{
  "best_metric": 2.917475700378418,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 1.0230179028132993,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005115089514066497,
      "grad_norm": 0.8756385445594788,
      "learning_rate": 1e-05,
      "loss": 4.9893,
      "step": 1
    },
    {
      "epoch": 0.005115089514066497,
      "eval_loss": 7.246866226196289,
      "eval_runtime": 5.1915,
      "eval_samples_per_second": 63.565,
      "eval_steps_per_second": 15.988,
      "step": 1
    },
    {
      "epoch": 0.010230179028132993,
      "grad_norm": 0.8946313858032227,
      "learning_rate": 2e-05,
      "loss": 5.4603,
      "step": 2
    },
    {
      "epoch": 0.015345268542199489,
      "grad_norm": 0.963908314704895,
      "learning_rate": 3e-05,
      "loss": 5.3047,
      "step": 3
    },
    {
      "epoch": 0.020460358056265986,
      "grad_norm": 0.9848924875259399,
      "learning_rate": 4e-05,
      "loss": 5.194,
      "step": 4
    },
    {
      "epoch": 0.02557544757033248,
      "grad_norm": 1.0486536026000977,
      "learning_rate": 5e-05,
      "loss": 5.6278,
      "step": 5
    },
    {
      "epoch": 0.030690537084398978,
      "grad_norm": 0.9693508148193359,
      "learning_rate": 6e-05,
      "loss": 5.5091,
      "step": 6
    },
    {
      "epoch": 0.03580562659846547,
      "grad_norm": 0.9206507205963135,
      "learning_rate": 7e-05,
      "loss": 5.7703,
      "step": 7
    },
    {
      "epoch": 0.04092071611253197,
      "grad_norm": 0.9566873908042908,
      "learning_rate": 8e-05,
      "loss": 5.4608,
      "step": 8
    },
    {
      "epoch": 0.04603580562659847,
      "grad_norm": 0.9202797412872314,
      "learning_rate": 9e-05,
      "loss": 5.6997,
      "step": 9
    },
    {
      "epoch": 0.05115089514066496,
      "grad_norm": 0.9266080856323242,
      "learning_rate": 0.0001,
      "loss": 5.5102,
      "step": 10
    },
    {
      "epoch": 0.056265984654731455,
      "grad_norm": 0.7484850883483887,
      "learning_rate": 9.999316524962345e-05,
      "loss": 5.4521,
      "step": 11
    },
    {
      "epoch": 0.061381074168797956,
      "grad_norm": 0.7800479531288147,
      "learning_rate": 9.997266286704631e-05,
      "loss": 5.2792,
      "step": 12
    },
    {
      "epoch": 0.06649616368286446,
      "grad_norm": 0.7275058627128601,
      "learning_rate": 9.993849845741524e-05,
      "loss": 5.4013,
      "step": 13
    },
    {
      "epoch": 0.07161125319693094,
      "grad_norm": 0.6656175851821899,
      "learning_rate": 9.989068136093873e-05,
      "loss": 5.5943,
      "step": 14
    },
    {
      "epoch": 0.07672634271099744,
      "grad_norm": 0.876819908618927,
      "learning_rate": 9.98292246503335e-05,
      "loss": 5.2865,
      "step": 15
    },
    {
      "epoch": 0.08184143222506395,
      "grad_norm": 0.8198714256286621,
      "learning_rate": 9.975414512725057e-05,
      "loss": 5.4947,
      "step": 16
    },
    {
      "epoch": 0.08695652173913043,
      "grad_norm": 0.8344730138778687,
      "learning_rate": 9.966546331768191e-05,
      "loss": 5.7214,
      "step": 17
    },
    {
      "epoch": 0.09207161125319693,
      "grad_norm": 0.9426600933074951,
      "learning_rate": 9.956320346634876e-05,
      "loss": 5.1823,
      "step": 18
    },
    {
      "epoch": 0.09718670076726342,
      "grad_norm": 0.9687696099281311,
      "learning_rate": 9.944739353007344e-05,
      "loss": 5.3089,
      "step": 19
    },
    {
      "epoch": 0.10230179028132992,
      "grad_norm": 0.9249797463417053,
      "learning_rate": 9.931806517013612e-05,
      "loss": 5.1934,
      "step": 20
    },
    {
      "epoch": 0.10741687979539642,
      "grad_norm": 1.0114589929580688,
      "learning_rate": 9.917525374361912e-05,
      "loss": 5.3811,
      "step": 21
    },
    {
      "epoch": 0.11253196930946291,
      "grad_norm": 0.9795160293579102,
      "learning_rate": 9.901899829374047e-05,
      "loss": 5.4308,
      "step": 22
    },
    {
      "epoch": 0.11764705882352941,
      "grad_norm": 0.9637768268585205,
      "learning_rate": 9.884934153917997e-05,
      "loss": 5.5837,
      "step": 23
    },
    {
      "epoch": 0.12276214833759591,
      "grad_norm": 0.9522973299026489,
      "learning_rate": 9.86663298624003e-05,
      "loss": 5.3143,
      "step": 24
    },
    {
      "epoch": 0.1278772378516624,
      "grad_norm": 0.9822115302085876,
      "learning_rate": 9.847001329696653e-05,
      "loss": 5.3466,
      "step": 25
    },
    {
      "epoch": 0.1329923273657289,
      "grad_norm": 0.8550523519515991,
      "learning_rate": 9.826044551386744e-05,
      "loss": 5.1107,
      "step": 26
    },
    {
      "epoch": 0.13810741687979539,
      "grad_norm": 1.9659918546676636,
      "learning_rate": 9.803768380684242e-05,
      "loss": 5.2732,
      "step": 27
    },
    {
      "epoch": 0.1432225063938619,
      "grad_norm": 0.9923157691955566,
      "learning_rate": 9.780178907671789e-05,
      "loss": 5.3098,
      "step": 28
    },
    {
      "epoch": 0.1483375959079284,
      "grad_norm": 0.9949304461479187,
      "learning_rate": 9.755282581475769e-05,
      "loss": 5.1649,
      "step": 29
    },
    {
      "epoch": 0.1534526854219949,
      "grad_norm": 1.1100773811340332,
      "learning_rate": 9.729086208503174e-05,
      "loss": 5.0877,
      "step": 30
    },
    {
      "epoch": 0.1585677749360614,
      "grad_norm": 0.9826070070266724,
      "learning_rate": 9.701596950580806e-05,
      "loss": 5.1172,
      "step": 31
    },
    {
      "epoch": 0.1636828644501279,
      "grad_norm": 0.9882199168205261,
      "learning_rate": 9.672822322997305e-05,
      "loss": 5.145,
      "step": 32
    },
    {
      "epoch": 0.16879795396419436,
      "grad_norm": 1.2074486017227173,
      "learning_rate": 9.642770192448536e-05,
      "loss": 5.196,
      "step": 33
    },
    {
      "epoch": 0.17391304347826086,
      "grad_norm": 1.1388556957244873,
      "learning_rate": 9.611448774886924e-05,
      "loss": 4.7332,
      "step": 34
    },
    {
      "epoch": 0.17902813299232737,
      "grad_norm": 1.35393488407135,
      "learning_rate": 9.578866633275288e-05,
      "loss": 4.6465,
      "step": 35
    },
    {
      "epoch": 0.18414322250639387,
      "grad_norm": 2.2784202098846436,
      "learning_rate": 9.545032675245813e-05,
      "loss": 4.3716,
      "step": 36
    },
    {
      "epoch": 0.18925831202046037,
      "grad_norm": 3.097393751144409,
      "learning_rate": 9.509956150664796e-05,
      "loss": 4.6442,
      "step": 37
    },
    {
      "epoch": 0.19437340153452684,
      "grad_norm": 3.9054620265960693,
      "learning_rate": 9.473646649103818e-05,
      "loss": 3.4198,
      "step": 38
    },
    {
      "epoch": 0.19948849104859334,
      "grad_norm": 3.546821355819702,
      "learning_rate": 9.43611409721806e-05,
      "loss": 4.6547,
      "step": 39
    },
    {
      "epoch": 0.20460358056265984,
      "grad_norm": 4.750877857208252,
      "learning_rate": 9.397368756032445e-05,
      "loss": 3.9999,
      "step": 40
    },
    {
      "epoch": 0.20971867007672634,
      "grad_norm": 4.424086570739746,
      "learning_rate": 9.357421218136386e-05,
      "loss": 3.1996,
      "step": 41
    },
    {
      "epoch": 0.21483375959079284,
      "grad_norm": 4.076082706451416,
      "learning_rate": 9.316282404787871e-05,
      "loss": 2.7845,
      "step": 42
    },
    {
      "epoch": 0.21994884910485935,
      "grad_norm": 3.142268180847168,
      "learning_rate": 9.273963562927695e-05,
      "loss": 3.76,
      "step": 43
    },
    {
      "epoch": 0.22506393861892582,
      "grad_norm": 3.921541690826416,
      "learning_rate": 9.230476262104677e-05,
      "loss": 3.3581,
      "step": 44
    },
    {
      "epoch": 0.23017902813299232,
      "grad_norm": 3.352081298828125,
      "learning_rate": 9.185832391312644e-05,
      "loss": 3.2676,
      "step": 45
    },
    {
      "epoch": 0.23529411764705882,
      "grad_norm": 2.9872987270355225,
      "learning_rate": 9.140044155740101e-05,
      "loss": 3.1869,
      "step": 46
    },
    {
      "epoch": 0.24040920716112532,
      "grad_norm": 3.149404287338257,
      "learning_rate": 9.093124073433463e-05,
      "loss": 2.9462,
      "step": 47
    },
    {
      "epoch": 0.24552429667519182,
      "grad_norm": 5.2785773277282715,
      "learning_rate": 9.045084971874738e-05,
      "loss": 10.4238,
      "step": 48
    },
    {
      "epoch": 0.2506393861892583,
      "grad_norm": 15.782574653625488,
      "learning_rate": 8.995939984474624e-05,
      "loss": 11.7345,
      "step": 49
    },
    {
      "epoch": 0.2557544757033248,
      "grad_norm": 16.099925994873047,
      "learning_rate": 8.945702546981969e-05,
      "loss": 11.7832,
      "step": 50
    },
    {
      "epoch": 0.2557544757033248,
      "eval_loss": 8.762343406677246,
      "eval_runtime": 5.1593,
      "eval_samples_per_second": 63.963,
      "eval_steps_per_second": 16.088,
      "step": 50
    },
    {
      "epoch": 0.2608695652173913,
      "grad_norm": 16.021198272705078,
      "learning_rate": 8.894386393810563e-05,
      "loss": 10.8511,
      "step": 51
    },
    {
      "epoch": 0.2659846547314578,
      "grad_norm": 15.505819320678711,
      "learning_rate": 8.842005554284296e-05,
      "loss": 9.9605,
      "step": 52
    },
    {
      "epoch": 0.2710997442455243,
      "grad_norm": 15.06153392791748,
      "learning_rate": 8.788574348801675e-05,
      "loss": 8.749,
      "step": 53
    },
    {
      "epoch": 0.27621483375959077,
      "grad_norm": 13.483952522277832,
      "learning_rate": 8.73410738492077e-05,
      "loss": 7.7458,
      "step": 54
    },
    {
      "epoch": 0.2813299232736573,
      "grad_norm": 9.045783996582031,
      "learning_rate": 8.678619553365659e-05,
      "loss": 6.0171,
      "step": 55
    },
    {
      "epoch": 0.2864450127877238,
      "grad_norm": 5.935639381408691,
      "learning_rate": 8.622126023955446e-05,
      "loss": 5.2473,
      "step": 56
    },
    {
      "epoch": 0.2915601023017903,
      "grad_norm": 3.589444875717163,
      "learning_rate": 8.564642241456986e-05,
      "loss": 4.8008,
      "step": 57
    },
    {
      "epoch": 0.2966751918158568,
      "grad_norm": 2.4138681888580322,
      "learning_rate": 8.506183921362443e-05,
      "loss": 4.7171,
      "step": 58
    },
    {
      "epoch": 0.30179028132992325,
      "grad_norm": 1.62509286403656,
      "learning_rate": 8.44676704559283e-05,
      "loss": 4.4878,
      "step": 59
    },
    {
      "epoch": 0.3069053708439898,
      "grad_norm": 1.4101530313491821,
      "learning_rate": 8.386407858128706e-05,
      "loss": 4.4446,
      "step": 60
    },
    {
      "epoch": 0.31202046035805625,
      "grad_norm": 1.1992028951644897,
      "learning_rate": 8.32512286056924e-05,
      "loss": 4.3495,
      "step": 61
    },
    {
      "epoch": 0.3171355498721228,
      "grad_norm": 1.206842303276062,
      "learning_rate": 8.262928807620843e-05,
      "loss": 4.2656,
      "step": 62
    },
    {
      "epoch": 0.32225063938618925,
      "grad_norm": 1.3343909978866577,
      "learning_rate": 8.199842702516583e-05,
      "loss": 4.1758,
      "step": 63
    },
    {
      "epoch": 0.3273657289002558,
      "grad_norm": 1.2275749444961548,
      "learning_rate": 8.135881792367686e-05,
      "loss": 4.3725,
      "step": 64
    },
    {
      "epoch": 0.33248081841432225,
      "grad_norm": 1.2183395624160767,
      "learning_rate": 8.07106356344834e-05,
      "loss": 3.909,
      "step": 65
    },
    {
      "epoch": 0.3375959079283887,
      "grad_norm": 1.0592944622039795,
      "learning_rate": 8.005405736415126e-05,
      "loss": 4.2697,
      "step": 66
    },
    {
      "epoch": 0.34271099744245526,
      "grad_norm": 1.2135802507400513,
      "learning_rate": 7.938926261462366e-05,
      "loss": 4.0459,
      "step": 67
    },
    {
      "epoch": 0.34782608695652173,
      "grad_norm": 0.9741806983947754,
      "learning_rate": 7.871643313414718e-05,
      "loss": 4.0884,
      "step": 68
    },
    {
      "epoch": 0.35294117647058826,
      "grad_norm": 0.9708122611045837,
      "learning_rate": 7.803575286758364e-05,
      "loss": 4.1056,
      "step": 69
    },
    {
      "epoch": 0.35805626598465473,
      "grad_norm": 0.9863829016685486,
      "learning_rate": 7.734740790612136e-05,
      "loss": 4.1001,
      "step": 70
    },
    {
      "epoch": 0.3631713554987212,
      "grad_norm": 0.7858729362487793,
      "learning_rate": 7.66515864363997e-05,
      "loss": 4.1358,
      "step": 71
    },
    {
      "epoch": 0.36828644501278773,
      "grad_norm": 0.8839144706726074,
      "learning_rate": 7.594847868906076e-05,
      "loss": 3.8212,
      "step": 72
    },
    {
      "epoch": 0.3734015345268542,
      "grad_norm": 0.8642678260803223,
      "learning_rate": 7.52382768867422e-05,
      "loss": 4.3446,
      "step": 73
    },
    {
      "epoch": 0.37851662404092073,
      "grad_norm": 0.8339263200759888,
      "learning_rate": 7.452117519152542e-05,
      "loss": 4.0435,
      "step": 74
    },
    {
      "epoch": 0.3836317135549872,
      "grad_norm": 0.8290123343467712,
      "learning_rate": 7.379736965185368e-05,
      "loss": 4.0089,
      "step": 75
    },
    {
      "epoch": 0.3887468030690537,
      "grad_norm": 0.834225058555603,
      "learning_rate": 7.30670581489344e-05,
      "loss": 4.3068,
      "step": 76
    },
    {
      "epoch": 0.3938618925831202,
      "grad_norm": 0.8786638975143433,
      "learning_rate": 7.233044034264034e-05,
      "loss": 4.0081,
      "step": 77
    },
    {
      "epoch": 0.3989769820971867,
      "grad_norm": 0.8940492272377014,
      "learning_rate": 7.158771761692464e-05,
      "loss": 4.3048,
      "step": 78
    },
    {
      "epoch": 0.4040920716112532,
      "grad_norm": 1.021957278251648,
      "learning_rate": 7.083909302476453e-05,
      "loss": 4.6177,
      "step": 79
    },
    {
      "epoch": 0.4092071611253197,
      "grad_norm": 1.0327810049057007,
      "learning_rate": 7.008477123264848e-05,
      "loss": 4.0423,
      "step": 80
    },
    {
      "epoch": 0.4143222506393862,
      "grad_norm": 2.2245564460754395,
      "learning_rate": 6.932495846462261e-05,
      "loss": 4.652,
      "step": 81
    },
    {
      "epoch": 0.4194373401534527,
      "grad_norm": 1.0273576974868774,
      "learning_rate": 6.855986244591104e-05,
      "loss": 4.1437,
      "step": 82
    },
    {
      "epoch": 0.42455242966751916,
      "grad_norm": 1.1362370252609253,
      "learning_rate": 6.778969234612584e-05,
      "loss": 4.9611,
      "step": 83
    },
    {
      "epoch": 0.4296675191815857,
      "grad_norm": 1.146104097366333,
      "learning_rate": 6.701465872208216e-05,
      "loss": 3.2239,
      "step": 84
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 1.5032472610473633,
      "learning_rate": 6.623497346023418e-05,
      "loss": 2.6577,
      "step": 85
    },
    {
      "epoch": 0.4398976982097187,
      "grad_norm": 1.3953113555908203,
      "learning_rate": 6.545084971874738e-05,
      "loss": 2.7042,
      "step": 86
    },
    {
      "epoch": 0.44501278772378516,
      "grad_norm": 1.5841379165649414,
      "learning_rate": 6.466250186922325e-05,
      "loss": 1.2121,
      "step": 87
    },
    {
      "epoch": 0.45012787723785164,
      "grad_norm": 1.4610826969146729,
      "learning_rate": 6.387014543809223e-05,
      "loss": 1.0477,
      "step": 88
    },
    {
      "epoch": 0.45524296675191817,
      "grad_norm": 1.432998776435852,
      "learning_rate": 6.307399704769099e-05,
      "loss": 1.2293,
      "step": 89
    },
    {
      "epoch": 0.46035805626598464,
      "grad_norm": 1.2084925174713135,
      "learning_rate": 6.227427435703997e-05,
      "loss": 1.4464,
      "step": 90
    },
    {
      "epoch": 0.46547314578005117,
      "grad_norm": 1.7171345949172974,
      "learning_rate": 6.147119600233758e-05,
      "loss": 2.2629,
      "step": 91
    },
    {
      "epoch": 0.47058823529411764,
      "grad_norm": 2.0102758407592773,
      "learning_rate": 6.066498153718735e-05,
      "loss": 1.746,
      "step": 92
    },
    {
      "epoch": 0.47570332480818417,
      "grad_norm": 1.8124990463256836,
      "learning_rate": 5.985585137257401e-05,
      "loss": 1.4494,
      "step": 93
    },
    {
      "epoch": 0.48081841432225064,
      "grad_norm": 1.9506274461746216,
      "learning_rate": 5.90440267166055e-05,
      "loss": 1.704,
      "step": 94
    },
    {
      "epoch": 0.4859335038363171,
      "grad_norm": 7.607231140136719,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 2.0192,
      "step": 95
    },
    {
      "epoch": 0.49104859335038364,
      "grad_norm": 4.36256217956543,
      "learning_rate": 5.74131823855921e-05,
      "loss": 8.8899,
      "step": 96
    },
    {
      "epoch": 0.4961636828644501,
      "grad_norm": 1.4300196170806885,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 3.8004,
      "step": 97
    },
    {
      "epoch": 0.5012787723785166,
      "grad_norm": 1.375491976737976,
      "learning_rate": 5.577423184847932e-05,
      "loss": 3.6296,
      "step": 98
    },
    {
      "epoch": 0.5063938618925832,
      "grad_norm": 1.313018560409546,
      "learning_rate": 5.495227651252315e-05,
      "loss": 3.6974,
      "step": 99
    },
    {
      "epoch": 0.5115089514066496,
      "grad_norm": 1.388663649559021,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 3.6027,
      "step": 100
    },
    {
      "epoch": 0.5115089514066496,
      "eval_loss": 3.444181203842163,
      "eval_runtime": 5.1428,
      "eval_samples_per_second": 64.168,
      "eval_steps_per_second": 16.139,
      "step": 100
    },
    {
      "epoch": 0.5166240409207161,
      "grad_norm": 1.2643789052963257,
      "learning_rate": 5.330452921628497e-05,
      "loss": 3.6121,
      "step": 101
    },
    {
      "epoch": 0.5217391304347826,
      "grad_norm": 1.1502041816711426,
      "learning_rate": 5.247918773366112e-05,
      "loss": 3.5931,
      "step": 102
    },
    {
      "epoch": 0.5268542199488491,
      "grad_norm": 1.1222349405288696,
      "learning_rate": 5.165316846586541e-05,
      "loss": 3.4811,
      "step": 103
    },
    {
      "epoch": 0.5319693094629157,
      "grad_norm": 0.9348773956298828,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 3.5376,
      "step": 104
    },
    {
      "epoch": 0.5370843989769821,
      "grad_norm": 0.8603200316429138,
      "learning_rate": 5e-05,
      "loss": 3.4854,
      "step": 105
    },
    {
      "epoch": 0.5421994884910486,
      "grad_norm": 0.9454997777938843,
      "learning_rate": 4.917330276168208e-05,
      "loss": 3.3169,
      "step": 106
    },
    {
      "epoch": 0.5473145780051151,
      "grad_norm": 0.8435836434364319,
      "learning_rate": 4.834683153413459e-05,
      "loss": 3.5956,
      "step": 107
    },
    {
      "epoch": 0.5524296675191815,
      "grad_norm": 0.8037415742874146,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 3.5245,
      "step": 108
    },
    {
      "epoch": 0.5575447570332481,
      "grad_norm": 0.7555946111679077,
      "learning_rate": 4.669547078371504e-05,
      "loss": 3.2983,
      "step": 109
    },
    {
      "epoch": 0.5626598465473146,
      "grad_norm": 0.7772465944290161,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 3.4548,
      "step": 110
    },
    {
      "epoch": 0.5677749360613811,
      "grad_norm": 0.8697976469993591,
      "learning_rate": 4.504772348747687e-05,
      "loss": 3.6988,
      "step": 111
    },
    {
      "epoch": 0.5728900255754475,
      "grad_norm": 0.7847548723220825,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 3.628,
      "step": 112
    },
    {
      "epoch": 0.578005115089514,
      "grad_norm": 0.8165000677108765,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 3.7546,
      "step": 113
    },
    {
      "epoch": 0.5831202046035806,
      "grad_norm": 0.7947345972061157,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 3.6399,
      "step": 114
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 0.8536720871925354,
      "learning_rate": 4.17702704859633e-05,
      "loss": 3.8268,
      "step": 115
    },
    {
      "epoch": 0.5933503836317136,
      "grad_norm": 0.8350384831428528,
      "learning_rate": 4.095597328339452e-05,
      "loss": 3.5219,
      "step": 116
    },
    {
      "epoch": 0.59846547314578,
      "grad_norm": 0.9872482419013977,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 3.5044,
      "step": 117
    },
    {
      "epoch": 0.6035805626598465,
      "grad_norm": 0.8847600221633911,
      "learning_rate": 3.933501846281267e-05,
      "loss": 3.5626,
      "step": 118
    },
    {
      "epoch": 0.6086956521739131,
      "grad_norm": 0.9034833908081055,
      "learning_rate": 3.852880399766243e-05,
      "loss": 3.4859,
      "step": 119
    },
    {
      "epoch": 0.6138107416879796,
      "grad_norm": 0.8705306053161621,
      "learning_rate": 3.772572564296005e-05,
      "loss": 3.3755,
      "step": 120
    },
    {
      "epoch": 0.618925831202046,
      "grad_norm": 0.9759798049926758,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 3.6691,
      "step": 121
    },
    {
      "epoch": 0.6240409207161125,
      "grad_norm": 1.5106711387634277,
      "learning_rate": 3.612985456190778e-05,
      "loss": 3.7934,
      "step": 122
    },
    {
      "epoch": 0.629156010230179,
      "grad_norm": 0.9518977999687195,
      "learning_rate": 3.533749813077677e-05,
      "loss": 3.6055,
      "step": 123
    },
    {
      "epoch": 0.6342710997442456,
      "grad_norm": 0.9911556839942932,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 3.8592,
      "step": 124
    },
    {
      "epoch": 0.639386189258312,
      "grad_norm": 0.9979852437973022,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 3.8736,
      "step": 125
    },
    {
      "epoch": 0.6445012787723785,
      "grad_norm": 0.909438967704773,
      "learning_rate": 3.298534127791785e-05,
      "loss": 4.016,
      "step": 126
    },
    {
      "epoch": 0.649616368286445,
      "grad_norm": 0.9996523261070251,
      "learning_rate": 3.221030765387417e-05,
      "loss": 3.582,
      "step": 127
    },
    {
      "epoch": 0.6547314578005116,
      "grad_norm": 0.9997970461845398,
      "learning_rate": 3.144013755408895e-05,
      "loss": 3.7646,
      "step": 128
    },
    {
      "epoch": 0.659846547314578,
      "grad_norm": 1.1803526878356934,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 3.9305,
      "step": 129
    },
    {
      "epoch": 0.6649616368286445,
      "grad_norm": 1.0851036310195923,
      "learning_rate": 2.991522876735154e-05,
      "loss": 3.9654,
      "step": 130
    },
    {
      "epoch": 0.670076726342711,
      "grad_norm": 1.2465025186538696,
      "learning_rate": 2.916090697523549e-05,
      "loss": 3.8419,
      "step": 131
    },
    {
      "epoch": 0.6751918158567775,
      "grad_norm": 0.9848014712333679,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 2.0624,
      "step": 132
    },
    {
      "epoch": 0.680306905370844,
      "grad_norm": 1.2557348012924194,
      "learning_rate": 2.766955965735968e-05,
      "loss": 2.4528,
      "step": 133
    },
    {
      "epoch": 0.6854219948849105,
      "grad_norm": 1.3081684112548828,
      "learning_rate": 2.693294185106562e-05,
      "loss": 2.09,
      "step": 134
    },
    {
      "epoch": 0.690537084398977,
      "grad_norm": 1.0094250440597534,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 0.6984,
      "step": 135
    },
    {
      "epoch": 0.6956521739130435,
      "grad_norm": 1.094954490661621,
      "learning_rate": 2.547882480847461e-05,
      "loss": 0.5682,
      "step": 136
    },
    {
      "epoch": 0.7007672634271099,
      "grad_norm": 0.8255013227462769,
      "learning_rate": 2.476172311325783e-05,
      "loss": 0.4677,
      "step": 137
    },
    {
      "epoch": 0.7058823529411765,
      "grad_norm": 0.9566165208816528,
      "learning_rate": 2.405152131093926e-05,
      "loss": 0.5419,
      "step": 138
    },
    {
      "epoch": 0.710997442455243,
      "grad_norm": 1.122994065284729,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 1.5137,
      "step": 139
    },
    {
      "epoch": 0.7161125319693095,
      "grad_norm": 1.3460452556610107,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 0.9566,
      "step": 140
    },
    {
      "epoch": 0.7212276214833759,
      "grad_norm": 1.302507758140564,
      "learning_rate": 2.196424713241637e-05,
      "loss": 0.8735,
      "step": 141
    },
    {
      "epoch": 0.7263427109974424,
      "grad_norm": 1.1895112991333008,
      "learning_rate": 2.128356686585282e-05,
      "loss": 0.7734,
      "step": 142
    },
    {
      "epoch": 0.731457800511509,
      "grad_norm": 1.4312363862991333,
      "learning_rate": 2.061073738537635e-05,
      "loss": 0.9322,
      "step": 143
    },
    {
      "epoch": 0.7365728900255755,
      "grad_norm": 3.6024553775787354,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 8.8607,
      "step": 144
    },
    {
      "epoch": 0.7416879795396419,
      "grad_norm": 1.0639846324920654,
      "learning_rate": 1.928936436551661e-05,
      "loss": 3.5509,
      "step": 145
    },
    {
      "epoch": 0.7468030690537084,
      "grad_norm": 1.1059770584106445,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 3.1506,
      "step": 146
    },
    {
      "epoch": 0.7519181585677749,
      "grad_norm": 1.0169697999954224,
      "learning_rate": 1.800157297483417e-05,
      "loss": 3.3809,
      "step": 147
    },
    {
      "epoch": 0.7570332480818415,
      "grad_norm": 0.968116819858551,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 3.4212,
      "step": 148
    },
    {
      "epoch": 0.7621483375959079,
      "grad_norm": 0.9484832286834717,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 3.2916,
      "step": 149
    },
    {
      "epoch": 0.7672634271099744,
      "grad_norm": 0.8964362144470215,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 3.3095,
      "step": 150
    },
    {
      "epoch": 0.7672634271099744,
      "eval_loss": 2.982480764389038,
      "eval_runtime": 5.1713,
      "eval_samples_per_second": 63.813,
      "eval_steps_per_second": 16.05,
      "step": 150
    },
    {
      "epoch": 0.7723785166240409,
      "grad_norm": 1.194737195968628,
      "learning_rate": 1.553232954407171e-05,
      "loss": 3.3856,
      "step": 151
    },
    {
      "epoch": 0.7774936061381074,
      "grad_norm": 0.9305944442749023,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 3.529,
      "step": 152
    },
    {
      "epoch": 0.782608695652174,
      "grad_norm": 0.8583710789680481,
      "learning_rate": 1.435357758543015e-05,
      "loss": 3.3139,
      "step": 153
    },
    {
      "epoch": 0.7877237851662404,
      "grad_norm": 0.8682720065116882,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 2.9918,
      "step": 154
    },
    {
      "epoch": 0.7928388746803069,
      "grad_norm": 0.9328944683074951,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 3.1411,
      "step": 155
    },
    {
      "epoch": 0.7979539641943734,
      "grad_norm": 0.7475224137306213,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 3.2944,
      "step": 156
    },
    {
      "epoch": 0.80306905370844,
      "grad_norm": 0.7422310709953308,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 3.4336,
      "step": 157
    },
    {
      "epoch": 0.8081841432225064,
      "grad_norm": 0.7506616115570068,
      "learning_rate": 1.157994445715706e-05,
      "loss": 3.3879,
      "step": 158
    },
    {
      "epoch": 0.8132992327365729,
      "grad_norm": 0.7556441426277161,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 3.4628,
      "step": 159
    },
    {
      "epoch": 0.8184143222506394,
      "grad_norm": 0.8010793924331665,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 3.6085,
      "step": 160
    },
    {
      "epoch": 0.8235294117647058,
      "grad_norm": 0.8776047825813293,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 3.3893,
      "step": 161
    },
    {
      "epoch": 0.8286445012787724,
      "grad_norm": 0.8556779623031616,
      "learning_rate": 9.549150281252633e-06,
      "loss": 3.2902,
      "step": 162
    },
    {
      "epoch": 0.8337595907928389,
      "grad_norm": 0.945901095867157,
      "learning_rate": 9.068759265665384e-06,
      "loss": 3.2806,
      "step": 163
    },
    {
      "epoch": 0.8388746803069054,
      "grad_norm": 0.8959454894065857,
      "learning_rate": 8.599558442598998e-06,
      "loss": 3.4154,
      "step": 164
    },
    {
      "epoch": 0.8439897698209718,
      "grad_norm": 0.8776317834854126,
      "learning_rate": 8.141676086873572e-06,
      "loss": 3.4694,
      "step": 165
    },
    {
      "epoch": 0.8491048593350383,
      "grad_norm": 0.8572931885719299,
      "learning_rate": 7.695237378953223e-06,
      "loss": 3.2485,
      "step": 166
    },
    {
      "epoch": 0.8542199488491049,
      "grad_norm": 1.0760451555252075,
      "learning_rate": 7.260364370723044e-06,
      "loss": 3.6148,
      "step": 167
    },
    {
      "epoch": 0.8593350383631714,
      "grad_norm": 0.8359008431434631,
      "learning_rate": 6.837175952121306e-06,
      "loss": 3.335,
      "step": 168
    },
    {
      "epoch": 0.8644501278772379,
      "grad_norm": 1.1002939939498901,
      "learning_rate": 6.425787818636131e-06,
      "loss": 3.4758,
      "step": 169
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 0.9957254528999329,
      "learning_rate": 6.026312439675552e-06,
      "loss": 3.7484,
      "step": 170
    },
    {
      "epoch": 0.8746803069053708,
      "grad_norm": 1.0247571468353271,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 3.5808,
      "step": 171
    },
    {
      "epoch": 0.8797953964194374,
      "grad_norm": 1.014377236366272,
      "learning_rate": 5.263533508961827e-06,
      "loss": 3.6026,
      "step": 172
    },
    {
      "epoch": 0.8849104859335039,
      "grad_norm": 1.0075761079788208,
      "learning_rate": 4.900438493352055e-06,
      "loss": 3.5915,
      "step": 173
    },
    {
      "epoch": 0.8900255754475703,
      "grad_norm": 1.0310754776000977,
      "learning_rate": 4.549673247541875e-06,
      "loss": 3.7095,
      "step": 174
    },
    {
      "epoch": 0.8951406649616368,
      "grad_norm": 0.9983537197113037,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 3.8526,
      "step": 175
    },
    {
      "epoch": 0.9002557544757033,
      "grad_norm": 1.092082142829895,
      "learning_rate": 3.885512251130763e-06,
      "loss": 3.3114,
      "step": 176
    },
    {
      "epoch": 0.9053708439897699,
      "grad_norm": 1.1465011835098267,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 3.7109,
      "step": 177
    },
    {
      "epoch": 0.9104859335038363,
      "grad_norm": 1.2844728231430054,
      "learning_rate": 3.271776770026963e-06,
      "loss": 3.9306,
      "step": 178
    },
    {
      "epoch": 0.9156010230179028,
      "grad_norm": 1.0861852169036865,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 3.5152,
      "step": 179
    },
    {
      "epoch": 0.9207161125319693,
      "grad_norm": 1.5091826915740967,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 3.8524,
      "step": 180
    },
    {
      "epoch": 0.9258312020460358,
      "grad_norm": 0.9985364675521851,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 1.4663,
      "step": 181
    },
    {
      "epoch": 0.9309462915601023,
      "grad_norm": 1.1269094944000244,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 1.465,
      "step": 182
    },
    {
      "epoch": 0.9360613810741688,
      "grad_norm": 4.568881988525391,
      "learning_rate": 1.962316193157593e-06,
      "loss": 0.4571,
      "step": 183
    },
    {
      "epoch": 0.9411764705882353,
      "grad_norm": 0.9197856187820435,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 0.5392,
      "step": 184
    },
    {
      "epoch": 0.9462915601023018,
      "grad_norm": 3.1013917922973633,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 0.5488,
      "step": 185
    },
    {
      "epoch": 0.9514066496163683,
      "grad_norm": 0.9130269885063171,
      "learning_rate": 1.333670137599713e-06,
      "loss": 0.6014,
      "step": 186
    },
    {
      "epoch": 0.9565217391304348,
      "grad_norm": 2.6520581245422363,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 1.7468,
      "step": 187
    },
    {
      "epoch": 0.9616368286445013,
      "grad_norm": 1.3507983684539795,
      "learning_rate": 9.810017062595322e-07,
      "loss": 0.919,
      "step": 188
    },
    {
      "epoch": 0.9667519181585678,
      "grad_norm": 1.8945026397705078,
      "learning_rate": 8.247462563808817e-07,
      "loss": 1.4609,
      "step": 189
    },
    {
      "epoch": 0.9718670076726342,
      "grad_norm": 1.5751993656158447,
      "learning_rate": 6.819348298638839e-07,
      "loss": 1.1708,
      "step": 190
    },
    {
      "epoch": 0.9769820971867008,
      "grad_norm": 13.107644081115723,
      "learning_rate": 5.526064699265753e-07,
      "loss": 0.9725,
      "step": 191
    },
    {
      "epoch": 0.9820971867007673,
      "grad_norm": 5.111643314361572,
      "learning_rate": 4.367965336512403e-07,
      "loss": 11.959,
      "step": 192
    },
    {
      "epoch": 0.9872122762148338,
      "grad_norm": 0.6885824203491211,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 3.1883,
      "step": 193
    },
    {
      "epoch": 0.9923273657289002,
      "grad_norm": 0.81787109375,
      "learning_rate": 2.458548727494292e-07,
      "loss": 3.4034,
      "step": 194
    },
    {
      "epoch": 0.9974424552429667,
      "grad_norm": 1.057052493095398,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 2.3718,
      "step": 195
    },
    {
      "epoch": 1.0025575447570332,
      "grad_norm": 1.9029217958450317,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 2.4876,
      "step": 196
    },
    {
      "epoch": 1.0076726342710998,
      "grad_norm": 0.6901929378509521,
      "learning_rate": 6.150154258476315e-08,
      "loss": 2.9838,
      "step": 197
    },
    {
      "epoch": 1.0127877237851663,
      "grad_norm": 0.6399970054626465,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 3.1486,
      "step": 198
    },
    {
      "epoch": 1.0179028132992327,
      "grad_norm": 0.7350558638572693,
      "learning_rate": 6.834750376549792e-09,
      "loss": 3.0224,
      "step": 199
    },
    {
      "epoch": 1.0230179028132993,
      "grad_norm": 0.8037768602371216,
      "learning_rate": 0.0,
      "loss": 2.9867,
      "step": 200
    },
    {
      "epoch": 1.0230179028132993,
      "eval_loss": 2.917475700378418,
      "eval_runtime": 5.1616,
      "eval_samples_per_second": 63.934,
      "eval_steps_per_second": 16.08,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "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": 1.373561255559168e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}