Please provide complete information as applicable to your setup.
• Hardware Platform (Jetson / GPU)
• DeepStream Version
• JetPack Version (valid for Jetson only)
4.5 (latest SD image)
• TensorRT Version
• NVIDIA GPU Driver Version (valid for GPU only)
• Issue Type( questions, new requirements, bugs)
• How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)
• Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)
I’m having a question regarding the config entry
model-engine-file. I’m having a configuration, in which I can use up to three cameras, dynamically from the command line, as in the Python “test-3” sample with the files.
As like with the files I’m programmatically change the
batch-size parameter according to the found sources (so 1 for one, 2 for two and so on).
This seems to clash with the fixed entry in the config, which always points to just one specific file., e.g./home/neil/jetson/models/primary-detector-nano/resnet10.caffemodel_b1_gpu0_fp16.engine
If now for instance this file was produced with (as the name “b1” says) batch-size 1, everything is OK and the engine loads it and boots up fast, if I start it with one source only. But if I start it with for instance 3 sources, the entire file is re-constructed, which slows down the start (and sometimes also causes a crash).
Is there a way to programmatically configure the engine file w/o having to adapt the config file all the time depending on the number of inputs?
I could als omit the dynamic change of the batch-size, but I don’t know about the consequences.