huggingface-KREW/EXAGIRL-7.8B-Instruct
LoRA-style vs LGAI-EXAONE/EXAONE-3.5-7.8B-Instruct. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of k_proj, q_proj, v_proj changed, in 32 of 32 layers; norms, biases and embeddings are untouched.
Tensors changed: Attention 96/128, Norms 0/65, MLP 0/96, Embeddings 0/1, lm_head 0/1
2 derivatives checked
- mradermacher/EXAGIRL-7.8B-Instruct-i1-GGUF: Quantization
- mradermacher/EXAGIRL-7.8B-Instruct-GGUF: Same weights, converted