trtexec --onnx=folded.onnx
&&&& RUNNING TensorRT.trtexec [TensorRT v8000] # trtexec --onnx=folded.onnx
[05/28/2021-10:00:07] [I] === Model Options ===
[05/28/2021-10:00:07] [I] Format: ONNX
[05/28/2021-10:00:07] [I] Model: folded.onnx
[05/28/2021-10:00:07] [I] Output:
[05/28/2021-10:00:07] [I] === Build Options ===
[05/28/2021-10:00:07] [I] Max batch: explicit
[05/28/2021-10:00:07] [I] Workspace: 16 MiB
[05/28/2021-10:00:07] [I] minTiming: 1
[05/28/2021-10:00:07] [I] avgTiming: 8
[05/28/2021-10:00:07] [I] Precision: FP32
[05/28/2021-10:00:07] [I] Calibration: 
[05/28/2021-10:00:07] [I] Refit: Disabled
[05/28/2021-10:00:07] [I] Sparsity: Disabled
[05/28/2021-10:00:07] [I] Safe mode: Disabled
[05/28/2021-10:00:07] [I] Enable serialization: Disabled
[05/28/2021-10:00:07] [I] Save engine: 
[05/28/2021-10:00:07] [I] Load engine: 
[05/28/2021-10:00:07] [I] NVTX verbosity: 0
[05/28/2021-10:00:07] [I] Tactic sources: Using default tactic sources
[05/28/2021-10:00:07] [I] timingCacheMode: local
[05/28/2021-10:00:07] [I] timingCacheFile: 
[05/28/2021-10:00:07] [I] Input(s)s format: fp32:CHW
[05/28/2021-10:00:07] [I] Output(s)s format: fp32:CHW
[05/28/2021-10:00:07] [I] Input build shapes: model
[05/28/2021-10:00:07] [I] Input calibration shapes: model
[05/28/2021-10:00:07] [I] === System Options ===
[05/28/2021-10:00:07] [I] Device: 0
[05/28/2021-10:00:07] [I] DLACore: 
[05/28/2021-10:00:07] [I] Plugins:
[05/28/2021-10:00:07] [I] === Inference Options ===
[05/28/2021-10:00:07] [I] Batch: Explicit
[05/28/2021-10:00:07] [I] Input inference shapes: model
[05/28/2021-10:00:07] [I] Iterations: 10
[05/28/2021-10:00:07] [I] Duration: 3s (+ 200ms warm up)
[05/28/2021-10:00:07] [I] Sleep time: 0ms
[05/28/2021-10:00:07] [I] Streams: 1
[05/28/2021-10:00:07] [I] ExposeDMA: Disabled
[05/28/2021-10:00:07] [I] Data transfers: Enabled
[05/28/2021-10:00:07] [I] Spin-wait: Disabled
[05/28/2021-10:00:07] [I] Multithreading: Disabled
[05/28/2021-10:00:07] [I] CUDA Graph: Disabled
[05/28/2021-10:00:07] [I] Separate profiling: Disabled
[05/28/2021-10:00:07] [I] Time Deserialize: Disabled
[05/28/2021-10:00:07] [I] Time Refit: Disabled
[05/28/2021-10:00:07] [I] Skip inference: Disabled
[05/28/2021-10:00:07] [I] Inputs:
[05/28/2021-10:00:07] [I] === Reporting Options ===
[05/28/2021-10:00:07] [I] Verbose: Disabled
[05/28/2021-10:00:07] [I] Averages: 10 inferences
[05/28/2021-10:00:07] [I] Percentile: 99
[05/28/2021-10:00:07] [I] Dump refittable layers:Disabled
[05/28/2021-10:00:07] [I] Dump output: Disabled
[05/28/2021-10:00:07] [I] Profile: Disabled
[05/28/2021-10:00:07] [I] Export timing to JSON file: 
[05/28/2021-10:00:07] [I] Export output to JSON file: 
[05/28/2021-10:00:07] [I] Export profile to JSON file: 
[05/28/2021-10:00:07] [I] 
[05/28/2021-10:00:07] [I] === Device Information ===
[05/28/2021-10:00:07] [I] Selected Device: GeForce RTX 2070 Super
[05/28/2021-10:00:07] [I] Compute Capability: 7.5
[05/28/2021-10:00:07] [I] SMs: 40
[05/28/2021-10:00:07] [I] Compute Clock Rate: 1.38 GHz
[05/28/2021-10:00:07] [I] Device Global Memory: 7973 MiB
[05/28/2021-10:00:07] [I] Shared Memory per SM: 64 KiB
[05/28/2021-10:00:07] [I] Memory Bus Width: 256 bits (ECC disabled)
[05/28/2021-10:00:07] [I] Memory Clock Rate: 7.001 GHz
[05/28/2021-10:00:07] [I] 
[05/28/2021-10:00:07] [I] TensorRT version: 8000
[05/28/2021-10:00:07] [I] [TRT] [MemUsageChange] Init CUDA: CPU +267, GPU +0, now: CPU 272, GPU 661 (MiB)
[05/28/2021-10:00:07] [I] [TRT] ----------------------------------------------------------------
[05/28/2021-10:00:07] [I] [TRT] Input filename:   folded.onnx
[05/28/2021-10:00:07] [I] [TRT] ONNX IR version:  0.0.7
[05/28/2021-10:00:07] [I] [TRT] Opset version:    11
[05/28/2021-10:00:07] [I] [TRT] Producer name:    
[05/28/2021-10:00:07] [I] [TRT] Producer version: 
[05/28/2021-10:00:07] [I] [TRT] Domain:           
[05/28/2021-10:00:07] [I] [TRT] Model version:    0
[05/28/2021-10:00:07] [I] [TRT] Doc string:       
[05/28/2021-10:00:07] [I] [TRT] ----------------------------------------------------------------
Unsupported ONNX data type: UINT8 (2)
[05/28/2021-10:00:07] [E] [TRT] ModelImporter.cpp:744: ERROR: input_tensor:248 In function importInput:
[8] Assertion failed: convertDtype(onnxDtype.elem_type(), &trtDtype) && "Failed to convert ONNX date type to TensorRT data type."
[05/28/2021-10:00:07] [E] Failed to parse onnx file
[05/28/2021-10:00:07] [E] Parsing model failed
[05/28/2021-10:00:07] [E] Engine creation failed
[05/28/2021-10:00:07] [E] Engine set up failed
&&&& FAILED TensorRT.trtexec [TensorRT v8000] # trtexec --onnx=folded.onnx