If you're looking to access Google's Gemma 4 models via HuggingFace, you'll need a HuggingFace access token. Here's a quick guide to get you started.
First, sign in to your HuggingFace account at huggingface.co. If you don't have an account, create one for free. Once logged in, go to your profile settings and click on "Access Tokens." You can generate a new token by clicking "New token." Choose a name for your token, select the appropriate permissions (typically "read" for model access), and click "Generate." Copy the token and keep it secure.
With the token, you can authenticate when downloading or using Gemma 4 models. For example, in Python with the transformers library, you can use HuggingFaceHub or set the token via environment variable HUGGINGFACEHUB_API_TOKEN. This token also works with the huggingface_hub library for model downloads.
Remember that some models may require you to accept terms of use on the model's HuggingFace page before access. Always review licensing agreements. With your token ready, you're all set to explore Gemma 4's capabilities.