note

When I ran the "Grug-12B" distillation of Gemma-4-12B, which "thinks briefly," on a coding benchmark, I encountered the paradox of token efficiency

This time, I'm looking at a model with a different character. Grug-12B—a so-called compact-reasoning model that fine-tuned google/gemma-4-12B-it to "shorten and densify inference traces." The author's ...
GitHub

longest_palindrome.py

LONGEST PALINDROME word in the file. The longest is "malayalam" with 9 characters. with open("sowpods.txt", "r") as file: words = [line.strip() for line in file ...
GitHub

Notes.md

This course provides a comprehensive introduction to data structures and algorithms used in computer science. Throughout the course, students will learn how to write efficient code, analyze algorithms ...