# SocialAI-tianji/Tianji

制作懂人情世故的大语言模型 | 涵盖提示词工程、RAG、Agent、LLM微调教程

Repository: https://github.com/SocialAI-tianji/Tianji
Canonical: https://ross.abutalabs.com/products/socialai-tianji-tianji
Homepage: https://socialai-tianji.github.io/socialai-web/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: finetuning, gpt, llm, prompt, rag, qwen
Last push: 2025-04-29T14:07:52+00:00

## Health v2 (maintenance only)
Score: 35/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 19, release rhythm 35, longevity 71
- inputs: {"age_days": 994, "days_push": 491, "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 1822, forks 157 (observed 2026-08-28T04:05:41.094089+00:00)

## What it is
Tianji is an open-source Chinese-language LLM application and tutorial project focused on social nuance ('renqing shigu') scenarios, covering prompt engineering, RAG knowledge bases, agents, and model fine-tuning. It ships datasets, fine-tuning code (LoRA/full fine-tuning on Qwen and similar models), and step-by-step lessons for building full-stack LLM applications.

## Use cases
- learn llm development from scratch
- fine-tune a chinese chat model with lora
- build a rag knowledge base chatbot
- learn prompt engineering
- create domain-specific finetuning datasets
- build llm agents with tool calling
- generate culturally appropriate chinese greetings and messages

## When to choose
- you want a hands-on end-to-end LLM tutorial covering prompts, RAG, agents, and fine-tuning
- you need Chinese social-context corpora and fine-tuned models like the wish-generating models
- you want to practice with Qwen, ChatGPT, DeepSeek, and other Chinese/online LLMs

## When to avoid
- you need a production-ready commercial chatbot rather than educational code
- your use case has no Chinese-language or social-etiquette component
- you only need a single component like a standalone RAG framework

## Facets
- artifact type: learning-resource
- maturity: active
- function: prompt-engineering, rag, agent-framework, llm-training, llm-inference, chatbot, data-generation
- domain: large-language-models, artificial-intelligence, tutorials
- platform: python, cross-platform
- tags: llm-tutorial, finetuning, chinese-social-context, qwen, langchain, llamaindex, lora, education, retrieval-augmented-generation, ai-agents, natural-language-processing

## Member repositories
- SocialAI-tianji/Tianji (main) score 35

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:41.094089+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 2026-08-30T03:19:47.670860+00:00, confidence not recorded.
  - readme: https://github.com/SocialAI-tianji/Tianji (fetched 2026-08-28T04:05:41.094089+00:00, sha 72c65dd89dcd)
  - homepage: https://socialai-tianji.github.io/socialai-web/ (fetched 2026-08-29T10:59:01.249292+00:00, sha dcd42a46abde)
- Data as of 2026-08-30T08:39:29.467469+00:00.