forked from CVLAB-Unibo/Real-time-self-adaptive-deep-stereo
-
Notifications
You must be signed in to change notification settings - Fork 1
/
sweep_config_example.yaml
47 lines (47 loc) · 959 Bytes
/
sweep_config_example.yaml
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
35
36
37
38
39
40
41
42
43
44
45
46
47
command:
- python
- ${program}
- ${args}
- "--log_tensorboard"
- "--sweep"
program: train_wandb.py
project: madnet-keras
method: grid
metric:
goal: minimize
name: loss
parameters:
search_range:
value: 2
lr:
values: [0.0005, 0.0001, 0.00001]
batch_size:
values: [1, 20]
num_epochs:
values: [100, 150, 200]
epoch_steps:
value: 1000
save_freq:
value: 10000
train_left_dir:
value: "/path/to/train/left_rectified"
train_right_dir:
value: "/path/to/train/right_rectified"
train_disp_dir:
value: "/path/to/train/left_disparity"
val_left_dir:
value: "/path/to/val/left_rectified"
val_right_dir:
value: "/path/to/val/right_rectified"
val_disp_dir:
value: "/path/to/val/left_disparity"
output_dir:
value: "/path/to/output_folder"
checkpoint_path:
value: "/path/to/weights.ckpt"
height:
value: 480
width:
value: 640
dataset_name:
value: "FlyingThings3D"