Skip to content

fix(model_loader): add autocast_adapter_dtype in Peft Model #40

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
cmd-HMN merged 2 commits into from
Oct 7, 2026
Merged

fix(model_loader): add autocast_adapter_dtype in Peft Model #40

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
828b780
fix(model_loader): add autocast_adapter_dtype in Peft Model
cmd-HMN Oct 7, 2026
ec2e96b
Merge branch 'main' into minor
cmd-HMN Oct 7, 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
6 changes: 5 additions & 1 deletion bbq/src/utils/model_loader.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -68,7 +68,9 @@ def resolve_torch_data_type(
return torch.float32


def _from_pretrained_fast(loader_cls_or_fn: Any, model_id_or_obj: Any, **kwargs: Any) -> Any:
def _from_pretrained_fast(
loader_cls_or_fn: Any, model_id_or_obj: Any, **kwargs: Any
) -> Any:
"""Load model/adapter/processor from local huggingface cache first to avoid network latency."""
try:
return loader_cls_or_fn.from_pretrained(
Expand Down Expand Up @@ -143,13 +145,15 @@ def load_model_with_lora_adapters(
config.lora_adapter_id,
config=lora_adapter_config,
local_files_only=True,
autocast_adapter_dtype=False,
)
except Exception:
model_to_use = PeftModel.from_pretrained(
base_model,
config.lora_adapter_id,
config=lora_adapter_config,
local_files_only=False,
autocast_adapter_dtype=False,
)
except Exception as exception_instance:
raise LoRAAdapterLoadError(
Expand Down
Loading