11
Euler a
1.1
WAI-illustrious-SDXL HSWQ EnhancedFast Like FP8. Stable Like Full Precision.A performance-focused merge built on the HSWQ FP8 edition of WAI-illustrious-SDXL v17.0, enhanced with DMD2, Stabilizer IL, and SPO-SDXL_4k-p_10ep.This model was created with a simple goal:Deliver high-quality anime image generation while maintaining the speed and efficiency advantages of FP8 quantization.By combining modern optimization techniques with carefully selected merges, this checkpoint aims to provide strong prompt adherence, improved stability, and excellent visual quality even at relatively low step counts.What's InsideBase ModelWAI-illustrious-SDXL v17.0Additional MergesDMD2Stabilizer ILSPO-SDXL_4k-p_10epEach component was selected to improve a different aspect of generation quality:DMD2 helps achieve stronger results at lower sampling steps.Stabilizer IL improves generation consistency and reliability.SPO-SDXL_4k-p_10ep enhances composition, structure, and detail retention.The result is a model optimized for both rapid iteration and everyday image generation.What is HSWQ?HSWQ (Hybrid Sensitivity-Weighted Quantization) is an advanced quantization technique designed to reduce model size and VRAM requirements while preserving image quality.Traditional FP8 quantization often applies the same level of compression throughout the entire network. While efficient, this can sometimes introduce quality degradation in layers that are particularly important for image generation.HSWQ takes a smarter approach.Instead of treating every layer equally, it analyzes layer sensitivity and applies quantization more selectively. Critical layers receive greater protection while less sensitive components can be compressed more aggressively.This allows the model to retain much of the visual quality and prompt understanding of the original checkpoint while benefiting from the efficiency of FP8.Benefits of HSWQLower VRAM requirementsFaster loading timesFaster inferenceBetter quality retention than conventional FP8 approachesMore accessible generation on a wider range of hardwareThink of HSWQ as a "smart FP8" approach that preserves precision where it matters most.Key Features⚡ Fast GenerationDesigned around an FP8 workflow for efficient image generation and reduced memory consumption.🎯 Improved Prompt AdherenceAdditional merges help maintain character details, clothing descriptions, and scene composition more accurately.🖼 Enhanced Visual QualityImproved detail retention, stronger compositions, and more consistent image structure.🚀 Low-Step PerformanceProduces attractive results even with relatively low step counts, making it ideal for online generators and rapid experimentation.💡 Balanced WorkflowRather than maximizing a single metric, this model focuses on balancing:SpeedStabilityDetailPrompt responsivenessResource efficiencyRecommended SettingsOnline Generation / Fast WorkflowCFG Scale: 1.0Steps: 10Sampler: Euler aResolution: 1024×1024These settings were the primary target during testing and offer an excellent balance between speed and quality.Higher Quality GenerationCFG Scale: 2.0–3.0Steps: 15–20Resolution: 1024×1024 or higherRecommended Use CasesThis model performs particularly well for:Anime illustrationsCharacter portraitsFull-body character artworkFantasy scenesVibrant lightingExpressive charactersDetailed costumesDynamic compositionsNotesThis is a merge model designed to provide a practical balance between performance and image quality.If you want the absolute fastest workflow possible while maintaining attractive anime-style outputs, this model was built with that goal in mind.Whether you're generating locally on limited hardware or using online generators with restricted resources, this checkpoint aims to make high-quality image creation more accessible.CreditsBase ModelWAI-illustrious-SDXL v17.0Merged ComponentsDMD2Stabilizer ILSPO-SDXL_4k-p_10epSpecial thanks to all original model creators and researchers whose work made this merge possible.Please respect the licenses and usage terms of all source models and components.
