Make attn windowing optional

config.json

@@ -31,6 +31,7 @@
     "lr": 0.001,            // Initial learning rate. If Noam decay is active, maximum learning rate.
     "lr_decay": false,      // if true, Noam learning rate decaying is applied through training.
     "warmup_steps": 4000,   // Noam decay steps to increase the learning rate from 0 to "lr"
+    "windowing": true,      // Enables attention windowing. Used only in eval mode.
 
     "batch_size": 32,       // Batch size for training. Lower values than 32 might cause hard to learn attention.
     "eval_batch_size":32,   

layers/attention.py

@@ -160,8 +160,7 @@ class AttentionRNNCell(nn.Module):
             mask = mask.view(memory.size(0), -1)
             alignment.masked_fill_(1 - mask, -float("inf"))
         # Windowing
-        if not self.training:
-            # print(" > Windowing active")
+        if not self.training and self.windowing:
             back_win = self.win_idx - self.win_back
             front_win = self.win_idx + self.win_front
             if back_win > 0: