WhiskyAKM/Gemma-4-E4B-it-qat-GGUF

Full fine-tune (fewer layers) vs google/gemma-4-E4B-it. WhiskyAKM/Gemma-4-E4B-it-qat-GGUF matches part of google/gemma-4-E4B-it (623 of its 2130 tensors; the rest of google/gemma-4-E4B-it has no counterpart). On the shared part: WhiskyAKM/Gemma-4-E4B-it-qat-GGUF is a GGUF (BF16) of different weights. Fine-tune: 100% of the weights changed, across 42 of 42 layers.

Tensors changed: Embeddings 2/2, Norms 184/278, MLP 126/126, other 85/85, Attention 132/132

On the Hub · Downloads on Model Pulse