Creative Commons Attribution 4.0 International

The model checkpoints in this repository are distributed under the
Creative Commons Attribution 4.0 International License:

https://creativecommons.org/licenses/by/4.0/

These checkpoints are derived from:

NVIDIA Corporation
nvidia/stt_en_fastconformer_ctc_large
https://huggingface.co/nvidia/stt_en_fastconformer_ctc_large

The original checkpoint is licensed under CC BY 4.0.

Modifications, fine-tuning, sparse Mixture-of-Experts implementation,
experiments, and checkpoint preparation were performed by Amelia Sasin
for the MSc thesis "Mixture-of-Experts Architectures for Inclusive and
Robust Automatic Speech Recognition", TU Delft, 2026.

These derivative models are not endorsed by NVIDIA.