DeepStream On the Fly Model Update network architecture

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU) Jetson Nano
• DeepStream Version 6.0
• JetPack Version (valid for Jetson only) 5.6
• TensorRT Version 8.0.1
• Issue Type( questions, new requirements, bugs) Question

According to the docs of On the Fly Model Update, the new model must have the same model architecture as the previous model. What exactly does this mean? For example, I found the function to work when I change between ResNet18 and ResNet34 (trafficCamNet and peopleNet), but not between some other architecture combinations.

This feature assumes that the model being updated has the same network parameters.

New model must have same network parameter configuration as of previous model (e.g. network resolution, network architecture, number of classes).

You can check whether such parameters of your models are the same.

I am talking about the parameter “network architecture”, what does it mean exactly? Does it refer to the backbone? Why does ResNet18 work with ResNet34?

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.

About network architecture, such as perceptron, feed-forward, Resnet, RNN and so on. Cause ResNet18 and ResNet34 have the same network architecture: Resnet. The two models that you use have the same resolution and same network architecture and so on.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.