[torchlib] Fix and improve quantization support #7424
| Job | Run time |
|---|---|
| 1m 3s | |
| 48s | |
| 7s | |
| 1m 1s | |
| 1m 4s | |
| 1m 2s | |
| 1m 6s | |
| 1m 1s | |
| 1m 4s | |
| 1m 0s | |
| 1m 5s | |
| 1m 1s | |
| 1m 5s | |
| 1m 1s | |
| 1m 4s | |
| 1m 4s | |
| 1m 4s | |
| 51s | |
| 51s | |
| 51s | |
| 50s | |
| 51s | |
| 50s | |
| 54s | |
| 52s | |
| 1m 7s | |
| 1m 7s | |
| 1m 6s | |
| 1m 3s | |
| 57s | |
| 54s | |
| 1m 8s | |
| 1m 7s | |
| 11s | |
| 32m 10s |