Downloads · 30 days
0
sokoban_adaptive_mixed_forward_rl_step30
novastar111/sokoban_adaptive_mixed_forward_rl_step30
sokoban_adaptive_mixed_forward_rl_step30 is a reinforcement learning model from novastar111. Use it for the reinforcement learning task on the model card, and read the license before you ship it in a product. The card lists the license as cc-by-nc-4.0.
BAGEL-7B-MoT Sokoban checkpoint after 30 IMP-agent RL steps.
Access
Public
Updated Aug 18, 2026
Repo size
29.2 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors29.2 GB · 100%
At a glance
- Task
- Reinforcement Learning
- License
- cc-by-nc-4.0
- Access
- Public
- Created
- Aug 18, 2026
- Updated
- Aug 18, 2026
- SHA
- e491c675
Model card
From the Hugging Face model README
sokoban_adaptive_mixed_forward_rl_step30
BAGEL-7B-MoT Sokoban checkpoint after 30 IMP-agent RL steps.
- arm: mixed-forward
- initialization:
novastar111/sokoban_adaptive_mixed_forward_sft3k - RL data: 2,000 mixed 3-box boards (certified deadlock + trivial)
- objective: environment success with a
-0.1malformed-output penalty - weights: converted BF16 EMA safetensors; optimizer/training state is not included
The mixed-forward and mixed-branch labels refer to the step-3000 SFT initializations used by the
RL experiment. Evaluation uses Sokoban-v8 q95/perseg, full self-rollout, move-only actions, and
stop-required success. Load with the public BAGEL-7B-MoT base/config and point the eval runner's
--bagel-path at this repository.
- Task
- Reinforcement Learning
- License
- cc-by-nc-4.0
- Created
- Aug 18, 2026
- Updated
- Aug 18, 2026