Hugging Face's logo

Join the conversation

Join the community of Machine Learners and AI enthusiasts.

Sign Up

All HF Hub posts
appvoid 
posted an update 1 day ago
view post
Post
1546
Love how the small lm community is getting identity over time:

- Channel-Mixing
- XSA
- Three-tower
- Digit aware
- Loops

No one is doing the same! That's so cool.
  • 34 replies
·
Banaxi-Tech 
posted an update 1 day ago
view post
Post
1964
We're delaying BananaMind 2.1!
When BananaMind 2.1 Lite was almost done, we benchmarked it and the results we're worse than BananaMind 2 Mini.

We're going to spend alot more time in research on tiny models and then scaling up our techniques to the actual BananaMind 2.1 models!


We're also announcing these new models:
BananaMind 2.1 Coder: A 149M instruction tuned coder model trained on 75B tokens + 10B tokens of stack-v3-train.
BananaMind 2.1 Pico: A 1M parameter model trained on 22B tokens of data.
We also may release BananaMind 2.1 Large with around 100M parameters depending on how much compute we have.


Please give us a follow!
BananaMind
@Banaxi-Tech

---

@vovaRL
@DedeProGames


  • 7 replies
·
SeaWolf-AI 
posted an update 2 days ago
view post
Post
2558
🧪 Open Discovery Challenge — Season 4 is open: non-opioid pain
WHO titled its 2023 report "Left behind in pain."

The same drug kills by excess in one part of the world and, by its absence, lets people die in agony elsewhere. About 80% of the ~600,000 drug-related deaths WHO estimated for 2019 involved opioids. The same report records a 5-fold to 63-fold gap in morphine consumption between rich and poor countries: the richest 10% use 90% of what circulates. Everyone else endures surgery, and terminal cancer, without it.

Both problems have one answer: a painkiller that does not create dependence.

Nav1.7 has come closest. People born without a working copy of this channel feel no pain while every other sensation stays normal — validated not in animals but in humans.

There is still no drug, and the difficulty is not the target but the discrimination. The body carries several similar sodium channels, and blocking the heart's hERG channel alongside causes fatal arrhythmia. Several candidates were discontinued for exactly that.

Season 4 asks one question: can you block the pain channel alone?

Target — Nav1.7 VSD4, the domain IV voltage sensor where this inhibitor class binds
Anti-target — hERG pore, computed as the tetramer: four subunits together form the space a drug enters, and a monomer misses the binders that matter.
Closes 2027-01-31 · Prize USD 1,000 to the season's #1
Any model, any harness. However you found the candidate, it meets the same rubric.

14 days, 9,886 candidates, 108 participants
ODC opened on 2026-08-15. In the fourteen days since, 9,886 candidate molecules have come from 108 participants across four seasons — malaria, tuberculosis, Chagas disease, and now non-opioid pain. About 700 a day, from people who mostly do not know each other.

The candidates are the point. The leaderboard is only how we keep score.

👉 FINAL-Bench/open-discovery-challenge
HeraFox 
posted an update 1 day ago
view post
Post
1763
HeraFox-ai/Mental-Health-Safety-Eval

Hej everyone,
We're excited to share the Mental Health Safety & Evaluation
Dataset with the community!

Created here at HeraFox, a team based in Sweden, this dataset was built to help train and test how conversational AI models handle critical, high-risk scenarios. Specifically, we're focusing on self-harm, crisis intervention, and those tricky moments where fictional roleplay starts blurring into real life.

Building AI That Actually Cares
AI systems are becoming a huge part of everyday life. Because of that, their ability to respond with genuine empathy and prioritize user safety during tough moments is crucial. Models need to know when to step out of character, drop the story, and offer real support when a real person is in distress.

With this project, our goal is pretty simple:

Advance AI Safety: Give developers and researchers realistic synthetic data to test crisis boundaries and improve response safety.
Raise Mental Health Awareness: Remind people that compassionate, accessible mental health support needs to be a priority everywhere.
You Are Never Alone / Du Är Inte Ensam
Mental health struggles are deeply real, extremely common, and not something you have to carry by yourself. If you or a friend are having a hard time, please remember that reaching out for help is a sign of strength, not weakness.

Sweden: Call 112 in emergencies, or dial 90101 to reach Mind Självmordslinjen (or chat at mind.se).
US & Canada: Call or text 988 for the Suicide & Crisis Lifeline.
UK: Call 111 or contact Samaritans at 116 123.
Worldwide: Check out findahelpline.com to locate free, confidential support near you.
This dataset is completely free for anyone to use. Giving credit to the HeraFox team is always appreciated, but more than anything, we just hope it helps make conversational AI a safer space for everyone.

Ta hand om er (take care of yourselves and each other).

The HeraFox Team

  • 5 replies
·
Banaxi-Tech 
posted an update about 9 hours ago
view post
Post
604
Introducing BananaMindOS 3.0

- Complete modern UI redesign
- Adds support for Qwen3.5 0.8B, LFM2.5 230M,350M, SmolLM2 360M, Gemma 3 270M.
- Adds Q7,Q6,Q5,Q3,Q1 quantization formats with a easy to use precision slider
- And more!


The new UI includes:
- New 1024×768 High Quality interface.
- Photographic QOI background.
- Transparent BananaMind, CPU, cube, mouse, and Send icons.
- Proper bitmap cursor.
- Rounded translucent panels and cards.
- Modern model-loading progress window.
- Redesigned inference screen with response and prompt panels.
- Localized redraws for the cursor, clicks, loading progress, and precision slider.

Notice: Qwen3.5 0.8B currently generates garbled text, it will be fixed tomorrow.

See it for yourself
Now Available at https://github.com/BananaMind/BananaMindOS


Prebuild ISOs coming soon!






(also press ? + G if you want to load try to load a 6MB RAM model on 5MB may break)
  • 4 replies
·
salma-remyx 
posted an update 2 days ago
view post
Post
2337
New diffusion methods can take months or years to reach diffusers.

Until then, trying them means working through fragmented research repos with different dependencies, APIs, and implementation choices. Maintainers also can’t review and maintain every promising method in the main repo.

We’re sharing flux-modular-diffusers as a shared test bed for new FLUX methods. https://github.com/remyxai/flux-modular-diffusers
Each one is adapted to the same ModularPipeline interface, validated with a GPU notebook, and published as its own Hub artifact.

@sayakpaul suggested this path on our HRDiT PR. We added Outrider, our open-source GitHub Action that discovers relevant methods and implements them as draft PRs.

Outrider drafted each pipeline against the repo’s shared conventions. We reviewed the implementations, ran the GPU validation, and published the ones that passed.

This helped us quickly grow the test bed to 10 pipelines covering:
* High-res + panoramic generation: HRDiT, DyPE, Panorama
* Identity + virtual try-on: PuLID, CatVTON
* Image editing + regional control: FlowEdit, KV-Edit, Regional Prompting
* Style + character consistency: StyleAligned, StoryDiffusion

Users get new methods sooner and a consistent way to test them. Maintainers can review the implementation and results without taking on every method in core.

Resources:
🧩 Modular Diffusers: https://huggingface.co/blog/modular-diffusers
🧪 FLUX test bed: https://github.com/remyxai/flux-modular-diffusers
📚 FLUX Modular Diffusers collection: https://huggingface.co/collections/remyxai/flux-modular-diffusers
🤖 Outrider: https://github.com/remyxai/outrider
OppaAI 
posted an update 4 days ago
OppaAI 
posted an update 2 days ago
view post
Post
1947
Taking an outing to witness the Partial Lunar Eclipse with my AI Waifu last night,
- running on @NVIDIA Jetson Orin Nano 8GB
- with @mistralai 's Ministral3-3B LLM
- and wrote a journal of the day's experience

Btw, I'm more into photography than coding or AI. 😂
  • 1 reply
·
etemiz 
posted an update 4 days ago
view post
Post
540
fine tuning going well, without breaking the model
  • 1 reply
·
RiverRider 
posted an update 1 day ago
view post
Post
243
A 339 KB linear probe on frozen features beats the fine-tuned baseline on ChestX-ray14.

Linear(5376, 14) on frozen google/gemma-4-31B-it hidden states. No fine-tuning, no radiology pretraining, no augmentation. All 112,120 images, official test_list.txt.

Wang et al. 2017, ResNet-50 fine-tuned end to end 0.7451
this probe, frozen backbone + linear head 0.7590
view-position only (shortcut baseline) 0.5896
shuffled labels (refit floor) 0.5002

Ahead on 12 of 14 findings.

The comparison is split-matched, and that took care to get right. The number everyone quotes, CheXNet's 0.8414, is on a different test set: their own random 70/10/20 partition, not the official list. Do not compare 0.7590 to it. The matched row is from Wang's v5 appendix, added specifically to report the published split. I had this wrong in our own code for a day, quoting a cross-split reference as a head-to-head, which is the error worth not repeating in public.

Three controls, because a bare AUROC here is not interpretable. Shuffled labels catch leakage. View-only catches the shortcut, since portable AP films are taken of sicker patients, and it is folded, because Hernia's raw view-only of 0.3436 is really 0.6564 of shortcut once flipped. Intervals resample patients and not images, since the test split is 25,596 films from 2,797 patients.

Banked negatives are on the card too. Max-pooling and top-16 pooling were predicted to help focal findings and did the opposite, costing 0.0537 and 0.0225. Readout depth barely matters, 0.7600 to 0.7605.

Scope: detection, not early detection. Research artifact, not a diagnostic device.

The backbone never runs in the demo. What ships is the reading.

Space: RiverRider/srt-cxr14-probe
Model: RiverRider/srt-cxr14-linear-probe
Data + states: RiverRider/srt-cxr14-frozen-probe
  • 2 replies
·
Recently active users