Fix Apr 05 integration test regressions (cuda sm_86) - #48170
Open
ydshieh wants to merge 8 commits into
Open
Conversation
…sts (cuda sm_86) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…uda sm_86) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…_decoding_non_distil (cuda sm_86) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…ce_bf16 (cuda sm_86) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…cuda sm_86) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…ation_generate_audio_context (cuda sm_86) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
… integration tests (cuda sm_86) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Contributor
|
[For maintainers] Suggested jobs to run (before merge) run-slow: dia, eomt_dinov3, glm, musicgen_melody, rag, stablelm, whisper |
Contributor
CI recapDashboard: View test results in Grafana |
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
…pr 04 CI report These models had success=0 or were missing from the April 04 CI report, making the April 05 regression classification unreliable. Whisper was also already fixed on main. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
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
2 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
Fixes integration test regressions that first appeared in the Apr 05, 2026 CI run and are still failing on Aug 20, 2026. All failures are on the cuda sm_86 (A10G) runner.
Fixed tests (3 tests across 2 models)
test_model_stablelm_3b_4e1t_logits+test_model_tiny_random_stablelm_2_logits— stale mean/slice logit valuestest_model_9b_eager— stale("cuda", 8)text ("DIY paper lantern" → "DIY paper flower")Not fixed in this PR
The items below had
success=0(or were missing entirely) in the April 04 CI report, meaning the CI job was not running correctly around that date. Their April 05 appearance is unreliable and may not represent a true regression.test_inference_bf16— same issue as the April 04 report (missing from report on April 04; failure existed since April 01).test_speculative_decoding_non_distil— same issue as the April 04 report (success=0). Also already fixed on main.higgs_audio_v2_tokenizer::test_cpu_offload— same issue as the April 04 report.llava_next::test_small_model_integration_test— same issue as the April 04 report.rag::test_rag_token_from_pretrained— same issue as the April 04 report.test_rag_token_generate_beam+test_rag_sequence_generate_beam— same issue as the April 04 report.Already fixed on main
qwen2_5_omni(2 tests) — fixed by PR [Qwen2.5-Omni] Update stale expected values for cuda sm_86 #48164🤖 Generated with Claude Code