PTeterwak
/

action-reward-models

Model card Files Files and versions

action-reward-models

169 kB

Ctrl+K

Ctrl+K

4 contributors

History: 12 commits

PTeterwak's picture

Claude Fable 5

distillation README: record full-pipeline verification (A-D real runs)

4d6dfff about 1 month ago

actor_distillation
distillation README: record full-pipeline verification (A-D real runs) about 1 month ago
data_generation
Action reward models: data-gen (2 actors), training (selection + BT), inference demos about 1 month ago
inference
selection_infer: candidates must be objects with .molmo_action/.thought (caught by sanity test) about 1 month ago
training
Action reward models: data-gen (2 actors), training (selection + BT), inference demos about 1 month ago
.gitignore

43 Bytes
README: serving checklist from end-to-end inference verification about 1 month ago
README.md

9.02 kB
Merge branch 'main' of github.com:piotr-teterwak/action-reward-models about 1 month ago
requirements-datagen.txt

210 Bytes
requirements files pinned from working envs (inference/training/datagen) about 1 month ago
requirements-inference.txt

355 Bytes
requirements files pinned from working envs (inference/training/datagen) about 1 month ago
requirements-training.txt

637 Bytes
requirements files pinned from working envs (inference/training/datagen) about 1 month ago