hubert

Model type	hubert
Class	Wav2Vec2Model
Task	audio-feature-extraction
Source	models/wav2vec2.py

Description

Wav2Vec2 encoder-only audio model.

Architecture: CNN feature extractor → feature projection → transformer encoder. Used for ASR (CTC), audio feature extraction, and audio classification.

Usage

mobius build --model <MODEL_ID> output_dir/

from mobius import build

model = build("<MODEL_ID>")