Ping navono

Anti-hype LLM reading list

Goals: Add links that are reasonable and good explanations of how stuff works. No hype and no vendor content if possible. Practical first-hand accounts of models in prod eagerly sought.

Foundational Concepts

Pre-Transformer Models

Ubuntu 22.04 for Deep Learning

In the name of God

This gist contains steps to setup Ubuntu 22.04 for deep learning.

Install Ubuntu 22.04

Hardware specs

Chinese Brand "ITworks" , Model TW891, distributed in France and Belgium by Darty

CPU: Intel(R) Atom(TM) CPU Z3735F @ 1.33GHz
Video: Intel® HD Graphics for Intel Atom® Processor Z3700 Series
Screen: 1280x800
WiFi + BT: Realtek RTL8723BS_BT
Disks: mmcblk1: mmc1:0001 DF4032 29.1 GiB
RAM: 2GB DDR3 @ 1333 MHz

Note:

When this guide is more complete, the plan is to move it into Prepack documentation.
For now I put it out as a gist to gather initial feedback.

A Gentle Introduction to Prepack (Part 1)

If you're building JavaScript apps, you might already be familiar with some tools that compile JavaScript code to equivalent JavaScript code:

Babel lets you use newer JavaScript language features, and outputs equivalent code that targets older JavaScript engines.

Tech Lead Book

Introduction
What is a Tech Lead?
The Tech Lead Mindset
Becoming a Tech Lead
Working Cross-Functionally
The Art of Delegation
Technical Best Practices
Effective Decision Making

Falsehoods programmers believe about time

This is a compiled list of falsehoods programmers tend to believe about working with time.

Don't re-invent a date time library yourself. If you think you understand everything about time, you're probably doing it wrong.

Falsehoods

There are always 24 hours in a day.
February is always 28 days long.
Any 24-hour period will always begin and end in the same day (or week, or month).

	# train_grpo.py
	#
	# See https://github.com/willccbb/verifiers for ongoing developments
	#
	import re
	import torch
	from datasets import load_dataset, Dataset
	from transformers import AutoTokenizer, AutoModelForCausalLM
	from peft import LoraConfig
	from trl import GRPOConfig, GRPOTrainer

	#!/bin/bash

	### steps ####
	# verify the system has a cuda-capable gpu
	# download and install the nvidia cuda toolkit and cudnn
	# setup environmental variables
	# verify the installation
	###

	### to verify your gpu is cuda enable check

	# Setup Ubuntu
	sudo apt update --yes
	sudo apt upgrade --yes

	# Get Miniforge and make it the main Python interpreter
	wget https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-Linux-aarch64.sh -O ~/miniforge.sh
	bash ~/miniforge.sh -b -p ~/miniforge
	rm ~/miniforge.sh

	echo "PATH=$PATH:$HOME/miniforge/bin" >> .bashrc