openlit 1.34.14__tar.gz → 1.34.15__tar.gz

{openlit-1.34.14 → openlit-1.34.15}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: openlit
-Version: 1.34.14
+Version: 1.34.15
 Summary: OpenTelemetry-native Auto instrumentation library for monitoring LLM Applications and GPUs, facilitating the integration of observability into your GenAI-driven projects
 License: Apache-2.0
 Keywords: OpenTelemetry,otel,otlp,llm,tracing,openai,anthropic,claude,cohere,llm monitoring,observability,monitoring,gpt,Generative AI,chatGPT,gpu

{openlit-1.34.14 → openlit-1.34.15}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "openlit"
-version = "1.34.14"
+version = "1.34.15"
 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"

{openlit-1.34.14 → openlit-1.34.15}/src/openlit/instrumentation/litellm/__init__.py

@@ -1,4 +1,3 @@
-# pylint: disable=useless-return, bad-staticmethod-argument, disable=duplicate-code
 """Initializer of Auto Instrumentation of LiteLLM Functions"""
 
 from typing import Collection
@@ -17,15 +16,15 @@ _instruments = ("litellm >= 1.52.6",)
 
 class LiteLLMInstrumentor(BaseInstrumentor):
     """
-    An instrumentor for LiteLLM's client library.
+    An instrumentor for LiteLLM client library.
     """
 
     def instrumentation_dependencies(self) -> Collection[str]:
         return _instruments
 
     def _instrument(self, **kwargs):
-        application_name = kwargs.get("application_name", "default_application")
-        environment = kwargs.get("environment", "default_environment")
+        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", {})
@@ -33,7 +32,7 @@ class LiteLLMInstrumentor(BaseInstrumentor):
         disable_metrics = kwargs.get("disable_metrics")
         version = importlib.metadata.version("litellm")
 
-        # completion
+        # Chat completions
         wrap_function_wrapper(
             "litellm",
             "completion",
@@ -41,6 +40,7 @@ class LiteLLMInstrumentor(BaseInstrumentor):
                   tracer, pricing_info, capture_message_content, metrics, disable_metrics),
         )
 
+        # Async chat completions
         wrap_function_wrapper(
             "litellm",
             "acompletion",
@@ -48,6 +48,7 @@ class LiteLLMInstrumentor(BaseInstrumentor):
                   tracer, pricing_info, capture_message_content, metrics, disable_metrics),
         )
 
+        # Embeddings
         wrap_function_wrapper(
             "litellm",
             "embedding",
@@ -55,6 +56,7 @@ class LiteLLMInstrumentor(BaseInstrumentor):
                   tracer, pricing_info, capture_message_content, metrics, disable_metrics),
         )
 
+        # Async embeddings
         wrap_function_wrapper(
             "litellm",
             "aembedding",
@@ -63,5 +65,4 @@ class LiteLLMInstrumentor(BaseInstrumentor):
         )
 
     def _uninstrument(self, **kwargs):
-        # Proper uninstrumentation logic to revert patched methods
         pass

openlit-1.34.15/src/openlit/instrumentation/litellm/async_litellm.py

@@ -0,0 +1,188 @@
+"""
+Module for monitoring LiteLLM API calls (async version).
+"""
+
+import time
+from opentelemetry.trace import SpanKind
+from openlit.__helpers import (
+    handle_exception,
+    set_server_address_and_port
+)
+from openlit.instrumentation.litellm.utils import (
+    process_chunk,
+    process_streaming_chat_response,
+    process_chat_response,
+    process_embedding_response
+)
+from openlit.semcov import SemanticConvention
+
+def acompletion(version, environment, application_name, tracer, pricing_info,
+    capture_message_content, metrics, disable_metrics):
+    """
+    Generates a telemetry wrapper for GenAI function call
+    """
+
+    class TracedAsyncStream:
+        """
+        Wrapper for async streaming responses to collect telemetry.
+        """
+
+        def __init__(
+                self,
+                wrapped,
+                span,
+                span_name,
+                kwargs,
+                server_address,
+                server_port,
+                **args,
+            ):
+            self.__wrapped__ = wrapped
+            self._span = span
+            self._span_name = span_name
+            self._llmresponse = ""
+            self._response_id = ""
+            self._response_model = ""
+            self._finish_reason = ""
+            self._response_service_tier = ""
+            self._tools = None
+            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
+
+        def __getattr__(self, name):
+            """Delegate attribute access to the wrapped object."""
+            return getattr(self.__wrapped__, name)
+
+        async def __anext__(self):
+            try:
+                chunk = await self.__wrapped__.__anext__()
+                process_chunk(self, chunk)
+                return chunk
+            except StopAsyncIteration:
+                try:
+                    with tracer.start_as_current_span(self._span_name, kind=SpanKind.CLIENT) as self._span:
+                        process_streaming_chat_response(
+                            self,
+                            pricing_info=pricing_info,
+                            environment=environment,
+                            application_name=application_name,
+                            metrics=metrics,
+                            capture_message_content=capture_message_content,
+                            disable_metrics=disable_metrics,
+                            version=version
+                        )
+
+                except Exception as e:
+                    handle_exception(self._span, e)
+
+                raise
+
+    async def wrapper(wrapped, instance, args, kwargs):
+        """
+        Wraps the GenAI function call.
+        """
+        # Check if streaming is enabled for the API call
+        streaming = kwargs.get("stream", False)
+        server_address, server_port = set_server_address_and_port(instance, "NOT_FOUND", "NOT_FOUND")
+        request_model = kwargs.get("model", "openai/gpt-4o")
+
+        span_name = f"{SemanticConvention.GEN_AI_OPERATION_TYPE_CHAT} {request_model}"
+
+        if streaming:
+            # Special handling for streaming response
+            awaited_wrapped = await wrapped(*args, **kwargs)
+            span = tracer.start_span(span_name, kind=SpanKind.CLIENT)
+            return TracedAsyncStream(awaited_wrapped, span, span_name, kwargs, server_address, server_port)
+        else:
+            # Handling for non-streaming responses
+            with tracer.start_as_current_span(span_name, kind=SpanKind.CLIENT) as span:
+                start_time = time.time()
+                response = await wrapped(*args, **kwargs)
+
+                try:
+                    response = process_chat_response(
+                        response=response,
+                        request_model=request_model,
+                        pricing_info=pricing_info,
+                        server_port=server_port,
+                        server_address=server_address,
+                        environment=environment,
+                        application_name=application_name,
+                        metrics=metrics,
+                        start_time=start_time,
+                        span=span,
+                        capture_message_content=capture_message_content,
+                        disable_metrics=disable_metrics,
+                        version=version,
+                        **kwargs
+                    )
+
+                except Exception as e:
+                    handle_exception(span, e)
+
+                return response
+
+    return wrapper
+
+def aembedding(version, environment, application_name, tracer, pricing_info,
+    capture_message_content, metrics, disable_metrics):
+    """
+    Generates a telemetry wrapper for GenAI embedding function call
+    """
+
+    async def wrapper(wrapped, instance, args, kwargs):
+        """
+        Wraps the GenAI embedding function call.
+        """
+        server_address, server_port = set_server_address_and_port(instance, "NOT_FOUND", "NOT_FOUND")
+        request_model = kwargs.get("model", "text-embedding-ada-002")
+
+        span_name = f"{SemanticConvention.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 = await wrapped(*args, **kwargs)
+
+            try:
+                response = process_embedding_response(
+                    response=response,
+                    request_model=request_model,
+                    pricing_info=pricing_info,
+                    server_port=server_port,
+                    server_address=server_address,
+                    environment=environment,
+                    application_name=application_name,
+                    metrics=metrics,
+                    start_time=start_time,
+                    span=span,
+                    capture_message_content=capture_message_content,
+                    disable_metrics=disable_metrics,
+                    version=version,
+                    **kwargs
+                )
+
+            except Exception as e:
+                handle_exception(span, e)
+
+            return response
+
+    return wrapper

openlit-1.34.15/src/openlit/instrumentation/litellm/litellm.py

@@ -0,0 +1,188 @@
+"""
+Module for monitoring LiteLLM API calls.
+"""
+
+import time
+from opentelemetry.trace import SpanKind
+from openlit.__helpers import (
+    handle_exception,
+    set_server_address_and_port
+)
+from openlit.instrumentation.litellm.utils import (
+    process_chunk,
+    process_streaming_chat_response,
+    process_chat_response,
+    process_embedding_response
+)
+from openlit.semcov import SemanticConvention
+
+def completion(version, environment, application_name, tracer, pricing_info,
+    capture_message_content, metrics, disable_metrics):
+    """
+    Generates a telemetry wrapper for GenAI function call
+    """
+
+    class TracedSyncStream:
+        """
+        Wrapper for streaming responses to collect telemetry.
+        """
+
+        def __init__(
+                self,
+                wrapped,
+                span,
+                span_name,
+                kwargs,
+                server_address,
+                server_port,
+                **args,
+            ):
+            self.__wrapped__ = wrapped
+            self._span = span
+            self._span_name = span_name
+            self._llmresponse = ""
+            self._response_id = ""
+            self._response_model = ""
+            self._finish_reason = ""
+            self._response_service_tier = ""
+            self._tools = None
+            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__()
+                process_chunk(self, chunk)
+                return chunk
+            except StopIteration:
+                try:
+                    with tracer.start_as_current_span(self._span_name, kind=SpanKind.CLIENT) as self._span:
+                        process_streaming_chat_response(
+                            self,
+                            pricing_info=pricing_info,
+                            environment=environment,
+                            application_name=application_name,
+                            metrics=metrics,
+                            capture_message_content=capture_message_content,
+                            disable_metrics=disable_metrics,
+                            version=version
+                        )
+
+                except Exception as e:
+                    handle_exception(self._span, e)
+
+                raise
+
+    def wrapper(wrapped, instance, args, kwargs):
+        """
+        Wraps the GenAI function call.
+        """
+        # Check if streaming is enabled for the API call
+        streaming = kwargs.get("stream", False)
+        server_address, server_port = set_server_address_and_port(instance, "NOT_FOUND", "NOT_FOUND")
+        request_model = kwargs.get("model", "openai/gpt-4o")
+
+        span_name = f"{SemanticConvention.GEN_AI_OPERATION_TYPE_CHAT} {request_model}"
+
+        if streaming:
+            # Special handling for streaming response
+            awaited_wrapped = wrapped(*args, **kwargs)
+            span = tracer.start_span(span_name, kind=SpanKind.CLIENT)
+            return TracedSyncStream(awaited_wrapped, span, span_name, kwargs, server_address, server_port)
+        else:
+            # Handling for non-streaming responses
+            with tracer.start_as_current_span(span_name, kind=SpanKind.CLIENT) as span:
+                start_time = time.time()
+                response = wrapped(*args, **kwargs)
+
+                try:
+                    response = process_chat_response(
+                        response=response,
+                        request_model=request_model,
+                        pricing_info=pricing_info,
+                        server_port=server_port,
+                        server_address=server_address,
+                        environment=environment,
+                        application_name=application_name,
+                        metrics=metrics,
+                        start_time=start_time,
+                        span=span,
+                        capture_message_content=capture_message_content,
+                        disable_metrics=disable_metrics,
+                        version=version,
+                        **kwargs
+                    )
+
+                except Exception as e:
+                    handle_exception(span, e)
+
+                return response
+
+    return wrapper
+
+def embedding(version, environment, application_name, tracer, pricing_info,
+    capture_message_content, metrics, disable_metrics):
+    """
+    Generates a telemetry wrapper for GenAI embedding function call
+    """
+
+    def wrapper(wrapped, instance, args, kwargs):
+        """
+        Wraps the GenAI embedding function call.
+        """
+        server_address, server_port = set_server_address_and_port(instance, "NOT_FOUND", "NOT_FOUND")
+        request_model = kwargs.get("model", "text-embedding-ada-002")
+
+        span_name = f"{SemanticConvention.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)
+
+            try:
+                response = process_embedding_response(
+                    response=response,
+                    request_model=request_model,
+                    pricing_info=pricing_info,
+                    server_port=server_port,
+                    server_address=server_address,
+                    environment=environment,
+                    application_name=application_name,
+                    metrics=metrics,
+                    start_time=start_time,
+                    span=span,
+                    capture_message_content=capture_message_content,
+                    disable_metrics=disable_metrics,
+                    version=version,
+                    **kwargs
+                )
+
+            except Exception as e:
+                handle_exception(span, e)
+
+            return response
+
+    return wrapper