{
  "best_metric": 1.9695662260055542,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.09754907938056334,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00048774539690281675,
      "grad_norm": 1.9052716493606567,
      "learning_rate": 1e-05,
      "loss": 4.2356,
      "step": 1
    },
    {
      "epoch": 0.00048774539690281675,
      "eval_loss": 5.431233882904053,
      "eval_runtime": 252.5492,
      "eval_samples_per_second": 13.673,
      "eval_steps_per_second": 3.421,
      "step": 1
    },
    {
      "epoch": 0.0009754907938056335,
      "grad_norm": 1.4416828155517578,
      "learning_rate": 2e-05,
      "loss": 4.2125,
      "step": 2
    },
    {
      "epoch": 0.0014632361907084503,
      "grad_norm": 1.152499794960022,
      "learning_rate": 3e-05,
      "loss": 4.1078,
      "step": 3
    },
    {
      "epoch": 0.001950981587611267,
      "grad_norm": 1.190706729888916,
      "learning_rate": 4e-05,
      "loss": 4.5125,
      "step": 4
    },
    {
      "epoch": 0.0024387269845140835,
      "grad_norm": 1.0438563823699951,
      "learning_rate": 5e-05,
      "loss": 4.0864,
      "step": 5
    },
    {
      "epoch": 0.0029264723814169005,
      "grad_norm": 1.0560365915298462,
      "learning_rate": 6e-05,
      "loss": 4.3029,
      "step": 6
    },
    {
      "epoch": 0.003414217778319717,
      "grad_norm": 1.0059778690338135,
      "learning_rate": 7e-05,
      "loss": 3.9783,
      "step": 7
    },
    {
      "epoch": 0.003901963175222534,
      "grad_norm": 1.0220872163772583,
      "learning_rate": 8e-05,
      "loss": 3.7404,
      "step": 8
    },
    {
      "epoch": 0.0043897085721253505,
      "grad_norm": 0.9616022706031799,
      "learning_rate": 9e-05,
      "loss": 3.6695,
      "step": 9
    },
    {
      "epoch": 0.004877453969028167,
      "grad_norm": 1.0607948303222656,
      "learning_rate": 0.0001,
      "loss": 3.2567,
      "step": 10
    },
    {
      "epoch": 0.005365199365930984,
      "grad_norm": 1.2449851036071777,
      "learning_rate": 9.999316524962345e-05,
      "loss": 3.7486,
      "step": 11
    },
    {
      "epoch": 0.005852944762833801,
      "grad_norm": 1.0133333206176758,
      "learning_rate": 9.997266286704631e-05,
      "loss": 2.7872,
      "step": 12
    },
    {
      "epoch": 0.0063406901597366176,
      "grad_norm": 1.1076979637145996,
      "learning_rate": 9.993849845741524e-05,
      "loss": 3.3464,
      "step": 13
    },
    {
      "epoch": 0.006828435556639434,
      "grad_norm": 1.5825839042663574,
      "learning_rate": 9.989068136093873e-05,
      "loss": 3.6083,
      "step": 14
    },
    {
      "epoch": 0.007316180953542251,
      "grad_norm": 1.579925298690796,
      "learning_rate": 9.98292246503335e-05,
      "loss": 3.557,
      "step": 15
    },
    {
      "epoch": 0.007803926350445068,
      "grad_norm": 1.6689749956130981,
      "learning_rate": 9.975414512725057e-05,
      "loss": 3.2903,
      "step": 16
    },
    {
      "epoch": 0.008291671747347884,
      "grad_norm": 1.827212929725647,
      "learning_rate": 9.966546331768191e-05,
      "loss": 3.2157,
      "step": 17
    },
    {
      "epoch": 0.008779417144250701,
      "grad_norm": 1.688193678855896,
      "learning_rate": 9.956320346634876e-05,
      "loss": 3.2331,
      "step": 18
    },
    {
      "epoch": 0.009267162541153518,
      "grad_norm": 1.6568336486816406,
      "learning_rate": 9.944739353007344e-05,
      "loss": 3.1171,
      "step": 19
    },
    {
      "epoch": 0.009754907938056334,
      "grad_norm": 1.7610609531402588,
      "learning_rate": 9.931806517013612e-05,
      "loss": 3.068,
      "step": 20
    },
    {
      "epoch": 0.010242653334959152,
      "grad_norm": 1.6493592262268066,
      "learning_rate": 9.917525374361912e-05,
      "loss": 3.02,
      "step": 21
    },
    {
      "epoch": 0.010730398731861967,
      "grad_norm": 1.8502589464187622,
      "learning_rate": 9.901899829374047e-05,
      "loss": 3.1663,
      "step": 22
    },
    {
      "epoch": 0.011218144128764785,
      "grad_norm": 1.6696405410766602,
      "learning_rate": 9.884934153917997e-05,
      "loss": 2.8192,
      "step": 23
    },
    {
      "epoch": 0.011705889525667602,
      "grad_norm": 1.7742797136306763,
      "learning_rate": 9.86663298624003e-05,
      "loss": 2.6178,
      "step": 24
    },
    {
      "epoch": 0.012193634922570418,
      "grad_norm": 1.8648438453674316,
      "learning_rate": 9.847001329696653e-05,
      "loss": 2.458,
      "step": 25
    },
    {
      "epoch": 0.012681380319473235,
      "grad_norm": 2.728337049484253,
      "learning_rate": 9.826044551386744e-05,
      "loss": 3.2729,
      "step": 26
    },
    {
      "epoch": 0.013169125716376053,
      "grad_norm": 2.4018361568450928,
      "learning_rate": 9.803768380684242e-05,
      "loss": 3.2133,
      "step": 27
    },
    {
      "epoch": 0.013656871113278868,
      "grad_norm": 2.372708559036255,
      "learning_rate": 9.780178907671789e-05,
      "loss": 3.0837,
      "step": 28
    },
    {
      "epoch": 0.014144616510181686,
      "grad_norm": 2.2014293670654297,
      "learning_rate": 9.755282581475769e-05,
      "loss": 2.9559,
      "step": 29
    },
    {
      "epoch": 0.014632361907084501,
      "grad_norm": 2.3799209594726562,
      "learning_rate": 9.729086208503174e-05,
      "loss": 3.0708,
      "step": 30
    },
    {
      "epoch": 0.015120107303987319,
      "grad_norm": 2.240283966064453,
      "learning_rate": 9.701596950580806e-05,
      "loss": 2.8685,
      "step": 31
    },
    {
      "epoch": 0.015607852700890136,
      "grad_norm": 2.5251808166503906,
      "learning_rate": 9.672822322997305e-05,
      "loss": 2.6582,
      "step": 32
    },
    {
      "epoch": 0.016095598097792953,
      "grad_norm": 3.1319077014923096,
      "learning_rate": 9.642770192448536e-05,
      "loss": 2.9253,
      "step": 33
    },
    {
      "epoch": 0.016583343494695767,
      "grad_norm": 2.4205679893493652,
      "learning_rate": 9.611448774886924e-05,
      "loss": 2.544,
      "step": 34
    },
    {
      "epoch": 0.017071088891598585,
      "grad_norm": 2.5769317150115967,
      "learning_rate": 9.578866633275288e-05,
      "loss": 2.9564,
      "step": 35
    },
    {
      "epoch": 0.017558834288501402,
      "grad_norm": 2.453679323196411,
      "learning_rate": 9.545032675245813e-05,
      "loss": 2.7366,
      "step": 36
    },
    {
      "epoch": 0.01804657968540422,
      "grad_norm": 2.8109004497528076,
      "learning_rate": 9.509956150664796e-05,
      "loss": 2.8361,
      "step": 37
    },
    {
      "epoch": 0.018534325082307037,
      "grad_norm": 2.4668376445770264,
      "learning_rate": 9.473646649103818e-05,
      "loss": 2.2735,
      "step": 38
    },
    {
      "epoch": 0.01902207047920985,
      "grad_norm": 2.8881049156188965,
      "learning_rate": 9.43611409721806e-05,
      "loss": 2.8365,
      "step": 39
    },
    {
      "epoch": 0.01950981587611267,
      "grad_norm": 3.0685033798217773,
      "learning_rate": 9.397368756032445e-05,
      "loss": 2.6331,
      "step": 40
    },
    {
      "epoch": 0.019997561273015486,
      "grad_norm": 3.2225987911224365,
      "learning_rate": 9.357421218136386e-05,
      "loss": 2.1921,
      "step": 41
    },
    {
      "epoch": 0.020485306669918303,
      "grad_norm": 2.9457056522369385,
      "learning_rate": 9.316282404787871e-05,
      "loss": 2.3838,
      "step": 42
    },
    {
      "epoch": 0.02097305206682112,
      "grad_norm": 3.490666389465332,
      "learning_rate": 9.273963562927695e-05,
      "loss": 2.5272,
      "step": 43
    },
    {
      "epoch": 0.021460797463723934,
      "grad_norm": 3.4128496646881104,
      "learning_rate": 9.230476262104677e-05,
      "loss": 2.1091,
      "step": 44
    },
    {
      "epoch": 0.021948542860626752,
      "grad_norm": 3.7460851669311523,
      "learning_rate": 9.185832391312644e-05,
      "loss": 2.3918,
      "step": 45
    },
    {
      "epoch": 0.02243628825752957,
      "grad_norm": 3.234468698501587,
      "learning_rate": 9.140044155740101e-05,
      "loss": 2.1729,
      "step": 46
    },
    {
      "epoch": 0.022924033654432387,
      "grad_norm": 3.8501033782958984,
      "learning_rate": 9.093124073433463e-05,
      "loss": 2.4277,
      "step": 47
    },
    {
      "epoch": 0.023411779051335204,
      "grad_norm": 4.398103713989258,
      "learning_rate": 9.045084971874738e-05,
      "loss": 2.7711,
      "step": 48
    },
    {
      "epoch": 0.02389952444823802,
      "grad_norm": 4.051817893981934,
      "learning_rate": 8.995939984474624e-05,
      "loss": 1.9342,
      "step": 49
    },
    {
      "epoch": 0.024387269845140835,
      "grad_norm": 7.170238018035889,
      "learning_rate": 8.945702546981969e-05,
      "loss": 2.8263,
      "step": 50
    },
    {
      "epoch": 0.024387269845140835,
      "eval_loss": 2.6282694339752197,
      "eval_runtime": 254.012,
      "eval_samples_per_second": 13.594,
      "eval_steps_per_second": 3.401,
      "step": 50
    },
    {
      "epoch": 0.024875015242043653,
      "grad_norm": 2.525137424468994,
      "learning_rate": 8.894386393810563e-05,
      "loss": 3.3201,
      "step": 51
    },
    {
      "epoch": 0.02536276063894647,
      "grad_norm": 1.9298559427261353,
      "learning_rate": 8.842005554284296e-05,
      "loss": 2.8963,
      "step": 52
    },
    {
      "epoch": 0.025850506035849288,
      "grad_norm": 1.562819480895996,
      "learning_rate": 8.788574348801675e-05,
      "loss": 2.8588,
      "step": 53
    },
    {
      "epoch": 0.026338251432752105,
      "grad_norm": 1.0883809328079224,
      "learning_rate": 8.73410738492077e-05,
      "loss": 2.6198,
      "step": 54
    },
    {
      "epoch": 0.02682599682965492,
      "grad_norm": 1.0330435037612915,
      "learning_rate": 8.678619553365659e-05,
      "loss": 2.5798,
      "step": 55
    },
    {
      "epoch": 0.027313742226557736,
      "grad_norm": 1.270867109298706,
      "learning_rate": 8.622126023955446e-05,
      "loss": 2.6996,
      "step": 56
    },
    {
      "epoch": 0.027801487623460554,
      "grad_norm": 1.1742101907730103,
      "learning_rate": 8.564642241456986e-05,
      "loss": 2.6916,
      "step": 57
    },
    {
      "epoch": 0.02828923302036337,
      "grad_norm": 1.0894042253494263,
      "learning_rate": 8.506183921362443e-05,
      "loss": 2.5709,
      "step": 58
    },
    {
      "epoch": 0.02877697841726619,
      "grad_norm": 1.244813084602356,
      "learning_rate": 8.44676704559283e-05,
      "loss": 2.6457,
      "step": 59
    },
    {
      "epoch": 0.029264723814169002,
      "grad_norm": 1.0746277570724487,
      "learning_rate": 8.386407858128706e-05,
      "loss": 2.2518,
      "step": 60
    },
    {
      "epoch": 0.02975246921107182,
      "grad_norm": 1.1548517942428589,
      "learning_rate": 8.32512286056924e-05,
      "loss": 2.3314,
      "step": 61
    },
    {
      "epoch": 0.030240214607974637,
      "grad_norm": 1.1649878025054932,
      "learning_rate": 8.262928807620843e-05,
      "loss": 2.4831,
      "step": 62
    },
    {
      "epoch": 0.030727960004877455,
      "grad_norm": 1.2175664901733398,
      "learning_rate": 8.199842702516583e-05,
      "loss": 2.0127,
      "step": 63
    },
    {
      "epoch": 0.031215705401780272,
      "grad_norm": 1.2224698066711426,
      "learning_rate": 8.135881792367686e-05,
      "loss": 2.3575,
      "step": 64
    },
    {
      "epoch": 0.031703450798683086,
      "grad_norm": 1.2590465545654297,
      "learning_rate": 8.07106356344834e-05,
      "loss": 2.2152,
      "step": 65
    },
    {
      "epoch": 0.03219119619558591,
      "grad_norm": 1.3008689880371094,
      "learning_rate": 8.005405736415126e-05,
      "loss": 2.2786,
      "step": 66
    },
    {
      "epoch": 0.03267894159248872,
      "grad_norm": 1.5196938514709473,
      "learning_rate": 7.938926261462366e-05,
      "loss": 2.2248,
      "step": 67
    },
    {
      "epoch": 0.033166686989391535,
      "grad_norm": 1.4997824430465698,
      "learning_rate": 7.871643313414718e-05,
      "loss": 2.3649,
      "step": 68
    },
    {
      "epoch": 0.033654432386294356,
      "grad_norm": 1.448543906211853,
      "learning_rate": 7.803575286758364e-05,
      "loss": 2.0617,
      "step": 69
    },
    {
      "epoch": 0.03414217778319717,
      "grad_norm": 1.5949437618255615,
      "learning_rate": 7.734740790612136e-05,
      "loss": 2.1539,
      "step": 70
    },
    {
      "epoch": 0.03462992318009999,
      "grad_norm": 1.6336168050765991,
      "learning_rate": 7.66515864363997e-05,
      "loss": 2.128,
      "step": 71
    },
    {
      "epoch": 0.035117668577002804,
      "grad_norm": 1.8401151895523071,
      "learning_rate": 7.594847868906076e-05,
      "loss": 2.2938,
      "step": 72
    },
    {
      "epoch": 0.03560541397390562,
      "grad_norm": 1.8767706155776978,
      "learning_rate": 7.52382768867422e-05,
      "loss": 2.2114,
      "step": 73
    },
    {
      "epoch": 0.03609315937080844,
      "grad_norm": 1.8363771438598633,
      "learning_rate": 7.452117519152542e-05,
      "loss": 2.0475,
      "step": 74
    },
    {
      "epoch": 0.03658090476771125,
      "grad_norm": 2.051151990890503,
      "learning_rate": 7.379736965185368e-05,
      "loss": 2.4243,
      "step": 75
    },
    {
      "epoch": 0.037068650164614074,
      "grad_norm": 2.10811185836792,
      "learning_rate": 7.30670581489344e-05,
      "loss": 2.0605,
      "step": 76
    },
    {
      "epoch": 0.03755639556151689,
      "grad_norm": 2.5270118713378906,
      "learning_rate": 7.233044034264034e-05,
      "loss": 2.4051,
      "step": 77
    },
    {
      "epoch": 0.0380441409584197,
      "grad_norm": 2.524273633956909,
      "learning_rate": 7.158771761692464e-05,
      "loss": 2.5405,
      "step": 78
    },
    {
      "epoch": 0.03853188635532252,
      "grad_norm": 2.4887049198150635,
      "learning_rate": 7.083909302476453e-05,
      "loss": 2.4866,
      "step": 79
    },
    {
      "epoch": 0.03901963175222534,
      "grad_norm": 2.3588764667510986,
      "learning_rate": 7.008477123264848e-05,
      "loss": 2.0816,
      "step": 80
    },
    {
      "epoch": 0.03950737714912816,
      "grad_norm": 2.476278781890869,
      "learning_rate": 6.932495846462261e-05,
      "loss": 2.1117,
      "step": 81
    },
    {
      "epoch": 0.03999512254603097,
      "grad_norm": 2.739741325378418,
      "learning_rate": 6.855986244591104e-05,
      "loss": 2.1998,
      "step": 82
    },
    {
      "epoch": 0.040482867942933785,
      "grad_norm": 2.7698280811309814,
      "learning_rate": 6.778969234612584e-05,
      "loss": 1.9936,
      "step": 83
    },
    {
      "epoch": 0.040970613339836606,
      "grad_norm": 2.5614287853240967,
      "learning_rate": 6.701465872208216e-05,
      "loss": 2.0132,
      "step": 84
    },
    {
      "epoch": 0.04145835873673942,
      "grad_norm": 3.0431385040283203,
      "learning_rate": 6.623497346023418e-05,
      "loss": 2.0992,
      "step": 85
    },
    {
      "epoch": 0.04194610413364224,
      "grad_norm": 3.1751694679260254,
      "learning_rate": 6.545084971874738e-05,
      "loss": 2.512,
      "step": 86
    },
    {
      "epoch": 0.042433849530545055,
      "grad_norm": 2.9213707447052,
      "learning_rate": 6.466250186922325e-05,
      "loss": 2.0407,
      "step": 87
    },
    {
      "epoch": 0.04292159492744787,
      "grad_norm": 2.7673933506011963,
      "learning_rate": 6.387014543809223e-05,
      "loss": 2.0313,
      "step": 88
    },
    {
      "epoch": 0.04340934032435069,
      "grad_norm": 3.18284010887146,
      "learning_rate": 6.307399704769099e-05,
      "loss": 2.1928,
      "step": 89
    },
    {
      "epoch": 0.043897085721253504,
      "grad_norm": 3.1404869556427,
      "learning_rate": 6.227427435703997e-05,
      "loss": 2.0266,
      "step": 90
    },
    {
      "epoch": 0.044384831118156325,
      "grad_norm": 3.900766372680664,
      "learning_rate": 6.147119600233758e-05,
      "loss": 2.2138,
      "step": 91
    },
    {
      "epoch": 0.04487257651505914,
      "grad_norm": 3.2854883670806885,
      "learning_rate": 6.066498153718735e-05,
      "loss": 2.0153,
      "step": 92
    },
    {
      "epoch": 0.04536032191196195,
      "grad_norm": 3.054872989654541,
      "learning_rate": 5.985585137257401e-05,
      "loss": 1.7195,
      "step": 93
    },
    {
      "epoch": 0.04584806730886477,
      "grad_norm": 3.2677552700042725,
      "learning_rate": 5.90440267166055e-05,
      "loss": 1.8897,
      "step": 94
    },
    {
      "epoch": 0.04633581270576759,
      "grad_norm": 3.4859201908111572,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 1.8254,
      "step": 95
    },
    {
      "epoch": 0.04682355810267041,
      "grad_norm": 3.594351053237915,
      "learning_rate": 5.74131823855921e-05,
      "loss": 1.9488,
      "step": 96
    },
    {
      "epoch": 0.04731130349957322,
      "grad_norm": 4.0036163330078125,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 2.01,
      "step": 97
    },
    {
      "epoch": 0.04779904889647604,
      "grad_norm": 4.132566452026367,
      "learning_rate": 5.577423184847932e-05,
      "loss": 2.0372,
      "step": 98
    },
    {
      "epoch": 0.04828679429337886,
      "grad_norm": 4.630672454833984,
      "learning_rate": 5.495227651252315e-05,
      "loss": 2.1507,
      "step": 99
    },
    {
      "epoch": 0.04877453969028167,
      "grad_norm": 6.598683834075928,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 2.1817,
      "step": 100
    },
    {
      "epoch": 0.04877453969028167,
      "eval_loss": 2.1385467052459717,
      "eval_runtime": 254.8774,
      "eval_samples_per_second": 13.548,
      "eval_steps_per_second": 3.39,
      "step": 100
    },
    {
      "epoch": 0.04926228508718449,
      "grad_norm": 1.0076099634170532,
      "learning_rate": 5.330452921628497e-05,
      "loss": 2.5588,
      "step": 101
    },
    {
      "epoch": 0.049750030484087306,
      "grad_norm": 1.001312017440796,
      "learning_rate": 5.247918773366112e-05,
      "loss": 2.3182,
      "step": 102
    },
    {
      "epoch": 0.050237775880990126,
      "grad_norm": 1.0313835144042969,
      "learning_rate": 5.165316846586541e-05,
      "loss": 2.2727,
      "step": 103
    },
    {
      "epoch": 0.05072552127789294,
      "grad_norm": 1.203776478767395,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 2.4243,
      "step": 104
    },
    {
      "epoch": 0.051213266674795754,
      "grad_norm": 1.0720765590667725,
      "learning_rate": 5e-05,
      "loss": 2.4688,
      "step": 105
    },
    {
      "epoch": 0.051701012071698575,
      "grad_norm": 0.9997125864028931,
      "learning_rate": 4.917330276168208e-05,
      "loss": 2.188,
      "step": 106
    },
    {
      "epoch": 0.05218875746860139,
      "grad_norm": 1.1665934324264526,
      "learning_rate": 4.834683153413459e-05,
      "loss": 2.25,
      "step": 107
    },
    {
      "epoch": 0.05267650286550421,
      "grad_norm": 1.3609803915023804,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 2.5379,
      "step": 108
    },
    {
      "epoch": 0.053164248262407024,
      "grad_norm": 1.0938054323196411,
      "learning_rate": 4.669547078371504e-05,
      "loss": 2.1572,
      "step": 109
    },
    {
      "epoch": 0.05365199365930984,
      "grad_norm": 1.1415295600891113,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 2.1127,
      "step": 110
    },
    {
      "epoch": 0.05413973905621266,
      "grad_norm": 1.366585373878479,
      "learning_rate": 4.504772348747687e-05,
      "loss": 2.3071,
      "step": 111
    },
    {
      "epoch": 0.05462748445311547,
      "grad_norm": 1.286671757698059,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 2.1031,
      "step": 112
    },
    {
      "epoch": 0.055115229850018294,
      "grad_norm": 1.2335736751556396,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 2.1239,
      "step": 113
    },
    {
      "epoch": 0.05560297524692111,
      "grad_norm": 1.2856693267822266,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 2.2264,
      "step": 114
    },
    {
      "epoch": 0.05609072064382392,
      "grad_norm": 1.4369779825210571,
      "learning_rate": 4.17702704859633e-05,
      "loss": 2.2642,
      "step": 115
    },
    {
      "epoch": 0.05657846604072674,
      "grad_norm": 1.4365304708480835,
      "learning_rate": 4.095597328339452e-05,
      "loss": 1.9535,
      "step": 116
    },
    {
      "epoch": 0.057066211437629556,
      "grad_norm": 1.7818381786346436,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 2.2405,
      "step": 117
    },
    {
      "epoch": 0.05755395683453238,
      "grad_norm": 1.5887640714645386,
      "learning_rate": 3.933501846281267e-05,
      "loss": 1.8797,
      "step": 118
    },
    {
      "epoch": 0.05804170223143519,
      "grad_norm": 1.5486586093902588,
      "learning_rate": 3.852880399766243e-05,
      "loss": 1.7751,
      "step": 119
    },
    {
      "epoch": 0.058529447628338005,
      "grad_norm": 1.8932163715362549,
      "learning_rate": 3.772572564296005e-05,
      "loss": 2.1617,
      "step": 120
    },
    {
      "epoch": 0.059017193025240826,
      "grad_norm": 1.8680821657180786,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 1.9447,
      "step": 121
    },
    {
      "epoch": 0.05950493842214364,
      "grad_norm": 1.9761908054351807,
      "learning_rate": 3.612985456190778e-05,
      "loss": 2.0518,
      "step": 122
    },
    {
      "epoch": 0.05999268381904646,
      "grad_norm": 1.890720248222351,
      "learning_rate": 3.533749813077677e-05,
      "loss": 1.8365,
      "step": 123
    },
    {
      "epoch": 0.060480429215949275,
      "grad_norm": 2.173712968826294,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 2.0536,
      "step": 124
    },
    {
      "epoch": 0.06096817461285209,
      "grad_norm": 2.3273727893829346,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 1.8561,
      "step": 125
    },
    {
      "epoch": 0.06145592000975491,
      "grad_norm": 2.0472538471221924,
      "learning_rate": 3.298534127791785e-05,
      "loss": 1.7752,
      "step": 126
    },
    {
      "epoch": 0.06194366540665772,
      "grad_norm": 2.443021297454834,
      "learning_rate": 3.221030765387417e-05,
      "loss": 1.8327,
      "step": 127
    },
    {
      "epoch": 0.062431410803560544,
      "grad_norm": 2.509351968765259,
      "learning_rate": 3.144013755408895e-05,
      "loss": 1.9403,
      "step": 128
    },
    {
      "epoch": 0.06291915620046336,
      "grad_norm": 2.4680798053741455,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 1.5282,
      "step": 129
    },
    {
      "epoch": 0.06340690159736617,
      "grad_norm": 2.428819417953491,
      "learning_rate": 2.991522876735154e-05,
      "loss": 1.6869,
      "step": 130
    },
    {
      "epoch": 0.06389464699426899,
      "grad_norm": 2.57232403755188,
      "learning_rate": 2.916090697523549e-05,
      "loss": 1.8523,
      "step": 131
    },
    {
      "epoch": 0.06438239239117181,
      "grad_norm": 2.801100730895996,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 1.8643,
      "step": 132
    },
    {
      "epoch": 0.06487013778807463,
      "grad_norm": 2.795808792114258,
      "learning_rate": 2.766955965735968e-05,
      "loss": 1.8388,
      "step": 133
    },
    {
      "epoch": 0.06535788318497744,
      "grad_norm": 2.9042930603027344,
      "learning_rate": 2.693294185106562e-05,
      "loss": 2.0097,
      "step": 134
    },
    {
      "epoch": 0.06584562858188026,
      "grad_norm": 2.932591438293457,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 1.7284,
      "step": 135
    },
    {
      "epoch": 0.06633337397878307,
      "grad_norm": 2.543088436126709,
      "learning_rate": 2.547882480847461e-05,
      "loss": 1.551,
      "step": 136
    },
    {
      "epoch": 0.0668211193756859,
      "grad_norm": 3.2886552810668945,
      "learning_rate": 2.476172311325783e-05,
      "loss": 2.0294,
      "step": 137
    },
    {
      "epoch": 0.06730886477258871,
      "grad_norm": 3.2782773971557617,
      "learning_rate": 2.405152131093926e-05,
      "loss": 1.7375,
      "step": 138
    },
    {
      "epoch": 0.06779661016949153,
      "grad_norm": 2.9647915363311768,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 1.8194,
      "step": 139
    },
    {
      "epoch": 0.06828435556639434,
      "grad_norm": 3.206841468811035,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 1.8198,
      "step": 140
    },
    {
      "epoch": 0.06877210096329715,
      "grad_norm": 3.204399824142456,
      "learning_rate": 2.196424713241637e-05,
      "loss": 1.625,
      "step": 141
    },
    {
      "epoch": 0.06925984636019998,
      "grad_norm": 3.584137201309204,
      "learning_rate": 2.128356686585282e-05,
      "loss": 2.0329,
      "step": 142
    },
    {
      "epoch": 0.0697475917571028,
      "grad_norm": 4.341381072998047,
      "learning_rate": 2.061073738537635e-05,
      "loss": 1.7938,
      "step": 143
    },
    {
      "epoch": 0.07023533715400561,
      "grad_norm": 3.4741930961608887,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 1.5957,
      "step": 144
    },
    {
      "epoch": 0.07072308255090842,
      "grad_norm": 3.2803332805633545,
      "learning_rate": 1.928936436551661e-05,
      "loss": 1.685,
      "step": 145
    },
    {
      "epoch": 0.07121082794781124,
      "grad_norm": 4.3190202713012695,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 1.9389,
      "step": 146
    },
    {
      "epoch": 0.07169857334471406,
      "grad_norm": 4.194921493530273,
      "learning_rate": 1.800157297483417e-05,
      "loss": 1.7211,
      "step": 147
    },
    {
      "epoch": 0.07218631874161688,
      "grad_norm": 5.248502731323242,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 1.9346,
      "step": 148
    },
    {
      "epoch": 0.07267406413851969,
      "grad_norm": 5.462099075317383,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 2.0417,
      "step": 149
    },
    {
      "epoch": 0.0731618095354225,
      "grad_norm": 5.787848472595215,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 1.6271,
      "step": 150
    },
    {
      "epoch": 0.0731618095354225,
      "eval_loss": 1.9856131076812744,
      "eval_runtime": 254.3168,
      "eval_samples_per_second": 13.578,
      "eval_steps_per_second": 3.397,
      "step": 150
    },
    {
      "epoch": 0.07364955493232532,
      "grad_norm": 0.8332124948501587,
      "learning_rate": 1.553232954407171e-05,
      "loss": 2.4971,
      "step": 151
    },
    {
      "epoch": 0.07413730032922815,
      "grad_norm": 0.9466613531112671,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 2.3997,
      "step": 152
    },
    {
      "epoch": 0.07462504572613096,
      "grad_norm": 0.883362889289856,
      "learning_rate": 1.435357758543015e-05,
      "loss": 2.3008,
      "step": 153
    },
    {
      "epoch": 0.07511279112303378,
      "grad_norm": 0.822169840335846,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 2.1958,
      "step": 154
    },
    {
      "epoch": 0.07560053651993659,
      "grad_norm": 0.9493765234947205,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 2.339,
      "step": 155
    },
    {
      "epoch": 0.0760882819168394,
      "grad_norm": 1.181384563446045,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 2.2073,
      "step": 156
    },
    {
      "epoch": 0.07657602731374223,
      "grad_norm": 0.9916157722473145,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 2.1347,
      "step": 157
    },
    {
      "epoch": 0.07706377271064505,
      "grad_norm": 1.1271525621414185,
      "learning_rate": 1.157994445715706e-05,
      "loss": 2.1515,
      "step": 158
    },
    {
      "epoch": 0.07755151810754786,
      "grad_norm": 1.0684483051300049,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 2.0984,
      "step": 159
    },
    {
      "epoch": 0.07803926350445067,
      "grad_norm": 1.0676051378250122,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 2.0952,
      "step": 160
    },
    {
      "epoch": 0.07852700890135349,
      "grad_norm": 1.1392630338668823,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 1.9164,
      "step": 161
    },
    {
      "epoch": 0.07901475429825632,
      "grad_norm": 1.260503888130188,
      "learning_rate": 9.549150281252633e-06,
      "loss": 2.1651,
      "step": 162
    },
    {
      "epoch": 0.07950249969515913,
      "grad_norm": 1.2723053693771362,
      "learning_rate": 9.068759265665384e-06,
      "loss": 2.1672,
      "step": 163
    },
    {
      "epoch": 0.07999024509206194,
      "grad_norm": 1.4457951784133911,
      "learning_rate": 8.599558442598998e-06,
      "loss": 2.1827,
      "step": 164
    },
    {
      "epoch": 0.08047799048896476,
      "grad_norm": 1.4315434694290161,
      "learning_rate": 8.141676086873572e-06,
      "loss": 1.9107,
      "step": 165
    },
    {
      "epoch": 0.08096573588586757,
      "grad_norm": 1.6096829175949097,
      "learning_rate": 7.695237378953223e-06,
      "loss": 2.181,
      "step": 166
    },
    {
      "epoch": 0.0814534812827704,
      "grad_norm": 1.6041358709335327,
      "learning_rate": 7.260364370723044e-06,
      "loss": 2.0235,
      "step": 167
    },
    {
      "epoch": 0.08194122667967321,
      "grad_norm": 1.8807945251464844,
      "learning_rate": 6.837175952121306e-06,
      "loss": 2.1797,
      "step": 168
    },
    {
      "epoch": 0.08242897207657603,
      "grad_norm": 1.759990930557251,
      "learning_rate": 6.425787818636131e-06,
      "loss": 1.9883,
      "step": 169
    },
    {
      "epoch": 0.08291671747347884,
      "grad_norm": 1.8521469831466675,
      "learning_rate": 6.026312439675552e-06,
      "loss": 2.0289,
      "step": 170
    },
    {
      "epoch": 0.08340446287038165,
      "grad_norm": 1.988706111907959,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 1.9114,
      "step": 171
    },
    {
      "epoch": 0.08389220826728448,
      "grad_norm": 1.9599580764770508,
      "learning_rate": 5.263533508961827e-06,
      "loss": 1.9462,
      "step": 172
    },
    {
      "epoch": 0.0843799536641873,
      "grad_norm": 2.191668748855591,
      "learning_rate": 4.900438493352055e-06,
      "loss": 2.0363,
      "step": 173
    },
    {
      "epoch": 0.08486769906109011,
      "grad_norm": 2.110227584838867,
      "learning_rate": 4.549673247541875e-06,
      "loss": 1.915,
      "step": 174
    },
    {
      "epoch": 0.08535544445799292,
      "grad_norm": 2.1333167552948,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 1.7748,
      "step": 175
    },
    {
      "epoch": 0.08584318985489574,
      "grad_norm": 2.453028678894043,
      "learning_rate": 3.885512251130763e-06,
      "loss": 1.8756,
      "step": 176
    },
    {
      "epoch": 0.08633093525179857,
      "grad_norm": 2.544600009918213,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 2.0746,
      "step": 177
    },
    {
      "epoch": 0.08681868064870138,
      "grad_norm": 2.625966787338257,
      "learning_rate": 3.271776770026963e-06,
      "loss": 2.0678,
      "step": 178
    },
    {
      "epoch": 0.0873064260456042,
      "grad_norm": 2.8309292793273926,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 1.9976,
      "step": 179
    },
    {
      "epoch": 0.08779417144250701,
      "grad_norm": 2.53399395942688,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 1.8744,
      "step": 180
    },
    {
      "epoch": 0.08828191683940982,
      "grad_norm": 2.6540863513946533,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 2.0297,
      "step": 181
    },
    {
      "epoch": 0.08876966223631265,
      "grad_norm": 2.830566644668579,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 1.9397,
      "step": 182
    },
    {
      "epoch": 0.08925740763321546,
      "grad_norm": 2.9761435985565186,
      "learning_rate": 1.962316193157593e-06,
      "loss": 1.9049,
      "step": 183
    },
    {
      "epoch": 0.08974515303011828,
      "grad_norm": 3.0352025032043457,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 1.8094,
      "step": 184
    },
    {
      "epoch": 0.09023289842702109,
      "grad_norm": 3.1714072227478027,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 1.7415,
      "step": 185
    },
    {
      "epoch": 0.0907206438239239,
      "grad_norm": 3.106734275817871,
      "learning_rate": 1.333670137599713e-06,
      "loss": 1.67,
      "step": 186
    },
    {
      "epoch": 0.09120838922082673,
      "grad_norm": 2.9521868228912354,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 1.8375,
      "step": 187
    },
    {
      "epoch": 0.09169613461772955,
      "grad_norm": 3.240025281906128,
      "learning_rate": 9.810017062595322e-07,
      "loss": 1.7915,
      "step": 188
    },
    {
      "epoch": 0.09218388001463236,
      "grad_norm": 3.5706686973571777,
      "learning_rate": 8.247462563808817e-07,
      "loss": 2.2209,
      "step": 189
    },
    {
      "epoch": 0.09267162541153517,
      "grad_norm": 4.059500694274902,
      "learning_rate": 6.819348298638839e-07,
      "loss": 2.1104,
      "step": 190
    },
    {
      "epoch": 0.09315937080843799,
      "grad_norm": 3.146223545074463,
      "learning_rate": 5.526064699265753e-07,
      "loss": 1.5803,
      "step": 191
    },
    {
      "epoch": 0.09364711620534082,
      "grad_norm": 3.539418935775757,
      "learning_rate": 4.367965336512403e-07,
      "loss": 1.5951,
      "step": 192
    },
    {
      "epoch": 0.09413486160224363,
      "grad_norm": 3.9369935989379883,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 1.6507,
      "step": 193
    },
    {
      "epoch": 0.09462260699914644,
      "grad_norm": 4.130059719085693,
      "learning_rate": 2.458548727494292e-07,
      "loss": 1.7946,
      "step": 194
    },
    {
      "epoch": 0.09511035239604926,
      "grad_norm": 4.931833267211914,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 1.9481,
      "step": 195
    },
    {
      "epoch": 0.09559809779295209,
      "grad_norm": 4.79127311706543,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 1.9826,
      "step": 196
    },
    {
      "epoch": 0.0960858431898549,
      "grad_norm": 4.758818626403809,
      "learning_rate": 6.150154258476315e-08,
      "loss": 2.0017,
      "step": 197
    },
    {
      "epoch": 0.09657358858675771,
      "grad_norm": 4.859172821044922,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 1.8742,
      "step": 198
    },
    {
      "epoch": 0.09706133398366053,
      "grad_norm": 7.261124134063721,
      "learning_rate": 6.834750376549792e-09,
      "loss": 2.229,
      "step": 199
    },
    {
      "epoch": 0.09754907938056334,
      "grad_norm": 9.038830757141113,
      "learning_rate": 0.0,
      "loss": 2.4418,
      "step": 200
    },
    {
      "epoch": 0.09754907938056334,
      "eval_loss": 1.9695662260055542,
      "eval_runtime": 255.1082,
      "eval_samples_per_second": 13.535,
      "eval_steps_per_second": 3.387,
      "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": 2.929090691806003e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}