Hello,
I tried the command:
python quantize.py --model resnet18 --block-size-cv 9 --block-size-pw 4 --n-centroids-cv 256 --n-centroids-pw 256 --n-centroids-fc 2048 --data-path YOUR_IMAGENET_PATH
However, after i got state_dict_compressed.pth, i run inference with that file, i received the accuracy is only 0.1
I installed the same version of torch, torchvision in requirements.txt
I got the correct accuracy for your file (resnet18_small_blocks.pth)
Could you check it?
Thanks you!
Hello, I tried the command: python quantize.py --model resnet18 --block-size-cv 9 --block-size-pw 4 --n-centroids-cv 256 --n-centroids-pw 256 --n-centroids-fc 2048 --data-path YOUR_IMAGENET_PATH However, after i got state_dict_compressed.pth, i run inference with that file, i received the accuracy is only 0.1 I installed the same version of torch, torchvision in requirements.txt I got the correct accuracy for your file (resnet18_small_blocks.pth) Could you check it? Thanks you!