mirror of
https://github.com/zhayujie/bot-on-anything.git
synced 2026-01-19 01:21:06 +08:00
feat: support terminal channel
This commit is contained in:
47
README.md
47
README.md
@@ -8,7 +8,7 @@
|
||||
|
||||
**应用:**
|
||||
|
||||
- [ ] 终端
|
||||
- [x] 终端
|
||||
- [ ] Web
|
||||
- [x] 个人微信
|
||||
- [x] 公众号 (个人/企业)
|
||||
@@ -36,13 +36,13 @@ cd bot-on-anything/
|
||||
|
||||
### 2.配置说明
|
||||
|
||||
核心配置文件为 `config.json`,项目中提供了模板文件 `config-template.json` ,可以从模板复制生成最终生效的 `config.json` 文件:
|
||||
核心配置文件为 `config.json`,在项目中提供了模板文件 `config-template.json` ,可以从模板复制生成最终生效的 `config.json` 文件:
|
||||
|
||||
```bash
|
||||
cp config-template.json config.json
|
||||
```
|
||||
|
||||
完整的配置文件结构如下:
|
||||
每一个模型和应用都有自己的配置块,最终组成完整的配置文件,整体结构如下:
|
||||
|
||||
```bash
|
||||
{
|
||||
@@ -63,9 +63,9 @@ cp config-template.json config.json
|
||||
}
|
||||
}
|
||||
```
|
||||
配置文件在最外层分成 `model` 和 `channel` 两部分,model 部分为模型配置,其中的 `type` 指定了选用哪个模型;`channel` 部分包含了应用渠道的配置,`type` 字段指定了接入哪个应用,同时下方对应的配置块也会生效。
|
||||
配置文件在最外层分成 `model` 和 `channel` 两部分,model部分为模型配置,其中的 `type` 指定了选用哪个模型;channel部分包含了应用渠道的配置,`type` 字段指定了接入哪个应用。
|
||||
|
||||
在使用时只需要更改 `model` 和 `channel` 配置块下的 `type` 字段,即可在任意模型和应用间完成切换,连接不同的通路。下面将依次介绍各个 模型 及 应用 的配置和运行过程。
|
||||
在使用时只需要更改 model 和 channel 配置块下的 type 字段,即可在任意模型和应用间完成切换,连接不同的通路。下面将依次介绍各个 模型 及 应用 的配置和运行过程。
|
||||
|
||||
|
||||
## 二、选择模型
|
||||
@@ -105,9 +105,17 @@ pip3 install --upgrade openai
|
||||
+ `character_desc` 配置中保存着你对机器人说的一段话,他会记住这段话并作为他的设定,你可以为他定制任何人格
|
||||
|
||||
|
||||
## 三、运行应用
|
||||
## 三、选择应用
|
||||
|
||||
### 1.个人微信
|
||||
### 1.命令行终端
|
||||
|
||||
配置模板中默认启动的应用即是终端,无需任何额外配置,直接在项目目录下通过命令行执行 `python3 app.py` 便可启动程序。用户通过命令行的输入与对话模型交互,且支持流式响应效果。
|
||||
|
||||

|
||||
|
||||
|
||||
|
||||
### 2.个人微信
|
||||
|
||||
与项目 [chatgpt-on-wechat](https://github.com/zhayujie/chatgpt-on-wechat) 的使用方式相同,目前接入个人微信可能导致账号被限制,暂时不建议使用。
|
||||
|
||||
@@ -131,9 +139,9 @@ pip3 install --upgrade openai
|
||||
|
||||
在项目根目录下执行 `python3 app.py` 即可启动程序,用手机扫码后完成登录,使用详情参考 [chatgpt-on-wechat](https://github.com/zhayujie/chatgpt-on-wechat)。
|
||||
|
||||
### 2.个人订阅号
|
||||
### 3.个人订阅号
|
||||
|
||||
**需要:**一台服务器,一个个人订阅号,一个已备案的域名。
|
||||
**需要:** 一台服务器,一个订阅号
|
||||
|
||||
#### 2.1 依赖安装
|
||||
|
||||
@@ -174,21 +182,28 @@ Hit Ctrl-C to quit.
|
||||
|
||||

|
||||
|
||||
- **服务器地址 (URL)**:在浏览器访问该URL需要能访问到服务器上运行的python程序 (默认监听8088端口)。由于公众号只能配置 80/443端口,所以需要在服务器进行端口转发 (如使用nginx),或改为直接监听80端口,并将对应的域名地址配置在url处 (仅用ip不行)。
|
||||
- **令牌 (Token)**:需和配置中的token一致。
|
||||
**服务器地址 (URL) 配置**: 如果在浏览器上通过配置的URL 能够访问到服务器上的Python程序 (默认监听8088端口),则说明配置有效。由于公众号只能配置 80/443端口,可以修改配置为直接监听 80 端口 (需要sudo权限),或者使用反向代理进行转发 (如nginx)。 根据官方文档说明,此处填写公网ip或域名均可。
|
||||
|
||||
**令牌 (Token) 配置**:需和 `config.json` 配置中的token一致。
|
||||
|
||||
详细操作过程参考 [官方文档](https://developers.weixin.qq.com/doc/offiaccount/Getting_Started/Getting_Started_Guide.html)
|
||||
|
||||
|
||||
#### 2.3 使用
|
||||
|
||||
用户关注订阅号后,发送消息即可。
|
||||
|
||||
> 注:用户发送消息后,微信后台会向配置的URL地址推送,但如果5s内未回复就会断开连接,同时重试3次,但往往请求openai接口不止5s。本项目中通过异步和缓存将5s超时限制优化至15s,但超出该时间仍无法正常回复。 同时每次5s连接断开时web框架会报错,待后续优化。
|
||||
|
||||
|
||||
### 3.企业服务号
|
||||
|
||||
在企业服务号中,通过先异步访问openai接口,再通过客服接口主动推送用户的方式,解决了个人订阅号的15s超时问题。
|
||||
|
||||
企业服务号配置只需修改type为`wechat_mp_service`,配置块仍复用 `wechat_mp`,在基础上增加了 `app_id` 和 `app_secret` 两个配置项。
|
||||
### 4.企业服务号
|
||||
|
||||
**需要:** 一个服务器、一个已微信认证的服务号
|
||||
|
||||
在企业服务号中,通过先异步访问openai接口,再通过客服接口主动推送给用户的方式,解决了个人订阅号的15s超时问题。服务号的开发者模式配置和上述订阅号类似,详情参考 [官方文档](https://developers.weixin.qq.com/doc/offiaccount/Getting_Started/Getting_Started_Guide.html)。
|
||||
|
||||
企业服务号的 `config.json` 配置只需修改type为`wechat_mp_service`,但配置块仍复用 `wechat_mp`,在此基础上需要增加 `app_id` 和 `app_secret` 两个配置项。
|
||||
|
||||
```bash
|
||||
"channel": {
|
||||
@@ -197,7 +212,7 @@ Hit Ctrl-C to quit.
|
||||
"wechat_mp": {
|
||||
"token": "YOUR TOKEN", # token值
|
||||
"port": "8088", # 程序启动监听的端口
|
||||
"app_id": "YOUR APP ID", # appID
|
||||
"app_id": "YOUR APP ID", # app ID
|
||||
"app_secret": "YOUR APP SECRET" # app secret
|
||||
}
|
||||
}
|
||||
|
||||
14
app.py
14
app.py
@@ -2,20 +2,24 @@
|
||||
|
||||
import config
|
||||
from channel import channel_factory
|
||||
from common.log import logger
|
||||
from common import log
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
try:
|
||||
# load config
|
||||
config.load_config()
|
||||
logger.info("[INIT] load config: {}".format(config.conf()))
|
||||
|
||||
model_type = config.conf().get("model").get("type")
|
||||
channel_type = config.conf().get("channel").get("type")
|
||||
|
||||
log.info("[INIT] Start up: {} on {}", model_type, channel_type)
|
||||
|
||||
# create channel
|
||||
channel = channel_factory.create_channel(config.conf().get("channel").get("type"))
|
||||
channel = channel_factory.create_channel(channel_type)
|
||||
|
||||
# startup channel
|
||||
channel.startup()
|
||||
except Exception as e:
|
||||
logger.error("App startup failed!")
|
||||
logger.exception(e)
|
||||
log.error("App startup failed!")
|
||||
log.exception(e)
|
||||
|
||||
@@ -9,6 +9,10 @@ def create_channel(channel_type):
|
||||
:param channel_type: channel type code
|
||||
:return: channel instance
|
||||
"""
|
||||
if channel_type== const.TERMINAL:
|
||||
from channel.terminal.terminal_channel import TerminalChannel
|
||||
return TerminalChannel()
|
||||
|
||||
if channel_type == const.WECHAT:
|
||||
from channel.wechat.wechat_channel import WechatChannel
|
||||
return WechatChannel()
|
||||
|
||||
@@ -0,0 +1,33 @@
|
||||
from channel.channel import Channel
|
||||
from common import log
|
||||
|
||||
import sys
|
||||
|
||||
class TerminalChannel(Channel):
|
||||
def startup(self):
|
||||
# close log
|
||||
log.close_log()
|
||||
context = {"from_user_id": "User", "stream": True}
|
||||
print("Please input your question\n")
|
||||
while True:
|
||||
try:
|
||||
prompt = self.get_input("User:\n")
|
||||
except KeyboardInterrupt:
|
||||
print("\nExiting...")
|
||||
sys.exit()
|
||||
|
||||
print("Bot:")
|
||||
sys.stdout.flush()
|
||||
for res in super().build_reply_content(prompt, context):
|
||||
print(res, end="")
|
||||
sys.stdout.flush()
|
||||
print("\n")
|
||||
|
||||
|
||||
def get_input(self, prompt):
|
||||
"""
|
||||
Multi-line input function
|
||||
"""
|
||||
print(prompt, end="")
|
||||
line = input()
|
||||
return line
|
||||
|
||||
@@ -18,8 +18,8 @@ class WechatServiceAccount(Channel):
|
||||
def startup(self):
|
||||
logger.info('[WX_Public] Wechat Public account service start!')
|
||||
robot.config['PORT'] = channel_conf(const.WECHAT_MP).get('port')
|
||||
robot.config["APP_ID"] = "YOUR APP ID"
|
||||
robot.config["APP_SECRET"] = "YOUR APP SECRET"
|
||||
robot.config["APP_ID"] = channel_conf(const.WECHAT_MP).get('app_id')
|
||||
robot.config["APP_SECRET"] = channel_conf(const.WECHAT_MP).get('app_secret')
|
||||
robot.run()
|
||||
|
||||
def handle(self, msg, count=0):
|
||||
|
||||
@@ -1,4 +1,5 @@
|
||||
# channel
|
||||
TERMINAL = "terminal"
|
||||
WECHAT = "wechat"
|
||||
WECHAT_MP = "wechat_mp"
|
||||
WECHAT_MP_SERVICE = "wechat_mp_service"
|
||||
|
||||
@@ -3,6 +3,7 @@
|
||||
import logging
|
||||
import sys
|
||||
|
||||
SWITCH = True
|
||||
|
||||
def _get_logger():
|
||||
log = logging.getLogger('log')
|
||||
@@ -13,6 +14,41 @@ def _get_logger():
|
||||
log.addHandler(console_handle)
|
||||
return log
|
||||
|
||||
def close_log():
|
||||
global SWITCH
|
||||
SWITCH = False
|
||||
|
||||
|
||||
def debug(arg, *args):
|
||||
if SWITCH:
|
||||
if len(args) == 0:
|
||||
logger.debug(arg)
|
||||
else:
|
||||
logger.debug(arg.format(*args))
|
||||
|
||||
def info(arg, *args):
|
||||
if SWITCH:
|
||||
if len(args) == 0:
|
||||
logger.info(arg)
|
||||
else:
|
||||
logger.info(arg.format(*args))
|
||||
|
||||
|
||||
def warn(arg, *args):
|
||||
if len(args) == 0:
|
||||
logger.warning(arg)
|
||||
else:
|
||||
logger.warning(arg.format(*args))
|
||||
|
||||
def error(arg, *args):
|
||||
if len(args) == 0:
|
||||
logger.error(arg)
|
||||
else:
|
||||
logger.error(arg.format(*args))
|
||||
|
||||
def exception(e):
|
||||
logger.exception(e)
|
||||
|
||||
|
||||
# 日志句柄
|
||||
logger = _get_logger()
|
||||
logger = _get_logger()
|
||||
|
||||
@@ -8,7 +8,7 @@
|
||||
}
|
||||
},
|
||||
"channel": {
|
||||
"type": "wechat_mp",
|
||||
"type": "terminal",
|
||||
"single_chat_prefix": ["bot", "@bot"],
|
||||
"single_chat_reply_prefix": "[bot] ",
|
||||
"group_chat_prefix": ["@bot"],
|
||||
|
||||
BIN
docs/images/terminal_demo.png
Normal file
BIN
docs/images/terminal_demo.png
Normal file
Binary file not shown.
|
After Width: | Height: | Size: 256 KiB |
Binary file not shown.
|
Before Width: | Height: | Size: 126 KiB After Width: | Height: | Size: 139 KiB |
@@ -3,7 +3,7 @@
|
||||
from model.model import Model
|
||||
from config import model_conf
|
||||
from common import const
|
||||
from common.log import logger
|
||||
from common import log
|
||||
import openai
|
||||
import time
|
||||
|
||||
@@ -18,17 +18,21 @@ class OpenAIModel(Model):
|
||||
def reply(self, query, context=None):
|
||||
# acquire reply content
|
||||
if not context or not context.get('type') or context.get('type') == 'TEXT':
|
||||
logger.info("[OPEN_AI] query={}".format(query))
|
||||
log.info("[OPEN_AI] query={}".format(query))
|
||||
from_user_id = context['from_user_id']
|
||||
if query == '#清除记忆':
|
||||
Session.clear_session(from_user_id)
|
||||
return '记忆已清除'
|
||||
|
||||
new_query = Session.build_session_query(query, from_user_id)
|
||||
logger.debug("[OPEN_AI] session query={}".format(new_query))
|
||||
log.debug("[OPEN_AI] session query={}".format(new_query))
|
||||
|
||||
if context.get('stream'):
|
||||
# reply in stream
|
||||
return self.reply_text_stream(query, new_query, from_user_id)
|
||||
|
||||
reply_content = self.reply_text(new_query, from_user_id, 0)
|
||||
logger.debug("[OPEN_AI] new_query={}, user={}, reply_cont={}".format(new_query, from_user_id, reply_content))
|
||||
log.debug("[OPEN_AI] new_query={}, user={}, reply_cont={}".format(new_query, from_user_id, reply_content))
|
||||
if reply_content and query:
|
||||
Session.save_session(query, reply_content, from_user_id)
|
||||
return reply_content
|
||||
@@ -49,45 +53,98 @@ class OpenAIModel(Model):
|
||||
stop=["\n\n\n"]
|
||||
)
|
||||
res_content = response.choices[0]['text'].strip().replace('<|endoftext|>', '')
|
||||
logger.info("[OPEN_AI] reply={}".format(res_content))
|
||||
log.info("[OPEN_AI] reply={}".format(res_content))
|
||||
return res_content
|
||||
except openai.error.RateLimitError as e:
|
||||
# rate limit exception
|
||||
logger.warn(e)
|
||||
log.warn(e)
|
||||
if retry_count < 1:
|
||||
time.sleep(5)
|
||||
logger.warn("[OPEN_AI] RateLimit exceed, 第{}次重试".format(retry_count+1))
|
||||
log.warn("[OPEN_AI] RateLimit exceed, 第{}次重试".format(retry_count+1))
|
||||
return self.reply_text(query, user_id, retry_count+1)
|
||||
else:
|
||||
return "提问太快啦,请休息一下再问我吧"
|
||||
except Exception as e:
|
||||
# unknown exception
|
||||
logger.exception(e)
|
||||
log.exception(e)
|
||||
Session.clear_session(user_id)
|
||||
return "请再问我一次吧"
|
||||
|
||||
|
||||
def reply_text_stream(self, query, new_query, user_id, retry_count=0):
|
||||
try:
|
||||
res = openai.Completion.create(
|
||||
model="text-davinci-003", # 对话模型的名称
|
||||
prompt=new_query,
|
||||
temperature=0.9, # 值在[0,1]之间,越大表示回复越具有不确定性
|
||||
max_tokens=1200, # 回复最大的字符数
|
||||
top_p=1,
|
||||
frequency_penalty=0.0, # [-2,2]之间,该值越大则更倾向于产生不同的内容
|
||||
presence_penalty=0.0, # [-2,2]之间,该值越大则更倾向于产生不同的内容
|
||||
stop=["\n\n\n"],
|
||||
stream=True
|
||||
)
|
||||
return self._process_reply_stream(query, res, user_id)
|
||||
|
||||
except openai.error.RateLimitError as e:
|
||||
# rate limit exception
|
||||
log.warn(e)
|
||||
if retry_count < 1:
|
||||
time.sleep(5)
|
||||
log.warn("[OPEN_AI] RateLimit exceed, 第{}次重试".format(retry_count+1))
|
||||
return self.reply_text(query, user_id, retry_count+1)
|
||||
else:
|
||||
return "提问太快啦,请休息一下再问我吧"
|
||||
except Exception as e:
|
||||
# unknown exception
|
||||
log.exception(e)
|
||||
Session.clear_session(user_id)
|
||||
return "请再问我一次吧"
|
||||
|
||||
|
||||
def _process_reply_stream(
|
||||
self,
|
||||
query: str,
|
||||
reply: dict,
|
||||
user_id: str
|
||||
) -> str:
|
||||
full_response = ""
|
||||
for response in reply:
|
||||
if response.get("choices") is None or len(response["choices"]) == 0:
|
||||
raise Exception("OpenAI API returned no choices")
|
||||
if response["choices"][0].get("finish_details") is not None:
|
||||
break
|
||||
if response["choices"][0].get("text") is None:
|
||||
raise Exception("OpenAI API returned no text")
|
||||
if response["choices"][0]["text"] == "<|endoftext|>":
|
||||
break
|
||||
yield response["choices"][0]["text"]
|
||||
full_response += response["choices"][0]["text"]
|
||||
if query and full_response:
|
||||
Session.save_session(query, full_response, user_id)
|
||||
|
||||
|
||||
def create_img(self, query, retry_count=0):
|
||||
try:
|
||||
logger.info("[OPEN_AI] image_query={}".format(query))
|
||||
log.info("[OPEN_AI] image_query={}".format(query))
|
||||
response = openai.Image.create(
|
||||
prompt=query, #图片描述
|
||||
n=1, #每次生成图片的数量
|
||||
size="256x256" #图片大小,可选有 256x256, 512x512, 1024x1024
|
||||
)
|
||||
image_url = response['data'][0]['url']
|
||||
logger.info("[OPEN_AI] image_url={}".format(image_url))
|
||||
log.info("[OPEN_AI] image_url={}".format(image_url))
|
||||
return image_url
|
||||
except openai.error.RateLimitError as e:
|
||||
logger.warn(e)
|
||||
log.warn(e)
|
||||
if retry_count < 1:
|
||||
time.sleep(5)
|
||||
logger.warn("[OPEN_AI] ImgCreate RateLimit exceed, 第{}次重试".format(retry_count+1))
|
||||
log.warn("[OPEN_AI] ImgCreate RateLimit exceed, 第{}次重试".format(retry_count+1))
|
||||
return self.reply_text(query, retry_count+1)
|
||||
else:
|
||||
return "提问太快啦,请休息一下再问我吧"
|
||||
except Exception as e:
|
||||
logger.exception(e)
|
||||
log.exception(e)
|
||||
return None
|
||||
|
||||
|
||||
@@ -125,8 +182,8 @@ class Session(object):
|
||||
conversation["question"] = query
|
||||
conversation["answer"] = answer
|
||||
session = user_session.get(user_id)
|
||||
logger.debug(conversation)
|
||||
logger.debug(session)
|
||||
log.debug(conversation)
|
||||
log.debug(session)
|
||||
if session:
|
||||
# append conversation
|
||||
session.append(conversation)
|
||||
|
||||
Reference in New Issue
Block a user