GitHub · Where software is built

fairseq v2
#4493 · cbalioglu opened on Jun 15, 2022
23

Labels Milestones New issue

Question regarding the model size, token limits and model optimizaion?

#5622

· sushanttwayana opened

on Jun 20, 2025

Why am I running nllb-200 on A100 with only 30% GPU utilization? Can you help me solve it

#5621

· hui-li-xf opened

on Jun 17, 2025

Installation fails on Windows: Missing stddef.h while building fairseq C++ extension (libbleu.cpp)

#5620

· pxD1311 opened

on Jun 13, 2025

Using RoBERTa with more than 512 tokens

#5618

· VicenteLavagnino opened

on Jun 5, 2025

payload['cfg'] has no 'model' attribute or it is None.

#5617

· ahmedimaddd opened

on Jun 1, 2025

Add an option to disable appending eos to the end of each document while preprocessing

#5615

· surprisedPikachu007 opened

on May 26, 2025

NLLB-200 & LLama 4: confusion about Nordic languages

#5614

· mindplay-dk opened

on May 16, 2025

How to save the fine-tune checkpoint

#5613

· asrays opened

on May 15, 2025

External torch.hub.load('pytorch/fairseq:main', 'xlmr.large') causes WeightsUnpickler error for PyTorch version 2.6

#5610

· vviitani-amd opened

on May 7, 2025

Wav2Vec2 pipeline feature extractor normalizes input over batch dimension, is it a feature or bug in design?

#5609

· ivan-alles opened

on May 1, 2025

Fine-tuning facebook/mms-tts-{lang_code} model using Fairseq

#5608

· IbratDO opened

on Apr 30, 2025

Question about fine-tuning NLLB for Mi'kmaq language (Indigenous Canadian language)

#5607

· bearsharktopus-dev opened

on Apr 28, 2025