fastconformer_rnnt

Model type	fastconformer_rnnt
Class	EncDecRNNTModel
Task	fastconformer-rnnt
Source	models/nemo_rnnt.py

Description

NeMo FastConformer-RNNT (transducer) speech-to-text model.

Holds the three sub-modules (:attr:encoder, :attr:prediction,

attr:

joint) that the :class:~mobius.tasks._rnnt.RNNTTask wires into three separate ONNX graphs.

Usage

mobius build --model <MODEL_ID> output_dir/

from mobius import build

model = build("<MODEL_ID>")