Dustin Franklin

> i tried using the command `--debug-steps=1` and I also command out the part from `voc_dataset.py` but Im not sure how to commit the change in the container. You would...

The ssd-mobilenet-v1 is the only network architecture from pytorch-ssd that I have tested & verified is working through the whole pipeline, including the ONNX export from PyTorch and import into...

@chromaowl can you provide the terminal log of the error you are getting? Are you sure you're providing the correct path to your dataset when you launch train_ssd.py?

Hi @miguemorales, I think that adding a class requires re-training across all of the data. For reference, see this similar discussion about adding a class to YOLO: https://github.com/ultralytics/yolov5/issues/1071

Hi @AgemennoN, that 90-class ssd-mobilenet-coco model came from TensorFlow, whereas the new models are trained in PyTorch, so that model in particular isn't capable of being extended (unless you were...

> 1. I will use the model as "--pretrained-ssd=mobilenet-v1-ssd-voc" while executing train_ssd.py right? > But what about the "--net" flag. I couldn't understand its purpose. There's a pretrained PyTorch checkpoint...

I'm not entirely sure, but I think the new labels.txt would need to have all the old classes (in the same order), in addition to your new classes. If you...

Hi @Cuzny, jetson-inference uses TensorRT underneath for the execution of the DNN inferencing. So the 3dconv operation would be the same.

Other than enabling INT8/FP16 mode, that would be my only suggestion. That and decreasing the resolution of the grid size if you can.

It looks like you are doing 30 inferencing runs in your benchmark, are you accounting for that in the total execution time? Regardless, the TensorRT forum may be a better...