Skip to content

检查点

在使用 Accelerate 训练 PyTorch 模型时,你可能经常需要保存和继续训练状态。这样做需要保存和加载模型、优化器、RNG 生成器和 GradScaler。Accelerate 内部提供了两个便捷函数来快速实现这一点:

  • 使用 [~Accelerator.save_state] 将上述所有内容保存到一个文件夹位置
  • 使用 [~Accelerator.load_state] 从之前的 save_state 加载所有存储的内容

为了进一步自定义通过 [~Accelerator.save_state] 保存状态的位置和方式,可以使用 [~utils.ProjectConfiguration] 类。例如,如果启用了 automatic_checkpoint_naming,每个保存的检查点将位于 Accelerator.project_dir/checkpoints/checkpoint_{checkpoint_number}

需要注意的是,这些状态应来自同一个训练脚本,不应来自两个不同的脚本。

  • 通过使用 [~Accelerator.register_for_checkpointing],你可以注册自定义对象以自动存储或从上述两个函数加载,只要该对象具有 state_dict load_state_dict 功能。这可以包括学习率调度器等对象。

以下是一个简短的示例,演示如何在训练过程中使用检查点保存和重新加载状态:

python
from accelerate import Accelerator
import torch

accelerator = Accelerator(project_dir="my/save/path")

my_scheduler = torch.optim.lr_scheduler.StepLR(my_optimizer, step_size=1, gamma=0.99)
my_model, my_optimizer, my_training_dataloader = accelerator.prepare(my_model, my_optimizer, my_training_dataloader)

# Register the LR scheduler
accelerator.register_for_checkpointing(my_scheduler)

# Save the starting state
accelerator.save_state()

device = accelerator.device
my_model.to(device)

# Perform training
for epoch in range(num_epochs):
    for batch in my_training_dataloader:
        my_optimizer.zero_grad()
        inputs, targets = batch
        inputs = inputs.to(device)
        targets = targets.to(device)
        outputs = my_model(inputs)
        loss = my_loss_function(outputs, targets)
        accelerator.backward(loss)
        my_optimizer.step()
    my_scheduler.step()

# Restore the previous state
accelerator.load_state("my/save/path/checkpointing/checkpoint_0")

恢复 DataLoader 的状态

在从检查点恢复后,如果在 epoch 的中途保存了状态,可能还需要从活动的 DataLoader 的某个特定点恢复。你可以使用 [~Accelerator.skip_first_batches] 来实现这一点。

python
from accelerate import Accelerator

accelerator = Accelerator(project_dir="my/save/path")

train_dataloader = accelerator.prepare(train_dataloader)
accelerator.load_state("my_state")

# Assume the checkpoint was saved 100 steps into the epoch
skipped_dataloader = accelerator.skip_first_batches(train_dataloader, 100)

# After the first iteration, go back to `train_dataloader`

# First epoch
for batch in skipped_dataloader:
    # Do something
    pass

# Second epoch
for batch in train_dataloader:
    # Do something
    pass