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]: Update GPT4All Instrumentation to align with OTel Semconv (#622)

Patcher2589d0d1051a
diff --git a/sdk/python/pyproject.toml b/sdk/python/pyproject.toml
index dde0383f..eea7ee41 100644
--- a/sdk/python/pyproject.toml
+++ b/sdk/python/pyproject.toml
@@ -1,6 +1,6 @@
[tool.poetry]
name = "openlit"
-version = "1.33.9"
+version = "1.33.10"
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/instrumentation/azure_ai_inference/async_azure_ai_inference.py b/sdk/python/src/openlit/instrumentation/azure_ai_inference/async_azure_ai_inference.py
index b19578da..dbe787e3 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
@@ -191,7 +191,7 @@ def async_complete(version, environment, application_name,
output_tokens)
self._span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
self._server_address)
- self._span.set_attribute(SemanticConvetion.GEN_AI_OPENAI_RESPONSE_SYSTEM_FINGERPRINT,
+ self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SYSTEM_FINGERPRINT,
self._system_fingerprint)
if isinstance(self._llmresponse, str):
self._span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
@@ -370,7 +370,7 @@ def async_complete(version, environment, application_name,
output_tokens)
span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
server_address)
- span.set_attribute(SemanticConvetion.GEN_AI_OPENAI_RESPONSE_SYSTEM_FINGERPRINT,
+ span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SYSTEM_FINGERPRINT,
response_dict.get('system_fingerprint'))
# Set base span attribues (Extras)
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 f5c56d1f..6e81c746 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
@@ -191,7 +191,7 @@ def complete(version, environment, application_name,
output_tokens)
self._span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
self._server_address)
- self._span.set_attribute(SemanticConvetion.GEN_AI_OPENAI_RESPONSE_SYSTEM_FINGERPRINT,
+ self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SYSTEM_FINGERPRINT,
self._system_fingerprint)
if isinstance(self._llmresponse, str):
self._span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
@@ -370,7 +370,7 @@ def complete(version, environment, application_name,
output_tokens)
span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
server_address)
- span.set_attribute(SemanticConvetion.GEN_AI_OPENAI_RESPONSE_SYSTEM_FINGERPRINT,
+ span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SYSTEM_FINGERPRINT,
response_dict.get('system_fingerprint'))
# Set base span attribues (Extras)
diff --git a/sdk/python/src/openlit/instrumentation/gpt4all/__init__.py b/sdk/python/src/openlit/instrumentation/gpt4all/__init__.py
index 54625787..4b4ba150 100644
--- a/sdk/python/src/openlit/instrumentation/gpt4all/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/gpt4all/__init__.py
@@ -34,7 +34,7 @@ class GPT4AllInstrumentor(BaseInstrumentor):
wrap_function_wrapper(
"gpt4all",
"GPT4All.generate",
- generate("gpt4all.generate", version, environment, application_name,
+ generate(version, environment, application_name,
tracer, pricing_info, trace_content, metrics, disable_metrics),
)
@@ -42,7 +42,7 @@ class GPT4AllInstrumentor(BaseInstrumentor):
wrap_function_wrapper(
"gpt4all",
"Embed4All.embed",
- embed("gpt4all.embed", version, environment, application_name,
+ embed(version, environment, application_name,
tracer, pricing_info, trace_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 5101bca5..f728ba34 100644
--- a/sdk/python/src/openlit/instrumentation/gpt4all/gpt4all.py
+++ b/sdk/python/src/openlit/instrumentation/gpt4all/gpt4all.py
@@ -1,207 +1,324 @@
-# pylint: disable=duplicate-code, broad-exception-caught, too-many-statements, unused-argument, possibly-used-before-assignment
"""
Module for monitoring GPT4All API calls.
"""
import logging
+import time
from opentelemetry.trace import SpanKind, Status, StatusCode
-from opentelemetry.sdk.resources import TELEMETRY_SDK_NAME
-from openlit.__helpers import handle_exception, general_tokens
+from opentelemetry.sdk.resources import SERVICE_NAME, TELEMETRY_SDK_NAME, DEPLOYMENT_ENVIRONMENT
+from openlit.__helpers import (
+ handle_exception,
+ general_tokens,
+ create_metrics_attributes,
+ set_server_address_and_port,
+ calculate_tbt,
+ calculate_ttft
+)
from openlit.semcov import SemanticConvetion
# Initialize logger for logging potential issues and operations
logger = logging.getLogger(__name__)
-def generate(gen_ai_endpoint, version, environment, application_name,
+def generate(version, environment, application_name,
tracer, pricing_info, trace_content, metrics, disable_metrics):
"""
- Generates a telemetry wrapper for generate to collect metrics.
+ Generates a telemetry wrapper for chat completions to collect metrics.
Args:
- gen_ai_endpoint: Endpoint identifier for logging and tracing.
version: Version of the monitoring package.
environment: Deployment environment (e.g., production, staging).
application_name: Name of the application using the GPT4All API.
tracer: OpenTelemetry tracer for creating spans.
- pricing_info: Information used for calculating the cost of GPT4All usage.
+ pricing_info: Information used for calculating GPT4All usage.
trace_content: Flag indicating whether to trace the actual content.
Returns:
- A function that wraps the generate method to add telemetry.
+ A function that wraps the chat completions method to add telemetry.
"""
+ class TracedSyncStream:
+ """
+ Wrapper for streaming responses to collect metrics and trace data.
+ Wraps the 'gpt4all.AsyncStream' 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.
+ """
+ def __init__(
+ self,
+ wrapped,
+ span,
+ kwargs,
+ server_address,
+ server_port,
+ request_model,
+ **args,
+ ):
+ self.__wrapped__ = wrapped
+ self._span = span
+ # Placeholder for aggregating streaming response
+ self._llmresponse = ""
+
+ self._args = args
+ self._kwargs = kwargs
+ self._start_time = time.time()
+ self._end_time = None
+ self._timestamps = []
+ self._ttft = 0
+ self._tbt = 0
+ self._server_address = server_address
+ self._server_port = server_port
+ self._request_model = request_model
+
+ def __enter__(self):
+ self.__wrapped__.__enter__()
+ return self
+
+ def __exit__(self, exc_type, exc_value, traceback):
+ self.__wrapped__.__exit__(exc_type, exc_value, traceback)
+
+ def __iter__(self):
+ return self
+
+ def __getattr__(self, name):
+ """Delegate attribute access to the wrapped object."""
+ return getattr(self.__wrapped__, 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)
+
+ self._llmresponse += chunk
+ return chunk
+ except StopIteration:
+ # Handling exception ensure LLM observability without disrupting operation
+ try:
+ self._end_time = time.time()
+
+ if len(self._timestamps) > 1:
+ self._tbt = calculate_tbt(self._timestamps)
+
+ prompt = self._kwargs.get("prompt") or self._args[0] or ""
+
+ # Calculate tokens using input prompt and aggregated response
+ input_tokens = general_tokens(prompt)
+ output_tokens = general_tokens(self._llmresponse)
+
+ # 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_GPT4ALL)
+ self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
+ self._request_model)
+ self._span.set_attribute(SemanticConvetion.SERVER_PORT,
+ self._server_port)
+ self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
+ self._kwargs.get("repeat_penalty", 1.18))
+ self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
+ self._kwargs.get("max_tokens", 200))
+ 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_TEMPERATURE,
+ self._kwargs.get("temp", 0.7))
+ self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
+ self._kwargs.get("top_p", 0.4))
+ self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_K,
+ self._kwargs.get("top_k", 40))
+ self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_MODEL,
+ self._request_model)
+ self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
+ input_tokens)
+ self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
+ 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,
+ input_tokens + output_tokens)
+ 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)
+ self._span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
+ 0)
+ 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_GPT4ALL,
+ request_model=self._request_model,
+ server_address=self._server_address,
+ server_port=self._server_port,
+ response_model=self._request_model,
+ )
+
+ metrics["genai_client_usage_tokens"].record(
+ input_tokens + 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(output_tokens, attributes)
+ metrics["genai_prompt_tokens"].add(input_tokens, attributes)
+ metrics["genai_cost"].record(0, 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 'generate' API call to add telemetry.
-
- This collects metrics such as execution time, cost, and token usage, and handles errors
+ Wraps the 'chat.completions' API call to add telemetry.
+
+ This collects metrics such as execution time, and token usage, and handles errors
gracefully, adding details to the trace for observability.
Args:
- wrapped: The original 'generate' method to be wrapped.
+ 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 'generate' method.
- kwargs: Keyword arguments for the 'generate' method.
+ args: Positional arguments for the 'chat.completions' method.
+ kwargs: Keyword arguments for the 'chat.completions' method.
Returns:
- The response from the original 'generate' method.
+ The response from the original 'chat.completions' method.
"""
# Check if streaming is enabled for the API call
streaming = kwargs.get("streaming", False)
+ server_address, server_port = set_server_address_and_port(instance, "localhost", 80)
+ request_model = str(instance.model.model_path).rsplit('/', maxsplit=1)[-1] or "orca-mini-3b-gguf2-q4_0.gguf"
+
+ span_name = f"{SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT} {request_model}"
+
# pylint: disable=no-else-return
if streaming:
# Special handling for streaming response to accommodate the nature of data flow
- def stream_generator():
- with tracer.start_as_current_span(gen_ai_endpoint, kind= SpanKind.CLIENT) as span:
- # Placeholder for aggregating streaming response
- llmresponse = ""
-
- # Loop through streaming events capturing relevant details
- for chunk in wrapped(*args, **kwargs):
- # Collect aggregated response from events
- llmresponse += chunk
-
- yield chunk
-
- # Handling exception ensure observability without disrupting operation
- try:
- # Calculate cost of the operation
- cost = 0
-
- # pylint: disable=line-too-long
- model = str(instance.model.model_path).rsplit('/', maxsplit=1)[-1] or "orca-mini-3b-gguf2-q4_0.gguf"
- prompt = kwargs.get("prompt") or args[0] or ""
-
- # Calculate cost of the operation
- cost = 0
- prompt_tokens = general_tokens(prompt)
- completion_tokens = general_tokens(llmresponse)
- total_tokens = prompt_tokens + completion_tokens
-
- # Set base span attribues
- span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
- span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_GPT4ALL)
- span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT)
- span.set_attribute(SemanticConvetion.GEN_AI_ENDPOINT,
- gen_ai_endpoint)
- span.set_attribute(SemanticConvetion.GEN_AI_ENVIRONMENT,
- environment)
- span.set_attribute(SemanticConvetion.GEN_AI_APPLICATION_NAME,
- application_name)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- model)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_K,
- kwargs.get("top_k", 40))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
- kwargs.get("top_p", 0.4))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
- kwargs.get("max_tokens", 200))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
- kwargs.get("temperature", 0.7))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
- kwargs.get("frequency_penalty", 1.18))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
- True)
- 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={
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: llmresponse,
- },
- )
-
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- prompt_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
- completion_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- total_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
-
- span.set_status(Status(StatusCode.OK))
-
- if disable_metrics is False:
- attributes = {
- TELEMETRY_SDK_NAME:
- "openlit",
- SemanticConvetion.GEN_AI_APPLICATION_NAME:
- application_name,
- SemanticConvetion.GEN_AI_SYSTEM:
- SemanticConvetion.GEN_AI_SYSTEM_GPT4ALL,
- SemanticConvetion.GEN_AI_ENVIRONMENT:
- environment,
- SemanticConvetion.GEN_AI_OPERATION:
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
- SemanticConvetion.GEN_AI_REQUEST_MODEL:
- model
- }
-
- metrics["genai_requests"].add(1, attributes)
- metrics["genai_total_tokens"].add(total_tokens, attributes)
- metrics["genai_completion_tokens"].add(completion_tokens, attributes)
- metrics["genai_prompt_tokens"].add(prompt_tokens, attributes)
- metrics["genai_cost"].record(cost, attributes)
-
- except Exception as e:
- handle_exception(span, e)
- logger.error("Error in trace creation: %s", e)
-
- return stream_generator()
+ awaited_wrapped = wrapped(*args, **kwargs)
+ span = tracer.start_span(span_name, kind=SpanKind.CLIENT)
+
+ return TracedSyncStream(awaited_wrapped, span, kwargs, server_address, server_port, request_model)
# Handling for non-streaming responses
else:
- # pylint: disable=line-too-long
- with tracer.start_as_current_span(gen_ai_endpoint, kind= SpanKind.CLIENT) as span:
+ 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()
- # pylint: disable=line-too-long
- model = str(instance.model.model_path).rsplit('/', maxsplit=1)[-1] or "orca-mini-3b-gguf2-q4_0.gguf"
- prompt = kwargs.get("prompt") or args[0] or ""
+ try:
+ prompt = kwargs.get("prompt") or args[0] or ""
- # Calculate cost of the operation
- cost = 0
- prompt_tokens = general_tokens(prompt)
- completion_tokens = general_tokens(response)
- total_tokens = prompt_tokens + completion_tokens
+ # Calculate tokens using input prompt and aggregated response
+ input_tokens = general_tokens(str(prompt))
+ output_tokens = general_tokens(str(response))
- try:
- # Set base span attribues
+ # Set Span attributes (OTel Semconv)
span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
- span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_GPT4ALL)
span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT)
- span.set_attribute(SemanticConvetion.GEN_AI_ENDPOINT,
- gen_ai_endpoint)
- span.set_attribute(SemanticConvetion.GEN_AI_ENVIRONMENT,
- environment)
- span.set_attribute(SemanticConvetion.GEN_AI_APPLICATION_NAME,
- application_name)
+ span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
+ SemanticConvetion.GEN_AI_SYSTEM_GPT4ALL)
span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- model)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_K,
- kwargs.get("top_k", 40))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
- kwargs.get("top_p", 0.4))
+ request_model)
+ span.set_attribute(SemanticConvetion.SERVER_PORT,
+ server_port)
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
+ kwargs.get("repeat_penalty", 1.18))
span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
kwargs.get("max_tokens", 200))
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_PRESENCE_PENALTY,
+ kwargs.get("presence_penalty", 0.0))
span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
- kwargs.get("temperature", 0.7))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
- kwargs.get("frequency_penalty", 1.18))
+ kwargs.get("temp", 0.7))
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
+ kwargs.get("top_p", 0.4))
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_K,
+ kwargs.get("top_k", 40))
+ 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)
+ if isinstance(response, str):
+ span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
+ "text")
+ else:
+ span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
+ "json")
+
+ # Set Span attributes (Extra)
+ 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_SERVER_TTFT,
+ end_time - start_time)
+ span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
+ version)
+ span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
+ 0)
if trace_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
@@ -216,38 +333,33 @@ def generate(gen_ai_endpoint, version, environment, application_name,
},
)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- prompt_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
- completion_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- total_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
-
span.set_status(Status(StatusCode.OK))
if disable_metrics is False:
- attributes = {
- TELEMETRY_SDK_NAME:
- "openlit",
- SemanticConvetion.GEN_AI_APPLICATION_NAME:
- application_name,
- SemanticConvetion.GEN_AI_SYSTEM:
- SemanticConvetion.GEN_AI_SYSTEM_GPT4ALL,
- SemanticConvetion.GEN_AI_ENVIRONMENT:
- environment,
- SemanticConvetion.GEN_AI_OPERATION:
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
- SemanticConvetion.GEN_AI_REQUEST_MODEL:
- model
- }
+ attributes = create_metrics_attributes(
+ service_name=application_name,
+ deployment_environment=environment,
+ operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
+ system=SemanticConvetion.GEN_AI_SYSTEM_GPT4ALL,
+ 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_total_tokens"].add(total_tokens, attributes)
- metrics["genai_completion_tokens"].add(completion_tokens, attributes)
- metrics["genai_prompt_tokens"].add(prompt_tokens, attributes)
- metrics["genai_cost"].record(cost, attributes)
+ metrics["genai_completion_tokens"].add(output_tokens, attributes)
+ metrics["genai_prompt_tokens"].add(input_tokens, attributes)
+ metrics["genai_cost"].record(0, attributes)
# Return original response
return response
@@ -261,20 +373,19 @@ def generate(gen_ai_endpoint, version, environment, application_name,
return wrapper
-def embed(gen_ai_endpoint, version, environment, application_name,
- tracer, pricing_info, trace_content, metrics, disable_metrics):
+def embed(version, environment, application_name,
+ tracer, pricing_info, trace_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for embeddings to collect metrics.
Args:
- gen_ai_endpoint: Endpoint identifier for logging and tracing.
version: Version of the monitoring package.
environment: Deployment environment (e.g., production, staging).
application_name: Name of the application using the GPT4All API.
tracer: OpenTelemetry tracer for creating spans.
- pricing_info: Information used for calculating the cost of GPT4All usage.
+ pricing_info: Information used for calculating GPT4All usage.
trace_content: Flag indicating whether to trace the actual content.
-
+
Returns:
A function that wraps the embeddings method to add telemetry.
"""
@@ -283,7 +394,7 @@ def embed(gen_ai_endpoint, version, environment, application_name,
"""
Wraps the 'embeddings' API call to add telemetry.
- This collects metrics such as execution time, cost, and token usage, and handles errors
+ This collects metrics such as execution time, and token usage, and handles errors
gracefully, adding details to the trace for observability.
Args:
@@ -296,68 +407,82 @@ def embed(gen_ai_endpoint, version, environment, application_name,
The response from the original 'embeddings' method.
"""
- with tracer.start_as_current_span(gen_ai_endpoint, kind= SpanKind.CLIENT) as span:
+ server_address, server_port = set_server_address_and_port(instance, "localhost", 80)
+
+ # pylint: disable=line-too-long
+ request_model = str(instance.gpt4all.model.model_path).rsplit('/', maxsplit=1)[-1] or "all-MiniLM-L6-v2.gguf2.f16.gguf"
+
+ span_name = f"{SemanticConvetion.GEN_AI_OPERATION_TYPE_EMBEDDING} {request_model}"
+
+ 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:
- # pylint: disable=line-too-long
- model = str(instance.gpt4all.model.model_path).rsplit('/', maxsplit=1)[-1] or "all-MiniLM-L6-v2.gguf2.f16.gguf"
prompt = kwargs.get("prompt") or args[0] or ""
+ input_tokens = general_tokens(prompt)
- # Calculate cost of the operation
- cost = 0
- prompt_tokens = general_tokens(prompt)
-
- # Set Span attributes
+ # Set Span attributes (OTel Semconv)
span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
- span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_GPT4ALL)
span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
SemanticConvetion.GEN_AI_OPERATION_TYPE_EMBEDDING)
- span.set_attribute(SemanticConvetion.GEN_AI_ENDPOINT,
- gen_ai_endpoint)
- span.set_attribute(SemanticConvetion.GEN_AI_ENVIRONMENT,
- environment)
- span.set_attribute(SemanticConvetion.GEN_AI_APPLICATION_NAME,
- application_name)
+ span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
+ SemanticConvetion.GEN_AI_SYSTEM_GPT4ALL)
span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- 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,
- prompt_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,
- prompt_tokens)
+ input_tokens)
+ span.set_attribute(SemanticConvetion.GEN_AI_SDK_VERSION,
+ version)
span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
+ 0)
+
if trace_content:
span.add_event(
name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
attributes={
- SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: str(kwargs.get("input", "")),
},
)
span.set_status(Status(StatusCode.OK))
if disable_metrics is False:
- attributes = {
- TELEMETRY_SDK_NAME:
- "openlit",
- SemanticConvetion.GEN_AI_APPLICATION_NAME:
- application_name,
- SemanticConvetion.GEN_AI_SYSTEM:
- SemanticConvetion.GEN_AI_SYSTEM_GPT4ALL,
- SemanticConvetion.GEN_AI_ENVIRONMENT:
- environment,
- SemanticConvetion.GEN_AI_OPERATION:
- SemanticConvetion.GEN_AI_OPERATION_TYPE_EMBEDDING,
- SemanticConvetion.GEN_AI_REQUEST_MODEL:
- model
- }
-
+ attributes = create_metrics_attributes(
+ service_name=application_name,
+ deployment_environment=environment,
+ operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_EMBEDDING,
+ system=SemanticConvetion.GEN_AI_SYSTEM_GPT4ALL,
+ 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_total_tokens"].add(prompt_tokens, attributes)
- metrics["genai_prompt_tokens"].add(prompt_tokens, attributes)
- metrics["genai_cost"].record(cost, attributes)
+ metrics["genai_prompt_tokens"].add(input_tokens, attributes)
+ metrics["genai_cost"].record(0, attributes)
+
# Return original response
return response
diff --git a/sdk/python/src/openlit/instrumentation/groq/__init__.py b/sdk/python/src/openlit/instrumentation/groq/__init__.py
index 69148a71..091b154f 100644
--- a/sdk/python/src/openlit/instrumentation/groq/__init__.py
+++ b/sdk/python/src/openlit/instrumentation/groq/__init__.py
@@ -33,7 +33,7 @@ class GroqInstrumentor(BaseInstrumentor):
wrap_function_wrapper(
"groq.resources.chat.completions",
"Completions.create",
- chat("groq.chat.completions", version, environment, application_name,
+ chat(version, environment, application_name,
tracer, pricing_info, trace_content, metrics, disable_metrics),
)
@@ -41,7 +41,7 @@ class GroqInstrumentor(BaseInstrumentor):
wrap_function_wrapper(
"groq.resources.chat.completions",
"AsyncCompletions.create",
- async_chat("groq.chat.completions", version, environment, application_name,
+ async_chat(version, environment, application_name,
tracer, pricing_info, trace_content, metrics, disable_metrics),
)
diff --git a/sdk/python/src/openlit/instrumentation/groq/async_groq.py b/sdk/python/src/openlit/instrumentation/groq/async_groq.py
index 1033d418..473a8cd5 100644
--- a/sdk/python/src/openlit/instrumentation/groq/async_groq.py
+++ b/sdk/python/src/openlit/instrumentation/groq/async_groq.py
@@ -1,24 +1,31 @@
-# pylint: disable=duplicate-code, broad-exception-caught, too-many-statements, unused-argument, used-before-assignment, too-many-branches
"""
Module for monitoring Groq API calls.
"""
import logging
+import time
from opentelemetry.trace import SpanKind, Status, StatusCode
-from opentelemetry.sdk.resources import TELEMETRY_SDK_NAME
-from openlit.__helpers import get_chat_model_cost, handle_exception
+from opentelemetry.sdk.resources import SERVICE_NAME, TELEMETRY_SDK_NAME, DEPLOYMENT_ENVIRONMENT
+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
+)
from openlit.semcov import SemanticConvetion
# Initialize logger for logging potential issues and operations
logger = logging.getLogger(__name__)
-def async_chat(gen_ai_endpoint, version, environment, application_name,
+def async_chat(version, environment, application_name,
tracer, pricing_info, trace_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
Args:
- gen_ai_endpoint: Endpoint identifier for logging and tracing.
version: Version of the monitoring package.
environment: Deployment environment (e.g., production, staging).
application_name: Name of the application using the Groq API.
@@ -30,10 +37,245 @@ def async_chat(gen_ai_endpoint, version, environment, application_name,
A function that wraps the chat completions method to add telemetry.
"""
+ class TracedAsyncStream:
+ """
+ Wrapper for streaming responses to collect metrics and trace data.
+ Wraps the 'Groq.AsyncStream' 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.
+ """
+ def __init__(
+ self,
+ wrapped,
+ span,
+ kwargs,
+ server_address,
+ server_port,
+ **args,
+ ):
+ self.__wrapped__ = wrapped
+ self._span = span
+ # Placeholder for aggregating streaming response
+ self._llmresponse = ""
+ self._response_id = ""
+ self._response_model = ""
+ self._finish_reason = ""
+ self._system_fingerprint = ""
+ self._input_tokens = 0
+ self._output_tokens = 0
+
+ self._args = args
+ self._kwargs = kwargs
+ self._start_time = time.time()
+ self._end_time = None
+ self._timestamps = []
+ self._ttft = 0
+ self._tbt = 0
+ self._server_address = server_address
+ self._server_port = server_port
+
+ async def __aenter__(self):
+ await self.__wrapped__.__aenter__()
+ return self
+
+ async def __aexit__(self, exc_type, exc_value, traceback):
+ await self.__wrapped__.__aexit__(exc_type, exc_value, traceback)
+
+ def __aiter__(self):
+ return self
+
+ async def __getattr__(self, name):
+ """Delegate attribute access to the wrapped object."""
+ return getattr(await self.__wrapped__, 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)
+ # Collect message IDs and aggregated response from events
+ 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._response_model = chunked.get('model')
+ self._finish_reason = chunked.get('choices')[0].get('finish_reason')
+ self._system_fingerprint = chunked.get('system_fingerprint')
+ 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_GROQ)
+ 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_completion_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", 1.0))
+ 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,
+ 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)
+ self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SERVICE_TIER,
+ self._kwargs.get("service_tier", "on_demand"))
+ self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SYSTEM_FINGERPRINT,
+ self._system_fingerprint)
+ 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_USER,
+ self._kwargs.get("user", ""))
+ 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_GROQ,
+ 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
+ )
+ 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.
@@ -49,148 +291,27 @@ def async_chat(gen_ai_endpoint, version, environment, application_name,
# 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.groq.com", 443)
+ request_model = kwargs.get("model", "gpt-4o")
+
+ span_name = f"{SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT} {request_model}"
# pylint: disable=no-else-return
if streaming:
# Special handling for streaming response to accommodate the nature of data flow
- async def stream_generator():
- with tracer.start_as_current_span(gen_ai_endpoint, kind= SpanKind.CLIENT) as span:
- # Placeholder for aggregating streaming response
- llmresponse = ""
-
- # Loop through streaming events capturing relevant details
- async for chunk in await wrapped(*args, **kwargs):
- # Collect message IDs and aggregated response from events
- if len(chunk.choices) > 0:
- # pylint: disable=line-too-long
- if hasattr(chunk.choices[0], "delta") and hasattr(chunk.choices[0].delta, "content"):
- content = chunk.choices[0].delta.content
- if content:
- llmresponse += content
- if chunk.x_groq is not None and chunk.x_groq.usage is not None:
- prompt_tokens = chunk.x_groq.usage.prompt_tokens
- completion_tokens = chunk.x_groq.usage.completion_tokens
- total_tokens = chunk.x_groq.usage.total_tokens
- response_id = chunk.x_groq.id
- yield chunk
-
- # Handling exception ensure observability without disrupting operation
- 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(
- # pylint: disable=line-too-long
- 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)
-
- # Calculate cost of the operation
- cost = get_chat_model_cost(kwargs.get("model", "gpt-3.5-turbo"),
- pricing_info, prompt_tokens,
- completion_tokens)
-
- # Set Span attributes
- span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
- span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_GROQ)
- span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT)
- span.set_attribute(SemanticConvetion.GEN_AI_ENDPOINT,
- gen_ai_endpoint)
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_ID,
- response_id)
- span.set_attribute(SemanticConvetion.GEN_AI_ENVIRONMENT,
- environment)
- span.set_attribute(SemanticConvetion.GEN_AI_APPLICATION_NAME,
- application_name)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- kwargs.get("model", "gpt-3.5-turbo"))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_USER,
- kwargs.get("user", ""))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
- kwargs.get("top_p", 1.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
- kwargs.get("max_tokens", -1))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
- kwargs.get("temperature", 1.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_PRESENCE_PENALTY,
- kwargs.get("presence_penalty", 0.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
- kwargs.get("frequency_penalty", 0.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
- kwargs.get("seed", ""))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
- True)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- prompt_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
- completion_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- prompt_tokens + completion_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
- 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={
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: llmresponse,
- },
- )
-
- span.set_status(Status(StatusCode.OK))
-
- if disable_metrics is False:
- attributes = {
- TELEMETRY_SDK_NAME:
- "openlit",
- SemanticConvetion.GEN_AI_APPLICATION_NAME:
- application_name,
- SemanticConvetion.GEN_AI_SYSTEM:
- SemanticConvetion.GEN_AI_SYSTEM_GROQ,
- SemanticConvetion.GEN_AI_ENVIRONMENT:
- environment,
- SemanticConvetion.GEN_AI_OPERATION:
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
- SemanticConvetion.GEN_AI_REQUEST_MODEL:
- kwargs.get("model", "gpt-3.5-turbo")
- }
-
- metrics["genai_requests"].add(1, attributes)
- metrics["genai_total_tokens"].add(
- total_tokens, attributes
- )
- metrics["genai_completion_tokens"].add(completion_tokens, attributes)
- metrics["genai_prompt_tokens"].add(prompt_tokens, attributes)
- metrics["genai_cost"].record(cost, attributes)
-
- except Exception as e:
- handle_exception(span, e)
- logger.error("Error in trace creation: %s", e)
+ awaited_wrapped = await wrapped(*args, **kwargs)
+ span = tracer.start_span(span_name, kind=SpanKind.CLIENT)
- return stream_generator()
+ return TracedAsyncStream(awaited_wrapped, span, kwargs, server_address, server_port)
# Handling for non-streaming responses
else:
- # pylint: disable=line-too-long
- with tracer.start_as_current_span(gen_ai_endpoint, kind= SpanKind.CLIENT) as span:
+ 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
@@ -202,7 +323,6 @@ def async_chat(gen_ai_endpoint, version, environment, application_name,
if isinstance(content, list):
content_str = ", ".join(
- # pylint: disable=line-too-long
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
@@ -212,38 +332,70 @@ def async_chat(gen_ai_endpoint, version, environment, application_name,
formatted_messages.append(f"{role}: {content}")
prompt = "\n".join(formatted_messages)
- # Set base span attribues
+ 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_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_GROQ)
span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT)
- span.set_attribute(SemanticConvetion.GEN_AI_ENDPOINT,
- gen_ai_endpoint)
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_ID,
- response.x_groq["id"])
- span.set_attribute(SemanticConvetion.GEN_AI_ENVIRONMENT,
- environment)
- span.set_attribute(SemanticConvetion.GEN_AI_APPLICATION_NAME,
- application_name)
+ span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
+ SemanticConvetion.GEN_AI_SYSTEM_GROQ)
span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- kwargs.get("model", "llama3-8b-8192"))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
- kwargs.get("top_p", 1.0))
+ 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_USER,
- kwargs.get("name", ""))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
- kwargs.get("temperature", 1.0))
+ kwargs.get("max_completion_tokens", -1))
span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_PRESENCE_PENALTY,
kwargs.get("presence_penalty", 0.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
- kwargs.get("frequency_penalty", 0.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
- kwargs.get("seed", ""))
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_STOP_SEQUENCES,
+ kwargs.get("stop", []))
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
+ kwargs.get("temperature", 1.0))
+ 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,
+ 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_REQUEST_SERVICE_TIER,
+ kwargs.get("service_tier", "on_demand"))
+ span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SYSTEM_FINGERPRINT,
+ response_dict.get('system_fingerprint'))
+
+ # 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_USER,
+ kwargs.get("user", ""))
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,
@@ -252,90 +404,54 @@ def async_chat(gen_ai_endpoint, version, environment, application_name,
},
)
- # Set span attributes when tools is not passed to the function call
- if "tools" not in kwargs:
- # Calculate cost of the operation
- cost = get_chat_model_cost(kwargs.get("model", "llama3-8b-8192"),
- pricing_info, response.usage.prompt_tokens,
- response.usage.completion_tokens)
-
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- response.usage.prompt_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
- response.usage.completion_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- response.usage.total_tokens)
+ for i in range(kwargs.get('n',1)):
span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
- [response.choices[0].finish_reason])
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
-
- # Set span attributes for when n = 1 (default)
- if "n" not in kwargs or kwargs["n"] == 1:
- if trace_content:
- span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: response.choices[0].message.content,
- },
- )
-
- # Set span attributes for when n > 0
- else:
- i = 0
- while i < kwargs["n"] and trace_content is True:
- attribute_name = f"gen_ai.completion.{i}"
- span.add_event(
- name=attribute_name,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: response.choices[i].message.content,
- },
- )
- i += 1
-
- # Return original response
- return response
-
- # Set span attributes when tools is passed to the function call
- elif "tools" in kwargs:
- # Calculate cost of the operation
- cost = get_chat_model_cost(kwargs.get("model", "gpt-3.5-turbo"),
- pricing_info, response.usage.prompt_tokens,
- response.usage.completion_tokens)
-
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_COMPLETION,
- "Function called with tools")
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- response.usage.prompt_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
- response.usage.completion_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- response.usage.total_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
+ [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 = {
- TELEMETRY_SDK_NAME:
- "openlit",
- SemanticConvetion.GEN_AI_APPLICATION_NAME:
- application_name,
- SemanticConvetion.GEN_AI_SYSTEM:
- SemanticConvetion.GEN_AI_SYSTEM_GROQ,
- SemanticConvetion.GEN_AI_ENVIRONMENT:
- environment,
- SemanticConvetion.GEN_AI_OPERATION:
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
- SemanticConvetion.GEN_AI_REQUEST_MODEL:
- kwargs.get("model", "gpt-3.5-turbo")
- }
+ attributes = create_metrics_attributes(
+ service_name=application_name,
+ deployment_environment=environment,
+ operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
+ system=SemanticConvetion.GEN_AI_SYSTEM_GROQ,
+ 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_total_tokens"].add(response.usage.total_tokens, attributes)
- metrics["genai_completion_tokens"].add(response.usage.completion_tokens, attributes)
- metrics["genai_prompt_tokens"].add(response.usage.prompt_tokens, 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
diff --git a/sdk/python/src/openlit/instrumentation/groq/groq.py b/sdk/python/src/openlit/instrumentation/groq/groq.py
index df89f15d..68bdc8ea 100644
--- a/sdk/python/src/openlit/instrumentation/groq/groq.py
+++ b/sdk/python/src/openlit/instrumentation/groq/groq.py
@@ -1,24 +1,31 @@
-# pylint: disable=duplicate-code, broad-exception-caught, too-many-statements, unused-argument, used-before-assignment, too-many-branches
"""
Module for monitoring Groq API calls.
"""
import logging
+import time
from opentelemetry.trace import SpanKind, Status, StatusCode
-from opentelemetry.sdk.resources import TELEMETRY_SDK_NAME
-from openlit.__helpers import get_chat_model_cost, handle_exception
+from opentelemetry.sdk.resources import SERVICE_NAME, TELEMETRY_SDK_NAME, DEPLOYMENT_ENVIRONMENT
+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
+)
from openlit.semcov import SemanticConvetion
# Initialize logger for logging potential issues and operations
logger = logging.getLogger(__name__)
-def chat(gen_ai_endpoint, version, environment, application_name,
+def chat(version, environment, application_name,
tracer, pricing_info, trace_content, metrics, disable_metrics):
"""
Generates a telemetry wrapper for chat completions to collect metrics.
Args:
- gen_ai_endpoint: Endpoint identifier for logging and tracing.
version: Version of the monitoring package.
environment: Deployment environment (e.g., production, staging).
application_name: Name of the application using the Groq API.
@@ -30,10 +37,245 @@ def chat(gen_ai_endpoint, version, environment, application_name,
A function that wraps the chat completions method to add telemetry.
"""
+ class TracedSyncStream:
+ """
+ Wrapper for streaming responses to collect metrics and trace data.
+ Wraps the 'Groq.AsyncStream' 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.
+ """
+ def __init__(
+ self,
+ wrapped,
+ span,
+ kwargs,
+ server_address,
+ server_port,
+ **args,
+ ):
+ self.__wrapped__ = wrapped
+ self._span = span
+ # Placeholder for aggregating streaming response
+ self._llmresponse = ""
+ self._response_id = ""
+ self._response_model = ""
+ self._finish_reason = ""
+ self._system_fingerprint = ""
+ self._input_tokens = 0
+ self._output_tokens = 0
+
+ self._args = args
+ self._kwargs = kwargs
+ self._start_time = time.time()
+ self._end_time = None
+ self._timestamps = []
+ self._ttft = 0
+ self._tbt = 0
+ self._server_address = server_address
+ self._server_port = server_port
+
+ def __enter__(self):
+ self.__wrapped__.__enter__()
+ return self
+
+ def __exit__(self, exc_type, exc_value, traceback):
+ self.__wrapped__.__exit__(exc_type, exc_value, traceback)
+
+ def __iter__(self):
+ return self
+
+ def __getattr__(self, name):
+ """Delegate attribute access to the wrapped object."""
+ return getattr(self.__wrapped__, 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)
+ # Collect message IDs and aggregated response from events
+ 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._response_model = chunked.get('model')
+ self._finish_reason = chunked.get('choices')[0].get('finish_reason')
+ self._system_fingerprint = chunked.get('system_fingerprint')
+ 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_GROQ)
+ 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_completion_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", 1.0))
+ 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,
+ 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)
+ self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SERVICE_TIER,
+ self._kwargs.get("service_tier", "on_demand"))
+ self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SYSTEM_FINGERPRINT,
+ self._system_fingerprint)
+ 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_USER,
+ self._kwargs.get("user", ""))
+ 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_GROQ,
+ 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
+ )
+ 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.
@@ -49,148 +291,27 @@ def chat(gen_ai_endpoint, version, environment, application_name,
# 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.groq.com", 443)
+ request_model = kwargs.get("model", "gpt-4o")
+
+ span_name = f"{SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT} {request_model}"
# pylint: disable=no-else-return
if streaming:
# Special handling for streaming response to accommodate the nature of data flow
- def stream_generator():
- with tracer.start_as_current_span(gen_ai_endpoint, kind= SpanKind.CLIENT) as span:
- # Placeholder for aggregating streaming response
- llmresponse = ""
-
- # Loop through streaming events capturing relevant details
- for chunk in wrapped(*args, **kwargs):
- # Collect message IDs and aggregated response from events
- if len(chunk.choices) > 0:
- # pylint: disable=line-too-long
- if hasattr(chunk.choices[0], "delta") and hasattr(chunk.choices[0].delta, "content"):
- content = chunk.choices[0].delta.content
- if content:
- llmresponse += content
- if chunk.x_groq is not None and chunk.x_groq.usage is not None:
- prompt_tokens = chunk.x_groq.usage.prompt_tokens
- completion_tokens = chunk.x_groq.usage.completion_tokens
- total_tokens = chunk.x_groq.usage.total_tokens
- response_id = chunk.x_groq.id
- yield chunk
-
- # Handling exception ensure observability without disrupting operation
- 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(
- # pylint: disable=line-too-long
- 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)
-
- # Calculate cost of the operation
- cost = get_chat_model_cost(kwargs.get("model", "gpt-3.5-turbo"),
- pricing_info, prompt_tokens,
- completion_tokens)
-
- # Set Span attributes
- span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
- span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_GROQ)
- span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT)
- span.set_attribute(SemanticConvetion.GEN_AI_ENDPOINT,
- gen_ai_endpoint)
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_ID,
- response_id)
- span.set_attribute(SemanticConvetion.GEN_AI_ENVIRONMENT,
- environment)
- span.set_attribute(SemanticConvetion.GEN_AI_APPLICATION_NAME,
- application_name)
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- kwargs.get("model", "gpt-3.5-turbo"))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_USER,
- kwargs.get("user", ""))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
- kwargs.get("top_p", 1.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
- kwargs.get("max_tokens", -1))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
- kwargs.get("temperature", 1.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_PRESENCE_PENALTY,
- kwargs.get("presence_penalty", 0.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
- kwargs.get("frequency_penalty", 0.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
- kwargs.get("seed", ""))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
- True)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- prompt_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
- completion_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- prompt_tokens + completion_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
- 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={
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: llmresponse,
- },
- )
-
- span.set_status(Status(StatusCode.OK))
-
- if disable_metrics is False:
- attributes = {
- TELEMETRY_SDK_NAME:
- "openlit",
- SemanticConvetion.GEN_AI_APPLICATION_NAME:
- application_name,
- SemanticConvetion.GEN_AI_SYSTEM:
- SemanticConvetion.GEN_AI_SYSTEM_GROQ,
- SemanticConvetion.GEN_AI_ENVIRONMENT:
- environment,
- SemanticConvetion.GEN_AI_OPERATION:
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
- SemanticConvetion.GEN_AI_REQUEST_MODEL:
- kwargs.get("model", "gpt-3.5-turbo")
- }
-
- metrics["genai_requests"].add(1, attributes)
- metrics["genai_total_tokens"].add(
- total_tokens, attributes
- )
- metrics["genai_completion_tokens"].add(completion_tokens, attributes)
- metrics["genai_prompt_tokens"].add(prompt_tokens, attributes)
- metrics["genai_cost"].record(cost, attributes)
-
- except Exception as e:
- handle_exception(span, e)
- logger.error("Error in trace creation: %s", e)
+ awaited_wrapped = wrapped(*args, **kwargs)
+ span = tracer.start_span(span_name, kind=SpanKind.CLIENT)
- return stream_generator()
+ return TracedSyncStream(awaited_wrapped, span, kwargs, server_address, server_port)
# Handling for non-streaming responses
else:
- # pylint: disable=line-too-long
- with tracer.start_as_current_span(gen_ai_endpoint, kind= SpanKind.CLIENT) as span:
+ 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
@@ -202,7 +323,6 @@ def chat(gen_ai_endpoint, version, environment, application_name,
if isinstance(content, list):
content_str = ", ".join(
- # pylint: disable=line-too-long
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
@@ -212,38 +332,70 @@ def chat(gen_ai_endpoint, version, environment, application_name,
formatted_messages.append(f"{role}: {content}")
prompt = "\n".join(formatted_messages)
- # Set base span attribues
+ 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_SYSTEM,
- SemanticConvetion.GEN_AI_SYSTEM_GROQ)
span.set_attribute(SemanticConvetion.GEN_AI_OPERATION,
SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT)
- span.set_attribute(SemanticConvetion.GEN_AI_ENDPOINT,
- gen_ai_endpoint)
- span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_ID,
- response.x_groq["id"])
- span.set_attribute(SemanticConvetion.GEN_AI_ENVIRONMENT,
- environment)
- span.set_attribute(SemanticConvetion.GEN_AI_APPLICATION_NAME,
- application_name)
+ span.set_attribute(SemanticConvetion.GEN_AI_SYSTEM,
+ SemanticConvetion.GEN_AI_SYSTEM_GROQ)
span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
- kwargs.get("model", "llama3-8b-8192"))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
- kwargs.get("top_p", 1.0))
+ 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_USER,
- kwargs.get("name", ""))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
- kwargs.get("temperature", 1.0))
+ kwargs.get("max_completion_tokens", -1))
span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_PRESENCE_PENALTY,
kwargs.get("presence_penalty", 0.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
- kwargs.get("frequency_penalty", 0.0))
- span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
- kwargs.get("seed", ""))
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_STOP_SEQUENCES,
+ kwargs.get("stop", []))
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
+ kwargs.get("temperature", 1.0))
+ 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,
+ 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_REQUEST_SERVICE_TIER,
+ kwargs.get("service_tier", "on_demand"))
+ span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SYSTEM_FINGERPRINT,
+ response_dict.get('system_fingerprint'))
+
+ # 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_USER,
+ kwargs.get("user", ""))
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,
@@ -252,90 +404,54 @@ def chat(gen_ai_endpoint, version, environment, application_name,
},
)
- # Set span attributes when tools is not passed to the function call
- if "tools" not in kwargs:
- # Calculate cost of the operation
- cost = get_chat_model_cost(kwargs.get("model", "llama3-8b-8192"),
- pricing_info, response.usage.prompt_tokens,
- response.usage.completion_tokens)
-
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- response.usage.prompt_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
- response.usage.completion_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- response.usage.total_tokens)
+ for i in range(kwargs.get('n',1)):
span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
- [response.choices[0].finish_reason])
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
-
- # Set span attributes for when n = 1 (default)
- if "n" not in kwargs or kwargs["n"] == 1:
- if trace_content:
- span.add_event(
- name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: response.choices[0].message.content,
- },
- )
-
- # Set span attributes for when n > 0
- else:
- i = 0
- while i < kwargs["n"] and trace_content is True:
- attribute_name = f"gen_ai.completion.{i}"
- span.add_event(
- name=attribute_name,
- attributes={
- SemanticConvetion.GEN_AI_CONTENT_COMPLETION: response.choices[i].message.content,
- },
- )
- i += 1
-
- # Return original response
- return response
-
- # Set span attributes when tools is passed to the function call
- elif "tools" in kwargs:
- # Calculate cost of the operation
- cost = get_chat_model_cost(kwargs.get("model", "gpt-3.5-turbo"),
- pricing_info, response.usage.prompt_tokens,
- response.usage.completion_tokens)
-
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_COMPLETION,
- "Function called with tools")
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_INPUT_TOKENS,
- response.usage.prompt_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_OUTPUT_TOKENS,
- response.usage.completion_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
- response.usage.total_tokens)
- span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
- cost)
+ [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 = {
- TELEMETRY_SDK_NAME:
- "openlit",
- SemanticConvetion.GEN_AI_APPLICATION_NAME:
- application_name,
- SemanticConvetion.GEN_AI_SYSTEM:
- SemanticConvetion.GEN_AI_SYSTEM_GROQ,
- SemanticConvetion.GEN_AI_ENVIRONMENT:
- environment,
- SemanticConvetion.GEN_AI_OPERATION:
- SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
- SemanticConvetion.GEN_AI_REQUEST_MODEL:
- kwargs.get("model", "gpt-3.5-turbo")
- }
+ attributes = create_metrics_attributes(
+ service_name=application_name,
+ deployment_environment=environment,
+ operation=SemanticConvetion.GEN_AI_OPERATION_TYPE_CHAT,
+ system=SemanticConvetion.GEN_AI_SYSTEM_GROQ,
+ 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_total_tokens"].add(response.usage.total_tokens, attributes)
- metrics["genai_completion_tokens"].add(response.usage.completion_tokens, attributes)
- metrics["genai_prompt_tokens"].add(response.usage.prompt_tokens, 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
diff --git a/sdk/python/src/openlit/instrumentation/openai/async_openai.py b/sdk/python/src/openlit/instrumentation/openai/async_openai.py
index 5d1acde7..bd25bdc5 100644
--- a/sdk/python/src/openlit/instrumentation/openai/async_openai.py
+++ b/sdk/python/src/openlit/instrumentation/openai/async_openai.py
@@ -197,11 +197,11 @@ def async_chat_completions(version, environment, application_name,
output_tokens)
self._span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
self._server_address)
- self._span.set_attribute(SemanticConvetion.GEN_AI_OPENAI_REQUEST_SERVICE_TIER,
+ self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SERVICE_TIER,
self._kwargs.get("service_tier", "auto"))
- self._span.set_attribute(SemanticConvetion.GEN_AI_OPENAI_RESPONSE_SERVICE_TIER,
+ self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SERVICE_TIER,
self._openai_response_service_tier)
- self._span.set_attribute(SemanticConvetion.GEN_AI_OPENAI_RESPONSE_SYSTEM_FINGERPRINT,
+ self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SYSTEM_FINGERPRINT,
self._openai_system_fingerprint)
if isinstance(self._llmresponse, str):
self._span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
@@ -382,11 +382,11 @@ def async_chat_completions(version, environment, application_name,
output_tokens)
span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
server_address)
- span.set_attribute(SemanticConvetion.GEN_AI_OPENAI_REQUEST_SERVICE_TIER,
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SERVICE_TIER,
kwargs.get("service_tier", "auto"))
- span.set_attribute(SemanticConvetion.GEN_AI_OPENAI_RESPONSE_SERVICE_TIER,
+ span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SERVICE_TIER,
response_dict.get('service_tier'))
- span.set_attribute(SemanticConvetion.GEN_AI_OPENAI_RESPONSE_SYSTEM_FINGERPRINT,
+ span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SYSTEM_FINGERPRINT,
response_dict.get('system_fingerprint'))
# Set base span attribues (Extras)
diff --git a/sdk/python/src/openlit/instrumentation/openai/openai.py b/sdk/python/src/openlit/instrumentation/openai/openai.py
index 7c39adb2..95378d88 100644
--- a/sdk/python/src/openlit/instrumentation/openai/openai.py
+++ b/sdk/python/src/openlit/instrumentation/openai/openai.py
@@ -197,11 +197,11 @@ def chat_completions(version, environment, application_name,
output_tokens)
self._span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
self._server_address)
- self._span.set_attribute(SemanticConvetion.GEN_AI_OPENAI_REQUEST_SERVICE_TIER,
+ self._span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SERVICE_TIER,
self._kwargs.get("service_tier", "auto"))
- self._span.set_attribute(SemanticConvetion.GEN_AI_OPENAI_RESPONSE_SERVICE_TIER,
+ self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SERVICE_TIER,
self._openai_response_service_tier)
- self._span.set_attribute(SemanticConvetion.GEN_AI_OPENAI_RESPONSE_SYSTEM_FINGERPRINT,
+ self._span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SYSTEM_FINGERPRINT,
self._openai_system_fingerprint)
if isinstance(self._llmresponse, str):
self._span.set_attribute(SemanticConvetion.GEN_AI_OUTPUT_TYPE,
@@ -382,11 +382,11 @@ def chat_completions(version, environment, application_name,
output_tokens)
span.set_attribute(SemanticConvetion.SERVER_ADDRESS,
server_address)
- span.set_attribute(SemanticConvetion.GEN_AI_OPENAI_REQUEST_SERVICE_TIER,
+ span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SERVICE_TIER,
kwargs.get("service_tier", "auto"))
- span.set_attribute(SemanticConvetion.GEN_AI_OPENAI_RESPONSE_SERVICE_TIER,
+ span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SERVICE_TIER,
response_dict.get('service_tier'))
- span.set_attribute(SemanticConvetion.GEN_AI_OPENAI_RESPONSE_SYSTEM_FINGERPRINT,
+ span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_SYSTEM_FINGERPRINT,
response_dict.get('system_fingerprint'))
# Set base span attribues (Extras)
diff --git a/sdk/python/src/openlit/semcov/__init__.py b/sdk/python/src/openlit/semcov/__init__.py
index 616722a5..84abc899 100644
--- a/sdk/python/src/openlit/semcov/__init__.py
+++ b/sdk/python/src/openlit/semcov/__init__.py
@@ -88,9 +88,9 @@ class SemanticConvetion:
GEN_AI_SYSTEM_XAI = "xai"
# GenAI OpenAI Attributes (OTel Semconv)
- GEN_AI_OPENAI_REQUEST_SERVICE_TIER = "gen_ai.openai.request.service_tier"
- GEN_AI_OPENAI_RESPONSE_SERVICE_TIER = "gen_ai.openai.response.service_tier"
- GEN_AI_OPENAI_RESPONSE_SYSTEM_FINGERPRINT = "gen_ai.openai.response.system_fingerprint"
+ GEN_AI_REQUEST_SERVICE_TIER = "gen_ai.request.service_tier"
+ GEN_AI_RESPONSE_SERVICE_TIER = "gen_ai.response.service_tier"
+ GEN_AI_RESPONSE_SYSTEM_FINGERPRINT = "gen_ai.response.system_fingerprint"
# GenAI System Names (Extra)
GEN_AI_SYSTEM_HUGGING_FACE = "huggingface"