> ## Documentation Index
> Fetch the complete documentation index at: https://training-docs.cerebras.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Loop

> Learn how to configure the training and validation loops of the Trainer using two `LoopCallback` subclasses.

## Prerequisites

Read the [Trainer Overview](../../../model-zoo/trainer-overview) and [Trainer Configuration Overview](../../../model-zoo/trainer-configuration-overview) for a basic overview of how to run Model Zoo models.

## Configure the Loop

The `loop` argument allows you to manage the training and/or validation loop.

The [`Trainer`](https://training-api.cerebras.ai/en/latest/wsc/Model-zoo/api/index.html) takes in a [`LoopCallback`](https://training-api.cerebras.ai/en/latest/wsc/Model-zoo/api/generated/cerebras.modelzoo.trainer.callbacks.html#cerebras.modelzoo.trainer.callbacks.LoopCallback "cerebras.modelzoo.trainer.callbacks.LoopCallback") subclass that is used to configure loop options such as number of steps/epochs to run for and how often to run validation.

A `LoopCallback` cannot be instantiated directly, [`TrainingLoop`](https://training-api.cerebras.ai/en/latest/wsc/Model-zoo/api/generated/cerebras.modelzoo.trainer.callbacks.html#cerebras.modelzoo.trainer.callbacks.TrainingLoop "cerebras.modelzoo.trainer.callbacks.TrainingLoop") or [`ValidationLoop`](https://training-api.cerebras.ai/en/latest/wsc/Model-zoo/api/generated/cerebras.modelzoo.trainer.callbacks.html#cerebras.modelzoo.trainer.callbacks.ValidationLoop "cerebras.modelzoo.trainer.callbacks.ValidationLoop") must be used instead.

### Configure for training

The `TrainingLoop` callback is used to configure the `Trainer` to run a `fit` task. The majority of loop arguments reference `step`. The `step` is simply a batch of training/validation data.

**Arguments**

* `num_steps`: The total number of steps to train for.

* `max_steps`: The maximum number of global steps to train for. `num_steps` supersedes this.

* `num_epochs`: The number of epochs to train for. Mutually exclusive with `num_steps`.

* `steps_per_epoch`: The number of steps to train for in each epoch.

* `eval_frequency`: The frequency at which validation is performed. See `LoopCallback` for more details on options.

* `eval_steps`: The number of validation steps to perform.

* `grad_accum_steps`: The number of steps to accumulate gradients before performing and optimizer step. Only relevant for `"CPU"` and `"GPU"` runs.

<Note>
  If you plan on running any kind of training (calling `fit`), you must use a `TrainingLoop`. If you plan on running **only** validation, you may use a `ValidationLoop`.
</Note>

In the example below, we configure the `Trainer` to run for 1000 steps and run validation for 50 steps every 100 training steps.

<CodeGroup>
  ```yaml YAML theme={null}

  trainer:
    init:
      ...
      loop:
        num_steps: 1000
        eval_steps: 50
        eval_frequency: 100
      ...
    fit:
      ...
  ```

  ```python Python theme={null}
  from cerebras.modelzoo import Trainer
  from cerebras.modelzoo.trainer.callbacks import TrainingLoop

  trainer = Trainer(
      ...,
      loop=TrainingLoop(
          num_steps=1000,
          eval_steps=50,
          eval_frequency=100,
      ),
      ...,
  )

  trainer.fit(...)
  ```
</CodeGroup>

### Configure for Validation

The `ValidationLoop` callback is used to configure the `Trainer` to run a `validate` or `validate_all` task.

**Arguments**

* `eval_steps`: The number of validation steps to perform.

* `hook`: The base name of the validation hooks to run. Used to extend validation functionality by implementing custom validation callbacks. See `EleutherEvalHarnessLoop` for an example. Defaults to `"validate"`.

<Note>
  `ValidationLoop` can only be used if you plan on running **only** validation tasks (calling `validate` or `validate_all`). Otherwise, use `TrainingLoop`.
</Note>

In the example below, we configure the `Trainer` to run validation for 100 steps. We do not need to set any training related options such as `num_steps` or `eval_frequency` since we are only running validation.

<CodeGroup>
  ```yaml YAML theme={null}

  trainer:
    init:
      ...
      loop:
        eval_frequency: 100
      ...
    validate:
      ...
  ```

  ```python Python theme={null}

  from cerebras.modelzoo import Trainer
  from cerebras.modelzoo.trainer.callbacks import ValidationLoop

  trainer = Trainer(
      ...,
      loop=ValidationLoop(
          eval_frequency=100,
      ),
      ...,
  )

  trainer.validate(...)

  ```
</CodeGroup>

<Note>
  `TrainingLoop` supports both training and validation because it instantiates a `ValidationLoop` on initalization.
</Note>

<Note>
  Everytime validation runs, we are restarting the validation dataloaders from scratch. This is not the same for training where we resume training from the where we left off in the training dataloader.
</Note>