DNN library is not found

[rober...@gmail.com]

Dear users,

 

It seems recently there was an update in SEPAL.  I want to train a TensorFlow model but I am getting an error “DNN library is not found”. When I checked the directory  “/usr/local/cuda-11.4/ “  I did not see the library

 libcudnn8.

Below is the error I get when I try to train the model seems that it doesn't load properly the old composites, it takes a long time and it's not possible to visualize them.

 

2023-12-29 21:28:59.517317: W tensorflow/core/framework/op_kernel.cc:1839] OP_REQUIRES failed at conv_ops_impl.h:1592 : UNIMPLEMENTED: DNN library is not found.

 

Thanks for your help.

[Masolele, Robert]

------

Dr. Robert N. Masolele

Postdoctoral Researcher | Laboratory of Geo-Information Science and Remote Sensing

Environmental Sciences Group, Wageningen University & Research 

Postbus 47, 6700AA, Wageningen 

Wageningen Campus, Building 101 (Gaia), Room C.319 

Droevendaalsesteeg 3, 6708PB, Wageningen 

Phone: +31 (0)622178455

[Andreas Vollrath]

Dear Robert, 

indeed there was an important update of the underlying OS, the whole Tensorflow dependencies is a bit tricky and it seems not everything is yet in place. We'll work to get this sorted asap. 

Apologies, and thanks for your patience,

Andreas

[Masolele, Robert]

Hi Andreas,

 

It looks like the issue of TensorFlow dependencies has not been solved. Do you have plan to work on it? i.e  “DNN library is not found.”?

 

Thanks for your response.

Robert Masolele

 

 

------

Dr. Robert N. Masolele

Postdoctoral Researcher | Laboratory of Geo-Information Science and Remote Sensing

Environmental Sciences Group, Wageningen University & Research 

Postbus 47, 6700AA, Wageningen 

Wageningen Campus, Building 101 (Gaia), Room C.319 

Droevendaalsesteeg 3, 6708PB, Wageningen 

Phone: +31 (0)622178455

 

--
You received this message because you are subscribed to the Google Groups "Sepal Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sepal-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sepal-users/e50e2f78-d1b4-4094-b976-595c88c1b9e9n%40googlegroups.com.

[John Dilger]

Hi,

Has there been any updates or workaround found? I'm running into the same issue with the DNN library.

Best,

John

[John Dilger]

A small update. I think the root issue is that CUDAv11.7 is not supported by Tesnsorflow. 

https://github.com/tensorflow/addons?tab=readme-ov-file#c-custom-op-compatibility-matrix

I haven't found a workaround for Tesnorflow but I did get Pytorch running and using the GPU.

https://pytorch.org/get-started/previous-versions/#linux-and-windows-12

conda create -n torch python=3.11 -y

conda activate torch

conda install pytorch==2.0.1 torchvision==0.15.2 torchaudio==2.0.2 pytorch-cuda=11.7 -c pytorch -c nvidia