Closed Leiyi-Hu closed 1 week ago
Sorry for our negligence, you can find the files here. https://drive.google.com/file/d/1IK0H4_pm85oGE7Dm9DXwEKhD2G6WY6J0/view?usp=sharing
Thanks for the files you provided! I found that the models.visual_encoder is not implemented in inference_sdf.py, and I wonder how to output the predictions by our own trained model and evaluate (It seems that there is no document about this evaluation pipeline in test.ipynb)?
We have just updated the 'inference_sdf.py' to solve that question. To evaluate your output, you should first use the 'inference_sdf.py' to generate midi output for the videos and then use the 'test.ipynb' and your generated midi to evaluate.
Hey, Thanks for your wonderful work, but when I run the training code, I got the Error "FileNotFoundError: [Errno 2] No such file or directory: 'data/train_split_pnt/pop909_new.pickle'. I could not find the file in the urls of data that you have provided, could you provide this file? Thanks!