Hugging Face is an innovative company leading the way in machine learning and AI, focused on natural language processing (NLP), and best known for its famous transformers library. Hugging Face is open source under the Apache-2.0 license, which means it’s free and can also be used commercially. That has encouraged collaboration and knowledge sharing across the global community and made the community very strong.
Many LLMs today are released on Hugging Face, so if you want to play with LLMs, learning how to use Hugging Face is a must.
Step 1. Find the repository you want to download
For example, I want to download a Chinese conversational model fine-tuned from Llama 2.
Step 2. Click “Files and versions”
Step 3. Download
There are two ways to download a model: download directly, or use Git clone. It depends on your needs and the nature of your project. If you only need part of the model or specific files, downloading directly is a good choice. If you plan to use it long term or need the full repository, Git clone is recommended.
Method a). Click to download
Good if you only want one specific file: just click to download it.
Method b). Git clone
Use git to clone the repository. Click “Clone repository” to see how to download it.
Just enter the commands above to download it easily:
git lfs install
git clone https://huggingface.co/hfl/chinese-llama-2-7bYou can choose to use GIT_LFS_SKIP_SMUDGE=1. This downloads only the small files; large files stored with git lfs are saved as text pointer files instead. So when there are many large files and you want to choose which to download, or network transfer is a concern, you can use this approach:
- First clone the repository as before
- Manually pull the files you want.
--include="*.bin"means download all .bin files; change it to suit your needs
git lfs install
GIT_LFS_SKIP_SMUDGE=1 git clone https://huggingface.co/hfl/chinese-llama-2-7b
git lfs pull --include="*.bin"Error
If you get the following error when cloning:
fatal: 'lfs' appears to be a git command, but we were not
able to execute it. Maybe git-lfs is broken?you need to install git-lfs. The commands below will get it working:
sudo apt install git-lfs
git lfs installWith these steps, you can successfully download the models you need from Hugging Face. Hugging Face not only offers a huge library of models; it has built a strong community that supports knowledge sharing and collaboration. Whether you’re a researcher, a developer or an enthusiast interested in AI, learning to use Hugging Face effectively will greatly improve your productivity and creativity in machine learning.
Support
If this article helped you, or you’d like to encourage me to keep writing, you can clap for it or buy me a coffee through the link below. Thank you for your support!