Automatic Speech Recognition
NeMo
PyTorch
English
speech
audio
CTC
FastConformer
Transformer
NeMo
hf-asr-leaderboard
Eval Results (legacy)
Eval Results
Instructions to use nvidia/stt_en_fastconformer_ctc_large with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- NeMo
How to use nvidia/stt_en_fastconformer_ctc_large with NeMo:
import nemo.collections.asr as nemo_asr asr_model = nemo_asr.models.ASRModel.from_pretrained("nvidia/stt_en_fastconformer_ctc_large") transcriptions = asr_model.transcribe(["file.wav"]) - Notebooks
- Google Colab
- Kaggle
KenLM Integration
#1
by puvvadasaikiran - opened
can we use this model with KenLM
is the size of emissions constant?
Yes you can use kenlm with any of the ctc models
Yes you can use kenlm with any of the ctc models
Hey @smajumdar94 , any particular Decoder implementation you would suggest?
have tested with pyctcdecoder(kenshoLM) and ctc_decoder(pytorch) but results are pretty bad when compared to a existing greedy decoder