Skip to content

Same agent Smaller bill. The tools and infrastructure to get more work from your AI. Fewer tokens. Less compute. Lower cost.

[Feat]: Use OpenTelemetry events for Content logging in Ollama and AI21 (#637)

Patcher6cb8b63e214d
diff --git a/docs/latest/features/tracing.mdx b/docs/latest/features/tracing.mdx
index 727d9c2b..8df856ae 100644
--- a/docs/latest/features/tracing.mdx
+++ b/docs/latest/features/tracing.mdx
@@ -57,7 +57,7 @@ However, you may want to disable this logging for privacy reasons, as they may c
Example:
```python
-openlit.init(trace_content=False)
+openlit.init(capture_message_content=False)
```
## Disable Batch
diff --git a/docs/latest/sdk-configuration.mdx b/docs/latest/sdk-configuration.mdx
index 29c6c5bd..b9eb0873 100644
--- a/docs/latest/sdk-configuration.mdx
+++ b/docs/latest/sdk-configuration.mdx
@@ -21,7 +21,7 @@ Below is a detailed overview of the configuration options available, allowing yo
| `otlp_endpoint` | Specifies the OTLP endpoint for transmitting telemetry data. | `None` | No |
| `otlp_headers` | Defines headers for the OTLP exporter, useful for backends requiring authentication. | `None` | No |
| `disable_batch` | A flag to disable batch span processing, favoring immediate dispatch. | `False` | No |
-| `trace_content` | Enables tracing of content for deeper insights. | `True` | No |
+| `capture_message_content` | Enables tracing of content for deeper insights. | `True` | No |
| `disabled_instrumentors`| List of instrumentors to disable. | `None` | No |
| `disable_metrics` | If set, disables the collection of metrics. | `False` | No |
| `pricing_json` | URL or file path of the pricing JSON file. | `https://github.com/openlit/openlit/blob/main/assets/pricing.json` | No |
diff --git a/sdk/python/.pylintrc b/sdk/python/.pylintrc
index 3bfcf9bd..91ba1442 100644
--- a/sdk/python/.pylintrc
+++ b/sdk/python/.pylintrc
@@ -7,7 +7,7 @@ max-locals=30
max-args=25
# Maximum number of positional arguments (see R0917).
-max-positional-arguments=12
+max-positional-arguments=15
# Maximum allowed line length
max-line-length=120
@@ -27,4 +27,5 @@ disable=
too-many-lines,
no-name-in-module,
protected-access,
- bare-except
\ No newline at end of file
+ bare-except,
+ expression-not-assigned
\ No newline at end of file
diff --git a/sdk/python/README.md b/sdk/python/README.md
index 761d322c..59803e03 100644
--- a/sdk/python/README.md
+++ b/sdk/python/README.md
@@ -187,7 +187,7 @@ Below is a detailed overview of the configuration options available, allowing yo
| `otlp_endpoint` | Specifies the OTLP endpoint for transmitting telemetry data. | `None` | No |
| `otlp_headers` | Defines headers for the OTLP exporter, useful for backends requiring authentication. | `None` | No |
| `disable_batch` | A flag to disable batch span processing, favoring immediate dispatch. | `False` | No |
-| `trace_content` | Enables tracing of content for deeper insights. | `True` | No |
+| `capture_message_content` | Enables tracing of content for deeper insights. | `True` | No |
| `disabled_instrumentors`| List of instrumentors to disable. | `None` | No |
| `disable_metrics` | If set, disables the collection of metrics. | `False` | No |
| `pricing_json` | URL or file path of the pricing JSON file. | `https://github.com/openlit/openlit/blob/main/assets/pricing.json` | No |
diff --git a/sdk/python/pyproject.toml b/sdk/python/pyproject.toml
index eea7ee41..e87cf8a4 100644
--- a/sdk/python/pyproject.toml
+++ b/sdk/python/pyproject.toml
@@ -1,6 +1,6 @@
[tool.poetry]
name = "openlit"
-version = "1.33.10"
+version = "1.33.11"
description = "OpenTelemetry-native Auto instrumentation library for monitoring LLM Applications and GPUs, facilitating the integration of observability into your GenAI-driven projects"
authors = ["OpenLIT"]
license = "Apache-2.0"
diff --git a/sdk/python/src/openlit/__helpers.py b/sdk/python/src/openlit/__helpers.py
index 5a7de73a..e287e8b8 100644
--- a/sdk/python/src/openlit/__helpers.py
+++ b/sdk/python/src/openlit/__helpers.py
@@ -11,6 +11,7 @@ import requests
import tiktoken
from opentelemetry.sdk.resources import SERVICE_NAME, TELEMETRY_SDK_NAME, DEPLOYMENT_ENVIRONMENT
from opentelemetry.trace import Status, StatusCode
+from opentelemetry._events import Event
from openlit.semcov import SemanticConvetion
# Set up logging
@@ -285,3 +286,75 @@ def set_server_address_and_port(client_instance: Any,
server_port = default_server_port
return server_address, server_port
+
+def otel_event(name, attributes, body):
+ """
+ Returns an OpenTelemetry Event object
+ """
+
+ return Event(
+ name=name,
+ attributes=attributes,
+ body=body,
+ )
+
+def extract_and_format_input(messages):
+ """
+ Process a list of messages to extract content and categorize
+ them into fixed roles like 'user', 'assistant', 'system'.
+ """
+
+ fixed_roles = ['user', 'assistant', 'system', 'tool'] # Ensure these are your fixed keys
+ # Initialize the dictionary with fixed keys and empty structures
+ formatted_messages = {role_key: {"role": "", "content": ""} for role_key in fixed_roles}
+
+ for message in messages:
+ # Normalize the message structure
+ message = response_as_dict(message)
+
+ # Extract role and content
+ role = message.get("role")
+ if role not in fixed_roles:
+ continue # Skip any role not in our predefined roles
+
+ content = message.get("content", "")
+
+ # Prepare content as a string
+ if isinstance(content, list):
+ content_str = ", ".join(
+ # pylint: disable=line-too-long
+ f'{item.get("type", "text")}: {item.get("text", item.get("image_url", "").get("url", "") if isinstance(item.get("image_url", ""), dict) else item.get("image_url", ""))}'
+ for item in content
+ )
+ else:
+ content_str = content
+
+ # Set the role in the formatted message and concatenate content
+ if not formatted_messages[role]["role"]:
+ formatted_messages[role]["role"] = role
+
+ if formatted_messages[role]["content"]:
+ formatted_messages[role]["content"] += " " + content_str
+ else:
+ formatted_messages[role]["content"] = content_str
+
+ return formatted_messages
+
+# To be removed one the change to log events (from span events) is complete
+def concatenate_all_contents(formatted_messages):
+ """
+ Concatenate all 'content' fields from the formatted messages
+ dictionary into a single string.
+
+ Parameters:
+ - formatted_messages: Dictionary with roles as keys and corresponding
+ role and content as values.
+
+ Returns:
+ - A single string with all content concatenated.
+ """
+ return " ".join(
+ message_data['content']
+ for message_data in formatted_messages.values()
+ if message_data['content']
+ )
diff --git a/sdk/python/src/openlit/__init__.py b/sdk/python/src/openlit/__init__.py
index 4a8850e2..4da6f21c 100644
--- a/sdk/python/src/openlit/__init__.py
+++ b/sdk/python/src/openlit/__init__.py
@@ -21,9 +21,9 @@ from opentelemetry.sdk.resources import SERVICE_NAME, DEPLOYMENT_ENVIRONMENT
from openlit.semcov import SemanticConvetion
from openlit.otel.tracing import setup_tracing
from openlit.otel.metrics import setup_meter
+from openlit.otel.events import setup_events
from openlit.__helpers import fetch_pricing_info, get_env_variable
-
# Instrumentors for various large language models.
from openlit.instrumentation.openai import OpenAIInstrumentor
from openlit.instrumentation.anthropic import AnthropicInstrumentor
@@ -85,10 +85,11 @@ class OpenlitConfig:
application_name (str): Name of the application using openLIT.
pricing_info (Dict[str, Any]): Pricing information.
tracer (Optional[Any]): Tracer instance for OpenTelemetry.
+ event_provider (Optional[Any]): Event logger provider for OpenTelemetry.
otlp_endpoint (Optional[str]): Endpoint for OTLP.
otlp_headers (Optional[Dict[str, str]]): Headers for OTLP.
disable_batch (bool): Flag to disable batch span processing in tracing.
- trace_content (bool): Flag to enable or disable tracing of content.
+ capture_message_content (bool): Flag to enable or disable tracing of content.
"""
_instance = None
@@ -107,11 +108,12 @@ class OpenlitConfig:
cls.application_name = "default"
cls.pricing_info = {}
cls.tracer = None
+ cls.event_provider = None
cls.metrics_dict = {}
cls.otlp_endpoint = None
cls.otlp_headers = None
cls.disable_batch = False
- cls.trace_content = True
+ cls.capture_message_content = True
cls.disable_metrics = False
@classmethod
@@ -120,10 +122,11 @@ class OpenlitConfig:
environment,
application_name,
tracer,
+ event_provider,
otlp_endpoint,
otlp_headers,
disable_batch,
- trace_content,
+ capture_message_content,
metrics_dict,
disable_metrics,
pricing_json,
@@ -135,22 +138,26 @@ class OpenlitConfig:
environment (str): Deployment environment.
application_name (str): Application name.
tracer: Tracer instance.
+ event_provider: Event logger provider instance.
meter: Metric Instance
otlp_endpoint (str): OTLP endpoint.
otlp_headers (Dict[str, str]): OTLP headers.
disable_batch (bool): Disable batch span processing flag.
- trace_content (bool): Enable or disable content tracing.
+ capture_message_content (bool): Enable or disable content tracing.
+ metrics_dict: Dictionary of metrics.
+ disable_metrics (bool): Flag to disable metrics.
pricing_json(str): path or url to the pricing json file
"""
cls.environment = environment
cls.application_name = application_name
cls.pricing_info = fetch_pricing_info(pricing_json)
cls.tracer = tracer
+ cls.event_provider = event_provider
cls.metrics_dict = metrics_dict
cls.otlp_endpoint = otlp_endpoint
cls.otlp_headers = otlp_headers
cls.disable_batch = disable_batch
- cls.trace_content = trace_content
+ cls.capture_message_content = capture_message_content
cls.disable_metrics = disable_metrics
@@ -187,8 +194,9 @@ def instrument_if_available(
environment=config.environment,
application_name=config.application_name,
tracer=config.tracer,
+ event_provider=config.event_provider,
pricing_info=config.pricing_info,
- trace_content=config.trace_content,
+ capture_message_content=config.capture_message_content,
metrics_dict=config.metrics_dict,
disable_metrics=config.disable_metrics,
)
@@ -207,10 +215,11 @@ def init(
environment="default",
application_name="default",
tracer=None,
+ event_logger=None,
otlp_endpoint=None,
otlp_headers=None,
disable_batch=False,
- trace_content=True,
+ capture_message_content=True,
disabled_instrumentors=None,
meter=None,
disable_metrics=False,
@@ -227,11 +236,12 @@ def init(
environment (str): Deployment environment.
application_name (str): Application name.
tracer: Tracer instance (Optional).
+ event_logger: EventLoggerProvider instance (Optional).
meter: OpenTelemetry Metrics Instance (Optional).
otlp_endpoint (str): OTLP endpoint for exporter (Optional).
otlp_headers (Dict[str, str]): OTLP headers for exporter (Optional).
disable_batch (bool): Flag to disable batch span processing (Optional).
- trace_content (bool): Flag to trace content (Optional).
+ capture_message_content (bool): Flag to trace content (Optional).
disabled_instrumentors (List[str]): Optional. List of instrumentor names to disable.
disable_metrics (bool): Flag to disable metrics (Optional).
pricing_json(str): File path or url to the pricing json (Optional).
@@ -308,9 +318,22 @@ def init(
)
if not tracer:
- logger.error("openLIT tracing setup failed. Tracing will not be available.")
+ logger.error("OpenLIT tracing setup failed. Tracing will not be available.")
return
+ # Setup events based on the provided or default configuration.
+ event_provider = setup_events(
+ application_name=application_name,
+ environment=environment,
+ event_logger=event_logger,
+ otlp_endpoint=None,
+ otlp_headers=None,
+ disable_batch=disable_batch,
+ )
+
+ if not event_provider:
+ logger.error("OpenLIT events setup failed. Events will not be available")
+
# Setup meter and receive metrics_dict instead of meter.
metrics_dict, err = setup_meter(
application_name=application_name,
@@ -326,15 +349,19 @@ def init(
)
return
+ if os.getenv("OTEL_INSTRUMENTATION_GENAI_CAPTURE_MESSAGE_CONTENT", "").lower == "false":
+ capture_message_content=False
+
# Update global configuration with the provided settings.
config.update_config(
environment,
application_name,
tracer,
+ event_provider,
otlp_endpoint,
otlp_headers,
disable_batch,
- trace_content,
+ capture_message_content,
metrics_dict,
disable_metrics,
pricing_json,
diff --git a/sdk/python/src/openlit/instrumentation/ag2/__init__.py b/sdk/python/src/openlit/instrumentation/ag2/__init__.py
index b952f804..7362213e 100644
--- a/sdk/python/src/openlit/instrumentation/ag2/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/ag2/__init__.py
@@ -1,4 +1,3 @@
-# pylint: disable=useless-return, bad-staticmethod-argument, disable=duplicate-code
"""Initializer of Auto Instrumentation of AG2 Functions"""
from typing import Collection
@@ -7,7 +6,7 @@ from opentelemetry.instrumentation.instrumentor import BaseInstrumentor
from wrapt import wrap_function_wrapper
from openlit.instrumentation.ag2.ag2 import (
- wrap_ag2
+ conversable_agent, agent_run
)
_instruments = ("ag2 >= 0.3.2",)
@@ -24,27 +23,27 @@ class AG2Instrumentor(BaseInstrumentor):
application_name = kwargs.get("application_name", "default_application")
environment = kwargs.get("environment", "default_environment")
tracer = kwargs.get("tracer")
+ event_provider = kwargs.get("event_provider")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("ag2")
wrap_function_wrapper(
"autogen.agentchat.conversable_agent",
- "ConversableAgent.initiate_chat",
- wrap_ag2("ag2.initiate_chat", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ "ConversableAgent.__init__",
+ conversable_agent(version, environment, application_name,
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"autogen.agentchat.conversable_agent",
- "ConversableAgent.generate_reply",
- wrap_ag2("ag2.generate_reply", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ "ConversableAgent.run",
+ agent_run(version, environment, application_name,
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics),
)
-
def _uninstrument(self, **kwargs):
# Proper uninstrumentation logic to revert patched methods
pass
diff --git a/sdk/python/src/openlit/instrumentation/ag2/ag2.py b/sdk/python/src/openlit/instrumentation/ag2/ag2.py
index 343807a6..53834635 100644
--- a/sdk/python/src/openlit/instrumentation/ag2/ag2.py
+++ b/sdk/python/src/openlit/instrumentation/ag2/ag2.py
@@ -1,98 +1,163 @@
-# pylint: disable=duplicate-code, broad-exception-caught, too-many-statements, unused-argument
"""
-Module for monitoring AG2.
+Module for monitoring AG2 API calls.
"""
import logging
+import time
from opentelemetry.trace import SpanKind, Status, StatusCode
from opentelemetry.sdk.resources import SERVICE_NAME, TELEMETRY_SDK_NAME, DEPLOYMENT_ENVIRONMENT
-from openlit.__helpers import handle_exception
+from openlit.__helpers import (
+ handle_exception,
+ get_chat_model_cost,
+ otel_event,
+)
from openlit.semcov import SemanticConvetion
# Initialize logger for logging potential issues and operations
logger = logging.getLogger(__name__)
-def wrap_ag2(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+AGENT_NAME = ''
+REQUEST_MODEL = ''
+SYSTEM_MESSAGE = ''
+MODEL_AND_NAME_SET = False
+
+def set_span_attributes(span, version, operation_name, environment,
+ application_name, server_address, server_port, request_model):
+ """
+ Set common attributes for the span.
+ """
+
+ # Set Span attributes (OTel Semconv)
+ span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
+ span.set_attribute(SemanticConvetion.GEN_AI_OPERATION, operation_name)
+ span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM, SemanticConvetion.GEN_AI_SYSTEM_AG2)
+ span.set_attribute(SemanticConvetion.GEN_AI_AGENT_NAME, AGENT_NAME)
+ span.set_attribute(SemanticConvetion.SERVER_ADDRESS, server_address)
+ span.set_attribute(SemanticConvetion.SERVER_PORT, server_port)
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL, request_model)
+
+ # Set Span attributes (Extras)
+ span.set_attribute(DEPLOYMENT_ENVIRONMENT, environment)
+ span.set_attribute(SERVICE_NAME, application_name)
+ span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION, version)
+
+def calculate_tokens_and_cost(response, request_model, pricing_info):
"""
- Creates a wrapper around a function call to trace and log its execution metrics.
-
- This function wraps any given function to measure its execution time,
- log its operation, and trace its execution using OpenTelemetry.
-
- Parameters:
- - gen_ai_endpoint (str): A descriptor or name for the endpoint being traced.
- - version (str): The version of the Langchain application.
- - environment (str): The deployment environment (e.g., 'production', 'development').
- - application_name (str): Name of the Langchain application.
- - tracer (opentelemetry.trace.Tracer): The tracer object used for OpenTelemetry tracing.
- - pricing_info (dict): Information about the pricing for internal metrics (currently not used).
- - trace_content (bool): Flag indicating whether to trace the content of the response.
-
- Returns:
- - function: A higher-order function that takes a function 'wrapped' and returns
- a new function that wraps 'wrapped' with additional tracing and logging.
+ Calculate the input, output tokens, and their respective costs.
"""
+ input_tokens = 0
+ output_tokens = 0
+ for usage_data in response.cost.values():
+ if isinstance(usage_data, dict):
+ for model_data in usage_data.values():
+ if isinstance(model_data, dict):
+ input_tokens += model_data.get('prompt_tokens', 0)
+ output_tokens += model_data.get('completion_tokens', 0)
+
+ cost = get_chat_model_cost(request_model, pricing_info, input_tokens, output_tokens)
+ return input_tokens, output_tokens, cost
+
+def emit_events(response, event_provider, capture_message_content):
+ """
+ Emit OpenTelemetry events for each chat history entry.
+ """
+ for chat in response.chat_history:
+ event_type = (
+ SemanticConvetion.GEN_AI_CHOICE if chat['role'] == 'user'
+ else SemanticConvetion.GEN_AI_USER_MESSAGE
+ )
+ choice_event = otel_event(
+ name=event_type,
+ attributes={
+ SemanticConvetion.GEN_AI_SYSTEM: SemanticConvetion.GEN_AI_SYSTEM_AG2
+ },
+ body={
+ "index": response.chat_history.index(chat),
+ "message": {
+ **({"content": chat['content']} if capture_message_content else {}),
+ "role": 'assistant' if chat['role'] == 'user' else 'user'
+ }
+ }
+ )
+ event_provider.emit(choice_event)
+
+def conversable_agent(version, environment, application_name,
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics):
+ """
+ Generates a telemetry wrapper for GenAI function call
+ """
def wrapper(wrapped, instance, args, kwargs):
- """
- An inner wrapper function that executes the wrapped function, measures execution
- time, and records trace data using OpenTelemetry.
-
- Parameters:
- - wrapped (Callable): The original function that this wrapper will execute.
- - instance (object): The instance to which the wrapped function belongs. This
- is used for instance methods. For static and classmethods,
- this may be None.
- - args (tuple): Positional arguments passed to the wrapped function.
- - kwargs (dict): Keyword arguments passed to the wrapped function.
-
- Returns:
- - The result of the wrapped function call.
-
- The wrapper initiates a span with the provided tracer, sets various attributes
- on the span based on the function's execution and response, and ensures
- errors are handled and logged appropriately.
- """
-
- with tracer.start_as_current_span(gen_ai_endpoint, kind= SpanKind.CLIENT) as span:
- response = wrapped(*args, **kwargs)
-
- if isinstance(instance.__dict__.get('llm_config'), dict):
- llm_model = instance.__dict__['llm_config'].get('model', 'gpt-4')
- else:
- # Fallback to default if 'llm_config' is not a dictionary
- llm_model = None
+ server_address, server_port = '127.0.0.1', 80
+ global AGENT_NAME, MODEL_AND_NAME_SET, REQUEST_MODEL, SYSTEM_MESSAGE
+
+ if not MODEL_AND_NAME_SET:
+ AGENT_NAME = kwargs.get("name", "NOT_FOUND")
+ REQUEST_MODEL = kwargs.get("llm_config", {}).get('model', 'gpt-4o')
+ SYSTEM_MESSAGE = kwargs.get('system_message', '')
+ MODEL_AND_NAME_SET = True
+ span_name = f"{SemanticConvetion.GEN_AI_OPERATION_TYPE_CREATE_AGENT} {AGENT_NAME}"
+
+ with tracer.start_as_current_span(span_name, kind=SpanKind.CLIENT) as span:
try:
- span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
- span.set_attribute(SemanticConvetion.GEN_AI_ENDPOINT,
- gen_ai_endpoint)
- span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_AG2)
- span.set_attribute(DEPLOYMENT_ENVIRONMENT,
- environment)
- span.set_attribute(SERVICE_NAME,
- application_name)
- span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
- SemanticConvetion.GEN_AI_OPERATION_TYPE_AGENT)
- span.set_attribute(SemanticConvetion.GEN_AI_AGENT_ROLE,
- instance.name)
- if llm_model:
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- llm_model)
+ start_time = time.time()
+ response = wrapped(*args, **kwargs)
+ end_time = time.time()
+ set_span_attributes(span, version, SemanticConvetion.GEN_AI_OPERATION_TYPE_CREATE_AGENT,
+ environment, application_name, server_address, server_port, REQUEST_MODEL)
+ span.set_attribute(SemanticConvetion.GEN_AI_AGENT_DESCRIPTION, SYSTEM_MESSAGE)
+ span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_MODEL, REQUEST_MODEL)
+ span.set_attribute(SemanticConvetion.GEN_AI_SERVER_TTFT, end_time - start_time)
span.set_status(Status(StatusCode.OK))
- # Return original response
return response
except Exception as e:
handle_exception(span, e)
logger.error("Error in trace creation: %s", e)
+ return response
+
+ return wrapper
+
+def agent_run(version, environment, application_name,
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics):
+ """
+ Generates a telemetry wrapper for GenAI function call
+ """
+ def wrapper(wrapped, instance, args, kwargs):
+ server_address, server_port = '127.0.0.1', 80
+
+ span_name = f"{SemanticConvetion.GEN_AI_OPERATION_TYPE_EXECUTE_AGENT_TASK} {AGENT_NAME}"
+
+ with tracer.start_as_current_span(span_name, kind=SpanKind.CLIENT) as span:
+ try:
+ start_time = time.time()
+ response = wrapped(*args, **kwargs)
+ end_time = time.time()
+
+ input_tokens, output_tokens, cost = calculate_tokens_and_cost(response, REQUEST_MODEL, pricing_info)
+ response_model = list(response.cost.get('usage_including_cached_inference', {}).keys())[1]
+
+ set_span_attributes(span, version, SemanticConvetion.GEN_AI_OPERATION_TYPE_EXECUTE_AGENT_TASK,
+ environment, application_name, server_address, server_port, REQUEST_MODEL)
+ span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_MODEL, response_model)
+ span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS, input_tokens)
+ span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS, output_tokens)
+ span.set_attribute(SemanticConvetion.GEN_AI_CLIENT_TOKEN_USAGE, input_tokens + output_tokens)
+ span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST, cost)
+ span.set_attribute(SemanticConvetion.GEN_AI_SERVER_TTFT, end_time - start_time)
+
+ emit_events(response, event_provider, capture_message_content)
+ span.set_status(Status(StatusCode.OK))
+
+ return response
- # Return original response
+ except Exception as e:
+ handle_exception(span, e)
+ logger.error("Error in trace creation: %s", e)
return response
return wrapper
diff --git a/sdk/python/src/openlit/instrumentation/ai21/__init__.py b/sdk/python/src/openlit/instrumentation/ai21/__init__.py
index 97099bbe..9606edc5 100644
--- a/sdk/python/src/openlit/instrumentation/ai21/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/ai21/__init__.py
@@ -27,9 +27,10 @@ class AI21Instrumentor(BaseInstrumentor):
application_name = kwargs.get("application_name", "default_application")
environment = kwargs.get("environment", "default_environment")
tracer = kwargs.get("tracer")
+ event_provider = kwargs.get("event_provider")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("ai21")
@@ -38,13 +39,13 @@ class AI21Instrumentor(BaseInstrumentor):
"ai21.clients.studio.resources.chat.chat_completions",
"ChatCompletions.create",
chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"ai21.clients.studio.resources.studio_conversational_rag",
"StudioConversationalRag.create",
chat_rag(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics),
)
#Async
@@ -52,13 +53,13 @@ class AI21Instrumentor(BaseInstrumentor):
"ai21.clients.studio.resources.chat.async_chat_completions",
"AsyncChatCompletions.create",
async_chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"ai21.clients.studio.resources.studio_conversational_rag",
"AsyncStudioConversationalRag.create",
async_chat_rag(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/ai21/ai21.py b/sdk/python/src/openlit/instrumentation/ai21/ai21.py
index d35a5739..16b7642f 100644
--- a/sdk/python/src/openlit/instrumentation/ai21/ai21.py
+++ b/sdk/python/src/openlit/instrumentation/ai21/ai21.py
@@ -4,55 +4,40 @@ Module for monitoring AI21 calls.
import logging
import time
-from opentelemetry.trace import SpanKind, Status, StatusCode
-from opentelemetry.sdk.resources import SERVICE_NAME, TELEMETRY_SDK_NAME, DEPLOYMENT_ENVIRONMENT
+from opentelemetry.trace import SpanKind
from openlit.__helpers import (
- get_chat_model_cost,
handle_exception,
response_as_dict,
- calculate_ttft,
- calculate_tbt,
- create_metrics_attributes,
set_server_address_and_port,
- general_tokens
)
+from openlit.instrumentation.ai21.utils import (
+ process_chunk,
+ process_chat_response,
+ process_streaming_chat_response,
+ process_chat_rag_response
+)
+
from openlit.semcov import SemanticConvetion
# Initialize logger for logging potential issues and operations
logger = logging.getLogger(__name__)
def chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics):
"""
- Generates a telemetry wrapper for chat completions to collect metrics.
-
- Args:
- version: Version of the monitoring package.
- environment: Deployment environment (e.g., production, staging).
- application_name: Name of the application using the AI21 SDK.
- tracer: OpenTelemetry tracer for creating spans.
- pricing_info: Information used for calculating the cost of AI21 usage.
- trace_content: Flag indicating whether to trace the actual content.
-
- Returns:
- A function that wraps the chat completions method to add telemetry.
+ Generates a telemetry wrapper for GenAI function call
"""
class TracedSyncStream:
"""
- Wrapper for streaming responses to collect metrics and trace data.
- Wraps the response to collect message IDs and aggregated response.
-
- This class implements the '__aiter__' and '__anext__' methods that
- handle asynchronous streaming responses.
-
- This class also implements '__aenter__' and '__aexit__' methods that
- handle asynchronous context management protocol.
+ Wrapper for streaming responses to collect telemetry.
"""
+
def __init__(
self,
wrapped,
span,
+ span_name,
kwargs,
server_address,
server_port,
@@ -60,12 +45,14 @@ def chat(version, environment, application_name,
):
self.__wrapped__ = wrapped
self._span = span
+ self._span_name = span_name
# Placeholder for aggregating streaming response
self._llmresponse = ""
self._response_id = ""
self._finish_reason = ""
self._input_tokens = 0
self._output_tokens = 0
+ self._choices = []
self._args = args
self._kwargs = kwargs
@@ -94,192 +81,36 @@ def chat(version, environment, application_name,
def __next__(self):
try:
chunk = self.__wrapped__.__next__()
- end_time = time.time()
- # Record the timestamp for the current chunk
- self._timestamps.append(end_time)
-
- if len(self._timestamps) == 1:
- # Calculate time to first chunk
- self._ttft = calculate_ttft(self._timestamps, self._start_time)
-
- chunked = response_as_dict(chunk)
- if (len(chunked.get('choices')) > 0 and ('delta' in chunked.get('choices')[0] and
- 'content' in chunked.get('choices')[0].get('delta'))):
-
- content = chunked.get('choices')[0].get('delta').get('content')
- if content:
- self._llmresponse += content
-
- if chunked.get('usage'):
- self._input_tokens = chunked.get('usage').get("prompt_tokens")
- self._output_tokens = chunked.get('usage').get("completion_tokens")
-
- self._response_id = chunked.get('id')
- self._finish_reason = chunked.get('choices')[0].get('finish_reason')
+ process_chunk(self, chunk)
return chunk
except StopIteration:
# Handling exception ensure observability without disrupting operation
try:
- self._end_time = time.time()
- if len(self._timestamps) > 1:
- self._tbt = calculate_tbt(self._timestamps)
-
- # Format 'messages' into a single string
- message_prompt = self._kwargs.get("messages", "")
- formatted_messages = []
- for message in message_prompt:
- role = message.role
- content = message.content
-
- if isinstance(content, list):
- content_str = ", ".join(
- f'{item["type"]}: {item["text"] if "text" in item else item["image_url"]}'
- if "type" in item else f'text: {item["text"]}'
- for item in content
- )
- formatted_messages.append(f"{role}: {content_str}")
- else:
- formatted_messages.append(f"{role}: {content}")
- prompt = "\n".join(formatted_messages)
-
- request_model = self._kwargs.get("model", "jamba-1.5-mini")
-
- # Calculate cost of the operation
- cost = get_chat_model_cost(request_model,
- pricing_info, self._input_tokens,
- self._output_tokens)
-
- # Set Span attributes (OTel Semconv)
- self._span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
- self._span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT)
- self._span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_AI21)
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- request_model)
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
- self._kwargs.get("seed", ""))
- self._span.set_attribute(SemanticConvetion.SERVER_PORT,
- self._server_port)
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
- self._kwargs.get("frequency_penalty", 0.0))
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
- self._kwargs.get("max_tokens", -1))
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_PRESENCE_PENALTY,
- self._kwargs.get("presence_penalty", 0.0))
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_STOP_SEQUENCES,
- self._kwargs.get("stop", []))
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
- self._kwargs.get("temperature", 0.4))
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
- self._kwargs.get("top_p", 1.0))
- self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
- [self._finish_reason])
- self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_ID,
- self._response_id)
- self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_MODEL,
- request_model)
- self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- self._input_tokens)
- self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
- self._output_tokens)
- self._span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
- self._server_address)
-
- if isinstance(self._llmresponse, str):
- self._span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- "text")
- else:
- self._span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- "json")
-
- # Set Span attributes (Extra)
- self._span.set_attribute(DEPLOYMENT_ENVIRONMENT,
- environment)
- self._span.set_attribute(SERVICE_NAME,
- application_name)
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
- True)
- self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- self._input_tokens + self._output_tokens)
- self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
- self._span.set_attribute(SemanticConvetion.GEN_AI_SERVER_TBT,
- self._tbt)
- self._span.set_attribute(SemanticConvetion.GEN_AI_SERVER_TTFT,
- self._ttft)
- self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
- version)
- if trace_content:
- self._span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
- },
- )
- self._span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: self._llmresponse,
- },
+ with tracer.start_as_current_span(self._span_name, kind= SpanKind.CLIENT) as self._span:
+ process_streaming_chat_response(
+ self,
+ pricing_info=pricing_info,
+ environment=environment,
+ application_name=application_name,
+ metrics=metrics,
+ event_provider=event_provider,
+ capture_message_content=capture_message_content,
+ disable_metrics=disable_metrics,
+ version=version
)
- self._span.set_status(Status(StatusCode.OK))
-
- if disable_metrics is False:
- attributes = create_metrics_attributes(
- service_name=application_name,
- deployment_environment=environment,
- operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
- system=SemanticConvetion.GEN_AI_SYSTEM_AI21,
- request_model=request_model,
- server_address=self._server_address,
- server_port=self._server_port,
- response_model=request_model,
- )
-
- metrics["genai_client_usage_tokens"].record(
- self._input_tokens + self._output_tokens, attributes
- )
- metrics["genai_client_operation_duration"].record(
- self._end_time - self._start_time, attributes
- )
- metrics["genai_server_tbt"].record(
- self._tbt, attributes
- )
- metrics["genai_server_ttft"].record(
- self._ttft, attributes
- )
- metrics["genai_requests"].add(1, attributes)
- metrics["genai_completion_tokens"].add(self._output_tokens, attributes)
- metrics["genai_prompt_tokens"].add(self._input_tokens, attributes)
- metrics["genai_cost"].record(cost, attributes)
-
except Exception as e:
handle_exception(self._span, e)
logger.error("Error in trace creation: %s", e)
- finally:
- self._span.end()
raise
def wrapper(wrapped, instance, args, kwargs):
"""
- Wraps the 'chat.completions' API call to add telemetry.
-
- This collects metrics such as execution time, cost, and token usage, and handles errors
- gracefully, adding details to the trace for observability.
-
- Args:
- wrapped: The original 'chat.completions' method to be wrapped.
- instance: The instance of the class where the original method is defined.
- args: Positional arguments for the 'chat.completions' method.
- kwargs: Keyword arguments for the 'chat.completions' method.
-
- Returns:
- The response from the original 'chat.completions' method.
+ Wraps the GenAI function call.
"""
# Check if streaming is enabled for the API call
streaming = kwargs.get("stream", False)
+
server_address, server_port = set_server_address_and_port(instance, "api.ai21.com", 443)
request_model = kwargs.get("model", "jamba-1.5-mini")
@@ -290,197 +121,44 @@ def chat(version, environment, application_name,
# Special handling for streaming response to accommodate the nature of data flow
awaited_wrapped = wrapped(*args, **kwargs)
span = tracer.start_span(span_name, kind=SpanKind.CLIENT)
-
- return TracedSyncStream(awaited_wrapped, span, kwargs, server_address, server_port)
+ return TracedSyncStream(awaited_wrapped, span, span_name, kwargs, server_address, server_port)
# Handling for non-streaming responses
else:
with tracer.start_as_current_span(span_name, kind= SpanKind.CLIENT) as span:
start_time = time.time()
response = wrapped(*args, **kwargs)
- end_time = time.time()
-
- response_dict = response_as_dict(response)
-
- try:
- # Format 'messages' into a single string
- message_prompt = kwargs.get("messages", "")
- formatted_messages = []
- for message in message_prompt:
- role = message.role
- content = message.content
-
- if isinstance(content, list):
- content_str = ", ".join(
- f'{item["type"]}: {item["text"] if "text" in item else item["image_url"]}'
- if "type" in item else f'text: {item["text"]}'
- for item in content
- )
- formatted_messages.append(f"{role}: {content_str}")
- else:
- formatted_messages.append(f"{role}: {content}")
- prompt = "\n".join(formatted_messages)
-
- input_tokens = response_dict.get('usage').get('prompt_tokens')
- output_tokens = response_dict.get('usage').get('completion_tokens')
-
- # Calculate cost of the operation
- cost = get_chat_model_cost(request_model,
- pricing_info, input_tokens,
- output_tokens)
-
- # Set base span attribues (OTel Semconv)
- span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
- span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT)
- span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_AI21)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- request_model)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
- kwargs.get("seed", ""))
- span.set_attribute(SemanticConvetion.SERVER_PORT,
- server_port)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
- kwargs.get("frequency_penalty", 0.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
- kwargs.get("max_tokens", -1))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_PRESENCE_PENALTY,
- kwargs.get("presence_penalty", 0.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_STOP_SEQUENCES,
- kwargs.get("stop", []))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
- kwargs.get("temperature", 0.4))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
- kwargs.get("top_p", 1.0))
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_ID,
- response_dict.get("id"))
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_MODEL,
- request_model)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- input_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
- output_tokens)
- span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
- server_address)
-
- # Set base span attribues (Extras)
- span.set_attribute(DEPLOYMENT_ENVIRONMENT,
- environment)
- span.set_attribute(SERVICE_NAME,
- application_name)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
- False)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- input_tokens + output_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
- span.set_attribute(SemanticConvetion.GEN_AI_SERVER_TTFT,
- end_time - start_time)
- span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
- version)
- if trace_content:
- span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
- },
- )
-
- for i in range(kwargs.get('n',1)):
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
- [response_dict.get('choices')[i].get('finish_reason')])
- if trace_content:
- span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
- attributes={
- # pylint: disable=line-too-long
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: str(response_dict.get('choices')[i].get('message').get('content')),
- },
- )
- if kwargs.get('tools'):
- span.set_attribute(SemanticConvetion.GEN_AI_TOOL_CALLS,
- str(response_dict.get('choices')[i].get('message').get('tool_calls')))
-
- if isinstance(response_dict.get('choices')[i].get('message').get('content'), str):
- span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- "text")
- elif response_dict.get('choices')[i].get('message').get('content') is not None:
- span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- "json")
-
- span.set_status(Status(StatusCode.OK))
-
- if disable_metrics is False:
- attributes = create_metrics_attributes(
- service_name=application_name,
- deployment_environment=environment,
- operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
- system=SemanticConvetion.GEN_AI_SYSTEM_AI21,
- request_model=request_model,
- server_address=server_address,
- server_port=server_port,
- response_model=request_model,
- )
-
- metrics["genai_client_usage_tokens"].record(
- input_tokens + output_tokens, attributes
- )
- metrics["genai_client_operation_duration"].record(
- end_time - start_time, attributes
- )
- metrics["genai_server_ttft"].record(
- end_time - start_time, attributes
- )
- metrics["genai_requests"].add(1, attributes)
- metrics["genai_completion_tokens"].add(output_tokens, attributes)
- metrics["genai_prompt_tokens"].add(input_tokens, attributes)
- metrics["genai_cost"].record(cost, attributes)
-
- # Return original response
- return response
-
- except Exception as e:
- handle_exception(span, e)
- logger.error("Error in trace creation: %s", e)
-
- # Return original response
- return response
+ response = process_chat_response(
+ response=response_as_dict(response),
+ request_model=request_model,
+ pricing_info=pricing_info,
+ server_port=server_port,
+ server_address=server_address,
+ environment=environment,
+ application_name=application_name,
+ metrics=metrics,
+ event_provider=event_provider,
+ start_time=start_time,
+ span=span,
+ capture_message_content=capture_message_content,
+ disable_metrics=disable_metrics,
+ version=version,
+ **kwargs
+ )
+
+ return response
return wrapper
def chat_rag(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics):
"""
- Generates a telemetry wrapper for chat completions to collect metrics.
-
- Args:
- version: Version of the monitoring package.
- environment: Deployment environment (e.g., production, staging).
- application_name: Name of the application using the AI21 SDK.
- tracer: OpenTelemetry tracer for creating spans.
- pricing_info: Information used for calculating the cost of AI21 usage.
- trace_content: Flag indicating whether to trace the actual content.
-
- Returns:
- A function that wraps the chat completions method to add telemetry.
+ Generates a telemetry wrapper for GenAI function call
"""
def wrapper(wrapped, instance, args, kwargs):
"""
- Wraps the 'chat.completions' API call to add telemetry.
-
- This collects metrics such as execution time, cost, and token usage, and handles errors
- gracefully, adding details to the trace for observability.
-
- Args:
- wrapped: The original 'chat.completions' method to be wrapped.
- instance: The instance of the class where the original method is defined.
- args: Positional arguments for the 'chat.completions' method.
- kwargs: Keyword arguments for the 'chat.completions' method.
-
- Returns:
- The response from the original 'chat.completions' method.
+ Wraps the GenAI function call.
"""
server_address, server_port = set_server_address_and_port(instance, "api.ai21.com", 443)
@@ -491,165 +169,24 @@ def chat_rag(version, environment, application_name,
with tracer.start_as_current_span(span_name, kind= SpanKind.CLIENT) as span:
start_time = time.time()
response = wrapped(*args, **kwargs)
- end_time = time.time()
-
- response_dict = response_as_dict(response)
-
- try:
- # Format 'messages' into a single string
- message_prompt = kwargs.get("messages", "")
- formatted_messages = []
- for message in message_prompt:
- role = message.role
- content = message.content
-
- if isinstance(content, list):
- content_str = ", ".join(
- f'{item["type"]}: {item["text"] if "text" in item else item["image_url"]}'
- if "type" in item else f'text: {item["text"]}'
- for item in content
- )
- formatted_messages.append(f"{role}: {content_str}")
- else:
- formatted_messages.append(f"{role}: {content}")
- prompt = "\n".join(formatted_messages)
-
- input_tokens = general_tokens(prompt)
-
- # Set base span attribues (OTel Semconv)
- span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
- span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT)
- span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_AI21)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- request_model)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
- kwargs.get("seed", ""))
- span.set_attribute(SemanticConvetion.SERVER_PORT,
- server_port)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
- kwargs.get("frequency_penalty", 0.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
- kwargs.get("max_tokens", -1))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_PRESENCE_PENALTY,
- kwargs.get("presence_penalty", 0.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_STOP_SEQUENCES,
- kwargs.get("stop", []))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
- kwargs.get("temperature", 0.4))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
- kwargs.get("top_p", 1.0))
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_ID,
- response_dict.get("id"))
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_MODEL,
- request_model)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- input_tokens)
- span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
- server_address)
-
- # Set base span attribues (Extras)
- span.set_attribute(DEPLOYMENT_ENVIRONMENT,
- environment)
- span.set_attribute(SERVICE_NAME,
- application_name)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
- False)
- span.set_attribute(SemanticConvetion.GEN_AI_SERVER_TTFT,
- end_time - start_time)
- span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
- version)
- span.set_attribute(SemanticConvetion.GEN_AI_RAG_MAX_SEGMENTS,
- kwargs.get("max_segments", -1))
- span.set_attribute(SemanticConvetion.GEN_AI_RAG_STRATEGY,
- kwargs.get("retrieval_strategy", "segments"))
- span.set_attribute(SemanticConvetion.GEN_AI_RAG_SIMILARITY_THRESHOLD,
- kwargs.get("retrieval_similarity_threshold", -1))
- span.set_attribute(SemanticConvetion.GEN_AI_RAG_MAX_NEIGHBORS,
- kwargs.get("max_neighbors", -1))
- span.set_attribute(SemanticConvetion.GEN_AI_RAG_FILE_IDS,
- str(kwargs.get("file_ids", "")))
- span.set_attribute(SemanticConvetion.GEN_AI_RAG_DOCUMENTS_PATH,
- kwargs.get("path", ""))
- if trace_content:
- span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
- },
- )
-
- output_tokens = 0
- for i in range(kwargs.get('n',1)):
- output_tokens += general_tokens(response_dict.get('choices')[i].get('content'))
-
- if trace_content:
- span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
- attributes={
- # pylint: disable=line-too-long
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: str(response_dict.get('choices')[i].get('content')),
- },
- )
- if kwargs.get('tools'):
- span.set_attribute(SemanticConvetion.GEN_AI_TOOL_CALLS,
- str(response_dict.get('choices')[i].get('message').get('tool_calls')))
-
- if isinstance(response_dict.get('choices')[i].get('content'), str):
- span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- "text")
- elif response_dict.get('choices')[i].get('content') is not None:
- span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- "json")
-
- # Calculate cost of the operation
- cost = get_chat_model_cost(request_model,
- pricing_info, input_tokens,
- output_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
- output_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- input_tokens + output_tokens)
-
- span.set_status(Status(StatusCode.OK))
-
- if disable_metrics is False:
- attributes = create_metrics_attributes(
- service_name=application_name,
- deployment_environment=environment,
- operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
- system=SemanticConvetion.GEN_AI_SYSTEM_AI21,
- request_model=request_model,
- server_address=server_address,
- server_port=server_port,
- response_model=request_model,
- )
-
- metrics["genai_client_usage_tokens"].record(
- input_tokens + output_tokens, attributes
- )
- metrics["genai_client_operation_duration"].record(
- end_time - start_time, attributes
- )
- metrics["genai_server_ttft"].record(
- end_time - start_time, attributes
- )
- metrics["genai_requests"].add(1, attributes)
- metrics["genai_completion_tokens"].add(output_tokens, attributes)
- metrics["genai_prompt_tokens"].add(input_tokens, attributes)
- metrics["genai_cost"].record(cost, attributes)
-
- # Return original response
- return response
-
- except Exception as e:
- handle_exception(span, e)
- logger.error("Error in trace creation: %s", e)
-
- # Return original response
- return response
+ response = process_chat_rag_response(
+ response=response_as_dict(response),
+ request_model=request_model,
+ pricing_info=pricing_info,
+ server_port=server_port,
+ server_address=server_address,
+ environment=environment,
+ application_name=application_name,
+ metrics=metrics,
+ event_provider=event_provider,
+ start_time=start_time,
+ span=span,
+ capture_message_content=capture_message_content,
+ disable_metrics=disable_metrics,
+ version=version,
+ **kwargs
+ )
+
+ return response
return wrapper
diff --git a/sdk/python/src/openlit/instrumentation/ai21/async_ai21.py b/sdk/python/src/openlit/instrumentation/ai21/async_ai21.py
index 5f471280..11359aa8 100644
--- a/sdk/python/src/openlit/instrumentation/ai21/async_ai21.py
+++ b/sdk/python/src/openlit/instrumentation/ai21/async_ai21.py
@@ -4,55 +4,40 @@ Module for monitoring AI21 calls.
import logging
import time
-from opentelemetry.trace import SpanKind, Status, StatusCode
-from opentelemetry.sdk.resources import SERVICE_NAME, TELEMETRY_SDK_NAME, DEPLOYMENT_ENVIRONMENT
+from opentelemetry.trace import SpanKind
from openlit.__helpers import (
- get_chat_model_cost,
handle_exception,
response_as_dict,
- calculate_ttft,
- calculate_tbt,
- create_metrics_attributes,
set_server_address_and_port,
- general_tokens
)
+from openlit.instrumentation.ai21.utils import (
+ process_chunk,
+ process_chat_response,
+ process_streaming_chat_response,
+ process_chat_rag_response
+)
+
from openlit.semcov import SemanticConvetion
# Initialize logger for logging potential issues and operations
logger = logging.getLogger(__name__)
def async_chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics):
"""
- Generates a telemetry wrapper for chat completions to collect metrics.
-
- Args:
- version: Version of the monitoring package.
- environment: Deployment environment (e.g., production, staging).
- application_name: Name of the application using the AI21 SDK.
- tracer: OpenTelemetry tracer for creating spans.
- pricing_info: Information used for calculating the cost of AI21 usage.
- trace_content: Flag indicating whether to trace the actual content.
-
- Returns:
- A function that wraps the chat completions method to add telemetry.
+ Generates a telemetry wrapper for GenAI function call
"""
class TracedAsyncStream:
"""
- Wrapper for streaming responses to collect metrics and trace data.
- Wraps the response to collect message IDs and aggregated response.
-
- This class implements the '__aiter__' and '__anext__' methods that
- handle asynchronous streaming responses.
-
- This class also implements '__aenter__' and '__aexit__' methods that
- handle asynchronous context management protocol.
+ Wrapper for streaming responses to collect telemetry.
"""
+
def __init__(
self,
wrapped,
span,
+ span_name,
kwargs,
server_address,
server_port,
@@ -60,12 +45,14 @@ def async_chat(version, environment, application_name,
):
self.__wrapped__ = wrapped
self._span = span
+ self._span_name = span_name
# Placeholder for aggregating streaming response
self._llmresponse = ""
self._response_id = ""
self._finish_reason = ""
self._input_tokens = 0
self._output_tokens = 0
+ self._choices = []
self._args = args
self._kwargs = kwargs
@@ -94,192 +81,36 @@ def async_chat(version, environment, application_name,
async def __anext__(self):
try:
chunk = await self.__wrapped__.__anext__()
- end_time = time.time()
- # Record the timestamp for the current chunk
- self._timestamps.append(end_time)
-
- if len(self._timestamps) == 1:
- # Calculate time to first chunk
- self._ttft = calculate_ttft(self._timestamps, self._start_time)
-
- chunked = response_as_dict(chunk)
- if (len(chunked.get('choices')) > 0 and ('delta' in chunked.get('choices')[0] and
- 'content' in chunked.get('choices')[0].get('delta'))):
-
- content = chunked.get('choices')[0].get('delta').get('content')
- if content:
- self._llmresponse += content
-
- if chunked.get('usage'):
- self._input_tokens = chunked.get('usage').get("prompt_tokens")
- self._output_tokens = chunked.get('usage').get("completion_tokens")
-
- self._response_id = chunked.get('id')
- self._finish_reason = chunked.get('choices')[0].get('finish_reason')
+ process_chunk(self, chunk)
return chunk
except StopAsyncIteration:
# Handling exception ensure observability without disrupting operation
try:
- self._end_time = time.time()
- if len(self._timestamps) > 1:
- self._tbt = calculate_tbt(self._timestamps)
-
- # Format 'messages' into a single string
- message_prompt = self._kwargs.get("messages", "")
- formatted_messages = []
- for message in message_prompt:
- role = message.role
- content = message.content
-
- if isinstance(content, list):
- content_str = ", ".join(
- f'{item["type"]}: {item["text"] if "text" in item else item["image_url"]}'
- if "type" in item else f'text: {item["text"]}'
- for item in content
- )
- formatted_messages.append(f"{role}: {content_str}")
- else:
- formatted_messages.append(f"{role}: {content}")
- prompt = "\n".join(formatted_messages)
-
- request_model = self._kwargs.get("model", "jamba-1.5-mini")
-
- # Calculate cost of the operation
- cost = get_chat_model_cost(request_model,
- pricing_info, self._input_tokens,
- self._output_tokens)
-
- # Set Span attributes (OTel Semconv)
- self._span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
- self._span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT)
- self._span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_AI21)
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- request_model)
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
- self._kwargs.get("seed", ""))
- self._span.set_attribute(SemanticConvetion.SERVER_PORT,
- self._server_port)
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
- self._kwargs.get("frequency_penalty", 0.0))
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
- self._kwargs.get("max_tokens", -1))
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_PRESENCE_PENALTY,
- self._kwargs.get("presence_penalty", 0.0))
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_STOP_SEQUENCES,
- self._kwargs.get("stop", []))
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
- self._kwargs.get("temperature", 0.4))
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
- self._kwargs.get("top_p", 1.0))
- self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
- [self._finish_reason])
- self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_ID,
- self._response_id)
- self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_MODEL,
- request_model)
- self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- self._input_tokens)
- self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
- self._output_tokens)
- self._span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
- self._server_address)
-
- if isinstance(self._llmresponse, str):
- self._span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- "text")
- else:
- self._span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- "json")
-
- # Set Span attributes (Extra)
- self._span.set_attribute(DEPLOYMENT_ENVIRONMENT,
- environment)
- self._span.set_attribute(SERVICE_NAME,
- application_name)
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
- True)
- self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- self._input_tokens + self._output_tokens)
- self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
- self._span.set_attribute(SemanticConvetion.GEN_AI_SERVER_TBT,
- self._tbt)
- self._span.set_attribute(SemanticConvetion.GEN_AI_SERVER_TTFT,
- self._ttft)
- self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
- version)
- if trace_content:
- self._span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
- },
- )
- self._span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: self._llmresponse,
- },
+ with tracer.start_as_current_span(self._span_name, kind= SpanKind.CLIENT) as self._span:
+ process_streaming_chat_response(
+ self,
+ pricing_info=pricing_info,
+ environment=environment,
+ application_name=application_name,
+ metrics=metrics,
+ event_provider=event_provider,
+ capture_message_content=capture_message_content,
+ disable_metrics=disable_metrics,
+ version=version
)
- self._span.set_status(Status(StatusCode.OK))
-
- if disable_metrics is False:
- attributes = create_metrics_attributes(
- service_name=application_name,
- deployment_environment=environment,
- operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
- system=SemanticConvetion.GEN_AI_SYSTEM_AI21,
- request_model=request_model,
- server_address=self._server_address,
- server_port=self._server_port,
- response_model=request_model,
- )
-
- metrics["genai_client_usage_tokens"].record(
- self._input_tokens + self._output_tokens, attributes
- )
- metrics["genai_client_operation_duration"].record(
- self._end_time - self._start_time, attributes
- )
- metrics["genai_server_tbt"].record(
- self._tbt, attributes
- )
- metrics["genai_server_ttft"].record(
- self._ttft, attributes
- )
- metrics["genai_requests"].add(1, attributes)
- metrics["genai_completion_tokens"].add(self._output_tokens, attributes)
- metrics["genai_prompt_tokens"].add(self._input_tokens, attributes)
- metrics["genai_cost"].record(cost, attributes)
-
except Exception as e:
handle_exception(self._span, e)
logger.error("Error in trace creation: %s", e)
- finally:
- self._span.end()
raise
async def wrapper(wrapped, instance, args, kwargs):
"""
- Wraps the 'chat.completions' API call to add telemetry.
-
- This collects metrics such as execution time, cost, and token usage, and handles errors
- gracefully, adding details to the trace for observability.
-
- Args:
- wrapped: The original 'chat.completions' method to be wrapped.
- instance: The instance of the class where the original method is defined.
- args: Positional arguments for the 'chat.completions' method.
- kwargs: Keyword arguments for the 'chat.completions' method.
-
- Returns:
- The response from the original 'chat.completions' method.
+ Wraps the GenAI function call.
"""
# Check if streaming is enabled for the API call
streaming = kwargs.get("stream", False)
+
server_address, server_port = set_server_address_and_port(instance, "api.ai21.com", 443)
request_model = kwargs.get("model", "jamba-1.5-mini")
@@ -290,197 +121,44 @@ def async_chat(version, environment, application_name,
# Special handling for streaming response to accommodate the nature of data flow
awaited_wrapped = await wrapped(*args, **kwargs)
span = tracer.start_span(span_name, kind=SpanKind.CLIENT)
-
- return TracedAsyncStream(awaited_wrapped, span, kwargs, server_address, server_port)
+ return TracedAsyncStream(awaited_wrapped, span, span_name, kwargs, server_address, server_port)
# Handling for non-streaming responses
else:
with tracer.start_as_current_span(span_name, kind= SpanKind.CLIENT) as span:
start_time = time.time()
response = await wrapped(*args, **kwargs)
- end_time = time.time()
-
- response_dict = response_as_dict(response)
-
- try:
- # Format 'messages' into a single string
- message_prompt = kwargs.get("messages", "")
- formatted_messages = []
- for message in message_prompt:
- role = message.role
- content = message.content
-
- if isinstance(content, list):
- content_str = ", ".join(
- f'{item["type"]}: {item["text"] if "text" in item else item["image_url"]}'
- if "type" in item else f'text: {item["text"]}'
- for item in content
- )
- formatted_messages.append(f"{role}: {content_str}")
- else:
- formatted_messages.append(f"{role}: {content}")
- prompt = "\n".join(formatted_messages)
-
- input_tokens = response_dict.get('usage').get('prompt_tokens')
- output_tokens = response_dict.get('usage').get('completion_tokens')
-
- # Calculate cost of the operation
- cost = get_chat_model_cost(request_model,
- pricing_info, input_tokens,
- output_tokens)
-
- # Set base span attribues (OTel Semconv)
- span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
- span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT)
- span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_AI21)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- request_model)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
- kwargs.get("seed", ""))
- span.set_attribute(SemanticConvetion.SERVER_PORT,
- server_port)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
- kwargs.get("frequency_penalty", 0.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
- kwargs.get("max_tokens", -1))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_PRESENCE_PENALTY,
- kwargs.get("presence_penalty", 0.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_STOP_SEQUENCES,
- kwargs.get("stop", []))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
- kwargs.get("temperature", 0.4))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
- kwargs.get("top_p", 1.0))
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_ID,
- response_dict.get("id"))
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_MODEL,
- request_model)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- input_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
- output_tokens)
- span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
- server_address)
-
- # Set base span attribues (Extras)
- span.set_attribute(DEPLOYMENT_ENVIRONMENT,
- environment)
- span.set_attribute(SERVICE_NAME,
- application_name)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
- False)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- input_tokens + output_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
- span.set_attribute(SemanticConvetion.GEN_AI_SERVER_TTFT,
- end_time - start_time)
- span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
- version)
- if trace_content:
- span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
- },
- )
-
- for i in range(kwargs.get('n',1)):
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
- [response_dict.get('choices')[i].get('finish_reason')])
- if trace_content:
- span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
- attributes={
- # pylint: disable=line-too-long
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: str(response_dict.get('choices')[i].get('message').get('content')),
- },
- )
- if kwargs.get('tools'):
- span.set_attribute(SemanticConvetion.GEN_AI_TOOL_CALLS,
- str(response_dict.get('choices')[i].get('message').get('tool_calls')))
-
- if isinstance(response_dict.get('choices')[i].get('message').get('content'), str):
- span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- "text")
- elif response_dict.get('choices')[i].get('message').get('content') is not None:
- span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- "json")
-
- span.set_status(Status(StatusCode.OK))
-
- if disable_metrics is False:
- attributes = create_metrics_attributes(
- service_name=application_name,
- deployment_environment=environment,
- operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
- system=SemanticConvetion.GEN_AI_SYSTEM_AI21,
- request_model=request_model,
- server_address=server_address,
- server_port=server_port,
- response_model=request_model,
- )
-
- metrics["genai_client_usage_tokens"].record(
- input_tokens + output_tokens, attributes
- )
- metrics["genai_client_operation_duration"].record(
- end_time - start_time, attributes
- )
- metrics["genai_server_ttft"].record(
- end_time - start_time, attributes
- )
- metrics["genai_requests"].add(1, attributes)
- metrics["genai_completion_tokens"].add(output_tokens, attributes)
- metrics["genai_prompt_tokens"].add(input_tokens, attributes)
- metrics["genai_cost"].record(cost, attributes)
-
- # Return original response
- return response
-
- except Exception as e:
- handle_exception(span, e)
- logger.error("Error in trace creation: %s", e)
-
- # Return original response
- return response
+ response = process_chat_response(
+ response=response_as_dict(response),
+ request_model=request_model,
+ pricing_info=pricing_info,
+ server_port=server_port,
+ server_address=server_address,
+ environment=environment,
+ application_name=application_name,
+ metrics=metrics,
+ event_provider=event_provider,
+ start_time=start_time,
+ span=span,
+ capture_message_content=capture_message_content,
+ disable_metrics=disable_metrics,
+ version=version,
+ **kwargs
+ )
+
+ return response
return wrapper
def async_chat_rag(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics):
"""
- Generates a telemetry wrapper for chat completions to collect metrics.
-
- Args:
- version: Version of the monitoring package.
- environment: Deployment environment (e.g., production, staging).
- application_name: Name of the application using the AI21 SDK.
- tracer: OpenTelemetry tracer for creating spans.
- pricing_info: Information used for calculating the cost of AI21 usage.
- trace_content: Flag indicating whether to trace the actual content.
-
- Returns:
- A function that wraps the chat completions method to add telemetry.
+ Generates a telemetry wrapper for GenAI function call
"""
async def wrapper(wrapped, instance, args, kwargs):
"""
- Wraps the 'chat.completions' API call to add telemetry.
-
- This collects metrics such as execution time, cost, and token usage, and handles errors
- gracefully, adding details to the trace for observability.
-
- Args:
- wrapped: The original 'chat.completions' method to be wrapped.
- instance: The instance of the class where the original method is defined.
- args: Positional arguments for the 'chat.completions' method.
- kwargs: Keyword arguments for the 'chat.completions' method.
-
- Returns:
- The response from the original 'chat.completions' method.
+ Wraps the GenAI function call.
"""
server_address, server_port = set_server_address_and_port(instance, "api.ai21.com", 443)
@@ -491,165 +169,24 @@ def async_chat_rag(version, environment, application_name,
with tracer.start_as_current_span(span_name, kind= SpanKind.CLIENT) as span:
start_time = time.time()
response = await wrapped(*args, **kwargs)
- end_time = time.time()
-
- response_dict = response_as_dict(response)
-
- try:
- # Format 'messages' into a single string
- message_prompt = kwargs.get("messages", "")
- formatted_messages = []
- for message in message_prompt:
- role = message.role
- content = message.content
-
- if isinstance(content, list):
- content_str = ", ".join(
- f'{item["type"]}: {item["text"] if "text" in item else item["image_url"]}'
- if "type" in item else f'text: {item["text"]}'
- for item in content
- )
- formatted_messages.append(f"{role}: {content_str}")
- else:
- formatted_messages.append(f"{role}: {content}")
- prompt = "\n".join(formatted_messages)
-
- input_tokens = general_tokens(prompt)
-
- # Set base span attribues (OTel Semconv)
- span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
- span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT)
- span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_AI21)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- request_model)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
- kwargs.get("seed", ""))
- span.set_attribute(SemanticConvetion.SERVER_PORT,
- server_port)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
- kwargs.get("frequency_penalty", 0.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
- kwargs.get("max_tokens", -1))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_PRESENCE_PENALTY,
- kwargs.get("presence_penalty", 0.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_STOP_SEQUENCES,
- kwargs.get("stop", []))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
- kwargs.get("temperature", 0.4))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
- kwargs.get("top_p", 1.0))
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_ID,
- response_dict.get("id"))
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_MODEL,
- request_model)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- input_tokens)
- span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
- server_address)
-
- # Set base span attribues (Extras)
- span.set_attribute(DEPLOYMENT_ENVIRONMENT,
- environment)
- span.set_attribute(SERVICE_NAME,
- application_name)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
- False)
- span.set_attribute(SemanticConvetion.GEN_AI_SERVER_TTFT,
- end_time - start_time)
- span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
- version)
- span.set_attribute(SemanticConvetion.GEN_AI_RAG_MAX_SEGMENTS,
- kwargs.get("max_segments", -1))
- span.set_attribute(SemanticConvetion.GEN_AI_RAG_STRATEGY,
- kwargs.get("retrieval_strategy", "segments"))
- span.set_attribute(SemanticConvetion.GEN_AI_RAG_SIMILARITY_THRESHOLD,
- kwargs.get("retrieval_similarity_threshold", -1))
- span.set_attribute(SemanticConvetion.GEN_AI_RAG_MAX_NEIGHBORS,
- kwargs.get("max_neighbors", -1))
- span.set_attribute(SemanticConvetion.GEN_AI_RAG_FILE_IDS,
- str(kwargs.get("file_ids", "")))
- span.set_attribute(SemanticConvetion.GEN_AI_RAG_DOCUMENTS_PATH,
- kwargs.get("path", ""))
- if trace_content:
- span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
- },
- )
-
- output_tokens = 0
- for i in range(kwargs.get('n',1)):
- output_tokens += general_tokens(response_dict.get('choices')[i].get('content'))
-
- if trace_content:
- span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
- attributes={
- # pylint: disable=line-too-long
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: str(response_dict.get('choices')[i].get('content')),
- },
- )
- if kwargs.get('tools'):
- span.set_attribute(SemanticConvetion.GEN_AI_TOOL_CALLS,
- str(response_dict.get('choices')[i].get('message').get('tool_calls')))
-
- if isinstance(response_dict.get('choices')[i].get('content'), str):
- span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- "text")
- elif response_dict.get('choices')[i].get('content') is not None:
- span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- "json")
-
- # Calculate cost of the operation
- cost = get_chat_model_cost(request_model,
- pricing_info, input_tokens,
- output_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
- output_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- input_tokens + output_tokens)
-
- span.set_status(Status(StatusCode.OK))
-
- if disable_metrics is False:
- attributes = create_metrics_attributes(
- service_name=application_name,
- deployment_environment=environment,
- operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
- system=SemanticConvetion.GEN_AI_SYSTEM_AI21,
- request_model=request_model,
- server_address=server_address,
- server_port=server_port,
- response_model=request_model,
- )
-
- metrics["genai_client_usage_tokens"].record(
- input_tokens + output_tokens, attributes
- )
- metrics["genai_client_operation_duration"].record(
- end_time - start_time, attributes
- )
- metrics["genai_server_ttft"].record(
- end_time - start_time, attributes
- )
- metrics["genai_requests"].add(1, attributes)
- metrics["genai_completion_tokens"].add(output_tokens, attributes)
- metrics["genai_prompt_tokens"].add(input_tokens, attributes)
- metrics["genai_cost"].record(cost, attributes)
-
- # Return original response
- return response
-
- except Exception as e:
- handle_exception(span, e)
- logger.error("Error in trace creation: %s", e)
-
- # Return original response
- return response
+ response = process_chat_rag_response(
+ response=response_as_dict(response),
+ request_model=request_model,
+ pricing_info=pricing_info,
+ server_port=server_port,
+ server_address=server_address,
+ environment=environment,
+ application_name=application_name,
+ metrics=metrics,
+ event_provider=event_provider,
+ start_time=start_time,
+ span=span,
+ capture_message_content=capture_message_content,
+ disable_metrics=disable_metrics,
+ version=version,
+ **kwargs
+ )
+
+ return response
return wrapper
diff --git a/sdk/python/src/openlit/instrumentation/ai21/utils.py b/sdk/python/src/openlit/instrumentation/ai21/utils.py
new file mode 100644
index 00000000..35cb9d24
--- /dev/null
+++ b/sdk/python/src/openlit/instrumentation/ai21/utils.py
@@ -0,0 +1,407 @@
+"""
+AI21 OpenTelemetry instrumentation utility functions
+"""
+
+import time
+
+from opentelemetry.sdk.resources import SERVICE_NAME, TELEMETRY_SDK_NAME, DEPLOYMENT_ENVIRONMENT
+from opentelemetry.trace import Status, StatusCode
+
+from openlit.__helpers import (
+ calculate_ttft,
+ response_as_dict,
+ calculate_tbt,
+ general_tokens,
+ extract_and_format_input,
+ get_chat_model_cost,
+ handle_exception,
+ create_metrics_attributes,
+ otel_event,
+ concatenate_all_contents
+)
+from openlit.semcov import SemanticConvetion
+
+def setup_common_span_attributes(span, request_model, kwargs, tokens,
+ server_port, server_address, environment,
+ application_name, extra_attrs):
+ """
+ Set common span attributes for both chat and RAG operations.
+ """
+
+ # Base attributes from SDK and operation settings.
+ span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
+ span.set_attribute(SemanticConvetion.GEN_AI_OPERATION, SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT)
+ span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM, SemanticConvetion.GEN_AI_SYSTEM_AI21)
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL, request_model)
+ span.set_attribute(SemanticConvetion.SERVER_PORT, server_port)
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED, kwargs.get("seed", ""))
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY, kwargs.get("frequency_penalty", 0.0))
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS, kwargs.get("max_tokens", -1))
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_PRESENCE_PENALTY, kwargs.get("presence_penalty", 0.0))
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_STOP_SEQUENCES, kwargs.get("stop", []))
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE, kwargs.get("temperature", 0.4))
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P, kwargs.get("top_p", 1.0))
+
+ # Add token-related attributes if available.
+ if "finish_reason" in tokens:
+ span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON, [tokens["finish_reason"]])
+ if "response_id" in tokens:
+ span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_ID, tokens["response_id"])
+ if "input_tokens" in tokens:
+ span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS, tokens["input_tokens"])
+ if "output_tokens" in tokens:
+ span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS, tokens["output_tokens"])
+ if "total_tokens" in tokens:
+ span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS, tokens["total_tokens"])
+
+ span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_MODEL, request_model)
+ span.set_attribute(SemanticConvetion.SERVER_ADDRESS, server_address)
+ # Environment and service identifiers.
+ span.set_attribute(DEPLOYMENT_ENVIRONMENT, environment)
+ span.set_attribute(SERVICE_NAME, application_name)
+ # Set any extra attributes passed in.
+ for key, value in extra_attrs.items():
+ span.set_attribute(key, value)
+
+def record_common_metrics(metrics, application_name, environment, request_model,
+ server_address, server_port, start_time, end_time,
+ input_tokens, output_tokens, cost, include_tbt=False, tbt_value=None):
+ """
+ Record common metrics for the operation.
+ """
+
+ attributes = create_metrics_attributes(
+ service_name=application_name,
+ deployment_environment=environment,
+ operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
+ system=SemanticConvetion.GEN_AI_SYSTEM_AI21,
+ request_model=request_model,
+ server_address=server_address,
+ server_port=server_port,
+ response_model=request_model,
+ )
+ metrics["genai_client_usage_tokens"].record(input_tokens + output_tokens, attributes)
+ metrics["genai_client_operation_duration"].record(end_time - start_time, attributes)
+ if include_tbt and tbt_value is not None:
+ metrics["genai_server_tbt"].record(tbt_value, attributes)
+ metrics["genai_server_ttft"].record(end_time - start_time, attributes)
+ metrics["genai_requests"].add(1, attributes)
+ metrics["genai_completion_tokens"].add(output_tokens, attributes)
+ metrics["genai_prompt_tokens"].add(input_tokens, attributes)
+ metrics["genai_cost"].record(cost, attributes)
+
+def emit_common_events(event_provider, choices, finish_reason, llmresponse, formatted_messages,
+ capture_message_content, n):
+ """
+ Emit events common to both chat and chat rag operations.
+ """
+
+ if n > 1:
+ for choice in choices:
+ choice_event_body = {
+ "finish_reason": finish_reason,
+ "index": choice.get('index', 0),
+ "message": {
+ **({"content": choice.get('message', {}).get('content', '')} if capture_message_content else {}),
+ "role": choice.get('message', {}).get('role', 'assistant')
+ }
+ }
+ # If tool calls exist, emit an event for each tool call.
+ tool_calls = choice.get('message', {}).get('tool_calls')
+ if tool_calls:
+ for tool_call in tool_calls:
+ choice_event_body["message"].update({
+ "tool_calls": {
+ "function": {
+ "name": tool_call.get('function', {}).get('name', ''),
+ "arguments": tool_call.get('function', {}).get('arguments', '')
+ },
+ "id": tool_call.get('id', ''),
+ "type": tool_call.get('type', 'function')
+ }
+ })
+ event = otel_event(
+ name=SemanticConvetion.GEN_AI_CHOICE,
+ attributes={SemanticConvetion.GEN_AI_SYSTEM: SemanticConvetion.GEN_AI_SYSTEM_AI21},
+ body=choice_event_body
+ )
+ event_provider.emit(event)
+ else:
+ event = otel_event(
+ name=SemanticConvetion.GEN_AI_CHOICE,
+ attributes={SemanticConvetion.GEN_AI_SYSTEM: SemanticConvetion.GEN_AI_SYSTEM_AI21},
+ body=choice_event_body
+ )
+ event_provider.emit(event)
+ else:
+ # Single choice case.
+ choice_event_body = {
+ "finish_reason": finish_reason,
+ "index": 0,
+ "message": {
+ **({"content": llmresponse} if capture_message_content else {}),
+ "role": 'assistant'
+ }
+ }
+ event = otel_event(
+ name=SemanticConvetion.GEN_AI_CHOICE,
+ attributes={SemanticConvetion.GEN_AI_SYSTEM: SemanticConvetion.GEN_AI_SYSTEM_AI21},
+ body=choice_event_body
+ )
+ event_provider.emit(event)
+
+ # Emit additional role-based events (if formatted messages are available).
+ for role in ['user', 'system', 'assistant', 'tool']:
+ msg = formatted_messages.get(role, {})
+ if msg.get('content', ''):
+ event_body = {
+ **({"content": msg.get('content', '')} if capture_message_content else {}),
+ "role": msg.get('role', [])
+ }
+ # For assistant messages, attach tool call details if they exist.
+ if role == 'assistant' and choices:
+ tool_calls = choices[0].get('message', {}).get('tool_calls', [])
+ if tool_calls:
+ event_body["tool_calls"] = {
+ "function": {
+ "name": tool_calls[0].get('function', {}).get('name', ''),
+ "arguments": tool_calls[0].get('function', {}).get('arguments', '')
+ },
+ "id": tool_calls[0].get('id', ''),
+ "type": "function"
+ }
+ if role == 'tool' and choices:
+ tool_calls = choices[0].get('message', {}).get('tool_calls', [])
+ if tool_calls:
+ event_body["id"] = tool_calls[0].get('id', '')
+ event = otel_event(
+ name=getattr(SemanticConvetion, f'GEN_AI_{role.upper()}_MESSAGE'),
+ attributes={SemanticConvetion.GEN_AI_SYSTEM: SemanticConvetion.GEN_AI_SYSTEM_AI21},
+ body=event_body
+ )
+ event_provider.emit(event)
+
+def process_chunk(self, chunk):
+ """
+ Process a chunk of response data and update state.
+ """
+
+ end_time = time.time()
+ # Record the timestamp for the current chunk.
+ self._timestamps.append(end_time)
+ if len(self._timestamps) == 1:
+ # Calculate time-to-first-chunk (TTFT).
+ self._ttft = calculate_ttft(self._timestamps, self._start_time)
+
+ chunked = response_as_dict(chunk)
+ if (len(chunked.get('choices')) > 0 and
+ 'delta' in chunked.get('choices')[0] and
+ 'content' in chunked.get('choices')[0].get('delta')):
+ content = chunked.get('choices')[0].get('delta').get('content')
+ if content:
+ self._llmresponse += content
+ if chunked.get('usage'):
+ self._input_tokens = chunked.get('usage').get("prompt_tokens")
+ self._output_tokens = chunked.get('usage').get("completion_tokens")
+ self._response_id = chunked.get('id')
+ self._choices += chunked.get('choices')
+ self._finish_reason = chunked.get('choices')[0].get('finish_reason')
+
+def common_chat_logic(scope, pricing_info, environment, application_name, metrics,
+ event_provider, capture_message_content, disable_metrics, version, is_stream):
+ """
+ Process chat request and generate Telemetry.
+ """
+
+ scope._end_time = time.time()
+ if len(scope._timestamps) > 1:
+ scope._tbt = calculate_tbt(scope._timestamps)
+
+ # Extract and format input messages.
+ formatted_messages = extract_and_format_input(scope._kwargs.get("messages", ""))
+ prompt = concatenate_all_contents(formatted_messages)
+ request_model = scope._kwargs.get("model", "jamba-1.5-mini")
+
+ # Calculate cost based on token usage.
+ cost = get_chat_model_cost(request_model, pricing_info, scope._input_tokens, scope._output_tokens)
+ # Prepare tokens dictionary.
+ tokens = {
+ "finish_reason": scope._finish_reason,
+ "response_id": scope._response_id,
+ "input_tokens": scope._input_tokens,
+ "output_tokens": scope._output_tokens,
+ "total_tokens": scope._input_tokens + scope._output_tokens,
+ }
+ extra_attrs = {
+ SemanticConvetion.GEN_AI_REQUEST_IS_STREAM: is_stream,
+ SemanticConvetion.GEN_AI_CLIENT_TOKEN_USAGE: scope._input_tokens + scope._output_tokens,
+ SemanticConvetion.GEN_AI_USAGE_COST: cost,
+ SemanticConvetion.GEN_AI_SERVER_TBT: scope._tbt,
+ SemanticConvetion.GEN_AI_SERVER_TTFT: scope._ttft,
+ SemanticConvetion.GEN_AI_SDK_VERSION: version,
+ SemanticConvetion.GEN_AI_OUTPUT_TYPE: "text" if isinstance(scope._llmresponse, str) else "json"
+ }
+ # Set span attributes.
+ setup_common_span_attributes(scope._span, request_model, scope._kwargs, tokens,
+ scope._server_port, scope._server_address, environment,
+ application_name, extra_attrs)
+
+ # Optionally add events capturing the prompt and completion.
+ if capture_message_content:
+ scope._span.add_event(
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
+ attributes={SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt},
+ )
+ scope._span.add_event(
+ name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
+ attributes={SemanticConvetion.GEN_AI_CONTENT_COMPLETION: scope._llmresponse},
+ )
+
+ # Emit events for each choice and message role.
+ n = scope._kwargs.get('n', 1)
+ emit_common_events(event_provider, scope._choices, scope._finish_reason, scope._llmresponse,
+ formatted_messages, capture_message_content, n)
+
+ scope._span.set_status(Status(StatusCode.OK))
+
+ if not disable_metrics:
+ record_common_metrics(metrics, application_name, environment, request_model,
+ scope._server_address, scope._server_port,
+ scope._start_time, scope._end_time,
+ scope._input_tokens, scope._output_tokens, cost,
+ include_tbt=True, tbt_value=scope._tbt)
+
+def process_streaming_chat_response(self, pricing_info, environment, application_name, metrics,
+ event_provider, capture_message_content=False, disable_metrics=False, version=''):
+ """
+ Process a streaming chat response and generate Telemetry.
+ """
+
+ common_chat_logic(self, pricing_info, environment, application_name, metrics,
+ event_provider, capture_message_content, disable_metrics, version, is_stream=True)
+
+def process_chat_response(response, request_model, pricing_info, server_port, server_address,
+ environment, application_name, metrics, event_provider, start_time,
+ span, capture_message_content=False, disable_metrics=False, version="1.0.0", **kwargs):
+ """
+ Process a synchronous chat response and generate Telemetry.
+ """
+
+ # Create a generic scope object to hold telemetry data.
+ self = type('GenericScope', (), {})()
+
+ # pylint: disable = no-member
+ self._start_time = start_time
+ self._end_time = time.time()
+
+ self._span = span
+ # Concatenate content from all choices.
+ self._llmresponse = ''.join(
+ (choice.get('message', {}).get('content') or '')
+ for choice in response.get('choices', [])
+ )
+ self._response_role = response.get('message', {}).get('role', 'assistant')
+ self._input_tokens = response.get('usage', {}).get('prompt_tokens', 0)
+ self._output_tokens = response.get('usage', {}).get('completion_tokens', 0)
+ self._response_id = response.get('id', '')
+ self._response_model = request_model
+ self._finish_reason = response.get('choices', [{}])[0].get('finish_reason')
+ self._timestamps = []
+ self._ttft, self._tbt = self._end_time - self._start_time, 0
+ self._server_address, self._server_port = server_address, server_port
+ self._kwargs = kwargs
+ self._choices = response.get('choices')
+
+ common_chat_logic(self, pricing_info, environment, application_name, metrics,
+ event_provider, capture_message_content, disable_metrics, version, is_stream=False)
+
+ return response
+
+def process_chat_rag_response(response, request_model, pricing_info, server_port, server_address,
+ environment, application_name, metrics, event_provider, start_time,
+ span, capture_message_content=False, disable_metrics=False, version="1.0.0", **kwargs):
+ """
+ Process a chat response and generate Telemetry.
+ """
+ end_time = time.time()
+ try:
+ # Format input messages into a single prompt string.
+ messages_input = kwargs.get("messages", "")
+ formatted_messages = extract_and_format_input(messages_input)
+ prompt = concatenate_all_contents(formatted_messages)
+ input_tokens = general_tokens(prompt)
+
+ # Create tokens dict and RAG-specific extra attributes.
+ tokens = {"response_id": response.get("id"), "input_tokens": input_tokens}
+ extra_attrs = {
+ SemanticConvetion.GEN_AI_REQUEST_IS_STREAM: False,
+ SemanticConvetion.GEN_AI_SERVER_TTFT: end_time - start_time,
+ SemanticConvetion.GEN_AI_SDK_VERSION: version,
+ SemanticConvetion.GEN_AI_RAG_MAX_SEGMENTS: kwargs.get("max_segments", -1),
+ SemanticConvetion.GEN_AI_RAG_STRATEGY: kwargs.get("retrieval_strategy", "segments"),
+ SemanticConvetion.GEN_AI_RAG_SIMILARITY_THRESHOLD: kwargs.get("retrieval_similarity_threshold", -1),
+ SemanticConvetion.GEN_AI_RAG_MAX_NEIGHBORS: kwargs.get("max_neighbors", -1),
+ SemanticConvetion.GEN_AI_RAG_FILE_IDS: str(kwargs.get("file_ids", "")),
+ SemanticConvetion.GEN_AI_RAG_DOCUMENTS_PATH: kwargs.get("path", "")
+ }
+ # Set common span attributes.
+ setup_common_span_attributes(span, request_model, kwargs, tokens,
+ server_port, server_address, environment, application_name,
+ extra_attrs)
+
+ # Record the prompt event if requested.
+ if capture_message_content:
+ span.add_event(
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
+ attributes={SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt},
+ )
+
+ output_tokens = 0
+ choices = response.get('choices', [])
+ # Instead of adding a separate event per choice, we aggregate all completion content.
+ aggregated_completion = []
+ for i in range(kwargs.get('n', 1)):
+ # Get the response content from each choice and count tokens.
+ content = choices[i].get('content', '')
+ aggregated_completion.append(content)
+ output_tokens += general_tokens(content)
+ if kwargs.get('tools'):
+ span.set_attribute(SemanticConvetion.GEN_AI_TOOL_CALLS,
+ str(choices[i].get('message', {}).get('tool_calls')))
+ # Set output type based on actual content type.
+ if isinstance(content, str):
+ span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE, "text")
+ elif content is not None:
+ span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE, "json")
+
+ # Concatenate completion responses.
+ llmresponse = ''.join(aggregated_completion)
+ tokens["output_tokens"] = output_tokens
+ tokens["total_tokens"] = input_tokens + output_tokens
+
+ cost = get_chat_model_cost(request_model, pricing_info, input_tokens, output_tokens)
+ span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST, cost)
+ span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS, output_tokens)
+ span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS, input_tokens + output_tokens)
+
+ span.set_status(Status(StatusCode.OK))
+ # Emit a single aggregated completion event.
+ if capture_message_content:
+ span.add_event(
+ name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
+ attributes={SemanticConvetion.GEN_AI_CONTENT_COMPLETION: llmresponse},
+ )
+ # Emit the rest of the events (choice and role-based events) as before.
+ n = kwargs.get('n', 1)
+ emit_common_events(event_provider, choices, choices[0].get('finish_reason', ''),
+ llmresponse, formatted_messages, capture_message_content, n)
+
+ if not disable_metrics:
+ record_common_metrics(metrics, application_name, environment, request_model,
+ server_address, server_port, start_time, end_time,
+ input_tokens, output_tokens, cost, include_tbt=False)
+ return response
+
+ except Exception as e:
+ handle_exception(span, e)
+ return response
diff --git a/sdk/python/src/openlit/instrumentation/anthropic/__init__.py b/sdk/python/src/openlit/instrumentation/anthropic/__init__.py
index 4079cba6..dd3b7091 100644
--- a/sdk/python/src/openlit/instrumentation/anthropic/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/anthropic/__init__.py
@@ -25,7 +25,7 @@ class AnthropicInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("anthropic")
@@ -34,7 +34,7 @@ class AnthropicInstrumentor(BaseInstrumentor):
"anthropic.resources.messages",
"Messages.create",
messages(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
#async
@@ -42,7 +42,7 @@ class AnthropicInstrumentor(BaseInstrumentor):
"anthropic.resources.messages",
"AsyncMessages.create",
async_messages(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/anthropic/anthropic.py b/sdk/python/src/openlit/instrumentation/anthropic/anthropic.py
index 27c0d5ba..ebdfbf1c 100644
--- a/sdk/python/src/openlit/instrumentation/anthropic/anthropic.py
+++ b/sdk/python/src/openlit/instrumentation/anthropic/anthropic.py
@@ -21,7 +21,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def messages(version, environment, application_name, tracer,
- pricing_info, trace_content, metrics, disable_metrics):
+ pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for messages to collect metrics.
@@ -31,7 +31,7 @@ def messages(version, environment, application_name, tracer,
application_name: Name of the application using the Anthropic API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Anthropic usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat method to add telemetry.
@@ -211,7 +211,7 @@ def messages(version, environment, application_name, tracer,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -386,7 +386,7 @@ def messages(version, environment, application_name, tracer,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/anthropic/async_anthropic.py b/sdk/python/src/openlit/instrumentation/anthropic/async_anthropic.py
index a9bfb1ec..bc960d69 100644
--- a/sdk/python/src/openlit/instrumentation/anthropic/async_anthropic.py
+++ b/sdk/python/src/openlit/instrumentation/anthropic/async_anthropic.py
@@ -21,7 +21,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def async_messages(version, environment, application_name, tracer,
- pricing_info, trace_content, metrics, disable_metrics):
+ pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for messages to collect metrics.
@@ -31,7 +31,7 @@ def async_messages(version, environment, application_name, tracer,
application_name: Name of the application using the Anthropic API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Anthropic usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat method to add telemetry.
@@ -211,7 +211,7 @@ def async_messages(version, environment, application_name, tracer,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -386,7 +386,7 @@ def async_messages(version, environment, application_name, tracer,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/assemblyai/__init__.py b/sdk/python/src/openlit/instrumentation/assemblyai/__init__.py
index 2280414f..8f3c25d7 100644
--- a/sdk/python/src/openlit/instrumentation/assemblyai/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/assemblyai/__init__.py
@@ -26,7 +26,7 @@ class AssemblyAIInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("assemblyai")
@@ -35,7 +35,7 @@ class AssemblyAIInstrumentor(BaseInstrumentor):
"assemblyai.transcriber",
"Transcriber.transcribe",
transcribe(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/assemblyai/assemblyai.py b/sdk/python/src/openlit/instrumentation/assemblyai/assemblyai.py
index 62db15c0..aabcce39 100644
--- a/sdk/python/src/openlit/instrumentation/assemblyai/assemblyai.py
+++ b/sdk/python/src/openlit/instrumentation/assemblyai/assemblyai.py
@@ -18,7 +18,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def transcribe(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for creating speech audio to collect metrics.
@@ -29,7 +29,7 @@ def transcribe(version, environment, application_name,
application_name: Name of the application using the Assembly AI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of generating speech audio.
- trace_content: Flag indicating whether to trace the input text and generated audio.
+ capture_message_content: Flag indicating whether to trace the input text and generated audio.
Returns:
A function that wraps the speech audio creation method to add telemetry.
@@ -96,7 +96,7 @@ def transcribe(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/astra/__init__.py b/sdk/python/src/openlit/instrumentation/astra/__init__.py
index 098eccd8..ccfde74d 100644
--- a/sdk/python/src/openlit/instrumentation/astra/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/astra/__init__.py
@@ -22,7 +22,7 @@ class AstraInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info")
- trace_content = kwargs.get("trace_content")
+ capture_message_content = kwargs.get("capture_message_content")
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("astrapy")
@@ -31,73 +31,73 @@ class AstraInstrumentor(BaseInstrumentor):
"astrapy.database",
"Database.create_collection",
general_wrap("astra.create_collection", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.database",
"Database.drop_collection",
general_wrap("astra.drop_collection", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"Collection.insert_one",
general_wrap("astra.insert_one", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"Collection.insert_many",
general_wrap("astra.insert_many", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"Collection.update_one",
general_wrap("astra.update_one", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"Collection.update_many",
general_wrap("astra.update_many", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"Collection.find_one_and_update",
general_wrap("astra.find_one_and_update", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"Collection.find",
general_wrap("astra.find", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"Collection.replace_one",
general_wrap("astra.replace_one", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"Collection.find_one_and_delete",
general_wrap("astra.find_one_and_delete", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"Collection.delete_one",
general_wrap("astra.delete_one", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"Collection.delete_many",
general_wrap("astra.delete_many", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# ASync
@@ -105,73 +105,73 @@ class AstraInstrumentor(BaseInstrumentor):
"astrapy.database",
"AsyncDatabase.create_collection",
general_wrap("astra.create_collection", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.database",
"AsyncDatabase.drop_collection",
general_wrap("astra.drop_collection", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"AsyncCollection.insert_one",
general_wrap("astra.insert_one", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"AsyncCollection.insert_many",
general_wrap("astra.insert_many", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"AsyncCollection.update_one",
general_wrap("astra.update_one", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"AsyncCollection.update_many",
general_wrap("astra.update_many", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"AsyncCollection.find_one_and_update",
general_wrap("astra.find_one_and_update", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"AsyncCollection.find",
general_wrap("astra.find", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"AsyncCollection.replace_one",
general_wrap("astra.replace_one", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"AsyncCollection.find_one_and_delete",
general_wrap("astra.find_one_and_delete", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"AsyncCollection.delete_one",
general_wrap("astra.delete_one", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"astrapy.collection",
"AsyncCollection.delete_many",
general_wrap("astra.delete_many", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
@staticmethod
diff --git a/sdk/python/src/openlit/instrumentation/astra/astra.py b/sdk/python/src/openlit/instrumentation/astra/astra.py
index ca9bf226..25b22560 100644
--- a/sdk/python/src/openlit/instrumentation/astra/astra.py
+++ b/sdk/python/src/openlit/instrumentation/astra/astra.py
@@ -20,7 +20,7 @@ def object_count(obj):
return len(obj) if obj else None
def general_wrap(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Wraps a AstraDB operation to trace and log its execution metrics.
@@ -37,7 +37,7 @@ def general_wrap(gen_ai_endpoint, version, environment, application_name,
- application_name (str): The name of the application performing the AstraDB operation.
- tracer (opentelemetry.trace.Tracer): An object used for OpenTelemetry tracing.
- pricing_info (dict): Information about pricing, not used in current implementation.
- - trace_content (bool): A flag indicating whether the content of responses should be traced.
+ - capture_message_content (bool): A flag indicating whether the content of responses should be traced.
Returns:
- function: A decorator function that, when applied, wraps the target function with
diff --git a/sdk/python/src/openlit/instrumentation/astra/async_astra.py b/sdk/python/src/openlit/instrumentation/astra/async_astra.py
index 464a297f..03896060 100644
--- a/sdk/python/src/openlit/instrumentation/astra/async_astra.py
+++ b/sdk/python/src/openlit/instrumentation/astra/async_astra.py
@@ -20,7 +20,7 @@ def object_count(obj):
return len(obj) if obj else None
def general_wrap(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Wraps a AstraDB operation to trace and log its execution metrics.
@@ -37,7 +37,7 @@ def general_wrap(gen_ai_endpoint, version, environment, application_name,
- application_name (str): The name of the application performing the AstraDB operation.
- tracer (opentelemetry.trace.Tracer): An object used for OpenTelemetry tracing.
- pricing_info (dict): Information about pricing, not used in current implementation.
- - trace_content (bool): A flag indicating whether the content of responses should be traced.
+ - capture_message_content (bool): A flag indicating whether the content of responses should be traced.
Returns:
- function: A decorator function that, when applied, wraps the target function with
diff --git a/sdk/python/src/openlit/instrumentation/azure_ai_inference/__init__.py b/sdk/python/src/openlit/instrumentation/azure_ai_inference/__init__.py
index cb499bed..165d6f29 100644
--- a/sdk/python/src/openlit/instrumentation/azure_ai_inference/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/azure_ai_inference/__init__.py
@@ -30,7 +30,7 @@ class AzureAIInferenceInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("azure-ai-inference")
@@ -39,7 +39,7 @@ class AzureAIInferenceInstrumentor(BaseInstrumentor):
"azure.ai.inference",
"ChatCompletionsClient.complete",
complete(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# sync embedding
@@ -47,7 +47,7 @@ class AzureAIInferenceInstrumentor(BaseInstrumentor):
"azure.ai.inference",
"EmbeddingsClient.embed",
embedding(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# async generate
@@ -55,7 +55,7 @@ class AzureAIInferenceInstrumentor(BaseInstrumentor):
"azure.ai.inference.aio",
"ChatCompletionsClient.complete",
async_complete(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# async embedding
@@ -63,7 +63,7 @@ class AzureAIInferenceInstrumentor(BaseInstrumentor):
"azure.ai.inference.aio",
"EmbeddingsClient.embed",
async_embedding(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/azure_ai_inference/async_azure_ai_inference.py b/sdk/python/src/openlit/instrumentation/azure_ai_inference/async_azure_ai_inference.py
index 7541a711..429473ed 100644
--- a/sdk/python/src/openlit/instrumentation/azure_ai_inference/async_azure_ai_inference.py
+++ b/sdk/python/src/openlit/instrumentation/azure_ai_inference/async_azure_ai_inference.py
@@ -23,7 +23,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def async_complete(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat to collect metrics.
@@ -33,7 +33,7 @@ def async_complete(version, environment, application_name,
application_name: Name of the application using the Azure AI Inference API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Azure AI Inference usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat method to add telemetry.
@@ -217,7 +217,7 @@ def async_complete(version, environment, application_name,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -388,7 +388,7 @@ def async_complete(version, environment, application_name,
end_time - start_time)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -399,7 +399,7 @@ def async_complete(version, environment, application_name,
for i in range(kwargs.get('n',1)):
span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
[response_dict.get('choices')[i].get('finish_reason')])
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
attributes={
@@ -459,7 +459,7 @@ def async_complete(version, environment, application_name,
return wrapper
def async_embedding(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for embeddings to collect metrics.
@@ -469,7 +469,7 @@ def async_embedding(version, environment, application_name,
application_name: Name of the application using the Azure Inference API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Azure Inference usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the embeddings method to add telemetry.
@@ -541,7 +541,7 @@ def async_embedding(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/azure_ai_inference/azure_ai_inference.py b/sdk/python/src/openlit/instrumentation/azure_ai_inference/azure_ai_inference.py
index 064c3908..14578044 100644
--- a/sdk/python/src/openlit/instrumentation/azure_ai_inference/azure_ai_inference.py
+++ b/sdk/python/src/openlit/instrumentation/azure_ai_inference/azure_ai_inference.py
@@ -23,7 +23,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def complete(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat to collect metrics.
@@ -33,7 +33,7 @@ def complete(version, environment, application_name,
application_name: Name of the application using the Azure AI Inference API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Azure AI Inference usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat method to add telemetry.
@@ -217,7 +217,7 @@ def complete(version, environment, application_name,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -388,7 +388,7 @@ def complete(version, environment, application_name,
end_time - start_time)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -399,7 +399,7 @@ def complete(version, environment, application_name,
for i in range(kwargs.get('n',1)):
span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
[response_dict.get('choices')[i].get('finish_reason')])
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
attributes={
@@ -459,7 +459,7 @@ def complete(version, environment, application_name,
return wrapper
def embedding(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for embeddings to collect metrics.
@@ -469,7 +469,7 @@ def embedding(version, environment, application_name,
application_name: Name of the application using the Azure Inference API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Azure Inference usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the embeddings method to add telemetry.
@@ -541,7 +541,7 @@ def embedding(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/bedrock/__init__.py b/sdk/python/src/openlit/instrumentation/bedrock/__init__.py
index 71a1f1b8..d0df6f7f 100644
--- a/sdk/python/src/openlit/instrumentation/bedrock/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/bedrock/__init__.py
@@ -24,7 +24,7 @@ class BedrockInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("boto3")
@@ -33,7 +33,7 @@ class BedrockInstrumentor(BaseInstrumentor):
"botocore.client",
"ClientCreator.create_client",
converse(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/bedrock/bedrock.py b/sdk/python/src/openlit/instrumentation/bedrock/bedrock.py
index 8cda881d..69f488b4 100644
--- a/sdk/python/src/openlit/instrumentation/bedrock/bedrock.py
+++ b/sdk/python/src/openlit/instrumentation/bedrock/bedrock.py
@@ -53,7 +53,7 @@ class CustomStreamWrapper(StreamingBody):
return data_chunk
def converse(version, environment, application_name, tracer,
- pricing_info, trace_content, metrics, disable_metrics):
+ pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for messages to collect metrics.
@@ -64,7 +64,7 @@ def converse(version, environment, application_name, tracer,
application_name: Name of the application using the Bedrock API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information for calculating Bedrock usage cost.
- trace_content: Whether to trace the actual content.
+ capture_message_content: Whether to trace the actual content.
metrics: Metrics collector.
disable_metrics: Flag to toggle metrics collection.
Returns:
@@ -194,7 +194,7 @@ def converse(version, environment, application_name, tracer,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/chroma/__init__.py b/sdk/python/src/openlit/instrumentation/chroma/__init__.py
index 320bc861..d18f8b88 100644
--- a/sdk/python/src/openlit/instrumentation/chroma/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/chroma/__init__.py
@@ -21,7 +21,7 @@ class ChromaInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info")
- trace_content = kwargs.get("trace_content")
+ capture_message_content = kwargs.get("capture_message_content")
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("chromadb")
@@ -29,55 +29,55 @@ class ChromaInstrumentor(BaseInstrumentor):
"chromadb.db",
"DB.create_collection",
general_wrap("chroma.create_collection", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"chromadb",
"Collection.add",
general_wrap("chroma.add", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"chromadb",
"Collection.get",
general_wrap("chroma.get", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"chromadb",
"Collection.peek",
general_wrap("chroma.peek", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"chromadb",
"Collection.query",
general_wrap("chroma.query", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"chromadb",
"Collection.update",
general_wrap("chroma.update", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"chromadb",
"Collection.upsert",
general_wrap("chroma.upsert", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"chromadb",
"Collection.delete",
general_wrap("chroma.delete", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
diff --git a/sdk/python/src/openlit/instrumentation/chroma/chroma.py b/sdk/python/src/openlit/instrumentation/chroma/chroma.py
index 3e45600b..714ecd71 100644
--- a/sdk/python/src/openlit/instrumentation/chroma/chroma.py
+++ b/sdk/python/src/openlit/instrumentation/chroma/chroma.py
@@ -25,7 +25,7 @@ def object_count(obj):
return cnt
def general_wrap(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Creates a wrapper around a function call to trace and log its execution metrics.
@@ -39,7 +39,7 @@ def general_wrap(gen_ai_endpoint, version, environment, application_name,
- application_name (str): Name of the Langchain application.
- tracer (opentelemetry.trace.Tracer): The tracer object used for OpenTelemetry tracing.
- pricing_info (dict): Information about the pricing for internal metrics (currently not used).
- - trace_content (bool): Flag indicating whether to trace the content of the response.
+ - capture_message_content (bool): Flag indicating whether to trace the content of the response.
Returns:
- function: A higher-order function that takes a function 'wrapped' and returns
diff --git a/sdk/python/src/openlit/instrumentation/cohere/__init__.py b/sdk/python/src/openlit/instrumentation/cohere/__init__.py
index 89f728de..b83449ea 100644
--- a/sdk/python/src/openlit/instrumentation/cohere/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/cohere/__init__.py
@@ -23,7 +23,7 @@ class CohereInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info")
- trace_content = kwargs.get("trace_content")
+ capture_message_content = kwargs.get("capture_message_content")
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("cohere")
@@ -32,19 +32,19 @@ class CohereInstrumentor(BaseInstrumentor):
"cohere.client_v2",
"ClientV2.chat",
chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"cohere.client_v2",
"ClientV2.chat_stream",
chat_stream(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"cohere.client_v2",
"ClientV2.embed",
embed(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# Async Client
@@ -52,21 +52,21 @@ class CohereInstrumentor(BaseInstrumentor):
"cohere.client_v2",
"AsyncClientV2.chat",
async_chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"cohere.client_v2",
"AsyncClientV2.chat_stream",
async_chat_stream(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"cohere.client_v2",
"AsyncClientV2.embed",
async_embed(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
@staticmethod
diff --git a/sdk/python/src/openlit/instrumentation/cohere/async_cohere.py b/sdk/python/src/openlit/instrumentation/cohere/async_cohere.py
index 45d58b16..0049d75f 100644
--- a/sdk/python/src/openlit/instrumentation/cohere/async_cohere.py
+++ b/sdk/python/src/openlit/instrumentation/cohere/async_cohere.py
@@ -22,7 +22,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def async_embed(version, environment, application_name, tracer,
- pricing_info, trace_content, metrics, disable_metrics):
+ pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for embeddings to collect metrics.
@@ -32,7 +32,7 @@ def async_embed(version, environment, application_name, tracer,
application_name: Name of the application using the Cohere API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Cohere usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the embeddings method to add telemetry.
@@ -105,7 +105,7 @@ def async_embed(version, environment, application_name, tracer,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -149,7 +149,7 @@ def async_embed(version, environment, application_name, tracer,
return wrapper
def async_chat(version, environment, application_name, tracer,
- pricing_info, trace_content, metrics, disable_metrics):
+ pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat to collect metrics.
@@ -159,7 +159,7 @@ def async_chat(version, environment, application_name, tracer,
application_name: Name of the application using the Cohere API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Cohere usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat method to add telemetry.
@@ -281,7 +281,7 @@ def async_chat(version, environment, application_name, tracer,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -336,7 +336,7 @@ def async_chat(version, environment, application_name, tracer,
return wrapper
def async_chat_stream(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat_stream to collect metrics.
@@ -346,7 +346,7 @@ def async_chat_stream(version, environment, application_name,
application_name: Name of the application using the Cohere API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Cohere usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat method to add telemetry.
@@ -547,7 +547,7 @@ def async_chat_stream(version, environment, application_name,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/cohere/cohere.py b/sdk/python/src/openlit/instrumentation/cohere/cohere.py
index 46c34333..a9db0dd5 100644
--- a/sdk/python/src/openlit/instrumentation/cohere/cohere.py
+++ b/sdk/python/src/openlit/instrumentation/cohere/cohere.py
@@ -22,7 +22,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def embed(version, environment, application_name, tracer,
- pricing_info, trace_content, metrics, disable_metrics):
+ pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for embeddings to collect metrics.
@@ -32,7 +32,7 @@ def embed(version, environment, application_name, tracer,
application_name: Name of the application using the Cohere API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Cohere usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the embeddings method to add telemetry.
@@ -105,7 +105,7 @@ def embed(version, environment, application_name, tracer,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -149,7 +149,7 @@ def embed(version, environment, application_name, tracer,
return wrapper
def chat(version, environment, application_name, tracer,
- pricing_info, trace_content, metrics, disable_metrics):
+ pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat to collect metrics.
@@ -159,7 +159,7 @@ def chat(version, environment, application_name, tracer,
application_name: Name of the application using the Cohere API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Cohere usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat method to add telemetry.
@@ -281,7 +281,7 @@ def chat(version, environment, application_name, tracer,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -336,7 +336,7 @@ def chat(version, environment, application_name, tracer,
return wrapper
def chat_stream(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat_stream to collect metrics.
@@ -346,7 +346,7 @@ def chat_stream(version, environment, application_name,
application_name: Name of the application using the Cohere API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Cohere usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat method to add telemetry.
@@ -547,7 +547,7 @@ def chat_stream(version, environment, application_name,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/controlflow/__init__.py b/sdk/python/src/openlit/instrumentation/controlflow/__init__.py
index 49eef1de..a7b97f7c 100644
--- a/sdk/python/src/openlit/instrumentation/controlflow/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/controlflow/__init__.py
@@ -26,7 +26,7 @@ class ControlFlowInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("controlflow")
@@ -34,21 +34,21 @@ class ControlFlowInstrumentor(BaseInstrumentor):
"controlflow.agents.agent",
"Agent.__init__",
wrap_controlflow("controlflow.create_agent", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"controlflow.tasks.task",
"Task.__init__",
wrap_controlflow("controlflow.create_task", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"controlflow",
"run",
wrap_controlflow("controlflow.run", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/controlflow/controlflow.py b/sdk/python/src/openlit/instrumentation/controlflow/controlflow.py
index a6be6ac4..232ca2e8 100644
--- a/sdk/python/src/openlit/instrumentation/controlflow/controlflow.py
+++ b/sdk/python/src/openlit/instrumentation/controlflow/controlflow.py
@@ -13,7 +13,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def wrap_controlflow(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Creates a wrapper around a function call to trace and log its execution metrics.
@@ -27,7 +27,7 @@ def wrap_controlflow(gen_ai_endpoint, version, environment, application_name,
- application_name (str): Name of the Langchain application.
- tracer (opentelemetry.trace.Tracer): The tracer object used for OpenTelemetry tracing.
- pricing_info (dict): Information about the pricing for internal metrics (currently not used).
- - trace_content (bool): Flag indicating whether to trace the content of the response.
+ - capture_message_content (bool): Flag indicating whether to trace the content of the response.
Returns:
- function: A higher-order function that takes a function 'wrapped' and returns
diff --git a/sdk/python/src/openlit/instrumentation/crawl4ai/__init__.py b/sdk/python/src/openlit/instrumentation/crawl4ai/__init__.py
index dc1816c4..13ac1d52 100644
--- a/sdk/python/src/openlit/instrumentation/crawl4ai/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/crawl4ai/__init__.py
@@ -29,7 +29,7 @@ class Crawl4AIInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("crawl4ai")
@@ -37,14 +37,14 @@ class Crawl4AIInstrumentor(BaseInstrumentor):
"crawl4ai.web_crawler",
"WebCrawler.run",
wrap_crawl("crawl4ai.web_crawl", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"crawl4ai.async_webcrawler",
"AsyncWebCrawler.arun",
async_wrap_crawl("crawl4ai.web_crawl", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/crawl4ai/async_crawl4ai.py b/sdk/python/src/openlit/instrumentation/crawl4ai/async_crawl4ai.py
index 2e95f0d6..890e7693 100644
--- a/sdk/python/src/openlit/instrumentation/crawl4ai/async_crawl4ai.py
+++ b/sdk/python/src/openlit/instrumentation/crawl4ai/async_crawl4ai.py
@@ -15,7 +15,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def async_wrap_crawl(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -26,7 +26,7 @@ def async_wrap_crawl(gen_ai_endpoint, version, environment, application_name,
application_name: Name of the application using the Crawl4AI Agent.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Crawl4AI usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
diff --git a/sdk/python/src/openlit/instrumentation/crawl4ai/crawl4ai.py b/sdk/python/src/openlit/instrumentation/crawl4ai/crawl4ai.py
index 3307d572..f4fe50b3 100644
--- a/sdk/python/src/openlit/instrumentation/crawl4ai/crawl4ai.py
+++ b/sdk/python/src/openlit/instrumentation/crawl4ai/crawl4ai.py
@@ -15,7 +15,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def wrap_crawl(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -26,7 +26,7 @@ def wrap_crawl(gen_ai_endpoint, version, environment, application_name,
application_name: Name of the application using the Crawl4AI Agent.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Crawl4AI usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
diff --git a/sdk/python/src/openlit/instrumentation/crewai/__init__.py b/sdk/python/src/openlit/instrumentation/crewai/__init__.py
index 83b34340..9957a81d 100644
--- a/sdk/python/src/openlit/instrumentation/crewai/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/crewai/__init__.py
@@ -26,7 +26,7 @@ class CrewAIInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("crewai")
@@ -34,14 +34,14 @@ class CrewAIInstrumentor(BaseInstrumentor):
"crewai.agent",
"Agent.execute_task",
crew_wrap("crewai.agent_execute_task", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"crewai.task",
"Task._execute_core",
crew_wrap("crewai.task_execute_core", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
diff --git a/sdk/python/src/openlit/instrumentation/crewai/crewai.py b/sdk/python/src/openlit/instrumentation/crewai/crewai.py
index f6a1da7d..75f946ae 100644
--- a/sdk/python/src/openlit/instrumentation/crewai/crewai.py
+++ b/sdk/python/src/openlit/instrumentation/crewai/crewai.py
@@ -28,7 +28,7 @@ def _parse_tools(tools):
return json.dumps(result)
def crew_wrap(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -39,7 +39,7 @@ def crew_wrap(gen_ai_endpoint, version, environment, application_name,
application_name: Name of the application using the CrewAI Agent.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of CrewAI usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
diff --git a/sdk/python/src/openlit/instrumentation/dynamiq/__init__.py b/sdk/python/src/openlit/instrumentation/dynamiq/__init__.py
index baee93e7..c493e94f 100644
--- a/sdk/python/src/openlit/instrumentation/dynamiq/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/dynamiq/__init__.py
@@ -26,7 +26,7 @@ class DynamiqInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("dynamiq")
@@ -34,28 +34,28 @@ class DynamiqInstrumentor(BaseInstrumentor):
"dynamiq.nodes.agents.base",
"Agent.run",
dynamiq_wrap("dynamiq.agent_run", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"dynamiq",
"Workflow.run",
dynamiq_wrap("dynamiq.workflow_run", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"dynamiq.memory",
"Memory.add",
dynamiq_wrap("dynamiq.memory_add", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"dynamiq.memory",
"Memory.search",
dynamiq_wrap("dynamiq.memory_search", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
diff --git a/sdk/python/src/openlit/instrumentation/dynamiq/dynamiq.py b/sdk/python/src/openlit/instrumentation/dynamiq/dynamiq.py
index 1ea74d89..c110e0e9 100644
--- a/sdk/python/src/openlit/instrumentation/dynamiq/dynamiq.py
+++ b/sdk/python/src/openlit/instrumentation/dynamiq/dynamiq.py
@@ -15,7 +15,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def dynamiq_wrap(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -26,7 +26,7 @@ def dynamiq_wrap(gen_ai_endpoint, version, environment, application_name,
application_name: Name of the application using the dynamiq Agent.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of dynamiq usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
diff --git a/sdk/python/src/openlit/instrumentation/elevenlabs/__init__.py b/sdk/python/src/openlit/instrumentation/elevenlabs/__init__.py
index e1b9477e..ba08ed6a 100644
--- a/sdk/python/src/openlit/instrumentation/elevenlabs/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/elevenlabs/__init__.py
@@ -29,7 +29,7 @@ class ElevenLabsInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("elevenlabs")
@@ -38,7 +38,7 @@ class ElevenLabsInstrumentor(BaseInstrumentor):
"elevenlabs.client",
"ElevenLabs.generate",
generate("elevenlabs.generate", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# sync text_to_speech.convert
@@ -46,7 +46,7 @@ class ElevenLabsInstrumentor(BaseInstrumentor):
"elevenlabs.text_to_speech.client",
"TextToSpeechClient.convert",
generate("elevenlabs.text_to_speech", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# async generate
@@ -54,7 +54,7 @@ class ElevenLabsInstrumentor(BaseInstrumentor):
"elevenlabs.client",
"AsyncElevenLabs.generate",
async_generate("elevenlabs.generate", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# sync text_to_speech.convert
@@ -62,7 +62,7 @@ class ElevenLabsInstrumentor(BaseInstrumentor):
"elevenlabs.text_to_speech.client",
"AsyncTextToSpeechClient.convert",
generate("elevenlabs.text_to_speech", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/elevenlabs/async_elevenlabs.py b/sdk/python/src/openlit/instrumentation/elevenlabs/async_elevenlabs.py
index 2a375d81..564fd9a7 100644
--- a/sdk/python/src/openlit/instrumentation/elevenlabs/async_elevenlabs.py
+++ b/sdk/python/src/openlit/instrumentation/elevenlabs/async_elevenlabs.py
@@ -18,7 +18,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def async_generate(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for creating speech audio to collect metrics.
@@ -28,7 +28,7 @@ def async_generate(gen_ai_endpoint, version, environment, application_name,
application_name: Name of the application using the ElevenLabs API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of generating speech audio.
- trace_content: Flag indicating whether to trace the input text and generated audio.
+ capture_message_content: Flag indicating whether to trace the input text and generated audio.
Returns:
A function that wraps the speech audio creation method to add telemetry.
@@ -105,7 +105,7 @@ def async_generate(gen_ai_endpoint, version, environment, application_name,
cost)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/elevenlabs/elevenlabs.py b/sdk/python/src/openlit/instrumentation/elevenlabs/elevenlabs.py
index 4dc6821a..c46ebd1d 100644
--- a/sdk/python/src/openlit/instrumentation/elevenlabs/elevenlabs.py
+++ b/sdk/python/src/openlit/instrumentation/elevenlabs/elevenlabs.py
@@ -18,7 +18,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def generate(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for creating speech audio to collect metrics.
@@ -28,7 +28,7 @@ def generate(gen_ai_endpoint, version, environment, application_name,
application_name: Name of the application using the ElevenLabs API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of generating speech audio.
- trace_content: Flag indicating whether to trace the input text and generated audio.
+ capture_message_content: Flag indicating whether to trace the input text and generated audio.
Returns:
A function that wraps the speech audio creation method to add telemetry.
@@ -106,7 +106,7 @@ def generate(gen_ai_endpoint, version, environment, application_name,
cost)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/embedchain/__init__.py b/sdk/python/src/openlit/instrumentation/embedchain/__init__.py
index 5177d286..da05f40e 100644
--- a/sdk/python/src/openlit/instrumentation/embedchain/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/embedchain/__init__.py
@@ -35,7 +35,7 @@ class EmbedChainInstrumentor(BaseInstrumentor):
environment = kwargs.get("environment")
tracer = kwargs.get("tracer")
pricing_info = kwargs.get("pricing_info")
- trace_content = kwargs.get("trace_content")
+ capture_message_content = kwargs.get("capture_message_content")
version = importlib.metadata.version("embedchain")
for wrapped_method in WRAPPED_METHODS:
@@ -47,7 +47,7 @@ class EmbedChainInstrumentor(BaseInstrumentor):
wrap_package,
wrap_object,
wrapper(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content),
+ tracer, pricing_info, capture_message_content),
)
@staticmethod
diff --git a/sdk/python/src/openlit/instrumentation/embedchain/embedchain.py b/sdk/python/src/openlit/instrumentation/embedchain/embedchain.py
index b6f13530..afa1c3e8 100644
--- a/sdk/python/src/openlit/instrumentation/embedchain/embedchain.py
+++ b/sdk/python/src/openlit/instrumentation/embedchain/embedchain.py
@@ -13,7 +13,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def evaluate(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content):
+ tracer, pricing_info, capture_message_content):
"""
Creates a wrapper around a function call to trace and log its execution metrics.
@@ -27,7 +27,7 @@ def evaluate(gen_ai_endpoint, version, environment, application_name,
- application_name (str): Name of the EmbedChain application.
- tracer (opentelemetry.trace.Tracer): The tracer object used for OpenTelemetry tracing.
- pricing_info (dict): Information about the pricing for internal metrics (currently not used).
- - trace_content (bool): Flag indicating whether to trace the content of the response.
+ - capture_message_content (bool): Flag indicating whether to trace the content of the response.
Returns:
- function: A higher-order function that takes a function 'wrapped' and returns
@@ -91,7 +91,7 @@ def evaluate(gen_ai_endpoint, version, environment, application_name,
return wrapper
def get_data_sources(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content):
+ tracer, pricing_info, capture_message_content):
"""
Creates a wrapper around a function call to trace and log its execution metrics.
@@ -105,7 +105,7 @@ def get_data_sources(gen_ai_endpoint, version, environment, application_name,
- application_name (str): Name of the EmbedChain application.
- tracer (opentelemetry.trace.Tracer): The tracer object used for OpenTelemetry tracing.
- pricing_info (dict): Information about the pricing for internal metrics (currently not used).
- - trace_content (bool): Flag indicating whether to trace the content of the response.
+ - capture_message_content (bool): Flag indicating whether to trace the content of the response.
Returns:
- function: A higher-order function that takes a function 'wrapped' and returns
diff --git a/sdk/python/src/openlit/instrumentation/firecrawl/__init__.py b/sdk/python/src/openlit/instrumentation/firecrawl/__init__.py
index bf6bf4c8..cf22a727 100644
--- a/sdk/python/src/openlit/instrumentation/firecrawl/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/firecrawl/__init__.py
@@ -26,7 +26,7 @@ class FireCrawlInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("firecrawl-py")
@@ -34,14 +34,14 @@ class FireCrawlInstrumentor(BaseInstrumentor):
"firecrawl.firecrawl",
"FirecrawlApp.scrape_url",
wrap_crawl("firecrawl.scrape_url", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"firecrawl.firecrawl",
"FirecrawlApp.crawl_url",
wrap_crawl("firecrawl.crawl_url", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/firecrawl/firecrawl.py b/sdk/python/src/openlit/instrumentation/firecrawl/firecrawl.py
index 4cbfd9c4..957ce856 100644
--- a/sdk/python/src/openlit/instrumentation/firecrawl/firecrawl.py
+++ b/sdk/python/src/openlit/instrumentation/firecrawl/firecrawl.py
@@ -15,7 +15,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def wrap_crawl(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -26,7 +26,7 @@ def wrap_crawl(gen_ai_endpoint, version, environment, application_name,
application_name: Name of the application using the FireCrawl Agent.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of FireCrawl usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
diff --git a/sdk/python/src/openlit/instrumentation/google_ai_studio/__init__.py b/sdk/python/src/openlit/instrumentation/google_ai_studio/__init__.py
index 2219b009..3e9f8450 100644
--- a/sdk/python/src/openlit/instrumentation/google_ai_studio/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/google_ai_studio/__init__.py
@@ -30,7 +30,7 @@ class GoogleAIStudioInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("google-genai")
@@ -39,7 +39,7 @@ class GoogleAIStudioInstrumentor(BaseInstrumentor):
"google.genai.models",
"Models.generate_content",
generate(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# async generate
@@ -47,7 +47,7 @@ class GoogleAIStudioInstrumentor(BaseInstrumentor):
"google.genai.models",
"AsyncModels.generate_content",
async_generate(version, environment,
- application_name, tracer, pricing_info, trace_content, metrics,
+ application_name, tracer, pricing_info, capture_message_content, metrics,
disable_metrics),
)
diff --git a/sdk/python/src/openlit/instrumentation/google_ai_studio/async_google_ai_studio.py b/sdk/python/src/openlit/instrumentation/google_ai_studio/async_google_ai_studio.py
index bb4721d4..acfe5533 100644
--- a/sdk/python/src/openlit/instrumentation/google_ai_studio/async_google_ai_studio.py
+++ b/sdk/python/src/openlit/instrumentation/google_ai_studio/async_google_ai_studio.py
@@ -19,7 +19,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def async_generate(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat to collect metrics.
@@ -30,7 +30,7 @@ def async_generate(version, environment, application_name,
application_name: Name of the application using the Google AI Studio API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Google AI Studio usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat method to add telemetry.
@@ -165,7 +165,7 @@ def async_generate(version, environment, application_name,
end_time - start_time)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/google_ai_studio/google_ai_studio.py b/sdk/python/src/openlit/instrumentation/google_ai_studio/google_ai_studio.py
index 7a54b4cf..b12974e8 100644
--- a/sdk/python/src/openlit/instrumentation/google_ai_studio/google_ai_studio.py
+++ b/sdk/python/src/openlit/instrumentation/google_ai_studio/google_ai_studio.py
@@ -19,7 +19,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def generate(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat to collect metrics.
@@ -30,7 +30,7 @@ def generate(version, environment, application_name,
application_name: Name of the application using the Google AI Studio API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Google AI Studio usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat method to add telemetry.
@@ -165,7 +165,7 @@ def generate(version, environment, application_name,
end_time - start_time)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/gpt4all/__init__.py b/sdk/python/src/openlit/instrumentation/gpt4all/__init__.py
index 4b4ba150..1b3021d6 100644
--- a/sdk/python/src/openlit/instrumentation/gpt4all/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/gpt4all/__init__.py
@@ -26,7 +26,7 @@ class GPT4AllInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("gpt4all")
@@ -35,7 +35,7 @@ class GPT4AllInstrumentor(BaseInstrumentor):
"gpt4all",
"GPT4All.generate",
generate(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# embed
@@ -43,7 +43,7 @@ class GPT4AllInstrumentor(BaseInstrumentor):
"gpt4all",
"Embed4All.embed",
embed(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
diff --git a/sdk/python/src/openlit/instrumentation/gpt4all/gpt4all.py b/sdk/python/src/openlit/instrumentation/gpt4all/gpt4all.py
index afb166d6..824aea6f 100644
--- a/sdk/python/src/openlit/instrumentation/gpt4all/gpt4all.py
+++ b/sdk/python/src/openlit/instrumentation/gpt4all/gpt4all.py
@@ -20,7 +20,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def generate(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -30,7 +30,7 @@ def generate(version, environment, application_name,
application_name: Name of the application using the GPT4All API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating GPT4All usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
@@ -168,7 +168,7 @@ def generate(version, environment, application_name,
version)
self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
0)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -319,7 +319,7 @@ def generate(version, environment, application_name,
version)
span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
0)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -374,7 +374,7 @@ def generate(version, environment, application_name,
return wrapper
def embed(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for embeddings to collect metrics.
@@ -384,7 +384,7 @@ def embed(version, environment, application_name,
application_name: Name of the application using the GPT4All API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating GPT4All usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the embeddings method to add telemetry.
@@ -452,7 +452,7 @@ def embed(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
0)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/groq/__init__.py b/sdk/python/src/openlit/instrumentation/groq/__init__.py
index 091b154f..1605bfed 100644
--- a/sdk/python/src/openlit/instrumentation/groq/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/groq/__init__.py
@@ -25,7 +25,7 @@ class GroqInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("groq")
@@ -34,7 +34,7 @@ class GroqInstrumentor(BaseInstrumentor):
"groq.resources.chat.completions",
"Completions.create",
chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
#async
@@ -42,7 +42,7 @@ class GroqInstrumentor(BaseInstrumentor):
"groq.resources.chat.completions",
"AsyncCompletions.create",
async_chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/groq/async_groq.py b/sdk/python/src/openlit/instrumentation/groq/async_groq.py
index e5c86f03..fe2bff90 100644
--- a/sdk/python/src/openlit/instrumentation/groq/async_groq.py
+++ b/sdk/python/src/openlit/instrumentation/groq/async_groq.py
@@ -21,7 +21,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def async_chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -31,7 +31,7 @@ def async_chat(version, environment, application_name,
application_name: Name of the application using the Groq API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Groq usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
@@ -221,7 +221,7 @@ def async_chat(version, environment, application_name,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -396,7 +396,7 @@ def async_chat(version, environment, application_name,
end_time - start_time)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -407,7 +407,7 @@ def async_chat(version, environment, application_name,
for i in range(kwargs.get('n',1)):
span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
[response_dict.get('choices')[i].get('finish_reason')])
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/groq/groq.py b/sdk/python/src/openlit/instrumentation/groq/groq.py
index c1e063ea..88f59834 100644
--- a/sdk/python/src/openlit/instrumentation/groq/groq.py
+++ b/sdk/python/src/openlit/instrumentation/groq/groq.py
@@ -21,7 +21,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -31,7 +31,7 @@ def chat(version, environment, application_name,
application_name: Name of the application using the Groq API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Groq usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
@@ -221,7 +221,7 @@ def chat(version, environment, application_name,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -396,7 +396,7 @@ def chat(version, environment, application_name,
end_time - start_time)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -407,7 +407,7 @@ def chat(version, environment, application_name,
for i in range(kwargs.get('n',1)):
span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
[response_dict.get('choices')[i].get('finish_reason')])
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/haystack/__init__.py b/sdk/python/src/openlit/instrumentation/haystack/__init__.py
index 35dccfbf..3553401e 100644
--- a/sdk/python/src/openlit/instrumentation/haystack/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/haystack/__init__.py
@@ -29,7 +29,7 @@ class HaystackInstrumentor(BaseInstrumentor):
environment = kwargs.get("environment")
tracer = kwargs.get("tracer")
pricing_info = kwargs.get("pricing_info")
- trace_content = kwargs.get("trace_content")
+ capture_message_content = kwargs.get("capture_message_content")
version = importlib.metadata.version("haystack-ai")
for wrapped_method in WRAPPED_METHODS:
@@ -41,7 +41,7 @@ class HaystackInstrumentor(BaseInstrumentor):
wrap_package,
wrap_object,
wrapper(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content),
+ tracer, pricing_info, capture_message_content),
)
@staticmethod
diff --git a/sdk/python/src/openlit/instrumentation/haystack/haystack.py b/sdk/python/src/openlit/instrumentation/haystack/haystack.py
index b9c9f6fa..6748e656 100644
--- a/sdk/python/src/openlit/instrumentation/haystack/haystack.py
+++ b/sdk/python/src/openlit/instrumentation/haystack/haystack.py
@@ -13,7 +13,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def join_data(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content):
+ tracer, pricing_info, capture_message_content):
"""
Creates a wrapper around a function call to trace and log its execution metrics.
@@ -27,7 +27,7 @@ def join_data(gen_ai_endpoint, version, environment, application_name,
- application_name (str): Name of the Haystack application.
- tracer (opentelemetry.trace.Tracer): The tracer object used for OpenTelemetry tracing.
- pricing_info (dict): Information about the pricing for internal metrics (currently not used).
- - trace_content (bool): Flag indicating whether to trace the content of the response.
+ - capture_message_content (bool): Flag indicating whether to trace the content of the response.
Returns:
- function: A higher-order function that takes a function 'wrapped' and returns
diff --git a/sdk/python/src/openlit/instrumentation/julep/__init__.py b/sdk/python/src/openlit/instrumentation/julep/__init__.py
index 6a581b9b..40d68a80 100644
--- a/sdk/python/src/openlit/instrumentation/julep/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/julep/__init__.py
@@ -30,7 +30,7 @@ class JulepInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("julep")
@@ -39,19 +39,19 @@ class JulepInstrumentor(BaseInstrumentor):
"julep.resources.agents.agents",
"AgentsResource.create",
wrap_julep("julep.agents_create", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"julep.resources.tasks",
"TasksResource.create",
wrap_julep("julep.task_create", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"julep.resources.executions.executions",
"ExecutionsResource.create",
wrap_julep("julep.execution_create", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# async
@@ -59,19 +59,19 @@ class JulepInstrumentor(BaseInstrumentor):
"julep.resources.agents.agents",
"AsyncAgentsResource.create",
async_wrap_julep("julep.agents_create", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"julep.resources.tasks",
"AsyncTasksResource.create",
async_wrap_julep("julep.task_create", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"julep.resources.executions.executions",
"AsyncExecutionsResource.create",
async_wrap_julep("julep.execution_create", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
diff --git a/sdk/python/src/openlit/instrumentation/julep/async_julep.py b/sdk/python/src/openlit/instrumentation/julep/async_julep.py
index bfe98195..29b1e1eb 100644
--- a/sdk/python/src/openlit/instrumentation/julep/async_julep.py
+++ b/sdk/python/src/openlit/instrumentation/julep/async_julep.py
@@ -13,7 +13,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def async_wrap_julep(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Creates a wrapper around a function call to trace and log its execution metrics.
@@ -27,7 +27,7 @@ def async_wrap_julep(gen_ai_endpoint, version, environment, application_name,
- application_name (str): Name of the application.
- tracer (opentelemetry.trace.Tracer): The tracer object used for OpenTelemetry tracing.
- pricing_info (dict): Information about the pricing for internal metrics (currently not used).
- - trace_content (bool): Flag indicating whether to trace the content of the response.
+ - capture_message_content (bool): Flag indicating whether to trace the content of the response.
Returns:
- function: A higher-order function that takes a function 'wrapped' and returns
@@ -88,7 +88,7 @@ def async_wrap_julep(gen_ai_endpoint, version, environment, application_name,
elif gen_ai_endpoint == "julep.execution_create":
span.set_attribute(SemanticConvetion.GEN_AI_AGENT_TASK_ID,
kwargs.get("task_id", ""))
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/julep/julep.py b/sdk/python/src/openlit/instrumentation/julep/julep.py
index 301bf864..fae7f2de 100644
--- a/sdk/python/src/openlit/instrumentation/julep/julep.py
+++ b/sdk/python/src/openlit/instrumentation/julep/julep.py
@@ -13,7 +13,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def wrap_julep(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Creates a wrapper around a function call to trace and log its execution metrics.
@@ -27,7 +27,7 @@ def wrap_julep(gen_ai_endpoint, version, environment, application_name,
- application_name (str): Name of the Langchain application.
- tracer (opentelemetry.trace.Tracer): The tracer object used for OpenTelemetry tracing.
- pricing_info (dict): Information about the pricing for internal metrics (currently not used).
- - trace_content (bool): Flag indicating whether to trace the content of the response.
+ - capture_message_content (bool): Flag indicating whether to trace the content of the response.
Returns:
- function: A higher-order function that takes a function 'wrapped' and returns
@@ -88,7 +88,7 @@ def wrap_julep(gen_ai_endpoint, version, environment, application_name,
elif gen_ai_endpoint == "julep.execution_create":
span.set_attribute(SemanticConvetion.GEN_AI_AGENT_TASK_ID,
kwargs.get("task_id", ""))
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/langchain/__init__.py b/sdk/python/src/openlit/instrumentation/langchain/__init__.py
index f4ecee85..9674f071 100644
--- a/sdk/python/src/openlit/instrumentation/langchain/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/langchain/__init__.py
@@ -96,7 +96,7 @@ class LangChainInstrumentor(BaseInstrumentor):
environment = kwargs.get("environment")
tracer = kwargs.get("tracer")
pricing_info = kwargs.get("pricing_info")
- trace_content = kwargs.get("trace_content")
+ capture_message_content = kwargs.get("capture_message_content")
metrics = kwargs.get("metrics_dict")
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("langchain")
@@ -110,7 +110,7 @@ class LangChainInstrumentor(BaseInstrumentor):
wrap_package,
wrap_object,
wrapper(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
@staticmethod
diff --git a/sdk/python/src/openlit/instrumentation/langchain/async_langchain.py b/sdk/python/src/openlit/instrumentation/langchain/async_langchain.py
index 24f4a8f5..b40a09ce 100644
--- a/sdk/python/src/openlit/instrumentation/langchain/async_langchain.py
+++ b/sdk/python/src/openlit/instrumentation/langchain/async_langchain.py
@@ -41,7 +41,7 @@ def get_attribute_from_instance_or_kwargs(instance, attribute_name, default=-1):
return default
def async_general_wrap(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Creates a wrapper around a function call to trace and log its execution metrics.
@@ -55,7 +55,7 @@ def async_general_wrap(gen_ai_endpoint, version, environment, application_name,
- application_name (str): Name of the Langchain application.
- tracer (opentelemetry.trace.Tracer): The tracer object used for OpenTelemetry tracing.
- pricing_info (dict): Information about the pricing for internal metrics (currently not used).
- - trace_content (bool): Flag indicating whether to trace the content of the response.
+ - capture_message_content (bool): Flag indicating whether to trace the content of the response.
Returns:
- function: A higher-order function that takes a function 'wrapped' and returns
@@ -114,7 +114,7 @@ def async_general_wrap(gen_ai_endpoint, version, environment, application_name,
return wrapper
def async_hub(gen_ai_endpoint, version, environment, application_name, tracer,
- pricing_info, trace_content, metrics, disable_metrics):
+ pricing_info, capture_message_content, metrics, disable_metrics):
"""
Creates a wrapper around Langchain hub operations for tracing and logging.
@@ -129,7 +129,7 @@ def async_hub(gen_ai_endpoint, version, environment, application_name, tracer,
- application_name (str): Name of the Langchain application.
- tracer (opentelemetry.trace.Tracer): The tracer for OpenTelemetry tracing.
- pricing_info (dict): Pricing information for the operation (not currently used).
- - trace_content (bool): Indicates if the content of the response should be traced.
+ - capture_message_content (bool): Indicates if the content of the response should be traced.
Returns:
- function: A new function that wraps the original hub operation call with added
@@ -189,7 +189,7 @@ def async_hub(gen_ai_endpoint, version, environment, application_name, tracer,
return wrapper
def async_chat(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Creates a wrapper around a function call to trace and log its execution metrics.
@@ -202,7 +202,7 @@ def async_chat(gen_ai_endpoint, version, environment, application_name,
- application_name (str): Name of the Langchain application.
- tracer (opentelemetry.trace.Tracer): The tracer object used for OpenTelemetry tracing.
- pricing_info (dict): Information about the pricing for internal metrics (currently not used).
- - trace_content (bool): Flag indicating whether to trace the content of the response.
+ - capture_message_content (bool): Flag indicating whether to trace the content of the response.
Returns:
- function: A higher-order function that takes a function 'wrapped' and returns
@@ -287,6 +287,11 @@ def async_chat(gen_ai_endpoint, version, environment, application_name,
pricing_info, input_tokens, output_tokens
)
+ try:
+ llm_response = response.content
+ except AttributeError:
+ llm_response = response
+
# Set base span attribues (OTel Semconv)
span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
@@ -328,18 +333,17 @@ def async_chat(gen_ai_endpoint, version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
},
)
- completion_content = getattr(response, 'content', "")
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
attributes={
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: completion_content,
+ SemanticConvetion.GEN_AI_CONTENT_COMPLETION: llm_response,
},
)
diff --git a/sdk/python/src/openlit/instrumentation/langchain/langchain.py b/sdk/python/src/openlit/instrumentation/langchain/langchain.py
index 32535bef..bdece4bb 100644
--- a/sdk/python/src/openlit/instrumentation/langchain/langchain.py
+++ b/sdk/python/src/openlit/instrumentation/langchain/langchain.py
@@ -41,7 +41,7 @@ def get_attribute_from_instance_or_kwargs(instance, attribute_name, default=-1):
return default
def general_wrap(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Creates a wrapper around a function call to trace and log its execution metrics.
@@ -55,7 +55,7 @@ def general_wrap(gen_ai_endpoint, version, environment, application_name,
- application_name (str): Name of the Langchain application.
- tracer (opentelemetry.trace.Tracer): The tracer object used for OpenTelemetry tracing.
- pricing_info (dict): Information about the pricing for internal metrics (currently not used).
- - trace_content (bool): Flag indicating whether to trace the content of the response.
+ - capture_message_content (bool): Flag indicating whether to trace the content of the response.
Returns:
- function: A higher-order function that takes a function 'wrapped' and returns
@@ -114,7 +114,7 @@ def general_wrap(gen_ai_endpoint, version, environment, application_name,
return wrapper
def hub(gen_ai_endpoint, version, environment, application_name, tracer,
- pricing_info, trace_content, metrics, disable_metrics):
+ pricing_info, capture_message_content, metrics, disable_metrics):
"""
Creates a wrapper around Langchain hub operations for tracing and logging.
@@ -129,7 +129,7 @@ def hub(gen_ai_endpoint, version, environment, application_name, tracer,
- application_name (str): Name of the Langchain application.
- tracer (opentelemetry.trace.Tracer): The tracer for OpenTelemetry tracing.
- pricing_info (dict): Pricing information for the operation (not currently used).
- - trace_content (bool): Indicates if the content of the response should be traced.
+ - capture_message_content (bool): Indicates if the content of the response should be traced.
Returns:
- function: A new function that wraps the original hub operation call with added
@@ -189,7 +189,7 @@ def hub(gen_ai_endpoint, version, environment, application_name, tracer,
return wrapper
def chat(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Creates a wrapper around a function call to trace and log its execution metrics.
@@ -202,7 +202,7 @@ def chat(gen_ai_endpoint, version, environment, application_name,
- application_name (str): Name of the Langchain application.
- tracer (opentelemetry.trace.Tracer): The tracer object used for OpenTelemetry tracing.
- pricing_info (dict): Information about the pricing for internal metrics (currently not used).
- - trace_content (bool): Flag indicating whether to trace the content of the response.
+ - capture_message_content (bool): Flag indicating whether to trace the content of the response.
Returns:
- function: A higher-order function that takes a function 'wrapped' and returns
@@ -287,6 +287,11 @@ def chat(gen_ai_endpoint, version, environment, application_name,
pricing_info, input_tokens, output_tokens
)
+ try:
+ llm_response = response.content
+ except AttributeError:
+ llm_response = response
+
# Set base span attribues (OTel Semconv)
span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
@@ -328,7 +333,7 @@ def chat(gen_ai_endpoint, version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -338,7 +343,7 @@ def chat(gen_ai_endpoint, version, environment, application_name,
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
attributes={
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: response,
+ SemanticConvetion.GEN_AI_CONTENT_COMPLETION: llm_response,
},
)
diff --git a/sdk/python/src/openlit/instrumentation/letta/__init__.py b/sdk/python/src/openlit/instrumentation/letta/__init__.py
index bb416a91..47af4145 100644
--- a/sdk/python/src/openlit/instrumentation/letta/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/letta/__init__.py
@@ -26,7 +26,7 @@ class LettaInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("letta")
@@ -34,42 +34,42 @@ class LettaInstrumentor(BaseInstrumentor):
"letta.client.client",
"LocalClient.create_agent",
create_agent("letta.create_agent", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"letta.client.client",
"LocalClient.get_agent",
create_agent("letta.get_agent", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"letta.client.client",
"LocalClient.send_message",
send_message("letta.send_message", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"letta.client.client",
"RESTClient.create_agent",
create_agent("letta.create_agent", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"letta.client.client",
"RESTClient.get_agent",
create_agent("letta.get_agent", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"letta.client.client",
"RESTClient.send_message",
send_message("letta.send_message", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/letta/letta.py b/sdk/python/src/openlit/instrumentation/letta/letta.py
index 97573126..d60830aa 100644
--- a/sdk/python/src/openlit/instrumentation/letta/letta.py
+++ b/sdk/python/src/openlit/instrumentation/letta/letta.py
@@ -15,7 +15,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def create_agent(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -26,7 +26,7 @@ def create_agent(gen_ai_endpoint, version, environment, application_name,
application_name: Name of the application using the Letta Agent.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Letta usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
@@ -94,7 +94,7 @@ def create_agent(gen_ai_endpoint, version, environment, application_name,
return wrapper
def send_message(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -105,7 +105,7 @@ def send_message(gen_ai_endpoint, version, environment, application_name,
application_name: Name of the application using the Letta Agent.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Letta usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
@@ -158,7 +158,7 @@ def send_message(gen_ai_endpoint, version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
cost)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/litellm/__init__.py b/sdk/python/src/openlit/instrumentation/litellm/__init__.py
index e85fa54f..47e8f085 100644
--- a/sdk/python/src/openlit/instrumentation/litellm/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/litellm/__init__.py
@@ -29,7 +29,7 @@ class LiteLLMInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("litellm")
@@ -38,28 +38,28 @@ class LiteLLMInstrumentor(BaseInstrumentor):
"litellm",
"completion",
completion(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"litellm",
"acompletion",
acompletion(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"litellm",
"embedding",
embedding(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"litellm",
"aembedding",
aembedding(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/litellm/async_litellm.py b/sdk/python/src/openlit/instrumentation/litellm/async_litellm.py
index 1db2f733..47328ee47 100644
--- a/sdk/python/src/openlit/instrumentation/litellm/async_litellm.py
+++ b/sdk/python/src/openlit/instrumentation/litellm/async_litellm.py
@@ -22,7 +22,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def acompletion(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -32,7 +32,7 @@ def acompletion(version, environment, application_name,
application_name: Name of the application using the LiteLLM SDK.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of LiteLLM usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
@@ -217,7 +217,7 @@ def acompletion(version, environment, application_name,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -393,7 +393,7 @@ def acompletion(version, environment, application_name,
end_time - start_time)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -404,7 +404,7 @@ def acompletion(version, environment, application_name,
for i in range(kwargs.get('n',1)):
span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
[response_dict.get('choices')[i].get('finish_reason')])
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
attributes={
@@ -464,7 +464,7 @@ def acompletion(version, environment, application_name,
return wrapper
def aembedding(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for embeddings to collect metrics.
@@ -474,7 +474,7 @@ def aembedding(version, environment, application_name,
application_name: Name of the application using the LiteLLM API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of LiteLLM usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the embeddings method to add telemetry.
@@ -548,7 +548,7 @@ def aembedding(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/litellm/litellm.py b/sdk/python/src/openlit/instrumentation/litellm/litellm.py
index ba24d1f8..120eeb41 100644
--- a/sdk/python/src/openlit/instrumentation/litellm/litellm.py
+++ b/sdk/python/src/openlit/instrumentation/litellm/litellm.py
@@ -22,7 +22,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def completion(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -32,7 +32,7 @@ def completion(version, environment, application_name,
application_name: Name of the application using the LiteLLM SDK.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of LiteLLM usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
@@ -217,7 +217,7 @@ def completion(version, environment, application_name,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -393,7 +393,7 @@ def completion(version, environment, application_name,
end_time - start_time)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -404,7 +404,7 @@ def completion(version, environment, application_name,
for i in range(kwargs.get('n',1)):
span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
[response_dict.get('choices')[i].get('finish_reason')])
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
attributes={
@@ -464,7 +464,7 @@ def completion(version, environment, application_name,
return wrapper
def embedding(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for embeddings to collect metrics.
@@ -474,7 +474,7 @@ def embedding(version, environment, application_name,
application_name: Name of the application using the LiteLLM API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of LiteLLM usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the embeddings method to add telemetry.
@@ -548,7 +548,7 @@ def embedding(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/llamaindex/__init__.py b/sdk/python/src/openlit/instrumentation/llamaindex/__init__.py
index 43c66875..d077aca8 100644
--- a/sdk/python/src/openlit/instrumentation/llamaindex/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/llamaindex/__init__.py
@@ -35,7 +35,7 @@ class LlamaIndexInstrumentor(BaseInstrumentor):
environment = kwargs.get("environment")
tracer = kwargs.get("tracer")
pricing_info = kwargs.get("pricing_info")
- trace_content = kwargs.get("trace_content")
+ capture_message_content = kwargs.get("capture_message_content")
version = importlib.metadata.version("llama-index")
for wrapped_method in WRAPPED_METHODS:
@@ -47,7 +47,7 @@ class LlamaIndexInstrumentor(BaseInstrumentor):
wrap_package,
wrap_object,
wrapper(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content),
+ tracer, pricing_info, capture_message_content),
)
@staticmethod
diff --git a/sdk/python/src/openlit/instrumentation/llamaindex/llamaindex.py b/sdk/python/src/openlit/instrumentation/llamaindex/llamaindex.py
index 381d56c2..3098a30a 100644
--- a/sdk/python/src/openlit/instrumentation/llamaindex/llamaindex.py
+++ b/sdk/python/src/openlit/instrumentation/llamaindex/llamaindex.py
@@ -13,7 +13,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def load_data(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content):
+ tracer, pricing_info, capture_message_content):
"""
Creates a wrapper around a function call to trace and log its execution metrics.
@@ -27,7 +27,7 @@ def load_data(gen_ai_endpoint, version, environment, application_name,
- application_name (str): Name of the LlamaIndex application.
- tracer (opentelemetry.trace.Tracer): The tracer object used for OpenTelemetry tracing.
- pricing_info (dict): Information about the pricing for internal metrics (currently not used).
- - trace_content (bool): Flag indicating whether to trace the content of the response.
+ - capture_message_content (bool): Flag indicating whether to trace the content of the response.
Returns:
- function: A higher-order function that takes a function 'wrapped' and returns
diff --git a/sdk/python/src/openlit/instrumentation/mem0/__init__.py b/sdk/python/src/openlit/instrumentation/mem0/__init__.py
index 1bf87afd..1cb7dfb2 100644
--- a/sdk/python/src/openlit/instrumentation/mem0/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/mem0/__init__.py
@@ -59,7 +59,7 @@ class Mem0Instrumentor(BaseInstrumentor):
environment = kwargs.get("environment")
tracer = kwargs.get("tracer")
pricing_info = kwargs.get("pricing_info")
- trace_content = kwargs.get("trace_content")
+ capture_message_content = kwargs.get("capture_message_content")
version = importlib.metadata.version("mem0ai")
for wrapped_method in WRAPPED_METHODS:
@@ -71,7 +71,7 @@ class Mem0Instrumentor(BaseInstrumentor):
wrap_package,
wrap_object,
wrapper(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content),
+ tracer, pricing_info, capture_message_content),
)
@staticmethod
diff --git a/sdk/python/src/openlit/instrumentation/mem0/mem0.py b/sdk/python/src/openlit/instrumentation/mem0/mem0.py
index 2c1719d3..0c2a26dd 100644
--- a/sdk/python/src/openlit/instrumentation/mem0/mem0.py
+++ b/sdk/python/src/openlit/instrumentation/mem0/mem0.py
@@ -13,7 +13,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def mem0_wrap(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content):
+ tracer, pricing_info, capture_message_content):
"""
Creates a wrapper around a function call to trace and log its execution metrics.
@@ -27,7 +27,7 @@ def mem0_wrap(gen_ai_endpoint, version, environment, application_name,
- application_name (str): Name of the mem0 application.
- tracer (opentelemetry.trace.Tracer): The tracer object used for OpenTelemetry tracing.
- pricing_info (dict): Information about the pricing for internal metrics (currently not used).
- - trace_content (bool): Flag indicating whether to trace the content of the response.
+ - capture_message_content (bool): Flag indicating whether to trace the content of the response.
Returns:
- function: A higher-order function that takes a function 'wrapped' and returns
diff --git a/sdk/python/src/openlit/instrumentation/milvus/__init__.py b/sdk/python/src/openlit/instrumentation/milvus/__init__.py
index c1e40472..66203730 100644
--- a/sdk/python/src/openlit/instrumentation/milvus/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/milvus/__init__.py
@@ -72,7 +72,7 @@ class MilvusInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info")
- trace_content = kwargs.get("trace_content")
+ capture_message_content = kwargs.get("capture_message_content")
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("pymilvus")
@@ -85,7 +85,7 @@ class MilvusInstrumentor(BaseInstrumentor):
wrap_package,
wrap_object,
wrapper(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
diff --git a/sdk/python/src/openlit/instrumentation/milvus/milvus.py b/sdk/python/src/openlit/instrumentation/milvus/milvus.py
index 1639491e..b0018595 100644
--- a/sdk/python/src/openlit/instrumentation/milvus/milvus.py
+++ b/sdk/python/src/openlit/instrumentation/milvus/milvus.py
@@ -25,7 +25,7 @@ def object_count(obj):
return cnt
def general_wrap(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Creates a wrapper around a function call to trace and log its execution metrics.
@@ -39,7 +39,7 @@ def general_wrap(gen_ai_endpoint, version, environment, application_name,
- application_name (str): Name of the Langchain application.
- tracer (opentelemetry.trace.Tracer): The tracer object used for OpenTelemetry tracing.
- pricing_info (dict): Information about the pricing for internal metrics (currently not used).
- - trace_content (bool): Flag indicating whether to trace the content of the response.
+ - capture_message_content (bool): Flag indicating whether to trace the content of the response.
Returns:
- function: A higher-order function that takes a function 'wrapped' and returns
diff --git a/sdk/python/src/openlit/instrumentation/mistral/__init__.py b/sdk/python/src/openlit/instrumentation/mistral/__init__.py
index 7bc0074d..e9b88b64 100644
--- a/sdk/python/src/openlit/instrumentation/mistral/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/mistral/__init__.py
@@ -23,7 +23,7 @@ class MistralInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info")
- trace_content = kwargs.get("trace_content")
+ capture_message_content = kwargs.get("capture_message_content")
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("mistralai")
@@ -32,7 +32,7 @@ class MistralInstrumentor(BaseInstrumentor):
"mistralai.chat",
"Chat.complete",
chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# sync
@@ -40,7 +40,7 @@ class MistralInstrumentor(BaseInstrumentor):
"mistralai.chat",
"Chat.stream",
chat_stream(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# sync
@@ -48,7 +48,7 @@ class MistralInstrumentor(BaseInstrumentor):
"mistralai.embeddings",
"Embeddings.create",
embeddings(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# Async
@@ -56,7 +56,7 @@ class MistralInstrumentor(BaseInstrumentor):
"mistralai.chat",
"Chat.complete_async",
async_chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# Async
@@ -64,7 +64,7 @@ class MistralInstrumentor(BaseInstrumentor):
"mistralai.chat",
"Chat.stream_async",
async_chat_stream(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
#sync
@@ -72,7 +72,7 @@ class MistralInstrumentor(BaseInstrumentor):
"mistralai.embeddings",
"Embeddings.create_async",
async_embeddings(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
@staticmethod
diff --git a/sdk/python/src/openlit/instrumentation/mistral/async_mistral.py b/sdk/python/src/openlit/instrumentation/mistral/async_mistral.py
index e92bc284..ff32ab52 100644
--- a/sdk/python/src/openlit/instrumentation/mistral/async_mistral.py
+++ b/sdk/python/src/openlit/instrumentation/mistral/async_mistral.py
@@ -22,7 +22,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def async_chat(version, environment, application_name, tracer,
- pricing_info, trace_content, metrics, disable_metrics):
+ pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat to collect metrics.
@@ -32,7 +32,7 @@ def async_chat(version, environment, application_name, tracer,
application_name: Name of the application using the Mistral API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Mistral usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat method to add telemetry.
@@ -144,7 +144,7 @@ def async_chat(version, environment, application_name, tracer,
end_time - start_time)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -155,7 +155,7 @@ def async_chat(version, environment, application_name, tracer,
for i in range(kwargs.get('n',1)):
span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
[response_dict.get('choices')[i].get('finish_reason')])
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
attributes={
@@ -215,7 +215,7 @@ def async_chat(version, environment, application_name, tracer,
return wrapper
def async_chat_stream(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat_stream to collect metrics.
@@ -225,7 +225,7 @@ def async_chat_stream(version, environment, application_name,
application_name: Name of the application using the Mistral API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Mistral usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat method to add telemetry.
@@ -422,7 +422,7 @@ def async_chat_stream(version, environment, application_name,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -485,7 +485,7 @@ def async_chat_stream(version, environment, application_name,
return wrapper
def async_embeddings(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for embeddings to collect metrics.
@@ -495,7 +495,7 @@ def async_embeddings(version, environment, application_name,
application_name: Name of the application using the Mistral API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Mistral usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the embeddings method to add telemetry.
@@ -567,7 +567,7 @@ def async_embeddings(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/mistral/mistral.py b/sdk/python/src/openlit/instrumentation/mistral/mistral.py
index 3dd363f6..5bbf9041 100644
--- a/sdk/python/src/openlit/instrumentation/mistral/mistral.py
+++ b/sdk/python/src/openlit/instrumentation/mistral/mistral.py
@@ -22,7 +22,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def chat(version, environment, application_name, tracer,
- pricing_info, trace_content, metrics, disable_metrics):
+ pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat to collect metrics.
@@ -32,7 +32,7 @@ def chat(version, environment, application_name, tracer,
application_name: Name of the application using the Mistral API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Mistral usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat method to add telemetry.
@@ -144,7 +144,7 @@ def chat(version, environment, application_name, tracer,
end_time - start_time)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -155,7 +155,7 @@ def chat(version, environment, application_name, tracer,
for i in range(kwargs.get('n',1)):
span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
[response_dict.get('choices')[i].get('finish_reason')])
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
attributes={
@@ -215,7 +215,7 @@ def chat(version, environment, application_name, tracer,
return wrapper
def chat_stream(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat_stream to collect metrics.
@@ -225,7 +225,7 @@ def chat_stream(version, environment, application_name,
application_name: Name of the application using the Mistral API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Mistral usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat method to add telemetry.
@@ -422,7 +422,7 @@ def chat_stream(version, environment, application_name,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -485,7 +485,7 @@ def chat_stream(version, environment, application_name,
return wrapper
def embeddings(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for embeddings to collect metrics.
@@ -495,7 +495,7 @@ def embeddings(version, environment, application_name,
application_name: Name of the application using the Mistral API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Mistral usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the embeddings method to add telemetry.
@@ -567,7 +567,7 @@ def embeddings(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/multion/__init__.py b/sdk/python/src/openlit/instrumentation/multion/__init__.py
index 65a4bb65..2487a5f0 100644
--- a/sdk/python/src/openlit/instrumentation/multion/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/multion/__init__.py
@@ -30,7 +30,7 @@ class MultiOnInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("multion")
@@ -39,19 +39,19 @@ class MultiOnInstrumentor(BaseInstrumentor):
"multion.client",
"MultiOn.browse",
multion_wrap("multion.browse", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"multion.client",
"MultiOn.retrieve",
multion_wrap("multion.retrieve", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"multion.sessions.client",
"SessionsClient.create",
multion_wrap("multion.sessions.create", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# Asynchronus
@@ -59,19 +59,19 @@ class MultiOnInstrumentor(BaseInstrumentor):
"multion.client",
"AsyncMultiOn.browse",
async_multion_wrap("multion.browse", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"multion.client",
"AsyncMultiOn.retrieve",
async_multion_wrap("multion.retrieve", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"multion.sessions.client",
"AsyncSessionsClient.create",
async_multion_wrap("multion.sessions.create", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
diff --git a/sdk/python/src/openlit/instrumentation/multion/async_multion.py b/sdk/python/src/openlit/instrumentation/multion/async_multion.py
index 2153f6b6..cff7bd9d 100644
--- a/sdk/python/src/openlit/instrumentation/multion/async_multion.py
+++ b/sdk/python/src/openlit/instrumentation/multion/async_multion.py
@@ -15,7 +15,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def async_multion_wrap(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -26,7 +26,7 @@ def async_multion_wrap(gen_ai_endpoint, version, environment, application_name,
application_name: Name of the application using the multion Agent.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of multion usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
@@ -75,7 +75,7 @@ def async_multion_wrap(gen_ai_endpoint, version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_AGENT_RESPONSE_TIME,
response.metadata.processing_time)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -92,7 +92,7 @@ def async_multion_wrap(gen_ai_endpoint, version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_AGENT_BROWSE_URL,
kwargs.get("url", ""))
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -110,7 +110,7 @@ def async_multion_wrap(gen_ai_endpoint, version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_AGENT_BROWSE_URL,
kwargs.get("url", ""))
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/multion/multion.py b/sdk/python/src/openlit/instrumentation/multion/multion.py
index f6ac6479..d6c34337 100644
--- a/sdk/python/src/openlit/instrumentation/multion/multion.py
+++ b/sdk/python/src/openlit/instrumentation/multion/multion.py
@@ -15,7 +15,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def multion_wrap(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -26,7 +26,7 @@ def multion_wrap(gen_ai_endpoint, version, environment, application_name,
application_name: Name of the application using the multion Agent.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of multion usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
@@ -75,7 +75,7 @@ def multion_wrap(gen_ai_endpoint, version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_AGENT_RESPONSE_TIME,
response.metadata.processing_time)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -92,7 +92,7 @@ def multion_wrap(gen_ai_endpoint, version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_AGENT_BROWSE_URL,
kwargs.get("url", ""))
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -110,7 +110,7 @@ def multion_wrap(gen_ai_endpoint, version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_AGENT_BROWSE_URL,
kwargs.get("url", ""))
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/ollama/__init__.py b/sdk/python/src/openlit/instrumentation/ollama/__init__.py
index d0f68237..85468c8f 100644
--- a/sdk/python/src/openlit/instrumentation/ollama/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/ollama/__init__.py
@@ -1,5 +1,6 @@
-# pylint: disable=useless-return, bad-staticmethod-argument, disable=duplicate-code
-"""Initializer of Auto Instrumentation of Ollama Functions"""
+"""
+Initializer of Auto Instrumentation of Ollama Functions
+"""
from typing import Collection
import importlib.metadata
@@ -27,9 +28,10 @@ class OllamaInstrumentor(BaseInstrumentor):
application_name = kwargs.get("application_name", "default_application")
environment = kwargs.get("environment", "default_environment")
tracer = kwargs.get("tracer")
+ event_provider = kwargs.get("event_provider")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("ollama")
@@ -38,13 +40,13 @@ class OllamaInstrumentor(BaseInstrumentor):
"ollama",
"chat",
chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"ollama",
"Client.chat",
chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics),
)
# sync embeddings
@@ -52,13 +54,13 @@ class OllamaInstrumentor(BaseInstrumentor):
"ollama",
"embeddings",
embeddings(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"ollama",
"Client.embeddings",
embeddings(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics),
)
# async chat
@@ -66,7 +68,7 @@ class OllamaInstrumentor(BaseInstrumentor):
"ollama",
"AsyncClient.chat",
async_chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics),
)
# async embeddings
@@ -74,7 +76,7 @@ class OllamaInstrumentor(BaseInstrumentor):
"ollama",
"AsyncClient.embeddings",
async_embeddings(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/ollama/async_ollama.py b/sdk/python/src/openlit/instrumentation/ollama/async_ollama.py
index 18c9c1d1..64367746 100644
--- a/sdk/python/src/openlit/instrumentation/ollama/async_ollama.py
+++ b/sdk/python/src/openlit/instrumentation/ollama/async_ollama.py
@@ -4,56 +4,37 @@ Module for monitoring Ollama API calls.
import logging
import time
-from opentelemetry.trace import SpanKind, Status, StatusCode
-from opentelemetry.sdk.resources import SERVICE_NAME, TELEMETRY_SDK_NAME, DEPLOYMENT_ENVIRONMENT
+from opentelemetry.trace import SpanKind
from openlit.__helpers import (
- get_chat_model_cost,
- get_embed_model_cost,
handle_exception,
- response_as_dict,
- general_tokens,
- calculate_ttft,
- calculate_tbt,
- create_metrics_attributes,
set_server_address_and_port
)
+from openlit.instrumentation.ollama.utils import (
+ process_chunk,
+ process_chat_response,
+ process_streaming_chat_response,
+ process_embedding_response
+)
from openlit.semcov import SemanticConvetion
-# Initialize logger for logging potential issues and operations
logger = logging.getLogger(__name__)
def async_chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics):
"""
- Generates a telemetry wrapper for chat completions to collect metrics.
-
- Args:
- version: Version of the monitoring package.
- environment: Deployment environment (e.g., production, staging).
- application_name: Name of the application using the Ollama API.
- tracer: OpenTelemetry tracer for creating spans.
- pricing_info: Information used for calculating the cost of Ollama usage.
- trace_content: Flag indicating whether to trace the actual content.
-
- Returns:
- A function that wraps the chat completions method to add telemetry.
+ Generates a telemetry wrapper for GenAI function call
"""
class TracedAsyncStream:
"""
- Wrapper for streaming responses to collect metrics and trace data.
- Wraps the response to collect message IDs and aggregated response.
-
- This class implements the '__aiter__' and '__anext__' methods that
- handle asynchronous streaming responses.
-
- This class also implements '__aenter__' and '__aexit__' methods that
- handle asynchronous context management protocol.
+ Wrapper for streaming responses to collect telemetry.
"""
+
def __init__(
self,
wrapped,
span,
+ span_name,
kwargs,
server_address,
server_port,
@@ -61,13 +42,14 @@ def async_chat(version, environment, application_name,
):
self.__wrapped__ = wrapped
self._span = span
- # Placeholder for aggregating streaming response
self._llmresponse = ""
self._response_model = ""
self._finish_reason = ""
+ self._tool_calls = []
self._input_tokens = 0
self._output_tokens = 0
-
+ self._response_role = ''
+ self._span_name = span_name
self._args = args
self._kwargs = kwargs
self._start_time = time.time()
@@ -95,194 +77,34 @@ def async_chat(version, environment, application_name,
async def __anext__(self):
try:
chunk = await self.__wrapped__.__anext__()
- end_time = time.time()
- # Record the timestamp for the current chunk
- self._timestamps.append(end_time)
-
- if len(self._timestamps) == 1:
- # Calculate time to first chunk
- self._ttft = calculate_ttft(self._timestamps, self._start_time)
-
- chunked = response_as_dict(chunk)
- self._llmresponse += chunked.get('message').get('content')
-
- if chunked.get('eval_count'):
- self._input_tokens = chunked.get('prompt_eval_count')
- self._output_tokens = chunked.get('eval_count')
- self._response_model = chunked.get('model')
- self._finish_reason = chunked.get('done_reason')
+ process_chunk(self, chunk)
return chunk
except StopAsyncIteration:
- # Handling exception ensure observability without disrupting operation
try:
- self._end_time = time.time()
- if len(self._timestamps) > 1:
- self._tbt = calculate_tbt(self._timestamps)
-
- # Format 'messages' into a single string
- message_prompt = self._kwargs.get("messages", "")
- formatted_messages = []
- for message in message_prompt:
- role = message["role"]
- content = message["content"]
-
- if isinstance(content, list):
- content_str_list = []
- for item in content:
- if item["type"] == "text":
- content_str_list.append(f'text: {item["text"]}')
- elif (item["type"] == "image_url" and
- not item["image_url"]["url"].startswith("data:")):
- content_str_list.append(f'image_url: {item["image_url"]["url"]}')
- content_str = ", ".join(content_str_list)
- formatted_messages.append(f"{role}: {content_str}")
- else:
- formatted_messages.append(f"{role}: {content}")
- prompt = "\n".join(formatted_messages)
-
- request_model = self._kwargs.get("model", "gpt-4o")
-
- # Calculate cost of the operation
- cost = get_chat_model_cost(request_model,
- pricing_info, self._input_tokens,
- self._output_tokens)
-
- # Set Span attributes (OTel Semconv)
- self._span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
- self._span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT)
- self._span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_OLLAMA)
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- request_model)
- self._span.set_attribute(SemanticConvetion.SERVER_PORT,
- self._server_port)
-
- # List of attributes and their config keys
- attributes = [
- (SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY, 'repeat_penalty'),
- (SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS, 'max_tokens'),
- (SemanticConvetion.GEN_AI_REQUEST_SEED, 'seed'),
- (SemanticConvetion.GEN_AI_REQUEST_STOP_SEQUENCES, 'stop'),
- (SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE, 'temperature'),
- (SemanticConvetion.GEN_AI_REQUEST_TOP_P, 'top_p'),
- (SemanticConvetion.GEN_AI_REQUEST_TOP_K, 'top_k'),
- ]
-
- # Safely get the options dictionary from kwargs
- options = self._kwargs.get('options', {})
-
- # Set each attribute if the corresponding value exists and is not None
- for attribute, key in attributes:
- # Use dictionary `get` to retrieve values from the options dictionary
- value = options.get(key)
- if value is not None:
- self._span.set_attribute(attribute, value)
-
- self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
- [self._finish_reason])
- self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_MODEL,
- self._response_model)
- self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- self._input_tokens)
- self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
- self._output_tokens)
- self._span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
- self._server_address)
- if isinstance(self._llmresponse, str):
- self._span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- "text")
- else:
- self._span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- "json")
-
- # Set Span attributes (Extra)
- self._span.set_attribute(DEPLOYMENT_ENVIRONMENT,
- environment)
- self._span.set_attribute(SERVICE_NAME,
- application_name)
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
- True)
- self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- self._input_tokens + self._output_tokens)
- self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
- self._span.set_attribute(SemanticConvetion.GEN_AI_SERVER_TBT,
- self._tbt)
- self._span.set_attribute(SemanticConvetion.GEN_AI_SERVER_TTFT,
- self._ttft)
- self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
- version)
- if trace_content:
- self._span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
- },
- )
- self._span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: self._llmresponse,
- },
- )
- self._span.set_status(Status(StatusCode.OK))
-
- if disable_metrics is False:
- attributes = create_metrics_attributes(
- service_name=application_name,
- deployment_environment=environment,
- operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
- system=SemanticConvetion.GEN_AI_SYSTEM_OLLAMA,
- request_model=request_model,
- server_address=self._server_address,
- server_port=self._server_port,
- response_model=self._response_model,
- )
-
- metrics["genai_client_usage_tokens"].record(
- self._input_tokens + self._output_tokens, attributes
- )
- metrics["genai_client_operation_duration"].record(
- self._end_time - self._start_time, attributes
- )
- metrics["genai_server_tbt"].record(
- self._tbt, attributes
+ with tracer.start_as_current_span(self._span_name, kind= SpanKind.CLIENT) as self._span:
+ process_streaming_chat_response(
+ self,
+ pricing_info=pricing_info,
+ environment=environment,
+ application_name=application_name,
+ metrics=metrics,
+ event_provider=event_provider,
+ capture_message_content=capture_message_content,
+ disable_metrics=disable_metrics,
+ version=version
)
- metrics["genai_server_ttft"].record(
- self._ttft, attributes
- )
- metrics["genai_requests"].add(1, attributes)
- metrics["genai_completion_tokens"].add(self._output_tokens, attributes)
- metrics["genai_prompt_tokens"].add(self._input_tokens, attributes)
- metrics["genai_cost"].record(cost, attributes)
-
except Exception as e:
handle_exception(self._span, e)
logger.error("Error in trace creation: %s", e)
- finally:
- self._span.end()
raise
async def wrapper(wrapped, instance, args, kwargs):
"""
- Wraps the 'chat.completions' API call to add telemetry.
-
- This collects metrics such as execution time, cost, and token usage, and handles errors
- gracefully, adding details to the trace for observability.
-
- Args:
- wrapped: The original 'chat.completions' method to be wrapped.
- instance: The instance of the class where the original method is defined.
- args: Positional arguments for the 'chat.completions' method.
- kwargs: Keyword arguments for the 'chat.completions' method.
-
- Returns:
- The response from the original 'chat.completions' method.
+ Wraps the GenAI function call.
"""
- # Check if streaming is enabled for the API call
streaming = kwargs.get("stream", False)
+
server_address, server_port = set_server_address_and_port(instance, "127.0.0.1", 11434)
request_model = kwargs.get("model", "gpt-4o")
@@ -290,202 +112,45 @@ def async_chat(version, environment, application_name,
# pylint: disable=no-else-return
if streaming:
- # Special handling for streaming response to accommodate the nature of data flow
awaited_wrapped = await wrapped(*args, **kwargs)
span = tracer.start_span(span_name, kind=SpanKind.CLIENT)
+ return TracedAsyncStream(awaited_wrapped, span, span_name, kwargs, server_address, server_port)
- return TracedAsyncStream(awaited_wrapped, span, kwargs, server_address, server_port)
-
- # Handling for non-streaming responses
else:
with tracer.start_as_current_span(span_name, kind= SpanKind.CLIENT) as span:
start_time = time.time()
response = await wrapped(*args, **kwargs)
- end_time = time.time()
-
- response_dict = response_as_dict(response)
-
- try:
- # Format 'messages' into a single string
- message_prompt = kwargs.get("messages", "")
- formatted_messages = []
- for message in message_prompt:
- role = message["role"]
- content = message["content"]
-
- if isinstance(content, list):
- content_str = ", ".join(
- f'{item["type"]}: {item["text"] if "text" in item else item["image_url"]}'
- if "type" in item else f'text: {item["text"]}'
- for item in content
- )
- formatted_messages.append(f"{role}: {content_str}")
- else:
- formatted_messages.append(f"{role}: {content}")
- prompt = "\n".join(formatted_messages)
-
- input_tokens = response_dict.get('prompt_eval_count')
- output_tokens = response_dict.get('eval_count')
-
- # Calculate cost of the operation
- cost = get_chat_model_cost(request_model,
- pricing_info, input_tokens,
- output_tokens)
-
- # Set base span attribues (OTel Semconv)
- span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
- span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT)
- span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_OLLAMA)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- request_model)
- span.set_attribute(SemanticConvetion.SERVER_PORT,
- server_port)
-
- # List of attributes and their config keys
- attributes = [
- (SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY, 'repeat_penalty'),
- (SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS, 'max_tokens'),
- (SemanticConvetion.GEN_AI_REQUEST_SEED, 'seed'),
- (SemanticConvetion.GEN_AI_REQUEST_STOP_SEQUENCES, 'stop'),
- (SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE, 'temperature'),
- (SemanticConvetion.GEN_AI_REQUEST_TOP_P, 'top_p'),
- (SemanticConvetion.GEN_AI_REQUEST_TOP_K, 'top_k'),
- ]
-
- # Safely get the options dictionary from kwargs
- options = kwargs.get('options', {})
-
- # Set each attribute if the corresponding value exists and is not None
- for attribute, key in attributes:
- # Use dictionary `get` to retrieve values from the options dictionary
- value = options.get(key)
- if value is not None:
- span.set_attribute(attribute, value)
-
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_MODEL,
- response_dict.get('model'))
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- input_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
- output_tokens)
- span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
- server_address)
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
- [response_dict.get('done_reason')])
- if kwargs.get('format'):
- span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- 'json')
- else:
- span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- 'text')
-
- # Set base span attribues (Extras)
- span.set_attribute(DEPLOYMENT_ENVIRONMENT,
- environment)
- span.set_attribute(SERVICE_NAME,
- application_name)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
- False)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- input_tokens + output_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
- span.set_attribute(SemanticConvetion.GEN_AI_SERVER_TTFT,
- end_time - start_time)
- span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
- version)
- if trace_content:
- span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
- },
- )
- span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
- attributes={
- # pylint: disable=line-too-long
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: str(response_dict.get('message').get('content')),
- },
- )
- if kwargs.get('tools'):
- span.set_attribute(SemanticConvetion.GEN_AI_TOOL_CALLS,
- str(response_dict.get('message').get('tool_calls')))
-
- span.set_status(Status(StatusCode.OK))
-
- if disable_metrics is False:
- attributes = create_metrics_attributes(
- service_name=application_name,
- deployment_environment=environment,
- operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
- system=SemanticConvetion.GEN_AI_SYSTEM_OLLAMA,
- request_model=request_model,
- server_address=server_address,
- server_port=server_port,
- response_model=response_dict.get('model'),
- )
-
- metrics["genai_client_usage_tokens"].record(
- input_tokens + output_tokens, attributes
- )
- metrics["genai_client_operation_duration"].record(
- end_time - start_time, attributes
- )
- metrics["genai_server_ttft"].record(
- end_time - start_time, attributes
- )
- metrics["genai_requests"].add(1, attributes)
- metrics["genai_completion_tokens"].add(output_tokens, attributes)
- metrics["genai_prompt_tokens"].add(input_tokens, attributes)
- metrics["genai_cost"].record(cost, attributes)
-
- # Return original response
- return response
-
- except Exception as e:
- handle_exception(span, e)
- logger.error("Error in trace creation: %s", e)
-
- # Return original response
- return response
+ response = process_chat_response(
+ response=response,
+ request_model=request_model,
+ pricing_info=pricing_info,
+ server_port=server_port,
+ server_address=server_address,
+ environment=environment,
+ application_name=application_name,
+ metrics=metrics,
+ event_provider=event_provider,
+ start_time=start_time,
+ span=span,
+ capture_message_content=capture_message_content,
+ disable_metrics=disable_metrics,
+ version=version,
+ **kwargs
+ )
+
+ return response
return wrapper
def async_embeddings(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics):
"""
- Generates a telemetry wrapper for embeddings to collect metrics.
-
- Args:
- version: Version of the monitoring package.
- environment: Deployment environment (e.g., production, staging).
- application_name: Name of the application using the Ollama API.
- tracer: OpenTelemetry tracer for creating spans.
- pricing_info: Information used for calculating the cost of Ollama usage.
- trace_content: Flag indicating whether to trace the actual content.
-
- Returns:
- A function that wraps the embeddings method to add telemetry.
+ Generates a telemetry wrapper for GenAI function call
"""
async def wrapper(wrapped, instance, args, kwargs):
"""
- Wraps the 'embeddings' API call to add telemetry.
-
- This collects metrics such as execution time, cost, and token usage, and handles errors
- gracefully, adding details to the trace for observability.
-
- Args:
- wrapped: The original 'embeddings' method to be wrapped.
- instance: The instance of the class where the original method is defined.
- args: Positional arguments for the 'embeddings' method.
- kwargs: Keyword arguments for the 'embeddings' method.
-
- Returns:
- The response from the original 'embeddings' method.
+ Wraps the GenAI function call.
"""
server_address, server_port = set_server_address_and_port(instance, '127.0.0.1', 11434)
@@ -496,83 +161,24 @@ def async_embeddings(version, environment, application_name,
with tracer.start_as_current_span(span_name, kind= SpanKind.CLIENT) as span:
start_time = time.time()
response = await wrapped(*args, **kwargs)
- end_time = time.time()
-
- try:
- input_tokens = general_tokens(str(kwargs.get('prompt')))
-
- # Calculate cost of the operation
- cost = get_embed_model_cost(request_model,
- pricing_info, input_tokens)
-
- # Set Span attributes (OTel Semconv)
- span.set_attribute(TELEMETRY_SDK_NAME, 'openlit')
- span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
- SemanticConvetion.GEN_AI_OPERATION_TYPE_EMBEDDING)
- span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_OLLAMA)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- request_model)
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_MODEL,
- request_model)
- span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
- server_address)
- span.set_attribute(SemanticConvetion.SERVER_PORT,
- server_port)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- input_tokens)
-
- # Set Span attributes (Extras)
- span.set_attribute(DEPLOYMENT_ENVIRONMENT,
- environment)
- span.set_attribute(SERVICE_NAME,
- application_name)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- input_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
- span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
- version)
-
- if trace_content:
- span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_PROMPT: str(kwargs.get('prompt', '')),
- },
- )
-
- span.set_status(Status(StatusCode.OK))
-
- if disable_metrics is False:
- attributes = create_metrics_attributes(
- service_name=application_name,
- deployment_environment=environment,
- operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_EMBEDDING,
- system=SemanticConvetion.GEN_AI_SYSTEM_OLLAMA,
- request_model=request_model,
- server_address=server_address,
- server_port=server_port,
- response_model=request_model,
- )
- metrics['genai_client_usage_tokens'].record(
- input_tokens, attributes
- )
- metrics['genai_client_operation_duration'].record(
- end_time - start_time, attributes
- )
- metrics['genai_requests'].add(1, attributes)
- metrics['genai_prompt_tokens'].add(input_tokens, attributes)
- metrics['genai_cost'].record(cost, attributes)
-
- # Return original response
- return response
-
- except Exception as e:
- handle_exception(span, e)
- logger.error('Error in trace creation: %s', e)
-
- # Return original response
- return response
+ response = process_embedding_response(
+ response=response,
+ request_model=request_model,
+ pricing_info=pricing_info,
+ server_port=server_port,
+ server_address=server_address,
+ environment=environment,
+ application_name=application_name,
+ metrics=metrics,
+ event_provider=event_provider,
+ start_time=start_time,
+ span=span,
+ capture_message_content=capture_message_content,
+ disable_metrics=disable_metrics,
+ version=version,
+ **kwargs
+ )
+
+ return response
return wrapper
diff --git a/sdk/python/src/openlit/instrumentation/ollama/ollama.py b/sdk/python/src/openlit/instrumentation/ollama/ollama.py
index 05f3b74a..4bfcbe4b 100644
--- a/sdk/python/src/openlit/instrumentation/ollama/ollama.py
+++ b/sdk/python/src/openlit/instrumentation/ollama/ollama.py
@@ -4,56 +4,37 @@ Module for monitoring Ollama API calls.
import logging
import time
-from opentelemetry.trace import SpanKind, Status, StatusCode
-from opentelemetry.sdk.resources import SERVICE_NAME, TELEMETRY_SDK_NAME, DEPLOYMENT_ENVIRONMENT
+from opentelemetry.trace import SpanKind
from openlit.__helpers import (
- get_chat_model_cost,
- get_embed_model_cost,
handle_exception,
- response_as_dict,
- general_tokens,
- calculate_ttft,
- calculate_tbt,
- create_metrics_attributes,
set_server_address_and_port
)
+from openlit.instrumentation.ollama.utils import (
+ process_chunk,
+ process_chat_response,
+ process_streaming_chat_response,
+ process_embedding_response
+)
from openlit.semcov import SemanticConvetion
-# Initialize logger for logging potential issues and operations
logger = logging.getLogger(__name__)
def chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics):
"""
- Generates a telemetry wrapper for chat completions to collect metrics.
-
- Args:
- version: Version of the monitoring package.
- environment: Deployment environment (e.g., production, staging).
- application_name: Name of the application using the Ollama API.
- tracer: OpenTelemetry tracer for creating spans.
- pricing_info: Information used for calculating the cost of Ollama usage.
- trace_content: Flag indicating whether to trace the actual content.
-
- Returns:
- A function that wraps the chat completions method to add telemetry.
+ Generates a telemetry wrapper for GenAI function call
"""
class TracedSyncStream:
"""
- Wrapper for streaming responses to collect metrics and trace data.
- Wraps the response to collect message IDs and aggregated response.
-
- This class implements the '__aiter__' and '__anext__' methods that
- handle asynchronous streaming responses.
-
- This class also implements '__aenter__' and '__aexit__' methods that
- handle asynchronous context management protocol.
+ Wrapper for streaming responses to collect telemetry.
"""
+
def __init__(
self,
wrapped,
span,
+ span_name,
kwargs,
server_address,
server_port,
@@ -61,13 +42,14 @@ def chat(version, environment, application_name,
):
self.__wrapped__ = wrapped
self._span = span
- # Placeholder for aggregating streaming response
self._llmresponse = ""
self._response_model = ""
self._finish_reason = ""
+ self._tool_calls = []
self._input_tokens = 0
self._output_tokens = 0
-
+ self._response_role = ''
+ self._span_name = span_name
self._args = args
self._kwargs = kwargs
self._start_time = time.time()
@@ -95,194 +77,34 @@ def chat(version, environment, application_name,
def __next__(self):
try:
chunk = self.__wrapped__.__next__()
- end_time = time.time()
- # Record the timestamp for the current chunk
- self._timestamps.append(end_time)
-
- if len(self._timestamps) == 1:
- # Calculate time to first chunk
- self._ttft = calculate_ttft(self._timestamps, self._start_time)
-
- chunked = response_as_dict(chunk)
- self._llmresponse += chunked.get('message').get('content')
-
- if chunked.get('eval_count'):
- self._input_tokens = chunked.get('prompt_eval_count')
- self._output_tokens = chunked.get('eval_count')
- self._response_model = chunked.get('model')
- self._finish_reason = chunked.get('done_reason')
+ process_chunk(self, chunk)
return chunk
except StopIteration:
- # Handling exception ensure observability without disrupting operation
try:
- self._end_time = time.time()
- if len(self._timestamps) > 1:
- self._tbt = calculate_tbt(self._timestamps)
-
- # Format 'messages' into a single string
- message_prompt = self._kwargs.get("messages", "")
- formatted_messages = []
- for message in message_prompt:
- role = message["role"]
- content = message["content"]
-
- if isinstance(content, list):
- content_str_list = []
- for item in content:
- if item["type"] == "text":
- content_str_list.append(f'text: {item["text"]}')
- elif (item["type"] == "image_url" and
- not item["image_url"]["url"].startswith("data:")):
- content_str_list.append(f'image_url: {item["image_url"]["url"]}')
- content_str = ", ".join(content_str_list)
- formatted_messages.append(f"{role}: {content_str}")
- else:
- formatted_messages.append(f"{role}: {content}")
- prompt = "\n".join(formatted_messages)
-
- request_model = self._kwargs.get("model", "gpt-4o")
-
- # Calculate cost of the operation
- cost = get_chat_model_cost(request_model,
- pricing_info, self._input_tokens,
- self._output_tokens)
-
- # Set Span attributes (OTel Semconv)
- self._span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
- self._span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT)
- self._span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_OLLAMA)
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- request_model)
- self._span.set_attribute(SemanticConvetion.SERVER_PORT,
- self._server_port)
-
- # List of attributes and their config keys
- attributes = [
- (SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY, 'repeat_penalty'),
- (SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS, 'max_tokens'),
- (SemanticConvetion.GEN_AI_REQUEST_SEED, 'seed'),
- (SemanticConvetion.GEN_AI_REQUEST_STOP_SEQUENCES, 'stop'),
- (SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE, 'temperature'),
- (SemanticConvetion.GEN_AI_REQUEST_TOP_P, 'top_p'),
- (SemanticConvetion.GEN_AI_REQUEST_TOP_K, 'top_k'),
- ]
-
- # Safely get the options dictionary from kwargs
- options = self._kwargs.get('options', {})
-
- # Set each attribute if the corresponding value exists and is not None
- for attribute, key in attributes:
- # Use dictionary `get` to retrieve values from the options dictionary
- value = options.get(key)
- if value is not None:
- self._span.set_attribute(attribute, value)
-
- self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
- [self._finish_reason])
- self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_MODEL,
- self._response_model)
- self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- self._input_tokens)
- self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
- self._output_tokens)
- self._span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
- self._server_address)
- if isinstance(self._llmresponse, str):
- self._span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- "text")
- else:
- self._span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- "json")
-
- # Set Span attributes (Extra)
- self._span.set_attribute(DEPLOYMENT_ENVIRONMENT,
- environment)
- self._span.set_attribute(SERVICE_NAME,
- application_name)
- self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
- True)
- self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- self._input_tokens + self._output_tokens)
- self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
- self._span.set_attribute(SemanticConvetion.GEN_AI_SERVER_TBT,
- self._tbt)
- self._span.set_attribute(SemanticConvetion.GEN_AI_SERVER_TTFT,
- self._ttft)
- self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
- version)
- if trace_content:
- self._span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
- },
- )
- self._span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: self._llmresponse,
- },
- )
- self._span.set_status(Status(StatusCode.OK))
-
- if disable_metrics is False:
- attributes = create_metrics_attributes(
- service_name=application_name,
- deployment_environment=environment,
- operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
- system=SemanticConvetion.GEN_AI_SYSTEM_OLLAMA,
- request_model=request_model,
- server_address=self._server_address,
- server_port=self._server_port,
- response_model=self._response_model,
- )
-
- metrics["genai_client_usage_tokens"].record(
- self._input_tokens + self._output_tokens, attributes
- )
- metrics["genai_client_operation_duration"].record(
- self._end_time - self._start_time, attributes
- )
- metrics["genai_server_tbt"].record(
- self._tbt, attributes
+ with tracer.start_as_current_span(self._span_name, kind= SpanKind.CLIENT) as self._span:
+ process_streaming_chat_response(
+ self,
+ pricing_info=pricing_info,
+ environment=environment,
+ application_name=application_name,
+ metrics=metrics,
+ event_provider=event_provider,
+ capture_message_content=capture_message_content,
+ disable_metrics=disable_metrics,
+ version=version
)
- metrics["genai_server_ttft"].record(
- self._ttft, attributes
- )
- metrics["genai_requests"].add(1, attributes)
- metrics["genai_completion_tokens"].add(self._output_tokens, attributes)
- metrics["genai_prompt_tokens"].add(self._input_tokens, attributes)
- metrics["genai_cost"].record(cost, attributes)
-
except Exception as e:
handle_exception(self._span, e)
logger.error("Error in trace creation: %s", e)
- finally:
- self._span.end()
raise
def wrapper(wrapped, instance, args, kwargs):
"""
- Wraps the 'chat.completions' API call to add telemetry.
-
- This collects metrics such as execution time, cost, and token usage, and handles errors
- gracefully, adding details to the trace for observability.
-
- Args:
- wrapped: The original 'chat.completions' method to be wrapped.
- instance: The instance of the class where the original method is defined.
- args: Positional arguments for the 'chat.completions' method.
- kwargs: Keyword arguments for the 'chat.completions' method.
-
- Returns:
- The response from the original 'chat.completions' method.
+ Wraps the GenAI function call.
"""
- # Check if streaming is enabled for the API call
streaming = kwargs.get("stream", False)
+
server_address, server_port = set_server_address_and_port(instance, "127.0.0.1", 11434)
request_model = kwargs.get("model", "gpt-4o")
@@ -290,202 +112,45 @@ def chat(version, environment, application_name,
# pylint: disable=no-else-return
if streaming:
- # Special handling for streaming response to accommodate the nature of data flow
awaited_wrapped = wrapped(*args, **kwargs)
span = tracer.start_span(span_name, kind=SpanKind.CLIENT)
+ return TracedSyncStream(awaited_wrapped, span, span_name, kwargs, server_address, server_port)
- return TracedSyncStream(awaited_wrapped, span, kwargs, server_address, server_port)
-
- # Handling for non-streaming responses
else:
with tracer.start_as_current_span(span_name, kind= SpanKind.CLIENT) as span:
start_time = time.time()
response = wrapped(*args, **kwargs)
- end_time = time.time()
-
- response_dict = response_as_dict(response)
-
- try:
- # Format 'messages' into a single string
- message_prompt = kwargs.get("messages", "")
- formatted_messages = []
- for message in message_prompt:
- role = message["role"]
- content = message["content"]
-
- if isinstance(content, list):
- content_str = ", ".join(
- f'{item["type"]}: {item["text"] if "text" in item else item["image_url"]}'
- if "type" in item else f'text: {item["text"]}'
- for item in content
- )
- formatted_messages.append(f"{role}: {content_str}")
- else:
- formatted_messages.append(f"{role}: {content}")
- prompt = "\n".join(formatted_messages)
-
- input_tokens = response_dict.get('prompt_eval_count')
- output_tokens = response_dict.get('eval_count')
-
- # Calculate cost of the operation
- cost = get_chat_model_cost(request_model,
- pricing_info, input_tokens,
- output_tokens)
-
- # Set base span attribues (OTel Semconv)
- span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
- span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT)
- span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_OLLAMA)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- request_model)
- span.set_attribute(SemanticConvetion.SERVER_PORT,
- server_port)
-
- # List of attributes and their config keys
- attributes = [
- (SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY, 'repeat_penalty'),
- (SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS, 'max_tokens'),
- (SemanticConvetion.GEN_AI_REQUEST_SEED, 'seed'),
- (SemanticConvetion.GEN_AI_REQUEST_STOP_SEQUENCES, 'stop'),
- (SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE, 'temperature'),
- (SemanticConvetion.GEN_AI_REQUEST_TOP_P, 'top_p'),
- (SemanticConvetion.GEN_AI_REQUEST_TOP_K, 'top_k'),
- ]
-
- # Safely get the options dictionary from kwargs
- options = kwargs.get('options', {})
-
- # Set each attribute if the corresponding value exists and is not None
- for attribute, key in attributes:
- # Use dictionary `get` to retrieve values from the options dictionary
- value = options.get(key)
- if value is not None:
- span.set_attribute(attribute, value)
-
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_MODEL,
- response_dict.get('model'))
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- input_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
- output_tokens)
- span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
- server_address)
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
- [response_dict.get('done_reason')])
- if kwargs.get('format'):
- span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- 'json')
- else:
- span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
- 'text')
-
- # Set base span attribues (Extras)
- span.set_attribute(DEPLOYMENT_ENVIRONMENT,
- environment)
- span.set_attribute(SERVICE_NAME,
- application_name)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
- False)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- input_tokens + output_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
- span.set_attribute(SemanticConvetion.GEN_AI_SERVER_TTFT,
- end_time - start_time)
- span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
- version)
- if trace_content:
- span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
- },
- )
- span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
- attributes={
- # pylint: disable=line-too-long
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: str(response_dict.get('message').get('content')),
- },
- )
- if kwargs.get('tools'):
- span.set_attribute(SemanticConvetion.GEN_AI_TOOL_CALLS,
- str(response_dict.get('message').get('tool_calls')))
-
- span.set_status(Status(StatusCode.OK))
-
- if disable_metrics is False:
- attributes = create_metrics_attributes(
- service_name=application_name,
- deployment_environment=environment,
- operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
- system=SemanticConvetion.GEN_AI_SYSTEM_OLLAMA,
- request_model=request_model,
- server_address=server_address,
- server_port=server_port,
- response_model=response_dict.get('model'),
- )
-
- metrics["genai_client_usage_tokens"].record(
- input_tokens + output_tokens, attributes
- )
- metrics["genai_client_operation_duration"].record(
- end_time - start_time, attributes
- )
- metrics["genai_server_ttft"].record(
- end_time - start_time, attributes
- )
- metrics["genai_requests"].add(1, attributes)
- metrics["genai_completion_tokens"].add(output_tokens, attributes)
- metrics["genai_prompt_tokens"].add(input_tokens, attributes)
- metrics["genai_cost"].record(cost, attributes)
-
- # Return original response
- return response
-
- except Exception as e:
- handle_exception(span, e)
- logger.error("Error in trace creation: %s", e)
-
- # Return original response
- return response
+ response = process_chat_response(
+ response=response,
+ request_model=request_model,
+ pricing_info=pricing_info,
+ server_port=server_port,
+ server_address=server_address,
+ environment=environment,
+ application_name=application_name,
+ metrics=metrics,
+ event_provider=event_provider,
+ start_time=start_time,
+ span=span,
+ capture_message_content=capture_message_content,
+ disable_metrics=disable_metrics,
+ version=version,
+ **kwargs
+ )
+
+ return response
return wrapper
def embeddings(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics):
"""
- Generates a telemetry wrapper for embeddings to collect metrics.
-
- Args:
- version: Version of the monitoring package.
- environment: Deployment environment (e.g., production, staging).
- application_name: Name of the application using the Ollama API.
- tracer: OpenTelemetry tracer for creating spans.
- pricing_info: Information used for calculating the cost of Ollama usage.
- trace_content: Flag indicating whether to trace the actual content.
-
- Returns:
- A function that wraps the embeddings method to add telemetry.
+ Generates a telemetry wrapper for GenAI function call
"""
def wrapper(wrapped, instance, args, kwargs):
"""
- Wraps the 'embeddings' API call to add telemetry.
-
- This collects metrics such as execution time, cost, and token usage, and handles errors
- gracefully, adding details to the trace for observability.
-
- Args:
- wrapped: The original 'embeddings' method to be wrapped.
- instance: The instance of the class where the original method is defined.
- args: Positional arguments for the 'embeddings' method.
- kwargs: Keyword arguments for the 'embeddings' method.
-
- Returns:
- The response from the original 'embeddings' method.
+ Wraps the GenAI function call.
"""
server_address, server_port = set_server_address_and_port(instance, '127.0.0.1', 11434)
@@ -496,83 +161,24 @@ def embeddings(version, environment, application_name,
with tracer.start_as_current_span(span_name, kind= SpanKind.CLIENT) as span:
start_time = time.time()
response = wrapped(*args, **kwargs)
- end_time = time.time()
-
- try:
- input_tokens = general_tokens(str(kwargs.get('prompt')))
-
- # Calculate cost of the operation
- cost = get_embed_model_cost(request_model,
- pricing_info, input_tokens)
-
- # Set Span attributes (OTel Semconv)
- span.set_attribute(TELEMETRY_SDK_NAME, 'openlit')
- span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
- SemanticConvetion.GEN_AI_OPERATION_TYPE_EMBEDDING)
- span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_OLLAMA)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- request_model)
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_MODEL,
- request_model)
- span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
- server_address)
- span.set_attribute(SemanticConvetion.SERVER_PORT,
- server_port)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- input_tokens)
-
- # Set Span attributes (Extras)
- span.set_attribute(DEPLOYMENT_ENVIRONMENT,
- environment)
- span.set_attribute(SERVICE_NAME,
- application_name)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- input_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
- span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
- version)
-
- if trace_content:
- span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_PROMPT: str(kwargs.get('prompt', '')),
- },
- )
-
- span.set_status(Status(StatusCode.OK))
-
- if disable_metrics is False:
- attributes = create_metrics_attributes(
- service_name=application_name,
- deployment_environment=environment,
- operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_EMBEDDING,
- system=SemanticConvetion.GEN_AI_SYSTEM_OLLAMA,
- request_model=request_model,
- server_address=server_address,
- server_port=server_port,
- response_model=request_model,
- )
- metrics['genai_client_usage_tokens'].record(
- input_tokens, attributes
- )
- metrics['genai_client_operation_duration'].record(
- end_time - start_time, attributes
- )
- metrics['genai_requests'].add(1, attributes)
- metrics['genai_prompt_tokens'].add(input_tokens, attributes)
- metrics['genai_cost'].record(cost, attributes)
-
- # Return original response
- return response
-
- except Exception as e:
- handle_exception(span, e)
- logger.error('Error in trace creation: %s', e)
-
- # Return original response
- return response
+ response = process_embedding_response(
+ response=response,
+ request_model=request_model,
+ pricing_info=pricing_info,
+ server_port=server_port,
+ server_address=server_address,
+ environment=environment,
+ application_name=application_name,
+ metrics=metrics,
+ event_provider=event_provider,
+ start_time=start_time,
+ span=span,
+ capture_message_content=capture_message_content,
+ disable_metrics=disable_metrics,
+ version=version,
+ **kwargs
+ )
+
+ return response
return wrapper
diff --git a/sdk/python/src/openlit/instrumentation/ollama/utils.py b/sdk/python/src/openlit/instrumentation/ollama/utils.py
new file mode 100644
index 00000000..4fdb617a
--- /dev/null
+++ b/sdk/python/src/openlit/instrumentation/ollama/utils.py
@@ -0,0 +1,333 @@
+"""
+Ollama OpenTelemetry instrumentation utility functions
+"""
+import time
+
+from opentelemetry.sdk.resources import SERVICE_NAME, TELEMETRY_SDK_NAME, DEPLOYMENT_ENVIRONMENT
+from opentelemetry.trace import Status, StatusCode
+
+from openlit.__helpers import (
+ calculate_ttft,
+ response_as_dict,
+ calculate_tbt,
+ general_tokens,
+ extract_and_format_input,
+ get_chat_model_cost,
+ get_embed_model_cost,
+ handle_exception,
+ create_metrics_attributes,
+ otel_event,
+ concatenate_all_contents
+)
+from openlit.semcov import SemanticConvetion
+
+def process_chunk(self, chunk):
+ """
+ Process a chunk of response data and update state.
+ """
+
+ end_time = time.time()
+ # Record the timestamp for the current chunk
+ self._timestamps.append(end_time)
+
+ if len(self._timestamps) == 1:
+ # Calculate time to first chunk
+ self._ttft = calculate_ttft(self._timestamps, self._start_time)
+
+ chunked = response_as_dict(chunk)
+ self._llmresponse += chunked.get('message', {}).get('content', '')
+
+ if chunked.get('message', {}).get('tool_calls'):
+ self._tool_calls = chunked['message']['tool_calls']
+
+ if chunked.get('eval_count'):
+ self._response_role = chunked.get('message', {}).get('role', '')
+ self._input_tokens = chunked.get('prompt_eval_count', 0)
+ self._output_tokens = chunked.get('eval_count', 0)
+ self._response_model = chunked.get('model', '')
+ self._finish_reason = chunked.get('done_reason', '')
+
+def common_chat_logic(scope, pricing_info, environment, application_name, metrics,
+ event_provider, capture_message_content, disable_metrics, version, is_stream):
+ """
+ Process chat request and generate Telemetry
+ """
+
+ scope._end_time = time.time()
+ if len(scope._timestamps) > 1:
+ scope._tbt = calculate_tbt(scope._timestamps)
+
+ formatted_messages = extract_and_format_input(scope._kwargs.get("messages", ""))
+ request_model = scope._kwargs.get("model", "gpt-4o")
+
+ cost = get_chat_model_cost(request_model, pricing_info, scope._input_tokens, scope._output_tokens)
+
+ # Set Span attributes (OTel Semconv)
+ scope._span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
+ scope._span.set_attribute(SemanticConvetion.GEN_AI_OPERATION, SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT)
+ scope._span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM, SemanticConvetion.GEN_AI_SYSTEM_OLLAMA)
+ scope._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL, request_model)
+ scope._span.set_attribute(SemanticConvetion.SERVER_PORT, scope._server_port)
+
+ options = scope._kwargs.get('options', {})
+ attributes = [
+ (SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY, 'repeat_penalty'),
+ (SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS, 'max_tokens'),
+ (SemanticConvetion.GEN_AI_REQUEST_SEED, 'seed'),
+ (SemanticConvetion.GEN_AI_REQUEST_STOP_SEQUENCES, 'stop'),
+ (SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE, 'temperature'),
+ (SemanticConvetion.GEN_AI_REQUEST_TOP_P, 'top_p'),
+ (SemanticConvetion.GEN_AI_REQUEST_TOP_K, 'top_k'),
+ ]
+
+ for attribute, key in attributes:
+ value = options.get(key)
+ if value is not None:
+ scope._span.set_attribute(attribute, value)
+
+ scope._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON, [scope._finish_reason])
+ scope._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_MODEL, scope._response_model)
+ scope._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS, scope._input_tokens)
+ scope._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS, scope._output_tokens)
+ scope._span.set_attribute(SemanticConvetion.SERVER_ADDRESS, scope._server_address)
+
+ scope._span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
+ "text" if isinstance(scope._llmresponse, str) else "json")
+
+ scope._span.set_attribute(DEPLOYMENT_ENVIRONMENT, environment)
+ scope._span.set_attribute(SERVICE_NAME, application_name)
+ scope._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM, is_stream)
+ scope._span.set_attribute(SemanticConvetion.GEN_AI_CLIENT_TOKEN_USAGE, scope._input_tokens + scope._output_tokens)
+ scope._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST, cost)
+ scope._span.set_attribute(SemanticConvetion.GEN_AI_SERVER_TBT, scope._tbt)
+ scope._span.set_attribute(SemanticConvetion.GEN_AI_SERVER_TTFT, scope._ttft)
+ scope._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION, version)
+
+ # To be removed one the change to log events (from span events) is complete
+ prompt = concatenate_all_contents(formatted_messages)
+ if capture_message_content:
+ scope._span.add_event(
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
+ attributes={
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
+ },
+ )
+ scope._span.add_event(
+ name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
+ attributes={
+ SemanticConvetion.GEN_AI_CONTENT_COMPLETION: scope._llmresponse,
+ },
+ )
+
+ choice_event_body = {
+ "finish_reason": scope._finish_reason,
+ "index": 0,
+ "message": {
+ **({"content": scope._llmresponse} if capture_message_content else {}),
+ "role": scope._response_role
+ }
+ }
+
+ if scope._tool_calls:
+ print(scope._tool_calls)
+ function_call = scope._tool_calls[0]
+ choice_event_body["message"].update({
+ "tool_calls": {
+ "function": {
+ "name": function_call.get('function', {}).get('name', ''),
+ "arguments": function_call.get('function', {}).get('arguments', '')
+ },
+ "id": function_call.get('id', ''),
+ "type": "function"
+ }
+ })
+
+ # Emit events
+ for role in ['user', 'system', 'assistant', 'tool']:
+ if formatted_messages.get(role, {}).get('content', ''):
+ event = otel_event(
+ name=getattr(SemanticConvetion, f'GEN_AI_{role.upper()}_MESSAGE'),
+ attributes={
+ SemanticConvetion.GEN_AI_SYSTEM: SemanticConvetion.GEN_AI_SYSTEM_OLLAMA
+ },
+ body = {
+ # pylint: disable=line-too-long
+ **({"content": formatted_messages.get(role, {}).get('content', '')} if capture_message_content else {}),
+ "role": formatted_messages.get(role, {}).get('role', []),
+ **({
+ "tool_calls": {
+ "function": {
+ # pylint: disable=line-too-long
+ "name": (scope._tool_calls[0].get('function', {}).get('name', '') if scope._tool_calls else ''),
+ "arguments": (scope._tool_calls[0].get('function', {}).get('arguments', '') if scope._tool_calls else '')
+ },
+ "id": (scope._tool_calls[0].get('id', '') if scope._tool_calls else ''),
+ "type": "function"
+ }
+ } if role == 'assistant' else {}),
+ **({
+ "id": (scope._tool_calls[0].get('id', '') if scope._tool_calls else '')
+ } if role == 'tool' else {})
+ }
+ )
+ event_provider.emit(event)
+
+ choice_event = otel_event(
+ name=SemanticConvetion.GEN_AI_CHOICE,
+ attributes={
+ SemanticConvetion.GEN_AI_SYSTEM: SemanticConvetion.GEN_AI_SYSTEM_OLLAMA
+ },
+ body=choice_event_body
+ )
+ event_provider.emit(choice_event)
+
+ scope._span.set_status(Status(StatusCode.OK))
+
+ if not disable_metrics:
+ metrics_attributes = create_metrics_attributes(
+ service_name=application_name,
+ deployment_environment=environment,
+ operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
+ system=SemanticConvetion.GEN_AI_SYSTEM_OLLAMA,
+ request_model=request_model,
+ server_address=scope._server_address,
+ server_port=scope._server_port,
+ response_model=scope._response_model,
+ )
+
+ metrics["genai_client_usage_tokens"].record(scope._input_tokens + scope._output_tokens, metrics_attributes)
+ metrics["genai_client_operation_duration"].record(scope._end_time - scope._start_time, metrics_attributes)
+ metrics["genai_server_tbt"].record(scope._tbt, metrics_attributes)
+ metrics["genai_server_ttft"].record(scope._ttft, metrics_attributes)
+ metrics["genai_requests"].add(1, metrics_attributes)
+ metrics["genai_completion_tokens"].add(scope._output_tokens, metrics_attributes)
+ metrics["genai_prompt_tokens"].add(scope._input_tokens, metrics_attributes)
+ metrics["genai_cost"].record(cost, metrics_attributes)
+
+def process_streaming_chat_response(self, pricing_info, environment, application_name, metrics,
+ event_provider, capture_message_content=False, disable_metrics=False, version=''):
+ """
+ Process chat request and generate Telemetry
+ """
+
+ common_chat_logic(self, pricing_info, environment, application_name, metrics,
+ event_provider, capture_message_content, disable_metrics, version, is_stream=True)
+
+def process_chat_response(response, request_model, pricing_info, server_port, server_address,
+ environment, application_name, metrics, event_provider, start_time,
+ span, capture_message_content=False, disable_metrics=False, version="1.0.0", **kwargs):
+ """
+ Process chat request and generate Telemetry
+ """
+
+ self = type('GenericScope', (), {})()
+
+ # pylint: disable = no-member
+ self._start_time = start_time
+ self._end_time = time.time()
+ self._span = span
+ self._llmresponse = response.get('message', {}).get('content', '')
+ self._response_role = response.get('message', {}).get('role', 'assistant')
+ self._input_tokens = response.get('prompt_eval_count')
+ self._output_tokens = response.get('eval_count')
+ self._response_model = response.get('model', '')
+ self._finish_reason = response.get('done_reason', '')
+ self._timestamps = []
+ self._ttft, self._tbt = self._end_time - self._start_time, 0
+ self._server_address, self._server_port = server_address, server_port
+ self._kwargs = kwargs
+ self._tool_calls = response.get('message', {}).get('tool_calls', [])
+
+ common_chat_logic(self, pricing_info, environment, application_name, metrics,
+ event_provider, capture_message_content, disable_metrics, version, is_stream=False)
+
+ return response
+
+def process_embedding_response(response, request_model, pricing_info, server_port, server_address,
+ environment, application_name, metrics, event_provider,
+ start_time, span, capture_message_content=False, disable_metrics=False, version="1.0.0", **kwargs):
+ """
+ Process embedding request and generate Telemetry
+ """
+
+ end_time = time.time()
+
+ try:
+ input_tokens = general_tokens(str(kwargs.get('prompt')))
+
+ # Calculate cost of the operation
+ cost = get_embed_model_cost(request_model,
+ pricing_info, input_tokens)
+
+ # Set Span attributes (OTel Semconv)
+ span.set_attribute(TELEMETRY_SDK_NAME, 'openlit')
+ span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
+ SemanticConvetion.GEN_AI_OPERATION_TYPE_EMBEDDING)
+ span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
+ SemanticConvetion.GEN_AI_SYSTEM_OLLAMA)
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
+ request_model)
+ span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_MODEL,
+ request_model)
+ span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
+ server_address)
+ span.set_attribute(SemanticConvetion.SERVER_PORT,
+ server_port)
+ span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
+ input_tokens)
+
+ # Set Span attributes (Extras)
+ span.set_attribute(DEPLOYMENT_ENVIRONMENT,
+ environment)
+ span.set_attribute(SERVICE_NAME,
+ application_name)
+ span.set_attribute(SemanticConvetion.GEN_AI_CLIENT_TOKEN_USAGE,
+ input_tokens)
+ span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
+ cost)
+ span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
+ version)
+
+ prompt_event = otel_event(
+ name=SemanticConvetion.GEN_AI_USER_MESSAGE,
+ attributes={
+ SemanticConvetion.GEN_AI_SYSTEM: SemanticConvetion.GEN_AI_SYSTEM_OLLAMA
+ },
+ body={
+ **({"content": kwargs.get('prompt', '')} if capture_message_content else {}),
+ "role": 'user'
+ }
+ )
+ event_provider.emit(prompt_event)
+
+ span.set_status(Status(StatusCode.OK))
+
+ if disable_metrics is False:
+ attributes = create_metrics_attributes(
+ service_name=application_name,
+ deployment_environment=environment,
+ operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_EMBEDDING,
+ system=SemanticConvetion.GEN_AI_SYSTEM_OLLAMA,
+ request_model=request_model,
+ server_address=server_address,
+ server_port=server_port,
+ response_model=request_model,
+ )
+ metrics['genai_client_usage_tokens'].record(
+ input_tokens, attributes
+ )
+ metrics['genai_client_operation_duration'].record(
+ end_time - start_time, attributes
+ )
+ metrics['genai_requests'].add(1, attributes)
+ metrics['genai_prompt_tokens'].add(input_tokens, attributes)
+ metrics['genai_cost'].record(cost, attributes)
+
+ # Return original response
+ return response
+
+ except Exception as e:
+ handle_exception(span, e)
+
+ # Return original response
+ return response
diff --git a/sdk/python/src/openlit/instrumentation/openai/__init__.py b/sdk/python/src/openlit/instrumentation/openai/__init__.py
index 112675d6..51df68fd 100644
--- a/sdk/python/src/openlit/instrumentation/openai/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/openai/__init__.py
@@ -25,7 +25,7 @@ class OpenAIInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info")
- trace_content = kwargs.get("trace_content")
+ capture_message_content = kwargs.get("capture_message_content")
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("openai")
@@ -33,7 +33,7 @@ class OpenAIInstrumentor(BaseInstrumentor):
"openai.resources.chat.completions",
"Completions.create",
chat_completions(version, environment, application_name,
- tracer, pricing_info, trace_content,
+ tracer, pricing_info, capture_message_content,
metrics, disable_metrics),
)
@@ -41,7 +41,7 @@ class OpenAIInstrumentor(BaseInstrumentor):
"openai.resources.chat.completions",
"AsyncCompletions.create",
async_chat_completions(version, environment, application_name,
- tracer, pricing_info, trace_content,
+ tracer, pricing_info, capture_message_content,
metrics, disable_metrics),
)
@@ -49,7 +49,7 @@ class OpenAIInstrumentor(BaseInstrumentor):
"openai.resources.images",
"Images.generate",
image_generate(version, environment, application_name,
- tracer, pricing_info, trace_content,
+ tracer, pricing_info, capture_message_content,
metrics, disable_metrics),
)
@@ -57,7 +57,7 @@ class OpenAIInstrumentor(BaseInstrumentor):
"openai.resources.images",
"AsyncImages.generate",
async_image_generate(version, environment, application_name,
- tracer, pricing_info, trace_content,
+ tracer, pricing_info, capture_message_content,
metrics, disable_metrics),
)
@@ -65,7 +65,7 @@ class OpenAIInstrumentor(BaseInstrumentor):
"openai.resources.embeddings",
"Embeddings.create",
embedding(version, environment, application_name,
- tracer, pricing_info, trace_content,
+ tracer, pricing_info, capture_message_content,
metrics, disable_metrics),
)
@@ -73,7 +73,7 @@ class OpenAIInstrumentor(BaseInstrumentor):
"openai.resources.embeddings",
"AsyncEmbeddings.create",
async_embedding(version, environment, application_name,
- tracer, pricing_info, trace_content,
+ tracer, pricing_info, capture_message_content,
metrics, disable_metrics),
)
@@ -82,7 +82,7 @@ class OpenAIInstrumentor(BaseInstrumentor):
"Images.create_variation",
image_variatons(version,
environment, application_name,
- tracer, pricing_info, trace_content,
+ tracer, pricing_info, capture_message_content,
metrics, disable_metrics),
)
@@ -91,7 +91,7 @@ class OpenAIInstrumentor(BaseInstrumentor):
"AsyncImages.create_variation",
async_image_variatons(version,
environment, application_name,
- tracer, pricing_info, trace_content,
+ tracer, pricing_info, capture_message_content,
metrics, disable_metrics),
)
@@ -99,7 +99,7 @@ class OpenAIInstrumentor(BaseInstrumentor):
"openai.resources.audio.speech",
"Speech.create",
audio_create(version, environment, application_name,
- tracer, pricing_info, trace_content,
+ tracer, pricing_info, capture_message_content,
metrics, disable_metrics),
)
@@ -107,7 +107,7 @@ class OpenAIInstrumentor(BaseInstrumentor):
"openai.resources.audio.speech",
"AsyncSpeech.create",
async_audio_create(version, environment, application_name,
- tracer, pricing_info, trace_content,
+ tracer, pricing_info, capture_message_content,
metrics, disable_metrics),
)
diff --git a/sdk/python/src/openlit/instrumentation/openai/async_openai.py b/sdk/python/src/openlit/instrumentation/openai/async_openai.py
index 8ec408b0..8ddb7846 100644
--- a/sdk/python/src/openlit/instrumentation/openai/async_openai.py
+++ b/sdk/python/src/openlit/instrumentation/openai/async_openai.py
@@ -25,7 +25,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def async_chat_completions(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -35,7 +35,7 @@ def async_chat_completions(version, environment, application_name,
application_name: Name of the application using the OpenAI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of OpenAI usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
@@ -229,7 +229,7 @@ def async_chat_completions(version, environment, application_name,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -387,7 +387,7 @@ def async_chat_completions(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SERVICE_TIER,
response_dict.get('service_tier'))
span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SYSTEM_FINGERPRINT,
- response_dict.get('system_fingerprint'))
+ response_dict.get('system_fingerprint', ''))
# Set base span attribues (Extras)
span.set_attribute(DEPLOYMENT_ENVIRONMENT,
@@ -406,7 +406,7 @@ def async_chat_completions(version, environment, application_name,
end_time - start_time)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -417,7 +417,7 @@ def async_chat_completions(version, environment, application_name,
for i in range(kwargs.get('n',1)):
span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
[response_dict.get('choices')[i].get('finish_reason')])
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
attributes={
@@ -477,7 +477,7 @@ def async_chat_completions(version, environment, application_name,
return wrapper
def async_embedding(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for embeddings to collect metrics.
@@ -487,7 +487,7 @@ def async_embedding(version, environment, application_name,
application_name: Name of the application using the OpenAI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of OpenAI usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the embeddings method to add telemetry.
@@ -561,7 +561,7 @@ def async_embedding(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -605,7 +605,7 @@ def async_embedding(version, environment, application_name,
return wrapper
def async_image_generate(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for image generation to collect metrics.
@@ -615,7 +615,7 @@ def async_image_generate(version, environment, application_name,
application_name: Name of the application using the OpenAI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of OpenAI image generation.
- trace_content: Flag indicating whether to trace the input prompt and generated images.
+ capture_message_content: Flag indicating whether to trace the input prompt and generated images.
Returns:
A function that wraps the image generation method to add telemetry.
@@ -700,7 +700,7 @@ def async_image_generate(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -752,7 +752,7 @@ def async_image_generate(version, environment, application_name,
return wrapper
def async_image_variatons(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for creating image variations to collect metrics.
@@ -762,7 +762,7 @@ def async_image_variatons(version, environment, application_name,
application_name: Name of the application using the OpenAI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of generating image variations.
- trace_content: Flag indicating whether to trace the input image and generated variations.
+ capture_message_content: Flag indicating whether to trace the input image and generated variations.
Returns:
A function that wraps the image variations creation method to add telemetry.
@@ -842,7 +842,7 @@ def async_image_variatons(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
attribute_name = f"{SemanticConvetion.GEN_AI_RESPONSE_IMAGE}.{images_count}"
span.add_event(
name=attribute_name,
@@ -888,7 +888,7 @@ def async_image_variatons(version, environment, application_name,
return wrapper
def async_audio_create(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for creating speech audio to collect metrics.
@@ -898,7 +898,7 @@ def async_audio_create(version, environment, application_name,
application_name: Name of the application using the OpenAI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of generating speech audio.
- trace_content: Flag indicating whether to trace the input text and generated audio.
+ capture_message_content: Flag indicating whether to trace the input text and generated audio.
Returns:
A function that wraps the speech audio creation method to add telemetry.
@@ -967,7 +967,7 @@ def async_audio_create(version, environment, application_name,
cost)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/openai/openai.py b/sdk/python/src/openlit/instrumentation/openai/openai.py
index 8c6378c4..3251b053 100644
--- a/sdk/python/src/openlit/instrumentation/openai/openai.py
+++ b/sdk/python/src/openlit/instrumentation/openai/openai.py
@@ -25,7 +25,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def chat_completions(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -35,7 +35,7 @@ def chat_completions(version, environment, application_name,
application_name: Name of the application using the OpenAI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of OpenAI usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
@@ -229,7 +229,7 @@ def chat_completions(version, environment, application_name,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -387,7 +387,7 @@ def chat_completions(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SERVICE_TIER,
response_dict.get('service_tier'))
span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SYSTEM_FINGERPRINT,
- response_dict.get('system_fingerprint'))
+ str(response_dict.get('system_fingerprint', '')))
# Set base span attribues (Extras)
span.set_attribute(DEPLOYMENT_ENVIRONMENT,
@@ -406,7 +406,7 @@ def chat_completions(version, environment, application_name,
end_time - start_time)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -417,7 +417,7 @@ def chat_completions(version, environment, application_name,
for i in range(kwargs.get('n',1)):
span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
[response_dict.get('choices')[i].get('finish_reason')])
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
attributes={
@@ -477,7 +477,7 @@ def chat_completions(version, environment, application_name,
return wrapper
def embedding(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for embeddings to collect metrics.
@@ -487,7 +487,7 @@ def embedding(version, environment, application_name,
application_name: Name of the application using the OpenAI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of OpenAI usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the embeddings method to add telemetry.
@@ -561,7 +561,7 @@ def embedding(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -605,7 +605,7 @@ def embedding(version, environment, application_name,
return wrapper
def image_generate(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for image generation to collect metrics.
@@ -615,7 +615,7 @@ def image_generate(version, environment, application_name,
application_name: Name of the application using the OpenAI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of OpenAI image generation.
- trace_content: Flag indicating whether to trace the input prompt and generated images.
+ capture_message_content: Flag indicating whether to trace the input prompt and generated images.
Returns:
A function that wraps the image generation method to add telemetry.
@@ -700,7 +700,7 @@ def image_generate(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -752,7 +752,7 @@ def image_generate(version, environment, application_name,
return wrapper
def image_variatons(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for creating image variations to collect metrics.
@@ -762,7 +762,7 @@ def image_variatons(version, environment, application_name,
application_name: Name of the application using the OpenAI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of generating image variations.
- trace_content: Flag indicating whether to trace the input image and generated variations.
+ capture_message_content: Flag indicating whether to trace the input image and generated variations.
Returns:
A function that wraps the image variations creation method to add telemetry.
@@ -842,7 +842,7 @@ def image_variatons(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
attribute_name = f"{SemanticConvetion.GEN_AI_RESPONSE_IMAGE}.{images_count}"
span.add_event(
name=attribute_name,
@@ -888,7 +888,7 @@ def image_variatons(version, environment, application_name,
return wrapper
def audio_create(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for creating speech audio to collect metrics.
@@ -898,7 +898,7 @@ def audio_create(version, environment, application_name,
application_name: Name of the application using the OpenAI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of generating speech audio.
- trace_content: Flag indicating whether to trace the input text and generated audio.
+ capture_message_content: Flag indicating whether to trace the input text and generated audio.
Returns:
A function that wraps the speech audio creation method to add telemetry.
@@ -967,7 +967,7 @@ def audio_create(version, environment, application_name,
cost)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/phidata/__init__.py b/sdk/python/src/openlit/instrumentation/phidata/__init__.py
index 03ee5657..ea6c433f 100644
--- a/sdk/python/src/openlit/instrumentation/phidata/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/phidata/__init__.py
@@ -26,7 +26,7 @@ class PhidataInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("phidata")
@@ -34,7 +34,7 @@ class PhidataInstrumentor(BaseInstrumentor):
"phi.agent",
"Agent.print_response",
phidata_wrap("phidata.print_response", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/phidata/phidata.py b/sdk/python/src/openlit/instrumentation/phidata/phidata.py
index f18dd444..17443a22 100644
--- a/sdk/python/src/openlit/instrumentation/phidata/phidata.py
+++ b/sdk/python/src/openlit/instrumentation/phidata/phidata.py
@@ -15,7 +15,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def phidata_wrap(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -26,7 +26,7 @@ def phidata_wrap(gen_ai_endpoint, version, environment, application_name,
application_name: Name of the application using the Phidata Agent.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Phidata usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
diff --git a/sdk/python/src/openlit/instrumentation/pinecone/__init__.py b/sdk/python/src/openlit/instrumentation/pinecone/__init__.py
index 5b9741d1..2575bf85 100644
--- a/sdk/python/src/openlit/instrumentation/pinecone/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/pinecone/__init__.py
@@ -21,7 +21,7 @@ class PineconeInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info")
- trace_content = kwargs.get("trace_content")
+ capture_message_content = kwargs.get("capture_message_content")
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("pinecone-client")
@@ -29,35 +29,35 @@ class PineconeInstrumentor(BaseInstrumentor):
"pinecone.control.pinecone",
"Pinecone.create_index",
general_wrap("pinecone.create_index", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"pinecone.data.index",
"Index.upsert",
general_wrap("pinecone.upsert", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"pinecone.data.index",
"Index.query",
general_wrap("pinecone.query", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"pinecone.data.index",
"Index.update",
general_wrap("pinecone.update", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"pinecone.data.index",
"Index.delete",
general_wrap("pinecone.delete", version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
diff --git a/sdk/python/src/openlit/instrumentation/pinecone/pinecone.py b/sdk/python/src/openlit/instrumentation/pinecone/pinecone.py
index 6f44dc2b..212e59fb 100644
--- a/sdk/python/src/openlit/instrumentation/pinecone/pinecone.py
+++ b/sdk/python/src/openlit/instrumentation/pinecone/pinecone.py
@@ -23,7 +23,7 @@ def object_count(obj):
return None
def general_wrap(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Wraps a Pinecone operation to trace and log its execution metrics.
@@ -40,7 +40,7 @@ def general_wrap(gen_ai_endpoint, version, environment, application_name,
- application_name (str): The name of the application performing the Pinecone operation.
- tracer (opentelemetry.trace.Tracer): An object used for OpenTelemetry tracing.
- pricing_info (dict): Information about pricing, not used in current implementation.
- - trace_content (bool): A flag indicating whether the content of responses should be traced.
+ - capture_message_content (bool): A flag indicating whether the content of responses should be traced.
Returns:
- function: A decorator function that, when applied, wraps the target function with
diff --git a/sdk/python/src/openlit/instrumentation/premai/__init__.py b/sdk/python/src/openlit/instrumentation/premai/__init__.py
index ca1218db..bd22aea1 100644
--- a/sdk/python/src/openlit/instrumentation/premai/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/premai/__init__.py
@@ -26,7 +26,7 @@ class PremAIInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("premai")
@@ -35,7 +35,7 @@ class PremAIInstrumentor(BaseInstrumentor):
"premai.api",
"ChatCompletionsModule.create",
chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# sync embedding
@@ -43,7 +43,7 @@ class PremAIInstrumentor(BaseInstrumentor):
"premai.api",
"EmbeddingsModule.create",
embedding(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/premai/premai.py b/sdk/python/src/openlit/instrumentation/premai/premai.py
index 20469b98..955532a7 100644
--- a/sdk/python/src/openlit/instrumentation/premai/premai.py
+++ b/sdk/python/src/openlit/instrumentation/premai/premai.py
@@ -22,7 +22,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -32,7 +32,7 @@ def chat(version, environment, application_name,
application_name: Name of the application using the PremAI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of PremAI usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
@@ -199,7 +199,7 @@ def chat(version, environment, application_name,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -363,7 +363,7 @@ def chat(version, environment, application_name,
end_time - start_time)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -429,7 +429,7 @@ def chat(version, environment, application_name,
return wrapper
def embedding(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for embeddings to collect metrics.
@@ -439,7 +439,7 @@ def embedding(version, environment, application_name,
application_name: Name of the application using the PremAI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of PremAI usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the embeddings method to add telemetry.
@@ -512,7 +512,7 @@ def embedding(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/qdrant/__init__.py b/sdk/python/src/openlit/instrumentation/qdrant/__init__.py
index 90a58f0c..8cff3cd4 100644
--- a/sdk/python/src/openlit/instrumentation/qdrant/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/qdrant/__init__.py
@@ -273,7 +273,7 @@ class QdrantInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info")
- trace_content = kwargs.get("trace_content")
+ capture_message_content = kwargs.get("capture_message_content")
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("qdrant-client")
@@ -286,7 +286,7 @@ class QdrantInstrumentor(BaseInstrumentor):
wrap_package,
wrap_object,
wrapper(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
diff --git a/sdk/python/src/openlit/instrumentation/qdrant/async_qdrant.py b/sdk/python/src/openlit/instrumentation/qdrant/async_qdrant.py
index e0abf62c..24d213ee 100644
--- a/sdk/python/src/openlit/instrumentation/qdrant/async_qdrant.py
+++ b/sdk/python/src/openlit/instrumentation/qdrant/async_qdrant.py
@@ -25,7 +25,7 @@ def object_count(obj):
return cnt
def async_general_wrap(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Creates a wrapper around a function call to trace and log its execution metrics.
@@ -39,7 +39,7 @@ def async_general_wrap(gen_ai_endpoint, version, environment, application_name,
- application_name (str): Name of the Langchain application.
- tracer (opentelemetry.trace.Tracer): The tracer object used for OpenTelemetry tracing.
- pricing_info (dict): Information about the pricing for internal metrics (currently not used).
- - trace_content (bool): Flag indicating whether to trace the content of the response.
+ - capture_message_content (bool): Flag indicating whether to trace the content of the response.
Returns:
- function: A higher-order function that takes a function 'wrapped' and returns
diff --git a/sdk/python/src/openlit/instrumentation/qdrant/qdrant.py b/sdk/python/src/openlit/instrumentation/qdrant/qdrant.py
index efe99962..4a07a51d 100644
--- a/sdk/python/src/openlit/instrumentation/qdrant/qdrant.py
+++ b/sdk/python/src/openlit/instrumentation/qdrant/qdrant.py
@@ -25,7 +25,7 @@ def object_count(obj):
return cnt
def general_wrap(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Creates a wrapper around a function call to trace and log its execution metrics.
@@ -39,7 +39,7 @@ def general_wrap(gen_ai_endpoint, version, environment, application_name,
- application_name (str): Name of the Langchain application.
- tracer (opentelemetry.trace.Tracer): The tracer object used for OpenTelemetry tracing.
- pricing_info (dict): Information about the pricing for internal metrics (currently not used).
- - trace_content (bool): Flag indicating whether to trace the content of the response.
+ - capture_message_content (bool): Flag indicating whether to trace the content of the response.
Returns:
- function: A higher-order function that takes a function 'wrapped' and returns
diff --git a/sdk/python/src/openlit/instrumentation/reka/__init__.py b/sdk/python/src/openlit/instrumentation/reka/__init__.py
index ef6f36eb..ff1a6495 100644
--- a/sdk/python/src/openlit/instrumentation/reka/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/reka/__init__.py
@@ -29,7 +29,7 @@ class RekaInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("reka-api")
@@ -38,7 +38,7 @@ class RekaInstrumentor(BaseInstrumentor):
"reka.chat.client",
"ChatClient.create",
chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# async chat
@@ -46,7 +46,7 @@ class RekaInstrumentor(BaseInstrumentor):
"reka.chat.client",
"AsyncChatClient.create",
async_chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/reka/async_reka.py b/sdk/python/src/openlit/instrumentation/reka/async_reka.py
index 791ebcd0..e35fc077 100644
--- a/sdk/python/src/openlit/instrumentation/reka/async_reka.py
+++ b/sdk/python/src/openlit/instrumentation/reka/async_reka.py
@@ -18,7 +18,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def async_chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat to collect metrics.
@@ -28,7 +28,7 @@ def async_chat(version, environment, application_name,
application_name: Name of the application using the Reka API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Reka usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat method to add telemetry.
@@ -142,7 +142,7 @@ def async_chat(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/reka/reka.py b/sdk/python/src/openlit/instrumentation/reka/reka.py
index dd93e376..755f799c 100644
--- a/sdk/python/src/openlit/instrumentation/reka/reka.py
+++ b/sdk/python/src/openlit/instrumentation/reka/reka.py
@@ -18,7 +18,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def chat(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat to collect metrics.
@@ -28,7 +28,7 @@ def chat(version, environment, application_name,
application_name: Name of the application using the Reka API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Reka usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat method to add telemetry.
@@ -142,7 +142,7 @@ def chat(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/together/__init__.py b/sdk/python/src/openlit/instrumentation/together/__init__.py
index 67f9398f..a60c5af9 100644
--- a/sdk/python/src/openlit/instrumentation/together/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/together/__init__.py
@@ -29,7 +29,7 @@ class TogetherInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("together")
@@ -38,7 +38,7 @@ class TogetherInstrumentor(BaseInstrumentor):
"together.resources.chat.completions",
"ChatCompletions.create",
completion(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# Image generate
@@ -46,7 +46,7 @@ class TogetherInstrumentor(BaseInstrumentor):
"together.resources.images",
"Images.generate",
image_generate(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# Chat completions
@@ -54,7 +54,7 @@ class TogetherInstrumentor(BaseInstrumentor):
"together.resources.chat.completions",
"AsyncChatCompletions.create",
async_completion(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
# Image generate
@@ -62,7 +62,7 @@ class TogetherInstrumentor(BaseInstrumentor):
"together.resources.images",
"AsyncImages.generate",
async_image_generate(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/together/async_together.py b/sdk/python/src/openlit/instrumentation/together/async_together.py
index 7c1905ff..8fd20a03 100644
--- a/sdk/python/src/openlit/instrumentation/together/async_together.py
+++ b/sdk/python/src/openlit/instrumentation/together/async_together.py
@@ -22,7 +22,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def async_completion(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -32,7 +32,7 @@ def async_completion(version, environment, application_name,
application_name: Name of the application using the Together AI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Together AI usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
@@ -218,7 +218,7 @@ def async_completion(version, environment, application_name,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -388,7 +388,7 @@ def async_completion(version, environment, application_name,
end_time - start_time)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -399,7 +399,7 @@ def async_completion(version, environment, application_name,
for i in range(kwargs.get('n',1)):
span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
[str(response_dict.get('choices')[i].get('finish_reason'))])
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
attributes={
@@ -459,7 +459,7 @@ def async_completion(version, environment, application_name,
return wrapper
def async_image_generate(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for image generation to collect metrics.
@@ -469,7 +469,7 @@ def async_image_generate(version, environment, application_name,
application_name: Name of the application using the Together AI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Together AI image generation.
- trace_content: Flag indicating whether to trace the input prompt and generated images.
+ capture_message_content: Flag indicating whether to trace the input prompt and generated images.
Returns:
A function that wraps the image generation method to add telemetry.
@@ -548,7 +548,7 @@ def async_image_generate(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/together/together.py b/sdk/python/src/openlit/instrumentation/together/together.py
index a19ebe8e..cd775962 100644
--- a/sdk/python/src/openlit/instrumentation/together/together.py
+++ b/sdk/python/src/openlit/instrumentation/together/together.py
@@ -22,7 +22,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def completion(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
@@ -32,7 +32,7 @@ def completion(version, environment, application_name,
application_name: Name of the application using the Together AI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Together AI usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat completions method to add telemetry.
@@ -218,7 +218,7 @@ def completion(version, environment, application_name,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -388,7 +388,7 @@ def completion(version, environment, application_name,
end_time - start_time)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -399,7 +399,7 @@ def completion(version, environment, application_name,
for i in range(kwargs.get('n',1)):
span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
[str(response_dict.get('choices')[i].get('finish_reason'))])
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
attributes={
@@ -459,7 +459,7 @@ def completion(version, environment, application_name,
return wrapper
def image_generate(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for image generation to collect metrics.
@@ -469,7 +469,7 @@ def image_generate(version, environment, application_name,
application_name: Name of the application using the Together AI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of Together AI image generation.
- trace_content: Flag indicating whether to trace the input prompt and generated images.
+ capture_message_content: Flag indicating whether to trace the input prompt and generated images.
Returns:
A function that wraps the image generation method to add telemetry.
@@ -548,7 +548,7 @@ def image_generate(version, environment, application_name,
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/transformers/__init__.py b/sdk/python/src/openlit/instrumentation/transformers/__init__.py
index 6ad105fb..d38d8f47 100644
--- a/sdk/python/src/openlit/instrumentation/transformers/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/transformers/__init__.py
@@ -21,7 +21,7 @@ class TransformersInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info")
- trace_content = kwargs.get("trace_content")
+ capture_message_content = kwargs.get("capture_message_content")
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("transformers")
@@ -29,7 +29,7 @@ class TransformersInstrumentor(BaseInstrumentor):
"transformers",
"TextGenerationPipeline.__call__",
text_wrap(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
@staticmethod
diff --git a/sdk/python/src/openlit/instrumentation/transformers/transformers.py b/sdk/python/src/openlit/instrumentation/transformers/transformers.py
index 9d9dd739..8a1550c8 100644
--- a/sdk/python/src/openlit/instrumentation/transformers/transformers.py
+++ b/sdk/python/src/openlit/instrumentation/transformers/transformers.py
@@ -19,7 +19,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def text_wrap(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Creates a wrapper around a function call to trace and log its execution metrics.
@@ -32,7 +32,7 @@ def text_wrap(version, environment, application_name,
- application_name (str): Name of the Langchain application.
- tracer (opentelemetry.trace.Tracer): The tracer object used for OpenTelemetry tracing.
- pricing_info (dict): Information about the pricing for internal metrics (currently not used).
- - trace_content (bool): Flag indicating whether to trace the content of the response.
+ - capture_message_content (bool): Flag indicating whether to trace the content of the response.
Returns:
- function: A higher-order function that takes a function 'wrapped' and returns
@@ -113,7 +113,7 @@ def text_wrap(version, environment, application_name,
end_time - start_time)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -128,7 +128,7 @@ def text_wrap(version, environment, application_name,
attribute_name = f"gen_ai.content.completion.{i}"
else:
attribute_name = SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT
- if trace_content:
+ if capture_message_content:
# pylint: disable=bare-except
try:
llm_response = completion.get('generated_text', '')
diff --git a/sdk/python/src/openlit/instrumentation/vertexai/__init__.py b/sdk/python/src/openlit/instrumentation/vertexai/__init__.py
index 24901f68..a7ba0b94 100644
--- a/sdk/python/src/openlit/instrumentation/vertexai/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/vertexai/__init__.py
@@ -30,7 +30,7 @@ class VertexAIInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("google-cloud-aiplatform")
@@ -39,28 +39,28 @@ class VertexAIInstrumentor(BaseInstrumentor):
"vertexai.generative_models",
"GenerativeModel.generate_content",
send_message(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"vertexai.generative_models",
"ChatSession.send_message",
send_message(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"vertexai.language_models",
"ChatSession.send_message",
send_message(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"vertexai.language_models",
"ChatSession.send_message_streaming",
send_message(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
#async
@@ -68,28 +68,28 @@ class VertexAIInstrumentor(BaseInstrumentor):
"vertexai.generative_models",
"GenerativeModel.generate_content_async",
async_send_message(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"vertexai.generative_models",
"ChatSession.send_message_async",
async_send_message(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"vertexai.language_models",
"ChatSession.send_message_async",
async_send_message(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
wrap_function_wrapper(
"vertexai.language_models",
"ChatSession.send_message_streaming_async",
async_send_message(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/vertexai/async_vertexai.py b/sdk/python/src/openlit/instrumentation/vertexai/async_vertexai.py
index e9900578..aa232ae4 100644
--- a/sdk/python/src/openlit/instrumentation/vertexai/async_vertexai.py
+++ b/sdk/python/src/openlit/instrumentation/vertexai/async_vertexai.py
@@ -19,7 +19,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def async_send_message(version, environment, application_name, tracer,
- pricing_info, trace_content, metrics, disable_metrics):
+ pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for messages to collect metrics.
@@ -29,7 +29,7 @@ def async_send_message(version, environment, application_name, tracer,
application_name: Name of the application using the VertexAI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of VertexAI usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat method to add telemetry.
@@ -200,7 +200,7 @@ def async_send_message(version, environment, application_name, tracer,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -397,7 +397,7 @@ def async_send_message(version, environment, application_name, tracer,
end_time - start_time)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/vertexai/vertexai.py b/sdk/python/src/openlit/instrumentation/vertexai/vertexai.py
index 9ba2fc89..12ca4c51 100644
--- a/sdk/python/src/openlit/instrumentation/vertexai/vertexai.py
+++ b/sdk/python/src/openlit/instrumentation/vertexai/vertexai.py
@@ -19,7 +19,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def send_message(version, environment, application_name, tracer,
- pricing_info, trace_content, metrics, disable_metrics):
+ pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for messages to collect metrics.
@@ -29,7 +29,7 @@ def send_message(version, environment, application_name, tracer,
application_name: Name of the application using the VertexAI API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of VertexAI usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the chat method to add telemetry.
@@ -200,7 +200,7 @@ def send_message(version, environment, application_name, tracer,
self._ttft)
self._span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
self._span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
@@ -397,7 +397,7 @@ def send_message(version, environment, application_name, tracer,
end_time - start_time)
span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
version)
- if trace_content:
+ if capture_message_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
diff --git a/sdk/python/src/openlit/instrumentation/vllm/__init__.py b/sdk/python/src/openlit/instrumentation/vllm/__init__.py
index 337a79a6..acd3b083 100644
--- a/sdk/python/src/openlit/instrumentation/vllm/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/vllm/__init__.py
@@ -26,7 +26,7 @@ class VLLMInstrumentor(BaseInstrumentor):
tracer = kwargs.get("tracer")
metrics = kwargs.get("metrics_dict")
pricing_info = kwargs.get("pricing_info", {})
- trace_content = kwargs.get("trace_content", False)
+ capture_message_content = kwargs.get("capture_message_content", False)
disable_metrics = kwargs.get("disable_metrics")
version = importlib.metadata.version("vllm")
@@ -35,7 +35,7 @@ class VLLMInstrumentor(BaseInstrumentor):
"vllm",
"LLM.generate",
generate(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics),
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics),
)
def _uninstrument(self, **kwargs):
diff --git a/sdk/python/src/openlit/instrumentation/vllm/vllm.py b/sdk/python/src/openlit/instrumentation/vllm/vllm.py
index 7e6facc7..5d884c76 100644
--- a/sdk/python/src/openlit/instrumentation/vllm/vllm.py
+++ b/sdk/python/src/openlit/instrumentation/vllm/vllm.py
@@ -19,7 +19,7 @@ from openlit.semcov import SemanticConvetion
logger = logging.getLogger(__name__)
def generate(version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+ tracer, pricing_info, capture_message_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for generate to collect metrics.
@@ -29,7 +29,7 @@ def generate(version, environment, application_name,
application_name: Name of the application using the vLLM API.
tracer: OpenTelemetry tracer for creating spans.
pricing_info: Information used for calculating the cost of vLLM usage.
- trace_content: Flag indicating whether to trace the actual content.
+ capture_message_content: Flag indicating whether to trace the actual content.
Returns:
A function that wraps the generate method to add telemetry.
@@ -97,7 +97,7 @@ def generate(version, environment, application_name,
output_tokens = 0
cost = 0
- if trace_content:
+ if capture_message_content:
prompt_attributes = {}
completion_attributes = {}
diff --git a/sdk/python/src/openlit/otel/events.py b/sdk/python/src/openlit/otel/events.py
new file mode 100644
index 00000000..1eb782b2
--- /dev/null
+++ b/sdk/python/src/openlit/otel/events.py
@@ -0,0 +1,85 @@
+"""
+Setups up OpenTelemetry events emitter
+"""
+
+import os
+from opentelemetry import _events, _logs
+from opentelemetry.sdk.resources import SERVICE_NAME, TELEMETRY_SDK_NAME, DEPLOYMENT_ENVIRONMENT
+from opentelemetry.sdk.resources import Resource
+from opentelemetry.sdk._events import EventLoggerProvider
+from opentelemetry.sdk._logs.export import BatchLogRecordProcessor, SimpleLogRecordProcessor
+from opentelemetry.sdk._logs import LoggerProvider
+from opentelemetry.sdk._logs.export import ConsoleLogExporter
+
+if os.environ.get("OTEL_EXPORTER_OTLP_PROTOCOL") == "grpc":
+ from opentelemetry.exporter.otlp.proto.grpc._log_exporter import OTLPLogExporter
+else:
+ from opentelemetry.exporter.otlp.proto.http._log_exporter import OTLPLogExporter
+
+# Global flag to check if the events provider initialization is complete.
+EVENTS_SET = False
+
+def setup_events(application_name, environment, event_logger, otlp_endpoint, otlp_headers, disable_batch):
+ """Setup OpenTelemetry events with the given configuration.
+
+ Args:
+ application_name: Name of the application
+ environment: Deployment environment
+ event_logger: Optional pre-configured event logger provider
+ otlp_endpoint: Optional OTLP endpoint for exporter
+ otlp_headers: Optional headers for OTLP exporter
+
+ Returns:
+ EventLoggerProvider: The configured event logger provider
+ """
+ # If an external events_logger is provided, return it immediately.
+ if event_logger:
+ return event_logger
+
+ # Proceed with setting up a new events or configuration only if EVENTS_SET is False.
+ global EVENTS_SET
+
+ try:
+ if not EVENTS_SET:
+ # Create resource with service and environment information
+ resource = Resource.create(attributes={
+ SERVICE_NAME: application_name,
+ DEPLOYMENT_ENVIRONMENT: environment,
+ TELEMETRY_SDK_NAME: "openlit"}
+ )
+
+ # Initialize the LoggerProvider with the created resource.
+ logger_provider = LoggerProvider(resource=resource)
+
+ # Only set environment variables if you have a non-None value.
+ if otlp_endpoint is not None:
+ os.environ["OTEL_EXPORTER_OTLP_ENDPOINT"] = otlp_endpoint
+
+ if otlp_headers is not None:
+ if isinstance(otlp_headers, dict):
+ headers_str = ','.join(f"{key}={value}" for key, value in otlp_headers.items())
+ else:
+ headers_str = otlp_headers
+
+ os.environ["OTEL_EXPORTER_OTLP_HEADERS"] = headers_str
+
+ # Configure the span exporter and processor based on whether the endpoint is effectively set.
+ if os.getenv("OTEL_EXPORTER_OTLP_ENDPOINT"):
+ event_exporter = OTLPLogExporter()
+ # pylint: disable=line-too-long
+ logger_provider.add_log_record_processor(SimpleLogRecordProcessor(event_exporter)) if disable_batch else logger_provider.add_log_record_processor(BatchLogRecordProcessor(event_exporter))
+ else:
+ event_exporter = ConsoleLogExporter()
+ logger_provider.add_log_record_processor(SimpleLogRecordProcessor(event_exporter))
+
+ _logs.set_logger_provider(logger_provider)
+ event_provider = EventLoggerProvider()
+ _events.set_event_logger_provider(event_provider)
+
+ EVENTS_SET = True
+
+ return _events.get_event_logger(__name__)
+
+ # pylint: disable=bare-except
+ except:
+ return None
diff --git a/sdk/python/src/openlit/otel/tracing.py b/sdk/python/src/openlit/otel/tracing.py
index d49017e8..a2aa9345 100644
--- a/sdk/python/src/openlit/otel/tracing.py
+++ b/sdk/python/src/openlit/otel/tracing.py
@@ -1,4 +1,3 @@
-# pylint: disable=duplicate-code, line-too-long, ungrouped-imports
"""
Setups up OpenTelemetry tracer
"""
@@ -21,17 +20,8 @@ TRACER_SET = False
def setup_tracing(application_name, environment, tracer, otlp_endpoint, otlp_headers, disable_batch):
"""
- Sets up tracing with OpenTelemetry. Initializes the tracer provider and configures the span processor and exporter.
-
- Params:
- application_name (str): The name of the application to be used in traces.
- tracer (Tracer): Optional custom tracer. If provided, it is immediately returned and no setup is performed.
- otlp_endpoint (str): The OTLP exporter endpoint. Falls back to the OTEL_EXPORTER_OTLP_ENDPOINT environment variable if not specified.
- otlp_headers (dict): Headers for the OTLP request. Falls back to the OTEL_EXPORTER_OTLP_HEADERS environment variable if not specified.
- disable_batch (bool): Flag to disable the batch span processor in favor of a simpler processor for exporting.
-
- Returns:
- The provided custom tracer if not None; otherwise, a tracer instance configured according to the given parameters or environment variables.
+ Sets up tracing with OpenTelemetry.
+ Initializes the tracer provider and configures the span processor and exporter.
"""
# If an external tracer is provided, return it immediately.
@@ -72,7 +62,7 @@ def setup_tracing(application_name, environment, tracer, otlp_endpoint, otlp_hea
# Configure the span exporter and processor based on whether the endpoint is effectively set.
if os.getenv("OTEL_EXPORTER_OTLP_ENDPOINT"):
span_exporter = OTLPSpanExporter()
-
+ # pylint: disable=line-too-long
span_processor = BatchSpanProcessor(span_exporter) if not disable_batch else SimpleSpanProcessor(span_exporter)
else:
span_exporter = ConsoleSpanExporter()
diff --git a/sdk/python/src/openlit/semcov/__init__.py b/sdk/python/src/openlit/semcov/__init__.py
index 84abc899..0d4300e5 100644
--- a/sdk/python/src/openlit/semcov/__init__.py
+++ b/sdk/python/src/openlit/semcov/__init__.py
@@ -28,6 +28,13 @@ class SemanticConvetion:
GEN_AI_SERVER_TBT = "gen_ai.server.time_per_output_token"
GEN_AI_SERVER_TTFT = "gen_ai.server.time_to_first_token"
+ # GenAI Event Names (OTel Semconv)
+ GEN_AI_USER_MESSAGE = "gen_ai.user.message"
+ GEN_AI_SYSTEM_MESSAGE = "gen_ai.system.message"
+ GEN_AI_ASSISTANT_MESSAGE = "gen_ai.assistant.message"
+ GEN_AI_TOOL_MESSAGE = "gen_ai.tools.message"
+ GEN_AI_CHOICE = "gen_ai.choice"
+
# GenAI Request Attributes (OTel Semconv)
GEN_AI_OPERATION = "gen_ai.operation.name"
GEN_AI_SYSTEM = "gen_ai.system"
@@ -64,6 +71,7 @@ class SemanticConvetion:
GEN_AI_OPERATION_TYPE_FRAMEWORK = "framework"
GEN_AI_OPERATION_TYPE_AGENT = "agent"
GEN_AI_OPERATION_TYPE_CREATE_AGENT = "create_agent"
+ GEN_AI_OPERATION_TYPE_EXECUTE_AGENT_TASK = "execute_task"
# GenAI Output Types (OTel Semconv)
GEN_AI_OUTPUT_TYPE_IMAGE = "image"
@@ -221,8 +229,12 @@ class SemanticConvetion:
DB_SYSTEM_MILVUS = "milvus"
DB_SYSTEM_ASTRA = "astra"
- # Agents
+ # GenAI Request Attributes (OTel Semconv)
GEN_AI_AGENT_ID = "gen_ai.agent.id"
+ GEN_AI_AGENT_NAME = "gen_ai.agent.name"
+ GEN_AI_AGENT_DESCRIPTION = "gen_ai.agent.description"
+
+
GEN_AI_AGENT_TYPE = "gen_ai.agent.type"
GEN_AI_AGENT_TASK_ID = "gen_ai.agent.task.id"
GEN_AI_AGENT_ROLE = "gen_ai.agent.role"
diff --git a/sdk/python/tests/test_ai21.py b/sdk/python/tests/test_ai21.py
index 32f76071..70fc1b60 100644
--- a/sdk/python/tests/test_ai21.py
+++ b/sdk/python/tests/test_ai21.py
@@ -43,7 +43,7 @@ def test_sync_ai21_chat():
model="jamba-1.5-mini",
max_tokens=1,
)
- assert isinstance(response.id, str)
+ assert isinstance(response['id'], str)
def test_sync_ai21_chat_stream():
"""
@@ -80,7 +80,7 @@ def test_sync_ai21_chat_rag():
max_neighbors=1,
max_tokens=1,
)
- assert isinstance(response.id, str)
+ assert isinstance(response['id'], str)
@pytest.mark.asyncio
async def test_async_ai21_test():
@@ -96,7 +96,7 @@ async def test_async_ai21_test():
model="jamba-1.5-mini",
max_tokens=1,
)
- assert isinstance(response.id, str)
+ assert isinstance(response['id'], str)
responses = await async_client.chat.completions.create(
messages=MESSAGES,
@@ -117,4 +117,4 @@ async def test_async_ai21_test():
max_neighbors=1,
max_tokens=1,
)
- assert isinstance(response.id, str)
+ assert isinstance(response['id'], str)