training script question

alinlab / ifseg

IFSeg: Image-free Semantic Segmentation via Vision-Language Model (CVPR 2023)

78 stars 9 forks source link

training script question #9

Open ruida opened 5 months ago

ruida commented 5 months ago

Hi,

coco_fine.sh: data=${data_dir}/fineseg_refined_val2017.tsv,${data_dir}/fineseg_refined_val2017.tsv coco_unseen.sh: data=${data_dir}/unseen_val2017.tsv,${data_dir}/unseen_val2017.tsv

Could you leave me input on why the two valid dataset TSV files were specified in the data variable?

Thanks,

--Ruida