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

10.3 KB
Raw Blame
# ========= Copyright 2023-2026 @ 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 2023-2026 @ CAMEL-AI.org. All Rights Reserved. =========
import json
import multiprocessing
import os
from typing import Any, Dict

from camel.agents import ChatAgent, TaskSpecifyAgent
from camel.generators import SystemMessageGenerator
from camel.messages import BaseMessage
from camel.types import RoleType, TaskType
from camel.utils import download_tasks


def init_chat(
    assistant_agent: ChatAgent,
    user_agent: ChatAgent,
    user_sys_msg: BaseMessage,
    assistant_sys_msg: BaseMessage,
):
    assistant_agent.reset()
    user_agent.reset()

    # Send the system messages again to the agents using chat messages
    assistant_msg = BaseMessage.make_assistant_message(
        role_name=assistant_agent.role_name,
        content=(
            f"{user_sys_msg.content}. "
            "Now start to give me instructions one by one. "
            "Only reply with Instruction and Input."
        ),
    )

    user_msg = BaseMessage.make_user_message(
        role_name=user_agent.role_name, content=f"{assistant_sys_msg.content}"
    )
    assistant_agent.step(user_msg)

    return assistant_msg


def generate_data(
    language_idx: int,
    language_name: str,
    domain_idx: int,
    domain_name: str,
    task_idx: int,
    task_prompt: str,
) -> None:
    max_num_messages = 40

    # Remove number from task prompt
    original_task_prompt = task_prompt.replace(f"{task_idx+1}. ", "")

    task_specify_agent = TaskSpecifyAgent(
        task_type=TaskType.CODE,
    )
    specified_task_prompt = task_specify_agent.run(
        original_task_prompt,
        meta_dict=dict(domain=domain_name, language=language_name),
    )

    print(f"Original Task: {original_task_prompt}")
    print(f"Specified Task: {specified_task_prompt}")

    sys_msg_generator = SystemMessageGenerator(task_type=TaskType.CODE)
    sys_msg_meta_dicts = [
        dict(
            language=language_name,
            domain=domain_name,
            task=specified_task_prompt,
        )
    ] * 2
    assistant_sys_msg, user_sys_msg = sys_msg_generator.from_dicts(
        sys_msg_meta_dicts,
        role_tuples=[
            (f"{language_name} Programmer", RoleType.ASSISTANT),
            (f"{domain_name} User", RoleType.USER),
        ],
    )

    assistant_agent = ChatAgent(
        assistant_sys_msg, message_window_size=max_num_messages
    )
    user_agent = ChatAgent(user_sys_msg, message_window_size=max_num_messages)

    input_assistant_msg = init_chat(
        assistant_agent, user_agent, user_sys_msg, assistant_sys_msg
    )

    print("Assistant System Message: ", assistant_sys_msg.content)
    print("User System Message: ", user_sys_msg.content)
    message_counter = 0
    message_dict: Dict[str, Any] = {}

    # Append roles to the dictionary
    # We start number from 1 not 0.
    message_dict["role_1"] = f"{language_name}_{assistant_agent.role_type!s}"
    message_dict["role_2"] = f"{domain_name}_{user_agent.role_type!s}"
    message_dict["id"] = (
        f"{(language_idx+1):03}_{(domain_idx+1):03}_{(task_idx+1):03}"
    )
    message_dict["original_task"] = original_task_prompt
    message_dict["specified_task"] = specified_task_prompt

    # Threshold to terminate the conversation if no end token appears
    repeat_word_counter = 0
    repeat_word_threshold = 4
    repeat_word_list = [
        "goodbye",
        "good bye",
        "thank",
        "bye",
        "welcome",
        "language model",
    ]

    assistant_instruct_counter = 0
    assistant_instruct_threshold = 1
    assistant_instruct_word = "Instruction:"

    user_no_instruct_counter = 0
    user_no_instruct_threshold = 3
    user_no_instruct_word = "Instruction:"

    # Set max number of messages for the chat

    while message_counter < max_num_messages:
        user_response = user_agent.step(input_assistant_msg)

        # Condition 1: User terminates the chat
        if user_response.terminated:
            message_dict["termination_reason"] = (
                f"{user_agent.role_type!s}: "
                f"{user_response.info['termination_reasons'][0]}"
            )
            break

        print(f"User:\n{user_response.msg.content}\n")

        assistant_response = assistant_agent.step(user_response.msg)

        # Condition 2: Assistant terminates the chat
        if assistant_response.terminated:
            message_dict["termination_reason"] = (
                f"{assistant_agent.role_type!s}: "
                f"{assistant_response.info['termination_reasons'][0]}"
            )
            break

        print(f"Assistant:\n{assistant_response.msg.content}\n")

        # Condition 3: Break if user does not give instruction
        if user_no_instruct_word not in user_response.msg.content:
            user_no_instruct_counter += 1
            if user_no_instruct_counter == user_no_instruct_threshold:
                message_dict['termination_reason'] = (
                    "user_no_instruct_threshold"
                )
                break
        else:
            user_no_instruct_counter = 0

        # Condition 4: Break if assistant gives instruction (flipped role)
        if assistant_instruct_word in assistant_response.msg.content:
            assistant_instruct_counter += 1
            if assistant_instruct_counter == assistant_instruct_threshold:
                message_dict['termination_reason'] = (
                    "assistant_instruct_threshold"
                )
                break
        else:
            assistant_instruct_counter = 0

        # Condition 5: Repeat word observed
        for repeat_word in repeat_word_list:
            if (
                repeat_word in user_response.msg.content.lower()
                or repeat_word in assistant_response.msg.content.lower()
            ):
                repeat_word_counter += 1
                if repeat_word_counter == repeat_word_threshold:
                    message_dict['termination_reason'] = (
                        "repeat_word_threshold"
                    )
                    break
            else:
                repeat_word_counter = 0

        # Save user message
        message_counter += 1
        message_dict[f"message_{message_counter}"] = (
            user_response.msg.to_dict()
        )

        # Condition 5: End token observed
        if "<CAMEL_TASK_DONE>" in user_response.msg.content:
            message_dict['termination_reason'] = "<CAMEL_TASK_DONE>"
            break

        # Save assistant message
        message_counter += 1
        message_dict[f"message_{message_counter}"] = (
            assistant_response.msg.to_dict()
        )

        input_assistant_msg = assistant_response.msg

    message_dict["num_messages"] = message_counter

    if message_dict["num_messages"] == max_num_messages:
        message_dict["termination_reason"] = "max_num_messages"

    with open(
        f"./camel_data/code/{message_dict['id']}.json", "w"
    ) as json_file:
        json.dump(message_dict, json_file, ensure_ascii=False)


def main() -> None:
    # Define the folder path
    folder_path = "./code_data/"

    # Check if the folder already exists
    if not os.path.exists(folder_path):
        os.makedirs(folder_path)

    # Check if the folder is empty
    if not os.listdir(folder_path):
        download_tasks(task=TaskType.CODE, folder_path=folder_path)

    # Chunk for parallel jobs
    try:
        slurm_array_task_id = os.environ.get('SLURM_ARRAY_TASK_ID')
        if not isinstance(slurm_array_task_id, str):
            raise TypeError()
        array_idx = int(slurm_array_task_id)
    except (TypeError, ValueError) as e:
        print(f"Error: {e}")
        array_idx = 0

    languages_per_chunk = 4

    # Parameters for filtering the generated task string
    start_token = "1."
    num_tasks = 50

    with open("./data/code/languages.txt", "r") as f:
        languages = f.read().splitlines()

    with open("./data/code/domains.txt", "r") as f:
        domains = f.read().splitlines()

    assert (array_idx + 1) * languages_per_chunk <= len(languages)
    languages = languages[
        array_idx * languages_per_chunk : (array_idx + 1) * languages_per_chunk
    ]

    pool = multiprocessing.Pool()

    for language_idx, language_name in enumerate(languages):
        language_idx += array_idx * languages_per_chunk
        language_name = " ".join(language_name.split(" ")[1:])
        for domain_idx, domain_name in enumerate(domains):
            domain_name = " ".join(domain_name.split(" ")[1:])
            # Load the task list assigned for assistant and user roles
            with open(
                f"./code_data/tasks/{language_name}_{domain_name}.txt", "r"
            ) as f:
                tasks = f.read().splitlines()

                # Filter out the generated response to include the tasks only
                for i, task in enumerate(tasks):
                    if start_token in task:
                        tasks = tasks[i : i + num_tasks]
                        break

                # Ensure exact number of tasks is generated
                assert str(num_tasks) in tasks[-1], print(tasks)

            for task_idx, task_prompt in enumerate(tasks):
                id = (
                    f"{(language_idx+1):03}_"
                    f"{(domain_idx+1):03}_{(task_idx+1):03}"
                )
                if not os.path.exists(f"./camel_data/code/{id}.json"):
                    pool.apply_async(
                        generate_data,
                        (
                            language_idx,
                            language_name,
                            domain_idx,
                            domain_name,
                            task_idx,
                            task_prompt,
                        ),
                    )

    pool.close()
    pool.join()


if __name__ == "__main__":
    main()