[Dev] Fix golden values mismatch and dependency error due to last pull main - #4123
Merged
Conversation
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Victarry
changed the title
Author
|
/ok to test c49a7ba |
Victarry
changed the title
ilml
approved these changes
Apr 3, 2026
yaox12
approved these changes
Apr 3, 2026
yaox12
enabled auto-merge
April 3, 2026 04:42
7 tasks
…erge The Sync main into dev (NVIDIA#4031) merge incorrectly resolved a conflict in pyproject.toml, keeping dev's flask[async]+openai instead of main's quart+openai[aiohttp]. The code (text_generation_server.py and rl/inference/megatron.py) correctly uses Quart and DefaultAioHttpClient, but the missing deps caused gpt_grpo_basic_function to fail in CI with: - WARNING: Could not import quart: No module named 'quart' - RuntimeError: To use the aiohttp client you must have installed the package with the `aiohttp` extra Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Author
|
/ok to test f7df0d7 |
Victarry
changed the title
Add quart and httpx-aiohttp package entries, add openai[aiohttp] optional dependency section, and update megatron-core dev extra to reference quart and openai[aiohttp] instead of flask[async] and openai. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Author
|
/ok to test dc1ca93 |
…erge Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Author
|
/ok to test 99e3c3c |
…orm build uv sync --locked fails when fast-hadamard-transform (a git source package in no-build-isolation-package) tries to build without setuptools available in the base environment. Pre-install wheel + setuptools before syncing, matching the fix from c2cbe54 that was never ported to this branch. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…d-transform build" This reverts commit 19be8f0.
Author
|
/ok to test c381ad8 |
…dagraph Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Author
|
/ok to test 10498f8 |
ko3n1g
approved these changes
Apr 3, 2026
Contributor
|
🔄 Merge queue validation started! You can track the progress here: https://github.com/NVIDIA/Megatron-LM/actions/runs/23945288668 |
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
5 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.
Summary
golden_values_dev_dgx_h100.jsonfor 4 tests that failed withnum-zerosgolden value mismatch in the merge queue of [Dev] Skip routed expert padding for graph-safe MoE #4071gpt/gpt3_mcore_te_tp2_pp2_mhcmoe/gpt3_mcore_te_tp2_pp1_te_a2a_ovlp_8experts_etp1_ep4moe/gpt3_mcore_te_tp2_pp1_te_8experts2parallel_overlap_grad_reduce_param_gather_groupedGEMMmoe/gpt3_mcore_te_tp2_pp1_te_8experts2parallel_ddp_average_in_collective🤖 Generated with Claude Code