Using abort_callback from ggml to interrupt llama computation #9209
Job | Run time |
---|---|
6m 11s | |
1m 29s | |
3m 39s | |
20s | |
21s | |
3m 32s | |
17s | |
23s | |
5m 18s | |
1s | |
5m 29s | |
1m 44s | |
2m 35s | |
2m 20s | |
2m 18s | |
6m 55s | |
5m 26s | |
2m 32s | |
16m 43s | |
5m 21s | |
4m 42s | |
22m 4s | |
5m 11s | |
9m 8s | |
16m 25s | |
11m 26s | |
16m 11s | |
6m 15s | |
3m 35s | |
5m 22s | |
4m 46s | |
4m 23s | |
4m 16s | |
0s | |
3h 6m 38s |