def test_missing_labels_file(self, fs: FakeFilesystem) -> None:
fs.add_real_directory('./testdata/test_model/test_instance')
fs.remove_object('./testdata/test_model/test_instance/label.vocab')
with pytest.raises(
Exception,
match=(r'Failure to load labels file from {0} with exception').
format('./testdata/test_model/test_instance/label.vocab')):
Classifier(self.BASE_CLASSIFIER_PATH, 'test_model')
def test_invalid_bert(self, fs: FakeFilesystem) -> None:
bad_bert_path = './bad/path/to/bert'
config = """
{
"bert": "%s",
"labels": "label.vocab",
"is_released": true,
"description": "This is the latest model from Sascha.",
"metadata": {
"thesaurus": "issues"
}
}
""" % (bad_bert_path)
fs.add_real_directory('./testdata/test_model/test_instance')
fs.remove_object('./testdata/test_model/test_instance/config.json')
fs.create_file('./testdata/test_model/test_instance/config.json',
contents=config)
with pytest.raises(Exception,
match='SavedModel file does not exist at'):
c = Classifier(self.BASE_CLASSIFIER_PATH, 'test_model')
# Bad bert is only used on uncached embed.
c.classify(['some string'])
def test_e2e(app: Flask, fs: FakeFilesystem) -> None:
seq_pattern = (
'react more swiftly to comply with international instruments %d')
instance_path = './testdata/test_model/test_instance'
fs.add_real_directory(instance_path)
fs.remove_object('./testdata/test_model/test_instance/quality.json')
fs.remove_object('./testdata/test_model/test_instance/thresholds.json')
client = app.test_client()
with app.test_request_context():
# initial classify returns no topics because we deleted thresholds.json.
resp = client.post('/classify?model=test_model',
data=json.dumps({'seqs': [seq_pattern % 1]}),
content_type='application/json')
assert resp.status == '200 OK'
assert len(json.loads(resp.data)[0]) == 0
# now we add training labels
assert client.put(
'/classification_sample?model=test_model',
data=json.dumps({
'samples': [{
'seq': seq_pattern % i,
'training_labels': [{
'topic': 'International instruments'
}]
} for i in range(20)]
}),
content_type='application/json').status == '200 OK'
resp = client.get('/classification_sample?model=test_model&seq=*')
assert resp.status == '200 OK'
assert len(json.loads(resp.data)) == 20
assert client.post('/task',
data=json.dumps({
'provider': 'RefreshThresholds',
'name': 'thres',
'model': 'test_model'
}),
content_type='application/json').status == '200 OK'
wait_for_task(client, 'thres')
for i in range(20):
resp = client.post('/classify?model=test_model',
data=json.dumps({'seqs': [seq_pattern % i]}),
content_type='application/json')
assert resp.status == '200 OK'
assert len(json.loads(resp.data)[0]) == 1
assert client.post('/task',
data=json.dumps({
'provider': 'RefreshPredictions',
'name': 'pred',
'model': 'test_model'
}),
content_type='application/json').status == '200 OK'
wait_for_task(client, 'pred')
resp = client.get('/classification_sample?model=test_model&seq=*')
assert resp.status == '200 OK'
data = json.loads(resp.data)
assert len(data) == 20
assert data[0]['predicted_labels'][0][
'topic'] == 'International instruments'
assert data[0]['predicted_labels'][0]['quality'] >= 0.5
