fairseq

fairseq copied to clipboard

Please Fix dependency issues?

2

# Before submitting - [ ] Was this discussed/approved via a Github issue? (no need for typos, doc improvements) - [ x] Did you read the [contributor guideline](https://github.com/pytorch/fairseq/blob/main/CONTRIBUTING.md)? - [x...

d8ahazard

## 🚀 Feature Request Add an option to disable appending EOS to the end of each document while preprocessing ### Motivation There are instances where adding EOS is not preferred...

surprisedPikachu007

I've been testing LLama 4 (both Maverick and Scout) and the both consistently mistranslate certain Danish words from English. For example, Llama 4, when asked to translate from English to...

mindplay-dk

I'm running ``` fairseq_cli/hydra_train.py ``` to fine tune hubert model. Script is running but its not saving the checkpoint. Am I missing something ? My base_10h.yaml looks like this ```...

asrays

Questions about the data used when training Hubert

1

When I start training using examples/hubert/README.md, first perform data preparation. Then it is said that the format of {train,valid}.tsv is.： ``` ... ``` In my understanding, that is.  But...

ghost

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

5

Hello NLLB/fairseq team, I'm interested in exploring how to fine-tune the NLLB model to support the Mi'kmaq language (also spelled Mi'gmaq), an Eastern Algonquian language spoken by approximately 11,000 people...

bearsharktopus-dev

Pt2.6 compatibility

2

# Before submitting - [x ] Was this discussed/approved via a Github issue? (no need for typos, doc improvements) - [ x] Did you read the [contributor guideline](https://github.com/pytorch/fairseq/blob/main/CONTRIBUTING.md)? - [...

vviitani-amd

## 🐛 Bug External torch.hub.load('pytorch/fairseq:main', 'xlmr.large') causes WeightsUnpickler error for PyTorch version >= 2.6. The error occurs in the torch.load() calls in the file fairseq/checkpoint_utils.py. The error appears in PyTorch...

vviitani-amd

I'm tryining to undestand the intuition of the input normalization using layer norm like this: `waveforms = nn.functional.layer_norm(waveforms, waveforms.shape)` [link](https://github.com/pytorch/audio/blob/d60ce09e2c532d5bf2e05619e700ab520543465e/src/torchaudio/pipelines/_wav2vec2/utils.py#L53) If the input is [B, L], this code will normalize...

ivan-alles

I am going to fine tune the facebook/mms-tts-{lang_code} model. 1. Is it possible to fine tune these models? 2. if yes, how to fine tune? what tools do I need?...

IbratDO