Update backup.app.py

backup.app.py

@@ -1,26 +1,26 @@
+import base64
+import datetime
 import gradio as gr
-from diffusers import DiffusionPipeline, LCMScheduler, AutoencoderTiny
-import torch
+import numpy as np
 import os
-import datetime
+import pytz
+import psutil
+import re
+import random
+import torch
 import time
+import time
+
 from PIL import Image
-import re
-import base64
 from io import BytesIO
-import pytz
+from PIL import Image
+from diffusers import DiffusionPipeline, LCMScheduler, AutoencoderTiny
 
 try:
     import intel_extension_for_pytorch as ipex
 except:
     pass
 
-from PIL import Image
-import numpy as np
-import gradio as gr
-import psutil
-import time
-
 SAFETY_CHECKER = os.environ.get("SAFETY_CHECKER", None)
 TORCH_COMPILE = os.environ.get("TORCH_COMPILE", None)
 HF_TOKEN = os.environ.get("HF_TOKEN", None)
@@ -76,6 +76,12 @@ def encode_image(image):
     buffered = BytesIO()
     #image.save(buffered, format="PNG")
     return base64.b64encode(buffered.getvalue()).decode()
+
+def fake_gan():
+    base_dir = os.getcwd()  # Get the current base directory
+    img_files = [file for file in os.listdir(base_dir) if file.lower().endswith((".png", ".jpg", ".jpeg"))]  # List all files ending with ".jpg" or ".jpeg"
+    images = [(random.choice(img_files), os.path.splitext(file)[0]) for file in img_files]
+    return images
     
 def predict(prompt, guidance, steps, seed=1231231):
     generator = torch.manual_seed(seed)
@@ -99,11 +105,6 @@ def predict(prompt, guidance, steps, seed=1231231):
     if nsfw_content_detected:
         nsfw=gr.Button("🕹️NSFW🎨", scale=1)
 
-    # Generate file name
-    #date_str = datetime.datetime.now().strftime("%Y%m%d")
-    #safe_prompt = prompt.replace(" ", "_")[:50]  # Truncate long prompts
-    #filename = f"{date_str}_{safe_prompt}.png"
-
     central = pytz.timezone('US/Central')
     safe_date_time = datetime.datetime.now().strftime("%Y%m%d")
     replaced_prompt = prompt.replace(" ", "_").replace("\n", "_")
@@ -116,8 +117,6 @@ def predict(prompt, guidance, steps, seed=1231231):
         image_path = os.path.join("", filename)  # Specify your directory
         results.images[0].save(image_path)
         print(f"#Image saved as {image_path}")
-        #filename = safe_filename(prompt)
-        #image.save(filename)
         encoded_image = encode_image(image)
         html_link = f'<a href="data:image/png;base64,{encoded_image}" download="{filename}">Download Image</a>'
         gr.Markdown(html_link)
@@ -151,7 +150,23 @@ with gr.Blocks(css=css) as demo:
                 )
                 generate_bt = gr.Button("Generate", scale=1)
 
+        # Image Result from last prompt
         image = gr.Image(type="filepath")
+
+        # Gallery of Generated Images with Image Names in Random Set to Download
+        with gr.Row(variant="compact"):
+            text = gr.Textbox(
+                label="Image Sets",
+                show_label=False,
+                max_lines=1,
+                placeholder="Enter your prompt",
+            )
+            btn = gr.Button("Generate Gallery of Saved Images")
+        gallery = gr.Gallery(
+            label="Generated Images", show_label=False, elem_id="gallery"
+        )
+
+        # Advanced Generate Options
         with gr.Accordion("Advanced options", open=False):
             guidance = gr.Slider(
                 label="Guidance", minimum=0.0, maximum=5, value=0.3, step=0.001
@@ -160,6 +175,8 @@ with gr.Blocks(css=css) as demo:
             seed = gr.Slider(
                 randomize=True, minimum=0, maximum=12013012031030, label="Seed", step=1
             )
+
+        # Diffusers
         with gr.Accordion("Run with diffusers"):
             gr.Markdown(
                 """## Running LCM-LoRAs it with `diffusers`
@@ -169,9 +186,11 @@ with gr.Blocks(css=css) as demo:
             
             ```py
             from diffusers import DiffusionPipeline, LCMScheduler
+
             pipe = DiffusionPipeline.from_pretrained("Lykon/dreamshaper-7").to("cuda") 
             pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)
             pipe.load_lora_weights("latent-consistency/lcm-lora-sdv1-5") #yes, it's a normal LoRA
+
             results = pipe(
                 prompt="ImageEditor",
                 num_inference_steps=4,
@@ -182,8 +201,10 @@ with gr.Blocks(css=css) as demo:
             """
             )
 
+        # Function IO Eventing and Controls
         inputs = [prompt, guidance, steps, seed]
         generate_bt.click(fn=predict, inputs=inputs, outputs=image, show_progress=False)
+        btn.click(fake_gan, None, gallery)
         prompt.input(fn=predict, inputs=inputs, outputs=image, show_progress=False)
         guidance.change(fn=predict, inputs=inputs, outputs=image, show_progress=False)
         steps.change(fn=predict, inputs=inputs, outputs=image, show_progress=False)