# BERT architecture optimization

M2-BERT-128 is an 80-million-parameter retrieval model checkpoint proposed in the paper 'Benchmarking and Building Long-Context Retrieval Models with LoCo and M2-BERT'

SimCSE model based on IndoBERT for generating semantic embedding vectors of Indonesian sentences

Indonesian text classification model fine-tuned based on indolem/indobert-base-uncased, achieving 79.54% accuracy on the evaluation set

Provides POS tagging functionality for ancient Chinese, with training data covering four historical periods of Chinese language development.

A Russian BERT model jointly trained by Sber AI team and MLSA Lab at Moscow State University's AI Institute, specializing in scientific text processing

This is a sentence-transformers based model that maps Vietnamese address sentences and paragraphs into a 768-dimensional dense vector space, primarily used for address standardization tasks.

A Spanish BERT model based on sentence-transformers for generating 768-dimensional vector representations of sentences and paragraphs

Legal-HeBERT is a BERT model tailored for Hebrew legal and legislative domains, aimed at advancing Hebrew legal natural language processing research and tool development.

BERT base model trained on Japanese Wikipedia dataset, suitable for masked language modeling tasks in Japanese text

An Arabic pre - trained language model based on the BERT architecture, supporting multiple Arabic NLP tasks

This model is a fine-tuned version of the Italian pre-trained model bert-base-italian-xxl-uncased on a specific dataset, suitable for Italian text processing tasks.

A Russian short-text sentiment classification model based on the RuBERT architecture, supporting neutral, positive, and negative sentiment classification

The state-of-the-art Hebrew language model based on the BERT architecture

Provides Traditional Chinese transformers models and natural language processing tools

SloBERTa is a monolingual BERT-like model specifically optimized for the Slovenian language, developed based on the Camembert architecture.

Chinese pre-trained model based on BigBird architecture, supporting 4096-length context processing

AraBERTMo is an Arabic pre-trained language model based on the BERT architecture, supporting fill-mask tasks.