prajjwal1/bert-medium is a compact pretrained BERT model distributed through Hugging Face for natural language processing tasks. Developers and researchers can load it with Transformers and run it locally for text representation and inference.
In the Other AI space, Bert Medium takes a focused approach. It focuses on using a compact pretrained language model for natural language processing without training a model from scratch. It is built as an open-source project for machine learning developers and NLP researchers. The project is open source (Apache-2.0). It ships for the web and API, and it can be self-hosted.
Behind Bert Medium is prajjwal1, and it first shipped in 2018. The GitHub repository has been archived. Among its 6 catalogued features are text embeddings, NLI classification, and pyTorch support.
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