kdexd commited on
Commit
8f84007
1 Parent(s): eb66921

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +10 -18
app.py CHANGED
@@ -16,15 +16,12 @@ def gen_show_caption(sub_prompt=None, cap_prompt=""):
16
  st.markdown(
17
  f"""
18
  <style>
19
- red{{
20
- color:#c62828
21
- }}
22
- blue{{
23
- color:#2a72d5
24
- }}
25
  </style>
26
 
27
- #### <red> r/{subreddit} </red> <blue> {cap_prompt} </blue> {caption}
28
  """,
29
  unsafe_allow_html=True,
30
  )
@@ -65,7 +62,7 @@ cap_prompt = st.sidebar.text_input("Write the start of your caption below", valu
65
  _ = st.sidebar.button("Regenerate Caption")
66
 
67
 
68
- st.sidebar.title("Advanced Options:")
69
  num_captions = st.sidebar.select_slider(
70
  "Number of Captions to Predict", options=[1, 2, 3, 4, 5], value=1
71
  )
@@ -76,15 +73,6 @@ nuc_size = st.sidebar.slider(
76
  value=0.8,
77
  step=0.05,
78
  )
79
- st.sidebar.markdown(
80
- """
81
- *Please note that this model was explicitly not trained on images of people, and as a result is not designed to caption images with humans.
82
-
83
- This demo accompanies our paper RedCaps.
84
-
85
- Created by Karan Desai, Gaurav Kaul, Zubin Aysola, Justin Johnson
86
- """
87
- )
88
  # ----------------------------------------------------------------------------
89
 
90
  virtexModel.model.decoder.nucleus_size = nuc_size
@@ -113,8 +101,12 @@ st.markdown("""
113
  Caption your own images or try out some of our sample images.
114
  You can also generate captions as if they are from specific subreddits,
115
  as if they start with a particular prompt, or even both.
116
-
117
  Tweet your results with `#redcaps`!
 
 
 
 
 
118
  """)
119
 
120
  _, center, _ = st.columns([1, 18, 1])
16
  st.markdown(
17
  f"""
18
  <style>
19
+ red {{ color:#c62828; font-size: 1.5rem }}
20
+ blue {{ color:#2a72d5; font-size: 1.5rem }}
21
+ remaining {{ color: black; font-size: 1.5rem }}
 
 
 
22
  </style>
23
 
24
+ - <red> r/{subreddit} </red> <blue> {cap_prompt} </blue><remaining> {caption} </remaining>
25
  """,
26
  unsafe_allow_html=True,
27
  )
62
  _ = st.sidebar.button("Regenerate Caption")
63
 
64
 
65
+ st.sidebar.title("Advanced Options")
66
  num_captions = st.sidebar.select_slider(
67
  "Number of Captions to Predict", options=[1, 2, 3, 4, 5], value=1
68
  )
73
  value=0.8,
74
  step=0.05,
75
  )
 
 
 
 
 
 
 
 
 
76
  # ----------------------------------------------------------------------------
77
 
78
  virtexModel.model.decoder.nucleus_size = nuc_size
101
  Caption your own images or try out some of our sample images.
102
  You can also generate captions as if they are from specific subreddits,
103
  as if they start with a particular prompt, or even both.
 
104
  Tweet your results with `#redcaps`!
105
+
106
+ **Note:** This model was not trained on images of people,
107
+ hence may not generate accurate captions describing humans.
108
+ For more details, visit [redcaps.xyz](https://redcaps.xyz) check out
109
+ our [NeurIPS 2021 paper](https://openreview.net/forum?id=VjJxBi1p9zh).
110
  """)
111
 
112
  _, center, _ = st.columns([1, 18, 1])