ihounie/huggy-2-1k-alpaca
LoRA-style vs huggyllama/llama-7b. 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 32 of 32 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/65, MLP 0/96, Attention 128/128, lm_head 0/1
2 derivatives checked
- mradermacher/huggy-2-1k-alpaca-i1-GGUF: Quantization
- mradermacher/huggy-2-1k-alpaca-GGUF: Same weights, converted