Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,98 @@
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
- zh
|
| 6 |
+
library_name: transformers
|
| 7 |
+
tags:
|
| 8 |
+
- text-generation-inference
|
| 9 |
---
|
| 10 |
+
### LucasInsight/Meta-Llama-3-8B-Instruct Model Card
|
| 11 |
+
|
| 12 |
+
**Model Overview**
|
| 13 |
+
|
| 14 |
+
The LucasInsight/Meta-Llama-3-8B-Instruct model is an enhanced version of the Meta-Llama3 project, incorporating the alpaca-gpt4-data-zh Chinese dataset. The model has been fine-tuned using Unsloth with 4-bit QLoRA and generates GGUF model files compatible with the Ollama inference engine.
|
| 15 |
+
|
| 16 |
+
**模型概述**
|
| 17 |
+
|
| 18 |
+
LucasInsight/Meta-Llama-3-8B-Instruct 模型是在 Meta-Llama3 工程的基础上,增加了 alpaca-gpt4-data-zh 中文数据集。该模型通过使用 Unsloth 的 4-bit QLoRA 进行微调,生成的 GGUF 模型文件支持 Ollama 推理引擎。
|
| 19 |
+
|
| 20 |
+
**License Information**
|
| 21 |
+
|
| 22 |
+
This project is governed by the licenses of the integrated components:
|
| 23 |
+
|
| 24 |
+
1. **Meta-Llama3 Project**
|
| 25 |
+
- Project URL: [https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct)
|
| 26 |
+
- License: Llama 3 Community License Agreement
|
| 27 |
+
|
| 28 |
+
**Citation:**
|
| 29 |
+
```
|
| 30 |
+
@article{llama3modelcard,
|
| 31 |
+
title={Llama 3 Model Card},
|
| 32 |
+
author={AI@Meta},
|
| 33 |
+
year={2024},
|
| 34 |
+
url = {https://github.com/meta-llama/llama3/blob/main/MODEL_CARD.md}
|
| 35 |
+
}
|
| 36 |
+
```
|
| 37 |
+
|
| 38 |
+
2. **Unsloth Project**
|
| 39 |
+
- License: Apache-2.0 License
|
| 40 |
+
- Project URL: [https://huggingface.co/unsloth/llama-3-8b-Instruct-bnb-4bit](https://huggingface.co/unsloth/llama-3-8b-Instruct-bnb-4bit)
|
| 41 |
+
|
| 42 |
+
3. **Chinese Dataset Integration**
|
| 43 |
+
- Dataset: alpaca-gpt4-data-zh
|
| 44 |
+
- License: CC BY NC 4.0 (for non-commercial research use only)
|
| 45 |
+
- Dataset URL: [https://huggingface.co/datasets/llm-wizard/alpaca-gpt4-data-zh](https://huggingface.co/datasets/llm-wizard/alpaca-gpt4-data-zh)
|
| 46 |
+
|
| 47 |
+
**Usage and License Notices:**
|
| 48 |
+
The data is intended and licensed for research use only. The dataset is CC BY NC 4.0, allowing only non-commercial use. Models trained using this dataset should not be used outside of research purposes.
|
| 49 |
+
|
| 50 |
+
**Citation:**
|
| 51 |
+
```
|
| 52 |
+
@article{peng2023gpt4llm,
|
| 53 |
+
title={Instruction Tuning with GPT-4},
|
| 54 |
+
author={Baolin Peng, Chunyuan Li, Pengcheng He, Michel Galley, Jianfeng Gao},
|
| 55 |
+
journal={arXiv preprint arXiv:2304.03277},
|
| 56 |
+
year={2023}
|
| 57 |
+
}
|
| 58 |
+
```
|
| 59 |
+
|
| 60 |
+
**许可证信息**
|
| 61 |
+
|
| 62 |
+
本项目的许可证由各集成工程的许可证构成:
|
| 63 |
+
|
| 64 |
+
1. **Meta-Llama3 项目**
|
| 65 |
+
- 项目地址:[https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct)
|
| 66 |
+
- 许可证:Llama 3 Community License Agreement
|
| 67 |
+
|
| 68 |
+
**引用说明:**
|
| 69 |
+
```
|
| 70 |
+
@article{llama3modelcard,
|
| 71 |
+
title={Llama 3 Model Card},
|
| 72 |
+
author={AI@Meta},
|
| 73 |
+
year={2024},
|
| 74 |
+
url = {https://github.com/meta-llama/llama3/blob/main/MODEL_CARD.md}
|
| 75 |
+
}
|
| 76 |
+
```
|
| 77 |
+
|
| 78 |
+
2. **Unsloth 项目**
|
| 79 |
+
- 许可证:Apache-2.0 许可证
|
| 80 |
+
- 项目地址:[https://huggingface.co/unsloth/llama-3-8b-Instruct-bnb-4bit](https://huggingface.co/unsloth/llama-3-8b-Instruct-bnb-4bit)
|
| 81 |
+
|
| 82 |
+
3. **中文数据集集成**
|
| 83 |
+
- 数据集:alpaca-gpt4-data-zh
|
| 84 |
+
- 许可证:CC BY NC 4.0(仅用于非商业的研究用途)
|
| 85 |
+
- 数据集地址:[https://huggingface.co/datasets/llm-wizard/alpaca-gpt4-data-zh](https://huggingface.co/datasets/llm-wizard/alpaca-gpt4-data-zh)
|
| 86 |
+
|
| 87 |
+
**使用和许可证通知:**
|
| 88 |
+
该数据仅限于研究使用,且基于 CC BY NC 4.0 许可证,只允许非商业用途。使用此数据集训练的模型不得用于研究用途以外的场合。
|
| 89 |
+
|
| 90 |
+
**引用说明:**
|
| 91 |
+
```
|
| 92 |
+
@article{peng2023gpt4llm,
|
| 93 |
+
title={Instruction Tuning with GPT-4},
|
| 94 |
+
author={Baolin Peng, Chunyuan Li, Pengcheng He, Michel Galley, Jianfeng Gao},
|
| 95 |
+
journal={arXiv preprint arXiv:2304.03277},
|
| 96 |
+
year={2023}
|
| 97 |
+
}
|
| 98 |
+
```
|