Implement a ChatGPT-like LLM in PyTorch from scratch, step by step
COMMITS
September 3, 2025
S
Update requirements for Intel macOS (#805)
Sebastian Raschka committed
September 2, 2025
H
added brief explanations about 2 different ways of RoPE implementations (#802)
Hayato Hongo committed
R
remove local config files
rasbt committed
S
Interactive qwen3 chat interface (#801)
Sebastian Raschka committed
August 31, 2025
S
Improve RoPE (#799)
Sebastian Raschka committed
S
Add KVCache variant of Qwen3 notebook (#800)
Sebastian Raschka committed
August 30, 2025
S
Update pixi powershell command section (#798)
Sebastian Raschka committed
August 28, 2025
S
reasoning-from-scratch (#793)
Sebastian Raschka committed
August 22, 2025
J
Improve MHA einsum (#781)
Jestine Paul committed
C
August 20, 2025
S
Minor cosmetic fixes in Gemma 3 nbs (#780)
Sebastian Raschka committed
August 19, 2025
S
Add Gemma3 KV cache variant (#776)
Sebastian Raschka committed
S
Improve MHA einsum (#775)
Sebastian Raschka committed
August 18, 2025
S
add HF equivalency tests for standalone nbs (#774)
Sebastian Raschka committed
August 17, 2025
S
Gemma 3 270M From Scratch (#771)
Sebastian Raschka committed
August 15, 2025
S
Fix qk_norm comment (#769)
Sebastian Raschka committed
August 14, 2025
S
Qwen3 and Llama3 equivalency teests with HF transformers (#768)
Sebastian Raschka committed
August 2, 2025
S
MoE Nb readability improvements (#761)
Sebastian Raschka committed
S
Qwen3 Coder Flash & MoE from Scratch (#760)
Sebastian Raschka committed
July 28, 2025
C
[Minor] Qwen3 typo & optim (#758)
casinca committed
July 23, 2025
S
Interleaved Q and K for RoPE in Llama 2 (#750)
Sebastian Raschka committed
July 22, 2025
S
Minor typo: pply -> Apply (#749)
Sebastian Raschka committed
S
Update Python dependency in pyproject.toml (#748)
Sebastian Raschka committed
July 16, 2025
S
get rid of redundant memory profiler import (#744)
Sebastian Raschka committed
July 13, 2025
S
Add link to official video course (#741)
Sebastian Raschka committed
July 10, 2025
M
Fix issue: 731 by resolving semantic error (#738)
Matthew Hernandez committed
S
Batched KV Cache Inference for Qwen3 (#735)
Sebastian Raschka committed
July 9, 2025
S
Qwen3 tokenizer sanity checks (#730)
Sebastian Raschka committed