10 repos
Pre-trained language models and seq2seq architectures optimized for Chinese-language natural language processing tasks. The cluster centers on BART variants (bart-large-chinese, bart-base-chinese), CPT models, and related transformer-based text generation systems designed for summarization, machine translation, and other sequence-to-sequence tasks in Chinese. Most repos are model weights, documentation, and reference implementations rather than frameworks, making this a practical resource for practitioners deploying Chinese NLP rather than a research toolkit cluster.