feat: upgrade pythone version, drop 3.10 and add VERBOSE env variable

.pre-commit-config.yaml

@@ -23,4 +23,4 @@ repos:
     rev: v3.3.1
     hooks:
     -   id: pyupgrade
-        args: ["--py310-plus"]
+        args: ["--py311-plus"]

Dockerfile

@@ -1,4 +1,4 @@
-FROM python:3.10-slim
+FROM python:3.11-slim
 WORKDIR /app
 
 RUN apt update && apt install -y libgomp1 && rm -rf /var/lib/apt/lists/*

VERSION

@@ -1 +1 @@
-3.3.0
+3.4.0

setup.cfg

@@ -12,7 +12,7 @@ max-complexity = 10
 per-file-ignores=__init__.py:F401
 
 [mypy]
-python_version = 3.10
+python_version = 3.11
 ignore_missing_imports = True
 files = src/
 plugins = numpy.typing.mypy_plugin

setup.py

@@ -40,7 +40,7 @@
     package_dir={"": "src"},
     include_package_data=True,
     zip_safe=True,
-    python_requires=">=3.10",
+    python_requires=">=3.11",
     install_requires=[
         "numpy~=1.23.0",
         "grpcio~=1.51",
@@ -60,7 +60,7 @@
         ]
     },
     classifiers=[
-        "Programming Language :: Python :: 3.10",
+        "Programming Language :: Python :: 3.11",
         "Programming Language :: Python :: 3 :: Only",
         "License :: OSI Approved :: GNU Lesser General Public License v3 or later (LGPLv3+)",
     ],

src/pythie_serving/abstract_wrapper.py

@@ -1,5 +1,6 @@
 import abc
 import logging
+import os
 import time
 from pathlib import Path
 from typing import Any, ClassVar, TypedDict
@@ -33,6 +34,7 @@ class AbstractPythieServingPredictionServiceServicer(prediction_service_pb2_grpc
 
     def __init__(self, *, model_server_config: ModelServerConfig):
         self.logger = logging.getLogger("pythie_serving")
+        self.verbose = bool(os.environ.get("VERBOSE", True))
 
         self.model_map = {}
         for model_config in model_server_config.model_config_list.config:
@@ -78,11 +80,12 @@ def Predict(self, request: predict_pb2.PredictRequest, context: grpc.RpcContext)
             self.logger.error(f"Failed to serve because: {e}")
             raise
         else:
-            duration = time.time() - start
-            self.logger.info(
-                f"Served model {request.model_spec.name}/{request.model_spec.signature_name}: "
-                f"{len(pred)} predictions in {duration:.2f} seconds ({len(pred) / duration:.2f} pred/sec) "
-            )
+            if self.verbose:
+                duration = time.time() - start
+                self.logger.info(
+                    f"Served model {request.model_spec.name}/{request.model_spec.signature_name}: "
+                    f"{len(pred)} predictions in {duration:.2f} seconds ({len(pred) / duration:.2f} pred/sec) "
+                )
             return predict_response
 
     @staticmethod
@@ -102,7 +105,6 @@ def _parse_samples(request: predict_pb2.PredictRequest, model_specs: ModelSpecs)
         samples = np.empty((nb_samples, nb_features), model_specs["samples_dtype"])
 
         for i, feature_name in enumerate(features_names):
-
             if request_inputs[feature_name].tensor_shape.dim[0].size != nb_samples:
                 raise PythieServingException(f"{feature_name} has invalid length.")