Could not load library cudnn_cnn_infer64_8.dll. Error code 126

[rcapp2506]

I'm trying to run your code but always got the following error:
"2021-12-12 23:48:46.254332: I tensorflow/stream_executor/cuda/cuda_dnn.cc:359] Loaded cuDNN version 8301
Could not load library cudnn_cnn_infer64_8.dll. Error code 126
Please make sure cudnn_cnn_infer64_8.dll is in your library path!"
from within cam_spy.py module.
I've tried many different versions of cudnn library (following some suggestions on the web) but always did not work.
Any idea?
Many thanks

[huckiyang]

@rcapp2506 hi, the error message is an indeed cuda issue not directly related to this setup.

Can you provide your hardware and drive setup? e.g., CUDA version, cudadnn version, tensorflow version and so. I can have a quick check (but not confident).

Noted the QML code is only tested in Linux environment.

[rcapp2506]

I do thank you for immediate answer! I’m running the code from my notebook with Python 3.9.7, tensorflow 2.5 (tried also with 2.7); GPU is a NVIDIA Geforce 960m with NVIDIA toolkit 15.5 with cudnn driver 8.2.4.1 (as suggested from some posts found with the same issue…). I could try to run from my academic machine (Power AI with TESLA P100)… My name is Roberto Cappuccio and teach quantum computation for the Master’s degree in Physics at Pisa, Italy. Many thanks Roberto From: ***@***.***> Sent: Monday, December 13, 2021 12:04 AM To: ***@***.***> Cc: Roberto ***@***.***>; ***@***.***> Subject: Re: [huckiyang/QuantumSpeech-QCNN] Could not load library cudnn_cnn_infer64_8.dll. Error code 126 (Issue #9) @rcapp2506<https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Frcapp2506&data=04%7C01%7C%7C235438c6c6a44553d56408d9bdc3af3b%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637749470417050016%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=qzv2rOKiAM1OZFOk9BI3r0nEJH2uPJxX0DK3a8o8VWM%3D&reserved=0> hi, the error message is an indeed cuda issue not directly related to this setup. Can you provide your hardware and drive setup? e.g., CUDA version, cudadnn version, tensorflow version and so. I can have a quick check (but not confident). Noted the QML code is only tested in Linux environment. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub<https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fhuckiyang%2FQuantumSpeech-QCNN%2Fissues%2F9%23issuecomment-991987796&data=04%7C01%7C%7C235438c6c6a44553d56408d9bdc3af3b%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637749470417050016%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=tdjN%2BhfLXrF4ZxbpFQulFnnNJexTSW1iLmpNlPvRZDg%3D&reserved=0>, or unsubscribe<https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FADYSWFHRFQ3TYFNLSEXLTMLUQUS57ANCNFSM5J4Y4O2A&data=04%7C01%7C%7C235438c6c6a44553d56408d9bdc3af3b%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637749470417050016%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=fIx%2BcVFNmQnRl9bYkh%2FsdVX%2B7KLI5QkIU8cdwly4VhE%3D&reserved=0>. Triage notifications on the go with GitHub Mobile for iOS<https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fapps.apple.com%2Fapp%2Fapple-store%2Fid1477376905%3Fct%3Dnotification-email%26mt%3D8%26pt%3D524675&data=04%7C01%7C%7C235438c6c6a44553d56408d9bdc3af3b%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637749470417050016%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=af3EWyVaoMhzEq9x8bv%2FPZ4UhhjbEAFZR%2BRyFcZQuJM%3D&reserved=0> or Android<https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fplay.google.com%2Fstore%2Fapps%2Fdetails%3Fid%3Dcom.github.android%26referrer%3Dutm_campaign%253Dnotification-email%2526utm_medium%253Demail%2526utm_source%253Dgithub&data=04%7C01%7C%7C235438c6c6a44553d56408d9bdc3af3b%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637749470417050016%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=Rl9nQMwGHxDhW%2Fq%2FRPvDtXsWavztoSfV%2B2pXZibWi9c%3D&reserved=0>.

[huckiyang]

@rcapp2506 hi, it looks like your driver environment is a bit too new and not properly setup tensorflow-gpu.

I recommend you can use ‘conda’ to create a virtual environment with tensorflow-gpu. conda will help you match a correct cuda and cudnn driver.

you can find the conda 101 here
https://carpentries-incubator.github.io/introduction-to-conda-for-data-scientists/aio/index.html

After installation, you should check your tensorflow-gpu working by simple testing

https://intellipaat.com/community/33459/how-to-tell-if-tensorflow-is-using-gpu-acceleration-from-inside-python-shell

In our setup, a tensorflow-gpu=2.0 is used (see the conda setup in readme)

https://github.com/huckiyang/QuantumSpeech-QCNN/blob/main/README.md#1-environment

Cheers,