Python SDK for AI agent monitoring, LLM cost tracking, benchmarking, and more. Integrates with most LLMs and agent frameworks including CrewAI, Agno, OpenAI Agents SDK, Langchain, Autogen, AG2, and CamelAI
Python agentagentopsagents-sdkaianthropicautogencost-estimationcrewaievalsevaluation-metricsgroqlangchainllmmistralollamaopenaiopenai-agents
Prints $0.00 now when token_cost is zero (#168)
kate yeh56648c8874c8
diff --git a/agentops/client.py b/agentops/client.pyindex d6a4c14a..2aa90e00 100644--- a/agentops/client.py+++ b/agentops/client.py@@ -15,12 +15,14 @@ from uuid import uuid4from typing import Optional, Listimport tracebackfrom .log_config import logger, set_logging_level_info+from decimal import Decimalimport inspectimport atexitimport signalimport sysimport threading+from .meta_client import MetaClientfrom .config import Configuration, ConfigurationErrorfrom .llm_tracker import LlmTracker@@ -33,16 +35,16 @@ class Client(metaclass=MetaClient):Args:- api_key (str, optional): API Key for AgentOps services. If none is provided, key will+ api_key (str, optional): API Key for AgentOps services. If none is provided, key willbe read from the AGENTOPS_API_KEY environment variable.- parent_key (str, optional): Organization key to give visibility of all user sessions the user's organization. If none is provided, key will+ parent_key (str, optional): Organization key to give visibility of all user sessions the user's organization. If none is provided, key willbe read from the AGENTOPS_PARENT_KEY environment variable.- endpoint (str, optional): The endpoint for the AgentOps service. If none is provided, key will+ endpoint (str, optional): The endpoint for the AgentOps service. If none is provided, key willbe read from the AGENTOPS_API_ENDPOINT environment variable. Defaults to 'https://api.agentops.ai'.- max_wait_time (int, optional): The maximum time to wait in milliseconds before flushing the queue.+ max_wait_time (int, optional): The maximum time to wait in milliseconds before flushing the queue.Defaults to 30,000 (30 seconds)max_queue_size (int, optional): The maximum size of the event queue. Defaults to 100.- tags (List[str], optional): Tags for the sessions that can be used for grouping or+ tags (List[str], optional): Tags for the sessions that can be used for grouping orsorting later (e.g. ["GPT-4"]).override (bool, optional): [Deprecated] Use `instrument_llm_calls` instead. Whether to instrument LLM calls and emit LLMEvents..instrument_llm_calls (bool): Whether to instrument LLM calls and emit LLMEvents..@@ -95,7 +97,7 @@ class Client(metaclass=MetaClient):def add_tags(self, tags: List[str]):"""- Append to session tags at runtime.+ Append to session tags at runtime.Args:tags (List[str]): The list of tags to append.@@ -111,7 +113,7 @@ class Client(metaclass=MetaClient):def set_tags(self, tags: List[str]):"""- Replace session tags at runtime.+ Replace session tags at runtime.Args:tags (List[str]): The list of tags to set.@@ -264,11 +266,12 @@ class Client(metaclass=MetaClient):self._session.video = videoself._session.end_session(end_state, end_state_reason)- token_cost = self._worker.end_session(self._session)+ token_cost = Decimal(self._worker.end_session(self._session))if token_cost == 'unknown':print('🖇 AgentOps: Could not determine cost of run.')else:- print('🖇 AgentOps: This run cost ${:.6f}'.format(float(token_cost)))++ print('🖇 AgentOps: This run cost ${}'.format('{:.2f}'.format(token_cost) if token_cost == 0 else '{:.6f}'.format(token_cost)))self._session = Noneself._worker = None@@ -305,7 +308,7 @@ class Client(metaclass=MetaClient):Args:exc_type (Type[BaseException]): The type of the exception.exc_value (BaseException): The exception instance.- exc_traceback (TracebackType): A traceback object encapsulating the call stack at the+ exc_traceback (TracebackType): A traceback object encapsulating the call stack at thepoint where the exception originally occurred."""formatted_traceback = ''.join(traceback.format_exception(exc_type, exc_value,