{
  "best_metric": 0.5493309497833252,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.7279344858962693,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.003639672429481347,
      "grad_norm": 1.627246379852295,
      "learning_rate": 1e-05,
      "loss": 4.1589,
      "step": 1
    },
    {
      "epoch": 0.003639672429481347,
      "eval_loss": 1.79654061794281,
      "eval_runtime": 46.9971,
      "eval_samples_per_second": 9.852,
      "eval_steps_per_second": 2.468,
      "step": 1
    },
    {
      "epoch": 0.007279344858962694,
      "grad_norm": 1.9117671251296997,
      "learning_rate": 2e-05,
      "loss": 5.4649,
      "step": 2
    },
    {
      "epoch": 0.01091901728844404,
      "grad_norm": 2.1369986534118652,
      "learning_rate": 3e-05,
      "loss": 6.0396,
      "step": 3
    },
    {
      "epoch": 0.014558689717925387,
      "grad_norm": 1.922481656074524,
      "learning_rate": 4e-05,
      "loss": 5.4704,
      "step": 4
    },
    {
      "epoch": 0.018198362147406732,
      "grad_norm": 2.276458501815796,
      "learning_rate": 5e-05,
      "loss": 7.0761,
      "step": 5
    },
    {
      "epoch": 0.02183803457688808,
      "grad_norm": 2.1037752628326416,
      "learning_rate": 6e-05,
      "loss": 5.9674,
      "step": 6
    },
    {
      "epoch": 0.025477707006369428,
      "grad_norm": 2.013711929321289,
      "learning_rate": 7e-05,
      "loss": 6.3611,
      "step": 7
    },
    {
      "epoch": 0.029117379435850774,
      "grad_norm": 2.357118606567383,
      "learning_rate": 8e-05,
      "loss": 7.0359,
      "step": 8
    },
    {
      "epoch": 0.03275705186533212,
      "grad_norm": 2.014467239379883,
      "learning_rate": 9e-05,
      "loss": 4.6014,
      "step": 9
    },
    {
      "epoch": 0.036396724294813464,
      "grad_norm": 2.6822941303253174,
      "learning_rate": 0.0001,
      "loss": 4.6165,
      "step": 10
    },
    {
      "epoch": 0.040036396724294813,
      "grad_norm": 2.9674253463745117,
      "learning_rate": 9.999316524962345e-05,
      "loss": 4.9409,
      "step": 11
    },
    {
      "epoch": 0.04367606915377616,
      "grad_norm": 3.10622501373291,
      "learning_rate": 9.997266286704631e-05,
      "loss": 5.2992,
      "step": 12
    },
    {
      "epoch": 0.047315741583257506,
      "grad_norm": 2.675551176071167,
      "learning_rate": 9.993849845741524e-05,
      "loss": 4.484,
      "step": 13
    },
    {
      "epoch": 0.050955414012738856,
      "grad_norm": 2.591759443283081,
      "learning_rate": 9.989068136093873e-05,
      "loss": 4.3192,
      "step": 14
    },
    {
      "epoch": 0.0545950864422202,
      "grad_norm": 2.9280741214752197,
      "learning_rate": 9.98292246503335e-05,
      "loss": 4.08,
      "step": 15
    },
    {
      "epoch": 0.05823475887170155,
      "grad_norm": 3.0319268703460693,
      "learning_rate": 9.975414512725057e-05,
      "loss": 4.6617,
      "step": 16
    },
    {
      "epoch": 0.06187443130118289,
      "grad_norm": 3.2446112632751465,
      "learning_rate": 9.966546331768191e-05,
      "loss": 4.1714,
      "step": 17
    },
    {
      "epoch": 0.06551410373066424,
      "grad_norm": 2.7704968452453613,
      "learning_rate": 9.956320346634876e-05,
      "loss": 3.8175,
      "step": 18
    },
    {
      "epoch": 0.06915377616014559,
      "grad_norm": 2.4823732376098633,
      "learning_rate": 9.944739353007344e-05,
      "loss": 3.0544,
      "step": 19
    },
    {
      "epoch": 0.07279344858962693,
      "grad_norm": 2.55143141746521,
      "learning_rate": 9.931806517013612e-05,
      "loss": 3.5428,
      "step": 20
    },
    {
      "epoch": 0.07643312101910828,
      "grad_norm": 3.101564645767212,
      "learning_rate": 9.917525374361912e-05,
      "loss": 3.285,
      "step": 21
    },
    {
      "epoch": 0.08007279344858963,
      "grad_norm": 3.0850398540496826,
      "learning_rate": 9.901899829374047e-05,
      "loss": 3.7318,
      "step": 22
    },
    {
      "epoch": 0.08371246587807098,
      "grad_norm": 2.3082125186920166,
      "learning_rate": 9.884934153917997e-05,
      "loss": 2.3845,
      "step": 23
    },
    {
      "epoch": 0.08735213830755233,
      "grad_norm": 3.455845355987549,
      "learning_rate": 9.86663298624003e-05,
      "loss": 3.6157,
      "step": 24
    },
    {
      "epoch": 0.09099181073703366,
      "grad_norm": 4.117661952972412,
      "learning_rate": 9.847001329696653e-05,
      "loss": 4.2201,
      "step": 25
    },
    {
      "epoch": 0.09463148316651501,
      "grad_norm": 3.8148367404937744,
      "learning_rate": 9.826044551386744e-05,
      "loss": 3.4827,
      "step": 26
    },
    {
      "epoch": 0.09827115559599636,
      "grad_norm": 2.942937135696411,
      "learning_rate": 9.803768380684242e-05,
      "loss": 2.6162,
      "step": 27
    },
    {
      "epoch": 0.10191082802547771,
      "grad_norm": 3.3898541927337646,
      "learning_rate": 9.780178907671789e-05,
      "loss": 3.1321,
      "step": 28
    },
    {
      "epoch": 0.10555050045495905,
      "grad_norm": 3.1848762035369873,
      "learning_rate": 9.755282581475769e-05,
      "loss": 2.636,
      "step": 29
    },
    {
      "epoch": 0.1091901728844404,
      "grad_norm": 4.092750549316406,
      "learning_rate": 9.729086208503174e-05,
      "loss": 3.2111,
      "step": 30
    },
    {
      "epoch": 0.11282984531392175,
      "grad_norm": 3.519073486328125,
      "learning_rate": 9.701596950580806e-05,
      "loss": 2.6637,
      "step": 31
    },
    {
      "epoch": 0.1164695177434031,
      "grad_norm": 3.4053587913513184,
      "learning_rate": 9.672822322997305e-05,
      "loss": 2.7886,
      "step": 32
    },
    {
      "epoch": 0.12010919017288443,
      "grad_norm": 3.211217164993286,
      "learning_rate": 9.642770192448536e-05,
      "loss": 2.3278,
      "step": 33
    },
    {
      "epoch": 0.12374886260236578,
      "grad_norm": 3.3157238960266113,
      "learning_rate": 9.611448774886924e-05,
      "loss": 2.406,
      "step": 34
    },
    {
      "epoch": 0.12738853503184713,
      "grad_norm": 3.215022563934326,
      "learning_rate": 9.578866633275288e-05,
      "loss": 2.1529,
      "step": 35
    },
    {
      "epoch": 0.13102820746132848,
      "grad_norm": 3.789199113845825,
      "learning_rate": 9.545032675245813e-05,
      "loss": 2.4993,
      "step": 36
    },
    {
      "epoch": 0.13466787989080983,
      "grad_norm": 3.265941619873047,
      "learning_rate": 9.509956150664796e-05,
      "loss": 2.3768,
      "step": 37
    },
    {
      "epoch": 0.13830755232029118,
      "grad_norm": 4.166137218475342,
      "learning_rate": 9.473646649103818e-05,
      "loss": 2.9483,
      "step": 38
    },
    {
      "epoch": 0.14194722474977253,
      "grad_norm": 3.466341018676758,
      "learning_rate": 9.43611409721806e-05,
      "loss": 2.518,
      "step": 39
    },
    {
      "epoch": 0.14558689717925385,
      "grad_norm": 3.070040702819824,
      "learning_rate": 9.397368756032445e-05,
      "loss": 1.857,
      "step": 40
    },
    {
      "epoch": 0.1492265696087352,
      "grad_norm": 3.779592752456665,
      "learning_rate": 9.357421218136386e-05,
      "loss": 3.0303,
      "step": 41
    },
    {
      "epoch": 0.15286624203821655,
      "grad_norm": 3.3679609298706055,
      "learning_rate": 9.316282404787871e-05,
      "loss": 2.3366,
      "step": 42
    },
    {
      "epoch": 0.1565059144676979,
      "grad_norm": 3.2041537761688232,
      "learning_rate": 9.273963562927695e-05,
      "loss": 2.0079,
      "step": 43
    },
    {
      "epoch": 0.16014558689717925,
      "grad_norm": 3.521491527557373,
      "learning_rate": 9.230476262104677e-05,
      "loss": 2.3795,
      "step": 44
    },
    {
      "epoch": 0.1637852593266606,
      "grad_norm": 3.4666640758514404,
      "learning_rate": 9.185832391312644e-05,
      "loss": 2.0071,
      "step": 45
    },
    {
      "epoch": 0.16742493175614195,
      "grad_norm": 3.5547351837158203,
      "learning_rate": 9.140044155740101e-05,
      "loss": 2.3359,
      "step": 46
    },
    {
      "epoch": 0.1710646041856233,
      "grad_norm": 4.037711143493652,
      "learning_rate": 9.093124073433463e-05,
      "loss": 2.4696,
      "step": 47
    },
    {
      "epoch": 0.17470427661510465,
      "grad_norm": 4.039661407470703,
      "learning_rate": 9.045084971874738e-05,
      "loss": 2.4595,
      "step": 48
    },
    {
      "epoch": 0.17834394904458598,
      "grad_norm": 4.1654744148254395,
      "learning_rate": 8.995939984474624e-05,
      "loss": 2.1335,
      "step": 49
    },
    {
      "epoch": 0.18198362147406733,
      "grad_norm": 5.677961349487305,
      "learning_rate": 8.945702546981969e-05,
      "loss": 2.9766,
      "step": 50
    },
    {
      "epoch": 0.18198362147406733,
      "eval_loss": 0.7961926460266113,
      "eval_runtime": 47.8431,
      "eval_samples_per_second": 9.677,
      "eval_steps_per_second": 2.425,
      "step": 50
    },
    {
      "epoch": 0.18562329390354868,
      "grad_norm": 6.720786094665527,
      "learning_rate": 8.894386393810563e-05,
      "loss": 3.6883,
      "step": 51
    },
    {
      "epoch": 0.18926296633303002,
      "grad_norm": 5.401752471923828,
      "learning_rate": 8.842005554284296e-05,
      "loss": 3.2575,
      "step": 52
    },
    {
      "epoch": 0.19290263876251137,
      "grad_norm": 3.6017062664031982,
      "learning_rate": 8.788574348801675e-05,
      "loss": 3.1164,
      "step": 53
    },
    {
      "epoch": 0.19654231119199272,
      "grad_norm": 3.036780834197998,
      "learning_rate": 8.73410738492077e-05,
      "loss": 3.214,
      "step": 54
    },
    {
      "epoch": 0.20018198362147407,
      "grad_norm": 2.9008591175079346,
      "learning_rate": 8.678619553365659e-05,
      "loss": 3.1117,
      "step": 55
    },
    {
      "epoch": 0.20382165605095542,
      "grad_norm": 2.769787549972534,
      "learning_rate": 8.622126023955446e-05,
      "loss": 3.4692,
      "step": 56
    },
    {
      "epoch": 0.20746132848043677,
      "grad_norm": 2.7811965942382812,
      "learning_rate": 8.564642241456986e-05,
      "loss": 2.9314,
      "step": 57
    },
    {
      "epoch": 0.2111010009099181,
      "grad_norm": 3.082224130630493,
      "learning_rate": 8.506183921362443e-05,
      "loss": 3.6847,
      "step": 58
    },
    {
      "epoch": 0.21474067333939945,
      "grad_norm": 2.673755645751953,
      "learning_rate": 8.44676704559283e-05,
      "loss": 3.1632,
      "step": 59
    },
    {
      "epoch": 0.2183803457688808,
      "grad_norm": 2.84306001663208,
      "learning_rate": 8.386407858128706e-05,
      "loss": 3.0326,
      "step": 60
    },
    {
      "epoch": 0.22202001819836215,
      "grad_norm": 2.7607879638671875,
      "learning_rate": 8.32512286056924e-05,
      "loss": 3.1445,
      "step": 61
    },
    {
      "epoch": 0.2256596906278435,
      "grad_norm": 3.034217357635498,
      "learning_rate": 8.262928807620843e-05,
      "loss": 3.0251,
      "step": 62
    },
    {
      "epoch": 0.22929936305732485,
      "grad_norm": 2.8413138389587402,
      "learning_rate": 8.199842702516583e-05,
      "loss": 3.0849,
      "step": 63
    },
    {
      "epoch": 0.2329390354868062,
      "grad_norm": 2.6723484992980957,
      "learning_rate": 8.135881792367686e-05,
      "loss": 2.9227,
      "step": 64
    },
    {
      "epoch": 0.23657870791628755,
      "grad_norm": 3.1689682006835938,
      "learning_rate": 8.07106356344834e-05,
      "loss": 3.0142,
      "step": 65
    },
    {
      "epoch": 0.24021838034576887,
      "grad_norm": 2.765841007232666,
      "learning_rate": 8.005405736415126e-05,
      "loss": 2.9293,
      "step": 66
    },
    {
      "epoch": 0.24385805277525022,
      "grad_norm": 3.059731960296631,
      "learning_rate": 7.938926261462366e-05,
      "loss": 3.1658,
      "step": 67
    },
    {
      "epoch": 0.24749772520473157,
      "grad_norm": 2.65315580368042,
      "learning_rate": 7.871643313414718e-05,
      "loss": 2.3629,
      "step": 68
    },
    {
      "epoch": 0.25113739763421294,
      "grad_norm": 2.504477024078369,
      "learning_rate": 7.803575286758364e-05,
      "loss": 2.4169,
      "step": 69
    },
    {
      "epoch": 0.25477707006369427,
      "grad_norm": 3.1260740756988525,
      "learning_rate": 7.734740790612136e-05,
      "loss": 3.2307,
      "step": 70
    },
    {
      "epoch": 0.2584167424931756,
      "grad_norm": 3.3141260147094727,
      "learning_rate": 7.66515864363997e-05,
      "loss": 3.1,
      "step": 71
    },
    {
      "epoch": 0.26205641492265697,
      "grad_norm": 2.1839711666107178,
      "learning_rate": 7.594847868906076e-05,
      "loss": 1.9521,
      "step": 72
    },
    {
      "epoch": 0.2656960873521383,
      "grad_norm": 2.9417312145233154,
      "learning_rate": 7.52382768867422e-05,
      "loss": 2.4712,
      "step": 73
    },
    {
      "epoch": 0.26933575978161967,
      "grad_norm": 2.796466827392578,
      "learning_rate": 7.452117519152542e-05,
      "loss": 2.5575,
      "step": 74
    },
    {
      "epoch": 0.272975432211101,
      "grad_norm": 3.0342817306518555,
      "learning_rate": 7.379736965185368e-05,
      "loss": 2.6345,
      "step": 75
    },
    {
      "epoch": 0.27661510464058237,
      "grad_norm": 2.476325035095215,
      "learning_rate": 7.30670581489344e-05,
      "loss": 2.0443,
      "step": 76
    },
    {
      "epoch": 0.2802547770700637,
      "grad_norm": 3.289766788482666,
      "learning_rate": 7.233044034264034e-05,
      "loss": 2.6388,
      "step": 77
    },
    {
      "epoch": 0.28389444949954507,
      "grad_norm": 3.0232796669006348,
      "learning_rate": 7.158771761692464e-05,
      "loss": 2.2449,
      "step": 78
    },
    {
      "epoch": 0.2875341219290264,
      "grad_norm": 3.855328321456909,
      "learning_rate": 7.083909302476453e-05,
      "loss": 2.8031,
      "step": 79
    },
    {
      "epoch": 0.2911737943585077,
      "grad_norm": 3.308868169784546,
      "learning_rate": 7.008477123264848e-05,
      "loss": 2.3914,
      "step": 80
    },
    {
      "epoch": 0.2948134667879891,
      "grad_norm": 3.645345687866211,
      "learning_rate": 6.932495846462261e-05,
      "loss": 2.6821,
      "step": 81
    },
    {
      "epoch": 0.2984531392174704,
      "grad_norm": 3.234999418258667,
      "learning_rate": 6.855986244591104e-05,
      "loss": 2.3851,
      "step": 82
    },
    {
      "epoch": 0.3020928116469518,
      "grad_norm": 3.897122383117676,
      "learning_rate": 6.778969234612584e-05,
      "loss": 2.7089,
      "step": 83
    },
    {
      "epoch": 0.3057324840764331,
      "grad_norm": 3.0344910621643066,
      "learning_rate": 6.701465872208216e-05,
      "loss": 2.0706,
      "step": 84
    },
    {
      "epoch": 0.3093721565059145,
      "grad_norm": 3.349313735961914,
      "learning_rate": 6.623497346023418e-05,
      "loss": 2.5606,
      "step": 85
    },
    {
      "epoch": 0.3130118289353958,
      "grad_norm": 3.3616232872009277,
      "learning_rate": 6.545084971874738e-05,
      "loss": 2.2757,
      "step": 86
    },
    {
      "epoch": 0.3166515013648772,
      "grad_norm": 3.2572226524353027,
      "learning_rate": 6.466250186922325e-05,
      "loss": 2.4387,
      "step": 87
    },
    {
      "epoch": 0.3202911737943585,
      "grad_norm": 2.5874180793762207,
      "learning_rate": 6.387014543809223e-05,
      "loss": 1.8548,
      "step": 88
    },
    {
      "epoch": 0.32393084622383983,
      "grad_norm": 3.1309866905212402,
      "learning_rate": 6.307399704769099e-05,
      "loss": 2.1818,
      "step": 89
    },
    {
      "epoch": 0.3275705186533212,
      "grad_norm": 2.5596563816070557,
      "learning_rate": 6.227427435703997e-05,
      "loss": 1.5533,
      "step": 90
    },
    {
      "epoch": 0.33121019108280253,
      "grad_norm": 3.5473990440368652,
      "learning_rate": 6.147119600233758e-05,
      "loss": 2.4156,
      "step": 91
    },
    {
      "epoch": 0.3348498635122839,
      "grad_norm": 2.545943021774292,
      "learning_rate": 6.066498153718735e-05,
      "loss": 1.6588,
      "step": 92
    },
    {
      "epoch": 0.33848953594176523,
      "grad_norm": 3.362652063369751,
      "learning_rate": 5.985585137257401e-05,
      "loss": 2.1111,
      "step": 93
    },
    {
      "epoch": 0.3421292083712466,
      "grad_norm": 2.7153704166412354,
      "learning_rate": 5.90440267166055e-05,
      "loss": 1.7483,
      "step": 94
    },
    {
      "epoch": 0.34576888080072793,
      "grad_norm": 3.3218350410461426,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 1.9325,
      "step": 95
    },
    {
      "epoch": 0.3494085532302093,
      "grad_norm": 2.5218467712402344,
      "learning_rate": 5.74131823855921e-05,
      "loss": 1.6064,
      "step": 96
    },
    {
      "epoch": 0.35304822565969063,
      "grad_norm": 3.5356359481811523,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 2.2088,
      "step": 97
    },
    {
      "epoch": 0.35668789808917195,
      "grad_norm": 3.6081345081329346,
      "learning_rate": 5.577423184847932e-05,
      "loss": 2.4458,
      "step": 98
    },
    {
      "epoch": 0.36032757051865333,
      "grad_norm": 4.0944085121154785,
      "learning_rate": 5.495227651252315e-05,
      "loss": 2.3938,
      "step": 99
    },
    {
      "epoch": 0.36396724294813465,
      "grad_norm": 3.5973987579345703,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 1.9108,
      "step": 100
    },
    {
      "epoch": 0.36396724294813465,
      "eval_loss": 0.64961838722229,
      "eval_runtime": 47.8372,
      "eval_samples_per_second": 9.679,
      "eval_steps_per_second": 2.425,
      "step": 100
    },
    {
      "epoch": 0.36760691537761603,
      "grad_norm": 3.3377132415771484,
      "learning_rate": 5.330452921628497e-05,
      "loss": 2.7465,
      "step": 101
    },
    {
      "epoch": 0.37124658780709735,
      "grad_norm": 3.747612953186035,
      "learning_rate": 5.247918773366112e-05,
      "loss": 3.3977,
      "step": 102
    },
    {
      "epoch": 0.37488626023657873,
      "grad_norm": 3.4456214904785156,
      "learning_rate": 5.165316846586541e-05,
      "loss": 3.0695,
      "step": 103
    },
    {
      "epoch": 0.37852593266606005,
      "grad_norm": 2.973848581314087,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 2.9707,
      "step": 104
    },
    {
      "epoch": 0.3821656050955414,
      "grad_norm": 2.9328627586364746,
      "learning_rate": 5e-05,
      "loss": 3.1883,
      "step": 105
    },
    {
      "epoch": 0.38580527752502275,
      "grad_norm": 3.184349536895752,
      "learning_rate": 4.917330276168208e-05,
      "loss": 3.2079,
      "step": 106
    },
    {
      "epoch": 0.38944494995450407,
      "grad_norm": 3.0407445430755615,
      "learning_rate": 4.834683153413459e-05,
      "loss": 3.5532,
      "step": 107
    },
    {
      "epoch": 0.39308462238398545,
      "grad_norm": 2.5558323860168457,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 3.0113,
      "step": 108
    },
    {
      "epoch": 0.39672429481346677,
      "grad_norm": 2.6029391288757324,
      "learning_rate": 4.669547078371504e-05,
      "loss": 2.8878,
      "step": 109
    },
    {
      "epoch": 0.40036396724294815,
      "grad_norm": 2.9034981727600098,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 2.9215,
      "step": 110
    },
    {
      "epoch": 0.40400363967242947,
      "grad_norm": 2.476382255554199,
      "learning_rate": 4.504772348747687e-05,
      "loss": 2.8013,
      "step": 111
    },
    {
      "epoch": 0.40764331210191085,
      "grad_norm": 2.6442599296569824,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 2.7091,
      "step": 112
    },
    {
      "epoch": 0.41128298453139217,
      "grad_norm": 2.788423776626587,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 2.9096,
      "step": 113
    },
    {
      "epoch": 0.41492265696087355,
      "grad_norm": 2.6069302558898926,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 2.8436,
      "step": 114
    },
    {
      "epoch": 0.41856232939035487,
      "grad_norm": 2.645390510559082,
      "learning_rate": 4.17702704859633e-05,
      "loss": 2.565,
      "step": 115
    },
    {
      "epoch": 0.4222020018198362,
      "grad_norm": 2.726881504058838,
      "learning_rate": 4.095597328339452e-05,
      "loss": 2.897,
      "step": 116
    },
    {
      "epoch": 0.42584167424931757,
      "grad_norm": 2.631944417953491,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 2.7293,
      "step": 117
    },
    {
      "epoch": 0.4294813466787989,
      "grad_norm": 2.9520139694213867,
      "learning_rate": 3.933501846281267e-05,
      "loss": 2.7533,
      "step": 118
    },
    {
      "epoch": 0.43312101910828027,
      "grad_norm": 2.4500296115875244,
      "learning_rate": 3.852880399766243e-05,
      "loss": 2.3622,
      "step": 119
    },
    {
      "epoch": 0.4367606915377616,
      "grad_norm": 2.336514711380005,
      "learning_rate": 3.772572564296005e-05,
      "loss": 2.1219,
      "step": 120
    },
    {
      "epoch": 0.44040036396724297,
      "grad_norm": 2.5286667346954346,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 2.5342,
      "step": 121
    },
    {
      "epoch": 0.4440400363967243,
      "grad_norm": 2.4934942722320557,
      "learning_rate": 3.612985456190778e-05,
      "loss": 1.9364,
      "step": 122
    },
    {
      "epoch": 0.44767970882620567,
      "grad_norm": 3.0001718997955322,
      "learning_rate": 3.533749813077677e-05,
      "loss": 2.7054,
      "step": 123
    },
    {
      "epoch": 0.451319381255687,
      "grad_norm": 3.175205945968628,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 2.6587,
      "step": 124
    },
    {
      "epoch": 0.4549590536851683,
      "grad_norm": 5.331387519836426,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 3.1747,
      "step": 125
    },
    {
      "epoch": 0.4585987261146497,
      "grad_norm": 2.996534824371338,
      "learning_rate": 3.298534127791785e-05,
      "loss": 2.3714,
      "step": 126
    },
    {
      "epoch": 0.462238398544131,
      "grad_norm": 2.782163619995117,
      "learning_rate": 3.221030765387417e-05,
      "loss": 2.3124,
      "step": 127
    },
    {
      "epoch": 0.4658780709736124,
      "grad_norm": 2.7269463539123535,
      "learning_rate": 3.144013755408895e-05,
      "loss": 2.0806,
      "step": 128
    },
    {
      "epoch": 0.4695177434030937,
      "grad_norm": 2.3673880100250244,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 1.7048,
      "step": 129
    },
    {
      "epoch": 0.4731574158325751,
      "grad_norm": 2.9457974433898926,
      "learning_rate": 2.991522876735154e-05,
      "loss": 1.9623,
      "step": 130
    },
    {
      "epoch": 0.4767970882620564,
      "grad_norm": 3.1078736782073975,
      "learning_rate": 2.916090697523549e-05,
      "loss": 2.1046,
      "step": 131
    },
    {
      "epoch": 0.48043676069153773,
      "grad_norm": 2.6460859775543213,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 1.7783,
      "step": 132
    },
    {
      "epoch": 0.4840764331210191,
      "grad_norm": 3.588235855102539,
      "learning_rate": 2.766955965735968e-05,
      "loss": 2.3552,
      "step": 133
    },
    {
      "epoch": 0.48771610555050043,
      "grad_norm": 2.959357738494873,
      "learning_rate": 2.693294185106562e-05,
      "loss": 2.0208,
      "step": 134
    },
    {
      "epoch": 0.4913557779799818,
      "grad_norm": 2.4479880332946777,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 1.8036,
      "step": 135
    },
    {
      "epoch": 0.49499545040946313,
      "grad_norm": 2.849869966506958,
      "learning_rate": 2.547882480847461e-05,
      "loss": 1.6477,
      "step": 136
    },
    {
      "epoch": 0.4986351228389445,
      "grad_norm": 2.9354405403137207,
      "learning_rate": 2.476172311325783e-05,
      "loss": 1.8989,
      "step": 137
    },
    {
      "epoch": 0.5022747952684259,
      "grad_norm": 2.8117895126342773,
      "learning_rate": 2.405152131093926e-05,
      "loss": 1.6576,
      "step": 138
    },
    {
      "epoch": 0.5059144676979072,
      "grad_norm": 3.178802013397217,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 2.0696,
      "step": 139
    },
    {
      "epoch": 0.5095541401273885,
      "grad_norm": 2.8615331649780273,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 1.934,
      "step": 140
    },
    {
      "epoch": 0.5131938125568699,
      "grad_norm": 3.041189432144165,
      "learning_rate": 2.196424713241637e-05,
      "loss": 1.7234,
      "step": 141
    },
    {
      "epoch": 0.5168334849863512,
      "grad_norm": 3.383711814880371,
      "learning_rate": 2.128356686585282e-05,
      "loss": 1.8509,
      "step": 142
    },
    {
      "epoch": 0.5204731574158326,
      "grad_norm": 2.709336042404175,
      "learning_rate": 2.061073738537635e-05,
      "loss": 1.7867,
      "step": 143
    },
    {
      "epoch": 0.5241128298453139,
      "grad_norm": 3.0257439613342285,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 1.6287,
      "step": 144
    },
    {
      "epoch": 0.5277525022747953,
      "grad_norm": 3.397047281265259,
      "learning_rate": 1.928936436551661e-05,
      "loss": 2.2298,
      "step": 145
    },
    {
      "epoch": 0.5313921747042766,
      "grad_norm": 3.4175384044647217,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 1.9722,
      "step": 146
    },
    {
      "epoch": 0.535031847133758,
      "grad_norm": 3.527061700820923,
      "learning_rate": 1.800157297483417e-05,
      "loss": 2.1503,
      "step": 147
    },
    {
      "epoch": 0.5386715195632393,
      "grad_norm": 3.472372531890869,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 2.0088,
      "step": 148
    },
    {
      "epoch": 0.5423111919927207,
      "grad_norm": 4.470870494842529,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 2.6901,
      "step": 149
    },
    {
      "epoch": 0.545950864422202,
      "grad_norm": 4.862389087677002,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 2.6104,
      "step": 150
    },
    {
      "epoch": 0.545950864422202,
      "eval_loss": 0.5726054906845093,
      "eval_runtime": 47.8363,
      "eval_samples_per_second": 9.679,
      "eval_steps_per_second": 2.425,
      "step": 150
    },
    {
      "epoch": 0.5495905368516834,
      "grad_norm": 2.3045661449432373,
      "learning_rate": 1.553232954407171e-05,
      "loss": 1.8687,
      "step": 151
    },
    {
      "epoch": 0.5532302092811647,
      "grad_norm": 2.3308658599853516,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 1.943,
      "step": 152
    },
    {
      "epoch": 0.556869881710646,
      "grad_norm": 3.019066572189331,
      "learning_rate": 1.435357758543015e-05,
      "loss": 2.7123,
      "step": 153
    },
    {
      "epoch": 0.5605095541401274,
      "grad_norm": 3.03554630279541,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 2.9098,
      "step": 154
    },
    {
      "epoch": 0.5641492265696088,
      "grad_norm": 2.8514564037323,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 2.6616,
      "step": 155
    },
    {
      "epoch": 0.5677888989990901,
      "grad_norm": 3.2490532398223877,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 3.0917,
      "step": 156
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 3.0110726356506348,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 2.5144,
      "step": 157
    },
    {
      "epoch": 0.5750682438580528,
      "grad_norm": 2.8920514583587646,
      "learning_rate": 1.157994445715706e-05,
      "loss": 2.3856,
      "step": 158
    },
    {
      "epoch": 0.5787079162875342,
      "grad_norm": 3.042830228805542,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 2.9325,
      "step": 159
    },
    {
      "epoch": 0.5823475887170154,
      "grad_norm": 2.9812355041503906,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 2.7771,
      "step": 160
    },
    {
      "epoch": 0.5859872611464968,
      "grad_norm": 3.1662824153900146,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 2.6821,
      "step": 161
    },
    {
      "epoch": 0.5896269335759782,
      "grad_norm": 3.039468288421631,
      "learning_rate": 9.549150281252633e-06,
      "loss": 2.8636,
      "step": 162
    },
    {
      "epoch": 0.5932666060054596,
      "grad_norm": 3.165109395980835,
      "learning_rate": 9.068759265665384e-06,
      "loss": 2.9832,
      "step": 163
    },
    {
      "epoch": 0.5969062784349408,
      "grad_norm": 2.8825981616973877,
      "learning_rate": 8.599558442598998e-06,
      "loss": 2.584,
      "step": 164
    },
    {
      "epoch": 0.6005459508644222,
      "grad_norm": 2.7540652751922607,
      "learning_rate": 8.141676086873572e-06,
      "loss": 2.5036,
      "step": 165
    },
    {
      "epoch": 0.6041856232939036,
      "grad_norm": 2.6940362453460693,
      "learning_rate": 7.695237378953223e-06,
      "loss": 2.1563,
      "step": 166
    },
    {
      "epoch": 0.607825295723385,
      "grad_norm": 2.829200506210327,
      "learning_rate": 7.260364370723044e-06,
      "loss": 2.2571,
      "step": 167
    },
    {
      "epoch": 0.6114649681528662,
      "grad_norm": 2.9042210578918457,
      "learning_rate": 6.837175952121306e-06,
      "loss": 2.2257,
      "step": 168
    },
    {
      "epoch": 0.6151046405823476,
      "grad_norm": 2.759990930557251,
      "learning_rate": 6.425787818636131e-06,
      "loss": 2.2483,
      "step": 169
    },
    {
      "epoch": 0.618744313011829,
      "grad_norm": 2.9242870807647705,
      "learning_rate": 6.026312439675552e-06,
      "loss": 2.3865,
      "step": 170
    },
    {
      "epoch": 0.6223839854413102,
      "grad_norm": 3.3415138721466064,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 2.8474,
      "step": 171
    },
    {
      "epoch": 0.6260236578707916,
      "grad_norm": 3.3262083530426025,
      "learning_rate": 5.263533508961827e-06,
      "loss": 2.4599,
      "step": 172
    },
    {
      "epoch": 0.629663330300273,
      "grad_norm": 3.0911431312561035,
      "learning_rate": 4.900438493352055e-06,
      "loss": 2.2419,
      "step": 173
    },
    {
      "epoch": 0.6333030027297544,
      "grad_norm": 3.0359034538269043,
      "learning_rate": 4.549673247541875e-06,
      "loss": 2.4275,
      "step": 174
    },
    {
      "epoch": 0.6369426751592356,
      "grad_norm": 2.591797351837158,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 1.9565,
      "step": 175
    },
    {
      "epoch": 0.640582347588717,
      "grad_norm": 2.9610791206359863,
      "learning_rate": 3.885512251130763e-06,
      "loss": 2.2195,
      "step": 176
    },
    {
      "epoch": 0.6442220200181984,
      "grad_norm": 3.2785117626190186,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 2.6549,
      "step": 177
    },
    {
      "epoch": 0.6478616924476797,
      "grad_norm": 3.319124937057495,
      "learning_rate": 3.271776770026963e-06,
      "loss": 2.5341,
      "step": 178
    },
    {
      "epoch": 0.651501364877161,
      "grad_norm": 2.5337727069854736,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 1.9203,
      "step": 179
    },
    {
      "epoch": 0.6551410373066424,
      "grad_norm": 2.702460289001465,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 2.0471,
      "step": 180
    },
    {
      "epoch": 0.6587807097361238,
      "grad_norm": 3.0306427478790283,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 2.2459,
      "step": 181
    },
    {
      "epoch": 0.6624203821656051,
      "grad_norm": 3.1413674354553223,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 2.0577,
      "step": 182
    },
    {
      "epoch": 0.6660600545950864,
      "grad_norm": 2.5585415363311768,
      "learning_rate": 1.962316193157593e-06,
      "loss": 1.6244,
      "step": 183
    },
    {
      "epoch": 0.6696997270245678,
      "grad_norm": 2.9799067974090576,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 2.115,
      "step": 184
    },
    {
      "epoch": 0.6733393994540491,
      "grad_norm": 2.8381919860839844,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 1.7643,
      "step": 185
    },
    {
      "epoch": 0.6769790718835305,
      "grad_norm": 2.8881571292877197,
      "learning_rate": 1.333670137599713e-06,
      "loss": 1.7219,
      "step": 186
    },
    {
      "epoch": 0.6806187443130118,
      "grad_norm": 3.4973676204681396,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 2.0544,
      "step": 187
    },
    {
      "epoch": 0.6842584167424932,
      "grad_norm": 2.6740148067474365,
      "learning_rate": 9.810017062595322e-07,
      "loss": 1.679,
      "step": 188
    },
    {
      "epoch": 0.6878980891719745,
      "grad_norm": 2.561082601547241,
      "learning_rate": 8.247462563808817e-07,
      "loss": 1.608,
      "step": 189
    },
    {
      "epoch": 0.6915377616014559,
      "grad_norm": 3.1520912647247314,
      "learning_rate": 6.819348298638839e-07,
      "loss": 2.1045,
      "step": 190
    },
    {
      "epoch": 0.6951774340309372,
      "grad_norm": 2.39721941947937,
      "learning_rate": 5.526064699265753e-07,
      "loss": 1.6011,
      "step": 191
    },
    {
      "epoch": 0.6988171064604186,
      "grad_norm": 3.330775022506714,
      "learning_rate": 4.367965336512403e-07,
      "loss": 2.0763,
      "step": 192
    },
    {
      "epoch": 0.7024567788898999,
      "grad_norm": 3.0983452796936035,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 1.9265,
      "step": 193
    },
    {
      "epoch": 0.7060964513193813,
      "grad_norm": 3.204805612564087,
      "learning_rate": 2.458548727494292e-07,
      "loss": 1.9357,
      "step": 194
    },
    {
      "epoch": 0.7097361237488626,
      "grad_norm": 2.6745471954345703,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 1.8047,
      "step": 195
    },
    {
      "epoch": 0.7133757961783439,
      "grad_norm": 3.268216609954834,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 2.1942,
      "step": 196
    },
    {
      "epoch": 0.7170154686078253,
      "grad_norm": 3.638026475906372,
      "learning_rate": 6.150154258476315e-08,
      "loss": 1.9774,
      "step": 197
    },
    {
      "epoch": 0.7206551410373067,
      "grad_norm": 3.1266729831695557,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 1.9979,
      "step": 198
    },
    {
      "epoch": 0.724294813466788,
      "grad_norm": 4.265519142150879,
      "learning_rate": 6.834750376549792e-09,
      "loss": 2.5902,
      "step": 199
    },
    {
      "epoch": 0.7279344858962693,
      "grad_norm": 4.284196376800537,
      "learning_rate": 0.0,
      "loss": 2.2774,
      "step": 200
    },
    {
      "epoch": 0.7279344858962693,
      "eval_loss": 0.5493309497833252,
      "eval_runtime": 47.8675,
      "eval_samples_per_second": 9.673,
      "eval_steps_per_second": 2.423,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.09187566075904e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}