RadLite: Multi-Task LoRA Fine-Tuning of Small Language Models for CPU-Deployable Radiology AI

• URL: http://arxiv.org/abs/2605.00421v2
• Date: Mon, 04 May 2026 16:31:36 GMT
• Title: RadLite: Multi-Task LoRA Fine-Tuning of Small Language Models for CPU-Deployable Radiology AI
• Abstract summary: Large language models (LLMs) show promise in radiology but their deployment is limited by computational requirements that preclude use in resource-constrained clinical environments.<n>We investigate whether small language models (SLMs) of 3-4 billion parameters can achieve strong multi-task radiology performance through LoRA fine-tuning, enabling deployment on consumer-grade CPUs.<n>Our work demonstrates that small, efficiently fine-tuned models can serve as practical multi-task radiology AI assistants deployable entirely on consumer hardware without GPU requirements.
• Score: 1.5161700234179216
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Large language models (LLMs) show promise in radiology but their deployment is limited by computational requirements that preclude use in resource-constrained clinical environments. We investigate whether small language models (SLMs) of 3-4 billion parameters can achieve strong multi-task radiology performance through LoRA fine-tuning, enabling deployment on consumer-grade CPUs. We train Qwen2.5-3B-Instruct and Qwen3-4B on 162K samples spanning 9 radiology tasks - RADS classification across 10 systems, impression generation, temporal comparison, radiology NLI, NER, abnormality detection, N/M staging, and radiology Q&A - compiled from 12 public datasets. Both models are evaluated on up to 500 held-out test samples per task with standardized metrics. Our key findings are: (1) LoRA fine-tuning dramatically improves performance over zero-shot baselines (RADS accuracy +53%, NLI +60%, N-staging +89%); (2) the two models exhibit complementary strengths - Qwen2.5 excels at structured generation tasks while Qwen3 dominates extractive tasks; (3) a task-outed oracle ensemble combining both models achieves the best performance across all tasks; (4) few-shot prompting with fine-tuned models hurts performance, demonstrating that LoRA adaptation is more effective than in-context learning for specialized domains; and (5) models can be quantized to GGUF format (~1.8-2.4GB) for CPU deployment at 4-8 tokens/second on consumer hardware. Our work demonstrates that small, efficiently fine-tuned models - which we collectively call RadLite - can serve as practical multi-task radiology AI assistants deployable entirely on consumer hardware without GPU requirements. Code and models are available at https://github.com/RadioX-Labs/RadLite

Related papers

• Embodied-R1.5: Evolving Physical Intelligence via Embodied Foundation Models [63.96252564303143]
  Embodied-R1.5 is a unified Embodied Foundation Model (EFM) that integrates comprehensive embodied reasoning capabilities.<n>We build a large-scale data system of over 15B tokens, and design a multi-task balanced RL recipe.<n>With only 8B parameters, Embodied-R1.5 SOTA on 16 out of 24 embodied VLM benchmarks, surpassing leading models like Gemini-Robotics-ER-1.5 and GPT-5.4.
  arXiv  Detail & Related papers  (2026-06-09T18:07:50Z)
• RADSeg: Unleashing Parameter and Compute Efficient Zero-Shot Open-Vocabulary Segmentation Using Agglomerative Models [6.977949425464]
  We leverage an overlooked agglomerative vision foundation model, RADIO, to improve zero-shot OVSS along three key axes simultaneously: mIoU, latency, and parameter efficiency.<n>Our approach, RADSeg, achieves 6-30% mIoU improvement in the base ViT class while being 3.95x faster and using 2.5x fewer parameters.
  arXiv  Detail & Related papers  (2025-11-24T21:15:01Z)
• One Model to Critique Them All: Rewarding Agentic Tool-Use via Efficient Reasoning [54.580646706013965]
  Reward models (RMs) play a critical role in aligning large language models with human preferences.<n>We introduce ToolRM, a family of lightweight generative RMs tailored for general tool-use scenarios.<n>To build these models, we propose a novel pipeline that constructs pairwise preference data using rule-based scoring and multidimensional sampling.
  arXiv  Detail & Related papers  (2025-10-30T06:08:27Z)
• CURA: Size Isnt All You Need - A Compact Universal Architecture for On-Device Intelligence [0.6244816393907943]
  We propose CURA, an architecture that provides a compact and lightweight solution for diverse machine learning tasks.<n>For compactness, it achieved equivalent accuracy using up to 2,500 times fewer parameters compared to baseline models.<n>For generalizability, it demonstrated consistent performance across four NLP benchmarks and one computer vision dataset.
  arXiv  Detail & Related papers  (2025-09-29T11:06:37Z)
• Structuring Radiology Reports: Challenging LLMs with Lightweight Models [5.01440254761063]
  Large language models (LLMs) have shown strong capabilities in reformatting clinical text, their high computational requirements, lack of transparency, and data privacy concerns hinder practical deployment.<n>We explore lightweight encoder-decoder models (300M parameters)-specifically T5 and BERT2BERT-for structuring radiology reports from the MIMIC-CXR and CheXpert Plus datasets.<n>Our best-performing lightweight model outperforms all LLMs adapted using prompt-based techniques on a human-annotated test set.
  arXiv  Detail & Related papers  (2025-05-30T20:12:51Z)
• EfficientLLM: Efficiency in Large Language Models [64.3537131208038]
  Large Language Models (LLMs) have driven significant progress, yet their growing counts and context windows incur prohibitive compute, energy, and monetary costs.<n>We introduce EfficientLLM, a novel benchmark and the first comprehensive empirical study evaluating efficiency techniques for LLMs at scale.
  arXiv  Detail & Related papers  (2025-05-20T02:27:08Z)
• Empowering Smaller Models: Tuning LLaMA and Gemma with Chain-of-Thought for Ukrainian Exam Tasks [0.0]
  Small or compact models, though more efficient, often lack sufficient support for underrepresented languages.<n>This work explores the potential of parameter-efficient fine-tuning of compact open-weight language models to handle reasoning-intensive tasks.<n> tuning method with joint task topic and step-by-step solution generation outperforms standard chain-of-thought tuning in matching tasks.
  arXiv  Detail & Related papers  (2025-03-18T07:44:49Z)
• AIvaluateXR: An Evaluation Framework for on-Device AI in XR with Benchmarking Results [55.33807002543901]
  We present AIvaluateXR, a comprehensive evaluation framework for benchmarking large language models (LLMs) running on XR devices.<n>We deploy 17 selected LLMs across four XR platforms: Magic Leap 2, Meta Quest 3, Vivo X100s Pro, and Apple Vision Pro, and conduct an extensive evaluation.<n>We propose a unified evaluation method based on the 3D Optimality theory to select the optimal device-model pairs from quality and speed objectives.
  arXiv  Detail & Related papers  (2025-02-13T20:55:48Z)
• PIVOT-R: Primitive-Driven Waypoint-Aware World Model for Robotic Manipulation [68.17081518640934]
  We propose a PrIrmitive-driVen waypOinT-aware world model for Robotic manipulation (PIVOT-R) PIVOT-R consists of a Waypoint-aware World Model (WAWM) and a lightweight action prediction module. Our PIVOT-R outperforms state-of-the-art open-source models on the SeaWave benchmark, achieving an average relative improvement of 19.45% across four levels of instruction tasks.
  arXiv  Detail & Related papers  (2024-10-14T11:30:18Z)
• Distributed Inference and Fine-tuning of Large Language Models Over The Internet [91.00270820533272]
  Large language models (LLMs) are useful in many NLP tasks and become more capable with size. These models require high-end hardware, making them inaccessible to most researchers. We develop fault-tolerant inference algorithms and load-balancing protocols that automatically assign devices to maximize the total system throughput.
  arXiv  Detail & Related papers  (2023-12-13T18:52:49Z)
• CodeGen2: Lessons for Training LLMs on Programming and Natural Languages [116.74407069443895]
  We unify encoder and decoder-based models into a single prefix-LM. For learning methods, we explore the claim of a "free lunch" hypothesis. For data distributions, the effect of a mixture distribution and multi-epoch training of programming and natural languages on model performance is explored.
  arXiv  Detail & Related papers  (2023-05-03T17:55:25Z)
• A Light Weight Model for Active Speaker Detection [7.253335671577093]
  We construct a lightweight active speaker detection architecture by reducing input candidates, splitting 2D and 3D convolutions for audio-visual feature extraction, and applying gated recurrent unit (GRU) with low computational complexity for cross-modal modeling. Experimental results on the AVA-ActiveSpeaker dataset show that our framework achieves competitive mAP performance (94.1% vs. 94.2%). Our framework also performs well on the Columbia dataset showing good robustness.
  arXiv  Detail & Related papers  (2023-03-08T08:40:56Z)
• LoRA: Low-Rank Adaptation of Large Language Models [71.75808607987281]
  Low-Rank Adaptation, or LoRA, freezes the pre-trained model weights and injects trainable rank decomposition into each layer of the Transformer architecture. For GPT-3, LoRA can reduce the number of trainable parameters by 10,000 times and the computation hardware requirement by 3 times compared to full fine-tuning.
  arXiv  Detail & Related papers  (2021-06-17T17:37:18Z)
• LiteMuL: A Lightweight On-Device Sequence Tagger using Multi-task Learning [1.3192560874022086]
  LiteMuL is a lightweight on-device sequence tagger that can efficiently process the user conversations using a Multi-Task Learning approach. Our model is competitive with other MTL approaches for NER and POS tasks while outshines them with a low memory footprint.
  arXiv  Detail & Related papers  (2020-12-15T19:15:54Z)
• REST: Robust and Efficient Neural Networks for Sleep Monitoring in the Wild [62.36144064259933]
  We propose REST, a new method that simultaneously tackles both issues via adversarial training and controlling the Lipschitz constant of the neural network. We demonstrate that REST produces highly-robust and efficient models that substantially outperform the original full-sized models in the presence of noise. By deploying these models to an Android application on a smartphone, we quantitatively observe that REST allows models to achieve up to 17x energy reduction and 9x faster inference.
  arXiv  Detail & Related papers  (2020-01-29T17:23:16Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.