W

Whisper Medium French

Developed by pierreguillou
A French speech recognition model fine-tuned on the common_voice_11_0 dataset based on openai/whisper-medium, achieving a standardized WER of 11.1406, outperforming the original model.
Downloads 260
Release Time : 12/6/2022

Model Overview

This model is an automatic speech recognition (ASR) model optimized specifically for French, capable of accurately transcribing French audio into text.

Model Features

High-Performance French Recognition
Achieves a standardized WER of 11.1406 on the Common Voice 11.0 test set, outperforming the original Whisper-medium model.
Optimized Training
Fine-tuned for 5000 steps on the Common Voice 11.0 dataset using linear learning rate scheduling and the Adam optimizer.
Standardized Output
Provides both standardized and non-standardized Word Error Rate (WER) evaluations to ensure transcription results better meet practical application needs.

Model Capabilities

French Speech Recognition
Audio Transcription
Automatic Speech-to-Text

Use Cases

Speech Transcription
French Meeting Minutes
Automatically transcribe French meeting recordings into written records
High accuracy with standardized WER 11.1406
French Media Subtitle Generation
Automatically generate subtitles for French video content
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
© 2025AIbase