Merge pull request #161 from pskrunner14/develop

Integration test expected cuda bug resolved

.gitignore

@@ -78,3 +78,10 @@ target/
 .vagrant
 
 vocab_pickle
+
+# VS Code config
+.vscode/*
+
+# local dev folders
+data/*
+experiment/*

Vagrantfile

@@ -11,7 +11,7 @@ Vagrant.configure("2") do |config|
   # https://docs.vagrantup.com.
 
   # Every Vagrant development environment requires a box. You can search for
-  # boxes at https://atlas.hashicorp.com/search.
+  # boxes at https://app.vagrantup.com/boxes/search.
   config.vm.box = "ubuntu/trusty64"
 
   # Disable automatic box update checking. If you disable this, then
@@ -48,7 +48,7 @@ Vagrant.configure("2") do |config|
      vb.gui = false
 
      # Customize the amount of memory on the VM:
-     vb.memory = "2048"
+     vb.memory = "4096"
    end
   #
   # View the documentation for the provider you are using for more

docs/source/conf.py

@@ -64,7 +64,7 @@
 
 # General information about the project.
 project = u'pytorch-seq2seq'
-copyright = u'2017, pytorch-seq2seq Contritors'
+copyright = u'2018, pytorch-seq2seq Contributors'
 author = u'pytorch-seq2seq Contributors'
 
 # The version info for the project you're documenting, acts as replacement for

scripts/integration_test.py

@@ -88,7 +88,7 @@ def len_filter(example):
                              eos_id=tgt.eos_id, sos_id=tgt.sos_id)
         seq2seq = Seq2seq(encoder, decoder)
         if torch.cuda.is_available():
-            seq2seq.cuda()
+            seq2seq = seq2seq.cuda()
 
         for param in seq2seq.parameters():
             param.data.uniform_(-0.08, 0.08)

seq2seq/models/TopKDecoder.py

@@ -2,6 +2,8 @@
 import torch.nn.functional as F
 from torch.autograd import Variable
 
+CUDA = torch.cuda.is_available()
+
 def _inflate(tensor, times, dim):
         """
         Given a tensor, 'inflates' it along the given dimension by replicating each slice specified number of times (in-place)
@@ -116,6 +118,10 @@ def forward(self, inputs=None, encoder_hidden=None, encoder_outputs=None, functi
 
         # Initialize the input vector
         input_var = Variable(torch.transpose(torch.LongTensor([[self.SOS] * batch_size * self.k]), 0, 1))
+        if CUDA:
+            self.pos_index = self.pos_index.cuda()
+            input_var = input_var.cuda()
+            sequence_scores = sequence_scores.cuda()
 
         # Store decisions for backtracking
         stored_outputs = list()
@@ -223,9 +229,14 @@ def _backtrack(self, nw_output, nw_hidden, predecessors, symbols, scores, b, hid
         # the last hidden state of decoding.
         if lstm:
             state_size = nw_hidden[0][0].size()
-            h_n = tuple([torch.zeros(state_size), torch.zeros(state_size)])
+            if CUDA:
+                h_n = tuple([torch.zeros(state_size).cuda(), torch.zeros(state_size).cuda()])
+            else:
+                h_n = tuple([torch.zeros(state_size), torch.zeros(state_size)])
         else:
             h_n = torch.zeros(nw_hidden[0].size())
+            if CUDA:
+                h_n = h_n.cuda()
         l = [[self.rnn.max_length] * self.k for _ in range(b)]  # Placeholder for lengths of top-k sequences
                                                                 # Similar to `h_n`
 

tests/test_checkpoint.py

@@ -27,7 +27,7 @@ def test_get_latest_checkpoint(self, mock_listdir):
                                      '2017_05_22_09_47_31',
                                      '2017_05_23_10_47_29']
         latest_checkpoint = Checkpoint.get_latest_checkpoint(self.EXP_DIR)
-        self.assertEquals(latest_checkpoint,
+        self.assertEqual(latest_checkpoint,
                           os.path.join(self.EXP_DIR,
                                        'checkpoints/2017_05_23_10_47_29'))
 
@@ -50,15 +50,15 @@ def test_save_checkpoint_calls_torch_save(self, mock_open, mock_dill, mock_torch
 
         path = chk_point.save(self._get_experiment_dir())
 
-        self.assertEquals(2, mock_torch.save.call_count)
+        self.assertEqual(2, mock_torch.save.call_count)
         mock_torch.save.assert_any_call(state_dict,
                                         os.path.join(chk_point.path, Checkpoint.TRAINER_STATE_NAME))
         mock_torch.save.assert_any_call(mock_model,
                                         os.path.join(chk_point.path, Checkpoint.MODEL_NAME))
-        self.assertEquals(2, mock_open.call_count)
+        self.assertEqual(2, mock_open.call_count)
         mock_open.assert_any_call(os.path.join(path, Checkpoint.INPUT_VOCAB_FILE), ANY)
         mock_open.assert_any_call(os.path.join(path, Checkpoint.OUTPUT_VOCAB_FILE), ANY)
-        self.assertEquals(2, mock_dill.dump.call_count)
+        self.assertEqual(2, mock_dill.dump.call_count)
         mock_dill.dump.assert_any_call(mock_vocab,
                                        mock_open.return_value.__enter__.return_value)
 
@@ -80,11 +80,11 @@ def test_load(self, mock_open, mock_dill, mock_torch):
         mock_torch.load.assert_any_call(
             os.path.join("mock_checkpoint_path", Checkpoint.MODEL_NAME))
 
-        self.assertEquals(loaded_chk_point.epoch, torch_dict['epoch'])
-        self.assertEquals(loaded_chk_point.optimizer, torch_dict['optimizer'])
-        self.assertEquals(loaded_chk_point.step, torch_dict['step'])
-        self.assertEquals(loaded_chk_point.input_vocab, dummy_vocabulary)
-        self.assertEquals(loaded_chk_point.output_vocab, dummy_vocabulary)
+        self.assertEqual(loaded_chk_point.epoch, torch_dict['epoch'])
+        self.assertEqual(loaded_chk_point.optimizer, torch_dict['optimizer'])
+        self.assertEqual(loaded_chk_point.step, torch_dict['step'])
+        self.assertEqual(loaded_chk_point.input_vocab, dummy_vocabulary)
+        self.assertEqual(loaded_chk_point.output_vocab, dummy_vocabulary)
 
     def _get_experiment_dir(self):
         root_dir = os.path.dirname(os.path.realpath(__file__))