Skip to content

Latest commit

 

History

History
29 lines (19 loc) · 714 Bytes

albef.md

File metadata and controls

29 lines (19 loc) · 714 Bytes

ALBEF with Selective Sampling

Pretrain

For pretraining, run the following commands:

cd ALBEF/
python -m torch.distributed.launch --nproc_per_node=1 --use_env Pretrain.py --config ./configs/Pretrain.yaml --output_dir output/Pretrain 

Check ./configs/Pretrain.yaml for selective sampling settings.

Finetuning for Image-to-Report and Report-to-Image Retrieval

For retrieval task, run the following command:

python -m torch.distributed.launch --nproc_per_node=1 --use_env Retrieval.py \
--config ./configs/Retrieval_coco.yaml \
--output_dir output/Retrieval \
--checkpoint [Pretrained checkpoint]

Check ./configs/Retrieval_coco.yaml for selective sampling settings.