Mungert/gemma-3n-E4B-it-GGUF
Full fine-tune (fewer layers) vs google/gemma-3n-E4B. Mungert/gemma-3n-E4B-it-GGUF matches part of google/gemma-3n-E4B (809 of its 1676 tensors; the rest of google/gemma-3n-E4B has no counterpart). On the shared part: Mungert/gemma-3n-E4B-it-GGUF is a GGUF (Q8_0) of different weights. Fine-tune: 100% of the weights changed, across 35 of 35 layers.
Tensors changed: Embeddings 1/1, other 246/246, Norms 247/317, MLP 105/105, Attention 125/140