Merge pull request #35 from bucolucas/refactor-ai-providers

Refactor AI Providers
2024-08-18 12:56:15 -05:00
parent 3be8137efe a6bea8dd89
commit f51190e3ce
3 changed files with 186 additions and 117 deletions
@@ -0,0 +1,51 @@
+# Telegram Inference Bot Refactoring
+
+This repository contains a refactored version of the Telegram Inference Bot, which now uses a more flexible and maintainable approach for handling different AI providers.
+
+## Changes
+
+1. Introduced an abstract `AIProvider` class and concrete implementations for Anthropic and OpenAI.
+2. Refactored the main bot code to use the new AI provider classes.
+3. Implemented a factory function `create_ai_provider` for easy provider instantiation.
+4. Updated command handlers to work with the new AI provider system.
+
+## How to Use
+
+1. Set up your environment variables in a `.env` file:
+   ```
+   TELEGRAM_BOT_TOKEN=your_telegram_bot_token
+   ANTHROPIC_API_KEY=your_anthropic_api_key
+   OPENAI_API_KEY=your_openai_api_key
+   ```
+
+2. Install the required dependencies:
+   ```
+   pip install -r requirements.txt
+   ```
+
+3. Run the bot:
+   ```
+   python telegram_inference_bot.py
+   ```
+
+## Commands
+
+- `/start`: Start the bot and receive a welcome message.
+- `/clear`: Clear the conversation history and any stored images.
+- `/switch`: Switch between smart and regular models (OpenAI only).
+- `/toggle`: Toggle between Anthropic and OpenAI providers.
+- `/status`: Display the current AI provider and model being used.
+
+## Extending the Bot
+
+To add a new AI provider:
+
+1. Create a new class in `ai_providers.py` that inherits from `AIProvider`.
+2. Implement the required methods: `get_chat_response`, `format_messages`, `format_tool_calls`, etc.
+3. Update the `create_ai_provider` function to include the new provider.
+
+## Future Improvements
+
+- Implement more robust error handling and logging.
+- Add unit tests for the AI provider classes and main bot functionality.
+- Extend the README with more detailed usage instructions and examples.
@@ -0,0 +1,100 @@
+import os
+import json
+import anthropic
+from openai import OpenAI
+from abc import ABC, abstractmethod
+
+class AIProvider(ABC):
+    @abstractmethod
+    def get_chat_response(self, messages):
+        pass
+
+    @abstractmethod
+    def format_messages(self, messages):
+        pass
+
+    @abstractmethod
+    def format_tool_calls(self, response):
+        pass
+
+class AnthropicProvider(AIProvider):
+    def __init__(self):
+        self.client = anthropic.Anthropic(
+            api_key=os.environ.get("ANTHROPIC_API_KEY"),
+            default_headers={"anthropic-beta": "max-tokens-3-5-sonnet-2024-07-15"}
+        )
+        self.model = "claude-3-5-sonnet-20240620"
+
+    def get_chat_response(self, messages):
+        try:
+            response = self.client.messages.create(
+                model=self.model,
+                system=messages[0]['content'],
+                messages=self.format_messages(messages[1:]),
+                max_tokens=8192,
+                tools=self.format_tools()
+            )
+            return response
+        except Exception as e:
+            logging.error(f"An error occurred: {str(e)}")
+            return None
+
+    def format_messages(self, messages):
+        return messages
+
+    def format_tool_calls(self, response):
+        tool_calls = []
+        for message in response.content:
+            if message.type == "tool_use":
+                tool_calls.append(message)
+        return tool_calls
+
+    def format_tools(self):
+        return [
+            {
+                "name": function['name'],
+                "description": function['description'],
+                "input_schema": function['parameters'] if function['parameters'] not in [None, {}] else {"type": "object", "properties": {"param1": {"type": "string", "description": "Unnecessary"}}, "required": []}
+            }
+            for function in functions  # This assumes 'functions' is globally accessible
+        ]
+
+class OpenAIProvider(AIProvider):
+    def __init__(self, use_smart_model=True):
+        self.client = OpenAI()
+        self.use_smart_model = use_smart_model
+        self.model = self.get_model()
+
+    def get_model(self):
+        return "gpt-4o" if self.use_smart_model else "gpt-4o-mini"
+
+    def get_chat_response(self, messages):
+        response = self.client.chat.completions.create(
+            model=self.model,
+            messages=self.format_messages(messages),
+            functions=functions,  # This assumes 'functions' is globally accessible
+            function_call="auto",
+            max_tokens=self.get_max_tokens()
+        )
+        return response
+
+    def format_messages(self, messages):
+        return messages
+
+    def format_tool_calls(self, response):
+        tool_calls = []
+        assistant_message = response.choices[0].message
+        if hasattr(assistant_message, 'function_call') and assistant_message.function_call is not None:
+            tool_calls.append(assistant_message.function_call)
+        return tool_calls
+
+    def get_max_tokens(self):
+        return 4096 if self.model == "gpt-4o" else 16384
+
+def create_ai_provider(provider_name="anthropic", use_smart_model=True):
+    if provider_name.lower() == "anthropic":
+        return AnthropicProvider()
+    elif provider_name.lower() == "openai":
+        return OpenAIProvider(use_smart_model)
+    else:
+        raise ValueError(f"Unknown provider: {provider_name}")
@@ -3,34 +3,15 @@ import os
 import importlib
 import inspect
 import logging
-import anthropic
 from telegram import Update
 from telegram.ext import Application, CommandHandler, MessageHandler, filters, ContextTypes
-from openai import OpenAI
 from dotenv import load_dotenv
 from tools.base_tool import BaseTool
+from ai_providers import create_ai_provider

 # Load environment variables
 load_dotenv()

-openai_client = OpenAI()
-
-anthropic_client = anthropic.Anthropic(
-    api_key=os.environ.get("ANTHROPIC_API_KEY"),
-    default_headers={"anthropic-beta": "max-tokens-3-5-sonnet-2024-07-15"}
-)
-
-GPT_4O = "gpt-4o"
-GPT_4O_MINI = "gpt-4o-mini"
-
-model_max_tokens = {
-    GPT_4O: 4096,
-    GPT_4O_MINI: 16384
-}
-
-use_smart_model = True
-use_anthropic = True
-
 # Set up logging to console and file
 logging.basicConfig(level=logging.WARNING, handlers=[
    logging.StreamHandler(),
@@ -63,6 +44,9 @@ functions = []
 for tool in tools:
    functions.extend(tool.get_functions())

+# Initialize AI provider
+ai_provider = create_ai_provider("anthropic")
+
 async def start(update: Update, context: ContextTypes.DEFAULT_TYPE) -> None:
    logging.info("Bot started")
    await update.message.reply_text("Hello! I'm your AI assistant. How can I help you today? You can send me images and then ask questions about them.")
@@ -91,138 +75,72 @@ async def handle_message(update: Update, context: ContextTypes.DEFAULT_TYPE) ->

        messages = conversation_history[user_id]

-        response = get_chat_response(messages)
-        tool_calls = []
-        if use_anthropic:
-            for message in response.content:
-                if message.type == "tool_use":
-                    tool_calls.append(message)
-                else:
-                    messages.append({"role": "assistant", "content": response.content})
-        else:
-            assistant_message = response.choices[0].message
-            if hasattr(assistant_message, 'function_call') and assistant_message.function_call is not None:
-                tool_calls.append(assistant_message.function_call)
+        response = ai_provider.get_chat_response([{"role": "system", "content": system_prompt}] + messages)
+        tool_calls = ai_provider.format_tool_calls(response)

        toolUseCount = 0

        while len(tool_calls) > 0 and toolUseCount < 50:
-
            tool_call = tool_calls.pop(0)
            function_name = tool_call.name
            
            tool_response = call_tool(tool_call)

-            formatted_result = {}
-
-            if use_anthropic:
-                formatted_result = {"role": "user", "content":[{"type": "tool_result", "tool_use_id": tool_call.id, "content": json.dumps(tool_response)}]}
-            else:
-                formatted_result = {"role": "function", "name": function_name, "content": json.dumps(tool_response)}
-
+            formatted_result = ai_provider.format_tool_result(tool_call, tool_response)
            messages.append(formatted_result)

-            response = get_chat_response(messages)
-            assistant_message = ""
-            if use_anthropic:
-                for message in response.content:
-                    if message.type == "tool_use":
-                        tool_calls.append(message)
-                    else:
-                        messages.append({"role": "assistant", "content": response.content})
-            else:
-                assistant_message = response.choices[0].message
-                conversation_history[user_id].append({"role": "assistant", "content": assistant_message})
-                if hasattr(assistant_message, 'function_call') and assistant_message.function_call is not None:
-                    tool_calls.append(assistant_message.function_call)
-                assistant_reply = assistant_message
+            response = ai_provider.get_chat_response([{"role": "system", "content": system_prompt}] + messages)
+            tool_calls = ai_provider.format_tool_calls(response)

            toolUseCount += 1

-
-        if (toolUseCount == 0):
-            if use_anthropic:
-                assistant_reply = response.content
-            else:
-                assistant_reply = assistant_message
+        if toolUseCount == 0:
+            assistant_reply = ai_provider.format_assistant_reply(response)
            conversation_history[user_id].append({"role": "assistant", "content": assistant_reply})
        
        if len(conversation_history[user_id]) > 20:
            conversation_history[user_id] = conversation_history[user_id][-20:]

-        if use_anthropic:
-            await update.message.reply_text(messages[-1]["content"][0].text)
-        else:
-            await update.message.reply_text(assistant_reply.content)
+        await update.message.reply_text(ai_provider.get_reply_text(response))

    except Exception as e:
        logging.error(f"An error occurred: {str(e)}")
        await update.message.reply_text("Sorry, an error occurred while processing your request.")

 def call_tool(function_call):
-    function_name = function_call.name if use_anthropic else function_call.name
-    function_args =  json.dumps(function_call.input) if use_anthropic else function_call.arguments
+    function_name = function_call.name
+    function_args = json.loads(function_call.arguments if hasattr(function_call, 'arguments') else json.dumps(function_call.input))
    for tool in tools:
        if function_name in [f["name"] for f in tool.get_functions()]:
-            return tool.execute(function_name, **json.loads(function_args))
-
-def get_chat_response(messages):
-    return get_claude_response(messages) if use_anthropic else get_openai_response(messages)
-
-def get_openai_response(messages):
-    model = GPT_4O if use_smart_model else GPT_4O_MINI
-    response = openai_client.chat.completions.create(
-        model=model,
-        messages = [{"role": "system", "content": system_prompt}] + messages,
-        functions=functions,
-        function_call="auto",
-        max_tokens=model_max_tokens[model]
-    )
-    return response
-
-def get_claude_response(messages):
-    anthropic_tools = [
-            {
-                "name": function['name'],
-                "description": function['description'],
-                "input_schema": function['parameters'] if function['parameters'] not in [None, {}] else {"type": "object", "properties": {"param1": {"type": "string", "description": "Unnecessary"}}, "required": []}
-            }
-            for function in functions
-        ]
-    try:
-        response = anthropic_client.messages.create(
-            model="claude-3-5-sonnet-20240620",
-            system=system_prompt,
-            messages=messages,
-            max_tokens=8192,
-            tools=anthropic_tools
-        )
-    except Exception as e:
-        logging.error(f"An error occurred: {str(e)}")
-        return None
-        
-    return response
+            return tool.execute(function_name, **function_args)

 async def switch(update: Update, context: ContextTypes.DEFAULT_TYPE) -> None:
-    global use_smart_model
-    use_smart_model = not use_smart_model
-    model = GPT_4O if use_smart_model else GPT_4O_MINI
+    global ai_provider
+    if isinstance(ai_provider, OpenAIProvider):
+        ai_provider.use_smart_model = not ai_provider.use_smart_model
+        model = ai_provider.get_model()
        logging.info(f"Switched to model: {model}")
        await update.message.reply_text(f"Switched to model: {model}")
+    else:
+        await update.message.reply_text("Switching models is only available for OpenAI provider.")

 async def switch_providers(update: Update, context: ContextTypes.DEFAULT_TYPE) -> None:
    await clear(update, context)
-    global use_anthropic
-    use_anthropic = not use_anthropic
-    logging.info("Using Anthropic" if use_anthropic else "Using OpenAI")
-    await update.message.reply_text("Using Anthropic" if use_anthropic else "Using OpenAI")
+    global ai_provider
+    if isinstance(ai_provider, AnthropicProvider):
+        ai_provider = create_ai_provider("openai")
+        logging.info("Switched to OpenAI provider")
+        await update.message.reply_text("Switched to OpenAI provider")
+    else:
+        ai_provider = create_ai_provider("anthropic")
+        logging.info("Switched to Anthropic provider")
+        await update.message.reply_text("Switched to Anthropic provider")

 async def status(update: Update, context: ContextTypes.DEFAULT_TYPE) -> None:
-    if use_anthropic:
-        await update.message.reply_text("Currently using claude-3-5-sonnet-20240620")
+    if isinstance(ai_provider, AnthropicProvider):
+        await update.message.reply_text(f"Currently using Anthropic: {ai_provider.model}")
    else:
-        model = GPT_4O if use_smart_model else GPT_4O_MINI
-        await update.message.reply_text(f"Currently using: {model}")
+        await update.message.reply_text(f"Currently using OpenAI: {ai_provider.get_model()}")

 def main() -> None:
    # Create the Application and pass it your bot's token