Having Trouble building firmware for Nicla Vision

I have been trying to follow this guide using edge impulse, OpenMV, and a Nicla vision

In the deploy stage I am unable to get GitHub to build the firmware successfully.
I Tried making the set smaller (went from 98x98 to 48x48, tried grayscale, made the training cycles smaller) I haven’t messed with the data set is that What I need to change to make it smaller?

I am also new to Github, so I’m unsure if I’m doing everything right.
here is my repository
Actions · whisleraaron/openmv · GitHub

I have been able to follow everything else but every time I put a dataset in the models it doesn’t build the firmware. I have just been doing it all in Github and deleting whats in models and replacing it with my files

any help is appreciated,
Thanks!

Hi,

If you click on any of the firmware builds it will tell you how much larger your model is than the firmware.

Also, you can save firmware space by just disabling features in the image library if you don’t need them:

Just comment out stuff there that you don’t need.

They control stuff here: openmv/py_image.c at master · openmv/openmv · GitHub

1 Like