Skip to content

Fix error for upgrading JAX to 0.7.2 #2368

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
merged 1 commit into from
Sep 20, 2025
Merged

Fix error for upgrading JAX to 0.7.2 #2368

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
2 changes: 1 addition & 1 deletion .github/workflows/run_tests_internal.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -67,4 +67,4 @@ jobs:
FINAL_PYTEST_MARKER="${{ inputs.pytest_marker }} and not scheduled_only"
fi
python3 -m pip install -e . --no-dependencies &&
python3 -m pytest -v -m "${FINAL_PYTEST_MARKER}" --durations=0
LIBTPU_INIT_ARGS='--xla_tpu_scoped_vmem_limit_kib=65536' python3 -m pytest -v -m "${FINAL_PYTEST_MARKER}" --durations=0
10 changes: 5 additions & 5 deletions tests/train_compile_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -493,7 +493,7 @@ def test_moe_deepseek_scanned_bf16(self):
"megablox=False",
"per_device_batch_size=2",
"max_target_length=1024",
"attention=dot_product", # Change to flash attention once it works for MLA
"attention=flash",
"dtype=bfloat16",
"weight_dtype=bfloat16",
"scan_layers=True",
Expand All @@ -518,7 +518,7 @@ def test_moe_deepseek_unscanned_bf16(self):
"megablox=False",
"per_device_batch_size=1",
"max_target_length=1024",
"attention=dot_product", # Change to flash attention once it works for MLA
"attention=flash",
"dtype=bfloat16",
"weight_dtype=bfloat16",
"scan_layers=False",
Expand All @@ -541,7 +541,7 @@ def test_moe_deepseek_with_device_limit(self):
"megablox=False",
"per_device_batch_size=1",
"max_target_length=1024",
"attention=dot_product", # Change to flash attention once it works for MLA
"attention=flash",
"dtype=bfloat16",
"weight_dtype=bfloat16",
"n_routing_groups=8",
Expand All @@ -565,7 +565,7 @@ def test_moe_deepseek_without_device_limit(self):
"megablox=False",
"per_device_batch_size=1",
"max_target_length=1024",
"attention=dot_product", # Change to flash attention once it works for MLA
"attention=flash",
"dtype=bfloat16",
"weight_dtype=bfloat16",
"n_routing_groups=-1",
Expand All @@ -585,7 +585,7 @@ def test_moe_deepseek_pipeline_subset(self):
"compile_topology_num_slices=8",
"use_iota_embed=true",
"model_name=deepseek3-671b",
"megablox=False", # dropless not yet supported (b/418313093)
"megablox=True",
"sparse_matmul=False",
"capacity_factor=1",
"per_device_batch_size=1",
Expand Down
Loading