bugczw

Thank a lot! And will the weight parameters obtained from the training of this model be published?

Also, I want to know how many is the input seq_length. What's more, how should I set the config of BertCapModel?

> Try setting config.add_cross_attention=True, I think it is a new thing, thus I didn't have it. Should I change BertConfig.max_position_embeddings?

I have the same problem,but I don't know how to solve. The haskell process log is `("Starting inferior stack GHCi process using stack" "haskell" nil "stack" "ghci" "--ghc-options=-ferror-spans") -> Prelude.putStrLn...

I also encounter the same problem and do't know how to solve. Is there any suitable solution?