ragaai-catalyst 2.1.7.2b0__py3-none-any.whl → 2.1.7.3b0__py3-none-any.whl

ragaai_catalyst/tracers/utils/rag_trace_json_converter.py

@@ -44,7 +44,7 @@ def rag_trace_json_converter(input_trace, custom_model_cost, trace_id, user_deta
                                             if isinstance(message, str):
                                                 human_index = message.find("Human:")
                                                 if human_index != -1:
-                                                    human_message = message[human_index:]
+                                                    human_message = message[human_index:].replace("Human:", "")
                                                     break
                                         return human_message if human_message else value
                                 except Exception as e:
@@ -212,14 +212,14 @@ def get_additional_metadata(spans, custom_model_cost, model_cost_dict, prompt=""
                     additional_metadata["tokens"]["prompt"] = span["attributes"]["llm.token_count.prompt"]
 
                 except:
-                    logger.warning("Warning: prompt token not found. using fallback strategies to get tokens.")
+                    logger.debug("Warning: prompt token not found. using fallback strategies to get tokens.")
                     try:
                         additional_metadata["tokens"]["prompt"] = num_tokens_from_messages(
                             model=additional_metadata["model_name"],
                             message=prompt
                         )
                     except Exception as e:
-                        logger.warning(f"Failed to count prompt tokens: {str(e)}. Using 'gpt-4o-mini' model count as fallback.")
+                        logger.debug(f"Failed to count prompt tokens: {str(e)}. Using 'gpt-4o-mini' model count as fallback.")
                         additional_metadata["tokens"]["prompt"] = num_tokens_from_messages(
                             model="gpt-4o-mini",
                             message=prompt
@@ -228,14 +228,14 @@ def get_additional_metadata(spans, custom_model_cost, model_cost_dict, prompt=""
                 try:
                     additional_metadata["tokens"]["completion"] = span["attributes"]["llm.token_count.completion"]
                 except:
-                    logger.warning("Warning: completion token not found. using fallback strategies to get tokens.")
+                    logger.debug("Warning: completion token not found. using fallback strategies to get tokens.")
                     try:
                         additional_metadata["tokens"]["completion"] = num_tokens_from_messages(
                             model=additional_metadata["model_name"],
                             message=response
                         )
                     except Exception as e:
-                        logger.warning(f"Failed to count completion tokens: {str(e)}. Using 'gpt-4o-mini' model count as fallback.")
+                        logger.debug(f"Failed to count completion tokens: {str(e)}. Using 'gpt-4o-mini' model count as fallback.")
                         additional_metadata["tokens"]["completion"] = num_tokens_from_messages(
                             model="gpt-4o-mini",
                             message=response
@@ -299,96 +299,47 @@ def get_additional_metadata(spans, custom_model_cost, model_cost_dict, prompt=""
 
 def num_tokens_from_messages(model, message):
     try:
-        # Handle None or empty message
         if not message:
-            logger.warning("Empty or None message provided to token counter")
+            logger.error("Empty or None message provided to token counter")
             return 0
-            
-        # GPT models
-        if re.match(r'^gpt-', model):
-            """Check if the model is any GPT model (pattern: ^gpt-)
-            This matches any model name that starts with 'gpt-'
-            """
-            def num_tokens_from_string(string: str, encoding_name: str) -> int:
-                """Returns the number of tokens in a text string."""
+        
+        def num_tokens_from_string(text_content: str, encoding_name: str) -> int:
+            """Returns the number of tokens in a text string."""
+            if isinstance(text_content, list):
+                list_str = str(text_content[0]) if text_content else ""
+                pattern = r"content=\'(.*?)\'(?:\s+additional_kwargs=|$)"
+                match = re.search(pattern, list_str, re.DOTALL)
+                if match:
+                    text_content = match.group(1)  # Extract content and process it for tokens
+                else:
+                    text_content = list_str 
+            try:
+                encoding = tiktoken.get_encoding(encoding_name)
+                return len(encoding.encode(text_content))
+            except Exception as e:
+                logger.warning(f"Error encoding with {encoding_name}: {str(e)}")
                 try:
-                    encoding = tiktoken.get_encoding(encoding_name)
-                    num_tokens = len(encoding.encode(string))
-                    return num_tokens
-                except Exception as e:
-                    logger.warning(f"Error encoding with {encoding_name}: {str(e)}")
-                    # Fallback to a different encoding if the requested one fails
-                    try:
-                        fallback_encoding = tiktoken.get_encoding("cl100k_base")
-                        return len(fallback_encoding.encode(string))
-                    except:
-                        logger.error("Failed to use fallback encoding")
-                        return 0
-            
-            if re.match(r'^gpt-4o.*', model):
-                """Check for GPT-4 Optimized models (pattern: ^gpt-4o.*)
-                Examples that match:
-                - gpt-4o
-                - gpt-4o-mini
-                - gpt-4o-2024-08-06
-                The .* allows for any characters after 'gpt-4o'
-                """
+                    fallback_encoding = tiktoken.get_encoding("cl100k_base")
+                    return len(fallback_encoding.encode(text_content))
+                except:
+                    logger.debug("Failed to use fallback encoding")
+                    return 0
+        
+        # Determine which encoding to use based on model name
+        encoding_name = "o200k_base"  
+        
+        if re.match(r'^gpt-', model):
+            if re.match(r'^gpt-(4o|4\.1).*', model):
+                # GPT-4o and GPT-4.1 models
                 encoding_name = "o200k_base"
-                return num_tokens_from_string(message, encoding_name)
-            
             elif re.match(r'^gpt-(4|3\.5).*', model):
-                """Check for GPT-4 and GPT-3.5 models (pattern: ^gpt-(4|3\.5).*)
-                Uses cl100k_base encoding for GPT-4 and GPT-3.5 models
-                Examples that match:
-                - gpt-4
-                - gpt-4-turbo
-                - gpt-4-2024-08-06
-                - gpt-3.5-turbo
-                - gpt-3.5-turbo-16k
-                """
+                # GPT-4 and GPT-3.5 models
                 encoding_name = "cl100k_base"
-                return num_tokens_from_string(message, encoding_name)
+        else:
+            logger.debug(f"Using default token counter for: {model}.")
             
-            else:
-                """Default case for any other GPT models
-                Uses o200k_base encoding as the default tokenizer
-                """
-                return num_tokens_from_string(message, encoding_name="o200k_base")
+        return num_tokens_from_string(message, encoding_name)
             
-
-        # Gemini models 
-        elif re.match(r'^gemini-', model):
-            try:
-                GOOGLE_API_KEY = os.getenv("GOOGLE_API_KEY")
-                if not GOOGLE_API_KEY:
-                    logger.warning("GOOGLE_API_KEY not found in environment variables")
-                    return 0
-                    
-                import google.generativeai as genai
-                client = genai.Client(api_key=GOOGLE_API_KEY)
-
-                response = client.models.count_tokens(
-                        model=model,
-                        contents=message,
-                    )
-                return response.total_tokens
-            except ImportError:
-                logger.warning("google.generativeai module not found. Install with pip install google-generativeai")
-                return 0
-            except Exception as e:
-                logger.warning(f"Error counting tokens for Gemini model: {str(e)}")
-                return 0
-        
-        # Default case for unknown models
-        else:
-            logger.warning(f"Unknown model type: {model}. Using default token counter.")
-            try:
-                # Use cl100k_base as a fallback for unknown models
-                encoding = tiktoken.get_encoding("cl100k_base")
-                return len(encoding.encode(message))
-            except:
-                logger.error("Failed to use fallback encoding for unknown model")
-                return 0
     except Exception as e:
         logger.error(f"Unexpected error in token counting: {str(e)}")
         return 0

{ragaai_catalyst-2.1.7.2b0.dist-info → ragaai_catalyst-2.1.7.3b0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ragaai_catalyst
-Version: 2.1.7.2b0
+Version: 2.1.7.3b0
 Summary: RAGA AI CATALYST
 Author-email: Kiran Scaria <kiran.scaria@raga.ai>, Kedar Gaikwad <kedar.gaikwad@raga.ai>, Dushyant Mahajan <dushyant.mahajan@raga.ai>, Siddhartha Kosti <siddhartha.kosti@raga.ai>, Ritika Goel <ritika.goel@raga.ai>, Vijay Chaurasia <vijay.chaurasia@raga.ai>, Tushar Kumar <tushar.kumar@raga.ai>
 Requires-Python: <=3.13.2,>=3.10

{ragaai_catalyst-2.1.7.2b0.dist-info → ragaai_catalyst-2.1.7.3b0.dist-info}/RECORD

@@ -85,11 +85,11 @@ ragaai_catalyst/tracers/utils/convert_llama_instru_callback.py,sha256=8qLo7x4Zsn
 ragaai_catalyst/tracers/utils/extraction_logic_llama_index.py,sha256=ZhPs0YhVtB82-Pq9o1BvCinKE_WPvVxPTEcZjlJbFYM,2371
 ragaai_catalyst/tracers/utils/langchain_tracer_extraction_logic.py,sha256=XS2_x2qneqEx9oAighLg-LRiueWcESLwIC2r7eJT-Ww,3117
 ragaai_catalyst/tracers/utils/model_prices_and_context_window_backup.json,sha256=C3uwkibJ08C9sOX-54kulZYmJlIpZ-SQpfE6HNGrjbM,343502
-ragaai_catalyst/tracers/utils/rag_trace_json_converter.py,sha256=Y1xWS9aPBUt4cg1Rsgg6m3K6B9LuWBp2BXPliCmGk3E,20615
+ragaai_catalyst/tracers/utils/rag_trace_json_converter.py,sha256=jwScK9bo0NGpaOgeII-m2l_ovcHNo9Pd2wkZHFBt0zc,18559
 ragaai_catalyst/tracers/utils/trace_json_converter.py,sha256=E0_QfciQMMpCtQYrNB4l8HJhlaFalr5bkMqkVRgQahY,14073
 ragaai_catalyst/tracers/utils/utils.py,sha256=ViygfJ7vZ7U0CTSA1lbxVloHp4NSlmfDzBRNCJuMhis,2374
-ragaai_catalyst-2.1.7.2b0.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-ragaai_catalyst-2.1.7.2b0.dist-info/METADATA,sha256=gZkHYlK9kJaM3_kNJtTQzFbIYV1T18LokNMGtqdwHzQ,17607
-ragaai_catalyst-2.1.7.2b0.dist-info/WHEEL,sha256=pxyMxgL8-pra_rKaQ4drOZAegBVuX-G_4nRHjjgWbmo,91
-ragaai_catalyst-2.1.7.2b0.dist-info/top_level.txt,sha256=HpgsdRgEJMk8nqrU6qdCYk3di7MJkDL0B19lkc7dLfM,16
-ragaai_catalyst-2.1.7.2b0.dist-info/RECORD,,
+ragaai_catalyst-2.1.7.3b0.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+ragaai_catalyst-2.1.7.3b0.dist-info/METADATA,sha256=5TXfNHsBByRy9NfLRp8vCsmedgGXJpq0qc_18g43HYA,17607
+ragaai_catalyst-2.1.7.3b0.dist-info/WHEEL,sha256=SmOxYU7pzNKBqASvQJ7DjX3XGUF92lrGhMb3R6_iiqI,91
+ragaai_catalyst-2.1.7.3b0.dist-info/top_level.txt,sha256=HpgsdRgEJMk8nqrU6qdCYk3di7MJkDL0B19lkc7dLfM,16
+ragaai_catalyst-2.1.7.3b0.dist-info/RECORD,,

{ragaai_catalyst-2.1.7.2b0.dist-info → ragaai_catalyst-2.1.7.3b0.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (79.0.0)
+Generator: setuptools (79.0.1)
 Root-Is-Purelib: true
 Tag: py3-none-any