Upload README.md with huggingface_hub

README.md

@@ -20,4 +20,24 @@ This repository contains the ONNX version of the SAM-Audio-Small model.
 
 ## Usage
 
-These models are intended to be used with the `sam-audio` inference pipeline.
+These models are intended to be used with the included `onnx_inference.py` script.
+
+### Installation
+
+```bash
+pip install onnxruntime transformers numpy librosa soundfile
+```
+
+### Running Inference
+
+```bash
+python onnx_inference.py --audio test_audio.wav --text "a person speaking" --output separated.wav
+```
+
+### End-to-End Test
+
+You can also run the end-to-end test comparing against the PyTorch implementation (requires `torch` and `sam-audio` installed):
+
+```bash
+python test_e2e.py --save-outputs
+```