-
Notifications
You must be signed in to change notification settings - Fork 0
/
readme_dev.txt
79 lines (55 loc) · 2.57 KB
/
readme_dev.txt
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
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
# Files organisation
`gaga_model`
Definition of the Discriminator and the Generator nets.
`gaga_trainer`
`gaga_helpers`
- gaga_penalty ?
- gaga_plot_helpers
- LOGGER
# Options and parameters
| loss_type | wasserstein non-saturating-bce |
| penalty_type | zero_penalty clamp_penalty gradient_penalty gradient_penalty_max |
| penalty_weight | |
| layer_norm | true/false |
| d_layers g_layers | |
| d_dim g_dim | |
| z_dim | |
| clamp_lower clamp_upper | |
|-------------------------+------------------------------------------------------------------|
| optimiser | adam RMSprop |
| shuffle | true/false |
"#": "adam optimiser: regularisation L2 (for adam only) ; zero if no regul",
"d_weight_decay": 0.5,
"g_weight_decay": 0.5,
"#": "Real and Fake instance Gaussian noise sigma. -1 for none. ",
"r_instance_noise_sigma": 0,
"f_instance_noise_sigma": 0,
"#": "adam optimiser: beta",
"beta_1": "0.9",
"beta_2": "0.999",
"#": "optimiser: learning rate",
"d_learning_rate": 1e-4,
"g_learning_rate": 1e-4,
"#": "optimiser: number of D and G update by epoch",
"d_nb_update": 2,
"g_nb_update": 1,
"#": "optimiser: max nb of epoch (iteration)",
"epoch": 80000,
"#": "optimiser: nb of samples by batch",
"batch_size": 1000,
"#": "Smooth fake/real labels instead of zero/one",
"#label_smoothing": 0.2,
"#": "---------------------------------------------------------------------",
"#": " DATA ",
"#": "---------------------------------------------------------------------",
"keys": "X dY",
"#": "---------------------------------------------------------------------",
"#": " GENERAL ",
"#": "---------------------------------------------------------------------",
"#": "gpu_mode: true false auto",
"gpu_mode": "auto",
"#": "save Generator and info every epoch",
"dump_epoch_start": 0,
"dump_epoch_every": 5000,
"dump_last_n_epoch": 0
}