FINAL_ecc_segformer / trainer_state.json
rishitunu's picture
End of training
8db463e
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 18.248175182481752,
"eval_steps": 500,
"global_step": 10000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.18,
"learning_rate": 5.9401e-05,
"loss": 0.1021,
"step": 100
},
{
"epoch": 0.36,
"learning_rate": 5.8802000000000004e-05,
"loss": 0.0544,
"step": 200
},
{
"epoch": 0.55,
"learning_rate": 5.8203e-05,
"loss": 0.0664,
"step": 300
},
{
"epoch": 0.73,
"learning_rate": 5.7604e-05,
"loss": 0.0563,
"step": 400
},
{
"epoch": 0.91,
"learning_rate": 5.7005e-05,
"loss": 0.0534,
"step": 500
},
{
"epoch": 1.0,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.2749723182564086,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.27359924384134937,
"eval_loss": 0.061391595751047134,
"eval_mean_accuracy": 0.2749723182564086,
"eval_mean_iou": 0.13679962192067469,
"eval_overall_accuracy": 0.2749723182564086,
"eval_runtime": 371.8778,
"eval_samples_per_second": 0.522,
"eval_steps_per_second": 0.261,
"step": 548
},
{
"epoch": 1.09,
"learning_rate": 5.6406e-05,
"loss": 0.0555,
"step": 600
},
{
"epoch": 1.28,
"learning_rate": 5.5806999999999996e-05,
"loss": 0.0437,
"step": 700
},
{
"epoch": 1.46,
"learning_rate": 5.5208000000000004e-05,
"loss": 0.0619,
"step": 800
},
{
"epoch": 1.64,
"learning_rate": 5.4609000000000005e-05,
"loss": 0.0452,
"step": 900
},
{
"epoch": 1.82,
"learning_rate": 5.401e-05,
"loss": 0.058,
"step": 1000
},
{
"epoch": 2.0,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.42382043908423217,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.4185528527078574,
"eval_loss": 0.10176104307174683,
"eval_mean_accuracy": 0.42382043908423217,
"eval_mean_iou": 0.2092764263539287,
"eval_overall_accuracy": 0.42382043908423217,
"eval_runtime": 378.5472,
"eval_samples_per_second": 0.512,
"eval_steps_per_second": 0.256,
"step": 1096
},
{
"epoch": 2.01,
"learning_rate": 5.3411e-05,
"loss": 0.0574,
"step": 1100
},
{
"epoch": 2.19,
"learning_rate": 5.2812e-05,
"loss": 0.0416,
"step": 1200
},
{
"epoch": 2.37,
"learning_rate": 5.2213e-05,
"loss": 0.0675,
"step": 1300
},
{
"epoch": 2.55,
"learning_rate": 5.1614000000000004e-05,
"loss": 0.0482,
"step": 1400
},
{
"epoch": 2.74,
"learning_rate": 5.1015e-05,
"loss": 0.0502,
"step": 1500
},
{
"epoch": 2.92,
"learning_rate": 5.0416e-05,
"loss": 0.0482,
"step": 1600
},
{
"epoch": 3.0,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.43149166759602975,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.35824652777777777,
"eval_loss": 0.05083267390727997,
"eval_mean_accuracy": 0.43149166759602975,
"eval_mean_iou": 0.17912326388888888,
"eval_overall_accuracy": 0.43149166759602975,
"eval_runtime": 380.6012,
"eval_samples_per_second": 0.51,
"eval_steps_per_second": 0.255,
"step": 1644
},
{
"epoch": 3.1,
"learning_rate": 4.9817e-05,
"loss": 0.0507,
"step": 1700
},
{
"epoch": 3.28,
"learning_rate": 4.9218e-05,
"loss": 0.0492,
"step": 1800
},
{
"epoch": 3.47,
"learning_rate": 4.8619e-05,
"loss": 0.0458,
"step": 1900
},
{
"epoch": 3.65,
"learning_rate": 4.8020000000000004e-05,
"loss": 0.0469,
"step": 2000
},
{
"epoch": 3.83,
"learning_rate": 4.7421000000000006e-05,
"loss": 0.0338,
"step": 2100
},
{
"epoch": 4.0,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.3716453311453471,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.3698312917782239,
"eval_loss": 0.05686204507946968,
"eval_mean_accuracy": 0.3716453311453471,
"eval_mean_iou": 0.18491564588911194,
"eval_overall_accuracy": 0.3716453311453471,
"eval_runtime": 385.9938,
"eval_samples_per_second": 0.503,
"eval_steps_per_second": 0.251,
"step": 2192
},
{
"epoch": 4.01,
"learning_rate": 4.6822e-05,
"loss": 0.051,
"step": 2200
},
{
"epoch": 4.2,
"learning_rate": 4.6223e-05,
"loss": 0.0353,
"step": 2300
},
{
"epoch": 4.38,
"learning_rate": 4.5624e-05,
"loss": 0.0503,
"step": 2400
},
{
"epoch": 4.56,
"learning_rate": 4.5025000000000003e-05,
"loss": 0.0425,
"step": 2500
},
{
"epoch": 4.74,
"learning_rate": 4.4426000000000005e-05,
"loss": 0.0528,
"step": 2600
},
{
"epoch": 4.93,
"learning_rate": 4.3827e-05,
"loss": 0.0395,
"step": 2700
},
{
"epoch": 5.0,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.3505922698233148,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.34898750111508686,
"eval_loss": 0.0596516877412796,
"eval_mean_accuracy": 0.3505922698233148,
"eval_mean_iou": 0.17449375055754343,
"eval_overall_accuracy": 0.3505922698233148,
"eval_runtime": 382.2611,
"eval_samples_per_second": 0.508,
"eval_steps_per_second": 0.254,
"step": 2740
},
{
"epoch": 5.11,
"learning_rate": 4.3228e-05,
"loss": 0.0445,
"step": 2800
},
{
"epoch": 5.29,
"learning_rate": 4.2629e-05,
"loss": 0.0495,
"step": 2900
},
{
"epoch": 5.47,
"learning_rate": 4.203e-05,
"loss": 0.0468,
"step": 3000
},
{
"epoch": 5.66,
"learning_rate": 4.1431e-05,
"loss": 0.0488,
"step": 3100
},
{
"epoch": 5.84,
"learning_rate": 4.0832e-05,
"loss": 0.0372,
"step": 3200
},
{
"epoch": 6.0,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.46350689852948207,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.459698930762766,
"eval_loss": 0.050851527601480484,
"eval_mean_accuracy": 0.46350689852948207,
"eval_mean_iou": 0.229849465381383,
"eval_overall_accuracy": 0.46350689852948207,
"eval_runtime": 381.5402,
"eval_samples_per_second": 0.508,
"eval_steps_per_second": 0.254,
"step": 3288
},
{
"epoch": 6.02,
"learning_rate": 4.0233e-05,
"loss": 0.0452,
"step": 3300
},
{
"epoch": 6.2,
"learning_rate": 3.9634e-05,
"loss": 0.0588,
"step": 3400
},
{
"epoch": 6.39,
"learning_rate": 3.9035e-05,
"loss": 0.0408,
"step": 3500
},
{
"epoch": 6.57,
"learning_rate": 3.8436e-05,
"loss": 0.0441,
"step": 3600
},
{
"epoch": 6.75,
"learning_rate": 3.7837000000000004e-05,
"loss": 0.0319,
"step": 3700
},
{
"epoch": 6.93,
"learning_rate": 3.7238000000000005e-05,
"loss": 0.0402,
"step": 3800
},
{
"epoch": 7.0,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.35071275511016936,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.3502962805306196,
"eval_loss": 0.061992477625608444,
"eval_mean_accuracy": 0.35071275511016936,
"eval_mean_iou": 0.1751481402653098,
"eval_overall_accuracy": 0.35071275511016936,
"eval_runtime": 381.7043,
"eval_samples_per_second": 0.508,
"eval_steps_per_second": 0.254,
"step": 3836
},
{
"epoch": 7.12,
"learning_rate": 3.6639e-05,
"loss": 0.0347,
"step": 3900
},
{
"epoch": 7.3,
"learning_rate": 3.604e-05,
"loss": 0.0356,
"step": 4000
},
{
"epoch": 7.48,
"learning_rate": 3.544100000000001e-05,
"loss": 0.046,
"step": 4100
},
{
"epoch": 7.66,
"learning_rate": 3.4842e-05,
"loss": 0.0417,
"step": 4200
},
{
"epoch": 7.85,
"learning_rate": 3.4243000000000004e-05,
"loss": 0.038,
"step": 4300
},
{
"epoch": 8.0,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.38145343094301143,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.381008030267165,
"eval_loss": 0.06807420402765274,
"eval_mean_accuracy": 0.38145343094301143,
"eval_mean_iou": 0.1905040151335825,
"eval_overall_accuracy": 0.38145343094301143,
"eval_runtime": 380.8562,
"eval_samples_per_second": 0.509,
"eval_steps_per_second": 0.255,
"step": 4384
},
{
"epoch": 8.03,
"learning_rate": 3.3644000000000005e-05,
"loss": 0.0471,
"step": 4400
},
{
"epoch": 8.21,
"learning_rate": 3.3045000000000006e-05,
"loss": 0.0504,
"step": 4500
},
{
"epoch": 8.39,
"learning_rate": 3.2446e-05,
"loss": 0.0406,
"step": 4600
},
{
"epoch": 8.58,
"learning_rate": 3.1847e-05,
"loss": 0.0336,
"step": 4700
},
{
"epoch": 8.76,
"learning_rate": 3.1248e-05,
"loss": 0.0322,
"step": 4800
},
{
"epoch": 8.94,
"learning_rate": 3.0649000000000004e-05,
"loss": 0.0393,
"step": 4900
},
{
"epoch": 9.0,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.4433271066006022,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.4425224978083646,
"eval_loss": 0.06852483004331589,
"eval_mean_accuracy": 0.4433271066006022,
"eval_mean_iou": 0.2212612489041823,
"eval_overall_accuracy": 0.4433271066006022,
"eval_runtime": 380.8095,
"eval_samples_per_second": 0.509,
"eval_steps_per_second": 0.255,
"step": 4932
},
{
"epoch": 9.12,
"learning_rate": 3.0050000000000002e-05,
"loss": 0.0475,
"step": 5000
},
{
"epoch": 9.31,
"learning_rate": 2.9451e-05,
"loss": 0.0335,
"step": 5100
},
{
"epoch": 9.49,
"learning_rate": 2.8851999999999998e-05,
"loss": 0.0402,
"step": 5200
},
{
"epoch": 9.67,
"learning_rate": 2.8253e-05,
"loss": 0.0405,
"step": 5300
},
{
"epoch": 9.85,
"learning_rate": 2.7653999999999996e-05,
"loss": 0.0376,
"step": 5400
},
{
"epoch": 10.0,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.39291446938677965,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.39239480430627627,
"eval_loss": 0.05900894105434418,
"eval_mean_accuracy": 0.39291446938677965,
"eval_mean_iou": 0.19619740215313813,
"eval_overall_accuracy": 0.39291446938677965,
"eval_runtime": 380.8979,
"eval_samples_per_second": 0.509,
"eval_steps_per_second": 0.255,
"step": 5480
},
{
"epoch": 10.04,
"learning_rate": 2.7054999999999998e-05,
"loss": 0.0282,
"step": 5500
},
{
"epoch": 10.22,
"learning_rate": 2.6455999999999995e-05,
"loss": 0.0246,
"step": 5600
},
{
"epoch": 10.4,
"learning_rate": 2.5857e-05,
"loss": 0.0362,
"step": 5700
},
{
"epoch": 10.58,
"learning_rate": 2.5258e-05,
"loss": 0.0333,
"step": 5800
},
{
"epoch": 10.77,
"learning_rate": 2.4659e-05,
"loss": 0.0437,
"step": 5900
},
{
"epoch": 10.95,
"learning_rate": 2.406e-05,
"loss": 0.0381,
"step": 6000
},
{
"epoch": 11.0,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.3801490034572307,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.37829945440961094,
"eval_loss": 0.06262247264385223,
"eval_mean_accuracy": 0.3801490034572307,
"eval_mean_iou": 0.18914972720480547,
"eval_overall_accuracy": 0.3801490034572307,
"eval_runtime": 379.9216,
"eval_samples_per_second": 0.511,
"eval_steps_per_second": 0.255,
"step": 6028
},
{
"epoch": 11.13,
"learning_rate": 2.3460999999999998e-05,
"loss": 0.0338,
"step": 6100
},
{
"epoch": 11.31,
"learning_rate": 2.2862e-05,
"loss": 0.0453,
"step": 6200
},
{
"epoch": 11.5,
"learning_rate": 2.2263e-05,
"loss": 0.0305,
"step": 6300
},
{
"epoch": 11.68,
"learning_rate": 2.1663999999999998e-05,
"loss": 0.0358,
"step": 6400
},
{
"epoch": 11.86,
"learning_rate": 2.1065e-05,
"loss": 0.034,
"step": 6500
},
{
"epoch": 12.0,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.4161940191501904,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.41221175296407997,
"eval_loss": 0.0623166486620903,
"eval_mean_accuracy": 0.4161940191501904,
"eval_mean_iou": 0.20610587648203998,
"eval_overall_accuracy": 0.4161940191501904,
"eval_runtime": 381.2562,
"eval_samples_per_second": 0.509,
"eval_steps_per_second": 0.254,
"step": 6576
},
{
"epoch": 12.04,
"learning_rate": 2.0465999999999997e-05,
"loss": 0.0294,
"step": 6600
},
{
"epoch": 12.23,
"learning_rate": 1.9866999999999998e-05,
"loss": 0.0294,
"step": 6700
},
{
"epoch": 12.41,
"learning_rate": 1.9267999999999996e-05,
"loss": 0.0452,
"step": 6800
},
{
"epoch": 12.59,
"learning_rate": 1.8669e-05,
"loss": 0.0347,
"step": 6900
},
{
"epoch": 12.77,
"learning_rate": 1.807e-05,
"loss": 0.0279,
"step": 7000
},
{
"epoch": 12.96,
"learning_rate": 1.7471e-05,
"loss": 0.0301,
"step": 7100
},
{
"epoch": 13.0,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.3669474007041917,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.3664464100154627,
"eval_loss": 0.08311110734939575,
"eval_mean_accuracy": 0.3669474007041917,
"eval_mean_iou": 0.18322320500773134,
"eval_overall_accuracy": 0.3669474007041917,
"eval_runtime": 380.1764,
"eval_samples_per_second": 0.51,
"eval_steps_per_second": 0.255,
"step": 7124
},
{
"epoch": 13.14,
"learning_rate": 1.6872e-05,
"loss": 0.0321,
"step": 7200
},
{
"epoch": 13.32,
"learning_rate": 1.6272999999999998e-05,
"loss": 0.0293,
"step": 7300
},
{
"epoch": 13.5,
"learning_rate": 1.5674e-05,
"loss": 0.0342,
"step": 7400
},
{
"epoch": 13.69,
"learning_rate": 1.5075000000000002e-05,
"loss": 0.0305,
"step": 7500
},
{
"epoch": 13.87,
"learning_rate": 1.4476e-05,
"loss": 0.034,
"step": 7600
},
{
"epoch": 14.0,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.4119352127710421,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.4118392523290055,
"eval_loss": 0.06355985999107361,
"eval_mean_accuracy": 0.4119352127710421,
"eval_mean_iou": 0.20591962616450274,
"eval_overall_accuracy": 0.4119352127710421,
"eval_runtime": 381.445,
"eval_samples_per_second": 0.509,
"eval_steps_per_second": 0.254,
"step": 7672
},
{
"epoch": 14.05,
"learning_rate": 1.3877e-05,
"loss": 0.0331,
"step": 7700
},
{
"epoch": 14.23,
"learning_rate": 1.3277999999999999e-05,
"loss": 0.0312,
"step": 7800
},
{
"epoch": 14.42,
"learning_rate": 1.2678999999999998e-05,
"loss": 0.0295,
"step": 7900
},
{
"epoch": 14.6,
"learning_rate": 1.2079999999999998e-05,
"loss": 0.0346,
"step": 8000
},
{
"epoch": 14.78,
"learning_rate": 1.1480999999999997e-05,
"loss": 0.0302,
"step": 8100
},
{
"epoch": 14.96,
"learning_rate": 1.0882000000000004e-05,
"loss": 0.0303,
"step": 8200
},
{
"epoch": 15.0,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.3864062724042889,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.38623359121586986,
"eval_loss": 0.07051627337932587,
"eval_mean_accuracy": 0.3864062724042889,
"eval_mean_iou": 0.19311679560793493,
"eval_overall_accuracy": 0.3864062724042889,
"eval_runtime": 379.3942,
"eval_samples_per_second": 0.511,
"eval_steps_per_second": 0.256,
"step": 8220
},
{
"epoch": 15.15,
"learning_rate": 1.0283000000000003e-05,
"loss": 0.0254,
"step": 8300
},
{
"epoch": 15.33,
"learning_rate": 9.684000000000002e-06,
"loss": 0.0325,
"step": 8400
},
{
"epoch": 15.51,
"learning_rate": 9.085000000000002e-06,
"loss": 0.0377,
"step": 8500
},
{
"epoch": 15.69,
"learning_rate": 8.486000000000001e-06,
"loss": 0.0246,
"step": 8600
},
{
"epoch": 15.88,
"learning_rate": 7.887000000000001e-06,
"loss": 0.0338,
"step": 8700
},
{
"epoch": 16.0,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.4206479917791196,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.42018106442487874,
"eval_loss": 0.06845217943191528,
"eval_mean_accuracy": 0.4206479917791196,
"eval_mean_iou": 0.21009053221243937,
"eval_overall_accuracy": 0.4206479917791196,
"eval_runtime": 380.9614,
"eval_samples_per_second": 0.509,
"eval_steps_per_second": 0.255,
"step": 8768
},
{
"epoch": 16.06,
"learning_rate": 7.2879999999999995e-06,
"loss": 0.0206,
"step": 8800
},
{
"epoch": 16.24,
"learning_rate": 6.688999999999999e-06,
"loss": 0.0259,
"step": 8900
},
{
"epoch": 16.42,
"learning_rate": 6.0899999999999984e-06,
"loss": 0.0333,
"step": 9000
},
{
"epoch": 16.61,
"learning_rate": 5.490999999999998e-06,
"loss": 0.0326,
"step": 9100
},
{
"epoch": 16.79,
"learning_rate": 4.891999999999997e-06,
"loss": 0.0325,
"step": 9200
},
{
"epoch": 16.97,
"learning_rate": 4.292999999999997e-06,
"loss": 0.0229,
"step": 9300
},
{
"epoch": 17.0,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.4204358978444087,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.4197487662985776,
"eval_loss": 0.07058978080749512,
"eval_mean_accuracy": 0.4204358978444087,
"eval_mean_iou": 0.2098743831492888,
"eval_overall_accuracy": 0.4204358978444087,
"eval_runtime": 383.066,
"eval_samples_per_second": 0.506,
"eval_steps_per_second": 0.253,
"step": 9316
},
{
"epoch": 17.15,
"learning_rate": 3.694000000000003e-06,
"loss": 0.0273,
"step": 9400
},
{
"epoch": 17.34,
"learning_rate": 3.0950000000000026e-06,
"loss": 0.0381,
"step": 9500
},
{
"epoch": 17.52,
"learning_rate": 2.496000000000002e-06,
"loss": 0.0262,
"step": 9600
},
{
"epoch": 17.7,
"learning_rate": 1.8970000000000013e-06,
"loss": 0.0242,
"step": 9700
},
{
"epoch": 17.88,
"learning_rate": 1.298000000000001e-06,
"loss": 0.0337,
"step": 9800
},
{
"epoch": 18.0,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.39679987095129604,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.39646904318930265,
"eval_loss": 0.07418648153543472,
"eval_mean_accuracy": 0.39679987095129604,
"eval_mean_iou": 0.19823452159465133,
"eval_overall_accuracy": 0.39679987095129604,
"eval_runtime": 380.0658,
"eval_samples_per_second": 0.51,
"eval_steps_per_second": 0.255,
"step": 9864
},
{
"epoch": 18.07,
"learning_rate": 6.990000000000005e-07,
"loss": 0.0243,
"step": 9900
},
{
"epoch": 18.25,
"learning_rate": 1e-07,
"loss": 0.0257,
"step": 10000
},
{
"epoch": 18.25,
"eval_accuracy_background": NaN,
"eval_accuracy_crack": 0.39388432615865027,
"eval_iou_background": 0.0,
"eval_iou_crack": 0.3935762911195175,
"eval_loss": 0.07488969713449478,
"eval_mean_accuracy": 0.39388432615865027,
"eval_mean_iou": 0.19678814555975874,
"eval_overall_accuracy": 0.39388432615865027,
"eval_runtime": 375.6599,
"eval_samples_per_second": 0.516,
"eval_steps_per_second": 0.258,
"step": 10000
},
{
"epoch": 18.25,
"step": 10000,
"total_flos": 7.976189631720849e+18,
"train_loss": 0.039843981289863585,
"train_runtime": 169492.4367,
"train_samples_per_second": 0.118,
"train_steps_per_second": 0.059
}
],
"logging_steps": 100,
"max_steps": 10000,
"num_train_epochs": 19,
"save_steps": 500,
"total_flos": 7.976189631720849e+18,
"trial_name": null,
"trial_params": null
}