nbeerbower/gemma2-gutenberg-27B
LoRA-style vs google/gemma-2-27b-it. 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 46 of 46 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/185, MLP 138/138, Attention 184/184
3 derivatives checked
- bartowski/gemma2-gutenberg-27B-GGUF: Same weights, converted
- mradermacher/gemma2-gutenberg-27B-i1-GGUF: Quantization
- mradermacher/gemma2-gutenberg-27B-GGUF: Quantization
1 other repos hold byte-for-byte identical weight files.