Neura Market
ChatYuan logo

ChatYuan

Free

1.9k

FreeFree tier
Inputs: textOutputs: text
Type
Open Source
Company
clue-ai
Visit Website GitHub (40 stars)

About ChatYuan

ChatYuan is an open-source large language model for dialogue developed by clue-ai, supporting both Chinese and English. The latest version, ChatYuan-large-v2, is designed as a functional dialogue model in the ChatGPT style, with lightweight architecture that enables inference on consumer-grade GPUs, PCs, and even mobile devices (INT4 quantization requires only 400MB). It builds on v1 with enhanced capabilities including improved contextual Q&A, creative writing, code generation, table generation, basic mathematical operations, and a refusal mechanism for harmful questions. The model supports a maximum token length of 4096 and can simulate various scenarios. It is available for free under an open-source license with online demos on Hugging Face and ModelScope.

Key Features

Bilingual support (Chinese and English)
Code generation functionality
Table generation with appropriate formatting
Enhanced scenario simulation abilities
Refusal to answer dangerous or harmful questions
Improved contextual question answering and creative writing
Basic mathematical computation capabilities
Maximum token length of 4096
Lightweight architecture: runs on consumer GPU, PC, or phone (INT4 mode ~400MB)
Open-source and free to use

Pros & Cons

Pros
  • Fully open-source and free, no usage limits
  • Lightweight enough to run on consumer hardware (including phones with INT4)
  • Supports both Chinese and English in a single model
  • Includes safety features like refusing harmful queries
  • Offers code, table, and math capabilities in addition to dialogue
  • Easy to use with transformers library and provided Gradio app
Cons
  • May not match the performance of much larger models like GPT-4 for complex tasks
  • Training data specifics and limitations are not fully disclosed
  • Code generation is basic and may not handle advanced programming scenarios
  • Relatively smaller community compared to some other open-source LLMs

Best For

Bilingual conversational AI assistantsCode generation and basic programming helpTable generation for structured data outputCreative writing and story generationScenario-based role-playing simulationsQuestion answering in Chinese and EnglishEducational tools for math problem solving

FAQ

What is ChatYuan?
ChatYuan is an open-source large language model for dialogue developed by clue-ai, supporting both Chinese and English. It is designed as a functional dialogue model in the ChatGPT style, with a lightweight architecture.
What languages does ChatYuan support?
ChatYuan-large-v2 supports both Chinese and English for dialogue, code generation, and other tasks.
Is ChatYuan free to use?
Yes, ChatYuan is fully open-source and free. The model weights are available on Hugging Face and ModelScope, and the code is available on GitHub under an open-source license.
How can I run ChatYuan locally?
You can run ChatYuan using the transformers library in Python. The README provides code examples for loading the model with AutoTokenizer and AutoModel, and a Gradio app is included for easy web interaction.
What hardware do I need to run ChatYuan?
ChatYuan-large-v2 is lightweight and can run on consumer-grade GPUs, PCs, and even mobile devices. In INT4 quantization mode, it requires only about 400MB of memory.
What are the main improvements in ChatYuan-large-v2 over v1?
Improvements include enhanced basic capabilities (contextual QA, creative writing), added bilingual support, refusal to answer harmful questions, code generation, table generation, enhanced math, increased token length to 4096, and better scenario simulation.