piskle/Phi-4-14B-IQ2_XXS-GGUF overview
Microsoft's Phi 4 14B, but heavily simplified. This model's raw weights were converted to GGUF compliant IQ2 XXS, ensuring some kind of stability with the heav…
Runs locally from ~4.24 GB disk (8 GB VRAM class GPUs with llama.cpp / guIDE).
Repository Files & Downloads
| File | Type | Quantization | Size | Link |
|---|---|---|---|---|
| Phi-4-14B-IQ2_XXS.gguf | GGUF | IQ2_XXS | 4.24 GB | Download |
Model Details
Model README
---
license: mit
base_model:
- microsoft/phi-4
tags:
- microsoft
- phi
- quantized
---
Microsoft's Phi 4 14B, but heavily simplified.
This model's raw weights were converted to GGUF-compliant IQ2_XXS, ensuring (some kind of) stability with the heaviest compression possible. For smooth inference, it is recommended to have at least of 16GB of RAM or a GPU with 6GB of VRAM, for example an NVIDIA RTX 3050. The model averages ~5-8tok/s on the Apple MacBook Pro (base M1, 8GB of unified memory, 256GB of storage).
To use this model more efficiently, use an extension that allows searching the web (for LM Studio users).
Run piskle/Phi-4-14B-IQ2_XXS-GGUF with guIDE
Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.
Source: Hugging Face · Compare models