Hubert Xlarge Ll60k

🚀 Hubert-Extra-Large

This is an extra large model pretrained on 16kHz sampled speech audio. It offers a powerful solution for various speech - related downstream tasks. When using this model, ensure your speech input is also sampled at 16kHz. And remember, this model needs to be fine - tuned on downstream tasks like Automatic Speech Recognition, Speaker Identification, Intent Classification, Emotion Recognition, etc.

✨ Features

• Pretrained on Large - Scale Data: The model was pretrained on [Libri - Light](https://github.com/facebookresearch/libri - light), which provides a solid foundation for speech representation learning.
• Addressing Speech Learning Challenges: As described in the paper, it tackles unique problems in self - supervised speech representation learning, such as multiple sound units in each utterance, lack of input sound unit lexicon during pre - training, and variable sound unit lengths without explicit segmentation.

📚 Documentation

Model Source

This model is based on [Facebook's Hubert](https://ai.facebook.com/blog/hubert - self - supervised - representation - learning - for - speech - recognition - generation - and - compression). The original model can be found under https://github.com/pytorch/fairseq/tree/master/examples/hubert.

Paper Details

• Title: Paper
• Authors: Wei - Ning Hsu, Benjamin Bolte, Yao - Hung Hubert Tsai, Kushal Lakhotia, Ruslan Salakhutdinov, Abdelrahman Mohamed

Abstract

Self - supervised approaches for speech representation learning are challenged by three unique problems: (1) there are multiple sound units in each input utterance, (2) there is no lexicon of input sound units during the pre - training phase, and (3) sound units have variable lengths with no explicit segmentation. To deal with these three problems, we propose the Hidden - Unit BERT (HuBERT) approach for self - supervised speech representation learning, which utilizes an offline clustering step to provide aligned target labels for a BERT - like prediction loss. A key ingredient of our approach is applying the prediction loss over the masked regions only, which forces the model to learn a combined acoustic and language model over the continuous inputs. HuBERT relies primarily on the consistency of the unsupervised clustering step rather than the intrinsic quality of the assigned cluster labels. Starting with a simple k - means teacher of 100 clusters, and using two iterations of clustering, the HuBERT model either matches or improves upon the state - of - the - art wav2vec 2.0 performance on the Librispeech (960h) and Libri - light (60,000h) benchmarks with 10min, 1h, 10h, 100h, and 960h fine - tuning subsets. Using a 1B parameter model, HuBERT shows up to 19% and 13% relative WER reduction on the more challenging dev - other and test - other evaluation subsets.

💻 Usage Examples

See [this blog](https://huggingface.co/blog/fine - tune - wav2vec2 - english) for more information on how to fine - tune the model. Note that the class Wav2Vec2ForCTC has to be replaced by HubertForCTC.

📄 License

This project is licensed under the Apache 2.0 license.

⚠️ Important Note

When using the model, make sure that your speech input is sampled at 16kHz. Also, this model should be fine - tuned on a downstream task.