Upgrade tensorflow to 2.x (#41)

Author: xuhdev

.travis.yml

@@ -1,6 +1,6 @@
 language: python
 python:
-  - 3.6
+  - 3.8
 services:
   - docker
 install:

Dockerfile

@@ -14,14 +14,14 @@
 # limitations under the License.
 #
 
-FROM quay.io/codait/max-base:v1.4.0
+FROM quay.io/codait/max-base:v1.5.1
 
 COPY requirements.txt .
 
 RUN pip install -r requirements.txt
- 
+
 COPY . .
-    
+
 EXPOSE 5000
 
 CMD python app.py

api/predict.py

@@ -16,7 +16,7 @@
 
 from core.model import ModelWrapper
 from maxfw.core import MAX_API, PredictAPI
-from flask_restplus import fields
+from flask_restx import fields
 from werkzeug.datastructures import FileStorage
 from config import DEFAULT_MODEL, MODELS
 

core/model.py

@@ -20,7 +20,7 @@
 from config import DEFAULT_MODEL_PATH, MODELS, MODEL_META_DATA as model_meta
 from keras.models import load_model
 import numpy as np
-from sklearn.externals import joblib
+import joblib
 
 logging.basicConfig()
 logger = logging.getLogger()
@@ -34,12 +34,19 @@ def load_array(input_data):
 class SingleModelWrapper(object):
 
     def __init__(self, model, path):
+        # The code was originally written for TF1 and doesn't work with eager mode.
+        tf.compat.v1.disable_eager_execution()
+        self.session = tf.compat.v1.Session()
+
         self.model_name = model
 
         # load model
         model_path = '{}/{}_model'.format(path, model)
         logger.info(model_path)
-        self.graph = tf.get_default_graph()
+        self.graph = tf.compat.v1.get_default_graph()
+        # See https://github.com/tensorflow/tensorflow/issues/28287#issuecomment-495005162
+        # We have to do this because we load 3 models in the process.
+        tf.compat.v1.keras.backend.set_session(self.session)
         self.model = load_model(model_path)
 
         # load scaler
@@ -96,6 +103,7 @@ def _rescale_preds(self, preds):
     def predict(self, x):
         reshaped_x = self._reshape_data(x)
         with self.graph.as_default():
+            tf.compat.v1.keras.backend.set_session(self.session)
             preds = self.model.predict(reshaped_x)
         rescaled_preds = self._rescale_preds(preds)
         return rescaled_preds

requirements.txt

@@ -1,5 +1,4 @@
-numpy==1.16.2
-tensorflow==1.15.2
-keras==2.2.4
-scikit-learn==0.22.1
-h5py==2.9.0
+tensorflow==2.6.0
+keras==2.6.0
+scikit-learn==0.22.2
+# numpy and h5py not specified here because tensorflow has specifies a major.minor version range for them