[PyTorch] Added support of inference (#361)

* [PyTorch] Added support of inference

README.md

@@ -23,6 +23,7 @@ DLI supports inference using the following frameworks:
 - [ONNX Runtime][onnx-runtime].
 - [MXNet][mxnet].
 - [OpenCV DNN][opencv-dnn] (C++ and Python API).
+- [PyTorch][pytorch].
 
 More information about DLI is available on the web-site
 ([here][dli-ru-web-page] (in Russian)
@@ -99,6 +100,9 @@ Novgorod State University Publishing House, 2021. – 423 p.
     - [`validation_results_mxnet_gluon_modelzoo.md`](results/validation/validation_results_mxnet_gluon_modelzoo.md)
       is a table that confirms correctness of inference implementation
       based on MXNet for [GluonCV-models][gluoncv-omz].
+    - [`validation_results_pytorch.md`](results/validation/validation_results_pytorch.md)
+      is a table that confirms correctness of inference implementation
+      based on PyTorch for [TorchVision][torchvision].
 
   - [`mxnet_models_checklist.md`](results/mxnet_models_checklist.md) contains a list
     of deep models inferred by MXNet checked in the DLI benchmark.
@@ -108,6 +112,8 @@ Novgorod State University Publishing House, 2021. – 423 p.
     of deep models inferred by TensorFlow checked in the DLI benchmark.
   - [`tflite_models_checklist.md`](results/tflite_models_checklist.md) contains a list
     of deep models inferred by TensorFlow Lite checked in the DLI benchmark.
+  - [`pytorch_models_checklist.md`](results/pytorch_models_checklist.md) contains a list
+    of deep models inferred by PyTorch checked in the DLI benchmark.
 
 - `src` directory contains benchmark sources.
 
@@ -184,11 +190,13 @@ Report questions, issues and suggestions, using:
 [onnx-runtime-github]: https://github.com/microsoft/onnxruntime
 [mxnet]: https://mxnet.apache.org
 [opencv-dnn]: https://docs.opencv.org/4.7.0/d2/d58/tutorial_table_of_content_dnn.html
+[pytorch]: https://pytorch.org
 [benchmark-app]: https://github.com/openvinotoolkit/openvino/tree/master/samples/cpp/benchmark_app
 [dli-ru-web-page]: http://hpc-education.unn.ru/dli-ru
 [dli-web-page]: http://hpc-education.unn.ru/dli
 [open-model-zoo]: https://github.com/opencv/open_model_zoo
 [gluoncv-omz]: https://cv.gluon.ai/model_zoo/index.html
+[torchvision]: https://pytorch.org/vision/stable/models.html
 [mmst-2021]: https://hpc-education.unn.ru/files/conference_hpc/2021/MMST2021_Proceedings.pdf
 [dli-wiki]: https://github.com/itlab-vision/dl-benchmark/wiki
 [dli-wiki-build]: https://github.com/itlab-vision/dl-benchmark/wiki#how-to-build

requirements_frameworks.txt

@@ -1,2 +1,3 @@
-openvino-dev[caffe,mxnet,tensorflow2]==2022.3.0
-gluoncv
+openvino-dev[caffe,mxnet,tensorflow2,pytorch]==2022.3.0
+gluoncv
+torchvision

results/pytorch_models_checklist.md

@@ -0,0 +1,88 @@
+# Model validation and performance analysis status for PyTorch
+
+## Public models
+
+The list of models is from [TorchVision][torchvision].
+
+### Image classification on ImageNet
+
+Model | Availability in [TorchVision][torchvision] (0.15.1)| Availability in the validation table |
+-|-|-|
+alexnet|+|+|
+densenet121|+|+|
+densenet161|+|+|
+densenet169|+|+|
+densenet201|+|+|
+googlenet|+|+|
+inception_v3|+|+|
+mnasnet0_5|+|+|
+mnasnet0_75|+|+|
+mnasnet1_0|+|+|
+mnasnet1_3|+|+|
+mobilenet_v2|+|+|
+resnext50_32x4d|+|+|
+resnext101_32x8d|+|+|
+resnet18|+|+|
+resnet34|+|+|
+resnet50|+|+|
+resnet101|+|+|
+resnet152|+|+|
+shufflenet_v2_x0_5|+|+|
+shufflenet_v2_x1_0|+|+|
+shufflenet_v2_x1_5|+|+|
+shufflenet_v2_x2_0|+|+|
+squeezenet1_0|+|+|
+squeezenet1_1|+|+|
+vgg11|+|+|
+vgg11_bn|+|+|
+vgg13|+|+|
+vgg13_bn|+|+|
+vgg16|+|+|
+vgg16_bn|+|+|
+vgg19|+|+|
+vgg19_bn|+|+|
+wide_resnet50_2|+|+|
+wide_resnet101_2|+|+|
+
+### Object detection
+
+Model | Availability in [TorchVision][torchvision] (0.15.1)| Availability in the validation table |
+-|-|-|
+fasterrcnn_resnet50_fpn|+|-|
+fasterrcnn_resnet50_fpn_v2|+|-|
+fasterrcnn_mobilenet_v3_large_fpn|+|-|
+fasterrcnn_mobilenet_v3_large_320_fpn|+|-|
+fcos_resnet50_fpn|+|-|
+retinanet_resnet50_fpn|+|-|
+retinanet_resnet50_fpn_v2|+|-|
+ssd300_vgg16|+|-|
+ssdlite320_mobilenet_v3_large|+|-|
+
+### Semantic segmentation
+
+Model | Availability in [TorchVision][torchvision] (0.15.1)| Availability in the validation table |
+-|-|-|
+deeplabv3_mobilenet_v3_large|+|-|
+deeplabv3_resnet50|+|-|
+deeplabv3_resnet101|+|-|
+fcn_resnet50|+|-|
+fcn_resnet101|+|-|
+lraspp_mobilenet_v3_large|+|-|
+
+### Instance segmentation
+
+Model | Availability in [TorchVision][torchvision] (0.15.1)| Availability in the validation table |
+-|-|-|
+maskrcnn_resnet50_fpn|+|-|
+maskrcnn_resnet50_fpn_v2|+|-|
+
+
+### Keypoint Detection
+
+Model | Availability in [TorchVision][torchvision] (0.15.1)| Availability in the validation table |
+-|-|-|
+keypointrcnn_resnet50_fpn|+|-|
+
+
+<!-- LINKS -->
+[torchvision]: https://pytorch.org/vision/stable/models.html