Upload 2 files

app.py

@@ -0,0 +1,64 @@
+# app.py
+
+import os
+import streamlit as st
+import soundfile as sf
+from TTS.api import TTS
+
+# Load XTTSv2 model (Coqui TTS)
+st.session_state.setdefault("tts", TTS(model_name="tts_models/multilingual/multi-dataset/xtts_v2", gpu=True))
+
+st.set_page_config(page_title="🧬 XTTSv2 Voice Cloner", layout="centered")
+st.title("🎙️ Clone Your Voice with XTTSv2")
+st.markdown("Upload your voice sample and generate cloned speech from text.")
+
+# Step 1: Upload voice sample
+st.subheader("🎤 Upload Your Voice Sample (.wav)")
+speaker_wav = st.file_uploader("Upload a clean voice sample (10–30 seconds)", type=["wav","mp3"], key="uploader")
+
+# Step 2: Choose or write text
+st.subheader("✍️ Choose or Write Your Text")
+predefined_texts = {
+    "Greeting": "Hi there! I'm excited to talk to you today.",
+    "Hindi-English": "Namaste! Mera naam Tanu hai aur main AI ke saath kaam karti hoon.",
+    "Narration": "Once upon a time, in a quiet village, there lived a young explorer named Zoya.",
+    "Funny": "Why did the neural net go to therapy? It couldn't stop overfitting!"
+}
+choice = st.selectbox("Pick a predefined sentence:", list(predefined_texts.keys()), key="text_selector")
+text = st.text_area("Or write your own:", predefined_texts[choice], height=120, key="custom_text")
+
+# Step 3: Generate and play audio
+generate_clicked = st.button("🔊 Generate Voice", key="generate_button")
+
+if generate_clicked and speaker_wav is not None and text.strip():
+    with st.spinner("Cloning your voice and synthesizing..."):
+        # Save uploaded speaker wav locally
+        with open("uploaded_speaker.wav", "wb") as f:
+            f.write(speaker_wav.read())
+
+        output_path = "xtts_output.wav"
+        st.session_state.tts.tts_to_file(
+            text=text,
+            speaker_wav="uploaded_speaker.wav",
+            language="hi",  # Use "hi" for Hindi-only text
+            file_path=output_path
+        )
+
+        st.success("✅ Voice cloned successfully!")
+        st.audio(output_path)
+
+elif generate_clicked:
+    st.warning("⚠️ Please upload a voice sample and enter some text.")
+
+# Optional demo samples
+st.markdown("---")
+st.subheader("🎧 Demo Samples (Optional)")
+
+demo_folder = "demo_samples"
+if os.path.exists(demo_folder):
+    demo_files = [f for f in os.listdir(demo_folder) if f.endswith(".wav")]
+    for demo in demo_files:
+        st.markdown(f"**{demo}**")
+        st.audio(os.path.join(demo_folder, demo))
+else:
+    st.info("📁 No demo samples found. Add `.wav` files in the `demo_samples/` folder.")

requirements.txt

@@ -0,0 +1,8 @@
+streamlit==1.33.0
+TTS==0.22.0
+torch==1.10.0  
+torchaudio==0.10.0 
+soundfile==0.12.1
+numpy>=1.21.0
+scipy>=1.5.0
+numba==0.58.1