Automatic Speech Recognition
NeMo
PyTorch
English
speech
audio
Transducer
TDT
FastConformer
Conformer
NeMo
hf-asr-leaderboard
Eval Results (legacy)
Eval Results
Instructions to use nvidia/parakeet-tdt-0.6b-v2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- NeMo
How to use nvidia/parakeet-tdt-0.6b-v2 with NeMo:
import nemo.collections.asr as nemo_asr asr_model = nemo_asr.models.ASRModel.from_pretrained("nvidia/parakeet-tdt-0.6b-v2") transcriptions = asr_model.transcribe(["file.wav"]) - Notebooks
- Google Colab
- Kaggle
Ignores repeated words
#8
by thecraig - opened
Unlike nvidia/parakeet-tdt-1.1b which has excellent accuracy, this version fails in detecting repeated words or phrases (false starts and disfluencies). Not sure if that's intended, but for absolute accuracy, this is a downgrade from tdt-1.1b. The word level timestamps however seem improved.
Thanks for the feedback!
Any plan to support this? btw This is an excellent model! in terms of speed and accuracy