Transformers: bert summarizer module import error

[eeegnu]

🐛

Running bert summarizer (run_summarization.py) gives the following error

Traceback (most recent call last):
  File "run_summarization.py", line 15, in <module>
    from .utils_summarization import (
ModuleNotFoundError: No module named '__main__.utils_summarization'; '__main__' is not a package

Information

I've managed to fix this issue personally by changing "from .utils_summarization import" to "from utils_summarization import", though I don't know if this is due to a convention change in python module imports.

The problem arises when using:

• [x] the official example scripts: (give details below)

Running the following command yields the error
python3 run_summarization.py --documents_dir ".../bertabs/dataset/input" --summaries_output_dir ".../bertabs/dataset/output" --no_cuda false --batch_size 4 --min_length 50 --max_length 200 --beam_size 5 --alpha 0.95 --block_trigram true

To reproduce

Steps to reproduce the behavior:

1. Followed the steps here https://github.com/huggingface/transformers/blob/5b396457e5035a8b16ddee14b205c098598fe6bb/examples/summarization/bertabs/README.md
   Though I skipped the 'Reproduce the authors' ROUGE score' section, that should not have any effect on usage for different inputs.
2. Created custom data paths for testing a single input article.
3. Run the command given above.

Traceback (most recent call last):
  File "run_summarization.py", line 15, in <module>
    from .utils_summarization import (
ModuleNotFoundError: No module named '__main__.utils_summarization'; '__main__' is not a package

Expected behavior

It should import the module utils_summarization

Environment info

• transformers version: 2.8.0
• Platform: Linux-4.15.0-74-generic-x86_64-with-Ubuntu-18.04-bionic
• Python version: 3.6.9
• PyTorch version (GPU?): 1.4.0 (False)
• Tensorflow version (GPU?): not installed (NA)
• Using GPU in script?: No
• Using distributed or parallel set-up in script?: No

[AppishettySHIVASAI]

HELLO eeegnu, I'm also facing the same issue.

Environment Info :
Platform: Windows 10 64bit
Python version: 3.6.10
PyTorch version (GPU?): 1.4.0
Tensorflow version (GPU?): not installed (NA)

[Nol01]

Hi!
Change: from .utils_summarization import ( CNNDMDataset, build_mask, compute_token_type_ids, encode_for_summarization, truncate_or_pad, )
To: from utils_summarization import ( CNNDMDataset, build_mask, compute_token_type_ids, encode_for_summarization, truncate_or_pad, )

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.