sam1120's picture
Training in progress, step 120
176aade
raw
history blame
21.7 kB
{
"best_metric": 0.8488205671310425,
"best_model_checkpoint": "/robodata/smodak/Projects/nspl/scripts/terrainseg/training/models/safety-utcustom-train-SF30-RGB-b0/checkpoint-120",
"epoch": 60.0,
"global_step": 120,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.5,
"learning_rate": 7.5e-07,
"loss": 1.1677,
"step": 1
},
{
"epoch": 1.0,
"learning_rate": 1.5e-06,
"loss": 1.1594,
"step": 2
},
{
"epoch": 1.5,
"learning_rate": 2.25e-06,
"loss": 1.1676,
"step": 3
},
{
"epoch": 2.0,
"learning_rate": 3e-06,
"loss": 1.1627,
"step": 4
},
{
"epoch": 2.5,
"learning_rate": 3.75e-06,
"loss": 1.1604,
"step": 5
},
{
"epoch": 3.0,
"learning_rate": 4.5e-06,
"loss": 1.1658,
"step": 6
},
{
"epoch": 3.5,
"learning_rate": 5.2500000000000006e-06,
"loss": 1.1583,
"step": 7
},
{
"epoch": 4.0,
"learning_rate": 6e-06,
"loss": 1.1687,
"step": 8
},
{
"epoch": 4.5,
"learning_rate": 6.75e-06,
"loss": 1.1569,
"step": 9
},
{
"epoch": 5.0,
"learning_rate": 7.5e-06,
"loss": 1.1527,
"step": 10
},
{
"epoch": 5.0,
"eval_accuracy_safe": 0.7703764811759412,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.14652629277186402,
"eval_iou_safe": 0.030663997549793695,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.14642944151194123,
"eval_loss": 1.1084816455841064,
"eval_mean_accuracy": 0.4584513869739026,
"eval_mean_iou": 0.059031146353911636,
"eval_overall_accuracy": 0.16635894775390625,
"eval_runtime": 1.4751,
"eval_samples_per_second": 5.423,
"eval_steps_per_second": 0.678,
"step": 10
},
{
"epoch": 5.5,
"learning_rate": 8.25e-06,
"loss": 1.1521,
"step": 11
},
{
"epoch": 6.0,
"learning_rate": 9e-06,
"loss": 1.1492,
"step": 12
},
{
"epoch": 6.5,
"learning_rate": 8.960526315789474e-06,
"loss": 1.1456,
"step": 13
},
{
"epoch": 7.0,
"learning_rate": 8.921052631578947e-06,
"loss": 1.1433,
"step": 14
},
{
"epoch": 7.5,
"learning_rate": 8.881578947368421e-06,
"loss": 1.1379,
"step": 15
},
{
"epoch": 8.0,
"learning_rate": 8.842105263157895e-06,
"loss": 1.1431,
"step": 16
},
{
"epoch": 8.5,
"learning_rate": 8.80263157894737e-06,
"loss": 1.1302,
"step": 17
},
{
"epoch": 9.0,
"learning_rate": 8.763157894736842e-06,
"loss": 1.1379,
"step": 18
},
{
"epoch": 9.5,
"learning_rate": 8.723684210526316e-06,
"loss": 1.1259,
"step": 19
},
{
"epoch": 10.0,
"learning_rate": 8.68421052631579e-06,
"loss": 1.1326,
"step": 20
},
{
"epoch": 10.0,
"eval_accuracy_safe": 0.9694915254237289,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.24695220149698446,
"eval_iou_safe": 0.04189444073553146,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.24695195825231817,
"eval_loss": 1.1090760231018066,
"eval_mean_accuracy": 0.6082218634603567,
"eval_mean_iou": 0.09628213299594988,
"eval_overall_accuracy": 0.26992225646972656,
"eval_runtime": 1.192,
"eval_samples_per_second": 6.711,
"eval_steps_per_second": 0.839,
"step": 20
},
{
"epoch": 10.5,
"learning_rate": 8.644736842105263e-06,
"loss": 1.1296,
"step": 21
},
{
"epoch": 11.0,
"learning_rate": 8.605263157894737e-06,
"loss": 1.1125,
"step": 22
},
{
"epoch": 11.5,
"learning_rate": 8.56578947368421e-06,
"loss": 1.1224,
"step": 23
},
{
"epoch": 12.0,
"learning_rate": 8.526315789473683e-06,
"loss": 1.1105,
"step": 24
},
{
"epoch": 12.5,
"learning_rate": 8.486842105263159e-06,
"loss": 1.108,
"step": 25
},
{
"epoch": 13.0,
"learning_rate": 8.447368421052632e-06,
"loss": 1.1265,
"step": 26
},
{
"epoch": 13.5,
"learning_rate": 8.407894736842106e-06,
"loss": 1.1047,
"step": 27
},
{
"epoch": 14.0,
"learning_rate": 8.36842105263158e-06,
"loss": 1.115,
"step": 28
},
{
"epoch": 14.5,
"learning_rate": 8.328947368421052e-06,
"loss": 1.092,
"step": 29
},
{
"epoch": 15.0,
"learning_rate": 8.289473684210526e-06,
"loss": 1.0981,
"step": 30
},
{
"epoch": 15.0,
"eval_accuracy_safe": 0.992200389980501,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.40550470282425555,
"eval_iou_safe": 0.053544543556037626,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.40550470282425555,
"eval_loss": 1.0979702472686768,
"eval_mean_accuracy": 0.6988525464023783,
"eval_mean_iou": 0.15301641546009773,
"eval_overall_accuracy": 0.42415618896484375,
"eval_runtime": 1.213,
"eval_samples_per_second": 6.595,
"eval_steps_per_second": 0.824,
"step": 30
},
{
"epoch": 15.5,
"learning_rate": 8.25e-06,
"loss": 1.0895,
"step": 31
},
{
"epoch": 16.0,
"learning_rate": 8.210526315789473e-06,
"loss": 1.0659,
"step": 32
},
{
"epoch": 16.5,
"learning_rate": 8.171052631578947e-06,
"loss": 1.0868,
"step": 33
},
{
"epoch": 17.0,
"learning_rate": 8.131578947368421e-06,
"loss": 1.079,
"step": 34
},
{
"epoch": 17.5,
"learning_rate": 8.092105263157896e-06,
"loss": 1.0884,
"step": 35
},
{
"epoch": 18.0,
"learning_rate": 8.052631578947368e-06,
"loss": 1.0499,
"step": 36
},
{
"epoch": 18.5,
"learning_rate": 8.013157894736842e-06,
"loss": 1.0726,
"step": 37
},
{
"epoch": 19.0,
"learning_rate": 7.973684210526316e-06,
"loss": 1.0718,
"step": 38
},
{
"epoch": 19.5,
"learning_rate": 7.934210526315789e-06,
"loss": 1.0653,
"step": 39
},
{
"epoch": 20.0,
"learning_rate": 7.894736842105263e-06,
"loss": 1.086,
"step": 40
},
{
"epoch": 20.0,
"eval_accuracy_safe": 0.9926503674816259,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.5103054348671892,
"eval_iou_safe": 0.06444250128047581,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.5103054348671892,
"eval_loss": 1.0822405815124512,
"eval_mean_accuracy": 0.7514779011744075,
"eval_mean_iou": 0.19158264538255498,
"eval_overall_accuracy": 0.525639533996582,
"eval_runtime": 1.1977,
"eval_samples_per_second": 6.679,
"eval_steps_per_second": 0.835,
"step": 40
},
{
"epoch": 20.5,
"learning_rate": 7.855263157894737e-06,
"loss": 1.0638,
"step": 41
},
{
"epoch": 21.0,
"learning_rate": 7.815789473684211e-06,
"loss": 1.0491,
"step": 42
},
{
"epoch": 21.5,
"learning_rate": 7.776315789473685e-06,
"loss": 1.058,
"step": 43
},
{
"epoch": 22.0,
"learning_rate": 7.736842105263158e-06,
"loss": 1.0102,
"step": 44
},
{
"epoch": 22.5,
"learning_rate": 7.697368421052632e-06,
"loss": 1.0523,
"step": 45
},
{
"epoch": 23.0,
"learning_rate": 7.657894736842106e-06,
"loss": 1.0807,
"step": 46
},
{
"epoch": 23.5,
"learning_rate": 7.618421052631579e-06,
"loss": 1.0437,
"step": 47
},
{
"epoch": 24.0,
"learning_rate": 7.578947368421053e-06,
"loss": 1.0586,
"step": 48
},
{
"epoch": 24.5,
"learning_rate": 7.539473684210527e-06,
"loss": 1.036,
"step": 49
},
{
"epoch": 25.0,
"learning_rate": 7.5e-06,
"loss": 1.0466,
"step": 50
},
{
"epoch": 25.0,
"eval_accuracy_safe": 0.9901604919754012,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.5916703521626885,
"eval_iou_safe": 0.0760597426269381,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.5916525776602614,
"eval_loss": 1.0541445016860962,
"eval_mean_accuracy": 0.7909154220690449,
"eval_mean_iou": 0.2225707734290665,
"eval_overall_accuracy": 0.6043386459350586,
"eval_runtime": 1.2115,
"eval_samples_per_second": 6.603,
"eval_steps_per_second": 0.825,
"step": 50
},
{
"epoch": 25.5,
"learning_rate": 7.460526315789474e-06,
"loss": 1.0143,
"step": 51
},
{
"epoch": 26.0,
"learning_rate": 7.421052631578948e-06,
"loss": 1.0743,
"step": 52
},
{
"epoch": 26.5,
"learning_rate": 7.381578947368421e-06,
"loss": 1.0126,
"step": 53
},
{
"epoch": 27.0,
"learning_rate": 7.342105263157895e-06,
"loss": 1.0943,
"step": 54
},
{
"epoch": 27.5,
"learning_rate": 7.302631578947368e-06,
"loss": 1.0216,
"step": 55
},
{
"epoch": 28.0,
"learning_rate": 7.263157894736842e-06,
"loss": 1.0104,
"step": 56
},
{
"epoch": 28.5,
"learning_rate": 7.223684210526317e-06,
"loss": 1.0259,
"step": 57
},
{
"epoch": 29.0,
"learning_rate": 7.18421052631579e-06,
"loss": 1.0378,
"step": 58
},
{
"epoch": 29.5,
"learning_rate": 7.144736842105263e-06,
"loss": 1.001,
"step": 59
},
{
"epoch": 30.0,
"learning_rate": 7.1052631578947375e-06,
"loss": 1.0533,
"step": 60
},
{
"epoch": 30.0,
"eval_accuracy_safe": 0.9862906854657267,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.6471891895618873,
"eval_iou_safe": 0.08607076662094097,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.6470951758858802,
"eval_loss": 1.024947166442871,
"eval_mean_accuracy": 0.816739937513807,
"eval_mean_iou": 0.2443886475022737,
"eval_overall_accuracy": 0.6579694747924805,
"eval_runtime": 1.2639,
"eval_samples_per_second": 6.33,
"eval_steps_per_second": 0.791,
"step": 60
},
{
"epoch": 30.5,
"learning_rate": 7.065789473684211e-06,
"loss": 1.0201,
"step": 61
},
{
"epoch": 31.0,
"learning_rate": 7.026315789473684e-06,
"loss": 0.9958,
"step": 62
},
{
"epoch": 31.5,
"learning_rate": 6.986842105263158e-06,
"loss": 1.0084,
"step": 63
},
{
"epoch": 32.0,
"learning_rate": 6.9473684210526315e-06,
"loss": 1.0021,
"step": 64
},
{
"epoch": 32.5,
"learning_rate": 6.907894736842106e-06,
"loss": 0.9748,
"step": 65
},
{
"epoch": 33.0,
"learning_rate": 6.86842105263158e-06,
"loss": 1.0624,
"step": 66
},
{
"epoch": 33.5,
"learning_rate": 6.828947368421053e-06,
"loss": 0.9931,
"step": 67
},
{
"epoch": 34.0,
"learning_rate": 6.7894736842105264e-06,
"loss": 1.0179,
"step": 68
},
{
"epoch": 34.5,
"learning_rate": 6.75e-06,
"loss": 1.0053,
"step": 69
},
{
"epoch": 35.0,
"learning_rate": 6.710526315789474e-06,
"loss": 0.9779,
"step": 70
},
{
"epoch": 35.0,
"eval_accuracy_safe": 0.9770661466926653,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.687357977071454,
"eval_iou_safe": 0.09514260217010605,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.6870926801315042,
"eval_loss": 1.0009846687316895,
"eval_mean_accuracy": 0.8322120618820597,
"eval_mean_iou": 0.26074509410053676,
"eval_overall_accuracy": 0.6965680122375488,
"eval_runtime": 1.3535,
"eval_samples_per_second": 5.91,
"eval_steps_per_second": 0.739,
"step": 70
},
{
"epoch": 35.5,
"learning_rate": 6.671052631578947e-06,
"loss": 1.0052,
"step": 71
},
{
"epoch": 36.0,
"learning_rate": 6.6315789473684205e-06,
"loss": 0.9521,
"step": 72
},
{
"epoch": 36.5,
"learning_rate": 6.592105263157895e-06,
"loss": 0.9952,
"step": 73
},
{
"epoch": 37.0,
"learning_rate": 6.552631578947369e-06,
"loss": 1.0184,
"step": 74
},
{
"epoch": 37.5,
"learning_rate": 6.513157894736842e-06,
"loss": 0.9709,
"step": 75
},
{
"epoch": 38.0,
"learning_rate": 6.473684210526316e-06,
"loss": 0.959,
"step": 76
},
{
"epoch": 38.5,
"learning_rate": 6.4342105263157896e-06,
"loss": 0.9497,
"step": 77
},
{
"epoch": 39.0,
"learning_rate": 6.394736842105263e-06,
"loss": 1.0239,
"step": 78
},
{
"epoch": 39.5,
"learning_rate": 6.355263157894737e-06,
"loss": 0.9677,
"step": 79
},
{
"epoch": 40.0,
"learning_rate": 6.31578947368421e-06,
"loss": 0.9161,
"step": 80
},
{
"epoch": 40.0,
"eval_accuracy_safe": 0.9634618269086546,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.7338942182201074,
"eval_iou_safe": 0.10909770285762813,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.7333943281692789,
"eval_loss": 0.9695309400558472,
"eval_mean_accuracy": 0.848678022564381,
"eval_mean_iou": 0.28083067700896897,
"eval_overall_accuracy": 0.7411923408508301,
"eval_runtime": 1.2732,
"eval_samples_per_second": 6.284,
"eval_steps_per_second": 0.785,
"step": 80
},
{
"epoch": 40.5,
"learning_rate": 6.276315789473684e-06,
"loss": 0.9493,
"step": 81
},
{
"epoch": 41.0,
"learning_rate": 6.236842105263159e-06,
"loss": 0.9665,
"step": 82
},
{
"epoch": 41.5,
"learning_rate": 6.197368421052632e-06,
"loss": 0.9623,
"step": 83
},
{
"epoch": 42.0,
"learning_rate": 6.157894736842105e-06,
"loss": 0.9901,
"step": 84
},
{
"epoch": 42.5,
"learning_rate": 6.118421052631579e-06,
"loss": 0.9523,
"step": 85
},
{
"epoch": 43.0,
"learning_rate": 6.078947368421053e-06,
"loss": 0.9034,
"step": 86
},
{
"epoch": 43.5,
"learning_rate": 6.039473684210526e-06,
"loss": 0.9353,
"step": 87
},
{
"epoch": 44.0,
"learning_rate": 6e-06,
"loss": 1.0231,
"step": 88
},
{
"epoch": 44.5,
"learning_rate": 5.9605263157894735e-06,
"loss": 0.9537,
"step": 89
},
{
"epoch": 45.0,
"learning_rate": 5.921052631578948e-06,
"loss": 0.9843,
"step": 90
},
{
"epoch": 45.0,
"eval_accuracy_safe": 0.9493925303734814,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.7768244190295703,
"eval_iou_safe": 0.1253679077487413,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.7759008194701261,
"eval_loss": 0.940314531326294,
"eval_mean_accuracy": 0.8631084747015259,
"eval_mean_iou": 0.3004229090729558,
"eval_overall_accuracy": 0.7823104858398438,
"eval_runtime": 1.3253,
"eval_samples_per_second": 6.036,
"eval_steps_per_second": 0.755,
"step": 90
},
{
"epoch": 45.5,
"learning_rate": 5.881578947368422e-06,
"loss": 0.9646,
"step": 91
},
{
"epoch": 46.0,
"learning_rate": 5.842105263157895e-06,
"loss": 0.9498,
"step": 92
},
{
"epoch": 46.5,
"learning_rate": 5.802631578947368e-06,
"loss": 0.9407,
"step": 93
},
{
"epoch": 47.0,
"learning_rate": 5.7631578947368425e-06,
"loss": 0.9684,
"step": 94
},
{
"epoch": 47.5,
"learning_rate": 5.723684210526316e-06,
"loss": 0.8967,
"step": 95
},
{
"epoch": 48.0,
"learning_rate": 5.684210526315789e-06,
"loss": 1.0137,
"step": 96
},
{
"epoch": 48.5,
"learning_rate": 5.644736842105263e-06,
"loss": 0.9542,
"step": 97
},
{
"epoch": 49.0,
"learning_rate": 5.6052631578947374e-06,
"loss": 0.8673,
"step": 98
},
{
"epoch": 49.5,
"learning_rate": 5.565789473684211e-06,
"loss": 0.9138,
"step": 99
},
{
"epoch": 50.0,
"learning_rate": 5.526315789473685e-06,
"loss": 0.9568,
"step": 100
},
{
"epoch": 50.0,
"eval_accuracy_safe": 0.9190940452977351,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.8135083196994605,
"eval_iou_safe": 0.14123580174066971,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.8116862567811934,
"eval_loss": 0.9070925116539001,
"eval_mean_accuracy": 0.8663011824985978,
"eval_mean_iou": 0.3176406861739544,
"eval_overall_accuracy": 0.8168649673461914,
"eval_runtime": 1.2557,
"eval_samples_per_second": 6.371,
"eval_steps_per_second": 0.796,
"step": 100
},
{
"epoch": 50.5,
"learning_rate": 5.486842105263158e-06,
"loss": 0.9128,
"step": 101
},
{
"epoch": 51.0,
"learning_rate": 5.4473684210526315e-06,
"loss": 0.8691,
"step": 102
},
{
"epoch": 51.5,
"learning_rate": 5.407894736842106e-06,
"loss": 0.9109,
"step": 103
},
{
"epoch": 52.0,
"learning_rate": 5.368421052631579e-06,
"loss": 0.8425,
"step": 104
},
{
"epoch": 52.5,
"learning_rate": 5.328947368421052e-06,
"loss": 0.895,
"step": 105
},
{
"epoch": 53.0,
"learning_rate": 5.289473684210527e-06,
"loss": 0.9433,
"step": 106
},
{
"epoch": 53.5,
"learning_rate": 5.2500000000000006e-06,
"loss": 0.9005,
"step": 107
},
{
"epoch": 54.0,
"learning_rate": 5.210526315789474e-06,
"loss": 0.8744,
"step": 108
},
{
"epoch": 54.5,
"learning_rate": 5.171052631578947e-06,
"loss": 0.8999,
"step": 109
},
{
"epoch": 55.0,
"learning_rate": 5.131578947368421e-06,
"loss": 0.8443,
"step": 110
},
{
"epoch": 55.0,
"eval_accuracy_safe": 0.8742462876856157,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.8570024260249537,
"eval_iou_safe": 0.1671743332931789,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.8537115815119297,
"eval_loss": 0.8626915812492371,
"eval_mean_accuracy": 0.8656243568552847,
"eval_mean_iou": 0.3402953049350362,
"eval_overall_accuracy": 0.8575506210327148,
"eval_runtime": 1.4008,
"eval_samples_per_second": 5.711,
"eval_steps_per_second": 0.714,
"step": 110
},
{
"epoch": 55.5,
"learning_rate": 5.092105263157895e-06,
"loss": 0.93,
"step": 111
},
{
"epoch": 56.0,
"learning_rate": 5.052631578947368e-06,
"loss": 0.8241,
"step": 112
},
{
"epoch": 56.5,
"learning_rate": 5.013157894736842e-06,
"loss": 0.919,
"step": 113
},
{
"epoch": 57.0,
"learning_rate": 4.973684210526316e-06,
"loss": 0.8839,
"step": 114
},
{
"epoch": 57.5,
"learning_rate": 4.9342105263157895e-06,
"loss": 0.8664,
"step": 115
},
{
"epoch": 58.0,
"learning_rate": 4.894736842105264e-06,
"loss": 0.8774,
"step": 116
},
{
"epoch": 58.5,
"learning_rate": 4.855263157894737e-06,
"loss": 0.8846,
"step": 117
},
{
"epoch": 59.0,
"learning_rate": 4.81578947368421e-06,
"loss": 0.931,
"step": 118
},
{
"epoch": 59.5,
"learning_rate": 4.7763157894736844e-06,
"loss": 0.8527,
"step": 119
},
{
"epoch": 60.0,
"learning_rate": 4.736842105263158e-06,
"loss": 0.8765,
"step": 120
},
{
"epoch": 60.0,
"eval_accuracy_safe": 0.8591420428978551,
"eval_accuracy_unlabeled": NaN,
"eval_accuracy_unsafe": 0.8658727336661936,
"eval_iou_safe": 0.17292191207636712,
"eval_iou_unlabeled": 0.0,
"eval_iou_unsafe": 0.8619790415433206,
"eval_loss": 0.8488205671310425,
"eval_mean_accuracy": 0.8625073882820243,
"eval_mean_iou": 0.34496698453989594,
"eval_overall_accuracy": 0.8656587600708008,
"eval_runtime": 1.4312,
"eval_samples_per_second": 5.59,
"eval_steps_per_second": 0.699,
"step": 120
}
],
"max_steps": 240,
"num_train_epochs": 120,
"total_flos": 2.1035001839616e+16,
"trial_name": null,
"trial_params": null
}