{
  "best_metric": 1.62697434425354,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.06157635467980296,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0003078817733990148,
      "grad_norm": 12.794803619384766,
      "learning_rate": 1e-05,
      "loss": 2.6401,
      "step": 1
    },
    {
      "epoch": 0.0003078817733990148,
      "eval_loss": 3.317500114440918,
      "eval_runtime": 95.783,
      "eval_samples_per_second": 57.119,
      "eval_steps_per_second": 14.282,
      "step": 1
    },
    {
      "epoch": 0.0006157635467980296,
      "grad_norm": 11.815323829650879,
      "learning_rate": 2e-05,
      "loss": 3.4245,
      "step": 2
    },
    {
      "epoch": 0.0009236453201970443,
      "grad_norm": 6.161344051361084,
      "learning_rate": 3e-05,
      "loss": 2.546,
      "step": 3
    },
    {
      "epoch": 0.0012315270935960591,
      "grad_norm": 7.512563228607178,
      "learning_rate": 4e-05,
      "loss": 2.6418,
      "step": 4
    },
    {
      "epoch": 0.001539408866995074,
      "grad_norm": 4.183810710906982,
      "learning_rate": 5e-05,
      "loss": 2.505,
      "step": 5
    },
    {
      "epoch": 0.0018472906403940886,
      "grad_norm": 4.402069091796875,
      "learning_rate": 6e-05,
      "loss": 2.4704,
      "step": 6
    },
    {
      "epoch": 0.0021551724137931034,
      "grad_norm": 4.252677917480469,
      "learning_rate": 7e-05,
      "loss": 2.7264,
      "step": 7
    },
    {
      "epoch": 0.0024630541871921183,
      "grad_norm": 3.8405380249023438,
      "learning_rate": 8e-05,
      "loss": 2.4306,
      "step": 8
    },
    {
      "epoch": 0.002770935960591133,
      "grad_norm": 4.374967575073242,
      "learning_rate": 9e-05,
      "loss": 2.5975,
      "step": 9
    },
    {
      "epoch": 0.003078817733990148,
      "grad_norm": 4.238419055938721,
      "learning_rate": 0.0001,
      "loss": 2.2356,
      "step": 10
    },
    {
      "epoch": 0.0033866995073891628,
      "grad_norm": 4.055003643035889,
      "learning_rate": 9.999316524962345e-05,
      "loss": 1.9499,
      "step": 11
    },
    {
      "epoch": 0.003694581280788177,
      "grad_norm": 3.8915112018585205,
      "learning_rate": 9.997266286704631e-05,
      "loss": 2.2006,
      "step": 12
    },
    {
      "epoch": 0.0040024630541871924,
      "grad_norm": 3.4351985454559326,
      "learning_rate": 9.993849845741524e-05,
      "loss": 2.3652,
      "step": 13
    },
    {
      "epoch": 0.004310344827586207,
      "grad_norm": 3.451555013656616,
      "learning_rate": 9.989068136093873e-05,
      "loss": 1.9924,
      "step": 14
    },
    {
      "epoch": 0.004618226600985222,
      "grad_norm": 3.781479597091675,
      "learning_rate": 9.98292246503335e-05,
      "loss": 1.9158,
      "step": 15
    },
    {
      "epoch": 0.0049261083743842365,
      "grad_norm": 3.2316930294036865,
      "learning_rate": 9.975414512725057e-05,
      "loss": 1.8755,
      "step": 16
    },
    {
      "epoch": 0.005233990147783251,
      "grad_norm": 3.4765806198120117,
      "learning_rate": 9.966546331768191e-05,
      "loss": 2.0399,
      "step": 17
    },
    {
      "epoch": 0.005541871921182266,
      "grad_norm": 3.318739175796509,
      "learning_rate": 9.956320346634876e-05,
      "loss": 2.0487,
      "step": 18
    },
    {
      "epoch": 0.005849753694581281,
      "grad_norm": 3.346731185913086,
      "learning_rate": 9.944739353007344e-05,
      "loss": 1.9341,
      "step": 19
    },
    {
      "epoch": 0.006157635467980296,
      "grad_norm": 3.489190101623535,
      "learning_rate": 9.931806517013612e-05,
      "loss": 1.9217,
      "step": 20
    },
    {
      "epoch": 0.00646551724137931,
      "grad_norm": 3.6199898719787598,
      "learning_rate": 9.917525374361912e-05,
      "loss": 1.9245,
      "step": 21
    },
    {
      "epoch": 0.0067733990147783255,
      "grad_norm": 3.2084062099456787,
      "learning_rate": 9.901899829374047e-05,
      "loss": 1.716,
      "step": 22
    },
    {
      "epoch": 0.00708128078817734,
      "grad_norm": 3.8445253372192383,
      "learning_rate": 9.884934153917997e-05,
      "loss": 2.0867,
      "step": 23
    },
    {
      "epoch": 0.007389162561576354,
      "grad_norm": 3.398552179336548,
      "learning_rate": 9.86663298624003e-05,
      "loss": 2.0083,
      "step": 24
    },
    {
      "epoch": 0.00769704433497537,
      "grad_norm": 3.760892629623413,
      "learning_rate": 9.847001329696653e-05,
      "loss": 2.1079,
      "step": 25
    },
    {
      "epoch": 0.008004926108374385,
      "grad_norm": 3.516383171081543,
      "learning_rate": 9.826044551386744e-05,
      "loss": 2.1386,
      "step": 26
    },
    {
      "epoch": 0.0083128078817734,
      "grad_norm": 3.6902809143066406,
      "learning_rate": 9.803768380684242e-05,
      "loss": 1.7452,
      "step": 27
    },
    {
      "epoch": 0.008620689655172414,
      "grad_norm": 3.3290133476257324,
      "learning_rate": 9.780178907671789e-05,
      "loss": 1.6427,
      "step": 28
    },
    {
      "epoch": 0.008928571428571428,
      "grad_norm": 3.692824125289917,
      "learning_rate": 9.755282581475769e-05,
      "loss": 1.924,
      "step": 29
    },
    {
      "epoch": 0.009236453201970444,
      "grad_norm": 4.465096473693848,
      "learning_rate": 9.729086208503174e-05,
      "loss": 2.3632,
      "step": 30
    },
    {
      "epoch": 0.009544334975369459,
      "grad_norm": 3.439237594604492,
      "learning_rate": 9.701596950580806e-05,
      "loss": 1.9511,
      "step": 31
    },
    {
      "epoch": 0.009852216748768473,
      "grad_norm": 3.878333330154419,
      "learning_rate": 9.672822322997305e-05,
      "loss": 1.8923,
      "step": 32
    },
    {
      "epoch": 0.010160098522167487,
      "grad_norm": 3.585252046585083,
      "learning_rate": 9.642770192448536e-05,
      "loss": 1.6174,
      "step": 33
    },
    {
      "epoch": 0.010467980295566502,
      "grad_norm": 3.739001989364624,
      "learning_rate": 9.611448774886924e-05,
      "loss": 1.9136,
      "step": 34
    },
    {
      "epoch": 0.010775862068965518,
      "grad_norm": 4.062669277191162,
      "learning_rate": 9.578866633275288e-05,
      "loss": 1.937,
      "step": 35
    },
    {
      "epoch": 0.011083743842364532,
      "grad_norm": 4.218130588531494,
      "learning_rate": 9.545032675245813e-05,
      "loss": 2.016,
      "step": 36
    },
    {
      "epoch": 0.011391625615763547,
      "grad_norm": 3.63144850730896,
      "learning_rate": 9.509956150664796e-05,
      "loss": 1.8001,
      "step": 37
    },
    {
      "epoch": 0.011699507389162561,
      "grad_norm": 5.773649215698242,
      "learning_rate": 9.473646649103818e-05,
      "loss": 2.0108,
      "step": 38
    },
    {
      "epoch": 0.012007389162561576,
      "grad_norm": 3.9894485473632812,
      "learning_rate": 9.43611409721806e-05,
      "loss": 1.5574,
      "step": 39
    },
    {
      "epoch": 0.012315270935960592,
      "grad_norm": 4.269454002380371,
      "learning_rate": 9.397368756032445e-05,
      "loss": 1.9702,
      "step": 40
    },
    {
      "epoch": 0.012623152709359606,
      "grad_norm": 4.614938259124756,
      "learning_rate": 9.357421218136386e-05,
      "loss": 1.7268,
      "step": 41
    },
    {
      "epoch": 0.01293103448275862,
      "grad_norm": 4.719139575958252,
      "learning_rate": 9.316282404787871e-05,
      "loss": 2.0262,
      "step": 42
    },
    {
      "epoch": 0.013238916256157635,
      "grad_norm": 4.625585556030273,
      "learning_rate": 9.273963562927695e-05,
      "loss": 1.7113,
      "step": 43
    },
    {
      "epoch": 0.013546798029556651,
      "grad_norm": 5.045303821563721,
      "learning_rate": 9.230476262104677e-05,
      "loss": 1.7249,
      "step": 44
    },
    {
      "epoch": 0.013854679802955665,
      "grad_norm": 4.845086097717285,
      "learning_rate": 9.185832391312644e-05,
      "loss": 1.8113,
      "step": 45
    },
    {
      "epoch": 0.01416256157635468,
      "grad_norm": 5.909614086151123,
      "learning_rate": 9.140044155740101e-05,
      "loss": 1.8854,
      "step": 46
    },
    {
      "epoch": 0.014470443349753694,
      "grad_norm": 5.605759620666504,
      "learning_rate": 9.093124073433463e-05,
      "loss": 1.6528,
      "step": 47
    },
    {
      "epoch": 0.014778325123152709,
      "grad_norm": 5.992961406707764,
      "learning_rate": 9.045084971874738e-05,
      "loss": 1.8714,
      "step": 48
    },
    {
      "epoch": 0.015086206896551725,
      "grad_norm": 6.28204870223999,
      "learning_rate": 8.995939984474624e-05,
      "loss": 1.2789,
      "step": 49
    },
    {
      "epoch": 0.01539408866995074,
      "grad_norm": 8.594783782958984,
      "learning_rate": 8.945702546981969e-05,
      "loss": 1.7533,
      "step": 50
    },
    {
      "epoch": 0.01539408866995074,
      "eval_loss": 1.8635733127593994,
      "eval_runtime": 95.5765,
      "eval_samples_per_second": 57.242,
      "eval_steps_per_second": 14.313,
      "step": 50
    },
    {
      "epoch": 0.015701970443349755,
      "grad_norm": 2.7793431282043457,
      "learning_rate": 8.894386393810563e-05,
      "loss": 2.2847,
      "step": 51
    },
    {
      "epoch": 0.01600985221674877,
      "grad_norm": 3.062514066696167,
      "learning_rate": 8.842005554284296e-05,
      "loss": 2.1553,
      "step": 52
    },
    {
      "epoch": 0.016317733990147784,
      "grad_norm": 2.4607667922973633,
      "learning_rate": 8.788574348801675e-05,
      "loss": 2.153,
      "step": 53
    },
    {
      "epoch": 0.0166256157635468,
      "grad_norm": 2.185328245162964,
      "learning_rate": 8.73410738492077e-05,
      "loss": 2.019,
      "step": 54
    },
    {
      "epoch": 0.016933497536945813,
      "grad_norm": 2.387446641921997,
      "learning_rate": 8.678619553365659e-05,
      "loss": 1.9588,
      "step": 55
    },
    {
      "epoch": 0.017241379310344827,
      "grad_norm": 2.2797012329101562,
      "learning_rate": 8.622126023955446e-05,
      "loss": 1.961,
      "step": 56
    },
    {
      "epoch": 0.017549261083743842,
      "grad_norm": 2.413177013397217,
      "learning_rate": 8.564642241456986e-05,
      "loss": 2.1657,
      "step": 57
    },
    {
      "epoch": 0.017857142857142856,
      "grad_norm": 2.538679838180542,
      "learning_rate": 8.506183921362443e-05,
      "loss": 2.0174,
      "step": 58
    },
    {
      "epoch": 0.01816502463054187,
      "grad_norm": 2.7959859371185303,
      "learning_rate": 8.44676704559283e-05,
      "loss": 2.1523,
      "step": 59
    },
    {
      "epoch": 0.01847290640394089,
      "grad_norm": 2.6453638076782227,
      "learning_rate": 8.386407858128706e-05,
      "loss": 2.22,
      "step": 60
    },
    {
      "epoch": 0.018780788177339903,
      "grad_norm": 2.543562889099121,
      "learning_rate": 8.32512286056924e-05,
      "loss": 1.8093,
      "step": 61
    },
    {
      "epoch": 0.019088669950738917,
      "grad_norm": 2.6459786891937256,
      "learning_rate": 8.262928807620843e-05,
      "loss": 2.1079,
      "step": 62
    },
    {
      "epoch": 0.01939655172413793,
      "grad_norm": 2.726716995239258,
      "learning_rate": 8.199842702516583e-05,
      "loss": 1.8566,
      "step": 63
    },
    {
      "epoch": 0.019704433497536946,
      "grad_norm": 2.8123903274536133,
      "learning_rate": 8.135881792367686e-05,
      "loss": 1.789,
      "step": 64
    },
    {
      "epoch": 0.02001231527093596,
      "grad_norm": 2.645667791366577,
      "learning_rate": 8.07106356344834e-05,
      "loss": 1.5838,
      "step": 65
    },
    {
      "epoch": 0.020320197044334975,
      "grad_norm": 2.9253270626068115,
      "learning_rate": 8.005405736415126e-05,
      "loss": 1.9532,
      "step": 66
    },
    {
      "epoch": 0.02062807881773399,
      "grad_norm": 2.6083531379699707,
      "learning_rate": 7.938926261462366e-05,
      "loss": 1.7559,
      "step": 67
    },
    {
      "epoch": 0.020935960591133004,
      "grad_norm": 3.166278123855591,
      "learning_rate": 7.871643313414718e-05,
      "loss": 2.0045,
      "step": 68
    },
    {
      "epoch": 0.021243842364532018,
      "grad_norm": 2.8139894008636475,
      "learning_rate": 7.803575286758364e-05,
      "loss": 2.0382,
      "step": 69
    },
    {
      "epoch": 0.021551724137931036,
      "grad_norm": 2.585330009460449,
      "learning_rate": 7.734740790612136e-05,
      "loss": 1.5838,
      "step": 70
    },
    {
      "epoch": 0.02185960591133005,
      "grad_norm": 2.923590660095215,
      "learning_rate": 7.66515864363997e-05,
      "loss": 1.8805,
      "step": 71
    },
    {
      "epoch": 0.022167487684729065,
      "grad_norm": 2.7168614864349365,
      "learning_rate": 7.594847868906076e-05,
      "loss": 1.7883,
      "step": 72
    },
    {
      "epoch": 0.02247536945812808,
      "grad_norm": 2.853916883468628,
      "learning_rate": 7.52382768867422e-05,
      "loss": 1.8399,
      "step": 73
    },
    {
      "epoch": 0.022783251231527094,
      "grad_norm": 2.4425501823425293,
      "learning_rate": 7.452117519152542e-05,
      "loss": 1.384,
      "step": 74
    },
    {
      "epoch": 0.023091133004926108,
      "grad_norm": 2.9630587100982666,
      "learning_rate": 7.379736965185368e-05,
      "loss": 1.7465,
      "step": 75
    },
    {
      "epoch": 0.023399014778325122,
      "grad_norm": 3.0901756286621094,
      "learning_rate": 7.30670581489344e-05,
      "loss": 1.7366,
      "step": 76
    },
    {
      "epoch": 0.023706896551724137,
      "grad_norm": 2.891655921936035,
      "learning_rate": 7.233044034264034e-05,
      "loss": 1.5632,
      "step": 77
    },
    {
      "epoch": 0.02401477832512315,
      "grad_norm": 3.0951004028320312,
      "learning_rate": 7.158771761692464e-05,
      "loss": 1.5723,
      "step": 78
    },
    {
      "epoch": 0.02432266009852217,
      "grad_norm": 2.9680230617523193,
      "learning_rate": 7.083909302476453e-05,
      "loss": 1.6155,
      "step": 79
    },
    {
      "epoch": 0.024630541871921183,
      "grad_norm": 3.263544797897339,
      "learning_rate": 7.008477123264848e-05,
      "loss": 1.7472,
      "step": 80
    },
    {
      "epoch": 0.024938423645320198,
      "grad_norm": 3.6640257835388184,
      "learning_rate": 6.932495846462261e-05,
      "loss": 1.603,
      "step": 81
    },
    {
      "epoch": 0.025246305418719212,
      "grad_norm": 2.8581202030181885,
      "learning_rate": 6.855986244591104e-05,
      "loss": 1.2952,
      "step": 82
    },
    {
      "epoch": 0.025554187192118227,
      "grad_norm": 3.441542387008667,
      "learning_rate": 6.778969234612584e-05,
      "loss": 1.5234,
      "step": 83
    },
    {
      "epoch": 0.02586206896551724,
      "grad_norm": 4.049309253692627,
      "learning_rate": 6.701465872208216e-05,
      "loss": 1.9077,
      "step": 84
    },
    {
      "epoch": 0.026169950738916255,
      "grad_norm": 3.866891622543335,
      "learning_rate": 6.623497346023418e-05,
      "loss": 1.7995,
      "step": 85
    },
    {
      "epoch": 0.02647783251231527,
      "grad_norm": 3.6575818061828613,
      "learning_rate": 6.545084971874738e-05,
      "loss": 1.5758,
      "step": 86
    },
    {
      "epoch": 0.026785714285714284,
      "grad_norm": 3.4970083236694336,
      "learning_rate": 6.466250186922325e-05,
      "loss": 1.5781,
      "step": 87
    },
    {
      "epoch": 0.027093596059113302,
      "grad_norm": 3.486997365951538,
      "learning_rate": 6.387014543809223e-05,
      "loss": 1.6358,
      "step": 88
    },
    {
      "epoch": 0.027401477832512317,
      "grad_norm": 3.8788833618164062,
      "learning_rate": 6.307399704769099e-05,
      "loss": 1.4384,
      "step": 89
    },
    {
      "epoch": 0.02770935960591133,
      "grad_norm": 3.9699721336364746,
      "learning_rate": 6.227427435703997e-05,
      "loss": 1.7356,
      "step": 90
    },
    {
      "epoch": 0.028017241379310345,
      "grad_norm": 4.85883092880249,
      "learning_rate": 6.147119600233758e-05,
      "loss": 1.7876,
      "step": 91
    },
    {
      "epoch": 0.02832512315270936,
      "grad_norm": 4.405791759490967,
      "learning_rate": 6.066498153718735e-05,
      "loss": 1.7133,
      "step": 92
    },
    {
      "epoch": 0.028633004926108374,
      "grad_norm": 4.197214603424072,
      "learning_rate": 5.985585137257401e-05,
      "loss": 1.6991,
      "step": 93
    },
    {
      "epoch": 0.02894088669950739,
      "grad_norm": 4.273960590362549,
      "learning_rate": 5.90440267166055e-05,
      "loss": 1.5906,
      "step": 94
    },
    {
      "epoch": 0.029248768472906403,
      "grad_norm": 4.965144157409668,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 1.7312,
      "step": 95
    },
    {
      "epoch": 0.029556650246305417,
      "grad_norm": 3.984543800354004,
      "learning_rate": 5.74131823855921e-05,
      "loss": 1.5332,
      "step": 96
    },
    {
      "epoch": 0.029864532019704435,
      "grad_norm": 5.208272457122803,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 1.5483,
      "step": 97
    },
    {
      "epoch": 0.03017241379310345,
      "grad_norm": 4.832465648651123,
      "learning_rate": 5.577423184847932e-05,
      "loss": 1.4693,
      "step": 98
    },
    {
      "epoch": 0.030480295566502464,
      "grad_norm": 5.779028415679932,
      "learning_rate": 5.495227651252315e-05,
      "loss": 1.3533,
      "step": 99
    },
    {
      "epoch": 0.03078817733990148,
      "grad_norm": 7.1983642578125,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 1.624,
      "step": 100
    },
    {
      "epoch": 0.03078817733990148,
      "eval_loss": 1.7067663669586182,
      "eval_runtime": 96.12,
      "eval_samples_per_second": 56.918,
      "eval_steps_per_second": 14.232,
      "step": 100
    },
    {
      "epoch": 0.031096059113300493,
      "grad_norm": 2.1641173362731934,
      "learning_rate": 5.330452921628497e-05,
      "loss": 2.3088,
      "step": 101
    },
    {
      "epoch": 0.03140394088669951,
      "grad_norm": 2.135202169418335,
      "learning_rate": 5.247918773366112e-05,
      "loss": 2.3254,
      "step": 102
    },
    {
      "epoch": 0.031711822660098525,
      "grad_norm": 2.2573812007904053,
      "learning_rate": 5.165316846586541e-05,
      "loss": 2.3562,
      "step": 103
    },
    {
      "epoch": 0.03201970443349754,
      "grad_norm": 2.27254581451416,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 2.1438,
      "step": 104
    },
    {
      "epoch": 0.032327586206896554,
      "grad_norm": 2.1876254081726074,
      "learning_rate": 5e-05,
      "loss": 1.704,
      "step": 105
    },
    {
      "epoch": 0.03263546798029557,
      "grad_norm": 2.172593832015991,
      "learning_rate": 4.917330276168208e-05,
      "loss": 1.939,
      "step": 106
    },
    {
      "epoch": 0.03294334975369458,
      "grad_norm": 2.1987357139587402,
      "learning_rate": 4.834683153413459e-05,
      "loss": 1.8684,
      "step": 107
    },
    {
      "epoch": 0.0332512315270936,
      "grad_norm": 2.2260749340057373,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 1.7967,
      "step": 108
    },
    {
      "epoch": 0.03355911330049261,
      "grad_norm": 2.287294387817383,
      "learning_rate": 4.669547078371504e-05,
      "loss": 1.9563,
      "step": 109
    },
    {
      "epoch": 0.033866995073891626,
      "grad_norm": 2.3295738697052,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 1.7977,
      "step": 110
    },
    {
      "epoch": 0.03417487684729064,
      "grad_norm": 2.4467082023620605,
      "learning_rate": 4.504772348747687e-05,
      "loss": 1.9297,
      "step": 111
    },
    {
      "epoch": 0.034482758620689655,
      "grad_norm": 2.196367025375366,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 1.6557,
      "step": 112
    },
    {
      "epoch": 0.03479064039408867,
      "grad_norm": 2.440333843231201,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 1.8117,
      "step": 113
    },
    {
      "epoch": 0.035098522167487683,
      "grad_norm": 2.7747368812561035,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 1.9072,
      "step": 114
    },
    {
      "epoch": 0.0354064039408867,
      "grad_norm": 2.844381093978882,
      "learning_rate": 4.17702704859633e-05,
      "loss": 1.941,
      "step": 115
    },
    {
      "epoch": 0.03571428571428571,
      "grad_norm": 2.613924980163574,
      "learning_rate": 4.095597328339452e-05,
      "loss": 1.6075,
      "step": 116
    },
    {
      "epoch": 0.03602216748768473,
      "grad_norm": 2.588918924331665,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 1.9404,
      "step": 117
    },
    {
      "epoch": 0.03633004926108374,
      "grad_norm": 2.356069803237915,
      "learning_rate": 3.933501846281267e-05,
      "loss": 1.6951,
      "step": 118
    },
    {
      "epoch": 0.036637931034482756,
      "grad_norm": 2.6977860927581787,
      "learning_rate": 3.852880399766243e-05,
      "loss": 1.6961,
      "step": 119
    },
    {
      "epoch": 0.03694581280788178,
      "grad_norm": 2.694115161895752,
      "learning_rate": 3.772572564296005e-05,
      "loss": 1.9051,
      "step": 120
    },
    {
      "epoch": 0.03725369458128079,
      "grad_norm": 2.7471978664398193,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 1.8305,
      "step": 121
    },
    {
      "epoch": 0.037561576354679806,
      "grad_norm": 2.5824384689331055,
      "learning_rate": 3.612985456190778e-05,
      "loss": 1.7062,
      "step": 122
    },
    {
      "epoch": 0.03786945812807882,
      "grad_norm": 3.92519474029541,
      "learning_rate": 3.533749813077677e-05,
      "loss": 1.7287,
      "step": 123
    },
    {
      "epoch": 0.038177339901477834,
      "grad_norm": 3.198096752166748,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 1.8632,
      "step": 124
    },
    {
      "epoch": 0.03848522167487685,
      "grad_norm": 2.9180526733398438,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 1.5767,
      "step": 125
    },
    {
      "epoch": 0.03879310344827586,
      "grad_norm": 3.2348389625549316,
      "learning_rate": 3.298534127791785e-05,
      "loss": 1.8087,
      "step": 126
    },
    {
      "epoch": 0.03910098522167488,
      "grad_norm": 3.234524726867676,
      "learning_rate": 3.221030765387417e-05,
      "loss": 1.7623,
      "step": 127
    },
    {
      "epoch": 0.03940886699507389,
      "grad_norm": 3.044912576675415,
      "learning_rate": 3.144013755408895e-05,
      "loss": 1.6405,
      "step": 128
    },
    {
      "epoch": 0.039716748768472906,
      "grad_norm": 2.8053579330444336,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 1.6898,
      "step": 129
    },
    {
      "epoch": 0.04002463054187192,
      "grad_norm": 2.8541269302368164,
      "learning_rate": 2.991522876735154e-05,
      "loss": 1.5833,
      "step": 130
    },
    {
      "epoch": 0.040332512315270935,
      "grad_norm": 2.8868765830993652,
      "learning_rate": 2.916090697523549e-05,
      "loss": 1.4404,
      "step": 131
    },
    {
      "epoch": 0.04064039408866995,
      "grad_norm": 3.489316940307617,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 1.4335,
      "step": 132
    },
    {
      "epoch": 0.040948275862068964,
      "grad_norm": 3.285278797149658,
      "learning_rate": 2.766955965735968e-05,
      "loss": 1.74,
      "step": 133
    },
    {
      "epoch": 0.04125615763546798,
      "grad_norm": 3.0024755001068115,
      "learning_rate": 2.693294185106562e-05,
      "loss": 1.5374,
      "step": 134
    },
    {
      "epoch": 0.04156403940886699,
      "grad_norm": 3.4362244606018066,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 1.618,
      "step": 135
    },
    {
      "epoch": 0.04187192118226601,
      "grad_norm": 3.6874582767486572,
      "learning_rate": 2.547882480847461e-05,
      "loss": 1.5676,
      "step": 136
    },
    {
      "epoch": 0.04217980295566502,
      "grad_norm": 4.311068534851074,
      "learning_rate": 2.476172311325783e-05,
      "loss": 1.6649,
      "step": 137
    },
    {
      "epoch": 0.042487684729064036,
      "grad_norm": 3.8568317890167236,
      "learning_rate": 2.405152131093926e-05,
      "loss": 1.2931,
      "step": 138
    },
    {
      "epoch": 0.04279556650246306,
      "grad_norm": 3.746522903442383,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 1.6454,
      "step": 139
    },
    {
      "epoch": 0.04310344827586207,
      "grad_norm": 4.780509948730469,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 1.817,
      "step": 140
    },
    {
      "epoch": 0.043411330049261086,
      "grad_norm": 3.572814464569092,
      "learning_rate": 2.196424713241637e-05,
      "loss": 1.3904,
      "step": 141
    },
    {
      "epoch": 0.0437192118226601,
      "grad_norm": 4.122625350952148,
      "learning_rate": 2.128356686585282e-05,
      "loss": 1.4035,
      "step": 142
    },
    {
      "epoch": 0.044027093596059115,
      "grad_norm": 4.572724342346191,
      "learning_rate": 2.061073738537635e-05,
      "loss": 1.3113,
      "step": 143
    },
    {
      "epoch": 0.04433497536945813,
      "grad_norm": 4.1004767417907715,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 1.3177,
      "step": 144
    },
    {
      "epoch": 0.044642857142857144,
      "grad_norm": 3.823925256729126,
      "learning_rate": 1.928936436551661e-05,
      "loss": 1.299,
      "step": 145
    },
    {
      "epoch": 0.04495073891625616,
      "grad_norm": 5.1147589683532715,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 1.5112,
      "step": 146
    },
    {
      "epoch": 0.04525862068965517,
      "grad_norm": 4.43835973739624,
      "learning_rate": 1.800157297483417e-05,
      "loss": 1.4704,
      "step": 147
    },
    {
      "epoch": 0.04556650246305419,
      "grad_norm": 5.0359697341918945,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 1.6739,
      "step": 148
    },
    {
      "epoch": 0.0458743842364532,
      "grad_norm": 4.905889987945557,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 1.5088,
      "step": 149
    },
    {
      "epoch": 0.046182266009852216,
      "grad_norm": 6.466745376586914,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 1.5168,
      "step": 150
    },
    {
      "epoch": 0.046182266009852216,
      "eval_loss": 1.643827199935913,
      "eval_runtime": 96.1003,
      "eval_samples_per_second": 56.93,
      "eval_steps_per_second": 14.235,
      "step": 150
    },
    {
      "epoch": 0.04649014778325123,
      "grad_norm": 2.0972900390625,
      "learning_rate": 1.553232954407171e-05,
      "loss": 2.1312,
      "step": 151
    },
    {
      "epoch": 0.046798029556650245,
      "grad_norm": 1.8464032411575317,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 1.8275,
      "step": 152
    },
    {
      "epoch": 0.04710591133004926,
      "grad_norm": 1.9081634283065796,
      "learning_rate": 1.435357758543015e-05,
      "loss": 1.9528,
      "step": 153
    },
    {
      "epoch": 0.04741379310344827,
      "grad_norm": 2.2396631240844727,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 2.1596,
      "step": 154
    },
    {
      "epoch": 0.04772167487684729,
      "grad_norm": 2.2289609909057617,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 1.9847,
      "step": 155
    },
    {
      "epoch": 0.0480295566502463,
      "grad_norm": 2.20233416557312,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 1.6343,
      "step": 156
    },
    {
      "epoch": 0.048337438423645324,
      "grad_norm": 2.1257739067077637,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 1.6882,
      "step": 157
    },
    {
      "epoch": 0.04864532019704434,
      "grad_norm": 2.0463664531707764,
      "learning_rate": 1.157994445715706e-05,
      "loss": 1.7972,
      "step": 158
    },
    {
      "epoch": 0.04895320197044335,
      "grad_norm": 2.06660795211792,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 1.9009,
      "step": 159
    },
    {
      "epoch": 0.04926108374384237,
      "grad_norm": 2.482626438140869,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 1.9218,
      "step": 160
    },
    {
      "epoch": 0.04956896551724138,
      "grad_norm": 2.069200277328491,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 1.4809,
      "step": 161
    },
    {
      "epoch": 0.049876847290640396,
      "grad_norm": 2.2237679958343506,
      "learning_rate": 9.549150281252633e-06,
      "loss": 1.793,
      "step": 162
    },
    {
      "epoch": 0.05018472906403941,
      "grad_norm": 2.7343242168426514,
      "learning_rate": 9.068759265665384e-06,
      "loss": 1.8613,
      "step": 163
    },
    {
      "epoch": 0.050492610837438424,
      "grad_norm": 2.5009348392486572,
      "learning_rate": 8.599558442598998e-06,
      "loss": 1.9039,
      "step": 164
    },
    {
      "epoch": 0.05080049261083744,
      "grad_norm": 2.5361435413360596,
      "learning_rate": 8.141676086873572e-06,
      "loss": 1.6445,
      "step": 165
    },
    {
      "epoch": 0.05110837438423645,
      "grad_norm": 2.428419351577759,
      "learning_rate": 7.695237378953223e-06,
      "loss": 1.5302,
      "step": 166
    },
    {
      "epoch": 0.05141625615763547,
      "grad_norm": 2.659372329711914,
      "learning_rate": 7.260364370723044e-06,
      "loss": 1.7918,
      "step": 167
    },
    {
      "epoch": 0.05172413793103448,
      "grad_norm": 2.6872763633728027,
      "learning_rate": 6.837175952121306e-06,
      "loss": 1.7508,
      "step": 168
    },
    {
      "epoch": 0.052032019704433496,
      "grad_norm": 2.592416286468506,
      "learning_rate": 6.425787818636131e-06,
      "loss": 1.5218,
      "step": 169
    },
    {
      "epoch": 0.05233990147783251,
      "grad_norm": 2.468825578689575,
      "learning_rate": 6.026312439675552e-06,
      "loss": 1.6022,
      "step": 170
    },
    {
      "epoch": 0.052647783251231525,
      "grad_norm": 2.6317803859710693,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 1.7892,
      "step": 171
    },
    {
      "epoch": 0.05295566502463054,
      "grad_norm": 2.583479881286621,
      "learning_rate": 5.263533508961827e-06,
      "loss": 1.4561,
      "step": 172
    },
    {
      "epoch": 0.053263546798029554,
      "grad_norm": 2.650202512741089,
      "learning_rate": 4.900438493352055e-06,
      "loss": 1.7939,
      "step": 173
    },
    {
      "epoch": 0.05357142857142857,
      "grad_norm": 2.969696044921875,
      "learning_rate": 4.549673247541875e-06,
      "loss": 1.5292,
      "step": 174
    },
    {
      "epoch": 0.05387931034482758,
      "grad_norm": 2.629815101623535,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 1.5375,
      "step": 175
    },
    {
      "epoch": 0.054187192118226604,
      "grad_norm": 3.1946606636047363,
      "learning_rate": 3.885512251130763e-06,
      "loss": 1.8596,
      "step": 176
    },
    {
      "epoch": 0.05449507389162562,
      "grad_norm": 2.9054579734802246,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 1.5599,
      "step": 177
    },
    {
      "epoch": 0.05480295566502463,
      "grad_norm": 3.0099689960479736,
      "learning_rate": 3.271776770026963e-06,
      "loss": 1.6706,
      "step": 178
    },
    {
      "epoch": 0.05511083743842365,
      "grad_norm": 2.8682191371917725,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 1.4405,
      "step": 179
    },
    {
      "epoch": 0.05541871921182266,
      "grad_norm": 2.836137294769287,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 1.6299,
      "step": 180
    },
    {
      "epoch": 0.055726600985221676,
      "grad_norm": 2.727856397628784,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 1.4651,
      "step": 181
    },
    {
      "epoch": 0.05603448275862069,
      "grad_norm": 2.5633797645568848,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 1.3224,
      "step": 182
    },
    {
      "epoch": 0.056342364532019705,
      "grad_norm": 3.656292676925659,
      "learning_rate": 1.962316193157593e-06,
      "loss": 1.5134,
      "step": 183
    },
    {
      "epoch": 0.05665024630541872,
      "grad_norm": 3.6075234413146973,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 1.7958,
      "step": 184
    },
    {
      "epoch": 0.056958128078817734,
      "grad_norm": 3.5210952758789062,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 1.6738,
      "step": 185
    },
    {
      "epoch": 0.05726600985221675,
      "grad_norm": 3.8826026916503906,
      "learning_rate": 1.333670137599713e-06,
      "loss": 1.768,
      "step": 186
    },
    {
      "epoch": 0.05757389162561576,
      "grad_norm": 3.267144203186035,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 1.4532,
      "step": 187
    },
    {
      "epoch": 0.05788177339901478,
      "grad_norm": 3.3945417404174805,
      "learning_rate": 9.810017062595322e-07,
      "loss": 1.4127,
      "step": 188
    },
    {
      "epoch": 0.05818965517241379,
      "grad_norm": 3.432589530944824,
      "learning_rate": 8.247462563808817e-07,
      "loss": 1.4505,
      "step": 189
    },
    {
      "epoch": 0.058497536945812806,
      "grad_norm": 3.739508628845215,
      "learning_rate": 6.819348298638839e-07,
      "loss": 1.4679,
      "step": 190
    },
    {
      "epoch": 0.05880541871921182,
      "grad_norm": 4.449434757232666,
      "learning_rate": 5.526064699265753e-07,
      "loss": 1.6403,
      "step": 191
    },
    {
      "epoch": 0.059113300492610835,
      "grad_norm": 3.986313581466675,
      "learning_rate": 4.367965336512403e-07,
      "loss": 1.4599,
      "step": 192
    },
    {
      "epoch": 0.05942118226600985,
      "grad_norm": 3.671109199523926,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 1.3629,
      "step": 193
    },
    {
      "epoch": 0.05972906403940887,
      "grad_norm": 4.419003486633301,
      "learning_rate": 2.458548727494292e-07,
      "loss": 1.6854,
      "step": 194
    },
    {
      "epoch": 0.060036945812807885,
      "grad_norm": 5.223719120025635,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 1.6689,
      "step": 195
    },
    {
      "epoch": 0.0603448275862069,
      "grad_norm": 4.668959140777588,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 1.2665,
      "step": 196
    },
    {
      "epoch": 0.060652709359605914,
      "grad_norm": 4.5939483642578125,
      "learning_rate": 6.150154258476315e-08,
      "loss": 1.4826,
      "step": 197
    },
    {
      "epoch": 0.06096059113300493,
      "grad_norm": 4.635838985443115,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 1.0832,
      "step": 198
    },
    {
      "epoch": 0.06126847290640394,
      "grad_norm": 5.430537223815918,
      "learning_rate": 6.834750376549792e-09,
      "loss": 1.587,
      "step": 199
    },
    {
      "epoch": 0.06157635467980296,
      "grad_norm": 7.641375541687012,
      "learning_rate": 0.0,
      "loss": 1.7084,
      "step": 200
    },
    {
      "epoch": 0.06157635467980296,
      "eval_loss": 1.62697434425354,
      "eval_runtime": 95.4658,
      "eval_samples_per_second": 57.308,
      "eval_steps_per_second": 14.33,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.54569648635904e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}