Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,111 @@
|
|
1 |
---
|
2 |
-
license:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
+
license: other
|
3 |
+
language:
|
4 |
+
- en
|
5 |
+
- zh
|
6 |
+
tags:
|
7 |
+
- qwen
|
8 |
+
- rk3588
|
9 |
+
- rkllm
|
10 |
+
- RockChip
|
11 |
---
|
12 |
+
---
|
13 |
+
license: other
|
14 |
+
license_name: tongyi-qianwen
|
15 |
+
license_link: LICENSE
|
16 |
+
language:
|
17 |
+
- en
|
18 |
+
- zh
|
19 |
+
tags:
|
20 |
+
- qwen
|
21 |
+
- rkllm
|
22 |
+
- rk3588
|
23 |
+
- RockChip
|
24 |
+
---
|
25 |
+
# 通义千问1.5-1_8B / Qwen1.5-1_8B-Chat_rkLLM
|
26 |
+
|
27 |
+
|
28 |
+
- [中文](#通义千问1.5-1_8b)
|
29 |
+
- [English](#qwen1.5-1_8b-chat_rkllm)
|
30 |
+
## 通义千问1.5-1_8B
|
31 |
+
### 介绍
|
32 |
+
这是从 [Qwen1.5-1_8B](https://huggingface.co/Qwen/Qwen1.5-1.8B) 转换而来的RKLLM模型,用于Rockchip设备的转换模型。
|
33 |
+
该模型运行在RK3588的NPU上。
|
34 |
+
|
35 |
+
- **模型名称**: Qwen1.5-1_8B
|
36 |
+
- **模型架构**: 同Qwen1.5-1_8B
|
37 |
+
- **发布者**: FydeOS
|
38 |
+
- **日期**: 2024-05-14
|
39 |
+
|
40 |
+
### 模型详情
|
41 |
+
Qwen1.5 是的预发布版本,是基于Transformer的仅解码器大语言模型, 在超大规模的预训练数据上进行训练得到。
|
42 |
+
### 要求
|
43 |
+
|
44 |
+
|
45 |
+
### 使用指南
|
46 |
+
> 本模型仅支持运行在搭载 rk3588 芯片的相关设备,请确认芯片信息并保证 npu 可用
|
47 |
+
#### OpenFyde 系统
|
48 |
+
> ⚠️:请确认你已经升级到openfyde的最新版本
|
49 |
+
- 点击下载模型文件 `XXX.rkllm`
|
50 |
+
- 新建文件夹 `model/` ,将模型文件放置在此文件夹下
|
51 |
+
- 启动 FydeOS AI ,在设置页面进行相关设置
|
52 |
+
|
53 |
+
#### 非 OpenFyde 系统
|
54 |
+
> ⚠️:请确认你进行了 rkllm 的 npu 相关内核更新
|
55 |
+
- 点击下载模型文件 `XXX.rkllm`
|
56 |
+
- 参考官方文档进行配置:[官方文档](https://github.com/airockchip/rknn-llm)
|
57 |
+
|
58 |
+
### FAQ
|
59 |
+
如遇到问题,敬请查阅issue区,如仍无法解决再提交issue。
|
60 |
+
|
61 |
+
### 限制与注意事项
|
62 |
+
说明模型的局限性和使用中的注意事项。
|
63 |
+
|
64 |
+
### 许可证
|
65 |
+
采用与[Qwen1.5-1_8B](https://huggingface.co/Qwen/Qwen1.5-1.8B)相同的许可证。
|
66 |
+
|
67 |
+
### 联系方式
|
68 |
+
如需更多信息,请联系:
|
69 |
+
- **电子邮件**: [email protected]
|
70 |
+
- **主页**: [FydeOS AI](https://fydeos.ai/zh/)
|
71 |
+
|
72 |
+
## Qwen1.5-1_8B-Chat_rkLLM
|
73 |
+
|
74 |
+
### Introduction
|
75 |
+
This is a conversion from [Qwen1.5-1_8B](https://huggingface.co/Qwen/Qwen1.5-1.8B) to the RKLLM format for Rockchip devices. The model runs on the NPU from RK3588.
|
76 |
+
|
77 |
+
- **Model Name**: Qwen1.5-1_8B
|
78 |
+
- **Model Architecture**: Same as Qwen1.5-1_8B
|
79 |
+
- **Publisher**: FydeOS
|
80 |
+
- **Date**: 2024-05-14
|
81 |
+
|
82 |
+
### Model Details
|
83 |
+
Qwen1.5 is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data.
|
84 |
+
### Requirements
|
85 |
+
|
86 |
+
### Usage Guide
|
87 |
+
> This model is only supported on devices equipped with the RK3588 chip. Ensure that the chip and NPU are available.
|
88 |
+
#### For OpenFyde Systems
|
89 |
+
> ⚠️: Please make sure you have upgraded to the latest version of OpenFyde.
|
90 |
+
- Click to download the model file `XXX.rkllm`.
|
91 |
+
- Create a new folder `model/` and place the model file in this folder.
|
92 |
+
- Start FydeOS AI and make the necessary settings on the settings page.
|
93 |
+
|
94 |
+
#### For Non-OpenFyde Systems
|
95 |
+
> ⚠️: Ensure you have updated the NPU related kernel for rkllm.
|
96 |
+
- Click to download the model file `XXX.rkllm`.
|
97 |
+
- Refer to the official documentation for setup: [Official Documentation](https://github.com/airockchip/rknn-llm)
|
98 |
+
|
99 |
+
### FAQ
|
100 |
+
Please consult the issue section for any problems, and submit a new issue if the problem remains unsolved.
|
101 |
+
|
102 |
+
### Limitations and Precautions
|
103 |
+
Details on the limitations and precautions to consider when using the model.
|
104 |
+
|
105 |
+
### License
|
106 |
+
Uses the same license as [Qwen1.5-1_8B](https://huggingface.co/Qwen/Qwen1.5-1.8B).
|
107 |
+
|
108 |
+
### Contact Information
|
109 |
+
For more information, please contact:
|
110 |
+
- **Email**: [email protected]
|
111 |
+
- **Homepage**: [FydeOS AI](https://fydeos.ai/zh/)
|