{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 21900,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0045662100456621,
      "grad_norm": 22.38092803955078,
      "learning_rate": 2.993150684931507e-05,
      "loss": 5.0082,
      "step": 50
    },
    {
      "epoch": 0.0091324200913242,
      "grad_norm": 19.525426864624023,
      "learning_rate": 2.9863013698630136e-05,
      "loss": 4.0853,
      "step": 100
    },
    {
      "epoch": 0.0136986301369863,
      "grad_norm": 29.21298599243164,
      "learning_rate": 2.9794520547945206e-05,
      "loss": 3.8879,
      "step": 150
    },
    {
      "epoch": 0.0182648401826484,
      "grad_norm": 29.60974884033203,
      "learning_rate": 2.9726027397260275e-05,
      "loss": 3.5946,
      "step": 200
    },
    {
      "epoch": 0.0228310502283105,
      "grad_norm": 20.75732421875,
      "learning_rate": 2.9657534246575345e-05,
      "loss": 3.1165,
      "step": 250
    },
    {
      "epoch": 0.0273972602739726,
      "grad_norm": 31.49699592590332,
      "learning_rate": 2.958904109589041e-05,
      "loss": 2.7346,
      "step": 300
    },
    {
      "epoch": 0.0319634703196347,
      "grad_norm": 43.215335845947266,
      "learning_rate": 2.952054794520548e-05,
      "loss": 2.3509,
      "step": 350
    },
    {
      "epoch": 0.0365296803652968,
      "grad_norm": 29.553388595581055,
      "learning_rate": 2.945205479452055e-05,
      "loss": 2.3374,
      "step": 400
    },
    {
      "epoch": 0.0410958904109589,
      "grad_norm": 30.272083282470703,
      "learning_rate": 2.938356164383562e-05,
      "loss": 2.0551,
      "step": 450
    },
    {
      "epoch": 0.045662100456621,
      "grad_norm": 23.157560348510742,
      "learning_rate": 2.9315068493150685e-05,
      "loss": 1.9659,
      "step": 500
    },
    {
      "epoch": 0.0502283105022831,
      "grad_norm": 39.72032165527344,
      "learning_rate": 2.9246575342465755e-05,
      "loss": 1.7552,
      "step": 550
    },
    {
      "epoch": 0.0547945205479452,
      "grad_norm": 24.441036224365234,
      "learning_rate": 2.9178082191780824e-05,
      "loss": 1.7992,
      "step": 600
    },
    {
      "epoch": 0.0593607305936073,
      "grad_norm": 32.07455825805664,
      "learning_rate": 2.910958904109589e-05,
      "loss": 1.6968,
      "step": 650
    },
    {
      "epoch": 0.0639269406392694,
      "grad_norm": 25.034423828125,
      "learning_rate": 2.9041095890410956e-05,
      "loss": 1.788,
      "step": 700
    },
    {
      "epoch": 0.0684931506849315,
      "grad_norm": 33.06129455566406,
      "learning_rate": 2.8972602739726026e-05,
      "loss": 1.8157,
      "step": 750
    },
    {
      "epoch": 0.0730593607305936,
      "grad_norm": 32.716251373291016,
      "learning_rate": 2.8904109589041095e-05,
      "loss": 1.7502,
      "step": 800
    },
    {
      "epoch": 0.0776255707762557,
      "grad_norm": 27.18515968322754,
      "learning_rate": 2.8835616438356165e-05,
      "loss": 1.5629,
      "step": 850
    },
    {
      "epoch": 0.0821917808219178,
      "grad_norm": 19.844303131103516,
      "learning_rate": 2.876712328767123e-05,
      "loss": 1.6421,
      "step": 900
    },
    {
      "epoch": 0.0867579908675799,
      "grad_norm": 160.48770141601562,
      "learning_rate": 2.86986301369863e-05,
      "loss": 1.6035,
      "step": 950
    },
    {
      "epoch": 0.091324200913242,
      "grad_norm": 35.13261032104492,
      "learning_rate": 2.863013698630137e-05,
      "loss": 1.4939,
      "step": 1000
    },
    {
      "epoch": 0.0958904109589041,
      "grad_norm": 24.28845977783203,
      "learning_rate": 2.856164383561644e-05,
      "loss": 1.5611,
      "step": 1050
    },
    {
      "epoch": 0.1004566210045662,
      "grad_norm": 30.07063865661621,
      "learning_rate": 2.8493150684931505e-05,
      "loss": 1.4715,
      "step": 1100
    },
    {
      "epoch": 0.1050228310502283,
      "grad_norm": 23.672300338745117,
      "learning_rate": 2.8424657534246575e-05,
      "loss": 1.534,
      "step": 1150
    },
    {
      "epoch": 0.1095890410958904,
      "grad_norm": 27.45283317565918,
      "learning_rate": 2.8356164383561644e-05,
      "loss": 1.5268,
      "step": 1200
    },
    {
      "epoch": 0.1141552511415525,
      "grad_norm": 26.83701515197754,
      "learning_rate": 2.8287671232876714e-05,
      "loss": 1.4089,
      "step": 1250
    },
    {
      "epoch": 0.1187214611872146,
      "grad_norm": 19.307844161987305,
      "learning_rate": 2.821917808219178e-05,
      "loss": 1.4929,
      "step": 1300
    },
    {
      "epoch": 0.1232876712328767,
      "grad_norm": 14.790838241577148,
      "learning_rate": 2.815068493150685e-05,
      "loss": 1.4687,
      "step": 1350
    },
    {
      "epoch": 0.1278538812785388,
      "grad_norm": 16.656993865966797,
      "learning_rate": 2.808219178082192e-05,
      "loss": 1.3456,
      "step": 1400
    },
    {
      "epoch": 0.1324200913242009,
      "grad_norm": 14.602429389953613,
      "learning_rate": 2.801369863013699e-05,
      "loss": 1.2725,
      "step": 1450
    },
    {
      "epoch": 0.136986301369863,
      "grad_norm": 23.853199005126953,
      "learning_rate": 2.7945205479452054e-05,
      "loss": 1.2684,
      "step": 1500
    },
    {
      "epoch": 0.1415525114155251,
      "grad_norm": 40.28663635253906,
      "learning_rate": 2.7876712328767124e-05,
      "loss": 1.2728,
      "step": 1550
    },
    {
      "epoch": 0.1461187214611872,
      "grad_norm": 11.038789749145508,
      "learning_rate": 2.7808219178082193e-05,
      "loss": 1.3578,
      "step": 1600
    },
    {
      "epoch": 0.1506849315068493,
      "grad_norm": 22.868593215942383,
      "learning_rate": 2.7739726027397263e-05,
      "loss": 1.3941,
      "step": 1650
    },
    {
      "epoch": 0.1552511415525114,
      "grad_norm": 23.89753532409668,
      "learning_rate": 2.767123287671233e-05,
      "loss": 1.3071,
      "step": 1700
    },
    {
      "epoch": 0.1598173515981735,
      "grad_norm": 40.02177429199219,
      "learning_rate": 2.76027397260274e-05,
      "loss": 1.3711,
      "step": 1750
    },
    {
      "epoch": 0.1643835616438356,
      "grad_norm": 40.53044128417969,
      "learning_rate": 2.7534246575342468e-05,
      "loss": 1.1694,
      "step": 1800
    },
    {
      "epoch": 0.1689497716894977,
      "grad_norm": 38.70663070678711,
      "learning_rate": 2.7465753424657537e-05,
      "loss": 1.2284,
      "step": 1850
    },
    {
      "epoch": 0.1735159817351598,
      "grad_norm": 114.71417236328125,
      "learning_rate": 2.7397260273972603e-05,
      "loss": 1.1634,
      "step": 1900
    },
    {
      "epoch": 0.1780821917808219,
      "grad_norm": 34.596923828125,
      "learning_rate": 2.732876712328767e-05,
      "loss": 1.2858,
      "step": 1950
    },
    {
      "epoch": 0.182648401826484,
      "grad_norm": 23.52661895751953,
      "learning_rate": 2.726027397260274e-05,
      "loss": 1.2273,
      "step": 2000
    },
    {
      "epoch": 0.1872146118721461,
      "grad_norm": 22.8360652923584,
      "learning_rate": 2.719178082191781e-05,
      "loss": 1.2622,
      "step": 2050
    },
    {
      "epoch": 0.1917808219178082,
      "grad_norm": 17.08322525024414,
      "learning_rate": 2.7123287671232875e-05,
      "loss": 1.3012,
      "step": 2100
    },
    {
      "epoch": 0.1963470319634703,
      "grad_norm": 12.923150062561035,
      "learning_rate": 2.7054794520547944e-05,
      "loss": 1.3847,
      "step": 2150
    },
    {
      "epoch": 0.2009132420091324,
      "grad_norm": 73.03246307373047,
      "learning_rate": 2.6986301369863014e-05,
      "loss": 1.3173,
      "step": 2200
    },
    {
      "epoch": 0.2054794520547945,
      "grad_norm": 16.656768798828125,
      "learning_rate": 2.6917808219178083e-05,
      "loss": 1.1779,
      "step": 2250
    },
    {
      "epoch": 0.2100456621004566,
      "grad_norm": 22.919973373413086,
      "learning_rate": 2.684931506849315e-05,
      "loss": 1.2497,
      "step": 2300
    },
    {
      "epoch": 0.2146118721461187,
      "grad_norm": 20.0987491607666,
      "learning_rate": 2.678082191780822e-05,
      "loss": 1.1787,
      "step": 2350
    },
    {
      "epoch": 0.2191780821917808,
      "grad_norm": 84.04356384277344,
      "learning_rate": 2.6712328767123288e-05,
      "loss": 1.1989,
      "step": 2400
    },
    {
      "epoch": 0.2237442922374429,
      "grad_norm": 12.585476875305176,
      "learning_rate": 2.6643835616438358e-05,
      "loss": 1.2324,
      "step": 2450
    },
    {
      "epoch": 0.228310502283105,
      "grad_norm": 18.206226348876953,
      "learning_rate": 2.6575342465753424e-05,
      "loss": 1.2541,
      "step": 2500
    },
    {
      "epoch": 0.2328767123287671,
      "grad_norm": 24.236295700073242,
      "learning_rate": 2.6506849315068493e-05,
      "loss": 1.1755,
      "step": 2550
    },
    {
      "epoch": 0.2374429223744292,
      "grad_norm": 21.91484260559082,
      "learning_rate": 2.6438356164383563e-05,
      "loss": 1.1436,
      "step": 2600
    },
    {
      "epoch": 0.2420091324200913,
      "grad_norm": 25.741945266723633,
      "learning_rate": 2.6369863013698632e-05,
      "loss": 1.146,
      "step": 2650
    },
    {
      "epoch": 0.2465753424657534,
      "grad_norm": 21.776227951049805,
      "learning_rate": 2.6301369863013698e-05,
      "loss": 1.1521,
      "step": 2700
    },
    {
      "epoch": 0.2511415525114155,
      "grad_norm": 30.281368255615234,
      "learning_rate": 2.6232876712328768e-05,
      "loss": 1.0747,
      "step": 2750
    },
    {
      "epoch": 0.2557077625570776,
      "grad_norm": 16.21628189086914,
      "learning_rate": 2.6164383561643837e-05,
      "loss": 1.2503,
      "step": 2800
    },
    {
      "epoch": 0.2602739726027397,
      "grad_norm": 17.022933959960938,
      "learning_rate": 2.6095890410958907e-05,
      "loss": 1.1719,
      "step": 2850
    },
    {
      "epoch": 0.2648401826484018,
      "grad_norm": 17.665477752685547,
      "learning_rate": 2.6027397260273973e-05,
      "loss": 1.091,
      "step": 2900
    },
    {
      "epoch": 0.2694063926940639,
      "grad_norm": 12.710230827331543,
      "learning_rate": 2.5958904109589042e-05,
      "loss": 1.1828,
      "step": 2950
    },
    {
      "epoch": 0.273972602739726,
      "grad_norm": 23.11914825439453,
      "learning_rate": 2.589041095890411e-05,
      "loss": 1.0025,
      "step": 3000
    },
    {
      "epoch": 0.2785388127853881,
      "grad_norm": 17.97935676574707,
      "learning_rate": 2.582191780821918e-05,
      "loss": 1.2145,
      "step": 3050
    },
    {
      "epoch": 0.2831050228310502,
      "grad_norm": 28.123554229736328,
      "learning_rate": 2.5753424657534247e-05,
      "loss": 1.2629,
      "step": 3100
    },
    {
      "epoch": 0.2876712328767123,
      "grad_norm": 17.924665451049805,
      "learning_rate": 2.5684931506849317e-05,
      "loss": 1.0439,
      "step": 3150
    },
    {
      "epoch": 0.2922374429223744,
      "grad_norm": 24.03557777404785,
      "learning_rate": 2.5616438356164386e-05,
      "loss": 1.127,
      "step": 3200
    },
    {
      "epoch": 0.2968036529680365,
      "grad_norm": 30.70620346069336,
      "learning_rate": 2.5547945205479452e-05,
      "loss": 1.1464,
      "step": 3250
    },
    {
      "epoch": 0.3013698630136986,
      "grad_norm": 21.80393409729004,
      "learning_rate": 2.5479452054794518e-05,
      "loss": 0.9603,
      "step": 3300
    },
    {
      "epoch": 0.3059360730593607,
      "grad_norm": 23.686079025268555,
      "learning_rate": 2.5410958904109588e-05,
      "loss": 1.0745,
      "step": 3350
    },
    {
      "epoch": 0.3105022831050228,
      "grad_norm": 10.258773803710938,
      "learning_rate": 2.5342465753424657e-05,
      "loss": 1.0776,
      "step": 3400
    },
    {
      "epoch": 0.3150684931506849,
      "grad_norm": 28.77837562561035,
      "learning_rate": 2.5273972602739727e-05,
      "loss": 1.1781,
      "step": 3450
    },
    {
      "epoch": 0.319634703196347,
      "grad_norm": 39.38608932495117,
      "learning_rate": 2.5205479452054793e-05,
      "loss": 0.9749,
      "step": 3500
    },
    {
      "epoch": 0.3242009132420091,
      "grad_norm": 29.596742630004883,
      "learning_rate": 2.5136986301369862e-05,
      "loss": 1.1099,
      "step": 3550
    },
    {
      "epoch": 0.3287671232876712,
      "grad_norm": 15.371848106384277,
      "learning_rate": 2.5068493150684932e-05,
      "loss": 1.1363,
      "step": 3600
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 27.008398056030273,
      "learning_rate": 2.5e-05,
      "loss": 1.2275,
      "step": 3650
    },
    {
      "epoch": 0.3378995433789954,
      "grad_norm": 54.85331726074219,
      "learning_rate": 2.4931506849315067e-05,
      "loss": 1.2297,
      "step": 3700
    },
    {
      "epoch": 0.3424657534246575,
      "grad_norm": 10.650010108947754,
      "learning_rate": 2.4863013698630137e-05,
      "loss": 1.1716,
      "step": 3750
    },
    {
      "epoch": 0.3470319634703196,
      "grad_norm": 21.537841796875,
      "learning_rate": 2.4794520547945206e-05,
      "loss": 1.2375,
      "step": 3800
    },
    {
      "epoch": 0.3515981735159817,
      "grad_norm": 31.040218353271484,
      "learning_rate": 2.4726027397260276e-05,
      "loss": 1.0511,
      "step": 3850
    },
    {
      "epoch": 0.3561643835616438,
      "grad_norm": 13.31843090057373,
      "learning_rate": 2.4657534246575342e-05,
      "loss": 1.1581,
      "step": 3900
    },
    {
      "epoch": 0.3607305936073059,
      "grad_norm": 22.53802490234375,
      "learning_rate": 2.458904109589041e-05,
      "loss": 1.0627,
      "step": 3950
    },
    {
      "epoch": 0.365296803652968,
      "grad_norm": 20.266468048095703,
      "learning_rate": 2.452054794520548e-05,
      "loss": 1.1103,
      "step": 4000
    },
    {
      "epoch": 0.3698630136986301,
      "grad_norm": 9.198668479919434,
      "learning_rate": 2.445205479452055e-05,
      "loss": 1.136,
      "step": 4050
    },
    {
      "epoch": 0.3744292237442922,
      "grad_norm": 13.770018577575684,
      "learning_rate": 2.4383561643835616e-05,
      "loss": 1.0984,
      "step": 4100
    },
    {
      "epoch": 0.3789954337899543,
      "grad_norm": 25.558441162109375,
      "learning_rate": 2.4315068493150686e-05,
      "loss": 1.062,
      "step": 4150
    },
    {
      "epoch": 0.3835616438356164,
      "grad_norm": 25.940107345581055,
      "learning_rate": 2.4246575342465755e-05,
      "loss": 1.2237,
      "step": 4200
    },
    {
      "epoch": 0.3881278538812785,
      "grad_norm": 11.530945777893066,
      "learning_rate": 2.4178082191780825e-05,
      "loss": 1.0717,
      "step": 4250
    },
    {
      "epoch": 0.3926940639269406,
      "grad_norm": 13.686588287353516,
      "learning_rate": 2.410958904109589e-05,
      "loss": 1.0011,
      "step": 4300
    },
    {
      "epoch": 0.3972602739726027,
      "grad_norm": 24.325557708740234,
      "learning_rate": 2.404109589041096e-05,
      "loss": 1.2033,
      "step": 4350
    },
    {
      "epoch": 0.4018264840182648,
      "grad_norm": 24.21858787536621,
      "learning_rate": 2.397260273972603e-05,
      "loss": 1.01,
      "step": 4400
    },
    {
      "epoch": 0.4063926940639269,
      "grad_norm": 14.875985145568848,
      "learning_rate": 2.39041095890411e-05,
      "loss": 1.0378,
      "step": 4450
    },
    {
      "epoch": 0.410958904109589,
      "grad_norm": 38.33986282348633,
      "learning_rate": 2.3835616438356165e-05,
      "loss": 1.0722,
      "step": 4500
    },
    {
      "epoch": 0.4155251141552511,
      "grad_norm": 10.655511856079102,
      "learning_rate": 2.376712328767123e-05,
      "loss": 1.1182,
      "step": 4550
    },
    {
      "epoch": 0.4200913242009132,
      "grad_norm": 26.336122512817383,
      "learning_rate": 2.36986301369863e-05,
      "loss": 1.0772,
      "step": 4600
    },
    {
      "epoch": 0.4246575342465753,
      "grad_norm": 26.254093170166016,
      "learning_rate": 2.363013698630137e-05,
      "loss": 0.9908,
      "step": 4650
    },
    {
      "epoch": 0.4292237442922374,
      "grad_norm": 7.081357002258301,
      "learning_rate": 2.3561643835616436e-05,
      "loss": 1.0079,
      "step": 4700
    },
    {
      "epoch": 0.4337899543378995,
      "grad_norm": 20.019088745117188,
      "learning_rate": 2.3493150684931506e-05,
      "loss": 1.1188,
      "step": 4750
    },
    {
      "epoch": 0.4383561643835616,
      "grad_norm": 30.188098907470703,
      "learning_rate": 2.3424657534246575e-05,
      "loss": 1.0616,
      "step": 4800
    },
    {
      "epoch": 0.4429223744292237,
      "grad_norm": 6.621674537658691,
      "learning_rate": 2.3356164383561645e-05,
      "loss": 1.0935,
      "step": 4850
    },
    {
      "epoch": 0.4474885844748858,
      "grad_norm": 21.145673751831055,
      "learning_rate": 2.328767123287671e-05,
      "loss": 1.0178,
      "step": 4900
    },
    {
      "epoch": 0.4520547945205479,
      "grad_norm": 26.22977066040039,
      "learning_rate": 2.321917808219178e-05,
      "loss": 1.1739,
      "step": 4950
    },
    {
      "epoch": 0.45662100456621,
      "grad_norm": 22.71933364868164,
      "learning_rate": 2.315068493150685e-05,
      "loss": 1.122,
      "step": 5000
    },
    {
      "epoch": 0.4611872146118721,
      "grad_norm": 16.020483016967773,
      "learning_rate": 2.308219178082192e-05,
      "loss": 1.0494,
      "step": 5050
    },
    {
      "epoch": 0.4657534246575342,
      "grad_norm": 10.217668533325195,
      "learning_rate": 2.3013698630136985e-05,
      "loss": 1.1577,
      "step": 5100
    },
    {
      "epoch": 0.4703196347031963,
      "grad_norm": 13.561128616333008,
      "learning_rate": 2.2945205479452055e-05,
      "loss": 1.1065,
      "step": 5150
    },
    {
      "epoch": 0.4748858447488584,
      "grad_norm": 24.07544708251953,
      "learning_rate": 2.2876712328767124e-05,
      "loss": 1.1721,
      "step": 5200
    },
    {
      "epoch": 0.4794520547945205,
      "grad_norm": 27.471532821655273,
      "learning_rate": 2.2808219178082194e-05,
      "loss": 1.0425,
      "step": 5250
    },
    {
      "epoch": 0.4840182648401826,
      "grad_norm": 39.17247772216797,
      "learning_rate": 2.273972602739726e-05,
      "loss": 1.0182,
      "step": 5300
    },
    {
      "epoch": 0.4885844748858447,
      "grad_norm": 11.253133773803711,
      "learning_rate": 2.267123287671233e-05,
      "loss": 1.0516,
      "step": 5350
    },
    {
      "epoch": 0.4931506849315068,
      "grad_norm": 19.816268920898438,
      "learning_rate": 2.26027397260274e-05,
      "loss": 0.9359,
      "step": 5400
    },
    {
      "epoch": 0.4977168949771689,
      "grad_norm": 21.89589500427246,
      "learning_rate": 2.253424657534247e-05,
      "loss": 1.0047,
      "step": 5450
    },
    {
      "epoch": 0.502283105022831,
      "grad_norm": 18.662717819213867,
      "learning_rate": 2.2465753424657534e-05,
      "loss": 1.1092,
      "step": 5500
    },
    {
      "epoch": 0.5068493150684932,
      "grad_norm": 13.65588092803955,
      "learning_rate": 2.2397260273972604e-05,
      "loss": 1.2777,
      "step": 5550
    },
    {
      "epoch": 0.5114155251141552,
      "grad_norm": 22.897354125976562,
      "learning_rate": 2.2328767123287673e-05,
      "loss": 1.0557,
      "step": 5600
    },
    {
      "epoch": 0.5159817351598174,
      "grad_norm": 22.30970573425293,
      "learning_rate": 2.2260273972602743e-05,
      "loss": 1.0701,
      "step": 5650
    },
    {
      "epoch": 0.5205479452054794,
      "grad_norm": 55.639892578125,
      "learning_rate": 2.219178082191781e-05,
      "loss": 1.1001,
      "step": 5700
    },
    {
      "epoch": 0.5251141552511416,
      "grad_norm": 11.327408790588379,
      "learning_rate": 2.212328767123288e-05,
      "loss": 1.1506,
      "step": 5750
    },
    {
      "epoch": 0.5296803652968036,
      "grad_norm": 31.288894653320312,
      "learning_rate": 2.2054794520547945e-05,
      "loss": 1.0774,
      "step": 5800
    },
    {
      "epoch": 0.5342465753424658,
      "grad_norm": 11.587857246398926,
      "learning_rate": 2.1986301369863014e-05,
      "loss": 1.0175,
      "step": 5850
    },
    {
      "epoch": 0.5388127853881278,
      "grad_norm": 27.174907684326172,
      "learning_rate": 2.191780821917808e-05,
      "loss": 1.1566,
      "step": 5900
    },
    {
      "epoch": 0.54337899543379,
      "grad_norm": 28.325786590576172,
      "learning_rate": 2.184931506849315e-05,
      "loss": 1.1089,
      "step": 5950
    },
    {
      "epoch": 0.547945205479452,
      "grad_norm": 21.70798683166504,
      "learning_rate": 2.178082191780822e-05,
      "loss": 1.0575,
      "step": 6000
    },
    {
      "epoch": 0.5525114155251142,
      "grad_norm": 18.79369354248047,
      "learning_rate": 2.171232876712329e-05,
      "loss": 1.1327,
      "step": 6050
    },
    {
      "epoch": 0.5570776255707762,
      "grad_norm": 31.291170120239258,
      "learning_rate": 2.1643835616438355e-05,
      "loss": 1.1412,
      "step": 6100
    },
    {
      "epoch": 0.5616438356164384,
      "grad_norm": 15.10459041595459,
      "learning_rate": 2.1575342465753424e-05,
      "loss": 1.1289,
      "step": 6150
    },
    {
      "epoch": 0.5662100456621004,
      "grad_norm": 23.595136642456055,
      "learning_rate": 2.1506849315068494e-05,
      "loss": 0.9419,
      "step": 6200
    },
    {
      "epoch": 0.5707762557077626,
      "grad_norm": 9.336952209472656,
      "learning_rate": 2.1438356164383563e-05,
      "loss": 0.9357,
      "step": 6250
    },
    {
      "epoch": 0.5753424657534246,
      "grad_norm": 34.959354400634766,
      "learning_rate": 2.136986301369863e-05,
      "loss": 1.1525,
      "step": 6300
    },
    {
      "epoch": 0.5799086757990868,
      "grad_norm": 25.974088668823242,
      "learning_rate": 2.13013698630137e-05,
      "loss": 1.0915,
      "step": 6350
    },
    {
      "epoch": 0.5844748858447488,
      "grad_norm": 26.807876586914062,
      "learning_rate": 2.1232876712328768e-05,
      "loss": 1.0783,
      "step": 6400
    },
    {
      "epoch": 0.589041095890411,
      "grad_norm": 11.836869239807129,
      "learning_rate": 2.1164383561643838e-05,
      "loss": 1.1106,
      "step": 6450
    },
    {
      "epoch": 0.593607305936073,
      "grad_norm": 20.5205078125,
      "learning_rate": 2.1095890410958904e-05,
      "loss": 1.0021,
      "step": 6500
    },
    {
      "epoch": 0.5981735159817352,
      "grad_norm": 63.918914794921875,
      "learning_rate": 2.1027397260273973e-05,
      "loss": 0.9942,
      "step": 6550
    },
    {
      "epoch": 0.6027397260273972,
      "grad_norm": 15.824874877929688,
      "learning_rate": 2.0958904109589043e-05,
      "loss": 1.0382,
      "step": 6600
    },
    {
      "epoch": 0.6073059360730594,
      "grad_norm": 23.539623260498047,
      "learning_rate": 2.0890410958904112e-05,
      "loss": 1.1323,
      "step": 6650
    },
    {
      "epoch": 0.6118721461187214,
      "grad_norm": 22.72905921936035,
      "learning_rate": 2.0821917808219178e-05,
      "loss": 1.096,
      "step": 6700
    },
    {
      "epoch": 0.6164383561643836,
      "grad_norm": 21.393850326538086,
      "learning_rate": 2.0753424657534248e-05,
      "loss": 1.0491,
      "step": 6750
    },
    {
      "epoch": 0.6210045662100456,
      "grad_norm": 10.425342559814453,
      "learning_rate": 2.0684931506849317e-05,
      "loss": 0.9894,
      "step": 6800
    },
    {
      "epoch": 0.6255707762557078,
      "grad_norm": 31.147842407226562,
      "learning_rate": 2.0616438356164387e-05,
      "loss": 1.0229,
      "step": 6850
    },
    {
      "epoch": 0.6301369863013698,
      "grad_norm": 13.265896797180176,
      "learning_rate": 2.0547945205479453e-05,
      "loss": 1.0372,
      "step": 6900
    },
    {
      "epoch": 0.634703196347032,
      "grad_norm": 19.299884796142578,
      "learning_rate": 2.0479452054794522e-05,
      "loss": 0.9861,
      "step": 6950
    },
    {
      "epoch": 0.639269406392694,
      "grad_norm": 15.215560913085938,
      "learning_rate": 2.041095890410959e-05,
      "loss": 1.0685,
      "step": 7000
    },
    {
      "epoch": 0.6438356164383562,
      "grad_norm": 12.113781929016113,
      "learning_rate": 2.034246575342466e-05,
      "loss": 1.0452,
      "step": 7050
    },
    {
      "epoch": 0.6484018264840182,
      "grad_norm": 53.0004768371582,
      "learning_rate": 2.0273972602739724e-05,
      "loss": 1.0177,
      "step": 7100
    },
    {
      "epoch": 0.6529680365296804,
      "grad_norm": 10.609742164611816,
      "learning_rate": 2.0205479452054793e-05,
      "loss": 0.9896,
      "step": 7150
    },
    {
      "epoch": 0.6575342465753424,
      "grad_norm": 13.45383071899414,
      "learning_rate": 2.0136986301369863e-05,
      "loss": 0.7549,
      "step": 7200
    },
    {
      "epoch": 0.6621004566210046,
      "grad_norm": 21.812686920166016,
      "learning_rate": 2.0068493150684932e-05,
      "loss": 1.0402,
      "step": 7250
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 7.622537612915039,
      "learning_rate": 1.9999999999999998e-05,
      "loss": 0.9468,
      "step": 7300
    },
    {
      "epoch": 0.6712328767123288,
      "grad_norm": 14.563289642333984,
      "learning_rate": 1.9931506849315068e-05,
      "loss": 1.0154,
      "step": 7350
    },
    {
      "epoch": 0.6757990867579908,
      "grad_norm": 29.9089298248291,
      "learning_rate": 1.9863013698630137e-05,
      "loss": 0.9872,
      "step": 7400
    },
    {
      "epoch": 0.680365296803653,
      "grad_norm": 7.758576393127441,
      "learning_rate": 1.9794520547945207e-05,
      "loss": 0.8618,
      "step": 7450
    },
    {
      "epoch": 0.684931506849315,
      "grad_norm": 15.849843978881836,
      "learning_rate": 1.9726027397260273e-05,
      "loss": 0.9981,
      "step": 7500
    },
    {
      "epoch": 0.6894977168949772,
      "grad_norm": 13.798691749572754,
      "learning_rate": 1.9657534246575342e-05,
      "loss": 1.0578,
      "step": 7550
    },
    {
      "epoch": 0.6940639269406392,
      "grad_norm": 12.977076530456543,
      "learning_rate": 1.9589041095890412e-05,
      "loss": 1.0081,
      "step": 7600
    },
    {
      "epoch": 0.6986301369863014,
      "grad_norm": 25.823368072509766,
      "learning_rate": 1.952054794520548e-05,
      "loss": 0.9483,
      "step": 7650
    },
    {
      "epoch": 0.7031963470319634,
      "grad_norm": 1.7260483503341675,
      "learning_rate": 1.9452054794520547e-05,
      "loss": 0.9396,
      "step": 7700
    },
    {
      "epoch": 0.7077625570776256,
      "grad_norm": 28.217559814453125,
      "learning_rate": 1.9383561643835617e-05,
      "loss": 1.0159,
      "step": 7750
    },
    {
      "epoch": 0.7123287671232876,
      "grad_norm": 15.566514015197754,
      "learning_rate": 1.9315068493150686e-05,
      "loss": 1.0673,
      "step": 7800
    },
    {
      "epoch": 0.7168949771689498,
      "grad_norm": 49.06792449951172,
      "learning_rate": 1.9246575342465756e-05,
      "loss": 0.9817,
      "step": 7850
    },
    {
      "epoch": 0.7214611872146118,
      "grad_norm": 35.709224700927734,
      "learning_rate": 1.9178082191780822e-05,
      "loss": 0.9818,
      "step": 7900
    },
    {
      "epoch": 0.726027397260274,
      "grad_norm": 20.94626808166504,
      "learning_rate": 1.910958904109589e-05,
      "loss": 0.9113,
      "step": 7950
    },
    {
      "epoch": 0.730593607305936,
      "grad_norm": 25.5698184967041,
      "learning_rate": 1.904109589041096e-05,
      "loss": 1.0525,
      "step": 8000
    },
    {
      "epoch": 0.7351598173515982,
      "grad_norm": 36.41669845581055,
      "learning_rate": 1.897260273972603e-05,
      "loss": 1.007,
      "step": 8050
    },
    {
      "epoch": 0.7397260273972602,
      "grad_norm": 17.15513038635254,
      "learning_rate": 1.8904109589041096e-05,
      "loss": 0.9568,
      "step": 8100
    },
    {
      "epoch": 0.7442922374429224,
      "grad_norm": 24.435543060302734,
      "learning_rate": 1.8835616438356166e-05,
      "loss": 1.0073,
      "step": 8150
    },
    {
      "epoch": 0.7488584474885844,
      "grad_norm": 26.5279483795166,
      "learning_rate": 1.8767123287671235e-05,
      "loss": 0.9768,
      "step": 8200
    },
    {
      "epoch": 0.7534246575342466,
      "grad_norm": 18.786693572998047,
      "learning_rate": 1.8698630136986305e-05,
      "loss": 1.0579,
      "step": 8250
    },
    {
      "epoch": 0.7579908675799086,
      "grad_norm": 22.201034545898438,
      "learning_rate": 1.863013698630137e-05,
      "loss": 1.003,
      "step": 8300
    },
    {
      "epoch": 0.7625570776255708,
      "grad_norm": 28.934587478637695,
      "learning_rate": 1.856164383561644e-05,
      "loss": 0.9753,
      "step": 8350
    },
    {
      "epoch": 0.7671232876712328,
      "grad_norm": 17.935937881469727,
      "learning_rate": 1.8493150684931506e-05,
      "loss": 0.9038,
      "step": 8400
    },
    {
      "epoch": 0.771689497716895,
      "grad_norm": 29.648881912231445,
      "learning_rate": 1.8424657534246576e-05,
      "loss": 1.0109,
      "step": 8450
    },
    {
      "epoch": 0.776255707762557,
      "grad_norm": 32.65835189819336,
      "learning_rate": 1.8356164383561642e-05,
      "loss": 1.0207,
      "step": 8500
    },
    {
      "epoch": 0.7808219178082192,
      "grad_norm": 21.674053192138672,
      "learning_rate": 1.828767123287671e-05,
      "loss": 1.0573,
      "step": 8550
    },
    {
      "epoch": 0.7853881278538812,
      "grad_norm": 19.038619995117188,
      "learning_rate": 1.821917808219178e-05,
      "loss": 1.042,
      "step": 8600
    },
    {
      "epoch": 0.7899543378995434,
      "grad_norm": 23.994884490966797,
      "learning_rate": 1.815068493150685e-05,
      "loss": 1.0316,
      "step": 8650
    },
    {
      "epoch": 0.7945205479452054,
      "grad_norm": 26.702028274536133,
      "learning_rate": 1.8082191780821916e-05,
      "loss": 0.9592,
      "step": 8700
    },
    {
      "epoch": 0.7990867579908676,
      "grad_norm": 7.485787391662598,
      "learning_rate": 1.8013698630136986e-05,
      "loss": 1.0175,
      "step": 8750
    },
    {
      "epoch": 0.8036529680365296,
      "grad_norm": 24.249893188476562,
      "learning_rate": 1.7945205479452055e-05,
      "loss": 1.0643,
      "step": 8800
    },
    {
      "epoch": 0.8082191780821918,
      "grad_norm": 31.136962890625,
      "learning_rate": 1.7876712328767125e-05,
      "loss": 0.8719,
      "step": 8850
    },
    {
      "epoch": 0.8127853881278538,
      "grad_norm": 18.02696418762207,
      "learning_rate": 1.780821917808219e-05,
      "loss": 0.9169,
      "step": 8900
    },
    {
      "epoch": 0.817351598173516,
      "grad_norm": 7.9297261238098145,
      "learning_rate": 1.773972602739726e-05,
      "loss": 0.9241,
      "step": 8950
    },
    {
      "epoch": 0.821917808219178,
      "grad_norm": 15.147187232971191,
      "learning_rate": 1.767123287671233e-05,
      "loss": 0.9451,
      "step": 9000
    },
    {
      "epoch": 0.8264840182648402,
      "grad_norm": 26.369773864746094,
      "learning_rate": 1.76027397260274e-05,
      "loss": 0.9198,
      "step": 9050
    },
    {
      "epoch": 0.8310502283105022,
      "grad_norm": 10.24289321899414,
      "learning_rate": 1.7534246575342465e-05,
      "loss": 0.9423,
      "step": 9100
    },
    {
      "epoch": 0.8356164383561644,
      "grad_norm": 17.285388946533203,
      "learning_rate": 1.7465753424657535e-05,
      "loss": 0.8636,
      "step": 9150
    },
    {
      "epoch": 0.8401826484018264,
      "grad_norm": 26.444520950317383,
      "learning_rate": 1.7397260273972604e-05,
      "loss": 1.0647,
      "step": 9200
    },
    {
      "epoch": 0.8447488584474886,
      "grad_norm": 37.353084564208984,
      "learning_rate": 1.7328767123287674e-05,
      "loss": 0.9484,
      "step": 9250
    },
    {
      "epoch": 0.8493150684931506,
      "grad_norm": 9.881476402282715,
      "learning_rate": 1.726027397260274e-05,
      "loss": 1.014,
      "step": 9300
    },
    {
      "epoch": 0.8538812785388128,
      "grad_norm": 27.257709503173828,
      "learning_rate": 1.719178082191781e-05,
      "loss": 0.9169,
      "step": 9350
    },
    {
      "epoch": 0.8584474885844748,
      "grad_norm": 20.704347610473633,
      "learning_rate": 1.712328767123288e-05,
      "loss": 0.9597,
      "step": 9400
    },
    {
      "epoch": 0.863013698630137,
      "grad_norm": 20.994293212890625,
      "learning_rate": 1.705479452054795e-05,
      "loss": 0.9328,
      "step": 9450
    },
    {
      "epoch": 0.867579908675799,
      "grad_norm": 31.617778778076172,
      "learning_rate": 1.6986301369863014e-05,
      "loss": 0.9988,
      "step": 9500
    },
    {
      "epoch": 0.8721461187214612,
      "grad_norm": 24.76189422607422,
      "learning_rate": 1.6917808219178084e-05,
      "loss": 0.9338,
      "step": 9550
    },
    {
      "epoch": 0.8767123287671232,
      "grad_norm": 27.70781135559082,
      "learning_rate": 1.6849315068493153e-05,
      "loss": 0.8994,
      "step": 9600
    },
    {
      "epoch": 0.8812785388127854,
      "grad_norm": 6.332718372344971,
      "learning_rate": 1.6780821917808223e-05,
      "loss": 0.8815,
      "step": 9650
    },
    {
      "epoch": 0.8858447488584474,
      "grad_norm": 25.22089195251465,
      "learning_rate": 1.6712328767123286e-05,
      "loss": 1.0243,
      "step": 9700
    },
    {
      "epoch": 0.8904109589041096,
      "grad_norm": 6.610196590423584,
      "learning_rate": 1.6643835616438355e-05,
      "loss": 0.856,
      "step": 9750
    },
    {
      "epoch": 0.8949771689497716,
      "grad_norm": 9.7604341506958,
      "learning_rate": 1.6575342465753425e-05,
      "loss": 0.9275,
      "step": 9800
    },
    {
      "epoch": 0.8995433789954338,
      "grad_norm": 13.222633361816406,
      "learning_rate": 1.6506849315068494e-05,
      "loss": 0.7912,
      "step": 9850
    },
    {
      "epoch": 0.9041095890410958,
      "grad_norm": 8.07390022277832,
      "learning_rate": 1.643835616438356e-05,
      "loss": 0.9296,
      "step": 9900
    },
    {
      "epoch": 0.908675799086758,
      "grad_norm": 8.25124740600586,
      "learning_rate": 1.636986301369863e-05,
      "loss": 0.9269,
      "step": 9950
    },
    {
      "epoch": 0.91324200913242,
      "grad_norm": 22.882007598876953,
      "learning_rate": 1.63013698630137e-05,
      "loss": 0.9694,
      "step": 10000
    },
    {
      "epoch": 0.9178082191780822,
      "grad_norm": 8.665299415588379,
      "learning_rate": 1.623287671232877e-05,
      "loss": 0.8643,
      "step": 10050
    },
    {
      "epoch": 0.9223744292237442,
      "grad_norm": 13.816045761108398,
      "learning_rate": 1.6164383561643835e-05,
      "loss": 0.9476,
      "step": 10100
    },
    {
      "epoch": 0.9269406392694064,
      "grad_norm": 11.227286338806152,
      "learning_rate": 1.6095890410958904e-05,
      "loss": 0.8805,
      "step": 10150
    },
    {
      "epoch": 0.9315068493150684,
      "grad_norm": 7.756823539733887,
      "learning_rate": 1.6027397260273974e-05,
      "loss": 0.9185,
      "step": 10200
    },
    {
      "epoch": 0.9360730593607306,
      "grad_norm": 3.170640230178833,
      "learning_rate": 1.5958904109589043e-05,
      "loss": 0.8761,
      "step": 10250
    },
    {
      "epoch": 0.9406392694063926,
      "grad_norm": 11.567336082458496,
      "learning_rate": 1.589041095890411e-05,
      "loss": 0.9298,
      "step": 10300
    },
    {
      "epoch": 0.9452054794520548,
      "grad_norm": 23.299612045288086,
      "learning_rate": 1.582191780821918e-05,
      "loss": 0.8361,
      "step": 10350
    },
    {
      "epoch": 0.9497716894977168,
      "grad_norm": 14.57529067993164,
      "learning_rate": 1.5753424657534248e-05,
      "loss": 0.9669,
      "step": 10400
    },
    {
      "epoch": 0.954337899543379,
      "grad_norm": 10.777979850769043,
      "learning_rate": 1.5684931506849318e-05,
      "loss": 0.9632,
      "step": 10450
    },
    {
      "epoch": 0.958904109589041,
      "grad_norm": 25.46247100830078,
      "learning_rate": 1.5616438356164384e-05,
      "loss": 0.8625,
      "step": 10500
    },
    {
      "epoch": 0.9634703196347032,
      "grad_norm": 8.925729751586914,
      "learning_rate": 1.5547945205479453e-05,
      "loss": 0.903,
      "step": 10550
    },
    {
      "epoch": 0.9680365296803652,
      "grad_norm": 11.709217071533203,
      "learning_rate": 1.5479452054794523e-05,
      "loss": 0.9446,
      "step": 10600
    },
    {
      "epoch": 0.9726027397260274,
      "grad_norm": 36.75910568237305,
      "learning_rate": 1.5410958904109592e-05,
      "loss": 0.9568,
      "step": 10650
    },
    {
      "epoch": 0.9771689497716894,
      "grad_norm": 30.059825897216797,
      "learning_rate": 1.5342465753424658e-05,
      "loss": 0.9603,
      "step": 10700
    },
    {
      "epoch": 0.9817351598173516,
      "grad_norm": 23.478477478027344,
      "learning_rate": 1.5273972602739728e-05,
      "loss": 0.9828,
      "step": 10750
    },
    {
      "epoch": 0.9863013698630136,
      "grad_norm": 34.233699798583984,
      "learning_rate": 1.5205479452054795e-05,
      "loss": 0.9867,
      "step": 10800
    },
    {
      "epoch": 0.9908675799086758,
      "grad_norm": 19.399288177490234,
      "learning_rate": 1.5136986301369865e-05,
      "loss": 0.8618,
      "step": 10850
    },
    {
      "epoch": 0.9954337899543378,
      "grad_norm": 9.319437026977539,
      "learning_rate": 1.5068493150684931e-05,
      "loss": 0.9352,
      "step": 10900
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.767410278320312,
      "learning_rate": 1.5e-05,
      "loss": 0.8511,
      "step": 10950
    },
    {
      "epoch": 1.004566210045662,
      "grad_norm": 15.091998100280762,
      "learning_rate": 1.4931506849315068e-05,
      "loss": 0.6653,
      "step": 11000
    },
    {
      "epoch": 1.009132420091324,
      "grad_norm": 23.510337829589844,
      "learning_rate": 1.4863013698630138e-05,
      "loss": 0.6163,
      "step": 11050
    },
    {
      "epoch": 1.0136986301369864,
      "grad_norm": 8.980317115783691,
      "learning_rate": 1.4794520547945205e-05,
      "loss": 0.7783,
      "step": 11100
    },
    {
      "epoch": 1.0182648401826484,
      "grad_norm": 36.67283248901367,
      "learning_rate": 1.4726027397260275e-05,
      "loss": 0.7681,
      "step": 11150
    },
    {
      "epoch": 1.0228310502283104,
      "grad_norm": 4.734091758728027,
      "learning_rate": 1.4657534246575343e-05,
      "loss": 0.6181,
      "step": 11200
    },
    {
      "epoch": 1.0273972602739727,
      "grad_norm": 7.513982772827148,
      "learning_rate": 1.4589041095890412e-05,
      "loss": 0.7297,
      "step": 11250
    },
    {
      "epoch": 1.0319634703196348,
      "grad_norm": 16.1832275390625,
      "learning_rate": 1.4520547945205478e-05,
      "loss": 0.6487,
      "step": 11300
    },
    {
      "epoch": 1.0365296803652968,
      "grad_norm": 25.557220458984375,
      "learning_rate": 1.4452054794520548e-05,
      "loss": 0.7089,
      "step": 11350
    },
    {
      "epoch": 1.0410958904109588,
      "grad_norm": 11.07717227935791,
      "learning_rate": 1.4383561643835615e-05,
      "loss": 0.6266,
      "step": 11400
    },
    {
      "epoch": 1.045662100456621,
      "grad_norm": 11.381722450256348,
      "learning_rate": 1.4315068493150685e-05,
      "loss": 0.6758,
      "step": 11450
    },
    {
      "epoch": 1.0502283105022832,
      "grad_norm": 9.080512046813965,
      "learning_rate": 1.4246575342465753e-05,
      "loss": 0.6865,
      "step": 11500
    },
    {
      "epoch": 1.0547945205479452,
      "grad_norm": 10.727092742919922,
      "learning_rate": 1.4178082191780822e-05,
      "loss": 0.6494,
      "step": 11550
    },
    {
      "epoch": 1.0593607305936072,
      "grad_norm": 4.969435691833496,
      "learning_rate": 1.410958904109589e-05,
      "loss": 0.5874,
      "step": 11600
    },
    {
      "epoch": 1.0639269406392695,
      "grad_norm": 12.03378963470459,
      "learning_rate": 1.404109589041096e-05,
      "loss": 0.6301,
      "step": 11650
    },
    {
      "epoch": 1.0684931506849316,
      "grad_norm": 8.797894477844238,
      "learning_rate": 1.3972602739726027e-05,
      "loss": 0.6839,
      "step": 11700
    },
    {
      "epoch": 1.0730593607305936,
      "grad_norm": 9.395676612854004,
      "learning_rate": 1.3904109589041097e-05,
      "loss": 0.6672,
      "step": 11750
    },
    {
      "epoch": 1.0776255707762556,
      "grad_norm": 26.268247604370117,
      "learning_rate": 1.3835616438356164e-05,
      "loss": 0.6398,
      "step": 11800
    },
    {
      "epoch": 1.0821917808219177,
      "grad_norm": 19.63582992553711,
      "learning_rate": 1.3767123287671234e-05,
      "loss": 0.7825,
      "step": 11850
    },
    {
      "epoch": 1.08675799086758,
      "grad_norm": 23.42495346069336,
      "learning_rate": 1.3698630136986302e-05,
      "loss": 0.7078,
      "step": 11900
    },
    {
      "epoch": 1.091324200913242,
      "grad_norm": 29.21826934814453,
      "learning_rate": 1.363013698630137e-05,
      "loss": 0.6973,
      "step": 11950
    },
    {
      "epoch": 1.095890410958904,
      "grad_norm": 13.69019603729248,
      "learning_rate": 1.3561643835616437e-05,
      "loss": 0.6325,
      "step": 12000
    },
    {
      "epoch": 1.1004566210045663,
      "grad_norm": 22.293821334838867,
      "learning_rate": 1.3493150684931507e-05,
      "loss": 0.6833,
      "step": 12050
    },
    {
      "epoch": 1.1050228310502284,
      "grad_norm": 9.701814651489258,
      "learning_rate": 1.3424657534246575e-05,
      "loss": 0.6518,
      "step": 12100
    },
    {
      "epoch": 1.1095890410958904,
      "grad_norm": 25.30548667907715,
      "learning_rate": 1.3356164383561644e-05,
      "loss": 0.7458,
      "step": 12150
    },
    {
      "epoch": 1.1141552511415524,
      "grad_norm": 11.800670623779297,
      "learning_rate": 1.3287671232876712e-05,
      "loss": 0.755,
      "step": 12200
    },
    {
      "epoch": 1.1187214611872145,
      "grad_norm": 15.019549369812012,
      "learning_rate": 1.3219178082191781e-05,
      "loss": 0.6871,
      "step": 12250
    },
    {
      "epoch": 1.1232876712328768,
      "grad_norm": 8.90066909790039,
      "learning_rate": 1.3150684931506849e-05,
      "loss": 0.738,
      "step": 12300
    },
    {
      "epoch": 1.1278538812785388,
      "grad_norm": 6.738426685333252,
      "learning_rate": 1.3082191780821919e-05,
      "loss": 0.6828,
      "step": 12350
    },
    {
      "epoch": 1.1324200913242009,
      "grad_norm": 7.866949558258057,
      "learning_rate": 1.3013698630136986e-05,
      "loss": 0.708,
      "step": 12400
    },
    {
      "epoch": 1.1369863013698631,
      "grad_norm": 26.066892623901367,
      "learning_rate": 1.2945205479452056e-05,
      "loss": 0.7078,
      "step": 12450
    },
    {
      "epoch": 1.1415525114155252,
      "grad_norm": 7.540081024169922,
      "learning_rate": 1.2876712328767124e-05,
      "loss": 0.6958,
      "step": 12500
    },
    {
      "epoch": 1.1461187214611872,
      "grad_norm": 13.667712211608887,
      "learning_rate": 1.2808219178082193e-05,
      "loss": 0.5233,
      "step": 12550
    },
    {
      "epoch": 1.1506849315068493,
      "grad_norm": 17.8378849029541,
      "learning_rate": 1.2739726027397259e-05,
      "loss": 0.669,
      "step": 12600
    },
    {
      "epoch": 1.1552511415525113,
      "grad_norm": 25.038570404052734,
      "learning_rate": 1.2671232876712329e-05,
      "loss": 0.5989,
      "step": 12650
    },
    {
      "epoch": 1.1598173515981736,
      "grad_norm": 2.4400246143341064,
      "learning_rate": 1.2602739726027396e-05,
      "loss": 0.745,
      "step": 12700
    },
    {
      "epoch": 1.1643835616438356,
      "grad_norm": 3.209836006164551,
      "learning_rate": 1.2534246575342466e-05,
      "loss": 0.6943,
      "step": 12750
    },
    {
      "epoch": 1.1689497716894977,
      "grad_norm": 12.491949081420898,
      "learning_rate": 1.2465753424657534e-05,
      "loss": 0.6115,
      "step": 12800
    },
    {
      "epoch": 1.17351598173516,
      "grad_norm": 11.220638275146484,
      "learning_rate": 1.2397260273972603e-05,
      "loss": 0.5898,
      "step": 12850
    },
    {
      "epoch": 1.178082191780822,
      "grad_norm": 16.73185920715332,
      "learning_rate": 1.2328767123287671e-05,
      "loss": 0.6734,
      "step": 12900
    },
    {
      "epoch": 1.182648401826484,
      "grad_norm": 14.307918548583984,
      "learning_rate": 1.226027397260274e-05,
      "loss": 0.7266,
      "step": 12950
    },
    {
      "epoch": 1.187214611872146,
      "grad_norm": 14.836384773254395,
      "learning_rate": 1.2191780821917808e-05,
      "loss": 0.7586,
      "step": 13000
    },
    {
      "epoch": 1.191780821917808,
      "grad_norm": 29.282228469848633,
      "learning_rate": 1.2123287671232878e-05,
      "loss": 0.7917,
      "step": 13050
    },
    {
      "epoch": 1.1963470319634704,
      "grad_norm": 26.005468368530273,
      "learning_rate": 1.2054794520547945e-05,
      "loss": 0.6413,
      "step": 13100
    },
    {
      "epoch": 1.2009132420091324,
      "grad_norm": 8.961703300476074,
      "learning_rate": 1.1986301369863015e-05,
      "loss": 0.792,
      "step": 13150
    },
    {
      "epoch": 1.2054794520547945,
      "grad_norm": 10.702567100524902,
      "learning_rate": 1.1917808219178083e-05,
      "loss": 0.6594,
      "step": 13200
    },
    {
      "epoch": 1.2100456621004567,
      "grad_norm": 13.418671607971191,
      "learning_rate": 1.184931506849315e-05,
      "loss": 0.7547,
      "step": 13250
    },
    {
      "epoch": 1.2146118721461188,
      "grad_norm": 17.413429260253906,
      "learning_rate": 1.1780821917808218e-05,
      "loss": 0.7908,
      "step": 13300
    },
    {
      "epoch": 1.2191780821917808,
      "grad_norm": 18.375572204589844,
      "learning_rate": 1.1712328767123288e-05,
      "loss": 0.7298,
      "step": 13350
    },
    {
      "epoch": 1.2237442922374429,
      "grad_norm": 12.126824378967285,
      "learning_rate": 1.1643835616438355e-05,
      "loss": 0.7482,
      "step": 13400
    },
    {
      "epoch": 1.228310502283105,
      "grad_norm": 6.292917728424072,
      "learning_rate": 1.1575342465753425e-05,
      "loss": 0.7049,
      "step": 13450
    },
    {
      "epoch": 1.2328767123287672,
      "grad_norm": 48.449954986572266,
      "learning_rate": 1.1506849315068493e-05,
      "loss": 0.5808,
      "step": 13500
    },
    {
      "epoch": 1.2374429223744292,
      "grad_norm": 11.724007606506348,
      "learning_rate": 1.1438356164383562e-05,
      "loss": 0.6694,
      "step": 13550
    },
    {
      "epoch": 1.2420091324200913,
      "grad_norm": 11.694669723510742,
      "learning_rate": 1.136986301369863e-05,
      "loss": 0.7223,
      "step": 13600
    },
    {
      "epoch": 1.2465753424657535,
      "grad_norm": 18.402746200561523,
      "learning_rate": 1.13013698630137e-05,
      "loss": 0.6794,
      "step": 13650
    },
    {
      "epoch": 1.2511415525114156,
      "grad_norm": 10.988481521606445,
      "learning_rate": 1.1232876712328767e-05,
      "loss": 0.618,
      "step": 13700
    },
    {
      "epoch": 1.2557077625570776,
      "grad_norm": 8.175172805786133,
      "learning_rate": 1.1164383561643837e-05,
      "loss": 0.6552,
      "step": 13750
    },
    {
      "epoch": 1.2602739726027397,
      "grad_norm": 17.85883903503418,
      "learning_rate": 1.1095890410958904e-05,
      "loss": 0.6645,
      "step": 13800
    },
    {
      "epoch": 1.2648401826484017,
      "grad_norm": 6.0199198722839355,
      "learning_rate": 1.1027397260273972e-05,
      "loss": 0.7725,
      "step": 13850
    },
    {
      "epoch": 1.269406392694064,
      "grad_norm": 12.200461387634277,
      "learning_rate": 1.095890410958904e-05,
      "loss": 0.6685,
      "step": 13900
    },
    {
      "epoch": 1.273972602739726,
      "grad_norm": 11.29808235168457,
      "learning_rate": 1.089041095890411e-05,
      "loss": 0.5974,
      "step": 13950
    },
    {
      "epoch": 1.278538812785388,
      "grad_norm": 20.522750854492188,
      "learning_rate": 1.0821917808219177e-05,
      "loss": 0.6497,
      "step": 14000
    },
    {
      "epoch": 1.2831050228310503,
      "grad_norm": 4.903714656829834,
      "learning_rate": 1.0753424657534247e-05,
      "loss": 0.805,
      "step": 14050
    },
    {
      "epoch": 1.2876712328767124,
      "grad_norm": 14.09408950805664,
      "learning_rate": 1.0684931506849315e-05,
      "loss": 0.6975,
      "step": 14100
    },
    {
      "epoch": 1.2922374429223744,
      "grad_norm": 5.306293964385986,
      "learning_rate": 1.0616438356164384e-05,
      "loss": 0.6181,
      "step": 14150
    },
    {
      "epoch": 1.2968036529680365,
      "grad_norm": 15.217981338500977,
      "learning_rate": 1.0547945205479452e-05,
      "loss": 0.6363,
      "step": 14200
    },
    {
      "epoch": 1.3013698630136985,
      "grad_norm": 2.86833119392395,
      "learning_rate": 1.0479452054794521e-05,
      "loss": 0.632,
      "step": 14250
    },
    {
      "epoch": 1.3059360730593608,
      "grad_norm": 12.824779510498047,
      "learning_rate": 1.0410958904109589e-05,
      "loss": 0.7499,
      "step": 14300
    },
    {
      "epoch": 1.3105022831050228,
      "grad_norm": 17.47176170349121,
      "learning_rate": 1.0342465753424659e-05,
      "loss": 0.581,
      "step": 14350
    },
    {
      "epoch": 1.3150684931506849,
      "grad_norm": 7.620934009552002,
      "learning_rate": 1.0273972602739726e-05,
      "loss": 0.5913,
      "step": 14400
    },
    {
      "epoch": 1.3196347031963471,
      "grad_norm": 11.1875,
      "learning_rate": 1.0205479452054796e-05,
      "loss": 0.6465,
      "step": 14450
    },
    {
      "epoch": 1.3242009132420092,
      "grad_norm": 24.72510528564453,
      "learning_rate": 1.0136986301369862e-05,
      "loss": 0.6796,
      "step": 14500
    },
    {
      "epoch": 1.3287671232876712,
      "grad_norm": 6.102226734161377,
      "learning_rate": 1.0068493150684931e-05,
      "loss": 0.7074,
      "step": 14550
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 6.332953929901123,
      "learning_rate": 9.999999999999999e-06,
      "loss": 0.6338,
      "step": 14600
    },
    {
      "epoch": 1.3378995433789953,
      "grad_norm": 28.30670738220215,
      "learning_rate": 9.931506849315069e-06,
      "loss": 0.7109,
      "step": 14650
    },
    {
      "epoch": 1.3424657534246576,
      "grad_norm": 40.727230072021484,
      "learning_rate": 9.863013698630136e-06,
      "loss": 0.7342,
      "step": 14700
    },
    {
      "epoch": 1.3470319634703196,
      "grad_norm": 11.026389122009277,
      "learning_rate": 9.794520547945206e-06,
      "loss": 0.6972,
      "step": 14750
    },
    {
      "epoch": 1.3515981735159817,
      "grad_norm": 16.95206642150879,
      "learning_rate": 9.726027397260274e-06,
      "loss": 0.6509,
      "step": 14800
    },
    {
      "epoch": 1.356164383561644,
      "grad_norm": 24.887845993041992,
      "learning_rate": 9.657534246575343e-06,
      "loss": 0.6608,
      "step": 14850
    },
    {
      "epoch": 1.360730593607306,
      "grad_norm": 5.1824421882629395,
      "learning_rate": 9.589041095890411e-06,
      "loss": 0.664,
      "step": 14900
    },
    {
      "epoch": 1.365296803652968,
      "grad_norm": 35.08380889892578,
      "learning_rate": 9.52054794520548e-06,
      "loss": 0.7375,
      "step": 14950
    },
    {
      "epoch": 1.36986301369863,
      "grad_norm": 13.273919105529785,
      "learning_rate": 9.452054794520548e-06,
      "loss": 0.6581,
      "step": 15000
    },
    {
      "epoch": 1.374429223744292,
      "grad_norm": 20.243751525878906,
      "learning_rate": 9.383561643835618e-06,
      "loss": 0.7028,
      "step": 15050
    },
    {
      "epoch": 1.3789954337899544,
      "grad_norm": 6.9884934425354,
      "learning_rate": 9.315068493150685e-06,
      "loss": 0.6699,
      "step": 15100
    },
    {
      "epoch": 1.3835616438356164,
      "grad_norm": 10.110499382019043,
      "learning_rate": 9.246575342465753e-06,
      "loss": 0.5804,
      "step": 15150
    },
    {
      "epoch": 1.3881278538812785,
      "grad_norm": 5.272585868835449,
      "learning_rate": 9.178082191780821e-06,
      "loss": 0.7046,
      "step": 15200
    },
    {
      "epoch": 1.3926940639269407,
      "grad_norm": 3.3239293098449707,
      "learning_rate": 9.10958904109589e-06,
      "loss": 0.6621,
      "step": 15250
    },
    {
      "epoch": 1.3972602739726028,
      "grad_norm": 9.12402057647705,
      "learning_rate": 9.041095890410958e-06,
      "loss": 0.5613,
      "step": 15300
    },
    {
      "epoch": 1.4018264840182648,
      "grad_norm": 9.260445594787598,
      "learning_rate": 8.972602739726028e-06,
      "loss": 0.6179,
      "step": 15350
    },
    {
      "epoch": 1.4063926940639269,
      "grad_norm": 40.01394271850586,
      "learning_rate": 8.904109589041095e-06,
      "loss": 0.5897,
      "step": 15400
    },
    {
      "epoch": 1.410958904109589,
      "grad_norm": 43.27082443237305,
      "learning_rate": 8.835616438356165e-06,
      "loss": 0.5804,
      "step": 15450
    },
    {
      "epoch": 1.4155251141552512,
      "grad_norm": 8.149184226989746,
      "learning_rate": 8.767123287671233e-06,
      "loss": 0.7865,
      "step": 15500
    },
    {
      "epoch": 1.4200913242009132,
      "grad_norm": 16.708332061767578,
      "learning_rate": 8.698630136986302e-06,
      "loss": 0.6983,
      "step": 15550
    },
    {
      "epoch": 1.4246575342465753,
      "grad_norm": 2.826059103012085,
      "learning_rate": 8.63013698630137e-06,
      "loss": 0.6238,
      "step": 15600
    },
    {
      "epoch": 1.4292237442922375,
      "grad_norm": 13.411745071411133,
      "learning_rate": 8.56164383561644e-06,
      "loss": 0.6448,
      "step": 15650
    },
    {
      "epoch": 1.4337899543378996,
      "grad_norm": 12.094820976257324,
      "learning_rate": 8.493150684931507e-06,
      "loss": 0.6868,
      "step": 15700
    },
    {
      "epoch": 1.4383561643835616,
      "grad_norm": 13.272956848144531,
      "learning_rate": 8.424657534246577e-06,
      "loss": 0.6758,
      "step": 15750
    },
    {
      "epoch": 1.4429223744292237,
      "grad_norm": 7.710869312286377,
      "learning_rate": 8.356164383561643e-06,
      "loss": 0.6154,
      "step": 15800
    },
    {
      "epoch": 1.4474885844748857,
      "grad_norm": 24.845901489257812,
      "learning_rate": 8.287671232876712e-06,
      "loss": 0.6228,
      "step": 15850
    },
    {
      "epoch": 1.452054794520548,
      "grad_norm": 10.101279258728027,
      "learning_rate": 8.21917808219178e-06,
      "loss": 0.7418,
      "step": 15900
    },
    {
      "epoch": 1.45662100456621,
      "grad_norm": 71.19440460205078,
      "learning_rate": 8.15068493150685e-06,
      "loss": 0.7128,
      "step": 15950
    },
    {
      "epoch": 1.461187214611872,
      "grad_norm": 6.2137041091918945,
      "learning_rate": 8.082191780821917e-06,
      "loss": 0.7121,
      "step": 16000
    },
    {
      "epoch": 1.4657534246575343,
      "grad_norm": 20.703536987304688,
      "learning_rate": 8.013698630136987e-06,
      "loss": 0.7115,
      "step": 16050
    },
    {
      "epoch": 1.4703196347031964,
      "grad_norm": 26.53441619873047,
      "learning_rate": 7.945205479452055e-06,
      "loss": 0.5534,
      "step": 16100
    },
    {
      "epoch": 1.4748858447488584,
      "grad_norm": 20.233125686645508,
      "learning_rate": 7.876712328767124e-06,
      "loss": 0.6204,
      "step": 16150
    },
    {
      "epoch": 1.4794520547945205,
      "grad_norm": 23.36627769470215,
      "learning_rate": 7.808219178082192e-06,
      "loss": 0.6914,
      "step": 16200
    },
    {
      "epoch": 1.4840182648401825,
      "grad_norm": 11.94163703918457,
      "learning_rate": 7.739726027397261e-06,
      "loss": 0.6062,
      "step": 16250
    },
    {
      "epoch": 1.4885844748858448,
      "grad_norm": 15.757901191711426,
      "learning_rate": 7.671232876712329e-06,
      "loss": 0.6142,
      "step": 16300
    },
    {
      "epoch": 1.4931506849315068,
      "grad_norm": 12.007556915283203,
      "learning_rate": 7.602739726027398e-06,
      "loss": 0.6708,
      "step": 16350
    },
    {
      "epoch": 1.4977168949771689,
      "grad_norm": 9.127739906311035,
      "learning_rate": 7.5342465753424655e-06,
      "loss": 0.6319,
      "step": 16400
    },
    {
      "epoch": 1.5022831050228311,
      "grad_norm": 14.81264877319336,
      "learning_rate": 7.465753424657534e-06,
      "loss": 0.6143,
      "step": 16450
    },
    {
      "epoch": 1.5068493150684932,
      "grad_norm": 8.986160278320312,
      "learning_rate": 7.397260273972603e-06,
      "loss": 0.7508,
      "step": 16500
    },
    {
      "epoch": 1.5114155251141552,
      "grad_norm": 23.52107810974121,
      "learning_rate": 7.328767123287671e-06,
      "loss": 0.6377,
      "step": 16550
    },
    {
      "epoch": 1.5159817351598175,
      "grad_norm": 4.14312219619751,
      "learning_rate": 7.260273972602739e-06,
      "loss": 0.595,
      "step": 16600
    },
    {
      "epoch": 1.5205479452054793,
      "grad_norm": 32.40318298339844,
      "learning_rate": 7.191780821917808e-06,
      "loss": 0.6177,
      "step": 16650
    },
    {
      "epoch": 1.5251141552511416,
      "grad_norm": 13.583507537841797,
      "learning_rate": 7.123287671232876e-06,
      "loss": 0.6513,
      "step": 16700
    },
    {
      "epoch": 1.5296803652968036,
      "grad_norm": 27.02487564086914,
      "learning_rate": 7.054794520547945e-06,
      "loss": 0.6136,
      "step": 16750
    },
    {
      "epoch": 1.5342465753424657,
      "grad_norm": 46.82355880737305,
      "learning_rate": 6.986301369863014e-06,
      "loss": 0.6545,
      "step": 16800
    },
    {
      "epoch": 1.538812785388128,
      "grad_norm": 11.266030311584473,
      "learning_rate": 6.917808219178082e-06,
      "loss": 0.6438,
      "step": 16850
    },
    {
      "epoch": 1.54337899543379,
      "grad_norm": 21.652450561523438,
      "learning_rate": 6.849315068493151e-06,
      "loss": 0.6617,
      "step": 16900
    },
    {
      "epoch": 1.547945205479452,
      "grad_norm": 18.254072189331055,
      "learning_rate": 6.780821917808219e-06,
      "loss": 0.645,
      "step": 16950
    },
    {
      "epoch": 1.5525114155251143,
      "grad_norm": 7.959854602813721,
      "learning_rate": 6.712328767123287e-06,
      "loss": 0.7444,
      "step": 17000
    },
    {
      "epoch": 1.5570776255707761,
      "grad_norm": 15.169598579406738,
      "learning_rate": 6.643835616438356e-06,
      "loss": 0.6788,
      "step": 17050
    },
    {
      "epoch": 1.5616438356164384,
      "grad_norm": 14.872618675231934,
      "learning_rate": 6.5753424657534245e-06,
      "loss": 0.7433,
      "step": 17100
    },
    {
      "epoch": 1.5662100456621004,
      "grad_norm": 12.479876518249512,
      "learning_rate": 6.506849315068493e-06,
      "loss": 0.676,
      "step": 17150
    },
    {
      "epoch": 1.5707762557077625,
      "grad_norm": 4.30610990524292,
      "learning_rate": 6.438356164383562e-06,
      "loss": 0.597,
      "step": 17200
    },
    {
      "epoch": 1.5753424657534247,
      "grad_norm": 9.52687931060791,
      "learning_rate": 6.3698630136986296e-06,
      "loss": 0.6084,
      "step": 17250
    },
    {
      "epoch": 1.5799086757990868,
      "grad_norm": 6.067666053771973,
      "learning_rate": 6.301369863013698e-06,
      "loss": 0.61,
      "step": 17300
    },
    {
      "epoch": 1.5844748858447488,
      "grad_norm": 15.737329483032227,
      "learning_rate": 6.232876712328767e-06,
      "loss": 0.7011,
      "step": 17350
    },
    {
      "epoch": 1.589041095890411,
      "grad_norm": 4.7880730628967285,
      "learning_rate": 6.1643835616438354e-06,
      "loss": 0.694,
      "step": 17400
    },
    {
      "epoch": 1.593607305936073,
      "grad_norm": 19.6992130279541,
      "learning_rate": 6.095890410958904e-06,
      "loss": 0.706,
      "step": 17450
    },
    {
      "epoch": 1.5981735159817352,
      "grad_norm": 10.835814476013184,
      "learning_rate": 6.027397260273973e-06,
      "loss": 0.6298,
      "step": 17500
    },
    {
      "epoch": 1.6027397260273972,
      "grad_norm": 13.219555854797363,
      "learning_rate": 5.958904109589041e-06,
      "loss": 0.6496,
      "step": 17550
    },
    {
      "epoch": 1.6073059360730593,
      "grad_norm": 22.237091064453125,
      "learning_rate": 5.890410958904109e-06,
      "loss": 0.6585,
      "step": 17600
    },
    {
      "epoch": 1.6118721461187215,
      "grad_norm": 12.173138618469238,
      "learning_rate": 5.821917808219178e-06,
      "loss": 0.712,
      "step": 17650
    },
    {
      "epoch": 1.6164383561643836,
      "grad_norm": 6.5948896408081055,
      "learning_rate": 5.753424657534246e-06,
      "loss": 0.6253,
      "step": 17700
    },
    {
      "epoch": 1.6210045662100456,
      "grad_norm": 5.447400093078613,
      "learning_rate": 5.684931506849315e-06,
      "loss": 0.5762,
      "step": 17750
    },
    {
      "epoch": 1.625570776255708,
      "grad_norm": 12.413744926452637,
      "learning_rate": 5.616438356164384e-06,
      "loss": 0.6272,
      "step": 17800
    },
    {
      "epoch": 1.6301369863013697,
      "grad_norm": 48.877052307128906,
      "learning_rate": 5.547945205479452e-06,
      "loss": 0.589,
      "step": 17850
    },
    {
      "epoch": 1.634703196347032,
      "grad_norm": 21.571834564208984,
      "learning_rate": 5.47945205479452e-06,
      "loss": 0.6592,
      "step": 17900
    },
    {
      "epoch": 1.639269406392694,
      "grad_norm": 7.417336940765381,
      "learning_rate": 5.410958904109589e-06,
      "loss": 0.5248,
      "step": 17950
    },
    {
      "epoch": 1.643835616438356,
      "grad_norm": 48.4299430847168,
      "learning_rate": 5.342465753424657e-06,
      "loss": 0.6622,
      "step": 18000
    },
    {
      "epoch": 1.6484018264840183,
      "grad_norm": 17.760732650756836,
      "learning_rate": 5.273972602739726e-06,
      "loss": 0.626,
      "step": 18050
    },
    {
      "epoch": 1.6529680365296804,
      "grad_norm": 14.000800132751465,
      "learning_rate": 5.2054794520547945e-06,
      "loss": 0.6084,
      "step": 18100
    },
    {
      "epoch": 1.6575342465753424,
      "grad_norm": 8.742088317871094,
      "learning_rate": 5.136986301369863e-06,
      "loss": 0.6029,
      "step": 18150
    },
    {
      "epoch": 1.6621004566210047,
      "grad_norm": 26.16733169555664,
      "learning_rate": 5.068493150684931e-06,
      "loss": 0.6828,
      "step": 18200
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 29.90041160583496,
      "learning_rate": 4.9999999999999996e-06,
      "loss": 0.5702,
      "step": 18250
    },
    {
      "epoch": 1.6712328767123288,
      "grad_norm": 15.568696022033691,
      "learning_rate": 4.931506849315068e-06,
      "loss": 0.6376,
      "step": 18300
    },
    {
      "epoch": 1.6757990867579908,
      "grad_norm": 15.59715747833252,
      "learning_rate": 4.863013698630137e-06,
      "loss": 0.6776,
      "step": 18350
    },
    {
      "epoch": 1.6803652968036529,
      "grad_norm": 11.07044506072998,
      "learning_rate": 4.7945205479452054e-06,
      "loss": 0.529,
      "step": 18400
    },
    {
      "epoch": 1.6849315068493151,
      "grad_norm": 5.349613666534424,
      "learning_rate": 4.726027397260274e-06,
      "loss": 0.692,
      "step": 18450
    },
    {
      "epoch": 1.6894977168949772,
      "grad_norm": 18.147794723510742,
      "learning_rate": 4.657534246575343e-06,
      "loss": 0.7057,
      "step": 18500
    },
    {
      "epoch": 1.6940639269406392,
      "grad_norm": 13.486499786376953,
      "learning_rate": 4.5890410958904105e-06,
      "loss": 0.6415,
      "step": 18550
    },
    {
      "epoch": 1.6986301369863015,
      "grad_norm": 10.16304874420166,
      "learning_rate": 4.520547945205479e-06,
      "loss": 0.6604,
      "step": 18600
    },
    {
      "epoch": 1.7031963470319633,
      "grad_norm": 29.04235076904297,
      "learning_rate": 4.452054794520548e-06,
      "loss": 0.6773,
      "step": 18650
    },
    {
      "epoch": 1.7077625570776256,
      "grad_norm": 9.932119369506836,
      "learning_rate": 4.383561643835616e-06,
      "loss": 0.617,
      "step": 18700
    },
    {
      "epoch": 1.7123287671232876,
      "grad_norm": 70.96830749511719,
      "learning_rate": 4.315068493150685e-06,
      "loss": 0.6157,
      "step": 18750
    },
    {
      "epoch": 1.7168949771689497,
      "grad_norm": 15.635278701782227,
      "learning_rate": 4.246575342465754e-06,
      "loss": 0.6127,
      "step": 18800
    },
    {
      "epoch": 1.721461187214612,
      "grad_norm": 2.3553667068481445,
      "learning_rate": 4.178082191780821e-06,
      "loss": 0.6847,
      "step": 18850
    },
    {
      "epoch": 1.726027397260274,
      "grad_norm": 8.880425453186035,
      "learning_rate": 4.10958904109589e-06,
      "loss": 0.5043,
      "step": 18900
    },
    {
      "epoch": 1.730593607305936,
      "grad_norm": 22.30252456665039,
      "learning_rate": 4.041095890410959e-06,
      "loss": 0.5546,
      "step": 18950
    },
    {
      "epoch": 1.7351598173515983,
      "grad_norm": 7.0828142166137695,
      "learning_rate": 3.972602739726027e-06,
      "loss": 0.5997,
      "step": 19000
    },
    {
      "epoch": 1.7397260273972601,
      "grad_norm": 25.635570526123047,
      "learning_rate": 3.904109589041096e-06,
      "loss": 0.6091,
      "step": 19050
    },
    {
      "epoch": 1.7442922374429224,
      "grad_norm": 9.004444122314453,
      "learning_rate": 3.8356164383561645e-06,
      "loss": 0.6257,
      "step": 19100
    },
    {
      "epoch": 1.7488584474885844,
      "grad_norm": 5.485713958740234,
      "learning_rate": 3.7671232876712327e-06,
      "loss": 0.6728,
      "step": 19150
    },
    {
      "epoch": 1.7534246575342465,
      "grad_norm": 35.63444519042969,
      "learning_rate": 3.6986301369863014e-06,
      "loss": 0.7696,
      "step": 19200
    },
    {
      "epoch": 1.7579908675799087,
      "grad_norm": 29.199115753173828,
      "learning_rate": 3.6301369863013696e-06,
      "loss": 0.59,
      "step": 19250
    },
    {
      "epoch": 1.7625570776255708,
      "grad_norm": 18.041336059570312,
      "learning_rate": 3.561643835616438e-06,
      "loss": 0.5744,
      "step": 19300
    },
    {
      "epoch": 1.7671232876712328,
      "grad_norm": 16.391035079956055,
      "learning_rate": 3.493150684931507e-06,
      "loss": 0.6831,
      "step": 19350
    },
    {
      "epoch": 1.771689497716895,
      "grad_norm": 9.0728759765625,
      "learning_rate": 3.4246575342465754e-06,
      "loss": 0.5779,
      "step": 19400
    },
    {
      "epoch": 1.776255707762557,
      "grad_norm": 18.102890014648438,
      "learning_rate": 3.3561643835616436e-06,
      "loss": 0.5507,
      "step": 19450
    },
    {
      "epoch": 1.7808219178082192,
      "grad_norm": 17.248735427856445,
      "learning_rate": 3.2876712328767123e-06,
      "loss": 0.7226,
      "step": 19500
    },
    {
      "epoch": 1.7853881278538812,
      "grad_norm": 27.942777633666992,
      "learning_rate": 3.219178082191781e-06,
      "loss": 0.7178,
      "step": 19550
    },
    {
      "epoch": 1.7899543378995433,
      "grad_norm": 5.3349809646606445,
      "learning_rate": 3.150684931506849e-06,
      "loss": 0.5566,
      "step": 19600
    },
    {
      "epoch": 1.7945205479452055,
      "grad_norm": 30.73387908935547,
      "learning_rate": 3.0821917808219177e-06,
      "loss": 0.7677,
      "step": 19650
    },
    {
      "epoch": 1.7990867579908676,
      "grad_norm": 21.230749130249023,
      "learning_rate": 3.0136986301369864e-06,
      "loss": 0.61,
      "step": 19700
    },
    {
      "epoch": 1.8036529680365296,
      "grad_norm": 10.363127708435059,
      "learning_rate": 2.9452054794520546e-06,
      "loss": 0.6385,
      "step": 19750
    },
    {
      "epoch": 1.808219178082192,
      "grad_norm": 1.5289000272750854,
      "learning_rate": 2.876712328767123e-06,
      "loss": 0.6747,
      "step": 19800
    },
    {
      "epoch": 1.8127853881278537,
      "grad_norm": 18.69597625732422,
      "learning_rate": 2.808219178082192e-06,
      "loss": 0.5934,
      "step": 19850
    },
    {
      "epoch": 1.817351598173516,
      "grad_norm": 10.11032772064209,
      "learning_rate": 2.73972602739726e-06,
      "loss": 0.5918,
      "step": 19900
    },
    {
      "epoch": 1.821917808219178,
      "grad_norm": 30.333818435668945,
      "learning_rate": 2.6712328767123286e-06,
      "loss": 0.6381,
      "step": 19950
    },
    {
      "epoch": 1.82648401826484,
      "grad_norm": 15.229024887084961,
      "learning_rate": 2.6027397260273973e-06,
      "loss": 0.6151,
      "step": 20000
    },
    {
      "epoch": 1.8310502283105023,
      "grad_norm": 28.18828773498535,
      "learning_rate": 2.5342465753424655e-06,
      "loss": 0.6968,
      "step": 20050
    },
    {
      "epoch": 1.8356164383561644,
      "grad_norm": 48.18195724487305,
      "learning_rate": 2.465753424657534e-06,
      "loss": 0.6256,
      "step": 20100
    },
    {
      "epoch": 1.8401826484018264,
      "grad_norm": 16.33424949645996,
      "learning_rate": 2.3972602739726027e-06,
      "loss": 0.6477,
      "step": 20150
    },
    {
      "epoch": 1.8447488584474887,
      "grad_norm": 25.994909286499023,
      "learning_rate": 2.3287671232876713e-06,
      "loss": 0.692,
      "step": 20200
    },
    {
      "epoch": 1.8493150684931505,
      "grad_norm": 8.117030143737793,
      "learning_rate": 2.2602739726027396e-06,
      "loss": 0.6933,
      "step": 20250
    },
    {
      "epoch": 1.8538812785388128,
      "grad_norm": 8.02834415435791,
      "learning_rate": 2.191780821917808e-06,
      "loss": 0.6413,
      "step": 20300
    },
    {
      "epoch": 1.8584474885844748,
      "grad_norm": 22.62827491760254,
      "learning_rate": 2.123287671232877e-06,
      "loss": 0.679,
      "step": 20350
    },
    {
      "epoch": 1.8630136986301369,
      "grad_norm": 9.562274932861328,
      "learning_rate": 2.054794520547945e-06,
      "loss": 0.6615,
      "step": 20400
    },
    {
      "epoch": 1.8675799086757991,
      "grad_norm": 12.407808303833008,
      "learning_rate": 1.9863013698630136e-06,
      "loss": 0.7476,
      "step": 20450
    },
    {
      "epoch": 1.8721461187214612,
      "grad_norm": 41.344093322753906,
      "learning_rate": 1.9178082191780823e-06,
      "loss": 0.5827,
      "step": 20500
    },
    {
      "epoch": 1.8767123287671232,
      "grad_norm": 10.044130325317383,
      "learning_rate": 1.8493150684931507e-06,
      "loss": 0.6566,
      "step": 20550
    },
    {
      "epoch": 1.8812785388127855,
      "grad_norm": 9.382560729980469,
      "learning_rate": 1.780821917808219e-06,
      "loss": 0.6259,
      "step": 20600
    },
    {
      "epoch": 1.8858447488584473,
      "grad_norm": 9.731813430786133,
      "learning_rate": 1.7123287671232877e-06,
      "loss": 0.5924,
      "step": 20650
    },
    {
      "epoch": 1.8904109589041096,
      "grad_norm": 13.417922973632812,
      "learning_rate": 1.6438356164383561e-06,
      "loss": 0.6229,
      "step": 20700
    },
    {
      "epoch": 1.8949771689497716,
      "grad_norm": 32.03701400756836,
      "learning_rate": 1.5753424657534245e-06,
      "loss": 0.5761,
      "step": 20750
    },
    {
      "epoch": 1.8995433789954337,
      "grad_norm": 6.067290782928467,
      "learning_rate": 1.5068493150684932e-06,
      "loss": 0.6463,
      "step": 20800
    },
    {
      "epoch": 1.904109589041096,
      "grad_norm": 19.67026710510254,
      "learning_rate": 1.4383561643835616e-06,
      "loss": 0.6671,
      "step": 20850
    },
    {
      "epoch": 1.908675799086758,
      "grad_norm": 50.498802185058594,
      "learning_rate": 1.36986301369863e-06,
      "loss": 0.5811,
      "step": 20900
    },
    {
      "epoch": 1.91324200913242,
      "grad_norm": 15.981374740600586,
      "learning_rate": 1.3013698630136986e-06,
      "loss": 0.6823,
      "step": 20950
    },
    {
      "epoch": 1.9178082191780823,
      "grad_norm": 19.175485610961914,
      "learning_rate": 1.232876712328767e-06,
      "loss": 0.6112,
      "step": 21000
    },
    {
      "epoch": 1.9223744292237441,
      "grad_norm": 8.795181274414062,
      "learning_rate": 1.1643835616438357e-06,
      "loss": 0.5474,
      "step": 21050
    },
    {
      "epoch": 1.9269406392694064,
      "grad_norm": 3.590404748916626,
      "learning_rate": 1.095890410958904e-06,
      "loss": 0.7026,
      "step": 21100
    },
    {
      "epoch": 1.9315068493150684,
      "grad_norm": 21.619140625,
      "learning_rate": 1.0273972602739725e-06,
      "loss": 0.701,
      "step": 21150
    },
    {
      "epoch": 1.9360730593607305,
      "grad_norm": 22.84990692138672,
      "learning_rate": 9.589041095890411e-07,
      "loss": 0.5942,
      "step": 21200
    },
    {
      "epoch": 1.9406392694063928,
      "grad_norm": 28.548683166503906,
      "learning_rate": 8.904109589041095e-07,
      "loss": 0.6343,
      "step": 21250
    },
    {
      "epoch": 1.9452054794520548,
      "grad_norm": 20.216583251953125,
      "learning_rate": 8.219178082191781e-07,
      "loss": 0.6836,
      "step": 21300
    },
    {
      "epoch": 1.9497716894977168,
      "grad_norm": 22.950048446655273,
      "learning_rate": 7.534246575342466e-07,
      "loss": 0.6815,
      "step": 21350
    },
    {
      "epoch": 1.954337899543379,
      "grad_norm": 20.81789207458496,
      "learning_rate": 6.84931506849315e-07,
      "loss": 0.7679,
      "step": 21400
    },
    {
      "epoch": 1.958904109589041,
      "grad_norm": 14.494473457336426,
      "learning_rate": 6.164383561643835e-07,
      "loss": 0.5561,
      "step": 21450
    },
    {
      "epoch": 1.9634703196347032,
      "grad_norm": 17.119592666625977,
      "learning_rate": 5.47945205479452e-07,
      "loss": 0.6881,
      "step": 21500
    },
    {
      "epoch": 1.9680365296803652,
      "grad_norm": 8.369721412658691,
      "learning_rate": 4.794520547945206e-07,
      "loss": 0.6692,
      "step": 21550
    },
    {
      "epoch": 1.9726027397260273,
      "grad_norm": 11.264369010925293,
      "learning_rate": 4.1095890410958903e-07,
      "loss": 0.6405,
      "step": 21600
    },
    {
      "epoch": 1.9771689497716896,
      "grad_norm": 3.320608615875244,
      "learning_rate": 3.424657534246575e-07,
      "loss": 0.671,
      "step": 21650
    },
    {
      "epoch": 1.9817351598173516,
      "grad_norm": 19.23822784423828,
      "learning_rate": 2.73972602739726e-07,
      "loss": 0.5475,
      "step": 21700
    },
    {
      "epoch": 1.9863013698630136,
      "grad_norm": 13.919721603393555,
      "learning_rate": 2.0547945205479452e-07,
      "loss": 0.7422,
      "step": 21750
    },
    {
      "epoch": 1.990867579908676,
      "grad_norm": 14.141480445861816,
      "learning_rate": 1.36986301369863e-07,
      "loss": 0.5855,
      "step": 21800
    },
    {
      "epoch": 1.9954337899543377,
      "grad_norm": 19.506189346313477,
      "learning_rate": 6.84931506849315e-08,
      "loss": 0.7103,
      "step": 21850
    },
    {
      "epoch": 2.0,
      "grad_norm": 16.25570297241211,
      "learning_rate": 0.0,
      "loss": 0.7072,
      "step": 21900
    },
    {
      "epoch": 2.0,
      "step": 21900,
      "total_flos": 3.8619551920019866e+17,
      "train_loss": 0.9270246051109,
      "train_runtime": 51747.0424,
      "train_samples_per_second": 3.386,
      "train_steps_per_second": 0.423
    }
  ],
  "logging_steps": 50,
  "max_steps": 21900,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.8619551920019866e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}