***

layout: overview
slug: nemo-curator/nemo\_curator/models/aesthetics
title: nemo\_curator.models.aesthetics
--------------------------------------

## Module Contents

### Classes

| Name                                                                 | Description                                                 |
| -------------------------------------------------------------------- | ----------------------------------------------------------- |
| [`AestheticScorer`](#nemo_curator-models-aesthetics-AestheticScorer) | Public interface for aesthetic scoring of video embeddings. |
| [`MLP`](#nemo_curator-models-aesthetics-MLP)                         | Multi-layer perceptron.                                     |

### Data

[`_AESTHETICS_MODEL_ID`](#nemo_curator-models-aesthetics-_AESTHETICS_MODEL_ID)

[`_AESTHETICS_MODEL_REVISION`](#nemo_curator-models-aesthetics-_AESTHETICS_MODEL_REVISION)

### API

<Anchor id="nemo_curator-models-aesthetics-AestheticScorer">
  <CodeBlock showLineNumbers={false} wordWrap={true}>
    ```python
    class nemo_curator.models.aesthetics.AestheticScorer(
        model_dir: str
    )
    ```
  </CodeBlock>
</Anchor>

<Indent>
  **Bases:** [ModelInterface](/nemo-curator/nemo_curator/models/base#nemo_curator-models-base-ModelInterface)

  Public interface for aesthetic scoring of video embeddings.

  This class provides a standardized interface for scoring the aesthetic quality
  of video embeddings using a pre-trained model.

  <ParamField path="device" type="= 'cuda' if torch.cuda.is_available() else 'cpu'" />

  <ParamField path="dtype" type="= torch.float32" />

  <ParamField path="model_id_names" type="list[str]">
    Get the model ID names associated with this aesthetic scorer.
  </ParamField>

  <Anchor id="nemo_curator-models-aesthetics-AestheticScorer-__call__">
    <CodeBlock showLineNumbers={false} wordWrap={true}>
      ```python
      nemo_curator.models.aesthetics.AestheticScorer.__call__(
          embeddings: torch.Tensor | numpy.typing.NDArray[numpy.float32]
      ) -> torch.Tensor
      ```
    </CodeBlock>
  </Anchor>

  <Indent>
    Score the aesthetics of input embeddings.

    **Parameters:**

    <ParamField path="embeddings" type="torch.Tensor | npt.NDArray[np.float32]">
      Input embeddings as either a torch tensor or numpy array.
    </ParamField>

    **Returns:** `torch.Tensor`

    Aesthetic scores for each input embedding.
  </Indent>

  <Anchor id="nemo_curator-models-aesthetics-AestheticScorer-download_weights_on_node">
    <CodeBlock showLineNumbers={false} wordWrap={true}>
      ```python
      nemo_curator.models.aesthetics.AestheticScorer.download_weights_on_node(
          model_dir: str
      ) -> None
      ```
    </CodeBlock>
  </Anchor>

  <Indent>
    <Badge>
      classmethod
    </Badge>

    Download the weights for the aesthetic scorer on the node.
  </Indent>

  <Anchor id="nemo_curator-models-aesthetics-AestheticScorer-get_weights_path">
    <CodeBlock showLineNumbers={false} wordWrap={true}>
      ```python
      nemo_curator.models.aesthetics.AestheticScorer.get_weights_path() -> str
      ```
    </CodeBlock>
  </Anchor>

  <Indent>
    Get the path to the weights for the aesthetic scorer.
  </Indent>

  <Anchor id="nemo_curator-models-aesthetics-AestheticScorer-setup">
    <CodeBlock showLineNumbers={false} wordWrap={true}>
      ```python
      nemo_curator.models.aesthetics.AestheticScorer.setup() -> None
      ```
    </CodeBlock>
  </Anchor>

  <Indent>
    Set up the aesthetic scoring model by loading weights.
  </Indent>
</Indent>

<Anchor id="nemo_curator-models-aesthetics-MLP">
  <CodeBlock showLineNumbers={false} wordWrap={true}>
    ```python
    class nemo_curator.models.aesthetics.MLP()
    ```
  </CodeBlock>
</Anchor>

<Indent>
  **Bases:** `Module`

  Multi-layer perceptron.

  A neural network that processes embeddings to predict aesthetic scores.

  <ParamField path="layers" />

  <Anchor id="nemo_curator-models-aesthetics-MLP-forward">
    <CodeBlock showLineNumbers={false} wordWrap={true}>
      ```python
      nemo_curator.models.aesthetics.MLP.forward(
          embed: torch.Tensor
      ) -> torch.Tensor
      ```
    </CodeBlock>
  </Anchor>

  <Indent>
    Forward pass through the MLP.

    **Parameters:**

    <ParamField path="embed" type="torch.Tensor">
      Input embeddings tensor.
    </ParamField>

    **Returns:** `torch.Tensor`

    Predicted aesthetic scores.
  </Indent>
</Indent>

<Anchor id="nemo_curator-models-aesthetics-_AESTHETICS_MODEL_ID">
  <CodeBlock showLineNumbers={false} wordWrap={true}>
    ```python
    nemo_curator.models.aesthetics._AESTHETICS_MODEL_ID = 'ttj/sac-logos-ava1-l14-linearMSE'
    ```
  </CodeBlock>
</Anchor>

<Anchor id="nemo_curator-models-aesthetics-_AESTHETICS_MODEL_REVISION">
  <CodeBlock showLineNumbers={false} wordWrap={true}>
    ```python
    nemo_curator.models.aesthetics._AESTHETICS_MODEL_REVISION = '1e77fa0'
    ```
  </CodeBlock>
</Anchor>