mirror of
https://github.com/hibobmaster/matrix_chatgpt_bot.git
synced 2024-10-01 05:35:36 -04:00
🎉 initial commit
This commit is contained in:
commit
fe1a53780d
163
.gitignore
vendored
Normal file
163
.gitignore
vendored
Normal file
@ -0,0 +1,163 @@
|
||||
# Byte-compiled / optimized / DLL files
|
||||
__pycache__/
|
||||
*.py[cod]
|
||||
*$py.class
|
||||
|
||||
# C extensions
|
||||
*.so
|
||||
|
||||
# Distribution / packaging
|
||||
.Python
|
||||
build/
|
||||
develop-eggs/
|
||||
dist/
|
||||
downloads/
|
||||
eggs/
|
||||
.eggs/
|
||||
bin/
|
||||
lib/
|
||||
lib64/
|
||||
parts/
|
||||
sdist/
|
||||
var/
|
||||
wheels/
|
||||
share/python-wheels/
|
||||
*.egg-info/
|
||||
.installed.cfg
|
||||
*.egg
|
||||
MANIFEST
|
||||
bot
|
||||
|
||||
# PyInstaller
|
||||
# Usually these files are written by a python script from a template
|
||||
# before PyInstaller builds the exe, so as to inject date/other infos into it.
|
||||
*.manifest
|
||||
*.spec
|
||||
|
||||
# Installer logs
|
||||
pip-log.txt
|
||||
pip-delete-this-directory.txt
|
||||
|
||||
# Unit test / coverage reports
|
||||
htmlcov/
|
||||
.tox/
|
||||
.nox/
|
||||
.coverage
|
||||
.coverage.*
|
||||
.cache
|
||||
nosetests.xml
|
||||
coverage.xml
|
||||
*.cover
|
||||
*.py,cover
|
||||
.hypothesis/
|
||||
.pytest_cache/
|
||||
cover/
|
||||
|
||||
# Translations
|
||||
*.mo
|
||||
*.pot
|
||||
|
||||
# Django stuff:
|
||||
*.log
|
||||
local_settings.py
|
||||
db.sqlite3
|
||||
db.sqlite3-journal
|
||||
|
||||
# Flask stuff:
|
||||
instance/
|
||||
.webassets-cache
|
||||
|
||||
# Scrapy stuff:
|
||||
.scrapy
|
||||
|
||||
# Sphinx documentation
|
||||
docs/_build/
|
||||
|
||||
# PyBuilder
|
||||
.pybuilder/
|
||||
target/
|
||||
|
||||
# Jupyter Notebook
|
||||
.ipynb_checkpoints
|
||||
|
||||
# IPython
|
||||
profile_default/
|
||||
ipython_config.py
|
||||
|
||||
# pyenv
|
||||
# For a library or package, you might want to ignore these files since the code is
|
||||
# intended to run in multiple environments; otherwise, check them in:
|
||||
# .python-version
|
||||
|
||||
# pipenv
|
||||
# According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
|
||||
# However, in case of collaboration, if having platform-specific dependencies or dependencies
|
||||
# having no cross-platform support, pipenv may install dependencies that don't work, or not
|
||||
# install all needed dependencies.
|
||||
#Pipfile.lock
|
||||
|
||||
# poetry
|
||||
# Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
|
||||
# This is especially recommended for binary packages to ensure reproducibility, and is more
|
||||
# commonly ignored for libraries.
|
||||
# https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
|
||||
#poetry.lock
|
||||
|
||||
# pdm
|
||||
# Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
|
||||
#pdm.lock
|
||||
# pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
|
||||
# in version control.
|
||||
# https://pdm.fming.dev/#use-with-ide
|
||||
.pdm.toml
|
||||
|
||||
# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
|
||||
__pypackages__/
|
||||
|
||||
# Celery stuff
|
||||
celerybeat-schedule
|
||||
celerybeat.pid
|
||||
|
||||
# SageMath parsed files
|
||||
*.sage.py
|
||||
|
||||
# Environments
|
||||
.env
|
||||
.venv
|
||||
env/
|
||||
venv/
|
||||
ENV/
|
||||
env.bak/
|
||||
venv.bak/
|
||||
config.json
|
||||
|
||||
# Spyder project settings
|
||||
.spyderproject
|
||||
.spyproject
|
||||
|
||||
# Rope project settings
|
||||
.ropeproject
|
||||
|
||||
# mkdocs documentation
|
||||
/site
|
||||
|
||||
# mypy
|
||||
.mypy_cache/
|
||||
.dmypy.json
|
||||
dmypy.json
|
||||
|
||||
# Pyre type checker
|
||||
.pyre/
|
||||
|
||||
# pytype static type analyzer
|
||||
.pytype/
|
||||
|
||||
# Cython debug symbols
|
||||
cython_debug/
|
||||
|
||||
# PyCharm
|
||||
# JetBrains specific template is maintained in a separate JetBrains.gitignore that can
|
||||
# be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
|
||||
# and can be added to the global gitignore or merged into this file. For a more nuclear
|
||||
# option (not recommended) you can uncomment the following to ignore the entire idea folder.
|
||||
.idea/
|
39
README.md
Normal file
39
README.md
Normal file
@ -0,0 +1,39 @@
|
||||
## Introduction
|
||||
This is a simple Matrix bot that uses OpenAI's GPT API and a Chatbot to generate responses to user inputs. The bot responds to two types of prompts: `!gpt` and `!chat`, depending on the first word of the prompt.
|
||||
![demo](https://i.imgur.com/kK4rnPf.jpeg "demo")
|
||||
|
||||
## Installation and Setup
|
||||
To run this application, follow the steps below:<br>
|
||||
1. Clone the repository:
|
||||
```
|
||||
git clone https://github.com/hibobmaster/matrix_chatgpt_bot.git
|
||||
```
|
||||
2. Install the required dependencies:<br>
|
||||
```
|
||||
pip install -r requirements.txt
|
||||
```
|
||||
3. Create a new config.json file and fill it with the necessary information:<br>
|
||||
```json
|
||||
{
|
||||
"homeserver": "YOUR_HOMESERVER",
|
||||
"user_id": "YOUR_USER_ID",
|
||||
"password": "YOUR_PASSWORD",
|
||||
"device_id": "YOUR_DEVICE_ID",
|
||||
"room_id": "YOUR_ROOM_ID",
|
||||
"api_key": "YOUR_API_KEY"
|
||||
}
|
||||
```
|
||||
4. Start the bot:
|
||||
```
|
||||
python main.py
|
||||
```
|
||||
## Usage
|
||||
To interact with the bot, simply send a message to the bot in the Matrix room with one of the two prompts:<br>
|
||||
- `!gpt` To generate a response using free_endpoint API:
|
||||
```
|
||||
!gpt What is the meaning of life?
|
||||
```
|
||||
- `!chat` To chat using official api with context associated support
|
||||
```
|
||||
!chat Can you tell me a joke?
|
||||
```
|
51
ask_gpt.py
Normal file
51
ask_gpt.py
Normal file
@ -0,0 +1,51 @@
|
||||
"""
|
||||
api_endpoint from https://github.com/ayaka14732/ChatGPTAPIFree
|
||||
"""
|
||||
import aiohttp
|
||||
import asyncio
|
||||
import json
|
||||
import time
|
||||
|
||||
api_endpoint_free = "https://chatgpt-api.shn.hk/v1/"
|
||||
headers = {'Content-Type': "application/json"}
|
||||
|
||||
|
||||
async def ask(prompt: str) -> str:
|
||||
jsons = {
|
||||
"model": "gpt-3.5-turbo",
|
||||
"messages": [
|
||||
{
|
||||
"role": "user",
|
||||
"content": prompt,
|
||||
},
|
||||
],
|
||||
}
|
||||
async with aiohttp.ClientSession() as session:
|
||||
|
||||
while True:
|
||||
try:
|
||||
async with session.post(url=api_endpoint_free,
|
||||
json=jsons, headers=headers, timeout=10) as response:
|
||||
status_code = response.status
|
||||
if not status_code == 200:
|
||||
# wait 2s
|
||||
time.sleep(2)
|
||||
continue
|
||||
|
||||
resp = await response.read()
|
||||
await session.close()
|
||||
return json.loads(resp)['choices'][0]['message']['content']
|
||||
except Exception as e:
|
||||
print(e)
|
||||
pass
|
||||
|
||||
|
||||
async def test() -> None:
|
||||
resp = await ask("Hello World")
|
||||
print(resp)
|
||||
# type: str
|
||||
print(type(resp))
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
asyncio.run(test())
|
82
bot.py
Normal file
82
bot.py
Normal file
@ -0,0 +1,82 @@
|
||||
import sys
|
||||
import re
|
||||
import os
|
||||
from typing import Optional
|
||||
from nio import AsyncClient, MatrixRoom, RoomMessageText, LoginResponse, AsyncClientConfig
|
||||
from nio.store.database import SqliteStore
|
||||
from ask_gpt import ask
|
||||
from send_message import send_room_message
|
||||
from v3 import Chatbot
|
||||
|
||||
|
||||
class Bot:
|
||||
def __init__(
|
||||
self,
|
||||
homeserver: str,
|
||||
user_id: str,
|
||||
password: str,
|
||||
device_id: str,
|
||||
api_key: str = "",
|
||||
room_id: Optional[str] = '',
|
||||
):
|
||||
self.homeserver = homeserver
|
||||
self.user_id = user_id
|
||||
self.password = password
|
||||
self.device_id = device_id
|
||||
self.room_id = room_id
|
||||
self.api_key = api_key
|
||||
# initialize AsyncClient object
|
||||
self.store_path = os.getcwd()
|
||||
self.config = AsyncClientConfig(store=SqliteStore,
|
||||
store_name="bot",
|
||||
store_sync_tokens=True,
|
||||
)
|
||||
self.client = AsyncClient(self.homeserver, user=self.user_id, device_id=self.device_id,
|
||||
config=self.config, store_path=self.store_path)
|
||||
# regular expression to match keyword [!gpt {prompt}] [!chat {prompt}]
|
||||
self.gpt_prog = re.compile(r"^\s*!gpt\s*(.+)$")
|
||||
self.chat_prog = re.compile(r"^\s*!chat\s*(.+)$")
|
||||
# initialize chatbot
|
||||
self.chatbot = Chatbot(api_key=self.api_key)
|
||||
|
||||
# message_callback event
|
||||
async def message_callback(self, room: MatrixRoom, event: RoomMessageText) -> None:
|
||||
# chatgpt
|
||||
m = self.gpt_prog.match(event.body)
|
||||
if m:
|
||||
# sending typing state
|
||||
await self.client.room_typing(self.room_id)
|
||||
prompt = m.group(1)
|
||||
text = await ask(prompt)
|
||||
text = text.strip()
|
||||
await send_room_message(self.client, self.room_id, send_text=text)
|
||||
|
||||
n = self.chat_prog.match(event.body)
|
||||
if n:
|
||||
# sending typing state
|
||||
await self.client.room_typing(self.room_id)
|
||||
prompt = n.group(1)
|
||||
try:
|
||||
text = self.chatbot.ask(prompt).strip()
|
||||
await send_room_message(self.client, self.room_id, send_text=text)
|
||||
except Exception as e:
|
||||
print(f"Error: {e}")
|
||||
pass
|
||||
|
||||
# print info to console
|
||||
# print(
|
||||
# f"Message received in room {room.display_name}\n"
|
||||
# f"{room.user_name(event.sender)} | {event.body}"
|
||||
# )
|
||||
|
||||
# bot login
|
||||
async def login(self) -> None:
|
||||
resp = await self.client.login(password=self.password)
|
||||
if not isinstance(resp, LoginResponse):
|
||||
print(f"Login Failed: {resp}")
|
||||
sys.exit(1)
|
||||
|
||||
# sync messages in the room
|
||||
async def sync_forever(self, timeout=30000):
|
||||
self.client.add_event_callback(self.message_callback, RoomMessageText)
|
||||
await self.client.sync_forever(timeout=timeout, full_state=True)
|
8
config.json.sample
Normal file
8
config.json.sample
Normal file
@ -0,0 +1,8 @@
|
||||
{
|
||||
"homeserver": "https://matrix.qqs.tw",
|
||||
"user_id": "@lullap:xxxxx.org",
|
||||
"password": "xxxxxxxxxxxxxxxxxx",
|
||||
"device_id": "ECYEOKVPLG",
|
||||
"room_id": "!FYCmBSkCRUNvZDBaDQ:matrix.qqs.tw",
|
||||
"api_key": "xxxxxxxxxxxxxxxxxxxxxxxx"
|
||||
}
|
30
main.py
Normal file
30
main.py
Normal file
@ -0,0 +1,30 @@
|
||||
#!/usr/bin/env python3
|
||||
import json
|
||||
import asyncio
|
||||
from bot import Bot
|
||||
import sys
|
||||
|
||||
|
||||
async def main():
|
||||
fp = open('config.json', 'r')
|
||||
config = json.load(fp)
|
||||
matrix_bot = Bot(homeserver=config['homeserver'],
|
||||
user_id=config['user_id'],
|
||||
password=config['password'],
|
||||
device_id=config['device_id'],
|
||||
room_id=config['room_id'],
|
||||
api_key=config['api_key'])
|
||||
await matrix_bot.login()
|
||||
await matrix_bot.sync_forever()
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
try:
|
||||
loop = asyncio.new_event_loop()
|
||||
asyncio.set_event_loop(loop)
|
||||
asyncio.run(main())
|
||||
except KeyboardInterrupt:
|
||||
loop.close()
|
||||
sys.exit(0)
|
||||
|
||||
# asyncio.get_event_loop().run_until_complete(main())
|
39
requirements.txt
Normal file
39
requirements.txt
Normal file
@ -0,0 +1,39 @@
|
||||
aiofiles==0.6.0
|
||||
aiohttp==3.8.4
|
||||
aiohttp-socks==0.7.1
|
||||
aiosignal==1.3.1
|
||||
async-timeout==4.0.2
|
||||
atomicwrites==1.4.1
|
||||
attrs==22.2.0
|
||||
blobfile==2.0.1
|
||||
cachetools==4.2.4
|
||||
certifi==2022.12.7
|
||||
cffi==1.15.1
|
||||
charset-normalizer==3.0.1
|
||||
filelock==3.9.0
|
||||
frozenlist==1.3.3
|
||||
future==0.18.3
|
||||
h11==0.12.0
|
||||
h2==4.1.0
|
||||
hpack==4.0.0
|
||||
hyperframe==6.0.1
|
||||
idna==3.4
|
||||
jsonschema==4.17.3
|
||||
Logbook==1.5.3
|
||||
lxml==4.9.2
|
||||
matrix-nio==0.20.1
|
||||
multidict==6.0.4
|
||||
peewee==3.16.0
|
||||
pycparser==2.21
|
||||
pycryptodome==3.17
|
||||
pycryptodomex==3.17
|
||||
pyrsistent==0.19.3
|
||||
python-olm==3.1.3
|
||||
python-socks==2.1.1
|
||||
regex==2022.10.31
|
||||
requests==2.28.2
|
||||
tiktoken==0.3.0
|
||||
unpaddedbase64==2.1.0
|
||||
urllib3==1.26.14
|
||||
wcwidth==0.2.6
|
||||
yarl==1.8.2
|
12
send_message.py
Normal file
12
send_message.py
Normal file
@ -0,0 +1,12 @@
|
||||
from nio import AsyncClient
|
||||
|
||||
|
||||
async def send_room_message(client: AsyncClient,
|
||||
room_id: str,
|
||||
send_text: str) -> None:
|
||||
await client.room_send(
|
||||
room_id,
|
||||
message_type="m.room.message",
|
||||
content={"msgtype": "m.text", "body": f"{send_text}"},
|
||||
)
|
||||
await client.room_typing(room_id, typing_state=False)
|
20
test.py
Normal file
20
test.py
Normal file
@ -0,0 +1,20 @@
|
||||
from v3 import Chatbot
|
||||
import asyncio
|
||||
from ask_gpt import ask
|
||||
import json
|
||||
fp = open("config.json", "r")
|
||||
config = json.load(fp)
|
||||
|
||||
|
||||
def test_v3(prompt: str):
|
||||
bot = Chatbot(api_key=config['api_key'])
|
||||
resp = bot.ask(prompt=prompt)
|
||||
print(resp)
|
||||
|
||||
|
||||
async def test_ask(prompt: str):
|
||||
print(await ask(prompt=prompt))
|
||||
|
||||
if __name__ == "__main__":
|
||||
test_v3("Hello World")
|
||||
asyncio.run(test_ask("Hello World"))
|
192
v3.py
Normal file
192
v3.py
Normal file
@ -0,0 +1,192 @@
|
||||
"""
|
||||
A simple wrapper for the official ChatGPT API
|
||||
https://github.com/acheong08/ChatGPT/blob/main/src/revChatGPT/V3.py
|
||||
"""
|
||||
|
||||
import json
|
||||
import os
|
||||
|
||||
|
||||
import requests
|
||||
import tiktoken
|
||||
|
||||
|
||||
ENGINE = os.environ.get("GPT_ENGINE") or "gpt-3.5-turbo"
|
||||
ENCODER = tiktoken.get_encoding("gpt2")
|
||||
|
||||
|
||||
class Chatbot:
|
||||
"""
|
||||
Official ChatGPT API
|
||||
"""
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
api_key: str = None,
|
||||
engine: str = None,
|
||||
proxy: str = None,
|
||||
max_tokens: int = 4096,
|
||||
temperature: float = 0.5,
|
||||
top_p: float = 1.0,
|
||||
reply_count: int = 1,
|
||||
system_prompt: str = "You are ChatGPT, a large language model trained by OpenAI. Respond conversationally",
|
||||
) -> None:
|
||||
"""
|
||||
Initialize Chatbot with API key (from https://platform.openai.com/account/api-keys)
|
||||
"""
|
||||
self.engine = engine or ENGINE
|
||||
self.session = requests.Session()
|
||||
self.api_key = api_key
|
||||
# self.proxy = proxy
|
||||
# if self.proxy:
|
||||
# proxies = {
|
||||
# "http": self.proxy,
|
||||
# "https": self.proxy,
|
||||
# }
|
||||
# self.session.proxies = proxies
|
||||
self.conversation: dict = {
|
||||
"default": [
|
||||
{
|
||||
"role": "system",
|
||||
"content": system_prompt,
|
||||
},
|
||||
],
|
||||
}
|
||||
self.system_prompt = system_prompt
|
||||
self.max_tokens = max_tokens
|
||||
self.temperature = temperature
|
||||
self.top_p = top_p
|
||||
self.reply_count = reply_count
|
||||
|
||||
initial_conversation = "\n".join(
|
||||
[x["content"] for x in self.conversation["default"]],
|
||||
)
|
||||
if len(ENCODER.encode(initial_conversation)) > self.max_tokens:
|
||||
raise Exception("System prompt is too long")
|
||||
|
||||
def add_to_conversation(
|
||||
self, message: str, role: str, convo_id: str = "default"
|
||||
) -> None:
|
||||
"""
|
||||
Add a message to the conversation
|
||||
"""
|
||||
self.conversation[convo_id].append({"role": role, "content": message})
|
||||
|
||||
def __truncate_conversation(self, convo_id: str = "default") -> None:
|
||||
"""
|
||||
Truncate the conversation
|
||||
"""
|
||||
while True:
|
||||
full_conversation = "".join(
|
||||
message["role"] + ": " + message["content"] + "\n"
|
||||
for message in self.conversation[convo_id]
|
||||
)
|
||||
if (
|
||||
len(ENCODER.encode(full_conversation)) > self.max_tokens
|
||||
and len(self.conversation[convo_id]) > 1
|
||||
):
|
||||
# Don't remove the first message
|
||||
self.conversation[convo_id].pop(1)
|
||||
else:
|
||||
break
|
||||
|
||||
def get_max_tokens(self, convo_id: str) -> int:
|
||||
"""
|
||||
Get max tokens
|
||||
"""
|
||||
full_conversation = "".join(
|
||||
message["role"] + ": " + message["content"] + "\n"
|
||||
for message in self.conversation[convo_id]
|
||||
)
|
||||
return 4000 - len(ENCODER.encode(full_conversation))
|
||||
|
||||
def ask_stream(
|
||||
self,
|
||||
prompt: str,
|
||||
role: str = "user",
|
||||
convo_id: str = "default",
|
||||
**kwargs,
|
||||
) -> str:
|
||||
"""
|
||||
Ask a question
|
||||
"""
|
||||
# Make conversation if it doesn't exist
|
||||
if convo_id not in self.conversation:
|
||||
self.reset(convo_id=convo_id, system_prompt=self.system_prompt)
|
||||
self.add_to_conversation(prompt, "user", convo_id=convo_id)
|
||||
self.__truncate_conversation(convo_id=convo_id)
|
||||
# Get response
|
||||
response = self.session.post(
|
||||
"https://api.openai.com/v1/chat/completions",
|
||||
headers={"Authorization": f"Bearer {kwargs.get('api_key', self.api_key)}"},
|
||||
json={
|
||||
"model": self.engine,
|
||||
"messages": self.conversation[convo_id],
|
||||
"stream": True,
|
||||
# kwargs
|
||||
"temperature": kwargs.get("temperature", self.temperature),
|
||||
"top_p": kwargs.get("top_p", self.top_p),
|
||||
"n": kwargs.get("n", self.reply_count),
|
||||
"user": role,
|
||||
# "max_tokens": self.get_max_tokens(convo_id=convo_id),
|
||||
},
|
||||
stream=True,
|
||||
)
|
||||
if response.status_code != 200:
|
||||
raise Exception(
|
||||
f"Error: {response.status_code} {response.reason} {response.text}",
|
||||
)
|
||||
response_role: str = None
|
||||
full_response: str = ""
|
||||
for line in response.iter_lines():
|
||||
if not line:
|
||||
continue
|
||||
# Remove "data: "
|
||||
line = line.decode("utf-8")[6:]
|
||||
if line == "[DONE]":
|
||||
break
|
||||
resp: dict = json.loads(line)
|
||||
choices = resp.get("choices")
|
||||
if not choices:
|
||||
continue
|
||||
delta = choices[0].get("delta")
|
||||
if not delta:
|
||||
continue
|
||||
if "role" in delta:
|
||||
response_role = delta["role"]
|
||||
if "content" in delta:
|
||||
content = delta["content"]
|
||||
full_response += content
|
||||
yield content
|
||||
self.add_to_conversation(full_response, response_role, convo_id=convo_id)
|
||||
|
||||
def ask(
|
||||
self, prompt: str, role: str = "user", convo_id: str = "default", **kwargs
|
||||
) -> str:
|
||||
"""
|
||||
Non-streaming ask
|
||||
"""
|
||||
response = self.ask_stream(
|
||||
prompt=prompt,
|
||||
role=role,
|
||||
convo_id=convo_id,
|
||||
**kwargs,
|
||||
)
|
||||
full_response: str = "".join(response)
|
||||
return full_response
|
||||
|
||||
def rollback(self, n: int = 1, convo_id: str = "default") -> None:
|
||||
"""
|
||||
Rollback the conversation
|
||||
"""
|
||||
for _ in range(n):
|
||||
self.conversation[convo_id].pop()
|
||||
|
||||
def reset(self, convo_id: str = "default", system_prompt: str = None) -> None:
|
||||
"""
|
||||
Reset the conversation
|
||||
"""
|
||||
self.conversation[convo_id] = [
|
||||
{"role": "system", "content": system_prompt or self.system_prompt},
|
||||
]
|
||||
|
Loading…
Reference in New Issue
Block a user