Update README.md

README.md

@@ -1,42 +1,25 @@
----
-title: Nnamdi TTS
-emoji: 🗣️
-colorFrom: indigo
-colorTo: blue
-sdk: gradio
-sdk_version: 5.29.0
-app_file: app.py
-pinned: false
-license: apache-2.0
----
+# Clone TTS Model
 
-# Nnamdi Text-to-Speech App
+This project is a simple **Text-to-Speech (TTS)** model built using TensorFlow, trained on the **Clone dataset**.
 
-This is a custom TTS (Text-to-Speech) app built using the Coqui XTTS v2 model and a custom voice dataset hosted on Hugging Face.
+### Files
+- `train.py`: Script to train the model.
+- `infer.py`: Script to generate speech from text.
+- `app.py`: Gradio interface for live testing.
+- `data/`: Dataset folder containing audio files and transcriptions.
+- `model/`: Folder to save the trained model.
 
-## Dataset
+### How to Train
+1. Prepare your dataset in the `data/` folder.
+2. Run `train.py` to train the model.
+3. The trained model will be saved in the `model/` folder.
 
-- **Name**: `Emmylahot12/nnamdi`
-- Contains one voice sample: `data/train/yourvoice.wav`
-- Matched with transcription in `data/train/transcriptions.csv`
+### How to Use Inference
+1. Run `infer.py` to generate speech from text.
+2. Optionally, use `app.py` for a Gradio interface.
 
-## Model
+### Requirements
+- `requirements.txt`: Contains all the Python dependencies needed.
 
-- Uses the multilingual Coqui XTTS v2 model
-- Clones the voice from the uploaded dataset sample
-- Runs inference on CPU (no GPU required)
-
-## How It Works
-
-1. Loads `yourvoice.wav` from the dataset
-2. Uses it as reference voice for synthesis
-3. Takes input text and language selection
-4. Outputs generated speech using Gradio
-
-## Usage
-
-Just type your text, select a language, and hear the result.
-
-## License
-
-Apache-2.0
+### License
+This project is licensed under [CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/).