app.py

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

# Load model and tokenizer
model_name = "Dorjzodovsuren/Mongolian_Llama3-v0.1"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)

def generate_response(input_text):
    # Tokenize the input text
    inputs = tokenizer(input_text, return_tensors="pt")
    
    # Generate response
    with torch.no_grad():
        outputs = model.generate(
            **inputs,
            max_length=100,  # Adjust for desired response length
            temperature=0.7,  # Adjust for creativity
            top_p=0.9        # Adjust for response diversity
        )
    # Decode the generated text
    response_text = tokenizer.decode(outputs[0], skip_special_tokens=True)
    return response_text

# Create Gradio interface
iface = gr.Interface(
    fn=generate_response,
    inputs="text",
    outputs="text",
    title="Mongolian Llama3 Chatbot",
    description="Ask anything in Mongolian!"
)

# Launch the Gradio app
iface.launch()