Is augmented data used for R2R task fine-tuning?

MarSaKi / VLN-BEVBert

[ICCV 2023} Official repo of "BEVBert: Multimodal Map Pre-training for Language-guided Navigation"

184 stars 5 forks source link

Is augmented data used for R2R task fine-tuning? #26

Closed DITINGmisa closed 1 week ago

DITINGmisa commented 1 week ago

Hello MarSaKi, thank you for your excellent work! I notice the arguments "--aug datasets/R2R/annotations/prevalent_aug_train_enc.json" at VLN-BEVBert/scripts/ft_r2r.bash. When you produce the result in your paper, is augmented data used for R2R task fine-tuning?

MarSaKi commented 1 week ago

yes "prevalent_aug_train_enc.json" has been used for fine-tuning