Spaces:
Build error
Build error
Dongxu Li
commited on
Commit
•
d2f0b33
1
Parent(s):
ee0e33d
disable image uploading.
Browse files
app.py
CHANGED
@@ -6,10 +6,6 @@ from torchvision.transforms.functional import InterpolationMode
|
|
6 |
|
7 |
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
|
8 |
|
9 |
-
|
10 |
-
|
11 |
-
|
12 |
-
|
13 |
import gradio as gr
|
14 |
|
15 |
from models.blip import blip_decoder
|
@@ -61,14 +57,23 @@ def inference(raw_image, model_n, question, strategy):
|
|
61 |
answer = model_vq(image_vq, question, train=False, inference='generate')
|
62 |
return 'answer: '+answer[0]
|
63 |
|
64 |
-
inputs = [
|
|
|
|
|
|
|
|
|
|
|
|
|
65 |
outputs = gr.outputs.Textbox(label="Output")
|
66 |
|
67 |
title = "BLIP"
|
68 |
|
69 |
description = "Gradio demo for BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation (Salesforce Research). To use it, simply upload your image, or click one of the examples to load them. Read more at the links below."
|
70 |
|
71 |
-
article = "
|
72 |
-
|
|
|
|
|
|
|
73 |
|
74 |
gr.Interface(inference, inputs, outputs, title=title, description=description, article=article, examples=[['starrynight.jpeg',"Image Captioning","None","Nucleus sampling"]]).launch(enable_queue=True)
|
|
|
6 |
|
7 |
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
|
8 |
|
|
|
|
|
|
|
|
|
9 |
import gradio as gr
|
10 |
|
11 |
from models.blip import blip_decoder
|
|
|
57 |
answer = model_vq(image_vq, question, train=False, inference='generate')
|
58 |
return 'answer: '+answer[0]
|
59 |
|
60 |
+
inputs = [
|
61 |
+
gr.Image(type='pil', interactive=False),
|
62 |
+
gr.inputs.Radio(choices=['Image Captioning',"Visual Question Answering"],
|
63 |
+
type="value",
|
64 |
+
default="Image Captioning",
|
65 |
+
label="Task"
|
66 |
+
),gr.inputs.Textbox(lines=2, label="Question"),gr.inputs.Radio(choices=['Beam search','Nucleus sampling'], type="value", default="Nucleus sampling", label="Caption Decoding Strategy")]
|
67 |
outputs = gr.outputs.Textbox(label="Output")
|
68 |
|
69 |
title = "BLIP"
|
70 |
|
71 |
description = "Gradio demo for BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation (Salesforce Research). To use it, simply upload your image, or click one of the examples to load them. Read more at the links below."
|
72 |
|
73 |
+
article = """
|
74 |
+
<p style='text-align: center'><a href='https://arxiv.org/abs/2201.12086' target='_blank'>BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation</a> | <a href='https://github.com/salesforce/BLIP' target='_blank'>Github Repo</a></p>
|
75 |
+
<p><strong>We have now disable image uploading as of March 23. 2023. </strong>
|
76 |
+
<p><strong>For example usage, see notebooks https://github.com/salesforce/LAVIS/tree/main/examples.</strong>
|
77 |
+
"""
|
78 |
|
79 |
gr.Interface(inference, inputs, outputs, title=title, description=description, article=article, examples=[['starrynight.jpeg',"Image Captioning","None","Nucleus sampling"]]).launch(enable_queue=True)
|