Mike Huffman michaelhuffman

Taxonomy of Python Sandbox Capabilities

Preamble: This taxonomy outlines the tasks achievable within this sandboxed Python environment (ChatGPT Code Interpreter). Its primary strengths lie in leveraging a rich set of pre-installed libraries for offline processing of uploaded data, rapid prototyping, generating outputs (including enhanced UI displays for certain types), and executing code without requiring local setup. Use this guide to determine if the tool fits your task, considering its inherent limitations: sessions are ephemeral (work must be downloaded each time), no internet access is available, computational resources and file upload sizes are capped, and only the pre-installed libraries can be used. Current as of March 29, 2025.

I. Information Transformation and Synthesis (Core Purpose: Using code to process data, perform computations, generate insights, and create new artifacts, including potentially triggering enhanced UI displays for specific outputs.)

**Decision Support (Wh

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

Conferences

Hydraconf

Summer School

Summer school on practice and theory of Distributed Computing

Books

Eric Elliott's "Composing Software" Series

A collection of links to the excellent "Composing Software" series of medium stories by Eric Elliott.

Edit: I see that each post in the series now has index, previous and next links. However, they don't follow a linear flow through all the articles with some pointing back to previous posts effectively locking you in a loop.

NGINX Tuning For Best Performance

For this configuration you can use web server you like, i decided, because i work mostly with it to use nginx.

Generally, properly configured nginx can handle up to 400K to 500K requests per second (clustered), most what i saw is 50K to 80K (non-clustered) requests per second and 30% CPU load, course, this was 2 x Intel Xeon with HyperThreading enabled, but it can work without problem on slower machines.

You must understand that this config is used in testing environment and not in production so you will need to find a way to implement most of those features best possible for your servers.

Stable version NGINX (deb/rpm)

	import collections
	import math
	import os
	import cv2
	import numpy as np
	import time

	MAX_LINES = 4000
	N_PINS = 36*8
	MIN_LOOP = 20 # To avoid getting stuck in a loop

	# Code from best solution in page below:
	# https://help.zoho.com/portal/community/topic/zoho-mail-servers-reject-python-smtp-module-communications

	import smtplib
	from email.mime.text import MIMEText
	from email.header import Header
	from email.utils import formataddr

	# Define to/from
	sender = '[email protected]'

	package main

	import (
	"log"
	"sync"
	"time"
	)

	const tasksPerSecond = 1

	// https://hashaam.com/2017/06/11/strip-html-tags-in-swift/

	let htmlString = "LCD Soundsystem was the musical project of producer <a href='http://www.last.fm/music/James+Murphy' class='bbcode_artist'>James Murphy</a>, co-founder of <a href='http://www.last.fm/tag/dance-punk' class='bbcode_tag' rel='tag'>dance-punk</a> label <a href='http://www.last.fm/label/DFA' class='bbcode_label'>DFA</a> Records. Formed in 2001 in New York City, New York, United States, the music of LCD Soundsystem can also be described as a mix of <a href='http://www.last.fm/tag/alternative%20dance' class='bbcode_tag' rel='tag'>alternative dance</a> and <a href='http://www.last.fm/tag/post%20punk' class='bbcode_tag' rel='tag'>post punk</a>, along with elements of <a href='http://www.last.fm/tag/disco' class='bbcode_tag' rel='tag'>disco</a> and other styles. <br />"

	let htmlStringData = htmlString.data(using: String.Encoding.utf8)!

	let options: [String: Any] = [
	NSDocumentTypeDocumentAttribute: NSHTMLTextDocumentType,
	NSCharacte

	onDrop = (acceptedFiles, rejectedFiles) => {
	const file = acceptedFiles.find(f => f)
	const i = new Image()

	i.onload = () => {
	let reader = new FileReader()
	reader.readAsDataURL(file)
	reader.onload = () => {
	console.log({
	src: file.preview,