Update evo_model.py

evo_model.py

@@ -1,8 +1,7 @@
 import torch
 import torch.nn as nn
-import torch.nn.functional as F
 
-class EvoTransformerV22(nn.Module):
+class EvoEncoder(nn.Module):
     def __init__(self, d_model=384, num_heads=6, ffn_dim=1024, num_layers=6, memory_enabled=False):
         super().__init__()
         self.embedding = nn.Embedding(30522, d_model)
@@ -17,12 +16,6 @@ class EvoTransformerV22(nn.Module):
         if memory_enabled:
             self.memory_token = nn.Parameter(torch.zeros(1, 1, d_model))
             self.memory_proj = nn.Linear(d_model, d_model)
-        self.pooling = nn.AdaptiveAvgPool1d(1)
-        self.classifier = nn.Sequential(
-            nn.Linear(d_model, 128),
-            nn.ReLU(),
-            nn.Linear(128, 2)
-        )
 
     def forward(self, input_ids):
         x = self.embedding(input_ids)
@@ -30,6 +23,23 @@ class EvoTransformerV22(nn.Module):
             mem = self.memory_token.expand(x.size(0), -1, -1)
             x = torch.cat([mem, x], dim=1)
         x = self.transformer(x)
-        x = x.permute(0, 2, 1)  # for pooling
+        return x
+
+class EvoTransformerV22(nn.Module):
+    def __init__(self):
+        super().__init__()
+        self.encoder = EvoEncoder(
+            d_model=384,
+            num_heads=6,
+            ffn_dim=1024,
+            num_layers=6,
+            memory_enabled=True
+        )
+        self.pooling = nn.AdaptiveAvgPool1d(1)
+        self.classifier = nn.Linear(384, 2)
+
+    def forward(self, input_ids):
+        x = self.encoder(input_ids)
+        x = x.permute(0, 2, 1)  # [B, D, T]
         x = self.pooling(x).squeeze(-1)
         return self.classifier(x)