Request a model addition in the Hugging Face collection on Microsoft Foundry

At the moment the Hugging Face collection on Microsoft Foundry and Azure Machine Learning contains +11,000 open models from the Hugging Face Hub, leveraging open-source inference solutions such as Text Generation Inference (TGI), vLLM, SGLang, Text Embeddings Inference (TEI), Transformers, Diffusers, Sentence Transformers, among many others.

Alternatively, you can also open an issue with the model or models you’d like to see on the Hugging Face collection on Microsoft Foundry and Azure Machine Learning.

Before requesting a model addition, you need to make sure that the model or models match the following criteria:

• Have any of the Transformers, Diffusers or Sentence-Transformers tags on the Hugging Face Hub, meaning that the model architecture is compatible with any of those. If the model you’d like to see in the collection doesn’t match this criteria, you can maybe check the Contributing a new model to Transformers guide on how to add new modular-based model architectures into Transformers.

• Make sure that the pipeline_tag tag i.e., the task, is any of the supported tasks as per Microsoft Foundry - Supported Tasks. If it’s not there nor listed among the upcoming tasks, then feel free to open an issue requesting the support for that task. Again, as per the point above, it needs to be a Transformers, Diffusers, or Sentence-Transformers compatible task.

• If you want to benefit from the production-like inference solutions and the OpenAI-compatible interfaces for text generation and embeddings, you should also make sure that the given model has either the text-generation-inference (shortened as tgi), vllm and/or sglang tags, or text-embeddings-inference (shortened as tei), respectively.

• There might be some cases where neither the “Deploy” nor “Deploy on Microsoft Foundry” buttons are enabled for a given model, in that case feel free to open an issue sharing the URL to the model on the Hugging Face Hub, so that we can help fix the metadata if applicable.

• The Hugging Face Hub models MUST be public (including gated models), as private models won’t be considered at the moment.

• The Hugging Face Hub models with trust_remote_code or custom_code tags are not allowed for security reasons, unless manually verified or coming from an already verified organization e.g. microsoft.

• And finally, the model weights need to be in Safetensors format and have passed the JFrog, ClamAV, and the rest of the security checks performed in the Hugging Face Hub. More information in Malware scanning.

Update on GitHub