distil-whisper/distil-large-v3 is an open-source distilled version of the Whisper large-v3 model, optimized for significantly faster automatic speech recognition while maintaining high accuracy. Hosted on Hugging Face, it supports the Transformers library and is suitable for local or edge deployment. It has been evaluated on open ASR leaderboards and is used for efficient transcription tasks.
In the Voice, TTS & speech space, Distil Large takes a focused approach. High computational cost and slow inference speed of large speech recognition models for real-time or on-device transcription needs. Distil Large is an open-source project aimed at developers and ML engineers. The project is open source (BSD-3-Clause). It ships for the web, the command line, and API.
Hugging Face builds and maintains Distil Large, and it first shipped in 2022. The project is developed in the open on GitHub with 24.5k stars and 83 commits in the last 90 days. Among its 3 catalogued features are Speech Recognition, Distilled Model, and Fast Inference.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Same category — not a similarity match