> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.nvidia.com/nemo/automodel/v0.4/nemo-automodel/nemo_automodel/components/attention/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.nvidia.com/nemo/automodel/_mcp/server.

# nemo_automodel.components.attention

## Submodules

* **[`nemo_automodel.components.attention.dflash_mask`](/nemo-automodel/nemo_automodel/components/attention/dflash_mask)**
* **[`nemo_automodel.components.attention.ffpa_attention`](/nemo-automodel/nemo_automodel/components/attention/ffpa_attention)**
* **[`nemo_automodel.components.attention.flex_attention`](/nemo-automodel/nemo_automodel/components/attention/flex_attention)**
* **[`nemo_automodel.components.attention.idlm_mask`](/nemo-automodel/nemo_automodel/components/attention/idlm_mask)**
* **[`nemo_automodel.components.attention.utils`](/nemo-automodel/nemo_automodel/components/attention/utils)**

## Docs

- [nemo_automodel.components.attention.dflash_mask](https://docs.nvidia.com/nemo/automodel/v0.4/nemo-automodel/nemo_automodel/components/attention/dflash_mask.md)
- [nemo_automodel.components.attention.ffpa_attention](https://docs.nvidia.com/nemo/automodel/v0.4/nemo-automodel/nemo_automodel/components/attention/ffpa_attention.md)
- [nemo_automodel.components.attention.flex_attention](https://docs.nvidia.com/nemo/automodel/v0.4/nemo-automodel/nemo_automodel/components/attention/flex_attention.md)
- [nemo_automodel.components.attention.idlm_mask](https://docs.nvidia.com/nemo/automodel/v0.4/nemo-automodel/nemo_automodel/components/attention/idlm_mask.md)
- [nemo_automodel.components.attention.utils](https://docs.nvidia.com/nemo/automodel/v0.4/nemo-automodel/nemo_automodel/components/attention/utils.md)