Enable Nvidia's ModelOpt fp8 quantized models #3573
Job | Run time |
---|---|
14m 17s | |
12m 6s | |
5m 46s | |
2m 11s | |
3m 40s | |
1s | |
5m 42s | |
3m 25s | |
11m 52s | |
2m 45s | |
6m 53s | |
10m 54s | |
0s | |
1h 19m 32s |
Job | Run time |
---|---|
14m 17s | |
12m 6s | |
5m 46s | |
2m 11s | |
3m 40s | |
1s | |
5m 42s | |
3m 25s | |
11m 52s | |
2m 45s | |
6m 53s | |
10m 54s | |
0s | |
1h 19m 32s |