MMAPI 04 example, only support dynamic batch(N=-1) onnx model


The help says only support dynamic batch(N=-1) onnx mode.
If I visualize the onnx model given as a sample with netron, it comes out as follows.
Is this dynamic batch also known as implicit batch?
Even if I read the onnx conversion tutorial in the official pytorch document, it didn’t come out to make dynamic batches.
Is there a way to make the already created onnx model into dynamic batch?

nvidia@tegra-ubuntu:/usr/src/jetson_multimedia_api/samples/04_video_dec_trt$ ./video_dec_trt

video_dec_trt [Channel-num] <in-file1> <in-file2> ... <in-format> [options]

        1-32, Number of file arguments should exactly match the number of channels specified

Supported formats:

        -h,--help            Prints this text
        --dbg-level <level>  Sets the debug level [Values 0-3]

         Caffe model:
        --trt-deployfile     set caffe deploy file name
        --trt-modelfile      set caffe model file name
         ONNX model:
        --trt-onnxmodel      set onnx model file name, only support dynamic batch(N=-1) onnx model
        --trt-mode           0 fp16 (if supported), 1 fp32, 2 int8
        --trt-enable-perf    1[default] to enable perf measurement, 0 otherwise

Thank you.


Batchsize = -1 indicates the dynamic batch. (decide in runtime)
But dynamic batch is explicit batch rather than the implicit one.


1 Like