openlit 1.34.17__py3-none-any.whl → 1.34.19__py3-none-any.whl

openlit/instrumentation/cohere/utils.py

@@ -0,0 +1,330 @@
+"""
+Cohere OpenTelemetry instrumentation utility functions
+"""
+import time
+
+from opentelemetry.trace import Status, StatusCode
+
+from openlit.__helpers import (
+    calculate_ttft,
+    response_as_dict,
+    calculate_tbt,
+    get_chat_model_cost,
+    get_embed_model_cost,
+    common_span_attributes,
+    record_completion_metrics,
+    record_embedding_metrics,
+)
+from openlit.semcov import SemanticConvention
+
+def format_content(messages):
+    """
+    Process a list of messages to extract content.
+    """
+
+    formatted_messages = []
+    for message in messages:
+        # Handle both dictionary and object formats
+        if isinstance(message, dict):
+            role = message.get("role", "user")
+            content = message.get("content", "")
+        else:
+            # Handle Cohere object format (e.g., cohere.UserChatMessageV2)
+            role = getattr(message, "role", "user")
+            content = getattr(message, "content", "")
+
+        if isinstance(content, list):
+            content_str = ", ".join(
+                f'{item["type"]}: {item["text"] if "text" in item else item.get("image_url", "")}'
+                if "type" in item else f'text: {item.get("text", "")}'
+                for item in content
+            )
+            formatted_messages.append(f"{role}: {content_str}")
+        else:
+            formatted_messages.append(f"{role}: {content}")
+
+    return "\n".join(formatted_messages)
+
+def process_chunk(scope, chunk):
+    """
+    Process a chunk of response data and update state.
+    """
+
+    end_time = time.time()
+    # Record the timestamp for the current chunk
+    scope._timestamps.append(end_time)
+
+    if len(scope._timestamps) == 1:
+        # Calculate time to first chunk
+        scope._ttft = calculate_ttft(scope._timestamps, scope._start_time)
+
+    chunked = response_as_dict(chunk)
+
+    # Handle different chunk types for Cohere streaming
+    if chunked.get("type") == "message-start":
+        scope._response_id = chunked.get("id")
+
+    if chunked.get("type") == "content-delta":
+        content = chunked.get("delta", {}).get("message", {}).get("content", {}).get("text")
+        if content:
+            scope._llmresponse += content
+
+    # Handle tool plan deltas
+    if chunked.get("type") == "tool-plan-delta":
+        tool_plan_text = chunked.get("delta", {}).get("message", {}).get("tool_plan", "")
+        if tool_plan_text:
+            if not hasattr(scope, "_tool_plan"):
+                scope._tool_plan = ""
+            scope._tool_plan += tool_plan_text
+
+    # Handle tool call start
+    if chunked.get("type") == "tool-call-start":
+        if not hasattr(scope, "_tools") or scope._tools is None:
+            scope._tools = []
+
+        index = chunked.get("index", 0)
+        tool_call = chunked.get("delta", {}).get("message", {}).get("tool_calls", {})
+
+        # Extend list if needed
+        scope._tools.extend([{}] * (index + 1 - len(scope._tools)))
+
+        # Initialize tool call
+        scope._tools[index] = {
+            "id": tool_call.get("id", ""),
+            "type": tool_call.get("type", "function"),
+            "function": {
+                "name": tool_call.get("function", {}).get("name", ""),
+                "arguments": ""
+            }
+        }
+
+    # Handle tool call deltas (arguments)
+    if chunked.get("type") == "tool-call-delta":
+        if hasattr(scope, "_tools") and scope._tools:
+            index = chunked.get("index", 0)
+            if index < len(scope._tools):
+                arguments = chunked.get("delta", {}).get("message", {}).get("tool_calls", {}).get("function", {}).get("arguments", "")
+                if arguments:
+                    scope._tools[index]["function"]["arguments"] += arguments
+
+    if chunked.get("type") == "message-end":
+        delta = chunked.get("delta", {})
+        scope._finish_reason = delta.get("finish_reason", "")
+        usage = delta.get("usage", {}).get("billed_units", {})
+        scope._input_tokens = usage.get("input_tokens", 0)
+        scope._output_tokens = usage.get("output_tokens", 0)
+        scope._end_time = time.time()
+
+def common_chat_logic(scope, pricing_info, environment, application_name, metrics,
+    capture_message_content, disable_metrics, version, is_stream):
+    """
+    Process chat request and generate Telemetry
+    """
+
+    if len(scope._timestamps) > 1:
+        scope._tbt = calculate_tbt(scope._timestamps)
+
+    prompt = format_content(scope._kwargs.get("messages", []))
+    request_model = scope._kwargs.get("model", "command-r-plus-08-2024")
+
+    cost = get_chat_model_cost(request_model, pricing_info, scope._input_tokens, scope._output_tokens)
+
+    # Common Span Attributes
+    common_span_attributes(scope,
+        SemanticConvention.GEN_AI_OPERATION_TYPE_CHAT, SemanticConvention.GEN_AI_SYSTEM_COHERE,
+        scope._server_address, scope._server_port, request_model, scope._response_model,
+        environment, application_name, is_stream, scope._tbt, scope._ttft, version)
+
+    # Span Attributes for Request parameters
+    scope._span.set_attribute(SemanticConvention.GEN_AI_REQUEST_SEED, scope._kwargs.get("seed", ""))
+    scope._span.set_attribute(SemanticConvention.GEN_AI_REQUEST_FREQUENCY_PENALTY, scope._kwargs.get("frequency_penalty", 0.0))
+    scope._span.set_attribute(SemanticConvention.GEN_AI_REQUEST_MAX_TOKENS, scope._kwargs.get("max_tokens", -1))
+    scope._span.set_attribute(SemanticConvention.GEN_AI_REQUEST_PRESENCE_PENALTY, scope._kwargs.get("presence_penalty", 0.0))
+    scope._span.set_attribute(SemanticConvention.GEN_AI_REQUEST_STOP_SEQUENCES, scope._kwargs.get("stop_sequences", []))
+    scope._span.set_attribute(SemanticConvention.GEN_AI_REQUEST_TEMPERATURE, scope._kwargs.get("temperature", 0.3))
+    scope._span.set_attribute(SemanticConvention.GEN_AI_REQUEST_TOP_K, scope._kwargs.get("k", 1.0))
+    scope._span.set_attribute(SemanticConvention.GEN_AI_REQUEST_TOP_P, scope._kwargs.get("p", 1.0))
+
+    # Span Attributes for Response parameters
+    scope._span.set_attribute(SemanticConvention.GEN_AI_RESPONSE_ID, scope._response_id)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_RESPONSE_FINISH_REASON, [scope._finish_reason])
+    scope._span.set_attribute(SemanticConvention.GEN_AI_OUTPUT_TYPE, "text" if isinstance(scope._llmresponse, str) else "json")
+
+    # Span Attributes for Cost and Tokens
+    scope._span.set_attribute(SemanticConvention.GEN_AI_USAGE_INPUT_TOKENS, scope._input_tokens)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_USAGE_OUTPUT_TOKENS, scope._output_tokens)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_CLIENT_TOKEN_USAGE, scope._input_tokens + scope._output_tokens)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_USAGE_COST, cost)
+
+    # Span Attributes for Tools - optimized
+    if scope._tools:
+        tools = scope._tools if isinstance(scope._tools, list) else [scope._tools]
+
+        names, ids, args = zip(*[
+            (t.get("function", {}).get("name", ""),
+             str(t.get("id", "")),
+             str(t.get("function", {}).get("arguments", "")))
+            for t in tools if isinstance(t, dict) and t
+        ]) if tools else ([], [], [])
+
+        scope._span.set_attribute(SemanticConvention.GEN_AI_TOOL_NAME, ", ".join(filter(None, names)))
+        scope._span.set_attribute(SemanticConvention.GEN_AI_TOOL_CALL_ID, ", ".join(filter(None, ids)))
+        scope._span.set_attribute(SemanticConvention.GEN_AI_TOOL_ARGS, ", ".join(filter(None, args)))
+
+    # Span Attributes for Tool Plan (Cohere specific)
+    if hasattr(scope, "_tool_plan") and scope._tool_plan:
+        scope._span.set_attribute(SemanticConvention.GEN_AI_CONTENT_REASONING, scope._tool_plan)
+
+    # 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._llmresponse)
+
+        # 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._llmresponse,
+            },
+        )
+
+    scope._span.set_status(Status(StatusCode.OK))
+
+    # Metrics
+    if not disable_metrics:
+        record_completion_metrics(metrics, SemanticConvention.GEN_AI_OPERATION_TYPE_CHAT, SemanticConvention.GEN_AI_SYSTEM_COHERE,
+            scope._server_address, scope._server_port, request_model, scope._response_model, environment,
+            application_name, scope._start_time, scope._end_time, scope._input_tokens, scope._output_tokens,
+            cost, scope._tbt, scope._ttft)
+
+def process_streaming_chat_response(scope, pricing_info, environment, application_name, metrics,
+    capture_message_content=False, disable_metrics=False, version=""):
+    """
+    Process streaming chat request and generate Telemetry
+    """
+
+    common_chat_logic(scope, pricing_info, environment, application_name, metrics,
+        capture_message_content, disable_metrics, version, is_stream=True)
+
+def process_chat_response(response, request_model, pricing_info, server_port, server_address,
+    environment, application_name, metrics, start_time, span, capture_message_content=False,
+    disable_metrics=False, version="1.0.0", **kwargs):
+    """
+    Process chat request and generate Telemetry
+    """
+
+    # Create scope object
+    scope = type("GenericScope", (), {})()
+    response_dict = response_as_dict(response)
+
+    scope._start_time = start_time
+    scope._end_time = time.time()
+    scope._span = span
+    # Extract response content - handle both text and tool responses
+    message = response_dict.get("message", {})
+    content_list = message.get("content", [])
+    if content_list and isinstance(content_list, list) and len(content_list) > 0:
+        scope._llmresponse = content_list[0].get("text", "")
+    else:
+        scope._llmresponse = ""
+    scope._response_id = response_dict.get("id")
+    scope._response_model = request_model
+    scope._input_tokens = response_dict.get("usage", {}).get("billed_units", {}).get("input_tokens", 0)
+    scope._output_tokens = response_dict.get("usage", {}).get("billed_units", {}).get("output_tokens", 0)
+    scope._timestamps = []
+    scope._ttft, scope._tbt = scope._end_time - scope._start_time, 0
+    scope._server_address, scope._server_port = server_address, server_port
+    scope._kwargs = kwargs
+    scope._finish_reason = response_dict.get("finish_reason", "")
+
+    # Handle tool calls
+    if scope._kwargs.get("tools"):
+        scope._tools = response_dict.get("message", {}).get("tool_calls")
+        # Handle tool plan if present
+        scope._tool_plan = response_dict.get("message", {}).get("tool_plan", "")
+    else:
+        scope._tools = None
+        scope._tool_plan = ""
+
+    common_chat_logic(scope, pricing_info, environment, application_name, metrics,
+        capture_message_content, disable_metrics, version, is_stream=False)
+
+    return response
+
+def common_embedding_logic(scope, pricing_info, environment, application_name, metrics,
+    capture_message_content, disable_metrics, version):
+    """
+    Process embedding request and generate Telemetry
+    """
+
+    request_model = scope._kwargs.get("model", "embed-english-v3.0")
+    inputs = scope._kwargs.get("texts", [])
+
+    cost = get_embed_model_cost(request_model, pricing_info, scope._input_tokens)
+
+    # Common Span Attributes
+    common_span_attributes(scope,
+        SemanticConvention.GEN_AI_OPERATION_TYPE_EMBEDDING, SemanticConvention.GEN_AI_SYSTEM_COHERE,
+        scope._server_address, scope._server_port, request_model, scope._response_model,
+        environment, application_name, False, 0, scope._ttft, version)
+
+    # Span Attributes for Request parameters
+    scope._span.set_attribute(SemanticConvention.GEN_AI_REQUEST_ENCODING_FORMATS, scope._kwargs.get("embedding_types", ["float"]))
+
+    # Span Attributes for Cost and Tokens
+    scope._span.set_attribute(SemanticConvention.GEN_AI_USAGE_INPUT_TOKENS, scope._input_tokens)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_CLIENT_TOKEN_USAGE, scope._input_tokens)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_USAGE_COST, cost)
+    scope._span.set_attribute(SemanticConvention.GEN_AI_OUTPUT_TYPE, scope._response_type)
+
+    # Span Attributes for Content
+    if capture_message_content:
+        scope._span.set_attribute(SemanticConvention.GEN_AI_CONTENT_PROMPT, str(inputs))
+
+        # 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: str(inputs),
+            },
+        )
+
+    scope._span.set_status(Status(StatusCode.OK))
+
+    # Metrics
+    if not disable_metrics:
+        record_embedding_metrics(metrics, SemanticConvention.GEN_AI_OPERATION_TYPE_EMBEDDING, SemanticConvention.GEN_AI_SYSTEM_COHERE,
+            scope._server_address, scope._server_port, request_model, scope._response_model, environment,
+            application_name, scope._start_time, scope._end_time, scope._input_tokens, cost)
+
+def process_embedding_response(response, request_model, 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 embedding request and generate Telemetry
+    """
+
+    # Create scope object
+    scope = type("GenericScope", (), {})()
+    response_dict = response_as_dict(response)
+
+    scope._start_time = start_time
+    scope._end_time = time.time()
+    scope._span = span
+    scope._input_tokens = response_dict.get("meta", {}).get("billed_units", {}).get("input_tokens", 0)
+    scope._response_model = request_model
+    scope._response_type = response_dict.get("response_type", "")
+    scope._ttft = scope._end_time - scope._start_time
+    scope._server_address, scope._server_port = server_address, server_port
+    scope._kwargs = kwargs
+
+    common_embedding_logic(scope, pricing_info, environment, application_name, metrics,
+        capture_message_content, disable_metrics, version)
+
+    return response

{openlit-1.34.17.dist-info → openlit-1.34.19.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: openlit
-Version: 1.34.17
+Version: 1.34.19
 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-1.34.17.dist-info → openlit-1.34.19.dist-info}/RECORD

@@ -18,10 +18,10 @@ openlit/instrumentation/ai21/__init__.py,sha256=tKX643fwxPWPJq1EXEZd0Xpd6B0jl_Vi
 openlit/instrumentation/ai21/ai21.py,sha256=zyQMfCLcOFG1tQWrZmGeMaVAmj8MtCUeXQtPHmlUAO0,6533
 openlit/instrumentation/ai21/async_ai21.py,sha256=q1Dhxru4tUJu0U1Px3PptNqrSGW0-VfRGcqkLKFR8vQ,6659
 openlit/instrumentation/ai21/utils.py,sha256=5zf69uw_TT8u-q-6R6rBeGm1bX0WpsbrAq-MTTZJ9Bk,14309
-openlit/instrumentation/anthropic/__init__.py,sha256=QEsiwdxcQDzzlVYR4_x7KTdf0-UJDJt8FjwNQMspnxM,1929
-openlit/instrumentation/anthropic/anthropic.py,sha256=NxJJjhsu9sSFIlBp322olGkPlLt9Bn5sndaugYA68dE,5149
-openlit/instrumentation/anthropic/async_anthropic.py,sha256=ivJGygKWVTS2hWWX12_g1tiq-5mpeHXETZsWoFZL3UE,5235
-openlit/instrumentation/anthropic/utils.py,sha256=g15QqkHdl2N5WsRCNvxWkMwOYpR_n-lrMcvlY71QbSs,11934
+openlit/instrumentation/anthropic/__init__.py,sha256=mpb15zoUk1jymxUOyyk4IZSKlhsra-3MjO7RLbm1o-I,1753
+openlit/instrumentation/anthropic/anthropic.py,sha256=e5by3_c8zr_6XsQX4R96Wzj840dDiQZe25BgM5melU0,4892
+openlit/instrumentation/anthropic/async_anthropic.py,sha256=fZLqZO5gOMBn_KtTl-5MEDfoOFyQctpNxo6saVaCw8I,4994
+openlit/instrumentation/anthropic/utils.py,sha256=W9paSh2npsCwsHyf-RyHeg0eRDcE509_f-pqb2IeO9c,9761
 openlit/instrumentation/assemblyai/__init__.py,sha256=-pW7c5Vxa493yETQABbebx4be_sTx5VwkvQrIHbhRbI,1404
 openlit/instrumentation/assemblyai/assemblyai.py,sha256=SJZ-O6k8adlRWJ2gMIP62vXobHJ3VI87PQOCFw9Ilng,2071
 openlit/instrumentation/assemblyai/utils.py,sha256=driBfwWBveWTqHyPRtl1R8oEG6m07-GXycyCnDfZ9PM,6089
@@ -38,9 +38,10 @@ openlit/instrumentation/bedrock/bedrock.py,sha256=kP9ESKzqhWu-dIWseyaeyendUyo6b7
 openlit/instrumentation/bedrock/utils.py,sha256=_mTUIbioEg4jfoxocUbfc7RgGjhm9ACelbxIoFu4jbM,11636
 openlit/instrumentation/chroma/__init__.py,sha256=4ZeHY1OInRKQbb4qg8BVvGJtWN1XdzW6mosqi7-6ruE,3353
 openlit/instrumentation/chroma/chroma.py,sha256=fPIX6vgcDFkBgKGxrkXZWDaBtZa5y-PV3jJjKsFqjDg,10640
-openlit/instrumentation/cohere/__init__.py,sha256=TIRq1obu-zqBji0HhMbFGfI2q5m-zw0nWbToKeZqpg4,2905
-openlit/instrumentation/cohere/async_cohere.py,sha256=cFIrQ-maKmM8JKRitX-mgD7Jm_1UDyO_kOp4X_0MQww,30745
-openlit/instrumentation/cohere/cohere.py,sha256=11QGsBTLiTfF-oH_fIQ9IBbIgtS2vf7M-VvbG8MuO_Q,30628
+openlit/instrumentation/cohere/__init__.py,sha256=FIJ_QokUZcN9UsVPWckRfHiLVJ5zj3jtcOjQmjQA5f0,2978
+openlit/instrumentation/cohere/async_cohere.py,sha256=wgctJa-BcKEkVguLorR7mYRShv7GiioF1_zxevvzXTw,6990
+openlit/instrumentation/cohere/cohere.py,sha256=7YAj6kwJ1MDLFY3WtbPRehZFIseDVIbvPQJZxG8Qc6A,6814
+openlit/instrumentation/cohere/utils.py,sha256=5l3Av-t0GzawPd0j5L9jcpdJDGOvLq2YCMpo3usS4OY,14533
 openlit/instrumentation/controlflow/__init__.py,sha256=Y5xSj6bwkGufG989YtRGujzj8qrV4T5kXr4hzGb2nGk,2168
 openlit/instrumentation/controlflow/controlflow.py,sha256=NNovai9DwRCSaa-shFxAyUIYiRaEBgXkyuXFk3lQlPU,5573
 openlit/instrumentation/crawl4ai/__init__.py,sha256=TTFFeZkTgPZJAD1Tlr2LK-CTH2A7NkZK-5SHKFfFOtU,1947
@@ -140,7 +141,7 @@ openlit/otel/events.py,sha256=VrMjTpvnLtYRBHCiFwJojTQqqNpRCxoD4yJYeQrtPsk,3560
 openlit/otel/metrics.py,sha256=GM2PDloBGRhBTkHHkYaqmOwIAQkY124ZhW4sEqW1Fgk,7086
 openlit/otel/tracing.py,sha256=tjV2bEbEDPUB1Z46gE-UsJsb04sRdFrfbhIDkxViZc0,3103
 openlit/semcov/__init__.py,sha256=8oIh2VC667NDh8FA3M-ESusHmeus1sgDUD8binx_nAc,13519
-openlit-1.34.17.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-openlit-1.34.17.dist-info/METADATA,sha256=1pKz7cmbdQZWl_nV785HCIDmqfvzimSRr4tEiPKxQ3s,23470
-openlit-1.34.17.dist-info/WHEEL,sha256=b4K_helf-jlQoXBBETfwnf4B04YC67LOev0jo4fX5m8,88
-openlit-1.34.17.dist-info/RECORD,,
+openlit-1.34.19.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+openlit-1.34.19.dist-info/METADATA,sha256=dxAiXXKjEgxrLrnGShuRAbVebbg1GOmPaa5CODpVqCA,23470
+openlit-1.34.19.dist-info/WHEEL,sha256=b4K_helf-jlQoXBBETfwnf4B04YC67LOev0jo4fX5m8,88
+openlit-1.34.19.dist-info/RECORD,,