Generate my own data based on calvin env

[zhanxch]

Hi! I am really like this project and build some vla tested on it. It is easy to get the results right now, but it is still hard to get more analysis about the bad cases.
Therefore, If there is any instruction for generating data what I need, that would be very helpful!
.

[XiangH9]

Hi, sorry to bother you.

I wanted to ask whether the validation set inside the task_ABC_D folder of your CALVIN dataset has corresponding language annotation files — is there a /lang_annotations folder?

This is very important for my work. Looking forward to your reply.

[zhanxch]

Hi, sorry to bother you.

I wanted to ask whether the validation set inside the task_ABC_D folder of your CALVIN dataset has corresponding language annotation files — is there a /lang_annotations folder?

This is very important for my work. Looking forward to your reply.

Do u mean calvin/dataset/task_ABC_D/validation/lang_annotations ? you can just find it in the dataset you download follow such path. And you can find auto_lang_ann.npy and embeddings.npy in the folder