whisper-finetune and whisper-prep
These are complementary tools designed to be used sequentially: whisper-prep handles the upstream data preparation stage, while whisper-finetune consumes that prepared data to perform the actual model fine-tuning.
About whisper-finetune
i4Ds/whisper-finetune
This repository contains code for fine-tuning the Whisper speech-to-text model.
This project helps machine learning engineers and researchers adapt the Whisper speech-to-text model for specific audio environments or accents. You provide an existing Whisper model and specialized audio datasets, and the project outputs a refined Whisper model that performs better on your unique data. It's designed for professionals working with speech recognition models.
About whisper-prep
i4Ds/whisper-prep
Data preparation utility for the finetuning of OpenAI's Whisper model.
This tool helps AI engineers, machine learning practitioners, and data scientists prepare audio and transcript data for training speech-to-text models like OpenAI's Whisper. It takes raw audio files, sentence-level datasets, or existing SRT/VTT transcripts and outputs meticulously segmented, cleaned, and formatted datasets ready for model fine-tuning. This ensures high-quality training data, leading to better performing transcription models.
Scores updated daily from GitHub, PyPI, and npm data. How scores work