Feat (gptq): optimizing CPU to GPU memory transfer #1611
| Job | Run time |
|---|---|
| 4m 2s | |
| 3m 22s | |
| 4m 16s | |
| 3m 44s | |
| 4m 28s | |
| 3m 38s | |
| 5m 12s | |
| 4m 18s | |
| 4m 23s | |
| 3m 5s | |
| 4m 17s | |
| 3m 26s | |
| 2h 22m 32s | |
| 1h 23m 37s | |
| 4m 0s | |
| 3m 36s | |
| 4m 16s | |
| 3m 35s | |
| 4m 26s | |
| 3m 31s | |
| 4m 59s | |
| 4m 25s | |
| 4m 16s | |
| 3m 28s | |
| 4m 13s | |
| 3m 22s | |
| 2h 22m 37s | |
| 1h 24m 25s | |
| 9h 9m 29s |