Feat (gptq): optimizing CPU to GPU memory transfer (#1009) #1738
| Job | Run time |
|---|---|
| 1m 59s | |
| 1m 32s | |
| 1m 51s | |
| 1m 59s | |
| 1m 25s | |
| 1m 45s | |
| 1m 24s | |
| 1m 36s | |
| 1m 52s | |
| 1m 58s | |
| 1m 56s | |
| 1m 35s | |
| 1m 53s | |
| 1m 34s | |
| 2m 1s | |
| 2m 15s | |
| 1m 42s | |
| 1m 46s | |
| 1m 24s | |
| 1m 29s | |
| 1m 53s | |
| 2m 9s | |
| 1m 33s | |
| 1m 41s | |
| 1m 26s | |
| 1m 28s | |
| 1m 59s | |
| 2m 7s | |
| 1m 32s | |
| 1m 43s | |
| 2m 4s | |
| 1m 27s | |
| 1m 59s | |
| 2m 10s | |
| 1m 39s | |
| 1m 45s | |
| 1m 29s | |
| 2m 6s | |
| 1m 57s | |
| 1m 34s | |
| 1m 58s | |
| 2m 12s | |
| 1m 39s | |
| 1m 43s | |
| 53s | |
| 53s | |
| 1m 56s | |
| 1m 55s | |
| 1m 38s | |
| 1m 52s | |
| 53s | |
| 1m 8s | |
| 2m 5s | |
| 2m 1s | |
| 1m 32s | |
| 1m 56s | |
| 51s | |
| 54s | |
| 1m 55s | |
| 2m 10s | |
| 1m 34s | |
| 2m 7s | |
| 56s | |
| 56s | |
| 2m 0s | |
| 2m 10s | |
| 1m 50s | |
| 1m 54s | |
| 55s | |
| 59s | |
| 2m 12s | |
| 2m 22s | |
| 1m 38s | |
| 1m 50s | |
| 58s | |
| 58s | |
| 2h 8m 0s |