{
  "best_metric": 1.3841888904571533,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.4158004158004158,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002079002079002079,
      "grad_norm": 28.46707534790039,
      "learning_rate": 1e-05,
      "loss": 6.7852,
      "step": 1
    },
    {
      "epoch": 0.002079002079002079,
      "eval_loss": 1.588921070098877,
      "eval_runtime": 16.5479,
      "eval_samples_per_second": 49.009,
      "eval_steps_per_second": 12.267,
      "step": 1
    },
    {
      "epoch": 0.004158004158004158,
      "grad_norm": 24.230926513671875,
      "learning_rate": 2e-05,
      "loss": 6.8426,
      "step": 2
    },
    {
      "epoch": 0.006237006237006237,
      "grad_norm": 24.949474334716797,
      "learning_rate": 3e-05,
      "loss": 7.2424,
      "step": 3
    },
    {
      "epoch": 0.008316008316008316,
      "grad_norm": 29.54012680053711,
      "learning_rate": 4e-05,
      "loss": 6.8775,
      "step": 4
    },
    {
      "epoch": 0.010395010395010396,
      "grad_norm": 24.96652603149414,
      "learning_rate": 5e-05,
      "loss": 6.7387,
      "step": 5
    },
    {
      "epoch": 0.012474012474012475,
      "grad_norm": 26.708059310913086,
      "learning_rate": 6e-05,
      "loss": 6.4657,
      "step": 6
    },
    {
      "epoch": 0.014553014553014554,
      "grad_norm": 24.244232177734375,
      "learning_rate": 7e-05,
      "loss": 6.8922,
      "step": 7
    },
    {
      "epoch": 0.016632016632016633,
      "grad_norm": 29.894004821777344,
      "learning_rate": 8e-05,
      "loss": 6.5568,
      "step": 8
    },
    {
      "epoch": 0.018711018711018712,
      "grad_norm": 28.081785202026367,
      "learning_rate": 9e-05,
      "loss": 6.2958,
      "step": 9
    },
    {
      "epoch": 0.02079002079002079,
      "grad_norm": 42.85635757446289,
      "learning_rate": 0.0001,
      "loss": 6.5382,
      "step": 10
    },
    {
      "epoch": 0.02286902286902287,
      "grad_norm": 25.6867618560791,
      "learning_rate": 9.999316524962345e-05,
      "loss": 6.1857,
      "step": 11
    },
    {
      "epoch": 0.02494802494802495,
      "grad_norm": 25.587675094604492,
      "learning_rate": 9.997266286704631e-05,
      "loss": 6.1499,
      "step": 12
    },
    {
      "epoch": 0.02702702702702703,
      "grad_norm": 29.740798950195312,
      "learning_rate": 9.993849845741524e-05,
      "loss": 6.1315,
      "step": 13
    },
    {
      "epoch": 0.029106029106029108,
      "grad_norm": 32.75001525878906,
      "learning_rate": 9.989068136093873e-05,
      "loss": 5.693,
      "step": 14
    },
    {
      "epoch": 0.031185031185031187,
      "grad_norm": 32.11300277709961,
      "learning_rate": 9.98292246503335e-05,
      "loss": 5.7046,
      "step": 15
    },
    {
      "epoch": 0.033264033264033266,
      "grad_norm": 33.56977081298828,
      "learning_rate": 9.975414512725057e-05,
      "loss": 5.6103,
      "step": 16
    },
    {
      "epoch": 0.035343035343035345,
      "grad_norm": 36.49180603027344,
      "learning_rate": 9.966546331768191e-05,
      "loss": 5.5331,
      "step": 17
    },
    {
      "epoch": 0.037422037422037424,
      "grad_norm": 34.360591888427734,
      "learning_rate": 9.956320346634876e-05,
      "loss": 5.5638,
      "step": 18
    },
    {
      "epoch": 0.0395010395010395,
      "grad_norm": 33.83558654785156,
      "learning_rate": 9.944739353007344e-05,
      "loss": 5.6453,
      "step": 19
    },
    {
      "epoch": 0.04158004158004158,
      "grad_norm": 35.50813293457031,
      "learning_rate": 9.931806517013612e-05,
      "loss": 5.2494,
      "step": 20
    },
    {
      "epoch": 0.04365904365904366,
      "grad_norm": 37.01751708984375,
      "learning_rate": 9.917525374361912e-05,
      "loss": 5.7079,
      "step": 21
    },
    {
      "epoch": 0.04573804573804574,
      "grad_norm": 53.92885971069336,
      "learning_rate": 9.901899829374047e-05,
      "loss": 4.9815,
      "step": 22
    },
    {
      "epoch": 0.04781704781704782,
      "grad_norm": 38.85811996459961,
      "learning_rate": 9.884934153917997e-05,
      "loss": 5.7015,
      "step": 23
    },
    {
      "epoch": 0.0498960498960499,
      "grad_norm": 38.31478500366211,
      "learning_rate": 9.86663298624003e-05,
      "loss": 5.4635,
      "step": 24
    },
    {
      "epoch": 0.05197505197505198,
      "grad_norm": 41.17340850830078,
      "learning_rate": 9.847001329696653e-05,
      "loss": 5.5956,
      "step": 25
    },
    {
      "epoch": 0.05405405405405406,
      "grad_norm": 36.10662841796875,
      "learning_rate": 9.826044551386744e-05,
      "loss": 5.2982,
      "step": 26
    },
    {
      "epoch": 0.056133056133056136,
      "grad_norm": 45.973751068115234,
      "learning_rate": 9.803768380684242e-05,
      "loss": 5.1937,
      "step": 27
    },
    {
      "epoch": 0.058212058212058215,
      "grad_norm": 37.021366119384766,
      "learning_rate": 9.780178907671789e-05,
      "loss": 5.4687,
      "step": 28
    },
    {
      "epoch": 0.060291060291060294,
      "grad_norm": 40.27494430541992,
      "learning_rate": 9.755282581475769e-05,
      "loss": 5.3546,
      "step": 29
    },
    {
      "epoch": 0.062370062370062374,
      "grad_norm": 35.98434066772461,
      "learning_rate": 9.729086208503174e-05,
      "loss": 5.4758,
      "step": 30
    },
    {
      "epoch": 0.06444906444906445,
      "grad_norm": 48.426021575927734,
      "learning_rate": 9.701596950580806e-05,
      "loss": 5.33,
      "step": 31
    },
    {
      "epoch": 0.06652806652806653,
      "grad_norm": 42.05217361450195,
      "learning_rate": 9.672822322997305e-05,
      "loss": 5.077,
      "step": 32
    },
    {
      "epoch": 0.06860706860706861,
      "grad_norm": 37.35602951049805,
      "learning_rate": 9.642770192448536e-05,
      "loss": 5.7777,
      "step": 33
    },
    {
      "epoch": 0.07068607068607069,
      "grad_norm": 51.590721130371094,
      "learning_rate": 9.611448774886924e-05,
      "loss": 5.6942,
      "step": 34
    },
    {
      "epoch": 0.07276507276507277,
      "grad_norm": 44.02403259277344,
      "learning_rate": 9.578866633275288e-05,
      "loss": 5.092,
      "step": 35
    },
    {
      "epoch": 0.07484407484407485,
      "grad_norm": 38.47739028930664,
      "learning_rate": 9.545032675245813e-05,
      "loss": 5.7395,
      "step": 36
    },
    {
      "epoch": 0.07692307692307693,
      "grad_norm": 40.031620025634766,
      "learning_rate": 9.509956150664796e-05,
      "loss": 5.3686,
      "step": 37
    },
    {
      "epoch": 0.079002079002079,
      "grad_norm": 41.19867706298828,
      "learning_rate": 9.473646649103818e-05,
      "loss": 5.5582,
      "step": 38
    },
    {
      "epoch": 0.08108108108108109,
      "grad_norm": 42.266849517822266,
      "learning_rate": 9.43611409721806e-05,
      "loss": 5.6561,
      "step": 39
    },
    {
      "epoch": 0.08316008316008316,
      "grad_norm": 37.43003845214844,
      "learning_rate": 9.397368756032445e-05,
      "loss": 5.8564,
      "step": 40
    },
    {
      "epoch": 0.08523908523908524,
      "grad_norm": 37.62900161743164,
      "learning_rate": 9.357421218136386e-05,
      "loss": 5.4157,
      "step": 41
    },
    {
      "epoch": 0.08731808731808732,
      "grad_norm": 39.51374816894531,
      "learning_rate": 9.316282404787871e-05,
      "loss": 5.5498,
      "step": 42
    },
    {
      "epoch": 0.0893970893970894,
      "grad_norm": 41.73848342895508,
      "learning_rate": 9.273963562927695e-05,
      "loss": 5.6719,
      "step": 43
    },
    {
      "epoch": 0.09147609147609148,
      "grad_norm": 33.238197326660156,
      "learning_rate": 9.230476262104677e-05,
      "loss": 6.0458,
      "step": 44
    },
    {
      "epoch": 0.09355509355509356,
      "grad_norm": 37.373695373535156,
      "learning_rate": 9.185832391312644e-05,
      "loss": 5.9867,
      "step": 45
    },
    {
      "epoch": 0.09563409563409564,
      "grad_norm": 40.94640350341797,
      "learning_rate": 9.140044155740101e-05,
      "loss": 6.3785,
      "step": 46
    },
    {
      "epoch": 0.09771309771309772,
      "grad_norm": 43.06721878051758,
      "learning_rate": 9.093124073433463e-05,
      "loss": 6.3831,
      "step": 47
    },
    {
      "epoch": 0.0997920997920998,
      "grad_norm": 48.167808532714844,
      "learning_rate": 9.045084971874738e-05,
      "loss": 6.3735,
      "step": 48
    },
    {
      "epoch": 0.10187110187110188,
      "grad_norm": 50.251670837402344,
      "learning_rate": 8.995939984474624e-05,
      "loss": 6.5997,
      "step": 49
    },
    {
      "epoch": 0.10395010395010396,
      "grad_norm": 69.40878295898438,
      "learning_rate": 8.945702546981969e-05,
      "loss": 5.6568,
      "step": 50
    },
    {
      "epoch": 0.10395010395010396,
      "eval_loss": 1.48008131980896,
      "eval_runtime": 16.6002,
      "eval_samples_per_second": 48.855,
      "eval_steps_per_second": 12.229,
      "step": 50
    },
    {
      "epoch": 0.10602910602910603,
      "grad_norm": 54.83725357055664,
      "learning_rate": 8.894386393810563e-05,
      "loss": 6.8567,
      "step": 51
    },
    {
      "epoch": 0.10810810810810811,
      "grad_norm": 56.870750427246094,
      "learning_rate": 8.842005554284296e-05,
      "loss": 6.4667,
      "step": 52
    },
    {
      "epoch": 0.1101871101871102,
      "grad_norm": 52.16533660888672,
      "learning_rate": 8.788574348801675e-05,
      "loss": 6.6183,
      "step": 53
    },
    {
      "epoch": 0.11226611226611227,
      "grad_norm": 56.41486740112305,
      "learning_rate": 8.73410738492077e-05,
      "loss": 6.7587,
      "step": 54
    },
    {
      "epoch": 0.11434511434511435,
      "grad_norm": 62.531707763671875,
      "learning_rate": 8.678619553365659e-05,
      "loss": 6.5008,
      "step": 55
    },
    {
      "epoch": 0.11642411642411643,
      "grad_norm": 63.76401138305664,
      "learning_rate": 8.622126023955446e-05,
      "loss": 6.6495,
      "step": 56
    },
    {
      "epoch": 0.11850311850311851,
      "grad_norm": 49.554752349853516,
      "learning_rate": 8.564642241456986e-05,
      "loss": 6.4602,
      "step": 57
    },
    {
      "epoch": 0.12058212058212059,
      "grad_norm": 72.33341217041016,
      "learning_rate": 8.506183921362443e-05,
      "loss": 6.4433,
      "step": 58
    },
    {
      "epoch": 0.12266112266112267,
      "grad_norm": 59.08444595336914,
      "learning_rate": 8.44676704559283e-05,
      "loss": 6.5198,
      "step": 59
    },
    {
      "epoch": 0.12474012474012475,
      "grad_norm": 52.813255310058594,
      "learning_rate": 8.386407858128706e-05,
      "loss": 6.2811,
      "step": 60
    },
    {
      "epoch": 0.12681912681912683,
      "grad_norm": 46.1800651550293,
      "learning_rate": 8.32512286056924e-05,
      "loss": 6.0972,
      "step": 61
    },
    {
      "epoch": 0.1288981288981289,
      "grad_norm": 49.58475875854492,
      "learning_rate": 8.262928807620843e-05,
      "loss": 5.8233,
      "step": 62
    },
    {
      "epoch": 0.13097713097713098,
      "grad_norm": 41.38501739501953,
      "learning_rate": 8.199842702516583e-05,
      "loss": 6.0824,
      "step": 63
    },
    {
      "epoch": 0.13305613305613306,
      "grad_norm": 43.60585403442383,
      "learning_rate": 8.135881792367686e-05,
      "loss": 5.8158,
      "step": 64
    },
    {
      "epoch": 0.13513513513513514,
      "grad_norm": 43.05984115600586,
      "learning_rate": 8.07106356344834e-05,
      "loss": 5.6642,
      "step": 65
    },
    {
      "epoch": 0.13721413721413722,
      "grad_norm": 43.71196365356445,
      "learning_rate": 8.005405736415126e-05,
      "loss": 5.5613,
      "step": 66
    },
    {
      "epoch": 0.1392931392931393,
      "grad_norm": 45.35591125488281,
      "learning_rate": 7.938926261462366e-05,
      "loss": 5.0937,
      "step": 67
    },
    {
      "epoch": 0.14137214137214138,
      "grad_norm": 37.6486930847168,
      "learning_rate": 7.871643313414718e-05,
      "loss": 5.5512,
      "step": 68
    },
    {
      "epoch": 0.14345114345114346,
      "grad_norm": 41.03268814086914,
      "learning_rate": 7.803575286758364e-05,
      "loss": 5.6187,
      "step": 69
    },
    {
      "epoch": 0.14553014553014554,
      "grad_norm": 40.384185791015625,
      "learning_rate": 7.734740790612136e-05,
      "loss": 5.4969,
      "step": 70
    },
    {
      "epoch": 0.14760914760914762,
      "grad_norm": 44.29239273071289,
      "learning_rate": 7.66515864363997e-05,
      "loss": 5.0777,
      "step": 71
    },
    {
      "epoch": 0.1496881496881497,
      "grad_norm": 41.99385452270508,
      "learning_rate": 7.594847868906076e-05,
      "loss": 5.5291,
      "step": 72
    },
    {
      "epoch": 0.15176715176715178,
      "grad_norm": 40.74198532104492,
      "learning_rate": 7.52382768867422e-05,
      "loss": 5.4119,
      "step": 73
    },
    {
      "epoch": 0.15384615384615385,
      "grad_norm": 40.196258544921875,
      "learning_rate": 7.452117519152542e-05,
      "loss": 5.6737,
      "step": 74
    },
    {
      "epoch": 0.15592515592515593,
      "grad_norm": 42.999290466308594,
      "learning_rate": 7.379736965185368e-05,
      "loss": 5.2002,
      "step": 75
    },
    {
      "epoch": 0.158004158004158,
      "grad_norm": 41.26256561279297,
      "learning_rate": 7.30670581489344e-05,
      "loss": 4.7373,
      "step": 76
    },
    {
      "epoch": 0.1600831600831601,
      "grad_norm": 37.30693435668945,
      "learning_rate": 7.233044034264034e-05,
      "loss": 5.1658,
      "step": 77
    },
    {
      "epoch": 0.16216216216216217,
      "grad_norm": 34.851322174072266,
      "learning_rate": 7.158771761692464e-05,
      "loss": 5.2531,
      "step": 78
    },
    {
      "epoch": 0.16424116424116425,
      "grad_norm": 37.84699630737305,
      "learning_rate": 7.083909302476453e-05,
      "loss": 4.964,
      "step": 79
    },
    {
      "epoch": 0.16632016632016633,
      "grad_norm": 40.300804138183594,
      "learning_rate": 7.008477123264848e-05,
      "loss": 5.0915,
      "step": 80
    },
    {
      "epoch": 0.1683991683991684,
      "grad_norm": 39.61699676513672,
      "learning_rate": 6.932495846462261e-05,
      "loss": 5.4426,
      "step": 81
    },
    {
      "epoch": 0.1704781704781705,
      "grad_norm": 40.3549919128418,
      "learning_rate": 6.855986244591104e-05,
      "loss": 5.197,
      "step": 82
    },
    {
      "epoch": 0.17255717255717257,
      "grad_norm": 34.17180633544922,
      "learning_rate": 6.778969234612584e-05,
      "loss": 4.9338,
      "step": 83
    },
    {
      "epoch": 0.17463617463617465,
      "grad_norm": 46.330501556396484,
      "learning_rate": 6.701465872208216e-05,
      "loss": 5.4305,
      "step": 84
    },
    {
      "epoch": 0.17671517671517672,
      "grad_norm": 43.79049301147461,
      "learning_rate": 6.623497346023418e-05,
      "loss": 5.5257,
      "step": 85
    },
    {
      "epoch": 0.1787941787941788,
      "grad_norm": 37.91574478149414,
      "learning_rate": 6.545084971874738e-05,
      "loss": 5.6571,
      "step": 86
    },
    {
      "epoch": 0.18087318087318088,
      "grad_norm": 70.67979431152344,
      "learning_rate": 6.466250186922325e-05,
      "loss": 5.048,
      "step": 87
    },
    {
      "epoch": 0.18295218295218296,
      "grad_norm": 33.63069152832031,
      "learning_rate": 6.387014543809223e-05,
      "loss": 5.2357,
      "step": 88
    },
    {
      "epoch": 0.18503118503118504,
      "grad_norm": 39.30858612060547,
      "learning_rate": 6.307399704769099e-05,
      "loss": 5.8054,
      "step": 89
    },
    {
      "epoch": 0.18711018711018712,
      "grad_norm": 38.35932922363281,
      "learning_rate": 6.227427435703997e-05,
      "loss": 5.6826,
      "step": 90
    },
    {
      "epoch": 0.1891891891891892,
      "grad_norm": 37.136924743652344,
      "learning_rate": 6.147119600233758e-05,
      "loss": 5.8794,
      "step": 91
    },
    {
      "epoch": 0.19126819126819128,
      "grad_norm": 38.46394729614258,
      "learning_rate": 6.066498153718735e-05,
      "loss": 6.0533,
      "step": 92
    },
    {
      "epoch": 0.19334719334719336,
      "grad_norm": 44.62322998046875,
      "learning_rate": 5.985585137257401e-05,
      "loss": 6.2104,
      "step": 93
    },
    {
      "epoch": 0.19542619542619544,
      "grad_norm": 32.984745025634766,
      "learning_rate": 5.90440267166055e-05,
      "loss": 5.7724,
      "step": 94
    },
    {
      "epoch": 0.19750519750519752,
      "grad_norm": 28.310604095458984,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 6.1788,
      "step": 95
    },
    {
      "epoch": 0.1995841995841996,
      "grad_norm": 31.271846771240234,
      "learning_rate": 5.74131823855921e-05,
      "loss": 6.748,
      "step": 96
    },
    {
      "epoch": 0.20166320166320167,
      "grad_norm": 29.09902572631836,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 6.3062,
      "step": 97
    },
    {
      "epoch": 0.20374220374220375,
      "grad_norm": 29.931982040405273,
      "learning_rate": 5.577423184847932e-05,
      "loss": 6.7834,
      "step": 98
    },
    {
      "epoch": 0.20582120582120583,
      "grad_norm": 26.51114845275879,
      "learning_rate": 5.495227651252315e-05,
      "loss": 6.5855,
      "step": 99
    },
    {
      "epoch": 0.2079002079002079,
      "grad_norm": 31.717966079711914,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 5.4724,
      "step": 100
    },
    {
      "epoch": 0.2079002079002079,
      "eval_loss": 1.4331369400024414,
      "eval_runtime": 16.5933,
      "eval_samples_per_second": 48.875,
      "eval_steps_per_second": 12.234,
      "step": 100
    },
    {
      "epoch": 0.20997920997921,
      "grad_norm": 26.54711151123047,
      "learning_rate": 5.330452921628497e-05,
      "loss": 6.167,
      "step": 101
    },
    {
      "epoch": 0.21205821205821207,
      "grad_norm": 25.854795455932617,
      "learning_rate": 5.247918773366112e-05,
      "loss": 6.344,
      "step": 102
    },
    {
      "epoch": 0.21413721413721415,
      "grad_norm": 25.46681785583496,
      "learning_rate": 5.165316846586541e-05,
      "loss": 6.4483,
      "step": 103
    },
    {
      "epoch": 0.21621621621621623,
      "grad_norm": 32.25164794921875,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 6.3607,
      "step": 104
    },
    {
      "epoch": 0.2182952182952183,
      "grad_norm": 28.720914840698242,
      "learning_rate": 5e-05,
      "loss": 6.2215,
      "step": 105
    },
    {
      "epoch": 0.2203742203742204,
      "grad_norm": 22.381977081298828,
      "learning_rate": 4.917330276168208e-05,
      "loss": 6.5318,
      "step": 106
    },
    {
      "epoch": 0.22245322245322247,
      "grad_norm": 25.075542449951172,
      "learning_rate": 4.834683153413459e-05,
      "loss": 5.9871,
      "step": 107
    },
    {
      "epoch": 0.22453222453222454,
      "grad_norm": 21.54170036315918,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 6.103,
      "step": 108
    },
    {
      "epoch": 0.22661122661122662,
      "grad_norm": 21.44270133972168,
      "learning_rate": 4.669547078371504e-05,
      "loss": 5.9033,
      "step": 109
    },
    {
      "epoch": 0.2286902286902287,
      "grad_norm": 21.39537239074707,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 5.7634,
      "step": 110
    },
    {
      "epoch": 0.23076923076923078,
      "grad_norm": 21.443811416625977,
      "learning_rate": 4.504772348747687e-05,
      "loss": 6.0082,
      "step": 111
    },
    {
      "epoch": 0.23284823284823286,
      "grad_norm": 21.43872833251953,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 5.9661,
      "step": 112
    },
    {
      "epoch": 0.23492723492723494,
      "grad_norm": 22.911455154418945,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 5.4578,
      "step": 113
    },
    {
      "epoch": 0.23700623700623702,
      "grad_norm": 23.506946563720703,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 5.2979,
      "step": 114
    },
    {
      "epoch": 0.2390852390852391,
      "grad_norm": 30.248619079589844,
      "learning_rate": 4.17702704859633e-05,
      "loss": 5.3445,
      "step": 115
    },
    {
      "epoch": 0.24116424116424118,
      "grad_norm": 24.931686401367188,
      "learning_rate": 4.095597328339452e-05,
      "loss": 5.1376,
      "step": 116
    },
    {
      "epoch": 0.24324324324324326,
      "grad_norm": 32.94673156738281,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 5.4534,
      "step": 117
    },
    {
      "epoch": 0.24532224532224534,
      "grad_norm": 25.710914611816406,
      "learning_rate": 3.933501846281267e-05,
      "loss": 5.1895,
      "step": 118
    },
    {
      "epoch": 0.24740124740124741,
      "grad_norm": 22.931718826293945,
      "learning_rate": 3.852880399766243e-05,
      "loss": 5.4679,
      "step": 119
    },
    {
      "epoch": 0.2494802494802495,
      "grad_norm": 31.658096313476562,
      "learning_rate": 3.772572564296005e-05,
      "loss": 5.0857,
      "step": 120
    },
    {
      "epoch": 0.2515592515592516,
      "grad_norm": 48.58084487915039,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 5.326,
      "step": 121
    },
    {
      "epoch": 0.25363825363825365,
      "grad_norm": 24.028602600097656,
      "learning_rate": 3.612985456190778e-05,
      "loss": 5.0892,
      "step": 122
    },
    {
      "epoch": 0.25571725571725573,
      "grad_norm": 25.648685455322266,
      "learning_rate": 3.533749813077677e-05,
      "loss": 5.5187,
      "step": 123
    },
    {
      "epoch": 0.2577962577962578,
      "grad_norm": 28.700227737426758,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 5.2872,
      "step": 124
    },
    {
      "epoch": 0.2598752598752599,
      "grad_norm": 23.682453155517578,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 5.2747,
      "step": 125
    },
    {
      "epoch": 0.26195426195426197,
      "grad_norm": 25.20798110961914,
      "learning_rate": 3.298534127791785e-05,
      "loss": 5.4999,
      "step": 126
    },
    {
      "epoch": 0.26403326403326405,
      "grad_norm": 19.902894973754883,
      "learning_rate": 3.221030765387417e-05,
      "loss": 5.191,
      "step": 127
    },
    {
      "epoch": 0.2661122661122661,
      "grad_norm": 25.95762062072754,
      "learning_rate": 3.144013755408895e-05,
      "loss": 4.7488,
      "step": 128
    },
    {
      "epoch": 0.2681912681912682,
      "grad_norm": 26.95562171936035,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 4.8508,
      "step": 129
    },
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 21.696176528930664,
      "learning_rate": 2.991522876735154e-05,
      "loss": 5.2093,
      "step": 130
    },
    {
      "epoch": 0.27234927234927236,
      "grad_norm": 26.578813552856445,
      "learning_rate": 2.916090697523549e-05,
      "loss": 5.1576,
      "step": 131
    },
    {
      "epoch": 0.27442827442827444,
      "grad_norm": 27.932218551635742,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 5.0432,
      "step": 132
    },
    {
      "epoch": 0.2765072765072765,
      "grad_norm": 24.860689163208008,
      "learning_rate": 2.766955965735968e-05,
      "loss": 5.1169,
      "step": 133
    },
    {
      "epoch": 0.2785862785862786,
      "grad_norm": 23.492446899414062,
      "learning_rate": 2.693294185106562e-05,
      "loss": 5.5201,
      "step": 134
    },
    {
      "epoch": 0.2806652806652807,
      "grad_norm": 74.35052490234375,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 5.3919,
      "step": 135
    },
    {
      "epoch": 0.28274428274428276,
      "grad_norm": 27.7348690032959,
      "learning_rate": 2.547882480847461e-05,
      "loss": 5.2264,
      "step": 136
    },
    {
      "epoch": 0.28482328482328484,
      "grad_norm": 27.24156379699707,
      "learning_rate": 2.476172311325783e-05,
      "loss": 5.0038,
      "step": 137
    },
    {
      "epoch": 0.2869022869022869,
      "grad_norm": 64.3189697265625,
      "learning_rate": 2.405152131093926e-05,
      "loss": 5.5356,
      "step": 138
    },
    {
      "epoch": 0.288981288981289,
      "grad_norm": 29.200817108154297,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 5.0124,
      "step": 139
    },
    {
      "epoch": 0.2910602910602911,
      "grad_norm": 33.68327713012695,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 5.1573,
      "step": 140
    },
    {
      "epoch": 0.29313929313929316,
      "grad_norm": 29.372739791870117,
      "learning_rate": 2.196424713241637e-05,
      "loss": 5.2473,
      "step": 141
    },
    {
      "epoch": 0.29521829521829523,
      "grad_norm": 35.02393341064453,
      "learning_rate": 2.128356686585282e-05,
      "loss": 5.8317,
      "step": 142
    },
    {
      "epoch": 0.2972972972972973,
      "grad_norm": 26.72066879272461,
      "learning_rate": 2.061073738537635e-05,
      "loss": 5.7935,
      "step": 143
    },
    {
      "epoch": 0.2993762993762994,
      "grad_norm": 28.788591384887695,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 5.5851,
      "step": 144
    },
    {
      "epoch": 0.30145530145530147,
      "grad_norm": 27.046279907226562,
      "learning_rate": 1.928936436551661e-05,
      "loss": 6.2411,
      "step": 145
    },
    {
      "epoch": 0.30353430353430355,
      "grad_norm": 75.47319793701172,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 6.5645,
      "step": 146
    },
    {
      "epoch": 0.30561330561330563,
      "grad_norm": 29.946130752563477,
      "learning_rate": 1.800157297483417e-05,
      "loss": 5.9446,
      "step": 147
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 35.889068603515625,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 6.0008,
      "step": 148
    },
    {
      "epoch": 0.3097713097713098,
      "grad_norm": 27.64640235900879,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 6.7218,
      "step": 149
    },
    {
      "epoch": 0.31185031185031187,
      "grad_norm": 32.522701263427734,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 5.5473,
      "step": 150
    },
    {
      "epoch": 0.31185031185031187,
      "eval_loss": 1.3959094285964966,
      "eval_runtime": 16.5955,
      "eval_samples_per_second": 48.869,
      "eval_steps_per_second": 12.232,
      "step": 150
    },
    {
      "epoch": 0.31392931392931395,
      "grad_norm": 35.28411102294922,
      "learning_rate": 1.553232954407171e-05,
      "loss": 6.3015,
      "step": 151
    },
    {
      "epoch": 0.316008316008316,
      "grad_norm": 32.234703063964844,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 5.9332,
      "step": 152
    },
    {
      "epoch": 0.3180873180873181,
      "grad_norm": 37.303131103515625,
      "learning_rate": 1.435357758543015e-05,
      "loss": 6.1876,
      "step": 153
    },
    {
      "epoch": 0.3201663201663202,
      "grad_norm": 25.75374984741211,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 6.075,
      "step": 154
    },
    {
      "epoch": 0.32224532224532226,
      "grad_norm": 29.31098747253418,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 6.1339,
      "step": 155
    },
    {
      "epoch": 0.32432432432432434,
      "grad_norm": 25.086851119995117,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 6.4355,
      "step": 156
    },
    {
      "epoch": 0.3264033264033264,
      "grad_norm": 36.77668762207031,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 6.0702,
      "step": 157
    },
    {
      "epoch": 0.3284823284823285,
      "grad_norm": 26.95594596862793,
      "learning_rate": 1.157994445715706e-05,
      "loss": 5.7807,
      "step": 158
    },
    {
      "epoch": 0.3305613305613306,
      "grad_norm": 21.836240768432617,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 5.9015,
      "step": 159
    },
    {
      "epoch": 0.33264033264033266,
      "grad_norm": 24.77936363220215,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 6.1678,
      "step": 160
    },
    {
      "epoch": 0.33471933471933474,
      "grad_norm": 26.354148864746094,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 5.6249,
      "step": 161
    },
    {
      "epoch": 0.3367983367983368,
      "grad_norm": 33.48122024536133,
      "learning_rate": 9.549150281252633e-06,
      "loss": 5.8313,
      "step": 162
    },
    {
      "epoch": 0.3388773388773389,
      "grad_norm": 28.284072875976562,
      "learning_rate": 9.068759265665384e-06,
      "loss": 5.3997,
      "step": 163
    },
    {
      "epoch": 0.340956340956341,
      "grad_norm": 25.544986724853516,
      "learning_rate": 8.599558442598998e-06,
      "loss": 5.2023,
      "step": 164
    },
    {
      "epoch": 0.34303534303534305,
      "grad_norm": 26.06259536743164,
      "learning_rate": 8.141676086873572e-06,
      "loss": 5.5258,
      "step": 165
    },
    {
      "epoch": 0.34511434511434513,
      "grad_norm": 28.504070281982422,
      "learning_rate": 7.695237378953223e-06,
      "loss": 5.4918,
      "step": 166
    },
    {
      "epoch": 0.3471933471933472,
      "grad_norm": 32.663543701171875,
      "learning_rate": 7.260364370723044e-06,
      "loss": 5.0942,
      "step": 167
    },
    {
      "epoch": 0.3492723492723493,
      "grad_norm": 34.26494216918945,
      "learning_rate": 6.837175952121306e-06,
      "loss": 5.3742,
      "step": 168
    },
    {
      "epoch": 0.35135135135135137,
      "grad_norm": 34.08233642578125,
      "learning_rate": 6.425787818636131e-06,
      "loss": 5.2209,
      "step": 169
    },
    {
      "epoch": 0.35343035343035345,
      "grad_norm": 26.16753578186035,
      "learning_rate": 6.026312439675552e-06,
      "loss": 4.6973,
      "step": 170
    },
    {
      "epoch": 0.35550935550935553,
      "grad_norm": 36.7252311706543,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 5.1574,
      "step": 171
    },
    {
      "epoch": 0.3575883575883576,
      "grad_norm": 28.787479400634766,
      "learning_rate": 5.263533508961827e-06,
      "loss": 4.8892,
      "step": 172
    },
    {
      "epoch": 0.3596673596673597,
      "grad_norm": 27.572086334228516,
      "learning_rate": 4.900438493352055e-06,
      "loss": 4.9546,
      "step": 173
    },
    {
      "epoch": 0.36174636174636177,
      "grad_norm": 26.785364151000977,
      "learning_rate": 4.549673247541875e-06,
      "loss": 4.9544,
      "step": 174
    },
    {
      "epoch": 0.36382536382536385,
      "grad_norm": 27.847000122070312,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 4.8326,
      "step": 175
    },
    {
      "epoch": 0.3659043659043659,
      "grad_norm": 25.241432189941406,
      "learning_rate": 3.885512251130763e-06,
      "loss": 5.6848,
      "step": 176
    },
    {
      "epoch": 0.367983367983368,
      "grad_norm": 28.408544540405273,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 5.2255,
      "step": 177
    },
    {
      "epoch": 0.3700623700623701,
      "grad_norm": 29.53036117553711,
      "learning_rate": 3.271776770026963e-06,
      "loss": 5.1757,
      "step": 178
    },
    {
      "epoch": 0.37214137214137216,
      "grad_norm": 33.13833236694336,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 5.024,
      "step": 179
    },
    {
      "epoch": 0.37422037422037424,
      "grad_norm": 29.38836097717285,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 5.0422,
      "step": 180
    },
    {
      "epoch": 0.3762993762993763,
      "grad_norm": 31.041500091552734,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 5.1083,
      "step": 181
    },
    {
      "epoch": 0.3783783783783784,
      "grad_norm": 26.82939910888672,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 4.9428,
      "step": 182
    },
    {
      "epoch": 0.3804573804573805,
      "grad_norm": 26.33706283569336,
      "learning_rate": 1.962316193157593e-06,
      "loss": 5.2242,
      "step": 183
    },
    {
      "epoch": 0.38253638253638256,
      "grad_norm": 35.151123046875,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 5.1276,
      "step": 184
    },
    {
      "epoch": 0.38461538461538464,
      "grad_norm": 43.028385162353516,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 5.0503,
      "step": 185
    },
    {
      "epoch": 0.3866943866943867,
      "grad_norm": 26.91993522644043,
      "learning_rate": 1.333670137599713e-06,
      "loss": 5.1356,
      "step": 186
    },
    {
      "epoch": 0.3887733887733888,
      "grad_norm": 33.323936462402344,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 5.3732,
      "step": 187
    },
    {
      "epoch": 0.3908523908523909,
      "grad_norm": 37.548187255859375,
      "learning_rate": 9.810017062595322e-07,
      "loss": 4.7597,
      "step": 188
    },
    {
      "epoch": 0.39293139293139295,
      "grad_norm": 29.744958877563477,
      "learning_rate": 8.247462563808817e-07,
      "loss": 5.4321,
      "step": 189
    },
    {
      "epoch": 0.39501039501039503,
      "grad_norm": 32.0604248046875,
      "learning_rate": 6.819348298638839e-07,
      "loss": 5.7061,
      "step": 190
    },
    {
      "epoch": 0.3970893970893971,
      "grad_norm": 27.980018615722656,
      "learning_rate": 5.526064699265753e-07,
      "loss": 5.1849,
      "step": 191
    },
    {
      "epoch": 0.3991683991683992,
      "grad_norm": 40.42841720581055,
      "learning_rate": 4.367965336512403e-07,
      "loss": 5.6513,
      "step": 192
    },
    {
      "epoch": 0.40124740124740127,
      "grad_norm": 31.531082153320312,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 5.7936,
      "step": 193
    },
    {
      "epoch": 0.40332640332640335,
      "grad_norm": 27.996227264404297,
      "learning_rate": 2.458548727494292e-07,
      "loss": 5.9676,
      "step": 194
    },
    {
      "epoch": 0.40540540540540543,
      "grad_norm": 39.96089172363281,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 5.8001,
      "step": 195
    },
    {
      "epoch": 0.4074844074844075,
      "grad_norm": 26.05563735961914,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 6.3064,
      "step": 196
    },
    {
      "epoch": 0.4095634095634096,
      "grad_norm": 36.754146575927734,
      "learning_rate": 6.150154258476315e-08,
      "loss": 6.1782,
      "step": 197
    },
    {
      "epoch": 0.41164241164241167,
      "grad_norm": 29.33064079284668,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 6.247,
      "step": 198
    },
    {
      "epoch": 0.41372141372141374,
      "grad_norm": 36.006195068359375,
      "learning_rate": 6.834750376549792e-09,
      "loss": 6.5253,
      "step": 199
    },
    {
      "epoch": 0.4158004158004158,
      "grad_norm": 31.561588287353516,
      "learning_rate": 0.0,
      "loss": 5.1503,
      "step": 200
    },
    {
      "epoch": 0.4158004158004158,
      "eval_loss": 1.3841888904571533,
      "eval_runtime": 16.5757,
      "eval_samples_per_second": 48.927,
      "eval_steps_per_second": 12.247,
      "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": 2.98048718635008e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}