{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.40100250626566414,
  "eval_steps": 500,
  "global_step": 320,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0012531328320802004,
      "grad_norm": 0.1519206017255783,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 1.8379,
      "step": 1
    },
    {
      "epoch": 0.002506265664160401,
      "grad_norm": 0.13846151530742645,
      "learning_rate": 4.800000000000001e-06,
      "loss": 1.6569,
      "step": 2
    },
    {
      "epoch": 0.0037593984962406013,
      "grad_norm": 0.13162349164485931,
      "learning_rate": 7.2e-06,
      "loss": 1.6083,
      "step": 3
    },
    {
      "epoch": 0.005012531328320802,
      "grad_norm": 0.1345846951007843,
      "learning_rate": 9.600000000000001e-06,
      "loss": 1.6603,
      "step": 4
    },
    {
      "epoch": 0.006265664160401002,
      "grad_norm": 0.1223267987370491,
      "learning_rate": 1.2e-05,
      "loss": 1.6579,
      "step": 5
    },
    {
      "epoch": 0.007518796992481203,
      "grad_norm": 0.09618138521909714,
      "learning_rate": 1.44e-05,
      "loss": 1.5689,
      "step": 6
    },
    {
      "epoch": 0.008771929824561403,
      "grad_norm": 0.09933976829051971,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 1.4663,
      "step": 7
    },
    {
      "epoch": 0.010025062656641603,
      "grad_norm": 0.1198974996805191,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 1.734,
      "step": 8
    },
    {
      "epoch": 0.011278195488721804,
      "grad_norm": 0.15424160659313202,
      "learning_rate": 2.16e-05,
      "loss": 1.6767,
      "step": 9
    },
    {
      "epoch": 0.012531328320802004,
      "grad_norm": 0.1701776534318924,
      "learning_rate": 2.4e-05,
      "loss": 1.5974,
      "step": 10
    },
    {
      "epoch": 0.013784461152882205,
      "grad_norm": 0.15073753893375397,
      "learning_rate": 2.64e-05,
      "loss": 1.5259,
      "step": 11
    },
    {
      "epoch": 0.015037593984962405,
      "grad_norm": 0.14940425753593445,
      "learning_rate": 2.88e-05,
      "loss": 1.6349,
      "step": 12
    },
    {
      "epoch": 0.016290726817042606,
      "grad_norm": 0.12604176998138428,
      "learning_rate": 3.12e-05,
      "loss": 1.6443,
      "step": 13
    },
    {
      "epoch": 0.017543859649122806,
      "grad_norm": 0.15434400737285614,
      "learning_rate": 3.3600000000000004e-05,
      "loss": 1.6378,
      "step": 14
    },
    {
      "epoch": 0.018796992481203006,
      "grad_norm": 0.12647251784801483,
      "learning_rate": 3.6e-05,
      "loss": 1.542,
      "step": 15
    },
    {
      "epoch": 0.020050125313283207,
      "grad_norm": 0.1258278489112854,
      "learning_rate": 3.8400000000000005e-05,
      "loss": 1.6091,
      "step": 16
    },
    {
      "epoch": 0.021303258145363407,
      "grad_norm": 0.09623159468173981,
      "learning_rate": 4.08e-05,
      "loss": 1.5555,
      "step": 17
    },
    {
      "epoch": 0.022556390977443608,
      "grad_norm": 0.10304850339889526,
      "learning_rate": 4.32e-05,
      "loss": 1.5517,
      "step": 18
    },
    {
      "epoch": 0.023809523809523808,
      "grad_norm": 0.09278815984725952,
      "learning_rate": 4.5600000000000004e-05,
      "loss": 1.6436,
      "step": 19
    },
    {
      "epoch": 0.02506265664160401,
      "grad_norm": 0.08530683070421219,
      "learning_rate": 4.8e-05,
      "loss": 1.5664,
      "step": 20
    },
    {
      "epoch": 0.02631578947368421,
      "grad_norm": 0.08876050263643265,
      "learning_rate": 5.04e-05,
      "loss": 1.5628,
      "step": 21
    },
    {
      "epoch": 0.02756892230576441,
      "grad_norm": 0.09901795536279678,
      "learning_rate": 5.28e-05,
      "loss": 1.5402,
      "step": 22
    },
    {
      "epoch": 0.02882205513784461,
      "grad_norm": 0.09156398475170135,
      "learning_rate": 5.520000000000001e-05,
      "loss": 1.606,
      "step": 23
    },
    {
      "epoch": 0.03007518796992481,
      "grad_norm": 0.09371288120746613,
      "learning_rate": 5.76e-05,
      "loss": 1.566,
      "step": 24
    },
    {
      "epoch": 0.03132832080200501,
      "grad_norm": 0.0892362967133522,
      "learning_rate": 6e-05,
      "loss": 1.6193,
      "step": 25
    },
    {
      "epoch": 0.03258145363408521,
      "grad_norm": 0.08784804493188858,
      "learning_rate": 5.99997522398708e-05,
      "loss": 1.5756,
      "step": 26
    },
    {
      "epoch": 0.03383458646616541,
      "grad_norm": 0.08121798187494278,
      "learning_rate": 5.999900896357553e-05,
      "loss": 1.5085,
      "step": 27
    },
    {
      "epoch": 0.03508771929824561,
      "grad_norm": 0.09490139782428741,
      "learning_rate": 5.999777018339115e-05,
      "loss": 1.5692,
      "step": 28
    },
    {
      "epoch": 0.03634085213032581,
      "grad_norm": 0.1026797667145729,
      "learning_rate": 5.999603591977901e-05,
      "loss": 1.5761,
      "step": 29
    },
    {
      "epoch": 0.03759398496240601,
      "grad_norm": 0.09289643168449402,
      "learning_rate": 5.999380620138454e-05,
      "loss": 1.5612,
      "step": 30
    },
    {
      "epoch": 0.03884711779448621,
      "grad_norm": 0.09448801726102829,
      "learning_rate": 5.9991081065036745e-05,
      "loss": 1.6029,
      "step": 31
    },
    {
      "epoch": 0.040100250626566414,
      "grad_norm": 0.09210563451051712,
      "learning_rate": 5.998786055574766e-05,
      "loss": 1.5229,
      "step": 32
    },
    {
      "epoch": 0.041353383458646614,
      "grad_norm": 0.09483816474676132,
      "learning_rate": 5.998414472671151e-05,
      "loss": 1.5753,
      "step": 33
    },
    {
      "epoch": 0.042606516290726815,
      "grad_norm": 0.0975150614976883,
      "learning_rate": 5.997993363930393e-05,
      "loss": 1.654,
      "step": 34
    },
    {
      "epoch": 0.043859649122807015,
      "grad_norm": 0.10024327784776688,
      "learning_rate": 5.997522736308089e-05,
      "loss": 1.5868,
      "step": 35
    },
    {
      "epoch": 0.045112781954887216,
      "grad_norm": 0.08656363934278488,
      "learning_rate": 5.9970025975777576e-05,
      "loss": 1.5087,
      "step": 36
    },
    {
      "epoch": 0.046365914786967416,
      "grad_norm": 0.09871859848499298,
      "learning_rate": 5.996432956330705e-05,
      "loss": 1.4932,
      "step": 37
    },
    {
      "epoch": 0.047619047619047616,
      "grad_norm": 0.11055205762386322,
      "learning_rate": 5.9958138219758926e-05,
      "loss": 1.5467,
      "step": 38
    },
    {
      "epoch": 0.04887218045112782,
      "grad_norm": 0.11448398977518082,
      "learning_rate": 5.995145204739774e-05,
      "loss": 1.5205,
      "step": 39
    },
    {
      "epoch": 0.05012531328320802,
      "grad_norm": 0.09594336152076721,
      "learning_rate": 5.994427115666128e-05,
      "loss": 1.4101,
      "step": 40
    },
    {
      "epoch": 0.05137844611528822,
      "grad_norm": 0.1022314578294754,
      "learning_rate": 5.993659566615878e-05,
      "loss": 1.5726,
      "step": 41
    },
    {
      "epoch": 0.05263157894736842,
      "grad_norm": 0.0918896496295929,
      "learning_rate": 5.9928425702668936e-05,
      "loss": 1.4416,
      "step": 42
    },
    {
      "epoch": 0.05388471177944862,
      "grad_norm": 0.11478869616985321,
      "learning_rate": 5.9919761401137845e-05,
      "loss": 1.4871,
      "step": 43
    },
    {
      "epoch": 0.05513784461152882,
      "grad_norm": 0.09935441613197327,
      "learning_rate": 5.991060290467671e-05,
      "loss": 1.5057,
      "step": 44
    },
    {
      "epoch": 0.05639097744360902,
      "grad_norm": 0.1036807969212532,
      "learning_rate": 5.990095036455958e-05,
      "loss": 1.6244,
      "step": 45
    },
    {
      "epoch": 0.05764411027568922,
      "grad_norm": 0.10617007315158844,
      "learning_rate": 5.989080394022074e-05,
      "loss": 1.6287,
      "step": 46
    },
    {
      "epoch": 0.05889724310776942,
      "grad_norm": 0.09776347130537033,
      "learning_rate": 5.988016379925215e-05,
      "loss": 1.5421,
      "step": 47
    },
    {
      "epoch": 0.06015037593984962,
      "grad_norm": 0.10930298268795013,
      "learning_rate": 5.986903011740067e-05,
      "loss": 1.6162,
      "step": 48
    },
    {
      "epoch": 0.06140350877192982,
      "grad_norm": 0.10414919257164001,
      "learning_rate": 5.985740307856512e-05,
      "loss": 1.5381,
      "step": 49
    },
    {
      "epoch": 0.06265664160401002,
      "grad_norm": 0.10629701614379883,
      "learning_rate": 5.984528287479328e-05,
      "loss": 1.6127,
      "step": 50
    },
    {
      "epoch": 0.06390977443609022,
      "grad_norm": 0.11536388099193573,
      "learning_rate": 5.983266970627869e-05,
      "loss": 1.5455,
      "step": 51
    },
    {
      "epoch": 0.06516290726817042,
      "grad_norm": 0.12118639051914215,
      "learning_rate": 5.9819563781357385e-05,
      "loss": 1.5043,
      "step": 52
    },
    {
      "epoch": 0.06641604010025062,
      "grad_norm": 0.11645355075597763,
      "learning_rate": 5.98059653165044e-05,
      "loss": 1.5473,
      "step": 53
    },
    {
      "epoch": 0.06766917293233082,
      "grad_norm": 0.12099536508321762,
      "learning_rate": 5.9791874536330225e-05,
      "loss": 1.5482,
      "step": 54
    },
    {
      "epoch": 0.06892230576441102,
      "grad_norm": 0.10795404016971588,
      "learning_rate": 5.97772916735771e-05,
      "loss": 1.5326,
      "step": 55
    },
    {
      "epoch": 0.07017543859649122,
      "grad_norm": 0.09979701787233353,
      "learning_rate": 5.9762216969115154e-05,
      "loss": 1.4133,
      "step": 56
    },
    {
      "epoch": 0.07142857142857142,
      "grad_norm": 0.12733447551727295,
      "learning_rate": 5.974665067193844e-05,
      "loss": 1.5839,
      "step": 57
    },
    {
      "epoch": 0.07268170426065163,
      "grad_norm": 0.11343546956777573,
      "learning_rate": 5.97305930391608e-05,
      "loss": 1.524,
      "step": 58
    },
    {
      "epoch": 0.07393483709273183,
      "grad_norm": 0.11789534986019135,
      "learning_rate": 5.971404433601165e-05,
      "loss": 1.4677,
      "step": 59
    },
    {
      "epoch": 0.07518796992481203,
      "grad_norm": 0.10624619573354721,
      "learning_rate": 5.969700483583159e-05,
      "loss": 1.4893,
      "step": 60
    },
    {
      "epoch": 0.07644110275689223,
      "grad_norm": 0.10631600022315979,
      "learning_rate": 5.967947482006786e-05,
      "loss": 1.573,
      "step": 61
    },
    {
      "epoch": 0.07769423558897243,
      "grad_norm": 0.12548910081386566,
      "learning_rate": 5.9661454578269724e-05,
      "loss": 1.5625,
      "step": 62
    },
    {
      "epoch": 0.07894736842105263,
      "grad_norm": 0.1474038064479828,
      "learning_rate": 5.964294440808368e-05,
      "loss": 1.6886,
      "step": 63
    },
    {
      "epoch": 0.08020050125313283,
      "grad_norm": 0.12192469835281372,
      "learning_rate": 5.962394461524854e-05,
      "loss": 1.5252,
      "step": 64
    },
    {
      "epoch": 0.08145363408521303,
      "grad_norm": 0.11869639158248901,
      "learning_rate": 5.960445551359037e-05,
      "loss": 1.5262,
      "step": 65
    },
    {
      "epoch": 0.08270676691729323,
      "grad_norm": 0.12624335289001465,
      "learning_rate": 5.958447742501735e-05,
      "loss": 1.5645,
      "step": 66
    },
    {
      "epoch": 0.08395989974937343,
      "grad_norm": 0.11716689169406891,
      "learning_rate": 5.9564010679514376e-05,
      "loss": 1.47,
      "step": 67
    },
    {
      "epoch": 0.08521303258145363,
      "grad_norm": 0.12144263088703156,
      "learning_rate": 5.954305561513769e-05,
      "loss": 1.6181,
      "step": 68
    },
    {
      "epoch": 0.08646616541353383,
      "grad_norm": 0.12713098526000977,
      "learning_rate": 5.9521612578009255e-05,
      "loss": 1.6227,
      "step": 69
    },
    {
      "epoch": 0.08771929824561403,
      "grad_norm": 0.1043514758348465,
      "learning_rate": 5.9499681922311046e-05,
      "loss": 1.4522,
      "step": 70
    },
    {
      "epoch": 0.08897243107769423,
      "grad_norm": 0.11064239591360092,
      "learning_rate": 5.947726401027921e-05,
      "loss": 1.4685,
      "step": 71
    },
    {
      "epoch": 0.09022556390977443,
      "grad_norm": 0.11946432292461395,
      "learning_rate": 5.945435921219806e-05,
      "loss": 1.4165,
      "step": 72
    },
    {
      "epoch": 0.09147869674185463,
      "grad_norm": 0.11144128441810608,
      "learning_rate": 5.943096790639398e-05,
      "loss": 1.5081,
      "step": 73
    },
    {
      "epoch": 0.09273182957393483,
      "grad_norm": 0.11953038722276688,
      "learning_rate": 5.9407090479229166e-05,
      "loss": 1.5773,
      "step": 74
    },
    {
      "epoch": 0.09398496240601503,
      "grad_norm": 0.12432811409235,
      "learning_rate": 5.938272732509525e-05,
      "loss": 1.528,
      "step": 75
    },
    {
      "epoch": 0.09523809523809523,
      "grad_norm": 0.13081413507461548,
      "learning_rate": 5.9357878846406776e-05,
      "loss": 1.5024,
      "step": 76
    },
    {
      "epoch": 0.09649122807017543,
      "grad_norm": 0.12342929095029831,
      "learning_rate": 5.933254545359456e-05,
      "loss": 1.478,
      "step": 77
    },
    {
      "epoch": 0.09774436090225563,
      "grad_norm": 0.12129031121730804,
      "learning_rate": 5.9306727565098925e-05,
      "loss": 1.531,
      "step": 78
    },
    {
      "epoch": 0.09899749373433583,
      "grad_norm": 0.12284113466739655,
      "learning_rate": 5.928042560736275e-05,
      "loss": 1.6204,
      "step": 79
    },
    {
      "epoch": 0.10025062656641603,
      "grad_norm": 0.14579783380031586,
      "learning_rate": 5.9253640014824466e-05,
      "loss": 1.4382,
      "step": 80
    },
    {
      "epoch": 0.10150375939849623,
      "grad_norm": 0.13060207664966583,
      "learning_rate": 5.9226371229910885e-05,
      "loss": 1.5713,
      "step": 81
    },
    {
      "epoch": 0.10275689223057644,
      "grad_norm": 0.12609316408634186,
      "learning_rate": 5.919861970302982e-05,
      "loss": 1.5312,
      "step": 82
    },
    {
      "epoch": 0.10401002506265664,
      "grad_norm": 0.12477698922157288,
      "learning_rate": 5.9170385892562755e-05,
      "loss": 1.5538,
      "step": 83
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 0.14683856070041656,
      "learning_rate": 5.914167026485719e-05,
      "loss": 1.5318,
      "step": 84
    },
    {
      "epoch": 0.10651629072681704,
      "grad_norm": 0.12768588960170746,
      "learning_rate": 5.9112473294218954e-05,
      "loss": 1.5587,
      "step": 85
    },
    {
      "epoch": 0.10776942355889724,
      "grad_norm": 0.1492781639099121,
      "learning_rate": 5.9082795462904396e-05,
      "loss": 1.5623,
      "step": 86
    },
    {
      "epoch": 0.10902255639097744,
      "grad_norm": 0.12561501562595367,
      "learning_rate": 5.905263726111241e-05,
      "loss": 1.5232,
      "step": 87
    },
    {
      "epoch": 0.11027568922305764,
      "grad_norm": 0.11631227284669876,
      "learning_rate": 5.902199918697634e-05,
      "loss": 1.505,
      "step": 88
    },
    {
      "epoch": 0.11152882205513784,
      "grad_norm": 0.126926988363266,
      "learning_rate": 5.899088174655571e-05,
      "loss": 1.5001,
      "step": 89
    },
    {
      "epoch": 0.11278195488721804,
      "grad_norm": 0.13932526111602783,
      "learning_rate": 5.8959285453827936e-05,
      "loss": 1.4141,
      "step": 90
    },
    {
      "epoch": 0.11403508771929824,
      "grad_norm": 0.13378114998340607,
      "learning_rate": 5.8927210830679785e-05,
      "loss": 1.5173,
      "step": 91
    },
    {
      "epoch": 0.11528822055137844,
      "grad_norm": 0.12687304615974426,
      "learning_rate": 5.889465840689878e-05,
      "loss": 1.3759,
      "step": 92
    },
    {
      "epoch": 0.11654135338345864,
      "grad_norm": 0.12497436255216599,
      "learning_rate": 5.886162872016442e-05,
      "loss": 1.6235,
      "step": 93
    },
    {
      "epoch": 0.11779448621553884,
      "grad_norm": 0.12597259879112244,
      "learning_rate": 5.882812231603937e-05,
      "loss": 1.4533,
      "step": 94
    },
    {
      "epoch": 0.11904761904761904,
      "grad_norm": 0.14128872752189636,
      "learning_rate": 5.879413974796033e-05,
      "loss": 1.4717,
      "step": 95
    },
    {
      "epoch": 0.12030075187969924,
      "grad_norm": 0.13640277087688446,
      "learning_rate": 5.8759681577229014e-05,
      "loss": 1.4978,
      "step": 96
    },
    {
      "epoch": 0.12155388471177944,
      "grad_norm": 0.11883015185594559,
      "learning_rate": 5.8724748373002805e-05,
      "loss": 1.5333,
      "step": 97
    },
    {
      "epoch": 0.12280701754385964,
      "grad_norm": 0.14055690169334412,
      "learning_rate": 5.868934071228539e-05,
      "loss": 1.5967,
      "step": 98
    },
    {
      "epoch": 0.12406015037593984,
      "grad_norm": 0.1388673484325409,
      "learning_rate": 5.8653459179917196e-05,
      "loss": 1.4698,
      "step": 99
    },
    {
      "epoch": 0.12531328320802004,
      "grad_norm": 0.14991509914398193,
      "learning_rate": 5.861710436856577e-05,
      "loss": 1.565,
      "step": 100
    },
    {
      "epoch": 0.12656641604010024,
      "grad_norm": 0.15015804767608643,
      "learning_rate": 5.8580276878715964e-05,
      "loss": 1.5374,
      "step": 101
    },
    {
      "epoch": 0.12781954887218044,
      "grad_norm": 0.14958204329013824,
      "learning_rate": 5.854297731866002e-05,
      "loss": 1.3789,
      "step": 102
    },
    {
      "epoch": 0.12907268170426064,
      "grad_norm": 0.1282397210597992,
      "learning_rate": 5.850520630448752e-05,
      "loss": 1.5317,
      "step": 103
    },
    {
      "epoch": 0.13032581453634084,
      "grad_norm": 0.12608109414577484,
      "learning_rate": 5.8466964460075225e-05,
      "loss": 1.4828,
      "step": 104
    },
    {
      "epoch": 0.13157894736842105,
      "grad_norm": 0.17321082949638367,
      "learning_rate": 5.8428252417076766e-05,
      "loss": 1.4825,
      "step": 105
    },
    {
      "epoch": 0.13283208020050125,
      "grad_norm": 0.135195791721344,
      "learning_rate": 5.838907081491219e-05,
      "loss": 1.5631,
      "step": 106
    },
    {
      "epoch": 0.13408521303258145,
      "grad_norm": 0.12379708886146545,
      "learning_rate": 5.8349420300757393e-05,
      "loss": 1.5213,
      "step": 107
    },
    {
      "epoch": 0.13533834586466165,
      "grad_norm": 0.13356727361679077,
      "learning_rate": 5.830930152953351e-05,
      "loss": 1.5834,
      "step": 108
    },
    {
      "epoch": 0.13659147869674185,
      "grad_norm": 0.13438117504119873,
      "learning_rate": 5.8268715163895984e-05,
      "loss": 1.5039,
      "step": 109
    },
    {
      "epoch": 0.13784461152882205,
      "grad_norm": 0.14895889163017273,
      "learning_rate": 5.82276618742237e-05,
      "loss": 1.5213,
      "step": 110
    },
    {
      "epoch": 0.13909774436090225,
      "grad_norm": 0.13232296705245972,
      "learning_rate": 5.818614233860789e-05,
      "loss": 1.4048,
      "step": 111
    },
    {
      "epoch": 0.14035087719298245,
      "grad_norm": 0.13256017863750458,
      "learning_rate": 5.8144157242840904e-05,
      "loss": 1.5077,
      "step": 112
    },
    {
      "epoch": 0.14160401002506265,
      "grad_norm": 0.13263653218746185,
      "learning_rate": 5.810170728040494e-05,
      "loss": 1.3955,
      "step": 113
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 0.12680189311504364,
      "learning_rate": 5.8058793152460524e-05,
      "loss": 1.4194,
      "step": 114
    },
    {
      "epoch": 0.14411027568922305,
      "grad_norm": 0.13666822016239166,
      "learning_rate": 5.801541556783501e-05,
      "loss": 1.4234,
      "step": 115
    },
    {
      "epoch": 0.14536340852130325,
      "grad_norm": 0.14599356055259705,
      "learning_rate": 5.7971575243010775e-05,
      "loss": 1.5034,
      "step": 116
    },
    {
      "epoch": 0.14661654135338345,
      "grad_norm": 0.14143531024456024,
      "learning_rate": 5.792727290211347e-05,
      "loss": 1.4257,
      "step": 117
    },
    {
      "epoch": 0.14786967418546365,
      "grad_norm": 0.14613430202007294,
      "learning_rate": 5.7882509276899995e-05,
      "loss": 1.5734,
      "step": 118
    },
    {
      "epoch": 0.14912280701754385,
      "grad_norm": 0.13908414542675018,
      "learning_rate": 5.7837285106746455e-05,
      "loss": 1.4259,
      "step": 119
    },
    {
      "epoch": 0.15037593984962405,
      "grad_norm": 0.14299297332763672,
      "learning_rate": 5.779160113863594e-05,
      "loss": 1.6024,
      "step": 120
    },
    {
      "epoch": 0.15162907268170425,
      "grad_norm": 0.1389327496290207,
      "learning_rate": 5.774545812714617e-05,
      "loss": 1.5475,
      "step": 121
    },
    {
      "epoch": 0.15288220551378445,
      "grad_norm": 0.12398528307676315,
      "learning_rate": 5.769885683443704e-05,
      "loss": 1.4205,
      "step": 122
    },
    {
      "epoch": 0.15413533834586465,
      "grad_norm": 0.1524042785167694,
      "learning_rate": 5.765179803023805e-05,
      "loss": 1.5462,
      "step": 123
    },
    {
      "epoch": 0.15538847117794485,
      "grad_norm": 0.13744120299816132,
      "learning_rate": 5.7604282491835546e-05,
      "loss": 1.396,
      "step": 124
    },
    {
      "epoch": 0.15664160401002505,
      "grad_norm": 0.1439579576253891,
      "learning_rate": 5.755631100405994e-05,
      "loss": 1.3751,
      "step": 125
    },
    {
      "epoch": 0.15789473684210525,
      "grad_norm": 0.1428079754114151,
      "learning_rate": 5.750788435927268e-05,
      "loss": 1.4661,
      "step": 126
    },
    {
      "epoch": 0.15914786967418545,
      "grad_norm": 0.15680895745754242,
      "learning_rate": 5.7459003357353214e-05,
      "loss": 1.5036,
      "step": 127
    },
    {
      "epoch": 0.16040100250626566,
      "grad_norm": 0.1669396311044693,
      "learning_rate": 5.740966880568579e-05,
      "loss": 1.5688,
      "step": 128
    },
    {
      "epoch": 0.16165413533834586,
      "grad_norm": 0.14206421375274658,
      "learning_rate": 5.735988151914606e-05,
      "loss": 1.4673,
      "step": 129
    },
    {
      "epoch": 0.16290726817042606,
      "grad_norm": 0.14038872718811035,
      "learning_rate": 5.730964232008765e-05,
      "loss": 1.4081,
      "step": 130
    },
    {
      "epoch": 0.16416040100250626,
      "grad_norm": 0.14589321613311768,
      "learning_rate": 5.72589520383286e-05,
      "loss": 1.4524,
      "step": 131
    },
    {
      "epoch": 0.16541353383458646,
      "grad_norm": 0.12820565700531006,
      "learning_rate": 5.72078115111376e-05,
      "loss": 1.513,
      "step": 132
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 0.14205814898014069,
      "learning_rate": 5.715622158322027e-05,
      "loss": 1.4528,
      "step": 133
    },
    {
      "epoch": 0.16791979949874686,
      "grad_norm": 0.14166466891765594,
      "learning_rate": 5.7104183106705065e-05,
      "loss": 1.4185,
      "step": 134
    },
    {
      "epoch": 0.16917293233082706,
      "grad_norm": 0.1772213876247406,
      "learning_rate": 5.705169694112929e-05,
      "loss": 1.6193,
      "step": 135
    },
    {
      "epoch": 0.17042606516290726,
      "grad_norm": 0.14686815440654755,
      "learning_rate": 5.6998763953424906e-05,
      "loss": 1.4533,
      "step": 136
    },
    {
      "epoch": 0.17167919799498746,
      "grad_norm": 0.1334281861782074,
      "learning_rate": 5.694538501790417e-05,
      "loss": 1.4905,
      "step": 137
    },
    {
      "epoch": 0.17293233082706766,
      "grad_norm": 0.1628011018037796,
      "learning_rate": 5.689156101624519e-05,
      "loss": 1.4211,
      "step": 138
    },
    {
      "epoch": 0.17418546365914786,
      "grad_norm": 0.15260601043701172,
      "learning_rate": 5.683729283747743e-05,
      "loss": 1.487,
      "step": 139
    },
    {
      "epoch": 0.17543859649122806,
      "grad_norm": 0.15220171213150024,
      "learning_rate": 5.6782581377966954e-05,
      "loss": 1.4547,
      "step": 140
    },
    {
      "epoch": 0.17669172932330826,
      "grad_norm": 0.13987644016742706,
      "learning_rate": 5.672742754140162e-05,
      "loss": 1.4544,
      "step": 141
    },
    {
      "epoch": 0.17794486215538846,
      "grad_norm": 0.1494978666305542,
      "learning_rate": 5.6671832238776246e-05,
      "loss": 1.3525,
      "step": 142
    },
    {
      "epoch": 0.17919799498746866,
      "grad_norm": 0.15626025199890137,
      "learning_rate": 5.661579638837744e-05,
      "loss": 1.4859,
      "step": 143
    },
    {
      "epoch": 0.18045112781954886,
      "grad_norm": 0.14824387431144714,
      "learning_rate": 5.655932091576849e-05,
      "loss": 1.487,
      "step": 144
    },
    {
      "epoch": 0.18170426065162906,
      "grad_norm": 0.15267407894134521,
      "learning_rate": 5.6502406753774104e-05,
      "loss": 1.507,
      "step": 145
    },
    {
      "epoch": 0.18295739348370926,
      "grad_norm": 0.14779289066791534,
      "learning_rate": 5.644505484246495e-05,
      "loss": 1.4844,
      "step": 146
    },
    {
      "epoch": 0.18421052631578946,
      "grad_norm": 0.153567373752594,
      "learning_rate": 5.638726612914217e-05,
      "loss": 1.5149,
      "step": 147
    },
    {
      "epoch": 0.18546365914786966,
      "grad_norm": 0.13506165146827698,
      "learning_rate": 5.632904156832169e-05,
      "loss": 1.4841,
      "step": 148
    },
    {
      "epoch": 0.18671679197994986,
      "grad_norm": 0.153153657913208,
      "learning_rate": 5.62703821217185e-05,
      "loss": 1.5633,
      "step": 149
    },
    {
      "epoch": 0.18796992481203006,
      "grad_norm": 0.14939852058887482,
      "learning_rate": 5.621128875823073e-05,
      "loss": 1.3896,
      "step": 150
    },
    {
      "epoch": 0.18922305764411027,
      "grad_norm": 0.1525796800851822,
      "learning_rate": 5.615176245392367e-05,
      "loss": 1.5226,
      "step": 151
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 0.15289181470870972,
      "learning_rate": 5.609180419201366e-05,
      "loss": 1.5143,
      "step": 152
    },
    {
      "epoch": 0.19172932330827067,
      "grad_norm": 0.15484128892421722,
      "learning_rate": 5.603141496285179e-05,
      "loss": 1.3956,
      "step": 153
    },
    {
      "epoch": 0.19298245614035087,
      "grad_norm": 0.1458120048046112,
      "learning_rate": 5.597059576390762e-05,
      "loss": 1.453,
      "step": 154
    },
    {
      "epoch": 0.19423558897243107,
      "grad_norm": 0.14109253883361816,
      "learning_rate": 5.590934759975267e-05,
      "loss": 1.4137,
      "step": 155
    },
    {
      "epoch": 0.19548872180451127,
      "grad_norm": 0.14820784330368042,
      "learning_rate": 5.584767148204379e-05,
      "loss": 1.4878,
      "step": 156
    },
    {
      "epoch": 0.19674185463659147,
      "grad_norm": 0.16532014310359955,
      "learning_rate": 5.578556842950651e-05,
      "loss": 1.4942,
      "step": 157
    },
    {
      "epoch": 0.19799498746867167,
      "grad_norm": 0.15438464283943176,
      "learning_rate": 5.572303946791819e-05,
      "loss": 1.353,
      "step": 158
    },
    {
      "epoch": 0.19924812030075187,
      "grad_norm": 0.150332972407341,
      "learning_rate": 5.566008563009107e-05,
      "loss": 1.469,
      "step": 159
    },
    {
      "epoch": 0.20050125313283207,
      "grad_norm": 0.1449950933456421,
      "learning_rate": 5.5596707955855215e-05,
      "loss": 1.4347,
      "step": 160
    },
    {
      "epoch": 0.20175438596491227,
      "grad_norm": 0.15944981575012207,
      "learning_rate": 5.553290749204134e-05,
      "loss": 1.512,
      "step": 161
    },
    {
      "epoch": 0.20300751879699247,
      "grad_norm": 0.14288978278636932,
      "learning_rate": 5.546868529246352e-05,
      "loss": 1.4669,
      "step": 162
    },
    {
      "epoch": 0.20426065162907267,
      "grad_norm": 0.16595055162906647,
      "learning_rate": 5.54040424179018e-05,
      "loss": 1.6102,
      "step": 163
    },
    {
      "epoch": 0.20551378446115287,
      "grad_norm": 0.1445576548576355,
      "learning_rate": 5.533897993608463e-05,
      "loss": 1.3255,
      "step": 164
    },
    {
      "epoch": 0.20676691729323307,
      "grad_norm": 0.15171661972999573,
      "learning_rate": 5.527349892167127e-05,
      "loss": 1.4527,
      "step": 165
    },
    {
      "epoch": 0.20802005012531327,
      "grad_norm": 0.1350892335176468,
      "learning_rate": 5.520760045623403e-05,
      "loss": 1.524,
      "step": 166
    },
    {
      "epoch": 0.20927318295739347,
      "grad_norm": 0.15125590562820435,
      "learning_rate": 5.514128562824039e-05,
      "loss": 1.5476,
      "step": 167
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 0.18517212569713593,
      "learning_rate": 5.507455553303506e-05,
      "loss": 1.4797,
      "step": 168
    },
    {
      "epoch": 0.21177944862155387,
      "grad_norm": 0.16194948554039001,
      "learning_rate": 5.5007411272821826e-05,
      "loss": 1.4256,
      "step": 169
    },
    {
      "epoch": 0.21303258145363407,
      "grad_norm": 0.16133227944374084,
      "learning_rate": 5.493985395664539e-05,
      "loss": 1.5882,
      "step": 170
    },
    {
      "epoch": 0.21428571428571427,
      "grad_norm": 0.15564045310020447,
      "learning_rate": 5.487188470037305e-05,
      "loss": 1.5051,
      "step": 171
    },
    {
      "epoch": 0.21553884711779447,
      "grad_norm": 0.1621345579624176,
      "learning_rate": 5.480350462667625e-05,
      "loss": 1.4765,
      "step": 172
    },
    {
      "epoch": 0.21679197994987467,
      "grad_norm": 0.16004936397075653,
      "learning_rate": 5.473471486501206e-05,
      "loss": 1.4995,
      "step": 173
    },
    {
      "epoch": 0.21804511278195488,
      "grad_norm": 0.1602533906698227,
      "learning_rate": 5.466551655160449e-05,
      "loss": 1.5091,
      "step": 174
    },
    {
      "epoch": 0.21929824561403508,
      "grad_norm": 0.15978209674358368,
      "learning_rate": 5.459591082942574e-05,
      "loss": 1.5351,
      "step": 175
    },
    {
      "epoch": 0.22055137844611528,
      "grad_norm": 0.15208299458026886,
      "learning_rate": 5.452589884817733e-05,
      "loss": 1.5023,
      "step": 176
    },
    {
      "epoch": 0.22180451127819548,
      "grad_norm": 0.1487240493297577,
      "learning_rate": 5.445548176427108e-05,
      "loss": 1.4458,
      "step": 177
    },
    {
      "epoch": 0.22305764411027568,
      "grad_norm": 0.14227648079395294,
      "learning_rate": 5.4384660740810074e-05,
      "loss": 1.4342,
      "step": 178
    },
    {
      "epoch": 0.22431077694235588,
      "grad_norm": 0.13798528909683228,
      "learning_rate": 5.431343694756935e-05,
      "loss": 1.3625,
      "step": 179
    },
    {
      "epoch": 0.22556390977443608,
      "grad_norm": 0.16410118341445923,
      "learning_rate": 5.424181156097666e-05,
      "loss": 1.4706,
      "step": 180
    },
    {
      "epoch": 0.22681704260651628,
      "grad_norm": 0.15224584937095642,
      "learning_rate": 5.416978576409301e-05,
      "loss": 1.5443,
      "step": 181
    },
    {
      "epoch": 0.22807017543859648,
      "grad_norm": 0.15167997777462006,
      "learning_rate": 5.409736074659311e-05,
      "loss": 1.5039,
      "step": 182
    },
    {
      "epoch": 0.22932330827067668,
      "grad_norm": 0.16224287450313568,
      "learning_rate": 5.402453770474575e-05,
      "loss": 1.4773,
      "step": 183
    },
    {
      "epoch": 0.23057644110275688,
      "grad_norm": 0.16024534404277802,
      "learning_rate": 5.395131784139401e-05,
      "loss": 1.5209,
      "step": 184
    },
    {
      "epoch": 0.23182957393483708,
      "grad_norm": 0.15539222955703735,
      "learning_rate": 5.3877702365935404e-05,
      "loss": 1.4137,
      "step": 185
    },
    {
      "epoch": 0.23308270676691728,
      "grad_norm": 0.146916463971138,
      "learning_rate": 5.380369249430191e-05,
      "loss": 1.4557,
      "step": 186
    },
    {
      "epoch": 0.23433583959899748,
      "grad_norm": 0.16157691180706024,
      "learning_rate": 5.37292894489399e-05,
      "loss": 1.5541,
      "step": 187
    },
    {
      "epoch": 0.23558897243107768,
      "grad_norm": 0.15176048874855042,
      "learning_rate": 5.36544944587899e-05,
      "loss": 1.5486,
      "step": 188
    },
    {
      "epoch": 0.23684210526315788,
      "grad_norm": 0.17215217649936676,
      "learning_rate": 5.357930875926636e-05,
      "loss": 1.5725,
      "step": 189
    },
    {
      "epoch": 0.23809523809523808,
      "grad_norm": 0.16552142798900604,
      "learning_rate": 5.3503733592237174e-05,
      "loss": 1.5372,
      "step": 190
    },
    {
      "epoch": 0.23934837092731828,
      "grad_norm": 0.1510804295539856,
      "learning_rate": 5.342777020600321e-05,
      "loss": 1.3789,
      "step": 191
    },
    {
      "epoch": 0.24060150375939848,
      "grad_norm": 0.1481272578239441,
      "learning_rate": 5.335141985527771e-05,
      "loss": 1.7376,
      "step": 192
    },
    {
      "epoch": 0.24185463659147868,
      "grad_norm": 0.1423473209142685,
      "learning_rate": 5.327468380116551e-05,
      "loss": 1.4623,
      "step": 193
    },
    {
      "epoch": 0.24310776942355888,
      "grad_norm": 0.15733863413333893,
      "learning_rate": 5.319756331114227e-05,
      "loss": 1.4554,
      "step": 194
    },
    {
      "epoch": 0.24436090225563908,
      "grad_norm": 0.1579495519399643,
      "learning_rate": 5.3120059659033484e-05,
      "loss": 1.3653,
      "step": 195
    },
    {
      "epoch": 0.24561403508771928,
      "grad_norm": 0.16561569273471832,
      "learning_rate": 5.304217412499349e-05,
      "loss": 1.4189,
      "step": 196
    },
    {
      "epoch": 0.24686716791979949,
      "grad_norm": 0.16004639863967896,
      "learning_rate": 5.296390799548426e-05,
      "loss": 1.4953,
      "step": 197
    },
    {
      "epoch": 0.24812030075187969,
      "grad_norm": 0.14528200030326843,
      "learning_rate": 5.2885262563254246e-05,
      "loss": 1.2997,
      "step": 198
    },
    {
      "epoch": 0.24937343358395989,
      "grad_norm": 0.15132363140583038,
      "learning_rate": 5.280623912731693e-05,
      "loss": 1.4261,
      "step": 199
    },
    {
      "epoch": 0.2506265664160401,
      "grad_norm": 0.15129469335079193,
      "learning_rate": 5.272683899292943e-05,
      "loss": 1.4778,
      "step": 200
    },
    {
      "epoch": 0.2518796992481203,
      "grad_norm": 0.16739892959594727,
      "learning_rate": 5.2647063471570914e-05,
      "loss": 1.4741,
      "step": 201
    },
    {
      "epoch": 0.2531328320802005,
      "grad_norm": 0.1433250457048416,
      "learning_rate": 5.256691388092095e-05,
      "loss": 1.398,
      "step": 202
    },
    {
      "epoch": 0.2543859649122807,
      "grad_norm": 0.16294027864933014,
      "learning_rate": 5.248639154483773e-05,
      "loss": 1.4912,
      "step": 203
    },
    {
      "epoch": 0.2556390977443609,
      "grad_norm": 0.16056182980537415,
      "learning_rate": 5.240549779333621e-05,
      "loss": 1.5168,
      "step": 204
    },
    {
      "epoch": 0.2568922305764411,
      "grad_norm": 0.1551842838525772,
      "learning_rate": 5.2324233962566154e-05,
      "loss": 1.4244,
      "step": 205
    },
    {
      "epoch": 0.2581453634085213,
      "grad_norm": 0.15926840901374817,
      "learning_rate": 5.224260139479004e-05,
      "loss": 1.5547,
      "step": 206
    },
    {
      "epoch": 0.2593984962406015,
      "grad_norm": 0.1555553674697876,
      "learning_rate": 5.21606014383609e-05,
      "loss": 1.4263,
      "step": 207
    },
    {
      "epoch": 0.2606516290726817,
      "grad_norm": 0.1518319994211197,
      "learning_rate": 5.207823544770006e-05,
      "loss": 1.4604,
      "step": 208
    },
    {
      "epoch": 0.2619047619047619,
      "grad_norm": 0.1934310644865036,
      "learning_rate": 5.199550478327474e-05,
      "loss": 1.5793,
      "step": 209
    },
    {
      "epoch": 0.2631578947368421,
      "grad_norm": 0.16186797618865967,
      "learning_rate": 5.1912410811575635e-05,
      "loss": 1.5374,
      "step": 210
    },
    {
      "epoch": 0.2644110275689223,
      "grad_norm": 0.15732304751873016,
      "learning_rate": 5.182895490509427e-05,
      "loss": 1.4555,
      "step": 211
    },
    {
      "epoch": 0.2656641604010025,
      "grad_norm": 0.17242950201034546,
      "learning_rate": 5.1745138442300396e-05,
      "loss": 1.5941,
      "step": 212
    },
    {
      "epoch": 0.2669172932330827,
      "grad_norm": 0.16432379186153412,
      "learning_rate": 5.16609628076192e-05,
      "loss": 1.4662,
      "step": 213
    },
    {
      "epoch": 0.2681704260651629,
      "grad_norm": 0.15895594656467438,
      "learning_rate": 5.15764293914084e-05,
      "loss": 1.4611,
      "step": 214
    },
    {
      "epoch": 0.2694235588972431,
      "grad_norm": 0.17354588210582733,
      "learning_rate": 5.149153958993537e-05,
      "loss": 1.4325,
      "step": 215
    },
    {
      "epoch": 0.2706766917293233,
      "grad_norm": 0.15314188599586487,
      "learning_rate": 5.140629480535396e-05,
      "loss": 1.4666,
      "step": 216
    },
    {
      "epoch": 0.2719298245614035,
      "grad_norm": 0.16267825663089752,
      "learning_rate": 5.132069644568143e-05,
      "loss": 1.4234,
      "step": 217
    },
    {
      "epoch": 0.2731829573934837,
      "grad_norm": 0.16543705761432648,
      "learning_rate": 5.123474592477517e-05,
      "loss": 1.5027,
      "step": 218
    },
    {
      "epoch": 0.2744360902255639,
      "grad_norm": 0.1630410999059677,
      "learning_rate": 5.1148444662309326e-05,
      "loss": 1.4812,
      "step": 219
    },
    {
      "epoch": 0.2756892230576441,
      "grad_norm": 0.17135612666606903,
      "learning_rate": 5.106179408375135e-05,
      "loss": 1.4541,
      "step": 220
    },
    {
      "epoch": 0.2769423558897243,
      "grad_norm": 0.1608368307352066,
      "learning_rate": 5.097479562033847e-05,
      "loss": 1.4371,
      "step": 221
    },
    {
      "epoch": 0.2781954887218045,
      "grad_norm": 0.17631706595420837,
      "learning_rate": 5.0887450709054065e-05,
      "loss": 1.4593,
      "step": 222
    },
    {
      "epoch": 0.2794486215538847,
      "grad_norm": 0.1584499627351761,
      "learning_rate": 5.079976079260391e-05,
      "loss": 1.3738,
      "step": 223
    },
    {
      "epoch": 0.2807017543859649,
      "grad_norm": 0.1915035843849182,
      "learning_rate": 5.071172731939233e-05,
      "loss": 1.5473,
      "step": 224
    },
    {
      "epoch": 0.2819548872180451,
      "grad_norm": 0.17202699184417725,
      "learning_rate": 5.062335174349831e-05,
      "loss": 1.4301,
      "step": 225
    },
    {
      "epoch": 0.2832080200501253,
      "grad_norm": 0.1597733050584793,
      "learning_rate": 5.053463552465145e-05,
      "loss": 1.3866,
      "step": 226
    },
    {
      "epoch": 0.2844611528822055,
      "grad_norm": 0.16445639729499817,
      "learning_rate": 5.0445580128207885e-05,
      "loss": 1.4294,
      "step": 227
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.17147259414196014,
      "learning_rate": 5.0356187025126035e-05,
      "loss": 1.4303,
      "step": 228
    },
    {
      "epoch": 0.2869674185463659,
      "grad_norm": 0.18276292085647583,
      "learning_rate": 5.026645769194236e-05,
      "loss": 1.4647,
      "step": 229
    },
    {
      "epoch": 0.2882205513784461,
      "grad_norm": 0.1690540462732315,
      "learning_rate": 5.017639361074694e-05,
      "loss": 1.5525,
      "step": 230
    },
    {
      "epoch": 0.2894736842105263,
      "grad_norm": 0.1618894338607788,
      "learning_rate": 5.0085996269159e-05,
      "loss": 1.4743,
      "step": 231
    },
    {
      "epoch": 0.2907268170426065,
      "grad_norm": 0.16552814841270447,
      "learning_rate": 4.9995267160302336e-05,
      "loss": 1.4404,
      "step": 232
    },
    {
      "epoch": 0.29197994987468673,
      "grad_norm": 0.17212465405464172,
      "learning_rate": 4.990420778278067e-05,
      "loss": 1.4921,
      "step": 233
    },
    {
      "epoch": 0.2932330827067669,
      "grad_norm": 0.17030447721481323,
      "learning_rate": 4.981281964065287e-05,
      "loss": 1.4577,
      "step": 234
    },
    {
      "epoch": 0.29448621553884713,
      "grad_norm": 0.16432829201221466,
      "learning_rate": 4.972110424340815e-05,
      "loss": 1.3663,
      "step": 235
    },
    {
      "epoch": 0.2957393483709273,
      "grad_norm": 0.17019811272621155,
      "learning_rate": 4.962906310594106e-05,
      "loss": 1.3605,
      "step": 236
    },
    {
      "epoch": 0.29699248120300753,
      "grad_norm": 0.15439601242542267,
      "learning_rate": 4.953669774852657e-05,
      "loss": 1.3821,
      "step": 237
    },
    {
      "epoch": 0.2982456140350877,
      "grad_norm": 0.16957609355449677,
      "learning_rate": 4.9444009696794836e-05,
      "loss": 1.5436,
      "step": 238
    },
    {
      "epoch": 0.29949874686716793,
      "grad_norm": 0.17840701341629028,
      "learning_rate": 4.9351000481706135e-05,
      "loss": 1.5132,
      "step": 239
    },
    {
      "epoch": 0.3007518796992481,
      "grad_norm": 0.16516822576522827,
      "learning_rate": 4.925767163952547e-05,
      "loss": 1.5301,
      "step": 240
    },
    {
      "epoch": 0.30200501253132833,
      "grad_norm": 0.16189689934253693,
      "learning_rate": 4.916402471179723e-05,
      "loss": 1.4475,
      "step": 241
    },
    {
      "epoch": 0.3032581453634085,
      "grad_norm": 0.1689058095216751,
      "learning_rate": 4.907006124531975e-05,
      "loss": 1.4622,
      "step": 242
    },
    {
      "epoch": 0.30451127819548873,
      "grad_norm": 0.16913363337516785,
      "learning_rate": 4.897578279211975e-05,
      "loss": 1.5433,
      "step": 243
    },
    {
      "epoch": 0.3057644110275689,
      "grad_norm": 0.17671072483062744,
      "learning_rate": 4.888119090942665e-05,
      "loss": 1.4205,
      "step": 244
    },
    {
      "epoch": 0.30701754385964913,
      "grad_norm": 0.1828470230102539,
      "learning_rate": 4.878628715964693e-05,
      "loss": 1.4191,
      "step": 245
    },
    {
      "epoch": 0.3082706766917293,
      "grad_norm": 0.17133688926696777,
      "learning_rate": 4.86910731103383e-05,
      "loss": 1.5249,
      "step": 246
    },
    {
      "epoch": 0.30952380952380953,
      "grad_norm": 0.167369082570076,
      "learning_rate": 4.859555033418374e-05,
      "loss": 1.4614,
      "step": 247
    },
    {
      "epoch": 0.3107769423558897,
      "grad_norm": 0.18615829944610596,
      "learning_rate": 4.8499720408965634e-05,
      "loss": 1.5662,
      "step": 248
    },
    {
      "epoch": 0.31203007518796994,
      "grad_norm": 0.15995121002197266,
      "learning_rate": 4.840358491753961e-05,
      "loss": 1.4518,
      "step": 249
    },
    {
      "epoch": 0.3132832080200501,
      "grad_norm": 0.17739561200141907,
      "learning_rate": 4.830714544780846e-05,
      "loss": 1.4187,
      "step": 250
    },
    {
      "epoch": 0.31453634085213034,
      "grad_norm": 0.16059748828411102,
      "learning_rate": 4.821040359269587e-05,
      "loss": 1.3574,
      "step": 251
    },
    {
      "epoch": 0.3157894736842105,
      "grad_norm": 0.1768496036529541,
      "learning_rate": 4.811336095012015e-05,
      "loss": 1.4048,
      "step": 252
    },
    {
      "epoch": 0.31704260651629074,
      "grad_norm": 0.17573674023151398,
      "learning_rate": 4.8016019122967826e-05,
      "loss": 1.542,
      "step": 253
    },
    {
      "epoch": 0.3182957393483709,
      "grad_norm": 0.15960711240768433,
      "learning_rate": 4.791837971906711e-05,
      "loss": 1.3686,
      "step": 254
    },
    {
      "epoch": 0.31954887218045114,
      "grad_norm": 0.18074101209640503,
      "learning_rate": 4.782044435116145e-05,
      "loss": 1.5344,
      "step": 255
    },
    {
      "epoch": 0.3208020050125313,
      "grad_norm": 0.2042127400636673,
      "learning_rate": 4.7722214636882806e-05,
      "loss": 1.5176,
      "step": 256
    },
    {
      "epoch": 0.32205513784461154,
      "grad_norm": 0.16601479053497314,
      "learning_rate": 4.762369219872494e-05,
      "loss": 1.3895,
      "step": 257
    },
    {
      "epoch": 0.3233082706766917,
      "grad_norm": 0.18349255621433258,
      "learning_rate": 4.752487866401667e-05,
      "loss": 1.4652,
      "step": 258
    },
    {
      "epoch": 0.32456140350877194,
      "grad_norm": 0.18353460729122162,
      "learning_rate": 4.742577566489493e-05,
      "loss": 1.5181,
      "step": 259
    },
    {
      "epoch": 0.3258145363408521,
      "grad_norm": 0.18172797560691833,
      "learning_rate": 4.732638483827784e-05,
      "loss": 1.4484,
      "step": 260
    },
    {
      "epoch": 0.32706766917293234,
      "grad_norm": 0.17833280563354492,
      "learning_rate": 4.722670782583769e-05,
      "loss": 1.4063,
      "step": 261
    },
    {
      "epoch": 0.3283208020050125,
      "grad_norm": 0.17507833242416382,
      "learning_rate": 4.7126746273973745e-05,
      "loss": 1.5928,
      "step": 262
    },
    {
      "epoch": 0.32957393483709274,
      "grad_norm": 0.17333512008190155,
      "learning_rate": 4.702650183378517e-05,
      "loss": 1.526,
      "step": 263
    },
    {
      "epoch": 0.3308270676691729,
      "grad_norm": 0.19539874792099,
      "learning_rate": 4.692597616104364e-05,
      "loss": 1.4775,
      "step": 264
    },
    {
      "epoch": 0.33208020050125314,
      "grad_norm": 0.1711081713438034,
      "learning_rate": 4.682517091616608e-05,
      "loss": 1.2613,
      "step": 265
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.17969585955142975,
      "learning_rate": 4.672408776418718e-05,
      "loss": 1.5688,
      "step": 266
    },
    {
      "epoch": 0.33458646616541354,
      "grad_norm": 0.17389164865016937,
      "learning_rate": 4.6622728374731936e-05,
      "loss": 1.3871,
      "step": 267
    },
    {
      "epoch": 0.3358395989974937,
      "grad_norm": 0.16892869770526886,
      "learning_rate": 4.652109442198804e-05,
      "loss": 1.4454,
      "step": 268
    },
    {
      "epoch": 0.33709273182957394,
      "grad_norm": 0.18089257180690765,
      "learning_rate": 4.641918758467824e-05,
      "loss": 1.4727,
      "step": 269
    },
    {
      "epoch": 0.3383458646616541,
      "grad_norm": 0.18539738655090332,
      "learning_rate": 4.6317009546032615e-05,
      "loss": 1.4823,
      "step": 270
    },
    {
      "epoch": 0.33959899749373434,
      "grad_norm": 0.1722467839717865,
      "learning_rate": 4.621456199376077e-05,
      "loss": 1.4174,
      "step": 271
    },
    {
      "epoch": 0.3408521303258145,
      "grad_norm": 0.17670480906963348,
      "learning_rate": 4.611184662002396e-05,
      "loss": 1.5394,
      "step": 272
    },
    {
      "epoch": 0.34210526315789475,
      "grad_norm": 0.16371813416481018,
      "learning_rate": 4.6008865121407126e-05,
      "loss": 1.4326,
      "step": 273
    },
    {
      "epoch": 0.3433583959899749,
      "grad_norm": 0.17700393497943878,
      "learning_rate": 4.590561919889091e-05,
      "loss": 1.4755,
      "step": 274
    },
    {
      "epoch": 0.34461152882205515,
      "grad_norm": 0.1750127226114273,
      "learning_rate": 4.58021105578235e-05,
      "loss": 1.3065,
      "step": 275
    },
    {
      "epoch": 0.3458646616541353,
      "grad_norm": 0.18933042883872986,
      "learning_rate": 4.5698340907892534e-05,
      "loss": 1.5368,
      "step": 276
    },
    {
      "epoch": 0.34711779448621555,
      "grad_norm": 0.19926317036151886,
      "learning_rate": 4.55943119630968e-05,
      "loss": 1.5401,
      "step": 277
    },
    {
      "epoch": 0.3483709273182957,
      "grad_norm": 0.1745910942554474,
      "learning_rate": 4.5490025441717946e-05,
      "loss": 1.4631,
      "step": 278
    },
    {
      "epoch": 0.34962406015037595,
      "grad_norm": 0.1677342802286148,
      "learning_rate": 4.538548306629209e-05,
      "loss": 1.4973,
      "step": 279
    },
    {
      "epoch": 0.3508771929824561,
      "grad_norm": 0.17847084999084473,
      "learning_rate": 4.5280686563581425e-05,
      "loss": 1.5144,
      "step": 280
    },
    {
      "epoch": 0.35213032581453635,
      "grad_norm": 0.18145380914211273,
      "learning_rate": 4.517563766454561e-05,
      "loss": 1.4917,
      "step": 281
    },
    {
      "epoch": 0.3533834586466165,
      "grad_norm": 0.19056649506092072,
      "learning_rate": 4.5070338104313224e-05,
      "loss": 1.4575,
      "step": 282
    },
    {
      "epoch": 0.35463659147869675,
      "grad_norm": 0.17569053173065186,
      "learning_rate": 4.4964789622153116e-05,
      "loss": 1.5196,
      "step": 283
    },
    {
      "epoch": 0.3558897243107769,
      "grad_norm": 0.18289506435394287,
      "learning_rate": 4.485899396144566e-05,
      "loss": 1.402,
      "step": 284
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 0.17359083890914917,
      "learning_rate": 4.4752952869653964e-05,
      "loss": 1.4144,
      "step": 285
    },
    {
      "epoch": 0.3583959899749373,
      "grad_norm": 0.1759563386440277,
      "learning_rate": 4.464666809829499e-05,
      "loss": 1.5265,
      "step": 286
    },
    {
      "epoch": 0.35964912280701755,
      "grad_norm": 0.17689688503742218,
      "learning_rate": 4.4540141402910645e-05,
      "loss": 1.5254,
      "step": 287
    },
    {
      "epoch": 0.3609022556390977,
      "grad_norm": 0.18115444481372833,
      "learning_rate": 4.443337454303881e-05,
      "loss": 1.6087,
      "step": 288
    },
    {
      "epoch": 0.36215538847117795,
      "grad_norm": 0.17893773317337036,
      "learning_rate": 4.432636928218419e-05,
      "loss": 1.4052,
      "step": 289
    },
    {
      "epoch": 0.3634085213032581,
      "grad_norm": 0.2004009187221527,
      "learning_rate": 4.4219127387789293e-05,
      "loss": 1.503,
      "step": 290
    },
    {
      "epoch": 0.36466165413533835,
      "grad_norm": 0.16822977364063263,
      "learning_rate": 4.411165063120514e-05,
      "loss": 1.3674,
      "step": 291
    },
    {
      "epoch": 0.3659147869674185,
      "grad_norm": 0.1785370260477066,
      "learning_rate": 4.400394078766208e-05,
      "loss": 1.5148,
      "step": 292
    },
    {
      "epoch": 0.36716791979949875,
      "grad_norm": 0.16503627598285675,
      "learning_rate": 4.389599963624042e-05,
      "loss": 1.4331,
      "step": 293
    },
    {
      "epoch": 0.3684210526315789,
      "grad_norm": 0.1841992437839508,
      "learning_rate": 4.3787828959841084e-05,
      "loss": 1.3116,
      "step": 294
    },
    {
      "epoch": 0.36967418546365916,
      "grad_norm": 0.18730437755584717,
      "learning_rate": 4.367943054515611e-05,
      "loss": 1.3779,
      "step": 295
    },
    {
      "epoch": 0.37092731829573933,
      "grad_norm": 0.17390833795070648,
      "learning_rate": 4.3570806182639185e-05,
      "loss": 1.4706,
      "step": 296
    },
    {
      "epoch": 0.37218045112781956,
      "grad_norm": 0.16994328796863556,
      "learning_rate": 4.346195766647606e-05,
      "loss": 1.4047,
      "step": 297
    },
    {
      "epoch": 0.37343358395989973,
      "grad_norm": 0.17841129004955292,
      "learning_rate": 4.335288679455487e-05,
      "loss": 1.4686,
      "step": 298
    },
    {
      "epoch": 0.37468671679197996,
      "grad_norm": 0.19922415912151337,
      "learning_rate": 4.3243595368436517e-05,
      "loss": 1.5165,
      "step": 299
    },
    {
      "epoch": 0.37593984962406013,
      "grad_norm": 0.16899648308753967,
      "learning_rate": 4.3134085193324865e-05,
      "loss": 1.4881,
      "step": 300
    },
    {
      "epoch": 0.37719298245614036,
      "grad_norm": 0.16627173125743866,
      "learning_rate": 4.302435807803693e-05,
      "loss": 1.4882,
      "step": 301
    },
    {
      "epoch": 0.37844611528822053,
      "grad_norm": 0.19958005845546722,
      "learning_rate": 4.291441583497296e-05,
      "loss": 1.5203,
      "step": 302
    },
    {
      "epoch": 0.37969924812030076,
      "grad_norm": 0.19517013430595398,
      "learning_rate": 4.280426028008661e-05,
      "loss": 1.4329,
      "step": 303
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 0.1873486489057541,
      "learning_rate": 4.269389323285484e-05,
      "loss": 1.3892,
      "step": 304
    },
    {
      "epoch": 0.38220551378446116,
      "grad_norm": 0.17650927603244781,
      "learning_rate": 4.258331651624791e-05,
      "loss": 1.4937,
      "step": 305
    },
    {
      "epoch": 0.38345864661654133,
      "grad_norm": 0.17674066126346588,
      "learning_rate": 4.247253195669925e-05,
      "loss": 1.4761,
      "step": 306
    },
    {
      "epoch": 0.38471177944862156,
      "grad_norm": 0.19194164872169495,
      "learning_rate": 4.236154138407533e-05,
      "loss": 1.4483,
      "step": 307
    },
    {
      "epoch": 0.38596491228070173,
      "grad_norm": 0.18375438451766968,
      "learning_rate": 4.2250346631645374e-05,
      "loss": 1.458,
      "step": 308
    },
    {
      "epoch": 0.38721804511278196,
      "grad_norm": 0.1865517795085907,
      "learning_rate": 4.213894953605113e-05,
      "loss": 1.4413,
      "step": 309
    },
    {
      "epoch": 0.38847117794486213,
      "grad_norm": 0.18141886591911316,
      "learning_rate": 4.2027351937276525e-05,
      "loss": 1.4391,
      "step": 310
    },
    {
      "epoch": 0.38972431077694236,
      "grad_norm": 0.18934974074363708,
      "learning_rate": 4.191555567861728e-05,
      "loss": 1.4008,
      "step": 311
    },
    {
      "epoch": 0.39097744360902253,
      "grad_norm": 0.1794135719537735,
      "learning_rate": 4.1803562606650374e-05,
      "loss": 1.3936,
      "step": 312
    },
    {
      "epoch": 0.39223057644110276,
      "grad_norm": 0.17992708086967468,
      "learning_rate": 4.169137457120373e-05,
      "loss": 1.3573,
      "step": 313
    },
    {
      "epoch": 0.39348370927318294,
      "grad_norm": 0.1869203895330429,
      "learning_rate": 4.157899342532545e-05,
      "loss": 1.4579,
      "step": 314
    },
    {
      "epoch": 0.39473684210526316,
      "grad_norm": 0.17959001660346985,
      "learning_rate": 4.146642102525336e-05,
      "loss": 1.6471,
      "step": 315
    },
    {
      "epoch": 0.39598997493734334,
      "grad_norm": 0.18178239464759827,
      "learning_rate": 4.1353659230384306e-05,
      "loss": 1.5083,
      "step": 316
    },
    {
      "epoch": 0.39724310776942356,
      "grad_norm": 0.17811906337738037,
      "learning_rate": 4.124070990324339e-05,
      "loss": 1.3851,
      "step": 317
    },
    {
      "epoch": 0.39849624060150374,
      "grad_norm": 0.18983860313892365,
      "learning_rate": 4.1127574909453276e-05,
      "loss": 1.4802,
      "step": 318
    },
    {
      "epoch": 0.39974937343358397,
      "grad_norm": 0.19141417741775513,
      "learning_rate": 4.1014256117703344e-05,
      "loss": 1.4496,
      "step": 319
    },
    {
      "epoch": 0.40100250626566414,
      "grad_norm": 0.18502330780029297,
      "learning_rate": 4.090075539971883e-05,
      "loss": 1.3621,
      "step": 320
    }
  ],
  "logging_steps": 1,
  "max_steps": 798,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 32,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.3421542911875482e+18,
  "train_batch_size": 12,
  "trial_name": null,
  "trial_params": null
}