File size: 3,169 Bytes
67b023b 70ad2ab 67b023b 70ad2ab 67b023b |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 |
from os import environ
import logging
from fastapi import FastAPI, Request
import subprocess
from huggingface_hub import InferenceClient, login
import langid
# Configure logging
logging.basicConfig(format="%(asctime)s - %(levelname)s - %(message)s",
level=logging.INFO)
logger = logging.getLogger(__name__)
# Get Hugging Face API token from environment variable
HF_HUB_TOKEN = environ.get("HUGGINGFACEHUB_API_TOKEN")
if not HF_HUB_TOKEN:
raise ValueError(
"Missing Hugging Face API token. Please set HUGGINGFACEHUB_API_TOKEN.")
# Login and initialize the client
login(token=HF_HUB_TOKEN)
client = InferenceClient(api_key=HF_HUB_TOKEN)
# Create FastAPI app
app = FastAPI()
# Function to detect language
def detect_language(user_input):
try:
lang, _ = langid.classify(user_input)
return "hebrew" if lang == "he" else "english" if lang == "en" else "unsupported"
except Exception as e:
logging.error(f"Language detection error: {e}")
return "unsupported"
# Function to generate response
def generate_response(text):
language = detect_language(text)
if language == "hebrew":
# content = "answer in hebrew, keep it short but tell your decision making process, " + text
content = "转砖诪讜专 注诇 转砖讜讘讛 拽爪专讛, 讗讘诇 转住驻专 讗讬讱 拽讬讘诇转 讗转 讛讛讞诇讟讛, " + text
model = "yam-peleg/Hebrew-Gemma-11B-V2" # too large.
# model = "google/gemma-2-2b-it"
#model = "microsoft/Phi-3.5-mini-instruct"
elif language == "english":
content = "keep it short but tell your decision making process, " + text
model = "mistralai/Mistral-Nemo-Instruct-2407"
else:
return "Sorry, I only support Hebrew and English."
messages = [{"role": "user", "content": content}]
completion = client.chat.completions.create(model=model,
messages=messages,
max_tokens=2048,
temperature=0.5,
top_p=0.7)
return completion.choices[0].message.content
@app.post("/generate_response")
async def generate_text(request: Request):
"""
Endpoint to generate a response from the chat model.
Expects a JSON with a "text" field.
"""
try:
data = await request.json()
text = data.get("text", "").strip()
if not text:
return {"error": "No text provided"}
response = generate_response(text)
return {"response": response}
except Exception as e:
logging.error(f"Error processing request: {e}")
return {"error": "An unexpected error occurred."}
@app.get("/")
async def root():
"""
Root endpoint to check that the API is running.
"""
return {"message": "Decision Helper API is running!"}
# Function to run bot.py
def run_bot():
logging.info("Starting Telegram bot...")
subprocess.Popen(["python3", "bot.py"])
if __name__ == "__main__":
run_bot()
# import uvicorn
# uvicorn.run(app, host="0.0.0.0", port=7860)
|