Spaces:
Runtime error
Runtime error
Commit
•
7518be4
1
Parent(s):
627fbe3
Update app.py
Browse files
app.py
CHANGED
@@ -75,13 +75,13 @@ def generate_answers(image, question):
|
|
75 |
|
76 |
answer_blip_large = generate_answer_blip(blip_processor_large, blip_model_large, image, question)
|
77 |
|
78 |
-
answer_vilt = generate_answer_vilt(vilt_processor, vilt_model, image, question)
|
79 |
|
80 |
-
return answer_git_base, answer_git_large, answer_blip_base, answer_blip_large
|
81 |
|
82 |
|
83 |
examples = [["cats.jpg", "How many cats are there?"], ["stop_sign.png", "What's behind the stop sign?"], ["astronaut.jpg", "What's the astronaut riding on?"]]
|
84 |
-
outputs = [gr.outputs.Textbox(label="Answer generated by GIT-base"), gr.outputs.Textbox(label="Answer generated by GIT-large"), gr.outputs.Textbox(label="Answer generated by BLIP-base"), gr.outputs.Textbox(label="Answer generated by BLIP-large")
|
85 |
|
86 |
title = "Interactive demo: comparing visual question answering (VQA) models"
|
87 |
description = "Gradio Demo to compare GIT, BLIP and ViLT, 3 state-of-the-art vision+language models. To use it, simply upload your image and click 'submit', or click one of the examples to load them. Read more at the links below."
|
|
|
75 |
|
76 |
answer_blip_large = generate_answer_blip(blip_processor_large, blip_model_large, image, question)
|
77 |
|
78 |
+
# answer_vilt = generate_answer_vilt(vilt_processor, vilt_model, image, question)
|
79 |
|
80 |
+
return answer_git_base, answer_git_large, answer_blip_base, answer_blip_large#, answer_vilt
|
81 |
|
82 |
|
83 |
examples = [["cats.jpg", "How many cats are there?"], ["stop_sign.png", "What's behind the stop sign?"], ["astronaut.jpg", "What's the astronaut riding on?"]]
|
84 |
+
outputs = [gr.outputs.Textbox(label="Answer generated by GIT-base"), gr.outputs.Textbox(label="Answer generated by GIT-large"), gr.outputs.Textbox(label="Answer generated by BLIP-base"), gr.outputs.Textbox(label="Answer generated by BLIP-large")]#, gr.outputs.Textbox(label="Answer generated by ViLT")]
|
85 |
|
86 |
title = "Interactive demo: comparing visual question answering (VQA) models"
|
87 |
description = "Gradio Demo to compare GIT, BLIP and ViLT, 3 state-of-the-art vision+language models. To use it, simply upload your image and click 'submit', or click one of the examples to load them. Read more at the links below."
|