mmBERT-base is a pre-trained multimodal BERT model hosted on Hugging Face. It supports fill-mask inference and can be used via the Transformers library or Hugging Face Inference API. The model is intended for researchers and developers working on language and multimodal AI tasks.
In the Multimodal & vision space, mmBERT Base takes a focused approach. It focuses on accessing and using a pre-trained multimodal BERT model for masked language modeling tasks without training from scratch. It is built as an open-source project for machine learning researchers and developers. The project is open source (Open Source). It runs on the web and API.
Johns Hopkins CLSP builds and maintains mmBERT Base. It exposes integrations via a public API.
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