app.py

import gradio as gr
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
import os
from threading import Thread
import spaces
import time
import subprocess

MIN_TOKENS=128
MAX_TOKENS=8192
DEFAULT_TOKENS=2048
DURATION=60

# Install flash attention
subprocess.run(
    "pip install flash-attn --no-build-isolation",
    env={"FLASH_ATTENTION_SKIP_CUDA_BUILD": "TRUE"},
    shell=True,
)

# Load model and tokenizer once when the app starts
model_token = os.environ["HF_TOKEN"]
model = AutoModelForCausalLM.from_pretrained(
    "microsoft/Phi-3-mini-128k-instruct",
    token=model_token,
    trust_remote_code=True,
)
tokenizer = AutoTokenizer.from_pretrained("microsoft/Phi-3-mini-128k-instruct", token=model_token)

# Set device (GPU or CPU)
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model.to(device)

# Define error handling function
def handle_error(error):
    return {"error": str(error)}

# Define chat function with input validation and error handling
@spaces.GPU(duration=DURATION)
def chat(message, history, temperature, do_sample, max_tokens):
    try:
        # Validate input
        if not message:
            raise ValueError("Please enter a message")
        if temperature < 0 or temperature > 1:
            raise ValueError("Temperature must be between 0 and 1")
        if max_tokens < MIN_TOKENS or max_tokens > MAX_TOKENS:
            raise ValueError(f"Max tokens must be between {MIN_TOKENS} and {MAX_TOKENS}")

        # Prepare chat history
        chat = []
        for item in history:
            chat.append({"role": "user", "content": item[0]})
            if item[1] is not None:
                chat.append({"role": "assistant", "content": item[1]})
        chat.append({"role": "user", "content": message})

        # Generate response
        messages = tokenizer.apply_chat_template(chat, tokenize=False, add_generation_prompt=True)
        model_inputs = tokenizer([messages], return_tensors="pt").to(device)
        streamer = TextIteratorStreamer(tokenizer, timeout=20.0, skip_prompt=True, skip_special_tokens=True)
        generate_kwargs = dict(
            model_inputs,
            streamer=streamer,
            max_new_tokens=max_tokens,
            do_sample=do_sample,
            temperature=temperature,
            eos_token_id=[tokenizer.eos_token_id],
        )

        # Generate response
        t = Thread(target=model.generate, kwargs=generate_kwargs)
        t.start()

        # Yield partial responses
        partial_text = ""
        for new_text in streamer:
            partial_text += new_text
            yield partial_text

        # Yield final response
        yield partial_text

    except Exception as e:
        yield handle_error(e)

# Create Gradio interface
demo = gr.ChatInterface(
    fn=chat,
    examples=[["Write me a poem about Machine Learning."]],
    additional_inputs_accordion=gr.Accordion(
        label="⚙️ Parameters", open=False, render=False
    ),
    additional_inputs=[
        gr.Slider(
            minimum=0, maximum=1, step=0.1, value=0.9, label="Temperature", render=False
        ),
        gr.Checkbox(label="Sampling", value=True),
        gr.Slider(
            minimum=MIN_TOKENS,
            maximum=MAX_TOKENS,
            step=1,
            value=DEFAULT_TOKENS,
            label="Max new tokens",
            render=False,
        ),
    ],
    stop_btn="Stop Generation",
    title="Chat With LLMs",
    description="Now Running [microsoft/Phi-3-mini-128k-instruct](https://huggingface.co/microsoft/Phi-3-mini-128k-instruct)",
)

# Launch Gradio app
demo.launch()