nodes/filliptm-ComfyUI_Fill-Nodes
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

added qwen image node F.U Spence lol.

Browse Source
This commit is contained in:
Fill
2025-10-12 22:00:13 -07:00
parent 2682ed26db
commit ff679e58d0
3 changed files with 110 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
__init__.py
+3
View File
@@ -207,6 +207,7 @@ from .nodes.wip.FL_KsamplerFractals import FL_FractalKSampler
from .nodes.wip.FL_TimeLine import FL_TimeLine
from .nodes.wip.FL_WF_Agent import FL_WF_Agent
from .nodes.wip.FL_WanFirstLastFrameToVideo import FL_WanFirstLastFrameToVideo
from .nodes.wip.FL_QwenImageEditStrength import FL_QwenImageEditStrength
NODE_CLASS_MAPPINGS = {
"FL_SaveWebM": FL_SaveWebM,
@@ -381,6 +382,7 @@ NODE_CLASS_MAPPINGS = {
"FL_Audio_Segment_Extractor": FL_Audio_Segment_Extractor,
"FL_Audio_Separation": FL_Audio_Separation,
"FL_Audio_Shot_Iterator": FL_Audio_Shot_Iterator,
"FL_QwenImageEditStrength": FL_QwenImageEditStrength,
}
NODE_DISPLAY_NAME_MAPPINGS = {
@@ -556,6 +558,7 @@ NODE_DISPLAY_NAME_MAPPINGS = {
"FL_Audio_Segment_Extractor": "FL Audio Segment Extractor",
"FL_Audio_Separation": "FL Audio Separation",
"FL_Audio_Shot_Iterator": "FL Audio Shot Iterator",
"FL_QwenImageEditStrength": "FL Qwen Image Edit with Strength",
}
nodes/wip/FL_QwenImageEditStrength.py
+106
View File
@@ -0,0 +1,106 @@
"""Qwen Image Edit with Per-Image Strength Control"""
import node_helpers
import comfy.utils
import math
import torch
class FL_QwenImageEditStrength:
"""
Enhanced version of TextEncodeQwenImageEditPlus that allows controlling
the strength/weight of each individual image in the conditioning.
"""
@classmethod
def INPUT_TYPES(s):
return {
"required": {
"clip": ("CLIP",),
"prompt": ("STRING", {"multiline": True, "dynamicPrompts": True}),
},
"optional": {
"vae": ("VAE",),
"image1": ("IMAGE",),
"image1_strength": ("FLOAT", {"default": 1.0, "min": -10.0, "max": 10.0, "step": 0.01}),
"image2": ("IMAGE",),
"image2_strength": ("FLOAT", {"default": 1.0, "min": -10.0, "max": 10.0, "step": 0.01}),
"image3": ("IMAGE",),
"image3_strength": ("FLOAT", {"default": 1.0, "min": -10.0, "max": 10.0, "step": 0.01}),
}
}
RETURN_TYPES = ("CONDITIONING",)
FUNCTION = "encode_with_strength"
CATEGORY = "🏵️Fill Nodes/WIP"
def encode_with_strength(self, clip, prompt, vae=None,
image1=None, image1_strength=1.0,
image2=None, image2_strength=1.0,
image3=None, image3_strength=1.0):
ref_latents = []
ref_strengths = []
images = [
(image1, image1_strength),
(image2, image2_strength),
(image3, image3_strength)
]
images_vl = []
llama_template = "<|im_start|>system\nDescribe the key features of the input image (color, shape, size, texture, objects, background), then explain how the user's text instruction should alter or modify the image. Generate a new image that meets the user's requirements while maintaining consistency with the original input where appropriate.<|im_end|>\n<|im_start|>user\n{}<|im_end|>\n<|im_start|>assistant\n"
image_prompt = ""
for i, (image, strength) in enumerate(images):
if image is not None:
samples = image.movedim(-1, 1)
# Resize for vision model (384x384)
total = int(384 * 384)
scale_by = math.sqrt(total / (samples.shape[3] * samples.shape[2]))
width = round(samples.shape[3] * scale_by)
height = round(samples.shape[2] * scale_by)
s = comfy.utils.common_upscale(samples, width, height, "area", "disabled")
images_vl.append(s.movedim(1, -1))
# Encode to latents if VAE is provided
if vae is not None:
total = int(1024 * 1024)
scale_by = math.sqrt(total / (samples.shape[3] * samples.shape[2]))
width = round(samples.shape[3] * scale_by / 8.0) * 8
height = round(samples.shape[2] * scale_by / 8.0) * 8
s = comfy.utils.common_upscale(samples, width, height, "area", "disabled")
latent = vae.encode(s.movedim(1, -1)[:, :, :, :3])
# Apply strength by scaling the latent
if strength != 1.0:
latent = latent * strength
ref_latents.append(latent)
ref_strengths.append(strength)
image_prompt += "Picture {}: <|vision_start|><|image_pad|><|vision_end|>".format(i + 1)
# Tokenize and encode
tokens = clip.tokenize(image_prompt + prompt, images=images_vl, llama_template=llama_template)
conditioning = clip.encode_from_tokens_scheduled(tokens)
# Add reference latents with applied strengths
if len(ref_latents) > 0:
conditioning = node_helpers.conditioning_set_values(
conditioning,
{"reference_latents": ref_latents},
append=True
)
return (conditioning,)
NODE_CLASS_MAPPINGS = {
"FL_QwenImageEditStrength": FL_QwenImageEditStrength,
}
NODE_DISPLAY_NAME_MAPPINGS = {
"FL_QwenImageEditStrength": "Qwen Image Edit with Strength 🏵️",
}
pyproject.toml
+1 -1
View File
@@ -1,7 +1,7 @@
[project]
name = "comfyui_fill-nodes"
description = "Fill-Nodes is a versatile collection of custom nodes for ComfyUI that extends functionality across multiple domains. Features include advanced image processing (pixelation, slicing, masking), visual effects generation (glitch, halftone, pixel art), comprehensive file handling (PDF creation/extraction, Google Drive integration), AI model interfaces (GPT, DALL-E, Hugging Face), utility nodes for workflow enhancement, and specialized tools for video processing, captioning, and batch operations. The pack provides both practical workflow solutions and creative tools within a unified node collection."
version = "1.9.7"
version = "1.9.8"
license = "LICENSE"
dependencies = ["diffusers", "librosa", "sounddevice", "glitch_this", "PyOpenGL", "glfw", "scipy>=1.13.1", "requests", "aiohttp", "moviepy", "matplotlib", "reportlab", "openai", "PyPDF2", "pdf2image", "PyMuPDF", "reportlab", "PyPDF2", "ollama", "kornia", "opencv-python", "gdown", "open_clip_torch", "google-genai"]