The engine plan file has a layer that utilizes loop API or boolean tensors, but the current platform does not support these layers due to mismatch of libnvrtc version.
Dameng23333 opened this issue · 18 comments
[01/24/2023-00:50:48] [TRT] [E] 9: The engine plan file has a layer that utilizes loop API or boolean tensors, but the current platform does not support these layers due to mismatch of libnvrtc version.
[01/24/2023-00:50:48] [TRT] [E] 1: [engine.cpp::nvinfer1::rt::Engine::deserialize::762] Error Code 1: Serialization (Serialization assertion postDeserializationCheck() failed.Post deserialization check failure)
[01/24/2023-00:50:48] [TRT] [E] 4: [runtime.cpp::nvinfer1::Runtime::deserializeCudaEngine::50] Error Code 4: Internal Error (Engine deserialization failed.)
Traceback (most recent call last):
File "D:\autodrive\script\main.py", line 54, in
clrnet = CLRNet("D:/autodrive/Perception/LaneDetection/weights/llamas_dla34.engine")
File "D:\autodrive/Perception\LaneDetection\clrnet_trt.py", line 75, in init
self.context = self.engine.create_execution_context()
AttributeError: 'NoneType' object has no attribute 'create_execution_context'
TensorRT 版本 8.4.3.1，CUDA 版本 10.2
请教大佬们如何解决此问题。万分感激。
你看我昨晚的求助，里面有办法。不能直接用大佬的engine文件，要自己转换。
是自己转换的engine文件
@Dameng23333 那不应该啊，我按你升级8522转的也识别不到。是用8424转的。然后就成功了。
8522转是同样报错吗
@Dameng23333 不是onnx_graphsurgeon的版本问题吗
已经更新到最新版，转换过程中没有任何报错
@Dameng23333 `but the current platform does not support these layers due to mismatch of libnvrtc version'是不是显卡不支持
1050ti不可以吗？还是我显卡驱动的问题？
@Dameng23333 我在网上搜了一下，还真没找到这个问题，我去问问
命令行输入 nvcc -V 看看cuda在Path里面配置的是哪个版本
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2019 NVIDIA Corporation
Built on Wed_Oct_23_19:32:27_Pacific_Daylight_Time_2019
Cuda compilation tools, release 10.2, V10.2.89
输入 nvidia-smi 呢
Wed Jan 25 20:07:57 2023
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 528.02 Driver Version: 528.02 CUDA Version: 12.0 |
|-------------------------------+----------------------+----------------------+
| GPU Name TCC/WDDM | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|===============================+======================+======================|
| 0 NVIDIA GeForce ... WDDM | 00000000:01:00.0 On | N/A |
|100% 46C P0 N/A / 100W | 2393MiB / 4096MiB | 55% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
贴一下 pip list
Package Version Editable project location
absl-py 1.4.0
addict 2.4.0
altair 4.2.0
appdirs 1.4.4
astor 0.8.1
astunparse 1.6.3
attrdict 2.0.1
attrs 22.2.0
Babel 2.11.0
bce-python-sdk 0.8.74
beautifulsoup4 4.11.1
blinker 1.5
cachetools 5.2.1
certifi 2022.12.7
charset-normalizer 3.0.1
click 8.1.3
colorama 0.4.6
coloredlogs 15.0.1
comtypes 1.1.14
contourpy 1.0.7
cssselect 1.2.0
cssutils 2.6.0
cycler 0.11.0
Cython 0.29.33
cython-bbox 0.1.3
decorator 5.1.1
dill 0.3.6
dxcam 0.0.5
entrypoints 0.4
et-xmlfile 1.1.0
fire 0.5.0
Flask 2.2.2
flask-babel 3.0.0
flatbuffers 23.1.4
fonttools 4.38.0
future 0.18.3
gast 0.4.0
gitdb 4.0.10
GitPython 3.1.30
google-auth 2.16.0
google-auth-oauthlib 0.4.6
google-pasta 0.2.0
graphsurgeon 0.4.6
grpcio 1.51.1
h5py 3.7.0
humanfriendly 10.0
idna 3.4
imageio 2.24.0
imgaug 0.4.0
importlib-metadata 6.0.0
itsdangerous 2.1.2
Jinja2 3.1.2
joblib 1.2.0
jsonschema 4.17.3
keras 2.11.0
kiwisolver 1.4.4
lap 0.4.0
libclang 15.0.6.1
lmdb 1.4.0
logitech-steering-wheel 1.0
lxml 4.9.2
Mako 1.2.4
Markdown 3.4.1
markdown-it-py 2.1.0
MarkupSafe 2.1.1
matplotlib 3.6.3
mdurl 0.1.2
mmcls 0.25.0 c:\users\azhen\downloads\mmclassification
mmcv 1.7.1
mpmath 1.2.1
mss 7.0.1
multiprocess 0.70.14
networkx 3.0
numpy 1.22.0
oauthlib 3.2.2
onnx 1.13.0
onnx-graphsurgeon 0.3.12
onnxruntime 1.13.1
opencv-contrib-python 4.7.0.68
opencv-python 4.7.0.68
openpyxl 3.0.10
opt-einsum 3.3.0
packaging 23.0
paddle-bfloat 0.1.7
paddleocr 2.6.1.2
paddlepaddle 2.4.1
paddlepaddle-gpu 2.4.1.post112
pandas 1.5.2
pdf2docx 0.5.6
Pillow 9.4.0
platformdirs 2.6.2
playsound 1.2.2
premailer 3.10.0
protobuf 3.20.3
pyarrow 10.0.1
pyasn1 0.4.8
pyasn1-modules 0.2.8
pyclipper 1.3.0.post4
pycocotools 2.0
pycryptodome 3.16.0
pycuda 2022.1+cuda116
pydeck 0.8.0
pygame 2.1.2
PyGetWindow 0.0.9
Pygments 2.14.0
Pympler 1.0.1
PyMuPDF 1.20.2
pynput 1.7.6
pyparsing 3.0.9
pyreadline3 3.4.1
PyRect 0.2.0
pyrsistent 0.19.3
python-dateutil 2.8.2
python-docx 0.8.11
pytools 2022.1.14
pytz 2022.7.1
pytz-deprecation-shim 0.1.0.post0
pyvjoy 1.0.1
PyWavelets 1.4.1
pywin32 305
PyYAML 6.0
rapidfuzz 2.13.7
regex 2022.10.31
requests 2.28.2
requests-oauthlib 1.3.1
rich 13.2.0
rsa 4.9
scikit-fuzzy 0.4.2
scikit-image 0.19.3
scikit-learn 1.2.0
scipy 1.10.0
semver 2.13.0
shapely 2.0.0
six 1.16.0
sklearn 0.0.post1
smmap 5.0.0
soupsieve 2.3.2.post1
streamlit 1.17.0
sv-ttk 2.4
sympy 1.11.1
tensorboard 2.11.2
tensorboard-data-server 0.6.1
tensorboard-plugin-wit 1.8.1
tensorflow 2.11.0
tensorflow-estimator 2.11.0
tensorflow-intel 2.11.0
tensorflow-io-gcs-filesystem 0.30.0
tensorrt 8.4.2.4
termcolor 2.2.0
tflite 2.10.0
threadpoolctl 3.1.0
tifffile 2022.10.10
tk 0.1.0
toml 0.10.2
toolz 0.12.0
torch 1.13.1+cu117
torchvision 0.1.6
tornado 6.2
tqdm 4.64.1
transitions 0.9.0
ttkwidgets 0.13.0
typing_extensions 4.4.0
tzdata 2022.7
tzlocal 4.2
urllib3 1.26.14
validators 0.20.0
vgamepad 0.0.8
visualdl 2.4.2
watchdog 2.2.1
Werkzeug 2.2.2
wheel 0.38.4
wrapt 1.14.1
yapf 0.32.0
zipp 3.11.0
感觉pip的包有点不太对劲，比如 pycuda 2022.1+cuda116，我本地的是 pycuda 2022.2.2。其他的没仔细看 torch 1.13.1+cu117 这都版本也跟cuda的版本不一样啊
感觉pip的包有点不太对劲，比如
pycuda 2022.1+cuda116，我本地的是pycuda 2022.2.2。其他的没仔细看torch 1.13.1+cu117这都版本也跟cuda的版本不一样啊
可不可以贴一下你的 各个所需依赖版本呀? 我的跟题主的差不多, 因为pycuda安装不上, 只能从网上下载下来然后本地安装.
目前我的环境应该是部署好了, 但是, 我不知道是对还是不对, 已经可以侦测到, 不过不知到是还要设置vjoy?