Hacker News: morrisjm

New comment by morrisjm in "Show HN: AsteroidOS 2.0 – Nobody asked, we shipped anyway"

morrisjm — Thu, 19 Feb 2026 00:39:38 +0000

Turns out you can play Asteroids on AsteroidOS https://github.com/moWerk/asteroid-blaster

New comment by morrisjm in "VALL-E: Microsoft’s new zero-shot text-to-speech model"

morrisjm — Mon, 09 Jan 2023 19:48:06 +0000

https://en.wikipedia.org/wiki/List_of_Nvidia_graphics_proces...

New comment by morrisjm in "VALL-E: Microsoft’s new zero-shot text-to-speech model"

morrisjm — Mon, 09 Jan 2023 18:09:48 +0000

rtx4000 only has 8gig memory which means reducing the batch size (much slowness) and/or how much text you can give it at once (meaning you have to break up text chunks not at sentence breaks)

rtx5000 maybe but not sure how much of a value improvement there is

New comment by morrisjm in "VALL-E: Microsoft’s new zero-shot text-to-speech model"

morrisjm — Mon, 09 Jan 2023 18:02:21 +0000

fixed thx

New comment by morrisjm in "VALL-E: Microsoft’s new zero-shot text-to-speech model"

morrisjm — Mon, 09 Jan 2023 17:54:53 +0000

James Betker in the tortoise-tts repo, which is similar, says he spent $15k for his home rig. I'm not finding right now how long it took to train the tortoise model but feel like I read him say weeks/months somewhere. Obvs all kinds of variations depending on coding efficiency and dataset size, but another datapoint https://nonint.com/2022/05/30/my-deep-learning-rig/ https://github.com/neonbjb/tortoise-tts

New comment by morrisjm in "VALL-E: Microsoft’s new zero-shot text-to-speech model"

morrisjm — Mon, 09 Jan 2023 16:55:39 +0000

you could also just skip step 1 https://twitter.com/lexman_ai

New comment by morrisjm in "VALL-E: Microsoft’s new zero-shot text-to-speech model"

morrisjm — Mon, 09 Jan 2023 16:52:24 +0000

New comment by morrisjm in "VALL-E: Microsoft’s new zero-shot text-to-speech model"

morrisjm — Mon, 09 Jan 2023 16:51:03 +0000

Open source tortoise-TTS has been able to do this for 6+ months now, which is also based on the same theory as DALL-E. From playing with tortoise a bit over the last couple of weeks it seems like the issue is not so much accuracy anymore, rather how GPU intensive it is to make a voice of any meaningful duration. Tortoise is ~5 seconds on a $1000 GPU (P5000) to do one second of spoken text. There's cloud options (collab, paperspace, runpod) but still https://github.com/neonbjb/tortoise-tts

New comment by morrisjm in "Squirt.io – Readability Meets Spritz Speed Reading"

morrisjm — Thu, 13 Mar 2014 16:37:30 +0000

http://endsoftpatents.org perhaps?