ccharnkij/gemma-4-31B-it-Uncensored-GGUF
LoRA-style (fewer layers) vs google/gemma-4-31B-it. ccharnkij/gemma-4-31B-it-Uncensored-GGUF matches part of google/gemma-4-31B-it (772 of its 1188 tensors; the rest of google/gemma-4-31B-it has no counterpart). On the shared part: ccharnkij/gemma-4-31B-it-Uncensored-GGUF is a GGUF (F16) of different weights. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, k_proj, o_proj, q_proj, up_proj, v_proj changed, in 60 of 60 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/361, MLP 180/180, Attention 230/230