Design of the IBM Granite 5.0 TurboCTC ASR Model

arXiv:2609.20104 · cs.CL · Submitted 2026-09-17 · Read on arXiv

cs.CL

Submitted: 2026-09-17

Updated: 2026-09-17

Comments: 5 pages, 2 figures, submitted to ICASSP 2027

Code: https://github.com/lucidrains/conformer

License: http://creativecommons.org/licenses/by-sa/4.0/

The gist: We describe the architecture, training methodology and inference speedups of Granite 5.0 Turbo CTC, a 470 million parameter encoder-only model with an excellent speed-accuracy tradeoff.

Terminology

Abstract

We describe the architecture, training methodology and inference speedups of Granite 5.0 Turbo CTC, a 470 million parameter encoder-only model with an excellent speed-accuracy tradeoff. The architecture uses pyramidal temporal subsampling within Conformer blocks using strided depthwise convolutions, block-diagonal (chunk-wise) self-attention, and conditioning on intermediate predictions from the middle layer. Training highlights are the use of only publicly available data, the novel use of a Muon optimizer, and balanced data sampling. Inference speedups include replacing 1 x 1 convolutions with linear layers and optimizing the attention computation in the Conformer blocks. Collectively, these result in a model that is on the speed-accuracy Pareto frontier of the Open ASR leaderboard for English short-form ASR while being twice as fast as the fastest competitor. The model can be used under a permissive license and downloaded from https://huggingface.co/ibm-granite/granite-speech-5.0-470m-turboctc.

Sources

Related papers