Adlik / Adlik
Projects that are alternatives of or similar to Adlik
Adlik
Adlik [ædlik] is an end-to-end optimizing framework for deep learning models. The goal of Adlik is to accelerate deep learning inference process both on cloud and embedded environment.
With Adlik framework, different deep learning models can be deployed to different platforms with high performance in a much flexible and easy way.
-
In cloud environment, the compiled model and Adlik Inference Engine should be built as a docker image, and deployed as a container.
-
In edge environment, Adlik Inference Engine should be deployed as a container. The compiled model should be transferred to edge environment, and the Adlik Inference Engine should automatically update and load model.
-
In device environment, Adlik Inference Engine and the compiled model should be compiled into a binary file (so or lib). Users who want to run model inference on device should link user defined AI function and Adlik binary file to the execution file, and run directly.
Inference performance of Adlik
We test the inference performance of Adlik on the same CPU or GPU using the simple CNN model (MNIST model), the ResNet50 model, and InceptionV3 with different serving engines. The test performance data of Adlik on different models are as follows:
- The test result of the MNIST model
- The test result of the ResNet50 model
- The test result of the InceptionV3 model
- The test result of the YoloV3 model
- The test result of the Bert model
Contents
Model Optimizer
Model optimizer focuses on specific hardware and runs on it to achieve acceleration. The proposed framework mainly consists of two categories of algorithm components, i.e. pruner and quantizer.
Model Compiler
Model compiler supports several optimizing technologies like pruning, quantization and structural compression, which can be easily used for models developed with TensorFlow, Keras, PyTorch, etc.
Serving Engine
Serving Engine provides deep learning models with optimized runtime based on the deployment environment. Put simply, based on a deep learning model, the users of Adlik can optimize it with model compiler and then deploy it to a certain platform with Adlik serving platform.
Docker images
All Adlik compiler images and serving images are stored in Alibaba Cloud. These images can be downloaded and used directly, users do not need to build the Adlik on Ubuntu. Users can use the compiler images to compile model from H5, CheckPoint, FrozenGraph, ONNX and SavedModel to Openvino, TensorFlow, TensorFlow Lite, TensorRT. Users also can use the serving images for model inference.
Docker pull command:
docker pull docker_image_name:tag
Compiler docker images
The compiler docker images can be used in CPU and GPU. In the CPU, you can compile the model from source type to TensorFlow model, OpenVino model and TensorFlow Lite model. And in the CPU, you can compile the model from source type to TensorFlow model, and TensorRT model. The names and labels of compiler mirrors are as follows, and the first half of label represents the version of TensorRT, the latter part of label represents the version of CUDA:
registry.cn-beijing.aliyuncs.com/adlik/model-compiler:7.2.1.6_11.0
registry.cn-beijing.aliyuncs.com/adlik/model-compiler:7.2.1.6_10.2
registry.cn-beijing.aliyuncs.com/adlik/model-compiler:7.2.0.11_11.0
registry.cn-beijing.aliyuncs.com/adlik/model-compiler:7.1.3.4_11.0
registry.cn-beijing.aliyuncs.com/adlik/model-compiler:7.1.3.4_10.2
registry.cn-beijing.aliyuncs.com/adlik/model-compiler:7.0.0.11_10.2
registry.cn-beijing.aliyuncs.com/adlik/model-compiler:7.0.0.11_10.0
Using model compiler image compile model
-
Run the image.
docker run -it --rm -v source_model:/home/john/model registry.cn-beijing.aliyuncs.com/adlik/model-compiler:latest bash
-
Configure the json file or environment variables required to compile the model.
The json_field.json describle the json file field information, and for the example, you can reference compiler_json_example.json. For the environment variable field description, see env_field.txt, for the example, reference compiler_env_example.txt.
Note: The checkpoint model must be given the input and output op names of the model when compiling, and other models can be compiled without the input and output op names of the model.
-
Compile the model.
Compilation instructions (json file mode):
python3 "-c" "import json; import model_compiler as compiler; file=open('/mnt/model/serving_model.json','r'); request = json.load(file);compiler.compile_model(request);file.close()"
Compilation instructions (environment variable mode):
python3 "-c" "import model_compiler.compiler as compiler;compiler.compile_from_env()"
Serving docker imaegs
The serving docker imaegs contains CPU and GPU mirrors. The label of openvino image represents the version of OpenVINO. And for the TensorRT image the first half of label represents the version of TensorRT, the latter part of label represents the version of CUDA. The names and labels of serving mirrors are as follows:
CPU:
registry.cn-beijing.aliyuncs.com/adlik/serving/tflite-cpu:latest
registry.cn-beijing.aliyuncs.com/adlik/serving/tensorflow-cpu:latest
registry.cn-beijing.aliyuncs.com/adlik/serving/openvino:2021.1.110
GPU:
registry.cn-beijing.aliyuncs.com/adlik/serving/tensorflow-gpu:latest
registry.cn-beijing.aliyuncs.com/adlik/serving/tensorrt:7.2.1.6_11.0
registry.cn-beijing.aliyuncs.com/adlik/serving/tensorrt:7.2.1.6_10.2
registry.cn-beijing.aliyuncs.com/adlik/serving/tensorrt:7.2.0.11_11.0
registry.cn-beijing.aliyuncs.com/adlik/serving/tensorrt:7.1.3.4_11.0
registry.cn-beijing.aliyuncs.com/adlik/serving/tensorrt:7.1.3.4_10.2
registry.cn-beijing.aliyuncs.com/adlik/serving/tensorrt:7.0.0.11_10.2
registry.cn-beijing.aliyuncs.com/adlik/serving/tensorrt:7.0.0.11_10.0
Using the serving images for model inference
-
Run the mirror and pay attention to mapping out the service port. example command line:
docker run -it --rm -p 8500:8500 -v compiled_model:/model adlik/serving-openvino:latest bash
-
Load the compiled model in the image and start the service.
example command line:
adlik-serving --grpc_port=8500 --http_port=8501 --model_base_path=/model
-
Install the client wheel package adlik serving package or adlik serving gpu package locally, execute the inference code, and perform inference.
Note: If you not mapping out the service port when you run the mirror, you need install the adlik serving package or adlik serving gpu package in the container. Then execute the inference code, and perform inference in the container.
Build
This guide is for building Adlik on Ubuntu systems.
Then, clone Adlik and change the working directory into the source directory:
git clone https://github.com/ZTE/Adlik.git
cd Adlik
Build clients
-
Install the following packages:
python3-setuptools
python3-wheel
-
Build clients:
bazel build //adlik_serving/clients/python:build_pip_package -c opt
-
Build pip package:
mkdir /tmp/pip-packages && bazel-bin/adlik_serving/clients/python/build_pip_package /tmp/pip-packages
Build serving
First, install the following packages:
automake
libtbb2
libtool
make
python3-six
Build serving with OpenVINO runtime
-
Install
intel-openvino-runtime-ubuntu<OS_VERSION>-<VERSION>
package from OpenVINO. -
Assume the installation path of OpenVINO is
/opt/intel/openvino_VERSION
, run the following command:export INTEL_CVSDK_DIR=/opt/intel/openvino_VERSION export InferenceEngine_DIR=$INTEL_CVSDK_DIR/deployment_tools/inference_engine/share bazel build //adlik_serving \ --config=openvino \ -c opt
Build serving with TensorFlow CPU runtime
Run the following command:
bazel build //adlik_serving \
--config=tensorflow-cpu \
-c opt
Build serving with TensorFlow GPU runtime
Assume building with CUDA version 11.0.
-
Install the following packages from here and here:
cuda-cupti-dev-11-0
libcublas-dev-11-0
libcudnn8=*+cuda11.0
libcudnn8-dev=*+cuda11.0
libcufft-dev-11-0
libcurand-dev-11-0
libcusolver-dev-11-0
libcusparse-dev-11-0
-
Run the following command:
env TF_CUDA_VERSION=11.0 \ bazel build //adlik_serving \ --config=tensorflow-gpu \ -c opt \ --incompatible_use_specific_tool_files=false
Build serving with TensorFlow Lite CPU runtime
Run the following command:
bazel build //adlik_serving \
--config=tensorflow-lite-cpu \
-c opt
Build serving with TensorRT runtime
Assume building with CUDA version 11.0.
-
Install the following packages from here and here:
cuda-cupti-dev-11-0
cuda-nvml-dev-11-0
cuda-nvrtc-11-0
libcublas-dev-11-0
libcudnn8=*+cuda11.0
libcudnn8-dev=*+cuda11.0
libcufft-dev-11-0
libcurand-dev-11-0
libcusolver-dev-11-0
libcusparse-dev-11-0
libnvinfer7=7.2.*+cuda11.0
libnvinfer-dev=7.2.*+cuda11.0
libnvonnxparsers7=7.2.*+cuda11.0
libnvonnxparsers-dev=7.2.*+cuda11.0
-
Run the following command:
env TF_CUDA_VERSION=11.0 \ bazel build //adlik_serving \ --config=TensorRT \ -c opt \ --action_env=LIBRARY_PATH=/usr/local/cuda-11.0/lib64/stubs \ --incompatible_use_specific_tool_files=false
Build in Docker
The ci/docker/build.sh
file can be used to build a Docker images that contains all the requirements for building
Adlik. You can build Adlik with the Docker image.
Note: If you build the runtime with GPU in a Docker image, you need to add the CUDA environment variables in the Dockerfile, such as:
ENV NVIDIA_VISIBLE_DEVICES all ENV NVIDIA_DRIVER_CAPABILITIES compute, utility
Getting Started
Release
The version of the service engine Adlik supports.
TensorFlow 1.14 | TensorFlow 2.x | OpenVINO 2021 | TensorRT 6 | TensorRT 7 | |
---|---|---|---|---|---|
Keras | ✓ | ✓ | ✓ | ✓ | ✓ |
TensorFlow | ✓ | ✓ | ✓ | ✓ | ✓ |
PyTorch | ✗ | ✗ | ✓ | ✓ | ✓ |
License
Apache License 2.0