aws / aws/amazon-sagemaker-examples
fairseq_translation issues
- Dominant language
- Jupyter Notebook
- Stars
- 11k
- Forks
- 7k
- Avg merge
- 8h 29m
- Merged PRs (30d)
- 8
Description
https://github.com/awslabs/amazon-sagemaker-examples/blob/master/advanced_functionality/fairseq_translation/
All of the notebooks are broken.
Issues:
1. Depends on this file, which is no longer there. https://github.com/pytorch/fairseq/blob/master/preprocess.py
1. All repeat the same preprocessing steps with the dataset. This should be one setup notebook instead.
Contributor guide
Research direction
Start in the advanced_functionality/fairseq_translation/ directory and inspect each notebook's preprocessing steps and its dependency on fairseq's preprocess.py. Confirm which notebooks are broken, then define the shared setup notebook and verify that the remaining notebooks no longer duplicate preprocessing and still work with the available fairseq workflow.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- aws, jupyter-notebook, pytorch
- Domain
- machine-learning
- Issue type
- Refactor
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 30/100