digital-garbageandClaude Opus 5
74671329f9
[Log] Info: FunPack Diffusion Model Loader | flux.safetensors loaded, attention sage
[Log] Alert: MiniMaxH3ImageToVideo node | no "first_image" input is connected
[Log] Warning: FunPack LoRA Loader | style.safetensors is empty and will not load
[Log] Critical: torch | cannot allocate 14 GB (3 GB free); the run will not start
The difference between the middle two is the one that matters. Alert is a
choice you may not have made on purpose. Warning is something that will NOT
do what its name says -- the level v4 needed and did not have, which is how
560 handlers came to swallow a failure and carry on, so a feature could be
switched on, do nothing, and never say so.
Info is not decoration either: "loaded on GPU using sage attention" is how
an attention setting that did nothing gets noticed months before the
pictures look wrong enough to investigate.
Every call site moved -- there is no note() left to be vague through -- and
the loaders now say what they actually loaded rather than only what they
were asked for. Records are kept in a bounded ring as well as printed, so a
long rental session cannot turn the log into a memory leak, and served at
/funpack/api/log so a person can read it somewhere other than a terminal
full of ComfyUI's own output. There is a page for it.
Also dropped the preflight this was traded for: min and max are already
declared and already enforced at both doors, and a second protection layer
around a check that exists is ceremony.
Two mistakes worth keeping: a log line reached into the model's shape to
describe it and could have failed the load it was only narrating; and both
new pages passed `children` to an element that takes `rows`, which the kit
accepted and silently dropped. It throws on an unknown element and not on an
unknown prop.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
66 lines
3.1 KiB
Python
66 lines
3.1 KiB
Python
"""The LoRA loader.
|
|
|
|
One LoRA per node, chained for more. v4 carried N LoRAs in a single node behind a
|
|
JSON list widget and a `FlexibleOptionalInputType` -- a dict subclass that lied to
|
|
ComfyUI's validator so arbitrary keys would pass. Chaining needs neither: it is
|
|
what ComfyUI's own LoraLoader does, every strength is a real socket, and the graph
|
|
shows how many LoRAs are actually applied instead of hiding the count in a string.
|
|
"""
|
|
|
|
import comfy.sd
|
|
import comfy.utils
|
|
import folder_paths
|
|
from comfy_api.latest import io
|
|
|
|
from ..._core import log
|
|
|
|
|
|
class FunPackLoraLoader(io.ComfyNode):
|
|
@classmethod
|
|
def define_schema(cls) -> io.Schema:
|
|
return io.Schema(
|
|
node_id="FunPackLoraLoader",
|
|
display_name="FunPack LoRA Loader",
|
|
category="FunPack/Loaders",
|
|
description="Apply one LoRA to a model, and optionally its text encoder. "
|
|
"Chain these for more than one.",
|
|
inputs=[
|
|
io.Model.Input("model"),
|
|
io.Combo.Input("lora_name", options=folder_paths.get_filename_list("loras")),
|
|
io.Float.Input("strength_model", default=1.0, min=-100.0, max=100.0, step=0.01),
|
|
io.Clip.Input("clip", optional=True,
|
|
tooltip="Leave unwired for a model-only LoRA."),
|
|
io.Float.Input("strength_clip", default=1.0, min=-100.0, max=100.0, step=0.01,
|
|
optional=True),
|
|
],
|
|
outputs=[
|
|
io.Model.Output(display_name="model"),
|
|
io.Clip.Output(display_name="clip"),
|
|
io.String.Output(display_name="status"),
|
|
],
|
|
)
|
|
|
|
@classmethod
|
|
def execute(cls, model, lora_name: str, strength_model: float,
|
|
clip=None, strength_clip: float = 1.0) -> io.NodeOutput:
|
|
# Doing nothing is a valid request, but it must return the SAME objects
|
|
# rather than clones: a clone here would silently drop patches a later
|
|
# loader in the chain applied to the original.
|
|
if strength_model == 0 and (clip is None or strength_clip == 0):
|
|
# Loud, because "applied a LoRA at strength 0" is indistinguishable
|
|
# from "forgot to set the strength" by looking at the picture.
|
|
log.alert("FunPack LoRA Loader",
|
|
f"{lora_name} has both strengths at 0 and was not applied")
|
|
return io.NodeOutput(model, clip, f"{lora_name}: both strengths are 0, not applied")
|
|
|
|
path = folder_paths.get_full_path_or_raise("loras", lora_name)
|
|
lora = comfy.utils.load_torch_file(path, safe_load=True)
|
|
patched_model, patched_clip = comfy.sd.load_lora_for_models(
|
|
model, clip, lora, strength_model, strength_clip)
|
|
|
|
applied = [f"model={strength_model}"]
|
|
applied.append(f"clip={strength_clip}" if clip is not None else "clip=not wired")
|
|
log.info("FunPack LoRA Loader", f"{lora_name} applied ({', '.join(applied)})")
|
|
return io.NodeOutput(patched_model, patched_clip,
|
|
f"FunPack LoRA Loader | {lora_name} ({', '.join(applied)})")
|