[pull] master from kserve:master

config/overlays/kubeflow/kustomization.yaml

@@ -27,11 +27,6 @@ replacements:
       name: kserve-controller-manager
       version: v1
     targets:
-      - select:
-          kind: CustomResourceDefinition
-          name: inferenceservices.serving.kserve.io
-        fieldPaths:
-          - spec.conversion.webhook.clientConfig.service.namespace
       - select:
           kind: MutatingWebhookConfiguration
           name: inferenceservice.serving.kserve.io

docs/samples/explanation/aif/germancredit/bias.yaml

@@ -9,7 +9,7 @@ spec:
     containerConcurrency: 0
     containers:
     - image: aipipeline/aifserver:predictor
-      name: kfserving-container
+      name: kserve-container
       resources:
         limits:
           cpu: "1"
@@ -61,4 +61,4 @@ spec:
           memory: 2Gi
         requests:
           cpu: "2"
-          memory: 2Gi
+          memory: 2Gi

docs/samples/explanation/aif/germancredit/server/model.py

@@ -1,3 +1,5 @@
+import argparse
+
 import kserve
 from typing import Dict, Union
 
@@ -7,7 +9,7 @@
     load_preproc_data_german,
 )
 
-from kserve import InferRequest, InferResponse
+from kserve import InferRequest, InferResponse, logging
 from kserve.protocol.grpc.grpc_predict_v2_pb2 import (
     ModelInferRequest,
     ModelInferResponse,
@@ -47,7 +49,12 @@ def predict(
         return {"predictions": predictions.tolist()}
 
 
+parser = argparse.ArgumentParser(parents=[kserve.model_server.parser])
+args, _ = parser.parse_known_args()
+
 if __name__ == "__main__":
-    model = KServeSampleModel("german-credit")
+    if args.configure_logging:
+        logging.configure_logging(args.log_config_file)
+    model = KServeSampleModel(args.model_name)
     model.load()
     kserve.ModelServer(workers=1).start([model])

docs/samples/explanation/aix/mnist/rfserver/rfserver/__main__.py

@@ -14,6 +14,7 @@
 import kserve
 import argparse
 
+from kserve import logging
 from .model import RFModel
 
 DEFAULT_MODEL_NAME = "rfserver"
@@ -27,6 +28,8 @@
 args, _ = parser.parse_known_args()
 
 if __name__ == "__main__":
+    if args.configure_logging:
+        logging.configure_logging(args.log_config_file)
     model = RFModel(args.model_name)
     model.load()
     kserve.ModelServer().start([model])

docs/samples/explanation/aix/mnist/rfserver/rfserver/model.py

@@ -11,7 +11,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import logging
 from typing import Dict, Union
 import pickle
 
@@ -23,6 +22,7 @@
     ModelInferRequest,
     ModelInferResponse,
 )
+from kserve.logging import logger
 
 
 class PipeStep(object):
@@ -64,7 +64,7 @@ def predict(
 
         try:
             inputs = np.asarray(instances)
-            logging.info("Calling predict on image of shape %s", (inputs.shape,))
+            logger.info("Calling predict on image of shape %s", (inputs.shape,))
         except Exception as e:
             raise Exception(
                 "Failed to initialize NumPy array from inputs: %s, %s" % (e, instances)

docs/samples/explanation/alibi/alibiexplainer/alibiexplainer/__main__.py

@@ -11,8 +11,7 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-
-import logging
+import argparse
 import os
 import sys
 
@@ -22,12 +21,15 @@
 from alibiexplainer.parser import parse_args
 
 import kserve
+from kserve import logging
 from kserve.storage import Storage
-
-logging.basicConfig(level=kserve.constants.KSERVE_LOGLEVEL)
+from kserve.logging import logger
 
 EXPLAINER_FILENAME = "explainer.dill"
 
+parser = argparse.ArgumentParser(parents=[kserve.model_server.parser])
+args, _ = parser.parse_known_args()
+
 
 def main():
     args, extra = parse_args(sys.argv[1:])
@@ -39,7 +41,7 @@ def main():
             Storage.download(args.storage_uri), EXPLAINER_FILENAME
         )
         with open(alibi_model, "rb") as f:
-            logging.info("Loading Alibi model")
+            logger.info("Loading Alibi model")
             alibi_model = dill.load(f)
 
     explainer = AlibiExplainer(
@@ -54,4 +56,6 @@ def main():
 
 
 if __name__ == "__main__":
+    if args.configure_logging:
+        logging.configure_logging(args.log_config_file)
     main()

docs/samples/explanation/alibi/alibiexplainer/alibiexplainer/anchor_images.py

@@ -11,16 +11,15 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-import kserve
-import logging
+
 import numpy as np
 import alibi
 from alibi.api.interfaces import Explanation
 from alibi.utils.wrappers import ArgmaxTransformer
 from alibiexplainer.explainer_wrapper import ExplainerWrapper
 from typing import Callable, List, Optional, Dict
 
-logging.basicConfig(level=kserve.constants.KSERVE_LOGLEVEL)
+from kserve.logging import logger
 
 
 class AnchorImages(ExplainerWrapper):
@@ -44,7 +43,7 @@ def explain(self, inputs: List, headers: Dict[str, str] = None) -> Explanation:
             self.anchors_image.predictor = self.predict_fn
         else:
             self.anchors_image.predictor = ArgmaxTransformer(self.predict_fn)
-        logging.info("Calling explain on image of shape %s", (arr.shape,))
-        logging.info("anchor image call with %s", self.kwargs)
+        logger.info("Calling explain on image of shape %s", (arr.shape,))
+        logger.info("anchor image call with %s", self.kwargs)
         anchor_exp = self.anchors_image.explain(arr[0], **self.kwargs)
         return anchor_exp

docs/samples/explanation/alibi/alibiexplainer/alibiexplainer/anchor_tabular.py

@@ -11,16 +11,15 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-import kserve
-import logging
+
 import numpy as np
 import alibi
 from alibi.api.interfaces import Explanation
 from alibi.utils.wrappers import ArgmaxTransformer
 from alibiexplainer.explainer_wrapper import ExplainerWrapper
 from typing import Callable, List, Optional, Dict
 
-logging.basicConfig(level=kserve.constants.KSERVE_LOGLEVEL)
+from kserve.logging import logger
 
 
 class AnchorTabular(ExplainerWrapper):
@@ -41,7 +40,7 @@ def explain(self, inputs: List, headers: Dict[str, str] = None) -> Explanation:
         arr = np.array(inputs)
         # set anchor_tabular predict function so it always returns predicted class
         # See anchor_tabular.__init__
-        logging.info("Arr shape %s ", (arr.shape,))
+        logger.info("Arr shape %s ", (arr.shape,))
 
         # check if predictor returns predicted class or prediction probabilities for each class
         # if needed adjust predictor so it returns the predicted class

docs/samples/explanation/alibi/alibiexplainer/alibiexplainer/anchor_text.py

@@ -11,8 +11,7 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-import kserve
-import logging
+
 import numpy as np
 import spacy
 import alibi
@@ -22,7 +21,7 @@
 from alibiexplainer.explainer_wrapper import ExplainerWrapper
 from typing import Callable, List, Optional, Dict
 
-logging.basicConfig(level=kserve.constants.KSERVE_LOGLEVEL)
+from kserve.logging import logger
 
 
 class AnchorText(ExplainerWrapper):
@@ -35,12 +34,12 @@ def __init__(
     ):
         self.predict_fn = predict_fn
         self.kwargs = kwargs
-        logging.info("Anchor Text args %s", self.kwargs)
+        logger.info("Anchor Text args %s", self.kwargs)
         if explainer is None:
-            logging.info("Loading Spacy Language model for %s", spacy_language_model)
+            logger.info("Loading Spacy Language model for %s", spacy_language_model)
             spacy_model(model=spacy_language_model)
             self.nlp = spacy.load(spacy_language_model)
-            logging.info("Language model loaded")
+            logger.info("Language model loaded")
         self.anchors_text = explainer
 
     def explain(self, inputs: List, headers: Dict[str, str] = None) -> Explanation:

docs/samples/explanation/alibi/alibiexplainer/alibiexplainer/explainer.py

@@ -11,25 +11,25 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
+
 import json
-import logging
 import asyncio
 from enum import Enum
 from typing import List, Any, Mapping, Union, Dict
 
-import kserve
 import numpy as np
+import nest_asyncio
+
 from alibiexplainer.anchor_images import AnchorImages
 from alibiexplainer.anchor_tabular import AnchorTabular
 from alibiexplainer.anchor_text import AnchorText
 from alibiexplainer.explainer_wrapper import ExplainerWrapper
 
-import nest_asyncio
+import kserve
+from kserve.logging import logger
 
 nest_asyncio.apply()
 
-logging.basicConfig(level=kserve.constants.KSERVE_LOGLEVEL)
-
 
 class ExplainerMethod(Enum):
     anchor_tabular = "AnchorTabular"
@@ -51,7 +51,7 @@ def __init__(  # pylint:disable=too-many-arguments
     ):
         super().__init__(name)
         self.predictor_host = predictor_host
-        logging.info("Predict URL set to %s", self.predictor_host)
+        logger.info("Predict URL set to %s", self.predictor_host)
         self.method = method
 
         if self.method is ExplainerMethod.anchor_tabular:
@@ -84,7 +84,7 @@ def explain(self, payload: Dict, headers: Dict[str, str] = None) -> Any:
         ):
             explanation = self.wrapper.explain(payload["instances"])
             explanationAsJsonStr = explanation.to_json()
-            logging.info("Explanation: %s", explanationAsJsonStr)
+            logger.info("Explanation: %s", explanationAsJsonStr)
             return json.loads(explanationAsJsonStr)
 
         raise NotImplementedError

docs/samples/explanation/alibi/alibiexplainer/alibiexplainer/parser.py

@@ -14,17 +14,16 @@
 
 import argparse
 import kserve
-import logging
 import os
-from alibiexplainer.explainer import ExplainerMethod  # pylint:disable=no-name-in-module
+from alibiexplainer.explainer import ExplainerMethod
 
-logging.basicConfig(level=kserve.constants.KSERVE_LOGLEVEL)
+from kserve.logging import logger
 
 DEFAULT_EXPLAINER_NAME = "explainer"
 ENV_STORAGE_URI = "STORAGE_URI"
 
 
-class GroupedAction(argparse.Action):  # pylint:disable=too-few-public-methods
+class GroupedAction(argparse.Action):
     def __call__(self, theparser, namespace, values, option_string=None):
         group, dest = self.dest.split(".", 2)
         groupspace = getattr(namespace, group, argparse.Namespace())
@@ -213,5 +212,5 @@ def parse_args(sys_args):
         extra = vars(args.explainer)
     else:
         extra = {}
-    logging.info("Extra args: %s", extra)
+    logger.info("Extra args: %s", extra)
     return args, extra

docs/samples/kafka/image_transformer/__main__.py

@@ -13,6 +13,8 @@
 
 import kserve
 import argparse
+
+from kserve import logging
 from .image_transformer import ImageTransformer
 
 DEFAULT_MODEL_NAME = "model"
@@ -30,6 +32,8 @@
 args, _ = parser.parse_known_args()
 
 if __name__ == "__main__":
+    if args.configure_logging:
+        logging.configure_logging(args.log_config_file)
     transformer = ImageTransformer(args.model_name, predictor_host=args.predictor_host)
     server = kserve.ModelServer()
     server.start(models=[transformer])

docs/samples/kafka/image_transformer/image_transformer.py

@@ -11,7 +11,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import logging
 from typing import Dict, Union
 
 import boto3
@@ -20,8 +19,7 @@
 import kserve
 from kserve import InferRequest, InferResponse
 from kserve.protocol.grpc.grpc_predict_v2_pb2 import ModelInferResponse
-
-logging.basicConfig(level=kserve.constants.KSERVE_LOGLEVEL)
+from kserve.logging import logger
 
 session = boto3.Session()
 client = session.client(
@@ -49,7 +47,7 @@ def __init__(self, name: str, predictor_host: str):
     async def preprocess(
         self, inputs: Union[Dict, InferRequest], headers: Dict[str, str] = None
     ) -> Union[Dict, InferRequest]:
-        logging.info("Received inputs %s", inputs)
+        logger.info("Received inputs %s", inputs)
         if inputs["EventName"] == "s3:ObjectCreated:Put":
             bucket = inputs["Records"][0]["s3"]["bucket"]["name"]
             key = inputs["Records"][0]["s3"]["object"]["key"]
@@ -64,10 +62,10 @@ async def postprocess(
         response: Union[Dict, InferResponse, ModelInferResponse],
         headers: Dict[str, str] = None,
     ) -> Union[Dict, ModelInferResponse]:
-        logging.info("response: %s", response)
+        logger.info("response: %s", response)
         index = response["predictions"][0]["classes"]
-        logging.info("digit:" + str(index))
+        logger.info("digit:" + str(index))
         upload_path = f"digit-{index}/{self._key}"
         client.upload_file("/tmp/" + self._key, digits_bucket, upload_path)
-        logging.info(f"Image {self._key} successfully uploaded to {upload_path}")
+        logger.info(f"Image {self._key} successfully uploaded to {upload_path}")
         return response