-
Notifications
You must be signed in to change notification settings - Fork 3
/
Copy patheval.sh
executable file
·14 lines (9 loc) · 1.4 KB
/
eval.sh
1
2
3
4
5
6
7
8
9
10
11
12
13
14
#! /bin/bash
source /zhouyuanen/anaconda3/bin/activate base
cd /zhouyuanen/new/cbtic
# For online evaluation
# python eval.py --input_json data/cocotest.json --input_fc_dir data/mscoco_VinVL/cocobu_test2014/cocobu_fc --input_att_dir data/mscoco_VinVL/cocobu_test2014/cocobu_att --input_label_h5 data/cocotalk_bw_label.h5 --language_eval 0 --model save/nsc-transformer-cb-VinVL-feat/model-best.pth --infos_path save/nsc-transformer-cb-VinVL-feat/infos_nsc-transformer-cb-VinVL-feat-best.pkl --batch_size 128 --beam_size 2 --id captions_test2014_cbtic_results
# For offline ensemble evaluation
# python eval_ensemble.py --ids nsc-transformer-cb-VinVL-feat nsc-transformer-cb-VinVL-feat-seed1 nsc-transformer-cb-VinVL-feat-seed2 nsc-transformer-cb-VinVL-feat-seed3 --weights 1 1 1 1 --beam_size 2 --split test
# For online ensemble evaluation
python eval_ensemble.py --split test --language_eval 0 --ids nsc-transformer-cb-VinVL-feat nsc-transformer-cb-VinVL-feat-seed1 nsc-transformer-cb-VinVL-feat-seed2 nsc-transformer-cb-VinVL-feat-seed3 --weights 1 1 1 1 --input_json data/cocotest.json --input_fc_dir data/mscoco_VinVL/cocobu_test2014/cocobu_fc --input_att_dir data/mscoco_VinVL/cocobu_test2014/cocobu_att --input_label_h5 data/cocotalk_bw_label.h5 --language_eval 0 --batch_size 128 --beam_size 2 --id captions_test2014_cbtic_results