{
  "best_metric": 0.12807011604309082,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.27164685908319186,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0013582342954159593,
      "grad_norm": 1.8889578580856323,
      "learning_rate": 1e-05,
      "loss": 2.2076,
      "step": 1
    },
    {
      "epoch": 0.0013582342954159593,
      "eval_loss": 0.9938458800315857,
      "eval_runtime": 66.1888,
      "eval_samples_per_second": 18.734,
      "eval_steps_per_second": 4.684,
      "step": 1
    },
    {
      "epoch": 0.0027164685908319186,
      "grad_norm": 2.3116564750671387,
      "learning_rate": 2e-05,
      "loss": 3.322,
      "step": 2
    },
    {
      "epoch": 0.0040747028862478775,
      "grad_norm": 3.3473684787750244,
      "learning_rate": 3e-05,
      "loss": 3.6327,
      "step": 3
    },
    {
      "epoch": 0.005432937181663837,
      "grad_norm": 3.8035199642181396,
      "learning_rate": 4e-05,
      "loss": 4.1776,
      "step": 4
    },
    {
      "epoch": 0.006791171477079796,
      "grad_norm": 4.566415309906006,
      "learning_rate": 5e-05,
      "loss": 4.4245,
      "step": 5
    },
    {
      "epoch": 0.008149405772495755,
      "grad_norm": 4.2230682373046875,
      "learning_rate": 6e-05,
      "loss": 4.3489,
      "step": 6
    },
    {
      "epoch": 0.009507640067911714,
      "grad_norm": 3.2478137016296387,
      "learning_rate": 7e-05,
      "loss": 2.3336,
      "step": 7
    },
    {
      "epoch": 0.010865874363327675,
      "grad_norm": 3.2353904247283936,
      "learning_rate": 8e-05,
      "loss": 2.3194,
      "step": 8
    },
    {
      "epoch": 0.012224108658743633,
      "grad_norm": 4.434475421905518,
      "learning_rate": 9e-05,
      "loss": 1.7946,
      "step": 9
    },
    {
      "epoch": 0.013582342954159592,
      "grad_norm": 2.9538373947143555,
      "learning_rate": 0.0001,
      "loss": 1.4935,
      "step": 10
    },
    {
      "epoch": 0.014940577249575551,
      "grad_norm": 3.012892961502075,
      "learning_rate": 9.999316524962345e-05,
      "loss": 1.4069,
      "step": 11
    },
    {
      "epoch": 0.01629881154499151,
      "grad_norm": 3.1674129962921143,
      "learning_rate": 9.997266286704631e-05,
      "loss": 0.967,
      "step": 12
    },
    {
      "epoch": 0.01765704584040747,
      "grad_norm": 3.2145333290100098,
      "learning_rate": 9.993849845741524e-05,
      "loss": 1.0964,
      "step": 13
    },
    {
      "epoch": 0.019015280135823428,
      "grad_norm": 4.188545227050781,
      "learning_rate": 9.989068136093873e-05,
      "loss": 1.8467,
      "step": 14
    },
    {
      "epoch": 0.02037351443123939,
      "grad_norm": 3.6614139080047607,
      "learning_rate": 9.98292246503335e-05,
      "loss": 1.4485,
      "step": 15
    },
    {
      "epoch": 0.02173174872665535,
      "grad_norm": 2.5848684310913086,
      "learning_rate": 9.975414512725057e-05,
      "loss": 1.1819,
      "step": 16
    },
    {
      "epoch": 0.023089983022071308,
      "grad_norm": 2.549119234085083,
      "learning_rate": 9.966546331768191e-05,
      "loss": 1.3204,
      "step": 17
    },
    {
      "epoch": 0.024448217317487267,
      "grad_norm": 1.8019766807556152,
      "learning_rate": 9.956320346634876e-05,
      "loss": 1.04,
      "step": 18
    },
    {
      "epoch": 0.025806451612903226,
      "grad_norm": 2.1322357654571533,
      "learning_rate": 9.944739353007344e-05,
      "loss": 0.9348,
      "step": 19
    },
    {
      "epoch": 0.027164685908319185,
      "grad_norm": 1.5636383295059204,
      "learning_rate": 9.931806517013612e-05,
      "loss": 0.7803,
      "step": 20
    },
    {
      "epoch": 0.028522920203735144,
      "grad_norm": 1.5372105836868286,
      "learning_rate": 9.917525374361912e-05,
      "loss": 0.9237,
      "step": 21
    },
    {
      "epoch": 0.029881154499151102,
      "grad_norm": 1.4937751293182373,
      "learning_rate": 9.901899829374047e-05,
      "loss": 0.7272,
      "step": 22
    },
    {
      "epoch": 0.03123938879456706,
      "grad_norm": 1.1647422313690186,
      "learning_rate": 9.884934153917997e-05,
      "loss": 0.5155,
      "step": 23
    },
    {
      "epoch": 0.03259762308998302,
      "grad_norm": 3.998044490814209,
      "learning_rate": 9.86663298624003e-05,
      "loss": 0.6064,
      "step": 24
    },
    {
      "epoch": 0.03395585738539898,
      "grad_norm": 1.3711620569229126,
      "learning_rate": 9.847001329696653e-05,
      "loss": 0.6581,
      "step": 25
    },
    {
      "epoch": 0.03531409168081494,
      "grad_norm": 1.5123319625854492,
      "learning_rate": 9.826044551386744e-05,
      "loss": 0.6528,
      "step": 26
    },
    {
      "epoch": 0.0366723259762309,
      "grad_norm": 1.4044246673583984,
      "learning_rate": 9.803768380684242e-05,
      "loss": 0.8442,
      "step": 27
    },
    {
      "epoch": 0.038030560271646856,
      "grad_norm": 1.2004684209823608,
      "learning_rate": 9.780178907671789e-05,
      "loss": 0.4215,
      "step": 28
    },
    {
      "epoch": 0.03938879456706282,
      "grad_norm": 1.6459702253341675,
      "learning_rate": 9.755282581475769e-05,
      "loss": 0.7792,
      "step": 29
    },
    {
      "epoch": 0.04074702886247878,
      "grad_norm": 1.5842430591583252,
      "learning_rate": 9.729086208503174e-05,
      "loss": 0.6471,
      "step": 30
    },
    {
      "epoch": 0.042105263157894736,
      "grad_norm": 1.0356532335281372,
      "learning_rate": 9.701596950580806e-05,
      "loss": 0.3588,
      "step": 31
    },
    {
      "epoch": 0.0434634974533107,
      "grad_norm": 1.0359015464782715,
      "learning_rate": 9.672822322997305e-05,
      "loss": 0.5072,
      "step": 32
    },
    {
      "epoch": 0.044821731748726654,
      "grad_norm": 1.0919171571731567,
      "learning_rate": 9.642770192448536e-05,
      "loss": 0.4494,
      "step": 33
    },
    {
      "epoch": 0.046179966044142616,
      "grad_norm": 1.2643227577209473,
      "learning_rate": 9.611448774886924e-05,
      "loss": 0.5515,
      "step": 34
    },
    {
      "epoch": 0.04753820033955857,
      "grad_norm": 1.2034884691238403,
      "learning_rate": 9.578866633275288e-05,
      "loss": 0.4593,
      "step": 35
    },
    {
      "epoch": 0.048896434634974534,
      "grad_norm": 5.325469017028809,
      "learning_rate": 9.545032675245813e-05,
      "loss": 0.3904,
      "step": 36
    },
    {
      "epoch": 0.05025466893039049,
      "grad_norm": 1.0181546211242676,
      "learning_rate": 9.509956150664796e-05,
      "loss": 0.3721,
      "step": 37
    },
    {
      "epoch": 0.05161290322580645,
      "grad_norm": 1.0899913311004639,
      "learning_rate": 9.473646649103818e-05,
      "loss": 0.3608,
      "step": 38
    },
    {
      "epoch": 0.052971137521222414,
      "grad_norm": 1.1829785108566284,
      "learning_rate": 9.43611409721806e-05,
      "loss": 0.4572,
      "step": 39
    },
    {
      "epoch": 0.05432937181663837,
      "grad_norm": 0.8924344778060913,
      "learning_rate": 9.397368756032445e-05,
      "loss": 0.3511,
      "step": 40
    },
    {
      "epoch": 0.05568760611205433,
      "grad_norm": 1.0210082530975342,
      "learning_rate": 9.357421218136386e-05,
      "loss": 0.3184,
      "step": 41
    },
    {
      "epoch": 0.05704584040747029,
      "grad_norm": 1.2711970806121826,
      "learning_rate": 9.316282404787871e-05,
      "loss": 0.4089,
      "step": 42
    },
    {
      "epoch": 0.05840407470288625,
      "grad_norm": 1.984538197517395,
      "learning_rate": 9.273963562927695e-05,
      "loss": 0.4726,
      "step": 43
    },
    {
      "epoch": 0.059762308998302205,
      "grad_norm": 1.6680279970169067,
      "learning_rate": 9.230476262104677e-05,
      "loss": 0.3792,
      "step": 44
    },
    {
      "epoch": 0.06112054329371817,
      "grad_norm": 1.7497315406799316,
      "learning_rate": 9.185832391312644e-05,
      "loss": 0.3635,
      "step": 45
    },
    {
      "epoch": 0.06247877758913412,
      "grad_norm": 1.80753755569458,
      "learning_rate": 9.140044155740101e-05,
      "loss": 0.4891,
      "step": 46
    },
    {
      "epoch": 0.06383701188455009,
      "grad_norm": 1.4111545085906982,
      "learning_rate": 9.093124073433463e-05,
      "loss": 0.4598,
      "step": 47
    },
    {
      "epoch": 0.06519524617996604,
      "grad_norm": 1.5910176038742065,
      "learning_rate": 9.045084971874738e-05,
      "loss": 0.3751,
      "step": 48
    },
    {
      "epoch": 0.06655348047538201,
      "grad_norm": 1.1267386674880981,
      "learning_rate": 8.995939984474624e-05,
      "loss": 0.2321,
      "step": 49
    },
    {
      "epoch": 0.06791171477079797,
      "grad_norm": 1.0913166999816895,
      "learning_rate": 8.945702546981969e-05,
      "loss": 0.2783,
      "step": 50
    },
    {
      "epoch": 0.06791171477079797,
      "eval_loss": 0.2754749059677124,
      "eval_runtime": 67.1086,
      "eval_samples_per_second": 18.478,
      "eval_steps_per_second": 4.619,
      "step": 50
    },
    {
      "epoch": 0.06926994906621392,
      "grad_norm": 3.659601926803589,
      "learning_rate": 8.894386393810563e-05,
      "loss": 2.3466,
      "step": 51
    },
    {
      "epoch": 0.07062818336162988,
      "grad_norm": 4.295755863189697,
      "learning_rate": 8.842005554284296e-05,
      "loss": 3.091,
      "step": 52
    },
    {
      "epoch": 0.07198641765704585,
      "grad_norm": 3.616570472717285,
      "learning_rate": 8.788574348801675e-05,
      "loss": 3.2909,
      "step": 53
    },
    {
      "epoch": 0.0733446519524618,
      "grad_norm": 3.0673322677612305,
      "learning_rate": 8.73410738492077e-05,
      "loss": 3.7464,
      "step": 54
    },
    {
      "epoch": 0.07470288624787776,
      "grad_norm": 2.9065418243408203,
      "learning_rate": 8.678619553365659e-05,
      "loss": 2.8627,
      "step": 55
    },
    {
      "epoch": 0.07606112054329371,
      "grad_norm": 3.1884777545928955,
      "learning_rate": 8.622126023955446e-05,
      "loss": 2.777,
      "step": 56
    },
    {
      "epoch": 0.07741935483870968,
      "grad_norm": 4.920656681060791,
      "learning_rate": 8.564642241456986e-05,
      "loss": 2.6136,
      "step": 57
    },
    {
      "epoch": 0.07877758913412564,
      "grad_norm": 1.7224739789962769,
      "learning_rate": 8.506183921362443e-05,
      "loss": 1.0914,
      "step": 58
    },
    {
      "epoch": 0.08013582342954159,
      "grad_norm": 1.2613626718521118,
      "learning_rate": 8.44676704559283e-05,
      "loss": 0.9451,
      "step": 59
    },
    {
      "epoch": 0.08149405772495756,
      "grad_norm": 2.074201822280884,
      "learning_rate": 8.386407858128706e-05,
      "loss": 0.9149,
      "step": 60
    },
    {
      "epoch": 0.08285229202037352,
      "grad_norm": 2.2877724170684814,
      "learning_rate": 8.32512286056924e-05,
      "loss": 1.2911,
      "step": 61
    },
    {
      "epoch": 0.08421052631578947,
      "grad_norm": 1.7160199880599976,
      "learning_rate": 8.262928807620843e-05,
      "loss": 0.6396,
      "step": 62
    },
    {
      "epoch": 0.08556876061120543,
      "grad_norm": 1.397225022315979,
      "learning_rate": 8.199842702516583e-05,
      "loss": 0.6306,
      "step": 63
    },
    {
      "epoch": 0.0869269949066214,
      "grad_norm": 1.5045193433761597,
      "learning_rate": 8.135881792367686e-05,
      "loss": 0.808,
      "step": 64
    },
    {
      "epoch": 0.08828522920203735,
      "grad_norm": 2.1164638996124268,
      "learning_rate": 8.07106356344834e-05,
      "loss": 1.0962,
      "step": 65
    },
    {
      "epoch": 0.08964346349745331,
      "grad_norm": 1.5968292951583862,
      "learning_rate": 8.005405736415126e-05,
      "loss": 0.7918,
      "step": 66
    },
    {
      "epoch": 0.09100169779286928,
      "grad_norm": 1.1926372051239014,
      "learning_rate": 7.938926261462366e-05,
      "loss": 0.568,
      "step": 67
    },
    {
      "epoch": 0.09235993208828523,
      "grad_norm": 1.2155282497406006,
      "learning_rate": 7.871643313414718e-05,
      "loss": 0.7005,
      "step": 68
    },
    {
      "epoch": 0.09371816638370119,
      "grad_norm": 1.3235710859298706,
      "learning_rate": 7.803575286758364e-05,
      "loss": 0.5867,
      "step": 69
    },
    {
      "epoch": 0.09507640067911714,
      "grad_norm": 1.535868763923645,
      "learning_rate": 7.734740790612136e-05,
      "loss": 0.5601,
      "step": 70
    },
    {
      "epoch": 0.09643463497453311,
      "grad_norm": 1.332809329032898,
      "learning_rate": 7.66515864363997e-05,
      "loss": 0.5128,
      "step": 71
    },
    {
      "epoch": 0.09779286926994907,
      "grad_norm": 1.1184101104736328,
      "learning_rate": 7.594847868906076e-05,
      "loss": 0.6584,
      "step": 72
    },
    {
      "epoch": 0.09915110356536502,
      "grad_norm": 1.3330540657043457,
      "learning_rate": 7.52382768867422e-05,
      "loss": 0.4603,
      "step": 73
    },
    {
      "epoch": 0.10050933786078098,
      "grad_norm": 1.7870532274246216,
      "learning_rate": 7.452117519152542e-05,
      "loss": 0.9252,
      "step": 74
    },
    {
      "epoch": 0.10186757215619695,
      "grad_norm": 1.4110106229782104,
      "learning_rate": 7.379736965185368e-05,
      "loss": 0.532,
      "step": 75
    },
    {
      "epoch": 0.1032258064516129,
      "grad_norm": 1.2677960395812988,
      "learning_rate": 7.30670581489344e-05,
      "loss": 0.369,
      "step": 76
    },
    {
      "epoch": 0.10458404074702886,
      "grad_norm": 1.5324734449386597,
      "learning_rate": 7.233044034264034e-05,
      "loss": 0.4731,
      "step": 77
    },
    {
      "epoch": 0.10594227504244483,
      "grad_norm": 1.3089126348495483,
      "learning_rate": 7.158771761692464e-05,
      "loss": 0.367,
      "step": 78
    },
    {
      "epoch": 0.10730050933786078,
      "grad_norm": 1.065536618232727,
      "learning_rate": 7.083909302476453e-05,
      "loss": 0.5767,
      "step": 79
    },
    {
      "epoch": 0.10865874363327674,
      "grad_norm": 1.0984853506088257,
      "learning_rate": 7.008477123264848e-05,
      "loss": 0.5426,
      "step": 80
    },
    {
      "epoch": 0.1100169779286927,
      "grad_norm": 1.1644034385681152,
      "learning_rate": 6.932495846462261e-05,
      "loss": 0.5205,
      "step": 81
    },
    {
      "epoch": 0.11137521222410866,
      "grad_norm": 0.6923614144325256,
      "learning_rate": 6.855986244591104e-05,
      "loss": 0.2668,
      "step": 82
    },
    {
      "epoch": 0.11273344651952462,
      "grad_norm": 1.040382981300354,
      "learning_rate": 6.778969234612584e-05,
      "loss": 0.4309,
      "step": 83
    },
    {
      "epoch": 0.11409168081494057,
      "grad_norm": 0.6914508938789368,
      "learning_rate": 6.701465872208216e-05,
      "loss": 0.1929,
      "step": 84
    },
    {
      "epoch": 0.11544991511035653,
      "grad_norm": 0.976138710975647,
      "learning_rate": 6.623497346023418e-05,
      "loss": 0.3347,
      "step": 85
    },
    {
      "epoch": 0.1168081494057725,
      "grad_norm": 0.8875433802604675,
      "learning_rate": 6.545084971874738e-05,
      "loss": 0.3286,
      "step": 86
    },
    {
      "epoch": 0.11816638370118845,
      "grad_norm": 0.7386589646339417,
      "learning_rate": 6.466250186922325e-05,
      "loss": 0.4692,
      "step": 87
    },
    {
      "epoch": 0.11952461799660441,
      "grad_norm": 1.2174679040908813,
      "learning_rate": 6.387014543809223e-05,
      "loss": 0.4497,
      "step": 88
    },
    {
      "epoch": 0.12088285229202038,
      "grad_norm": 0.8012718558311462,
      "learning_rate": 6.307399704769099e-05,
      "loss": 0.3505,
      "step": 89
    },
    {
      "epoch": 0.12224108658743633,
      "grad_norm": 0.8060907125473022,
      "learning_rate": 6.227427435703997e-05,
      "loss": 0.3401,
      "step": 90
    },
    {
      "epoch": 0.12359932088285229,
      "grad_norm": 0.612433135509491,
      "learning_rate": 6.147119600233758e-05,
      "loss": 0.2253,
      "step": 91
    },
    {
      "epoch": 0.12495755517826825,
      "grad_norm": 0.798200786113739,
      "learning_rate": 6.066498153718735e-05,
      "loss": 0.3208,
      "step": 92
    },
    {
      "epoch": 0.12631578947368421,
      "grad_norm": 0.9495358467102051,
      "learning_rate": 5.985585137257401e-05,
      "loss": 0.5265,
      "step": 93
    },
    {
      "epoch": 0.12767402376910017,
      "grad_norm": 1.0848101377487183,
      "learning_rate": 5.90440267166055e-05,
      "loss": 0.6237,
      "step": 94
    },
    {
      "epoch": 0.12903225806451613,
      "grad_norm": 0.6925517916679382,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 0.3498,
      "step": 95
    },
    {
      "epoch": 0.13039049235993208,
      "grad_norm": 0.6562709808349609,
      "learning_rate": 5.74131823855921e-05,
      "loss": 0.2197,
      "step": 96
    },
    {
      "epoch": 0.13174872665534804,
      "grad_norm": 1.1780856847763062,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 0.449,
      "step": 97
    },
    {
      "epoch": 0.13310696095076402,
      "grad_norm": 0.8803690075874329,
      "learning_rate": 5.577423184847932e-05,
      "loss": 0.2064,
      "step": 98
    },
    {
      "epoch": 0.13446519524617997,
      "grad_norm": 1.1821833848953247,
      "learning_rate": 5.495227651252315e-05,
      "loss": 0.4137,
      "step": 99
    },
    {
      "epoch": 0.13582342954159593,
      "grad_norm": 1.2370933294296265,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 0.312,
      "step": 100
    },
    {
      "epoch": 0.13582342954159593,
      "eval_loss": 0.18762627243995667,
      "eval_runtime": 67.3144,
      "eval_samples_per_second": 18.421,
      "eval_steps_per_second": 4.605,
      "step": 100
    },
    {
      "epoch": 0.13718166383701189,
      "grad_norm": 2.157839775085449,
      "learning_rate": 5.330452921628497e-05,
      "loss": 1.8393,
      "step": 101
    },
    {
      "epoch": 0.13853989813242784,
      "grad_norm": 3.5639190673828125,
      "learning_rate": 5.247918773366112e-05,
      "loss": 2.5582,
      "step": 102
    },
    {
      "epoch": 0.1398981324278438,
      "grad_norm": 2.826010227203369,
      "learning_rate": 5.165316846586541e-05,
      "loss": 2.4372,
      "step": 103
    },
    {
      "epoch": 0.14125636672325975,
      "grad_norm": 2.775473117828369,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 2.3188,
      "step": 104
    },
    {
      "epoch": 0.14261460101867574,
      "grad_norm": 2.6814448833465576,
      "learning_rate": 5e-05,
      "loss": 1.9215,
      "step": 105
    },
    {
      "epoch": 0.1439728353140917,
      "grad_norm": 1.5364346504211426,
      "learning_rate": 4.917330276168208e-05,
      "loss": 0.9285,
      "step": 106
    },
    {
      "epoch": 0.14533106960950765,
      "grad_norm": 1.3806895017623901,
      "learning_rate": 4.834683153413459e-05,
      "loss": 0.7553,
      "step": 107
    },
    {
      "epoch": 0.1466893039049236,
      "grad_norm": 1.292453408241272,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 0.4807,
      "step": 108
    },
    {
      "epoch": 0.14804753820033956,
      "grad_norm": 1.7320868968963623,
      "learning_rate": 4.669547078371504e-05,
      "loss": 0.8362,
      "step": 109
    },
    {
      "epoch": 0.1494057724957555,
      "grad_norm": 1.613324522972107,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 0.5316,
      "step": 110
    },
    {
      "epoch": 0.15076400679117147,
      "grad_norm": 1.395429015159607,
      "learning_rate": 4.504772348747687e-05,
      "loss": 0.4296,
      "step": 111
    },
    {
      "epoch": 0.15212224108658742,
      "grad_norm": 1.6080973148345947,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 0.449,
      "step": 112
    },
    {
      "epoch": 0.1534804753820034,
      "grad_norm": 2.063976764678955,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 0.7207,
      "step": 113
    },
    {
      "epoch": 0.15483870967741936,
      "grad_norm": 1.6148802042007446,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 0.8924,
      "step": 114
    },
    {
      "epoch": 0.15619694397283532,
      "grad_norm": 1.8477810621261597,
      "learning_rate": 4.17702704859633e-05,
      "loss": 0.7061,
      "step": 115
    },
    {
      "epoch": 0.15755517826825127,
      "grad_norm": 1.478446125984192,
      "learning_rate": 4.095597328339452e-05,
      "loss": 0.6664,
      "step": 116
    },
    {
      "epoch": 0.15891341256366723,
      "grad_norm": 1.2406508922576904,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 0.5284,
      "step": 117
    },
    {
      "epoch": 0.16027164685908318,
      "grad_norm": 1.1124491691589355,
      "learning_rate": 3.933501846281267e-05,
      "loss": 0.5255,
      "step": 118
    },
    {
      "epoch": 0.16162988115449914,
      "grad_norm": 0.9964424967765808,
      "learning_rate": 3.852880399766243e-05,
      "loss": 0.3398,
      "step": 119
    },
    {
      "epoch": 0.16298811544991512,
      "grad_norm": 1.0388567447662354,
      "learning_rate": 3.772572564296005e-05,
      "loss": 0.4075,
      "step": 120
    },
    {
      "epoch": 0.16434634974533108,
      "grad_norm": 1.1458494663238525,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 0.4073,
      "step": 121
    },
    {
      "epoch": 0.16570458404074703,
      "grad_norm": 0.9584537148475647,
      "learning_rate": 3.612985456190778e-05,
      "loss": 0.3457,
      "step": 122
    },
    {
      "epoch": 0.167062818336163,
      "grad_norm": 1.1488529443740845,
      "learning_rate": 3.533749813077677e-05,
      "loss": 0.2939,
      "step": 123
    },
    {
      "epoch": 0.16842105263157894,
      "grad_norm": 0.8016781806945801,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 0.2453,
      "step": 124
    },
    {
      "epoch": 0.1697792869269949,
      "grad_norm": 0.8860985040664673,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 0.2834,
      "step": 125
    },
    {
      "epoch": 0.17113752122241085,
      "grad_norm": 1.3960487842559814,
      "learning_rate": 3.298534127791785e-05,
      "loss": 0.3573,
      "step": 126
    },
    {
      "epoch": 0.17249575551782684,
      "grad_norm": 1.4462261199951172,
      "learning_rate": 3.221030765387417e-05,
      "loss": 0.4583,
      "step": 127
    },
    {
      "epoch": 0.1738539898132428,
      "grad_norm": 0.9451635479927063,
      "learning_rate": 3.144013755408895e-05,
      "loss": 0.2419,
      "step": 128
    },
    {
      "epoch": 0.17521222410865875,
      "grad_norm": 0.9558318257331848,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 0.4274,
      "step": 129
    },
    {
      "epoch": 0.1765704584040747,
      "grad_norm": 1.0216695070266724,
      "learning_rate": 2.991522876735154e-05,
      "loss": 0.413,
      "step": 130
    },
    {
      "epoch": 0.17792869269949066,
      "grad_norm": 0.8177182674407959,
      "learning_rate": 2.916090697523549e-05,
      "loss": 0.2995,
      "step": 131
    },
    {
      "epoch": 0.17928692699490661,
      "grad_norm": 0.8251747488975525,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 0.2619,
      "step": 132
    },
    {
      "epoch": 0.18064516129032257,
      "grad_norm": 0.8147902488708496,
      "learning_rate": 2.766955965735968e-05,
      "loss": 0.328,
      "step": 133
    },
    {
      "epoch": 0.18200339558573855,
      "grad_norm": 1.024512767791748,
      "learning_rate": 2.693294185106562e-05,
      "loss": 0.2928,
      "step": 134
    },
    {
      "epoch": 0.1833616298811545,
      "grad_norm": 0.9274868965148926,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 0.2619,
      "step": 135
    },
    {
      "epoch": 0.18471986417657046,
      "grad_norm": 1.010575294494629,
      "learning_rate": 2.547882480847461e-05,
      "loss": 0.3417,
      "step": 136
    },
    {
      "epoch": 0.18607809847198642,
      "grad_norm": 0.9014824628829956,
      "learning_rate": 2.476172311325783e-05,
      "loss": 0.4541,
      "step": 137
    },
    {
      "epoch": 0.18743633276740237,
      "grad_norm": 0.9789355993270874,
      "learning_rate": 2.405152131093926e-05,
      "loss": 0.3065,
      "step": 138
    },
    {
      "epoch": 0.18879456706281833,
      "grad_norm": 0.7785047888755798,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 0.3614,
      "step": 139
    },
    {
      "epoch": 0.19015280135823429,
      "grad_norm": 0.809662938117981,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 0.2764,
      "step": 140
    },
    {
      "epoch": 0.19151103565365024,
      "grad_norm": 1.2895207405090332,
      "learning_rate": 2.196424713241637e-05,
      "loss": 0.6835,
      "step": 141
    },
    {
      "epoch": 0.19286926994906622,
      "grad_norm": 0.8904023766517639,
      "learning_rate": 2.128356686585282e-05,
      "loss": 0.4472,
      "step": 142
    },
    {
      "epoch": 0.19422750424448218,
      "grad_norm": 0.7771594524383545,
      "learning_rate": 2.061073738537635e-05,
      "loss": 0.293,
      "step": 143
    },
    {
      "epoch": 0.19558573853989814,
      "grad_norm": 0.9814310669898987,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 0.2739,
      "step": 144
    },
    {
      "epoch": 0.1969439728353141,
      "grad_norm": 1.117093563079834,
      "learning_rate": 1.928936436551661e-05,
      "loss": 0.379,
      "step": 145
    },
    {
      "epoch": 0.19830220713073005,
      "grad_norm": 0.7254476547241211,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 0.2855,
      "step": 146
    },
    {
      "epoch": 0.199660441426146,
      "grad_norm": 0.716963529586792,
      "learning_rate": 1.800157297483417e-05,
      "loss": 0.2806,
      "step": 147
    },
    {
      "epoch": 0.20101867572156196,
      "grad_norm": 0.7899037003517151,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 0.2052,
      "step": 148
    },
    {
      "epoch": 0.20237691001697794,
      "grad_norm": 0.9633182287216187,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 0.1877,
      "step": 149
    },
    {
      "epoch": 0.2037351443123939,
      "grad_norm": 1.0001466274261475,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 0.2413,
      "step": 150
    },
    {
      "epoch": 0.2037351443123939,
      "eval_loss": 0.14431816339492798,
      "eval_runtime": 67.1884,
      "eval_samples_per_second": 18.456,
      "eval_steps_per_second": 4.614,
      "step": 150
    },
    {
      "epoch": 0.20509337860780985,
      "grad_norm": 1.2680031061172485,
      "learning_rate": 1.553232954407171e-05,
      "loss": 1.549,
      "step": 151
    },
    {
      "epoch": 0.2064516129032258,
      "grad_norm": 1.2180172204971313,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 2.0092,
      "step": 152
    },
    {
      "epoch": 0.20780984719864176,
      "grad_norm": 1.5306624174118042,
      "learning_rate": 1.435357758543015e-05,
      "loss": 2.173,
      "step": 153
    },
    {
      "epoch": 0.20916808149405772,
      "grad_norm": 1.5853345394134521,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 2.1808,
      "step": 154
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 1.7286503314971924,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 1.8567,
      "step": 155
    },
    {
      "epoch": 0.21188455008488966,
      "grad_norm": 1.7660788297653198,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 1.3088,
      "step": 156
    },
    {
      "epoch": 0.2132427843803056,
      "grad_norm": 0.83577561378479,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 0.6152,
      "step": 157
    },
    {
      "epoch": 0.21460101867572157,
      "grad_norm": 1.0349199771881104,
      "learning_rate": 1.157994445715706e-05,
      "loss": 0.525,
      "step": 158
    },
    {
      "epoch": 0.21595925297113752,
      "grad_norm": 1.2438606023788452,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 0.2478,
      "step": 159
    },
    {
      "epoch": 0.21731748726655348,
      "grad_norm": 2.1919026374816895,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 0.9311,
      "step": 160
    },
    {
      "epoch": 0.21867572156196943,
      "grad_norm": 2.6150472164154053,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 0.8594,
      "step": 161
    },
    {
      "epoch": 0.2200339558573854,
      "grad_norm": 2.1280641555786133,
      "learning_rate": 9.549150281252633e-06,
      "loss": 0.3983,
      "step": 162
    },
    {
      "epoch": 0.22139219015280137,
      "grad_norm": 0.9619987607002258,
      "learning_rate": 9.068759265665384e-06,
      "loss": 0.3423,
      "step": 163
    },
    {
      "epoch": 0.22275042444821733,
      "grad_norm": 1.3266946077346802,
      "learning_rate": 8.599558442598998e-06,
      "loss": 0.807,
      "step": 164
    },
    {
      "epoch": 0.22410865874363328,
      "grad_norm": 0.9932201504707336,
      "learning_rate": 8.141676086873572e-06,
      "loss": 0.4736,
      "step": 165
    },
    {
      "epoch": 0.22546689303904924,
      "grad_norm": 1.1188908815383911,
      "learning_rate": 7.695237378953223e-06,
      "loss": 0.4146,
      "step": 166
    },
    {
      "epoch": 0.2268251273344652,
      "grad_norm": 1.243120551109314,
      "learning_rate": 7.260364370723044e-06,
      "loss": 0.645,
      "step": 167
    },
    {
      "epoch": 0.22818336162988115,
      "grad_norm": 1.0500059127807617,
      "learning_rate": 6.837175952121306e-06,
      "loss": 0.3818,
      "step": 168
    },
    {
      "epoch": 0.2295415959252971,
      "grad_norm": 1.0347051620483398,
      "learning_rate": 6.425787818636131e-06,
      "loss": 0.3161,
      "step": 169
    },
    {
      "epoch": 0.23089983022071306,
      "grad_norm": 0.9902946949005127,
      "learning_rate": 6.026312439675552e-06,
      "loss": 0.3734,
      "step": 170
    },
    {
      "epoch": 0.23225806451612904,
      "grad_norm": 1.1293691396713257,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 0.3349,
      "step": 171
    },
    {
      "epoch": 0.233616298811545,
      "grad_norm": 0.8362142443656921,
      "learning_rate": 5.263533508961827e-06,
      "loss": 0.2666,
      "step": 172
    },
    {
      "epoch": 0.23497453310696095,
      "grad_norm": 0.7939189076423645,
      "learning_rate": 4.900438493352055e-06,
      "loss": 0.1747,
      "step": 173
    },
    {
      "epoch": 0.2363327674023769,
      "grad_norm": 1.3641140460968018,
      "learning_rate": 4.549673247541875e-06,
      "loss": 0.6154,
      "step": 174
    },
    {
      "epoch": 0.23769100169779286,
      "grad_norm": 1.0040560960769653,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 0.3224,
      "step": 175
    },
    {
      "epoch": 0.23904923599320882,
      "grad_norm": 0.8438626527786255,
      "learning_rate": 3.885512251130763e-06,
      "loss": 0.2955,
      "step": 176
    },
    {
      "epoch": 0.24040747028862477,
      "grad_norm": 0.7685846090316772,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 0.2529,
      "step": 177
    },
    {
      "epoch": 0.24176570458404076,
      "grad_norm": 0.860312283039093,
      "learning_rate": 3.271776770026963e-06,
      "loss": 0.3214,
      "step": 178
    },
    {
      "epoch": 0.2431239388794567,
      "grad_norm": 1.2312915325164795,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 0.4232,
      "step": 179
    },
    {
      "epoch": 0.24448217317487267,
      "grad_norm": 1.0875355005264282,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 0.3892,
      "step": 180
    },
    {
      "epoch": 0.24584040747028862,
      "grad_norm": 0.8430658578872681,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 0.3758,
      "step": 181
    },
    {
      "epoch": 0.24719864176570458,
      "grad_norm": 1.005794882774353,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 0.3334,
      "step": 182
    },
    {
      "epoch": 0.24855687606112054,
      "grad_norm": 1.1433653831481934,
      "learning_rate": 1.962316193157593e-06,
      "loss": 0.3489,
      "step": 183
    },
    {
      "epoch": 0.2499151103565365,
      "grad_norm": 1.028061866760254,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 0.3125,
      "step": 184
    },
    {
      "epoch": 0.25127334465195245,
      "grad_norm": 0.9788212180137634,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 0.3202,
      "step": 185
    },
    {
      "epoch": 0.25263157894736843,
      "grad_norm": 0.7344953417778015,
      "learning_rate": 1.333670137599713e-06,
      "loss": 0.1761,
      "step": 186
    },
    {
      "epoch": 0.25398981324278436,
      "grad_norm": 0.8415098190307617,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 0.2663,
      "step": 187
    },
    {
      "epoch": 0.25534804753820034,
      "grad_norm": 0.7864528894424438,
      "learning_rate": 9.810017062595322e-07,
      "loss": 0.2404,
      "step": 188
    },
    {
      "epoch": 0.2567062818336163,
      "grad_norm": 0.8776137828826904,
      "learning_rate": 8.247462563808817e-07,
      "loss": 0.5342,
      "step": 189
    },
    {
      "epoch": 0.25806451612903225,
      "grad_norm": 1.2769297361373901,
      "learning_rate": 6.819348298638839e-07,
      "loss": 0.3738,
      "step": 190
    },
    {
      "epoch": 0.25942275042444823,
      "grad_norm": 1.3412901163101196,
      "learning_rate": 5.526064699265753e-07,
      "loss": 0.4223,
      "step": 191
    },
    {
      "epoch": 0.26078098471986416,
      "grad_norm": 1.203078269958496,
      "learning_rate": 4.367965336512403e-07,
      "loss": 0.5816,
      "step": 192
    },
    {
      "epoch": 0.26213921901528014,
      "grad_norm": 0.7777627110481262,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 0.2591,
      "step": 193
    },
    {
      "epoch": 0.2634974533106961,
      "grad_norm": 1.0466378927230835,
      "learning_rate": 2.458548727494292e-07,
      "loss": 0.5292,
      "step": 194
    },
    {
      "epoch": 0.26485568760611206,
      "grad_norm": 0.8496146202087402,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 0.266,
      "step": 195
    },
    {
      "epoch": 0.26621392190152804,
      "grad_norm": 0.8443109393119812,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 0.2021,
      "step": 196
    },
    {
      "epoch": 0.26757215619694397,
      "grad_norm": 0.6287698745727539,
      "learning_rate": 6.150154258476315e-08,
      "loss": 0.1533,
      "step": 197
    },
    {
      "epoch": 0.26893039049235995,
      "grad_norm": 0.7238965630531311,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 0.2349,
      "step": 198
    },
    {
      "epoch": 0.2702886247877759,
      "grad_norm": 0.7158587574958801,
      "learning_rate": 6.834750376549792e-09,
      "loss": 0.2074,
      "step": 199
    },
    {
      "epoch": 0.27164685908319186,
      "grad_norm": 0.7276369333267212,
      "learning_rate": 0.0,
      "loss": 0.149,
      "step": 200
    },
    {
      "epoch": 0.27164685908319186,
      "eval_loss": 0.12807011604309082,
      "eval_runtime": 67.31,
      "eval_samples_per_second": 18.422,
      "eval_steps_per_second": 4.606,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.6826331452512666e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}