GODELEV/PretrainingDatasets
0 Bytes
🔄 In a Training Loop
Mr Munk
GODELEV
AI & ML interests
High schooler by day, LLM builder by night. Driven by a deep love for both Physics and AI. Currently spending my runtime building on Hugging Face, experimenting with transformer architectures, and training custom LLMs.
Recent Activity
reacted to TobiasLogic's post with 🔥 about 4 hours ago
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 leads the models in our comparison on ArithMark-3, scoring 43.80, about 4.2 points ahead of the next result.
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:
https://huggingface.co/bench-labs/cagliostro-v3
Built by Bench Labs.
Open SLM Leaderboard:
https://huggingface.co/spaces/AxiomicLabs/Open_SLM_Leaderboard
reacted to Datdanboi25's post with 🔥 about 7 hours ago
THE SLM FRONTIER ADVANCES!
https://huggingface.co/bench-labs/cagliostro-v3 just hit an Intelligence Index of 26.13 on the https://huggingface.co/spaces/AxiomicLabs/Open_SLM_Leaderboard a 146M-param model trained completely from scratch on a single consumer GPU. That's 2nd place overall, and as far as I can tell, the most capable SLM trained on consumer hardware to date. Beating SmolLM-135m on 1/8th of the data is just silly levels of efficiency.
Big congrats to the @BenchLabs team and specifically @TobiasLogic !
new activity about 7 hours ago
bench-labs/cagliostro-v3:Hats Off !!Organizations
