Fixed typeset in (optional) policy gradient theorem

[vicha-w]

Hi,

I am currently following HuggingFace's Deep RL course. The course is great so far, but I can't help but notice an error in the math typeset in optional chapter about the policy gradient theorem. I have fixed that in this pull request.

However, I believe we can have better typeset on the tutorial just like in calculus textbook, but I need to understand the typeset system you are using in the tutorials to work further, especially how does it show mathematical equations. I have an extensive experience in typesetting with LaTeX, so I think I can help if the syntax closely resembles that language.

Please let me know if you would like me to help with the typeset. I would love to see a nice, textbook-like typeset in this optional chapter.

Thanks! Vichayanun

[simoninithomas]

Hey there

Our latex is like this \( YOUR LATEX HERE //)

But your change lead to a problem, you removed the content in "we can rewrite as"

[simoninithomas]

I run a build PR so that you can see what it looks like (it will take 5min to build)

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

[vicha-w]

Hi Thomas,

I have checked the latest build, and I do not see the problem with my change, so let me explain what I fixed in this PR.

Here is the text in the current build, live on HuggingFace.

As you can see, there is excess text after "Thus we can rewrite the sum as" which I have highlighted in the red box and should be moved to the previous line. Here is the fixed version on CI:

Please let me know if that does not actually work in this case.

Thanks! Vichayanun

[simoninithomas]

Oh yes my bad, I merge it in that case thanks for the update 🤗