-
Notifications
You must be signed in to change notification settings - Fork 353
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
No GPU/TPU found, falling back to CPU. #30
Comments
Same error here
Using device: cuda
No GPU/TPU found, falling back to CPU. (Set TF_CPP_MIN_LOG_LEVEL=0 and rerun for more info.) |
stupid me. It is tensorflow!
and fixed |
For people who have no problems with their tensorflow installations - get a jaxlib compiled against the appropriate CUDA and CUDNN libraries, e.g.:
|
Hey all! Super sorry to hear it's been so tough to install JAX on GPU - I would definitely recommend reading through the official JAX installation guide to see if there are any hardware specific instructions: https://github.com/google/jax#installation If your problems persist, by all means open an issue on the JAX repo and request help for your GPU set-up. You can quickly test that you've got a proper installation of JAX on GPU/TPU with the following Python code: import jax
num_devices = jax.device_count()
device_type = jax.devices()[0].device_kind
print(f"Found {num_devices} JAX devices of type {device_type}.") I would only proceed to install Whisper JAX once this check gives you the correct output |
Simply using CUDA and CUDNN installed from pip wheels worked for me (NVIDIA GPU):
|
This is how I installed JAX previously on GPU as well @luisroque - you just need to match your JAX installation to your CUDA version! |
Assuming you have matching video drivers, cuda, and cudnn from Nvidia website. Don't install torch! It will currently downgrade cudnn to an incompatible version.
This is why we use venv or conda. Put torch in a separate container. |
Configuration:
GPU:3090
Driver Version: 525.105.17
CUDA Version: 12.0
but I find the problem that never find the GPU, How to solve this problem?
The text was updated successfully, but these errors were encountered: