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_ctc-110m with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- NeMo
How to use nvidia/parakeet-tdt_ctc-110m with NeMo:
import nemo.collections.asr as nemo_asr asr_model = nemo_asr.models.ASRModel.from_pretrained("nvidia/parakeet-tdt_ctc-110m") transcriptions = asr_model.transcribe(["file.wav"]) - Notebooks
- Google Colab
- Kaggle
Can we use this without going through Nemo?
#1
by bhavnicksm - opened
I want to use this model through huggingface without the dependencies of Nemo -- if it's open sourced it should be usable via huggingface directly or even pytorch.
But there are no instructions on this.
Hey I am looking for the same answer. Currently implementation is kind of bloated.
I might now even need GPU support for inferencing.
Is there any way to run this without nemo??