# OFA-Sys/Chinese-CLIP

Chinese version of CLIP which achieves Chinese cross-modal retrieval and representation generation.

Repository: https://github.com/OFA-Sys/Chinese-CLIP
Canonical: https://ross.abutalabs.com/products/chinese-clip
Language: Jupyter Notebook
License: MIT
License Family: permissive
Topics: chinese, computer-vision, multi-modal-learning, nlp, pytorch, vision-and-language-pre-training, image-text-retrieval, clip, pretrained-models, vision-language, deep-learning, multi-modal, contrastive-loss, transformers, coreml-models
Last push: 2026-03-31T13:27:44+00:00

## Health v2 (maintenance only)
Score: 66/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 75, release rhythm 35, longevity 100
- inputs: {"age_days": 1517, "days_push": 155, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 5998, forks 552 (observed 2026-08-28T04:09:34.130783+00:00)

## What it is
Chinese-CLIP is a Chinese version of the CLIP model trained on ~200 million Chinese image-text pairs, built on open_clip. It provides APIs, training, and evaluation code for cross-modal retrieval, image-text feature extraction and similarity, and zero-shot image classification.

## Use cases
- compute image and text embeddings and similarity in Chinese
- build Chinese image-text cross-modal search
- zero-shot image classification with Chinese labels
- finetune a CLIP model on Chinese image-text data
- deploy CLIP models with ONNX, TensorRT, or CoreML

## When to choose
- you need Chinese-language vision-language embeddings or retrieval
- you want pretrained CLIP variants from RN50 to ViT-H-14 with training code
- you need zero-shot classification for Chinese datasets

## When to avoid
- your application is English-only - use standard OpenAI CLIP or open_clip
- you need a non-PyTorch training framework
- you need a lightweight model without GPU resources

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, image-processing, nlp, search-engine
- domain: machine-learning, computer-vision, deep-learning, large-language-models
- platform: python, cross-platform
- tags: clip, vision-language, multimodal, image-text-retrieval, chinese, contrastive-learning, pretrained-models, zero-shot-classification, pytorch, coreml, natural-language-processing, gpu

## Member repositories
- OFA-Sys/Chinese-CLIP (main) score 66

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:34.130783+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:49:49.574741+00:00, confidence not recorded.
  - readme: https://github.com/OFA-Sys/Chinese-CLIP (fetched 2026-08-28T04:09:34.130783+00:00, sha 1f4b23fbc747)
- Data as of 2026-08-30T08:39:29.467469+00:00.