From 95865de24f2e6a18e9df1ea0a567933c3bb57379 Mon Sep 17 00:00:00 2001
From: Yihao Wang <42559837+AgainstEntropy@users.noreply.github.com>
Date: Fri, 24 Jul 2026 18:55:57 -0700
Subject: [PATCH] [diffusion] CI: read consistency GT from ci-data-diffusion at
 per-platform commit (#32297)

---
 python/sglang/multimodal_gen/test/test_utils.py | 17 ++++++++---------
 .../test/unit/test_consistency_metrics.py       | 16 +++++++++++-----
 2 files changed, 19 insertions(+), 14 deletions(-)

diff --git a/python/sglang/multimodal_gen/test/test_utils.py b/python/sglang/multimodal_gen/test/test_utils.py
index 2c9d3093e..09300734e 100644
--- a/python/sglang/multimodal_gen/test/test_utils.py
+++ b/python/sglang/multimodal_gen/test/test_utils.py
@@ -35,17 +35,16 @@ if TYPE_CHECKING:
 logger = init_logger(__name__)
 
 # GT is read from <repo>@<revision>. A given SHA only exists in the repo it was
-# committed to, so REPO and REVISION must be bumped together. New GT now
-# publishes to sgl-project/ci-data-diffusion; the pins below still resolve
-# against sgl-project/ci-data (readable but frozen). When you bump REVISION to a
-# freshly generated commit, switch REPO to sgl-project/ci-data-diffusion too
-# (and update the URL assertion in the consistency-metrics unit test).
-SGL_TEST_FILES_CI_DATA_REPO = "sgl-project/ci-data"
-
-SGL_TEST_FILES_CI_DATA_REVISION = "320949ecc2587474a2f535229ffc8f47ed16ee51"
+# committed to, so REPO and REVISION must be bumped together. All GT (CUDA and
+# NPU/ascend) is read from sgl-project/ci-data-diffusion, where the GT-gen workflows
+# publish.
+SGL_TEST_FILES_CI_DATA_REPO = "sgl-project/ci-data-diffusion"
+SGL_TEST_FILES_CI_DATA_REVISION = "6d689f4833e8c106ff0d853865f50496d1f2b453"
 
+# The NPU pin is kept as a separate branch so ascend GT can be bumped independently
+# when it's regenerated on its own cadence.
 if current_platform.is_npu():
-    SGL_TEST_FILES_CI_DATA_REVISION = "6b62f4b6825c76a25fd2ba28248df68f2b400e65"
+    SGL_TEST_FILES_CI_DATA_REVISION = "6d689f4833e8c106ff0d853865f50496d1f2b453"
 
 SGL_TEST_FILES_CONSISTENCY_GT_ROOT = (
     "https://raw.githubusercontent.com/"
diff --git a/python/sglang/multimodal_gen/test/unit/test_consistency_metrics.py b/python/sglang/multimodal_gen/test/unit/test_consistency_metrics.py
index f35f7b70e..bf8f7c354 100644
--- a/python/sglang/multimodal_gen/test/unit/test_consistency_metrics.py
+++ b/python/sglang/multimodal_gen/test/unit/test_consistency_metrics.py
@@ -33,11 +33,17 @@ def _disable_remote_official_gt_case_map(monkeypatch):
 
 
 def test_consistency_gt_urls_are_pinned_to_ci_data_revision():
-    revision_path = f"/ci-data/{test_utils.SGL_TEST_FILES_CI_DATA_REVISION}/"
-
-    assert "/ci-data/main/" not in test_utils.SGL_TEST_FILES_CONSISTENCY_GT_ROOT
-    assert revision_path in test_utils.SGL_TEST_FILES_OFFICIAL_CONSISTENCY_GT_BASE
-    assert revision_path in test_utils.SGL_TEST_FILES_SGLANG_CONSISTENCY_GT_BASE
+    # GT must be pinned to an immutable commit (not a moving branch) so results are
+    # reproducible and the per-URL download cache invalidates on regeneration.
+    assert test_utils.SGL_TEST_FILES_CI_DATA_REVISION != "main"
+    pinned_revision_path = (
+        f"/{test_utils.SGL_TEST_FILES_CI_DATA_REPO}/"
+        f"{test_utils.SGL_TEST_FILES_CI_DATA_REVISION}/"
+    )
+    assert (
+        pinned_revision_path in test_utils.SGL_TEST_FILES_OFFICIAL_CONSISTENCY_GT_BASE
+    )
+    assert pinned_revision_path in test_utils.SGL_TEST_FILES_SGLANG_CONSISTENCY_GT_BASE
 
 
 def test_remote_file_exists_returns_false_for_definitive_404(monkeypatch):