gio 280649f481 Add --save-keep-latest=N for checkpoint rotation ... Pre-save deletion of oldest model_*.pt + optim_*.pt + meta_*.json so disk peak stays bounded. Default -1 = keep all (no behavior change). Used by the d26 1-week dense run to bound live checkpoints across many save-every cycles. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>		2026-05-10 13:19:26 -05:00
..
base_eval.py	delete autocast, an unnecessary thorn in my side, manage dtypes directly	2026-03-04 23:55:30 +00:00
base_train.py	Add --save-keep-latest=N for checkpoint rotation	2026-05-10 13:19:26 -05:00
chat_cli.py	delete autocast, an unnecessary thorn in my side, manage dtypes directly	2026-03-04 23:55:30 +00:00
chat_eval.py	delete autocast, an unnecessary thorn in my side, manage dtypes directly	2026-03-04 23:55:30 +00:00
chat_rl.py	delete autocast, an unnecessary thorn in my side, manage dtypes directly	2026-03-04 23:55:30 +00:00
chat_sft.py	Merge pull request #634 from 2bitbit/fix-docs-and-comments	2026-03-25 14:31:49 -07:00
chat_web.py	delete autocast, an unnecessary thorn in my side, manage dtypes directly	2026-03-04 23:55:30 +00:00
tok_eval.py	initial commit	2025-10-13 06:49:24 -07:00
tok_train.py	fix: correct minor typos in help text, README, and comments	2026-03-12 17:03:26 +08:00