new dataset pipeline, test_engine_train_new_dataset_pipeline

Seems to work. At least for this simple case.
rwth-i6#292

tests/test_TFEngine.py

@@ -246,6 +246,36 @@ def test_engine_train():
   engine.finalize()
 
 
+
+def test_engine_train_new_dataset_pipeline():
+  from GeneratingDataset import DummyDataset
+  seq_len = 5
+  n_data_dim = 2
+  n_classes_dim = 3
+  train_data = DummyDataset(input_dim=n_data_dim, output_dim=n_classes_dim, num_seqs=5, seq_len=seq_len)
+  train_data.init_seq_order(epoch=1)
+  cv_data = DummyDataset(input_dim=n_data_dim, output_dim=n_classes_dim, num_seqs=3, seq_len=seq_len)
+  cv_data.init_seq_order(epoch=1)
+
+  config = Config()
+  config.update({
+    "model": "%s/model" % _get_tmp_dir(),
+    "num_outputs": n_classes_dim,
+    "num_inputs": n_data_dim,
+    "network": {"output": {"class": "softmax", "loss": "ce"}},
+    "start_epoch": 1,
+    "num_epochs": 2,
+    "max_seqs": 2,
+    "dataset_pipeline": True
+  })
+  _cleanup_old_models(config)
+  engine = Engine(config=config)
+  engine.init_train_from_config(config=config, train_data=train_data, dev_data=cv_data, eval_data=None)
+  assert engine.dataset_provider
+  engine.train()
+  engine.finalize()
+
+
 def test_engine_train_uneven_batches():
   rnd = numpy.random.RandomState(42)
   from GeneratingDataset import StaticDataset