app.py

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

# Load model and tokenizer
model_name = "rkwsuper/lora_model"

# Use an environment variable or secret for the token
auth_token = os.getenv("HF_TOKEN")  # Automatically fetches the token if set as an environment variable

tokenizer = AutoTokenizer.from_pretrained(model_name, use_auth_token=auth_token)
model = AutoModelForCausalLM.from_pretrained(model_name, use_auth_token=auth_token)

# Define the function for inference
def generate_text(prompt, max_length=100, temperature=1.0):
    inputs = tokenizer(prompt, return_tensors="pt")
    outputs = model.generate(
        inputs["input_ids"],
        max_length=max_length,
        temperature=temperature,
        pad_token_id=tokenizer.eos_token_id
    )
    return tokenizer.decode(outputs[0], skip_special_tokens=True)

# Create the Gradio interface
iface = gr.Interface(
    fn=generate_text,
    inputs=[
        gr.Textbox(label="Enter Prompt"),
        gr.Slider(10, 300, value=100, step=10, label="Max Length"),
        gr.Slider(0.1, 2.0, value=1.0, step=0.1, label="Temperature")
    ],
    outputs="text",
    title="Hugging Face Model Text Generator",
    description="This interface generates text based on your input using a fine-tuned Hugging Face model."
)

# Launch
if __name__ == "__main__":
    iface.launch()