Update core/train_eval.py

core/train_eval.py

@@ -22,22 +22,23 @@ def mean_absolute_percentage_error(y_true, y_pred):
     y_true, y_pred = np.array(y_true), np.array(y_pred)
     non_zero = np.abs(y_true) > 0
     if np.sum(non_zero) == 0:
-        return np.nan  # Return NaN if all true values are zero
+        return np.nan
     return np.mean(np.abs((y_true[non_zero] - y_pred[non_zero]) / y_true[non_zero])) * 100
 
 
 def train_and_evaluate(
     df,
+    future_df,  # Added
     model_cls,
     horizon=1,
     hidden=64,
     layers=1,
     epochs=50,
     lr=0.001,
-    beta1=0.9,  # Added
-    beta2=0.999,  # Added
-    weight_decay=0.01,  # Added
-    dropout=0.2,  # Added
+    beta1=0.9,
+    beta2=0.999,
+    weight_decay=0.01,
+    dropout=0.2,
     window=30,
     test_split=0.2,
     device="cuda" if torch.cuda.is_available() else "cpu",
@@ -76,6 +77,7 @@ def train_and_evaluate(
     model = model_cls(input_size=input_dim, hidden_size=hidden, num_layers=layers, output_size=horizon, dropout=dropout).to(device)
     optimizer = torch.optim.AdamW(model.parameters(), lr=lr, betas=(beta1, beta2), weight_decay=weight_decay)
     loss_fn = nn.MSELoss()
+    scheduler = torch.optim.lr_scheduler.ReduceLROnPlateau(optimizer, mode='min', patience=5, factor=0.5, threshold=1e-4, verbose=True)
 
     train_losses = []
     val_losses = []
@@ -108,8 +110,10 @@ def train_and_evaluate(
         val_loss /= len(val_loader)
         val_losses.append(val_loss)
 
+        scheduler.step(val_loss)
+
         if verbose and (epoch + 1) % 10 == 0:
-            print(f"Epoch {epoch+1}/{epochs} - Train Loss: {train_losses[-1]:.4f}, Val Loss: {val_losses[-1]:.4f}")
+            print(f"Epoch {epoch+1}/{epochs} - Train Loss: {train_losses[-1]:.4f}, Val Loss: {val_losses[-1]:.4f}, LR: {optimizer.param_groups[0]['lr']:.6f}")
 
         if val_loss < best_val_loss:
             best_val_loss = val_loss
@@ -173,4 +177,8 @@ def train_and_evaluate(
 
     result["latest_prediction"] = future_pred_inv[0].tolist()
 
+    # Include actual future values if available
+    if not future_df.empty:
+        result["future_actuals"] = future_df['value'].values.tolist()[:horizon]
+
     return result