Source code for camel.models.base_model

# =========== Copyright 2023 @ CAMEL-AI.org. All Rights Reserved. ===========
# Licensed under the Apache License, Version 2.0 (the “License”);
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an “AS IS” BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# =========== Copyright 2023 @ CAMEL-AI.org. All Rights Reserved. ===========
from abc import ABC, abstractmethod
from typing import Any, Dict, List, Optional, Union

from openai import Stream

from camel.messages import OpenAIMessage
from camel.types import ChatCompletion, ChatCompletionChunk, ModelType
from camel.utils import BaseTokenCounter




class BaseModelBackend(ABC):
    r"""Base class for different model backends.
    May be OpenAI API, a local LLM, a stub for unit tests, etc.
    """

    def __init__(
        self,
        model_type: ModelType,
        model_config_dict: Dict[str, Any],
        api_key: Optional[str] = None,
        url: Optional[str] = None,
        token_counter: Optional[BaseTokenCounter] = None,
    ) -> None:
        r"""Constructor for the model backend.

        Args:
            model_type (ModelType): Model for which a backend is created.
            model_config_dict (Dict[str, Any]): A config dictionary.
            api_key (Optional[str]): The API key for authenticating with the
                model service.
            url (Optional[str]): The url to the model service.
            token_counter (Optional[BaseTokenCounter]): Token counter to use
                for the model. If not provided, `OpenAITokenCounter` will
                be used.
        """
        self.model_type = model_type
        self.model_config_dict = model_config_dict
        self._api_key = api_key
        self._url = url
        self.check_model_config()
        self._token_counter = token_counter

    @property
    @abstractmethod
    def token_counter(self) -> BaseTokenCounter:
        r"""Initialize the token counter for the model backend.

        Returns:
            BaseTokenCounter: The token counter following the model's
                tokenization style.
        """
        pass



    @abstractmethod
    def run(
        self,
        messages: List[OpenAIMessage],
    ) -> Union[ChatCompletion, Stream[ChatCompletionChunk]]:
        r"""Runs the query to the backend model.

        Args:
            messages (List[OpenAIMessage]): Message list with the chat history
                in OpenAI API format.

        Returns:
            Union[ChatCompletion, Stream[ChatCompletionChunk]]:
                `ChatCompletion` in the non-stream mode, or
                `Stream[ChatCompletionChunk]` in the stream mode.
        """
        pass




    @abstractmethod
    def check_model_config(self):
        r"""Check whether the input model configuration contains unexpected
        arguments

        Raises:
            ValueError: If the model configuration dictionary contains any
                unexpected argument for this model class.
        """
        pass




    def count_tokens_from_messages(self, messages: List[OpenAIMessage]) -> int:
        r"""Count the number of tokens in the messages using the specific
        tokenizer.

        Args:
            messages (List[Dict]): message list with the chat history
                in OpenAI API format.

        Returns:
            int: Number of tokens in the messages.
        """
        return self.token_counter.count_tokens_from_messages(messages)


    @property
    def token_limit(self) -> int:
        r"""Returns the maximum token limit for a given model.

        Returns:
            int: The maximum token limit for the given model.
        """
        return (
            self.model_config_dict.get("max_tokens")
            or self.model_type.token_limit
        )

    @property
    def stream(self) -> bool:
        r"""Returns whether the model is in stream mode,
            which sends partial results each time.

        Returns:
            bool: Whether the model is in stream mode.
        """
        return False