[torchlib] Fix and improve quantization support #7420
| Job | Run time |
|---|---|
| 2m 26s | |
| 7s | |
| 2m 54s | |
| 16m 59s | |
| 8m 35s | |
| 5m 13s | |
| 2m 5s | |
| 8m 30s | |
| 3m 4s | |
| 8m 23s | |
| 17m 7s | |
| 8m 13s | |
| 16m 24s | |
| 8m 29s | |
| 16m 59s | |
| 8m 29s | |
| 17m 1s | |
| 2m 47s | |
| 3m 15s | |
| 2m 53s | |
| 3m 2s | |
| 9m 41s | |
| 7m 44s | |
| 9m 10s | |
| 9m 5s | |
| 1m 51s | |
| 4m 7s | |
| 1m 45s | |
| 1m 42s | |
| 4m 37s | |
| 2m 59s | |
| 5m 32s | |
| 5m 27s | |
| 24s | |
| 3h 46m 59s |