mozilla-ai
/

whisper-small-fr

Automatic Speech Recognition

Eval Results (legacy)

Model card Files Files and versions

Metrics Training metrics Community

Finetuned openai/whisper-small on 20000 French training audio samples from mozilla-foundation/common_voice_17_0.

This model was created from the Mozilla.ai Blueprint: speech-to-text-finetune.

Evaluation results on 5000 audio samples of French:

Baseline model (before finetuning) on French

Word Error Rate: 30.304
Loss: 1.155

Finetuned model (after finetuning) on French

Word Error Rate: 23.511
Loss: 0.449

Downloads last month: 40

Safetensors

Model size

0.2B params

Tensor type

F32

·

Model tree for mozilla-ai/whisper-small-fr

Base model

openai/whisper-small

Finetuned

(3783)

this model

Dataset used to train mozilla-ai/whisper-small-fr

Collection including mozilla-ai/whisper-small-fr

Common Voice Whisper

Whisper models finetuned on the Common Voice dataset using https://github.com/mozilla-ai/speech-to-text-finetune • 9 items • Updated Jul 8, 2025 • 2

Evaluation results

wer on Common Voice (French)
self-reported

23.511