Claude/great allen 51nxz7 - #194

Open

tinchen777 wants to merge 17 commits into

locuslab:mainfrom

tinchen777:claude/great-allen-51nxz7

Commits on Jun 14, 2026

new: ignore whl
tinchen777
committed
1bc277f

Commits on Jun 15, 2026

Add usage demos and LoRA finetuning support
claude
committed
ac81280
Merge pull request #1 from tinchen777/claude/great-allen-51nxz7
tinchen777
authored
4c51d99
new: add ignore
tinchen777
committed
6378210
new: change LLama-3.2-1B model
tinchen777
committed
fca38eb
Add unlearning demos and a custom unlearning trainer example
claude
committed
94f9b2b
Fix bf16 numpy crash in eval and respect HF_HOME hub layout
claude
committed
75b4113
docs: add runtime/troubleshooting section to demos README
claude
committed
0ab6b8c
docs: correct setup_data flag to --eval_logs in demos README
claude
committed
54b6940
demos: pin a single GPU to avoid DataParallel OOM on shared clusters
claude
committed
c383c61

Commits on Jun 16, 2026

fix: apply_chat_template returns BatchEncoding on new transformers
claude
committed
cb3f9c3
compat: adapt to transformers 5.x deprecations
claude
committed
7674163
logging: add rich job_logging config and make it the default
claude
committed
8bad8a3
logging: show logger name in rich format as [%(name)s]
claude
committed
f3253d5
logging: give logger name its own color via custom Rich formatter
claude
committed
ec6991b
logging: add rich config for hydra_logging group too
claude
committed
ca1d8fd

Commits on Jun 21, 2026

docs: add Chinese Hydra guide with priority/compose mermaid diagram
claude
committed
21bad31