Hi,
I have a question about the model related to SimplerEnv.
It seems that the model qwbu/univla-7b-224-sft-simpler-bridge was fine-tuned on the Bridge_OXE dataset.
Will you release the corresponding finetune code? Is it based on vla-scripts/finetune_realworld.py, or another script?
Any guidance on reproducing the post-training process would be appreciated.
Thanks in advance.