Make attn_implementation configurable for huggingface models #873
Job | Run time |
---|---|
5m 8s | |
5m 0s | |
4m 54s | |
5m 6s | |
1s | |
1s | |
1s | |
1s | |
4m 36s | |
4m 5s | |
5m 4s | |
3m 55s | |
0s | |
37m 52s |
Job | Run time |
---|---|
5m 8s | |
5m 0s | |
4m 54s | |
5m 6s | |
1s | |
1s | |
1s | |
1s | |
4m 36s | |
4m 5s | |
5m 4s | |
3m 55s | |
0s | |
37m 52s |