PreSampling (Advanced)¶
Documentation¶
- Class name:
easy preSamplingAdvanced - Category:
EasyUse/PreSampling - Output node:
True
This node is designed to configure advanced pre-sampling settings for image generation processes. It allows for detailed customization of the sampling process, enabling users to fine-tune parameters such as noise levels, scheduler types, denoise settings, and seed values to achieve desired image qualities and variations. The node serves as a crucial component for preparing the image generation pipeline, offering flexibility and control over the pre-sampling phase.
Input types¶
Required¶
pipe- The pipeline object that is being configured for the pre-sampling process. It is essential for defining the context and settings for the image generation workflow.
- Comfy dtype:
PIPE_LINE - Python dtype:
dict
steps- Specifies the number of steps to be executed in the pre-sampling process. This parameter directly influences the granularity and quality of the generated images.
- Comfy dtype:
INT - Python dtype:
int
cfg- Configures the conditioning factor for the generation process, affecting the influence of the conditioning on the generated images.
- Comfy dtype:
FLOAT - Python dtype:
float
sampler_name- Determines the sampling algorithm to be used, allowing for selection from a variety of predefined samplers to optimize the generation process.
- Comfy dtype:
COMBO[STRING] - Python dtype:
str
scheduler- Specifies the scheduler algorithm to be used during the pre-sampling process. This parameter allows for the selection from a predefined list of schedulers, including custom ones, to control the sampling behavior and efficiency.
- Comfy dtype:
COMBO[STRING] - Python dtype:
str
start_at_step- Defines the starting step for the pre-sampling process, allowing for control over the initial state of image generation.
- Comfy dtype:
INT - Python dtype:
int
end_at_step- Sets the ending step for the pre-sampling process, determining the final state of the generated images.
- Comfy dtype:
INT - Python dtype:
int
add_noise- Enables or disables the addition of noise during the pre-sampling process, affecting the texture and details of the generated images.
- Comfy dtype:
COMBO[STRING] - Python dtype:
bool
seed- Sets the seed value for the random number generator used in the pre-sampling process. This ensures reproducibility of results by initializing the generator to a known state.
- Comfy dtype:
INT - Python dtype:
int
return_with_leftover_noise- Determines whether the generated images should retain some level of noise, offering a balance between clarity and naturalness.
- Comfy dtype:
COMBO[STRING] - Python dtype:
bool
Optional¶
image_to_latent- Optional input for providing an image to be converted into a latent representation, enabling direct manipulation of the latent space.
- Comfy dtype:
IMAGE - Python dtype:
object
latent- Optional input for providing a latent representation, offering an advanced level of customization by directly manipulating the latent space prior to sampling.
- Comfy dtype:
LATENT - Python dtype:
object
Output types¶
pipe- Comfy dtype:
PIPE_LINE - Outputs a configured pipeline object, encapsulating all the specified pre-sampling settings and ready for integration into the image generation workflow.
- Python dtype:
dict
- Comfy dtype:
Usage tips¶
- Infra type:
CPU - Common nodes: unknown
Source code¶
class samplerSettingsAdvanced:
def __init__(self):
pass
@classmethod
def INPUT_TYPES(cls):
return {"required":
{"pipe": ("PIPE_LINE",),
"steps": ("INT", {"default": 20, "min": 1, "max": 10000}),
"cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0}),
"sampler_name": (comfy.samplers.KSampler.SAMPLERS,),
"scheduler": (comfy.samplers.KSampler.SCHEDULERS + new_schedulers,),
"start_at_step": ("INT", {"default": 0, "min": 0, "max": 10000}),
"end_at_step": ("INT", {"default": 10000, "min": 0, "max": 10000}),
"add_noise": (["enable", "disable"],),
"seed": ("INT", {"default": 0, "min": 0, "max": MAX_SEED_NUM}),
"return_with_leftover_noise": (["disable", "enable"], ),
},
"optional": {
"image_to_latent": ("IMAGE",),
"latent": ("LATENT",)
},
"hidden":
{"prompt": "PROMPT", "extra_pnginfo": "EXTRA_PNGINFO", "my_unique_id": "UNIQUE_ID"},
}
RETURN_TYPES = ("PIPE_LINE", )
RETURN_NAMES = ("pipe",)
OUTPUT_NODE = True
FUNCTION = "settings"
CATEGORY = "EasyUse/PreSampling"
def settings(self, pipe, steps, cfg, sampler_name, scheduler, start_at_step, end_at_step, add_noise, seed, return_with_leftover_noise, image_to_latent=None, latent=None, prompt=None, extra_pnginfo=None, my_unique_id=None):
# 图生图转换
vae = pipe["vae"]
batch_size = pipe["loader_settings"]["batch_size"] if "batch_size" in pipe["loader_settings"] else 1
if image_to_latent is not None:
_, height, width, _ = image_to_latent.shape
if height == 1 and width == 1:
samples = pipe["samples"]
images = pipe["images"]
else:
samples = {"samples": vae.encode(image_to_latent[:, :, :, :3])}
samples = RepeatLatentBatch().repeat(samples, batch_size)[0]
images = image_to_latent
elif latent is not None:
samples = latent
images = pipe["images"]
else:
samples = pipe["samples"]
images = pipe["images"]
force_full_denoise = True
if return_with_leftover_noise == "enable":
force_full_denoise = False
new_pipe = {
"model": pipe['model'],
"positive": pipe['positive'],
"negative": pipe['negative'],
"vae": pipe['vae'],
"clip": pipe['clip'],
"samples": samples,
"images": images,
"seed": seed,
"loader_settings": {
**pipe["loader_settings"],
"steps": steps,
"cfg": cfg,
"sampler_name": sampler_name,
"scheduler": scheduler,
"start_step": start_at_step,
"last_step": end_at_step,
"denoise": 1.0,
"add_noise": add_noise,
"force_full_denoise": force_full_denoise
}
}
del pipe
return {"ui": {"value": [seed]}, "result": (new_pipe,)}