Implement a ChatGPT-like LLM in PyTorch from scratch, step by step
COMMITS
/ appendix-D/01_main-chapter-code/previous_chapters.py June 19, 2024
D
fixed num_workers (#229)
Daniel Kleine committed
June 8, 2024
S
Remove leftover instances of self.tokenizer (#201)
Sebastian Raschka committed
June 3, 2024
R
restore file
rasbt committed
R
add number of workers to data loader
rasbt committed
April 28, 2024
S
Rename drop_resid to drop_shortcut (#136)
Sebastian Raschka committed
April 20, 2024
S
Return nan if val loader is empty (#124)
Sebastian Raschka committed
April 13, 2024
S
Make datesets and loaders compatible with multiprocessing (#118)
Sebastian Raschka committed
April 11, 2024
J
Use instance tokenizer (#116)
James Holcombe committed
April 4, 2024
S
Rename variable to context_length to make it easier on readers (#106)
Sebastian Raschka committed
S
Remove reundant dropout in MLP module (#105)
Sebastian Raschka committed
April 1, 2024
R
figure scaling
rasbt committed
R
make figures for appendix d
rasbt committed
R
make figures for appendix d
rasbt committed
March 27, 2024
R
make batch loss calculatution more efficient
rasbt committed
R
simplify calc_loss_loader
rasbt committed
March 25, 2024
R
simplify .view code
rasbt committed
March 19, 2024
S
Ch05 supplementary code (#81)
Sebastian Raschka committed
March 18, 2024
S
Update pep8 (#78)
Sebastian Raschka committed
March 11, 2024
R
Add appendix D
rasbt committed