CronusVLA
Collection
Paper, Data and Checkpoints for ''CronusVLA: Towards Efficient and Robust Manipulation
via Multi-Frame Vision-Language-Action Modeling''
β’
12 items
β’
Updated
β’
2
Weights
checkpoints/step-055000-epoch-04-loss=0.0286.pt: Complete model checkpoint of CronusVLA-7B for direct evaluation. This checkpoint has a high average performance across both WidowX (VM) and Google Robot (VM and VA) settting of SimplerEnv.If you want to evaluate or further finetune with this checkpoint, please refer to CronusVLA for more details.
Evaluation Results
final_result_of_SimplerEnv.log: Results on SimplerEnv.If you find this model useful, please cite our work:
@article{li2025cronusvla,
title={CronusVLA: Transferring Latent Motion Across Time for Multi-Frame Prediction in Manipulation},
author={Li, Hao and Yang, Shuai and Chen, Yilun and Tian, Yang and Yang, Xiaoda and Chen, Xinyi and Wang, Hanqing and Wang, Tai and Zhao, Feng and Lin, Dahua and others},
journal={arXiv preprint arXiv:2506.19816},
year={2025}
}
Base model
CogACT/CogACT-Base