{
  "best_metric": 0.7749578952789307,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.1832340815391663,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0009161704076958314,
      "grad_norm": 10.420132637023926,
      "learning_rate": 1e-05,
      "loss": 3.9341,
      "step": 1
    },
    {
      "epoch": 0.0009161704076958314,
      "eval_loss": 1.0264543294906616,
      "eval_runtime": 141.3224,
      "eval_samples_per_second": 13.013,
      "eval_steps_per_second": 3.255,
      "step": 1
    },
    {
      "epoch": 0.0018323408153916628,
      "grad_norm": 11.907942771911621,
      "learning_rate": 2e-05,
      "loss": 4.0648,
      "step": 2
    },
    {
      "epoch": 0.002748511223087494,
      "grad_norm": 7.917791843414307,
      "learning_rate": 3e-05,
      "loss": 3.9942,
      "step": 3
    },
    {
      "epoch": 0.0036646816307833257,
      "grad_norm": 6.13906717300415,
      "learning_rate": 4e-05,
      "loss": 3.9128,
      "step": 4
    },
    {
      "epoch": 0.004580852038479157,
      "grad_norm": 7.859758377075195,
      "learning_rate": 5e-05,
      "loss": 3.7754,
      "step": 5
    },
    {
      "epoch": 0.005497022446174988,
      "grad_norm": 6.809757232666016,
      "learning_rate": 6e-05,
      "loss": 3.8282,
      "step": 6
    },
    {
      "epoch": 0.00641319285387082,
      "grad_norm": 5.6679277420043945,
      "learning_rate": 7e-05,
      "loss": 3.7432,
      "step": 7
    },
    {
      "epoch": 0.007329363261566651,
      "grad_norm": 5.149811267852783,
      "learning_rate": 8e-05,
      "loss": 3.5363,
      "step": 8
    },
    {
      "epoch": 0.008245533669262483,
      "grad_norm": 4.5849080085754395,
      "learning_rate": 9e-05,
      "loss": 3.5204,
      "step": 9
    },
    {
      "epoch": 0.009161704076958314,
      "grad_norm": 4.159773826599121,
      "learning_rate": 0.0001,
      "loss": 3.4225,
      "step": 10
    },
    {
      "epoch": 0.010077874484654145,
      "grad_norm": 3.5201876163482666,
      "learning_rate": 9.999316524962345e-05,
      "loss": 3.4388,
      "step": 11
    },
    {
      "epoch": 0.010994044892349977,
      "grad_norm": 3.9413065910339355,
      "learning_rate": 9.997266286704631e-05,
      "loss": 3.3177,
      "step": 12
    },
    {
      "epoch": 0.011910215300045808,
      "grad_norm": 3.331254482269287,
      "learning_rate": 9.993849845741524e-05,
      "loss": 3.4013,
      "step": 13
    },
    {
      "epoch": 0.01282638570774164,
      "grad_norm": 3.327693223953247,
      "learning_rate": 9.989068136093873e-05,
      "loss": 3.3396,
      "step": 14
    },
    {
      "epoch": 0.013742556115437472,
      "grad_norm": 3.3501760959625244,
      "learning_rate": 9.98292246503335e-05,
      "loss": 3.2285,
      "step": 15
    },
    {
      "epoch": 0.014658726523133303,
      "grad_norm": 3.223546028137207,
      "learning_rate": 9.975414512725057e-05,
      "loss": 3.3571,
      "step": 16
    },
    {
      "epoch": 0.015574896930829134,
      "grad_norm": 3.100618600845337,
      "learning_rate": 9.966546331768191e-05,
      "loss": 3.2902,
      "step": 17
    },
    {
      "epoch": 0.016491067338524967,
      "grad_norm": 3.023641586303711,
      "learning_rate": 9.956320346634876e-05,
      "loss": 3.405,
      "step": 18
    },
    {
      "epoch": 0.017407237746220796,
      "grad_norm": 2.963796615600586,
      "learning_rate": 9.944739353007344e-05,
      "loss": 3.2498,
      "step": 19
    },
    {
      "epoch": 0.01832340815391663,
      "grad_norm": 3.0240752696990967,
      "learning_rate": 9.931806517013612e-05,
      "loss": 3.3706,
      "step": 20
    },
    {
      "epoch": 0.01923957856161246,
      "grad_norm": 2.942776679992676,
      "learning_rate": 9.917525374361912e-05,
      "loss": 3.1782,
      "step": 21
    },
    {
      "epoch": 0.02015574896930829,
      "grad_norm": 2.983973264694214,
      "learning_rate": 9.901899829374047e-05,
      "loss": 3.352,
      "step": 22
    },
    {
      "epoch": 0.021071919377004124,
      "grad_norm": 3.050541877746582,
      "learning_rate": 9.884934153917997e-05,
      "loss": 3.26,
      "step": 23
    },
    {
      "epoch": 0.021988089784699953,
      "grad_norm": 3.0281982421875,
      "learning_rate": 9.86663298624003e-05,
      "loss": 3.3342,
      "step": 24
    },
    {
      "epoch": 0.022904260192395786,
      "grad_norm": 2.8630337715148926,
      "learning_rate": 9.847001329696653e-05,
      "loss": 3.4261,
      "step": 25
    },
    {
      "epoch": 0.023820430600091615,
      "grad_norm": 2.897817611694336,
      "learning_rate": 9.826044551386744e-05,
      "loss": 3.3987,
      "step": 26
    },
    {
      "epoch": 0.024736601007787448,
      "grad_norm": 2.754214286804199,
      "learning_rate": 9.803768380684242e-05,
      "loss": 3.2664,
      "step": 27
    },
    {
      "epoch": 0.02565277141548328,
      "grad_norm": 2.7787299156188965,
      "learning_rate": 9.780178907671789e-05,
      "loss": 3.2235,
      "step": 28
    },
    {
      "epoch": 0.02656894182317911,
      "grad_norm": 2.800663948059082,
      "learning_rate": 9.755282581475769e-05,
      "loss": 3.2619,
      "step": 29
    },
    {
      "epoch": 0.027485112230874943,
      "grad_norm": 2.7739968299865723,
      "learning_rate": 9.729086208503174e-05,
      "loss": 3.2289,
      "step": 30
    },
    {
      "epoch": 0.028401282638570773,
      "grad_norm": 2.771052598953247,
      "learning_rate": 9.701596950580806e-05,
      "loss": 3.2396,
      "step": 31
    },
    {
      "epoch": 0.029317453046266605,
      "grad_norm": 2.7918596267700195,
      "learning_rate": 9.672822322997305e-05,
      "loss": 3.3144,
      "step": 32
    },
    {
      "epoch": 0.030233623453962438,
      "grad_norm": 2.72592830657959,
      "learning_rate": 9.642770192448536e-05,
      "loss": 3.2155,
      "step": 33
    },
    {
      "epoch": 0.031149793861658268,
      "grad_norm": 2.864677906036377,
      "learning_rate": 9.611448774886924e-05,
      "loss": 3.2692,
      "step": 34
    },
    {
      "epoch": 0.0320659642693541,
      "grad_norm": 2.894106149673462,
      "learning_rate": 9.578866633275288e-05,
      "loss": 3.3678,
      "step": 35
    },
    {
      "epoch": 0.03298213467704993,
      "grad_norm": 2.9205408096313477,
      "learning_rate": 9.545032675245813e-05,
      "loss": 3.2992,
      "step": 36
    },
    {
      "epoch": 0.03389830508474576,
      "grad_norm": 2.8021671772003174,
      "learning_rate": 9.509956150664796e-05,
      "loss": 3.3055,
      "step": 37
    },
    {
      "epoch": 0.03481447549244159,
      "grad_norm": 2.945963144302368,
      "learning_rate": 9.473646649103818e-05,
      "loss": 3.3908,
      "step": 38
    },
    {
      "epoch": 0.03573064590013743,
      "grad_norm": 2.7552454471588135,
      "learning_rate": 9.43611409721806e-05,
      "loss": 3.1012,
      "step": 39
    },
    {
      "epoch": 0.03664681630783326,
      "grad_norm": 2.993711471557617,
      "learning_rate": 9.397368756032445e-05,
      "loss": 3.1608,
      "step": 40
    },
    {
      "epoch": 0.03756298671552909,
      "grad_norm": 2.9157307147979736,
      "learning_rate": 9.357421218136386e-05,
      "loss": 3.1934,
      "step": 41
    },
    {
      "epoch": 0.03847915712322492,
      "grad_norm": 2.9856550693511963,
      "learning_rate": 9.316282404787871e-05,
      "loss": 3.271,
      "step": 42
    },
    {
      "epoch": 0.03939532753092075,
      "grad_norm": 3.02144455909729,
      "learning_rate": 9.273963562927695e-05,
      "loss": 3.1494,
      "step": 43
    },
    {
      "epoch": 0.04031149793861658,
      "grad_norm": 3.14676570892334,
      "learning_rate": 9.230476262104677e-05,
      "loss": 3.1785,
      "step": 44
    },
    {
      "epoch": 0.04122766834631241,
      "grad_norm": 2.8963444232940674,
      "learning_rate": 9.185832391312644e-05,
      "loss": 3.2633,
      "step": 45
    },
    {
      "epoch": 0.04214383875400825,
      "grad_norm": 3.0370452404022217,
      "learning_rate": 9.140044155740101e-05,
      "loss": 3.4015,
      "step": 46
    },
    {
      "epoch": 0.04306000916170408,
      "grad_norm": 3.0973758697509766,
      "learning_rate": 9.093124073433463e-05,
      "loss": 3.1235,
      "step": 47
    },
    {
      "epoch": 0.043976179569399906,
      "grad_norm": 3.1558079719543457,
      "learning_rate": 9.045084971874738e-05,
      "loss": 3.1534,
      "step": 48
    },
    {
      "epoch": 0.04489234997709574,
      "grad_norm": 3.6855924129486084,
      "learning_rate": 8.995939984474624e-05,
      "loss": 3.0983,
      "step": 49
    },
    {
      "epoch": 0.04580852038479157,
      "grad_norm": 4.269506454467773,
      "learning_rate": 8.945702546981969e-05,
      "loss": 2.8739,
      "step": 50
    },
    {
      "epoch": 0.04580852038479157,
      "eval_loss": 0.857905387878418,
      "eval_runtime": 143.2493,
      "eval_samples_per_second": 12.838,
      "eval_steps_per_second": 3.211,
      "step": 50
    },
    {
      "epoch": 0.0467246907924874,
      "grad_norm": 4.632282257080078,
      "learning_rate": 8.894386393810563e-05,
      "loss": 3.3263,
      "step": 51
    },
    {
      "epoch": 0.04764086120018323,
      "grad_norm": 3.454362154006958,
      "learning_rate": 8.842005554284296e-05,
      "loss": 3.4455,
      "step": 52
    },
    {
      "epoch": 0.04855703160787907,
      "grad_norm": 2.2924325466156006,
      "learning_rate": 8.788574348801675e-05,
      "loss": 3.1865,
      "step": 53
    },
    {
      "epoch": 0.049473202015574896,
      "grad_norm": 2.3359978199005127,
      "learning_rate": 8.73410738492077e-05,
      "loss": 3.2213,
      "step": 54
    },
    {
      "epoch": 0.050389372423270726,
      "grad_norm": 2.6054296493530273,
      "learning_rate": 8.678619553365659e-05,
      "loss": 3.4569,
      "step": 55
    },
    {
      "epoch": 0.05130554283096656,
      "grad_norm": 2.411471366882324,
      "learning_rate": 8.622126023955446e-05,
      "loss": 3.4058,
      "step": 56
    },
    {
      "epoch": 0.05222171323866239,
      "grad_norm": 2.5202689170837402,
      "learning_rate": 8.564642241456986e-05,
      "loss": 3.1101,
      "step": 57
    },
    {
      "epoch": 0.05313788364635822,
      "grad_norm": 2.3162035942077637,
      "learning_rate": 8.506183921362443e-05,
      "loss": 3.2279,
      "step": 58
    },
    {
      "epoch": 0.05405405405405406,
      "grad_norm": 2.292923927307129,
      "learning_rate": 8.44676704559283e-05,
      "loss": 3.2233,
      "step": 59
    },
    {
      "epoch": 0.054970224461749886,
      "grad_norm": 2.277437925338745,
      "learning_rate": 8.386407858128706e-05,
      "loss": 3.2655,
      "step": 60
    },
    {
      "epoch": 0.055886394869445716,
      "grad_norm": 2.4472885131835938,
      "learning_rate": 8.32512286056924e-05,
      "loss": 3.356,
      "step": 61
    },
    {
      "epoch": 0.056802565277141545,
      "grad_norm": 2.4935200214385986,
      "learning_rate": 8.262928807620843e-05,
      "loss": 3.3415,
      "step": 62
    },
    {
      "epoch": 0.05771873568483738,
      "grad_norm": 2.3300907611846924,
      "learning_rate": 8.199842702516583e-05,
      "loss": 3.2716,
      "step": 63
    },
    {
      "epoch": 0.05863490609253321,
      "grad_norm": 2.2986490726470947,
      "learning_rate": 8.135881792367686e-05,
      "loss": 3.2727,
      "step": 64
    },
    {
      "epoch": 0.05955107650022904,
      "grad_norm": 2.3526642322540283,
      "learning_rate": 8.07106356344834e-05,
      "loss": 3.2887,
      "step": 65
    },
    {
      "epoch": 0.060467246907924876,
      "grad_norm": 2.36346697807312,
      "learning_rate": 8.005405736415126e-05,
      "loss": 3.1683,
      "step": 66
    },
    {
      "epoch": 0.061383417315620706,
      "grad_norm": 2.218932628631592,
      "learning_rate": 7.938926261462366e-05,
      "loss": 3.3704,
      "step": 67
    },
    {
      "epoch": 0.062299587723316535,
      "grad_norm": 2.328066110610962,
      "learning_rate": 7.871643313414718e-05,
      "loss": 3.2611,
      "step": 68
    },
    {
      "epoch": 0.06321575813101236,
      "grad_norm": 2.4511618614196777,
      "learning_rate": 7.803575286758364e-05,
      "loss": 3.2324,
      "step": 69
    },
    {
      "epoch": 0.0641319285387082,
      "grad_norm": 2.322838544845581,
      "learning_rate": 7.734740790612136e-05,
      "loss": 3.1691,
      "step": 70
    },
    {
      "epoch": 0.06504809894640404,
      "grad_norm": 2.294978380203247,
      "learning_rate": 7.66515864363997e-05,
      "loss": 3.0479,
      "step": 71
    },
    {
      "epoch": 0.06596426935409987,
      "grad_norm": 2.4404642581939697,
      "learning_rate": 7.594847868906076e-05,
      "loss": 3.1572,
      "step": 72
    },
    {
      "epoch": 0.0668804397617957,
      "grad_norm": 2.3828468322753906,
      "learning_rate": 7.52382768867422e-05,
      "loss": 3.0854,
      "step": 73
    },
    {
      "epoch": 0.06779661016949153,
      "grad_norm": 2.5379884243011475,
      "learning_rate": 7.452117519152542e-05,
      "loss": 3.3354,
      "step": 74
    },
    {
      "epoch": 0.06871278057718735,
      "grad_norm": 2.5061211585998535,
      "learning_rate": 7.379736965185368e-05,
      "loss": 3.3212,
      "step": 75
    },
    {
      "epoch": 0.06962895098488318,
      "grad_norm": 2.3468308448791504,
      "learning_rate": 7.30670581489344e-05,
      "loss": 3.2781,
      "step": 76
    },
    {
      "epoch": 0.07054512139257901,
      "grad_norm": 2.3657217025756836,
      "learning_rate": 7.233044034264034e-05,
      "loss": 3.2078,
      "step": 77
    },
    {
      "epoch": 0.07146129180027486,
      "grad_norm": 2.379638671875,
      "learning_rate": 7.158771761692464e-05,
      "loss": 3.1101,
      "step": 78
    },
    {
      "epoch": 0.07237746220797069,
      "grad_norm": 2.3256547451019287,
      "learning_rate": 7.083909302476453e-05,
      "loss": 3.1999,
      "step": 79
    },
    {
      "epoch": 0.07329363261566652,
      "grad_norm": 2.40804123878479,
      "learning_rate": 7.008477123264848e-05,
      "loss": 3.2508,
      "step": 80
    },
    {
      "epoch": 0.07420980302336234,
      "grad_norm": 2.463440418243408,
      "learning_rate": 6.932495846462261e-05,
      "loss": 3.1665,
      "step": 81
    },
    {
      "epoch": 0.07512597343105817,
      "grad_norm": 2.5652084350585938,
      "learning_rate": 6.855986244591104e-05,
      "loss": 3.2652,
      "step": 82
    },
    {
      "epoch": 0.076042143838754,
      "grad_norm": 2.472507953643799,
      "learning_rate": 6.778969234612584e-05,
      "loss": 3.2888,
      "step": 83
    },
    {
      "epoch": 0.07695831424644985,
      "grad_norm": 2.509007453918457,
      "learning_rate": 6.701465872208216e-05,
      "loss": 3.2122,
      "step": 84
    },
    {
      "epoch": 0.07787448465414568,
      "grad_norm": 2.4399054050445557,
      "learning_rate": 6.623497346023418e-05,
      "loss": 3.1636,
      "step": 85
    },
    {
      "epoch": 0.0787906550618415,
      "grad_norm": 2.5819153785705566,
      "learning_rate": 6.545084971874738e-05,
      "loss": 3.356,
      "step": 86
    },
    {
      "epoch": 0.07970682546953733,
      "grad_norm": 2.5383763313293457,
      "learning_rate": 6.466250186922325e-05,
      "loss": 3.2435,
      "step": 87
    },
    {
      "epoch": 0.08062299587723316,
      "grad_norm": 2.4706544876098633,
      "learning_rate": 6.387014543809223e-05,
      "loss": 3.346,
      "step": 88
    },
    {
      "epoch": 0.081539166284929,
      "grad_norm": 2.4567182064056396,
      "learning_rate": 6.307399704769099e-05,
      "loss": 3.2462,
      "step": 89
    },
    {
      "epoch": 0.08245533669262482,
      "grad_norm": 2.5329692363739014,
      "learning_rate": 6.227427435703997e-05,
      "loss": 3.1573,
      "step": 90
    },
    {
      "epoch": 0.08337150710032067,
      "grad_norm": 2.5082852840423584,
      "learning_rate": 6.147119600233758e-05,
      "loss": 3.1434,
      "step": 91
    },
    {
      "epoch": 0.0842876775080165,
      "grad_norm": 2.5499184131622314,
      "learning_rate": 6.066498153718735e-05,
      "loss": 3.1346,
      "step": 92
    },
    {
      "epoch": 0.08520384791571232,
      "grad_norm": 2.5437309741973877,
      "learning_rate": 5.985585137257401e-05,
      "loss": 3.1203,
      "step": 93
    },
    {
      "epoch": 0.08612001832340815,
      "grad_norm": 2.560912609100342,
      "learning_rate": 5.90440267166055e-05,
      "loss": 2.9664,
      "step": 94
    },
    {
      "epoch": 0.08703618873110398,
      "grad_norm": 2.5366506576538086,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 2.9797,
      "step": 95
    },
    {
      "epoch": 0.08795235913879981,
      "grad_norm": 2.702955722808838,
      "learning_rate": 5.74131823855921e-05,
      "loss": 3.1788,
      "step": 96
    },
    {
      "epoch": 0.08886852954649564,
      "grad_norm": 2.691070795059204,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 3.0511,
      "step": 97
    },
    {
      "epoch": 0.08978469995419149,
      "grad_norm": 2.8578526973724365,
      "learning_rate": 5.577423184847932e-05,
      "loss": 3.0515,
      "step": 98
    },
    {
      "epoch": 0.09070087036188731,
      "grad_norm": 3.083177089691162,
      "learning_rate": 5.495227651252315e-05,
      "loss": 3.1002,
      "step": 99
    },
    {
      "epoch": 0.09161704076958314,
      "grad_norm": 3.706963300704956,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 3.1031,
      "step": 100
    },
    {
      "epoch": 0.09161704076958314,
      "eval_loss": 0.8161656856536865,
      "eval_runtime": 143.009,
      "eval_samples_per_second": 12.859,
      "eval_steps_per_second": 3.217,
      "step": 100
    },
    {
      "epoch": 0.09253321117727897,
      "grad_norm": 3.368330478668213,
      "learning_rate": 5.330452921628497e-05,
      "loss": 3.0879,
      "step": 101
    },
    {
      "epoch": 0.0934493815849748,
      "grad_norm": 2.6313602924346924,
      "learning_rate": 5.247918773366112e-05,
      "loss": 3.3085,
      "step": 102
    },
    {
      "epoch": 0.09436555199267063,
      "grad_norm": 2.066396713256836,
      "learning_rate": 5.165316846586541e-05,
      "loss": 3.0479,
      "step": 103
    },
    {
      "epoch": 0.09528172240036646,
      "grad_norm": 1.9643681049346924,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 3.0732,
      "step": 104
    },
    {
      "epoch": 0.0961978928080623,
      "grad_norm": 2.020611047744751,
      "learning_rate": 5e-05,
      "loss": 3.2907,
      "step": 105
    },
    {
      "epoch": 0.09711406321575813,
      "grad_norm": 2.078235387802124,
      "learning_rate": 4.917330276168208e-05,
      "loss": 3.3846,
      "step": 106
    },
    {
      "epoch": 0.09803023362345396,
      "grad_norm": 2.1009562015533447,
      "learning_rate": 4.834683153413459e-05,
      "loss": 3.1361,
      "step": 107
    },
    {
      "epoch": 0.09894640403114979,
      "grad_norm": 2.0752336978912354,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 3.2305,
      "step": 108
    },
    {
      "epoch": 0.09986257443884562,
      "grad_norm": 2.054835557937622,
      "learning_rate": 4.669547078371504e-05,
      "loss": 3.0906,
      "step": 109
    },
    {
      "epoch": 0.10077874484654145,
      "grad_norm": 2.1959903240203857,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 3.1616,
      "step": 110
    },
    {
      "epoch": 0.1016949152542373,
      "grad_norm": 2.148083209991455,
      "learning_rate": 4.504772348747687e-05,
      "loss": 3.2299,
      "step": 111
    },
    {
      "epoch": 0.10261108566193312,
      "grad_norm": 2.0367064476013184,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 3.1561,
      "step": 112
    },
    {
      "epoch": 0.10352725606962895,
      "grad_norm": 2.0471811294555664,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 3.121,
      "step": 113
    },
    {
      "epoch": 0.10444342647732478,
      "grad_norm": 2.1069531440734863,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 3.2607,
      "step": 114
    },
    {
      "epoch": 0.10535959688502061,
      "grad_norm": 2.12760329246521,
      "learning_rate": 4.17702704859633e-05,
      "loss": 3.3148,
      "step": 115
    },
    {
      "epoch": 0.10627576729271644,
      "grad_norm": 2.1181223392486572,
      "learning_rate": 4.095597328339452e-05,
      "loss": 3.1579,
      "step": 116
    },
    {
      "epoch": 0.10719193770041227,
      "grad_norm": 2.195899486541748,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 3.2684,
      "step": 117
    },
    {
      "epoch": 0.10810810810810811,
      "grad_norm": 2.182040214538574,
      "learning_rate": 3.933501846281267e-05,
      "loss": 3.2551,
      "step": 118
    },
    {
      "epoch": 0.10902427851580394,
      "grad_norm": 2.209686279296875,
      "learning_rate": 3.852880399766243e-05,
      "loss": 3.1189,
      "step": 119
    },
    {
      "epoch": 0.10994044892349977,
      "grad_norm": 2.234920024871826,
      "learning_rate": 3.772572564296005e-05,
      "loss": 3.2586,
      "step": 120
    },
    {
      "epoch": 0.1108566193311956,
      "grad_norm": 2.41520619392395,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 3.1355,
      "step": 121
    },
    {
      "epoch": 0.11177278973889143,
      "grad_norm": 2.281057596206665,
      "learning_rate": 3.612985456190778e-05,
      "loss": 3.1295,
      "step": 122
    },
    {
      "epoch": 0.11268896014658726,
      "grad_norm": 2.2140228748321533,
      "learning_rate": 3.533749813077677e-05,
      "loss": 3.1915,
      "step": 123
    },
    {
      "epoch": 0.11360513055428309,
      "grad_norm": 2.25614070892334,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 3.2605,
      "step": 124
    },
    {
      "epoch": 0.11452130096197893,
      "grad_norm": 2.2511088848114014,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 3.1568,
      "step": 125
    },
    {
      "epoch": 0.11543747136967476,
      "grad_norm": 2.2680823802948,
      "learning_rate": 3.298534127791785e-05,
      "loss": 3.2275,
      "step": 126
    },
    {
      "epoch": 0.11635364177737059,
      "grad_norm": 2.2072861194610596,
      "learning_rate": 3.221030765387417e-05,
      "loss": 2.9066,
      "step": 127
    },
    {
      "epoch": 0.11726981218506642,
      "grad_norm": 2.1895344257354736,
      "learning_rate": 3.144013755408895e-05,
      "loss": 3.1004,
      "step": 128
    },
    {
      "epoch": 0.11818598259276225,
      "grad_norm": 2.2173779010772705,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 3.0738,
      "step": 129
    },
    {
      "epoch": 0.11910215300045808,
      "grad_norm": 2.416811227798462,
      "learning_rate": 2.991522876735154e-05,
      "loss": 3.1071,
      "step": 130
    },
    {
      "epoch": 0.12001832340815392,
      "grad_norm": 2.3208141326904297,
      "learning_rate": 2.916090697523549e-05,
      "loss": 3.214,
      "step": 131
    },
    {
      "epoch": 0.12093449381584975,
      "grad_norm": 2.2636945247650146,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 3.1249,
      "step": 132
    },
    {
      "epoch": 0.12185066422354558,
      "grad_norm": 2.357179880142212,
      "learning_rate": 2.766955965735968e-05,
      "loss": 3.0948,
      "step": 133
    },
    {
      "epoch": 0.12276683463124141,
      "grad_norm": 2.3882765769958496,
      "learning_rate": 2.693294185106562e-05,
      "loss": 3.0909,
      "step": 134
    },
    {
      "epoch": 0.12368300503893724,
      "grad_norm": 2.405975580215454,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 3.1359,
      "step": 135
    },
    {
      "epoch": 0.12459917544663307,
      "grad_norm": 2.392596960067749,
      "learning_rate": 2.547882480847461e-05,
      "loss": 3.191,
      "step": 136
    },
    {
      "epoch": 0.1255153458543289,
      "grad_norm": 2.443600654602051,
      "learning_rate": 2.476172311325783e-05,
      "loss": 3.2062,
      "step": 137
    },
    {
      "epoch": 0.12643151626202473,
      "grad_norm": 2.3901736736297607,
      "learning_rate": 2.405152131093926e-05,
      "loss": 3.0655,
      "step": 138
    },
    {
      "epoch": 0.12734768666972057,
      "grad_norm": 2.3279871940612793,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 2.9941,
      "step": 139
    },
    {
      "epoch": 0.1282638570774164,
      "grad_norm": 2.3471009731292725,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 3.0925,
      "step": 140
    },
    {
      "epoch": 0.12918002748511223,
      "grad_norm": 2.407135009765625,
      "learning_rate": 2.196424713241637e-05,
      "loss": 3.0499,
      "step": 141
    },
    {
      "epoch": 0.13009619789280807,
      "grad_norm": 2.432713270187378,
      "learning_rate": 2.128356686585282e-05,
      "loss": 3.0151,
      "step": 142
    },
    {
      "epoch": 0.1310123683005039,
      "grad_norm": 2.4331960678100586,
      "learning_rate": 2.061073738537635e-05,
      "loss": 3.0783,
      "step": 143
    },
    {
      "epoch": 0.13192853870819973,
      "grad_norm": 2.552966833114624,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 3.0234,
      "step": 144
    },
    {
      "epoch": 0.13284470911589555,
      "grad_norm": 2.4917969703674316,
      "learning_rate": 1.928936436551661e-05,
      "loss": 3.1001,
      "step": 145
    },
    {
      "epoch": 0.1337608795235914,
      "grad_norm": 2.5499484539031982,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 2.9641,
      "step": 146
    },
    {
      "epoch": 0.1346770499312872,
      "grad_norm": 2.6731204986572266,
      "learning_rate": 1.800157297483417e-05,
      "loss": 3.1391,
      "step": 147
    },
    {
      "epoch": 0.13559322033898305,
      "grad_norm": 2.690629243850708,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 3.0358,
      "step": 148
    },
    {
      "epoch": 0.1365093907466789,
      "grad_norm": 3.047454833984375,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 3.0926,
      "step": 149
    },
    {
      "epoch": 0.1374255611543747,
      "grad_norm": 3.6542186737060547,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 2.7824,
      "step": 150
    },
    {
      "epoch": 0.1374255611543747,
      "eval_loss": 0.7817953824996948,
      "eval_runtime": 142.9742,
      "eval_samples_per_second": 12.862,
      "eval_steps_per_second": 3.217,
      "step": 150
    },
    {
      "epoch": 0.13834173156207055,
      "grad_norm": 1.9468052387237549,
      "learning_rate": 1.553232954407171e-05,
      "loss": 2.8115,
      "step": 151
    },
    {
      "epoch": 0.13925790196976637,
      "grad_norm": 2.077725410461426,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 3.0368,
      "step": 152
    },
    {
      "epoch": 0.1401740723774622,
      "grad_norm": 2.08034610748291,
      "learning_rate": 1.435357758543015e-05,
      "loss": 3.1402,
      "step": 153
    },
    {
      "epoch": 0.14109024278515803,
      "grad_norm": 2.0894968509674072,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 3.292,
      "step": 154
    },
    {
      "epoch": 0.14200641319285387,
      "grad_norm": 2.0221424102783203,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 3.1796,
      "step": 155
    },
    {
      "epoch": 0.1429225836005497,
      "grad_norm": 1.996311902999878,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 3.2408,
      "step": 156
    },
    {
      "epoch": 0.14383875400824553,
      "grad_norm": 1.9280788898468018,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 3.1141,
      "step": 157
    },
    {
      "epoch": 0.14475492441594137,
      "grad_norm": 1.931397557258606,
      "learning_rate": 1.157994445715706e-05,
      "loss": 3.0242,
      "step": 158
    },
    {
      "epoch": 0.1456710948236372,
      "grad_norm": 1.927922010421753,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 3.1927,
      "step": 159
    },
    {
      "epoch": 0.14658726523133303,
      "grad_norm": 2.027775287628174,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 3.0717,
      "step": 160
    },
    {
      "epoch": 0.14750343563902885,
      "grad_norm": 2.054180383682251,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 3.2456,
      "step": 161
    },
    {
      "epoch": 0.1484196060467247,
      "grad_norm": 2.0460307598114014,
      "learning_rate": 9.549150281252633e-06,
      "loss": 3.1453,
      "step": 162
    },
    {
      "epoch": 0.14933577645442053,
      "grad_norm": 2.093560218811035,
      "learning_rate": 9.068759265665384e-06,
      "loss": 3.1613,
      "step": 163
    },
    {
      "epoch": 0.15025194686211635,
      "grad_norm": 2.083407163619995,
      "learning_rate": 8.599558442598998e-06,
      "loss": 3.0278,
      "step": 164
    },
    {
      "epoch": 0.1511681172698122,
      "grad_norm": 2.0652873516082764,
      "learning_rate": 8.141676086873572e-06,
      "loss": 3.0578,
      "step": 165
    },
    {
      "epoch": 0.152084287677508,
      "grad_norm": 2.1485202312469482,
      "learning_rate": 7.695237378953223e-06,
      "loss": 3.0824,
      "step": 166
    },
    {
      "epoch": 0.15300045808520385,
      "grad_norm": 2.159834861755371,
      "learning_rate": 7.260364370723044e-06,
      "loss": 3.0621,
      "step": 167
    },
    {
      "epoch": 0.1539166284928997,
      "grad_norm": 2.172588348388672,
      "learning_rate": 6.837175952121306e-06,
      "loss": 3.2574,
      "step": 168
    },
    {
      "epoch": 0.1548327989005955,
      "grad_norm": 2.107384443283081,
      "learning_rate": 6.425787818636131e-06,
      "loss": 3.1769,
      "step": 169
    },
    {
      "epoch": 0.15574896930829135,
      "grad_norm": 2.1264374256134033,
      "learning_rate": 6.026312439675552e-06,
      "loss": 3.1278,
      "step": 170
    },
    {
      "epoch": 0.15666513971598717,
      "grad_norm": 2.0993690490722656,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 3.094,
      "step": 171
    },
    {
      "epoch": 0.157581310123683,
      "grad_norm": 2.096158504486084,
      "learning_rate": 5.263533508961827e-06,
      "loss": 3.201,
      "step": 172
    },
    {
      "epoch": 0.15849748053137883,
      "grad_norm": 2.153181791305542,
      "learning_rate": 4.900438493352055e-06,
      "loss": 3.091,
      "step": 173
    },
    {
      "epoch": 0.15941365093907467,
      "grad_norm": 2.22326397895813,
      "learning_rate": 4.549673247541875e-06,
      "loss": 3.054,
      "step": 174
    },
    {
      "epoch": 0.1603298213467705,
      "grad_norm": 2.1712334156036377,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 3.0009,
      "step": 175
    },
    {
      "epoch": 0.16124599175446633,
      "grad_norm": 2.244997501373291,
      "learning_rate": 3.885512251130763e-06,
      "loss": 3.1269,
      "step": 176
    },
    {
      "epoch": 0.16216216216216217,
      "grad_norm": 2.256870746612549,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 2.9981,
      "step": 177
    },
    {
      "epoch": 0.163078332569858,
      "grad_norm": 2.2340736389160156,
      "learning_rate": 3.271776770026963e-06,
      "loss": 3.2058,
      "step": 178
    },
    {
      "epoch": 0.16399450297755383,
      "grad_norm": 2.2496109008789062,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 3.2021,
      "step": 179
    },
    {
      "epoch": 0.16491067338524965,
      "grad_norm": 2.1700618267059326,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 2.9377,
      "step": 180
    },
    {
      "epoch": 0.1658268437929455,
      "grad_norm": 2.3376107215881348,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 3.2303,
      "step": 181
    },
    {
      "epoch": 0.16674301420064133,
      "grad_norm": 2.299830913543701,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 3.2369,
      "step": 182
    },
    {
      "epoch": 0.16765918460833715,
      "grad_norm": 2.3322975635528564,
      "learning_rate": 1.962316193157593e-06,
      "loss": 3.1317,
      "step": 183
    },
    {
      "epoch": 0.168575355016033,
      "grad_norm": 2.2672934532165527,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 3.177,
      "step": 184
    },
    {
      "epoch": 0.1694915254237288,
      "grad_norm": 2.281609296798706,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 3.1484,
      "step": 185
    },
    {
      "epoch": 0.17040769583142465,
      "grad_norm": 2.321932315826416,
      "learning_rate": 1.333670137599713e-06,
      "loss": 3.1907,
      "step": 186
    },
    {
      "epoch": 0.17132386623912046,
      "grad_norm": 2.4272027015686035,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 3.257,
      "step": 187
    },
    {
      "epoch": 0.1722400366468163,
      "grad_norm": 2.361142158508301,
      "learning_rate": 9.810017062595322e-07,
      "loss": 3.1788,
      "step": 188
    },
    {
      "epoch": 0.17315620705451215,
      "grad_norm": 2.381781816482544,
      "learning_rate": 8.247462563808817e-07,
      "loss": 3.0534,
      "step": 189
    },
    {
      "epoch": 0.17407237746220797,
      "grad_norm": 2.3604490756988525,
      "learning_rate": 6.819348298638839e-07,
      "loss": 3.0835,
      "step": 190
    },
    {
      "epoch": 0.1749885478699038,
      "grad_norm": 2.4307730197906494,
      "learning_rate": 5.526064699265753e-07,
      "loss": 3.1297,
      "step": 191
    },
    {
      "epoch": 0.17590471827759963,
      "grad_norm": 2.4940223693847656,
      "learning_rate": 4.367965336512403e-07,
      "loss": 3.1289,
      "step": 192
    },
    {
      "epoch": 0.17682088868529547,
      "grad_norm": 2.526481866836548,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 3.0664,
      "step": 193
    },
    {
      "epoch": 0.17773705909299128,
      "grad_norm": 2.5835301876068115,
      "learning_rate": 2.458548727494292e-07,
      "loss": 3.1121,
      "step": 194
    },
    {
      "epoch": 0.17865322950068713,
      "grad_norm": 2.5882251262664795,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 2.9637,
      "step": 195
    },
    {
      "epoch": 0.17956939990838297,
      "grad_norm": 2.683682441711426,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 3.1245,
      "step": 196
    },
    {
      "epoch": 0.18048557031607879,
      "grad_norm": 2.789461612701416,
      "learning_rate": 6.150154258476315e-08,
      "loss": 3.0841,
      "step": 197
    },
    {
      "epoch": 0.18140174072377463,
      "grad_norm": 2.9657254219055176,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 2.8502,
      "step": 198
    },
    {
      "epoch": 0.18231791113147044,
      "grad_norm": 3.0290040969848633,
      "learning_rate": 6.834750376549792e-09,
      "loss": 2.834,
      "step": 199
    },
    {
      "epoch": 0.1832340815391663,
      "grad_norm": 3.8789222240448,
      "learning_rate": 0.0,
      "loss": 3.0643,
      "step": 200
    },
    {
      "epoch": 0.1832340815391663,
      "eval_loss": 0.7749578952789307,
      "eval_runtime": 143.3394,
      "eval_samples_per_second": 12.83,
      "eval_steps_per_second": 3.209,
      "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": 2.890619231592776e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}