metadata
library_name: lerobot
tags:
- molmoact2
- robotics
- lerobot
- vla
- quantile-normalization
base_model: allenai/MolmoAct2
molmoact2_block_stack_base_quantile_12k
Fine-tuned MolmoAct2 (action-expert-only) on block_stack with QUANTILES normalization (q01/q99). Intermediate checkpoint at step 12000 while 30k training continues on Isambard.
| Policy | MolmoAct2 (policy.type=molmoact2) |
| Init checkpoint | allenai/MolmoAct2 |
| Dataset | villekuosmanen/armnetbench_block_stack |
| Task | block_stack |
| Local run | molmoact2_block_stack_base_quantile |
| Checkpoint step | 012000 (12k / 30k target) |
| Normalization | QUANTILES (action + state + gripper), IDENTITY (visual) |
| Training | Isambard GH200, batch 64 (16/GPU x 4 DDP), bf16, no gradient checkpointing |
Checkpoints
Step 012000 lives at the repository root for direct loading.
Usage
from lerobot.policies.molmoact2.modeling_molmoact2 import MolmoAct2Policy
policy = MolmoAct2Policy.from_pretrained("pravsels/molmoact2_block_stack_base_quantile_12k")