hotdogs/Qwen3.8-27B-abliterated-sme-preview
LoRA-style vs hotdogs/Qwen3.8-27B-abliterated. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, in_proj_a, in_proj_b, in_proj_qkv, in_proj_z, k_proj, o_proj, out_proj, q_proj, up_proj, v_proj changed, in 64 of 64 layers; norms, biases and embeddings are untouched.
Tensors changed: Norms 0/215, Attention 304/452, MLP 192/195, Embeddings 0/2, lm_head 0/1, other 0/1
1 derivatives checked
- hotdogs/Qwen3.8-27B-abliterated-sme-preview-mtp-GGUF: Same weights, converted