DmitryDB/Gemma-3-12B-IT-QAT-Heretic-MOE-v1
Full fine-tune (fewer layers) vs google/gemma-3-12b-it-qat-q4_0-unquantized. DmitryDB/Gemma-3-12B-IT-QAT-Heretic-MOE-v1 matches part of google/gemma-3-12b-it-qat-q4_0-unquantized (626 of its 1065 tensors; the rest of google/gemma-3-12b-it-qat-q4_0-unquantized has no counterpart). On the shared part: DmitryDB/Gemma-3-12B-IT-QAT-Heretic-MOE-v1 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