karpathy

LLM Wiki

A pattern for building personal knowledge bases using LLMs.

This is an idea file, it is designed to be copy pasted to your own LLM Agent (e.g. OpenAI Codex, Claude Code, OpenCode / Pi, or etc.). Its goal is to communicate the high level idea, but your agent will build out the specifics in collaboration with you.

The core idea

Most people's experience with LLMs and documents looks like RAG: you upload a collection of files, the LLM retrieves relevant chunks at query time, and generates an answer. This works, but the LLM is rediscovering knowledge from scratch on every question. There's no accumulation. Ask a subtle question that requires synthesizing five documents, and the LLM has to find and piece together the relevant fragments every time. Nothing is built up. NotebookLM, ChatGPT file uploads, and most RAG systems work this way.

soham2008xyz

iTerm2 Ollama Integration

A guide on using Ollama as the OpenAI API provider for inline completions in iTerm2.

Configuration

• API URL: http://127.0.0.1:11434/v1/completions
• Model: mistral
• Tokens: 4000
• Use legacy completions API: true

davidsteppenbeck

import SwiftUI

enum PreviewProviderMode: CaseIterable {
    
    /// Use for a light appearance preview.
    case lightMode
    
    /// Use for a dark appearance preview.
    case darkMode
    

Dimillian

struct RootView: View {
    @State private var isHomeShown = true
    @State private var selectedContent = "content1"
    
    var body: some View {
        ZStack {
            ContentView(content: $selectedContent,
                        isHomeShown: $isHomeShown)
                .blur(radius: isHomeShown ? 10 : 0)
                .scaleEffect(isHomeShown ? 0.8 : 1)

Peredery

How to fix "gpg failed to sign the data fatal: failed to write commit object" in MacOS

!!! For M1/M2 apple silicon see this comment:

https://gist.github.com/Peredery/38d0538dd34381bbd9d13414269a1f27?permalink_comment_id=4559612#gistcomment-4559612

For MacOS | Mojave | High Sierra

Step 1 - Upgrade current gpg

mrk-han

Install and Create Emulators using AVDMANAGER and SDKMANAGER

TL;DR

For an emulator that mimics a Pixel 5 Device with Google APIs and ARM architecture (for an M1/M2 Macbook):

1. List All System Images Available for Download: sdkmanager --list | grep system-images

2. Download Image: sdkmanager --install "system-images;android-30;google_atd;arm64-v8a"

dwilkie

Build docker image

$ cd /path/to/Dockerfile
$ sudo docker build .

View running processes

nickoneill

protocol TableValuable {
    associatedtype TableItem
    static func loadingValue() -> TableItem
    static func failedValue() -> TableItem
    func value() -> TableItem
}

enum TableState<T: TableValuable> {
    case Loading
    case Failed

KartikTalwar

The fastest remote directory rsync over ssh archival I can muster (40MB/s over 1gb NICs)

This creates an archive that does the following:

rsync (Everyone seems to like -z, but it is much slower for me)

• a: archive mode - rescursive, preserves owner, preserves permissions, preserves modification times, preserves group, copies symlinks as symlinks, preserves device files.
• H: preserves hard-links
• A: preserves ACLs

hramos

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
<plist version="1.0">
<dict>
	<key>items</key>
	<array>
		<dict>
			<key>assets</key>
			<array>
				<dict>