From 4a6024921fa142f28e8d0034ae28693713b3bfd0 Mon Sep 17 00:00:00 2001 From: Vaibhav Srivastav Date: Wed, 5 Jun 2024 12:56:11 +0200 Subject: [PATCH] doc: add info about wav2vec2 bert in older wav2vec2 models. (#31120) * doc: add info about wav2vec2 bert in older wav2vec2 models. * apply suggestions from review. * forward contrib credits from review --------- Co-authored-by: Sanchit Gandhi --- docs/source/en/model_doc/wav2vec2-conformer.md | 2 ++ docs/source/en/model_doc/wav2vec2.md | 2 ++ docs/source/en/model_doc/xlsr_wav2vec2.md | 2 ++ 3 files changed, 6 insertions(+) diff --git a/docs/source/en/model_doc/wav2vec2-conformer.md b/docs/source/en/model_doc/wav2vec2-conformer.md index c32c03bb0cb7ac..0b30cf5fa43145 100644 --- a/docs/source/en/model_doc/wav2vec2-conformer.md +++ b/docs/source/en/model_doc/wav2vec2-conformer.md @@ -27,6 +27,8 @@ The Wav2Vec2-Conformer weights were released by the Meta AI team within the [Fai This model was contributed by [patrickvonplaten](https://huggingface.co/patrickvonplaten). The original code can be found [here](https://github.com/pytorch/fairseq/tree/main/examples/wav2vec). +Note: Meta (FAIR) released a new version of [Wav2Vec2-BERT 2.0](https://huggingface.co/docs/transformers/en/model_doc/wav2vec2-bert) - it's pretrained on 4.5M hours of audio. We especially recommend using it for fine-tuning tasks, e.g. as per [this guide](https://huggingface.co/blog/fine-tune-w2v2-bert). + ## Usage tips - Wav2Vec2-Conformer follows the same architecture as Wav2Vec2, but replaces the *Attention*-block with a *Conformer*-block diff --git a/docs/source/en/model_doc/wav2vec2.md b/docs/source/en/model_doc/wav2vec2.md index c573db69c4d9e5..5ef3fdbb1eaa66 100644 --- a/docs/source/en/model_doc/wav2vec2.md +++ b/docs/source/en/model_doc/wav2vec2.md @@ -33,6 +33,8 @@ recognition with limited amounts of labeled data.* This model was contributed by [patrickvonplaten](https://huggingface.co/patrickvonplaten). +Note: Meta (FAIR) released a new version of [Wav2Vec2-BERT 2.0](https://huggingface.co/docs/transformers/en/model_doc/wav2vec2-bert) - it's pretrained on 4.5M hours of audio. We especially recommend using it for fine-tuning tasks, e.g. as per [this guide](https://huggingface.co/blog/fine-tune-w2v2-bert). + ## Usage tips - Wav2Vec2 is a speech model that accepts a float array corresponding to the raw waveform of the speech signal. diff --git a/docs/source/en/model_doc/xlsr_wav2vec2.md b/docs/source/en/model_doc/xlsr_wav2vec2.md index d1b5444c2469bd..6369d068850a26 100644 --- a/docs/source/en/model_doc/xlsr_wav2vec2.md +++ b/docs/source/en/model_doc/xlsr_wav2vec2.md @@ -36,6 +36,8 @@ XLSR-53, a large model pretrained in 53 languages.* The original code can be found [here](https://github.com/pytorch/fairseq/tree/master/fairseq/models/wav2vec). +Note: Meta (FAIR) released a new version of [Wav2Vec2-BERT 2.0](https://huggingface.co/docs/transformers/en/model_doc/wav2vec2-bert) - it's pretrained on 4.5M hours of audio. We especially recommend using it for fine-tuning tasks, e.g. as per [this guide](https://huggingface.co/blog/fine-tune-w2v2-bert). + ## Usage tips - XLSR-Wav2Vec2 is a speech model that accepts a float array corresponding to the raw waveform of the speech signal.