Give example on how to handle gradient accumulation with cross-entropy #5283
| Job | Run time |
|---|---|
| 1m 48s | |
| 3m 12s | |
| 2m 29s | |
| 2m 23s | |
| 2m 20s | |
| 2m 27s | |
| 2m 22s | |
| 4m 2s | |
| 3m 34s | |
| 9m 57s | |
| 2m 46s | |
| 4m 11s | |
| 3m 45s | |
| 3m 40s | |
| 3m 27s | |
| 3m 26s | |
| 3m 22s | |
| 5m 10s | |
| 4m 29s | |
| 10m 37s | |
| 1h 19m 27s |
| Job | Run time |
|---|---|
| 1m 48s | |
| 3m 12s | |
| 2m 29s | |
| 2m 23s | |
| 2m 20s | |
| 2m 27s | |
| 2m 22s | |
| 4m 2s | |
| 3m 34s | |
| 9m 57s | |
| 2m 46s | |
| 4m 11s | |
| 3m 45s | |
| 3m 40s | |
| 3m 27s | |
| 3m 26s | |
| 3m 22s | |
| 5m 10s | |
| 4m 29s | |
| 10m 37s | |
| 1h 19m 27s |