File size: 930 Bytes
773aa58
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
---
license: cc-by-nc-4.0
tags: [bagel, vlm-gym, sokoban, reinforcement-learning, adaptive-thinking]
---

# sokoban_adaptive_mixed_branch_rl_step70

BAGEL-7B-MoT Sokoban checkpoint after **70 IMP-agent RL steps**.

- arm: **mixed-branch**
- initialization: [`novastar111/sokoban_adaptive_mixed_branch_sft3k`](https://huggingface.co/novastar111/sokoban_adaptive_mixed_branch_sft3k)
- RL data: 2,000 mixed 3-box boards (certified deadlock + trivial)
- objective: environment success with a `-0.1` malformed-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.