Merge pull request #17 from hookla/11-add-meeting-object

.github/workflows/style.yml

@@ -9,7 +9,7 @@ jobs:
       matrix:
         python-version: ["3.11"]
     steps:
-    - uses: actions/checkout@v3
+    - uses: actions/checkout@v4
     - name: Set up Python ${{ matrix.python-version }}
       uses: actions/setup-python@v4
       with:
@@ -20,6 +20,9 @@ jobs:
         python -m pip install --upgrade pip
         pip install poetry
         poetry install
-    - name: Analysing the code with pylint
+    - name: Analysing the code
       run: |
         make style
+    - name: Run tests
+      run: |
+        make test

dream_team_gpt/agents/agent.py

@@ -1,6 +1,6 @@
 from textwrap import dedent
+from typing import Callable
 
-from dream_team_gpt.clients.base import AIClient
 from dream_team_gpt.constants import NO_COMMENT
 
 DEFAULT_SYSTEM_PROMPT = dedent(
@@ -16,14 +16,14 @@
 class Agent:
     def __init__(
         self,
-        client: AIClient,
+        client_factory: Callable,
         name: str,
         user_prompt: str,
         system_prompt: str = DEFAULT_SYSTEM_PROMPT,
     ):
         self.name = name
 
-        self.client = client
+        self.client = client_factory()
         self.system_prompt = system_prompt
         self.user_prompt = user_prompt
         self.client.common_instructions = system_prompt

dream_team_gpt/agents/chairman.py

@@ -1,23 +1,25 @@
+from typing import Callable
+
 from loguru import logger
 
 from dream_team_gpt.agents.agent import Agent
 from dream_team_gpt.agents.sme import SME
-from dream_team_gpt.clients.base import AIClient
 
 
 class Chairman(Agent):
-    def __init__(self, client: AIClient, executives: list[SME], name: str = "Chairman"):
+    def __init__(self, client_factory: Callable, executives: list[SME], name: str = "Chairman"):
         # Construct the user_prompt string with details of the executives
         self.user_prompt = self.update_user_prompt(executives)
 
         system_prompt = f"Answer with only the name and nothing else."
 
         # Call the superclass constructor with the constructed user_prompt
-        super().__init__(client, name, self.user_prompt, system_prompt)
+        super().__init__(client_factory, name, self.user_prompt, system_prompt)
 
         self.executives = executives
 
-    def update_user_prompt(self, SMEs: list[SME]) -> str:
+    @staticmethod
+    def update_user_prompt(SMEs: list[SME]) -> str:
         frequency_info_list = []
         for sme in SMEs:
             frequency_info_list.append(
@@ -31,12 +33,10 @@ def update_user_prompt(self, SMEs: list[SME]) -> str:
             f"Participants:\n{''.join(frequency_info_list)} "
         )
 
-    def decide_if_meeting_over(self, transcript: list) -> bool:
+    def decide_if_meeting_over(self, transcript: str) -> bool:
         return False
 
-    def decide_next_speaker(self, transcript_list: list[str]) -> SME:
-        transcript = " ".join(transcript_list)
-
+    def decide_next_speaker(self, transcript: str) -> SME:
         while True:
             next_speaker = self.query_gpt(transcript).strip().rstrip(".")
             logger.info(f"Chairman called speaker: {next_speaker}")

dream_team_gpt/agents/idea_refiner.py

@@ -1,7 +1,7 @@
 from textwrap import dedent
+from typing import Callable
 
 from dream_team_gpt.agents.agent import Agent
-from dream_team_gpt.clients import AIClient
 
 REFINER_PROMPT = dedent(
     """\
@@ -14,9 +14,9 @@
 
 
 class IdeaRefiner(Agent):
-    def __init__(self, client: AIClient, name: str = "Refiner"):
+    def __init__(self, client_factory: Callable, name: str = "Refiner"):
         # Call the superclass constructor with the constructed user_prompt
-        super().__init__(client, name, REFINER_PROMPT)
+        super().__init__(client_factory, name, REFINER_PROMPT)
 
     def refine_idea(self, idea: str) -> str:
         return self.query_gpt(idea)

dream_team_gpt/agents/sme.py

@@ -1,7 +1,7 @@
 from textwrap import dedent
+from typing import Callable
 
 from dream_team_gpt.agents.agent import Agent
-from dream_team_gpt.clients.base import AIClient
 
 USER_PROMPT_TEMPLATE = dedent(
     """\
@@ -18,18 +18,17 @@
 
 
 class SME(Agent):
-    def __init__(self, client: AIClient, name: str, expertise: str, concerns: list[str]):
+    def __init__(self, client_factory: Callable, name: str, expertise: str, concerns: list[str]):
         # Construct the user_prompt string
         user_prompt = USER_PROMPT_TEMPLATE.format(
             name=name, expertise=expertise, concerns=", ".join(concerns)
         )
 
         # Call the superclass constructor with the constructed user_prompt
-        super().__init__(client, name, user_prompt)
+        super().__init__(client_factory, name, user_prompt)
         self.expertise = expertise
         self.concerns = concerns
         self.spoken_count = 0
 
-    def opinion(self, transcript_list: list[str]) -> str:
-        transcript = " ".join(transcript_list)
+    def opinion(self, transcript: str) -> str:
         return self.query_gpt(transcript)

dream_team_gpt/clients/__init__.py

@@ -1,3 +1,4 @@
 from .base import AIClient
-from .config import AIClientConfig
-from .get_client import AIClientType, GPTClient, get_ai_client
+from .config import AIClientConfig, AIClientType
+from .get_client import GPTClient, ai_client_factory, get_ai_client
+from .gpt_client import Models

dream_team_gpt/clients/config.py

@@ -1,7 +1,15 @@
 from dataclasses import dataclass
+from enum import Enum
+
+from .gpt_client import Models
+
+
+class AIClientType(str, Enum):
+    ChatGPT = "ChatGPT"
 
 
 @dataclass
 class AIClientConfig:
+    client_type: AIClientType
     api_key: str
-    model: str | None
+    model: Models | None

dream_team_gpt/clients/get_client.py

@@ -1,16 +1,50 @@
-from enum import Enum
+from dataclasses import dataclass
+from typing import Any, Callable
 
-from dream_team_gpt.clients.base import AIClient
-from dream_team_gpt.clients.config import AIClientConfig
-from dream_team_gpt.clients.gpt_client import GPTClient
+from .base import AIClient
+from .config import AIClientConfig, AIClientType
+from .gpt_client import GPTClient, Models
 
 
-class AIClientType(str, Enum):
-    ChatGPT = "ChatGPT"
-
-
-def get_ai_client(client_type: AIClientType, config: AIClientConfig) -> AIClient:
-    if client_type == AIClientType.ChatGPT:
+def get_ai_client(config: AIClientConfig) -> AIClient:
+    if config.client_type == AIClientType.ChatGPT:
         return GPTClient(config.api_key)
     else:
-        raise ValueError(f"Unknown AI client type: {client_type}")
+        raise ValueError(f"Unknown AI client type: {config.client_type}")
+
+
+def ai_client_factory(config: AIClientConfig) -> Callable[[Any], AIClient]:
+    return lambda _: get_ai_client(config)
+
+
+@dataclass
+class AIClientFactory:
+    """Callable factory for AIClient.
+
+    Usage:
+    factory = AIClientFactory(config=AIClientConfig(...))
+    Agent(factory)
+
+    or
+
+    factory.config.client_type=<AIClientType>
+    factory.config.model=<Models>
+    Agent(factory)
+
+    or update these config params in Agent on calling the factory:
+
+    factory(client_type=<AIClientType>,model=<Models>)
+
+    """
+
+    config: AIClientConfig
+
+    def __call__(
+        self, client_type: AIClientType = None, model: Models = None
+    ) -> Callable[[Any], AIClient]:
+        if client_type:
+            self.config.client_type = client_type
+        if model:
+            self.config.model = model
+
+        return lambda _: get_ai_client(self.config)

dream_team_gpt/clients/gpt_client.py

@@ -3,18 +3,22 @@
 import time
 
 from loguru import logger
+import backoff
 import openai
 
 from .base import AIClient
 
+MAX_RETRIES = 6  # Number of retries
+RETRY_DELAY = 10  # Delay between retries in seconds
+
 
 class Models(str, Enum):
     GPT3 = "gpt-3.5-turbo"
     GPT4 = "gpt-4"
 
 
 class GPTClient(AIClient):
-    def __init__(self, api_key: str, model: str = Models.GPT3.value):
+    def __init__(self, api_key: str, model: str = Models.GPT3):
         openai.api_key = api_key
         self._system_instructions = None
         self._user_prompt = None
@@ -28,63 +32,55 @@ def __init__(self, api_key: str, model: str = Models.GPT3.value):
         logger.info(f"Temperature: {self.temperature}")
 
     @property
-    def system_instructions(self):
+    def system_instructions(self) -> str:
         return self._system_instructions
 
     @system_instructions.setter
-    def system_instructions(self, value):
+    def system_instructions(self, value) -> None:
         logger.debug(f"Setting system instructions: {self._system_instructions}")
         self._system_instructions = value
 
     @property
-    def user_prompt(self):
+    def user_prompt(self) -> str:
         return self._user_prompt
 
     @user_prompt.setter
-    def user_prompt(self, value):
+    def user_prompt(self, value: str) -> None:
         logger.debug(f"Setting user prompt: {self._user_prompt}")
         self._user_prompt = value
 
     def query(self, transcript: str) -> str:
-        if self._system_instructions is None:
+        if not self._system_instructions:
             logger.error("self._system_instructions is None. Aborting the query.")
             raise RuntimeError("self._system_instructions is None, cannot proceed with query.")
-        if self._user_prompt is None:
+        if not self._user_prompt:
             logger.error("self._user_prompt is None. Aborting the query.")
             raise RuntimeError("self._user_prompt is None, cannot proceed with query.")
 
-        max_retries = 6  # Number of retries
-        retry_delay = 10  # Delay between retries in seconds
-
-        # TODO: use backoff decorator
-        for i in range(max_retries):
-            try:
-                start_time = time.time()
-                messages = [
-                    {"role": "system", "content": self._system_instructions},
-                    {"role": "user", "content": self._user_prompt},
-                    {"role": "assistant", "content": transcript},
-                ]
-                logger.info(json.dumps(messages, indent=4).replace("\\n", "\n"))
-
-                response = openai.ChatCompletion.create(
-                    model=self.model,
-                    temperature=self.temperature,
-                    messages=messages,
-                )
-
-                elapsed_time = time.time() - start_time
-
-                # Log the time taken and token usage
-                logger.info(f"GPT query took {elapsed_time:.2f} seconds")
-                logger.info(f"Tokens used in the request: {response['usage']}")
-
-                return response.choices[0].message.content.strip()
-            except openai.error.RateLimitError as e:
-                logger.warning(
-                    f"Rate limit reached. Retrying in {retry_delay} seconds. Details: {e}"
-                )
-                time.sleep(retry_delay)
-
-        logger.error(f"Max retries reached. Could not complete the GPT query.")
-        return "Error in GPT client that could not be resolved by retrying."
+        return self._query(transcript)
+
+    @backoff.on_exception(
+        backoff.constant, openai.error.RateLimitError, max_tries=MAX_RETRIES, interval=RETRY_DELAY
+    )
+    def _query(self, transcript: str) -> str:
+        start_time = time.time()
+        messages = [
+            {"role": "system", "content": self._system_instructions},
+            {"role": "user", "content": self._user_prompt},
+            {"role": "assistant", "content": transcript},
+        ]
+        logger.info(json.dumps(messages, indent=4).replace("\\n", "\n"))
+
+        response = openai.ChatCompletion.create(
+            model=self.model,
+            temperature=self.temperature,
+            messages=messages,
+        )
+
+        elapsed_time = time.time() - start_time
+
+        # Log the time taken and token usage
+        logger.info(f"GPT query took {elapsed_time:.2f} seconds")
+        logger.info(f"Tokens used in the request: {response['usage']}")
+
+        return response.choices[0].message.content.strip()