userGagan's picture
Upload trainer_state.json
3aa90dc
raw
history blame
45.2 kB
{
"best_metric": 0.15978053212165833,
"best_model_checkpoint": "segformer-b0-finetuned-segments-sidewalk-outputs/checkpoint-240",
"epoch": 10.0,
"global_step": 320,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.03,
"learning_rate": 5.98125e-05,
"loss": 0.0848,
"step": 1
},
{
"epoch": 0.06,
"learning_rate": 5.9625000000000006e-05,
"loss": 0.1057,
"step": 2
},
{
"epoch": 0.09,
"learning_rate": 5.94375e-05,
"loss": 0.1762,
"step": 3
},
{
"epoch": 0.12,
"learning_rate": 5.9250000000000004e-05,
"loss": 0.1069,
"step": 4
},
{
"epoch": 0.16,
"learning_rate": 5.90625e-05,
"loss": 0.4235,
"step": 5
},
{
"epoch": 0.19,
"learning_rate": 5.8875e-05,
"loss": 0.1128,
"step": 6
},
{
"epoch": 0.22,
"learning_rate": 5.8687500000000003e-05,
"loss": 0.0893,
"step": 7
},
{
"epoch": 0.25,
"learning_rate": 5.85e-05,
"loss": 0.0531,
"step": 8
},
{
"epoch": 0.28,
"learning_rate": 5.83125e-05,
"loss": 0.0611,
"step": 9
},
{
"epoch": 0.31,
"learning_rate": 5.8125e-05,
"loss": 0.0709,
"step": 10
},
{
"epoch": 0.34,
"learning_rate": 5.79375e-05,
"loss": 0.2623,
"step": 11
},
{
"epoch": 0.38,
"learning_rate": 5.775e-05,
"loss": 0.1398,
"step": 12
},
{
"epoch": 0.41,
"learning_rate": 5.75625e-05,
"loss": 0.2857,
"step": 13
},
{
"epoch": 0.44,
"learning_rate": 5.7375000000000005e-05,
"loss": 0.0934,
"step": 14
},
{
"epoch": 0.47,
"learning_rate": 5.71875e-05,
"loss": 0.1849,
"step": 15
},
{
"epoch": 0.5,
"learning_rate": 5.6999999999999996e-05,
"loss": 0.0929,
"step": 16
},
{
"epoch": 0.53,
"learning_rate": 5.6812500000000005e-05,
"loss": 0.1765,
"step": 17
},
{
"epoch": 0.56,
"learning_rate": 5.6625e-05,
"loss": 0.1466,
"step": 18
},
{
"epoch": 0.59,
"learning_rate": 5.64375e-05,
"loss": 0.1918,
"step": 19
},
{
"epoch": 0.62,
"learning_rate": 5.625e-05,
"loss": 0.239,
"step": 20
},
{
"epoch": 0.62,
"eval_loss": 0.41225576400756836,
"eval_mean_accuracy": 0.7466867022012313,
"eval_mean_iou": 0.49318473255214207,
"eval_overall_accuracy": 0.8299140242615486,
"eval_per_category_accuracy": [
0.9102603868063844,
NaN,
0.5831130175960781
],
"eval_per_category_iou": [
0.9025356636572145,
0.0,
0.5770185339992117
],
"eval_runtime": 20.0948,
"eval_samples_per_second": 0.796,
"eval_steps_per_second": 0.398,
"step": 20
},
{
"epoch": 0.66,
"learning_rate": 5.60625e-05,
"loss": 0.0813,
"step": 21
},
{
"epoch": 0.69,
"learning_rate": 5.5875e-05,
"loss": 0.3204,
"step": 22
},
{
"epoch": 0.72,
"learning_rate": 5.56875e-05,
"loss": 0.1948,
"step": 23
},
{
"epoch": 0.75,
"learning_rate": 5.550000000000001e-05,
"loss": 0.3204,
"step": 24
},
{
"epoch": 0.78,
"learning_rate": 5.53125e-05,
"loss": 0.2348,
"step": 25
},
{
"epoch": 0.81,
"learning_rate": 5.5125e-05,
"loss": 0.1096,
"step": 26
},
{
"epoch": 0.84,
"learning_rate": 5.49375e-05,
"loss": 0.1375,
"step": 27
},
{
"epoch": 0.88,
"learning_rate": 5.475e-05,
"loss": 0.1218,
"step": 28
},
{
"epoch": 0.91,
"learning_rate": 5.4562500000000005e-05,
"loss": 0.2683,
"step": 29
},
{
"epoch": 0.94,
"learning_rate": 5.4375e-05,
"loss": 0.0548,
"step": 30
},
{
"epoch": 0.97,
"learning_rate": 5.4187499999999996e-05,
"loss": 0.1978,
"step": 31
},
{
"epoch": 1.0,
"learning_rate": 5.4000000000000005e-05,
"loss": 0.1099,
"step": 32
},
{
"epoch": 1.03,
"learning_rate": 5.38125e-05,
"loss": 0.1159,
"step": 33
},
{
"epoch": 1.06,
"learning_rate": 5.3625e-05,
"loss": 0.1443,
"step": 34
},
{
"epoch": 1.09,
"learning_rate": 5.3437500000000005e-05,
"loss": 0.065,
"step": 35
},
{
"epoch": 1.12,
"learning_rate": 5.325e-05,
"loss": 0.953,
"step": 36
},
{
"epoch": 1.16,
"learning_rate": 5.30625e-05,
"loss": 0.1242,
"step": 37
},
{
"epoch": 1.19,
"learning_rate": 5.2875e-05,
"loss": 0.1423,
"step": 38
},
{
"epoch": 1.22,
"learning_rate": 5.268750000000001e-05,
"loss": 0.3064,
"step": 39
},
{
"epoch": 1.25,
"learning_rate": 5.25e-05,
"loss": 0.0384,
"step": 40
},
{
"epoch": 1.25,
"eval_loss": 0.21261490881443024,
"eval_mean_accuracy": 0.9342305073015774,
"eval_mean_iou": 0.6157699255432437,
"eval_overall_accuracy": 0.9372036331047264,
"eval_per_category_accuracy": [
0.9400738425234326,
NaN,
0.9283871720797222
],
"eval_per_category_iou": [
0.9324786248273802,
0.0,
0.914831151802351
],
"eval_runtime": 20.0231,
"eval_samples_per_second": 0.799,
"eval_steps_per_second": 0.4,
"step": 40
},
{
"epoch": 1.28,
"learning_rate": 5.23125e-05,
"loss": 0.1001,
"step": 41
},
{
"epoch": 1.31,
"learning_rate": 5.2125e-05,
"loss": 0.0732,
"step": 42
},
{
"epoch": 1.34,
"learning_rate": 5.19375e-05,
"loss": 0.0736,
"step": 43
},
{
"epoch": 1.38,
"learning_rate": 5.1750000000000004e-05,
"loss": 0.145,
"step": 44
},
{
"epoch": 1.41,
"learning_rate": 5.15625e-05,
"loss": 0.0704,
"step": 45
},
{
"epoch": 1.44,
"learning_rate": 5.1375e-05,
"loss": 0.0602,
"step": 46
},
{
"epoch": 1.47,
"learning_rate": 5.1187500000000004e-05,
"loss": 0.2405,
"step": 47
},
{
"epoch": 1.5,
"learning_rate": 5.1e-05,
"loss": 0.0717,
"step": 48
},
{
"epoch": 1.53,
"learning_rate": 5.08125e-05,
"loss": 0.1311,
"step": 49
},
{
"epoch": 1.56,
"learning_rate": 5.0625000000000004e-05,
"loss": 0.1064,
"step": 50
},
{
"epoch": 1.59,
"learning_rate": 5.04375e-05,
"loss": 0.1129,
"step": 51
},
{
"epoch": 1.62,
"learning_rate": 5.025e-05,
"loss": 0.1423,
"step": 52
},
{
"epoch": 1.66,
"learning_rate": 5.00625e-05,
"loss": 0.2291,
"step": 53
},
{
"epoch": 1.69,
"learning_rate": 4.9875000000000006e-05,
"loss": 0.191,
"step": 54
},
{
"epoch": 1.72,
"learning_rate": 4.96875e-05,
"loss": 0.1337,
"step": 55
},
{
"epoch": 1.75,
"learning_rate": 4.95e-05,
"loss": 0.1408,
"step": 56
},
{
"epoch": 1.78,
"learning_rate": 4.93125e-05,
"loss": 0.1698,
"step": 57
},
{
"epoch": 1.81,
"learning_rate": 4.9125e-05,
"loss": 0.1266,
"step": 58
},
{
"epoch": 1.84,
"learning_rate": 4.8937500000000004e-05,
"loss": 0.1028,
"step": 59
},
{
"epoch": 1.88,
"learning_rate": 4.875e-05,
"loss": 0.1145,
"step": 60
},
{
"epoch": 1.88,
"eval_loss": 0.2426888346672058,
"eval_mean_accuracy": 0.8655657176277471,
"eval_mean_iou": 0.5680959283423686,
"eval_overall_accuracy": 0.8885364845945652,
"eval_per_category_accuracy": [
0.9107121058324004,
NaN,
0.8204193294230938
],
"eval_per_category_iou": [
0.9087610217218509,
0.0,
0.7955267633052547
],
"eval_runtime": 20.2645,
"eval_samples_per_second": 0.79,
"eval_steps_per_second": 0.395,
"step": 60
},
{
"epoch": 1.91,
"learning_rate": 4.85625e-05,
"loss": 0.0419,
"step": 61
},
{
"epoch": 1.94,
"learning_rate": 4.8375000000000004e-05,
"loss": 0.157,
"step": 62
},
{
"epoch": 1.97,
"learning_rate": 4.81875e-05,
"loss": 0.707,
"step": 63
},
{
"epoch": 2.0,
"learning_rate": 4.8e-05,
"loss": 0.0843,
"step": 64
},
{
"epoch": 2.03,
"learning_rate": 4.7812500000000003e-05,
"loss": 0.1084,
"step": 65
},
{
"epoch": 2.06,
"learning_rate": 4.7625e-05,
"loss": 0.2629,
"step": 66
},
{
"epoch": 2.09,
"learning_rate": 4.74375e-05,
"loss": 0.0796,
"step": 67
},
{
"epoch": 2.12,
"learning_rate": 4.7249999999999997e-05,
"loss": 0.1253,
"step": 68
},
{
"epoch": 2.16,
"learning_rate": 4.7062500000000006e-05,
"loss": 0.0744,
"step": 69
},
{
"epoch": 2.19,
"learning_rate": 4.6875e-05,
"loss": 0.0775,
"step": 70
},
{
"epoch": 2.22,
"learning_rate": 4.6687499999999996e-05,
"loss": 0.194,
"step": 71
},
{
"epoch": 2.25,
"learning_rate": 4.6500000000000005e-05,
"loss": 0.4576,
"step": 72
},
{
"epoch": 2.28,
"learning_rate": 4.63125e-05,
"loss": 0.0502,
"step": 73
},
{
"epoch": 2.31,
"learning_rate": 4.6125e-05,
"loss": 0.0402,
"step": 74
},
{
"epoch": 2.34,
"learning_rate": 4.59375e-05,
"loss": 0.4612,
"step": 75
},
{
"epoch": 2.38,
"learning_rate": 4.575e-05,
"loss": 0.0382,
"step": 76
},
{
"epoch": 2.41,
"learning_rate": 4.55625e-05,
"loss": 0.3068,
"step": 77
},
{
"epoch": 2.44,
"learning_rate": 4.5375e-05,
"loss": 0.2821,
"step": 78
},
{
"epoch": 2.47,
"learning_rate": 4.51875e-05,
"loss": 0.175,
"step": 79
},
{
"epoch": 2.5,
"learning_rate": 4.5e-05,
"loss": 0.163,
"step": 80
},
{
"epoch": 2.5,
"eval_loss": 0.22554153203964233,
"eval_mean_accuracy": 0.8740183132328878,
"eval_mean_iou": 0.5773767943990825,
"eval_overall_accuracy": 0.9256960064443283,
"eval_per_category_accuracy": [
0.9755848478903867,
NaN,
0.7724517785753888
],
"eval_per_category_iou": [
0.9662459147578681,
0.0,
0.7658844684393793
],
"eval_runtime": 20.1088,
"eval_samples_per_second": 0.796,
"eval_steps_per_second": 0.398,
"step": 80
},
{
"epoch": 2.53,
"learning_rate": 4.48125e-05,
"loss": 0.2007,
"step": 81
},
{
"epoch": 2.56,
"learning_rate": 4.4625e-05,
"loss": 0.1948,
"step": 82
},
{
"epoch": 2.59,
"learning_rate": 4.44375e-05,
"loss": 0.2429,
"step": 83
},
{
"epoch": 2.62,
"learning_rate": 4.4250000000000005e-05,
"loss": 0.0465,
"step": 84
},
{
"epoch": 2.66,
"learning_rate": 4.40625e-05,
"loss": 0.3231,
"step": 85
},
{
"epoch": 2.69,
"learning_rate": 4.3874999999999996e-05,
"loss": 0.1687,
"step": 86
},
{
"epoch": 2.72,
"learning_rate": 4.3687500000000005e-05,
"loss": 0.103,
"step": 87
},
{
"epoch": 2.75,
"learning_rate": 4.35e-05,
"loss": 0.2989,
"step": 88
},
{
"epoch": 2.78,
"learning_rate": 4.33125e-05,
"loss": 0.9179,
"step": 89
},
{
"epoch": 2.81,
"learning_rate": 4.3125e-05,
"loss": 0.072,
"step": 90
},
{
"epoch": 2.84,
"learning_rate": 4.29375e-05,
"loss": 0.3553,
"step": 91
},
{
"epoch": 2.88,
"learning_rate": 4.275e-05,
"loss": 0.0849,
"step": 92
},
{
"epoch": 2.91,
"learning_rate": 4.25625e-05,
"loss": 0.1466,
"step": 93
},
{
"epoch": 2.94,
"learning_rate": 4.237500000000001e-05,
"loss": 0.1367,
"step": 94
},
{
"epoch": 2.97,
"learning_rate": 4.21875e-05,
"loss": 0.3345,
"step": 95
},
{
"epoch": 3.0,
"learning_rate": 4.2e-05,
"loss": 0.0762,
"step": 96
},
{
"epoch": 3.03,
"learning_rate": 4.18125e-05,
"loss": 0.2443,
"step": 97
},
{
"epoch": 3.06,
"learning_rate": 4.1625e-05,
"loss": 0.2698,
"step": 98
},
{
"epoch": 3.09,
"learning_rate": 4.1437500000000004e-05,
"loss": 0.3754,
"step": 99
},
{
"epoch": 3.12,
"learning_rate": 4.125e-05,
"loss": 0.0832,
"step": 100
},
{
"epoch": 3.12,
"eval_loss": 0.18825271725654602,
"eval_mean_accuracy": 0.9523870826679519,
"eval_mean_iou": 0.6270557614254478,
"eval_overall_accuracy": 0.9553877902482951,
"eval_per_category_accuracy": [
0.958284626685812,
NaN,
0.9464895386500919
],
"eval_per_category_iou": [
0.9550280003976388,
0.0,
0.9261392838787048
],
"eval_runtime": 19.8632,
"eval_samples_per_second": 0.806,
"eval_steps_per_second": 0.403,
"step": 100
},
{
"epoch": 3.16,
"learning_rate": 4.1062499999999995e-05,
"loss": 0.0569,
"step": 101
},
{
"epoch": 3.19,
"learning_rate": 4.0875000000000004e-05,
"loss": 0.3998,
"step": 102
},
{
"epoch": 3.22,
"learning_rate": 4.06875e-05,
"loss": 0.1192,
"step": 103
},
{
"epoch": 3.25,
"learning_rate": 4.05e-05,
"loss": 0.0834,
"step": 104
},
{
"epoch": 3.28,
"learning_rate": 4.0312500000000004e-05,
"loss": 0.1567,
"step": 105
},
{
"epoch": 3.31,
"learning_rate": 4.0125e-05,
"loss": 0.1221,
"step": 106
},
{
"epoch": 3.34,
"learning_rate": 3.99375e-05,
"loss": 0.1699,
"step": 107
},
{
"epoch": 3.38,
"learning_rate": 3.975e-05,
"loss": 0.0803,
"step": 108
},
{
"epoch": 3.41,
"learning_rate": 3.9562500000000006e-05,
"loss": 0.1208,
"step": 109
},
{
"epoch": 3.44,
"learning_rate": 3.9375e-05,
"loss": 0.1283,
"step": 110
},
{
"epoch": 3.47,
"learning_rate": 3.91875e-05,
"loss": 0.4262,
"step": 111
},
{
"epoch": 3.5,
"learning_rate": 3.9e-05,
"loss": 0.0729,
"step": 112
},
{
"epoch": 3.53,
"learning_rate": 3.88125e-05,
"loss": 0.1058,
"step": 113
},
{
"epoch": 3.56,
"learning_rate": 3.8625000000000004e-05,
"loss": 0.1903,
"step": 114
},
{
"epoch": 3.59,
"learning_rate": 3.84375e-05,
"loss": 0.2635,
"step": 115
},
{
"epoch": 3.62,
"learning_rate": 3.825e-05,
"loss": 0.0744,
"step": 116
},
{
"epoch": 3.66,
"learning_rate": 3.8062500000000004e-05,
"loss": 0.2714,
"step": 117
},
{
"epoch": 3.69,
"learning_rate": 3.7875e-05,
"loss": 0.1265,
"step": 118
},
{
"epoch": 3.72,
"learning_rate": 3.76875e-05,
"loss": 0.2965,
"step": 119
},
{
"epoch": 3.75,
"learning_rate": 3.7500000000000003e-05,
"loss": 0.047,
"step": 120
},
{
"epoch": 3.75,
"eval_loss": 0.19484524428844452,
"eval_mean_accuracy": 0.9191032313533529,
"eval_mean_iou": 0.605987572284888,
"eval_overall_accuracy": 0.9443237844948401,
"eval_per_category_accuracy": [
0.9686713143218681,
NaN,
0.8695351483848376
],
"eval_per_category_iou": [
0.9642150125223343,
0.0,
0.8537477043323296
],
"eval_runtime": 21.0542,
"eval_samples_per_second": 0.76,
"eval_steps_per_second": 0.38,
"step": 120
},
{
"epoch": 3.78,
"learning_rate": 3.73125e-05,
"loss": 0.2113,
"step": 121
},
{
"epoch": 3.81,
"learning_rate": 3.7125e-05,
"loss": 0.1938,
"step": 122
},
{
"epoch": 3.84,
"learning_rate": 3.6937499999999997e-05,
"loss": 0.3208,
"step": 123
},
{
"epoch": 3.88,
"learning_rate": 3.6750000000000006e-05,
"loss": 0.0783,
"step": 124
},
{
"epoch": 3.91,
"learning_rate": 3.65625e-05,
"loss": 0.1378,
"step": 125
},
{
"epoch": 3.94,
"learning_rate": 3.6374999999999996e-05,
"loss": 0.1342,
"step": 126
},
{
"epoch": 3.97,
"learning_rate": 3.6187500000000005e-05,
"loss": 0.0532,
"step": 127
},
{
"epoch": 4.0,
"learning_rate": 3.6e-05,
"loss": 0.2819,
"step": 128
},
{
"epoch": 4.03,
"learning_rate": 3.58125e-05,
"loss": 0.1006,
"step": 129
},
{
"epoch": 4.06,
"learning_rate": 3.5625e-05,
"loss": 0.1431,
"step": 130
},
{
"epoch": 4.09,
"learning_rate": 3.54375e-05,
"loss": 0.197,
"step": 131
},
{
"epoch": 4.12,
"learning_rate": 3.525e-05,
"loss": 0.1781,
"step": 132
},
{
"epoch": 4.16,
"learning_rate": 3.50625e-05,
"loss": 0.17,
"step": 133
},
{
"epoch": 4.19,
"learning_rate": 3.4875e-05,
"loss": 0.2306,
"step": 134
},
{
"epoch": 4.22,
"learning_rate": 3.46875e-05,
"loss": 0.1139,
"step": 135
},
{
"epoch": 4.25,
"learning_rate": 3.45e-05,
"loss": 0.0544,
"step": 136
},
{
"epoch": 4.28,
"learning_rate": 3.43125e-05,
"loss": 0.1489,
"step": 137
},
{
"epoch": 4.31,
"learning_rate": 3.4125e-05,
"loss": 0.0706,
"step": 138
},
{
"epoch": 4.34,
"learning_rate": 3.3937500000000005e-05,
"loss": 0.0953,
"step": 139
},
{
"epoch": 4.38,
"learning_rate": 3.375e-05,
"loss": 0.1989,
"step": 140
},
{
"epoch": 4.38,
"eval_loss": 0.18885326385498047,
"eval_mean_accuracy": 0.9465842428924998,
"eval_mean_iou": 0.6241060965774974,
"eval_overall_accuracy": 0.9592237852473473,
"eval_per_category_accuracy": [
0.9714258028937567,
NaN,
0.9217426828912428
],
"eval_per_category_iou": [
0.9659579975137067,
0.0,
0.9063602922187853
],
"eval_runtime": 19.9313,
"eval_samples_per_second": 0.803,
"eval_steps_per_second": 0.401,
"step": 140
},
{
"epoch": 4.41,
"learning_rate": 3.3562499999999996e-05,
"loss": 0.0458,
"step": 141
},
{
"epoch": 4.44,
"learning_rate": 3.3375000000000005e-05,
"loss": 0.2325,
"step": 142
},
{
"epoch": 4.47,
"learning_rate": 3.31875e-05,
"loss": 0.2194,
"step": 143
},
{
"epoch": 4.5,
"learning_rate": 3.3e-05,
"loss": 0.0531,
"step": 144
},
{
"epoch": 4.53,
"learning_rate": 3.28125e-05,
"loss": 0.0904,
"step": 145
},
{
"epoch": 4.56,
"learning_rate": 3.2625e-05,
"loss": 0.1224,
"step": 146
},
{
"epoch": 4.59,
"learning_rate": 3.24375e-05,
"loss": 0.1236,
"step": 147
},
{
"epoch": 4.62,
"learning_rate": 3.225e-05,
"loss": 0.0396,
"step": 148
},
{
"epoch": 4.66,
"learning_rate": 3.206250000000001e-05,
"loss": 0.0329,
"step": 149
},
{
"epoch": 4.69,
"learning_rate": 3.1875e-05,
"loss": 0.0514,
"step": 150
},
{
"epoch": 4.72,
"learning_rate": 3.16875e-05,
"loss": 0.7069,
"step": 151
},
{
"epoch": 4.75,
"learning_rate": 3.15e-05,
"loss": 0.1664,
"step": 152
},
{
"epoch": 4.78,
"learning_rate": 3.13125e-05,
"loss": 0.1071,
"step": 153
},
{
"epoch": 4.81,
"learning_rate": 3.1125000000000004e-05,
"loss": 0.0525,
"step": 154
},
{
"epoch": 4.84,
"learning_rate": 3.09375e-05,
"loss": 0.1084,
"step": 155
},
{
"epoch": 4.88,
"learning_rate": 3.0749999999999995e-05,
"loss": 0.3014,
"step": 156
},
{
"epoch": 4.91,
"learning_rate": 3.0562500000000004e-05,
"loss": 0.0506,
"step": 157
},
{
"epoch": 4.94,
"learning_rate": 3.0375e-05,
"loss": 0.1682,
"step": 158
},
{
"epoch": 4.97,
"learning_rate": 3.0187500000000002e-05,
"loss": 0.1597,
"step": 159
},
{
"epoch": 5.0,
"learning_rate": 3e-05,
"loss": 0.1113,
"step": 160
},
{
"epoch": 5.0,
"eval_loss": 0.28806212544441223,
"eval_mean_accuracy": 0.8278887930983827,
"eval_mean_iou": 0.5469872417793716,
"eval_overall_accuracy": 0.8720594437681833,
"eval_per_category_accuracy": [
0.9147011030779117,
NaN,
0.7410764831188538
],
"eval_per_category_iou": [
0.9109685952665123,
0.0,
0.7299931300716025
],
"eval_runtime": 20.3787,
"eval_samples_per_second": 0.785,
"eval_steps_per_second": 0.393,
"step": 160
},
{
"epoch": 5.03,
"learning_rate": 2.9812500000000003e-05,
"loss": 0.0783,
"step": 161
},
{
"epoch": 5.06,
"learning_rate": 2.9625000000000002e-05,
"loss": 0.0648,
"step": 162
},
{
"epoch": 5.09,
"learning_rate": 2.94375e-05,
"loss": 0.0512,
"step": 163
},
{
"epoch": 5.12,
"learning_rate": 2.925e-05,
"loss": 0.1477,
"step": 164
},
{
"epoch": 5.16,
"learning_rate": 2.90625e-05,
"loss": 0.1665,
"step": 165
},
{
"epoch": 5.19,
"learning_rate": 2.8875e-05,
"loss": 0.1202,
"step": 166
},
{
"epoch": 5.22,
"learning_rate": 2.8687500000000003e-05,
"loss": 0.0873,
"step": 167
},
{
"epoch": 5.25,
"learning_rate": 2.8499999999999998e-05,
"loss": 0.0529,
"step": 168
},
{
"epoch": 5.28,
"learning_rate": 2.83125e-05,
"loss": 0.0848,
"step": 169
},
{
"epoch": 5.31,
"learning_rate": 2.8125e-05,
"loss": 0.1113,
"step": 170
},
{
"epoch": 5.34,
"learning_rate": 2.79375e-05,
"loss": 0.1652,
"step": 171
},
{
"epoch": 5.38,
"learning_rate": 2.7750000000000004e-05,
"loss": 0.0554,
"step": 172
},
{
"epoch": 5.41,
"learning_rate": 2.75625e-05,
"loss": 0.2955,
"step": 173
},
{
"epoch": 5.44,
"learning_rate": 2.7375e-05,
"loss": 0.1975,
"step": 174
},
{
"epoch": 5.47,
"learning_rate": 2.71875e-05,
"loss": 0.2093,
"step": 175
},
{
"epoch": 5.5,
"learning_rate": 2.7000000000000002e-05,
"loss": 0.0438,
"step": 176
},
{
"epoch": 5.53,
"learning_rate": 2.68125e-05,
"loss": 0.0997,
"step": 177
},
{
"epoch": 5.56,
"learning_rate": 2.6625e-05,
"loss": 0.0872,
"step": 178
},
{
"epoch": 5.59,
"learning_rate": 2.64375e-05,
"loss": 0.1928,
"step": 179
},
{
"epoch": 5.62,
"learning_rate": 2.625e-05,
"loss": 0.0877,
"step": 180
},
{
"epoch": 5.62,
"eval_loss": 0.20857231318950653,
"eval_mean_accuracy": 0.9132418593317307,
"eval_mean_iou": 0.5999547882669239,
"eval_overall_accuracy": 0.9410679367076951,
"eval_per_category_accuracy": [
0.9679307991151817,
NaN,
0.8585529195482798
],
"eval_per_category_iou": [
0.9588284256593376,
0.0,
0.841035939141434
],
"eval_runtime": 20.2448,
"eval_samples_per_second": 0.79,
"eval_steps_per_second": 0.395,
"step": 180
},
{
"epoch": 5.66,
"learning_rate": 2.60625e-05,
"loss": 0.5519,
"step": 181
},
{
"epoch": 5.69,
"learning_rate": 2.5875000000000002e-05,
"loss": 0.0631,
"step": 182
},
{
"epoch": 5.72,
"learning_rate": 2.56875e-05,
"loss": 0.1026,
"step": 183
},
{
"epoch": 5.75,
"learning_rate": 2.55e-05,
"loss": 0.1227,
"step": 184
},
{
"epoch": 5.78,
"learning_rate": 2.5312500000000002e-05,
"loss": 0.2057,
"step": 185
},
{
"epoch": 5.81,
"learning_rate": 2.5125e-05,
"loss": 0.1111,
"step": 186
},
{
"epoch": 5.84,
"learning_rate": 2.4937500000000003e-05,
"loss": 0.1093,
"step": 187
},
{
"epoch": 5.88,
"learning_rate": 2.475e-05,
"loss": 0.2486,
"step": 188
},
{
"epoch": 5.91,
"learning_rate": 2.45625e-05,
"loss": 0.3687,
"step": 189
},
{
"epoch": 5.94,
"learning_rate": 2.4375e-05,
"loss": 0.1365,
"step": 190
},
{
"epoch": 5.97,
"learning_rate": 2.4187500000000002e-05,
"loss": 0.1974,
"step": 191
},
{
"epoch": 6.0,
"learning_rate": 2.4e-05,
"loss": 0.0645,
"step": 192
},
{
"epoch": 6.03,
"learning_rate": 2.38125e-05,
"loss": 0.1426,
"step": 193
},
{
"epoch": 6.06,
"learning_rate": 2.3624999999999998e-05,
"loss": 0.1137,
"step": 194
},
{
"epoch": 6.09,
"learning_rate": 2.34375e-05,
"loss": 0.072,
"step": 195
},
{
"epoch": 6.12,
"learning_rate": 2.3250000000000003e-05,
"loss": 0.0737,
"step": 196
},
{
"epoch": 6.16,
"learning_rate": 2.30625e-05,
"loss": 0.1085,
"step": 197
},
{
"epoch": 6.19,
"learning_rate": 2.2875e-05,
"loss": 0.0563,
"step": 198
},
{
"epoch": 6.22,
"learning_rate": 2.26875e-05,
"loss": 0.0642,
"step": 199
},
{
"epoch": 6.25,
"learning_rate": 2.25e-05,
"loss": 0.0688,
"step": 200
},
{
"epoch": 6.25,
"eval_loss": 0.2616099715232849,
"eval_mean_accuracy": 0.881410709224524,
"eval_mean_iou": 0.5805529011665603,
"eval_overall_accuracy": 0.8979442578505313,
"eval_per_category_accuracy": [
0.9139054885999133,
NaN,
0.8489159298491348
],
"eval_per_category_iou": [
0.9093434184699977,
0.0,
0.8323152850296831
],
"eval_runtime": 20.1245,
"eval_samples_per_second": 0.795,
"eval_steps_per_second": 0.398,
"step": 200
},
{
"epoch": 6.28,
"learning_rate": 2.23125e-05,
"loss": 0.0385,
"step": 201
},
{
"epoch": 6.31,
"learning_rate": 2.2125000000000002e-05,
"loss": 0.0416,
"step": 202
},
{
"epoch": 6.34,
"learning_rate": 2.1937499999999998e-05,
"loss": 0.035,
"step": 203
},
{
"epoch": 6.38,
"learning_rate": 2.175e-05,
"loss": 0.1747,
"step": 204
},
{
"epoch": 6.41,
"learning_rate": 2.15625e-05,
"loss": 0.1221,
"step": 205
},
{
"epoch": 6.44,
"learning_rate": 2.1375e-05,
"loss": 0.1319,
"step": 206
},
{
"epoch": 6.47,
"learning_rate": 2.1187500000000003e-05,
"loss": 0.1493,
"step": 207
},
{
"epoch": 6.5,
"learning_rate": 2.1e-05,
"loss": 0.1611,
"step": 208
},
{
"epoch": 6.53,
"learning_rate": 2.08125e-05,
"loss": 0.1956,
"step": 209
},
{
"epoch": 6.56,
"learning_rate": 2.0625e-05,
"loss": 0.1174,
"step": 210
},
{
"epoch": 6.59,
"learning_rate": 2.0437500000000002e-05,
"loss": 0.0776,
"step": 211
},
{
"epoch": 6.62,
"learning_rate": 2.025e-05,
"loss": 0.2234,
"step": 212
},
{
"epoch": 6.66,
"learning_rate": 2.00625e-05,
"loss": 0.1132,
"step": 213
},
{
"epoch": 6.69,
"learning_rate": 1.9875e-05,
"loss": 0.1123,
"step": 214
},
{
"epoch": 6.72,
"learning_rate": 1.96875e-05,
"loss": 0.078,
"step": 215
},
{
"epoch": 6.75,
"learning_rate": 1.95e-05,
"loss": 0.1747,
"step": 216
},
{
"epoch": 6.78,
"learning_rate": 1.9312500000000002e-05,
"loss": 0.2602,
"step": 217
},
{
"epoch": 6.81,
"learning_rate": 1.9125e-05,
"loss": 0.1234,
"step": 218
},
{
"epoch": 6.84,
"learning_rate": 1.89375e-05,
"loss": 0.2718,
"step": 219
},
{
"epoch": 6.88,
"learning_rate": 1.8750000000000002e-05,
"loss": 0.0785,
"step": 220
},
{
"epoch": 6.88,
"eval_loss": 0.18271270394325256,
"eval_mean_accuracy": 0.950766931450973,
"eval_mean_iou": 0.6214100814010263,
"eval_overall_accuracy": 0.9585103367613023,
"eval_per_category_accuracy": [
0.9659856998391196,
NaN,
0.9355481630628265
],
"eval_per_category_iou": [
0.949549266878738,
0.0,
0.9146809773243406
],
"eval_runtime": 21.0499,
"eval_samples_per_second": 0.76,
"eval_steps_per_second": 0.38,
"step": 220
},
{
"epoch": 6.91,
"learning_rate": 1.85625e-05,
"loss": 0.0806,
"step": 221
},
{
"epoch": 6.94,
"learning_rate": 1.8375000000000003e-05,
"loss": 0.0847,
"step": 222
},
{
"epoch": 6.97,
"learning_rate": 1.8187499999999998e-05,
"loss": 0.259,
"step": 223
},
{
"epoch": 7.0,
"learning_rate": 1.8e-05,
"loss": 0.0936,
"step": 224
},
{
"epoch": 7.03,
"learning_rate": 1.78125e-05,
"loss": 0.0977,
"step": 225
},
{
"epoch": 7.06,
"learning_rate": 1.7625e-05,
"loss": 0.0597,
"step": 226
},
{
"epoch": 7.09,
"learning_rate": 1.74375e-05,
"loss": 0.0843,
"step": 227
},
{
"epoch": 7.12,
"learning_rate": 1.725e-05,
"loss": 0.0709,
"step": 228
},
{
"epoch": 7.16,
"learning_rate": 1.70625e-05,
"loss": 0.0542,
"step": 229
},
{
"epoch": 7.19,
"learning_rate": 1.6875e-05,
"loss": 0.2108,
"step": 230
},
{
"epoch": 7.22,
"learning_rate": 1.6687500000000002e-05,
"loss": 0.0454,
"step": 231
},
{
"epoch": 7.25,
"learning_rate": 1.65e-05,
"loss": 0.1047,
"step": 232
},
{
"epoch": 7.28,
"learning_rate": 1.63125e-05,
"loss": 0.03,
"step": 233
},
{
"epoch": 7.31,
"learning_rate": 1.6125e-05,
"loss": 0.204,
"step": 234
},
{
"epoch": 7.34,
"learning_rate": 1.59375e-05,
"loss": 0.1506,
"step": 235
},
{
"epoch": 7.38,
"learning_rate": 1.575e-05,
"loss": 0.07,
"step": 236
},
{
"epoch": 7.41,
"learning_rate": 1.5562500000000002e-05,
"loss": 0.0615,
"step": 237
},
{
"epoch": 7.44,
"learning_rate": 1.5374999999999998e-05,
"loss": 0.1664,
"step": 238
},
{
"epoch": 7.47,
"learning_rate": 1.51875e-05,
"loss": 0.0805,
"step": 239
},
{
"epoch": 7.5,
"learning_rate": 1.5e-05,
"loss": 0.1514,
"step": 240
},
{
"epoch": 7.5,
"eval_loss": 0.15978053212165833,
"eval_mean_accuracy": 0.9622730849481105,
"eval_mean_iou": 0.6345740505797159,
"eval_overall_accuracy": 0.9649951570787098,
"eval_per_category_accuracy": [
0.9676230031857829,
NaN,
0.956923166710438
],
"eval_per_category_iou": [
0.9619821347955928,
0.0,
0.941740016943555
],
"eval_runtime": 19.855,
"eval_samples_per_second": 0.806,
"eval_steps_per_second": 0.403,
"step": 240
},
{
"epoch": 7.53,
"learning_rate": 1.4812500000000001e-05,
"loss": 0.155,
"step": 241
},
{
"epoch": 7.56,
"learning_rate": 1.4625e-05,
"loss": 0.1469,
"step": 242
},
{
"epoch": 7.59,
"learning_rate": 1.44375e-05,
"loss": 0.1164,
"step": 243
},
{
"epoch": 7.62,
"learning_rate": 1.4249999999999999e-05,
"loss": 0.1728,
"step": 244
},
{
"epoch": 7.66,
"learning_rate": 1.40625e-05,
"loss": 0.1716,
"step": 245
},
{
"epoch": 7.69,
"learning_rate": 1.3875000000000002e-05,
"loss": 0.0593,
"step": 246
},
{
"epoch": 7.72,
"learning_rate": 1.36875e-05,
"loss": 0.1736,
"step": 247
},
{
"epoch": 7.75,
"learning_rate": 1.3500000000000001e-05,
"loss": 0.1599,
"step": 248
},
{
"epoch": 7.78,
"learning_rate": 1.33125e-05,
"loss": 0.1509,
"step": 249
},
{
"epoch": 7.81,
"learning_rate": 1.3125e-05,
"loss": 0.3715,
"step": 250
},
{
"epoch": 7.84,
"learning_rate": 1.2937500000000001e-05,
"loss": 0.1907,
"step": 251
},
{
"epoch": 7.88,
"learning_rate": 1.275e-05,
"loss": 0.062,
"step": 252
},
{
"epoch": 7.91,
"learning_rate": 1.25625e-05,
"loss": 0.0986,
"step": 253
},
{
"epoch": 7.94,
"learning_rate": 1.2375e-05,
"loss": 0.056,
"step": 254
},
{
"epoch": 7.97,
"learning_rate": 1.21875e-05,
"loss": 0.0457,
"step": 255
},
{
"epoch": 8.0,
"learning_rate": 1.2e-05,
"loss": 0.2367,
"step": 256
},
{
"epoch": 8.03,
"learning_rate": 1.1812499999999999e-05,
"loss": 0.0705,
"step": 257
},
{
"epoch": 8.06,
"learning_rate": 1.1625000000000001e-05,
"loss": 0.2302,
"step": 258
},
{
"epoch": 8.09,
"learning_rate": 1.14375e-05,
"loss": 0.0406,
"step": 259
},
{
"epoch": 8.12,
"learning_rate": 1.125e-05,
"loss": 0.0928,
"step": 260
},
{
"epoch": 8.12,
"eval_loss": 0.1784609854221344,
"eval_mean_accuracy": 0.9522921660717218,
"eval_mean_iou": 0.626629108700037,
"eval_overall_accuracy": 0.9567627283904122,
"eval_per_category_accuracy": [
0.961078539736312,
NaN,
0.9435057924071318
],
"eval_per_category_iou": [
0.956243026139247,
0.0,
0.9236442999608637
],
"eval_runtime": 20.2853,
"eval_samples_per_second": 0.789,
"eval_steps_per_second": 0.394,
"step": 260
},
{
"epoch": 8.16,
"learning_rate": 1.1062500000000001e-05,
"loss": 0.7041,
"step": 261
},
{
"epoch": 8.19,
"learning_rate": 1.0875e-05,
"loss": 0.3392,
"step": 262
},
{
"epoch": 8.22,
"learning_rate": 1.06875e-05,
"loss": 0.4754,
"step": 263
},
{
"epoch": 8.25,
"learning_rate": 1.05e-05,
"loss": 0.1551,
"step": 264
},
{
"epoch": 8.28,
"learning_rate": 1.03125e-05,
"loss": 0.3086,
"step": 265
},
{
"epoch": 8.31,
"learning_rate": 1.0125e-05,
"loss": 0.098,
"step": 266
},
{
"epoch": 8.34,
"learning_rate": 9.9375e-06,
"loss": 0.0402,
"step": 267
},
{
"epoch": 8.38,
"learning_rate": 9.75e-06,
"loss": 0.1219,
"step": 268
},
{
"epoch": 8.41,
"learning_rate": 9.5625e-06,
"loss": 0.13,
"step": 269
},
{
"epoch": 8.44,
"learning_rate": 9.375000000000001e-06,
"loss": 0.1765,
"step": 270
},
{
"epoch": 8.47,
"learning_rate": 9.187500000000001e-06,
"loss": 0.0731,
"step": 271
},
{
"epoch": 8.5,
"learning_rate": 9e-06,
"loss": 0.0719,
"step": 272
},
{
"epoch": 8.53,
"learning_rate": 8.8125e-06,
"loss": 0.1989,
"step": 273
},
{
"epoch": 8.56,
"learning_rate": 8.625e-06,
"loss": 0.5729,
"step": 274
},
{
"epoch": 8.59,
"learning_rate": 8.4375e-06,
"loss": 0.0705,
"step": 275
},
{
"epoch": 8.62,
"learning_rate": 8.25e-06,
"loss": 0.0826,
"step": 276
},
{
"epoch": 8.66,
"learning_rate": 8.0625e-06,
"loss": 0.1006,
"step": 277
},
{
"epoch": 8.69,
"learning_rate": 7.875e-06,
"loss": 0.211,
"step": 278
},
{
"epoch": 8.72,
"learning_rate": 7.687499999999999e-06,
"loss": 0.1274,
"step": 279
},
{
"epoch": 8.75,
"learning_rate": 7.5e-06,
"loss": 0.0743,
"step": 280
},
{
"epoch": 8.75,
"eval_loss": 0.24013127386569977,
"eval_mean_accuracy": 0.8523734940480358,
"eval_mean_iou": 0.5607729730415726,
"eval_overall_accuracy": 0.8913641199553799,
"eval_per_category_accuracy": [
0.9290050639080298,
NaN,
0.7757419241880417
],
"eval_per_category_iou": [
0.9208480465549874,
0.0,
0.7614708725697303
],
"eval_runtime": 20.1953,
"eval_samples_per_second": 0.792,
"eval_steps_per_second": 0.396,
"step": 280
},
{
"epoch": 8.78,
"learning_rate": 7.3125e-06,
"loss": 0.1298,
"step": 281
},
{
"epoch": 8.81,
"learning_rate": 7.1249999999999995e-06,
"loss": 0.1123,
"step": 282
},
{
"epoch": 8.84,
"learning_rate": 6.937500000000001e-06,
"loss": 0.2529,
"step": 283
},
{
"epoch": 8.88,
"learning_rate": 6.750000000000001e-06,
"loss": 0.0566,
"step": 284
},
{
"epoch": 8.91,
"learning_rate": 6.5625e-06,
"loss": 0.1256,
"step": 285
},
{
"epoch": 8.94,
"learning_rate": 6.375e-06,
"loss": 0.0894,
"step": 286
},
{
"epoch": 8.97,
"learning_rate": 6.1875e-06,
"loss": 0.0488,
"step": 287
},
{
"epoch": 9.0,
"learning_rate": 6e-06,
"loss": 0.0572,
"step": 288
},
{
"epoch": 9.03,
"learning_rate": 5.812500000000001e-06,
"loss": 0.0958,
"step": 289
},
{
"epoch": 9.06,
"learning_rate": 5.625e-06,
"loss": 0.0657,
"step": 290
},
{
"epoch": 9.09,
"learning_rate": 5.4375e-06,
"loss": 0.2126,
"step": 291
},
{
"epoch": 9.12,
"learning_rate": 5.25e-06,
"loss": 0.0488,
"step": 292
},
{
"epoch": 9.16,
"learning_rate": 5.0625e-06,
"loss": 0.0482,
"step": 293
},
{
"epoch": 9.19,
"learning_rate": 4.875e-06,
"loss": 0.0917,
"step": 294
},
{
"epoch": 9.22,
"learning_rate": 4.6875000000000004e-06,
"loss": 0.3122,
"step": 295
},
{
"epoch": 9.25,
"learning_rate": 4.5e-06,
"loss": 0.098,
"step": 296
},
{
"epoch": 9.28,
"learning_rate": 4.3125e-06,
"loss": 0.0843,
"step": 297
},
{
"epoch": 9.31,
"learning_rate": 4.125e-06,
"loss": 0.0494,
"step": 298
},
{
"epoch": 9.34,
"learning_rate": 3.9375e-06,
"loss": 0.0898,
"step": 299
},
{
"epoch": 9.38,
"learning_rate": 3.75e-06,
"loss": 0.1218,
"step": 300
},
{
"epoch": 9.38,
"eval_loss": 0.2210669219493866,
"eval_mean_accuracy": 0.8813038618860116,
"eval_mean_iou": 0.5779585539907081,
"eval_overall_accuracy": 0.9082890817298062,
"eval_per_category_accuracy": [
0.934340193350943,
NaN,
0.8282675304210803
],
"eval_per_category_iou": [
0.9255767862587662,
0.0,
0.8082988757133581
],
"eval_runtime": 19.925,
"eval_samples_per_second": 0.803,
"eval_steps_per_second": 0.402,
"step": 300
},
{
"epoch": 9.41,
"learning_rate": 3.5624999999999998e-06,
"loss": 0.0511,
"step": 301
},
{
"epoch": 9.44,
"learning_rate": 3.3750000000000003e-06,
"loss": 0.1314,
"step": 302
},
{
"epoch": 9.47,
"learning_rate": 3.1875e-06,
"loss": 0.0482,
"step": 303
},
{
"epoch": 9.5,
"learning_rate": 3e-06,
"loss": 0.1668,
"step": 304
},
{
"epoch": 9.53,
"learning_rate": 2.8125e-06,
"loss": 0.0771,
"step": 305
},
{
"epoch": 9.56,
"learning_rate": 2.625e-06,
"loss": 0.0385,
"step": 306
},
{
"epoch": 9.59,
"learning_rate": 2.4375e-06,
"loss": 0.0811,
"step": 307
},
{
"epoch": 9.62,
"learning_rate": 2.25e-06,
"loss": 0.0863,
"step": 308
},
{
"epoch": 9.66,
"learning_rate": 2.0625e-06,
"loss": 0.0905,
"step": 309
},
{
"epoch": 9.69,
"learning_rate": 1.875e-06,
"loss": 0.1298,
"step": 310
},
{
"epoch": 9.72,
"learning_rate": 1.6875000000000001e-06,
"loss": 0.18,
"step": 311
},
{
"epoch": 9.75,
"learning_rate": 1.5e-06,
"loss": 0.0558,
"step": 312
},
{
"epoch": 9.78,
"learning_rate": 1.3125e-06,
"loss": 0.1444,
"step": 313
},
{
"epoch": 9.81,
"learning_rate": 1.125e-06,
"loss": 0.2402,
"step": 314
},
{
"epoch": 9.84,
"learning_rate": 9.375e-07,
"loss": 0.0606,
"step": 315
},
{
"epoch": 9.88,
"learning_rate": 7.5e-07,
"loss": 0.156,
"step": 316
},
{
"epoch": 9.91,
"learning_rate": 5.625e-07,
"loss": 0.1229,
"step": 317
},
{
"epoch": 9.94,
"learning_rate": 3.75e-07,
"loss": 0.0599,
"step": 318
},
{
"epoch": 9.97,
"learning_rate": 1.875e-07,
"loss": 0.1281,
"step": 319
},
{
"epoch": 10.0,
"learning_rate": 0.0,
"loss": 0.0554,
"step": 320
},
{
"epoch": 10.0,
"eval_loss": 0.21498346328735352,
"eval_mean_accuracy": 0.9014076839216982,
"eval_mean_iou": 0.5929254854495193,
"eval_overall_accuracy": 0.9170195985123291,
"eval_per_category_accuracy": [
0.9320910980952277,
NaN,
0.8707242697481689
],
"eval_per_category_iou": [
0.9254394052507448,
0.0,
0.853337051097813
],
"eval_runtime": 20.094,
"eval_samples_per_second": 0.796,
"eval_steps_per_second": 0.398,
"step": 320
}
],
"max_steps": 320,
"num_train_epochs": 10,
"total_flos": 1.12186676477952e+16,
"trial_name": null,
"trial_params": null
}