HackerTrans
トップ新着トレンドコメント過去質問紹介求人

mrkn1

255 カルマ登録 6 年前
@kouhxp on twitter and bluesky, follow for more

投稿

Self-proclaimed King of Switzerland uses loophole to build his empire for free

france24.com
4 ポイント·投稿者 mrkn1·14 日前·0 コメント

Self-proclaimed King of Switzerland uses loophole to build his empire for free

france24.com
6 ポイント·投稿者 mrkn1·15 日前·0 コメント

Show HN: Recreate Thinking Machines 276B voice demo with duct tape and 8B model

github.com
3 ポイント·投稿者 mrkn1·29 日前·0 コメント

Show HN: Realtime voice agent that sees, hears, and interrupts – on a CPU laptop

github.com
1 ポイント·投稿者 mrkn1·30 日前·1 コメント

Show HN: A CLI that fact-checks text using a 500MB local model and web search

github.com
1 ポイント·投稿者 mrkn1·先月·0 コメント

Show HN: Free summary, fact check, ELI5 of any text on CPU

github.com
2 ポイント·投稿者 mrkn1·先月·0 コメント

Show HN: fftext - quick summary, fact check, ELI5 on CPU

github.com
1 ポイント·投稿者 mrkn1·先月·0 コメント

Show HN: Fact check, summarize, explain any article or webpage on local CPU

github.com
5 ポイント·投稿者 mrkn1·先月·0 コメント

Show HN: Local-first fast CPU image to text for screenshots, PDFs, webpages

github.com
19 ポイント·投稿者 mrkn1·先月·17 コメント

Show HN: CPU-only fact-check, summarize, explain, translate any text

github.com
2 ポイント·投稿者 mrkn1·先月·0 コメント

Show HN: CPU-only fact-check, summarize, explain, translate anything

github.com
3 ポイント·投稿者 mrkn1·先月·0 コメント

Show HN: CPU-only fact-check, summarize, explain, translate anything

github.com
3 ポイント·投稿者 mrkn1·先月·0 コメント

Show HN: CPU-only fast OCR for screenshots, images, PDFs, webpages

github.com
9 ポイント·投稿者 mrkn1·先月·8 コメント

Show HN: CPU model for fact-checking, summarizing, explaining text locally

github.com
1 ポイント·投稿者 mrkn1·先月·0 コメント

Show HN: Local CPU OCR for images, PDFs, webpages

github.com
3 ポイント·投稿者 mrkn1·先月·0 コメント

Show HN: Local CPU model for fact-checking, summarizing, explaining text

github.com
6 ポイント·投稿者 mrkn1·先月·0 コメント

Show HN: CPU transcription for YouTube/TikTok/X, now 3x faster and diarization

github.com
3 ポイント·投稿者 mrkn1·先月·0 コメント

Advancing Mathematics Research with AI-Driven Formal Proof Search

arxiv.org
1 ポイント·投稿者 mrkn1·2 か月前·0 コメント

Show HN: CPU-only fast OCR for screenshots, images, PDFs, webpages

github.com
4 ポイント·投稿者 mrkn1·2 か月前·9 コメント

Show HN: CPU-only transcription for YouTube, TikTok, X, Instagram videos

github.com
97 ポイント·投稿者 mrkn1·2 か月前·49 コメント

コメント

mrkn1
·3 日前·議論
fwiw kimchi-derived probiotic bacterium (Leuconostoc mesenteroides CBA3656) was shown to bind nanoplastics and help mice excrete more of them. But it’s not yet proven that eating kimchi removes microplastics from humans
mrkn1
·13 日前·議論
Has anyone tried the others listed? Any feedback?
mrkn1
·18 日前·議論
This runs for free on CPU https://github.com/kouhxp/textsnap
mrkn1
·25 日前·議論
[flagged]
mrkn1
·26 日前·議論
The key moments feat is neat. Been working on a free opensource offline transcriber that runs fast on CPU and does diarization too

https://github.com/kouhxp/yapsnap
mrkn1
·先月·議論
No rigorous eval, and I love Tesseract. Here's the example that motivated me to build textsnap (which is in the github's README), parsed with Tesseract:

https://imgur.com/a/i2eQra8
mrkn1
·先月·議論
No reason other than their Q4 model working reasonably well and fast on my CPU laptop. Should work with any ONNX VLM model
mrkn1
·先月·議論
109 languages, including other alphabets.
mrkn1
·先月·議論
Just ran

  textsnap "https://i.ytimg.com/vi/LBNDfxjEYlA/maxresdefault.jpg"
and got this

  $('.count').each(function () {
  $('this').prop('Counter', 0).animate({
    Counter: $('this').text()
  }, {
      duration: 4000,
      easing: 'swing',
      step: 'function (now) {
          $('this").text(Math.ceil(now));
      }
    }); 
  });
mrkn1
·先月·議論
thank you!
mrkn1
·先月·議論
thank you! what is it about?
mrkn1
·先月·議論
I have not. My experience has been a few seconds for an 1024x1024 with medium density of text, FWIW. Feel free to try it on a few test images, model is pretty small and fast, but yeah no formal evals on CPU.
mrkn1
·先月·議論
It should support 109 languages. More info here: https://huggingface.co/PaddlePaddle/PaddleOCR-VL
mrkn1
·先月·議論
No, simply, my laptop only has a CPU.
mrkn1
·先月·議論
I haven't. But the evals of the underlying model are published here, including on Omnibench. https://huggingface.co/PaddlePaddle/PaddleOCR-VL-1.5
mrkn1
·先月·議論
https://github.com/kouhxp/fftext
mrkn1
·先月·議論
For 100% local CPU fact checking, I made this: https://news.ycombinator.com/item?id=48301003
mrkn1
·先月·議論
A lot of my queries are summarize/explain/fact check, and these are covered 100% on my CPU locally [0], reducing frontier model reliance

[0] https://news.ycombinator.com/item?id=48301003
mrkn1
·先月·議論
just released new version that implements your ideas
mrkn1
·先月·議論
new version released is up to 3x faster on CPU. Let me know!