Upload README.md with huggingface_hub

README.md

@@ -0,0 +1,64 @@
+---
+license: mit
+tags:
+  - audio
+  - source-separation
+  - drums
+  - onnx
+pipeline_tag: audio-to-audio
+---
+
+# DrumSep ONNX
+
+Drum element separation model converted to ONNX format for use with Splitzo.
+
+## Model Details
+
+- **Original Author**: [inagoy/drumsep](https://github.com/inagoy/drumsep)
+- **License**: MIT
+- **Format**: ONNX
+- **Size**: ~335 MB
+
+## Stems
+
+Separates drum tracks into 4 elements:
+1. **Kick** (bombo)
+2. **Snare** (redoblante)
+3. **Cymbals** (platillos)
+4. **Toms**
+
+## Technical Specs
+
+| Parameter | Value |
+|-----------|-------|
+| Sample Rate | 44100 Hz |
+| Segment Length | 40 seconds (1,764,000 samples) |
+| STFT n_fft | 4096 |
+| STFT hop_length | 1024 |
+
+## Usage
+
+This model is automatically downloaded by the Splitzo app on first use.
+
+**Input**: Pre-separated drum stem (stereo, 44.1kHz)
+**Output**: 4 separated drum elements
+
+## Attribution
+
+This ONNX conversion is based on the original DrumSep model by [@inagoy](https://github.com/inagoy).
+
+```
+MIT License
+
+Copyright (c) 2023 inagoy
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software...
+```
+
+## Conversion
+
+Converted from PyTorch checkpoint using `tools/convert_drumsep_onnx.py` with STFT operations moved outside the model for ONNX compatibility.