Update app.py

app.py

@@ -48,17 +48,20 @@
 # )
 # interface.launch(debug=False)
 
+
 import warnings
 warnings.filterwarnings("ignore")
 import gradio as gr
 from src.video_model import describe_video  # Assuming this function processes the video and query
 
-def process_video_and_questions(video, query, sitting, hands, location, screen):
+# --- Function to construct the final query --- 
+def process_video_and_questions(video, sitting, hands, location, screen):
+    query = "Here are some frames of a video. Describe this video in detail."
     additional_info = []
     if sitting:
         additional_info.append("Is the subject in the video standing or sitting?")
     if hands:
-        additional_info.append("Is the subject holding any object in their hands, if so the hands are not free else they are free.?")
+        additional_info.append("Is the subject holding any object in their hands, if so the hands are not free else they are free?")
     if location:
         additional_info.append("Is the subject present indoors or outdoors?")
     if screen:
@@ -71,7 +74,6 @@ def process_video_and_questions(video, query, sitting, hands, location, screen):
 
 # Video and text inputs for the interface
 video = gr.Video(label="Video")
-query = gr.Textbox(label="Question", value="Here are some frames of a video. Describe this video in detail and answer the below questions.", placeholder="Enter your question here")
 
 # Options as checkboxes
 sitting = gr.Checkbox(label="Sitting/Standing")
@@ -86,14 +88,11 @@ response = gr.Textbox(label="Predicted answer", show_label=True, show_copy_butto
 examples = [
     [
         "videos/2016-01-01_0100_US_KNBC_Channel_4_News_1867.16-1871.38_now.mp4",
-        "Here are some frames of a video. Describe this video in detail."
     ],
     [
         "videos/2016-01-01_0200_US_KNBC_Channel_4_News_1329.12-1333.29_tonight.mp4",
-        "Here are some frames of a video. Describe this video in detail."
     ],
     [   "videos/2016-01-01_0830_US_KNBC_Tonight_Show_with_Jimmy_Fallon_725.45-729.76_tonight.mp4", 
-        "Here are some frames of a video. Describe this video in detail."
     ]
 ]
 
@@ -105,7 +104,7 @@ article = "<p style='text-align: center'><a href='https://github.com/OpenBMB/Min
 # Launch the interface
 interface = gr.Interface(
     fn=process_video_and_questions, # Updated function to handle the query construction
-    inputs=[video, query, sitting, hands, location, screen], 
+    inputs=[video, sitting, hands, location, screen], 
     outputs=response,
     examples=examples,
     title=title,