ComfyUI Node
⚙️ Advanced CFG Controller (Expert)
A ComfyUI node in model_patches/Advanced_CFG_Controller with 53 inputs and 2 outputs.
⚙️ Advanced CFG Controller (Expert)
- model
- attention_modifiers_positive
- attention_modifiers_negative
- attention_modifiers_fake_negative
- attention_modifiers_global
- MODEL
- STRING
◄adaptive_cfgHard Intensity►
◄adaptive_cfg_topk0.25►
◄adaptive_cfg_ref8.0►
◄skip_uncondtrue►
◄uncond_sigma_start1000.0►
◄uncond_sigma_end1.0►
◄fake_uncond_startfalse►
◄fake_uncond_expfalse►
◄fake_uncond_exp_methodUse Positive►
◄fake_uncond_exp_value2.0►
◄fake_uncond_exp_normalizefalse►
◄fake_uncond_multiplier1►
◄fake_uncond_sigma_start1000.0►
◄fake_uncond_sigma_end1.0►
◄lerp_uncondfalse►
◄lerp_uncond_strength2.0►
◄lerp_uncond_sigma_start1000.0►
◄lerp_uncond_sigma_end1.0►
◄subtract_latent_meanfalse►
◄subtract_latent_mean_sigma_start1000.0►
◄subtract_latent_mean_sigma_end1.0►
◄latent_intensity_rescalefalse►
◄latent_intensity_rescale_methodHard Measurement►
◄latent_intensity_rescale_cfg8.0►
◄latent_intensity_rescale_sigma_start1000.0►
◄latent_intensity_rescale_sigma_end3.0►
◄cond_expfalse►
◄cond_exp_normalizefalse►
◄cond_exp_sigma_start1000.0►
◄cond_exp_sigma_end1.0►
◄cond_exp_methodAmplify►
◄cond_exp_value2.0►
◄uncond_expfalse►
◄uncond_exp_normalizefalse►
◄uncond_exp_sigma_start1000.0►
◄uncond_exp_sigma_end1.0►
◄uncond_exp_methodAmplify►
◄uncond_exp_value2.0►
◄attention_modifiers_global_enabledfalse►
◄disable_condfalse►
◄disable_cond_sigma_start1000.0►
◄disable_cond_sigma_end0.0►
◄enable_performance_monitoringfalse►
◄adaptive_noise_strength0.00►
◄spectral_norm_target1.0►
◄frequency_filter_cutoff0.50►
◄dynamic_range_compression1.00►
◄args_filter—►
Categorymodel_patches/Advanced_CFG_Controller
Inputs (53)
| Name | Type | Default | Description |
|---|---|---|---|
| model | MODEL | — | |
| adaptive_cfg | COMBO | Hard Intensity | 6 options: Disabled, Soft Intensity, Hard Intensity, Hard Squared, Range Based, Spectral Analysis |
| adaptive_cfg_topk | FLOAT | 0.250–0.5 | Top-K percentage for intensity measurement (lower = more precise) |
| adaptive_cfg_ref | FLOAT | 8.00–100 | Reference CFG value for adaptive scaling |
| skip_uncond | BOOLEAN | true | Skip negative prompt calculation for speed boost |
| uncond_sigma_start | FLOAT | 1000.00–10000 | Start skipping negative at this sigma level |
| uncond_sigma_end | FLOAT | 1.00–10000 | Stop skipping negative at this sigma level |
| fake_uncond_start | BOOLEAN | false | Use fake negative prompt when skipping |
| fake_uncond_exp | BOOLEAN | false | Apply experimental modifications to fake negative |
| fake_uncond_exp_method | COMBO | Use Positive | Method for generating fake negative prompt |
| fake_uncond_exp_value | FLOAT | 2.00–1000 | Strength value for fake negative method |
| fake_uncond_exp_normalize | BOOLEAN | false | Normalize fake negative to preserve magnitude |
| fake_uncond_multiplier | INT | 1-1–1 | Multiplier for fake negative (-1, 0, or 1) |
| fake_uncond_sigma_start | FLOAT | 1000.00–10000 | Start using fake negative at this sigma |
| fake_uncond_sigma_end | FLOAT | 1.00–10000 | Stop using fake negative at this sigma |
| lerp_uncond | BOOLEAN | false | Blend positive and negative prompts |
| lerp_uncond_strength | FLOAT | 2.00–10 | Blending strength (higher = more negative influence) |
| lerp_uncond_sigma_start | FLOAT | 1000.00–10000 | Start blending at this sigma level |
| lerp_uncond_sigma_end | FLOAT | 1.00–10000 | Stop blending at this sigma level |
| subtract_latent_mean | BOOLEAN | false | Fix color drift by centering latent values |
| subtract_latent_mean_sigma_start | FLOAT | 1000.00–10000 | Start color correction at this sigma |
| subtract_latent_mean_sigma_end | FLOAT | 1.00–10000 | Stop color correction at this sigma |
| latent_intensity_rescale | BOOLEAN | false | Rescale latent intensity to prevent oversaturation |
| latent_intensity_rescale_method | COMBO | Hard Measurement | Method for measuring latent intensity |
| latent_intensity_rescale_cfg | FLOAT | 8.00–100 | Target intensity level for rescaling |
| latent_intensity_rescale_sigma_start | FLOAT | 1000.00–10000 | Start rescaling at this sigma |
| latent_intensity_rescale_sigma_end | FLOAT | 3.00–10000 | Stop rescaling at this sigma |
| cond_exp | BOOLEAN | false | Apply experimental modifications to positive prompt |
| cond_exp_normalize | BOOLEAN | false | Normalize positive prompt to preserve magnitude |
| cond_exp_sigma_start | FLOAT | 1000.00–10000 | Start modifications at this sigma |
| cond_exp_sigma_end | FLOAT | 1.00–10000 | Stop modifications at this sigma |
| cond_exp_method | COMBO | Amplify | Modification method for positive prompt |
| cond_exp_value | FLOAT | 2.00–100 | Strength value for positive modification |
| uncond_exp | BOOLEAN | false | Apply experimental modifications to negative prompt |
| uncond_exp_normalize | BOOLEAN | false | Normalize negative prompt to preserve magnitude |
| uncond_exp_sigma_start | FLOAT | 1000.00–10000 | Start modifications at this sigma |
| uncond_exp_sigma_end | FLOAT | 1.00–10000 | Stop modifications at this sigma |
| uncond_exp_method | COMBO | Amplify | Modification method for negative prompt |
| uncond_exp_value | FLOAT | 2.00–100 | Strength value for negative modification |
| attention_modifiers_global_enabled | BOOLEAN | false | Enable global attention modifications |
| disable_cond | BOOLEAN | false | Disable positive prompt in specified range |
| disable_cond_sigma_start | FLOAT | 1000.00–10000 | Start disabling positive at this sigma |
| disable_cond_sigma_end | FLOAT | 0.00–10000 | Stop disabling positive at this sigma |
| enable_performance_monitoring | BOOLEAN | false | Track performance metrics during generation |
| adaptive_noise_strength | FLOAT | 0.000–1 | Strength of adaptive noise injection |
| spectral_norm_target | FLOAT | 1.00.1–5 | Target norm for spectral normalization |
| frequency_filter_cutoff | FLOAT | 0.500.1–1 | Cutoff frequency for frequency domain filtering |
| dynamic_range_compression | FLOAT | 1.000.1–1 | Dynamic range compression ratio (1.0 = disabled) |
| args_filteropt | STRING | Comma-separated list of parameters to include in output | |
| attention_modifiers_positiveopt | ATTNMOD | Attention modifications for positive prompt | |
| attention_modifiers_negativeopt | ATTNMOD | Attention modifications for negative prompt | |
| attention_modifiers_fake_negativeopt | ATTNMOD | Attention modifications for fake negative | |
| attention_modifiers_globalopt | ATTNMOD | Global attention modifications |
Outputs (2)
| Name | Type | Description |
|---|---|---|
| MODEL | MODEL | — |
| STRING | STRING | — |