xinyun/ComfyUI
1
0
Fork 0
mirror of https://git.datalinker.icu/comfyanonymous/ComfyUI synced 2026-09-03 01:57:05 +08:00
Code Issues Packages Projects Releases Wiki Activity

Support Kohya-ss structure

Browse Source
This commit is contained in:
Kohaku-Blueleaf 2025-07-07 16:34:17 +08:00
parent 3e3adfe804
commit 8d87ec78ec
1 changed files with 21 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
comfy_extras/nodes_train.py
View File

@ -255,13 +255,23 @@ class LoadImageTextSetFromFolderNode:
sub_input_dir = os.path.join(folder_paths.get_input_directory(), folder) sub_input_dir = os.path.join(folder_paths.get_input_directory(), folder)
valid_extensions = [".png", ".jpg", ".jpeg", ".webp"] valid_extensions = [".png", ".jpg", ".jpeg", ".webp"]
image_files = [
f image_files = []
for f in os.listdir(sub_input_dir) for item in os.listdir(sub_input_dir):
if any(f.lower().endswith(ext) for ext in valid_extensions) path = os.path.join(sub_input_dir, item)
] if any(item.lower().endswith(ext) for ext in valid_extensions):
image_files.append(path)
elif os.path.isdir(path):
# Support kohya-ss/sd-scripts folder structure
repeat = 1
if item.split("_")[0].isdigit():
repeat = int(item.split("_")[0])
image_files.extend([
os.path.join(path, f) for f in os.listdir(path) if any(f.lower().endswith(ext) for ext in valid_extensions)
] * repeat)
caption_file_path = [ caption_file_path = [
os.path.basename(f).replace(os.path.splitext(f)[1], ".txt") f.replace(os.path.splitext(f)[1], ".txt")
for f in image_files for f in image_files
] ]
captions = [] captions = []
@ -280,10 +290,15 @@ class LoadImageTextSetFromFolderNode:
logging.info(f"Loaded {len(output_tensor)} images from {sub_input_dir}.") logging.info(f"Loaded {len(output_tensor)} images from {sub_input_dir}.")
logging.info(f"Encoding captions from {sub_input_dir}.")
conditions = [] conditions = []
empty_cond = clip.encode_from_tokens_scheduled(clip.tokenize(""))
for text in captions: for text in captions:
if text == "":
conditions.append(empty_cond)
tokens = clip.tokenize(text) tokens = clip.tokenize(text)
conditions.extend(clip.encode_from_tokens_scheduled(tokens)) conditions.extend(clip.encode_from_tokens_scheduled(tokens))
logging.info(f"Encoded {len(conditions)} captions from {sub_input_dir}.")
return (output_tensor, conditions) return (output_tensor, conditions)