{
  "best_metric": 1.3963215351104736,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.045493318168893945,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0002274665908444697,
      "grad_norm": 33.85075759887695,
      "learning_rate": 1e-05,
      "loss": 8.9316,
      "step": 1
    },
    {
      "epoch": 0.0002274665908444697,
      "eval_loss": 2.5400564670562744,
      "eval_runtime": 564.9072,
      "eval_samples_per_second": 13.107,
      "eval_steps_per_second": 3.277,
      "step": 1
    },
    {
      "epoch": 0.0004549331816889394,
      "grad_norm": 44.51757049560547,
      "learning_rate": 2e-05,
      "loss": 10.7919,
      "step": 2
    },
    {
      "epoch": 0.0006823997725334092,
      "grad_norm": 32.20150375366211,
      "learning_rate": 3e-05,
      "loss": 7.4136,
      "step": 3
    },
    {
      "epoch": 0.0009098663633778788,
      "grad_norm": 23.51431655883789,
      "learning_rate": 4e-05,
      "loss": 6.958,
      "step": 4
    },
    {
      "epoch": 0.0011373329542223485,
      "grad_norm": 21.65468978881836,
      "learning_rate": 5e-05,
      "loss": 5.6753,
      "step": 5
    },
    {
      "epoch": 0.0013647995450668184,
      "grad_norm": 19.024845123291016,
      "learning_rate": 6e-05,
      "loss": 5.5647,
      "step": 6
    },
    {
      "epoch": 0.001592266135911288,
      "grad_norm": 20.26327133178711,
      "learning_rate": 7e-05,
      "loss": 5.218,
      "step": 7
    },
    {
      "epoch": 0.0018197327267557577,
      "grad_norm": 18.784332275390625,
      "learning_rate": 8e-05,
      "loss": 5.284,
      "step": 8
    },
    {
      "epoch": 0.0020471993176002273,
      "grad_norm": 18.434711456298828,
      "learning_rate": 9e-05,
      "loss": 5.0946,
      "step": 9
    },
    {
      "epoch": 0.002274665908444697,
      "grad_norm": 17.953128814697266,
      "learning_rate": 0.0001,
      "loss": 4.7867,
      "step": 10
    },
    {
      "epoch": 0.002502132499289167,
      "grad_norm": 14.284416198730469,
      "learning_rate": 9.999316524962345e-05,
      "loss": 4.88,
      "step": 11
    },
    {
      "epoch": 0.0027295990901336368,
      "grad_norm": 14.370080947875977,
      "learning_rate": 9.997266286704631e-05,
      "loss": 5.0883,
      "step": 12
    },
    {
      "epoch": 0.0029570656809781064,
      "grad_norm": 14.492742538452148,
      "learning_rate": 9.993849845741524e-05,
      "loss": 5.1898,
      "step": 13
    },
    {
      "epoch": 0.003184532271822576,
      "grad_norm": 15.091177940368652,
      "learning_rate": 9.989068136093873e-05,
      "loss": 4.8526,
      "step": 14
    },
    {
      "epoch": 0.0034119988626670457,
      "grad_norm": 14.381218910217285,
      "learning_rate": 9.98292246503335e-05,
      "loss": 4.8072,
      "step": 15
    },
    {
      "epoch": 0.0036394654535115154,
      "grad_norm": 14.923418998718262,
      "learning_rate": 9.975414512725057e-05,
      "loss": 4.682,
      "step": 16
    },
    {
      "epoch": 0.003866932044355985,
      "grad_norm": 14.486419677734375,
      "learning_rate": 9.966546331768191e-05,
      "loss": 4.6145,
      "step": 17
    },
    {
      "epoch": 0.004094398635200455,
      "grad_norm": 18.26784896850586,
      "learning_rate": 9.956320346634876e-05,
      "loss": 5.047,
      "step": 18
    },
    {
      "epoch": 0.004321865226044924,
      "grad_norm": 16.599550247192383,
      "learning_rate": 9.944739353007344e-05,
      "loss": 5.9347,
      "step": 19
    },
    {
      "epoch": 0.004549331816889394,
      "grad_norm": 17.720232009887695,
      "learning_rate": 9.931806517013612e-05,
      "loss": 5.6102,
      "step": 20
    },
    {
      "epoch": 0.004776798407733864,
      "grad_norm": 14.871539115905762,
      "learning_rate": 9.917525374361912e-05,
      "loss": 5.3959,
      "step": 21
    },
    {
      "epoch": 0.005004264998578334,
      "grad_norm": 14.669853210449219,
      "learning_rate": 9.901899829374047e-05,
      "loss": 5.8429,
      "step": 22
    },
    {
      "epoch": 0.005231731589422804,
      "grad_norm": 15.016830444335938,
      "learning_rate": 9.884934153917997e-05,
      "loss": 5.659,
      "step": 23
    },
    {
      "epoch": 0.0054591981802672735,
      "grad_norm": 15.089632034301758,
      "learning_rate": 9.86663298624003e-05,
      "loss": 6.2289,
      "step": 24
    },
    {
      "epoch": 0.005686664771111743,
      "grad_norm": 13.98339557647705,
      "learning_rate": 9.847001329696653e-05,
      "loss": 5.7122,
      "step": 25
    },
    {
      "epoch": 0.005914131361956213,
      "grad_norm": 14.297969818115234,
      "learning_rate": 9.826044551386744e-05,
      "loss": 5.8346,
      "step": 26
    },
    {
      "epoch": 0.0061415979528006825,
      "grad_norm": 15.060986518859863,
      "learning_rate": 9.803768380684242e-05,
      "loss": 5.8084,
      "step": 27
    },
    {
      "epoch": 0.006369064543645152,
      "grad_norm": 15.347947120666504,
      "learning_rate": 9.780178907671789e-05,
      "loss": 6.8402,
      "step": 28
    },
    {
      "epoch": 0.006596531134489622,
      "grad_norm": 18.604143142700195,
      "learning_rate": 9.755282581475769e-05,
      "loss": 7.1201,
      "step": 29
    },
    {
      "epoch": 0.0068239977253340914,
      "grad_norm": 14.261972427368164,
      "learning_rate": 9.729086208503174e-05,
      "loss": 6.1269,
      "step": 30
    },
    {
      "epoch": 0.007051464316178561,
      "grad_norm": 18.89975357055664,
      "learning_rate": 9.701596950580806e-05,
      "loss": 7.467,
      "step": 31
    },
    {
      "epoch": 0.007278930907023031,
      "grad_norm": 14.834335327148438,
      "learning_rate": 9.672822322997305e-05,
      "loss": 7.11,
      "step": 32
    },
    {
      "epoch": 0.0075063974978675,
      "grad_norm": 15.745987892150879,
      "learning_rate": 9.642770192448536e-05,
      "loss": 7.5026,
      "step": 33
    },
    {
      "epoch": 0.00773386408871197,
      "grad_norm": 15.845263481140137,
      "learning_rate": 9.611448774886924e-05,
      "loss": 8.2636,
      "step": 34
    },
    {
      "epoch": 0.00796133067955644,
      "grad_norm": 18.213821411132812,
      "learning_rate": 9.578866633275288e-05,
      "loss": 8.6304,
      "step": 35
    },
    {
      "epoch": 0.00818879727040091,
      "grad_norm": 15.348206520080566,
      "learning_rate": 9.545032675245813e-05,
      "loss": 8.6848,
      "step": 36
    },
    {
      "epoch": 0.008416263861245379,
      "grad_norm": 16.648630142211914,
      "learning_rate": 9.509956150664796e-05,
      "loss": 7.5992,
      "step": 37
    },
    {
      "epoch": 0.008643730452089849,
      "grad_norm": 17.715679168701172,
      "learning_rate": 9.473646649103818e-05,
      "loss": 7.9322,
      "step": 38
    },
    {
      "epoch": 0.008871197042934318,
      "grad_norm": 15.76171588897705,
      "learning_rate": 9.43611409721806e-05,
      "loss": 8.4198,
      "step": 39
    },
    {
      "epoch": 0.009098663633778788,
      "grad_norm": 16.3723201751709,
      "learning_rate": 9.397368756032445e-05,
      "loss": 9.4072,
      "step": 40
    },
    {
      "epoch": 0.009326130224623258,
      "grad_norm": 23.01369285583496,
      "learning_rate": 9.357421218136386e-05,
      "loss": 8.1971,
      "step": 41
    },
    {
      "epoch": 0.009553596815467727,
      "grad_norm": 19.986509323120117,
      "learning_rate": 9.316282404787871e-05,
      "loss": 9.1774,
      "step": 42
    },
    {
      "epoch": 0.009781063406312199,
      "grad_norm": 19.7462100982666,
      "learning_rate": 9.273963562927695e-05,
      "loss": 8.7603,
      "step": 43
    },
    {
      "epoch": 0.010008529997156668,
      "grad_norm": 25.48119354248047,
      "learning_rate": 9.230476262104677e-05,
      "loss": 9.4801,
      "step": 44
    },
    {
      "epoch": 0.010235996588001138,
      "grad_norm": 18.16246795654297,
      "learning_rate": 9.185832391312644e-05,
      "loss": 7.6469,
      "step": 45
    },
    {
      "epoch": 0.010463463178845608,
      "grad_norm": 19.562881469726562,
      "learning_rate": 9.140044155740101e-05,
      "loss": 9.2923,
      "step": 46
    },
    {
      "epoch": 0.010690929769690077,
      "grad_norm": 21.285625457763672,
      "learning_rate": 9.093124073433463e-05,
      "loss": 9.0842,
      "step": 47
    },
    {
      "epoch": 0.010918396360534547,
      "grad_norm": 26.481782913208008,
      "learning_rate": 9.045084971874738e-05,
      "loss": 10.1887,
      "step": 48
    },
    {
      "epoch": 0.011145862951379017,
      "grad_norm": 31.1817626953125,
      "learning_rate": 8.995939984474624e-05,
      "loss": 10.891,
      "step": 49
    },
    {
      "epoch": 0.011373329542223486,
      "grad_norm": 26.394760131835938,
      "learning_rate": 8.945702546981969e-05,
      "loss": 10.6882,
      "step": 50
    },
    {
      "epoch": 0.011373329542223486,
      "eval_loss": 1.733583927154541,
      "eval_runtime": 570.6179,
      "eval_samples_per_second": 12.975,
      "eval_steps_per_second": 3.244,
      "step": 50
    },
    {
      "epoch": 0.011600796133067956,
      "grad_norm": 22.06222152709961,
      "learning_rate": 8.894386393810563e-05,
      "loss": 6.8888,
      "step": 51
    },
    {
      "epoch": 0.011828262723912426,
      "grad_norm": 22.462017059326172,
      "learning_rate": 8.842005554284296e-05,
      "loss": 7.0561,
      "step": 52
    },
    {
      "epoch": 0.012055729314756895,
      "grad_norm": 22.672849655151367,
      "learning_rate": 8.788574348801675e-05,
      "loss": 6.0168,
      "step": 53
    },
    {
      "epoch": 0.012283195905601365,
      "grad_norm": 13.247769355773926,
      "learning_rate": 8.73410738492077e-05,
      "loss": 5.0592,
      "step": 54
    },
    {
      "epoch": 0.012510662496445835,
      "grad_norm": 11.172239303588867,
      "learning_rate": 8.678619553365659e-05,
      "loss": 4.5667,
      "step": 55
    },
    {
      "epoch": 0.012738129087290304,
      "grad_norm": 15.57034969329834,
      "learning_rate": 8.622126023955446e-05,
      "loss": 5.059,
      "step": 56
    },
    {
      "epoch": 0.012965595678134774,
      "grad_norm": 13.847099304199219,
      "learning_rate": 8.564642241456986e-05,
      "loss": 4.7772,
      "step": 57
    },
    {
      "epoch": 0.013193062268979244,
      "grad_norm": 12.406902313232422,
      "learning_rate": 8.506183921362443e-05,
      "loss": 4.1958,
      "step": 58
    },
    {
      "epoch": 0.013420528859823713,
      "grad_norm": 12.617752075195312,
      "learning_rate": 8.44676704559283e-05,
      "loss": 4.4648,
      "step": 59
    },
    {
      "epoch": 0.013647995450668183,
      "grad_norm": 12.333085060119629,
      "learning_rate": 8.386407858128706e-05,
      "loss": 4.0935,
      "step": 60
    },
    {
      "epoch": 0.013875462041512653,
      "grad_norm": 11.767393112182617,
      "learning_rate": 8.32512286056924e-05,
      "loss": 3.8735,
      "step": 61
    },
    {
      "epoch": 0.014102928632357122,
      "grad_norm": 10.841387748718262,
      "learning_rate": 8.262928807620843e-05,
      "loss": 3.6947,
      "step": 62
    },
    {
      "epoch": 0.014330395223201592,
      "grad_norm": 10.124613761901855,
      "learning_rate": 8.199842702516583e-05,
      "loss": 3.4769,
      "step": 63
    },
    {
      "epoch": 0.014557861814046062,
      "grad_norm": 10.16076374053955,
      "learning_rate": 8.135881792367686e-05,
      "loss": 3.6365,
      "step": 64
    },
    {
      "epoch": 0.014785328404890531,
      "grad_norm": 10.758759498596191,
      "learning_rate": 8.07106356344834e-05,
      "loss": 3.8966,
      "step": 65
    },
    {
      "epoch": 0.015012794995735,
      "grad_norm": 14.040424346923828,
      "learning_rate": 8.005405736415126e-05,
      "loss": 4.5235,
      "step": 66
    },
    {
      "epoch": 0.01524026158657947,
      "grad_norm": 13.657550811767578,
      "learning_rate": 7.938926261462366e-05,
      "loss": 4.7853,
      "step": 67
    },
    {
      "epoch": 0.01546772817742394,
      "grad_norm": 13.6283597946167,
      "learning_rate": 7.871643313414718e-05,
      "loss": 5.7456,
      "step": 68
    },
    {
      "epoch": 0.01569519476826841,
      "grad_norm": 12.536165237426758,
      "learning_rate": 7.803575286758364e-05,
      "loss": 4.564,
      "step": 69
    },
    {
      "epoch": 0.01592266135911288,
      "grad_norm": 15.896062850952148,
      "learning_rate": 7.734740790612136e-05,
      "loss": 6.1405,
      "step": 70
    },
    {
      "epoch": 0.01615012794995735,
      "grad_norm": 11.829387664794922,
      "learning_rate": 7.66515864363997e-05,
      "loss": 4.8092,
      "step": 71
    },
    {
      "epoch": 0.01637759454080182,
      "grad_norm": 12.196739196777344,
      "learning_rate": 7.594847868906076e-05,
      "loss": 5.3058,
      "step": 72
    },
    {
      "epoch": 0.01660506113164629,
      "grad_norm": 11.067307472229004,
      "learning_rate": 7.52382768867422e-05,
      "loss": 4.6056,
      "step": 73
    },
    {
      "epoch": 0.016832527722490758,
      "grad_norm": 12.510245323181152,
      "learning_rate": 7.452117519152542e-05,
      "loss": 6.0812,
      "step": 74
    },
    {
      "epoch": 0.017059994313335228,
      "grad_norm": 14.12991714477539,
      "learning_rate": 7.379736965185368e-05,
      "loss": 6.0048,
      "step": 75
    },
    {
      "epoch": 0.017287460904179697,
      "grad_norm": 14.953682899475098,
      "learning_rate": 7.30670581489344e-05,
      "loss": 7.1154,
      "step": 76
    },
    {
      "epoch": 0.017514927495024167,
      "grad_norm": 13.34103775024414,
      "learning_rate": 7.233044034264034e-05,
      "loss": 6.6286,
      "step": 77
    },
    {
      "epoch": 0.017742394085868637,
      "grad_norm": 12.986444473266602,
      "learning_rate": 7.158771761692464e-05,
      "loss": 6.1797,
      "step": 78
    },
    {
      "epoch": 0.017969860676713106,
      "grad_norm": 14.475767135620117,
      "learning_rate": 7.083909302476453e-05,
      "loss": 6.5806,
      "step": 79
    },
    {
      "epoch": 0.018197327267557576,
      "grad_norm": 13.516197204589844,
      "learning_rate": 7.008477123264848e-05,
      "loss": 6.506,
      "step": 80
    },
    {
      "epoch": 0.018424793858402046,
      "grad_norm": 15.601790428161621,
      "learning_rate": 6.932495846462261e-05,
      "loss": 6.6756,
      "step": 81
    },
    {
      "epoch": 0.018652260449246515,
      "grad_norm": 17.589967727661133,
      "learning_rate": 6.855986244591104e-05,
      "loss": 7.1867,
      "step": 82
    },
    {
      "epoch": 0.018879727040090985,
      "grad_norm": 14.674684524536133,
      "learning_rate": 6.778969234612584e-05,
      "loss": 7.0511,
      "step": 83
    },
    {
      "epoch": 0.019107193630935455,
      "grad_norm": 13.449907302856445,
      "learning_rate": 6.701465872208216e-05,
      "loss": 6.461,
      "step": 84
    },
    {
      "epoch": 0.019334660221779928,
      "grad_norm": 12.952549934387207,
      "learning_rate": 6.623497346023418e-05,
      "loss": 6.5646,
      "step": 85
    },
    {
      "epoch": 0.019562126812624397,
      "grad_norm": 14.450105667114258,
      "learning_rate": 6.545084971874738e-05,
      "loss": 8.1201,
      "step": 86
    },
    {
      "epoch": 0.019789593403468867,
      "grad_norm": 15.296058654785156,
      "learning_rate": 6.466250186922325e-05,
      "loss": 7.2419,
      "step": 87
    },
    {
      "epoch": 0.020017059994313337,
      "grad_norm": 15.565361976623535,
      "learning_rate": 6.387014543809223e-05,
      "loss": 7.9268,
      "step": 88
    },
    {
      "epoch": 0.020244526585157806,
      "grad_norm": 13.930686950683594,
      "learning_rate": 6.307399704769099e-05,
      "loss": 8.3019,
      "step": 89
    },
    {
      "epoch": 0.020471993176002276,
      "grad_norm": 14.909489631652832,
      "learning_rate": 6.227427435703997e-05,
      "loss": 7.9257,
      "step": 90
    },
    {
      "epoch": 0.020699459766846746,
      "grad_norm": 13.126533508300781,
      "learning_rate": 6.147119600233758e-05,
      "loss": 8.0766,
      "step": 91
    },
    {
      "epoch": 0.020926926357691215,
      "grad_norm": 14.795088768005371,
      "learning_rate": 6.066498153718735e-05,
      "loss": 8.0016,
      "step": 92
    },
    {
      "epoch": 0.021154392948535685,
      "grad_norm": 15.316478729248047,
      "learning_rate": 5.985585137257401e-05,
      "loss": 8.5793,
      "step": 93
    },
    {
      "epoch": 0.021381859539380155,
      "grad_norm": 13.869293212890625,
      "learning_rate": 5.90440267166055e-05,
      "loss": 7.9948,
      "step": 94
    },
    {
      "epoch": 0.021609326130224624,
      "grad_norm": 14.535714149475098,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 7.3371,
      "step": 95
    },
    {
      "epoch": 0.021836792721069094,
      "grad_norm": 13.993658065795898,
      "learning_rate": 5.74131823855921e-05,
      "loss": 6.7171,
      "step": 96
    },
    {
      "epoch": 0.022064259311913564,
      "grad_norm": 19.669038772583008,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 8.0931,
      "step": 97
    },
    {
      "epoch": 0.022291725902758033,
      "grad_norm": 18.46149253845215,
      "learning_rate": 5.577423184847932e-05,
      "loss": 7.5802,
      "step": 98
    },
    {
      "epoch": 0.022519192493602503,
      "grad_norm": 18.072834014892578,
      "learning_rate": 5.495227651252315e-05,
      "loss": 8.4263,
      "step": 99
    },
    {
      "epoch": 0.022746659084446973,
      "grad_norm": 28.934932708740234,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 11.156,
      "step": 100
    },
    {
      "epoch": 0.022746659084446973,
      "eval_loss": 1.5281273126602173,
      "eval_runtime": 569.1178,
      "eval_samples_per_second": 13.01,
      "eval_steps_per_second": 3.252,
      "step": 100
    },
    {
      "epoch": 0.022974125675291442,
      "grad_norm": 14.376147270202637,
      "learning_rate": 5.330452921628497e-05,
      "loss": 5.8557,
      "step": 101
    },
    {
      "epoch": 0.023201592266135912,
      "grad_norm": 18.546865463256836,
      "learning_rate": 5.247918773366112e-05,
      "loss": 6.5242,
      "step": 102
    },
    {
      "epoch": 0.02342905885698038,
      "grad_norm": 22.680837631225586,
      "learning_rate": 5.165316846586541e-05,
      "loss": 5.8747,
      "step": 103
    },
    {
      "epoch": 0.02365652544782485,
      "grad_norm": 13.674721717834473,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 4.9718,
      "step": 104
    },
    {
      "epoch": 0.02388399203866932,
      "grad_norm": 10.834354400634766,
      "learning_rate": 5e-05,
      "loss": 4.634,
      "step": 105
    },
    {
      "epoch": 0.02411145862951379,
      "grad_norm": 13.663738250732422,
      "learning_rate": 4.917330276168208e-05,
      "loss": 4.8656,
      "step": 106
    },
    {
      "epoch": 0.02433892522035826,
      "grad_norm": 11.865979194641113,
      "learning_rate": 4.834683153413459e-05,
      "loss": 4.7424,
      "step": 107
    },
    {
      "epoch": 0.02456639181120273,
      "grad_norm": 10.258040428161621,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 3.8621,
      "step": 108
    },
    {
      "epoch": 0.0247938584020472,
      "grad_norm": 9.784005165100098,
      "learning_rate": 4.669547078371504e-05,
      "loss": 3.4615,
      "step": 109
    },
    {
      "epoch": 0.02502132499289167,
      "grad_norm": 11.660787582397461,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 4.375,
      "step": 110
    },
    {
      "epoch": 0.02524879158373614,
      "grad_norm": 9.96675968170166,
      "learning_rate": 4.504772348747687e-05,
      "loss": 3.5918,
      "step": 111
    },
    {
      "epoch": 0.02547625817458061,
      "grad_norm": 10.535229682922363,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 3.9647,
      "step": 112
    },
    {
      "epoch": 0.025703724765425078,
      "grad_norm": 10.414861679077148,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 3.7578,
      "step": 113
    },
    {
      "epoch": 0.025931191356269548,
      "grad_norm": 10.012444496154785,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 3.758,
      "step": 114
    },
    {
      "epoch": 0.026158657947114018,
      "grad_norm": 10.83643627166748,
      "learning_rate": 4.17702704859633e-05,
      "loss": 4.0217,
      "step": 115
    },
    {
      "epoch": 0.026386124537958487,
      "grad_norm": 10.590777397155762,
      "learning_rate": 4.095597328339452e-05,
      "loss": 3.9621,
      "step": 116
    },
    {
      "epoch": 0.026613591128802957,
      "grad_norm": 10.981629371643066,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 3.6573,
      "step": 117
    },
    {
      "epoch": 0.026841057719647426,
      "grad_norm": 10.608504295349121,
      "learning_rate": 3.933501846281267e-05,
      "loss": 3.9342,
      "step": 118
    },
    {
      "epoch": 0.027068524310491896,
      "grad_norm": 10.594474792480469,
      "learning_rate": 3.852880399766243e-05,
      "loss": 4.2988,
      "step": 119
    },
    {
      "epoch": 0.027295990901336366,
      "grad_norm": 14.070845603942871,
      "learning_rate": 3.772572564296005e-05,
      "loss": 4.8005,
      "step": 120
    },
    {
      "epoch": 0.027523457492180835,
      "grad_norm": 11.825883865356445,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 4.7208,
      "step": 121
    },
    {
      "epoch": 0.027750924083025305,
      "grad_norm": 10.942339897155762,
      "learning_rate": 3.612985456190778e-05,
      "loss": 4.7025,
      "step": 122
    },
    {
      "epoch": 0.027978390673869775,
      "grad_norm": 14.623804092407227,
      "learning_rate": 3.533749813077677e-05,
      "loss": 4.4894,
      "step": 123
    },
    {
      "epoch": 0.028205857264714244,
      "grad_norm": 12.45561695098877,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 4.9834,
      "step": 124
    },
    {
      "epoch": 0.028433323855558714,
      "grad_norm": 14.46573257446289,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 6.5757,
      "step": 125
    },
    {
      "epoch": 0.028660790446403184,
      "grad_norm": 10.923254013061523,
      "learning_rate": 3.298534127791785e-05,
      "loss": 5.1281,
      "step": 126
    },
    {
      "epoch": 0.028888257037247653,
      "grad_norm": 13.554349899291992,
      "learning_rate": 3.221030765387417e-05,
      "loss": 6.6561,
      "step": 127
    },
    {
      "epoch": 0.029115723628092123,
      "grad_norm": 12.60084342956543,
      "learning_rate": 3.144013755408895e-05,
      "loss": 6.1441,
      "step": 128
    },
    {
      "epoch": 0.029343190218936593,
      "grad_norm": 12.465630531311035,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 5.3392,
      "step": 129
    },
    {
      "epoch": 0.029570656809781062,
      "grad_norm": 14.023859977722168,
      "learning_rate": 2.991522876735154e-05,
      "loss": 6.109,
      "step": 130
    },
    {
      "epoch": 0.029798123400625532,
      "grad_norm": 12.172150611877441,
      "learning_rate": 2.916090697523549e-05,
      "loss": 6.4841,
      "step": 131
    },
    {
      "epoch": 0.03002558999147,
      "grad_norm": 11.973929405212402,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 6.5025,
      "step": 132
    },
    {
      "epoch": 0.03025305658231447,
      "grad_norm": 12.626691818237305,
      "learning_rate": 2.766955965735968e-05,
      "loss": 5.9404,
      "step": 133
    },
    {
      "epoch": 0.03048052317315894,
      "grad_norm": 12.803211212158203,
      "learning_rate": 2.693294185106562e-05,
      "loss": 6.2931,
      "step": 134
    },
    {
      "epoch": 0.03070798976400341,
      "grad_norm": 14.181547164916992,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 6.6704,
      "step": 135
    },
    {
      "epoch": 0.03093545635484788,
      "grad_norm": 13.545026779174805,
      "learning_rate": 2.547882480847461e-05,
      "loss": 6.7224,
      "step": 136
    },
    {
      "epoch": 0.03116292294569235,
      "grad_norm": 12.032373428344727,
      "learning_rate": 2.476172311325783e-05,
      "loss": 6.25,
      "step": 137
    },
    {
      "epoch": 0.03139038953653682,
      "grad_norm": 13.542147636413574,
      "learning_rate": 2.405152131093926e-05,
      "loss": 7.1351,
      "step": 138
    },
    {
      "epoch": 0.03161785612738129,
      "grad_norm": 14.927430152893066,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 6.9813,
      "step": 139
    },
    {
      "epoch": 0.03184532271822576,
      "grad_norm": 14.990326881408691,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 7.0015,
      "step": 140
    },
    {
      "epoch": 0.03207278930907023,
      "grad_norm": 14.314873695373535,
      "learning_rate": 2.196424713241637e-05,
      "loss": 7.53,
      "step": 141
    },
    {
      "epoch": 0.0323002558999147,
      "grad_norm": 16.397851943969727,
      "learning_rate": 2.128356686585282e-05,
      "loss": 8.4307,
      "step": 142
    },
    {
      "epoch": 0.03252772249075917,
      "grad_norm": 14.191055297851562,
      "learning_rate": 2.061073738537635e-05,
      "loss": 7.5042,
      "step": 143
    },
    {
      "epoch": 0.03275518908160364,
      "grad_norm": 14.179217338562012,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 7.2985,
      "step": 144
    },
    {
      "epoch": 0.03298265567244811,
      "grad_norm": 24.625808715820312,
      "learning_rate": 1.928936436551661e-05,
      "loss": 5.7651,
      "step": 145
    },
    {
      "epoch": 0.03321012226329258,
      "grad_norm": 14.160369873046875,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 7.54,
      "step": 146
    },
    {
      "epoch": 0.03343758885413705,
      "grad_norm": 16.07367515563965,
      "learning_rate": 1.800157297483417e-05,
      "loss": 8.1794,
      "step": 147
    },
    {
      "epoch": 0.033665055444981516,
      "grad_norm": 15.326973915100098,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 7.2608,
      "step": 148
    },
    {
      "epoch": 0.03389252203582599,
      "grad_norm": 17.90393829345703,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 7.4912,
      "step": 149
    },
    {
      "epoch": 0.034119988626670456,
      "grad_norm": 32.38639831542969,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 8.0163,
      "step": 150
    },
    {
      "epoch": 0.034119988626670456,
      "eval_loss": 1.427915096282959,
      "eval_runtime": 570.6727,
      "eval_samples_per_second": 12.974,
      "eval_steps_per_second": 3.244,
      "step": 150
    },
    {
      "epoch": 0.03434745521751493,
      "grad_norm": 12.664813041687012,
      "learning_rate": 1.553232954407171e-05,
      "loss": 5.0047,
      "step": 151
    },
    {
      "epoch": 0.034574921808359395,
      "grad_norm": 17.22288703918457,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 5.5107,
      "step": 152
    },
    {
      "epoch": 0.03480238839920387,
      "grad_norm": 13.586400985717773,
      "learning_rate": 1.435357758543015e-05,
      "loss": 4.5216,
      "step": 153
    },
    {
      "epoch": 0.035029854990048334,
      "grad_norm": 10.968194007873535,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 3.7325,
      "step": 154
    },
    {
      "epoch": 0.03525732158089281,
      "grad_norm": 13.027766227722168,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 4.4968,
      "step": 155
    },
    {
      "epoch": 0.03548478817173727,
      "grad_norm": 10.571267127990723,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 3.9322,
      "step": 156
    },
    {
      "epoch": 0.03571225476258175,
      "grad_norm": 9.579731941223145,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 3.4191,
      "step": 157
    },
    {
      "epoch": 0.03593972135342621,
      "grad_norm": 9.993508338928223,
      "learning_rate": 1.157994445715706e-05,
      "loss": 3.3861,
      "step": 158
    },
    {
      "epoch": 0.036167187944270686,
      "grad_norm": 9.353436470031738,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 3.4702,
      "step": 159
    },
    {
      "epoch": 0.03639465453511515,
      "grad_norm": 10.042057991027832,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 3.6784,
      "step": 160
    },
    {
      "epoch": 0.036622121125959625,
      "grad_norm": 9.229738235473633,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 3.4632,
      "step": 161
    },
    {
      "epoch": 0.03684958771680409,
      "grad_norm": 9.604883193969727,
      "learning_rate": 9.549150281252633e-06,
      "loss": 3.9405,
      "step": 162
    },
    {
      "epoch": 0.037077054307648565,
      "grad_norm": 9.514839172363281,
      "learning_rate": 9.068759265665384e-06,
      "loss": 3.8746,
      "step": 163
    },
    {
      "epoch": 0.03730452089849303,
      "grad_norm": 9.144609451293945,
      "learning_rate": 8.599558442598998e-06,
      "loss": 3.4946,
      "step": 164
    },
    {
      "epoch": 0.037531987489337504,
      "grad_norm": 10.798237800598145,
      "learning_rate": 8.141676086873572e-06,
      "loss": 3.865,
      "step": 165
    },
    {
      "epoch": 0.03775945408018197,
      "grad_norm": 9.650134086608887,
      "learning_rate": 7.695237378953223e-06,
      "loss": 4.176,
      "step": 166
    },
    {
      "epoch": 0.03798692067102644,
      "grad_norm": 9.184988975524902,
      "learning_rate": 7.260364370723044e-06,
      "loss": 3.6171,
      "step": 167
    },
    {
      "epoch": 0.03821438726187091,
      "grad_norm": 10.25433349609375,
      "learning_rate": 6.837175952121306e-06,
      "loss": 3.6888,
      "step": 168
    },
    {
      "epoch": 0.03844185385271538,
      "grad_norm": 10.409635543823242,
      "learning_rate": 6.425787818636131e-06,
      "loss": 4.2133,
      "step": 169
    },
    {
      "epoch": 0.038669320443559856,
      "grad_norm": 10.504165649414062,
      "learning_rate": 6.026312439675552e-06,
      "loss": 4.3117,
      "step": 170
    },
    {
      "epoch": 0.03889678703440432,
      "grad_norm": 10.35623836517334,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 4.8457,
      "step": 171
    },
    {
      "epoch": 0.039124253625248795,
      "grad_norm": 11.593437194824219,
      "learning_rate": 5.263533508961827e-06,
      "loss": 4.8739,
      "step": 172
    },
    {
      "epoch": 0.03935172021609326,
      "grad_norm": 13.0822172164917,
      "learning_rate": 4.900438493352055e-06,
      "loss": 6.0011,
      "step": 173
    },
    {
      "epoch": 0.039579186806937734,
      "grad_norm": 12.846392631530762,
      "learning_rate": 4.549673247541875e-06,
      "loss": 5.8659,
      "step": 174
    },
    {
      "epoch": 0.0398066533977822,
      "grad_norm": 10.912185668945312,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 4.494,
      "step": 175
    },
    {
      "epoch": 0.040034119988626674,
      "grad_norm": 11.526544570922852,
      "learning_rate": 3.885512251130763e-06,
      "loss": 5.6956,
      "step": 176
    },
    {
      "epoch": 0.04026158657947114,
      "grad_norm": 12.195563316345215,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 6.047,
      "step": 177
    },
    {
      "epoch": 0.04048905317031561,
      "grad_norm": 11.480687141418457,
      "learning_rate": 3.271776770026963e-06,
      "loss": 5.52,
      "step": 178
    },
    {
      "epoch": 0.04071651976116008,
      "grad_norm": 13.064004898071289,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 6.3017,
      "step": 179
    },
    {
      "epoch": 0.04094398635200455,
      "grad_norm": 11.711867332458496,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 5.9174,
      "step": 180
    },
    {
      "epoch": 0.04117145294284902,
      "grad_norm": 13.297533988952637,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 6.5242,
      "step": 181
    },
    {
      "epoch": 0.04139891953369349,
      "grad_norm": 14.621213912963867,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 7.245,
      "step": 182
    },
    {
      "epoch": 0.04162638612453796,
      "grad_norm": 12.36281681060791,
      "learning_rate": 1.962316193157593e-06,
      "loss": 6.5108,
      "step": 183
    },
    {
      "epoch": 0.04185385271538243,
      "grad_norm": 12.767642974853516,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 6.9014,
      "step": 184
    },
    {
      "epoch": 0.0420813193062269,
      "grad_norm": 11.755709648132324,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 6.4528,
      "step": 185
    },
    {
      "epoch": 0.04230878589707137,
      "grad_norm": 12.436304092407227,
      "learning_rate": 1.333670137599713e-06,
      "loss": 6.2917,
      "step": 186
    },
    {
      "epoch": 0.042536252487915836,
      "grad_norm": 13.730623245239258,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 7.7719,
      "step": 187
    },
    {
      "epoch": 0.04276371907876031,
      "grad_norm": 15.618932723999023,
      "learning_rate": 9.810017062595322e-07,
      "loss": 7.0378,
      "step": 188
    },
    {
      "epoch": 0.042991185669604776,
      "grad_norm": 13.797956466674805,
      "learning_rate": 8.247462563808817e-07,
      "loss": 6.8731,
      "step": 189
    },
    {
      "epoch": 0.04321865226044925,
      "grad_norm": 14.05947494506836,
      "learning_rate": 6.819348298638839e-07,
      "loss": 7.7444,
      "step": 190
    },
    {
      "epoch": 0.043446118851293715,
      "grad_norm": 12.968502044677734,
      "learning_rate": 5.526064699265753e-07,
      "loss": 7.2223,
      "step": 191
    },
    {
      "epoch": 0.04367358544213819,
      "grad_norm": 14.379146575927734,
      "learning_rate": 4.367965336512403e-07,
      "loss": 7.3876,
      "step": 192
    },
    {
      "epoch": 0.043901052032982654,
      "grad_norm": 14.21398639678955,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 7.6726,
      "step": 193
    },
    {
      "epoch": 0.04412851862382713,
      "grad_norm": 20.439437866210938,
      "learning_rate": 2.458548727494292e-07,
      "loss": 6.0773,
      "step": 194
    },
    {
      "epoch": 0.044355985214671594,
      "grad_norm": 11.196659088134766,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 4.8987,
      "step": 195
    },
    {
      "epoch": 0.04458345180551607,
      "grad_norm": 15.034622192382812,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 8.6126,
      "step": 196
    },
    {
      "epoch": 0.04481091839636053,
      "grad_norm": 14.39323616027832,
      "learning_rate": 6.150154258476315e-08,
      "loss": 6.2352,
      "step": 197
    },
    {
      "epoch": 0.045038384987205006,
      "grad_norm": 15.825337409973145,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 7.2781,
      "step": 198
    },
    {
      "epoch": 0.04526585157804947,
      "grad_norm": 14.904879570007324,
      "learning_rate": 6.834750376549792e-09,
      "loss": 7.2065,
      "step": 199
    },
    {
      "epoch": 0.045493318168893945,
      "grad_norm": 21.38407325744629,
      "learning_rate": 0.0,
      "loss": 9.4298,
      "step": 200
    },
    {
      "epoch": 0.045493318168893945,
      "eval_loss": 1.3963215351104736,
      "eval_runtime": 570.0303,
      "eval_samples_per_second": 12.989,
      "eval_steps_per_second": 3.247,
      "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": 2.995657493352284e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}