first commit

app.py

@@ -0,0 +1,70 @@
+import os
+from huggingface_hub import notebook_login
+import gradio as gr
+from transformers import pipeline
+from deep_translator import GoogleTranslator
+from gtts import gTTS
+import shutil
+from pytube import YouTube
+
+os.environ["HF_TOKEN"] = "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"
+notebook_login()
+
+pipe = pipeline(model="nelanbu/ID2223_Lab2_Whisper")  # change to "your-username/the-name-you-picked"
+
+
+def transcribe(audio_input, link_input, lang):
+    try:
+        if link_input:
+          video=YouTube(link_input).streams.filter(only_audio=True).all()
+          audio=video[0].download()
+
+        elif audio_input:
+          audio = audio_input
+          # input_path = "input.mp3"
+          # shutil.copyfile(audio, input_path)
+          # print(f"Input audio is saved to {input_path}")
+
+        result = pipe(audio)
+        text = result['text']
+        print(f"Transcribed text: {text}")
+
+        if lang == 'english':
+          target_lang = 'en'
+        elif lang == 'swedish':
+          target_lang = 'sv'
+        elif lang == 'italian':
+          target_lang = 'it'
+        elif lang == 'german':
+          target_lang = 'de'
+        elif lang == 'french':
+          target_lang = 'fr'
+
+        translator = GoogleTranslator(source='auto', target=target_lang)
+        translated_text = translator.translate(text)
+        print(f"Translated text: {translated_text}")
+        # Convert translated text to speech
+        tts = gTTS(text=translated_text, tld='com', slow=False, lang=target_lang)
+        output_audio_path = "test2.mp3"
+        tts.save(output_audio_path)
+        print(f"Saved TTS audio to {output_audio_path}")
+
+        return output_audio_path
+    except Exception as e:
+        print(f"An error occurred: {e}")
+        raise
+
+demo = gr.Interface(
+    fn=transcribe,
+    inputs=[gr.Audio(type="filepath", label="Translate from microphone/MP3 input"),
+            gr.Text(max_lines=1, label="Translate from YouTube URL"),
+            gr.Dropdown(
+            ["english", "swedish", "italian", "german", "french"], label="Language", info="pick the language you want to translate your auido"),
+            ],
+    outputs=gr.Audio(type="filepath"),
+    title="Turkish Audio Translator",
+    description="You can upload YouTube link of a video in Turkish or use the microphone to record your voice or upload an MP3 file to translate Turkish audio to other languages."
+)
+
+if __name__ == "__main__":
+    demo.launch(debug=True)

requirements.txt

@@ -0,0 +1,6 @@
+gradio
+transformers
+gTTS
+deep-translator
+ttpx==0.24.1
+pytube