Adds selected_seed outputs for CADE Easy and MG_SuperSimple while preserving existing output slot order.
Adds regression coverage for output contracts.
Set MSVC build tool env vars for Windows builds
DZRobo
7533dde
Adds environment variables DISTUTILS_USE_SDK and MSSdk when running on Windows to ensure distutils reuses existing MSVC build tools. This helps with compiler detection in plain shells and improves reliability of source builds.
Skip NAG when context key dim mismatches layer
DZRobo
cdc8e19
Adds a check to bypass NAG cross-attention if the context's last dimension does not match the expected key dimension for the layer, preventing matmul crashes with mismatched context inputs.
Update README
DZRobo
e06d499
Update README
Add Z_image support and Improve latent/channel handling
DZRobo
e69f3b7
Adds functions to harmonize latent channel counts and condition token lengths to prevent mismatches, especially for models like FLUX/Z_image. Enhances error reporting with debug output and traceback printing. Updates mg_combinode to better validate VAE/CLIP presence for checkpoint and input selection. Fixes hybrid sigma schedule alignment in mg_zesmart_sampler_v1_1.
Added mg_i2i_Easy-Workflow.json to provide a new ComfyUI workflow for image-to-image processing with MagicNodes CADE 2.5 integration.
Refined multiple parameters in mg_cade25.cfg for improved image generation, including steps, denoise, scaling, thresholds, and scheduler settings across all steps.
Tweaked some param for more consistency
DZRobo
57e5af5
Tweaked some param for more consistency
Add smart seed diversity and config options
DZRobo
0f11662
Introduces 'smart_seed_k', 'smart_seed_steps', and 'smart_seed_diversity' parameters to the ComfyAdaptiveDetailEnhancer25 class and updates the smart seed selection logic to support diversity-based scoring. Updates mg_cade25.cfg to include these new smart seed options for each step. Also improves LoRA slot gating in mg_combinode.py to ignore empty, None, or disabled slots more strictly.
cfg update
DZRobo
7ca7bce
cfg tweaking
Upd cfg, micro fix
DZRobo
aa11bf4
Upd cfg, micro fix
Add NAG fallback and fp32 decode options, update presets
DZRobo
0503191
Introduces NAG fallback logic to both easy and hard modules for improved guidance when CrossAttention patch is inactive. Adds an option to force VAE decode output to fp32 in the easy module. Updates mg_cade25.cfg with new default values for steps, cfg, denoise, buffer, guidance, and other parameters to improve quality and stability.
Add CWN and AGC guidance options to CADE25 nodes
DZRobo
e42ce45
Introduces Conditioning Weight Normalization (CWN) and Adaptive Guidance Clipping (AGC) to both easy and hard CADE25 modules, with configurable parameters exposed in the code and presets. Updates mg_cade25.cfg to include default values for these new options and adjusts some scheduler and denoise settings for improved stability and control.
Improve depth model auto-detection and fallback logic
DZRobo
42c372b
Enhanced depth model path resolution to scan additional directories and prefer local weights, including comfyui_controlnet_aux locations. Updated fallback to use a constant mid-gray tensor for pseudo-depth. Added warnings when no local weights are found and improved robustness of path detection. Fix memory leaking.
Tweaking param
DZRobo
5b20491
Tweaking param
Update README
DZRobo
c5117cb
Update README
per‑step CFG scheduling
DZRobo
c86b5e0
Introduces per‑step CFG schedules (cosine, warmup, U‑shape) driven by log‑sigma progress; schedules stack with existing cfg_curve.
Presset finetune.
Add configurable CFG scheduling to guidance wrapper
DZRobo
7dd5fc1
Introduces per-step CFG scheduling options (cosine, warmup, U-shape) to _wrap_model_with_guidance and exposes related parameters in ComfyAdaptiveDetailEnhancer25. Updates mg_cade25.cfg presets to utilize new scheduling controls and adjusts detail and denoise settings for improved flexibility.
Add latent buffer stabilization to enhancer
DZRobo
df38cb0
Introduces a latent buffer mechanism to ComfyAdaptiveDetailEnhancer25 for improved multi-iteration stability, including EMA, anchor rebasing, and masked injection. Updates presets to enable and tune latent buffer parameters, and adjusts detail and mid-frequency settings for better results. Minor README clarification and workflow file update.
param finetune
DZRobo
aeb570c
param finetune
finetune gen param
DZRobo
4eca43b
finetune gen param, more smooth flow an quality up
Tune presets and workflow parameters for MG pipeline
DZRobo
b28e4ed
Adjusted multiple parameters in mg_cade25.cfg and mg_controlfusion.cfg for improved guidance, blending, and stabilization. Updated workflow JSON to reflect new step settings and parameter values for better image quality and control. These changes aim to optimize generation results and enhance overall pipeline performance.
Speed optimisation
DZRobo
4bfcccd
Speed optimisation for CADE_Easy and SuperSImple workflow.
corrected a typo in Readme
DZRobo
5595de7
corrected a typo in Readme
Add preview downscale and mid-frequency stabilizer
DZRobo
92ca926
Introduces a 'preview_downscale' option to cap output image size for reduced RAM usage in mg_cade25_easy.py. Adds mid-frequency stabilizer controls to mg_cade25.py for improved hand/object stability at high resolution. Updates README with new module notes and workflow documentation, and revises mg_Easy-Workflow.json for node order and structure.
update README
DZRobo
06ac6de
update README
Optimize memory usage
DZRobo
bb62070
Improves memory management by releasing temporary tensors throughout the CADE2.5 easy and hard modules, reducing RAM/VRAM peaks. Adds optional auto-save functionality for final images with configurable compression. Updates README with memory tips and clarifies workflow steps. Adjusts presets for sharpen, guidance, NAG, and mid-frequency stabilizer parameters to further optimize performance and output quality.
Adjust denoise parameter in mg_cade25.cfg
DZRobo
3d335ed
Reduced the denoise value from 0.60 to 0.45 to refine output quality in the mg_cade25 preset configuration.
Update README
DZRobo
9e3dffe
Update README
Add MG_CleanUp node and improve memory management
DZRobo
9068525
Introduces the MG_CleanUp node for aggressive RAM/VRAM cleanup and system cache trimming. Improves safe_encode and safe_decode in mg_cade25_easy.py to better release VRAM and avoid autograd graphs. Updates MG_ControlFusion nodes to clear caches and intermediates after execution. Adjusts blend and edge parameters in mg_controlfusion.cfg for better mask blending. Updates workflow to integrate new cleanup and parameter changes.
Increase denoise value in mg_cade25 preset
DZRobo
55ddc32
Adjusted the denoise parameter from 0.45 to 0.60 in the mg_cade25.cfg preset to modify image generation behavior.
Bugfixes (#11)
DZRobo
Unverified signature
478379a
* Add MagicLatentAdapter node
Introduces MagicLatentAdapter for generating or adapting latents to match model expectations, including support for 5D (NCDHW) layouts and experimental FLUX/Qwen models. Adds mask alignment utilities to CADE nodes for robust mask blending, improves Gaussian blur to support 5D tensors, and updates preset defaults for improved results. Documentation and workflow updated to reflect new node and features.
* Bugfixes
1. Adjust edge mask gain logic and preset parameters. Fix depth path for CF (hard).
2. Memory leaking fix.
3. Quality improvement at all steps (Easy and SuperSimple)
Add MagicLatentAdapter node (experimental support FLUX/Qwen models) (#10)
DZRobo
Unverified signature
ccadebd
Introduces MagicLatentAdapter (experimental support FLUX/Qwen models) for generating or adapting latents to match model expectations. Adds mask alignment utilities to CADE nodes for robust mask blending, improves Gaussian blur to support 5D tensors, and updates preset defaults for improved results. Documentation and workflow updated to reflect new node and features.
Update tested RAM specification in README
DZRobo
6cce3ef
Changed the tested RAM value from 92Gb to 128Gb in the hardware section to reflect updated hardware requirements.
Update hardware requirements section in README
DZRobo
bcae778
Revised the hardware requirements section to provide more practical guidance, including tested hardware configurations and recommendations for managing starting latency to avoid errors during upscaling.
Remove bundled negative LoRA model and update README (#7)
DZRobo
Unverified signature
8ca0a45
Deleted mg_7lambda_negative.safetensors from the repository and updated the README to reference downloading the model from Hugging Face instead. Cleaned up related documentation and directory structure notes to reflect this change.
BUG FIX and QoL. Fix cancel support to MG nodes (#6)
DZRobo
Unverified signature
dc64aaa
Implements cooperative cancellation and user interrupt propagation in mg_cade25_easy.py, mg_supersimple_easy.py, mg_cade25.py, and related modules. Adds interruptible sampler and callback wrappers, sentinel handling, and frequent cancel checks throughout long-running loops and workflows. Also improves LoRA selection robustness in mg_combinode.py by handling missing files gracefully. Updates workflow JSON to match new node logic.
Update SageAttention installation note in README
DZRobo
b163497
Added clarification that SageAttention v2.2.0 is required for the pipeline to work, and that version 1.0.6 is not suitable.
Update comments and documentation for clarity
DZRobo
8e7e41a
Removed outdated reference to arXiv draft in CADE25 documentation and updated a comment in check_sageattention.ps1 from Russian to English for better clarity.
Update README image layout and funding config
DZRobo
bf37973
Replaced HTML-based image sections in README.md with Markdown tables for improved formatting and consistency. Removed custom funding links from .github/FUNDING.yml to streamline funding configuration.
Add step to inject HF metadata into README
DZRobo
b72e135
Introduces a workflow step that prepends Hugging Face metadata to the README file before pushing to the HF mirror. This ensures the repository contains required metadata for Hugging Face model repositories.
Track more image formats with Git LFS
DZRobo
714ae4b
Added additional image extensions to .gitattributes for LFS tracking and updated the hf-mirror workflow to rewrite history for all image formats, ensuring no plain binaries remain in the mirror.
Track image assets via LFS for HF mirror
DZRobo
a53096e
Adds a workflow step to track .jpg, .jpeg, and .png files with Git LFS and rewrites history for Hugging Face mirror compliance. This ensures binary assets are stored as LFS pointers in the HF mirror while preserving GitHub history.
Add Git LFS support to HF mirror workflow
DZRobo
e109c36
Enabled Git LFS in the checkout step and added steps to ensure LFS objects are present before pushing to the Hugging Face mirror. This ensures large files tracked by LFS are properly handled during mirroring.
Initial project structure and core files added (#1)
DZRobo
Unverified signature
695fbf0
Add MagicNodes repository with documentation, licensing, assets, workflows, presets, and main Python modules for Easy/Hard nodes, CADE 2.5, ControlFusion, QSilk, and supporting utilities. Includes sample images, negative LoRA, Depth Anything v2 vendor code, and setup scripts for ComfyUI integration.
