feat: Add SambaNova Cloud API support (#930)

camel/configs/__init__.py

@@ -21,8 +21,10 @@
 from .openai_config import OPENAI_API_PARAMS, ChatGPTConfig, OpenSourceConfig
 from .reka_config import REKA_API_PARAMS, RekaConfig
 from .samba_config import (
+    SAMBA_CLOUD_API_PARAMS,
     SAMBA_FAST_API_PARAMS,
     SAMBA_VERSE_API_PARAMS,
+    SambaCloudAPIConfig,
     SambaFastAPIConfig,
     SambaVerseAPIConfig,
 )
@@ -57,6 +59,8 @@
     'SAMBA_FAST_API_PARAMS',
     'SambaVerseAPIConfig',
     'SAMBA_VERSE_API_PARAMS',
+    'SambaCloudAPIConfig',
+    'SAMBA_CLOUD_API_PARAMS',
     'TogetherAIConfig',
     'TOGETHERAI_API_PARAMS',
 ]

camel/configs/samba_config.py

@@ -13,7 +13,10 @@
 # =========== Copyright 2023 @ CAMEL-AI.org. All Rights Reserved. ===========
 from __future__ import annotations
 
-from typing import Any, Dict, Optional, Union
+from typing import Any, Dict, Optional, Sequence, Union
+
+from openai._types import NOT_GIVEN, NotGiven
+from pydantic import Field
 
 from camel.configs.base_config import BaseConfig
 
@@ -107,3 +110,98 @@ def as_dict(self) -> dict[str, Any]:
 SAMBA_VERSE_API_PARAMS = {
     param for param in SambaVerseAPIConfig().model_fields.keys()
 }
+
+
+class SambaCloudAPIConfig(BaseConfig):
+    r"""Defines the parameters for generating chat completions using the
+    OpenAI API.
+
+    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:`0.2`)
+        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:`1.0`)
+        n (int, optional): How many chat completion choices to generate for
+            each input message. (default: :obj:`1`)
+        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:`False`)
+        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:`0.0`)
+        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:`0.0`)
+        logit_bias (dict, optional): Modify the likelihood of specified tokens
+            appearing in the completion. Accepts a json object that maps tokens
+            (specified by their token ID in the tokenizer) to an associated
+            bias value from :obj:`-100` to :obj:`100`. Mathematically, the bias
+            is added to the logits generated by the model prior to sampling.
+            The exact effect will vary per model, but values between:obj:` -1`
+            and :obj:`1` should decrease or increase likelihood of selection;
+            values like :obj:`-100` or :obj:`100` should result in a ban or
+            exclusive selection of the relevant token. (default: :obj:`{}`)
+        user (str, optional): A unique identifier representing your end-user,
+            which can help OpenAI to monitor and detect abuse.
+            (default: :obj:`""`)
+        tools (list[OpenAIFunction], 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: float = 0.2  # openai default: 1.0
+    top_p: float = 1.0
+    n: int = 1
+    stream: bool = False
+    stop: Union[str, Sequence[str], NotGiven] = NOT_GIVEN
+    max_tokens: Union[int, NotGiven] = NOT_GIVEN
+    presence_penalty: float = 0.0
+    response_format: Union[dict, NotGiven] = NOT_GIVEN
+    frequency_penalty: float = 0.0
+    logit_bias: dict = Field(default_factory=dict)
+    user: str = ""
+    tool_choice: Optional[Union[dict[str, str], str]] = None
+
+
+SAMBA_CLOUD_API_PARAMS = {
+    param for param in SambaCloudAPIConfig().model_fields.keys()
+}

camel/models/ollama_model.py

@@ -76,10 +76,10 @@ def _start_server(self) -> None:
             )
             print(
                 f"Ollama server started on http://localhost:11434/v1 "
-                f"for {self.model_type} model"
+                f"for {self.model_type} model."
             )
         except Exception as e:
-            print(f"Failed to start Ollama server: {e}")
+            print(f"Failed to start Ollama server: {e}.")
 
     @property
     def token_counter(self) -> BaseTokenCounter:

camel/models/samba_model.py

@@ -20,7 +20,11 @@
 import httpx
 from openai import OpenAI, Stream
 
-from camel.configs import SAMBA_FAST_API_PARAMS, SAMBA_VERSE_API_PARAMS
+from camel.configs import (
+    SAMBA_CLOUD_API_PARAMS,
+    SAMBA_FAST_API_PARAMS,
+    SAMBA_VERSE_API_PARAMS,
+)
 from camel.messages import OpenAIMessage
 from camel.types import (
     ChatCompletion,
@@ -59,9 +63,10 @@ def __init__(
                 SambaNova service. (default: :obj:`None`)
             url (Optional[str]): The url to the SambaNova service. Current
                 support SambaNova Fast API: :obj:`"https://fast-api.snova.ai/
-                v1/chat/ completions"` and SambaVerse API: :obj:`"https://
-                sambaverse.sambanova.ai/api/predict"`. (default::obj:`"https://
-                fast-api.snova.ai/v1/chat/completions"`)
+                v1/chat/ completions"`, SambaVerse API: :obj:`"https://
+                sambaverse.sambanova.ai/api/predict"` and SambaNova Cloud:
+                :obj:`"https://api.sambanova.ai/v1"`
+                (default::obj:`"https://fast-api.snova.ai/v1/chat/completions"`)
             token_counter (Optional[BaseTokenCounter]): Token counter to use
                 for the model. If not provided, `OpenAITokenCounter(ModelType.
                 GPT_4O_MINI)` will be used.
@@ -76,6 +81,14 @@ def __init__(
         self.model_config_dict = model_config_dict
         self.check_model_config()
 
+        if self._url == "https://api.sambanova.ai/v1":
+            self._client = OpenAI(
+                timeout=60,
+                max_retries=3,
+                base_url=self._url,
+                api_key=self._api_key,
+            )
+
     @property
     def token_counter(self) -> BaseTokenCounter:
         r"""Initialize the token counter for the model backend.
@@ -111,6 +124,14 @@ def check_model_config(self):
                         "input into SambaVerse API."
                     )
 
+        elif self._url == "https://api.sambanova.ai/v1":
+            for param in self.model_config_dict:
+                if param not in SAMBA_CLOUD_API_PARAMS:
+                    raise ValueError(
+                        f"Unexpected argument `{param}` is "
+                        "input into SambaCloud API."
+                    )
+
         else:
             raise ValueError(
                 f"{self._url} is not supported, please check the url to the"
@@ -141,7 +162,7 @@ def run(  # type: ignore[misc]
     def _run_streaming(  # type: ignore[misc]
         self, messages: List[OpenAIMessage]
     ) -> Stream[ChatCompletionChunk]:
-        r"""Handles streaming inference with SambaNova FastAPI.
+        r"""Handles streaming inference with SambaNova's API.
 
         Args:
             messages (List[OpenAIMessage]): A list of messages representing the
@@ -189,6 +210,15 @@ def _run_streaming(  # type: ignore[misc]
             except httpx.HTTPError as e:
                 raise RuntimeError(f"HTTP request failed: {e!s}")
 
+        # Handle SambaNova's Cloud API
+        elif self._url == "https://api.sambanova.ai/v1":
+            response = self._client.chat.completions.create(
+                messages=messages,
+                model=self.model_type,
+                **self.model_config_dict,
+            )
+            return response
+
         elif self._url == "https://sambaverse.sambanova.ai/api/predict":
             raise ValueError(
                 "https://sambaverse.sambanova.ai/api/predict doesn't support"
@@ -198,7 +228,7 @@ def _run_streaming(  # type: ignore[misc]
     def _run_non_streaming(
         self, messages: List[OpenAIMessage]
     ) -> ChatCompletion:
-        r"""Handles non-streaming inference with SambaNova FastAPI.
+        r"""Handles non-streaming inference with SambaNova's API.
 
         Args:
             messages (List[OpenAIMessage]): A list of messages representing the
@@ -251,6 +281,15 @@ def _run_non_streaming(
             except json.JSONDecodeError as e:
                 raise ValueError(f"Failed to decode JSON response: {e!s}")
 
+        # Handle SambaNova's Cloud API
+        elif self._url == "https://api.sambanova.ai/v1":
+            response = self._client.chat.completions.create(
+                messages=messages,
+                model=self.model_type,
+                **self.model_config_dict,
+            )
+            return response
+
         # Handle SambaNova's Sambaverse API
         else:
             headers = {

camel/models/vllm_model.py

@@ -80,10 +80,10 @@ def _start_server(self) -> None:
             )
             print(
                 f"vllm server started on http://localhost:8000/v1 "
-                f"for {self.model_type} model"
+                f"for {self.model_type} model."
             )
         except Exception as e:
-            print(f"Failed to start vllm server: {e}")
+            print(f"Failed to start vllm server: {e}.")
 
     @property
     def token_counter(self) -> BaseTokenCounter:

examples/models/samba_model_example.py

@@ -12,7 +12,11 @@
 # limitations under the License.
 # =========== Copyright 2023 @ CAMEL-AI.org. All Rights Reserved. ===========
 from camel.agents import ChatAgent
-from camel.configs import SambaFastAPIConfig, SambaVerseAPIConfig
+from camel.configs import (
+    SambaCloudAPIConfig,
+    SambaFastAPIConfig,
+    SambaVerseAPIConfig,
+)
 from camel.messages import BaseMessage
 from camel.models import ModelFactory
 from camel.types import ModelPlatformType
@@ -21,7 +25,7 @@
     model_platform=ModelPlatformType.SAMBA,
     model_type="llama3-70b",
     model_config_dict=SambaFastAPIConfig(max_tokens=800).as_dict(),
-    api_key="Your Fast API Key",
+    api_key="Your SambaNova Fast API Key",
     url="https://fast-api.snova.ai/v1/chat/completions",
 )
 
@@ -33,6 +37,15 @@
     url="https://sambaverse.sambanova.ai/api/predict",
 )
 
+sambacloud_api_model = ModelFactory.create(
+    model_platform=ModelPlatformType.SAMBA,
+    model_type="Meta-Llama-3.1-405B-Instruct",
+    model_config_dict=SambaCloudAPIConfig(max_tokens=800).as_dict(),
+    api_key="Your SambaNova Cloud API Key",
+    url="https://api.sambanova.ai/v1",
+)
+
+
 # Define system message
 sys_msg = BaseMessage.make_assistant_message(
     role_name="Assistant",
@@ -50,10 +63,15 @@
 camel_agent_fast_api = ChatAgent(
     system_message=sys_msg, model=samba_fast_api_model
 )
+
 camel_agent_sambaverse_api = ChatAgent(
     system_message=sys_msg, model=sambaverse_api_model
 )
 
+camel_agent_sambacloud_api = ChatAgent(
+    system_message=sys_msg, model=sambacloud_api_model
+)
+
 # Get response information
 response = camel_agent_fast_api.step(user_msg)
 print(response.msgs[0].content)
@@ -83,3 +101,20 @@
 of assistance.
 ===============================================================================
 '''
+
+# Get response information
+response = camel_agent_sambacloud_api.step(user_msg)
+print(response.msgs[0].content)
+'''
+===============================================================================
+Hello to the CAMEL AI community.  It's great to see open-source communities 
+like yours working on autonomous and communicative agents, as this field has 
+the potential to revolutionize many areas of our lives, from customer service 
+to healthcare and beyond.
+
+What specific projects or initiatives is the CAMEL AI community currently 
+working on? Are there any exciting developments or breakthroughs that you'd 
+like to share? I'm all ears (or rather, all text) and happy to learn more 
+about your work!
+===============================================================================
+'''