whisper.cpp-wsl-nvidia is a Bash script that automates the setup of Whisper.cpp on WSL Ubuntu with NVIDIA GPU acceleration. This script configures a WSL 2 environment to leverage NVIDIA's GPU capabilities for faster, more efficient performance while running Whisper.cpp, an efficient implementation of OpenAI's Whisper speech offline recognition model.
- Automated Setup: The script simplifies the installation process for setting up Whisper.cpp on WSL Ubuntu.
- GPU Acceleration: Enables NVIDIA GPU support on WSL 2 for running Whisper.cpp, speeding up transcription and other tasks.
- Pre-configured Environment: Installs all necessary dependencies, such as CUDA and cuDNN, to ensure seamless GPU usage with Whisper.cpp.
- Easy-to-Use: Run a single Bash script to configure everything you need to get started with Whisper.cpp on WSL with NVIDIA hardware.
Before running the script, ensure you have:
- WSL 2 installed and configured on your Windows machine.
- An Ubuntu 24.04.1 LTS installation in WSL.
- A compatible NVIDIA GPU and NVIDIA drivers installed on your Windows machine.
- The machine must be running on the x86_64 architecture.
-
Run the setup script:
curl -s https://raw.githubusercontent.com/2g4/whisper.cpp-wsl-nvidia/main/nvidia.sh | bash -
Follow the on-screen instructions to complete the setup. The script will:
- Install dependencies (like build-essential, nvidia-cuda-toolkit, etc.).
- Clone the Whisper.cpp repository. Set up NVIDIA GPU support for Whisper.cpp on WSL 2.
- Download the base.en model.
-
Once complete, you can run Whisper.cpp with GPU acceleration in your WSL Ubuntu environment.
To use Whisper.cpp with NVIDIA GPU acceleration, simply navigate to the Whisper.cpp directory and run the desired commands as you normally would. The GPU will automatically be utilized for model inference and processing, greatly improving performance.
- Ubuntu 24.04.1 LTS.
- NVIDIA GeForce RTX 4080 SUPER and 4090 GPUs.
This repository is licensed under the MIT License.