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

anyscale/llm/dataset/_private/sdk.py

@@ -1,147 +0,0 @@
-from typing import Optional
-
-from anyscale._private.models.model_base import ListResponse
-from anyscale._private.sdk import sdk_command_v2
-from anyscale._private.sdk.base_sdk import BaseSDK
-from anyscale.llm.dataset._private import docs
-from anyscale.llm.dataset._private.models import Dataset
-
-
-@sdk_command_v2(
-    doc_py_example=docs.GET_PY_EXAMPLE, arg_docstrings=docs.GET_PY_ARG_DOCSTRINGS,
-)
-def get(
-    name: str, version: Optional[int] = None, project: Optional[str] = None
-) -> Dataset:
-    """Retrieves metadata about a dataset.
-
-    :param name: Name of the dataset.
-    :param 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.
-    :param project: Name of the Anyscale project that the dataset belongs to. If not provided, all projects will be searched.
-
-    Example usage:
-    ```python
-    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)
-    ```
-
-    Return:
-        Dataset: The `Dataset` object.
-    """
-    _sdk = BaseSDK()
-    dataset = _sdk.client.get_dataset(name, version, project)
-    return dataset
-
-
-@sdk_command_v2(
-    doc_py_example=docs.UPLOAD_PY_EXAMPLE, arg_docstrings=docs.UPLOAD_PY_ARG_DOCSTRINGS,
-)
-def upload(
-    dataset_file: str,
-    name: str,
-    *,
-    description: Optional[str] = None,
-    cloud: Optional[str] = None,
-    project: Optional[str] = None,
-) -> Dataset:
-    """Uploads a dataset, or a new version of a dataset, to your Anyscale cloud.
-
-    :param dataset_file: Path to the dataset file to upload.
-    :param name: Name of a new dataset, or an existing dataset, to upload a new version of.
-    :param description: Description of the dataset version.
-    :param cloud: Name of the Anyscale cloud to upload a new dataset to. If not provided, the default cloud will be used.
-    :param project: Name of the Anyscale project to upload a new dataset to. If not provided, the default project of the cloud will be used.
-
-    Example usage:
-    ```python
-    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")
-    ```
-    Return:
-        Dataset: The `Dataset` object representing the uploaded dataset.
-
-    NOTE:
-    If you are uploading a new version, have run this from within an Anyscale workspace,
-    and neither `cloud` nor `project` are provided, the cloud and project of the workspace will be used.
-    """
-    _sdk = BaseSDK()
-    dataset = _sdk.client.upload_dataset(
-        dataset_file, name, description, cloud, project,
-    )
-    return dataset
-
-
-@sdk_command_v2(
-    doc_py_example=docs.DOWNLOAD_PY_EXAMPLE,
-    arg_docstrings=docs.DOWNLOAD_PY_ARG_DOCSTRINGS,
-)
-def download(
-    name: str, version: Optional[int] = None, project: Optional[str] = None
-) -> bytes:
-    """Downloads a dataset from your Anyscale cloud.
-
-    :param name: Name of the dataset to download.
-    :param 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.
-    :param project: Name of the Anyscale project to download the dataset from. If not provided, all projects will be searched.
-
-    Example usage:
-    ```python
-    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)
-    ```
-
-    Returns:
-        bytes: The contents of the dataset file.
-    """
-    _sdk = BaseSDK()
-    dataset_bytes = _sdk.client.download_dataset(name, version, project)
-    return dataset_bytes
-
-
-@sdk_command_v2(
-    doc_py_example=docs.LIST_PY_EXAMPLE, arg_docstrings=docs.LIST_PY_ARG_DOCSTRINGS
-)
-def list(  # noqa: A001
-    *,
-    # Pagination
-    limit: Optional[int] = None,
-    after: Optional[str] = None,  # Unique ID to start listing after
-    # Filtering
-    name_contains: Optional[str] = None,  # Substring in name, case insensitive
-    cloud: Optional[str] = None,
-    project: Optional[str] = None,
-) -> ListResponse[Dataset]:
-    """
-    Lists datasets.
-
-    :param limit: Maximum number of datasets to return. Default: 1000.
-    :param after: ID of the dataset to start the listing from. If provided, the list will start from the dataset after this ID.
-    :param name_contains: Filter datasets by name. If provided, only datasets with name containing this string will be returned.
-    :param cloud: Name of the Anyscale cloud to search in. If not provided, all clouds will be searched.
-    :param project: Name of the Anyscale project to search in. If not provided, all projects will be searched.
-
-    Example usage:
-    ```
-    datasets = anyscale.llm.dataset.list(limit=10)
-    for d in datasets:
-        print(f"Dataset name: '{d.name}'")  # Prints 10 dataset names
-    ```
-
-    Returns:
-        ListResponse[Dataset]: List of `Dataset` objects.
-    """
-    _sdk = BaseSDK()
-    list_response = _sdk.client.list_datasets(
-        limit=limit,
-        after=after,
-        name_contains=name_contains,
-        cloud=cloud,
-        project=project,
-    )
-    return list_response

anyscale/llm/model/__init__.py

@@ -1,2 +0,0 @@
-from anyscale.llm.model.commands import delete, get, list
-from anyscale.llm.model.sdk import LLMModelsSDK

anyscale/llm/model/_private/models_sdk.py

@@ -1,62 +0,0 @@
-from typing import Optional
-
-from anyscale._private.anyscale_client import AnyscaleClientInterface
-from anyscale._private.sdk.base_sdk import BaseSDK
-from anyscale._private.sdk.timer import Timer
-from anyscale.cli_logger import BlockLogger
-from anyscale.client.openapi_client.models import FineTunedModel as APIFineTunedModel
-from anyscale.llm.model.models import DeletedFineTunedModel, FineTunedModel
-
-
-class PrivateLLMModelsSDK(BaseSDK):
-    def __init__(
-        self,
-        *,
-        logger: Optional[BlockLogger] = None,
-        client: Optional[AnyscaleClientInterface] = None,
-        timer: Optional[Timer] = None,
-    ):
-        super().__init__(logger=logger, client=client, timer=timer)
-
-    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 list(
-        self,
-        *,
-        cloud_id: Optional[str] = None,
-        project_id: Optional[str] = None,
-        max_items: int = 20,
-    ):
-        finetuned_models = self.client.list_finetuned_models(
-            cloud_id, project_id, max_items
-        )
-        parsed_models = [
-            self._parse_response_model_get(model) for model in finetuned_models
-        ]
-        return parsed_models
-
-    def get(
-        self, *, model_id: Optional[str] = None, job_id: Optional[str] = None
-    ) -> FineTunedModel:
-        model = self.client.get_finetuned_model(model_id, job_id)
-        return self._parse_response_model_get(model)
-
-    def delete(self, model_id) -> DeletedFineTunedModel:
-        deleted_model = self.client.delete_finetuned_model(model_id).to_dict()
-        deleted_model["deleted_at"] = int(deleted_model["deleted_at"].timestamp())
-        return DeletedFineTunedModel(**deleted_model)

anyscale/llm/model/commands.py

@@ -1,93 +0,0 @@
-from typing import List, Optional
-
-from anyscale._private.sdk import sdk_command
-from anyscale.llm.model._private.models_sdk import PrivateLLMModelsSDK
-from anyscale.llm.model.models import DeletedFineTunedModel, FineTunedModel
-
-
-_LLM_MODELS_SDK_SINGLETON_KEY = "llm_models_sdk"
-
-
-_RETRIEVE_EXAMPLE = """
-import anyscale
-
-anyscale.llm.model.get(model_id="my-model-id")
-anyscale.llm.model.get(job_ib="prodjob_123")
-"""
-
-_RETRIEVE_ARG_DOCSTRINGS = {
-    "model_id": " ID of the finetuned model that is being retrieved.",
-    "job_id": " ID of the Anyscale job corresponding to the fine-tuning run.",
-}
-
-
-_DELETE_EXAMPLE = """
-import anyscale
-
-anyscale.llm.model.delete(model_id="my-model-id")
-"""
-
-_DELETE_ARG_DOCSTRINGS = {"model_id": " ID of the finetuned model to delete."}
-
-_LIST_EXAMPLE = """
-import anyscale
-
-anyscale.llm.model.list()
-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")
-anyscale.llm.model.list(project_id="prj_123", max_items=10)
-"""
-
-_LIST_ARG_DOCSTRINGS = {
-    "cloud_id": "Cloud ID to filter by. If not specified, all models from all the clouds visible to the user (filtered optionally by `project_id`)  are listed.",
-    "project_id": "Project ID to filter by. If not specified, all the models from all visible projects to the user (filtered optionally by `cloud_id`) are listed.",
-    "max_items": "Maximum number of items to show in the list. By default, the 20 most recently created models are fetched.",
-}
-
-
-@sdk_command(
-    _LLM_MODELS_SDK_SINGLETON_KEY,
-    PrivateLLMModelsSDK,
-    doc_py_example=_RETRIEVE_EXAMPLE,
-    arg_docstrings=_RETRIEVE_ARG_DOCSTRINGS,
-)
-def get(
-    *,
-    model_id: Optional[str] = None,
-    job_id: Optional[str] = None,
-    _sdk: PrivateLLMModelsSDK,
-) -> FineTunedModel:
-    """Retrieves model card for a finetuned model."""
-    return _sdk.get(model_id=model_id, job_id=job_id)
-
-
-@sdk_command(
-    _LLM_MODELS_SDK_SINGLETON_KEY,
-    PrivateLLMModelsSDK,
-    doc_py_example=_DELETE_EXAMPLE,
-    arg_docstrings=_DELETE_ARG_DOCSTRINGS,
-)
-def delete(model_id: str, _sdk: PrivateLLMModelsSDK,) -> DeletedFineTunedModel:
-    """Deletes a finetuned model. Requires owner permission for the corresponding Anyscale project."""
-    return _sdk.delete(model_id)
-
-
-@sdk_command(
-    _LLM_MODELS_SDK_SINGLETON_KEY,
-    PrivateLLMModelsSDK,
-    doc_py_example=_LIST_EXAMPLE,
-    arg_docstrings=_LIST_ARG_DOCSTRINGS,
-)
-def list(  # noqa: A001
-    *,
-    cloud_id: Optional[str] = None,
-    project_id: Optional[str] = None,
-    max_items: int = 20,
-    _sdk: PrivateLLMModelsSDK,
-) -> List[FineTunedModel]:
-    """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`.
-    """
-    return _sdk.list(cloud_id=cloud_id, project_id=project_id, max_items=max_items)

anyscale/llm/model/models.py

@@ -1,171 +0,0 @@
-from dataclasses import dataclass, field
-from typing import Any, Dict, Optional
-
-from anyscale._private.models.model_base import ModelBase, ModelEnum
-
-
-class FineTuningType(ModelEnum):
-    LORA = "LORA"
-    FULL_PARAM = "FULL_PARAM"
-
-    __docstrings__ = {
-        LORA: "Low-Rank Adaptation (LoRA) fine-tuning method.",
-        FULL_PARAM: "Full parameter fine-tuning method.",
-    }
-
-
-@dataclass(frozen=True)
-class FineTunedModel(ModelBase):
-    """Represents a fine-tuned model with its associated metadata."""
-
-    __doc_py_example__ = """\
-import anyscale
-from anyscale.llm.model.models import FineTunedModel
-model: FineTunedModel = anyscale.llm.model.get(model_id="my-model-id")
-"""
-
-    __doc_cli_example__ = """\
-$ anyscale llm model get --model-id my-model-id
-id: my-model-id
-base_model_id: meta-llama/Llama-3-8B-Instruct
-cloud_id: cloud_abc123
-created_at: 1725473924
-job_id: prodjob_xyz789
-ft_type: LORA
-...
-"""
-
-    id: str = field(metadata={"docstring": "Unique ID/tag for the fine-tuned model."})
-
-    def _validate_id(self, id: str):  # noqa: A002
-        if not isinstance(id, str):
-            raise TypeError("'id' must be a string.")
-
-    base_model_id: str = field(
-        metadata={"docstring": "Base model ID used for fine-tuning."}
-    )
-
-    def _validate_base_model_id(self, base_model_id: str):
-        if not isinstance(base_model_id, str):
-            raise TypeError("'base_model_id' must be a string.")
-
-    cloud_id: str = field(
-        metadata={
-            "docstring": "ID for the Anyscale Cloud corresponding to the fine-tuning run."
-        }
-    )
-
-    def _validate_cloud_id(self, cloud_id: str):
-        if not isinstance(cloud_id, str):
-            raise TypeError("'cloud_id' must be a string.")
-
-    created_at: int = field(
-        metadata={"docstring": "Time at which the fine-tuned model was created."}
-    )
-
-    def _validate_created_at(self, created_at: int):
-        if not isinstance(created_at, int):
-            raise TypeError("'created_at' must be an integer (Unix timestamp).")
-
-        if created_at < 0:
-            raise ValueError("'created_at' must be a positive integer")
-
-    creator: Optional[str] = field(
-        metadata={"docstring": "Email address for the user who created the model."}
-    )
-
-    def _validate_creator(self, creator: Optional[str]):
-        if creator is not None and not isinstance(creator, str):
-            raise TypeError("'creator' must be a string or None.")
-
-    ft_type: FineTuningType = field(metadata={"docstring": "Fine-tuning type."})
-
-    def _validate_ft_type(self, ft_type: FineTuningType) -> FineTuningType:
-        return FineTuningType.validate(ft_type)
-
-    generation_config: Optional[Dict[str, Any]] = field(
-        metadata={
-            "docstring": "Inference generation config with chat-templating parameters and stopping sequences."
-        }
-    )
-
-    def _validate_generation_config(self, generation_config: Optional[Dict[str, Any]]):
-        if generation_config is not None and not isinstance(generation_config, dict):
-            raise TypeError("'generation_config' must be a dictionary or None.")
-
-    job_id: Optional[str] = field(
-        metadata={
-            "docstring": "ID for the Anyscale job corresponding to the fine-tuning run, if applicable."
-        }
-    )
-
-    def _validate_job_id(self, job_id: Optional[str]):
-        if job_id is not None and not isinstance(job_id, str):
-            raise TypeError("'job_id' must be a string or None.")
-
-    project_id: Optional[str] = field(
-        metadata={
-            "docstring": "ID for the Anyscale Project corresponding to the fine-tuning run."
-        }
-    )
-
-    def _validate_project_id(self, project_id: Optional[str]):
-        if project_id is not None and not isinstance(project_id, str):
-            raise TypeError("'project_id' must be a string.")
-
-    storage_uri: str = field(
-        metadata={
-            "docstring": "URI at which the fine-tuned model checkpoint is stored."
-        }
-    )
-
-    def _validate_storage_uri(self, storage_uri: str):
-        if not isinstance(storage_uri, str):
-            raise TypeError("'storage_uri' must be a string.")
-
-    workspace_id: Optional[str] = field(
-        default=None,
-        metadata={
-            "docstring": "ID for the Anyscale Workspace in which the model was fine-tuned, if any."
-        },
-    )
-
-    def _validate_workspace_id(self, workspace_id: Optional[str]):
-        if workspace_id is not None and not isinstance(workspace_id, str):
-            raise TypeError("'workspace_id' must be a string or None.")
-
-
-@dataclass(frozen=True)
-class DeletedFineTunedModel(ModelBase):
-    """Represents a deleted fine-tuned model with its deletion metadata."""
-
-    __doc_py_example__ = """\
-import anyscale
-from anyscale.llm.model.models import DeletedFineTunedModel
-deleted_model: DeletedFineTunedModel = anyscale.llm.model.delete("my-model-id")
-"""
-
-    __doc_cli_example__ = """\
-$ anyscale models get-deleted my-model-id
-id: my-model-id
-deleted_at: 1725473924
-"""
-
-    id: str = field(metadata={"docstring": "Unique ID/tag for the fine-tuned model."})
-
-    def _validate_id(self, id: str):  # noqa: A002
-        if not isinstance(id, str):
-            raise TypeError("'id' must be a string.")
-
-    deleted_at: int = field(
-        metadata={
-            "docstring": "Unix timestamp (in seconds) at which the fine-tuned model was deleted."
-        }
-    )
-
-    def _validate_deleted_at(self, deleted_at: int):
-        if not isinstance(deleted_at, int):
-            raise TypeError("'deleted_at' must be an integer (Unix timestamp).")
-
-        if deleted_at < 0:
-            raise ValueError("'deleted_at' must be a positive integer (Unix timestamp)")

anyscale/llm/model/sdk.py

@@ -1,62 +0,0 @@
-from typing import Any, Optional
-
-from anyscale._private.anyscale_client import AnyscaleClientInterface
-from anyscale._private.sdk import sdk_docs
-from anyscale._private.sdk.base_sdk import BaseSDK, Timer
-from anyscale.cli_logger import BlockLogger
-from anyscale.llm.model._private.models_sdk import PrivateLLMModelsSDK
-from anyscale.llm.model.commands import (
-    _DELETE_ARG_DOCSTRINGS,
-    _DELETE_EXAMPLE,
-    _LIST_ARG_DOCSTRINGS,
-    _LIST_EXAMPLE,
-    _RETRIEVE_ARG_DOCSTRINGS,
-    _RETRIEVE_EXAMPLE,
-)
-
-
-class LLMModelsSDK(BaseSDK):
-    def __init__(
-        self,
-        *,
-        logger: Optional[BlockLogger] = None,
-        client: Optional[AnyscaleClientInterface] = None,
-        timer: Optional[Timer] = None,
-    ):
-        self._private_sdk = PrivateLLMModelsSDK(
-            logger=logger, client=client, timer=timer
-        )
-
-    @sdk_docs(
-        doc_py_example=_RETRIEVE_EXAMPLE, arg_docstrings=_RETRIEVE_ARG_DOCSTRINGS,
-    )
-    def get(
-        self, *, model_id: Optional[str] = None, job_id: Optional[str] = None
-    ) -> Any:
-        """Retrives model card for a finetuned model."""
-        return self._private_sdk.get(model_id=model_id, job_id=job_id)
-
-    @sdk_docs(
-        doc_py_example=_DELETE_EXAMPLE, arg_docstrings=_DELETE_ARG_DOCSTRINGS,
-    )
-    def delete(self, model_id: str) -> Any:
-        """Deletes a finetuned model. Requires owner permission for the corresponding Anyscale project."""
-        return self._private_sdk.delete(model_id)
-
-    @sdk_docs(
-        doc_py_example=_LIST_EXAMPLE, arg_docstrings=_LIST_ARG_DOCSTRINGS,
-    )
-    def list(
-        self,
-        *,
-        cloud_id: Optional[str] = None,
-        project_id: Optional[str] = None,
-        max_items: int = 20,
-    ) -> Any:
-        """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`.
-        """
-        return self._private_sdk.list(
-            cloud_id=cloud_id, project_id=project_id, max_items=max_items
-        )

anyscale/llm/sdk.py

@@ -1,27 +0,0 @@
-from typing import Optional
-
-import anyscale
-from anyscale._private.anyscale_client import AnyscaleClientInterface
-from anyscale._private.sdk.base_sdk import BaseSDK
-from anyscale._private.sdk.timer import Timer
-from anyscale.cli_logger import BlockLogger
-from anyscale.llm.model import LLMModelsSDK
-
-
-class LLMSDK(BaseSDK):
-    def __init__(
-        self,
-        *,
-        logger: Optional[BlockLogger] = None,
-        client: Optional[AnyscaleClientInterface] = None,
-        timer: Optional[Timer] = None,
-    ):
-        self._model = LLMModelsSDK(logger=logger, client=client, timer=timer)
-
-    @property
-    def model(self):
-        return self._model
-
-    @property
-    def dataset(self):
-        return anyscale.llm.dataset