Classification on Higgs (train)
76.6AccuracyAdam
Evaluation Results
| Method | Links | |
|---|---|---|
| AdamArchitecture=4 x 256, Steps (k)=26.0 ± 4.9, Time (s)=26.8 ± 4.7, Time/Step (ms)=1.06 ± 0.022025.06 | 76.6 | |
| AdamArchitecture=4 x 256 + Skip, Steps (k)=22.6 ± 1.7, Time (s)=25.2 ± 1.9, Time/Step (ms)=1.15 ± 0.022025.06 | 76.2 | |
| AdamArchitecture=1 x 256, Steps (k)=29.0 ± 3.9, Time (s)=27.3 ± 3.8, Time/Step (ms)=0.97 ± 0.012025.06 | 74.6 | |
| FAArchitecture=1 x 256, Steps (k)=59.0 ± 18.9, Time (s)=52.1 ± 17.1, Time/Step (ms)=0.89 ± 0.012025.06 | 71.3 | |
| SGDArchitecture=4 x 256, Steps (k)=178.8 ± 37.7, Time (s)=176.0 ± 38.7, Time/Step (ms)=0.98 ± 0.012025.06 | 71.1 | |
| FF (Adam)Architecture=1 x 256, Steps (k)=55.6 ± 12.6, Time (s)=66.0 ± 15.5, Time/Step (ms)=1.18 ± 0.032025.06 | 71 | |
| FF (Adam)Architecture=4 x 256, Steps (k)=33.6 ± 5.1, Time (s)=44.3 ± 7.5, Time/Step (ms)=1.33 ± 0.012025.06 | 69.4 | |
| SGDArchitecture=4 x 256 + Skip, Steps (k)=73.0 ± 10.3, Time (s)=72.4 ± 10.9, Time/Step (ms)=0.99 ± 0.012025.06 | 68.9 | |
| DRArchitecture=1 x 256, Steps (k)=24.8 ± 5.1, Time (s)=0.9 ± 0.2, Time/Step (ms)=0.04 ± 0.002025.06 | 67 | |
| SGDArchitecture=1 x 256, Steps (k)=63.6 ± 11.4, Time (s)=55.5 ± 10.3, Time/Step (ms)=0.88 ± 0.012025.06 | 66.9 | |
| APArchitecture=1 x 256, Steps (k)=27.4 ± 13.5, Time (s)=0.6 ± 0.3, Time/Step (ms)=0.02 ± 0.002025.06 | 65.3 | |
| CPArchitecture=1 x 256, Steps (k)=25.4 ± 10.8, Time (s)=0.7 ± 0.3, Time/Step (ms)=0.03 ± 0.002025.06 | 65.2 | |
| DRArchitecture=4 x 256 + Skip, Steps (k)=10.0 ± 3.6, Time (s)=6.4 ± 2.3, Time/Step (ms)=0.64 ± 0.002025.06 | 62.9 | |
| FAArchitecture=4 x 256, Steps (k)=6.6 ± 2.0, Time (s)=6.1 ± 1.8, Time/Step (ms)=0.92 ± 0.022025.06 | 62.4 | |
| DRArchitecture=4 x 256, Steps (k)=10.8 ± 5.8, Time (s)=5.4 ± 2.9, Time/Step (ms)=0.50 ± 0.002025.06 | 61.1 | |
| CPArchitecture=4 x 256 + Skip, Steps (k)=14.0 ± 10.4, Time (s)=6.7 ± 5.0, Time/Step (ms)=0.48 ± 0.002025.06 | 58.4 | |
| APArchitecture=4 x 256 + Skip, Steps (k)=13.0 ± 10.4, Time (s)=3.9 ± 3.1, Time/Step (ms)=0.30 ± 0.002025.06 | 58.3 | |
| FF (SGD)Architecture=1 x 256, Steps (k)=152.8 ± 191.6, Time (s)=168.1 ± 210.9, Time/Step (ms)=1.07 ± 0.032025.06 | 57.7 | |
| FF (SGD)Architecture=4 x 256, Steps (k)=35.4 ± 19.6, Time (s)=41.4 ± 23.2, Time/Step (ms)=1.16 ± 0.022025.06 | 57.3 | |
| APArchitecture=4 x 256, Steps (k)=3.4 ± 6.8, Time (s)=1.0 ± 2.0, Time/Step (ms)=0.29 ± 0.002025.06 | 51.7 | |
| CPArchitecture=4 x 256, Steps (k)=3.4 ± 6.8, Time (s)=1.9 ± 3.9, Time/Step (ms)=0.57 ± 0.002025.06 | 51.7 |