Newer
Older
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
import os
import pickle
import logging
from flask import Flask, Response, request
from apache_beam.io.filesystems import FileSystems
LOGGER = logging.getLogger(__name__)
def get_model_path():
return os.environ['AUTOCUT_MODEL_PATH']
def load_model(file_path):
with FileSystems.open(file_path) as fp:
return pickle.load(fp)
def create_app():
app = Flask(__name__)
model = load_model(get_model_path())
LOGGER.debug('loaded model: %s', model)
@app.route('/api/autocut', methods=['GET', 'POST'])
def _autocut():
if request.method == 'POST':
value = request.get_data()
else:
value = request.args.get('value')
LOGGER.debug('value: %s', value)
output_value = model.predict([value])[0]
return Response(output_value)
return app
def main():
debug_enabled = False
if os.environ.get('AUTOCUT_DEBUG') == '1':
logging.root.setLevel('DEBUG')
debug_enabled = True
create_app().run(debug=debug_enabled)
if __name__ == "__main__":
logging.basicConfig(level='INFO')
main()