Runqi Yang comments

Results 43 comments of


                                            Runqi Yang

why the best score can't get 89.2 when I use quora.json5 to train re2?

Did you download the Quora dataset from the [link](https://drive.google.com/file/d/0B0PlTAo--BnaQWlsZl9FZ3l1c28/view?usp=sharing) in ReadMe and prepare the data using "prepare_quora.py"?

why the best score can't get 89.2 when I use quora.json5 to train re2?

Then that's strange... What's your pytorch version? I'll rerun the experiment later

why the best score can't get 89.2 when I use quora.json5 to train re2?

@TheMnBN Hi, sorry for the late reply. In pytorch 1.5, the return type of `torch.nn.utils.clip_grad_norm_` has change from "float" to "torch.Tensor". Change line 82 of model.py to ```python 'gnorm': grad_norm.item(),...

the accuracy

https://github.com/alibaba-edu/simple-effective-text-matching/blob/master/configs/data/quora.json5

Wrong with checkpoint file

Same error with tf1.8rc0, py2 ``` NotFoundError (see above for traceback): Key lstm/basic_lstm_cell/bias not found in checkpoint [[Node: save/RestoreV2 = RestoreV2[dtypes=[DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, ..., DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT,...

data preprocess

+1. I would like to know what tagger is used for POS tagging, Spacy, NLTK or Stanford NLP?

data preprocess

Thanks for sharing! ❤️

Cant do "bash"

I haven't tested my code on Windows. I've no idea what's happening. I'll reply to you maybe next week after I try my code on a windows machine.

Cant do "bash"

Bash is only for Linux-based systems. Maybe you need WSL or something like that to make the script work.

Cant do "bash"

Glad to see that. :)