Implement a ChatGPT-like LLM in PyTorch from scratch, step by step
COMMITS
July 9, 2025
S
Add more sophisticated Qwen3 tokenizer (#729)
Sebastian Raschka committed
July 8, 2025
S
Simplify KV cache usage (#728)
Sebastian Raschka committed
S
Update Qwen3 tokenizer test (#727)
Sebastian Raschka committed
M
Fix issue 724: unused args (#726)
Matthew Hernandez committed
July 6, 2025
S
Update README.md
Sebastian Raschka committed
S
Add prerequisite section (#723)
Sebastian Raschka committed
July 3, 2025
S
Fix `pip uv` typo in installation instructions (#722)
Sebastian Raschka committed
June 30, 2025
S
Handle other Qwen3 tokenizer settings (#716)
Sebastian Raschka committed
June 28, 2025
S
Fix d_out code comment in bonus materials (#715)
Sebastian Raschka committed
S
Support different Qwen3 sizes in pkg (#714)
Sebastian Raschka committed
S
Use test mode arg in ch07 (#713)
Sebastian Raschka committed
June 25, 2025
S
Remove unused params for hparam script (#710)
Sebastian Raschka committed
S
Add Qwen3 1.7, 4B, 8B, and 32B support to from-scratch nb (#709)
Sebastian Raschka committed
June 24, 2025
S
Link the other KV cache sections (#708)
Sebastian Raschka committed
S
Add link to free exercise PDF (#706)
Sebastian Raschka committed
June 23, 2025
S
Update Llama 3 table for consistency with Qwen3
Sebastian Raschka committed
S
Improve KV cache code for torch.compile (#705)
Sebastian Raschka committed
M
Fix bug in masking when kv cache is used. (#697)
Martin Ma committed
S
CPU compile performance for Qwen3 models (#704)
Sebastian Raschka committed
S
Ch06 classifier function asserts (#703)
Sebastian Raschka committed
S
Update README.md (#702)
Shamik committed
M
Fix issue #684: Minor docstring edit (#699)
Matthew Hernandez committed
June 22, 2025
S
Fix code comment: embed_dim -> d_out (#698)
Sebastian Raschka committed
S
Use more recent sentencepiece tokenizer API (#696)
Sebastian Raschka committed
S
Fix some wording issues in the notes (#695)
Sebastian Raschka committed
S
Fix: Typo in `appendix_d.py` comments. (#682)
Sajjad Baloch committed
C
fix issue #664 - inverted token and pos emb layers (#665)
casinca committed
June 21, 2025
S
Qwen3 KV cache (#688)
Sebastian Raschka committed
D
added pkg fixes (#676)
Daniel Kleine committed
S
Add GPT-2 KV cache to pkg (#687)
Sebastian Raschka committed