feat: minamo model 修改为余弦退火+重启调度

minamo/model/loss.py

@@ -1,7 +1,7 @@
 import torch.nn as nn
 
 class MinamoLoss(nn.Module):
-    def __init__(self, vision_weight=1, topo_weight=0):
+    def __init__(self, vision_weight=0.4, topo_weight=0.6):
         super().__init__()
         self.vision_weight = vision_weight
         self.topo_weight = topo_weight

minamo/model/vision.py

@@ -14,21 +14,21 @@ class MinamoVisionModel(nn.Module):
             nn.Conv2d(embedding_dim, conv_channels, 3, padding=1),
             nn.BatchNorm2d(conv_channels),
             CBAM(conv_channels),
-            nn.ReLU(),
+            nn.GELU(),
             nn.MaxPool2d(2),
-            nn.Dropout2d(0.3),
+            nn.Dropout2d(0.4),
             
             nn.Conv2d(conv_channels, conv_channels*2, 3, padding=1),
             nn.BatchNorm2d(conv_channels*2),
             CBAM(conv_channels*2),
-            nn.ReLU(),
+            nn.GELU(),
             nn.MaxPool2d(2),
-            nn.Dropout2d(0.3),
+            nn.Dropout2d(0.4),
             
             nn.Conv2d(conv_channels*2, conv_channels*4, 3, padding=1),
             nn.BatchNorm2d(conv_channels*4),
             CBAM(conv_channels*4),
-            nn.ReLU(),
+            nn.GELU(),
             
             nn.AdaptiveMaxPool2d(1)
         )

minamo/train.py

@@ -12,7 +12,7 @@ device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 os.makedirs("result", exist_ok=True)
 os.makedirs("result/minamo_checkpoint", exist_ok=True)
 
-epochs = 100
+epochs = 150
 
 def collate_fn(batch):
     """动态处理不同尺寸地图的批处理"""
@@ -52,8 +52,8 @@ def train():
     )
     
     # 设定优化器与调度器
-    optimizer = optim.AdamW(model.parameters(), lr=5e-4, weight_decay=1e-2)
+    optimizer = optim.AdamW(model.parameters(), lr=1e-3, weight_decay=1e-2)
-    scheduler = optim.lr_scheduler.CosineAnnealingLR(optimizer, T_max=epochs, eta_min=1e-6)
+    scheduler = optim.lr_scheduler.CosineAnnealingWarmRestarts(optimizer, T_0=10, T_mult=2, eta_min=1e-6)
     criterion = MinamoLoss()
     
     # 开始训练

shared/attention.py

@@ -9,7 +9,7 @@ class ChannelAttention(nn.Module):
         self.channel_att = nn.Sequential(
             nn.AdaptiveAvgPool2d(1),
             nn.Conv2d(channels, channels//reduction, 1),
-            nn.ReLU(),
+            nn.GELU(),
             nn.Conv2d(channels//reduction, channels, 1),
             nn.Sigmoid()
         )