Image Classification on CIFAR-10 OOD DEEPNETS-1M (test)
88.9Accuracy (WIDE)SGD
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| SGD#upd+=25x10^3, training_duration=50 epochs2021.10 | 88.9 | 84.5 | 87.3 | 45.6 | 93.5 | 75.7 | |
| Adam#upd+=25x10^3, training_duration=50 epochs2021.10 | 82 | 76.2 | 84.8 | 38.8 | 91.5 | 79.4 | |
| SGD#upd+=2.5x10^3, training_duration=5 epochs2021.10 | 71.2 | 56.7 | 67.8 | 29 | 78.2 | 52.5 | |
| GHN-2#upd+=12021.10 | 64 | 60.5 | 65.8 | 36.8 | 58.6 | 11.4 | |
| SGD#upd+=0.5x10^3, training_duration=1 epoch2021.10 | 47.2 | 34.2 | 45.3 | 18 | 61.8 | 34.5 | |
| GHN-1#upd+=12021.10 | 43.1 | 48.3 | 51.8 | 13.7 | 19.2 | 18.2 | |
| MLP#upd+=12021.10 | 22.3 | 37.9 | 44.8 | 23.9 | 17.7 | 10 |