<br />

|---|
| [![](https://www.tensorflow.org/images/GitHub-Mark-32px.png) View source on GitHub](https://github.com/tensorflow/models/blob/v2.20.0/official/core/actions.py#L76-L123) |

Eval action to save checkpoint with average weights when EMA is used.

    tfm.core.actions.EMACheckpointing(
        export_dir: str,
        optimizer: tf_keras.optimizers.Optimizer,
        checkpoint: tf.train.Checkpoint,
        max_to_keep: int = 1
    )

This action swaps the weights of the model with the average weights, then it
saves the checkpoint under export_dir/ema_checkpoints. Checkpointing is
expensive for large models, so doing this action in eval is more efficient
than training.

<br />

<br />

<br />

| ## Args ||
|---|---|
| `export_dir` | `str` for the export directory of the EMA average weights. |
| `optimizer` | `tf_keras.optimizers.Optimizer` optimizer instance used for training. This will be used to swap the model weights with the average weigths. |
| `checkpoint` | [`tf.train.Checkpoint`](https://www.tensorflow.org/api_docs/python/tf/train/Checkpoint) instance. |
| `max_to_keep` | `int` for max checkpoints to keep in ema_checkpoints subdir. |

<br />

## Methods

### `__call__`

[View source](https://github.com/tensorflow/models/blob/v2.20.0/official/core/actions.py#L115-L123)

    __call__(
        output: orbit.runner.Output
    )

Swaps model weights, and saves the checkpoint.

<br />

<br />

<br />

| Args ||
|---|---|
| `output` | The train or eval output. |

<br />