Lukas Blecher

Hi @kermitt2, > So the idea with Grobid was rather to identify the area of formulas, and then send the area as bitmap (after cropping/rasterizing) to am equation recognition library...

I've had the same issue. Try adding a pre_tokenizer: ```python from tokenizers.pre_tokenizers import Whitespace tokenizer = Tokenizer(BPE()) tokenizer.pre_tokenizer = Whitespace() trainer = BpeTrainer(vocab_size=280) tokenizer.train(trainer, ["preprocessing/corpus/corpus.txt"]) save_to_filepath = 'preprocessing/tokenizer.json' tokenizer.save(save_to_filepath) tokenizer...

Hello, this has nothing to do with the incompatibility with Blender 2.9. Open a new issue for that. Check out existing answers to your problem since it has nothing to...

Thank you for the contribution. I'm still getting the same error as described in #9 What python version are you using?

Hi, sorry for the late reply. I've got university stuff going on right now, so I won't be able to improve the addon much further atm. The `"TypeError: 'NoneType' object...

There is not much plain math going on in the code I wrote iirc. The only thing you should be afraid of are bad practices

I've replaced the FlowNet2 with LiteFlowNet and it works on windows. check it out https://github.com/lukas-blecher/Deep-Flow-Guided-Video-Inpainting

I can't help you with the flownet2 modules. I've also had problems with them. So I removed them and implemented a similar network in its place: The LiteFlowNet ([pytorch](https://github.com/sniklaus/pytorch-liteflownet), [original](https://github.com/twhui/LiteFlowNet))....

seems like there was missing a keyword argument. Try again with `interpolation=cv2.INTER_NEAREST` or pull the changes.

@yuyang16101066 I've linked them in the README of the fork or in [this issue](https://github.com/lukas-blecher/Deep-Flow-Guided-Video-Inpainting/issues/3#issuecomment-636317366)