refactor(config): 统一配置文件命名并调整模型参数

更新多个模型的配置文件命名格式,统一使用大驼峰格式
调整SolarEnergy和BJTaxi数据集的输入维度和批量大小
删除旧命名格式的配置文件并添加新的配置文件
修改训练脚本中的模型和数据集列表用于调试
This commit is contained in:
czzhangheng 2025-12-15 21:33:28 +08:00
parent 5e52f23c8d
commit 659b41f612
16 changed files with 10 additions and 66 deletions

View File

@ -6,11 +6,11 @@ basic:
seed: 2023
data:
batch_size: 512
batch_size: 64
column_wise: false
days_per_week: 7
horizon: 24
input_dim: 6
input_dim: 1
lag: 24
normalizer: std
num_nodes: 137
@ -25,7 +25,7 @@ model:
train:
batch_size: 512
batch_size: 64
debug: false
early_stop: true
early_stop_patience: 15

View File

@ -10,7 +10,7 @@ data:
column_wise: false
days_per_week: 7
horizon: 24
input_dim: 6
input_dim: 1
lag: 24
normalizer: std
num_nodes: 137

5
config/REPST/BJTaxi-InFlow.yaml Normal file → Executable file
View File

@ -1,6 +1,6 @@
basic:
dataset: BJTaxi-InFlow
device: cuda:1
device: cuda:0
mode: train
model: REPST
seed: 2023
@ -27,7 +27,6 @@ model:
input_dim: 1
n_heads: 1
num_nodes: 1024
output_dim: 1
patch_len: 6
pred_len: 24
seq_len: 24
@ -41,7 +40,7 @@ train:
early_stop_patience: 15
epochs: 100
grad_norm: false
log_step: 1000
log_step: 100
loss_func: mae
lr_decay: true
lr_decay_rate: 0.3

View File

@ -1,55 +0,0 @@
basic:
dataset: BJTaxi-InFlow
device: cuda:0
mode: train
model: REPST
seed: 2023
data:
batch_size: 16
column_wise: false
days_per_week: 7
horizon: 24
input_dim: 1
lag: 24
normalizer: std
num_nodes: 1024
steps_per_day: 48
test_ratio: 0.2
val_ratio: 0.2
model:
d_ff: 128
d_model: 64
dropout: 0.2
gpt_layers: 9
gpt_path: ./GPT-2
input_dim: 1
n_heads: 1
num_nodes: 1024
patch_len: 6
pred_len: 24
seq_len: 24
stride: 7
word_num: 1000
train:
batch_size: 16
debug: false
early_stop: true
early_stop_patience: 15
epochs: 100
grad_norm: false
log_step: 100
loss_func: mae
lr_decay: true
lr_decay_rate: 0.3
lr_decay_step: 5,20,40,70
lr_init: 0.003
mae_thresh: None
mape_thresh: 0.001
max_grad_norm: 5
output_dim: 1
plot: false
real_value: true
weight_decay: 0

View File

@ -89,10 +89,10 @@ def main(model, data, debug=False):
if __name__ == "__main__":
# 调试用
# model_list = ["iTransformer", "PatchTST", "HI"]
model_list = ["iTransformer", "PatchTST", "HI"]
# model_list = ["ASTRA_v2", "GWN", "REPST", "STAEFormer", "MTGNN"]
model_list = ["iTransformer"]
dataset_list = ["AirQuality", "SolarEnergy", "PEMS-BAY", "METR-LA", "BJTaxi-InFlow", "BJTaxi-OutFlow", "NYCBike-InFlow", "NYCBike-OutFlow"]
# model_list = ["iTransformer"]
# dataset_list = ["AirQuality", "SolarEnergy", "PEMS-BAY", "METR-LA", "BJTaxi-InFlow", "BJTaxi-OutFlow", "NYCBike-InFlow", "NYCBike-OutFlow"]
# dataset_list = ["AirQuality"]
# dataset_list = ["AirQuality", "SolarEnergy", "METR-LA", "NYCBike-InFlow", "NYCBike-OutFlow"]
dataset_list = ["AirQuality", "SolarEnergy", "METR-LA", "NYCBike-InFlow", "NYCBike-OutFlow"]
main(model_list, dataset_list, debug = True)