main.py

import fastapi
import json
import markdown
import uvicorn
from fastapi.responses import HTMLResponse
from fastapi.middleware.cors import CORSMiddleware
from sse_starlette.sse import EventSourceResponse
from ctransformers import AutoModelForCausalLM
from ctransformers.langchain import CTransformers
from pydantic import BaseModel

llm = AutoModelForCausalLM.from_pretrained("TheBloke/gorilla-7B-GGML",
                                           model_file="Gorilla-7B.ggmlv3.q4_0.bin",
                                           model_type="llama")
app = fastapi.FastAPI()
app.add_middleware(
    CORSMiddleware,
    allow_origins=["*"],
    allow_credentials=True,
    allow_methods=["*"],
    allow_headers=["*"],
)

@app.get("/")
async def index():
    with open("README.md", "r", encoding="utf-8") as readme_file:
        md_template_string = readme_file.read()
    html_content = markdown.markdown(md_template_string)
    return HTMLResponse(content=html_content, status_code=200)

class ChatCompletionRequest(BaseModel):
    prompt: str

@app.post("/v1/chat/completions")
async def chat(request: ChatCompletionRequest, response_mode=None):
    tokens = llm.tokenize(prompt)
    async def server_sent_events(chat_chunks, llm):
        yield prompt
        for chat_chunk in llm.generate(chat_chunks):
            yield llm.detokenize(chat_chunk)
        yield ""

    return EventSourceResponse(server_sent_events(tokens, llm))

if __name__ == "__main__":
  uvicorn.run(app, host="0.0.0.0", port=8000)