Update app.py

app.py

@@ -1,54 +1,74 @@
-import json
-import torch
 import gradio as gr
+import json
 from transformers import pipeline
+import torch
 from gtts import gTTS
+import tempfile
 import os
 
-# Load FLORES-200 language codes from lang_code.json
+# Load language codes from JSON
 with open("lang_code.json", "r", encoding="utf-8") as f:
-    LANG_CODES = json.load(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
 
-# Translation pipeline
-translator = pipeline("translation", model="facebook/nllb-200-distilled-600M")
+# Load translation pipeline
+translator = pipeline(
+    "translation",
+    model="facebook/nllb-200-distilled-600M",
+    device=0 if torch.cuda.is_available() else -1
+)
 
-# Translation + TTS function
-def translate_and_speak(text, src_lang, tgt_lang):
-    try:
-        src_code = LANG_CODES[src_lang]
-        tgt_code = LANG_CODES[tgt_lang]
+# 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']
 
-        # Translate
-        result = translator(text, src_lang=src_code, tgt_lang=tgt_code)
-        translated_text = 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"]
 
-        # Generate speech for translated text
-        tts = gTTS(translated_text, lang=tgt_lang.split("_")[0])
-        audio_path = "translated_audio.mp3"
-        tts.save(audio_path)
+    # Translate the text
+    translated_text = translate_text(transcription, src_lang, tgt_lang)
 
-        return translated_text, audio_path
-    except Exception as e:
-        return f"Error: {e}", None
+    # 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(theme="soft") as demo:
-    gr.HTML("<h1 style='color:black; text-align:center;'>@GenAi MultiLanguage Translator Model</h1>")
-
-    with gr.Row():
-        with gr.Column():
-            src_text = gr.Textbox(label="Enter Text", placeholder="Type something here...")
-            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")
-            translate_btn = gr.Button("Translate")
-
-        with gr.Column():
-            output_text = gr.Textbox(label="Translated Text")
-            output_audio = gr.Audio(label="Speech Output", type="filepath")
-
-    translate_btn.click(fn=translate_and_speak,
-                        inputs=[src_text, src_lang, tgt_lang],
-                        outputs=[output_text, output_audio])
-
-# Run app
-demo.launch()
+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()