# ymcui/Chinese-LLaMA-Alpaca-2

中文LLaMA-2 & Alpaca-2大模型二期项目 + 64K超长上下文模型 (Chinese LLaMA-2 & Alpaca-2 LLMs with 64K long context models)

Repository: https://github.com/ymcui/Chinese-LLaMA-Alpaca-2
Canonical: https://ross.abutalabs.com/products/chinese-llama-alpaca-2
Language: Python
License: Apache-2.0
License Family: permissive
Topics: alpaca, llama, llm, llama-2, large-language-models, nlp, alpaca-2, flash-attention, llama2, alpaca2, 64k, yarn, rlhf
Last push: 2026-04-19T00:58:50+00:00

## Health v2 (maintenance only)
Score: 54/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 78, release rhythm 8, longevity 81
- inputs: {"age_days": 1142, "days_push": 137, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 7120, forks 561 (observed 2026-08-28T04:09:56.020870+00:00)

## What it is
A project releasing Chinese-adapted LLaMA-2 base models and Alpaca-2 instruction-tuned chat models with an extended Chinese vocabulary, including 16K/64K long-context variants and RLHF-aligned versions. It provides pretraining and instruction-tuning scripts plus local quantized deployment guides compatible with the LLaMA ecosystem (transformers, llama.cpp, vLLM, LangChain).

## Use cases
- run a chinese instruction-tuned llm locally
- fine-tune llama-2 on chinese text
- deploy a chinese chatbot on cpu or gpu
- process long documents with 64k context llm
- quantize a chinese llama model for local inference
- align an llm with human preferences using rlhf

## When to choose
- you need strong Chinese language understanding or generation from an open Llama-2 model
- you want long-context (16K/64K) Chinese models
- you want to further pretrain or instruction-tune on your own Chinese data
- you need locally deployable quantized Chinese chat models

## When to avoid
- you need a general-purpose English-first LLM
- you want the newest generation - Chinese-LLaMA-Alpaca-3 supersedes this project
- you need multimodal (vision) capabilities
- you cannot run models locally and prefer a hosted API

## Facets
- artifact type: library
- maturity: stable
- function: llm-training, llm-inference, nlp, machine-learning
- domain: large-language-models, machine-learning, artificial-intelligence
- platform: python, cross-platform
- tags: chinese-llm, llama-2, alpaca, long-context, rlhf, quantization, huggingface, pretrained-models, natural-language-processing, gpu, linux

## Member repositories
- ymcui/Chinese-LLaMA-Alpaca-2 (main) score 54

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:56.020870+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 2026-08-29T17:40:05.661144+00:00, confidence not recorded.
  - readme: https://github.com/ymcui/Chinese-LLaMA-Alpaca-2 (fetched 2026-08-28T04:09:56.020870+00:00, sha 8e0b30d8d828)
- Data as of 2026-08-30T08:39:29.467469+00:00.
