有什么方法可以将 PyTorch 中可用的预训练模型下载到特定路径? [英] Is there any way I can download the pre-trained models available in PyTorch to a specific path?

问题描述

我指的是可以在这里找到的模型:https://pytorch.org/docs/stable/torchvision/models.html#torchvision-models

I am referring to the models that can be found here: https://pytorch.org/docs/stable/torchvision/models.html#torchvision-models

推荐答案

作为，@dennlinger 在他的中提到href="https://stackoverflow.com/a/52631758/8085890">答案:torch.utils.model_zoo，在您加载预训练模型时被内部调用.

As, @dennlinger mentioned in his answer : torch.utils.model_zoo, is being internally called when you load a pre-trained model.

更具体地说，每次加载预训练模型时都会调用方法:torch.utils.model_zoo.load_url().相同的文档中提到:

More specifically, the method: torch.utils.model_zoo.load_url() is being called every time a pre-trained model is loaded. The documentation for the same, mentions:

model_dir 的默认值为 $TORCH_HOME/models 其中$TORCH_HOME 默认为 ~/.torch.

The default value of model_dir is $TORCH_HOME/models where $TORCH_HOME defaults to ~/.torch.

默认目录可以用 $TORCH_HOME 覆盖环境变量.

The default directory can be overridden with the $TORCH_HOME environment variable.

这可以按如下方式完成:

This can be done as follows:

import torch 
import torchvision
import os

# Suppose you are trying to load pre-trained resnet model in directory- models
esnet

os.environ['TORCH_HOME'] = 'models\resnet' #setting the environment variable
resnet = torchvision.models.resnet18(pretrained=True)

我通过在 PyTorch 的 GitHub 存储库中提出问题来遇到上述解决方案:https://github.com/pytorch/vision/issues/616

I came across the above solution by raising an issue in the PyTorch's GitHub repository: https://github.com/pytorch/vision/issues/616

这导致了文档的改进，即上述解决方案.

This led to an improvement in the documentation i.e. the solution mentioned above.

这篇关于有什么方法可以将 PyTorch 中可用的预训练模型下载到特定路径?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！