Spaces:
Runtime error
Runtime error
File size: 2,217 Bytes
3c4ad65 19f09f4 71a635a 1d97cff 6134bda 1d97cff 3c4ad65 1d97cff 3c4ad65 d2cfc63 3c4ad65 d2cfc63 3c4ad65 d2cfc63 3c4ad65 1d97cff 3c4ad65 1d97cff 19f09f4 3c4ad65 1d97cff 3c4ad65 1d97cff 3c4ad65 d2cfc63 3c4ad65 d2cfc63 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 |
# Transform an audio to text script with language detection.
# Author: Pratiksha Patel
# Description: This script record the audio, transform it to text, detect the language of the file and save it to a txt file.
# import required modules
import os
import streamlit as st
from audio_recorder_streamlit import audio_recorder
from langdetect import detect
import numpy as np
# Use a pipeline as a high-level helper
#from transformers import pipeline
#pipe = pipeline("automatic-speech-recognition", model="openai/whisper-large")
# Load model directly
from transformers import AutoProcessor, AutoModelForSpeechSeq2Seq
def transcribe_audio(audio_bytes):
processor = AutoProcessor.from_pretrained("openai/whisper-large")
model = AutoModelForSpeechSeq2Seq.from_pretrained("openai/whisper-large")
audio_array = np.frombuffer(audio_bytes, dtype=np.int16)
# Cast audio array to double precision and normalize
audio_tensor = torch.tensor(audio_array, dtype=torch.float64) / 32768.0
input_values = processor(audio_tensor, return_tensors="pt", sampling_rate=16000).input_values
logits = model(input_values).logits
predicted_ids = torch.argmax(logits, dim=-1)
transcription = processor.decode(predicted_ids[0])
return transcription
# Function to open a file
#def startfile(fn):
# os.system('open %s' % fn)
# Function to create and open a txt file
#def create_and_open_txt(text, filename):
# Create and write the text to a txt file
# with open(filename, "w") as file:
# file.write(text)
# startfile(filename)
# Streamlit app
st.title("Audio to Text Transcription..")
audio_bytes = audio_recorder(pause_threshold=3.0, sample_rate=16_000)
if audio_bytes:
st.audio(audio_bytes, format="audio/wav")
transcription = transcribe_audio(audio_bytes)
if transcription:
st.write("Transcription:")
st.write(transcription)
else:
st.write("Error: Failed to transcribe audio.")
else:
st.write("No audio recorded.")
# Detect the language
#language = detect(transcription)
#st.write(f"Detected language: {language}")
# Create and open a txt file with the text
#create_and_open_txt(transcription, f"output_{language}.txt")
|