【代码记录】pytorch推理及与onnx推理精度对比

最新推荐文章于 2024-09-19 11:43:10 发布

甜苦酱汁

最新推荐文章于 2024-09-19 11:43:10 发布

阅读量1.1k

点赞数 1

分类专栏：代码记录文章标签： pytorch python numpy

本文链接：https://blog.youkuaiyun.com/weixin_45392674/article/details/127671658

版权

1.pytorch推理

import cv2
import sys
import numpy as np
import torch, os
from torch import nn
import torchvision.models as models
import torchvision.transforms as transforms
from PIL import Image

def inference(mat):
    # preprocess
   
    #预处理要保持一致
    mean_nom = [0.485, 0.456, 0.406]
    std_norm = [0.229, 0.229, 0.229]

    transform_test = transforms.Compose([transforms.Resize((224, 224)),transforms.ToTensor(