Hugging Face's logo
Activity Feed

AI & ML interests

None defined yet.

Recent Activity
caizhongang  authored a paper 10 days ago
SenseNova-U1.5: Towards Native Unified Visual Intelligence
yuntian-deng  authored a paper 16 days ago
Compile by Training: Turning Natural-Language Specifications into Local Neural Functions
yuntian-deng  authored a paper 16 days ago
Interactive Training 2: Auditable Control Plane for Live Model Training
View all activity
mrfakename 
posted an update 6 days ago
view post
Post
307
We’ve been working with LAION on a voice acting arena. You listen to two models doing the same scene and compare how well they pull it off.

It’s ready to try now - would love to hear what you think 🙂

TTS-AGI/voice-acting-arena
appvoid 
posted an update 7 days ago
view post
Post
104
LLMS for edge devices?
What if we go the reliable route instead of the speed route?
What if we make it run on sbcs with few megabytes available?

That's the idea for the next model.

Keep in tune.
appvoid 
posted an update 11 days ago
view post
Post
124
We trained a 10.9M byte-level recurrent Transformer on L3 and L6. (Loop 3 and Loop 6)

Yet L4/L5 improved too, L8 held up, and the L3→L6 gain grew during training.

Same weights. More compute. Better predictions.

This is a new architecture for effective compute after several steps beyond original training!

We mixed and matched components like time and mhc into an ouro-like byte-level language model and the result is BET, a byte-level step-elastic transformer that can run computation steps without significant degradation.

One of the coolest parts of this training was discovering how Gradient Descent decided to use the first layer as what we would consider a scratchpad! Totally destroyed for the decoder but somehow makes total sense for the next layer!

I believe looped-transformers are the future of edge computing and this is a first step towards it.

Blogpost: https://medium.com/@appvoidofficial/byte-level-elasticity-182fe2ed1d2f

appvoid/bet-10m
appvoid 
posted an update 16 days ago
view post
Post
3866
We got gpt6 before gta6
  • 15 replies
·
yuntian-deng 
authored 2 papers 16 days ago
yuntian-deng 
submitted a paper to Daily Papers 17 days ago
appvoid 
posted an update 18 days ago
view post
Post
125
Any thoughts on Nvidia acquiring this website?

I don't know what to feel about it. But would be great if huggingface gets something similar to Kaggle with free GPU hours (or even days) for training.
  • 3 replies
·
marcusinthesky 
authored 2 papers 18 days ago
marcusinthesky 
submitted a paper to Daily Papers 18 days ago
codelion 
in zero-gpu-explorers/README 18 days ago

ZeroGPU clone fails during tensor packing with Errno 28 — 82.1 GB pack target

1
#185 opened 18 days ago by
Thananon
marcusinthesky 
submitted a paper to Daily Papers 18 days ago
yuntian-deng 
authored a paper 19 days ago
yuntian-deng 
submitted a paper to Daily Papers 19 days ago
appvoid 
posted an update 23 days ago
view post
Post
3303
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.
  • 36 replies
·
appvoid 
posted an update 25 days ago
view post
Post
2646
Nobody knows what is doing, when you train a model, you are experimenting to advance the frontier, so keep failing 🫵
  • 17 replies
·
appvoid 
posted an update 27 days ago
view post
Post
2601
I hope that after this OpenAI disaster on Plus users, more people start realizing why Open Weights were always the only way.
  • 5 replies
·
appvoid 
posted an update 30 days ago
view post
Post
1003
Byte-level state-space models. That sounded pretty scary for a scientist decades ago. Now we have:

1. Knowledge that deeper layers train smoothly.
2. Knowledge that Transformers work but is quadratic on sequence length.
3. Knowledge that SSMs work even better. Numerically unstable sometimes.
4. Speculative-decoding.
5. Open high-quality data.
6. Knowledge that KD works.

It slowly feels like is no longer a bad idea.
  • 3 replies
·
BestWishYsh 
authored a paper about 1 month ago