YOYO-AI/YOYO-O1-32B-V4-preview4

Full fine-tune vs Qwen/Qwen2.5-Coder-32B. Fine-tune: 100% of the weights changed, across 64 of 64 layers.

Tensors changed: Norms 129/129, MLP 192/192, Attention 448/448

2 derivatives checked