In the documentation, it is mentioned that there is support for all pre-trained models from torchvision. How can I use a backbone model like Inception v3?
What are the possible ways that exist for me to do to use a BNInception backbone? Are there any other options than building the model or loading a checkpoint?
You can use torchvision://inception_v3 to use the pre-trained inception backbone. However, for now, we do not implement inception models in MMCV. We may support this in the future release.
BNInception
backbone? Are there any other options than building the model or loading a checkpoint?Thank you