Spec file for EfficientNet-B0 for Detectnet_V2 Object Detection

Please provide the following information when requesting support.

• Hardware (T4/V100/Xavier/Nano/etc) - Xavier
• Network Type (Detectnet_v2/Faster_rcnn/Yolo_v4/LPRnet/Mask_rcnn/Classification/etc) - Detectnet_V2
• TLT Version (Please run “tlt info --verbose” and share “docker_tag” here) - 3.22.02

Hello everyone,
I would like to train an object detection model using efficientnet_b0 model. When I run the following command

!ngc registry model list nvidia/tao/pretrained_detectnet_v2:*

I get two efficientnet_b0 models - efficientnet_b0_swish and efficientnet_b0_relu

However, when I am creating the model config section in the spec file, the arch parameters take only the following values : resnet, vgg, mobilenet _v1, mobilenet _v2, googlenet

Can you share a sample spec file for efficientnet_b0 models which can be used as reference?

Looking forward to your replies !!

The detectnet_v2 network support below backbones. It does not support efficientnet backbone.
See DetectNet_v2 — TAO Toolkit 3.22.02 documentation

Thanks for the information !!