Update app.py

app.py

@@ -28,6 +28,9 @@ import clip
 from guided_diffusion.script_util import create_model_and_diffusion, model_and_diffusion_defaults
 import numpy as np
 import imageio
+
+torch.hub.download_url_to_file('https://images.pexels.com/photos/68767/divers-underwater-ocean-swim-68767.jpeg', 'coralreef.jpeg')
+
 def fetch(url_or_path):
     if str(url_or_path).startswith('http://') or str(url_or_path).startswith('https://'):
         r = requests.get(url_or_path)
@@ -219,5 +222,5 @@ def inference(text, init_image, skip_timesteps, clip_guidance_scale, tv_scale, r
 title = "CLIP Guided Diffusion HQ"
 description = "Gradio demo for CLIP Guided Diffusion. To use it, simply add your text, or click one of the examples to load them. Read more at the links below."
 article = "<p style='text-align: center'> By Katherine Crowson (https://github.com/crowsonkb, https://twitter.com/RiversHaveWings). It uses OpenAI's 256x256 unconditional ImageNet diffusion model (https://github.com/openai/guided-diffusion) together with CLIP (https://github.com/openai/CLIP) to connect text prompts with images. | <a href='https://colab.research.google.com/drive/12a_Wrfi2_gwwAuN3VvMTwVMz9TfqctNj' target='_blank'>Colab</a></p>"
-iface = gr.Interface(inference, inputs=["text",gr.inputs.Image(type="file", label='initial image (optional)', optional=True),gr.inputs.Slider(minimum=0, maximum=45, step=1, default=10, label="skip_timesteps"), gr.inputs.Slider(minimum=0, maximum=3000, step=1, default=600, label="clip guidance scale (Controls how much the image should look like the prompt)"), gr.inputs.Slider(minimum=0, maximum=1000, step=1, default=0, label="tv_scale (Controls the smoothness of the final output)"), gr.inputs.Slider(minimum=0, maximum=1000, step=1, default=0, label="range_scale (Controls how far out of range RGB values are allowed to be)"), gr.inputs.Slider(minimum=0, maximum=1000, step=1, default=0, label="init_scale (This enhances the effect of the init image)"), gr.inputs.Number(default=0, label="Seed"), gr.inputs.Image(type="file", label='image prompt (optional)', optional=True), gr.inputs.Slider(minimum=50, maximum=500, step=1, default=50, label="timestep respacing"),gr.inputs.Slider(minimum=1, maximum=64, step=1, default=32, label="cutn")], outputs=["image","video"], title=title, description=description, article=article, examples=[["coral reef city by artistation artists", "", 0, 1000, 150, 50, 0, 0, "", 90, 32]])
+iface = gr.Interface(inference, inputs=["text",gr.inputs.Image(type="file", label='initial image (optional)', optional=True),gr.inputs.Slider(minimum=0, maximum=45, step=1, default=10, label="skip_timesteps"), gr.inputs.Slider(minimum=0, maximum=3000, step=1, default=600, label="clip guidance scale (Controls how much the image should look like the prompt)"), gr.inputs.Slider(minimum=0, maximum=1000, step=1, default=0, label="tv_scale (Controls the smoothness of the final output)"), gr.inputs.Slider(minimum=0, maximum=1000, step=1, default=0, label="range_scale (Controls how far out of range RGB values are allowed to be)"), gr.inputs.Slider(minimum=0, maximum=1000, step=1, default=0, label="init_scale (This enhances the effect of the init image)"), gr.inputs.Number(default=0, label="Seed"), gr.inputs.Image(type="file", label='image prompt (optional)', optional=True), gr.inputs.Slider(minimum=50, maximum=500, step=1, default=50, label="timestep respacing"),gr.inputs.Slider(minimum=1, maximum=64, step=1, default=32, label="cutn")], outputs=["image","video"], title=title, description=description, article=article, examples=[["coral reef city by artistation artists", "coralreef.jpeg", 0, 1000, 150, 50, 0, 0, "coralreef.jpeg", 90, 32]])
 iface.launch(enable_queue=True,cache_examples=True)