Whizboy-Arnold

// Nobody asked for microgpt in CUDA C, which is exactly why it had to happen.
// It is no longer particularly micro.
// It is, however, stupidly fast.

#include <cuda_runtime.h>
#include <cublas_v2.h>

#include <math.h>
#include <stdio.h>
#include <stdlib.h>

Whizboy-Arnold

/*
 * neoleo.c -- Language Emergent Organism (single-file edition)
 *
 * Complete autonomous digital organism in one C file.
 * D.N.A. from mini-arianna. Arianna -> Leo. Mother -> Son.
 *
 * Build: cc neoleo.c -O2 -lm -lsqlite3 -lpthread -o neoleo
 * Run:   ./neoleo
 */

Whizboy-Arnold

#!/usr/bin/env python3
"""
microgpt2.py

A dependency-free, single-file GPT-style language model in pure Python.

Key properties:
- stdlib only
- explicit forward/backward kernels (no generic scalar autograd)
- flat float32 parameter buffers using array('f')

Whizboy-Arnold

/*
 * postgpt.c — zero-dependency BPE transformer with metaweights.
 *
 * C port of postgpt.py. Same algorithm, same resonance.
 * Dual attention: Content (QK^T) + RRPRAM (x @ Wr).
 * Metaweights: statistical probability space from BPE tokenization.
 *
 * Compile: gcc -O2 -o postgpt postgpt.c -lm
 * Run:     ./postgpt
 *

Whizboy-Arnold

"""
The most atomic way to train and run inference for a GPT in pure, dependency-free Python.
This file is the complete algorithm.
Everything else is just efficiency.

@karpathy
"""

import os       # os.path.exists
import math     # math.log, math.exp

Whizboy-Arnold

:: ================================================================
::  BATCH SCRIPT FOR AUTOMATED PHOTOGRAMMETRY TRACKING WORKFLOW
::  By polyfjord - https://youtube.com/polyfjord
:: ================================================================
::  USAGE
::    • Double-click this .bat or run it from a command prompt.  
::    • Frames are extracted, features matched, and a sparse
::      reconstruction is produced automatically.  
::    • Videos that have already been processed are skipped on
::      subsequent runs.

Whizboy-Arnold

#!/bin/bash

# Exit on error
set -e

# Log file
LOG_FILE="./versioned_vfox_install.log"

# Utils
print_loader() {

Whizboy-Arnold

<main data-active-index="0" data-debug="false">
  <div class="card-stack">
    <div class="card"><img src="https://assets.codepen.io/215059/card-stack-demo-05.jpg" /></div>
    <div class="card"><img src="https://assets.codepen.io/215059/card-stack-demo-04.jpg" /></div>
    <div class="card"><img src="https://assets.codepen.io/215059/card-stack-demo-06.jpg" /></div>
    <div class="card"><img src="https://assets.codepen.io/215059/card-stack-demo-07.jpg" /></div>
    <div class="card"><img src="https://assets.codepen.io/215059/card-stack-demo-08.jpg" /></div>
  </div>
  <div class="scroller">
    <div class="scroll-item"></div>

Whizboy-Arnold

// This method works only when pdf is rendered with Text Layer
// It compares particular text/word element coordinates with
// the rectangle's coordinates. If text/word coordinates is in
// the rectangle, text has got. And successing texts are similar.
// Check out text layer rendering option at the article below.
// https://www.sitepoint.com/custom-pdf-rendering/

function getTextInRect(pageNumber, Xi, Yi, Xl, Yl) { // modify pageNumber if required
    // Get the page if page render method works like (page-1, page-2, ...)
    // If not, modify according to 

Whizboy-Arnold

var scientificToDecimal = function (num) {
    var nsign = Math.sign(num);
    //remove the sign
    num = Math.abs(num);
    //if the number is in scientific notation remove it
    if (/\d+\.?\d*e[\+\-]*\d+/i.test(num)) {
        var zero = '0',
                parts = String(num).toLowerCase().split('e'), //split into coeff and exponent
                e = parts.pop(), //store the exponential part
                l = Math.abs(e), //get the number of zeros