fairseq
https://github.com/facebookresearch/fairseq
Python
Facebook AI Research Sequence-to-Sequence Toolkit written in Python.
Triage Issues!
When you volunteer to triage issues, you'll receive an email each day with a link to an open issue that needs help in this project. You'll also receive instructions on how to triage issues.
Triage Docs!
Receive a documented method or class from your favorite GitHub repos in your inbox every day. If you're really pro, receive undocumented methods or classes and supercharge your commit history.
Python not yet supported14 Subscribers
Add a CodeTriage badge to fairseq
Help out
- Issues
- Can BART and/or RoBERTa support running on CPU?
- Typo in wav2vec.py (ConvAggegator)
- Inference using fairseq-generate producing just 200 chars and truncating the recognition for longer wav files
- Providing Models in SafeTensors format
- Issue about doing MMS ASR inference with multiple GPUs
- About gradient accumulation
- How to train a multilingual model on parallel Hindi to English corpus in both directions and also non-parallel data in English and Hindi? For the non-parallel corpora, I need to train the model to copy the source to the target by masking x% tokens in the source. Is it possible to do all these tasks in single training?
- raise InvalidRequirement(str(e)) from e pkg_resources.extern.packaging.requirements.InvalidRequirement: Expected closing RIGHT_PARENTHESIS PyYAML (>=5.1.*) ~~~~~~^
- Update dump_mfcc_feature.py
- data2vec2.0 finetune maxtokens can only set to 320000 in A100, is there any problem?
- Docs
- Python not yet supported