> For clean Markdown of any page, append .md to the page URL. > For a complete documentation index, see https://docs.nvidia.com/nemo/curator/v26.04/nemo-curator/nemo_curator/backends/ray_data/llms.txt. > For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.nvidia.com/nemo/curator/_mcp/server. # nemo_curator.backends.ray_data ## Submodules * **[`nemo_curator.backends.ray_data.adapter`](/nemo/curator/nemo-curator/nemo_curator/backends/ray_data/adapter)** * **[`nemo_curator.backends.ray_data.diagnostics`](/nemo/curator/nemo-curator/nemo_curator/backends/ray_data/diagnostics)** * **[`nemo_curator.backends.ray_data.executor`](/nemo/curator/nemo-curator/nemo_curator/backends/ray_data/executor)** * **[`nemo_curator.backends.ray_data.utils`](/nemo/curator/nemo-curator/nemo_curator/backends/ray_data/utils)** ## Package Contents ### Classes | Name | Description | | ----------------------------------------------------------------------------- | ----------------------------------------------- | | [`RayDataExecutor`](#nemo_curator-backends-ray_data-executor-RayDataExecutor) | Ray Data-based executor for pipeline execution. | ### API ```python class nemo_curator.backends.ray_data.RayDataExecutor( config: dict[str, typing.Any] | None = None, ignore_head_node: bool = False ) ``` **Bases:** [BaseExecutor](/nemo/curator/nemo-curator/nemo_curator/backends/base#nemo_curator-backends-base-BaseExecutor) Ray Data-based executor for pipeline execution. This executor: 1. Executes setup on Ray nodes for all stages 2. Converts initial tasks to Ray Data dataset 3. Applies each stage as a Ray Data transformation (as a task or actor in map\_batches) 4. Returns final results as a list of tasks ```python nemo_curator.backends.ray_data.RayDataExecutor._dataset_to_tasks( dataset: ray.data.Dataset ) -> list[nemo_curator.tasks.Task] ``` Convert Ray Data dataset back to list of tasks. **Parameters:** **`dataset`** `Dataset` Ray Data dataset containing Task objects --- **Returns:** `list[` [`Task`](/nemo/curator/nemo-curator/nemo_curator/tasks/tasks#nemo_curator-tasks-tasks-Task) `]` List of Task objects ```python nemo_curator.backends.ray_data.RayDataExecutor._tasks_to_dataset( tasks: list[nemo_curator.tasks.Task] ) -> ray.data.Dataset ``` Convert list of tasks to Ray Data dataset. **Parameters:** **`tasks`** `list[Task]` `list[` [`Task`](/nemo/curator/nemo-curator/nemo_curator/tasks/tasks#nemo_curator-tasks-tasks-Task) `]` — List of Task objects --- **Returns:** `Dataset` Ray Data dataset containing Task objects directly ```python nemo_curator.backends.ray_data.RayDataExecutor.execute( stages: list[nemo_curator.stages.base.ProcessingStage], initial_tasks: list[nemo_curator.tasks.Task] | None = None ) -> list[nemo_curator.tasks.Task] ``` Execute the pipeline stages using Ray Data. **Parameters:** **`stages`** `list[ProcessingStage]` `list[` [`ProcessingStage`](/nemo/curator/nemo-curator/nemo_curator/stages/base#nemo_curator-stages-base-ProcessingStage) `]` — List of processing stages to execute --- **`initial_tasks`** `list[Task]` — default: None `list[` [`Task`](/nemo/curator/nemo-curator/nemo_curator/tasks/tasks#nemo_curator-tasks-tasks-Task) `]` — Initial tasks to process (can be None for empty start) --- **Returns:** `list[` [`Task`](/nemo/curator/nemo-curator/nemo_curator/tasks/tasks#nemo_curator-tasks-tasks-Task) `]` list\[Task]: List of final processed tasks ## Docs - [nemo_curator.backends.ray_data.adapter](https://docs.nvidia.com/nemo/curator/v26.04/nemo-curator/nemo_curator/backends/ray_data/adapter.md) - [nemo_curator.backends.ray_data.diagnostics](https://docs.nvidia.com/nemo/curator/v26.04/nemo-curator/nemo_curator/backends/ray_data/diagnostics.md) - [nemo_curator.backends.ray_data.executor](https://docs.nvidia.com/nemo/curator/v26.04/nemo-curator/nemo_curator/backends/ray_data/executor.md) - [nemo_curator.backends.ray_data.utils](https://docs.nvidia.com/nemo/curator/v26.04/nemo-curator/nemo_curator/backends/ray_data/utils.md)