MEscriva/french-education-speech

French Education Speech - Transcribed Dataset High-quality French educational speech dataset transcribed with OpenAI Whisper API, prepared for training automatic speech recognition (ASR) models. Dataset Summary This dataset contains 3,933 transcribed audio segments from the French educational domain, totaling approximately 12.82 hours of audio. All transcriptions were performed using OpenAI Whisper API (optimized Whisper-1 model) to ensure maximum accuracy… See the full description on the dataset page: https://huggingface.co/datasets/MEscriva/french-education-speech.