Nvidia Nemotron 3.5 Transcribe: open ASR for 40 languages

Re:Nvidia Nemotron 3.5 Transcribe: open ASR for 40 languages by Farmadupe2 on Wednesday October 28, 2026 @07:50PM (#50282593)

Almost all ML models come with reference python code that will run on most architectures.

If you want an implementation optimized for a certain architecture the cost is between 10-30USD of opus tokens and is oneshottable: ("implement Nvidia nemotron 3.5 transcribe for the intel npu on this machine. Use the hf transformers implementation as reference. The finished artifact takes pcm on stdin and utf8 chunks on stdout".)

You can substitute your old intel CPU by modifying the prompt by replacing "intel npu" with "the target cou on this machine ... Your task is complete when the model decode performance is 80% of the memory read speed of this machine"

[ Parent | on HN ]

Re:Nvidia Nemotron 3.5 Transcribe: open ASR for 40 languages by throwaway81523 on Thursday October 29, 2026 @12:42AM (#50284654)

Thanks, that's interesting, I hoped maybe it would use pytorch or something. But what I'm wondering is whether the result would run at usable speed. I don't currently use anything like opus and only want to use open weight models, and at the moment I don't have a GPU machine, though I guess I can spin up an hourly VM with one sometime.

[ Parent | on HN ]

Re:Nvidia Nemotron 3.5 Transcribe: open ASR for 40 languages by roel5059 on Thursday October 29, 2026 @06:04AM (#50286710)

Parakeet does really, really well in CPU only. I would try it first.

[ Parent | on HN ]