Support scikit-learn 0.24.1

ci/buildspec.yml

@@ -2,7 +2,7 @@ version: 0.2
 
 env:
   variables:
-    FRAMEWORK_VERSION: "0.23-1"
+    FRAMEWORK_VERSION: "0.24-1"
 
 phases:
   install:

docker/0.24-1/base/Dockerfile.cpu

@@ -0,0 +1,106 @@
+ARG UBUNTU_VERSION=18.04
+ARG UBUNTU_IMAGE_DIGEST=646942475da61b4ce9cc5b3fadb42642ea90e5d0de46111458e100ff2c7031e6
+
+FROM ubuntu:${UBUNTU_VERSION}@sha256:${UBUNTU_IMAGE_DIGEST}
+
+ARG MINICONDA_VERSION=4.8.3
+ARG CONDA_PY_VERSION=37
+ARG CONDA_PKG_VERSION=4.9.0
+ARG PYTHON_VERSION=3.7.10
+ARG PYARROW_VERSION=0.16.0
+ARG MLIO_VERSION=0.6.0
+
+# Install python and other scikit-learn runtime dependencies
+# Dependency list from http://scikit-learn.org/stable/developers/advanced_installation.html#installing-build-dependencies
+RUN apt-get update && \
+    apt-get -y install --no-install-recommends \
+        build-essential \
+        curl \
+        git \
+        jq \
+        libatlas-base-dev \
+        nginx \
+        openjdk-8-jdk-headless \
+        unzip \
+        wget \
+        && \
+    # MLIO build dependencies
+    # Official Ubuntu APT repositories do not contain an up-to-date version of CMake required to build MLIO.
+    # Kitware contains the latest version of CMake.
+    apt-get -y install --no-install-recommends \
+        apt-transport-https \
+        ca-certificates \
+        gnupg \
+        software-properties-common \
+        && \
+    wget -O - https://apt.kitware.com/keys/kitware-archive-latest.asc 2>/dev/null | \
+        gpg --dearmor - | \
+        tee /etc/apt/trusted.gpg.d/kitware.gpg >/dev/null && \
+    apt-add-repository 'deb https://apt.kitware.com/ubuntu/ bionic main' && \
+    apt-get update && \
+    apt-get install -y --no-install-recommends \
+        autoconf \
+        automake \
+        build-essential \
+        cmake=3.18.4-0kitware1 \
+        cmake-data=3.18.4-0kitware1 \
+        doxygen \
+        kitware-archive-keyring \
+        libcurl4-openssl-dev \
+        libssl-dev \
+        libtool \
+        ninja-build \
+        python3-dev \
+        python3-distutils \
+        python3-pip \
+        zlib1g-dev \
+        && \
+    rm /etc/apt/trusted.gpg.d/kitware.gpg && \
+    rm -rf /var/lib/apt/lists/*
+
+RUN cd /tmp && \
+    curl -L --output /tmp/Miniconda3.sh https://repo.anaconda.com/miniconda/Miniconda3-py${CONDA_PY_VERSION}_${MINICONDA_VERSION}-Linux-x86_64.sh && \
+    bash /tmp/Miniconda3.sh -bfp /miniconda3 && \
+    rm /tmp/Miniconda3.sh
+
+ENV PATH=/miniconda3/bin:${PATH}
+
+# Install MLIO with Apache Arrow integration
+# We could install mlio-py from conda, but it comes  with extra support such as image reader that increases image size
+# which increases training time. We build from source to minimize the image size.
+RUN echo "conda ${CONDA_PKG_VERSION}" >> /miniconda3/conda-meta/pinned && \
+    # Conda configuration see https://conda.io/projects/conda/en/latest/configuration.html
+    conda config --system --set auto_update_conda false && \
+    conda config --system --set show_channel_urls true && \
+    echo "python ${PYTHON_VERSION}.*" >> /miniconda3/conda-meta/pinned && \
+    conda install -c conda-forge python=${PYTHON_VERSION} && \
+    conda install conda=${CONDA_PKG_VERSION} && \
+    conda update -y conda && \
+    conda install -c conda-forge pyarrow=${PYARROW_VERSION} && \
+    cd /tmp && \
+    git clone --branch v${MLIO_VERSION} https://github.com/awslabs/ml-io.git mlio && \
+    cd mlio && \
+    build-tools/build-dependency build/third-party all && \
+    mkdir -p build/release && \
+    cd build/release && \
+    cmake -GNinja -DCMAKE_BUILD_TYPE=RelWithDebInfo -DCMAKE_PREFIX_PATH="$(pwd)/../third-party" ../.. && \
+    cmake --build . && \
+    cmake --build . --target install && \
+    cmake -DMLIO_INCLUDE_PYTHON_EXTENSION=ON -DMLIO_INCLUDE_ARROW_INTEGRATION=ON ../.. && \
+    cmake --build . --target mlio-py && \
+    cmake --build . --target mlio-arrow && \
+    cd ../../src/mlio-py && \
+    python3 setup.py bdist_wheel && \
+    python3 -m pip install dist/*.whl && \
+    cp -r /tmp/mlio/build/third-party/lib/intel64/gcc4.7/* /usr/local/lib/ && \
+    ldconfig && \
+    rm -rf /tmp/mlio
+
+# Python won’t try to write .pyc or .pyo files on the import of source modules
+# Force stdin, stdout and stderr to be totally unbuffered. Good for logging
+ENV PYTHONDONTWRITEBYTECODE=1 PYTHONUNBUFFERED=1 PYTHONIOENCODING=UTF-8 LANG=C.UTF-8 LC_ALL=C.UTF-8
+
+# Install Scikit-Learn
+# Scikit-learn 0.20 was the last version to support Python 2.7 and Python 3.4.
+# Scikit-learn now requires Python 3.6 or newer.
+RUN python -m pip install --no-cache -I scikit-learn==0.24.1

docker/0.24-1/extension/Dockerfile.cpu

@@ -0,0 +1,9 @@
+FROM preprod-sklearn:0.24-1-cpu-py3
+
+RUN pip freeze | grep -q 'scikit-learn==0.24.1'; \
+	if [ $? -eq 0 ]; \
+		then echo 'scikit-learn version 0.24.1 requirement met'; \
+		else echo 'ERROR: Expected scikit-learn version is 0.24.1, check base images for scikit-learn version' && \
+			 exit 1; fi
+
+RUN pip install --upgrade --no-cache --no-deps sagemaker-scikit-learn-extension==2.1.0

docker/0.24-1/extension/README.md

@@ -0,0 +1,56 @@
+# SageMaker Scikit-learn Extension Container
+
+The SageMaker Scikit-learn Extension Container is used in SageMaker Autopilot.
+
+The SageMaker Scikit-learn Extension Container is built in 3 steps. The first 2 steps should be the same as building the [sagemaker-scikit-learn-container](https://github.com/aws/sagemaker-scikit-learn-container) image.
+
+### Step 1: Base Image
+
+The "base" Dockerfile encompass the installation of the framework and all of the dependencies needed.
+
+Tagging scheme is based on <Scikit-learn_version>-<SageMaker_version>-cpu-py<python_version>. (e.g. 0.24-1-cpu-py3)
+
+All "final" Dockerfiles build images using base images that use the tagging scheme above.
+
+```
+docker build -t sklearn-base:0.24-1-cpu-py3 -f docker/0.24-1/base/Dockerfile.cpu .
+```
+
+Notice that this Dockerfile has the updated version of sklearn (0.24.1) installed.
+
+### Step 2: Final Image
+
+The "final" Dockerfiles encompass the installation of the SageMaker specific support code.
+
+All "final" Dockerfiles use base images for building.
+
+These "base" images are specified with the naming convention of sklearn-base:<Scikit-learn_version>-<SageMaker_version>-cpu-py<python_version>.
+
+Before building "final" images:
+
+Build your "base" image. Make sure it is named and tagged in accordance with your "final" Dockerfile.
+
+```
+# Create the SageMaker Scikit-learn Container Python package.
+python setup.py bdist_wheel
+```
+
+Then build the final image, like in the sagemaker-sklearn-container
+
+```
+docker build -t preprod-sklearn:0.24-1-cpu-py3 -f docker/0.24-1/final/Dockerfile.cpu .
+```
+
+### Step 3: Build the extension image for SageMaker Scikit-learn Extension Container
+
+The "extension" Dockerfiles encompass the installation of the SageMaker Autopilot specific support code.
+
+The "extension" Dockerfiles use final images for building.
+
+Build the third additional Dockerfile needed for SageMaker Scikit-learn Extension Container. This Dockerfile specifies a hard dependecy on a certain version of scikit-learn (i.e. v0.24.1).
+
+Tagging scheme is based on extension-<Scikit-learn-Extension_version>-<SageMaker_version>-cpu-py<python_version>. (e.g. extension-0.2-1-cpu-py3). Make sure the "extension" image is tagged in accordance with the  `extension` (i.e. `extension-0.2-1-cpu-py3`).
+
+```
+docker build -t preprod-sklearn-extension:0.2-1-cpu-py3 -f  docker/0.24-1/extension/Dockerfile.cpu .
+```

docker/0.24-1/final/Dockerfile.cpu

@@ -0,0 +1,57 @@
+FROM sklearn-base:0.24-1-cpu-py3
+ENV SAGEMAKER_SKLEARN_VERSION 0.24-1
+
+LABEL com.amazonaws.sagemaker.capabilities.accept-bind-to-port=true
+
+COPY requirements.txt /requirements.txt
+RUN python -m pip install -r /requirements.txt && \
+    rm /requirements.txt
+
+COPY dist/sagemaker_sklearn_container-2.0-py3-none-any.whl /sagemaker_sklearn_container-2.0-py3-none-any.whl
+# https://github.com/googleapis/google-cloud-python/issues/6647
+RUN rm -rf /miniconda3/lib/python3.7/site-packages/numpy-1.19.4.dist-info && \
+    pip install --no-cache /sagemaker_sklearn_container-2.0-py3-none-any.whl && \
+    rm /sagemaker_sklearn_container-2.0-py3-none-any.whl
+
+ENV SAGEMAKER_TRAINING_MODULE sagemaker_sklearn_container.training:main
+ENV SAGEMAKER_SERVING_MODULE sagemaker_sklearn_container.serving:main
+
+#######
+# MMS #
+#######
+# Create MMS user directory
+RUN useradd -m model-server
+RUN mkdir -p /home/model-server/tmp
+RUN chown -R model-server /home/model-server
+
+# Copy MMS configs
+COPY docker/$SAGEMAKER_SKLEARN_VERSION/resources/mms/config.properties.tmp /home/model-server
+ENV SKLEARN_MMS_CONFIG=/home/model-server/config.properties
+
+# Copy execution parameters endpoint plugin for MMS
+RUN mkdir -p /tmp/plugins
+COPY docker/$SAGEMAKER_SKLEARN_VERSION/resources/mms/endpoints-1.0.jar /tmp/plugins
+RUN chmod +x /tmp/plugins/endpoints-1.0.jar
+
+# Create directory for models
+RUN mkdir -p /opt/ml/models
+RUN chmod +rwx /opt/ml/models
+
+#####################
+# Required ENV vars #
+#####################
+# Set SageMaker training environment variables
+ENV SM_INPUT /opt/ml/input
+ENV SM_INPUT_TRAINING_CONFIG_FILE $SM_INPUT/config/hyperparameters.json
+ENV SM_INPUT_DATA_CONFIG_FILE $SM_INPUT/config/inputdataconfig.json
+ENV SM_CHECKPOINT_CONFIG_FILE $SM_INPUT/config/checkpointconfig.json
+
+# Set SageMaker serving environment variables
+ENV SM_MODEL_DIR /opt/ml/model
+
+EXPOSE 8080
+ENV TEMP=/home/model-server/tmp
+
+# Required label for multi-model loading
+LABEL com.amazonaws.sagemaker.capabilities.multi-models=true
+

docker/0.24-1/resources/mms/ExecutionParameters.java

@@ -0,0 +1,98 @@
+package software.amazon.ai.mms.plugins.endpoint;
+
+import com.google.gson.GsonBuilder;
+import com.google.gson.annotations.SerializedName;
+import java.io.IOException;
+import java.nio.charset.StandardCharsets;
+import java.util.Properties;
+import software.amazon.ai.mms.servingsdk.Context;
+import software.amazon.ai.mms.servingsdk.ModelServerEndpoint;
+import software.amazon.ai.mms.servingsdk.annotations.Endpoint;
+import software.amazon.ai.mms.servingsdk.annotations.helpers.EndpointTypes;
+import software.amazon.ai.mms.servingsdk.http.Request;
+import software.amazon.ai.mms.servingsdk.http.Response;
+
+/**
+The modified endpoint source code for the jar used in this container.
+You can create this endpoint by moving it by cloning the MMS repo:
+> git clone https://github.com/awslabs/mxnet-model-server.git
+
+Copy this file into plugins/endpoints/src/main/java/software/amazon/ai/mms/plugins/endpoints/
+and then from the plugins directory, run:
+
+> ./gradlew fJ
+
+Modify file in plugins/endpoint/resources/META-INF/services/* to specify this file location
+
+Then build the JAR:
+
+> ./gradlew build
+
+The jar should be available in plugins/endpoints/build/libs as endpoints-1.0.jar
+**/
+@Endpoint(
+        urlPattern = "execution-parameters",
+        endpointType = EndpointTypes.INFERENCE,
+        description = "Execution parameters endpoint")
+public class ExecutionParameters extends ModelServerEndpoint {
+
+    @Override
+    public void doGet(Request req, Response rsp, Context ctx) throws IOException {
+        Properties prop = ctx.getConfig();
+        // 6 * 1024 * 1024
+        int maxRequestSize = Integer.parseInt(prop.getProperty("max_request_size", "6291456"));
+        SagemakerXgboostResponse response = new SagemakerXgboostResponse();
+        response.setMaxConcurrentTransforms(Integer.parseInt(prop.getProperty("NUM_WORKERS", "1")));
+        response.setBatchStrategy("MULTI_RECORD");
+        response.setMaxPayloadInMB(maxRequestSize / (1024 * 1024));
+        rsp.getOutputStream()
+                .write(
+                        new GsonBuilder()
+                                .setPrettyPrinting()
+                                .create()
+                                .toJson(response)
+                                .getBytes(StandardCharsets.UTF_8));
+    }
+
+    /** Response for Model server endpoint */
+    public static class SagemakerXgboostResponse {
+        @SerializedName("MaxConcurrentTransforms")
+        private int maxConcurrentTransforms;
+
+        @SerializedName("BatchStrategy")
+        private String batchStrategy;
+
+        @SerializedName("MaxPayloadInMB")
+        private int maxPayloadInMB;
+
+        public SagemakerXgboostResponse() {
+            maxConcurrentTransforms = 4;
+            batchStrategy = "MULTI_RECORD";
+            maxPayloadInMB = 6;
+        }
+
+        public int getMaxConcurrentTransforms() {
+            return maxConcurrentTransforms;
+        }
+
+        public String getBatchStrategy() {
+            return batchStrategy;
+        }
+
+        public int getMaxPayloadInMB() {
+            return maxPayloadInMB;
+        }
+
+        public void setMaxConcurrentTransforms(int newMaxConcurrentTransforms) {
+            maxConcurrentTransforms = newMaxConcurrentTransforms;
+        }
+
+        public void setBatchStrategy(String newBatchStrategy) {
+            batchStrategy = newBatchStrategy;
+        }
+
+        public void setMaxPayloadInMB(int newMaxPayloadInMB) {
+            maxPayloadInMB = newMaxPayloadInMB;
+        }
+    }
+}

docker/0.24-1/resources/mms/config.properties.tmp

@@ -0,0 +1,12 @@
+model_store=$$SAGEMAKER_MMS_MODEL_STORE$$
+load_models=$$SAGEMAKER_MMS_LOAD_MODELS$$
+plugins_path=/tmp/plugins
+inference_address=http://0.0.0.0:$$SAGEMAKER_BIND_TO_PORT$$
+management_address=http://0.0.0.0:$$SAGEMAKER_BIND_TO_PORT$$
+default_workers_per_model=$$SAGEMAKER_NUM_MODEL_WORKERS$$
+max_request_size=$$SAGEMAKER_MAX_REQUEST_SIZE$$
+decode_input_request=false
+default_service_handler=$$SAGEMAKER_MMS_DEFAULT_HANDLER$$
+job_queue_size=$$SAGEMAKER_MODEL_JOB_QUEUE_SIZE$$
+preload_model=true
+

requirements.txt

@@ -13,6 +13,6 @@ retrying==1.3.3
 sagemaker-containers==2.8.6.post2
 sagemaker-inference==1.2.0
 sagemaker-training==3.6.2
-scikit-learn==0.23.2
+scikit-learn==0.24.1
 scipy==1.5.3
 six==1.15.0

tox.ini

@@ -6,7 +6,7 @@ max-line-length = 120
 
 [testenv]
 deps =
-    sklearn0.23: scikit-learn==0.23.2
+    sklearn0.24: scikit-learn==0.24.1
     -r{toxinidir}/requirements.txt
     -r{toxinidir}/test-requirements.txt
 conda_deps=