after the Image Retrieval training run today, some errors appeared when running subs.py like this:

RuntimeError: the Error (s) in loading state_dict for DataParallel: </ strong>
Missing key (s) in state_dict: “Module. Backbone. 0. Weight”, “module. Backbone. 1. The weight”, “module. Backbone. 1. The bias”, “module. Backbone. 1. Running_mean”, “module. Backbone. 1. Running_var”, “module. Backbone. 4.0. Conv1. Weight”. “Module. Backbone. 4.0. Bn1. Weight”, “module. Backbone. 4.0. Bn1. Bias”, “module. Backbone. 4.0. Bn1. Running_mean”, “module. Backbone. 4.0. Bn1. Running_var”, “module. Backbone. 4.0. Conv2. Weight”. “Module. Backbone. 4.0. Bn2. Weight”, “module. Backbone. 4.0. Bn2. Bias”, “module. Backbone. 4.0. Bn2. Running_mean”, “module. Backbone. 4.0. Bn2. Running_var”, “module. Backbone. 4.0. Conv3. Weight”. “Module. Backbone. 4.0. Bn3. Weight”, “module. Backbone. 4.0. Bn3. Bias”, “module. Backbone. 4.0. Bn3. Running_mean”, “module. Backbone. 4.0. Bn3. Running_var”, “module. Backbone. 4.0. Downsample. 0. Weight”. “Module. Backbone. 4.0. The downsample. 1. The weight”, “module. Backbone. 4.0. The downsample. 1. The bias”, “module. Backbone. 4.0. The downsample. 1. Running_mean”, “module. Backbone. 4.0. The downsample. 1. Running_var”, “Module. Backbone. 4.1. Conv1. Weight”, “module. Backbone. 4.1. Bn1. Weight”, “layer4.2. Bn3. Running_var”, “layer4.2. Bn3. Weight”, “layer4.2. Bn3. Bias”, “fc. Weight”, “fc. Bias,”… And so on.

is incorrectly located to line 95 and traced back to a path not generated by the training. PTH file.

solution:

find their own training after the file path is ok.

</ p>