LeBenchmark

AI & ML interests

We release models related to speech processing in the French language. In particular, we developed a benchmark to compare Self-supervised models including E2E and HMM-DNN Automatic Speech Recognition and Spoken Language Understanding.

Recent Activity

mzboito  authored a paper about 1 month ago

From TOWER to SPIRE: Adding the Speech Modality to a Text-Only LLM

mzboito  new activity 3 months ago

LeBenchmark/wav2vec2-FR-2.6K-base:Adding `safetensors` variant of this model

mzboito  new activity 3 months ago

LeBenchmark/wav2vec2-FR-1K-base:Adding `safetensors` variant of this model

View all activity

Team members 17