Update app.py

app.py

@@ -1,13 +1,8 @@
-from fastapi import FastAPI, File, UploadFile
-import uvicorn
-from pydantic import BaseModel
+import gradio as gr
 import whisper
 from transformers import MarianMTModel, MarianTokenizer
 import subprocess
 import os
-from pathlib import Path
-
-app = FastAPI()
 
 # Load models
 def load_models():
@@ -23,66 +18,84 @@ def load_models():
         lang: MarianTokenizer.from_pretrained(f"Helsinki-NLP/opus-mt-{lang}-en")
         for lang in translation_models.keys()
     }
-    return translation_models, translation_tokenizers
 
-translation_models, translation_tokenizers = load_models()
+load_models()
 
-# Whisper endpoint
-@app.post("/transcribe")
-async def transcribe(file: UploadFile = File(...), language: str = "en"):
+# Transcribe function
+def transcribe_audio(file, language="en"):
     try:
-        # Save the file temporarily
-        temp_file = f"temp/{file.filename}"
-        Path("temp").mkdir(parents=True, exist_ok=True)
-        with open(temp_file, "wb") as f:
-            f.write(await file.read())
-
-        # Transcription using Whisper
-        result = whisper_model.transcribe(temp_file, language=language)
+        result = whisper_model.transcribe(file, language=language)
         transcription = result["text"]
-        os.remove(temp_file)  # Clean up
-        return {"success": True, "transcription": transcription}
-
+        return transcription
     except Exception as e:
-        return {"success": False, "error": str(e)}
+        return f"Error: {str(e)}"
 
-# Translation endpoint
-@app.post("/translate")
-async def translate(text: str, source_lang: str, target_lang: str):
+# Translate function
+def translate_text(text, source_lang, target_lang="en"):
     try:
         if source_lang not in translation_models or target_lang != "en":
-            return {"success": False, "error": "Unsupported language."}
+            return "Unsupported language."
 
-        # Tokenize and translate
         tokenizer = translation_tokenizers[source_lang]
         model = translation_models[source_lang]
         inputs = tokenizer(text, return_tensors="pt", padding=True)
         translated_tokens = model.generate(**inputs)
         translated_text = tokenizer.decode(translated_tokens[0], skip_special_tokens=True)
-        return {"success": True, "translation": translated_text}
-
+        return translated_text
     except Exception as e:
-        return {"success": False, "error": str(e)}
+        return f"Error: {str(e)}"
 
-# TTS endpoint
-@app.post("/tts")
-async def text_to_speech(text: str, speaker: str = "male", speed: str = "normal"):
+# Text-to-Speech function
+def text_to_speech(text, speaker="male", speed="normal"):
     try:
         output_file = "output.wav"
-
-        # Coqui TTS command
         tts_command = [
             "tts",
             f"--text={text}",
             "--model_name=tts_models/en/ljspeech/tacotron2-DCA",
             f"--out_path={output_file}",
         ]
-        subprocess.run(tts_command)
+        subprocess.run(tts_command, check=True)
+        return output_file
+    except Exception as e:
+        return f"Error: {str(e)}"
 
-        return {"success": True, "audio_file": output_file}
+# Gradio Interface
+def tts_interface(text):
+    audio_file = text_to_speech(text)
+    return audio_file if isinstance(audio_file, str) and os.path.exists(audio_file) else None
 
-    except Exception as e:
-        return {"success": False, "error": str(e)}
+with gr.Blocks() as demo:
+    gr.Markdown("### Audio Transcription, Translation, and TTS App")
+    
+    # Transcription section
+    with gr.Row():
+        with gr.Column():
+            audio_input = gr.Audio(label="Upload Audio File", type="file")
+            lang_input = gr.Dropdown(["en", "es", "fr", "ar"], label="Language", value="en")
+            transcribe_btn = gr.Button("Transcribe")
+        transcription_output = gr.Textbox(label="Transcription Output")
+
+    transcribe_btn.click(transcribe_audio, inputs=[audio_input, lang_input], outputs=transcription_output)
+
+    # Translation section
+    with gr.Row():
+        with gr.Column():
+            text_input = gr.Textbox(label="Input Text", lines=3)
+            source_lang = gr.Dropdown(["en", "es", "fr", "ar"], label="Source Language", value="en")
+            translate_btn = gr.Button("Translate")
+        translation_output = gr.Textbox(label="Translation Output")
+
+    translate_btn.click(translate_text, inputs=[text_input, source_lang], outputs=translation_output)
+
+    # TTS section
+    with gr.Row():
+        with gr.Column():
+            tts_input = gr.Textbox(label="Text for TTS", lines=2)
+            tts_btn = gr.Button("Generate Audio")
+        tts_output = gr.Audio(label="Generated Audio")
+
+    tts_btn.click(tts_interface, inputs=tts_input, outputs=tts_output)
 
-if __name__ == "__main__":
-    uvicorn.run(app, host="0.0.0.0", port=8000)
+# Launch Gradio App
+demo.launch()