Hugging Face's logo
Activity Feed

AI & ML interests

None defined yet.

Recent Activity
saisatwik1234567ย  new activity 15 days ago
blog-explorers/README:[Join request] Verification and provenance for AI agents
kgrabkoย  new activity 21 days ago
blog-explorers/README:[Support] Community Articles
saisatwik1234567ย  new activity 22 days ago
blog-explorers/README:Join request pending for a week, anything else needed?
View all activity
apsysย 
posted an update 11 days ago
anakin87ย 
posted an update 13 days ago
view post
Post
3442
I made a 1.1M ModernBERT encoder play Doom in real time on a CPU

Some time ago, VAGO Solutions released SauerkrautLM-Doom-MultiVec-1.3M, a tiny model trained to play Doom Defend the Center scenario from 31k human gameplay examples.

My first thought: cool! I love both Doom and Small Language Models.

Then another idea: I bet I can do better :-)

What I did?
- evaluated the original model and found it's better than reported
- changed a bit the architecture
- generated SFT data with a scripted oracle
- SFT + PPO refinement on consumer hardware

Got a smaller, faster and killer model
Can even fit a floppy with int8 quantization ๐Ÿ’พ

Watch it play/read the article: anakin87/tiny-doom-defender
danielhanchenย 
posted an update 13 days ago
saisatwik1234567ย 
in blog-explorers/README 15 days ago

[Join request] Verification and provenance for AI agents

2
#29 opened 15 days ago by
AiOpsEnabler
aryopgย 
authored a paper 20 days ago
kgrabkoย 
in blog-explorers/README 21 days ago

[Support] Community Articles

๐Ÿš€๐Ÿค 2
113
#5 opened over 2 years ago by
victor
saisatwik1234567ย 
in blog-explorers/README 22 days ago

Join request pending for a week, anything else needed?

1
#28 opened about 2 months ago by
yuncheol-ncdtech

[Support] Community Articles

๐Ÿ”ฅ๐Ÿš€ 2
113
#5 opened over 2 years ago by
victor
jasoncorkillย 
posted an update 25 days ago
view post
Post
674
Researchers shouldnโ€™t have to limit experiments because human feedback is too expensive/slow to collect.

So, @RapidataAI is offering up to $50,000 in human annotation credits per research project (can be more for selected topics).

Learn more on our website, and apply via email, we'll do our best to get back to you within 2 days :).
jasoncorkillย 
posted an update 27 days ago
view post
Post
2592
Most public benchmarks collapse model performance into one broad preference signal.

That makes it hard to understand which capabilities differentiate between models. It's also almost impossible to inspect the evidence behind it. So @RapidataAI is releasing Benchmark.AI.

We started with an SVG generation benchmark including 42 models, 500 prompts, 1.9M+ human judgements, 300K+ match-ups.

We evaluate models separately on Preference, Alignment and Coherence, while making the prompts, outputs, match-ups and methodology public.

Full dataset: Rapidata/svg-benchmark
Full benchmark: https://www.benchmark.ai/svg

Methodology feedback and benchmark suggestions very welcome!


  • 13 replies
ยท
tomaarsenย 
posted an update about 1 month ago
view post
Post
3972
๐Ÿšจ I've just published Sentence Transformers v6.0, introducing MultiVectorEncoder: ColBERT-style late interaction models are now a fourth model type, for training, inference, and interpretation, alongside the dense, sparse, and reranker models! Details:

Where a regular embedding model compresses a whole text into one vector, a multi-vector model keeps one vector per token and scores query against document with the MaxSim operator. That preserves token-level matching information that a single vector has to average away. It is also the state of the art for visual document retrieval, where a text query is matched against page images directly, charts and tables included, with no OCR step in between.

Any PyLate, Stanford ColBERT, or ColPali checkpoint loads straight into the same familiar API: model.encode_query(), model.encode_document(), and model.similarity() just work, whether the documents are texts or page images.

Does it help? LightOn trained LateOn (multi-vector) and DenseOn (dense) on the same data with the same 149M ModernBERT backbone, and the multi-vector model wins on 9 of the 13 NanoBEIR datasets: 0.6868 vs 0.6764 mean NDCG@10. The price is a bigger index, and the new HierarchicalTokenPooling module halves it at roughly no retrieval cost.

Antoine Chaffin, Raphaรซl Sourty, and I wrote a blog post walking through multi-vector models in practice: loading the various checkpoint formats, encoding and scoring, plugging them into a search stack, running them on page images, and keeping the index affordable. Check it out if you want to get started, or just point your Agent to the URL: https://huggingface.co/blog/multi-vector-encoder

pip install sentence-transformers==6.0.0

Release notes: https://github.com/huggingface/sentence-transformers/releases/tag/v6.0.0
danielhanchenย 
posted an update about 1 month ago
view post
Post
5862
Introducing Unsloth Desktop ๐Ÿฆฅ
The first desktop app to run and train models locally.

โ€ข Open-source. Runs on Mac, Windows and Linux
โ€ข Supports MLX, diffusion image/video, audio, GGUF
โ€ข Connect Claude Code and Codex to local LLMs
โ€ข 50% more accurate, self-healing tool calls + sandboxed code exec
โ€ข Works for CPU + multiGPU setups - NVIDIA, AMD, Intel, Mac
โ€ข Train models 2ร— faster with 70% less VRAM
โ€ข Private web search, deep research, RAG, MCP and exports (NVFP4, GGUF)
โ€ข Use Unslothโ€™s OpenAI-compatible API and cloud models
โ€ข Securely deploy LLMs remotely and access anywhere

Unsloth Desktop is now available on http://unsloth.ai
and GitHub.

GitHub: https://github.com/unslothai/unsloth
Blog and Guide: https://unsloth.ai/docs/desktop
  • 6 replies
ยท
danielhanchenย 
posted an update about 2 months ago
oopereย 
authored a paper about 2 months ago
oopereย 
submitted a paper to Daily Papers about 2 months ago
danielhanchenย 
posted an update about 2 months ago
view post
Post
2367
We compared 1-bit Kimi K3 to Claude Opus 5 and GPT 5.6. ๐Ÿคฏ

We gave 4 models the same prompt: Create a glass aquarium whose side panel develops a visible crack and then bursts...

1-bit Kimi K3 GGUF ran locally on 4x B200s at 36 tok/s.

GGUF: unsloth/Kimi-K3-GGUF
GitHub repo: https://github.com/unslothai/unsloth
  • 1 reply
ยท
danielhanchenย 
posted an update about 2 months ago
view post
Post
4725
Kimi K3 can now be run locally! โœจ

The 1-bit model retains ~78.9% accuracy after we shrunk it from 1.56TB to 594GB (-62% size).

Run on a Mac Studio connected with 128GB RAM device. Kimi K3 is the strongest open model to date.

GGUF: unsloth/Kimi-K3-GGUF
Guide: https://unsloth.ai/docs/models/kimi-k3
  • 5 replies
ยท
christopherย 
in blog-explorers/README about 2 months ago

[Support] Community Articles

๐Ÿ”ฅ๐Ÿค 2
113
#5 opened over 2 years ago by
victor
juiceb0xc0deย 
in blog-explorers/README about 2 months ago

[Support] Community Articles

๐Ÿค๐Ÿ”ฅ 2
113
#5 opened over 2 years ago by
victor
jasoncorkillย 
posted an update 2 months ago
view post
Post
100
Are you an AI Researcher, Enthusiast, Founder or else and live in or close by to Zรผrich? Drop in for an Aperol Spritz and Euromaxx on our Balcony on the 14th of August.
  • 1 reply
ยท