tf.test.is_gpu_available( cuda_only=False, min_cuda_compute_capability=None
)

If nothing else works (tested on Python 3.11)
1. In conda's base env run:	conda install nvidia::cuda	conda install anaconda::cudnn
2. Export cuda path (anaconda / miniconda):	LD_LIBRARY_PATH=$HOME/anaconda3/lib:$LD_LIBRARY_PATH	or	LD_LIBRARY_PATH=$HOME/miniconda3/lib:$LD_LIBRARY_PATH
3. Create new conda env and install tensorflow:	pip install tensorflow[and-cuda]
4. Verify with:	python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))"
5. Enjoy!

import tensorflow as tf
print("Num GPUs Available: ", len(tf.config.experimental.list_physical_devices('GPU')))

python -c "import tensorflow as tf; print('Num GPUs Available: ', len(tf.config.experimental.list_physical_devices('GPU')))"

{"tags":[{"tag":"textarea","content":"I have tried alot to install tf-gpu but I always get into errors!\nSo after a lot of brainstorming here is few steps for you to install tensorflow-gpu\n\nWindows 10 or 11 Only...\n\n1. Install Nvidia Driver \n\ta. Download Automatic Driver Updates File form here :-\n \thttps://www.nvidia.com/en-in/geforce/drivers/\n It will automatically finds best suitable driver for your system\n \n2. Install wsl(Windows Subsystem for linux) on your windows 10 or greater\n\ta. open powershell or cmd in adminstrator\n b. wsl --install\n c. Enter credentials and usernames(Imp to remember)\n d. Reboot System\n \n3. Install Anaconda\n\ta. Open cmd as user not adminstrator and type wsl\n b. type \"curl https://repo.anaconda.com/archive/Anaconda3-2021.11-Linux-x86_64.sh --output anaconda.sh\"\n c. type \"bash anaconda.sh\" in wsl\n d. Enter and say yes to all like[Do you approve the license terms? [yes|no], Do you wish the installer to initialize Anaconda3]\n e. activate installation by type \"source ~/.bashrc\"\n f. Now you will get (base)\n \n4. Environment and Cuda, Cudnn(GO on pasting every line in (base))\n\ta. conda create --name tf python=3.9\n b. conda activate tf\n c. nvidia-smi (To Check GPU )\n d. conda install -c conda-forge cudatoolkit=11.8.0\n e. pip install nvidia-cudnn-cu11==8.6.0.163\n f. mkdir -p $CONDA_PREFIX/etc/conda/activate.d\n g. CUDNN_PATH=$(dirname $(python -c \"import nvidia.cudnn;print(nvidia.cudnn.__file__)\"))\n h. echo 'export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CONDA_PREFIX/lib/:$CUDNN_PATH/lib' > $CONDA_PREFIX/etc/conda/activate.d/env_vars.sh\n\t\n5. Installing Tensorflow in env\n\ta. pip install --upgrade pip\n b. pip install tensorflow==2.12.*\n c. python3 -c \"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\" (To Check GPU)\n \n6. Install Jupyter in base and not in environment\n\ta. conda install -y jupyter\n \nProblems:-\n\ta. No GPU detected after running 5. and c. command:-\n \tRerun 4. g. and h. every time when you launch or activate environment i.e tf\n ","code_language":"whatever"}]}

from __future__ import absolute_import, division, print_function, unicode_literalsimport tensorflow as tfprint("Num GPUs Available: ", len(tf.config.experimental.list_physical_devices('GPU')))