Your ONNX model has been generated with INT64 weights

Bug Report

Which model does this pertain to?

tiny-yolov3

Describe the bug

[TensorRT] WARNING: onnx2trt_utils.cpp:217: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.
[TensorRT] WARNING: onnx2trt_utils.cpp:243: One or more weights outside the range of INT32 was clamped
[TensorRT] WARNING: onnx2trt_utils.cpp:243: One or more weights outside the range of INT32 was clamped
[TensorRT] WARNING: onnx2trt_utils.cpp:243: One or more weights outside the range of INT32 was clamped
[TensorRT] WARNING: onnx2trt_utils.cpp:243: One or more weights outside the range of INT32 was clamped
[TensorRT] WARNING: onnx2trt_utils.cpp:243: One or more weights outside the range of INT32 was clamped
[TensorRT] WARNING: onnx2trt_utils.cpp:243: One or more weights outside the range of INT32 was clamped
[TensorRT] WARNING: onnx2trt_utils.cpp:243: One or more weights outside the range of INT32 was clamped
[TensorRT] WARNING: onnx2trt_utils.cpp:243: One or more weights outside the range of INT32 was clamped
[TensorRT] WARNING: onnx2trt_utils.cpp:243: One or more weights outside the range of INT32 was clamped
[TensorRT] WARNING: onnx2trt_utils.cpp:243: One or more weights outside the range of INT32 was clamped
[TensorRT] ERROR: (Unnamed Layer* 0) [Convolution]: image size is smaller than filter size
[TensorRT] ERROR: conv2d_1: image size is smaller than filter size
[TensorRT] ERROR: conv2d_1: image size is smaller than filter size
[TensorRT] ERROR: conv2d_1: image size is smaller than filter size
[TensorRT] ERROR: conv2d_1: image size is smaller than filter size
[TensorRT] ERROR: conv2d_1: image size is smaller than filter size
[TensorRT] ERROR: conv2d_1: image size is smaller than filter size
ERROR: Failed to parse the ONNX file.
In node -1 (scaleHelper): UNSUPPORTED_NODE: Assertion failed: dims.nbDims == 4 || dims.nbDims == 5

Reproduction instructions

System Information

OS Platform and Distribution (e.g. Linux Ubuntu 16.04): jetpack 4.4
ONNX version (e.g. 1.6): 1.4.1
Backend/Runtime version (e.g. ONNX Runtime 1.1, PyTorch 1.2): pytorch 1.5

Provide a code snippet to reproduce your errors.

import onnx
model = onnx.load('model.onnx')
...

Notes

I used my own .pth model to Onnx. and Onnx to TRT ,also ouucr
[TensorRT] WARNING: onnx2trt_utils.cpp:217: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.

Reproduce the tiny-yolov3 conversion using the reported JetPack 4.4, ONNX 1.4.1, and PyTorch 1.5 context, starting with model.onnx and the ONNX-to-TensorRT parse errors. The issue does not name repository files or tests; done would require identifying a reproducible cause and verified conversion outcome for the reported warnings and convolution-shape errors.