Add prompt size checking and use default llm

apps/slackbot/reflection.py

@@ -1,48 +1,65 @@
-import openai
 from os import environ
-from langchain.chat_models import ChatOpenAI
+from typing import List
+
+from langchain.base_language import BaseLanguageModel
+from langchain.schema import BaseMessage
+from langchain.tools import BaseTool
+from loguru import logger
+
 from prompt_generator import PromptGenerator
 
-class Reflection():
-    def __init__(self, tools, action_list = []):
+
+class Reflection:
+    def __init__(
+        self, llm: BaseLanguageModel, tools: List[BaseTool], action_list: List = []
+    ):
+        self.llm = llm
         self.action_list = action_list
+
+        self.token_counter = self.llm.get_num_tokens
         prompt = PromptGenerator()
         self.format = prompt.response_format
         self.commands = [
-                f"{i + 1}. {prompt._generate_command_string(item)}"
-                for i, item in enumerate(tools)
-            ]
-
+            f"{i + 1}. {prompt._generate_command_string(item)}"
+            for i, item in enumerate(tools)
+        ]
+
+    def create_message_history(
+        self, messages: List[BaseMessage], max_token=2000
+    ) -> str:
+        result = ""
+        current_tokens = 0
+
+        for message in reversed(messages):
+            current_tokens = current_tokens + self.token_counter(message.content)
+            if current_tokens > max_token:
+                break
+            result = message.type + ": " + message.content + "\n" + result
+
+        return result
+
     # we will also include previous_messages in the sherpa system
     def evaluate_action(self, action, assistant_reply, task, previous_message):
         self.action_list.append(action)
         if len(self.action_list) == 1:  # first action, no previous action
             return assistant_reply
         else:
             previous_action = self.action_list[-2]
-            if previous_action == action:   # duplicate action
+            message_history = self.create_message_history(previous_message)
+            if previous_action == action:  # duplicate action
                 instruction = (
                     f"You want to solve the task: {task}."
                     f"The original reply is: {assistant_reply}"
                     f"Here is all the commands you can choose to use: {self.commands}"
-                    f"Here is previous messages: {previous_message}"
+                    f"Here is previous messages: \n{message_history}\n"
                     f"We need a new reply by changing neither command.name or command.args.query."
                     f"Make sure the new reply is different from the original reply by name or query."
                     f"You should only respond in JSON format as described below without any extra text. Do not return the TaskAction object."
                     f"Format for the new reply: {self.format}"
                     f"Ensure the response can be parsed by Python json.loads"
                     f"New reply:\n\n"
                 )
-                openai.api_key = environ.get("OPENAI_API_KEY")
-                response = openai.Completion.create(
-                            engine="text-davinci-003",
-                            prompt= instruction,
-                            temperature=0.7,
-                            max_tokens=1024,
-                            top_p=1,
-                            frequency_penalty=0,
-                            presence_penalty=0
-                        )
-                return response['choices'][0]['text']
+                logger.warning(f"tokens used: {self.token_counter(instruction)}")
+                return self.llm.predict(instruction)
             else:
-                return assistant_reply
+                return assistant_reply

apps/slackbot/task_agent.py

@@ -1,5 +1,4 @@
 import json
-from loguru import logger
 from os import environ
 from typing import List, Optional
 
@@ -11,6 +10,7 @@
 from langchain.tools.base import BaseTool
 from langchain.tools.human.tool import HumanInputRun
 from langchain.vectorstores.base import VectorStoreRetriever
+from loguru import logger
 from pydantic import ValidationError
 
 from action_planner import SelectiveActionPlanner
@@ -97,7 +97,7 @@ def run(self, task: str) -> str:
 
         # Interaction Loop
 
-        reflection = Reflection(self.tools, [])
+        reflection = Reflection(self.llm, self.tools, [])
         while True:
             # Discontinue if continuous limit is reached
             loop_count = self.loop_count
@@ -169,7 +169,9 @@ def run(self, task: str) -> str:
             action = self.output_parser.parse(assistant_reply)
             logger.debug("action:", action)
             tools = {t.name: t for t in self.tools}
-            new_reply = reflection.evaluate_action(action, assistant_reply, task, self.previous_message)
+            new_reply = reflection.evaluate_action(
+                action, assistant_reply, task, self.previous_message
+            )
             action = self.output_parser.parse(new_reply)
             logger.debug("action after reflection: ", action)