Dear author,
Thank you for your precious blogs and repo. I am learning about quantization and your tutorial is very helpful. However, I just wonder why we need to fuse this module. # Fuse the model in place rather manually. fused_model = torch.quantization.fuse_modules(fused_model, [["conv1", "bn1", "relu"]], inplace=True) for module_name, module in fused_model.named_children(): if "layer" in module_name: for basic_block_name, basic_block in module.named_children(): torch.quantization.fuse_modules(basic_block, [["conv1", "bn1", "relu1"], ["conv2", "bn2"]], inplace=True) for sub_block_name, sub_block in basic_block.named_children(): if sub_block_name == "downsample": torch.quantization.fuse_modules(sub_block, [["0", "1"]], inplace=True). Can you address it for me. Thank you
Dear author, Thank you for your precious blogs and repo. I am learning about quantization and your tutorial is very helpful. However, I just wonder why we need to fuse this module.
# Fuse the model in place rather manually. fused_model = torch.quantization.fuse_modules(fused_model, [["conv1", "bn1", "relu"]], inplace=True) for module_name, module in fused_model.named_children(): if "layer" in module_name: for basic_block_name, basic_block in module.named_children(): torch.quantization.fuse_modules(basic_block, [["conv1", "bn1", "relu1"], ["conv2", "bn2"]], inplace=True) for sub_block_name, sub_block in basic_block.named_children(): if sub_block_name == "downsample": torch.quantization.fuse_modules(sub_block, [["0", "1"]], inplace=True). Can you address it for me. Thank you