# How to use GPU’s when training a model?

## post by benthehen on May 8, 2023

I am sure this is a very simple question for most, so apologies for the dumb question.

I am starting work on a lambda vector GPU workstation.

My question is: given some code to train a model (in this case using tensorflow), how do I make sure the GPU’s are utilized while training this code?

Thank you for the help!

## post by Calvin on May 9, 2023

Hi,

You may find our tutorials in our blog helpful in this regard, especially since we have a short series specifically for Tensorflow.

Also, here is a Python command to use Tensorflow to ensure Tensorflow can see your GPUs:

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

Otherwise you’ll need to edit your code/model to ensure the GPU(s) are being utilized.

If this doesn’t answer your question, let us know here or you can also open a support ticket here:

Regards,

Calvin Wallace

Linux Support Engineer

## post by benthehen on May 9, 2023

Running the code snippet shows:

Num GPUs Available: 0

How do I go about ensuring that tensorflow can see my GPU’s??

## post by yanos on May 11, 2023

Can you see the GPU(s) when you run “nvidia-smi”?  
Also, do you by any chance have overridden any of the Lambda Stack installed Python packages?  
Can you share the output of “pip -v list”?