SPB Git
spboucher.ai

Repository Public Research preview

spbwpai0002_uqo

History

Latest commit B2: MPS cache release and padded shapes 3b399f3

NameLast commitUpdated
..
__init__.py WP-AI-0002: repository scaffold, licence governance, PMC S3 ingest, JATS parser, segmentation,...
data.py D19: mask same-passage negatives in InfoNCE
denoise.py Denoising pretraining stage (unlabelled originals), --init-from fine-tuning, no-repeat n-gram...
hf_finetune.py B2: MPS cache release and padded shapes
trainer.py Checkpoints always carry the tokenizer (partial runs stay loadable)

Clone wpai0002_uqo

Anonymous clone and fetch. Pushing requires the owner's personal access token.

Snapshot of main

ZIP tar.gz
Files main
Menu

Theme
↑↓navigate ↵open escclose /⌘Kanywhere