Upload 3 files

app.py

@@ -0,0 +1,136 @@
+import streamlit as st
+from engine import Translate
+import os
+import shutil
+import json
+
+translator = Translate()
+# Creating the list of the tasks the seamlessM4t model can perform
+tasks = ['S2ST (Speech to Speech translation)',
+        'S2TT (Speech to Text translation)',
+        'T2ST (Text to Speech translation)',
+        'T2TT (Text to Text translation)']
+
+def move_audio_file(source_path, destination_path):
+    try:
+        shutil.move(source_path, destination_path)
+        return True  # Return True if the move was successful
+    except Exception as e:
+        print(f"Error moving the audio file: {str(e)}")
+        return False  # Return False if there was an error
+
+# List of language options
+languages = [
+    "Catalan",
+    "Czech",
+    "Danish",
+    "Dutch",
+    "English",
+    "Estonian",
+    "Finnish",
+    "French",
+    "German",
+    "Hindi",
+    "Indonesian",
+    "Italian",
+    "Japanese",
+    "Korean",
+    "Maltese",
+    "Mandarin Chinese",
+    "Modern Standard Arabic",
+    "Northern Uzbek",
+    "Polish",
+    "Portuguese",
+    "Romanian",
+    "Russian",
+    "Slovak",
+    "Spanish",
+    "Swahili",
+    "Swedish",
+    "Tagalog",
+    "Telugu",
+    "Thai",
+    "Turkish",
+    "Ukrainian",
+    "Urdu",
+    "Vietnamese",
+    "Welsh",
+    "Western Persian"
+]
+
+st.title("Seamless-M4t Translator")
+task = st.selectbox('Select Task:', tasks)
+
+if task == 'S2ST (Speech to Speech translation)':
+    upload_audio = st.file_uploader('Upload Audio file', type=['mp3', 'wav', 'ogg'])
+    target_language = st.selectbox('Target language:', languages)
+
+    button_clicked = st.button('Convert')
+    if button_clicked:
+        if upload_audio and target_language:
+            result = translator.translate(task=task, audio=upload_audio, target_language=target_language)
+            
+            # Set the source and destination paths for moving the audio file
+            source_path = result.replace("\\", "/")
+            audio_paths = source_path.split(',')
+            for audio_path in audio_paths:
+                try:
+                    audio_path = audio_path.lstrip("['")
+                    st.audio(audio_path.strip('"'), format="audio/*")
+                    
+                except Exception as e:
+                    # audio_content = audio_path.lstrip(']')
+                    audio_content=audio_path.rstrip(']')
+                    st.write(audio_content)
+
+
+elif task == 'S2TT (Speech to Text translation)':
+    upload_audio = st.file_uploader('Upload Audio file', type=['mp3', 'wav', 'ogg'])
+    # Create a button
+    target_language = st.selectbox('Target language : ', languages)
+    button_clicked = st.button("Convert")
+
+    if button_clicked:
+        if upload_audio:
+            result = translator.translate(task=task, audio=upload_audio, target_language=target_language)
+            result = json.loads(result)
+            st.write(result)
+
+
+elif task == 'T2ST (Text to Speech translation)':
+    input_language = st.selectbox("Input Language:", languages)
+    text = st.text_area("Enter the Text Here:")
+    target_language = st.selectbox('Target language:', languages)
+
+    button_clicked = st.button('Convert')
+
+    if button_clicked:
+        if text and input_language and target_language:
+            result = translator.translate(task=task, text=text, input_language=input_language, target_language=target_language)
+            
+            # Set the source and destination paths for moving the audio file
+            source_path = result.replace("\\", "/")
+            audio_paths = source_path.split(',')
+            for audio_path in audio_paths:
+                try:
+                    audio_path = audio_path.lstrip("['")
+                    st.audio(audio_path.strip('"'), format="audio/wav")
+                    
+                except Exception as e:
+                    # audio_content = audio_path.lstrip(']')
+                    audio_content=audio_path.rstrip(']')
+                    st.write(audio_content)
+
+                
+elif task == 'T2TT (Text to Text translation)':
+    input_language = st.selectbox("Input Language:", languages)
+    text = st.text_area("Enter the Text Here:")
+    target_language = st.selectbox('Target language:', languages)
+
+    button_clicked = st.button('Convert')
+
+    if button_clicked:
+        if text and input_language and target_language:
+            result = translator.translate(task=task, text=text, input_language=input_language, target_language=target_language)
+            conv = json.loads(result)
+            st.write(conv[1])

engine.py

@@ -0,0 +1,48 @@
+from gradio_client import Client
+import json
+import base64
+
+class Translate:
+    
+    def translate(self, task=None, audio=None, text=None, input_language=None, target_language=None):
+        client = Client("https://facebook-seamless-m4t.hf.space/")
+        
+        audio_content = None  # Initialize to None
+        result = None  # Initialize result
+        
+        if audio is not None:
+            # Handle the uploaded audio file
+            audio_content = audio.read()  # Read the binary content of the uploaded audio
+            audio.close()  # Close the uploaded file
+            # Convert audio content to base64-encoded string
+            audio_content = base64.b64encode(audio_content).decode('utf-8')
+        
+            # Call the Gradio predict method and store the result
+            result = client.predict(
+                task,
+                audio_content,  # Pass the audio content as base64-encoded string
+                "https://github.com/gradio-app/gradio/raw/main/test/test_files/audio_sample.wav",
+                "https://github.com/gradio-app/gradio/raw/main/test/test_files/audio_sample.wav",
+                text,
+                input_language,
+                target_language,
+                api_name="/run"
+            )
+        else:
+            result = client.predict(
+                task,
+                audio_content,  # Pass the audio content as base64-encoded string
+                "https://github.com/gradio-app/gradio/raw/main/test/test_files/audio_sample.wav",
+                "https://github.com/gradio-app/gradio/raw/main/test/test_files/audio_sample.wav",
+                text,
+                input_language,
+                target_language,
+                api_name="/run"
+            )
+
+        
+
+        # Serialize the dictionary to a JSON-serializable string
+        result_str = json.dumps(result)
+        
+        return result_str