{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9971469329529244,
  "eval_steps": 500,
  "global_step": 233,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0042796005706134095,
      "grad_norm": 34.48759460449219,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 4.9483,
      "step": 1
    },
    {
      "epoch": 0.008559201141226819,
      "grad_norm": 35.18870544433594,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 5.1789,
      "step": 2
    },
    {
      "epoch": 0.012838801711840228,
      "grad_norm": 34.3349494934082,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 4.8881,
      "step": 3
    },
    {
      "epoch": 0.017118402282453638,
      "grad_norm": 36.14989471435547,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 5.0836,
      "step": 4
    },
    {
      "epoch": 0.021398002853067047,
      "grad_norm": 33.93682861328125,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 4.9619,
      "step": 5
    },
    {
      "epoch": 0.025677603423680456,
      "grad_norm": 34.47760772705078,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 4.9723,
      "step": 6
    },
    {
      "epoch": 0.029957203994293864,
      "grad_norm": 34.52993392944336,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 4.9278,
      "step": 7
    },
    {
      "epoch": 0.034236804564907276,
      "grad_norm": 33.86513900756836,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 4.9141,
      "step": 8
    },
    {
      "epoch": 0.03851640513552068,
      "grad_norm": 32.446311950683594,
      "learning_rate": 4.5000000000000003e-07,
      "loss": 4.7811,
      "step": 9
    },
    {
      "epoch": 0.042796005706134094,
      "grad_norm": 33.34199905395508,
      "learning_rate": 5.000000000000001e-07,
      "loss": 4.8566,
      "step": 10
    },
    {
      "epoch": 0.047075606276747506,
      "grad_norm": 34.60908508300781,
      "learning_rate": 5.5e-07,
      "loss": 4.9406,
      "step": 11
    },
    {
      "epoch": 0.05135520684736091,
      "grad_norm": 34.570987701416016,
      "learning_rate": 6.000000000000001e-07,
      "loss": 4.8281,
      "step": 12
    },
    {
      "epoch": 0.05563480741797432,
      "grad_norm": 32.070953369140625,
      "learning_rate": 6.5e-07,
      "loss": 4.6165,
      "step": 13
    },
    {
      "epoch": 0.05991440798858773,
      "grad_norm": 33.586509704589844,
      "learning_rate": 7.000000000000001e-07,
      "loss": 4.8012,
      "step": 14
    },
    {
      "epoch": 0.06419400855920114,
      "grad_norm": 30.96600341796875,
      "learning_rate": 7.5e-07,
      "loss": 4.5845,
      "step": 15
    },
    {
      "epoch": 0.06847360912981455,
      "grad_norm": 30.152454376220703,
      "learning_rate": 8.000000000000001e-07,
      "loss": 4.4356,
      "step": 16
    },
    {
      "epoch": 0.07275320970042796,
      "grad_norm": 28.61970329284668,
      "learning_rate": 8.500000000000001e-07,
      "loss": 4.359,
      "step": 17
    },
    {
      "epoch": 0.07703281027104136,
      "grad_norm": 27.507383346557617,
      "learning_rate": 9.000000000000001e-07,
      "loss": 4.4086,
      "step": 18
    },
    {
      "epoch": 0.08131241084165478,
      "grad_norm": 25.110597610473633,
      "learning_rate": 9.500000000000001e-07,
      "loss": 4.1411,
      "step": 19
    },
    {
      "epoch": 0.08559201141226819,
      "grad_norm": 23.737529754638672,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 4.0639,
      "step": 20
    },
    {
      "epoch": 0.0898716119828816,
      "grad_norm": 22.502748489379883,
      "learning_rate": 1.0500000000000001e-06,
      "loss": 3.9744,
      "step": 21
    },
    {
      "epoch": 0.09415121255349501,
      "grad_norm": 21.957761764526367,
      "learning_rate": 1.1e-06,
      "loss": 3.8115,
      "step": 22
    },
    {
      "epoch": 0.09843081312410841,
      "grad_norm": 20.552080154418945,
      "learning_rate": 1.1500000000000002e-06,
      "loss": 3.6059,
      "step": 23
    },
    {
      "epoch": 0.10271041369472182,
      "grad_norm": 21.490442276000977,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 3.5285,
      "step": 24
    },
    {
      "epoch": 0.10699001426533523,
      "grad_norm": 20.63368034362793,
      "learning_rate": 1.25e-06,
      "loss": 3.3484,
      "step": 25
    },
    {
      "epoch": 0.11126961483594865,
      "grad_norm": 20.170631408691406,
      "learning_rate": 1.3e-06,
      "loss": 3.0973,
      "step": 26
    },
    {
      "epoch": 0.11554921540656206,
      "grad_norm": 20.217880249023438,
      "learning_rate": 1.3500000000000002e-06,
      "loss": 2.9758,
      "step": 27
    },
    {
      "epoch": 0.11982881597717546,
      "grad_norm": 18.303476333618164,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 2.8659,
      "step": 28
    },
    {
      "epoch": 0.12410841654778887,
      "grad_norm": 15.922847747802734,
      "learning_rate": 1.45e-06,
      "loss": 2.6369,
      "step": 29
    },
    {
      "epoch": 0.12838801711840228,
      "grad_norm": 14.529817581176758,
      "learning_rate": 1.5e-06,
      "loss": 2.5532,
      "step": 30
    },
    {
      "epoch": 0.1326676176890157,
      "grad_norm": 14.177539825439453,
      "learning_rate": 1.5500000000000002e-06,
      "loss": 2.4497,
      "step": 31
    },
    {
      "epoch": 0.1369472182596291,
      "grad_norm": 15.144468307495117,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 2.3432,
      "step": 32
    },
    {
      "epoch": 0.14122681883024252,
      "grad_norm": 14.920907974243164,
      "learning_rate": 1.6500000000000003e-06,
      "loss": 2.14,
      "step": 33
    },
    {
      "epoch": 0.14550641940085593,
      "grad_norm": 15.700516700744629,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 2.0247,
      "step": 34
    },
    {
      "epoch": 0.14978601997146934,
      "grad_norm": 15.600655555725098,
      "learning_rate": 1.75e-06,
      "loss": 1.8893,
      "step": 35
    },
    {
      "epoch": 0.15406562054208273,
      "grad_norm": 15.281895637512207,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 1.7768,
      "step": 36
    },
    {
      "epoch": 0.15834522111269614,
      "grad_norm": 13.99860668182373,
      "learning_rate": 1.85e-06,
      "loss": 1.5486,
      "step": 37
    },
    {
      "epoch": 0.16262482168330955,
      "grad_norm": 13.295161247253418,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 1.427,
      "step": 38
    },
    {
      "epoch": 0.16690442225392296,
      "grad_norm": 12.667908668518066,
      "learning_rate": 1.9500000000000004e-06,
      "loss": 1.2251,
      "step": 39
    },
    {
      "epoch": 0.17118402282453637,
      "grad_norm": 12.730255126953125,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.1402,
      "step": 40
    },
    {
      "epoch": 0.1754636233951498,
      "grad_norm": 13.135544776916504,
      "learning_rate": 2.05e-06,
      "loss": 0.9867,
      "step": 41
    },
    {
      "epoch": 0.1797432239657632,
      "grad_norm": 12.429509162902832,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 0.8455,
      "step": 42
    },
    {
      "epoch": 0.1840228245363766,
      "grad_norm": 11.61819076538086,
      "learning_rate": 2.15e-06,
      "loss": 0.6935,
      "step": 43
    },
    {
      "epoch": 0.18830242510699002,
      "grad_norm": 10.11789608001709,
      "learning_rate": 2.2e-06,
      "loss": 0.5486,
      "step": 44
    },
    {
      "epoch": 0.19258202567760344,
      "grad_norm": 8.705474853515625,
      "learning_rate": 2.25e-06,
      "loss": 0.4345,
      "step": 45
    },
    {
      "epoch": 0.19686162624821682,
      "grad_norm": 7.407937526702881,
      "learning_rate": 2.3000000000000004e-06,
      "loss": 0.3223,
      "step": 46
    },
    {
      "epoch": 0.20114122681883023,
      "grad_norm": 5.553786754608154,
      "learning_rate": 2.35e-06,
      "loss": 0.2376,
      "step": 47
    },
    {
      "epoch": 0.20542082738944364,
      "grad_norm": 3.248605966567993,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.198,
      "step": 48
    },
    {
      "epoch": 0.20970042796005706,
      "grad_norm": 3.1356797218322754,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 0.1706,
      "step": 49
    },
    {
      "epoch": 0.21398002853067047,
      "grad_norm": 2.7913248538970947,
      "learning_rate": 2.5e-06,
      "loss": 0.1436,
      "step": 50
    },
    {
      "epoch": 0.21825962910128388,
      "grad_norm": 2.2059645652770996,
      "learning_rate": 2.55e-06,
      "loss": 0.144,
      "step": 51
    },
    {
      "epoch": 0.2225392296718973,
      "grad_norm": 1.5722575187683105,
      "learning_rate": 2.6e-06,
      "loss": 0.1203,
      "step": 52
    },
    {
      "epoch": 0.2268188302425107,
      "grad_norm": 1.3645284175872803,
      "learning_rate": 2.6500000000000005e-06,
      "loss": 0.1309,
      "step": 53
    },
    {
      "epoch": 0.23109843081312412,
      "grad_norm": 1.3658019304275513,
      "learning_rate": 2.7000000000000004e-06,
      "loss": 0.1122,
      "step": 54
    },
    {
      "epoch": 0.23537803138373753,
      "grad_norm": 1.1719582080841064,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.1239,
      "step": 55
    },
    {
      "epoch": 0.2396576319543509,
      "grad_norm": 1.030941367149353,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.1124,
      "step": 56
    },
    {
      "epoch": 0.24393723252496433,
      "grad_norm": 0.7212232947349548,
      "learning_rate": 2.85e-06,
      "loss": 0.0848,
      "step": 57
    },
    {
      "epoch": 0.24821683309557774,
      "grad_norm": 0.8292589783668518,
      "learning_rate": 2.9e-06,
      "loss": 0.0976,
      "step": 58
    },
    {
      "epoch": 0.2524964336661912,
      "grad_norm": 1.0893069505691528,
      "learning_rate": 2.95e-06,
      "loss": 0.0954,
      "step": 59
    },
    {
      "epoch": 0.25677603423680456,
      "grad_norm": 0.8184618353843689,
      "learning_rate": 3e-06,
      "loss": 0.1029,
      "step": 60
    },
    {
      "epoch": 0.26105563480741795,
      "grad_norm": 0.551844596862793,
      "learning_rate": 3.05e-06,
      "loss": 0.0846,
      "step": 61
    },
    {
      "epoch": 0.2653352353780314,
      "grad_norm": 0.6214303970336914,
      "learning_rate": 3.1000000000000004e-06,
      "loss": 0.0855,
      "step": 62
    },
    {
      "epoch": 0.26961483594864477,
      "grad_norm": 0.6127054691314697,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 0.0858,
      "step": 63
    },
    {
      "epoch": 0.2738944365192582,
      "grad_norm": 0.6121392250061035,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.0913,
      "step": 64
    },
    {
      "epoch": 0.2781740370898716,
      "grad_norm": 0.5743774771690369,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.0813,
      "step": 65
    },
    {
      "epoch": 0.28245363766048504,
      "grad_norm": 0.6720462441444397,
      "learning_rate": 3.3000000000000006e-06,
      "loss": 0.0871,
      "step": 66
    },
    {
      "epoch": 0.2867332382310984,
      "grad_norm": 0.5387166142463684,
      "learning_rate": 3.3500000000000005e-06,
      "loss": 0.0722,
      "step": 67
    },
    {
      "epoch": 0.29101283880171186,
      "grad_norm": 0.4062807261943817,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.0699,
      "step": 68
    },
    {
      "epoch": 0.29529243937232524,
      "grad_norm": 0.43954649567604065,
      "learning_rate": 3.45e-06,
      "loss": 0.0808,
      "step": 69
    },
    {
      "epoch": 0.2995720399429387,
      "grad_norm": 0.5136083960533142,
      "learning_rate": 3.5e-06,
      "loss": 0.0782,
      "step": 70
    },
    {
      "epoch": 0.30385164051355207,
      "grad_norm": 0.43194347620010376,
      "learning_rate": 3.5500000000000003e-06,
      "loss": 0.0883,
      "step": 71
    },
    {
      "epoch": 0.30813124108416545,
      "grad_norm": 0.42371511459350586,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.0875,
      "step": 72
    },
    {
      "epoch": 0.3124108416547789,
      "grad_norm": 0.4384445250034332,
      "learning_rate": 3.65e-06,
      "loss": 0.0842,
      "step": 73
    },
    {
      "epoch": 0.3166904422253923,
      "grad_norm": 0.3909939229488373,
      "learning_rate": 3.7e-06,
      "loss": 0.0718,
      "step": 74
    },
    {
      "epoch": 0.3209700427960057,
      "grad_norm": 0.33483919501304626,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.0766,
      "step": 75
    },
    {
      "epoch": 0.3252496433666191,
      "grad_norm": 0.6500667929649353,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 0.0841,
      "step": 76
    },
    {
      "epoch": 0.32952924393723254,
      "grad_norm": 0.42915841937065125,
      "learning_rate": 3.85e-06,
      "loss": 0.0743,
      "step": 77
    },
    {
      "epoch": 0.3338088445078459,
      "grad_norm": 0.5390797853469849,
      "learning_rate": 3.900000000000001e-06,
      "loss": 0.0784,
      "step": 78
    },
    {
      "epoch": 0.33808844507845937,
      "grad_norm": 0.4562499225139618,
      "learning_rate": 3.95e-06,
      "loss": 0.0768,
      "step": 79
    },
    {
      "epoch": 0.34236804564907275,
      "grad_norm": 0.45602527260780334,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.076,
      "step": 80
    },
    {
      "epoch": 0.3466476462196862,
      "grad_norm": 0.3802862763404846,
      "learning_rate": 4.05e-06,
      "loss": 0.0776,
      "step": 81
    },
    {
      "epoch": 0.3509272467902996,
      "grad_norm": 0.3189656138420105,
      "learning_rate": 4.1e-06,
      "loss": 0.0695,
      "step": 82
    },
    {
      "epoch": 0.35520684736091296,
      "grad_norm": 0.37464478611946106,
      "learning_rate": 4.15e-06,
      "loss": 0.0732,
      "step": 83
    },
    {
      "epoch": 0.3594864479315264,
      "grad_norm": 0.5200878977775574,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.0646,
      "step": 84
    },
    {
      "epoch": 0.3637660485021398,
      "grad_norm": 0.3902634382247925,
      "learning_rate": 4.25e-06,
      "loss": 0.0751,
      "step": 85
    },
    {
      "epoch": 0.3680456490727532,
      "grad_norm": 0.4341444969177246,
      "learning_rate": 4.3e-06,
      "loss": 0.0739,
      "step": 86
    },
    {
      "epoch": 0.3723252496433666,
      "grad_norm": 0.3288861811161041,
      "learning_rate": 4.350000000000001e-06,
      "loss": 0.0631,
      "step": 87
    },
    {
      "epoch": 0.37660485021398005,
      "grad_norm": 0.33849674463272095,
      "learning_rate": 4.4e-06,
      "loss": 0.0663,
      "step": 88
    },
    {
      "epoch": 0.38088445078459343,
      "grad_norm": 0.36165380477905273,
      "learning_rate": 4.450000000000001e-06,
      "loss": 0.0618,
      "step": 89
    },
    {
      "epoch": 0.38516405135520687,
      "grad_norm": 0.5453753471374512,
      "learning_rate": 4.5e-06,
      "loss": 0.0734,
      "step": 90
    },
    {
      "epoch": 0.38944365192582026,
      "grad_norm": 0.4055081605911255,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 0.0684,
      "step": 91
    },
    {
      "epoch": 0.39372325249643364,
      "grad_norm": 0.4177473783493042,
      "learning_rate": 4.600000000000001e-06,
      "loss": 0.0605,
      "step": 92
    },
    {
      "epoch": 0.3980028530670471,
      "grad_norm": 0.35140708088874817,
      "learning_rate": 4.65e-06,
      "loss": 0.0695,
      "step": 93
    },
    {
      "epoch": 0.40228245363766046,
      "grad_norm": 0.5472511053085327,
      "learning_rate": 4.7e-06,
      "loss": 0.0743,
      "step": 94
    },
    {
      "epoch": 0.4065620542082739,
      "grad_norm": 0.35946714878082275,
      "learning_rate": 4.75e-06,
      "loss": 0.0658,
      "step": 95
    },
    {
      "epoch": 0.4108416547788873,
      "grad_norm": 0.3542482852935791,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.0707,
      "step": 96
    },
    {
      "epoch": 0.41512125534950073,
      "grad_norm": 0.5081551671028137,
      "learning_rate": 4.85e-06,
      "loss": 0.0774,
      "step": 97
    },
    {
      "epoch": 0.4194008559201141,
      "grad_norm": 0.3265087902545929,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.0715,
      "step": 98
    },
    {
      "epoch": 0.42368045649072755,
      "grad_norm": 0.4452495574951172,
      "learning_rate": 4.95e-06,
      "loss": 0.064,
      "step": 99
    },
    {
      "epoch": 0.42796005706134094,
      "grad_norm": 0.33817920088768005,
      "learning_rate": 5e-06,
      "loss": 0.0674,
      "step": 100
    },
    {
      "epoch": 0.4322396576319544,
      "grad_norm": 0.4477202296257019,
      "learning_rate": 4.9999926774868305e-06,
      "loss": 0.0701,
      "step": 101
    },
    {
      "epoch": 0.43651925820256776,
      "grad_norm": 0.3426607847213745,
      "learning_rate": 4.999970709990216e-06,
      "loss": 0.0637,
      "step": 102
    },
    {
      "epoch": 0.44079885877318115,
      "grad_norm": 0.4300517737865448,
      "learning_rate": 4.999934097638842e-06,
      "loss": 0.0716,
      "step": 103
    },
    {
      "epoch": 0.4450784593437946,
      "grad_norm": 0.29154735803604126,
      "learning_rate": 4.9998828406471855e-06,
      "loss": 0.066,
      "step": 104
    },
    {
      "epoch": 0.44935805991440797,
      "grad_norm": 0.38544416427612305,
      "learning_rate": 4.99981693931551e-06,
      "loss": 0.0735,
      "step": 105
    },
    {
      "epoch": 0.4536376604850214,
      "grad_norm": 0.42879173159599304,
      "learning_rate": 4.999736394029866e-06,
      "loss": 0.0724,
      "step": 106
    },
    {
      "epoch": 0.4579172610556348,
      "grad_norm": 0.45247069001197815,
      "learning_rate": 4.999641205262088e-06,
      "loss": 0.0624,
      "step": 107
    },
    {
      "epoch": 0.46219686162624823,
      "grad_norm": 0.45127418637275696,
      "learning_rate": 4.9995313735697935e-06,
      "loss": 0.0635,
      "step": 108
    },
    {
      "epoch": 0.4664764621968616,
      "grad_norm": 0.4134274125099182,
      "learning_rate": 4.999406899596378e-06,
      "loss": 0.0654,
      "step": 109
    },
    {
      "epoch": 0.47075606276747506,
      "grad_norm": 0.5010129809379578,
      "learning_rate": 4.99926778407101e-06,
      "loss": 0.0678,
      "step": 110
    },
    {
      "epoch": 0.47503566333808844,
      "grad_norm": 0.7583813667297363,
      "learning_rate": 4.999114027808632e-06,
      "loss": 0.0729,
      "step": 111
    },
    {
      "epoch": 0.4793152639087018,
      "grad_norm": 0.39662906527519226,
      "learning_rate": 4.998945631709948e-06,
      "loss": 0.0688,
      "step": 112
    },
    {
      "epoch": 0.48359486447931527,
      "grad_norm": 0.33845070004463196,
      "learning_rate": 4.998762596761424e-06,
      "loss": 0.0689,
      "step": 113
    },
    {
      "epoch": 0.48787446504992865,
      "grad_norm": 0.6589358448982239,
      "learning_rate": 4.998564924035282e-06,
      "loss": 0.0664,
      "step": 114
    },
    {
      "epoch": 0.4921540656205421,
      "grad_norm": 0.4025883078575134,
      "learning_rate": 4.99835261468949e-06,
      "loss": 0.0731,
      "step": 115
    },
    {
      "epoch": 0.4964336661911555,
      "grad_norm": 0.32868894934654236,
      "learning_rate": 4.998125669967758e-06,
      "loss": 0.0657,
      "step": 116
    },
    {
      "epoch": 0.5007132667617689,
      "grad_norm": 0.3705316483974457,
      "learning_rate": 4.997884091199531e-06,
      "loss": 0.0668,
      "step": 117
    },
    {
      "epoch": 0.5049928673323824,
      "grad_norm": 0.32613426446914673,
      "learning_rate": 4.997627879799981e-06,
      "loss": 0.0654,
      "step": 118
    },
    {
      "epoch": 0.5092724679029957,
      "grad_norm": 0.42335274815559387,
      "learning_rate": 4.997357037269996e-06,
      "loss": 0.0695,
      "step": 119
    },
    {
      "epoch": 0.5135520684736091,
      "grad_norm": 0.39254000782966614,
      "learning_rate": 4.997071565196175e-06,
      "loss": 0.0663,
      "step": 120
    },
    {
      "epoch": 0.5178316690442225,
      "grad_norm": 0.33641666173934937,
      "learning_rate": 4.996771465250814e-06,
      "loss": 0.0626,
      "step": 121
    },
    {
      "epoch": 0.5221112696148359,
      "grad_norm": 0.38578540086746216,
      "learning_rate": 4.996456739191905e-06,
      "loss": 0.064,
      "step": 122
    },
    {
      "epoch": 0.5263908701854494,
      "grad_norm": 0.31561926007270813,
      "learning_rate": 4.996127388863116e-06,
      "loss": 0.0626,
      "step": 123
    },
    {
      "epoch": 0.5306704707560628,
      "grad_norm": 0.27469775080680847,
      "learning_rate": 4.995783416193782e-06,
      "loss": 0.0617,
      "step": 124
    },
    {
      "epoch": 0.5349500713266762,
      "grad_norm": 0.2643572986125946,
      "learning_rate": 4.9954248231989016e-06,
      "loss": 0.0592,
      "step": 125
    },
    {
      "epoch": 0.5392296718972895,
      "grad_norm": 0.32749131321907043,
      "learning_rate": 4.995051611979115e-06,
      "loss": 0.0597,
      "step": 126
    },
    {
      "epoch": 0.543509272467903,
      "grad_norm": 0.4380977153778076,
      "learning_rate": 4.994663784720698e-06,
      "loss": 0.0663,
      "step": 127
    },
    {
      "epoch": 0.5477888730385164,
      "grad_norm": 0.30976516008377075,
      "learning_rate": 4.994261343695546e-06,
      "loss": 0.0634,
      "step": 128
    },
    {
      "epoch": 0.5520684736091298,
      "grad_norm": 0.4299394488334656,
      "learning_rate": 4.9938442912611625e-06,
      "loss": 0.065,
      "step": 129
    },
    {
      "epoch": 0.5563480741797432,
      "grad_norm": 0.3520357012748718,
      "learning_rate": 4.993412629860646e-06,
      "loss": 0.059,
      "step": 130
    },
    {
      "epoch": 0.5606276747503567,
      "grad_norm": 0.44977450370788574,
      "learning_rate": 4.9929663620226734e-06,
      "loss": 0.0639,
      "step": 131
    },
    {
      "epoch": 0.5649072753209701,
      "grad_norm": 0.3896942436695099,
      "learning_rate": 4.992505490361487e-06,
      "loss": 0.0608,
      "step": 132
    },
    {
      "epoch": 0.5691868758915835,
      "grad_norm": 0.3307473063468933,
      "learning_rate": 4.992030017576876e-06,
      "loss": 0.0649,
      "step": 133
    },
    {
      "epoch": 0.5734664764621968,
      "grad_norm": 0.27266785502433777,
      "learning_rate": 4.991539946454166e-06,
      "loss": 0.058,
      "step": 134
    },
    {
      "epoch": 0.5777460770328102,
      "grad_norm": 0.43257489800453186,
      "learning_rate": 4.991035279864199e-06,
      "loss": 0.0651,
      "step": 135
    },
    {
      "epoch": 0.5820256776034237,
      "grad_norm": 0.33029407262802124,
      "learning_rate": 4.990516020763317e-06,
      "loss": 0.0601,
      "step": 136
    },
    {
      "epoch": 0.5863052781740371,
      "grad_norm": 0.40888652205467224,
      "learning_rate": 4.989982172193346e-06,
      "loss": 0.0676,
      "step": 137
    },
    {
      "epoch": 0.5905848787446505,
      "grad_norm": 0.2740240693092346,
      "learning_rate": 4.989433737281576e-06,
      "loss": 0.0599,
      "step": 138
    },
    {
      "epoch": 0.5948644793152639,
      "grad_norm": 0.3733709156513214,
      "learning_rate": 4.988870719240744e-06,
      "loss": 0.0571,
      "step": 139
    },
    {
      "epoch": 0.5991440798858774,
      "grad_norm": 0.33992213010787964,
      "learning_rate": 4.988293121369016e-06,
      "loss": 0.0615,
      "step": 140
    },
    {
      "epoch": 0.6034236804564908,
      "grad_norm": 0.34187227487564087,
      "learning_rate": 4.987700947049966e-06,
      "loss": 0.0625,
      "step": 141
    },
    {
      "epoch": 0.6077032810271041,
      "grad_norm": 0.35290244221687317,
      "learning_rate": 4.987094199752558e-06,
      "loss": 0.0606,
      "step": 142
    },
    {
      "epoch": 0.6119828815977175,
      "grad_norm": 0.2595992982387543,
      "learning_rate": 4.986472883031124e-06,
      "loss": 0.0551,
      "step": 143
    },
    {
      "epoch": 0.6162624821683309,
      "grad_norm": 0.43302714824676514,
      "learning_rate": 4.9858370005253435e-06,
      "loss": 0.0643,
      "step": 144
    },
    {
      "epoch": 0.6205420827389444,
      "grad_norm": 0.3430291414260864,
      "learning_rate": 4.985186555960223e-06,
      "loss": 0.0576,
      "step": 145
    },
    {
      "epoch": 0.6248216833095578,
      "grad_norm": 0.5924923419952393,
      "learning_rate": 4.984521553146074e-06,
      "loss": 0.0679,
      "step": 146
    },
    {
      "epoch": 0.6291012838801712,
      "grad_norm": 0.3352043330669403,
      "learning_rate": 4.98384199597849e-06,
      "loss": 0.0636,
      "step": 147
    },
    {
      "epoch": 0.6333808844507846,
      "grad_norm": 0.31009000539779663,
      "learning_rate": 4.983147888438324e-06,
      "loss": 0.0499,
      "step": 148
    },
    {
      "epoch": 0.637660485021398,
      "grad_norm": 0.41210371255874634,
      "learning_rate": 4.982439234591665e-06,
      "loss": 0.0632,
      "step": 149
    },
    {
      "epoch": 0.6419400855920114,
      "grad_norm": 0.30768534541130066,
      "learning_rate": 4.9817160385898145e-06,
      "loss": 0.0669,
      "step": 150
    },
    {
      "epoch": 0.6462196861626248,
      "grad_norm": 0.4492046535015106,
      "learning_rate": 4.980978304669263e-06,
      "loss": 0.0603,
      "step": 151
    },
    {
      "epoch": 0.6504992867332382,
      "grad_norm": 0.3134528696537018,
      "learning_rate": 4.9802260371516635e-06,
      "loss": 0.0553,
      "step": 152
    },
    {
      "epoch": 0.6547788873038516,
      "grad_norm": 0.33358660340309143,
      "learning_rate": 4.979459240443806e-06,
      "loss": 0.061,
      "step": 153
    },
    {
      "epoch": 0.6590584878744651,
      "grad_norm": 0.2705594301223755,
      "learning_rate": 4.978677919037594e-06,
      "loss": 0.0525,
      "step": 154
    },
    {
      "epoch": 0.6633380884450785,
      "grad_norm": 0.4429916739463806,
      "learning_rate": 4.977882077510018e-06,
      "loss": 0.0595,
      "step": 155
    },
    {
      "epoch": 0.6676176890156919,
      "grad_norm": 0.37663739919662476,
      "learning_rate": 4.977071720523125e-06,
      "loss": 0.0596,
      "step": 156
    },
    {
      "epoch": 0.6718972895863052,
      "grad_norm": 0.5795411467552185,
      "learning_rate": 4.9762468528239945e-06,
      "loss": 0.0629,
      "step": 157
    },
    {
      "epoch": 0.6761768901569187,
      "grad_norm": 0.2995544373989105,
      "learning_rate": 4.975407479244711e-06,
      "loss": 0.0577,
      "step": 158
    },
    {
      "epoch": 0.6804564907275321,
      "grad_norm": 0.2688761055469513,
      "learning_rate": 4.974553604702332e-06,
      "loss": 0.0612,
      "step": 159
    },
    {
      "epoch": 0.6847360912981455,
      "grad_norm": 0.32369598746299744,
      "learning_rate": 4.9736852341988655e-06,
      "loss": 0.054,
      "step": 160
    },
    {
      "epoch": 0.6890156918687589,
      "grad_norm": 0.24816952645778656,
      "learning_rate": 4.972802372821235e-06,
      "loss": 0.0631,
      "step": 161
    },
    {
      "epoch": 0.6932952924393724,
      "grad_norm": 0.33390095829963684,
      "learning_rate": 4.97190502574125e-06,
      "loss": 0.0629,
      "step": 162
    },
    {
      "epoch": 0.6975748930099858,
      "grad_norm": 0.2538573443889618,
      "learning_rate": 4.9709931982155805e-06,
      "loss": 0.0585,
      "step": 163
    },
    {
      "epoch": 0.7018544935805991,
      "grad_norm": 0.337729275226593,
      "learning_rate": 4.970066895585722e-06,
      "loss": 0.0534,
      "step": 164
    },
    {
      "epoch": 0.7061340941512125,
      "grad_norm": 0.3311474621295929,
      "learning_rate": 4.9691261232779644e-06,
      "loss": 0.0592,
      "step": 165
    },
    {
      "epoch": 0.7104136947218259,
      "grad_norm": 0.4538433849811554,
      "learning_rate": 4.968170886803361e-06,
      "loss": 0.0607,
      "step": 166
    },
    {
      "epoch": 0.7146932952924394,
      "grad_norm": 0.3233712911605835,
      "learning_rate": 4.9672011917577e-06,
      "loss": 0.0557,
      "step": 167
    },
    {
      "epoch": 0.7189728958630528,
      "grad_norm": 0.41257205605506897,
      "learning_rate": 4.966217043821462e-06,
      "loss": 0.0618,
      "step": 168
    },
    {
      "epoch": 0.7232524964336662,
      "grad_norm": 0.36823365092277527,
      "learning_rate": 4.965218448759797e-06,
      "loss": 0.0619,
      "step": 169
    },
    {
      "epoch": 0.7275320970042796,
      "grad_norm": 0.38311564922332764,
      "learning_rate": 4.964205412422487e-06,
      "loss": 0.0668,
      "step": 170
    },
    {
      "epoch": 0.7318116975748931,
      "grad_norm": 0.31038492918014526,
      "learning_rate": 4.963177940743908e-06,
      "loss": 0.0573,
      "step": 171
    },
    {
      "epoch": 0.7360912981455064,
      "grad_norm": 0.2835666835308075,
      "learning_rate": 4.962136039742998e-06,
      "loss": 0.0649,
      "step": 172
    },
    {
      "epoch": 0.7403708987161198,
      "grad_norm": 0.4525282680988312,
      "learning_rate": 4.961079715523228e-06,
      "loss": 0.0593,
      "step": 173
    },
    {
      "epoch": 0.7446504992867332,
      "grad_norm": 0.33394986391067505,
      "learning_rate": 4.9600089742725545e-06,
      "loss": 0.0601,
      "step": 174
    },
    {
      "epoch": 0.7489300998573466,
      "grad_norm": 0.2642901837825775,
      "learning_rate": 4.95892382226339e-06,
      "loss": 0.0623,
      "step": 175
    },
    {
      "epoch": 0.7532097004279601,
      "grad_norm": 0.34347841143608093,
      "learning_rate": 4.95782426585257e-06,
      "loss": 0.0568,
      "step": 176
    },
    {
      "epoch": 0.7574893009985735,
      "grad_norm": 0.30288106203079224,
      "learning_rate": 4.956710311481303e-06,
      "loss": 0.0569,
      "step": 177
    },
    {
      "epoch": 0.7617689015691869,
      "grad_norm": 0.285087913274765,
      "learning_rate": 4.955581965675148e-06,
      "loss": 0.0534,
      "step": 178
    },
    {
      "epoch": 0.7660485021398002,
      "grad_norm": 0.2861518859863281,
      "learning_rate": 4.954439235043966e-06,
      "loss": 0.0536,
      "step": 179
    },
    {
      "epoch": 0.7703281027104137,
      "grad_norm": 0.4016513526439667,
      "learning_rate": 4.953282126281884e-06,
      "loss": 0.0575,
      "step": 180
    },
    {
      "epoch": 0.7746077032810271,
      "grad_norm": 0.4838986098766327,
      "learning_rate": 4.9521106461672605e-06,
      "loss": 0.0666,
      "step": 181
    },
    {
      "epoch": 0.7788873038516405,
      "grad_norm": 0.270673006772995,
      "learning_rate": 4.950924801562636e-06,
      "loss": 0.0579,
      "step": 182
    },
    {
      "epoch": 0.7831669044222539,
      "grad_norm": 0.31113702058792114,
      "learning_rate": 4.9497245994147e-06,
      "loss": 0.0546,
      "step": 183
    },
    {
      "epoch": 0.7874465049928673,
      "grad_norm": 0.2960825264453888,
      "learning_rate": 4.948510046754251e-06,
      "loss": 0.0553,
      "step": 184
    },
    {
      "epoch": 0.7917261055634808,
      "grad_norm": 0.3720844089984894,
      "learning_rate": 4.947281150696151e-06,
      "loss": 0.0607,
      "step": 185
    },
    {
      "epoch": 0.7960057061340942,
      "grad_norm": 0.2569741904735565,
      "learning_rate": 4.946037918439285e-06,
      "loss": 0.0552,
      "step": 186
    },
    {
      "epoch": 0.8002853067047075,
      "grad_norm": 0.3115193843841553,
      "learning_rate": 4.944780357266522e-06,
      "loss": 0.0544,
      "step": 187
    },
    {
      "epoch": 0.8045649072753209,
      "grad_norm": 0.2828395366668701,
      "learning_rate": 4.943508474544667e-06,
      "loss": 0.0541,
      "step": 188
    },
    {
      "epoch": 0.8088445078459344,
      "grad_norm": 0.29760900139808655,
      "learning_rate": 4.942222277724424e-06,
      "loss": 0.0576,
      "step": 189
    },
    {
      "epoch": 0.8131241084165478,
      "grad_norm": 0.3851959705352783,
      "learning_rate": 4.940921774340347e-06,
      "loss": 0.059,
      "step": 190
    },
    {
      "epoch": 0.8174037089871612,
      "grad_norm": 0.35299891233444214,
      "learning_rate": 4.939606972010798e-06,
      "loss": 0.0627,
      "step": 191
    },
    {
      "epoch": 0.8216833095577746,
      "grad_norm": 0.30920645594596863,
      "learning_rate": 4.9382778784379036e-06,
      "loss": 0.053,
      "step": 192
    },
    {
      "epoch": 0.8259629101283881,
      "grad_norm": 0.31357741355895996,
      "learning_rate": 4.936934501407507e-06,
      "loss": 0.0523,
      "step": 193
    },
    {
      "epoch": 0.8302425106990015,
      "grad_norm": 0.28295785188674927,
      "learning_rate": 4.935576848789127e-06,
      "loss": 0.0521,
      "step": 194
    },
    {
      "epoch": 0.8345221112696148,
      "grad_norm": 0.3303717076778412,
      "learning_rate": 4.934204928535905e-06,
      "loss": 0.0564,
      "step": 195
    },
    {
      "epoch": 0.8388017118402282,
      "grad_norm": 0.31010597944259644,
      "learning_rate": 4.932818748684565e-06,
      "loss": 0.0495,
      "step": 196
    },
    {
      "epoch": 0.8430813124108416,
      "grad_norm": 0.33654817938804626,
      "learning_rate": 4.931418317355362e-06,
      "loss": 0.0536,
      "step": 197
    },
    {
      "epoch": 0.8473609129814551,
      "grad_norm": 0.35488635301589966,
      "learning_rate": 4.9300036427520396e-06,
      "loss": 0.054,
      "step": 198
    },
    {
      "epoch": 0.8516405135520685,
      "grad_norm": 0.37233811616897583,
      "learning_rate": 4.928574733161775e-06,
      "loss": 0.0527,
      "step": 199
    },
    {
      "epoch": 0.8559201141226819,
      "grad_norm": 0.3253094553947449,
      "learning_rate": 4.927131596955137e-06,
      "loss": 0.0545,
      "step": 200
    },
    {
      "epoch": 0.8601997146932953,
      "grad_norm": 0.345102995634079,
      "learning_rate": 4.92567424258603e-06,
      "loss": 0.0606,
      "step": 201
    },
    {
      "epoch": 0.8644793152639088,
      "grad_norm": 0.32971855998039246,
      "learning_rate": 4.924202678591655e-06,
      "loss": 0.0542,
      "step": 202
    },
    {
      "epoch": 0.8687589158345221,
      "grad_norm": 0.30636993050575256,
      "learning_rate": 4.922716913592447e-06,
      "loss": 0.0509,
      "step": 203
    },
    {
      "epoch": 0.8730385164051355,
      "grad_norm": 0.27384838461875916,
      "learning_rate": 4.921216956292033e-06,
      "loss": 0.0496,
      "step": 204
    },
    {
      "epoch": 0.8773181169757489,
      "grad_norm": 0.30968862771987915,
      "learning_rate": 4.919702815477179e-06,
      "loss": 0.0608,
      "step": 205
    },
    {
      "epoch": 0.8815977175463623,
      "grad_norm": 0.25974956154823303,
      "learning_rate": 4.918174500017739e-06,
      "loss": 0.0543,
      "step": 206
    },
    {
      "epoch": 0.8858773181169758,
      "grad_norm": 0.30472415685653687,
      "learning_rate": 4.916632018866599e-06,
      "loss": 0.0567,
      "step": 207
    },
    {
      "epoch": 0.8901569186875892,
      "grad_norm": 0.3091549277305603,
      "learning_rate": 4.91507538105963e-06,
      "loss": 0.0511,
      "step": 208
    },
    {
      "epoch": 0.8944365192582026,
      "grad_norm": 0.26851367950439453,
      "learning_rate": 4.9135045957156356e-06,
      "loss": 0.0536,
      "step": 209
    },
    {
      "epoch": 0.8987161198288159,
      "grad_norm": 0.25963643193244934,
      "learning_rate": 4.911919672036291e-06,
      "loss": 0.0509,
      "step": 210
    },
    {
      "epoch": 0.9029957203994294,
      "grad_norm": 0.28279975056648254,
      "learning_rate": 4.910320619306094e-06,
      "loss": 0.0538,
      "step": 211
    },
    {
      "epoch": 0.9072753209700428,
      "grad_norm": 0.26387840509414673,
      "learning_rate": 4.908707446892315e-06,
      "loss": 0.0504,
      "step": 212
    },
    {
      "epoch": 0.9115549215406562,
      "grad_norm": 0.2996926009654999,
      "learning_rate": 4.907080164244934e-06,
      "loss": 0.0497,
      "step": 213
    },
    {
      "epoch": 0.9158345221112696,
      "grad_norm": 0.4072028398513794,
      "learning_rate": 4.905438780896589e-06,
      "loss": 0.053,
      "step": 214
    },
    {
      "epoch": 0.920114122681883,
      "grad_norm": 0.5946587324142456,
      "learning_rate": 4.903783306462522e-06,
      "loss": 0.0577,
      "step": 215
    },
    {
      "epoch": 0.9243937232524965,
      "grad_norm": 0.34546515345573425,
      "learning_rate": 4.9021137506405205e-06,
      "loss": 0.0554,
      "step": 216
    },
    {
      "epoch": 0.9286733238231099,
      "grad_norm": 0.3417337238788605,
      "learning_rate": 4.9004301232108585e-06,
      "loss": 0.0529,
      "step": 217
    },
    {
      "epoch": 0.9329529243937232,
      "grad_norm": 0.31937581300735474,
      "learning_rate": 4.8987324340362445e-06,
      "loss": 0.0536,
      "step": 218
    },
    {
      "epoch": 0.9372325249643366,
      "grad_norm": 0.28485214710235596,
      "learning_rate": 4.897020693061758e-06,
      "loss": 0.0488,
      "step": 219
    },
    {
      "epoch": 0.9415121255349501,
      "grad_norm": 0.3833935558795929,
      "learning_rate": 4.895294910314797e-06,
      "loss": 0.0577,
      "step": 220
    },
    {
      "epoch": 0.9457917261055635,
      "grad_norm": 0.2772054076194763,
      "learning_rate": 4.8935550959050135e-06,
      "loss": 0.0506,
      "step": 221
    },
    {
      "epoch": 0.9500713266761769,
      "grad_norm": 0.35764405131340027,
      "learning_rate": 4.891801260024262e-06,
      "loss": 0.0518,
      "step": 222
    },
    {
      "epoch": 0.9543509272467903,
      "grad_norm": 0.30019721388816833,
      "learning_rate": 4.890033412946527e-06,
      "loss": 0.0525,
      "step": 223
    },
    {
      "epoch": 0.9586305278174037,
      "grad_norm": 0.3105527460575104,
      "learning_rate": 4.888251565027879e-06,
      "loss": 0.0552,
      "step": 224
    },
    {
      "epoch": 0.9629101283880172,
      "grad_norm": 0.35943159461021423,
      "learning_rate": 4.8864557267064e-06,
      "loss": 0.0543,
      "step": 225
    },
    {
      "epoch": 0.9671897289586305,
      "grad_norm": 0.3295605480670929,
      "learning_rate": 4.88464590850213e-06,
      "loss": 0.0565,
      "step": 226
    },
    {
      "epoch": 0.9714693295292439,
      "grad_norm": 0.31729599833488464,
      "learning_rate": 4.882822121017004e-06,
      "loss": 0.0524,
      "step": 227
    },
    {
      "epoch": 0.9757489300998573,
      "grad_norm": 0.4707350432872772,
      "learning_rate": 4.880984374934788e-06,
      "loss": 0.0589,
      "step": 228
    },
    {
      "epoch": 0.9800285306704708,
      "grad_norm": 0.35285064578056335,
      "learning_rate": 4.879132681021017e-06,
      "loss": 0.0509,
      "step": 229
    },
    {
      "epoch": 0.9843081312410842,
      "grad_norm": 0.32979413866996765,
      "learning_rate": 4.877267050122934e-06,
      "loss": 0.051,
      "step": 230
    },
    {
      "epoch": 0.9885877318116976,
      "grad_norm": 0.3415399491786957,
      "learning_rate": 4.8753874931694254e-06,
      "loss": 0.0507,
      "step": 231
    },
    {
      "epoch": 0.992867332382311,
      "grad_norm": 0.3240598738193512,
      "learning_rate": 4.8734940211709535e-06,
      "loss": 0.0524,
      "step": 232
    },
    {
      "epoch": 0.9971469329529244,
      "grad_norm": 0.31305283308029175,
      "learning_rate": 4.871586645219499e-06,
      "loss": 0.0528,
      "step": 233
    }
  ],
  "logging_steps": 1,
  "max_steps": 1398,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 233,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.219533199339684e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}