Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Dec 27, 2025

• Cover Python 3.12 (#933)

  rasbt

  authored
  d85ba93

  Copy full SHA for d85ba93

Commits on Dec 21, 2025

• upload saved nb

  rasbt

  committed
  1cea30f

  Copy full SHA for 1cea30f

• Update memory efficient loading nb

  rasbt

  committed
  2b9a67c

  Copy full SHA for 2b9a67c

Commits on Dec 20, 2025

• update submodule

  rasbt

  committed
  695ecb6

  Copy full SHA for 695ecb6

Commits on Dec 19, 2025

• Add some appendix E runtimes (#927)

  Show description for 1c9f49c
  rasbt

  authored
  1c9f49c

  Copy full SHA for 1c9f49c

• Gated DeltaNet updates (#926)

  rasbt

  authored
  57430d2

  Copy full SHA for 57430d2

Commits on Dec 16, 2025

• Sliding window KV Cache bug fix (#925)

  Show description for d7f178d
  talentJay-ux

  authored
  d7f178d

  Copy full SHA for d7f178d

Commits on Nov 25, 2025

• Remove persistent flag from cache buffers (#916)

  rasbt

  authored
  a11965f

  Copy full SHA for a11965f

Commits on Nov 23, 2025

• Add Olmo 3 README (#915)

  Show description for c195338
  rasbt

  authored
  c195338

  Copy full SHA for c195338

• Olmo 3 from scratch (#914)

  Show description for bc6f335
  rasbt

  authored
  bc6f335

  Copy full SHA for bc6f335

Commits on Nov 17, 2025

• RoPE decay plot (#910)

  Show description for 398b079
  rasbt

  authored
  398b079

  Copy full SHA for 398b079

• Update README wrt multi-query attention

  Show description for 28a8408
  rasbt

  authored
  28a8408

  Copy full SHA for 28a8408

Commits on Nov 13, 2025

• Write-up on how to get the most out of this book (#909)

  rasbt

  authored
  a409447

  Copy full SHA for a409447

Commits on Nov 9, 2025

• fix(GatedDeltaNet): Init param A from log of a uniform distrib (#906)

  casinca

  authored
  7d92267

  Copy full SHA for 7d92267

Commits on Nov 6, 2025

• Use consistent title case

  rasbt

  committed
  35354fa

  Copy full SHA for 35354fa

• Fix empty device issue (#904)

  rasbt

  authored
  58f45ae

  Copy full SHA for 58f45ae

• n_heads × d_head -> d_head × d_head in DeltaNet (#903)

  Show description for bcc73f7
  rasbt

  authored
  bcc73f7

  Copy full SHA for bcc73f7

Commits on Nov 3, 2025

• Image resizing

  rasbt

  authored
  488bef7

  Copy full SHA for 488bef7

• Gated DeltaNet write-up (#901)

  Show description for c6b8332
  rasbt

  authored
  c6b8332

  Copy full SHA for c6b8332

Commits on Nov 1, 2025

• Training on MPS in PyTorch 2.9 (#900)

  Show description for d6c3990
  rasbt

  authored
  d6c3990

  Copy full SHA for d6c3990

• Fix MHAEinsum weight dimension bug when d_in != d_out (#857) (#893)

  Show description for 27d52d6
  

  aviralgarg05

  and

  rasbt

  authored
  27d52d6

  Copy full SHA for 27d52d6

• simplify uv command (#898)

  rasbt

  authored
  b1db33b

  Copy full SHA for b1db33b

Commits on Oct 29, 2025

• Add bonus dependencies to pyproject (#897)

  Show description for 760f4c9
  rasbt

  authored
  760f4c9

  Copy full SHA for 760f4c9

Commits on Oct 22, 2025

• Fix ffn link (#892)

  Show description for 0adb5b8
  rasbt

  authored
  0adb5b8

  Copy full SHA for 0adb5b8

• Make quote style consistent (#891)

  rasbt

  authored
  7ca7c47

  Copy full SHA for 7ca7c47

Commits on Oct 21, 2025

• - docs(moe): correct arXiv link for DeepSeekMoE (#890)

  Show description for 9276edb
  casinca

  authored
  9276edb

  Copy full SHA for 9276edb

Commits on Oct 20, 2025

• Mixture-of-Experts intro (#888)

  rasbt

  authored
  218221a

  Copy full SHA for 218221a

Commits on Oct 17, 2025

• Make it easier to toggle between thinking and instruct variants (#887)

  rasbt

  authored
  27b6dfa

  Copy full SHA for 27b6dfa

Commits on Oct 14, 2025

• Update the compression rate comment in MLA (#883)

  Show description for 7fe4874
  rasbt

  authored
  7fe4874

  Copy full SHA for 7fe4874

Commits on Oct 13, 2025

• Use figure numbers in ch05-7 (#881)

  rasbt

  authored
  b969b3e

  Copy full SHA for b969b3e

• Add alternative attention structure (#880)

  rasbt

  authored
  bf039ff

  Copy full SHA for bf039ff

• sliding window attention (#879)

  rasbt

  authored
  6eb6adf

  Copy full SHA for 6eb6adf

• Add other appendices for completeness (#878)

  Show description for 21f0617
  rasbt

  authored
  21f0617

  Copy full SHA for 21f0617

Commits on Oct 12, 2025

• rm plot

  rasbt

  committed
  44eda53

  Copy full SHA for 44eda53

• Multi-Head Latent Attention (#876)

  Show description for 9b95866
  rasbt

  authored
  9b95866

  Copy full SHA for 9b95866

Pagination

Previous

Next