Skip to content

Fix AttributeError on missing use_sft in grain data processing #5137

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

Fix AttributeError on missing use_sft in grain data processing #5137

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
14 changes: 7 additions & 7 deletions src/maxtext/input_pipeline/grain_data_processing.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -732,11 +732,11 @@ def vision_sft_preprocessing_pipeline(

def _get_pipeline_fn(config):
"""Returns the appropriate preprocessing pipeline function based on config."""
if config.use_sft and config.use_multimodal:
if getattr(config, "use_sft", False) and getattr(config, "use_multimodal", False):
return vision_sft_preprocessing_pipeline
if config.use_dpo:
if getattr(config, "use_dpo", False):
return dpo_preprocessing_pipeline
if config.use_sft:
if getattr(config, "use_sft", False):
return sft_preprocessing_pipeline
return pretrain_preprocessing_pipeline

Expand Down Expand Up @@ -831,8 +831,8 @@ def make_grain_train_iterator(
"grain_worker_count": config.grain_worker_count,
"grain_per_worker_buffer_size": config.grain_per_worker_buffer_size,
}
if config.use_sft and config.use_multimodal:
preprocessing_fn_kwargs["image_column"] = config.train_image_column
if getattr(config, "use_sft", False) and getattr(config, "use_multimodal", False):
preprocessing_fn_kwargs["image_column"] = getattr(config, "train_image_column", None)

preprocessing_fn = functools.partial(
pipeline_fn,
Expand Down Expand Up @@ -977,8 +977,8 @@ def make_grain_eval_iterator(
"grain_worker_count": config.grain_worker_count_eval,
"grain_per_worker_buffer_size": config.grain_per_worker_buffer_size_eval,
}
if config.use_sft and config.use_multimodal:
eval_preprocessing_fn_kwargs["image_column"] = config.eval_image_column
if getattr(config, "use_sft", False) and getattr(config, "use_multimodal", False):
eval_preprocessing_fn_kwargs["image_column"] = getattr(config, "eval_image_column", None)

preprocessing_fn = functools.partial(
pipeline_fn,
Expand Down
2 changes: 2 additions & 0 deletions tests/unit/mmap_data_processing_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1906,6 +1906,8 @@ def _config(mmap_npy_split):
tokenize_eval_data=False,
colocated_python_data_input=False,
generate_padding_batch_eval=False,
use_sft=False,
use_multimodal=False,
)

@staticmethod
Expand Down
Loading