Spaces:
Running
Running
import gradio as gr | |
from transformers import pipeline | |
import os | |
import azure.cognitiveservices.speech as speechsdk | |
dialects = {"Palestinian/Jordanian": "P", "Syrian": "S", "Lebanese": "L", "Egyptian": "E"} | |
translator_en2ar = pipeline(task="translation", model="guymorlan/English2Dialect") | |
translator_ar2en = pipeline(task="translation", model="guymorlan/Shami2English") | |
transliterator = pipeline(task="translation", model="guymorlan/DialectTransliterator") | |
speech_config = speechsdk.SpeechConfig(subscription=os.environ.get('SPEECH_KEY'), region=os.environ.get('SPEECH_REGION')) | |
def translate_english(input_text, include): | |
if not input_text: | |
return "", "", "", "" | |
inputs = [f"{val} {input_text}" for val in dialects.values()] | |
sy, lb, eg = "Syrian" in include, "Lebanese" in include, "Egyptian" in include | |
# remove 2nd element if sy is false | |
if not eg: | |
inputs.pop() | |
if not lb: | |
inputs.pop() | |
if not sy: | |
inputs.pop() | |
result = translator_en2ar(inputs) | |
pal_out = result[0]["translation_text"] | |
sy_out = result[1]["translation_text"] if sy else "" | |
lb_out = result[1 + sy]["translation_text"] if lb else "" | |
eg_out = result[1 + sy + lb]["translation_text"] if eg else "" | |
return pal_out, sy_out, lb_out, eg_out | |
def translate_arabic(input_text): | |
if not input_text: | |
return "" | |
result = translator_ar2en([input_text]) | |
return result[0]["translation_text"] | |
def get_audio(input_text): | |
audio_config = speechsdk.audio.AudioOutputConfig(filename=f"{input_text}.wav") | |
speech_config.speech_synthesis_voice_name='ar-SY-AmanyNeural' | |
speech_synthesizer = speechsdk.SpeechSynthesizer(speech_config=speech_config, audio_config=audio_config) | |
speech_synthesis_result = speech_synthesizer.speak_text_async(input_text).get() | |
return f"{input_text}.wav" | |
def get_transliteration(input_text, include=["Transliteration"]): | |
if "Transliteration" not in include: | |
return "" | |
result = transliterator([input_text]) | |
return result[0]["translation_text"] | |
css = """ | |
#liter textarea, #trans textarea { font-size: 25px;} | |
#trans textarea { direction: rtl; }; | |
""" | |
with gr.Blocks(title = "English to Levantine Arabic", css=css, theme="default") as demo: | |
gr.Markdown("# Levantine Arabic Translator") | |
with gr.Tab('En -> Ar'): | |
with gr.Row(): | |
with gr.Column(): | |
input_text = gr.Textbox(label="Input", placeholder="Enter English text", lines=1) | |
gr.Examples(["I wanted to go to the store yesterday, but it rained", "How are you feeling today?", "Let's drink coffee"], input_text) | |
btn = gr.Button("Translate", label="Translate") | |
with gr.Row(): | |
include = gr.CheckboxGroup(["Transliteration", "Syrian", "Lebanese", "Egyptian"], | |
label="Disable features to speed up translation", | |
value=["Transliteration", "Syrian", "Lebanese", "Egyptian"]) | |
gr.Markdown("Built by [Guy Mor-Lan](mailto:guy.mor@mail.huji.ac.il). Pronunciation model is specifically tailored to urban Palestinian Arabic. Text-to-speech uses Microsoft Azure's API and may provide different result from the transliterated pronunciation.") | |
with gr.Column(): | |
pal = gr.Textbox(lines=1, label="Palestinian", elem_id="trans") | |
pal_translit = gr.Textbox(lines=1, label="Palestinian Pronunciation", elem_id="liter") | |
sy = gr.Textbox(lines=1, label="Syrian", elem_id="trans") | |
lb = gr.Textbox(lines=1, label="Lebanese", elem_id="trans") | |
eg = gr.Textbox(lines=1, label="Egyptian", elem_id="trans") | |
with gr.Row(): | |
audio = gr.Audio(label="Audio - Palestinian", interactive=False) | |
audio_button = gr.Button("Get Audio", label="Click Here to Get Audio") | |
audio_button.click(get_audio, inputs=[pal], outputs=[audio]) | |
btn.click(translate_english,inputs=[input_text, include], outputs=[pal, sy, lb, eg]) | |
input_text.submit(translate_english, inputs=[input_text, include], outputs=[pal, sy, lb, eg]) | |
pal.change(get_transliteration, inputs=[pal, include], outputs=[pal_translit]) | |
with gr.Tab('Ar -> En'): | |
with gr.Row(): | |
with gr.Column(): | |
input_text = gr.Textbox(label="Input", placeholder="Enter Levantine Arabic text", lines=1, elem_id="trans") | |
gr.Examples(["ุฎูููุง ูุฏูุฑ ุนูู ู ุทุนู ุชุงูู", "ูุงู ุจุฏู ุงููู ุงุดู ูุจู ู ุง ูุฑูุญ"], input_text) | |
btn = gr.Button("Translate", label="Translate") | |
gr.Markdown("Built by [Guy Mor-Lan](mailto:guy.mor@mail.huji.ac.il).") | |
with gr.Column(): | |
eng = gr.Textbox(label="English", lines=1, elem_id="liter") | |
btn.click(translate_arabic,inputs=input_text, outputs=[eng]) | |
with gr.Tab("Transliterate"): | |
with gr.Row(): | |
with gr.Column(): | |
input_text = gr.Textbox(label="Input", placeholder="Enter Levantine Arabic text", lines=1) | |
gr.Examples(["ุฎูููุง ูุฏูุฑ ุนูู ู ุทุนู ุชุงูู", "ูุงู ุจุฏู ุงููู ุงุดู ูุจู ู ุง ูุฑูุญ"], input_text) | |
btn = gr.Button("Transliterate", label="Transliterate") | |
gr.Markdown("Built by [Guy Mor-Lan](mailto:guy.mor@mail.huji.ac.il)") | |
with gr.Column(): | |
translit = gr.Textbox(label="Transliteration", lines=1, elem_id="liter") | |
btn.click(get_transliteration, inputs=input_text, outputs=[translit]) | |
demo.launch() | |