Update app.py

app.py

@@ -6,6 +6,9 @@ from transformers import WhisperProcessor, WhisperForConditionalGeneration
 processor = WhisperProcessor.from_pretrained("openai/whisper-tiny.en")
 model = WhisperForConditionalGeneration.from_pretrained("openai/whisper-tiny.en")
 
+# Title
+st.title('Audio to Text Transcription')
+
 # Sidebar for file upload
 st.sidebar.title("Upload your audio file")
 uploaded_file = st.sidebar.file_uploader("Choose an audio file", type=["mp3", "wav", "mp4", "m4a"])
@@ -21,8 +24,6 @@ if uploaded_file:
     segment_duration = 120  # Segment duration in seconds (2 minutes)
     num_segments = len(resampled_waveform[0]) // (segment_duration * 16000)
     segment_transcriptions = []
-    current_speaker = None
-    current_paragraph = ""
 
     # Transcribe each segment
     for i in range(num_segments):
@@ -43,21 +44,7 @@ if uploaded_file:
     # Combine segment transcriptions into the full transcript
     full_transcript = " ".join(segment_transcriptions)
 
-    # Separate transcript by speakers and apply paragraphs
-    for line in full_transcript.split('. '):  # You can adjust the separator depending on your audio content
-        if line.endswith((':', 'said')):
-            if current_paragraph:
-                st.write(f"Speaker: {current_speaker}")
-                st.write(current_paragraph + ".")
-            current_speaker = line
-            current_paragraph = ""
-        else:
-            current_paragraph += line + ". "
-    # Adding the last speaker paragraph
-    if current_paragraph:
-        st.write(f"Speaker: {current_speaker}")
-        st.write(current_paragraph + ".")
-
     # Display the transcript
     st.header("Transcription")
-    st.write(full_transcript)
+    st.write(full_transcript)
+