hsienchen commited on
Commit
902323a
1 Parent(s): 4a98f1e

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +10 -16
app.py CHANGED
@@ -46,6 +46,7 @@ def llm_response(history,text,img):
46
  return history
47
 
48
  # Function that takes User Inputs and displays it on ChatUI
 
49
  def output_query_message(txt,img):
50
  if not img:
51
  return txt
@@ -71,33 +72,26 @@ def output_llm_response(text,img):
71
  def sentence_builder(animal, place):
72
  return f"""how many {animal}s from the {place} are shown in the picture?"""
73
 
74
-
75
  # gradio block
76
 
77
  with gr.Blocks(theme='snehilsanyal/scikit-learn') as app1:
78
- title = gr.Markdown("## COT ##")
79
  with gr.Column():
80
- outputbox = gr.Textbox(label="AI prediction here...")
81
- text_box = gr.Dropdown(
82
- ["what is in the image",
83
- "provide alternative title for the image",
84
- "how many birds can be seen in the picture?"],
85
- label="Prompts", info="Will add more animals later!"
86
- )
87
  image_box = gr.Image(type="filepath")
88
 
89
- btn = gr.Button("Submit")
90
  clicked = btn.click(output_query_message,
91
- [text_box,image_box],
92
  outputbox
93
  ).then(output_llm_response,
94
- [text_box,image_box],
95
  outputbox
96
  )
97
  gr.Markdown("""
98
- # Multimodal Chain-of-Thought Reasoning in Language Models
99
 
100
- <h5 align="center"><i>"Imagine learning a textbook without figures or tables."</i></h5>
101
 
102
  Multimodal-CoT incorporates vision features in a decoupled training framework. The framework consists of two training stages: (i) rationale generation and (ii) answer inference. Both stages share the same model architecture but differ in the input and output.
103
  """)
@@ -128,8 +122,8 @@ with gr.Blocks(theme='snehilsanyal/scikit-learn') as app2:
128
  chatbot
129
  )
130
  with gr.Blocks(theme='snehilsanyal/scikit-learn') as demo:
131
- gr.Markdown("# DEMO #")
132
- gr.TabbedInterface([app1, app2], ["APP #1", "APP #2"])
133
 
134
  demo.queue()
135
  demo.launch()
 
46
  return history
47
 
48
  # Function that takes User Inputs and displays it on ChatUI
49
+ text_box_01 = "what is in the image"
50
  def output_query_message(txt,img):
51
  if not img:
52
  return txt
 
72
  def sentence_builder(animal, place):
73
  return f"""how many {animal}s from the {place} are shown in the picture?"""
74
 
 
75
  # gradio block
76
 
77
  with gr.Blocks(theme='snehilsanyal/scikit-learn') as app1:
78
+ title = 'line clearance'
79
  with gr.Column():
80
+ outputbox = gr.Textbox(label="result here...")
 
 
 
 
 
 
81
  image_box = gr.Image(type="filepath")
82
 
83
+ btn = gr.Button("Check This")
84
  clicked = btn.click(output_query_message,
85
+ [text_box_01,image_box],
86
  outputbox
87
  ).then(output_llm_response,
88
+ [text_box_01,image_box],
89
  outputbox
90
  )
91
  gr.Markdown("""
92
+ ## SOP-302: Line Clearance ##
93
 
94
+ <h5 align="center"><i>"XXXX here here."</i></h5>
95
 
96
  Multimodal-CoT incorporates vision features in a decoupled training framework. The framework consists of two training stages: (i) rationale generation and (ii) answer inference. Both stages share the same model architecture but differ in the input and output.
97
  """)
 
122
  chatbot
123
  )
124
  with gr.Blocks(theme='snehilsanyal/scikit-learn') as demo:
125
+ gr.Markdown("## SOP Camera ##")
126
+ gr.TabbedInterface([app1, app2], ["Check #1", "Check #2"])
127
 
128
  demo.queue()
129
  demo.launch()