app.py

from fastapi import FastAPI
from fastapi.middleware.cors import CORSMiddleware
from pydantic import BaseModel
import torch
import joblib
import os
from transformers import AutoTokenizer
from model_inference import load_model, predict_from_input

# ✅ FIX: Set Hugging Face cache to a writable directory
os.environ["TRANSFORMERS_CACHE"] = "/tmp/hf_cache"
os.environ["HF_HOME"] = "/tmp/hf_cache"

app = FastAPI(title="Question Difficulty/Discrimination Predictor")

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

# ✅ Load model on startup
model, device = load_model()

encoder = joblib.load("assets/onehot_encoder.pkl")
scaler = joblib.load("assets/scaler.pkl")

tok_mcq = AutoTokenizer.from_pretrained("microsoft/BiomedNLP-PubMedBERT-base-uncased-abstract")
tok_clin = AutoTokenizer.from_pretrained("emilyalsentzer/Bio_ClinicalBERT")

class QuestionInput(BaseModel):
    StemText: str
    LeadIn: str
    OptionA: str
    OptionB: str
    OptionC: str
    OptionD: str
    DepartmentName: str
    CourseName: str
    BloomLevel: str

@app.get("/health")
def health():
    return {"status": "ok"}

@app.post("/predict")
def predict(input_data: QuestionInput):
    pred = predict_from_input(
        input_data.dict(), model, device,
        tok_mcq, tok_clin, encoder, scaler
    )
    return pred