microsoft/Phi-4-reasoning-plus
Full fine-tune vs microsoft/phi-4. Fine-tune: 100% of the weights changed, across 40 of 40 layers.
Tensors changed: Embeddings 1/1, Norms 81/81, MLP 80/80, Attention 80/80, lm_head 1/1
24 derivatives checked
- lmstudio-community/Phi-4-reasoning-plus-MLX-4bit: Quantization
- unsloth/Phi-4-reasoning-plus-GGUF: Same weights, converted
- lmstudio-community/Phi-4-reasoning-plus-GGUF: Quantization
- bartowski/microsoft_Phi-4-reasoning-plus-GGUF: Same weights, converted
- amd/Phi-4-reasoning-plus-w4a16-llmcompressor: Quantization
- mradermacher/Phi-4-reasoning-plus-i1-GGUF: Quantization
- second-state/Phi-4-reasoning-plus-GGUF: Same weights, converted
- nvidia/Phi-4-reasoning-plus-NVFP4: Quantization
- mradermacher/Phi-4-reasoning-plus-GGUF: Quantization
- amd/Phi-4-reasoning-plus-w4a16-asym-torchao-v0.17.0: Quantization
- DevQuasar/microsoft.Phi-4-reasoning-plus-GGUF: Same weights, converted
- jkim96/Phi-4-reasoning-plus-DASHQ-Q2-GGUF: Quantization
- ghostoverflow/Phi-4-reasoning-plus-Q4_K_M-GGUF: Quantization
- gaianet/Phi-4-reasoning-plus-GGUF: Same weights, converted
- nvidia/Phi-4-reasoning-plus-FP8: Quantization
- mmnga/Phi-4-reasoning-plus-gguf: Quantization
- mlx-community/Phi-4-reasoning-plus-4bit: Quantization
- pbhappliedsystems/phi-4-reasoning-plus-gguf-Q4-K-M: Quantization
- unsloth/Phi-4-reasoning-plus-unsloth-bnb-4bit: Quantization
- amd/Phi-4-reasoning-plus-w8a8-llmcompressor: Quantization
- ghostoverflow/Phi-4-reasoning-plus-Q8_0-GGUF: Quantization
- mlx-community/Phi-4-reasoning-plus-8bit: Quantization
- rubricreward/R3-Phi-4-reasoning-plus-14k: Full fine-tune
- rubricreward/R3-Phi-4-reasoning-plus-LoRA-5K-v1.1: LoRA-style
6 other repos hold byte-for-byte identical weight files.