Skip to content

rewrite fill_default_value logic for robust #1165

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

Open
xin3he wants to merge 1 commit into
base: main
Choose a base branch
from
Open

rewrite fill_default_value logic for robust #1165

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
14 changes: 8 additions & 6 deletions auto_round/compressors/utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -334,13 +334,15 @@ def normalize_item(item: Union[str, dict, "QuantizationScheme"], layer_name: str

# In AutoScheme with mixed gguf:q4_k_m, the super_group_size of gguf:q8_0 layer is None,
# which should not be filled by default q4km again
if fill_default_value:
tmp_scheme_keys = scheme_keys
else:
tmp_scheme_keys = extra_scheme_keys
for cfg in layer_config.values():
for key in tmp_scheme_keys:
cfg.setdefault(key, copy.deepcopy(default_dict.get(key)))
for key in scheme_keys:
if fill_default_value:
cfg.setdefault(key, copy.deepcopy(default_dict.get(key)))
else:
if key in extra_scheme_keys:
cfg.setdefault(key, copy.deepcopy(default_dict.get(key)))
else:
cfg.setdefault(key, None)

# 5. collect supported modules
embedding_types = (torch.nn.Embedding,)
Expand Down