提交 2e46a63c 编写于 作者: C cuicheng01

update PP-LCNet.md

上级 ebcd9710
......@@ -242,10 +242,10 @@ MobileNetV3_large_x0_75 | 64.53 | 151 |
### 2.1 安装 paddleclas
使用如下命令快速安装 paddleclas
使用如下命令快速安装 paddlepaddle, paddleclas
```
pip3 install paddleclas
pip3 install paddlepaddle paddleclas
```
<a name="2.2"></a>
......@@ -260,9 +260,9 @@ paddleclas --model_name=PPLCNet_x1_0 --infer_imgs="docs/images/inference_deploy
结果如下:
```
>>> result
filename: docs/images/inference_deployment/whl_demo.jpg, top-5, class_ids: [8, 7, 86, 81, 85], scores: [0.91347, 0.03779, 0.0036, 0.00117, 0.00112], label_names: ['hen', 'cock', 'partridge', 'ptarmigan', 'quail']
class_ids: [8, 7, 86, 81, 85], scores: [0.91347, 0.03779, 0.0036, 0.00117, 0.00112], label_names: ['hen', 'cock', 'partridge', 'ptarmigan', 'quail'], filename: docs/images/inference_deployment/whl_demo.jpg
Predict complete!
```
```
**备注**: 更换 PPLCNet 的其他 scale 的模型时,只需替换 `model_name`,如将此时的模型改为 `PPLCNet_x2_0` 时,只需要将 `--model_name=PPLCNet_x1_0` 改为 `--model_name=PPLCNet_x2_0` 即可。
......@@ -281,10 +281,9 @@ print(next(result))
```
>>> result
filename: docs/images/inference_deployment/whl_demo.jpg, top-5, class_ids: [8, 7, 86, 81, 85], scores: [0.91347, 0.03779, 0.0036, 0.00117, 0.00112], label_names: ['hen', 'cock', 'partridge', 'ptarmigan', 'quail']
[{'class_ids': [8, 7, 86, 81, 85], 'scores': [0.91347, 0.03779, 0.0036, 0.00117, 0.00112], 'label_names': ['hen', 'cock', 'partridge', 'ptarmigan', 'quail'], 'filename': 'docs/images/inference_deployment/whl_demo.jpg'}]
```
<a name="3"></a>
## 3. 模型训练、评估和预测
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册