{
  "best_metric": 1.4893066883087158,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.13499831252109348,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006749915626054674,
      "grad_norm": 4.020172595977783,
      "learning_rate": 1e-05,
      "loss": 6.0694,
      "step": 1
    },
    {
      "epoch": 0.0006749915626054674,
      "eval_loss": 1.9079341888427734,
      "eval_runtime": 15.821,
      "eval_samples_per_second": 157.702,
      "eval_steps_per_second": 39.441,
      "step": 1
    },
    {
      "epoch": 0.0013499831252109348,
      "grad_norm": 3.990826368331909,
      "learning_rate": 2e-05,
      "loss": 5.8031,
      "step": 2
    },
    {
      "epoch": 0.0020249746878164025,
      "grad_norm": 3.4698331356048584,
      "learning_rate": 3e-05,
      "loss": 5.9971,
      "step": 3
    },
    {
      "epoch": 0.0026999662504218697,
      "grad_norm": 3.5152740478515625,
      "learning_rate": 4e-05,
      "loss": 6.7254,
      "step": 4
    },
    {
      "epoch": 0.0033749578130273373,
      "grad_norm": 3.191807508468628,
      "learning_rate": 5e-05,
      "loss": 6.8242,
      "step": 5
    },
    {
      "epoch": 0.004049949375632805,
      "grad_norm": 3.091366767883301,
      "learning_rate": 6e-05,
      "loss": 6.9487,
      "step": 6
    },
    {
      "epoch": 0.004724940938238272,
      "grad_norm": 3.197777032852173,
      "learning_rate": 7e-05,
      "loss": 6.6845,
      "step": 7
    },
    {
      "epoch": 0.005399932500843739,
      "grad_norm": 3.2111337184906006,
      "learning_rate": 8e-05,
      "loss": 7.0961,
      "step": 8
    },
    {
      "epoch": 0.006074924063449207,
      "grad_norm": 3.5811140537261963,
      "learning_rate": 9e-05,
      "loss": 6.7339,
      "step": 9
    },
    {
      "epoch": 0.006749915626054675,
      "grad_norm": 3.24065899848938,
      "learning_rate": 0.0001,
      "loss": 6.8642,
      "step": 10
    },
    {
      "epoch": 0.007424907188660142,
      "grad_norm": 3.195089340209961,
      "learning_rate": 9.999316524962345e-05,
      "loss": 6.8105,
      "step": 11
    },
    {
      "epoch": 0.00809989875126561,
      "grad_norm": 3.6922385692596436,
      "learning_rate": 9.997266286704631e-05,
      "loss": 6.9582,
      "step": 12
    },
    {
      "epoch": 0.008774890313871077,
      "grad_norm": 3.2148289680480957,
      "learning_rate": 9.993849845741524e-05,
      "loss": 7.0517,
      "step": 13
    },
    {
      "epoch": 0.009449881876476544,
      "grad_norm": 3.1709163188934326,
      "learning_rate": 9.989068136093873e-05,
      "loss": 6.9415,
      "step": 14
    },
    {
      "epoch": 0.010124873439082012,
      "grad_norm": 3.276898145675659,
      "learning_rate": 9.98292246503335e-05,
      "loss": 7.3105,
      "step": 15
    },
    {
      "epoch": 0.010799865001687479,
      "grad_norm": 3.4331533908843994,
      "learning_rate": 9.975414512725057e-05,
      "loss": 7.2449,
      "step": 16
    },
    {
      "epoch": 0.011474856564292946,
      "grad_norm": 3.6121511459350586,
      "learning_rate": 9.966546331768191e-05,
      "loss": 7.2851,
      "step": 17
    },
    {
      "epoch": 0.012149848126898413,
      "grad_norm": 3.413409948348999,
      "learning_rate": 9.956320346634876e-05,
      "loss": 7.266,
      "step": 18
    },
    {
      "epoch": 0.01282483968950388,
      "grad_norm": 3.5942392349243164,
      "learning_rate": 9.944739353007344e-05,
      "loss": 7.0644,
      "step": 19
    },
    {
      "epoch": 0.01349983125210935,
      "grad_norm": 3.194371461868286,
      "learning_rate": 9.931806517013612e-05,
      "loss": 6.9005,
      "step": 20
    },
    {
      "epoch": 0.014174822814714817,
      "grad_norm": 3.298105478286743,
      "learning_rate": 9.917525374361912e-05,
      "loss": 6.9625,
      "step": 21
    },
    {
      "epoch": 0.014849814377320284,
      "grad_norm": 3.4364609718322754,
      "learning_rate": 9.901899829374047e-05,
      "loss": 6.9719,
      "step": 22
    },
    {
      "epoch": 0.015524805939925751,
      "grad_norm": 3.29854416847229,
      "learning_rate": 9.884934153917997e-05,
      "loss": 7.0532,
      "step": 23
    },
    {
      "epoch": 0.01619979750253122,
      "grad_norm": 3.172783851623535,
      "learning_rate": 9.86663298624003e-05,
      "loss": 6.6702,
      "step": 24
    },
    {
      "epoch": 0.016874789065136685,
      "grad_norm": 3.43276309967041,
      "learning_rate": 9.847001329696653e-05,
      "loss": 6.6978,
      "step": 25
    },
    {
      "epoch": 0.017549780627742154,
      "grad_norm": 3.337679147720337,
      "learning_rate": 9.826044551386744e-05,
      "loss": 6.9754,
      "step": 26
    },
    {
      "epoch": 0.01822477219034762,
      "grad_norm": 3.2819411754608154,
      "learning_rate": 9.803768380684242e-05,
      "loss": 7.0357,
      "step": 27
    },
    {
      "epoch": 0.01889976375295309,
      "grad_norm": 3.738227128982544,
      "learning_rate": 9.780178907671789e-05,
      "loss": 7.3609,
      "step": 28
    },
    {
      "epoch": 0.019574755315558554,
      "grad_norm": 3.6656219959259033,
      "learning_rate": 9.755282581475769e-05,
      "loss": 7.3442,
      "step": 29
    },
    {
      "epoch": 0.020249746878164023,
      "grad_norm": 3.7280478477478027,
      "learning_rate": 9.729086208503174e-05,
      "loss": 7.4316,
      "step": 30
    },
    {
      "epoch": 0.020924738440769492,
      "grad_norm": 3.73734450340271,
      "learning_rate": 9.701596950580806e-05,
      "loss": 6.2404,
      "step": 31
    },
    {
      "epoch": 0.021599730003374958,
      "grad_norm": 3.7774674892425537,
      "learning_rate": 9.672822322997305e-05,
      "loss": 7.0264,
      "step": 32
    },
    {
      "epoch": 0.022274721565980427,
      "grad_norm": 3.7192978858947754,
      "learning_rate": 9.642770192448536e-05,
      "loss": 6.9916,
      "step": 33
    },
    {
      "epoch": 0.022949713128585892,
      "grad_norm": 4.261127948760986,
      "learning_rate": 9.611448774886924e-05,
      "loss": 7.1033,
      "step": 34
    },
    {
      "epoch": 0.02362470469119136,
      "grad_norm": 4.1082892417907715,
      "learning_rate": 9.578866633275288e-05,
      "loss": 6.7066,
      "step": 35
    },
    {
      "epoch": 0.024299696253796826,
      "grad_norm": 4.049247741699219,
      "learning_rate": 9.545032675245813e-05,
      "loss": 6.6038,
      "step": 36
    },
    {
      "epoch": 0.024974687816402295,
      "grad_norm": 4.457817077636719,
      "learning_rate": 9.509956150664796e-05,
      "loss": 6.5418,
      "step": 37
    },
    {
      "epoch": 0.02564967937900776,
      "grad_norm": 4.287579536437988,
      "learning_rate": 9.473646649103818e-05,
      "loss": 6.5526,
      "step": 38
    },
    {
      "epoch": 0.02632467094161323,
      "grad_norm": 4.18779993057251,
      "learning_rate": 9.43611409721806e-05,
      "loss": 6.6445,
      "step": 39
    },
    {
      "epoch": 0.0269996625042187,
      "grad_norm": 4.44959831237793,
      "learning_rate": 9.397368756032445e-05,
      "loss": 7.0148,
      "step": 40
    },
    {
      "epoch": 0.027674654066824164,
      "grad_norm": 4.411748886108398,
      "learning_rate": 9.357421218136386e-05,
      "loss": 6.7453,
      "step": 41
    },
    {
      "epoch": 0.028349645629429633,
      "grad_norm": 4.816530227661133,
      "learning_rate": 9.316282404787871e-05,
      "loss": 6.4025,
      "step": 42
    },
    {
      "epoch": 0.0290246371920351,
      "grad_norm": 5.458149433135986,
      "learning_rate": 9.273963562927695e-05,
      "loss": 7.3557,
      "step": 43
    },
    {
      "epoch": 0.029699628754640568,
      "grad_norm": 5.505343437194824,
      "learning_rate": 9.230476262104677e-05,
      "loss": 6.9874,
      "step": 44
    },
    {
      "epoch": 0.030374620317246033,
      "grad_norm": 6.1009602546691895,
      "learning_rate": 9.185832391312644e-05,
      "loss": 6.7907,
      "step": 45
    },
    {
      "epoch": 0.031049611879851502,
      "grad_norm": 5.795923233032227,
      "learning_rate": 9.140044155740101e-05,
      "loss": 6.7509,
      "step": 46
    },
    {
      "epoch": 0.03172460344245697,
      "grad_norm": 6.259619235992432,
      "learning_rate": 9.093124073433463e-05,
      "loss": 7.04,
      "step": 47
    },
    {
      "epoch": 0.03239959500506244,
      "grad_norm": 7.380779266357422,
      "learning_rate": 9.045084971874738e-05,
      "loss": 6.8176,
      "step": 48
    },
    {
      "epoch": 0.0330745865676679,
      "grad_norm": 8.422913551330566,
      "learning_rate": 8.995939984474624e-05,
      "loss": 7.1606,
      "step": 49
    },
    {
      "epoch": 0.03374957813027337,
      "grad_norm": 15.977940559387207,
      "learning_rate": 8.945702546981969e-05,
      "loss": 7.7214,
      "step": 50
    },
    {
      "epoch": 0.03374957813027337,
      "eval_loss": 1.7495349645614624,
      "eval_runtime": 15.4146,
      "eval_samples_per_second": 161.859,
      "eval_steps_per_second": 40.481,
      "step": 50
    },
    {
      "epoch": 0.03442456969287884,
      "grad_norm": 24.358673095703125,
      "learning_rate": 8.894386393810563e-05,
      "loss": 6.8574,
      "step": 51
    },
    {
      "epoch": 0.03509956125548431,
      "grad_norm": 22.186777114868164,
      "learning_rate": 8.842005554284296e-05,
      "loss": 7.0588,
      "step": 52
    },
    {
      "epoch": 0.03577455281808977,
      "grad_norm": 17.49462890625,
      "learning_rate": 8.788574348801675e-05,
      "loss": 6.4566,
      "step": 53
    },
    {
      "epoch": 0.03644954438069524,
      "grad_norm": 10.939961433410645,
      "learning_rate": 8.73410738492077e-05,
      "loss": 6.1007,
      "step": 54
    },
    {
      "epoch": 0.03712453594330071,
      "grad_norm": 8.027196884155273,
      "learning_rate": 8.678619553365659e-05,
      "loss": 6.7321,
      "step": 55
    },
    {
      "epoch": 0.03779952750590618,
      "grad_norm": 6.052131652832031,
      "learning_rate": 8.622126023955446e-05,
      "loss": 6.6765,
      "step": 56
    },
    {
      "epoch": 0.038474519068511646,
      "grad_norm": 4.8165740966796875,
      "learning_rate": 8.564642241456986e-05,
      "loss": 6.1377,
      "step": 57
    },
    {
      "epoch": 0.03914951063111711,
      "grad_norm": 4.70635461807251,
      "learning_rate": 8.506183921362443e-05,
      "loss": 6.6126,
      "step": 58
    },
    {
      "epoch": 0.03982450219372258,
      "grad_norm": 4.0797014236450195,
      "learning_rate": 8.44676704559283e-05,
      "loss": 6.4919,
      "step": 59
    },
    {
      "epoch": 0.040499493756328046,
      "grad_norm": 3.740966558456421,
      "learning_rate": 8.386407858128706e-05,
      "loss": 6.3549,
      "step": 60
    },
    {
      "epoch": 0.041174485318933515,
      "grad_norm": 3.810687780380249,
      "learning_rate": 8.32512286056924e-05,
      "loss": 6.5293,
      "step": 61
    },
    {
      "epoch": 0.041849476881538984,
      "grad_norm": 3.6058359146118164,
      "learning_rate": 8.262928807620843e-05,
      "loss": 6.4259,
      "step": 62
    },
    {
      "epoch": 0.042524468444144446,
      "grad_norm": 3.443782329559326,
      "learning_rate": 8.199842702516583e-05,
      "loss": 6.6893,
      "step": 63
    },
    {
      "epoch": 0.043199460006749915,
      "grad_norm": 3.471390962600708,
      "learning_rate": 8.135881792367686e-05,
      "loss": 6.8213,
      "step": 64
    },
    {
      "epoch": 0.043874451569355384,
      "grad_norm": 3.128981351852417,
      "learning_rate": 8.07106356344834e-05,
      "loss": 6.8962,
      "step": 65
    },
    {
      "epoch": 0.04454944313196085,
      "grad_norm": 3.0716471672058105,
      "learning_rate": 8.005405736415126e-05,
      "loss": 6.3879,
      "step": 66
    },
    {
      "epoch": 0.045224434694566315,
      "grad_norm": 2.913241386413574,
      "learning_rate": 7.938926261462366e-05,
      "loss": 6.9029,
      "step": 67
    },
    {
      "epoch": 0.045899426257171784,
      "grad_norm": 2.8057708740234375,
      "learning_rate": 7.871643313414718e-05,
      "loss": 6.4862,
      "step": 68
    },
    {
      "epoch": 0.04657441781977725,
      "grad_norm": 3.0844922065734863,
      "learning_rate": 7.803575286758364e-05,
      "loss": 6.5822,
      "step": 69
    },
    {
      "epoch": 0.04724940938238272,
      "grad_norm": 3.082979917526245,
      "learning_rate": 7.734740790612136e-05,
      "loss": 6.8447,
      "step": 70
    },
    {
      "epoch": 0.04792440094498819,
      "grad_norm": 2.983363151550293,
      "learning_rate": 7.66515864363997e-05,
      "loss": 6.8932,
      "step": 71
    },
    {
      "epoch": 0.04859939250759365,
      "grad_norm": 2.947859525680542,
      "learning_rate": 7.594847868906076e-05,
      "loss": 6.7243,
      "step": 72
    },
    {
      "epoch": 0.04927438407019912,
      "grad_norm": 2.921736478805542,
      "learning_rate": 7.52382768867422e-05,
      "loss": 6.9261,
      "step": 73
    },
    {
      "epoch": 0.04994937563280459,
      "grad_norm": 3.0818352699279785,
      "learning_rate": 7.452117519152542e-05,
      "loss": 6.7607,
      "step": 74
    },
    {
      "epoch": 0.05062436719541006,
      "grad_norm": 2.9901797771453857,
      "learning_rate": 7.379736965185368e-05,
      "loss": 6.7537,
      "step": 75
    },
    {
      "epoch": 0.05129935875801552,
      "grad_norm": 3.0485782623291016,
      "learning_rate": 7.30670581489344e-05,
      "loss": 6.5602,
      "step": 76
    },
    {
      "epoch": 0.05197435032062099,
      "grad_norm": 2.9166438579559326,
      "learning_rate": 7.233044034264034e-05,
      "loss": 6.5486,
      "step": 77
    },
    {
      "epoch": 0.05264934188322646,
      "grad_norm": 3.095290422439575,
      "learning_rate": 7.158771761692464e-05,
      "loss": 6.5728,
      "step": 78
    },
    {
      "epoch": 0.05332433344583193,
      "grad_norm": 3.1079509258270264,
      "learning_rate": 7.083909302476453e-05,
      "loss": 6.2796,
      "step": 79
    },
    {
      "epoch": 0.0539993250084374,
      "grad_norm": 3.2958786487579346,
      "learning_rate": 7.008477123264848e-05,
      "loss": 6.4534,
      "step": 80
    },
    {
      "epoch": 0.05467431657104286,
      "grad_norm": 3.347139835357666,
      "learning_rate": 6.932495846462261e-05,
      "loss": 6.3848,
      "step": 81
    },
    {
      "epoch": 0.05534930813364833,
      "grad_norm": 3.417102098464966,
      "learning_rate": 6.855986244591104e-05,
      "loss": 6.5057,
      "step": 82
    },
    {
      "epoch": 0.0560242996962538,
      "grad_norm": 3.473957061767578,
      "learning_rate": 6.778969234612584e-05,
      "loss": 6.8314,
      "step": 83
    },
    {
      "epoch": 0.056699291258859266,
      "grad_norm": 3.4834492206573486,
      "learning_rate": 6.701465872208216e-05,
      "loss": 6.7918,
      "step": 84
    },
    {
      "epoch": 0.05737428282146473,
      "grad_norm": 3.4572885036468506,
      "learning_rate": 6.623497346023418e-05,
      "loss": 6.8511,
      "step": 85
    },
    {
      "epoch": 0.0580492743840702,
      "grad_norm": 3.816117763519287,
      "learning_rate": 6.545084971874738e-05,
      "loss": 6.0809,
      "step": 86
    },
    {
      "epoch": 0.058724265946675666,
      "grad_norm": 3.756046772003174,
      "learning_rate": 6.466250186922325e-05,
      "loss": 6.5657,
      "step": 87
    },
    {
      "epoch": 0.059399257509281135,
      "grad_norm": 3.8619883060455322,
      "learning_rate": 6.387014543809223e-05,
      "loss": 6.3795,
      "step": 88
    },
    {
      "epoch": 0.060074249071886604,
      "grad_norm": 4.117077350616455,
      "learning_rate": 6.307399704769099e-05,
      "loss": 6.8149,
      "step": 89
    },
    {
      "epoch": 0.060749240634492066,
      "grad_norm": 4.581713676452637,
      "learning_rate": 6.227427435703997e-05,
      "loss": 6.4477,
      "step": 90
    },
    {
      "epoch": 0.061424232197097535,
      "grad_norm": 4.078260898590088,
      "learning_rate": 6.147119600233758e-05,
      "loss": 6.5158,
      "step": 91
    },
    {
      "epoch": 0.062099223759703004,
      "grad_norm": 4.048325538635254,
      "learning_rate": 6.066498153718735e-05,
      "loss": 6.3714,
      "step": 92
    },
    {
      "epoch": 0.06277421532230847,
      "grad_norm": 4.1303863525390625,
      "learning_rate": 5.985585137257401e-05,
      "loss": 6.6528,
      "step": 93
    },
    {
      "epoch": 0.06344920688491394,
      "grad_norm": 4.609678745269775,
      "learning_rate": 5.90440267166055e-05,
      "loss": 6.8852,
      "step": 94
    },
    {
      "epoch": 0.0641241984475194,
      "grad_norm": 4.879993438720703,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 5.955,
      "step": 95
    },
    {
      "epoch": 0.06479919001012488,
      "grad_norm": 5.801650524139404,
      "learning_rate": 5.74131823855921e-05,
      "loss": 6.3698,
      "step": 96
    },
    {
      "epoch": 0.06547418157273034,
      "grad_norm": 5.091540336608887,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 6.6189,
      "step": 97
    },
    {
      "epoch": 0.0661491731353358,
      "grad_norm": 6.6563215255737305,
      "learning_rate": 5.577423184847932e-05,
      "loss": 6.1839,
      "step": 98
    },
    {
      "epoch": 0.06682416469794128,
      "grad_norm": 6.991439342498779,
      "learning_rate": 5.495227651252315e-05,
      "loss": 6.6778,
      "step": 99
    },
    {
      "epoch": 0.06749915626054674,
      "grad_norm": 15.223173141479492,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 5.9777,
      "step": 100
    },
    {
      "epoch": 0.06749915626054674,
      "eval_loss": 1.5713826417922974,
      "eval_runtime": 15.4214,
      "eval_samples_per_second": 161.788,
      "eval_steps_per_second": 40.463,
      "step": 100
    },
    {
      "epoch": 0.06817414782315222,
      "grad_norm": 5.9351487159729,
      "learning_rate": 5.330452921628497e-05,
      "loss": 5.8061,
      "step": 101
    },
    {
      "epoch": 0.06884913938575768,
      "grad_norm": 5.7407050132751465,
      "learning_rate": 5.247918773366112e-05,
      "loss": 5.6303,
      "step": 102
    },
    {
      "epoch": 0.06952413094836314,
      "grad_norm": 6.292360305786133,
      "learning_rate": 5.165316846586541e-05,
      "loss": 6.0706,
      "step": 103
    },
    {
      "epoch": 0.07019912251096862,
      "grad_norm": 5.218698501586914,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 5.991,
      "step": 104
    },
    {
      "epoch": 0.07087411407357408,
      "grad_norm": 4.594059944152832,
      "learning_rate": 5e-05,
      "loss": 6.3916,
      "step": 105
    },
    {
      "epoch": 0.07154910563617954,
      "grad_norm": 3.864203453063965,
      "learning_rate": 4.917330276168208e-05,
      "loss": 6.2858,
      "step": 106
    },
    {
      "epoch": 0.07222409719878502,
      "grad_norm": 3.4061079025268555,
      "learning_rate": 4.834683153413459e-05,
      "loss": 6.3032,
      "step": 107
    },
    {
      "epoch": 0.07289908876139048,
      "grad_norm": 3.3191843032836914,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 6.215,
      "step": 108
    },
    {
      "epoch": 0.07357408032399596,
      "grad_norm": 3.3675599098205566,
      "learning_rate": 4.669547078371504e-05,
      "loss": 6.5061,
      "step": 109
    },
    {
      "epoch": 0.07424907188660142,
      "grad_norm": 3.0164263248443604,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 6.183,
      "step": 110
    },
    {
      "epoch": 0.07492406344920688,
      "grad_norm": 2.910047769546509,
      "learning_rate": 4.504772348747687e-05,
      "loss": 6.0549,
      "step": 111
    },
    {
      "epoch": 0.07559905501181235,
      "grad_norm": 2.7586965560913086,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 6.1783,
      "step": 112
    },
    {
      "epoch": 0.07627404657441782,
      "grad_norm": 2.886554002761841,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 6.5103,
      "step": 113
    },
    {
      "epoch": 0.07694903813702329,
      "grad_norm": 3.0475223064422607,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 6.4468,
      "step": 114
    },
    {
      "epoch": 0.07762402969962875,
      "grad_norm": 2.837944269180298,
      "learning_rate": 4.17702704859633e-05,
      "loss": 5.7243,
      "step": 115
    },
    {
      "epoch": 0.07829902126223422,
      "grad_norm": 2.7370758056640625,
      "learning_rate": 4.095597328339452e-05,
      "loss": 6.056,
      "step": 116
    },
    {
      "epoch": 0.07897401282483969,
      "grad_norm": 2.7509467601776123,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 6.3956,
      "step": 117
    },
    {
      "epoch": 0.07964900438744515,
      "grad_norm": 2.635509729385376,
      "learning_rate": 3.933501846281267e-05,
      "loss": 6.243,
      "step": 118
    },
    {
      "epoch": 0.08032399595005063,
      "grad_norm": 2.864286422729492,
      "learning_rate": 3.852880399766243e-05,
      "loss": 6.6887,
      "step": 119
    },
    {
      "epoch": 0.08099898751265609,
      "grad_norm": 2.9439237117767334,
      "learning_rate": 3.772572564296005e-05,
      "loss": 6.0783,
      "step": 120
    },
    {
      "epoch": 0.08167397907526155,
      "grad_norm": 2.7667555809020996,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 6.4751,
      "step": 121
    },
    {
      "epoch": 0.08234897063786703,
      "grad_norm": 3.032123565673828,
      "learning_rate": 3.612985456190778e-05,
      "loss": 6.537,
      "step": 122
    },
    {
      "epoch": 0.08302396220047249,
      "grad_norm": 2.868000030517578,
      "learning_rate": 3.533749813077677e-05,
      "loss": 6.4228,
      "step": 123
    },
    {
      "epoch": 0.08369895376307797,
      "grad_norm": 2.971174955368042,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 6.3492,
      "step": 124
    },
    {
      "epoch": 0.08437394532568343,
      "grad_norm": 2.86604905128479,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 6.3523,
      "step": 125
    },
    {
      "epoch": 0.08504893688828889,
      "grad_norm": 2.9495882987976074,
      "learning_rate": 3.298534127791785e-05,
      "loss": 6.4747,
      "step": 126
    },
    {
      "epoch": 0.08572392845089437,
      "grad_norm": 2.965926170349121,
      "learning_rate": 3.221030765387417e-05,
      "loss": 6.3026,
      "step": 127
    },
    {
      "epoch": 0.08639892001349983,
      "grad_norm": 3.2158634662628174,
      "learning_rate": 3.144013755408895e-05,
      "loss": 6.5068,
      "step": 128
    },
    {
      "epoch": 0.08707391157610529,
      "grad_norm": 3.074702262878418,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 6.296,
      "step": 129
    },
    {
      "epoch": 0.08774890313871077,
      "grad_norm": 3.2940545082092285,
      "learning_rate": 2.991522876735154e-05,
      "loss": 6.2292,
      "step": 130
    },
    {
      "epoch": 0.08842389470131623,
      "grad_norm": 3.1414380073547363,
      "learning_rate": 2.916090697523549e-05,
      "loss": 6.2769,
      "step": 131
    },
    {
      "epoch": 0.0890988862639217,
      "grad_norm": 3.420393228530884,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 6.1765,
      "step": 132
    },
    {
      "epoch": 0.08977387782652717,
      "grad_norm": 3.522200107574463,
      "learning_rate": 2.766955965735968e-05,
      "loss": 5.8471,
      "step": 133
    },
    {
      "epoch": 0.09044886938913263,
      "grad_norm": 3.513277053833008,
      "learning_rate": 2.693294185106562e-05,
      "loss": 6.3863,
      "step": 134
    },
    {
      "epoch": 0.0911238609517381,
      "grad_norm": 3.4822897911071777,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 6.2767,
      "step": 135
    },
    {
      "epoch": 0.09179885251434357,
      "grad_norm": 3.4105777740478516,
      "learning_rate": 2.547882480847461e-05,
      "loss": 6.0907,
      "step": 136
    },
    {
      "epoch": 0.09247384407694904,
      "grad_norm": 3.446399450302124,
      "learning_rate": 2.476172311325783e-05,
      "loss": 6.0758,
      "step": 137
    },
    {
      "epoch": 0.0931488356395545,
      "grad_norm": 3.736788511276245,
      "learning_rate": 2.405152131093926e-05,
      "loss": 6.4812,
      "step": 138
    },
    {
      "epoch": 0.09382382720215997,
      "grad_norm": 3.8830013275146484,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 6.386,
      "step": 139
    },
    {
      "epoch": 0.09449881876476544,
      "grad_norm": 4.05012321472168,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 5.9687,
      "step": 140
    },
    {
      "epoch": 0.0951738103273709,
      "grad_norm": 4.038466453552246,
      "learning_rate": 2.196424713241637e-05,
      "loss": 6.0553,
      "step": 141
    },
    {
      "epoch": 0.09584880188997638,
      "grad_norm": 4.241274833679199,
      "learning_rate": 2.128356686585282e-05,
      "loss": 6.1175,
      "step": 142
    },
    {
      "epoch": 0.09652379345258184,
      "grad_norm": 4.591134548187256,
      "learning_rate": 2.061073738537635e-05,
      "loss": 6.5335,
      "step": 143
    },
    {
      "epoch": 0.0971987850151873,
      "grad_norm": 4.445840835571289,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 6.0318,
      "step": 144
    },
    {
      "epoch": 0.09787377657779278,
      "grad_norm": 4.963098526000977,
      "learning_rate": 1.928936436551661e-05,
      "loss": 6.5899,
      "step": 145
    },
    {
      "epoch": 0.09854876814039824,
      "grad_norm": 5.24894380569458,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 5.8648,
      "step": 146
    },
    {
      "epoch": 0.09922375970300372,
      "grad_norm": 6.118842124938965,
      "learning_rate": 1.800157297483417e-05,
      "loss": 6.4922,
      "step": 147
    },
    {
      "epoch": 0.09989875126560918,
      "grad_norm": 6.529957294464111,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 5.4712,
      "step": 148
    },
    {
      "epoch": 0.10057374282821464,
      "grad_norm": 8.207730293273926,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 6.9457,
      "step": 149
    },
    {
      "epoch": 0.10124873439082012,
      "grad_norm": 15.415940284729004,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 6.4275,
      "step": 150
    },
    {
      "epoch": 0.10124873439082012,
      "eval_loss": 1.5056450366973877,
      "eval_runtime": 15.3995,
      "eval_samples_per_second": 162.018,
      "eval_steps_per_second": 40.521,
      "step": 150
    },
    {
      "epoch": 0.10192372595342558,
      "grad_norm": 3.269871473312378,
      "learning_rate": 1.553232954407171e-05,
      "loss": 5.1203,
      "step": 151
    },
    {
      "epoch": 0.10259871751603104,
      "grad_norm": 3.434751033782959,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 5.154,
      "step": 152
    },
    {
      "epoch": 0.10327370907863652,
      "grad_norm": 3.432587146759033,
      "learning_rate": 1.435357758543015e-05,
      "loss": 5.8263,
      "step": 153
    },
    {
      "epoch": 0.10394870064124198,
      "grad_norm": 3.3248212337493896,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 5.7504,
      "step": 154
    },
    {
      "epoch": 0.10462369220384746,
      "grad_norm": 3.362471580505371,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 5.7939,
      "step": 155
    },
    {
      "epoch": 0.10529868376645292,
      "grad_norm": 3.074429750442505,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 5.9893,
      "step": 156
    },
    {
      "epoch": 0.10597367532905838,
      "grad_norm": 2.9347994327545166,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 6.0377,
      "step": 157
    },
    {
      "epoch": 0.10664866689166386,
      "grad_norm": 3.037919759750366,
      "learning_rate": 1.157994445715706e-05,
      "loss": 5.9979,
      "step": 158
    },
    {
      "epoch": 0.10732365845426932,
      "grad_norm": 3.0758988857269287,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 6.2559,
      "step": 159
    },
    {
      "epoch": 0.1079986500168748,
      "grad_norm": 2.9436991214752197,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 6.1055,
      "step": 160
    },
    {
      "epoch": 0.10867364157948026,
      "grad_norm": 2.866224765777588,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 5.9472,
      "step": 161
    },
    {
      "epoch": 0.10934863314208572,
      "grad_norm": 2.8063206672668457,
      "learning_rate": 9.549150281252633e-06,
      "loss": 6.0211,
      "step": 162
    },
    {
      "epoch": 0.1100236247046912,
      "grad_norm": 2.7773265838623047,
      "learning_rate": 9.068759265665384e-06,
      "loss": 6.2256,
      "step": 163
    },
    {
      "epoch": 0.11069861626729666,
      "grad_norm": 2.6896560192108154,
      "learning_rate": 8.599558442598998e-06,
      "loss": 6.1602,
      "step": 164
    },
    {
      "epoch": 0.11137360782990213,
      "grad_norm": 2.7699813842773438,
      "learning_rate": 8.141676086873572e-06,
      "loss": 6.172,
      "step": 165
    },
    {
      "epoch": 0.1120485993925076,
      "grad_norm": 2.883126974105835,
      "learning_rate": 7.695237378953223e-06,
      "loss": 5.9709,
      "step": 166
    },
    {
      "epoch": 0.11272359095511306,
      "grad_norm": 2.8404297828674316,
      "learning_rate": 7.260364370723044e-06,
      "loss": 6.5008,
      "step": 167
    },
    {
      "epoch": 0.11339858251771853,
      "grad_norm": 2.6605091094970703,
      "learning_rate": 6.837175952121306e-06,
      "loss": 6.2458,
      "step": 168
    },
    {
      "epoch": 0.114073574080324,
      "grad_norm": 2.715022563934326,
      "learning_rate": 6.425787818636131e-06,
      "loss": 6.5226,
      "step": 169
    },
    {
      "epoch": 0.11474856564292946,
      "grad_norm": 2.846055030822754,
      "learning_rate": 6.026312439675552e-06,
      "loss": 6.3761,
      "step": 170
    },
    {
      "epoch": 0.11542355720553493,
      "grad_norm": 2.8522748947143555,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 6.1735,
      "step": 171
    },
    {
      "epoch": 0.1160985487681404,
      "grad_norm": 2.817167282104492,
      "learning_rate": 5.263533508961827e-06,
      "loss": 5.9526,
      "step": 172
    },
    {
      "epoch": 0.11677354033074587,
      "grad_norm": 3.126849412918091,
      "learning_rate": 4.900438493352055e-06,
      "loss": 6.1235,
      "step": 173
    },
    {
      "epoch": 0.11744853189335133,
      "grad_norm": 2.836930513381958,
      "learning_rate": 4.549673247541875e-06,
      "loss": 6.1303,
      "step": 174
    },
    {
      "epoch": 0.1181235234559568,
      "grad_norm": 2.900245189666748,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 6.17,
      "step": 175
    },
    {
      "epoch": 0.11879851501856227,
      "grad_norm": 3.005939245223999,
      "learning_rate": 3.885512251130763e-06,
      "loss": 6.0476,
      "step": 176
    },
    {
      "epoch": 0.11947350658116773,
      "grad_norm": 2.9445931911468506,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 6.3625,
      "step": 177
    },
    {
      "epoch": 0.12014849814377321,
      "grad_norm": 2.978919267654419,
      "learning_rate": 3.271776770026963e-06,
      "loss": 6.6415,
      "step": 178
    },
    {
      "epoch": 0.12082348970637867,
      "grad_norm": 3.075558662414551,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 6.5928,
      "step": 179
    },
    {
      "epoch": 0.12149848126898413,
      "grad_norm": 2.996297836303711,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 5.9654,
      "step": 180
    },
    {
      "epoch": 0.12217347283158961,
      "grad_norm": 3.0895910263061523,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 6.0993,
      "step": 181
    },
    {
      "epoch": 0.12284846439419507,
      "grad_norm": 3.230306386947632,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 6.0599,
      "step": 182
    },
    {
      "epoch": 0.12352345595680055,
      "grad_norm": 3.189131736755371,
      "learning_rate": 1.962316193157593e-06,
      "loss": 6.3485,
      "step": 183
    },
    {
      "epoch": 0.12419844751940601,
      "grad_norm": 3.424772024154663,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 6.2527,
      "step": 184
    },
    {
      "epoch": 0.12487343908201147,
      "grad_norm": 3.5832743644714355,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 6.468,
      "step": 185
    },
    {
      "epoch": 0.12554843064461693,
      "grad_norm": 3.5637094974517822,
      "learning_rate": 1.333670137599713e-06,
      "loss": 6.1272,
      "step": 186
    },
    {
      "epoch": 0.1262234222072224,
      "grad_norm": 3.6116654872894287,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 5.9774,
      "step": 187
    },
    {
      "epoch": 0.12689841376982788,
      "grad_norm": 4.070122718811035,
      "learning_rate": 9.810017062595322e-07,
      "loss": 6.2395,
      "step": 188
    },
    {
      "epoch": 0.12757340533243333,
      "grad_norm": 3.9304137229919434,
      "learning_rate": 8.247462563808817e-07,
      "loss": 6.0866,
      "step": 189
    },
    {
      "epoch": 0.1282483968950388,
      "grad_norm": 3.9991202354431152,
      "learning_rate": 6.819348298638839e-07,
      "loss": 5.7749,
      "step": 190
    },
    {
      "epoch": 0.12892338845764428,
      "grad_norm": 4.074830532073975,
      "learning_rate": 5.526064699265753e-07,
      "loss": 6.3257,
      "step": 191
    },
    {
      "epoch": 0.12959838002024976,
      "grad_norm": 4.401994228363037,
      "learning_rate": 4.367965336512403e-07,
      "loss": 6.2388,
      "step": 192
    },
    {
      "epoch": 0.1302733715828552,
      "grad_norm": 4.41982889175415,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 6.1065,
      "step": 193
    },
    {
      "epoch": 0.13094836314546068,
      "grad_norm": 4.914302349090576,
      "learning_rate": 2.458548727494292e-07,
      "loss": 6.1827,
      "step": 194
    },
    {
      "epoch": 0.13162335470806616,
      "grad_norm": 5.346338748931885,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 6.1515,
      "step": 195
    },
    {
      "epoch": 0.1322983462706716,
      "grad_norm": 5.664148807525635,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 6.0913,
      "step": 196
    },
    {
      "epoch": 0.13297333783327708,
      "grad_norm": 6.124570846557617,
      "learning_rate": 6.150154258476315e-08,
      "loss": 6.699,
      "step": 197
    },
    {
      "epoch": 0.13364832939588256,
      "grad_norm": 7.166781425476074,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 5.8216,
      "step": 198
    },
    {
      "epoch": 0.134323320958488,
      "grad_norm": 8.77023696899414,
      "learning_rate": 6.834750376549792e-09,
      "loss": 6.7824,
      "step": 199
    },
    {
      "epoch": 0.13499831252109348,
      "grad_norm": 22.04912757873535,
      "learning_rate": 0.0,
      "loss": 6.478,
      "step": 200
    },
    {
      "epoch": 0.13499831252109348,
      "eval_loss": 1.4893066883087158,
      "eval_runtime": 15.7092,
      "eval_samples_per_second": 158.824,
      "eval_steps_per_second": 39.722,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "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": 3978324338540544.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}