Feat (gptq): optimizing CPU to GPU memory transfer #933
Job | Run time |
---|---|
2m 6s | |
1m 22s | |
2m 13s | |
1m 28s | |
2m 17s | |
1m 23s | |
2m 17s | |
1m 55s | |
2m 13s | |
1m 46s | |
2m 29s | |
1m 44s | |
2m 19s | |
1m 32s | |
2m 2s | |
1m 25s | |
2m 11s | |
1m 32s | |
2m 10s | |
1m 37s | |
2m 8s | |
1m 37s | |
2m 27s | |
1m 56s | |
2m 59s | |
1m 49s | |
2m 33s | |
1m 35s | |
55m 5s |