Hacker Newsnew | past | comments | ask | show | jobs | submit | gpjt's submissionslogin
1.
Extending Raschka's GPT-2: an MoE trained from scratch on an RTX 3090 (gilesthomas.com)
1 point by gpjt 6 days ago | past | discuss
2.
Putting my Jax-trained models on the Hugging Face Hub (gilesthomas.com)
2 points by gpjt 13 days ago | past | discuss
3.
Why do OpenAI's GPT-2 weights beat mine? Part four: digging into dropout (gilesthomas.com)
1 point by gpjt 20 days ago | past
4.
First patient to undergo live AI-assisted brain surgery has tumour removed (bbc.com)
6 points by gpjt 21 days ago | past
5.
Adding diagrams to my static site generator with D2 (gilesthomas.com)
3 points by gpjt 22 days ago | past
6.
Use the built-in GELU, don't roll your own (gilesthomas.com)
1 point by gpjt 28 days ago | past
7.
A Quick(ish) Chinchilla Check (gilesthomas.com)
1 point by gpjt 40 days ago | past
8.
I use AI on this blog (gilesthomas.com)
1 point by gpjt 47 days ago | past | 1 comment
9.
Why do OpenAI's GPT-2 weights beat mine? Part three: testing overtraining (gilesthomas.com)
2 points by gpjt 48 days ago | past
10.
Why do OpenAI's GPT-2 weights beat mine? Part two: the bugfix (gilesthomas.com)
8 points by gpjt 48 days ago | past
11.
Why do OpenAI's GPT-2 weights beat mine? (gilesthomas.com)
4 points by gpjt 50 days ago | past
12.
Benchmarking Qwen 3.6 35B MoE (3B active) on an RTX 3090 (gilesthomas.com)
16 points by gpjt 54 days ago | past
13.
Building intuition about LLM parameter counts (gilesthomas.com)
2 points by gpjt 68 days ago | past
14.
Poppy the training box, part 1: the beginnings (gilesthomas.com)
3 points by gpjt 70 days ago | past
15.
From bigrams to GPT-2, one component at a time (in Jax) (gilesthomas.com)
1 point by gpjt 70 days ago | past
16.
Building a Jax training loop for an LLM training run (gilesthomas.com)
2 points by gpjt 78 days ago | past
17.
Thoughts on Role Confusion (gilesthomas.com)
3 points by gpjt 84 days ago | past
18.
Flax debugging: making a hash of things (gilesthomas.com)
2 points by gpjt 3 months ago | past
19.
10Gb/s Ethernet: switching to a Broadcom SFP+ module (gilesthomas.com)
195 points by gpjt 3 months ago | past | 170 comments
20.
Jax: Commitment Issues (gilesthomas.com)
4 points by gpjt 3 months ago | past
21.
Jax Back Ends and Devices (gilesthomas.com)
2 points by gpjt 3 months ago | past
22.
Using Safetensors with Flax (gilesthomas.com)
2 points by gpjt 3 months ago | past
23.
First Looking into Jax (gilesthomas.com)
3 points by gpjt 3 months ago | past
24.
10Gb/s Ethernet: using mini-heatsinks with a 10GBASE-T SFP+ module (gilesthomas.com)
3 points by gpjt 4 months ago | past
25.
10Gb/s Ethernet: what I did to get it working in my home (gilesthomas.com)
232 points by gpjt 4 months ago | past | 177 comments
26.
10Gb Ethernet: what I had to (re)learn (gilesthomas.com)
1 point by gpjt 4 months ago | past | 1 comment
27.
LLM from scratch, part 33 – what I learned from the appendices (gilesthomas.com)
5 points by gpjt 4 months ago | past
28.
LLM from scratch (32l) – Interventions: updated instruction fine-tuning results (gilesthomas.com)
1 point by gpjt 4 months ago | past
29.
How an LLM becomes more coherent as we train it (gilesthomas.com)
3 points by gpjt 5 months ago | past
30.
LLM from scratch, part 32k – Interventions: gradient accumulation (gilesthomas.com)
2 points by gpjt 5 months ago | past

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: