The steps to install and configure the gpt-oss-recipes repository are as follows:
- clone warehouse: by ordering
git clone https://github.com/huggingface/gpt-oss-recipes.gitClone the repository to local. - Creating a Virtual Environment: Creating a virtual environment with Python 3.11 is recommended.
uvtool, the command isuv venv gpt-oss --python 3.11, and activate the environment. - Installation of dependencies: Install the necessary packages such as PyTorch and Transformers after upgrading pip with the command
uv pip install torch==2.8.0 torchvision torchaudio --index-url https://download.pytorch.org/whl/test/cu128cap (a poem)uv pip install -U transformers accelerateThe - Optional installation of Triton kernel: If the hardware supports MXFP4 quantization (e.g., H100 or RTX 50xx), it can be quantized with the command
uv pip install git+https://github.com/triton-lang/triton.git@main#subdirectory=python/triton_kernelsInstallation.
Once you have completed the above steps, you can start using the scripts and tutorials in the repository.
This answer comes from the articleCollection of scripts and tutorials for fine-tuning OpenAI GPT OSS modelsThe
May not be reproduced without permission:AI productivity tools " How to install and configure the environment for the gpt-oss-recipes repository?
