AttributeError: 'NoneType' object has no attribute 'create_execution_context'

[wrz57661]

Hello,SceanAvery! When I run the demo.py,something went wrong.How can I solve it?
File "demo.py", line 23, in main
processor = Processor(model=args['model'])
File "/home/waveshare/yolov5-tensorrt/python/lib/Processor.py", line 19, in init
self.context = engine.create_execution_context()
AttributeError: 'NoneType' object has no attribute 'create_execution_context'

[seanavery]

hey @wrz57661 , this is due to differing TRT version from the compiled model.

I have updated the export_tensorrt.py script.

Will work if you recompile.

[seanavery]

uploaded new compiled model with trt version 7.1.3.0

[sunanlin13174]

hi,SeanAvery,i meets the same problem, my jetpack is 4.4, and tensorRT may be 7.1.3.0,i run this demo.py ,and got the same error, did i do some things wrong?

[sunanlin13174]

yes , my tensorRT verson is 7.1.3.0 ,and i try it in jetson xavier nx devices

[sunanlin13174]

@seanavery

[seanavery]

Thank you for your patience @sunanlin13174 🙏

Have not tried with nx but also using jetpack 4.4

Can you send me full error stack?

[sunanlin13174]

@seanavery ,Thank you for your quick reply. Your work is very useful for me to learn how to use python-tensorRT. Frist,i checked my environment of running,and then i followed your readme file to run demo.py .The error messages is as follows:

`/usr/bin/python3.6 /home/qyg/Desktop/yolov5-tensorrt-master/python/lib/demo.py
trtbin /home/qyg/Desktop/yolov5-tensorrt-master/python/lib/models/yolov5s-simple-32.trt
[TensorRT] ERROR: INVALID_CONFIG: The engine plan file is generated on an incompatible device, expecting compute 7.2 got compute 5.3, please rebuild.
[TensorRT] ERROR: engine.cpp (1546) - Serialization Error in deserialize: 0 (Core engine deserialization failure)
[TensorRT] ERROR: INVALID_STATE: std::exception
[TensorRT] ERROR: INVALID_CONFIG: Deserialize the cuda engine failed.
Traceback (most recent call last):
File "/home/qyg/Desktop/yolov5-tensorrt-master/python/lib/demo.py", line 49, in
main()
File "/home/qyg/Desktop/yolov5-tensorrt-master/python/lib/demo.py", line 21, in main
processor = Processor(model=args['model'])
File "/home/qyg/Desktop/yolov5-tensorrt-master/python/lib/Processor.py", line 19, in init
self.context = engine.create_execution_context()
AttributeError: 'NoneType' object has no attribute 'create_execution_context'

Process finished with exit code 1
`

[sunanlin13174]

i also meets another little problem. i just run export_tensorrt.py in jetson xavier nx devices,and i got this meassages:

[TensorRT] VERBOSE: *************** Autotuning format combination: Float(1,320,102400,3276800) -> Float(1,320,102400,3276800) *************** [TensorRT] ERROR: /home/jenkins/workspace/TensorRT/helpers/rel-7.1/L1_Nightly_Internal/build/source/rtSafe/resources.h (462) - Cuda Error in loadKernel: 500 (named symbol not found) [TensorRT] WARNING: GPU memory allocation error during getBestTactic: LeakyRelu_42 [TensorRT] VERBOSE: *************** Autotuning format combination: Float(1,320,102400:32,102400) -> Float(1,320,102400:32,102400) ***************
i am sorry for that brought you trouble,if you are not busy, i hope you can give me the solution.

[lnn523716]

i also meets the problem. i just run demo.py using yolov5s-simple-16.trt in jetson TX2 devices,the result is OK.But, i use yolov5s-simple-32.trt and got this meassages:
trtbin /home/nvidia/Documents/yolov5-tensorrt-master/python/lib/models/yolov5s-simple-32.trt
[TensorRT] ERROR: INVALID_CONFIG: The engine plan file is generated on an incompatible device, expecting compute 6.2 got compute 5.3, please rebuild.
[TensorRT] ERROR: engine.cpp (1546) - Serialization Error in deserialize: 0 (Core engine deserialization failure)
[TensorRT] ERROR: INVALID_STATE: std::exception
[TensorRT] ERROR: INVALID_CONFIG: Deserialize the cuda engine failed.
Traceback (most recent call last):
File "demo_video.py", line 37, in
main()
File "demo_video.py", line 15, in main
processor = Processor(model='yolov5s-simple-32.trt')
File "/home/nvidia/Documents/yolov5-tensorrt-master/python/lib/Processor.py", line 19, in init
self.context = engine.create_execution_context()
AttributeError: 'NoneType' object has no attribute 'create_execution_context'

[naotar]

It is interesting that we are in 2021 and this error still exists with latest downloads and model.

AttributeErrorTraceback (most recent call last)
in
1 from jetbot import ObjectDetector
2
----> 3 model = ObjectDetector('saved_model.pb')

/usr/local/lib/python3.6/dist-packages/jetbot-0.4.3-py3.6.egg/jetbot/object_detection.py in init(self, engine_path, preprocess_fn)
27 load_plugins()
28 self.trt_model = TRTModel(engine_path, input_names=[TRT_INPUT_NAME],
---> 29 output_names=[TRT_OUTPUT_NAME, TRT_OUTPUT_NAME + '_1'])
30 self.preprocess_fn = preprocess_fn
31

/usr/local/lib/python3.6/dist-packages/jetbot-0.4.3-py3.6.egg/jetbot/tensorrt_model.py in init(self, engine_path, input_names, output_names, final_shapes)
57 with open(engine_path, 'rb') as f:
58 self.engine = self.runtime.deserialize_cuda_engine(f.read())
---> 59 self.context = self.engine.create_execution_context()
60
61 if input_names is None:

AttributeError: 'NoneType' object has no attribute 'create_execution_context'