[Warning] missing keys: {'refine_model.post0.weight', ....

How do I fixed this warning?

[2019-12-02 15:32:54,256-rk0-load_helper.py# 31] load pretrained model from checkpoint_e10.pth
[2019-12-02 15:32:57,294-rk0-load_helper.py# 25] remove prefix 'module.'
[2019-12-02 15:32:57,297-rk0-load_helper.py# 13] [Warning] missing keys: {'refine_model.post0.weight', 'refine_model.v2.0.weight', 'refine_model.post2.bias', 'refine_model.v2.2.bias', 'refine_model.post1.bias', 'refine_model.h2.0.bias', 'refine_model.post1.weight', 'refine_model.v0.2.weight', 'refine_model.post0.bias', 'refine_model.v1.0.weight', 'refine_model.v1.2.bias', 'refine_model.h1.0.weight', 'refine_model.v1.2.weight', 'refine_model.h0.2.bias', 'refine_model.v2.0.bias', 'refine_model.h1.2.bias', 'refine_model.v0.0.weight', 'refine_model.h2.2.bias', 'refine_model.v1.0.bias', 'refine_model.h1.2.weight', 'refine_model.v2.2.weight', 'refine_model.h0.0.bias', 'refine_model.v0.2.bias', 'refine_model.h1.0.bias', 'refine_model.deconv.weight', 'refine_model.h0.0.weight', 'refine_model.v0.0.bias', 'refine_model.post2.weight', 'refine_model.h0.2.weight', 'refine_model.h2.0.weight', 'refine_model.h2.2.weight', 'refine_model.deconv.bias'}
[2019-12-02 15:32:57,297-rk0-load_helper.py# 14] missing keys:32
[2019-12-02 15:32:57,297-rk0-load_helper.py# 18] used keys:324

Start with load_helper.py and checkpoint_e10.pth, then compare the checkpoint keys with the model parameters named under refine_model. Check how the loader reports missing and used keys and whether the checkpoint matches the selected model configuration. Done means the warning's cause is identified and loading completes without unexpected missing parameters.

Written by the indexing model from the issue text.