Spaces:
Sleeping
Sleeping
import streamlit as st | |
from transformers import pipeline | |
# Initialize the text generation pipeline with optimizations | |
pipe = pipeline( | |
"text-generation", | |
model="Qwen/Qwen2.5-0.5B-Instruct", | |
device=-1, # Ensure it runs on CPU | |
use_fast=True, # Use fast tokenizer | |
) | |
# Streamlit app | |
st.title("Qwen Model Chat") | |
# Text input from the user | |
user_input = st.text_input("Enter your message:", "Delete this and write your query?") | |
# Generate text when the button is clicked | |
if st.button("Generate"): | |
messages = [{"role": "user", "content": user_input}] | |
# Reduce max_new_tokens for faster generation | |
output = pipe(messages, max_new_tokens=150) # Adjust as needed for speed | |
generated_text = output[0]['generated_text'] | |
# Display the generated text | |
st.write("Generated Response:") | |
st.write(generated_text) |