Sakalti/SJT-14B

Partial fine-tune vs djuna/Q2.5-Veltha-14B. Partial change: 13% of the weights changed (attention, embeddings, mlp, norms), in 25 of 48 layers.

Tensors changed: lm_head 0/1, Embeddings 1/1, Norms 6/97, MLP 13/144, Attention 16/336

1 derivatives checked