Pith. sign in

REVIEW 6 cited by

Progressive Knowledge Distillation Of Stable Diffusion XL Using Layer Level Loss

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2401.02677 v1 pith:HWD5WEDM submitted 2024-01-05 cs.CV cs.AI

classification cs.CVcs.AI
keywords sdxldiffusionknowledgemodelmodelsstabledistillationgenerative
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Stable Diffusion XL (SDXL) has become the best open source text-to-image model (T2I) for its versatility and top-notch image quality. Efficiently addressing the computational demands of SDXL models is crucial for wider reach and applicability. In this work, we introduce two scaled-down variants, Segmind Stable Diffusion (SSD-1B) and Segmind-Vega, with 1.3B and 0.74B parameter UNets, respectively, achieved through progressive removal using layer-level losses focusing on reducing the model size while preserving generative quality. We release these models weights at https://hf.co/Segmind. Our methodology involves the elimination of residual networks and transformer blocks from the U-Net structure of SDXL, resulting in significant reductions in parameters, and latency. Our compact models effectively emulate the original SDXL by capitalizing on transferred knowledge, achieving competitive results against larger multi-billion parameter SDXL. Our work underscores the efficacy of knowledge distillation coupled with layer-level losses in reducing model size while preserving the high-quality generative capabilities of SDXL, thus facilitating more accessible deployment in resource-constrained environments.

Discussion (0). Sign in to comment.

Forward citations

Cited by 6 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. DynEval: Holistic Evaluations of T2I Generative Models in the Wild

    cs.CV 2026-07 conditional novelty 6.5 of 10

    DynEval distills a 235B teacher VLM into 2B/4B evaluators via 250K synthetic instruction triplets, yielding higher human correlation than existing T2I metrics while enabling open-set dynamic QA and scene-graph quality checks.

  2. OC-Distill: Ontology-aware Contrastive Learning with Cross-Modal Distillation for ICU Risk Prediction

    cs.LG 2026-04 unverdicted novelty 6.0 of 10

    Ontology-aware contrastive pretraining plus note-to-vitals distillation improves MIMIC ICU risk and length-of-stay prediction using only vital signs at inference.

  3. EC-Diff: Fast and High-Quality Edge-Cloud Collaborative Inference for Diffusion Models

    cs.CV 2025-07 reject novelty 5.0 of 10

    EC-Diff accelerates edge-cloud diffusion inference with a k-step noise approximation strategy and a two-stage greedy search for the cloud-edge handoff point, claiming about 2x speedup with preserved quality.

  4. Memory-Efficient Personalization of Text-to-Image Diffusion Models via Selective Optimization Strategies

    cs.CV 2025-07 conditional novelty 5.0 of 10

    A hybrid of low-resolution backpropagation and high-resolution zeroth-order optimization, scheduled by a dynamic timestep-dependent probability, matches full-resolution fine-tuning quality while cutting training memory.

  5. Temporally Consistent Unsupervised Segmentation for Mobile Robot Perception

    cs.CV 2025-07 conditional novelty 4.0 of 10

    Frontier-Seg clusters DINOv2 superpixel features locally per window and globally across a whole video to produce unsupervised, temporally stable terrain segmentations.

  6. The Hidden Cost of an Image: Quantifying the Energy Consumption of AI Image Generation

    cs.LG 2025-06

Pith tools