Feat (gptq): optimizing CPU to GPU memory transfer #934
Job | Run time |
---|---|
2m 13s | |
1m 29s | |
1m 56s | |
1m 29s | |
2m 24s | |
1m 55s | |
2m 16s | |
1m 35s | |
2m 23s | |
1m 38s | |
2m 17s | |
1m 37s | |
2m 18s | |
1m 42s | |
2m 5s | |
1m 50s | |
2m 12s | |
1m 37s | |
2m 19s | |
1m 46s | |
2m 13s | |
1m 45s | |
2m 18s | |
1m 41s | |
2m 32s | |
1m 52s | |
2m 41s | |
2m 7s | |
56m 10s |