app.py

#import librosa
import torch
from transformers import Wav2Vec2ForCTC, Wav2Vec2Tokenizer
import streamlit as st
from audio_recorder_streamlit import audio_recorder

audio_bytes = audio_recorder(pause_threshold=3.0, sample_rate=16_000)
if audio_bytes:
    st.audio(audio_bytes, format="audio/wav")
    
#load pre-trained model and tokenizer
tokenizer = Wav2Vec2Tokenizer.from_pretrained("facebook/wav2vec2-base-960h")
model = Wav2Vec2ForCTC.from_pretrained("facebook/wav2vec2-base-960h")

#load audio file
#speech, rate = librosa.load("/hip-voice.m4a",sr=16000)

#import IPython.display as display
#display.Audio("batman1.wav", autoplay=True)

input_values = tokenizer(audio_bytes, return_tensors = 'pt').input_values

#input_values = tokenizer(speech, return_tensors = 'pt').input_values
logits = model(input_values).logits

predicted_ids = torch.argmax(logits, dim =-1)

#decode the audio to generate text
transcriptions = tokenizer.decode(predicted_ids[0])

print(transcriptions)