diff --git a/changelog/index.mdx b/changelog/index.mdx index ca3a706a..6a590932 100644 --- a/changelog/index.mdx +++ b/changelog/index.mdx @@ -4,6 +4,46 @@ description: "Track ComfyUI's latest features, improvements, and bug fixes" icon: "clock-rotate-left" --- + + +**Revolutionary Image Editing Models & Enhanced Workflow Control** + +This release introduces groundbreaking new image editing capabilities and significant improvements to workflow reliability: + +## Major New Image Editing Models +- **Flux Kontext**: Revolutionary image editing model now integrated with full workflow support, enabling precise image modifications while maintaining consistency with your original content +- **Omnigen2**: Advanced image editing model implementation brings professional-grade editing capabilities directly into your ComfyUI workflows + +## Enhanced Model Compatibility & Performance +- **Flux ControlNet Integration**: Now works seamlessly with SD3 text encoders, expanding your creative control options across different model architectures +- **Flux Optimizations**: Performance improvements reduce generation time and memory usage for Flux-based workflows +- **Cosmos Predict2 Enhancements**: Added model merging nodes for advanced workflow composition and fixed noise mask exceptions for more reliable video generation + +## Improved Workflow Reliability +- **Better Error Handling**: Enhanced out-of-memory error messages provide clearer guidance when workflows exceed system limits +- **Longer Render Support**: Improved logging for renders exceeding 10 minutes, helping track progress on complex, time-intensive workflows +- **PyTorch Warning Fixes**: Cleaner console output with resolved compatibility warnings + +## New Node Capabilities +- **Resize and Pad Image Node**: Streamlined image preprocessing for consistent input dimensions across your workflows +- **ImageStitch Enhancements**: White padding option provides more flexible image composition tools +- **Singlestep DPM++ SDE**: New sampler option for Rectified Flow models, expanding your sampling toolkit + +## Developer & System Improvements +- **Frontend Updates**: Multiple frontend improvements (v1.22.2, v1.23.3, v1.23.4) enhance UI responsiveness and workflow management +- **Template Updates**: Latest template version (0.1.29) includes improved custom node development tools +- **DirectML Guidance**: Updated documentation clarifies hardware recommendations for optimal performance +- **Transformers Compatibility**: Minimum version bump ensures compatibility with latest model implementations + +## Technical Refinements +- **Terminology Updates**: "UNet" references updated to "Diffusion Model" for clearer technical communication +- **FP16 Compatibility**: Omnigen2 FP16 disabled on older PyTorch versions to prevent compatibility issues +- **Requirements Updates**: Dependency updates ensure stable operation with the latest Python packages + +These enhancements make ComfyUI more powerful for AI artists working with advanced image editing workflows while maintaining the stability and performance that creators depend on for professional projects. + + + **Advanced Sampling & Model Control Enhancements** @@ -91,9 +131,7 @@ These changes enhance the reliability of advanced model workflows while maintain - **Python Version Warnings**: Added alerts for outdated Python installations to prevent compatibility issues - **WebcamCapture Fixes**: Corrected IS_CHANGED signature for reliable live input workflows -This release significantly expands ComfyUI's model ecosystem support while delivering crucial stability improvements and enhanced hardware compatibility across different platforms. - - +This release significantly expands ComfyUI's model ecosystem support while delivering crucial stability This release brings powerful new workflow utilities and performance optimizations for ComfyUI creators: diff --git a/zh-CN/changelog/index.mdx b/zh-CN/changelog/index.mdx index 4715759d..a88158bb 100644 --- a/zh-CN/changelog/index.mdx +++ b/zh-CN/changelog/index.mdx @@ -1,9 +1,48 @@ --- title: "更新日志" -description: "跟踪ComfyUI的最新功能、改进和错误修复" +description: "追踪ComfyUI的最新功能、改进和错误修复" icon: "clock-rotate-left" --- + + +**革命性图像编辑模型与增强的工作流控制** + +此版本引入了突破性的新图像编辑功能和工作流可靠性的重大改进： + +## 主要的新图像编辑模型 +- **Flux Kontext**: 革命性图像编辑模型现已集成完整工作流支持，能够进行精确的图像修改，同时保持与原始内容的一致性 +- **Omnigen2**: 先进的图像编辑模型实现为您的ComfyUI工作流带来专业级编辑功能 + +## 增强的模型兼容性与性能 +- **Flux ControlNet集成**: 现在与SD3文本编码器无缝协作，扩展了跨不同模型架构的创意控制选项 +- **Flux优化**: 性能改进减少了基于Flux的工作流的生成时间和内存使用量 +- **Cosmos Predict2增强**: 为高级工作流组合添加了模型合并节点，并修复了噪声遮罩异常以提高视频生成的可靠性 + +## 改进的工作流可靠性 +- **更好的错误处理**: 增强的内存不足错误消息在工作流超出系统限制时提供更清晰的指导 +- **更长渲染支持**: 改进了超过10分钟渲染的日志记录，帮助追踪复杂、耗时工作流的进度 +- **PyTorch警告修复**: 通过解决兼容性警告实现更清洁的控制台输出 + +## 新的节点功能 +- **调整和填充图像节点**: 简化的图像预处理，为您的工作流提供一致的输入尺寸 +- **ImageStitch增强**: 白色填充选项提供更灵活的图像合成工具 +- **Singlestep DPM++ SDE**: 为Rectified Flow模型提供新的采样器选项，扩展您的采样工具包 + +## 开发者与系统改进 +- **前端更新**: 多项前端改进（v1.22.2、v1.23.3、v1.23.4）增强了UI响应性和工作流管理 +- **模板更新**: 最新模板版本（0.1.29）包含改进的自定义节点开发工具 +- **DirectML指南**: 更新的文档阐明了获得最佳性能的硬件建议 +- **Transformers兼容性**: 最低版本要求提升确保与最新模型实现的兼容性 + +## 技术细节优化 +- **术语更新**: "UNet"引用更新为"Diffusion Model"以提供更清晰的技术交流 +- **FP16兼容性**: 在较旧的PyTorch版本上禁用Omnigen2 FP16以防止兼容性问题 +- **依赖更新**: 依赖项更新确保与最新Python包的稳定运行 + +这些增强使ComfyUI对于使用高级图像编辑工作流的AI艺术家更加强大，同时保持了创作者在专业项目中依赖的稳定性和性能。 + + **高级采样和模型控制增强**