Release plan for model weights, training data, and training pipeline

[FanchengDing00]

Hi Qwen-VLA team,

Thanks for your great work.

I would like to ask whether there is a release plan for the Qwen-VLA model weights, training data, and training pipeline.

I am particularly interested in the VLN / navigation part. Could you please share whether the VLN training recipe and dataset configuration will be released, including the datasets used for pretraining and post-training?

These details would be very helpful for reproducing the VLN results and comparing with existing VLN-CE methods.

Thank you.