Streaming local LLM with FastAPI, Llama.cpp and Langchain

题意：

使用FastAPI、Llama.cpp和Langchain流式传输本地大型语言模型

问题背景：

I have setup FastAPI with Llama.cpp and Langchain. Now I want to enable streaming in the FastAPI responses. Streaming works with Llama.cpp in my terminal, but I wasn't able to implement it with a FastAPI response.

我已经使用Llama.cpp和Langchain设置了FastAPI。现在我想在FastAPI响应中启用流式传输。在我的终端中，流式传输与Llama.cpp一起工作正常，但我无法将其与FastAPI响应一起实现。

Most tutorials focused on enabling streaming with an OpenAI model, but I am using a local LLM (quantized Mistral) with llama.cpp. I think I have to modify the Callbackhandler, but no tutorial worked. Here is my code:

大多数教程都集中在如何使用OpenAI模型启用流式传输，但我正在使用带有llama.cpp的本地大型语言模型（量化的Mistral）。我认为我需要修改Callbackhandler，但我没有找到任何可行的教程。以下是我的代码：

from fastapi import FastAPI, Request, Response
from langchain_community.llms import LlamaCpp
from langchain.callbacks.manager import CallbackManager
from langchain.callbacks.streaming_stdout import StreamingStdOutCallbackHandler
import copy
from langchain.chains import LLMChain
from langchain.prompts import PromptTemplate

model_path = "../modelle/mixtral-8x7b-instruct-v0.1.Q5_K_M.gguf"

prompt= """
<s> [INST] Im folgenden bekommst du eine Aufgabe. Erledige diese anhand des User Inputs.

### Hier die Aufgabe: ###
{typescript_string}

### Hier der User Input: ###
{input}

Antwort: [/INST]
"""

def model_response_prompt():
    return PromptTemplate(template=prompt, input_variables=['input', 'typescript_string'])

def build_llm(model_path, callback=None):
        callback_manager = CallbackManager([StreamingStdOutCallbackHandler()])
        #callback_manager = CallbackManager(callback)
        
        n_gpu_layers = 1 # Metal set to 1 is enough. # ausprobiert mit mehreren
        n_batch = 512#1024 # Should be between 1 and n_ctx, consider the amount of RAM of your Apple Silicon Chip.
   
        llm = LlamaCpp(
                max_tokens =1000,
                n_threads = 6,
                model_path=model_path,
                temperature= 0.8,
                f16_kv=True,
                n_ctx=28000, 
                n_gpu_layers=n_gpu_layers,
                n_batch=n_batch,
                callback_manager=callback_manager, 
                verbose=True,
                top_p=0.75,
                top_k=40,
                repeat_penalty = 1.1,
                streaming=True,
                model_kwargs={
                        'mirostat': 2,
                },
        )
        
        return llm

# caching LLM
@lru_cache(maxsize=100)
def get_cached_llm():
        chat = build_llm(model_path)
        return chat

chat = get_cached_llm()

app = FastAPI(
    title="Inference API for Mistral and Mixtral",
    description="A simple API that use Mistral or Mixtral",
    version="1.0",
)

app.add_middleware(
    CORSMiddleware,
    allow_origins=["*"],
    allow_credentials=True,
    allow_methods=["*"],
    allow_headers=["*"],
)

def bullet_point_model():          
    llm = build_llm(model_path=model_path)
    llm_chain = LLMChain(
        llm=llm,
        prompt=model_response_prompt(),
        verbose=True,
    )
    return llm_chain

@app.get('/model_response')
async def model(question : str, prompt: str):
    model = bullet_point_model()
    res = model({"typescript_string": prompt, "input": question})
    result = copy.deepcopy(res)
    return result

In a example notebook, I am calling FastAPI like this:

在一个示例笔记本中，我像这样调用FastAPI：

import  subprocess
import urllib.parse
import shlex
query = input("Insert your bullet points here: ")
task = input("Insert the task here: ")
#Safe Encode url string
encodedquery =  urllib.parse.quote(query)
encodedtask =  urllib.parse.quote(task)
#Join the curl command textx
command = f"curl -X 'GET' 'http://127.0.0.1:8000/model_response?question={encodedquery}&prompt={encodedtask}' -H 'accept: application/json'"
print(command)
args = shlex.split(command)
process = subprocess.Popen(args, shell=False, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
stdout, stderr = process.communicate()
print(stdout)

So with this code, getting responses from the API works. But I only see streaming in my terminal (I think this is because of the StreamingStdOutCallbackHandler. After the streaming in the terminal is complete, I am getting my FastAPI response.

所以，使用这段代码，从API获取响应是可行的。但我只能在终端中看到流式传输（我认为这是因为使用了StreamingStdOutCallbackHandler）。在终端中的流式传输完成后，我才能收到FastAPI的响应。

What do I have to change now that I can stream token by token with FastAPI and a local llama.cpp model?

我现在可以使用FastAPI和本地的llama.cpp模型逐令牌（token-by-token）地进行流式传输，那么我还需要改变什么？

问题解决：

I was doing the same and hit similar issue that FastAPI was not streaming the response even I am using the StreamingResponse API and eventually I got the following code work. There are three important part:

我之前也做了同样的事情，并遇到了类似的问题，即即使我使用了StreamingResponse API，FastAPI也没有流式传输响应。但最终我得到了以下可以工作的代码。这里有三个重要的部分：

• Make sure using StreamingResponse to wrap an Iterator.

确保使用StreamingResponse来包装一个迭代器

• Make sure the Iterator sends newline character \n in each streaming response.

确保迭代器在每个流式响应中发送换行符 \n。

• Make sure using streaming APIs to connect to your LLMs. For example, _client.chat function in my example is using httpx to connect to REST APIs for LLMs. If you use requests package, it won't work as it doesn't support streaming.

确保使用流式API来连接您的大型语言模型（LLMs）。例如，在我的示例中，_client.chat 函数使用 httpx 来连接到LLMs的REST API。如果您使用 requests 包，那么它将无法工作，因为 requests 不支持流式传输。

async def chat(self, request: Request):
"""
Generate a chat response using the requested model.
"""

# Passing request body JSON to parameters of function _chat
# Request body follows ollama API's chat request format for now.
params = await request.json()
self.logger.debug("Request data: %s", params)

chat_response = self._client.chat(**params)

# Always return as streaming
if isinstance(chat_response, Iterator):
    def generate_response():
        for response in chat_response:
            yield json.dumps(response) + "\n"
    return StreamingResponse(generate_response(), media_type="application/x-ndjson")
elif chat_response is not None:
    return json.dumps(chat_response)