Skip to content

Add NVTE_BACKWARD_MODE=default|unquant|dequant #2644

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
zianglih wants to merge 47 commits into
base: main
Choose a base branch
from
Open

Add NVTE_BACKWARD_MODE=default|unquant|dequant #2644

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
47 commits
Select commit Hold shift + click to select a range
7ce112c
Add NVTE_KEEP_BACKWARD_UNQUANTIZED
zianglih Feb 3, 2026
2a274a1
[pre-commit.ci] auto fixes from pre-commit.com hooks
pre-commit-ci[bot] Feb 3, 2026
fc38587
Disable ub and clean up
zianglih Feb 3, 2026
28adf90
Drop fuser changes
zianglih Feb 3, 2026
3de128f
Replace use_quantized_bwd with use_fp8_bwd
zianglih Feb 3, 2026
b35f4b2
[pre-commit.ci] auto fixes from pre-commit.com hooks
pre-commit-ci[bot] Feb 3, 2026
d069a7c
Ignore keep_backward_unquantized if delayed scaling
zianglih Feb 3, 2026
7b3d3c0
Refactor ignoring NVTE_KEEP_BACKWARD_UNQUANTIZED when delayed scaling…
zianglih Feb 3, 2026
f75de0c
Add back missing ctx.debug
zianglih Feb 3, 2026
460f3a9
Refactor changes under fused
zianglih Feb 3, 2026
4a9ae12
Clean up
zianglih Feb 3, 2026
d589225
Refactor high-precision overwrite if keep_backward_unquantized
zianglih Feb 3, 2026
33af723
Clean up
zianglih Feb 3, 2026
e187506
Drop redundant fp8_recipe_bwd
zianglih Feb 4, 2026
5612c6e
[pre-commit.ci] auto fixes from pre-commit.com hooks
pre-commit-ci[bot] Feb 4, 2026
d56d8dc
Drop redundant ub changes
zianglih Feb 4, 2026
880e6ba
Drop more redundant ub changes
zianglih Feb 4, 2026
4ae52cc
Drop redundant delayed scaling changes
zianglih Feb 4, 2026
f1c85cd
[pre-commit.ci] auto fixes from pre-commit.com hooks
pre-commit-ci[bot] Feb 4, 2026
2154afc
Drop unneeded backwards_needs_fc1_input
zianglih Feb 4, 2026
4627465
Drop and disallow LayerNormMLP implementation
zianglih Feb 4, 2026
067bd0e
[pre-commit.ci] auto fixes from pre-commit.com hooks
pre-commit-ci[bot] Feb 4, 2026
3ad5e02
Move interface changes to recipe
zianglih Feb 5, 2026
8b037a6
[pre-commit.ci] auto fixes from pre-commit.com hooks
pre-commit-ci[bot] Feb 5, 2026
6d51788
Move ub overrides to fwd
zianglih Feb 5, 2026
7f9d7fe
Remove duplication
zianglih Feb 5, 2026
b750890
Simplify use_fp8_bwd logic in bwd
zianglih Feb 5, 2026
ec8cb53
[pre-commit.ci] auto fixes from pre-commit.com hooks
pre-commit-ci[bot] Feb 5, 2026
1d0ba9d
Set grad quantizers to none if keep bwd unquantized
zianglih Feb 5, 2026
a344da6
[pre-commit.ci] auto fixes from pre-commit.com hooks
pre-commit-ci[bot] Feb 5, 2026
90e9706
Drop delayed scaling change
zianglih Feb 6, 2026
645b80f
Simplify env var logic
zianglih Feb 9, 2026
3b36235
Move validation check to recipe
zianglih Feb 9, 2026
44dfb91
Simplify effective_enabled
zianglih Feb 9, 2026
675037e
Fix inverted assertion logic
zianglih Feb 9, 2026
530b421
Simplify changes under ops
zianglih Feb 9, 2026
33d52ac
[pre-commit.ci] auto fixes from pre-commit.com hooks
pre-commit-ci[bot] Feb 9, 2026
1167b2e
Simplify ctx.keep_backward_unquantized
zianglih Feb 9, 2026
ac6f77e
Fix missing attribute
zianglih Feb 9, 2026
8e5223b
Add unit tests
zianglih Feb 10, 2026
b2bf241
Fix bias errors in unit test
zianglih Feb 10, 2026
3b37175
[pre-commit.ci] auto fixes from pre-commit.com hooks
pre-commit-ci[bot] Feb 10, 2026
188ba19
Add more shapes to unit test
zianglih Feb 10, 2026
885f127
Refator interface to `NVTE_BACKWARD_MODE=default|unquant|dequant`
zianglih Feb 24, 2026
6006867
Fix override and clean up
zianglih Feb 25, 2026
d5f0195
Clean up unit test
zianglih Mar 5, 2026
1edb2c3
Clean up unit test
zianglih Mar 5, 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
1 change: 1 addition & 0 deletions qa/L0_pytorch_unittest/test.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -42,6 +42,7 @@ python3 -m pytest --tb=auto --junitxml=$XML_LOG_DIR/pytest_test_gqa.xml $TE_PATH
python3 -m pytest --tb=auto --junitxml=$XML_LOG_DIR/pytest_test_fused_optimizer.xml $TE_PATH/tests/pytorch/test_fused_optimizer.py || test_fail "test_fused_optimizer.py"
python3 -m pytest --tb=auto --junitxml=$XML_LOG_DIR/pytest_test_multi_tensor.xml $TE_PATH/tests/pytorch/test_multi_tensor.py || test_fail "test_multi_tensor.py"
python3 -m pytest --tb=auto --junitxml=$XML_LOG_DIR/pytest_test_fusible_ops.xml $TE_PATH/tests/pytorch/test_fusible_ops.py || test_fail "test_fusible_ops.py"
python3 -m pytest --tb=auto --junitxml=$XML_LOG_DIR/pytest_test_backward_mode.xml $TE_PATH/tests/pytorch/test_backward_mode.py || test_fail "test_backward_mode.py"
python3 -m pytest --tb=auto --junitxml=$XML_LOG_DIR/pytest_test_permutation.xml $TE_PATH/tests/pytorch/test_permutation.py || test_fail "test_permutation.py"
python3 -m pytest --tb=auto --junitxml=$XML_LOG_DIR/pytest_test_parallel_cross_entropy.xml $TE_PATH/tests/pytorch/test_parallel_cross_entropy.py || test_fail "test_parallel_cross_entropy.py"
python3 -m pytest --tb=auto --junitxml=$XML_LOG_DIR/pytest_test_cpu_offloading.xml $TE_PATH/tests/pytorch/test_cpu_offloading.py || test_fail "test_cpu_offloading.py"
Expand Down
Loading