README.md

English | 简体中文

ResNet Ready-to-deploy Model

• ResNet Deployment is based on the code of Torchvision and Pre-trained Models on ImageNet2012.

  • （1）Deployment is conducted after Export ONNX Model by the *.pt provided by Official Repository；
  • （2）The ResNet Model trained by personal data should Export ONNX Model. Please refer to Detailed Deployment Tutorials for deployment.

Export the ONNX Model

Import Torchvision, load the pre-trained model, and conduct model transformation as the following steps.

  import torch
  import torchvision.models as models

  model = models.resnet50(pretrained=True)
  batch_size = 1  #Batch size
  input_shape = (3, 224, 224)   #Input data, and change to personal input shape
  # #set the model to inference mode
  model.eval()
  x = torch.randn(batch_size, *input_shape)	# Generate tensor
  export_onnx_file = "ResNet50.onnx"			# Purpose ONNX file name
  torch.onnx.export(model,
                      x,
                      export_onnx_file,
                      opset_version=12,
                      input_names=["input"],	# Input name
                      output_names=["output"],	# Output name
                      dynamic_axes={"input":{0:"batch_size"},  # Batch variables
                                      "output":{0:"batch_size"}})

Download Pre-trained ONNX Model

For developers' testing, models exported by ResNet are provided below. Developers can download them directly. (The model accuracy in the following table is derived from the source official repository)

Model	Size	Accuracy
ResNet-18	45MB
ResNet-34	84MB
ResNet-50	98MB
ResNet-101	170MB

Detailed Deployment Documents

• Python Deployment
• C++ Deployment

Release Note

• Document and code are based on Torchvision v0.12.0