{
  "best_metric": 1.908612608909607,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.06705783738474434,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00033528918692372173,
      "grad_norm": 3.0645880699157715,
      "learning_rate": 5e-06,
      "loss": 2.5227,
      "step": 1
    },
    {
      "epoch": 0.00033528918692372173,
      "eval_loss": 4.411320209503174,
      "eval_runtime": 413.4616,
      "eval_samples_per_second": 12.149,
      "eval_steps_per_second": 6.076,
      "step": 1
    },
    {
      "epoch": 0.0006705783738474435,
      "grad_norm": 4.630534648895264,
      "learning_rate": 1e-05,
      "loss": 3.4584,
      "step": 2
    },
    {
      "epoch": 0.001005867560771165,
      "grad_norm": 4.6843414306640625,
      "learning_rate": 1.5e-05,
      "loss": 3.2337,
      "step": 3
    },
    {
      "epoch": 0.001341156747694887,
      "grad_norm": 5.27373743057251,
      "learning_rate": 2e-05,
      "loss": 3.4903,
      "step": 4
    },
    {
      "epoch": 0.0016764459346186086,
      "grad_norm": 4.094366073608398,
      "learning_rate": 2.5e-05,
      "loss": 3.4112,
      "step": 5
    },
    {
      "epoch": 0.00201173512154233,
      "grad_norm": 4.469041347503662,
      "learning_rate": 3e-05,
      "loss": 3.9968,
      "step": 6
    },
    {
      "epoch": 0.002347024308466052,
      "grad_norm": 3.391988754272461,
      "learning_rate": 3.5e-05,
      "loss": 2.8687,
      "step": 7
    },
    {
      "epoch": 0.002682313495389774,
      "grad_norm": 4.106904029846191,
      "learning_rate": 4e-05,
      "loss": 3.1234,
      "step": 8
    },
    {
      "epoch": 0.0030176026823134955,
      "grad_norm": 4.628849029541016,
      "learning_rate": 4.5e-05,
      "loss": 2.5943,
      "step": 9
    },
    {
      "epoch": 0.003352891869237217,
      "grad_norm": 4.383242130279541,
      "learning_rate": 5e-05,
      "loss": 3.0693,
      "step": 10
    },
    {
      "epoch": 0.0036881810561609388,
      "grad_norm": 5.0880446434021,
      "learning_rate": 5.500000000000001e-05,
      "loss": 2.3701,
      "step": 11
    },
    {
      "epoch": 0.00402347024308466,
      "grad_norm": 4.474156856536865,
      "learning_rate": 6e-05,
      "loss": 2.3279,
      "step": 12
    },
    {
      "epoch": 0.0043587594300083824,
      "grad_norm": 3.052427053451538,
      "learning_rate": 6.500000000000001e-05,
      "loss": 2.382,
      "step": 13
    },
    {
      "epoch": 0.004694048616932104,
      "grad_norm": 3.632539749145508,
      "learning_rate": 7e-05,
      "loss": 2.3547,
      "step": 14
    },
    {
      "epoch": 0.005029337803855826,
      "grad_norm": 3.813579559326172,
      "learning_rate": 7.500000000000001e-05,
      "loss": 2.0444,
      "step": 15
    },
    {
      "epoch": 0.005364626990779548,
      "grad_norm": 3.673614025115967,
      "learning_rate": 8e-05,
      "loss": 2.4787,
      "step": 16
    },
    {
      "epoch": 0.005699916177703269,
      "grad_norm": 3.4429280757904053,
      "learning_rate": 8.5e-05,
      "loss": 2.0714,
      "step": 17
    },
    {
      "epoch": 0.006035205364626991,
      "grad_norm": 6.227026462554932,
      "learning_rate": 9e-05,
      "loss": 2.6872,
      "step": 18
    },
    {
      "epoch": 0.006370494551550712,
      "grad_norm": 2.7409024238586426,
      "learning_rate": 9.5e-05,
      "loss": 2.2862,
      "step": 19
    },
    {
      "epoch": 0.006705783738474434,
      "grad_norm": 3.4296722412109375,
      "learning_rate": 0.0001,
      "loss": 2.4185,
      "step": 20
    },
    {
      "epoch": 0.007041072925398156,
      "grad_norm": 3.6439619064331055,
      "learning_rate": 9.999238475781957e-05,
      "loss": 2.4921,
      "step": 21
    },
    {
      "epoch": 0.0073763621123218775,
      "grad_norm": 2.5447402000427246,
      "learning_rate": 9.99695413509548e-05,
      "loss": 2.1374,
      "step": 22
    },
    {
      "epoch": 0.0077116512992456,
      "grad_norm": 3.3165926933288574,
      "learning_rate": 9.99314767377287e-05,
      "loss": 2.3914,
      "step": 23
    },
    {
      "epoch": 0.00804694048616932,
      "grad_norm": 2.4111523628234863,
      "learning_rate": 9.987820251299122e-05,
      "loss": 2.1302,
      "step": 24
    },
    {
      "epoch": 0.008382229673093043,
      "grad_norm": 2.1574089527130127,
      "learning_rate": 9.980973490458728e-05,
      "loss": 2.1361,
      "step": 25
    },
    {
      "epoch": 0.008717518860016765,
      "grad_norm": 2.2673230171203613,
      "learning_rate": 9.972609476841367e-05,
      "loss": 1.9017,
      "step": 26
    },
    {
      "epoch": 0.009052808046940487,
      "grad_norm": 2.552058219909668,
      "learning_rate": 9.962730758206611e-05,
      "loss": 2.1296,
      "step": 27
    },
    {
      "epoch": 0.009388097233864207,
      "grad_norm": 2.4075942039489746,
      "learning_rate": 9.951340343707852e-05,
      "loss": 2.1832,
      "step": 28
    },
    {
      "epoch": 0.00972338642078793,
      "grad_norm": 2.210503101348877,
      "learning_rate": 9.938441702975689e-05,
      "loss": 2.1556,
      "step": 29
    },
    {
      "epoch": 0.010058675607711651,
      "grad_norm": 2.603090286254883,
      "learning_rate": 9.924038765061042e-05,
      "loss": 1.9968,
      "step": 30
    },
    {
      "epoch": 0.010393964794635373,
      "grad_norm": 2.427295446395874,
      "learning_rate": 9.908135917238321e-05,
      "loss": 2.0991,
      "step": 31
    },
    {
      "epoch": 0.010729253981559096,
      "grad_norm": 2.822526216506958,
      "learning_rate": 9.890738003669029e-05,
      "loss": 2.1003,
      "step": 32
    },
    {
      "epoch": 0.011064543168482816,
      "grad_norm": 2.231940269470215,
      "learning_rate": 9.871850323926177e-05,
      "loss": 2.2889,
      "step": 33
    },
    {
      "epoch": 0.011399832355406538,
      "grad_norm": 2.446948766708374,
      "learning_rate": 9.851478631379982e-05,
      "loss": 2.3449,
      "step": 34
    },
    {
      "epoch": 0.01173512154233026,
      "grad_norm": 2.6945621967315674,
      "learning_rate": 9.829629131445342e-05,
      "loss": 2.1744,
      "step": 35
    },
    {
      "epoch": 0.012070410729253982,
      "grad_norm": 2.202670097351074,
      "learning_rate": 9.806308479691595e-05,
      "loss": 2.4228,
      "step": 36
    },
    {
      "epoch": 0.012405699916177704,
      "grad_norm": 2.983804225921631,
      "learning_rate": 9.781523779815179e-05,
      "loss": 2.2251,
      "step": 37
    },
    {
      "epoch": 0.012740989103101424,
      "grad_norm": 3.077720880508423,
      "learning_rate": 9.755282581475769e-05,
      "loss": 2.274,
      "step": 38
    },
    {
      "epoch": 0.013076278290025146,
      "grad_norm": 1.9877910614013672,
      "learning_rate": 9.727592877996585e-05,
      "loss": 1.9802,
      "step": 39
    },
    {
      "epoch": 0.013411567476948869,
      "grad_norm": 2.4598207473754883,
      "learning_rate": 9.698463103929542e-05,
      "loss": 2.2072,
      "step": 40
    },
    {
      "epoch": 0.01374685666387259,
      "grad_norm": 2.6784284114837646,
      "learning_rate": 9.667902132486009e-05,
      "loss": 2.0905,
      "step": 41
    },
    {
      "epoch": 0.014082145850796313,
      "grad_norm": 2.057436943054199,
      "learning_rate": 9.635919272833938e-05,
      "loss": 1.9063,
      "step": 42
    },
    {
      "epoch": 0.014417435037720033,
      "grad_norm": 2.43062686920166,
      "learning_rate": 9.602524267262203e-05,
      "loss": 2.3002,
      "step": 43
    },
    {
      "epoch": 0.014752724224643755,
      "grad_norm": 2.5436949729919434,
      "learning_rate": 9.567727288213005e-05,
      "loss": 2.5027,
      "step": 44
    },
    {
      "epoch": 0.015088013411567477,
      "grad_norm": 2.3301496505737305,
      "learning_rate": 9.53153893518325e-05,
      "loss": 2.2546,
      "step": 45
    },
    {
      "epoch": 0.0154233025984912,
      "grad_norm": 2.0488905906677246,
      "learning_rate": 9.493970231495835e-05,
      "loss": 2.2543,
      "step": 46
    },
    {
      "epoch": 0.01575859178541492,
      "grad_norm": 1.8271653652191162,
      "learning_rate": 9.45503262094184e-05,
      "loss": 2.0725,
      "step": 47
    },
    {
      "epoch": 0.01609388097233864,
      "grad_norm": 1.8139671087265015,
      "learning_rate": 9.414737964294636e-05,
      "loss": 1.9978,
      "step": 48
    },
    {
      "epoch": 0.016429170159262364,
      "grad_norm": 3.030299186706543,
      "learning_rate": 9.373098535696979e-05,
      "loss": 2.083,
      "step": 49
    },
    {
      "epoch": 0.016764459346186086,
      "grad_norm": 2.121826410293579,
      "learning_rate": 9.330127018922194e-05,
      "loss": 2.1672,
      "step": 50
    },
    {
      "epoch": 0.016764459346186086,
      "eval_loss": 2.070281505584717,
      "eval_runtime": 415.3934,
      "eval_samples_per_second": 12.092,
      "eval_steps_per_second": 6.047,
      "step": 50
    },
    {
      "epoch": 0.017099748533109808,
      "grad_norm": 2.837768077850342,
      "learning_rate": 9.285836503510562e-05,
      "loss": 1.3665,
      "step": 51
    },
    {
      "epoch": 0.01743503772003353,
      "grad_norm": 2.6767208576202393,
      "learning_rate": 9.24024048078213e-05,
      "loss": 1.97,
      "step": 52
    },
    {
      "epoch": 0.017770326906957252,
      "grad_norm": 1.704293966293335,
      "learning_rate": 9.193352839727121e-05,
      "loss": 1.6777,
      "step": 53
    },
    {
      "epoch": 0.018105616093880974,
      "grad_norm": 1.6387865543365479,
      "learning_rate": 9.145187862775209e-05,
      "loss": 1.6942,
      "step": 54
    },
    {
      "epoch": 0.018440905280804692,
      "grad_norm": 1.7911442518234253,
      "learning_rate": 9.09576022144496e-05,
      "loss": 1.6699,
      "step": 55
    },
    {
      "epoch": 0.018776194467728415,
      "grad_norm": 1.837194800376892,
      "learning_rate": 9.045084971874738e-05,
      "loss": 2.002,
      "step": 56
    },
    {
      "epoch": 0.019111483654652137,
      "grad_norm": 2.0472233295440674,
      "learning_rate": 8.993177550236464e-05,
      "loss": 1.9996,
      "step": 57
    },
    {
      "epoch": 0.01944677284157586,
      "grad_norm": 1.6679171323776245,
      "learning_rate": 8.940053768033609e-05,
      "loss": 1.6527,
      "step": 58
    },
    {
      "epoch": 0.01978206202849958,
      "grad_norm": 1.514108419418335,
      "learning_rate": 8.885729807284856e-05,
      "loss": 1.6827,
      "step": 59
    },
    {
      "epoch": 0.020117351215423303,
      "grad_norm": 2.0194151401519775,
      "learning_rate": 8.83022221559489e-05,
      "loss": 2.168,
      "step": 60
    },
    {
      "epoch": 0.020452640402347025,
      "grad_norm": 1.5403530597686768,
      "learning_rate": 8.773547901113862e-05,
      "loss": 1.6586,
      "step": 61
    },
    {
      "epoch": 0.020787929589270747,
      "grad_norm": 1.7011923789978027,
      "learning_rate": 8.715724127386972e-05,
      "loss": 1.7703,
      "step": 62
    },
    {
      "epoch": 0.02112321877619447,
      "grad_norm": 1.5274035930633545,
      "learning_rate": 8.656768508095853e-05,
      "loss": 1.7972,
      "step": 63
    },
    {
      "epoch": 0.02145850796311819,
      "grad_norm": 1.4921406507492065,
      "learning_rate": 8.596699001693255e-05,
      "loss": 1.5572,
      "step": 64
    },
    {
      "epoch": 0.02179379715004191,
      "grad_norm": 1.6368732452392578,
      "learning_rate": 8.535533905932738e-05,
      "loss": 1.8478,
      "step": 65
    },
    {
      "epoch": 0.02212908633696563,
      "grad_norm": 2.0866599082946777,
      "learning_rate": 8.473291852294987e-05,
      "loss": 2.0398,
      "step": 66
    },
    {
      "epoch": 0.022464375523889354,
      "grad_norm": 1.8268249034881592,
      "learning_rate": 8.409991800312493e-05,
      "loss": 1.9158,
      "step": 67
    },
    {
      "epoch": 0.022799664710813076,
      "grad_norm": 1.592197060585022,
      "learning_rate": 8.345653031794292e-05,
      "loss": 1.8296,
      "step": 68
    },
    {
      "epoch": 0.023134953897736798,
      "grad_norm": 1.9506229162216187,
      "learning_rate": 8.280295144952536e-05,
      "loss": 2.001,
      "step": 69
    },
    {
      "epoch": 0.02347024308466052,
      "grad_norm": 1.7068800926208496,
      "learning_rate": 8.213938048432697e-05,
      "loss": 1.5299,
      "step": 70
    },
    {
      "epoch": 0.023805532271584242,
      "grad_norm": 1.5702540874481201,
      "learning_rate": 8.146601955249188e-05,
      "loss": 1.935,
      "step": 71
    },
    {
      "epoch": 0.024140821458507964,
      "grad_norm": 2.0129430294036865,
      "learning_rate": 8.07830737662829e-05,
      "loss": 2.0506,
      "step": 72
    },
    {
      "epoch": 0.024476110645431686,
      "grad_norm": 1.8408924341201782,
      "learning_rate": 8.009075115760243e-05,
      "loss": 2.2472,
      "step": 73
    },
    {
      "epoch": 0.024811399832355408,
      "grad_norm": 1.6532615423202515,
      "learning_rate": 7.938926261462366e-05,
      "loss": 1.7969,
      "step": 74
    },
    {
      "epoch": 0.025146689019279127,
      "grad_norm": 1.6880450248718262,
      "learning_rate": 7.86788218175523e-05,
      "loss": 1.9276,
      "step": 75
    },
    {
      "epoch": 0.02548197820620285,
      "grad_norm": 1.74911367893219,
      "learning_rate": 7.795964517353735e-05,
      "loss": 1.9406,
      "step": 76
    },
    {
      "epoch": 0.02581726739312657,
      "grad_norm": 1.6293855905532837,
      "learning_rate": 7.723195175075136e-05,
      "loss": 1.8643,
      "step": 77
    },
    {
      "epoch": 0.026152556580050293,
      "grad_norm": 1.846153974533081,
      "learning_rate": 7.649596321166024e-05,
      "loss": 1.9918,
      "step": 78
    },
    {
      "epoch": 0.026487845766974015,
      "grad_norm": 1.8405578136444092,
      "learning_rate": 7.575190374550272e-05,
      "loss": 2.0667,
      "step": 79
    },
    {
      "epoch": 0.026823134953897737,
      "grad_norm": 1.7012615203857422,
      "learning_rate": 7.500000000000001e-05,
      "loss": 2.2575,
      "step": 80
    },
    {
      "epoch": 0.02715842414082146,
      "grad_norm": 1.868895173072815,
      "learning_rate": 7.424048101231686e-05,
      "loss": 2.0323,
      "step": 81
    },
    {
      "epoch": 0.02749371332774518,
      "grad_norm": 1.862149953842163,
      "learning_rate": 7.347357813929454e-05,
      "loss": 2.1767,
      "step": 82
    },
    {
      "epoch": 0.027829002514668903,
      "grad_norm": 1.9703943729400635,
      "learning_rate": 7.269952498697734e-05,
      "loss": 2.0881,
      "step": 83
    },
    {
      "epoch": 0.028164291701592625,
      "grad_norm": 1.7949926853179932,
      "learning_rate": 7.191855733945387e-05,
      "loss": 2.2016,
      "step": 84
    },
    {
      "epoch": 0.028499580888516344,
      "grad_norm": 1.8068915605545044,
      "learning_rate": 7.113091308703498e-05,
      "loss": 2.2612,
      "step": 85
    },
    {
      "epoch": 0.028834870075440066,
      "grad_norm": 2.0017426013946533,
      "learning_rate": 7.033683215379002e-05,
      "loss": 2.0378,
      "step": 86
    },
    {
      "epoch": 0.029170159262363788,
      "grad_norm": 1.7997188568115234,
      "learning_rate": 6.953655642446368e-05,
      "loss": 1.8471,
      "step": 87
    },
    {
      "epoch": 0.02950544844928751,
      "grad_norm": 2.1644763946533203,
      "learning_rate": 6.873032967079561e-05,
      "loss": 2.0321,
      "step": 88
    },
    {
      "epoch": 0.029840737636211232,
      "grad_norm": 1.910346269607544,
      "learning_rate": 6.7918397477265e-05,
      "loss": 2.0989,
      "step": 89
    },
    {
      "epoch": 0.030176026823134954,
      "grad_norm": 1.8434735536575317,
      "learning_rate": 6.710100716628344e-05,
      "loss": 1.9885,
      "step": 90
    },
    {
      "epoch": 0.030511316010058676,
      "grad_norm": 2.194366216659546,
      "learning_rate": 6.627840772285784e-05,
      "loss": 2.1688,
      "step": 91
    },
    {
      "epoch": 0.0308466051969824,
      "grad_norm": 1.953049659729004,
      "learning_rate": 6.545084971874738e-05,
      "loss": 2.1352,
      "step": 92
    },
    {
      "epoch": 0.03118189438390612,
      "grad_norm": 2.183722734451294,
      "learning_rate": 6.461858523613684e-05,
      "loss": 2.3673,
      "step": 93
    },
    {
      "epoch": 0.03151718357082984,
      "grad_norm": 1.875823736190796,
      "learning_rate": 6.378186779084995e-05,
      "loss": 2.045,
      "step": 94
    },
    {
      "epoch": 0.03185247275775356,
      "grad_norm": 1.9691027402877808,
      "learning_rate": 6.294095225512603e-05,
      "loss": 2.3289,
      "step": 95
    },
    {
      "epoch": 0.03218776194467728,
      "grad_norm": 1.7443435192108154,
      "learning_rate": 6.209609477998338e-05,
      "loss": 2.033,
      "step": 96
    },
    {
      "epoch": 0.032523051131601005,
      "grad_norm": 1.6460736989974976,
      "learning_rate": 6.124755271719325e-05,
      "loss": 1.9156,
      "step": 97
    },
    {
      "epoch": 0.03285834031852473,
      "grad_norm": 1.6413999795913696,
      "learning_rate": 6.0395584540887963e-05,
      "loss": 2.1261,
      "step": 98
    },
    {
      "epoch": 0.03319362950544845,
      "grad_norm": 1.6792007684707642,
      "learning_rate": 5.9540449768827246e-05,
      "loss": 2.001,
      "step": 99
    },
    {
      "epoch": 0.03352891869237217,
      "grad_norm": 2.3478386402130127,
      "learning_rate": 5.868240888334653e-05,
      "loss": 2.215,
      "step": 100
    },
    {
      "epoch": 0.03352891869237217,
      "eval_loss": 1.971840500831604,
      "eval_runtime": 415.4443,
      "eval_samples_per_second": 12.091,
      "eval_steps_per_second": 6.047,
      "step": 100
    },
    {
      "epoch": 0.03386420787929589,
      "grad_norm": 2.282325029373169,
      "learning_rate": 5.782172325201155e-05,
      "loss": 1.6507,
      "step": 101
    },
    {
      "epoch": 0.034199497066219615,
      "grad_norm": 2.8105344772338867,
      "learning_rate": 5.695865504800327e-05,
      "loss": 2.0721,
      "step": 102
    },
    {
      "epoch": 0.03453478625314334,
      "grad_norm": 1.6801016330718994,
      "learning_rate": 5.6093467170257374e-05,
      "loss": 1.5186,
      "step": 103
    },
    {
      "epoch": 0.03487007544006706,
      "grad_norm": 1.7351365089416504,
      "learning_rate": 5.522642316338268e-05,
      "loss": 1.7305,
      "step": 104
    },
    {
      "epoch": 0.03520536462699078,
      "grad_norm": 1.8039164543151855,
      "learning_rate": 5.435778713738292e-05,
      "loss": 1.5557,
      "step": 105
    },
    {
      "epoch": 0.035540653813914504,
      "grad_norm": 1.6865015029907227,
      "learning_rate": 5.348782368720626e-05,
      "loss": 1.5957,
      "step": 106
    },
    {
      "epoch": 0.035875943000838226,
      "grad_norm": 1.7986878156661987,
      "learning_rate": 5.26167978121472e-05,
      "loss": 1.8013,
      "step": 107
    },
    {
      "epoch": 0.03621123218776195,
      "grad_norm": 1.8831130266189575,
      "learning_rate": 5.174497483512506e-05,
      "loss": 1.9391,
      "step": 108
    },
    {
      "epoch": 0.03654652137468566,
      "grad_norm": 1.7376582622528076,
      "learning_rate": 5.0872620321864185e-05,
      "loss": 1.7032,
      "step": 109
    },
    {
      "epoch": 0.036881810561609385,
      "grad_norm": 1.6134065389633179,
      "learning_rate": 5e-05,
      "loss": 1.6744,
      "step": 110
    },
    {
      "epoch": 0.03721709974853311,
      "grad_norm": 1.428790807723999,
      "learning_rate": 4.912737967813583e-05,
      "loss": 1.561,
      "step": 111
    },
    {
      "epoch": 0.03755238893545683,
      "grad_norm": 1.5310112237930298,
      "learning_rate": 4.825502516487497e-05,
      "loss": 1.399,
      "step": 112
    },
    {
      "epoch": 0.03788767812238055,
      "grad_norm": 1.5675244331359863,
      "learning_rate": 4.738320218785281e-05,
      "loss": 1.6196,
      "step": 113
    },
    {
      "epoch": 0.03822296730930427,
      "grad_norm": 1.6006245613098145,
      "learning_rate": 4.6512176312793736e-05,
      "loss": 1.9188,
      "step": 114
    },
    {
      "epoch": 0.038558256496227995,
      "grad_norm": 1.804556965827942,
      "learning_rate": 4.564221286261709e-05,
      "loss": 1.8277,
      "step": 115
    },
    {
      "epoch": 0.03889354568315172,
      "grad_norm": 1.5680475234985352,
      "learning_rate": 4.477357683661734e-05,
      "loss": 1.7761,
      "step": 116
    },
    {
      "epoch": 0.03922883487007544,
      "grad_norm": 1.500326156616211,
      "learning_rate": 4.390653282974264e-05,
      "loss": 1.721,
      "step": 117
    },
    {
      "epoch": 0.03956412405699916,
      "grad_norm": 1.2612425088882446,
      "learning_rate": 4.3041344951996746e-05,
      "loss": 1.5461,
      "step": 118
    },
    {
      "epoch": 0.039899413243922884,
      "grad_norm": 1.5856456756591797,
      "learning_rate": 4.2178276747988446e-05,
      "loss": 1.9058,
      "step": 119
    },
    {
      "epoch": 0.040234702430846606,
      "grad_norm": 1.386609435081482,
      "learning_rate": 4.131759111665349e-05,
      "loss": 1.6732,
      "step": 120
    },
    {
      "epoch": 0.04056999161777033,
      "grad_norm": 1.7134383916854858,
      "learning_rate": 4.045955023117276e-05,
      "loss": 1.8074,
      "step": 121
    },
    {
      "epoch": 0.04090528080469405,
      "grad_norm": 1.5574049949645996,
      "learning_rate": 3.960441545911204e-05,
      "loss": 1.9288,
      "step": 122
    },
    {
      "epoch": 0.04124056999161777,
      "grad_norm": 1.315619945526123,
      "learning_rate": 3.875244728280676e-05,
      "loss": 1.5456,
      "step": 123
    },
    {
      "epoch": 0.041575859178541494,
      "grad_norm": 1.625394582748413,
      "learning_rate": 3.790390522001662e-05,
      "loss": 1.7575,
      "step": 124
    },
    {
      "epoch": 0.041911148365465216,
      "grad_norm": 1.4131178855895996,
      "learning_rate": 3.705904774487396e-05,
      "loss": 1.835,
      "step": 125
    },
    {
      "epoch": 0.04224643755238894,
      "grad_norm": 1.5563520193099976,
      "learning_rate": 3.6218132209150045e-05,
      "loss": 1.8175,
      "step": 126
    },
    {
      "epoch": 0.04258172673931266,
      "grad_norm": 1.9665088653564453,
      "learning_rate": 3.5381414763863166e-05,
      "loss": 2.1785,
      "step": 127
    },
    {
      "epoch": 0.04291701592623638,
      "grad_norm": 1.935386300086975,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 2.0485,
      "step": 128
    },
    {
      "epoch": 0.0432523051131601,
      "grad_norm": 1.9241557121276855,
      "learning_rate": 3.372159227714218e-05,
      "loss": 1.9806,
      "step": 129
    },
    {
      "epoch": 0.04358759430008382,
      "grad_norm": 1.8136560916900635,
      "learning_rate": 3.289899283371657e-05,
      "loss": 1.8733,
      "step": 130
    },
    {
      "epoch": 0.04392288348700754,
      "grad_norm": 1.7437769174575806,
      "learning_rate": 3.2081602522734986e-05,
      "loss": 1.8345,
      "step": 131
    },
    {
      "epoch": 0.04425817267393126,
      "grad_norm": 1.565887451171875,
      "learning_rate": 3.12696703292044e-05,
      "loss": 1.7127,
      "step": 132
    },
    {
      "epoch": 0.044593461860854985,
      "grad_norm": 1.7738462686538696,
      "learning_rate": 3.046344357553632e-05,
      "loss": 1.7927,
      "step": 133
    },
    {
      "epoch": 0.04492875104777871,
      "grad_norm": 2.0633041858673096,
      "learning_rate": 2.9663167846209998e-05,
      "loss": 2.2202,
      "step": 134
    },
    {
      "epoch": 0.04526404023470243,
      "grad_norm": 1.881186604499817,
      "learning_rate": 2.886908691296504e-05,
      "loss": 1.9712,
      "step": 135
    },
    {
      "epoch": 0.04559932942162615,
      "grad_norm": 1.7657580375671387,
      "learning_rate": 2.8081442660546125e-05,
      "loss": 1.9912,
      "step": 136
    },
    {
      "epoch": 0.045934618608549874,
      "grad_norm": 1.7930335998535156,
      "learning_rate": 2.7300475013022663e-05,
      "loss": 2.2414,
      "step": 137
    },
    {
      "epoch": 0.046269907795473596,
      "grad_norm": 1.7103720903396606,
      "learning_rate": 2.6526421860705473e-05,
      "loss": 2.0095,
      "step": 138
    },
    {
      "epoch": 0.04660519698239732,
      "grad_norm": 1.8686071634292603,
      "learning_rate": 2.575951898768315e-05,
      "loss": 2.0788,
      "step": 139
    },
    {
      "epoch": 0.04694048616932104,
      "grad_norm": 1.8127381801605225,
      "learning_rate": 2.500000000000001e-05,
      "loss": 2.005,
      "step": 140
    },
    {
      "epoch": 0.04727577535624476,
      "grad_norm": 1.8599308729171753,
      "learning_rate": 2.4248096254497288e-05,
      "loss": 1.8135,
      "step": 141
    },
    {
      "epoch": 0.047611064543168484,
      "grad_norm": 1.8119877576828003,
      "learning_rate": 2.350403678833976e-05,
      "loss": 2.024,
      "step": 142
    },
    {
      "epoch": 0.047946353730092206,
      "grad_norm": 1.7392181158065796,
      "learning_rate": 2.2768048249248648e-05,
      "loss": 2.0515,
      "step": 143
    },
    {
      "epoch": 0.04828164291701593,
      "grad_norm": 1.9801397323608398,
      "learning_rate": 2.2040354826462668e-05,
      "loss": 2.233,
      "step": 144
    },
    {
      "epoch": 0.04861693210393965,
      "grad_norm": 1.7392927408218384,
      "learning_rate": 2.132117818244771e-05,
      "loss": 2.0363,
      "step": 145
    },
    {
      "epoch": 0.04895222129086337,
      "grad_norm": 1.7619372606277466,
      "learning_rate": 2.061073738537635e-05,
      "loss": 2.0353,
      "step": 146
    },
    {
      "epoch": 0.049287510477787094,
      "grad_norm": 2.0845959186553955,
      "learning_rate": 1.9909248842397584e-05,
      "loss": 2.0969,
      "step": 147
    },
    {
      "epoch": 0.049622799664710816,
      "grad_norm": 1.4988843202590942,
      "learning_rate": 1.9216926233717085e-05,
      "loss": 1.8349,
      "step": 148
    },
    {
      "epoch": 0.04995808885163453,
      "grad_norm": 1.5146304368972778,
      "learning_rate": 1.8533980447508137e-05,
      "loss": 1.9017,
      "step": 149
    },
    {
      "epoch": 0.050293378038558254,
      "grad_norm": 1.787965178489685,
      "learning_rate": 1.7860619515673033e-05,
      "loss": 1.9605,
      "step": 150
    },
    {
      "epoch": 0.050293378038558254,
      "eval_loss": 1.9188876152038574,
      "eval_runtime": 415.8121,
      "eval_samples_per_second": 12.08,
      "eval_steps_per_second": 6.041,
      "step": 150
    },
    {
      "epoch": 0.050628667225481976,
      "grad_norm": 1.7575668096542358,
      "learning_rate": 1.7197048550474643e-05,
      "loss": 1.6145,
      "step": 151
    },
    {
      "epoch": 0.0509639564124057,
      "grad_norm": 1.691274881362915,
      "learning_rate": 1.6543469682057106e-05,
      "loss": 1.8129,
      "step": 152
    },
    {
      "epoch": 0.05129924559932942,
      "grad_norm": 1.617201328277588,
      "learning_rate": 1.5900081996875083e-05,
      "loss": 1.9304,
      "step": 153
    },
    {
      "epoch": 0.05163453478625314,
      "grad_norm": 1.6812304258346558,
      "learning_rate": 1.526708147705013e-05,
      "loss": 2.0492,
      "step": 154
    },
    {
      "epoch": 0.051969823973176864,
      "grad_norm": 1.481422781944275,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 1.5662,
      "step": 155
    },
    {
      "epoch": 0.052305113160100586,
      "grad_norm": 1.2591601610183716,
      "learning_rate": 1.4033009983067452e-05,
      "loss": 1.2785,
      "step": 156
    },
    {
      "epoch": 0.05264040234702431,
      "grad_norm": 1.589030385017395,
      "learning_rate": 1.3432314919041478e-05,
      "loss": 1.8481,
      "step": 157
    },
    {
      "epoch": 0.05297569153394803,
      "grad_norm": 1.5457546710968018,
      "learning_rate": 1.2842758726130283e-05,
      "loss": 1.6473,
      "step": 158
    },
    {
      "epoch": 0.05331098072087175,
      "grad_norm": 1.4942293167114258,
      "learning_rate": 1.22645209888614e-05,
      "loss": 1.7979,
      "step": 159
    },
    {
      "epoch": 0.053646269907795474,
      "grad_norm": 1.363032341003418,
      "learning_rate": 1.1697777844051105e-05,
      "loss": 1.6266,
      "step": 160
    },
    {
      "epoch": 0.053981559094719196,
      "grad_norm": 1.531604528427124,
      "learning_rate": 1.1142701927151456e-05,
      "loss": 1.6577,
      "step": 161
    },
    {
      "epoch": 0.05431684828164292,
      "grad_norm": 1.5591459274291992,
      "learning_rate": 1.0599462319663905e-05,
      "loss": 1.8455,
      "step": 162
    },
    {
      "epoch": 0.05465213746856664,
      "grad_norm": 1.8428808450698853,
      "learning_rate": 1.006822449763537e-05,
      "loss": 1.7508,
      "step": 163
    },
    {
      "epoch": 0.05498742665549036,
      "grad_norm": 1.633676290512085,
      "learning_rate": 9.549150281252633e-06,
      "loss": 1.756,
      "step": 164
    },
    {
      "epoch": 0.055322715842414084,
      "grad_norm": 1.5237594842910767,
      "learning_rate": 9.042397785550405e-06,
      "loss": 1.8059,
      "step": 165
    },
    {
      "epoch": 0.055658005029337806,
      "grad_norm": 1.5941179990768433,
      "learning_rate": 8.548121372247918e-06,
      "loss": 1.8792,
      "step": 166
    },
    {
      "epoch": 0.05599329421626153,
      "grad_norm": 1.7206653356552124,
      "learning_rate": 8.066471602728803e-06,
      "loss": 1.9089,
      "step": 167
    },
    {
      "epoch": 0.05632858340318525,
      "grad_norm": 1.8432869911193848,
      "learning_rate": 7.597595192178702e-06,
      "loss": 1.8875,
      "step": 168
    },
    {
      "epoch": 0.056663872590108966,
      "grad_norm": 1.4452455043792725,
      "learning_rate": 7.1416349648943894e-06,
      "loss": 1.6278,
      "step": 169
    },
    {
      "epoch": 0.05699916177703269,
      "grad_norm": 1.4318848848342896,
      "learning_rate": 6.698729810778065e-06,
      "loss": 1.6474,
      "step": 170
    },
    {
      "epoch": 0.05733445096395641,
      "grad_norm": 1.6797986030578613,
      "learning_rate": 6.269014643030213e-06,
      "loss": 1.8615,
      "step": 171
    },
    {
      "epoch": 0.05766974015088013,
      "grad_norm": 1.742579698562622,
      "learning_rate": 5.852620357053651e-06,
      "loss": 1.9825,
      "step": 172
    },
    {
      "epoch": 0.058005029337803854,
      "grad_norm": 1.7321466207504272,
      "learning_rate": 5.449673790581611e-06,
      "loss": 2.0054,
      "step": 173
    },
    {
      "epoch": 0.058340318524727576,
      "grad_norm": 1.7640982866287231,
      "learning_rate": 5.060297685041659e-06,
      "loss": 2.0576,
      "step": 174
    },
    {
      "epoch": 0.0586756077116513,
      "grad_norm": 1.6197388172149658,
      "learning_rate": 4.684610648167503e-06,
      "loss": 2.0079,
      "step": 175
    },
    {
      "epoch": 0.05901089689857502,
      "grad_norm": 1.7537802457809448,
      "learning_rate": 4.322727117869951e-06,
      "loss": 1.901,
      "step": 176
    },
    {
      "epoch": 0.05934618608549874,
      "grad_norm": 1.588670253753662,
      "learning_rate": 3.974757327377981e-06,
      "loss": 1.7803,
      "step": 177
    },
    {
      "epoch": 0.059681475272422464,
      "grad_norm": 1.9405856132507324,
      "learning_rate": 3.6408072716606346e-06,
      "loss": 2.0009,
      "step": 178
    },
    {
      "epoch": 0.060016764459346186,
      "grad_norm": 1.615829586982727,
      "learning_rate": 3.3209786751399187e-06,
      "loss": 1.8363,
      "step": 179
    },
    {
      "epoch": 0.06035205364626991,
      "grad_norm": 2.093332529067993,
      "learning_rate": 3.0153689607045845e-06,
      "loss": 2.0939,
      "step": 180
    },
    {
      "epoch": 0.06068734283319363,
      "grad_norm": 1.9198389053344727,
      "learning_rate": 2.724071220034158e-06,
      "loss": 2.1039,
      "step": 181
    },
    {
      "epoch": 0.06102263202011735,
      "grad_norm": 1.8478941917419434,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 1.7815,
      "step": 182
    },
    {
      "epoch": 0.061357921207041075,
      "grad_norm": 1.6894562244415283,
      "learning_rate": 2.1847622018482283e-06,
      "loss": 2.0035,
      "step": 183
    },
    {
      "epoch": 0.0616932103939648,
      "grad_norm": 2.0324554443359375,
      "learning_rate": 1.9369152030840556e-06,
      "loss": 2.1021,
      "step": 184
    },
    {
      "epoch": 0.06202849958088852,
      "grad_norm": 1.6658598184585571,
      "learning_rate": 1.70370868554659e-06,
      "loss": 2.0876,
      "step": 185
    },
    {
      "epoch": 0.06236378876781224,
      "grad_norm": 1.7874583005905151,
      "learning_rate": 1.4852136862001764e-06,
      "loss": 2.1589,
      "step": 186
    },
    {
      "epoch": 0.06269907795473596,
      "grad_norm": 1.7210146188735962,
      "learning_rate": 1.2814967607382432e-06,
      "loss": 2.2337,
      "step": 187
    },
    {
      "epoch": 0.06303436714165968,
      "grad_norm": 1.880820870399475,
      "learning_rate": 1.0926199633097157e-06,
      "loss": 2.1527,
      "step": 188
    },
    {
      "epoch": 0.0633696563285834,
      "grad_norm": 1.5491559505462646,
      "learning_rate": 9.186408276168013e-07,
      "loss": 1.9314,
      "step": 189
    },
    {
      "epoch": 0.06370494551550712,
      "grad_norm": 1.7793771028518677,
      "learning_rate": 7.596123493895991e-07,
      "loss": 2.1539,
      "step": 190
    },
    {
      "epoch": 0.06404023470243085,
      "grad_norm": 1.826148271560669,
      "learning_rate": 6.15582970243117e-07,
      "loss": 2.02,
      "step": 191
    },
    {
      "epoch": 0.06437552388935457,
      "grad_norm": 1.6579688787460327,
      "learning_rate": 4.865965629214819e-07,
      "loss": 1.8212,
      "step": 192
    },
    {
      "epoch": 0.0647108130762783,
      "grad_norm": 1.6833443641662598,
      "learning_rate": 3.7269241793390085e-07,
      "loss": 1.767,
      "step": 193
    },
    {
      "epoch": 0.06504610226320201,
      "grad_norm": 1.7201486825942993,
      "learning_rate": 2.7390523158633554e-07,
      "loss": 1.9714,
      "step": 194
    },
    {
      "epoch": 0.06538139145012574,
      "grad_norm": 1.8961586952209473,
      "learning_rate": 1.9026509541272275e-07,
      "loss": 2.0871,
      "step": 195
    },
    {
      "epoch": 0.06571668063704945,
      "grad_norm": 1.6697744131088257,
      "learning_rate": 1.2179748700879012e-07,
      "loss": 2.0173,
      "step": 196
    },
    {
      "epoch": 0.06605196982397318,
      "grad_norm": 2.0735931396484375,
      "learning_rate": 6.852326227130834e-08,
      "loss": 2.3844,
      "step": 197
    },
    {
      "epoch": 0.0663872590108969,
      "grad_norm": 1.8878953456878662,
      "learning_rate": 3.04586490452119e-08,
      "loss": 1.8719,
      "step": 198
    },
    {
      "epoch": 0.06672254819782061,
      "grad_norm": 1.749005675315857,
      "learning_rate": 7.615242180436522e-09,
      "loss": 2.0864,
      "step": 199
    },
    {
      "epoch": 0.06705783738474434,
      "grad_norm": 1.9532133340835571,
      "learning_rate": 0.0,
      "loss": 2.0195,
      "step": 200
    },
    {
      "epoch": 0.06705783738474434,
      "eval_loss": 1.908612608909607,
      "eval_runtime": 416.1885,
      "eval_samples_per_second": 12.069,
      "eval_steps_per_second": 6.036,
      "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": 3.017027157491712e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}