MMClassification Deployment

• MMSegmentation Deployment

  • Installation

    • Install mmcls

    • Install mmdeploy

  • Convert model

  • Model Specification

  • Model inference

    • Backend model inference

    • SDK model inference

  • Supported models

---

MMClassification aka mmcls is an open-source image classification toolbox based on PyTorch. It is a part of the OpenMMLab project.

Installation

Install mmcls

Please follow this quick guide to install mmcls.

Install mmdeploy

There are several methods to install mmdeploy, among which you can choose an appropriate one according to your target platform and device.

Method I: Install precompiled package

TODO. MMDeploy hasn’t released based on 1.x branch.

Method II: Build using scripts

If your target platform is Ubuntu 18.04 or later version, we encourage you to run scripts. For example, the following commands install mmdeploy as well as inference engine - ONNX Runtime.

git clone --recursive -b 1.x https://github.com/open-mmlab/mmdeploy.git
cd mmdeploy
python3 tools/scripts/build_ubuntu_x64_ort.py $(nproc)
export PYTHONPATH=$(pwd)/build/lib:$PYTHONPATH
export LD_LIBRARY_PATH=$(pwd)/../mmdeploy-dep/onnxruntime-linux-x64-1.8.1/lib/:$LD_LIBRARY_PATH

Method III: Build from source

If neither I nor II meets your requirements, building mmdeploy from source is the last option.

Convert model

You can use tools/deploy.py to convert mmcls models to the specified backend models. Its detailed usage can be learned from here.

The command below shows an example about converting resnet18 model to onnx model that can be inferred by ONNX Runtime.

cd mmdeploy

# download resnet18 model from mmcls model zoo
mim download mmcls --config resnet18_8xb32_in1k --dest .

# convert mmcls model to onnxruntime model with dynamic shape
python tools/deploy.py \
    configs/mmcls/classification_onnxruntime_dynamic.py \
    resnet18_8xb32_in1k.py \
    resnet18_8xb32_in1k_20210831-fbbb1da6.pth \
    tests/data/tiger.jpeg \
    --work-dir mmdeploy_models/mmcls/ort \
    --device cpu \
    --show \
    --dump-info

It is crucial to specify the correct deployment config during model conversion. We’ve already provided builtin deployment config files of all supported backends for mmclassification. The config filename pattern is:

classification_{backend}-{precision}_{static | dynamic}_{shape}.py

• {backend}: inference backend, such as onnxruntime, tensorrt, pplnn, ncnn, openvino, coreml and etc.

• {precision}: fp16, int8. When it’s empty, it means fp32

• {static | dynamic}: static shape or dynamic shape

• {shape}: input shape or shape range of a model

Therefore, in the above example, you can also convert resnet18 to other backend models by changing the deployment config file classification_onnxruntime_dynamic.py to others, e.g., converting to tensorrt-fp16 model by classification_tensorrt-fp16_dynamic-224x224-224x224.py.

Tip

When converting mmcls models to tensorrt models, –device should be set to “cuda”

Model Specification

Before moving on to model inference chapter, let’s know more about the converted model structure which is very important for model inference.

The converted model locates in the working directory like mmdeploy_models/mmcls/ort in the previous example. It includes:

mmdeploy_models/mmcls/ort
├── deploy.json
├── detail.json
├── end2end.onnx
└── pipeline.json

in which,

• end2end.onnx: backend model which can be inferred by ONNX Runtime

• *.json: the necessary information for mmdeploy SDK

The whole package mmdeploy_models/mmcls/ort is defined as mmdeploy SDK model, i.e., mmdeploy SDK model includes both backend model and inference meta information.

Model inference

Backend model inference

Take the previous converted end2end.onnx model as an example, you can use the following code to inference the model.

from mmdeploy.apis.utils import build_task_processor
from mmdeploy.utils import get_input_shape, load_config
import torch

deploy_cfg = 'configs/mmcls/classification_onnxruntime_dynamic.py'
model_cfg = './resnet18_8xb32_in1k.py'
device = 'cpu'
backend_model = ['./mmdeploy_models/mmcls/ort/end2end.onnx']
image = 'tests/data/tiger.jpeg'

# read deploy_cfg and model_cfg
deploy_cfg, model_cfg = load_config(deploy_cfg, model_cfg)

# build task and backend model
task_processor = build_task_processor(model_cfg, deploy_cfg, device)
model = task_processor.build_backend_model(backend_model)

# process input image
input_shape = get_input_shape(deploy_cfg)
model_inputs, _ = task_processor.create_input(image, input_shape)

# do model inference
with torch.no_grad():
    result = model.test_step(model_inputs)

# visualize results
task_processor.visualize(
    image=image,
    model=model,
    result=result[0],
    window_name='visualize',
    output_file='output_classification.png')

SDK model inference

You can also perform SDK model inference like following,

from mmdeploy_python import Classifier
import cv2

img = cv2.imread('tests/data/tiger.jpeg')

# create a classifier
classifier = Classifier(model_path='./mmdeploy_models/mmcls/ort', device_name='cpu', device_id=0)
# perform inference
result = classifier(img)
# show inference result
for label_id, score in result:
    print(label_id, score)

Besides python API, mmdeploy SDK also provides other FFI (Foreign Function Interface), such as C, C++, C#, Java and so on. You can learn their usage from demos.

Supported models

Model	TorchScript	ONNX Runtime	TensorRT	ncnn	PPLNN	OpenVINO
ResNet	Y	Y	Y	Y	Y	Y
ResNeXt	Y	Y	Y	Y	Y	Y
SE-ResNet	Y	Y	Y	Y	Y	Y
MobileNetV2	Y	Y	Y	Y	Y	Y
ShuffleNetV1	Y	Y	Y	Y	Y	Y
ShuffleNetV2	Y	Y	Y	Y	Y	Y
VisionTransformer	Y	Y	Y	Y	?	Y
SwinTransformer	Y	Y	Y	N	?	N