ComfyUI-ClybsChromaNodes

ChromaNAG

A model patch for Normalized Attention Guidance on Chroma's double blocks. It derives context from the supplied conditioning, patches each double-block forward method in a cloned model, and returns the patched MODEL; when nag_scale is 0, it returns the input model unchanged. The patch exposes nag_scale, nag_alpha, and nag_tau controls.

Advanced/Model Patches · entradas 2 · parâmetros 3 · saída 1

ClybAdaptiveLoraLoader

Apply multiple LoRAs adaptively by merging patches into a single model clone.

Loaders · saídas 2

ClybAdaptiveLoraLoaderModelOnly

File: clybModelLoader.py Category: loaders Same as ClybAdaptiveLoraLoader, but the CLIP input is omitted from the schema and the internal strengthclip is forced to 0.0 for every slot, leaving only the diffusion-model patches applied. Use this for MODEL-only pipelines (e.g. unconditional sampling, flows without a text encoder, or cases where CLIP is wired in separately). | Input | Type | Default | Range | Description | |---|---|---|---|---| | model | MODEL | — | — | Diffusion model to patch | | loraname1 | ENUM | — | loras list | First LoRA (set to "none" to skip) | | strengthmodel1 | FLOAT | 1.0 | -100, 100 | Diffusion-model strength (negative allowed) | | loraname2..20 | ENUM | "none" | loras list | Additional LoRAs (revealed as you fill slots) | | strengthmodel2..20 | FLOAT | 1.0 | -100, 100 | Per-slot diffusion-model strength | Returns: MODEL.

saída 1

ClybGuidance

File: clybGuidance.py Category: sampling/customsampling A pre-CFG model patch that rewires how the conditional and unconditional predictions are combined at every sampling step. Stacks the following features on top of standard CFG: - Project-and-scale (eta) — the guidance vector is split into a component parallel to the conditional and a component orthogonal to it. The parallel component is scaled by eta, the orthogonal component is left alone. eta = 1.0 recovers default CFG. - Norm clamping (normthreshold) — if the L2 norm of the guided output exceeds the conditional's norm times normthreshold, the guided output is rescaled back down. Disabled at 0.0. - Momentum (momentum, momentumbeta, momentumrenorm) — adds a fraction of a running-average guidance vector to the current guidance, optionally re-normalized back to its original norm. momentum = 0 disables it. - Scalar projection (scalarprojection, scalarlogsumexp) — projects the conditional onto the unconditional as a scalar (logsumexp or sum reduction), then scales the unconditional by that scalar. - STD/var rescale (rescalephi, varrescale) — blends the guided output toward an output whose standard deviation (or variance) matches the conditional's. rescalephi = 0 disables it. - Sine-bell schedule (scaleupratio, scaleupshift) — animates the effective CFG scale from 1.0 at the start, up to the configured CFG scale at the middle of diffusion, and back down to 1.0 at the end. scaleupratio = 0 disables it. scaleupshift 1.0 earlier. - atan2/sin blend (atan2sinratio) — blends the unconditional with uncond.atan().sin() / cond.atan…

Sampling/Custom Sampling · entrada 1 · parâmetros 12 · saída 1

InverseSquaredScheduler

File: clybSchedulers.py Category: sampling/customsampling/schedulers A sigma scheduler that biases the schedule toward the end of diffusion. It uses (1 - t²)² (i.e. the inverse of t mapped through (1-t)²) to pick sigma indices — fine-grained near the end, coarser at the start. The scheduler is also registered into SCHEDULERHANDLERS under the name inversesquared, so it can be used as a string in any node that accepts a scheduler name. | Input | Type | Default | Range | Description | |---|---|---|---|---| | model | MODEL | — | — | Model to derive the sigma range from | | steps | INT | 20 | 3, 1000 | Number of steps | | denoise | FLOAT | 1.0 | 0, 1 | Denoise strength (< 1.0 enables img2img-style short schedules) | Returns: SIGMAS.

Sampling/Custom Sampling/Schedulers · entrada 1 · parâmetros 2 · saída 1

PrintSigmas

File: clybSchedulers.py Category: sampling/customsampling/schedulers A debug helper that prints the incoming SIGMAS tensor to the console and passes it through unchanged. Useful for inspecting schedules from other nodes without modifying them. | Input | Type | Description | |---|---|---| | sigmas | SIGMAS | Sigma tensor to print | Returns: SIGMAS (passthrough).

Sampling/Custom Sampling/Schedulers · entrada 1 · parâmetros 0 · saída 1

SamplerClyb_BDF

File: clybSamplers.py Category: sampling/customsampling/samplers A backward-differentiation-formula-style sampler that takes a single model evaluation at the start of the step, then synthesizes a refined denoised prediction at the sigmadown point and combines them with one of three scalar fusions: - projection — projects the half-step prediction back onto the line spanned by the full-step prediction. - atan2sin — uses atan2(sin(half), cos(full)) to blend the two predictions in angle space. - atan2sin+projection (default) — the atan2/sin blend followed by a projection rescaling, combining both stabilizations. The sampler detects flow-matching models (FLUX, SD3, Chroma) automatically and switches to the flow-style ancestral update with alphaip1/alphadown/renoisecoeff. | Input | Type | Default | Range | Description | |---|---|---|---|---| | scalar | ENUM | atan2sin+projection | projection, atan2sin, atan2sin+projection | Scalar fusion mode | | eta | FLOAT | 1.0 | 0, 100 | Ancestral stochasticity | | snoise | FLOAT | 1.0 | 0, 100 | Noise scaling factor |

Sampling/Custom Sampling/Samplers · entrada 1 · parâmetros 2 · saída 1

SamplerClyb_GeomExtrap

Geometric-Midpoint Extrapolation sampler. 3-NFE per step sampler that uses two geometric midpoints between sigma_cur and sigma_down to build a 3-point quadratic extrapolation of the denoised prediction at sigma_down, then integrates that prediction into the noisy x latent. Compatible with both flow-matching (FLUX, SD3, Chroma) and non-flow models. Parameters: - eta: Ancestral sampling stochasticity (0=deterministic, 1=full stochastic) - s_noise: Noise scaling factor - sigma_calc: Ancestral sigma calculation method (clyb, taylor-expansion, ancestral, adaptive)

Sampling/Custom Sampling/Samplers · entradas 0 · parâmetros 3 · saída 1

SamplerTaylorFlow

File: clybSamplers.py Category: sampling/customsampling/samplers A multi-step Taylor-expansion sampler (registered as taylorflow in the ComfyUI sampler list). Implements the algorithm from "Leveraging Previous Steps: A Training-free Fast Solver for Flow Diffusion" (Nov 2024): 1. Maintain a rolling history of (sigma, denoised) pairs from the previous order steps. 2. At each step, perform one model evaluation at the current state. 3. Build a Vandermonde matrix from the historical sigma values. 4. Solve for Taylor coefficients B that predict the latent at sigmanext. 5. Apply the Euler step plus a correction term built from the history. 6. Inject ancestral noise as usual. The four sigmacalc modes control how the sigmadown / sigmaup pair is computed for the ancestral noise injection: - clyb (default) — logarithmic scaling, the original Clyb scheme. - taylor-expansion — exponential factor with a quadratic correction based on the step ratio. - ancestral — standard k-diffusion getancestralstep. - adaptive — converges toward the standard scheme based on a normalized variance of the denoised history (small history variance ⇒ less noise). The Vandermonde solve supports two methods internally (iterative two-sided equilibration with Tikhonov regularization, and diagonal-dominant extraction) and falls back to a lstsq solve if the regularized system is singular. | Input | Type | Default | Range | Description | |---|---|---|---|---| | order | INT | 8 | 1, 16 | Taylor expansion order (history length) | | eta | FLOAT | 1.0 | 0, 1 | Ancestral stochasticity | | snoise | FLOAT | 1.0 | 0, 2 | N…

Sampling/Custom Sampling/Samplers · entradas 0 · parâmetros 4 · saída 1

SamplerWrapperCFGPP

File: clybSamplers.py Category: sampling/customsampling/samplers A sampler wrapper — takes any other SAMPLER as input and returns a new sampler that runs the inner sampler but with a CFG++-style denoised recomputation. CFG++ replaces the standard CFG blend with a closed-form denoised that uses the unconditional prediction from the next sigma: denoisedstar = (sigma alphat denoisedguided - sigmanext alphas unconddenoised) / (sigma - sigmanext) where alphas = sigma exp(lambda(sigma)) and alphat = sigmanext exp(lambda(sigmanext)), with lambda(s) = sigmatohalflogsnr(s, modelsampling). The wrapper installs a postcfgfunction hook on the model to capture unconddenoised from the inner sampler, then uses a CFGPPProxyModel to perform the recombination at every step. The wrapper itself is not added to the standard KSampler dropdown — it is only reachable through this node (or any node that constructs it via comfy.samplers.ksampler("cfgpp", {...})). | Input | Type | Description | |---|---|---| | sampler | SAMPLER | Inner sampler to wrap with CFG++ |

Sampling/Custom Sampling/Samplers · entrada 1 · parâmetros 0 · saída 1