ZHANG45

Hi Han8931. I think doublehead model is trained to solve two tasks: seq2seq and seq2label. You can check the code in their transformer. In `interact.py`, we can't predict the label,...

Hi, Hannabrahman. I think you can edit the functions `build_input_from_segments`, and `MODEL_INPUTS` in file `train.py`. And add your classification in the function `update`

ekQ Thank you for your reply! I used the size you suggested and now I can get at most **65.16** F0.5 score. I am considering whether there still has some...

Hi! Jmallins and casaro. Thank you for your valuable information. I tried the setting you mentioned, but my valid loss was still larger than 0.6. So I checked the clang8...

ekQ, thank you for your reply! I used the default flag vaue ` --tokenize_text=True` as you set in `run.sh` I also checked the file `targets/clang8_en.detokenized.tsv`, and I found some sentences...

ekQ and Jmallins , Thank you for your reply and suggestions. I can achieve 72.31 F0.5 score on bea test data by my reproduced T5-large. So I think my high...

Hi, @mrqorib. I am sorry that I reproduced this work by fairseq.

@mrqorib This is setting I used. But you need to edit the save_path, data_dir and arch by yourself. Additionally, you need to add a new arch hf_t5 according to this...