{
  "best_metric": 0.06726283580064774,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 2.867383512544803,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.014336917562724014,
      "grad_norm": 4.397572040557861,
      "learning_rate": 1e-05,
      "loss": 1.4088,
      "step": 1
    },
    {
      "epoch": 0.014336917562724014,
      "eval_loss": 2.468245267868042,
      "eval_runtime": 8.862,
      "eval_samples_per_second": 13.315,
      "eval_steps_per_second": 3.385,
      "step": 1
    },
    {
      "epoch": 0.02867383512544803,
      "grad_norm": 4.511147975921631,
      "learning_rate": 2e-05,
      "loss": 1.561,
      "step": 2
    },
    {
      "epoch": 0.043010752688172046,
      "grad_norm": 4.272532939910889,
      "learning_rate": 3e-05,
      "loss": 1.4871,
      "step": 3
    },
    {
      "epoch": 0.05734767025089606,
      "grad_norm": 4.45158052444458,
      "learning_rate": 4e-05,
      "loss": 1.5105,
      "step": 4
    },
    {
      "epoch": 0.07168458781362007,
      "grad_norm": 3.4182779788970947,
      "learning_rate": 5e-05,
      "loss": 1.2247,
      "step": 5
    },
    {
      "epoch": 0.08602150537634409,
      "grad_norm": 4.2475266456604,
      "learning_rate": 6e-05,
      "loss": 1.2059,
      "step": 6
    },
    {
      "epoch": 0.1003584229390681,
      "grad_norm": 2.925811290740967,
      "learning_rate": 7e-05,
      "loss": 1.2442,
      "step": 7
    },
    {
      "epoch": 0.11469534050179211,
      "grad_norm": 2.881817102432251,
      "learning_rate": 8e-05,
      "loss": 0.8519,
      "step": 8
    },
    {
      "epoch": 0.12903225806451613,
      "grad_norm": 2.842601776123047,
      "learning_rate": 9e-05,
      "loss": 1.0145,
      "step": 9
    },
    {
      "epoch": 0.14336917562724014,
      "grad_norm": 3.05037260055542,
      "learning_rate": 0.0001,
      "loss": 1.0831,
      "step": 10
    },
    {
      "epoch": 0.15770609318996415,
      "grad_norm": 3.3153631687164307,
      "learning_rate": 9.999316524962345e-05,
      "loss": 1.1391,
      "step": 11
    },
    {
      "epoch": 0.17204301075268819,
      "grad_norm": 2.482717514038086,
      "learning_rate": 9.997266286704631e-05,
      "loss": 0.9365,
      "step": 12
    },
    {
      "epoch": 0.1863799283154122,
      "grad_norm": 3.5255157947540283,
      "learning_rate": 9.993849845741524e-05,
      "loss": 1.7734,
      "step": 13
    },
    {
      "epoch": 0.2007168458781362,
      "grad_norm": 4.361560821533203,
      "learning_rate": 9.989068136093873e-05,
      "loss": 1.7419,
      "step": 14
    },
    {
      "epoch": 0.21505376344086022,
      "grad_norm": 4.079258918762207,
      "learning_rate": 9.98292246503335e-05,
      "loss": 1.8014,
      "step": 15
    },
    {
      "epoch": 0.22939068100358423,
      "grad_norm": 7.080872058868408,
      "learning_rate": 9.975414512725057e-05,
      "loss": 2.249,
      "step": 16
    },
    {
      "epoch": 0.24372759856630824,
      "grad_norm": 5.132293701171875,
      "learning_rate": 9.966546331768191e-05,
      "loss": 2.1615,
      "step": 17
    },
    {
      "epoch": 0.25806451612903225,
      "grad_norm": 4.942999362945557,
      "learning_rate": 9.956320346634876e-05,
      "loss": 1.2654,
      "step": 18
    },
    {
      "epoch": 0.2724014336917563,
      "grad_norm": 2.344113826751709,
      "learning_rate": 9.944739353007344e-05,
      "loss": 1.245,
      "step": 19
    },
    {
      "epoch": 0.2867383512544803,
      "grad_norm": 2.540186882019043,
      "learning_rate": 9.931806517013612e-05,
      "loss": 1.1376,
      "step": 20
    },
    {
      "epoch": 0.3010752688172043,
      "grad_norm": 2.5160861015319824,
      "learning_rate": 9.917525374361912e-05,
      "loss": 1.0088,
      "step": 21
    },
    {
      "epoch": 0.3154121863799283,
      "grad_norm": 1.9440033435821533,
      "learning_rate": 9.901899829374047e-05,
      "loss": 1.0809,
      "step": 22
    },
    {
      "epoch": 0.32974910394265233,
      "grad_norm": 1.7402596473693848,
      "learning_rate": 9.884934153917997e-05,
      "loss": 0.7672,
      "step": 23
    },
    {
      "epoch": 0.34408602150537637,
      "grad_norm": 1.8408048152923584,
      "learning_rate": 9.86663298624003e-05,
      "loss": 0.6858,
      "step": 24
    },
    {
      "epoch": 0.35842293906810035,
      "grad_norm": 2.228644371032715,
      "learning_rate": 9.847001329696653e-05,
      "loss": 0.7948,
      "step": 25
    },
    {
      "epoch": 0.3727598566308244,
      "grad_norm": 2.7800474166870117,
      "learning_rate": 9.826044551386744e-05,
      "loss": 1.2284,
      "step": 26
    },
    {
      "epoch": 0.3870967741935484,
      "grad_norm": 2.4595513343811035,
      "learning_rate": 9.803768380684242e-05,
      "loss": 0.9438,
      "step": 27
    },
    {
      "epoch": 0.4014336917562724,
      "grad_norm": 2.3911242485046387,
      "learning_rate": 9.780178907671789e-05,
      "loss": 0.7394,
      "step": 28
    },
    {
      "epoch": 0.4157706093189964,
      "grad_norm": 4.901264667510986,
      "learning_rate": 9.755282581475769e-05,
      "loss": 0.8924,
      "step": 29
    },
    {
      "epoch": 0.43010752688172044,
      "grad_norm": 2.3451790809631348,
      "learning_rate": 9.729086208503174e-05,
      "loss": 0.9093,
      "step": 30
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 3.2730343341827393,
      "learning_rate": 9.701596950580806e-05,
      "loss": 1.117,
      "step": 31
    },
    {
      "epoch": 0.45878136200716846,
      "grad_norm": 5.669613838195801,
      "learning_rate": 9.672822322997305e-05,
      "loss": 1.7776,
      "step": 32
    },
    {
      "epoch": 0.4731182795698925,
      "grad_norm": 5.203555107116699,
      "learning_rate": 9.642770192448536e-05,
      "loss": 1.2931,
      "step": 33
    },
    {
      "epoch": 0.4874551971326165,
      "grad_norm": 3.978545665740967,
      "learning_rate": 9.611448774886924e-05,
      "loss": 1.0465,
      "step": 34
    },
    {
      "epoch": 0.5017921146953405,
      "grad_norm": 1.8522157669067383,
      "learning_rate": 9.578866633275288e-05,
      "loss": 1.0426,
      "step": 35
    },
    {
      "epoch": 0.5161290322580645,
      "grad_norm": 2.190333127975464,
      "learning_rate": 9.545032675245813e-05,
      "loss": 0.7932,
      "step": 36
    },
    {
      "epoch": 0.5304659498207885,
      "grad_norm": 2.331299066543579,
      "learning_rate": 9.509956150664796e-05,
      "loss": 1.0709,
      "step": 37
    },
    {
      "epoch": 0.5448028673835126,
      "grad_norm": 1.870590329170227,
      "learning_rate": 9.473646649103818e-05,
      "loss": 0.9257,
      "step": 38
    },
    {
      "epoch": 0.5591397849462365,
      "grad_norm": 1.6125731468200684,
      "learning_rate": 9.43611409721806e-05,
      "loss": 0.5924,
      "step": 39
    },
    {
      "epoch": 0.5734767025089605,
      "grad_norm": 1.4967056512832642,
      "learning_rate": 9.397368756032445e-05,
      "loss": 0.7173,
      "step": 40
    },
    {
      "epoch": 0.5878136200716846,
      "grad_norm": 2.7689082622528076,
      "learning_rate": 9.357421218136386e-05,
      "loss": 0.5854,
      "step": 41
    },
    {
      "epoch": 0.6021505376344086,
      "grad_norm": 1.8559356927871704,
      "learning_rate": 9.316282404787871e-05,
      "loss": 0.6817,
      "step": 42
    },
    {
      "epoch": 0.6164874551971327,
      "grad_norm": 2.5627405643463135,
      "learning_rate": 9.273963562927695e-05,
      "loss": 0.9151,
      "step": 43
    },
    {
      "epoch": 0.6308243727598566,
      "grad_norm": 2.89658522605896,
      "learning_rate": 9.230476262104677e-05,
      "loss": 0.745,
      "step": 44
    },
    {
      "epoch": 0.6451612903225806,
      "grad_norm": 2.8034560680389404,
      "learning_rate": 9.185832391312644e-05,
      "loss": 0.7944,
      "step": 45
    },
    {
      "epoch": 0.6594982078853047,
      "grad_norm": 2.722273588180542,
      "learning_rate": 9.140044155740101e-05,
      "loss": 0.463,
      "step": 46
    },
    {
      "epoch": 0.6738351254480287,
      "grad_norm": 3.258568525314331,
      "learning_rate": 9.093124073433463e-05,
      "loss": 0.7951,
      "step": 47
    },
    {
      "epoch": 0.6881720430107527,
      "grad_norm": 2.5290963649749756,
      "learning_rate": 9.045084971874738e-05,
      "loss": 0.7495,
      "step": 48
    },
    {
      "epoch": 0.7025089605734767,
      "grad_norm": 3.949486017227173,
      "learning_rate": 8.995939984474624e-05,
      "loss": 0.7659,
      "step": 49
    },
    {
      "epoch": 0.7168458781362007,
      "grad_norm": 5.4951958656311035,
      "learning_rate": 8.945702546981969e-05,
      "loss": 1.1883,
      "step": 50
    },
    {
      "epoch": 0.7168458781362007,
      "eval_loss": 0.6906208395957947,
      "eval_runtime": 9.0646,
      "eval_samples_per_second": 13.018,
      "eval_steps_per_second": 3.31,
      "step": 50
    },
    {
      "epoch": 0.7311827956989247,
      "grad_norm": 6.160062313079834,
      "learning_rate": 8.894386393810563e-05,
      "loss": 0.9228,
      "step": 51
    },
    {
      "epoch": 0.7455197132616488,
      "grad_norm": 2.088334560394287,
      "learning_rate": 8.842005554284296e-05,
      "loss": 0.9856,
      "step": 52
    },
    {
      "epoch": 0.7598566308243727,
      "grad_norm": 1.7493679523468018,
      "learning_rate": 8.788574348801675e-05,
      "loss": 0.7416,
      "step": 53
    },
    {
      "epoch": 0.7741935483870968,
      "grad_norm": 2.019888401031494,
      "learning_rate": 8.73410738492077e-05,
      "loss": 0.8308,
      "step": 54
    },
    {
      "epoch": 0.7885304659498208,
      "grad_norm": 2.5577056407928467,
      "learning_rate": 8.678619553365659e-05,
      "loss": 0.7716,
      "step": 55
    },
    {
      "epoch": 0.8028673835125448,
      "grad_norm": 1.479081630706787,
      "learning_rate": 8.622126023955446e-05,
      "loss": 0.4901,
      "step": 56
    },
    {
      "epoch": 0.8172043010752689,
      "grad_norm": 1.7726107835769653,
      "learning_rate": 8.564642241456986e-05,
      "loss": 0.7134,
      "step": 57
    },
    {
      "epoch": 0.8315412186379928,
      "grad_norm": 1.445792317390442,
      "learning_rate": 8.506183921362443e-05,
      "loss": 0.3662,
      "step": 58
    },
    {
      "epoch": 0.8458781362007168,
      "grad_norm": 1.3742926120758057,
      "learning_rate": 8.44676704559283e-05,
      "loss": 0.3006,
      "step": 59
    },
    {
      "epoch": 0.8602150537634409,
      "grad_norm": 1.7996090650558472,
      "learning_rate": 8.386407858128706e-05,
      "loss": 0.568,
      "step": 60
    },
    {
      "epoch": 0.8745519713261649,
      "grad_norm": 1.690359115600586,
      "learning_rate": 8.32512286056924e-05,
      "loss": 0.3693,
      "step": 61
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 1.7139132022857666,
      "learning_rate": 8.262928807620843e-05,
      "loss": 0.3151,
      "step": 62
    },
    {
      "epoch": 0.9032258064516129,
      "grad_norm": 2.8020946979522705,
      "learning_rate": 8.199842702516583e-05,
      "loss": 0.5861,
      "step": 63
    },
    {
      "epoch": 0.9175627240143369,
      "grad_norm": 3.937044858932495,
      "learning_rate": 8.135881792367686e-05,
      "loss": 1.0549,
      "step": 64
    },
    {
      "epoch": 0.931899641577061,
      "grad_norm": 2.774298667907715,
      "learning_rate": 8.07106356344834e-05,
      "loss": 0.6135,
      "step": 65
    },
    {
      "epoch": 0.946236559139785,
      "grad_norm": 3.47070574760437,
      "learning_rate": 8.005405736415126e-05,
      "loss": 0.6331,
      "step": 66
    },
    {
      "epoch": 0.9605734767025089,
      "grad_norm": 6.632359981536865,
      "learning_rate": 7.938926261462366e-05,
      "loss": 1.5964,
      "step": 67
    },
    {
      "epoch": 0.974910394265233,
      "grad_norm": 21.868614196777344,
      "learning_rate": 7.871643313414718e-05,
      "loss": 1.1257,
      "step": 68
    },
    {
      "epoch": 0.989247311827957,
      "grad_norm": 1.4519816637039185,
      "learning_rate": 7.803575286758364e-05,
      "loss": 0.4882,
      "step": 69
    },
    {
      "epoch": 1.003584229390681,
      "grad_norm": 2.7172324657440186,
      "learning_rate": 7.734740790612136e-05,
      "loss": 0.6589,
      "step": 70
    },
    {
      "epoch": 1.017921146953405,
      "grad_norm": 1.5232489109039307,
      "learning_rate": 7.66515864363997e-05,
      "loss": 0.7278,
      "step": 71
    },
    {
      "epoch": 1.032258064516129,
      "grad_norm": 1.5466943979263306,
      "learning_rate": 7.594847868906076e-05,
      "loss": 0.4031,
      "step": 72
    },
    {
      "epoch": 1.0465949820788532,
      "grad_norm": 1.3596998453140259,
      "learning_rate": 7.52382768867422e-05,
      "loss": 0.3218,
      "step": 73
    },
    {
      "epoch": 1.060931899641577,
      "grad_norm": 1.84239661693573,
      "learning_rate": 7.452117519152542e-05,
      "loss": 0.2823,
      "step": 74
    },
    {
      "epoch": 1.075268817204301,
      "grad_norm": 1.2260679006576538,
      "learning_rate": 7.379736965185368e-05,
      "loss": 0.1385,
      "step": 75
    },
    {
      "epoch": 1.0896057347670252,
      "grad_norm": 1.0787522792816162,
      "learning_rate": 7.30670581489344e-05,
      "loss": 0.1525,
      "step": 76
    },
    {
      "epoch": 1.103942652329749,
      "grad_norm": 0.7422471642494202,
      "learning_rate": 7.233044034264034e-05,
      "loss": 0.0784,
      "step": 77
    },
    {
      "epoch": 1.118279569892473,
      "grad_norm": 4.13701868057251,
      "learning_rate": 7.158771761692464e-05,
      "loss": 0.0828,
      "step": 78
    },
    {
      "epoch": 1.1326164874551972,
      "grad_norm": 1.8935657739639282,
      "learning_rate": 7.083909302476453e-05,
      "loss": 0.2191,
      "step": 79
    },
    {
      "epoch": 1.146953405017921,
      "grad_norm": 1.5780242681503296,
      "learning_rate": 7.008477123264848e-05,
      "loss": 0.1151,
      "step": 80
    },
    {
      "epoch": 1.1612903225806452,
      "grad_norm": 1.499728798866272,
      "learning_rate": 6.932495846462261e-05,
      "loss": 0.1268,
      "step": 81
    },
    {
      "epoch": 1.1756272401433692,
      "grad_norm": 2.2440967559814453,
      "learning_rate": 6.855986244591104e-05,
      "loss": 0.1854,
      "step": 82
    },
    {
      "epoch": 1.189964157706093,
      "grad_norm": 2.2842183113098145,
      "learning_rate": 6.778969234612584e-05,
      "loss": 0.2224,
      "step": 83
    },
    {
      "epoch": 1.2043010752688172,
      "grad_norm": 0.9249707460403442,
      "learning_rate": 6.701465872208216e-05,
      "loss": 0.0577,
      "step": 84
    },
    {
      "epoch": 1.2186379928315412,
      "grad_norm": 3.4146389961242676,
      "learning_rate": 6.623497346023418e-05,
      "loss": 0.3028,
      "step": 85
    },
    {
      "epoch": 1.2329749103942653,
      "grad_norm": 2.9881715774536133,
      "learning_rate": 6.545084971874738e-05,
      "loss": 0.2318,
      "step": 86
    },
    {
      "epoch": 1.2473118279569892,
      "grad_norm": 3.518266439437866,
      "learning_rate": 6.466250186922325e-05,
      "loss": 0.364,
      "step": 87
    },
    {
      "epoch": 1.2616487455197132,
      "grad_norm": 2.379763603210449,
      "learning_rate": 6.387014543809223e-05,
      "loss": 0.4057,
      "step": 88
    },
    {
      "epoch": 1.2759856630824373,
      "grad_norm": 1.829189658164978,
      "learning_rate": 6.307399704769099e-05,
      "loss": 0.229,
      "step": 89
    },
    {
      "epoch": 1.2903225806451613,
      "grad_norm": 3.376854419708252,
      "learning_rate": 6.227427435703997e-05,
      "loss": 0.4505,
      "step": 90
    },
    {
      "epoch": 1.3046594982078852,
      "grad_norm": 1.835128903388977,
      "learning_rate": 6.147119600233758e-05,
      "loss": 0.216,
      "step": 91
    },
    {
      "epoch": 1.3189964157706093,
      "grad_norm": 1.298094630241394,
      "learning_rate": 6.066498153718735e-05,
      "loss": 0.1513,
      "step": 92
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 1.650740623474121,
      "learning_rate": 5.985585137257401e-05,
      "loss": 0.1077,
      "step": 93
    },
    {
      "epoch": 1.3476702508960574,
      "grad_norm": 0.939468264579773,
      "learning_rate": 5.90440267166055e-05,
      "loss": 0.0623,
      "step": 94
    },
    {
      "epoch": 1.3620071684587813,
      "grad_norm": 2.0720038414001465,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 0.0929,
      "step": 95
    },
    {
      "epoch": 1.3763440860215055,
      "grad_norm": 1.627712368965149,
      "learning_rate": 5.74131823855921e-05,
      "loss": 0.0955,
      "step": 96
    },
    {
      "epoch": 1.3906810035842294,
      "grad_norm": 1.253098726272583,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 0.1399,
      "step": 97
    },
    {
      "epoch": 1.4050179211469533,
      "grad_norm": 1.75249183177948,
      "learning_rate": 5.577423184847932e-05,
      "loss": 0.1307,
      "step": 98
    },
    {
      "epoch": 1.4193548387096775,
      "grad_norm": 2.218088150024414,
      "learning_rate": 5.495227651252315e-05,
      "loss": 0.1233,
      "step": 99
    },
    {
      "epoch": 1.4336917562724014,
      "grad_norm": 1.65668523311615,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 0.0858,
      "step": 100
    },
    {
      "epoch": 1.4336917562724014,
      "eval_loss": 0.24534527957439423,
      "eval_runtime": 9.0659,
      "eval_samples_per_second": 13.016,
      "eval_steps_per_second": 3.309,
      "step": 100
    },
    {
      "epoch": 1.4480286738351253,
      "grad_norm": 1.6351250410079956,
      "learning_rate": 5.330452921628497e-05,
      "loss": 0.1017,
      "step": 101
    },
    {
      "epoch": 1.4623655913978495,
      "grad_norm": 1.4549205303192139,
      "learning_rate": 5.247918773366112e-05,
      "loss": 0.0528,
      "step": 102
    },
    {
      "epoch": 1.4767025089605734,
      "grad_norm": 3.6143550872802734,
      "learning_rate": 5.165316846586541e-05,
      "loss": 0.3822,
      "step": 103
    },
    {
      "epoch": 1.4910394265232976,
      "grad_norm": 5.097026824951172,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 0.2987,
      "step": 104
    },
    {
      "epoch": 1.5053763440860215,
      "grad_norm": 1.1341527700424194,
      "learning_rate": 5e-05,
      "loss": 0.2981,
      "step": 105
    },
    {
      "epoch": 1.5197132616487457,
      "grad_norm": 1.0630345344543457,
      "learning_rate": 4.917330276168208e-05,
      "loss": 0.1787,
      "step": 106
    },
    {
      "epoch": 1.5340501792114696,
      "grad_norm": 1.177541732788086,
      "learning_rate": 4.834683153413459e-05,
      "loss": 0.2361,
      "step": 107
    },
    {
      "epoch": 1.5483870967741935,
      "grad_norm": 0.8346509337425232,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 0.1039,
      "step": 108
    },
    {
      "epoch": 1.5627240143369177,
      "grad_norm": 0.5888307094573975,
      "learning_rate": 4.669547078371504e-05,
      "loss": 0.0521,
      "step": 109
    },
    {
      "epoch": 1.5770609318996416,
      "grad_norm": 0.7871260643005371,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 0.086,
      "step": 110
    },
    {
      "epoch": 1.5913978494623655,
      "grad_norm": 1.8992468118667603,
      "learning_rate": 4.504772348747687e-05,
      "loss": 0.1583,
      "step": 111
    },
    {
      "epoch": 1.6057347670250897,
      "grad_norm": 0.7448009252548218,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 0.0613,
      "step": 112
    },
    {
      "epoch": 1.6200716845878136,
      "grad_norm": 0.8584257364273071,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 0.0603,
      "step": 113
    },
    {
      "epoch": 1.6344086021505375,
      "grad_norm": 1.1192561388015747,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 0.047,
      "step": 114
    },
    {
      "epoch": 1.6487455197132617,
      "grad_norm": 2.2471425533294678,
      "learning_rate": 4.17702704859633e-05,
      "loss": 0.0816,
      "step": 115
    },
    {
      "epoch": 1.6630824372759858,
      "grad_norm": 2.3649044036865234,
      "learning_rate": 4.095597328339452e-05,
      "loss": 0.1115,
      "step": 116
    },
    {
      "epoch": 1.6774193548387095,
      "grad_norm": 1.4149099588394165,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 0.0971,
      "step": 117
    },
    {
      "epoch": 1.6917562724014337,
      "grad_norm": 2.165285348892212,
      "learning_rate": 3.933501846281267e-05,
      "loss": 0.116,
      "step": 118
    },
    {
      "epoch": 1.7060931899641578,
      "grad_norm": 2.6265251636505127,
      "learning_rate": 3.852880399766243e-05,
      "loss": 0.114,
      "step": 119
    },
    {
      "epoch": 1.7204301075268817,
      "grad_norm": 4.279717445373535,
      "learning_rate": 3.772572564296005e-05,
      "loss": 0.2685,
      "step": 120
    },
    {
      "epoch": 1.7347670250896057,
      "grad_norm": 5.294130325317383,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 0.0826,
      "step": 121
    },
    {
      "epoch": 1.7491039426523298,
      "grad_norm": 1.007126808166504,
      "learning_rate": 3.612985456190778e-05,
      "loss": 0.1608,
      "step": 122
    },
    {
      "epoch": 1.7634408602150538,
      "grad_norm": 0.9493341445922852,
      "learning_rate": 3.533749813077677e-05,
      "loss": 0.0896,
      "step": 123
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 1.1440672874450684,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 0.0875,
      "step": 124
    },
    {
      "epoch": 1.7921146953405018,
      "grad_norm": 1.3371349573135376,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 0.0676,
      "step": 125
    },
    {
      "epoch": 1.8064516129032258,
      "grad_norm": 0.7539826035499573,
      "learning_rate": 3.298534127791785e-05,
      "loss": 0.0332,
      "step": 126
    },
    {
      "epoch": 1.8207885304659497,
      "grad_norm": 0.33484286069869995,
      "learning_rate": 3.221030765387417e-05,
      "loss": 0.0144,
      "step": 127
    },
    {
      "epoch": 1.8351254480286738,
      "grad_norm": 1.6466560363769531,
      "learning_rate": 3.144013755408895e-05,
      "loss": 0.0616,
      "step": 128
    },
    {
      "epoch": 1.849462365591398,
      "grad_norm": 1.9090458154678345,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 0.0569,
      "step": 129
    },
    {
      "epoch": 1.863799283154122,
      "grad_norm": 1.0547633171081543,
      "learning_rate": 2.991522876735154e-05,
      "loss": 0.0328,
      "step": 130
    },
    {
      "epoch": 1.8781362007168458,
      "grad_norm": 1.2474104166030884,
      "learning_rate": 2.916090697523549e-05,
      "loss": 0.0182,
      "step": 131
    },
    {
      "epoch": 1.89247311827957,
      "grad_norm": 0.9929400086402893,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 0.0502,
      "step": 132
    },
    {
      "epoch": 1.906810035842294,
      "grad_norm": 1.7062305212020874,
      "learning_rate": 2.766955965735968e-05,
      "loss": 0.0833,
      "step": 133
    },
    {
      "epoch": 1.9211469534050178,
      "grad_norm": 1.6501058340072632,
      "learning_rate": 2.693294185106562e-05,
      "loss": 0.129,
      "step": 134
    },
    {
      "epoch": 1.935483870967742,
      "grad_norm": 0.9860913753509521,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 0.0311,
      "step": 135
    },
    {
      "epoch": 1.949820788530466,
      "grad_norm": 2.3620800971984863,
      "learning_rate": 2.547882480847461e-05,
      "loss": 0.0973,
      "step": 136
    },
    {
      "epoch": 1.9641577060931898,
      "grad_norm": 1.6745470762252808,
      "learning_rate": 2.476172311325783e-05,
      "loss": 0.0537,
      "step": 137
    },
    {
      "epoch": 1.978494623655914,
      "grad_norm": 1.2220505475997925,
      "learning_rate": 2.405152131093926e-05,
      "loss": 0.0539,
      "step": 138
    },
    {
      "epoch": 1.9928315412186381,
      "grad_norm": 0.7096848487854004,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 0.0229,
      "step": 139
    },
    {
      "epoch": 2.007168458781362,
      "grad_norm": 2.7335457801818848,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 0.1315,
      "step": 140
    },
    {
      "epoch": 2.021505376344086,
      "grad_norm": 1.0886074304580688,
      "learning_rate": 2.196424713241637e-05,
      "loss": 0.08,
      "step": 141
    },
    {
      "epoch": 2.03584229390681,
      "grad_norm": 0.9922390580177307,
      "learning_rate": 2.128356686585282e-05,
      "loss": 0.0281,
      "step": 142
    },
    {
      "epoch": 2.050179211469534,
      "grad_norm": 0.5020017623901367,
      "learning_rate": 2.061073738537635e-05,
      "loss": 0.0237,
      "step": 143
    },
    {
      "epoch": 2.064516129032258,
      "grad_norm": 0.5839765667915344,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 0.0383,
      "step": 144
    },
    {
      "epoch": 2.078853046594982,
      "grad_norm": 0.09325302392244339,
      "learning_rate": 1.928936436551661e-05,
      "loss": 0.0041,
      "step": 145
    },
    {
      "epoch": 2.0931899641577063,
      "grad_norm": 1.7177072763442993,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 0.0599,
      "step": 146
    },
    {
      "epoch": 2.10752688172043,
      "grad_norm": 0.23047403991222382,
      "learning_rate": 1.800157297483417e-05,
      "loss": 0.0084,
      "step": 147
    },
    {
      "epoch": 2.121863799283154,
      "grad_norm": 0.3760106563568115,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 0.0125,
      "step": 148
    },
    {
      "epoch": 2.1362007168458783,
      "grad_norm": 1.6124687194824219,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 0.0325,
      "step": 149
    },
    {
      "epoch": 2.150537634408602,
      "grad_norm": 1.5602785348892212,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 0.0162,
      "step": 150
    },
    {
      "epoch": 2.150537634408602,
      "eval_loss": 0.09235768765211105,
      "eval_runtime": 9.0495,
      "eval_samples_per_second": 13.039,
      "eval_steps_per_second": 3.315,
      "step": 150
    },
    {
      "epoch": 2.164874551971326,
      "grad_norm": 0.5806900262832642,
      "learning_rate": 1.553232954407171e-05,
      "loss": 0.0398,
      "step": 151
    },
    {
      "epoch": 2.1792114695340503,
      "grad_norm": 0.2753426730632782,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 0.0098,
      "step": 152
    },
    {
      "epoch": 2.193548387096774,
      "grad_norm": 0.5121511220932007,
      "learning_rate": 1.435357758543015e-05,
      "loss": 0.0084,
      "step": 153
    },
    {
      "epoch": 2.207885304659498,
      "grad_norm": 1.5801509618759155,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 0.0287,
      "step": 154
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.7592269778251648,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 0.0196,
      "step": 155
    },
    {
      "epoch": 2.236559139784946,
      "grad_norm": 0.948007345199585,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 0.0271,
      "step": 156
    },
    {
      "epoch": 2.25089605734767,
      "grad_norm": 1.3356605768203735,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 0.0285,
      "step": 157
    },
    {
      "epoch": 2.2652329749103943,
      "grad_norm": 0.8015672564506531,
      "learning_rate": 1.157994445715706e-05,
      "loss": 0.071,
      "step": 158
    },
    {
      "epoch": 2.279569892473118,
      "grad_norm": 0.6799566745758057,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 0.0374,
      "step": 159
    },
    {
      "epoch": 2.293906810035842,
      "grad_norm": 0.2814430594444275,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 0.0102,
      "step": 160
    },
    {
      "epoch": 2.3082437275985663,
      "grad_norm": 0.7916092276573181,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 0.0204,
      "step": 161
    },
    {
      "epoch": 2.3225806451612905,
      "grad_norm": 0.15406902134418488,
      "learning_rate": 9.549150281252633e-06,
      "loss": 0.0037,
      "step": 162
    },
    {
      "epoch": 2.336917562724014,
      "grad_norm": 0.5541296005249023,
      "learning_rate": 9.068759265665384e-06,
      "loss": 0.0093,
      "step": 163
    },
    {
      "epoch": 2.3512544802867383,
      "grad_norm": 0.35052183270454407,
      "learning_rate": 8.599558442598998e-06,
      "loss": 0.0092,
      "step": 164
    },
    {
      "epoch": 2.3655913978494625,
      "grad_norm": 0.12663590908050537,
      "learning_rate": 8.141676086873572e-06,
      "loss": 0.0045,
      "step": 165
    },
    {
      "epoch": 2.379928315412186,
      "grad_norm": 0.14168612658977509,
      "learning_rate": 7.695237378953223e-06,
      "loss": 0.0031,
      "step": 166
    },
    {
      "epoch": 2.3942652329749103,
      "grad_norm": 0.1643611192703247,
      "learning_rate": 7.260364370723044e-06,
      "loss": 0.003,
      "step": 167
    },
    {
      "epoch": 2.4086021505376345,
      "grad_norm": 0.22675709426403046,
      "learning_rate": 6.837175952121306e-06,
      "loss": 0.0055,
      "step": 168
    },
    {
      "epoch": 2.4229390681003586,
      "grad_norm": 0.17917326092720032,
      "learning_rate": 6.425787818636131e-06,
      "loss": 0.0036,
      "step": 169
    },
    {
      "epoch": 2.4372759856630823,
      "grad_norm": 0.26549574732780457,
      "learning_rate": 6.026312439675552e-06,
      "loss": 0.0033,
      "step": 170
    },
    {
      "epoch": 2.4516129032258065,
      "grad_norm": 0.10524545609951019,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 0.0028,
      "step": 171
    },
    {
      "epoch": 2.4659498207885306,
      "grad_norm": 1.0003790855407715,
      "learning_rate": 5.263533508961827e-06,
      "loss": 0.0118,
      "step": 172
    },
    {
      "epoch": 2.4802867383512543,
      "grad_norm": 0.2203439176082611,
      "learning_rate": 4.900438493352055e-06,
      "loss": 0.0057,
      "step": 173
    },
    {
      "epoch": 2.4946236559139785,
      "grad_norm": 1.8931493759155273,
      "learning_rate": 4.549673247541875e-06,
      "loss": 0.0702,
      "step": 174
    },
    {
      "epoch": 2.5089605734767026,
      "grad_norm": 0.815489649772644,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 0.0228,
      "step": 175
    },
    {
      "epoch": 2.5232974910394264,
      "grad_norm": 1.1504946947097778,
      "learning_rate": 3.885512251130763e-06,
      "loss": 0.0451,
      "step": 176
    },
    {
      "epoch": 2.5376344086021505,
      "grad_norm": 0.3113609254360199,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 0.0137,
      "step": 177
    },
    {
      "epoch": 2.5519713261648747,
      "grad_norm": 0.14661143720149994,
      "learning_rate": 3.271776770026963e-06,
      "loss": 0.0058,
      "step": 178
    },
    {
      "epoch": 2.5663082437275984,
      "grad_norm": 0.23132531344890594,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 0.0047,
      "step": 179
    },
    {
      "epoch": 2.5806451612903225,
      "grad_norm": 0.28742915391921997,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 0.0048,
      "step": 180
    },
    {
      "epoch": 2.5949820788530467,
      "grad_norm": 0.21918118000030518,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 0.0044,
      "step": 181
    },
    {
      "epoch": 2.6093189964157704,
      "grad_norm": 0.8730088472366333,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 0.0205,
      "step": 182
    },
    {
      "epoch": 2.6236559139784945,
      "grad_norm": 0.046389706432819366,
      "learning_rate": 1.962316193157593e-06,
      "loss": 0.0015,
      "step": 183
    },
    {
      "epoch": 2.6379928315412187,
      "grad_norm": 0.6478707194328308,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 0.0182,
      "step": 184
    },
    {
      "epoch": 2.652329749103943,
      "grad_norm": 0.5956841111183167,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 0.0154,
      "step": 185
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.265004962682724,
      "learning_rate": 1.333670137599713e-06,
      "loss": 0.0053,
      "step": 186
    },
    {
      "epoch": 2.6810035842293907,
      "grad_norm": 0.08419239521026611,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 0.0024,
      "step": 187
    },
    {
      "epoch": 2.695340501792115,
      "grad_norm": 0.18191705644130707,
      "learning_rate": 9.810017062595322e-07,
      "loss": 0.0043,
      "step": 188
    },
    {
      "epoch": 2.709677419354839,
      "grad_norm": 0.36635175347328186,
      "learning_rate": 8.247462563808817e-07,
      "loss": 0.0059,
      "step": 189
    },
    {
      "epoch": 2.7240143369175627,
      "grad_norm": 0.18067115545272827,
      "learning_rate": 6.819348298638839e-07,
      "loss": 0.0054,
      "step": 190
    },
    {
      "epoch": 2.738351254480287,
      "grad_norm": 0.5659708976745605,
      "learning_rate": 5.526064699265753e-07,
      "loss": 0.0434,
      "step": 191
    },
    {
      "epoch": 2.752688172043011,
      "grad_norm": 0.3749392032623291,
      "learning_rate": 4.367965336512403e-07,
      "loss": 0.0317,
      "step": 192
    },
    {
      "epoch": 2.7670250896057347,
      "grad_norm": 0.6255916357040405,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 0.0391,
      "step": 193
    },
    {
      "epoch": 2.781362007168459,
      "grad_norm": 1.4000941514968872,
      "learning_rate": 2.458548727494292e-07,
      "loss": 0.0303,
      "step": 194
    },
    {
      "epoch": 2.795698924731183,
      "grad_norm": 0.14650675654411316,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 0.0059,
      "step": 195
    },
    {
      "epoch": 2.8100358422939067,
      "grad_norm": 0.22659583389759064,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 0.005,
      "step": 196
    },
    {
      "epoch": 2.824372759856631,
      "grad_norm": 0.2989434599876404,
      "learning_rate": 6.150154258476315e-08,
      "loss": 0.0048,
      "step": 197
    },
    {
      "epoch": 2.838709677419355,
      "grad_norm": 0.0759616494178772,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 0.0019,
      "step": 198
    },
    {
      "epoch": 2.8530465949820787,
      "grad_norm": 0.19209331274032593,
      "learning_rate": 6.834750376549792e-09,
      "loss": 0.0074,
      "step": 199
    },
    {
      "epoch": 2.867383512544803,
      "grad_norm": 0.16473375260829926,
      "learning_rate": 0.0,
      "loss": 0.0045,
      "step": 200
    },
    {
      "epoch": 2.867383512544803,
      "eval_loss": 0.06726283580064774,
      "eval_runtime": 9.0539,
      "eval_samples_per_second": 13.033,
      "eval_steps_per_second": 3.313,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.017027157491712e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}