Skip to content

[v1.0] fix argparse.ArgumentError #3616

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

Merged
zjjlivein merged 3 commits into from
Jan 19, 2026
Merged

[v1.0] fix argparse.ArgumentError #3616

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
120f41c
fix
miao200years Jan 19, 2026
cfa2807
fix
miao200years Jan 19, 2026
32d3ba9
fix
miao200years Jan 19, 2026
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
7 changes: 0 additions & 7 deletions examples/experiments/deepseek_v3_pretrain/run_pretrain.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -239,10 +239,6 @@ class ModelArguments:
"help": "Pre-training from existing paddleformers model weights. Default False and model will train from scratch. If set True, the model_name_or_path argument must exist in the paddleformers models."
},
)
num_hidden_layers: Optional[int] = field(
default=None,
metadata={"help": "num_hidden_layers."},
)


def create_pretrained_dataset(
Expand Down Expand Up @@ -482,9 +478,6 @@ def main():
config.vocab_size = max(config.vocab_size, ((tokenizer.vocab_size - 1) // 128 + 1) * 128)
logger.info(f"Reset vocab size to {config.vocab_size} for batter amp peformance.")

config.num_hidden_layers = (
model_args.num_hidden_layers if model_args.num_hidden_layers is not None else config.num_hidden_layers
)
# Config for model using dropout, such as GPT.
if hasattr(config, "use_dualpipev"):
# NOTE(zhangyuqin): In Paddle, the segmentation and scheduling of pipeline parallel
Expand Down
7 changes: 0 additions & 7 deletions paddleformers/cli/train/pretrain/workflow.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -176,10 +176,6 @@ class ModelArguments:
"help": "Pre-training from existing paddleformers model weights. Default False and model will train from scratch. If set True, the model_name_or_path argument must exist in the paddleformers models."
},
)
num_hidden_layers: Optional[int] = field(
default=None,
metadata={"help": "num_hidden_layers."},
)


def create_pretrained_dataset(
Expand Down Expand Up @@ -418,9 +414,6 @@ def run_dsv3_pretrain(model_args, data_args, generating_args, training_args):
config.vocab_size = max(config.vocab_size, ((tokenizer.vocab_size - 1) // 128 + 1) * 128)
logger.info(f"Reset vocab size to {config.vocab_size} for batter amp peformance.")

config.num_hidden_layers = (
model_args.num_hidden_layers if model_args.num_hidden_layers is not None else config.num_hidden_layers
)
# Config for model using dropout, such as GPT.
if hasattr(config, "use_dualpipev"):
# NOTE(zhangyuqin): In Paddle, the segmentation and scheduling of pipeline parallel
Expand Down
Loading