AWATERA / app.py
denis-kazakov's picture
Upload 2 files
01a1731 verified
raw
history blame
899 Bytes
from transformers import pipeline
import gradio as gr
model = pipeline(
"automatic-speech-recognition",
model='openai/whisper-large-v3',
chunk_length_s=30,
generate_kwargs={"task": "transcribe"}
)
def transcribe_audio(mic=None, file=None, return_timestamps=False):
if mic is not None:
audio = mic
elif file is not None:
audio = file
else:
return "You must either provide a mic recording or a file"
result = model(audio, return_timestamps=return_timestamps, batch_size=8)
if return_timestamps:
return result['chunks']
else:
return result['text']
gr.Interface(
fn=transcribe_audio,
inputs=[
gr.Audio(sources=["microphone"], type="filepath"),
gr.Audio(sources=["upload"], type="filepath"),
gr.Checkbox(label="Add timestamps?")
],
outputs="text",
).launch()