Neural network engineering built with CubeAI 7.10, and then initialized in MDK5 to error（type=33）

[JZhou.6]

I built the LSTM network with keras 2.6.0 and generated the MDK project with CubeAI 7.10. However, in running the program AIInit() reported an error with an error type of 33.Do you know what the problem is? Thank you very much!

[Andrew Neil]

Doesn't the documentation of that function tell you what error codes it can return, and why?

Also try a 'Go To Definition' on AI_ERROR_NONE - there you should also find the definitions of other error codes ...

[JZhou.6]

I looked at the meaning of the error type. This is (AI_ERROR_CREATE_FAILED), but there is no documentation telling me how to solve this problem. I wouldn't have this problem with CubeAI 6.0.If you know where to help documentation or workarounds, please let me know. Thank you very much!

[Andrew Neil]

so now you can look at the source of that function, and see where that code is returned.

that'll tell you what it was trying to create; should also show why it failed.

[JZhou.6]

I've been using GO TO define refe's method to locate the error. But when it comes to this API, it appears undefine.

[Andrew Neil]

Try stepping into it, then.

[JZhou.6]

I can't get into this function. But I found that the value of the (AI_NETWORK_DATA_CONFIG) parameter generated by CubeAI7.1 is NULL. This is also the input parameter of this function. Could it be a problem here. Do you have a reference document for reporting errors, thank you!

[Andrew Neil]

@JZhou.6​ "I can't get into this function"

How do you mean?

Are you sure that's the one returning the error?

"Do you have a reference document"

Here?

[JZhou.6]

Thank you�?I have also checked a lot of documents, and I have not seen the error analysis.