ab-ai/PII-Model-Phi3-Mini
LoRA-style vs microsoft/Phi-3-mini-4k-instruct. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of qkv_proj changed, in 32 of 32 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/65, MLP 0/64, Attention 32/64, lm_head 0/1
1 derivatives checked
- mradermacher/PII-Model-Phi3-Mini-GGUF: Same weights, converted