BioMistral/BioMistral-DARE-NS

Full fine-tune vs BioMistral/BioMistral-7B-DARE. Fine-tune: 100% of the weights changed, across 32 of 32 layers.

Tensors changed: Embeddings 1/1, Norms 65/65, MLP 96/96, Attention 128/128, lm_head 1/1

2 derivatives checked

• mradermacher/BioMistral-DARE-NS-GGUF: Same weights, converted
• jsfs11/MixtureofMerges-MoE-4x7bRP-v11: MoE built from dense models

On the Hub · Downloads on Model Pulse