Vikram Dutt vdt

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

Here is my personal opinion about the questions I posed in this tweet:

I think that fine-tuning is still very valuable in many situations. I’ve done some more digging and I find that people who say that fine-tuning isn't useful are indeed often working on products where fine-tuning isn't likely to be useful:

They are making developer tools - foundation models have been trained extensively on coding tasks.
They are building foundation models and testing for the most general cases. But the foundation models themselves are also being trained for the most general cases.
They are building a personal assistant that isn’t scoped to any type of domain or use case, and is essentially similar to the same folks building foundation models.

Mock Exam

1

Q1. The metric node_cpu_temp_celcius reports the current temperature of a nodes CPU in celsius. What query will return the average temperature across all CPUs on a per node basis? The query should return {instance=“node1”} 23.5 //average temp across all CPUs on node1 {instance=“node2”} 33.5 //average temp across all CPUs on node2.

node_cpu_temp_celsius{instance="node1", cpu="0"} 28
node_cpu_temp_celsius{instance="node1", cpu="1"} 19
node_cpu_temp_celsius{instance="node2", cpu="0"} 36
node_cpu_temp_celsius{instance="node2", cpu="1"} 31

During the past days, this great article by Sam Pruden has been making the rounds around the gamedev community. While the article provides an in-depth analysis, its a bit easy to miss the point and exert the wrong conclusions from it. As such, and in many cases, users unfamiliar with Godot internals have used it points such as following:

Godot C# support is inefficient
Godot API and binding system is designed around GDScript
Godot is not production ready

In this brief article, I will shed a bit more light about how the Godot binding system works and some detail on the Godot

This worked on 14/May/23. The instructions will probably require updating in the future.

llama is a text prediction model similar to GPT-2, and the version of GPT-3 that has not been fine tuned yet. It is also possible to run fine tuned versions (like alpaca or vicuna with this. I think. Those versions are more focused on answering questions)

It is possible to run LLama 13B with a 6GB graphics card now! (e.g. a RTX 2060). Thanks to the amazing work involved in llama.cpp. The latest change is CUDA/cuBLAS which allows you pick an arbitrary number of the transformer layers to be run on the GPU. This is perfect for low VRAM.

Clone llama.cpp from git, I am on commit 08737ef720f0510c7ec2aa84d7f70c691073c35d.
- git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp

macOS Internals

Understand your Mac and iPhone more deeply by tracing the evolution of Mac OS X from prelease to Swift. John Siracusa delivers the details.

Starting Points

How to use this gist

You've got two main options:

The reverse turing test

I asked GPT-4 to come up with 10 questions to determine if the answerer was AI or human.

I provided my own answers for these questions and I also asked ChatGPT to answer them.

The result is that GPT-4 was able to correctly differentiate between AI and Human.

	type term =
	\| Lam of (term -> term)
	\| Pi of term * (term -> term)
	\| Appl of term * term
	\| Ann of term * term
	\| FreeVar of int
	\| Star
	\| Box

	let unfurl lvl f = f (FreeVar lvl)

	# coding=utf-8
	# Copyright 2023 The HuggingFace Inc. team. All rights reserved.
	#
	# Licensed under the Apache License, Version 2.0 (the "License");
	# you may not use this file except in compliance with the License.
	# You may obtain a copy of the License at
	#
	# http://www.apache.org/licenses/LICENSE-2.0
	#
	# Unless required by applicable law or agreed to in writing, software