djuna/G2-GSHT-32K
LoRA-style vs Nekuromento/Hematoma-Gemma-Model-Stock-9B. 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 42 of 42 layers; norms, biases and embeddings are untouched.
Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/169, MLP 126/126, Attention 168/168
1 derivatives checked
- mradermacher/G2-GSHT-32K-GGUF: Same weights, converted
1 other repos hold byte-for-byte identical weight files.