together 1.5.17__py3-none-any.whl → 2.0.0a8__py3-none-any.whl

together/types/dedicated_endpoint.py

@@ -0,0 +1,44 @@
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from datetime import datetime
+from typing_extensions import Literal
+
+from .._models import BaseModel
+from .autoscaling import Autoscaling
+
+__all__ = ["DedicatedEndpoint"]
+
+
+class DedicatedEndpoint(BaseModel):
+    id: str
+    """Unique identifier for the endpoint"""
+
+    autoscaling: Autoscaling
+    """Configuration for automatic scaling of the endpoint"""
+
+    created_at: datetime
+    """Timestamp when the endpoint was created"""
+
+    display_name: str
+    """Human-readable name for the endpoint"""
+
+    hardware: str
+    """The hardware configuration used for this endpoint"""
+
+    model: str
+    """The model deployed on this endpoint"""
+
+    name: str
+    """System name for the endpoint"""
+
+    object: Literal["endpoint"]
+    """The type of object"""
+
+    owner: str
+    """The owner of this endpoint"""
+
+    state: Literal["PENDING", "STARTING", "STARTED", "STOPPING", "STOPPED", "ERROR"]
+    """Current state of the endpoint"""
+
+    type: Literal["dedicated"]
+    """The type of endpoint"""

together/types/embedding.py

@@ -0,0 +1,24 @@
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from typing import List
+from typing_extensions import Literal
+
+from .._models import BaseModel
+
+__all__ = ["Embedding", "Data"]
+
+
+class Data(BaseModel):
+    embedding: List[float]
+
+    index: int
+
+    object: Literal["embedding"]
+
+
+class Embedding(BaseModel):
+    data: List[Data]
+
+    model: str
+
+    object: Literal["list"]

together/types/embedding_create_params.py

@@ -0,0 +1,31 @@
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from __future__ import annotations
+
+from typing import Union
+from typing_extensions import Literal, Required, TypedDict
+
+from .._types import SequenceNotStr
+
+__all__ = ["EmbeddingCreateParams"]
+
+
+class EmbeddingCreateParams(TypedDict, total=False):
+    input: Required[Union[str, SequenceNotStr[str]]]
+    """A string providing the text for the model to embed."""
+
+    model: Required[
+        Union[
+            Literal[
+                "WhereIsAI/UAE-Large-V1",
+                "BAAI/bge-large-en-v1.5",
+                "BAAI/bge-base-en-v1.5",
+                "togethercomputer/m2-bert-80M-8k-retrieval",
+            ],
+            str,
+        ]
+    ]
+    """The name of the embedding model to use.
+
+    [See all of Together AI's embedding models](https://docs.together.ai/docs/serverless-models#embedding-models)
+    """

together/types/endpoint_create_params.py

@@ -0,0 +1,43 @@
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from __future__ import annotations
+
+from typing import Optional
+from typing_extensions import Literal, Required, TypedDict
+
+from .autoscaling_param import AutoscalingParam
+
+__all__ = ["EndpointCreateParams"]
+
+
+class EndpointCreateParams(TypedDict, total=False):
+    autoscaling: Required[AutoscalingParam]
+    """Configuration for automatic scaling of the endpoint"""
+
+    hardware: Required[str]
+    """The hardware configuration to use for this endpoint"""
+
+    model: Required[str]
+    """The model to deploy on this endpoint"""
+
+    availability_zone: str
+    """Create the endpoint in a specified availability zone (e.g., us-central-4b)"""
+
+    disable_prompt_cache: bool
+    """Whether to disable the prompt cache for this endpoint"""
+
+    disable_speculative_decoding: bool
+    """Whether to disable speculative decoding for this endpoint"""
+
+    display_name: str
+    """A human-readable name for the endpoint"""
+
+    inactive_timeout: Optional[int]
+    """
+    The number of minutes of inactivity after which the endpoint will be
+    automatically stopped. Set to null, omit or set to 0 to disable automatic
+    timeout.
+    """
+
+    state: Literal["STARTED", "STOPPED"]
+    """The desired state of the endpoint"""

together/types/endpoint_list_avzones_response.py

@@ -0,0 +1,11 @@
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from typing import List
+
+from .._models import BaseModel
+
+__all__ = ["EndpointListAvzonesResponse"]
+
+
+class EndpointListAvzonesResponse(BaseModel):
+    avzones: List[str]

together/types/endpoint_list_params.py

@@ -0,0 +1,18 @@
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from __future__ import annotations
+
+from typing_extensions import Literal, TypedDict
+
+__all__ = ["EndpointListParams"]
+
+
+class EndpointListParams(TypedDict, total=False):
+    mine: bool
+    """If true, return only endpoints owned by the caller"""
+
+    type: Literal["dedicated", "serverless"]
+    """Filter endpoints by type"""
+
+    usage_type: Literal["on-demand", "reserved"]
+    """Filter endpoints by usage type"""

together/types/endpoint_list_response.py

@@ -0,0 +1,41 @@
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from typing import List
+from datetime import datetime
+from typing_extensions import Literal
+
+from .._models import BaseModel
+
+__all__ = ["EndpointListResponse", "Data"]
+
+
+class Data(BaseModel):
+    id: str
+    """Unique identifier for the endpoint"""
+
+    created_at: datetime
+    """Timestamp when the endpoint was created"""
+
+    model: str
+    """The model deployed on this endpoint"""
+
+    name: str
+    """System name for the endpoint"""
+
+    object: Literal["endpoint"]
+    """The type of object"""
+
+    owner: str
+    """The owner of this endpoint"""
+
+    state: Literal["PENDING", "STARTING", "STARTED", "STOPPING", "STOPPED", "ERROR"]
+    """Current state of the endpoint"""
+
+    type: Literal["serverless", "dedicated"]
+    """The type of endpoint"""
+
+
+class EndpointListResponse(BaseModel):
+    data: List[Data]
+
+    object: Literal["list"]

together/types/endpoint_update_params.py

@@ -0,0 +1,27 @@
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from __future__ import annotations
+
+from typing import Optional
+from typing_extensions import Literal, TypedDict
+
+from .autoscaling_param import AutoscalingParam
+
+__all__ = ["EndpointUpdateParams"]
+
+
+class EndpointUpdateParams(TypedDict, total=False):
+    autoscaling: AutoscalingParam
+    """New autoscaling configuration for the endpoint"""
+
+    display_name: str
+    """A human-readable name for the endpoint"""
+
+    inactive_timeout: Optional[int]
+    """
+    The number of minutes of inactivity after which the endpoint will be
+    automatically stopped. Set to 0 to disable automatic timeout.
+    """
+
+    state: Literal["STARTED", "STOPPED"]
+    """The desired state of the endpoint"""

together/types/eval_create_params.py

@@ -0,0 +1,263 @@
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from __future__ import annotations
+
+from typing import Union
+from typing_extensions import Literal, Required, TypeAlias, TypedDict
+
+from .._types import SequenceNotStr
+
+__all__ = [
+    "EvalCreateParams",
+    "Parameters",
+    "ParametersEvaluationClassifyParameters",
+    "ParametersEvaluationClassifyParametersJudge",
+    "ParametersEvaluationClassifyParametersModelToEvaluate",
+    "ParametersEvaluationClassifyParametersModelToEvaluateEvaluationModelRequest",
+    "ParametersEvaluationScoreParameters",
+    "ParametersEvaluationScoreParametersJudge",
+    "ParametersEvaluationScoreParametersModelToEvaluate",
+    "ParametersEvaluationScoreParametersModelToEvaluateEvaluationModelRequest",
+    "ParametersEvaluationCompareParameters",
+    "ParametersEvaluationCompareParametersJudge",
+    "ParametersEvaluationCompareParametersModelA",
+    "ParametersEvaluationCompareParametersModelAEvaluationModelRequest",
+    "ParametersEvaluationCompareParametersModelB",
+    "ParametersEvaluationCompareParametersModelBEvaluationModelRequest",
+]
+
+
+class EvalCreateParams(TypedDict, total=False):
+    parameters: Required[Parameters]
+    """Type-specific parameters for the evaluation"""
+
+    type: Required[Literal["classify", "score", "compare"]]
+    """The type of evaluation to perform"""
+
+
+class ParametersEvaluationClassifyParametersJudge(TypedDict, total=False):
+    model: Required[str]
+    """Name of the judge model"""
+
+    model_source: Required[Literal["serverless", "dedicated", "external"]]
+    """Source of the judge model."""
+
+    system_template: Required[str]
+    """System prompt template for the judge"""
+
+    external_api_token: str
+    """Bearer/API token for external judge models."""
+
+    external_base_url: str
+    """Base URL for external judge models. Must be OpenAI-compatible base URL."""
+
+
+class ParametersEvaluationClassifyParametersModelToEvaluateEvaluationModelRequest(TypedDict, total=False):
+    input_template: Required[str]
+    """Input prompt template"""
+
+    max_tokens: Required[int]
+    """Maximum number of tokens to generate"""
+
+    model: Required[str]
+    """Name of the model to evaluate"""
+
+    model_source: Required[Literal["serverless", "dedicated", "external"]]
+    """Source of the model."""
+
+    system_template: Required[str]
+    """System prompt template"""
+
+    temperature: Required[float]
+    """Sampling temperature"""
+
+    external_api_token: str
+    """Bearer/API token for external models."""
+
+    external_base_url: str
+    """Base URL for external models. Must be OpenAI-compatible base URL"""
+
+
+ParametersEvaluationClassifyParametersModelToEvaluate: TypeAlias = Union[
+    str, ParametersEvaluationClassifyParametersModelToEvaluateEvaluationModelRequest
+]
+
+
+class ParametersEvaluationClassifyParameters(TypedDict, total=False):
+    input_data_file_path: Required[str]
+    """Data file ID"""
+
+    judge: Required[ParametersEvaluationClassifyParametersJudge]
+
+    labels: Required[SequenceNotStr[str]]
+    """List of possible classification labels"""
+
+    pass_labels: Required[SequenceNotStr[str]]
+    """List of labels that are considered passing"""
+
+    model_to_evaluate: ParametersEvaluationClassifyParametersModelToEvaluate
+    """Field name in the input data"""
+
+
+class ParametersEvaluationScoreParametersJudge(TypedDict, total=False):
+    model: Required[str]
+    """Name of the judge model"""
+
+    model_source: Required[Literal["serverless", "dedicated", "external"]]
+    """Source of the judge model."""
+
+    system_template: Required[str]
+    """System prompt template for the judge"""
+
+    external_api_token: str
+    """Bearer/API token for external judge models."""
+
+    external_base_url: str
+    """Base URL for external judge models. Must be OpenAI-compatible base URL."""
+
+
+class ParametersEvaluationScoreParametersModelToEvaluateEvaluationModelRequest(TypedDict, total=False):
+    input_template: Required[str]
+    """Input prompt template"""
+
+    max_tokens: Required[int]
+    """Maximum number of tokens to generate"""
+
+    model: Required[str]
+    """Name of the model to evaluate"""
+
+    model_source: Required[Literal["serverless", "dedicated", "external"]]
+    """Source of the model."""
+
+    system_template: Required[str]
+    """System prompt template"""
+
+    temperature: Required[float]
+    """Sampling temperature"""
+
+    external_api_token: str
+    """Bearer/API token for external models."""
+
+    external_base_url: str
+    """Base URL for external models. Must be OpenAI-compatible base URL"""
+
+
+ParametersEvaluationScoreParametersModelToEvaluate: TypeAlias = Union[
+    str, ParametersEvaluationScoreParametersModelToEvaluateEvaluationModelRequest
+]
+
+
+class ParametersEvaluationScoreParameters(TypedDict, total=False):
+    input_data_file_path: Required[str]
+    """Data file ID"""
+
+    judge: Required[ParametersEvaluationScoreParametersJudge]
+
+    max_score: Required[float]
+    """Maximum possible score"""
+
+    min_score: Required[float]
+    """Minimum possible score"""
+
+    pass_threshold: Required[float]
+    """Score threshold for passing"""
+
+    model_to_evaluate: ParametersEvaluationScoreParametersModelToEvaluate
+    """Field name in the input data"""
+
+
+class ParametersEvaluationCompareParametersJudge(TypedDict, total=False):
+    model: Required[str]
+    """Name of the judge model"""
+
+    model_source: Required[Literal["serverless", "dedicated", "external"]]
+    """Source of the judge model."""
+
+    system_template: Required[str]
+    """System prompt template for the judge"""
+
+    external_api_token: str
+    """Bearer/API token for external judge models."""
+
+    external_base_url: str
+    """Base URL for external judge models. Must be OpenAI-compatible base URL."""
+
+
+class ParametersEvaluationCompareParametersModelAEvaluationModelRequest(TypedDict, total=False):
+    input_template: Required[str]
+    """Input prompt template"""
+
+    max_tokens: Required[int]
+    """Maximum number of tokens to generate"""
+
+    model: Required[str]
+    """Name of the model to evaluate"""
+
+    model_source: Required[Literal["serverless", "dedicated", "external"]]
+    """Source of the model."""
+
+    system_template: Required[str]
+    """System prompt template"""
+
+    temperature: Required[float]
+    """Sampling temperature"""
+
+    external_api_token: str
+    """Bearer/API token for external models."""
+
+    external_base_url: str
+    """Base URL for external models. Must be OpenAI-compatible base URL"""
+
+
+ParametersEvaluationCompareParametersModelA: TypeAlias = Union[
+    str, ParametersEvaluationCompareParametersModelAEvaluationModelRequest
+]
+
+
+class ParametersEvaluationCompareParametersModelBEvaluationModelRequest(TypedDict, total=False):
+    input_template: Required[str]
+    """Input prompt template"""
+
+    max_tokens: Required[int]
+    """Maximum number of tokens to generate"""
+
+    model: Required[str]
+    """Name of the model to evaluate"""
+
+    model_source: Required[Literal["serverless", "dedicated", "external"]]
+    """Source of the model."""
+
+    system_template: Required[str]
+    """System prompt template"""
+
+    temperature: Required[float]
+    """Sampling temperature"""
+
+    external_api_token: str
+    """Bearer/API token for external models."""
+
+    external_base_url: str
+    """Base URL for external models. Must be OpenAI-compatible base URL"""
+
+
+ParametersEvaluationCompareParametersModelB: TypeAlias = Union[
+    str, ParametersEvaluationCompareParametersModelBEvaluationModelRequest
+]
+
+
+class ParametersEvaluationCompareParameters(TypedDict, total=False):
+    input_data_file_path: Required[str]
+    """Data file name"""
+
+    judge: Required[ParametersEvaluationCompareParametersJudge]
+
+    model_a: ParametersEvaluationCompareParametersModelA
+    """Field name in the input data"""
+
+    model_b: ParametersEvaluationCompareParametersModelB
+    """Field name in the input data"""
+
+
+Parameters: TypeAlias = Union[
+    ParametersEvaluationClassifyParameters, ParametersEvaluationScoreParameters, ParametersEvaluationCompareParameters
+]

together/types/eval_create_response.py

@@ -0,0 +1,16 @@
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from typing import Optional
+from typing_extensions import Literal
+
+from .._models import BaseModel
+
+__all__ = ["EvalCreateResponse"]
+
+
+class EvalCreateResponse(BaseModel):
+    status: Optional[Literal["pending"]] = None
+    """Initial status of the job"""
+
+    workflow_id: Optional[str] = None
+    """The ID of the created evaluation job"""

together/types/eval_list_params.py

@@ -0,0 +1,21 @@
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from __future__ import annotations
+
+from typing_extensions import Annotated, TypedDict
+
+from .._utils import PropertyInfo
+
+__all__ = ["EvalListParams"]
+
+
+class EvalListParams(TypedDict, total=False):
+    limit: int
+
+    status: str
+
+    user_id: Annotated[str, PropertyInfo(alias="userId")]
+    """Admin users can specify a user ID to filter jobs.
+
+    Pass empty string to get all jobs.
+    """

together/types/eval_list_response.py

@@ -0,0 +1,10 @@
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from typing import List
+from typing_extensions import TypeAlias
+
+from .evaluation_job import EvaluationJob
+
+__all__ = ["EvalListResponse"]
+
+EvalListResponse: TypeAlias = List[EvaluationJob]

together/types/eval_status_response.py

@@ -0,0 +1,100 @@
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from typing import Union, Optional
+from typing_extensions import Literal, TypeAlias
+
+from pydantic import Field as FieldInfo
+
+from .._models import BaseModel
+
+__all__ = [
+    "EvalStatusResponse",
+    "Results",
+    "ResultsEvaluationClassifyResults",
+    "ResultsEvaluationScoreResults",
+    "ResultsEvaluationScoreResultsAggregatedScores",
+    "ResultsEvaluationCompareResults",
+]
+
+
+class ResultsEvaluationClassifyResults(BaseModel):
+    generation_fail_count: Optional[float] = None
+    """Number of failed generations."""
+
+    invalid_label_count: Optional[float] = None
+    """Number of invalid labels"""
+
+    judge_fail_count: Optional[float] = None
+    """Number of failed judge generations"""
+
+    label_counts: Optional[str] = None
+    """JSON string representing label counts"""
+
+    pass_percentage: Optional[float] = None
+    """Pecentage of pass labels."""
+
+    result_file_id: Optional[str] = None
+    """Data File ID"""
+
+
+class ResultsEvaluationScoreResultsAggregatedScores(BaseModel):
+    mean_score: Optional[float] = None
+
+    pass_percentage: Optional[float] = None
+
+    std_score: Optional[float] = None
+
+
+class ResultsEvaluationScoreResults(BaseModel):
+    aggregated_scores: Optional[ResultsEvaluationScoreResultsAggregatedScores] = None
+
+    failed_samples: Optional[float] = None
+    """number of failed samples generated from model"""
+
+    generation_fail_count: Optional[float] = None
+    """Number of failed generations."""
+
+    invalid_score_count: Optional[float] = None
+    """number of invalid scores generated from model"""
+
+    judge_fail_count: Optional[float] = None
+    """Number of failed judge generations"""
+
+    result_file_id: Optional[str] = None
+    """Data File ID"""
+
+
+class ResultsEvaluationCompareResults(BaseModel):
+    a_wins: Optional[int] = FieldInfo(alias="A_wins", default=None)
+    """Number of times model A won"""
+
+    b_wins: Optional[int] = FieldInfo(alias="B_wins", default=None)
+    """Number of times model B won"""
+
+    generation_fail_count: Optional[float] = None
+    """Number of failed generations."""
+
+    judge_fail_count: Optional[float] = None
+    """Number of failed judge generations"""
+
+    num_samples: Optional[int] = None
+    """Total number of samples compared"""
+
+    result_file_id: Optional[str] = None
+    """Data File ID"""
+
+    ties: Optional[int] = FieldInfo(alias="Ties", default=None)
+    """Number of ties"""
+
+
+Results: TypeAlias = Union[
+    ResultsEvaluationClassifyResults, ResultsEvaluationScoreResults, ResultsEvaluationCompareResults
+]
+
+
+class EvalStatusResponse(BaseModel):
+    results: Optional[Results] = None
+    """The results of the evaluation job"""
+
+    status: Optional[Literal["completed", "error", "user_error", "running", "queued", "pending"]] = None
+    """The status of the evaluation job"""