Skip to content

fix(args): stop a custom TrainPipelineConfig from clearing the model family #150

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
Rockdu wants to merge 1 commit into
base: main
Choose a base branch
from
Draft

fix(args): stop a custom TrainPipelineConfig from clearing the model family #150

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 4 additions & 1 deletion miles/rollout/encoder_hub/__init__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,4 +14,7 @@ def get_encoder(family: str | None):
from miles.rollout.encoder_hub import wan2_2

return wan2_2
raise ValueError(f"no encoder_hub entry for model family {family!r}")
raise ValueError(
f"no encoder_hub entry for model family {family!r}; set --diffusion-model-family to one "
"that has an entry, or add a module here for it"
)
15 changes: 7 additions & 8 deletions miles/utils/arguments.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1472,12 +1472,15 @@ def miles_validate_args(args):

from miles.utils.misc import load_function

if args.diffusion_model_family is not None:
# Downstream lookups compare this exactly (encoder_hub.get_encoder), so normalize once
# here rather than at every reader.
args.diffusion_model_family = args.diffusion_model_family.strip().lower()

if args.train_pipeline_config_path is not None:
if args.diffusion_model_family is not None:
raise ValueError("--train-pipeline-config-path and --diffusion-model-family both name a config; pass one.")
# Explicit config path IS the identity (custom classes never need registering).
# The path names the config class; the family stays whatever was passed, since it is a
# separate axis (encoder_hub) and may be unset for a model that customizes that too.
cfg_cls = load_function(args.train_pipeline_config_path)
args.diffusion_model_family = None
else:
from miles.backends.fsdp_utils.configs.train_pipeline_config import (
get_train_pipeline_config_cls,
Expand All @@ -1486,10 +1489,6 @@ def miles_validate_args(args):

if args.diffusion_model_family is None:
args.diffusion_model_family = resolve_diffusion_model_family(args.hf_checkpoint)
else:
# Downstream lookups compare this exactly (encoder_hub.get_encoder), so normalize
# here rather than at every reader.
args.diffusion_model_family = args.diffusion_model_family.strip().lower()
cfg_cls = get_train_pipeline_config_cls(args.diffusion_model_family)
args.train_pipeline_config_path = f"{cfg_cls.__module__}.{cfg_cls.__qualname__}"
if args.model_backend_path is None:
Expand Down
Loading