diff --git a/misc/utils.py b/misc/utils.py
index 611dbff..fa28a20 100644
--- a/misc/utils.py
+++ b/misc/utils.py
@@ -233,7 +233,7 @@ def __getattr__(self, name):
 class ReduceLROnPlateau(object):
     "Optim wrapper that implements rate."
     def __init__(self, optimizer, mode='min', factor=0.1, patience=10, verbose=False, threshold=0.0001, threshold_mode='rel', cooldown=0, min_lr=0, eps=1e-08):
-        self.scheduler = optim.lr_scheduler.ReduceLROnPlateau(optimizer, mode, factor, patience, verbose, threshold, threshold_mode, cooldown, min_lr, eps)
+        self.scheduler = optim.lr_scheduler.ReduceLROnPlateau(optimizer, mode, factor, patience, threshold, threshold_mode, cooldown, min_lr, eps, verbose)
         self.optimizer = optimizer
         self.current_lr = get_lr(optimizer)
         
diff --git a/models/AttModel.py b/models/AttModel.py
index 02c3c9a..bda0e1a 100644
--- a/models/AttModel.py
+++ b/models/AttModel.py
@@ -31,6 +31,8 @@
 
 def sort_pack_padded_sequence(input, lengths):
     sorted_lengths, indices = torch.sort(lengths, descending=True)
+    sorted_lengths=sorted_lengths.to(int)
+    tmp = pack_padded_sequence(input[indices], sorted_lengths.cpu(), batch_first=True)
     tmp = pack_padded_sequence(input[indices], sorted_lengths, batch_first=True)
     inv_ix = indices.clone()
     inv_ix[indices] = torch.arange(0,len(indices)).type_as(inv_ix)
@@ -791,4 +793,4 @@ def core(self, xt, fc_feats, att_feats, p_att_feats, state, att_masks):
     def _prepare_feature(self, fc_feats, att_feats, att_masks):
         fc_feats = self.fc_embed(fc_feats)
 
-        return fc_feats, None, None, None
\ No newline at end of file
+        return fc_feats, None, None, None