Prepare a model for distributed training

You've decided to use the Hugging Face Accelerator library to train your language translation model. Now it's time to prepare your model for distributed training!

Some data has been pre-loaded:

accelerator is an instance of Accelerator
model, optimizer, train_dataloader, and lr_scheduler have been defined

Acest exercițiu face parte din cursul

Efficient AI Model Training with PyTorch

Vezi cursul

Instrucțiuni pentru exercițiu

Call a method to prepare objects for distributed training.
Pass the training objects as positional arguments to accelerator.prepare(), matching the order of the output.

Exercițiu interactiv practic

Încearcă acest exercițiu completând acest cod de exemplu.

# Prepare objects for distributed training
model, optimizer, train_dataloader, lr_scheduler = ____.____(
    # Pass in the training objects matching the order of the output
    ____,
    ____,
    ____,
    ____)

Editează și rulează codul