I am an experienced ML programmer, but new to DL and Python.
During part 1 of the course i used google.colab, but now i’m looking to really dig in and start coding more myself. Seems like in 2024 I should be using a tool with an AI coding assistant such as copilot or openAI codex. What is the best setup available these days?
I have a MacbookPro w/Apple M1 Max. Can I run everything (py, jupyter, gpu) locally via VScode/copilot? or do any of the GPU cloud hosts offer a copilot?
I have tried a bunch of cloud dev environments like sagemaker, lightning ai, etc. I still think the best way is to do as much as possible locally and then spin up a gpu ec2/sagemaker instance and connect with the machine from local vs code. I use cursor as my local vs code ide which makes ai powered dev much easier and has a lot of quality of life features.
So i took your advice and set up VS code with github copilot. I have set up a local python environment and paperspace remote for GPU. I have to say as an experienced dev, but new to python, copilot is truly game changing! it’s just table stakes for learning and coding. The initial setup is a bit painful, but once you get it going its very high ROI.
I wil try cursor next.