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.

rymiwe

Claude Code Config Backup

Automatically backup your Claude Code configuration to a private GitHub repo using chezmoi. Your settings, agents, rules, and project configs are version-controlled and synced across machines.

What you get:

• Auto-backup of memory files, rules, settings, agents, and skills
• Auto-push to your dotfiles repo via chezmoi's native autoPush
• Auto-pull on session start to keep multi-machine setups in sync
• Restore your full setup on any machine with two commands

stenuto

#!/bin/bash

# Function to display usage information
usage() {
    echo "Usage: $0 /path/to/input.mp4 [ /path/to/output_directory ]"
    exit 1
}

# Check if at least one argument (input file) is provided
if [ $# -lt 1 ]; then

Artefact2

Which GGUF is right for me? (Opinionated)

Good question! I am collecting human data on how quantization affects outputs. See here for more information: ggml-org/llama.cpp#5962

In the meantime, use the largest that fully fits in your GPU. If you can comfortably fit Q4_K_S, try using a model with more parameters.

llama.cpp feature matrix

See the wiki upstream: https://github.com/ggerganov/llama.cpp/wiki/Feature-matrix

darinpope

Pipeline version 1

pipeline {
  agent { label "linux" }
  stages {
    stage('Hello') {
      steps {
        sh '''
          aws --version
        '''

ashishrawat2911

getStringValuesSF() async {
  SharedPreferences prefs = await SharedPreferences.getInstance();
  //Return String
  String stringValue = prefs.getString('stringValue');
  return stringValue;
}
getBoolValuesSF() async {
  SharedPreferences prefs = await SharedPreferences.getInstance();
  //Return bool
  bool boolValue = prefs.getBool('boolValue');

yunyu

import "package:flutter/widgets.dart";
import "dart:math";

class SnappingListView extends StatefulWidget {
  final Axis scrollDirection;
  final ScrollController controller;

  final IndexedWidgetBuilder itemBuilder;
  final List<Widget> children;
  final int itemCount;

MithraTalluri

wkliwk

# BetterTouchTool reset trial time 
# ** All preference will reset

echo "remove ~/Library/Preferences/com.hegenberg.BetterTouchTool.plist"
rm -rf ~/Library/Preferences/com.hegenberg.BetterTouchTool.plist
echo "Done"

echo "remove ~/Library/Application\ Support/BetterTouchTool/"
rm -rf ~/Library/Application\ Support/BetterTouchTool/
echo "Done"

boeledi

import 'dart:async';

import 'package:flutter/material.dart';
import 'package:shared_preferences/shared_preferences.dart';

List<String> _languages = <String>['de','en','es','fr','it','nl','pt'];

class TestPage extends StatefulWidget {
  @override
  _TestPageState createState() => _TestPageState();