MobileBERT is an open-weight, compact BERT model published by Google for natural language processing on resource-limited devices. Developers can download and run it with Transformers, PyTorch, or TensorFlow for pretraining and downstream NLP tasks.
In the Other models space, Mobilebert Uncased takes a focused approach. It focuses on running BERT-based natural language processing on resource-limited devices. Mobilebert Uncased is an open-source project aimed at machine learning developers and NLP researchers. Mobilebert Uncased is open source under the Open Source license. It ships for the web and API, and it can be self-hosted.
Google builds and maintains Mobilebert Uncased. Key capabilities include text encoding, pretraining, and transformers 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