sam1120's picture
Training in progress, step 240
74483d1
raw
history blame
43.8 kB
{
"best_metric": 0.5627052187919617,
"best_model_checkpoint": "/robodata/smodak/Projects/nspl/scripts/terrainseg/training/models/safety-utcustom-train-SF-RGB-b5/checkpoint-240",
"epoch": 21.818181818181817,
"global_step": 240,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.09,
"learning_rate": 1.764705882352941e-07,
"loss": 1.2293,
"step": 1
},
{
"epoch": 0.18,
"learning_rate": 3.529411764705882e-07,
"loss": 1.2349,
"step": 2
},
{
"epoch": 0.27,
"learning_rate": 5.294117647058824e-07,
"loss": 1.2357,
"step": 3
},
{
"epoch": 0.36,
"learning_rate": 7.058823529411765e-07,
"loss": 1.2242,
"step": 4
},
{
"epoch": 0.45,
"learning_rate": 8.823529411764706e-07,
"loss": 1.2247,
"step": 5
},
{
"epoch": 0.55,
"learning_rate": 1.0588235294117648e-06,
"loss": 1.227,
"step": 6
},
{
"epoch": 0.64,
"learning_rate": 1.2352941176470588e-06,
"loss": 1.2292,
"step": 7
},
{
"epoch": 0.73,
"learning_rate": 1.411764705882353e-06,
"loss": 1.2257,
"step": 8
},
{
"epoch": 0.82,
"learning_rate": 1.5882352941176472e-06,
"loss": 1.2173,
"step": 9
},
{
"epoch": 0.91,
"learning_rate": 1.7647058823529412e-06,
"loss": 1.2239,
"step": 10
},
{
"epoch": 0.91,
"eval_accuracy_safe": 0.39920432072351303,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.29510386893492613,
"eval_iou_safe": 0.031418797454432365,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.29392505413655395,
"eval_loss": 1.1102662086486816,
"eval_mean_accuracy": 0.34715409482921955,
"eval_mean_iou": 0.10844795053032878,
"eval_overall_accuracy": 0.2981788293639226,
"eval_runtime": 9.3061,
"eval_samples_per_second": 7.2,
"eval_steps_per_second": 0.537,
"step": 10
},
{
"epoch": 1.0,
"learning_rate": 1.9411764705882353e-06,
"loss": 1.221,
"step": 11
},
{
"epoch": 1.09,
"learning_rate": 2.1176470588235296e-06,
"loss": 1.2149,
"step": 12
},
{
"epoch": 1.18,
"learning_rate": 2.2941176470588234e-06,
"loss": 1.2197,
"step": 13
},
{
"epoch": 1.27,
"learning_rate": 2.4705882352941177e-06,
"loss": 1.2064,
"step": 14
},
{
"epoch": 1.36,
"learning_rate": 2.647058823529412e-06,
"loss": 1.2042,
"step": 15
},
{
"epoch": 1.45,
"learning_rate": 2.823529411764706e-06,
"loss": 1.2046,
"step": 16
},
{
"epoch": 1.55,
"learning_rate": 3e-06,
"loss": 1.2054,
"step": 17
},
{
"epoch": 1.64,
"learning_rate": 2.990415335463259e-06,
"loss": 1.2002,
"step": 18
},
{
"epoch": 1.73,
"learning_rate": 2.9808306709265177e-06,
"loss": 1.1982,
"step": 19
},
{
"epoch": 1.82,
"learning_rate": 2.9712460063897764e-06,
"loss": 1.1948,
"step": 20
},
{
"epoch": 1.82,
"eval_accuracy_safe": 0.5218599774094934,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.3705393407485172,
"eval_iou_safe": 0.0439841259320199,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.36893471456827454,
"eval_loss": 1.0963102579116821,
"eval_mean_accuracy": 0.4461996590790053,
"eval_mean_iou": 0.13763961350009815,
"eval_overall_accuracy": 0.37500910972481344,
"eval_runtime": 8.3942,
"eval_samples_per_second": 7.982,
"eval_steps_per_second": 0.596,
"step": 20
},
{
"epoch": 1.91,
"learning_rate": 2.9616613418530352e-06,
"loss": 1.1776,
"step": 21
},
{
"epoch": 2.0,
"learning_rate": 2.952076677316294e-06,
"loss": 1.1936,
"step": 22
},
{
"epoch": 2.09,
"learning_rate": 2.942492012779553e-06,
"loss": 1.1664,
"step": 23
},
{
"epoch": 2.18,
"learning_rate": 2.9329073482428116e-06,
"loss": 1.1809,
"step": 24
},
{
"epoch": 2.27,
"learning_rate": 2.9233226837060704e-06,
"loss": 1.1731,
"step": 25
},
{
"epoch": 2.36,
"learning_rate": 2.913738019169329e-06,
"loss": 1.1794,
"step": 26
},
{
"epoch": 2.45,
"learning_rate": 2.904153354632588e-06,
"loss": 1.1684,
"step": 27
},
{
"epoch": 2.55,
"learning_rate": 2.8945686900958468e-06,
"loss": 1.1594,
"step": 28
},
{
"epoch": 2.64,
"learning_rate": 2.884984025559105e-06,
"loss": 1.146,
"step": 29
},
{
"epoch": 2.73,
"learning_rate": 2.8753993610223644e-06,
"loss": 1.1661,
"step": 30
},
{
"epoch": 2.73,
"eval_accuracy_safe": 0.5863200218966003,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.49882181393718666,
"eval_iou_safe": 0.06472096392925002,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.49613792944888824,
"eval_loss": 1.0516184568405151,
"eval_mean_accuracy": 0.5425709179168935,
"eval_mean_iou": 0.1869529644593794,
"eval_overall_accuracy": 0.5014063707038537,
"eval_runtime": 9.3615,
"eval_samples_per_second": 7.157,
"eval_steps_per_second": 0.534,
"step": 30
},
{
"epoch": 2.82,
"learning_rate": 2.865814696485623e-06,
"loss": 1.153,
"step": 31
},
{
"epoch": 2.91,
"learning_rate": 2.856230031948882e-06,
"loss": 1.149,
"step": 32
},
{
"epoch": 3.0,
"learning_rate": 2.8466453674121407e-06,
"loss": 1.164,
"step": 33
},
{
"epoch": 3.09,
"learning_rate": 2.8370607028753995e-06,
"loss": 1.1479,
"step": 34
},
{
"epoch": 3.18,
"learning_rate": 2.8274760383386583e-06,
"loss": 1.1428,
"step": 35
},
{
"epoch": 3.27,
"learning_rate": 2.817891373801917e-06,
"loss": 1.1398,
"step": 36
},
{
"epoch": 3.36,
"learning_rate": 2.8083067092651755e-06,
"loss": 1.1415,
"step": 37
},
{
"epoch": 3.45,
"learning_rate": 2.7987220447284347e-06,
"loss": 1.1146,
"step": 38
},
{
"epoch": 3.55,
"learning_rate": 2.7891373801916935e-06,
"loss": 1.1234,
"step": 39
},
{
"epoch": 3.64,
"learning_rate": 2.7795527156549523e-06,
"loss": 1.1112,
"step": 40
},
{
"epoch": 3.64,
"eval_accuracy_safe": 0.5458517893146133,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.5794390867479824,
"eval_iou_safe": 0.0899865428032596,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.5753581085678546,
"eval_loss": 1.0047743320465088,
"eval_mean_accuracy": 0.5626454380312979,
"eval_mean_iou": 0.22178155045703807,
"eval_overall_accuracy": 0.5784469718363747,
"eval_runtime": 9.38,
"eval_samples_per_second": 7.143,
"eval_steps_per_second": 0.533,
"step": 40
},
{
"epoch": 3.73,
"learning_rate": 2.769968051118211e-06,
"loss": 1.1212,
"step": 41
},
{
"epoch": 3.82,
"learning_rate": 2.7603833865814694e-06,
"loss": 1.1117,
"step": 42
},
{
"epoch": 3.91,
"learning_rate": 2.7507987220447286e-06,
"loss": 1.1051,
"step": 43
},
{
"epoch": 4.0,
"learning_rate": 2.7412140575079874e-06,
"loss": 1.1382,
"step": 44
},
{
"epoch": 4.09,
"learning_rate": 2.731629392971246e-06,
"loss": 1.1075,
"step": 45
},
{
"epoch": 4.18,
"learning_rate": 2.722044728434505e-06,
"loss": 1.1096,
"step": 46
},
{
"epoch": 4.27,
"learning_rate": 2.7124600638977634e-06,
"loss": 1.0974,
"step": 47
},
{
"epoch": 4.36,
"learning_rate": 2.7028753993610226e-06,
"loss": 1.0851,
"step": 48
},
{
"epoch": 4.45,
"learning_rate": 2.6932907348242814e-06,
"loss": 1.0851,
"step": 49
},
{
"epoch": 4.55,
"learning_rate": 2.6837060702875398e-06,
"loss": 1.0907,
"step": 50
},
{
"epoch": 4.55,
"eval_accuracy_safe": 0.5993018531154467,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.6366664738420048,
"eval_iou_safe": 0.10939424133801141,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.6320960572783577,
"eval_loss": 0.9689871668815613,
"eval_mean_accuracy": 0.6179841634787258,
"eval_mean_iou": 0.24716343287212306,
"eval_overall_accuracy": 0.6355627828569554,
"eval_runtime": 9.8119,
"eval_samples_per_second": 6.828,
"eval_steps_per_second": 0.51,
"step": 50
},
{
"epoch": 4.64,
"learning_rate": 2.674121405750799e-06,
"loss": 1.0789,
"step": 51
},
{
"epoch": 4.73,
"learning_rate": 2.6645367412140573e-06,
"loss": 1.0566,
"step": 52
},
{
"epoch": 4.82,
"learning_rate": 2.654952076677316e-06,
"loss": 1.0718,
"step": 53
},
{
"epoch": 4.91,
"learning_rate": 2.6453674121405753e-06,
"loss": 1.0808,
"step": 54
},
{
"epoch": 5.0,
"learning_rate": 2.6357827476038337e-06,
"loss": 1.0808,
"step": 55
},
{
"epoch": 5.09,
"learning_rate": 2.626198083067093e-06,
"loss": 1.0625,
"step": 56
},
{
"epoch": 5.18,
"learning_rate": 2.6166134185303517e-06,
"loss": 1.0508,
"step": 57
},
{
"epoch": 5.27,
"learning_rate": 2.60702875399361e-06,
"loss": 1.0521,
"step": 58
},
{
"epoch": 5.36,
"learning_rate": 2.5974440894568693e-06,
"loss": 1.052,
"step": 59
},
{
"epoch": 5.45,
"learning_rate": 2.5878594249201277e-06,
"loss": 1.047,
"step": 60
},
{
"epoch": 5.45,
"eval_accuracy_safe": 0.6691570194409427,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.669918812994849,
"eval_iou_safe": 0.11592410370525706,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.6655783206367077,
"eval_loss": 0.9437160491943359,
"eval_mean_accuracy": 0.6695379162178958,
"eval_mean_iou": 0.2605008081139883,
"eval_overall_accuracy": 0.6698963108347423,
"eval_runtime": 9.3506,
"eval_samples_per_second": 7.165,
"eval_steps_per_second": 0.535,
"step": 60
},
{
"epoch": 5.55,
"learning_rate": 2.5782747603833865e-06,
"loss": 1.0456,
"step": 61
},
{
"epoch": 5.64,
"learning_rate": 2.5686900958466457e-06,
"loss": 1.0296,
"step": 62
},
{
"epoch": 5.73,
"learning_rate": 2.559105431309904e-06,
"loss": 1.0334,
"step": 63
},
{
"epoch": 5.82,
"learning_rate": 2.5495207667731633e-06,
"loss": 1.0525,
"step": 64
},
{
"epoch": 5.91,
"learning_rate": 2.5399361022364216e-06,
"loss": 1.0397,
"step": 65
},
{
"epoch": 6.0,
"learning_rate": 2.5303514376996804e-06,
"loss": 1.0575,
"step": 66
},
{
"epoch": 6.09,
"learning_rate": 2.5207667731629396e-06,
"loss": 1.0245,
"step": 67
},
{
"epoch": 6.18,
"learning_rate": 2.511182108626198e-06,
"loss": 1.0284,
"step": 68
},
{
"epoch": 6.27,
"learning_rate": 2.501597444089457e-06,
"loss": 1.0127,
"step": 69
},
{
"epoch": 6.36,
"learning_rate": 2.4920127795527156e-06,
"loss": 1.0112,
"step": 70
},
{
"epoch": 6.36,
"eval_accuracy_safe": 0.6673046750012529,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.718865515124044,
"eval_iou_safe": 0.13494764234639625,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.7136723036690169,
"eval_loss": 0.9084014892578125,
"eval_mean_accuracy": 0.6930850950626484,
"eval_mean_iou": 0.28287331533847104,
"eval_overall_accuracy": 0.7173424905805446,
"eval_runtime": 9.0845,
"eval_samples_per_second": 7.375,
"eval_steps_per_second": 0.55,
"step": 70
},
{
"epoch": 6.45,
"learning_rate": 2.4824281150159744e-06,
"loss": 1.0107,
"step": 71
},
{
"epoch": 6.55,
"learning_rate": 2.4728434504792336e-06,
"loss": 1.0122,
"step": 72
},
{
"epoch": 6.64,
"learning_rate": 2.463258785942492e-06,
"loss": 1.0087,
"step": 73
},
{
"epoch": 6.73,
"learning_rate": 2.4536741214057507e-06,
"loss": 0.991,
"step": 74
},
{
"epoch": 6.82,
"learning_rate": 2.44408945686901e-06,
"loss": 1.0042,
"step": 75
},
{
"epoch": 6.91,
"learning_rate": 2.4345047923322683e-06,
"loss": 0.9788,
"step": 76
},
{
"epoch": 7.0,
"learning_rate": 2.424920127795527e-06,
"loss": 1.0371,
"step": 77
},
{
"epoch": 7.09,
"learning_rate": 2.415335463258786e-06,
"loss": 0.9884,
"step": 78
},
{
"epoch": 7.18,
"learning_rate": 2.4057507987220447e-06,
"loss": 0.9759,
"step": 79
},
{
"epoch": 7.27,
"learning_rate": 2.396166134185304e-06,
"loss": 0.9925,
"step": 80
},
{
"epoch": 7.27,
"eval_accuracy_safe": 0.6842282797676185,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.7664939888573942,
"eval_iou_safe": 0.14521835177647324,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.7605469001045981,
"eval_loss": 0.8647122979164124,
"eval_mean_accuracy": 0.7253611343125064,
"eval_mean_iou": 0.3019217506270238,
"eval_overall_accuracy": 0.7640639917174382,
"eval_runtime": 9.9787,
"eval_samples_per_second": 6.714,
"eval_steps_per_second": 0.501,
"step": 80
},
{
"epoch": 7.36,
"learning_rate": 2.3865814696485623e-06,
"loss": 0.9784,
"step": 81
},
{
"epoch": 7.45,
"learning_rate": 2.376996805111821e-06,
"loss": 0.9806,
"step": 82
},
{
"epoch": 7.55,
"learning_rate": 2.36741214057508e-06,
"loss": 0.973,
"step": 83
},
{
"epoch": 7.64,
"learning_rate": 2.3578274760383387e-06,
"loss": 0.9551,
"step": 84
},
{
"epoch": 7.73,
"learning_rate": 2.3482428115015974e-06,
"loss": 0.9484,
"step": 85
},
{
"epoch": 7.82,
"learning_rate": 2.3386581469648562e-06,
"loss": 0.939,
"step": 86
},
{
"epoch": 7.91,
"learning_rate": 2.329073482428115e-06,
"loss": 0.9533,
"step": 87
},
{
"epoch": 8.0,
"learning_rate": 2.319488817891374e-06,
"loss": 0.9533,
"step": 88
},
{
"epoch": 8.09,
"learning_rate": 2.3099041533546326e-06,
"loss": 0.9469,
"step": 89
},
{
"epoch": 8.18,
"learning_rate": 2.3003194888178914e-06,
"loss": 0.9395,
"step": 90
},
{
"epoch": 8.18,
"eval_accuracy_safe": 0.6817822598987667,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.7920860651953089,
"eval_iou_safe": 0.16203325137990066,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.785611270080723,
"eval_loss": 0.8319001197814941,
"eval_mean_accuracy": 0.7369341625470378,
"eval_mean_iou": 0.3158815071535412,
"eval_overall_accuracy": 0.788827867650274,
"eval_runtime": 9.682,
"eval_samples_per_second": 6.92,
"eval_steps_per_second": 0.516,
"step": 90
},
{
"epoch": 8.27,
"learning_rate": 2.29073482428115e-06,
"loss": 0.9422,
"step": 91
},
{
"epoch": 8.36,
"learning_rate": 2.281150159744409e-06,
"loss": 0.9012,
"step": 92
},
{
"epoch": 8.45,
"learning_rate": 2.2715654952076678e-06,
"loss": 0.937,
"step": 93
},
{
"epoch": 8.55,
"learning_rate": 2.2619808306709266e-06,
"loss": 0.9122,
"step": 94
},
{
"epoch": 8.64,
"learning_rate": 2.2523961661341854e-06,
"loss": 0.9272,
"step": 95
},
{
"epoch": 8.73,
"learning_rate": 2.242811501597444e-06,
"loss": 0.9122,
"step": 96
},
{
"epoch": 8.82,
"learning_rate": 2.233226837060703e-06,
"loss": 0.9261,
"step": 97
},
{
"epoch": 8.91,
"learning_rate": 2.2236421725239617e-06,
"loss": 0.9004,
"step": 98
},
{
"epoch": 9.0,
"learning_rate": 2.2140575079872205e-06,
"loss": 0.8981,
"step": 99
},
{
"epoch": 9.09,
"learning_rate": 2.2044728434504793e-06,
"loss": 0.8902,
"step": 100
},
{
"epoch": 9.09,
"eval_accuracy_safe": 0.6805698513112902,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.8141590718977455,
"eval_iou_safe": 0.1769983948474704,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.8072069422555583,
"eval_loss": 0.8013909459114075,
"eval_mean_accuracy": 0.7473644616045179,
"eval_mean_iou": 0.32806844570100957,
"eval_overall_accuracy": 0.8102130605213678,
"eval_runtime": 9.6492,
"eval_samples_per_second": 6.944,
"eval_steps_per_second": 0.518,
"step": 100
},
{
"epoch": 9.18,
"learning_rate": 2.194888178913738e-06,
"loss": 0.9085,
"step": 101
},
{
"epoch": 9.27,
"learning_rate": 2.185303514376997e-06,
"loss": 0.8976,
"step": 102
},
{
"epoch": 9.36,
"learning_rate": 2.1757188498402557e-06,
"loss": 0.8909,
"step": 103
},
{
"epoch": 9.45,
"learning_rate": 2.1661341853035145e-06,
"loss": 0.8878,
"step": 104
},
{
"epoch": 9.55,
"learning_rate": 2.1565495207667733e-06,
"loss": 0.8909,
"step": 105
},
{
"epoch": 9.64,
"learning_rate": 2.146964856230032e-06,
"loss": 0.8529,
"step": 106
},
{
"epoch": 9.73,
"learning_rate": 2.137380191693291e-06,
"loss": 0.9097,
"step": 107
},
{
"epoch": 9.82,
"learning_rate": 2.1277955271565496e-06,
"loss": 0.8778,
"step": 108
},
{
"epoch": 9.91,
"learning_rate": 2.1182108626198084e-06,
"loss": 0.8819,
"step": 109
},
{
"epoch": 10.0,
"learning_rate": 2.1086261980830672e-06,
"loss": 0.9057,
"step": 110
},
{
"epoch": 10.0,
"eval_accuracy_safe": 0.6984244470915687,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.8178638868312451,
"eval_iou_safe": 0.17330681673606707,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.8109252159162895,
"eval_loss": 0.7867299914360046,
"eval_mean_accuracy": 0.7581441669614069,
"eval_mean_iou": 0.3280773442174522,
"eval_overall_accuracy": 0.814335837293027,
"eval_runtime": 9.8362,
"eval_samples_per_second": 6.812,
"eval_steps_per_second": 0.508,
"step": 110
},
{
"epoch": 10.09,
"learning_rate": 2.099041533546326e-06,
"loss": 0.8509,
"step": 111
},
{
"epoch": 10.18,
"learning_rate": 2.089456869009585e-06,
"loss": 0.8746,
"step": 112
},
{
"epoch": 10.27,
"learning_rate": 2.0798722044728436e-06,
"loss": 0.8532,
"step": 113
},
{
"epoch": 10.36,
"learning_rate": 2.0702875399361024e-06,
"loss": 0.8509,
"step": 114
},
{
"epoch": 10.45,
"learning_rate": 2.060702875399361e-06,
"loss": 0.8559,
"step": 115
},
{
"epoch": 10.55,
"learning_rate": 2.05111821086262e-06,
"loss": 0.8435,
"step": 116
},
{
"epoch": 10.64,
"learning_rate": 2.0415335463258783e-06,
"loss": 0.8465,
"step": 117
},
{
"epoch": 10.73,
"learning_rate": 2.0319488817891376e-06,
"loss": 0.8345,
"step": 118
},
{
"epoch": 10.82,
"learning_rate": 2.0223642172523963e-06,
"loss": 0.8403,
"step": 119
},
{
"epoch": 10.91,
"learning_rate": 2.012779552715655e-06,
"loss": 0.8321,
"step": 120
},
{
"epoch": 10.91,
"eval_accuracy_safe": 0.6743748096576343,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.8493713583566551,
"eval_iou_safe": 0.18617264978095113,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.8413265356526651,
"eval_loss": 0.7439588308334351,
"eval_mean_accuracy": 0.7618730840071447,
"eval_mean_iou": 0.3424997284778721,
"eval_overall_accuracy": 0.8442022409012069,
"eval_runtime": 9.4999,
"eval_samples_per_second": 7.053,
"eval_steps_per_second": 0.526,
"step": 120
},
{
"epoch": 11.0,
"learning_rate": 2.003194888178914e-06,
"loss": 0.8343,
"step": 121
},
{
"epoch": 11.09,
"learning_rate": 1.9936102236421723e-06,
"loss": 0.8413,
"step": 122
},
{
"epoch": 11.18,
"learning_rate": 1.9840255591054315e-06,
"loss": 0.8037,
"step": 123
},
{
"epoch": 11.27,
"learning_rate": 1.9744408945686903e-06,
"loss": 0.8284,
"step": 124
},
{
"epoch": 11.36,
"learning_rate": 1.9648562300319487e-06,
"loss": 0.8023,
"step": 125
},
{
"epoch": 11.45,
"learning_rate": 1.955271565495208e-06,
"loss": 0.8176,
"step": 126
},
{
"epoch": 11.55,
"learning_rate": 1.9456869009584667e-06,
"loss": 0.8078,
"step": 127
},
{
"epoch": 11.64,
"learning_rate": 1.936102236421725e-06,
"loss": 0.8074,
"step": 128
},
{
"epoch": 11.73,
"learning_rate": 1.9265175718849843e-06,
"loss": 0.8139,
"step": 129
},
{
"epoch": 11.82,
"learning_rate": 1.9169329073482426e-06,
"loss": 0.8152,
"step": 130
},
{
"epoch": 11.82,
"eval_accuracy_safe": 0.6688004286799203,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.8589712690862681,
"eval_iou_safe": 0.20063653201077156,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.8506617582762998,
"eval_loss": 0.7270171046257019,
"eval_mean_accuracy": 0.7638858488830942,
"eval_mean_iou": 0.3504327634290238,
"eval_overall_accuracy": 0.8533539273845616,
"eval_runtime": 11.0671,
"eval_samples_per_second": 6.054,
"eval_steps_per_second": 0.452,
"step": 130
},
{
"epoch": 11.91,
"learning_rate": 1.9073482428115018e-06,
"loss": 0.8102,
"step": 131
},
{
"epoch": 12.0,
"learning_rate": 1.8977635782747604e-06,
"loss": 0.8192,
"step": 132
},
{
"epoch": 12.09,
"learning_rate": 1.888178913738019e-06,
"loss": 0.8246,
"step": 133
},
{
"epoch": 12.18,
"learning_rate": 1.878594249201278e-06,
"loss": 0.7864,
"step": 134
},
{
"epoch": 12.27,
"learning_rate": 1.8690095846645368e-06,
"loss": 0.7718,
"step": 135
},
{
"epoch": 12.36,
"learning_rate": 1.8594249201277954e-06,
"loss": 0.8074,
"step": 136
},
{
"epoch": 12.45,
"learning_rate": 1.8498402555910544e-06,
"loss": 0.7569,
"step": 137
},
{
"epoch": 12.55,
"learning_rate": 1.8402555910543132e-06,
"loss": 0.7989,
"step": 138
},
{
"epoch": 12.64,
"learning_rate": 1.8306709265175722e-06,
"loss": 0.7987,
"step": 139
},
{
"epoch": 12.73,
"learning_rate": 1.8210862619808307e-06,
"loss": 0.7929,
"step": 140
},
{
"epoch": 12.73,
"eval_accuracy_safe": 0.6659920354971646,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.8656634973410731,
"eval_iou_safe": 0.20852726465106847,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.8572303764577973,
"eval_loss": 0.7045122981071472,
"eval_mean_accuracy": 0.7658277664191189,
"eval_mean_iou": 0.3552525470362886,
"eval_overall_accuracy": 0.8597655225155959,
"eval_runtime": 9.4502,
"eval_samples_per_second": 7.09,
"eval_steps_per_second": 0.529,
"step": 140
},
{
"epoch": 12.82,
"learning_rate": 1.8115015974440893e-06,
"loss": 0.7703,
"step": 141
},
{
"epoch": 12.91,
"learning_rate": 1.8019169329073483e-06,
"loss": 0.768,
"step": 142
},
{
"epoch": 13.0,
"learning_rate": 1.7923322683706071e-06,
"loss": 0.8397,
"step": 143
},
{
"epoch": 13.09,
"learning_rate": 1.7827476038338657e-06,
"loss": 0.775,
"step": 144
},
{
"epoch": 13.18,
"learning_rate": 1.7731629392971247e-06,
"loss": 0.7718,
"step": 145
},
{
"epoch": 13.27,
"learning_rate": 1.7635782747603833e-06,
"loss": 0.7544,
"step": 146
},
{
"epoch": 13.36,
"learning_rate": 1.7539936102236423e-06,
"loss": 0.7585,
"step": 147
},
{
"epoch": 13.45,
"learning_rate": 1.744408945686901e-06,
"loss": 0.742,
"step": 148
},
{
"epoch": 13.55,
"learning_rate": 1.7348242811501597e-06,
"loss": 0.7629,
"step": 149
},
{
"epoch": 13.64,
"learning_rate": 1.7252396166134187e-06,
"loss": 0.7568,
"step": 150
},
{
"epoch": 13.64,
"eval_accuracy_safe": 0.657079193989229,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.8838084544735693,
"eval_iou_safe": 0.21846087789622765,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.874847055157359,
"eval_loss": 0.6744459867477417,
"eval_mean_accuracy": 0.7704438242313991,
"eval_mean_iou": 0.3644359776845289,
"eval_overall_accuracy": 0.8771112356612931,
"eval_runtime": 10.8117,
"eval_samples_per_second": 6.197,
"eval_steps_per_second": 0.462,
"step": 150
},
{
"epoch": 13.73,
"learning_rate": 1.7156549520766772e-06,
"loss": 0.7311,
"step": 151
},
{
"epoch": 13.82,
"learning_rate": 1.706070287539936e-06,
"loss": 0.7352,
"step": 152
},
{
"epoch": 13.91,
"learning_rate": 1.696485623003195e-06,
"loss": 0.7661,
"step": 153
},
{
"epoch": 14.0,
"learning_rate": 1.6869009584664536e-06,
"loss": 0.7335,
"step": 154
},
{
"epoch": 14.09,
"learning_rate": 1.6773162939297126e-06,
"loss": 0.7491,
"step": 155
},
{
"epoch": 14.18,
"learning_rate": 1.6677316293929714e-06,
"loss": 0.715,
"step": 156
},
{
"epoch": 14.27,
"learning_rate": 1.65814696485623e-06,
"loss": 0.7349,
"step": 157
},
{
"epoch": 14.36,
"learning_rate": 1.648562300319489e-06,
"loss": 0.7127,
"step": 158
},
{
"epoch": 14.45,
"learning_rate": 1.6389776357827476e-06,
"loss": 0.718,
"step": 159
},
{
"epoch": 14.55,
"learning_rate": 1.6293929712460064e-06,
"loss": 0.7085,
"step": 160
},
{
"epoch": 14.55,
"eval_accuracy_safe": 0.6519423595128777,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.8934174588611713,
"eval_iou_safe": 0.2259934305828019,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.8842020655315154,
"eval_loss": 0.65559983253479,
"eval_mean_accuracy": 0.7726799091870244,
"eval_mean_iou": 0.3700651653714391,
"eval_overall_accuracy": 0.8862846716126399,
"eval_runtime": 9.325,
"eval_samples_per_second": 7.185,
"eval_steps_per_second": 0.536,
"step": 160
},
{
"epoch": 14.64,
"learning_rate": 1.6198083067092654e-06,
"loss": 0.753,
"step": 161
},
{
"epoch": 14.73,
"learning_rate": 1.610223642172524e-06,
"loss": 0.7376,
"step": 162
},
{
"epoch": 14.82,
"learning_rate": 1.600638977635783e-06,
"loss": 0.7256,
"step": 163
},
{
"epoch": 14.91,
"learning_rate": 1.5910543130990415e-06,
"loss": 0.7132,
"step": 164
},
{
"epoch": 15.0,
"learning_rate": 1.5814696485623003e-06,
"loss": 0.722,
"step": 165
},
{
"epoch": 15.09,
"learning_rate": 1.5718849840255593e-06,
"loss": 0.7015,
"step": 166
},
{
"epoch": 15.18,
"learning_rate": 1.562300319488818e-06,
"loss": 0.7176,
"step": 167
},
{
"epoch": 15.27,
"learning_rate": 1.5527156549520765e-06,
"loss": 0.7147,
"step": 168
},
{
"epoch": 15.36,
"learning_rate": 1.5431309904153355e-06,
"loss": 0.7217,
"step": 169
},
{
"epoch": 15.45,
"learning_rate": 1.5335463258785943e-06,
"loss": 0.7147,
"step": 170
},
{
"epoch": 15.45,
"eval_accuracy_safe": 0.6561000150346374,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.8963761831582403,
"eval_iou_safe": 0.22827906452703248,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.8872234626234047,
"eval_loss": 0.6508952975273132,
"eval_mean_accuracy": 0.7762380990964388,
"eval_mean_iou": 0.3718341757168124,
"eval_overall_accuracy": 0.889278810415695,
"eval_runtime": 10.5537,
"eval_samples_per_second": 6.348,
"eval_steps_per_second": 0.474,
"step": 170
},
{
"epoch": 15.55,
"learning_rate": 1.5239616613418533e-06,
"loss": 0.7212,
"step": 171
},
{
"epoch": 15.64,
"learning_rate": 1.5143769968051119e-06,
"loss": 0.7196,
"step": 172
},
{
"epoch": 15.73,
"learning_rate": 1.5047923322683706e-06,
"loss": 0.6797,
"step": 173
},
{
"epoch": 15.82,
"learning_rate": 1.4952076677316294e-06,
"loss": 0.6626,
"step": 174
},
{
"epoch": 15.91,
"learning_rate": 1.4856230031948882e-06,
"loss": 0.6783,
"step": 175
},
{
"epoch": 16.0,
"learning_rate": 1.476038338658147e-06,
"loss": 0.7436,
"step": 176
},
{
"epoch": 16.09,
"learning_rate": 1.4664536741214058e-06,
"loss": 0.7123,
"step": 177
},
{
"epoch": 16.18,
"learning_rate": 1.4568690095846646e-06,
"loss": 0.7241,
"step": 178
},
{
"epoch": 16.27,
"learning_rate": 1.4472843450479234e-06,
"loss": 0.6914,
"step": 179
},
{
"epoch": 16.36,
"learning_rate": 1.4376996805111822e-06,
"loss": 0.6991,
"step": 180
},
{
"epoch": 16.36,
"eval_accuracy_safe": 0.6619847263503225,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.8964291023808605,
"eval_iou_safe": 0.22672413565462626,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.8874142561414083,
"eval_loss": 0.6502141952514648,
"eval_mean_accuracy": 0.7792069143655915,
"eval_mean_iou": 0.37137946393201154,
"eval_overall_accuracy": 0.889503991425927,
"eval_runtime": 9.3648,
"eval_samples_per_second": 7.154,
"eval_steps_per_second": 0.534,
"step": 180
},
{
"epoch": 16.45,
"learning_rate": 1.428115015974441e-06,
"loss": 0.6552,
"step": 181
},
{
"epoch": 16.55,
"learning_rate": 1.4185303514376998e-06,
"loss": 0.6621,
"step": 182
},
{
"epoch": 16.64,
"learning_rate": 1.4089456869009586e-06,
"loss": 0.6742,
"step": 183
},
{
"epoch": 16.73,
"learning_rate": 1.3993610223642173e-06,
"loss": 0.6462,
"step": 184
},
{
"epoch": 16.82,
"learning_rate": 1.3897763578274761e-06,
"loss": 0.6759,
"step": 185
},
{
"epoch": 16.91,
"learning_rate": 1.3801916932907347e-06,
"loss": 0.6773,
"step": 186
},
{
"epoch": 17.0,
"learning_rate": 1.3706070287539937e-06,
"loss": 0.7158,
"step": 187
},
{
"epoch": 17.09,
"learning_rate": 1.3610223642172525e-06,
"loss": 0.6958,
"step": 188
},
{
"epoch": 17.18,
"learning_rate": 1.3514376996805113e-06,
"loss": 0.6747,
"step": 189
},
{
"epoch": 17.27,
"learning_rate": 1.3418530351437699e-06,
"loss": 0.6357,
"step": 190
},
{
"epoch": 17.27,
"eval_accuracy_safe": 0.6611674588764114,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.9051011079830231,
"eval_iou_safe": 0.24110413837507133,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.895975219611748,
"eval_loss": 0.6229549646377563,
"eval_mean_accuracy": 0.7831342834297172,
"eval_mean_iou": 0.3790264526622731,
"eval_overall_accuracy": 0.8978956991167211,
"eval_runtime": 10.2719,
"eval_samples_per_second": 6.523,
"eval_steps_per_second": 0.487,
"step": 190
},
{
"epoch": 17.36,
"learning_rate": 1.3322683706070287e-06,
"loss": 0.6769,
"step": 191
},
{
"epoch": 17.45,
"learning_rate": 1.3226837060702877e-06,
"loss": 0.6396,
"step": 192
},
{
"epoch": 17.55,
"learning_rate": 1.3130990415335465e-06,
"loss": 0.6567,
"step": 193
},
{
"epoch": 17.64,
"learning_rate": 1.303514376996805e-06,
"loss": 0.6538,
"step": 194
},
{
"epoch": 17.73,
"learning_rate": 1.2939297124600638e-06,
"loss": 0.6654,
"step": 195
},
{
"epoch": 17.82,
"learning_rate": 1.2843450479233228e-06,
"loss": 0.6637,
"step": 196
},
{
"epoch": 17.91,
"learning_rate": 1.2747603833865816e-06,
"loss": 0.6417,
"step": 197
},
{
"epoch": 18.0,
"learning_rate": 1.2651757188498402e-06,
"loss": 0.6707,
"step": 198
},
{
"epoch": 18.09,
"learning_rate": 1.255591054313099e-06,
"loss": 0.6229,
"step": 199
},
{
"epoch": 18.18,
"learning_rate": 1.2460063897763578e-06,
"loss": 0.6815,
"step": 200
},
{
"epoch": 18.18,
"eval_accuracy_safe": 0.6484207848080771,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.9178010760554833,
"eval_iou_safe": 0.2593923131195133,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.908185465668198,
"eval_loss": 0.5992904305458069,
"eval_mean_accuracy": 0.7831109304317803,
"eval_mean_iou": 0.3891925929292371,
"eval_overall_accuracy": 0.9098440141820195,
"eval_runtime": 9.7351,
"eval_samples_per_second": 6.882,
"eval_steps_per_second": 0.514,
"step": 200
},
{
"epoch": 18.27,
"learning_rate": 1.2364217252396168e-06,
"loss": 0.6677,
"step": 201
},
{
"epoch": 18.36,
"learning_rate": 1.2268370607028754e-06,
"loss": 0.6658,
"step": 202
},
{
"epoch": 18.45,
"learning_rate": 1.2172523961661342e-06,
"loss": 0.649,
"step": 203
},
{
"epoch": 18.55,
"learning_rate": 1.207667731629393e-06,
"loss": 0.6342,
"step": 204
},
{
"epoch": 18.64,
"learning_rate": 1.198083067092652e-06,
"loss": 0.6441,
"step": 205
},
{
"epoch": 18.73,
"learning_rate": 1.1884984025559105e-06,
"loss": 0.6429,
"step": 206
},
{
"epoch": 18.82,
"learning_rate": 1.1789137380191693e-06,
"loss": 0.6155,
"step": 207
},
{
"epoch": 18.91,
"learning_rate": 1.1693290734824281e-06,
"loss": 0.6347,
"step": 208
},
{
"epoch": 19.0,
"learning_rate": 1.159744408945687e-06,
"loss": 0.6611,
"step": 209
},
{
"epoch": 19.09,
"learning_rate": 1.1501597444089457e-06,
"loss": 0.6398,
"step": 210
},
{
"epoch": 19.09,
"eval_accuracy_safe": 0.6413602877398313,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.9258418644556835,
"eval_iou_safe": 0.26823363821482293,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.9159351131912487,
"eval_loss": 0.5784907341003418,
"eval_mean_accuracy": 0.7836010760977574,
"eval_mean_iou": 0.39472291713535723,
"eval_overall_accuracy": 0.9174387348231985,
"eval_runtime": 11.2723,
"eval_samples_per_second": 5.944,
"eval_steps_per_second": 0.444,
"step": 210
},
{
"epoch": 19.18,
"learning_rate": 1.1405750798722045e-06,
"loss": 0.6403,
"step": 211
},
{
"epoch": 19.27,
"learning_rate": 1.1309904153354633e-06,
"loss": 0.6271,
"step": 212
},
{
"epoch": 19.36,
"learning_rate": 1.121405750798722e-06,
"loss": 0.679,
"step": 213
},
{
"epoch": 19.45,
"learning_rate": 1.1118210862619809e-06,
"loss": 0.6239,
"step": 214
},
{
"epoch": 19.55,
"learning_rate": 1.1022364217252397e-06,
"loss": 0.6527,
"step": 215
},
{
"epoch": 19.64,
"learning_rate": 1.0926517571884984e-06,
"loss": 0.6228,
"step": 216
},
{
"epoch": 19.73,
"learning_rate": 1.0830670926517572e-06,
"loss": 0.6227,
"step": 217
},
{
"epoch": 19.82,
"learning_rate": 1.073482428115016e-06,
"loss": 0.6347,
"step": 218
},
{
"epoch": 19.91,
"learning_rate": 1.0638977635782748e-06,
"loss": 0.6374,
"step": 219
},
{
"epoch": 20.0,
"learning_rate": 1.0543130990415336e-06,
"loss": 0.5845,
"step": 220
},
{
"epoch": 20.0,
"eval_accuracy_safe": 0.6426151017151052,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.9286315640516787,
"eval_iou_safe": 0.26982054857603477,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.9187221572756564,
"eval_loss": 0.5641139149665833,
"eval_mean_accuracy": 0.7856233328833919,
"eval_mean_iou": 0.39618090195056377,
"eval_overall_accuracy": 0.9201830963590252,
"eval_runtime": 9.4874,
"eval_samples_per_second": 7.062,
"eval_steps_per_second": 0.527,
"step": 220
},
{
"epoch": 20.09,
"learning_rate": 1.0447284345047924e-06,
"loss": 0.606,
"step": 221
},
{
"epoch": 20.18,
"learning_rate": 1.0351437699680512e-06,
"loss": 0.6491,
"step": 222
},
{
"epoch": 20.27,
"learning_rate": 1.02555910543131e-06,
"loss": 0.6396,
"step": 223
},
{
"epoch": 20.36,
"learning_rate": 1.0159744408945688e-06,
"loss": 0.617,
"step": 224
},
{
"epoch": 20.45,
"learning_rate": 1.0063897763578276e-06,
"loss": 0.614,
"step": 225
},
{
"epoch": 20.55,
"learning_rate": 9.968051118210861e-07,
"loss": 0.6281,
"step": 226
},
{
"epoch": 20.64,
"learning_rate": 9.872204472843452e-07,
"loss": 0.6374,
"step": 227
},
{
"epoch": 20.73,
"learning_rate": 9.77635782747604e-07,
"loss": 0.6394,
"step": 228
},
{
"epoch": 20.82,
"learning_rate": 9.680511182108625e-07,
"loss": 0.6351,
"step": 229
},
{
"epoch": 20.91,
"learning_rate": 9.584664536741213e-07,
"loss": 0.6062,
"step": 230
},
{
"epoch": 20.91,
"eval_accuracy_safe": 0.6519963300064379,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.9252195649054266,
"eval_iou_safe": 0.2640576051310034,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.9156001170929594,
"eval_loss": 0.5693235397338867,
"eval_mean_accuracy": 0.7886079474559322,
"eval_mean_iou": 0.3932192407413209,
"eval_overall_accuracy": 0.9171489886383513,
"eval_runtime": 11.0762,
"eval_samples_per_second": 6.049,
"eval_steps_per_second": 0.451,
"step": 230
},
{
"epoch": 21.0,
"learning_rate": 9.488817891373802e-07,
"loss": 0.6456,
"step": 231
},
{
"epoch": 21.09,
"learning_rate": 9.39297124600639e-07,
"loss": 0.6055,
"step": 232
},
{
"epoch": 21.18,
"learning_rate": 9.297124600638977e-07,
"loss": 0.6134,
"step": 233
},
{
"epoch": 21.27,
"learning_rate": 9.201277955271566e-07,
"loss": 0.5718,
"step": 234
},
{
"epoch": 21.36,
"learning_rate": 9.105431309904154e-07,
"loss": 0.5939,
"step": 235
},
{
"epoch": 21.45,
"learning_rate": 9.009584664536742e-07,
"loss": 0.5917,
"step": 236
},
{
"epoch": 21.55,
"learning_rate": 8.913738019169329e-07,
"loss": 0.599,
"step": 237
},
{
"epoch": 21.64,
"learning_rate": 8.817891373801916e-07,
"loss": 0.6239,
"step": 238
},
{
"epoch": 21.73,
"learning_rate": 8.722044728434505e-07,
"loss": 0.5946,
"step": 239
},
{
"epoch": 21.82,
"learning_rate": 8.626198083067093e-07,
"loss": 0.6071,
"step": 240
},
{
"epoch": 21.82,
"eval_accuracy_safe": 0.6591840432380754,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.9282916959179331,
"eval_iou_safe": 0.267531459702011,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.9188227497665283,
"eval_loss": 0.5627052187919617,
"eval_mean_accuracy": 0.7937378695780043,
"eval_mean_iou": 0.39545140315617977,
"eval_overall_accuracy": 0.9203426873506005,
"eval_runtime": 9.7383,
"eval_samples_per_second": 6.88,
"eval_steps_per_second": 0.513,
"step": 240
}
],
"max_steps": 330,
"num_train_epochs": 30,
"total_flos": 1.3448116877241876e+18,
"trial_name": null,
"trial_params": null
}