search

amirgholami / ZeroQ

[CVPR'20] ZeroQ: A Novel Zero Shot Quantization Framework
275 stars 56 forks source link

The accuracy of pytorch official model is 0.3% lower than full precision #14

Open talenz opened 3 years ago

talenz commented 3 years ago

I've used the pytorch official resnet50 (https://pytorch.org/docs/stable/torchvision/models.html) in your ZeroQ, and it only gives 75.85 (full precision is 76.13)

All I do is change the model loading line in uniform_test.py as follows

model = ptcv_get_model(args.model, pretrained=True) model = torchvision.models.resnet50(pretrained=True)

Any hypermeters are allowed to be tuned for better performance?

jiangaojie commented 2 years ago

I've used the pytorch official resnet50 (https://pytorch.org/docs/stable/torchvision/models.html) in your ZeroQ, and it only gives 75.85 (full precision is 76.13)

All I do is change the model loading line in uniform_test.py as follows

model = ptcv_get_model(args.model, pretrained=True) model = torchvision.models.resnet50(pretrained=True)

Any hypermeters are allowed to be tuned for better performance?

I meet the same issue . Do you slove this problem? Thanks