Fix (gptq): Caching quant_inp values for quant_weight #1278
Job | Run time |
---|---|
4m 29s | |
2m 30s | |
7m 0s | |
4m 40s | |
2m 34s | |
3m 52s | |
4m 8s | |
3m 35s | |
5m 14s | |
4m 23s | |
2m 38s | |
7m 20s | |
4m 19s | |
2m 37s | |
4m 10s | |
3m 16s | |
1m 49s | |
3m 22s | |
3m 19s | |
2m 2s | |
3m 19s | |
3m 16s | |
2m 0s | |
3m 16s | |
3m 2s | |
2m 3s | |
4m 42s | |
3m 5s | |
2m 1s | |
5m 31s | |
1h 49m 32s |