Alniyat5f/WuQi-V3-Onyx-27B
LoRA-style vs Qwen/Qwen3.8-27B. 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 63 of 64 layers; norms, biases and embeddings are untouched.
Tensors changed: Norms 0/325, Attention 64/560, MLP 192/307, Embeddings 0/5, lm_head 0/1, other 0/1
3 derivatives checked
- mradermacher/WuQi-V3-Onyx-27B-GGUF: Quantization
- mradermacher/WuQi-V3-Onyx-27B-i1-GGUF: Quantization
- Alniyat5f/WuQi-V3-Onyx-27b-GGUF: Quantization