Readme might be missing some steps

[westerdaled]

I have connected my local VSCode to my CodeSpace instance.
Going through the initial set up, I found a couple of issues:

• Do we generate a classic or fine grained token - I chose fine grained
• For the python file that reads the token, I created a test.py in the project root.
• in the terminal
  - pip install python-dotenv however, on running the script I am getting an error on from dotenv import load_dotenv
  I have seen a error referencing installing .NET 9.0 sdk so I have went through these instructions: install .net
• closed VSCode and will retry tomorrow.

[github-actions]

👋 Thanks for contributing @westerdaled! We will review the issue and get back to you soon.

[PaskalSunari]

I was facing the same issue.. Clarification is inevitable in Readme about the file name for accessing the token while running the python scripts

[westerdaled]

@PaskalSunari , I agree with you. I am getting a bit further using in Codespaces. I think the confusion can be that things that are optional for a local install seem already installed in codespaces such as MInicoda. However, the setup, next suggests Jupyter notebook which I am not running in a separate tab . Hopefully, why I need this will become clearer as I explore the course 😕.

[slacker-kb]

I can't get any of the example notebooks to execute properly locally at all, even after installing all the required packages. Codespaces help but not everything runs there either and it's slow. The video content is fine but trying things out on your own does not really work. Still can't get tiktoken to work at all locally.

[koreyspace]

Thanks for reporting @westerdaled - will look to improve onboarding here.

@slacker-kb can you share what errors you are seeing?

[slacker-kb]

For example, lesson 04 I keep getting errors for items I have installed.

oai-assignment.ipynb

Exercise 1, import error on tiktoken. Not found. I've properly installed.
Exercise 2, not found openai. I've properly installed.

I've tried installing the rust compiler first (which was a reported issue with tiktoken). I've installed with pip and with conda. I've created my .env file properly. If the first lesson that uses these items doesn't work, it doesn't motivate learners to move on.

EDIT: If I run 'pip list' the modules are there. Same with 'conda list'.