nvidia

/

stt_en_fastconformer_hybrid_large_streaming_multi

like 27

Follow

NVIDIA 67.2k

We have plans to release models on other languages. What languages you are looing for?

Hi @vnoroozi ; thanks for the reply! In addition to English, we're working with Filipino (Tagalog), Hindi and Spanish.

@vnoroozi is there a plan for supporting custom language models ?

any future arabic version?
am struggling on building the right tokenizer from scratch and training recipes

Hi @vnoroozi ; thanks for the reply! In addition to English, we're working with Filipino (Tagalog), Hindi and Spanish.

hey! we've also been planning to do something similar! i would love to know if u guys did any work in finetuning this model or any tips ! would appreciate anything!

@vnoroozi Enabling the model to support batch inference with streaming transcription would be very useful, in my opinion.😃