{
  "best_metric": 2.3417510986328125,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.8213552361396304,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004106776180698152,
      "grad_norm": 1.0435107946395874,
      "learning_rate": 1e-05,
      "loss": 2.5899,
      "step": 1
    },
    {
      "epoch": 0.004106776180698152,
      "eval_loss": 2.63228440284729,
      "eval_runtime": 44.5972,
      "eval_samples_per_second": 9.193,
      "eval_steps_per_second": 2.31,
      "step": 1
    },
    {
      "epoch": 0.008213552361396304,
      "grad_norm": 1.1768908500671387,
      "learning_rate": 2e-05,
      "loss": 2.6126,
      "step": 2
    },
    {
      "epoch": 0.012320328542094456,
      "grad_norm": 1.1626293659210205,
      "learning_rate": 3e-05,
      "loss": 2.6151,
      "step": 3
    },
    {
      "epoch": 0.01642710472279261,
      "grad_norm": 0.9760255813598633,
      "learning_rate": 4e-05,
      "loss": 2.6293,
      "step": 4
    },
    {
      "epoch": 0.02053388090349076,
      "grad_norm": 0.54732745885849,
      "learning_rate": 5e-05,
      "loss": 2.455,
      "step": 5
    },
    {
      "epoch": 0.024640657084188913,
      "grad_norm": 1.1945987939834595,
      "learning_rate": 6e-05,
      "loss": 2.5308,
      "step": 6
    },
    {
      "epoch": 0.028747433264887063,
      "grad_norm": 1.3102564811706543,
      "learning_rate": 7e-05,
      "loss": 2.4742,
      "step": 7
    },
    {
      "epoch": 0.03285420944558522,
      "grad_norm": 0.7673408389091492,
      "learning_rate": 8e-05,
      "loss": 2.4759,
      "step": 8
    },
    {
      "epoch": 0.03696098562628337,
      "grad_norm": 0.49131035804748535,
      "learning_rate": 9e-05,
      "loss": 2.4432,
      "step": 9
    },
    {
      "epoch": 0.04106776180698152,
      "grad_norm": 0.6802359819412231,
      "learning_rate": 0.0001,
      "loss": 2.4229,
      "step": 10
    },
    {
      "epoch": 0.045174537987679675,
      "grad_norm": 0.5661302804946899,
      "learning_rate": 9.999316524962345e-05,
      "loss": 2.4795,
      "step": 11
    },
    {
      "epoch": 0.049281314168377825,
      "grad_norm": 0.32107627391815186,
      "learning_rate": 9.997266286704631e-05,
      "loss": 2.3257,
      "step": 12
    },
    {
      "epoch": 0.053388090349075976,
      "grad_norm": 0.5493288636207581,
      "learning_rate": 9.993849845741524e-05,
      "loss": 2.4467,
      "step": 13
    },
    {
      "epoch": 0.057494866529774126,
      "grad_norm": 0.48505455255508423,
      "learning_rate": 9.989068136093873e-05,
      "loss": 2.3687,
      "step": 14
    },
    {
      "epoch": 0.061601642710472276,
      "grad_norm": 0.3593801259994507,
      "learning_rate": 9.98292246503335e-05,
      "loss": 2.4058,
      "step": 15
    },
    {
      "epoch": 0.06570841889117043,
      "grad_norm": 0.28538331389427185,
      "learning_rate": 9.975414512725057e-05,
      "loss": 2.4039,
      "step": 16
    },
    {
      "epoch": 0.06981519507186858,
      "grad_norm": 0.3594168424606323,
      "learning_rate": 9.966546331768191e-05,
      "loss": 2.3937,
      "step": 17
    },
    {
      "epoch": 0.07392197125256673,
      "grad_norm": 0.3700706660747528,
      "learning_rate": 9.956320346634876e-05,
      "loss": 2.4102,
      "step": 18
    },
    {
      "epoch": 0.07802874743326489,
      "grad_norm": 0.3427894413471222,
      "learning_rate": 9.944739353007344e-05,
      "loss": 2.4358,
      "step": 19
    },
    {
      "epoch": 0.08213552361396304,
      "grad_norm": 0.3393927812576294,
      "learning_rate": 9.931806517013612e-05,
      "loss": 2.3442,
      "step": 20
    },
    {
      "epoch": 0.08624229979466119,
      "grad_norm": 0.36971014738082886,
      "learning_rate": 9.917525374361912e-05,
      "loss": 2.4027,
      "step": 21
    },
    {
      "epoch": 0.09034907597535935,
      "grad_norm": 0.33891844749450684,
      "learning_rate": 9.901899829374047e-05,
      "loss": 2.417,
      "step": 22
    },
    {
      "epoch": 0.0944558521560575,
      "grad_norm": 0.3384864628314972,
      "learning_rate": 9.884934153917997e-05,
      "loss": 2.3861,
      "step": 23
    },
    {
      "epoch": 0.09856262833675565,
      "grad_norm": 0.31759172677993774,
      "learning_rate": 9.86663298624003e-05,
      "loss": 2.4335,
      "step": 24
    },
    {
      "epoch": 0.1026694045174538,
      "grad_norm": 0.3097175657749176,
      "learning_rate": 9.847001329696653e-05,
      "loss": 2.3641,
      "step": 25
    },
    {
      "epoch": 0.10677618069815195,
      "grad_norm": 0.3116050064563751,
      "learning_rate": 9.826044551386744e-05,
      "loss": 2.3446,
      "step": 26
    },
    {
      "epoch": 0.11088295687885011,
      "grad_norm": 0.3219049870967865,
      "learning_rate": 9.803768380684242e-05,
      "loss": 2.334,
      "step": 27
    },
    {
      "epoch": 0.11498973305954825,
      "grad_norm": 0.31973138451576233,
      "learning_rate": 9.780178907671789e-05,
      "loss": 2.413,
      "step": 28
    },
    {
      "epoch": 0.11909650924024641,
      "grad_norm": 0.3318912982940674,
      "learning_rate": 9.755282581475769e-05,
      "loss": 2.364,
      "step": 29
    },
    {
      "epoch": 0.12320328542094455,
      "grad_norm": 0.3966512382030487,
      "learning_rate": 9.729086208503174e-05,
      "loss": 2.4181,
      "step": 30
    },
    {
      "epoch": 0.1273100616016427,
      "grad_norm": 0.3277226686477661,
      "learning_rate": 9.701596950580806e-05,
      "loss": 2.4132,
      "step": 31
    },
    {
      "epoch": 0.13141683778234087,
      "grad_norm": 0.35295799374580383,
      "learning_rate": 9.672822322997305e-05,
      "loss": 2.3735,
      "step": 32
    },
    {
      "epoch": 0.13552361396303902,
      "grad_norm": 0.34467580914497375,
      "learning_rate": 9.642770192448536e-05,
      "loss": 2.4029,
      "step": 33
    },
    {
      "epoch": 0.13963039014373715,
      "grad_norm": 0.3466915190219879,
      "learning_rate": 9.611448774886924e-05,
      "loss": 2.3832,
      "step": 34
    },
    {
      "epoch": 0.1437371663244353,
      "grad_norm": 0.3206815719604492,
      "learning_rate": 9.578866633275288e-05,
      "loss": 2.3467,
      "step": 35
    },
    {
      "epoch": 0.14784394250513347,
      "grad_norm": 0.3314959704875946,
      "learning_rate": 9.545032675245813e-05,
      "loss": 2.4252,
      "step": 36
    },
    {
      "epoch": 0.15195071868583163,
      "grad_norm": 0.36066117882728577,
      "learning_rate": 9.509956150664796e-05,
      "loss": 2.4813,
      "step": 37
    },
    {
      "epoch": 0.15605749486652978,
      "grad_norm": 0.3375574052333832,
      "learning_rate": 9.473646649103818e-05,
      "loss": 2.3545,
      "step": 38
    },
    {
      "epoch": 0.1601642710472279,
      "grad_norm": 0.3462226986885071,
      "learning_rate": 9.43611409721806e-05,
      "loss": 2.4314,
      "step": 39
    },
    {
      "epoch": 0.16427104722792607,
      "grad_norm": 0.3462670147418976,
      "learning_rate": 9.397368756032445e-05,
      "loss": 2.4575,
      "step": 40
    },
    {
      "epoch": 0.16837782340862423,
      "grad_norm": 0.34911662340164185,
      "learning_rate": 9.357421218136386e-05,
      "loss": 2.4089,
      "step": 41
    },
    {
      "epoch": 0.17248459958932238,
      "grad_norm": 0.3775201141834259,
      "learning_rate": 9.316282404787871e-05,
      "loss": 2.4123,
      "step": 42
    },
    {
      "epoch": 0.17659137577002054,
      "grad_norm": 0.37096771597862244,
      "learning_rate": 9.273963562927695e-05,
      "loss": 2.43,
      "step": 43
    },
    {
      "epoch": 0.1806981519507187,
      "grad_norm": 0.3602532148361206,
      "learning_rate": 9.230476262104677e-05,
      "loss": 2.3749,
      "step": 44
    },
    {
      "epoch": 0.18480492813141683,
      "grad_norm": 0.3737677037715912,
      "learning_rate": 9.185832391312644e-05,
      "loss": 2.48,
      "step": 45
    },
    {
      "epoch": 0.188911704312115,
      "grad_norm": 0.36593466997146606,
      "learning_rate": 9.140044155740101e-05,
      "loss": 2.3724,
      "step": 46
    },
    {
      "epoch": 0.19301848049281314,
      "grad_norm": 0.4230540096759796,
      "learning_rate": 9.093124073433463e-05,
      "loss": 2.3996,
      "step": 47
    },
    {
      "epoch": 0.1971252566735113,
      "grad_norm": 0.39244401454925537,
      "learning_rate": 9.045084971874738e-05,
      "loss": 2.4115,
      "step": 48
    },
    {
      "epoch": 0.20123203285420946,
      "grad_norm": 0.40585052967071533,
      "learning_rate": 8.995939984474624e-05,
      "loss": 2.3191,
      "step": 49
    },
    {
      "epoch": 0.2053388090349076,
      "grad_norm": 0.43886420130729675,
      "learning_rate": 8.945702546981969e-05,
      "loss": 2.379,
      "step": 50
    },
    {
      "epoch": 0.2053388090349076,
      "eval_loss": 2.397188186645508,
      "eval_runtime": 45.3847,
      "eval_samples_per_second": 9.034,
      "eval_steps_per_second": 2.269,
      "step": 50
    },
    {
      "epoch": 0.20944558521560575,
      "grad_norm": 0.5794761180877686,
      "learning_rate": 8.894386393810563e-05,
      "loss": 2.4273,
      "step": 51
    },
    {
      "epoch": 0.2135523613963039,
      "grad_norm": 0.39014923572540283,
      "learning_rate": 8.842005554284296e-05,
      "loss": 2.4078,
      "step": 52
    },
    {
      "epoch": 0.21765913757700206,
      "grad_norm": 0.3453044891357422,
      "learning_rate": 8.788574348801675e-05,
      "loss": 2.4478,
      "step": 53
    },
    {
      "epoch": 0.22176591375770022,
      "grad_norm": 0.3471531867980957,
      "learning_rate": 8.73410738492077e-05,
      "loss": 2.2902,
      "step": 54
    },
    {
      "epoch": 0.22587268993839835,
      "grad_norm": 0.28497612476348877,
      "learning_rate": 8.678619553365659e-05,
      "loss": 2.3663,
      "step": 55
    },
    {
      "epoch": 0.2299794661190965,
      "grad_norm": 0.27953407168388367,
      "learning_rate": 8.622126023955446e-05,
      "loss": 2.4229,
      "step": 56
    },
    {
      "epoch": 0.23408624229979466,
      "grad_norm": 0.2591160237789154,
      "learning_rate": 8.564642241456986e-05,
      "loss": 2.3187,
      "step": 57
    },
    {
      "epoch": 0.23819301848049282,
      "grad_norm": 0.2484808713197708,
      "learning_rate": 8.506183921362443e-05,
      "loss": 2.3382,
      "step": 58
    },
    {
      "epoch": 0.24229979466119098,
      "grad_norm": 0.2553481161594391,
      "learning_rate": 8.44676704559283e-05,
      "loss": 2.3109,
      "step": 59
    },
    {
      "epoch": 0.2464065708418891,
      "grad_norm": 0.26387447118759155,
      "learning_rate": 8.386407858128706e-05,
      "loss": 2.405,
      "step": 60
    },
    {
      "epoch": 0.25051334702258726,
      "grad_norm": 0.24189330637454987,
      "learning_rate": 8.32512286056924e-05,
      "loss": 2.3397,
      "step": 61
    },
    {
      "epoch": 0.2546201232032854,
      "grad_norm": 0.2473788559436798,
      "learning_rate": 8.262928807620843e-05,
      "loss": 2.3868,
      "step": 62
    },
    {
      "epoch": 0.2587268993839836,
      "grad_norm": 0.2503631114959717,
      "learning_rate": 8.199842702516583e-05,
      "loss": 2.2831,
      "step": 63
    },
    {
      "epoch": 0.26283367556468173,
      "grad_norm": 0.24004723131656647,
      "learning_rate": 8.135881792367686e-05,
      "loss": 2.3741,
      "step": 64
    },
    {
      "epoch": 0.2669404517453799,
      "grad_norm": 0.2441152185201645,
      "learning_rate": 8.07106356344834e-05,
      "loss": 2.3254,
      "step": 65
    },
    {
      "epoch": 0.27104722792607805,
      "grad_norm": 0.2660163938999176,
      "learning_rate": 8.005405736415126e-05,
      "loss": 2.2818,
      "step": 66
    },
    {
      "epoch": 0.2751540041067762,
      "grad_norm": 0.254585862159729,
      "learning_rate": 7.938926261462366e-05,
      "loss": 2.2796,
      "step": 67
    },
    {
      "epoch": 0.2792607802874743,
      "grad_norm": 0.2543337941169739,
      "learning_rate": 7.871643313414718e-05,
      "loss": 2.2795,
      "step": 68
    },
    {
      "epoch": 0.28336755646817247,
      "grad_norm": 0.2710717022418976,
      "learning_rate": 7.803575286758364e-05,
      "loss": 2.2723,
      "step": 69
    },
    {
      "epoch": 0.2874743326488706,
      "grad_norm": 0.2980392873287201,
      "learning_rate": 7.734740790612136e-05,
      "loss": 2.3951,
      "step": 70
    },
    {
      "epoch": 0.2915811088295688,
      "grad_norm": 0.2811497449874878,
      "learning_rate": 7.66515864363997e-05,
      "loss": 2.2955,
      "step": 71
    },
    {
      "epoch": 0.29568788501026694,
      "grad_norm": 0.27456995844841003,
      "learning_rate": 7.594847868906076e-05,
      "loss": 2.3536,
      "step": 72
    },
    {
      "epoch": 0.2997946611909651,
      "grad_norm": 0.27245110273361206,
      "learning_rate": 7.52382768867422e-05,
      "loss": 2.3628,
      "step": 73
    },
    {
      "epoch": 0.30390143737166325,
      "grad_norm": 0.28173360228538513,
      "learning_rate": 7.452117519152542e-05,
      "loss": 2.3959,
      "step": 74
    },
    {
      "epoch": 0.3080082135523614,
      "grad_norm": 0.288441002368927,
      "learning_rate": 7.379736965185368e-05,
      "loss": 2.3812,
      "step": 75
    },
    {
      "epoch": 0.31211498973305957,
      "grad_norm": 0.293016254901886,
      "learning_rate": 7.30670581489344e-05,
      "loss": 2.3411,
      "step": 76
    },
    {
      "epoch": 0.3162217659137577,
      "grad_norm": 0.2989495098590851,
      "learning_rate": 7.233044034264034e-05,
      "loss": 2.3633,
      "step": 77
    },
    {
      "epoch": 0.3203285420944558,
      "grad_norm": 0.2963714897632599,
      "learning_rate": 7.158771761692464e-05,
      "loss": 2.3702,
      "step": 78
    },
    {
      "epoch": 0.324435318275154,
      "grad_norm": 0.32292625308036804,
      "learning_rate": 7.083909302476453e-05,
      "loss": 2.3771,
      "step": 79
    },
    {
      "epoch": 0.32854209445585214,
      "grad_norm": 0.3097062408924103,
      "learning_rate": 7.008477123264848e-05,
      "loss": 2.3655,
      "step": 80
    },
    {
      "epoch": 0.3326488706365503,
      "grad_norm": 0.3158957064151764,
      "learning_rate": 6.932495846462261e-05,
      "loss": 2.3183,
      "step": 81
    },
    {
      "epoch": 0.33675564681724846,
      "grad_norm": 0.32411056756973267,
      "learning_rate": 6.855986244591104e-05,
      "loss": 2.3707,
      "step": 82
    },
    {
      "epoch": 0.3408624229979466,
      "grad_norm": 0.32971060276031494,
      "learning_rate": 6.778969234612584e-05,
      "loss": 2.3214,
      "step": 83
    },
    {
      "epoch": 0.34496919917864477,
      "grad_norm": 0.3237762451171875,
      "learning_rate": 6.701465872208216e-05,
      "loss": 2.2978,
      "step": 84
    },
    {
      "epoch": 0.3490759753593429,
      "grad_norm": 0.30338796973228455,
      "learning_rate": 6.623497346023418e-05,
      "loss": 2.3783,
      "step": 85
    },
    {
      "epoch": 0.3531827515400411,
      "grad_norm": 0.33562251925468445,
      "learning_rate": 6.545084971874738e-05,
      "loss": 2.4431,
      "step": 86
    },
    {
      "epoch": 0.35728952772073924,
      "grad_norm": 0.3178534209728241,
      "learning_rate": 6.466250186922325e-05,
      "loss": 2.4342,
      "step": 87
    },
    {
      "epoch": 0.3613963039014374,
      "grad_norm": 0.32157230377197266,
      "learning_rate": 6.387014543809223e-05,
      "loss": 2.2591,
      "step": 88
    },
    {
      "epoch": 0.3655030800821355,
      "grad_norm": 0.343168705701828,
      "learning_rate": 6.307399704769099e-05,
      "loss": 2.335,
      "step": 89
    },
    {
      "epoch": 0.36960985626283366,
      "grad_norm": 0.34791627526283264,
      "learning_rate": 6.227427435703997e-05,
      "loss": 2.3692,
      "step": 90
    },
    {
      "epoch": 0.3737166324435318,
      "grad_norm": 0.3411727249622345,
      "learning_rate": 6.147119600233758e-05,
      "loss": 2.2979,
      "step": 91
    },
    {
      "epoch": 0.37782340862423,
      "grad_norm": 0.3406916856765747,
      "learning_rate": 6.066498153718735e-05,
      "loss": 2.4069,
      "step": 92
    },
    {
      "epoch": 0.38193018480492813,
      "grad_norm": 0.34638410806655884,
      "learning_rate": 5.985585137257401e-05,
      "loss": 2.2895,
      "step": 93
    },
    {
      "epoch": 0.3860369609856263,
      "grad_norm": 0.34439417719841003,
      "learning_rate": 5.90440267166055e-05,
      "loss": 2.3482,
      "step": 94
    },
    {
      "epoch": 0.39014373716632444,
      "grad_norm": 0.3640625774860382,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 2.3401,
      "step": 95
    },
    {
      "epoch": 0.3942505133470226,
      "grad_norm": 0.376729816198349,
      "learning_rate": 5.74131823855921e-05,
      "loss": 2.4102,
      "step": 96
    },
    {
      "epoch": 0.39835728952772076,
      "grad_norm": 0.3897058665752411,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 2.4631,
      "step": 97
    },
    {
      "epoch": 0.4024640657084189,
      "grad_norm": 0.36534807085990906,
      "learning_rate": 5.577423184847932e-05,
      "loss": 2.3228,
      "step": 98
    },
    {
      "epoch": 0.406570841889117,
      "grad_norm": 0.42786386609077454,
      "learning_rate": 5.495227651252315e-05,
      "loss": 2.3745,
      "step": 99
    },
    {
      "epoch": 0.4106776180698152,
      "grad_norm": 0.4445665180683136,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 2.3634,
      "step": 100
    },
    {
      "epoch": 0.4106776180698152,
      "eval_loss": 2.362854480743408,
      "eval_runtime": 45.2342,
      "eval_samples_per_second": 9.064,
      "eval_steps_per_second": 2.277,
      "step": 100
    },
    {
      "epoch": 0.41478439425051333,
      "grad_norm": 0.3759523332118988,
      "learning_rate": 5.330452921628497e-05,
      "loss": 2.4067,
      "step": 101
    },
    {
      "epoch": 0.4188911704312115,
      "grad_norm": 0.3261904716491699,
      "learning_rate": 5.247918773366112e-05,
      "loss": 2.3491,
      "step": 102
    },
    {
      "epoch": 0.42299794661190965,
      "grad_norm": 0.331424742937088,
      "learning_rate": 5.165316846586541e-05,
      "loss": 2.4312,
      "step": 103
    },
    {
      "epoch": 0.4271047227926078,
      "grad_norm": 0.2765870988368988,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 2.2758,
      "step": 104
    },
    {
      "epoch": 0.43121149897330596,
      "grad_norm": 0.2854355573654175,
      "learning_rate": 5e-05,
      "loss": 2.4131,
      "step": 105
    },
    {
      "epoch": 0.4353182751540041,
      "grad_norm": 0.2687312960624695,
      "learning_rate": 4.917330276168208e-05,
      "loss": 2.3662,
      "step": 106
    },
    {
      "epoch": 0.4394250513347023,
      "grad_norm": 0.2524164021015167,
      "learning_rate": 4.834683153413459e-05,
      "loss": 2.324,
      "step": 107
    },
    {
      "epoch": 0.44353182751540043,
      "grad_norm": 0.24188047647476196,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 2.3438,
      "step": 108
    },
    {
      "epoch": 0.44763860369609854,
      "grad_norm": 0.25191670656204224,
      "learning_rate": 4.669547078371504e-05,
      "loss": 2.3429,
      "step": 109
    },
    {
      "epoch": 0.4517453798767967,
      "grad_norm": 0.24199290573596954,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 2.3142,
      "step": 110
    },
    {
      "epoch": 0.45585215605749485,
      "grad_norm": 0.24496878683567047,
      "learning_rate": 4.504772348747687e-05,
      "loss": 2.3662,
      "step": 111
    },
    {
      "epoch": 0.459958932238193,
      "grad_norm": 0.2524467408657074,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 2.3438,
      "step": 112
    },
    {
      "epoch": 0.46406570841889117,
      "grad_norm": 0.2448316067457199,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 2.268,
      "step": 113
    },
    {
      "epoch": 0.4681724845995893,
      "grad_norm": 0.24713213741779327,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 2.3439,
      "step": 114
    },
    {
      "epoch": 0.4722792607802875,
      "grad_norm": 0.24555900692939758,
      "learning_rate": 4.17702704859633e-05,
      "loss": 2.3202,
      "step": 115
    },
    {
      "epoch": 0.47638603696098564,
      "grad_norm": 0.2566995620727539,
      "learning_rate": 4.095597328339452e-05,
      "loss": 2.3097,
      "step": 116
    },
    {
      "epoch": 0.4804928131416838,
      "grad_norm": 0.2605609893798828,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 2.3743,
      "step": 117
    },
    {
      "epoch": 0.48459958932238195,
      "grad_norm": 0.2803257703781128,
      "learning_rate": 3.933501846281267e-05,
      "loss": 2.2706,
      "step": 118
    },
    {
      "epoch": 0.4887063655030801,
      "grad_norm": 0.26998603343963623,
      "learning_rate": 3.852880399766243e-05,
      "loss": 2.3701,
      "step": 119
    },
    {
      "epoch": 0.4928131416837782,
      "grad_norm": 0.2674187421798706,
      "learning_rate": 3.772572564296005e-05,
      "loss": 2.3414,
      "step": 120
    },
    {
      "epoch": 0.49691991786447637,
      "grad_norm": 0.2632690370082855,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 2.3486,
      "step": 121
    },
    {
      "epoch": 0.5010266940451745,
      "grad_norm": 0.2781819701194763,
      "learning_rate": 3.612985456190778e-05,
      "loss": 2.3087,
      "step": 122
    },
    {
      "epoch": 0.5051334702258727,
      "grad_norm": 0.2736590802669525,
      "learning_rate": 3.533749813077677e-05,
      "loss": 2.3172,
      "step": 123
    },
    {
      "epoch": 0.5092402464065708,
      "grad_norm": 0.29256224632263184,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 2.2894,
      "step": 124
    },
    {
      "epoch": 0.5133470225872689,
      "grad_norm": 0.27834874391555786,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 2.3079,
      "step": 125
    },
    {
      "epoch": 0.5174537987679672,
      "grad_norm": 0.2883935868740082,
      "learning_rate": 3.298534127791785e-05,
      "loss": 2.3878,
      "step": 126
    },
    {
      "epoch": 0.5215605749486653,
      "grad_norm": 0.30138325691223145,
      "learning_rate": 3.221030765387417e-05,
      "loss": 2.3325,
      "step": 127
    },
    {
      "epoch": 0.5256673511293635,
      "grad_norm": 0.28910502791404724,
      "learning_rate": 3.144013755408895e-05,
      "loss": 2.3467,
      "step": 128
    },
    {
      "epoch": 0.5297741273100616,
      "grad_norm": 0.27907320857048035,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 2.3272,
      "step": 129
    },
    {
      "epoch": 0.5338809034907598,
      "grad_norm": 0.28163498640060425,
      "learning_rate": 2.991522876735154e-05,
      "loss": 2.3003,
      "step": 130
    },
    {
      "epoch": 0.5379876796714579,
      "grad_norm": 0.2866162061691284,
      "learning_rate": 2.916090697523549e-05,
      "loss": 2.3134,
      "step": 131
    },
    {
      "epoch": 0.5420944558521561,
      "grad_norm": 0.28740113973617554,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 2.3408,
      "step": 132
    },
    {
      "epoch": 0.5462012320328542,
      "grad_norm": 0.32099780440330505,
      "learning_rate": 2.766955965735968e-05,
      "loss": 2.3434,
      "step": 133
    },
    {
      "epoch": 0.5503080082135524,
      "grad_norm": 0.3175574541091919,
      "learning_rate": 2.693294185106562e-05,
      "loss": 2.3162,
      "step": 134
    },
    {
      "epoch": 0.5544147843942505,
      "grad_norm": 0.31221118569374084,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 2.2849,
      "step": 135
    },
    {
      "epoch": 0.5585215605749486,
      "grad_norm": 0.3160749673843384,
      "learning_rate": 2.547882480847461e-05,
      "loss": 2.3527,
      "step": 136
    },
    {
      "epoch": 0.5626283367556468,
      "grad_norm": 0.31787121295928955,
      "learning_rate": 2.476172311325783e-05,
      "loss": 2.3746,
      "step": 137
    },
    {
      "epoch": 0.5667351129363449,
      "grad_norm": 0.32089346647262573,
      "learning_rate": 2.405152131093926e-05,
      "loss": 2.3183,
      "step": 138
    },
    {
      "epoch": 0.5708418891170431,
      "grad_norm": 0.3264188766479492,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 2.4364,
      "step": 139
    },
    {
      "epoch": 0.5749486652977412,
      "grad_norm": 0.32564181089401245,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 2.4302,
      "step": 140
    },
    {
      "epoch": 0.5790554414784395,
      "grad_norm": 0.33926892280578613,
      "learning_rate": 2.196424713241637e-05,
      "loss": 2.3719,
      "step": 141
    },
    {
      "epoch": 0.5831622176591376,
      "grad_norm": 0.34851282835006714,
      "learning_rate": 2.128356686585282e-05,
      "loss": 2.4361,
      "step": 142
    },
    {
      "epoch": 0.5872689938398358,
      "grad_norm": 0.34841641783714294,
      "learning_rate": 2.061073738537635e-05,
      "loss": 2.3556,
      "step": 143
    },
    {
      "epoch": 0.5913757700205339,
      "grad_norm": 0.35201162099838257,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 2.3423,
      "step": 144
    },
    {
      "epoch": 0.5954825462012321,
      "grad_norm": 0.35167983174324036,
      "learning_rate": 1.928936436551661e-05,
      "loss": 2.3508,
      "step": 145
    },
    {
      "epoch": 0.5995893223819302,
      "grad_norm": 0.3647037148475647,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 2.4173,
      "step": 146
    },
    {
      "epoch": 0.6036960985626283,
      "grad_norm": 0.37230825424194336,
      "learning_rate": 1.800157297483417e-05,
      "loss": 2.3651,
      "step": 147
    },
    {
      "epoch": 0.6078028747433265,
      "grad_norm": 0.3980330228805542,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 2.4056,
      "step": 148
    },
    {
      "epoch": 0.6119096509240246,
      "grad_norm": 0.41046053171157837,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 2.3764,
      "step": 149
    },
    {
      "epoch": 0.6160164271047228,
      "grad_norm": 0.4375952482223511,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 2.3456,
      "step": 150
    },
    {
      "epoch": 0.6160164271047228,
      "eval_loss": 2.344496011734009,
      "eval_runtime": 45.479,
      "eval_samples_per_second": 9.015,
      "eval_steps_per_second": 2.265,
      "step": 150
    },
    {
      "epoch": 0.6201232032854209,
      "grad_norm": 0.2772764265537262,
      "learning_rate": 1.553232954407171e-05,
      "loss": 2.3712,
      "step": 151
    },
    {
      "epoch": 0.6242299794661191,
      "grad_norm": 0.3079642653465271,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 2.4293,
      "step": 152
    },
    {
      "epoch": 0.6283367556468172,
      "grad_norm": 0.3389173150062561,
      "learning_rate": 1.435357758543015e-05,
      "loss": 2.3476,
      "step": 153
    },
    {
      "epoch": 0.6324435318275154,
      "grad_norm": 0.2954648435115814,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 2.3365,
      "step": 154
    },
    {
      "epoch": 0.6365503080082136,
      "grad_norm": 0.28537774085998535,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 2.3594,
      "step": 155
    },
    {
      "epoch": 0.6406570841889117,
      "grad_norm": 0.2851802408695221,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 2.3352,
      "step": 156
    },
    {
      "epoch": 0.6447638603696099,
      "grad_norm": 0.2677701711654663,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 2.3096,
      "step": 157
    },
    {
      "epoch": 0.648870636550308,
      "grad_norm": 0.2570532560348511,
      "learning_rate": 1.157994445715706e-05,
      "loss": 2.3084,
      "step": 158
    },
    {
      "epoch": 0.6529774127310062,
      "grad_norm": 0.26534196734428406,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 2.3495,
      "step": 159
    },
    {
      "epoch": 0.6570841889117043,
      "grad_norm": 0.24905315041542053,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 2.2499,
      "step": 160
    },
    {
      "epoch": 0.6611909650924025,
      "grad_norm": 0.25606027245521545,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 2.3516,
      "step": 161
    },
    {
      "epoch": 0.6652977412731006,
      "grad_norm": 0.24965108931064606,
      "learning_rate": 9.549150281252633e-06,
      "loss": 2.3206,
      "step": 162
    },
    {
      "epoch": 0.6694045174537988,
      "grad_norm": 0.2460511177778244,
      "learning_rate": 9.068759265665384e-06,
      "loss": 2.3313,
      "step": 163
    },
    {
      "epoch": 0.6735112936344969,
      "grad_norm": 0.24727021157741547,
      "learning_rate": 8.599558442598998e-06,
      "loss": 2.2926,
      "step": 164
    },
    {
      "epoch": 0.6776180698151951,
      "grad_norm": 0.250802218914032,
      "learning_rate": 8.141676086873572e-06,
      "loss": 2.2717,
      "step": 165
    },
    {
      "epoch": 0.6817248459958932,
      "grad_norm": 0.26117438077926636,
      "learning_rate": 7.695237378953223e-06,
      "loss": 2.2706,
      "step": 166
    },
    {
      "epoch": 0.6858316221765913,
      "grad_norm": 0.2591541111469269,
      "learning_rate": 7.260364370723044e-06,
      "loss": 2.3383,
      "step": 167
    },
    {
      "epoch": 0.6899383983572895,
      "grad_norm": 0.26453521847724915,
      "learning_rate": 6.837175952121306e-06,
      "loss": 2.2882,
      "step": 168
    },
    {
      "epoch": 0.6940451745379876,
      "grad_norm": 0.25713303685188293,
      "learning_rate": 6.425787818636131e-06,
      "loss": 2.3399,
      "step": 169
    },
    {
      "epoch": 0.6981519507186859,
      "grad_norm": 0.2757847309112549,
      "learning_rate": 6.026312439675552e-06,
      "loss": 2.2972,
      "step": 170
    },
    {
      "epoch": 0.702258726899384,
      "grad_norm": 0.27346155047416687,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 2.3254,
      "step": 171
    },
    {
      "epoch": 0.7063655030800822,
      "grad_norm": 0.2759503424167633,
      "learning_rate": 5.263533508961827e-06,
      "loss": 2.3119,
      "step": 172
    },
    {
      "epoch": 0.7104722792607803,
      "grad_norm": 0.28355729579925537,
      "learning_rate": 4.900438493352055e-06,
      "loss": 2.2971,
      "step": 173
    },
    {
      "epoch": 0.7145790554414785,
      "grad_norm": 0.2665920853614807,
      "learning_rate": 4.549673247541875e-06,
      "loss": 2.3807,
      "step": 174
    },
    {
      "epoch": 0.7186858316221766,
      "grad_norm": 0.28380122780799866,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 2.2972,
      "step": 175
    },
    {
      "epoch": 0.7227926078028748,
      "grad_norm": 0.30623114109039307,
      "learning_rate": 3.885512251130763e-06,
      "loss": 2.2766,
      "step": 176
    },
    {
      "epoch": 0.7268993839835729,
      "grad_norm": 0.29208096861839294,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 2.2743,
      "step": 177
    },
    {
      "epoch": 0.731006160164271,
      "grad_norm": 0.2811971604824066,
      "learning_rate": 3.271776770026963e-06,
      "loss": 2.2922,
      "step": 178
    },
    {
      "epoch": 0.7351129363449692,
      "grad_norm": 0.29797789454460144,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 2.3994,
      "step": 179
    },
    {
      "epoch": 0.7392197125256673,
      "grad_norm": 0.3097155690193176,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 2.2718,
      "step": 180
    },
    {
      "epoch": 0.7433264887063655,
      "grad_norm": 0.3049456477165222,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 2.3288,
      "step": 181
    },
    {
      "epoch": 0.7474332648870636,
      "grad_norm": 0.3047029972076416,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 2.3522,
      "step": 182
    },
    {
      "epoch": 0.7515400410677618,
      "grad_norm": 0.30717289447784424,
      "learning_rate": 1.962316193157593e-06,
      "loss": 2.3686,
      "step": 183
    },
    {
      "epoch": 0.75564681724846,
      "grad_norm": 0.32809892296791077,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 2.2879,
      "step": 184
    },
    {
      "epoch": 0.7597535934291582,
      "grad_norm": 0.32817089557647705,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 2.3531,
      "step": 185
    },
    {
      "epoch": 0.7638603696098563,
      "grad_norm": 0.3201553225517273,
      "learning_rate": 1.333670137599713e-06,
      "loss": 2.3368,
      "step": 186
    },
    {
      "epoch": 0.7679671457905544,
      "grad_norm": 0.3280608355998993,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 2.3651,
      "step": 187
    },
    {
      "epoch": 0.7720739219712526,
      "grad_norm": 0.3260633945465088,
      "learning_rate": 9.810017062595322e-07,
      "loss": 2.3264,
      "step": 188
    },
    {
      "epoch": 0.7761806981519507,
      "grad_norm": 0.33519792556762695,
      "learning_rate": 8.247462563808817e-07,
      "loss": 2.2996,
      "step": 189
    },
    {
      "epoch": 0.7802874743326489,
      "grad_norm": 0.32681143283843994,
      "learning_rate": 6.819348298638839e-07,
      "loss": 2.3973,
      "step": 190
    },
    {
      "epoch": 0.784394250513347,
      "grad_norm": 0.3333568572998047,
      "learning_rate": 5.526064699265753e-07,
      "loss": 2.2841,
      "step": 191
    },
    {
      "epoch": 0.7885010266940452,
      "grad_norm": 0.34127864241600037,
      "learning_rate": 4.367965336512403e-07,
      "loss": 2.3543,
      "step": 192
    },
    {
      "epoch": 0.7926078028747433,
      "grad_norm": 0.35863664746284485,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 2.3654,
      "step": 193
    },
    {
      "epoch": 0.7967145790554415,
      "grad_norm": 0.3504040837287903,
      "learning_rate": 2.458548727494292e-07,
      "loss": 2.2604,
      "step": 194
    },
    {
      "epoch": 0.8008213552361396,
      "grad_norm": 0.3843472898006439,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 2.4066,
      "step": 195
    },
    {
      "epoch": 0.8049281314168378,
      "grad_norm": 0.3833732008934021,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 2.3927,
      "step": 196
    },
    {
      "epoch": 0.8090349075975359,
      "grad_norm": 0.39364683628082275,
      "learning_rate": 6.150154258476315e-08,
      "loss": 2.3541,
      "step": 197
    },
    {
      "epoch": 0.813141683778234,
      "grad_norm": 0.4244134724140167,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 2.3894,
      "step": 198
    },
    {
      "epoch": 0.8172484599589322,
      "grad_norm": 0.4756809175014496,
      "learning_rate": 6.834750376549792e-09,
      "loss": 2.4361,
      "step": 199
    },
    {
      "epoch": 0.8213552361396304,
      "grad_norm": 0.5187717080116272,
      "learning_rate": 0.0,
      "loss": 2.3908,
      "step": 200
    },
    {
      "epoch": 0.8213552361396304,
      "eval_loss": 2.3417510986328125,
      "eval_runtime": 45.1998,
      "eval_samples_per_second": 9.071,
      "eval_steps_per_second": 2.279,
      "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": 4.2577795760101786e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}