matthewagi
/

HeAR-s

Feature Extraction

vision-transformer

Model card Files Files and versions

HeAR-s / preprocessor_config.json

matthewagi's picture

Initial upload of distilled ViT-S Canon model

73629b7 verified 8 months ago

Raw History Blame Contribute Delete

259 Bytes

	{
	"clip_duration_seconds": 2.0,
	"do_normalize": false,
	"feature_extractor_type": "HearCanonFeatureExtractor",
	"image_size": [
	192,
	128
	],
	"input_channels": 1,
	"num_audio_samples": 32000,
	"num_mel_bins": 128,
	"sampling_rate": 16000
	}