Post
117
Reality has not limits if you know how to shape it, one step at a time.
- 1 reply
🤏 quecto mode
appvoid
appvoid
AI & ML interests
singularity through byte-level tokens and small language models - creating applications and ideas out of the void
Recent Activity
new activity about 11 hours ago
AxiomicLabs/Open_SLM_Leaderboard:Only PR's will be added for the next little while updated a dataset about 14 hours ago
appvoid/void-byte-data liked a model about 22 hours ago
bench-labs/tinctura-v1Organizations
posted an update 1 day ago
reacted to TobiasLogic's post with 🚀 2 days ago
Post
3473
We’ve been cooking something new at Bench Labs.
Introducing Cagliostro-v3, our new 146M parameter language model trained completely from scratch.
The run isn’t even finished yet.
At the current checkpoint:
• 146M parameters
• 72.7B / 75B tokens trained
• 26.27 Open SLM Index
• 43.80 ArithMark-3
• Trained on a single RTX 5090
• ~90K to 103K tokens/sec during training
• ~9 days for the full run
• Apache 2.0
For some context, SmolLM2-135M scores 27.13 on the same Index after being trained on roughly 2 trillion tokens.
Cagliostro-v3 is currently at 26.27 with only ~72.7B.
That’s around 27x fewer training tokens.
The model also currently Hold the number 3rd spot for ArithMark-3, scoring 43.80
This wasn’t achieved by just throwing more tokens at the model. A huge part of v3 has been figuring out architecture, data mixture, and training dynamics at this scale.
The model uses a custom 30-layer decoder architecture with grouped-query attention and cross-head subspace attenuation, SwiGLU, RMSNorm, RoPE, tied embeddings, and a warmup-stable-decay training schedule.
During cooldown we also substantially shifted the data mixture toward higher-quality synthetic textbook and mathematics data, with the mathematics share increasing from 10% to 28%.
And everything is open.
The repository contains the training history with checkpoints pushed roughly every 30 minutes, so you can inspect how the model evolved throughout training rather than only seeing the final weights.
This is still a pre-final checkpoint. We have roughly 2.3B tokens left and the learning-rate cooldown is still running.
So 26.27 isn’t the final number.
Really excited to see where the last part of the run lands.
Cagliostro-v3:
bench-labs/cagliostro-v3
Built by Bench Labs.
Open SLM Leaderboard:
AxiomicLabs/Open_SLM_Leaderboard
Introducing Cagliostro-v3, our new 146M parameter language model trained completely from scratch.
The run isn’t even finished yet.
At the current checkpoint:
• 146M parameters
• 72.7B / 75B tokens trained
• 26.27 Open SLM Index
• 43.80 ArithMark-3
• Trained on a single RTX 5090
• ~90K to 103K tokens/sec during training
• ~9 days for the full run
• Apache 2.0
For some context, SmolLM2-135M scores 27.13 on the same Index after being trained on roughly 2 trillion tokens.
Cagliostro-v3 is currently at 26.27 with only ~72.7B.
That’s around 27x fewer training tokens.
The model also currently Hold the number 3rd spot for ArithMark-3, scoring 43.80
This wasn’t achieved by just throwing more tokens at the model. A huge part of v3 has been figuring out architecture, data mixture, and training dynamics at this scale.
The model uses a custom 30-layer decoder architecture with grouped-query attention and cross-head subspace attenuation, SwiGLU, RMSNorm, RoPE, tied embeddings, and a warmup-stable-decay training schedule.
During cooldown we also substantially shifted the data mixture toward higher-quality synthetic textbook and mathematics data, with the mathematics share increasing from 10% to 28%.
And everything is open.
The repository contains the training history with checkpoints pushed roughly every 30 minutes, so you can inspect how the model evolved throughout training rather than only seeing the final weights.
This is still a pre-final checkpoint. We have roughly 2.3B tokens left and the learning-rate cooldown is still running.
So 26.27 isn’t the final number.
Really excited to see where the last part of the run lands.
Cagliostro-v3:
bench-labs/cagliostro-v3
Built by Bench Labs.
Open SLM Leaderboard:
AxiomicLabs/Open_SLM_Leaderboard
- 8 replies
posted an update 3 days ago
Post
62
How old were you when you discovered good AI agents favor the Strategy Design Pattern over any other?
- 1 reply
posted an update 14 days ago
Post
126
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.
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.
congrats!
posted an update 17 days ago
Post
137
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
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
reacted to KlondikeDev's post with ❤️ 19 days ago
Post
2152
The SLM Consortium has begun work on a safety dataset for Small Language Models, with the creation of the dataset being headed by @wayneworkman2012
The dataset will focus on refusals and redirects surrounding dangerous or extreme sexual content, designed to be shaped sized appropriately for SLMs, without significantly lowering benchmark performance.
More info will be out soon!
slmconsortium
The dataset will focus on refusals and redirects surrounding dangerous or extreme sexual content, designed to be shaped sized appropriately for SLMs, without significantly lowering benchmark performance.
More info will be out soon!
- 3 replies
replied to their post 21 days ago
Well I don’t care if it’s AGI or not I just find it super helpful in real practice
It would be for me if my credits wouldn't drain even light mode 🥲
replied to their post 21 days ago
it is not AGI
For me it is not, but it's pretty close for me though
replied to their post 22 days ago
Ill wait for the glm distilled xD
I think we are years away though
replied to their post 22 days ago
Wild if true
replied to their post 24 days ago
Feeling the same :)
posted an update 24 days ago
Post
126
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.
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
replied to Banaxi-Tech's post 28 days ago
Imagine having some kind of server so it always feels updated with your latest integrated models!
reacted to Banaxi-Tech's post with 🚀 28 days ago
Post
3452
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)
- 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)
- 7 replies
replied to their post 28 days ago
byte-level tokenizer is the part that caught my attention, keep it up sir!
replied to their post 28 days ago
Let's go!!!
replied to their post 28 days ago
Yes sir!
replied to their post 28 days ago
Do you mean like the architecture? Could you point at the model?