---
title: "LFM2.5 Q4_0 Checkpoints from Quantization-Aware Distillation"
url: https://daily.dev/posts/lfm2-5-q4-0-checkpoints-from-quantization-aware-distillation-qyvxdigkf
source_url: https://huggingface.co/blog/LiquidAI/qad
type: article
source: "Hugging Face"
published: 2026-08-19T13:51:37.915Z
updated: 2026-08-19T13:54:31.169Z
tags: ["data-science", "edge-computing", "llama-cpp"]
reading_time: 2
upvotes: 1
comments: 0
language: en
---

> ## Documentation Index
> Fetch the complete documentation index at: https://daily.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# LFM2.5 Q4_0 Checkpoints from Quantization-Aware Distillation

**[Hugging Face](https://daily.dev/sources/huggingface)** · 2 min read · 1 upvotes · 0 comments

## Summary

Liquid AI released Q4_0 GGUF checkpoints for its LFM2.5 model family (230M, 350M, 1.2B-Instruct, 2.6B) trained using Quantization-Aware Distillation (QAD), where a high-precision teacher model is distilled directly into a quantized student. These checkpoints recover 96.5-97.4% of BF16 accuracy while keeping the same memory footprint and throughput as native Q4_0, outperforming standard post-training quantization. On edge hardware (MacBook Pro, NucBox EVO-X2, Samsung Galaxy S26 Ultra, Raspberry Pi 5), the QAD checkpoints match Q5_K_M or Q4_K_M quality at 3-33% higher decode throughput, and are compatible with llama.cpp and other GGUF runtimes.

## Full article

daily.dev links to this article rather than hosting it. Read it at the original source: <https://huggingface.co/blog/LiquidAI/qad>

## Questions this post answers

### What is quantization-aware distillation (QAD) and how does it improve Q4_0 quantized models?

Quantization-aware distillation trains a quantized student model directly from a high-precision teacher, rather than quantizing an already-trained model after the fact. Applied to Liquid AI's LFM2.5 models (230M, 350M, 1.2B-Instruct, 2.6B), QAD Q4_0 checkpoints recover 96.5-97.4% of BF16 accuracy while keeping the same memory footprint and throughput as standard Q4_0 GGUFs.

_daily.dev surfaces edge-deployment quantization techniques for engineers picking model formats._

### How much faster are LFM2.5 QAD Q4_0 checkpoints compared to Q5_K_M or Q4_K_M quantization on edge devices?

The 230M and 350M QAD Q4_0 checkpoints match Q5_K_M quality within evaluation variance while achieving 4-33% higher decode throughput. The 1.2B and 2.6B QAD Q4_0 checkpoints match Q4_K_M quality with 3-14% higher throughput, measured across MacBook Pro, NucBox EVO-X2, Samsung Galaxy S26 Ultra, and Raspberry Pi 5.

_Developers benchmarking on-device model formats can track results like these on daily.dev._

### How do I run the LFM2.5 QAD Q4_0 GGUF checkpoints with llama.cpp?

Use llama-cli with the Hugging Face model reference, for example: llama-cli -hf LiquidAI/LFM2.5-350M --hf-file LFM2.5-350M-QAD-Q4_0.gguf -p "prompt text". Any runtime supporting GGUF Q4_0 artifacts can load these checkpoints, which are hosted on Hugging Face for the 230M, 350M, 1.2B-Instruct, and 2.6B LFM2.5 variants.

_daily.dev helps developers keep up with practical steps for deploying quantized edge models._

---

Tags: [#data-science](https://daily.dev/tags/data-science), [#edge-computing](https://daily.dev/tags/edge-computing), [#llama-cpp](https://daily.dev/tags/llama-cpp)

[View this post on daily.dev](https://daily.dev/posts/lfm2-5-q4-0-checkpoints-from-quantization-aware-distillation-qyvxdigkf)