mvpmaster/MistralDpoPearl-7b-slerp
Full fine-tune vs NousResearch/Nous-Hermes-2-Mistral-7B-DPO. Fine-tune: 97% of the weights changed, across 32 of 32 layers.
Tensors changed: Norms 65/65, MLP 93/96, Attention 124/128
2 derivatives checked
- mradermacher/MistralDpoPearl-7b-slerp-GGUF: Same weights, converted
- mvpmaster/PearlMath-lafted-7b-slerp: Full fine-tune