Hugging Face's logo

carbon-internal
non-profit
Activity Feed

AI & ML interests

None defined yet.
loubnabnlย 
posted an update 5 days ago
view post
Post
4,213
๐ˆ๐ง๐ญ๐ซ๐จ๐๐ฎ๐œ๐ข๐ง๐  ๐Ÿงฌ ๐…๐ข๐ง๐ž๐ƒ๐๐€: ๐ญ๐ก๐ž ๐›๐ž๐ฌ๐ญ ๐๐š๐ญ๐š๐ฌ๐ž๐ญ ๐ญ๐จ ๐ญ๐ž๐š๐œ๐ก ๐š๐ง๐ฒ ๐ฅ๐š๐ง๐ ๐ฎ๐š๐ ๐ž ๐ฆ๐จ๐๐ž๐ฅ ๐ƒ๐๐€.
Training Qwen3-14B on it produces Carbon-X, the strongest open model on DNA benchmarks, on par with GPT-6.1 Sol.

It contains:
- Midtraining: 300B+ tokens of genomic sequences, papers and web pages mixing DNA with text
- SFT: 200K samples, from variant effect prediction to sequence question answering
- RL: 98K prompts with verifiable rewards

Data, models and the full recipe are open.
HuggingFaceBio/FineDNA

Enjoy this late Halloween treat! ๐ŸŽƒ
  • 5 replies
ยท
lewtunย 
submitted 2 papers to Daily Papers 9 months ago
cgeorgiawย 
posted an update 12 months ago
lvwerraย 
authored a paper about 1 year ago
cgeorgiawย 
posted an update about 1 year ago
view post
Post
6,209
๐Ÿš€๐Ÿš€๐Ÿš€ The largest ever dataset of co-folded 3D protein-ligand structures just dropped on HF!!

Meet SAIR (Structurally Augmented ICโ‚…โ‚€ Repository): 5M+ AI-generated complexes with experimentally measured drug potency data from SandboxAQ. ๐Ÿš€๐Ÿš€๐Ÿš€

Check it out and explore here: SandboxAQ/SAIR

  • 3 replies
ยท
cgeorgiawย 
posted an update about 1 year ago
lvwerraย 
authored a paper over 1 year ago
cgeorgiawย 
posted an update over 1 year ago
cgeorgiawย 
posted an update over 1 year ago
view post
Post
1,665
Snooping on HF is the best because sometimes you just discover that someone (in this case, Earth Species Project) is about to drop terabytes of sick (high quality animal sounds) data...

EarthSpeciesProject/NatureLM-audio-training
loubnabnlย 
authored a paper over 1 year ago
danaaubakirovaย 
posted an update over 1 year ago
danaaubakirovaย 
authored 2 papers over 1 year ago
cgeorgiawย 
posted an update over 1 year ago
view post
Post
597
Just dropped two bigger physics datasets (both on photonics)!

NUMBA 1: SIB-CL
This dataset of Surrogate- and Invariance-Boosted Contrastive Learning (SIB-CL) datasets for two scientific problems:
- PhC2D: 2D photonic crystal density-of-states (DOS) and bandstructure data.
- TISE: 3D time-independent Schrรถdinger equation eigenvalue and eigenvector solutions.

NUMBA2: 2D Photonic Topology
Symmetry-driven analysis of 2D photonic crystals: 10k random unit cells across 11 symmetries, 2 polarizations, 5 contrasts. Includes time-reversal breaking cases for 4 symmetries at high contrast.

Check them out: cgeorgiaw/sib-cl & cgeorgiaw/2d-photonic-topology
cgeorgiawย 
authored 4 papers over 1 year ago
loubnabnlย 
posted an update over 1 year ago