Skip to content

armgit / camel

Public Imported from GitHub
0

🐫 CAMEL: The first and the best multi-agent framework. Finding the Scaling Law of Agents. https://www.camel-ai.org

Remove `ChatGPTVisionConfig` and refactor configs (#539)

Zecheng Zhang16cd2c113d57
diff --git a/camel/agents/chat_agent.py b/camel/agents/chat_agent.py
index 8a8713563..5e6feb5ba 100644
--- a/camel/agents/chat_agent.py
+++ b/camel/agents/chat_agent.py
@@ -19,7 +19,7 @@ from dataclasses import dataclass
from typing import TYPE_CHECKING, Any, Callable, Dict, List, Optional, Tuple
from camel.agents.base import BaseAgent
-from camel.configs import ChatGPTConfig, ChatGPTVisionConfig
+from camel.configs import ChatGPTConfig
from camel.memories import (
AgentMemory,
ChatHistoryMemory,
@@ -132,27 +132,7 @@ class ChatAgent(BaseAgent):
for func in function_list:
self.func_dict[func.get_function_name()] = func.func
- self.model_config: BaseConfig
- if self.model_type == ModelType.GPT_4_TURBO_VISION:
- if model_config is not None and not isinstance(
- model_config, ChatGPTVisionConfig
- ):
- raise ValueError(
- "Please use `ChatGPTVisionConfig` as "
- "the `model_config` when `model_type` "
- "is `GPT_4_TURBO_VISION`"
- )
- self.model_config = model_config or ChatGPTVisionConfig()
- else:
- if model_config is not None and isinstance(
- model_config, ChatGPTVisionConfig
- ):
- raise ValueError(
- "Please don't use `ChatGPTVisionConfig` as "
- "the `model_config` when `model_type` "
- "is not `GPT_4_TURBO_VISION`"
- )
- self.model_config = model_config or ChatGPTConfig()
+ self.model_config = model_config or ChatGPTConfig()
self.model_backend: BaseModelBackend = ModelFactory.create(
self.model_type, self.model_config.__dict__
diff --git a/camel/configs/__init__.py b/camel/configs/__init__.py
new file mode 100644
index 000000000..211079018
--- /dev/null
+++ b/camel/configs/__init__.py
@@ -0,0 +1,33 @@
+# =========== Copyright 2023 @ CAMEL-AI.org. All Rights Reserved. ===========
+# Licensed under the Apache License, Version 2.0 (the ā€œLicenseā€);
+# 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 ā€œAS ISā€ 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 .anthropic_config import ANTHROPIC_API_PARAMS, AnthropicConfig
+from .base_config import BaseConfig
+from .openai_config import (
+ OPENAI_API_PARAMS,
+ OPENAI_API_PARAMS_WITH_FUNCTIONS,
+ ChatGPTConfig,
+ FunctionCallingConfig,
+ OpenSourceConfig,
+)
+
+__all__ = [
+ 'BaseConfig',
+ 'ChatGPTConfig',
+ 'FunctionCallingConfig',
+ 'OPENAI_API_PARAMS',
+ 'OPENAI_API_PARAMS_WITH_FUNCTIONS',
+ 'AnthropicConfig',
+ 'ANTHROPIC_API_PARAMS',
+ 'OpenSourceConfig',
+]
diff --git a/camel/configs/anthropic_config.py b/camel/configs/anthropic_config.py
new file mode 100644
index 000000000..8c64ff1ce
--- /dev/null
+++ b/camel/configs/anthropic_config.py
@@ -0,0 +1,73 @@
+# =========== Copyright 2023 @ CAMEL-AI.org. All Rights Reserved. ===========
+# Licensed under the Apache License, Version 2.0 (the ā€œLicenseā€);
+# 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 ā€œAS ISā€ 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 __future__ import annotations
+
+from dataclasses import asdict, dataclass
+
+from anthropic._types import NOT_GIVEN, NotGiven
+
+from camel.configs.base_config import BaseConfig
+
+
+@dataclass(frozen=True)
+class AnthropicConfig(BaseConfig):
+ r"""Defines the parameters for generating chat completions using the
+ Anthropic API.
+
+ See: https://docs.anthropic.com/claude/reference/complete_post
+ Args:
+ max_tokens (int, optional): The maximum number of tokens to
+ generate before stopping. Note that Anthropic models may stop
+ before reaching this maximum. This parameter only specifies the
+ absolute maximum number of tokens to generate.
+ (default: :obj:`256`)
+ stop_sequences (List[str], optional): Sequences that will cause the
+ model to stop generating completion text. Anthropic models stop
+ on "\n\nHuman:", and may include additional built-in stop sequences
+ in the future. By providing the stop_sequences parameter, you may
+ include additional strings that will cause the model to stop
+ generating.
+ temperature (float, optional): Amount of randomness injected into the
+ response. Defaults to 1. Ranges from 0 to 1. Use temp closer to 0
+ for analytical / multiple choice, and closer to 1 for creative
+ and generative tasks.
+ (default: :obj:`1`)
+ top_p (float, optional): Use nucleus sampling. In nucleus sampling, we
+ compute the cumulative distribution over all the options for each
+ subsequent token in decreasing probability order and cut it off
+ once it reaches a particular probability specified by `top_p`.
+ You should either alter `temperature` or `top_p`,
+ but not both.
+ (default: :obj:`0.7`)
+ top_k (int, optional): Only sample from the top K options for each
+ subsequent token. Used to remove "long tail" low probability
+ responses.
+ (default: :obj:`5`)
+ metadata: An object describing metadata about the request.
+ stream (bool, optional): Whether to incrementally stream the response
+ using server-sent events.
+ (default: :obj:`False`)
+
+ """
+
+ max_tokens: int = 256
+ stop_sequences: list[str] | NotGiven = NOT_GIVEN
+ temperature: float = 1
+ top_p: float | NotGiven = NOT_GIVEN
+ top_k: int | NotGiven = NOT_GIVEN
+ metadata: NotGiven = NOT_GIVEN
+ stream: bool = False
+
+
+ANTHROPIC_API_PARAMS = {param for param in asdict(AnthropicConfig()).keys()}
diff --git a/camel/configs/base_config.py b/camel/configs/base_config.py
new file mode 100644
index 000000000..9a28a3cb6
--- /dev/null
+++ b/camel/configs/base_config.py
@@ -0,0 +1,22 @@
+# =========== Copyright 2023 @ CAMEL-AI.org. All Rights Reserved. ===========
+# Licensed under the Apache License, Version 2.0 (the ā€œLicenseā€);
+# 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 ā€œAS ISā€ 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 __future__ import annotations
+
+from abc import ABC
+from dataclasses import dataclass
+
+
+@dataclass(frozen=True)
+class BaseConfig(ABC): # noqa: B024
+ pass
diff --git a/camel/configs.py b/camel/configs/openai_config.py
similarity index 55%
rename from camel/configs.py
rename to camel/configs/openai_config.py
index 26db8183d..993f64e97 100644
--- a/camel/configs.py
+++ b/camel/configs/openai_config.py
@@ -13,21 +13,17 @@
# =========== Copyright 2023 @ CAMEL-AI.org. All Rights Reserved. ===========
from __future__ import annotations
-from abc import ABC
from dataclasses import asdict, dataclass, field
-from typing import TYPE_CHECKING, Any, Dict, List, Optional, Sequence, Union
+from typing import TYPE_CHECKING, Any, Sequence
-from anthropic._types import NOT_GIVEN, NotGiven
+from openai._types import NOT_GIVEN, NotGiven
+
+from camel.configs.base_config import BaseConfig
if TYPE_CHECKING:
from camel.functions import OpenAIFunction
-@dataclass(frozen=True)
-class BaseConfig(ABC): # noqa: B024
- pass
-
-
@dataclass(frozen=True)
class ChatGPTConfig(BaseConfig):
r"""Defines the parameters for generating chat completions using the
@@ -82,61 +78,11 @@ class ChatGPTConfig(BaseConfig):
top_p: float = 1.0
n: int = 1
stream: bool = False
- stop: Optional[Union[str, Sequence[str]]] = None
- max_tokens: Optional[int] = None
- presence_penalty: float = 0.0
- frequency_penalty: float = 0.0
- logit_bias: Dict = field(default_factory=dict)
- user: str = ""
-
-
-@dataclass(frozen=True)
-class ChatGPTVisionConfig(BaseConfig):
- r"""Defines the parameters for generating chat completions with
- vision model using the OpenAI API. The vision config here is the
- subset of the :class:`ChatGPTConfig`.
-
- 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`)
- stream (bool, optional): If True, partial message deltas will be sent
- as data-only server-sent events as they become available.
- (default: :obj:`False`)
- 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:`4096`)
- 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`)
- user (str, optional): A unique identifier representing your end-user,
- which can help OpenAI to monitor and detect abuse.
- (default: :obj:`""`)
- """
-
- temperature: float = 0.2 # openai default: 1.0
- top_p: float = 1.0
- n: int = 1
- stream: bool = False
- max_tokens: int = 4096
+ stop: str | Sequence[str] | NotGiven = NOT_GIVEN
+ max_tokens: int | NotGiven = NOT_GIVEN
presence_penalty: float = 0.0
frequency_penalty: float = 0.0
+ logit_bias: dict = field(default_factory=dict)
user: str = ""
@@ -157,15 +103,15 @@ class FunctionCallingConfig(ChatGPTConfig):
function. (default: :obj:`"auto"`)
"""
- functions: List[Dict[str, Any]] = field(default_factory=list)
- function_call: Union[Dict[str, str], str] = "auto"
+ functions: list[dict[str, Any]] = field(default_factory=list)
+ function_call: dict[str, str] | str = "auto"
@classmethod
def from_openai_function_list(
cls,
- function_list: List[OpenAIFunction],
- function_call: Union[Dict[str, str], str] = "auto",
- kwargs: Optional[Dict[str, Any]] = None,
+ function_list: list[OpenAIFunction],
+ function_call: dict[str, str] | str = "auto",
+ kwargs: dict[str, Any] | None = None,
):
r"""Class method for creating an instance given the function-related
arguments.
@@ -193,6 +139,12 @@ class FunctionCallingConfig(ChatGPTConfig):
)
+OPENAI_API_PARAMS = {param for param in asdict(ChatGPTConfig()).keys()}
+OPENAI_API_PARAMS_WITH_FUNCTIONS = {
+ param for param in asdict(FunctionCallingConfig()).keys()
+}
+
+
@dataclass(frozen=True)
class OpenSourceConfig(BaseConfig):
r"""Defines parameters for setting up open-source models and includes
@@ -210,62 +162,3 @@ class OpenSourceConfig(BaseConfig):
model_path: str
server_url: str
api_params: ChatGPTConfig = field(default_factory=ChatGPTConfig)
-
-
-OPENAI_API_PARAMS = {param for param in asdict(ChatGPTConfig()).keys()}
-OPENAI_API_PARAMS_WITH_FUNCTIONS = {
- param for param in asdict(FunctionCallingConfig()).keys()
-}
-
-
-@dataclass(frozen=True)
-class AnthropicConfig(BaseConfig):
- r"""Defines the parameters for generating chat completions using the
- Anthropic API.
-
- See: https://docs.anthropic.com/claude/reference/complete_post
- Args:
- max_tokens_to_sample (int, optional): The maximum number of tokens to
- generate before stopping. Note that Anthropic models may stop
- before reaching this maximum. This parameter only specifies the
- absolute maximum number of tokens to generate.
- (default: :obj:`256`)
- stop_sequences (List[str], optional): Sequences that will cause the
- model to stop generating completion text. Anthropic models stop
- on "\n\nHuman:", and may include additional built-in stop sequences
- in the future. By providing the stop_sequences parameter, you may
- include additional strings that will cause the model to stop
- generating.
- temperature (float, optional): Amount of randomness injected into the
- response. Defaults to 1. Ranges from 0 to 1. Use temp closer to 0
- for analytical / multiple choice, and closer to 1 for creative
- and generative tasks.
- (default: :obj:`1`)
- top_p (float, optional): Use nucleus sampling. In nucleus sampling, we
- compute the cumulative distribution over all the options for each
- subsequent token in decreasing probability order and cut it off
- once it reaches a particular probability specified by `top_p`.
- You should either alter `temperature` or `top_p`,
- but not both.
- (default: :obj:`0.7`)
- top_k (int, optional): Only sample from the top K options for each
- subsequent token. Used to remove "long tail" low probability
- responses.
- (default: :obj:`5`)
- metadata: An object describing metadata about the request.
- stream (bool, optional): Whether to incrementally stream the response
- using server-sent events.
- (default: :obj:`False`)
-
- """
-
- max_tokens: int = 256
- stop_sequences: Union[List[str], NotGiven] = NOT_GIVEN
- temperature: float = 1
- top_p: Union[float, NotGiven] = NOT_GIVEN
- top_k: Union[int, NotGiven] = NOT_GIVEN
- metadata: NotGiven = NOT_GIVEN
- stream: bool = False
-
-
-ANTHROPIC_API_PARAMS = {param for param in asdict(AnthropicConfig()).keys()}
diff --git a/camel/types/enums.py b/camel/types/enums.py
index 37c8a21b1..e85483d86 100644
--- a/camel/types/enums.py
+++ b/camel/types/enums.py
@@ -28,7 +28,6 @@ class ModelType(Enum):
GPT_4 = "gpt-4"
GPT_4_32K = "gpt-4-32k"
GPT_4_TURBO = "gpt-4-turbo"
- GPT_4_TURBO_VISION = "gpt-4-turbo"
STUB = "stub"
@@ -59,7 +58,6 @@ class ModelType(Enum):
ModelType.GPT_4,
ModelType.GPT_4_32K,
ModelType.GPT_4_TURBO,
- ModelType.GPT_4_TURBO_VISION,
}
@property
@@ -101,8 +99,6 @@ class ModelType(Enum):
return 32768
elif self is ModelType.GPT_4_TURBO:
return 128000
- elif self is ModelType.GPT_4_TURBO_VISION:
- return 128000
elif self is ModelType.STUB:
return 4096
elif self is ModelType.LLAMA_2:
diff --git a/examples/vision/object_recognition.py b/examples/vision/object_recognition.py
index dde15fb66..dbefb6dbe 100644
--- a/examples/vision/object_recognition.py
+++ b/examples/vision/object_recognition.py
@@ -25,6 +25,7 @@ parser.add_argument(
"--image_path",
type=str,
help="Path to the image for object detection.",
+ required=True,
)
@@ -42,7 +43,7 @@ def detect_image_obj(image_path: str) -> None:
)
agent = ChatAgent(
assistant_sys_msg,
- model_type=ModelType.GPT_4_TURBO_VISION,
+ model_type=ModelType.GPT_4_TURBO,
)
user_msg = BaseMessage.make_user_message(
role_name="User",
diff --git a/test/agents/test_chat_agent.py b/test/agents/test_chat_agent.py
index 5d833e86a..d2e8c3c4f 100644
--- a/test/agents/test_chat_agent.py
+++ b/test/agents/test_chat_agent.py
@@ -23,11 +23,7 @@ from PIL import Image
from camel.agents import ChatAgent
from camel.agents.chat_agent import FunctionCallingRecord
-from camel.configs import (
- ChatGPTConfig,
- ChatGPTVisionConfig,
- FunctionCallingConfig,
-)
+from camel.configs import ChatGPTConfig, FunctionCallingConfig
from camel.functions import MATH_FUNCS
from camel.generators import SystemMessageGenerator
from camel.memories import MemoryRecord
@@ -449,10 +445,10 @@ def test_chat_agent_vision():
meta_dict=None,
content="You are a help assistant.",
)
- model_config = ChatGPTVisionConfig(temperature=0, max_tokens=200)
+ model_config = ChatGPTConfig(temperature=0, max_tokens=200, stop="")
agent = ChatAgent(
system_message=system_message,
- model_type=ModelType.GPT_4_TURBO_VISION,
+ model_type=ModelType.GPT_4_TURBO,
model_config=model_config,
)
@@ -488,9 +484,9 @@ def test_chat_agent_vision():
)
],
created=123456,
- model='gpt-4-1106-vision-preview',
+ model='gpt-4-turbo-2024-04-09',
object='chat.completion',
- system_fingerprint=None,
+ system_fingerprint='fp_5d12056990',
usage=CompletionUsage(
completion_tokens=2, prompt_tokens=113, total_tokens=115
),
diff --git a/test/models/test_openai_model.py b/test/models/test_openai_model.py
index ae04b1e60..e4f5b38f5 100644
--- a/test/models/test_openai_model.py
+++ b/test/models/test_openai_model.py
@@ -29,7 +29,6 @@ from camel.utils import OpenAITokenCounter
ModelType.GPT_4,
ModelType.GPT_4_32K,
ModelType.GPT_4_TURBO,
- ModelType.GPT_4_TURBO_VISION,
],
)
def test_openai_model(model_type):