macadeliccc/magistrate-3.2-3b-base
Partial fine-tune vs meta-llama/Llama-3.2-3B. Partial change: 47% of the weights changed (attention, embeddings, mlp), in 27 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 0/57, MLP 42/84, Attention 14/112
3 derivatives checked
- macadeliccc/magistrate-3.2-3b-it-GGUF: Partial fine-tune
- mradermacher/magistrate-3.2-3b-base-GGUF: Same weights, converted
- macadeliccc/magistrate-3.2-3b-it: Partial fine-tune