Skip to content

CHANGELOG.md

Latest commit

 

History

History
189 lines (119 loc) · 5.03 KB

CHANGELOG.md

File metadata and controls

189 lines (119 loc) · 5.03 KB

更新日志

本项目的所有重要更改都将记录在此文件中。

本项目遵循 语义化版本。

v5.6.0

✨ New Features

  • Batch Image Generation:
    • Added support for generating multiple images in a single batch (1-10 images per request).
    • Configurable "Batch Size" (生成数量) slider in Generation Quality settings.
    • Real-time sequential generation display with progress tracking (e.g., (1/4), (2/4)) directly in the chat interface.

v5.5.0

  • Sampling Methods:
    • Select the specific algorithm used to denoise the image during generation (Euler, Euler a, DPM++ 2M, LCM, DDIM, TCD, etc.).
    • Dynamically switches sampler directly in the stable-diffusion.cpp native backend.

v5.4.0

  • Modify Android application name

v5.3.0

  • add copy & retry

v5.2.0

✨ New Features

  • Image Metadata Injection:
    • Saved PNG images now automatically embed generation parameters (prompt, negative prompt, seed, steps, cfg, model, etc.) as tEXt chunks.
    • Seamlessly supported across all platforms (Android, Desktop, iOS).
  • LoRA Support:
    • Load external LoRA models (.safetensors) to adjust image generation style and details.
    • Added intuitive UI in settings for managing multiple LoRAs (enable/disable, adjust strength, remove).
    • Fully integrated with the native generation pipeline across Android, Desktop, and iOS platforms.

v5.1.0

  • Fix: com/sun/security/auth/module/UnixSystem error on Linux

v5.0.0

  • IOS support

v4.4.0

  • fix linux sdloader error

v4.3.0

  • translation modify

v4.2.1

  • android enable mmap default

v4.2.0

  • modify: offloadToCpu && keepClipOnCpu && keepVaeOnCpu

v4.1.0

  • fix: linux CXXABL1.3.13'not found problem

v4.0.0

  • fix linux system error:The following packages have unmet dependencies
  • fix android ui visible problem and exception

v3.2.0

  • add clipLPath、clipGPath、t5xxlPath

v3.1.0

  • Fix FLUX Model loading issue

v3.0.0

🎨 Advanced Settings Overhaul

New Features

  • CPU Offloading Controls: Added three independent toggles for fine-grained memory management

    • offloadToCpu: Offload model computations to CPU to save GPU/memory
    • keepClipOnCpu: Keep CLIP model on CPU to reduce GPU memory usage
    • keepVaeOnCpu: Keep VAE decoder on CPU to reduce GPU memory usage
    • All parameters now user-controllable in Advanced Settings (previously hardcoded)

  • Expanded Quantization Options: Increased from 6 to 13 quantization types

    • Added K-series variants: Q2_K, Q3_K, Q4_K, Q5_K, Q6_K (better quality at same bit-depth)
    • Added BF16 (Brain Float 16) for modern AI hardware optimization
    • New "Auto" mode (default) that lets the library select optimal quantization
    • Full list: Auto → F32 → F16 → BF16 → Q8_0 → Q6_K → Q5_K → Q5_0 → Q4_K → Q4_0 → Q3_K → Q2_K

UI/UX Improvements

  • Enhanced Flash Attention Documentation

    • Detailed benefits info box showing 30-50% memory reduction
    • Compatibility notes for NVIDIA GPUs (compute capability ≥7.5)
    • Clear recommendations: Enable for devices with <8GB RAM
    • Dynamic info box (shows only when enabled)

  • Comprehensive Quantization Guide

    • Warning box alerting users about re-initialization requirements
    • Detailed info box with memory usage and quality tradeoffs for all 13 types
    • Example: "Q4_K: 4-bit K-variant, ~1GB, better than Q4_0"
    • Note explaining K-variants offer superior quality at same bit-depth

  • Visual Design Enhancements

    • Color-coded info boxes:
      • 🟧 Orange warning box for quantization risks
      • 🔵 Blue info box for technical details
      • 🟢 Green success box for Flash Attention benefits
    • Monospace font for technical specifications
    • Improved readability with proper line spacing

Technical Changes

  • Default Value Changes

    • wtype default changed from 0 (F32) to -1 (Auto)
    • C++ JNI layer now skips wtype assignment when value is -1
    • Allows underlying stable-diffusion.cpp library to use its optimal defaults

  • Internationalization

    • Added 15+ new string resources (English + Chinese)
    • All new settings fully localized

Developer Notes

  • Updated DiffusionLoader interface across all platforms (common, Android, Desktop)
  • Modified ChatViewModel to include new mutable state properties
  • Enhanced AdvancedSettingScreen.kt with conditional rendering and info boxes
  • C++ native code (diffusion_loader_jni.cpp) updated to handle new parameters

v2.3.0

  • Flash Attention Switch
  • Quantization Type

v2.2.0

  • add image to save

v2.1.0

  • Android file selection optimization

v2.0.0

  • support z_image_turbo model

v1.3.2

  • ggml and sd backend use METAL in

v1.3.1

  • fix dylib copy in MacOS

v1.3.0

  • jvmArgs change

v1.2.0

  • adapt to android soft keyboard input
  • fix android 16kb size
  • android app icon add

v1.1.0

  • Custom generated parameters

v1.0.0

  • MineStableDiffusion First Release
  • Core:
    • Multiplatform (Android,DeskTop)
    • txt2Img
    • MultiLanguage