Update app.py

app.py

@@ -6,8 +6,6 @@ from gtts import gTTS
 import tempfile
 import shutil
 import re
-import speech_recognition as sr  # Library for voice recognition
-import sounddevice as sd
 
 def translate_to_japanese(api_key, text):
     """
@@ -31,7 +29,7 @@ def translate_to_japanese(api_key, text):
     try:
         # Call the OpenAI API to get the Japanese translation
         response_translation = openai.ChatCompletion.create(
-            model="gpt-4",  # Ensure using the correct endpoint for the GPT model
+            model="gpt-4o",  # Use the correct endpoint for chat models
             messages=messages_translation,
             max_tokens=300,
             temperature=0.5
@@ -48,7 +46,7 @@ def translate_to_japanese(api_key, text):
         
         # Call the OpenAI API to get the pronunciation
         response_pronunciation = openai.ChatCompletion.create(
-            model="gpt-4",
+            model="gpt-4o",
             messages=messages_pronunciation,
             max_tokens=300,
             temperature=0.5
@@ -78,39 +76,12 @@ def generate_audio_from_text(text):
     tts.save(temp_audio_file.name)
     return temp_audio_file.name
 
-def transcribe_voice_input():
-    recognizer = sr.Recognizer()
-
-    # Set parameters for sounddevice (e.g., 44100 Hz, 1 channel)
-    duration = 5  # Seconds to record
-    fs = 44100  # Sampling frequency
-
-    st.info("Please speak now...")
-
-    # Record the audio using sounddevice
-    audio_data = sd.rec(int(duration * fs), samplerate=fs, channels=1, dtype='int16')
-    sd.wait()  # Wait until recording is finished
-
-    # Convert the numpy array to audio and recognize the speech
-    with sr.AudioData(audio_data.tobytes(), fs, 2) as source:
-        try:
-            text = recognizer.recognize_google(source)
-            st.success(f"Transcribed text: {text}")
-            return text
-        except sr.UnknownValueError:
-            st.error("Sorry, I could not understand the audio. Please try again.")
-            return None
-        except sr.RequestError:
-            st.error("Could not request results from Google Speech Recognition service.")
-            return None
-
 # Streamlit UI
 st.title("English to Japanese Translator with Pronunciation")
-st.markdown("Translate English text into Japanese and get its pronunciation (Romaji). You can input text or use voice.")
+st.markdown("Translate English text into Japanese and get its pronunciation (Romaji) using OpenAI's API.")
 
-# Display an image if you have one
 translateimg = Image.open("Untitled.png")  # Ensure the file is in the correct directory
-st.image(translateimg, use_container_width=True)
+st.image(translateimg, use_container_width=True)  # Adjust the size as per preference
 
 # Access the API key from Hugging Face Secrets
 api_key = os.getenv("OPENAI_API_KEY")
@@ -118,17 +89,19 @@ api_key = os.getenv("OPENAI_API_KEY")
 # Input field for the text
 english_text = st.text_area("Enter the English text to translate")
 
-# Button to trigger the translation from text input
-if st.button("Translate from Text"):
+# Button to trigger the translation
+if st.button("Translate"):
     if api_key and english_text:
         japanese_text, pronunciation = translate_to_japanese(api_key, english_text)
         if pronunciation:
+            # Clean pronunciation (remove unnecessary parts)
             cleaned_pronunciation = clean_pronunciation(pronunciation)
+
             st.markdown("### Translation Result:")
             st.write(f"**English Text:** {english_text}")
             st.write(f"**Japanese Output:** {japanese_text}")
             st.write(f"**Pronunciation:** {cleaned_pronunciation}")
-
+            
             # Save the result in a text file
             result_text = f"English Text: {english_text}\n\nJapanese Translation: {japanese_text}\nPronunciation: {cleaned_pronunciation}"
 
@@ -147,34 +120,16 @@ if st.button("Translate from Text"):
 
             # Generate audio for pronunciation
             audio_file_path = generate_audio_from_text(cleaned_pronunciation)
+
+            # Provide a button to play the pronunciation audio
             st.audio(audio_file_path, format="audio/mp3")
+
+            translateimg2 = Image.open("v3.png")  # Ensure the file is in the correct directory
+            st.image(translateimg2, width=150)  # Adjust the size as per preference
         else:
             st.error(japanese_text)  # Display error message if API call fails
     else:
         if not api_key:
             st.error("API key is missing. Please add it as a secret in Hugging Face Settings.")
         else:
-            st.error("Please provide text to translate.")
-
-# Button to trigger translation from voice input
-if st.button("Translate from Voice"):
-    if api_key:
-        voice_input = transcribe_voice_input()
-        if voice_input:
-            japanese_text, pronunciation = translate_to_japanese(api_key, voice_input)
-            if pronunciation:
-                cleaned_pronunciation = clean_pronunciation(pronunciation)
-                st.markdown("### Translation Result from Voice:")
-                st.write(f"**English Text (from Voice):** {voice_input}")
-                st.write(f"**Japanese Output:** {japanese_text}")
-                st.write(f"**Pronunciation:** {cleaned_pronunciation}")
-
-                # Generate audio for pronunciation
-                audio_file_path = generate_audio_from_text(cleaned_pronunciation)
-                st.audio(audio_file_path, format="audio/mp3")
-            else:
-                st.error(japanese_text)  # Display error message if API call fails
-        else:
-            st.error("Could not transcribe voice input.")
-    else:
-        st.error("API key is missing. Please add it as a secret in Hugging Face Settings.")
+            st.error("Please provide text to translate.")