modify installation_other_devices.md for export (#2133)

docs/other_devices_support/installation_other_devices.md

@@ -99,4 +99,13 @@ paddlex --install --platform gitee.com
 All packages are installed.
 ```
 ## 2、使用
-基于昇腾 NPU、寒武纪 MLU、昆仑 XPU、海光DCU 硬件平台的 PaddleX 模型产线开发工具使用方法与 GPU 相同，只需根据所属硬件平台，修改配置设备的参数，详细的使用教程可以查阅[PaddleX产线开发工具本地使用教程](../pipeline_usage/pipeline_develop_guide.md)
+基于昇腾 NPU、寒武纪 MLU、昆仑 XPU、海光DCU 硬件平台的 PaddleX 模型产线开发工具使用方法与 GPU 相同，只需根据所属硬件平台，修改配置设备的参数，详细的使用教程可以查阅[PaddleX产线开发工具本地使用教程](../pipeline_usage/pipeline_develop_guide.md)
+
+需要注意的是，这些硬件平台暂不支持 PaddleX 的边训练边导出静态图的功能，在您训练完之后，体验模块和产线的能力或准备部署时，需要先手动导出静态图。例如，在昇腾 NPU 平台下，导出训练好的 PP-YOLOE_plus-S 模型，可以运行如下命令：
+
+```
+python main.py -c paddlex/configs/object_detection/PP-YOLOE_plus-S.yaml \
+    -o Global.mode=export \
+    -o Global.device=npu \
+    -o Export.weight_path=output/best_model/best_model.pdparams
+```

docs/other_devices_support/installation_other_devices_en.md

@@ -103,3 +103,12 @@ All packages are installed.
 ## 2. Usage
 
 The usage of PaddleX model pipeline development tool on hardware platforms such as Ascend NPU, Cambricon MLU, Kunlun XPU, and Hygon DCU is identical to that on GPU. You only need to modify the device configuration parameters according to your hardware platform. For detailed usage tutorials, please refer to [PaddleX Pipeline Development Tool Local Usage Guide](../pipeline_usage/pipeline_develop_guide_en.md).
+
+It is important to note that these hardware platforms currently do not support PaddleX’s simultaneous training and static graph export feature. After completing your training, when you are ready to test the capabilities in a module or prepare for deployment on the production line, you will need to manually export the static graph. For example, to export the trained PP-YOLOE_plus-S model on the Ascend NPU platform, you can run the following command:
+
+```
+python main.py -c paddlex/configs/object_detection/PP-YOLOE_plus-S.yaml \
+    -o Global.mode=export \
+    -o Global.device=npu \
+    -o Export.weight_path=output/best_model/best_model.pdparams
+```