upstage platform support

camel/configs/__init__.py

@@ -43,6 +43,7 @@
 from .sglang_config import SGLANG_API_PARAMS, SGLangConfig
 from .siliconflow_config import SILICONFLOW_API_PARAMS, SiliconFlowConfig
 from .togetherai_config import TOGETHERAI_API_PARAMS, TogetherAIConfig
+from .upstage_config import UPSTAGE_API_PARAMS, UpstageConfig
 from .vllm_config import VLLM_API_PARAMS, VLLMConfig
 from .watsonx_config import WATSONX_API_PARAMS, WatsonXConfig
 from .yi_config import YI_API_PARAMS, YiConfig
@@ -108,6 +109,8 @@
     'AIML_API_PARAMS',
     'OpenRouterConfig',
     'OPENROUTER_API_PARAMS',
+    'UpstageConfig',
+    'UPSTAGE_API_PARAMS',
     'LMSTUDIO_API_PARAMS',
     'LMStudioConfig',
     'WatsonXConfig',

camel/configs/upstage_config.py

@@ -0,0 +1,103 @@
+# ========= Copyright 2023-2024 @ 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 2023-2024 @ CAMEL-AI.org. All Rights Reserved. =========
+from __future__ import annotations
+
+from typing import Optional, Sequence, Union
+
+from camel.configs.base_config import BaseConfig
+
+
+class UpstageConfig(BaseConfig):
+    r"""Defines the parameters for generating chat completions using OpenAI
+    compatibility.
+
+    Reference: https://console.upstage.ai/api/chat
+
+    Args:
+        temperature (float, optional): Sampling temperature to use, between
+            :obj:`0` and :obj:`2`. Higher values make the output more random,
+            while lower values make it more focused and deterministic.
+            (default: :obj:`None`)
+        top_p (float, optional): An alternative to sampling with temperature,
+            called nucleus sampling, where the model considers the results of
+            the tokens with top_p probability mass. So :obj:`0.1` means only
+            the tokens comprising the top 10% probability mass are considered.
+            (default: :obj:`None`)
+        n (int, optional): How many chat completion choices to generate for
+            each input message. (default: :obj:`None`)
+        response_format (object, optional): An object specifying the format
+            that the model must output. Compatible with GPT-4 Turbo and all
+            GPT-3.5 Turbo models newer than gpt-3.5-turbo-1106. Setting to
+            {"type": "json_object"} enables JSON mode, which guarantees the
+            message the model generates is valid JSON. Important: when using
+            JSON mode, you must also instruct the model to produce JSON
+            yourself via a system or user message. Without this, the model
+            may generate an unending stream of whitespace until the generation
+            reaches the token limit, resulting in a long-running and seemingly
+            "stuck" request. Also note that the message content may be
+            partially cut off if finish_reason="length", which indicates the
+            generation exceeded max_tokens or the conversation exceeded the
+            max context length.
+        stream (bool, optional): If True, partial message deltas will be sent
+            as data-only server-sent events as they become available.
+            (default: :obj:`None`)
+        stop (str or list, optional): Up to :obj:`4` sequences where the API
+            will stop generating further tokens. (default: :obj:`None`)
+        max_tokens (int, optional): The maximum number of tokens to generate
+            in the chat completion. The total length of input tokens and
+            generated tokens is limited by the model's context length.
+            (default: :obj:`None`)
+        presence_penalty (float, optional): Number between :obj:`-2.0` and
+            :obj:`2.0`. Positive values penalize new tokens based on whether
+            they appear in the text so far, increasing the model's likelihood
+            to talk about new topics. See more information about frequency and
+            presence penalties. (default: :obj:`None`)
+        frequency_penalty (float, optional): Number between :obj:`-2.0` and
+            :obj:`2.0`. Positive values penalize new tokens based on their
+            existing frequency in the text so far, decreasing the model's
+            likelihood to repeat the same line verbatim. See more information
+            about frequency and presence penalties. (default: :obj:`None`)
+        user (str, optional): A unique identifier representing your end-user,
+            which can help OpenAI to monitor and detect abuse.
+            (default: :obj:`None`)
+        tools (list[FunctionTool], optional): A list of tools the model may
+            call. Currently, only functions are supported as a tool. Use this
+            to provide a list of functions the model may generate JSON inputs
+            for. A max of 128 functions are supported.
+        tool_choice (Union[dict[str, str], str], optional): Controls which (if
+            any) tool is called by the model. :obj:`"none"` means the model
+            will not call any tool and instead generates a message.
+            :obj:`"auto"` means the model can pick between generating a
+            message or calling one or more tools.  :obj:`"required"` means the
+            model must call one or more tools. Specifying a particular tool
+            via {"type": "function", "function": {"name": "my_function"}}
+            forces the model to call that tool. :obj:`"none"` is the default
+            when no tools are present. :obj:`"auto"` is the default if tools
+            are present.
+    """
+
+    temperature: Optional[float] = None
+    top_p: Optional[float] = None
+    n: Optional[int] = None
+    stream: Optional[bool] = None
+    stop: Optional[Union[str, Sequence[str]]] = None
+    max_tokens: Optional[int] = None
+    presence_penalty: Optional[float] = None
+    response_format: Optional[dict] = None
+    frequency_penalty: Optional[float] = None
+    user: Optional[str] = None
+    tool_choice: Optional[Union[dict[str, str], str]] = None
+
+
+UPSTAGE_API_PARAMS = {param for param in UpstageConfig.model_fields.keys()}

camel/models/__init__.py

@@ -47,6 +47,7 @@
 from .siliconflow_model import SiliconFlowModel
 from .stub_model import StubModel
 from .togetherai_model import TogetherAIModel
+from .upstage_model import UpstageModel
 from .vllm_model import VLLMModel
 from .volcano_model import VolcanoModel
 from .watsonx_model import WatsonXModel
@@ -94,6 +95,7 @@
     'BaseAudioModel',
     'SiliconFlowModel',
     'VolcanoModel',
+    'UpstageModel',
     'LMStudioModel',
     'WatsonXModel',
 ]

camel/models/model_factory.py

@@ -46,6 +46,7 @@
 from camel.models.siliconflow_model import SiliconFlowModel
 from camel.models.stub_model import StubModel
 from camel.models.togetherai_model import TogetherAIModel
+from camel.models.upstage_model import UpstageModel
 from camel.models.vllm_model import VLLMModel
 from camel.models.volcano_model import VolcanoModel
 from camel.models.watsonx_model import WatsonXModel
@@ -163,6 +164,8 @@ def create(
             model_class = LMStudioModel
         elif model_platform.is_openrouter and model_type.is_openrouter:
             model_class = OpenRouterModel
+        elif model_platform.is_upstage and model_type.is_upstage:
+            model_class = UpstageModel
         elif model_platform.is_zhipuai and model_type.is_zhipuai:
             model_class = ZhipuAIModel
         elif model_platform.is_gemini and model_type.is_gemini:

camel/models/upstage_model.py

@@ -0,0 +1,210 @@
+# ========= Copyright 2023-2024 @ 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 2023-2024 @ CAMEL-AI.org. All Rights Reserved. =========
+import os
+from typing import Any, Dict, List, Optional, Type, Union
+
+from openai import AsyncOpenAI, AsyncStream, OpenAI, Stream
+from pydantic import BaseModel
+
+from camel.configs import UPSTAGE_API_PARAMS, UpstageConfig
+from camel.messages import OpenAIMessage
+from camel.models import BaseModelBackend
+from camel.models._utils import try_modify_message_with_format
+from camel.types import (
+    ChatCompletion,
+    ChatCompletionChunk,
+    ModelType,
+)
+from camel.utils import (
+    BaseTokenCounter,
+    OpenAITokenCounter,
+    api_keys_required,
+)
+
+
+class UpstageModel(BaseModelBackend):
+    r"""LLM API served by Upstage in a unified BaseModelBackend interface.
+
+    Args:
+        model_type (Union[ModelType, str]): Model for which a backend is
+            created.
+        model_config_dict (Optional[Dict[str, Any]], optional): A dictionary
+            that will be fed into:obj:`openai.ChatCompletion.create()`.
+            If:obj:`None`, :obj:`UpstageConfig().as_dict()` will be used.
+            (default: :obj:`None`)
+        api_key (Optional[str], optional): The API key for authenticating
+            with the Upstage service. (default: :obj:`None`).
+        url (Optional[str], optional): The url to the Upstage service.
+            (default: :obj:`None`)
+        token_counter (Optional[BaseTokenCounter], optional): Token counter to
+            use for the model. If not provided, :obj:`OpenAITokenCounter(
+            ModelType.GPT_4O_MINI)` will be used.
+            (default: :obj:`None`)
+        timeout (Optional[float], optional): The timeout value in seconds for
+            API calls. If not provided, will fall back to the MODEL_TIMEOUT
+            environment variable or default to 180 seconds.
+            (default: :obj:`None`)
+    """
+
+    @api_keys_required([("api_key", "UPSTAGE_API_KEY")])
+    def __init__(
+        self,
+        model_type: Union[ModelType, str],
+        model_config_dict: Optional[Dict[str, Any]] = None,
+        api_key: Optional[str] = None,
+        url: Optional[str] = None,
+        token_counter: Optional[BaseTokenCounter] = None,
+        timeout: Optional[float] = None,
+    ) -> None:
+        if model_config_dict is None:
+            model_config_dict = UpstageConfig().as_dict()
+        api_key = api_key or os.environ.get("UPSTAGE_API_KEY")
+        url = url or os.environ.get(
+            "UPSTAGE_API_BASE_URL", "https://api.upstage.ai/v1"
+        )
+        timeout = timeout or float(os.environ.get("MODEL_TIMEOUT", 180))
+        super().__init__(
+            model_type, model_config_dict, api_key, url, token_counter, timeout
+        )
+        self._client = OpenAI(
+            timeout=self._timeout,
+            max_retries=3,
+            api_key=self._api_key,
+            base_url=self._url,
+        )
+        self._async_client = AsyncOpenAI(
+            timeout=self._timeout,
+            max_retries=3,
+            api_key=self._api_key,
+            base_url=self._url,
+        )
+
+    @property
+    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.
+        """
+        if not self._token_counter:
+            self._token_counter = OpenAITokenCounter(ModelType.GPT_4O_MINI)
+        return self._token_counter
+
+    def _prepare_request(
+        self,
+        messages: List[OpenAIMessage],
+        response_format: Optional[Type[BaseModel]] = None,
+        tools: Optional[List[Dict[str, Any]]] = None,
+    ) -> Dict[str, Any]:
+        request_config = self.model_config_dict.copy()
+        if tools:
+            request_config["tools"] = tools
+        elif response_format:
+            try_modify_message_with_format(messages[-1], response_format)
+            request_config["response_format"] = {"type": "json_object"}
+
+        return request_config
+
+    def _run(
+        self,
+        messages: List[OpenAIMessage],
+        response_format: Optional[type[BaseModel]] = None,
+        tools: Optional[List[Dict[str, Any]]] = None,
+    ) -> Union[ChatCompletion, Stream[ChatCompletionChunk]]:
+        r"""Runs inference of Upstage chat completion.
+
+        Args:
+            messages (List[OpenAIMessage]): Message list with the chat history
+                in OpenAI API format.
+            response_format (Optional[Type[BaseModel]]): The format of the
+                response.
+            tools (Optional[List[Dict[str, Any]]]): The schema of the tools to
+                use for the request.
+
+        Returns:
+            Union[ChatCompletion, Stream[ChatCompletionChunk]]:
+                `ChatCompletion` in the non-stream mode, or
+                `Stream[ChatCompletionChunk]` in the stream mode.
+        """
+        request_config = self._prepare_request(
+            messages, response_format, tools
+        )
+
+        response = self._client.chat.completions.create(
+            messages=messages,
+            model=self.model_type,
+            **request_config,
+        )
+
+        return response
+
+    async def _arun(
+        self,
+        messages: List[OpenAIMessage],
+        response_format: Optional[type[BaseModel]] = None,
+        tools: Optional[List[Dict[str, Any]]] = None,
+    ) -> Union[ChatCompletion, AsyncStream[ChatCompletionChunk]]:
+        r"""Runs inference of Upstage chat completion asynchronously.
+
+        Args:
+            messages (List[OpenAIMessage]): Message list with the chat history
+                in OpenAI API format.
+            response_format (Optional[Type[BaseModel]]): The format of the
+                response.
+            tools (Optional[List[Dict[str, Any]]]): The schema of the tools to
+                use for the request.
+
+        Returns:
+            Union[ChatCompletion, AsyncStream[ChatCompletionChunk]]:
+                `ChatCompletion` in the non-stream mode, or
+                `AsyncStream[ChatCompletionChunk]` in the stream mode.
+        """
+        request_config = self._prepare_request(
+            messages, response_format, tools
+        )
+
+        response = await self._async_client.chat.completions.create(
+            messages=messages,
+            model=self.model_type,
+            **request_config,
+        )
+
+        return response
+
+    def check_model_config(self):
+        r"""Check whether the model configuration contains any unexpected
+        arguments to Upstage API. But Upstage API does not have any additional
+        arguments to check.
+
+        Raises:
+            ValueError: If the model configuration dictionary contains any
+                unexpected arguments to Upstage API.
+        """
+        for param in self.model_config_dict:
+            if param not in UPSTAGE_API_PARAMS:
+                raise ValueError(
+                    f"Unexpected argument `{param}` is "
+                    "input into Upstage model backend."
+                )
+
+    @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 self.model_config_dict.get("stream", False)