题意:“在 LangChain 和 Flask 中,如何将 OpenAI 更改为 ChatOpenAI?”
问题背景:
This is an implementation based on langchain and flask and refers to an implementation to be able to stream responses from the OpenAI server in langchain to a page with javascript that can show the streamed response.
“这是一个基于 LangChain 和 Flask 的实现,用于将 OpenAI 服务器的响应流式传输到一个带有 JavaScript 的页面,该页面可以显示流式响应。”
I tried all ways to modify the code below to replace the langchain library from openai to chatopenai without success, i upload below both implementations (the one with openai working) and the one chatopenai with error. thank you to all the community and those who can help me to understand the problem, it would be very helpful if you could also show me how to solve it since I have been trying for days and the error it shows has really no significance.
“我尝试了所有方法修改下面的代码,将 LangChain 库从 OpenAI 替换为 ChatOpenAI,但没有成功。下面上传了两个实现(一个是使用 OpenAI 正常工作的版本,另一个是带有错误的 ChatOpenAI 版本)。感谢所有社区成员以及能够帮助我理解问题的人。如果你们能告诉我如何解决这个问题,那将非常有帮助,因为我已经尝试了好几天,而显示的错误并没有什么实际意义。”
Code version with library that works but reports as deprecated:
“这是使用库的代码版本,它可以正常工作但报告为已弃用:”
from flask import Flask, Response
import threading
import queue
from langchain.llms import OpenAI
from langchain.callbacks.base import BaseCallbackManager
from langchain.callbacks.streaming_stdout import StreamingStdOutCallbackHandler
app = Flask(__name__)
@app.route('/')
def index():
return Response('''<!DOCTYPE html>
<html>
<head><title>Flask Streaming Langchain Example</title></head>
<body>
<div id="output"></div>
<script>
const outputEl = document.getElementById('output');
(async function() {
try {
const controller = new AbortController();
const signal = controller.signal;
const timeout = 120000; // Imposta il timeout su 120 secondi
setTimeout(() => controller.abort(), timeout);
const response = await fetch('/chain', {method: 'POST', signal});
const reader = response.body.getReader();
const decoder = new TextDecoder();
let buffer = '';
while (true) {
const { done, value } = await reader.read();
if (done) { break; }
const text = decoder.decode(value, {stream: true});
outputEl.innerHTML += text;
}
} catch (err) {
console.error(err);
}
})();
</script>
</body>
</html>''', mimetype='text/html')
class ThreadedGenerator:
def __init__(self):
self.queue = queue.Queue()
def __iter__(self):
return self
def __next__(self):
item = self.queue.get()
if item is StopIteration: raise item
return item
def send(self, data):
self.queue.put(data)
def close(self):
self.queue.put(StopIteration)
class ChainStreamHandler(StreamingStdOutCallbackHandler):
def __init__(self, gen):
super().__init__()
self.gen = gen
def on_llm_new_token(self, token: str, **kwargs):
self.gen.send(token)
def llm_thread(g, prompt):
try:
llm = OpenAI(
model_name="gpt-4",
verbose=True,
streaming=True,
callback_manager=BaseCallbackManager([ChainStreamHandler(g)]),
temperature=0.7,
)
llm(prompt)
finally:
g.close()
def chain(prompt):
g = ThreadedGenerator()
threading.Thread(target=llm_thread, args=(g, prompt)).start()
return g
@app.route('/chain', methods=['POST'])
def _chain():
return Response(chain("Create a poem about the meaning of life \n\n"), mimetype='text/plain')
if __name__ == '__main__':
app.run(threaded=True, debug=True)
Version with error (OpenAI replaced with ChatOpenAI)“
这是带有错误的版本(将 OpenAI 替换为 ChatOpenAI):”
import threading
import queue
from langchain.chat_models import ChatOpenAI
from langchain.callbacks.base import BaseCallbackManager
from langchain.callbacks.streaming_stdout import StreamingStdOutCallbackHandler
app = Flask(__name__)
@app.route('/')
def index():
return Response('''<!DOCTYPE html>
<html>
<head><title>Flask Streaming Langchain Example</title></head>
<body>
<div id="output"></div>
<script>
const outputEl = document.getElementById('output');
(async function() {
try {
const controller = new AbortController();
const signal = controller.signal;
const timeout = 120000; // Imposta il timeout su 120 secondi
setTimeout(() => controller.abort(), timeout);
const response = await fetch('/chain', {method: 'POST', signal});
const reader = response.body.getReader();
const decoder = new TextDecoder();
let buffer = '';
while (true) {
const { done, value } = await reader.read();
if (done) { break; }
const text = decoder.decode(value, {stream: true});
outputEl.innerHTML += text;
}
} catch (err) {
console.error(err);
}
})();
</script>
</body>
</html>''', mimetype='text/html')
class ThreadedGenerator:
def __init__(self):
self.queue = queue.Queue()
def __iter__(self):
return self
def __next__(self):
item = self.queue.get()
if item is StopIteration: raise item
return item
def send(self, data):
self.queue.put(data)
def close(self):
self.queue.put(StopIteration)
class ChainStreamHandler(StreamingStdOutCallbackHandler):
def __init__(self, gen):
super().__init__()
self.gen = gen
def on_llm_new_token(self, token: str, **kwargs):
self.gen.send(token)
def on_chat_model_start(self, token: str):
print("started")
def llm_thread(g, prompt):
try:
llm = ChatOpenAI(
model_name="gpt-4",
verbose=True,
streaming=True,
callback_manager=BaseCallbackManager([ChainStreamHandler(g)]),
temperature=0.7,
)
llm(prompt)
finally:
g.close()
def chain(prompt):
g = ThreadedGenerator()
threading.Thread(target=llm_thread, args=(g, prompt)).start()
return g
@app.route('/chain', methods=['POST'])
def _chain():
return Response(chain("parlami dei 5 modi di dire in inglese che gli italiani conoscono meno \n\n"), mimetype='text/plain')
if __name__ == '__main__':
app.run(threaded=True, debug=True)
Error showing the console at startup and at the time I enter the web page:
“启动时和进入网页时控制台显示的错误:”
Error in ChainStreamHandler.on_chat_model_start callback: ChainStreamHandler.on_chat_model_start() got an unexpected keyword argument 'run_id'
Exception in thread Thread-4 (llm_thread):
127.0.0.1 - - [09/Sep/2023 18:09:29] "POST /chain HTTP/1.1" 200 -
Traceback (most recent call last):
File "C:\Users\user22\Desktop\Work\TESTPROJ\env\Lib\site-packages\langchain\callbacks\manager.py", line 300, in _handle_event
getattr(handler, event_name)(*args, **kwargs)
File "C:\Users\user22\Desktop\Work\TESTPROJ\env\Lib\site-packages\langchain\callbacks\base.py", line 168, in on_chat_model_start
raise NotImplementedError(
NotImplementedError: StdOutCallbackHandler does not implement `on_chat_model_start`
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "C:\Users\user22\AppData\Local\Programs\Python\Python311\Lib\threading.py", line 1038, in _bootstrap_inner
self.run()
File "C:\Users\user22\AppData\Local\Programs\Python\Python311\Lib\threading.py", line 975, in run
self._target(*self._args, **self._kwargs)
File "c:\Users\user22\Desktop\Work\TESTPROJ\streamresp.py", line 90, in llm_thread
llm(prompt)
File "C:\Users\user22\Desktop\Work\TESTPROJ\env\Lib\site-packages\langchain\chat_models\base.py", line 552, in __call__
generation = self.generate(
^^^^^^^^^^^^^^
File "C:\Users\user22\Desktop\Work\TESTPROJ\env\Lib\site-packages\langchain\chat_models\base.py", line 293, in generate
run_managers = callback_manager.on_chat_model_start(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\user22\Desktop\Work\TESTPROJ\env\Lib\site-packages\langchain\callbacks\manager.py", line 1112, in on_chat_model_start
_handle_event(
File "C:\Users\user22\Desktop\Work\TESTPROJ\env\Lib\site-packages\langchain\callbacks\manager.py", line 304, in _handle_event
message_strings = [get_buffer_string(m) for m in args[1]]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\user22\Desktop\Work\TESTPROJ\env\Lib\site-packages\langchain\callbacks\manager.py", line 304, in <listcomp>
message_strings = [get_buffer_string(m) for m in args[1]]
^^^^^^^^^^^^^^^^^^^^
File "C:\Users\user22\Desktop\Work\TESTPROJ\env\Lib\site-packages\langchain\schema\messages.py", line 52, in get_buffer_string
raise ValueError(f"Got unsupported message type: {m}")
ValueError: Got unsupported message type: p
thank you very much for the support!
“非常感谢您的支持!”
问题解决:
Thanks to python273 user on github I've resolved:
“感谢 GitHub 上的用户 python273,我已经解决了这个问题。”
import os
os.environ["OPENAI_API_KEY"] = ""
from flask import Flask, Response, request
import threading
import queue
from langchain.chat_models import ChatOpenAI
from langchain.callbacks.streaming_stdout import StreamingStdOutCallbackHandler
from langchain.schema import AIMessage, HumanMessage, SystemMessage
app = Flask(__name__)
@app.route('/')
def index():
# just for the example, html is included directly, move to .html file
return Response('''
<!DOCTYPE html>
<html>
<head><title>Flask Streaming Langchain Example</title></head>
<body>
<form id="form">
<input name="prompt" value="write a short koan story about seeing beyond"/>
<input type="submit"/>
</form>
<div id="output"></div>
<script>
const formEl = document.getElementById('form');
const outputEl = document.getElementById('output');
let aborter = new AbortController();
async function run() {
aborter.abort(); // cancel previous request
outputEl.innerText = '';
aborter = new AbortController();
const prompt = new FormData(formEl).get('prompt');
try {
const response = await fetch(
'/chain', {
signal: aborter.signal,
method: 'POST',
headers: {'Content-Type': 'application/json'},
body: JSON.stringify({
prompt
}),
}
);
const reader = response.body.getReader();
const decoder = new TextDecoder();
while (true) {
const { done, value } = await reader.read();
if (done) { break; }
const decoded = decoder.decode(value, {stream: true});
outputEl.innerText += decoded;
}
} catch (err) {
console.error(err);
}
}
run(); // run on initial prompt
formEl.addEventListener('submit', function(event) {
event.preventDefault();
run();
});
</script>
</body>
</html>
''', mimetype='text/html')
class ThreadedGenerator:
def __init__(self):
self.queue = queue.Queue()
def __iter__(self):
return self
def __next__(self):
item = self.queue.get()
if item is StopIteration: raise item
return item
def send(self, data):
self.queue.put(data)
def close(self):
self.queue.put(StopIteration)
class ChainStreamHandler(StreamingStdOutCallbackHandler):
def __init__(self, gen):
super().__init__()
self.gen = gen
def on_llm_new_token(self, token: str, **kwargs):
self.gen.send(token)
def llm_thread(g, prompt):
try:
chat = ChatOpenAI(
verbose=True,
streaming=True,
callbacks=[ChainStreamHandler(g)],
temperature=0.7,
)
chat([HumanMessage(content=prompt)])
finally:
g.close()
def chain(prompt):
g = ThreadedGenerator()
threading.Thread(target=llm_thread, args=(g, prompt)).start()
return g
@app.route('/chain', methods=['POST'])
def _chain():
return Response(chain(request.json['prompt']), mimetype='text/plain')
if __name__ == '__main__':
app.run(threaded=True, debug=True)
Link to the original reply: https://gist.github.com/python273/563177b3ad5b9f74c0f8f3299ec13850
“原始回复的链接: https://gist.github.com/python273/563177b3ad5b9f74c0f8f3299ec13850”