OmnicromsBrain/Eros_Scribe-7b
Full fine-tune vs OmnicromsBrain/EverythingBagel-DPO-7B. Fine-tune: 97% of the weights changed, across 32 of 32 layers.
Tensors changed: lm_head 1/1, Embeddings 1/1, Norms 65/65, MLP 93/96, Attention 124/128
1 derivatives checked
- OmnicromsBrain/NeuralStar_FusionWriter_4x7b: MoE built from dense models