mfielding92/Grok-3-reasoning-gemma3-12B-distilled-GGUF
Full fine-tune (fewer layers) vs unsloth/gemma-3-12b-it. mfielding92/Grok-3-reasoning-gemma3-12B-distilled-GGUF matches part of unsloth/gemma-3-12b-it (626 of its 1065 tensors; the rest of unsloth/gemma-3-12b-it has no counterpart). On the shared part: mfielding92/Grok-3-reasoning-gemma3-12B-distilled-GGUF is a GGUF (Q8_0) of different weights. Fine-tune: 100% of the weights changed, across 48 of 48 layers.
Tensors changed: Embeddings 1/1, Norms 0/289, MLP 144/144, Attention 192/192