GGML format conversion of Japanese Whisper models for use with
whisper.cpp and other GGML-compatible inference engines.
Both models are fine-tuned from OpenAI Whisper and achieve strong performance on Japanese ASR tasks, including anime, TEDx, Common Voice, and other benchmarks.
All credit goes to
@efwkjn for training and releasing these models. This repository only performs a lossless format conversion to GGML. If you find these models useful, please visit the original repos and give them a ⭐.
This repository inherits the license of the original models, which are based on
OpenAI Whisper (MIT License).
