-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathtests.sh
35 lines (25 loc) · 1.35 KB
/
tests.sh
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
# python get_file_info_from_dump.py jay_imagenet_train_0.1.yaml
python clickme_prepare_maps_for_modeling.py jay_imagenet_val_0.1.yaml
python clickme_prepare_maps_for_modeling.py jay_imagenet_for_co3d_val_0.1.yaml
python clickme_prepare_maps_for_modeling.py co3d_val.yaml
python compute_human_ceiling_split_half.py jay_imagenet_train_0.1.yaml
python compute_human_ceiling_split_half.py jay_imagenet_val_0.1.yaml
python compute_human_ceiling_split_half.py jay_imagenet_for_co3d_val_0.1.yaml
python compute_human_ceiling_split_half.py co3d_val.yaml
python clickme_prepare_maps_for_modeling.py jay_imagenet_train_0.1.yaml
# python compute_human_ceiling_hold_one_out.py jay_imagenet_train_0.1.yaml
# python compute_human_ceiling_hold_one_out.py jay_imagenet_val_0.1.yaml
# python compute_human_ceiling_hold_one_out.py jay_imagenet_for_co3d_val_0.1.yaml
# python compute_human_ceiling_hold_one_out.py co3d_val.yaml
# Train:
# Mean human correlation full set: 0.08664774537280222
# Null correlations full set: 0.027026184604559984
# Val:
# Mean human correlation full set: 0.09120038285655054
# Null correlations full set: 0.028687646741805756
# Val_imagenet_co3d:
# Mean human correlation full set: 0.09161272922789762
# Null correlations full set: 0.02913157979315741
# co3d:
# Mean human correlation full set: 0.16571727314019297
# Null correlations full set: 0.023708055710504