Update app.py

app.py

@@ -34,7 +34,8 @@ sys.path.append('clip-interrogator')
 
 import gradio as gr
 from clip_interrogator import Config, Interrogator
-
+import io 
+from PIL import Image
 config = Config()
 config.device = 'cuda' if torch.cuda.is_available() else 'cpu'
 config.blip_offload = False if torch.cuda.is_available() else True
@@ -44,31 +45,20 @@ config.blip_num_beams = 64
 
 ci = Interrogator(config)
 
-def inference(image, mode, best_max_flavors):
-    image = image.convert('RGB')
-    if mode == 'best':
-        
-        prompt_result = ci.interrogate(image, max_flavors=int(best_max_flavors))
-        
-        print("mode best: " + prompt_result)
-        
-        return prompt_result, gr.update(visible=True), gr.update(visible=True), gr.update(visible=True)
-    
-    elif mode == 'classic':
-        
-        prompt_result = ci.interrogate_classic(image)
-        
-        print("mode classic: " + prompt_result)
-        
-        return prompt_result, gr.update(visible=True), gr.update(visible=True), gr.update(visible=True)
-    
-    else:
-        
-        prompt_result = ci.interrogate_fast(image)
-        
-        print("mode fast: " + prompt_result)
-        
-        return prompt_result, gr.update(visible=True), gr.update(visible=True), gr.update(visible=True)
+def inference(input_images, mode, best_max_flavors):
+    # Process each image in the list and generate prompt results
+    prompt_results = []
+    for image_bytes in input_images:
+        image = Image.open(io.BytesIO(image_bytes)).convert('RGB')
+        if mode == 'best':
+            prompt_result = ci.interrogate(image, max_flavors=int(best_max_flavors))
+        elif mode == 'classic':
+            prompt_result = ci.interrogate_classic(image)
+        else:
+            prompt_result = ci.interrogate_fast(image)
+        prompt_results.append((image, prompt_result))  # Use dictionary to set image labels
+    return prompt_results
+
 
 title = """
     <div style="text-align: center; max-width: 500px; margin: 0 auto;">
@@ -109,7 +99,7 @@ article = """
 """
 
 css = '''
-#col-container {max-width: 700px; margin-left: auto; margin-right: auto;}
+#col-container {width: width: 80%;; margin-left: auto; margin-right: auto;}
 a {text-decoration-line: underline; font-weight: 600;}
 .animate-spin {
     animation: spin 1s linear infinite;
@@ -144,27 +134,24 @@ with gr.Blocks(css=css) as block:
     with gr.Column(elem_id="col-container"):
         gr.HTML(title)
 
-        input_image = gr.Image(type='pil', elem_id="input-img")
+        input_image = gr.inputs.File(file_count="multiple", type='bytes')
+        # NUM_IMAGES = len(input_image)
         with gr.Row():
             mode_input = gr.Radio(['best', 'classic', 'fast'], label='Select mode', value='best')
             flavor_input = gr.Slider(minimum=2, maximum=24, step=2, value=4, label='best mode max flavors')
         
         submit_btn = gr.Button("Submit")
         
-        output_text = gr.Textbox(label="Description Output", elem_id="output-txt")
-        
+        # rows, cols = NUM_IMAGES //3, 
+        gallery = gr.Gallery(
+            label="Images", show_label=True, elem_id="gallery"
+        ).style( object_fit="contain", height="auto")
+
         with gr.Group(elem_id="share-btn-container"):
-            community_icon = gr.HTML(community_icon_html, visible=False)
             loading_icon = gr.HTML(loading_icon_html, visible=False)
-            share_button = gr.Button("Share to community", elem_id="share-btn", visible=False)
 
-        examples=[['27E894C4-9375-48A1-A95D-CB2425416B4B.png', "best",4], ['DB362F56-BA98-4CA1-A999-A25AA94B723B.png',"fast",4]]
-        ex = gr.Examples(examples=examples, fn=inference, inputs=[input_image, mode_input, flavor_input], outputs=[output_text, share_button, community_icon, loading_icon], cache_examples=True, run_on_click=True)
-        ex.dataset.headers = [""]
-        
         gr.HTML(article)
+    submit_btn.click(fn=inference, inputs=[input_image,mode_input,flavor_input], outputs=[gallery], api_name="clipi2")
 
-    submit_btn.click(fn=inference, inputs=[input_image,mode_input,flavor_input], outputs=[output_text, share_button, community_icon, loading_icon], api_name="clipi2")
-    share_button.click(None, [], [], _js=share_js)
     
 block.queue(max_size=32,concurrency_count=10).launch(show_api=False)