Project
Systems of Ashish
Math Explainer Agent
This was converted into research by University of Waterloo
pythonjavascript
Systems of Ashish
This was converted into research by University of Waterloo
More things from the same systems rabbit hole.
Gemma-style Per-Layer Embeddings trained from scratch. Every transformer layer gets its own tiny token table, streamed per-token from a memory-mapped file. Beats dense at matched inference VRAM — 149.88 PPL vs 160.39 PPL.
A Rust-based LLM inference engine focused on efficient model execution, token generation, and low-level systems performance.
🚀 Introducing NamasteScreen: The Launcher That Fights Phone Addiction! 🇮🇳 🔥 Inspired by my conversation with Hon'ble Prime Minister of India Narendra Modi about Digital Well-Being