Day 4/100 — Gradient Descent.
Quick question —
Do you think you need to build an LLM
from scratch to become an AI engineer?
You don't.
Your job is to use the tools.
And solve real problems.
Unless you want to be a research scientist.
That's a different path.
Code → in bio
Day 3/100 — Cost Function.
How does a model know it's wrong?
It measures the gap. Squares it. Averages it.
Start : 0.500
End : 0.121
Reduction : 75.6%
1000 iterations. No instructions. Just math.
Code → [GitHub link]
#100DaysOfML#buildinpublic