Python learning roadmap for someone with a machine learning background

Community
1

Python experts - seeking your advice, I have ML experience but need to improve my Python skills. For those who transitioned from ML to Python proficiency, what was the optimal learning roadmap? What resources and projects were most important for quickly gaining competency? Want to avoid pitfalls and structure my learning effectively. Thanks :grin:

2 Likes
2

it depends much on your current level of python knowledge. not much on your ML knowledge. I enjoyed studying “Fluent Python” by Ramalho, but I recenlty looked up the newer edition, and I think it’s become a bit too heavy.

If you want to advance your allround skills in python, an advanced learner book is a good idea I’d say.

2 Likes
3

Adding on to what @foschi has said, ChatGPT is an excellent python tutor, especially when used into combination with a good tutorial.

1 Like
4

Agreed,

Focus on making projects, it’s a great way to learn and you’ll have something to show at the end :laughing:

1 Like
5

Oh yeah, nice advice… But am into frontend at the same time, currently learning JavaScript and python together
:bomb: , feel like focusing on python before coming back to JS, or what do you think?:eyes:

1 Like
6

I strongly agree with that👍

1 Like
7

(Am I the only one that “Aug 21” looks like “OMG someone bumped a 2021 thread?”)

I used jailbroken Bing to write out to max_tokens for me then continue as an experiment. I could have it finish more than the three chapters I did.

image
image569×760 44.7 KB

image
image512×910 36 KB

You can have gpt-4 write the TOC for such a course and see where you need to pick up, pytorch transformers or such.

2 Likes
8

:joy: , this was my first post here, at that time it got no reply, don’t know how it came up now.
hmm, nice, Thanks by the way, I’ll try gpt-4, don’t know how to jailbreak Bing :sweat_smile:

1 Like
9

There are a couple more angles here.

One is writing AI systems, from scratch, directly in Python. For example:

After months of developing on this initial toy example, I was training RRN’s, converted them to use embeddings instead of 1-hot encodings, and served inference from my own custom Lambda function in AWS.

Another route, that doesn’t go a deep into one topic, but gives you a general overview, would be check out books like " Generative Deep Learning: Teaching Machines To Paint, Write, Compose, and Play 2nd Edition" by David Foster.

Here it’s more survey oriented and it comes with Tensorflow/Keras code (Python code). So you can build advanced models, but it’s going to be hard to dig deeper to get the understanding since these frameworks are at such a high level of abstraction.

2 Likes
10

Thank you for this. To be frank, I have a preference for courses over books. Could you please suggest any courses?

1 Like
11

AI courses or Python courses?

For Python, I remember liking the courses from Dr. Fred Baptiste on Udemy. Like his Python 3: Deep Dive series, here is Part 1:

https://www.udemy.com/course/python-3-deep-dive-part-1/

If new to programming, then maybe try his Fundamentals class:

https://www.udemy.com/course/python3-fundamentals/

As for AI courses, there are so many on Udemy, take your pick! Don’t forget about YouTube as well!

2 Likes
12

Thank you so much for the recommendations. I’ll definitely look into both AI and Python courses.

2 Likes