Skip to content

Speeding the tests for T5 #888

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Mar 21, 2023
Merged

Speeding the tests for T5 #888

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
51 changes: 26 additions & 25 deletions keras_nlp/models/t5/t5_backbone_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -25,7 +25,7 @@

class T5Test(tf.test.TestCase, parameterized.TestCase):
def setUp(self):
self.model = T5Backbone(
self.backbone = T5Backbone(
vocabulary_size=4,
num_layers=2,
num_heads=2,
Expand Down Expand Up @@ -53,12 +53,20 @@ def setUp(self):
).batch(2)

def test_valid_call_t5(self):
self.model(self.input_batch)
self.backbone(self.input_batch)

def test_token_embedding(self):
output = self.backbone.token_embedding(
self.input_batch["encoder_token_ids"]
)
self.assertEqual(output.shape, (2, 3, 4))

def test_name(self):
# Check default name passed through
self.assertRegexpMatches(self.model.name, "t5_backbone")
self.assertRegexpMatches(self.backbone.name, "t5_backbone")

def test_variable_sequence_length_call_t5(self):
for seq_length in (4, 5, 6):
for seq_length in (2, 3, 4):
input_data = {
"encoder_token_ids": tf.ones(
(self.batch_size, seq_length), dtype="int32"
Expand All @@ -73,36 +81,29 @@ def test_variable_sequence_length_call_t5(self):
(self.batch_size, seq_length), dtype="int32"
),
}
outputs = self.model(input_data)
outputs = self.backbone(input_data)
self.assertIn("encoder_sequence_output", outputs)
self.assertIn("decoder_sequence_output", outputs)

@parameterized.named_parameters(
("jit_compile_false", False), ("jit_compile_true", True)
)
def test_t5_compile(self, jit_compile):
self.model.compile(jit_compile=jit_compile)
outputs = self.model.predict(self.input_batch)
self.assertIn("encoder_sequence_output", outputs)
self.assertIn("decoder_sequence_output", outputs)
def test_predict(self):
self.backbone.predict(self.input_batch)
self.backbone.predict(self.input_dataset)

@parameterized.named_parameters(
("jit_compile_false", False), ("jit_compile_true", True)
)
def test_t5_compile_batched_ds(self, jit_compile):
self.model.compile(jit_compile=jit_compile)
outputs = self.model.predict(self.input_dataset)
self.assertIn("encoder_sequence_output", outputs)
self.assertIn("decoder_sequence_output", outputs)
def test_serialization(self):
new_backbone = keras.utils.deserialize_keras_object(
keras.utils.serialize_keras_object(self.backbone)
)
self.assertEqual(new_backbone.get_config(), self.backbone.get_config())

@parameterized.named_parameters(
("tf_format", "tf", "model"),
("keras_format", "keras_v3", "model.keras"),
)
@pytest.mark.large # Saving is slow, so mark these large.
def test_saved_model(self, save_format, filename):
outputs = self.model(self.input_batch)
outputs = self.backbone(self.input_batch)
save_path = os.path.join(self.get_temp_dir(), filename)
self.model.save(save_path, save_format=save_format)
self.backbone.save(save_path, save_format=save_format)
restored_model = keras.models.load_model(save_path)

# Check we got the real object back.
Expand All @@ -119,7 +120,7 @@ def test_saved_model(self, save_format, filename):
class T5BackboneTPUTest(tf.test.TestCase, parameterized.TestCase):
def setUp(self):
with self.tpu_strategy.scope():
self.model = T5Backbone(
self.backbone = T5Backbone(
vocabulary_size=4,
num_layers=2,
num_heads=2,
Expand All @@ -136,6 +137,6 @@ def setUp(self):

def test_predict(self):
self.model.compile()
outputs = self.model.predict(self.input_dataset)
outputs = self.backbone.predict(self.input_dataset)
self.assertIn("encoder_sequence_output", outputs)
self.assertIn("decoder_sequence_output", outputs)
10 changes: 10 additions & 0 deletions keras_nlp/models/t5/t5_tokenizer_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,6 +17,7 @@
import io
import os

import pytest
import sentencepiece
import tensorflow as tf
from absl.testing import parameterized
Expand Down Expand Up @@ -81,10 +82,19 @@ def test_errors_missing_special_tokens(self):
with self.assertRaises(ValueError):
T5Tokenizer(proto=bytes_io.getvalue())

def test_serialization(self):
config = keras.utils.serialize_keras_object(self.tokenizer)
new_tokenizer = keras.utils.deserialize_keras_object(config)
self.assertEqual(
new_tokenizer.get_config(),
self.tokenizer.get_config(),
)

@parameterized.named_parameters(
("tf_format", "tf", "model"),
("keras_format", "keras_v3", "model.keras"),
)
@pytest.mark.large # Saving is slow, so mark these large.
def test_saved_model(self, save_format, filename):
input_data = tf.constant(["the quick brown fox"])

Expand Down