Redo custom attention processor to support other attention types #2294
Job | Run time |
---|---|
4m 36s | |
4m 33s | |
4m 58s | |
4m 29s | |
9m 20s | |
4m 25s | |
5m 45s | |
4m 16s | |
5m 49s | |
10m 22s | |
58m 33s |
Job | Run time |
---|---|
4m 36s | |
4m 33s | |
4m 58s | |
4m 29s | |
9m 20s | |
4m 25s | |
5m 45s | |
4m 16s | |
5m 49s | |
10m 22s | |
58m 33s |