Upload app.py

app.py

@@ -3,6 +3,7 @@ import streamlit as st
 import time
 import os
 
+
 os.environ['COQUI_TOS_AGREED'] = '1'
 os.environ['BARK_TOS_AGREED'] = '1'
 
@@ -18,9 +19,11 @@ def model_coqui():
 
 # @st.cache_resource()
 def model_bark():
-    device = "cpu"
-    tts = TTS("tts_models/multilingual/multi-dataset/bark").to(device)
-    return tts
+    from transformers import AutoProcessor, BarkModel
+    processor = AutoProcessor.from_pretrained("suno/bark")
+    model = BarkModel.from_pretrained("suno/bark")
+
+    return processor, model
 
 
 col1, col2 = st.columns(2)
@@ -29,11 +32,11 @@ with col1:
     sentence = st.text_area("Konuşmak istediğiniz metni girin")
 
 with col2:
-    st.selectbox("Dil Seçin", ["Ceylin", "Funda"])
-    if st.selectbox == "tr":
-        speaker_wav ="sounds/glinda.mp3"
+    language = st.selectbox("Dil Seçin", ["Ceylin", "Funda"])
+    if language == "Ceylin":
+        speaker_wav = "sounds/glinda.mp3"
     else:
-        speaker_wav_2 ="sounds/alp.mp3"
+        speaker_wav = "sounds/alp.mp3"
 
 speaker=st.button("Seslendir")
 
@@ -47,7 +50,7 @@ if speaker:
     with col11:
         tts_coqui = model_coqui()
         start_time_coqui = time.time()  # Start the timer
-        output_path_coqui = tts_coqui.tts_to_file(text=sentence, speaker_wav=speaker_wav_2 ,language="tr")
+        output_path_coqui = tts_coqui.tts_to_file(text=sentence, speaker_wav=speaker_wav ,language="tr")
         end_time = time.time()  # End the timer
 
         execution_time = round(end_time - start_time_coqui,2)  # Calculate the time taken
@@ -56,13 +59,19 @@ if speaker:
         st.audio(output_path_coqui, format="audio/mp3")
 
     with col22:
-        tts_bark = model_bark()
 
-        start_time_bark = time.time()
-        output_path_bark = tts_bark.tts_to_file(text=sentence, speaker_wav=speaker_wav, language="tr")
-        end_time = time.time()
+        processor, model = model_bark()
+        voice_preset = "tr_speaker_1"
+        inputs = processor(f"{sentence}", voice_preset=voice_preset)
+
+        start_time_bark = time.time()  # Start the timer
+        audio_array = model.generate(**inputs)
+        end_time = time.time()  # End the timer
+
+        audio_array = audio_array.cpu().numpy().squeeze()
+        sample_rate = model.generation_config.sample_rate
 
-        execution_time = round(end_time - start_time_bark,2)
-        st.write(f"Bark içi tamamlanma süresi: {execution_time} saniye")
+        execution_time = round(end_time - start_time_bark,2)  # Calculate the time taken
+        st.write(f"Bark için tamamlanma süresi: {execution_time} saniye")  # Display the time taken
 
-        st.audio(output_path_bark, format="audio/mp3")
+        st.audio(audio_array, format="audio/wav", sample_rate=sample_rate)