🐛 CI failure: test_paged_stashing.py::TestPagedStashingOverBudget::test_overload_factor_and_over_budget
- Dominant language
- Python
- Stars
- 17.9k
- Forks
- 4.5k
- Avg merge
- 4d 6h
- Merged PRs (30d)
- 271
Description
**Describe the bug**
CI test \`tests/unit_tests/transformer/moe/test_paged_stashing.py::TestPagedStashingOverBudget::test_overload_factor_and_over_budget\` failed in job [tests/unit_tests/transformer/moe/**/*.py - latest](https://github.com/NVIDIA/Megatron-LM/actions/runs/24514641750/job/71660734791).
Tag the [@mcore-oncall](https://github.com/orgs/NVIDIA/teams/mcore-oncall) to get oncall's attention to this issue.
Same root cause as #4339 — \`TEGroupedMLP\` asserts \`_is_fused_impl_supported()\` when \`use_transformer_engine_op_fuser=True\`, but the combination with \`moe_paged_stash=True\` makes that check return \`False\`.
**Failing run**
| Field | Value |
|-------|-------|
| PR | [#4332: build: bump TransformerEngine to release_v2.14 (dev)](https://github.com/NVIDIA/Megatron-LM/pull/4332) |
| Run | [24514641750](https://github.com/NVIDIA/Megatron-LM/actions/runs/24514641750) |
| Job | [71660734791](https://github.com/NVIDIA/Megatron-LM/actions/runs/24514641750/job/71660734791) |
**Error**
```
AssertionError: Fused GroupedMLP is not supported for this configuration.
megatron/core/transformer/moe/experts.py:730: AssertionError
FAILED tests/unit_tests/transformer/moe/test_paged_stashing.py::TestPagedStashingOverBudget::test_overload_factor_and_over_budget
```
**Steps/Code to reproduce bug**
```bash
pytest tests/unit_tests/transformer/moe/test_paged_stashing.py::TestPagedStashingOverBudget::test_overload_factor_and_over_budget
```
**Additional context**
Triaged automatically via \`/triage-issue\`.
Contributor guide
Assessment
This issue has not been assessed yet.