ggml-org/gemma-4-E4B-it-GGUF
Same weights, converted (fewer layers) vs google/gemma-4-E4B-it. ggml-org/gemma-4-E4B-it-GGUF matches part of google/gemma-4-E4B-it (677 of its 2130 tensors; the rest of google/gemma-4-E4B-it has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → BF16/F32), plus 43 new tensors.
Tensors changed: Embeddings 0/2, Norms 0/296, MLP 0/126, other 0/85, Attention 0/168
1 derivatives checked
- trek47/gemma-4-E4B-it-Ampere: Same weights, repacked