Hi i geting ERRor Negative Slope ExponentNonFixable

[vinaygouda.ttssl]

Negative Slope ExponentNonFixable: Quantization failed in layer custom_violation_model_700epoch_v2/conv83 due to unsupported required slope. Desired shift is 19.0, but op has only 8 data bits. This error raises when the data or weight range are not balanced. Mostly happens when using random calibration-set/weights, the calibration-set is not normalized properly or batch-normalization was not used during training. @Nadav @omria @Omer

[vinaygouda.ttssl]

plz any one tell me why this error ocured an how to solve it

[nina-vilela]

Hi @vinaygouda.ttssl,

As the error message says, this issue usually happens when there are issues with normalization/calibset/weights. Could you share the following details with us:

1. Model architecture
2. Model script used
3. How calibset was created

[vinaygouda.ttssl]

Model architecture yolov8
Calibration set has been created as per the hailo tutorials

[nina-vilela]

please try this solution:

Problem With Model Optimization

I might have a solution, but I’m not sure if it’s general, but I think that it worth checking. It seems that on datasets that have low number of classes and/or low number of pictures, some nodes on the YOLOv8 are getting almost nullified. I’ve used this command to force a wider dynamic range on the outputs: quantization_param([conv42, conv53, conv63], force_range_out=[0.0, 1.0]) This is relevant for YOLOV8s trained from our retraining dockers. These are the end nodes before the NMS layer.