whisper-jax
Whisper model
An optimized implementation of OpenAI's Whisper Model for speech recognition and speech-to-text tasks using JAX.
JAX implementation of OpenAI's Whisper model for up to 70x speed-up on TPU.
4k stars
43 watching
385 forks
Language: Jupyter Notebook
last commit: 8 months ago
Linked from 1 awesome list
deep-learningjaxspeech-recognitionspeech-to-textwhisper
sindresorhus/awesome-whisperRelated projects:
| Repository | Description | Stars |
|---|---|---|
 ggerganov/whisper.cpp | A high-performance implementation of the OpenAI Whisper ASR model in C++ | 35,706 |
 systran/faster-whisper | A fast speech recognition system built on top of the CTranslate2 transformer model | 12,506 |
 openai/whisper | A general-purpose speech recognition system trained on large-scale weak supervision | 71,257 |
 const-me/whisper | An implementation of OpenAI's Whisper ASR model using DirectCompute for GPGPU inference | 8,460 |
 huggingface/distil-whisper | A machine learning model that uses audio input to generate text transcriptions at high speeds and with good accuracy. | 3,613 |
 vaibhavs10/insanely-fast-whisper | A command-line tool for fast audio transcription using the Whisper AI model | 7,731 |
 softcatala/whisper-ctranslate2 | A Whisper client compatible with the CTranslate2 model, providing faster transcription and translation capabilities than OpenAI Whisper | 914 |
 jax-ml/jax | Accelerates numerical computing by automatically differentiating and compiling Python functions for high-performance execution on GPUs and TPUs. | 30,499 |
 purfview/whisper-standalone-win | Executable standalone versions of Whisper and Faster-Whisper speech recognition tools | 1,326 |
 mahmoudashraf97/whisper-diarization | Automates speaker diarization from audio recordings using OpenAI Whisper ASR and additional neural networks. | 3,718 |
 m-bain/whisperx | An automatic speech recognition system with word-level timestamps and speaker diarization. | 12,489 |
 xenova/whisper-web | An open-source speech recognition system built using machine learning models and JavaScript. | 2,578 |
 xai-org/grok-1 | An implementation of a Mixture of Experts (MoE) model with large parameters and specialized features for natural language processing tasks. | 49,566 |
 arthurfdlr/whisper-youtube | Transcribes Youtube videos using OpenAI's Whisper speech recognition model | 362 |
 illyism/openai-whisper-api | An OpenAI speech-to-text API service built with Node.js and Typescript, running on Docker and Google Cloud Run. | 108 |