wav2vec2-large-xlsr-53-th

Home/Models/airesearch/wav2vec2-large-xlsr-53-th

airesearch/wav2vec2-large-xlsr-53-th

kind: original
quant: bf16 (16-bit)
family: misc
modality: automatic-speech-recognition
root model: —
provider: airesearch
license: cc-by-sa-4.0
language: —
context: —
library: —
vRAM (GB): —
size on disk: —
downloads: 1,065,202
likes: 28
created: —
last modified: 2022-03-23T18:24:45+00:00 (2022)
huggingface: https://huggingface.co/airesearch/wav2vec2-large-xlsr-53-th
tags: transformerspytorchwav2vec2automatic-speech-recognitionaudiohf-asr-leaderboardrobust-speech-eventspeechxlsr-fine-tuningthdataset:common_voicedoi:10.57967/hf/0404license:cc-by-sa-4.0model-indexendpoints_compatibleregion:us

benchmark	category	score	context	conf	evidence
WER	speech-asr	95.24		low	`\| Only Tokenization \| 0.9524% \| 2.5316% \| 1.2346% \| 0.1623% \|`
WER Common Voice	speech-asr	23.04	[Kaldi from scratch](https://github.com/vistec-AI/commonvoice-th)	medium	`\| [Kaldi from scratch](https://github.com/vistec-AI/commonvoice-th) \| 23.04 \| \| 7.57 \|`
CER	speech-asr	16.23		low	`\| Only Tokenization \| 0.9524% \| 2.5316% \| 1.2346% \| 0.1623% \|`
WER Common Voice	speech-asr	7.0		medium	`Common Voice 7.0`

none scraped.