Post
3,724
Tommorow we are releasing Supra2-Pro(100M params), our flagship, get ready!
(follow SupraLabs to get tuned in!)
SupraLabs
(follow SupraLabs to get tuned in!)
- 9 replies
🏗️ Building on HF
AxionLab
AxionLab-official
AI & ML interests
Owner of SupraLabs
Recent Activity
reacted to theirpost with 🚀 about 3 hours ago
Tommorow we are releasing Supra2-Pro(100M params), our flagship, get ready!
(follow SupraLabs to get tuned in!)
https://huggingface.co/SupraLabs liked a model about 22 hours ago
AuroraAI-Research/Aurora1.0-150M liked a model 1 day ago
tiiuae/Falcon-H1-Tiny-90M-Instruct-GGUFOrganizations
reacted to their post with 🚀 about 3 hours ago
reacted to DedeProGames's post with 🧠🚀🔥 5 days ago
Post
3,432
Meet Supra2.5-Ultra-Preview-001. 230M parameters, ~40.4B training tokens.
SupraLabs has released the first preview of its Supra2.5-Ultra family: a compact, Llama-style base model for text completion and small-model experimentation.
What's inside?
• 230M parameters
• 32 transformer layers
• 2,048-token context window
• Grouped Query Attention: 12 attention heads, 4 KV heads
• 49,152-token vocabulary
• SwiGLU activation, RoPE and tied embeddings
• BF16 weights and Hugging Face Transformers support
• Apache 2.0 license
Training
The model was trained on approximately 40.4B tokens from FineWeb-Edu, DCLM-Baseline and FineMath using a single A100 SXM.
Reported zero-shot results
• HellaSwag: 37.18% normalized accuracy
• PIQA: 65.78% accuracy
• ARC-Easy: 57.11% accuracy
• ARC-Challenge: 26.28% normalized accuracy
Want to experiment?
Code on REPO.
Model, weights and usage example:
SupraLabs/Supra2.5-Ultra-Preview-001
SupraLabs has released the first preview of its Supra2.5-Ultra family: a compact, Llama-style base model for text completion and small-model experimentation.
What's inside?
• 230M parameters
• 32 transformer layers
• 2,048-token context window
• Grouped Query Attention: 12 attention heads, 4 KV heads
• 49,152-token vocabulary
• SwiGLU activation, RoPE and tied embeddings
• BF16 weights and Hugging Face Transformers support
• Apache 2.0 license
Training
The model was trained on approximately 40.4B tokens from FineWeb-Edu, DCLM-Baseline and FineMath using a single A100 SXM.
Reported zero-shot results
• HellaSwag: 37.18% normalized accuracy
• PIQA: 65.78% accuracy
• ARC-Easy: 57.11% accuracy
• ARC-Challenge: 26.28% normalized accuracy
Want to experiment?
Code on REPO.
Model, weights and usage example:
SupraLabs/Supra2.5-Ultra-Preview-001
- 3 replies
reacted to mithulaartigala's post with 🔥 5 days ago
replied to DedeProGames's post 5 days ago
Wishing you a strong fresh start with SupraLabs. The focused direction sounds promising—looking forward to seeing what you build next.
You are so cool bro ngl
replied to Banaxi-Tech's post 6 days ago
@LH-Tech-AI @AxionLab-official which model are you talking about?
That BGA shitty thing
replied to Banaxi-Tech's post 6 days ago
If you are going to make the same slop shit you already did, you should stop. Please stop using AI to make AI slop.
reacted to Banaxi-Tech's post with 😔 7 days ago
Post
4,523
Hi everyone, we're back!
And we're releasing something major again now, if you have seen Grok Bot, its very powerful, but there isn't really a local alternative yet.
Today we're releasing Minds, they're personal AI bots that live on your computer.
They can run anywhere, on your phone, on your mac, on windows, and even on linux.
They support OpenAI Oauth as the main integration because most people already have a ChatGPT sub.
But locally, It supports ANY openai compatible api which allows you to setup ollama and llama.cpp and everything else.
It uses Jev for command approval on OpenRouter and Muse Spark Transcribe for transcription.
Its fully Open Source!
Check it out: https://github.com/BananaMind/Minds
And we're releasing something major again now, if you have seen Grok Bot, its very powerful, but there isn't really a local alternative yet.
Today we're releasing Minds, they're personal AI bots that live on your computer.
They can run anywhere, on your phone, on your mac, on windows, and even on linux.
They support OpenAI Oauth as the main integration because most people already have a ChatGPT sub.
But locally, It supports ANY openai compatible api which allows you to setup ollama and llama.cpp and everything else.
It uses Jev for command approval on OpenRouter and Muse Spark Transcribe for transcription.
Its fully Open Source!
Check it out: https://github.com/BananaMind/Minds
- 9 replies
reacted to DedeProGames's post with 🔥🚀🤗❤️ 7 days ago
Post
5,129
I’ve just left 
BananaMind to focus my work on SupraLabs .
Following some internal pressure from SupraLabs and issues @Banaxi-Tech encountered with certain architectures, I decided it was time to change direction.
I want to thank BananaMind for inviting me to join their organization and for helping shape my knowledge and passion regarding SLMs.
Now, it’s just me and SupraLabs .
Following some internal pressure from SupraLabs and issues @Banaxi-Tech encountered with certain architectures, I decided it was time to change direction.
I want to thank
Now, it’s just me and
- 8 replies
replied to Banaxi-Tech's post 8 days ago
Bananamind3 is getting delayed lmao
reacted to Banaxi-Tech's post with 😔 8 days ago
Post
3,226
We have released BGA!
And wow, It provides 256x (and 512x at the end of 1M) yes 256x LESS attention compute at 1M context window.
That means you can train a 1M context window at the compute of a ~4K context window.
Check IT OUT: BananaMind/blog
The Accuracy Should BE WAy better than DSA but untested yet.
And, now some updates on BananaMind 3:
BananaMind 3 Will start training Soon!
Sizes: 10M, 25M, 50M, 100M, 150M
And the context windows ARE INSANE: 10M, 16K context, 25M 16k context, 50M 32K context, 100M and 150M, 64K context!!!!
And wow, It provides 256x (and 512x at the end of 1M) yes 256x LESS attention compute at 1M context window.
That means you can train a 1M context window at the compute of a ~4K context window.
Check IT OUT: BananaMind/blog
The Accuracy Should BE WAy better than DSA but untested yet.
And, now some updates on BananaMind 3:
BananaMind 3 Will start training Soon!
Sizes: 10M, 25M, 50M, 100M, 150M
And the context windows ARE INSANE: 10M, 16K context, 25M 16k context, 50M 32K context, 100M and 150M, 64K context!!!!
- 13 replies
reacted to mithulaartigala's post with 🤗🔥 9 days ago
Post
4,603
I am LOVING HuggingFace Pro <3
- 7 replies
L banaxi 😢
reacted to AtAndDev's post with 😔 9 days ago
Post
5,214
@Banaxi-Tech stop hiding my comments. AND STOP STEALING PAPERS AND SPREADING MISINFORMATION.
your BGA blog is a copy of NSA (deepseek, 2025) branded under your name. literally the same top16 selected blocks, 512 local window, router over block summaries, all you did was change block size from 64 to 128.
you didnt cite NSA once but you put a “please cite BGA” bibtex at the bottom.
i commented under your post and said that there is no way that you can support claims like: “The Accuracy Should BE WAy better than DSA but untested yet.” you didnt run a single experiment. and the 256x isnt from BGA, its just n/2k with k=2048 so the exact same k DSA uses. if opus wrote this for you, at least read it before posting.
i commented again after you hid my comment despite it having constructive and correct feedback and you hid that too. and again.
you can hide the truth and just try to get hf post likes..... but is it really the thing that needs to be done? do you really want to take papers and make them yours while barely even changing the params?
admitting your mistakes and doing something about them needs humbleness, intelligence, humanness.
i encourage you to admit your mistakes and try to do better next time (at least read what blog your ai wrote or do proper experiments to back your stuff up).
your BGA blog is a copy of NSA (deepseek, 2025) branded under your name. literally the same top16 selected blocks, 512 local window, router over block summaries, all you did was change block size from 64 to 128.
you didnt cite NSA once but you put a “please cite BGA” bibtex at the bottom.
i commented under your post and said that there is no way that you can support claims like: “The Accuracy Should BE WAy better than DSA but untested yet.” you didnt run a single experiment. and the 256x isnt from BGA, its just n/2k with k=2048 so the exact same k DSA uses. if opus wrote this for you, at least read it before posting.
i commented again after you hid my comment despite it having constructive and correct feedback and you hid that too. and again.
you can hide the truth and just try to get hf post likes..... but is it really the thing that needs to be done? do you really want to take papers and make them yours while barely even changing the params?
admitting your mistakes and doing something about them needs humbleness, intelligence, humanness.
i encourage you to admit your mistakes and try to do better next time (at least read what blog your ai wrote or do proper experiments to back your stuff up).
- 49 replies
reacted to DedeProGames's post with 🧠 21 days ago
Post
2,892
🧱 SLM Tetris Arena: can a small language model play Tetris without ever being trained on it?
I built an arena where tiny decoder-only LMs (50K–250M params) play Tetris zero-shot. There is no fine-tuning and no game data. They only use what they picked up from pre-training on text.
How it works:
- For every piece, the engine simulates each legal placement and describes the result in plain English ("clears one line, creates no new holes, keeps the stack low…").
- The model never sees the grid. It reads each description, and the arena compares log P(" good move") with log P(" bad move"). The best-rated placement is played.
- Every player gets the same piece sequence, so it's a fair race.
- There are two protocols: Guided (the rules are in the prompt) and Blind (no rules, only pre-training knowledge).
Two ways to play:
- Match: pick any models (even your own, custom architectures welcome) and watch them play side by side on retro 8-bit boards.
- Ranked: press Play and the arena picks up to 4 models at random from a curated pool of 29. Nobody chooses their opponents, so Elo can't be farmed. Matches run on the server and count even if you close the tab.
First results (~225 ranked matches):
- gpt2 (124M) leads with 1283 Elo, but SupraNeo-4M (4M) is right behind at 1239. Next come LowOnMind-5M and BananaMind-2.1-Pico (1.5M!).
- Model size barely predicts Elo (r ≈ 0.06). Survival does (r ≈ 0.9): the models that avoid holes and keep the stack low are the ones that win.
Every ranked match (seed, model commit SHAs, scores, Elo before/after) is logged in a public dataset.
▶ Play: DedeProGames/SLM-Tetris-Arena
📊 Results: DedeProGames/lm-tetris-arena-results
Want your model in the Ranked pool? Drop it in the comments!
I built an arena where tiny decoder-only LMs (50K–250M params) play Tetris zero-shot. There is no fine-tuning and no game data. They only use what they picked up from pre-training on text.
How it works:
- For every piece, the engine simulates each legal placement and describes the result in plain English ("clears one line, creates no new holes, keeps the stack low…").
- The model never sees the grid. It reads each description, and the arena compares log P(" good move") with log P(" bad move"). The best-rated placement is played.
- Every player gets the same piece sequence, so it's a fair race.
- There are two protocols: Guided (the rules are in the prompt) and Blind (no rules, only pre-training knowledge).
Two ways to play:
- Match: pick any models (even your own, custom architectures welcome) and watch them play side by side on retro 8-bit boards.
- Ranked: press Play and the arena picks up to 4 models at random from a curated pool of 29. Nobody chooses their opponents, so Elo can't be farmed. Matches run on the server and count even if you close the tab.
First results (~225 ranked matches):
- gpt2 (124M) leads with 1283 Elo, but SupraNeo-4M (4M) is right behind at 1239. Next come LowOnMind-5M and BananaMind-2.1-Pico (1.5M!).
- Model size barely predicts Elo (r ≈ 0.06). Survival does (r ≈ 0.9): the models that avoid holes and keep the stack low are the ones that win.
Every ranked match (seed, model commit SHAs, scores, Elo before/after) is logged in a public dataset.
▶ Play: DedeProGames/SLM-Tetris-Arena
📊 Results: DedeProGames/lm-tetris-arena-results
Want your model in the Ranked pool? Drop it in the comments!
- 1 reply