Dongxu Li
commited on
Commit
•
6f077c8
1
Parent(s):
f6fc05a
update ui.
Browse files
app.py
CHANGED
@@ -126,9 +126,12 @@ def inference_caption(
|
|
126 |
|
127 |
|
128 |
title = """<h1 align="center">BLIP-2</h1>"""
|
129 |
-
description = """Gradio demo for BLIP-2, image-to-text generation from Salesforce Research. To use it, simply upload your image, or click one of the examples to load them
|
130 |
<p> <strong>Disclaimer</strong>: This is a research prototype and is not intended for production use. No data including but not restricted to text and images is collected. </p>"""
|
131 |
-
article = "<
|
|
|
|
|
|
|
132 |
|
133 |
endpoint = Endpoint()
|
134 |
|
@@ -204,7 +207,7 @@ with gr.Blocks() as iface:
|
|
204 |
)
|
205 |
|
206 |
with gr.Column():
|
207 |
-
chat_input = gr.Textbox(lines=1, label="Chat Input")
|
208 |
with gr.Row():
|
209 |
chatbot = gr.Chatbot(label="Chat Output (from FlanT5)")
|
210 |
image_input.change(lambda: (None, "", "", []), [], [chatbot, chat_input, caption_output, state])
|
|
|
126 |
|
127 |
|
128 |
title = """<h1 align="center">BLIP-2</h1>"""
|
129 |
+
description = """Gradio demo for BLIP-2, image-to-text generation from Salesforce Research. To use it, simply upload your image, or click one of the examples to load them.</p>
|
130 |
<p> <strong>Disclaimer</strong>: This is a research prototype and is not intended for production use. No data including but not restricted to text and images is collected. </p>"""
|
131 |
+
article = """<strong>Paper</strong>: <a href='https://arxiv.org/abs/2301.12597' target='_blank'>BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models</a>
|
132 |
+
<br> <strong>Code</strong>: BLIP2 is now integrated into GitHub repo: <a href='https://github.com/salesforce/LAVIS' target='_blank'>LAVIS: a One-stop Library for Language and Vision</a>
|
133 |
+
<br> <strong>Project Page</strong>: <a href='https://github.com/salesforce/LAVIS/tree/main/projects/blip2' target='_blank'> BLIP2 on LAVIS</a>
|
134 |
+
"""
|
135 |
|
136 |
endpoint = Endpoint()
|
137 |
|
|
|
207 |
)
|
208 |
|
209 |
with gr.Column():
|
210 |
+
chat_input = gr.Textbox(lines=1, label="Chat Input (recommend prompt for QA, Question: {} Answer:)")
|
211 |
with gr.Row():
|
212 |
chatbot = gr.Chatbot(label="Chat Output (from FlanT5)")
|
213 |
image_input.change(lambda: (None, "", "", []), [], [chatbot, chat_input, caption_output, state])
|