merge

main.py

@@ -25,7 +25,7 @@ if __name__ == '__main__':
     dropout = 0.2
     epochs = 100
     nb_log_epoch = 5
-    nb_words = 200
+    nb_words = 100
     temperature = 1.0
 
     print("device :", device)
@@ -35,6 +35,7 @@ if __name__ == '__main__':
     ###############################################################################
     corpus = data.Corpus(data_path, model_path is not None, split_train_test_valid)
     train_data = data.batchify(corpus.train, batch_size, device)
+    test_data = data.batchify(corpus.test, batch_size, device)
     val_data = data.batchify(corpus.valid, batch_size, device)
 
     ###############################################################################
@@ -58,6 +59,7 @@ if __name__ == '__main__':
             criterion=criterion,
             ntokens=ntokens,
             train_data=train_data,
+            test_data=test_data,
             val_data=val_data,
             sequence_length=sequence_length,
             lr=learning_rate,

musicgenerator.py

@@ -94,6 +94,7 @@ def train(model: MusicGenerator,
           criterion: nn.NLLLoss,
           ntokens: int,
           train_data: torch.Tensor,
+          test_data: torch.Tensor,
           val_data: torch.Tensor,
           sequence_length: int,
           lr: float,
@@ -123,6 +124,12 @@ def train(model: MusicGenerator,
             else:
                 # Anneal the learning rate if no improvement has been seen in the validation dataset.
                 lr /= 4.0
+        test_loss = evaluate(model, criterion, test_data, ntokens, sequence_length)
+        print('=' * 89)
+        print('| End of training | test loss {:5.2f} | test ppl {:8.2f}'.format(
+            test_loss, math.exp(test_loss)))
+        print('=' * 89)
+
 
     except KeyboardInterrupt:
         print('-' * 89)
@@ -152,7 +159,6 @@ def __train(model: MusicGenerator,
         loss = criterion(output, targets)
         loss.backward()
 
-        # `clip_grad_norm` helps prevent the exploding gradient problem in RNNs / LSTMs.
         clip = 0.25
         torch.nn.utils.clip_grad_norm_(model.parameters(), clip)
         for p in model.parameters():