You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
modify the requirements.txt to explicitly install the GPU-compatible version of TensorFlow by specifying tensorflow-gpu or tensorflow==2.x+nv instead of tensorflow-cpu. Ensure your Dockerfile uses a GPU-enabled base image, such as nvidia/cuda:12.2.0-cudnn8-runtime-ubuntu20.04, and install TensorFlow accordingly. Additionally, use the --gpus all flag when running the container to enable GPU support, and verify that TensorFlow detects the GPU with tf.config.list_physical_devices('GPU').
why the cuda and cudnn installed and the requirements file in docker install tensorflow-cpu version???
The text was updated successfully, but these errors were encountered: