Installing Python, Jupyter Lab and Nvidia CUDA on Your Gaming PC

troll constructing foundation of an ancient castle, outdoors, evening, enchanting, digital paint

Getting my PC ready to do some AI LLM training. First step, make sure the essentials, Python, Jupyter Lab and Nvidia CUDA, are installed.

Python is the easiest thing to install. Simply visit the python website and use the easy-to-use installer for windows.

Download Python | Python.org

The main thing is to ensure that it is actually installed. Also, I would recommend you leave everything on defaults. (in the VS Code terminal or any cmd line)

PS C:\Users\vijay\Videos\PythonPrivate> python 
Python 3.12.3 (tags/v3.12.3:f6650f9, Apr  9 2024, 14:05:25) [MSC v.1938 64 bit (AMD64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> exit()
PS C:\Users\vijay\Videos\PythonPrivate> 

Once that is done, let’s take care of Jupyter Lab. I recently wrote about the many ways you can run the JLab.

The Myriad Ways to Run Jupyter Lab Notebook for Python

Now that python is there, you can simply use the pip option.

pip install jupyterlab

and then, simply launch it.

 python -m jupyterlab 

I want to note something here. some sites ask you to do this.

jupyter lab //this does not work. use the command above. 

so, do the command that works, and you should have something like this.

screenshot of running Jupyter Lab

But, normally, I just VS Code to use Jupyter notebooks, but, sometimes, I also use Jupyter Lab.

The last item, for training LLMs (and of course, there are more things you will need, but, I have only gotten till here in my AI education) are the CUDA drivers.

CUDA Toolkit 12.5 Downloads | NVIDIA Developer

One thing I do want to mention is, sometimes, the server is simply too slow. It took me many days to actually download this driver. Perhaps, it was just my bad luck, or I don’t know. Also, I recommend the network version instead of the local version.

The local version just did not download for me after multiple attempts.

use the exe network option.

Once done, you have to make sure, it is all working. so, once again, terminal VS Code or any cmd line.

PS C:\Users\vijay\Videos\PythonPrivate> nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2024 NVIDIA Corporation
Built on Wed_Apr_17_19:36:51_Pacific_Daylight_Time_2024
Cuda compilation tools, release 12.5, V12.5.40
Build cuda_12.5.r12.5/compiler.34177558_0
PS C:\Users\vijay\Videos\PythonPrivate>

And that’s all there is to it. I will post most stuff as my AI journey continues.

Hire me at UpWork or Fiverr or codementor or GitHub. more about me on my personal profile, and also on my personal website.