app.py

# Import required libraries
import gradio as gr
from huggingface_hub import InferenceClient
from transformers import pipeline
import torch
import huggingfacehub as infer
import threading
import time
import tensorflow as tf
from transformers import pipeline






# Initialize the text generation pipeline with the specified model
pipe = pipeline("text-generation", model="chargoddard/Yi-34B-Llama", device=0)

def respond(
    message,

    response = ""

    # Generate the response using the pipeline
    result = pipe(
        messages[-1]["content"],
        max_length=max_tokens,
        num_return_sequences=1,
        temperature=temperature,
        top_p=top_p,
    )


    response = result[0]['generated_text']
    yield response)

# Gradio interface setup


demo = gr.ChatInterface(
    respond,
    additional_inputs=[
    ],
)


if __name__ == "__main__":
    demo.launch()