Remove pipeline parallelism (graph_pp) from AutoParallel#439
Merged
Conversation
meta-cla
Bot
added
the
CLA Signed
Graph-based pipeline parallelism is moving to its own package. This
removes all PP modules (api_pp, graph_pp_runner, graph_partition,
split_fsdp_collectives, split_di_dw_graph, graph_multiplex), the two
PP examples (example_ds3_pp, example_pp_graph_passes), the numerics
comparison script (run_ds3_numerics_check), and their test
(test_api_pp).
Files edited:
- __init__.py: removed AutoParallelPP export
- placement_options.py: removed log_pp_model_weights/log_pp_grads
from NumericsLogger (only called from the deleted PP example)
- dsv3.py: removed DeepSeekV3Stage{0,I,N} classes and simplified
_init_weights_* type annotations from Union[Model, Stage] to Model
- test_cuda.yml: removed commented-out PP CI entries
activation_checkpointing.py, module_construction.py, and graph_utils.py
are shared with the non-PP codepath and left untouched.
Validated: pytest tests/ passes (327 tests, 1 xfail).
stack-info: PR: #439, branch: sanketpurandare/stack/1
sanketpurandare
force-pushed
the
sanketpurandare/stack/1
branch
from
84d89de to
bbedceb
Compare
This was referenced May 1, 2026
xmfan
approved these changes
May 1, 2026
aditvenk
approved these changes
May 1, 2026
fmassa
approved these changes
May 1, 2026
This was referenced May 4, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Stacked PRs:
Remove pipeline parallelism (graph_pp) from AutoParallel
Graph-based pipeline parallelism is moving to its own package. This
removes all PP modules (api_pp, graph_pp_runner, graph_partition,
split_fsdp_collectives, split_di_dw_graph, graph_multiplex), the two
PP examples (example_ds3_pp, example_pp_graph_passes), the numerics
comparison script (run_ds3_numerics_check), and their test
(test_api_pp).
Files edited:
from NumericsLogger (only called from the deleted PP example)
init_weights* type annotations from Union[Model, Stage] to Model
activation_checkpointing.py, module_construction.py, and graph_utils.py
are shared with the non-PP codepath and left untouched.
Validated: pytest tests/ passes (327 tests, 1 xfail).