sighthough

Here are 8 public repositories matching this topic...

in this repository we discuss the way to train a prefixed LLM network size .

training optimization gemini claude llm chatgpt qwen deepseek nano-banana sighthough

a fast way to do zero drift calculations using base 2520

fast finance optimization optimized calculations drift precision zero high sighthough

so the main idea was what if instead of binary we had 4bit dibit utilizing 2 transistors at a time so it can run on existing hardware

optimization wizard speedtest bandwidth tweak speedup 4bit wizard-steps 4bits wizardry wizard-of-oz wizard101 gemini-ai sighthough

its good

fast algorithms gemini sorting-algorithms vibe-coding sighthough

This is an llm training engine that makes an llm learn things by adding neurons and layers dynamically as it learns

this is a sceduler optimization so that it runs smooth while still saving power (improves stability on 90-95-99 percent lows too in 3d)

linux kernel optimization tweaks kernel-driver tweak optimization-algorithms windows-11 vibe-coding sighthough sceduler

its a physics based ai neural network concept trying to simulate a human brain

ai wizard wizardry wizard-of-oz wizarding-world wizard101 gemini-ai sighthough

A continuous physics-based alternative to traditional grid-search algorithms.

pathfinding gemini physics-simulation djkstra vibe-coding sighthough

To associate your repository with the sighthough topic, visit your repo's landing page and select "manage topics."