cafepm/Ministral-3-14B-Instruct-2512-Q3Q4-GGUF
Full fine-tune (fewer layers) vs mistralai/Ministral-3-14B-Base-2512. cafepm/Ministral-3-14B-Instruct-2512-Q3Q4-GGUF matches part of mistralai/Ministral-3-14B-Base-2512 (363 of its 585 tensors; the rest of mistralai/Ministral-3-14B-Base-2512 has no counterpart). On the shared part: cafepm/Ministral-3-14B-Instruct-2512-Q3Q4-GGUF is a GGUF (Q2Q3) of different weights. Fine-tune: 100% of the weights changed, across 40 of 40 layers.
Tensors changed: Embeddings 1/1, Norms 54/81, MLP 120/120, Attention 160/160, lm_head 1/1