app-sd.py

import gradio as gr
from gtts import gTTS
from moviepy.editor import *
from PIL import Image
import numpy as np
import torch
from diffusers import StableDiffusionPipeline
import tempfile
import os

# -----------------------------------------------------
# Load Stable Diffusion model (this will work on GPU)
# -----------------------------------------------------
device = "cuda" if torch.cuda.is_available() else "cpu"
sd_pipeline = StableDiffusionPipeline.from_pretrained(
    "runwayml/stable-diffusion-v1-5",
    torch_dtype=torch.float16 if device == "cuda" else torch.float32
).to(device)

# -----------------------------------------------------
# Function: Generate background image with Stable Diffusion
# -----------------------------------------------------
def generate_background(prompt: str):
    image = sd_pipeline(prompt).images[0]
    return image

# -----------------------------------------------------
# Function: Generate video with voice + background
# -----------------------------------------------------
def generate_video(prompt, affirmation, with_visuals):
    # Create speech
    text = f"{prompt}. {affirmation}" if affirmation else prompt
    tts = gTTS(text=text, lang="en")
    temp_audio = tempfile.NamedTemporaryFile(delete=False, suffix=".mp3")
    tts.save(temp_audio.name)

    # Generate background
    if with_visuals:
        bg_img = generate_background(prompt)
    else:
        # fallback: plain white
        bg_img = Image.new("RGB", (720, 480), color=(255, 255, 255))

    bg_path = tempfile.NamedTemporaryFile(delete=False, suffix=".png").name
    bg_img.save(bg_path)

    # Make video
    clip = ImageClip(bg_path).set_duration(15)  # 15 sec video
    audioclip = AudioFileClip(temp_audio.name)
    final = clip.set_audio(audioclip)

    out_path = tempfile.NamedTemporaryFile(delete=False, suffix=".mp4").name
    final.write_videofile(out_path, fps=24)

    return out_path

# -----------------------------------------------------
# Gradio Interface
# -----------------------------------------------------
with gr.Blocks() as demo:
    gr.Markdown("## ✨ Future Journal & Affirmation App with AI Visuals ✨")

    with gr.Row():
        prompt_in = gr.Textbox(label="Journal Prompt / Affirmation")
        affirmation_in = gr.Textbox(label="Extra Affirmation (optional)")

    with gr.Row():
        visuals_in = gr.Checkbox(label="Add AI Visuals (Stable Diffusion)", value=True)

    generate_btn = gr.Button("🎬 Generate Video")

    output_video = gr.Video(label="Generated Video")

    def run(prompt, affirmation, visuals):
        return generate_video(prompt, affirmation, visuals)

    generate_btn.click(run, [prompt_in, affirmation_in, visuals_in], output_video)

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