498
    },
    {
      "epoch": 0.10211285619276615,
      "grad_norm": 12.529196739196777,
      "learning_rate": 1.0709167935385455e-09,
      "loss": 12.7862,
      "step": 499
    },
    {
      "epoch": 0.10231749117511639,
      "grad_norm": 19.44767951965332,
      "learning_rate": 0.0,
      "loss": 13.4285,
      "step": 500
    },
    {
      "epoch": 0.10231749117511639,
      "eval_loss": 3.130511999130249,
      "eval_runtime": 47.3503,
      "eval_samples_per_second": 173.811,
      "eval_steps_per_second": 43.463,
      "step": 500
    }
  ],
  "logging_steps": 1,
  "max_steps": 500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9442642165235712.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}