ml advice
- Start with cs229 and math
- learn fundamentals and core
- learn math - algebra, calculus, probability and statistics, etc
- don't neglect math, math is core
- don't get in because of the hype
- knowing how to build a rag system doesn't make you an ml engineer
- learn to read research papers
- implement papers from scratch
- the hype about building agents is not ML, it's applied ai
- you don't become a core ml engineer until you do a masters or research thesis, some exceptions might include
- If you just wanna follow one path, do fastAI course
- if you want to learn about llms, watch karpathy
- learn data analysis and visualization
- learn transformers
- learn mlops
- finding ml jobs are much harder than sde/swe jobs
- build something (period)
- train an ml model from scratch
- write an algorithm from scratch
Read sankalp's blog for indepth advice, my version is still under construction