App Debugging

App Execution

For details about how to run an application, see Image Classification with Caffe ResNet-50 (Synchronous Inference).

Pay attention to the following points:

1. Model conversion. For details, see ATC Instructions.
2. Before running an executable file, upload the directories of the pyACL initialization configuration file (acl.json), test image, and .om file to the same directory on the host.

   If no parameter is passed to the acl.init API during pyACL initialization, the directory of the pyACL initialization configuration file (acl.json) does not need to be uploaded to the host.

3. To run the code, directly run the corresponding Python script. For example:

1

python3 main.py

Troubleshooting

If an error occurs during application running, you can obtain the log file by referring to Log Reference to view the error details in the log file. Take the following steps to fix the error:

• If the API calling logic is incorrect due to API restrictions, refer to Restrictions and the API-specific restrictions to adjust your API calling logic.
• If the error is reported during operator running on the AI Core, you need to further locate the cause by calling the pyACL API to obtain the description of the error operator for further analysis. For details, see AI Core Troubleshooting.
• For typical cases and solutions of Atlas 200/300/500 Inference Product , see Troubleshooting.