Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- zh
|
| 4 |
+
base_model:
|
| 5 |
+
- Qwen/Qwen3-4B-Instruct-2507
|
| 6 |
+
- Qwen/Qwen3-1.7B
|
| 7 |
+
- Qwen/Qwen3-0.6B
|
| 8 |
+
pipeline_tag: text-generation
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
# 用于Nano电子鹦鹉端上部署的Qwen3语言模型
|
| 12 |
+
|
| 13 |
+
本仓库是对Qwen3系列的3个开源模型的二次封装和量化,仅适用于[Nano电子鹦鹉](https://github.com/bd4sur/Nano)。包括以下3个模型:
|
| 14 |
+
|
| 15 |
+
- Qwen3-0.6B
|
| 16 |
+
- Qwen3-1.7B
|
| 17 |
+
- Qwen3-4B-Instruct-2507
|
| 18 |
+
|
| 19 |
+
有FP32(原始精度强制转换到FP32)和Q80(W8A32)两个量化规格。
|
| 20 |
+
|
| 21 |
+
推理引擎的主线实现只支持Q80量化规格。
|
| 22 |
+
|
| 23 |
+
具体用例请转到[GitHub](https://github.com/bd4sur/Nano),查看视频或在线体验。
|