Pretrained Model Error: block_1a_conv_1 has incorrect shape in pretrained model

Hi, I am running training with Yolov4 (224, 224) using a pretrained model downloaded with:
wget https://api.ngc.nvidia.com/v2/models/nvidia/tao/pretrained_object_detection/versions/resnet50/files/resnet_50.hdf5

and running into the error:
block_1a_conv_1 has incorrect shape in pretrained model.

any idea on how to solve this?
Thank you!

Can you share your training spec file?

Hi Morganh.
You are correct, I had the wrong number of layers in the training spec! Thank you!

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.