app.py

import os
import random
import uuid
import json
import time
import asyncio
from threading import Thread

import gradio as gr
import spaces
import torch
import numpy as np
from PIL import Image
import edge_tts

from transformers import (
    AutoModelForCausalLM,
    AutoTokenizer,
    TextIteratorStreamer,
    Qwen2VLForConditionalGeneration,
    AutoProcessor,
)
from transformers.image_utils import load_image
from diffusers import StableDiffusionXLPipeline, EulerAncestralDiscreteScheduler

DESCRIPTION = """
# QwQ Edge 💬
"""

css = '''
h1 {
  text-align: center;
  display: block;
}

#duplicate-button {
  margin: auto;
  color: #fff;
  background: #1565c0;
  border-radius: 100vh;
}
'''

MAX_MAX_NEW_TOKENS = 2048
DEFAULT_MAX_NEW_TOKENS = 1024
MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))

device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")

# Load text-only model and tokenizer
model_id = "prithivMLmods/FastThink-0.5B-Tiny"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    device_map="auto",
    torch_dtype=torch.bfloat16,
)
model.eval()

TTS_VOICES = [
    "en-US-JennyNeural",  # @tts1
    "en-US-GuyNeural",    # @tts2
]

MODEL_ID = "prithivMLmods/Qwen2-VL-OCR-2B-Instruct" 
processor = AutoProcessor.from_pretrained(MODEL_ID, trust_remote_code=True)
model_m = Qwen2VLForConditionalGeneration.from_pretrained(
    MODEL_ID,
    trust_remote_code=True,
    torch_dtype=torch.float16
).to("cuda").eval()

async def text_to_speech(text: str, voice: str, output_file="output.mp3"):
    """Convert text to speech using Edge TTS and save as MP3"""
    communicate = edge_tts.Communicate(text, voice)
    await communicate.save(output_file)
    return output_file

def clean_chat_history(chat_history):
    """
    Filter out any chat entries whose "content" is not a string.
    This helps prevent errors when concatenating previous messages.
    """
    cleaned = []
    for msg in chat_history:
        if isinstance(msg, dict) and isinstance(msg.get("content"), str):
            cleaned.append(msg)
    return cleaned

# A helper function to render a progress bar using HTML.
def render_progress_bar(label: str, progress: int, output_text: str = "") -> str:
    """
    Returns an HTML snippet containing a label, a progress bar (red background with a green inner bar),
    and optionally some output text.
    """
    return f'''
    <div style="margin-bottom: 10px;">
      <div style="font-weight: bold; margin-bottom: 5px;">{label}</div>
      <div style="width: 100%; background-color: red; border-radius: 5px; overflow: hidden; height: 10px;">
        <div style="width: {progress}%; background-color: green; height: 100%; transition: width 0.3s;"></div>
      </div>
      <div style="margin-top: 10px;">{output_text}</div>
    </div>
    '''

# Environment variables and parameters for Stable Diffusion XL
MODEL_ID_SD = os.getenv("MODEL_VAL_PATH")  # SDXL Model repository path via env variable
MAX_IMAGE_SIZE = int(os.getenv("MAX_IMAGE_SIZE", "4096"))
USE_TORCH_COMPILE = os.getenv("USE_TORCH_COMPILE", "0") == "1"
ENABLE_CPU_OFFLOAD = os.getenv("ENABLE_CPU_OFFLOAD", "0") == "1"
BATCH_SIZE = int(os.getenv("BATCH_SIZE", "1"))  # For batched image generation

# Load the SDXL pipeline
sd_pipe = StableDiffusionXLPipeline.from_pretrained(
    MODEL_ID_SD,
    torch_dtype=torch.float16 if torch.cuda.is_available() else torch.float32,
    use_safetensors=True,
    add_watermarker=False,
).to(device)
sd_pipe.scheduler = EulerAncestralDiscreteScheduler.from_config(sd_pipe.scheduler.config)

# Ensure that the text encoder is in half-precision if using CUDA.
if torch.cuda.is_available():
    sd_pipe.text_encoder = sd_pipe.text_encoder.half()

# Optional: compile the model for speedup if enabled
if USE_TORCH_COMPILE:
    sd_pipe.compile()

# Optional: offload parts of the model to CPU if needed
if ENABLE_CPU_OFFLOAD:
    sd_pipe.enable_model_cpu_offload()

MAX_SEED = np.iinfo(np.int32).max

def save_image(img: Image.Image) -> str:
    """Save a PIL image with a unique filename and return the path."""
    unique_name = str(uuid.uuid4()) + ".png"
    img.save(unique_name)
    return unique_name

def randomize_seed_fn(seed: int, randomize_seed: bool) -> int:
    if randomize_seed:
        seed = random.randint(0, MAX_SEED)
    return seed

@spaces.GPU(duration=60, enable_queue=True)
def generate_image_fn(
    prompt: str,
    negative_prompt: str = "",
    use_negative_prompt: bool = False,
    seed: int = 1,
    width: int = 1024,
    height: int = 1024,
    guidance_scale: float = 3,
    num_inference_steps: int = 25,
    randomize_seed: bool = False,
    use_resolution_binning: bool = True,
    num_images: int = 1,
    progress=gr.Progress(track_tqdm=True),
):
    """Generate images using the SDXL pipeline."""
    seed = int(randomize_seed_fn(seed, randomize_seed))
    generator = torch.Generator(device=device).manual_seed(seed)

    options = {
        "prompt": [prompt] * num_images,
        "negative_prompt": [negative_prompt] * num_images if use_negative_prompt else None,
        "width": width,
        "height": height,
        "guidance_scale": guidance_scale,
        "num_inference_steps": num_inference_steps,
        "generator": generator,
        "output_type": "pil",
    }
    if use_resolution_binning:
        options["use_resolution_binning"] = True

    images = []
    # Process in batches
    for i in range(0, num_images, BATCH_SIZE):
        batch_options = options.copy()
        batch_options["prompt"] = options["prompt"][i:i+BATCH_SIZE]
        if "negative_prompt" in batch_options and batch_options["negative_prompt"] is not None:
            batch_options["negative_prompt"] = options["negative_prompt"][i:i+BATCH_SIZE]
        if device.type == "cuda":
            with torch.autocast("cuda", dtype=torch.float16):
                outputs = sd_pipe(**batch_options)
        else:
            outputs = sd_pipe(**batch_options)
        images.extend(outputs.images)
    image_paths = [save_image(img) for img in images]
    return image_paths, seed

@spaces.GPU
def generate(
    input_dict: dict,
    chat_history: list[dict],
    max_new_tokens: int = 1024,
    temperature: float = 0.6,
    top_p: float = 0.9,
    top_k: int = 50,
    repetition_penalty: float = 1.2,
):
    """
    Generates chatbot responses with support for multimodal input, TTS, and image generation.
    Special commands:
      - "@tts1" or "@tts2": triggers text-to-speech.
      - "@image": triggers image generation using the SDXL pipeline.
      
    Instead of yielding a simple "Thinking..." text, an animated progress bar is shown (via an HTML snippet)
    that goes from red to green. When the inference is complete the progress bar is replaced by the final result.
    """
    text = input_dict["text"]
    files = input_dict.get("files", [])

    # Image generation branch
    if text.strip().lower().startswith("@image"):
        prompt = text[len("@image"):].strip()
        # Use a container to capture the result from the thread.
        result_container = []
        def run_image():
            result_container.append(generate_image_fn(
                prompt=prompt,
                negative_prompt="",
                use_negative_prompt=False,
                seed=1,
                width=1024,
                height=1024,
                guidance_scale=3,
                num_inference_steps=25,
                randomize_seed=True,
                use_resolution_binning=True,
                num_images=1,
            ))
        thread = Thread(target=run_image)
        thread.start()
        start_time = time.time()
        # Simulate progress bar updates while image generation is running.
        while thread.is_alive():
            progress = min(95, int((time.time() - start_time) / 5 * 95))
            yield render_progress_bar("Generating Image", progress)
            time.sleep(0.5)
        thread.join()
        # Final update before showing the result.
        yield render_progress_bar("Generating Image", 100)
        image_paths, used_seed = result_container[0]
        yield gr.Image(image_paths[0])
        return  # Exit early

    tts_prefix = "@tts"
    is_tts = any(text.strip().lower().startswith(f"{tts_prefix}{i}") for i in range(1, 3))
    voice_index = next((i for i in range(1, 3) if text.strip().lower().startswith(f"{tts_prefix}{i}")), None)
    
    if is_tts and voice_index:
        voice = TTS_VOICES[voice_index - 1]
        text = text.replace(f"{tts_prefix}{voice_index}", "").strip()
        # Clear previous chat history for a fresh TTS request.
        conversation = [{"role": "user", "content": text}]
    else:
        voice = None
        # Remove any stray @tts tags and build the conversation history.
        text = text.replace(tts_prefix, "").strip()
        conversation = clean_chat_history(chat_history)
        conversation.append({"role": "user", "content": text})

    # Multimodal (image + text) branch
    if files:
        if len(files) > 1:
            images = [load_image(image) for image in files]
        elif len(files) == 1:
            images = [load_image(files[0])]
        else:
            images = []
        messages = [{
            "role": "user",
            "content": [
                *[{"type": "image", "image": image} for image in images],
                {"type": "text", "text": text},
            ]
        }]
        prompt = processor.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
        inputs = processor(text=[prompt], images=images, return_tensors="pt", padding=True).to("cuda")
        streamer = TextIteratorStreamer(processor, skip_prompt=True, skip_special_tokens=True)
        generation_kwargs = {**inputs, "streamer": streamer, "max_new_tokens": max_new_tokens}
        thread = Thread(target=model_m.generate, kwargs=generation_kwargs)
        thread.start()

        buffer = ""
        start_time = time.time()
        # Initial progress bar for multimodal inference.
        yield render_progress_bar("Thinking...", 0)
        for new_text in streamer:
            buffer += new_text
            buffer = buffer.replace("<|im_end|>", "")
            progress = min(95, int((time.time() - start_time) / 5 * 95))
            yield render_progress_bar("Thinking...", progress, output_text=buffer)
        # Final progress update (100%).
        yield render_progress_bar("Thinking...", 100, output_text=buffer)
        # Then yield final response (progress bar update no longer shown).
        yield buffer
    else:
        # Text-only generation branch.
        input_ids = tokenizer.apply_chat_template(conversation, add_generation_prompt=True, return_tensors="pt")
        if input_ids.shape[1] > MAX_INPUT_TOKEN_LENGTH:
            input_ids = input_ids[:, -MAX_INPUT_TOKEN_LENGTH:]
            gr.Warning(f"Trimmed input from conversation as it was longer than {MAX_INPUT_TOKEN_LENGTH} tokens.")
        input_ids = input_ids.to(model.device)
        streamer = TextIteratorStreamer(tokenizer, timeout=20.0, skip_prompt=True, skip_special_tokens=True)
        generation_kwargs = {
            "input_ids": input_ids,
            "streamer": streamer,
            "max_new_tokens": max_new_tokens,
            "do_sample": True,
            "top_p": top_p,
            "top_k": top_k,
            "temperature": temperature,
            "num_beams": 1,
            "repetition_penalty": repetition_penalty,
        }
        t = Thread(target=model.generate, kwargs=generation_kwargs)
        t.start()

        outputs = []
        start_time = time.time()
        # Initial progress bar update.
        yield render_progress_bar("Thinking...", 0)
        for new_text in streamer:
            outputs.append(new_text)
            current_text = "".join(outputs)
            progress = min(95, int((time.time() - start_time) / 5 * 95))
            yield render_progress_bar("Thinking...", progress, output_text=current_text)
        final_response = "".join(outputs)
        # Final update (100% progress).
        yield render_progress_bar("Thinking...", 100, output_text=final_response)
        # Finally, yield the final plain response so the progress bar disappears.
        yield final_response

        # If TTS was requested, convert the final response to speech.
        if is_tts and voice:
            output_file = asyncio.run(text_to_speech(final_response, voice))
            yield gr.Audio(output_file, autoplay=True)

demo = gr.ChatInterface(
    fn=generate,
    additional_inputs=[
        gr.Slider(label="Max new tokens", minimum=1, maximum=MAX_MAX_NEW_TOKENS, step=1, value=DEFAULT_MAX_NEW_TOKENS),
        gr.Slider(label="Temperature", minimum=0.1, maximum=4.0, step=0.1, value=0.6),
        gr.Slider(label="Top-p (nucleus sampling)", minimum=0.05, maximum=1.0, step=0.05, value=0.9),
        gr.Slider(label="Top-k", minimum=1, maximum=1000, step=1, value=50),
        gr.Slider(label="Repetition penalty", minimum=1.0, maximum=2.0, step=0.05, value=1.2),
    ],
    examples=[
        ["@tts1 Who is Nikola Tesla, and why did he die?"],
        [{"text": "Extract JSON from the image", "files": ["examples/document.jpg"]}],
        [{"text": "summarize the letter", "files": ["examples/1.png"]}],
        ["@image Chocolate dripping from a donut against a yellow background, in the style of brocore, hyper-realistic"],
        ["Write a Python function to check if a number is prime."],
        ["@tts2 What causes rainbows to form?"],
    ],
    cache_examples=False,
    type="messages",
    description=DESCRIPTION,
    css=css,
    fill_height=True,
    textbox=gr.MultimodalTextbox(label="Query Input", file_types=["image"], file_count="multiple"),
    stop_btn="Stop Generation",
    multimodal=True,
)

if __name__ == "__main__":
    demo.queue(max_size=20).launch(share=True)