hell0ks/Solar-Open-100B-jailbreak
LoRA-style vs upstage/Solar-Open-100B. 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 21 of 48 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, MLP 230/2709, Attention 84/192, Norms 0/97, lm_head 0/1
1 derivatives checked
- hell0ks/Solar-Open-100B-jailbreak-gguf: Quantization