TitleOS/HomePhi4_4B_FP16
LoRA-style vs microsoft/Phi-4-mini-reasoning. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_up_proj, o_proj, qkv_proj changed, in 32 of 32 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/65, MLP 64/64, Attention 64/64
3 derivatives checked
- mradermacher/HomePhi4_4B_FP16-i1-GGUF: Quantization
- mradermacher/HomePhi4_4B_FP16-GGUF: Same weights, converted
- TitleOS/HomePhi4_4B_Q4_K_M-GGUF: Quantization