Tensors intuition lab
A gentle lab for **tensors** as multi-dimensional arrays—shapes, batching, and why frameworks care—without requiring GPU expertise.
What it is
A gentle lab for tensors as multi-dimensional arrays—shapes, batching, and why frameworks care—without requiring GPU expertise.
Why it matters
Unblocks reading DL code and Course 05/07 technical tabs.
How it works (plain)
Scalar → vector → matrix → batch of matrices. Practice reshaping mentally; run a tiny numpy/torch snippet if installed.
Try it
Write shapes for: 32 RGB 64×64 images (32,3,64,64) and explain each dim.
Myths
- ⚠️ Myth: You must master all of linear algebra first.
- ✓ Reality: Shape literacy unlocks most labs.
Sources
- Course 04 vectors/matrices; Course 21 setup
- NumPy/PyTorch quickstarts (cite specifically)