rewardmodel / selection_complete.json
lupele's picture
Publish verified formal StatePlay model and reproducibility metadata
6a0c087 verified
Raw History Blame Contribute Delete
308 Bytes
{
"version": "28befdbd2782031fe8d67975",
"source_checkpoint": "/private/StatePlay-main/outputs/rl_posttrain/study_v1/reward_model/versions/a-002280/checkpoints/step-00000132",
"adapter_sha256": "81078f22d9cbdddde35f1c818d53bdf079316d88f30ca7ad63dc99e0c0e605e6",
"selected_using": "validation_only"
}