<!-- mobian-agent-page publisher="dailydev" canonical="https://daily.dev/posts/lfm2-5-q4-0-checkpoints-from-quantization-aware-distillation-qyvxdigkf" -->

---
title: LFM2.5 Q4_0 Checkpoints from Quantization-Aware Distillation
description: 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),...
canonical: https://daily.dev/posts/lfm2-5-q4-0-checkpoints-from-quantization-aware-distillation-qyvxdigkf
twitter:card: summary_large_image
twitter:site: @dailydotdev
og:type: website
og:site_name: daily.dev
og:title: LFM2.5 Q4_0 Checkpoints from Quantization-Aware Distillation | daily.dev
og:description: 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),...
og:url: https://daily.dev/posts/lfm2-5-q4-0-checkpoints-from-quantization-aware-distillation-qyvxdigkf
og:image: https://api.daily.dev/og/posts/QYvXDiGkF.png
og:image:alt: LFM2.5 Q4_0 Checkpoints from Quantization-Aware Distillation
og:image:width: 1200
og:image:height: 630
og:locale: 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 GGUF models?

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

_Developers optimizing on-device model deployment can track quantization advances like this on daily.dev._

### How does LFM2.5 QAD Q4_0 performance compare to Q4_K_M and Q5_K_M quantization on edge hardware?

The 230M and 350M QAD Q4_0 checkpoints match Q5_K_M quality within evaluation variance while delivering 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, tested on MacBook Pro, NucBox EVO-X2, Samsung Galaxy S26 Ultra, and Raspberry Pi 5 hardware.

_Anyone choosing a quantization format for edge deployment can weigh trade-offs like these via daily.dev._

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

Use llama-cli with the -hf flag pointing to the LiquidAI Hugging Face repo and the --hf-file flag specifying the QAD Q4_0 gguf filename, for example: llama-cli -hf LiquidAI/LFM2.5-350M --hf-file LFM2.5-350M-QAD-Q4_0.gguf -p "your prompt". Any runtime supporting GGUF Q4_0 artifacts works.

_Developers wiring up local model inference can find setup steps like this through daily.dev._

## Similar posts on daily.dev

- [Unsloth Dynamic 2.0 GGUFs](https://daily.dev/posts/unsloth-dynamic-2-0-ggufs-pggi40vtt) · Hacker News · 2 upvotes · 0 comments
- [Up to 3.2x Faster Inference with LFM2.5-DSpark](https://daily.dev/posts/up-to-3-2x-faster-inference-with-lfm2-5-dspark-nwilr6rly) · Hugging Face · 0 upvotes · 0 comments
- [Untitled](https://daily.dev/posts/untitled-qawtuhuih) · SitePoint · 1 upvotes · 0 comments

---

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)

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]}
{"@context":"https://schema.org","@type":"TechArticle","headline":"LFM2.5 Q4_0 Checkpoints from Quantization-Aware Distillation","url":"https://daily.dev/posts/lfm2-5-q4-0-checkpoints-from-quantization-aware-distillation-qyvxdigkf","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/lfm2-5-q4-0-checkpoints-from-quantization-aware-distillation-qyvxdigkf"},"datePublished":"2026-08-19T13:51:37.915Z","dateModified":"2026-09-14T06:43:37.417Z","description":"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),...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/f4ff5d1e3e210b6bfd78a662f64aae32?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/f4ff5d1e3e210b6bfd78a662f64aae32?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Hugging Face","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Hugging Face","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/f1f55c67d81a4330acf5b90b26b0c8e1","url":"https://daily.dev/sources/huggingface"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/lfm2-5-q4-0-checkpoints-from-quantization-aware-distillation-qyvxdigkf","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"data-science,edge-computing,llama-cpp","timeRequired":"PT2M"}
{"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Hugging Face","item":"https://daily.dev/sources/huggingface"},{"@type":"ListItem","position":3,"name":"LFM2.5 Q4_0 Checkpoints from Quantization-Aware Distillation"}]}
{"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/lfm2-5-q4-0-checkpoints-from-quantization-aware-distillation-qyvxdigkf#faq","mainEntity":[{"@type":"Question","name":"What is quantization-aware distillation (QAD) and how does it improve Q4_0 GGUF models?","acceptedAnswer":{"@type":"Answer","text":"Quantization-aware distillation trains a quantized student model directly from a high-precision teacher model, rather than quantizing an already-trained model after the fact. Applied to LFM2.5 models (230M, 350M, 1.2B-Instruct, 2.6B), QAD Q4_0 checkpoints recover 96.5-97.4% of BF16 baseline accuracy while keeping the same memory and speed as standard post-training quantized Q4_0 GGUFs. Developers optimizing on-device model deployment can track quantization advances like this on daily.dev."}},{"@type":"Question","name":"How does LFM2.5 QAD Q4_0 performance compare to Q4_K_M and Q5_K_M quantization on edge hardware?","acceptedAnswer":{"@type":"Answer","text":"The 230M and 350M QAD Q4_0 checkpoints match Q5_K_M quality within evaluation variance while delivering 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, tested on MacBook Pro, NucBox EVO-X2, Samsung Galaxy S26 Ultra, and Raspberry Pi 5 hardware. Anyone choosing a quantization format for edge deployment can weigh trade-offs like these via daily.dev."}},{"@type":"Question","name":"How do I run the LFM2.5 QAD Q4_0 GGUF checkpoints with llama.cpp?","acceptedAnswer":{"@type":"Answer","text":"Use llama-cli with the -hf flag pointing to the LiquidAI Hugging Face repo and the --hf-file flag specifying the QAD Q4_0 gguf filename, for example: llama-cli -hf LiquidAI/LFM2.5-350M --hf-file LFM2.5-350M-QAD-Q4_0.gguf -p \"your prompt\". Any runtime supporting GGUF Q4_0 artifacts works. Developers wiring up local model inference can find setup steps like this through daily.dev."}}]}
```