search

cshizhe / VLN-DUET

Official implementation of Think Global, Act Local: Dual-scale GraphTransformer for Vision-and-Language Navigation (CVPR'22 Oral).
114 stars 7 forks source link

Could you make the trained model weight public? #4

Closed heyzude closed 2 years ago

heyzude commented 2 years ago

Hi, Thank you for the amazing work you made :)

I wonder whether you can make the model weight public as also.

Thank you :)

cshizhe commented 2 years ago

We have already released the trained models (see instructions here).

heyzude commented 2 years ago

it is best_val_unseen from Dropbox, under / trained_models?