This Hugging Face repository by MaziyarPanahi contains GGUF quantized versions of Microsoft's Phi-4 language model. The files support various quantization levels (Q2_K through Q8_0 and FP16) for running the model locally using tools like llama.cpp or Ollama. It enables efficient inference of a capable small language model on CPUs and consumer GPUs without requiring high-end hardware or cloud services.
Phi 4 is a Foundation models & chat project. It focuses on running powerful language models efficiently on consumer-grade hardware without cloud dependency. Phi 4 is an open-source project aimed at developers and AI enthusiasts. The project is open source (MIT). Phi 4 is available on the web, the command line, and API.
Maziyar Panahi builds and maintains Phi 4, and it first shipped in 2023. The project is developed in the open on GitHub with 122.5k stars and 1.2k commits in the last 90 days. Key capabilities include Quantized Models, Local Inference, and Multiple Precision Levels.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do