stt_en_fastconformer_ctc_large
NVIDIA🇺🇸 USMajor lab·
Audioopen weights
English speech recognition model (CTC, FastConformer architecture, around 115M parameters) that transcribes speech into lowercase letters. The model card reports 2.1 percent WER on LibriSpeech test-clean with greedy decoding.
Summary of the model card
- License
- cc-by-4.0
- Modalities
- text → text
1 variant
| Large | HF |
Sources
HuggingFaceDeutsch
Englischsprachiges Spracherkennungsmodell mit FastConformer Architektur und CTC Decoder (rund 115 Mio. Parameter), das Sprache in Kleinbuchstaben transkribiert. Die Modelcard nennt 2,1 Prozent WER auf LibriSpeech test-clean bei Greedy Decoding.
Benchmarks
| Benchmark | Domain | Value | Evidence |
|---|---|---|---|
| Earnings-22Open ASR Leaderboard (Langform) | audio-stt | 24.29 % WER | 68 |
| Open ASR LeaderboardOpen ASR Leaderboard (Langform) | audio-stt | 21.52 % WER | 58 |
| TED-LIUM 3Open ASR Leaderboard (Langform) | audio-stt | 5.26 % WER | 49 |
Vendor-reported numbers are marked and count with factor 0.7 in rankings. The evidence score says how much a benchmark still tells you today.
Provenance
- Detected
- Aug 25, 2026
- First source
- openasr
- Description
- Summary of the model card
- Editorially reviewed
- —