Cifar10 上的 ReNet18 SiLU 线性化

本文主要介绍在 cifar10 的试验 resnet18 的性能。

导入一些必要包：

import logging
import torch
from torch.fx.node import Argument
from typing import Any
from torch import nn, fx
from torchvision.models import resnet18, ResNet18_Weights
from torch_book.vision.classifier import Classifier, evaluate_accuracy
from torch_book.datasets.cifar10 import Cifar10
torch.cuda.empty_cache() # 清空 GPU 缓存
from torch_book.transforms.cutout import Cutout

model = torch.jit.load("params/resnet18_cifar10_silu_cutout.pt")

dataset = Cifar10(root="../data", batch_size=64, num_workers=4)
train_iter = dataset.train_loader()
test_iter = dataset.val_loader()

Files already downloaded and verified
Files already downloaded and verified

valid_acc = evaluate_accuracy(model, test_iter, device=torch.device("cuda:0"))
valid_acc

0.9463