kuzaai/kuza-gemma-4-e2b
LoRA-style (fewer layers) vs unsloth/gemma-4-E2B-it-qat-q4_0-unquantized. kuzaai/kuza-gemma-4-e2b matches part of unsloth/gemma-4-E2B-it-qat-q4_0-unquantized (505 of its 1951 tensors; the rest of unsloth/gemma-4-E2B-it-qat-q4_0-unquantized has no counterpart). On the shared part: kuzaai/kuza-gemma-4-e2b is a GGUF (BF16) of different weights. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, k_proj, o_proj, q_proj, up_proj, v_proj changed, in 35 of 35 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/2, Norms 0/227, MLP 105/105, other 0/71, Attention 100/100