Unable to parse custom pytorch UNET onnx model with python deepstream-segmentation-app

• Hardware Platform (Jetson / GPU) GPU
• DeepStream Version 6.1 Triton Container
• TensorRT Version 8.2.5-1
• NVIDIA GPU Driver Version (valid for GPU only) 510.47.03
• Issue Type( questions, new requirements, bugs) Questions
• 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)

python3 deepstream_segmentation.py seg_onnx.txt …/…/…/…/streams/sample_720p.mjpeg output

  • I am using existing deepstream-seg app with custom pytorch unet onnx model.
  • Here is my pgie config file
[property]
gpu-id=0
net-scale-factor=1.0
model-color-format=0
onnx-file=/app/unet_seg/unet_converted_1280x1918.onnx
batch-size=2
## 0=FP32, 1=INT8, 2=FP16 mode
network-mode=0
num-detected-classes=2
interval=0
gie-unique-id=1
network-type=2
output-blob-names=outc
segmentation-threshold=0.0


[class-attrs-all]
roi-top-offset=0
roi-bottom-offset=0
detected-min-w=0
detected-min-h=0
detected-max-w=0
detected-max-h=0

getting error while nvifer parsing this model:

Starting pipeline 

0:00:03.323590595   463      0x6146750 INFO                 nvinfer gstnvinfer.cpp:646:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::buildModel() <nvdsinfer_context_impl.cpp:1914> [UID = 1]: Trying to create engine from model files
WARNING: [TRT]: onnx2trt_utils.cpp:366: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.
WARNING: [TRT]: onnx2trt_utils.cpp:392: One or more weights outside the range of INT32 was clamped
ERROR: [TRT]: [shuffleNode.cpp::symbolicExecute::387] Error Code 4: Internal Error (Reshape_75: IShuffleLayer applied to shape tensor must have 0 or 1 reshape dimensions: dimensions were [-1,2])
ERROR: [TRT]: ModelImporter.cpp:773: While parsing node number 86 [Pad -> "onnx::Concat_218"]:
ERROR: [TRT]: ModelImporter.cpp:774: --- Begin node ---
ERROR: [TRT]: ModelImporter.cpp:775: input: "x1"
input: "onnx::Pad_216"
input: "onnx::Pad_217"
output: "onnx::Concat_218"
name: "Pad_86"
op_type: "Pad"
attribute {
  name: "mode"
  s: "constant"
  type: STRING
}

ERROR: [TRT]: ModelImporter.cpp:776: --- End node ---
ERROR: [TRT]: ModelImporter.cpp:779: ERROR: ModelImporter.cpp:179 In function parseGraph:
[6] Invalid Node - Pad_86
[shuffleNode.cpp::symbolicExecute::387] Error Code 4: Internal Error (Reshape_75: IShuffleLayer applied to shape tensor must have 0 or 1 reshape dimensions: dimensions were [-1,2])
ERROR: ../nvdsinfer/nvdsinfer_model_builder.cpp:315 Failed to parse onnx file
ERROR: ../nvdsinfer/nvdsinfer_model_builder.cpp:966 failed to build network since parsing model errors.
ERROR: ../nvdsinfer/nvdsinfer_model_builder.cpp:799 failed to build network.
0:00:31.772357377   463      0x6146750 ERROR                nvinfer gstnvinfer.cpp:640:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Error in NvDsInferContextImpl::buildModel() <nvdsinfer_context_impl.cpp:1934> [UID = 1]: build engine file failed
0:00:31.796943142   463      0x6146750 ERROR                nvinfer gstnvinfer.cpp:640:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Error in NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:2020> [UID = 1]: build backend context failed
0:00:31.796998851   463      0x6146750 ERROR                nvinfer gstnvinfer.cpp:640:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Error in NvDsInferContextImpl::initialize() <nvdsinfer_context_impl.cpp:1257> [UID = 1]: generate backend failed, check config file settings
0:00:31.797042630   463      0x6146750 WARN                 nvinfer gstnvinfer.cpp:846:gst_nvinfer_start:<primary-nvinference-engine> error: Failed to create NvDsInferContext instance
0:00:31.797135561   463      0x6146750 WARN                 nvinfer gstnvinfer.cpp:846:gst_nvinfer_start:<primary-nvinference-engine> error: Config file path: /app/unet_seg/seg_onnx.txt, NvDsInfer Error: NVDSINFER_CONFIG_FAILED
Error: gst-resource-error-quark: Failed to create NvDsInferContext instance (1): gstnvinfer.cpp(846): gst_nvinfer_start (): /GstPipeline:pipeline0/GstNvInfer:primary-nvinference-engine:
Config file path: /app/unet_seg/seg_onnx.txt, NvDsInfer Error: NVDSINFER_CONFIG_FAILED

We have trained on custom dataset on this network
GitHub - milesial/Pytorch-UNet: PyTorch implementation of the U-Net for image semantic segmentation with high quality images.

Onnx Model: unet_converted_1280x1918.onnx - Google Drive

Netron Graph of the custom trained onnx model:

Hi,
We recommend you to raise this query in TRITON Inference Server Github instance issues section.

Thanks!

@NVES
I am just using Deepstream triton container and using nvinfer only for the inference which is using tensorrt.
ERROR: [TRT]: [shuffleNode.cpp::symbolicExecute::387] Error Code 4: Internal Error (Reshape_75: IShuffleLayer applied to shape tensor must have 0 or 1 reshape dimensions: dimensions were [-1,2])

This issue is not related to triton.

Hi,

Please refer to the following similar posts, which may help you.

Thank you.

@spolisetty
I performed the steps given in the post you shared.

polygraphy surgeon sanitize --fold-constants unet_seg/unet_converted_1280x1918.onnx -o foldeddd.onnx

Doing some fold operation using onnx graphsurgeon o trt8.2.5.1 container:

root@0689aec570f5:/workspace# polygraphy surgeon sanitize --fold-constants unet_seg/unet_converted_1280x1918.onnx -o foldeddd.onnx

[I] Loading model: /tmp/tmp_polygraphy_f24ecdc817cd1e57e562dfc0dd542563e5ea6ffe88741367.onnx
[I] Original Model:
    Name: torch-jit-export | ONNX Opset: 11

    ---- 1 Graph Input(s) ----
    {inc [dtype=float32, shape=('batch_size', 3, 'width', 'height')]}

    ---- 1 Graph Output(s) ----
    {outc [dtype=float32, shape=('batch_size', 2, 'width', 'height')]}

    ---- 50 Initializer(s) ----

    ---- 297 Node(s) ----

[I] Folding Constants | Pass 1
[I]     Total Nodes | Original:   297, After Folding:   197 |   100 Nodes Folded
[I] Folding Constants | Pass 2
[I]     Total Nodes | Original:   197, After Folding:   197 |     0 Nodes Folded
[I] Saving ONNX model to: foldeddd.onnx
[I] New Model:
    Name: torch-jit-export | ONNX Opset: 11

    ---- 1 Graph Input(s) ----
    {inc [dtype=float32, shape=('batch_size', 3, 'width', 'height')]}

    ---- 1 Graph Output(s) ----
    {outc [dtype=float32, shape=('batch_size', 2, 'width', 'height')]}

    ---- 110 Initializer(s) ----

    ---- 197 Node(s) ----

After the above operation used the model in DeepStream, still it’s same.

0:00:00.365195508  4367      0x6739d50 INFO                 nvinfer gstnvinfer.cpp:646:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::buildModel() <nvdsinfer_context_impl.cpp:1914> [UID = 1]: Trying to create engine from model files
WARNING: [TRT]: onnx2trt_utils.cpp:366: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.
WARNING: [TRT]: onnx2trt_utils.cpp:392: One or more weights outside the range of INT32 was clamped
WARNING: [TRT]: onnx2trt_utils.cpp:392: One or more weights outside the range of INT32 was clamped
WARNING: [TRT]: onnx2trt_utils.cpp:392: One or more weights outside the range of INT32 was clamped
WARNING: [TRT]: onnx2trt_utils.cpp:392: One or more weights outside the range of INT32 was clamped
ERROR: [TRT]: [shuffleNode.cpp::symbolicExecute::387] Error Code 4: Internal Error (Reshape_75: IShuffleLayer applied to shape tensor must have 0 or 1 reshape dimensions: dimensions were [-1,2])
ERROR: [TRT]: ModelImporter.cpp:773: While parsing node number 61 [Pad -> "onnx::Concat_218"]:
ERROR: [TRT]: ModelImporter.cpp:774: --- Begin node ---
ERROR: [TRT]: ModelImporter.cpp:775: input: "x1"
input: "onnx::Pad_216"
input: "onnx::Pad_217"
output: "onnx::Concat_218"
name: "Pad_86"
op_type: "Pad"
attribute {
  name: "mode"
  s: "constant"
  type: STRING
}

ERROR: [TRT]: ModelImporter.cpp:776: --- End node ---
ERROR: [TRT]: ModelImporter.cpp:779: ERROR: ModelImporter.cpp:179 In function parseGraph:
[6] Invalid Node - Pad_86
[shuffleNode.cpp::symbolicExecute::387] Error Code 4: Internal Error (Reshape_75: IShuffleLayer applied to shape tensor must have 0 or 1 reshape dimensions: dimensions were [-1,2])
ERROR: ../nvdsinfer/nvdsinfer_model_builder.cpp:315 Failed to parse onnx file
ERROR: ../nvdsinfer/nvdsinfer_model_builder.cpp:966 failed to build network since parsing model errors.
ERROR: ../nvdsinfer/nvdsinfer_model_builder.cpp:799 failed to build network.
0:00:19.977414541  4367      0x6739d50 ERROR                nvinfer gstnvinfer.cpp:640:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Error in NvDsInferContextImpl::buildModel() <nvdsinfer_context_impl.cpp:1934> [UID = 1]: build engine file failed
0:00:20.008574321  4367      0x6739d50 ERROR                nvinfer gstnvinfer.cpp:640:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Error in NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:2020> [UID = 1]: build backend context failed
0:00:20.008649896  4367      0x6739d50 ERROR                nvinfer gstnvinfer.cpp:640:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Error in NvDsInferContextImpl::initialize() <nvdsinfer_context_impl.cpp:1257> [UID = 1]: generate backend failed, check config file settings
0:00:20.008713164  4367      0x6739d50 WARN                 nvinfer gstnvinfer.cpp:846:gst_nvinfer_start:<primary-nvinference-engine> error: Failed to create NvDsInferContext instance
0:00:20.008838534  4367      0x6739d50 WARN                 nvinfer gstnvinfer.cpp:846:gst_nvinfer_start:<primary-nvinference-engine> error: Config file path: fold.txt, NvDsInfer Error: NVDSINFER_CONFIG_FAILED
Error: gst-resource-error-quark: Failed to create NvDsInferContext instance (1): gstnvinfer.cpp(846): gst_nvinfer_start (): /GstPipeline:pipeline0/GstNvInfer:primary-nvinference-engine:

Hi,

Which version of the PyTorch are you using to generate the ONNX model?
Could you please use the latest PyTorch version and Opset version to generate the ONNX model.
And try again the same steps above mentioned with the latest ONNX model.

Thank you.

pytorch is 1.11.0 &
opset 11

trying with opset 13

root@0689aec570f5:/workspace/unet_seg# polygraphy surgeon sanitize --fold-constants unet_converted_1280x1918_opset_13.onnx -o folded_opset13.onnx
[I] Loading model: /tmp/tmp_polygraphy_8640236436daecdf60a21757a7e2234c36b34db29e77c98f.onnx
[I] Original Model:
    Name: torch-jit-export | ONNX Opset: 13

    ---- 1 Graph Input(s) ----
    {inc [dtype=float32, shape=('batch_size', 3, 'width', 'height')]}

    ---- 1 Graph Output(s) ----
    {outc [dtype=float32, shape=('batch_size', 2, 'width', 'height')]}

    ---- 50 Initializer(s) ----

    ---- 329 Node(s) ----

[I] Folding Constants | Pass 1
[I]     Total Nodes | Original:   329, After Folding:   197 |   132 Nodes Folded
[I] Folding Constants | Pass 2
[I]     Total Nodes | Original:   197, After Folding:   197 |     0 Nodes Folded
[I] Saving ONNX model to: folded_opset13.onnx
[I] New Model:
    Name: torch-jit-export | ONNX Opset: 13

    ---- 1 Graph Input(s) ----
    {inc [dtype=float32, shape=('batch_size', 3, 'width', 'height')]}

    ---- 1 Graph Output(s) ----
    {outc [dtype=float32, shape=('batch_size', 2, 'width', 'height')]}

    ---- 126 Initializer(s) ----

    ---- 197 Node(s) ----

still it’s same with opset 12/13 folded model.

Adding elements to Pipeline 

Linking elements in the Pipeline 

Now playing...
1 :  sample_720p.mjpeg
Starting pipeline 

0:00:07.000338400 10562      0x4e26b50 INFO                 nvinfer gstnvinfer.cpp:646:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::buildModel() <nvdsinfer_context_impl.cpp:1914> [UID = 1]: Trying to create engine from model files
WARNING: [TRT]: onnx2trt_utils.cpp:366: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.
WARNING: [TRT]: onnx2trt_utils.cpp:392: One or more weights outside the range of INT32 was clamped
WARNING: [TRT]: onnx2trt_utils.cpp:392: One or more weights outside the range of INT32 was clamped
WARNING: [TRT]: onnx2trt_utils.cpp:392: One or more weights outside the range of INT32 was clamped
WARNING: [TRT]: onnx2trt_utils.cpp:392: One or more weights outside the range of INT32 was clamped
ERROR: [TRT]: [shuffleNode.cpp::symbolicExecute::387] Error Code 4: Internal Error (Reshape_83: IShuffleLayer applied to shape tensor must have 0 or 1 reshape dimensions: dimensions were [-1,2])
ERROR: [TRT]: ModelImporter.cpp:773: While parsing node number 61 [Pad -> "onnx::Concat_226"]:
ERROR: [TRT]: ModelImporter.cpp:774: --- Begin node ---
ERROR: [TRT]: ModelImporter.cpp:775: input: "x1"
input: "onnx::Pad_224"
input: "onnx::Pad_225"
output: "onnx::Concat_226"
name: "Pad_94"
op_type: "Pad"
attribute {
  name: "mode"
  s: "constant"
  type: STRING
}

ERROR: [TRT]: ModelImporter.cpp:776: --- End node ---
ERROR: [TRT]: ModelImporter.cpp:779: ERROR: ModelImporter.cpp:179 In function parseGraph:
[6] Invalid Node - Pad_94
[shuffleNode.cpp::symbolicExecute::387] Error Code 4: Internal Error (Reshape_83: IShuffleLayer applied to shape tensor must have 0 or 1 reshape dimensions: dimensions were [-1,2])
ERROR: ../nvdsinfer/nvdsinfer_model_builder.cpp:315 Failed to parse onnx file
ERROR: ../nvdsinfer/nvdsinfer_model_builder.cpp:966 failed to build network since parsing model errors.
ERROR: ../nvdsinfer/nvdsinfer_model_builder.cpp:799 failed to build network.
0:00:23.187896384 10562      0x4e26b50 ERROR                nvinfer gstnvinfer.cpp:640:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Error in NvDsInferContextImpl::buildModel() <nvdsinfer_context_impl.cpp:1934> [UID = 1]: build engine file failed
0:00:23.214256270 10562      0x4e26b50 ERROR                nvinfer gstnvinfer.cpp:640:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Error in NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:2020> [UID = 1]: build backend context failed
0:00:23.214333580 10562      0x4e26b50 ERROR                nvinfer gstnvinfer.cpp:640:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Error in NvDsInferContextImpl::initialize() <nvdsinfer_context_impl.cpp:1257> [UID = 1]: generate backend failed, check config file settings
0:00:23.214363958 10562      0x4e26b50 WARN                 nvinfer gstnvinfer.cpp:846:gst_nvinfer_start:<primary-nvinference-engine> error: Failed to create NvDsInferContext instance
0:00:23.214462597 10562      0x4e26b50 WARN                 nvinfer gstnvinfer.cpp:846:gst_nvinfer_start:<primary-nvinference-engine> error: Config file path: fold.txt, NvDsInfer Error: NVDSINFER_CONFIG_FAILED
Error: gst-resource-error-quark: Failed to create NvDsInferContext instance (1): gstnvinfer.cpp(846): gst_nvinfer_start (): /GstPipeline:pipeline0/GstNvInfer:primary-nvinference-engine:
Config file path: fold.txt, NvDsInfer Error: NVDSINFER_CONFIG_FAILED

I have tried with tensorrt8.4.1 earlier I was using 8.2.5.
still the same using trtexec as well

root@f44959378758:/workspace#  ./tensorrt/bin/trtexec --onnx=unet_seg/unet_converted_1280x1918.onnx --workspace=4096 --saveEngine=unet_model.engine
&&&& RUNNING TensorRT.trtexec [TensorRT v8401] # ./tensorrt/bin/trtexec --onnx=unet_seg/unet_converted_1280x1918.onnx --workspace=4096 --saveEngine=unet_model.engine
[08/25/2022-05:55:28] [W] --workspace flag has been deprecated by --memPoolSize flag.
[08/25/2022-05:55:28] [I] === Model Options ===
[08/25/2022-05:55:28] [I] Format: ONNX
[08/25/2022-05:55:28] [I] Model: unet_seg/unet_converted_1280x1918.onnx
[08/25/2022-05:55:28] [I] Output:
[08/25/2022-05:55:28] [I] === Build Options ===
[08/25/2022-05:55:28] [I] Max batch: explicit batch
[08/25/2022-05:55:28] [I] Memory Pools: workspace: 4096 MiB, dlaSRAM: default, dlaLocalDRAM: default, dlaGlobalDRAM: default
[08/25/2022-05:55:28] [I] minTiming: 1
[08/25/2022-05:55:28] [I] avgTiming: 8
[08/25/2022-05:55:28] [I] Precision: FP32
[08/25/2022-05:55:28] [I] LayerPrecisions:
[08/25/2022-05:55:28] [I] Calibration:
[08/25/2022-05:55:28] [I] Refit: Disabled
[08/25/2022-05:55:28] [I] Sparsity: Disabled
[08/25/2022-05:55:28] [I] Safe mode: Disabled
[08/25/2022-05:55:28] [I] DirectIO mode: Disabled
[08/25/2022-05:55:28] [I] Restricted mode: Disabled
[08/25/2022-05:55:28] [I] Build only: Disabled
[08/25/2022-05:55:28] [I] Save engine: unet_model.engine
[08/25/2022-05:55:28] [I] Load engine:
[08/25/2022-05:55:28] [I] Profiling verbosity: 0
[08/25/2022-05:55:28] [I] Tactic sources: Using default tactic sources
[08/25/2022-05:55:28] [I] timingCacheMode: local
[08/25/2022-05:55:28] [I] timingCacheFile:
[08/25/2022-05:55:28] [I] Input(s)s format: fp32:CHW
[08/25/2022-05:55:28] [I] Output(s)s format: fp32:CHW
[08/25/2022-05:55:28] [I] Input build shapes: model
[08/25/2022-05:55:28] [I] Input calibration shapes: model
[08/25/2022-05:55:28] [I] === System Options ===
[08/25/2022-05:55:28] [I] Device: 0
[08/25/2022-05:55:28] [I] DLACore:
[08/25/2022-05:55:28] [I] Plugins:
[08/25/2022-05:55:28] [I] === Inference Options ===
[08/25/2022-05:55:28] [I] Batch: Explicit
[08/25/2022-05:55:28] [I] Input inference shapes: model
[08/25/2022-05:55:28] [I] Iterations: 10
[08/25/2022-05:55:28] [I] Duration: 3s (+ 200ms warm up)
[08/25/2022-05:55:28] [I] Sleep time: 0ms
[08/25/2022-05:55:28] [I] Idle time: 0ms
[08/25/2022-05:55:28] [I] Streams: 1
[08/25/2022-05:55:28] [I] ExposeDMA: Disabled
[08/25/2022-05:55:28] [I] Data transfers: Enabled
[08/25/2022-05:55:28] [I] Spin-wait: Disabled
[08/25/2022-05:55:28] [I] Multithreading: Disabled
[08/25/2022-05:55:28] [I] CUDA Graph: Disabled
[08/25/2022-05:55:28] [I] Separate profiling: Disabled
[08/25/2022-05:55:28] [I] Time Deserialize: Disabled
[08/25/2022-05:55:28] [I] Time Refit: Disabled
[08/25/2022-05:55:28] [I] Inputs:
[08/25/2022-05:55:28] [I] === Reporting Options ===
[08/25/2022-05:55:28] [I] Verbose: Disabled
[08/25/2022-05:55:28] [I] Averages: 10 inferences
[08/25/2022-05:55:28] [I] Percentile: 99
[08/25/2022-05:55:28] [I] Dump refittable layers:Disabled
[08/25/2022-05:55:28] [I] Dump output: Disabled
[08/25/2022-05:55:28] [I] Profile: Disabled
[08/25/2022-05:55:28] [I] Export timing to JSON file:
[08/25/2022-05:55:28] [I] Export output to JSON file:
[08/25/2022-05:55:28] [I] Export profile to JSON file:
[08/25/2022-05:55:28] [I]
[08/25/2022-05:55:28] [I] === Device Information ===
[08/25/2022-05:55:28] [I] Selected Device: Tesla T4
[08/25/2022-05:55:28] [I] Compute Capability: 7.5
[08/25/2022-05:55:28] [I] SMs: 40
[08/25/2022-05:55:28] [I] Compute Clock Rate: 1.59 GHz
[08/25/2022-05:55:28] [I] Device Global Memory: 14910 MiB
[08/25/2022-05:55:28] [I] Shared Memory per SM: 64 KiB
[08/25/2022-05:55:28] [I] Memory Bus Width: 256 bits (ECC enabled)
[08/25/2022-05:55:28] [I] Memory Clock Rate: 5.001 GHz
[08/25/2022-05:55:28] [I]
[08/25/2022-05:55:28] [I] TensorRT version: 8.4.1
[08/25/2022-05:55:40] [I] [TRT] [MemUsageChange] Init CUDA: CPU +311, GPU +0, now: CPU 319, GPU 4622 (MiB)
[08/25/2022-05:55:53] [I] [TRT] [MemUsageChange] Init builder kernel library: CPU +207, GPU +68, now: CPU 545, GPU 4690 (MiB)
[08/25/2022-05:55:53] [I] Start parsing network model
[08/25/2022-05:55:53] [I] [TRT] ----------------------------------------------------------------
[08/25/2022-05:55:53] [I] [TRT] Input filename:   unet_seg/unet_converted_1280x1918.onnx
[08/25/2022-05:55:53] [I] [TRT] ONNX IR version:  0.0.6
[08/25/2022-05:55:53] [I] [TRT] Opset version:    11
[08/25/2022-05:55:53] [I] [TRT] Producer name:    pytorch
[08/25/2022-05:55:53] [I] [TRT] Producer version: 1.11.0
[08/25/2022-05:55:53] [I] [TRT] Domain:
[08/25/2022-05:55:53] [I] [TRT] Model version:    0
[08/25/2022-05:55:53] [I] [TRT] Doc string:
[08/25/2022-05:55:53] [I] [TRT] ----------------------------------------------------------------
[08/25/2022-05:55:53] [W] [TRT] parsers/onnx/onnx2trt_utils.cpp:367: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.
[08/25/2022-05:55:54] [W] [TRT] parsers/onnx/onnx2trt_utils.cpp:395: One or more weights outside the range of INT32 was clamped
[08/25/2022-05:55:55] [E] Error[4]: [shuffleNode.cpp::symbolicExecute::392] Error Code 4: Internal Error (Reshape_75: IShuffleLayer applied to shape tensor must have 0 or 1 reshape dimensions: dimensions were [-1,2])
[08/25/2022-05:55:55] [E] [TRT] parsers/onnx/ModelImporter.cpp:773: While parsing node number 86 [Pad -> "onnx::Concat_218"]:
[08/25/2022-05:55:55] [E] [TRT] parsers/onnx/ModelImporter.cpp:774: --- Begin node ---
[08/25/2022-05:55:55] [E] [TRT] parsers/onnx/ModelImporter.cpp:775: input: "x1"
input: "onnx::Pad_216"
input: "onnx::Pad_217"
output: "onnx::Concat_218"
name: "Pad_86"
op_type: "Pad"
attribute {
  name: "mode"
  s: "constant"
  type: STRING
}



[08/25/2022-05:55:55] [E] [TRT] parsers/onnx/ModelImporter.cpp:776: --- End node ---
[08/25/2022-05:55:55] [E] [TRT] parsers/onnx/ModelImporter.cpp:778: ERROR: parsers/onnx/ModelImporter.cpp:180 In function parseGraph:
[6] Invalid Node - Pad_86
[shuffleNode.cpp::symbolicExecute::392] Error Code 4: Internal Error (Reshape_75: IShuffleLayer applied to shape tensor must have 0 or 1 reshape dimensions: dimensions were [-1,2])
[08/25/2022-05:55:55] [E] Failed to parse onnx file
[08/25/2022-05:55:55] [I] Finish parsing network model
[08/25/2022-05:55:55] [E] Parsing model failed
[08/25/2022-05:55:55] [E] Failed to create engine from model or file.
[08/25/2022-05:55:55] [E] Engine set up failed
&&&& FAILED TensorRT.trtexec [TensorRT v8401] # ./tensorrt/bin/trtexec --onnx=unet_seg/unet_converted_1280x1918.onnx --workspace=4096 --saveEngine=unet_model.engine

Can you try the TAO unet model? PeopleSemSegnet | NVIDIA NGC