class DummyAttentionGeneratorConsumerTest(TestCase):
def setUp(self):
criteriondim = 30
datadim = 20
innerdim = 25
batsize = 33
seqlen = 11
self.attgenshape = (batsize, seqlen)
self.attconshape = (batsize, datadim)
self.attgenc = self.getattgenc()
self.attgen = self.attgenc(indim=criteriondim + datadim,
innerdim=innerdim)
self.attgenparams = self.getattgenparams()
self.attcon = WeightedSumAttCon()
self.att = Attention(self.attgen, self.attcon)
self.criterion_val = np.random.random(
(batsize, criteriondim)).astype("float32")
self.data_val = np.random.random(
(batsize, seqlen, datadim)).astype("float32")
def getattgenc(self):
return LinearSumAttentionGenerator
def getattgenparams(self):
return {self.attgen.W}
def test_generator_shape(self):
pred = self.attgen.predict(self.criterion_val, self.data_val)
self.assertEqual(pred.shape, self.attgenshape)
def test_generator_param_prop(self):
self.attgen.predict(self.criterion_val, self.data_val)
allparams = self.attgen.output.allparams
self.assertSetEqual(allparams, self.attgenparams)
def test_consumer_shape(self):
pred = self.att.predict(self.criterion_val, self.data_val)
self.assertEqual(pred.shape, self.attconshape)
def test_consumer_param_prop(self):
self.att.predict(self.criterion_val, self.data_val)
allparams = self.att.output.allparams
self.assertSetEqual(allparams, self.attgenparams)
class DummyAttentionGeneratorConsumerTest(TestCase):
def setUp(self):
criteriondim = 30
datadim = 20
innerdim = 25
batsize = 33
seqlen = 11
self.attgenshape = (batsize, seqlen)
self.attconshape = (batsize, datadim)
self.attgenc = self.getattgenc()
self.attgen = self.attgenc(indim=criteriondim + datadim, innerdim=innerdim)
self.attgenparams = self.getattgenparams()
self.attcon = WeightedSumAttCon()
self.att = Attention(self.attgen, self.attcon)
self.criterion_val = np.random.random((batsize, criteriondim)).astype("float32")
self.data_val = np.random.random((batsize, seqlen, datadim)).astype("float32")
def getattgenc(self):
return LinearSumAttentionGenerator
def getattgenparams(self):
return {self.attgen.W}
def test_generator_shape(self):
pred = self.attgen.predict(self.criterion_val, self.data_val)
self.assertEqual(pred.shape, self.attgenshape)
def test_generator_param_prop(self):
self.attgen.predict(self.criterion_val, self.data_val)
allparams = self.attgen.output.allparams
self.assertSetEqual(allparams, self.attgenparams)
def test_consumer_shape(self):
pred = self.att.predict(self.criterion_val, self.data_val)
self.assertEqual(pred.shape, self.attconshape)
def test_consumer_param_prop(self):
self.att.predict(self.criterion_val, self.data_val)
allparams = self.att.output.allparams
self.assertSetEqual(allparams, self.attgenparams)
class DummyAttentionGeneratorConsumerTest(TestCase):
def setUp(self):
criteriondim = 20
datadim = 20
innerdim = 30
batsize = 33
seqlen = 11
self.attgenshape = (batsize, seqlen)
self.attconshape = (batsize, datadim)
self.attgen = self.getattgenc(critdim=criteriondim, datadim=datadim, attdim=innerdim)
self.attgenparams = self.getattgenparams()
self.attcon = WeightedSumAttCon()
self.att = Attention(self.attgen, self.attcon)
self.criterion_val = np.random.random((batsize, criteriondim)).astype("float32")
self.data_val = np.random.random((batsize, seqlen, datadim)).astype("float32")
def getattgenc(self, critdim=None, datadim=None, attdim=None):
return AttGen(DotDistance())
def getattgenparams(self):
return set()
def test_generator_shape(self):
pred = self.attgen.predict(self.criterion_val, self.data_val)
self.assertEqual(pred.shape, self.attgenshape)
def test_generator_param_prop(self):
_, outps = self.attgen.autobuild(self.criterion_val, self.data_val)
allparams = outps[0].allparams
self.assertSetEqual(allparams, self.attgenparams)
def test_consumer_shape(self):
pred = self.att.predict(self.criterion_val, self.data_val)
self.assertEqual(pred.shape, self.attconshape)
def test_consumer_param_prop(self):
_, outps = self.att.autobuild(self.criterion_val, self.data_val)
allparams = outps[0].allparams
self.assertSetEqual(allparams, self.attgenparams)
