Skip to content

Make lora_alpha a float #10004

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
tastelikefeet merged 1 commit into from
Sep 3, 2026
Merged

Make lora_alpha a float #10004

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
4 changes: 2 additions & 2 deletions swift/arguments/tuner_args.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -52,7 +52,7 @@ class TunerArguments:
modules_to_save (List[str]): List of modules to save. Default is an empty list.

lora_rank (int): Rank for LoRA. Default is 8.
lora_alpha (int): Alpha value for LoRA. Default is 32.
lora_alpha (float): Alpha value for LoRA. Default is 32.0.
lora_dropout (float): Dropout rate for LoRA. Default is 0.05.
lora_bias (Literal['none', 'all']): The possible values are 'none' and 'all'. If set to 'all', all biases
will be trainable. Default is 'none'.
Expand Down Expand Up @@ -131,7 +131,7 @@ class TunerArguments:

# lora
lora_rank: int = 8
lora_alpha: int = 32
lora_alpha: float = 32.0
lora_dropout: float = 0.05
lora_bias: Literal['none', 'all'] = 'none'
lora_dtype: Literal['float16', 'bfloat16', 'float32', None] = None
Expand Down
2 changes: 1 addition & 1 deletion swift/megatron/arguments/megatron_args.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -463,7 +463,7 @@ class MegatronTunerMixin:

# lora
lora_rank: int = 8
lora_alpha: int = 32
lora_alpha: float = 32.0
lora_dropout: float = 0.05
lora_bias: Literal['none', 'all'] = 'none'
lora_dtype: Literal['float16', 'bfloat16', 'float32', None] = None
Expand Down
4 changes: 2 additions & 2 deletions swift/tuners/lora_layers.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -517,7 +517,7 @@ def __init__(
adapter_name: str,
module_key: str,
r: int,
lora_alpha: int,
lora_alpha: float,
lora_dropout: float,
merge_weights: bool,
):
Expand All @@ -544,7 +544,7 @@ def __init__(self,
module_key: str,
base_layer: nn.Linear,
r: int = 0,
lora_alpha: int = 1,
lora_alpha: float = 1.0,
lora_dropout: float = 0.,
enable_lora: List[bool] = [False],
fan_in_fan_out: bool = False,
Expand Down
2 changes: 1 addition & 1 deletion tests/megatron/test_opsd.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,7 +19,7 @@
vllm_max_model_len=10240,
tuner_type='lora',
lora_rank=64,
lora_alpha=128,
lora_alpha=128.0,
sleep_level=1,
lmbda=1.0,
beta=0.5,
Expand Down
2 changes: 1 addition & 1 deletion tests/train/test_opsd.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ def train():
teacher_model='Qwen/Qwen3.5-4B',
tuner_type='lora',
lora_rank=64,
lora_alpha=128,
lora_alpha=128.0,
target_modules=['all-linear'],
use_vllm=True,
vllm_mode='colocate',
Expand Down
Loading