Create app.py

app.py

@@ -1,21 +1,36 @@
+import streamlit as st
 from speechbrain.pretrained import Tacotron2, HIFIGAN
 from scipy.io.wavfile import write
-import gradio as gr
 
-# Load TTS and vocoder models
-tacotron2 = Tacotron2.from_hparams(source="speechbrain/tts-tacotron2-ljspeech", savedir="tmpdir_tts")
-hifi_gan = HIFIGAN.from_hparams(source="speechbrain/tts-hifigan-ljspeech", savedir="tmpdir_vocoder")
+# Load the TTS and Vocoder models
+@st.cache_resource
+def load_models():
+    tacotron2 = Tacotron2.from_hparams(source="speechbrain/tts-tacotron2-ljspeech", savedir="tmpdir_tts")
+    hifi_gan = HIFIGAN.from_hparams(source="speechbrain/tts-hifigan-ljspeech", savedir="tmpdir_vocoder")
+    return tacotron2, hifi_gan
 
-# TTS function
+tacotron2, hifi_gan = load_models()
+
+# Text-to-Speech function
 def text_to_speech(text):
     mel_output, mel_length, alignment = tacotron2.encode_text(text)
     waveforms, _, _ = hifi_gan.decode_batch(mel_output)
-    audio_path = "tts_output.wav"
+    audio_path = "output.wav"
     write(audio_path, 22050, waveforms.squeeze(1).cpu().numpy())
     return audio_path
 
-# Gradio interface
-def chatbot_response(text):
-    return text_to_speech(text)
+# Streamlit App UI
+st.title("Text-to-Speech Chatbot")
+
+# Input text box
+text = st.text_input("Enter text to convert to speech:", "")
+
+if st.button("Generate Speech"):
+    if text.strip():
+        st.write("Generating speech...")
+        audio_file = text_to_speech(text)
+        st.audio(audio_file, format="audio/wav")
+    else:
+        st.warning("Please enter some text.")
 
-gr.Interface(fn=chatbot_response, inputs="text", outputs="audio").launch()
+st.write("Powered by SpeechBrain and Streamlit.")