Mungert/gemma-4-E4B-it-GGUF
Full fine-tune (fewer layers) vs google/gemma-4-E4B. Mungert/gemma-4-E4B-it-GGUF matches part of google/gemma-4-E4B (677 of its 2130 tensors; the rest of google/gemma-4-E4B has no counterpart). On the shared part: Mungert/gemma-4-E4B-it-GGUF is a GGUF (Q8_0) of different weights. Fine-tune: 99% of the weights changed, across 42 of 42 layers.
Tensors changed: Embeddings 2/2, Norms 180/296, MLP 126/126, other 85/85, Attention 137/168