app.py

import gradio as gr
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

model_id = "srujanamadiraju/merged-nl-sql-gemma2b"  # your uploaded model

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.float16, device_map="auto")

def generate_sql(prompt):
    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
    output = model.generate(**inputs, max_new_tokens=128, temperature=0.2, do_sample=True)
    return tokenizer.decode(output[0], skip_special_tokens=True)

interface = gr.Interface(fn=generate_sql, inputs="text", outputs="text", 
                         title="Gemma NL2SQL API", 
                         description="Convert natural language to SQL with merged Gemma model")

interface.launch()