FCNet/config.py at main · oucailab/FCNet · GitHub

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
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
"""
Author: 爱吃菠萝 1690608011@qq.com
Date: 2024-09-26 18:28:47
LastEditors: 爱吃菠萝 1690608011@qq.com
LastEditTime: 2024-10-16 14:38:03
FilePath: /root/OSI-SAF/config.py
Description:

Copyright (c) 2024 by ${git_name_email}, All Rights Reserved.
"""


class Configs:
    def __init__(self):
        # self.model = "ConvLSTM"
        # self.model = "PredRNN"
        # self.model = "SimVP"
        # self.model = "TAU"
        # self.model = "Swin_Transformer"
        self.model = "FCNet"
        # self.model = "SICFN"

        # paths
        self.data_paths = "data/data_path.txt"
        self.train_log_path = "train_logs"
        self.test_results_path = "test_results"

        # trainer related
        self.batch_size_vali = 16
        self.batch_size = 4
        self.lr = 5e-4
        self.weight_decay = 1e-2
        self.num_epochs = 200
        self.early_stop = True
        self.patience = self.num_epochs // 10
        self.gradient_clip = True
        self.clip_threshold = 1.0
        self.num_workers = 32

        # data related
        self.img_size = (432, 432)

        # 这种数据的加载方式有些不方便，有更好的方式，可以联系我（爱吃菠萝）
        self.input_dim = 1  #  input_dim: 输入张量对应的通道数
        self.output_dim = 1  #  output_dim: 输入张量对应的通道数

        self.input_length = 14  # 每轮训练输入多少张数据
        self.pred_length = 14  # 每轮训练输出多少张数据

        self.input_gap = 1  # 每张输入数据之间的间隔
        self.pred_gap = 1  # 每张输出数据之间的间隔

        self.pred_shift = self.pred_gap * self.pred_length

        self.train_period = (19910101, 20100101)
        self.eval_period = (20100101, 20151231)

        # model related
        if self.model == "FCNet":
            self.hid_S = 64
            self.hid_T = 256
            self.N_T = 8
            self.N_S = 4
            self.spatio_kernel_enc = 3
            self.spatio_kernel_dec = 3
            self.mlp_ratio = 4.0
            self.drop = 0.05
            self.drop_path = 0.05
            self.dropcls = 0.05
            self.patch_embed_size = (8, 8)
            self.patch_size = (2, 2)

        elif self.model == "SICFN":
            self.hid_S = 64
            self.hid_T = 256
            self.N_T = 8
            self.N_S = 4
            self.spatio_kernel_enc = 3
            self.spatio_kernel_dec = 3
            self.mlp_ratio = 4.0
            self.drop = 0.05
            self.drop_path = 0.05
            self.dropcls = 0.05
            self.patch_embed_size = (8, 8)
            self.patch_size = (2, 2)
            self.fno_blocks = 8
            self.fno_bias = True
            self.fno_softshrink = 0.05

        elif self.model == "ConvLSTM":
            self.kernel_size = (3, 3)
            self.patch_size = (2, 2)
            self.hidden_dim = (
                64,
                64,
                64,
                64,
            )  # hidden_dim: 隐藏状态的神经单元个数，也就是隐藏层的节点数。

        elif self.model == "PredRNN":
            self.kernel_size = (3, 3)
            self.patch_size = (2, 2)
            self.hidden_dim = (
                64,
                64,
                64,
                64,
            )  # hidden_dim: 隐藏状态的神经单元个数，也就是隐藏层的节点数。
            self.layer_norm = False

        elif self.model == "SimVP":
            self.hid_S = 64
            self.hid_T = 256
            self.N_T = 8
            self.N_S = 4
            self.spatio_kernel_enc = 3
            self.spatio_kernel_dec = 3
            self.patch_size = (2, 2)

        elif self.model == "SimVP":
            self.hid_S = 64
            self.hid_T = 256
            self.N_T = 8
            self.N_S = 4
            self.spatio_kernel_enc = 3
            self.spatio_kernel_dec = 3
            self.patch_size = (2, 2)
            self.mlp_ratio = 4.0
            self.drop = 0.05
            self.drop_path = 0.05

        elif self.model == "TAU":
            self.hid_S = 64
            self.hid_T = 256
            self.N_T = 8
            self.N_S = 4
            self.spatio_kernel_enc = 3
            self.spatio_kernel_dec = 3
            self.patch_size = (2, 2)
            self.mlp_ratio = 4.0
            self.drop = 0.05
            self.drop_path = 0.05

        elif self.model == "Swin_Transformer":
            self.hid_S = 64
            self.hid_T = 256
            self.N_T = 8
            self.N_S = 4
            self.spatio_kernel_enc = 3
            self.spatio_kernel_dec = 3
            self.patch_size = (2, 2)
            self.mlp_ratio = 4.0
            self.drop = 0.05
            self.drop_path = 0.05


configs = Configs()