{
  "best_metric": 1.425579309463501,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.9378663540445487,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004689331770222743,
      "grad_norm": 21.725034713745117,
      "learning_rate": 1e-05,
      "loss": 15.1751,
      "step": 1
    },
    {
      "epoch": 0.004689331770222743,
      "eval_loss": 1.7309983968734741,
      "eval_runtime": 16.9956,
      "eval_samples_per_second": 84.492,
      "eval_steps_per_second": 5.295,
      "step": 1
    },
    {
      "epoch": 0.009378663540445486,
      "grad_norm": 7.985870361328125,
      "learning_rate": 2e-05,
      "loss": 13.8567,
      "step": 2
    },
    {
      "epoch": 0.01406799531066823,
      "grad_norm": 12.209921836853027,
      "learning_rate": 3e-05,
      "loss": 15.2047,
      "step": 3
    },
    {
      "epoch": 0.01875732708089097,
      "grad_norm": 9.236440658569336,
      "learning_rate": 4e-05,
      "loss": 15.253,
      "step": 4
    },
    {
      "epoch": 0.023446658851113716,
      "grad_norm": 34.01106262207031,
      "learning_rate": 5e-05,
      "loss": 13.8273,
      "step": 5
    },
    {
      "epoch": 0.02813599062133646,
      "grad_norm": 9.674768447875977,
      "learning_rate": 6e-05,
      "loss": 13.2691,
      "step": 6
    },
    {
      "epoch": 0.032825322391559206,
      "grad_norm": 7.357479572296143,
      "learning_rate": 7e-05,
      "loss": 13.7893,
      "step": 7
    },
    {
      "epoch": 0.03751465416178194,
      "grad_norm": 7.434540271759033,
      "learning_rate": 8e-05,
      "loss": 13.2297,
      "step": 8
    },
    {
      "epoch": 0.04220398593200469,
      "grad_norm": 7.8318562507629395,
      "learning_rate": 9e-05,
      "loss": 12.0348,
      "step": 9
    },
    {
      "epoch": 0.04689331770222743,
      "grad_norm": 12.641838073730469,
      "learning_rate": 0.0001,
      "loss": 12.2262,
      "step": 10
    },
    {
      "epoch": 0.05158264947245018,
      "grad_norm": 15.809853553771973,
      "learning_rate": 9.999316524962345e-05,
      "loss": 12.2885,
      "step": 11
    },
    {
      "epoch": 0.05627198124267292,
      "grad_norm": 33.80238342285156,
      "learning_rate": 9.997266286704631e-05,
      "loss": 14.3757,
      "step": 12
    },
    {
      "epoch": 0.06096131301289566,
      "grad_norm": 27.908411026000977,
      "learning_rate": 9.993849845741524e-05,
      "loss": 13.5397,
      "step": 13
    },
    {
      "epoch": 0.06565064478311841,
      "grad_norm": 12.773577690124512,
      "learning_rate": 9.989068136093873e-05,
      "loss": 10.6613,
      "step": 14
    },
    {
      "epoch": 0.07033997655334115,
      "grad_norm": 15.075891494750977,
      "learning_rate": 9.98292246503335e-05,
      "loss": 10.8771,
      "step": 15
    },
    {
      "epoch": 0.07502930832356389,
      "grad_norm": 9.669127464294434,
      "learning_rate": 9.975414512725057e-05,
      "loss": 9.5139,
      "step": 16
    },
    {
      "epoch": 0.07971864009378664,
      "grad_norm": 9.47426700592041,
      "learning_rate": 9.966546331768191e-05,
      "loss": 9.5607,
      "step": 17
    },
    {
      "epoch": 0.08440797186400938,
      "grad_norm": 6.41294002532959,
      "learning_rate": 9.956320346634876e-05,
      "loss": 9.3721,
      "step": 18
    },
    {
      "epoch": 0.08909730363423213,
      "grad_norm": 8.135649681091309,
      "learning_rate": 9.944739353007344e-05,
      "loss": 9.2091,
      "step": 19
    },
    {
      "epoch": 0.09378663540445487,
      "grad_norm": 6.92952823638916,
      "learning_rate": 9.931806517013612e-05,
      "loss": 8.9764,
      "step": 20
    },
    {
      "epoch": 0.0984759671746776,
      "grad_norm": 6.828975677490234,
      "learning_rate": 9.917525374361912e-05,
      "loss": 9.0491,
      "step": 21
    },
    {
      "epoch": 0.10316529894490035,
      "grad_norm": 8.996785163879395,
      "learning_rate": 9.901899829374047e-05,
      "loss": 9.3138,
      "step": 22
    },
    {
      "epoch": 0.10785463071512309,
      "grad_norm": 9.205068588256836,
      "learning_rate": 9.884934153917997e-05,
      "loss": 9.4437,
      "step": 23
    },
    {
      "epoch": 0.11254396248534584,
      "grad_norm": 12.9137544631958,
      "learning_rate": 9.86663298624003e-05,
      "loss": 10.2754,
      "step": 24
    },
    {
      "epoch": 0.11723329425556858,
      "grad_norm": 21.076824188232422,
      "learning_rate": 9.847001329696653e-05,
      "loss": 12.0929,
      "step": 25
    },
    {
      "epoch": 0.12192262602579132,
      "grad_norm": 8.461442947387695,
      "learning_rate": 9.826044551386744e-05,
      "loss": 8.2324,
      "step": 26
    },
    {
      "epoch": 0.12661195779601406,
      "grad_norm": 4.547621726989746,
      "learning_rate": 9.803768380684242e-05,
      "loss": 7.9646,
      "step": 27
    },
    {
      "epoch": 0.13130128956623682,
      "grad_norm": 4.0150580406188965,
      "learning_rate": 9.780178907671789e-05,
      "loss": 8.2603,
      "step": 28
    },
    {
      "epoch": 0.13599062133645956,
      "grad_norm": 5.907535076141357,
      "learning_rate": 9.755282581475769e-05,
      "loss": 7.9537,
      "step": 29
    },
    {
      "epoch": 0.1406799531066823,
      "grad_norm": 5.424262523651123,
      "learning_rate": 9.729086208503174e-05,
      "loss": 7.991,
      "step": 30
    },
    {
      "epoch": 0.14536928487690504,
      "grad_norm": 4.939754486083984,
      "learning_rate": 9.701596950580806e-05,
      "loss": 8.1568,
      "step": 31
    },
    {
      "epoch": 0.15005861664712777,
      "grad_norm": 14.181144714355469,
      "learning_rate": 9.672822322997305e-05,
      "loss": 8.5081,
      "step": 32
    },
    {
      "epoch": 0.15474794841735054,
      "grad_norm": 20.627016067504883,
      "learning_rate": 9.642770192448536e-05,
      "loss": 8.6137,
      "step": 33
    },
    {
      "epoch": 0.15943728018757328,
      "grad_norm": 7.873172760009766,
      "learning_rate": 9.611448774886924e-05,
      "loss": 8.3533,
      "step": 34
    },
    {
      "epoch": 0.16412661195779601,
      "grad_norm": 10.927459716796875,
      "learning_rate": 9.578866633275288e-05,
      "loss": 8.5853,
      "step": 35
    },
    {
      "epoch": 0.16881594372801875,
      "grad_norm": 10.551491737365723,
      "learning_rate": 9.545032675245813e-05,
      "loss": 8.7065,
      "step": 36
    },
    {
      "epoch": 0.1735052754982415,
      "grad_norm": 21.820234298706055,
      "learning_rate": 9.509956150664796e-05,
      "loss": 9.7237,
      "step": 37
    },
    {
      "epoch": 0.17819460726846426,
      "grad_norm": 17.7884464263916,
      "learning_rate": 9.473646649103818e-05,
      "loss": 9.95,
      "step": 38
    },
    {
      "epoch": 0.182883939038687,
      "grad_norm": 4.263730049133301,
      "learning_rate": 9.43611409721806e-05,
      "loss": 7.5972,
      "step": 39
    },
    {
      "epoch": 0.18757327080890973,
      "grad_norm": 4.870667457580566,
      "learning_rate": 9.397368756032445e-05,
      "loss": 7.5254,
      "step": 40
    },
    {
      "epoch": 0.19226260257913247,
      "grad_norm": 4.490108013153076,
      "learning_rate": 9.357421218136386e-05,
      "loss": 7.5034,
      "step": 41
    },
    {
      "epoch": 0.1969519343493552,
      "grad_norm": 5.4115681648254395,
      "learning_rate": 9.316282404787871e-05,
      "loss": 7.4857,
      "step": 42
    },
    {
      "epoch": 0.20164126611957797,
      "grad_norm": 5.6862568855285645,
      "learning_rate": 9.273963562927695e-05,
      "loss": 7.6218,
      "step": 43
    },
    {
      "epoch": 0.2063305978898007,
      "grad_norm": 5.7528395652771,
      "learning_rate": 9.230476262104677e-05,
      "loss": 7.8176,
      "step": 44
    },
    {
      "epoch": 0.21101992966002345,
      "grad_norm": 6.320159912109375,
      "learning_rate": 9.185832391312644e-05,
      "loss": 7.9366,
      "step": 45
    },
    {
      "epoch": 0.21570926143024619,
      "grad_norm": 6.65614652633667,
      "learning_rate": 9.140044155740101e-05,
      "loss": 8.0427,
      "step": 46
    },
    {
      "epoch": 0.22039859320046892,
      "grad_norm": 6.75200080871582,
      "learning_rate": 9.093124073433463e-05,
      "loss": 7.8885,
      "step": 47
    },
    {
      "epoch": 0.2250879249706917,
      "grad_norm": 8.359070777893066,
      "learning_rate": 9.045084971874738e-05,
      "loss": 7.7627,
      "step": 48
    },
    {
      "epoch": 0.22977725674091443,
      "grad_norm": 13.23523998260498,
      "learning_rate": 8.995939984474624e-05,
      "loss": 8.6928,
      "step": 49
    },
    {
      "epoch": 0.23446658851113716,
      "grad_norm": 23.538938522338867,
      "learning_rate": 8.945702546981969e-05,
      "loss": 9.9074,
      "step": 50
    },
    {
      "epoch": 0.23446658851113716,
      "eval_loss": 1.5064642429351807,
      "eval_runtime": 7.4192,
      "eval_samples_per_second": 193.551,
      "eval_steps_per_second": 12.131,
      "step": 50
    },
    {
      "epoch": 0.2391559202813599,
      "grad_norm": 4.802016735076904,
      "learning_rate": 8.894386393810563e-05,
      "loss": 6.9284,
      "step": 51
    },
    {
      "epoch": 0.24384525205158264,
      "grad_norm": 4.636072635650635,
      "learning_rate": 8.842005554284296e-05,
      "loss": 7.0464,
      "step": 52
    },
    {
      "epoch": 0.2485345838218054,
      "grad_norm": 4.342709064483643,
      "learning_rate": 8.788574348801675e-05,
      "loss": 7.1974,
      "step": 53
    },
    {
      "epoch": 0.2532239155920281,
      "grad_norm": 6.190311908721924,
      "learning_rate": 8.73410738492077e-05,
      "loss": 7.2172,
      "step": 54
    },
    {
      "epoch": 0.25791324736225085,
      "grad_norm": 4.974456310272217,
      "learning_rate": 8.678619553365659e-05,
      "loss": 7.3369,
      "step": 55
    },
    {
      "epoch": 0.26260257913247365,
      "grad_norm": 4.5510101318359375,
      "learning_rate": 8.622126023955446e-05,
      "loss": 7.3572,
      "step": 56
    },
    {
      "epoch": 0.2672919109026964,
      "grad_norm": 5.4362077713012695,
      "learning_rate": 8.564642241456986e-05,
      "loss": 7.5915,
      "step": 57
    },
    {
      "epoch": 0.2719812426729191,
      "grad_norm": 5.323110103607178,
      "learning_rate": 8.506183921362443e-05,
      "loss": 7.6154,
      "step": 58
    },
    {
      "epoch": 0.27667057444314186,
      "grad_norm": 6.725597381591797,
      "learning_rate": 8.44676704559283e-05,
      "loss": 7.8977,
      "step": 59
    },
    {
      "epoch": 0.2813599062133646,
      "grad_norm": 8.55213737487793,
      "learning_rate": 8.386407858128706e-05,
      "loss": 7.8923,
      "step": 60
    },
    {
      "epoch": 0.28604923798358733,
      "grad_norm": 10.243842124938965,
      "learning_rate": 8.32512286056924e-05,
      "loss": 8.0286,
      "step": 61
    },
    {
      "epoch": 0.29073856975381007,
      "grad_norm": 14.788423538208008,
      "learning_rate": 8.262928807620843e-05,
      "loss": 8.0237,
      "step": 62
    },
    {
      "epoch": 0.2954279015240328,
      "grad_norm": 11.049989700317383,
      "learning_rate": 8.199842702516583e-05,
      "loss": 8.221,
      "step": 63
    },
    {
      "epoch": 0.30011723329425555,
      "grad_norm": 4.678815841674805,
      "learning_rate": 8.135881792367686e-05,
      "loss": 6.9101,
      "step": 64
    },
    {
      "epoch": 0.3048065650644783,
      "grad_norm": 4.617557048797607,
      "learning_rate": 8.07106356344834e-05,
      "loss": 7.209,
      "step": 65
    },
    {
      "epoch": 0.3094958968347011,
      "grad_norm": 5.434515953063965,
      "learning_rate": 8.005405736415126e-05,
      "loss": 7.055,
      "step": 66
    },
    {
      "epoch": 0.3141852286049238,
      "grad_norm": 4.113319396972656,
      "learning_rate": 7.938926261462366e-05,
      "loss": 7.0158,
      "step": 67
    },
    {
      "epoch": 0.31887456037514655,
      "grad_norm": 4.233543872833252,
      "learning_rate": 7.871643313414718e-05,
      "loss": 7.3929,
      "step": 68
    },
    {
      "epoch": 0.3235638921453693,
      "grad_norm": 5.7781596183776855,
      "learning_rate": 7.803575286758364e-05,
      "loss": 7.1164,
      "step": 69
    },
    {
      "epoch": 0.32825322391559203,
      "grad_norm": 6.291742324829102,
      "learning_rate": 7.734740790612136e-05,
      "loss": 7.4607,
      "step": 70
    },
    {
      "epoch": 0.33294255568581477,
      "grad_norm": 6.425448417663574,
      "learning_rate": 7.66515864363997e-05,
      "loss": 7.3888,
      "step": 71
    },
    {
      "epoch": 0.3376318874560375,
      "grad_norm": 8.985451698303223,
      "learning_rate": 7.594847868906076e-05,
      "loss": 7.5019,
      "step": 72
    },
    {
      "epoch": 0.34232121922626024,
      "grad_norm": 28.634096145629883,
      "learning_rate": 7.52382768867422e-05,
      "loss": 7.4658,
      "step": 73
    },
    {
      "epoch": 0.347010550996483,
      "grad_norm": 11.263056755065918,
      "learning_rate": 7.452117519152542e-05,
      "loss": 7.7625,
      "step": 74
    },
    {
      "epoch": 0.3516998827667057,
      "grad_norm": 20.739336013793945,
      "learning_rate": 7.379736965185368e-05,
      "loss": 9.4905,
      "step": 75
    },
    {
      "epoch": 0.3563892145369285,
      "grad_norm": 3.723982810974121,
      "learning_rate": 7.30670581489344e-05,
      "loss": 6.7071,
      "step": 76
    },
    {
      "epoch": 0.36107854630715125,
      "grad_norm": 4.178149700164795,
      "learning_rate": 7.233044034264034e-05,
      "loss": 6.9562,
      "step": 77
    },
    {
      "epoch": 0.365767878077374,
      "grad_norm": 3.768890619277954,
      "learning_rate": 7.158771761692464e-05,
      "loss": 6.5697,
      "step": 78
    },
    {
      "epoch": 0.3704572098475967,
      "grad_norm": 4.64286994934082,
      "learning_rate": 7.083909302476453e-05,
      "loss": 7.0919,
      "step": 79
    },
    {
      "epoch": 0.37514654161781946,
      "grad_norm": 4.023698806762695,
      "learning_rate": 7.008477123264848e-05,
      "loss": 6.9348,
      "step": 80
    },
    {
      "epoch": 0.3798358733880422,
      "grad_norm": 5.7482709884643555,
      "learning_rate": 6.932495846462261e-05,
      "loss": 7.3517,
      "step": 81
    },
    {
      "epoch": 0.38452520515826494,
      "grad_norm": 4.70426607131958,
      "learning_rate": 6.855986244591104e-05,
      "loss": 7.2307,
      "step": 82
    },
    {
      "epoch": 0.3892145369284877,
      "grad_norm": 5.754972457885742,
      "learning_rate": 6.778969234612584e-05,
      "loss": 7.2414,
      "step": 83
    },
    {
      "epoch": 0.3939038686987104,
      "grad_norm": 5.299173355102539,
      "learning_rate": 6.701465872208216e-05,
      "loss": 7.2156,
      "step": 84
    },
    {
      "epoch": 0.39859320046893315,
      "grad_norm": 6.2285685539245605,
      "learning_rate": 6.623497346023418e-05,
      "loss": 7.2268,
      "step": 85
    },
    {
      "epoch": 0.40328253223915594,
      "grad_norm": 10.629804611206055,
      "learning_rate": 6.545084971874738e-05,
      "loss": 7.8437,
      "step": 86
    },
    {
      "epoch": 0.4079718640093787,
      "grad_norm": 13.943224906921387,
      "learning_rate": 6.466250186922325e-05,
      "loss": 7.414,
      "step": 87
    },
    {
      "epoch": 0.4126611957796014,
      "grad_norm": 10.789519309997559,
      "learning_rate": 6.387014543809223e-05,
      "loss": 8.0949,
      "step": 88
    },
    {
      "epoch": 0.41735052754982416,
      "grad_norm": 3.6843442916870117,
      "learning_rate": 6.307399704769099e-05,
      "loss": 6.7614,
      "step": 89
    },
    {
      "epoch": 0.4220398593200469,
      "grad_norm": 3.655954360961914,
      "learning_rate": 6.227427435703997e-05,
      "loss": 6.7671,
      "step": 90
    },
    {
      "epoch": 0.42672919109026963,
      "grad_norm": 3.750074863433838,
      "learning_rate": 6.147119600233758e-05,
      "loss": 6.8703,
      "step": 91
    },
    {
      "epoch": 0.43141852286049237,
      "grad_norm": 3.6812822818756104,
      "learning_rate": 6.066498153718735e-05,
      "loss": 7.0981,
      "step": 92
    },
    {
      "epoch": 0.4361078546307151,
      "grad_norm": 3.8529052734375,
      "learning_rate": 5.985585137257401e-05,
      "loss": 6.7913,
      "step": 93
    },
    {
      "epoch": 0.44079718640093785,
      "grad_norm": 5.374138355255127,
      "learning_rate": 5.90440267166055e-05,
      "loss": 7.2675,
      "step": 94
    },
    {
      "epoch": 0.4454865181711606,
      "grad_norm": 4.7173638343811035,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 7.1592,
      "step": 95
    },
    {
      "epoch": 0.4501758499413834,
      "grad_norm": 5.084648132324219,
      "learning_rate": 5.74131823855921e-05,
      "loss": 7.2873,
      "step": 96
    },
    {
      "epoch": 0.4548651817116061,
      "grad_norm": 5.910208225250244,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 6.9238,
      "step": 97
    },
    {
      "epoch": 0.45955451348182885,
      "grad_norm": 8.065096855163574,
      "learning_rate": 5.577423184847932e-05,
      "loss": 7.1742,
      "step": 98
    },
    {
      "epoch": 0.4642438452520516,
      "grad_norm": 10.980594635009766,
      "learning_rate": 5.495227651252315e-05,
      "loss": 7.7012,
      "step": 99
    },
    {
      "epoch": 0.46893317702227433,
      "grad_norm": 20.036052703857422,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 8.2287,
      "step": 100
    },
    {
      "epoch": 0.46893317702227433,
      "eval_loss": 1.4547187089920044,
      "eval_runtime": 7.4168,
      "eval_samples_per_second": 193.614,
      "eval_steps_per_second": 12.135,
      "step": 100
    },
    {
      "epoch": 0.47362250879249707,
      "grad_norm": 3.57832932472229,
      "learning_rate": 5.330452921628497e-05,
      "loss": 6.6729,
      "step": 101
    },
    {
      "epoch": 0.4783118405627198,
      "grad_norm": 4.062009811401367,
      "learning_rate": 5.247918773366112e-05,
      "loss": 6.6085,
      "step": 102
    },
    {
      "epoch": 0.48300117233294254,
      "grad_norm": 3.4798083305358887,
      "learning_rate": 5.165316846586541e-05,
      "loss": 6.765,
      "step": 103
    },
    {
      "epoch": 0.4876905041031653,
      "grad_norm": 4.188394546508789,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 6.7609,
      "step": 104
    },
    {
      "epoch": 0.492379835873388,
      "grad_norm": 3.8594112396240234,
      "learning_rate": 5e-05,
      "loss": 6.9061,
      "step": 105
    },
    {
      "epoch": 0.4970691676436108,
      "grad_norm": 3.694976806640625,
      "learning_rate": 4.917330276168208e-05,
      "loss": 6.8423,
      "step": 106
    },
    {
      "epoch": 0.5017584994138335,
      "grad_norm": 5.059836387634277,
      "learning_rate": 4.834683153413459e-05,
      "loss": 6.8287,
      "step": 107
    },
    {
      "epoch": 0.5064478311840562,
      "grad_norm": 5.029688835144043,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 7.1168,
      "step": 108
    },
    {
      "epoch": 0.511137162954279,
      "grad_norm": 4.955503940582275,
      "learning_rate": 4.669547078371504e-05,
      "loss": 7.2739,
      "step": 109
    },
    {
      "epoch": 0.5158264947245017,
      "grad_norm": 6.338578224182129,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 6.811,
      "step": 110
    },
    {
      "epoch": 0.5205158264947245,
      "grad_norm": 8.37576961517334,
      "learning_rate": 4.504772348747687e-05,
      "loss": 7.6846,
      "step": 111
    },
    {
      "epoch": 0.5252051582649473,
      "grad_norm": 12.16775894165039,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 7.6839,
      "step": 112
    },
    {
      "epoch": 0.52989449003517,
      "grad_norm": 13.318696975708008,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 8.2105,
      "step": 113
    },
    {
      "epoch": 0.5345838218053928,
      "grad_norm": 3.2152092456817627,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 6.8794,
      "step": 114
    },
    {
      "epoch": 0.5392731535756154,
      "grad_norm": 3.318774700164795,
      "learning_rate": 4.17702704859633e-05,
      "loss": 6.5404,
      "step": 115
    },
    {
      "epoch": 0.5439624853458382,
      "grad_norm": 4.0433573722839355,
      "learning_rate": 4.095597328339452e-05,
      "loss": 6.4892,
      "step": 116
    },
    {
      "epoch": 0.5486518171160609,
      "grad_norm": 5.393244743347168,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 6.8966,
      "step": 117
    },
    {
      "epoch": 0.5533411488862837,
      "grad_norm": 3.667290449142456,
      "learning_rate": 3.933501846281267e-05,
      "loss": 6.5456,
      "step": 118
    },
    {
      "epoch": 0.5580304806565064,
      "grad_norm": 4.325159072875977,
      "learning_rate": 3.852880399766243e-05,
      "loss": 6.8948,
      "step": 119
    },
    {
      "epoch": 0.5627198124267292,
      "grad_norm": 4.588203430175781,
      "learning_rate": 3.772572564296005e-05,
      "loss": 6.9063,
      "step": 120
    },
    {
      "epoch": 0.567409144196952,
      "grad_norm": 5.554940223693848,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 7.2434,
      "step": 121
    },
    {
      "epoch": 0.5720984759671747,
      "grad_norm": 5.843456745147705,
      "learning_rate": 3.612985456190778e-05,
      "loss": 7.0151,
      "step": 122
    },
    {
      "epoch": 0.5767878077373975,
      "grad_norm": 6.406453609466553,
      "learning_rate": 3.533749813077677e-05,
      "loss": 7.3105,
      "step": 123
    },
    {
      "epoch": 0.5814771395076201,
      "grad_norm": 9.872526168823242,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 7.2374,
      "step": 124
    },
    {
      "epoch": 0.5861664712778429,
      "grad_norm": 14.861069679260254,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 7.7956,
      "step": 125
    },
    {
      "epoch": 0.5908558030480656,
      "grad_norm": 2.7287437915802,
      "learning_rate": 3.298534127791785e-05,
      "loss": 6.575,
      "step": 126
    },
    {
      "epoch": 0.5955451348182884,
      "grad_norm": 3.332763910293579,
      "learning_rate": 3.221030765387417e-05,
      "loss": 6.6199,
      "step": 127
    },
    {
      "epoch": 0.6002344665885111,
      "grad_norm": 3.004457950592041,
      "learning_rate": 3.144013755408895e-05,
      "loss": 6.538,
      "step": 128
    },
    {
      "epoch": 0.6049237983587339,
      "grad_norm": 3.19492244720459,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 6.4676,
      "step": 129
    },
    {
      "epoch": 0.6096131301289566,
      "grad_norm": 3.649566650390625,
      "learning_rate": 2.991522876735154e-05,
      "loss": 6.6255,
      "step": 130
    },
    {
      "epoch": 0.6143024618991794,
      "grad_norm": 4.453381538391113,
      "learning_rate": 2.916090697523549e-05,
      "loss": 6.7886,
      "step": 131
    },
    {
      "epoch": 0.6189917936694022,
      "grad_norm": 4.349698543548584,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 6.9771,
      "step": 132
    },
    {
      "epoch": 0.6236811254396248,
      "grad_norm": 5.081732749938965,
      "learning_rate": 2.766955965735968e-05,
      "loss": 6.8414,
      "step": 133
    },
    {
      "epoch": 0.6283704572098476,
      "grad_norm": 5.801446437835693,
      "learning_rate": 2.693294185106562e-05,
      "loss": 7.0234,
      "step": 134
    },
    {
      "epoch": 0.6330597889800703,
      "grad_norm": 6.952717304229736,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 7.0204,
      "step": 135
    },
    {
      "epoch": 0.6377491207502931,
      "grad_norm": 8.456244468688965,
      "learning_rate": 2.547882480847461e-05,
      "loss": 6.9329,
      "step": 136
    },
    {
      "epoch": 0.6424384525205158,
      "grad_norm": 11.700654029846191,
      "learning_rate": 2.476172311325783e-05,
      "loss": 7.687,
      "step": 137
    },
    {
      "epoch": 0.6471277842907386,
      "grad_norm": 17.627973556518555,
      "learning_rate": 2.405152131093926e-05,
      "loss": 7.7505,
      "step": 138
    },
    {
      "epoch": 0.6518171160609613,
      "grad_norm": 2.8758513927459717,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 6.4286,
      "step": 139
    },
    {
      "epoch": 0.6565064478311841,
      "grad_norm": 3.151440382003784,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 6.5013,
      "step": 140
    },
    {
      "epoch": 0.6611957796014069,
      "grad_norm": 3.158478260040283,
      "learning_rate": 2.196424713241637e-05,
      "loss": 6.5987,
      "step": 141
    },
    {
      "epoch": 0.6658851113716295,
      "grad_norm": 3.5037615299224854,
      "learning_rate": 2.128356686585282e-05,
      "loss": 6.6357,
      "step": 142
    },
    {
      "epoch": 0.6705744431418523,
      "grad_norm": 3.7461934089660645,
      "learning_rate": 2.061073738537635e-05,
      "loss": 6.6647,
      "step": 143
    },
    {
      "epoch": 0.675263774912075,
      "grad_norm": 3.9924912452697754,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 6.9375,
      "step": 144
    },
    {
      "epoch": 0.6799531066822978,
      "grad_norm": 4.6558356285095215,
      "learning_rate": 1.928936436551661e-05,
      "loss": 6.7515,
      "step": 145
    },
    {
      "epoch": 0.6846424384525205,
      "grad_norm": 5.49552059173584,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 7.3371,
      "step": 146
    },
    {
      "epoch": 0.6893317702227433,
      "grad_norm": 5.995530128479004,
      "learning_rate": 1.800157297483417e-05,
      "loss": 7.1971,
      "step": 147
    },
    {
      "epoch": 0.694021101992966,
      "grad_norm": 7.859855651855469,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 7.241,
      "step": 148
    },
    {
      "epoch": 0.6987104337631888,
      "grad_norm": 9.96211051940918,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 7.3072,
      "step": 149
    },
    {
      "epoch": 0.7033997655334114,
      "grad_norm": 17.28060531616211,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 8.3215,
      "step": 150
    },
    {
      "epoch": 0.7033997655334114,
      "eval_loss": 1.435927391052246,
      "eval_runtime": 7.4096,
      "eval_samples_per_second": 193.804,
      "eval_steps_per_second": 12.146,
      "step": 150
    },
    {
      "epoch": 0.7080890973036342,
      "grad_norm": 2.880777597427368,
      "learning_rate": 1.553232954407171e-05,
      "loss": 6.4905,
      "step": 151
    },
    {
      "epoch": 0.712778429073857,
      "grad_norm": 2.9923276901245117,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 6.9008,
      "step": 152
    },
    {
      "epoch": 0.7174677608440797,
      "grad_norm": 3.4475700855255127,
      "learning_rate": 1.435357758543015e-05,
      "loss": 6.4464,
      "step": 153
    },
    {
      "epoch": 0.7221570926143025,
      "grad_norm": 3.19404935836792,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 6.2517,
      "step": 154
    },
    {
      "epoch": 0.7268464243845252,
      "grad_norm": 3.7995076179504395,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 6.6069,
      "step": 155
    },
    {
      "epoch": 0.731535756154748,
      "grad_norm": 4.021190166473389,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 6.8354,
      "step": 156
    },
    {
      "epoch": 0.7362250879249707,
      "grad_norm": 4.269179821014404,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 6.6669,
      "step": 157
    },
    {
      "epoch": 0.7409144196951934,
      "grad_norm": 5.380557537078857,
      "learning_rate": 1.157994445715706e-05,
      "loss": 7.1822,
      "step": 158
    },
    {
      "epoch": 0.7456037514654161,
      "grad_norm": 6.300451755523682,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 7.0165,
      "step": 159
    },
    {
      "epoch": 0.7502930832356389,
      "grad_norm": 7.254987716674805,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 6.8956,
      "step": 160
    },
    {
      "epoch": 0.7549824150058617,
      "grad_norm": 8.63143539428711,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 7.3455,
      "step": 161
    },
    {
      "epoch": 0.7596717467760844,
      "grad_norm": 15.235586166381836,
      "learning_rate": 9.549150281252633e-06,
      "loss": 7.0031,
      "step": 162
    },
    {
      "epoch": 0.7643610785463072,
      "grad_norm": 11.362581253051758,
      "learning_rate": 9.068759265665384e-06,
      "loss": 7.8937,
      "step": 163
    },
    {
      "epoch": 0.7690504103165299,
      "grad_norm": 2.7481682300567627,
      "learning_rate": 8.599558442598998e-06,
      "loss": 6.487,
      "step": 164
    },
    {
      "epoch": 0.7737397420867527,
      "grad_norm": 2.7565269470214844,
      "learning_rate": 8.141676086873572e-06,
      "loss": 6.4092,
      "step": 165
    },
    {
      "epoch": 0.7784290738569754,
      "grad_norm": 3.129683494567871,
      "learning_rate": 7.695237378953223e-06,
      "loss": 6.5016,
      "step": 166
    },
    {
      "epoch": 0.7831184056271981,
      "grad_norm": 3.2038040161132812,
      "learning_rate": 7.260364370723044e-06,
      "loss": 6.512,
      "step": 167
    },
    {
      "epoch": 0.7878077373974208,
      "grad_norm": 3.2864785194396973,
      "learning_rate": 6.837175952121306e-06,
      "loss": 6.5959,
      "step": 168
    },
    {
      "epoch": 0.7924970691676436,
      "grad_norm": 3.555297374725342,
      "learning_rate": 6.425787818636131e-06,
      "loss": 6.5844,
      "step": 169
    },
    {
      "epoch": 0.7971864009378663,
      "grad_norm": 4.174851894378662,
      "learning_rate": 6.026312439675552e-06,
      "loss": 6.6676,
      "step": 170
    },
    {
      "epoch": 0.8018757327080891,
      "grad_norm": 4.666030406951904,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 6.6626,
      "step": 171
    },
    {
      "epoch": 0.8065650644783119,
      "grad_norm": 5.393040180206299,
      "learning_rate": 5.263533508961827e-06,
      "loss": 7.0817,
      "step": 172
    },
    {
      "epoch": 0.8112543962485346,
      "grad_norm": 8.227474212646484,
      "learning_rate": 4.900438493352055e-06,
      "loss": 7.4928,
      "step": 173
    },
    {
      "epoch": 0.8159437280187574,
      "grad_norm": 9.667816162109375,
      "learning_rate": 4.549673247541875e-06,
      "loss": 7.3913,
      "step": 174
    },
    {
      "epoch": 0.82063305978898,
      "grad_norm": 12.965902328491211,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 8.0083,
      "step": 175
    },
    {
      "epoch": 0.8253223915592028,
      "grad_norm": 3.255803108215332,
      "learning_rate": 3.885512251130763e-06,
      "loss": 6.492,
      "step": 176
    },
    {
      "epoch": 0.8300117233294255,
      "grad_norm": 2.8813278675079346,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 6.1729,
      "step": 177
    },
    {
      "epoch": 0.8347010550996483,
      "grad_norm": 2.9110350608825684,
      "learning_rate": 3.271776770026963e-06,
      "loss": 6.2831,
      "step": 178
    },
    {
      "epoch": 0.839390386869871,
      "grad_norm": 3.2865688800811768,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 6.5589,
      "step": 179
    },
    {
      "epoch": 0.8440797186400938,
      "grad_norm": 3.622809410095215,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 6.7911,
      "step": 180
    },
    {
      "epoch": 0.8487690504103166,
      "grad_norm": 4.170779228210449,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 6.8754,
      "step": 181
    },
    {
      "epoch": 0.8534583821805393,
      "grad_norm": 4.805901527404785,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 6.8648,
      "step": 182
    },
    {
      "epoch": 0.8581477139507621,
      "grad_norm": 5.259015083312988,
      "learning_rate": 1.962316193157593e-06,
      "loss": 6.8166,
      "step": 183
    },
    {
      "epoch": 0.8628370457209847,
      "grad_norm": 6.820671081542969,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 7.1662,
      "step": 184
    },
    {
      "epoch": 0.8675263774912075,
      "grad_norm": 6.595808029174805,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 6.8149,
      "step": 185
    },
    {
      "epoch": 0.8722157092614302,
      "grad_norm": 8.34425163269043,
      "learning_rate": 1.333670137599713e-06,
      "loss": 7.031,
      "step": 186
    },
    {
      "epoch": 0.876905041031653,
      "grad_norm": 11.208433151245117,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 7.3296,
      "step": 187
    },
    {
      "epoch": 0.8815943728018757,
      "grad_norm": 11.770706176757812,
      "learning_rate": 9.810017062595322e-07,
      "loss": 7.2013,
      "step": 188
    },
    {
      "epoch": 0.8862837045720985,
      "grad_norm": 2.7891805171966553,
      "learning_rate": 8.247462563808817e-07,
      "loss": 6.2056,
      "step": 189
    },
    {
      "epoch": 0.8909730363423212,
      "grad_norm": 2.8778746128082275,
      "learning_rate": 6.819348298638839e-07,
      "loss": 6.3772,
      "step": 190
    },
    {
      "epoch": 0.895662368112544,
      "grad_norm": 2.9058592319488525,
      "learning_rate": 5.526064699265753e-07,
      "loss": 6.4887,
      "step": 191
    },
    {
      "epoch": 0.9003516998827668,
      "grad_norm": 3.0471181869506836,
      "learning_rate": 4.367965336512403e-07,
      "loss": 6.3516,
      "step": 192
    },
    {
      "epoch": 0.9050410316529894,
      "grad_norm": 4.441810131072998,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 6.6395,
      "step": 193
    },
    {
      "epoch": 0.9097303634232122,
      "grad_norm": 4.076968193054199,
      "learning_rate": 2.458548727494292e-07,
      "loss": 6.8421,
      "step": 194
    },
    {
      "epoch": 0.9144196951934349,
      "grad_norm": 4.745524883270264,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 6.7579,
      "step": 195
    },
    {
      "epoch": 0.9191090269636577,
      "grad_norm": 4.947898864746094,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 6.7419,
      "step": 196
    },
    {
      "epoch": 0.9237983587338804,
      "grad_norm": 5.085257530212402,
      "learning_rate": 6.150154258476315e-08,
      "loss": 6.8408,
      "step": 197
    },
    {
      "epoch": 0.9284876905041032,
      "grad_norm": 7.77234411239624,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 7.1667,
      "step": 198
    },
    {
      "epoch": 0.9331770222743259,
      "grad_norm": 9.531180381774902,
      "learning_rate": 6.834750376549792e-09,
      "loss": 6.8948,
      "step": 199
    },
    {
      "epoch": 0.9378663540445487,
      "grad_norm": 17.240276336669922,
      "learning_rate": 0.0,
      "loss": 7.6948,
      "step": 200
    },
    {
      "epoch": 0.9378663540445487,
      "eval_loss": 1.425579309463501,
      "eval_runtime": 7.4229,
      "eval_samples_per_second": 193.457,
      "eval_steps_per_second": 12.125,
      "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": 4.678847714820096e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}