Error Codes | Value | Description | Solution |
INNER | 2 | - The model is not correctly initialized.
- The input data is not in the little endian format.
- The input data is not set to Direct ByteBuffer.
- Failed to load the model file.
| - The status of the inference engine is abnormal. Check whether the inference engine has ever exited abnormally.
- Set ByteBuffer to the little endian format byteBuffer.order(ByteOrder.nativeOrder()).
- Call the ByteBuffer.allocateDirect API when creating a ByteBuffer.
- Check whether the model file path is correct.
|
INACTIVE | 3 | The model file is incorrect or does not exist. | - The model file does not exist.
- Failed to load the model file. Check whether the model conversion parameters are correct.
- Unmatched inference engine version. Upgrade the SDK.
|
The input or output parameters of the model are empty. | Check whether the input or output parameters of the model are correctly set. |
The data type is not supported. | Check whether the data type is correctly set. |
ILLEGAL_PARAMETER | 5 | - Indexes of the model input parameters are incorrectly set.
- The actual number of input parameters does not match that required by the model.
- The length of the input parameter of the inference object does not match that required by the model.
- The length of the inference result does not match that of the output parameter.
| - Check whether the index value of a model input parameter exceeds the number of model input parameters.
- Check whether the actual number of input parameters matches that required by the model.
- Check whether the length of the input parameter matches with that of the original model.
- Check whether the model inference result matches the configured model output length.
|
If the error persists, you can submit a ticket online. Huawei technical support will handle it in time.