General Introduction
Whisper is a GitHub open source project developed by Const-me focused on high performance inference using GPGPU to implement OpenAI's Whisper automatic speech recognition (ASR) model. This project is released under the MPL-2.0 license, with the latest version 1.12 released on 7/22/2023. In terms of code, it is written primarily in C++, but also includes other languages such as C, C# and HLSL.
Dependent language module download: https://huggingface.co/ggerganov/whisper.cpp/tree/main
Function List
Providing High Performance GPGPU Inference for OpenAI Whisper Models
Supports the latest ASR technology
Sourced from reliable open source communities
Includes rich development language types and version history
Using Help
Download WhisperDesktop.zip from the "Releases" section of this repository, extract the ZIP file, and then run WhisperDesktop.exe.
On the first screen, it will ask you to download a model.
I would recommend using ggml-medium.bin (size 1.42GB) as I use that model most of the time to test software.
Visit the GitHub project for source code and documentation
Project deployment and use according to Readme
Modification and distribution following MPL-2.0 license specifications
Keep track of the latest version updates and discussions via GitHub
Whisper GPGPU Client Download