app.py

from fastapi import FastAPI
from pydantic import BaseModel
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

app = FastAPI()

# Load model & tokenizer
MODEL_PATH = "./"  # since it's inside the same repo
tokenizer = AutoTokenizer.from_pretrained(MODEL_PATH)
model = AutoModelForCausalLM.from_pretrained(
    MODEL_PATH,
    torch_dtype=torch.float16,
    device_map="auto"
)

class RequestBody(BaseModel):
    prompt: str
    max_length: int = 100

@app.post("/generate")
def generate_text(req: RequestBody):
    inputs = tokenizer(req.prompt, return_tensors="pt").to(model.device)
    outputs = model.generate(**inputs, max_length=req.max_length)
    text = tokenizer.decode(outputs[0], skip_special_tokens=True)
    return {"generated_text": text}

@app.get("/")
def root():
    return {"message": "FastAPI Hugging Face Space is running!"}