 Tobi Lütke
|
cbeeb1f89b
Add wall-clock checkpoints and full eval defaults
|
3 maanden geleden |
|
Tobi Lutke
|
1d7d167b29
finetune: strict Pydantic schema, one canonical data format
|
3 maanden geleden |
|
Tobi Lütke
|
bf1b8fc90a
lots of training stuff
|
3 maanden geleden |
|
Tobi Lutke
|
38073799c0
chore: clean up finetune folder and fix training workflow
|
3 maanden geleden |
|
Tobi Lutke
|
785620467a
refactor: reorder output format to put hyde line first
|
3 maanden geleden |
|
Tobi Lutke
|
32706a720f
Refactor finetune folder: train/rl scripts with YAML configs
|
4 maanden geleden |
