Skip Navigation

InitialsDiceBearhttps://github.com/dicebear/dicebearhttps://creativecommons.org/publicdomain/zero/1.0/„Initials” (https://github.com/dicebear/dicebear) by „DiceBear”, licensed under „CC0 1.0” (https://creativecommons.org/publicdomain/zero/1.0/)S

solrize

@ solrize @lemmy.ml

Posts
6
Comments
35
Joined
3 yr. ago

  • Current SOTA in local FOSS speech to text?

    Jump

  • Thanks, I'm more interested in quality than speed, but will try the smaller models to see if the results are usable. My CPU has Intel Integrated Graphics 4000 (this is in an i7-3770 which is way old by now) and I kind of doubt Pytorch supports that, though it's possible. I can imagine upgrading to a newer processor but probably not a real GPU. Actually it looks like everything is way more expensive than it used to be, no big surprise.

  • Current SOTA in local FOSS speech to text?

    Jump

  • Thanks, I don't have a GPU so if the program required one, I'd rent an hourly one (vast.ai has lots of affordable rentals). But it's easier if I can run cpu-only. If I were doing this a lot (I don't expect to), I could see getting a Ryzen APU server, if the GPU in those is supported.

  • Current SOTA in local FOSS speech to text?

    Jump

  • Thanks. My old i5-something server is probably in the same speed range as your laptop. It's good to hear about the transcription quality. If conversion is slower than real time, I can live with it. I can just throw a bunch of files at it and let it run overnight. Faster is always nicer of course.

  • Current SOTA in local FOSS speech to text?

    Jump

  • Thanks, I have heard of it and the configuration does look manageable. Do the more accurate models run at tolerable speed without a GPU? Or even with one for that matter? I'd hope to get transcriptions as good as Youtube's (they have lots of errors but they're fairly usable). Is that realistic?

  • Current SOTA in local FOSS speech to text?

    Jump

  • Thanks, that's interesting and I might try it, though it's an Android app. I was hoping for a Linux CLI tool that I could run on a server.

    Added: the phone app says it wants microphone and video permissions, which is a bit annoying. I don't care at all about live transcribing. I only want to convert files. It also wants network permissions and special protocols, which seems weird. I can understand if it wants to download models, but I wonder what else it wants the network for.

  • Free and Open Source Software @beehaw.org
    solrize @lemmy.ml

    Current SOTA in local FOSS speech to text?

  • Komi Store: Android app store to install and update apps hosted on GitHub, Codeberg & Forgejo

    Jump

  • I don't see the point then. I can install direct from GitHub if I want that. I don't want a random intermediary that's another possible attack vector.

  • Komi Store: Android app store to install and update apps hosted on GitHub, Codeberg & Forgejo

    Jump

  • What's wrong with F-droid?

  • There Ought be a Word for "FOSS" Projects Whose Infrastructure Involves a Lot of Proprietary Services

    Jump

  • "This program has features you might not like" --F-droid warning.

  • This AI Tool Rips Off Open Source Software Without Violating Copyright

    Jump

  • It's obviously not clean room since the original code was in the training data.

  • Not The Onion @lemmy.world
    solrize @lemmy.ml

    Someone may have maliciously caused a huge rodent invasion in California

    www.sfgate.com /bayarea/article/california-nutria-reintroduced-22198180.php

  • Request for messenger suggestions

    Jump

  • Well start with a few not-that-private letters to check for evidence of their being opened. What happens with ordinary email by the way?

  • Request for messenger suggestions

    Jump

  • I think it's hard, and even if there is something that works, its use can probably be detected somehow, and that could get your family in trouble.

    Tbh I'd probably use snail mail letters for anything private on the theory that the RU govt doesn't have the resources to open all the envelopes, and you can use special phrases for particularly private meanings. All that stuff like media attachments is asking for trouble. You could also send microSD cards by snail mail though that might attract attention.

    Remember that Osama bin Laden's compound in Pakistan had no internet connectivity at all. If he wanted to send an email, he'd write it to a USB drive and have a guy on a motorcycle take it to a café 70 km away or something like that. Replies would be brought to him the same way. They still managed to find him and kill him in his bedroom.

    Today with AI analysis of massive amounts of traffic logs, I'm sure signal ID is far easier than it was in 2011.

  • FEMA Official Says He Teleported to Waffle House. Experts Are Dubious.

    Jump

  • Experts are dubious that he teleported to Waffle House. CNN investigation reveals it was really an IHOP.

  • Data Finds Republicans are Obsessed with Searching for Transgender Porn

    Jump

  • Article says there are more searches for trans-related keywords in GOP areas. However it doesn't say anything about trans porn searches vs cis porn searches. It wouldn't surprise me if conservative areas have more searches for porn in general, so all the numbers are higher. Anyway the article uses a lot of statistical jargon without seeming to check this basic issue. FAIL.

  • Hidden Valley hiring “Ranch-bassadors” to travel Europe and test ranch on local food

    Jump

  • omg this is literally about salad dressing. The photo looks AI too.

  • Petition · Rename the 'Zavikon Island Bridge' to the 'President Donald J. Trump Bridge' - Canada · Change.org

    Jump

  • I'm unfamiliar with the Zavikon Island Bridge but if the renaming doesn't work out, the original Tacoma Narrows bridge could be another good candidate.

  • Vending machine in Hiroshima offers cans of air from the prefecture

    Jump

  • CAH Gives You Your Fucking Money Back

    Jump

  • The Man Who Stole Infinity | Quanta Magazine

    Jump

  • Disgraced Andrew insists royal staff call him 'sir' and fetch beer

    Jump

  • He is getting used to it I guess. "Sir" is a pretty big demotion from "your royal highness".

  • Science @mander.xyz
    solrize @lemmy.ml

    Incorrect Citation Association for Articles in Online-Only Springer Nature Journals

    arxiv.org /abs/2511.01675
  • Not The Onion @lemmy.world
    solrize @lemmy.ml

    Fact check: Is Mamdani introducing Arabic numerals to New York schools?

    www.newsweek.com /fact-check-mamdani-arabic-numerals-new-york-schools-11079540
  • News @lemmy.world
    solrize @lemmy.ml
    melatonin

    Common Supplement Shows Concerning Link to Heart Failure

    www.sciencealert.com /common-supplement-shows-concerning-link-to-heart-failure
  • Fuck AI @lemmy.world
    solrize @lemmy.ml

    arxiv.org stops accepting non-peer-reviewed/accepted CS reviews/position papers, due to AI slop

    mas.to /@carnage4life/115475618755048102