changed description and made GIF faster

app.py

@@ -55,10 +55,10 @@ import warnings
 TITLE="""<h1 style="font-size: 42px;" align="center">Word-To-Image: Morphing Arabic Text to a Visual Representation</h1>"""
 
 
-DESCRIPTION="""This demo builds on the [Word-As-Image for Semantic Typography](https://wordasimage.github.io/Word-As-Image-Page/) work to support Arabic fonts and morphing whole words into semantic concepts. It is part of an ongoing project with the [ARBML](https://arbml.github.io/website/) community."""
+DESCRIPTION="""This demo builds on the [Word-As-Image for Semantic Typography](https://wordasimage.github.io/Word-As-Image-Page/) work to support Arabic fonts and morphing whole words and phrases to a visual representation of a semantic concept. This is part of an ongoing effort with the [ARBML](https://arbml.github.io/website/) community to build open-source Arabic tools using machine learning."""
 
 # DESCRIPTION += '\n<p>This demo is licensed under a <a rel="license" href="http://creativecommons.org/licenses/by-sa/4.0/"> Creative Commons Attribution-ShareAlike 4.0 International License</a>.</p>'
-DESCRIPTION += '\n<p>For faster inference without waiting in queue, you can <a href="https://colab.research.google.com/drive/1wobOAsnLpkIzaRxG5yac8NcV7iCrlycP"><img style="display: inline; margin-top: 0em; margin-bottom: 0em" src="https://colab.research.google.com/assets/colab-badge.svg" alt="Open In Colab"/></a></p>'
+DESCRIPTION += '\n<p>Note: it takes about 5 minutes for 500 iterations to generate the final GIF. For faster inference without waiting in queue, you can <a href="https://colab.research.google.com/drive/1wobOAsnLpkIzaRxG5yac8NcV7iCrlycP"><img style="display: inline; margin-top: 0em; margin-bottom: 0em" src="https://colab.research.google.com/assets/colab-badge.svg" alt="Open In Colab"/></a></p>'
 
 if (SPACE_ID := os.getenv('SPACE_ID')) is not None:
  DESCRIPTION = DESCRIPTION.replace("</p>", " ")
@@ -213,6 +213,7 @@ def run_main_app(semantic_concept, word, prompt_suffix, font_name, num_steps, se
  # training loop
  t_range = tqdm(range(num_iter))
  gif_frames = []
+ skip = 5
  for step in t_range:
  optim.zero_grad()
 
@@ -224,7 +225,9 @@ def run_main_app(semantic_concept, word, prompt_suffix, font_name, num_steps, se
  img = img[:, :, 3:4] * img[:, :, :3] + torch.ones(img.shape[0], img.shape[1], 3, device=device) * (
  1 - img[:, :, 3:4])
  img = img[:, :, :3]
- gif_frames += [img.detach().cpu().numpy()*255]
+
+ if step % skip == 0:
+ gif_frames += [img.detach().cpu().numpy()*255]
 
  
  filename = os.path.join(cfg.experiment_dir, "video-svg", f"iter{step:04d}.svg")
@@ -365,10 +368,10 @@ with gr.Blocks() as demo:
  with gr.Row():
  # examples
  examples = [
- ["قطة", "Cat", 5, 42],
- # ["كلب", "Dog", 5, 42],
- # ["حصان", "Horse", 5, 42],
- # ["أخطبوط", "Octopus", 5, 42],
+ ["قطة", "Cat", 250, 42],
+ ["جمل جميل", "Horse", 250, 42],
+ ["كلب", "Dog", 250, 42],
+ ["أخطبوط", "Octopus", 250, 42],
  ]
  demo.queue(max_size=10, concurrency_count=1)
  gr.Examples(examples=examples,