Update app.py

app.py

@@ -13,46 +13,31 @@ config.blip_num_beams = 64
 
 ci = Interrogator(config)
 
-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['data']:
-        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(prompt_result)  # Append text prompt result
-    return prompt_results
-
-# Function to convert image bytes to file format
-def convert_to_file(image_bytes):
-    file_dict = {"filename": ["image.jpg"], "data": [image_bytes]}
-    return file_dict
-
-# Function to handle image upload and conversion
-def handle_image_upload(image):
-    image_bytes = image.read()
-    file_dict = convert_to_file(image_bytes)
-    return file_dict
+def inference(input_image, mode, best_max_flavors):
+    image_bytes = input_image.read()
+    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)
+        
+    return prompt_result
 
 with gr.Blocks() as block:
     with gr.Column(elem_id="col-container"):
-        
-        
-
-        input_image_upload = gr.Image(label="Upload Image")  # Using gr.Image for image upload
-        mode_input = gr.Radio (['classic', 'fast'], label='Select mode', value='fast')
-        flavor_input = gr.Slider(minimum=2, maximum=24, step=2, value=4, label='best mode max flavors')
+        input_image = gr.Image(label="Upload Image")  # Using gr.Image for image upload
+        mode_input = gr.Radio(['classic', 'fast', 'best'], label='Select mode', default='fast')
+        flavor_input = gr.Slider(minimum=2, maximum=24, step=2, default=4, label='Best mode max flavors')
         submit_btn = gr.Button("Submit")
-        output_text = gr.Textbox(label="Output", elem_id="output-txt")  # Changed the variable name here
+        output_text = gr.Textbox(label="Output", elem_id="output-txt")  # Output Textbox
 
-        def process_image(image):
-            file_dict = handle_image_upload(image)
-            output_text = gr.Textbox(label="Output", elem_id="output-txt")  # Output Textbox
-
-            submit_btn.click(fn=inference, inputs=[file_dict, mode_input, flavor_input], outputs=[output_text], api_name="clipi2")
+        def process_image():
+            prompt_result = inference(input_image, mode_input, flavor_input)
+            output_text = prompt_result
         
+        submit_btn.click(process_image)
+
 block.queue(max_size=32, concurrency_count=10).launch(show_api=False)