search

analogdevicesinc / ai8x-training

Model Training for ADI's MAX78000 and MAX78002 Edge AI Devices
Apache License 2.0
86 stars 78 forks source link

Setting up a Custom Model Issue #152

Closed Doruk-Dilmen closed 2 years ago

Doruk-Dilmen commented 2 years ago

Hello,

I create custemize model: image

Then, I trained the model but I can't do it Synthesis part. After tranning these files were created: image

I used this command line: python quantize.py proj/best.pth.tar proj/proj_q8.pth.tar --device MAX78000 proj_q8.pth.tar folder is created.

I wrote the yaml file for my custom model: image image

After that, I runned this command line: python ai8xize.py --verbose --test-dir demos --prefix doruk-kws20 --checkpoint-file proj/best.pth.tar --config-file networks/kws20-doruk.yaml --device MAX78000 --compact-data --mexpress --softmax But result: image

Can you help me? What am I doing wrong?

aniktash commented 2 years ago

For quantization, use qat_best.pth.tar. Check the kernel size in YAML to make it the same as the model.

github-actions[bot] commented 2 years ago

This issue has been marked stale because it has been open for over 30 days with no activity. It will be closed automatically in 10 days unless a comment is added or the "Stale" label is removed.