{
  "best_metric": 0.49114030599594116,
  "best_model_checkpoint": "ckpt/origin/pedes_appearance_v2/checkpoint-96",
  "epoch": 5.0,
  "eval_steps": 6,
  "global_step": 120,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04,
      "learning_rate": 5e-06,
      "loss": 0.5465,
      "step": 1
    },
    {
      "epoch": 0.08,
      "learning_rate": 1e-05,
      "loss": 0.6047,
      "step": 2
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.998228055617264e-06,
      "loss": 0.5432,
      "step": 3
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.99291347838381e-06,
      "loss": 0.6246,
      "step": 4
    },
    {
      "epoch": 0.21,
      "learning_rate": 9.984060035153752e-06,
      "loss": 0.5991,
      "step": 5
    },
    {
      "epoch": 0.25,
      "learning_rate": 9.971674001050687e-06,
      "loss": 0.5292,
      "step": 6
    },
    {
      "epoch": 0.25,
      "eval_loss": 0.5087342262268066,
      "eval_runtime": 34.2619,
      "eval_samples_per_second": 1.926,
      "eval_steps_per_second": 1.926,
      "step": 6
    },
    {
      "epoch": 0.29,
      "learning_rate": 9.955764155020037e-06,
      "loss": 0.561,
      "step": 7
    },
    {
      "epoch": 0.33,
      "learning_rate": 9.936341773606723e-06,
      "loss": 0.4865,
      "step": 8
    },
    {
      "epoch": 0.38,
      "learning_rate": 9.913420622962606e-06,
      "loss": 0.4836,
      "step": 9
    },
    {
      "epoch": 0.42,
      "learning_rate": 9.887016949089334e-06,
      "loss": 0.5487,
      "step": 10
    },
    {
      "epoch": 0.46,
      "learning_rate": 9.85714946632355e-06,
      "loss": 0.4804,
      "step": 11
    },
    {
      "epoch": 0.5,
      "learning_rate": 9.823839344072582e-06,
      "loss": 0.5086,
      "step": 12
    },
    {
      "epoch": 0.5,
      "eval_loss": 0.5009010434150696,
      "eval_runtime": 34.2235,
      "eval_samples_per_second": 1.929,
      "eval_steps_per_second": 1.929,
      "step": 12
    },
    {
      "epoch": 0.54,
      "learning_rate": 9.787110191810027e-06,
      "loss": 0.5672,
      "step": 13
    },
    {
      "epoch": 0.58,
      "learning_rate": 9.746988042341907e-06,
      "loss": 0.5379,
      "step": 14
    },
    {
      "epoch": 0.62,
      "learning_rate": 9.703501333355167e-06,
      "loss": 0.5142,
      "step": 15
    },
    {
      "epoch": 0.67,
      "learning_rate": 9.656680887261693e-06,
      "loss": 0.4473,
      "step": 16
    },
    {
      "epoch": 0.71,
      "learning_rate": 9.606559889352065e-06,
      "loss": 0.481,
      "step": 17
    },
    {
      "epoch": 0.75,
      "learning_rate": 9.553173864274567e-06,
      "loss": 0.6152,
      "step": 18
    },
    {
      "epoch": 0.75,
      "eval_loss": 0.4982697367668152,
      "eval_runtime": 34.2148,
      "eval_samples_per_second": 1.929,
      "eval_steps_per_second": 1.929,
      "step": 18
    },
    {
      "epoch": 0.79,
      "learning_rate": 9.496560650856097e-06,
      "loss": 0.4646,
      "step": 19
    },
    {
      "epoch": 0.83,
      "learning_rate": 9.436760375282858e-06,
      "loss": 0.5232,
      "step": 20
    },
    {
      "epoch": 0.88,
      "learning_rate": 9.373815422659806e-06,
      "loss": 0.4288,
      "step": 21
    },
    {
      "epoch": 0.92,
      "learning_rate": 9.307770406969032e-06,
      "loss": 0.5026,
      "step": 22
    },
    {
      "epoch": 0.96,
      "learning_rate": 9.238672139448354e-06,
      "loss": 0.4804,
      "step": 23
    },
    {
      "epoch": 1.0,
      "learning_rate": 9.166569595412576e-06,
      "loss": 0.4183,
      "step": 24
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.4975961148738861,
      "eval_runtime": 34.1526,
      "eval_samples_per_second": 1.933,
      "eval_steps_per_second": 1.933,
      "step": 24
    },
    {
      "epoch": 1.04,
      "learning_rate": 9.091513879540845e-06,
      "loss": 0.4532,
      "step": 25
    },
    {
      "epoch": 1.08,
      "learning_rate": 9.013558189654819e-06,
      "loss": 0.5137,
      "step": 26
    },
    {
      "epoch": 1.12,
      "learning_rate": 8.932757779013214e-06,
      "loss": 0.5173,
      "step": 27
    },
    {
      "epoch": 1.17,
      "learning_rate": 8.849169917149532e-06,
      "loss": 0.5057,
      "step": 28
    },
    {
      "epoch": 1.21,
      "learning_rate": 8.762853849280692e-06,
      "loss": 0.5461,
      "step": 29
    },
    {
      "epoch": 1.25,
      "learning_rate": 8.673870754315336e-06,
      "loss": 0.3704,
      "step": 30
    },
    {
      "epoch": 1.25,
      "eval_loss": 0.4955504536628723,
      "eval_runtime": 34.1868,
      "eval_samples_per_second": 1.931,
      "eval_steps_per_second": 1.931,
      "step": 30
    },
    {
      "epoch": 1.29,
      "learning_rate": 8.582283701491576e-06,
      "loss": 0.4297,
      "step": 31
    },
    {
      "epoch": 1.33,
      "learning_rate": 8.488157605674924e-06,
      "loss": 0.4957,
      "step": 32
    },
    {
      "epoch": 1.38,
      "learning_rate": 8.391559181348081e-06,
      "loss": 0.4564,
      "step": 33
    },
    {
      "epoch": 1.42,
      "learning_rate": 8.292556895325195e-06,
      "loss": 0.5101,
      "step": 34
    },
    {
      "epoch": 1.46,
      "learning_rate": 8.191220918224102e-06,
      "loss": 0.4718,
      "step": 35
    },
    {
      "epoch": 1.5,
      "learning_rate": 8.08762307473096e-06,
      "loss": 0.4089,
      "step": 36
    },
    {
      "epoch": 1.5,
      "eval_loss": 0.49461573362350464,
      "eval_runtime": 34.188,
      "eval_samples_per_second": 1.931,
      "eval_steps_per_second": 1.931,
      "step": 36
    },
    {
      "epoch": 1.54,
      "learning_rate": 7.981836792692508e-06,
      "loss": 0.4593,
      "step": 37
    },
    {
      "epoch": 1.58,
      "learning_rate": 7.873937051072037e-06,
      "loss": 0.477,
      "step": 38
    },
    {
      "epoch": 1.62,
      "learning_rate": 7.764000326805967e-06,
      "loss": 0.4715,
      "step": 39
    },
    {
      "epoch": 1.67,
      "learning_rate": 7.652104540598712e-06,
      "loss": 0.4718,
      "step": 40
    },
    {
      "epoch": 1.71,
      "learning_rate": 7.5383290016942e-06,
      "loss": 0.515,
      "step": 41
    },
    {
      "epoch": 1.75,
      "learning_rate": 7.422754351663252e-06,
      "loss": 0.5042,
      "step": 42
    },
    {
      "epoch": 1.75,
      "eval_loss": 0.4928542673587799,
      "eval_runtime": 34.2625,
      "eval_samples_per_second": 1.926,
      "eval_steps_per_second": 1.926,
      "step": 42
    },
    {
      "epoch": 1.79,
      "learning_rate": 7.30546250724663e-06,
      "loss": 0.49,
      "step": 43
    },
    {
      "epoch": 1.83,
      "learning_rate": 7.186536602294278e-06,
      "loss": 0.4668,
      "step": 44
    },
    {
      "epoch": 1.88,
      "learning_rate": 7.066060928841891e-06,
      "loss": 0.5409,
      "step": 45
    },
    {
      "epoch": 1.92,
      "learning_rate": 6.944120877366605e-06,
      "loss": 0.4741,
      "step": 46
    },
    {
      "epoch": 1.96,
      "learning_rate": 6.820802876264112e-06,
      "loss": 0.4585,
      "step": 47
    },
    {
      "epoch": 2.0,
      "learning_rate": 6.6961943305901515e-06,
      "loss": 0.5924,
      "step": 48
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.4919312298297882,
      "eval_runtime": 34.1613,
      "eval_samples_per_second": 1.932,
      "eval_steps_per_second": 1.932,
      "step": 48
    },
    {
      "epoch": 2.04,
      "learning_rate": 6.570383560109745e-06,
      "loss": 0.4469,
      "step": 49
    },
    {
      "epoch": 2.08,
      "learning_rate": 6.443459736698106e-06,
      "loss": 0.4503,
      "step": 50
    },
    {
      "epoch": 2.12,
      "learning_rate": 6.315512821137606e-06,
      "loss": 0.4587,
      "step": 51
    },
    {
      "epoch": 2.17,
      "learning_rate": 6.186633499355576e-06,
      "loss": 0.4068,
      "step": 52
    },
    {
      "epoch": 2.21,
      "learning_rate": 6.056913118148122e-06,
      "loss": 0.4505,
      "step": 53
    },
    {
      "epoch": 2.25,
      "learning_rate": 5.926443620435572e-06,
      "loss": 0.5053,
      "step": 54
    },
    {
      "epoch": 2.25,
      "eval_loss": 0.49188804626464844,
      "eval_runtime": 34.2448,
      "eval_samples_per_second": 1.927,
      "eval_steps_per_second": 1.927,
      "step": 54
    },
    {
      "epoch": 2.29,
      "learning_rate": 5.795317480095361e-06,
      "loss": 0.4201,
      "step": 55
    },
    {
      "epoch": 2.33,
      "learning_rate": 5.663627636418611e-06,
      "loss": 0.5029,
      "step": 56
    },
    {
      "epoch": 2.38,
      "learning_rate": 5.531467428236827e-06,
      "loss": 0.4954,
      "step": 57
    },
    {
      "epoch": 2.42,
      "learning_rate": 5.398930527765416e-06,
      "loss": 0.5362,
      "step": 58
    },
    {
      "epoch": 2.46,
      "learning_rate": 5.266110874210893e-06,
      "loss": 0.4561,
      "step": 59
    },
    {
      "epoch": 2.5,
      "learning_rate": 5.133102607188875e-06,
      "loss": 0.4752,
      "step": 60
    },
    {
      "epoch": 2.5,
      "eval_loss": 0.4920966327190399,
      "eval_runtime": 34.2375,
      "eval_samples_per_second": 1.928,
      "eval_steps_per_second": 1.928,
      "step": 60
    },
    {
      "epoch": 2.54,
      "learning_rate": 5e-06,
      "loss": 0.3937,
      "step": 61
    },
    {
      "epoch": 2.58,
      "learning_rate": 4.866897392811127e-06,
      "loss": 0.5296,
      "step": 62
    },
    {
      "epoch": 2.62,
      "learning_rate": 4.7338891257891085e-06,
      "loss": 0.4653,
      "step": 63
    },
    {
      "epoch": 2.67,
      "learning_rate": 4.601069472234584e-06,
      "loss": 0.4839,
      "step": 64
    },
    {
      "epoch": 2.71,
      "learning_rate": 4.468532571763174e-06,
      "loss": 0.4785,
      "step": 65
    },
    {
      "epoch": 2.75,
      "learning_rate": 4.336372363581391e-06,
      "loss": 0.4627,
      "step": 66
    },
    {
      "epoch": 2.75,
      "eval_loss": 0.4910169541835785,
      "eval_runtime": 34.25,
      "eval_samples_per_second": 1.927,
      "eval_steps_per_second": 1.927,
      "step": 66
    },
    {
      "epoch": 2.79,
      "learning_rate": 4.204682519904641e-06,
      "loss": 0.4955,
      "step": 67
    },
    {
      "epoch": 2.83,
      "learning_rate": 4.073556379564429e-06,
      "loss": 0.4617,
      "step": 68
    },
    {
      "epoch": 2.88,
      "learning_rate": 3.9430868818518786e-06,
      "loss": 0.5626,
      "step": 69
    },
    {
      "epoch": 2.92,
      "learning_rate": 3.813366500644426e-06,
      "loss": 0.4044,
      "step": 70
    },
    {
      "epoch": 2.96,
      "learning_rate": 3.6844871788623946e-06,
      "loss": 0.5331,
      "step": 71
    },
    {
      "epoch": 3.0,
      "learning_rate": 3.5565402633018963e-06,
      "loss": 0.4079,
      "step": 72
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.49220502376556396,
      "eval_runtime": 34.2442,
      "eval_samples_per_second": 1.927,
      "eval_steps_per_second": 1.927,
      "step": 72
    },
    {
      "epoch": 3.04,
      "learning_rate": 3.4296164398902576e-06,
      "loss": 0.4431,
      "step": 73
    },
    {
      "epoch": 3.08,
      "learning_rate": 3.3038056694098485e-06,
      "loss": 0.4309,
      "step": 74
    },
    {
      "epoch": 3.12,
      "learning_rate": 3.1791971237358893e-06,
      "loss": 0.52,
      "step": 75
    },
    {
      "epoch": 3.17,
      "learning_rate": 3.0558791226333974e-06,
      "loss": 0.5226,
      "step": 76
    },
    {
      "epoch": 3.21,
      "learning_rate": 2.9339390711581105e-06,
      "loss": 0.4657,
      "step": 77
    },
    {
      "epoch": 3.25,
      "learning_rate": 2.8134633977057236e-06,
      "loss": 0.4638,
      "step": 78
    },
    {
      "epoch": 3.25,
      "eval_loss": 0.49126145243644714,
      "eval_runtime": 34.2128,
      "eval_samples_per_second": 1.929,
      "eval_steps_per_second": 1.929,
      "step": 78
    },
    {
      "epoch": 3.29,
      "learning_rate": 2.69453749275337e-06,
      "loss": 0.4113,
      "step": 79
    },
    {
      "epoch": 3.33,
      "learning_rate": 2.57724564833675e-06,
      "loss": 0.4887,
      "step": 80
    },
    {
      "epoch": 3.38,
      "learning_rate": 2.461670998305802e-06,
      "loss": 0.5027,
      "step": 81
    },
    {
      "epoch": 3.42,
      "learning_rate": 2.3478954594012884e-06,
      "loss": 0.4334,
      "step": 82
    },
    {
      "epoch": 3.46,
      "learning_rate": 2.2359996731940348e-06,
      "loss": 0.471,
      "step": 83
    },
    {
      "epoch": 3.5,
      "learning_rate": 2.1260629489279662e-06,
      "loss": 0.466,
      "step": 84
    },
    {
      "epoch": 3.5,
      "eval_loss": 0.4917301535606384,
      "eval_runtime": 34.2178,
      "eval_samples_per_second": 1.929,
      "eval_steps_per_second": 1.929,
      "step": 84
    },
    {
      "epoch": 3.54,
      "learning_rate": 2.0181632073074925e-06,
      "loss": 0.4021,
      "step": 85
    },
    {
      "epoch": 3.58,
      "learning_rate": 1.912376925269041e-06,
      "loss": 0.5112,
      "step": 86
    },
    {
      "epoch": 3.62,
      "learning_rate": 1.808779081775901e-06,
      "loss": 0.4545,
      "step": 87
    },
    {
      "epoch": 3.67,
      "learning_rate": 1.7074431046748075e-06,
      "loss": 0.4638,
      "step": 88
    },
    {
      "epoch": 3.71,
      "learning_rate": 1.6084408186519195e-06,
      "loss": 0.4908,
      "step": 89
    },
    {
      "epoch": 3.75,
      "learning_rate": 1.511842394325077e-06,
      "loss": 0.4834,
      "step": 90
    },
    {
      "epoch": 3.75,
      "eval_loss": 0.491076797246933,
      "eval_runtime": 34.1746,
      "eval_samples_per_second": 1.931,
      "eval_steps_per_second": 1.931,
      "step": 90
    },
    {
      "epoch": 3.79,
      "learning_rate": 1.4177162985084242e-06,
      "loss": 0.4876,
      "step": 91
    },
    {
      "epoch": 3.83,
      "learning_rate": 1.3261292456846648e-06,
      "loss": 0.4917,
      "step": 92
    },
    {
      "epoch": 3.88,
      "learning_rate": 1.2371461507193077e-06,
      "loss": 0.4748,
      "step": 93
    },
    {
      "epoch": 3.92,
      "learning_rate": 1.1508300828504682e-06,
      "loss": 0.4598,
      "step": 94
    },
    {
      "epoch": 3.96,
      "learning_rate": 1.0672422209867879e-06,
      "loss": 0.4163,
      "step": 95
    },
    {
      "epoch": 4.0,
      "learning_rate": 9.86441810345183e-07,
      "loss": 0.4581,
      "step": 96
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.49114030599594116,
      "eval_runtime": 34.1668,
      "eval_samples_per_second": 1.932,
      "eval_steps_per_second": 1.932,
      "step": 96
    },
    {
      "epoch": 4.04,
      "learning_rate": 9.08486120459155e-07,
      "loss": 0.4276,
      "step": 97
    },
    {
      "epoch": 4.08,
      "learning_rate": 8.334304045874248e-07,
      "loss": 0.4776,
      "step": 98
    },
    {
      "epoch": 4.12,
      "learning_rate": 7.613278605516455e-07,
      "loss": 0.503,
      "step": 99
    },
    {
      "epoch": 4.17,
      "learning_rate": 6.922295930309691e-07,
      "loss": 0.5257,
      "step": 100
    },
    {
      "epoch": 4.21,
      "learning_rate": 6.261845773401936e-07,
      "loss": 0.5144,
      "step": 101
    },
    {
      "epoch": 4.25,
      "learning_rate": 5.632396247171429e-07,
      "loss": 0.5178,
      "step": 102
    },
    {
      "epoch": 4.25,
      "eval_loss": 0.4920191764831543,
      "eval_runtime": 34.1838,
      "eval_samples_per_second": 1.931,
      "eval_steps_per_second": 1.931,
      "step": 102
    },
    {
      "epoch": 4.29,
      "learning_rate": 5.034393491439044e-07,
      "loss": 0.4231,
      "step": 103
    },
    {
      "epoch": 4.33,
      "learning_rate": 4.468261357254339e-07,
      "loss": 0.4624,
      "step": 104
    },
    {
      "epoch": 4.38,
      "learning_rate": 3.934401106479352e-07,
      "loss": 0.4505,
      "step": 105
    },
    {
      "epoch": 4.42,
      "learning_rate": 3.433191127383079e-07,
      "loss": 0.4092,
      "step": 106
    },
    {
      "epoch": 4.46,
      "learning_rate": 2.9649866664483387e-07,
      "loss": 0.5014,
      "step": 107
    },
    {
      "epoch": 4.5,
      "learning_rate": 2.530119576580936e-07,
      "loss": 0.4354,
      "step": 108
    },
    {
      "epoch": 4.5,
      "eval_loss": 0.4916991591453552,
      "eval_runtime": 34.1588,
      "eval_samples_per_second": 1.932,
      "eval_steps_per_second": 1.932,
      "step": 108
    },
    {
      "epoch": 4.54,
      "learning_rate": 2.1288980818997272e-07,
      "loss": 0.5014,
      "step": 109
    },
    {
      "epoch": 4.58,
      "learning_rate": 1.7616065592742038e-07,
      "loss": 0.4265,
      "step": 110
    },
    {
      "epoch": 4.62,
      "learning_rate": 1.4285053367645074e-07,
      "loss": 0.4602,
      "step": 111
    },
    {
      "epoch": 4.67,
      "learning_rate": 1.1298305091066664e-07,
      "loss": 0.4799,
      "step": 112
    },
    {
      "epoch": 4.71,
      "learning_rate": 8.657937703739516e-08,
      "loss": 0.5299,
      "step": 113
    },
    {
      "epoch": 4.75,
      "learning_rate": 6.365822639327724e-08,
      "loss": 0.4875,
      "step": 114
    },
    {
      "epoch": 4.75,
      "eval_loss": 0.49080830812454224,
      "eval_runtime": 34.1733,
      "eval_samples_per_second": 1.931,
      "eval_steps_per_second": 1.931,
      "step": 114
    },
    {
      "epoch": 4.79,
      "learning_rate": 4.423584497996458e-08,
      "loss": 0.4093,
      "step": 115
    },
    {
      "epoch": 4.83,
      "learning_rate": 2.8325998949314536e-08,
      "loss": 0.4361,
      "step": 116
    },
    {
      "epoch": 4.88,
      "learning_rate": 1.593996484624938e-08,
      "loss": 0.4469,
      "step": 117
    },
    {
      "epoch": 4.92,
      "learning_rate": 7.0865216161902785e-09,
      "loss": 0.4857,
      "step": 118
    },
    {
      "epoch": 4.96,
      "learning_rate": 1.7719443827368677e-09,
      "loss": 0.4774,
      "step": 119
    },
    {
      "epoch": 5.0,
      "learning_rate": 0.0,
      "loss": 0.4127,
      "step": 120
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.4913090765476227,
      "eval_runtime": 34.1855,
      "eval_samples_per_second": 1.931,
      "eval_steps_per_second": 1.931,
      "step": 120
    },
    {
      "epoch": 5.0,
      "step": 120,
      "total_flos": 3.2229528504631296e+17,
      "train_loss": 0.48162667627135913,
      "train_runtime": 7150.9925,
      "train_samples_per_second": 0.802,
      "train_steps_per_second": 0.017
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 120,
  "num_train_epochs": 5,
  "save_steps": 12,
  "total_flos": 3.2229528504631296e+17,
  "trial_name": null,
  "trial_params": null
}