{
  "best_metric": 1.2995595932006836,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.032416224320272295,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00016208112160136148,
      "grad_norm": 21.450843811035156,
      "learning_rate": 1e-05,
      "loss": 9.9546,
      "step": 1
    },
    {
      "epoch": 0.00016208112160136148,
      "eval_loss": 2.8611927032470703,
      "eval_runtime": 1024.5411,
      "eval_samples_per_second": 10.143,
      "eval_steps_per_second": 2.536,
      "step": 1
    },
    {
      "epoch": 0.00032416224320272295,
      "grad_norm": 20.973268508911133,
      "learning_rate": 2e-05,
      "loss": 10.6012,
      "step": 2
    },
    {
      "epoch": 0.00048624336480408443,
      "grad_norm": 18.114707946777344,
      "learning_rate": 3e-05,
      "loss": 10.833,
      "step": 3
    },
    {
      "epoch": 0.0006483244864054459,
      "grad_norm": 12.959110260009766,
      "learning_rate": 4e-05,
      "loss": 10.37,
      "step": 4
    },
    {
      "epoch": 0.0008104056080068074,
      "grad_norm": 14.474491119384766,
      "learning_rate": 5e-05,
      "loss": 9.5092,
      "step": 5
    },
    {
      "epoch": 0.0009724867296081689,
      "grad_norm": 14.347698211669922,
      "learning_rate": 6e-05,
      "loss": 8.3771,
      "step": 6
    },
    {
      "epoch": 0.0011345678512095303,
      "grad_norm": 11.730208396911621,
      "learning_rate": 7e-05,
      "loss": 8.6556,
      "step": 7
    },
    {
      "epoch": 0.0012966489728108918,
      "grad_norm": 10.065325736999512,
      "learning_rate": 8e-05,
      "loss": 8.7092,
      "step": 8
    },
    {
      "epoch": 0.0014587300944122533,
      "grad_norm": 8.963871955871582,
      "learning_rate": 9e-05,
      "loss": 8.0248,
      "step": 9
    },
    {
      "epoch": 0.0016208112160136148,
      "grad_norm": 9.340985298156738,
      "learning_rate": 0.0001,
      "loss": 7.8928,
      "step": 10
    },
    {
      "epoch": 0.0017828923376149762,
      "grad_norm": 7.826351165771484,
      "learning_rate": 9.999316524962345e-05,
      "loss": 7.7345,
      "step": 11
    },
    {
      "epoch": 0.0019449734592163377,
      "grad_norm": 7.2508697509765625,
      "learning_rate": 9.997266286704631e-05,
      "loss": 7.6428,
      "step": 12
    },
    {
      "epoch": 0.002107054580817699,
      "grad_norm": 6.509117603302002,
      "learning_rate": 9.993849845741524e-05,
      "loss": 6.6641,
      "step": 13
    },
    {
      "epoch": 0.0022691357024190607,
      "grad_norm": 7.647972106933594,
      "learning_rate": 9.989068136093873e-05,
      "loss": 7.2049,
      "step": 14
    },
    {
      "epoch": 0.002431216824020422,
      "grad_norm": 6.6707000732421875,
      "learning_rate": 9.98292246503335e-05,
      "loss": 6.8951,
      "step": 15
    },
    {
      "epoch": 0.0025932979456217836,
      "grad_norm": 6.7753095626831055,
      "learning_rate": 9.975414512725057e-05,
      "loss": 7.1702,
      "step": 16
    },
    {
      "epoch": 0.002755379067223145,
      "grad_norm": 5.891744136810303,
      "learning_rate": 9.966546331768191e-05,
      "loss": 6.3936,
      "step": 17
    },
    {
      "epoch": 0.0029174601888245066,
      "grad_norm": 6.441070556640625,
      "learning_rate": 9.956320346634876e-05,
      "loss": 6.748,
      "step": 18
    },
    {
      "epoch": 0.003079541310425868,
      "grad_norm": 5.625812530517578,
      "learning_rate": 9.944739353007344e-05,
      "loss": 6.2807,
      "step": 19
    },
    {
      "epoch": 0.0032416224320272295,
      "grad_norm": 5.964963436126709,
      "learning_rate": 9.931806517013612e-05,
      "loss": 6.8812,
      "step": 20
    },
    {
      "epoch": 0.003403703553628591,
      "grad_norm": 6.139531135559082,
      "learning_rate": 9.917525374361912e-05,
      "loss": 6.2387,
      "step": 21
    },
    {
      "epoch": 0.0035657846752299525,
      "grad_norm": 5.533466815948486,
      "learning_rate": 9.901899829374047e-05,
      "loss": 6.1352,
      "step": 22
    },
    {
      "epoch": 0.003727865796831314,
      "grad_norm": 5.690276622772217,
      "learning_rate": 9.884934153917997e-05,
      "loss": 6.4498,
      "step": 23
    },
    {
      "epoch": 0.0038899469184326754,
      "grad_norm": 5.708102703094482,
      "learning_rate": 9.86663298624003e-05,
      "loss": 6.3784,
      "step": 24
    },
    {
      "epoch": 0.004052028040034037,
      "grad_norm": 5.579735279083252,
      "learning_rate": 9.847001329696653e-05,
      "loss": 6.3002,
      "step": 25
    },
    {
      "epoch": 0.004214109161635398,
      "grad_norm": 5.780838966369629,
      "learning_rate": 9.826044551386744e-05,
      "loss": 5.8777,
      "step": 26
    },
    {
      "epoch": 0.00437619028323676,
      "grad_norm": 5.95778226852417,
      "learning_rate": 9.803768380684242e-05,
      "loss": 6.3759,
      "step": 27
    },
    {
      "epoch": 0.004538271404838121,
      "grad_norm": 5.736791133880615,
      "learning_rate": 9.780178907671789e-05,
      "loss": 6.1456,
      "step": 28
    },
    {
      "epoch": 0.004700352526439483,
      "grad_norm": 5.662046909332275,
      "learning_rate": 9.755282581475769e-05,
      "loss": 6.1055,
      "step": 29
    },
    {
      "epoch": 0.004862433648040844,
      "grad_norm": 5.7258806228637695,
      "learning_rate": 9.729086208503174e-05,
      "loss": 6.2523,
      "step": 30
    },
    {
      "epoch": 0.005024514769642206,
      "grad_norm": 5.707207202911377,
      "learning_rate": 9.701596950580806e-05,
      "loss": 5.7521,
      "step": 31
    },
    {
      "epoch": 0.005186595891243567,
      "grad_norm": 5.748448371887207,
      "learning_rate": 9.672822322997305e-05,
      "loss": 6.0935,
      "step": 32
    },
    {
      "epoch": 0.005348677012844929,
      "grad_norm": 5.671350479125977,
      "learning_rate": 9.642770192448536e-05,
      "loss": 6.2013,
      "step": 33
    },
    {
      "epoch": 0.00551075813444629,
      "grad_norm": 6.076589584350586,
      "learning_rate": 9.611448774886924e-05,
      "loss": 6.2627,
      "step": 34
    },
    {
      "epoch": 0.005672839256047652,
      "grad_norm": 6.076786041259766,
      "learning_rate": 9.578866633275288e-05,
      "loss": 6.2,
      "step": 35
    },
    {
      "epoch": 0.005834920377649013,
      "grad_norm": 6.442121982574463,
      "learning_rate": 9.545032675245813e-05,
      "loss": 6.097,
      "step": 36
    },
    {
      "epoch": 0.005997001499250375,
      "grad_norm": 6.375777244567871,
      "learning_rate": 9.509956150664796e-05,
      "loss": 6.3335,
      "step": 37
    },
    {
      "epoch": 0.006159082620851736,
      "grad_norm": 6.780538558959961,
      "learning_rate": 9.473646649103818e-05,
      "loss": 6.4989,
      "step": 38
    },
    {
      "epoch": 0.006321163742453098,
      "grad_norm": 6.880222797393799,
      "learning_rate": 9.43611409721806e-05,
      "loss": 6.6263,
      "step": 39
    },
    {
      "epoch": 0.006483244864054459,
      "grad_norm": 6.89586877822876,
      "learning_rate": 9.397368756032445e-05,
      "loss": 6.6472,
      "step": 40
    },
    {
      "epoch": 0.0066453259856558205,
      "grad_norm": 7.324431896209717,
      "learning_rate": 9.357421218136386e-05,
      "loss": 6.9357,
      "step": 41
    },
    {
      "epoch": 0.006807407107257182,
      "grad_norm": 7.074251174926758,
      "learning_rate": 9.316282404787871e-05,
      "loss": 6.4737,
      "step": 42
    },
    {
      "epoch": 0.0069694882288585435,
      "grad_norm": 7.916954040527344,
      "learning_rate": 9.273963562927695e-05,
      "loss": 6.9543,
      "step": 43
    },
    {
      "epoch": 0.007131569350459905,
      "grad_norm": 8.427472114562988,
      "learning_rate": 9.230476262104677e-05,
      "loss": 6.6134,
      "step": 44
    },
    {
      "epoch": 0.007293650472061266,
      "grad_norm": 8.175810813903809,
      "learning_rate": 9.185832391312644e-05,
      "loss": 6.5496,
      "step": 45
    },
    {
      "epoch": 0.007455731593662628,
      "grad_norm": 8.738851547241211,
      "learning_rate": 9.140044155740101e-05,
      "loss": 6.8652,
      "step": 46
    },
    {
      "epoch": 0.007617812715263989,
      "grad_norm": 9.828368186950684,
      "learning_rate": 9.093124073433463e-05,
      "loss": 6.7311,
      "step": 47
    },
    {
      "epoch": 0.007779893836865351,
      "grad_norm": 10.030229568481445,
      "learning_rate": 9.045084971874738e-05,
      "loss": 6.7401,
      "step": 48
    },
    {
      "epoch": 0.007941974958466713,
      "grad_norm": 12.767650604248047,
      "learning_rate": 8.995939984474624e-05,
      "loss": 7.082,
      "step": 49
    },
    {
      "epoch": 0.008104056080068074,
      "grad_norm": 14.701695442199707,
      "learning_rate": 8.945702546981969e-05,
      "loss": 5.9265,
      "step": 50
    },
    {
      "epoch": 0.008104056080068074,
      "eval_loss": 1.5559238195419312,
      "eval_runtime": 1031.1317,
      "eval_samples_per_second": 10.078,
      "eval_steps_per_second": 2.52,
      "step": 50
    },
    {
      "epoch": 0.008266137201669436,
      "grad_norm": 8.963892936706543,
      "learning_rate": 8.894386393810563e-05,
      "loss": 6.6975,
      "step": 51
    },
    {
      "epoch": 0.008428218323270797,
      "grad_norm": 5.794774055480957,
      "learning_rate": 8.842005554284296e-05,
      "loss": 6.5483,
      "step": 52
    },
    {
      "epoch": 0.008590299444872159,
      "grad_norm": 4.353773593902588,
      "learning_rate": 8.788574348801675e-05,
      "loss": 6.2493,
      "step": 53
    },
    {
      "epoch": 0.00875238056647352,
      "grad_norm": 4.177701950073242,
      "learning_rate": 8.73410738492077e-05,
      "loss": 5.725,
      "step": 54
    },
    {
      "epoch": 0.008914461688074882,
      "grad_norm": 3.7573206424713135,
      "learning_rate": 8.678619553365659e-05,
      "loss": 5.5812,
      "step": 55
    },
    {
      "epoch": 0.009076542809676243,
      "grad_norm": 4.3062591552734375,
      "learning_rate": 8.622126023955446e-05,
      "loss": 5.6443,
      "step": 56
    },
    {
      "epoch": 0.009238623931277605,
      "grad_norm": 3.8506991863250732,
      "learning_rate": 8.564642241456986e-05,
      "loss": 5.6099,
      "step": 57
    },
    {
      "epoch": 0.009400705052878966,
      "grad_norm": 3.8697073459625244,
      "learning_rate": 8.506183921362443e-05,
      "loss": 5.0964,
      "step": 58
    },
    {
      "epoch": 0.009562786174480328,
      "grad_norm": 3.971761465072632,
      "learning_rate": 8.44676704559283e-05,
      "loss": 5.5613,
      "step": 59
    },
    {
      "epoch": 0.009724867296081689,
      "grad_norm": 3.820019245147705,
      "learning_rate": 8.386407858128706e-05,
      "loss": 5.6731,
      "step": 60
    },
    {
      "epoch": 0.009886948417683051,
      "grad_norm": 4.112672805786133,
      "learning_rate": 8.32512286056924e-05,
      "loss": 5.4865,
      "step": 61
    },
    {
      "epoch": 0.010049029539284412,
      "grad_norm": 3.847612142562866,
      "learning_rate": 8.262928807620843e-05,
      "loss": 5.548,
      "step": 62
    },
    {
      "epoch": 0.010211110660885774,
      "grad_norm": 3.959521770477295,
      "learning_rate": 8.199842702516583e-05,
      "loss": 5.6983,
      "step": 63
    },
    {
      "epoch": 0.010373191782487134,
      "grad_norm": 3.779252052307129,
      "learning_rate": 8.135881792367686e-05,
      "loss": 5.1694,
      "step": 64
    },
    {
      "epoch": 0.010535272904088497,
      "grad_norm": 3.9358041286468506,
      "learning_rate": 8.07106356344834e-05,
      "loss": 5.6322,
      "step": 65
    },
    {
      "epoch": 0.010697354025689857,
      "grad_norm": 3.8914718627929688,
      "learning_rate": 8.005405736415126e-05,
      "loss": 4.9709,
      "step": 66
    },
    {
      "epoch": 0.01085943514729122,
      "grad_norm": 4.115433216094971,
      "learning_rate": 7.938926261462366e-05,
      "loss": 5.2434,
      "step": 67
    },
    {
      "epoch": 0.01102151626889258,
      "grad_norm": 4.220749855041504,
      "learning_rate": 7.871643313414718e-05,
      "loss": 5.4555,
      "step": 68
    },
    {
      "epoch": 0.011183597390493943,
      "grad_norm": 4.126224040985107,
      "learning_rate": 7.803575286758364e-05,
      "loss": 5.3821,
      "step": 69
    },
    {
      "epoch": 0.011345678512095303,
      "grad_norm": 4.412211894989014,
      "learning_rate": 7.734740790612136e-05,
      "loss": 5.6406,
      "step": 70
    },
    {
      "epoch": 0.011507759633696666,
      "grad_norm": 3.9712724685668945,
      "learning_rate": 7.66515864363997e-05,
      "loss": 5.4385,
      "step": 71
    },
    {
      "epoch": 0.011669840755298026,
      "grad_norm": 4.128697395324707,
      "learning_rate": 7.594847868906076e-05,
      "loss": 5.7166,
      "step": 72
    },
    {
      "epoch": 0.011831921876899389,
      "grad_norm": 4.208035945892334,
      "learning_rate": 7.52382768867422e-05,
      "loss": 5.5222,
      "step": 73
    },
    {
      "epoch": 0.01199400299850075,
      "grad_norm": 4.283714294433594,
      "learning_rate": 7.452117519152542e-05,
      "loss": 5.4174,
      "step": 74
    },
    {
      "epoch": 0.012156084120102112,
      "grad_norm": 4.24643087387085,
      "learning_rate": 7.379736965185368e-05,
      "loss": 5.5042,
      "step": 75
    },
    {
      "epoch": 0.012318165241703472,
      "grad_norm": 4.205446720123291,
      "learning_rate": 7.30670581489344e-05,
      "loss": 5.0924,
      "step": 76
    },
    {
      "epoch": 0.012480246363304835,
      "grad_norm": 4.379979133605957,
      "learning_rate": 7.233044034264034e-05,
      "loss": 5.3984,
      "step": 77
    },
    {
      "epoch": 0.012642327484906195,
      "grad_norm": 4.69776725769043,
      "learning_rate": 7.158771761692464e-05,
      "loss": 5.6712,
      "step": 78
    },
    {
      "epoch": 0.012804408606507557,
      "grad_norm": 4.539688587188721,
      "learning_rate": 7.083909302476453e-05,
      "loss": 5.2928,
      "step": 79
    },
    {
      "epoch": 0.012966489728108918,
      "grad_norm": 4.4620490074157715,
      "learning_rate": 7.008477123264848e-05,
      "loss": 5.2381,
      "step": 80
    },
    {
      "epoch": 0.01312857084971028,
      "grad_norm": 4.704196929931641,
      "learning_rate": 6.932495846462261e-05,
      "loss": 5.2825,
      "step": 81
    },
    {
      "epoch": 0.013290651971311641,
      "grad_norm": 4.632774829864502,
      "learning_rate": 6.855986244591104e-05,
      "loss": 5.4664,
      "step": 82
    },
    {
      "epoch": 0.013452733092913003,
      "grad_norm": 4.884334087371826,
      "learning_rate": 6.778969234612584e-05,
      "loss": 5.3945,
      "step": 83
    },
    {
      "epoch": 0.013614814214514364,
      "grad_norm": 4.739119529724121,
      "learning_rate": 6.701465872208216e-05,
      "loss": 5.7381,
      "step": 84
    },
    {
      "epoch": 0.013776895336115726,
      "grad_norm": 5.047530174255371,
      "learning_rate": 6.623497346023418e-05,
      "loss": 5.7925,
      "step": 85
    },
    {
      "epoch": 0.013938976457717087,
      "grad_norm": 5.00535249710083,
      "learning_rate": 6.545084971874738e-05,
      "loss": 5.4451,
      "step": 86
    },
    {
      "epoch": 0.01410105757931845,
      "grad_norm": 5.174424171447754,
      "learning_rate": 6.466250186922325e-05,
      "loss": 5.6002,
      "step": 87
    },
    {
      "epoch": 0.01426313870091981,
      "grad_norm": 5.37460470199585,
      "learning_rate": 6.387014543809223e-05,
      "loss": 5.7764,
      "step": 88
    },
    {
      "epoch": 0.014425219822521172,
      "grad_norm": 5.565922260284424,
      "learning_rate": 6.307399704769099e-05,
      "loss": 6.0238,
      "step": 89
    },
    {
      "epoch": 0.014587300944122533,
      "grad_norm": 5.54338264465332,
      "learning_rate": 6.227427435703997e-05,
      "loss": 5.8409,
      "step": 90
    },
    {
      "epoch": 0.014749382065723895,
      "grad_norm": 5.88922119140625,
      "learning_rate": 6.147119600233758e-05,
      "loss": 5.8109,
      "step": 91
    },
    {
      "epoch": 0.014911463187325256,
      "grad_norm": 6.442539691925049,
      "learning_rate": 6.066498153718735e-05,
      "loss": 5.9126,
      "step": 92
    },
    {
      "epoch": 0.015073544308926618,
      "grad_norm": 6.439504146575928,
      "learning_rate": 5.985585137257401e-05,
      "loss": 5.8973,
      "step": 93
    },
    {
      "epoch": 0.015235625430527979,
      "grad_norm": 6.778526306152344,
      "learning_rate": 5.90440267166055e-05,
      "loss": 5.9212,
      "step": 94
    },
    {
      "epoch": 0.015397706552129341,
      "grad_norm": 7.527666091918945,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 6.1104,
      "step": 95
    },
    {
      "epoch": 0.015559787673730702,
      "grad_norm": 8.4906644821167,
      "learning_rate": 5.74131823855921e-05,
      "loss": 6.6445,
      "step": 96
    },
    {
      "epoch": 0.015721868795332064,
      "grad_norm": 8.908556938171387,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 6.2731,
      "step": 97
    },
    {
      "epoch": 0.015883949916933426,
      "grad_norm": 10.14052963256836,
      "learning_rate": 5.577423184847932e-05,
      "loss": 6.0604,
      "step": 98
    },
    {
      "epoch": 0.016046031038534785,
      "grad_norm": 12.570101737976074,
      "learning_rate": 5.495227651252315e-05,
      "loss": 6.3531,
      "step": 99
    },
    {
      "epoch": 0.016208112160136148,
      "grad_norm": 14.668326377868652,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 6.2624,
      "step": 100
    },
    {
      "epoch": 0.016208112160136148,
      "eval_loss": 1.3847177028656006,
      "eval_runtime": 1029.8455,
      "eval_samples_per_second": 10.091,
      "eval_steps_per_second": 2.523,
      "step": 100
    },
    {
      "epoch": 0.01637019328173751,
      "grad_norm": 4.2005791664123535,
      "learning_rate": 5.330452921628497e-05,
      "loss": 5.9015,
      "step": 101
    },
    {
      "epoch": 0.016532274403338872,
      "grad_norm": 3.48933744430542,
      "learning_rate": 5.247918773366112e-05,
      "loss": 5.3167,
      "step": 102
    },
    {
      "epoch": 0.01669435552494023,
      "grad_norm": 3.365427255630493,
      "learning_rate": 5.165316846586541e-05,
      "loss": 5.4929,
      "step": 103
    },
    {
      "epoch": 0.016856436646541594,
      "grad_norm": 3.3781585693359375,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 4.9119,
      "step": 104
    },
    {
      "epoch": 0.017018517768142956,
      "grad_norm": 3.499401807785034,
      "learning_rate": 5e-05,
      "loss": 5.0897,
      "step": 105
    },
    {
      "epoch": 0.017180598889744318,
      "grad_norm": 3.4389567375183105,
      "learning_rate": 4.917330276168208e-05,
      "loss": 5.0629,
      "step": 106
    },
    {
      "epoch": 0.017342680011345677,
      "grad_norm": 3.493772029876709,
      "learning_rate": 4.834683153413459e-05,
      "loss": 4.9737,
      "step": 107
    },
    {
      "epoch": 0.01750476113294704,
      "grad_norm": 3.5260846614837646,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 4.9115,
      "step": 108
    },
    {
      "epoch": 0.017666842254548402,
      "grad_norm": 3.542769193649292,
      "learning_rate": 4.669547078371504e-05,
      "loss": 5.0926,
      "step": 109
    },
    {
      "epoch": 0.017828923376149764,
      "grad_norm": 3.4510931968688965,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 5.1305,
      "step": 110
    },
    {
      "epoch": 0.017991004497751123,
      "grad_norm": 3.4425296783447266,
      "learning_rate": 4.504772348747687e-05,
      "loss": 5.0381,
      "step": 111
    },
    {
      "epoch": 0.018153085619352485,
      "grad_norm": 3.593886137008667,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 5.0233,
      "step": 112
    },
    {
      "epoch": 0.018315166740953848,
      "grad_norm": 3.619311809539795,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 4.9667,
      "step": 113
    },
    {
      "epoch": 0.01847724786255521,
      "grad_norm": 3.6048331260681152,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 5.0864,
      "step": 114
    },
    {
      "epoch": 0.01863932898415657,
      "grad_norm": 3.5712852478027344,
      "learning_rate": 4.17702704859633e-05,
      "loss": 4.9163,
      "step": 115
    },
    {
      "epoch": 0.01880141010575793,
      "grad_norm": 4.003663539886475,
      "learning_rate": 4.095597328339452e-05,
      "loss": 5.2713,
      "step": 116
    },
    {
      "epoch": 0.018963491227359294,
      "grad_norm": 3.878546953201294,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 4.9272,
      "step": 117
    },
    {
      "epoch": 0.019125572348960656,
      "grad_norm": 3.715549945831299,
      "learning_rate": 3.933501846281267e-05,
      "loss": 4.9297,
      "step": 118
    },
    {
      "epoch": 0.019287653470562015,
      "grad_norm": 3.861307144165039,
      "learning_rate": 3.852880399766243e-05,
      "loss": 5.0783,
      "step": 119
    },
    {
      "epoch": 0.019449734592163377,
      "grad_norm": 3.8011298179626465,
      "learning_rate": 3.772572564296005e-05,
      "loss": 4.9478,
      "step": 120
    },
    {
      "epoch": 0.01961181571376474,
      "grad_norm": 3.8588430881500244,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 5.0608,
      "step": 121
    },
    {
      "epoch": 0.019773896835366102,
      "grad_norm": 4.052944660186768,
      "learning_rate": 3.612985456190778e-05,
      "loss": 5.1274,
      "step": 122
    },
    {
      "epoch": 0.01993597795696746,
      "grad_norm": 3.9233617782592773,
      "learning_rate": 3.533749813077677e-05,
      "loss": 5.0635,
      "step": 123
    },
    {
      "epoch": 0.020098059078568823,
      "grad_norm": 3.8510828018188477,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 5.0339,
      "step": 124
    },
    {
      "epoch": 0.020260140200170185,
      "grad_norm": 4.156836032867432,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 5.1914,
      "step": 125
    },
    {
      "epoch": 0.020422221321771548,
      "grad_norm": 4.0888471603393555,
      "learning_rate": 3.298534127791785e-05,
      "loss": 5.4011,
      "step": 126
    },
    {
      "epoch": 0.020584302443372907,
      "grad_norm": 4.1924028396606445,
      "learning_rate": 3.221030765387417e-05,
      "loss": 5.1957,
      "step": 127
    },
    {
      "epoch": 0.02074638356497427,
      "grad_norm": 4.259236812591553,
      "learning_rate": 3.144013755408895e-05,
      "loss": 5.1416,
      "step": 128
    },
    {
      "epoch": 0.02090846468657563,
      "grad_norm": 4.339385509490967,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 5.3724,
      "step": 129
    },
    {
      "epoch": 0.021070545808176994,
      "grad_norm": 4.363649845123291,
      "learning_rate": 2.991522876735154e-05,
      "loss": 5.3802,
      "step": 130
    },
    {
      "epoch": 0.021232626929778353,
      "grad_norm": 4.244985103607178,
      "learning_rate": 2.916090697523549e-05,
      "loss": 4.8814,
      "step": 131
    },
    {
      "epoch": 0.021394708051379715,
      "grad_norm": 4.618222713470459,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 5.4262,
      "step": 132
    },
    {
      "epoch": 0.021556789172981077,
      "grad_norm": 4.758816242218018,
      "learning_rate": 2.766955965735968e-05,
      "loss": 5.2982,
      "step": 133
    },
    {
      "epoch": 0.02171887029458244,
      "grad_norm": 4.659364223480225,
      "learning_rate": 2.693294185106562e-05,
      "loss": 5.2061,
      "step": 134
    },
    {
      "epoch": 0.0218809514161838,
      "grad_norm": 4.976042747497559,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 5.2508,
      "step": 135
    },
    {
      "epoch": 0.02204303253778516,
      "grad_norm": 4.979991912841797,
      "learning_rate": 2.547882480847461e-05,
      "loss": 5.3716,
      "step": 136
    },
    {
      "epoch": 0.022205113659386523,
      "grad_norm": 4.848566055297852,
      "learning_rate": 2.476172311325783e-05,
      "loss": 5.3942,
      "step": 137
    },
    {
      "epoch": 0.022367194780987885,
      "grad_norm": 5.319351673126221,
      "learning_rate": 2.405152131093926e-05,
      "loss": 5.2437,
      "step": 138
    },
    {
      "epoch": 0.022529275902589244,
      "grad_norm": 5.115017890930176,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 5.4039,
      "step": 139
    },
    {
      "epoch": 0.022691357024190607,
      "grad_norm": 5.578060626983643,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 5.2967,
      "step": 140
    },
    {
      "epoch": 0.02285343814579197,
      "grad_norm": 5.800137042999268,
      "learning_rate": 2.196424713241637e-05,
      "loss": 5.972,
      "step": 141
    },
    {
      "epoch": 0.02301551926739333,
      "grad_norm": 6.258816242218018,
      "learning_rate": 2.128356686585282e-05,
      "loss": 6.232,
      "step": 142
    },
    {
      "epoch": 0.02317760038899469,
      "grad_norm": 6.430749893188477,
      "learning_rate": 2.061073738537635e-05,
      "loss": 5.7707,
      "step": 143
    },
    {
      "epoch": 0.023339681510596053,
      "grad_norm": 6.416401386260986,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 5.7204,
      "step": 144
    },
    {
      "epoch": 0.023501762632197415,
      "grad_norm": 7.269473075866699,
      "learning_rate": 1.928936436551661e-05,
      "loss": 6.021,
      "step": 145
    },
    {
      "epoch": 0.023663843753798777,
      "grad_norm": 7.365699768066406,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 5.6886,
      "step": 146
    },
    {
      "epoch": 0.023825924875400136,
      "grad_norm": 7.84915828704834,
      "learning_rate": 1.800157297483417e-05,
      "loss": 5.7911,
      "step": 147
    },
    {
      "epoch": 0.0239880059970015,
      "grad_norm": 9.239398956298828,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 6.2228,
      "step": 148
    },
    {
      "epoch": 0.02415008711860286,
      "grad_norm": 10.24454402923584,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 5.6922,
      "step": 149
    },
    {
      "epoch": 0.024312168240204223,
      "grad_norm": 14.783937454223633,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 5.2486,
      "step": 150
    },
    {
      "epoch": 0.024312168240204223,
      "eval_loss": 1.313240647315979,
      "eval_runtime": 1031.1636,
      "eval_samples_per_second": 10.078,
      "eval_steps_per_second": 2.519,
      "step": 150
    },
    {
      "epoch": 0.024474249361805582,
      "grad_norm": 3.495206356048584,
      "learning_rate": 1.553232954407171e-05,
      "loss": 5.3843,
      "step": 151
    },
    {
      "epoch": 0.024636330483406944,
      "grad_norm": 3.2122416496276855,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 5.6559,
      "step": 152
    },
    {
      "epoch": 0.024798411605008307,
      "grad_norm": 3.382153034210205,
      "learning_rate": 1.435357758543015e-05,
      "loss": 5.4387,
      "step": 153
    },
    {
      "epoch": 0.02496049272660967,
      "grad_norm": 3.1782279014587402,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 5.2725,
      "step": 154
    },
    {
      "epoch": 0.025122573848211028,
      "grad_norm": 3.148488759994507,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 4.7697,
      "step": 155
    },
    {
      "epoch": 0.02528465496981239,
      "grad_norm": 3.3467092514038086,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 5.0527,
      "step": 156
    },
    {
      "epoch": 0.025446736091413753,
      "grad_norm": 3.38024640083313,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 5.2292,
      "step": 157
    },
    {
      "epoch": 0.025608817213015115,
      "grad_norm": 3.4166059494018555,
      "learning_rate": 1.157994445715706e-05,
      "loss": 5.0702,
      "step": 158
    },
    {
      "epoch": 0.025770898334616474,
      "grad_norm": 3.418752670288086,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 4.9772,
      "step": 159
    },
    {
      "epoch": 0.025932979456217836,
      "grad_norm": 3.2902719974517822,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 4.678,
      "step": 160
    },
    {
      "epoch": 0.0260950605778192,
      "grad_norm": 3.5988080501556396,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 5.0181,
      "step": 161
    },
    {
      "epoch": 0.02625714169942056,
      "grad_norm": 3.4244163036346436,
      "learning_rate": 9.549150281252633e-06,
      "loss": 4.946,
      "step": 162
    },
    {
      "epoch": 0.02641922282102192,
      "grad_norm": 3.396667957305908,
      "learning_rate": 9.068759265665384e-06,
      "loss": 4.9048,
      "step": 163
    },
    {
      "epoch": 0.026581303942623282,
      "grad_norm": 3.550051212310791,
      "learning_rate": 8.599558442598998e-06,
      "loss": 5.0899,
      "step": 164
    },
    {
      "epoch": 0.026743385064224644,
      "grad_norm": 3.6404800415039062,
      "learning_rate": 8.141676086873572e-06,
      "loss": 5.2303,
      "step": 165
    },
    {
      "epoch": 0.026905466185826007,
      "grad_norm": 3.5576372146606445,
      "learning_rate": 7.695237378953223e-06,
      "loss": 4.7082,
      "step": 166
    },
    {
      "epoch": 0.02706754730742737,
      "grad_norm": 3.565706491470337,
      "learning_rate": 7.260364370723044e-06,
      "loss": 4.7021,
      "step": 167
    },
    {
      "epoch": 0.027229628429028728,
      "grad_norm": 3.6336700916290283,
      "learning_rate": 6.837175952121306e-06,
      "loss": 4.8047,
      "step": 168
    },
    {
      "epoch": 0.02739170955063009,
      "grad_norm": 3.8932554721832275,
      "learning_rate": 6.425787818636131e-06,
      "loss": 5.3234,
      "step": 169
    },
    {
      "epoch": 0.027553790672231453,
      "grad_norm": 3.937469720840454,
      "learning_rate": 6.026312439675552e-06,
      "loss": 5.1347,
      "step": 170
    },
    {
      "epoch": 0.027715871793832815,
      "grad_norm": 3.751491069793701,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 4.7766,
      "step": 171
    },
    {
      "epoch": 0.027877952915434174,
      "grad_norm": 3.9108495712280273,
      "learning_rate": 5.263533508961827e-06,
      "loss": 5.2754,
      "step": 172
    },
    {
      "epoch": 0.028040034037035536,
      "grad_norm": 3.858539342880249,
      "learning_rate": 4.900438493352055e-06,
      "loss": 4.8552,
      "step": 173
    },
    {
      "epoch": 0.0282021151586369,
      "grad_norm": 3.9055700302124023,
      "learning_rate": 4.549673247541875e-06,
      "loss": 5.0243,
      "step": 174
    },
    {
      "epoch": 0.02836419628023826,
      "grad_norm": 4.093253135681152,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 5.0397,
      "step": 175
    },
    {
      "epoch": 0.02852627740183962,
      "grad_norm": 4.234896659851074,
      "learning_rate": 3.885512251130763e-06,
      "loss": 4.8942,
      "step": 176
    },
    {
      "epoch": 0.028688358523440982,
      "grad_norm": 4.321315765380859,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 5.0831,
      "step": 177
    },
    {
      "epoch": 0.028850439645042344,
      "grad_norm": 4.054682731628418,
      "learning_rate": 3.271776770026963e-06,
      "loss": 4.8865,
      "step": 178
    },
    {
      "epoch": 0.029012520766643707,
      "grad_norm": 4.357540130615234,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 5.0379,
      "step": 179
    },
    {
      "epoch": 0.029174601888245066,
      "grad_norm": 4.804247856140137,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 5.2027,
      "step": 180
    },
    {
      "epoch": 0.029336683009846428,
      "grad_norm": 4.579477787017822,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 5.4622,
      "step": 181
    },
    {
      "epoch": 0.02949876413144779,
      "grad_norm": 4.408065319061279,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 5.1606,
      "step": 182
    },
    {
      "epoch": 0.029660845253049153,
      "grad_norm": 4.742163181304932,
      "learning_rate": 1.962316193157593e-06,
      "loss": 5.2205,
      "step": 183
    },
    {
      "epoch": 0.02982292637465051,
      "grad_norm": 4.853640079498291,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 5.5836,
      "step": 184
    },
    {
      "epoch": 0.029985007496251874,
      "grad_norm": 4.554466724395752,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 4.947,
      "step": 185
    },
    {
      "epoch": 0.030147088617853236,
      "grad_norm": 4.986450672149658,
      "learning_rate": 1.333670137599713e-06,
      "loss": 5.2926,
      "step": 186
    },
    {
      "epoch": 0.0303091697394546,
      "grad_norm": 4.943232536315918,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 5.4558,
      "step": 187
    },
    {
      "epoch": 0.030471250861055958,
      "grad_norm": 4.997854709625244,
      "learning_rate": 9.810017062595322e-07,
      "loss": 5.1901,
      "step": 188
    },
    {
      "epoch": 0.03063333198265732,
      "grad_norm": 5.163132667541504,
      "learning_rate": 8.247462563808817e-07,
      "loss": 5.2417,
      "step": 189
    },
    {
      "epoch": 0.030795413104258682,
      "grad_norm": 5.660645961761475,
      "learning_rate": 6.819348298638839e-07,
      "loss": 5.8253,
      "step": 190
    },
    {
      "epoch": 0.030957494225860045,
      "grad_norm": 6.080727577209473,
      "learning_rate": 5.526064699265753e-07,
      "loss": 5.9154,
      "step": 191
    },
    {
      "epoch": 0.031119575347461403,
      "grad_norm": 5.952897548675537,
      "learning_rate": 4.367965336512403e-07,
      "loss": 5.6952,
      "step": 192
    },
    {
      "epoch": 0.031281656469062766,
      "grad_norm": 6.578916072845459,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 5.8682,
      "step": 193
    },
    {
      "epoch": 0.03144373759066413,
      "grad_norm": 7.020766735076904,
      "learning_rate": 2.458548727494292e-07,
      "loss": 6.2454,
      "step": 194
    },
    {
      "epoch": 0.03160581871226549,
      "grad_norm": 6.9942121505737305,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 5.9022,
      "step": 195
    },
    {
      "epoch": 0.03176789983386685,
      "grad_norm": 7.8461833000183105,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 5.6423,
      "step": 196
    },
    {
      "epoch": 0.031929980955468215,
      "grad_norm": 7.985757350921631,
      "learning_rate": 6.150154258476315e-08,
      "loss": 5.7497,
      "step": 197
    },
    {
      "epoch": 0.03209206207706957,
      "grad_norm": 9.962211608886719,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 5.9707,
      "step": 198
    },
    {
      "epoch": 0.03225414319867093,
      "grad_norm": 10.860705375671387,
      "learning_rate": 6.834750376549792e-09,
      "loss": 5.6974,
      "step": 199
    },
    {
      "epoch": 0.032416224320272295,
      "grad_norm": 13.134469032287598,
      "learning_rate": 0.0,
      "loss": 5.8554,
      "step": 200
    },
    {
      "epoch": 0.032416224320272295,
      "eval_loss": 1.2995595932006836,
      "eval_runtime": 1031.1144,
      "eval_samples_per_second": 10.078,
      "eval_steps_per_second": 2.52,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.656168479842632e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}