proxectonos/Carvalho-Salamandra-Instruct
Partial fine-tune vs BSC-LT/salamandra-7b-instruct. Partial change: 87% of the weights changed (attention, lm_head, mlp, norms), in 32 of 32 layers.
Tensors changed: Embeddings 0/1, Norms 55/65, MLP 96/96, Attention 128/128, lm_head 1/1
1 derivatives checked
- mradermacher/Carvalho-Salamandra-Instruct-GGUF: Same weights, converted