search

naver-ai / pcmepp

Official Pytorch implementation of "Improved Probabilistic Image-Text Representations" (ICLR 2024)
https://naver-ai.github.io/pcmepp/
Other
51 stars 1 forks source link

Config file for VSE_infty with ViT-L/14 #8

Closed jinhyunj closed 1 month ago

jinhyunj commented 1 month ago

Hi, thanks for the great work! I'm doing an experiment with your code and want to know the exact configuration for the results of VSE_infty with ViT-L/14 backbone. Could you please provide the corresponding config file? Thank you!

SanghyukChun commented 1 month ago

It does not need a new configuration file. The following command will work (for 8 GPUs)

python3 train.py ./configs/others/vse_infty.yaml --dataloader__data_path /path/to/dataset --model__backbone_source clip_ViT-L/14 --model__img_dim 1024 --dataloader__batch_size 16 --train__all_gather 1