Writing an LLM from scratch, part 34b -- from bigrams to GPT-2, one component at a time (in JAX) :: Giles' blog
el.dataset.currentDropdown = '')<br>}">
Giles' blog
Writing the post that I wished I'd found when I started learning whatever it was...
About
Contact
Archives
Categories
Blogroll
July 2026 (1)
June 2026 (7)
May 2026 (2)
April 2026 (11)
March 2026 (3)
February 2026 (4)
January 2026 (4)
December 2025 (1)
November 2025 (3)
October 2025 (9)
September 2025 (3)
August 2025 (5)
July 2025 (1)
June 2025 (2)
May 2025 (3)
April 2025 (2)
March 2025 (7)
February 2025 (10)
January 2025 (6)
December 2024 (7)
September 2024 (1)
August 2024 (2)
July 2024 (2)
May 2024 (2)
April 2024 (2)
February 2024 (2)
April 2023 (1)
March 2023 (2)
September 2022 (1)
February 2022 (1)
November 2021 (1)
March 2021 (1)
February 2021 (2)
August 2019 (1)
November 2018 (1)
May 2017 (1)
December 2016 (1)
April 2016 (1)
August 2015 (1)
December 2014 (1)
August 2014 (1)
March 2014 (1)
December 2013 (1)
October 2013 (3)
September 2013 (4)
August 2013 (2)
July 2013 (1)
June 2013 (1)
February 2013 (1)
October 2012 (1)
June 2012 (1)
May 2012 (1)
April 2012 (1)
February 2012 (1)
October 2011 (1)
June 2011 (1)
May 2011 (1)
April 2011 (1)
March 2011 (1)
February 2011 (1)
January 2011 (1)
December 2010 (3)
November 2010 (1)
October 2010 (1)
September 2010 (1)
August 2010 (1)
July 2010 (1)
May 2010 (3)
April 2010 (1)
March 2010 (2)
February 2010 (3)
January 2010 (4)
December 2009 (2)
November 2009 (5)
October 2009 (2)
September 2009 (2)
August 2009 (3)
July 2009 (1)
May 2009 (1)
April 2009 (1)
March 2009 (5)
February 2009 (5)
January 2009 (5)
December 2008 (3)
November 2008 (7)
October 2008 (4)
September 2008 (2)
August 2008 (1)
July 2008 (1)
June 2008 (1)
May 2008 (1)
April 2008 (1)
January 2008 (4)
December 2007 (3)
March 2007 (3)
February 2007 (1)
January 2007 (2)
December 2006 (4)
November 2006 (18)
AI (88)
TIL deep dives (77)
Python (72)
LLM from scratch (48)
Resolver One (34)
PyTorch (21)
TIL (21)
Blogkeeping (18)
PythonAnywhere (17)
Linux (16)
Startups (15)
Hugging Face (13)
NSLU2 offsite backup project (13)
Gadgets (12)
Funny (11)
Musings (11)
Finance (10)
Fine-tuning LLMs (10)
C (9)
Personal (8)
Robotics (8)
Website design (8)
JAX (7)
3D (5)
Quick links (5)
Rants (5)
Cryptography (4)
JavaScript (4)
Music (4)
Oddities (4)
Talks (4)
Dirigible (3)
Eee (3)
Memes (3)
Politics (3)
Django (2)
GPU Computing (2)
LaTeX (2)
MathML (2)
OLPC XO (2)
Retro Language Models (2)
Space (2)
VoIP (2)
Copyright (1)
Golang (1)
Microprojects (1)
Raspberry Pi (1)
Software development tools (1)
Agile Abstractions
Astral Codex Ten
:: (Bloggable a) => a -> IO ()
David Friedman's Substack
Econ & Energy
Entrepreneurial Geekiness
For some value of "Magic"
Hackaday
kaleidic.ai newsletter
Knowing.NET
Language Log
Millennium Hand
ntoll.org
Obey the Testing Goat!
PK
PythonAnywhere News
Simon Willison's Weblog
Societive
Software Deviser
Some opinions, held with varying degrees of certainty
tartley.com
Writing an LLM from scratch, part 34b -- from bigrams to GPT-2, one component at a time (in JAX)
Posted on 8 July 2026
in
AI,
LLM from scratch,
TIL deep dives,
JAX
This post is the capstone of<br>the most long-running series on my blog.<br>In December 2024 (!), I started<br>reading Sebastian Raschka's book<br>"Build a Large Language Model (from Scratch)",<br>and worked through it carefully. Being who I am, despite trying to apply a strict "no side quests"<br>policy, I found myself zooming off and digging into all kinds of things.
It's time to wrap it up. I had decided that the endpoint would be to build and train an LLM from scratch just using my notes --<br>no reference to the book, no reference to the model code I'd written when following the book.<br>After an X/Twitter poll, I decided to<br>use JAX for that, just to make sure that I really was building it from scratch and<br>not regurgitating bits of PyTorch code like a bad coding LLM spitting out half-digested<br>lumps of Stack Overflow.
In my last post,<br>I showed how I built a JAX training script that mirrored what I had built for the original<br>PyTorch version of the model. To test it as I went along, I used it to train a really dumb "LLM", which instead<br>of trying to predict the next token for every token in an input sequence, instead<br>predicted the input -- that is, if you fed it
The fat cat sat on the mat
It would return the same thing. I called that an A-to-A model.
In this post, I'll show you how I turned it into a GPT-2 model, and then trained<br>it from scratch on my RTX 3090 (using the parameter counts for the original paper's "small" size). What turned out<br>really well with this is that I found a route that meant that almost every component I added<br>made the model better! That's not guaranteed -- sometimes different aspects of an AI model<br>depend on each other, so...