Tensor Flow not working in Jupyter

General Discussion
,
#1

I’m in Linux Ubuntu 22.04, I have an Nvidia RTX 4070 TI which I bought on purpose to use in a Data Science Bootcamp I’m enrolled in.

As you can see here, I’ve installed TensorFlow, and everything seems to be OK.

image
image1408×281 90.5 KB

But I don’t know why is not being used or detected.

image
image1401×740 278 KB

Who can help me to fix it?

Thanks

#3

Hi @EMR_TF_DSA, Could you please let us know which version of tensorflow you are using. And steps followed to install tensorflow and cuda. Thank You.

#5

Hi, I’m using Version: 2.15.0.post1.

I executed both in the terminal and Jupyter next command: pip install tensorflow

I installed CUDA also. To be honest, I mixed different sources as I’m not clear at all following the installation guide in Instalar TensorFlow con pip, what does it mean the [and-cuda] in pip install tensorflow[and-cuda]?

I found also other installation guides in TensorFlow mixing version topics, etc. :frowning:

Could you please share some clear steps?

Thanks

#7

Hi @EMR_TF_DSA, Could you please try to create a new environment and follow the below steps

"!pip install --extra-index-url https://pypi.nvidia.com tensorrt-bindings==8.6.1 tensorrt-libs==8.6.1
!pip install -U tensorflow[and-cuda]==2.15.0
import tensorflow as tf; print(tf.__version__); print(tf.config.list_physical_devices('GPU'))"

Thank You.

#9

I’ll try and let you know, this means that I’ll have to install again all the modules I’m using in the BootCamp, right?

I’m also using Rapids, can I install them both in the same environment?

Thanks

#10

Hi,

I created a new environment in anaconda and ran JupyterLab.

  1. !pip install --extra-index-url https://pypi.nvidia.com tensorrt-bindings==8.6.1 tensorrt-libs==8.6.1

Result apparently OK.

  1. !pip install -U tensorflow[and-cuda]==2.15.0

Result NOK.

  1. But:
    !pip install -U tensorflow==2.15.0

After all the downloads, etc seemed to be OK. Removed the “[and-cuda]”, which I don’t know what’s for or how to execute it…

  1. Unfortunatelly:

import tensorflow as tf; print(tf.__version__); print(tf.config.list_physical_devices('GPU'))

2023-12-20 23:42:10.248207: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
2023-12-20 23:42:10.248232: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
2023-12-20 23:42:10.249226: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
2023-12-20 23:42:10.254802: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
To enable the following instructions: AVX2 FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
2023-12-20 23:42:10.876303: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
2.15.0
[]
2023-12-20 23:42:11.305619: I external/local_xla/xla/stream_executor/cuda/cuda_executor.cc:901] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
2023-12-20 23:42:11.334770: W tensorflow/core/common_runtime/gpu/gpu_device.cc:2256] Cannot dlopen some GPU libraries. Please make sure the missing libraries mentioned above are installed properly if you would like to use GPU. Follow the guide at https://www.tensorflow.org/install/gpu for how to download and setup the required libraries for your platform.
Skipping registering GPU devices...

So result was NOK.

Any suggestion further?

Thanks!

PS: I can’t add screenshots…sorry, if not I’d be showing the results in screen.

#12

Hi, who can help further?

I don’t know what to do and I’ve spent a lot of money on this to be unable to use it :disappointed_relieved:

Thanks!