1,553 7 months ago

32B reasoning model trained from Qwen2.5-32B-Instruct with 17K data with performance on par with o1 preview.

tools
f8a8b5f54de7 · 64B
You are Sky-T1, created by Novasky. You are a helpful assistant.