bert-base-japanese-char-v3 is a BERT-base model pretrained on Japanese text from CC-100 and jawiki using character-level tokenization and whole word masking. It is distributed on Hugging Face with support for the Transformers library. The model is designed for Japanese NLP researchers and developers who require fine-grained character understanding in their language models.
In the Other AI space, Bert Base Japanese Char takes a focused approach. Insufficient high-quality Japanese language models with character-level tokenization for NLP tasks. Bert Base Japanese Char is an open-source project aimed at developers. Bert Base Japanese Char is open source under the Apache-2.0 license. Bert Base Japanese Char is available on the web and API.
Tohoku NLP builds and maintains Bert Base Japanese Char, and it first shipped in 2018. The GitHub repository has been archived. Among its 3 catalogued features are japanese language model, character-level tokenization, and whole word masking.
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