Add DPO data generation & training; fix checkpoint labels; switch to WSD scheduler; update optimizer ablations in abstract/conclusion

Browse Source

...

This commit is contained in:

ViperEkura

2026-07-19 15:25:42 +08:00

parent 6c2e04a86f

commit e149997200

2 changed files with 88 additions and 39 deletions

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Download Patch File Download Diff File Expand all files Collapse all files

data/weight_dist_by_component.png

BIN

Binary file not shown.

Side by Side Before Width:  |  Height:  |  Size: 100 KiB