Chinese embedding use
The record is tagged feature-extraction with a zh marker, and the card presents a Chinese retrieval instruction for passage search.
Open Source Model Profile · BAAI
bge-large-zh
bge-large-zh is a 326M-parameter BERT Chinese embedding model from BAAI. The recorded license is MIT with documented Transformers usage.
- Publisher
- BAAI
- Task
- feature-extraction
- Model type
- bert
- License
- mit
- Library
- transformers
- Publication status
- Approved for indexing
Model overview
bge-large-zh is published by BAAI as a feature-extraction model. The captured configuration identifies BertModel with model type bert, and Safetensors metadata reports 325,522,944 parameters. According to the model card, it is a Chinese BAAI General Embedding model, and the recorded license is mit.
Recorded capabilities
Four-stack usage docs
According to the model card, examples cover FlagEmbedding, Sentence-Transformers, Langchain, and Transformers with [CLS] pooling and normalized embeddings.
RetroMAE plus contrastive training
According to the model card, BAAI pre-trains with RetroMAE on large-scale pairs data using contrastive learning.
v1.5 successor pointer
According to the model card, users are directed to BAAI/bge-large-zh-v1.5 for improved similarity distribution.
Use cases in the source record
- Chinese passage retrieval where queries carry the documented retrieval instruction and scores come from normalized embedding dot products.
- Retrieval-plus-rerank pipelines where a bge embedding model returns top candidates for a bge reranker to reorder, as described in the card.
Limitations and unknowns
- No Ethen-verified evaluation scores were extracted; benchmark tables in the card concern related reranker and v1.5 models.
- No context-window value was extracted from this record.
- Provider state is historical snapshot data and should be refreshed before being presented as current.
Source and provenance
Source: BAAI/bge-large-zh
Captured: Unknown. Processed: 2026-09-07T19:34:29.726930+00:00.
Recommend switching to newest BAAI/bge-large-zh-v1.5 , which has more reasonable similarity distribution and same method of usage. FlagEmbedding Model List | FAQ | Usage | Evaluation | Train | Contact | Citation | License More details please refer to our Github: FlagEmbedding . English | 中文 FlagEmbedding can map any text to a low-dimensional dense vector which can be used for tasks like retrieval, classification, clustering, or semantic search. And it also can be used in vector databases for LLMs. ************* 🌟 Updates 🌟 ************* 10/12/2023: Release LLM-Embedder , a unified embedding model to support diverse retrieval augme…
F001F002F003F004F005F006F007F009F010F011F016F020F024F025F028F029F031F032F034