frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Show HN: LocalGPT – A local-first AI assistant in Rust with persistent memory

https://github.com/localgpt-app/localgpt
167•yi_wang•6h ago•58 comments

Haskell for all: Beyond agentic coding

https://haskellforall.com/2026/02/beyond-agentic-coding
82•RebelPotato•5h ago•21 comments

SectorC: A C Compiler in 512 bytes (2023)

https://xorvoid.com/sectorc.html
273•valyala•14h ago•52 comments

OpenClaw Is Changing My Life

https://reorx.com/blog/openclaw-is-changing-my-life/
16•novoreorx•1h ago•28 comments

Software factories and the agentic moment

https://factory.strongdm.ai/
213•mellosouls•16h ago•360 comments

LLMs as the new high level language

https://federicopereiro.com/llm-high/
82•swah•4d ago•151 comments

Speed up responses with fast mode

https://code.claude.com/docs/en/fast-mode
172•surprisetalk•13h ago•172 comments

LineageOS 23.2

https://lineageos.org/Changelog-31/
20•pentagrama•2h ago•0 comments

Hoot: Scheme on WebAssembly

https://www.spritely.institute/hoot/
185•AlexeyBrin•19h ago•35 comments

The world heard JD Vance being booed at the Olympics. Except for viewers in USA

https://www.theguardian.com/sport/2026/feb/07/jd-vance-boos-winter-olympics
72•treetalker•39m ago•15 comments

Brookhaven Lab's RHIC concludes 25-year run with final collisions

https://www.hpcwire.com/off-the-wire/brookhaven-labs-rhic-concludes-25-year-run-with-final-collis...
76•gnufx•12h ago•60 comments

Stories from 25 Years of Software Development

https://susam.net/twenty-five-years-of-computing.html
178•vinhnx•16h ago•18 comments

The Architecture of Open Source Applications (Volume 1) Berkeley DB

https://aosabook.org/en/v1/bdb.html
12•grep_it•5d ago•0 comments

Vocal Guide – belt sing without killing yourself

https://jesperordrup.github.io/vocal-guide/
337•jesperordrup•1d ago•102 comments

First Proof

https://arxiv.org/abs/2602.05192
139•samasblack•16h ago•81 comments

Show HN: I saw this cool navigation reveal, so I made a simple HTML+CSS version

https://github.com/Momciloo/fun-with-clip-path
89•momciloo•13h ago•18 comments

Substack confirms data breach affects users’ email addresses and phone numbers

https://techcrunch.com/2026/02/05/substack-confirms-data-breach-affecting-email-addresses-and-pho...
34•witnessme•3h ago•10 comments

Wood Gas Vehicles: Firewood in the Fuel Tank (2010)

https://solar.lowtechmagazine.com/2010/01/wood-gas-vehicles-firewood-in-the-fuel-tank/
39•Rygian•2d ago•13 comments

uLauncher

https://github.com/jrpie/launcher
11•dtj1123•4d ago•0 comments

Vouch

https://twitter.com/mitchellh/status/2020252149117313349
86•chwtutha•4h ago•23 comments

Al Lowe on model trains, funny deaths and working with Disney

https://spillhistorie.no/2026/02/06/interview-with-sierra-veteran-al-lowe/
109•thelok•15h ago•24 comments

Start all of your commands with a comma (2009)

https://rhodesmill.org/brandon/2009/commands-with-comma/
593•theblazehen•3d ago•216 comments

Show HN: A luma dependent chroma compression algorithm (image compression)

https://www.bitsnbites.eu/a-spatial-domain-variable-block-size-luma-dependent-chroma-compression-...
42•mbitsnbites•3d ago•6 comments

The AI boom is causing shortages everywhere else

https://www.washingtonpost.com/technology/2026/02/07/ai-spending-economy-shortages/
318•1vuio0pswjnm7•20h ago•523 comments

FDA intends to take action against non-FDA-approved GLP-1 drugs

https://www.fda.gov/news-events/press-announcements/fda-intends-take-action-against-non-fda-appro...
117•randycupertino•9h ago•245 comments

OpenCiv3: Open-source, cross-platform reimagining of Civilization III

https://openciv3.org/
909•klaussilveira•1d ago•277 comments

Where did all the starships go?

https://www.datawrapper.de/blog/science-fiction-decline
165•speckx•4d ago•247 comments

Selection rather than prediction

https://voratiq.com/blog/selection-rather-than-prediction/
37•languid-photic•4d ago•18 comments

Show HN: Look Ma, No Linux: Shell, App Installer, Vi, Cc on ESP32-S3 / BreezyBox

https://github.com/valdanylchuk/breezydemo
305•isitcontent•1d ago•39 comments

Unseen Footage of Atari Battlezone Arcade Cabinet Production

https://arcadeblogger.com/2026/02/02/unseen-footage-of-atari-battlezone-cabinet-production/
149•videotopia•4d ago•49 comments
Open in hackernews

Show HN: Chonky – a neural text semantic chunking goes multilingual

https://huggingface.co/mirth/chonky_mmbert_small_multilingual_1
43•hessdalenlight•3mo ago
TLDR: I’m expanding the family of text-splitting Chonky models with new multilingual model.

You can learn more about this neural approach in a previous post: https://news.ycombinator.com/item?id=43652968

Since the release of the first distilbert-based model I’ve released two more models based on a ModernBERT. All these models were pre-trained and fine-tuned primary on English texts.

But recently mmBERT(https://huggingface.co/blog/mmbert) has been released. This model pre-trained on massive dataset that contains 1833 languages. So I had an idea of fine-tuning a new multilingual Chonky model.

I’ve expanded training dataset (that previously contained bookcorpus and minipile datasets) with Project Gutenberg dataset which contains books in some widespread languages.

To make the model more robust for real-world data I’ve removed punctuation for last word for every training chunk with probability of 0.15 (no ablation was made for this technique though).

The hard part is evaluation. The real-world data are typically OCR'ed markdown, transcripts of calls, meeting notes etc. and not a clean book paragraphs. I didn’t find such labeled datasets. So I used what I had: already mentioned bookcorpus and Project Gutenberg validation, Paul Graham essays, concatenated 20_newsgroups.

I also tried to fine-tune the bigger mmBERT model (mmbert-base) but unfortunately it didn’t go well — metrics are weirdly lower in comparison with a small model.

Please give it a try. I'll appreciate a feedback.

The new multilingual model: https://huggingface.co/mirth/chonky_mmbert_small_multilingua...

All the Chonky models: https://huggingface.co/mirth

Chonky wrapper library: https://github.com/mirth/chonky

Comments

kamranjon•3mo ago
This is interesting! I once trained a t5 model by removing newlines from Wikipedia text and it worked surprisingly well / at the time the context length was the biggest issue.

Another, not so easy to solve issue was conversational dialogue type data, which wasn’t super well represented in the training data.

I’ve always wanted to come back to working on the problem again, because I think it’s very interesting and we will have a bunch of unstructured text as a result of STT models like whisper that do a great job of transcribing/translating but generally don’t format anything.

nvdnadj92•3mo ago
In case you need conversational data for the experiment you want to try, I developed an open-source cli tool [1] that create transcripts from voice chats on discord. Feel free to try it out!

[1] https://github.com/naveedn/audio-transcriber

CjHuber•3mo ago
Took me a minute to realize this is not about Chonkie. I would be interested in how this compares to the other's semantic chunking approach
jimmySixDOF•3mo ago
you can read the labels this (-y) uses modernBERT and even has an eval comparison to the (-ie) in it's GitHub so you can see the improvement as tested -- although if you want to do vanilla rules based chinking for whatever reason your data needs then (-ie) is still good.
TZubiri•3mo ago
That example looks terribly useless. Maybe there's an actually useful application you had in mind? I don't know say

Chonk("Hey I forgot my password, this is Tom from X Company") = ("Hey", "I forgot my password", "this is Tom from X Company")

Even then it doesn't quite look helpful.

freakynit•3mo ago
This is absolutely useless. Tried a few examples yesterday using hf demo. Fcking retarded af.

It literally splitted the text in-between of related texts while at the same time kept unrelated texts together, even though the embedding limit was far off.

I genuinely wanted this to work. I mean this. But nop. This shit did not work at all.

RAG is still fcked because if chunking issues. GraphRAG doesn't work correctly either unless you are willing to throw a lot of money during ingestion time.