Template dequant kernel on output type, add bf16/fp32 native output #2158
| Job | Run time |
|---|---|
| 1m 39s | |
| 49s | |
| 6m 15s | |
| 14s | |
| 15s | |
| 13s | |
| 5m 1s | |
| 4m 39s | |
| 4m 17s | |
| 4m 59s | |
| 4m 23s | |
| 5m 32s | |
| 4m 25s | |
| 4m 12s | |
| 4m 6s | |
| 4m 11s | |
| 4m 15s | |
| 4m 13s | |
| 6m 10s | |
| 6m 11s | |
| 6m 2s | |
| 4m 4s | |
| 6m 11s | |
| 5m 51s | |
| 6m 11s | |
| 3m 38s | |
| 3m 54s | |
| 4m 44s | |
| 4m 3s | |
| 3m 35s | |
| 4m 6s | |
| 5m 56s | |
| 3m 38s | |
| 6m 6s | |
| 4m 4s | |
| 5m 46s | |
| 4m 31s | |
| 4m 49s | |
| 5m 47s | |
| 6m 11s | |
| 5m 40s | |
| 3m 56s | |
| 3m 40s | |
| 3m 40s | |
| 6m 11s | |
| 1s | |
| 1s | |
| 1s | |
| 1s | |
| 3h 18m 17s |