{
  "best_metric": 2.0797932147979736,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.6872852233676976,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.003436426116838488,
      "grad_norm": 1699.055908203125,
      "learning_rate": 6e-06,
      "loss": 7.2109,
      "step": 1
    },
    {
      "epoch": 0.003436426116838488,
      "eval_loss": 2.100292682647705,
      "eval_runtime": 3.3852,
      "eval_samples_per_second": 144.749,
      "eval_steps_per_second": 36.335,
      "step": 1
    },
    {
      "epoch": 0.006872852233676976,
      "grad_norm": 1523.189208984375,
      "learning_rate": 1.2e-05,
      "loss": 7.1845,
      "step": 2
    },
    {
      "epoch": 0.010309278350515464,
      "grad_norm": 1488.359619140625,
      "learning_rate": 1.8e-05,
      "loss": 7.171,
      "step": 3
    },
    {
      "epoch": 0.013745704467353952,
      "grad_norm": 2312.205078125,
      "learning_rate": 2.4e-05,
      "loss": 7.1277,
      "step": 4
    },
    {
      "epoch": 0.01718213058419244,
      "grad_norm": 1413.8890380859375,
      "learning_rate": 3e-05,
      "loss": 6.9779,
      "step": 5
    },
    {
      "epoch": 0.020618556701030927,
      "grad_norm": 1860.9564208984375,
      "learning_rate": 3.6e-05,
      "loss": 7.1177,
      "step": 6
    },
    {
      "epoch": 0.024054982817869417,
      "grad_norm": 1768.141357421875,
      "learning_rate": 4.2e-05,
      "loss": 7.0749,
      "step": 7
    },
    {
      "epoch": 0.027491408934707903,
      "grad_norm": 1243.787841796875,
      "learning_rate": 4.8e-05,
      "loss": 6.9072,
      "step": 8
    },
    {
      "epoch": 0.030927835051546393,
      "grad_norm": 1541.1134033203125,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 6.9315,
      "step": 9
    },
    {
      "epoch": 0.03436426116838488,
      "grad_norm": 1837.2386474609375,
      "learning_rate": 6e-05,
      "loss": 7.1346,
      "step": 10
    },
    {
      "epoch": 0.037800687285223365,
      "grad_norm": 2413.2001953125,
      "learning_rate": 5.999589914977407e-05,
      "loss": 6.8781,
      "step": 11
    },
    {
      "epoch": 0.041237113402061855,
      "grad_norm": 1226.415283203125,
      "learning_rate": 5.998359772022778e-05,
      "loss": 7.131,
      "step": 12
    },
    {
      "epoch": 0.044673539518900345,
      "grad_norm": 1650.582763671875,
      "learning_rate": 5.996309907444915e-05,
      "loss": 7.0264,
      "step": 13
    },
    {
      "epoch": 0.048109965635738834,
      "grad_norm": 1872.6573486328125,
      "learning_rate": 5.9934408816563236e-05,
      "loss": 6.9366,
      "step": 14
    },
    {
      "epoch": 0.05154639175257732,
      "grad_norm": 3318.892578125,
      "learning_rate": 5.98975347902001e-05,
      "loss": 6.8686,
      "step": 15
    },
    {
      "epoch": 0.054982817869415807,
      "grad_norm": 2973.668701171875,
      "learning_rate": 5.9852487076350345e-05,
      "loss": 7.1014,
      "step": 16
    },
    {
      "epoch": 0.058419243986254296,
      "grad_norm": 1740.9530029296875,
      "learning_rate": 5.979927799060915e-05,
      "loss": 7.3723,
      "step": 17
    },
    {
      "epoch": 0.061855670103092786,
      "grad_norm": 2309.447021484375,
      "learning_rate": 5.9737922079809257e-05,
      "loss": 7.2998,
      "step": 18
    },
    {
      "epoch": 0.06529209621993128,
      "grad_norm": 1850.541259765625,
      "learning_rate": 5.9668436118044054e-05,
      "loss": 7.0061,
      "step": 19
    },
    {
      "epoch": 0.06872852233676977,
      "grad_norm": 4049.42138671875,
      "learning_rate": 5.959083910208167e-05,
      "loss": 7.2665,
      "step": 20
    },
    {
      "epoch": 0.07216494845360824,
      "grad_norm": 1948.1513671875,
      "learning_rate": 5.9505152246171474e-05,
      "loss": 7.543,
      "step": 21
    },
    {
      "epoch": 0.07560137457044673,
      "grad_norm": 2752.593994140625,
      "learning_rate": 5.941139897624428e-05,
      "loss": 7.0936,
      "step": 22
    },
    {
      "epoch": 0.07903780068728522,
      "grad_norm": 2305.347900390625,
      "learning_rate": 5.9309604923507984e-05,
      "loss": 7.416,
      "step": 23
    },
    {
      "epoch": 0.08247422680412371,
      "grad_norm": 2590.050537109375,
      "learning_rate": 5.9199797917440176e-05,
      "loss": 7.9452,
      "step": 24
    },
    {
      "epoch": 0.0859106529209622,
      "grad_norm": 1818.8985595703125,
      "learning_rate": 5.908200797817991e-05,
      "loss": 7.2278,
      "step": 25
    },
    {
      "epoch": 0.08934707903780069,
      "grad_norm": 2739.406005859375,
      "learning_rate": 5.895626730832046e-05,
      "loss": 7.5109,
      "step": 26
    },
    {
      "epoch": 0.09278350515463918,
      "grad_norm": 2736.05126953125,
      "learning_rate": 5.882261028410545e-05,
      "loss": 7.4654,
      "step": 27
    },
    {
      "epoch": 0.09621993127147767,
      "grad_norm": 2838.5927734375,
      "learning_rate": 5.8681073446030734e-05,
      "loss": 7.6614,
      "step": 28
    },
    {
      "epoch": 0.09965635738831616,
      "grad_norm": 3893.25048828125,
      "learning_rate": 5.853169548885461e-05,
      "loss": 7.6923,
      "step": 29
    },
    {
      "epoch": 0.10309278350515463,
      "grad_norm": 2812.3095703125,
      "learning_rate": 5.8374517251019035e-05,
      "loss": 8.0422,
      "step": 30
    },
    {
      "epoch": 0.10652920962199312,
      "grad_norm": 4552.505859375,
      "learning_rate": 5.820958170348484e-05,
      "loss": 7.2086,
      "step": 31
    },
    {
      "epoch": 0.10996563573883161,
      "grad_norm": 3343.3740234375,
      "learning_rate": 5.8036933937983825e-05,
      "loss": 7.7071,
      "step": 32
    },
    {
      "epoch": 0.1134020618556701,
      "grad_norm": 1772.2581787109375,
      "learning_rate": 5.7856621154691217e-05,
      "loss": 7.2992,
      "step": 33
    },
    {
      "epoch": 0.11683848797250859,
      "grad_norm": 2802.28466796875,
      "learning_rate": 5.766869264932154e-05,
      "loss": 7.1957,
      "step": 34
    },
    {
      "epoch": 0.12027491408934708,
      "grad_norm": 2161.58447265625,
      "learning_rate": 5.747319979965172e-05,
      "loss": 8.1493,
      "step": 35
    },
    {
      "epoch": 0.12371134020618557,
      "grad_norm": 2594.178466796875,
      "learning_rate": 5.727019605147488e-05,
      "loss": 8.7289,
      "step": 36
    },
    {
      "epoch": 0.12714776632302405,
      "grad_norm": 4844.7353515625,
      "learning_rate": 5.7059736903988775e-05,
      "loss": 8.657,
      "step": 37
    },
    {
      "epoch": 0.13058419243986255,
      "grad_norm": 2519.898193359375,
      "learning_rate": 5.684187989462291e-05,
      "loss": 8.4565,
      "step": 38
    },
    {
      "epoch": 0.13402061855670103,
      "grad_norm": 4387.62646484375,
      "learning_rate": 5.661668458330836e-05,
      "loss": 9.1231,
      "step": 39
    },
    {
      "epoch": 0.13745704467353953,
      "grad_norm": 2012.0771484375,
      "learning_rate": 5.638421253619467e-05,
      "loss": 8.1266,
      "step": 40
    },
    {
      "epoch": 0.140893470790378,
      "grad_norm": 4595.40771484375,
      "learning_rate": 5.614452730881832e-05,
      "loss": 7.7234,
      "step": 41
    },
    {
      "epoch": 0.14432989690721648,
      "grad_norm": 7514.57177734375,
      "learning_rate": 5.589769442872722e-05,
      "loss": 9.3852,
      "step": 42
    },
    {
      "epoch": 0.14776632302405499,
      "grad_norm": 4701.0615234375,
      "learning_rate": 5.5643781377566175e-05,
      "loss": 7.941,
      "step": 43
    },
    {
      "epoch": 0.15120274914089346,
      "grad_norm": 7156.04833984375,
      "learning_rate": 5.538285757262806e-05,
      "loss": 8.9046,
      "step": 44
    },
    {
      "epoch": 0.15463917525773196,
      "grad_norm": 4677.55810546875,
      "learning_rate": 5.5114994347875856e-05,
      "loss": 9.0037,
      "step": 45
    },
    {
      "epoch": 0.15807560137457044,
      "grad_norm": 7397.443359375,
      "learning_rate": 5.48402649344406e-05,
      "loss": 7.6486,
      "step": 46
    },
    {
      "epoch": 0.16151202749140894,
      "grad_norm": 4792.80322265625,
      "learning_rate": 5.455874444060078e-05,
      "loss": 9.5176,
      "step": 47
    },
    {
      "epoch": 0.16494845360824742,
      "grad_norm": 8214.7734375,
      "learning_rate": 5.427050983124843e-05,
      "loss": 8.8844,
      "step": 48
    },
    {
      "epoch": 0.16838487972508592,
      "grad_norm": 10067.8759765625,
      "learning_rate": 5.397563990684774e-05,
      "loss": 9.3324,
      "step": 49
    },
    {
      "epoch": 0.1718213058419244,
      "grad_norm": 10682.091796875,
      "learning_rate": 5.367421528189181e-05,
      "loss": 12.6026,
      "step": 50
    },
    {
      "epoch": 0.1718213058419244,
      "eval_loss": 2.1914706230163574,
      "eval_runtime": 3.3857,
      "eval_samples_per_second": 144.726,
      "eval_steps_per_second": 36.329,
      "step": 50
    },
    {
      "epoch": 0.17525773195876287,
      "grad_norm": 3756.540771484375,
      "learning_rate": 5.336631836286338e-05,
      "loss": 7.4742,
      "step": 51
    },
    {
      "epoch": 0.17869415807560138,
      "grad_norm": 1808.46044921875,
      "learning_rate": 5.3052033325705774e-05,
      "loss": 7.1845,
      "step": 52
    },
    {
      "epoch": 0.18213058419243985,
      "grad_norm": 1342.697509765625,
      "learning_rate": 5.2731446092810044e-05,
      "loss": 7.1671,
      "step": 53
    },
    {
      "epoch": 0.18556701030927836,
      "grad_norm": 1255.91162109375,
      "learning_rate": 5.240464430952462e-05,
      "loss": 6.8797,
      "step": 54
    },
    {
      "epoch": 0.18900343642611683,
      "grad_norm": 1505.02294921875,
      "learning_rate": 5.207171732019395e-05,
      "loss": 7.296,
      "step": 55
    },
    {
      "epoch": 0.19243986254295534,
      "grad_norm": 1200.6898193359375,
      "learning_rate": 5.1732756143732675e-05,
      "loss": 7.0165,
      "step": 56
    },
    {
      "epoch": 0.1958762886597938,
      "grad_norm": 1603.5499267578125,
      "learning_rate": 5.1387853448741916e-05,
      "loss": 7.0639,
      "step": 57
    },
    {
      "epoch": 0.19931271477663232,
      "grad_norm": 1951.0045166015625,
      "learning_rate": 5.103710352817465e-05,
      "loss": 7.128,
      "step": 58
    },
    {
      "epoch": 0.2027491408934708,
      "grad_norm": 1422.667236328125,
      "learning_rate": 5.068060227355698e-05,
      "loss": 6.9692,
      "step": 59
    },
    {
      "epoch": 0.20618556701030927,
      "grad_norm": 1536.1329345703125,
      "learning_rate": 5.0318447148772234e-05,
      "loss": 7.0729,
      "step": 60
    },
    {
      "epoch": 0.20962199312714777,
      "grad_norm": 1215.6038818359375,
      "learning_rate": 4.995073716341545e-05,
      "loss": 7.1131,
      "step": 61
    },
    {
      "epoch": 0.21305841924398625,
      "grad_norm": 1439.0755615234375,
      "learning_rate": 4.957757284572506e-05,
      "loss": 7.0884,
      "step": 62
    },
    {
      "epoch": 0.21649484536082475,
      "grad_norm": 1372.0079345703125,
      "learning_rate": 4.91990562150995e-05,
      "loss": 7.1223,
      "step": 63
    },
    {
      "epoch": 0.21993127147766323,
      "grad_norm": 1942.8409423828125,
      "learning_rate": 4.881529075420611e-05,
      "loss": 7.3073,
      "step": 64
    },
    {
      "epoch": 0.22336769759450173,
      "grad_norm": 2515.036865234375,
      "learning_rate": 4.8426381380690036e-05,
      "loss": 7.0794,
      "step": 65
    },
    {
      "epoch": 0.2268041237113402,
      "grad_norm": 2236.493408203125,
      "learning_rate": 4.8032434418490753e-05,
      "loss": 7.2286,
      "step": 66
    },
    {
      "epoch": 0.23024054982817868,
      "grad_norm": 1951.1087646484375,
      "learning_rate": 4.7633557568774194e-05,
      "loss": 7.4805,
      "step": 67
    },
    {
      "epoch": 0.23367697594501718,
      "grad_norm": 1184.3919677734375,
      "learning_rate": 4.722985988048831e-05,
      "loss": 7.3142,
      "step": 68
    },
    {
      "epoch": 0.23711340206185566,
      "grad_norm": 1259.7314453125,
      "learning_rate": 4.6821451720550184e-05,
      "loss": 7.6081,
      "step": 69
    },
    {
      "epoch": 0.24054982817869416,
      "grad_norm": 2070.515869140625,
      "learning_rate": 4.640844474367282e-05,
      "loss": 7.4036,
      "step": 70
    },
    {
      "epoch": 0.24398625429553264,
      "grad_norm": 1795.15380859375,
      "learning_rate": 4.5990951861839815e-05,
      "loss": 7.2923,
      "step": 71
    },
    {
      "epoch": 0.24742268041237114,
      "grad_norm": 2915.548583984375,
      "learning_rate": 4.5569087213436455e-05,
      "loss": 7.3081,
      "step": 72
    },
    {
      "epoch": 0.2508591065292096,
      "grad_norm": 2263.517333984375,
      "learning_rate": 4.514296613204532e-05,
      "loss": 7.5925,
      "step": 73
    },
    {
      "epoch": 0.2542955326460481,
      "grad_norm": 2741.153076171875,
      "learning_rate": 4.471270511491525e-05,
      "loss": 7.8143,
      "step": 74
    },
    {
      "epoch": 0.25773195876288657,
      "grad_norm": 2714.728271484375,
      "learning_rate": 4.427842179111221e-05,
      "loss": 7.8965,
      "step": 75
    },
    {
      "epoch": 0.2611683848797251,
      "grad_norm": 2927.20458984375,
      "learning_rate": 4.3840234889360634e-05,
      "loss": 7.6063,
      "step": 76
    },
    {
      "epoch": 0.2646048109965636,
      "grad_norm": 3291.874755859375,
      "learning_rate": 4.33982642055842e-05,
      "loss": 8.0501,
      "step": 77
    },
    {
      "epoch": 0.26804123711340205,
      "grad_norm": 3684.526611328125,
      "learning_rate": 4.2952630570154785e-05,
      "loss": 9.1159,
      "step": 78
    },
    {
      "epoch": 0.27147766323024053,
      "grad_norm": 2882.729736328125,
      "learning_rate": 4.250345581485871e-05,
      "loss": 8.8894,
      "step": 79
    },
    {
      "epoch": 0.27491408934707906,
      "grad_norm": 3075.021240234375,
      "learning_rate": 4.205086273958909e-05,
      "loss": 8.3385,
      "step": 80
    },
    {
      "epoch": 0.27835051546391754,
      "grad_norm": 3509.171875,
      "learning_rate": 4.1594975078773565e-05,
      "loss": 8.1994,
      "step": 81
    },
    {
      "epoch": 0.281786941580756,
      "grad_norm": 3701.130126953125,
      "learning_rate": 4.113591746754662e-05,
      "loss": 9.2697,
      "step": 82
    },
    {
      "epoch": 0.2852233676975945,
      "grad_norm": 4614.2958984375,
      "learning_rate": 4.06738154076755e-05,
      "loss": 8.6035,
      "step": 83
    },
    {
      "epoch": 0.28865979381443296,
      "grad_norm": 4621.25146484375,
      "learning_rate": 4.020879523324929e-05,
      "loss": 7.9529,
      "step": 84
    },
    {
      "epoch": 0.2920962199312715,
      "grad_norm": 4300.3115234375,
      "learning_rate": 3.974098407614051e-05,
      "loss": 8.7308,
      "step": 85
    },
    {
      "epoch": 0.29553264604810997,
      "grad_norm": 3846.455810546875,
      "learning_rate": 3.927050983124842e-05,
      "loss": 8.7034,
      "step": 86
    },
    {
      "epoch": 0.29896907216494845,
      "grad_norm": 6396.93603515625,
      "learning_rate": 3.8797501121533946e-05,
      "loss": 9.2579,
      "step": 87
    },
    {
      "epoch": 0.3024054982817869,
      "grad_norm": 5468.02490234375,
      "learning_rate": 3.832208726285534e-05,
      "loss": 9.0404,
      "step": 88
    },
    {
      "epoch": 0.30584192439862545,
      "grad_norm": 5713.4375,
      "learning_rate": 3.784439822861459e-05,
      "loss": 8.6309,
      "step": 89
    },
    {
      "epoch": 0.30927835051546393,
      "grad_norm": 7521.90966796875,
      "learning_rate": 3.7364564614223976e-05,
      "loss": 8.98,
      "step": 90
    },
    {
      "epoch": 0.3127147766323024,
      "grad_norm": 5354.02001953125,
      "learning_rate": 3.688271760140255e-05,
      "loss": 8.8258,
      "step": 91
    },
    {
      "epoch": 0.3161512027491409,
      "grad_norm": 9768.9306640625,
      "learning_rate": 3.6398988922312406e-05,
      "loss": 7.5365,
      "step": 92
    },
    {
      "epoch": 0.31958762886597936,
      "grad_norm": 6491.6865234375,
      "learning_rate": 3.591351082354441e-05,
      "loss": 10.3068,
      "step": 93
    },
    {
      "epoch": 0.3230240549828179,
      "grad_norm": 8386.2021484375,
      "learning_rate": 3.54264160299633e-05,
      "loss": 10.0165,
      "step": 94
    },
    {
      "epoch": 0.32646048109965636,
      "grad_norm": 6590.6767578125,
      "learning_rate": 3.493783770842202e-05,
      "loss": 8.7206,
      "step": 95
    },
    {
      "epoch": 0.32989690721649484,
      "grad_norm": 9552.482421875,
      "learning_rate": 3.444790943135526e-05,
      "loss": 8.9128,
      "step": 96
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 7169.98779296875,
      "learning_rate": 3.3956765140262074e-05,
      "loss": 9.4693,
      "step": 97
    },
    {
      "epoch": 0.33676975945017185,
      "grad_norm": 14819.2490234375,
      "learning_rate": 3.346453910908759e-05,
      "loss": 9.6341,
      "step": 98
    },
    {
      "epoch": 0.3402061855670103,
      "grad_norm": 11971.1591796875,
      "learning_rate": 3.297136590751389e-05,
      "loss": 10.5735,
      "step": 99
    },
    {
      "epoch": 0.3436426116838488,
      "grad_norm": 8081.7119140625,
      "learning_rate": 3.247738036416998e-05,
      "loss": 13.7919,
      "step": 100
    },
    {
      "epoch": 0.3436426116838488,
      "eval_loss": 2.2170331478118896,
      "eval_runtime": 3.4016,
      "eval_samples_per_second": 144.052,
      "eval_steps_per_second": 36.16,
      "step": 100
    },
    {
      "epoch": 0.3470790378006873,
      "grad_norm": 6338.203125,
      "learning_rate": 3.1982717529770985e-05,
      "loss": 8.1504,
      "step": 101
    },
    {
      "epoch": 0.35051546391752575,
      "grad_norm": 3960.850341796875,
      "learning_rate": 3.148751264019667e-05,
      "loss": 7.7718,
      "step": 102
    },
    {
      "epoch": 0.3539518900343643,
      "grad_norm": 3253.1552734375,
      "learning_rate": 3.099190107951924e-05,
      "loss": 7.5944,
      "step": 103
    },
    {
      "epoch": 0.35738831615120276,
      "grad_norm": 2214.524658203125,
      "learning_rate": 3.049601834299076e-05,
      "loss": 7.5333,
      "step": 104
    },
    {
      "epoch": 0.36082474226804123,
      "grad_norm": 2329.060791015625,
      "learning_rate": 3e-05,
      "loss": 7.5675,
      "step": 105
    },
    {
      "epoch": 0.3642611683848797,
      "grad_norm": 1914.9527587890625,
      "learning_rate": 2.9503981657009246e-05,
      "loss": 7.3778,
      "step": 106
    },
    {
      "epoch": 0.36769759450171824,
      "grad_norm": 1756.5228271484375,
      "learning_rate": 2.9008098920480752e-05,
      "loss": 7.5316,
      "step": 107
    },
    {
      "epoch": 0.3711340206185567,
      "grad_norm": 1628.3492431640625,
      "learning_rate": 2.851248735980333e-05,
      "loss": 7.1625,
      "step": 108
    },
    {
      "epoch": 0.3745704467353952,
      "grad_norm": 1857.13671875,
      "learning_rate": 2.801728247022902e-05,
      "loss": 7.4011,
      "step": 109
    },
    {
      "epoch": 0.37800687285223367,
      "grad_norm": 1966.443359375,
      "learning_rate": 2.7522619635830034e-05,
      "loss": 7.3879,
      "step": 110
    },
    {
      "epoch": 0.38144329896907214,
      "grad_norm": 1875.7662353515625,
      "learning_rate": 2.702863409248612e-05,
      "loss": 7.6974,
      "step": 111
    },
    {
      "epoch": 0.3848797250859107,
      "grad_norm": 3008.91162109375,
      "learning_rate": 2.6535460890912416e-05,
      "loss": 7.8305,
      "step": 112
    },
    {
      "epoch": 0.38831615120274915,
      "grad_norm": 2057.904052734375,
      "learning_rate": 2.604323485973793e-05,
      "loss": 7.8253,
      "step": 113
    },
    {
      "epoch": 0.3917525773195876,
      "grad_norm": 2149.185546875,
      "learning_rate": 2.555209056864474e-05,
      "loss": 7.2513,
      "step": 114
    },
    {
      "epoch": 0.3951890034364261,
      "grad_norm": 2182.825439453125,
      "learning_rate": 2.5062162291577978e-05,
      "loss": 7.7119,
      "step": 115
    },
    {
      "epoch": 0.39862542955326463,
      "grad_norm": 2759.802490234375,
      "learning_rate": 2.4573583970036712e-05,
      "loss": 7.7795,
      "step": 116
    },
    {
      "epoch": 0.4020618556701031,
      "grad_norm": 2170.989990234375,
      "learning_rate": 2.4086489176455595e-05,
      "loss": 7.8915,
      "step": 117
    },
    {
      "epoch": 0.4054982817869416,
      "grad_norm": 2467.65576171875,
      "learning_rate": 2.36010110776876e-05,
      "loss": 7.9465,
      "step": 118
    },
    {
      "epoch": 0.40893470790378006,
      "grad_norm": 2266.196044921875,
      "learning_rate": 2.3117282398597456e-05,
      "loss": 8.0087,
      "step": 119
    },
    {
      "epoch": 0.41237113402061853,
      "grad_norm": 2992.08349609375,
      "learning_rate": 2.263543538577603e-05,
      "loss": 7.7058,
      "step": 120
    },
    {
      "epoch": 0.41580756013745707,
      "grad_norm": 1708.01318359375,
      "learning_rate": 2.215560177138541e-05,
      "loss": 8.1656,
      "step": 121
    },
    {
      "epoch": 0.41924398625429554,
      "grad_norm": 4836.68505859375,
      "learning_rate": 2.167791273714467e-05,
      "loss": 7.6497,
      "step": 122
    },
    {
      "epoch": 0.422680412371134,
      "grad_norm": 2311.93994140625,
      "learning_rate": 2.1202498878466062e-05,
      "loss": 8.5898,
      "step": 123
    },
    {
      "epoch": 0.4261168384879725,
      "grad_norm": 3336.768310546875,
      "learning_rate": 2.072949016875158e-05,
      "loss": 7.8291,
      "step": 124
    },
    {
      "epoch": 0.42955326460481097,
      "grad_norm": 2548.131103515625,
      "learning_rate": 2.0259015923859498e-05,
      "loss": 7.5833,
      "step": 125
    },
    {
      "epoch": 0.4329896907216495,
      "grad_norm": 3319.1611328125,
      "learning_rate": 1.979120476675071e-05,
      "loss": 7.9073,
      "step": 126
    },
    {
      "epoch": 0.436426116838488,
      "grad_norm": 2070.8388671875,
      "learning_rate": 1.9326184592324503e-05,
      "loss": 7.893,
      "step": 127
    },
    {
      "epoch": 0.43986254295532645,
      "grad_norm": 2963.176025390625,
      "learning_rate": 1.8864082532453373e-05,
      "loss": 8.0254,
      "step": 128
    },
    {
      "epoch": 0.44329896907216493,
      "grad_norm": 3843.35107421875,
      "learning_rate": 1.840502492122644e-05,
      "loss": 8.276,
      "step": 129
    },
    {
      "epoch": 0.44673539518900346,
      "grad_norm": 3847.33349609375,
      "learning_rate": 1.7949137260410924e-05,
      "loss": 8.1098,
      "step": 130
    },
    {
      "epoch": 0.45017182130584193,
      "grad_norm": 3402.124267578125,
      "learning_rate": 1.7496544185141295e-05,
      "loss": 8.5796,
      "step": 131
    },
    {
      "epoch": 0.4536082474226804,
      "grad_norm": 3649.834228515625,
      "learning_rate": 1.7047369429845216e-05,
      "loss": 8.0409,
      "step": 132
    },
    {
      "epoch": 0.4570446735395189,
      "grad_norm": 2720.489013671875,
      "learning_rate": 1.6601735794415806e-05,
      "loss": 8.0999,
      "step": 133
    },
    {
      "epoch": 0.46048109965635736,
      "grad_norm": 2908.25537109375,
      "learning_rate": 1.615976511063937e-05,
      "loss": 8.0204,
      "step": 134
    },
    {
      "epoch": 0.4639175257731959,
      "grad_norm": 3314.382080078125,
      "learning_rate": 1.5721578208887793e-05,
      "loss": 8.3157,
      "step": 135
    },
    {
      "epoch": 0.46735395189003437,
      "grad_norm": 4135.548828125,
      "learning_rate": 1.5287294885084766e-05,
      "loss": 8.3494,
      "step": 136
    },
    {
      "epoch": 0.47079037800687284,
      "grad_norm": 4054.155517578125,
      "learning_rate": 1.4857033867954697e-05,
      "loss": 7.8104,
      "step": 137
    },
    {
      "epoch": 0.4742268041237113,
      "grad_norm": 7448.6650390625,
      "learning_rate": 1.4430912786563554e-05,
      "loss": 8.2547,
      "step": 138
    },
    {
      "epoch": 0.47766323024054985,
      "grad_norm": 2514.72705078125,
      "learning_rate": 1.4009048138160195e-05,
      "loss": 8.582,
      "step": 139
    },
    {
      "epoch": 0.48109965635738833,
      "grad_norm": 3390.73779296875,
      "learning_rate": 1.3591555256327199e-05,
      "loss": 8.3898,
      "step": 140
    },
    {
      "epoch": 0.4845360824742268,
      "grad_norm": 5241.2099609375,
      "learning_rate": 1.3178548279449822e-05,
      "loss": 8.1913,
      "step": 141
    },
    {
      "epoch": 0.4879725085910653,
      "grad_norm": 4858.0283203125,
      "learning_rate": 1.2770140119511693e-05,
      "loss": 8.7944,
      "step": 142
    },
    {
      "epoch": 0.49140893470790376,
      "grad_norm": 5469.38037109375,
      "learning_rate": 1.2366442431225809e-05,
      "loss": 8.8084,
      "step": 143
    },
    {
      "epoch": 0.4948453608247423,
      "grad_norm": 6011.06591796875,
      "learning_rate": 1.1967565581509248e-05,
      "loss": 8.8881,
      "step": 144
    },
    {
      "epoch": 0.49828178694158076,
      "grad_norm": 5176.05078125,
      "learning_rate": 1.1573618619309965e-05,
      "loss": 8.9016,
      "step": 145
    },
    {
      "epoch": 0.5017182130584192,
      "grad_norm": 6610.90478515625,
      "learning_rate": 1.1184709245793889e-05,
      "loss": 8.6331,
      "step": 146
    },
    {
      "epoch": 0.5051546391752577,
      "grad_norm": 10126.125,
      "learning_rate": 1.0800943784900502e-05,
      "loss": 7.9256,
      "step": 147
    },
    {
      "epoch": 0.5085910652920962,
      "grad_norm": 5469.77978515625,
      "learning_rate": 1.042242715427494e-05,
      "loss": 10.2646,
      "step": 148
    },
    {
      "epoch": 0.5120274914089347,
      "grad_norm": 4336.35205078125,
      "learning_rate": 1.004926283658455e-05,
      "loss": 10.6199,
      "step": 149
    },
    {
      "epoch": 0.5154639175257731,
      "grad_norm": 7235.44677734375,
      "learning_rate": 9.681552851227774e-06,
      "loss": 13.1587,
      "step": 150
    },
    {
      "epoch": 0.5154639175257731,
      "eval_loss": 2.1347427368164062,
      "eval_runtime": 3.3947,
      "eval_samples_per_second": 144.344,
      "eval_steps_per_second": 36.233,
      "step": 150
    },
    {
      "epoch": 0.5189003436426117,
      "grad_norm": 2705.450439453125,
      "learning_rate": 9.319397726443026e-06,
      "loss": 7.5747,
      "step": 151
    },
    {
      "epoch": 0.5223367697594502,
      "grad_norm": 3762.812744140625,
      "learning_rate": 8.962896471825342e-06,
      "loss": 7.5746,
      "step": 152
    },
    {
      "epoch": 0.5257731958762887,
      "grad_norm": 1911.854736328125,
      "learning_rate": 8.61214655125809e-06,
      "loss": 7.438,
      "step": 153
    },
    {
      "epoch": 0.5292096219931272,
      "grad_norm": 1469.72119140625,
      "learning_rate": 8.267243856267331e-06,
      "loss": 7.2063,
      "step": 154
    },
    {
      "epoch": 0.5326460481099656,
      "grad_norm": 1573.3465576171875,
      "learning_rate": 7.928282679806052e-06,
      "loss": 7.2888,
      "step": 155
    },
    {
      "epoch": 0.5360824742268041,
      "grad_norm": 1803.7042236328125,
      "learning_rate": 7.595355690475393e-06,
      "loss": 7.4952,
      "step": 156
    },
    {
      "epoch": 0.5395189003436426,
      "grad_norm": 1138.683837890625,
      "learning_rate": 7.268553907189964e-06,
      "loss": 7.2513,
      "step": 157
    },
    {
      "epoch": 0.5429553264604811,
      "grad_norm": 1677.6661376953125,
      "learning_rate": 6.947966674294236e-06,
      "loss": 7.5826,
      "step": 158
    },
    {
      "epoch": 0.5463917525773195,
      "grad_norm": 1217.319091796875,
      "learning_rate": 6.6336816371366305e-06,
      "loss": 7.5123,
      "step": 159
    },
    {
      "epoch": 0.5498281786941581,
      "grad_norm": 1235.499755859375,
      "learning_rate": 6.325784718108196e-06,
      "loss": 7.6381,
      "step": 160
    },
    {
      "epoch": 0.5532646048109966,
      "grad_norm": 1441.3863525390625,
      "learning_rate": 6.0243600931522595e-06,
      "loss": 7.5427,
      "step": 161
    },
    {
      "epoch": 0.5567010309278351,
      "grad_norm": 1629.83935546875,
      "learning_rate": 5.72949016875158e-06,
      "loss": 7.4876,
      "step": 162
    },
    {
      "epoch": 0.5601374570446735,
      "grad_norm": 1260.6104736328125,
      "learning_rate": 5.44125555939923e-06,
      "loss": 7.7081,
      "step": 163
    },
    {
      "epoch": 0.563573883161512,
      "grad_norm": 1765.477294921875,
      "learning_rate": 5.159735065559399e-06,
      "loss": 7.521,
      "step": 164
    },
    {
      "epoch": 0.5670103092783505,
      "grad_norm": 2019.567626953125,
      "learning_rate": 4.885005652124144e-06,
      "loss": 7.6643,
      "step": 165
    },
    {
      "epoch": 0.570446735395189,
      "grad_norm": 1756.0921630859375,
      "learning_rate": 4.617142427371934e-06,
      "loss": 7.7939,
      "step": 166
    },
    {
      "epoch": 0.5738831615120275,
      "grad_norm": 1463.5172119140625,
      "learning_rate": 4.3562186224338265e-06,
      "loss": 7.6615,
      "step": 167
    },
    {
      "epoch": 0.5773195876288659,
      "grad_norm": 1783.8548583984375,
      "learning_rate": 4.102305571272783e-06,
      "loss": 7.4967,
      "step": 168
    },
    {
      "epoch": 0.5807560137457045,
      "grad_norm": 1875.7342529296875,
      "learning_rate": 3.855472691181678e-06,
      "loss": 7.4725,
      "step": 169
    },
    {
      "epoch": 0.584192439862543,
      "grad_norm": 2582.930908203125,
      "learning_rate": 3.615787463805331e-06,
      "loss": 8.2118,
      "step": 170
    },
    {
      "epoch": 0.5876288659793815,
      "grad_norm": 1897.54638671875,
      "learning_rate": 3.383315416691646e-06,
      "loss": 8.2408,
      "step": 171
    },
    {
      "epoch": 0.5910652920962199,
      "grad_norm": 2314.038330078125,
      "learning_rate": 3.158120105377096e-06,
      "loss": 7.6892,
      "step": 172
    },
    {
      "epoch": 0.5945017182130584,
      "grad_norm": 1786.5667724609375,
      "learning_rate": 2.940263096011233e-06,
      "loss": 8.3423,
      "step": 173
    },
    {
      "epoch": 0.5979381443298969,
      "grad_norm": 2857.995849609375,
      "learning_rate": 2.729803948525125e-06,
      "loss": 7.616,
      "step": 174
    },
    {
      "epoch": 0.6013745704467354,
      "grad_norm": 2287.089111328125,
      "learning_rate": 2.526800200348275e-06,
      "loss": 8.6003,
      "step": 175
    },
    {
      "epoch": 0.6048109965635738,
      "grad_norm": 1499.41015625,
      "learning_rate": 2.3313073506784575e-06,
      "loss": 7.7733,
      "step": 176
    },
    {
      "epoch": 0.6082474226804123,
      "grad_norm": 2719.6171875,
      "learning_rate": 2.143378845308791e-06,
      "loss": 8.3946,
      "step": 177
    },
    {
      "epoch": 0.6116838487972509,
      "grad_norm": 2311.212158203125,
      "learning_rate": 1.9630660620161777e-06,
      "loss": 8.9659,
      "step": 178
    },
    {
      "epoch": 0.6151202749140894,
      "grad_norm": 2324.716552734375,
      "learning_rate": 1.790418296515165e-06,
      "loss": 8.1793,
      "step": 179
    },
    {
      "epoch": 0.6185567010309279,
      "grad_norm": 2077.054443359375,
      "learning_rate": 1.625482748980961e-06,
      "loss": 8.5296,
      "step": 180
    },
    {
      "epoch": 0.6219931271477663,
      "grad_norm": 3390.5927734375,
      "learning_rate": 1.4683045111453942e-06,
      "loss": 8.2995,
      "step": 181
    },
    {
      "epoch": 0.6254295532646048,
      "grad_norm": 2156.932373046875,
      "learning_rate": 1.3189265539692707e-06,
      "loss": 8.547,
      "step": 182
    },
    {
      "epoch": 0.6288659793814433,
      "grad_norm": 2244.710205078125,
      "learning_rate": 1.1773897158945557e-06,
      "loss": 7.3688,
      "step": 183
    },
    {
      "epoch": 0.6323024054982818,
      "grad_norm": 3128.6435546875,
      "learning_rate": 1.0437326916795432e-06,
      "loss": 9.2353,
      "step": 184
    },
    {
      "epoch": 0.6357388316151202,
      "grad_norm": 3524.7509765625,
      "learning_rate": 9.179920218200888e-07,
      "loss": 8.7623,
      "step": 185
    },
    {
      "epoch": 0.6391752577319587,
      "grad_norm": 4061.834716796875,
      "learning_rate": 8.002020825598277e-07,
      "loss": 8.2449,
      "step": 186
    },
    {
      "epoch": 0.6426116838487973,
      "grad_norm": 4144.5556640625,
      "learning_rate": 6.90395076492022e-07,
      "loss": 8.4433,
      "step": 187
    },
    {
      "epoch": 0.6460481099656358,
      "grad_norm": 3348.039306640625,
      "learning_rate": 5.886010237557194e-07,
      "loss": 7.4949,
      "step": 188
    },
    {
      "epoch": 0.6494845360824743,
      "grad_norm": 2904.456298828125,
      "learning_rate": 4.94847753828529e-07,
      "loss": 7.7912,
      "step": 189
    },
    {
      "epoch": 0.6529209621993127,
      "grad_norm": 3423.37353515625,
      "learning_rate": 4.091608979183303e-07,
      "loss": 8.2337,
      "step": 190
    },
    {
      "epoch": 0.6563573883161512,
      "grad_norm": 2644.664794921875,
      "learning_rate": 3.315638819559452e-07,
      "loss": 7.876,
      "step": 191
    },
    {
      "epoch": 0.6597938144329897,
      "grad_norm": 4483.470703125,
      "learning_rate": 2.6207792019074414e-07,
      "loss": 8.5568,
      "step": 192
    },
    {
      "epoch": 0.6632302405498282,
      "grad_norm": 8579.669921875,
      "learning_rate": 2.0072200939085573e-07,
      "loss": 8.8126,
      "step": 193
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 7409.6220703125,
      "learning_rate": 1.475129236496575e-07,
      "loss": 7.9544,
      "step": 194
    },
    {
      "epoch": 0.6701030927835051,
      "grad_norm": 4703.0517578125,
      "learning_rate": 1.0246520979990459e-07,
      "loss": 8.26,
      "step": 195
    },
    {
      "epoch": 0.6735395189003437,
      "grad_norm": 4779.853515625,
      "learning_rate": 6.559118343676396e-08,
      "loss": 8.2973,
      "step": 196
    },
    {
      "epoch": 0.6769759450171822,
      "grad_norm": 6850.4453125,
      "learning_rate": 3.690092555085789e-08,
      "loss": 8.3535,
      "step": 197
    },
    {
      "epoch": 0.6804123711340206,
      "grad_norm": 5519.74365234375,
      "learning_rate": 1.640227977221853e-08,
      "loss": 8.6194,
      "step": 198
    },
    {
      "epoch": 0.6838487972508591,
      "grad_norm": 7148.4296875,
      "learning_rate": 4.1008502259298755e-09,
      "loss": 10.8181,
      "step": 199
    },
    {
      "epoch": 0.6872852233676976,
      "grad_norm": 3729.072509765625,
      "learning_rate": 0.0,
      "loss": 13.0084,
      "step": 200
    },
    {
      "epoch": 0.6872852233676976,
      "eval_loss": 2.0797932147979736,
      "eval_runtime": 3.4064,
      "eval_samples_per_second": 143.848,
      "eval_steps_per_second": 36.109,
      "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": 4,
        "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": 6628506133856256.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}