This model is a successor of s1-32B with slightly better performance. Thanks to Bespoke Labs (Ryan Marten) for helping generate r1 traces for s1K with Curator.

Use

Logs: <a href="https://wandb.ai/tikatoka-snu/s1/runs/dfnv09nk" target="_blank" rel="noopener noreferrer" class="text-blue-400 hover:text-blue-300 underline underline-offset-2 transition-colors">https://wandb.ai/tikatoka-snu/s1/runs/dfnv09nk
Repository: <a href="https://github.com/simplescaling/s1" target="_blank" rel="noopener noreferrer" class="text-blue-400 hover:text-blue-300 underline underline-offset-2 transition-colors">simplescaling/s1
Paper: <a href="https://arxiv.org/abs/2501.19393" target="_blank" rel="noopener noreferrer" class="text-blue-400 hover:text-blue-300 underline underline-offset-2 transition-colors">https://arxiv.org/abs/2501.19393

The model usage is documented here.

Model is trained with block_size 20000

TikaToka/s1.1-1.5B-20k-bf16

Model Summary

Use

TikaToka/
s1.1-1.5B-20k-bf16