def testSequenceClassifierWithSelfAttentionEncoder(self):
# SelfAttentionEncoder does not return a state, so test that the classifier
# does not crash on this.
model = models.SequenceClassifier(
inputters.WordEmbedder(10),
encoders.SelfAttentionEncoder(num_layers=2, num_units=16, num_heads=4, ffn_inner_dim=32))
features_file, labels_file, data_config = self._makeToyClassifierData()
model.initialize(data_config)
dataset = model.examples_inputter.make_training_dataset(features_file, labels_file, 16)
features, labels = iter(dataset).next()
model(features, labels, training=True)
def testSequenceClassifier(self, mode):
model = models.SequenceClassifier(inputters.WordEmbedder(10),
encoders.MeanEncoder())
features_file, labels_file, data_config = self._makeToyClassifierData()
params = {"optimizer": "SGD", "learning_rate": 0.1}
self._testGenericModel(model,
mode,
features_file,
labels_file,
data_config,
prediction_heads=["classes"],
metrics=["accuracy"],
params=params)
def testSequenceClassifier(self, mode):
model = models.SequenceClassifier(
inputters.WordEmbedder("source_vocabulary", 10),
encoders.MeanEncoder(), "target_vocabulary")
features_file, labels_file, metadata = self._makeToyClassifierData()
params = {
"optimizer": "GradientDescentOptimizer",
"learning_rate": 0.1
}
self._testGenericModel(model,
mode,
features_file,
labels_file,
metadata,
prediction_heads=["classes"],
metrics=["accuracy"],
params=params)
