{
  "best_metric": 1.3407422304153442,
  "best_model_checkpoint": "miner_id_24/checkpoint-100",
  "epoch": 2.9197080291970803,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.014598540145985401,
      "grad_norm": 0.33876606822013855,
      "learning_rate": 1e-05,
      "loss": 1.181,
      "step": 1
    },
    {
      "epoch": 0.014598540145985401,
      "eval_loss": 1.5133534669876099,
      "eval_runtime": 9.3214,
      "eval_samples_per_second": 12.444,
      "eval_steps_per_second": 3.111,
      "step": 1
    },
    {
      "epoch": 0.029197080291970802,
      "grad_norm": 0.22813493013381958,
      "learning_rate": 2e-05,
      "loss": 1.4452,
      "step": 2
    },
    {
      "epoch": 0.043795620437956206,
      "grad_norm": 0.38017651438713074,
      "learning_rate": 3e-05,
      "loss": 1.4234,
      "step": 3
    },
    {
      "epoch": 0.058394160583941604,
      "grad_norm": 0.2631050646305084,
      "learning_rate": 4e-05,
      "loss": 1.4075,
      "step": 4
    },
    {
      "epoch": 0.072992700729927,
      "grad_norm": 0.25626707077026367,
      "learning_rate": 5e-05,
      "loss": 1.3441,
      "step": 5
    },
    {
      "epoch": 0.08759124087591241,
      "grad_norm": 0.26295986771583557,
      "learning_rate": 6e-05,
      "loss": 1.3305,
      "step": 6
    },
    {
      "epoch": 0.10218978102189781,
      "grad_norm": 0.2986043691635132,
      "learning_rate": 7e-05,
      "loss": 1.4307,
      "step": 7
    },
    {
      "epoch": 0.11678832116788321,
      "grad_norm": 0.3112764060497284,
      "learning_rate": 8e-05,
      "loss": 1.5268,
      "step": 8
    },
    {
      "epoch": 0.13138686131386862,
      "grad_norm": 0.3284042179584503,
      "learning_rate": 9e-05,
      "loss": 1.5571,
      "step": 9
    },
    {
      "epoch": 0.145985401459854,
      "grad_norm": 0.31436583399772644,
      "learning_rate": 0.0001,
      "loss": 1.5346,
      "step": 10
    },
    {
      "epoch": 0.16058394160583941,
      "grad_norm": 0.34263306856155396,
      "learning_rate": 9.999316524962345e-05,
      "loss": 1.377,
      "step": 11
    },
    {
      "epoch": 0.17518248175182483,
      "grad_norm": 0.34792596101760864,
      "learning_rate": 9.997266286704631e-05,
      "loss": 1.5105,
      "step": 12
    },
    {
      "epoch": 0.1897810218978102,
      "grad_norm": 0.3962007164955139,
      "learning_rate": 9.993849845741524e-05,
      "loss": 1.6025,
      "step": 13
    },
    {
      "epoch": 0.20437956204379562,
      "grad_norm": 0.463651567697525,
      "learning_rate": 9.989068136093873e-05,
      "loss": 1.6562,
      "step": 14
    },
    {
      "epoch": 0.21897810218978103,
      "grad_norm": 0.519864559173584,
      "learning_rate": 9.98292246503335e-05,
      "loss": 1.5119,
      "step": 15
    },
    {
      "epoch": 0.23357664233576642,
      "grad_norm": 0.7638342976570129,
      "learning_rate": 9.975414512725057e-05,
      "loss": 1.5523,
      "step": 16
    },
    {
      "epoch": 0.24817518248175183,
      "grad_norm": 1.083395004272461,
      "learning_rate": 9.966546331768191e-05,
      "loss": 1.5634,
      "step": 17
    },
    {
      "epoch": 0.26277372262773724,
      "grad_norm": 0.22973737120628357,
      "learning_rate": 9.956320346634876e-05,
      "loss": 1.1169,
      "step": 18
    },
    {
      "epoch": 0.2773722627737226,
      "grad_norm": 0.2813929617404938,
      "learning_rate": 9.944739353007344e-05,
      "loss": 1.3161,
      "step": 19
    },
    {
      "epoch": 0.291970802919708,
      "grad_norm": 0.32428091764450073,
      "learning_rate": 9.931806517013612e-05,
      "loss": 1.4935,
      "step": 20
    },
    {
      "epoch": 0.30656934306569344,
      "grad_norm": 0.2738804221153259,
      "learning_rate": 9.917525374361912e-05,
      "loss": 1.4167,
      "step": 21
    },
    {
      "epoch": 0.32116788321167883,
      "grad_norm": 0.26207801699638367,
      "learning_rate": 9.901899829374047e-05,
      "loss": 1.3319,
      "step": 22
    },
    {
      "epoch": 0.3357664233576642,
      "grad_norm": 0.27858981490135193,
      "learning_rate": 9.884934153917997e-05,
      "loss": 1.2409,
      "step": 23
    },
    {
      "epoch": 0.35036496350364965,
      "grad_norm": 0.2742447257041931,
      "learning_rate": 9.86663298624003e-05,
      "loss": 1.2964,
      "step": 24
    },
    {
      "epoch": 0.36496350364963503,
      "grad_norm": 0.268431156873703,
      "learning_rate": 9.847001329696653e-05,
      "loss": 1.3255,
      "step": 25
    },
    {
      "epoch": 0.3795620437956204,
      "grad_norm": 0.27622514963150024,
      "learning_rate": 9.826044551386744e-05,
      "loss": 1.5215,
      "step": 26
    },
    {
      "epoch": 0.39416058394160586,
      "grad_norm": 0.284374475479126,
      "learning_rate": 9.803768380684242e-05,
      "loss": 1.4688,
      "step": 27
    },
    {
      "epoch": 0.40875912408759124,
      "grad_norm": 0.30457139015197754,
      "learning_rate": 9.780178907671789e-05,
      "loss": 1.4599,
      "step": 28
    },
    {
      "epoch": 0.4233576642335766,
      "grad_norm": 0.30239981412887573,
      "learning_rate": 9.755282581475769e-05,
      "loss": 1.2839,
      "step": 29
    },
    {
      "epoch": 0.43795620437956206,
      "grad_norm": 0.33081021904945374,
      "learning_rate": 9.729086208503174e-05,
      "loss": 1.5109,
      "step": 30
    },
    {
      "epoch": 0.45255474452554745,
      "grad_norm": 0.40180522203445435,
      "learning_rate": 9.701596950580806e-05,
      "loss": 1.5508,
      "step": 31
    },
    {
      "epoch": 0.46715328467153283,
      "grad_norm": 0.4250384569168091,
      "learning_rate": 9.672822322997305e-05,
      "loss": 1.3818,
      "step": 32
    },
    {
      "epoch": 0.48175182481751827,
      "grad_norm": 0.5974124073982239,
      "learning_rate": 9.642770192448536e-05,
      "loss": 1.2643,
      "step": 33
    },
    {
      "epoch": 0.49635036496350365,
      "grad_norm": 1.155615210533142,
      "learning_rate": 9.611448774886924e-05,
      "loss": 1.7648,
      "step": 34
    },
    {
      "epoch": 0.5109489051094891,
      "grad_norm": 0.23765483498573303,
      "learning_rate": 9.578866633275288e-05,
      "loss": 1.0542,
      "step": 35
    },
    {
      "epoch": 0.5255474452554745,
      "grad_norm": 0.3011747896671295,
      "learning_rate": 9.545032675245813e-05,
      "loss": 1.243,
      "step": 36
    },
    {
      "epoch": 0.5401459854014599,
      "grad_norm": 0.35093218088150024,
      "learning_rate": 9.509956150664796e-05,
      "loss": 1.3931,
      "step": 37
    },
    {
      "epoch": 0.5547445255474452,
      "grad_norm": 0.31631869077682495,
      "learning_rate": 9.473646649103818e-05,
      "loss": 1.3067,
      "step": 38
    },
    {
      "epoch": 0.5693430656934306,
      "grad_norm": 0.26237165927886963,
      "learning_rate": 9.43611409721806e-05,
      "loss": 1.4526,
      "step": 39
    },
    {
      "epoch": 0.583941605839416,
      "grad_norm": 0.29758620262145996,
      "learning_rate": 9.397368756032445e-05,
      "loss": 1.427,
      "step": 40
    },
    {
      "epoch": 0.5985401459854015,
      "grad_norm": 0.2432779222726822,
      "learning_rate": 9.357421218136386e-05,
      "loss": 1.2568,
      "step": 41
    },
    {
      "epoch": 0.6131386861313869,
      "grad_norm": 0.23281298577785492,
      "learning_rate": 9.316282404787871e-05,
      "loss": 1.2953,
      "step": 42
    },
    {
      "epoch": 0.6277372262773723,
      "grad_norm": 0.27693718671798706,
      "learning_rate": 9.273963562927695e-05,
      "loss": 1.6696,
      "step": 43
    },
    {
      "epoch": 0.6423357664233577,
      "grad_norm": 0.24839429557323456,
      "learning_rate": 9.230476262104677e-05,
      "loss": 1.5105,
      "step": 44
    },
    {
      "epoch": 0.656934306569343,
      "grad_norm": 0.2595592737197876,
      "learning_rate": 9.185832391312644e-05,
      "loss": 1.5493,
      "step": 45
    },
    {
      "epoch": 0.6715328467153284,
      "grad_norm": 0.2965114116668701,
      "learning_rate": 9.140044155740101e-05,
      "loss": 1.5033,
      "step": 46
    },
    {
      "epoch": 0.6861313868613139,
      "grad_norm": 0.30737271904945374,
      "learning_rate": 9.093124073433463e-05,
      "loss": 1.6587,
      "step": 47
    },
    {
      "epoch": 0.7007299270072993,
      "grad_norm": 0.326045423746109,
      "learning_rate": 9.045084971874738e-05,
      "loss": 1.3648,
      "step": 48
    },
    {
      "epoch": 0.7153284671532847,
      "grad_norm": 0.4043254852294922,
      "learning_rate": 8.995939984474624e-05,
      "loss": 1.4356,
      "step": 49
    },
    {
      "epoch": 0.7299270072992701,
      "grad_norm": 0.5170460343360901,
      "learning_rate": 8.945702546981969e-05,
      "loss": 1.5433,
      "step": 50
    },
    {
      "epoch": 0.7299270072992701,
      "eval_loss": 1.3493229150772095,
      "eval_runtime": 9.5607,
      "eval_samples_per_second": 12.133,
      "eval_steps_per_second": 3.033,
      "step": 50
    },
    {
      "epoch": 0.7445255474452555,
      "grad_norm": 0.8487922549247742,
      "learning_rate": 8.894386393810563e-05,
      "loss": 1.8449,
      "step": 51
    },
    {
      "epoch": 0.7591240875912408,
      "grad_norm": 0.19726397097110748,
      "learning_rate": 8.842005554284296e-05,
      "loss": 1.2332,
      "step": 52
    },
    {
      "epoch": 0.7737226277372263,
      "grad_norm": 0.19429665803909302,
      "learning_rate": 8.788574348801675e-05,
      "loss": 1.141,
      "step": 53
    },
    {
      "epoch": 0.7883211678832117,
      "grad_norm": 0.260591059923172,
      "learning_rate": 8.73410738492077e-05,
      "loss": 1.2771,
      "step": 54
    },
    {
      "epoch": 0.8029197080291971,
      "grad_norm": 0.24296972155570984,
      "learning_rate": 8.678619553365659e-05,
      "loss": 1.3942,
      "step": 55
    },
    {
      "epoch": 0.8175182481751825,
      "grad_norm": 0.24839623272418976,
      "learning_rate": 8.622126023955446e-05,
      "loss": 1.3292,
      "step": 56
    },
    {
      "epoch": 0.8321167883211679,
      "grad_norm": 0.23443883657455444,
      "learning_rate": 8.564642241456986e-05,
      "loss": 1.2817,
      "step": 57
    },
    {
      "epoch": 0.8467153284671532,
      "grad_norm": 0.2386799305677414,
      "learning_rate": 8.506183921362443e-05,
      "loss": 1.3752,
      "step": 58
    },
    {
      "epoch": 0.8613138686131386,
      "grad_norm": 0.30478376150131226,
      "learning_rate": 8.44676704559283e-05,
      "loss": 1.5058,
      "step": 59
    },
    {
      "epoch": 0.8759124087591241,
      "grad_norm": 0.3062061667442322,
      "learning_rate": 8.386407858128706e-05,
      "loss": 1.5203,
      "step": 60
    },
    {
      "epoch": 0.8905109489051095,
      "grad_norm": 0.28531453013420105,
      "learning_rate": 8.32512286056924e-05,
      "loss": 1.3249,
      "step": 61
    },
    {
      "epoch": 0.9051094890510949,
      "grad_norm": 0.2942626476287842,
      "learning_rate": 8.262928807620843e-05,
      "loss": 1.3588,
      "step": 62
    },
    {
      "epoch": 0.9197080291970803,
      "grad_norm": 0.2991859018802643,
      "learning_rate": 8.199842702516583e-05,
      "loss": 1.4485,
      "step": 63
    },
    {
      "epoch": 0.9343065693430657,
      "grad_norm": 0.32852524518966675,
      "learning_rate": 8.135881792367686e-05,
      "loss": 1.6162,
      "step": 64
    },
    {
      "epoch": 0.948905109489051,
      "grad_norm": 0.3827306628227234,
      "learning_rate": 8.07106356344834e-05,
      "loss": 1.5128,
      "step": 65
    },
    {
      "epoch": 0.9635036496350365,
      "grad_norm": 0.4586751163005829,
      "learning_rate": 8.005405736415126e-05,
      "loss": 1.4783,
      "step": 66
    },
    {
      "epoch": 0.9781021897810219,
      "grad_norm": 0.658205509185791,
      "learning_rate": 7.938926261462366e-05,
      "loss": 1.2937,
      "step": 67
    },
    {
      "epoch": 0.9927007299270073,
      "grad_norm": 0.8376323580741882,
      "learning_rate": 7.871643313414718e-05,
      "loss": 1.502,
      "step": 68
    },
    {
      "epoch": 1.0072992700729928,
      "grad_norm": 0.39524346590042114,
      "learning_rate": 7.803575286758364e-05,
      "loss": 1.9244,
      "step": 69
    },
    {
      "epoch": 1.0218978102189782,
      "grad_norm": 0.20059272646903992,
      "learning_rate": 7.734740790612136e-05,
      "loss": 1.2563,
      "step": 70
    },
    {
      "epoch": 1.0364963503649636,
      "grad_norm": 0.18347716331481934,
      "learning_rate": 7.66515864363997e-05,
      "loss": 1.2003,
      "step": 71
    },
    {
      "epoch": 1.051094890510949,
      "grad_norm": 0.22512517869472504,
      "learning_rate": 7.594847868906076e-05,
      "loss": 1.31,
      "step": 72
    },
    {
      "epoch": 1.0656934306569343,
      "grad_norm": 0.23011034727096558,
      "learning_rate": 7.52382768867422e-05,
      "loss": 1.2662,
      "step": 73
    },
    {
      "epoch": 1.0802919708029197,
      "grad_norm": 0.2077198177576065,
      "learning_rate": 7.452117519152542e-05,
      "loss": 1.2224,
      "step": 74
    },
    {
      "epoch": 1.094890510948905,
      "grad_norm": 0.20596177875995636,
      "learning_rate": 7.379736965185368e-05,
      "loss": 1.1483,
      "step": 75
    },
    {
      "epoch": 1.1094890510948905,
      "grad_norm": 0.22836129367351532,
      "learning_rate": 7.30670581489344e-05,
      "loss": 1.1611,
      "step": 76
    },
    {
      "epoch": 1.1240875912408759,
      "grad_norm": 0.2444555014371872,
      "learning_rate": 7.233044034264034e-05,
      "loss": 1.3633,
      "step": 77
    },
    {
      "epoch": 1.1386861313868613,
      "grad_norm": 0.2522647976875305,
      "learning_rate": 7.158771761692464e-05,
      "loss": 1.474,
      "step": 78
    },
    {
      "epoch": 1.1532846715328466,
      "grad_norm": 0.27095985412597656,
      "learning_rate": 7.083909302476453e-05,
      "loss": 1.4392,
      "step": 79
    },
    {
      "epoch": 1.167883211678832,
      "grad_norm": 0.29701387882232666,
      "learning_rate": 7.008477123264848e-05,
      "loss": 1.3007,
      "step": 80
    },
    {
      "epoch": 1.1824817518248176,
      "grad_norm": 0.30774107575416565,
      "learning_rate": 6.932495846462261e-05,
      "loss": 1.2798,
      "step": 81
    },
    {
      "epoch": 1.197080291970803,
      "grad_norm": 0.3699590563774109,
      "learning_rate": 6.855986244591104e-05,
      "loss": 1.5073,
      "step": 82
    },
    {
      "epoch": 1.2116788321167884,
      "grad_norm": 0.40869346261024475,
      "learning_rate": 6.778969234612584e-05,
      "loss": 1.2186,
      "step": 83
    },
    {
      "epoch": 1.2262773722627738,
      "grad_norm": 0.48988431692123413,
      "learning_rate": 6.701465872208216e-05,
      "loss": 1.1217,
      "step": 84
    },
    {
      "epoch": 1.2408759124087592,
      "grad_norm": 0.650515079498291,
      "learning_rate": 6.623497346023418e-05,
      "loss": 1.0204,
      "step": 85
    },
    {
      "epoch": 1.2554744525547445,
      "grad_norm": 0.511817455291748,
      "learning_rate": 6.545084971874738e-05,
      "loss": 1.0432,
      "step": 86
    },
    {
      "epoch": 1.27007299270073,
      "grad_norm": 0.26496750116348267,
      "learning_rate": 6.466250186922325e-05,
      "loss": 1.1708,
      "step": 87
    },
    {
      "epoch": 1.2846715328467153,
      "grad_norm": 0.33906814455986023,
      "learning_rate": 6.387014543809223e-05,
      "loss": 1.314,
      "step": 88
    },
    {
      "epoch": 1.2992700729927007,
      "grad_norm": 0.3010120093822479,
      "learning_rate": 6.307399704769099e-05,
      "loss": 1.2973,
      "step": 89
    },
    {
      "epoch": 1.313868613138686,
      "grad_norm": 0.29846590757369995,
      "learning_rate": 6.227427435703997e-05,
      "loss": 1.3966,
      "step": 90
    },
    {
      "epoch": 1.3284671532846715,
      "grad_norm": 0.27580320835113525,
      "learning_rate": 6.147119600233758e-05,
      "loss": 1.3393,
      "step": 91
    },
    {
      "epoch": 1.343065693430657,
      "grad_norm": 0.26557210087776184,
      "learning_rate": 6.066498153718735e-05,
      "loss": 1.0822,
      "step": 92
    },
    {
      "epoch": 1.3576642335766422,
      "grad_norm": 0.295459121465683,
      "learning_rate": 5.985585137257401e-05,
      "loss": 1.2982,
      "step": 93
    },
    {
      "epoch": 1.3722627737226278,
      "grad_norm": 0.34873393177986145,
      "learning_rate": 5.90440267166055e-05,
      "loss": 1.3999,
      "step": 94
    },
    {
      "epoch": 1.3868613138686132,
      "grad_norm": 0.364439457654953,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 1.3001,
      "step": 95
    },
    {
      "epoch": 1.4014598540145986,
      "grad_norm": 0.3429001569747925,
      "learning_rate": 5.74131823855921e-05,
      "loss": 1.2162,
      "step": 96
    },
    {
      "epoch": 1.416058394160584,
      "grad_norm": 0.3794465661048889,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 1.3517,
      "step": 97
    },
    {
      "epoch": 1.4306569343065694,
      "grad_norm": 0.42539718747138977,
      "learning_rate": 5.577423184847932e-05,
      "loss": 1.3532,
      "step": 98
    },
    {
      "epoch": 1.4452554744525548,
      "grad_norm": 0.45871689915657043,
      "learning_rate": 5.495227651252315e-05,
      "loss": 1.4441,
      "step": 99
    },
    {
      "epoch": 1.4598540145985401,
      "grad_norm": 0.4918805956840515,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 1.2127,
      "step": 100
    },
    {
      "epoch": 1.4598540145985401,
      "eval_loss": 1.3407422304153442,
      "eval_runtime": 9.5733,
      "eval_samples_per_second": 12.117,
      "eval_steps_per_second": 3.029,
      "step": 100
    },
    {
      "epoch": 1.4744525547445255,
      "grad_norm": 0.6055482625961304,
      "learning_rate": 5.330452921628497e-05,
      "loss": 1.0081,
      "step": 101
    },
    {
      "epoch": 1.489051094890511,
      "grad_norm": 0.8676618933677673,
      "learning_rate": 5.247918773366112e-05,
      "loss": 1.1944,
      "step": 102
    },
    {
      "epoch": 1.5036496350364965,
      "grad_norm": 0.6699618697166443,
      "learning_rate": 5.165316846586541e-05,
      "loss": 0.998,
      "step": 103
    },
    {
      "epoch": 1.5182481751824817,
      "grad_norm": 0.270437091588974,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 1.1106,
      "step": 104
    },
    {
      "epoch": 1.5328467153284673,
      "grad_norm": 0.406057208776474,
      "learning_rate": 5e-05,
      "loss": 1.3345,
      "step": 105
    },
    {
      "epoch": 1.5474452554744524,
      "grad_norm": 0.3463612198829651,
      "learning_rate": 4.917330276168208e-05,
      "loss": 1.2775,
      "step": 106
    },
    {
      "epoch": 1.562043795620438,
      "grad_norm": 0.3449850380420685,
      "learning_rate": 4.834683153413459e-05,
      "loss": 1.2543,
      "step": 107
    },
    {
      "epoch": 1.5766423357664232,
      "grad_norm": 0.3121253550052643,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 1.2477,
      "step": 108
    },
    {
      "epoch": 1.5912408759124088,
      "grad_norm": 0.3038662374019623,
      "learning_rate": 4.669547078371504e-05,
      "loss": 1.2077,
      "step": 109
    },
    {
      "epoch": 1.6058394160583942,
      "grad_norm": 0.28545626997947693,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 1.1897,
      "step": 110
    },
    {
      "epoch": 1.6204379562043796,
      "grad_norm": 0.3149450719356537,
      "learning_rate": 4.504772348747687e-05,
      "loss": 1.3912,
      "step": 111
    },
    {
      "epoch": 1.635036496350365,
      "grad_norm": 0.3550225496292114,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 1.4719,
      "step": 112
    },
    {
      "epoch": 1.6496350364963503,
      "grad_norm": 0.3512927293777466,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 1.2396,
      "step": 113
    },
    {
      "epoch": 1.6642335766423357,
      "grad_norm": 0.37811553478240967,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 1.2828,
      "step": 114
    },
    {
      "epoch": 1.6788321167883211,
      "grad_norm": 0.3926678001880646,
      "learning_rate": 4.17702704859633e-05,
      "loss": 1.4642,
      "step": 115
    },
    {
      "epoch": 1.6934306569343067,
      "grad_norm": 0.44688135385513306,
      "learning_rate": 4.095597328339452e-05,
      "loss": 1.4418,
      "step": 116
    },
    {
      "epoch": 1.7080291970802919,
      "grad_norm": 0.5358821749687195,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 1.2599,
      "step": 117
    },
    {
      "epoch": 1.7226277372262775,
      "grad_norm": 0.6397238373756409,
      "learning_rate": 3.933501846281267e-05,
      "loss": 1.1426,
      "step": 118
    },
    {
      "epoch": 1.7372262773722627,
      "grad_norm": 0.8133037686347961,
      "learning_rate": 3.852880399766243e-05,
      "loss": 1.0914,
      "step": 119
    },
    {
      "epoch": 1.7518248175182483,
      "grad_norm": 0.4947717785835266,
      "learning_rate": 3.772572564296005e-05,
      "loss": 0.7536,
      "step": 120
    },
    {
      "epoch": 1.7664233576642334,
      "grad_norm": 0.4615175724029541,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 1.1761,
      "step": 121
    },
    {
      "epoch": 1.781021897810219,
      "grad_norm": 0.29529380798339844,
      "learning_rate": 3.612985456190778e-05,
      "loss": 1.2687,
      "step": 122
    },
    {
      "epoch": 1.7956204379562044,
      "grad_norm": 0.31780120730400085,
      "learning_rate": 3.533749813077677e-05,
      "loss": 1.342,
      "step": 123
    },
    {
      "epoch": 1.8102189781021898,
      "grad_norm": 0.29632455110549927,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 1.2432,
      "step": 124
    },
    {
      "epoch": 1.8248175182481752,
      "grad_norm": 0.29321733117103577,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 1.1672,
      "step": 125
    },
    {
      "epoch": 1.8394160583941606,
      "grad_norm": 0.2935655117034912,
      "learning_rate": 3.298534127791785e-05,
      "loss": 1.2178,
      "step": 126
    },
    {
      "epoch": 1.854014598540146,
      "grad_norm": 0.31564828753471375,
      "learning_rate": 3.221030765387417e-05,
      "loss": 1.3567,
      "step": 127
    },
    {
      "epoch": 1.8686131386861313,
      "grad_norm": 0.3216329514980316,
      "learning_rate": 3.144013755408895e-05,
      "loss": 1.5563,
      "step": 128
    },
    {
      "epoch": 1.883211678832117,
      "grad_norm": 0.34779438376426697,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 1.4827,
      "step": 129
    },
    {
      "epoch": 1.897810218978102,
      "grad_norm": 0.33754420280456543,
      "learning_rate": 2.991522876735154e-05,
      "loss": 1.1582,
      "step": 130
    },
    {
      "epoch": 1.9124087591240877,
      "grad_norm": 0.364000529050827,
      "learning_rate": 2.916090697523549e-05,
      "loss": 1.2544,
      "step": 131
    },
    {
      "epoch": 1.9270072992700729,
      "grad_norm": 0.391092449426651,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 1.3224,
      "step": 132
    },
    {
      "epoch": 1.9416058394160585,
      "grad_norm": 0.45297062397003174,
      "learning_rate": 2.766955965735968e-05,
      "loss": 1.3419,
      "step": 133
    },
    {
      "epoch": 1.9562043795620438,
      "grad_norm": 0.5484812259674072,
      "learning_rate": 2.693294185106562e-05,
      "loss": 1.3331,
      "step": 134
    },
    {
      "epoch": 1.9708029197080292,
      "grad_norm": 0.6513543128967285,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 1.2833,
      "step": 135
    },
    {
      "epoch": 1.9854014598540146,
      "grad_norm": 0.977857768535614,
      "learning_rate": 2.547882480847461e-05,
      "loss": 1.2161,
      "step": 136
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7950093150138855,
      "learning_rate": 2.476172311325783e-05,
      "loss": 1.8114,
      "step": 137
    },
    {
      "epoch": 2.0145985401459856,
      "grad_norm": 0.2949065864086151,
      "learning_rate": 2.405152131093926e-05,
      "loss": 1.0359,
      "step": 138
    },
    {
      "epoch": 2.0291970802919708,
      "grad_norm": 0.2553476095199585,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 1.2407,
      "step": 139
    },
    {
      "epoch": 2.0437956204379564,
      "grad_norm": 0.269107460975647,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 1.3298,
      "step": 140
    },
    {
      "epoch": 2.0583941605839415,
      "grad_norm": 0.2857801914215088,
      "learning_rate": 2.196424713241637e-05,
      "loss": 1.2393,
      "step": 141
    },
    {
      "epoch": 2.072992700729927,
      "grad_norm": 0.29582342505455017,
      "learning_rate": 2.128356686585282e-05,
      "loss": 1.1772,
      "step": 142
    },
    {
      "epoch": 2.0875912408759123,
      "grad_norm": 0.2814674973487854,
      "learning_rate": 2.061073738537635e-05,
      "loss": 1.1288,
      "step": 143
    },
    {
      "epoch": 2.102189781021898,
      "grad_norm": 0.3022535741329193,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 1.214,
      "step": 144
    },
    {
      "epoch": 2.116788321167883,
      "grad_norm": 0.3051672875881195,
      "learning_rate": 1.928936436551661e-05,
      "loss": 1.1678,
      "step": 145
    },
    {
      "epoch": 2.1313868613138687,
      "grad_norm": 0.3158007264137268,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 1.2408,
      "step": 146
    },
    {
      "epoch": 2.145985401459854,
      "grad_norm": 0.3425518274307251,
      "learning_rate": 1.800157297483417e-05,
      "loss": 1.2748,
      "step": 147
    },
    {
      "epoch": 2.1605839416058394,
      "grad_norm": 0.34279200434684753,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 1.0408,
      "step": 148
    },
    {
      "epoch": 2.1751824817518246,
      "grad_norm": 0.40088874101638794,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 1.3625,
      "step": 149
    },
    {
      "epoch": 2.18978102189781,
      "grad_norm": 0.41201409697532654,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 1.3051,
      "step": 150
    },
    {
      "epoch": 2.18978102189781,
      "eval_loss": 1.3498927354812622,
      "eval_runtime": 9.8402,
      "eval_samples_per_second": 11.788,
      "eval_steps_per_second": 2.947,
      "step": 150
    },
    {
      "epoch": 2.204379562043796,
      "grad_norm": 0.47057774662971497,
      "learning_rate": 1.553232954407171e-05,
      "loss": 1.3305,
      "step": 151
    },
    {
      "epoch": 2.218978102189781,
      "grad_norm": 0.6118239760398865,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 1.0581,
      "step": 152
    },
    {
      "epoch": 2.2335766423357666,
      "grad_norm": 0.6453813910484314,
      "learning_rate": 1.435357758543015e-05,
      "loss": 0.7983,
      "step": 153
    },
    {
      "epoch": 2.2481751824817517,
      "grad_norm": 0.9582388401031494,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 0.7787,
      "step": 154
    },
    {
      "epoch": 2.2627737226277373,
      "grad_norm": 0.25991177558898926,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 1.1092,
      "step": 155
    },
    {
      "epoch": 2.2773722627737225,
      "grad_norm": 0.3064647912979126,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 1.2848,
      "step": 156
    },
    {
      "epoch": 2.291970802919708,
      "grad_norm": 0.31570515036582947,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 1.2677,
      "step": 157
    },
    {
      "epoch": 2.3065693430656933,
      "grad_norm": 0.3369019031524658,
      "learning_rate": 1.157994445715706e-05,
      "loss": 1.2989,
      "step": 158
    },
    {
      "epoch": 2.321167883211679,
      "grad_norm": 0.35968828201293945,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 1.228,
      "step": 159
    },
    {
      "epoch": 2.335766423357664,
      "grad_norm": 0.3660455048084259,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 1.1944,
      "step": 160
    },
    {
      "epoch": 2.3503649635036497,
      "grad_norm": 0.3260854482650757,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 1.0763,
      "step": 161
    },
    {
      "epoch": 2.3649635036496353,
      "grad_norm": 0.346077024936676,
      "learning_rate": 9.549150281252633e-06,
      "loss": 1.2568,
      "step": 162
    },
    {
      "epoch": 2.3795620437956204,
      "grad_norm": 0.3681202530860901,
      "learning_rate": 9.068759265665384e-06,
      "loss": 1.4127,
      "step": 163
    },
    {
      "epoch": 2.394160583941606,
      "grad_norm": 0.3981098532676697,
      "learning_rate": 8.599558442598998e-06,
      "loss": 1.1965,
      "step": 164
    },
    {
      "epoch": 2.408759124087591,
      "grad_norm": 0.4082994759082794,
      "learning_rate": 8.141676086873572e-06,
      "loss": 1.226,
      "step": 165
    },
    {
      "epoch": 2.423357664233577,
      "grad_norm": 0.4200907051563263,
      "learning_rate": 7.695237378953223e-06,
      "loss": 1.2993,
      "step": 166
    },
    {
      "epoch": 2.437956204379562,
      "grad_norm": 0.45979371666908264,
      "learning_rate": 7.260364370723044e-06,
      "loss": 1.1564,
      "step": 167
    },
    {
      "epoch": 2.4525547445255476,
      "grad_norm": 0.5479443073272705,
      "learning_rate": 6.837175952121306e-06,
      "loss": 1.2589,
      "step": 168
    },
    {
      "epoch": 2.4671532846715327,
      "grad_norm": 0.5844144225120544,
      "learning_rate": 6.425787818636131e-06,
      "loss": 1.0876,
      "step": 169
    },
    {
      "epoch": 2.4817518248175183,
      "grad_norm": 0.6877148747444153,
      "learning_rate": 6.026312439675552e-06,
      "loss": 0.7598,
      "step": 170
    },
    {
      "epoch": 2.4963503649635035,
      "grad_norm": 0.9213687777519226,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 0.6701,
      "step": 171
    },
    {
      "epoch": 2.510948905109489,
      "grad_norm": 0.24933631718158722,
      "learning_rate": 5.263533508961827e-06,
      "loss": 1.1072,
      "step": 172
    },
    {
      "epoch": 2.5255474452554747,
      "grad_norm": 0.22774094343185425,
      "learning_rate": 4.900438493352055e-06,
      "loss": 0.9733,
      "step": 173
    },
    {
      "epoch": 2.54014598540146,
      "grad_norm": 0.27643153071403503,
      "learning_rate": 4.549673247541875e-06,
      "loss": 1.1547,
      "step": 174
    },
    {
      "epoch": 2.554744525547445,
      "grad_norm": 0.29355645179748535,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 1.1371,
      "step": 175
    },
    {
      "epoch": 2.5693430656934306,
      "grad_norm": 0.3061830699443817,
      "learning_rate": 3.885512251130763e-06,
      "loss": 1.1845,
      "step": 176
    },
    {
      "epoch": 2.5839416058394162,
      "grad_norm": 0.3109261989593506,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 1.1253,
      "step": 177
    },
    {
      "epoch": 2.5985401459854014,
      "grad_norm": 0.31706884503364563,
      "learning_rate": 3.271776770026963e-06,
      "loss": 1.1852,
      "step": 178
    },
    {
      "epoch": 2.613138686131387,
      "grad_norm": 0.3444359004497528,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 1.217,
      "step": 179
    },
    {
      "epoch": 2.627737226277372,
      "grad_norm": 0.35367968678474426,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 1.4271,
      "step": 180
    },
    {
      "epoch": 2.6423357664233578,
      "grad_norm": 0.36658063530921936,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 1.2979,
      "step": 181
    },
    {
      "epoch": 2.656934306569343,
      "grad_norm": 0.3930165469646454,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 1.0503,
      "step": 182
    },
    {
      "epoch": 2.6715328467153285,
      "grad_norm": 0.4118279218673706,
      "learning_rate": 1.962316193157593e-06,
      "loss": 1.1624,
      "step": 183
    },
    {
      "epoch": 2.686131386861314,
      "grad_norm": 0.4913877248764038,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 1.1907,
      "step": 184
    },
    {
      "epoch": 2.7007299270072993,
      "grad_norm": 0.5602124333381653,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 1.0478,
      "step": 185
    },
    {
      "epoch": 2.7153284671532845,
      "grad_norm": 0.6251142024993896,
      "learning_rate": 1.333670137599713e-06,
      "loss": 0.8332,
      "step": 186
    },
    {
      "epoch": 2.72992700729927,
      "grad_norm": 0.7298576831817627,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 0.7257,
      "step": 187
    },
    {
      "epoch": 2.7445255474452557,
      "grad_norm": 0.9666867256164551,
      "learning_rate": 9.810017062595322e-07,
      "loss": 0.9913,
      "step": 188
    },
    {
      "epoch": 2.759124087591241,
      "grad_norm": 0.22270461916923523,
      "learning_rate": 8.247462563808817e-07,
      "loss": 0.9232,
      "step": 189
    },
    {
      "epoch": 2.7737226277372264,
      "grad_norm": 0.2743498980998993,
      "learning_rate": 6.819348298638839e-07,
      "loss": 1.248,
      "step": 190
    },
    {
      "epoch": 2.7883211678832116,
      "grad_norm": 0.29309284687042236,
      "learning_rate": 5.526064699265753e-07,
      "loss": 1.217,
      "step": 191
    },
    {
      "epoch": 2.802919708029197,
      "grad_norm": 0.32170671224594116,
      "learning_rate": 4.367965336512403e-07,
      "loss": 1.2521,
      "step": 192
    },
    {
      "epoch": 2.8175182481751824,
      "grad_norm": 0.3107758164405823,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 1.1975,
      "step": 193
    },
    {
      "epoch": 2.832116788321168,
      "grad_norm": 0.3339341878890991,
      "learning_rate": 2.458548727494292e-07,
      "loss": 1.1492,
      "step": 194
    },
    {
      "epoch": 2.846715328467153,
      "grad_norm": 0.3442275822162628,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 1.2463,
      "step": 195
    },
    {
      "epoch": 2.8613138686131387,
      "grad_norm": 0.3328962028026581,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 1.2965,
      "step": 196
    },
    {
      "epoch": 2.875912408759124,
      "grad_norm": 0.3658321499824524,
      "learning_rate": 6.150154258476315e-08,
      "loss": 1.3809,
      "step": 197
    },
    {
      "epoch": 2.8905109489051095,
      "grad_norm": 0.3719880282878876,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 1.2631,
      "step": 198
    },
    {
      "epoch": 2.905109489051095,
      "grad_norm": 0.3807631731033325,
      "learning_rate": 6.834750376549792e-09,
      "loss": 1.2368,
      "step": 199
    },
    {
      "epoch": 2.9197080291970803,
      "grad_norm": 0.4068572521209717,
      "learning_rate": 0.0,
      "loss": 1.1797,
      "step": 200
    },
    {
      "epoch": 2.9197080291970803,
      "eval_loss": 1.3501931428909302,
      "eval_runtime": 9.84,
      "eval_samples_per_second": 11.789,
      "eval_steps_per_second": 2.947,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 2
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.105049416617165e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}