Update app.py

app.py

@@ -45,7 +45,7 @@ def speech_to_text(mic=None, file=None):
 
 def zero_shot(image, text_input):
     results = zero(image, text_input)
-    return([results])
+    return {dic["label"]: dic["score"] for dic in results}
     
 with gr.Blocks() as demo:    
     with gr.Row():
@@ -65,16 +65,16 @@ with gr.Blocks() as demo:
             image = gr.outputs.Image(type="pil", label="Your result")
             img = [image,gr.outputs.Carousel(label="Individual images",components=["image"]),gr.outputs.Textbox(label="Error")]
             with gr.Row():
-                get_image_latent = gr.Button("Generate Image go brr", css={"margin-top": "1em"})         
+                get_image_latent = gr.Button("Generate Image go brr")         
         with gr.Column():
             text_input = gr.Textbox(placeholder="input a list of labels separated by commas")
             label = gr.Label()
             with gr.Row():
-                zero_shot_clf = gr.Button("Classify Image go brr", css={"margin-top": "1em"})
+                zero_shot_clf = gr.Button("Classify Image go brr")
 
         
     speech_to_text.click(speech_to_text, inputs=audio_file, outputs=text)
     get_image_latent.click(text2image_latent, inputs=[text,steps,width,height,images,diversity], outputs=img)
-    zero_shot_clf.click(zero_shot, inputs=[img,text_input], outputs = label)
+    zero_shot_clf.click(zero_shot, inputs=[img[0],text_input], outputs=label)
 
 demo.launch(enable_queue=False)