model checkpoint size mismatch

[wangpei7001]

/video-translation-main/src/lip_sync_src/facerender/animate.py:199 in load_cpk_mapping 
❱ 199    mapping.load_state_dict(checkpoint["mapping"])  
                  
RuntimeError: Error(s) in loading state_dict for MappingNet:
        size mismatch for first.0.weight: copying a param with shape torch.Size([1024, 73, 7]) from checkpoint, the shape in current 
model is torch.Size([1024, 70, 7]).

How to address this size mismatch when loading the checkpoint?

[samhita-alla]

@wangpei7001 here's the updated code: https://github.com/unionai/unionai-examples/tree/main/tutorials/video_translation. could you give this a try? you should be able to run it on Union serverless: https://signup.union.ai/. let me know if it still doesn't work for you.