add summary topic sentiment features

app.py

@@ -1,16 +1,28 @@
 import gradio as gr
-from transcribe import transcribe
+import requests
+# from transcribe import transcribe
+from sentiment_analysis import sentiment_analyser
+from summary import summarizer
+from topic import topic_gen
+
+def transcribe2():
+  response = requests.post("https://dwarkesh-whisper-speaker-recognition.hf.space/run/predict", json={
+    "data": [
+      {"name":"audio.wav","data":"data:audio/wav;base64,UklGRiQAAABXQVZFZm10IBAAAAABAAEARKwAAIhYAQACABAAZGF0YQAAAAA="},
+      2,
+  ]}).json()
+
+  data = response["data"]
 
 def main(audio_file, number_of_speakers):
   # Audio to Text Converter
-  text_data = transcribe(audio_file, number_of_speakers)
-  print(text_data)
-  title = "ss"
-  short_summary = "dsa"
-  sentiment_analysis = "gyn"
-  quality = "dsdww"
-  detailed_summary = "jbjbjbjs"
-  return title, short_summary, sentiment_analysis, quality, detailed_summary
+  # text_data = transcribe(audio_file, number_of_speakers)
+  # print(text_data)
+  text_data = ""
+  topic = topic_gen(text_data)
+  summary = summarizer(text_data)
+  sentiment_analysis = sentiment_analyser(text_data)
+  return topic, summary, sentiment_analysis
 
 # UI Interface on the Hugging Face Page
 with gr.Blocks() as demo:
@@ -23,19 +35,17 @@ with gr.Blocks() as demo:
           btn_clear = gr.ClearButton(value="Clear", components=[audio_file, number_of_speakers])
           btn_submit = gr.Button(value="Submit")
       with gr.Column():
-        title = gr.Textbox(label="Title", placeholder="Title for Conversation")
-        short_summary = gr.Textbox(label="Short Summary", placeholder="Short Summary for Conversation")
+        topic = gr.Textbox(label="Title", placeholder="Title for Conversation")
+        summary = gr.Textbox(label="Short Summary", placeholder="Short Summary for Conversation")
         sentiment_analysis = gr.Textbox(label="Sentiment Analysis", placeholder="Sentiment Analysis for Conversation")
-        quality = gr.Textbox(label="Quality of Conversation", placeholder="Quality of Conversation")
-        detailed_summary = gr.Textbox(label="Detailed Summary", placeholder="Detailed Summary for Conversation")
-      btn_submit.click(fn=main, inputs=[audio_file, number_of_speakers], outputs=[title, short_summary, sentiment_analysis, quality, detailed_summary])
+      btn_submit.click(fn=main, inputs=[audio_file, number_of_speakers], outputs=[topic, summary, sentiment_analysis])
     gr.Markdown("## Examples")
     gr.Examples( 
       examples=[
         ["./examples/sample4.wav", 2],
       ],
       inputs=[audio_file, number_of_speakers],
-      outputs=[title, short_summary, sentiment_analysis, quality, detailed_summary],
+      outputs=[topic, summary, sentiment_analysis],
       fn=main,
     )
   gr.Markdown(

sentiment_analysis.py

@@ -0,0 +1,5 @@
+from transformers import pipeline
+
+def sentiment_analyser(text):
+  sent = pipeline("sentiment-analysis",model="siebert/sentiment-roberta-large-english")
+  return sent(text)

summary.py

@@ -0,0 +1,5 @@
+from transformers import pipeline
+
+def summarizer(text):
+  summ = pipeline("summarization", model="knkarthick/MEETING-SUMMARY-BART-LARGE-XSUM-SAMSUM-DIALOGSUM")
+  return summ(text)

topic.py

@@ -0,0 +1,5 @@
+from transformers import pipeline
+
+def topic_gen(text):
+  topic = pipeline(" Text2Text Generation", model="knkarthick/TOPIC-DIALOGSUM")
+  return topic(text)