VirtuoTuring/justina_clarus-24b
LoRA-style vs VirtuoTuring/chat_noir-24b. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of k_proj, o_proj, q_proj, v_proj changed, in 40 of 40 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/81, MLP 0/120, Attention 160/160, lm_head 0/1
2 derivatives checked
- mradermacher/justina_clarus-24b-i1-GGUF: Quantization
- mradermacher/justina_clarus-24b-GGUF: Quantization