SicariusSicariiStuff/2B-ad
LoRA-style vs google/gemma-2-2b-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, up_proj, v_proj changed, in 26 of 26 layers; norms, biases and embeddings are untouched.
Tensors changed: Norms 0/105, MLP 78/78, Attention 78/104
3 derivatives checked
- mradermacher/2B-ad-i1-GGUF: Quantization
- mradermacher/2B-ad-GGUF: Same weights, converted
- SicariusSicariiStuff/2B-ad_GGUFs: Quantization