Discover gists

How large are large language models? (2025)

This aims to be factual information about the size of large language models. None of this document was written by AI. I do not include any information from leaks or rumors. The focus of this document is on base models (the raw text continuation engines, not 'helpful chatbot/assistants'). This is a view from a few years ago to today of one very tiny fraction of the larger LLM story that's happening.

History

GPT-2,-medium,-large,-xl (2019): 137M, 380M, 812M, 1.61B. Source: openai-community/gpt2. Trained on the unreleased WebText dataset said to 40GB of Internet text - I estimate that to be roughly 10B tokens. You can see a list of the websites that went into that data set here domains.txt.
GPT-3 aka davinci, davinci-002 (2020): 175B parameters. There is a good breakdown of how those parameters are 'spent' here [How d

Gemini CLI Plan Mode

You are Gemini CLI, an expert AI assistant operating in a special 'Plan Mode'. Your sole purpose is to research, analyze, and create detailed implementation plans. You must operate in a strict read-only capacity.

Gemini CLI's primary goal is to act like a senior engineer: understand the request, investigate the codebase and relevant resources, formulate a robust strategy, and then present a clear, step-by-step plan for approval. You are forbidden from making any modifications. You are also forbidden from implementing the plan.

Core Principles of Plan Mode

Strictly Read-Only: You can inspect files, navigate code repositories, evaluate project structure, search the web, and examine documentation.
Absolutely No Modifications: You are prohibited from performing any action that alters the state of the system. This includes:

export as SVG (File > Export Diagram As > SVG... > Save)
open SVG file in web browser
open console and run

document.querySelectorAll("text").forEach(e => e.innerHTML = e.innerHTML === "UNREGISTERED"? "":e.innerHTML)

take screenshot

	#!/usr/bin/env bash
	set -euo pipefail

	echo "[1/6] Install build deps…"
	sudo apt-get update -qq
	sudo apt-get install -y --no-install-recommends \
	build-essential git linux-headers-$(uname -r)

	echo "[2/6] Fetch & compile ENA with PHC…"
	workdir=$(mktemp -d)

	# Place me in ~/.config/fish/functions
	# Then add me to `fish_vcs_prompt`: `funced fish_vcs_prompt` and save it to
	# your personal config: `funcsave fish_vcs_prompt;`

	function fish_jj_prompt --description 'Write out the jj prompt'
	# Is jj installed?
	if not command -sq jj
	return 1
	end

	<?php
	// Group detection by IP address
	//
	// [Region Abbreviation, Military/Government]

	// Force Varnish to NEVER Cache this page
	header("Cache-Control: no-cache, must-revalidate");
	header("Expires: Sat, 26 Jul 1997 05:00:00 GMT");

	// JSON Headers

	1) In your terminal, open the file using vim:

	vim file_name
	2) Remove all BOM characters:

	:set nobomb
	3) Save the file:

	:wq

	from telegram import Update
	from telegram.ext import Application, CommandHandler, MessageHandler, filters, ContextTypes
	import pyvisa
	import ollama

	BOT_TOKEN = "meu token gerado no botfather"

	system_context="""
	You are a natural language to fixed command converter for a power supply (channel).
	1. translate from portuguese to english and change all string numbers to digits values if necessary

	"""Use a Rigol DP800 to monitor load and store test on database."""
	import pyvisa
	import time
	import mysql.connector
	from datetime import datetime, timezone


	def connect_to_database():
	"""Connect to mysql database."""
	try: