---
license: apache-2.0
language:
- zh
- en
base_model:
- hfl/llama-3-chinese-8b-instruct-v3
---

# Llama-3-Chinese-8B-Instruct-v3-GGUF

<p align="center">
    <a href="https://github.com/ymcui/Chinese-LLaMA-Alpaca-3"><img src="https://ymcui.com/images/chinese-llama-alpaca-3-banner.png" width="600"/></a>
</p>

[**[👉👉👉 Chat with Llama-3-Chinese-8B-Instruct-v3 @ HF Space]**](https://huggingface.co/spaces/hfl-rc/llama-3-chinese-8b-instruct-demo)

This repository contains **Llama-3-Chinese-8B-Instruct-v3-GGUF** (llama.cpp/ollama/tgw, etc. compatible), which is the quantized version of [Llama-3-Chinese-8B-Instruct-v3](https://huggingface.co/hfl/llama-3-chinese-8b-instruct-v3).

**Note: this is an instruction (chat) model, which can be used for conversation, QA, etc.** 

Further details (performance, usage, etc.) should refer to GitHub project page: https://github.com/ymcui/Chinese-LLaMA-Alpaca-3

## Performance

Metric: PPL, lower is better

*Note: Unless constrained by memory, we suggest using Q8_0 or Q6_K for better performance.*

| Quant |     Size |                 PPL |
| :---: | -------: | ------------------: |
| Q2_K  |  2.96 GB | 10.0534 +/- 0.13135 |
| Q3_K  |  3.74 GB |  6.3295 +/- 0.07816 |
| Q4_0  |  4.34 GB |  6.3200 +/- 0.07893 |
| Q4_K  |  4.58 GB |  6.0042 +/- 0.07431 |
| Q5_0  |  5.21 GB |  6.0437 +/- 0.07526 |
| Q5_K  |  5.34 GB |  5.9484 +/- 0.07399 |
| Q6_K  |  6.14 GB |  5.9469 +/- 0.07404 |
| Q8_0  |  7.95 GB |  5.8933 +/- 0.07305 |
|  F16  | 14.97 GB |  5.8902 +/- 0.07303 |

## Others

- For full model, please see: https://huggingface.co/hfl/llama-3-chinese-8b-instruct-v3
- If you have questions/issues regarding this model, please submit an issue through https://github.com/ymcui/Chinese-LLaMA-Alpaca-3