{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9943851768669285,
  "eval_steps": 500,
  "global_step": 666,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004491858506457047,
      "grad_norm": 5.744781241940902,
      "learning_rate": 4e-05,
      "loss": 0.8286,
      "step": 1
    },
    {
      "epoch": 0.008983717012914094,
      "grad_norm": 19.505498037626097,
      "learning_rate": 4e-05,
      "loss": 1.1828,
      "step": 2
    },
    {
      "epoch": 0.01347557551937114,
      "grad_norm": 6.070353334110269,
      "learning_rate": 4e-05,
      "loss": 0.9659,
      "step": 3
    },
    {
      "epoch": 0.017967434025828188,
      "grad_norm": 5.468112961485104,
      "learning_rate": 4e-05,
      "loss": 0.8547,
      "step": 4
    },
    {
      "epoch": 0.022459292532285232,
      "grad_norm": 2.1085384841685646,
      "learning_rate": 4e-05,
      "loss": 0.783,
      "step": 5
    },
    {
      "epoch": 0.02695115103874228,
      "grad_norm": 3.446476631557721,
      "learning_rate": 4e-05,
      "loss": 0.7203,
      "step": 6
    },
    {
      "epoch": 0.031443009545199324,
      "grad_norm": 2.238632491312605,
      "learning_rate": 4e-05,
      "loss": 0.6916,
      "step": 7
    },
    {
      "epoch": 0.035934868051656375,
      "grad_norm": 2.571418301311506,
      "learning_rate": 4e-05,
      "loss": 0.6661,
      "step": 8
    },
    {
      "epoch": 0.04042672655811342,
      "grad_norm": 1.5383198728688738,
      "learning_rate": 4e-05,
      "loss": 0.6438,
      "step": 9
    },
    {
      "epoch": 0.044918585064570464,
      "grad_norm": 1.9036968819623286,
      "learning_rate": 4e-05,
      "loss": 0.6306,
      "step": 10
    },
    {
      "epoch": 0.049410443571027515,
      "grad_norm": 1.6141877530301105,
      "learning_rate": 4e-05,
      "loss": 0.6285,
      "step": 11
    },
    {
      "epoch": 0.05390230207748456,
      "grad_norm": 1.4143552939291812,
      "learning_rate": 4e-05,
      "loss": 0.6089,
      "step": 12
    },
    {
      "epoch": 0.058394160583941604,
      "grad_norm": 1.5750614054371128,
      "learning_rate": 4e-05,
      "loss": 0.5951,
      "step": 13
    },
    {
      "epoch": 0.06288601909039865,
      "grad_norm": 1.0413334981594176,
      "learning_rate": 4e-05,
      "loss": 0.5897,
      "step": 14
    },
    {
      "epoch": 0.06737787759685569,
      "grad_norm": 1.2817386337807046,
      "learning_rate": 4e-05,
      "loss": 0.5786,
      "step": 15
    },
    {
      "epoch": 0.07186973610331275,
      "grad_norm": 0.9716086429768103,
      "learning_rate": 4e-05,
      "loss": 0.576,
      "step": 16
    },
    {
      "epoch": 0.0763615946097698,
      "grad_norm": 1.1604927323149636,
      "learning_rate": 4e-05,
      "loss": 0.5839,
      "step": 17
    },
    {
      "epoch": 0.08085345311622684,
      "grad_norm": 1.1266587957324712,
      "learning_rate": 4e-05,
      "loss": 0.5694,
      "step": 18
    },
    {
      "epoch": 0.08534531162268388,
      "grad_norm": 1.0614584514172463,
      "learning_rate": 4e-05,
      "loss": 0.5604,
      "step": 19
    },
    {
      "epoch": 0.08983717012914093,
      "grad_norm": 1.1844349845119013,
      "learning_rate": 4e-05,
      "loss": 0.566,
      "step": 20
    },
    {
      "epoch": 0.09432902863559797,
      "grad_norm": 0.8677961971351155,
      "learning_rate": 4e-05,
      "loss": 0.551,
      "step": 21
    },
    {
      "epoch": 0.09882088714205503,
      "grad_norm": 0.7735015956378571,
      "learning_rate": 4e-05,
      "loss": 0.5487,
      "step": 22
    },
    {
      "epoch": 0.10331274564851207,
      "grad_norm": 0.9225528379919399,
      "learning_rate": 4e-05,
      "loss": 0.5502,
      "step": 23
    },
    {
      "epoch": 0.10780460415496912,
      "grad_norm": 1.39780610045094,
      "learning_rate": 4e-05,
      "loss": 0.5431,
      "step": 24
    },
    {
      "epoch": 0.11229646266142616,
      "grad_norm": 0.6722572434921797,
      "learning_rate": 4e-05,
      "loss": 0.5441,
      "step": 25
    },
    {
      "epoch": 0.11678832116788321,
      "grad_norm": 1.4026250706835734,
      "learning_rate": 4e-05,
      "loss": 0.526,
      "step": 26
    },
    {
      "epoch": 0.12128017967434025,
      "grad_norm": 0.8456481431582107,
      "learning_rate": 4e-05,
      "loss": 0.5373,
      "step": 27
    },
    {
      "epoch": 0.1257720381807973,
      "grad_norm": 0.9930171737711396,
      "learning_rate": 4e-05,
      "loss": 0.536,
      "step": 28
    },
    {
      "epoch": 0.13026389668725435,
      "grad_norm": 1.0525927406993234,
      "learning_rate": 4e-05,
      "loss": 0.5379,
      "step": 29
    },
    {
      "epoch": 0.13475575519371139,
      "grad_norm": 0.8284359012875363,
      "learning_rate": 4e-05,
      "loss": 0.5363,
      "step": 30
    },
    {
      "epoch": 0.13924761370016844,
      "grad_norm": 0.8268233353771701,
      "learning_rate": 4e-05,
      "loss": 0.5372,
      "step": 31
    },
    {
      "epoch": 0.1437394722066255,
      "grad_norm": 0.8877230726819368,
      "learning_rate": 4e-05,
      "loss": 0.5299,
      "step": 32
    },
    {
      "epoch": 0.14823133071308253,
      "grad_norm": 0.8579498414777771,
      "learning_rate": 4e-05,
      "loss": 0.5172,
      "step": 33
    },
    {
      "epoch": 0.1527231892195396,
      "grad_norm": 0.8463267777631354,
      "learning_rate": 4e-05,
      "loss": 0.5248,
      "step": 34
    },
    {
      "epoch": 0.15721504772599662,
      "grad_norm": 0.7816768076849624,
      "learning_rate": 4e-05,
      "loss": 0.5212,
      "step": 35
    },
    {
      "epoch": 0.16170690623245368,
      "grad_norm": 0.7798297675298368,
      "learning_rate": 4e-05,
      "loss": 0.513,
      "step": 36
    },
    {
      "epoch": 0.16619876473891074,
      "grad_norm": 0.8468618470979631,
      "learning_rate": 4e-05,
      "loss": 0.5121,
      "step": 37
    },
    {
      "epoch": 0.17069062324536777,
      "grad_norm": 0.8267671349893145,
      "learning_rate": 4e-05,
      "loss": 0.5199,
      "step": 38
    },
    {
      "epoch": 0.17518248175182483,
      "grad_norm": 0.6860322515619023,
      "learning_rate": 4e-05,
      "loss": 0.5171,
      "step": 39
    },
    {
      "epoch": 0.17967434025828186,
      "grad_norm": 0.9030633176085179,
      "learning_rate": 4e-05,
      "loss": 0.5211,
      "step": 40
    },
    {
      "epoch": 0.18416619876473891,
      "grad_norm": 0.5488539279808397,
      "learning_rate": 4e-05,
      "loss": 0.5186,
      "step": 41
    },
    {
      "epoch": 0.18865805727119594,
      "grad_norm": 0.5773732737635846,
      "learning_rate": 4e-05,
      "loss": 0.5065,
      "step": 42
    },
    {
      "epoch": 0.193149915777653,
      "grad_norm": 0.7589333789633681,
      "learning_rate": 4e-05,
      "loss": 0.5062,
      "step": 43
    },
    {
      "epoch": 0.19764177428411006,
      "grad_norm": 1.2111633655783423,
      "learning_rate": 4e-05,
      "loss": 0.5137,
      "step": 44
    },
    {
      "epoch": 0.2021336327905671,
      "grad_norm": 0.9249209621906633,
      "learning_rate": 4e-05,
      "loss": 0.5011,
      "step": 45
    },
    {
      "epoch": 0.20662549129702415,
      "grad_norm": 0.5565215966506204,
      "learning_rate": 4e-05,
      "loss": 0.5083,
      "step": 46
    },
    {
      "epoch": 0.21111734980348118,
      "grad_norm": 0.5762805653750149,
      "learning_rate": 4e-05,
      "loss": 0.5013,
      "step": 47
    },
    {
      "epoch": 0.21560920830993824,
      "grad_norm": 0.618758493692195,
      "learning_rate": 4e-05,
      "loss": 0.5073,
      "step": 48
    },
    {
      "epoch": 0.2201010668163953,
      "grad_norm": 0.5228632839606239,
      "learning_rate": 4e-05,
      "loss": 0.4943,
      "step": 49
    },
    {
      "epoch": 0.22459292532285233,
      "grad_norm": 0.7019492212411,
      "learning_rate": 4e-05,
      "loss": 0.5158,
      "step": 50
    },
    {
      "epoch": 0.22908478382930939,
      "grad_norm": 0.7048591666385562,
      "learning_rate": 4e-05,
      "loss": 0.5006,
      "step": 51
    },
    {
      "epoch": 0.23357664233576642,
      "grad_norm": 1.0810927518154445,
      "learning_rate": 4e-05,
      "loss": 0.5066,
      "step": 52
    },
    {
      "epoch": 0.23806850084222347,
      "grad_norm": 1.1704953957598856,
      "learning_rate": 4e-05,
      "loss": 0.5007,
      "step": 53
    },
    {
      "epoch": 0.2425603593486805,
      "grad_norm": 0.5828568248897151,
      "learning_rate": 4e-05,
      "loss": 0.5045,
      "step": 54
    },
    {
      "epoch": 0.24705221785513756,
      "grad_norm": 0.49535386557755867,
      "learning_rate": 4e-05,
      "loss": 0.4837,
      "step": 55
    },
    {
      "epoch": 0.2515440763615946,
      "grad_norm": 0.8745705539662333,
      "learning_rate": 4e-05,
      "loss": 0.4989,
      "step": 56
    },
    {
      "epoch": 0.25603593486805165,
      "grad_norm": 1.3873723619178653,
      "learning_rate": 4e-05,
      "loss": 0.4982,
      "step": 57
    },
    {
      "epoch": 0.2605277933745087,
      "grad_norm": 0.802886260074987,
      "learning_rate": 4e-05,
      "loss": 0.4893,
      "step": 58
    },
    {
      "epoch": 0.26501965188096577,
      "grad_norm": 0.868266747892591,
      "learning_rate": 4e-05,
      "loss": 0.5043,
      "step": 59
    },
    {
      "epoch": 0.26951151038742277,
      "grad_norm": 1.3849995126640307,
      "learning_rate": 4e-05,
      "loss": 0.4973,
      "step": 60
    },
    {
      "epoch": 0.27400336889387983,
      "grad_norm": 0.7933512431625963,
      "learning_rate": 4e-05,
      "loss": 0.4885,
      "step": 61
    },
    {
      "epoch": 0.2784952274003369,
      "grad_norm": 0.821749600355561,
      "learning_rate": 4e-05,
      "loss": 0.4952,
      "step": 62
    },
    {
      "epoch": 0.28298708590679394,
      "grad_norm": 0.8605698822435636,
      "learning_rate": 4e-05,
      "loss": 0.5057,
      "step": 63
    },
    {
      "epoch": 0.287478944413251,
      "grad_norm": 0.911415814643325,
      "learning_rate": 4e-05,
      "loss": 0.4914,
      "step": 64
    },
    {
      "epoch": 0.291970802919708,
      "grad_norm": 0.8945457591090099,
      "learning_rate": 4e-05,
      "loss": 0.4952,
      "step": 65
    },
    {
      "epoch": 0.29646266142616506,
      "grad_norm": 0.9470945660790681,
      "learning_rate": 4e-05,
      "loss": 0.4822,
      "step": 66
    },
    {
      "epoch": 0.3009545199326221,
      "grad_norm": 1.141332334299757,
      "learning_rate": 4e-05,
      "loss": 0.4953,
      "step": 67
    },
    {
      "epoch": 0.3054463784390792,
      "grad_norm": 0.5927788091774613,
      "learning_rate": 4e-05,
      "loss": 0.4823,
      "step": 68
    },
    {
      "epoch": 0.30993823694553624,
      "grad_norm": 1.1208197341053938,
      "learning_rate": 4e-05,
      "loss": 0.491,
      "step": 69
    },
    {
      "epoch": 0.31443009545199324,
      "grad_norm": 0.7484424664511564,
      "learning_rate": 4e-05,
      "loss": 0.486,
      "step": 70
    },
    {
      "epoch": 0.3189219539584503,
      "grad_norm": 0.7198849807955354,
      "learning_rate": 4e-05,
      "loss": 0.486,
      "step": 71
    },
    {
      "epoch": 0.32341381246490736,
      "grad_norm": 0.8171556771898378,
      "learning_rate": 4e-05,
      "loss": 0.5018,
      "step": 72
    },
    {
      "epoch": 0.3279056709713644,
      "grad_norm": 0.7321498711363629,
      "learning_rate": 4e-05,
      "loss": 0.4878,
      "step": 73
    },
    {
      "epoch": 0.3323975294778215,
      "grad_norm": 1.2175783198978702,
      "learning_rate": 4e-05,
      "loss": 0.4915,
      "step": 74
    },
    {
      "epoch": 0.3368893879842785,
      "grad_norm": 1.0028811052364257,
      "learning_rate": 4e-05,
      "loss": 0.494,
      "step": 75
    },
    {
      "epoch": 0.34138124649073553,
      "grad_norm": 0.6633234889481054,
      "learning_rate": 4e-05,
      "loss": 0.4833,
      "step": 76
    },
    {
      "epoch": 0.3458731049971926,
      "grad_norm": 0.7686382161519945,
      "learning_rate": 4e-05,
      "loss": 0.4962,
      "step": 77
    },
    {
      "epoch": 0.35036496350364965,
      "grad_norm": 1.121541318437043,
      "learning_rate": 4e-05,
      "loss": 0.4783,
      "step": 78
    },
    {
      "epoch": 0.35485682201010665,
      "grad_norm": 0.7851132535171547,
      "learning_rate": 4e-05,
      "loss": 0.49,
      "step": 79
    },
    {
      "epoch": 0.3593486805165637,
      "grad_norm": 0.8702836918121081,
      "learning_rate": 4e-05,
      "loss": 0.4916,
      "step": 80
    },
    {
      "epoch": 0.36384053902302077,
      "grad_norm": 1.0654983424826385,
      "learning_rate": 4e-05,
      "loss": 0.4921,
      "step": 81
    },
    {
      "epoch": 0.36833239752947783,
      "grad_norm": 0.6838772355526056,
      "learning_rate": 4e-05,
      "loss": 0.4951,
      "step": 82
    },
    {
      "epoch": 0.3728242560359349,
      "grad_norm": 0.5450715252765912,
      "learning_rate": 4e-05,
      "loss": 0.4869,
      "step": 83
    },
    {
      "epoch": 0.3773161145423919,
      "grad_norm": 0.7882292503623511,
      "learning_rate": 4e-05,
      "loss": 0.4862,
      "step": 84
    },
    {
      "epoch": 0.38180797304884895,
      "grad_norm": 0.8399149392888215,
      "learning_rate": 4e-05,
      "loss": 0.4899,
      "step": 85
    },
    {
      "epoch": 0.386299831555306,
      "grad_norm": 0.8104469928788055,
      "learning_rate": 4e-05,
      "loss": 0.4729,
      "step": 86
    },
    {
      "epoch": 0.39079169006176306,
      "grad_norm": 0.786503783799548,
      "learning_rate": 4e-05,
      "loss": 0.4723,
      "step": 87
    },
    {
      "epoch": 0.3952835485682201,
      "grad_norm": 0.6810115594198747,
      "learning_rate": 4e-05,
      "loss": 0.4773,
      "step": 88
    },
    {
      "epoch": 0.3997754070746771,
      "grad_norm": 0.655336062155516,
      "learning_rate": 4e-05,
      "loss": 0.4864,
      "step": 89
    },
    {
      "epoch": 0.4042672655811342,
      "grad_norm": 0.643705881932075,
      "learning_rate": 4e-05,
      "loss": 0.484,
      "step": 90
    },
    {
      "epoch": 0.40875912408759124,
      "grad_norm": 0.43126942074014457,
      "learning_rate": 4e-05,
      "loss": 0.4845,
      "step": 91
    },
    {
      "epoch": 0.4132509825940483,
      "grad_norm": 0.6313212298730237,
      "learning_rate": 4e-05,
      "loss": 0.4762,
      "step": 92
    },
    {
      "epoch": 0.41774284110050536,
      "grad_norm": 0.5013582116787559,
      "learning_rate": 4e-05,
      "loss": 0.4785,
      "step": 93
    },
    {
      "epoch": 0.42223469960696236,
      "grad_norm": 0.5530967702907004,
      "learning_rate": 4e-05,
      "loss": 0.4858,
      "step": 94
    },
    {
      "epoch": 0.4267265581134194,
      "grad_norm": 0.7626916934517558,
      "learning_rate": 4e-05,
      "loss": 0.4853,
      "step": 95
    },
    {
      "epoch": 0.4312184166198765,
      "grad_norm": 0.9503864921122651,
      "learning_rate": 4e-05,
      "loss": 0.4675,
      "step": 96
    },
    {
      "epoch": 0.43571027512633353,
      "grad_norm": 1.0773660088775425,
      "learning_rate": 4e-05,
      "loss": 0.487,
      "step": 97
    },
    {
      "epoch": 0.4402021336327906,
      "grad_norm": 0.8726741655437852,
      "learning_rate": 4e-05,
      "loss": 0.481,
      "step": 98
    },
    {
      "epoch": 0.4446939921392476,
      "grad_norm": 1.021704071734518,
      "learning_rate": 4e-05,
      "loss": 0.4602,
      "step": 99
    },
    {
      "epoch": 0.44918585064570465,
      "grad_norm": 0.956327119910157,
      "learning_rate": 4e-05,
      "loss": 0.471,
      "step": 100
    },
    {
      "epoch": 0.4536777091521617,
      "grad_norm": 0.6563826653084724,
      "learning_rate": 4e-05,
      "loss": 0.4832,
      "step": 101
    },
    {
      "epoch": 0.45816956765861877,
      "grad_norm": 0.7010259922738414,
      "learning_rate": 4e-05,
      "loss": 0.4793,
      "step": 102
    },
    {
      "epoch": 0.4626614261650758,
      "grad_norm": 0.7440214322577411,
      "learning_rate": 4e-05,
      "loss": 0.4797,
      "step": 103
    },
    {
      "epoch": 0.46715328467153283,
      "grad_norm": 0.88720325020729,
      "learning_rate": 4e-05,
      "loss": 0.4699,
      "step": 104
    },
    {
      "epoch": 0.4716451431779899,
      "grad_norm": 1.233382631804206,
      "learning_rate": 4e-05,
      "loss": 0.4672,
      "step": 105
    },
    {
      "epoch": 0.47613700168444695,
      "grad_norm": 0.7317344289397154,
      "learning_rate": 4e-05,
      "loss": 0.4735,
      "step": 106
    },
    {
      "epoch": 0.480628860190904,
      "grad_norm": 0.6532990606262099,
      "learning_rate": 4e-05,
      "loss": 0.4723,
      "step": 107
    },
    {
      "epoch": 0.485120718697361,
      "grad_norm": 0.6915122089199583,
      "learning_rate": 4e-05,
      "loss": 0.4832,
      "step": 108
    },
    {
      "epoch": 0.48961257720381807,
      "grad_norm": 0.8116242277752135,
      "learning_rate": 4e-05,
      "loss": 0.483,
      "step": 109
    },
    {
      "epoch": 0.4941044357102751,
      "grad_norm": 0.7549235192200487,
      "learning_rate": 4e-05,
      "loss": 0.4769,
      "step": 110
    },
    {
      "epoch": 0.4985962942167322,
      "grad_norm": 0.7547758862775902,
      "learning_rate": 4e-05,
      "loss": 0.4783,
      "step": 111
    },
    {
      "epoch": 0.5030881527231892,
      "grad_norm": 0.9140526922250276,
      "learning_rate": 4e-05,
      "loss": 0.4686,
      "step": 112
    },
    {
      "epoch": 0.5075800112296462,
      "grad_norm": 0.8724131718569759,
      "learning_rate": 4e-05,
      "loss": 0.4848,
      "step": 113
    },
    {
      "epoch": 0.5120718697361033,
      "grad_norm": 0.763556479522536,
      "learning_rate": 4e-05,
      "loss": 0.479,
      "step": 114
    },
    {
      "epoch": 0.5165637282425604,
      "grad_norm": 0.690090038210421,
      "learning_rate": 4e-05,
      "loss": 0.4662,
      "step": 115
    },
    {
      "epoch": 0.5210555867490174,
      "grad_norm": 0.5658638909252569,
      "learning_rate": 4e-05,
      "loss": 0.4861,
      "step": 116
    },
    {
      "epoch": 0.5255474452554745,
      "grad_norm": 0.4967500611059774,
      "learning_rate": 4e-05,
      "loss": 0.4718,
      "step": 117
    },
    {
      "epoch": 0.5300393037619315,
      "grad_norm": 0.4268932003542312,
      "learning_rate": 4e-05,
      "loss": 0.4783,
      "step": 118
    },
    {
      "epoch": 0.5345311622683886,
      "grad_norm": 0.4961761307046008,
      "learning_rate": 4e-05,
      "loss": 0.471,
      "step": 119
    },
    {
      "epoch": 0.5390230207748455,
      "grad_norm": 0.5929083615034594,
      "learning_rate": 4e-05,
      "loss": 0.4718,
      "step": 120
    },
    {
      "epoch": 0.5435148792813026,
      "grad_norm": 0.720442042366462,
      "learning_rate": 4e-05,
      "loss": 0.4796,
      "step": 121
    },
    {
      "epoch": 0.5480067377877597,
      "grad_norm": 0.7425929255776511,
      "learning_rate": 4e-05,
      "loss": 0.4715,
      "step": 122
    },
    {
      "epoch": 0.5524985962942167,
      "grad_norm": 0.6442984183770077,
      "learning_rate": 4e-05,
      "loss": 0.4554,
      "step": 123
    },
    {
      "epoch": 0.5569904548006738,
      "grad_norm": 0.5832639730755836,
      "learning_rate": 4e-05,
      "loss": 0.4827,
      "step": 124
    },
    {
      "epoch": 0.5614823133071308,
      "grad_norm": 0.7079302759728495,
      "learning_rate": 4e-05,
      "loss": 0.4836,
      "step": 125
    },
    {
      "epoch": 0.5659741718135879,
      "grad_norm": 0.5538421407332266,
      "learning_rate": 4e-05,
      "loss": 0.4622,
      "step": 126
    },
    {
      "epoch": 0.570466030320045,
      "grad_norm": 0.4449970363159924,
      "learning_rate": 4e-05,
      "loss": 0.4639,
      "step": 127
    },
    {
      "epoch": 0.574957888826502,
      "grad_norm": 0.5183308175456901,
      "learning_rate": 4e-05,
      "loss": 0.4683,
      "step": 128
    },
    {
      "epoch": 0.5794497473329591,
      "grad_norm": 0.5881505553174785,
      "learning_rate": 4e-05,
      "loss": 0.4739,
      "step": 129
    },
    {
      "epoch": 0.583941605839416,
      "grad_norm": 0.4955515258075932,
      "learning_rate": 4e-05,
      "loss": 0.4594,
      "step": 130
    },
    {
      "epoch": 0.5884334643458731,
      "grad_norm": 0.46521840912965134,
      "learning_rate": 4e-05,
      "loss": 0.4742,
      "step": 131
    },
    {
      "epoch": 0.5929253228523301,
      "grad_norm": 0.5486407755832184,
      "learning_rate": 4e-05,
      "loss": 0.4576,
      "step": 132
    },
    {
      "epoch": 0.5974171813587872,
      "grad_norm": 0.5475916676615601,
      "learning_rate": 4e-05,
      "loss": 0.4636,
      "step": 133
    },
    {
      "epoch": 0.6019090398652442,
      "grad_norm": 0.5882929789659717,
      "learning_rate": 4e-05,
      "loss": 0.4701,
      "step": 134
    },
    {
      "epoch": 0.6064008983717013,
      "grad_norm": 0.7123845942758071,
      "learning_rate": 4e-05,
      "loss": 0.4705,
      "step": 135
    },
    {
      "epoch": 0.6108927568781584,
      "grad_norm": 0.8358345161218314,
      "learning_rate": 4e-05,
      "loss": 0.4742,
      "step": 136
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 0.8111728940958627,
      "learning_rate": 4e-05,
      "loss": 0.4689,
      "step": 137
    },
    {
      "epoch": 0.6198764738910725,
      "grad_norm": 0.8737394790926428,
      "learning_rate": 4e-05,
      "loss": 0.4665,
      "step": 138
    },
    {
      "epoch": 0.6243683323975294,
      "grad_norm": 1.169818588236249,
      "learning_rate": 4e-05,
      "loss": 0.4741,
      "step": 139
    },
    {
      "epoch": 0.6288601909039865,
      "grad_norm": 0.9021740823310876,
      "learning_rate": 4e-05,
      "loss": 0.4633,
      "step": 140
    },
    {
      "epoch": 0.6333520494104435,
      "grad_norm": 0.78603189466474,
      "learning_rate": 4e-05,
      "loss": 0.4738,
      "step": 141
    },
    {
      "epoch": 0.6378439079169006,
      "grad_norm": 1.121182356081946,
      "learning_rate": 4e-05,
      "loss": 0.4665,
      "step": 142
    },
    {
      "epoch": 0.6423357664233577,
      "grad_norm": 0.8666919602803608,
      "learning_rate": 4e-05,
      "loss": 0.4581,
      "step": 143
    },
    {
      "epoch": 0.6468276249298147,
      "grad_norm": 0.6035212308601808,
      "learning_rate": 4e-05,
      "loss": 0.4586,
      "step": 144
    },
    {
      "epoch": 0.6513194834362718,
      "grad_norm": 0.7205923360001546,
      "learning_rate": 4e-05,
      "loss": 0.4758,
      "step": 145
    },
    {
      "epoch": 0.6558113419427288,
      "grad_norm": 0.775262754139686,
      "learning_rate": 4e-05,
      "loss": 0.47,
      "step": 146
    },
    {
      "epoch": 0.6603032004491859,
      "grad_norm": 0.741054429383634,
      "learning_rate": 4e-05,
      "loss": 0.467,
      "step": 147
    },
    {
      "epoch": 0.664795058955643,
      "grad_norm": 0.9677391733410978,
      "learning_rate": 4e-05,
      "loss": 0.4667,
      "step": 148
    },
    {
      "epoch": 0.6692869174620999,
      "grad_norm": 0.8740978793386832,
      "learning_rate": 4e-05,
      "loss": 0.4699,
      "step": 149
    },
    {
      "epoch": 0.673778775968557,
      "grad_norm": 0.5123947626090699,
      "learning_rate": 4e-05,
      "loss": 0.4615,
      "step": 150
    },
    {
      "epoch": 0.678270634475014,
      "grad_norm": 1.0110045676882082,
      "learning_rate": 4e-05,
      "loss": 0.4774,
      "step": 151
    },
    {
      "epoch": 0.6827624929814711,
      "grad_norm": 1.0658702971598875,
      "learning_rate": 4e-05,
      "loss": 0.4844,
      "step": 152
    },
    {
      "epoch": 0.6872543514879281,
      "grad_norm": 0.4495371315120995,
      "learning_rate": 4e-05,
      "loss": 0.47,
      "step": 153
    },
    {
      "epoch": 0.6917462099943852,
      "grad_norm": 0.7016570929728161,
      "learning_rate": 4e-05,
      "loss": 0.4642,
      "step": 154
    },
    {
      "epoch": 0.6962380685008422,
      "grad_norm": 0.6775824457187086,
      "learning_rate": 4e-05,
      "loss": 0.4689,
      "step": 155
    },
    {
      "epoch": 0.7007299270072993,
      "grad_norm": 0.452549912045293,
      "learning_rate": 4e-05,
      "loss": 0.4662,
      "step": 156
    },
    {
      "epoch": 0.7052217855137564,
      "grad_norm": 0.5914173735385079,
      "learning_rate": 4e-05,
      "loss": 0.459,
      "step": 157
    },
    {
      "epoch": 0.7097136440202133,
      "grad_norm": 0.5605586838803835,
      "learning_rate": 4e-05,
      "loss": 0.4649,
      "step": 158
    },
    {
      "epoch": 0.7142055025266704,
      "grad_norm": 0.6628213547699484,
      "learning_rate": 4e-05,
      "loss": 0.4776,
      "step": 159
    },
    {
      "epoch": 0.7186973610331274,
      "grad_norm": 0.7204819349880731,
      "learning_rate": 4e-05,
      "loss": 0.4717,
      "step": 160
    },
    {
      "epoch": 0.7231892195395845,
      "grad_norm": 0.61438257614997,
      "learning_rate": 4e-05,
      "loss": 0.4604,
      "step": 161
    },
    {
      "epoch": 0.7276810780460415,
      "grad_norm": 0.5198941680539216,
      "learning_rate": 4e-05,
      "loss": 0.4627,
      "step": 162
    },
    {
      "epoch": 0.7321729365524986,
      "grad_norm": 0.4598102713775601,
      "learning_rate": 4e-05,
      "loss": 0.4465,
      "step": 163
    },
    {
      "epoch": 0.7366647950589557,
      "grad_norm": 0.5700664637785826,
      "learning_rate": 4e-05,
      "loss": 0.4522,
      "step": 164
    },
    {
      "epoch": 0.7411566535654127,
      "grad_norm": 0.5939196641822652,
      "learning_rate": 4e-05,
      "loss": 0.4684,
      "step": 165
    },
    {
      "epoch": 0.7456485120718698,
      "grad_norm": 0.8428789042093132,
      "learning_rate": 4e-05,
      "loss": 0.4689,
      "step": 166
    },
    {
      "epoch": 0.7501403705783268,
      "grad_norm": 1.2497594176367328,
      "learning_rate": 4e-05,
      "loss": 0.4741,
      "step": 167
    },
    {
      "epoch": 0.7546322290847838,
      "grad_norm": 0.633381196128311,
      "learning_rate": 4e-05,
      "loss": 0.4571,
      "step": 168
    },
    {
      "epoch": 0.7591240875912408,
      "grad_norm": 0.6874907087362482,
      "learning_rate": 4e-05,
      "loss": 0.4625,
      "step": 169
    },
    {
      "epoch": 0.7636159460976979,
      "grad_norm": 0.6179181500050008,
      "learning_rate": 4e-05,
      "loss": 0.4613,
      "step": 170
    },
    {
      "epoch": 0.768107804604155,
      "grad_norm": 0.7341308022124572,
      "learning_rate": 4e-05,
      "loss": 0.4689,
      "step": 171
    },
    {
      "epoch": 0.772599663110612,
      "grad_norm": 0.9447484619669985,
      "learning_rate": 4e-05,
      "loss": 0.4605,
      "step": 172
    },
    {
      "epoch": 0.7770915216170691,
      "grad_norm": 0.8954773191151448,
      "learning_rate": 4e-05,
      "loss": 0.4512,
      "step": 173
    },
    {
      "epoch": 0.7815833801235261,
      "grad_norm": 1.093005343712751,
      "learning_rate": 4e-05,
      "loss": 0.4659,
      "step": 174
    },
    {
      "epoch": 0.7860752386299832,
      "grad_norm": 0.8179940618852544,
      "learning_rate": 4e-05,
      "loss": 0.4686,
      "step": 175
    },
    {
      "epoch": 0.7905670971364402,
      "grad_norm": 0.4614130371065188,
      "learning_rate": 4e-05,
      "loss": 0.4671,
      "step": 176
    },
    {
      "epoch": 0.7950589556428973,
      "grad_norm": 0.5905315826579326,
      "learning_rate": 4e-05,
      "loss": 0.448,
      "step": 177
    },
    {
      "epoch": 0.7995508141493542,
      "grad_norm": 0.687134068257525,
      "learning_rate": 4e-05,
      "loss": 0.4562,
      "step": 178
    },
    {
      "epoch": 0.8040426726558113,
      "grad_norm": 0.5340584555365142,
      "learning_rate": 4e-05,
      "loss": 0.4601,
      "step": 179
    },
    {
      "epoch": 0.8085345311622684,
      "grad_norm": 0.40415584073843613,
      "learning_rate": 4e-05,
      "loss": 0.4652,
      "step": 180
    },
    {
      "epoch": 0.8130263896687254,
      "grad_norm": 0.40922076778892336,
      "learning_rate": 4e-05,
      "loss": 0.4733,
      "step": 181
    },
    {
      "epoch": 0.8175182481751825,
      "grad_norm": 0.3871061243116146,
      "learning_rate": 4e-05,
      "loss": 0.4559,
      "step": 182
    },
    {
      "epoch": 0.8220101066816395,
      "grad_norm": 0.49008015607191197,
      "learning_rate": 4e-05,
      "loss": 0.4618,
      "step": 183
    },
    {
      "epoch": 0.8265019651880966,
      "grad_norm": 0.5187486486772962,
      "learning_rate": 4e-05,
      "loss": 0.454,
      "step": 184
    },
    {
      "epoch": 0.8309938236945537,
      "grad_norm": 0.5382429601293272,
      "learning_rate": 4e-05,
      "loss": 0.4535,
      "step": 185
    },
    {
      "epoch": 0.8354856822010107,
      "grad_norm": 0.4714530439381728,
      "learning_rate": 4e-05,
      "loss": 0.4587,
      "step": 186
    },
    {
      "epoch": 0.8399775407074677,
      "grad_norm": 0.5014643025897988,
      "learning_rate": 4e-05,
      "loss": 0.4614,
      "step": 187
    },
    {
      "epoch": 0.8444693992139247,
      "grad_norm": 0.7040028866670897,
      "learning_rate": 4e-05,
      "loss": 0.453,
      "step": 188
    },
    {
      "epoch": 0.8489612577203818,
      "grad_norm": 0.7025567080097845,
      "learning_rate": 4e-05,
      "loss": 0.454,
      "step": 189
    },
    {
      "epoch": 0.8534531162268388,
      "grad_norm": 0.6940845302575863,
      "learning_rate": 4e-05,
      "loss": 0.4685,
      "step": 190
    },
    {
      "epoch": 0.8579449747332959,
      "grad_norm": 0.6888968080405814,
      "learning_rate": 4e-05,
      "loss": 0.4689,
      "step": 191
    },
    {
      "epoch": 0.862436833239753,
      "grad_norm": 0.6769620719610056,
      "learning_rate": 4e-05,
      "loss": 0.4655,
      "step": 192
    },
    {
      "epoch": 0.86692869174621,
      "grad_norm": 0.8051053387958492,
      "learning_rate": 4e-05,
      "loss": 0.4615,
      "step": 193
    },
    {
      "epoch": 0.8714205502526671,
      "grad_norm": 1.028641450260458,
      "learning_rate": 4e-05,
      "loss": 0.4729,
      "step": 194
    },
    {
      "epoch": 0.8759124087591241,
      "grad_norm": 1.1069726931738983,
      "learning_rate": 4e-05,
      "loss": 0.4486,
      "step": 195
    },
    {
      "epoch": 0.8804042672655812,
      "grad_norm": 0.7717753364463334,
      "learning_rate": 4e-05,
      "loss": 0.4517,
      "step": 196
    },
    {
      "epoch": 0.8848961257720381,
      "grad_norm": 0.7156471809354176,
      "learning_rate": 4e-05,
      "loss": 0.4595,
      "step": 197
    },
    {
      "epoch": 0.8893879842784952,
      "grad_norm": 0.8102351330882908,
      "learning_rate": 4e-05,
      "loss": 0.4676,
      "step": 198
    },
    {
      "epoch": 0.8938798427849522,
      "grad_norm": 0.6744340740980704,
      "learning_rate": 4e-05,
      "loss": 0.4579,
      "step": 199
    },
    {
      "epoch": 0.8983717012914093,
      "grad_norm": 0.5796572051869849,
      "learning_rate": 4e-05,
      "loss": 0.445,
      "step": 200
    },
    {
      "epoch": 0.9028635597978664,
      "grad_norm": 0.8595562833407072,
      "learning_rate": 4e-05,
      "loss": 0.4574,
      "step": 201
    },
    {
      "epoch": 0.9073554183043234,
      "grad_norm": 0.8640330914817986,
      "learning_rate": 4e-05,
      "loss": 0.4734,
      "step": 202
    },
    {
      "epoch": 0.9118472768107805,
      "grad_norm": 0.7897305589178268,
      "learning_rate": 4e-05,
      "loss": 0.4523,
      "step": 203
    },
    {
      "epoch": 0.9163391353172375,
      "grad_norm": 0.6092847549633084,
      "learning_rate": 4e-05,
      "loss": 0.4521,
      "step": 204
    },
    {
      "epoch": 0.9208309938236946,
      "grad_norm": 0.6469620658263115,
      "learning_rate": 4e-05,
      "loss": 0.462,
      "step": 205
    },
    {
      "epoch": 0.9253228523301515,
      "grad_norm": 0.75489251931131,
      "learning_rate": 4e-05,
      "loss": 0.4673,
      "step": 206
    },
    {
      "epoch": 0.9298147108366086,
      "grad_norm": 0.626362559467887,
      "learning_rate": 4e-05,
      "loss": 0.4554,
      "step": 207
    },
    {
      "epoch": 0.9343065693430657,
      "grad_norm": 0.4969098225428687,
      "learning_rate": 4e-05,
      "loss": 0.4528,
      "step": 208
    },
    {
      "epoch": 0.9387984278495227,
      "grad_norm": 0.6298837222109367,
      "learning_rate": 4e-05,
      "loss": 0.4632,
      "step": 209
    },
    {
      "epoch": 0.9432902863559798,
      "grad_norm": 0.798659528060675,
      "learning_rate": 4e-05,
      "loss": 0.4664,
      "step": 210
    },
    {
      "epoch": 0.9477821448624368,
      "grad_norm": 0.7695772110246359,
      "learning_rate": 4e-05,
      "loss": 0.4565,
      "step": 211
    },
    {
      "epoch": 0.9522740033688939,
      "grad_norm": 0.6089136922828309,
      "learning_rate": 4e-05,
      "loss": 0.4501,
      "step": 212
    },
    {
      "epoch": 0.956765861875351,
      "grad_norm": 0.8462642791869017,
      "learning_rate": 4e-05,
      "loss": 0.4498,
      "step": 213
    },
    {
      "epoch": 0.961257720381808,
      "grad_norm": 0.742195679410513,
      "learning_rate": 4e-05,
      "loss": 0.4609,
      "step": 214
    },
    {
      "epoch": 0.9657495788882651,
      "grad_norm": 0.8339350657097956,
      "learning_rate": 4e-05,
      "loss": 0.4392,
      "step": 215
    },
    {
      "epoch": 0.970241437394722,
      "grad_norm": 1.0412421614509542,
      "learning_rate": 4e-05,
      "loss": 0.4559,
      "step": 216
    },
    {
      "epoch": 0.9747332959011791,
      "grad_norm": 0.7416785850604267,
      "learning_rate": 4e-05,
      "loss": 0.4513,
      "step": 217
    },
    {
      "epoch": 0.9792251544076361,
      "grad_norm": 0.4419843388051339,
      "learning_rate": 4e-05,
      "loss": 0.4571,
      "step": 218
    },
    {
      "epoch": 0.9837170129140932,
      "grad_norm": 0.5549728329047522,
      "learning_rate": 4e-05,
      "loss": 0.4561,
      "step": 219
    },
    {
      "epoch": 0.9882088714205502,
      "grad_norm": 0.5569110222151862,
      "learning_rate": 4e-05,
      "loss": 0.4699,
      "step": 220
    },
    {
      "epoch": 0.9927007299270073,
      "grad_norm": 0.3777694667085498,
      "learning_rate": 4e-05,
      "loss": 0.4571,
      "step": 221
    },
    {
      "epoch": 0.9971925884334644,
      "grad_norm": 0.4899375061696797,
      "learning_rate": 4e-05,
      "loss": 0.4555,
      "step": 222
    },
    {
      "epoch": 1.0030881527231892,
      "grad_norm": 0.734091692716981,
      "learning_rate": 4e-05,
      "loss": 0.73,
      "step": 223
    },
    {
      "epoch": 1.0075800112296462,
      "grad_norm": 1.039004916800266,
      "learning_rate": 4e-05,
      "loss": 0.4365,
      "step": 224
    },
    {
      "epoch": 1.0120718697361033,
      "grad_norm": 1.4775793758217903,
      "learning_rate": 4e-05,
      "loss": 0.414,
      "step": 225
    },
    {
      "epoch": 1.0165637282425604,
      "grad_norm": 0.7594021952977889,
      "learning_rate": 4e-05,
      "loss": 0.4201,
      "step": 226
    },
    {
      "epoch": 1.0210555867490174,
      "grad_norm": 1.672592418338108,
      "learning_rate": 4e-05,
      "loss": 0.4236,
      "step": 227
    },
    {
      "epoch": 1.0255474452554745,
      "grad_norm": 0.8815121169077754,
      "learning_rate": 4e-05,
      "loss": 0.4405,
      "step": 228
    },
    {
      "epoch": 1.0300393037619315,
      "grad_norm": 1.5620999986447677,
      "learning_rate": 4e-05,
      "loss": 0.4279,
      "step": 229
    },
    {
      "epoch": 1.0345311622683886,
      "grad_norm": 1.1398064434513893,
      "learning_rate": 4e-05,
      "loss": 0.4433,
      "step": 230
    },
    {
      "epoch": 1.0390230207748457,
      "grad_norm": 1.6074086336723372,
      "learning_rate": 4e-05,
      "loss": 0.4371,
      "step": 231
    },
    {
      "epoch": 1.0435148792813027,
      "grad_norm": 1.392278654528992,
      "learning_rate": 4e-05,
      "loss": 0.4534,
      "step": 232
    },
    {
      "epoch": 1.0480067377877598,
      "grad_norm": 0.9943738868623905,
      "learning_rate": 4e-05,
      "loss": 0.4279,
      "step": 233
    },
    {
      "epoch": 1.0524985962942168,
      "grad_norm": 0.909374767430801,
      "learning_rate": 4e-05,
      "loss": 0.4299,
      "step": 234
    },
    {
      "epoch": 1.0569904548006739,
      "grad_norm": 0.8243691433987034,
      "learning_rate": 4e-05,
      "loss": 0.4348,
      "step": 235
    },
    {
      "epoch": 1.0614823133071307,
      "grad_norm": 0.8647530310523033,
      "learning_rate": 4e-05,
      "loss": 0.4266,
      "step": 236
    },
    {
      "epoch": 1.0659741718135878,
      "grad_norm": 0.7347921921688428,
      "learning_rate": 4e-05,
      "loss": 0.4123,
      "step": 237
    },
    {
      "epoch": 1.0704660303200448,
      "grad_norm": 0.8226947057061942,
      "learning_rate": 4e-05,
      "loss": 0.4298,
      "step": 238
    },
    {
      "epoch": 1.074957888826502,
      "grad_norm": 0.7022283311703869,
      "learning_rate": 4e-05,
      "loss": 0.4226,
      "step": 239
    },
    {
      "epoch": 1.079449747332959,
      "grad_norm": 0.6911144716630917,
      "learning_rate": 4e-05,
      "loss": 0.4157,
      "step": 240
    },
    {
      "epoch": 1.083941605839416,
      "grad_norm": 0.7507173473567265,
      "learning_rate": 4e-05,
      "loss": 0.419,
      "step": 241
    },
    {
      "epoch": 1.088433464345873,
      "grad_norm": 0.5588078415587642,
      "learning_rate": 4e-05,
      "loss": 0.4304,
      "step": 242
    },
    {
      "epoch": 1.0929253228523301,
      "grad_norm": 0.5354393113569151,
      "learning_rate": 4e-05,
      "loss": 0.4219,
      "step": 243
    },
    {
      "epoch": 1.0974171813587872,
      "grad_norm": 0.6053169619440386,
      "learning_rate": 4e-05,
      "loss": 0.4108,
      "step": 244
    },
    {
      "epoch": 1.1019090398652442,
      "grad_norm": 0.539163056501405,
      "learning_rate": 4e-05,
      "loss": 0.4193,
      "step": 245
    },
    {
      "epoch": 1.1064008983717013,
      "grad_norm": 0.5643849629135265,
      "learning_rate": 4e-05,
      "loss": 0.4423,
      "step": 246
    },
    {
      "epoch": 1.1108927568781584,
      "grad_norm": 0.5784006424125349,
      "learning_rate": 4e-05,
      "loss": 0.4045,
      "step": 247
    },
    {
      "epoch": 1.1153846153846154,
      "grad_norm": 0.5523140938972508,
      "learning_rate": 4e-05,
      "loss": 0.4205,
      "step": 248
    },
    {
      "epoch": 1.1198764738910725,
      "grad_norm": 0.5276860866313803,
      "learning_rate": 4e-05,
      "loss": 0.4244,
      "step": 249
    },
    {
      "epoch": 1.1243683323975295,
      "grad_norm": 0.5365506074726554,
      "learning_rate": 4e-05,
      "loss": 0.4418,
      "step": 250
    },
    {
      "epoch": 1.1288601909039866,
      "grad_norm": 0.4823955956060016,
      "learning_rate": 4e-05,
      "loss": 0.4202,
      "step": 251
    },
    {
      "epoch": 1.1333520494104437,
      "grad_norm": 0.46273764286196334,
      "learning_rate": 4e-05,
      "loss": 0.3969,
      "step": 252
    },
    {
      "epoch": 1.1378439079169007,
      "grad_norm": 0.45845204095899983,
      "learning_rate": 4e-05,
      "loss": 0.4403,
      "step": 253
    },
    {
      "epoch": 1.1423357664233578,
      "grad_norm": 0.47826341004954265,
      "learning_rate": 4e-05,
      "loss": 0.4048,
      "step": 254
    },
    {
      "epoch": 1.1468276249298146,
      "grad_norm": 0.49950178611444324,
      "learning_rate": 4e-05,
      "loss": 0.433,
      "step": 255
    },
    {
      "epoch": 1.1513194834362717,
      "grad_norm": 0.43653485928074476,
      "learning_rate": 4e-05,
      "loss": 0.4022,
      "step": 256
    },
    {
      "epoch": 1.1558113419427287,
      "grad_norm": 0.425511772659085,
      "learning_rate": 4e-05,
      "loss": 0.4349,
      "step": 257
    },
    {
      "epoch": 1.1603032004491858,
      "grad_norm": 0.4594995105333513,
      "learning_rate": 4e-05,
      "loss": 0.4163,
      "step": 258
    },
    {
      "epoch": 1.1647950589556428,
      "grad_norm": 0.4015475541306957,
      "learning_rate": 4e-05,
      "loss": 0.4331,
      "step": 259
    },
    {
      "epoch": 1.1692869174621,
      "grad_norm": 0.4338125471271982,
      "learning_rate": 4e-05,
      "loss": 0.4055,
      "step": 260
    },
    {
      "epoch": 1.173778775968557,
      "grad_norm": 0.33198692837258637,
      "learning_rate": 4e-05,
      "loss": 0.398,
      "step": 261
    },
    {
      "epoch": 1.178270634475014,
      "grad_norm": 0.37633532999936237,
      "learning_rate": 4e-05,
      "loss": 0.4425,
      "step": 262
    },
    {
      "epoch": 1.182762492981471,
      "grad_norm": 0.38194118061867677,
      "learning_rate": 4e-05,
      "loss": 0.4145,
      "step": 263
    },
    {
      "epoch": 1.1872543514879281,
      "grad_norm": 0.46415902851925883,
      "learning_rate": 4e-05,
      "loss": 0.4244,
      "step": 264
    },
    {
      "epoch": 1.1917462099943852,
      "grad_norm": 0.4272476761062711,
      "learning_rate": 4e-05,
      "loss": 0.4101,
      "step": 265
    },
    {
      "epoch": 1.1962380685008422,
      "grad_norm": 0.477272824875684,
      "learning_rate": 4e-05,
      "loss": 0.4257,
      "step": 266
    },
    {
      "epoch": 1.2007299270072993,
      "grad_norm": 0.46052805433754523,
      "learning_rate": 4e-05,
      "loss": 0.4191,
      "step": 267
    },
    {
      "epoch": 1.2052217855137564,
      "grad_norm": 0.42671718172921996,
      "learning_rate": 4e-05,
      "loss": 0.4187,
      "step": 268
    },
    {
      "epoch": 1.2097136440202134,
      "grad_norm": 0.41618581662880394,
      "learning_rate": 4e-05,
      "loss": 0.4276,
      "step": 269
    },
    {
      "epoch": 1.2142055025266705,
      "grad_norm": 0.6319986543616546,
      "learning_rate": 4e-05,
      "loss": 0.4036,
      "step": 270
    },
    {
      "epoch": 1.2186973610331275,
      "grad_norm": 0.7187057649197472,
      "learning_rate": 4e-05,
      "loss": 0.438,
      "step": 271
    },
    {
      "epoch": 1.2231892195395846,
      "grad_norm": 0.7990746791193173,
      "learning_rate": 4e-05,
      "loss": 0.3992,
      "step": 272
    },
    {
      "epoch": 1.2276810780460417,
      "grad_norm": 0.7612046061078259,
      "learning_rate": 4e-05,
      "loss": 0.4151,
      "step": 273
    },
    {
      "epoch": 1.2321729365524985,
      "grad_norm": 0.6564491799957413,
      "learning_rate": 4e-05,
      "loss": 0.4287,
      "step": 274
    },
    {
      "epoch": 1.2366647950589555,
      "grad_norm": 0.5794594161593019,
      "learning_rate": 4e-05,
      "loss": 0.4351,
      "step": 275
    },
    {
      "epoch": 1.2411566535654126,
      "grad_norm": 0.5186369540271984,
      "learning_rate": 4e-05,
      "loss": 0.3966,
      "step": 276
    },
    {
      "epoch": 1.2456485120718697,
      "grad_norm": 0.5095617200902736,
      "learning_rate": 4e-05,
      "loss": 0.4295,
      "step": 277
    },
    {
      "epoch": 1.2501403705783267,
      "grad_norm": 0.3390707225901379,
      "learning_rate": 4e-05,
      "loss": 0.4053,
      "step": 278
    },
    {
      "epoch": 1.2546322290847838,
      "grad_norm": 0.4739120543268497,
      "learning_rate": 4e-05,
      "loss": 0.4207,
      "step": 279
    },
    {
      "epoch": 1.2591240875912408,
      "grad_norm": 0.5360068240062349,
      "learning_rate": 4e-05,
      "loss": 0.4165,
      "step": 280
    },
    {
      "epoch": 1.263615946097698,
      "grad_norm": 0.522231052719427,
      "learning_rate": 4e-05,
      "loss": 0.4208,
      "step": 281
    },
    {
      "epoch": 1.268107804604155,
      "grad_norm": 0.5099140566646764,
      "learning_rate": 4e-05,
      "loss": 0.4189,
      "step": 282
    },
    {
      "epoch": 1.272599663110612,
      "grad_norm": 0.5226412830377559,
      "learning_rate": 4e-05,
      "loss": 0.4175,
      "step": 283
    },
    {
      "epoch": 1.277091521617069,
      "grad_norm": 0.4153315222700753,
      "learning_rate": 4e-05,
      "loss": 0.4197,
      "step": 284
    },
    {
      "epoch": 1.2815833801235261,
      "grad_norm": 0.4160640077198146,
      "learning_rate": 4e-05,
      "loss": 0.4212,
      "step": 285
    },
    {
      "epoch": 1.2860752386299832,
      "grad_norm": 0.40898685842353266,
      "learning_rate": 4e-05,
      "loss": 0.4198,
      "step": 286
    },
    {
      "epoch": 1.2905670971364402,
      "grad_norm": 0.3585190992846669,
      "learning_rate": 4e-05,
      "loss": 0.4335,
      "step": 287
    },
    {
      "epoch": 1.2950589556428973,
      "grad_norm": 0.45846966617810186,
      "learning_rate": 4e-05,
      "loss": 0.4186,
      "step": 288
    },
    {
      "epoch": 1.2995508141493544,
      "grad_norm": 0.47187726606199826,
      "learning_rate": 4e-05,
      "loss": 0.399,
      "step": 289
    },
    {
      "epoch": 1.3040426726558114,
      "grad_norm": 0.4794317977249347,
      "learning_rate": 4e-05,
      "loss": 0.4207,
      "step": 290
    },
    {
      "epoch": 1.3085345311622683,
      "grad_norm": 0.5333626541847482,
      "learning_rate": 4e-05,
      "loss": 0.4218,
      "step": 291
    },
    {
      "epoch": 1.3130263896687255,
      "grad_norm": 0.5787006953110693,
      "learning_rate": 4e-05,
      "loss": 0.4145,
      "step": 292
    },
    {
      "epoch": 1.3175182481751824,
      "grad_norm": 0.5172431079368286,
      "learning_rate": 4e-05,
      "loss": 0.4244,
      "step": 293
    },
    {
      "epoch": 1.3220101066816397,
      "grad_norm": 0.46351702390373356,
      "learning_rate": 4e-05,
      "loss": 0.4187,
      "step": 294
    },
    {
      "epoch": 1.3265019651880965,
      "grad_norm": 0.3829661937309283,
      "learning_rate": 4e-05,
      "loss": 0.4273,
      "step": 295
    },
    {
      "epoch": 1.3309938236945535,
      "grad_norm": 0.4561734799207676,
      "learning_rate": 4e-05,
      "loss": 0.4207,
      "step": 296
    },
    {
      "epoch": 1.3354856822010106,
      "grad_norm": 0.5606394080279372,
      "learning_rate": 4e-05,
      "loss": 0.4136,
      "step": 297
    },
    {
      "epoch": 1.3399775407074677,
      "grad_norm": 0.5838378084287339,
      "learning_rate": 4e-05,
      "loss": 0.4147,
      "step": 298
    },
    {
      "epoch": 1.3444693992139247,
      "grad_norm": 0.6803150483301132,
      "learning_rate": 4e-05,
      "loss": 0.4147,
      "step": 299
    },
    {
      "epoch": 1.3489612577203818,
      "grad_norm": 0.9385386759145683,
      "learning_rate": 4e-05,
      "loss": 0.4315,
      "step": 300
    },
    {
      "epoch": 1.3534531162268388,
      "grad_norm": 0.9510589960287957,
      "learning_rate": 4e-05,
      "loss": 0.4233,
      "step": 301
    },
    {
      "epoch": 1.357944974733296,
      "grad_norm": 0.8155983646660458,
      "learning_rate": 4e-05,
      "loss": 0.4156,
      "step": 302
    },
    {
      "epoch": 1.362436833239753,
      "grad_norm": 0.662019772956271,
      "learning_rate": 4e-05,
      "loss": 0.4126,
      "step": 303
    },
    {
      "epoch": 1.36692869174621,
      "grad_norm": 0.5379494785301945,
      "learning_rate": 4e-05,
      "loss": 0.4113,
      "step": 304
    },
    {
      "epoch": 1.371420550252667,
      "grad_norm": 0.4450957694128562,
      "learning_rate": 4e-05,
      "loss": 0.4105,
      "step": 305
    },
    {
      "epoch": 1.3759124087591241,
      "grad_norm": 0.5385460450601072,
      "learning_rate": 4e-05,
      "loss": 0.4156,
      "step": 306
    },
    {
      "epoch": 1.3804042672655812,
      "grad_norm": 0.7176270884592646,
      "learning_rate": 4e-05,
      "loss": 0.4013,
      "step": 307
    },
    {
      "epoch": 1.3848961257720382,
      "grad_norm": 0.7598696051062441,
      "learning_rate": 4e-05,
      "loss": 0.431,
      "step": 308
    },
    {
      "epoch": 1.3893879842784953,
      "grad_norm": 0.6626198841091498,
      "learning_rate": 4e-05,
      "loss": 0.4116,
      "step": 309
    },
    {
      "epoch": 1.3938798427849521,
      "grad_norm": 0.6042171732048263,
      "learning_rate": 4e-05,
      "loss": 0.4128,
      "step": 310
    },
    {
      "epoch": 1.3983717012914094,
      "grad_norm": 0.5138254611835871,
      "learning_rate": 4e-05,
      "loss": 0.399,
      "step": 311
    },
    {
      "epoch": 1.4028635597978663,
      "grad_norm": 0.4696181244920852,
      "learning_rate": 4e-05,
      "loss": 0.418,
      "step": 312
    },
    {
      "epoch": 1.4073554183043235,
      "grad_norm": 0.46692894480025615,
      "learning_rate": 4e-05,
      "loss": 0.4102,
      "step": 313
    },
    {
      "epoch": 1.4118472768107804,
      "grad_norm": 0.5419897398234483,
      "learning_rate": 4e-05,
      "loss": 0.4433,
      "step": 314
    },
    {
      "epoch": 1.4163391353172374,
      "grad_norm": 0.6866743397264817,
      "learning_rate": 4e-05,
      "loss": 0.4107,
      "step": 315
    },
    {
      "epoch": 1.4208309938236945,
      "grad_norm": 0.6716789995098363,
      "learning_rate": 4e-05,
      "loss": 0.4102,
      "step": 316
    },
    {
      "epoch": 1.4253228523301515,
      "grad_norm": 0.6786750166206234,
      "learning_rate": 4e-05,
      "loss": 0.4377,
      "step": 317
    },
    {
      "epoch": 1.4298147108366086,
      "grad_norm": 0.5728572359681285,
      "learning_rate": 4e-05,
      "loss": 0.4159,
      "step": 318
    },
    {
      "epoch": 1.4343065693430657,
      "grad_norm": 0.442057009559478,
      "learning_rate": 4e-05,
      "loss": 0.4266,
      "step": 319
    },
    {
      "epoch": 1.4387984278495227,
      "grad_norm": 0.4670014125561427,
      "learning_rate": 4e-05,
      "loss": 0.4186,
      "step": 320
    },
    {
      "epoch": 1.4432902863559798,
      "grad_norm": 0.5428091819635465,
      "learning_rate": 4e-05,
      "loss": 0.4308,
      "step": 321
    },
    {
      "epoch": 1.4477821448624368,
      "grad_norm": 0.6450388057295665,
      "learning_rate": 4e-05,
      "loss": 0.4109,
      "step": 322
    },
    {
      "epoch": 1.452274003368894,
      "grad_norm": 0.6843072478502371,
      "learning_rate": 4e-05,
      "loss": 0.4298,
      "step": 323
    },
    {
      "epoch": 1.456765861875351,
      "grad_norm": 0.5952359220178496,
      "learning_rate": 4e-05,
      "loss": 0.4214,
      "step": 324
    },
    {
      "epoch": 1.461257720381808,
      "grad_norm": 0.643420064566965,
      "learning_rate": 4e-05,
      "loss": 0.424,
      "step": 325
    },
    {
      "epoch": 1.465749578888265,
      "grad_norm": 0.5543956323979068,
      "learning_rate": 4e-05,
      "loss": 0.423,
      "step": 326
    },
    {
      "epoch": 1.4702414373947221,
      "grad_norm": 0.3782160373779078,
      "learning_rate": 4e-05,
      "loss": 0.3989,
      "step": 327
    },
    {
      "epoch": 1.4747332959011792,
      "grad_norm": 0.47137361199689054,
      "learning_rate": 4e-05,
      "loss": 0.4383,
      "step": 328
    },
    {
      "epoch": 1.479225154407636,
      "grad_norm": 0.5067830081166466,
      "learning_rate": 4e-05,
      "loss": 0.4057,
      "step": 329
    },
    {
      "epoch": 1.4837170129140933,
      "grad_norm": 0.569479595480312,
      "learning_rate": 4e-05,
      "loss": 0.4309,
      "step": 330
    },
    {
      "epoch": 1.4882088714205501,
      "grad_norm": 0.5403747506981664,
      "learning_rate": 4e-05,
      "loss": 0.422,
      "step": 331
    },
    {
      "epoch": 1.4927007299270074,
      "grad_norm": 0.6232818314028562,
      "learning_rate": 4e-05,
      "loss": 0.4115,
      "step": 332
    },
    {
      "epoch": 1.4971925884334643,
      "grad_norm": 0.5217123578214004,
      "learning_rate": 4e-05,
      "loss": 0.4184,
      "step": 333
    },
    {
      "epoch": 1.5016844469399215,
      "grad_norm": 0.37723894166241295,
      "learning_rate": 4e-05,
      "loss": 0.4231,
      "step": 334
    },
    {
      "epoch": 1.5061763054463784,
      "grad_norm": 0.5290176852661544,
      "learning_rate": 4e-05,
      "loss": 0.4218,
      "step": 335
    },
    {
      "epoch": 1.5106681639528357,
      "grad_norm": 0.5460142809451128,
      "learning_rate": 4e-05,
      "loss": 0.4286,
      "step": 336
    },
    {
      "epoch": 1.5151600224592925,
      "grad_norm": 0.3658348001695562,
      "learning_rate": 4e-05,
      "loss": 0.4136,
      "step": 337
    },
    {
      "epoch": 1.5196518809657495,
      "grad_norm": 0.470838084751463,
      "learning_rate": 4e-05,
      "loss": 0.4071,
      "step": 338
    },
    {
      "epoch": 1.5241437394722066,
      "grad_norm": 0.5162069919323958,
      "learning_rate": 4e-05,
      "loss": 0.4203,
      "step": 339
    },
    {
      "epoch": 1.5286355979786637,
      "grad_norm": 0.39725404880080073,
      "learning_rate": 4e-05,
      "loss": 0.4158,
      "step": 340
    },
    {
      "epoch": 1.5331274564851207,
      "grad_norm": 0.4194286976473832,
      "learning_rate": 4e-05,
      "loss": 0.4224,
      "step": 341
    },
    {
      "epoch": 1.5376193149915778,
      "grad_norm": 0.48253154256108516,
      "learning_rate": 4e-05,
      "loss": 0.4143,
      "step": 342
    },
    {
      "epoch": 1.5421111734980348,
      "grad_norm": 0.4785038648117691,
      "learning_rate": 4e-05,
      "loss": 0.4296,
      "step": 343
    },
    {
      "epoch": 1.546603032004492,
      "grad_norm": 0.5114749438333243,
      "learning_rate": 4e-05,
      "loss": 0.4211,
      "step": 344
    },
    {
      "epoch": 1.551094890510949,
      "grad_norm": 0.6224813300129919,
      "learning_rate": 4e-05,
      "loss": 0.4097,
      "step": 345
    },
    {
      "epoch": 1.5555867490174058,
      "grad_norm": 0.6230580890782437,
      "learning_rate": 4e-05,
      "loss": 0.4244,
      "step": 346
    },
    {
      "epoch": 1.560078607523863,
      "grad_norm": 0.5134467787355581,
      "learning_rate": 4e-05,
      "loss": 0.4146,
      "step": 347
    },
    {
      "epoch": 1.56457046603032,
      "grad_norm": 0.606897722271038,
      "learning_rate": 4e-05,
      "loss": 0.4046,
      "step": 348
    },
    {
      "epoch": 1.5690623245367772,
      "grad_norm": 0.6290535779488216,
      "learning_rate": 4e-05,
      "loss": 0.4172,
      "step": 349
    },
    {
      "epoch": 1.573554183043234,
      "grad_norm": 0.708340602995717,
      "learning_rate": 4e-05,
      "loss": 0.4346,
      "step": 350
    },
    {
      "epoch": 1.5780460415496913,
      "grad_norm": 0.8056428974253329,
      "learning_rate": 4e-05,
      "loss": 0.4219,
      "step": 351
    },
    {
      "epoch": 1.5825379000561481,
      "grad_norm": 0.903400987094472,
      "learning_rate": 4e-05,
      "loss": 0.4207,
      "step": 352
    },
    {
      "epoch": 1.5870297585626054,
      "grad_norm": 0.730665072472088,
      "learning_rate": 4e-05,
      "loss": 0.3962,
      "step": 353
    },
    {
      "epoch": 1.5915216170690623,
      "grad_norm": 0.4941087521343281,
      "learning_rate": 4e-05,
      "loss": 0.4351,
      "step": 354
    },
    {
      "epoch": 1.5960134755755195,
      "grad_norm": 0.5763323928451937,
      "learning_rate": 4e-05,
      "loss": 0.4267,
      "step": 355
    },
    {
      "epoch": 1.6005053340819764,
      "grad_norm": 0.6040841604196749,
      "learning_rate": 4e-05,
      "loss": 0.3991,
      "step": 356
    },
    {
      "epoch": 1.6049971925884334,
      "grad_norm": 0.528214304828605,
      "learning_rate": 4e-05,
      "loss": 0.4126,
      "step": 357
    },
    {
      "epoch": 1.6094890510948905,
      "grad_norm": 0.3908959757632829,
      "learning_rate": 4e-05,
      "loss": 0.4281,
      "step": 358
    },
    {
      "epoch": 1.6139809096013475,
      "grad_norm": 0.5706963628700729,
      "learning_rate": 4e-05,
      "loss": 0.4284,
      "step": 359
    },
    {
      "epoch": 1.6184727681078046,
      "grad_norm": 0.6595006111421147,
      "learning_rate": 4e-05,
      "loss": 0.4149,
      "step": 360
    },
    {
      "epoch": 1.6229646266142617,
      "grad_norm": 0.45922917893401166,
      "learning_rate": 4e-05,
      "loss": 0.4182,
      "step": 361
    },
    {
      "epoch": 1.6274564851207187,
      "grad_norm": 0.4787813652443957,
      "learning_rate": 4e-05,
      "loss": 0.4213,
      "step": 362
    },
    {
      "epoch": 1.6319483436271758,
      "grad_norm": 0.47309691076200305,
      "learning_rate": 4e-05,
      "loss": 0.3969,
      "step": 363
    },
    {
      "epoch": 1.6364402021336328,
      "grad_norm": 0.4497755632460707,
      "learning_rate": 4e-05,
      "loss": 0.4237,
      "step": 364
    },
    {
      "epoch": 1.6409320606400897,
      "grad_norm": 0.4175916391032968,
      "learning_rate": 4e-05,
      "loss": 0.404,
      "step": 365
    },
    {
      "epoch": 1.645423919146547,
      "grad_norm": 0.557284717051137,
      "learning_rate": 4e-05,
      "loss": 0.4155,
      "step": 366
    },
    {
      "epoch": 1.6499157776530038,
      "grad_norm": 0.5488169683654379,
      "learning_rate": 4e-05,
      "loss": 0.4184,
      "step": 367
    },
    {
      "epoch": 1.654407636159461,
      "grad_norm": 0.7206046850685944,
      "learning_rate": 4e-05,
      "loss": 0.3982,
      "step": 368
    },
    {
      "epoch": 1.658899494665918,
      "grad_norm": 0.8551804521574082,
      "learning_rate": 4e-05,
      "loss": 0.428,
      "step": 369
    },
    {
      "epoch": 1.6633913531723752,
      "grad_norm": 0.7210896730492177,
      "learning_rate": 4e-05,
      "loss": 0.411,
      "step": 370
    },
    {
      "epoch": 1.667883211678832,
      "grad_norm": 0.6846399484679692,
      "learning_rate": 4e-05,
      "loss": 0.4118,
      "step": 371
    },
    {
      "epoch": 1.6723750701852893,
      "grad_norm": 0.7202545710292472,
      "learning_rate": 4e-05,
      "loss": 0.4284,
      "step": 372
    },
    {
      "epoch": 1.6768669286917461,
      "grad_norm": 0.5433528460968153,
      "learning_rate": 4e-05,
      "loss": 0.432,
      "step": 373
    },
    {
      "epoch": 1.6813587871982034,
      "grad_norm": 0.44097726264478215,
      "learning_rate": 4e-05,
      "loss": 0.4015,
      "step": 374
    },
    {
      "epoch": 1.6858506457046603,
      "grad_norm": 0.4124921565500918,
      "learning_rate": 4e-05,
      "loss": 0.4106,
      "step": 375
    },
    {
      "epoch": 1.6903425042111173,
      "grad_norm": 0.42374502044167894,
      "learning_rate": 4e-05,
      "loss": 0.4221,
      "step": 376
    },
    {
      "epoch": 1.6948343627175744,
      "grad_norm": 0.5283287945870614,
      "learning_rate": 4e-05,
      "loss": 0.4198,
      "step": 377
    },
    {
      "epoch": 1.6993262212240314,
      "grad_norm": 0.5842987086876869,
      "learning_rate": 4e-05,
      "loss": 0.4181,
      "step": 378
    },
    {
      "epoch": 1.7038180797304885,
      "grad_norm": 0.5760357519612713,
      "learning_rate": 4e-05,
      "loss": 0.4107,
      "step": 379
    },
    {
      "epoch": 1.7083099382369455,
      "grad_norm": 0.5676663134290156,
      "learning_rate": 4e-05,
      "loss": 0.4137,
      "step": 380
    },
    {
      "epoch": 1.7128017967434026,
      "grad_norm": 0.6153122107224821,
      "learning_rate": 4e-05,
      "loss": 0.43,
      "step": 381
    },
    {
      "epoch": 1.7172936552498597,
      "grad_norm": 0.644075331574322,
      "learning_rate": 4e-05,
      "loss": 0.4124,
      "step": 382
    },
    {
      "epoch": 1.7217855137563167,
      "grad_norm": 0.5029326578799416,
      "learning_rate": 4e-05,
      "loss": 0.4371,
      "step": 383
    },
    {
      "epoch": 1.7262773722627736,
      "grad_norm": 0.3843931572164035,
      "learning_rate": 4e-05,
      "loss": 0.4081,
      "step": 384
    },
    {
      "epoch": 1.7307692307692308,
      "grad_norm": 0.4235561773896018,
      "learning_rate": 4e-05,
      "loss": 0.4173,
      "step": 385
    },
    {
      "epoch": 1.7352610892756877,
      "grad_norm": 0.5494238937866214,
      "learning_rate": 4e-05,
      "loss": 0.4303,
      "step": 386
    },
    {
      "epoch": 1.739752947782145,
      "grad_norm": 0.5914098327914491,
      "learning_rate": 4e-05,
      "loss": 0.4155,
      "step": 387
    },
    {
      "epoch": 1.7442448062886018,
      "grad_norm": 0.5053679608492333,
      "learning_rate": 4e-05,
      "loss": 0.4073,
      "step": 388
    },
    {
      "epoch": 1.748736664795059,
      "grad_norm": 0.49526310371529964,
      "learning_rate": 4e-05,
      "loss": 0.4281,
      "step": 389
    },
    {
      "epoch": 1.753228523301516,
      "grad_norm": 0.6931321847226084,
      "learning_rate": 4e-05,
      "loss": 0.4095,
      "step": 390
    },
    {
      "epoch": 1.7577203818079732,
      "grad_norm": 0.7008072365751862,
      "learning_rate": 4e-05,
      "loss": 0.4417,
      "step": 391
    },
    {
      "epoch": 1.76221224031443,
      "grad_norm": 0.5705239684281941,
      "learning_rate": 4e-05,
      "loss": 0.4073,
      "step": 392
    },
    {
      "epoch": 1.7667040988208873,
      "grad_norm": 0.5586859579572065,
      "learning_rate": 4e-05,
      "loss": 0.4227,
      "step": 393
    },
    {
      "epoch": 1.7711959573273441,
      "grad_norm": 0.4180760209527486,
      "learning_rate": 4e-05,
      "loss": 0.4074,
      "step": 394
    },
    {
      "epoch": 1.7756878158338012,
      "grad_norm": 0.4235751919300548,
      "learning_rate": 4e-05,
      "loss": 0.408,
      "step": 395
    },
    {
      "epoch": 1.7801796743402583,
      "grad_norm": 0.44786465160299255,
      "learning_rate": 4e-05,
      "loss": 0.4051,
      "step": 396
    },
    {
      "epoch": 1.7846715328467153,
      "grad_norm": 0.40582132557294276,
      "learning_rate": 4e-05,
      "loss": 0.4041,
      "step": 397
    },
    {
      "epoch": 1.7891633913531724,
      "grad_norm": 0.415501797589053,
      "learning_rate": 4e-05,
      "loss": 0.4181,
      "step": 398
    },
    {
      "epoch": 1.7936552498596294,
      "grad_norm": 0.5688319430116934,
      "learning_rate": 4e-05,
      "loss": 0.4063,
      "step": 399
    },
    {
      "epoch": 1.7981471083660865,
      "grad_norm": 0.5222786916523878,
      "learning_rate": 4e-05,
      "loss": 0.4225,
      "step": 400
    },
    {
      "epoch": 1.8026389668725435,
      "grad_norm": 0.39175912016483655,
      "learning_rate": 4e-05,
      "loss": 0.4233,
      "step": 401
    },
    {
      "epoch": 1.8071308253790006,
      "grad_norm": 0.3899723576702837,
      "learning_rate": 4e-05,
      "loss": 0.4062,
      "step": 402
    },
    {
      "epoch": 1.8116226838854577,
      "grad_norm": 0.44899099339006415,
      "learning_rate": 4e-05,
      "loss": 0.4293,
      "step": 403
    },
    {
      "epoch": 1.8161145423919147,
      "grad_norm": 0.6515908652092102,
      "learning_rate": 4e-05,
      "loss": 0.4278,
      "step": 404
    },
    {
      "epoch": 1.8206064008983716,
      "grad_norm": 0.7361432974904519,
      "learning_rate": 4e-05,
      "loss": 0.4121,
      "step": 405
    },
    {
      "epoch": 1.8250982594048288,
      "grad_norm": 0.727810932046529,
      "learning_rate": 4e-05,
      "loss": 0.4168,
      "step": 406
    },
    {
      "epoch": 1.8295901179112857,
      "grad_norm": 0.9147354600024983,
      "learning_rate": 4e-05,
      "loss": 0.4176,
      "step": 407
    },
    {
      "epoch": 1.834081976417743,
      "grad_norm": 0.9384772456277503,
      "learning_rate": 4e-05,
      "loss": 0.4008,
      "step": 408
    },
    {
      "epoch": 1.8385738349241998,
      "grad_norm": 0.785653505953151,
      "learning_rate": 4e-05,
      "loss": 0.4119,
      "step": 409
    },
    {
      "epoch": 1.843065693430657,
      "grad_norm": 0.6290608242503178,
      "learning_rate": 4e-05,
      "loss": 0.4306,
      "step": 410
    },
    {
      "epoch": 1.847557551937114,
      "grad_norm": 0.4093006721495116,
      "learning_rate": 4e-05,
      "loss": 0.4061,
      "step": 411
    },
    {
      "epoch": 1.8520494104435712,
      "grad_norm": 0.4428210352972703,
      "learning_rate": 4e-05,
      "loss": 0.412,
      "step": 412
    },
    {
      "epoch": 1.856541268950028,
      "grad_norm": 0.5543478216737765,
      "learning_rate": 4e-05,
      "loss": 0.4338,
      "step": 413
    },
    {
      "epoch": 1.861033127456485,
      "grad_norm": 0.6723269151630776,
      "learning_rate": 4e-05,
      "loss": 0.3931,
      "step": 414
    },
    {
      "epoch": 1.8655249859629421,
      "grad_norm": 0.6554716382008948,
      "learning_rate": 4e-05,
      "loss": 0.4123,
      "step": 415
    },
    {
      "epoch": 1.8700168444693992,
      "grad_norm": 0.6814087999195951,
      "learning_rate": 4e-05,
      "loss": 0.4449,
      "step": 416
    },
    {
      "epoch": 1.8745087029758563,
      "grad_norm": 0.6293444361131807,
      "learning_rate": 4e-05,
      "loss": 0.4093,
      "step": 417
    },
    {
      "epoch": 1.8790005614823133,
      "grad_norm": 0.5239372432818226,
      "learning_rate": 4e-05,
      "loss": 0.4152,
      "step": 418
    },
    {
      "epoch": 1.8834924199887704,
      "grad_norm": 0.7016221391462388,
      "learning_rate": 4e-05,
      "loss": 0.4268,
      "step": 419
    },
    {
      "epoch": 1.8879842784952274,
      "grad_norm": 0.7491045489395309,
      "learning_rate": 4e-05,
      "loss": 0.4042,
      "step": 420
    },
    {
      "epoch": 1.8924761370016845,
      "grad_norm": 0.5373005289145114,
      "learning_rate": 4e-05,
      "loss": 0.4148,
      "step": 421
    },
    {
      "epoch": 1.8969679955081415,
      "grad_norm": 0.36242363600835353,
      "learning_rate": 4e-05,
      "loss": 0.4118,
      "step": 422
    },
    {
      "epoch": 1.9014598540145986,
      "grad_norm": 0.4034854156076673,
      "learning_rate": 4e-05,
      "loss": 0.4147,
      "step": 423
    },
    {
      "epoch": 1.9059517125210554,
      "grad_norm": 0.4625091597649348,
      "learning_rate": 4e-05,
      "loss": 0.4289,
      "step": 424
    },
    {
      "epoch": 1.9104435710275127,
      "grad_norm": 0.6636893621508165,
      "learning_rate": 4e-05,
      "loss": 0.4021,
      "step": 425
    },
    {
      "epoch": 1.9149354295339696,
      "grad_norm": 0.7731482192209793,
      "learning_rate": 4e-05,
      "loss": 0.4059,
      "step": 426
    },
    {
      "epoch": 1.9194272880404268,
      "grad_norm": 0.6321068474163406,
      "learning_rate": 4e-05,
      "loss": 0.4367,
      "step": 427
    },
    {
      "epoch": 1.9239191465468837,
      "grad_norm": 0.6431637505047455,
      "learning_rate": 4e-05,
      "loss": 0.4116,
      "step": 428
    },
    {
      "epoch": 1.928411005053341,
      "grad_norm": 0.6173795038682337,
      "learning_rate": 4e-05,
      "loss": 0.4161,
      "step": 429
    },
    {
      "epoch": 1.9329028635597978,
      "grad_norm": 0.5410869916051849,
      "learning_rate": 4e-05,
      "loss": 0.4055,
      "step": 430
    },
    {
      "epoch": 1.937394722066255,
      "grad_norm": 0.564711766247606,
      "learning_rate": 4e-05,
      "loss": 0.4153,
      "step": 431
    },
    {
      "epoch": 1.941886580572712,
      "grad_norm": 0.652114690543729,
      "learning_rate": 4e-05,
      "loss": 0.4226,
      "step": 432
    },
    {
      "epoch": 1.946378439079169,
      "grad_norm": 0.5841671627139245,
      "learning_rate": 4e-05,
      "loss": 0.41,
      "step": 433
    },
    {
      "epoch": 1.950870297585626,
      "grad_norm": 0.5718142593238181,
      "learning_rate": 4e-05,
      "loss": 0.4052,
      "step": 434
    },
    {
      "epoch": 1.955362156092083,
      "grad_norm": 0.7003394398741115,
      "learning_rate": 4e-05,
      "loss": 0.4237,
      "step": 435
    },
    {
      "epoch": 1.9598540145985401,
      "grad_norm": 0.6875693015818767,
      "learning_rate": 4e-05,
      "loss": 0.4234,
      "step": 436
    },
    {
      "epoch": 1.9643458731049972,
      "grad_norm": 0.59324395073825,
      "learning_rate": 4e-05,
      "loss": 0.4158,
      "step": 437
    },
    {
      "epoch": 1.9688377316114543,
      "grad_norm": 0.49665663577166813,
      "learning_rate": 4e-05,
      "loss": 0.4126,
      "step": 438
    },
    {
      "epoch": 1.9733295901179113,
      "grad_norm": 0.44669237828063063,
      "learning_rate": 4e-05,
      "loss": 0.4069,
      "step": 439
    },
    {
      "epoch": 1.9778214486243684,
      "grad_norm": 0.3686434306003568,
      "learning_rate": 4e-05,
      "loss": 0.4204,
      "step": 440
    },
    {
      "epoch": 1.9823133071308254,
      "grad_norm": 0.4971677321111303,
      "learning_rate": 4e-05,
      "loss": 0.4084,
      "step": 441
    },
    {
      "epoch": 1.9868051656372825,
      "grad_norm": 0.4795175264200768,
      "learning_rate": 4e-05,
      "loss": 0.402,
      "step": 442
    },
    {
      "epoch": 1.9912970241437393,
      "grad_norm": 0.4509517431549584,
      "learning_rate": 4e-05,
      "loss": 0.4271,
      "step": 443
    },
    {
      "epoch": 1.9957888826501966,
      "grad_norm": 0.45803780139694994,
      "learning_rate": 4e-05,
      "loss": 0.4118,
      "step": 444
    },
    {
      "epoch": 2.0016844469399215,
      "grad_norm": 0.8776500584068143,
      "learning_rate": 4e-05,
      "loss": 0.6659,
      "step": 445
    },
    {
      "epoch": 2.0061763054463784,
      "grad_norm": 1.2438371042596672,
      "learning_rate": 4e-05,
      "loss": 0.3809,
      "step": 446
    },
    {
      "epoch": 2.0106681639528357,
      "grad_norm": 0.73907980667939,
      "learning_rate": 4e-05,
      "loss": 0.3729,
      "step": 447
    },
    {
      "epoch": 2.0151600224592925,
      "grad_norm": 0.6729157247984111,
      "learning_rate": 4e-05,
      "loss": 0.364,
      "step": 448
    },
    {
      "epoch": 2.0196518809657498,
      "grad_norm": 0.7461339868111228,
      "learning_rate": 4e-05,
      "loss": 0.371,
      "step": 449
    },
    {
      "epoch": 2.0241437394722066,
      "grad_norm": 0.8241260594853907,
      "learning_rate": 4e-05,
      "loss": 0.384,
      "step": 450
    },
    {
      "epoch": 2.028635597978664,
      "grad_norm": 0.9341600031948939,
      "learning_rate": 4e-05,
      "loss": 0.3829,
      "step": 451
    },
    {
      "epoch": 2.0331274564851207,
      "grad_norm": 1.063783125407379,
      "learning_rate": 4e-05,
      "loss": 0.3688,
      "step": 452
    },
    {
      "epoch": 2.0376193149915776,
      "grad_norm": 0.7168872805058292,
      "learning_rate": 4e-05,
      "loss": 0.3637,
      "step": 453
    },
    {
      "epoch": 2.042111173498035,
      "grad_norm": 0.7005003667873583,
      "learning_rate": 4e-05,
      "loss": 0.3784,
      "step": 454
    },
    {
      "epoch": 2.0466030320044917,
      "grad_norm": 1.1050176835966414,
      "learning_rate": 4e-05,
      "loss": 0.3822,
      "step": 455
    },
    {
      "epoch": 2.051094890510949,
      "grad_norm": 0.7439076131304606,
      "learning_rate": 4e-05,
      "loss": 0.3631,
      "step": 456
    },
    {
      "epoch": 2.055586749017406,
      "grad_norm": 0.6248674143163359,
      "learning_rate": 4e-05,
      "loss": 0.3716,
      "step": 457
    },
    {
      "epoch": 2.060078607523863,
      "grad_norm": 0.7821050807840862,
      "learning_rate": 4e-05,
      "loss": 0.3833,
      "step": 458
    },
    {
      "epoch": 2.06457046603032,
      "grad_norm": 0.9053874337400066,
      "learning_rate": 4e-05,
      "loss": 0.3886,
      "step": 459
    },
    {
      "epoch": 2.069062324536777,
      "grad_norm": 0.8174200262761315,
      "learning_rate": 4e-05,
      "loss": 0.3566,
      "step": 460
    },
    {
      "epoch": 2.073554183043234,
      "grad_norm": 0.581350974251612,
      "learning_rate": 4e-05,
      "loss": 0.361,
      "step": 461
    },
    {
      "epoch": 2.0780460415496913,
      "grad_norm": 0.5866318972509964,
      "learning_rate": 4e-05,
      "loss": 0.3731,
      "step": 462
    },
    {
      "epoch": 2.082537900056148,
      "grad_norm": 0.6030350039620676,
      "learning_rate": 4e-05,
      "loss": 0.372,
      "step": 463
    },
    {
      "epoch": 2.0870297585626054,
      "grad_norm": 0.5652825787356963,
      "learning_rate": 4e-05,
      "loss": 0.3711,
      "step": 464
    },
    {
      "epoch": 2.0915216170690623,
      "grad_norm": 0.6459679458072997,
      "learning_rate": 4e-05,
      "loss": 0.365,
      "step": 465
    },
    {
      "epoch": 2.0960134755755195,
      "grad_norm": 0.6160045648614417,
      "learning_rate": 4e-05,
      "loss": 0.3599,
      "step": 466
    },
    {
      "epoch": 2.1005053340819764,
      "grad_norm": 0.4686170942467444,
      "learning_rate": 4e-05,
      "loss": 0.3739,
      "step": 467
    },
    {
      "epoch": 2.1049971925884337,
      "grad_norm": 0.42052418076040404,
      "learning_rate": 4e-05,
      "loss": 0.3705,
      "step": 468
    },
    {
      "epoch": 2.1094890510948905,
      "grad_norm": 0.5231744142675778,
      "learning_rate": 4e-05,
      "loss": 0.3599,
      "step": 469
    },
    {
      "epoch": 2.1139809096013478,
      "grad_norm": 0.49578197670221213,
      "learning_rate": 4e-05,
      "loss": 0.3862,
      "step": 470
    },
    {
      "epoch": 2.1184727681078046,
      "grad_norm": 0.48834559315100046,
      "learning_rate": 4e-05,
      "loss": 0.3632,
      "step": 471
    },
    {
      "epoch": 2.1229646266142614,
      "grad_norm": 0.5993065891394365,
      "learning_rate": 4e-05,
      "loss": 0.3728,
      "step": 472
    },
    {
      "epoch": 2.1274564851207187,
      "grad_norm": 0.5200187083740063,
      "learning_rate": 4e-05,
      "loss": 0.3761,
      "step": 473
    },
    {
      "epoch": 2.1319483436271756,
      "grad_norm": 0.44735477174778177,
      "learning_rate": 4e-05,
      "loss": 0.381,
      "step": 474
    },
    {
      "epoch": 2.136440202133633,
      "grad_norm": 0.35622605081751896,
      "learning_rate": 4e-05,
      "loss": 0.3665,
      "step": 475
    },
    {
      "epoch": 2.1409320606400897,
      "grad_norm": 0.48700689851053464,
      "learning_rate": 4e-05,
      "loss": 0.3688,
      "step": 476
    },
    {
      "epoch": 2.145423919146547,
      "grad_norm": 0.47733707677458204,
      "learning_rate": 4e-05,
      "loss": 0.3691,
      "step": 477
    },
    {
      "epoch": 2.149915777653004,
      "grad_norm": 0.4849992866388392,
      "learning_rate": 4e-05,
      "loss": 0.3649,
      "step": 478
    },
    {
      "epoch": 2.154407636159461,
      "grad_norm": 0.4619611163929782,
      "learning_rate": 4e-05,
      "loss": 0.3696,
      "step": 479
    },
    {
      "epoch": 2.158899494665918,
      "grad_norm": 0.34403467186103076,
      "learning_rate": 4e-05,
      "loss": 0.3702,
      "step": 480
    },
    {
      "epoch": 2.163391353172375,
      "grad_norm": 0.5444183060261978,
      "learning_rate": 4e-05,
      "loss": 0.3845,
      "step": 481
    },
    {
      "epoch": 2.167883211678832,
      "grad_norm": 0.5881304938527785,
      "learning_rate": 4e-05,
      "loss": 0.341,
      "step": 482
    },
    {
      "epoch": 2.1723750701852893,
      "grad_norm": 0.4519994650509173,
      "learning_rate": 4e-05,
      "loss": 0.3961,
      "step": 483
    },
    {
      "epoch": 2.176866928691746,
      "grad_norm": 0.3092773006691979,
      "learning_rate": 4e-05,
      "loss": 0.3639,
      "step": 484
    },
    {
      "epoch": 2.1813587871982034,
      "grad_norm": 0.5074146674725775,
      "learning_rate": 4e-05,
      "loss": 0.3867,
      "step": 485
    },
    {
      "epoch": 2.1858506457046603,
      "grad_norm": 0.5916931618899153,
      "learning_rate": 4e-05,
      "loss": 0.3861,
      "step": 486
    },
    {
      "epoch": 2.1903425042111175,
      "grad_norm": 0.6578071662341124,
      "learning_rate": 4e-05,
      "loss": 0.3595,
      "step": 487
    },
    {
      "epoch": 2.1948343627175744,
      "grad_norm": 0.630816880417875,
      "learning_rate": 4e-05,
      "loss": 0.3818,
      "step": 488
    },
    {
      "epoch": 2.199326221224031,
      "grad_norm": 0.5042854765278244,
      "learning_rate": 4e-05,
      "loss": 0.368,
      "step": 489
    },
    {
      "epoch": 2.2038180797304885,
      "grad_norm": 0.42115378319149216,
      "learning_rate": 4e-05,
      "loss": 0.3617,
      "step": 490
    },
    {
      "epoch": 2.2083099382369458,
      "grad_norm": 0.3908254000467483,
      "learning_rate": 4e-05,
      "loss": 0.3667,
      "step": 491
    },
    {
      "epoch": 2.2128017967434026,
      "grad_norm": 0.4448611128689832,
      "learning_rate": 4e-05,
      "loss": 0.3645,
      "step": 492
    },
    {
      "epoch": 2.2172936552498594,
      "grad_norm": 0.4944868221564945,
      "learning_rate": 4e-05,
      "loss": 0.3849,
      "step": 493
    },
    {
      "epoch": 2.2217855137563167,
      "grad_norm": 0.4666560741381937,
      "learning_rate": 4e-05,
      "loss": 0.3573,
      "step": 494
    },
    {
      "epoch": 2.2262773722627736,
      "grad_norm": 0.3215553793551666,
      "learning_rate": 4e-05,
      "loss": 0.3608,
      "step": 495
    },
    {
      "epoch": 2.230769230769231,
      "grad_norm": 0.3359867397822268,
      "learning_rate": 4e-05,
      "loss": 0.3653,
      "step": 496
    },
    {
      "epoch": 2.2352610892756877,
      "grad_norm": 0.4068587788112034,
      "learning_rate": 4e-05,
      "loss": 0.3711,
      "step": 497
    },
    {
      "epoch": 2.239752947782145,
      "grad_norm": 0.4815590365381342,
      "learning_rate": 4e-05,
      "loss": 0.3651,
      "step": 498
    },
    {
      "epoch": 2.244244806288602,
      "grad_norm": 0.4602607684295312,
      "learning_rate": 4e-05,
      "loss": 0.3584,
      "step": 499
    },
    {
      "epoch": 2.248736664795059,
      "grad_norm": 0.3561749106254481,
      "learning_rate": 4e-05,
      "loss": 0.3833,
      "step": 500
    },
    {
      "epoch": 2.253228523301516,
      "grad_norm": 0.405044748429829,
      "learning_rate": 4e-05,
      "loss": 0.3704,
      "step": 501
    },
    {
      "epoch": 2.257720381807973,
      "grad_norm": 0.5009329482609307,
      "learning_rate": 4e-05,
      "loss": 0.386,
      "step": 502
    },
    {
      "epoch": 2.26221224031443,
      "grad_norm": 0.674235167783579,
      "learning_rate": 4e-05,
      "loss": 0.3607,
      "step": 503
    },
    {
      "epoch": 2.2667040988208873,
      "grad_norm": 0.6650958745577789,
      "learning_rate": 4e-05,
      "loss": 0.3954,
      "step": 504
    },
    {
      "epoch": 2.271195957327344,
      "grad_norm": 0.608405829439968,
      "learning_rate": 4e-05,
      "loss": 0.3681,
      "step": 505
    },
    {
      "epoch": 2.2756878158338014,
      "grad_norm": 0.48113375615285886,
      "learning_rate": 4e-05,
      "loss": 0.3627,
      "step": 506
    },
    {
      "epoch": 2.2801796743402583,
      "grad_norm": 0.4550605757845848,
      "learning_rate": 4e-05,
      "loss": 0.3721,
      "step": 507
    },
    {
      "epoch": 2.2846715328467155,
      "grad_norm": 0.48004125396671943,
      "learning_rate": 4e-05,
      "loss": 0.4006,
      "step": 508
    },
    {
      "epoch": 2.2891633913531724,
      "grad_norm": 0.5810520147150838,
      "learning_rate": 4e-05,
      "loss": 0.3687,
      "step": 509
    },
    {
      "epoch": 2.293655249859629,
      "grad_norm": 0.7428879222537429,
      "learning_rate": 4e-05,
      "loss": 0.3788,
      "step": 510
    },
    {
      "epoch": 2.2981471083660865,
      "grad_norm": 0.6603266079916377,
      "learning_rate": 4e-05,
      "loss": 0.3808,
      "step": 511
    },
    {
      "epoch": 2.3026389668725433,
      "grad_norm": 0.40515352315014136,
      "learning_rate": 4e-05,
      "loss": 0.384,
      "step": 512
    },
    {
      "epoch": 2.3071308253790006,
      "grad_norm": 0.363014934330548,
      "learning_rate": 4e-05,
      "loss": 0.3873,
      "step": 513
    },
    {
      "epoch": 2.3116226838854574,
      "grad_norm": 0.37884395992894115,
      "learning_rate": 4e-05,
      "loss": 0.366,
      "step": 514
    },
    {
      "epoch": 2.3161145423919147,
      "grad_norm": 0.4887405223356135,
      "learning_rate": 4e-05,
      "loss": 0.3732,
      "step": 515
    },
    {
      "epoch": 2.3206064008983716,
      "grad_norm": 0.5845601743968118,
      "learning_rate": 4e-05,
      "loss": 0.3633,
      "step": 516
    },
    {
      "epoch": 2.325098259404829,
      "grad_norm": 0.5725302745850596,
      "learning_rate": 4e-05,
      "loss": 0.3786,
      "step": 517
    },
    {
      "epoch": 2.3295901179112857,
      "grad_norm": 0.3989281774410526,
      "learning_rate": 4e-05,
      "loss": 0.3816,
      "step": 518
    },
    {
      "epoch": 2.334081976417743,
      "grad_norm": 0.40249195920107184,
      "learning_rate": 4e-05,
      "loss": 0.3712,
      "step": 519
    },
    {
      "epoch": 2.3385738349242,
      "grad_norm": 0.5019774339978715,
      "learning_rate": 4e-05,
      "loss": 0.362,
      "step": 520
    },
    {
      "epoch": 2.343065693430657,
      "grad_norm": 0.5477071017560632,
      "learning_rate": 4e-05,
      "loss": 0.4083,
      "step": 521
    },
    {
      "epoch": 2.347557551937114,
      "grad_norm": 0.5822508257782679,
      "learning_rate": 4e-05,
      "loss": 0.3575,
      "step": 522
    },
    {
      "epoch": 2.352049410443571,
      "grad_norm": 0.4627829061556483,
      "learning_rate": 4e-05,
      "loss": 0.383,
      "step": 523
    },
    {
      "epoch": 2.356541268950028,
      "grad_norm": 0.319209824854133,
      "learning_rate": 4e-05,
      "loss": 0.3671,
      "step": 524
    },
    {
      "epoch": 2.3610331274564853,
      "grad_norm": 0.32402038977952646,
      "learning_rate": 4e-05,
      "loss": 0.3833,
      "step": 525
    },
    {
      "epoch": 2.365524985962942,
      "grad_norm": 0.3938134783010219,
      "learning_rate": 4e-05,
      "loss": 0.366,
      "step": 526
    },
    {
      "epoch": 2.370016844469399,
      "grad_norm": 0.4031112340351704,
      "learning_rate": 4e-05,
      "loss": 0.3663,
      "step": 527
    },
    {
      "epoch": 2.3745087029758563,
      "grad_norm": 0.34298557504905686,
      "learning_rate": 4e-05,
      "loss": 0.3902,
      "step": 528
    },
    {
      "epoch": 2.3790005614823135,
      "grad_norm": 0.33021318915425457,
      "learning_rate": 4e-05,
      "loss": 0.3763,
      "step": 529
    },
    {
      "epoch": 2.3834924199887704,
      "grad_norm": 0.37649942069966086,
      "learning_rate": 4e-05,
      "loss": 0.395,
      "step": 530
    },
    {
      "epoch": 2.387984278495227,
      "grad_norm": 0.35009991290019565,
      "learning_rate": 4e-05,
      "loss": 0.3581,
      "step": 531
    },
    {
      "epoch": 2.3924761370016845,
      "grad_norm": 0.3711353359692777,
      "learning_rate": 4e-05,
      "loss": 0.3892,
      "step": 532
    },
    {
      "epoch": 2.3969679955081413,
      "grad_norm": 0.48980700002588906,
      "learning_rate": 4e-05,
      "loss": 0.3672,
      "step": 533
    },
    {
      "epoch": 2.4014598540145986,
      "grad_norm": 0.5533988152172018,
      "learning_rate": 4e-05,
      "loss": 0.3759,
      "step": 534
    },
    {
      "epoch": 2.4059517125210554,
      "grad_norm": 0.6346413857630363,
      "learning_rate": 4e-05,
      "loss": 0.3903,
      "step": 535
    },
    {
      "epoch": 2.4104435710275127,
      "grad_norm": 0.6983066200100682,
      "learning_rate": 4e-05,
      "loss": 0.3727,
      "step": 536
    },
    {
      "epoch": 2.4149354295339696,
      "grad_norm": 0.7258313793497868,
      "learning_rate": 4e-05,
      "loss": 0.3792,
      "step": 537
    },
    {
      "epoch": 2.419427288040427,
      "grad_norm": 0.6619573955404704,
      "learning_rate": 4e-05,
      "loss": 0.3865,
      "step": 538
    },
    {
      "epoch": 2.4239191465468837,
      "grad_norm": 0.5612054943179736,
      "learning_rate": 4e-05,
      "loss": 0.3492,
      "step": 539
    },
    {
      "epoch": 2.428411005053341,
      "grad_norm": 0.41763006199015507,
      "learning_rate": 4e-05,
      "loss": 0.3862,
      "step": 540
    },
    {
      "epoch": 2.432902863559798,
      "grad_norm": 0.6215162049820858,
      "learning_rate": 4e-05,
      "loss": 0.3787,
      "step": 541
    },
    {
      "epoch": 2.437394722066255,
      "grad_norm": 0.7168416860410493,
      "learning_rate": 4e-05,
      "loss": 0.3661,
      "step": 542
    },
    {
      "epoch": 2.441886580572712,
      "grad_norm": 0.595528691914441,
      "learning_rate": 4e-05,
      "loss": 0.374,
      "step": 543
    },
    {
      "epoch": 2.446378439079169,
      "grad_norm": 0.5173570518101819,
      "learning_rate": 4e-05,
      "loss": 0.3756,
      "step": 544
    },
    {
      "epoch": 2.450870297585626,
      "grad_norm": 0.4665593460846923,
      "learning_rate": 4e-05,
      "loss": 0.3768,
      "step": 545
    },
    {
      "epoch": 2.4553621560920833,
      "grad_norm": 0.4601788670788593,
      "learning_rate": 4e-05,
      "loss": 0.353,
      "step": 546
    },
    {
      "epoch": 2.45985401459854,
      "grad_norm": 0.45683096399298145,
      "learning_rate": 4e-05,
      "loss": 0.3767,
      "step": 547
    },
    {
      "epoch": 2.464345873104997,
      "grad_norm": 0.4763232868754955,
      "learning_rate": 4e-05,
      "loss": 0.3956,
      "step": 548
    },
    {
      "epoch": 2.4688377316114543,
      "grad_norm": 0.45803723650510947,
      "learning_rate": 4e-05,
      "loss": 0.3639,
      "step": 549
    },
    {
      "epoch": 2.473329590117911,
      "grad_norm": 0.38471171900114937,
      "learning_rate": 4e-05,
      "loss": 0.3907,
      "step": 550
    },
    {
      "epoch": 2.4778214486243684,
      "grad_norm": 0.3420018199154973,
      "learning_rate": 4e-05,
      "loss": 0.3866,
      "step": 551
    },
    {
      "epoch": 2.482313307130825,
      "grad_norm": 0.4008009597126848,
      "learning_rate": 4e-05,
      "loss": 0.3697,
      "step": 552
    },
    {
      "epoch": 2.4868051656372825,
      "grad_norm": 0.4052005273407867,
      "learning_rate": 4e-05,
      "loss": 0.3768,
      "step": 553
    },
    {
      "epoch": 2.4912970241437393,
      "grad_norm": 0.4191650915252312,
      "learning_rate": 4e-05,
      "loss": 0.3686,
      "step": 554
    },
    {
      "epoch": 2.4957888826501966,
      "grad_norm": 0.41529071133374573,
      "learning_rate": 4e-05,
      "loss": 0.381,
      "step": 555
    },
    {
      "epoch": 2.5002807411566534,
      "grad_norm": 0.3807878537132418,
      "learning_rate": 4e-05,
      "loss": 0.3547,
      "step": 556
    },
    {
      "epoch": 2.5047725996631107,
      "grad_norm": 0.41554779251839863,
      "learning_rate": 4e-05,
      "loss": 0.3998,
      "step": 557
    },
    {
      "epoch": 2.5092644581695676,
      "grad_norm": 0.3777677080246465,
      "learning_rate": 4e-05,
      "loss": 0.3482,
      "step": 558
    },
    {
      "epoch": 2.513756316676025,
      "grad_norm": 0.4321715431601564,
      "learning_rate": 4e-05,
      "loss": 0.3689,
      "step": 559
    },
    {
      "epoch": 2.5182481751824817,
      "grad_norm": 0.4634955973829025,
      "learning_rate": 4e-05,
      "loss": 0.3919,
      "step": 560
    },
    {
      "epoch": 2.522740033688939,
      "grad_norm": 0.5399560563715813,
      "learning_rate": 4e-05,
      "loss": 0.3666,
      "step": 561
    },
    {
      "epoch": 2.527231892195396,
      "grad_norm": 0.46855988828506545,
      "learning_rate": 4e-05,
      "loss": 0.3728,
      "step": 562
    },
    {
      "epoch": 2.531723750701853,
      "grad_norm": 0.42199242457805825,
      "learning_rate": 4e-05,
      "loss": 0.3633,
      "step": 563
    },
    {
      "epoch": 2.53621560920831,
      "grad_norm": 0.3402367603287058,
      "learning_rate": 4e-05,
      "loss": 0.3848,
      "step": 564
    },
    {
      "epoch": 2.5407074677147667,
      "grad_norm": 0.5140486096270049,
      "learning_rate": 4e-05,
      "loss": 0.3927,
      "step": 565
    },
    {
      "epoch": 2.545199326221224,
      "grad_norm": 0.5456782039815428,
      "learning_rate": 4e-05,
      "loss": 0.3856,
      "step": 566
    },
    {
      "epoch": 2.5496911847276813,
      "grad_norm": 0.32315101829523196,
      "learning_rate": 4e-05,
      "loss": 0.3629,
      "step": 567
    },
    {
      "epoch": 2.554183043234138,
      "grad_norm": 0.45120691699097315,
      "learning_rate": 4e-05,
      "loss": 0.3588,
      "step": 568
    },
    {
      "epoch": 2.558674901740595,
      "grad_norm": 0.5392767185875726,
      "learning_rate": 4e-05,
      "loss": 0.416,
      "step": 569
    },
    {
      "epoch": 2.5631667602470523,
      "grad_norm": 0.4590345888001034,
      "learning_rate": 4e-05,
      "loss": 0.3689,
      "step": 570
    },
    {
      "epoch": 2.5676586187535095,
      "grad_norm": 0.47113782741905963,
      "learning_rate": 4e-05,
      "loss": 0.3711,
      "step": 571
    },
    {
      "epoch": 2.5721504772599664,
      "grad_norm": 0.4612119783675237,
      "learning_rate": 4e-05,
      "loss": 0.3779,
      "step": 572
    },
    {
      "epoch": 2.576642335766423,
      "grad_norm": 0.4732937554235207,
      "learning_rate": 4e-05,
      "loss": 0.3657,
      "step": 573
    },
    {
      "epoch": 2.5811341942728805,
      "grad_norm": 0.6540944933052659,
      "learning_rate": 4e-05,
      "loss": 0.4067,
      "step": 574
    },
    {
      "epoch": 2.5856260527793373,
      "grad_norm": 0.6610565743703379,
      "learning_rate": 4e-05,
      "loss": 0.3622,
      "step": 575
    },
    {
      "epoch": 2.5901179112857946,
      "grad_norm": 0.4237120349440777,
      "learning_rate": 4e-05,
      "loss": 0.3765,
      "step": 576
    },
    {
      "epoch": 2.5946097697922514,
      "grad_norm": 0.40702669651456436,
      "learning_rate": 4e-05,
      "loss": 0.3835,
      "step": 577
    },
    {
      "epoch": 2.5991016282987087,
      "grad_norm": 0.3816693893194654,
      "learning_rate": 4e-05,
      "loss": 0.3651,
      "step": 578
    },
    {
      "epoch": 2.6035934868051656,
      "grad_norm": 0.3405308565619867,
      "learning_rate": 4e-05,
      "loss": 0.3817,
      "step": 579
    },
    {
      "epoch": 2.608085345311623,
      "grad_norm": 0.37005528629482604,
      "learning_rate": 4e-05,
      "loss": 0.3783,
      "step": 580
    },
    {
      "epoch": 2.6125772038180797,
      "grad_norm": 0.5376994805335993,
      "learning_rate": 4e-05,
      "loss": 0.3797,
      "step": 581
    },
    {
      "epoch": 2.6170690623245365,
      "grad_norm": 0.5290587996066913,
      "learning_rate": 4e-05,
      "loss": 0.3889,
      "step": 582
    },
    {
      "epoch": 2.621560920830994,
      "grad_norm": 0.366157582334466,
      "learning_rate": 4e-05,
      "loss": 0.347,
      "step": 583
    },
    {
      "epoch": 2.626052779337451,
      "grad_norm": 0.4058194560050475,
      "learning_rate": 4e-05,
      "loss": 0.3867,
      "step": 584
    },
    {
      "epoch": 2.630544637843908,
      "grad_norm": 0.3973333333613782,
      "learning_rate": 4e-05,
      "loss": 0.3663,
      "step": 585
    },
    {
      "epoch": 2.6350364963503647,
      "grad_norm": 0.3177180771406964,
      "learning_rate": 4e-05,
      "loss": 0.3927,
      "step": 586
    },
    {
      "epoch": 2.639528354856822,
      "grad_norm": 0.42952724457437175,
      "learning_rate": 4e-05,
      "loss": 0.3778,
      "step": 587
    },
    {
      "epoch": 2.6440202133632793,
      "grad_norm": 0.40073393837897797,
      "learning_rate": 4e-05,
      "loss": 0.3808,
      "step": 588
    },
    {
      "epoch": 2.648512071869736,
      "grad_norm": 0.4157373428885615,
      "learning_rate": 4e-05,
      "loss": 0.3637,
      "step": 589
    },
    {
      "epoch": 2.653003930376193,
      "grad_norm": 0.4245514116872641,
      "learning_rate": 4e-05,
      "loss": 0.3871,
      "step": 590
    },
    {
      "epoch": 2.6574957888826503,
      "grad_norm": 0.3619326789289283,
      "learning_rate": 4e-05,
      "loss": 0.3812,
      "step": 591
    },
    {
      "epoch": 2.661987647389107,
      "grad_norm": 0.34579695526873777,
      "learning_rate": 4e-05,
      "loss": 0.3746,
      "step": 592
    },
    {
      "epoch": 2.6664795058955644,
      "grad_norm": 0.38367943278436895,
      "learning_rate": 4e-05,
      "loss": 0.3693,
      "step": 593
    },
    {
      "epoch": 2.670971364402021,
      "grad_norm": 0.4325822957536181,
      "learning_rate": 4e-05,
      "loss": 0.3792,
      "step": 594
    },
    {
      "epoch": 2.6754632229084785,
      "grad_norm": 0.33450888910929266,
      "learning_rate": 4e-05,
      "loss": 0.3833,
      "step": 595
    },
    {
      "epoch": 2.6799550814149353,
      "grad_norm": 0.42871681012786866,
      "learning_rate": 4e-05,
      "loss": 0.3747,
      "step": 596
    },
    {
      "epoch": 2.6844469399213926,
      "grad_norm": 0.4707140064198212,
      "learning_rate": 4e-05,
      "loss": 0.3793,
      "step": 597
    },
    {
      "epoch": 2.6889387984278494,
      "grad_norm": 0.3859152330171797,
      "learning_rate": 4e-05,
      "loss": 0.3684,
      "step": 598
    },
    {
      "epoch": 2.6934306569343067,
      "grad_norm": 0.4952832605276495,
      "learning_rate": 4e-05,
      "loss": 0.3807,
      "step": 599
    },
    {
      "epoch": 2.6979225154407636,
      "grad_norm": 0.6160465616875055,
      "learning_rate": 4e-05,
      "loss": 0.3616,
      "step": 600
    },
    {
      "epoch": 2.702414373947221,
      "grad_norm": 0.5414831874139456,
      "learning_rate": 4e-05,
      "loss": 0.3837,
      "step": 601
    },
    {
      "epoch": 2.7069062324536777,
      "grad_norm": 0.5781173627018455,
      "learning_rate": 4e-05,
      "loss": 0.3716,
      "step": 602
    },
    {
      "epoch": 2.7113980909601345,
      "grad_norm": 0.5432774782419044,
      "learning_rate": 4e-05,
      "loss": 0.3788,
      "step": 603
    },
    {
      "epoch": 2.715889949466592,
      "grad_norm": 0.38846250517317776,
      "learning_rate": 4e-05,
      "loss": 0.3767,
      "step": 604
    },
    {
      "epoch": 2.720381807973049,
      "grad_norm": 0.6080526017734954,
      "learning_rate": 4e-05,
      "loss": 0.3922,
      "step": 605
    },
    {
      "epoch": 2.724873666479506,
      "grad_norm": 0.5074245540222534,
      "learning_rate": 4e-05,
      "loss": 0.3571,
      "step": 606
    },
    {
      "epoch": 2.7293655249859627,
      "grad_norm": 0.326454216704089,
      "learning_rate": 4e-05,
      "loss": 0.3719,
      "step": 607
    },
    {
      "epoch": 2.73385738349242,
      "grad_norm": 0.5059072588721313,
      "learning_rate": 4e-05,
      "loss": 0.3724,
      "step": 608
    },
    {
      "epoch": 2.7383492419988773,
      "grad_norm": 0.48494644882426696,
      "learning_rate": 4e-05,
      "loss": 0.3831,
      "step": 609
    },
    {
      "epoch": 2.742841100505334,
      "grad_norm": 0.4444102577725821,
      "learning_rate": 4e-05,
      "loss": 0.3771,
      "step": 610
    },
    {
      "epoch": 2.747332959011791,
      "grad_norm": 0.4704368031457099,
      "learning_rate": 4e-05,
      "loss": 0.3623,
      "step": 611
    },
    {
      "epoch": 2.7518248175182483,
      "grad_norm": 0.455314589636396,
      "learning_rate": 4e-05,
      "loss": 0.3778,
      "step": 612
    },
    {
      "epoch": 2.756316676024705,
      "grad_norm": 0.4594336148176704,
      "learning_rate": 4e-05,
      "loss": 0.3742,
      "step": 613
    },
    {
      "epoch": 2.7608085345311624,
      "grad_norm": 0.41062833592312387,
      "learning_rate": 4e-05,
      "loss": 0.3712,
      "step": 614
    },
    {
      "epoch": 2.765300393037619,
      "grad_norm": 0.43687381124967845,
      "learning_rate": 4e-05,
      "loss": 0.3791,
      "step": 615
    },
    {
      "epoch": 2.7697922515440765,
      "grad_norm": 0.3548705083924286,
      "learning_rate": 4e-05,
      "loss": 0.3614,
      "step": 616
    },
    {
      "epoch": 2.7742841100505333,
      "grad_norm": 0.376269124875261,
      "learning_rate": 4e-05,
      "loss": 0.3921,
      "step": 617
    },
    {
      "epoch": 2.7787759685569906,
      "grad_norm": 0.4331020564556367,
      "learning_rate": 4e-05,
      "loss": 0.3806,
      "step": 618
    },
    {
      "epoch": 2.7832678270634474,
      "grad_norm": 0.5335610461149791,
      "learning_rate": 4e-05,
      "loss": 0.3739,
      "step": 619
    },
    {
      "epoch": 2.7877596855699043,
      "grad_norm": 0.49952139437160936,
      "learning_rate": 4e-05,
      "loss": 0.3589,
      "step": 620
    },
    {
      "epoch": 2.7922515440763616,
      "grad_norm": 0.30045987849767986,
      "learning_rate": 4e-05,
      "loss": 0.3881,
      "step": 621
    },
    {
      "epoch": 2.796743402582819,
      "grad_norm": 0.3713455962887138,
      "learning_rate": 4e-05,
      "loss": 0.379,
      "step": 622
    },
    {
      "epoch": 2.8012352610892757,
      "grad_norm": 0.5095231631055466,
      "learning_rate": 4e-05,
      "loss": 0.3881,
      "step": 623
    },
    {
      "epoch": 2.8057271195957325,
      "grad_norm": 0.5765190036815673,
      "learning_rate": 4e-05,
      "loss": 0.3714,
      "step": 624
    },
    {
      "epoch": 2.81021897810219,
      "grad_norm": 0.5808766978925778,
      "learning_rate": 4e-05,
      "loss": 0.3746,
      "step": 625
    },
    {
      "epoch": 2.814710836608647,
      "grad_norm": 0.40839384219488034,
      "learning_rate": 4e-05,
      "loss": 0.3773,
      "step": 626
    },
    {
      "epoch": 2.819202695115104,
      "grad_norm": 0.40810322934921667,
      "learning_rate": 4e-05,
      "loss": 0.3626,
      "step": 627
    },
    {
      "epoch": 2.8236945536215607,
      "grad_norm": 0.5243658974563962,
      "learning_rate": 4e-05,
      "loss": 0.3887,
      "step": 628
    },
    {
      "epoch": 2.828186412128018,
      "grad_norm": 0.5079531761789325,
      "learning_rate": 4e-05,
      "loss": 0.3811,
      "step": 629
    },
    {
      "epoch": 2.832678270634475,
      "grad_norm": 0.5296759137943663,
      "learning_rate": 4e-05,
      "loss": 0.3739,
      "step": 630
    },
    {
      "epoch": 2.837170129140932,
      "grad_norm": 0.5264572005810496,
      "learning_rate": 4e-05,
      "loss": 0.3876,
      "step": 631
    },
    {
      "epoch": 2.841661987647389,
      "grad_norm": 0.48054871678821137,
      "learning_rate": 4e-05,
      "loss": 0.3727,
      "step": 632
    },
    {
      "epoch": 2.8461538461538463,
      "grad_norm": 0.5302278156888595,
      "learning_rate": 4e-05,
      "loss": 0.3813,
      "step": 633
    },
    {
      "epoch": 2.850645704660303,
      "grad_norm": 0.42397882790251423,
      "learning_rate": 4e-05,
      "loss": 0.3852,
      "step": 634
    },
    {
      "epoch": 2.8551375631667604,
      "grad_norm": 0.4113157242754436,
      "learning_rate": 4e-05,
      "loss": 0.3895,
      "step": 635
    },
    {
      "epoch": 2.859629421673217,
      "grad_norm": 0.35804448693832264,
      "learning_rate": 4e-05,
      "loss": 0.3685,
      "step": 636
    },
    {
      "epoch": 2.8641212801796745,
      "grad_norm": 0.5157414567775632,
      "learning_rate": 4e-05,
      "loss": 0.3706,
      "step": 637
    },
    {
      "epoch": 2.8686131386861313,
      "grad_norm": 0.576203121931043,
      "learning_rate": 4e-05,
      "loss": 0.384,
      "step": 638
    },
    {
      "epoch": 2.8731049971925886,
      "grad_norm": 0.5217046502954408,
      "learning_rate": 4e-05,
      "loss": 0.3903,
      "step": 639
    },
    {
      "epoch": 2.8775968556990454,
      "grad_norm": 0.36621915396770777,
      "learning_rate": 4e-05,
      "loss": 0.3706,
      "step": 640
    },
    {
      "epoch": 2.8820887142055023,
      "grad_norm": 0.32201588702933054,
      "learning_rate": 4e-05,
      "loss": 0.3588,
      "step": 641
    },
    {
      "epoch": 2.8865805727119596,
      "grad_norm": 0.3982060575455867,
      "learning_rate": 4e-05,
      "loss": 0.3968,
      "step": 642
    },
    {
      "epoch": 2.891072431218417,
      "grad_norm": 0.7079261557763602,
      "learning_rate": 4e-05,
      "loss": 0.3749,
      "step": 643
    },
    {
      "epoch": 2.8955642897248737,
      "grad_norm": 0.758614660292448,
      "learning_rate": 4e-05,
      "loss": 0.3744,
      "step": 644
    },
    {
      "epoch": 2.9000561482313305,
      "grad_norm": 0.5291524554803048,
      "learning_rate": 4e-05,
      "loss": 0.3806,
      "step": 645
    },
    {
      "epoch": 2.904548006737788,
      "grad_norm": 0.44286311181518556,
      "learning_rate": 4e-05,
      "loss": 0.3798,
      "step": 646
    },
    {
      "epoch": 2.909039865244245,
      "grad_norm": 0.38810472267759816,
      "learning_rate": 4e-05,
      "loss": 0.3687,
      "step": 647
    },
    {
      "epoch": 2.913531723750702,
      "grad_norm": 0.4138048136293526,
      "learning_rate": 4e-05,
      "loss": 0.3794,
      "step": 648
    },
    {
      "epoch": 2.9180235822571587,
      "grad_norm": 0.42388237886562125,
      "learning_rate": 4e-05,
      "loss": 0.3815,
      "step": 649
    },
    {
      "epoch": 2.922515440763616,
      "grad_norm": 0.3810507958404275,
      "learning_rate": 4e-05,
      "loss": 0.3762,
      "step": 650
    },
    {
      "epoch": 2.927007299270073,
      "grad_norm": 0.3191891626789646,
      "learning_rate": 4e-05,
      "loss": 0.3767,
      "step": 651
    },
    {
      "epoch": 2.93149915777653,
      "grad_norm": 0.3634778667506236,
      "learning_rate": 4e-05,
      "loss": 0.3714,
      "step": 652
    },
    {
      "epoch": 2.935991016282987,
      "grad_norm": 0.45400053242994165,
      "learning_rate": 4e-05,
      "loss": 0.3752,
      "step": 653
    },
    {
      "epoch": 2.9404828747894443,
      "grad_norm": 0.39290702253242793,
      "learning_rate": 4e-05,
      "loss": 0.3652,
      "step": 654
    },
    {
      "epoch": 2.944974733295901,
      "grad_norm": 0.32237692876989965,
      "learning_rate": 4e-05,
      "loss": 0.361,
      "step": 655
    },
    {
      "epoch": 2.9494665918023584,
      "grad_norm": 0.38200445268951994,
      "learning_rate": 4e-05,
      "loss": 0.3958,
      "step": 656
    },
    {
      "epoch": 2.953958450308815,
      "grad_norm": 0.3788057779796393,
      "learning_rate": 4e-05,
      "loss": 0.3702,
      "step": 657
    },
    {
      "epoch": 2.958450308815272,
      "grad_norm": 0.3736082722255773,
      "learning_rate": 4e-05,
      "loss": 0.3769,
      "step": 658
    },
    {
      "epoch": 2.9629421673217293,
      "grad_norm": 0.38499781125815363,
      "learning_rate": 4e-05,
      "loss": 0.3883,
      "step": 659
    },
    {
      "epoch": 2.9674340258281866,
      "grad_norm": 0.3506934822312221,
      "learning_rate": 4e-05,
      "loss": 0.3723,
      "step": 660
    },
    {
      "epoch": 2.9719258843346434,
      "grad_norm": 0.4322668092983293,
      "learning_rate": 4e-05,
      "loss": 0.3788,
      "step": 661
    },
    {
      "epoch": 2.9764177428411003,
      "grad_norm": 0.4082763714115294,
      "learning_rate": 4e-05,
      "loss": 0.3778,
      "step": 662
    },
    {
      "epoch": 2.9809096013475576,
      "grad_norm": 0.4516039147761112,
      "learning_rate": 4e-05,
      "loss": 0.3793,
      "step": 663
    },
    {
      "epoch": 2.985401459854015,
      "grad_norm": 0.729468953523648,
      "learning_rate": 4e-05,
      "loss": 0.3704,
      "step": 664
    },
    {
      "epoch": 2.9898933183604717,
      "grad_norm": 0.8936941793085164,
      "learning_rate": 4e-05,
      "loss": 0.3751,
      "step": 665
    },
    {
      "epoch": 2.9943851768669285,
      "grad_norm": 0.9158738572591631,
      "learning_rate": 4e-05,
      "loss": 0.3822,
      "step": 666
    },
    {
      "epoch": 2.9943851768669285,
      "step": 666,
      "total_flos": 7.06867120956244e+18,
      "train_loss": 0.4317017092450603,
      "train_runtime": 110001.8378,
      "train_samples_per_second": 3.108,
      "train_steps_per_second": 0.006
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 666,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.06867120956244e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}