C3DS/Windy-Qwen3.5-27B
LoRA-style vs Qwen/Qwen3.5-27B. 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 64 of 64 layers; norms, biases and embeddings are untouched.
Tensors changed: lm_head 0/1, Embeddings 0/5, MLP 192/307, Attention 64/560, other 0/1, Norms 0/325
2 derivatives checked
- mradermacher/Windy-Qwen3.5-27B-i1-GGUF: Quantization
- mradermacher/Windy-Qwen3.5-27B-GGUF: Quantization