anyscale 0.26.47__py3-none-any.whl → 0.26.49__py3-none-any.whl

anyscale/commands/llm/dataset_commands.py

@@ -1,269 +0,0 @@
-import os
-from typing import Optional
-
-import click
-from dateutil import tz
-from rich import print as rprint
-import tabulate
-
-import anyscale
-from anyscale.commands import command_examples
-from anyscale.commands.util import AnyscaleCommand
-
-
-@click.group("dataset", help="Dataset files stored on your Anyscale cloud.")
-def dataset_cli():
-    pass
-
-
-@dataset_cli.command(
-    name="get",
-    short_help="Retrieves metadata about a dataset.",
-    cls=AnyscaleCommand,
-    is_alpha=True,
-    example=command_examples.LLM_DATASET_GET_EXAMPLE,
-)
-@click.argument(
-    "name", required=True,
-)
-@click.option(
-    "--version",
-    "-v",
-    required=False,
-    type=int,
-    help="Version of the dataset. "
-    "If a negative integer is provided, the dataset returned is this many versions back of the latest version. "
-    "Default: Latest version.",
-)
-@click.option(
-    "--project",
-    required=False,
-    help="Name of the Anyscale project that the dataset belongs to. "
-    "If not provided, all projects will be searched.",
-)
-def get_dataset(
-    name: str, version: Optional[int], project: Optional[str],
-):
-    """
-    Retrieves metadata about a dataset.
-
-    NAME = Name of the dataset
-
-    Example usage:
-
-        anyscale llm dataset get my_first_dataset
-
-    Retrieve the second latest version of the dataset:
-
-        anyscale llm dataset get my_first_dataset -v -1
-    """
-    dataset = anyscale.llm.dataset.get(name, version, project)
-    rprint(dataset)
-
-
-@dataset_cli.command(
-    name="upload",
-    short_help="Upload a dataset to your Anyscale cloud.",
-    cls=AnyscaleCommand,
-    is_alpha=True,
-    example=command_examples.LLM_DATASET_UPLOAD_EXAMPLE,
-)
-@click.argument("dataset_file", required=True)
-@click.option(
-    "--name",
-    "-n",
-    required=True,
-    help="Name of a new dataset, or an existing dataset, to upload a new version of.",
-)
-@click.option(
-    "--description", required=False, help="Description of the dataset version.",
-)
-@click.option(
-    "--cloud",
-    required=False,
-    help="Name of the Anyscale cloud to upload a new dataset to. "
-    "If not provided, the default cloud will be used.",
-)
-@click.option(
-    "--project",
-    required=False,
-    help="Name of the Anyscale project to upload a new dataset to. "
-    "If not provided, the default project of the cloud will be used.",
-)
-def upload_dataset(
-    dataset_file: str,
-    name: str,
-    description: Optional[str],
-    cloud: Optional[str],
-    project: Optional[str],
-):
-    """
-    Uploads a dataset, or a new version of a dataset, to your Anyscale cloud.
-
-    DATASET_FILE = Path to the dataset file to upload
-
-    Example usage:
-
-        anyscale llm dataset upload path/to/my_dataset.jsonl -n my_first_dataset
-
-        anyscale llm dataset upload my_dataset.jsonl -n second_dataset.jsonl
-
-        anyscale llm dataset upload my_dataset2.jsonl -n second_dataset.jsonl --description 'added 3 lines'
-
-    \b
-    NOTE:
-    If you are uploading a new version, have run this from within an Anyscale workspace,
-    and neither `--cloud` nor `--project` is provided, the cloud and project of the workspace will be used.
-    """
-    dataset = anyscale.llm.dataset.upload(
-        dataset_file, name, description=description, cloud=cloud, project=project,
-    )
-    rprint(dataset)
-
-
-@dataset_cli.command(
-    name="download",
-    short_help="Download a dataset.",
-    cls=AnyscaleCommand,
-    is_alpha=True,
-    example=command_examples.LLM_DATASET_DOWNLOAD_EXAMPLE,
-)
-@click.argument(
-    "name", required=True,
-)
-@click.option(
-    "--version",
-    "-v",
-    required=False,
-    type=int,
-    help="Version of the dataset to download. "
-    "If a negative integer is provided, the dataset returned is this many versions back of the latest version. "
-    "Default: Latest version.",
-)
-@click.option(
-    "--project",
-    required=False,
-    help="Name of the Anyscale project to download the dataset from. "
-    "If not provided, all projects will be searched.",
-)
-@click.option(
-    "--output",
-    "-o",
-    required=False,
-    help="Path to save the downloaded dataset to."
-    "If not provided, the dataset contents will be printed to the terminal.",
-)
-def download_dataset(
-    name: str, version: Optional[int], project: Optional[str], output: Optional[str],
-):
-    """
-    Downloads a dataset from your Anyscale cloud.
-
-    NAME = Name of the dataset to download
-
-    Prints the dataset contents to the terminal by default.
-
-    Example usage:
-
-        anyscale llm dataset download my_first_dataset.jsonl
-
-    Save the dataset to a file:
-
-        anyscale llm dataset download my_dataset.jsonl -o ~/Downloads/my_dataset.jsonl
-
-    Retrieve the second latest version of the dataset:
-
-        anyscale llm dataset download my_dataset.jsonl -v -1
-    """
-    if output:
-        output = os.path.expanduser(output)
-        if os.path.exists(output):
-            raise click.ClickException(f"File already exists at '{output}'")
-    downloaded_contents = anyscale.llm.dataset.download(name, version, project)
-    if output:
-        # Create the parent directory if it doesn't exist
-        parent_dir = os.path.dirname(output)
-        if parent_dir:
-            os.makedirs(parent_dir, exist_ok=True)
-        with open(output, "wb") as output_file:
-            output_file.write(downloaded_contents)
-        rprint(f"Dataset '{name}' downloaded to '{output}'")
-        return
-    print(downloaded_contents.decode())
-
-
-@dataset_cli.command(
-    name="list",
-    short_help="List datasets.",
-    cls=AnyscaleCommand,
-    is_alpha=True,
-    example=command_examples.LLM_DATASET_LIST_EXAMPLE,
-)
-@click.option(
-    "--limit",
-    "-l",
-    required=False,
-    default=10,
-    type=int,
-    help="Maximum number of datasets to return. Default: 10.",
-)
-@click.option(
-    "--after",
-    required=False,
-    help="ID of the dataset to start the listing from. If provided, the list will start from the dataset after this ID.",
-)
-@click.option(
-    "--name-contains",
-    "-n",
-    required=False,
-    help="Filter datasets by name. If provided, only datasets with name containing this string will be returned.",
-)
-@click.option(
-    "--cloud",
-    required=False,
-    help="Name of the Anyscale cloud to search in. If not provided, all clouds will be searched.",
-)
-@click.option(
-    "--project",
-    required=False,
-    help="Name of the Anyscale project to search in. If not provided, all projects will be searched.",
-)
-def list_datasets(
-    limit: int,
-    after: Optional[str],
-    name_contains: Optional[str],
-    cloud: Optional[str],
-    project: Optional[str],
-):
-    """
-    Lists datasets.
-
-    Example usage:
-
-        anyscale llm dataset list
-
-    List datasets in a specific project:
-
-        anyscale llm dataset list --project my_project
-    """
-    datasets = anyscale.llm.dataset.list(
-        limit=limit,
-        after=after,
-        name_contains=name_contains,
-        cloud=cloud,
-        project=project,
-    )
-    table = tabulate.tabulate(
-        [
-            (
-                d.id,
-                d.name,
-                d.description,
-                d.created_at.astimezone(tz=tz.tzlocal()).strftime("%m/%d/%Y %I:%M %p"),
-                d.num_versions,
-            )
-            for d in datasets
-        ],
-        headers=["ID", "Name", "Description", "Created At", "Num Versions"],
-    )
-    rprint(table)

anyscale/commands/llm/group.py

@@ -1,15 +0,0 @@
-import click
-
-from anyscale.commands.llm.dataset_commands import dataset_cli
-from anyscale.commands.llm.models_commands import models_cli
-
-
-@click.group(
-    "llm", help="Interact with Anyscale's LLM APIs.",
-)
-def llm_cli():
-    pass
-
-
-llm_cli.add_command(dataset_cli)
-llm_cli.add_command(models_cli)

anyscale/commands/llm/models_commands.py

@@ -1,123 +0,0 @@
-from typing import Optional
-
-import click
-
-from anyscale.commands import command_examples
-from anyscale.commands.util import AnyscaleCommand
-from anyscale.controllers.llm.models_controller import ModelsController
-from anyscale.util import validate_non_negative_arg
-
-
-@click.group("model", help="Finetuned models stored on your Anyscale cloud.")
-def models_cli():
-    pass
-
-
-@models_cli.command(
-    name="get",
-    short_help="Retrieves information for a model in your Anyscale cloud.",
-    cls=AnyscaleCommand,
-    is_alpha=True,
-    example=command_examples.LLM_MODELS_GET_EXAMPLE,
-)
-@click.option(
-    "--model-id",
-    required=False,
-    type=str,
-    default=None,
-    help="ID for the model of interest",
-)
-@click.option(
-    "--job-id",
-    required=False,
-    type=str,
-    default=None,
-    help="ID for the Anyscale job corresponding to the fine-tuning run",
-)
-def get_model(model_id: Optional[str], job_id: Optional[str]) -> None:
-    """
-    Gets the model card for the given model ID or corresponding job ID.
-
-    Example usage:
-
-        anyscale llm model get --model-id my-model-id
-
-        anyscale llm model get --job-id job_123
-    """
-    ModelsController().get_model(model_id=model_id, job_id=job_id)
-
-
-@models_cli.command(
-    name="delete",
-    short_help="Delete a fine-tuned model in your Anyscale cloud.",
-    cls=AnyscaleCommand,
-    is_alpha=True,
-    example=command_examples.LLM_MODELS_DELETE_EXAMPLE,
-)
-@click.argument("model_id", required=True)
-def delete_model(model_id: str) -> None:
-    """
-    Deletes the model for the given model ID. Requires owner permission for the corresponding Anyscale project.
-
-    MODEL_ID = ID for the model of interest
-
-    Example usage:
-
-        anyscale llm model delete my-model-id
-    """
-    ModelsController().delete_model(model_id)
-
-
-@models_cli.command(
-    name="list",
-    short_help="Lists fine-tuned models available to the user.",
-    cls=AnyscaleCommand,
-    is_alpha=True,
-    example=command_examples.LLM_MODELS_LIST_EXAMPLE,
-)
-@click.option(
-    "--cloud-id",
-    required=False,
-    type=str,
-    help="Cloud ID to filter by. If not specified, all models from all visible clouds (filtered optionally by `project_id`) are listed.",
-)
-@click.option(
-    "--project-id",
-    required=False,
-    type=str,
-    help="Project ID to filter by. If not specified, all the models from all visible projects (filtered optionally by `cloud_id`) are listed.",
-)
-@click.option(
-    "--max-items",
-    required=False,
-    type=int,
-    default=20,
-    help="Maximum number of items to show in the list. By default, the 20 most recently created models are fetched.",
-    callback=validate_non_negative_arg,
-)
-def list_models(cloud_id: Optional[str], project_id: Optional[str], max_items: int):
-    """
-    Lists fine-tuned models available to the user.
-
-    By default, all models in all visible clouds under all visible projects to the user are listed. This is optionally filtered by `project_id` and/or `cloud_id`.
-
-    Example usage:
-
-        anyscale llm model list
-
-        anyscale llm model list --max-items 50
-
-        anyscale llm model list --cloud-id cld_123
-
-        anyscale llm model list --project-id prj_123
-
-        anyscale llm model list --cloud-id cld_123 --project-id prj_123
-
-
-    NOTE:
-    If you are running this from within an Anyscale workspace, and neither `cloud_id` nor `project_id` are provided, the cloud and project of the workspace will be used.
-    """
-    controller = ModelsController()
-    controller.list_models(
-        cloud_id=cloud_id, project_id=project_id, max_items=max_items
-    )

anyscale/controllers/llm/models_controller.py

@@ -1,144 +0,0 @@
-import json
-from typing import Any, Dict, Optional
-
-from rich import print as rprint
-
-from anyscale.api_utils.common_utils import (
-    get_current_workspace_id,
-    source_cloud_id_and_project_id,
-)
-from anyscale.cli_logger import BlockLogger
-from anyscale.client.openapi_client import FinetunedmodelListResponse
-from anyscale.client.openapi_client.models import FineTunedModel as APIFineTunedModel
-from anyscale.controllers.base_controller import BaseController
-from anyscale.llm.model.models import DeletedFineTunedModel, FineTunedModel
-
-
-LIST_ENDPOINT_COUNT = 20
-
-
-class ModelsController(BaseController):
-    def __init__(
-        self, log: Optional[BlockLogger] = None, initialize_auth_api_client: bool = True
-    ):
-        if log is None:
-            log = BlockLogger()
-
-        super().__init__(initialize_auth_api_client=initialize_auth_api_client)
-
-        self.log = log
-        self.log.open_block("Output")
-
-    def _parse_response_model_get(self, model: APIFineTunedModel) -> FineTunedModel:
-        return FineTunedModel(
-            id=model.id,
-            base_model_id=model.base_model_id,
-            cloud_id=model.cloud_id,
-            # model.created_at is a datetime object, convert to unix timestamp
-            created_at=int(model.created_at.timestamp()),
-            # `creator` is a MiniUser object, just retrieve email
-            creator=model.creator.email if model.creator is not None else None,
-            ft_type=model.ft_type,
-            generation_config=model.generation_config,
-            job_id=model.job_id,
-            project_id=model.project_id,
-            storage_uri=model.storage_uri,
-            workspace_id=model.workspace_id,
-        )
-
-    def _truncate(self, val: str, limit=50):
-        return val[:limit] + "..." if len(val) > limit else val
-
-    def _format_as_dict(
-        self, model: FineTunedModel, truncate: bool = False
-    ) -> Dict[str, Any]:
-        output_map = {
-            "id": model.id,
-            "base_model_id": model.base_model_id,
-            "storage_uri": self._truncate(model.storage_uri)
-            if truncate
-            else model.storage_uri,
-            "ft_type": str(model.ft_type),
-            "cloud_id": model.cloud_id,
-            "project_id": model.project_id if model.project_id else "N/A",
-            "created_at": model.created_at,
-            "creator": model.creator if model.creator else "N/A",
-            "job_id": model.job_id if model.job_id else "N/A",
-            "workspace_id": model.workspace_id if model.workspace_id else "N/A",
-            "generation_config": self._truncate(json.dumps(model.generation_config))
-            if truncate
-            else model.generation_config,
-        }
-        return output_map
-
-    def get_model(self, model_id: Optional[str], job_id: Optional[str]):
-        """Retrieves model information given model id"""
-        if model_id:
-            model = self.api_client.get_model_api_v2_llm_models_model_id_get(
-                model_id
-            ).result
-        elif job_id:
-            model = self.api_client.get_model_by_job_id_api_v2_llm_models_get_by_job_id_job_id_get(
-                job_id
-            ).result
-        else:
-            raise ValueError("Atleast one of `model-id` or `job-id` should be provided")
-
-        model = self._parse_response_model_get(model)
-        formatted_model = self._format_as_dict(model)
-        rprint(formatted_model)
-        return
-
-    def delete_model(self, model_id: str):
-        deleted_model = self.api_client.delete_model_api_v2_llm_models_model_id_delete(
-            model_id
-        ).result
-        deleted_model_dict = deleted_model.to_dict()
-        deleted_model_dict["deleted_at"] = int(
-            deleted_model_dict["deleted_at"].timestamp()
-        )
-        deleted_model = DeletedFineTunedModel.from_dict(deleted_model_dict)
-        rprint(deleted_model.to_dict())
-        return
-
-    def list_models(
-        self, *, cloud_id: Optional[str], project_id: Optional[str], max_items: int
-    ):
-        """Lists fine-tuned models optionally filtered by `cloud_id` and `project_id`"""
-        if get_current_workspace_id() is not None:
-            # Resolve `cloud_id` and `project_id`. If not provided and if this is being run in a workspace,
-            # we use the `cloud_id` and `project_id` of the workspace
-            cloud_id, project_id = source_cloud_id_and_project_id(
-                internal_api=self.api_client,
-                external_api=self.anyscale_api_client,
-                cloud_id=cloud_id,
-                project_id=project_id,
-            )
-        paging_token = None
-        results = []
-        while True:
-            count = min(LIST_ENDPOINT_COUNT, max_items)
-            resp: FinetunedmodelListResponse = self.api_client.list_models_api_v2_llm_models_get(
-                cloud_id=cloud_id,
-                project_id=project_id,
-                paging_token=paging_token,
-                count=count,
-            )
-            models = resp.results
-            results.extend(models)
-            if not len(models) or not resp.metadata.next_paging_token:
-                break
-
-            if max_items and len(results) >= max_items:
-                break
-            paging_token = resp.metadata.next_paging_token
-
-        results = results[:max_items] if max_items else results
-        parsed_results = [self._parse_response_model_get(result) for result in results]
-        # get formatted dict with truncated strings for a nicer print
-        models_as_dicts = [
-            self._format_as_dict(model, truncate=True) for model in parsed_results
-        ]
-        print("MODELS:")
-        rprint(models_as_dicts)
-        return

anyscale/llm/__init__.py

@@ -1,2 +0,0 @@
-from anyscale.llm import dataset, model
-from anyscale.llm.sdk import LLMSDK

anyscale/llm/dataset/__init__.py

@@ -1,2 +0,0 @@
-from anyscale.llm.dataset._private.models import Dataset
-from anyscale.llm.dataset._private.sdk import download, get, list, upload

anyscale/llm/dataset/_private/docs.py

@@ -1,63 +0,0 @@
-GET_PY_EXAMPLE = """
-import anyscale
-from anyscale.llm.dataset import Dataset
-
-dataset: Dataset = anyscale.llm.dataset.get("my_first_dataset")
-print(f"Dataset name: '{dataset.name}'")  # Dataset name: 'my_first_dataset'
-
-# Get the second latest version of the dataset
-prev_dataset = anyscale.llm.dataset.get("my_first_dataset", version=-1)
-"""
-
-GET_PY_ARG_DOCSTRINGS = {
-    "name": "Name of the dataset",
-    "version": "Version of the dataset. If a negative integer is provided, the dataset returned is this many versions back of the latest version. Default: Latest version.",
-    "project": "Name of the Anyscale project that the dataset belongs to. If not provided, all projects will be searched.",
-}
-
-UPLOAD_PY_EXAMPLE = """
-import anyscale
-
-anyscale.llm.dataset.upload("path/to/my_first_dataset.jsonl", name="my_first_dataset")
-anyscale.llm.dataset.upload("my_dataset.jsonl", "second_dataset")
-anyscale.llm.dataset.upload("my_dataset2.jsonl", "second_dataset", description="added 3 lines")
-"""
-
-UPLOAD_PY_ARG_DOCSTRINGS = {
-    "dataset_file": "Path to the dataset file to upload.",
-    "name": "Name of a new dataset, or an existing dataset, to upload a new version of.",
-    "description": "Description of the dataset version.",
-    "cloud": "Name of the Anyscale cloud to upload a new dataset to. If not provided, the default cloud will be used.",
-    "project": "Name of the Anyscale project to upload a new dataset to. If not provided, the default project of the cloud will be used.",
-}
-
-DOWNLOAD_PY_EXAMPLE = """
-import anyscale
-
-dataset_contents: bytes = anyscale.llm.dataset.download("my_first_dataset.jsonl")
-jsonl_obj = [json.loads(line) for line in dataset_contents.decode().splitlines()]
-
-prev_dataset_contents = anyscale.llm.dataset.download("my_first_dataset.jsonl", version=-1)
-"""
-
-DOWNLOAD_PY_ARG_DOCSTRINGS = {
-    "name": "Name of the dataset to download.",
-    "version": "Version of the dataset to download. If a negative integer is provided, the dataset returned is this many versions back of the latest version. Default: Latest version.",
-    "project": "Name of the Anyscale project to download the dataset from. If not provided, all projects will be searched.",
-}
-
-LIST_PY_EXAMPLE = """
-import anyscale
-
-datasets = anyscale.llm.dataset.list(limit=10)
-for d in datasets:
-    print(f"Dataset name: '{d.name}'")  # Prints 10 dataset names
-"""
-
-LIST_PY_ARG_DOCSTRINGS = {
-    "limit": "Maximum number of datasets to return. Default: 1000.",
-    "after": "ID of the dataset to start the listing from. If provided, the list will start from the dataset after this ID.",
-    "name_contains": "Filter datasets by name. If provided, only datasets with name containing this string will be returned.",
-    "cloud": "Name of the Anyscale cloud to search in. If not provided, all clouds will be searched.",
-    "project": "Name of the Anyscale project to search in. If not provided, all projects will be searched.",
-}

anyscale/llm/dataset/_private/models.py

@@ -1,71 +0,0 @@
-from dataclasses import dataclass, field
-from datetime import datetime
-from typing import Optional
-
-from anyscale._private.models.model_base import ModelBase
-from anyscale.client.openapi_client.models import Dataset as InternalDataset
-from anyscale.commands import command_examples
-from anyscale.llm.dataset._private import docs
-
-
-@dataclass(frozen=True)
-class Dataset(ModelBase):
-    """
-    Metadata about a dataset, which is a file uploaded by a user to their Anyscale cloud.
-    """
-
-    __ignore_validation__ = True
-
-    __doc_py_example__ = docs.GET_PY_EXAMPLE
-    __doc_cli_example__ = command_examples.LLM_DATASET_GET_EXAMPLE
-
-    id: str = field(metadata={"docstring": "The ID of the dataset."})
-    name: str = field(metadata={"docstring": "The name of the dataset."})
-    filename: str = field(
-        metadata={"docstring": "The file name of the uploaded dataset."}
-    )
-    storage_uri: str = field(
-        metadata={
-            "docstring": "The URI at which the dataset is stored (eg. `s3://bucket/path/to/test.jsonl`)."
-        }
-    )
-    version: int = field(metadata={"docstring": "The version of the dataset."})
-    num_versions: int = field(
-        metadata={"docstring": "Number of versions of the dataset."}
-    )
-    created_at: datetime = field(
-        metadata={"docstring": "The time at which the dataset was uploaded."}
-    )
-    creator_id: str = field(
-        metadata={"docstring": "The ID of the Anyscale user who uploaded the dataset."}
-    )
-    project_id: str = field(
-        metadata={
-            "docstring": "The ID of the Anyscale project that the dataset belongs to."
-        }
-    )
-    cloud_id: str = field(
-        metadata={
-            "docstring": "The ID of the Anyscale cloud that the dataset belongs to."
-        }
-    )
-    description: Optional[str] = field(
-        default=None,
-        metadata={"docstring": "The description of the current dataset version."},
-    )
-
-    @classmethod
-    def parse_from_internal_model(cls, internal_model: InternalDataset) -> "Dataset":
-        return Dataset(
-            id=internal_model.id,
-            name=internal_model.name,
-            filename=internal_model.filename,
-            storage_uri=internal_model.storage_uri,
-            version=internal_model.version,
-            num_versions=internal_model.num_versions,
-            created_at=internal_model.created_at,
-            creator_id=internal_model.creator_id,
-            project_id=internal_model.project_id,
-            cloud_id=internal_model.cloud_id,
-            description=internal_model.description,
-        )