Issue with --input-data-type --inputs-ch-position --output-data-type in ST Edge AI 2.1.0

[qiqi]

Hello everyone,

I encountered an issue while using ST Edge AI 2.1.0 and I would appreciate some help.

Here is my original CUSTOM_MPOOLS configuration, which worked perfectly:

"CUSTOM_MPOOLS": { "memory_pool": "./my_mpools/MY_CUSTOM_MPOOL.mpool", "options" : "--enable-epoch-controller -O3 --all-buffers-info --mvei --cache-maintenance --native-float --enable-virtual-mem-pools --Omax-ca-pipe 4 --Ocache-opt --Os" }

However, when I added the following options to the options section:

--input-data-type uint8 --inputs-ch-position chlast --output-data-type float32

I received the following error:

[AI:network] E103(CliRuntimeError): Error calling the Neural Art compiler - ['', 'The following arguments were not expected: float32 --output-data-type chlast --inputs-ch-position uint8 --input-data-type', 'Run with --help for more information.', '']

I referred to the official documentation and saw that it does mention using these commands to modify the input/output requirements of the model. However, it seems that these commands are not fully supported in ST Edge AI 2.1.0, which is causing the error above.

My questions are:

• Why does adding --input-data-type uint8 --inputs-ch-position chlast --output-data-type float32 cause this error?

• Are these commands incompatible with ST Edge AI 2.1.0, or is there another configuration issue?

• Is there another way to modify the input/output data formats, or are there specific things I need to be aware of?

Thank you in advance for your help!

[Julian E. · Accepted solution]

Hello @qiqi,

These three option are front end option, meaning that they are used in the stedgeai command, not in the aton profiles.

Remove these from the profiles in user_neuralart.json and use them instead like this:

stedgeai.exe generate --model .... --target stm32n6 --st-neural-art --input-data-type ... --output-data-type ... --input-ch-position ...

Have a good day,

Julian