PyTorch轉TensorRT

• 安裝TensorRT
  按照官網的安裝教程即可，我裝的是TensorRT-6.0.1.5。
• 安裝onnx

sudo apt-get install protobuf-compiler libprotoc-dev
pip install onnx==1.5.0

• 安裝 onnx-tensorrt

git clone --recursive https://github.com/onnx/onnx-tensorrt.git
mkdir build
cd build
cmake .. -DTENSORRT_ROOT=<tensorrt_install_dir>
OR
cmake .. -DTENSORRT_ROOT=<tensorrt_install_dir> -DGPU_ARCHS="61"
make -j8
sudo make install

如果遇到bug, `git checkout v5.0，然後刪除build文件夾裏的內容重新build.

• 準備PyTorch model

簡單起見,以torch.save(net,’./model_300.pkl’)形式保存

• PyTorch model 轉 ONNX

import torch
​
model = './model_300.pkl'
​
dummy_input = torch.randn(batch_size, 3, 300, 300, device='cuda')
model = torch.load(model)
torch.onnx.export(model, dummy_input,"mymodel.onnx" , verbose=False)

• 測試ONNX model 是否與PyTorch model 輸出一致
  你應該在onnx_tensorrt 目錄下

import onnx_tensorrt.backend as backend
import cv2
import onnx
import numpy as np
​
​
model = onnx.load("model_300.onnx")
engine = backend.prepare(model, device='CUDA:0')
path = '../Net/test.jpg'
img = cv2.imread(path)
print(img.shape)
img = cv2.resize(img,(300,300))
img = img.transpose(2,0,1)
img = np.ascontiguousarray(img)
img = img[np.newaxis,:]
print(img.shape)
input_data= img.astype(np.float32)
data =np.random.random(size=(3, 300, 300))
#data = data.transpose(2,0,1)
data = data[np.newaxis,:]
#input_data=data.astype(np.float32)
#input_data = np.random.random(size=(1, 3, 300, 300)).astype(np.float32)
output_data = engine.run(input_data)
print(output_data)

• ONNX 轉 TensorRT engine

onnx2trt model_300.onnx -o my_engine.trt

ref:
https://zhuanlan.zhihu.com/p/74144263