DarkArtsForge/Magistaroth-24B-v1.2

Full fine-tune vs Darkhn/Magistral-2509-24B-Text-Only. Fine-tune: 100% of the weights changed, across 40 of 40 layers.

Tensors changed: Embeddings 1/1, Norms 7/81, MLP 120/120, Attention 160/160, lm_head 1/1

3 derivatives checked