Biogpt is a GPT-style causal language model hosted on Hugging Face under the microsoft organization. It belongs to the class of foundation models and was pretrained specifically on biomedical literature for text generation tasks in that domain.
The model can be loaded and run through the Transformers library using either a high-level pipeline for text generation or direct instantiation of AutoTokenizer and AutoModelForCausalLM. It supports device_map auto for hardware placement. Inference can also be performed by serving the model with vLLM, which starts an OpenAI-compatible API server accessible via HTTP POST requests to a local completions endpoint.
Biogpt carries an MIT license. It is distributed as open model weights on the Hugging Face hub, where users can access files, versions, and community contributions. Notebooks for Google Colab and Kaggle are referenced for experimentation, and local deployment is supported through the listed libraries and vLLM installation via pip.
The repository lists it under English language with a Transformers and PyTorch implementation. No pricing information appears because the model weights and code are provided under the open MIT license for free use, modification, and deployment.
Biogpt is a Text generation project. It focuses on generating and understanding high-quality text in the biomedical domain. It is built as an open-source project for biomedical researchers. The project is open source (Open Source). It ships for the web, the command line, and API.
It is developed by Microsoft (United States).
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do