fix 'LazyValue' object has no attribute 'keys' in qwen3 moe + deepep + eplb#21820
Closed
Evgueni-Petrov-aka-espetrov wants to merge 14 commits intosgl-project:mainfrom
Closed
fix 'LazyValue' object has no attribute 'keys' in qwen3 moe + deepep + eplb#21820Evgueni-Petrov-aka-espetrov wants to merge 14 commits intosgl-project:mainfrom
Evgueni-Petrov-aka-espetrov wants to merge 14 commits intosgl-project:mainfrom
Conversation
temporary tensors get stuck in vram for too long hit this oom with qwen3 coder 480b instruct fp8 on 2 mi355x
File "/sgl-workspace/sglang/python/sglang/srt/eplb/eplb_manager.py", line 110, in _compute_update_layer_ids_chunks
list(self._model_runner.model.routed_experts_weights_of_layer.keys())
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
AttributeError: 'LazyValue' object has no attribute 'keys'
Contributor
There was a problem hiding this comment.
Code Review
This pull request fixes a crash in the Qwen3 MoE model by accessing the .value property of a LazyValue wrapper during initialization. The review feedback points out that while this change resolves the immediate issue, it renders the LazyValue wrapper redundant because the evaluation is no longer deferred; a refactor is suggested to remove the wrapper entirely for better maintainability.
| ) | ||
| } | ||
| ) | ||
| ).value |
Contributor
There was a problem hiding this comment.
While adding .value fixes the reported crash by ensuring self.routed_experts_weights_of_layer is a dictionary, it makes the LazyValue wrapper and the associated lambda (starting at line 1187) redundant. Since get_moe_weights() is a lightweight operation that merely returns references to existing tensors, the lazy evaluation provides no benefit here. For better maintainability, consider refactoring this block to remove LazyValue and assign the dictionary comprehension directly.
vroomfondel
added a commit
to vroomfondel/dgxarley
that referenced
this pull request
Apr 7, 2026
**Upstream status** as of 2026-04-06: - Qwen3.5: fixed via [PR #19767](sgl-project/sglang#19767) (merged 2026-03-09, included in v0.5.10) - Qwen3: [PR #21461](sgl-project/sglang#21461) — closed without merge 2026-03-30 (CI failure), superseded by #21822 - Qwen3: [PR #21822](sgl-project/sglang#21822) — new fix opened 2026-03-26, addresses `AttributeError: 'LazyValue' object has no attribute 'keys'` in `eplb_manager.py` for Qwen3 MoE. Code review 2026-04-04 by `Fridge003` and `Evgueni-Petrov-aka-espetrov`. Alternative `LazyValue.__getattr__` approach proposed (avoids modifying the model class). **Approved** by `Fridge003` on 2026-04-06, CI rerun triggered — awaiting merge. (Duplicate [PR #21820](sgl-project/sglang#21820) was closed same day in favour of #21822.) Not in v0.5.10 When `--enable-eplb` is active with EP, the `EPLBManager` crashes after its first rebalance interval (default: 1000 forward passes): - SGLang PR #17137 — non-Marlin WNA16MoE port (does not fix EP bug) - SGLang #14158 — update_weights_from_tensor for WNA16MoE (unrelated) - SGLang [PR #13715](sgl-project/sglang#13715) — fix EPLB + FP4 weight tensor filtering (merged, different issue) - SGLang [PR #20963](sgl-project/sglang#20963) — Nvidia modelopt refactoring (1/N). Under active review: reviewer `Edwardf0t1` asked for end-to-end verification 2026-03-31, author `wenscarl` responded 2026-04-01 and posted 3 further inline review responses 2026-04-06. Not stalled but awaiting approval. Migrates the NVFP4 code as-is — expected vehicle for EP-awareness fixes (#20869, #21630). Watch this PR for resolution of the NVFP4 input_scale and CutlassMoEParams bugs - SGLang [PR #21822](sgl-project/sglang#21822) — new EPLB/Qwen3 fix (opened 2026-03-26). Addresses `LazyValue.keys()` AttributeError. Code review 2026-04-04 by `Fridge003` and `Evgueni-Petrov-aka-espetrov`. Alternative `LazyValue.__getattr__` approach proposed. **Approved** by `Fridge003` on 2026-04-06, CI rerun triggered — awaiting merge "Good code is like humor: when you have to explain it, it’s bad." - Cory House P.S.: Code reviews and approvals are crucial for maintaining high-quality software.
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
1 participant
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.
Motivation
found this typo while tuning sglang performance for qwen3 coder 480b in disaggregated prefill-decode setup
Modifications
add missing attribute name
Accuracy Tests
n/a
Speed Tests and Profiling
n/a
Checklist
Review and Merge Process
/tag-and-rerun-ci,/tag-run-ci-label,/rerun-failed-ci