File size: 2,910 Bytes
68f0220 42b415b 68f0220 42b415b 52c0097 42b415b 68f0220 42b415b 0fed57d 42b415b 68f0220 42b415b 68f0220 42b415b 68f0220 42b415b 68f0220 42b415b 68f0220 42b415b 52c0097 0fed57d 42b415b |
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 64 65 66 67 68 69 70 71 72 73 74 75 |
import gradio as gr
import json
from transformers import pipeline
import torch
from gtts import gTTS
import tempfile
import os
# Load language codes from JSON
with open("lang_code.json", "r", encoding="utf-8") as f:
data = json.load(f)
# Convert list → dict if needed
if isinstance(data, list):
LANG_CODES = {item["Language"]: item["FLORES-200 code"] for item in data}
else:
LANG_CODES = data
# Load translation pipeline
translator = pipeline(
"translation",
model="facebook/nllb-200-distilled-600M",
device=0 if torch.cuda.is_available() else -1
)
# Text translation function
def translate_text(text, src_lang, tgt_lang):
src_code = LANG_CODES[src_lang]
tgt_code = LANG_CODES[tgt_lang]
result = translator(text, src_lang=src_code, tgt_lang=tgt_code)
return result[0]['translation_text']
# Speech translation function
def translate_speech(audio, src_lang, tgt_lang):
if audio is None:
return None, "Please provide an audio file."
# Convert speech to text (using Whisper)
asr = pipeline("automatic-speech-recognition", model="openai/whisper-base")
transcription = asr(audio)["text"]
# Translate the text
translated_text = translate_text(transcription, src_lang, tgt_lang)
# Convert translated text to speech
tts = gTTS(text=translated_text, lang='en') # gTTS uses ISO-639-1, adjust as needed
temp_file = tempfile.NamedTemporaryFile(delete=False, suffix=".mp3")
tts.save(temp_file.name)
return temp_file.name, translated_text
# Gradio UI
with gr.Blocks() as demo:
gr.Markdown("## 🌐 Multilingual Text & Speech Translator")
with gr.Tab("Text Translation"):
src_lang = gr.Dropdown(choices=list(LANG_CODES.keys()), value="English (Latin script)", label="Source Language")
tgt_lang = gr.Dropdown(choices=list(LANG_CODES.keys()), value="French", label="Target Language")
input_text = gr.Textbox(label="Enter text to translate")
output_text = gr.Textbox(label="Translated text")
translate_btn = gr.Button("Translate")
translate_btn.click(translate_text, inputs=[input_text, src_lang, tgt_lang], outputs=output_text)
with gr.Tab("Speech Translation"):
src_lang_s = gr.Dropdown(choices=list(LANG_CODES.keys()), value="English (Latin script)", label="Source Language")
tgt_lang_s = gr.Dropdown(choices=list(LANG_CODES.keys()), value="French", label="Target Language")
audio_input = gr.Audio(sources=["microphone", "upload"], type="filepath")
audio_output = gr.Audio(label="Translated Speech")
translated_text_output = gr.Textbox(label="Translated Text")
translate_speech_btn = gr.Button("Translate Speech")
translate_speech_btn.click(translate_speech, inputs=[audio_input, src_lang_s, tgt_lang_s], outputs=[audio_output, translated_text_output])
if __name__ == "__main__":
demo.launch()
|