{
  "best_metric": 2.077235460281372,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.11761246692149367,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0005880623346074683,
      "grad_norm": 4.865276336669922,
      "learning_rate": 1e-05,
      "loss": 7.461,
      "step": 1
    },
    {
      "epoch": 0.0005880623346074683,
      "eval_loss": 2.480590581893921,
      "eval_runtime": 17.8688,
      "eval_samples_per_second": 160.279,
      "eval_steps_per_second": 40.07,
      "step": 1
    },
    {
      "epoch": 0.0011761246692149367,
      "grad_norm": 5.674622535705566,
      "learning_rate": 2e-05,
      "loss": 9.0449,
      "step": 2
    },
    {
      "epoch": 0.0017641870038224052,
      "grad_norm": 4.216952323913574,
      "learning_rate": 3e-05,
      "loss": 8.8929,
      "step": 3
    },
    {
      "epoch": 0.0023522493384298734,
      "grad_norm": 4.285717010498047,
      "learning_rate": 4e-05,
      "loss": 7.3488,
      "step": 4
    },
    {
      "epoch": 0.002940311673037342,
      "grad_norm": 3.7261674404144287,
      "learning_rate": 5e-05,
      "loss": 9.3057,
      "step": 5
    },
    {
      "epoch": 0.0035283740076448105,
      "grad_norm": 3.2327747344970703,
      "learning_rate": 6e-05,
      "loss": 8.5458,
      "step": 6
    },
    {
      "epoch": 0.004116436342252279,
      "grad_norm": 3.429197311401367,
      "learning_rate": 7e-05,
      "loss": 8.9012,
      "step": 7
    },
    {
      "epoch": 0.004704498676859747,
      "grad_norm": 3.6553750038146973,
      "learning_rate": 8e-05,
      "loss": 9.3626,
      "step": 8
    },
    {
      "epoch": 0.0052925610114672155,
      "grad_norm": 4.141356945037842,
      "learning_rate": 9e-05,
      "loss": 8.7039,
      "step": 9
    },
    {
      "epoch": 0.005880623346074684,
      "grad_norm": 4.04561185836792,
      "learning_rate": 0.0001,
      "loss": 9.0502,
      "step": 10
    },
    {
      "epoch": 0.006468685680682152,
      "grad_norm": 3.977367639541626,
      "learning_rate": 9.999316524962345e-05,
      "loss": 9.1719,
      "step": 11
    },
    {
      "epoch": 0.007056748015289621,
      "grad_norm": 3.731240749359131,
      "learning_rate": 9.997266286704631e-05,
      "loss": 9.0438,
      "step": 12
    },
    {
      "epoch": 0.007644810349897089,
      "grad_norm": 3.447376012802124,
      "learning_rate": 9.993849845741524e-05,
      "loss": 8.9509,
      "step": 13
    },
    {
      "epoch": 0.008232872684504558,
      "grad_norm": 3.764333724975586,
      "learning_rate": 9.989068136093873e-05,
      "loss": 9.4756,
      "step": 14
    },
    {
      "epoch": 0.008820935019112026,
      "grad_norm": 3.6373300552368164,
      "learning_rate": 9.98292246503335e-05,
      "loss": 8.7075,
      "step": 15
    },
    {
      "epoch": 0.009408997353719493,
      "grad_norm": 3.9035732746124268,
      "learning_rate": 9.975414512725057e-05,
      "loss": 10.0699,
      "step": 16
    },
    {
      "epoch": 0.009997059688326962,
      "grad_norm": 3.6659138202667236,
      "learning_rate": 9.966546331768191e-05,
      "loss": 8.4551,
      "step": 17
    },
    {
      "epoch": 0.010585122022934431,
      "grad_norm": 3.843367099761963,
      "learning_rate": 9.956320346634876e-05,
      "loss": 8.87,
      "step": 18
    },
    {
      "epoch": 0.0111731843575419,
      "grad_norm": 3.698371648788452,
      "learning_rate": 9.944739353007344e-05,
      "loss": 8.7306,
      "step": 19
    },
    {
      "epoch": 0.011761246692149369,
      "grad_norm": 3.4617395401000977,
      "learning_rate": 9.931806517013612e-05,
      "loss": 9.4107,
      "step": 20
    },
    {
      "epoch": 0.012349309026756836,
      "grad_norm": 3.9732842445373535,
      "learning_rate": 9.917525374361912e-05,
      "loss": 9.2916,
      "step": 21
    },
    {
      "epoch": 0.012937371361364304,
      "grad_norm": 4.051387786865234,
      "learning_rate": 9.901899829374047e-05,
      "loss": 8.9073,
      "step": 22
    },
    {
      "epoch": 0.013525433695971773,
      "grad_norm": 4.409919738769531,
      "learning_rate": 9.884934153917997e-05,
      "loss": 9.2753,
      "step": 23
    },
    {
      "epoch": 0.014113496030579242,
      "grad_norm": 4.249014854431152,
      "learning_rate": 9.86663298624003e-05,
      "loss": 9.4633,
      "step": 24
    },
    {
      "epoch": 0.014701558365186709,
      "grad_norm": 3.861879587173462,
      "learning_rate": 9.847001329696653e-05,
      "loss": 9.6005,
      "step": 25
    },
    {
      "epoch": 0.015289620699794178,
      "grad_norm": 3.8618669509887695,
      "learning_rate": 9.826044551386744e-05,
      "loss": 9.1563,
      "step": 26
    },
    {
      "epoch": 0.01587768303440165,
      "grad_norm": 4.00145959854126,
      "learning_rate": 9.803768380684242e-05,
      "loss": 8.4538,
      "step": 27
    },
    {
      "epoch": 0.016465745369009115,
      "grad_norm": 4.04928731918335,
      "learning_rate": 9.780178907671789e-05,
      "loss": 8.338,
      "step": 28
    },
    {
      "epoch": 0.017053807703616582,
      "grad_norm": 4.237801551818848,
      "learning_rate": 9.755282581475769e-05,
      "loss": 8.7512,
      "step": 29
    },
    {
      "epoch": 0.017641870038224053,
      "grad_norm": 4.405162811279297,
      "learning_rate": 9.729086208503174e-05,
      "loss": 9.8008,
      "step": 30
    },
    {
      "epoch": 0.01822993237283152,
      "grad_norm": 4.705152988433838,
      "learning_rate": 9.701596950580806e-05,
      "loss": 8.6011,
      "step": 31
    },
    {
      "epoch": 0.018817994707438987,
      "grad_norm": 4.47661018371582,
      "learning_rate": 9.672822322997305e-05,
      "loss": 8.519,
      "step": 32
    },
    {
      "epoch": 0.019406057042046457,
      "grad_norm": 4.636375904083252,
      "learning_rate": 9.642770192448536e-05,
      "loss": 8.2032,
      "step": 33
    },
    {
      "epoch": 0.019994119376653924,
      "grad_norm": 4.961963176727295,
      "learning_rate": 9.611448774886924e-05,
      "loss": 8.2178,
      "step": 34
    },
    {
      "epoch": 0.020582181711261395,
      "grad_norm": 4.983580112457275,
      "learning_rate": 9.578866633275288e-05,
      "loss": 8.5388,
      "step": 35
    },
    {
      "epoch": 0.021170244045868862,
      "grad_norm": 5.467924118041992,
      "learning_rate": 9.545032675245813e-05,
      "loss": 9.5947,
      "step": 36
    },
    {
      "epoch": 0.02175830638047633,
      "grad_norm": 6.1575026512146,
      "learning_rate": 9.509956150664796e-05,
      "loss": 9.3919,
      "step": 37
    },
    {
      "epoch": 0.0223463687150838,
      "grad_norm": 7.173287391662598,
      "learning_rate": 9.473646649103818e-05,
      "loss": 9.6322,
      "step": 38
    },
    {
      "epoch": 0.022934431049691267,
      "grad_norm": 6.465355396270752,
      "learning_rate": 9.43611409721806e-05,
      "loss": 9.8151,
      "step": 39
    },
    {
      "epoch": 0.023522493384298737,
      "grad_norm": 6.790063381195068,
      "learning_rate": 9.397368756032445e-05,
      "loss": 9.1537,
      "step": 40
    },
    {
      "epoch": 0.024110555718906204,
      "grad_norm": 6.897984027862549,
      "learning_rate": 9.357421218136386e-05,
      "loss": 9.7276,
      "step": 41
    },
    {
      "epoch": 0.02469861805351367,
      "grad_norm": 6.992136478424072,
      "learning_rate": 9.316282404787871e-05,
      "loss": 8.6977,
      "step": 42
    },
    {
      "epoch": 0.025286680388121142,
      "grad_norm": 7.082433700561523,
      "learning_rate": 9.273963562927695e-05,
      "loss": 9.1598,
      "step": 43
    },
    {
      "epoch": 0.02587474272272861,
      "grad_norm": 8.282886505126953,
      "learning_rate": 9.230476262104677e-05,
      "loss": 9.7373,
      "step": 44
    },
    {
      "epoch": 0.02646280505733608,
      "grad_norm": 9.197856903076172,
      "learning_rate": 9.185832391312644e-05,
      "loss": 10.3421,
      "step": 45
    },
    {
      "epoch": 0.027050867391943546,
      "grad_norm": 9.766355514526367,
      "learning_rate": 9.140044155740101e-05,
      "loss": 10.3045,
      "step": 46
    },
    {
      "epoch": 0.027638929726551013,
      "grad_norm": 11.853132247924805,
      "learning_rate": 9.093124073433463e-05,
      "loss": 8.2403,
      "step": 47
    },
    {
      "epoch": 0.028226992061158484,
      "grad_norm": 15.254680633544922,
      "learning_rate": 9.045084971874738e-05,
      "loss": 9.5235,
      "step": 48
    },
    {
      "epoch": 0.02881505439576595,
      "grad_norm": 18.938486099243164,
      "learning_rate": 8.995939984474624e-05,
      "loss": 9.0537,
      "step": 49
    },
    {
      "epoch": 0.029403116730373418,
      "grad_norm": 49.306671142578125,
      "learning_rate": 8.945702546981969e-05,
      "loss": 8.2417,
      "step": 50
    },
    {
      "epoch": 0.029403116730373418,
      "eval_loss": 2.3457820415496826,
      "eval_runtime": 18.0941,
      "eval_samples_per_second": 158.284,
      "eval_steps_per_second": 39.571,
      "step": 50
    },
    {
      "epoch": 0.02999117906498089,
      "grad_norm": 32.40308380126953,
      "learning_rate": 8.894386393810563e-05,
      "loss": 9.4112,
      "step": 51
    },
    {
      "epoch": 0.030579241399588356,
      "grad_norm": 24.540586471557617,
      "learning_rate": 8.842005554284296e-05,
      "loss": 9.1602,
      "step": 52
    },
    {
      "epoch": 0.031167303734195826,
      "grad_norm": 17.45161247253418,
      "learning_rate": 8.788574348801675e-05,
      "loss": 8.9991,
      "step": 53
    },
    {
      "epoch": 0.0317553660688033,
      "grad_norm": 13.077903747558594,
      "learning_rate": 8.73410738492077e-05,
      "loss": 8.1767,
      "step": 54
    },
    {
      "epoch": 0.032343428403410764,
      "grad_norm": 10.01981258392334,
      "learning_rate": 8.678619553365659e-05,
      "loss": 8.9059,
      "step": 55
    },
    {
      "epoch": 0.03293149073801823,
      "grad_norm": 7.11581563949585,
      "learning_rate": 8.622126023955446e-05,
      "loss": 8.8765,
      "step": 56
    },
    {
      "epoch": 0.0335195530726257,
      "grad_norm": 5.8939924240112305,
      "learning_rate": 8.564642241456986e-05,
      "loss": 8.7036,
      "step": 57
    },
    {
      "epoch": 0.034107615407233165,
      "grad_norm": 5.505342960357666,
      "learning_rate": 8.506183921362443e-05,
      "loss": 9.1436,
      "step": 58
    },
    {
      "epoch": 0.03469567774184063,
      "grad_norm": 4.9250969886779785,
      "learning_rate": 8.44676704559283e-05,
      "loss": 9.4932,
      "step": 59
    },
    {
      "epoch": 0.035283740076448106,
      "grad_norm": 4.490935802459717,
      "learning_rate": 8.386407858128706e-05,
      "loss": 8.656,
      "step": 60
    },
    {
      "epoch": 0.03587180241105557,
      "grad_norm": 3.4468441009521484,
      "learning_rate": 8.32512286056924e-05,
      "loss": 9.0581,
      "step": 61
    },
    {
      "epoch": 0.03645986474566304,
      "grad_norm": 3.4602138996124268,
      "learning_rate": 8.262928807620843e-05,
      "loss": 9.4731,
      "step": 62
    },
    {
      "epoch": 0.03704792708027051,
      "grad_norm": 3.181739091873169,
      "learning_rate": 8.199842702516583e-05,
      "loss": 8.2736,
      "step": 63
    },
    {
      "epoch": 0.037635989414877974,
      "grad_norm": 3.4806437492370605,
      "learning_rate": 8.135881792367686e-05,
      "loss": 8.4785,
      "step": 64
    },
    {
      "epoch": 0.03822405174948545,
      "grad_norm": 3.207629680633545,
      "learning_rate": 8.07106356344834e-05,
      "loss": 8.9628,
      "step": 65
    },
    {
      "epoch": 0.038812114084092915,
      "grad_norm": 2.971317768096924,
      "learning_rate": 8.005405736415126e-05,
      "loss": 8.7241,
      "step": 66
    },
    {
      "epoch": 0.03940017641870038,
      "grad_norm": 3.1931509971618652,
      "learning_rate": 7.938926261462366e-05,
      "loss": 8.6304,
      "step": 67
    },
    {
      "epoch": 0.03998823875330785,
      "grad_norm": 3.18424129486084,
      "learning_rate": 7.871643313414718e-05,
      "loss": 8.5659,
      "step": 68
    },
    {
      "epoch": 0.040576301087915316,
      "grad_norm": 3.107588529586792,
      "learning_rate": 7.803575286758364e-05,
      "loss": 8.3743,
      "step": 69
    },
    {
      "epoch": 0.04116436342252279,
      "grad_norm": 3.182344913482666,
      "learning_rate": 7.734740790612136e-05,
      "loss": 8.4377,
      "step": 70
    },
    {
      "epoch": 0.04175242575713026,
      "grad_norm": 3.2035622596740723,
      "learning_rate": 7.66515864363997e-05,
      "loss": 8.6552,
      "step": 71
    },
    {
      "epoch": 0.042340488091737724,
      "grad_norm": 3.233053684234619,
      "learning_rate": 7.594847868906076e-05,
      "loss": 8.7825,
      "step": 72
    },
    {
      "epoch": 0.04292855042634519,
      "grad_norm": 3.306154251098633,
      "learning_rate": 7.52382768867422e-05,
      "loss": 8.8326,
      "step": 73
    },
    {
      "epoch": 0.04351661276095266,
      "grad_norm": 3.2203564643859863,
      "learning_rate": 7.452117519152542e-05,
      "loss": 8.9463,
      "step": 74
    },
    {
      "epoch": 0.04410467509556013,
      "grad_norm": 3.2799510955810547,
      "learning_rate": 7.379736965185368e-05,
      "loss": 9.1004,
      "step": 75
    },
    {
      "epoch": 0.0446927374301676,
      "grad_norm": 3.168398857116699,
      "learning_rate": 7.30670581489344e-05,
      "loss": 8.901,
      "step": 76
    },
    {
      "epoch": 0.045280799764775066,
      "grad_norm": 3.2739171981811523,
      "learning_rate": 7.233044034264034e-05,
      "loss": 8.7111,
      "step": 77
    },
    {
      "epoch": 0.04586886209938253,
      "grad_norm": 3.6047799587249756,
      "learning_rate": 7.158771761692464e-05,
      "loss": 8.7528,
      "step": 78
    },
    {
      "epoch": 0.04645692443399,
      "grad_norm": 3.6243784427642822,
      "learning_rate": 7.083909302476453e-05,
      "loss": 8.902,
      "step": 79
    },
    {
      "epoch": 0.047044986768597474,
      "grad_norm": 3.6083507537841797,
      "learning_rate": 7.008477123264848e-05,
      "loss": 8.9888,
      "step": 80
    },
    {
      "epoch": 0.04763304910320494,
      "grad_norm": 3.4579484462738037,
      "learning_rate": 6.932495846462261e-05,
      "loss": 8.6996,
      "step": 81
    },
    {
      "epoch": 0.04822111143781241,
      "grad_norm": 4.177857875823975,
      "learning_rate": 6.855986244591104e-05,
      "loss": 8.6282,
      "step": 82
    },
    {
      "epoch": 0.048809173772419875,
      "grad_norm": 4.470897197723389,
      "learning_rate": 6.778969234612584e-05,
      "loss": 8.007,
      "step": 83
    },
    {
      "epoch": 0.04939723610702734,
      "grad_norm": 4.531810760498047,
      "learning_rate": 6.701465872208216e-05,
      "loss": 9.3965,
      "step": 84
    },
    {
      "epoch": 0.049985298441634816,
      "grad_norm": 4.726022243499756,
      "learning_rate": 6.623497346023418e-05,
      "loss": 9.7157,
      "step": 85
    },
    {
      "epoch": 0.050573360776242283,
      "grad_norm": 5.001683712005615,
      "learning_rate": 6.545084971874738e-05,
      "loss": 9.161,
      "step": 86
    },
    {
      "epoch": 0.05116142311084975,
      "grad_norm": 4.754822254180908,
      "learning_rate": 6.466250186922325e-05,
      "loss": 8.5483,
      "step": 87
    },
    {
      "epoch": 0.05174948544545722,
      "grad_norm": 4.863503932952881,
      "learning_rate": 6.387014543809223e-05,
      "loss": 8.8508,
      "step": 88
    },
    {
      "epoch": 0.052337547780064685,
      "grad_norm": 5.733709812164307,
      "learning_rate": 6.307399704769099e-05,
      "loss": 9.8378,
      "step": 89
    },
    {
      "epoch": 0.05292561011467216,
      "grad_norm": 5.767556667327881,
      "learning_rate": 6.227427435703997e-05,
      "loss": 9.4093,
      "step": 90
    },
    {
      "epoch": 0.053513672449279626,
      "grad_norm": 6.011463642120361,
      "learning_rate": 6.147119600233758e-05,
      "loss": 8.6997,
      "step": 91
    },
    {
      "epoch": 0.05410173478388709,
      "grad_norm": 6.228724002838135,
      "learning_rate": 6.066498153718735e-05,
      "loss": 8.9712,
      "step": 92
    },
    {
      "epoch": 0.05468979711849456,
      "grad_norm": 7.02557897567749,
      "learning_rate": 5.985585137257401e-05,
      "loss": 9.2499,
      "step": 93
    },
    {
      "epoch": 0.05527785945310203,
      "grad_norm": 7.3237810134887695,
      "learning_rate": 5.90440267166055e-05,
      "loss": 8.7102,
      "step": 94
    },
    {
      "epoch": 0.055865921787709494,
      "grad_norm": 8.604790687561035,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 9.2058,
      "step": 95
    },
    {
      "epoch": 0.05645398412231697,
      "grad_norm": 8.26591968536377,
      "learning_rate": 5.74131823855921e-05,
      "loss": 8.7815,
      "step": 96
    },
    {
      "epoch": 0.057042046456924435,
      "grad_norm": 10.319204330444336,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 9.5898,
      "step": 97
    },
    {
      "epoch": 0.0576301087915319,
      "grad_norm": 11.870678901672363,
      "learning_rate": 5.577423184847932e-05,
      "loss": 9.1623,
      "step": 98
    },
    {
      "epoch": 0.05821817112613937,
      "grad_norm": 16.48353385925293,
      "learning_rate": 5.495227651252315e-05,
      "loss": 8.4069,
      "step": 99
    },
    {
      "epoch": 0.058806233460746836,
      "grad_norm": 31.682035446166992,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 8.6315,
      "step": 100
    },
    {
      "epoch": 0.058806233460746836,
      "eval_loss": 2.1445388793945312,
      "eval_runtime": 17.9901,
      "eval_samples_per_second": 159.199,
      "eval_steps_per_second": 39.8,
      "step": 100
    },
    {
      "epoch": 0.05939429579535431,
      "grad_norm": 6.193917274475098,
      "learning_rate": 5.330452921628497e-05,
      "loss": 7.9086,
      "step": 101
    },
    {
      "epoch": 0.05998235812996178,
      "grad_norm": 5.619760513305664,
      "learning_rate": 5.247918773366112e-05,
      "loss": 7.5205,
      "step": 102
    },
    {
      "epoch": 0.060570420464569244,
      "grad_norm": 5.22607421875,
      "learning_rate": 5.165316846586541e-05,
      "loss": 7.6454,
      "step": 103
    },
    {
      "epoch": 0.06115848279917671,
      "grad_norm": 4.238805294036865,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 8.6684,
      "step": 104
    },
    {
      "epoch": 0.06174654513378418,
      "grad_norm": 4.190961837768555,
      "learning_rate": 5e-05,
      "loss": 8.9169,
      "step": 105
    },
    {
      "epoch": 0.06233460746839165,
      "grad_norm": 3.7844080924987793,
      "learning_rate": 4.917330276168208e-05,
      "loss": 8.5661,
      "step": 106
    },
    {
      "epoch": 0.06292266980299911,
      "grad_norm": 3.943453073501587,
      "learning_rate": 4.834683153413459e-05,
      "loss": 8.2442,
      "step": 107
    },
    {
      "epoch": 0.0635107321376066,
      "grad_norm": 3.5105576515197754,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 8.2983,
      "step": 108
    },
    {
      "epoch": 0.06409879447221406,
      "grad_norm": 3.7738616466522217,
      "learning_rate": 4.669547078371504e-05,
      "loss": 8.2154,
      "step": 109
    },
    {
      "epoch": 0.06468685680682153,
      "grad_norm": 3.175321578979492,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 8.1635,
      "step": 110
    },
    {
      "epoch": 0.065274919141429,
      "grad_norm": 2.8794965744018555,
      "learning_rate": 4.504772348747687e-05,
      "loss": 8.3374,
      "step": 111
    },
    {
      "epoch": 0.06586298147603646,
      "grad_norm": 2.8476223945617676,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 8.7048,
      "step": 112
    },
    {
      "epoch": 0.06645104381064393,
      "grad_norm": 2.700763463973999,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 8.5285,
      "step": 113
    },
    {
      "epoch": 0.0670391061452514,
      "grad_norm": 2.5856125354766846,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 8.3342,
      "step": 114
    },
    {
      "epoch": 0.06762716847985886,
      "grad_norm": 2.5090999603271484,
      "learning_rate": 4.17702704859633e-05,
      "loss": 7.7179,
      "step": 115
    },
    {
      "epoch": 0.06821523081446633,
      "grad_norm": 2.675593852996826,
      "learning_rate": 4.095597328339452e-05,
      "loss": 8.5672,
      "step": 116
    },
    {
      "epoch": 0.0688032931490738,
      "grad_norm": 2.7386438846588135,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 9.2258,
      "step": 117
    },
    {
      "epoch": 0.06939135548368126,
      "grad_norm": 3.1391453742980957,
      "learning_rate": 3.933501846281267e-05,
      "loss": 8.2759,
      "step": 118
    },
    {
      "epoch": 0.06997941781828874,
      "grad_norm": 2.859008550643921,
      "learning_rate": 3.852880399766243e-05,
      "loss": 9.2972,
      "step": 119
    },
    {
      "epoch": 0.07056748015289621,
      "grad_norm": 2.8385045528411865,
      "learning_rate": 3.772572564296005e-05,
      "loss": 9.2444,
      "step": 120
    },
    {
      "epoch": 0.07115554248750368,
      "grad_norm": 2.785433053970337,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 8.2695,
      "step": 121
    },
    {
      "epoch": 0.07174360482211115,
      "grad_norm": 3.413860559463501,
      "learning_rate": 3.612985456190778e-05,
      "loss": 8.8478,
      "step": 122
    },
    {
      "epoch": 0.07233166715671861,
      "grad_norm": 3.047285556793213,
      "learning_rate": 3.533749813077677e-05,
      "loss": 9.2273,
      "step": 123
    },
    {
      "epoch": 0.07291972949132608,
      "grad_norm": 3.2367634773254395,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 8.5344,
      "step": 124
    },
    {
      "epoch": 0.07350779182593355,
      "grad_norm": 3.2533440589904785,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 9.3347,
      "step": 125
    },
    {
      "epoch": 0.07409585416054101,
      "grad_norm": 3.4966914653778076,
      "learning_rate": 3.298534127791785e-05,
      "loss": 8.6104,
      "step": 126
    },
    {
      "epoch": 0.07468391649514848,
      "grad_norm": 3.0202932357788086,
      "learning_rate": 3.221030765387417e-05,
      "loss": 8.7115,
      "step": 127
    },
    {
      "epoch": 0.07527197882975595,
      "grad_norm": 3.3886260986328125,
      "learning_rate": 3.144013755408895e-05,
      "loss": 8.6994,
      "step": 128
    },
    {
      "epoch": 0.07586004116436343,
      "grad_norm": 3.5800437927246094,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 9.3688,
      "step": 129
    },
    {
      "epoch": 0.0764481034989709,
      "grad_norm": 3.5207343101501465,
      "learning_rate": 2.991522876735154e-05,
      "loss": 8.6903,
      "step": 130
    },
    {
      "epoch": 0.07703616583357836,
      "grad_norm": 3.4582135677337646,
      "learning_rate": 2.916090697523549e-05,
      "loss": 8.4372,
      "step": 131
    },
    {
      "epoch": 0.07762422816818583,
      "grad_norm": 4.331637382507324,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 8.1935,
      "step": 132
    },
    {
      "epoch": 0.0782122905027933,
      "grad_norm": 4.02256965637207,
      "learning_rate": 2.766955965735968e-05,
      "loss": 8.5141,
      "step": 133
    },
    {
      "epoch": 0.07880035283740076,
      "grad_norm": 4.460637092590332,
      "learning_rate": 2.693294185106562e-05,
      "loss": 9.6442,
      "step": 134
    },
    {
      "epoch": 0.07938841517200823,
      "grad_norm": 4.595434188842773,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 8.7553,
      "step": 135
    },
    {
      "epoch": 0.0799764775066157,
      "grad_norm": 4.413257122039795,
      "learning_rate": 2.547882480847461e-05,
      "loss": 8.6695,
      "step": 136
    },
    {
      "epoch": 0.08056453984122317,
      "grad_norm": 4.718889236450195,
      "learning_rate": 2.476172311325783e-05,
      "loss": 9.3965,
      "step": 137
    },
    {
      "epoch": 0.08115260217583063,
      "grad_norm": 4.938948631286621,
      "learning_rate": 2.405152131093926e-05,
      "loss": 9.2762,
      "step": 138
    },
    {
      "epoch": 0.08174066451043811,
      "grad_norm": 5.82443904876709,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 9.3252,
      "step": 139
    },
    {
      "epoch": 0.08232872684504558,
      "grad_norm": 5.855938911437988,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 9.7101,
      "step": 140
    },
    {
      "epoch": 0.08291678917965305,
      "grad_norm": 5.540469169616699,
      "learning_rate": 2.196424713241637e-05,
      "loss": 8.6131,
      "step": 141
    },
    {
      "epoch": 0.08350485151426051,
      "grad_norm": 6.2768940925598145,
      "learning_rate": 2.128356686585282e-05,
      "loss": 8.8135,
      "step": 142
    },
    {
      "epoch": 0.08409291384886798,
      "grad_norm": 6.498032093048096,
      "learning_rate": 2.061073738537635e-05,
      "loss": 9.0499,
      "step": 143
    },
    {
      "epoch": 0.08468097618347545,
      "grad_norm": 7.061596393585205,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 8.7917,
      "step": 144
    },
    {
      "epoch": 0.08526903851808292,
      "grad_norm": 9.55113410949707,
      "learning_rate": 1.928936436551661e-05,
      "loss": 8.3972,
      "step": 145
    },
    {
      "epoch": 0.08585710085269038,
      "grad_norm": 8.871682167053223,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 8.9225,
      "step": 146
    },
    {
      "epoch": 0.08644516318729785,
      "grad_norm": 10.061549186706543,
      "learning_rate": 1.800157297483417e-05,
      "loss": 8.6611,
      "step": 147
    },
    {
      "epoch": 0.08703322552190532,
      "grad_norm": 13.46516227722168,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 9.0124,
      "step": 148
    },
    {
      "epoch": 0.0876212878565128,
      "grad_norm": 14.108145713806152,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 8.4522,
      "step": 149
    },
    {
      "epoch": 0.08820935019112026,
      "grad_norm": 29.471630096435547,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 9.0066,
      "step": 150
    },
    {
      "epoch": 0.08820935019112026,
      "eval_loss": 2.0842721462249756,
      "eval_runtime": 17.9332,
      "eval_samples_per_second": 159.704,
      "eval_steps_per_second": 39.926,
      "step": 150
    },
    {
      "epoch": 0.08879741252572773,
      "grad_norm": 7.597183704376221,
      "learning_rate": 1.553232954407171e-05,
      "loss": 7.2178,
      "step": 151
    },
    {
      "epoch": 0.0893854748603352,
      "grad_norm": 3.146782398223877,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 8.4071,
      "step": 152
    },
    {
      "epoch": 0.08997353719494267,
      "grad_norm": 3.1675562858581543,
      "learning_rate": 1.435357758543015e-05,
      "loss": 8.3765,
      "step": 153
    },
    {
      "epoch": 0.09056159952955013,
      "grad_norm": 2.7412240505218506,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 7.441,
      "step": 154
    },
    {
      "epoch": 0.0911496618641576,
      "grad_norm": 2.777986764907837,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 8.2939,
      "step": 155
    },
    {
      "epoch": 0.09173772419876507,
      "grad_norm": 2.605185031890869,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 8.0201,
      "step": 156
    },
    {
      "epoch": 0.09232578653337253,
      "grad_norm": 2.6774957180023193,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 8.461,
      "step": 157
    },
    {
      "epoch": 0.09291384886798,
      "grad_norm": 2.7459444999694824,
      "learning_rate": 1.157994445715706e-05,
      "loss": 8.2426,
      "step": 158
    },
    {
      "epoch": 0.09350191120258747,
      "grad_norm": 2.675412178039551,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 8.5952,
      "step": 159
    },
    {
      "epoch": 0.09408997353719495,
      "grad_norm": 2.484586238861084,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 8.9638,
      "step": 160
    },
    {
      "epoch": 0.09467803587180242,
      "grad_norm": 2.480360746383667,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 8.2939,
      "step": 161
    },
    {
      "epoch": 0.09526609820640988,
      "grad_norm": 2.4721195697784424,
      "learning_rate": 9.549150281252633e-06,
      "loss": 8.5063,
      "step": 162
    },
    {
      "epoch": 0.09585416054101735,
      "grad_norm": 2.6462907791137695,
      "learning_rate": 9.068759265665384e-06,
      "loss": 8.4315,
      "step": 163
    },
    {
      "epoch": 0.09644222287562482,
      "grad_norm": 2.6958720684051514,
      "learning_rate": 8.599558442598998e-06,
      "loss": 8.542,
      "step": 164
    },
    {
      "epoch": 0.09703028521023228,
      "grad_norm": 2.7175683975219727,
      "learning_rate": 8.141676086873572e-06,
      "loss": 8.6202,
      "step": 165
    },
    {
      "epoch": 0.09761834754483975,
      "grad_norm": 2.7758874893188477,
      "learning_rate": 7.695237378953223e-06,
      "loss": 9.1191,
      "step": 166
    },
    {
      "epoch": 0.09820640987944722,
      "grad_norm": 2.7156481742858887,
      "learning_rate": 7.260364370723044e-06,
      "loss": 8.6151,
      "step": 167
    },
    {
      "epoch": 0.09879447221405468,
      "grad_norm": 2.9561545848846436,
      "learning_rate": 6.837175952121306e-06,
      "loss": 8.0691,
      "step": 168
    },
    {
      "epoch": 0.09938253454866215,
      "grad_norm": 2.950094699859619,
      "learning_rate": 6.425787818636131e-06,
      "loss": 8.1961,
      "step": 169
    },
    {
      "epoch": 0.09997059688326963,
      "grad_norm": 2.8107621669769287,
      "learning_rate": 6.026312439675552e-06,
      "loss": 8.8672,
      "step": 170
    },
    {
      "epoch": 0.1005586592178771,
      "grad_norm": 2.8233821392059326,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 8.93,
      "step": 171
    },
    {
      "epoch": 0.10114672155248457,
      "grad_norm": 3.132800817489624,
      "learning_rate": 5.263533508961827e-06,
      "loss": 9.6831,
      "step": 172
    },
    {
      "epoch": 0.10173478388709203,
      "grad_norm": 2.8715932369232178,
      "learning_rate": 4.900438493352055e-06,
      "loss": 8.3045,
      "step": 173
    },
    {
      "epoch": 0.1023228462216995,
      "grad_norm": 2.900547742843628,
      "learning_rate": 4.549673247541875e-06,
      "loss": 8.748,
      "step": 174
    },
    {
      "epoch": 0.10291090855630697,
      "grad_norm": 3.2423956394195557,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 9.7763,
      "step": 175
    },
    {
      "epoch": 0.10349897089091444,
      "grad_norm": 2.9825217723846436,
      "learning_rate": 3.885512251130763e-06,
      "loss": 8.869,
      "step": 176
    },
    {
      "epoch": 0.1040870332255219,
      "grad_norm": 3.1538853645324707,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 8.8533,
      "step": 177
    },
    {
      "epoch": 0.10467509556012937,
      "grad_norm": 3.2650489807128906,
      "learning_rate": 3.271776770026963e-06,
      "loss": 8.6512,
      "step": 178
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 3.273656129837036,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 8.797,
      "step": 179
    },
    {
      "epoch": 0.10585122022934432,
      "grad_norm": 3.34405779838562,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 8.7568,
      "step": 180
    },
    {
      "epoch": 0.10643928256395178,
      "grad_norm": 3.5786080360412598,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 8.6351,
      "step": 181
    },
    {
      "epoch": 0.10702734489855925,
      "grad_norm": 3.4921109676361084,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 8.2808,
      "step": 182
    },
    {
      "epoch": 0.10761540723316672,
      "grad_norm": 3.555251121520996,
      "learning_rate": 1.962316193157593e-06,
      "loss": 8.1811,
      "step": 183
    },
    {
      "epoch": 0.10820346956777419,
      "grad_norm": 3.8983936309814453,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 9.3426,
      "step": 184
    },
    {
      "epoch": 0.10879153190238165,
      "grad_norm": 4.291001319885254,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 8.7215,
      "step": 185
    },
    {
      "epoch": 0.10937959423698912,
      "grad_norm": 4.339574337005615,
      "learning_rate": 1.333670137599713e-06,
      "loss": 8.3155,
      "step": 186
    },
    {
      "epoch": 0.10996765657159659,
      "grad_norm": 4.75957727432251,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 8.3999,
      "step": 187
    },
    {
      "epoch": 0.11055571890620405,
      "grad_norm": 5.064064979553223,
      "learning_rate": 9.810017062595322e-07,
      "loss": 8.8486,
      "step": 188
    },
    {
      "epoch": 0.11114378124081152,
      "grad_norm": 4.923590660095215,
      "learning_rate": 8.247462563808817e-07,
      "loss": 8.8091,
      "step": 189
    },
    {
      "epoch": 0.11173184357541899,
      "grad_norm": 5.2403106689453125,
      "learning_rate": 6.819348298638839e-07,
      "loss": 8.7856,
      "step": 190
    },
    {
      "epoch": 0.11231990591002647,
      "grad_norm": 5.671937465667725,
      "learning_rate": 5.526064699265753e-07,
      "loss": 8.801,
      "step": 191
    },
    {
      "epoch": 0.11290796824463394,
      "grad_norm": 6.104888439178467,
      "learning_rate": 4.367965336512403e-07,
      "loss": 8.8457,
      "step": 192
    },
    {
      "epoch": 0.1134960305792414,
      "grad_norm": 5.921628952026367,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 9.3598,
      "step": 193
    },
    {
      "epoch": 0.11408409291384887,
      "grad_norm": 7.0785417556762695,
      "learning_rate": 2.458548727494292e-07,
      "loss": 9.2931,
      "step": 194
    },
    {
      "epoch": 0.11467215524845634,
      "grad_norm": 7.207335948944092,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 8.948,
      "step": 195
    },
    {
      "epoch": 0.1152602175830638,
      "grad_norm": 8.11253833770752,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 9.2911,
      "step": 196
    },
    {
      "epoch": 0.11584827991767127,
      "grad_norm": 8.847803115844727,
      "learning_rate": 6.150154258476315e-08,
      "loss": 7.7041,
      "step": 197
    },
    {
      "epoch": 0.11643634225227874,
      "grad_norm": 13.241528511047363,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 9.5575,
      "step": 198
    },
    {
      "epoch": 0.1170244045868862,
      "grad_norm": 16.475034713745117,
      "learning_rate": 6.834750376549792e-09,
      "loss": 7.865,
      "step": 199
    },
    {
      "epoch": 0.11761246692149367,
      "grad_norm": 29.06346321105957,
      "learning_rate": 0.0,
      "loss": 7.5398,
      "step": 200
    },
    {
      "epoch": 0.11761246692149367,
      "eval_loss": 2.077235460281372,
      "eval_runtime": 18.2502,
      "eval_samples_per_second": 156.93,
      "eval_steps_per_second": 39.232,
      "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": 4062969537232896.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}