{
  "best_metric": 1.8256925344467163,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.43360433604336046,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002168021680216802,
      "grad_norm": 29.12859344482422,
      "learning_rate": 1e-05,
      "loss": 6.7917,
      "step": 1
    },
    {
      "epoch": 0.002168021680216802,
      "eval_loss": 2.9564599990844727,
      "eval_runtime": 9.2682,
      "eval_samples_per_second": 83.835,
      "eval_steps_per_second": 21.04,
      "step": 1
    },
    {
      "epoch": 0.004336043360433604,
      "grad_norm": 22.322139739990234,
      "learning_rate": 2e-05,
      "loss": 8.8215,
      "step": 2
    },
    {
      "epoch": 0.0065040650406504065,
      "grad_norm": 25.346025466918945,
      "learning_rate": 3e-05,
      "loss": 7.7766,
      "step": 3
    },
    {
      "epoch": 0.008672086720867209,
      "grad_norm": 27.97523307800293,
      "learning_rate": 4e-05,
      "loss": 9.6007,
      "step": 4
    },
    {
      "epoch": 0.01084010840108401,
      "grad_norm": 50.85835647583008,
      "learning_rate": 5e-05,
      "loss": 8.8435,
      "step": 5
    },
    {
      "epoch": 0.013008130081300813,
      "grad_norm": 22.54094696044922,
      "learning_rate": 6e-05,
      "loss": 8.705,
      "step": 6
    },
    {
      "epoch": 0.015176151761517615,
      "grad_norm": 21.17547607421875,
      "learning_rate": 7e-05,
      "loss": 8.4503,
      "step": 7
    },
    {
      "epoch": 0.017344173441734417,
      "grad_norm": 25.830350875854492,
      "learning_rate": 8e-05,
      "loss": 9.4576,
      "step": 8
    },
    {
      "epoch": 0.01951219512195122,
      "grad_norm": 27.319122314453125,
      "learning_rate": 9e-05,
      "loss": 9.6108,
      "step": 9
    },
    {
      "epoch": 0.02168021680216802,
      "grad_norm": 30.378042221069336,
      "learning_rate": 0.0001,
      "loss": 9.102,
      "step": 10
    },
    {
      "epoch": 0.023848238482384824,
      "grad_norm": 39.58894348144531,
      "learning_rate": 9.999316524962345e-05,
      "loss": 9.3861,
      "step": 11
    },
    {
      "epoch": 0.026016260162601626,
      "grad_norm": 50.28947448730469,
      "learning_rate": 9.997266286704631e-05,
      "loss": 9.2129,
      "step": 12
    },
    {
      "epoch": 0.028184281842818428,
      "grad_norm": 53.55912399291992,
      "learning_rate": 9.993849845741524e-05,
      "loss": 9.4558,
      "step": 13
    },
    {
      "epoch": 0.03035230352303523,
      "grad_norm": 114.91228485107422,
      "learning_rate": 9.989068136093873e-05,
      "loss": 9.4151,
      "step": 14
    },
    {
      "epoch": 0.032520325203252036,
      "grad_norm": 73.47563934326172,
      "learning_rate": 9.98292246503335e-05,
      "loss": 9.156,
      "step": 15
    },
    {
      "epoch": 0.034688346883468835,
      "grad_norm": 56.4515266418457,
      "learning_rate": 9.975414512725057e-05,
      "loss": 9.5922,
      "step": 16
    },
    {
      "epoch": 0.03685636856368564,
      "grad_norm": 60.218231201171875,
      "learning_rate": 9.966546331768191e-05,
      "loss": 9.5008,
      "step": 17
    },
    {
      "epoch": 0.03902439024390244,
      "grad_norm": 58.83360290527344,
      "learning_rate": 9.956320346634876e-05,
      "loss": 9.4523,
      "step": 18
    },
    {
      "epoch": 0.041192411924119245,
      "grad_norm": 71.57213592529297,
      "learning_rate": 9.944739353007344e-05,
      "loss": 9.0671,
      "step": 19
    },
    {
      "epoch": 0.04336043360433604,
      "grad_norm": 68.44664764404297,
      "learning_rate": 9.931806517013612e-05,
      "loss": 10.3774,
      "step": 20
    },
    {
      "epoch": 0.04552845528455285,
      "grad_norm": 125.75868225097656,
      "learning_rate": 9.917525374361912e-05,
      "loss": 10.2201,
      "step": 21
    },
    {
      "epoch": 0.04769647696476965,
      "grad_norm": 70.28474426269531,
      "learning_rate": 9.901899829374047e-05,
      "loss": 10.1407,
      "step": 22
    },
    {
      "epoch": 0.04986449864498645,
      "grad_norm": 63.97534942626953,
      "learning_rate": 9.884934153917997e-05,
      "loss": 10.2355,
      "step": 23
    },
    {
      "epoch": 0.05203252032520325,
      "grad_norm": 65.8618392944336,
      "learning_rate": 9.86663298624003e-05,
      "loss": 9.0292,
      "step": 24
    },
    {
      "epoch": 0.05420054200542006,
      "grad_norm": 63.874446868896484,
      "learning_rate": 9.847001329696653e-05,
      "loss": 9.6761,
      "step": 25
    },
    {
      "epoch": 0.056368563685636856,
      "grad_norm": 82.74284362792969,
      "learning_rate": 9.826044551386744e-05,
      "loss": 10.2796,
      "step": 26
    },
    {
      "epoch": 0.05853658536585366,
      "grad_norm": 68.2499008178711,
      "learning_rate": 9.803768380684242e-05,
      "loss": 10.0398,
      "step": 27
    },
    {
      "epoch": 0.06070460704607046,
      "grad_norm": 72.59913635253906,
      "learning_rate": 9.780178907671789e-05,
      "loss": 9.8312,
      "step": 28
    },
    {
      "epoch": 0.06287262872628727,
      "grad_norm": 63.439430236816406,
      "learning_rate": 9.755282581475769e-05,
      "loss": 9.8374,
      "step": 29
    },
    {
      "epoch": 0.06504065040650407,
      "grad_norm": 83.34996795654297,
      "learning_rate": 9.729086208503174e-05,
      "loss": 9.8564,
      "step": 30
    },
    {
      "epoch": 0.06720867208672086,
      "grad_norm": 70.10639953613281,
      "learning_rate": 9.701596950580806e-05,
      "loss": 10.2465,
      "step": 31
    },
    {
      "epoch": 0.06937669376693767,
      "grad_norm": 102.28328704833984,
      "learning_rate": 9.672822322997305e-05,
      "loss": 10.5723,
      "step": 32
    },
    {
      "epoch": 0.07154471544715447,
      "grad_norm": 88.00032806396484,
      "learning_rate": 9.642770192448536e-05,
      "loss": 10.5202,
      "step": 33
    },
    {
      "epoch": 0.07371273712737128,
      "grad_norm": 103.43360137939453,
      "learning_rate": 9.611448774886924e-05,
      "loss": 10.4615,
      "step": 34
    },
    {
      "epoch": 0.07588075880758807,
      "grad_norm": 142.5089569091797,
      "learning_rate": 9.578866633275288e-05,
      "loss": 13.2118,
      "step": 35
    },
    {
      "epoch": 0.07804878048780488,
      "grad_norm": 136.34939575195312,
      "learning_rate": 9.545032675245813e-05,
      "loss": 10.4707,
      "step": 36
    },
    {
      "epoch": 0.08021680216802168,
      "grad_norm": 132.55331420898438,
      "learning_rate": 9.509956150664796e-05,
      "loss": 9.9917,
      "step": 37
    },
    {
      "epoch": 0.08238482384823849,
      "grad_norm": 244.61973571777344,
      "learning_rate": 9.473646649103818e-05,
      "loss": 10.6348,
      "step": 38
    },
    {
      "epoch": 0.08455284552845528,
      "grad_norm": 242.89825439453125,
      "learning_rate": 9.43611409721806e-05,
      "loss": 12.4025,
      "step": 39
    },
    {
      "epoch": 0.08672086720867209,
      "grad_norm": 342.6029052734375,
      "learning_rate": 9.397368756032445e-05,
      "loss": 13.344,
      "step": 40
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 272.9972839355469,
      "learning_rate": 9.357421218136386e-05,
      "loss": 11.3174,
      "step": 41
    },
    {
      "epoch": 0.0910569105691057,
      "grad_norm": 263.7679443359375,
      "learning_rate": 9.316282404787871e-05,
      "loss": 11.4468,
      "step": 42
    },
    {
      "epoch": 0.09322493224932249,
      "grad_norm": 230.99366760253906,
      "learning_rate": 9.273963562927695e-05,
      "loss": 9.3454,
      "step": 43
    },
    {
      "epoch": 0.0953929539295393,
      "grad_norm": 312.9385986328125,
      "learning_rate": 9.230476262104677e-05,
      "loss": 9.2028,
      "step": 44
    },
    {
      "epoch": 0.0975609756097561,
      "grad_norm": 334.3559875488281,
      "learning_rate": 9.185832391312644e-05,
      "loss": 7.8907,
      "step": 45
    },
    {
      "epoch": 0.0997289972899729,
      "grad_norm": 363.6707458496094,
      "learning_rate": 9.140044155740101e-05,
      "loss": 6.8039,
      "step": 46
    },
    {
      "epoch": 0.1018970189701897,
      "grad_norm": 266.79193115234375,
      "learning_rate": 9.093124073433463e-05,
      "loss": 7.2706,
      "step": 47
    },
    {
      "epoch": 0.1040650406504065,
      "grad_norm": 309.787109375,
      "learning_rate": 9.045084971874738e-05,
      "loss": 5.2873,
      "step": 48
    },
    {
      "epoch": 0.10623306233062331,
      "grad_norm": 336.9679260253906,
      "learning_rate": 8.995939984474624e-05,
      "loss": 5.0446,
      "step": 49
    },
    {
      "epoch": 0.10840108401084012,
      "grad_norm": 222.40145874023438,
      "learning_rate": 8.945702546981969e-05,
      "loss": 4.4654,
      "step": 50
    },
    {
      "epoch": 0.10840108401084012,
      "eval_loss": 2.6827785968780518,
      "eval_runtime": 9.2082,
      "eval_samples_per_second": 84.381,
      "eval_steps_per_second": 21.177,
      "step": 50
    },
    {
      "epoch": 0.11056910569105691,
      "grad_norm": 152.42861938476562,
      "learning_rate": 8.894386393810563e-05,
      "loss": 6.4405,
      "step": 51
    },
    {
      "epoch": 0.11273712737127371,
      "grad_norm": 174.4921875,
      "learning_rate": 8.842005554284296e-05,
      "loss": 11.5187,
      "step": 52
    },
    {
      "epoch": 0.11490514905149052,
      "grad_norm": 119.81610107421875,
      "learning_rate": 8.788574348801675e-05,
      "loss": 13.9507,
      "step": 53
    },
    {
      "epoch": 0.11707317073170732,
      "grad_norm": 107.01353454589844,
      "learning_rate": 8.73410738492077e-05,
      "loss": 11.5721,
      "step": 54
    },
    {
      "epoch": 0.11924119241192412,
      "grad_norm": 119.64827728271484,
      "learning_rate": 8.678619553365659e-05,
      "loss": 11.2909,
      "step": 55
    },
    {
      "epoch": 0.12140921409214092,
      "grad_norm": 93.13258361816406,
      "learning_rate": 8.622126023955446e-05,
      "loss": 11.1453,
      "step": 56
    },
    {
      "epoch": 0.12357723577235773,
      "grad_norm": 96.87866973876953,
      "learning_rate": 8.564642241456986e-05,
      "loss": 10.4206,
      "step": 57
    },
    {
      "epoch": 0.12574525745257453,
      "grad_norm": 89.92967224121094,
      "learning_rate": 8.506183921362443e-05,
      "loss": 9.8702,
      "step": 58
    },
    {
      "epoch": 0.12791327913279132,
      "grad_norm": 108.23527526855469,
      "learning_rate": 8.44676704559283e-05,
      "loss": 9.9942,
      "step": 59
    },
    {
      "epoch": 0.13008130081300814,
      "grad_norm": 105.78068542480469,
      "learning_rate": 8.386407858128706e-05,
      "loss": 9.8182,
      "step": 60
    },
    {
      "epoch": 0.13224932249322494,
      "grad_norm": 99.80757141113281,
      "learning_rate": 8.32512286056924e-05,
      "loss": 9.7978,
      "step": 61
    },
    {
      "epoch": 0.13441734417344173,
      "grad_norm": 84.73035430908203,
      "learning_rate": 8.262928807620843e-05,
      "loss": 9.1806,
      "step": 62
    },
    {
      "epoch": 0.13658536585365855,
      "grad_norm": 93.6194076538086,
      "learning_rate": 8.199842702516583e-05,
      "loss": 9.8823,
      "step": 63
    },
    {
      "epoch": 0.13875338753387534,
      "grad_norm": 100.37662506103516,
      "learning_rate": 8.135881792367686e-05,
      "loss": 8.9169,
      "step": 64
    },
    {
      "epoch": 0.14092140921409213,
      "grad_norm": 83.19640350341797,
      "learning_rate": 8.07106356344834e-05,
      "loss": 9.8207,
      "step": 65
    },
    {
      "epoch": 0.14308943089430895,
      "grad_norm": 80.64735412597656,
      "learning_rate": 8.005405736415126e-05,
      "loss": 10.1058,
      "step": 66
    },
    {
      "epoch": 0.14525745257452574,
      "grad_norm": 94.49461364746094,
      "learning_rate": 7.938926261462366e-05,
      "loss": 9.7281,
      "step": 67
    },
    {
      "epoch": 0.14742547425474256,
      "grad_norm": 91.92240142822266,
      "learning_rate": 7.871643313414718e-05,
      "loss": 9.4357,
      "step": 68
    },
    {
      "epoch": 0.14959349593495935,
      "grad_norm": 69.79247283935547,
      "learning_rate": 7.803575286758364e-05,
      "loss": 10.6283,
      "step": 69
    },
    {
      "epoch": 0.15176151761517614,
      "grad_norm": 72.7747573852539,
      "learning_rate": 7.734740790612136e-05,
      "loss": 10.0402,
      "step": 70
    },
    {
      "epoch": 0.15392953929539296,
      "grad_norm": 68.02134704589844,
      "learning_rate": 7.66515864363997e-05,
      "loss": 9.9393,
      "step": 71
    },
    {
      "epoch": 0.15609756097560976,
      "grad_norm": 98.18658447265625,
      "learning_rate": 7.594847868906076e-05,
      "loss": 10.1389,
      "step": 72
    },
    {
      "epoch": 0.15826558265582655,
      "grad_norm": 73.68216705322266,
      "learning_rate": 7.52382768867422e-05,
      "loss": 10.3997,
      "step": 73
    },
    {
      "epoch": 0.16043360433604337,
      "grad_norm": 69.40960693359375,
      "learning_rate": 7.452117519152542e-05,
      "loss": 10.1879,
      "step": 74
    },
    {
      "epoch": 0.16260162601626016,
      "grad_norm": 73.4581298828125,
      "learning_rate": 7.379736965185368e-05,
      "loss": 10.4545,
      "step": 75
    },
    {
      "epoch": 0.16476964769647698,
      "grad_norm": 70.93222045898438,
      "learning_rate": 7.30670581489344e-05,
      "loss": 9.8091,
      "step": 76
    },
    {
      "epoch": 0.16693766937669377,
      "grad_norm": 99.13206481933594,
      "learning_rate": 7.233044034264034e-05,
      "loss": 9.5552,
      "step": 77
    },
    {
      "epoch": 0.16910569105691056,
      "grad_norm": 62.624671936035156,
      "learning_rate": 7.158771761692464e-05,
      "loss": 9.6037,
      "step": 78
    },
    {
      "epoch": 0.17127371273712738,
      "grad_norm": 134.58192443847656,
      "learning_rate": 7.083909302476453e-05,
      "loss": 10.0818,
      "step": 79
    },
    {
      "epoch": 0.17344173441734417,
      "grad_norm": 85.55967712402344,
      "learning_rate": 7.008477123264848e-05,
      "loss": 8.9797,
      "step": 80
    },
    {
      "epoch": 0.17560975609756097,
      "grad_norm": 75.0257568359375,
      "learning_rate": 6.932495846462261e-05,
      "loss": 9.0529,
      "step": 81
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 91.2174301147461,
      "learning_rate": 6.855986244591104e-05,
      "loss": 9.4115,
      "step": 82
    },
    {
      "epoch": 0.17994579945799458,
      "grad_norm": 69.63128662109375,
      "learning_rate": 6.778969234612584e-05,
      "loss": 9.3689,
      "step": 83
    },
    {
      "epoch": 0.1821138211382114,
      "grad_norm": 100.88275146484375,
      "learning_rate": 6.701465872208216e-05,
      "loss": 10.4355,
      "step": 84
    },
    {
      "epoch": 0.1842818428184282,
      "grad_norm": 111.9276123046875,
      "learning_rate": 6.623497346023418e-05,
      "loss": 10.0252,
      "step": 85
    },
    {
      "epoch": 0.18644986449864498,
      "grad_norm": 112.8913345336914,
      "learning_rate": 6.545084971874738e-05,
      "loss": 9.4463,
      "step": 86
    },
    {
      "epoch": 0.1886178861788618,
      "grad_norm": 99.43082427978516,
      "learning_rate": 6.466250186922325e-05,
      "loss": 8.8526,
      "step": 87
    },
    {
      "epoch": 0.1907859078590786,
      "grad_norm": 131.60076904296875,
      "learning_rate": 6.387014543809223e-05,
      "loss": 9.8509,
      "step": 88
    },
    {
      "epoch": 0.19295392953929538,
      "grad_norm": 107.29341888427734,
      "learning_rate": 6.307399704769099e-05,
      "loss": 7.5175,
      "step": 89
    },
    {
      "epoch": 0.1951219512195122,
      "grad_norm": 156.77398681640625,
      "learning_rate": 6.227427435703997e-05,
      "loss": 8.271,
      "step": 90
    },
    {
      "epoch": 0.197289972899729,
      "grad_norm": 134.43824768066406,
      "learning_rate": 6.147119600233758e-05,
      "loss": 5.7906,
      "step": 91
    },
    {
      "epoch": 0.1994579945799458,
      "grad_norm": 138.85414123535156,
      "learning_rate": 6.066498153718735e-05,
      "loss": 6.0581,
      "step": 92
    },
    {
      "epoch": 0.2016260162601626,
      "grad_norm": 194.57774353027344,
      "learning_rate": 5.985585137257401e-05,
      "loss": 4.162,
      "step": 93
    },
    {
      "epoch": 0.2037940379403794,
      "grad_norm": 148.40780639648438,
      "learning_rate": 5.90440267166055e-05,
      "loss": 3.3101,
      "step": 94
    },
    {
      "epoch": 0.20596205962059622,
      "grad_norm": 100.26160430908203,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 3.2583,
      "step": 95
    },
    {
      "epoch": 0.208130081300813,
      "grad_norm": 93.37901306152344,
      "learning_rate": 5.74131823855921e-05,
      "loss": 2.3772,
      "step": 96
    },
    {
      "epoch": 0.2102981029810298,
      "grad_norm": 132.9501190185547,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 2.5803,
      "step": 97
    },
    {
      "epoch": 0.21246612466124662,
      "grad_norm": 169.74169921875,
      "learning_rate": 5.577423184847932e-05,
      "loss": 1.8584,
      "step": 98
    },
    {
      "epoch": 0.2146341463414634,
      "grad_norm": 103.20438385009766,
      "learning_rate": 5.495227651252315e-05,
      "loss": 0.8846,
      "step": 99
    },
    {
      "epoch": 0.21680216802168023,
      "grad_norm": 124.05949401855469,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 2.7046,
      "step": 100
    },
    {
      "epoch": 0.21680216802168023,
      "eval_loss": 2.0304088592529297,
      "eval_runtime": 9.2411,
      "eval_samples_per_second": 84.081,
      "eval_steps_per_second": 21.101,
      "step": 100
    },
    {
      "epoch": 0.21897018970189702,
      "grad_norm": 42.41352081298828,
      "learning_rate": 5.330452921628497e-05,
      "loss": 5.1714,
      "step": 101
    },
    {
      "epoch": 0.22113821138211381,
      "grad_norm": 55.999114990234375,
      "learning_rate": 5.247918773366112e-05,
      "loss": 12.1885,
      "step": 102
    },
    {
      "epoch": 0.22330623306233063,
      "grad_norm": 44.16761016845703,
      "learning_rate": 5.165316846586541e-05,
      "loss": 9.4398,
      "step": 103
    },
    {
      "epoch": 0.22547425474254743,
      "grad_norm": 49.66854476928711,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 10.3506,
      "step": 104
    },
    {
      "epoch": 0.22764227642276422,
      "grad_norm": 47.262901306152344,
      "learning_rate": 5e-05,
      "loss": 9.0541,
      "step": 105
    },
    {
      "epoch": 0.22981029810298104,
      "grad_norm": 40.28150177001953,
      "learning_rate": 4.917330276168208e-05,
      "loss": 8.2982,
      "step": 106
    },
    {
      "epoch": 0.23197831978319783,
      "grad_norm": 46.546836853027344,
      "learning_rate": 4.834683153413459e-05,
      "loss": 8.5468,
      "step": 107
    },
    {
      "epoch": 0.23414634146341465,
      "grad_norm": 50.914913177490234,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 9.1811,
      "step": 108
    },
    {
      "epoch": 0.23631436314363144,
      "grad_norm": 46.76639938354492,
      "learning_rate": 4.669547078371504e-05,
      "loss": 9.2346,
      "step": 109
    },
    {
      "epoch": 0.23848238482384823,
      "grad_norm": 41.451637268066406,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 9.4973,
      "step": 110
    },
    {
      "epoch": 0.24065040650406505,
      "grad_norm": 43.00974655151367,
      "learning_rate": 4.504772348747687e-05,
      "loss": 9.1342,
      "step": 111
    },
    {
      "epoch": 0.24281842818428184,
      "grad_norm": 50.22190856933594,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 9.7044,
      "step": 112
    },
    {
      "epoch": 0.24498644986449863,
      "grad_norm": 51.052486419677734,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 8.4034,
      "step": 113
    },
    {
      "epoch": 0.24715447154471545,
      "grad_norm": 49.655677795410156,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 9.4296,
      "step": 114
    },
    {
      "epoch": 0.24932249322493225,
      "grad_norm": 48.18131637573242,
      "learning_rate": 4.17702704859633e-05,
      "loss": 8.6055,
      "step": 115
    },
    {
      "epoch": 0.25149051490514907,
      "grad_norm": 54.04130172729492,
      "learning_rate": 4.095597328339452e-05,
      "loss": 8.7869,
      "step": 116
    },
    {
      "epoch": 0.25365853658536586,
      "grad_norm": 42.525760650634766,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 8.5799,
      "step": 117
    },
    {
      "epoch": 0.25582655826558265,
      "grad_norm": 51.273563385009766,
      "learning_rate": 3.933501846281267e-05,
      "loss": 8.7128,
      "step": 118
    },
    {
      "epoch": 0.25799457994579944,
      "grad_norm": 54.508750915527344,
      "learning_rate": 3.852880399766243e-05,
      "loss": 9.1462,
      "step": 119
    },
    {
      "epoch": 0.2601626016260163,
      "grad_norm": 57.53263473510742,
      "learning_rate": 3.772572564296005e-05,
      "loss": 9.1607,
      "step": 120
    },
    {
      "epoch": 0.2623306233062331,
      "grad_norm": 54.50743865966797,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 9.35,
      "step": 121
    },
    {
      "epoch": 0.26449864498644987,
      "grad_norm": 46.642276763916016,
      "learning_rate": 3.612985456190778e-05,
      "loss": 9.2498,
      "step": 122
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 57.96341323852539,
      "learning_rate": 3.533749813077677e-05,
      "loss": 9.3575,
      "step": 123
    },
    {
      "epoch": 0.26883468834688345,
      "grad_norm": 54.722747802734375,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 9.1893,
      "step": 124
    },
    {
      "epoch": 0.27100271002710025,
      "grad_norm": 67.47887420654297,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 9.6908,
      "step": 125
    },
    {
      "epoch": 0.2731707317073171,
      "grad_norm": 77.53616333007812,
      "learning_rate": 3.298534127791785e-05,
      "loss": 9.5023,
      "step": 126
    },
    {
      "epoch": 0.2753387533875339,
      "grad_norm": 59.79452896118164,
      "learning_rate": 3.221030765387417e-05,
      "loss": 9.3832,
      "step": 127
    },
    {
      "epoch": 0.2775067750677507,
      "grad_norm": 77.19922637939453,
      "learning_rate": 3.144013755408895e-05,
      "loss": 9.4195,
      "step": 128
    },
    {
      "epoch": 0.27967479674796747,
      "grad_norm": 55.471431732177734,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 8.5907,
      "step": 129
    },
    {
      "epoch": 0.28184281842818426,
      "grad_norm": 93.79061889648438,
      "learning_rate": 2.991522876735154e-05,
      "loss": 9.2269,
      "step": 130
    },
    {
      "epoch": 0.2840108401084011,
      "grad_norm": 61.34229278564453,
      "learning_rate": 2.916090697523549e-05,
      "loss": 8.7878,
      "step": 131
    },
    {
      "epoch": 0.2861788617886179,
      "grad_norm": 59.42024230957031,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 8.5367,
      "step": 132
    },
    {
      "epoch": 0.2883468834688347,
      "grad_norm": 75.08689880371094,
      "learning_rate": 2.766955965735968e-05,
      "loss": 9.0569,
      "step": 133
    },
    {
      "epoch": 0.2905149051490515,
      "grad_norm": 62.514442443847656,
      "learning_rate": 2.693294185106562e-05,
      "loss": 8.2505,
      "step": 134
    },
    {
      "epoch": 0.2926829268292683,
      "grad_norm": 126.19690704345703,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 10.0156,
      "step": 135
    },
    {
      "epoch": 0.2948509485094851,
      "grad_norm": 72.40161895751953,
      "learning_rate": 2.547882480847461e-05,
      "loss": 9.924,
      "step": 136
    },
    {
      "epoch": 0.2970189701897019,
      "grad_norm": 120.88567352294922,
      "learning_rate": 2.476172311325783e-05,
      "loss": 7.6603,
      "step": 137
    },
    {
      "epoch": 0.2991869918699187,
      "grad_norm": 196.6604766845703,
      "learning_rate": 2.405152131093926e-05,
      "loss": 7.5255,
      "step": 138
    },
    {
      "epoch": 0.3013550135501355,
      "grad_norm": 116.19119262695312,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 8.4667,
      "step": 139
    },
    {
      "epoch": 0.3035230352303523,
      "grad_norm": 134.3529815673828,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 7.1047,
      "step": 140
    },
    {
      "epoch": 0.3056910569105691,
      "grad_norm": 119.5435791015625,
      "learning_rate": 2.196424713241637e-05,
      "loss": 3.4751,
      "step": 141
    },
    {
      "epoch": 0.30785907859078593,
      "grad_norm": 181.33877563476562,
      "learning_rate": 2.128356686585282e-05,
      "loss": 4.3411,
      "step": 142
    },
    {
      "epoch": 0.3100271002710027,
      "grad_norm": 103.76376342773438,
      "learning_rate": 2.061073738537635e-05,
      "loss": 2.6192,
      "step": 143
    },
    {
      "epoch": 0.3121951219512195,
      "grad_norm": 107.93944549560547,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 2.0389,
      "step": 144
    },
    {
      "epoch": 0.3143631436314363,
      "grad_norm": 97.89171600341797,
      "learning_rate": 1.928936436551661e-05,
      "loss": 1.7801,
      "step": 145
    },
    {
      "epoch": 0.3165311653116531,
      "grad_norm": 48.800628662109375,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 0.9013,
      "step": 146
    },
    {
      "epoch": 0.31869918699186994,
      "grad_norm": 82.11100769042969,
      "learning_rate": 1.800157297483417e-05,
      "loss": 1.0817,
      "step": 147
    },
    {
      "epoch": 0.32086720867208673,
      "grad_norm": 98.84628295898438,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 1.0696,
      "step": 148
    },
    {
      "epoch": 0.3230352303523035,
      "grad_norm": 73.2131576538086,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 0.8876,
      "step": 149
    },
    {
      "epoch": 0.3252032520325203,
      "grad_norm": 70.06466674804688,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 2.0163,
      "step": 150
    },
    {
      "epoch": 0.3252032520325203,
      "eval_loss": 1.8626056909561157,
      "eval_runtime": 9.2517,
      "eval_samples_per_second": 83.985,
      "eval_steps_per_second": 21.077,
      "step": 150
    },
    {
      "epoch": 0.3273712737127371,
      "grad_norm": 51.9277458190918,
      "learning_rate": 1.553232954407171e-05,
      "loss": 6.6969,
      "step": 151
    },
    {
      "epoch": 0.32953929539295396,
      "grad_norm": 75.45313262939453,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 10.0399,
      "step": 152
    },
    {
      "epoch": 0.33170731707317075,
      "grad_norm": 42.13958740234375,
      "learning_rate": 1.435357758543015e-05,
      "loss": 10.1013,
      "step": 153
    },
    {
      "epoch": 0.33387533875338754,
      "grad_norm": 53.16179656982422,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 9.7405,
      "step": 154
    },
    {
      "epoch": 0.33604336043360433,
      "grad_norm": 44.29389572143555,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 8.5341,
      "step": 155
    },
    {
      "epoch": 0.3382113821138211,
      "grad_norm": 36.9524040222168,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 8.7209,
      "step": 156
    },
    {
      "epoch": 0.3403794037940379,
      "grad_norm": 56.26239776611328,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 9.1412,
      "step": 157
    },
    {
      "epoch": 0.34254742547425476,
      "grad_norm": 53.07481002807617,
      "learning_rate": 1.157994445715706e-05,
      "loss": 8.5697,
      "step": 158
    },
    {
      "epoch": 0.34471544715447155,
      "grad_norm": 41.179908752441406,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 8.284,
      "step": 159
    },
    {
      "epoch": 0.34688346883468835,
      "grad_norm": 43.900657653808594,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 8.4032,
      "step": 160
    },
    {
      "epoch": 0.34905149051490514,
      "grad_norm": 40.793426513671875,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 8.4764,
      "step": 161
    },
    {
      "epoch": 0.35121951219512193,
      "grad_norm": 44.538753509521484,
      "learning_rate": 9.549150281252633e-06,
      "loss": 8.7613,
      "step": 162
    },
    {
      "epoch": 0.3533875338753388,
      "grad_norm": 42.16035461425781,
      "learning_rate": 9.068759265665384e-06,
      "loss": 9.2068,
      "step": 163
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 41.48743438720703,
      "learning_rate": 8.599558442598998e-06,
      "loss": 8.5854,
      "step": 164
    },
    {
      "epoch": 0.35772357723577236,
      "grad_norm": 46.81573486328125,
      "learning_rate": 8.141676086873572e-06,
      "loss": 8.6576,
      "step": 165
    },
    {
      "epoch": 0.35989159891598915,
      "grad_norm": 35.49210739135742,
      "learning_rate": 7.695237378953223e-06,
      "loss": 8.8878,
      "step": 166
    },
    {
      "epoch": 0.36205962059620594,
      "grad_norm": 46.1544189453125,
      "learning_rate": 7.260364370723044e-06,
      "loss": 8.4476,
      "step": 167
    },
    {
      "epoch": 0.3642276422764228,
      "grad_norm": 47.492454528808594,
      "learning_rate": 6.837175952121306e-06,
      "loss": 8.6836,
      "step": 168
    },
    {
      "epoch": 0.3663956639566396,
      "grad_norm": 44.350013732910156,
      "learning_rate": 6.425787818636131e-06,
      "loss": 8.242,
      "step": 169
    },
    {
      "epoch": 0.3685636856368564,
      "grad_norm": 42.295711517333984,
      "learning_rate": 6.026312439675552e-06,
      "loss": 9.4691,
      "step": 170
    },
    {
      "epoch": 0.37073170731707317,
      "grad_norm": 43.10215759277344,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 9.6269,
      "step": 171
    },
    {
      "epoch": 0.37289972899728996,
      "grad_norm": 54.865089416503906,
      "learning_rate": 5.263533508961827e-06,
      "loss": 9.5286,
      "step": 172
    },
    {
      "epoch": 0.37506775067750675,
      "grad_norm": 62.83253479003906,
      "learning_rate": 4.900438493352055e-06,
      "loss": 10.1959,
      "step": 173
    },
    {
      "epoch": 0.3772357723577236,
      "grad_norm": 50.127498626708984,
      "learning_rate": 4.549673247541875e-06,
      "loss": 8.9541,
      "step": 174
    },
    {
      "epoch": 0.3794037940379404,
      "grad_norm": 48.99895477294922,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 9.45,
      "step": 175
    },
    {
      "epoch": 0.3815718157181572,
      "grad_norm": 40.754093170166016,
      "learning_rate": 3.885512251130763e-06,
      "loss": 9.3567,
      "step": 176
    },
    {
      "epoch": 0.383739837398374,
      "grad_norm": 55.58876037597656,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 9.8759,
      "step": 177
    },
    {
      "epoch": 0.38590785907859076,
      "grad_norm": 68.02944946289062,
      "learning_rate": 3.271776770026963e-06,
      "loss": 9.0546,
      "step": 178
    },
    {
      "epoch": 0.3880758807588076,
      "grad_norm": 66.05428314208984,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 10.0226,
      "step": 179
    },
    {
      "epoch": 0.3902439024390244,
      "grad_norm": 44.68964385986328,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 8.6251,
      "step": 180
    },
    {
      "epoch": 0.3924119241192412,
      "grad_norm": 69.2607421875,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 9.5753,
      "step": 181
    },
    {
      "epoch": 0.394579945799458,
      "grad_norm": 52.203182220458984,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 8.05,
      "step": 182
    },
    {
      "epoch": 0.3967479674796748,
      "grad_norm": 46.09951400756836,
      "learning_rate": 1.962316193157593e-06,
      "loss": 8.4656,
      "step": 183
    },
    {
      "epoch": 0.3989159891598916,
      "grad_norm": 69.3664779663086,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 9.1121,
      "step": 184
    },
    {
      "epoch": 0.4010840108401084,
      "grad_norm": 115.72722625732422,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 9.7761,
      "step": 185
    },
    {
      "epoch": 0.4032520325203252,
      "grad_norm": 89.68798065185547,
      "learning_rate": 1.333670137599713e-06,
      "loss": 10.7616,
      "step": 186
    },
    {
      "epoch": 0.405420054200542,
      "grad_norm": 81.19366455078125,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 9.3529,
      "step": 187
    },
    {
      "epoch": 0.4075880758807588,
      "grad_norm": 72.34278869628906,
      "learning_rate": 9.810017062595322e-07,
      "loss": 5.9872,
      "step": 188
    },
    {
      "epoch": 0.4097560975609756,
      "grad_norm": 141.31683349609375,
      "learning_rate": 8.247462563808817e-07,
      "loss": 5.523,
      "step": 189
    },
    {
      "epoch": 0.41192411924119243,
      "grad_norm": 176.84629821777344,
      "learning_rate": 6.819348298638839e-07,
      "loss": 7.7485,
      "step": 190
    },
    {
      "epoch": 0.4140921409214092,
      "grad_norm": 175.30377197265625,
      "learning_rate": 5.526064699265753e-07,
      "loss": 3.3017,
      "step": 191
    },
    {
      "epoch": 0.416260162601626,
      "grad_norm": 240.5296630859375,
      "learning_rate": 4.367965336512403e-07,
      "loss": 2.4354,
      "step": 192
    },
    {
      "epoch": 0.4184281842818428,
      "grad_norm": 304.1140441894531,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 3.5376,
      "step": 193
    },
    {
      "epoch": 0.4205962059620596,
      "grad_norm": 125.94408416748047,
      "learning_rate": 2.458548727494292e-07,
      "loss": 2.3871,
      "step": 194
    },
    {
      "epoch": 0.42276422764227645,
      "grad_norm": 94.66616821289062,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 1.0802,
      "step": 195
    },
    {
      "epoch": 0.42493224932249324,
      "grad_norm": 178.35890197753906,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 1.3666,
      "step": 196
    },
    {
      "epoch": 0.42710027100271003,
      "grad_norm": 156.51934814453125,
      "learning_rate": 6.150154258476315e-08,
      "loss": 1.3191,
      "step": 197
    },
    {
      "epoch": 0.4292682926829268,
      "grad_norm": 129.61354064941406,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 1.3543,
      "step": 198
    },
    {
      "epoch": 0.4314363143631436,
      "grad_norm": 101.63610076904297,
      "learning_rate": 6.834750376549792e-09,
      "loss": 0.8991,
      "step": 199
    },
    {
      "epoch": 0.43360433604336046,
      "grad_norm": 94.34125518798828,
      "learning_rate": 0.0,
      "loss": 2.7274,
      "step": 200
    },
    {
      "epoch": 0.43360433604336046,
      "eval_loss": 1.8256925344467163,
      "eval_runtime": 9.1784,
      "eval_samples_per_second": 84.655,
      "eval_steps_per_second": 21.246,
      "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": 1.5703441863081984e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}