rhaymison/Llama-portuguese-13b-Luana-v0.2
LoRA-style vs meta-llama/Llama-2-13b-chat-hf. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of 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 80/160, lm_head 0/1
2 derivatives checked
- mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF: Quantization
- mradermacher/Llama-portuguese-13b-Luana-v0.2-GGUF: Quantization