{
  "best_metric": 2.317739963531494,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.4548038658328596,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0022740193291642978,
      "grad_norm": 2.0814766883850098,
      "learning_rate": 1e-05,
      "loss": 3.7319,
      "step": 1
    },
    {
      "epoch": 0.0022740193291642978,
      "eval_loss": 4.409133434295654,
      "eval_runtime": 53.5885,
      "eval_samples_per_second": 13.828,
      "eval_steps_per_second": 3.471,
      "step": 1
    },
    {
      "epoch": 0.0045480386583285955,
      "grad_norm": 1.5011945962905884,
      "learning_rate": 2e-05,
      "loss": 3.3846,
      "step": 2
    },
    {
      "epoch": 0.006822057987492893,
      "grad_norm": 1.4566481113433838,
      "learning_rate": 3e-05,
      "loss": 3.6299,
      "step": 3
    },
    {
      "epoch": 0.009096077316657191,
      "grad_norm": 0.997739851474762,
      "learning_rate": 4e-05,
      "loss": 3.2553,
      "step": 4
    },
    {
      "epoch": 0.01137009664582149,
      "grad_norm": 1.075835943222046,
      "learning_rate": 5e-05,
      "loss": 3.4026,
      "step": 5
    },
    {
      "epoch": 0.013644115974985787,
      "grad_norm": 1.0173403024673462,
      "learning_rate": 6e-05,
      "loss": 3.1608,
      "step": 6
    },
    {
      "epoch": 0.015918135304150087,
      "grad_norm": 0.8207800388336182,
      "learning_rate": 7e-05,
      "loss": 3.3522,
      "step": 7
    },
    {
      "epoch": 0.018192154633314382,
      "grad_norm": 0.7397863268852234,
      "learning_rate": 8e-05,
      "loss": 3.0094,
      "step": 8
    },
    {
      "epoch": 0.02046617396247868,
      "grad_norm": 0.7712698578834534,
      "learning_rate": 9e-05,
      "loss": 2.9865,
      "step": 9
    },
    {
      "epoch": 0.02274019329164298,
      "grad_norm": 1.059030532836914,
      "learning_rate": 0.0001,
      "loss": 3.2459,
      "step": 10
    },
    {
      "epoch": 0.025014212620807278,
      "grad_norm": 1.129819393157959,
      "learning_rate": 9.999316524962345e-05,
      "loss": 3.0385,
      "step": 11
    },
    {
      "epoch": 0.027288231949971573,
      "grad_norm": 1.1459600925445557,
      "learning_rate": 9.997266286704631e-05,
      "loss": 3.2707,
      "step": 12
    },
    {
      "epoch": 0.029562251279135872,
      "grad_norm": 1.1167596578598022,
      "learning_rate": 9.993849845741524e-05,
      "loss": 2.8996,
      "step": 13
    },
    {
      "epoch": 0.031836270608300174,
      "grad_norm": 0.7198836207389832,
      "learning_rate": 9.989068136093873e-05,
      "loss": 2.8833,
      "step": 14
    },
    {
      "epoch": 0.03411028993746447,
      "grad_norm": 0.775762140750885,
      "learning_rate": 9.98292246503335e-05,
      "loss": 2.7893,
      "step": 15
    },
    {
      "epoch": 0.036384309266628764,
      "grad_norm": 0.6547066569328308,
      "learning_rate": 9.975414512725057e-05,
      "loss": 3.0829,
      "step": 16
    },
    {
      "epoch": 0.038658328595793066,
      "grad_norm": 0.6285343170166016,
      "learning_rate": 9.966546331768191e-05,
      "loss": 2.9581,
      "step": 17
    },
    {
      "epoch": 0.04093234792495736,
      "grad_norm": 0.7529861927032471,
      "learning_rate": 9.956320346634876e-05,
      "loss": 3.0242,
      "step": 18
    },
    {
      "epoch": 0.04320636725412166,
      "grad_norm": 0.7179023623466492,
      "learning_rate": 9.944739353007344e-05,
      "loss": 3.0135,
      "step": 19
    },
    {
      "epoch": 0.04548038658328596,
      "grad_norm": 0.6654857397079468,
      "learning_rate": 9.931806517013612e-05,
      "loss": 2.7648,
      "step": 20
    },
    {
      "epoch": 0.047754405912450254,
      "grad_norm": 0.6120819449424744,
      "learning_rate": 9.917525374361912e-05,
      "loss": 2.9448,
      "step": 21
    },
    {
      "epoch": 0.050028425241614556,
      "grad_norm": 0.6705901622772217,
      "learning_rate": 9.901899829374047e-05,
      "loss": 2.9723,
      "step": 22
    },
    {
      "epoch": 0.05230244457077885,
      "grad_norm": 0.7630836367607117,
      "learning_rate": 9.884934153917997e-05,
      "loss": 3.2882,
      "step": 23
    },
    {
      "epoch": 0.054576463899943146,
      "grad_norm": 0.7805302143096924,
      "learning_rate": 9.86663298624003e-05,
      "loss": 2.9121,
      "step": 24
    },
    {
      "epoch": 0.05685048322910745,
      "grad_norm": 0.779008150100708,
      "learning_rate": 9.847001329696653e-05,
      "loss": 2.86,
      "step": 25
    },
    {
      "epoch": 0.059124502558271744,
      "grad_norm": 0.7268081307411194,
      "learning_rate": 9.826044551386744e-05,
      "loss": 3.0592,
      "step": 26
    },
    {
      "epoch": 0.061398521887436046,
      "grad_norm": 0.7348185181617737,
      "learning_rate": 9.803768380684242e-05,
      "loss": 2.8172,
      "step": 27
    },
    {
      "epoch": 0.06367254121660035,
      "grad_norm": 0.7491097450256348,
      "learning_rate": 9.780178907671789e-05,
      "loss": 2.7001,
      "step": 28
    },
    {
      "epoch": 0.06594656054576464,
      "grad_norm": 0.7652377486228943,
      "learning_rate": 9.755282581475769e-05,
      "loss": 2.7501,
      "step": 29
    },
    {
      "epoch": 0.06822057987492894,
      "grad_norm": 0.7634243369102478,
      "learning_rate": 9.729086208503174e-05,
      "loss": 2.8482,
      "step": 30
    },
    {
      "epoch": 0.07049459920409323,
      "grad_norm": 0.759520947933197,
      "learning_rate": 9.701596950580806e-05,
      "loss": 3.0724,
      "step": 31
    },
    {
      "epoch": 0.07276861853325753,
      "grad_norm": 0.8122732639312744,
      "learning_rate": 9.672822322997305e-05,
      "loss": 2.6431,
      "step": 32
    },
    {
      "epoch": 0.07504263786242182,
      "grad_norm": 0.8141663670539856,
      "learning_rate": 9.642770192448536e-05,
      "loss": 3.0307,
      "step": 33
    },
    {
      "epoch": 0.07731665719158613,
      "grad_norm": 0.8564486503601074,
      "learning_rate": 9.611448774886924e-05,
      "loss": 3.0299,
      "step": 34
    },
    {
      "epoch": 0.07959067652075043,
      "grad_norm": 0.9348270297050476,
      "learning_rate": 9.578866633275288e-05,
      "loss": 2.9455,
      "step": 35
    },
    {
      "epoch": 0.08186469584991472,
      "grad_norm": 1.0203107595443726,
      "learning_rate": 9.545032675245813e-05,
      "loss": 2.7799,
      "step": 36
    },
    {
      "epoch": 0.08413871517907902,
      "grad_norm": 0.9717430472373962,
      "learning_rate": 9.509956150664796e-05,
      "loss": 2.7277,
      "step": 37
    },
    {
      "epoch": 0.08641273450824331,
      "grad_norm": 1.0820761919021606,
      "learning_rate": 9.473646649103818e-05,
      "loss": 2.8131,
      "step": 38
    },
    {
      "epoch": 0.08868675383740762,
      "grad_norm": 1.2270339727401733,
      "learning_rate": 9.43611409721806e-05,
      "loss": 2.7115,
      "step": 39
    },
    {
      "epoch": 0.09096077316657192,
      "grad_norm": 1.2938342094421387,
      "learning_rate": 9.397368756032445e-05,
      "loss": 3.0858,
      "step": 40
    },
    {
      "epoch": 0.09323479249573621,
      "grad_norm": 1.3103678226470947,
      "learning_rate": 9.357421218136386e-05,
      "loss": 2.7187,
      "step": 41
    },
    {
      "epoch": 0.09550881182490051,
      "grad_norm": 1.405442476272583,
      "learning_rate": 9.316282404787871e-05,
      "loss": 2.7994,
      "step": 42
    },
    {
      "epoch": 0.0977828311540648,
      "grad_norm": 1.621390461921692,
      "learning_rate": 9.273963562927695e-05,
      "loss": 2.8012,
      "step": 43
    },
    {
      "epoch": 0.10005685048322911,
      "grad_norm": 1.7072330713272095,
      "learning_rate": 9.230476262104677e-05,
      "loss": 3.1509,
      "step": 44
    },
    {
      "epoch": 0.10233086981239341,
      "grad_norm": 1.847292423248291,
      "learning_rate": 9.185832391312644e-05,
      "loss": 2.9792,
      "step": 45
    },
    {
      "epoch": 0.1046048891415577,
      "grad_norm": 2.0138838291168213,
      "learning_rate": 9.140044155740101e-05,
      "loss": 2.926,
      "step": 46
    },
    {
      "epoch": 0.106878908470722,
      "grad_norm": 2.8807947635650635,
      "learning_rate": 9.093124073433463e-05,
      "loss": 3.1747,
      "step": 47
    },
    {
      "epoch": 0.10915292779988629,
      "grad_norm": 3.2223868370056152,
      "learning_rate": 9.045084971874738e-05,
      "loss": 3.2006,
      "step": 48
    },
    {
      "epoch": 0.1114269471290506,
      "grad_norm": 3.2357280254364014,
      "learning_rate": 8.995939984474624e-05,
      "loss": 3.3191,
      "step": 49
    },
    {
      "epoch": 0.1137009664582149,
      "grad_norm": 5.290281772613525,
      "learning_rate": 8.945702546981969e-05,
      "loss": 3.9757,
      "step": 50
    },
    {
      "epoch": 0.1137009664582149,
      "eval_loss": 2.780275344848633,
      "eval_runtime": 53.7506,
      "eval_samples_per_second": 13.786,
      "eval_steps_per_second": 3.46,
      "step": 50
    },
    {
      "epoch": 0.11597498578737919,
      "grad_norm": 2.353924512863159,
      "learning_rate": 8.894386393810563e-05,
      "loss": 2.8536,
      "step": 51
    },
    {
      "epoch": 0.11824900511654349,
      "grad_norm": 2.1533467769622803,
      "learning_rate": 8.842005554284296e-05,
      "loss": 2.5879,
      "step": 52
    },
    {
      "epoch": 0.12052302444570778,
      "grad_norm": 1.7365756034851074,
      "learning_rate": 8.788574348801675e-05,
      "loss": 2.649,
      "step": 53
    },
    {
      "epoch": 0.12279704377487209,
      "grad_norm": 1.2279722690582275,
      "learning_rate": 8.73410738492077e-05,
      "loss": 2.3847,
      "step": 54
    },
    {
      "epoch": 0.12507106310403637,
      "grad_norm": 0.8106296062469482,
      "learning_rate": 8.678619553365659e-05,
      "loss": 2.4474,
      "step": 55
    },
    {
      "epoch": 0.1273450824332007,
      "grad_norm": 0.7888613939285278,
      "learning_rate": 8.622126023955446e-05,
      "loss": 2.3418,
      "step": 56
    },
    {
      "epoch": 0.129619101762365,
      "grad_norm": 1.0792253017425537,
      "learning_rate": 8.564642241456986e-05,
      "loss": 2.5672,
      "step": 57
    },
    {
      "epoch": 0.1318931210915293,
      "grad_norm": 0.8654270172119141,
      "learning_rate": 8.506183921362443e-05,
      "loss": 2.4619,
      "step": 58
    },
    {
      "epoch": 0.13416714042069358,
      "grad_norm": 0.6500675082206726,
      "learning_rate": 8.44676704559283e-05,
      "loss": 2.1112,
      "step": 59
    },
    {
      "epoch": 0.13644115974985788,
      "grad_norm": 0.7029934525489807,
      "learning_rate": 8.386407858128706e-05,
      "loss": 2.3759,
      "step": 60
    },
    {
      "epoch": 0.13871517907902217,
      "grad_norm": 0.660019040107727,
      "learning_rate": 8.32512286056924e-05,
      "loss": 2.2432,
      "step": 61
    },
    {
      "epoch": 0.14098919840818647,
      "grad_norm": 0.6818193793296814,
      "learning_rate": 8.262928807620843e-05,
      "loss": 2.3159,
      "step": 62
    },
    {
      "epoch": 0.14326321773735076,
      "grad_norm": 0.774668276309967,
      "learning_rate": 8.199842702516583e-05,
      "loss": 2.1564,
      "step": 63
    },
    {
      "epoch": 0.14553723706651506,
      "grad_norm": 0.7120164632797241,
      "learning_rate": 8.135881792367686e-05,
      "loss": 2.2556,
      "step": 64
    },
    {
      "epoch": 0.14781125639567935,
      "grad_norm": 0.7783941626548767,
      "learning_rate": 8.07106356344834e-05,
      "loss": 2.2787,
      "step": 65
    },
    {
      "epoch": 0.15008527572484365,
      "grad_norm": 0.797565221786499,
      "learning_rate": 8.005405736415126e-05,
      "loss": 2.3226,
      "step": 66
    },
    {
      "epoch": 0.15235929505400797,
      "grad_norm": 0.7392576336860657,
      "learning_rate": 7.938926261462366e-05,
      "loss": 2.4609,
      "step": 67
    },
    {
      "epoch": 0.15463331438317227,
      "grad_norm": 0.7914451360702515,
      "learning_rate": 7.871643313414718e-05,
      "loss": 2.238,
      "step": 68
    },
    {
      "epoch": 0.15690733371233656,
      "grad_norm": 0.8438661694526672,
      "learning_rate": 7.803575286758364e-05,
      "loss": 2.2895,
      "step": 69
    },
    {
      "epoch": 0.15918135304150086,
      "grad_norm": 0.7578096389770508,
      "learning_rate": 7.734740790612136e-05,
      "loss": 2.2941,
      "step": 70
    },
    {
      "epoch": 0.16145537237066515,
      "grad_norm": 0.8900478482246399,
      "learning_rate": 7.66515864363997e-05,
      "loss": 2.4431,
      "step": 71
    },
    {
      "epoch": 0.16372939169982945,
      "grad_norm": 0.8806887865066528,
      "learning_rate": 7.594847868906076e-05,
      "loss": 2.5127,
      "step": 72
    },
    {
      "epoch": 0.16600341102899374,
      "grad_norm": 0.9947291612625122,
      "learning_rate": 7.52382768867422e-05,
      "loss": 2.4604,
      "step": 73
    },
    {
      "epoch": 0.16827743035815804,
      "grad_norm": 0.926886796951294,
      "learning_rate": 7.452117519152542e-05,
      "loss": 2.5279,
      "step": 74
    },
    {
      "epoch": 0.17055144968732233,
      "grad_norm": 0.9283239841461182,
      "learning_rate": 7.379736965185368e-05,
      "loss": 2.8139,
      "step": 75
    },
    {
      "epoch": 0.17282546901648663,
      "grad_norm": 0.8995282649993896,
      "learning_rate": 7.30670581489344e-05,
      "loss": 2.4341,
      "step": 76
    },
    {
      "epoch": 0.17509948834565095,
      "grad_norm": 0.8015967011451721,
      "learning_rate": 7.233044034264034e-05,
      "loss": 2.3167,
      "step": 77
    },
    {
      "epoch": 0.17737350767481525,
      "grad_norm": 0.9189391732215881,
      "learning_rate": 7.158771761692464e-05,
      "loss": 2.4371,
      "step": 78
    },
    {
      "epoch": 0.17964752700397954,
      "grad_norm": 0.9155376553535461,
      "learning_rate": 7.083909302476453e-05,
      "loss": 2.4785,
      "step": 79
    },
    {
      "epoch": 0.18192154633314384,
      "grad_norm": 1.0044804811477661,
      "learning_rate": 7.008477123264848e-05,
      "loss": 2.5576,
      "step": 80
    },
    {
      "epoch": 0.18419556566230813,
      "grad_norm": 0.9002287983894348,
      "learning_rate": 6.932495846462261e-05,
      "loss": 2.6381,
      "step": 81
    },
    {
      "epoch": 0.18646958499147243,
      "grad_norm": 0.9138435125350952,
      "learning_rate": 6.855986244591104e-05,
      "loss": 2.2561,
      "step": 82
    },
    {
      "epoch": 0.18874360432063672,
      "grad_norm": 0.9708390235900879,
      "learning_rate": 6.778969234612584e-05,
      "loss": 2.4614,
      "step": 83
    },
    {
      "epoch": 0.19101762364980102,
      "grad_norm": 1.0929266214370728,
      "learning_rate": 6.701465872208216e-05,
      "loss": 2.4536,
      "step": 84
    },
    {
      "epoch": 0.1932916429789653,
      "grad_norm": 1.071815013885498,
      "learning_rate": 6.623497346023418e-05,
      "loss": 2.7662,
      "step": 85
    },
    {
      "epoch": 0.1955656623081296,
      "grad_norm": 1.1642342805862427,
      "learning_rate": 6.545084971874738e-05,
      "loss": 2.8042,
      "step": 86
    },
    {
      "epoch": 0.19783968163729393,
      "grad_norm": 1.2180472612380981,
      "learning_rate": 6.466250186922325e-05,
      "loss": 2.3794,
      "step": 87
    },
    {
      "epoch": 0.20011370096645822,
      "grad_norm": 1.2788559198379517,
      "learning_rate": 6.387014543809223e-05,
      "loss": 2.5128,
      "step": 88
    },
    {
      "epoch": 0.20238772029562252,
      "grad_norm": 1.2693041563034058,
      "learning_rate": 6.307399704769099e-05,
      "loss": 2.8704,
      "step": 89
    },
    {
      "epoch": 0.20466173962478681,
      "grad_norm": 1.2078670263290405,
      "learning_rate": 6.227427435703997e-05,
      "loss": 2.748,
      "step": 90
    },
    {
      "epoch": 0.2069357589539511,
      "grad_norm": 1.4330995082855225,
      "learning_rate": 6.147119600233758e-05,
      "loss": 2.6692,
      "step": 91
    },
    {
      "epoch": 0.2092097782831154,
      "grad_norm": 1.36742103099823,
      "learning_rate": 6.066498153718735e-05,
      "loss": 2.6047,
      "step": 92
    },
    {
      "epoch": 0.2114837976122797,
      "grad_norm": 1.488307237625122,
      "learning_rate": 5.985585137257401e-05,
      "loss": 2.7858,
      "step": 93
    },
    {
      "epoch": 0.213757816941444,
      "grad_norm": 1.8383708000183105,
      "learning_rate": 5.90440267166055e-05,
      "loss": 2.6478,
      "step": 94
    },
    {
      "epoch": 0.2160318362706083,
      "grad_norm": 1.8237890005111694,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 2.6868,
      "step": 95
    },
    {
      "epoch": 0.21830585559977259,
      "grad_norm": 2.238405704498291,
      "learning_rate": 5.74131823855921e-05,
      "loss": 2.6585,
      "step": 96
    },
    {
      "epoch": 0.2205798749289369,
      "grad_norm": 2.7208950519561768,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 1.7769,
      "step": 97
    },
    {
      "epoch": 0.2228538942581012,
      "grad_norm": 3.2288146018981934,
      "learning_rate": 5.577423184847932e-05,
      "loss": 2.3164,
      "step": 98
    },
    {
      "epoch": 0.2251279135872655,
      "grad_norm": 3.477285861968994,
      "learning_rate": 5.495227651252315e-05,
      "loss": 3.0284,
      "step": 99
    },
    {
      "epoch": 0.2274019329164298,
      "grad_norm": 6.095785140991211,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 3.9517,
      "step": 100
    },
    {
      "epoch": 0.2274019329164298,
      "eval_loss": 2.4707589149475098,
      "eval_runtime": 53.7455,
      "eval_samples_per_second": 13.787,
      "eval_steps_per_second": 3.461,
      "step": 100
    },
    {
      "epoch": 0.2296759522455941,
      "grad_norm": 1.3073029518127441,
      "learning_rate": 5.330452921628497e-05,
      "loss": 2.37,
      "step": 101
    },
    {
      "epoch": 0.23194997157475838,
      "grad_norm": 1.3657740354537964,
      "learning_rate": 5.247918773366112e-05,
      "loss": 2.3999,
      "step": 102
    },
    {
      "epoch": 0.23422399090392268,
      "grad_norm": 1.253461480140686,
      "learning_rate": 5.165316846586541e-05,
      "loss": 2.3091,
      "step": 103
    },
    {
      "epoch": 0.23649801023308697,
      "grad_norm": 1.0211058855056763,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 2.3138,
      "step": 104
    },
    {
      "epoch": 0.23877202956225127,
      "grad_norm": 0.8298658728599548,
      "learning_rate": 5e-05,
      "loss": 2.0126,
      "step": 105
    },
    {
      "epoch": 0.24104604889141557,
      "grad_norm": 0.7955043911933899,
      "learning_rate": 4.917330276168208e-05,
      "loss": 2.2312,
      "step": 106
    },
    {
      "epoch": 0.2433200682205799,
      "grad_norm": 0.6754152774810791,
      "learning_rate": 4.834683153413459e-05,
      "loss": 2.0947,
      "step": 107
    },
    {
      "epoch": 0.24559408754974418,
      "grad_norm": 0.6781653761863708,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 2.0386,
      "step": 108
    },
    {
      "epoch": 0.24786810687890848,
      "grad_norm": 0.6811093688011169,
      "learning_rate": 4.669547078371504e-05,
      "loss": 2.0824,
      "step": 109
    },
    {
      "epoch": 0.25014212620807275,
      "grad_norm": 0.8755627870559692,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 2.2296,
      "step": 110
    },
    {
      "epoch": 0.25241614553723707,
      "grad_norm": 0.829145610332489,
      "learning_rate": 4.504772348747687e-05,
      "loss": 2.361,
      "step": 111
    },
    {
      "epoch": 0.2546901648664014,
      "grad_norm": 0.7885571122169495,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 2.138,
      "step": 112
    },
    {
      "epoch": 0.25696418419556566,
      "grad_norm": 0.6788994669914246,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 2.0723,
      "step": 113
    },
    {
      "epoch": 0.25923820352473,
      "grad_norm": 0.7365356683731079,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 2.3084,
      "step": 114
    },
    {
      "epoch": 0.26151222285389425,
      "grad_norm": 0.7680405378341675,
      "learning_rate": 4.17702704859633e-05,
      "loss": 2.1175,
      "step": 115
    },
    {
      "epoch": 0.2637862421830586,
      "grad_norm": 0.8954876065254211,
      "learning_rate": 4.095597328339452e-05,
      "loss": 2.0908,
      "step": 116
    },
    {
      "epoch": 0.26606026151222284,
      "grad_norm": 0.8231341242790222,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 2.2758,
      "step": 117
    },
    {
      "epoch": 0.26833428084138716,
      "grad_norm": 0.7816740274429321,
      "learning_rate": 3.933501846281267e-05,
      "loss": 2.3994,
      "step": 118
    },
    {
      "epoch": 0.27060830017055143,
      "grad_norm": 0.7686096429824829,
      "learning_rate": 3.852880399766243e-05,
      "loss": 2.367,
      "step": 119
    },
    {
      "epoch": 0.27288231949971575,
      "grad_norm": 0.8274964690208435,
      "learning_rate": 3.772572564296005e-05,
      "loss": 2.3618,
      "step": 120
    },
    {
      "epoch": 0.27515633882888,
      "grad_norm": 0.8358203172683716,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 2.1838,
      "step": 121
    },
    {
      "epoch": 0.27743035815804434,
      "grad_norm": 1.0523433685302734,
      "learning_rate": 3.612985456190778e-05,
      "loss": 2.4431,
      "step": 122
    },
    {
      "epoch": 0.27970437748720867,
      "grad_norm": 0.8998702168464661,
      "learning_rate": 3.533749813077677e-05,
      "loss": 2.5009,
      "step": 123
    },
    {
      "epoch": 0.28197839681637293,
      "grad_norm": 0.8731322884559631,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 2.3193,
      "step": 124
    },
    {
      "epoch": 0.28425241614553726,
      "grad_norm": 0.9921669363975525,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 2.4642,
      "step": 125
    },
    {
      "epoch": 0.2865264354747015,
      "grad_norm": 0.9523817896842957,
      "learning_rate": 3.298534127791785e-05,
      "loss": 2.2686,
      "step": 126
    },
    {
      "epoch": 0.28880045480386585,
      "grad_norm": 0.9201560020446777,
      "learning_rate": 3.221030765387417e-05,
      "loss": 2.3529,
      "step": 127
    },
    {
      "epoch": 0.2910744741330301,
      "grad_norm": 1.007020354270935,
      "learning_rate": 3.144013755408895e-05,
      "loss": 2.2531,
      "step": 128
    },
    {
      "epoch": 0.29334849346219444,
      "grad_norm": 0.9516400694847107,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 2.2752,
      "step": 129
    },
    {
      "epoch": 0.2956225127913587,
      "grad_norm": 0.9625292420387268,
      "learning_rate": 2.991522876735154e-05,
      "loss": 2.286,
      "step": 130
    },
    {
      "epoch": 0.297896532120523,
      "grad_norm": 0.9690910577774048,
      "learning_rate": 2.916090697523549e-05,
      "loss": 2.3002,
      "step": 131
    },
    {
      "epoch": 0.3001705514496873,
      "grad_norm": 0.9300558567047119,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 2.3713,
      "step": 132
    },
    {
      "epoch": 0.3024445707788516,
      "grad_norm": 1.0282573699951172,
      "learning_rate": 2.766955965735968e-05,
      "loss": 2.4672,
      "step": 133
    },
    {
      "epoch": 0.30471859010801594,
      "grad_norm": 1.1509264707565308,
      "learning_rate": 2.693294185106562e-05,
      "loss": 2.5568,
      "step": 134
    },
    {
      "epoch": 0.3069926094371802,
      "grad_norm": 1.0650074481964111,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 2.6764,
      "step": 135
    },
    {
      "epoch": 0.30926662876634453,
      "grad_norm": 1.258753776550293,
      "learning_rate": 2.547882480847461e-05,
      "loss": 2.2947,
      "step": 136
    },
    {
      "epoch": 0.3115406480955088,
      "grad_norm": 1.158823847770691,
      "learning_rate": 2.476172311325783e-05,
      "loss": 2.2928,
      "step": 137
    },
    {
      "epoch": 0.3138146674246731,
      "grad_norm": 1.1734964847564697,
      "learning_rate": 2.405152131093926e-05,
      "loss": 2.4324,
      "step": 138
    },
    {
      "epoch": 0.3160886867538374,
      "grad_norm": 1.3184300661087036,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 2.3776,
      "step": 139
    },
    {
      "epoch": 0.3183627060830017,
      "grad_norm": 1.259461760520935,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 2.7089,
      "step": 140
    },
    {
      "epoch": 0.320636725412166,
      "grad_norm": 1.4400967359542847,
      "learning_rate": 2.196424713241637e-05,
      "loss": 2.3936,
      "step": 141
    },
    {
      "epoch": 0.3229107447413303,
      "grad_norm": 1.4404484033584595,
      "learning_rate": 2.128356686585282e-05,
      "loss": 2.6887,
      "step": 142
    },
    {
      "epoch": 0.3251847640704946,
      "grad_norm": 1.6879733800888062,
      "learning_rate": 2.061073738537635e-05,
      "loss": 2.7207,
      "step": 143
    },
    {
      "epoch": 0.3274587833996589,
      "grad_norm": 1.8530641794204712,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 2.6506,
      "step": 144
    },
    {
      "epoch": 0.3297328027288232,
      "grad_norm": 1.977170467376709,
      "learning_rate": 1.928936436551661e-05,
      "loss": 2.6748,
      "step": 145
    },
    {
      "epoch": 0.3320068220579875,
      "grad_norm": 2.2924273014068604,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 2.463,
      "step": 146
    },
    {
      "epoch": 0.3342808413871518,
      "grad_norm": 2.352557420730591,
      "learning_rate": 1.800157297483417e-05,
      "loss": 1.8491,
      "step": 147
    },
    {
      "epoch": 0.3365548607163161,
      "grad_norm": 2.775940418243408,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 2.1364,
      "step": 148
    },
    {
      "epoch": 0.3388288800454804,
      "grad_norm": 4.207704544067383,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 2.5829,
      "step": 149
    },
    {
      "epoch": 0.34110289937464466,
      "grad_norm": 8.697121620178223,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 3.9042,
      "step": 150
    },
    {
      "epoch": 0.34110289937464466,
      "eval_loss": 2.341630697250366,
      "eval_runtime": 54.0251,
      "eval_samples_per_second": 13.716,
      "eval_steps_per_second": 3.443,
      "step": 150
    },
    {
      "epoch": 0.343376918703809,
      "grad_norm": 0.6939475536346436,
      "learning_rate": 1.553232954407171e-05,
      "loss": 2.5413,
      "step": 151
    },
    {
      "epoch": 0.34565093803297325,
      "grad_norm": 0.7881330251693726,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 2.0515,
      "step": 152
    },
    {
      "epoch": 0.3479249573621376,
      "grad_norm": 0.6707624197006226,
      "learning_rate": 1.435357758543015e-05,
      "loss": 1.8278,
      "step": 153
    },
    {
      "epoch": 0.3501989766913019,
      "grad_norm": 0.9109097123146057,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 2.0773,
      "step": 154
    },
    {
      "epoch": 0.35247299602046617,
      "grad_norm": 0.8475165367126465,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 2.2884,
      "step": 155
    },
    {
      "epoch": 0.3547470153496305,
      "grad_norm": 0.8534179329872131,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 2.0617,
      "step": 156
    },
    {
      "epoch": 0.35702103467879476,
      "grad_norm": 0.7414795756340027,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 2.1371,
      "step": 157
    },
    {
      "epoch": 0.3592950540079591,
      "grad_norm": 0.7311276793479919,
      "learning_rate": 1.157994445715706e-05,
      "loss": 2.0273,
      "step": 158
    },
    {
      "epoch": 0.36156907333712335,
      "grad_norm": 0.6954187750816345,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 2.1665,
      "step": 159
    },
    {
      "epoch": 0.36384309266628767,
      "grad_norm": 0.7674466967582703,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 2.2331,
      "step": 160
    },
    {
      "epoch": 0.36611711199545194,
      "grad_norm": 0.6683114171028137,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 1.9872,
      "step": 161
    },
    {
      "epoch": 0.36839113132461626,
      "grad_norm": 0.6786666512489319,
      "learning_rate": 9.549150281252633e-06,
      "loss": 2.0092,
      "step": 162
    },
    {
      "epoch": 0.3706651506537806,
      "grad_norm": 0.9155296087265015,
      "learning_rate": 9.068759265665384e-06,
      "loss": 2.2505,
      "step": 163
    },
    {
      "epoch": 0.37293916998294485,
      "grad_norm": 0.7727603316307068,
      "learning_rate": 8.599558442598998e-06,
      "loss": 2.1084,
      "step": 164
    },
    {
      "epoch": 0.3752131893121092,
      "grad_norm": 0.8741194605827332,
      "learning_rate": 8.141676086873572e-06,
      "loss": 2.2731,
      "step": 165
    },
    {
      "epoch": 0.37748720864127344,
      "grad_norm": 0.8787727355957031,
      "learning_rate": 7.695237378953223e-06,
      "loss": 2.2042,
      "step": 166
    },
    {
      "epoch": 0.37976122797043776,
      "grad_norm": 0.8821845054626465,
      "learning_rate": 7.260364370723044e-06,
      "loss": 2.047,
      "step": 167
    },
    {
      "epoch": 0.38203524729960203,
      "grad_norm": 0.9485935568809509,
      "learning_rate": 6.837175952121306e-06,
      "loss": 2.2568,
      "step": 168
    },
    {
      "epoch": 0.38430926662876636,
      "grad_norm": 0.8914996981620789,
      "learning_rate": 6.425787818636131e-06,
      "loss": 2.0959,
      "step": 169
    },
    {
      "epoch": 0.3865832859579306,
      "grad_norm": 0.9226830005645752,
      "learning_rate": 6.026312439675552e-06,
      "loss": 2.3705,
      "step": 170
    },
    {
      "epoch": 0.38885730528709495,
      "grad_norm": 0.9367247819900513,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 2.1546,
      "step": 171
    },
    {
      "epoch": 0.3911313246162592,
      "grad_norm": 0.9327746629714966,
      "learning_rate": 5.263533508961827e-06,
      "loss": 2.0204,
      "step": 172
    },
    {
      "epoch": 0.39340534394542354,
      "grad_norm": 0.8738958239555359,
      "learning_rate": 4.900438493352055e-06,
      "loss": 2.2742,
      "step": 173
    },
    {
      "epoch": 0.39567936327458786,
      "grad_norm": 0.9805490970611572,
      "learning_rate": 4.549673247541875e-06,
      "loss": 2.3245,
      "step": 174
    },
    {
      "epoch": 0.3979533826037521,
      "grad_norm": 1.0276888608932495,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 2.5121,
      "step": 175
    },
    {
      "epoch": 0.40022740193291645,
      "grad_norm": 0.9224311709403992,
      "learning_rate": 3.885512251130763e-06,
      "loss": 2.3173,
      "step": 176
    },
    {
      "epoch": 0.4025014212620807,
      "grad_norm": 0.9187234044075012,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 2.3004,
      "step": 177
    },
    {
      "epoch": 0.40477544059124504,
      "grad_norm": 0.9445127844810486,
      "learning_rate": 3.271776770026963e-06,
      "loss": 2.3174,
      "step": 178
    },
    {
      "epoch": 0.4070494599204093,
      "grad_norm": 0.9968290328979492,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 2.364,
      "step": 179
    },
    {
      "epoch": 0.40932347924957363,
      "grad_norm": 1.039773941040039,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 2.2251,
      "step": 180
    },
    {
      "epoch": 0.4115974985787379,
      "grad_norm": 1.0480740070343018,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 2.3684,
      "step": 181
    },
    {
      "epoch": 0.4138715179079022,
      "grad_norm": 1.0252336263656616,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 2.3768,
      "step": 182
    },
    {
      "epoch": 0.4161455372370665,
      "grad_norm": 1.0584899187088013,
      "learning_rate": 1.962316193157593e-06,
      "loss": 2.4886,
      "step": 183
    },
    {
      "epoch": 0.4184195565662308,
      "grad_norm": 1.1426482200622559,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 2.6109,
      "step": 184
    },
    {
      "epoch": 0.42069357589539513,
      "grad_norm": 1.0886224508285522,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 2.3443,
      "step": 185
    },
    {
      "epoch": 0.4229675952245594,
      "grad_norm": 1.2631548643112183,
      "learning_rate": 1.333670137599713e-06,
      "loss": 2.616,
      "step": 186
    },
    {
      "epoch": 0.4252416145537237,
      "grad_norm": 1.2152425050735474,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 2.7226,
      "step": 187
    },
    {
      "epoch": 0.427515633882888,
      "grad_norm": 1.182738184928894,
      "learning_rate": 9.810017062595322e-07,
      "loss": 2.8306,
      "step": 188
    },
    {
      "epoch": 0.4297896532120523,
      "grad_norm": 1.398547887802124,
      "learning_rate": 8.247462563808817e-07,
      "loss": 2.415,
      "step": 189
    },
    {
      "epoch": 0.4320636725412166,
      "grad_norm": 1.4223021268844604,
      "learning_rate": 6.819348298638839e-07,
      "loss": 2.5084,
      "step": 190
    },
    {
      "epoch": 0.4343376918703809,
      "grad_norm": 1.538164734840393,
      "learning_rate": 5.526064699265753e-07,
      "loss": 2.6586,
      "step": 191
    },
    {
      "epoch": 0.43661171119954517,
      "grad_norm": 1.709108829498291,
      "learning_rate": 4.367965336512403e-07,
      "loss": 2.7876,
      "step": 192
    },
    {
      "epoch": 0.4388857305287095,
      "grad_norm": 1.65629243850708,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 2.4725,
      "step": 193
    },
    {
      "epoch": 0.4411597498578738,
      "grad_norm": 1.861850380897522,
      "learning_rate": 2.458548727494292e-07,
      "loss": 2.6326,
      "step": 194
    },
    {
      "epoch": 0.4434337691870381,
      "grad_norm": 2.0244126319885254,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 2.6424,
      "step": 195
    },
    {
      "epoch": 0.4457077885162024,
      "grad_norm": 2.3574278354644775,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 2.3744,
      "step": 196
    },
    {
      "epoch": 0.4479818078453667,
      "grad_norm": 2.3692092895507812,
      "learning_rate": 6.150154258476315e-08,
      "loss": 2.4769,
      "step": 197
    },
    {
      "epoch": 0.450255827174531,
      "grad_norm": 2.747682571411133,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 1.8513,
      "step": 198
    },
    {
      "epoch": 0.45252984650369527,
      "grad_norm": 3.3357551097869873,
      "learning_rate": 6.834750376549792e-09,
      "loss": 2.9307,
      "step": 199
    },
    {
      "epoch": 0.4548038658328596,
      "grad_norm": 6.950160503387451,
      "learning_rate": 0.0,
      "loss": 4.593,
      "step": 200
    },
    {
      "epoch": 0.4548038658328596,
      "eval_loss": 2.317739963531494,
      "eval_runtime": 54.0297,
      "eval_samples_per_second": 13.715,
      "eval_steps_per_second": 3.443,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.8899887529590784e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}