feat: add example script for CUDA usage and update README instructions

Author: thewh1teagle

README.md

@@ -36,9 +36,9 @@ pip install -U kokoro-onnx
 
 1. Install [uv](https://docs.astral.sh/uv/getting-started/installation) for isolated Python (Recommend).
 
-Basically open the terminal (PowerShell / Bash) and run the command listed in their website.
-
-_Note: you don't have to use `uv`. but it just make things much simpler. You can use regular Python as well._
+```console
+pip install uv
+```
 
 2. Create new project folder (you name it)
 3. Run in the project folder

examples/with_cuda.py

@@ -0,0 +1,35 @@
+"""
+Note:
+    On Linux you need to run this as well: apt-get install portaudio19-dev
+    gpu version is sufficient only for Linux and Windows. macOS works with GPU by default.
+    You can see the used execution provider by enable debug log. see with_log.py
+    The script tested on CUDA 12.1 with CUDNN 9.1.0 on RTX4060 TI
+    You might need to install the current CUDA version along with the CUDNN version (12.1, 9.1.0)
+    See https://developer.nvidia.com/cuda-12-1-0-download-archive and https://developer.nvidia.com/cudnn-9-1-0-download-archive
+
+Setup:
+    pip install -U kokoro-onnx[gpu] soudfile
+    wget https://github.com/thewh1teagle/kokoro-onnx/releases/download/model-files-v1.0/kokoro-v1.0.onnx
+    wget https://github.com/thewh1teagle/kokoro-onnx/releases/download/model-files-v1.0/voices-v1.0.bin
+
+Run with Python:
+    python examples/with_cuda.py
+
+Run with uv (if you cloned the repo):
+    uv run --extra gpu ./examples/with_cuda.py 
+"""
+
+import soundfile as sf
+from kokoro_onnx import Kokoro
+import onnxruntime as ort
+
+privders = ort.get_available_providers()
+print("Available providers:", privders)  # Make sure CUDAExecutionProvider is listed
+print(f'Is CUDA available: {"CUDAExecutionProvider" in privders}')
+
+kokoro = Kokoro("kokoro-v1.0.onnx", "voices-v1.0.bin")
+samples, sample_rate = kokoro.create(
+    "Hello. This audio generated by kokoro!", voice="af_sarah", speed=1.0, lang="en-us"
+)
+sf.write('audio.wav', samples, sample_rate)
+print('Created audio.wav')