escucha / app.py
avilaroman's picture
Update app.py
eb17f5b
raw
history blame contribute delete
No virus
3.33 kB
import torch
import gradio as gr
import pytube as pt
from transformers import pipeline
from huggingface_hub import model_info
MODEL_NAME = "openai/whisper-small" #脡sto es la base de con lo que trabaja el script entonces hay que cargarlo en la linea 8.
lang = "es"
device = 0 if torch.cuda.is_available() else "cpu"
pipe = pipeline(
task="automatic-speech-recognition",
model=MODEL_NAME,
chunk_length_s=30,
device=device,
)
pipe.model.config.forced_decoder_ids = pipe.tokenizer.get_decoder_prompt_ids(language=lang, task="transcribe")
def transcribe(microphone, file_upload):
warn_output = ""
if (microphone is not None) and (file_upload is not None):
warn_output = (
"Consejo: Subiste un Archivo de Audio y usado el micr贸fono. "
"el Archivo Grabado del Microfono Local ser谩 enviado a Analizar. OK!\n"
)
elif (microphone is None) and (file_upload is None):
return "Advertencia: Tu tienes los 2 usa el Micr贸fono o un archivo subido desde tu PC."
file = microphone if microphone is not None else file_upload
text = pipe(file)["text"]
return warn_output + text
def _return_yt_html_embed(yt_url):
video_id = yt_url.split("?v=")[-1]
HTML_str = (
f'<center> <iframe width="500" height="320" src="https://www.youtube.com/embed/{video_id}"> </iframe>'
" </center>"
)
return HTML_str
def yt_transcribe(yt_url):
yt = pt.YouTube(yt_url)
html_embed_str = _return_yt_html_embed(yt_url)
stream = yt.streams.filter(only_audio=True)[0]
stream.download(filename="audio.mp3")
text = pipe("audio.mp3")["text"]
return html_embed_str, text
demo = gr.Blocks()
mf_transcribe = gr.Interface(
fn=transcribe,
inputs=[
gr.inputs.Audio(source="microphone", type="filepath", optional=True),
gr.inputs.Audio(source="upload", type="filepath", optional=True),
],
outputs="text",
layout="horizontal",
theme="huggingface",
title="Escucha",
description=(
"Transcribe Largas capturas de audio de Radio 160.310 Mgh por micr贸fono o grabaciones desde archivos a 1 click y gratis! 脡sta version est谩 utilizando por defecto fine-tuned para el mejor aprovechamiento del contexto."
f" checkpoint [{MODEL_NAME}](https://huggingface.co/{MODEL_NAME}) and 馃 Transformers to transcribe audio files"
" of arbitrary length."
),
allow_flagging="never",
)
yt_transcribe = gr.Interface(
fn=yt_transcribe,
inputs=[gr.inputs.Textbox(lines=1, placeholder="Paste the URL to a YouTube video here", label="YouTube URL")],
outputs=["html", "text"],
layout="horizontal",
theme="huggingface",
title="Escucha: Transcribe los videos de YouTube",
description=(
"Transcribe Largas capturas de audio de Radio 160.310 Mgh por micr贸fono o grabaciones desde archivos a 1 click y gratis! 脡sta version est谩 utilizando por defecto fine-tuned para el mejor aprovechamiento del contexto."
f" [{MODEL_NAME}](https://huggingface.co/{MODEL_NAME}) and 馃 Transformers to transcribe audio files of"
" arbitrary length."
),
allow_flagging="never",
)
with demo:
gr.TabbedInterface([mf_transcribe, yt_transcribe], ["Transcribe Audio", "Transcribe YouTube"])
demo.launch(enable_queue=True)