app.py

from transformers import AutoModelForCausalLM, AutoTokenizer
from fastapi import FastAPI


MODEL_ID = "rasyosef/Phi-1_5-Instruct-v0.1"


tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
model = AutoModelForCausalLM.from_pretrained(MODEL_ID)

app = FastAPI()

@app.get("/chat")
def chat(query: str):
    """
    GET /chat?query=Your+question
    Returns JSON: {"answer": "...model’s reply..."}
    """
    prompt expected by Phi‐1.5 Instruct
    prompt = (
        "<|im_start|>system\nYou are a helpful assistant.<|im_end|>"
        "<|im_start|>user\n" + query + "<|im_end|>"
        "<|im_start|>assistant\n"
    )
    inputs = tokenizer(prompt, return_tensors="pt")
    outputs = model.generate(**inputs, max_new_tokens=200)
    # Only decode newly generated tokens (skip the “prompt” tokens)
    response = tokenizer.decode(
        outputs[0][inputs.input_ids.shape[-1]:], 
        skip_special_tokens=True
    )
    return {"answer": response.strip()}