langsmith.schemas.TracerSessionResult

class langsmith.schemas.TracerSessionResult

A project, hydrated with additional information.

Sessions are also referred to as “Projects” in the UI.

Initialize a Run object.

Fields:

• completion_cost (decimal.Decimal | None)

• completion_tokens (int | None)

• error_rate (float | None)

• feedback_stats (Dict[str, Any] | None)

• first_token_p50 (datetime.timedelta | None)

• first_token_p99 (datetime.timedelta | None)

• last_run_start_time (datetime.datetime | None)

• latency_p50 (datetime.timedelta | None)

• latency_p99 (datetime.timedelta | None)

• prompt_cost (decimal.Decimal | None)

• prompt_tokens (int | None)

• run_count (int | None)

• run_facets (List[Dict[str, Any]] | None)

• total_cost (decimal.Decimal | None)

• total_tokens (int | None)

param completion_cost: Decimal | None = None

The estimated cost associated with the completion tokens.

param completion_tokens: int | None = None

The total number of completion tokens consumed in the project.

param error_rate: float | None = None

The error rate for the project.

param feedback_stats: Dict[str, Any] | None = None

Feedback stats for the project.

param first_token_p50: timedelta | None = None

The median (50th percentile) time to process the first token.

param first_token_p99: timedelta | None = None

The 99th percentile time to process the first token.

param last_run_start_time: datetime | None = None

The start time of the last run in the project.

param latency_p50: timedelta | None = None

The median (50th percentile) latency for the project.

param latency_p99: timedelta | None = None

The 99th percentile latency for the project.

param prompt_cost: Decimal | None = None

The estimated cost associated with the prompt (input) tokens.

param prompt_tokens: int | None = None

The total number of prompt tokens consumed in the project.

param run_count: int | None = None

The number of runs in the project.

param run_facets: List[Dict[str, Any]] | None = None

Facets for the runs in the project.

param total_cost: Decimal | None = None

The total estimated LLM cost associated with the completion tokens.

param total_tokens: int | None = None

The total number of tokens consumed in the project.