PIP安装火炬闪电

确保你的父模型使用pl.LightningModule而不是nn。模块

使用pytorch闪电保存和加载检查点

import pytorch_lightning as pl

model = MyLightningModule(hparams)
trainer.fit(model)
trainer.save_checkpoint("example.ckpt")
new_model = MyModel.load_from_checkpoint(checkpoint_path="example.ckpt")

保存在本地

如何保存模型取决于将来如何访问它。如果你可以调用模型类的一个新实例，那么你所需要做的就是用model.state_dict()保存/加载模型的权重:

# Save:
torch.save(old_model.state_dict(), PATH)

# Load:
new_model = TheModelClass(*args, **kwargs)
new_model.load_state_dict(torch.load(PATH))

如果你因为任何原因(或者更喜欢简单的语法)不能，那么你可以使用torch.save()保存整个模型(实际上是对定义模型的文件的引用，以及它的state_dict):

# Save:
torch.save(old_model, PATH)

# Load:
new_model = torch.load(PATH)

但是由于这是对定义模型类的文件位置的引用，所以这段代码是不可移植的，除非这些文件也移植到相同的目录结构中。

保存到云- TorchHub

如果您希望您的模型是可移植的，您可以轻松地使用torch.hub导入它。如果你在github repo中添加了一个适当定义的hubconf.py文件，这可以很容易地在PyTorch中调用，使用户能够加载你的模型，带/不带权重:

hubconf.py (github.com/repo_owner/repo_name)

dependencies = ['torch']
from my_module import mymodel as _mymodel

def mymodel(pretrained=False, **kwargs):
    return _mymodel(pretrained=pretrained, **kwargs)

加载模型:

new_model = torch.hub.load('repo_owner/repo_name', 'mymodel')
new_model_pretrained = torch.hub.load('repo_owner/repo_name', 'mymodel', pretrained=True)

如果您想保存模型，并希望稍后恢复训练:

单一的GPU: 拯救策略:

state = {
        'epoch': epoch,
        'state_dict': model.state_dict(),
        'optimizer': optimizer.state_dict(),
}
savepath='checkpoint.t7'
torch.save(state,savepath)

负载:

checkpoint = torch.load('checkpoint.t7')
model.load_state_dict(checkpoint['state_dict'])
optimizer.load_state_dict(checkpoint['optimizer'])
epoch = checkpoint['epoch']

多个GPU: 保存

state = {
        'epoch': epoch,
        'state_dict': model.module.state_dict(),
        'optimizer': optimizer.state_dict(),
}
savepath='checkpoint.t7'
torch.save(state,savepath)

负载:

checkpoint = torch.load('checkpoint.t7')
model.load_state_dict(checkpoint['state_dict'])
optimizer.load_state_dict(checkpoint['optimizer'])
epoch = checkpoint['epoch']

#Don't call DataParallel before loading the model otherwise you will get an error

model = nn.DataParallel(model) #ignore the line if you want to load on Single GPU

现在所有内容都写在官方教程中: https://pytorch.org/tutorials/beginner/saving_loading_models.html

关于如何保存和保存什么，您有几个选项，所有这些都在本教程中进行了解释。

在他们的github回购上找到了这个页面:

Recommended approach for saving a model There are two main approaches for serializing and restoring a model. The first (recommended) saves and loads only the model parameters: torch.save(the_model.state_dict(), PATH) Then later: the_model = TheModelClass(*args, **kwargs) the_model.load_state_dict(torch.load(PATH)) The second saves and loads the entire model: torch.save(the_model, PATH) Then later: the_model = torch.load(PATH) However in this case, the serialized data is bound to the specific classes and the exact directory structure used, so it can break in various ways when used in other projects, or after some serious refactors.

请参见官方PyTorch教程中的保存和加载模型部分。

PIP安装火炬闪电

确保你的父模型使用pl.LightningModule而不是nn。模块

使用pytorch闪电保存和加载检查点

import pytorch_lightning as pl

model = MyLightningModule(hparams)
trainer.fit(model)
trainer.save_checkpoint("example.ckpt")
new_model = MyModel.load_from_checkpoint(checkpoint_path="example.ckpt")