learning assistant usecase

community_usecase/learning-assistant/README.md

@@ -0,0 +1,23 @@
+# Learning Assistant
+
+This code example searches the internet for relevant learning materials depending on the user's knowledge level and devises a concrete learning roadmap all tailored to the user.
+
+## How to use
+
+1. Set up the OPENAI api key in the .env file
+
+```bash
+OPENAI_API_KEY = 'xxx'
+```
+
+2. Copy the python script to the owl/examples folder.
+
+3. Run the script
+
+```bash
+python run_gpt4o.py
+```
+
+4. You can find the entire thought process of the agent within the log file.
+
+5. Video demo - https://drive.google.com/drive/folders/1msrNNwjeZ0DKhSXCi2w1ljz_hULSusa_?usp=sharing

community_usecase/learning-assistant/run_gpt4o.py

@@ -0,0 +1,169 @@
+import os
+import logging
+import json
+
+from dotenv import load_dotenv
+from camel.models import ModelFactory
+from camel.types import ModelPlatformType
+
+from camel.toolkits import (
+    SearchToolkit,
+    BrowserToolkit,
+)
+from camel.societies import RolePlaying
+from camel.logger import set_log_level, get_logger
+
+import pathlib
+
+base_dir = pathlib.Path(__file__).parent.parent
+env_path = base_dir / "owl" / ".env"
+load_dotenv(dotenv_path=str(env_path))
+
+set_log_level(level="DEBUG")
+logger = get_logger(__name__)
+file_handler = logging.FileHandler("learning_journey.log")
+file_handler.setLevel(logging.DEBUG)
+formatter = logging.Formatter("%(asctime)s - %(name)s - %(levelname)s - %(message)s")
+file_handler.setFormatter(formatter)
+logger.addHandler(file_handler)
+
+root_logger = logging.getLogger()
+root_logger.addHandler(file_handler)
+
+
+def construct_learning_society(task: str) -> RolePlaying:
+    """Construct a society of agents for the learning journey companion.
+
+    Args:
+        task (str): The learning task description including what the user wants to learn and what they already know.
+
+    Returns:
+        RolePlaying: A configured society of agents for the learning companion.
+    """
+    models = {
+        "user": ModelFactory.create(
+            model_platform=ModelPlatformType.OPENAI_COMPATIBLE_MODEL,
+            model_type="gpt-4o",
+            api_key=os.getenv("OPENAI_API_KEY"),
+            model_config_dict={"temperature": 0.4},
+        ),
+        "assistant": ModelFactory.create(
+            model_platform=ModelPlatformType.OPENAI_COMPATIBLE_MODEL,
+            model_type="gpt-4o",
+            api_key=os.getenv("OPENAI_API_KEY"),
+            model_config_dict={"temperature": 0.4},
+        ),
+        "content_researcher": ModelFactory.create(
+            model_platform=ModelPlatformType.OPENAI_COMPATIBLE_MODEL,
+            model_type="gpt-4o",
+            api_key=os.getenv("OPENAI_API_KEY"),
+            model_config_dict={"temperature": 0.2},
+        ),
+        "planning": ModelFactory.create(
+            model_platform=ModelPlatformType.OPENAI_COMPATIBLE_MODEL,
+            model_type="gpt-4o",
+            api_key=os.getenv("OPENAI_API_KEY"),
+            model_config_dict={"temperature": 0.3},
+        ),
+    }
+
+    browser_toolkit = BrowserToolkit(
+        headless=False,
+        web_agent_model=models["content_researcher"],
+        planning_agent_model=models["planning"],
+    )
+
+    tools = [
+        *browser_toolkit.get_tools(),
+        SearchToolkit().search_duckduckgo,
+    ]
+
+    user_agent_kwargs = {
+        "model": models["user"],
+    }
+    
+    assistant_agent_kwargs = {
+        "model": models["assistant"],
+        "tools": tools,
+
+    }
+
+    task_kwargs = {
+        "task_prompt": task,
+        "with_task_specify": False,
+    }
+
+    society = RolePlaying(
+        **task_kwargs,
+        user_role_name="learner",
+        user_agent_kwargs=user_agent_kwargs,
+        assistant_role_name="learning_companion",
+        assistant_agent_kwargs=assistant_agent_kwargs,
+    )
+
+    return society
+
+
+def analyze_chat_history(chat_history):
+    """Analyze chat history and extract tool call information."""
+    print("\n============ Tool Call Analysis ============")
+    logger.info("========== Starting tool call analysis ==========")
+
+    tool_calls = []
+    for i, message in enumerate(chat_history):
+        if message.get("role") == "assistant" and "tool_calls" in message:
+            for tool_call in message.get("tool_calls", []):
+                if tool_call.get("type") == "function":
+                    function = tool_call.get("function", {})
+                    tool_info = {
+                        "call_id": tool_call.get("id"),
+                        "name": function.get("name"),
+                        "arguments": function.get("arguments"),
+                        "message_index": i,
+                    }
+                    tool_calls.append(tool_info)
+                    print(f"Tool Call: {function.get('name')} Args: {function.get('arguments')}")
+                    logger.info(f"Tool Call: {function.get('name')} Args: {function.get('arguments')}")
+
+        elif message.get("role") == "tool" and "tool_call_id" in message:
+            for tool_call in tool_calls:
+                if tool_call.get("call_id") == message.get("tool_call_id"):
+                    result = message.get("content", "")
+                    result_summary = result[:100] + "..." if len(result) > 100 else result
+                    print(f"Tool Result: {tool_call.get('name')} Return: {result_summary}")
+                    logger.info(f"Tool Result: {tool_call.get('name')} Return: {result_summary}")
+
+    print(f"Total tool calls found: {len(tool_calls)}")
+    logger.info(f"Total tool calls found: {len(tool_calls)}")
+    logger.info("========== Finished tool call analysis ==========")
+
+    with open("learning_journey_history.json", "w", encoding="utf-8") as f:
+        json.dump(chat_history, f, ensure_ascii=False, indent=2)
+
+    print("Records saved to learning_journey_history.json")
+    print("============ Analysis Complete ============\n")
+
+
+def run_learning_companion(task: str = None):
+    """Run the learning companion with the given task.
+    
+    Args:
+        task (str, optional): The learning task description. Defaults to an example task.
+    """
+    task = """
+    I want to learn about the transformers architecture in an llm.  
+    I've also taken a basic statistics course. 
+    I have about 10 hours per week to dedicate to learning. Devise a roadmap for me .
+    """
+    
+    society = construct_learning_society(task)
+    
+    from owl.utils import run_society
+    answer, chat_history, token_count = run_society(society, round_limit = 5)
+
+    # Record tool usage history
+    analyze_chat_history(chat_history)
+    print(f"\033[94mAnswer: {answer}\033[0m")
+
+if __name__ == "__main__":
+    run_learning_companion()