LFOSTM

Results 10 comments of LFOSTM

Hello, It depends on the model size, but if no external memory is needed, the model should be capable of running on this device. STM32H745XIH6 is not part of the...

Hello, Could you start from a clean environment following installation recommendations from the main readme: git clone https://github.com/STMicroelectronics/stm32ai-modelzoo.git cd stm32ai-modelzoo python -m venv st_zoo source st_zoo/bin/activate pip install -r requirements.txt...

Could you ensure as well you are using python 3.10.x version? Thanks

Hello, Can you confirm that the issue is solved on your side? Then I will close the ticket. Regards, Laurent

Hello, Actually general.model_path and training.model should be mutually exclusive (as it is the case for Image Classification). We will fix it in future updates.

"How do we determine the appropriate rescaling parameters (preprocessing.rescaling) when starting with an existing model ? Should we rescale our images based on the scale used by this model ?...

Hello, boards like STM32F746G-DISCO is supported for the benchmark service but the deployment examples provided are not for this board.

Hello, Could you provide more information of what you did? Thanks

Hello, Could you make a try using the latest version of the model zoo, and pointing on stedgeai version 10.0.0, please? Thanks & Regards,

Hello, Thanks for your remarks. You right, and good results are the ones in the documentation you reproduce by retraining. You have better results evaluating the trained model because we...