-
Notifications
You must be signed in to change notification settings - Fork 10
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Merge pull request #5 from camel-ai/twitter_simu
add twitter gpt example
- Loading branch information
Showing
6 changed files
with
233 additions
and
11 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,11 @@ | ||
| # OBJECTIVE | ||
| You're a Twitter user, and I'll present you with some posts. After you see the posts, choose some actions from the following functions. | ||
| Suppose you are a real Twitter user. Please simulate real behavior. | ||
|
|
||
| - do_nothing: Most of the time, you just don't feel like reposting or liking a post, and you just want to look at it. In such cases, choose this action "do_nothing" | ||
| - repost: Repost a post. | ||
| - Arguments: "post_id" (integer) - The ID of the post to be reposted. You can `repost` when you want to spread it. | ||
| - like_post: Likes a specified post. | ||
| - Arguments: "post_id" (integer) - The ID of the tweet to be liked. You can `like` when you feel something interesting or you agree with. | ||
| - follow: Follow a user specified by 'followee_id'. You can `follow' when you respect someone, love someone, or care about someone. | ||
| - Arguments: "followee_id" (integer) - The ID of the user to be followed. |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,21 @@ | ||
| --- | ||
| data: | ||
| db_path: data/simu_db/yaml_gpt/False_Business_0.db | ||
| csv_path: data/twitter_dataset/anonymous_topic_200_1h/False_Business_0.csv | ||
| simulation: | ||
| num_timesteps: 10 | ||
| clock_factor: 60 | ||
| recsys_type: twhin-bert | ||
| model: | ||
| num_agents: 111 | ||
| model_random_seed: 42 | ||
| cfgs: | ||
| - model_type: gpt-4o-mini | ||
| num: 111 | ||
| server_url: null | ||
| model_path: null | ||
| stop_tokens: null | ||
| temperature: null | ||
| inference: | ||
| model_type: gpt-4o-mini # Name of the OpenAI model | ||
| is_openai_model: true # Whether it is an OpenAI model |
185 changes: 185 additions & 0 deletions
185
scripts/twitter_gpt_example/twitter_simulation_large.py
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,185 @@ | ||
| # =========== 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 __future__ import annotations | ||
|
|
||
| import argparse | ||
| import asyncio | ||
| import logging | ||
| import os | ||
| import random | ||
| from datetime import datetime | ||
| from pathlib import Path | ||
| from typing import Any | ||
|
|
||
| import pandas as pd | ||
| from colorama import Back | ||
| from yaml import safe_load | ||
|
|
||
| from oasis.clock.clock import Clock | ||
| from oasis.social_agent.agents_generator import generate_agents | ||
| from oasis.social_platform.channel import Channel | ||
| from oasis.social_platform.platform import Platform | ||
| from oasis.social_platform.typing import ActionType | ||
|
|
||
| social_log = logging.getLogger(name="social") | ||
| social_log.setLevel("DEBUG") | ||
|
|
||
| file_handler = logging.FileHandler("social.log") | ||
| file_handler.setLevel("DEBUG") | ||
| file_handler.setFormatter( | ||
| logging.Formatter("%(levelname)s - %(asctime)s - %(name)s - %(message)s")) | ||
| social_log.addHandler(file_handler) | ||
| stream_handler = logging.StreamHandler() | ||
| stream_handler.setLevel("DEBUG") | ||
| stream_handler.setFormatter( | ||
| logging.Formatter("%(levelname)s - %(asctime)s - %(name)s - %(message)s")) | ||
| social_log.addHandler(stream_handler) | ||
|
|
||
| parser = argparse.ArgumentParser(description="Arguments for script.") | ||
| parser.add_argument( | ||
| "--config_path", | ||
| type=str, | ||
| help="Path to the YAML config file.", | ||
| required=False, | ||
| default="", | ||
| ) | ||
|
|
||
| DATA_DIR = os.path.join( | ||
| os.path.dirname(os.path.dirname(__file__)), | ||
| "data/twitter_dataset/anonymous_topic_200_1h", | ||
| ) | ||
| DEFAULT_DB_PATH = ":memory:" | ||
| DEFAULT_CSV_PATH = os.path.join(DATA_DIR, "False_Business_0.csv") | ||
|
|
||
|
|
||
| async def running( | ||
| db_path: str | None = DEFAULT_DB_PATH, | ||
| csv_path: str | None = DEFAULT_CSV_PATH, | ||
| num_timesteps: int = 3, | ||
| clock_factor: int = 60, | ||
| recsys_type: str = "twhin-bert", | ||
| model_configs: dict[str, Any] | None = None, | ||
| inference_configs: dict[str, Any] | None = None, | ||
| action_space_file_path: str = None, | ||
| ) -> None: | ||
| db_path = DEFAULT_DB_PATH if db_path is None else db_path | ||
| csv_path = DEFAULT_CSV_PATH if csv_path is None else csv_path | ||
| if os.path.exists(db_path): | ||
| os.remove(db_path) | ||
| Path(db_path).parent.mkdir(parents=True, exist_ok=True) | ||
|
|
||
| if recsys_type == "reddit": | ||
| start_time = datetime.now() | ||
| else: | ||
| start_time = 0 | ||
| social_log.info(f"Start time: {start_time}") | ||
| clock = Clock(k=clock_factor) | ||
| twitter_channel = Channel() | ||
| infra = Platform( | ||
| db_path, | ||
| twitter_channel, | ||
| clock, | ||
| start_time, | ||
| recsys_type=recsys_type, | ||
| refresh_rec_post_count=2, | ||
| max_rec_post_len=2, | ||
| following_post_count=3, | ||
| ) | ||
| inference_channel = Channel() | ||
| twitter_task = asyncio.create_task(infra.running()) | ||
| if inference_configs["model_type"][:3] == "gpt": | ||
| is_openai_model = True | ||
|
|
||
| try: | ||
| all_topic_df = pd.read_csv("data/twitter_dataset/all_topics.csv") | ||
| if "False" in csv_path or "True" in csv_path: | ||
| if "-" not in csv_path: | ||
| topic_name = csv_path.split("/")[-1].split(".")[0] | ||
| else: | ||
| topic_name = csv_path.split("/")[-1].split(".")[0].split( | ||
| "-")[0] | ||
| source_post_time = ( | ||
| all_topic_df[all_topic_df["topic_name"] == | ||
| topic_name]["start_time"].item().split(" ")[1]) | ||
| start_hour = int(source_post_time.split(":")[0]) + float( | ||
| int(source_post_time.split(":")[1]) / 60) | ||
| except Exception: | ||
| print("No real-world data, let start_hour be 1PM") | ||
| start_hour = 13 | ||
|
|
||
| model_configs = model_configs or {} | ||
| with open(action_space_file_path, "r", encoding="utf-8") as file: | ||
| action_space = file.read() | ||
| agent_graph = await generate_agents( | ||
| agent_info_path=csv_path, | ||
| twitter_channel=twitter_channel, | ||
| inference_channel=inference_channel, | ||
| start_time=start_time, | ||
| recsys_type=recsys_type, | ||
| twitter=infra, | ||
| action_space_prompt=action_space, | ||
| is_openai_model=is_openai_model, | ||
| **model_configs, | ||
| ) | ||
| # agent_graph.visualize("initial_social_graph.png") | ||
|
|
||
| for timestep in range(1, num_timesteps + 1): | ||
| os.environ["SANDBOX_TIME"] = str(timestep * 3) | ||
| social_log.info(f"timestep:{timestep}") | ||
| db_file = db_path.split("/")[-1] | ||
| print(Back.GREEN + f"DB:{db_file} timestep:{timestep}" + Back.RESET) | ||
| # if you want to disable recsys, please comment this line | ||
| await infra.update_rec_table() | ||
|
|
||
| # 0.05 * timestep here means 3 minutes / timestep | ||
| simulation_time_hour = start_hour + 0.05 * timestep | ||
| tasks = [] | ||
| for node_id, agent in agent_graph.get_agents(): | ||
| if agent.user_info.is_controllable is False: | ||
| agent_ac_prob = random.random() | ||
| threshold = agent.user_info.profile["other_info"][ | ||
| "active_threshold"][int(simulation_time_hour % 24)] | ||
| if agent_ac_prob < threshold: | ||
| tasks.append(agent.perform_action_by_llm()) | ||
| else: | ||
| await agent.perform_action_by_hci() | ||
|
|
||
| await asyncio.gather(*tasks) | ||
| # agent_graph.visualize(f"timestep_{timestep}_social_graph.png") | ||
|
|
||
| await twitter_channel.write_to_receive_queue((None, None, ActionType.EXIT)) | ||
| await twitter_task | ||
|
|
||
|
|
||
| if __name__ == "__main__": | ||
| args = parser.parse_args() | ||
| os.environ["SANDBOX_TIME"] = str(0) | ||
| if os.path.exists(args.config_path): | ||
| with open(args.config_path, "r") as f: | ||
| cfg = safe_load(f) | ||
| data_params = cfg.get("data") | ||
| simulation_params = cfg.get("simulation") | ||
| model_configs = cfg.get("model") | ||
| inference_configs = cfg.get("inference") | ||
|
|
||
| asyncio.run( | ||
| running(**data_params, | ||
| **simulation_params, | ||
| model_configs=model_configs, | ||
| inference_configs=inference_configs, | ||
| action_space_file_path=("scripts/twitter_gpt_example/" | ||
| "action_space_prompt.txt"))) | ||
| else: | ||
| asyncio.run(running()) | ||
| social_log.info("Simulation finished.") |