search

NVIDIA / NeMo

A scalable generative AI framework built for researchers and developers working on Large Language Models, Multimodal, and Speech AI (Automatic Speech Recognition and Text-to-Speech)
https://docs.nvidia.com/nemo-framework/user-guide/latest/overview.html
Apache License 2.0
10.96k stars 2.29k forks source link

FastConformer-Longformer #9434

Open GrigorKirakosyan opened 1 month ago

GrigorKirakosyan commented 1 month ago

Hi NeMo team, Do you plan to release En Large FastConformer-Long-CTC-BPE model trained with Local Attention and Global token?

nithinraok commented 1 month ago

we released 1B model but not the large model here: https://huggingface.co/spaces/nvidia/parakeet-tdt_ctc-1.1b

GrigorKirakosyan commented 1 month ago

Thanks, I have seen this. I was interested to know about Large-FastConformer(~120M) model trained with this config: https://github.com/NVIDIA/NeMo/blob/main/examples/asr/conf/fastconformer/long_fastconformer/fast-conformer-long_ctc_bpe.yaml