HF Model & Benchmark Matrix

24,078 variants · 14,984 scores · 120 benchmarks

parakeet-tdt-1.1b

nvidia/parakeet-tdt-1.1b

kind
original
quant
bf16 (16-bit)
family
misc
modality
automatic-speech-recognition
root model
nvidia/parakeet-tdt-1.1b
provider
nvidia
license
cc-by-4.0
language
en
context
—
library
nemo
vRAM (GB)
—
size on disk
—
downloads
5,523
likes
136
created
—
last modified
2025-12-03T22:27:57+00:00 (2025)
huggingface
https://huggingface.co/nvidia/parakeet-tdt-1.1b
tags
nemoautomatic-speech-recognitionspeechaudioTransducerTDTFastConformerConformerpytorchNeMohf-asr-leaderboardendataset:librispeech_asrdataset:fisher_corpusdataset:Switchboard-1dataset:WSJ-0dataset:WSJ-1dataset:National-Singapore-Corpus-Part-1dataset:National-Singapore-Corpus-Part-6dataset:vctkdataset:voxpopulidataset:europarldataset:multilingual_librispeechdataset:mozilla-foundation/common_voice_8_0

Benchmark scores (8)

benchmarkcategoryscorecontextconfevidence
WERspeech-asr17.18% WERmedium| % WER | 17.18 | 14.61 | 19.06 | 37.57 |
WER AMIspeech-asr15.9low| 1.22.0 | SentencePiece Unigram | 1024 | 15.90 | 14.65 | 9.55 | 1.39 | 2.62 | 3.42 | 3.56 | 5.48 | 5.97 |
WER Earnings-22speech-asr14.65low| 1.22.0 | SentencePiece Unigram | 1024 | 15.90 | 14.65 | 9.55 | 1.39 | 2.62 | 3.42 | 3.56 | 5.48 | 5.97 |
WER GigaSpeechspeech-asr9.55low| 1.22.0 | SentencePiece Unigram | 1024 | 15.90 | 14.65 | 9.55 | 1.39 | 2.62 | 3.42 | 3.56 | 5.48 | 5.97 |
WER Common Voicespeech-asr5.48low| 1.22.0 | SentencePiece Unigram | 1024 | 15.90 | 14.65 | 9.55 | 1.39 | 2.62 | 3.42 | 3.56 | 5.48 | 5.97 |
WER VoxPopulispeech-asr3.56low| 1.22.0 | SentencePiece Unigram | 1024 | 15.90 | 14.65 | 9.55 | 1.39 | 2.62 | 3.42 | 3.56 | 5.48 | 5.97 |
WER TED-LIUMspeech-asr3.42low| 1.22.0 | SentencePiece Unigram | 1024 | 15.90 | 14.65 | 9.55 | 1.39 | 2.62 | 3.42 | 3.56 | 5.48 | 5.97 |
WER SPGISpeechspeech-asr2.62low| 1.22.0 | SentencePiece Unigram | 1024 | 15.90 | 14.65 | 9.55 | 1.39 | 2.62 | 3.42 | 3.56 | 5.48 | 5.97 |

Discussions (0)

none scraped.

Hugging Face model & benchmark data scraped & matrix-ified. 24,078 variants · 14,984 scores · 120 benchmarks.