Closed Jzjerry closed 1 month ago
Hi,
I didn't find any code related to batch norm layers in the inference code, and I'd like to know how they are handled. Are they fused with Conv2D layers?
Currently, the batch normalization layers are fused with the convolutional layers before proceeding with fine-tuning the model's parameters.
Got it, thanks!
Hi,
I didn't find any code related to batch norm layers in the inference code, and I'd like to know how they are handled. Are they fused with Conv2D layers?