{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.496,
  "eval_steps": 156,
  "global_step": 312,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 0.0001,
      "loss": 1.2612,
      "step": 1
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5839,
      "step": 2
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0001994638069705094,
      "loss": 1.2718,
      "step": 3
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.00019892761394101877,
      "loss": 0.9663,
      "step": 4
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.00019839142091152817,
      "loss": 0.8353,
      "step": 5
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.00019785522788203755,
      "loss": 0.4235,
      "step": 6
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.00019731903485254693,
      "loss": 0.3499,
      "step": 7
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0001967828418230563,
      "loss": 0.8885,
      "step": 8
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.00019624664879356568,
      "loss": 0.6318,
      "step": 9
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.00019571045576407506,
      "loss": 0.7978,
      "step": 10
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.00019517426273458447,
      "loss": 0.3304,
      "step": 11
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.00019463806970509384,
      "loss": 0.8138,
      "step": 12
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.00019410187667560322,
      "loss": 0.3049,
      "step": 13
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0001935656836461126,
      "loss": 0.9467,
      "step": 14
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.00019302949061662198,
      "loss": 0.4926,
      "step": 15
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.00019249329758713138,
      "loss": 0.2896,
      "step": 16
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.00019195710455764076,
      "loss": 0.314,
      "step": 17
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.00019142091152815014,
      "loss": 0.6319,
      "step": 18
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.00019088471849865952,
      "loss": 0.8047,
      "step": 19
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0001903485254691689,
      "loss": 0.6664,
      "step": 20
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.00018981233243967828,
      "loss": 0.7509,
      "step": 21
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.00018927613941018768,
      "loss": 0.572,
      "step": 22
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.00018873994638069706,
      "loss": 0.6726,
      "step": 23
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.00018820375335120644,
      "loss": 0.2929,
      "step": 24
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.00018766756032171581,
      "loss": 0.5821,
      "step": 25
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0001871313672922252,
      "loss": 0.5325,
      "step": 26
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.00018659517426273457,
      "loss": 0.4568,
      "step": 27
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.00018605898123324398,
      "loss": 0.2517,
      "step": 28
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.00018552278820375335,
      "loss": 0.5343,
      "step": 29
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00018498659517426273,
      "loss": 0.3203,
      "step": 30
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0001844504021447721,
      "loss": 0.0,
      "step": 31
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0001839142091152815,
      "loss": 0.2905,
      "step": 32
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0001833780160857909,
      "loss": 0.5738,
      "step": 33
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0001828418230563003,
      "loss": 0.5445,
      "step": 34
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.00018230563002680968,
      "loss": 0.6075,
      "step": 35
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.00018176943699731906,
      "loss": 0.6869,
      "step": 36
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.00018123324396782843,
      "loss": 0.2808,
      "step": 37
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0001806970509383378,
      "loss": 0.6005,
      "step": 38
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0001801608579088472,
      "loss": 0.5839,
      "step": 39
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0001796246648793566,
      "loss": 0.5602,
      "step": 40
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.00017908847184986597,
      "loss": 0.5986,
      "step": 41
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.00017855227882037535,
      "loss": 0.3701,
      "step": 42
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.00017801608579088473,
      "loss": 0.2545,
      "step": 43
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0001774798927613941,
      "loss": 0.6094,
      "step": 44
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0001769436997319035,
      "loss": 0.0,
      "step": 45
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0001764075067024129,
      "loss": 0.4668,
      "step": 46
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.00017587131367292227,
      "loss": 0.2302,
      "step": 47
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.00017533512064343165,
      "loss": 0.2882,
      "step": 48
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.00017479892761394103,
      "loss": 0.2486,
      "step": 49
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0001742627345844504,
      "loss": 0.2614,
      "step": 50
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0001737265415549598,
      "loss": 0.237,
      "step": 51
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0001731903485254692,
      "loss": 0.5986,
      "step": 52
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.00017265415549597856,
      "loss": 0.0,
      "step": 53
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.00017211796246648794,
      "loss": 0.5648,
      "step": 54
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00017158176943699732,
      "loss": 0.5096,
      "step": 55
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.00017104557640750673,
      "loss": 0.5139,
      "step": 56
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0001705093833780161,
      "loss": 0.2226,
      "step": 57
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.00016997319034852548,
      "loss": 0.4839,
      "step": 58
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.00016943699731903486,
      "loss": 0.5375,
      "step": 59
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.00016890080428954424,
      "loss": 0.5792,
      "step": 60
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.00016836461126005362,
      "loss": 0.6151,
      "step": 61
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.00016782841823056302,
      "loss": 0.5358,
      "step": 62
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0001672922252010724,
      "loss": 0.4993,
      "step": 63
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.00016675603217158178,
      "loss": 0.2054,
      "step": 64
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.00016621983914209116,
      "loss": 0.2973,
      "step": 65
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.00016568364611260053,
      "loss": 0.4578,
      "step": 66
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.00016514745308310994,
      "loss": 0.4827,
      "step": 67
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.00016461126005361932,
      "loss": 0.2362,
      "step": 68
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0001640750670241287,
      "loss": 0.2993,
      "step": 69
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.00016353887399463807,
      "loss": 0.6205,
      "step": 70
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.00016300268096514745,
      "loss": 0.5475,
      "step": 71
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.00016246648793565683,
      "loss": 0.4761,
      "step": 72
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.00016193029490616624,
      "loss": 0.2454,
      "step": 73
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.00016139410187667561,
      "loss": 0.559,
      "step": 74
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.000160857908847185,
      "loss": 0.4535,
      "step": 75
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.00016032171581769437,
      "loss": 0.5401,
      "step": 76
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.00015978552278820375,
      "loss": 0.4684,
      "step": 77
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.00015924932975871313,
      "loss": 0.2701,
      "step": 78
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.00015871313672922253,
      "loss": 0.2443,
      "step": 79
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0001581769436997319,
      "loss": 0.3073,
      "step": 80
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0001576407506702413,
      "loss": 0.3229,
      "step": 81
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.00015710455764075067,
      "loss": 0.4567,
      "step": 82
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.00015656836461126004,
      "loss": 0.3031,
      "step": 83
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.00015603217158176945,
      "loss": 0.2604,
      "step": 84
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.00015549597855227883,
      "loss": 0.5654,
      "step": 85
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0001549597855227882,
      "loss": 0.2723,
      "step": 86
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00015442359249329758,
      "loss": 0.5312,
      "step": 87
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00015388739946380696,
      "loss": 0.5638,
      "step": 88
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.00015335120643431634,
      "loss": 0.2928,
      "step": 89
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00015281501340482574,
      "loss": 0.5724,
      "step": 90
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.00015227882037533512,
      "loss": 0.5019,
      "step": 91
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0001517426273458445,
      "loss": 0.5651,
      "step": 92
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.00015120643431635388,
      "loss": 0.5588,
      "step": 93
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.00015067024128686326,
      "loss": 0.6167,
      "step": 94
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.00015013404825737266,
      "loss": 0.5652,
      "step": 95
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.00014959785522788207,
      "loss": 0.0,
      "step": 96
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.00014906166219839145,
      "loss": 0.2309,
      "step": 97
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.00014852546916890082,
      "loss": 0.5339,
      "step": 98
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0001479892761394102,
      "loss": 0.522,
      "step": 99
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.00014745308310991958,
      "loss": 0.2473,
      "step": 100
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.00014691689008042896,
      "loss": 0.2674,
      "step": 101
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.00014638069705093836,
      "loss": 0.4833,
      "step": 102
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.00014584450402144774,
      "loss": 0.5153,
      "step": 103
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.00014530831099195712,
      "loss": 0.2366,
      "step": 104
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0001447721179624665,
      "loss": 0.5037,
      "step": 105
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.00014423592493297588,
      "loss": 0.4913,
      "step": 106
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00014369973190348528,
      "loss": 0.4979,
      "step": 107
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00014316353887399466,
      "loss": 0.518,
      "step": 108
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.00014262734584450404,
      "loss": 0.0,
      "step": 109
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.00014209115281501342,
      "loss": 0.5866,
      "step": 110
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0001415549597855228,
      "loss": 0.4525,
      "step": 111
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.00014101876675603217,
      "loss": 0.2802,
      "step": 112
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.00014048257372654158,
      "loss": 0.4686,
      "step": 113
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.00013994638069705096,
      "loss": 0.4963,
      "step": 114
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.00013941018766756033,
      "loss": 0.4853,
      "step": 115
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0001388739946380697,
      "loss": 0.6237,
      "step": 116
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0001383378016085791,
      "loss": 0.582,
      "step": 117
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0001378016085790885,
      "loss": 0.5814,
      "step": 118
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.00013726541554959787,
      "loss": 0.4702,
      "step": 119
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.00013672922252010725,
      "loss": 0.2196,
      "step": 120
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.00013619302949061663,
      "loss": 0.5335,
      "step": 121
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.000135656836461126,
      "loss": 0.5734,
      "step": 122
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.00013512064343163539,
      "loss": 0.0,
      "step": 123
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0001345844504021448,
      "loss": 0.385,
      "step": 124
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.00013404825737265417,
      "loss": 0.4994,
      "step": 125
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.00013351206434316355,
      "loss": 0.2143,
      "step": 126
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.00013297587131367293,
      "loss": 0.609,
      "step": 127
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0001324396782841823,
      "loss": 0.2692,
      "step": 128
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.00013190348525469168,
      "loss": 0.2581,
      "step": 129
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0001313672922252011,
      "loss": 0.3353,
      "step": 130
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.00013083109919571046,
      "loss": 0.4328,
      "step": 131
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.00013029490616621984,
      "loss": 0.2286,
      "step": 132
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.00012975871313672922,
      "loss": 0.4495,
      "step": 133
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0001292225201072386,
      "loss": 0.5591,
      "step": 134
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.000128686327077748,
      "loss": 0.2977,
      "step": 135
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.00012815013404825738,
      "loss": 0.4241,
      "step": 136
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.00012761394101876676,
      "loss": 0.3706,
      "step": 137
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.00012707774798927614,
      "loss": 0.4937,
      "step": 138
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.00012654155495978552,
      "loss": 0.4193,
      "step": 139
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0001260053619302949,
      "loss": 0.4165,
      "step": 140
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0001254691689008043,
      "loss": 0.4585,
      "step": 141
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.00012493297587131368,
      "loss": 0.4031,
      "step": 142
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.00012439678284182306,
      "loss": 0.1778,
      "step": 143
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.00012386058981233243,
      "loss": 0.4911,
      "step": 144
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0001233243967828418,
      "loss": 0.489,
      "step": 145
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.00012278820375335122,
      "loss": 0.3833,
      "step": 146
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0001222520107238606,
      "loss": 0.4249,
      "step": 147
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.00012171581769436997,
      "loss": 0.1813,
      "step": 148
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.00012117962466487935,
      "loss": 0.2181,
      "step": 149
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.00012064343163538874,
      "loss": 0.4004,
      "step": 150
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.00012010723860589812,
      "loss": 0.5599,
      "step": 151
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0001195710455764075,
      "loss": 0.2171,
      "step": 152
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.00011903485254691689,
      "loss": 0.4072,
      "step": 153
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.00011849865951742627,
      "loss": 0.4229,
      "step": 154
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.00011796246648793565,
      "loss": 0.3922,
      "step": 155
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.00011742627345844504,
      "loss": 0.4483,
      "step": 156
    },
    {
      "epoch": 1.25,
      "eval_loss": 0.4745557904243469,
      "eval_runtime": 507.5138,
      "eval_samples_per_second": 1.97,
      "eval_steps_per_second": 0.124,
      "step": 156
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.00011689008042895442,
      "loss": 0.3986,
      "step": 157
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.00011635388739946382,
      "loss": 0.3715,
      "step": 158
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0001158176943699732,
      "loss": 0.3707,
      "step": 159
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.00011528150134048259,
      "loss": 0.2226,
      "step": 160
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.00011474530831099197,
      "loss": 0.1913,
      "step": 161
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.00011420911528150135,
      "loss": 0.0,
      "step": 162
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.00011367292225201074,
      "loss": 0.4847,
      "step": 163
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.00011313672922252012,
      "loss": 0.0,
      "step": 164
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.00011260053619302951,
      "loss": 0.2191,
      "step": 165
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.00011206434316353889,
      "loss": 0.4678,
      "step": 166
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.00011152815013404827,
      "loss": 0.4135,
      "step": 167
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.00011099195710455766,
      "loss": 0.4971,
      "step": 168
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.00011045576407506704,
      "loss": 0.2149,
      "step": 169
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.00010991957104557641,
      "loss": 0.2638,
      "step": 170
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0001093833780160858,
      "loss": 0.5197,
      "step": 171
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.00010884718498659518,
      "loss": 0.4409,
      "step": 172
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.00010831099195710456,
      "loss": 0.4943,
      "step": 173
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.00010777479892761395,
      "loss": 0.4369,
      "step": 174
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.00010723860589812333,
      "loss": 0.281,
      "step": 175
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.00010670241286863272,
      "loss": 0.5053,
      "step": 176
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0001061662198391421,
      "loss": 0.4787,
      "step": 177
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.00010563002680965148,
      "loss": 0.4353,
      "step": 178
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.00010509383378016087,
      "loss": 0.437,
      "step": 179
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.00010455764075067025,
      "loss": 0.2177,
      "step": 180
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.00010402144772117963,
      "loss": 0.0,
      "step": 181
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.00010348525469168902,
      "loss": 0.2128,
      "step": 182
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0001029490616621984,
      "loss": 0.622,
      "step": 183
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.00010241286863270778,
      "loss": 0.0,
      "step": 184
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.00010187667560321717,
      "loss": 0.285,
      "step": 185
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.00010134048257372655,
      "loss": 0.4149,
      "step": 186
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.00010080428954423592,
      "loss": 0.426,
      "step": 187
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.00010026809651474532,
      "loss": 0.5177,
      "step": 188
    },
    {
      "epoch": 1.51,
      "learning_rate": 9.97319034852547e-05,
      "loss": 0.5009,
      "step": 189
    },
    {
      "epoch": 1.52,
      "learning_rate": 9.919571045576409e-05,
      "loss": 0.2128,
      "step": 190
    },
    {
      "epoch": 1.53,
      "learning_rate": 9.865951742627346e-05,
      "loss": 0.2701,
      "step": 191
    },
    {
      "epoch": 1.54,
      "learning_rate": 9.812332439678284e-05,
      "loss": 0.4293,
      "step": 192
    },
    {
      "epoch": 1.54,
      "learning_rate": 9.758713136729223e-05,
      "loss": 0.4286,
      "step": 193
    },
    {
      "epoch": 1.55,
      "learning_rate": 9.705093833780161e-05,
      "loss": 0.4106,
      "step": 194
    },
    {
      "epoch": 1.56,
      "learning_rate": 9.651474530831099e-05,
      "loss": 0.6385,
      "step": 195
    },
    {
      "epoch": 1.57,
      "learning_rate": 9.597855227882038e-05,
      "loss": 0.2073,
      "step": 196
    },
    {
      "epoch": 1.58,
      "learning_rate": 9.544235924932976e-05,
      "loss": 0.5092,
      "step": 197
    },
    {
      "epoch": 1.58,
      "learning_rate": 9.490616621983914e-05,
      "loss": 0.1404,
      "step": 198
    },
    {
      "epoch": 1.59,
      "learning_rate": 9.436997319034853e-05,
      "loss": 0.0,
      "step": 199
    },
    {
      "epoch": 1.6,
      "learning_rate": 9.383378016085791e-05,
      "loss": 0.5251,
      "step": 200
    },
    {
      "epoch": 1.61,
      "learning_rate": 9.329758713136729e-05,
      "loss": 0.4321,
      "step": 201
    },
    {
      "epoch": 1.62,
      "learning_rate": 9.276139410187668e-05,
      "loss": 0.5201,
      "step": 202
    },
    {
      "epoch": 1.62,
      "learning_rate": 9.222520107238606e-05,
      "loss": 0.4687,
      "step": 203
    },
    {
      "epoch": 1.63,
      "learning_rate": 9.168900804289545e-05,
      "loss": 0.0,
      "step": 204
    },
    {
      "epoch": 1.64,
      "learning_rate": 9.115281501340484e-05,
      "loss": 0.4181,
      "step": 205
    },
    {
      "epoch": 1.65,
      "learning_rate": 9.061662198391422e-05,
      "loss": 0.39,
      "step": 206
    },
    {
      "epoch": 1.66,
      "learning_rate": 9.00804289544236e-05,
      "loss": 0.4724,
      "step": 207
    },
    {
      "epoch": 1.66,
      "learning_rate": 8.954423592493299e-05,
      "loss": 0.2146,
      "step": 208
    },
    {
      "epoch": 1.67,
      "learning_rate": 8.900804289544236e-05,
      "loss": 0.4829,
      "step": 209
    },
    {
      "epoch": 1.68,
      "learning_rate": 8.847184986595176e-05,
      "loss": 0.2066,
      "step": 210
    },
    {
      "epoch": 1.69,
      "learning_rate": 8.793565683646113e-05,
      "loss": 0.48,
      "step": 211
    },
    {
      "epoch": 1.7,
      "learning_rate": 8.739946380697051e-05,
      "loss": 0.3879,
      "step": 212
    },
    {
      "epoch": 1.7,
      "learning_rate": 8.68632707774799e-05,
      "loss": 0.3477,
      "step": 213
    },
    {
      "epoch": 1.71,
      "learning_rate": 8.632707774798928e-05,
      "loss": 0.5459,
      "step": 214
    },
    {
      "epoch": 1.72,
      "learning_rate": 8.579088471849866e-05,
      "loss": 0.0,
      "step": 215
    },
    {
      "epoch": 1.73,
      "learning_rate": 8.525469168900805e-05,
      "loss": 0.26,
      "step": 216
    },
    {
      "epoch": 1.74,
      "learning_rate": 8.471849865951743e-05,
      "loss": 0.4626,
      "step": 217
    },
    {
      "epoch": 1.74,
      "learning_rate": 8.418230563002681e-05,
      "loss": 0.4755,
      "step": 218
    },
    {
      "epoch": 1.75,
      "learning_rate": 8.36461126005362e-05,
      "loss": 0.4189,
      "step": 219
    },
    {
      "epoch": 1.76,
      "learning_rate": 8.310991957104558e-05,
      "loss": 0.2322,
      "step": 220
    },
    {
      "epoch": 1.77,
      "learning_rate": 8.257372654155497e-05,
      "loss": 0.3844,
      "step": 221
    },
    {
      "epoch": 1.78,
      "learning_rate": 8.203753351206435e-05,
      "loss": 0.1986,
      "step": 222
    },
    {
      "epoch": 1.78,
      "learning_rate": 8.150134048257373e-05,
      "loss": 0.3902,
      "step": 223
    },
    {
      "epoch": 1.79,
      "learning_rate": 8.096514745308312e-05,
      "loss": 0.4943,
      "step": 224
    },
    {
      "epoch": 1.8,
      "learning_rate": 8.04289544235925e-05,
      "loss": 0.2464,
      "step": 225
    },
    {
      "epoch": 1.81,
      "learning_rate": 7.989276139410187e-05,
      "loss": 0.6758,
      "step": 226
    },
    {
      "epoch": 1.82,
      "learning_rate": 7.935656836461127e-05,
      "loss": 0.4641,
      "step": 227
    },
    {
      "epoch": 1.82,
      "learning_rate": 7.882037533512064e-05,
      "loss": 0.4743,
      "step": 228
    },
    {
      "epoch": 1.83,
      "learning_rate": 7.828418230563002e-05,
      "loss": 0.4186,
      "step": 229
    },
    {
      "epoch": 1.84,
      "learning_rate": 7.774798927613941e-05,
      "loss": 0.4553,
      "step": 230
    },
    {
      "epoch": 1.85,
      "learning_rate": 7.721179624664879e-05,
      "loss": 0.4656,
      "step": 231
    },
    {
      "epoch": 1.86,
      "learning_rate": 7.667560321715817e-05,
      "loss": 0.4051,
      "step": 232
    },
    {
      "epoch": 1.86,
      "learning_rate": 7.613941018766756e-05,
      "loss": 0.2216,
      "step": 233
    },
    {
      "epoch": 1.87,
      "learning_rate": 7.560321715817694e-05,
      "loss": 0.186,
      "step": 234
    },
    {
      "epoch": 1.88,
      "learning_rate": 7.506702412868633e-05,
      "loss": 0.4749,
      "step": 235
    },
    {
      "epoch": 1.89,
      "learning_rate": 7.453083109919572e-05,
      "loss": 0.4781,
      "step": 236
    },
    {
      "epoch": 1.9,
      "learning_rate": 7.39946380697051e-05,
      "loss": 0.4187,
      "step": 237
    },
    {
      "epoch": 1.9,
      "learning_rate": 7.345844504021448e-05,
      "loss": 0.1483,
      "step": 238
    },
    {
      "epoch": 1.91,
      "learning_rate": 7.292225201072387e-05,
      "loss": 0.1983,
      "step": 239
    },
    {
      "epoch": 1.92,
      "learning_rate": 7.238605898123325e-05,
      "loss": 0.3519,
      "step": 240
    },
    {
      "epoch": 1.93,
      "learning_rate": 7.184986595174264e-05,
      "loss": 0.196,
      "step": 241
    },
    {
      "epoch": 1.94,
      "learning_rate": 7.131367292225202e-05,
      "loss": 0.4758,
      "step": 242
    },
    {
      "epoch": 1.94,
      "learning_rate": 7.07774798927614e-05,
      "loss": 0.5811,
      "step": 243
    },
    {
      "epoch": 1.95,
      "learning_rate": 7.024128686327079e-05,
      "loss": 0.2388,
      "step": 244
    },
    {
      "epoch": 1.96,
      "learning_rate": 6.970509383378017e-05,
      "loss": 0.1879,
      "step": 245
    },
    {
      "epoch": 1.97,
      "learning_rate": 6.916890080428954e-05,
      "loss": 0.4715,
      "step": 246
    },
    {
      "epoch": 1.98,
      "learning_rate": 6.863270777479894e-05,
      "loss": 0.4777,
      "step": 247
    },
    {
      "epoch": 1.98,
      "learning_rate": 6.809651474530831e-05,
      "loss": 0.1942,
      "step": 248
    },
    {
      "epoch": 1.99,
      "learning_rate": 6.756032171581769e-05,
      "loss": 0.3845,
      "step": 249
    },
    {
      "epoch": 2.0,
      "learning_rate": 6.702412868632708e-05,
      "loss": 0.3878,
      "step": 250
    },
    {
      "epoch": 2.01,
      "learning_rate": 6.648793565683646e-05,
      "loss": 0.3244,
      "step": 251
    },
    {
      "epoch": 2.02,
      "learning_rate": 6.595174262734584e-05,
      "loss": 0.1966,
      "step": 252
    },
    {
      "epoch": 2.02,
      "learning_rate": 6.541554959785523e-05,
      "loss": 0.3773,
      "step": 253
    },
    {
      "epoch": 2.03,
      "learning_rate": 6.487935656836461e-05,
      "loss": 0.2275,
      "step": 254
    },
    {
      "epoch": 2.04,
      "learning_rate": 6.4343163538874e-05,
      "loss": 0.3814,
      "step": 255
    },
    {
      "epoch": 2.05,
      "learning_rate": 6.380697050938338e-05,
      "loss": 0.3665,
      "step": 256
    },
    {
      "epoch": 2.06,
      "learning_rate": 6.327077747989276e-05,
      "loss": 0.2171,
      "step": 257
    },
    {
      "epoch": 2.06,
      "learning_rate": 6.273458445040215e-05,
      "loss": 0.3475,
      "step": 258
    },
    {
      "epoch": 2.07,
      "learning_rate": 6.219839142091153e-05,
      "loss": 0.3976,
      "step": 259
    },
    {
      "epoch": 2.08,
      "learning_rate": 6.16621983914209e-05,
      "loss": 0.3182,
      "step": 260
    },
    {
      "epoch": 2.09,
      "learning_rate": 6.11260053619303e-05,
      "loss": 0.1861,
      "step": 261
    },
    {
      "epoch": 2.1,
      "learning_rate": 6.0589812332439676e-05,
      "loss": 0.1793,
      "step": 262
    },
    {
      "epoch": 2.1,
      "learning_rate": 6.005361930294906e-05,
      "loss": 0.2179,
      "step": 263
    },
    {
      "epoch": 2.11,
      "learning_rate": 5.9517426273458446e-05,
      "loss": 0.1979,
      "step": 264
    },
    {
      "epoch": 2.12,
      "learning_rate": 5.8981233243967824e-05,
      "loss": 0.3276,
      "step": 265
    },
    {
      "epoch": 2.13,
      "learning_rate": 5.844504021447721e-05,
      "loss": 0.3906,
      "step": 266
    },
    {
      "epoch": 2.14,
      "learning_rate": 5.79088471849866e-05,
      "loss": 0.2071,
      "step": 267
    },
    {
      "epoch": 2.14,
      "learning_rate": 5.7372654155495985e-05,
      "loss": 0.4533,
      "step": 268
    },
    {
      "epoch": 2.15,
      "learning_rate": 5.683646112600537e-05,
      "loss": 0.3243,
      "step": 269
    },
    {
      "epoch": 2.16,
      "learning_rate": 5.6300268096514755e-05,
      "loss": 0.2265,
      "step": 270
    },
    {
      "epoch": 2.17,
      "learning_rate": 5.5764075067024133e-05,
      "loss": 0.1803,
      "step": 271
    },
    {
      "epoch": 2.18,
      "learning_rate": 5.522788203753352e-05,
      "loss": 0.3864,
      "step": 272
    },
    {
      "epoch": 2.18,
      "learning_rate": 5.46916890080429e-05,
      "loss": 0.2345,
      "step": 273
    },
    {
      "epoch": 2.19,
      "learning_rate": 5.415549597855228e-05,
      "loss": 0.3619,
      "step": 274
    },
    {
      "epoch": 2.2,
      "learning_rate": 5.3619302949061666e-05,
      "loss": 0.359,
      "step": 275
    },
    {
      "epoch": 2.21,
      "learning_rate": 5.308310991957105e-05,
      "loss": 0.3599,
      "step": 276
    },
    {
      "epoch": 2.22,
      "learning_rate": 5.2546916890080436e-05,
      "loss": 0.3484,
      "step": 277
    },
    {
      "epoch": 2.22,
      "learning_rate": 5.2010723860589814e-05,
      "loss": 0.4419,
      "step": 278
    },
    {
      "epoch": 2.23,
      "learning_rate": 5.14745308310992e-05,
      "loss": 0.3972,
      "step": 279
    },
    {
      "epoch": 2.24,
      "learning_rate": 5.0938337801608584e-05,
      "loss": 0.1137,
      "step": 280
    },
    {
      "epoch": 2.25,
      "learning_rate": 5.040214477211796e-05,
      "loss": 0.1447,
      "step": 281
    },
    {
      "epoch": 2.26,
      "learning_rate": 4.986595174262735e-05,
      "loss": 0.1893,
      "step": 282
    },
    {
      "epoch": 2.26,
      "learning_rate": 4.932975871313673e-05,
      "loss": 0.3628,
      "step": 283
    },
    {
      "epoch": 2.27,
      "learning_rate": 4.879356568364612e-05,
      "loss": 0.4312,
      "step": 284
    },
    {
      "epoch": 2.28,
      "learning_rate": 4.8257372654155495e-05,
      "loss": 0.2004,
      "step": 285
    },
    {
      "epoch": 2.29,
      "learning_rate": 4.772117962466488e-05,
      "loss": 0.2917,
      "step": 286
    },
    {
      "epoch": 2.3,
      "learning_rate": 4.7184986595174265e-05,
      "loss": 0.2032,
      "step": 287
    },
    {
      "epoch": 2.3,
      "learning_rate": 4.664879356568364e-05,
      "loss": 0.4294,
      "step": 288
    },
    {
      "epoch": 2.31,
      "learning_rate": 4.611260053619303e-05,
      "loss": 0.3369,
      "step": 289
    },
    {
      "epoch": 2.32,
      "learning_rate": 4.557640750670242e-05,
      "loss": 0.5108,
      "step": 290
    },
    {
      "epoch": 2.33,
      "learning_rate": 4.50402144772118e-05,
      "loss": 0.1781,
      "step": 291
    },
    {
      "epoch": 2.34,
      "learning_rate": 4.450402144772118e-05,
      "loss": 0.3727,
      "step": 292
    },
    {
      "epoch": 2.34,
      "learning_rate": 4.396782841823057e-05,
      "loss": 0.3751,
      "step": 293
    },
    {
      "epoch": 2.35,
      "learning_rate": 4.343163538873995e-05,
      "loss": 0.3441,
      "step": 294
    },
    {
      "epoch": 2.36,
      "learning_rate": 4.289544235924933e-05,
      "loss": 0.1936,
      "step": 295
    },
    {
      "epoch": 2.37,
      "learning_rate": 4.2359249329758715e-05,
      "loss": 0.3061,
      "step": 296
    },
    {
      "epoch": 2.38,
      "learning_rate": 4.18230563002681e-05,
      "loss": 0.1177,
      "step": 297
    },
    {
      "epoch": 2.38,
      "learning_rate": 4.1286863270777485e-05,
      "loss": 0.3463,
      "step": 298
    },
    {
      "epoch": 2.39,
      "learning_rate": 4.075067024128686e-05,
      "loss": 0.2356,
      "step": 299
    },
    {
      "epoch": 2.4,
      "learning_rate": 4.021447721179625e-05,
      "loss": 0.1803,
      "step": 300
    },
    {
      "epoch": 2.41,
      "learning_rate": 3.967828418230563e-05,
      "loss": 0.3799,
      "step": 301
    },
    {
      "epoch": 2.42,
      "learning_rate": 3.914209115281501e-05,
      "loss": 0.1402,
      "step": 302
    },
    {
      "epoch": 2.42,
      "learning_rate": 3.8605898123324396e-05,
      "loss": 0.388,
      "step": 303
    },
    {
      "epoch": 2.43,
      "learning_rate": 3.806970509383378e-05,
      "loss": 0.4519,
      "step": 304
    },
    {
      "epoch": 2.44,
      "learning_rate": 3.7533512064343166e-05,
      "loss": 0.2871,
      "step": 305
    },
    {
      "epoch": 2.45,
      "learning_rate": 3.699731903485255e-05,
      "loss": 0.4283,
      "step": 306
    },
    {
      "epoch": 2.46,
      "learning_rate": 3.6461126005361935e-05,
      "loss": 0.2182,
      "step": 307
    },
    {
      "epoch": 2.46,
      "learning_rate": 3.592493297587132e-05,
      "loss": 0.4494,
      "step": 308
    },
    {
      "epoch": 2.47,
      "learning_rate": 3.53887399463807e-05,
      "loss": 0.1482,
      "step": 309
    },
    {
      "epoch": 2.48,
      "learning_rate": 3.485254691689008e-05,
      "loss": 0.5202,
      "step": 310
    },
    {
      "epoch": 2.49,
      "learning_rate": 3.431635388739947e-05,
      "loss": 0.4222,
      "step": 311
    },
    {
      "epoch": 2.5,
      "learning_rate": 3.3780160857908846e-05,
      "loss": 0.1812,
      "step": 312
    },
    {
      "epoch": 2.5,
      "eval_loss": 0.4782574772834778,
      "eval_runtime": 507.7885,
      "eval_samples_per_second": 1.969,
      "eval_steps_per_second": 0.124,
      "step": 312
    }
  ],
  "logging_steps": 1,
  "max_steps": 375,
  "num_train_epochs": 3,
  "save_steps": 156,
  "total_flos": 2.0372969489301504e+17,
  "trial_name": null,
  "trial_params": null
}