app.py

import gradio as gr
import torch
from peft import PeftModel, PeftConfig
from transformers import AutoModelForCausalLM, AutoTokenizer


peft_model_id = f"IThinkUPC/SQLGenerator-AI"
config = PeftConfig.from_pretrained(peft_model_id)
model = AutoModelForCausalLM.from_pretrained(config.base_model_name_or_path, return_dict=True, load_in_8bit=True, device_map='auto')
tokenizer = AutoTokenizer.from_pretrained(config.base_model_name_or_path)
# Load the Lora model
model = PeftModel.from_pretrained(model, peft_model_id)

def greet(name):
    return "Hello " + name + "!!"

def make_inference(prompt):  
  batch = tokenizer(f"### Question:\n{prompt}: \n\n### Query", return_tensors='pt')
  with torch.cuda.amp.autocast():
    output_tokens = model.generate(**batch, max_new_tokens=50)
  return tokenizer.decode(output_tokens[0], skip_special_tokens=True)
    

#iface = gr.Interface(fn=greet, inputs="text", outputs="text")
iface = gr.Interface(fn=make_inference, inputs="text", outputs="text")
iface.launch()