Update app.py

app.py

@@ -1,32 +1,20 @@
-import sounddevice as sd
-from scipy.io.wavfile import write
 import whisper
-import numpy as np
 import streamlit as st
 
-# Streamlit interface
 st.title("Audio Transcription with Whisper")
 
-# Record audio
-def record_audio():
-    fs = 44100
-    seconds = st.slider("Select recording duration (seconds):", 1, 10, 8)
-    st.write("Recording Audio - Speak now!")
-    myrecording = sd.rec(int(seconds * fs), samplerate=fs, channels=2)
-    sd.wait()
-    write('output.mp3', fs, myrecording)
-    st.audio('output.mp3', format='audio/mp3')
-    st.write("Audio recording complete, play audio")
+# File uploader for audio
+uploaded_file = st.file_uploader("Upload an audio file", type=["wav", "mp3", "m4a"])
 
-# Transcribe audio
-def transcribe_audio():
-    model = whisper.load_model("base")
+if uploaded_file is not None:
+    with open("output.mp3", "wb") as f:
+        f.write(uploaded_file.getbuffer())
+    st.audio("output.mp3")
 
-    # Load audio and pad/trim it to fit 30 seconds
+    # Transcribe audio
+    model = whisper.load_model("base")
     audio = whisper.load_audio("output.mp3")
     audio = whisper.pad_or_trim(audio)
-
-    # Make log-Mel spectrogram and move to the same device as the model
     mel = whisper.log_mel_spectrogram(audio).to(model.device)
 
     # Detect the spoken language
@@ -39,16 +27,3 @@ def transcribe_audio():
 
     # Display the recognized text
     st.write(result.text)
-
-# Main Streamlit application
-def main():
-    st.header("Audio Recorder")
-    if st.button("Start Recording"):
-        record_audio()
-    
-    st.header("Transcription")
-    if st.button("Transcribe Audio"):
-        transcribe_audio()
-
-if __name__ == "__main__":
-    main()