openlit 1.34.8__py3-none-any.whl → 1.34.11__py3-none-any.whl

openlit/__helpers.py

@@ -333,3 +333,72 @@ def format_and_concatenate(messages):
         for message_data in formatted_messages.values()
         if message_data['content']
     )
+
+def common_span_attributes(scope, gen_ai_operation, gen_ai_system, server_address, server_port,
+    request_model, response_model, environment, application_name, is_stream, tbt, ttft, version):
+    """
+    Set common span attributes for both chat and RAG operations.
+    """
+
+    scope._span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
+    scope._span.set_attribute(SemanticConvention.GEN_AI_OPERATION, gen_ai_operation)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_SYSTEM, gen_ai_system)
+    scope._span.set_attribute(SemanticConvention.SERVER_ADDRESS, server_address)
+    scope._span.set_attribute(SemanticConvention.SERVER_PORT, server_port)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_REQUEST_MODEL, request_model)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_RESPONSE_MODEL, scope._response_model)
+    scope._span.set_attribute(DEPLOYMENT_ENVIRONMENT, environment)
+    scope._span.set_attribute(SERVICE_NAME, application_name)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_REQUEST_IS_STREAM, is_stream)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_SERVER_TBT, scope._tbt)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_SERVER_TTFT, scope._ttft)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_SDK_VERSION, version)
+
+def record_completion_metrics(metrics, gen_ai_operation, gen_ai_system, server_address, server_port,
+    request_model, response_model, environment, application_name, start_time, end_time, cost,
+    input_tokens, output_tokens, tbt, ttft):
+    """
+    Record completion metrics for the operation.
+    """
+
+    attributes = create_metrics_attributes(
+        operation=gen_ai_operation,
+        system=gen_ai_system,
+        server_address=server_address,
+        server_port=server_port,
+        request_model=request_model,
+        response_model=response_model,
+        service_name=application_name,
+        deployment_environment=environment,
+    )
+    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_tbt"].record(tbt, attributes)
+    metrics["genai_server_ttft"].record(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(cost, attributes)
+
+def record_embedding_metrics(metrics, gen_ai_operation, gen_ai_system, server_address, server_port,
+    request_model, response_model, environment, application_name, start_time, end_time,
+    input_tokens, cost):
+    """
+    Record embedding-specific metrics for the operation.
+    """
+
+    attributes = create_metrics_attributes(
+        operation=gen_ai_operation,
+        system=gen_ai_system,
+        server_address=server_address,
+        server_port=server_port,
+        request_model=request_model,
+        response_model=response_model,
+        service_name=application_name,
+        deployment_environment=environment,
+    )
+    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)

openlit/instrumentation/assemblyai/__init__.py

@@ -5,39 +5,35 @@ import importlib.metadata
 from opentelemetry.instrumentation.instrumentor import BaseInstrumentor
 from wrapt import wrap_function_wrapper
 
-from openlit.instrumentation.assemblyai.assemblyai import (
-    transcribe
-)
+from openlit.instrumentation.assemblyai.assemblyai import transcribe
 
-_instruments = ('assemblyai >= 0.35.1',)
+_instruments = ("assemblyai >= 0.35.1",)
 
 class AssemblyAIInstrumentor(BaseInstrumentor):
     """
-    An instrumentor for AssemblyAI's client library.
+    An instrumentor for AssemblyAI client library.
     """
 
     def instrumentation_dependencies(self) -> Collection[str]:
         return _instruments
 
     def _instrument(self, **kwargs):
-        application_name = kwargs.get('application_name', 'default')
-        environment = kwargs.get('environment', 'default')
-        tracer = kwargs.get('tracer')
-        event_provider = kwargs.get('event_provider')
-        metrics = kwargs.get('metrics_dict')
-        pricing_info = kwargs.get('pricing_info', {})
-        capture_message_content = kwargs.get('capture_message_content', False)
-        disable_metrics = kwargs.get('disable_metrics')
-        version = importlib.metadata.version('assemblyai')
+        application_name = kwargs.get("application_name", "default")
+        environment = kwargs.get("environment", "default")
+        tracer = kwargs.get("tracer")
+        metrics = kwargs.get("metrics_dict")
+        pricing_info = kwargs.get("pricing_info", {})
+        capture_message_content = kwargs.get("capture_message_content", False)
+        disable_metrics = kwargs.get("disable_metrics")
+        version = importlib.metadata.version("assemblyai")
 
         # sync transcribe
         wrap_function_wrapper(
-            'assemblyai.transcriber',
-            'Transcriber.transcribe',
+            "assemblyai.transcriber",
+            "Transcriber.transcribe",
             transcribe(version, environment, application_name,
-                  tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics),
+                  tracer, pricing_info, capture_message_content, metrics, disable_metrics),
         )
 
     def _uninstrument(self, **kwargs):
-        # Proper uninstrumentation logic to revert patched methods
         pass

openlit/instrumentation/assemblyai/assemblyai.py

@@ -1,150 +1,59 @@
 """
-Module for monitoring Assembly AI API calls.
+Module for monitoring AssemblyAI 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 (
-    get_audio_model_cost,
-    handle_exception,
-    create_metrics_attributes,
-    set_server_address_and_port,
-    otel_event
-)
+from opentelemetry.trace import SpanKind
+from openlit.__helpers import handle_exception, set_server_address_and_port
+from openlit.instrumentation.assemblyai.utils import process_audio_response
 from openlit.semcov import SemanticConvention
 
 # Initialize logger for logging potential issues and operations
 logger = logging.getLogger(__name__)
 
 def transcribe(version, environment, application_name,
-                 tracer, event_provider, pricing_info, capture_message_content, metrics, disable_metrics):
+                 tracer, pricing_info, capture_message_content, metrics, disable_metrics):
     """
-    Generates a telemetry wrapper for GenAI function call
+    Generates a telemetry wrapper for AssemblyAI transcribe function call
     """
 
-
     def wrapper(wrapped, instance, args, kwargs):
         """
-        Wraps the GenAI function call.
+        Wraps the AssemblyAI transcribe function call.
         """
 
-        server_address, server_port = set_server_address_and_port(instance, 'api.assemblyai.com', 443)
-        request_model = kwargs.get('speech_model', 'best')
+        server_address, server_port = set_server_address_and_port(instance, "api.assemblyai.com", 443)
+        request_model = kwargs.get("speech_model", "best")
 
-        span_name = f'{SemanticConvention.GEN_AI_OPERATION_TYPE_AUDIO} {request_model}'
+        span_name = f"{SemanticConvention.GEN_AI_OPERATION_TYPE_AUDIO} {request_model}"
 
-        with tracer.start_as_current_span(span_name, 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()
 
             try:
-                # Calculate cost of the operation
-                cost = get_audio_model_cost(request_model,
-                                            pricing_info, None, response.audio_duration)
-
-                # Set Span attributes (OTel Semconv)
-                span.set_attribute(TELEMETRY_SDK_NAME, 'openlit')
-                span.set_attribute(SemanticConvention.GEN_AI_OPERATION,
-                                    SemanticConvention.GEN_AI_OPERATION_TYPE_AUDIO)
-                span.set_attribute(SemanticConvention.GEN_AI_SYSTEM,
-                                    SemanticConvention.GEN_AI_SYSTEM_ASSEMBLYAI)
-                span.set_attribute(SemanticConvention.GEN_AI_REQUEST_MODEL,
-                                    request_model)
-                span.set_attribute(SemanticConvention.SERVER_ADDRESS,
-                                    server_address)
-                span.set_attribute(SemanticConvention.SERVER_PORT,
-                                    server_port)
-                span.set_attribute(SemanticConvention.GEN_AI_RESPONSE_MODEL,
-                                    request_model)
-                span.set_attribute(SemanticConvention.GEN_AI_OUTPUT_TYPE,
-                                    'text')
-
-                # Set Span attributes (Extras)
-                span.set_attribute(DEPLOYMENT_ENVIRONMENT,
-                                    environment)
-                span.set_attribute(SERVICE_NAME,
-                                    application_name)
-                span.set_attribute(SemanticConvention.GEN_AI_REQUEST_AUDIO_DURATION,
-                                    response.audio_duration)
-                span.set_attribute(SemanticConvention.GEN_AI_USAGE_COST,
-                                    cost)
-                span.set_attribute(SemanticConvention.GEN_AI_SDK_VERSION,
-                                    version)
-
-                # To be removed one the change to log events (from span events) is complete
-                if capture_message_content:
-                    span.add_event(
-                        name=SemanticConvention.GEN_AI_CONTENT_PROMPT_EVENT,
-                        attributes={
-                            SemanticConvention.GEN_AI_CONTENT_PROMPT: response.audio_url,
-                        },
-                    )
-                    span.add_event(
-                        name=SemanticConvention.GEN_AI_CONTENT_COMPLETION_EVENT,
-                        attributes={
-                            SemanticConvention.GEN_AI_CONTENT_COMPLETION: response.text,
-                        },
-                    )
-
-                input_event = otel_event(
-                    name=SemanticConvention.GEN_AI_USER_MESSAGE,
-                    attributes={
-                        SemanticConvention.GEN_AI_SYSTEM: SemanticConvention.GEN_AI_SYSTEM_ASSEMBLYAI
-                    },
-                    body={
-                        **({'content': response.audio_url} if capture_message_content else {}),
-                        'role': 'user'
-                    }
+                response = wrapped(*args, **kwargs)
+
+                response = process_audio_response(
+                    response=response,
+                    gen_ai_endpoint="assemblyai.transcribe",
+                    pricing_info=pricing_info,
+                    server_port=server_port,
+                    server_address=server_address,
+                    environment=environment,
+                    application_name=application_name,
+                    metrics=metrics,
+                    start_time=start_time,
+                    span=span,
+                    capture_message_content=capture_message_content,
+                    disable_metrics=disable_metrics,
+                    version=version,
+                    **kwargs
                 )
-                event_provider.emit(input_event)
-
-                output_event = otel_event(
-                    name=SemanticConvention.GEN_AI_CHOICE,
-                    attributes={
-                        SemanticConvention.GEN_AI_SYSTEM: SemanticConvention.GEN_AI_SYSTEM_ASSEMBLYAI
-                    },
-                    body={
-                        'finish_reason': 'stop',
-                        'index': 0,
-                        'message': {
-                            **({'content': response.text} if capture_message_content else {}),
-                            'role': 'assistant'
-                        }
-                    }
-                )
-                event_provider.emit(output_event)
-
-                span.set_status(Status(StatusCode.OK))
-
-                if disable_metrics is False:
-                    attributes = create_metrics_attributes(
-                        service_name=application_name,
-                        deployment_environment=environment,
-                        operation=SemanticConvention.GEN_AI_OPERATION_TYPE_AUDIO,
-                        system=SemanticConvention.GEN_AI_SYSTEM_ASSEMBLYAI,
-                        request_model=request_model,
-                        server_address=server_address,
-                        server_port=server_port,
-                        response_model=request_model,
-                    )
-
-                    metrics['genai_client_operation_duration'].record(
-                        end_time - start_time, attributes
-                    )
-                    metrics['genai_requests'].add(1, 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
+            return response
 
     return wrapper

openlit/instrumentation/assemblyai/utils.py

@@ -0,0 +1,142 @@
+"""
+AssemblyAI 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 (
+    get_audio_model_cost,
+    create_metrics_attributes,
+)
+from openlit.semcov import SemanticConvention
+
+def format_audio_url(audio_url):
+    """
+    Process audio URL input to extract content.
+    """
+    return str(audio_url) if audio_url else ""
+
+def common_span_attributes(scope, gen_ai_operation, gen_ai_system, server_address, server_port,
+    request_model, response_model, environment, application_name, is_stream, tbt, ttft, version):
+    """
+    Set common span attributes for both chat and RAG operations.
+    """
+
+    scope._span.set_attribute(TELEMETRY_SDK_NAME, "openlit")
+    scope._span.set_attribute(SemanticConvention.GEN_AI_OPERATION, gen_ai_operation)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_SYSTEM, gen_ai_system)
+    scope._span.set_attribute(SemanticConvention.SERVER_ADDRESS, server_address)
+    scope._span.set_attribute(SemanticConvention.SERVER_PORT, server_port)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_REQUEST_MODEL, request_model)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_RESPONSE_MODEL, scope._response_model)
+    scope._span.set_attribute(DEPLOYMENT_ENVIRONMENT, environment)
+    scope._span.set_attribute(SERVICE_NAME, application_name)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_REQUEST_IS_STREAM, is_stream)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_SERVER_TBT, scope._tbt)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_SERVER_TTFT, scope._ttft)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_SDK_VERSION, version)
+
+def record_audio_metrics(metrics, gen_ai_operation, gen_ai_system, server_address, server_port,
+    request_model, response_model, environment, application_name, start_time, end_time, cost):
+    """
+    Record audio metrics for the operation.
+    """
+
+    attributes = create_metrics_attributes(
+        operation=gen_ai_operation,
+        system=gen_ai_system,
+        server_address=server_address,
+        server_port=server_port,
+        request_model=request_model,
+        response_model=response_model,
+        service_name=application_name,
+        deployment_environment=environment,
+    )
+    metrics["genai_client_operation_duration"].record(end_time - start_time, attributes)
+    metrics["genai_requests"].add(1, attributes)
+    metrics["genai_cost"].record(cost, attributes)
+
+def common_audio_logic(scope, gen_ai_endpoint, pricing_info, environment, application_name,
+    metrics, capture_message_content, disable_metrics, version):
+    """
+    Process audio transcription request and generate Telemetry
+    """
+
+    prompt = scope._response.audio_url
+    request_model = scope._kwargs.get("speech_model", "best")
+    is_stream = False
+
+    # Calculate cost based on audio duration
+    cost = get_audio_model_cost(request_model, pricing_info, prompt, scope._response.audio_duration)
+
+    # Common Span Attributes
+    common_span_attributes(scope,
+        SemanticConvention.GEN_AI_OPERATION_TYPE_AUDIO, SemanticConvention.GEN_AI_SYSTEM_ASSEMBLYAI,
+        scope._server_address, scope._server_port, request_model, request_model,
+        environment, application_name, is_stream, scope._tbt, scope._ttft, version)
+
+    # Span Attributes for Response parameters
+    scope._span.set_attribute(SemanticConvention.GEN_AI_OUTPUT_TYPE, "text")
+    scope._span.set_attribute(SemanticConvention.GEN_AI_RESPONSE_ID, scope._response.id)
+
+    # Span Attributes for Cost
+    scope._span.set_attribute(SemanticConvention.GEN_AI_USAGE_COST, cost)
+
+    # Audio-specific span attributes
+    scope._span.set_attribute(SemanticConvention.GEN_AI_REQUEST_AUDIO_DURATION, scope._response.audio_duration)
+
+    # Span Attributes for Content
+    if capture_message_content:
+        scope._span.set_attribute(SemanticConvention.GEN_AI_CONTENT_PROMPT, prompt)
+        scope._span.set_attribute(SemanticConvention.GEN_AI_CONTENT_COMPLETION, scope._response.text)
+
+        # To be removed once the change to span_attributes (from span events) is complete
+        scope._span.add_event(
+            name=SemanticConvention.GEN_AI_CONTENT_PROMPT_EVENT,
+            attributes={
+                SemanticConvention.GEN_AI_CONTENT_PROMPT: prompt,
+            },
+        )
+        scope._span.add_event(
+            name=SemanticConvention.GEN_AI_CONTENT_COMPLETION_EVENT,
+            attributes={
+                SemanticConvention.GEN_AI_CONTENT_COMPLETION: scope._response.text,
+            },
+        )
+
+    scope._span.set_status(Status(StatusCode.OK))
+
+    # Metrics
+    if not disable_metrics:
+        record_audio_metrics(metrics, SemanticConvention.GEN_AI_OPERATION_TYPE_AUDIO,
+            SemanticConvention.GEN_AI_SYSTEM_ASSEMBLYAI, scope._server_address, scope._server_port,
+            request_model, request_model, environment, application_name, scope._start_time,
+            scope._end_time, cost)
+
+def process_audio_response(response, gen_ai_endpoint, pricing_info, server_port, server_address,
+    environment, application_name, metrics, start_time, span, capture_message_content=False,
+    disable_metrics=False, version="1.0.0", **kwargs):
+    """
+    Process audio transcription request and generate Telemetry
+    """
+
+    scope = type("GenericScope", (), {})()
+
+    scope._start_time = start_time
+    scope._end_time = time.time()
+    scope._span = span
+    scope._server_address, scope._server_port = server_address, server_port
+    scope._kwargs = kwargs
+    scope._response = response
+
+    # Initialize streaming and timing values for AssemblyAI transcription
+    scope._response_model = kwargs.get("speech_model", "best")
+    scope._tbt = 0.0
+    scope._ttft = scope._end_time - scope._start_time
+
+    common_audio_logic(scope, gen_ai_endpoint, pricing_info, environment, application_name,
+        metrics, capture_message_content, disable_metrics, version)
+
+    return response

openlit/instrumentation/elevenlabs/__init__.py

@@ -1,4 +1,3 @@
-# pylint: disable=useless-return, bad-staticmethod-argument, disable=duplicate-code
 """Initializer of Auto Instrumentation of ElevenLabs Functions"""
 
 from typing import Collection
@@ -6,18 +5,14 @@ import importlib.metadata
 from opentelemetry.instrumentation.instrumentor import BaseInstrumentor
 from wrapt import wrap_function_wrapper
 
-from openlit.instrumentation.elevenlabs.elevenlabs import (
-    generate
-)
-from openlit.instrumentation.elevenlabs.async_elevenlabs import (
-    async_generate
-)
+from openlit.instrumentation.elevenlabs.elevenlabs import generate
+from openlit.instrumentation.elevenlabs.async_elevenlabs import async_generate
 
 _instruments = ("elevenlabs >= 1.4.0",)
 
 class ElevenLabsInstrumentor(BaseInstrumentor):
     """
-    An instrumentor for ElevenLabs's client library.
+    An instrumentor for ElevenLabs client library.
     """
 
     def instrumentation_dependencies(self) -> Collection[str]:
@@ -33,14 +28,6 @@ class ElevenLabsInstrumentor(BaseInstrumentor):
         disable_metrics = kwargs.get("disable_metrics")
         version = importlib.metadata.version("elevenlabs")
 
-        # sync generate
-        wrap_function_wrapper(
-            "elevenlabs.client",
-            "ElevenLabs.generate",
-            generate("elevenlabs.generate", version, environment, application_name,
-                  tracer, pricing_info, capture_message_content, metrics, disable_metrics),
-        )
-
         # sync text_to_speech.convert
         wrap_function_wrapper(
             "elevenlabs.text_to_speech.client",
@@ -49,22 +36,13 @@ class ElevenLabsInstrumentor(BaseInstrumentor):
                   tracer, pricing_info, capture_message_content, metrics, disable_metrics),
         )
 
-        # async generate
-        wrap_function_wrapper(
-            "elevenlabs.client",
-            "AsyncElevenLabs.generate",
-            async_generate("elevenlabs.generate", version, environment, application_name,
-                  tracer, pricing_info, capture_message_content, metrics, disable_metrics),
-        )
-
-        # sync text_to_speech.convert
+        # async text_to_speech.convert
         wrap_function_wrapper(
             "elevenlabs.text_to_speech.client",
             "AsyncTextToSpeechClient.convert",
-            generate("elevenlabs.text_to_speech", version, environment, application_name,
+            async_generate("elevenlabs.text_to_speech", version, environment, application_name,
                   tracer, pricing_info, capture_message_content, metrics, disable_metrics),
         )
 
     def _uninstrument(self, **kwargs):
-        # Proper uninstrumentation logic to revert patched methods
         pass