Create README.md

README.md

@@ -0,0 +1,13 @@
+---
+pipeline_tag: audio-classification
+---
+This model is an ONNX version of a YAMNet-based classifier trained to recognize environmental and human-made sounds (e.g., speech, music, silence, barking, etc.).
+
+Format: ONNX (.onnx)
+
+Input: float32 mono audio, sampled at 16,000 Hz
+
+Output: Score matrix per class [num_frames, num_classes]
+
+Classes: See mapping in yamnet_class_map.csv
+