search

getalp / disambiguate

Disambiguate is a tool for training and using state of the art neural WSD models
https://arxiv.org/abs/1905.05677
MIT License
58 stars 17 forks source link

Another max_size = sequences[0].size() IndexError: list index out of range ERROR #11

Closed MarshallCharles closed 4 years ago

MarshallCharles commented 4 years ago

I have experienced the same error twice in a row now when trying to train the ELMO model on my computer.

As I understand, this problem has been specified as an issue on the github but the solution proposed did not work for me.

Here's what I did:

I ran the ./prepare_data.sh and pointed it to an empty directory called ./new_model with the --data_path argument, and used the --train argument to point it to the semcor.xml corpus. Also I used the --dir argument to point it towards the semcor.xml file.

I then ran ./train.sh, and used the --data_path flag, pointing it to the ./new_model directory mentioned previously. As well as with the --model_path argument, and pointed it to an empty directory called ./new_trained_model.

I keep getting the same error. Are you able to guide me at all on this?

Thanks

Here's the error:

    pad_batch_x(batch_x, input_clear_text)
  File "/Users/CharlesMarshall/Desktop/m2s1/nlp/practical/WSD/disambiguate-master/python/getalp/wsd/common.py", line 158, in pad_batch_x
    batch_x[i] = pad_sequence(batch_x[i], batch_first=True)
  File "/Users/CharlesMarshall/Desktop/m2s1/nlp/practical/WSD/disambiguate-master/env/lib/python3.7/site-packages/torch/nn/utils/rnn.py", line 327, in pad_sequence
    max_size = sequences[0].size()
IndexError: list index out of range