moonshine-streaming-medium

Home/Models/UsefulSensors/moonshine-streaming-medium

UsefulSensors/moonshine-streaming-medium

kind: original
quant: bf16 (16-bit)
family: misc
modality: automatic-speech-recognition
root model: UsefulSensors/moonshine-streaming-medium
provider: UsefulSensors
license: mit
language: en
context: —
library: transformers
vRAM (GB): —
size on disk: —
downloads: 6,713
likes: 30
created: —
last modified: 2026-02-11T21:06:28+00:00 (2026)
huggingface: https://huggingface.co/UsefulSensors/moonshine-streaming-medium
tags: transformerssafetensorsmoonshine_streamingautomatic-speech-recognitionenlicense:miteval-resultsendpoints_compatibleregion:us

benchmark	category	score	context	conf	evidence
WER Earnings-22	speech-asr	20.27	Earnings-22	medium	`\| Earnings-22 \| 20.27 \| 13.53 \| 11.90 \|`
WER AMI	speech-asr	19.03	AMI	medium	`\| AMI \| 19.03 \| 12.54 \| 10.68 \|`
WER VoxPopuli	speech-asr	14.02	VoxPopuli	medium	`\| VoxPopuli \| 14.02 \| 9.98 \| 8.54 \|`
WER GigaSpeech	speech-asr	13.9	GigaSpeech	medium	`\| GigaSpeech \| 13.90 \| 10.41 \| 9.46 \|`
WER LibriSpeech	speech-asr	12.09	LibriSpeech (other)	medium	`\| LibriSpeech (other) \| 12.09 \| 6.78 \| 5.00 \|`
WER SPGISpeech	speech-asr	6.16	SPGISpeech	medium	`\| SPGISpeech \| 6.16 \| 3.19 \| 2.58 \|`
WER TED-LIUM	speech-asr	6.12	TED-LIUM	medium	`\| TED-LIUM \| 6.12 \| 3.77 \| 2.99 \|`
WER LibriSpeech	speech-asr	4.49	LibriSpeech (clean)	medium	`\| LibriSpeech (clean) \| 4.49 \| 2.49 \| 2.08 \|`

none scraped.