Upload 4 files

README.md

@@ -0,0 +1,3 @@
+# Multilingual Realtime Translator
+
+Translate between English and Nigerian languages using speech and text.

gradio_app.py

@@ -0,0 +1,21 @@
+
+import gradio as gr
+from transformers import pipeline, MarianMTModel, MarianTokenizer
+
+asr = pipeline("automatic-speech-recognition", model="openai/whisper-small")
+mt_model_name = "Helsinki-NLP/opus-mt-yo-en"
+tokenizer = MarianTokenizer.from_pretrained(mt_model_name)
+model = MarianMTModel.from_pretrained(mt_model_name)
+
+def translate_speech(audio):
+    transcription = asr(audio)["text"]
+    inputs = tokenizer(transcription, return_tensors="pt", padding=True)
+    translated = model.generate(**inputs)
+    translation = tokenizer.decode(translated[0], skip_special_tokens=True)
+    return transcription, translation
+
+iface = gr.Interface(fn=translate_speech,
+                     inputs=gr.Audio(source="microphone", type="filepath"),
+                     outputs=["text", "text"],
+                     title="Yoruba to English Speech Translator")
+iface.launch()

main.py

@@ -0,0 +1,42 @@
+
+import speech_recognition as sr
+from transformers import pipeline
+import edge_tts
+import asyncio
+
+# Load Whisper model
+asr = pipeline("automatic-speech-recognition", model="openai/whisper-small")
+
+# Load translation model (Yoruba → English)
+from transformers import MarianMTModel, MarianTokenizer
+mt_model_name = "Helsinki-NLP/opus-mt-yo-en"
+tokenizer = MarianTokenizer.from_pretrained(mt_model_name)
+model = MarianMTModel.from_pretrained(mt_model_name)
+
+# TTS
+async def speak(text):
+    communicate = edge_tts.Communicate(text, "en-US-GuyNeural")
+    await communicate.save("output.mp3")
+
+def translate_text(text):
+    inputs = tokenizer(text, return_tensors="pt", padding=True)
+    translated = model.generate(**inputs)
+    return tokenizer.decode(translated[0], skip_special_tokens=True)
+
+# Real-time mic input
+recognizer = sr.Recognizer()
+with sr.Microphone() as source:
+    print("Speak now...")
+    audio = recognizer.listen(source)
+    print("Processing...")
+
+    # Speech to Text
+    result = asr(audio.get_wav_data())["text"]
+    print("Transcribed:", result)
+
+    # Translate
+    translation = translate_text(result)
+    print("Translated:", translation)
+
+    # Speak
+    asyncio.run(speak(translation))

requirements.txt

@@ -0,0 +1,6 @@
+transformers
+torch
+openai-whisper
+gradio
+edge-tts
+speechrecognition