{
  "best_metric": 4.333376884460449,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.3766478342749529,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0018832391713747645,
      "grad_norm": 0.4404832422733307,
      "learning_rate": 1e-05,
      "loss": 4.8125,
      "step": 1
    },
    {
      "epoch": 0.0018832391713747645,
      "eval_loss": 5.504382133483887,
      "eval_runtime": 26.0352,
      "eval_samples_per_second": 34.376,
      "eval_steps_per_second": 8.604,
      "step": 1
    },
    {
      "epoch": 0.003766478342749529,
      "grad_norm": 0.6471639275550842,
      "learning_rate": 2e-05,
      "loss": 5.0705,
      "step": 2
    },
    {
      "epoch": 0.005649717514124294,
      "grad_norm": 0.6837315559387207,
      "learning_rate": 3e-05,
      "loss": 5.0085,
      "step": 3
    },
    {
      "epoch": 0.007532956685499058,
      "grad_norm": 0.8111864328384399,
      "learning_rate": 4e-05,
      "loss": 5.0946,
      "step": 4
    },
    {
      "epoch": 0.009416195856873822,
      "grad_norm": 0.8689448237419128,
      "learning_rate": 5e-05,
      "loss": 4.959,
      "step": 5
    },
    {
      "epoch": 0.011299435028248588,
      "grad_norm": 0.869692862033844,
      "learning_rate": 6e-05,
      "loss": 4.9862,
      "step": 6
    },
    {
      "epoch": 0.013182674199623353,
      "grad_norm": 0.8494324684143066,
      "learning_rate": 7e-05,
      "loss": 5.0878,
      "step": 7
    },
    {
      "epoch": 0.015065913370998116,
      "grad_norm": 0.8554010987281799,
      "learning_rate": 8e-05,
      "loss": 4.982,
      "step": 8
    },
    {
      "epoch": 0.01694915254237288,
      "grad_norm": 0.8704142570495605,
      "learning_rate": 9e-05,
      "loss": 5.0892,
      "step": 9
    },
    {
      "epoch": 0.018832391713747645,
      "grad_norm": 0.8269019722938538,
      "learning_rate": 0.0001,
      "loss": 4.9553,
      "step": 10
    },
    {
      "epoch": 0.02071563088512241,
      "grad_norm": 0.8658633232116699,
      "learning_rate": 9.999316524962345e-05,
      "loss": 4.8072,
      "step": 11
    },
    {
      "epoch": 0.022598870056497175,
      "grad_norm": 0.8111575841903687,
      "learning_rate": 9.997266286704631e-05,
      "loss": 4.9093,
      "step": 12
    },
    {
      "epoch": 0.02448210922787194,
      "grad_norm": 0.9275814890861511,
      "learning_rate": 9.993849845741524e-05,
      "loss": 4.9829,
      "step": 13
    },
    {
      "epoch": 0.026365348399246705,
      "grad_norm": 0.9555933475494385,
      "learning_rate": 9.989068136093873e-05,
      "loss": 4.7234,
      "step": 14
    },
    {
      "epoch": 0.02824858757062147,
      "grad_norm": 1.1188527345657349,
      "learning_rate": 9.98292246503335e-05,
      "loss": 4.9422,
      "step": 15
    },
    {
      "epoch": 0.030131826741996232,
      "grad_norm": 0.9981339573860168,
      "learning_rate": 9.975414512725057e-05,
      "loss": 5.0508,
      "step": 16
    },
    {
      "epoch": 0.032015065913371,
      "grad_norm": 1.0851399898529053,
      "learning_rate": 9.966546331768191e-05,
      "loss": 4.6557,
      "step": 17
    },
    {
      "epoch": 0.03389830508474576,
      "grad_norm": 1.0240473747253418,
      "learning_rate": 9.956320346634876e-05,
      "loss": 4.944,
      "step": 18
    },
    {
      "epoch": 0.035781544256120526,
      "grad_norm": 1.0421777963638306,
      "learning_rate": 9.944739353007344e-05,
      "loss": 4.621,
      "step": 19
    },
    {
      "epoch": 0.03766478342749529,
      "grad_norm": 1.1771303415298462,
      "learning_rate": 9.931806517013612e-05,
      "loss": 4.4454,
      "step": 20
    },
    {
      "epoch": 0.03954802259887006,
      "grad_norm": 1.3488545417785645,
      "learning_rate": 9.917525374361912e-05,
      "loss": 4.5895,
      "step": 21
    },
    {
      "epoch": 0.04143126177024482,
      "grad_norm": 1.1995290517807007,
      "learning_rate": 9.901899829374047e-05,
      "loss": 4.6091,
      "step": 22
    },
    {
      "epoch": 0.04331450094161959,
      "grad_norm": 1.287221074104309,
      "learning_rate": 9.884934153917997e-05,
      "loss": 4.6541,
      "step": 23
    },
    {
      "epoch": 0.04519774011299435,
      "grad_norm": 1.298819661140442,
      "learning_rate": 9.86663298624003e-05,
      "loss": 4.7365,
      "step": 24
    },
    {
      "epoch": 0.047080979284369114,
      "grad_norm": 1.3010953664779663,
      "learning_rate": 9.847001329696653e-05,
      "loss": 4.5698,
      "step": 25
    },
    {
      "epoch": 0.04896421845574388,
      "grad_norm": 1.2813111543655396,
      "learning_rate": 9.826044551386744e-05,
      "loss": 4.3968,
      "step": 26
    },
    {
      "epoch": 0.05084745762711865,
      "grad_norm": 1.3757303953170776,
      "learning_rate": 9.803768380684242e-05,
      "loss": 4.5344,
      "step": 27
    },
    {
      "epoch": 0.05273069679849341,
      "grad_norm": 1.2203274965286255,
      "learning_rate": 9.780178907671789e-05,
      "loss": 4.2782,
      "step": 28
    },
    {
      "epoch": 0.054613935969868174,
      "grad_norm": 1.3546189069747925,
      "learning_rate": 9.755282581475769e-05,
      "loss": 4.5428,
      "step": 29
    },
    {
      "epoch": 0.05649717514124294,
      "grad_norm": 1.406325101852417,
      "learning_rate": 9.729086208503174e-05,
      "loss": 4.5178,
      "step": 30
    },
    {
      "epoch": 0.0583804143126177,
      "grad_norm": 1.4857614040374756,
      "learning_rate": 9.701596950580806e-05,
      "loss": 4.1759,
      "step": 31
    },
    {
      "epoch": 0.060263653483992465,
      "grad_norm": 1.390303134918213,
      "learning_rate": 9.672822322997305e-05,
      "loss": 4.1171,
      "step": 32
    },
    {
      "epoch": 0.062146892655367235,
      "grad_norm": 1.440464973449707,
      "learning_rate": 9.642770192448536e-05,
      "loss": 4.0926,
      "step": 33
    },
    {
      "epoch": 0.064030131826742,
      "grad_norm": 1.4431653022766113,
      "learning_rate": 9.611448774886924e-05,
      "loss": 4.3693,
      "step": 34
    },
    {
      "epoch": 0.06591337099811675,
      "grad_norm": 1.4864574670791626,
      "learning_rate": 9.578866633275288e-05,
      "loss": 4.2244,
      "step": 35
    },
    {
      "epoch": 0.06779661016949153,
      "grad_norm": 1.362823247909546,
      "learning_rate": 9.545032675245813e-05,
      "loss": 4.275,
      "step": 36
    },
    {
      "epoch": 0.0696798493408663,
      "grad_norm": 1.5752381086349487,
      "learning_rate": 9.509956150664796e-05,
      "loss": 4.0307,
      "step": 37
    },
    {
      "epoch": 0.07156308851224105,
      "grad_norm": 1.6116201877593994,
      "learning_rate": 9.473646649103818e-05,
      "loss": 4.084,
      "step": 38
    },
    {
      "epoch": 0.07344632768361582,
      "grad_norm": 1.554311752319336,
      "learning_rate": 9.43611409721806e-05,
      "loss": 3.8963,
      "step": 39
    },
    {
      "epoch": 0.07532956685499058,
      "grad_norm": 1.5103185176849365,
      "learning_rate": 9.397368756032445e-05,
      "loss": 3.9708,
      "step": 40
    },
    {
      "epoch": 0.07721280602636535,
      "grad_norm": 1.5464398860931396,
      "learning_rate": 9.357421218136386e-05,
      "loss": 4.2411,
      "step": 41
    },
    {
      "epoch": 0.07909604519774012,
      "grad_norm": 1.6189881563186646,
      "learning_rate": 9.316282404787871e-05,
      "loss": 4.1771,
      "step": 42
    },
    {
      "epoch": 0.08097928436911488,
      "grad_norm": 1.5227197408676147,
      "learning_rate": 9.273963562927695e-05,
      "loss": 3.8867,
      "step": 43
    },
    {
      "epoch": 0.08286252354048965,
      "grad_norm": 1.7572941780090332,
      "learning_rate": 9.230476262104677e-05,
      "loss": 4.266,
      "step": 44
    },
    {
      "epoch": 0.0847457627118644,
      "grad_norm": 1.7709587812423706,
      "learning_rate": 9.185832391312644e-05,
      "loss": 4.5049,
      "step": 45
    },
    {
      "epoch": 0.08662900188323917,
      "grad_norm": 2.164048671722412,
      "learning_rate": 9.140044155740101e-05,
      "loss": 3.9513,
      "step": 46
    },
    {
      "epoch": 0.08851224105461393,
      "grad_norm": 1.8082555532455444,
      "learning_rate": 9.093124073433463e-05,
      "loss": 4.1858,
      "step": 47
    },
    {
      "epoch": 0.0903954802259887,
      "grad_norm": 1.7886871099472046,
      "learning_rate": 9.045084971874738e-05,
      "loss": 4.3065,
      "step": 48
    },
    {
      "epoch": 0.09227871939736347,
      "grad_norm": 2.2777397632598877,
      "learning_rate": 8.995939984474624e-05,
      "loss": 3.9524,
      "step": 49
    },
    {
      "epoch": 0.09416195856873823,
      "grad_norm": 2.2879488468170166,
      "learning_rate": 8.945702546981969e-05,
      "loss": 4.1489,
      "step": 50
    },
    {
      "epoch": 0.09416195856873823,
      "eval_loss": 4.684483051300049,
      "eval_runtime": 26.0146,
      "eval_samples_per_second": 34.404,
      "eval_steps_per_second": 8.611,
      "step": 50
    },
    {
      "epoch": 0.096045197740113,
      "grad_norm": 5.314813137054443,
      "learning_rate": 8.894386393810563e-05,
      "loss": 5.1513,
      "step": 51
    },
    {
      "epoch": 0.09792843691148775,
      "grad_norm": 6.118640422821045,
      "learning_rate": 8.842005554284296e-05,
      "loss": 5.2192,
      "step": 52
    },
    {
      "epoch": 0.09981167608286252,
      "grad_norm": 5.554737567901611,
      "learning_rate": 8.788574348801675e-05,
      "loss": 5.2439,
      "step": 53
    },
    {
      "epoch": 0.1016949152542373,
      "grad_norm": 4.769510746002197,
      "learning_rate": 8.73410738492077e-05,
      "loss": 5.2584,
      "step": 54
    },
    {
      "epoch": 0.10357815442561205,
      "grad_norm": 3.551276922225952,
      "learning_rate": 8.678619553365659e-05,
      "loss": 4.9679,
      "step": 55
    },
    {
      "epoch": 0.10546139359698682,
      "grad_norm": 2.761868953704834,
      "learning_rate": 8.622126023955446e-05,
      "loss": 5.1733,
      "step": 56
    },
    {
      "epoch": 0.10734463276836158,
      "grad_norm": 1.8833056688308716,
      "learning_rate": 8.564642241456986e-05,
      "loss": 4.892,
      "step": 57
    },
    {
      "epoch": 0.10922787193973635,
      "grad_norm": 1.3549888134002686,
      "learning_rate": 8.506183921362443e-05,
      "loss": 4.6912,
      "step": 58
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 1.072119951248169,
      "learning_rate": 8.44676704559283e-05,
      "loss": 4.6438,
      "step": 59
    },
    {
      "epoch": 0.11299435028248588,
      "grad_norm": 0.9934912919998169,
      "learning_rate": 8.386407858128706e-05,
      "loss": 4.4821,
      "step": 60
    },
    {
      "epoch": 0.11487758945386065,
      "grad_norm": 0.9702751636505127,
      "learning_rate": 8.32512286056924e-05,
      "loss": 4.5895,
      "step": 61
    },
    {
      "epoch": 0.1167608286252354,
      "grad_norm": 1.0550380945205688,
      "learning_rate": 8.262928807620843e-05,
      "loss": 4.5458,
      "step": 62
    },
    {
      "epoch": 0.11864406779661017,
      "grad_norm": 1.2004576921463013,
      "learning_rate": 8.199842702516583e-05,
      "loss": 4.6984,
      "step": 63
    },
    {
      "epoch": 0.12052730696798493,
      "grad_norm": 1.3918439149856567,
      "learning_rate": 8.135881792367686e-05,
      "loss": 4.5142,
      "step": 64
    },
    {
      "epoch": 0.1224105461393597,
      "grad_norm": 1.37985098361969,
      "learning_rate": 8.07106356344834e-05,
      "loss": 4.5775,
      "step": 65
    },
    {
      "epoch": 0.12429378531073447,
      "grad_norm": 1.306896686553955,
      "learning_rate": 8.005405736415126e-05,
      "loss": 4.3314,
      "step": 66
    },
    {
      "epoch": 0.12617702448210924,
      "grad_norm": 1.4172680377960205,
      "learning_rate": 7.938926261462366e-05,
      "loss": 4.5173,
      "step": 67
    },
    {
      "epoch": 0.128060263653484,
      "grad_norm": 1.2359857559204102,
      "learning_rate": 7.871643313414718e-05,
      "loss": 4.3144,
      "step": 68
    },
    {
      "epoch": 0.12994350282485875,
      "grad_norm": 1.3013797998428345,
      "learning_rate": 7.803575286758364e-05,
      "loss": 4.6413,
      "step": 69
    },
    {
      "epoch": 0.1318267419962335,
      "grad_norm": 1.1234915256500244,
      "learning_rate": 7.734740790612136e-05,
      "loss": 4.3396,
      "step": 70
    },
    {
      "epoch": 0.1337099811676083,
      "grad_norm": 1.1252583265304565,
      "learning_rate": 7.66515864363997e-05,
      "loss": 4.3902,
      "step": 71
    },
    {
      "epoch": 0.13559322033898305,
      "grad_norm": 1.1471351385116577,
      "learning_rate": 7.594847868906076e-05,
      "loss": 4.1393,
      "step": 72
    },
    {
      "epoch": 0.1374764595103578,
      "grad_norm": 1.1607205867767334,
      "learning_rate": 7.52382768867422e-05,
      "loss": 4.7234,
      "step": 73
    },
    {
      "epoch": 0.1393596986817326,
      "grad_norm": 1.203909158706665,
      "learning_rate": 7.452117519152542e-05,
      "loss": 4.163,
      "step": 74
    },
    {
      "epoch": 0.14124293785310735,
      "grad_norm": 1.2247579097747803,
      "learning_rate": 7.379736965185368e-05,
      "loss": 4.6043,
      "step": 75
    },
    {
      "epoch": 0.1431261770244821,
      "grad_norm": 1.2941570281982422,
      "learning_rate": 7.30670581489344e-05,
      "loss": 4.6977,
      "step": 76
    },
    {
      "epoch": 0.14500941619585686,
      "grad_norm": 1.5688599348068237,
      "learning_rate": 7.233044034264034e-05,
      "loss": 4.1314,
      "step": 77
    },
    {
      "epoch": 0.14689265536723164,
      "grad_norm": 1.1933225393295288,
      "learning_rate": 7.158771761692464e-05,
      "loss": 4.1808,
      "step": 78
    },
    {
      "epoch": 0.1487758945386064,
      "grad_norm": 1.246169090270996,
      "learning_rate": 7.083909302476453e-05,
      "loss": 3.9987,
      "step": 79
    },
    {
      "epoch": 0.15065913370998116,
      "grad_norm": 1.4048924446105957,
      "learning_rate": 7.008477123264848e-05,
      "loss": 4.2167,
      "step": 80
    },
    {
      "epoch": 0.15254237288135594,
      "grad_norm": 1.4151579141616821,
      "learning_rate": 6.932495846462261e-05,
      "loss": 4.4665,
      "step": 81
    },
    {
      "epoch": 0.1544256120527307,
      "grad_norm": 1.6417031288146973,
      "learning_rate": 6.855986244591104e-05,
      "loss": 4.558,
      "step": 82
    },
    {
      "epoch": 0.15630885122410546,
      "grad_norm": 1.3999319076538086,
      "learning_rate": 6.778969234612584e-05,
      "loss": 4.4642,
      "step": 83
    },
    {
      "epoch": 0.15819209039548024,
      "grad_norm": 1.7804279327392578,
      "learning_rate": 6.701465872208216e-05,
      "loss": 4.0368,
      "step": 84
    },
    {
      "epoch": 0.160075329566855,
      "grad_norm": 1.3715370893478394,
      "learning_rate": 6.623497346023418e-05,
      "loss": 4.1679,
      "step": 85
    },
    {
      "epoch": 0.16195856873822975,
      "grad_norm": 1.4773646593093872,
      "learning_rate": 6.545084971874738e-05,
      "loss": 4.1398,
      "step": 86
    },
    {
      "epoch": 0.1638418079096045,
      "grad_norm": 1.4795664548873901,
      "learning_rate": 6.466250186922325e-05,
      "loss": 4.1285,
      "step": 87
    },
    {
      "epoch": 0.1657250470809793,
      "grad_norm": 1.4087882041931152,
      "learning_rate": 6.387014543809223e-05,
      "loss": 4.2199,
      "step": 88
    },
    {
      "epoch": 0.16760828625235405,
      "grad_norm": 1.5499064922332764,
      "learning_rate": 6.307399704769099e-05,
      "loss": 4.2428,
      "step": 89
    },
    {
      "epoch": 0.1694915254237288,
      "grad_norm": 1.529011845588684,
      "learning_rate": 6.227427435703997e-05,
      "loss": 4.0152,
      "step": 90
    },
    {
      "epoch": 0.1713747645951036,
      "grad_norm": 1.5831725597381592,
      "learning_rate": 6.147119600233758e-05,
      "loss": 3.9022,
      "step": 91
    },
    {
      "epoch": 0.17325800376647835,
      "grad_norm": 1.588663935661316,
      "learning_rate": 6.066498153718735e-05,
      "loss": 4.1288,
      "step": 92
    },
    {
      "epoch": 0.1751412429378531,
      "grad_norm": 2.129978895187378,
      "learning_rate": 5.985585137257401e-05,
      "loss": 4.1492,
      "step": 93
    },
    {
      "epoch": 0.17702448210922786,
      "grad_norm": 1.656499981880188,
      "learning_rate": 5.90440267166055e-05,
      "loss": 3.6049,
      "step": 94
    },
    {
      "epoch": 0.17890772128060264,
      "grad_norm": 1.8470999002456665,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 4.2987,
      "step": 95
    },
    {
      "epoch": 0.1807909604519774,
      "grad_norm": 1.5709669589996338,
      "learning_rate": 5.74131823855921e-05,
      "loss": 4.2057,
      "step": 96
    },
    {
      "epoch": 0.18267419962335216,
      "grad_norm": 1.9587165117263794,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 4.1114,
      "step": 97
    },
    {
      "epoch": 0.18455743879472694,
      "grad_norm": 1.8411070108413696,
      "learning_rate": 5.577423184847932e-05,
      "loss": 4.1232,
      "step": 98
    },
    {
      "epoch": 0.1864406779661017,
      "grad_norm": 1.8720942735671997,
      "learning_rate": 5.495227651252315e-05,
      "loss": 4.5076,
      "step": 99
    },
    {
      "epoch": 0.18832391713747645,
      "grad_norm": 2.048525094985962,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 4.3497,
      "step": 100
    },
    {
      "epoch": 0.18832391713747645,
      "eval_loss": 4.46586275100708,
      "eval_runtime": 26.0362,
      "eval_samples_per_second": 34.375,
      "eval_steps_per_second": 8.603,
      "step": 100
    },
    {
      "epoch": 0.1902071563088512,
      "grad_norm": 2.4281506538391113,
      "learning_rate": 5.330452921628497e-05,
      "loss": 4.6306,
      "step": 101
    },
    {
      "epoch": 0.192090395480226,
      "grad_norm": 3.126316785812378,
      "learning_rate": 5.247918773366112e-05,
      "loss": 5.0779,
      "step": 102
    },
    {
      "epoch": 0.19397363465160075,
      "grad_norm": 3.111471652984619,
      "learning_rate": 5.165316846586541e-05,
      "loss": 5.0164,
      "step": 103
    },
    {
      "epoch": 0.1958568738229755,
      "grad_norm": 3.085820436477661,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 4.9277,
      "step": 104
    },
    {
      "epoch": 0.1977401129943503,
      "grad_norm": 2.9912514686584473,
      "learning_rate": 5e-05,
      "loss": 5.0844,
      "step": 105
    },
    {
      "epoch": 0.19962335216572505,
      "grad_norm": 2.4392735958099365,
      "learning_rate": 4.917330276168208e-05,
      "loss": 5.0779,
      "step": 106
    },
    {
      "epoch": 0.2015065913370998,
      "grad_norm": 2.4553072452545166,
      "learning_rate": 4.834683153413459e-05,
      "loss": 4.7664,
      "step": 107
    },
    {
      "epoch": 0.2033898305084746,
      "grad_norm": 1.8265289068222046,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 4.9092,
      "step": 108
    },
    {
      "epoch": 0.20527306967984935,
      "grad_norm": 1.5107651948928833,
      "learning_rate": 4.669547078371504e-05,
      "loss": 4.6377,
      "step": 109
    },
    {
      "epoch": 0.2071563088512241,
      "grad_norm": 1.4672855138778687,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 4.7664,
      "step": 110
    },
    {
      "epoch": 0.20903954802259886,
      "grad_norm": 1.1098840236663818,
      "learning_rate": 4.504772348747687e-05,
      "loss": 4.4854,
      "step": 111
    },
    {
      "epoch": 0.21092278719397364,
      "grad_norm": 0.9786215424537659,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 4.4619,
      "step": 112
    },
    {
      "epoch": 0.2128060263653484,
      "grad_norm": 0.914353609085083,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 4.425,
      "step": 113
    },
    {
      "epoch": 0.21468926553672316,
      "grad_norm": 1.0050040483474731,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 4.5643,
      "step": 114
    },
    {
      "epoch": 0.21657250470809794,
      "grad_norm": 1.0604749917984009,
      "learning_rate": 4.17702704859633e-05,
      "loss": 4.2492,
      "step": 115
    },
    {
      "epoch": 0.2184557438794727,
      "grad_norm": 0.993858814239502,
      "learning_rate": 4.095597328339452e-05,
      "loss": 4.3766,
      "step": 116
    },
    {
      "epoch": 0.22033898305084745,
      "grad_norm": 1.1347594261169434,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 4.151,
      "step": 117
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 1.2211852073669434,
      "learning_rate": 3.933501846281267e-05,
      "loss": 4.2032,
      "step": 118
    },
    {
      "epoch": 0.224105461393597,
      "grad_norm": 1.202661395072937,
      "learning_rate": 3.852880399766243e-05,
      "loss": 4.7303,
      "step": 119
    },
    {
      "epoch": 0.22598870056497175,
      "grad_norm": 1.207620620727539,
      "learning_rate": 3.772572564296005e-05,
      "loss": 4.3258,
      "step": 120
    },
    {
      "epoch": 0.2278719397363465,
      "grad_norm": 1.3141003847122192,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 4.7308,
      "step": 121
    },
    {
      "epoch": 0.2297551789077213,
      "grad_norm": 1.4259010553359985,
      "learning_rate": 3.612985456190778e-05,
      "loss": 4.235,
      "step": 122
    },
    {
      "epoch": 0.23163841807909605,
      "grad_norm": 1.3282722234725952,
      "learning_rate": 3.533749813077677e-05,
      "loss": 4.0164,
      "step": 123
    },
    {
      "epoch": 0.2335216572504708,
      "grad_norm": 1.407547950744629,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 4.2051,
      "step": 124
    },
    {
      "epoch": 0.23540489642184556,
      "grad_norm": 1.3534866571426392,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 3.9517,
      "step": 125
    },
    {
      "epoch": 0.23728813559322035,
      "grad_norm": 1.348299264907837,
      "learning_rate": 3.298534127791785e-05,
      "loss": 4.3554,
      "step": 126
    },
    {
      "epoch": 0.2391713747645951,
      "grad_norm": 1.460729956626892,
      "learning_rate": 3.221030765387417e-05,
      "loss": 4.1674,
      "step": 127
    },
    {
      "epoch": 0.24105461393596986,
      "grad_norm": 1.4006423950195312,
      "learning_rate": 3.144013755408895e-05,
      "loss": 3.7897,
      "step": 128
    },
    {
      "epoch": 0.24293785310734464,
      "grad_norm": 1.503464698791504,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 4.288,
      "step": 129
    },
    {
      "epoch": 0.2448210922787194,
      "grad_norm": 1.2829421758651733,
      "learning_rate": 2.991522876735154e-05,
      "loss": 3.9962,
      "step": 130
    },
    {
      "epoch": 0.24670433145009416,
      "grad_norm": 1.4906039237976074,
      "learning_rate": 2.916090697523549e-05,
      "loss": 4.2905,
      "step": 131
    },
    {
      "epoch": 0.24858757062146894,
      "grad_norm": 1.3797855377197266,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 4.2297,
      "step": 132
    },
    {
      "epoch": 0.2504708097928437,
      "grad_norm": 1.5242537260055542,
      "learning_rate": 2.766955965735968e-05,
      "loss": 4.1111,
      "step": 133
    },
    {
      "epoch": 0.2523540489642185,
      "grad_norm": 1.470442295074463,
      "learning_rate": 2.693294185106562e-05,
      "loss": 4.4072,
      "step": 134
    },
    {
      "epoch": 0.2542372881355932,
      "grad_norm": 1.4676114320755005,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 4.3991,
      "step": 135
    },
    {
      "epoch": 0.256120527306968,
      "grad_norm": 1.4559814929962158,
      "learning_rate": 2.547882480847461e-05,
      "loss": 4.3998,
      "step": 136
    },
    {
      "epoch": 0.2580037664783427,
      "grad_norm": 1.5525643825531006,
      "learning_rate": 2.476172311325783e-05,
      "loss": 3.8932,
      "step": 137
    },
    {
      "epoch": 0.2598870056497175,
      "grad_norm": 1.4801933765411377,
      "learning_rate": 2.405152131093926e-05,
      "loss": 3.8249,
      "step": 138
    },
    {
      "epoch": 0.2617702448210923,
      "grad_norm": 1.4712077379226685,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 3.6408,
      "step": 139
    },
    {
      "epoch": 0.263653483992467,
      "grad_norm": 1.5557100772857666,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 4.0423,
      "step": 140
    },
    {
      "epoch": 0.2655367231638418,
      "grad_norm": 1.6846270561218262,
      "learning_rate": 2.196424713241637e-05,
      "loss": 4.1841,
      "step": 141
    },
    {
      "epoch": 0.2674199623352166,
      "grad_norm": 1.8425194025039673,
      "learning_rate": 2.128356686585282e-05,
      "loss": 3.7632,
      "step": 142
    },
    {
      "epoch": 0.2693032015065913,
      "grad_norm": 1.6457120180130005,
      "learning_rate": 2.061073738537635e-05,
      "loss": 3.9849,
      "step": 143
    },
    {
      "epoch": 0.2711864406779661,
      "grad_norm": 1.670451045036316,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 3.9007,
      "step": 144
    },
    {
      "epoch": 0.2730696798493409,
      "grad_norm": 1.7191650867462158,
      "learning_rate": 1.928936436551661e-05,
      "loss": 4.2836,
      "step": 145
    },
    {
      "epoch": 0.2749529190207156,
      "grad_norm": 1.6189135313034058,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 3.8274,
      "step": 146
    },
    {
      "epoch": 0.2768361581920904,
      "grad_norm": 1.9642584323883057,
      "learning_rate": 1.800157297483417e-05,
      "loss": 3.8536,
      "step": 147
    },
    {
      "epoch": 0.2787193973634652,
      "grad_norm": 1.8275858163833618,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 4.0992,
      "step": 148
    },
    {
      "epoch": 0.2806026365348399,
      "grad_norm": 2.0902316570281982,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 3.9778,
      "step": 149
    },
    {
      "epoch": 0.2824858757062147,
      "grad_norm": 2.241020679473877,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 4.2315,
      "step": 150
    },
    {
      "epoch": 0.2824858757062147,
      "eval_loss": 4.364166736602783,
      "eval_runtime": 26.0267,
      "eval_samples_per_second": 34.388,
      "eval_steps_per_second": 8.607,
      "step": 150
    },
    {
      "epoch": 0.2843691148775895,
      "grad_norm": 0.9599138498306274,
      "learning_rate": 1.553232954407171e-05,
      "loss": 4.7042,
      "step": 151
    },
    {
      "epoch": 0.2862523540489642,
      "grad_norm": 1.393892765045166,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 4.885,
      "step": 152
    },
    {
      "epoch": 0.288135593220339,
      "grad_norm": 1.3492121696472168,
      "learning_rate": 1.435357758543015e-05,
      "loss": 4.5615,
      "step": 153
    },
    {
      "epoch": 0.2900188323917137,
      "grad_norm": 1.32272207736969,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 4.7243,
      "step": 154
    },
    {
      "epoch": 0.2919020715630885,
      "grad_norm": 1.6310317516326904,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 4.7071,
      "step": 155
    },
    {
      "epoch": 0.2937853107344633,
      "grad_norm": 1.5734827518463135,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 4.6195,
      "step": 156
    },
    {
      "epoch": 0.295668549905838,
      "grad_norm": 1.6394295692443848,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 4.9109,
      "step": 157
    },
    {
      "epoch": 0.2975517890772128,
      "grad_norm": 1.5407096147537231,
      "learning_rate": 1.157994445715706e-05,
      "loss": 4.6784,
      "step": 158
    },
    {
      "epoch": 0.2994350282485876,
      "grad_norm": 1.5206984281539917,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 4.8502,
      "step": 159
    },
    {
      "epoch": 0.3013182674199623,
      "grad_norm": 1.1881388425827026,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 4.6115,
      "step": 160
    },
    {
      "epoch": 0.3032015065913371,
      "grad_norm": 1.3782165050506592,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 4.6922,
      "step": 161
    },
    {
      "epoch": 0.3050847457627119,
      "grad_norm": 1.2255258560180664,
      "learning_rate": 9.549150281252633e-06,
      "loss": 4.4166,
      "step": 162
    },
    {
      "epoch": 0.3069679849340866,
      "grad_norm": 1.528144359588623,
      "learning_rate": 9.068759265665384e-06,
      "loss": 4.6867,
      "step": 163
    },
    {
      "epoch": 0.3088512241054614,
      "grad_norm": 1.249307632446289,
      "learning_rate": 8.599558442598998e-06,
      "loss": 4.5781,
      "step": 164
    },
    {
      "epoch": 0.3107344632768362,
      "grad_norm": 1.2240351438522339,
      "learning_rate": 8.141676086873572e-06,
      "loss": 4.3626,
      "step": 165
    },
    {
      "epoch": 0.3126177024482109,
      "grad_norm": 1.3092036247253418,
      "learning_rate": 7.695237378953223e-06,
      "loss": 4.5233,
      "step": 166
    },
    {
      "epoch": 0.3145009416195857,
      "grad_norm": 1.1483343839645386,
      "learning_rate": 7.260364370723044e-06,
      "loss": 4.223,
      "step": 167
    },
    {
      "epoch": 0.3163841807909605,
      "grad_norm": 1.2056214809417725,
      "learning_rate": 6.837175952121306e-06,
      "loss": 4.1604,
      "step": 168
    },
    {
      "epoch": 0.3182674199623352,
      "grad_norm": 1.3199313879013062,
      "learning_rate": 6.425787818636131e-06,
      "loss": 4.3708,
      "step": 169
    },
    {
      "epoch": 0.32015065913371,
      "grad_norm": 1.2276356220245361,
      "learning_rate": 6.026312439675552e-06,
      "loss": 4.1844,
      "step": 170
    },
    {
      "epoch": 0.3220338983050847,
      "grad_norm": 1.2125377655029297,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 4.2332,
      "step": 171
    },
    {
      "epoch": 0.3239171374764595,
      "grad_norm": 1.1636658906936646,
      "learning_rate": 5.263533508961827e-06,
      "loss": 4.4785,
      "step": 172
    },
    {
      "epoch": 0.3258003766478343,
      "grad_norm": 1.2514711618423462,
      "learning_rate": 4.900438493352055e-06,
      "loss": 4.2767,
      "step": 173
    },
    {
      "epoch": 0.327683615819209,
      "grad_norm": 1.2968021631240845,
      "learning_rate": 4.549673247541875e-06,
      "loss": 3.9597,
      "step": 174
    },
    {
      "epoch": 0.3295668549905838,
      "grad_norm": 1.1687623262405396,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 4.4077,
      "step": 175
    },
    {
      "epoch": 0.3314500941619586,
      "grad_norm": 1.3746123313903809,
      "learning_rate": 3.885512251130763e-06,
      "loss": 4.1864,
      "step": 176
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 1.2834721803665161,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 4.2694,
      "step": 177
    },
    {
      "epoch": 0.3352165725047081,
      "grad_norm": 1.3037786483764648,
      "learning_rate": 3.271776770026963e-06,
      "loss": 4.3656,
      "step": 178
    },
    {
      "epoch": 0.3370998116760829,
      "grad_norm": 1.340135097503662,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 4.053,
      "step": 179
    },
    {
      "epoch": 0.3389830508474576,
      "grad_norm": 1.3368055820465088,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 4.0245,
      "step": 180
    },
    {
      "epoch": 0.3408662900188324,
      "grad_norm": 1.3717410564422607,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 4.0554,
      "step": 181
    },
    {
      "epoch": 0.3427495291902072,
      "grad_norm": 1.3427554368972778,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 4.2376,
      "step": 182
    },
    {
      "epoch": 0.3446327683615819,
      "grad_norm": 1.4679207801818848,
      "learning_rate": 1.962316193157593e-06,
      "loss": 4.0917,
      "step": 183
    },
    {
      "epoch": 0.3465160075329567,
      "grad_norm": 1.5706599950790405,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 4.0967,
      "step": 184
    },
    {
      "epoch": 0.3483992467043315,
      "grad_norm": 1.4587688446044922,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 4.0887,
      "step": 185
    },
    {
      "epoch": 0.3502824858757062,
      "grad_norm": 1.4248045682907104,
      "learning_rate": 1.333670137599713e-06,
      "loss": 3.9085,
      "step": 186
    },
    {
      "epoch": 0.352165725047081,
      "grad_norm": 1.5704526901245117,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 4.0975,
      "step": 187
    },
    {
      "epoch": 0.3540489642184557,
      "grad_norm": 1.58639657497406,
      "learning_rate": 9.810017062595322e-07,
      "loss": 3.9859,
      "step": 188
    },
    {
      "epoch": 0.3559322033898305,
      "grad_norm": 1.6287637948989868,
      "learning_rate": 8.247462563808817e-07,
      "loss": 3.9519,
      "step": 189
    },
    {
      "epoch": 0.3578154425612053,
      "grad_norm": 1.7928752899169922,
      "learning_rate": 6.819348298638839e-07,
      "loss": 4.2057,
      "step": 190
    },
    {
      "epoch": 0.35969868173258,
      "grad_norm": 1.664750099182129,
      "learning_rate": 5.526064699265753e-07,
      "loss": 4.0619,
      "step": 191
    },
    {
      "epoch": 0.3615819209039548,
      "grad_norm": 1.7163918018341064,
      "learning_rate": 4.367965336512403e-07,
      "loss": 4.0208,
      "step": 192
    },
    {
      "epoch": 0.3634651600753296,
      "grad_norm": 1.8046633005142212,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 4.3861,
      "step": 193
    },
    {
      "epoch": 0.3653483992467043,
      "grad_norm": 1.8558590412139893,
      "learning_rate": 2.458548727494292e-07,
      "loss": 3.8195,
      "step": 194
    },
    {
      "epoch": 0.3672316384180791,
      "grad_norm": 1.8654998540878296,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 4.6079,
      "step": 195
    },
    {
      "epoch": 0.3691148775894539,
      "grad_norm": 1.906052589416504,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 3.9341,
      "step": 196
    },
    {
      "epoch": 0.3709981167608286,
      "grad_norm": 2.1169979572296143,
      "learning_rate": 6.150154258476315e-08,
      "loss": 3.7939,
      "step": 197
    },
    {
      "epoch": 0.3728813559322034,
      "grad_norm": 2.054453134536743,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 4.2131,
      "step": 198
    },
    {
      "epoch": 0.3747645951035782,
      "grad_norm": 2.3749969005584717,
      "learning_rate": 6.834750376549792e-09,
      "loss": 3.9909,
      "step": 199
    },
    {
      "epoch": 0.3766478342749529,
      "grad_norm": 2.8953471183776855,
      "learning_rate": 0.0,
      "loss": 4.4298,
      "step": 200
    },
    {
      "epoch": 0.3766478342749529,
      "eval_loss": 4.333376884460449,
      "eval_runtime": 26.0404,
      "eval_samples_per_second": 34.37,
      "eval_steps_per_second": 8.602,
      "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": 5.4428946137088e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}