Loading a nvidia-tao model trained in python code

Hello All!,

I have managed to train a Fater rcnn model using the nvidia-tao framework, I wanted to ask about the way to load this model inside python code for experimenting with intermediate features of the model and looking for the confidences of the predictions and not only the final detection metrics (a yielded by the command line ‘evaluate’ tool)

I would be so glad for any reference on this!

There is no update from you for a period, assuming this is not an issue anymore.
Hence we are closing this topic. If need further support, please open a new one.

Currently, there is not explicit python code yet. You can refer to the preprocessing deepstream_tao_apps/pgie_frcnn_tao_config.txt at master · NVIDIA-AI-IOT/deepstream_tao_apps · GitHub and postprocessing in deepstream_tao_apps/nvdsinfer_custombboxparser_tao.cpp at master · NVIDIA-AI-IOT/deepstream_tao_apps · GitHub . You can run inference against faster_rcnn model via deepstream application. GitHub - NVIDIA-AI-IOT/deepstream_tao_apps: Sample apps to demonstrate how to deploy models trained with TAO on DeepStream